INTRUMENTAL

ANALYSIS

David Harvey
DePauw University
Instrumental Analysis

David Harvey
DePauw University
This text is disseminated via the Open Education Resource (OER) LibreTexts Project (https://LibreTexts.org) and like the hundreds
of other texts available within this powerful platform, it is freely available for reading, printing and "consuming." Most, but not all,
pages in the library have licenses that may allow individuals to make changes, save, and print this book. Carefully
consult the applicable license(s) before pursuing such effects.
Instructors can adopt existing LibreTexts texts or Remix them to quickly build course-specific resources to meet the needs of their
students. Unlike traditional textbooks, LibreTexts’ web based origins allow powerful integration of advanced features and new
technologies to support learning.

The LibreTexts mission is to unite students, faculty and scholars in a cooperative effort to develop an easy-to-use online platform
for the construction, customization, and dissemination of OER content to reduce the burdens of unreasonable textbook costs to our
students and society. The LibreTexts project is a multi-institutional collaborative venture to develop the next generation of open-
access texts to improve postsecondary education at all levels of higher learning by developing an Open Access Resource
environment. The project currently consists of 14 independently operating and interconnected libraries that are constantly being
optimized by students, faculty, and outside experts to supplant conventional paper-based books. These free textbook alternatives are
organized within a central environment that is both vertically (from advance to basic level) and horizontally (across different fields)
integrated.
The LibreTexts libraries are Powered by MindTouch® and are supported by the Department of Education Open Textbook Pilot
Project, the UC Davis Office of the Provost, the UC Davis Library, the California State University Affordable Learning Solutions
Program, and Merlot. This material is based upon work supported by the National Science Foundation under Grant No. 1246120,
1525057, and 1413739. Unless otherwise noted, LibreTexts content is licensed by CC BY-NC-SA 3.0.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not
necessarily reflect the views of the National Science Foundation nor the US Department of Education.
Have questions or comments? For information about adoptions or adaptions contact [email protected]. More information on our
activities can be found via Facebook (https://facebook.com/Libretexts), Twitter (https://twitter.com/libretexts), or our blog
(http://Blog.Libretexts.org).

This text was compiled on 12/05/2022

TABLE OF CONTENTS
Licensing

1: Introduction
1.1: Classi cation of Analytical Methods
1.2: Types of Instrumental Methods
1.3: Instruments For Analysis
1.4: Selecting an Analytical Method
1.5: Calibration of Instrumental Methods

2: Electrical Components and Circuits

2.1: Basic Terminology and Laws of Electricity
2.2: Direct Current (DC) Circuits
2.3: Alternating Current Circuits
2.4: Semiconductors

3: Operational Ampli ers

3.1: Operational Ampli ers
3.2: Operational Ampli er Circuits
3.3: Ampli cation and Measurement of Signals
3.4: Mathematical Operations Using Operational Ampli ers

4: Analog and Digital Data

4.1: Analog and Digital Data
4.2: Working With Binary Numbers
4.3: Cleaning Up Signals and Counting Events

5: Signals and Noise

5.1: The Signal-to-Noise Ratio
5.2: Sources of Instrumental Noise
5.3: Signal-to-Noise Enhancement

6: An Introduction to Spectrophotometric Methods

6.1: General Properties of Electromagnetic Radiation
6.2: Wave Properties of Electromagnetic Radiation
6.3: Quantum Mechanical Properties of Electromagnetic Radiation
6.4: Emission and Absorbance Spectra
6.5: Quantitative Considerations

7: Components of Optical Instruments

7.1: General Design of Optical Instruments
7.2: Sources of Radiation
7.3: Wavelength Selectors
7.4: Sample Containers
7.5: Radiation Transducers

1 https://chem.libretexts.org/@go/page/332634
 7.6: Fiber Optics
7.7: Fourier Transform Optical Spectroscopy

8: An Introduction to Optical Atomic Spectroscopy

8.1: Optical Atomic Spectra
8.2: Atomization Methods
8.3: Sample Introduction Methods

9: Atomic Absorption Spectrometry

9.1: Sample Atomization Techniques
9.2: Atomic Absorption Instrumentation
9.3: Interferences in Absorption Spectroscopy
9.4: Atomic Absorption Techniques

10: Atomic Emission Spectrometry

10.1: Emission Spectroscopy Based on Flame and Plasma Sources
10.2: Emission Spectroscopy Based on Arc and Spark Sources

11: Atomic Mass Spectrometry

11.1: General Features of Atomic Mass Spectrometry
11.2: Mass Spectrometers
11.3: Inductively Coupled Plasma Mass Spectrometer
11.4: Other Forms of Atomic Mass Spectrometry

12: Atomic X-Ray Spectrometry

12.1: Fundamental Principles
12.2: Instrument Components
12.3: Atomic X-Ray Fluorescence Methods
12.4: Other X-Ray Methods

13: Introduction to Ultraviolet/Visible Absorption Spectrometry

13.1: Transmittance and Absorbance
13.2: Beer's Law
13.3: Effect of Noise on Transmittance and Absorbance Measurements
13.4: Instrumentation

14: Applications of Ultraviolet/Visible Molecular Absorption Spectrometry

14.2: Absorbing Species
14.3: Qualitative and Characterization Applications
14.4: Quantitative Applications
14.5: Photometric Titrations

15: Molecular Luminescence

15.1: Theory of Fluorescence and Phosphorescence
15.2: Instruments for Measuring Fluorescence and Phosphorescence
15.3: Applications and Photoluminescence methods
15.4: Chemiluminscence
15.5: Evaluation of Molecular Luminescence

2 https://chem.libretexts.org/@go/page/332634
16: An Introduction to Infrared Spectrometry
16.1: Theory of Infrared Absorption Spectrometry
16.2: Infrared Sources and Transducers
16.3: Infrared Instruments

17: Applications of Infrared Spectrometry

17.1: Mid-Infrared Absorption Spectometry
17.2: Mid-Infrared Re ection Spectrometry
17.3: Near-Infrared and Far-Infrared Spectroscopy

18: Raman Spectroscopy

18.1: Theory of Raman Spectroscopy
18.2: Instrumentation
18.3: Applications of Raman Spectroscopy
18.4: Other Types of Raman Spectroscopy

19: Nuclear Magnetic Resonance Spectroscopy

19.1: Theory of Nuclear Magnetic Resonance
19.2: Environmental Effects on NMR Spectra
19.3: NMR Spectrometers
19.4: Applications of Proton NMR
19.5: Carbon-13 NMR
19.6: Two-Dimensional Fourier Transform NMR

20: Molecular Mass Spectrometry

20.1: Molecular Mass Spectra
20.2: Ion Sources
20.3: Mass Spectrometers
20.4: Applications of Molecular Mass Spectrometry

21: Surface Characterization by Spectroscopy and Microscopy

21.1: Introduction to the Study of Surfaces
21.2: Spectroscopic Surface Methods
21.3: Scanning Electron Microscopy
21.4: Scanning Probe Microscopes

22: An Introduction to Electroanalytical Chemistry

22.1: Electrochemical Cells
22.2: Potentials in Electroanalytical Cells
22.3: Electrode Potentials
22.4: Calculation of Cell Potentials from Electrode Potentials
22.5: Currents in Electrochemical Cells
22.6: Types of Electroanalytical Methods

23: Potentiometry
23.1: Reference Electrodes
23.2: Metallic Indicator Electrodes
23.3: Membrane Ion-Selective Electrodes

3 https://chem.libretexts.org/@go/page/332634
 23.4: Molecular-Selective Electrode Systems
23.5: Instruments for Measuring Cell Potentials
23.6: Quantitative Potentiometry

24: Coulometry
24.1: Introduction to Coulometry
24.2: Controlled-Potential Coulometry
24.3: Controlled-Current Coulometry

25: Voltammetry
25.1: Potential Excitation Signals and Currents in Voltammetry
25.2: Voltammetric Instrumentation
25.3: Linear Sweep Voltammetry
25.4: Cyclic Voltammetry
25.5: Polarography
25.6: Stripping Methods
25.7: Applications of Voltammetry

26: Introduction to Chromatographic Separations

26.1: A General Description of Chromatography
26.2: Migration Rates of Solutes
26.3: Zone Broadening and Column Ef ciency
26.4: Optimization and Column Performance
26.5: Summary of Important Relationships for Chromatography
26.6: Applications of Chromatography

27: Gas Chromatography

27.1: Principles of Gas Chromatography
27.2: Instruments for Gas Chromatography
27.3: Gas Chromatographic Columns and Stationary Phases
27.4: Applications of Gas Chromatography

28: High-Performance Liquid Chromatography

28.1: Scope of HPLC
28.2: Column Ef ciency in Liquid Chromatography
28.3: Instruments for Liquid Chromatography
28.4: Partition Chromatography
28.5: Adsorption Chromatography
28.6: Ion-Exchange Chromatography
28.7: Size-Exclusion Chromatography

29: Supercritical Fluid Chromatography

29.1: Properties of Supercritical Fluids
29.2: Supercritical Fluid Chromatography

30: Capillary Electrophoresis and Capillary Electrochromatography

30.1: An Overview of Electrophoresis
30.2: Capillary Electrophoresis
30.3: Applications of Capillary Electrophoresis

4 https://chem.libretexts.org/@go/page/332634
31: Thermal Methods
31.1: Thermogravimetry
31.2: Differential Thermal Analysis and Differential Scanning Calorimetry

32: Radiochemical Methods

32.1: Radioactive Isotopes
32.2: Instrumentation
32.3: Neutron Activation Methods
32.4: Isotope Dilution Methods

33: Automated Methods of Analysis

33.1: Overview of Automated Methods of Analysis
33.2: Flow-Injection Analysis
33.3: Other Automated Methods of Analysis

34: Particle Size Determination

34.1: Overview
34.2: Measuring Particle Size Using Sieves
34.3: Measuring Particle Size by Sedimentation
34.4: Measuring Particle Size Using Image Analysis
34.5: Measuring Particle Size Using Light Scattering

35: Appendicies
35.1: Evaluation of Analytical Data
35.2: Single-Sided Normal Distribution
35.3: Critical Values for t-Test
35.4: Critical Values for F-Test
35.5: Critical Values for Dixon's Q-Test
35.6: Critical Values for Grubb's Test
35.7: Activity Coef cients
35.8: Standard Reduction Potentials & Polarographic Half-wave Potentials
35.9: Recommended Primary Standards
35.10: Acronyms and Abbreviations

Index
Detailed Licensing

5 https://chem.libretexts.org/@go/page/332634
Licensing
A detailed breakdown of this resource's licensing can be found in Back Matter/Detailed Licensing.

1 https://chem.libretexts.org/@go/page/417081
 CHAPTER OVERVIEW
1: Introduction
Analytical chemistry is the science of how to make good measurements that we can use to solve a chemical problem. Many
problems in analytical chemistry begin with the need to identify what is present in a sample. This is the scope of a qualitative
analysis, examples of which include identifying the products of a chemical reaction, screening an athlete’s urine for a performance-
enhancing drug, or determining the spatial distribution of Pb on the surface of an airborne particulate. An early challenge for
analytical chemists was developing simple chemical tests to identify inorganic ions and organic functional groups. The classical
laboratory courses in inorganic and organic qualitative analysis, still taught at some schools, are based on this work.
Perhaps the most common analytical problem is a quantitative analysis, examples of which include the elemental analysis of a
newly synthesized compound, measuring the concentration of glucose in blood, or determining the difference between the bulk and
the surface concentrations of Cr in steel. Much of the analytical work in clinical, pharmaceutical, environmental, and industrial labs
involves developing new quantitative methods to detect trace amounts of chemical species in complex samples. Most of the
examples in this text are of quantitative analyses.
Another important area of analytical chemistry, which receives more limited attention in this text, are methods for characterizing
physical and chemical properties. The determination of chemical structure, particle size, and surface structure are examples of a
characterization analysis.
The purpose of a qualitative, a quantitative, or a characterization analysis is to solve a problem associated with a particular sample.
The purpose of a fundamental analysis, on the other hand, is to improve our understanding of the theory that supports an analytical
method and to understand better an analytical method’s limitations.
Like all areas of chemistry, analytical chemistry is so broad in scope and so much in flux that it is difficult to find a simple
definition more revealing than this quote attributed to C. N. Reilly (1925-1981), who was a professor of chemistry at the University
of North Carolina at Chapel Hill and one of the most influential analytical chemists of the last half of the twentieth century:
"Analytical chemistry is what analytical chemists do."
In this chapter we expand upon this simple definition by introducing approaches to making analytical measurements and by
developing a shared language for discussing analytical chemistry, more generally, and instrumentation, more specifically.
1.1: Classification of Analytical Methods
1.2: Types of Instrumental Methods
1.3: Instruments For Analysis
1.4: Selecting an Analytical Method
1.5: Calibration of Instrumental Methods

This page titled 1: Introduction is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

1
1.1: Classification of Analytical Methods
Analytical chemistry has a long history. On the bookshelf of my office, for example, there is a copy of the first American edition of
Fresenius's A System of Instruction in Quantitative Chemical Analysis, which was published by John Wiley & Sons in 1886.
Nearby are many newer texts, such as Bard and Faulkner's Electrochemical Methods: Fundamentals and Applications, the most
recent edition of which was published by Wiley in 2000. In 883 pages, Fresnius's text covers essentially all that was known in the
1880s about analytical chemistry and what we now call classical methods of analysis. Bard and Faulkner's text, which is 864 pages,
covers just one category of what we now call modern instrumental methods of analysis. Whether a classical method of analysis or a
modern instrumental method analysis, the species of interest, which we call the analyte, is probed in a way that provides qualitative
or quantitative information.

Classical Methods of Analysis

The distinguishing feature of a classical method of analysis is that the principal measurements are observations of reactions (Did a
precipitate form? Did the solution change color?) or the measurement of one of a small number of physical properties, such as mass
or volume. Because these measurements are not selective for a single analyte, a classical method of analysis usually required
extensive work to isolate the analyte of interest from other species that would interfere in the analysis. As we see in Figure 1.1.1,
Fresenius's method for determining the amount of nickel in ores required 58 hours, most of which was spent bringing the ore into
solution and then isolating the analyte from interferents by a sequence of precipitations and filtrations. The final determination of
the amount of nickel in the ore was derived from two measurements of mass: the combined mass of Co and Ni, and the mass of Co.
Although of historic interest, we will not consider further classical methods of analysis in this text.

1.1.1 https://chem.libretexts.org/@go/page/332624
 Figure 1.1.1 : Fresenius’ analytical scheme for the gravimetric analysis of Ni in ores. After each step, the solid and the solution are
separated by gravity filtration. Note that the mass of nickel is not determined directly. Instead, Co and Ni first are isolated and
weighed together (mass A), and then Co is isolated and weighed separately (mass B). The timeline shows that it takes
approximately 58 hours to analyze a single sample; although several sample could be prepared in parallel, the throughput of
samples was limited to just a few per day.

Modern Instrumental Methods of Analysis

The distinguishing feature of modern instrumental methods of analysis is that it extends measurements to many more physical
properties, such as current, potential, the absorption or emission of light, and mass-to-charge ratios, to name a few. Instrumental
methods for separating analytes, such as chromatographic separations, and instrumental methods that allow for the simultaneous
analysis of multiple analytes make for a much more rapid analysis. By the 1970s, flame atomic absorption spectrometry (FAAS)
replaced gravimetry as the standard method for analyzing nickel in ores [see, for example, Van Loon, J. C. Analytical Atomic
Absorption Spectroscopy, Academic Press: New York, 1980]. Because FAAS is much more selective than precipitation, there is
less need to chemically isolate the analyte; as a result, the time to analyze a single sample decreased to a few hours and the
throughput of samples increased to hundreds per day.

This page titled 1.1: Classification of Analytical Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

1.1.2 https://chem.libretexts.org/@go/page/332624
1.2: Types of Instrumental Methods
It is useful to organize instrumental methods of analysis into several groups based on the chemical or physical properties that we
use to generate a signal that we can measure and relate to the analyte of interest to us. One group of instrumental methods is based
on the interaction of photons of electromagnetic radiation with matter, which we call collectively spectroscopy. We can divide
spectroscopy into two broad classes of techniques. In one class of techniques there is a transfer of energy between the photon and
the sample. Table 1.2.1 provides a list of several representative examples.
Table 1.2.1 . Examples of Spectroscopic Instrumental Methods That Involve an Exchange of Energy Between a Photon and the Sample
type of energy transfer region of electromagnetic spectrum spectroscopic technique

absorption γ -ray Mossbauer spectroscopy

X-ray X-ray absorption spectroscopy

UV/Vis UV/Vis spectroscopy

IR infrared spectroscopy

microwave raman spectroscopy

electron spin resonance

radio wave
nuclear magnetic resonance

emission (thermal excitation) UV/Vis atomic emission spectroscopy

photoluminescence X-ray X-ray fluorescence

fluorescence spectroscopy
UV/Vis phosphorescence spectroscopy
atomic fluorescence spectroscopy

chemiluminescence UV/Vis chemiluminescence spectroscopy

In the second broad class of spectroscopic techniques, the electromagnetic radiation undergoes a change in amplitude, phase angle,
polarization, or direction of propagation as a result of its refraction, reflection, scattering, diffraction, or dispersion by the sample.
Several representative spectroscopic techniques are listed in Table 1.2.2.
Table 1.2.2 . Examples of Other Spectroscopic Instrumental Methods
region of electromagnetic spectrum type of interaction spectroscopic technique

X-ray diffraction X-ray diffraction

UV/Vis refraction refractrometry

nephelometry
scattering
turbidimetry

dispersion optical rotary dispersion

A second group of instrumental methods is based on the measurement of current, charge, or potential at the surface of an electrode,
sometimes while controlling one or both of the other two variables, and sometime while stirring the solution. Figure 1.2.1 provides
a visual introduction to these methods.

1.2.1 https://chem.libretexts.org/@go/page/332832
 Figure 1.2.1 Family tree that highlights the similarities and differences between a number of interfacial electrochemical techniques.
The specific instrumental methods are shown in red, the experimental conditions are shown in blue, and the analytical signals are
shown in green.
Our third group of instrumental methods gathers together a variety of other measurements that can provide a useful analytical
signal; these are summarized in Table 1.2.3.
Table 1.2.3 . Additional Examples of Instrumental Methods of Analysis
type of measurement or phenomenon instrumental method

piezoelectric effect quartz crystal microbalance

mass-to-charge ratio mass spectrometry

kinetic methods
flow injection analysis
rate of chemical reaction or physical process
neutron activation analysis
isotope diution analysis
thermal gravimetry
thermal energy differential thermal analysis
differential scanning calorimetry

Our last group of instrumental methods are used to separate mixtures based on either the equilibrium partitioning of species between two
phases or the migration of species in response to an applied electrical field. These methods usally are paired with a suitable instrumental
method from Table 1.2.1 , Table 1.2.2 , Table 1.2.3 , or Figure 1.2.1 to provide a way to follow the separation.
Table 1.2.4 . Examples of Instrumental Methods for Separating Mixtures.
basis of separation instrumental method
gas chromatography
equilibrium partitioning between two phases liquid chromatography
supercritical fluid chromatography
migration in response to applied electrical field electrophoresis

This page titled 1.2: Types of Instrumental Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1.2.2 https://chem.libretexts.org/@go/page/332832
1.3: Instruments For Analysis
An early example of a colorimetric analysis is Nessler’s method for ammonia, which was introduced in 1856. Nessler found that
adding an alkaline solution of HgI2 and KI to a dilute solution of ammonia produced a yellow-to-reddish brown colloid in which
the colloid’s color depended on the concentration of ammonia. In addition to the sample, Nessler prepared a series of standard
solutions, each containing a known amount of ammonia, and placed each in a glass tube with a flat bottom. Allowing sunlight to
pass through the tubes from bottom-to-top, Nessler observed them from above, as seen in Figure 1.3.1. By visually comparing the
color of the sample to the colors of the standards, Nessler was able to estimate the concentration of ammonia in the sample.

Figure 1.3.1 . Nessler’s original method for comparing the color of two solutions. Natural sunlight passes upwards through the
samples and standards and the analyst views the solutions by looking down toward the light source. The top view, shown on the
right, is what the analyst sees. To estimate the analyte’s concentration, the analyst exchanges standards until the sample's color falls
between the color of two standards.
Nessler's method converts a sample's chemical and/or physical properties—the color that forms when NH3 reacts with HgI2 and KI
—into a signal that we can detect, process, and report as a relative measure of the amount of NH3 in the sample. Although we
might not think of a Nessler tube as an instrument, the process of probing a sample in a way that converts its chemical or physical
properties into a form of information that we can report is the essence of any instrument.
The basic components of an instrument include a probe that interacts with the sample, an input transducer that converts the
sample's chemical and/or physical properties into an electrical signal, a signal processor that converts the electrical signal into a
form that an output transducer can convert into a numerical or visual output that we can understand. We can represent this as a
sequence of actions that take place within the instrument

probe → sample → input transducer → raw data → signal processor → output transducer

and as a general flow of information

chemical and/or physical information → electrical information → numerical or visual response

In Nessler’s method, the probe is sunlight, the analyst’s eye is the input transducer, the raw data is the response of the eye's optic
nerve to the attenuation of light, the signal processor is the brain, and the output is a visual report of the sample's color relative to
the standards.

sunlight → sample → eye → response of optic never → brain → visual report of color

Ways to Encode Information

As suggested above, information is encoded in two broad ways: as electrical information (such as currents and potentials) and as
information in other, non-electrical forms (such as chemical and physical properties).

Non-electrical Information
Nessler's method begins and ends with non-electrical forms of information: the sample has a color and we use that color to report
that the concentration of NH3 in our sample is greater than 0.50 mg/L and less than 1.00 mg/L. Other non-electrical ways to encode
information are the observation that a precipitate forms when we add Ag+ to a solution of NaCl, the balance beam scale that my
doctor uses to measure my weight, the percentage of light that passes through a sample, and the volume and moles of Cu(NO3)2 in
a graduated cylinder.

1.3.1 https://chem.libretexts.org/@go/page/332834
Electrical Information
Although my doctor's balance beam scale encodes my mass by the position of two movable weights along a signal arm–a decidedly
non-electrical means of encoding information—the electronic analytical balance that is found in almost all chemistry labs encodes
the mass in the form of electrical information (Figure 1.3.1). An electromagnet levitates the sample pan above a permanent
cylindrical magnet. When we place an object on the sample pan, it displaces the sample pan downward by a force equal to the
product of the sample’s mass and its acceleration due to gravity. The balance detects this downward movement and generates a
counterbalancing force by increasing the current to the electromagnet. The current needed to return the balance to its original
position is proportional to the object’s mass.

Figure 1.3.1 : The photo shows a typical electronic balance capable of determining mass to the nearest ±0.1 mg. The sticker inside
the balance’s wind shield is its annual calibration certification. For a review of other types of electronic balances, see Schoonover,
R. M. Anal. Chem. 1982, 54, 973A-980A.

Although we tend to use interchangeably, the terms “weight” and “mass,” there is an important distinction between them.
Mass is the absolute amount of matter in an object, measured in grams. Weight, W, is a measure of the gravitational force, g,
acting on that mass, m:

W = m ×g

An object has a fixed mass but its weight depends upon the acceleration due to gravity, which varies subtly from location-to-
location.
A balance measures an object’s weight, not its mass. Because weight and mass are proportional to each other, we can calibrate
a balance using a standard weight whose mass is traceable to the standard prototype for the kilogram. A properly calibrated
balance gives an accurate value for an object’s mass.

Electrical information comes in three domains: analog, time, and digital. In the analog domain, the signal shows the amplitude of
the electrical signal—say current or potential—as a function of an independent variable, which might be wavelength when
recording a spectrum, applied potential in a cyclic voltammetry experiment, or time when separating a mixture by gas
chromatography. A time domain signal shows the frequency with which the electrical signal rises above or below a threshold value,
as when counting the rate at which ionizing radiation, such as alpha or beta particles, are detected by a Geiger counter. Finally, in
the digital domain, the signal is a count of discrete events, such as counting the number of drops dispensed by an autotitrator by
allowing the drops to disrupt a beam of light.

Input Transducers, Detectors, and Sensors

As defined above, a transducer is a device that converts information from a non-electrical form to an electrical form (the input
transducer) or from an electrical form to a non-electrical form (the output transducer). Detector is a much broader term that
includes all aspects of the instrument from the input transducer to the output transducer; thus, a visible spectrometer is a detector
that uses an input transducer to convert the attenuation of the source radiation to a reported absorbance. A sensor is a detector
designed to monitor a particular analyte, such as a pH electrode.

Output Transducers and Readout Devices

An instrument's output transducer converts the information carried in electrical form into a non-electrical form that we can
understand. Common examples of output transducers, or readout devices, are a simple meter, a digital display, a physical trace of
the signal as a function of a dependent variable, such as a spectrum or a chromatogram, or a photographic plate.

1.3.2 https://chem.libretexts.org/@go/page/332834
Computers in Instruments
Many instruments include a computer that provides us with the ability to control the instrument and, perhaps of greater importance,
to process the data both by modifying the electrical signal as it passes from the input transducer to the output transducer, and by
providing tools for processing the data after it leaves the output transducer.

This page titled 1.3: Instruments For Analysis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1.3.3 https://chem.libretexts.org/@go/page/332834
1.4: Selecting an Analytical Method
The analysis of a sample generates a chemical or a physical signal that is proportional to the amount of analyte in the sample. This
signal may be anything we can measure, such the examples described in Section 1.2. It is convenient to divide analytical techniques
into two general classes based on whether the signal is directly proportional to the mass or moles of analyte, or is directly
proportional to the analyte’s concentration.
Consider the two graduated cylinders in Figure 1.4.1, each of which contains a solution of 0.010 M Cu(NO3)2. The cylinder on the
left contains 10 mL, or 1.0 × 10 moles of Cu2+, and the cylinder on the right contains 20 mL, or 2.0 × 10 moles of Cu2+. If a
−4 −4

technique responds to the absolute amount of analyte in the sample, then the signal due to the analyte SA is given as

SA = kA nA (1.4.1)

where nA is the moles or grams of analyte in the sample, and kA is a proportionality constant. Because the cylinder on the right
contains twice as many moles of Cu2+ as the cylinder on the left, analyzing its contents gives a signal twice as large as that for the
other cylinder.

Figure 1.4.1 : Two graduated cylinders, each containing 0.10 M Cu(NO3)2. Although the cylinders contain the same concentration
of Cu2+, the cylinder on the left contains 1.0 × 10 mol Cu2+ and the cylinder on the right contains 2.0 × 10 mol Cu2+.
−4 −4

A second class of analytical techniques are those that respond to the analyte’s concentration, CA
SA = kA CA (1.4.2)

In this case, an analysis of the contents of the two cylinders gives the same result. As most instruments respond to the analyte's
concentration, we will imit ourselves to using Equation 1.4.2 for the remainder of this section.

Defining the Problem

To select an appropriate analytical method for a particular problem we need to consider our needs and compare them to the
strengths and weaknesses of the available analytical methods. If we are screening samples on a production line to determine if an
analyte exceeds a threshold so that we can set them aside for a more careful analysis, then we may wish to give more consideration
to speed than to accuracy or precision. On the other hand, if we our analyte is part of a complex mixture, then we may wish to give
more consideration to analytical methods that provide for greater selectivity. Or, if we expect that our samples will vary
substantially in the concentration of analyte, then we may give more consideration to an analytical method for which Equation
1.4.2 applies over a wide range of concentrations.

Performance Characteristics of Instruments

As suggested above, when we choose an analytical method, we match the its performance characteristics (or figures of merit) to our
needs. Some of these characteristics are quantitative (accuracy, precision, sensitivity, detection limit, selectivity, dynamic range,
and selectivity) and others are more qualitative (robustness, ruggedness, scale of operation, time, and cost).

Accuracy
Accuracy, or bias, is a measure of how close the result of an experiment is to the “true” or expected result. We can express accuracy
as an absolute error, e
e = x −μ

1.4.1 https://chem.libretexts.org/@go/page/332836
where x is the experimental result and μ is the expected result, or as a percentage relative error, %er
x −μ
%er = × 100
μ

A method’s accuracy depends on many things, including the signal’s source, the value of kA in Equation 1.4.2 , and the ease of
handling samples without loss or contamination.

Because it is unlikely that we know the true result, we can use an expected or accepted result to evaluate accuracy. For
example, we might use a standard reference material, which has an accepted value for our analyte, to establish the analytical
method’s accuracy. You will find a more detailed treatment of accuracy in, including a discussion of sources of errors, in
Appendix 1.

Precision
When we analyze a sample several times, the individual results vary from trial-to-trial. Precision is a measure of this variability.
The closer the agreement between individual analyses, the more precise the results. For example, the results shown in the upper
half of Figure 1.4.2 for the concentration of potassium in a sample of serum are more precise than those in the lower half of Figure
1.4.2. It is important to understand that precision does not imply accuracy. That the data in the upper half of Figure 1.4.2 are more

precise does not mean that the first set of results is more accurate. In fact, neither set of results may be accurate.

Figure 1.4.2: Two determinations of the concentration of potassium in serum, showing the
effect of precision on the distribution of individual results. The data in (a) are less scattered
and, therefore, more precise than the data in (b).
A method’s precision depends on several factors, including the uncertainty in measuring the signal and the ease of handling
samples reproducibly, and is reported as an absolute standard deviation, s
−−−−−−−−−−−−−
n ¯¯¯
¯
2
∑i=1 (Xi − X )
s =√ (1.4.3)
n−1

or a relative standard deviation, sr

s
sr = (1.4.4)
¯¯¯
¯
X

¯¯¯
¯
where X is the average, or mean value of the individual measurements.
n

¯¯¯
¯
∑ Xi
i=1
X = (1.4.5)
n

Confusing accuracy and precision is a common mistake. See Ryder, J.; Clark, A. U. Chem. Ed. 2002, 6, 1–3, and Tomlinson,
J.; Dyson, P. J.; Garratt, J. U. Chem. Ed. 2001, 5, 16–23 for discussions of this and other common misconceptions about the
meaning of error. You will find a more detailed treatment of precision in Appendix 1, including a discussion of sources of
errors.

1.4.2 https://chem.libretexts.org/@go/page/332836
Sensitivity
The ability to demonstrate that two samples have different amounts of analyte is an essential part of many analyses. A method’s
sensitivity is a measure of its ability to establish that such a difference is significant. Sensitivity is often confused with a method’s
detection limit, which is the smallest amount of analyte we can determine with confidence.

See Pardue, H. L. Clin. Chem. 1997, 43, 1831-1837 for an explanation for why a method's sensitivity is not the same as its
detection limit.

Sensitivity is equivalent to the proportionality constant, kA, in Equation 1.4.2 [IUPAC Compendium of Chemical Terminology,
Electronic version]. If ΔS is the smallest difference we can measure between two signals, then the smallest detectable difference
A

in the analyte's concentration is

ΔSA
ΔCA =
kA

Suppose, for example, that our analytical signal is a measurement for which the smallest detectable increment is ±0.001 (arbitrary
units). If our method’s sensitivity is 0.200M , then our method can conceivably detect a difference in concentration of as little as
−1

±0.001 −1
ΔCA = = ±0.005  M
−1
0.200 M

For two methods with the same ΔS , the method with the greater sensitivity—that is, the method with the larger kA—is better able
A

to discriminate between smaller amounts of analyte.

Detection Limit
The International Union of Pure and Applied Chemistry (IUPAC) defines a method’s detection limit as the smallest concentration
or absolute amount of analyte that has a signal significantly larger than the signal from a suitable blank [IUPAC Compendium of
Chemical Technology, Electronic Version]. Although our interest is in the amount of analyte, in this section we will define the
detection limit in terms of the analyte’s signal. Knowing the signal, we can calculate the analyte’s concentration, CA, using
Equation 1.4.2, S = k C where k is the method’s sensitivity.
A A A

Let’s translate the IUPAC definition of the detection limit into a mathematical form by letting Smb represent the average signal for a
method blank, and letting σ represent the method blank’s standard deviation. To detect the analyte, its signal must exceed Smb by
mb

a suitable amount; thus,

(SA )DL = Smb ± zσmb (1.4.6)

where (S A )DL is the analyte’s detection limit.

The value we choose for z depends on our tolerance for reporting the analyte’s concentration even if it is absent from the sample
(what is called a type 1 error). Typically, z is set to three, which corresponds to a probability, α , of 0.00135, or 0.135%. As shown
in Figure 1.4.3a, there is only a 0.135% probability of detecting the analyte in a sample that actually is analyte-free.

Figure 1.4.3 : Normal distribution curves showing the probability of type 1 and type 2 errors for the IUPAC detection limit. (a) The
normal distribution curve for the method blank, with Smb = 0 and σ = 1. The minimum detectable signal for the analyte, (SA)DL,
mb

has a type 1 error of 0.135%. (b) The normal distribution curve for the analyte at its detection limit, (SA)DL= 3, is superimposed on
the normal distribution curve for the method blank. The standard deviation for the analyte’s signal, σ , is 0.8, The area in green
A

represents the probability of a type 2 error, which is 50%. The inset shows, in blue, the probability of a type 1 error, which is
0.135%.

1.4.3 https://chem.libretexts.org/@go/page/332836
A detection limit also is subject to a type 2 error in which we fail to find evidence for the analyte even though it is present in the
sample. Consider, for example, the situation shown in Figure 1.4.3b where the signal for a sample that contains the analyte is
exactly equal to (SA)DL. In this case the probability of a type 2 error is 50% because half of the sample’s possible signals are below
the detection limit. We correctly detect the analyte at the IUPAC detection limit only half the time. The IUPAC definition for the
detection limit is the smallest signal for which we can say, at a significance level of α , that an analyte is present in the sample;
however, failing to detect the analyte does not mean it is not present in the sample.
The detection limit often is represented, particularly when discussing public policy issues, as a distinct line that separates detectable
concentrations of analytes from concentrations we cannot detect. This use of a detection limit is incorrect [Rogers, L. B. J. Chem.
Educ. 1986, 63, 3–6]. As suggested by Figure 1.4.3, for an analyte whose concentration is near the detection limit there is a high
probability that we will fail to detect the analyte.
An alternative expression for the detection limit, the limit of identification, minimizes both type 1 and type 2 errors [Long, G. L.;
Winefordner, J. D. Anal. Chem. 1983, 55, 712A–724A]. The analyte’s signal at the limit of identification, (SA)LOI, includes an
additional term, zσ , to account for the distribution of the analyte’s signal.
A

(SA )LOI = (SA )DL + zσA = Smb + zσmb + zσA (1.4.7)

As shown in Figure 1.4.4, the limit of identification provides an equal probability of a type 1 and a type 2 error at the detection
limit. When the analyte’s concentration is at its limit of identification, there is only a 0.135% probability that its signal is
indistinguishable from that of the method blank.

Figure : Normal distribution curves for a method blank and for a sample at the limit of identification: Smb = 0; σ = 1 ;
1.4.4 mb

σA = 0.8 ; and (SA)LOI = 0 + 3 × 1 + 3 × 0.8 = 5.4. The inset shows that the probability of a type 1 error (0.135%) is the same as
the probability of a type 2 error (0.135%).
The ability to detect the analyte with confidence is not the same as the ability to report with confidence its concentration, or to
distinguish between its concentration in two samples. For this reason the American Chemical Society’s Committee on
Environmental Analytical Chemistry recommends the limit of quantitation, (SA)LOQ [“Guidelines for Data Acquisition and Data
Quality Evaluation in Environmental Chemistry,” Anal. Chem. 1980, 52, 2242–2249 ].
(SA )LOQ = Smb + 10 σmb (1.4.8)

Dynamic Range
A method's dynamic range (or linear range) runs from its limit of quantication, (Equation 1.4.8, to the highest concentration for
which the sensitivity, kA, remains constant, resulting in a straight-line relationship between S and C . This upper limit is called
A A

the limit of linearity, LOL. Between the LOQ and the LOL we can use Equation 1.4.2 to convert a measured signal into the
corresponding concentration of the analyte. Above the LOQ the relationship between the signal and the analyte's concentration no
longer is a straight-line.

Selectivity
An analytical method is specific if its signal depends only on the analyte [Persson, B-A; Vessman, J. Trends Anal. Chem. 1998, 17,
117–119; Persson, B-A; Vessman, J. Trends Anal. Chem. 2001, 20, 526–532]. Although specificity is the ideal, few analytical
methods are free from interferences. When an interferent, I, contributes to the signal, we expand 1.4.1 and Equation 1.4.2 to
include its contribution to the sample’s signal, Ssamp

1.4.4 https://chem.libretexts.org/@go/page/332836
 Ssamp = SA + SI = kA CA + kI CI (1.4.9)

where SI is the interferent’s contribution to the signal, kI is the interferent’s sensitivity, and CI is the concentration of interferent in
the sample.
Selectivity is a measure of a method’s freedom from interferences [Valcárcel, M.; Gomez-Hens, A.; Rubio, S. Trends Anal. Chem.
2001, 20, 386–393]. A method’s selectivity for an interferent relative to the analyte is defined by a selectivity coefficient, KA,I
kI
KA,I = (1.4.10)
kA

which may be positive or negative depending on the signs of kI and kA. The selectivity coefficient is greater than +1 or less than –1
when the method is more selective for the interferent than for the analyte.

Although kA and kI usually are positive, they can be negative. For example, some analytical methods work by measuring the
concentration of a species that remains after is reacts with the analyte. As the analyte’s concentration increases, the
concentration of the species that produces the signal decreases, and the signal becomes smaller. If the signal in the absence of
analyte is assigned a value of zero, then the subsequent signals are negative.

Determining the selectivity coefficient’s value is easy if we already know the values for kA and kI. As shown by Example 1.4.1, we
also can determine KA,I by measuring Ssamp in the presence of and in the absence of the interferent.

 Example 1.4.1

A method for the analysis of Ca2+ in water suffers from an interference in the presence of Zn2+. When the concentration of
Ca2+ is 100 times greater than that of Zn2+, an analysis for Ca2+ has a relative error of +0.5%. What is the selectivity
coefficient for this method?
Solution
Since only relative concentrations are reported, we can arbitrarily assign absolute concentrations. To make the calculations
easy, we will let CCa = 100 (arbitrary units) and CZn = 1. A relative error of +0.5% means the signal in the presence of Zn2+ is
0.5% greater than the signal in the absence of Zn2+. Again, we can assign values to make the calculation easier. If the signal for
Cu2+ in the absence of Zn2+ is 100 (arbitrary units), then the signal in the presence of Zn2+ is 100.5.
The value of kCa is determined using Equation 1.4.2
SCa 100
kCa = = =1
CCa 100

In the presence of Zn2+ the signal is given by Equation 1.4.9; thus

Ssamp = 100.5 = kCa CCa + kZn CZn = (1 × 100) + kZn × 1

Solving for kZn gives its value as 0.5. The selectivity coefficient is
kZn 0.5
KCa,Zn = = = 0.5
kCa 1

If you are unsure why, in the above example, the signal in the presence of zinc is 100.5, note that the percentage relative error
for this problem is given by
obtained result − 100
× 100 = +0.5%
100

Solving gives an obtained result of 100.5.

A selectivity coefficient provides us with a useful way to evaluate an interferent’s potential effect on an analysis. Solving Equation
1.4.10 for kI

kI = KA,I × kA (1.4.11)

1.4.5 https://chem.libretexts.org/@go/page/332836
and substituting in Equation 1.4.2 and simplifying gives
Ssamp = kA { CA + KA,I × CI } (1.4.12)

An interferent will not pose a problem as long as the term K A,I × CI in Equation 1.4.12 is significantly smaller than than CA.

 Example 1.4.2

Barnett and colleagues developed a method to determine the concentration of codeine (structure shown below) in poppy plants
[Barnett, N. W.; Bowser, T. A.; Geraldi, R. D.; Smith, B. Anal. Chim. Acta 1996, 318, 309– 317]. As part of their study they
evaluated the effect of several interferents. For example, the authors found that equimolar solutions of codeine and the
interferent 6-methoxycodeine gave signals, respectively of 40 and 6 (arbitrary units).

(a) What is the selectivity coefficient for the interferent, 6-methoxycodeine, relative to that for the analyte, codeine.
(b) If we need to know the concentration of codeine with an accuracy of ±0.50%, what is the maximum relative concentration
of 6-methoxy-codeine that we can tolerate?
Solution
(a) The signals due to the analyte, SA, and the interferent, SI, are

SA = kA CA SI = kI CI

Solving these equations for kA and for kI, and substituting into Equation 1.4.10 gives
SI / CI
KA,I =
SA / CI

Because the concentrations of analyte and interferent are equimolar (CA = CI), the selectivity coefficient is
SI 6
KA,I = = = 0.15
SA 40

(b) To achieve an accuracy of better than ±0.50% the term K A,I × CI in Equation 1.4.12 must be less than 0.50% of CA; thus

KA,I × CI ≤ 0.0050 × CA

Solving this inequality for the ratio CI/CA and substituting in the value for KA,I from part (a) gives
CI 0.0050 0.0050
≤ = = 0.033
CA KA,I 0.15

Therefore, the concentration of 6-methoxycodeine must be less than 3.3% of codeine’s concentration.

Problems with selectivity also are more likely when the analyte is present at a very low concentration [Rodgers, L. B. J. Chem.
Educ. 1986, 63, 3–6].

Robustness and Ruggedness

For a method to be useful it must provide reliable results. Unfortunately, methods are subject to a variety of chemical and physical
interferences that contribute uncertainty to the analysis. If a method is relatively free from chemical interferences, we can use it to
analyze an analyte in a wide variety of sample matrices. Such methods are considered robust.
Random variations in experimental conditions introduces uncertainty. If a method’s sensitivity, k, is too dependent on experimental
conditions, such as temperature, acidity, or reaction time, then a slight change in any of these conditions may give a significantly
different result. A rugged method is relatively insensitive to changes in experimental conditions.

1.4.6 https://chem.libretexts.org/@go/page/332836
Scale of Operation
Another way to narrow the choice of methods is to consider three potential limitations: the amount of sample available for the
analysis, the expected concentration of analyte in the samples, and the minimum amount of analyte that will produce a measurable
signal. Collectively, these limitations define the analytical method’s scale of operations.
We can display the scale of operations visually (Figure 1.4.5) by plotting the sample’s size on the x-axis and the analyte’s
concentration on the y-axis. For convenience, we divide samples into macro (>0.1 g), meso (10 mg–100 mg), micro (0.1 mg–10
mg), and ultramicro (<0.1 mg) sizes, and we divide analytes into major (>1% w/w), minor (0.01% w/w–1% w/w), trace (10–7%
w/w–0.01% w/w), and ultratrace (<10–7% w/w) components. Together, the analyte’s concentration and the sample’s size provide a
characteristic description for an analysis. For example, in a microtrace analysis the sample weighs between 0.1 mg and 10 mg and
contains a concentration of analyte between 10–7% w/w and 10–2% w/w.

Figure 1.4.5 : Scale of operations for analytical methods. The shaded areas define different types of analyses. The boxed area, for
example, represents a microtrace analysis. The diagonal lines show combinations of sample size and analyte concentration that
contain the same mass of analyte. The three filled circles (•), for example, indicate analyses that use 10 mg of analyte. See Sandell,
E. B.; Elving, P. J. in Kolthoff, I. M.; Elving, P. J., eds. Treatise on Analytical Chem-istry, Interscience: New York, Part I, Vol. 1,
Chapter 1, pp. 3–6; (b) Potts, L. W. Quantitative Analysis–Theory and Practice, Harper and Row: New York, 1987, pp. 12 for more
details.
The diagonal lines connecting the axes show combinations of sample size and analyte concentration that contain the same absolute
mass of analyte. As shown in Figure 1.4.5, for example, a 1-g sample that is 1% w/w analyte has the same amount of analyte (10
mg) as a 100-mg sample that is 10% w/w analyte, or a 10-mg sample that is 100% w/w analyte.
We can use Figure 1.4.5 to establish limits for analytical methods. If a method’s minimum detectable signal is equivalent to 10 mg
of analyte, then it is best suited to a major analyte in a macro or meso sample. Extending the method to an analyte with a
concentration of 0.1% w/w requires a sample of 10 g, which rarely is practical due to the complications of carrying such a large
amount of material through the analysis. On the other hand, a small sample that contains a trace amount of analyte places
significant restrictions on an analysis. For example, a 1-mg sample that is 10–4% w/w in analyte contains just 1 ng of analyte. If we
isolate the analyte in 1 mL of solution, then we need an analytical method that reliably can detect it at a concentration of 1 ng/mL.

Equipment, Time, and Cost

Finally, we can compare analytical methods with respect to their equipment needs, the time needed to complete an analysis, and the
cost per sample. Methods that rely on instrumentation are equipment-intensive and may require significant operator training. For
example, the graphite furnace atomic absorption spectroscopic method for determining lead in water requires a significant capital
investment in the instrument and an experienced operator to obtain reliable results. Other methods, such as titrimetry, require less
expensive equipment and less training.
The time to complete an analysis for one sample often is fairly similar from method-to-method. This is somewhat misleading,
however, because much of this time is spent preparing samples, preparing reagents, and gathering together equipment. Once the
samples, reagents, and equipment are in place, the sampling rate may differ substantially. For example, it takes just a few minutes
to analyze a single sample for lead using graphite furnace atomic absorption spectroscopy, but several hours to analyze the same
sample using gravimetry. This is a significant factor in selecting a method for a laboratory that handles a high volume of samples.
The cost of an analysis depends on many factors, including the cost of equipment and reagents, the cost of hiring analysts, and the
number of samples that can be processed per hour. In general, methods that rely on instruments cost more per sample then other

1.4.7 https://chem.libretexts.org/@go/page/332836
methods.

Making the Final Choice

Unfortunately, the design criteria discussed in this section are not mutually independent [Valcárcel, M.; Ríos, A. Anal. Chem. 1993,
65, 781A–787A]. Working with smaller samples or improving selectivity often comes at the expense of precision. Minimizing cost
and analysis time may decrease accuracy. Selecting a method requires carefully balancing the various design criteria. Usually, the
most important design criterion is accuracy, and the best method is the one that gives the most accurate result. When the need for a
result is urgent, as is often the case in clinical labs, analysis time may become the critical factor.
In some cases it is the sample’s properties that determine the best method. A sample with a complex matrix, for example, may
require a method with excellent selectivity to avoid interferences. Samples in which the analyte is present at a trace or ultratrace
concentration usually require a concentration method. If the quantity of sample is limited, then the method must not require a large
amount of sample.

This page titled 1.4: Selecting an Analytical Method is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1.4.8 https://chem.libretexts.org/@go/page/332836
1.5: Calibration of Instrumental Methods
To standardize an analytical method we also must determine the analyte’s sensitivity, kA, in the following equation
Stotal = kA CA + Sblank (1.5.1)

where S total is the measured signal, C is the analyte's concentration, and S

A
is the signal in the absence of the analyte. In
reag

principle, it is possible to derive the value of kA for any analytical method if we understand fully all the chemical reactions and
physical processes responsible for the signal. Unfortunately, such calculations are not feasible if we lack a sufficiently developed
theoretical model of the physical processes or if the chemical reaction’s evince non-ideal behavior. In such situations we must
determine the value of kA by analyzing one or more standard solutions, each of which contains a known amount of analyte. In this
section we consider several approaches for determining the value of kA. For simplicity we assume that S is accounted for by a
blank

proper reagent blank, allowing us to replace Stotal in with the analyte’s signal, SA.

SA = kA CA (1.5.2)

Single-Point Versus Multiple-Point Standardization

The simplest way to determine the value of kA in Equation 1.5.2 is to use a single-point standardization in which we measure the
signal for a standard, Sstd, that contains a known concentration of analyte, Cstd. Substituting these values into Equation 1.5.2 and
rearrange
Sstd
kA = (1.5.3)
Cstd

to give us the value for kA. Having determined kA, we can calculate the concentration of analyte in a sample by measuring its signal,
Ssamp, and calculate CA as
Ssamp
CA = (1.5.4)
kA

A single-point standardization is the least desirable method for standardizing a method. There are two reasons for this. First, any
error in our determination of kA carries over into our calculation of CA. Second, our experimental value for kA is based on a single
concentration of analyte. To extend this value of kA to other concentrations of analyte requires that we assume a linear relationship
between the signal and the analyte’s concentration, an assumption that often is not true [Cardone, M. J.; Palmero, P. J.; Sybrandt, L.
B. Anal. Chem. 1980, 52, 1187–1191]. Figure 1.5.1 shows how assuming a constant value of kA leads to a determinate error in CA
if kA becomes smaller at higher concentrations of analyte. Despite these limitations, single-point standardizations find routine use
when the expected range for the analyte’s concentrations is small. Under these conditions it often is safe to assume that kA is
constant (although you should verify this assumption experimentally). This is the case, for example, in clinical labs where many
automated analyzers use only a single standard.

Figure 1.5.1 : Example showing how a single-point standardization leads to a determinate error in an analyte’s reported
concentration if we incorrectly assume that kA is constant. The assumed relationship between Ssamp and CA is based on a single
standard and is a straight-line; the actual relationship between Ssamp and CA becomes curved for larger concentrations of analyte.
The better way to standardize a method is to prepare a series of standards, each of which contains a different concentration of
analyte. Standards are chosen such that they bracket the expected range for the analyte’s concentration. A multiple-point
standardization should include at least three standards, although more are preferable. A plot of Sstd versus Cstd is called a calibration
curve. The exact standardization, or calibration relationship, is determined by an appropriate curve-fitting algorithm.

1.5.1 https://chem.libretexts.org/@go/page/332838
 Linear regression, which also is known as the method of least squares, is one such algorithm. Its use is covered in Appendix 1.

There are two advantages to a multiple-point standardization. First, although a determinate error in one standard introduces a
determinate error, its effect is minimized by the remaining standards. Second, because we measure the signal for several
concentrations of analyte, we no longer must assume kA is independent of the analyte’s concentration. Instead, we can construct a
calibration curve similar to the “actual relationship” in Figure 1.5.1.

External Standards
The most common method of standardization uses one or more external standards, each of which contains a known concentration
of analyte. We call these standards “external” because they are prepared and analyzed separate from the samples.

Appending the adjective “external” to the noun “standard” might strike you as odd at this point, as it seems reasonable to
assume that standards and samples are analyzed separately. As we will soon learn, however, we can add standards to our
samples and analyze both simultaneously.

Single External Standard

With a single external standard we determine kA using EEquation 1.5.3 and then calculate the concentration of analyte, CA, using
Equation 1.5.4.

 Example 1.5.1

A spectrophotometric method for the quantitative analysis of Pb2+ in blood yields an Sstd of 0.474 for a single standard for
which the concentration of lead is 1.75 ppb. What is the concentration of Pb2+ in a sample of blood for which Ssamp is 0.361?
Solution
Equation 1.5.3 allows us to calculate the value of kA using the data for the single external standard.
Sstd 0.474 −1
kA = = = 0.2709  ppb
Cstd 1.75 ppb

Having determined the value of kA, we calculate the concentration of Pb2+ in the sample of blood is calculated using Equation
1.5.4.

Ssamp 0.361
CA = = = 1.33 ppb
−1
kA 0.2709 ppb

Multiple External Standards

Figure 1.5.2 shows a typical multiple-point external standardization. The volumetric flask on the left contains a reagent blank and
the remaining volumetric flasks contain increasing concentrations of Cu2+. Shown below the volumetric flasks is the resulting
calibration curve. Because this is the most common method of standardization, the resulting relationship is called a normal
calibration curve.

1.5.2 https://chem.libretexts.org/@go/page/332838
 Figure 1.5.2 : The photo at the top of the figure shows a reagent blank (far left) and a set of five external standards for Cu2+ with
concentrations that increase from left-to-right. Shown below the external standards is the resulting normal calibration curve. The
absorbance of each standard, Sstd, is shown by the filled circles.
When a calibration curve is a straight-line, as it is in Figure 1.5.2, the slope of the line gives the value of kA. This is the most
desirable situation because the method’s sensitivity remains constant throughout the analyte’s concentration range. When the
calibration curve is not a straight-line, the method’s sensitivity is a function of the analyte’s concentration. In Figure 1.5.1, for
example, the value of kA is greatest when the analyte’s concentration is small and it decreases continuously for higher
concentrations of analyte. The value of kA at any point along the calibration curve in Figure 1.5.1 is the slope at that point. In either
case, a calibration curve allows to relate Ssamp to the analyte’s concentration.

 Example 1.5.2

A second spectrophotometric method for the quantitative analysis of Pb2+ in blood has a normal calibration curve for which
−1
Sstd = (0.296  ppb × Cstd ) + 0.003

What is the concentration of Pb2+ in a sample of blood if Ssamp is 0.397?

Solution
To determine the concentration of Pb2+ in the sample of blood, we replace Sstd in the calibration equation with Ssamp and solve
for CA.
Ssamp − 0.003 0.397 − 0.003
CA = = = 1.33 ppb
−1 −1
0.296 ppb 0.296 ppb

It is worth noting that the calibration equation in this problem includes an extra term that does not appear in Equation 1.5.4.
Ideally we expect our calibration curve to have a signal of zero when CA is zero. This is the purpose of using a reagent blank to
correct the measured signal. The extra term of +0.003 in our calibration equation results from the uncertainty in measuring the
signal for the reagent blank and the standards.

An external standardization allows us to analyze a series of samples using a single calibration curve. This is an important advantage
when we have many samples to analyze. Not surprisingly, many of the most common quantitative analytical methods use an
external standardization.
There is a serious limitation, however, to an external standardization. When we determine the value of kA using Equation 1.5.3, the
analyte is present in the external standard’s matrix, which usually is a much simpler matrix than that of our samples. When we use
an external standardization we assume the matrix does not affect the value of kA. If this is not true, then we introduce a proportional
determinate error into our analysis. This is not the case in Figure 1.5.3, for instance, where we show calibration curves for an
analyte in the sample’s matrix and in the standard’s matrix. In this case, using the calibration curve for the external standards leads
to a negative determinate error in analyte’s reported concentration. If we expect that matrix effects are important, then we try to
match the standard’s matrix to that of the sample, a process known as matrix matching. If we are unsure of the sample’s matrix,
then we must show that matrix effects are negligible or use an alternative method of standardization. Both approaches are discussed
in the following section.

1.5.3 https://chem.libretexts.org/@go/page/332838
 2 +
The matrix for the external standards in Figure 1.5.2, for example, is dilute ammonia. Because the Cu(NH ) complex 3 4

absorbs more strongly than Cu2+, adding ammonia increases the signal’s magnitude. If we fail to add the same amount of
ammonia to our samples, then we will introduce a proportional determinate error into our analysis.

Figure 1.5.3 : Calibration curves for an analyte in the standard’s matrix and in the sample’s matrix. If the matrix affects the value of
kA, as is the case here, then we introduce a proportional determinate error into our analysis if we use a normal calibration curve.

Standard Additions
We can avoid the complication of matching the matrix of the standards to the matrix of the sample if we carry out the
standardization in the sample. This is known as the method of standard additions.

Single Standard Addition

The simplest version of a standard addition is shown in Figure 1.5.4. First we add a portion of the sample, Vo, to a volumetric flask,
dilute it to volume, Vf, and measure its signal, Ssamp. Next, we add a second identical portion of sample to an equivalent volumetric
flask along with a spike, Vstd, of an external standard whose concentration is Cstd. After we dilute the spiked sample to the same
final volume, we measure its signal, Sspike.

Figure 1.5.4 : Illustration showing the method of standard additions. The volumetric flask on the left contains a portion of the
sample, Vo, and the volumetric flask on the right contains an identical portion of the sample and a spike, Vstd, of a standard solution
of the analyte. Both flasks are diluted to the same final volume, Vf. The concentration of analyte in each flask is shown at the
bottom of the figure where CA is the analyte’s concentration in the original sample and Cstd is the concentration of analyte in the
external standard.
The following two equations relate Ssamp and Sspike to the concentration of analyte, CA, in the original sample.
Vo
Ssamp = kA CA (1.5.5)
Vf

Vo Vstd
Sspike = kA ( CA + Cstd ) (1.5.6)
Vf Vf

As long as Vstd is small relative to Vo, the effect of the standard’s matrix on the sample’s matrix is insignificant. Under these
conditions the value of kA is the same in Equation 1.5.5 and Equation 1.5.6. Solving both equations for kA and equating gives

1.5.4 https://chem.libretexts.org/@go/page/332838
 Ssamp Sspike
= (1.5.7)
Vo Vo Vstd
CA CA + Cstd
Vf Vf Vf

which we can solve for the concentration of analyte, CA, in the original sample.

 Example 1.5.3
A third spectrophotometric method for the quantitative analysis of Pb2+ in blood yields an Ssamp of 0.193 when a 1.00 mL
sample of blood is diluted to 5.00 mL. A second 1.00 mL sample of blood is spiked with 1.00 mL of a 1560-ppb Pb2+ external
standard and diluted to 5.00 mL, yielding an Sspike of 0.419. What is the concentration of Pb2+ in the original sample of blood?
Solution
We begin by making appropriate substitutions into Equation 1.5.7 and solving for CA. Note that all volumes must be in the
same units; thus, we first covert Vstd from 1.00 mL to 1.00 × 10  mL . −3

0.193 0.419
=
−3
1.00 mL 1.00 mL 1.00×10  mL
CA CA + 1560 ppb
5.00 mL 5.00 mL 5.00 mL

0.193 0.419
=
0.200CA 0.200 CA + 0.3120 ppb

0.0386 CA + 0.0602 ppb = 0.0838 CA

0.0452 CA = 0.0602 ppb

CA = 1.33 ppb

The concentration of Pb2+ in the original sample of blood is 1.33 ppb.

It also is possible to add the standard addition directly to the sample, measuring the signal both before and after the spike (Figure
1.5.5). In this case the final volume after the standard addition is Vo + Vstd and Equation 1.5.5, Equation 1.5.6, and Equation 1.5.7

become

Ssamp = kA CA (1.5.8)

Vo Vstd
Sspike = kA ( CA + Cstd ) (1.5.9)
Vo + Vstd Vo + Vstd

Ssamp Sspike
= (1.5.10)
Vo Vstd
CA
CA + Cstd
Vo +Vstd Vo +Vstd

Figure 1.5.5 : Illustration showing an alternative form of the method of standard additions. In this case we add the spike of external
standard directly to the sample without any further adjust in the volume.

 Example 1.5.4

A fourth spectrophotometric method for the quantitative analysis of Pb2+ in blood yields an Ssamp of 0.712 for a 5.00 mL
sample of blood. After spiking the blood sample with 5.00 mL of a 1560-ppb Pb2+ external standard, an Sspike of 1.546 is
measured. What is the concentration of Pb2+ in the original sample of blood?

1.5.5 https://chem.libretexts.org/@go/page/332838
 Solution
0.712 1.546
=
−3
CA 5.00 mL 5.00×10  mL
CA + 1560 ppb
5.005 mL 5.005 mL

0.712 1.546
=
CA 0.9990 CA + 1.558 ppb

0.7113 CA + 1.109 ppb = 1.546 CA

CA = 1.33 ppb

The concentration of Pb2+ in the original sample of blood is 1.33 ppb.

Multiple Standard Additions

We can adapt a single-point standard addition into a multiple-point standard addition by preparing a series of samples that contain
increasing amounts of the external standard. Figure 1.5.6 shows two ways to plot a standard addition calibration curve based on
Equation 1.5.6. In Figure 1.5.6a we plot Sspike against the volume of the spikes, Vstd. If kA is constant, then the calibration curve is a
straight-line. It is easy to show that the x-intercept is equivalent to –CAVo/Cstd.

Figure 1.5.6 : Shown at the top of the figure is a set of six standard additions for the determination of Mn2+. The flask on the left is
a 25.00 mL sample diluted to 50.00 mL with water. The remaining flasks contain 25.00 mL of sample and, from left-to-right, 1.00,
2.00, 3.00, 4.00, and 5.00 mL spikes of an external standard that is 100.6 mg/L Mn2+. Shown below are two ways to plot the
standard additions calibration curve. The absorbance for each standard addition, Sspike, is shown by the filled circles.

 Example 1.5.5

Beginning with Equation 1.5.6 show that the equations in Figure 1.5.6 a for the slope, the y-intercept, and the x-intercept are
correct.
Solution
We begin by rewriting Equation 1.5.6 as
kA CA Vo kA Cstd
Sspike = + × Vstd
Vf Vf

which is in the form of the equation for a straight-line

y = y-intercept + slope × x-intercept

where y is Sspike and x is Vstd. The slope of the line, therefore, is kACstd/Vf and the y-intercept is kACAVo/Vf. The x-intercept is
the value of x when y is zero, or

1.5.6 https://chem.libretexts.org/@go/page/332838
 kA CA Vo kA Cstd
0 = + × x-intercept
Vf Vf

kA CA Vo / Vf CA Vo
x-intercept = − =−
KA Cstd / Vf Cstd

Because we know the volume of the original sample, Vo, and the concentration of the external standard, Cstd, we can calculate the
analyte’s concentrations from the x-intercept of a multiple-point standard additions.

 Example 1.5.6
A fifth spectrophotometric method for the quantitative analysis of Pb2+ in blood uses a multiple-point standard addition based
on Equation 1.5.6. The original blood sample has a volume of 1.00 mL and the standard used for spiking the sample has a
concentration of 1560 ppb Pb2+. All samples were diluted to 5.00 mL before measuring the signal. A calibration curve of Sspike
versus Vstd has the following equation
−1
Sspike = 0.266 + 312  mL × Vstd

What is the concentration of Pb2+ in the original sample of blood?

Solution
To find the x-intercept we set Sspike equal to zero.
−1
Sspike = 0.266 + 312  mL × Vstd

Solving for Vstd, we obtain a value of −8.526 × 10

−4
 mL for the x-intercept. Substituting the x-intercept’s value into the
equation from Figure 1.5.6a
CA Vo CA × 1.00 mL
−4
−8.526 × 10  mL = − =−
Cstd 1560 ppb

and solving for CA gives the concentration of Pb2+ in the blood sample as 1.33 ppb.

Since we construct a standard additions calibration curve in the sample, we can not use the calibration equation for other samples.
Each sample, therefore, requires its own standard additions calibration curve. This is a serious drawback if you have many samples.
For example, suppose you need to analyze 10 samples using a five-point calibration curve. For a normal calibration curve you need
to analyze only 15 solutions (five standards and ten samples). If you use the method of standard additions, however, you must
analyze 50 solutions (each of the ten samples is analyzed five times, once before spiking and after each of four spikes).

Using a Standard Addition to Identify Matrix Effects

We can use the method of standard additions to validate an external standardization when matrix matching is not feasible. First, we
prepare a normal calibration curve of Sstd versus Cstd and determine the value of kA from its slope. Next, we prepare a standard
additions calibration curve using Equation 1.5.6, plotting the data as shown in Figure 1.5.6b. The slope of this standard additions
calibration curve provides an independent determination of kA. If there is no significant difference between the two values of kA,
then we can ignore the difference between the sample’s matrix and that of the external standards. When the values of kA are
significantly different, then using a normal calibration curve introduces a proportional determinate error.

Internal Standards
To use an external standardization or the method of standard additions, we must be able to treat identically all samples and
standards. When this is not possible, the accuracy and precision of our standardization may suffer. For example, if our analyte is in
a volatile solvent, then its concentration will increase if we lose solvent to evaporation. Suppose we have a sample and a standard
with identical concentrations of analyte and identical signals. If both experience the same proportional loss of solvent, then their
respective concentrations of analyte and signals remain identical. In effect, we can ignore evaporation if the samples and the
standards experience an equivalent loss of solvent. If an identical standard and sample lose different amounts of solvent, however,
then their respective concentrations and signals are no longer equal. In this case a simple external standardization or standard
addition is not possible.

1.5.7 https://chem.libretexts.org/@go/page/332838
We can still complete a standardization if we reference the analyte’s signal to a signal from another species that we add to all
samples and standards. The species, which we call an internal standard, must be different than the analyte.
Because the analyte and the internal standard receive the same treatment, the ratio of their signals is unaffected by any lack of
reproducibility in the procedure. If a solution contains an analyte of concentration CA and an internal standard of concentration CIS,
then the signals due to the analyte, SA, and the internal standard, SIS, are

SA = kA CA

SIS = kSI CIS

where k and k are the sensitivities for the analyte and the internal standard, respectively. Taking the ratio of the two signals
A IS

gives the fundamental equation for an internal standardization.

SA kA CA CA
= =K× (1.5.11)
SIS kIS CIS CIS

Because K is a ratio of the analyte’s sensitivity and the internal standard’s sensitivity, it is not necessary to determine independently
values for either kA or kIS.

Single Internal Standard

In a single-point internal standardization, we prepare a single standard that contains the analyte and the internal standard, and use it
to determine the value of K in Equation 1.5.11.
CIS SA
K =( ) ×( ) (1.5.12)
CA SIS
std std

Having standardized the method, the analyte’s concentration is given by

CIS SA
CA = ×( )
K SIS
samp

 Example 1.5.7
A sixth spectrophotometric method for the quantitative analysis of Pb2+ in blood uses Cu2+ as an internal standard. A standard
that is 1.75 ppb Pb2+ and 2.25 ppb Cu2+ yields a ratio of (SA/SIS)std of 2.37. A sample of blood spiked with the same
concentration of Cu2+ gives a signal ratio, (SA/SIS)samp, of 1.80. What is the concentration of Pb2+ in the sample of blood?
Solution
Equation 1.5.12 allows us to calculate the value of K using the data for the standard
2 + 2 +
CIS SA 2.25 ppb Cu ppb Cu
K =( ) ×( ) = × 2.37 = 3.05
2 + 2 +
CA SIS 1.75 ppb Pb ppb Pb
std std

The concentration of Pb2+, therefore, is

2 +
CIS SA 2.25 ppb Cu 2 +
CA = ×( ) = × 1.80 = 1.33 ppb Pb
2 +
K SIS ppb Cu
samp
3.05 2 +
ppb Pb

Multiple Internal Standards

A single-point internal standardization has the same limitations as a single-point normal calibration. To construct an internal
standard calibration curve we prepare a series of standards, each of which contains the same concentration of internal standard and
a different concentrations of analyte. Under these conditions a calibration curve of (SA/SIS)std versus CA is linear with a slope of
K/CIS.

Although the usual practice is to prepare the standards so that each contains an identical amount of the internal standard, this is
not a requirement.

1.5.8 https://chem.libretexts.org/@go/page/332838
  Example 1.5.8
A seventh spectrophotometric method for the quantitative analysis of Pb2+ in blood gives a linear internal standards calibration
curve for which
SA
−1
( ) = (2.11  ppb × CA ) − 0.006
SIS
std

What is the ppb Pb2+ in a sample of blood if (SA/SIS)samp is 2.80?

Solution
To determine the concentration of Pb2+ in the sample of blood we replace (SA/SIS)std in the calibration equation with
(SA/SIS)samp and solve for CA.
SA
( ) + 0.006
SIS
samp 2.80 + 0.006
2 +
CA = = = 1.33 ppb Pb
−1 −1
2.11 ppb 2.11 ppb

The concentration of Pb2+ in the sample of blood is 1.33 ppb.

In some circumstances it is not possible to prepare the standards so that each contains the same concentration of internal standard.
This is the case, for example, when we prepare samples by mass instead of volume. We can still prepare a calibration curve,
however, by plotting (S /S ) versus CA/CIS, giving a linear calibration curve with a slope of K.
A IS std

You might wonder if it is possible to include an internal standard in the method of standard additions to correct for both matrix
effects and uncontrolled variations between samples; well, the answer is yes as described in the paper “Standard Dilution
Analysis,” the full reference for which is Jones, W. B.; Donati, G. L.; Calloway, C. P.; Jones, B. T. Anal. Chem. 2015, 87,
2321-2327.

This page titled 1.5: Calibration of Instrumental Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

1.5.9 https://chem.libretexts.org/@go/page/332838
 CHAPTER OVERVIEW
2: Electrical Components and Circuits
2.1: Basic Terminology and Laws of Electricity
2.2: Direct Current (DC) Circuits
2.3: Alternating Current Circuits
2.4: Semiconductors

This page titled 2: Electrical Components and Circuits is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1
2.1: Basic Terminology and Laws of Electricity
Current, I , is a movement of charge over time and is expressed in amperes, A , where 1 ampere is equivalent to 1 coulomb/sec. In
this section we review the convention used to describe currents in electrical circuits and review four laws of electricity.

Conventional Currents
If we connect one end of a wire to the positive terminal of a battery and connect the other end to the negative terminal of the same
battery, then electrons will move through the and a current will flow through the wire. The electrons move from the battery's
negative terminal through the wire to the battery's positive terminal. The direction of the current, however, runs from the battery's
positive terminal to the battery's negative terminal; that is, current is treated as if it is the movement of positive charge. This
probably strikes you as odd, but it simply reflects the original understanding of current from a time before the electron was
identified. Figure 2.1.1 shows the difference between these two ways of thinking about current.

Figure 2.1.1 . Two descriptions of current in an electrical circuit. The convention for currents is

Laws of Electricity
There are four basic laws of electricity that are important to us in this chapter: Ohm's law, Kirchhoff's laws, and the power law.
Let's take a brief look at each.

Ohm's Law
Ohm's law explains the relationship between current, I , measured in amps (A ), resistance, R , measured in ohms (Ω), and potential,
V , measured in volts (V ), and is written as

V = I ×R (2.1.1)

The voltage is measured between any two points in a circuit using a voltmeter.

Kirchhoff's Two Laws

The first of Kirchoff's two laws states that the sum of the currents at any point in a circuit must equal zero.

∑I = 0 (2.1.2)

The second law states that the sum of the voltages in a closed loop must equal zero.

∑V = 0 (2.1.3)

Power Law
When a current passes through a resistor, the temperature of the resistor increases and power (energy per unit time) is lost. The
amount of power lost, P , is the product of current and voltage, with units of joules/sec
P = I ×V (2.1.4)

or, substituting in Ohm's law (Equation 2.1.1), we can express power as

2
V
2
P =I ×R = (2.1.5)
R

2.1.1 https://chem.libretexts.org/@go/page/406376
  Note

An excellent resource for this section and other sections in this chapter is Principles of Electronic Instrumentation by A.
James Diefenderfer and published by W. B. Saunders Company, 1972.

This page titled 2.1: Basic Terminology and Laws of Electricity is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

2.1.2 https://chem.libretexts.org/@go/page/406376
2.2: Direct Current (DC) Circuits
A direct current, which is the focus of this section, is one that flows in one direction. An alternating current, which is the focus of
the next section, is one that periodically switches direction.

Basic Direct Current (DC) Circuits

There are two basic direct current circuits of importance to us: those with two or more resistors connected in a series, and those
with two or more resistors arranged parallel to each other. Other direct current circuits can be understood in terms of these two
basic circuits.

Resistors in Series
Figure 2.2.1 shows an example of a simple DC circuit in which three resistors, with resistances of R1, R2, and R3, are connected in
series to the two ends of battery that has a potential of V. A switch is included in the circuit that is used to close the loop and allow
a current to flow from the battery's positive terminal to its negative terminal.

Figure 2.2.1 . Example of a DC circuit that consists of three resistors in series.

Kirchoff's first law requires that the sum of the currents at any point in the circuit is zero. Consider point b. If the current that
arrives from point a is I , then the current that leaves b is −I , where the sign tells us about the direction of the current with respect
to the point. This requires that I − I = 0 , which means that the current has the same absolute value at all points in the circuit.
a c

Application of Kirchhoff's second law requires that the sum of the voltages in this circuit equal 0, which is true if the sum of the
voltage across each of the three resistors is equal to the voltage of the battery. The voltage across each resistor is given by Ohm's
law

V = I R1 + I R2 + I R3 (2.2.1)

If we divide both sides of Equation 2.2.1 by the current, then we have

V = I × (R1 + R2 + R3 ) = I Rs (2.2.2)

where R is the circuit's effective resistance, which is the sum of the resistances of the individual resistors.
s

One of the useful properties of this circuit is that the voltage drop across an individual resistor is proportional to that resistor's
contribution to R . Consider the points in Figure 2.2.1 labeled a and b that are on opposite sides of the first resistor in this series.
s

The drop in voltage across this resister, V , is

ab

Vab = I R1 (2.2.3)

Dividing Equation 2.2.3 by Equation 2.2.1

Vab I R1 R1
= = (2.2.4)
V I Rs Rs

and
R1
Vab = V × (2.2.5)
Rs

2.2.1 https://chem.libretexts.org/@go/page/402750
The circuit in Figure 2.2.1 is an example of a simple voltage divider in that it divides the battery's voltage into parts and allows us
to use a single battery to select one of several possible voltages. For example, the voltage at between points a and b is
R1
Vab = V × (2.2.6)
Rs

the voltage at between points a and c is

R1 + R2
Vac = V × (2.2.7)
Rs

and the voltage at between points a and d is

R1 + R2 + R3 Rs
Vad = V × =V × =V (2.2.8)
Rs Rs

Parallel Circuits
Figure 2.2.2 shows an example of a simple DC circuit in which three resistors, with resistances of R1, R2, and R3, are connected
parallel to each other. A switch is included in the circuit that is used to close the loop and allow a current to flow from the battery's
positive terminal to its negative terminal.

Figure 2.2.2 . Example of a DC circuit that consists of three resistors in parallel.

If we apply Kirchoff's first law to the current at the point identified as a, then the sum of the currents must equal zero and

∑ I = 0 = I − I1 − I2 − I3 (2.2.9)

where I is the current entering point a and I , I , and I are the currents passing through the three resistors. Rearranging Equation
1 2 3

2.2.9 and substituting into Ohm's law gives

V V V V
= + + (2.2.10)
Rp R1 R2 R3

where R is the circuit's effective resistance, which is equivalent to

p

1 1 1 1
= + + (2.2.11)
Rp R1 R2 R3

or to
Gp = G1 + G2 + G3 (2.2.12)

where G is a resistor's conductance, which is the inverse of its resistance.

One of the useful properties of this circuit is that it serves as a current divider. The current passing through the resistor R is 1

I1 V /R1 1/R1 G1
= = = (2.2.13)
I V /Pp 1/Rp Gp

Multiplying through by the total current gives

G1
I1 = I × (2.2.14)
Gp

2.2.2 https://chem.libretexts.org/@go/page/402750
More Complex Circuits
The treatment above considers circuits that contain only resistors in series or resistors in parallel. A circuit that contains both
resistors in series and resistors in parallel can often be simplified to an equivalent circuit that has only resistors in series or in
parallel, or that consists of single resistor. Figure 2.2.3 provides an example. The circuit at the far left shows two parallel resistors,
R and R , that, together, are in series with a third resistor, R .
2 3 1

Figure 2.2.3 . Example (on left) of a DC circuit that has both resistors in series and in parallel, and two equivalent circuits.
Using Equation 2.2.11 we can replace the two resistors in parallel with a single resistor, R , where 4

1 1 1
= + (2.2.15)
R4 R2 R3

giving the equivalent circuit shown in the middle. Finally, we can use Equation 2.2.2 to replace the two resistors in series with the
single resistor, R , as shown on the far right, where
5

R5 = R1 + R4 (2.2.16)

Measuring Voltage and Current

Figure 2.2.4 shows a digital multimeter that is used to measure voltage or current (amongst other possible measurements that we
will not consider here). The measurement of voltages and currents always contains some error, the magnitude of which we consider
here.

Figure 2.2.1 . Photos of (left) a digital multimeter and (right) a close-up of the digital multimeter's dial. Symbols: V is voltage, Ω is
resistance, F is capacitance, Hz is frequency, A is current, H is inductance. Symbols with a wavy line on top (~) are for AC circuits
and symbols with solid or dashed lines (— or ---) on top are for DC circuits.

Errors in Measuring Voltage

To measure an unknown voltage of V with an internal resistance of R , we include the meter with its internal resistance of R as
x x m

part of a voltage divider circuit. We read the voltage displayed on the meter, V , and use Equation 2.2.5 to determine V
m x

Rm
Vm = Vx × (2.2.17)
Rm + Rx

If we do not know the value of R , which is often the case, then we can still report an accurate value for V if R
x x m >> Rx , as we
can then write
Rm Rm
Vm = Vx × ≈ Vx × ≈ Vx (2.2.18)
Rm + Rx R,

2.2.3 https://chem.libretexts.org/@go/page/402750
The percent error, E , in V
x x

Vm − Vx Rm
Ex = × 100 = − × 100 (2.2.19)
Vx Rm + Rx

For example, suppose that R m = 10

3
× Rx , then the measurement error is
Rx 1
E =− × 100 = − × 100 = −0.0999% (2.2.20)
3 3
(10 × Rx ) + Rx 10 +1

or approximately –0.1%.

Errors in Measuring Current

To measure an unknown current, I , we include the meter in a current divider circuit in which some of I is drawn through a load
x x

resistor, R , of known value, and the remaining current is drawn through a known standard resistance set by the meter, R . Using
l m

Equation 2.2.13 for a current divider, the fraction of I that passes through the meter is
x

Im Rm + Rl
= (2.2.21)
Ix Rm

Solving for I gives

x

Rm Rm
Ix = Im × ( ) = Im × (1 + ) (2.2.22)
Rm + Rl Rl

Rm
If the resistors are selected such that Rl
<< 1 , then the current displayed on the meter, Im , is an accurate measure of Ix . The
percent error in the reported current is
Rm
Ex = − × 100 (2.2.23)
Rm + Rl

For example, suppose that R m = 10

−3
× Rl , then the measurement error is
−3 −3
10 × Rl 10
E =− × 100 = − × 100 = −0.0999% (2.2.24)
−3 −3
(10 × Rl ) + Rl 10 +1

or approximately −0.1%.

This page titled 2.2: Direct Current (DC) Circuits is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

2.2.4 https://chem.libretexts.org/@go/page/402750
2.3: Alternating Current Circuits
A direct current has a fixed value that is independent of time. An alternating current, on the other hand, has a value that changes
with time. This change in current follows a pattern that we can characterize by it period—the time, t , for one complete cycle—or
p

by its frequency, f , which is the reciprocal of its period

1
f = (2.3.1)
tp

Frequency is reported in hertz (Hz), which is equivalent to one cycle per second.

Sinusoidal Currents and Voltages

Although we can draw many periodic signals—and will do so in later chapters—the simplest periodic signal is a sine wave: as
shown on the right side of Figure 2.3.1, the sine is a propagating wave whose amplitude, A , is a function of time, t , which we
write as a(t) .

Figure 2.3.1 . Illustration showing how a sine wave can be explained by a rotating vector. See text for an explanation of the
symbols.
The left side of Figure 2.3.1 provides a rotating vector representation of the sine wave (a representation we will encounter again in
Chapter 19 on NMR spectroscopy). The vector is the arrow that extends from the center of the circle to the circle's edge. It is
rotating to the left with an angular velocity given by ω and that is expressed in radians per the sine wave's period, t ; thus
p

2π
ω = = 2πf (2.3.2)
tp

where f is the frequency. The amplitude of the sine wave as a function of time, a(t) , is equivalent to the projection of the rotating
vector onto the x-axis; thus

a(t) = A sin ωt = A sin 2πf t (2.3.3)

In the context of this chapter, the amplitude is either a current, i, or a voltage, v .

i(t) = I sin ωt = I sin 2πf t (2.3.4)

v(t) = V sin ωt = V sin 2πf t (2.3.5)

where I is the maximum, or peak current, and V is the maximum, or peak voltage.
Equations 2.3.4 and 2.3.5 require that a sine wave's time-dependent amplitude, a(t) , has a value of zero when t = nπ , where n is
an integer. There is no reason to insist on this and two sine waves can be separated from each other in time, as shown in Figure
2.3.2, by a phase angle, Φ . The equation for the sine wave when Φ ≠ 0 becomes

a(t) = A sin (ωt + Φ) = A sin (2πf t + Φ) (2.3.6)

2.3.1 https://chem.libretexts.org/@go/page/402753
 Figure 2.3.2 . Two sine waves that are out-of-phase with each other. The angle, Φ , between the two rotating vectors is the phase
angle.
One complication of an alternating current is that the net current over the course of a single cycle is zero. This is a problem for us
because the equation for power in a resistor is
2
I
P = ≠0 (2.3.7)
R

Figure 2.3.3 shows several ways to report current in AC circuits.

Figure 2.3.3 . Different ways to report current: Ip is the maximum absolute current, or peak current; Ipp is the peak-to-peak current;
Irms is the root-mean-square current; and Iavg is the numerical average current.
The root-mean-square current, I rms , is defined as
−−
−
2
Ip – Ip
Irms = √ = √2 × = 0.707 × Ip (2.3.8)
2 2

and yields the same power in an AC circuit as a direct current of equal value in a DC circuit. The average current, I avg , is
π
1 2Ip
Iavg = ∫ Ip sin ωt dt = = 0.6371 × Ip (2.3.9)
π 0
π

Capacitors
A capacitor is a component of circuits that is capable of storing charge. Figure 2.3.4 shows the design of a typical capacitor and its
symbol when constructing an electrical circuit. The capacitor consists of two conducting plates separated by a thin layer of an
insulating, or dielectric material. The plates have areas of A and are separated by a distance, d . The dielectric material has a
dielectric constant, ϵ. A simple capacitor might consist of two pieces of a metal foil separated by air, which serves as as the
dielectric material. A capacitor's ability to store charge, Q, is given by
Q = C ×V (2.3.10)

where V is the voltage applied across the two plates and where C is the capacitor's capacitance, which, in turn, is defined as
ϵA
C = (2.3.11)
d

Capacitance is measured in units of farads, where one farad is equal to one coulomb per volt.

2.3.2 https://chem.libretexts.org/@go/page/402753
 Figure 2.3.4 . Schematic diagram of a capacitor. In an electrical circuit, a capacitor is represented by the symbol at the lower right.

Resistor and Capacitor in Series

Figure 2.3.5 shows a resistor, with a resistance of R , and a capacitor, with a capacitance of C , in series with a voltage source, with
a voltage of V .

Figure 2.3.5 . A simple circuit that consists of a resistor and capacitor in series.
When the switch it closed, current flows as the capacitor builds up a charge. From Kirchoff's laws, we know that
Q
V = vR + vC = iR + (2.3.12)
C

where v and v are, respectively, the time-dependent voltages across the resistor and the capacitor. Because V has a fixed value,
R C

any increase in v as the capacitor is charged is offset by a decrease in V . Given that the values of v and v —and the associated
C r C R

currents—are time-dependent, we can differentiate Equation 2.3.12 with respect to time

dV di 1 dq di i
= 0 = (R × ) +( × ) = (R × )+ (2.3.13)
dt dt C dt dt C

Rearranging Equation 2.3.13 gives

di 1
=− dt (2.3.14)
i RC

Integrating both sides of this equation

i t
1 1
∫ di = − ∫ dt (2.3.15)
I0
i RC 0

leads to the following relationship between the current at time t and the initial current, I 0

−t/RC
it = I0 × e (2.3.16)

which tells us that the current decreases exponentially as the capacitor becomes fully charged. Replacing the current in equation
2.3.16 with and substituting back into Equation 2.3.12
V

−t/RC
vC = V0 (1 − e ) (2.3.17)

shows us that during the time the capacitor is being charged, the current flowing through the capacitor is decreasing exponentially
to its limit of zero, and the voltage across the capacitor is increasing exponentially to its limit of the applied voltage.

2.3.3 https://chem.libretexts.org/@go/page/402753
Time Constant
The value RC in Equation 2.3.16 and in Equation 2.3.17 is the circuit's time constant. It takes approximately five time constants
for the capacitor to fully charge or fully discharge. Figure 2.3.6 shows the voltage across the capacitor, v , as it is allowed to
C

charge and to discharge. Time is shown in terms of the number of elapsed time constants, and voltage is expressed as a fraction of
the maximum voltage. The dashed line shows that the time constant, RC , is equivalent to 0.63× the maximum voltage.

Figure 2.3.6 . Voltage across a capacitor as it is charged and then discharged. The scale on the x-axis is time in increments of the
time constant, RC , and the scale on the y-axis is the fraction of the applied voltage. The dashed line shows that the time constant is
equivalent to 0.63× the applied voltage.

Response of a Series RC Circuit to a Sinusoidal Input

If we replace the DC voltage source in Figure 2.3.5 with an AC source, then the capacitor will undergo a continuous fluctuation in
its voltage and current as a function of time. We know, form Equation 2.3.10 that charge, Q, is the product of capacitance, C , and
voltage,V , which we can write as a derivative with respect to time.
dq dv
=C × (2.3.18)
dt dt

Phase Shift in an AC Circuit

In an AC circuit, as we learned earlier in Equation 2.3.4, the current, which is equivalent to dq/dt is
i = Ip sin 2πf t (2.3.19)

where I is the peak current. Substituting into Equation 2.3.18 gives

p

dv
i = Ip sin 2πf t = C × (2.3.20)
dt

Rearranging this equation and integrating over time gives the time-dependent voltage across the capacitor, v , as C

Ip
vC = (− cos 2πf t) (2.3.21)
2πf C

We can rewrite this equation in terms of a sine function instead of a cosine function by recognizing that the two are 90° out of
phase with each other; thus
Ip
vC = (sin 2πf t − 90) = Vp (sin 2πf t − 90) (2.3.22)
2πf C

Comparing Equation 2.3.19 and Equation 2.3.22 , we see that the current and the voltage are 90° out-of-phase with each other;
Figure 2.3.7 shows this visually.

2.3.4 https://chem.libretexts.org/@go/page/402753
 Figure 2.3.7 . Response of a capacitor's voltage to a sinusoidal input current showing that they are 90° (π/2) out-of-phase.

Capacitive Reactance, Resistance, and Impedence

From Equation 2.3.22 we see that
Ip
Vp = (2.3.23)
2πf t

Dividing both sides by I gives

p

Vp 1
= XC = (2.3.24)
Ip 2πf t

where X is the capacitor's reactance, which, like a resistor's resistance, has units of ohms. Unlike a resistor, however, a capacitor's
C

reactance is frequency dependent and, given the reciprocal relationship between X and f , it becomes smaller at higher
C

frequencies.
In a RC circuit, both the resistor and the capacitor contribute to the circuit's impedence of the alternating current. Because the
contribution of the capacitor is 90° out-of-phase to the contribution from the resistor, the net impedence, Z , is
−−−−−−−
2 2
Z = √R +X (2.3.25)
C

as shown in Figure 2.3.8 where the vector that represents Z is the hypotonus of a right triangle defined by the resistor's resistance
and the capacitor's reactance.

Figure 2.3.8 : The relationship between a resistor's resistance, a capacitor's reactance, and the impedence of the RC circuit.
Substituting in Equation 2.3.24 shows the effect of frequency on impedence.
−−−−−−−−−−−−
2
1
2
Z = √R +( ) (2.3.26)
2πf t

Writing Ohm's law in terms of impedence, Vp = Ip × Z , and substituting it into Equation , defines
2.3.26 Ip and Vp in terms of
impedence.
−−−−−−−−−−−−
2
1
2
Vp = Ip × √R +( ) (2.3.27)
2πf t

Vp
Ip = −−−−−−−−−−− (2.3.28)
2
2 1
√R +( )
2πf t

2.3.5 https://chem.libretexts.org/@go/page/402753
Filters Based on RC Circuits
The frequency dependence of an RC circuit provides us with the ability to attenuate some frequencies and to pass other
frequencies. This allows for the selective filtering of an input signal. Here we consider the design of a low-pass filter that removes
higher frequency signals, and the design of a high-pass filter that removes lower frequency signals. Figure 2.3.9 shows that (a) a
low-pass filter places the resistor before the capacitor and measures the output voltage, V , across the capacitor, and that (b) a
out

high-pass filter places the capacitor before the resistor and measures the output voltage, V , across the resistor.
out

Figure 2.3.9 . Circuits for (a) a low-pass filter and (b) a high-pass filter.

Low-Pass Filter
For the low-pass filter in Figure 2.3.9a, the ratio of the voltage across the capacitor, (V ) , to the peak input voltage, (V ) , is
p out p in

equal to the fraction of the circuit's impedence, Z , attributed to the capacitor's reactance, X , as expected for a voltage divider that
C

consist of elements in series.

−1
(Vp )out XC (2πf C )
= = −−−−−−−−−−− (2.3.29)
(Vp )in Z 2
2 1
√R +( )
2πf C

Figure 2.3.10a shows the frequency response for a low-pass filter with a 6
1 × 10  Hz resistor and a −6
1 × 10  F capacitor,
removing all frequencies greater than approximately 10 Hz. 1

Figure 2.3.10 . The output of (a) a low-pass filter and (b) a high-pass filter.

High-Pass Filter
For the high-pass filter in Figure 2.3.9b, the ratio of the voltage across the resistor, (V ) , to the peak input voltage, (V ) , is
p out p in

equal to the fraction of the circuit's impedence, Z , attributed to the resistor's resistance, R , as expected for a voltage divider that
consist of elements in series.
(Vp )out R R
= = −−−−−−−−−−− (2.3.30)
(Vp )in Z 2
2 1
√R +( )
2πf C

Figure 2.3.10b shows the frequency response for a low-pass filter with a 5
1 × 10  Hz resistor and a −7
1 × 10  F capacitor,
removing all frequencies less than approximately 10 . −1

This page titled 2.3: Alternating Current Circuits is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

2.3.6 https://chem.libretexts.org/@go/page/402753
2.4: Semiconductors
A semiconductor is a material whose resistivity to the movement of charge falls somewhere between that of a conductor, through
which we can move a charge easily, and an insulator, which resists the movement of charge. Some semiconductors are elemental,
such as silicon and germanium (both of which we examine more closely in this section) and some are multielemental, such as
silicon carbide.

Properties of Silicon and Germanium Semiconductors

A conductor, such as aluminum or copper, has a resistivity on the order of 10 − 10  Ω ⋅ m , which means that its resistance to
−8 −6

the movement of electrons is sufficiently small that it carries a current without much effort. An insulator, such as the mineral
quartz, SiO , has a resistivity on the order of 10 − 10  Ω ⋅ m . Silicon, on the other hand, has a resistivity of approximately
2
15 19

640 Ω ⋅ m and germainum has a resistivity of about 0.46 Ω ⋅ m.

 Note
The inverse of resistivity is conductivity.

Silicon and germanium are in the same group as carbon. If we use a simplified view of atoms, we can treat silicon and germanium
as having four valence electrons and an effective nuclear charge, Z , of ef f

(Zef f )Si = Z −  number of core electrons

= 14 − 10

= +4

(Zef f )Ge = Z −  number of core electrons

= 32 − 28

= +4

We can increase the conductivity of silicon and germanium by adding to them—this is called doping—a small amount of an
impurity. Adding a small amount of In or Ga, which have three valence electrons instead of four valence electrons, leaves a small
number of vacancies, or holes, in which an electron is missing. Adding a small amount of As or Sb, which have five valence
electrons instead of four valence electrons, leaves a small number of extra electrons. Figure 2.4.1 shows all three possibilities.

Figure 2.4.1 . Illustration showing a two-dimensional cross-section through (left) a single crystal of silicon or germanium; (middle)
a single crystal of silicon or germanium doped with indium or gallium; and (right) a single crystal of silicon or germanium doped
with arsenic or antimony. Doping with indium or gallium leaves a hole where an electron is missing, and doping with arsenic or
antimony leaves an extra electron. Patterned after figures in Diefenderfer, A. J. Principles of Electronic Instrumentation, W. B.
Saunders (1972).
If we apply a potential across the semiconductor doped with As or Sb, the extra electron moves toward the positive pole, creating a
small current, and leaving behind a vacancy, or hole. If we apply a potential across the semiconductor doped with In or Ga,
electrons enter the semiconductor from the negative pole, occupying the vacancies, or holes, and creating a small current. In both
cases, electrons move toward the positive pole and holes move toward the negative pole. We call an As or Sb doped semiconductor

2.4.1 https://chem.libretexts.org/@go/page/402755
an n-type semiconductor because the primary carrier of charge is an electron; we call an In or Ga doped semiconductor an p-type
semiconductor because the primary carrier of charge is the hole.

Semiconductor Diodes
A diode is an electrical device that is more conductive in one direction than in the opposite direction. A diode takes advantage of
the properties of the junction between a p-type and an n-type semiconductor.

Properties of pn Junctions
Let's use Figure 2.4.2 to make sense of how a semiconductor diode works. The figure is divided into two parts: the left side of the
figure, parts (a), (b), and (c), show the behavior of the semiconductor diode when a foward bias is applied, and the right side of the
figure, parts (d), (e), and (f), show its behavior when a reverse bias is applied. For both, the semiconductor diode consists of a
junction between a n-type semiconductor, which has an excess of electrons and carries a negative charge, and a p-type
semiconductor, which has an excess of holes and, thus, a positive charge; this is shown in (a) and (d). How the semiconductor is
manufactured is not of important to us.

Figure 2.4.2 . The behavior of a semiconductor diode under forward bias and reverse bias, and symbol for a semiconductor diode.
See the text for additional details.
To effect a forward bias, we apply a positive potential to the p-type region and apply a negative potential to the n-type region. As
we see in (b), the holes in the p-region move toward the junction and the electrons in the n-region move toward the junction as
well. When holes and electrons meet they combine and are eliminated, which is why we see fewer holes and electrons in (c).
Additional electrons flow into the n-region and electrons are pulled away from the p-region, as seen in (c), resulting in a current. To
effect a reverse bias, we switch the applied potentials so that the p-region has the negative potential and the n-region has the
positive potential. The result, as seen in (e) is a brief current as the holes and electrons move away from each other. leaving behind,
in (f), a depletion zone that has essentially no electrons or holes.

Current-Voltage Curves for Semiconductor Diodes

Figure 2.4.3 shows a plot of current as a function of voltage for a semiconductor diode. In forward bias mode the current increases
exponentially with an increase in applied voltage, but remains at essentially zero when operated under a reverse bias. The use of a
sufficiently large negative potential, however, does result in an sudden and dramatic increase in current; the potential at which this
happens is called the breakdown voltage.

2.4.2 https://chem.libretexts.org/@go/page/402755
 Figure 2.4.3 . Current-voltage curve for a semiconductor diode under forward bias and under reverse bias.

This page titled 2.4: Semiconductors is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

2.4.3 https://chem.libretexts.org/@go/page/402755
 CHAPTER OVERVIEW
3: Operational Amplifiers
3.1: Operational Amplifiers
3.2: Operational Amplifier Circuits
3.3: Amplification and Measurement of Signals
3.4: Mathematical Operations Using Operational Amplifiers

This page titled 3: Operational Amplifiers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1
3.1: Operational Amplifiers
An operational amplifier (or op amp, for short) is an electrical circuit that has a variety of uses, a few of which we consider in this
section: how to amplify and measure the signal from a transducer (detector), and how to perform mathematical operations on
signals. In this section we will provide a basic overview of operational amplifiers without worrying about the specific internal
details of its electrical circuit.

 Note
An excellent resource for this section and other sections in this chapter is Principles of Electronic Instrumentation by A.
James Diefenderfer and published by W. B. Saunders Company, 1972.

Symbolic Representation of an Operational Amplifier

Figure 3.1.1 provides a symbolic representation of an operational amplifier. The large triangular shape is the operational amplifier,
which is an extensive circuit whose exact design is not of interest to us; thus, the simple shape. The operational amplifier has two
voltage inputs that are identified as v and as v and labeled as − and + on the op amp. The difference between v and v is
− + − +

defined as v . The operational amplifier also has a single voltage output that is identified as v . All voltages are measured relative
s out

to a circuit common of 0 V, represented by the small triangle at the bottom of the figure, that provides a shared reference; the circuit
common is understood to be present even when it is not shown. Not included in this figure are the connections to a power supply,
which are necessary for its operation.

Figure 3.1.1 . Representation of an operational amplifier. See the text for details.

Inverting and Noninverting Inputs

The minus sign and the plus sign that appear as labels on the op amp in Figure 3.1.1 do not mean that one input has a positive value
and that the other input has a negative value. Instead, an input to the lead with a negative sign is inverted: if v is a negative DC
−

voltage, then the output voltage, v , is a positive DC voltage, and if v is a positive DC voltage, then the output voltage, v , is a
out − out

negative DC voltage. For an AC input to v , the output is 180° out-of-phase, which implies a reversal in sign. The other input to
−

the op amp is noninverting, which means that applying a positive voltage to v results in a positive signal at v .
+ out

Key Properties of Operational Amplifiers

The ideal operational amplifier has several important properties that derive from its internal circuitry. The first of these properties is
that the op amp's gain, A , which is defined as the ratio of the output voltage to the input voltage
vout vout
Aop = − =− (3.1.1)
vs v− − v+

is very large, typically on the order of 104 – 106. We need to be careful when we use the term gain as there can be a significant
difference between the gain of the operational amplifier and the gain of the circuit that contains the operational amplifier. The gain
of the operational amplifier, which is what we mean by Equation 3.1.1, is called the open-loop gain. The gain of a circuit that
contains an operational amplifier is called a closed-loop gain. Where there is ambiguity, we will be careful to refer to the op amp's
gain, A , or to the circuit's gain, A , as these are more descriptive.
op c

A second property of an operational amplifier is that regardless of the specific values of v and v , the op amp's internal circuitry
− +

is designed such that the current between the two inputs is effectively zero; in essence, the impedence, Z , between the two inputs is
so large that from Ohm's law, V = I × Z , the current between these two inputs is I ≈ 0 . A large input impedence means we can

3.1.1 https://chem.libretexts.org/@go/page/406437
connect our op amp to a high voltage source and know that it will draw a small current instead of overloading the circuit that
includes the op amp.
A third property of an operational amplifier is that its output impedence is very small, which means we can draw a current from the
circuit that meets our needs—this current is drawn from the op amp's power supply—even if the current into the op amp is zero.
For example, if the circuit's gain is small, we can use the operational amplifier to provide a large gain in current.

This page titled 3.1: Operational Amplifiers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

3.1.2 https://chem.libretexts.org/@go/page/406437
3.2: Operational Amplifier Circuits
In the last section we noted that an operational amplifier magnifies the difference between two voltage inputs
vout
Aop = − (3.2.1)
v− − v+

where the gain, A , is typically between 104 and 106. To better control the gain—that is, to make the gain something we can adjust
op

to meet our needs—the operational amplifier is incorporated into a circuit that allows for feedback between the output and the
inputs. In this section, we examine two feedback circuits.

The Inverting Amplifier Circuit

Figure 3.2.1 is an example of an operational amplifier circuit with a negative feedback loop that consists of a resistor, R , that f

connects the op amp's output to its input at a summing point, S . Because the feedback loop is connected to the op amp's inverting
input, the effect is called negative feedback.

Figure 3.2.1 . Operational amplifier circuit with a negative feedback loop.

We can analyze this circuit using the laws of electricity from Chapter 2. Let's begin by rearranging Equation 3.2.1 to solve for v out

vout = −Aop × (v− − v+ ) (3.2.2)

and then expand the right side of this equation

vout = −Aop × (v− − v+ ) = −Aop × v− + Aop × v+ (3.2.3)

and solve for v−

vout
v− = v+ − (3.2.4)
Aop

Because the op amp's gain A op is so large—recall that it is typically in the range 104 and 106—we can simplify Equation 3.2.4 to
v− ≈ v+ (3.2.5)

One consequence of Equation 3.2.5 is that for this circuit v − ≈ 0 V as it is at the circuit common.
From Kirchoff's laws, we know that the total current that enters the summing point must equal the total current that leaves the
summing point, or
Iin = Is + If (3.2.6)

where I is the current between the op amp's two inputs. As we noted in Chapter 3.1, an operational amplifier's internal circuitry is
s

designed such at I ≈ 0 ; thus

s

Iin = If (3.2.7)

Substituting in Ohm's law (V = I ×R ) gives

vin − v− v− − vout
= (3.2.8)
Rin Rf

From Equation 3.2.5, we know that v − ≈0 , which allows us to simplify Equation 3.2.8 to

3.2.1 https://chem.libretexts.org/@go/page/406439
 vin vout
=− (3.2.9)
Rin Rf

Rearranging, we find that the gain for the circuit, A , is

c

vout Rf
Ac = =− (3.2.10)
vin Rin

Equation 3.2.10 shows us that circuit in Figure 3.2.1 returns a voltage, v , that has the opposite sign of
out vin with a gain for the
circuit that depends on only the relative values of the two resistors, R and R .f in

The Voltage-Follower Circuit

Figure 3.2.2 shows another operational amplifier with a feedback loop. In this case the input to the op amp, vin , is made to the
noninverting lead and the output is feed back into the op amp's inverting lead.

Figure 3.2.2 . Operational amplifier circuit for a voltage follower.

From Kirchoff's voltage law, we know that the op amp's output voltage is equal to the sum of the input voltage and the difference,
v between the voltage applied to the op amp's two leads; thus
s

Vout = vin + vs (3.2.11)

The op amp's gain, A , is defined in terms of v and v

op s out

vout
−Aop = (3.2.12)
vs

where the minus sign is due to the change in sign between the output voltage and the voltage applied to the inverting lead.
Substituting Equation 3.2.12 into Equation 3.2.11 gives
Vout
Vin − = Vout (3.2.13)
Aop

Because the operational amplifier's gain—which is not the same thing as the circuit's gain—is large, Equation 3.2.13 becomes

vin = vout (3.2.14)

Our analysis of this circuit shows that it returns the original voltage without any gain. It does, however, allow us to draw that
voltage from the circuit with more current than the original voltage source might be able to handle.

This page titled 3.2: Operational Amplifier Circuits is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

3.2.2 https://chem.libretexts.org/@go/page/406439
3.3: Amplification and Measurement of Signals
In Chapter 1.3 we identified the basic components of an instrument as a probe that interacts with the sample, an input transducer
that converts the sample's chemical and/or physical properties into an electrical signal, a signal processor that converts the electrical
signal into a form that an output transducer can convert into a numerical or a visual output that we can understand. We can
represent this as a sequence of actions that take place within the instrument

probe → sample → input transducer → raw data → signal processor → output transducer

creating the following general flow of information

chemical and/or physical information → electrical information → numerical or visual response

As suggested above, information is encoded in two broad ways: as electrical information (such as currents and potentials) and as
information in other, non-electrical forms (such as chemical and physical properties). In this section we will consider how we can
measure electrical signals.

Current Measurements
In Chapter 7 we will introduce the phototube, seen here in Figure 3.3.1, as a transducer for converting photons of light into an
electrical current that we can measure. A photon of light strikes a photoemissive cathode and ejects an electron that is then drawn
to an anode that is held a positive potential. The resulting current is our analytical signal. As each photon generates a single
electron, the resulting current is small and needs amplifying if it is to be useful to us. Operational amplifiers provide a way to
accomplish this amplification.

Figure 3.3.1 . Schematic illustration of a phototube. See Chapter 7 for additional details.
Figure 3.3.2 shows a simple electrical circuit that we can use to amplify and measure a small current. If you compare this to the
inverting amplifier circuit in Chapter 3.2, you will see that we are replacing the input voltage and input resistor with the input
current, I , that we wish to measure.
x

Figure 3.3.2 . Operational amplifier circuit for converting a current into a voltage.
From Kirchoff's current law, we know that at the summing point, S , the current from the transducer is equal to the sum of the
current through the feedback loop, I , and the current to the op amp's inverting input, I .
f s

Ix = Is + If (3.3.1)

As we learned in the last section, I s ≈0 , which means I x ≈ If . From Ohm's law, we have
Vout = −If × Rf = −Ix × Rf (3.3.2)

Rearranging to solve for I x

3.3.1 https://chem.libretexts.org/@go/page/406440
 Vout
Ix = − = kVout (3.3.3)
Rf

shows us that there is a linear relationship between the voltage we measure from the circuit's output and the current that enters the
circuit. By choosing to make R large, a small current is converted into a voltage that is easy to measure. In addition, we know
f

from Chapter 2 that the error in measuring the current from the transducer, E , is x

Rm
Ex = − × 100 (3.3.4)
Rm + Rl

where R is the resistance of the measuring circuit and R is the resistance of the source, which generally is large. The resistance
m l

of the measuring circuit is

Rr
Rm = (3.3.5)
Aop

If we choose R such that it is similar in magnitude to the op amp's gain, A , then R

f op m is small and the relative error is small as
well.

Potential Measurements
From Chapter 2 we know that the error in measuring voltage, E , is a function of the resistance of the measuring circuit, R , and
x m

the resistance of the source, R . x

Vm − Vx Rm
Ex = × 100 = − × 100 (3.3.6)
Vx Rm + Rx

To maintain a small measurement error requires that R << R . This creates a complication when the voltage source has a high
x m

internal resistance, as is the case, for example, when we measure pH using a glass electrode where the internal resistance is on the
order of 10 − 10 Ω (see Chapter 23 for details about glass electrodes). The inverting amplifier circuit discussed in Chapter 3.2
7 8

has a resistance of perhaps 10 Ω. To increase R , the voltage we wish to measure, V , is first run through a voltage follower
5
m x

circuit, where the internal resistance is on the order of 10 Ω, and the output is then run through the inverting amplifier, as seen in
12

Figure 3.3.3. The result is an amplified output voltage measured under conditions where the relative error is small.

Figure 3.3.3 . Circuit for measuring voltage that combines a voltage follower and an inverting amplifier.

Comparison of Transducer Outputs

In Chapter 13 we will cover molecular absorption spectroscopy in which we measure the absorbance of sample relative to the
absorbance of a reference. A difference amplifier, such as that shown in Figure 3.3.4, allows us to amplify and measure the
difference between two voltages. In this circuit, the two voltages, v and v , are fed into the op amp's two inputs, v and v ,
1 2 − +

passing through identical resistors. A feedback loop with a resistor connects v to v and a resistor identical to that in the
1 out

feedback loop connects v to the circuit common.

2

3.3.2 https://chem.libretexts.org/@go/page/406440
 Figure 3.3.4 . Circuit used to amplify and measure the difference between two voltages.
We can use Ohm's law to define the currents I and I as
1 f

v1 − v−
I1 = (3.3.7)
Ri

v− − vout
If = (3.3.8)
Rf

By now you should see that the currents I and I are approximately the same because the op amp's high internal impedence
1 f

prevents current from flowing into the op amp. Combining Equation 3.3.7 and Equation 3.3.8 gives
v1 − v− v− − vout
= (3.3.9)
Ri Rf

which we can solve for the voltage at the op amp's inverting input

Rf v1 − Rf v− = Ri v− − Ri Vout (3.3.10)

Ri v− + Rf v− = Rf v1 + Ri vout (3.3.11)

Rf v1 + Ri vout
v− = (3.3.12)
Ri + Rf

The input to the op amp's noniverting lead is the output of a voltage divider (see Chapter 2) acting on v 2

Rf
v+ = v2 × ( ) (3.3.13)
Ri + Rf

The feedback loop works to ensure that the inputs to v and to v are identical; thus
− +

Rf v1 + Ri vout Rf
= v2 × ( ) (3.3.14)
Ri + Rf Ri + Rf

which we can simply to

V1 Rf + Vout Ri = V2 Rf (3.3.15)

Rf
Vout = × (V2 − V1 ) (3.3.16)
Ri

The output voltage from the circuit is equal to the difference between the two input voltages, but amplified by the ratio of the
resistance of R to R .
f i

This page titled 3.3: Amplification and Measurement of Signals is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

3.3.3 https://chem.libretexts.org/@go/page/406440
3.4: Mathematical Operations Using Operational Amplifiers
The circuit for comparing two voltages is an example of using an operational amplifier to complete a mathematical operation. In
this section we will examine several additional examples of mathematical operations completed using operational amplifiers.

Multiplication and Division by a Constant

The inverting amplifier that we considered earlier, and that is reproduced here in Figure 3.4.1, returns an output voltage, v out that
multiplies the input voltage by an amount that depends on the ratio of the resistors R and R . in f

Rf
vout = −vin × (3.4.1)
Rin

Multiplication takes place when R f > Rin and division takes place when R in > Rf . Note that there is a reversal in the sign of the
voltage.

Figure 3.4.1 . Operational amplifier circuit that multiples or divides an input voltage by a constant.

Addition or Subtraction
Figure 3.4.2 shows an operational amplifier circuit that adds together four separate input voltages. From our earlier analysis of
circuits, you should see that

If = I1 + I2 + I3 + I4 (3.4.2)

We can replace I in this equation using Ohm's law; thus

f

v1 v2 v3 v4
vout = −Rf × ( + + + ) (3.4.3)
R1 R2 R3 R4

If all five of the resistors are identical, then v out is a simple summation of the four input voltages.
vout = −(v1 + v2 + v3 + v4 ) (3.4.4)

If we choose Rf such that it is 0.25 × R1 and set R1 = R2 = R3 = R4 , then the output voltage is the average of the input
voltages
v1 + v2 + v3 + v4
vout = − (3.4.5)
4

Figure 3.4.2 . Operational amplifier circuit that adds together four voltages.

3.4.1 https://chem.libretexts.org/@go/page/406445
The voltage comparator covered in the last section subtracts one voltage from another. When more than two voltages are involved,
then we can adapt the voltage adder circuit in Figure 3.4.2 to include subtraction by first running the voltage we wish to subtract
through the inverting amplifier introduced in Chapter 3.2. Figure 3.4.3 shows this where v = −(v + v + v − v ) . out 4 3 2 1

Figure 3.4.3 . Operational amplifier circuit that adds together three voltages and subtracts a fourth voltage from the sum.

Integration
Figure 3.4.4 shows an operational amplifier circuit that we can use to integrate a time-dependent signal. The circuit has a feedback
loop, but it is built around a capacitor instead of a resistor because it stores charge over time. The circuit also has two switches that
allow us to use the circuit over a specific period of time. When the hold switch is open, the input voltage cannot enter the circuit.
Closing the hold switch sets t = t . As long as the reset switch is open, current moves through the feedback loop. Opening the hold
0

switch sets t = t , where t is the total elapsed time. When this cycle is over, closing the reset switch drains the capacitor so that it
f f

is ready for its next use.

Figure 3.4.4 . Operational amplifier circuit that integrates the input signal.
As we have seen several times, the current into the summing point, I in is equal to the current in the feedback loop.
Iin = If (3.4.6)

dvout Vin
From Chapter 2, we know that the current in the feedback loop is If = −Cf
dt
and, we know from Ohm's law that iin =
Rin
.
Substituting both relationships into Equation 3.4.6 gives
Vin dvout
= −Cf (3.4.7)
Rin dt

Rearranging this equation

vin
dvout = − dt (3.4.8)
Ri Cf

and integrating over time gives

vout,2 t2
1
∫ dvout = − ∫ vin dt (3.4.9)
vout,1
Rin Cf t1

If we begin the integration having previously discharged the capacitor and define t as the the moment we close the hold switch
1

and define t as the moment we reopen the hold switch, then Equation 3.4.9 becomes
2

3.4.2 https://chem.libretexts.org/@go/page/406445
 t
1
vout = − ∫ vin dt (3.4.10)
Rin Cf 0

and the output voltage is the integral of the input voltage multiplied by (−R in Cf
−1
) .

Differentiation
Reversing the capacitor and the resistor in the circuit in Figure 3.4.4 coverts the circuit from one that returns the integral of the
input signal, into one that returns the derivative of the input signal; Figure 3.4.5 shows the resulting circuit.

Figure 3.4.5 . Operational amplifier circuit that returns the derivative of the input signal.
dvin Vout
For this circuit we have I
in =C ×
dt
and I f =−
R
. Given that I in = If , we are left with
Vout dvin
− =C × (3.4.11)
R dt

dvin
vout = −RC × (3.4.12)
dt

This page titled 3.4: Mathematical Operations Using Operational Amplifiers is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

3.4.3 https://chem.libretexts.org/@go/page/406445
 CHAPTER OVERVIEW
4: Analog and Digital Data
4.1: Analog and Digital Data
4.2: Working With Binary Numbers
4.3: Cleaning Up Signals and Counting Events

This page titled 4: Analog and Digital Data is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1
4.1: Analog and Digital Data
Figure 4.1.1 shows an xy-recorder that we can use to provide a permanent record of a cyclic voltammetry experiment. In this
particular experiment, we apply a variable potential to an electrochemical cell and measure the current that flows in response to
this potential (see Chapter 25 for a discussion of cyclic voltammetry). The potential and the current, which is converted into a
voltage for the purpose of recording the cyclic voltammogram, are fed into the recorder using the cables on the right side of the
recorder. The Y1 Range and the X Range controls allow us to adjust the scales of the axes.

Figure 4.1.1 . Photograph (a) of an xy-recorder and (b) close-up view of the recorder's controls.
The vertical bar on the xy-recorder moves toward the recorder's left or right based on the applied potential, and a pen attached to
the vertical bar moves toward the recorder's top or bottom based on the measured current. The applied potential and the current are
continuous variables within the instrument's range; the resulting cyclic voltammogram in Figure 4.1.2 is an analog record of the
experiment.

Figure 4.1.2 . Example of the type of output obtained with the xy-recorder for a cyclic voltammetry experiment.
Although the analog trace in Figure 4.1.2 provides a permanent record of an experiment, it is not in form that gives us access to the
raw data. We can take the image and use digitizing software (see here for an open-source digitizer) to extract a digital version of the
data, or we can design our instruments to collect the data in digital form by sampling the analog signal at preset intervals and then
saving the data. Such files often are in a format that includes metadata that explains how to extract the data from the file. For
example, xy-coordinate data for a wide variety of spectroscopy experiments is often stored digitally using a format established by
the Joint Committee on Atomic and Molecular Physical Data (JCAMP). Such files have the extension .jdx and can be opened using
a variety of different software programs.
Figure 4.1.3 is a screenshot that illustrates how we can work with digitized data using data analysis software, such as R and
RStudio. The upper left panel shows some of the contents of a .jdx file that contains the IR spectrum of methanol (in this case,
digitized by NIST from an analog hard copy). The lines preceded by double hashtags (##) are metadata that provide information
about the x-axis scale (minimum and maximum limits and increments between values), the y-axis scale (minimum and maximum
values), and the number of data points. This is followed by multiple lines of digitized data. Each line of data contains one value of x

4.1.1 https://chem.libretexts.org/@go/page/407087
and five values of y. The R package readJDX was used to extract the information from the .jdx file and to store it in a variable
given the name methanol (see upper right panel). Code written in R (see lower left panel) was used to plot (see lower right panel)
the spectrum.

Figure 4.1.3 : Screenshot from RStudio showing the analysis of a digitized spectrum. See text for details.
Although the spectrum for methanol in Figure 4.1.4—with its smooth, continuous line—looks like an analog spectrum, this is a
result of choosing to plot the data as a sequence of lines that connect individual points without actually displaying the individual
points themselves. Figure 4.1.4, in which we plot only the individual data points, shows us that the spectrum actually consists of
discrete, digitized data.

Figure 4.1.4 . The IR spectrum for methanol from the .jdx data in Figure 4.1.2 plotted here using individual points instead of a line
that connects the individual points.

This page titled 4.1: Analog and Digital Data is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

4.1.2 https://chem.libretexts.org/@go/page/407087
4.2: Working With Binary Numbers
In Chapter 7 we will examine several transducers for counting photons. The transducers are made of an array—some use a linear
array and some use a two-dimensional array—of individual detecting units. We will worry about the details of how these
transducers work in Chapter 7, but if you take a quick look at Figure 7.5.4 – 7.5.6 you will see that the number of individual
detecting units are interesting: a linear array of 1024 individual units; another linear array, but with 2048 units, and a two-
dimensional array that has 1024 × 1024 = 1, 048, 576 individual units. What is interesting about these numbers is that each is a
power of two: 1024 = 2 , 2048 = 2 , and 1, 048, 576 = 10 .
10 11 20

Humans are comfortable working with numbers expressed using a decimal notation that relies on 10 unique digits (0, 1, 2, 3, 4, 5,
6, 7, 8, and 9), but computers work with information using a binary notation that is limited to just two unique digits (0 and 1).
Although we will not complete calculations using binary numbers, you will see examples of instrumental methods, such as FT-
NMR, where the data analysis algorithms (the Fourier transform in this case) require that the number of data points be a power of
two. It is useful, therefore, to be familiar with how we represent numbers in both decimal and binary form.

Decimal Representation of Numbers

My university was founded in 1837, which is a decimal expression of the year. Each of these four digits represents a power of 10, a
fact that is clear when we read the number out loud: one thousand—eight hundred— thirty—seven, or, when we write it out this
way

(1 × 1000) + (8 × 100) + (3 × 10) + (7 × 1) = 1837

or this way
3 2 1 0
(1 × 10 ) + (8 × 10 ) + (3 × 10 ) + (7 × 10 ) = 1837

We refer to the 7 being in the ones place (10 = 1 ), the 3 in the tens place (10 = 10 ), the 8 in the hundreds place (10 = 100),
0 1 2

and the 1 in the thousandths place (10 = 1000). Figure 4.2.1a shows these three ways of representing a number using a decimal
3

notation.

Figure 4.2.1 . Three notations used to represent a number using (a) decimal digits and powers of 10 and (b) binary digits and
powers of 2. The same number is represented in both (a) and (b).

Binary Representation of Numbers

The decimal number 1837 is 11100101101 in binary notation. We can see that this is true if we follow the pattern for decimal
numbers in reverse. There are eleven binary digits, so we begin by expressing the number as multiples of the powers of two from
2
10
to 2 , beginning with the digit furthest to the left and moving to the right
0

10 9 8 7 6 5 4 3 2 1
(1 × 2 ) + (1 × 2 ) + (1 × 2 ) + (0 × 2 ) + (0 × 2 ) + (1 × 2 ) + (0 × 2 ) + (1 × 2 ) + (1 × 2 ) + (0 × 2 )

0
+ (1 × 2 ) = 1837

Each power of two has a decimal equivalent—2 is the same as 2 × 2 × 2 × 2 = 16 , for example—which we can express here as
4

(1 × 1024) + (1 × 512) + (1 × 256) + (0 × 128) + (0 × 64) + (1 × 32) + (0 × 16) + (1 × 8) + (1 × 4) + (0 × 2)

+ (1 × 1) = 1837

Each power of two represents a place as well; thus, the second 0 from the right is in the sixteenths place. Figure 4.2.1b provides a
visual representation of these ways of expressing a binary number.

4.2.1 https://chem.libretexts.org/@go/page/407089
Converting Between Decimal and Binary Representations of Numbers
There are lots of on-line calculators that you can use to convert between decimal and binary representations of numbers, such as the
one here. Still, it is useful to be comfortable with converting numbers by hand. Converting a binary number into its decimal
equivalent is straightforward, as we showed above for the binary representation of the year in which my university was founded
11100101101 = (1 × 1024) + (1 × 512) + (1 × 256) + (0 × 128) + (0 × 64) + (1 × 32) + (0 × 16) + (1 × 8)

+ (1 × 4) + (0 × 2) + (1 × 1) = 1837

Converting a decimal number, such as 1837, into its binary equivalent requires a bit more work; Table 4.2.1 will help us organize
the conversion. We begin by writing the dividend, which is 1837, in the left-most column and divide it by 2, writing the quotient of
918 in the second column and the remainder of 1 in the third column; note that dividing by 2 gives a remainder of 0 if the dividend
is even or a remainder of 1 if the dividend is odd. The remainder is the exponent for the first place in the binary notation. In this
case, we have 2 = 1 . The quotient becomes the dividend for the next cycle, with the process continuing until we achieve a
0

quotient of 0. The binary equivalent of the original decimal is given by reading the remainders from bottom-to-top as 11100101101.
Table 4.2.1 . Converting a binary number into its decimal equivalent.
dividend quotient remainder binary notation

1837 918 1 2
0
= 1

918 459 0 2
1
= 1

459 229 1 2
2
= 1

229 114 1 2
3
= 1

114 57 0 2
4
= 1

57 28 1 2
5
= 1

28 14 0 2
6
= 1

14 7 0 2
7
= 1

7 3 1 2
8
= 1

3 1 1 2
9
= 1

1 0 1 2
10
= 1

This page titled 4.2: Working With Binary Numbers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

4.2.2 https://chem.libretexts.org/@go/page/407089
4.3: Cleaning Up Signals and Counting Events
How an instrument handles signals depends on what is being measured, so we cannot develop here a single model that applies to all
instruments. Broadly speaking, however, an instrument is likely to include one or more of the following: the ability to clean up the
raw signal and convert it into a form that we can analyze; the ability to count events in binary form; the ability to convert binary
information into a digital information; and the ability to convert between digital and analog signals. In this section we will cover
the first two of these topics.

Cleaning Up a Signal
Suppose our instrument is designed to count discrete events, perhaps a Geiger counter that detects the emission of β particles, or a
photodiode that detects photons. Even though a time-dependent count of particles is a digital signal, the raw signal (a voltage)
likely consists of digital pulses superimposed on a background signal that contains noise, as seen in Figure 4.3.1. The total signal,
therefore, is in analog form.

Figure 4.3.1 . The raw signal (voltage) for an experiment in which we are interested in counting the number of events in a defined
period of time. In this case, there are five discrete pulses, two of which partially overlap each other. The individual pulses are
superimposed on a background signal that shows a modest amount of random noise.
To clean up this signal we want to accomplish two things: remove the noise and ensure that each pulse is counted. A simple way to
accomplish this is to set a threshold signal and use a voltage follower operational amplifier (see Chapter 3) to set all voltages below
the threshold to a logical value of 0 and all voltages above the threshold to a logical value of 1. As seen in Figure 4.3.2, the choice
of the threshold voltage must be chosen carefully if we are to resolve closely spaced pulses and discriminate against noise. Note
that the peak-shaped pulses become rectangular pulses.

4.3.1 https://chem.libretexts.org/@go/page/407090
 Figure 4.3.2 . Two attempts at cleaning up the data from Figure 4.3.1 . At the top, setting the threshold voltage to 5 results in a
noise-free set of rectangular pulses that are well-separated from each other. Setting the threshold to 0.5 volts merges the two
separate pulses at times between 1.0 and 1.5 into a single pulse and captures a number of noise spikes between times of 1.5 and 2.0,
and between times of 2.5 and 3.0.

Binary Pulse Counter

To count the pulses in Figure 4.3.2 we can send them though a binary pulse counter (BPC). Figure 4.3.3 shows how such a counter
works. In this case, the BPC has three registers, each of which can be in a logical state of 0 or 1. With three registers, we are limited
to counting no more than 2 = 8 pulses; a more useful BPC would have more registers. We can treat the pulses as entering the BPC
3

from the right. When a pulse enters a register, it flips each register from 1 to 0 or from 0 to 1, stopping after if first flips a register
from 0 to 1. For example, the second pulse flips the right-most register from 1 to 0 and the middle register from 0 to 1; because the
middle register initially was at 0, the counting of this pulse comes to an end.

Figure 4.3.3 . Operation of a binary pulse counter showing how the contents of three registers respond to the measurement of five
pulses.

This page titled 4.3: Cleaning Up Signals and Counting Events is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

4.3.2 https://chem.libretexts.org/@go/page/407090
 CHAPTER OVERVIEW
5: Signals and Noise
When we try to calibrate an analytical method or to optimize an analytical system, our ability to do so successfully is limited by the
uncertainty, or noise, in our measurements and by background signals that interfere with our ability to measure the signal of interest
to us. In this chapter we will consider how we characterize noise, example of sources of noise, and ways to clean up our data by
decreasing the contribution of noise to our measurements and by correcting for the presence of background signals.
5.1: The Signal-to-Noise Ratio
5.2: Sources of Instrumental Noise
5.3: Signal-to-Noise Enhancement

This page titled 5: Signals and Noise is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

1
5.1: The Signal-to-Noise Ratio
When we make a measurement it is the sum of two parts, a determinate, or fixed contribution that arises from the analyte and an
indeterminate, or random, contribution that arises from uncertainty in the measurement process. We call the first of these the signal
and we call the latter the noise. There are two broad categories of noise: that associated with obtaining samples and that associated
with making measurements. Our interest here is in the latter.

What is Noise?
Noise is a random event characterized by a mean and standard deviation. There are many types of noise, but we will limit ourselves
for now to noise that is stationary, in that its mean and its standard deviation are independent of time, and that is heteroscedastic, in
that its mean and its variance (and its standard deviation) are independent of the signal's magnitude. Figure 5.1.1a shows an
example of a noisy signal that meets these criteria. The x-axis here is shown as time—perhaps a chromatogram—but other units,
such as wavelength (spectroscopy) or potential (electrochemistry), are possible. Figure 5.1.1b shows the underlying noise and
Figure 5.1.1c shows the underlying signal. Note that the noise in Figure 5.1.1b appears consistent in its central tendency (mean)
and its spread (variance) along the x-axis and is independent of the signal's strength.

Figure 5.1.1 : Plots showing (a) the signal and the noise in blue with the signal superimposed as a smooth line; (b) the noise only;
and (c) the signal only. The signal consists of three peaks at times of 250, 500, and 750, and with maximum values of 100, 60, and
30, respectively. The noise is drawn at random from a normal distribution with a mean of 0 and a standard deviation of 10.

How Do We Characterize the Signal and the Noise?

Although we characterize noise by its mean and its standard deviation, the most important benchmark is the signal-to-noise ratio,
S/N , which we define as

Sanalyte
S/N =
snoise

where S analyte is the signal's value at particular location on the x-axis and s is the standard deviation of the noise using a
noise

signal-free portion of the data. As general rules-of-thumb, we can measure the signal with some confidence when S/N ≥ 3 and we
can detect the signal with some confidence when 3 ≥ S/N ≥ 2 . For the data in Figure 5.1.1, and using the information in the
figure caption, the signal-to-noise ratios are, from left-to-right, 10, 6, and 3.

5.1.1 https://chem.libretexts.org/@go/page/364553
  Note

To measure the signal with confidence implies we can use the signal's value in a calculation, such as constructing a calibration
curve. To detect the signal with confidence means we are certain that a signal is present (and that an analyte responsible for the
signal is present) even if we cannot measure the signal with sufficient confidence to allow for a meaningful calculation.

This page titled 5.1: The Signal-to-Noise Ratio is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

5.1.2 https://chem.libretexts.org/@go/page/364553
5.2: Sources of Instrumental Noise
When we make an analytical measurement, we are interested in both the accuracy and the precision of our results. Noise, as we
learned in the previous section, is a random fluctuation in the signal that limits our ability to detect the presence of the underlying
signal. There are a variety of ways in which noise can enter into our measurements. Some of these sources of noise are related to
the process of collecting and processing samples for analysis; these sources of noise, which we might collectively call chemical
sources of noise, are important and receive consideration in those sections of this textbook that consider the application of
analytical methods. In this chapter, we will limit ourselves to considering sources of noise that arise from the instruments we use to
make measurements. We call these sources of instrumental noise.

Thermal Noise
Even when an external voltage is not applied to an electrical circuit, a small current is present due to the random motion of
electrons that arises from the temperature of the surroundings; we can this thermal (or, sometimes, Johnson) noise. The magnitude
of this noise in any electrical element increases with temperature, of course, but it also is affected by its resistance, and by how
quickly it responds to a change in the signal. Mathematically, we express this as the root-mean-square voltage, ν , which is given
rms

as
−−−−−−−
νrms = √4kT RΔf (5.2.1)

where k is Boltzmann's constant, T is the temperature in Kelvin, R is the resistance in ohms, and Δf is the bandwidth. The latter
term is a measure of how quickly the electrical element responds to a change in its input by changing its output from 10% to 90%
of its final value, which is called the rise time, t , where
r

1
Δf =
3tr

For example, if a change in the input increases the output by 1, then the rise time is how long it takes the output to increase from
0.1 to 0.9.
A close look at Equation 5.2.1 shows that we can reduce thermal noise by decreasing the temperature, by decreasing the resistance
of the electrical circuit, and by decreasing the bandwidth; the latter, of course, comes at the cost of an increase in the response time,
which means the instrument responds more slowly to a change in the signal. Of these, it is often easiest to reduce the temperature
by cooling, for example, the instrument's detector.

Shot Noise
As its name implies, shot noise is a discrete event that happens in response to an event, such as the movement of an electron
through the space between two surfaces of opposite charge. These events are random and quantized, and generate random
furcations in the current that have a root-mean-square value, i , which is given by
rms

−−− −−−
irms = √2I eΔf (5.2.2)

where I is the average current, e is the charge on the electron in Coulombs, and Δf is the bandwidth. Of these terms, the only one
under our control is the bandwidth; again, decreasing the bandwidth comes at the cost of an instrument that responds more slowly
to a change in the signal.

Flicker Noise
Unlike thermal noise or shot noise, flicker noise is related to the frequency of the signal being measured, f , instead of the signal's
bandwidth. The sources of flicker noise are not well understood, but it is known that it is inversely proportional to the signal's
frequency; thus, flicker noise is sometimes called 1/f noise. Because of the inverse relationship, flicker noise is more important at
low frequencies, where it appears as a long-term drift in the signal. It is less important at higher frequencies where thermal noise
and shot noise are more important.

Environmental Noise
Our instruments normally do not operate in an environment free from external signals, each of which has a frequency that can be
picked up by the instrument. Television signals, cell-phone signals, radio signals, power lines are obvious examples of high-to-

5.2.1 https://chem.libretexts.org/@go/page/364554
moderate frequencies that can serve as noise. Less obvious are lower frequency sources of noise, such as the change in temperature
during the day or through the year.

This page titled 5.2: Sources of Instrumental Noise is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

5.2.2 https://chem.libretexts.org/@go/page/364554
5.3: Signal-to-Noise Enhancement
There are two broad approaches we can use to improve the signal-to-noise ratio: hardware and software. Hardware approaches are
built into the instrument and include decisions on how the instrument is set-up for making measurements (for example, the choice
of a scan rate or a slit width), and how the signal is processed by the instrument (for example, using electronic filters). A few
approaches are briefly considered here; others are included with the discussion of individual instruments. Software solutions are
computational approaches in which we manipulate the data either while we are collecting it or after data acquisition is complete.

Hardware Solutions
One way to reduce noise is to focus on the circuitry, or hardware, used to measure the signal.

Shielding
One way to reduce environmental noise is to prevent it from entering into the instrument's electronic circuitry. One approach is to
use a Faraday cage in which the instrument sits within a room or space covered with a conductive material. Electromagnetic
radiation from the environment is absorbed by the conductive material and then shunted away to the ground. Rather than encasing
the entire instrument in a Faraday cage, particularly sensitive portions of the circuitry can be shielded.

Differential Amplifier
A difference amplifier (see Chapter 3) is an electrical circuit used to determine the difference between two input voltages or
currents and to return that difference as a larger voltage or current. As the magnitude of the noise in the two input signals is
generally similar in value—that is, it is in phase—while the signal of interest is not, much of the noise's contribution to the signal is
subtracted out.

Filtering
When the frequency of the noise is quite different from the frequency of the signal, a simple electrical circuit can be used to remove
the high frequency noise and pass the low frequency signal; this is called a low-pass filter. See Chapter 2 for details on low-pass
filters.

Modulation
When the signal of interest has a low frequency, the effect of flicker noise becomes significant because a technique that removes
low frequency noise will remove the signal as well. Modulation is a process of increasing the frequency of the signal. When
complete, a high-pass filter is used to remove the noise. Reversing the modulation returns the original signal, but with much of the
noise removed.

Software Solutions
In this section we will consider three common computational tools for improving the signal-to-noise ratio: signal averaging, digital
smoothing, and Fourier filtering.

Signal Averaging
The most important difference between the signal and the noise is that a signal is determinate (fixed in value) and the noise is
indeterminate (random in value). If we measure a pure signal several times, we expect its value to be the same each time; thus, if
we add together n scans, we expect that the net signal, S , is defined as
n

Sn = nS

where S is the signal for a single scan. Because noise is random, its value varies from one run to the next, sometimes with a value
that is larger and sometimes with a value that is smaller, and sometimes with a value that is positive and sometimes with a value
that is negative. On average, the standard deviation of the noise increases as we make more scans, but it does so at a slower rate
than for the signal
−
sn = √n s

where s is the standard deviation for a single scan and s is the standard deviation after n scans. Combining these two equations,
n

shows us that the signal-to-noise ratio, S/N , after n scans increases as

5.3.1 https://chem.libretexts.org/@go/page/364556
 Sn nS −
(S/N )n = = − = √n (S/N )n=1
sn √n s

where (S/N ) n=1 is the signal-to-noise ratio for the initial scan. Thus, when n = 4 the signal-to-noise ratio improves by a factor of
2, and when n = 16 the signal-to-noise ratio increases by a factor of 4. Figure 5.3.1 shows the improvement in the signal-to-noise
ratio for 1, 2, 4, and 8 scans.

Figure 5.3.1 : Improvement in the signal-to-noise ratio through signal averaging using 1, 2, 4, and 8 scans. Each plot shows the
noisy signal in blue with the pure signal superimposed in black. The total signal is divided by the number of scans so that each y-
axis has the same scale.
Signal averaging works well when the time it takes to collect a single scan is short and when the analyte's signal is stable with
respect to time both because the sample is stable and the instrument is stable; when this is not the case, then we risk a time-
dependent change in S and/or s
analyte noiseBecause the equation for (S/N ) is proportional to the √−
n n , the relative improvement in

the signal-to-noise ratio decreases as n increases; for example, 16 scans gives a 4× improvement in the signal-to-noise ratio, but it
takes an additional 48 scans (for a total of 64 scans) to achieve a 8× improvement in the signal-to-noise ratio.

Digital Smoothing Filters

One characteristic of noise is that its magnitude fluctuates rapidly in contrast to the underlying signal. We see this, for example, in
Figure 5.3.1 where the underlying signal either remains constant or steadily increases or decreases while the noise fluctuates
chaotically. Digital smoothing filters take advantage of this by using a mathematical function to average the data for a small range
of consecutive data points, replacing the range's middle value with the average signal over that range.
Moving Average Filters
For a moving average filter, also called a boxcar filter, we replace each point by the average signal for that point and an equal
number of points on either side; thus, a moving average filter has a width, w, of 3, 5, 7, ... points. For example, suppose the first
five points in a sequence are

0.80 0.30 0.80 0.20 1.00

then a three-point moving average (w = 3) returns values of

NA 0.63 0.43 0.67 NA

where, for example, 0.63 is the average of 0.80, 0.30, and 0.80. Note that we lose (w − 1)/2 = (3 − 1)/2 = 1 points at each end
of the data set because we do not have a sufficient number of data points to complete a calculation for the first and the last point.
Figure 5.3.2 shows the improvement in the S/N ratio when using moving average filters with widths of 5, 9, and 13.

5.3.2 https://chem.libretexts.org/@go/page/364556
 Figure 5.3.2 : Improvement in the signal-to-noise ratio using moving average filters with ranges of 5, 9, and 13 on the original data
shown in the upper left quadrant. Each plot shows the noisy signal in blue with the pure signal superimposed in black.

One limitation to a moving average filter is that it distorts the original data by removing points from both ends, although this is not
a serious concern if the points in question are just noise. Of greater concern is the distortion in a signal's height if we use a range
that is too wide; for example, Figure 5.3.3, shows how a 23-point moving average filter (shown in blue) applied to the noisy signal
in the upper left quadrant of Figure 5.3.2, reduces the height of the original signal (shown in black). Because the filter's width—
shown by the red bar—is similar to the peak's width, as the filter passes through the peak it systematically reduces the signal by
averaging together values that are mostly smaller than the maximum signal.

Figure 5.3.3 : Example that shows how a moving average filter can distort the signal when applied to the noisy signal in the upper
left quadrant of Figure 5.3.2 . The original pure signal is shown in black and the signal after applying a 23-point moving average
filter is shown in blue. The width of the moving average filter is shown by the red bar.
Savitzky-Golay Filters
A moving average filter weights all points equally; that is, points near the edges of the filter contribute to the average as a level
equal to points near the filter's center. A Savitzky-Golay filter uses a polynomial model that weights each point differently, placing
more weight on points near the center of the filter and less weight on points at the edge of the filter. Specific values depend on the
size of the window and the polynomial model; for example, a five-point filter using a second-order polynomial has weights of

−3/35 12/35 17/35 12/35 − 3/35

For example, suppose the first five points in a sequence are

5.3.3 https://chem.libretexts.org/@go/page/364556
 0.80 0.30 0.80 0.20 1.00

then this Savitzky-Golay filter returns values of

NA NA 0.41 NA NA

where, for example, the value for the middle point is

−3 12 17 12 −3
0.80 × + 0.30 × + 0.80 × + 0.20 × + 1.00 × = 0.406 ≈ 0.41
35 35 35 35 35

Note that we lose (w − 1)/2 = (5 − 1)/2 = 2 points at each end of the data set, where w is the filter's range, because we do not
have a sufficient number of data points to complete the calculations. For other Savitzky-Golay smoothing filters, see Savitzky, A.;
Golay, M. J. E. Anal Chem, 1964, 36, 1627-1639. Figure 5.3.4 shows the improvement in the S/N ratio when using Savitzky-
Golay filters using a second-order polynomial with 5, 9, and 13 points.

Figure 5.3.4 : Improvement in the signal-to-noise ratio using Savitzky-Golay filters of ranges of 5, 9, and 13 on the original data
shown in the upper left quadrant. Each plot shows the noisy signal in blue with the pure signal superimposed in black.
Because a Savitzky-Golay filter weights points differently than does a moving average smoothing filter, a Savitzky-Golay filter
introduces less distortion to the signal, as we see in the following figure.

5.3.4 https://chem.libretexts.org/@go/page/364556
 Figure 5.3.5 : A Savitzky-Golay filter is less aggressive than a moving average filter. Applying a 23-point Savitzky-Golay filter to
the noisy signal in the upper left quadrant of Figure 5.3.4 results in little distortion of the signal. Contrast this with Figure 5.3.3
where a 23-point moving average filter results in substantial distortion of the signal. The original pure signal is shown in black and
the signal after applying a 23-point Savitzky-Golay filter is shown in blue. The width of the Savitzky-Golay filter is shown by the
red bar.
Fourier Filtering
This approach to improving the signal-to-noise ratio takes advantage of a mathematical technique called a Fourier transform (FT).
The basis of a Fourier transform is that we can express a signal in two separate domains. In the first domain the signal is
characterized by one or more peaks, each defined by its position, its width, and its area; this is called the frequency domain. In the
second domain, which is called the time domain, the signal consists of a set of oscillations, each defined by its frequency, its
amplitude, and its decay rate. The Fourier transform—and the inverse Fourier transform—allow us to move between these two
domains.

 Note

The mathematical details behind the Fourier transform are beyond the level of this textbook; for a more in-depth treatment,
consult this series of articles from the Journal of Chemical Education:
Glasser, L. “Fourier Transforms for Chemists: Part I. Introduction to the Fourier Transform,” J. Chem. Educ. 1987, 64,
A228–A233.
Glasser, L. “Fourier Transforms for Chemists: Part II. Fourier Transforms in Chemistry and Spectroscopy,” J. Chem.
Educ. 1987, 64, A260–A266.
Glasser, L. “Fourier Transforms for Chemists: Part III. Fourier Transforms in Data Treatment,” J. Chem. Educ. 1987, 64,
A306–A313.

Figure 5.3.6a shows a single peak in the frequency domain and Figure 5.3.6b shows its equivalent time domain signal. There are
correlations between the two domains:
the further a peak in the frequency domain is from the origin, the greater it corresponding oscillation frequency in the time
domain
the broader a peak's width in the frequency domain, the faster its decay rate in the time domain
the greater the area under a peak in the frequency domain, the higher its initial intensity in the time domain

5.3.5 https://chem.libretexts.org/@go/page/364556
 Figure 5.3.6 : The plot in (a) shows a frequency domain consisting of a single peak defined by its position along the x-axis, its
width, and its area. The plot in (b) shows the corresponding time domain that consists of a single oscillating signal defined by its
oscillation frequency, its initial intensity, and its decay rate.
We can use a Fourier transform to improve the signal-to-noise ratio because the signal is a single broad peak and the noise appears
as a multitude of very narrow peaks. As noted above, a broad peak in the frequency domain has a fast decaying signal in the time
domain, which means that while the beginning of the time domain signal includes contributions from the signal and the noise, the
latter part of the time domain signal includes contributions from noise only. The figure below shows how we can take advantage of
this to reduce the noise and improve the signal-to-noise ratio for the noisy signal in Figure 5.3.7a, which has 256 points along the
x-axis and has a signal-to-noise ratio of 5.1. First, we use the Fourier transform to convert its original domain into the new domain,
the first 128 points of which are shown in Figure 5.3.7b (note: the first half of the data contains the same information as the second
half of the data, so we only need to look at the first half of the data). The points at the beginning are dominated by the signal, which
is why there is a systematic decrease in the intensity of the oscillations; the remaining points are dominated by noise, which is why
the variation in intensity is random. To filter out the noise we retain the first 24 points as they are and set the intensities of the
remaining points to zero (the choice of how many points to retain may require some adjustment). As shown in Figure 5.3.7c, we
repeat this for the remaining 128 points, retaining the last 24 points as they are. Finally, we use an inverse Fourier transform to
return to our original domain, with the result in Figure 5.3.7d, with the signal-to-noise ratio improving from 5. 1 for the original
noisy signal to 11.2 for the filtered signal.

Figure 5.3.7 : Example of removing noise using a Fourier filter. The original noisy signal with S/N = 5.1 is shown in (a) and is
similar to the noisy signal in Figure 5.3.2a and Figure 5.3.4a . The first half of the Fourier transformed data is shown in (b). The
Fourier transformed data is shown in (c) after zeroing out all but the first and the last 24 points. Returning to the original domain
gives the final filtered signal with S/N = 11.1 .

5.3.6 https://chem.libretexts.org/@go/page/364556
This page titled 5.3: Signal-to-Noise Enhancement is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

5.3.7 https://chem.libretexts.org/@go/page/364556
 CHAPTER OVERVIEW
6: An Introduction to Spectrophotometric Methods
6.1: General Properties of Electromagnetic Radiation
6.2: Wave Properties of Electromagnetic Radiation
6.3: Quantum Mechanical Properties of Electromagnetic Radiation
6.4: Emission and Absorbance Spectra
6.5: Quantitative Considerations

This page titled 6: An Introduction to Spectrophotometric Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

1
6.1: General Properties of Electromagnetic Radiation
Electromagnetic radiation—light—is a form of energy whose behavior is described by the properties of both waves and particles.
Some properties of electromagnetic radiation, such as its refraction when it passes from one medium to another (F igure6.1.1), are
explained best when we describe light as a wave. Other properties, such as absorption and emission, are better described by treating
light as a particle. The exact nature of electromagnetic radiation remains unclear, as it has since the development of quantum
mechanics in the first quarter of the 20th century [Home, D.; Gribbin, J. New Scientist 1991, 2 Nov. 30–33]. Nevertheless, this dual
model of wave and particle behavior provides a useful description for electromagnetic radiation.

Figure 6.1.1 . The Golden Gate bridge as seen through rain drops. Refraction of light by the rain drops produces the distorted
images. Source: Brocken Inaglory (commons. Wikipedia.org).

The Electromagnetic Spectrum

The frequency and the wavelength of electromagnetic radiation vary over many orders of magnitude. For convenience, we divide
electromagnetic radiation into different regions—the electromagnetic spectrum—based on the type of atomic or molecular
transitions that gives rise to the absorption or emission of photons (Figure 6.1.2). The boundaries between the regions of the
electromagnetic spectrum are not rigid and overlap between spectral regions is possible.

Figure 6.1.2 . The electromagnetic spectrum showing the boundaries between different regions and the type of atomic or molecular
transitions responsible for the change in energy. The colored inset shows the visible spectrum. Source: modified from Zedh
(www.commons.Wikipedia.org).

This page titled 6.1: General Properties of Electromagnetic Radiation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

6.1.1 https://chem.libretexts.org/@go/page/364619
6.2: Wave Properties of Electromagnetic Radiation
Ways to Characterize a Wave
Electromagnetic radiation consists of oscillating electric and magnetic fields that propagate through space along a linear path and
with a constant velocity. The oscillations in the electric field and the magnetic field are perpendicular to each other and to the
direction of the wave’s propagation. Figure 6.2.1 shows an example of plane-polarized electromagnetic radiation, which consists of
a single oscillating electric field and a single oscillating magnetic field.

Figure 6.2.1 . Plane-polarized electromagnetic radiation showing the oscillating electric field in blue and the oscillating magnetic
field in red. The radiation’s amplitude, A, and its wavelength, λ , are shown. Normally, electromagnetic radiation is unpolarized,
with oscillating electric and magnetic fields present in all possible planes perpendicular to the direction of propagation.

Measurable Properties
An electromagnetic wave is characterized by several fundamental properties, including its velocity, amplitude, frequency, phase
angle, polarization, and direction of propagation [Ball, D. W. Spectroscopy 1994, 9(5), 24–25]. Focusing on the oscillations in the
electric field, amplitude is the maximum displacement of the electrical field. The wave's frequency, ν , is the number of oscillations
in the electric field per unit time. Wavelength, λ is defined as the distance between successive maxima . Figure 6.2.1 shows the
initial amplitude as 0; the phase angle Φ accounts for the fact that the initial amplitude need not be zero, which we can accomplish
by shifting the wave along the direction of propagation.
There is a relationship between wavelength and frequency, which is
c
λ =
ν

where c is the speed of light in a vacuum. Another unit useful unit is the wavenumber, ν̄ , which is the reciprocal of the wavelength
¯
¯

1
¯
¯¯
ν =
λ

Wavenumbers frequently are used to characterize infrared radiation, with the units given in cm–1. Power, P , and intensity, I , are
two additional properties of light, both related to the square of the amplitude; power is the energy transferred per second and
intensity is the power transferred to a given area.
In a vacuum, electromagnetic radiation travels at the speed of light, c, which is 2.99792 × 10 m/s. When electromagnetic
8

radiation moves through a medium other than a vacuum, its velocity, v, is less than the speed of light in a vacuum. The difference
between v and c is sufficiently small (<0.1%) that the speed of light to three significant figures, 3.00 × 10 m/s, is accurate enough
8

for most purposes.

When electromagnetic radiation moves between different media—for example, when it moves from air into water—its
frequency, ν , remains constant. Because its velocity depends upon the medium in which it is traveling, the electromagnetic
radiation’s wavelength, λ , changes. If we replace the speed of light in a vacuum, c, with its speed in the medium, v , then the
wavelength is
v
λ =
ν

6.2.1 https://chem.libretexts.org/@go/page/364621
 This change in wavelength as light passes between two media explains the refraction of electromagnetic radiation seen in the
photograph of light passing through rain drop, what was included in the previous section. This is discussed in more detail later
in this section.

 Example 6.2.1
In 1817, Josef Fraunhofer studied the spectrum of solar radiation, observing a continuous spectrum with numerous dark lines.
Fraunhofer labeled the most prominent of the dark lines with letters. In 1859, Gustav Kirchhoff showed that the D line in the
sun’s spectrum was due to the absorption of solar radiation by sodium atoms. The wavelength of the sodium D line is 589 nm.
What are the frequency and the wavenumber for this line?

Solution
The frequency and wavenumber of the sodium D line are
8
c 3.00 × 10  m/s
14 −1
ν = = = 5.09 × 10  s
−9
λ 589 × 10  m

1 1 1 m 4 −1
¯
¯¯
ν = = × = 1.70 × 10  cm
−9
λ 589 × 10  m 100 cm

 Exercise 6.2.1

Another historically important series of spectral lines is the Balmer series of emission lines from hydrogen. One of its lines has
a wavelength of 656.3 nm. What are the frequency and the wavenumber for this line?

Answer
The frequency and wavenumber for the line are
8
c 3.00 × 10  m/s
14 −1
ν = = = 4.57 × 10  s
−9
λ 656.3 × 10  m

1 1 1 m
¯
¯¯ 4 −1
ν = = × = 1.524 × 10  cm
−9
λ 656.3 × 10  m 100 cm

Polarization
Figure 6.2.1 shows a single oscillating electrical field and, perpendicular to that, a single oscillating magnetic field. This is an
example of plane polarized light in which oscillation of the electrical field occurs at just one angle. Normally electromagnetic
radiation oscillates simultaneously at all possible angles. Figure 6.2.2 shows the difference in these two cases. If we observe the
plane polarized light as it oscillates toward us, we see the single line at the top of the figure where blue indicates a positive
amplitude and red indicates a negative amplitude, and where the opacity of the shading shows the change in the amplitudes. The
vertical dashed lines show nodes where the amplitude is zero and where no light is seen. With ordinary light, we see a circular
beam of radiation because the electrical field is oscillating at all angles. The amplitude's sign and magnitude, and the presence of
nodes where the amplitude is zero, remain evident to us. Note that if we observe the source's intensity, then the each of the lines
and circles in Figure 6.2.2 will appear blue (positive values as intensity is proportion to the square of the amplitude); we continue
to observe fluctuations in the intensity and the presence of the nodes.

6.2.2 https://chem.libretexts.org/@go/page/364621
 Figure 6.2.2 : Illustration showing the difference between plane polarized light (top) and ordinary light (bottom). Blue is used to
show positive amplitudes and red is used to show negative amplitudes. The opacity of the colors—which have values of 100%,
67%, 33%, and 0%—indicate the relative magnitudes of the amplitudes. When the opacity is 0%, the amplitude is zero and there is
a node.

Mathematical Representation of Waves

We can describe the oscillations in the electric field as a sine wave

At = Ae sin(2πν t + Φ)

where At is the magnitude of the electric field at time t, Ae is the field’s maximum amplitude, ν is the wave's frequency, and Φ is a
phase angle that accounts for the fact that A need not have a value of zero at time t = 0 . The identical equation for the magnetic
t

field is

At = Am sin(2πν t + Φ)

where Am is the magnetic field’s maximum amplitude.

One of the important features of waves is that adding or subtracting together two (or more) gives a new wave. Figure 6.2.3 shows
one example. The superposition of waves explains why two identical waves that are completely out-of-phase with each other
produce a signal in which the amplitude is zero at all points.

6.2.3 https://chem.libretexts.org/@go/page/364621
 Figure 6.2.3 : Illustration of the principle of superposition. The x-axis scale runs from zero to 2π. The equation for the wave shown
as a dashed blue line is y = sin(x) and the equation for the wave shown as a dashed red line is y = 2sin(π/16 + 3x) . The sum of
these two waves is the solid black line.
Another important consequence of the superposition of waves is that if we can add together a series of waves to produce a new
wave, then there is a corresponding mathematical process that takes a complex wave and determines the underlying set of sine
waves of which it is comprised. This process is called a Fourier transform, which we will revisit in later chapters.

Interactions of Waves With Matter

When light encounters matter—perhaps a particle, a solution, or a thin film—it can interact with it in several ways. In this section
we consider two such interactions: refraction and reflection. Three additional types of interactions—the scattering of light, the
diffraction of light, and the transmission of light—are considered in later chapters where they play an important role in specific
instrumental methods of analysis.

Refraction
When light passes from one medium (perhaps air) into another medium (perhaps water) that has a different density, the light
experiences a change in direction that is a consequence of a difference in its velocity in the two media. This bending of light is
called refraction, the extent of which is given by Snell's law
sin θ1 η2 v1
= =
sin θ2 η1 v2

where ηi is the refractive index of a medium and vi is the velocity in a medium, and where the angles, θi , are shown in Figure
6.2.4.

6.2.4 https://chem.libretexts.org/@go/page/364621
 Figure 6.2.4 : Illustration of the refraction of light as it moves across the interface of two media with different refractive indexes.
The angles of the light are given by θ and are measured relative to a surface normal, which is shown by the dashed line. The denser
medium has the smaller value of θ and the larger the refractive index.

Reflection
In addition to refraction, when light crosses an interface that separates media with different refractive indexes, some of the light is
reflected back. When then angle of incidence is 0° (that is, the light is perpendicular to the interface), then the fraction of light that
is reflected is given by
2
Ir (η2 − η1 )
=
2
I0 (η2 + η1 )

where I is the intensity of light that is reflected, I is the intensity of light from the source that enters the interface, and η is the
r 0 i

refractive index of the media. If light crosses more than one interface—as is the case when light passes through a sample cell—then
the total fraction of reflected light is the sum of the fraction of light reflected at each interface.

This page titled 6.2: Wave Properties of Electromagnetic Radiation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

6.2.5 https://chem.libretexts.org/@go/page/364621
6.3: Quantum Mechanical Properties of Electromagnetic Radiation
In the last section, we considered properties of electromagnetic radiation that are consistent with identifying light as a wave. Other
properties of light, however, cannot be explained by a model that treats it as a wave; instead, we need to consider a model that
treats light as a system of discrete particles, which we call photons.

The Photoelectric Effect

As shown in Figure 6.3.1, in a photoelectric cell, a metal, such as sodium, is held under vacuum and exposed to electromagnetic
radiation, which enters the cell through an optical window. If the frequency of the radiation is sufficient, electrons escape from the
metal with a kinetic energy that we can measure; we call these photoelectrons. If the photocell's anode is held at a potential that is
positive relative to the potential applied to the cathode, the photoelectrons move from the cathode to the anode, generating a current
that is measured by an ammeter. If the voltage applied to the anode is made sufficiently negative, the electrons eventually fail to
reach the anode and the current decreases to zero. The voltage needed to stop the flow of electrons is called the stopping voltage.

Figure 6.3.1 : Illustration of a photoelectric cell. The photocathode is fashioned from or coated with a metal and placed in an
evacuated cell along with an anode. Electromagnetic radiation is focused on the photocathode and, if it has sufficient energy, the
metal emits electrons with a kinetic energy that depends on the metal and the frequency of the electromagnetic radiation. The
potential applied to the photocathode and the anode determine whether the electrons reach the anode; if they do, then a current
flows, which is measured by the ammeter.
In a photoelectron spectrum we vary the frequency and intensity of the electromagnetic radiation and observe their effect on either
the number of photoelectrons released (measured as a current) or the energy of the photoelectrons released (measured by their
kinetic energy). A typical set of experiments are shown in Figure 6.3.2a using Na and in Figure 6.3.2b using Na, Zn, and Cu. The
data show several interesting features. First, we see in Figure 6.3.2a that the intensity of the light source has no effect on the
minimum frequency of light needed to eject a photoelectron from Na—we call this the threshold frequency—but that a high
intensity source of electromagnetic radiation results in the release of a greater number of photoelectrons and, therefore, a greater
current than for a lower intensity source. Second, we see in Figure 6.3.2b that different metals have different threshold frequencies,
but that once we exceed each metal's threshold frequency, the change in the kinetic energy of the photoelectrons with increasing
frequency yields lines of equal slopes.

6.3.1 https://chem.libretexts.org/@go/page/364623
 Figure 6.3.2 : The two types of photoelectric effect experiments. In (a) we measure the current that flows as we change the intensity
and the frequency of the electromagnetic radiation. In (b) we measure the kinetic energy of the photoelectrons as a function of the
frequency of the electromagnetic radiation.
We can explain these experimental observations if we assume that the source of electromagnetic energy has an energy, E , that ER

does two things: it overcomes the energy that binds the photoelectron to the metal, E , and it imparts the remaining energy into
BE

the photoelectron's kinetic energy, E , where ER means electromagnetic radiation, BE means binding energy and KE means
KE

kinetic energy.

EKE = EER − EBE

A wave model for electromagnetic radiation is insufficient to explain the photoelectric effect because when it strikes the metal the
radiation's energy would be distributed across all atoms on the surface, none of which would then receive an energy that exceeds
the photoelectron's binding energy. Instead, the results in Figure 6.3.2 make sense only if we assume that light consists of discrete
particles with energies that are a function of frequency or wavelength
hc
EER = hν = (6.3.1)
λ

where h is Plank's constant. This leave us with the following equation relating kinetic energy, the energy of the photon, and the
binding energy of the electron.

EKE = hν − EBE

Note that the slope of the lines in Figure 6.3.2b is Plank's constant.

Energy States
Equation 6.3.1 is central to the particle, or quantum mechanical model of the atom in which we understand that chemical species—
atoms, ions, molecules—exist only in discrete states, each with a single, well-defined energy. A wave, on the other hand, can take
on any energy. A simple image is the possible energies of a ball as it rolls down a ramp (wave) or a staircase (particle), as in Figure
6.3.3.

Figure 6.3.3 . The energy of a ball is measured by its height relative to its final position, shown in blue. When a ball rolls down a
ramp, as in (a), it can take on any energy between its initial and final energies. When a ball rolls down a staircase, as in (b), it can
take only certain discrete energies. The ball in (b) is said to be quantized.
When an atom, ion, or molecule moves between two of these discrete states, the difference in energy, ΔE, is given by

6.3.2 https://chem.libretexts.org/@go/page/364623
 hc
ΔE = hν =
λ

In absorption spectroscopy a photon is absorbed by an atom, ion, or molecule, which undergoes a transition from a lower-energy
state to a higher-energy, or excited state (Figure 6.3.4a). The reverse process, in which an atom, ion, or molecule emits a photon as
it moves from a higher-energy state to a lower energy state (6.3.4b), is called emission.

Figure 6.3.4 . A simplified energy diagram that shows the absorption and emission of a photon by an atom or a molecule. When a
photon of energy hν strikes the atom, ion, or molecule, absorption may occur if the difference in energy, ΔE, between the ground
state and the excited state is equal to the photon’s energy. An atom, ion, or molecule in an excited state may emit a photon and
return to the ground state. The photon’s energy, hν , equals the difference in energy, ΔE, between the two states.
The types of energy states involved in emission and absorption depend on the energy of the electromagnetic radiation. In general,
γ-rays involve transitions between nuclear states, X-rays probe the energies of core-level electrons, ultraviolet-visible radiation

probes the energies of valence electrons, infrared radiation provides information on vibrational energy states, microwave radiation
probes rotational energy levels and electron spins, and radio waves provide information on nuclear spins. While infrared
spectroscopy may provide information on a molecule's vibrational energy states, the energies available in ultraviolet-visible
spectroscopy provide information on both the molecule's electronic states and its vibrational states, as shown in Figure 6.3.5.

Figure 6.3.5 . Diagram showing two electronic energy levels (E0 and E1), each with five vibrational energy levels ν . Absorption
0−4

of ultraviolet and visible radiation (shown by the blue arrows) leads to a change in the analyte’s electronic energy levels and,
possibly, a change in vibrational energy as well. A change in vibrational energy without a change in electronic energy levels occurs
with the absorption of infrared radiation (shown by the red arrows).

This page titled 6.3: Quantum Mechanical Properties of Electromagnetic Radiation is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by David Harvey.

6.3.3 https://chem.libretexts.org/@go/page/364623
6.4: Emission and Absorbance Spectra
In the last section we considered the source of emission and absorption. In this section we consider the types of emission and
absorbance spectra that we will form the basis for many of the chapters that follow.

Emission Spectra
When an atom, ion, or molecule moves from a higher-energy state to a lower-energy state it emits photons with energies equal to
the difference in energy between the two states. The result is an emission spectrum that shows the intensity of emission as a
function of wavelength. The shapes of these emission spectra fall into two broad types: line spectra and band spectra.

Line Spectra
When the energy states are well separated from each other, and when there is just one type of transition between the energy states,
the result is a line spectrum that consists of a small number of narrow bands. Figure 6.4.1, for example, shows the emission
spectrum from gas phase Cu atoms, which consists of seven lines, two of which are too close to each other to resolve them from
each other. The individual emissions lines are very narrow, as we might expect, because the atom's energy levels have precise
values.

Figure 6.4.1 . Emission spectrum from a Cu hollow cathode lamp in which Cu atoms are present in the gas phase. This spectrum
consists of seven distinct emission lines (the first two differ by only 0.4 nm and are not resolved at the scale shown in this
spectrum). Each emission line has a width of approximately 0.01 nm at 1⁄2 of its maximum intensity. See Chapter 9 for more
information on hollow cathode lamps.

Band Spectra
The emission spectrum for a gas phase atom is relatively simple because the number of possible transitions is small and because
their individuals energies are well-separated from each other. When a molecule in a solvent emits light, the number of possible
changes in energy levels can be quite large if the molecule undergoes transitions between electronic, vibrational, and rotational
energy levels. The resulting spectrum has so many emission individual emission lines that we see a single broad peak, or band, that
we call a band spectrum. Figure 6.4.2 shows the emission spectrum for the dye coumarin 343, which is incorporated in a reverse
micelle and suspended in cyclohexanol.

6.4.1 https://chem.libretexts.org/@go/page/365610
 Figure 6.4.2 . Emission spectrum of the dye coumarin 343, which is incorporated in a reverse micelle suspended in cyclohexanol.
The sharp peak at 409 nm is from the laser source used to excite coumarin 343. The broad band centered at approximately 500 nm
is the dye’s emission band. Source: data from Bridget Gourley, Department of Chemistry & Biochemistry, DePauw University.

 Note

When considering sources of electromagnetic radiation for spectroscopic instruments, we usually describe them as line sources
and continuous sources depending on on whether they emit discrete lines, as is the case for the hollow cathode lamp in Figure
6.4.1, or exhibit emission over a broad range of wavelengths without any gaps, as is the case for a green light-emitting diode

(LED), whose spectrum is shown in Figure 6.4.3.

Figure 6.4.3 . Spectrum showing the emission from a green LED, which provides continuous emission over a wavelength range
of approximately 530–640 nm.

Absorbance Spectra
When an atom, ion, or molecule moves from a lower-energy state to a higher-energy state it absorbs photons with energies equal to
the difference in energy between the two states. The result is an absorbance spectrum that shows the intensity of emission as a
function of wavelength. As is the case for emission spectra, absorbance spectra range from narrow lines to broad bands. The atomic
absorption spectrum for Na is shown in Figure 6.4.4, and is typical of that found for most atoms. The most obvious feature of this
spectrum is that it consists of a small number of discrete absorption lines that correspond to transitions between the ground state
(the 3s atomic orbital) and the 3p and the 4p atomic orbitals.

6.4.2 https://chem.libretexts.org/@go/page/365610
 Figure 6.4.4 . Atomic absorption spectrum for sodium. Note that the scale on the x-axis includes a break.
Another feature of the atomic absorption spectrum in Figure 6.4.4 is the narrow width of the absorption lines, which is a
consequence of the fixed difference in energy between the ground state and the excited state, and the lack of vibrational and
rotational energy levels. Natural line widths for atomic absorption, which are governed by the uncertainty principle, are
approximately 10–5 nm. Other contributions to broadening increase this line width to approximately 10–3 nm.
The absorbance spectra for molecules consists of broad bands for the same reasons discussed above for emission spectra. The
UV/Vis spectrum for cranberry juice in Figure 6.4.5 shows a single broad band for the anthocyanin dyes that are responsible for its
red color. The IR spectrum for ethanol in Figure 6.4.6 shows multiple absorption bands, some broader and some narrower. The
narrow bands, however, are still much broader than the lines in the atomic absorption spectrum for Na.

Figure 6.4.5 . Visible absorbance spectrum for cranberry juice. The anthocyanin dyes in cranberry juice absorb visible light with
blue, green, and yellow wavelengths; as a result, the juice appears red.

Figure 6.4.6 . Infrared spectrum of ethanol.

This page titled 6.4: Emission and Absorbance Spectra is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

6.4.3 https://chem.libretexts.org/@go/page/365610
6.5: Quantitative Considerations
An important part of the chapters that follow is a consideration of how we can use the emission or absorbance of photons to
determine the concentration of an analyte in a sample. Here we provide a brief summary of quantitative spectroscopic methods of
analysis in Table 6.5.1, leaving more specific details for later chapters.
Table 6.5.1 . Quantitative spectroscopic methods of analysis.
What is the relationship between
What happens to the photons? What do we measure? what we measure and What are some examples?
concentration, C ?

flame atomic emission, molecular

emitted the power, P , of emitted light
e Pe = kC
fluorescence and phosphorescence
nephelometry, turbidity, Raman
scattered the power, P , of scattered light
sc Psc = kC
spectroscopy

the power, P , of transmitted light

t
Pt flame atomic absorbance,
absorbed relative to the power, P , of the − log( )
0 P0
molecular absorbance
light source

This page titled 6.5: Quantitative Considerations is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

6.5.1 https://chem.libretexts.org/@go/page/407712
 CHAPTER OVERVIEW
7: Components of Optical Instruments
An early example of a colorimetric analysis is Nessler’s method for ammonia, which was introduced in 1856. Nessler found that
adding an alkaline solution of HgI2 and KI to a dilute solution of ammonia produced a yellow-to-reddish brown colloid, in which
the colloid’s color depended on the concentration of ammonia. By visually comparing the color of a sample to the colors of a series
of standards, Nessler was able to determine the concentration of ammonia. Colorimetry, in which a sample absorbs visible light, is
one example of a spectroscopic method of analysis. At the end of the nineteenth century, spectroscopy was limited to the
absorption, emission, and scattering of visible, ultraviolet, and infrared electromagnetic radiation. Since then, spectroscopy has
expanded to include other forms of electromagnetic radiation—such as X-rays, microwaves, and radio waves—and other energetic
particles—such as electrons and ions.
7.1: General Design of Optical Instruments
7.2: Sources of Radiation
7.3: Wavelength Selectors
7.4: Sample Containers
7.5: Radiation Transducers
7.6: Fiber Optics
7.7: Fourier Transform Optical Spectroscopy

Thumbnail: A diffraction grating monochromator.

This page titled 7: Components of Optical Instruments is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1
7.1: General Design of Optical Instruments
The spectroscopic techniques in the chapters that follow use instruments that share several common basic components: a source of
energy; a means for holding the sample of interest to us; a device that can isolate a narrow range of wavelengths; a detector for
measuring the signal; and a signal processor that displays the signal in a form convenient for the analyst. Figure 7.1.1 shows four
common ways of stringing together these units.

Figure 7.1.1 : Four common optical benches for spectroscopy. Each of (a), (b), and (c) includes a source of energy either in the form
of electromagnetic radiation (a, b) or thermal energy (c), a means for introducing the sample, a means for selecting the wavelengths
of electromagnetic radiation to monitor, a detector, and a signal processor. In (d), the source of energy comes from a chemical
reaction. The optical bench in (a) is suitable for absorbance measurements; the optical bench in (b) is suitable for fluorescence
measurements; the optical bench in (c) is suitable for emission measurements, and the optical bench in (d) is suitable for
chemiluminescence measurements.
The remaining sections of this chapter provide general information on each of these units. More specific details appear in the
chapters on individual methods.

This page titled 7.1: General Design of Optical Instruments is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

7.1.1 https://chem.libretexts.org/@go/page/366605
7.2: Sources of Radiation
All forms of spectroscopy require a source of energy to place the analyte in an excited state. In absorption and scattering
spectroscopy this energy is supplied by photons. Emission and photoluminescence spectroscopy use thermal energy, radiant
(photon) energy, or chemical energy to promote the analyte to a suitable excited state. In this section we consider the sources of
radiant energy.

Sources of Electromagnetic Radiation

A source of electromagnetic radiation must provide an output that is both intense and stable in the region of interest. Sources of
electromagnetic radiation are classified as either continuum or line sources. Table 7.2.1 provides a list of some common sources of
electromagnetic radiation.
Table 7.2.1 . Common sources of electromagnetic radiation.
source wavelength region useful for...

H2 and D2 lamp continuum source from 160–380 nm molecular absorption

tungsten lamp continuum source from 320–2400 nm molecular absorption

Xe arc lamp continuum source from 200–1000 nm molecular fluorescence

nernst glower continuum source from 0.4–20 µm molecular absorption

globar continuum source from 1–40 µm molecular absorption

nichrome wire continuum source from 0.75–20 µm molecular absorption

hollow cathode lamp line source in UV/Vis atomic absorption

Hg vapor lamp line source in UV/Vis molecular fluorescence

atomic and molecular absorption, fluorescence,

laser line source in UV/Vis/IR
and scattering

Continuum sources emits radiation over a broad range of wavelengths, with a relatively smooth variation in intensity (Figure
7.2.1), and are used for molecular absorbance using UV/Vis and IR radiation. Further details on these sources are in Chapters 13

and 16, respectively.

Figure 7.2.1 . Spectrum showing the emission from a green LED, which provides continuous emission over a wavelength range of
approximately 530–640 nm.
A line source, on the other hand, emits radiation at discrete wavelengths, with broad regions showing no emission lines (Figure
7.2.2), and are used for atomic absorption, atomic and molecular fluorescence, and Raman spectroscopy. Further details on hollow

cathode lamps are included in Chapter 9.

7.2.1 https://chem.libretexts.org/@go/page/366607
 Figure 7.2.2 . Emission spectrum from a Cu hollow cathode lamp. This spectrum consists of seven distinct emission lines (the first
two differ by only 0.4 nm and are not resolved at the scale shown in this spectrum). Each emission line has a width of
approximately 0.01 nm at 1⁄2 of its maximum intensity.

Laser Sources
An important line source of radiation is a laser, which is an acronym for light amplification by stimulated emission of radiation.
Laser emission is monochromatic with a narrow bandwidth of just a few micrometers. As suggested by the term amplification, a
laser provides a source of high intensity emission. The source of this intensity is embedded in the term stimulated emission, to
which we now turn our attention.

How a Laser Works

To understand how a laser works, we need to consider four key ideas: pumping, population inversion, stimulated emission, and
light amplification.
Pumping
Emission cannot occur unless we first populate higher energy levels with electrons, which we can accomplish by, for example, the
absorption of photons, as shown in Figure 7.2.3. Emission occurs when an electron in a higher energy state relaxes back to a lower
energy state by emitting a photon with an energy equal to the difference in the energy between these two states. The process of
populating the excited states with electrons is called pumping and is accomplished by using an electrical discharge, by passing an
electrical current through the lasing medium, or by absorption of high energy photons. The goal of pumping is to create a large
population of excited states.

Figure 7.2.3 . A general scheme showing the difference between absorption (on the left) and emission (on the right).
Population Inversion
Normally the majority of the species we are studying are in their ground electronic state with only a small number of species in an
excited electronic state. For a laser to achieve a high intensity of emission, it is necessary to create a situation in which there are
more species in the excited state than in the ground state, as shown in Figure 7.2.4 where the non-inverted population has four
species in the ground state and two species in the excited state, and where the inverted population has four species in the excited
state and two species in the ground state.

7.2.2 https://chem.libretexts.org/@go/page/366607
 Figure 7.2.4 . Illustration of (a) a non-inverted population and (b) an inverted population. In the non-inverted population there are
four electrons with energies of E1 and two electrons with energies of E2; for the inverted population there are four electrons with
energies of E2 and two electrons with energies of E1.
Stimulated Emission and Light Amplification
Figure 7.2.3 shows emission of a photon following absorption of a photon of equal energy. No more than one photon is emitted for
each photon that is absorbed, with some species in an excited state relaxing to the ground state through non-radiative pathways.
This spontaneous emission is a random process, which means that the timing of emission and the direction in which emission
occurs are random.
Emission in a laser, as depicted in Figure 7.2.5, is stimulated by a photon with an energy equal to that of the difference in energy
between the excited state and the ground state. The interaction of the incoming photon with the excited state results in the excited
state's immediate relaxation to the ground state by the emission of a photon. The original photon and the emitted photon are
coherent, with identical energies, identical directions, and identical phases. Because two coherent photons are emitted, the
amplitude of the emitted radiation is doubled, as we see in Figure 7.2.5; this is what we mean by light amplification.

Figure 7.2.5 . In stimulated emission, a photo with an energy equivalent to E − E stimulates the emission of a photon of equal
2 1

energy. The two photons are coherent with each other, producing an emission with an amplitude that is double that of the individual
photons.

Laser Systems
As the previous sections suggest, creating a population inversion is the limiting factor in generating radiation from a laser. The two-
level system in Figure 7.2.5, which involves a single excited state and a single ground state, cannot create a population inversion
because when the ground state and excited state are equal in population, the rate at which excited states are produced through
pumping equals the rate at which excited states are lost through emission. To achieve stimulated emission, laser systems use three-
level or four-level systems, as outlined in Figure 7.2.6.

7.2.3 https://chem.libretexts.org/@go/page/366607
 Figure 7.2.6 . Examples of (a) a three-level system and (b) a four-level system for generating emission from a laser.
In a three-level system, pumping is used to populate the excited states in level two. From level two, an efficient pathway for non-
radiative relaxation populates the excited state in level three, which is sufficiently stable to allow for a population inversion. In a
four-level system, the population inversion is achieved between level three and level four.

Types of Lasers
Lasers are categorized by the nature of the lasing medium: solid-state crystals, gases, dyes, and semiconductors. Solid-state lasers
use a crystalline material, such as aluminum oxide, that contains trace amounts of an element, such as chromium or neodymium,
which serves as the actual lasing medium. Gas lasers use gas phase atoms, ions, or molecules as a lasing medium. The lasing
medium in a dye laser is a solution of an organic dye molecule. A dye laser typically is capable of emitting light over a broad range
of wavelengths, but is tunable to a specific wavelength within that range. Finally, a semiconductor laser uses modified light-
emitting diodes as a lasing medium.
Table 7.2.2 . Examples of Lasers.
category lasing medium wavelengths

solid state ruby (0.05% Cr(III) in Al2O3) 694.3 nm

Nd:YAG (neodymium ion in yttrium

1054 nm; 532 nm
aluminum garnet

gas He/Ne 632.8 nm

Ar+ 514.5 nm, 488 nm

N2 337.1 nm

CO2 10.6 µm

dye rhodamine 540–680 nm

fluorescein 530–560 nm

cumarin 490–620 nm

semiconductor indium gallium nitride 405 nm

aluminum gallium indium phosphide 635 nm

This page titled 7.2: Sources of Radiation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

7.2.4 https://chem.libretexts.org/@go/page/366607
7.3: Wavelength Selectors
In Nessler’s original colorimetric method for ammonia, which was described at the beginning of the chapter, the sample and several
standard solutions of ammonia are placed in separate tall, flat-bottomed tubes. As shown in Figure 7.3.1, after adding the reagents
and allowing the color to develop, the analyst evaluates the color by passing ambient light through the bottom of the tubes and
looking down through the solutions. By matching the sample’s color to that of a standard, the analyst is able to determine the
concentration of ammonia in the sample.

Figure 7.3.1 . Nessler’s original method for comparing the color of two solutions. Natural light passes upwards through the samples
and standards and the analyst views the solutions by looking down toward the light source. The top view, shown on the right, is
what the analyst sees. To determine the analyte’s concentration, the analyst exchanges standards until the two colors match.
In Figure 7.3.1 every wavelength of light from the source passes through the sample. This is not a problem if there is only one
absorbing species in the sample. If the sample contains two components, then a quantitative analysis using Nessler’s original
method is impossible unless the standards contains the second component at the same concentration as in the sample.
To overcome this problem, we want to select a wavelength that only the analyte absorbs. Unfortunately, we can not isolate a single
wavelength of radiation from a continuum source, although we can narrow the range of wavelengths that reach the sample. As seen
in Figure 7.3.2, a wavelength selector always passes a narrow band of radiation characterized by a nominal wavelength, an
effective bandwidth, and a maximum throughput of radiation. The effective bandwidth is defined as the width of the radiation at
half of its maximum throughput.

Figure 7.3.2 . Radiation exiting a wavelength selector showing the band’s nominal wavelength and its effective bandwidth.
The ideal wavelength selector has a high throughput of radiation and a narrow effective bandwidth. A high throughput is desirable
because the more photons that pass through the wavelength selector, the stronger the signal and the smaller the background noise.
A narrow effective bandwidth provides a higher resolution, with spectral features separated by more than twice the effective
bandwidth being resolved. As shown in Figure 7.3.3, these two features of a wavelength selector often are in opposition. A larger
effective bandwidth favors a higher throughput of radiation, but provide less resolution. Decreasing the effective bandwidth
improves resolution, but at the cost of a noisier signal [Jiang, S.; Parker, G. A. Am. Lab. 1981, October, 38–43]. For a qualitative
analysis, resolution usually is more important than noise and a smaller effective bandwidth is desirable; however, in a quantitative
analysis less noise usually is desirable.

7.3.1 https://chem.libretexts.org/@go/page/366609
 Figure 7.3.3 . Example showing the effect of the wavelength selector’s effective bandwidth on resolution and noise. The spectrum
with the smaller effective bandwidth (on the right) has a better resolution, allowing us to see the presence of three peaks, but at the
expense of a noisier signal. The spectrum with the larger effective bandwidth (on the left) has less noise, but at the expense of less
resolution between the three peaks.

Filters
The simplest method for isolating a narrow band of radiation is to use an absorption or interference filter.

Absorption Filters
As their name suggests, absorption filters work by selectively absorbing radiation from a narrow region of the electromagnetic
spectrum. A simple example of an absorption filter is a piece of colored glass or polymer film. A purple filter, for example,
removes the complementary color green from 500–560 nm. Commercially available absorption filters provide effective bandwidths
of 30–250 nm, although the throughput at the low end of this range often is only 10% of the source’s emission intensity.
Interference filters are more expensive than absorption filters, but have narrower effective bandwidths, typically 10–20 nm, with
maximum throughputs of at least 40%. The latter value suggests that an important limitation to an absorption filter is that it may
significantly reduce the amount of light from the source that reaches the sample and the detector. Figure 7.3.4 shows an example of
a filter holder with filters that pass bands of light centered at 440 nm, 490 nm, or 550 nm.

Figure 7.3.4 : The photo on the left shows a filter holder with three filters in place. The photo on the right shows a filter that passes
green light centered at 550 nm and a filter that passes blue light centered at 440 nm.

Interference Filters
An interference filter consists of a transparent dielectric material, such as CaF2, which is sandwiched between two glass plates,
each coated with a thin, semitransparent metal film (7.3.5a). When a continuous source of light passes through the interference
filter it undergoes constructive and destructive interference that isolates and passes a narrow band of light centered at a wavelength
that satisfies Equation 7.3.1
2nb
λ = (7.3.1)
m

where n is the refractive index of the dielectric material, b is the thickness of the dielectric material, and m is the order of the
interference (typically first-order). Figure 7.3.5b shows the result of passing the emission from a green LED—a continuous source

7.3.2 https://chem.libretexts.org/@go/page/366609
that emits light from approximately 500 nm to 650 nm—through an interference filter that produces an effective bandwidth of a
few nanometers. In this case, a 210 nm thick film with a refractive index of 1.35 passes light centered at a wavelength of
2 × 1.35 × 210 nm
λ = = 567 nm
1

Figure 7.3.5 . The structure of an Interference filter is shown in (a) and consists of a thin dielectric film sandwiched between two
glass plates coated with a thin metal film that allows some of the source's light to pass through the film while reflecting the
remainder of the light. Because the light traverses different distances through the dielectric, only light with a wavelength that
satisfies Equation 7.3.1 passes through the filter, resulting in a narrow effective bandwith.

Monochromators
A filter has one significant limitation—because a filter has a fixed nominal wavelength, if we need to make measurements at two
wavelengths, then we must use two filters. A monochromator is an alternative method for selecting a narrow band of radiation that
also allows us to continuously adjust the band’s nominal wavelength. Monochromators are classified as either fixed-wavelength or
scanning. In a fixed-wavelength monochromator we select the wavelength by manually rotating the grating. Normally a fixed-
wavelength monochromator is used for a quantitative analysis where measurements are made at one or two wavelengths. A
scanning monochromator includes a drive mechanism that continuously rotates the grating, which allows successive wavelengths
of light to exit from the monochromator. A scanning monochromator is used to acquire a spectrum, and, when operated in a fixed-
wavelength mode, for a quantitative analysis.
The construction of a typical monochromator is shown in Figure 7.3.6. Radiation from the source enters the monochromator
through an entrance slit. The radiation is collected by a collimating mirror or lens, which focuses a parallel beam of radiation to a
diffraction grating (left) or a prism (right), that disperses the radiation in space. A second mirror or lens focuses the radiation onto a
planar surface that contains an exit slit. Radiation exits the monochromator and passes to the detector. As shown in Figure 7.3.6, a
monochromator converts a polychromatic source of radiation at the entrance slit to a monochromatic source of finite effective
bandwidth at the exit slit. The choice of which wavelength exits the monochromator is determined by rotating the diffraction
grating or prism. A narrower exit slit provides a smaller effective bandwidth and better resolution than does a wider exit slit, but at
the cost of a smaller throughput of radiation.

7.3.3 https://chem.libretexts.org/@go/page/366609
 Figure 7.3.6 . Examples of two types of monochromators: a diffraction grating monochromator an the left and a prism
monochromator on the right.

Polychromatic means many colored. Polychromatic radiation contains many different wavelengths of light. Monochromatic
means one color, or one wavelength. Although the light exiting a monochromator is not strictly of a single wavelength, its
narrow effective bandwidth allows us to think of it as monochromatic.

Monochromators Based on Prisms

Although prism monochromators were once in common use, they have mostly been replaced by diffraction gratings. There are
several reasons for this. One reason is that diffraction gratings are much less expensive to manufacture. A second reason is that a
diffraction grating provides a linear dispersion of of wavelengths along the focal plane of the exit slit, which means the resolution
between adjacent wavelengths is the same throughout the source's optical range. A prism, on the other hand, provides a greater
resolution at shorter wavelengths than it does a longer wavelengths.

Monochromators Based on Diffraction Gratings

The inset in the diffraction grating monochromator in Figure 7.3.6 shows the general saw-toothed pattern of a diffraction grating,
which consists of a series of grooves with broad surfaces exposed to light from the source. As shown in Figure 7.3.7, parallel
beams of source radiation (shown in blue) from the monochromator's collimating mirror strike the surface of the diffraction grating
and are reflected back (shown in green) toward the monochromator's focusing mirror and the detector. The parallel beams from the
source strike the diffraction grating at an incident angle i relative to the grating normal, which is a line perpendicular to the
diffraction grating's base. The parallel beams bounce back toward the detector do so at a reflected angle r to the grating normal.

Figure 7.3.7 . Illustration of how a monochromator's diffraction grating works to convert a beam of polychromatic light at the
monochromator's entrance slit into a monochromatic beam of light at the monochromator's exit slit. Rotating the diffraction grating,
changes the angles i and r and, therefore, the wavelength of light that exits the monochromator. The diffraction grating illustrated
here is known as an echellette-type grating.

7.3.4 https://chem.libretexts.org/@go/page/366609
Constructive interference between the reflected beams occurs if their path lengths differ by an integer multiple of the incident
beam's wavelength (nλ), where n is the diffraction order. A close examination of Figure 7.3.7 shows that the difference in the
distance traveled by two parallel beams of light, identified as 1 and 2, that strike adjacent grooves on the diffraction grating is equal
¯
¯¯¯¯¯¯
¯ ¯
¯¯¯¯¯¯
¯
to the sum of the line segments C B and BD, both shown in red; thus
¯
¯¯¯¯¯¯
¯ ¯
¯¯¯¯¯¯
¯
nλ = C B + BD (7.3.2)

The incident angle, i, is equal to the angle CAB and the reflected angle, r, is equal to the angle DAB, which means we can write
the following two equations
¯
¯¯¯¯¯¯
¯
C B = d sin i (7.3.3)

¯
¯¯¯¯¯¯
¯
BD = d sin r (7.3.4)

where d is the distance between the diffraction grating's grooves. Substituting back gives

nλ = d(sin i + sin r) (7.3.5)

which allows us to calculate the angle at which we can detect a wavelength of interest, r, given the angle of incidence from the
source, i, and the number of grooves per mm (or the distance between grooves).

 Example 7.3.1

At what angle can we detect light of 650 nm using a diffraction grating with 1500 gooves per mm if the incident radiation is at
an angle of 50 to the grating normal? Assume that this is a first-order diffraction.
∘

Solution
The distance between the grooves is
6
1 mm 10  nm
d = × = 666.7 nm
1500 grooves mm

To find the angle, we begin with

nλ = 1 × 650 nm = d(sin i + sin r) = 666.7 nm × (sin (50) + sin r

0.9750 = 0.7660 + sin r

0.2090 = sin r

∘
r = 12.1

Performance Characteristics of a Monochromator

The quality of a monochromator depends on several key factors: the purity of the light that emerges from the exit slit, the power of
the light that emerges from the exit slit, and the resolution between adjacent wavelengths.

Spectral Purity
The radiation that emerges from a monochromator is pure if it (a) arises from the source and if it (b) follows the optical path from
the entrance slit to the exit slit. Stray radiation that enters the monochromator from openings other than the entrance slit—perhaps
through small imperfections in the joints—or that reaches the exit slit after scattering from imperfections in the optical components
or dust, serves as a contaminant in that the power measured at the detector has a component at the monochromator's analytical
wavelength and a component from the stray radiation that includes radiation at other wavelengths.

Power
The amount of radiant energy that exits the monochromator and reaches the detector in a unit time is power. The greater the power,
the better the resulting signal-to-noise ratio. The more radiation that enters the monochromator and is gathered by the collimating
mirror, the greater the amount of radiation that exits the monochromator and the greater the power at the detector. The ability of a

7.3.5 https://chem.libretexts.org/@go/page/366609
monochromator to collect radiation is defined by its f /number. As shown in Figure 7.3.8, the smaller the f /number, the greater
the area and the greater the power. The light-gathering power increases as the inverse square of the f /number; thus, a
monochromator rated as f /2 gathers 4× as much radiation as a monochromator rated as f /4.

Figure 7.3.8 . Relationship between the light-gathering power of a monochromator and its f /number.

Resolution
To separate two wavelengths of light and detect them separately, it is necessary to to disperse them over a sufficient distance. The
angular dispersion of a monochromator is defined as the change in the angle of reflection (see the angle r in Figure 7.3.7) for a
change in wavelength, or dr/dλ. Taking the derivative of Equation 7.3.5 for a fixed angle of incidence (see the angle i in Figure
7.3.7) gives the angular dispersion as

dr n
= (7.3.6)
dλ d cos r

where n is the diffraction order. The linear dispersion of radiation, D, gives the change in wavelength as a function of y , the
distance along the focal plane of the monochromator's exit slit; this is related to the angular dispersion by
dy F dr
D = = (7.3.7)
dλ dλ

where F is the focal length. Because we are interested in wavelength, it is convenient to take the inverse of Equation 7.3.7

−1
dλ 1 dλ
D = = × (7.3.8)
dy F dr

where D −1
is the reciprocal linear dispersion. Substituting Equation 7.3.6 into Equation 7.3.8 gives
dλ d cos r
−1
D = = (7.3.9)
dy nF

which simplifies to
d
−1
D = (7.3.10)
nF

for angles r < 20 where cos r ≈ 1 . Because the linear dispersion of radiation along the monochromator's exit slit is independent
∘

of wavelength, the ability to resolve two wavelengths is the same across the spectrum of wavelengths.
Another way to report a monochromator's ability to distinguish between two closely spaced wavelengths is its resolving power, R ,
which is defined as
λ
R = = nN
Δλ

where λ is the average of the two wavelengths, Δλ is the difference in their values and N is the number of grooves on the
diffraction grating that are exposed to the radiation from the collimating mirror. The greater the number of grooves, the greater the
resolving power.

Monochromator Slits
A monochromator has two sets of slits: an entrance slit that brings radiation from the source into the monochromator and an exit
slit that passes the radiation from the monochromator to the detector. Each slit consists of two metal plates with sharp, beveled
edges separated by a narrow gap that forms a rectangular window and which is aligned with the focal plane of the collimating
mirror. Figure 7.3.9 shows a set of four slits from a monochromator taken from an atomic absorption spectrophotometer. From
bottom-to-top, the slits have widths, w, of 2.0 mm, 1.0 mm, 0.5 mm, and 0.2 mm.

7.3.6 https://chem.libretexts.org/@go/page/366609
 Figure 7.3.9 . Photo showing a set of four entrance slits to a monochromator from an atomic absorption spectrometer. From bottom-
to-top they have widths of 2.0 mm, 1.0 mm, 0.5 mm, and 0.2 mm. The monochromator's exit slits are identical and. For this
monochromator, the operator selects the slit width for the entrance slits and the exit is automatically set to the same width.

Effect of Slits on Monochromatic Radiation

Suppose we have a source of monochromatic radiation with a wavelength of 400.0 nm and that we pass this beam of radiation
through a monochromator that has entrance and exit slits with a width, w, of 1.0 mm and a reciprocal linear dispersion of 1.2
nm/mm. The product of these two variables is called the monochromator's effective bandwidth, Δλ , and is given as
eff

−1
Δλeff = w D = 1.0 mm × 1.2 nm/mm = 1.2 nm (7.3.11)

The width of the beam in units of wavelength, therefore, is 1.2 nm. In this case, as shown in Figure 7.3.10, if we scan the
monochromator, our beam of monochromatic radiation will first enter the exit slit at a wavelength setting of 398.8 nm and will
fully exit the slit at a wavelength setting of 401.2 nm. In between these limits a portion of the beam is blocked and only a portion of
the beam passes through the exit slit and reaches the detector. For example, when the monochromator is set to 399.4 nm or 400.6
nm, half of ther beam reaches the detector with a power of 0.5 × P . If we monitor the power at the detector as a function of
wavelength, we obtain the profile shown at the bottom of Figure 7.3.10. The monochromator's bandwidth encompasses the range
of wavelengths over which some portion of the beam of radiation passes through the exit slit.

Figure 7.3.10 : Illustration showing the relationship between effective bandwidth of a monochromatic source of radiation and the
width of a monochromator's exit slit.

Effect of Slit Width on Resolution

Suppose we have a source of radiation that consists of precisely three wavelengths—399.4 nm, 400.0 nm, and 400.6 nm—and we
pass them through a monochromator with an effective bandwidth of 1.2 nm. Using the analysis from the previous section, the
radiation with a wavelength of 399.4 nm passes through the monochromator's exit slit for any wavelength setting between 398.8
and 400.0 nm, which means it overlaps with the radiation with a wavelength of 400.0 nm. The same is true for the radiation with a
wavelength of 400.6 nm, which also overlaps with the radiation with a wavelength of 400.0 nm. As shown in Figure 7.3.11a, we

7.3.7 https://chem.libretexts.org/@go/page/366609
cannot resolve the three monochromatic sources of radiation, which appear as a single broad band of radiation. Decreasing the
effective bandwidth to one-half of the difference in the wavelengths of the adjacent sources of radiation produces, as shown in
Figure 7.3.11b, baseline resolution of the individual sources of wavelength. To resolve the sources of radiation with wavelengths of
399.4 nm and 400.0 nm using a monochromator with a reciprical linear dispersion of 1.2 nm/mm requires an effective bandwidth
of

Δλeff = 0.5 × (400.0 nm − 399.4 nm) = 0.3 nm

and a slit width of

Δλeff 0.3 nm
w = = = 0.25 mm
−1
D 1.2nm/mm

Figure 7.3.11 : Illustration showing (a) the inability of a monochromator with an effective bandwidth of 1.2 nm to resolve three
lines with wavelengths of 399.4 nm, 400.0 nm, and 400.6 nm. Decreasing the effective bandwidth to 0.3 nm, as in (b), allows for
complete resolution of the three lines.

Choosing a Slit Width

The choice of slit width always involves a trade-off between increasing the radiant power that reaches the detector by using a wide
slit width, which improves the signal-to-noise ratio, and improving the resolution between closely spaced peaks, which requires a
narrow slit width. Figure 7.3.3 illustrates this trade-off. Ultimately, the needs of the analyst will dictate the choice of slit width.

This page titled 7.3: Wavelength Selectors is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

7.3.8 https://chem.libretexts.org/@go/page/366609
7.4: Sample Containers
The sample compartment provides a light-tight environment that limits stray radiation. Samples normally are in a liquid or solution
state, and are placed in cells constructed with UV/Vis transparent materials, such as quartz, glass, and plastic (Figure 7.4.1). A
quartz or fused-silica cell is required when working at a wavelength <300 nm where other materials show a significant absorption.
The most common pathlength is 1 cm (10 mm), although cells with shorter (as little as 0.1 cm) and longer pathlengths (up to 10
cm) are available. Longer pathlength cells are useful when analyzing a very dilute solution or for gas samples. The highest quality
cells allow the radiation to strike a flat surface at a 90o angle, minimizing the loss of radiation to reflection. A test tube often is used
as a sample cell with simple, single-beam instruments, although differences in the cell’s pathlength and optical properties add an
additional source of error to the analysis.

Figure 7.4.1 . Examples of sample cells for UV/Vis spectroscopy. From left to right (with path lengths in parentheses): rectangular
plastic cuvette (10.0 mm), rectangular quartz cuvette (5.000 mm), rectangular quartz cuvette (1.000 mm), cylindrical quartz cuvette
(10.00 mm), cylindrical glass cuvette with quartz windows (100.0 mm). Cells often are available as a matched pair, which is
important when using a double-beam instrument.
Infrared spectroscopy routinely is used to analyze gas, liquid, and solid samples. Sample cells are made from materials, such as
NaCl and KBr, that are transparent to infrared radiation. Gases are analyzed using a cell with a pathlength of approximately 10 cm.
Longer pathlengths are obtained by using mirrors to pass the beam of radiation through the sample several times.
A liquid sample may be analyzed using a variety of different sample cells (Figure 7.4.2). For non-volatile liquids a suitable sample
is prepared by placing a drop of the liquid between two NaCl plates, forming a thin film that typically is less than 0.01 mm thick.
Volatile liquids are placed in a sealed cell to prevent their evaporation.

Figure 7.4.2 . Three examples of IR sample cells: (a) NaCl salts plates; (b) fixed pathlength (0.5 mm) sample cell with NaCl
windows; (c) disposable card with a polyethylene window that is IR transparent with the exception of strong absorption bands at
2918 cm–1 and 2849 cm–1.

This page titled 7.4: Sample Containers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

7.4.1 https://chem.libretexts.org/@go/page/366611
7.5: Radiation Transducers
Introduction
In Nessler’s original method for determining ammonia (see Section 7.4) the analyst’s eye serves as the detector, matching the
sample’s color to that of a standard. The human eye, of course, has a poor range—it responds only to visible light—and it is not
particularly sensitive or accurate. Modern detectors use a sensitive transducer to convert a signal consisting of photons into an
easily measured electrical signal. Ideally the detector’s signal, S, is a linear function of the electromagnetic radiation’s power, P,
S = kP + D (7.5.1)

where k is the detector’s sensitivity, and D is the detector’s dark current, or the background current when we prevent the source’s
radiation from reaching the detector. There are two broad classes of spectroscopic transducers: photon transducers and thermal
transducers, although we will subdivide the photon transducers given their rich variety. Table 7.5.1 provides several representative
examples of each class of transducers.

Transducer is a general term that refers to any device that converts a chemical or a physical property into an easily measured
electrical signal. The retina in your eye, for example, is a transducer that converts photons into an electrical nerve impulse;
your eardrum is a transducer that converts sound waves into a different electrical nerve impulse.

Table 7.5.1 . Examples of Transducers for Spectroscopy

transducer class wavelength range output signal

photovoltaic cell photon 350–750 nm current

phototube photon 200–1000 nm current

photomultiplier photon 110–1000 nm current

Si photodiode photon 250–1100 nm current

photoconductor photon 750–6000 nm change in resistance

photovoltaic cell photon 400–5000 nm current or voltage

thermocouple thermal 0.8–40 µm voltage

thermistor thermal 0.8–40 µm change in resistance

pneumatic thermal 0.8–1000 µm membrane displacement

pyroelectric thermal 0.3–1000 µm current

Photon Transducers
A photon transducer takes a photon and converts it into an electrical signal, such as a current, a change in resistance, or a voltage.
Many such detectors use a semiconductor as the photosensitive surface. When the semiconductor absorbs photons, valence
electrons move to the semiconductor’s conduction band, producing a measurable current.

Photovoltaic Cells
A photovoltaic cell (Figure 7.5.1) consists of a thin film of a semiconducting material, such as selenium sandwiched between two
electrodes: a base electrode of iron or copper and a thin semi-transparent layer of silver or gold that serves as the collector
electrode. When a photon of visible light falls on the photovoltaic cell it generates an electron and a hole with a positive charge
within the semiconductor. Movement of the electrons from the collector electrode to the base electrode generates a current that is
proportional to the power of the incoming radiation and that serves as the signal.

7.5.1 https://chem.libretexts.org/@go/page/366613
 Figure 7.5.1 : Schematic diagram of a photovoltaic cell that consists of a semiconducting thin film of selenium placed between two
electrodes. When a photon strikes the semiconductor, it generates an electron and a hole that carries a positive charge. The
movement of the electron from the collector electrode to the base electrode generates a measureable current.

Phototubes and Photomultipliers

Phototubes and photomultipliers use a photosensitive surface that absorbs radiation in the ultraviolet, visible, or near IR to produce
an electrical current that is proportional to the number of photons that reach the transducer (see Figure 7.5.2). The current results
from applying a negative potential to the photoemissive surface and a positive potential to a wire that serves as the anode. In a
photomultiplier tube, a series of positively charged dynodes serves to amplify the current, producing 106–107 electrons per photon.

Figure 7.5.2 .Schematic of a phototube (left) and a photomultiplier tube (right). In both, a photon strikes the photoemissive cathode
producing electrons. In the phototube, the electrons that are drawn toward a positively charged anode, generating a current. In the
photomultiplier tube, the electrons accelerate toward a positively charged dynode. Collision of these electrons with the dynode
generates additional electrons, which accelerate toward the next dynode. A total of 106–107 electrons per photon eventually reach
the anode, generating an electrical current.

Silicon Photodiodes
Applying a reverse biased voltage to the pn junction of a silicon semiconductor creates a depletion zone in which conductance is
close to zero (see Chapter 2 for an earlier discussion of semiconductors). When a photon of light of sufficient energy impinges on
the depletion zone, an electron-hole pair is formed. Movement of the electron through the n–region and of the hole through the p–
region generates a current that is proportional to the number of photons reaching the detector. A silicon photodiode has a wide
spectral range from approximately 190 nm to 1100 nm, which makes it versatile; however, a photodiode is less sensitive than a
photomultiplier.

Figure 7.5.3 : Illustration of silicon photodiode. See the text for a description of how the deterctor works.

7.5.2 https://chem.libretexts.org/@go/page/366613
Multichannel Photon Transducers
The photon transducers discussed above detect light at a single wavelength passed by the monochromator to the detector. If we
wish to record a complete spectrum then we must continually adjust the monochromator either manually or by using a servo motor.
In a multichannel instrument we create a one-dimensional or two-dimensional array of detectors that allow us to monitor
simultaneously radiation spanning a broad range of wavelengths.

Photodiode Arrays
An individual silicon photodiode is quite small, typically with a width of approximately 0.025 mm. As a result, a linear (one-
dimensional) array that consists of 1024 individual photodiodes has a width of just 25.6 mm. Figure 7.5.4, for example, shows the
UV detector from an HPLC. Light from the deuterium lamp passes through a flow cell, is dispersed by a diffraction grating, and
then focused onto a linear array of photodiodes. The close-up on the right shows the active protion of the photodiode array covered
by an optical window. The active width of this photodiode array is approximately 6 mm and includes more than 200 individual
photodiodes, sufficient to provide 1 nm resolution from 180 nm to 400 nm.

Figure 7.5.4 : The photo on the left shows the UV detector from an HPLC and the photo on the right shows a close-up view of the
its linear photodiode array deterctor. The bright line, from reflected light, is the photodiode array. The active portion of this detector
is covered by an optically transparent window.

Charge-Transfer Devices
One way to increase the sensitivity of a detector is to collect and store charges before counting them. This is the approach taken
with two types of charge-transfer devices: charged-coupled detectors and charge-injection detectors. Individual detectors, or pixels,
consist of a layer of silicon dioxide coated on top of semiconductor. When a photon impinges on the detector it creates an electron-
hole pair. An electrode on top of the silicon dioxide layer collects and stores either the negatively charged electrons or the
positively charged holes. After a sufficient time, during which 10,000-100,000 charges are collected, the total accumulated charge
is measured. Because individual pixels are small, typically 10 µm, they can be arranged in either a linear, one-dimensional array or
a two-dimensional array. A charge-transfer device with 1024 x 1024 pixels will be approximately 10 mm x 10 mm in size.

7.5.3 https://chem.libretexts.org/@go/page/366613
  Note
There are two important charge-transfer devices used as detectors: a charge-coupled device (CCD), which is discussed below,
and a charge-injection device (CID), which is discussed in Chapter 10. Both types of devices use a two-dimensional arrays of
individual detectors that store charge. The two devices differ primarily in how the accumulated charges are read.

Figure 7.5.5 shows a cross-section of a single detector (pixel) in a charge-coupled device (CCD) where individual pixels are
arranged in a two-dimensional array. Electron-hole pairs are created in a layer of p-doped silicon. The holes migrate to the n-doped
silicon layer and the electrons are drawn to the area below a positively charged electrode. When it is time to record the accumulated
charges, the charge is read in the upper-right corner of the array with charges in the same row measured by shifting them from left-
to-right. When the first row is read, the charges in the remaining rows are shifted up and recorded. In a charge-injection device, the
roles of the electrons and holes are reversed and the accumulated positive charged are recorded.

Figure 7.5.5 : Schematic diagram of a charge-coupled device (CCD). The figure on the left shows a single detector, or pixel. When
a photon impinges on the detector it generates an electron-hole pair in the p-doped layer. The electrons migrate and accumulate
beneath a positively charged electrode. Individual pixels are arranged in an array, as shown on the right. This array may be a single
row (a linear, or one-dimensional array) or in a series of rows. The charge in each pixels is recorded by shifting the pixels from left-
to-right and by moving rows of pixels up.
Figure 7.5.6 shows an example of spectrophotometer equipped with a linear CCD detector that includes 2048 individual elements
with a wavelength range from 200 nm to 1100 nm. The spectrometer is housed in a compact space of 90 mm x 60 mm

Figure 7.5.6 : Example of a spectrophotometer equipped with a linear CCD detector.

Thermal Transducers
Infrared photons do not have enough energy to produce a measurable current with a photon transducer. A thermal transducer,
therefore, is used for infrared spectroscopy. The absorption of infrared photons increases a thermal transducer’s temperature,

7.5.4 https://chem.libretexts.org/@go/page/366613
changing one or more of its characteristic properties. A pneumatic transducer, for example, is a small tube of xenon gas with an IR
transparent window at one end and a flexible membrane at the other end. Photons enter the tube and are absorbed by a blackened
surface, increasing the temperature of the gas. As the temperature inside the tube fluctuates, the gas expands and contracts and the
flexible membrane moves in and out. Monitoring the membrane’s displacement produces an electrical signal.

This page titled 7.5: Radiation Transducers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

7.5.5 https://chem.libretexts.org/@go/page/366613
7.6: Fiber Optics
If we need to monitor an analyte’s concentration over time, it may not be possible to remove samples for analysis. This often is the
case, for example, when monitoring an industrial production line or waste line, when monitoring a patient’s blood, or when
monitoring an environmental system, such as stream. With a fiber-optic probe we can analyze samples in situ. An example of a
remote sensing fiber-optic probe is shown in Figure 7.6.1. The probe consists of two bundles of fiber-optic cable. One bundle
transmits radiation from the source to the probe’s tip, which is designed to allow the sample to flow through the sample cell.
Radiation from the source passes through the solution and is reflected back by a mirror. The second bundle of fiber-optic cable
transmits the nonabsorbed radiation to the wavelength selector. Another design replaces the flow cell shown in Figure 7.6.1 with a
membrane that contains a reagent that reacts with the analyte. When the analyte diffuses into the membrane it reacts with the
reagent, producing a product that absorbs UV or visible radiation. The nonabsorbed radiation from the source is reflected or
scattered back to the detector. Fiber optic probes that show chemical selectivity are called optrodes.

Figure 7.6.1 : Example of a fiber-optic probe. The inset photographs at the bottom of the figure provide close-up views of the
probe’s flow cell and the reflecting mirror.

This page titled 7.6: Fiber Optics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

7.6.1 https://chem.libretexts.org/@go/page/366615
7.7: Fourier Transform Optical Spectroscopy
Thus far, the optical benches described in this chapter either use a single detector and a monochromator to pass a single wavelength
of light to the detector, or use a multichannel array of detectors and a diffraction grating to disperse the light across the detectors.
Both of these approaches have advantages and limitations. For the first of these designs, we can improve resolution by using a
smaller slit width, although this comes with a decrease in the throughput of light that reaches the detector, which increases noise.
Recording a complete spectrum requires scanning the monochromator; a slow scan rate can improve resolution by reducing the
range of wavelengths reaching the detector per unit time, but at the expense of a longer analysis time, which is a problem if the
composition of our samples changes with time. For the second of these designs, resolution is limited by the size of the array; for
example, a spectral range of 190 nm to 800 nm and a photodiode array with 512 individual elements has a digital resolution of
800 − 190
= 1.2 nm/diode
512

although the optical resolution—defined by the actual number of individual diodes over which a wavelength of light is dispersed—
is greater and may vary with wavelength. Because a photodiode array allows for the simultaneous detection of radiation by each
diode in the array, data acquisition is fast and a complete spectrum is acquired in approximately one second.

Interferometers
We can overcome the limitations described above if we can find a way to avoid dispersing the source radiation in time by scanning
the monochromator, or dispersing the source radiation in space across an array of sensors. An interferometer, Figure 7.7.1, provides
one way to accomplish this. Radiation from the source is collected by a collimating mirror and passed to a beam splitter where half
of the radiation is directed toward a mirror set at a fixed distance from the beam splitter, and the other half of the radiation is passed
through to a mirror that moves back and forth. The radiation from the two mirrors is recombined at the beam splitter and half of it
is passed along to the detector.

Figure 7.7.1 : Schematic diagram of an interferometer for use in optical spectroscopy. The distance of the fixed mirror and the
moving mirror from the beam splitter is the same when the moving mirror is in the position shown in black.

Time Domain and Frequency Domain

When the radiation recombines at the beam splitter, constructive and destructive interference determines, for each wavelength, the
intensity of light that reaches the detector. As the moving mirror changes position, the wavelength of light that experiences
maximum constructive interference and maximum destructive interference also changes. The signal at the detector shows intensity
as a function of the moving mirror’s position, expressed in units of distance or time. The result is called an interferogram or a time
domain spectrum. The time domain spectrum is converted mathematically, by a process called a Fourier transform, to a spectrum (a
frequency domain) that shows intensity as a function of the radiation’s frequency.

7.7.1 https://chem.libretexts.org/@go/page/366617
Figure 7.7.2 shows the relationship between the time domain spectrum and the frequency domain spectrum. The spectra in the first
row show the relationship between (a) the time domain spectrum and (b) the corresponding frequency domain spectrum for a
monochromatic source of radiation with a frequency, ν , of 1 and an amplitude, A , of 1.0. In the time domain we see a simple
1 1

cosine function with the general form

S = A1 × cos (2π ν1 t) (7.7.1)

where S is the signal and t is the time. The spectra in the second row show the same information for a second monochromatic
source of radiation with a frequency, ν , of 1.2 and an amplitude, A , of 1.5, which is given by the equation
2 2

S = A2 × cos (2π ν2 t) (7.7.2)

If we have a source that emits just these two frequencies of light, then the corresponding time domain and frequency domain
spectra in the last row, where
S = A1 × cos (2π ν1 t) + A2 × cos (2π ν2 t) (7.7.3)

Although the time domain spectrum in panel (e) is more complex than those in panels (a) and (c), there is a clear repeating pattern,
one cycle of which is shown by the arrow. Note that for each of these three examples, the time domain spectrum and the frequency
domain spectrum encode the same information about the source radiation.

Figure 7.7.2 : Time domain and frequency domain spectra for (a) and (b) a monochromatic source with a frequency of ν and an
1

amplitude of A , for (c) and (d) a monochromatic source with a frequency of ν and an amplitude of A , and (e) and (f) a
1 2 2

polychromatic source that consists of the two monochromatic sources.

The two monochromatic signals in Figure 7.7.2 are line spectra with line widths that are essentially zero. But what if our signal has
a measurable linewidth? We might consider such a signal to be the sum of a series of cosine functions, each with an amplitude and
a frequency. Figure 7.7.3a shows a frequency domain that contains a single peak with a finite width and Figure 7.7.3b shows the
corresponding time domain spectrum, which consists of an oscillating signal with an amplitude that decays over time. In general,
Figure 7.7.2 and Figure 7.7.3 show that
the further a peak in the frequency domain is from the origin, the greater its corresponding oscillation frequency in the
time domain
the broader a peak's width in the frequency domain, the faster its decay rate in the time domain
the greater the area under a peak in the frequency domain, the higher its initial intensity in the time domain

7.7.2 https://chem.libretexts.org/@go/page/366617
 Figure 7.7.3 : The plot in (a) shows a frequency domain consisting of a single peak defined by its position along the x-axis, its
width, and its area. The plot in (b) shows the corresponding time domain that consists of a single oscillating signal defined by its
oscillation frequency, its initial intensity, and its decay rate.
The mathematical process of converting between the time domain and the frequency domain is called a Fourier transform. The
details of the mathematics are sufficiently complex that calculations by hand are impractical.

Advantages of Fourier Transform Spectrometry

In comparison to a monochromator, an interferometer has several significant advantages. The first advantage, which is termed
Jacquinot’s advantage, is the greater throughput of source radiation. Because an interferometer does not use slits and has fewer
optical components from which radiation is scattered and lost, the throughput of radiation reaching the detector is 80 − 200×
greater than that for a monochromator. The result is less noise. A second advantage, which is called Fellgett’s advantage, is a
savings in the time needed to obtain a spectrum. Because the detector monitors all frequencies simultaneously, a spectrum takes
approximately one second to record, as compared to 10–15 minutes when using a scanning monochromator. A third advantage is
that increased resolution is achieved by increasing the distance traveled by the moving mirror, which we can achieve without the
need to decrease a scanniing monochromator's slit width or without increasing the size of an array detector.

This page titled 7.7: Fourier Transform Optical Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

7.7.3 https://chem.libretexts.org/@go/page/366617
 CHAPTER OVERVIEW
8: An Introduction to Optical Atomic Spectroscopy
8.1: Optical Atomic Spectra
8.2: Atomization Methods
8.3: Sample Introduction Methods

This page titled 8: An Introduction to Optical Atomic Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

1
8.1: Optical Atomic Spectra
Energy Level Diagrams
The energy of ultraviolet and visible electromagnetic radiation is sufficient to cause a change in an atom’s valence electron
configuration. Sodium, for example, has a single valence electron in its 3s atomic orbital. As shown in Figure 8.1.1, unoccupied,
higher energy atomic orbitals also exist. The valence shell energy level diagram in Figure 8.1.1 might strike you as odd because it
shows the 3p orbitals split into two groups of slightly different energy (the two lines differ by just 0.6 nm). The cause of this
splitting is a consequence of an electron's angular momentum and its spin. When these two are in the opposite direction, then the
energy is slightly smaller than when the two are in the same direction. The effect is largest for p orbitals and sufficiently smaller for
d and f orbitals that we do not bother to show the difference in their energies in this diagram.

Figure 8.1.1 . Valence shell energy level diagram for sodium. The wavelengths (in nm) corresponding to several transitions are
shown.
Absorption of a photon is accompanied by the excitation of an electron from a lower-energy atomic orbital to an atomic orbital of
higher energy. Not all possible transitions between atomic orbitals are allowed. For sodium the only allowed transitions are those in
which there is a change of ±1 in the orbital quantum number (l); thus transitions from s → p orbitals are allowed, but transitions
from s → s and from s → d orbitals are forbidden.

Atomic Absorption Spectra

The atomic absorption spectrum for Na is shown in Figure 8.1.2, and is typical of that found for most atoms. The most obvious
feature of this spectrum is that it consists of a small number of discrete absorption lines that correspond to transitions between the
ground state (the 3s atomic orbital) and the 3p and the 4p atomic orbitals. Absorption from excited states, such as the 3p → 4s and
the 3p → 3d transitions included in Figure 8.1.1, are too weak to detect. Because an excited state’s lifetime is short—an excited
state atom typically returns to a lower energy state in 10–7 to 10–8 seconds—an atom in the exited state is likely to return to the
ground state before it has an opportunity to absorb a photon.

Figure 8.1.2 . Atomic absorption spectrum for sodium. Note that the scale on the x-axis includes a break.

8.1.1 https://chem.libretexts.org/@go/page/366310
Atomic Emission Spectra
Atomic emission occurs when electrons in higher energy orbitals return to a lower energy state, releasing the excess energy as a
photon. The ground state electron configuration for Na of 1s 2s 2p 3s places a single electron in the 3s valence shell.
2 2 6 1

Introducing a solution of NaCl to a flame results in the formation of Na atoms (more on this in Chapter 9) and provides sufficient
energy to promote the valence electron in the 3s orbital to higher energy excited states, such as the 3p orbitals identified in the
energy level diagram for sodium in Figure 8.1.1. When an electron returns to its ground state, the excess energy is released as a
photon. As seen in Figure 8.1.3, the emission spectrum for Na is dominated by the pair of lines with wavelengths of 589.0 and
589.6 nm.

Figure 8.1.3 : Atomic emission spectrum for Na at a temperature of 3000 K. The inset shows that there are two closely spaced
emission lines with wavelengths of 589.0 and 589.6 nm. The spectra were simulated using the R package atomicEmission.

Atomic Fluorescence Spectra

When an atom in an excited state emits a photon as a means of returning to a lower energy state, how we describe the process
depends on the source of energy creating the excited state. When excitation is the result of thermal energy, as is the case for the
spectrum in Figure 8.1.3, we call the process atomic emission spectroscopy. When excitation is the result of the absorption of a
photon, we call the process atomic fluorescence spectroscopy. The absorption spectrum for Na in Figure 8.1.2 and its emission
spectrum in Figure 8.1.2 shows that Na has both strong absorption and emission lines at 589.0 and 589.6 nm. If we use a source of
light at 589.6 nm to move the 3s valence electron to a 3p excited state, we can then measure the emission of light at the same
wavelength, making the measurement at 90° to avoid an interference from the original light source.
Fluorescence also may occur when an electron in an excited state first loses energy by a process other than the emission of a photon
—we call this a radiationless transition—reaching a lower energy excited state from which it then emits a photon. For example, a
ground state Na atom may first absorb a photon with a wavelength of 330.2 nm (a 3s → 4p transition), which then loses energy
through a radiationless transition to the 3p orbital where it then emits a photon to reach the 3s orbital.

Atomic Line Widths

Another feature of the atomic absorption spectrum in Figure 8.1.2 and the atomic emission spectrum in Figure 8.1.3 is the narrow
width of the absorption and emission lines, which is a consequence of the fixed difference in energy between the ground state and
the excited state, and the lack of vibrational and rotational energy levels. The width of an atomic absorption or emission line arises
from several factors that we consider here.

Broadening Due to Uncertainty Principle

From the uncertainty principle, the product of the uncertainty of the frequency of light and the uncertainty in time must be greater
than 1.

Δν × Δt > 1

To determine the frequency with infinite precision, Δν = 0 , requires that the lifetime of an electron in a particular orbital must be
infinitely large. While this may be essentially true for an electron in the ground state, it is not true for an electron in an excited state

8.1.2 https://chem.libretexts.org/@go/page/366310
where the average lifetime—how long it takes before it returns to the ground state—may be on the order of 10
−7 −8
 to 10  s . For
example, if Δt = 5 × 10  s for the emission of a photon with a wavelength of 500.0 nm, then
−8

7 −1
Δν = 2 × 10  s

To convert this an uncertainty in wavelength, Δλ, we begin with the relationship

c
ν =
λ

and take the derivative of ν with respect to wavelength

c
dν = − dλ
2
λ

Rearranging to solve for the uncertainty in wavelength, and letting Δν and Δλ serve as estimates for dν and dλ leaves us with
2 −9 2 7 −1
Δν × λ (500.0 × 10  m ) × (2 × 10 s )
−14
|Δλ| = = = 1.7 × 10  m
8
c 2.998 × 10  m/s

or 1.7 × 10−5
 nm . Natural line widths for atomic spectra are approximately 10–5 nm.

Doppler Broadening and Pressure Broadening

When an atom emits a photon, the frequency (and, thus, the wavelength) of the photon depends on whether the emitting atom is
moving toward the detector or moving away from the detector. When the atom is moving toward the detector, as in Figure 8.1.4a,
its emitted light reaches the detector at a greater frequency—a shorter wavelength—than when the light source is stationary, as in
Figure 8.1.4b. An atom moving away from the detector, as in Figure 8.1.4c emits light that reaches the detector with a smaller
frequency and a longer wavelength.

Figure 8.1.4 : Illustration showing the original of Doppler broadening. In (a) and (c) the atom is moving toward or away from the
detector, respectively. In (b), the atom is stationary relative to the detector.
Atoms are in constant motion, which means that they also experience constant collisions, each of which results in a small change in
the energy of an electron in the ground state or in an excited state, and a corresponding change in the wavelength emitted or
absorbed. This effect is called pressure (or collisional) broadening. As is the case for Doppler broadening, pressure broadening
increases with temperature. Together, Doppler broadening and pressure broadening result in an approximately 100-fold increase in
line width, with line widths on the order of approximately 10–3 nm.

Effect of Temperature on Atomic Spectra

As noted in the previous section, temperature contributes to the broadening of atomic absorption and atomic emission lines.
Temperature also has an effect on the intensity of emission lines as it determines the relative population of an atom's various

8.1.3 https://chem.libretexts.org/@go/page/366310
excited states. The Boltzmann distribution
Ni Pi −Ei /kT
= e
N0 P0

gives the relative number of atoms in a specific excited state, N , relative to the number of atoms in the ground state, N , as a
i 0

function of the difference in their energies, E , Boltzmann's constant, k , the temperature in Kelvin, T , and P and P are statistical
i i 0

factors that account for the number of equivalent energy states for the excited state and the ground state. Figure 8.1.5 shows how
temperature affects the atomic emission spectrum for sodium's two intense emission lines at 589.0 and 589.6 nm for temperatures
from 2500 K to 7500 K. Note that the emission at 2500 K is too small to appear using a y-axis scale of absolute intensities. A
change in temperature from 5500 K to 4500 K reduces the emission intensity by 62%. As you might guess from this, a small
change in temperature—perhaps as little as 10 K can result in a measurable decrease in emission intensity of a few percent.

Figure 8.1.5 : Atomic emission spectra for Na as a function of temperature. The spectra were simulated using the R package
atomicEmission.
An increase in temperature may also change the relative emission intensity of different lines. Figure 8.1.6, for example, shows the
atomic emission spectra for copper at 5000 K and 7000 K. At the higher temperature, the most intense emission line changes from
510.55 nm to 521.82 nm, and several additional peaks between 400 nm and 500 nm become more intense.

Figure 8.1.6 : Atomic emission spectra for Cu at 5000 K and 7000 K. The spectra were simulated using the R package
atomicEmission.

8.1.4 https://chem.libretexts.org/@go/page/366310
Band and Continuum Spectra
The atomic emission spectra for sodium in Figure 8.1.3 consists of discrete, narrow lines because they arise from the transition
between the discrete, well-defined energy levels seen in Figure 8.1.1. Atomic emission from a flame also include contributions
from two additional sources: the emission from molecular species that form in the flame and emission from the flame. A sample of
water, for example, is likely to contain a variety of ions, such as Ca2+, that form molecular species, such as CaOH in the flame, and
that emit photon over a much broader range of wavelengths than do atoms. The flame, itself, emits photons throughout the range of
wavelengths used in UV/Vis atomic emission.

This page titled 8.1: Optical Atomic Spectra is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

8.1.5 https://chem.libretexts.org/@go/page/366310
8.2: Atomization Methods
Atomic methods require that the sample consist of individual gas phase atoms or gas phase atomic ions. With rare exceptions, this
is not the form in which we obtain samples. If we are interested in analyzing seawater for the concentration of sodium, we need to
find a way to convert the solution of aqueous sodium ions, Na+(aq), into gas phase sodium atoms, Na(g), or gas phase sodium ions,
Na+(g). The process by which this happens is called atomization and requires a source of thermal energy. Examples of atomization
methods include the use of flames, resistive heating, plasmas, and electric arcs and sparks. More details on specific atomization
methods appear in the chapters that follow.

This page titled 8.2: Atomization Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

8.2.1 https://chem.libretexts.org/@go/page/366312
8.3: Sample Introduction Methods
In addition to a method of atomization, atomic spectroscopic methods require a means of placing the sample within the device used
for atomization. The analysis of seawater for sodium ions requires a means for working with a sample that is in solution. The
analysis of a salt-substitute for sodium, on the other hand, requires a means for working with solid samples, which could be first
bringing it into solution or working directly with the solid. How a sample is introduced also depends on the method of atomization.
Examples of different methods of sample introduction include aspirating a solution directly into a flame, injecting a small aliquot of
solution onto a resistive heating mechanism, or exposing a solid sample to a laser or electric spark. More details on specific
methods for introducing samples appear in the chapters that follow.

This page titled 8.3: Sample Introduction Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

8.3.1 https://chem.libretexts.org/@go/page/366314
 CHAPTER OVERVIEW
9: Atomic Absorption Spectrometry
9.1: Sample Atomization Techniques
9.2: Atomic Absorption Instrumentation
9.3: Interferences in Absorption Spectroscopy
9.4: Atomic Absorption Techniques

This page titled 9: Atomic Absorption Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1
9.1: Sample Atomization Techniques
The process of converting an analyte to a free gaseous atom is called atomization. Converting an aqueous analyte into a free atom
requires that we strip away the solvent, volatilize the analyte, and, if necessary, dissociate the analyte into free atoms. Desolvating
an aqueous solution of CuCl2, for example, leaves us with solid particulates of CuCl2. Converting the particulate CuCl2 to gas
phases atoms of Cu and Cl requires thermal energy.

CuCl2 (aq) → CuCl2 (s) → Cu(g) + 2Cl(g)

There are two common atomization methods: flame atomization and electrothermal atomization, although a few elements are
atomized using other methods.

Flame Atomization
Figure 9.1.1 shows a typical flame atomization assembly with close-up views of several key components. In the unit shown here,
the aqueous sample is drawn into the assembly by passing a high-pressure stream of compressed air past the end of a capillary tube
immersed in the sample. When the sample exits the nebulizer it strikes a glass impact bead, which converts it into a fine aerosol
mist within the spray chamber. The aerosol mist is swept through the spray chamber by the combustion gases—compressed air and
acetylene in this case—to the burner head where the flame’s thermal energy desolvates the aerosol mist to a dry aerosol of small,
solid particulates. The flame’s thermal energy then volatilizes the particles, producing a vapor that consists of molecular species,
ionic species, and free atoms.

Figure 9.1.1 . Flame atomization assembly with expanded views of (a) the burner head showing the burner slot where the flame is
located; (b) the nebulizer’s impact bead; and (c) the interior of the spray chamber. Although the unit shown here is from an
instrument dating to the 1970s, the basic components of a modern flame AA spectrometer are the same.
Burner. The slot burner in Figure 9.1.1a provides a long optical pathlength and a stable flame. Because absorbance is directly
proportional to pathlength, a long pathlength provides greater sensitivity. A stable flame minimizes uncertainty due to fluctuations
in the flame.
The burner is mounted on an adjustable stage that allows the entire assembly to move horizontally and vertically. Horizontal
adjustments ensure the flame is aligned with the instrument’s optical path. Vertical adjustments change the height within the flame
from which absorbance is monitored. This is important because two competing processes affect the concentration of free atoms in
the flame. The more time an analyte spends in the flame the greater the atomization efficiency; thus, the production of free atoms
increases with height. On the other hand, a longer residence time allows more opportunity for the free atoms to combine with
oxygen to form a molecular oxide. As seen in Figure 9.1.2, for a metal this is easy to oxidize, such as Cr, the concentration of free

9.1.1 https://chem.libretexts.org/@go/page/366341
atoms is greatest just above the burner head. For a metal, such as Ag, which is difficult to oxidize, the concentration of free atoms
increases steadily with height.

Figure 9.1.2 . Absorbance versus height profiles for Ag and Cr in flame atomic absorption spectroscopy.
Flame. The flame’s temperature, which affects the efficiency of atomization, depends on the fuel–oxidant mixture, several
examples of which are listed in Table 9.1.1. Of these, the air–acetylene and the nitrous oxide–acetylene flames are the most
popular. Normally the fuel and oxidant are mixed in an approximately stoichiometric ratio; however, a fuel-rich mixture may be
necessary for easily oxidized analytes.
Table 9.1.1 . Fuels and Oxidants Used for Flame Combustion
fuel oxidant temperature range (oC)

natural gas air 1700–1900

hydrogen air 2000–2100

acetylene air 2100–2400

acetylene nitrous oxide 2600–2800

acetylene oxygen 3050–3150

Figure 9.1.3 shows a cross-section through the flame, looking down the source radiation’s optical path. The primary combustion
zone usually is rich in gas combustion products that emit radiation, limiting is useful- ness for atomic absorption. The interzonal
region generally is rich in free atoms and provides the best location for measuring atomic absorption. The hottest part of the flame
typically is 2–3 cm above the primary combustion zone. As atoms approach the flame’s secondary combustion zone, the decrease
in temperature allows for formation of stable molecular species.

Figure 9.1.3 . Profile of typical flame using a slot burner. The relative size of each zone depends on many factors, including the
choice of fuel and oxidant, and their relative proportions.
Sample Introduction. The most common means for introducing a sample into a flame atomizer is a continuous aspiration in which
the sample flows through the burner while we monitor absorbance. Continuous aspiration is sample intensive, typically requiring
from 2–5 mL of sample.
Flame microsampling allows us to introduce a discrete sample of fixed volume, and is useful if we have a limited amount of sample
or when the sample’s matrix is incompatible with the flame atomizer. For example, continuously aspirating a sample that has a high
concentration of dissolved solids—sea water, for example, comes to mind—may build-up a solid de- posit on the burner head that

9.1.2 https://chem.libretexts.org/@go/page/366341
obstructs the flame and that lowers the absorbance. Flame microsampling is accomplished using a micropipet to place 50–250 μL
of sample in a Teflon funnel connected to the nebulizer, or by dipping the nebulizer tubing into the sample for a short time. Dip
sampling usually is accomplished with an automatic sampler. The signal for flame microsampling is a transitory peak whose height
or area is proportional to the amount of analyte that is injected.
Advantages and Disadvantages of Flame Atomization. The principal advantage of flame atomization is the reproducibility with
which the sample is introduced into the spectrophotometer; a significant disadvantage is that the efficiency of atomization is quite
poor. There are two reasons for poor atomization efficiency. First, the majority of the aerosol droplets produced during nebulization
are too large to be carried to the flame by the combustion gases. Consequently, as much as 95% of the sample never reaches the
flame, which is the reason for the waste line shown at the bottom of the spray chamber in Figure 9.1.1. A second reason for poor
atomization efficiency is that the large volume of combustion gases significantly dilutes the sample. Together, these contributions
to the efficiency of atomization reduce sensitivity because the analyte’s concentration in the flame may be a factor of 2.5 × 10 −6

less than that in solution [Ingle, J. D.; Crouch, S. R. Spectrochemical Analysis, Prentice-Hall: Englewood Cliffs, NJ, 1988; p. 275].

Electrothermal Atomization
A significant improvement in sensitivity is achieved by using the resistive heating of a graphite tube in place of a flame. A typical
electrothermal atomizer, also known as a graphite furnace, consists of a cylindrical graphite tube approximately 1–3 cm in length
and 3–8 mm in diameter. As shown in Figure 9.1.4, the graphite tube is housed in an sealed assembly that has an optically
transparent window at each end. A continuous stream of inert gas is passed through the furnace, which protects the graphite tube
from oxidation and removes the gaseous products produced during atomization. A power supply is used to pass a current through
the graphite tube, resulting in resistive heating.

Figure 9.1.4 . Diagram showing a cross-section of an electrothermal atomizer, which also is known as a graphite furnace.
Samples of between 5–50 μL are injected into the graphite tube through a small hole at the top of the tube. Atomization is achieved
in three stages. In the first stage the sample is dried to a solid residue using a current that raises the temperature of the graphite tube
to about 110oC. In the second stage, which is called ashing, the temperature is increased to between 350–1200oC. At these
temperatures organic material in the sample is converted to CO2 and H2O, and volatile inorganic materials are vaporized. These
gases are removed by the inert gas flow. In the final stage the sample is atomized by rapidly increasing the temperature to between
2000–3000oC. The result is a transient absorbance peak whose height or area is proportional to the absolute amount of analyte
injected into the graphite tube. Together, the three stages take approximately 45–90 s, with most of this time used for drying and
ashing the sample.
Electrothermal atomization provides a significant improvement in sensitivity by trapping the gaseous analyte in the small volume
within the graphite tube. The analyte’s concentration in the resulting vapor phase is as much as 1000× greater than in a flame
atomization [Parsons, M. L.; Major, S.; Forster, A. R. Appl. Spectrosc. 1983, 37, 411–418]. This improvement in sensitivity—and
the resulting improvement in detection limits—is offset by a significant decrease in precision. Atomization efficiency is influenced
strongly by the sample’s contact with the graphite tube, which is difficult to control reproducibly.

Specialized Atomization Techniques

A few elements are atomized by using a chemical reaction to produce a volatile product. Elements such as As, Se, Sb, Bi, Ge, Sn,
Te, and Pb, for example, form volatile hydrides when they react with NaBH4 in the presence of acid. An inert gas carries the
volatile hydride to either a flame or to a heated quartz observation tube situated in the optical path. Mercury is determined by the
cold-vapor method in which it is reduced to elemental mercury with SnCl2. The volatile Hg is carried by an inert gas to an unheated
observation tube situated in the instrument’s optical path.

9.1.3 https://chem.libretexts.org/@go/page/366341
Flame or Electrothermal Atomization?
The most important factor in choosing a method of atomization is the analyte’s concentration. Because of its greater sensitivity, it
takes less analyte to achieve a given absorbance when using electrothermal atomization. Table 9.1.2 which compares the amount of
analyte needed to achieve an absorbance of 0.20 when using flame atomization and electrothermal atomization, is useful when
selecting an atomization method. For example, flame atomization is the method of choice if our samples contain 1–10 mg Zn2+/L,
but electrothermal atomization is the best choice for samples that contain 1–10 μg Zn2+/L.
Table 9.1.2 . Concentration of Analyte (in mg/L) That Yields an Absorbance of 0.20
element flame atomization electrothermal atomization

Ag 1.5 0.0035

Al 40 0.015

As 40 0.050

Ca 0.8 0.003

Cd 0.6 0.001

Co 2.5 0.021

Cr 2.5 0.0075

Cu 1.5 0.012

Fe 2.5 0.006

Hg 70 0.52

Mg 0.15 0.00075

Mn 1 0.003

Na 0.3 0.00023

Ni 2 0.024

Pb 5 0.080

Pt 70 0.29

Sn 50 0.023

Zn 0.3 0.00071

Source: Varian Cookbook, SpectraAA Software Version 4.00 Pro.

As: 10 mg/L by hydride vaporization; Hg: 11.5 mg/L by cold-vapor; and Sn:18 mg/L by hydride vaporization

This page titled 9.1: Sample Atomization Techniques is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

9.1.4 https://chem.libretexts.org/@go/page/366341
9.2: Atomic Absorption Instrumentation
Atomic absorption spectrophotometers use optical benches similar to those described earlier in Chapter 7, including a source of
radiation, a method for introducing the sample (covered in the previous section), a means for isolating the wavelengths of interest,
and a way to measure the amount of light absorbed or emitted.

Radiation Sources
Because atomic absorption lines are narrow, we need to use a line source instead of a continuum source to record atomic absorption
spectra. Figure 9.2.1 will help us understand why this is necessary. As discussed in Chapter 7, a typical continuum source has an
effective bandwidth on the order of 1 nm after passing through a monochromator. An atomic absorption line, as we learned in
Chapter 8, has an effective line width on the order of 0.002 nm due to Doppler broadening and pressure broadening that takes place
in a flame. If we pass the radiation from a continuum source through the flame, the incident power from the source, P , and the 0

power that reaches the detector, P , are essentially identical, leading to an absorbance of zero. A line source, which operates a
t

temperature that is lower than a flame, has a line width on the order of 0.001 nm. Passing this source radiation through the flame
results in a measurable P and a measurable absorbance.
T

Figure 9.2.1 : Illustration showing the effect on absorbance of using a continuum source and a line source for atomic absorption.
Note that the peaks in the top row have the same width; thus, the x-axis scale is different for the continuum spectra than for the line
source spectra.
The source for atomic absorption is a hollow cathode lamp that consists of a cathode and anode enclosed within a glass tube filled
with a low pressure of an inert gas, such as Ne or Ar (Figure 9.2.2). Applying a potential across the electrodes ionizes the filler gas.
The positively charged gas ions collide with the negatively charged cathode, sputtering atoms from the cathode’s surface. Some of
the sputtered atoms are in the excited state and emit radiation characteristic of the metal(s) from which the cathode is
manufactured. By fashioning the cathode from the metallic analyte, a hollow cathode lamp provides emission lines that correspond
to the analyte’s absorption spectrum.

9.2.1 https://chem.libretexts.org/@go/page/366343
 Figure 9.2.2 . Photo of a typical multielemental hollow cathode lamp. The cathode in this lamp is fashioned from an alloy of Co,
Cr, Cu, Fe, Mn, and Ni, and is surrounded by a glass shield that isolates it from the anode. The lamp is filled with Ne gas. Also
shown is the process that leads to atomic emission. Note the black deposit of sputtered metal on the outer wall of the hollow
cathode lamp. See the text for additional details.
Each element in a hollow cathode lamp provides several atomic emission lines that we can use for atomic absorption. Usually the
wavelength that provides the best sensitivity is the one we choose to use, although a less sensitive wavelength may be more
appropriate for a sample that has higher concentration of analyte. For the Cr hollow cathode lamp in Table 9.2.1, the best
sensitivity is obtained using a wavelength of 357.9 nm as this line requires the smallest concentration of analyte to achieve an
absorbance of 0.20.
Table 9.2.1 . Atomic Emission Lines for a Cr Hollow Cathode Lamp
wavelength (nm) slit width (nm) mg Cr/L giving A = 0.20 P0 (relative)

357.9 0.2 2.5 40

425.4 0.2 12 85

429.0 0.5 20 100

520.5 0.2 1500 15

520.8 0.2 500 20

Another consideration is the emission line's intensity, P . If several emission lines meet our requirements for sensitivity, we may
0

wish to use the emission line with the largest relative P0 because there is less uncertainty in measuring P0 and PT. When analyzing
a sample that is ≈10 mg Cr/L, for example, the first three wavelengths in Table 9.2.1 provide good sensitivity; the wavelengths of
425.4 nm and 429.0 nm, however, have a greater P0 and will provide less uncertainty in the measured absorbance.
The emission spectrum for a hollow cathode lamp includes, in addition to the analyte's emission lines, additional emission lines
from impurities present in the metallic cathode and from the filler gas. These additional lines are a potential source of stray
radiation that could result in an instrumental deviation from Beer’s law. The monochromator’s slit width is set as wide as possible
to improve the throughput of radiation and narrow enough to eliminate these sources of stray radiation.

Optical Benches
Atomic absorption spectrometers are available using either a single-beam and a double-beam optical bench. Figure 9.2.3 shows a
typical single-beam spectrometer, which consists of a hollow cathode lamp as a source, a flame, a grating monochromator, a
detector (usually a photomultiplier tube), and a signal processor. Also included in this design is a chopper that periodically blocks
light from the hollow cathode lamp from passing through the flame and reaching the detector. The purpose of the chopper is to
provide a means for discriminating against the emission of light from the flame, which will otherwise contribute to the total amount
of light that reaches the detector. As shown in Figure 9.2.3, when the chopper is closed, the only light reaching the detector is from
the flame; emission from the flame and light from the lamp after it passes through the flame reach the detector when the chopper is
open. The difference between the two signals gives the amount of light that reaches the detector after being absorbed by the
sample. An alternative method that accomplishes the same thing is to modulate the amount of radiation emitted by the hollow
cathode lamp.

9.2.2 https://chem.libretexts.org/@go/page/366343
 Figure 9.2.3 : Typical single-beam optical bench for atomic absorption.
Figure 9.2.4 shows the typical arrangement of a double-beam instrument for atomic absorption spectroscopy. In this design, the
chopper alternates between two optical paths: one in which light from the hollow cathode lamp bypasses the flame and that
measures the total emission of radiation from the flame and the lamp, and one that passes the hollow light from the hollow cathode
lamp through the flame and that measures the emission of light from the flame and the amount of light from the hollow cathode
lamp that is not absorbed by the sample. The difference between the two signals gives the amount of light that reaches the detector
after being absorbed by the sample.

Figure 9.2.4 : Typical double-beam optical bench for atomic absorption.

This page titled 9.2: Atomic Absorption Instrumentation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

9.2.3 https://chem.libretexts.org/@go/page/366343
9.3: Interferences in Absorption Spectroscopy
In describing the optical benches for atomic absorption spectroscopy, we noted the need to modulate the radiation from the source
in order to discriminate against emission of radiation from the flame. In this section we consider additional sources of interference
and discuss ways to compensate for them.

Spectral Interferences
A spectral interference occurs when an analyte’s absorption line overlaps with an interferent’s absorption line or band. Because
atomic absorption lines are so narrow, the overlap of two such lines seldom is a problem. On the other hand, a molecule’s broad
absorption band or the scattering of source radiation is a potentially serious spectral interference.
An important consideration when using a flame as an atomization source is its effect on the measured absorbance. Among the
products of combustion are molecular species that exhibit broad absorption bands and particulates that scatter radiation from the
source. If we fail to compensate for these spectral interferences, then the intensity of transmitted radiation is smaller than expected.
The result is an apparent increase in the sample’s absorbance. Fortunately, absorption and scattering of radiation by the flame are
corrected by analyzing a blank that does not contain the sample.
Spectral interferences also occur when components of the sample’s matrix other than the analyte react to form molecular species,
such as oxides and hydroxides. The resulting absorption and scattering constitutes the sample’s background and may present a
significant problem, particularly at wavelengths below 300 nm where the scattering of radiation becomes more important. If we
know the composition of the sample’s matrix, then we can prepare our samples using an identical matrix. In this case the
background absorption is the same for both the samples and the standards. Alternatively, if the background is due to a known
matrix component, then we can add that component in excess to all samples and standards so that the contribution of the naturally
occurring interferent is insignificant. Finally, many interferences due to the sample’s matrix are eliminated by increasing the
atomization temperature. For example, switching to a higher temperature flame helps prevents the formation of interfering oxides
and hydroxides.
If the identity of the matrix interference is unknown, or if it is not possible to adjust the flame or furnace conditions to eliminate the
interference, then we must find another method to compensate for the background interference. Several methods have been
developed to compensate for matrix interferences, and most atomic absorption spectrophotometers include one or more of these
methods.
One of the most common methods for background correction is to use a continuum source, such as a D2 lamp. Because a D2 lamp
is a continuum source, absorbance of its radiation by the analyte’s narrow absorption line is negligible. Only the background,
therefore, absorbs radiation from the D2 lamp. Both the analyte and the background, on the other hand, absorb the hollow cathode’s
radiation. Subtracting the absorbance for the D2 lamp from that for the hollow cathode lamp gives a corrected absorbance that
compensates for the background interference. Although this method of background correction is effective, it does assume that the
background absorbance is constant over the range of wavelengths passed by the monochromator. If this is not true, then subtracting
the two absorbances underestimates or overestimates the background. A typical optical arrangement is shown in Figure 9.3.1.

Figure 9.3.1 : Illustration showing a modification to the optical bench to allow for background correction using a continuous
source, such as a D2-lamp. The chopper alternates between allowing light from the hollow cathode lamp and light from the D2-
lamp to pass through the flame and reach the detector.

9.3.1 https://chem.libretexts.org/@go/page/366345
Another approach to removing the background is to take advantage of the Zeeman effect. The basis of the technique is outlined in
Figure 9.3.2 and described below in more detail. In the absence of an applied magnetic field—B = 0, where B is the strength of the
magnetic field—a p → d absorbance by the analyte takes place between two well-defined energy levels and yields a single well-
defined absorption line, as seen on the left side of panel (a). When a magnetic field is applied, B > 0, the three equal energy p-
orbitals split into three closely spaced energy levels and the five equal energy d-orbitals split into five closely spaced energy levels.
The allowed transitions between these energy levels of ΔM = 0, ±1 yields three well-defined absorption lines, as seen on the
l

right-side of panel (a), the central one of which (ΔM = 0 ) is at the same wavelength as the absorption line in the absence of the
l

applied magnetic field. This central band is the only wavelength at which the analyte absorbs.
As we see in Figure 9.3.2b, c, we apply a magnetic field to the instrument's electrothermal atomizer and place a rotating polarizer
between it and the hollow cathode lamp. When the rotating polarizer is in one position, radiation from the hollow cathode light is
absorbed only by the central absorption line, giving a measure of absorption by both the background and the analyte. when the
rotating polarizer is in the other position, radiation from the hollow cathode lamp is absorbed only by the two outside lines,
providing a measure of absorption by the background only. The difference in these two absorption values is a function of the
analyte's concentration.

Figure 9.3.2 : Illustration of using the Zeeman effect to compensate for background absorption when using an electrothermal
atomizer: panel (a) explains the origin of the Zeemen effect; panel (b) shows the modification to the instrument; and panel (c)
shows the resulting absorption lines. See the text for more details.
A third method for compensating for background absorption is to take advantage of what happens to the emission intensity of a
hollow cathode lamp when it is operated at a high current. As seen in Figure 9.3.3, when using a high current the emission band
become significantly broader than when using a normal (low) current and, at the analytical wavelength, the emission intensity from
the lamp decreases due to self-absorption, a process in which the ground state atoms in the hollow cathode lamp absorb photons
emitted by the excited state atoms in the hollow cathode lamp. When using a low current we measure absorption from both the
analyte and the background; when using a high current, absorption is due almost exclusively to the background. This approach is
called Smith-Hieftje background corrections.

9.3.2 https://chem.libretexts.org/@go/page/366345
 Figure 9.3.1 : Illustration showing the basis of Smith-Hieftje background correction.

Chemical Interferences
The quantitative analysis of some elements is complicated by chemical interferences that occur during atomization. The most
common chemical interferences are the formation of nonvolatile compounds that contain the analyte and ionization of the analyte.
One example of the formation of a nonvolatile compound is the effect of PO or Al3+ on the flame atomic absorption analysis of
3−

Ca2+. In one study, for example, adding 100 ppm Al3+ to a solution of 5 ppm Ca2+ decreased calcium ion’s absorbance from 0.50 to
0.14, while adding 500 ppm PO to a similar solution of Ca2+ decreased the absorbance from 0.50 to 0.38. These interferences
3−
4

are attributed to the formation of nonvolatile particles of Ca3(PO4)2 and an Al–Ca–O oxide [Hosking, J. W.; Snell, N. B.; Sturman,
B. T. J. Chem. Educ. 1977, 54, 128–130].
When using flame atomization, we can minimize the formation of non-volatile compounds by increasing the flame’s temperature
by changing the fuel-to-oxidant ratio or by switching to a different combination of fuel and oxidant. Another approach is to add a
releasing agent or a protecting agent to the sample. A releasing agent is a species that reacts preferentially with the interferent,
releasing the analyte during atomization. For example, Sr2+ and La3+ serve as releasing agents for the analysis of Ca2+ in the
presence of PO or Al3+. Adding 2000 ppm SrCl2 to the Ca2+/ PO and to the Ca2+/Al3+ mixtures described in the previous
3−

4
3−

paragraph increased the absorbance to 0.48. A protecting agent reacts with the analyte to form a stable volatile complex. Adding
1% w/w EDTA to the Ca2+/ PO solution described in the previous paragraph increased the absorbance to 0.52.
3−
4

An ionization interference occurs when thermal energy from the flame or the electrothermal atomizer is sufficient to ionize the
analyte
+ −
M(s) ⇌  M (aq) + e (9.3.1)

where M is the analyte. Because the absorption spectra for M and M+ are different, the position of the equilibrium in reaction 9.3.1
affects the absorbance at wavelengths where M absorbs. To limit ionization we add a high concentration of an ionization
suppressor, which is a species that ionizes more easily than the analyte. If the ionization suppressor's concentration is sufficient,
then the increased concentration of electrons in the flame pushes reaction 9.3.1 to the left, preventing the analyte’s ionization.
Potassium and cesium frequently are used as an ionization suppressor because of their low ionization energy.

This page titled 9.3: Interferences in Absorption Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

9.3.3 https://chem.libretexts.org/@go/page/366345
9.4: Atomic Absorption Techniques
Preparing the Sample
Flame and electrothermal atomization require that the analyte is in solution. Solid samples are brought into solution by dissolving
in an appropriate solvent. If the sample is not soluble it is digested, either on a hot-plate or by microwave, using HNO3, H2SO4, or
HClO4. Alternatively, we can extract the analyte using a Soxhlet extractor. Liquid samples are analyzed directly or the analytes
extracted if the matrix is in- compatible with the method of atomization. A serum sample, for instance, is difficult to aspirate when
using flame atomization and may produce an unacceptably high background absorbance when using electrothermal atomization. A
liquid–liquid extraction using an organic solvent and a chelating agent frequently is used to concentrate analytes. Dilute solutions
of Cd2+, Co2+, Cu2+, Fe3+, Pb2+, Ni2+, and Zn2+, for example, are concentrated by extracting with a solution of ammonium
pyrrolidine dithiocarbamate in methyl isobutyl ketone.

Standardizing the Method

Because Beer’s law also applies to atomic absorption, we might expect atomic absorption calibration curves to be linear. In
practice, however, most atomic absorption calibration curves are nonlinear or linear over a limited range of concentrations.
Nonlinearity in atomic absorption is a consequence of instrumental limitations, including stray radiation from the hollow cathode
lamp and the variation in molar absorptivity across the absorption line. Accurate quantitative work, therefore, requires a suitable
means for computing the calibration curve from a set of standards.
When possible, a quantitative analysis is best conducted using external standards. Unfortunately, matrix interferences are a frequent
problem, particularly when using electrothermal atomization. For this reason the method of standard additions often is used. One
limitation to this method of standardization, however, is the requirement of a linear relationship between absorbance and
concentration.

Most instruments include several different algorithms for computing the calibration curve. The instrument in my lab, for
example, includes five algorithms. Three of the algorithms fit absorbance data using linear, quadratic, or cubic polynomial
functions of the analyte’s concentration. It also includes two algorithms that fit the concentrations of the standards to quadratic
functions of the absorbance.

Evaluation of Atomic Absorption Spectroscopy

Scale of Operation
Atomic absorption spectroscopy is ideally suited for the analysis of trace and ultratrace analytes, particularly when using
electrothermal atomization. For minor and major analytes, sample are diluted before the analysis. Most analyses use a macro or a
meso sample. The small volume requirement for electrothermal atomization or for flame microsampling, however, makes practical
the analysis of micro and ultramicro samples.

Accuracy
If spectral and chemical interferences are minimized, an accuracy of 0.5–5% is routinely attainable. When the calibration curve is
nonlinear, accuracy is improved by using a pair of standards whose absorbances closely bracket the sample’s absorbance and
assuming that the change in absorbance is linear over this limited concentration range. Determinate errors for electrothermal
atomization often are greater than those obtained with flame atomization due to more serious matrix interferences.

Precision
For an absorbance greater than 0.1–0.2, the relative standard deviation for atomic absorption is 0.3–1% for flame atomization and
1–5% for electrothermal atomization. The principle limitation is the uncertainty in the concentration of free analyte atoms that
result from variations in the rate of aspiration, nebulization, and atomization for a flame atomizer, and the consistency of injecting
samples for electrothermal atomization.

Sensitivity
The sensitivity of a flame atomic absorption analysis is influenced by the flame’s composition and by the position in the flame from
which we monitor the absorbance. Normally the sensitivity of an analysis is optimized by aspirating a standard solution of analyte

9.4.1 https://chem.libretexts.org/@go/page/366365
and adjusting the fuel-to-oxidant ratio, the nebulizer flow rate, and the height of the burner, to give the greatest absorbance. With
electrothermal atomization, sensitivity is influenced by the drying and ashing stages that precede atomization. The temperature and
time at each stage is optimized for each type of sample.
Sensitivity also is influenced by the sample’s matrix. We already noted, for example, that sensitivity is decreased by a chemical
interference. An increase in sensitivity may be realized by adding a low molecular weight alcohol, ester, or ketone to the solution,
or by using an organic solvent.

Selectivity
Due to the narrow width of absorption lines, atomic absorption provides excellent selectivity. Atomic absorption is used for the
analysis of over 60 elements at concentrations at or below the level of μg/L.

Time, Cost, and Equipment

The analysis time when using flame atomization is short, with sample throughputs of 250–350 determinations per hour when using
a fully automated system. Electrothermal atomization requires substantially more time per analysis, with maximum sample
throughputs of 20–30 determinations per hour. The cost of a new instrument ranges from between $10,000– $50,000 for flame
atomization, and from $18,000–$70,000 for electrothermal atomization. The more expensive instruments in each price range
include double-beam optics, automatic samplers, and can be programmed for multielemental analysis by allowing the wavelength
and hollow cathode lamp to be changed automatically.

This page titled 9.4: Atomic Absorption Techniques is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

9.4.2 https://chem.libretexts.org/@go/page/366365
 CHAPTER OVERVIEW
10: Atomic Emission Spectrometry
10.1: Emission Spectroscopy Based on Flame and Plasma Sources
10.2: Emission Spectroscopy Based on Arc and Spark Sources

This page titled 10: Atomic Emission Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1
10.1: Emission Spectroscopy Based on Flame and Plasma Sources
What is Emission?
An analyte in an excited state possesses an energy, E2, that is greater than its energy when it is in a lower energy state, E1. When
the analyte returns to its lower energy state—a process we call relaxation—the excess energy, ΔE, is

ΔE = E2 − E1

There are several ways in which an atom may end up in an excited state, including thermal energy, which is the focus of this
chapter. The amount of time an atom, A, spends in its excited state—what we call the excited state's lifetime—is short, typically
10
−5
to 10 s for an electronic excited state. Relaxation of the atom's excited state, A*, occurs through several mechanisms,
−9

including collisions with other species in the sample and the emission of photons. In the first process, which we call nonradiative
relaxation, the excess energy is released as heat.
∗
A ⟶ A +  heat 

In the second mechanism, the excess energy is released as a photon of electromagnetic radiation.
∗
A ⟶ A + hν

The release of a photon following thermal excitation is called emission. The focus of this chapter is on the emission of ultraviolet
and visible radiation following the thermal excitation of atoms. Atomic emission spectroscopy has a long history. Qualitative
applications based on the color of flames were used in the smelting of ores as early as 1550 and were more fully developed around
1830 with the observation of atomic spectra generated by flame emission and spark emission [Dawson, J. B. J. Anal. At. Spectrosc.
1991, 6, 93–98]. Quantitative applications based on the atomic emission from electric sparks were developed by Lockyer in the
early 1870 and quantitative applications based on flame emission were pioneered by Lundegardh in 1930. Atomic emission based
on emission from a plasma was introduced in 1964.

Atomic Emission Spectra

Atomic emission occurs when a valence electron in a higher energy atomic orbital returns to a lower energy atomic orbital. Figure
10.1.1 shows a portion of the energy level diagram for sodium, which consists of a series of discrete lines at wavelengths that

correspond to the difference in energy between two atomic orbitals.

FFigure 10.1.1 . Valence shell energy level diagram for sodium. The wavelengths corresponding to several transitions are shown.
The intensity of an atomic emission line, Ie, is proportional to the number of atoms, N , that populate the excited state
∗

∗
Ie = kN (10.1.1)

where k is a constant that accounts for the efficiency of the transition. If a system of atoms is in thermal equilibrium, the population
of excited state i is related to the total concentration of atoms, N, by the Boltzmann distribution. For many elements at temperatures
of less than 5000 K the Boltzmann distribution is approximated as

10.1.1 https://chem.libretexts.org/@go/page/366444
 gi
∗ −Ei /kT
N =N ( )e (10.1.2)
g0

where gi and g0 are statistical factors that account for the number of equivalent energy levels for the excited state and the ground
state, Ei is the energy of the excited state relative to the ground state, E0, k is Boltzmann’s constant (1.3807 × 10 J/K), and T is
−23

the temperature in Kelvin. From Equation 10.1.2 we expect that excited states with lower energies have larger populations and
more intense emission lines. We also expect emission intensity to increase with temperature. The emission spectrum for sodium is
shown in Figure 10.1.2

Figure 10.1.2 : Visible emission spectrum for sodium. The inset shows a close up of the emission lines at 589.0 and 589.6 nm that
appear in the energy level diagram in Figure 10.1.1 .
An atomic emission spectrometer is similar in design to the instrumentation for atomic absorption. In fact, it is easy to adapt most
flame atomic absorption spectrometers for atomic emission by turning off the hollow cathode lamp and monitoring the difference
between the emission intensity when aspirating the sample and when aspirating a blank. Many atomic emission spectrometers,
however, are dedicated instruments designed to take advantage of features unique to atomic emission, including the use of plasmas,
arcs, sparks, and lasers as atomization and excitation sources, and an enhanced capability for multielemental analysis.

Flames as a Source
Atomization and excitation in flame atomic emission is accomplished with the same nebulization and spray chamber assembly used
in atomic absorption (see Chapter 9). The burner head consists of a single or multiple slots, or a Meker-style burner. Older atomic
emission instruments often used a total consumption burner in which the sample is drawn through a capillary tube and injected
directly into the flame.

A Meker burner is similar to the more common Bunsen burner found in most laboratories; it is designed to allow for higher
temperatures and for a larger diameter flame.

The Inductively Coupled Plasma Source

A plasma is a hot, partially ionized gas that contains an abundant concentration of cations and electrons. The plasma used in atomic
emission is formed by ionizing a flowing stream of argon gas, producing argon ions and electrons. A plasma’s high temperature
results from resistive heating as the electrons and argon ions move through the gas. Because a plasma operates at a much higher
temperature than a flame, it provides for a better atomization efficiency and a higher population of excited states.
A schematic diagram of the inductively coupled plasma source (ICP) is shown in Figure 10.1.3. The ICP torch consists of three
concentric quartz tubes, surrounded at the top by a radio-frequency induction coil. The sample is mixed with a stream of Ar using a
nebulizer, and is carried to the plasma through the torch’s central capillary tube. Plasma formation is initiated by a spark from a
Tesla coil. An alternating radio-frequency current in the induction coil creates a fluctuating magnetic field that induces the argon
ions and the electrons to move in a circular path. The resulting collisions with the abundant unionized gas give rise to resistive
heating, providing temperatures as high as 10000 K at the base of the plasma, and between 6000 and 8000 K at a height of 15–20
mm above the coil, where emission usually is measured. At these high temperatures the outer quartz tube must be thermally

10.1.2 https://chem.libretexts.org/@go/page/366444
isolated from the plasma. This is accomplished by the tangential flow of argon shown in the schematic diagram. Samples are
brought into the ICP using the same basic types of nebulization described in Chapter 8 for flame atomic absorption spectroscopy.

Figure 10.1.3 . Schematic diagram of an inductively coupled plasma torch for atomic emission.

The Direct Current Plasma Source

An alternative to the inductively coupled plasma source is the direct current (dc) plasma jet, one example of which is illustrated in
Figure 10.1.4. The argon plasma (shown here in blue) forms between two graphite anodes and a tungsten cathode. The sample is
aspirated into the plasma's excitation region where it undergoes atomization, excitation, and emission at temperatures of 5000 K.

Figure 10.1.4 : Illustration showing a schematic diagram of a direct current plasma source of atomic emission.

Flame and Plasma Spectrometers

One advantage of atomic emission over atomic absorption is the ease of analyzing samples for multiple analytes. This additional
capability arises because atomic emission, unlike atomic absorption, does not need an analyte-specific source of radiation. The two
most common types of spectrometers are sequential and multichannel. In a sequential spectrometer the instrument has a single
detector and uses the monochromator to move from one emission line to the next. A multichannel spectrometer uses the
monochromator to disperse the emission across a field of detectors, each of which measures the emission intensity at a different
wavelength.

Sequential Instruments
A sequential instrument uses a programmable scanning monochromator, such as those described in Chapter 7, to rapidly move the
monochromator's grating over wavelength regions that are not of interest, and then pauses and scans slowly over the emission lines

10.1.3 https://chem.libretexts.org/@go/page/366444
of the analytes. Sampling rates of 300 determinations per hour are possible with this configuration. Another option, which is less
common, is to move the exit slit and the detector across the monochromator's focal plane, pausing and recording the emission at the
desired wavelengths.

Multichannel Instruments
Another approach to a multielemental analysis is to use a multichannel instrument that allows us to monitor simultaneously many
analytes. A simple design for a multichannel spectrometer, shown in Figure 10.1.5, couples a monochromator with multiple
detectors that are positioned in a semicircular array around the monochromator at positions that correspond to the wavelengths for
the analytes. A sample throughput of 3000 determinations per hour are possible using a multichannel ICP.

Figure 10.1.5 . Schematic diagram of a multichannel atomic emission spectrometer for the simultaneous analysis of several
elements. Instruments may contain as many as 48–60 detectors.
Another option for a multichannel instrument takes advantage of the charge-injection device, or CID, as a detector (see Chapter 7
for discussion of the charge-coupled device, another type of charge-transfer device used as a detector). Light from the plasma
source is dispersed across the CID in two dimensions. The surface of the CID has in excess of 90000 detecting elements, or
pixels, that allows for a resolution between detecting elements on the order of 0.04 nm. Light from the atomic emission source is
distributed across the detector's surface by a diffraction grating such that each element of interest is detected using its own set of
pixels, called a read window. Figure 10.1.6 shows that individual read windows consist of a set of detecting elements, nine of
which collect photons from the spectral line and 30 of which provide a measurement of the source's background.

Figure 10.1.6 : Illustration showing a charge-injection detector. The read window consists of nine central elements that measure the
analyte's emission intensity (shown here using blue shading) and 30 of which measure the background emission from the plasma
source.

10.1.4 https://chem.libretexts.org/@go/page/366444
Application of Flame and Plasma Sources
Atomic emission is used widely for the analysis of trace metals in a variety of sample matrices. The development of a quantitative
atomic emission method requires several considerations, including choosing a source for atomization and excitation, selecting a
wavelength and slit width, preparing the sample for analysis, minimizing spectral and chemical interferences, and selecting a
method of standardization.

Choice of Atomization and Excitation Source

Except for the alkali metals, detection limits when using an ICP are significantly better than those obtained with flame emission
(Table 10.1.1). Plasmas also are subject to fewer spectral and chemical interferences. For these reasons a plasma emission source is
usually the better choice.
Table 10.1.1 . Detection Limits for Atomic Emission
element detection limit (µg/mL): flame emission detection limit (µg/mL): ICP

Ag 2 0.2

Al 3 0.2

As 2000 2

Ca 0.1 0.0001

Cd 300 0.07

Co 5 0.1

Cr 1 0.08

Fe 10 0.09

Hg 150 1

K 0.01 30

Li 0.001 0.02

Mg 1 0.02

Mn 1 0.01

Na 0.01 0.1

Ni 10 0.2

Pb 0.2 1

Pt 2000 0.9

Sn 100 3

Zn 1000 0.1

Source: Parsons, M. L.; Major, S.; Forster, A. R.; App. Spectrosc. 1983, 37, 411–418.

Selecting the Wavelength and Slit Width

The choice of wavelength is dictated by the need for sensitivity and the need to avoid interferences from the emission lines of other
constituents in the sample. Because an analyte’s atomic emission spectrum has an abundance of emission lines—particularly when
using a high temperature plasma source—it is inevitable that there will be some overlap between emission lines. For example, an
analysis for Ni using the atomic emission line at 349.30 nm is complicated by the atomic emission line for Fe at 349.06 nm.
A narrower slit width provides better resolution, but at the cost of less radiation reaching the detector. The easiest approach to
selecting a wavelength is to record the sample’s emission spectrum and look for an emission line that provides an intense signal and
is resolved from other emission lines.

10.1.5 https://chem.libretexts.org/@go/page/366444
Preparing the Sample
Flame and plasma sources are best suited for samples in solution and in liquid form. Although a solid sample can be analyzed by
directly inserting it into the flame or plasma, they usually are first brought into solution by digestion or extraction.

Minimizing Spectral Interferences

The most important spectral interference is broad, background emission from the flame or plasma and emission bands from
molecular species. This background emission is particularly severe for flames because the temperature is insufficient to break down
refractory compounds, such as oxides and hydroxides. Background corrections for flame emission are made by scanning over the
emission line and drawing a baseline (Figure 10.1.7). Because a plasma’s temperature is much higher, a background interference
due to molecular emission is less of a problem. Although emission from the plasma’s core is strong, it is insignificant at a height of
10–30 mm above the core where measurements normally are made.

Figure 10.1.7 . Method for correcting an analyte’s emission for the flame’s background emission.

Minimizing Chemical Interferences

Flame emission is subject to the same types of chemical interferences as atomic absorption; they are minimized using the same
methods: by adjusting the flame’s composition and by adding protecting agents, releasing agents, or ionization suppressors. An
additional chemical interference results from self-absorption. Because the flame’s temperature is greatest at its center, the
concentration of analyte atoms in an excited state is greater at the flame’s center than at its outer edges. If an excited state atom in
the flame’s center emits a photon, then a ground state atom in the cooler, outer regions of the flame may absorb the photon, which
decreases the emission intensity. For higher concentrations of analyte self-absorption may invert the center of the emission band
(Figure 10.1.8).

Figure 10.1.8 . Atomic emission lines for (a) a low concentration of analyte, and (b) a high concentration of analyte showing the
effect of self-absorption.
Chemical interferences when using a plasma source generally are not significant because the plasma’s higher temperature limits the
formation of nonvolatile species. For example, PO is a significant interferent when analyzing samples for Ca2+ by flame
3−
4

emission, but has a negligible effect when using a plasma source. In addition, the high concentration of electrons from the
ionization of argon minimizes ionization interferences.

10.1.6 https://chem.libretexts.org/@go/page/366444
Standardizing the Method
From Equation 10.1.1 we know that emission intensity is proportional to the population of the analyte’s excited state, N . If the∗

flame or plasma is in thermal equilibrium, then the excited state population is proportional to the analyte’s total population, N,
through the Boltzmann distribution (Equation 10.1.2).
A calibration curve for flame emission usually is linear over two to three orders of magnitude, with ionization limiting linearity
when the analyte’s concentrations is small and self-absorption limiting linearity at higher concentrations of analyte. When using a
plasma, which suffers from fewer chemical interferences, the calibration curve often is linear over four to five orders of magnitude
and is not affected significantly by changes in the matrix of the standards.
Emission intensity is affected significantly by many parameters, including the temperature of the excitation source and the
efficiency of atomization. An increase in temperature of 10 K, for example, produces a 4% increase in the fraction of Na atoms in
the 3p excited state, an uncertainty in the signal that may limit the use of external standards. The method of internal standards is
used when the variations in source parameters are difficult to control. To compensate for changes in the temperature of the
excitation source, the internal standard is selected so that its emission line is close to the analyte’s emission line. In addition, the
internal standard should be subject to the same chemical interferences to compensate for changes in atomization efficiency. To
accurately correct for these errors the analyte and internal standard emission lines are monitored simultaneously.

This page titled 10.1: Emission Spectroscopy Based on Flame and Plasma Sources is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

10.1.7 https://chem.libretexts.org/@go/page/366444
10.2: Emission Spectroscopy Based on Arc and Spark Sources
Arc Source
An arc source consists of two electrodes separated by a gap of up to 20 mm (see Figure 10.2.1 for one configuration). A potential
of 50 V (or more) is applied and a continuous current in the range of 2–30 A is maintained throughout the analysis. If the sample is
a metal, then it can be fashioned into the electrodes. For nonmetallic samples, the electrodes typically are fashioned from graphite
and a cup-like depression is drilled into one of the electrodes. The sample is ground into a powder and packed into the sample cup.
The plasma generated by an arc source typically has a temperature of 4000 K to 5000 K and has an abundance of emission lines for
the analyte with a relatively small background emission.

Figure 10.2.1 : Example of an arc source in which the electrodes are fashioned from graphite. The lower electrode is fashioned to
contain a cup-like depression in which a sample can be placed. The orange starburst shows the location where the plasma forms
and where emission takes place.

Spark Source
Unlike an arc source, which generates a continuous emission of electromagnetic radiation, a spark source generates a series of short
emissions, each lasting on the order of a few µs. The sample serves as one of the two electrodes, with the other electrode fashioned
from tungsten (see Figure 10.2.2). The two electrodes are separated by a gap of 3–6 mm. A potential as small as 300–500 V and as
large as 1020 KV. The frequency of the spark is in the range of 100–500 per second. The temperature within the plasma can be
quite intense, which gives rise to both emission lines from the atoms, but also emission from ions formed in the plasma.

Figure 10.2.2 : Example of spark source. The orange starburst shows the location where the plasma forms and where emission takes
place.

Instrumentation
For both the arc source and the spark source, emission from the plasma is collected and analyzed using the same types of optical
benches discussed in the previous section on atomic emission from flames and plasma sources. Figure 10.2.3 shows an emission
spectrum for a sample of the alkaline earth metals, which shows a single intense emission line for Ca at 422.673nm and a single

10.2.1 https://chem.libretexts.org/@go/page/366446
intense emission line for Sr at 460.7331 nm. Mg exhibits three closely spaced emission lines at 516.7322 nm, 517.2684 nm, and
518.3604 nm. Finally, Ba has a single strong emission line at 553.5481 nm, but also many less intense emission lines above 600
nm. The presence of faint, but measurable emission lines can create complications when trying to identify the elements present in a
sample.

Figure 10.2.3 : Simulated emission spectrum using an arc source. The sample shows strong emission lines for Ca, Sr, Mg, and Ba
that can serve for qualitative identification of these elements. The faint lines above 600 nm—you may have difficulty seeing that
they are there—are additional emission lines from Ba. What appears as a single line for Mg is actually three closely spaced
emission lines at 516.7322 nm, 517.2684 nm, and 518.3604 nm.

This page titled 10.2: Emission Spectroscopy Based on Arc and Spark Sources is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

10.2.2 https://chem.libretexts.org/@go/page/366446
 CHAPTER OVERVIEW
11: Atomic Mass Spectrometry
11.1: General Features of Atomic Mass Spectrometry
11.2: Mass Spectrometers
11.3: Inductively Coupled Plasma Mass Spectrometer
11.4: Other Forms of Atomic Mass Spectrometry

This page titled 11: Atomic Mass Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1
11.1: General Features of Atomic Mass Spectrometry
In mass spectrometry—whether of atoms, which is covered in this chapter, or of molecules, which is covered in Chapter 20—we
convert the analyte into ions and then separate these ions based on the ratio of their masses to their charges. In this section we give
careful attention to what we mean by mass, by charge, and by mass-to-charge ratio. We also give brief consideration to how we
generate and measure ions, topics covered in greater detail in subsequent sections.

Atomic Weights in Mass Spectrometry

We trace the modern era of chemistry to John Dalton’s development of atomic theory, which made three hypotheses:
1. Elements, which are the smallest division of matter with distinct chemical properties, are composed of atoms. All atoms of a
given element are identical—This is not strictly true, as we will see shortly, but we won’t hold that against Dalton!—and
different from the atoms of other elements. The element carbon is made of carbon atoms, which are different from the atoms of
oxygen that make up elemental oxygen.
2. Compounds are composed of atoms from two or more elements. Because atoms cannot be subdivided, the elements that make
up a compound are always present in ratios of whole numbers. A compound containing carbon and oxygen, for example, can
have 1 carbon atom and 1 oxygen atom (CO) or 1 carbon atom and 2 oxygen atoms (CO2), but it cannot have 1.5 carbon atoms.
3. In a chemical reaction, the elements that make up the reactants rearrange to make new compounds as products. The atoms that
make up these compounds, however, are not destroyed, nor are new atoms created.
Dalton’s first hypothesis simply recognized the atom as the basic building block of chemistry. Water, for example, is made from
atoms of hydrogen and oxygen. The second hypothesis recognizes that for every compound there is a fixed combination of atoms.
Regardless of its source (rain, tears, or a bottle of Evian) a molecule of water always consists of two hydrogen atoms for every
atom of oxygen. Dalton’s third hypothesis is a statement that atoms are conserved in a reaction; this is more commonly known as
the conservation of mass.

The Structure of the Atom

Although Dalton believed that atoms were indivisible, we know now that they are made from three smaller subatomic particles: the
electron, the proton, and the neutron. The atom, however, remains the smallest division of matter with distinct chemical properties.
Electrons, Protons, and Neutrons. The characteristic properties of electrons, protons, and neutrons, are shown in Table 11.1.1.
Table 11.1.1 . Mass and Charge of Subatomic Particles
particle mass (g) unit charge charge (in Coulombs, C)

electron 9.10939 × 10
−28
−1 −1.6022 × 10
−19

proton 1.67262 × 10
−24
+1 +1.6022 × 10
−19

neutron 1.67493 × 10
−24
0 0

The proton and the neutron make up the atom’s nucleus, which is located at the center of the atom and has a radius of
approximately 5 × 10  pm . The remainder of the atom, which has radius of approximately 100 pm, is mostly empty space in
−3

which the electrons are free to move. Of the three subatomic particles, only the electron and the proton carry a charge, which we
can express as a relative unit charge, such as +1 or −2, or as an absolute charge in Coulombs. Because elements have no net
charge (that is, they are neutral), the number of electrons and protons in an element must be the same.
Atomic Numbers. Why is an atom of carbon different from an atom of hydrogen or helium? One possible explanation is that
carbon and hydrogen and helium have different numbers of electrons, protons, or neutrons; Table 11.1.2 provides the relevant
numbers.
Table 11.1.2 . Comparison of the Elements Hydrogen, Helium, and Carbon
element number of protons number of neutrons 1 number of electrons

hydrogen 1 0, 1, or 2 2

helium 2 2 2

carbon 6 6, 7, or 8 6

11.1.1 https://chem.libretexts.org/@go/page/387717
 element number of protons number of neutrons 1 number of electrons
1 Only the number of neutrons for the most important naturally occurring forms of these elements are shown here.

Note that although Table 11.1.2 shows that a helium atom has two neutrons, an atom of hydrogen or carbon has three possibilities
for the numbers of neutrons. It is even possible for a hydrogen atom to exist without a neutron. Clearly the number of neutrons is
not crucial to determining if an atom is carbon, hydrogen, or helium. Although hydrogen, helium, and carbon have different
numbers of electrons, the number is not critical to an element's identity. For example, it is possible to strip an electron away from
helium to form a helium ion with a charge of +1 that has the same number of electrons as hydrogen; nevertheless, it is still helium.
What makes an atom carbon is the presence of six protons, whereas every atom of hydrogen has one proton and every atom of
helium has two protons. The number of protons in an atom is called its atomic number, which we represent as Z.
Atomic Mass and Isotopes. Protons and neutrons are of similar mass and much heavier than electrons (see Table 11.1.1); thus,
most of an atom’s mass is in its nucleus. Because not all of an element’s atoms necessarily have the same number of neutrons, it is
possible for two atoms of an element to differ in mass. For this reason, the sum of an atom’s protons and neutrons is known as its
mass number (A). Carbon, for example, can have a mass number of 12, 13, or 14 (six protons and six, seven, or eight neutrons), and
hydrogen can have a mass number of 1, 2, or 3 (one proton and zero, one, or two neutrons).
Atoms of the same element (same Z), but with a different number of neutrons (different A) are called isotopes. Hydrogen, for
example has three isotopes (see Table 11.1.2). The isotope with 0 neutrons is the most abundant, accounting for 99.985% of all
stable hydrogen atoms, and is known, somewhat self-referentially, as hydrogen. Deuterium, which accounts for 0.015% of all stable
hydrogen atoms, has 1 neutron. The isotope of hydrogen with two neutrons is called tritium. Because tritium is radioactive it is
unstable and disappears with time.
The usual way to represent isotopes is with the symbol X where X is the atomic symbol for the element. The three isotopes of
A
Z

hydrogen, which has an elemental symbol of H, are H , H , and H . Because the elemental symbol (X) and the atomic number (Z)
1
1
2
1
3
1

provide redundant information, we often omit the atomic number; thus, deuterium becomes H . Unlike hydrogen, the isotopes of
2

other elements do not have specific names. Instead they are named by taking the element’s name and appending the atomic mass.
For example, the isotopes of carbon are called carbon-12, carbon-13, and carbon-14.

Atomic Mass
Individual atoms weigh very little, typically about 10  g to 10  g . This amount is so small that there is no easy way to
−24 −22

measure the mass of a single atom. To assign masses to atoms it is necessary to assign a mass to one atom and to report the masses
of all other atoms relative to that absolute standard. By agreement, atomic mass is stated in terms of atomic mass units (amu) or
Daltons (Da), where 1 amu and 1 Da are defined as 1/12 of the mass of an atom of carbon-12. The atomic mass of carbon-12,
therefore, is exactly 12 amu. The atomic mass of carbon-13 is 13.00335 amu because the mass of an atom of carbon-13 is
1.0836125×greater than the mass of an atom of carbon-12.

 Note

If you calculate the masses of carbon-12 and carbon-13 by adding together the masses of each isotope’s electrons, neutrons,
and protons from Table 11.1.1 you will obtain a mass ratio of 1.08336, not 1.0836125. The reason for this is that the masses in
Table 11.1.1 are for “free” electrons, protons, and neutrons; that is, for electrons, protons, and neutrons that are not in an atom.
When an atom forms, some of the mass is lost. “Where does it go?,” you ask. Remember Einstein and E = mc ? Mass can be 2

converted to energy and the lost mass is the nuclear binding energy that holds the nucleus together.

Average Atomic Mass. Because carbon exists in several isotopes, the atomic mass of an “average” carbon atom is not exactly 12
amu. Instead it is usually reported on periodic tables as 12.01 or 12.011, values that are closer to 12.0 because 98.90% of all carbon
atoms are carbon-12. The IUPAC's Commission on Isotopic Abundances and Atomic Weights currently reports its mass as
[12.0096, 12.0116] amu where the values in the brackets are the lower and the upper estimates for the average mass in a variety of
naturally occurring materials. As shown in the following example, if you know the percent abundance and atomic masses of an
element’s isotopes, then you can calculate it’s average atomic mass.

11.1.2 https://chem.libretexts.org/@go/page/387717
  Example 11.1.1

The element magnesium, Mg, has three stable isotopes with the following atomic masses and percent abundances:

isotope mass (amu) percent abundance

24
Mg 23.9924 78.70
25
Mg 24.9938 10.13
26
Mg 25.9898 11.17

Calculate the average atomic mass for magnesium.

Solution
To find the average atomic mass we multiply each isotopes’ atomic mass by its fractional abundance (the decimal equivalent of
its percent abundance) and add together the results; thus
avg. amu = (0.7870)(23.994 amu) + (0.1013)(24.9938 amu) + (0.1117)(25.9898 amu) avg. amu = 24.32 amu

As the next example shows, we also can work such problems in reverse, using an element’s average atomic mass and the atomic
masses of its isotopes to find each isotope’s percent abundance.

 Example 11.1.2

The element gallium, Ga, has two naturally occurring isotopes. The isotope Ga has an atomic mass of 68.926 amu and the
69

isotope Ga has an atomic mass of 70.926 amu. The average atomic mass for gallium is 69.723. Find the percent abundances
71

for gallium’s two isotopes.

Solution
If we let x be the fractional abundance of Ga, then the fractional abundance of Ga is 1 – x (that is, the total amounts of
69 71

Ga and Ga must add up to one). Usingthe same general approach as Example 11.1.1, we find that
69 71

69.723 amu = (x)(68.926 amu) + (1 – x)(70.926 amu)

69.723 amu = 68.926x amu + 70.926 amu – 70.926x amu
2.000x amu = 1.203 amu
x = 0.6015
1 – x = 1 – 0.6015 = 0.3985
Thus, 60.15% of naturally occurring gallium is 69
Ga and 39.85% is 71
Ga .

 Note

Although many periodic tables report atomic masses to two decimal places—the periodic table I consult most frequently, for
example gives the average atomic mass of carbon as 12.01 amu—the high resolving power of some mass spectrometers allows
us to report masses to three or four decimal places.

Mass-to-Charge Ratio
As we will learn later, a mass spectrometer separates ions on the basis of their mass-to-charge ratio (m/z), and not on their mass
only or their charge only. As most ions that form during mass spectrometry are singly charged, spectra are often reported using
masses (m) instead of mass-to-charge ratios; be sure to remain alert for this when looking at mass spectra.

11.1.3 https://chem.libretexts.org/@go/page/387717
This page titled 11.1: General Features of Atomic Mass Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

11.1.4 https://chem.libretexts.org/@go/page/387717
11.2: Mass Spectrometers
A mass spectrometer has three essential needs: a means for producing ions, in this case (mostly) singly charged atoms; a means for
separating these ions in space or in time by their mass-to-charge ratios; and a means for counting the number of ions for each mass-
to-charge ratio. Figure 11.2.1 provides a general view of a mass spectrometer in the same way that we first introduced optical
instruments in Chapter 7. The ionization of the sample is analagous to the source of photons in optical spectroscopy as it generates
the particles (ions, instead of photons) that ultimately make up the measured signal. The separation of the resulting ions by their
mass-to-charge ratios, which is accomplished using a mass analyzer, is analagous to the role of a monochromator in optical
spectroscopy. The means for counting ions serves the same role as, for example, a photomultiplier tube in optical spectroscopy.
Note that the mass spectrometer is held under vacuum as this allows the ions to travel great distances without undergoing collisions
that might alter their charge or energy.

Figure 11.2.1 : The basic components of a mass spectrometer. The blue arrows show the direction of sample-to-signal in which the
sample is ionized, the ions separated by their mass-to-charge ratio, and ions with the same mass-to-charge ratio counted. The text in
red at the top of the figure relates these components to their equivalent in optical spectroscopy.

Sources of Ionization
The most common means for generating ions are plasmas of various sorts, lasers, electrical sparks, and other ions. We will give
greater attention to these in the next several sections as we consider specific examples of atomic mass spectrometry.

Transducers for Counting Ions

The transducer for mass spectrometry must be able to report the number of ions that emerge from the mass analyzer. Here we
consider two common types of transducers.

Electron Multipliers
In Chapter 7 we introduced the photomultiplier tube as a way to convert photons into electrons, amplifying the signal so that a
single photon produces 106 to 107 electrons, which generates a measurable current. An electron multiplier serves the same role in
mass spectrometry. Figure 11.2.2 shows two versions of this transducer. The electron multiplier in Figure 11.2.2a uses a set of
individual dynodes. When an ion strikes the first dynode, it generates several electrons, each of which is passed along to the next
dynode before arriving at a collecting plate where the current is measured. The result is an amplification, or gain, in the signal of
approximately 10 ×. The electron multiplier in Figure 11.2.2b uses a horn-shaped cylinder—typically made from glass coated
7

with a thin layer of a semiconducting material—whose surface acts as a single, continuous dynode. When an ion strikes the
continuous dynode it generates several electrons that are reflected toward the collector plate where the current is measured. The
result is an amplification of 10  to 10 × .
5 8

11.2.1 https://chem.libretexts.org/@go/page/387720
 Figure 11.2.2 : Electron multipliers using (a) individual, or discrete dynodes, and (b) a single, continuous dynode. The ion beam is
represented by the blue arrows and the electrons are represented by the green lines. The increasing thickness of the green lines
shows the amplification of the signal.

Faraday Cup
A Faraday cup, as its name suggests, is a simple device shaped like a cup. Ions enter the cup where they strike a collector electrode.
A current is directed to the collector plate that is sufficient to neutralize the charge of the ions. The magnitude of this current is
proportional to the number of ions. A Faraday cup has the advantage of simplicity, but is less sensitive than an electron multiplier
because it lacks the amplification provided by the dynodes.

Separating Ions
Before we can detect the ions, we need to separate them so that we can generate a spectrum that shows the intensity of ions as a
function of their mass-to-charge ratio. In this section we consider the three most common mass analyzers for atomic mass
spectrometry.

Quadrupole Mass Analyzers

The quadrupole mass analyzer is the most important of the mass analyzers included in this chapter: it is compact in size, low in
cost, easy to use, and easy to maintain. As shown in Figure 11.2.3, a quadrapole mass analyzer consists of four cylindrical rods,
two of which are connected to the positive terminal of a variable direct current (dc) power supply and two of which are connected
to the power supply's negative terminal; the two positive rods are positioned opposite of each other and the two negative rods are
positioned opposite of each other. Each pair of rods is also connected to a variable alternating current (ac) source operated such that
the alternating currents are 180° out-of-phase with each other. An ion beam from the source is drawn into the channel between the
quadrupoles and, depending on the applied dc and ac voltages, ions with only one mass-to-charge ratio successfully travel the
length of the mass analyzer and reach the transducer; all other ions collide with one of the four rods and are destroyed.

Figure 11.2.3 : Basic arrangement of the quadrupole mass analyzer. The plus and the minus signs on each rod indicate which are
connected to the positive terminal of the dc power supply. Not shown here is the circuitry for applying the alternating current. The
ion beam enters the channel between the quadrupoles. Depending on the applied dc and ac voltages, some of the ions emerge from
this channel and reach the transducer.
To understand how a quadrupole mass analyzer achieves this separation of ions, it helps to consider the movement of an ion
relative to just two of the four rods, as shown in Figure 11.2.4 for the poles that carry a positive dc voltage. When the ion beam
enters the channel between the rods, the ac voltage causes the ion to begin to oscillate. If, as in the top diagram, the ion is able to

11.2.2 https://chem.libretexts.org/@go/page/387720
maintain a stable oscillation, it will pass through the mass analyzer and reach the transducer. If, as in the middle diagram, the ion is
unable to maintain a stable oscillation, then the ion eventually collides with one of the rods and is destroyed. When the rods have a
positive dc voltage, as they do here, ions with larger mass-to-charge ratios will be slow to respond to the alternating ac voltage and
will pass through the transducer. The result is shown in the figure at the bottom (and repeated in Figure 11.2.5a) where we see that
ions with a sufficiently large mass-to-charge ratios successfully pass through the transducer; ions with smaller mass-to-charge
ratios do not. In this case, the quadrupole mass analyzer acts as a high-pass filter.

Figure 11.2.4 : Illustration of how a quadrupole mass analyzer achieves separation of ions by their mass-to-charge ratio. See the
text for additional details.
We can extend this to the behavior of the ions when they interact with rods that carry a negative dc voltage. In this case, the ions
are attracted to the rods, but those ions that have a sufficiently small mass-to-charge ratio are able to respond to the alternating
current's voltage and remain in the channel between the rods. The ions with larger mass-to-charge ratios move more sluggishly and
eventually collide with one of the rods. As shown in Figure 11.2.5b, in this case, the quadrupole mass analyzer acts as a low-pass
filter. Together, as we see in Figure 11.2.5c, a quadrupole mass analyzer operates as both a high-pass and a low-pass filter, allowing
a narrow band of mass-to-charge ratios to pass through the transducer. By varying the applied dc voltage and the applied ac
voltage, we can obtain a full mass spectrum.

Figure 11.2.5 : Illustration that shows how a quadrupole mass analyzer acts as both (a) a high-pass filter and (b) a low pass filter,
with the result that it passes only ions with a narrow range of mass-to-charge ratios.
Quadrupole mass analyzers provide a modest mass-to-charge resolution of about 1 amu and extend to m/z ratios of approximately
2000. Quadrupole mass analyzers are particularly useful for sources based on plasmas.

Time-of-Flight (TOF) Mass Analyzers

In a time-of-flight mass analyzers, ions are created in small clusters by applying a periodic pulse of energy to the sample using a
laser beam or a beam of energetic particles to ionize the sample. The small cluster of ions are then drawn into a tube by applying an
electric field and then allowed to drift through the tube in the absence of any additional applied field; the tube, for obvious reasons,
is called a drift tube. All of the ions in the cluster enter the drift tube with the same kinetic energy, KE, which means, given

11.2.3 https://chem.libretexts.org/@go/page/387720
 1 2
KE = mv (11.2.1)
2

that the square of an ion's velocity is inversely proportional to the ion's mass. As a result, lighter ions move more quickly than
heavier ions. Flight times are typically less than 30 µs. A time-of-flight mass analyzer provide better resolution than a quadrupole
mass analyzer, but is limited to sources that can be pulsed.

Double-Focusing Mass Analyzers

In a double-focusing mass analyzer, two mechanisms are used to focus a beam of ions onto the transducer. One of the mechanisms
is an electrotatic analyzer that serves to confine the kinetic energy of the ions to a narrow range of energies. The second mechanism
is a magnetic sector analyzer that uses an applied magnetic field to separate the ions by their mass-to-charge ratio. The combination
of two analyzers allows for a significant resolution. More details on this type of mass analyzer is included in Chapter 20.

This page titled 11.2: Mass Spectrometers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

11.2.4 https://chem.libretexts.org/@go/page/387720
11.3: Inductively Coupled Plasma Mass Spectrometer
In Chapter 10 we introduced the inductively coupled plasma (ICP) as a source for atomic emission. The plasma in ICP is formed by
ionizing a flowing stream of argon gas, producing argon ions and electrons. The sample is introduced into the plasma where the
high operating temperature of 6000–8000 K is sufficient to atomize and ionize the sample. In optical ICP we measure the emission
of photons from the atoms and ions that are in excited states. In ICP-MS we use the plasma as a source of ions that we can send to a
mass spectrometer for analysis.

Instrumentation for ICP-MS

An ICP torch operates at room pressure and at an elevated temperature, and a mass spectrometer, as noted in Section 11.2, operates
under a vacuum and at room temperature. This difference in pressure and temperature means that a coupling together of these two
instrument requires an interface than can bring the pressure and temprature in line with the demands of the mass spectrometer.
Figure 11.3.1 provides a schematic diagram of a typical ICP-MS instrument with the ICP torch on the right and the mass
spectrometer's quadrupole mass analyzer and a continuous electron multiplier on the left. In between the two is a two-stage
interface. Note that none of the components in Figure 11.3.1 are drawn to scale.

Figure Schematic diagram of an instrument for an inductively coupled plasma mass

11.3.1 .
spectrometer (ICP-MS), showing the torch and the sample nebulizer that make up the ICP on
the right, and the mass analyzer and the transducer that make up the MS on the left. Also
shown is the interface between the two. The first stage of the interface (1) samples the plasma
from the ICP torch and then further reduces the number of particles that are passed along to
the second stage (2) a set of ion lenses are used to remove electrons, neutral species, and
photons and to focus the ion beam onto the quadrupole.
The first stage of the interface consists of two cone-shaped openings: a sampler cone and a skimmer cone. The hot plasma from the
ICP torch enters the first stage of the interface through the sampler cone, which is a pin-hole with a diameter of approximately 1
mm. Samples in solution form are drawn directly into the ICP torch using a nebulizer. Solid samples are vaporized using a laser (a
process called laser ablation) and the vapor drawn directly into the ICP torch.
A pump is used to drop the pressure in the first stage to approximately 1 torr. The expansion of the plasma as it enters the first stage
results in some cooling of the plasma. The skimmer cone allows a small portion of the plasma in the first stage to pass into the
second stage, which is held at the mass spectrometer's operating pressure of approximately 10–5 torr. A series of ion lenses are used
to narrow the conical dispersion of the plasma, to isolate positive ions from electrons, neutral species, and photons—all of which
will generate a signal if they reach the transducer—and to focus the ion beam onto the quadrupole's entrance.

Atomic Mass Spectra and Interferences

Figure 11.3.2 shows an example of an ICP-MS spectrum for the analysis of a metal coating using laser ablation to volatilize the
sample. The quadrupole mass analyzer operates over a mass-to-charge range of approximately 3 to 300 and can resolve lines that
differ by ±1 m/z. Data are collected either by scanning the quadrupole to provide a survey spectrum of all ions generated in the
plasma, as is the case in Figure 11.3.2, or by peak hopping in which we gather data for just a few discrete mass-to-charge ratios,
adjusting the quadrupole so that it passes only a single mass-to-charge ratio and count the ions for a set period of time before
moving to the next mass-to-charge ratio.

11.3.1 https://chem.libretexts.org/@go/page/387722
 Figure 11.3.2 . Example of an ICP-MS spectrum of a metal coating using laser ablation to vaporize the sample before drawing it
into the ICP torch. The intensity of ions is reported in kilocounts per seconds. The major peaks are identified by element. The data
used to construct this figure were extrapolated from the source linked to here.

Spectroscopic Interferences
An ICP-MS spectrum is much simpler than the corresponding ICP atomic emission spectrum because each element in the latter has
many emission lines and because the plasma itself has many emission lines. Still, an ICP-MS is not free from interferences, the two
most important of which are isobaric ions and polyatomic ions.
Isobaric Ions. Iso- means same and -baric means weight; thus, isobaric means same weight and refers to two (or more) species that
have—within the resolution of the mass spectrometer—identical weights and that both contribute to the same peak in the mass
spectrum. The source of this interference is the existence of isotopes. For example, the most abundant ions for argon and for
calcium are 40Ar and 40Ca, and, given the resolving power of a quadrupole mass analyzer, the two ions appear as a single peak at
40
m/z = 40 even though the mass of Ar is 39.962383 amu and the mass of 40Ca is 39.962591 amu. We can correct for this
interference because the second most abundant isotope of calcium, 44Ca, does not share a m/z with argon (or with another
element). Figure 11.3.3 shows the ICP-MS spectrum for a sample that contains calcium and argon, and Example 11.3.1 shows how
we can use this spectrum to determine the contribution of each element.

Figure 11.3.3 : ICP-MS spectrum for a mixture of calcium and argon. The peak at m/z = 40 includes contributions from both 40Ca
and 40Ar. The other two peaks are for 42Ca and for 44Ca and do not include contributions from argon. The peaks for 36Ar, 38Ar,
43Ca, 46Ca, and 48Ca are too small to appear as peaks in this spectrum.

 Example 11.3.1

For the spectrum in Figure 11.3.3, the intensity at m/z = 40 is 972.07 cps and the intensity at m/z = 44 is 18.77 cps. Given
that the istopic abundance of 40Ca is 96.941% and the isotopic abundance of 44Ca is 2.086%, what is the counts-per-second at
m/z = 40 for Ca and for Ar.

11.3.2 https://chem.libretexts.org/@go/page/387722
 Solution
Given that only 44Ca contributes to the peak at m/z = 44 we can use the relative abundances of 40
Ca and 44Ca to determine
the expected contribution of 40Ca to the total intensity at m/z = 40 .
96.941
18.77 cps × = 872.28 cps
2.086

Subtracting this result from the total intensity gives the intensity at m/z = 40 for argon as

972.07 cps − 872.28 cps = 99.79 cps

Polyatomic Ions. Compensating for isobaric ions is relatively straightforward because we can rely on the known isotopic
abundances of the elements. A more difficult problem is an interference between the isotope of an elemental analyte and a
polyatomic ion that has the same mass. Such polyatomic ions may arise from the sample's matrix or from the plasma. For example,
the ion 40Ar16O+ has a mass-to-charge ratio of 56, which overlaps with peak for 56Fe, the most abundant isotope of iron. Although
we could choose to monitor iron at a different mass-to-charge ratio, we will lose sensitivity as we are using a less abundant isotope.
Corrections can be made using the method outlined in Example 11.3.1, although it may require using multiple peaks, which
increases the uncertainty of the final result.

Matrix Effects
A matrix effect occurs when the sample's matrix affects the relationship between the signal and the concentration of the analyte.
Matrix effects are common in ICP-MS and may lead to either a suppression or an enhancement in the signal. Although not always
well understood, matrix effects likely result from how easily an ionizable element affects the ability to ionize other elements.
Matrix matching, using the method of standard additions, or using an internal standard can help minimize matrix effects for
quantitative work.

Applications of ICP-MS
ICP-MS finds application for analytes in a wide variety of matrices, including both solutions and solids. Solution samples with high
concentrations of dissolved ions may present problems due to the deposition of the salts onto the sampler and skimmer cones,
which reduces the size of the pinhole that provides entry into the interface between the ICP torch and the mass spectrometer. The
use of laser ablation makes it possible to analyzer surfaces—such as glasses, metals, and ceramics—without additional sample
preparation.
Qualitative and Semiquantitative Applications. One of the strengths of ICP-MS is its ability to provide a survey scan, such as
that in Figure 11.3.1, that allows for the identification of the elements present in a sample. Analysis of a single sample that contains
known concentrations of these elements is suitable for providing a rough estimate of their concentration in the sample.
Quantitative Analysis. For a more accurate and precise quantitative analysis, one can prepare multiple external standards and
prepare a calibration curve. Linearly across approximately six orders of magnitude with detection limits of less than 1 ppb.
Including an internal standard in the external standards can help reduce matrix effects. The ideal internal standard will not produce
isobaric ions and its primary ionization potential should be similar to that for the analyte; when working with several analytes, it
may be necessary to choose a different internal standard for each analyte.
Isotope Ratios. An important advantage of ICP-MS over other analytical methods is its ability to monitor multiple isotopes for a
single element.

This page titled 11.3: Inductively Coupled Plasma Mass Spectrometer is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

11.3.3 https://chem.libretexts.org/@go/page/387722
11.4: Other Forms of Atomic Mass Spectrometry
Although ICP-MS is the most widely used method of atomic mass spectrometry, there are other forms of atomic mass
spectrometry, three of which we highlight here.

Spark Source Mass Spectrometry (SSMS)

In SSMS, a solid sample is vaporized using a spark source, as described in Chapter 10.2 for atomic emission. Because the spark is
generated in an evacuated housing the interface between the spark source and the mass spectrometer is simpler. Because the spark
generates ions with a large distribution of kinetic energies a quadrupole mass analyzer is not practicable; instead, the mass
spectrum is recorded using a double-focusing mass analyzer (see Chapter 20 for more details about this type of mass spectrometer).
One advantage of the double-focusing mass analyzer is that it is capable or resolving small differences in masses. For example, in
ICP-MS the peaks for 56Fe+ and the polyatomic ion 40Ar16O+ overlap, appearing as a single peak. A double-focusing mass analyzer
can separate these two ions, which have, respectively, masses of 55.934942 amu and 55.957298 amu.

Glow Discharge Mass Spectrometry (GDMS)

A glow discharge source generates ions in manner similar to that used to generate the emission of photons in a hollow cathode
lamp (see Chapter 9.2 for a discussion of the hollow cathode lamp). The sample serves as the cathode in a cell that contains a very
low pressure of argon gas. The application of a high voltage pulse between the cathode and an anode that also is in the cell,
converts some of the Ar to Ar+ ion, which then collide with the cathode, sputtering some of the solid sample into a mixture of gas-
phase atoms and ions, the later of which are drawn into the mass spectrometer for analysis.

Elemental Surface Analysis by Mass Spectrometry

When analyzing a solid sample, we often are interested in how its composition varies either across the surface or as a function of
depth. We can gather information across a surface if we can focus the ion source to a small spot and then raster that spot across the
surface, and we can gather information as a function of depth if we can use sputter away a portion of the surface. See Chapter 21
for a discussion of two such techniques: secondary ion mass spectometry (SIMS) and laser microprove mass spectrometry.

This page titled 11.4: Other Forms of Atomic Mass Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

11.4.1 https://chem.libretexts.org/@go/page/387724
 CHAPTER OVERVIEW
12: Atomic X-Ray Spectrometry
12.1: Fundamental Principles
12.2: Instrument Components
12.3: Atomic X-Ray Fluorescence Methods
12.4: Other X-Ray Methods

This page titled 12: Atomic X-Ray Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1
12.1: Fundamental Principles
In Chapter 6 we introduced the electromagnetic spectrum and the characteristic properties of photons, such as the wavelengths, the
frequencies, and the energies of ultraviolet, visible, and infrared light. The wavelength range for photons of X-ray radiation extends
from approximately 0.01 nm to 10 nm. Although we are used to reporting a photon's wavelength in nanometers, for historical
reasons the wavelength of an X-ray photon usually is reported in angstroms (for which the symbol is Å) where 1 Å = 0.1 nm; thus
the wavelength range of 0.01 nm to 10 nm for X-ray radiation also is expressed as 0.1 Å to 100 Å. This range of wavelengths
corresponds to a range of frequencies from approximately 3 × 10  s to 3 × 10  s , and a range of energies from 2 × 10  J
19 −1 16 −1 −19

to 2 × 10  J .
−17

Sources of X-Rays
There are three routine ways to generate X-rays, each of which is covered in this section: we can bombard a suitable metal with a
beam of high-energy electrons, we can use one X-ray to stimulate the emission of additional X-rays through fluorescence, and we
can use a radioactive isotope that emits X-rays as it decays.

Obtaining X-Rays From Electron Beam Sources

An electron beam is created by heating a tungsten wire filament to a temperature at which it releases electrons. These electrons are
pulled toward a metal target by applying an accelerating voltage between the metal target and the tungsten wire. The result is the
broad continuum of X-ray emission in Figure 12.1.1. The source of this continuous emission spectrum is the reduction in the
kinetic energy of the electrons as they collide with the metal target. The loss of kinetic energy results in the production of photons
over a broad range of wavelengths and is known as Bremsstrahlung, or braking radiation.

 Note
In earlier chapters we divided the sources of photons into two broad groups: continuous sources, such as a tungsten lamp, that
produce photons at all wavelengths between a lower limit and an upper limit, and line sources, such as a hollow cathode lamp,
that produce photons for one or more discrete wavelengths. The sources used to generate X-rays also generate continuum
and/or line spectra.

Figure 12.1.1 : X-ray emission from a metal target using an electron beam. The voltage listed above each curve is the accelerating
voltage used to create the electron beam. Each accelerating voltage has a minimum wavelength, λ , below which emission does not
0

occur. The dashed line shows λ for an accelerating voltage of 40 kV.

0

The lower wavelength limit for X-ray emission, identified here as λ0 , is the maximum possible loss of kinetic energy, KE, and is
equal to
hc
KE = =V e (12.1.1)
λ0

where h is Planck's constant, c is the speed of light, V is the accelerating voltage, and e is the charge on the electron. The product of
the accelerating voltage and the charge on the electron is the kinetic energy of the electrons. Solving Equation 12.1.1 for λ gives 0

12.1.1 https://chem.libretexts.org/@go/page/389929
 hc 12.398 kV Å
λ0 = = (12.1.2)
V e V

where λ is in angstroms and V is in kilovolts. Note that Equation 12.1.1 and Equation 12.1.2 do not include any terms that depend
0

on the target metal, which means that for any accelerating voltage, λ is the same for all metal targets. Table 12.1.1 gives values of
0

λ that span the range of accelerating voltages in Figure 12.1.1.

0

Table 12.1.1 . Shortest wavelength of continuous X-ray emission when using an electron beam as a function of its accelerating voltage.
accelerating voltage (kV) λ0 (Å)

20 0.62

25 0.50

30 0.41

35 0.35

40 0.31

45 0.28

50 0.25

If we apply a sufficiently large accelerating voltage, then the emission spectrum will consist of both a continuum spectrum and a
line spectrum, as we see in Figure 12.1.2 with molybdenum as the target metal. The spectrum consists of both a continuum similar
to that in Figure 12.1.1, and two lines, one at a wavelength of 0.63 Å and one at a wavelength of 0.71 Å. The source of these lines
is the emission of X-rays from excited state ions that form when a sufficiently high-energy electron from the electron beam
removes an electron from an atomic orbital close to the nucleus. As electrons in atomic orbitals at a greater distance from the
nucleus drop into the atomic orbital with a vacancy, they release their extra energy as a photon.
Although the background emission from the continuum is the same for all metal targets, the energy for the lines have values that
are characteristic for different metals because the energy to remove an electron varies from element-to-element, increasing with
atomic number. For example, an accelerating voltage of at least

12.398 kV Å
V = = 20 kV
0.61 Å

is needed to generate the line spectrum for molybdenum in Figure 12.1.2.

Figure 12.1.2 : The X-ray emission spectrum for molybdenum using an electron beam with an accelerating voltage of 35 kV. The
spectrum consists of both a continuum similar to that in Figure 12.1.1 and two lines, one at a wavelength of 0.63 Å and one at a
wavelength of 0.71 Å.

12.1.2 https://chem.libretexts.org/@go/page/389929
The characteristic emission lines for molybdenum in Figure 12.1.2 are identified as K and K , a notation with which you may
α β

not be familiar. The simplified energy level diagram in Figure 12.1.3 will help us understand this notation. Each arrow in this
energy-level diagram shows a transition in which an electron moves from an orbital at greater distance from the nucleus to an
orbital closer to the nucleus. The letters K, L, and M correspond to the principal quantum number n, which has values of 1, 2, 3...
that indicate the initial vacancy created by the collision of the ion beam with the target metal. The Greek symbols α , β, and γ
indicate the source of the electron that fills this vacancy in terms of its change in the principal quantum number, Δn. An electron
moving from n = 2 to n = 1 and an electron moving from n = 4 to n = 3 have the same designation of α . The emission line in Figure
12.1.2 identified as K , therefore, is the result of an electron in the n = 3 shell moving into a vacancy in the n = 1 shell K .
β

Figure 12.1.3 : Simplified energy-level diagram showing the nomenclature for identifying characteristic X-ray lines.

 Note

Why is Figure 12.1.3 a simplified energy-level diagram? For each n > 1 there is more than one atomic orbital. When n = 2
there are three energy levels: one that corresponds to l = 0, one that corresponds to l = 1 and ml = 0, and one that corresponds to
l = 1 and ml = ±1. The allowed transitions to the n = 1 energy levels requires a change in the value for l; thus, we expect to find
two emission lines from n = 2 to n = 1 instead of the one shown in Figure 12.1.3. These two lines, which we can identify as
K α1 and  K , generally are sufficiently close in value that they are not resolved in the X-ray emission spectrum. For example,
α2

K α1 = 0.709 and  K = 0.714 for molybdenum. You can find a table of X-ray emission lines here.
α2

Obtaining X-Rays From Fluorescent Sources

When an atom in an excited state emits a photon as a means of returning to a lower energy state, how we describe the process
depends on the source of energy that created the excited state. When excitation is the result of thermal energy, we call the process
atomic emission. When excitation is the result of the absorption of a photon, we call the process atomic fluorescence. In X-ray
fluorescence, excitation is brought about using photons from a source of continuous X-ray radiation. More details on X-ray
fluorescence are provided later in this chapter.

Obtaining X-Rays From Radioactive Sources

Atoms that have the same number of protons but a different number of neutrons are isotopes. To identify an isotope we use the
notation E , where E is the element’s atomic symbol, Z is the element’s atomic number, and A is the element’s atomic mass
A
Z

number. Although an element’s different isotopes have the same chemical properties, their nuclear properties are not identical. The
most important difference between isotopes is their stability. The nuclear configuration of a stable isotope remains constant with
time. Unstable isotopes, however, disintegrate spontaneously, emitting radioactive decay particles as they transform into a more
stable form.

An element’s atomic number, Z, is equal to the number of protons and its atomic mass, A, is equal to the sum of the number of
protons and neutrons. We represent an isotope of carbon-13 as C because carbon has six protons and seven neutrons.
13
6

Sometimes we omit Z from this notation—identifying the element and the atomic number is repetitive because all isotopes of

12.1.3 https://chem.libretexts.org/@go/page/389929
 carbon have six protons and any atom that has six protons is an isotope of carbon. Thus, 13C and C–13 are alternative notations
for this isotope of carbon.

Radioactive particles can decay in several ways, one of which results in the emission of X-rays. For example, 55Fe can capture an
electron and undergo a process in which a proton becomes a neutron, becoming 55Mn and releasing the excess energy as a K X- α

ray. We will not give further consideration to radioactive sources of atomic X-ray emission; see Chapter 32, however, for a further
discussion of radioactive methods of analysis.

X-Ray Absorption
Figure 12.1.4 shows a portion of molybdenum's X-ray absorption spectrum over the same range of wavelengths as shown in Figure
12.1.2 for its emission spectrum. Both spectra are relatively simple: the emission spectrum consists of two lines superimposed on a

continuum background, and the absorbance spectrum consists of a single line, identified here as the K edge.

Figure 12.1.4 : A portion of the X-ray absorption spectrum for molybdenum.

The Absorption Process

If an X-ray photon is of sufficient energy, then its absorbance by an atom results in the ejection of an electron from one of the
atom's innermost atomic orbitals, which you may recognize as the production of a photoelectron. For molybdenum, a wavelength
of 0.62 Å (an energy of 20.0 kV) is needed to eject a photoelectron from the K shell (n = 1). At this wavelength the probability of
absorption is at is greatest. At shorter wavelengths (greater energies) there is sufficient energy to eject the electron, however, the
probability of absorption decreases and the relative absorbance decreases slowly. The abrupt decrease in absorbance for
wavelengths larger than 0.62 Å—this abrupt decrease is the source of the term edge—happens because the photons no longer have
sufficient energy to eject an electron from the K shell. The slow increasing absorbance at wavelengths greater than the K edge is
the result of ejecting electrons from the L shell, which has edges at 4.3 Å, 4.7 Å, and 4.9 Å.

 Note
The simplified energy level diagram in 12.1.3 shows only one energy level for n = 2 (the L shell). As we noted earlier, there
are three energy levels when n = 2: one that corresponds to l = 0, one that corresponds to l = 1 and ml = 0, and one that
corresponds to l = 1 and ml = ±1. The three edges corresponding to these energy levels are identified as LI, LII, and LIII.

Beer's Law and X-Ray Absorption

When a source of X-rays passes through a sample with a thickness of x, the following equation holds
P
A = − ln = μM ρx (12.1.3)
P0

where A is the absorbance, P is the power of the X-ray source incident on the sample, P is the power of the X-ray source after it
0

passes through the sample, μ is the sample's mass absorption coefficient and ρ is the sample's density. You may have noticed the
M

similarity between this equation and the equation for Beer's law that we first encountered in Chapter 6

12.1.4 https://chem.libretexts.org/@go/page/389929
 P
A = − ln = ϵbC (12.1.4)
P0

where ϵ is the molar absorptivity, b is the pathlength, and C is molar concentration. Note that both density (g/mL) and molarity
(mol/L) are a measure of concentration that expresses the amount of the absorbing material present in the sample.

X-Ray Fluorescence
When an electron is ejected from a shell near the nucleus by the absorption of an X-ray, the vacancy created is eventually filled
when an electron at a greater distance from the nucleus moves down. Because it takes more energy to eject an electron and create a
vacancy than is returned by the movement of other electrons into the vacancy, the resulting fluorescent emission of X-rays is
always at wavelengths that are longer (lower energy) than the wavelength that was absorbed. We see this in Figure 12.1.4 and
Figure 12.1.1 for molybdenum where it absorbs an X-ray with a wavelength of 0.62 Å and emits X-rays with wavelengths of 0.63
Å and 0.71 Å.

X-Ray Diffraction
When an X-ray beam is focused onto a sample that has a regular (crystalline) pattern of atoms in three dimensions, some of the
radiation scatters from the surface and some of the radiation passes through to the next layer of atoms where the combination of
scattering and passing through continues. As a result of this process, the radiation undergoes diffraction in which X-rays of some
wavelengths appear to reflect off the surface while X-rays of other wavelengths do not. The conditions the result in diffraction are
easy to understand using the diagram in Figure 12.1.5.

Figure 12.1.5 : Illustration showing the diffraction of X-rays from an ordered crystalline sample. See the text for additional details.
The red and green arrows are two parallel beams of X-rays that are focused on an ordered crystalline solid that consists of a layered
repeatable pattern of atoms shown by the blue circles. The two beams of X-rays encounter the solid at an angle of θ . The X-ray
shown in red scatters off of the first layer, exiting at the same angle of θ . The X-ray shown in green penetrates to the second layer
where it undergoes scattering, exiting at the same angle of θ . We know from the superposition of waves (see Chapter 6) that the
two beams of X-rays will remain in phase, and thus experience constructive interference, only if the additional distance traveled by
the green wave—the sum of the line segments bc and cd —is an integer multiple of the wavelength; thus
¯¯¯
¯ ¯
¯¯¯
¯

¯¯¯
¯ ¯
¯¯¯
¯
bc + cd = nλ (12.1.5)

¯¯¯
¯ ¯
¯¯¯
¯
We also know that the length of the line segments bc and cd are given by
¯¯¯
¯ ¯
¯¯¯
¯
bc = cd = d sin θ (12.1.6)

where d is the distance between the crystal's layers. Combining Equation 12.1.5 and Equation 12.1.6 gives
nλ = 2d sin θ (12.1.7)

Rearranging Equation 12.1.7 shows that we will observe diffraction only at angles that satisfy the equation
nλ
sin θ = (12.1.8)
2d

This page titled 12.1: Fundamental Principles is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

12.1.5 https://chem.libretexts.org/@go/page/389929
12.2: Instrument Components
Atomic X-ray spectrometry has the same needs as other forms of optical spectroscopy: a source of X-rays, a means for isolating a
desired range of wavelengths of X-rays, a means for detecting the X-rays, and a means for converting the signal at the transducer
into a meaningful number.

X-Ray Sources
The most important source of X-rays is the X-ray tube, a basic diagram of which is shown in Figure 12.2.1. A beam of electrons
(shown in red) from a heated tungsten filament (shown in orange) serves as a cathode with a negative potential. The electrons are
drawn toward an anode that has a positive potential. The tip of the anode is made from a metal target (shown in blue) that will
produce X-rays (shown in green) with the desired wavelengths when struck by the electron beam. Typical metal targets include
tungsten, molybdenum, silver, copper, iron, and cobalt. The filament and the target metal are housed inside an evacuated tube. The
emitted X-rays exit the tube through an optical window.

Figure 12.2.1 : Schematic diagram showing the key parts of an X-ray tube. See text for details.
Any material that is naturally radioactive emits characteristic X-rays that potentially can serve as a source of X-rays that another
species can absorb. For example, in the absorption spectrum for molybdenum (see Figure 18.1.4) the K line has a wavelength of
α

0.62 Å, which corresponds to an energy of 20.0 kV. A radioactive source with an emission line that has a wavelength slightly
longer than 0.62 Å (between, for example, 0.5 Å and 0.6 Å) is sufficient. One possibility is 109Cd, which emits X-rays with a
wavelength of 0.56 Å, or an energy of 22 kV.

X-Ray Filters and Monochromators

A filter and a monochromator are designed to take a broad range of emission from a source and narrow the range of wavelengths
that reach the sample. Figure 12.2.2 shows how to accomplish this using an absorption filter. The blue line shows the emission
spectrum for a sample that includes two lines—the K line and the K line—superimposed on a broad continuum. The green line
α β

shows the absorption spectrum for a different element whose K edge falls in between the sample's K and K lines. In this case the
α β

K edge filter removes most of the continuum and the K line, allowing just the K line and a small amount of the continuum to
β α

reach the sample.

12.2.1 https://chem.libretexts.org/@go/page/389930
 Figure 12.2.2 : Illustration of how an X-ray filter works by comparing the absorbance spectrum for the filter (in green) to the
emission spectrum of the source (in blue). See the text for additional details.
Figure 12.2.3 shows the basic design for an X-ray monochromator, which can operate in either an absorption mode, in which X-
rays from the source pass through the sample before entering the monochromator, or in an emission mode, in which X-rays from
the source excite the sample and fluorescent emission is sampled at 90°.

Figure 12.2.3 : Illustration of an X-ray monochromator. The black circle shows the location of the sample and the X-ray source,
which can take one of two configurations and experiments: absorption mode or emission mode. See text for additional details.
In either mode, the X-rays pass through a collimator that focuses them onto a crystal where the X-rays undergo diffraction. X-rays
are collected by a second collimator before arriving at the transducer. To scan the source, the crystal rotates through an angle of θ ;
the transducer must rotate twice as fast, traversing an angle of 2θ to maintain an identical angle between the source and the
transducer.
An X-ray monochromator's effective range is determined by the properties of the crystal used for diffraction. We know from
Chapter 12.1 that
nλ = 2d sin θ (12.2.1)

where n is the diffraction order, λ is the wavelength, θ is the X-ray's angle of incidence, and d is the spacing between the crystal's
layers. The practical limit for the angle depends on the monochromator's design, but typically θ is 7.5° to 75° (or 2θ angles of 15°
to 150°). A common crystal is LiF, which has a spacing of 2.01 Å; thus, it provides a wavelength range from a lower limit of
∘
λ = 2d sin θ = 2 × 2.01  Å × sin(7.5 ) = 0.52  Å

to an upper limit of
∘
λ = 2d sin θ = 2 × 2.01  Å × sin(75 Å
) = 3.9 

when n = 1 . This range of wavelengths is sufficient to study the elements K to Cd using their K lines.α

12.2.2 https://chem.libretexts.org/@go/page/389930
X-Ray Transducers
The most common transducers for atomic X-ray spectrometry are the flow proportional counter, the scintillation counter, and the
Si(Li) semiconductor. All three transducers act as photon counters.

Photon Counting
The most common transducer for measuring atomic absorbance and atomic emission of ultraviolet and visible light is a
photomultiplier tube. As we learned in Chapter 7, a photon strikes a photosensitive surface and generates several electrons. These
electrons collide with a series of dynodes, each collision of which generates additional electrons. This amplification of one photon
into 106–107 electrons results in a steady-state current that we can measure. When the intensity of radiation from the source is
smaller, as it is with X-rays, then it is possible to store the electrons in a capacitor that, when discharged, provides a pulsed signal
that carries information about the photons.

Flow Proportional Counters

Figure 12.2.4 shows the basic structure of a flow proportional counter. The transducer's cell has an inlet and an outlet for creating
the flow of argon gas. The cell has windows made from an X-ray transparent materials, such as beryllium. X-rays enter the cell
and, as shown by the reaction in the upper left, ionizes the argon, generating a photoelectron. This photoelectron is sufficiently
energetic that it further ionizes the argon, as shown by the reaction in the lower right. The result is an amplification of a single
photon into as many as 10,000 electrons. These electrons are drawn to a tungsten wire that is held at a positive charge, and then
flow into a capacitor. Discharging the capacitor gives a pulsed signal whose height is proportional to the initial number of electrons
and, therefore, to the energy, frequency, and wavelength of the photons.

Figure 12.2.4 : Schematic diagram of a flow proportional cell for photon counting. See text for details.

Scintillation Counters
A flow proportional counter is not an efficient transducer for shorter wavelength (lower energy) X-rays that are likely to pass
through the cell without being absorbed by the argon gas, leading to a reduction in the signal. In this case we can use a scintillation
counter. Figure 12.2.5 shows how this works. X-ray photons are focused onto a single crystal of NaI that is doped with a small
amount, approximately 0.2%, of Tl+ as an iodide salt. Absorption of the X-rays results in the fluorescent emission of multiple
photons of visible light with a wavelength of 410 nm. Each of these photons falls on the photocathode of a photomultiplier,
eventually producing a voltage pulse. Each pulse corresponds to a single photon with an energy that is proportional to the pulse's
height.

Figure 12.2.5 : Schematic diagram of a scintillation counter. See text for details.

12.2.3 https://chem.libretexts.org/@go/page/389930
Semiconductor Transducers
In Chapter 7.5 we introduced the use of the pn junction of a silicon semiconductor as a transducer for optical spectroscopy.
Absorption of a photon of sufficient energy results in the formation of an electron-hole pair. Movement of the electron through the
n-layer and movement of the hole through the p-region generates a current that is proportional to the number of photons reaching
the detector. Figure 12.2.6 shows the structure of the semiconductor used in monitoring X-rays, which consists of a p-type layer
and an n-type layer on either side of single crystal of silicon doped with lithium or germanium. The Si(Li) layer has the same role
here as Ar has in the flow proportional counter. An X-ray photon that enters into the Si(Li) layer generates electron-hole pairs
leading to a measurable current that is proportional to the energy of the X-ray.

Figure 12.2.6 : Structure of a Si(Li) semiconductor transducer.

X-Ray Signal Processors

The flow proportional counter, scintillation counter, and semiconductor transducers pass a stream of pulses to the signal processor
where pulse-height selector is used to isolate only those pulses of interest and a pulse-height analyzer is used to summarize the
distribution of pulses.

Pulse-Height Selectors
Not all pulses measured by the transducer are of interest. For example, pulses with small heights are likely to be noise and pulses
with large heights may be a higher-order (n > 1 ) diffraction of shorter, and more energetic wavelengths. Figure 12.2.7 shows the
basic details of how pulse-height selector works. The pulse-height selector is set to pass only those pulse heights that are between a
lower limit and an upper limit. The figure shows three pulses, one that is too small (in blue), one that is too large (in red), and one
that we wish to keep (in green). The pulses run through two channels, one that removes only the blue signal and one that retains
only the red signal. The latter signal is inverted and combined with the signal from the other channel. Because the red signal has a
different sign in the two channels, it, too, is removed, leaving only the one pulse height that meets the criteria for selection.

Figure 12.2.7 : Illustration of how a pulse-height selector works. Three pulses arrive at the signal processor shown in blue, green,
and red with heights greater than the baseline (shown as 0). The pulse-height selector is set to remove all pulses that have a height
less than that shown by the low pass limit, and to remove all pulses that have a height greater than the high pass limit. In this case,
the low pass filter removes the pulse in blue, and the high pass filter removes all but the pulse in green. The latter is then inverted
and then added to the former giving a final signal that contains the one pulse that falls within the two limits.
Having removed pulses with heights that are too small or too large, the remaining pulses are analyzed by counting the number of
pulses that share a range of pulse heights. Each unique range of pulse heights is called a channel and corresponds to a specific
energy of the photons. A spectrum is a plot showing the count of pulses as function of the energy of the photons.

12.2.4 https://chem.libretexts.org/@go/page/389930
This page titled 12.2: Instrument Components is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

12.2.5 https://chem.libretexts.org/@go/page/389930
12.3: Atomic X-Ray Fluorescence Methods
In X-ray fluorescence a source of X-rays—emission from an X-ray tube or emission from a radioactive element—is used to excite
the atoms of an analyte in a sample. These excited-state atoms return to their ground state by emitting X-rays, the process we know
as fluorescence. The wavelengths of these emission lines are characteristic of the elements that make up the sample; thus, atomic
X-ray fluorescence is a useful method for both a qualitative analysis and a quantitative analysis.

Instruments
In the previous section we covered the basic components that make up an atomic X-ray spectrometer: a source of X-rays, a means
for isolating those wavelengths of interest, a transducer to measure the intensity of fluorescence, and a signal processor to convert
the transducer's signal into a useful measurement. How we string these units together is the subject of this section in which we
consider two ways to acquire a sample's spectrum: wavelength dispersive instruments and energy dispersive instruments.

Wavelength Dispersive Instruments

A wavelength dispersive instrument relies on diffraction using a monochromator, such as that in Figure 12.2.3, to select the
analytical wavelength. A sequential wavelength dispersive instrument uses a single monochromator. The monochromator's crystal
and transducer are set to the desired angles—θ for the diffracting crystal and 2θ for the transducer—for the analyte of interest and
the fluorescence intensity measured for 1-100 s. The monochromator is adjusted for the next analyte and the process repeated until
the analysis for all analytes is complete. Analyzing a sample for 20 analytes may take 30 min or more.
A simultaneous, or multichannel, wavelength dispersive instrument contains as many as 30 crystals and transducers, each at a fixed
angle that is preset for an analyte of interest. Each individual channel has a dedicated transducer and pulse-height selector and
analyzer. Analysis of a complex sample with many analytes requires less than a minute. This is similar to the multichannel ICP
used in atomic emission (see Figure 10.1.5).

Energy Dispersive Instruments

An energy dispersive instrument eschews a scanning monochromator and, instead, uses a semiconductor transducer to analyze the
fluorescent emission by the determining the energies of the emitted photons. As each photon reaches the transducer as a pulse of
electrons, its height is measured and converted into the photon's energy. The result is a spectrum showing a count of photons with
the same energy as a function of the energy. The collection of data is very fast: if it takes 25 µs to complete the collection and
processing of a single photon, then the instrument can count 40,000 photons each second (40 kcps, or kilo counts per second). One
limitation to an energy dispersive instrument is its limited resolution with respect to energy. An instrument that operates with 2048
channels—that is, an instrument that divides the energies into 2048 bins—and that processes photons with energies up to 20 keV,
has a resolution of approximately 10 eV per channel. Because it does not rely on a monochromator, an energy dispersive instrument
occupies a smaller footprint, and portable, hand-held versions are available.

Qualitative Analysis
Figure 12.3.1 shows the X-ray fluorescence spectrum for the yellow pigment known as naples yellow, the major elements of which
are zinc, lead, and antimony. It is easy to identify the major elements in the sample by matching the energies of the individual lines
to the published emission lines of the elements, which are available in many on-line sources. For example, the first line highlighted
in this spectrum is at an energy of 8.66 KeV, which is close to the K line for Zn at 8.64 KeV, and the last highlighted line is at an
α

energy of 29.97 KeV, which is close to the K line for Sb of 29.7 KeV.
β

12.3.1 https://chem.libretexts.org/@go/page/389931
 Figure 12.3.1 : Energy dispersive XRF of the pigment naples yellow, a lead antimonate that also contains zinc. The spectrometer's
energy scale was calibrated using the K line for 26Fe at 6.4 KeV and the K line for 42Mo at 17.44 eV. The emission lines for zinc
α α

(K and K ) are shown in green, the lines for lead (L , L , and L ) are shown in red, and the lines for antimony (K and K ) are
α β α β γ α β

shown in blue. The original data is available here.

Quantitative Analysis
A semi-quantitative analysis is possible if we assume that there is a linear relationship between the intensity of an element's
emission line and its %w/w concentration in the sample. The intensity of emission from a pure sample or the element, I , is pure

measured along with the intensity of emission for the element in a sample, I , and the %w/w calculated as
sample

Isample
%w/w = (12.3.1)
Ipure

Equation 12.3.1 is essentially a one-point standardization that makes the significant assumption that the intensity of fluorescent
emission is independent of the matrix in which the analyte sits. When this is not true, then errors of 2-3× are likely.

Matrix Effects
For fluorescent emission to occur, the analyte must first absorb a photon that can eject a photoelectron. For Equation 12.3.1 to
hold, the photons that initiate the fluorescent emission must come from the source only. If other elements within the sample's
matrix produced fluorescent emission with sufficient energy to eject photoelectrons from the analyte, then the total fluorescence
increases and we overestimate the analyte's concentration. If an element in the matrix absorbs the X-rays from the source more
strongly than the analyte, then the analyte's total fluorescence becomes smaller and we underestimate the analyte's concentration.
There are three common strategies for compensating for matrix effects.
External Standards with Matrix Matching. Instead of using a single, pure sample for the calibration, we prepare a series of
standards with different concentrations of the analyte. By matching, as best we can, the matrix of the standards to the matrix of the

12.3.2 https://chem.libretexts.org/@go/page/389931
samples, we can improve the accuracy of a quantitative analysis. This assumes, of course, that we have sufficient knowledge of our
sample's matrix.
Internal Standards. An internal standard is an element that we add to the standards and samples so that its concentration is the
same in each. If the analyte and the internal standard experience similar matrix effects, then the ratio of their intensities is
proportional to the ratio of their concentrations
Ianalyte, sample Canalyte, sample
=K× (12.3.2)
Iint std, sample Cint std, sample

Dilution. A third approach is to dilute the samples and standards by adding a quantity of non-absorbing or poorly absorbing
material. Dilution has the effect of minimizing the difference in the matrix of the original samples and standards.

This page titled 12.3: Atomic X-Ray Fluorescence Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

12.3.3 https://chem.libretexts.org/@go/page/389931
12.4: Other X-Ray Methods
The application of X-rays to the analysis of materials can take forms other than X-ray fluorescence. In X-ray absorption
spectrometry, the ability of a sample to absorb radiation from an X-ray source is measured. Absorption follows Beer's law (see
Section 12.1) and, compared to emission, is relatively free of matrix effects. X-ray absorption, however, is a less selective
technique than atomic fluorescence because we are not measuring the emission from an analyte's characteristic lines. X-ray
absorption finds its greatest utility for the quantitative analysis of samples that contain just one or two major analytes.
In powder X-ray diffraction we focus the radiation from an X-ray tube line source on a powdered sample and measure the intensity
of diffracted radiation as a function of the transducer's angle (2θ). A typical powder X-ray diffraction spectrum is in Figure 12.4.1
for the mineral calcite (CaCO3). Qualitative identification is obtained by matching the 2θ peaks to those in published databases. A
quantitative analysis for the compound—not the elements that make up the compound—is possible using the intensity of a unique
diffraction line in a sample to that for a pure sample. Figure 12.4.2 for a mixture of calcite and magnesite (MgCO3) shows that a
simultaneous quantitative analysis for both compounds is possible using the diffraction line at a 2θ of 29.44 for calcite and of 32.65
for magnesite.

Figure 12.4.1 : Powder X-ray diffraction spectrum for the mineral calcite (CaCO3). The original data used to construct this
spectrum is here.

Figure 12.4.2 : Powder diffraction spectra for calcite (CaCO3) and magnesite (MgCO3). A quantitative analysis for calcite is
possible using its most intense line at a 2θ of 29.44 and a quantitative analysis for magnesite is possible using its most intense line
at a 2θ of 32.65. The original data for calcite is here and that for magnesite is here.

This page titled 12.4: Other X-Ray Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

12.4.1 https://chem.libretexts.org/@go/page/389932
 CHAPTER OVERVIEW
13: Introduction to Ultraviolet/Visible Absorption Spectrometry
13.1: Transmittance and Absorbance
13.2: Beer's Law
13.3: Effect of Noise on Transmittance and Absorbance Measurements
13.4: Instrumentation

This page titled 13: Introduction to Ultraviolet/Visible Absorption Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

1
13.1: Transmittance and Absorbance
As light passes through a sample, its power decreases as some of it is absorbed. This attenuation of radiation is described
quantitatively by two separate, but related terms: transmittance and absorbance. As shown in Figure 13.1.1a, transmittance is the
ratio of the source radiation’s power as it exits the sample, PT, to that incident on the sample, P0.
PT
T = (13.1.1)
P0

Multiplying the transmittance by 100 gives the percent transmittance, %T, which varies between 100% (no absorption) and 0%
(complete absorption). All methods of detecting photons—including the human eye and modern photoelectric transducers—
measure the transmittance of electromagnetic radiation.

Figure 13.1.1b. (a) Schematic diagram showing the attenuation of radiation passing through a sample; P0 is the source’s radiant
power and PT is the radiant power transmitted by the sample. (b) Schematic diagram showing how we redefine P0 as the radiant
power transmitted by the blank. Redefining P0 in this way corrects the transmittance in (a) for the loss of radiation due to
scattering, reflection, absorption by the sample’s container, and absorption by the sample’s matrix.
Equation 13.1.1 does not distinguish between different mechanisms that prevent a photon emitted by the source from reaching the
detector. In addition to absorption by the analyte, several additional phenomena contribute to the attenuation of radiation, including
reflection and absorption by the sample’s container, absorption by other components in the sample’s matrix, and the scattering of
radiation. To compensate for this loss of the radiation’s power, we use a method blank. As shown in Figure 13.1.1b, we redefine P0
as the power exiting the method blank.
An alternative method for expressing the attenuation of electromagnetic radiation is absorbance, A, which we define as
PT
A = − log T = − log (13.1.2)
P0

Absorbance is the more common unit for expressing the attenuation of radiation because—as we will see in the next section—it is a
linear function of the analyte’s concentration.

 Example 13.1.1
A sample has a percent transmittance of 50%. What is its absorbance?

Solution
A percent transmittance of 50.0% is the same as a transmittance of 0.500. Substituting into Equation 13.1.2 gives

A = − log T = − log(0.500) = 0.301

 Exercise 13.1.1
What is the %T for a sample if its absorbance is 1.27?

13.1.1 https://chem.libretexts.org/@go/page/365723
 Answer
To find the transmittance, T , we begin by noting that

A = 1.27 = − log T

Solving for T

−1.27 = log T

−1.27
10 =T

gives a transmittance of 0.054, or a %T of 5.4%.

This page titled 13.1: Transmittance and Absorbance is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

13.1.2 https://chem.libretexts.org/@go/page/365723
13.2: Beer's Law
Absorbance and Concentration
When monochromatic electromagnetic radiation passes through an infinitesimally thin layer of sample of thickness dx, it
experiences a decrease in its power of dP (Figure 13.2.1).

Figure 13.2.1 . Factors used to derive the Beer’s law.

This fractional decrease in power is proportional to the sample’s thickness and to the analyte’s concentration, C; thus
dP
− = αC dx (13.2.1)
P

where P is the power incident on the thin layer of sample and α is a proportionality constant. Integrating the left side of Equation
13.2.1 over the sample’s full thickness

P =Pt x=b
dP
−∫ = αC ∫ dx
P =P0
P x=0

P0
ln = αbC
PT

converting from ln to log, and substituting into the equation relating transmittance to absorbance
PT
A = −logT = −log
P0

gives
A = abC (13.2.2)

where a is the analyte’s absorptivity with units of cm–1 conc–1. If we express the concentration using molarity, then we replace a
with the molar absorptivity, ε , which has units of cm–1 M–1.
A = εbC (13.2.3)

The absorptivity and the molar absorptivity are proportional to the probability that the analyte absorbs a photon of a given energy.
As a result, values for both a and ε depend on the wavelength of the absorbed photon.

 Example 13.2.1

A 5.00 × 10 M solution of analyte is placed in a sample cell that has a pathlength of 1.00 cm. At a wavelength of 490 nm,
−4

the solution’s absorbance is 0.338. What is the analyte’s molar absorptivity at this wavelength?

Solution
Solving Equation 13.2.3 for ϵ and making appropriate substitutions gives
A 0.338 −1 −1
ε = = = 676 cm  M
−4
bC (1.00 cm) (5.00 × 10  M)

13.2.1 https://chem.libretexts.org/@go/page/365725
 Exercise 13.2.1
A solution of the analyte from Example 13.2.1 has an absorbance of 0.228 in a 1.00-cm sample cell. What is the analyte’s
concentration?

Answer
Making appropriate substitutions into Beer’s law
−1 −1
A = 0.228 = εbC = (676 M  cm ) (1 cm)C

and solving for C gives a concentration of 3.37 × 10 −4

M.

Equation 13.2.2 and Equation 13.2.3, which establish the linear relationship between absorbance and concentration, are known as
Beer’s law. Calibration curves based on Beer’s law are common in quantitative analyses.

As is often the case, the formulation of a law is more complicated than its name suggests. This is the case, for example, with
Beer’s law, which also is known as the Beer-Lambert law or the Beer-Lambert-Bouguer law. Pierre Bouguer, in 1729, and
Johann Lambert, in 1760, noted that the transmittance of light decreases exponentially with an increase in the sample’s
thickness.
−b
T ∝e

Later, in 1852, August Beer noted that the transmittance of light decreases exponentially as the concentration of the absorbing
species increases.
−C
T ∝e

Together, and when written in terms of absorbance instead of transmittance, these two relationships make up what we know as
Beer’s law.

Beer's Law and Multicomponent Samples

We can extend Beer’s law to a sample that contains several absorbing components. If there are no interactions between the
components, then the individual absorbances, Ai, are additive. For a two-component mixture of analyte’s X and Y, the total
absorbance, Atot, is

Atot = AX + AY = εX b CX + εY b CY

Generalizing, the absorbance for a mixture of n components, Amix, is

n n

Amix = ∑ Ai = ∑ εi b Ci (13.2.4)

i=1 i=1

Limitations to Beer's Law

Beer’s law suggests that a plot of absorbance vs. concentration—we will call this a Beer’s law plot—is a straight line with a y-
intercept of zero and a slope of ab or εb . In some cases a Beer’s law plot deviates from this ideal behavior (see Figure 13.2.2), and
such deviations from linearity are divided into three categories: fundamental, chemical, and instrumental.

13.2.2 https://chem.libretexts.org/@go/page/365725
 Figure 13.2.2 . Plots of absorbance vs. concentration showing positive and negative deviations from the ideal Beer’s law
relationship, which is a straight line.

Fundamental Limitations to Beer's Law

Beer’s law is a limiting law that is valid only for low concentrations of analyte. There are two contributions to this fundamental
limitation to Beer’s law. At higher concentrations the individual particles of analyte no longer are independent of each other. The
resulting interaction between particles of analyte may change the analyte’s absorptivity. A second contribution is that an analyte’s
absorptivity depends on the solution’s refractive index. Because a solution’s refractive index varies with the analyte’s
concentration, values of a and ε may change. For sufficiently low concentrations of analyte, the refractive index essentially is
constant and a Beer’s law plot is linear.

Chemical Limitations to Beer's Law

A chemical deviation from Beer’s law may occur if the analyte is involved in an equilibrium reaction. Consider, for example, the
weak acid, HA. To construct a Beer’s law plot we prepare a series of standard solutions—each of which contains a known total
concentration of HA—and then measure each solution’s absorbance at the same wavelength. Because HA is a weak acid, it is in
equilibrium with its conjugate weak base, A–.

In the equations that follow, the conjugate weak base A– is sometimes written as A as it is easy to mistake the symbol for
anionic charge as a minus sign; thus, we will write C instead of C .
A
−
A

+ −
HA(aq) + H2 O(l) ⇌ H3 O (aq) + A (aq)

–
If both HA and A absorb at the selected wavelength, then Beer’s law is
A = εHA b CHA + εA b CA (13.2.5)

Because the weak acid’s total concentration, Ctotal, is

Ctotal = CHA + CA

we can write the concentrations of HA and A– as

CHA = αHA Ctotal (13.2.6)

CA = (1 − αHA )Ctotal (13.2.7)

where α HA is the fraction of weak acid present as HA. Substituting Equation 13.2.6 and Equation 13.2.7 into Equation 13.2.5 and
rearranging, gives
A = (εHA αHA + εA − εA αA ) b Ctotal (13.2.8)

To obtain a linear Beer’s law plot, we must satisfy one of two conditions. If εHA and εA have the same value at the selected
wavelength, then Equation 13.2.8 simplifies to

A = εA b Ctotal = εHA b Ctotal

Alternatively, if α has the same value for all standard solutions, then each term within the parentheses of Equation
HA 13.2.8 is
constant—which we replace with k—and a linear calibration curve is obtained at any wavelength.

13.2.3 https://chem.libretexts.org/@go/page/365725
 A = kbCtotal

Because HA is a weak acid, the value of α HAvaries with pH. To hold α constant we buffer each standard solution to the same
HA

pH. Depending on the relative values of α and α , the calibration curve has a positive or a negative deviation from Beer’s law
HA A

if we do not buffer the standards to the same pH.

Instrumental Limitations to Beer's Law

There are two principal instrumental limitations to Beer’s law: stray radiation and non-polychormatic radiation.
Stray radiation is the first contribution to instrumental deviations from Beer’s law. Stray radiation arises from imperfections in the
wavelength selector that allow light to enter the instrument and to reach the detector without passing through the sample. Stray
radiation adds an additional contribution, Pstray, to the radiant power that reaches the detector; thus
PT + P stray 
A = − log
P0 + P stray 

For a small concentration of analyte, Pstray is significantly smaller than P0 and PT, and the absorbance is unaffected by the stray
radiation. For higher concentrations of analyte, less light passes through the sample and PT and Pstray become similar in magnitude.
This results is an absorbance that is smaller than expected, and a negative deviation from Beer’s law.
The second limitation is that Beer’s law assumes that radiation reaching the sample is of a single wavelength—that is, it assumes a
purely monochromatic source of radiation. Even the best wavelength selector, however, passes radiation with a small, but finite
effective bandwidth. Let's assume we have a line source that emits light at two wavelengths, λ and λ . When treated separately, ′ ′′

the absorbances at these wavelengths, A′ and A′′, are

′ ′′
Pr P
′ ′ ′′ T ′′
A = − log = ε bC A = − log = ε bC
′ ′′
P P
0 0

If both wavelengths are measured simultaneously the absorbance is

′ ′′
(P +P )
T T
A = − log
′ ′′
(P +P )
0 0

Expanding the logarithmic function of the equation's right side gives

′ ′′ ′ ′′
A = log(P +P ) − log(P +P ) (13.2.9)
0 0 T T

Next, we need to find a relationship between P and P for any wavelength. To do this, we start with Beer's law
T 0

PT
A = − log = ϵbC
P0

and then solve for P in terms of \P

T 0

PT
log = −ϵbC
P0

PT −ϵbC
= 10
P0

−ϵbC
PT = P0 × 10

Substituting this general relationship back into our wavelength-specific equation for absorbance, 13.2.9, we obtain
′ ′′ ′ −ϵbC ′′ −ϵbC
A = log(P +P ) − log(P × 10 +P × 10 ) (13.2.10)
0 0 0 0

For monochromatic radiation, we have ϵ ′ ′′

=ϵ =ϵ and Equation 13.2.10 simplifies to Beer's law
−ϵbC
A = − log(10 ) = ϵbC

For non-monochromatic radiation, Equation 13.2.10 predicts that the absorbance is smaller than expected if ϵ   > ϵ . ′ ′′

Polychromatic radiation always gives a deviation from Beer’s law, but the effect is smaller if the value of ε essentially is constant
over the wavelength range passed by the wavelength selector. For this reason, as shown in Figure 13.2.3, it is better to make

13.2.4 https://chem.libretexts.org/@go/page/365725
absorbance measurements at the top of a broad absorption peak. In addition, the deviation from Beer’s law is less serious if the
source’s effective bandwidth is less than one-tenth of the absorbing species’ natural bandwidth [(a) Strong, F. C., III Anal. Chem.
1984, 56, 16A–34A; Gilbert, D. D. J. Chem. Educ. 1991, 68, A278–A281]. When measurements must be made on a slope, linearity
is improved by using a narrower effective bandwidth.

Figure 13.2.3 . Effect of wavelength selection on the linearity of a Beer’s law plot. Another reason for measuring absorbance at the
top of an absorbance peak is that it provides for a more sensitive analysis. Note that the green Beer’s law plot has a steeper slope—
and, therefore, a greater sensitivity—than the red Beer’s law plot. A Beer’s law plot, of course, is equivalent to a calibration curve.

This page titled 13.2: Beer's Law is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

13.2.5 https://chem.libretexts.org/@go/page/365725
13.3: Effect of Noise on Transmittance and Absorbance Measurements
In absorption spectroscopy, precision is limited by indeterminate errors—primarily instrumental noise—which are introduced when
we measure absorbance. Precision generally is worse for low absorbances where P0 ≈ PT, and for high absorbances where PT
approaches 0. We might expect, therefore, that precision will vary with transmittance.
We can derive an expression between precision and transmittance by rewriting Beer's law as
1
C =− log T (13.3.1)
εb

and completing a propagation of uncertainty (see the Appendicies for a discussion of propagation of error), which gives
0.4343 sT
sc = − × (13.3.2)
εb T

where sT is the absolute uncertainty in the transmittance. Dividing Equation 13.3.2 by Equation 13.3.1 gives the relative
uncertainty in concentration, sC/C, as
sc 0.4343sT
=
C T log T

If we know the transmittance’s absolute uncertainty, then we can determine the relative uncertainty in concentration for any
measured transmittance.
Determining the relative uncertainty in concentration is complicated because sT is a function of the transmittance. As shown in
Table 13.3.1, three categories of indeterminate instrumental error are observed [Rothman, L. D.; Crouch, S. R.; Ingle, J. D. Jr. Anal.
Chem. 1975, 47, 1226–1233].
Table 13.3.1 . Effect of Indeterminate Errors on Relative Uncertainty in Concentration
category sources of indeterminate error relative uncertainty in concentration

%T readout resolution sC 0.4343k1

sT = k1 =
noise in thermal detectors C T log T

−−−−−
−−− −−− sC 0.4343k2
sT = k2 √T
2
+T noise in photon detectors C
=
log T
√1 +
1

positioning of sample cell sC 0.4343k3

sT = k3 T =
fluctuations in source intensity C log T

A constant sT is observed for the uncertainty associated with reading %T on a meter’s analog or digital scale, both common on less-
expensive spectrophotometers. Typical values are ±0.2–0.3% (a k1 of ±0.002–0.003) for an analog scale and ±0.001% a (k1 of
±0.00001) for a digital scale. A constant sT also is observed for the thermal transducers used in infrared spectrophotometers. The
effect of a constant sT on the relative uncertainty in concentration is shown by curve A in Figure 13.3.1. Note that the relative
uncertainty is very large for both high absorbances and low absorbances, reaching a minimum when the absorbance is 0.4343. This
source of indeterminate error is important for infrared spectrophotometers and for inexpensive UV/Vis spectrophotometers. To
obtain a relative uncertainty in concentration of ±1–2%, the absorbance is kept within the range 0.1–1.

13.3.1 https://chem.libretexts.org/@go/page/365727
 Figure 13.3.1 . Percent relative uncertainty in concentration as a function of absorbance for the categories of indeterminate errors in
Table 13.3.10.3 . A: k1 = ±0.0030; B: k2 = ±0.0030; and C: k3= ±0.0130. The dashed lines correspond to the minimum uncertainty
for curve A (absorbance of 0.4343) and for curve B (absorbance of 0.963).
Values of sT are a complex function of transmittance when indeterminate errors are dominated by the noise associated with photon
detectors. Curve B in Figure 13.3.1 shows that the relative uncertainty in concentration is very large for low absorbances, but is
smaller at higher absorbances. Although the relative uncertainty reaches a minimum when the absorbance is 0.963, there is little
change in the relative uncertainty for absorbances between 0.5 and 2. This source of indeterminate error generally limits the
precision of high quality UV/Vis spectrophotometers for mid-to-high absorbances.
Finally, the value of sT is directly proportional to transmittance for indeterminate errors that result from fluctuations in the source’s
intensity and from uncertainty in positioning the sample within the spectrometer. The latter is particularly important because the
optical properties of a sample cell are not uniform. As a result, repositioning the sample cell may lead to a change in the intensity of
transmitted radiation. As shown by curve C in Figure 13.3.1, the effect is important only at low absorbances. This source of
indeterminate errors usually is the limiting factor for high quality UV/Vis spectrophotometers when the absorbance is relatively
small.
When the relative uncertainty in concentration is limited by the %T readout resolution, it is possible to improve the precision of the
analysis by redefining 100% T and 0% T. Normally 100% T is established using a blank and 0% T is established while preventing
the source’s radiation from reaching the detector. If the absorbance is too high, precision is improved by resetting 100% T using a
standard solution of analyte whose concentration is less than that of the sample (Figure 13.3.2a). For a sample whose absorbance is
too low, precision is improved by redefining 0% T using a standard solution of the analyte whose concentration is greater than that
of the analyte (Figure 13.3.2b). In this case a calibration curve is required because a linear relationship between absorbance and
concentration no longer exists. Precision is further increased by combining these two methods (Figure 13.3.2c). Again, a
calibration curve is necessary since the relationship between absorbance and concentration is no longer linear.

Figure 13.3.17 . Methods for improving the precision of absorption methods: (a) high-absorbance method; (b) low-absorbance
method; (c) maximum precision method.

13.3.2 https://chem.libretexts.org/@go/page/365727
This page titled 13.3: Effect of Noise on Transmittance and Absorbance Measurements is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by David Harvey.

13.3.3 https://chem.libretexts.org/@go/page/365727
13.4: Instrumentation
Basic Components
As covered in Chapter 7, the basic instrumentation for absorbance measurements consists of a source of radiation, a means for
selecting the wavelengths to use, a means for detecting the amount of light absorbed by the sample, and a means for processing and
displaying the data. In this section we consider two other essential components of an instrument for measuring the absorbance of
UV/Vis radiation by molecules: the optical path that connects the source to the detector and a means for placing the sample in this
optical path.

Instrument Designs for Molecular UV/Vis Absorption

Frequently an analyst must select from among several different optical paths, the one that is best suited for a particular analysis. In
this section we examine several different instruments for molecular absorption spectroscopy with an emphasis on their advantages
and limitations.

Filter Photometer
The simplest instrument for molecular UV/Vis absorption is a filter photometer (Figure 13.4.1), which uses an absorption or
interference filter to isolate a band of radiation. The filter is placed between the source and the sample to prevent the sample from
decomposing when exposed to higher energy radiation. A filter photometer has a single optical path between the source and
detector, and is called a single-beam instrument. The instrument is calibrated to 0% T while using a shutter to block the source
radiation from the detector. After opening the shutter, the instrument is calibrated to 100% T using an appropriate blank. The blank
is then replaced with the sample and its transmittance measured. Because the source’s incident power and the sensitivity of the
detector vary with wavelength, the photometer is recalibrated whenever the filter is changed. Photometers have the advantage of
being relatively inexpensive, rugged, and easy to maintain. Another advantage of a photometer is its portability, making it easy to
take into the field. Disadvantages of a photometer include the inability to record an absorption spectrum and the source’s relatively
large effective bandwidth, which limits the calibration curve’s linearity.

The percent transmittance varies between 0% and 100%. We use a blank to determine P0, which corresponds to 100%T. Even
in the absence of light the detector records a signal. Closing the shutter allows us to assign 0%T to this signal. Together, setting
0% T and 100%T calibrates the instrument. The amount of light that passes through a sample produces a signal that is greater
than or equal to 0%T and smaller than or equal to 100%T.

Figure 13.4.1. Schematic diagram of a filter photometer. The analyst either inserts a removable filter or the filters are placed in a
carousel, an example of which is shown in the photographic inset. The analyst selects a filter by rotating it into place.

13.4.1 https://chem.libretexts.org/@go/page/365729
Single-Beam Spectrophotometer
An instrument that uses a monochromator for wavelength selection is called a spectrophotometer. The simplest spectrophotometer
is a single-beam instrument equipped with a fixed-wavelength monochromator (Figure 13.4.2). Single-beam spectrophotometers
are calibrated and used in the same manner as a photometer. One example of a single-beam spectrophotometer is Thermo
Scientific’s Spectronic 20D+, which is shown in the photographic insert to Figure 13.4.2. The Spectronic 20D+ has a wavelength
range of 340–625 nm (950 nm when using a red-sensitive detector), and a fixed effective bandwidth of 20 nm. Battery-operated,
hand-held single-beam spectrophotometers are available, which are easy to transport into the field. Other single-beam
spectrophotometers also are available with effective bandwidths of 2–8 nm. Fixed wavelength single-beam spectrophotometers are
not practical for recording spectra because manually adjusting the wavelength and recalibrating the spectrophotometer is awkward
and time-consuming. The accuracy of a single-beam spectrophotometer is limited by the stability of its source and detector over
time.

Figure 13.4.2 . Schematic diagram of a fixed-wavelength, single-beam spectrophotometer. The photographic inset shows a typical
instrument. The shutter remains closed until the sample or blank is placed in the sample compartment. The analyst manually selects
the wavelength by adjusting the wavelength dial. Inset photo modified from: Adi (www.commons.Wikipedia.org).

Double-Beam Spectrophotometer
The limitations of a fixed-wavelength, single-beam spectrophotometer is minimized by using a double-beam spectrophotometer
(Figure 13.4.3). A chopper controls the radiation’s path, alternating it between the sample, the blank, and a shutter. The signal
processor uses the chopper’s speed of rotation to resolve the signal that reaches the detector into the transmission of the blank, P0,
and the sample, PT. By including an opaque surface as a shutter, it also is possible to continuously adjust 0%T. The effective
bandwidth of a double-beam spectrophotometer is controlled by adjusting the monochromator’s entrance and exit slits. Effective
bandwidths of 0.2–3.0 nm are common. A scanning monochromator allows for the automated recording of spectra. Double-beam
instruments are more versatile than single-beam instruments, being useful for both quantitative and qualitative analyses, but also
are more expensive and not particularly portable.

13.4.2 https://chem.libretexts.org/@go/page/365729
 Figure 13.4.3 . Schematic diagram of a scanning, double-beam spectrophotometer. A chopper directs the source’s radiation, using a
transparent window to pass radiation to the sample and a mirror to reflect radiation to the blank. The chopper’s opaque surface
serves as a shutter, which allows for a constant adjustment of the spectrophotometer’s 0%T. The photographic insert shows a
typical instrument. The module in the middle of the photo is a temperature control unit that makes it possible to heat or cool the
sample to a constant temperature.

Diode Array Spectrometer

An instrument with a single detector can monitor only one wavelength at a time. If we replace a single photomultiplier with an
array of photodiodes, we can use the resulting detector to record a full spectrum in as little as 0.1 s. In a diode array spectrometer
the source radiation passes through the sample and is dispersed by a grating (Figure 13.4.4). The photodiode array detector is
situated at the grating’s focal plane, with each diode recording the radiant power over a narrow range of wavelengths. Because we
replace a full monochromator with just a grating, a diode array spectrometer is small and compact.

Figure 13.4.4 . Schematic diagram of a diode array spectrophotometer. The photographic insert shows a typical instrument. Note
that the 50-mL beaker provides a sense of scale. Because the spectrometer is small and compact, it is easy to transport into the
field.
One advantage of a diode array spectrometer is the speed of data acquisition, which allows us to collect multiple spectra for a
single sample. Individual spectra are added and averaged to obtain the final spectrum. This signal averaging improves a spectrum’s
signal-to-noise ratio. If we add together n spectra, the sum of the signal at any point, x, increases as nSx, where Sx is the signal. The
−
noise at any point, Nx, is a random event, which increases as √nN when we add together n spectra. The signal-to-noise ratio
x

after n scans, (S/N)n is

S nSx − Sx
( ) = − = √n
N √n Nx Nx
n

where Sx/Nx is the signal-to-noise ratio for a single scan. The impact of signal averaging is shown in Figure 13.4.5. The first
spectrum shows the signal after one scan, which consists of a single, noisy peak. Signal averaging using 4 scans and 16 scans

13.4.3 https://chem.libretexts.org/@go/page/365729
decreases the noise and improves the signal-to-noise ratio. One disadvantage of a photodiode array is that the effective bandwidth
per diode is roughly an order of magnitude larger than that for a high quality monochromator.

Figure 13.4.5 . Effect of signal averaging on a spectrum’s signal-to-noise ratio. From top to bottom: spectrum for a single scan;
average spectrum after four scans; and average spectrum after adding 16 scans.

Sample Cells
The sample compartment provides a light-tight environment that limits stray radiation. Samples normally are in a liquid or solution
state, and are placed in cells constructed with UV/Vis transparent materials, such as quartz, glass, and plastic (Figure 13.4.6). A
quartz or fused-silica cell is required when working at a wavelength <300 nm where other materials show a significant absorption.
The most common pathlength is 1 cm (10 mm), although cells with shorter (as little as 0.1 cm) and longer pathlengths (up to 10
cm) are available. Longer pathlength cells are useful when analyzing a very dilute solution or for gas samples. The highest quality
cells allow the radiation to strike a flat surface at a 90o angle, minimizing the loss of radiation to reflection. A test tube often is used
as a sample cell with simple, single-beam instruments, although differences in the cell’s pathlength and optical properties add an
additional source of error to the analysis.

FFigure 13.4.6 . Examples of sample cells for UV/Vis spectroscopy. From left to right (with path lengths in parentheses):
rectangular plastic cuvette (10.0 mm), rectangular quartz cuvette (5.000 mm), rectangular quartz cuvette (1.000 mm), cylindrical
quartz cuvette (10.00 mm), cylindrical glass cuvette with quartz windows (100.0 mm). Cells often are available as a matched pair,
which is important when using a double-beam instrument.
If we need to monitor an analyte’s concentration over time, it may not be possible to remove samples for analysis. This often is the
case, for example, when monitoring an industrial production line or waste line, when monitoring a patient’s blood, or when
monitoring an environmental system, such as stream. With a fiber-optic probe we can analyze samples in situ. An example of a
remote sensing fiber-optic probe is shown in FFigure 13.4.7. The probe consists of two bundles of fiber-optic cable. One bundle
transmits radiation from the source to the probe’s tip, which is designed to allow the sample to flow through the sample cell.
Radiation from the source passes through the solution and is reflected back by a mirror. The second bundle of fiber-optic cable
transmits the nonabsorbed radiation to the wavelength selector. Another design replaces the flow cell shown in Figure 13.4.7 with a
membrane that contains a reagent that reacts with the analyte. When the analyte diffuses into the membrane it reacts with the
reagent, producing a product that absorbs UV or visible radiation. The nonabsorbed radiation from the source is reflected or
scattered back to the detector. Fiber optic probes that show chemical selectivity are called optrodes [(a) Seitz, W. R. Anal. Chem.
1984, 56, 16A–34A; (b) Angel, S. M. Spectroscopy 1987, 2(2), 38–48].

13.4.4 https://chem.libretexts.org/@go/page/365729
 Figure 13.4.7 . Example of a fiber-optic probe. The inset photographs at the bottom of the figure provide close-up views of the
probe’s flow cell and the reflecting mirror.

This page titled 13.4: Instrumentation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

13.4.5 https://chem.libretexts.org/@go/page/365729
 CHAPTER OVERVIEW
14: Applications of Ultraviolet/Visible Molecular Absorption Spectrometry
14.1: What is Molar Absorptivity?
14.2: Absorbing Species
14.3: Qualitative and Characterization Applications
14.4: Quantitative Applications
14.5: Photometric Titrations

This page titled 14: Applications of Ultraviolet/Visible Molecular Absorption Spectrometry is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by David Harvey.

1
14.1: What is Molar Absorptivity?
Beer's law, as we learned in Chapter 13, gives the relationship between the amount of light absorbed by a sample, A , the
concentration of the species absorbing light, C , the distance (path length) the light travels through the sample, b , and the molar
absorptivity of the species absorbing light, ϵ

A = ϵbC

The meaning of path length and concentration are self-evident, and their effect on the extent of absorbance also are self-evident: the
more absorbing species that are present (concentration) and the more opportunity for any one molecule to absorb light (path
length), the greater the absorbance. The meaning of molar absorptivity—what it represents—is less intuitive. It is, of course, a
proportionality constant that converts the product of path length and concentration, bC , into absorbance, but that is not a
particularly satisfying definition. Maximum values for ϵ are on the order of 10 L/(mol•cm) for simple molecules. and are
5

proportional to the cross-sectional area of the absorbing species and the probability that a photon passing through this cross-
sectional area is absorbed. Here we have a self-evident relationship: the greater the cross-sectional area—the more space occupied
by the absorbing species—the greater the opportunity for absorbance; and the more favorable the probability of absorption—with
probabilities ranging from 0 to 1—the greater the absorbance.
Although molar absorptivity values are often reported in the literature, their values usually vary significantly from study-to-study,
presumably due to differences in the purity of the reagents, the solvents used to prepare solutions, the precision with which path
length is measured, and the instrument used for the measurements. For this reason, molar absorptivity values are usually calculated
as needed by making careful measurements of A , b , and C , or by simply reducing Beer's law to A = kC where k is determined
from a calibration curve.

This page titled 14.1: What is Molar Absorptivity? is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

14.1.1 https://chem.libretexts.org/@go/page/365731
14.2: Absorbing Species
There are two general requirements for an analyte’s absorption of electromagnetic radiation. First, there must be a mechanism by
which the radiation’s electric field or magnetic field can interact with the analyte. For ultraviolet and visible radiation, absorption of
a photon changes the energy of the analyte’s valence electrons. The second requirement is that the photon’s energy, hν , must
exactly equal the difference in energy, ΔE, between two of the analyte’s quantized energy states. We can use the energy level
diagram in Figure 14.2.1 to explain an absorbance spectrum. The lines labeled E0 and E1 represent the analyte’s ground (lowest)
electronic state and its first electronic excited state. Superimposed on each electronic energy level is a series of lines representing
vibrational energy levels.

Figure 14.2.1 . Diagram showing two electronic energy levels (E0 and E1), each with five vibrational energy levels ν . 0−4

Absorption of ultraviolet and visible radiation (shown by the blue arrows) leads to a change in the analyte’s electronic energy levels
and, possibly, a change in vibrational energy as well. A change in vibrational energy without a change in electronic energy levels
occurs with the absorption of infrared radiation (shown by the red arrows).

UV/Vis Spectra for Molecules and Ions

The valence electrons in organic molecules and polyatomic ions, such as CO , occupy quantized sigma bonding (σ), pi bonding (
2−
3

π), and non-bonding (n) molecular orbitals (MOs). Unoccupied sigma antibonding (σ ) and pi antibonding (π ) molecular orbitals
∗ ∗

are slightly higher in energy. Because the difference in energy between the highest-energy occupied molecular orbitals (HOMO)
and the lowest-energy unoccupied molecular orbitals (LUMO) corresponds to ultraviolet and visible radiation, absorption of a
photon is possible.
Four types of transitions between quantized energy levels account for most molecular UV/Vis spectra. Table 14.2.1 lists the
approximate wavelength ranges for these transitions, as well as a partial list of bonds, functional groups, or molecules responsible
for these transitions. Of these transitions, the most important are n → π and π → π because they involve important functional
∗ ∗

groups that are characteristic of many analytes and because the wavelengths are easily accessible. The bonds and functional groups
that give rise to the absorption of ultraviolet and visible radiation are called chromophores.
Table 14.2.1 . Electronic Transitions Involving n, σ , and π Molecular Orbitals
transition wavelength range examples

σ → σ
∗
<200 nm C—C, C—H

n → σ
∗
160–260 nm H2O, CH3OH, CH3Cl

π → π
∗
200–500 nm C=C, C=O, C=N, C≡C

n → π
∗
250–600 nm C=O, C=N, N=N, N=O

Many transition metal ions, such as Cu2+ and Co2+, form colorful solutions because the metal ion absorbs visible light. The
transitions that give rise to this absorption are valence electrons in the metal ion’s d-orbitals, which are shown in Figure 14.2.2. For
a free metal ion, the five d-orbitals are of equal energy.

14.2.1 https://chem.libretexts.org/@go/page/365737
 Figure 14.2.2 : Models showing the 3d orbitals: (A) dz2; (B) dxz; (C) dyz; (D) dxy; and (E) dx2–y2. The first and last of these orbitals
have electron density aligned along the x-, y-, and z-axes.
In the presence of a complexing ligand or solvent molecule, however, the d-orbitals split into two or more groups that differ in
energy. For example, in an octahedral complex of Cu(H O) the six water molecules, which are aligned with the metal rods in
2
2+

Figure 14.2.2, perturb the d-orbitals into the two groups shown in Figure 14.2.3. The magnitude of the splitting of the d -orbitals is
called the octahedral field strength, Δ . oct

Figure 14.2.3 . Splitting of the d-orbitals in an octahedral field. The energy gap, hν , is called Δ
oct .
Although the magnitude of the resulting d → d transitions for transition metal ions are relatively weak, solutions of the metal-
ligand complexes show distinct colors that depend on the metal ion and the ligand, which affect the magnitude of Δ . Figure oct

14.2.4 shows the variation in color for a series of seven octahedral complexes of Co
3+. The spectra for three of these complexes are

shown in Figure 14.2.5, which we can use to estimate the relative size of Δ . Each of the spectra shows two absorption bands,
oct

one near 400 nm and one a somewhat longer wavelength: a shoulder at about 470 nm for phenanthroline, a peak at about 550 nm
for glycine, and a peak at about 620 nm for oxalate. Because Δ is inversely proportional to wavelength, the relative magnitude
oct

of Δ increases from Co(phen)33+ to Co(glycine)33+ to Co(oxalate)33–. In Figure 14.2.4, the octahedral field strengths of the
oct

ligands decrease from Co(NO2)63– to Co(CO3)33–.

Figure 14.2.4 : Metal-ligand complexes of Co3+ with nitrite, phenanthroline, ethylenediamine, glycine, water, oxalate, and
carbonate. The ligands are arranged from largest octahedral field strength on the left to the smallest on the right.

14.2.2 https://chem.libretexts.org/@go/page/365737
 Figure 14.2.5 : Visible spectra for the metal-ligand complexes of Co3+ with phenanthroline, glycine, and oxalate. See Figure 14.2.4
to see the solution colors.

A more important source of UV/Vis absorption for inorganic metal–ligand complexes is charge transfer, in which absorption of a
photon produces an excited state in which there is transfer of an electron from the metal, M, to the ligand, L.
+ − ∗
M − L + hν → (M −L )

Charge-transfer absorption is important because it produces very large absorbances. One important example of a charge-transfer
complex is that of o-phenanthroline with Fe2+, the UV/Vis spectrum for which is shown in Figure 14.2.7. Charge-transfer
absorption in which an electron moves from the ligand to the metal also is possible.

Figure 14.2.6 . UV/Vis spectrum for the metal–ligand complex Fe(phen) , where phen is the ligand o-phenanthroline.
2+

This page titled 14.2: Absorbing Species is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

14.2.3 https://chem.libretexts.org/@go/page/365737
14.3: Qualitative and Characterization Applications
Qualitative Applications
As discussed in Chapter 14.2, ultraviolet, visible, and infrared absorption bands result from the absorption of electromagnetic
radiation by specific valence electrons or bonds. The energy at which the absorption occurs, and the intensity of that absorption, is
determined by the chemical environment of the absorbing moiety. For example, benzene has several ultraviolet absorption bands
due to π → π transitions. The position and intensity of two of these bands, 203.5 nm (ϵ = 7400 M–1cm–1) and 254 nm (ϵ = 204
∗

M–1 cm–1), are sensitive to substitution. For benzoic acid, in which a carboxylic acid group replaces one of the aromatic hydrogens,
the two bands shift to 230 nm (ϵ = 11600 M–1 cm–1) and 273 nm (ϵ = 970 M–1 cm–1). A variety of rules have been developed to aid
in correlating UV/Vis absorption bands to chemical structure. With the availability of computerized data acquisition and storage it
is possible to build digital libraries of standard reference spectra. The identity of an a unknown compound often can be determined
by comparing its spectrum against a library of reference spectra, a process known as spectral searching.

Characterization Applications
Molecular absorption, particularly in the UV/Vis range, has been used for a variety of different characterization studies, including
determining the stoichiometry of metal–ligand complexes and determining equilibrium constants. Both of these examples are
examined in this section.

Stoichiometry of a Metal-Ligand Complex

We can determine the stoichiometry of the metal–ligand complexation reaction

M + yL ⇌ MLy

using one of three methods: the method of continuous variations, the mole-ratio method, and the slope-ratio method. Of these
approaches, the method of continuous variations, also called Job’s method, is the most popular. In this method a series of solutions
is prepared such that the total moles of metal and of ligand, ntotal, in each solution is the same. If (nM)i and (nL)i are, respectively,
the moles of metal and ligand in solution i, then

n total  =  (nM )i   +  (nL )i

The relative amount of ligand and metal in each solution is expressed as the mole fraction of ligand, (XL)i, and the mole fraction of
metal, (XM)i,
(nL )
i
(XL ) =
i
ntotal

(nL ) (nM )
i i
(XM ) =1− =
i
n total  n total 

The concentration of the metal–ligand complex in any solution is determined by the limiting reagent, with the greatest
concentration occurring when the metal and the ligand are mixed stoichiometrically. If we monitor the complexation reaction at a
wavelength where only the metal–ligand complex absorbs, a graph of absorbance versus the mole fraction of ligand has two linear
branches—one when the ligand is the limiting reagent and a second when the metal is the limiting reagent. The intersection of the
two branches represents a stoichiometric mixing of the metal and the ligand. We use the mole fraction of ligand at the intersection
to determine the value of y for the metal–ligand complex MLy.
nL XL XL
y = = =
nM XM 1 − XL

You also can plot the data as absorbance versus the mole fraction of metal. In this case, y is equal to (1 – XM)/XM.

 Example 14.3.1

To determine the formula for the complex between Fe2+ and o-phenanthroline, a series of solutions is prepared in which the
total concentration of metal and ligand is held constant at 3.15 × 10 M. The absorbance of each solution is measured at a
−4

14.3.1 https://chem.libretexts.org/@go/page/365735
 wavelength of 510 nm. Using the following data, determine the formula for the complex.

XL absorbance XL absorbance

0.000 0.000 0.600 0.693

0.100 0.116 0.700 0.809

0.200 0.231 0.800 0.693

0.300 0.347 0.900 0.347

0.400 0.462 1.000 0.000

0.500 0.578

Solution
A plot of absorbance versus the mole fraction of ligand is shown in Figure 14.3.1. To find the maximum absorbance, we
extrapolate the two linear portions of the plot. The two lines intersect at a mole fraction of ligand of 0.75. Solving for y gives
XL 0.75
y = = =3
1 − XL 1 − 0.75

The formula for the metal–ligand complex is Fe(phen) 2+

3
.

Figure 14.3.1 . Continuous variations plot for Example 14.3.1 . The photo shows the solutions used to gather the data. Each solution
is displayed directly below its corresponding point on the continuous variations plot. To prepare these solutions I first prepared a
solution of 3.15 × 10 M Fe2+ and a solution of 3.15 × 10 M o-phenanthroline. Because the two stock solutions have the same
−4 −4

concentration, diluting a portion of one solution with the other solution gives a mixture in which the combined concentration of o-
phenanthroline and Fe2+ is 3.15 × 10 M. Because each solution has the same volume, each solution also contains the same total
−4

moles of metal and ligand.

 Exercise 14.3.1

Use the continuous variations data in the following table to determine the formula for the complex between Fe2+ and SCN–.
The data for this problem is adapted from Meloun, M.; Havel, J.; Högfeldt, E. Computation of Solution Equilibria, Ellis
Horwood: Chichester, England, 1988, p. 236.

14.3.2 https://chem.libretexts.org/@go/page/365735
 XL absorbance XL absorbance XL absorbance XL absorbance

0.0200 0.068 0.2951 0.670 0.5811 0.790 0.8923 0.324

0.0870 0.262 0.3887 0.767 0.6860 0.701 0.9787 0.071

0.1792 0.471 0.4964 0.807 0.7885 0.540

Answer
The figure below shows a continuous variations plot for the data in this exercise. Although the individual data points show
substantial curvature—enough curvature that there is little point in trying to draw linear branches for excess metal and
excess ligand—the maximum absorbance clearly occurs at XL ≈ 0.5. The complex’s stoichiometry, therefore, is Fe(SCN)2+.

Several precautions are necessary when using the method of continuous variations. First, the metal and the ligand must form only
one metal–ligand complex. To determine if this condition is true, plots of absorbance versus XL are constructed at several different
wavelengths and for several different values of ntotal. If the maximum absorbance does not occur at the same value of XL for each
set of conditions, then more than one metal–ligand complex is present. A second precaution is that the metal–ligand complex’s
absorbance must obey Beer’s law. Third, if the metal–ligand complex’s formation constant is relatively small, a plot of absorbance
versus XL may show significant curvature. In this case it often is difficult to determine the stoichiometry by extrapolation. Finally,
because the stability of a metal–ligand complex may be influenced by solution conditions, it is necessary to control carefully the
composition of the solutions. When the ligand is a weak base, for example, each solutions must be buffered to the same pH.
In the mole-ratio method the moles of one reactant, usually the metal, is held constant, while the moles of the other reactant is
varied. The absorbance is monitored at a wavelength where the metal–ligand complex absorbs. A plot of absorbance as a function
of the ligand-to-metal mole ratio, nL/nM, has two linear branches that intersect at a mole–ratio corresponding to the complex’s
formula. Figure 14.3.2a shows a mole-ratio plot for the formation of a 1:1 complex in which the absorbance is monitored at a
wavelength where only the complex absorbs. Figure 14.3.2b shows a mole-ratio plot for a 1:2 complex in which all three species—
the metal, the ligand, and the complex—absorb at the selected wavelength. Unlike the method of continuous variations, the mole-
ratio method can be used for complexation reactions that occur in a stepwise fashion if there is a difference in the molar
absorptivities of the metal–ligand complexes, and if the formation constants are sufficiently different. A typical mole-ratio plot for
the step-wise formation of ML and ML2 is shown in Figure 14.3.2c.

Figure 14.3.2a . Mole-ratio plots for: (a) a 1:1 metal–ligand complex in which only the complex absorbs; (b) a 1:2 metal–ligand
complex in which the metal, the ligand, and the complex absorb; and (c) the stepwise formation of a 1:1 and a 1:2 metal–ligand
complex.

14.3.3 https://chem.libretexts.org/@go/page/365735
For both the method of continuous variations and the mole-ratio method, we determine the complex’s stoichiometry by
extrapolating absorbance data from conditions in which there is a linear relationship between absorbance and the relative amounts
of metal and ligand. If a metal–ligand complex is very weak, a plot of absorbance versus XL or nL/nM becomes so curved that it is
impossible to determine the stoichiometry by extrapolation. In this case the slope-ratio is used.
In the slope-ratio method two sets of solutions are prepared. The first set of solutions contains a constant amount of metal and a
variable amount of ligand, chosen such that the total concentration of metal, CM, is much larger than the total concentration of
ligand, CL. Under these conditions we may assume that essentially all the ligand reacts to form the metal–ligand complex. The
concentration of the complex, which has the general form MxLy, is
CL
[ Mx Ly ] =
y

If we monitor the absorbance at a wavelength where only MxLy absorbs, then

εbCL
A = εb [ Mx Ly ] =
y

and a plot of absorbance versus CL is linear with a slope, sL, of

\[s_{\mathrm{L}}=\frac{\varepsilon b}{y}\ ]
A second set of solutions is prepared with a fixed concentration of ligand that is much greater than a variable concentration of
metal; thus
CM
[ Mx Ly ] =
x

εbCM
A = εb [ Mx Ly ] =
x

εb
sM =
x

A ratio of the slopes provides the relative values of x and y.

sM εb/x y
= =
sL εb/y x

An important assumption in the slope-ratio method is that the complexation reaction continues to completion in the presence of a
sufficiently large excess of metal or ligand. The slope-ratio method also is limited to systems in which only a single complex forms
and for which Beer’s law is obeyed.

Determination of Equilibrium Constants

Another important application of molecular absorption spectroscopy is the determination of equilibrium constants. Let’s consider,
as a simple example, an acid–base reaction of the general form
+ −
HIn(aq) +  H2 O(l) ⇌  H3 O (aq) + In (aq)

where HIn and In– are the conjugate weak acid and weak base forms of an acid–base indicator. The equilibrium constant for this
reaction is
+ −
[ H3 O ] [A ]
Ka =
[HA]

To determine the equilibrium constant’s value, we prepare a solution in which the reaction is in a state of equilibrium and determine
the equilibrium concentration for H3O+, HIn, and In–. The concentration of H3O+ is easy to determine by measuring the solution’s
pH. To determine the concentration of HIn and In– we can measure the solution’s absorbance.
If both HIn and In– absorb at the selected wavelength, then, from Beer's law, we know that
−
A = εHln b[HIn] + εln b[ In ] (14.3.1)

14.3.4 https://chem.libretexts.org/@go/page/365735
where εHIn and ε In are the molar absorptivities for HIn and In–. The indicator’s total concentration, C, is given by a mass balance
equation
−
C = [HIn] + [ In ] (14.3.2)

Solving Equation 14.3.2 for [HIn] and substituting into Equation 14.3.1 gives
− −
A = εHln b (C − [ In ]) + εln b [ In ]

which we simplify to
− −
A = εHln bC − εHln b [ In ] + εln b [ In ]

−
A = AHIn + b [ In ] (εln − εHIn ) (14.3.3)

where AHIn, which is equal to ε bC, is the absorbance when the pH is acidic enough that essentially all the indicator is present as
HIn

HIn. Solving Equation 14.3.3 for the concentration of In– gives

−
A − AHln
[ In ] = (14.3.4)
b (εln − εHIn )

Proceeding in the same fashion, we derive a similar equation for the concentration of HIn
AIn − A
[HIn] = (14.3.5)
b (εln − εHln )

where AIn, which is equal to ε bC , is the absorbance when the pH is basic enough that only In– contributes to the absorbance.
In

Substituting Equation 14.3.4 and Equation 14.3.5 into the equilibrium constant expression for HIn gives
+ −
[ H3 O ][ In ] A − AHIn
+
Ka = = [ H3 O ]× (14.3.6)
[HIn] AIn − A

We can use Equation 14.3.6 to determine Ka in one of two ways. The simplest approach is to prepare three solutions, each of which
contains the same amount, C, of indicator. The pH of one solution is made sufficiently acidic such that [HIn] >> [In–]. The
absorbance of this solution gives AHIn. The value of AIn is determined by adjusting the pH of the second solution such that [In–] >>
[HIn]. Finally, the pH of the third solution is adjusted to an intermediate value, and the pH and absorbance, A, recorded. The value
of Ka is calculated using Equation 14.3.6.

 Example 14.3.2

The acidity constant for an acid–base indicator is determined by preparing three solutions, each of which has a total
concentration of indicator equal to 5.00 × 10 M. The first solution is made strongly acidic with HCl and has an absorbance
−5

of 0.250. The second solution is made strongly basic and has an absorbance of 1.40. The pH of the third solution is 2.91 and
has an absorbance of 0.662. What is the value of Ka for the indicator?

Solution
The value of Ka is determined by making appropriate substitutions into 10.20 where [H3O+] is 1.23 × 10 −3
; thus

−3
0.662 − 0.250 −4
Ka = (1.23 × 10 )× = 6.87 × 10
1.40 − 0.662

 Exercise 14.3.2

To determine the Ka of a merocyanine dye, the absorbance of a solution of 3.5 × 10 M dye was measured at a pH of 2.00, a
−4

pH of 6.00, and a pH of 12.00, yielding absorbances of 0.000, 0.225, and 0.680, respectively. What is the value of Ka for this
dye? The data for this problem is adapted from Lu, H.; Rutan, S. C. Anal. Chem., 1996, 68, 1381–1386.

Answer

14.3.5 https://chem.libretexts.org/@go/page/365735
 The value of Ka is

−6
0.225 − 0.000 −7
Ka = (1.00 × 10 )× = 4.95 × 10
0.680 − 0.225

A second approach for determining Ka is to prepare a series of solutions, each of which contains the same amount of indicator. Two
solutions are used to determine values for AHIn and AIn. Taking the log of both sides of Equation 14.3.6 and rearranging leave us
with the following equation.
A − AHin
log = pH − pKa (14.3.7)
A − −A
ln

A plot of log[(A – AHIn)/(AIn – A)] versus pH is a straight-line with a slope of +1 and a y-intercept of –pKa.

 Exercise 14.3.3

To determine the Ka for the indicator bromothymol blue, the absorbance of each a series of solutions that contain the same
concentration of bromothymol blue is measured at pH levels of 3.35, 3.65, 3.94, 4.30, and 4.64, yielding absorbance values of
0.170, 0.287, 0.411, 0.562, and 0.670, respectively. Acidifying the first solution to a pH of 2 changes its absorbance to 0.006,
and adjusting the pH of the last solution to 12 changes its absorbance to 0.818. What is the value of Ka for bromothymol blue?
The data for this problem is from Patterson, G. S. J. Chem. Educ., 1999, 76, 395–398.

Answer
To determine Ka we use Equation 14.3.7, plotting log[(A – AHIn)/(AIn – A)] versus pH, as shown below.

Fitting a straight-line to the data gives a regression model of

A − AHIn
log = −3.80 + 0.962pH
Aln − A

The y-intercept is –pKa; thus, the pKa is 3.80 and the Ka is 1.58 × 10 . −4

In developing these approaches for determining Ka we considered a relatively simple system in which the absorbance of HIn and
In– are easy to measure and for which it is easy to determine the concentration of H3O+. In addition to acid–base reactions, we can
adapt these approaches to any reaction of the general form

X(aq) + Y (aq) ⇌ Z(aq)

including metal–ligand complexation reactions and redox reactions, provided we can determine spectrophotometrically the
concentration of the product, Z, and one of the reactants, either X or Y, and that we can determine the concentration of the other
reactant by some other method. With appropriate modifications, a more complicated system in which we cannot determine the
concentration of one or more of the reactants or products also is possible [Ramette, R. W. Chemical Equilibrium and Analysis,
Addison-Wesley: Reading, MA, 1981, Chapter 13].

This page titled 14.3: Qualitative and Characterization Applications is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

14.3.6 https://chem.libretexts.org/@go/page/365735
14.4: Quantitative Applications
Scope
The determination of an analyte’s concentration based on its absorption of ultraviolet or visible radiation is one of the most
frequently encountered quantitative analytical methods. One reason for its popularity is that many organic and inorganic
compounds have strong absorption bands in the UV/Vis region of the electromagnetic spectrum. In addition, if an analyte does not
absorb UV/Vis radiation—or if its absorbance is too weak—we often can react it with another species that is strongly absorbing.
For example, a dilute solution of Fe2+ does not absorb visible light. Reacting Fe2+ with o-phenanthroline, however, forms an
orange–red complex of Fe(phen) that has a strong, broad absorbance band near 500 nm. An additional advantage to UV/Vis
2+

absorption is that in most cases it is relatively easy to adjust experimental and instrumental conditions so that Beer’s law is obeyed.

Environmental Applications
The analysis of waters and wastewaters often relies on the absorption of ultraviolet and visible radiation. Many of these methods
are outlined in Table 14.4.1. Several of these methods are described here in more detail.
Table 14.4.1 . Examples of Molecular UV/Vis Analysis of Waters and Wastewaters
analyte method λ (nm)

trace metals

react with Eriochrome cyanide R dye at pH 6;

aluminum 535
forms red to pink complex

reduce to AsH3 using Zn and react with silver

arsenic 535
diethyldithiocarbamate; forms red complex

extract into CHCl3 containing dithizone from a

cadmium sample made basic with NaOH; forms pink to 518
red complex

oxidize to Cr(VI) and react with

chromium diphenylcarbazide; forms 540 red-violet 540
product

react with neocuprine in neutral to slightly

copper acid solution and extract into CHCl3/CH3OH; 457
forms yellow complex

reduce to Fe2+ and react with o-

iron 510
phenanthroline; forms orange-red complex

extract into CHCl3 containing dithizone from

lead sample made basic with NH3/ NH4+ buffer; 510
forms cherry red complex

oxidize to MnO4– with persulfate; forms

manganese 525
purple solution

extract into CHCl3 containing dithizone from

mercury 492
acidic sample; forms orange complex

zinc react with zincon at pH 9; forms blue complex 620

inorganic nonmetals

reaction with hypochlorite and phenol using a

ammonia manganous 630 salt catalyst; forms blue 630
indophenol as product

react with chloroamine-T to form CNCl and

cyanide then with a pyridine-barbituric acid; forms a 578
red-blue dye

14.4.1 https://chem.libretexts.org/@go/page/365739
 analyte method λ (nm)

react with red Zr-SPADNS lake; formation of

fluoride 570
ZrF62– decreases color of the red lake

react with leuco crystal violet; forms blue

chlorine (residual) 592
product

react with Cd to form NO2– and then react

nitrate with sulfanilamide and N-(1-napthyl)- 543
ethylenediamine; forms red azo 543 dye

react with ammonium molybdate and then

phosphate 690
reduce with SnCl2; forms molybdenum blue

organics

react with 4-aminoantipyrine and K3Fe(CN)6;

phenol 460
forms yellow antipyrine dye

react with cationic methylene blue dye and

anionic surfactants 652
extract into CHCl3; forms blue ion pair

Although the quantitative analysis of metals in waters and wastewaters is accomplished primarily by atomic absorption or atomic
emission spectroscopy, many metals also can be analyzed following the formation of a colorful metal–ligand complex. One
advantage to these spectroscopic methods is that they easily are adapted to the analysis of samples in the field using a filter
photometer. One ligand used for the analysis of several metals is diphenylthiocarbazone, also known as dithizone. Dithizone is not
soluble in water, but when a solution of dithizone in CHCl3 is shaken with an aqueous solution that contains an appropriate metal
ion, a colored metal–dithizonate complex forms that is soluble in CHCl3. The selectivity of dithizone is controlled by adjusting the
sample’s pH. For example, Cd2+ is extracted from solutions made strongly basic with NaOH, Pb2+ from solutions made basic with
an NH3/ NH4+ buffer, and Hg2+ from solutions that are slightly acidic.

The structure of dithizone is shown below.

When chlorine is added to water the portion available for disinfection is called the chlorine residual. There are two forms of
chlorine residual. The free chlorine residual includes Cl2, HOCl, and OCl–. The combined chlorine residual, which forms from the
reaction of NH3 with HOCl, consists of monochloramine, NH2Cl, dichloramine, NHCl2, and trichloramine, NCl3. Because the free
chlorine residual is more efficient as a disinfectant, there is an interest in methods that can distinguish between the total chlorine
residual’s different forms. One such method is the leuco crystal violet method. The free residual chlorine is determined by adding
leuco crystal violet to the sample, which instantaneously oxidizes to give a blue-colored compound that is monitored at 592 nm.
Completing the analysis in less than five minutes prevents a possible interference from the combined chlorine residual. The total
chlorine residual (free + combined) is determined by reacting a separate sample with iodide, which reacts with both chlorine
residuals to form HOI. When the reaction is complete, leuco crystal violet is added and oxidized by HOI, giving the same blue-
colored product. The combined chlorine residual is determined by difference.
The concentration of fluoride in drinking water is determined indirectly by its ability to form a complex with zirconium. In the
presence of the dye SPADNS, a solution of zirconium forms a red colored compound, called a lake, that absorbs at 570 nm. When
fluoride is added, the formation of the stable ZrF complex causes a portion of the lake to dissociate, decreasing the absorbance.
2−

A plot of absorbance versus the concentration of fluoride, therefore, has a negative slope.

14.4.2 https://chem.libretexts.org/@go/page/365739
 SPADNS, the structure of which is shown below, is an abbreviation for the sodium salt of 2-(4-sulfophenylazo)-1,8-dihydroxy-
3,6-napthalenedisulfonic acid, which is a mouthful to say.

Spectroscopic methods also are used to determine organic constituents in water. For example, the combined concentrations of
phenol and ortho- and meta-substituted phenols are determined by using steam distillation to separate the phenols from nonvolatile
impurities. The distillate reacts with 4-aminoantipyrine at pH 7.9 ± 0.1 in the presence of K3Fe(CN)6 to a yellow colored antipyrine
dye. After extracting the dye into CHCl3, its absorbance is monitored at 460 nm. A calibration curve is prepared using only the
unsubstituted phenol, C6H5OH. Because the molar absorptivity of substituted phenols generally are less than that for phenol, the
reported concentration represents the minimum concentration of phenolic compounds.

4-aminoantipyrene

Molecular absorption also is used for the analysis of environmentally significant airborne pollutants. In many cases the analysis is
carried out by collecting the sample in water, converting the analyte to an aqueous form that can be analyzed by methods such as
those described in Table 14.4.1. For example, the concentration of NO2 is determined by oxidizing NO2 to NO . The −

concentration of NO is then determined by first reducing it to NO with Cd, and then reacting NO with sulfanilamide and N-
−

3
−

2
−

(1-naphthyl)-ethylenediamine to form a red azo dye. Another important application is the analysis for SO2, which is determined by
collecting the sample in an aqueous solution of HgCl where it reacts to form Hg(SO ) . Addition of p-rosaniline and
2−
4 3
2−

formaldehyde produces a purple complex that is monitored at 569 nm. Infrared absorption is useful for the analysis of organic
vapors, including HCN, SO2, nitrobenzene, methyl mercaptan, and vinyl chloride. Frequently, these analyses are accomplished
using portable, dedicated infrared photometers.

Clinical Applications
The analysis of clinical samples often is complicated by the complexity of the sample’s matrix, which may contribute a significant
background absorption at the desired wavelength. The determination of serum barbiturates provides one example of how this
problem is overcome. The barbiturates are first extracted from a sample of serum with CHCl3 and then extracted from the CHCl3
into 0.45 M NaOH (pH ≈ 13). The absorbance of the aqueous extract is measured at 260 nm, and includes contributions from the
barbiturates as well as other components extracted from the serum sample. The pH of the sample is then lowered to approximately
10 by adding NH4Cl and the absorbance remeasured. Because the barbiturates do not absorb at this pH, we can use the absorbance
at pH 10, ApH 10, to correct the absor-ance at pH 13, ApH 13
Vsamp + VNH Cl
4
Abarb = ApH 13 − × ApH 10
Vsamp

where Abarb is the absorbance due to the serum barbiturates and Vsamp and V are the volumes of sample and NH4Cl,
NH4 Cl

respectively. Table 14.4.2 provides a summary of several other methods for analyzing clinical samples.
Table 14.4.2 . Examples of the Molecular UV/Vis Analysis of Clinical Samples
analyte method λ (nm)

react with NaOH and Cu2+; forms blue-violet

total serum protein 540
complex

14.4.3 https://chem.libretexts.org/@go/page/365739
 analyte method λ (nm)

react withFe3+ in presence of isopropanol,

serum cholesterol acetic acid, and H2SO4; forms blue-violet 540
complex

react with phosphotungstic acid; forms

uric acid 710
tungsten blue

extract into CHCl3 to isolate from interferents

serum barbituates 260
and then extract into 0.45 M NaOH

react with o-toludine at 100oC; forms blue-

glucose 630
green complex

decompose protein to release iodide, which

protein-bound iodine catalyzes redox reaction between Ce3+ and 420
As3+; forms yellow colored Ce4+

Industrial Applications
UV/Vis molecular absorption is used for the analysis of a diverse array of industrial samples including pharmaceuticals, food,
paint, glass, and metals. In many cases the methods are similar to those described in Table 14.4.1 and in Table 14.4.2. For example,
the amount of iron in food is determined by bringing the iron into solution and analyzing using the o-phenanthroline method listed
in Table 14.4.1.
Many pharmaceutical compounds contain chromophores that make them suitable for analysis by UV/Vis absorption. Products
analyzed in this fashion include antibiotics, hormones, vitamins, and analgesics. One example of the use of UV absorption is in
determining the purity of aspirin tablets, for which the active ingredient is acetylsalicylic acid. Salicylic acid, which is produced by
the hydrolysis of acetylsalicylic acid, is an undesirable impurity in aspirin tablets, and should not be present at more than 0.01%
w/w. Samples are screened for unacceptable levels of salicylic acid by monitoring the absorbance at a wavelength of 312 nm.
Acetylsalicylic acid absorbs at 280 nm, but absorbs poorly at 312 nm. Conditions for preparing the sample are chosen such that an
absorbance of greater than 0.02 signifies an unacceptable level of salicylic acid.

Forensic Applications
UV/Vis molecular absorption routinely is used for the analysis of narcotics and for drug testing. One interesting forensic
application is the determination of blood alcohol using the Breathalyzer test. In this test a 52.5-mL breath sample is bubbled
through an acidified solution of K2Cr2O7, which oxidizes ethanol to acetic acid. The concentration of ethanol in the breath sample
is determined by a decrease in the absorbance at 440 nm where the dichromate ion absorbs. A blood alcohol content of 0.10%,
which is above the legal limit, corresponds to 0.025 mg of ethanol in the breath sample.

Developing a Quantitative Method for a Single Component

To develop a quantitative analytical method, the conditions under which Beer’s law is obeyed must be established. First, the most
appropriate wavelength for the analysis is determined from an absorption spectrum. In most cases the best wavelength corresponds
to an absorption maximum because it provides greater sensitivity and is less susceptible to instrumental limitations. Second, if the
instrument has adjustable slits, then an appropriate slit width is chosen. The absorption spectrum also aids in selecting a slit width
by choosing a width that is narrow enough to avoid instrumental limitaions to Beer’s law, but wide enough to increase the
throughput of source radiation. Finally, a calibration curve is constructed to determine the range of concentrations for which Beer’s
law is valid. Additional considerations that are important in any quantitative method are the effect of potential interferents and
establishing an appropriate blank.

Quantitative Analysis for a Single Sample

To determine the concentration of an analyte we measure its absorbance and apply Beer’s law using any of the standardization
methods described in Chapter 5. The most common methods are a normal calibration curve using external standards and the
method of standard additions. A single point standardization also is possible, although we must first verify that Beer’s law holds for
the concentration of analyte in the samples and the standard.

 Example 14.4.1

14.4.4 https://chem.libretexts.org/@go/page/365739
 The determination of iron in an industrial waste stream is carried out by the o-phenanthroline described in Representative
Method 10.3.1. Using the data in the following table, determine the mg Fe/L in the waste stream.

mg Fe/L absorbance

0.00 0.000

1.00 0.183

2.00 0.364

3.00 0.546

4.00 0.727

sample 0.269

Solution
Linear regression of absorbance versus the concentration of Fe in the standards gives the calibration curve and calibration
equation shown here

−1
A = 0.0006 + (0.1817 mg L) × (mgFe/L)

Substituting the sample’s absorbance into the calibration equation gives the concentration of Fe in the waste stream as 1.48 mg
Fe/L

 Exercise 14.4.1

The concentration of Cu2+ in a sample is determined by reacting it with the ligand cuprizone and measuring its absorbance at
606 nm in a 1.00-cm cell. When a 5.00-mL sample is treated with cuprizone and diluted to 10.00 mL, the resulting solution has
an absorbance of 0.118. A second 5.00-mL sample is mixed with 1.00 mL of a 20.00 mg/L standard of Cu2+, treated with
cuprizone and diluted to 10.00 mL, giving an absorbance of 0.162. Report the mg Cu2+/L in the sample.

Answer
For this standard addition we write equations that relate absorbance to the concentration of Cu2+ in the sample before the
standard addition
5.00 mL
0.118 = εb [CCu × ]
10.00 mL

and after the standard addition

5.00 mL 20.00 mg Cu 1.00 mL
0.162 = εb ( CCu × + × )
10.00 mL L 10.00 mL

in each case accounting for the dilution of the original sample and for the standard. The value of εb is the same in both
equation. Solving each equation for εb and equating

14.4.5 https://chem.libretexts.org/@go/page/365739
 0.162 0.118
=
5.00 mL 20.00 mg Cu 1.00 mL 5.00 mL
CCu × + × CCu ×
10.00 mL L 10.00 mL 10.00 mL

leaves us with an equation in which CCu is the only variable. Solving for CCu gives its value as
0.162 0.118
=
0.500 × CCu + 2.00 mg Cu/L 0.500 × CCu

0.0810 × CCu = 0.0590 × CCa + 0.236 mg Cu/L

0.0220 × CCu = 0.236 mg Cu/L

CCu = 10.7 mg Cu/L

Quantitative Analysis of Mixtures

Suppose we need to determine the concentration of two analytes, X and Y, in a sample. If each analyte has a wavelength where the
other analyte does not absorb, then we can proceed using the approach in Example 14.4.1 . Unfortunately, UV/Vis absorption bands
are so broad that frequently it is not possible to find suitable wavelengths. Because Beer’s law is additive the mixture’s absorbance,
Amix, is
(Amix ) = (εx ) b CX + (εY ) b CY (14.4.1)
λ1 λ1 λ1

where λ is the wavelength at which we measure the absorbance. Because Equation 14.4.1 includes terms for the concentration of
1

both X and Y, the absorbance at one wavelength does not provide enough information to determine either CX or CY. If we measure
the absorbance at a second wavelength
(Amix ) = (εx ) b CX + (εY ) b CY (14.4.2)
λ2 λ2 λ2

then we can determine CX and CY by solving simultaneously Equation 14.4.1 and Equation 14.4.2. Of course, we also must
determine the value for ε and ε at each wavelength. For a mixture of n components, we must measure the absorbance at n
X Y

different wavelengths.

 Example 14.4.2
The concentrations of Fe3+ and Cu2+ in a mixture are determined following their reaction with hexacyanoruthenate (II),
, which forms a purple-blue complex with Fe3+ (λ = 550 nm) and a pale-green complex with Cu2+ (λ
4−
Ru(CN)
6 max = 396 max

nm) [DiTusa, M. R.; Schlit, A. A. J. Chem. Educ. 1985, 62, 541–542]. The molar absorptivities (M–1 cm–1) for the metal
complexes at the two wavelengths are summarized in the following table.

analyte ε550 ε396

Fe3+ 9970 84

Cu2+ 34 856

When a sample that contains Fe3+ and Cu2+ is analyzed in a cell with a pathlength of 1.00 cm, the absorbance at 550 nm is
0.183 and the absorbance at 396 nm is 0.109. What are the molar concentrations of Fe3+ and Cu2+ in the sample?

Solution
Substituting known values into Equation 14.4.1 and Equation 14.4.2 gives
A550 = 0.183 = 9970 CFe + 34 CCu

A396 = 0.109 = 84 CFe + 856 CCu

To determine CFe and CCu we solve the first equation for CCu
0.183 − 9970CFe
CCu =
34

14.4.6 https://chem.libretexts.org/@go/page/365739
 and substitute the result into the second equation.
0.183 − 9970CFe
0.109 = 84 CFe + 856 ×
34
5
= 4.607 − (2.51 × 10 ) CFe

Solving for CFe gives the concentration of Fe3+ as 1.8 × 10 M. Substituting this concentration back into the equation for the
−5

mixture’s absorbance at 396 nm gives the concentration of Cu2+ as 1.3 × 10 M. −4

Another approach to solving Example 14.4.2 is to multiply the first equation by 856/34 giving

4.607 = 251009 CFe + 856 CCu

Subtracting the second equation from this equation

4.607 = 251009 CFe + 856 CCu

−0.109 = 84 CFe + 856 CCu

gives

4.498 = 250925CFe

and we find that CFe is 1.8 × 10 −5

. Having determined CFe we can substitute back into one of the other equations to solve for
CCu, which is 1.3 × 10 .
−5

 Exercise 14.4.2

The absorbance spectra for Cr3+ and Co2+ overlap significantly. To determine the concentration of these analytes in a mixture,
its absorbance is measured at 400 nm and at 505 nm, yielding values of 0.336 and 0.187, respectively. The individual molar
absorptivities (M–1 cm–1) for Cr3+ are 15.2 at 400 nm and 0.533 at 505 nm; the values for Co2+ are 5.60 at 400 nm and 5.07 at
505 nm.

Answer
Substituting into Equation 14.4.1and Equation 14.4.2gives

A400 = 0.336 = 15.2 CCr + 5.60 CCo

A400 = 0187 = 0.533 CCr + 5.07 CCo

To determine CCr and CCo we solve the first equation for CCo
0.336 − 15.2CCo
CCo =
5.60

and substitute the result into the second equation.

0.336 − 15.2CCo
0.187 = 0.533 CCr + 5.07 ×
5.60

0.187 = 0.3042 − 13.23CCr

Solving for CCr gives the concentration of Cr3+ as 8.86 × 10 M. Substituting this concentration back into the equation
−3

for the mixture’s absorbance at 400 nm gives the concentration of Co2+ as 3.60 × 10 M. −2

To obtain results with good accuracy and precision the two wavelengths should be selected so that ε > ε at one wavelength and
X Y

εX <ε Yat the other wavelength. It is easy to appreciate why this is true. Because the absorbance at each wavelength is dominated
by one analyte, any uncertainty in the concentration of the other analyte has less of an impact. Figure 14.4.q shows that the choice
of wavelengths for Practice Exercise 14.4.2 are reasonable. When the choice of wavelengths is not obvious, one method for

14.4.7 https://chem.libretexts.org/@go/page/365739
locating the optimum wavelengths is to plot ε /ε as function of wavelength, and determine the wavelengths where ε
X y X / εy reaches
maximum and minimum values [Mehra, M. C.; Rioux, J. J. Chem. Educ. 1982, 59, 688–689].

Figure 14.4.1 . Visible absorption spectra for 0.0250 M Cr3+, 0.0750 M Co2+, and for a mixture of Cr3+ and Co2+. The two
wavelengths used to analyze the mixture of Cr3+ and Co2+ are shown by the dashed lines. The data for the two standards are from
Brewer, S. Solving Problems in Analytical Chemistry, John Wiley & Sons: New York, 1980.
When the analyte’s spectra overlap severely, such that ε ≈ ε at all wavelengths, other computational methods may provide
X Y

better accuracy and precision. In a multiwavelength linear regression analysis, for example, a mixture’s absorbance is compared to
that for a set of standard solutions at several wavelengths [Blanco, M.; Iturriaga, H.; Maspoch, S.; Tarin, P. J. Chem. Educ. 1989,
66, 178–180]. If ASX and ASY are the absorbance values for standard solutions of components X and Y at any wavelength, then
ASX = εX b CSX (14.4.3)

ASY = εY b CSY (14.4.4)

where CSX and CSY are the known concentrations of X and Y in the standard solutions. Solving Equation 14.4.3 and Equation 14.4.4
for ε and for ε , substituting into Equation 14.4.1, and rearranging, gives
X Y

Amix CX CY ASY
= + ×
ASX CSX CSY ASX

To determine CX and CY the mixture’s absorbance and the absorbances of the standard solutions are measured at several
wavelengths. Graphing Amix/ASX versus ASY/ASX gives a straight line with a slope of CY/CSY and a y-intercept of CX/CSX. This
approach is particularly helpful when it is not possible to find wavelengths where ε > ε and ε < ε .
X Y X Y

The approach outlined here for a multiwavelength linear regression uses a single standard solution for each analyte. A more
rigorous approach uses multiple standards for each analyte. The math behind the analysis of such data—which we call a
multiple linear regression—is beyond the level of this text. For more details about multiple linear regression see Brereton, R.
G. Chemometrics: Data Analysis for the Laboratory and Chemical Plant, Wiley: Chichester, England, 2003.

 Example 14.4.3

Figure 14.4.1 shows visible absorbance spectra for a standard solution of 0.0250 M Cr3+, a standard solution of 0.0750 M
Co2+, and a mixture that contains unknown concentrations of each ion. The data for these spectra are shown here.

λ (nm) ACr ACu Amix λ (nm) ACr ACu Amix

375 0.26 0.01 0.53 520 0.19 0.38 0.63

400 0.43 0.03 0.88 530 0.24 0.33 0.70

425 0.39 0.07 0.83 540 0.28 0.26 0.73

440 0.29 0.13 0.67 550 0.32 0.18 0.76

455 0.20 0.21 0.54 570 0.38 0.08 0.81

470 0.14 0.28 0.47 575 0.39 0.06 0.82

14.4.8 https://chem.libretexts.org/@go/page/365739
 480 0.12 0.30 0.44 580 0.38 0.05 0.79

490 0.11 0.34 0.45 600 0.34 0.03 0.70

500 0.13 0.38 0.51 625 0.24 0.02 0.49

Use a multiwavelength regression analysis to determine the composition of the unknown.

Solution
First we need to calculate values for Amix/ASX and for ASY/ASX. Let’s define X as Co2+ and Y as Cr3+. For example, at a
wavelength of 375 nm Amix/ASX is 0.53/0.01, or 53 and ASY/ASX is 0.26/0.01, or 26. Completing the calculation for all
wavelengths and graphing Amix/ASX versus ASY/ASX gives the calibration curve shown in Figure 14.4.2. Fitting a straight-line to
the data gives a regression model of
Amix ASY
= 0.636 + 2.01 ×
ASX ASX

Using the y-intercept, the concentration of Co2+ is

2+
CX [ Co ]
= = 0.636
CSX 0.0750M

or [Co2+] = 0.048 M; using the slope the concentration of Cr3+ is

3+
CY [ Cr ]
= = 2.01
CSY 0.0250M

or [Cr3+] = 0.050 M.

Figure 14.4.2 . Multiwavelength linear regression analysis for the data in Example 14.4.3 .

 Exercise 14.4.3

A mixture of MnO and Cr O , and standards of 0.10 mM KMnO4 and of 0.10 mM K2Cr2O7 give the results shown in the
−

4 2
2−

following table. Determine the composition of the mixture. The data for this problem is from Blanco, M. C.; Iturriaga, H.;
Maspoch, S.; Tarin, P. J. Chem. Educ. 1989, 66, 178–180.

λ (nm) AMn ACr Amix

266 0.042 0.410 0.766

288 0.082 0.283 0.571

320 0.168 0.158 0.422

350 0.125 0.318 0.672

360 0.036 0.181 0.366

14.4.9 https://chem.libretexts.org/@go/page/365739
 Answer
Letting X represent MnO and letting Y represent Cr
−
4
2−
2 O7 , we plot the equation
Amix CX CY ASY
= + ×
ASX CSX CSY ASX

placing Amix/ASX on the y-axis and ASY/ASX on the x-axis. For example, at a wavelength of 266 nm the value Amix/ASX of is
0.766/0.042, or 18.2, and the value of ASY/ASX is 0.410/0.042, or 9.76. Completing the calculations for all wavelengths and
plotting the data gives the result shown here

Fitting a straight-line to the data gives a regression model of

A mix  ASY
= 0.8147 + 1.7839 ×
A SX  ASX

Using the y-intercept, the concentration of MnO is−

4

−
CX [ MnO 4 ]
= 0.8147 =
−4 −
CSX 1.0 × 10  M MnO
4

or 8.15 × 10 −5
M MnO , and using the slope, the concentration of Cr
−
4 2 O7
2−
is
2−
CY [ Cr2 O ]
7
= 1.7839 =
−4 2−
CSY 1.00 × 10  M Cr2 O7

or 1.78 × 10 −4
M Cr 2−
2 O7 .

Derivative Spectroscopy
Sometimes our signal is superimposed on a background signal, which complicates our analysis because the measure absorbance has
contributions from both our analyte and from the background. For example, the following figure shows a Gaussian signal with a
maximum value of 50 centered at x = 125 that is superimposed on an exponential background. The dotted line is the Gaussian
signal, which has a maximum value of 50 at x = 125, and the solid line is the signal as measured, which has a maximum value of
57 at x = 125.

14.4.10 https://chem.libretexts.org/@go/page/365739
 Figure 14.4.3 : A Gaussian signal (dotted line) superimposed on an exponential background, which gives rise to the measured
signal (solid line).
If the background signal is consistent across all samples, then we can analyze the data without first removing its contribution. For
example, the following figure shows a set of calibration standards and their resulting calibration curve, for which the y-intercept of
7 gives the offset introduced by the background.

Figure 14.4.4 : When the background is the same for all calibration standards and samples, then we can construct a calibration
curve without taking into account the presence of the background.
But background signals often are not consistent across samples, particularly when the source of the background is a property of the
samples we collect (natural water samples, for example, may have variations in color due to differences in the concentration of
dissolved organic matter) or a property of the instrument we are using (such as a variation in source intensity over time). When
true, our data may look more like what we see in the following figure, which leads to a calibration curve with a greater uncertainty.

Figure 14.4.5 : When the background is not the same for all calibration standards, the quality of the calibration curves suffers,
making it less useful for the analysis of samples.

14.4.11 https://chem.libretexts.org/@go/page/365739
Because the background changes gradually with the values for x while the analyte's signal changes quickly, we can use a derivative
Δy
to the distinguish between the two. One approach is to calculate and plot the derivative, , as a function of x, as shown in Figure
Δx

14.4.6. The calibration signal in this case is the difference between the maximum signal and the minimum signal, which are shown

by the dotted red lines in the top part of the figure. The fit of the calibration curve to the data and the calibration curve's y-intercept
of zero shows that we have successfully compensated for the background signals.

Figure 14.4.6 : Taking the derivative of the data in Figure 14.4.5 removes the background signal.

This page titled 14.4: Quantitative Applications is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

14.4.12 https://chem.libretexts.org/@go/page/365739
14.5: Photometric Titrations
If at least one species in a titration absorbs electromagnetic radiation, then we can identify the end point by monitoring the titrand’s
absorbance at a carefully selected wavelength. For example, we can identify the end point for a titration of Cu2+ with EDTA in the
presence of NH3 by monitoring the titrand’s absorbance at a wavelength of 745 nm, where the Cu(NH ) complex absorbs 3
4+
2

strongly. At the beginning of the titration the absorbance is at a maximum. As we add EDTA, however, the reaction
2+ 4− 2−
Cu(NH ) (aq) + Y ⇌ CuY (aq) + 4 NH3 (aq)
3 4

decreases the concentration of Cu(NH ) 3

and decreases the absorbance until we reach the equivalence point. After the
4+
2

equivalence point the absorbance essentially remains unchanged. The resulting spectrophotometric titration curve is shown in
Figure 14.5.1a. Note that the titration curve’s y-axis is not the measured absorbance, Ameas, but a corrected absorbance, Acorr
VEDTA + VCu
Acorr = Ameas ×
VCu

where VEDTA and VCu are, respectively, the volumes of EDTA and Cu. Correcting the absorbance for the titrand’s dilution ensures
that the spectrophotometric titration curve consists of linear segments that we can extrapolate to find the end point. Other common
spectrophotometric titration curves are shown in Figures 14.5.1b − f .

Figure 14.5.1 . Examples of spectrophotometric titration curves: (a) only the titrand absorbs; (b) only the titrant absorbs; (c) only
the product of the titration reaction absorbs; (d) both the titrand and the titrant absorb; (e) both the titration reaction’s product and
the titrant absorb; (f ) only the indicator absorbs. The red arrows indicate the end points for each titration curve.

This page titled 14.5: Photometric Titrations is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

14.5.1 https://chem.libretexts.org/@go/page/365741
 CHAPTER OVERVIEW
15: Molecular Luminescence
15.1: Theory of Fluorescence and Phosphorescence
15.2: Instruments for Measuring Fluorescence and Phosphorescence
15.3: Applications and Photoluminescence methods
15.4: Chemiluminscence
15.5: Evaluation of Molecular Luminescence

This page titled 15: Molecular Luminescence is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1
15.1: Theory of Fluorescence and Phosphorescence
The use of molecular fluorescence for qualitative analysis and for semi-quantitative analysis dates to the early to mid 1800s, with
more accurate quantitative methods appearing in the 1920s. Instrumentation for fluorescence spectroscopy using a filter or a
monochromator for wavelength selection appeared in, respectively, the 1930s and 1950s. Although the discovery of
phosphorescence preceded that of fluorescence by almost 200 years, qualitative and quantitative applications of molecular
phosphorescence did not receive much attention until after the development of fluorescence instrumentation.

Source of Fluorescence and Phosphorescence

Photoluminescence is divided into two categories: fluorescence and phosphorescence. A pair of electrons that occupy the same
electronic ground state have opposite spins and are in a singlet spin state (Figure 15.1.1a).

Figure 15.1.1a . Electron configurations for (a) a singlet ground state; (b) a singlet excited state; and (c) a triplet excited state.
When an analyte absorbs an ultraviolet or a visible photon, one of its valence electrons moves from the ground state to an excited
state with a conservation of the electron’s spin (Figure 15.1.1b). Emission of a photon from a singlet excited state to the singlet
ground state—or between any two energy levels with the same spin—is called fluorescence. The probability of fluorescence is very
high and the average lifetime of an electron in the excited state is only 10–5–10–8 s. Fluorescence, therefore, rapidly decays once the
source of excitation is removed.
In some cases an electron in a singlet excited state is transformed to a triplet excited state (Figure 15.1.1c) in which its spin no is
longer paired with the ground state. Emission between a triplet excited state and a singlet ground state—or between any two energy
levels that differ in their respective spin states–is called phosphorescence. Because the average lifetime for phosphorescence can be
quite long—it ranges from 10–4–104 seconds—phosphorescence may continue for some time after we remove the excitation source.
To appreciate the origin of fluorescence and phosphorescence we must consider what happens to a molecule following the
absorption of a photon. Let’s assume the molecule initially occupies the lowest vibrational energy level of its electronic ground
state, which is the singlet state labeled S0 in Figure 15.1.2. Absorption of a photon excites the molecule to one of several
vibrational energy levels in the first excited electronic state, S1, or the second electronic excited state, S2, both of which are singlet
states. Relaxation to the ground state occurs by a number of mechanisms, some of which result in the emission of a photon and
others that occur without the emission of a photon. These relaxation mechanisms are shown in Figure 15.1.2. The most likely
relaxation pathway from any excited state is the one with the shortest lifetime.

15.1.1 https://chem.libretexts.org/@go/page/366514
 Figure 15.1.2 . Energy level diagram for a molecule that shows pathways for the deactivation of an excited state: vr is vibrational
relaxation; ic is internal conversion; ec is external conversion; and isc is an intersystem crossing. The lowest vibrational energy for
each electronic state is indicated by the thicker line. The electronic ground state is shown in black and the three electronic excited
states are shown in green. The absorption, fluorescence, and phosphorescence of photons also are shown.

Deactivation Processes
A molecule in an excited state can return to its ground state in a variety of ways that we collectively call deactivation processes.

Radiationless Deactivation
When a molecule relaxes without emitting a photon we call the process radiationless deactivation. One example of radiationless
deactivation is vibrational relaxation, in which a molecule in an excited vibrational energy level loses energy by moving to a lower
vibrational energy level in the same electronic state. Vibrational relaxation is very rapid, with an average lifetime of <10–12 s.
Because vibrational relaxation is so efficient, a molecule in one of its excited state’s higher vibrational energy levels quickly returns
to the excited state’s lowest vibrational energy level.
Another form of radiationless deactivation is an internal conversion in which a molecule in the ground vibrational level of an
excited state passes directly into a higher vibrational energy level of a lower energy electronic state of the same spin state. By a
combination of internal conversions and vibrational relaxations, a molecule in an excited electronic state may return to the ground
electronic state without emitting a photon. A related form of radiationless deactivation is an external conversion in which excess
energy is transferred to the solvent or to another component of the sample’s matrix.

Let’s use Figure 15.1.2 to illustrate how a molecule can relax back to its ground state without emitting a photon. Suppose our
molecule is in the highest vibrational energy level of the second electronic excited state. After a series of vibrational
relaxations brings the molecule to the lowest vibrational energy level of S2, it undergoes an internal conversion into a higher
vibrational energy level of the first excited electronic state. Vibrational relaxations bring the molecule to the lowest vibrational
energy level of S1. Following an internal conversion into a higher vibrational energy level of the ground state, the molecule
continues to undergo vibrational relaxation until it reaches the lowest vibrational energy level of S0.

A final form of radiationless deactivation is an intersystem crossing in which a molecule in the ground vibrational energy level of
an excited electronic state passes into one of the higher vibrational energy levels of a lower energy electronic state with a different
spin state. For example, an intersystem crossing is shown in Figure 15.1.2 between the singlet excited state S1 and the triplet
excited state T1.

Variables that Affect Fluorescence

Fluorescence occurs when a molecule in an excited state’s lowest vibrational energy level returns to a lower energy electronic state
by emitting a photon. Because molecules return to their ground state by the fastest mechanism, fluorescence is observed only if it is
a more efficient means of relaxation than a combination of internal conversions and vibrational relaxations.

15.1.2 https://chem.libretexts.org/@go/page/366514
A quantitative expression of fluorescence efficiency is the fluorescent quantum yield, Φ , which is the fraction of excited state
f

molecules that return to the ground state by fluorescence. The fluorescent quantum yields range from 1 when every molecule in an
excited state undergoes fluorescence, to 0 when fluorescence does not occur.
The intensity of fluorescence, If, is proportional to the amount of radiation absorbed by the sample, P0 – PT, and the fluorescence
quantum yield

If = kΦf (P0 − PT ) (15.1.1)

where k is a constant that accounts for the efficiency of collecting and detecting the fluorescent emission. From Beer’s law we
know that
PT
−εbC
= 10 (15.1.2)
P0

where C is the concentration of the fluorescing species. Solving Equation 15.1.2 for PT and substituting into Equation 15.1.1 gives,
after simplifying
−εbC
If = kΦf P0 (1 − 10 ) (15.1.3)

When εbC < 0.01, which often is the case when the analyte's concentration is small, Equation 15.1.3 simplifies to
′
If = 2.303kΦf εbC P0 = k P0 (15.1.4)

where k′ is a collection of constants. The intensity of fluorescence, therefore, increases with an increase in the quantum efficiency,
the source’s incident power, and the molar absorptivity and the concentration of the fluorescing species.
Fluorescence generally is observed when the molecule’s lowest energy absorption is a π → π transition, although some n → π
∗ ∗

transitions show weak fluorescence. Many unsubstituted, nonheterocyclic aromatic compounds have a favorable fluorescence
quantum yield, although substitutions on the aromatic ring can effect Φ significantly. For example, the presence of an electron-
f

withdrawing group, such as –NO2, decreases Φ , while adding an electron-donating group, such as –OH, increases Φ .
f f

Fluorrescence also increases for aromatic ring systems and for aromatic molecules with rigid planar structures. Figure 15.1.3
shows the fluorescence of quinine under a UV lamp.

Figure 15.1.3 . Tonic water, which contains quinine, is fluorescent when placed under a UV lamp. Source: Splarka
(commons.Wikipedia.org).
A molecule’s fluorescent quantum yield also is influenced by external variables, such as temperature and solvent. Increasing the
temperature generally decreases Φ because more frequent collisions between the molecule and the solvent increases external
f

conversion. A decrease in the solvent’s viscosity decreases Φ for similar reasons. For an analyte with acidic or basic functional
f

groups, a change in pH may change the analyte’s structure and its fluorescent properties.
As shown in Figure 15.1.3, fluorescence may return the molecule to any of several vibrational energy levels in the ground
electronic state. Fluorescence, therefore, occurs over a range of wavelengths. Because the change in energy for fluorescent

15.1.3 https://chem.libretexts.org/@go/page/366514
emission generally is less than that for absorption, a molecule’s fluorescence spectrum is shifted to higher wavelengths than its
absorption spectrum.

Variables that Affect Phosphorescence

A molecule in a triplet electronic excited state’s lowest vibrational energy level normally relaxes to the ground state by an
intersystem crossing to a singlet state or by an external conversion. Phosphorescence occurs when the molecule relaxes by emitting
a photon. As shown in Figure 15.1.2, phosphorescence occurs over a range of wavelengths, all of which are at lower energies than
the molecule’s absorption band. The intensity of phosphorescence, I , is given by an equation similar to Equation 15.1.4 for
p

fluorescence

IP = 2.303kΦP εbC P0

′
= k P0 (15.1.5)

where Φ is the phosphorescence quantum yield.

p

Phosphorescence is most favorable for molecules with n → π transitions, which have a higher probability for an intersystem
∗

crossing than π → π transitions. For example, phosphorescence is observed with aromatic molecules that contain carbonyl groups
∗

or heteroatoms. Aromatic compounds that contain halide atoms also have a higher efficiency for phosphorescence. In general, an
increase in phosphorescence corresponds to a decrease in fluorescence.
Because the average lifetime for phosphorescence can be quite long, ranging from 10–4–104 s, the phosphorescent quantum yield
usually is quite small. An improvement in Φ is realized by decreasing the efficiency of external conversion. This is accomplished
p

in several ways, including lowering the temperature, using a more viscous solvent, depositing the sample on a solid substrate, or
trapping the molecule in solution. Figure 15.1.4 shows an example of phosphorescence.

Figure 15.1.4 . An europium doped strontium silicate-aluminum oxide powder under (a) natural light, (b) a long-wave UV lamp,
and (c) in total darkness. The photo taken in total darkness shows the phosphorescent emission. Source: modified from Splarka
(commons.Wikipedia.org).

Emission and Excitation Spectra

Photoluminescence spectra are recorded by measuring the intensity of emitted radiation as a function of either the excitation
wavelength or the emission wavelength. An excitation spectrum is obtained by monitoring emission at a fixed wavelength while
varying the excitation wavelength. When corrected for variations in the source’s intensity and the detector’s response, a sample’s
excitation spectrum is nearly identical to its absorbance spectrum. The excitation spectrum provides a convenient means for
selecting the best excitation wavelength for a quantitative or qualitative analysis.
In an emission spectrum a fixed wavelength is used to excite the sample and the intensity of emitted radiation is monitored as
function of wavelength. Although a molecule has a single excitation spectrum, it has two emission spectra, one for fluorescence
and one for phosphorescence. Figure 15.1.5 shows the UV absorption spectrum and the UV fluorescence emission spectrum for
quinine.

15.1.4 https://chem.libretexts.org/@go/page/366514
 Figure 15.1.5 . Absorbance spectrum and fluorescence emission spectrum for quinine in 0.05 M H2SO4. The emission spectrum
uses an excitation wavelength of 350 nm with a bandwidth of 20 nm. Both spectra are normalized so that the maximum absorbance
is 1.00 and the maximum emission is 1.00. The actual maximum absorbance is 0.444 and the actual maximum emission is 126747.
Source: data from Daniel Scott, Department of Chemistry & Biochemistry, DePauw University.

This page titled 15.1: Theory of Fluorescence and Phosphorescence is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

15.1.5 https://chem.libretexts.org/@go/page/366514
15.2: Instruments for Measuring Fluorescence and Phosphorescence
Instrumentation
The basic instrumentation for monitoring fluorescence and phosphorescence—a source of radiation, a means of selecting a narrow
band of radiation, and a detector—are the same as those for absorption spectroscopy. The unique demands of fluorescence and
phosphorescence, however, require some modifications to the instrument designs discussed in earlier chapters: the filter
photometer, the single-beam spectrophotometer, the double-beam spectrophotometer, and the diode array spectrometer. The most
important difference is that the detector cannot be placed directly across from the source. Figure 15.2.1 shows why this is the case.
If we place the detector along the source’s axis it receives both the transmitted source radiation, PT, and the fluorescent, If, or
phosphorescent, Ip, radiation. Instead, we rotate the director and place it at 90o to the source.

Figure 15.2.1 . Schematic diagram showing the orientation of the source and the detector when measuring fluorescence and
phosphorescence. Contrast this to Figure 13.4.2, which shows the orientation for absorption spectroscopy.

Instruments for Measuring Fluorescence

Figure 15.2.2 shows the basic design of an instrument for measuring fluorescence, which includes two wavelength selectors, one
for selecting the source's excitation wavelength and one for selecting the analyte's emission wavelength. In a fluorometer the
excitation and emission wavelengths are selected using absorption or interference filters. The excitation source for a fluorometer
usually is a low-pressure Hg vapor lamp that provides intense emission lines distributed throughout the ultraviolet and visible
region. When a monochromator is used to select the excitation and the emission wavelengths, the instrument is called a
spectrofluorometer. With a monochromator the excitation source usually is a high-pressure Xe arc lamp, which has a continuous
emission spectrum. Either instrumental design is appropriate for quantitative work, although only a spectrofluorometer can record
an excitation or emission spectrum.

A Hg vapor lamp has emission lines at 254, 312, 365, 405, 436, 546, 577, 691, and 773 nm.

15.2.1 https://chem.libretexts.org/@go/page/366515
 Figure 15.2.2 . Schematic diagram for measuring fluorescence showing the placement of the wavelength selectors for excitation and
emission. When a filter is used the instrument is called a fluorometer and when a monochromator is used the instrument is called a
spectrofluorometer.
The sample cells for molecular fluorescence are similar to those for molecular absorption. Remote sensing using a fiber optic
probe is possible using with either a fluorometer or spectrofluorometer. An analyte that is fluorescent is monitored directly. For an
analyte that is not fluorescent, a suitable fluorescent probe molecule is incorporated into the tip of the fiber optic probe. The
analyte’s reaction with the probe molecule leads to an increase or decrease in fluorescence.

Instruments for Measuring Phosphorescence

An instrument for molecular phosphorescence must discriminate between phosphorescence and fluorescence. Because the lifetime
for fluorescence is shorter than that for phosphorescence, discrimination is achieved by incorporating a delay between exciting the
sample and measuring the phosphorescent emission. Figure 15.2.3 shows how two out-of-phase choppers allow us to block
fluorescent emission from reaching the detector when the sample is being excited and to prevent the source radiation from causing
fluorescence when we are measuring the phosphorescent emission.

Figure 15.2.3 . Schematic diagram showing how choppers are used to prevent fluorescent emission from interfering with the
measurement of phosphorescent emission.
Because phosphorescence is such a slow process, we must prevent the excited state from relaxing by external conversion. One way
this is accomplished is by dissolving the sample in a suitable organic solvent, usually a mixture of ethanol, isopentane, and
diethylether. The resulting solution is frozen at liquid-N2 temperatures to form an optically clear solid. The solid matrix minimizes
external conversion due to collisions between the analyte and the solvent. External conversion also is minimized by immobilizing
the sample on a solid substrate, making possible room temperature measurements. One approach is to place a drop of a solution
that contains the analyte on a small disc of filter paper. After drying the sample under a heat lamp, the sample is placed in the

15.2.2 https://chem.libretexts.org/@go/page/366515
spectrofluorometer for analysis. Other solid substrates include silica gel, alumina, sodium acetate, and sucrose. This approach is
particularly useful for the analysis of thin layer chromatography plates.

This page titled 15.2: Instruments for Measuring Fluorescence and Phosphorescence is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by David Harvey.

15.2.3 https://chem.libretexts.org/@go/page/366515
15.3: Applications and Photoluminescence methods
Quantitative Applications
Molecular fluorescence and, to a lesser extent, phosphorescence are used for the direct or indirect quantitative analysis of analytes
in a variety of matrices. A direct quantitative analysis is possible when the analyte’s fluorescent or phosphorescent quantum yield is
favorable. If the analyte is not fluorescent or phosphorescent, or if the quantum yield is unfavorable, then an indirect analysis may
be feasible. One approach is to react the analyte with a reagent to form a product that is fluorescent or phosphorescent. Another
approach is to measure a decrease in fluorescence or phosphorescence when the analyte is added to a solution that contains a
fluorescent or phosphorescent probe molecule. A decrease in emission is observed when the reaction between the analyte and the
probe molecule enhances radiationless deactivation or results in a nonemitting product. The application of fluorescence and
phosphorescence to inorganic and organic analytes are considered in this section.

Inorganic Analytes
Except for a few metal ions, most notably UO , most inorganic ions are not sufficiently fluorescent for a direct analysis. Many
+
2

metal ions are determined indirectly by reacting with an organic ligand to form a fluorescent or, less commonly, a phosphorescent
metal–ligand complex. One example is the reaction of Al3+ with the sodium salt of 2, 4, 3′-trihydroxyazobenzene-5′-sulfonic acid
—also known as alizarin garnet R—which forms a fluorescent metal–ligand complex (Figure 15.3.1). The analysis is carried out
using an excitation wavelength of 470 nm, with fluorescence monitored at 500 nm. Table 15.3.1 provides additional examples of
chelating reagents that form fluorescent metal–ligand complexes with metal ions. A few inorganic nonmetals are determined by
their ability to decrease, or quench, the fluorescence of another species. One example is the analysis for F– based on its ability to
quench the fluorescence of the Al3+–alizarin garnet R complex.

Figure 15.3.1 . Structure of alizarin garnet R and its metal–ligand complex with Al3+.
Table 15.3.1 . Chelating Agents for the Fluorescent Analysis of Metal Ions
chelating agent metal ions

8-hydroxyquinoline Al3+, Be2+, Zn2+, Li+, Mg2+ (and others)

flavonal Zr2+, Sn4+

benzoin B4 O
2−
6
, Zn2+
′ ′ ′
2 , 3 , 4 , 5, 7 − pentahydroxylflavone Be2+

2-(o-hydroxyphenyl) benzoxazole Cd2+

Organic Analytes
As noted earlier, organic compounds that contain aromatic rings generally are fluorescent and aromatic heterocycles often are
phosphorescent. Table 15.3.2 provides examples of several important biochemical, pharmaceutical, and environmental compounds
that are analyzed quantitatively by fluorimetry or phosphorimetry. If an organic analyte is not naturally fluorescent or
phosphorescent, it may be possible to incorporate it into a chemical reaction that produces a fluorescent or phosphorescent product.
For example, the enzyme creatine phosphokinase is determined by using it to catalyze the formation of creatine from
phosphocreatine. Reacting the creatine with ninhydrin produces a fluorescent product of unknown structure.
Table 15.3.2 . Examples of Naturally Photoluminescent Organic Analytes
class compounds (F = fluorescence, P = phosphorescence)

phenylalanine (F)
aromatic amino acids tyrosine (F)
tryptophan (F, P)

15.3.1 https://chem.libretexts.org/@go/page/366516
 class compounds (F = fluorescence, P = phosphorescence)

vitamin A (F)
vitamin B2 (F)
vitamin B6 (F)
vitamins
vitamin B12 (F)
vitamin E (F)
folic acid (F)
dopamine (F)
catecholamines
norepinephrine (F)
quinine (F)
salicylic acid (F, P)
morphine (F)
barbiturates (F)
pharmaceuticals and drugs
LSD (F)
codeine (P)
caffeine (P)
sulfanilamide (P)
pyrene (F)
benzo[a]pyrene (F)
environmental pollutants organothiophosphorous pesticides (F)
carbamate insecticides (F)
DDT (P)

Standardizing the Method

In Section 15.1 we showed that the intensity of fluorescence or phosphorescence is a linear function of the analyte’s concentration
provided that the sample’s absorbance of source radiation (A = εbC ) is less than approximately 0.01. Calibration curves often are
linear over four to six orders of magnitude for fluorescence and over two to four orders of magnitude for phosphorescence. For
higher concentrations of analyte the calibration curve becomes nonlinear because the assumption that absorbance is negligible no
longer apply. Nonlinearity may be observed for smaller concentrations of analyte fluorescent or phosphorescent contaminants are
present. As discussed earlier, quantum efficiency is sensitive to temperature and sample matrix, both of which must be controlled
when using external standards. In addition, emission intensity depends on the molar absorptivity of the photoluminescent species,
which is sensitive to the sample matrix.

Representative Method: Determination of Quinine in Urine

The best way to appreciate the theoretical and the practical details discussed in this section is to carefully examine a typical
analytical method. Although each method is unique, the following description of the determination of quinine in urine provides an
instructive example of a typical procedure. The description here is based on Mule, S. J.; Hushin, P. L. Anal. Chem. 1971, 43, 708–
711, and O’Reilly, J. E.; J. Chem. Educ. 1975, 52, 610–612.
Description of the Method
Quinine is an alkaloid used to treat malaria. It is a strongly fluorescent compound in dilute solutions of H2SO4 (Φ = 0.55). f

Quinine’s excitation spectrum has absorption bands at 250 nm and 350 nm and its emission spectrum has a single emission band at
450 nm. Quinine is excreted rapidly from the body in urine and is determined by measuring its fluorescence following its extraction
from the urine sample.
Procedure
Transfer a 2.00-mL sample of urine to a 15-mL test tube and use 3.7 M NaOH to adjust its pH to between 9 and 10. Add 4 mL of a
3:1 (v/v) mixture of chloroform and isopropanol and shake the contents of the test tube for one minute. Allow the organic and the
aqueous (urine) layers to separate and transfer the organic phase to a clean test tube. Add 2.00 mL of 0.05 M H2SO4 to the organic
phase and shake the contents for one minute. Allow the organic and the aqueous layers to separate and transfer the aqueous phase
to the sample cell. Measure the fluorescent emission at 450 nm using an excitation wavelength of 350 nm. Determine the
concentration of quinine in the urine sample using a set of external standards in 0.05 M H2SO4, prepared from a 100.0 ppm solution
of quinine in 0.05 M H2SO4. Use distilled water as a blank.

15.3.2 https://chem.libretexts.org/@go/page/366516
Questions
1. Chloride ion quenches the intensity of quinine’s fluorescent emission. For example, in the presence of 100 ppm NaCl (61 ppm
Cl–) quinine’s emission intensity is only 83% of its emission intensity in the absence of chloride. The presence of 1000 ppm NaCl
(610 ppm Cl–) further reduces quinine’s fluorescent emission to less than 30% of its emission intensity in the absence of chloride.
The concentration of chloride in urine typically ranges from 4600–6700 ppm Cl–. Explain how this procedure prevents an
interference from chloride.
The procedure uses two extractions. In the first of these extractions, quinine is separated from urine by extracting it into a
mixture of chloroform and isopropanol, leaving the chloride ion behind in the original sample.
2. Samples of urine may contain small amounts of other fluorescent compounds, which will interfere with the analysis if they are
carried through the two extractions. Explain how you can modify the procedure to take this into account?
One approach is to prepare a blank that uses a sample of urine known to be free of quinine. Subtracting the blank’s
fluorescent signal from the measured fluorescence from urine samples corrects for the interfering compounds.
3. The fluorescent emission for quinine at 450 nm can be induced using an excitation frequency of either 250 nm or 350 nm. The
fluorescent quantum efficiency is the same for either excitation wavelength. Quinine’s absorption spectrum shows that ε is 250

greater than ε . Given that quinine has a stronger absorbance at 250 nm, explain why its fluorescent emission intensity is greater
350

when using 350 nm as the excitation wavelength.

We know that If is a function of the following terms: k, Φ , P0, ε , b, and C. We know that Φ , b, and C are the same for both
f f

excitation wavelengths and that ε is larger for a wavelength of 250 nm; we can, therefore, ignore these terms. The greater
emission intensity when using an excitation wavelength of 350 nm must be due to a larger value for P0 or k . In fact, P0 at
350 nm for a high-pressure Xe arc lamp is about 170% of that at 250 nm. In addition, the sensitivity of a typical
photomultiplier detector (which contributes to the value of k) at 350 nm is about 140% of that at 250 nm.

 Example

To evaluate the method described iabove, a series of external standard are prepared and analyzed, providing the results shown
in the following table. All fluorescent intensities are corrected using a blank prepared from a quinine-free sample of urine. The
fluorescent intensities are normalized by setting If for the highest concentration standard to 100.

[quinine] (µg/mL) If

1.00 10.11

3.00 30.20

5.00 49.84

7.00 69.89

10.00 100.0

After ingesting 10.0 mg of quinine, a volunteer provides a urine sample 24-h later. Analysis of the urine sample gives a relative
emission intensity of 28.16. Report the concentration of quinine in the sample in mg/L and the percent recovery for the
ingested quinine.

Solution
Linear regression of the relative emission intensity versus the concentration of quinine in the standards gives the calibration
curve shown below and the following calibration equation.

15.3.3 https://chem.libretexts.org/@go/page/366516
 g quinine 
If = 0.122 + 9.978 ×
mL

Substituting the sample’s relative emission intensity into the calibration equation gives the concentration of quinine as 2.81
μg/mL. Because the volume of urine taken, 2.00 mL, is the same as the volume of 0.05 M H2SO4 used to extract the quinine,
the concentration of quinine in the urine also is 2.81 μg/mL. The recovery of the ingested quinine is
2.81 μg quinine  1mg
× 2.00 mL urine  ×
mL urine  1000 μg
× 100 = 0.0562%
10.0 mg quinine ingested 

It can take 10–11 days for the body to completely excrete quinine so it is not surprising that such a small amount of quinine is
recovered from this sample of urine.

This page titled 15.3: Applications and Photoluminescence methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

15.3.4 https://chem.libretexts.org/@go/page/366516
15.4: Chemiluminscence
The focus of this chapter has been on molecular luminescence methods in which emission from the analyte's excited state is
achieved following its absorption of a photon. In Chapter 10 we considered atomic emission following excitation of the analyte by
thermal energy. An exothermic reaction may also serve as a source of energy. In chemiluminescence the analyte is raised to a
higher-energy state by means of a chemical reaction, emitting characteristic radiation when it returns to a lower-energy state. When
the chemical reaction results from a biological or enzymatic reaction, the emission of radiation is called bioluminescence.
Commercially available “light sticks” and the flash of light from a firefly are examples of chemiluminescence and
bioluminescence.
The intensity of emitted light, I , is proportional to the quantum yield for chemiluminescent emission, Φ , which is, itself the
CL

product of the quantum yield for creating excited states, Φ , and the quantum yield for emission through emission of a photon,
EX

ΦEM . The intensity also depends on the rate of the chemical reaction(s) responsible for creating the excited state; thus
dC
I = ΦC l ×
dt

where dC /dt is the rate of the chemical reaction.

Chemiluminescent measurements require less equipment than do other forms of molecular emission because there is no need for a
source of photons and no need for a monochromator as the only source of photons are those arising from the chemiluminescent
reaction. A sample cell to hold the reaction mixture and a photomultiplier tube may be sufficient for the optical bench. Because
chemiluminescent emission depends on the reaction's rate, and because the rate decreases with time, the intensity of emission is
time-dependent. As a result, the analytical signal is often the integrated emission intensity over a fixed interval of time.

This page titled 15.4: Chemiluminscence is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

15.4.1 https://chem.libretexts.org/@go/page/366517
15.5: Evaluation of Molecular Luminescence
Scale of Operation
Photoluminescence spectroscopy is used for the routine analysis of trace and ultratrace analytes in macro and meso samples.
Detection limits for fluorescence spectroscopy are influenced by the analyte’s quantum yield. For an analyte with Φ > 0.5 , a f

picomolar detection limit is possible when using a high quality spectrofluorometer. For example, the detection limit for quinine
sulfate, for which Φ is 0.55, generally is between 1 part per billion and 1 part per trillion. Detection limits for phosphorescence are
somewhat higher, with typical values in the nanomolar range for low-temperature phosphorimetry and in the micromolar range for
room-temperature phosphorimetry using a solid substrate.

Accuracy
The accuracy of a fluorescence method generally is between 1–5% when spectral and chemical interferences are insignificant.
Accuracy is limited by the same types of problems that affect other optical spectroscopic methods. In addition, accuracy is affected
by interferences that affect the fluorescent quantum yield. The accuracy of phosphorescence is somewhat greater than that for
fluorescence.

Precision
The relative standard deviation for fluorescence usually is between 0.5–2% when the analyte’s concentration is well above its
detection limit. Precision usually is limited by the stability of the excitation source. The precision for phosphorescence often is
limited by reproducibility in preparing samples for analysis, with relative standard deviations of 5–10% being common.

Sensitivity
The sensitivity of a fluorescent or a phosphorescent method is affected by a number of parameters. We already have considered the
importance of quantum yield and the effect of temperature and solution composition on Φ and Φ . Besides quantum yield,
f p

sensitivity is improved by using an excitation source that has a greater emission intensity, P0, at the desired wavelength, and by
selecting an excitation wavelength for which the analyte has a greater molar absorptivity, ε . Another approach for improving
sensitivity is to increase the volume from which emission is monitored. Figure 15.5.1 shows how rotating a monochromator’s slits
from their usual vertical orientation to a horizontal orientation increases the sampling volume. The result can increase the emission
from the sample by 5 − 30× .

Figure 15.5.1 . Use of slit orientation to change the volume from which fluorescence is measured: (a) vertical slit orientation; (b)
horizontal slit orientation. Suppose the slit’s dimensions are 0.1 mm × 3 mm. In (a) the dimensions of the sampling volume are 0.1
mm × 0.1 mm × 3 mm, or 0.03 mm3. For (b) the dimensions of the sampling volume are 0.1 mm × 3 mm × 3 mm, or 0.9 mm3, a
30-fold increase in the sampling volume.

Selectivity
The selectivity of fluorescence and phosphorescence is superior to that of absorption spectrophotometry for two reasons: first, not
every compound that absorbs radiation is fluorescent or phosphorescent; and, second, selectivity between an analyte and an
interferent is possible if there is a difference in either their excitation or their emission spectra. The total emission intensity is a
linear sum of that from each fluorescent or phosphorescent species. The analysis of a sample that contains n analytes, therefore, is
accomplished by measuring the total emission intensity at n wavelengths.

15.5.1 https://chem.libretexts.org/@go/page/366539
Time, Cost, and Equipment
As with other optical spectroscopic methods, fluorescent and phosphorescent methods provide a rapid means for analyzing samples
and are capable of automation. Fluorometers are relatively inexpensive, ranging from several hundred to several thousand dollars,
and often are satisfactory for quantitative work. Spectrofluorometers are more expensive, with models often exceeding $50,000.

This page titled 15.5: Evaluation of Molecular Luminescence is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

15.5.2 https://chem.libretexts.org/@go/page/366539
 CHAPTER OVERVIEW
16: An Introduction to Infrared Spectrometry
16.1: Theory of Infrared Absorption Spectrometry
16.2: Infrared Sources and Transducers
16.3: Infrared Instruments

This page titled 16: An Introduction to Infrared Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

1
16.1: Theory of Infrared Absorption Spectrometry
Understanding the IR Spectrum
Figure 16.1.1 shows the infrared spectrum for ethanol. Unlike a UV/Vis absorbance spectrum, the y-axis is displayed as percent
transmittance (%T) instead of absorbance, reflecting the fact that IR is used more for qualitative purposes than for quantitative
purposes, where Beer's law, which is a linear function of concentration (A = ϵbC ) makes absorbance the more useful
measurement. The x-axis for an IR spectrum usually is given in wavenumbers, ν = λ , with units of cm–1. The peaks in an IR
¯
¯¯ −1

spectrum are inverted relative to absorbance spectrum; that is, they descend from a baseline of 100%T instead of rising from a
baseline of zero absorbance.

Figure 16.1.1 : Infrared (IR) spectrum for ethanol.

Dipole Changes
The energy of a photon of infrared radiation (see Figure 16.1.2) is not sufficient to affect a change in the electronic energy levels of
electrons, as in the UV/Vis atomic or molecular absorption or emission spectroscopies covered in Chapters 9, 10, and 12–15.
Instead, infrared radiation is confined to changes in the vibrational energy states of molecules and molecular ions. To absorb an IR
photon, the absorbing species must experience a change in its dipole moment, which allows the oscillation in the photon's electrical
field to interact with an oscillation in charge within the absorbing species. If the two oscillations have the same frequency, then
absorption is possible.

 Note

Each vibrational energy state in Figure 16.1.2 also has a set of rotational energy states, which means that the peak for a
particular change in vibrational energy levels may consist of a series of closely spaced lines, one for each of several changes in
rotational energy. Because rotation is difficult for analytes that in liquid or solid forms, we usually see just a single, broad
absorption line; for this reason, we will consider only vibrational transitions in this chapter.

16.1.1 https://chem.libretexts.org/@go/page/366599
 Figure 16.1.2 : Diagram showing two electronic energy levels, E and E , each with five vibrational energy levels, ν − ν .
0 1 0 1

Absorption of ultraviolet and visible radiation (shown by the blue arrows) leads to a change in the analyte's electronic energy levels
and, possibly, a change in vibrational energy as well. A change in vibrational energy without a change in electronic energy levels
occurs with the absorption of infrared radiation (shown by the red arrows).

Types of Molecular Vibrations

Although we tend to think of the atoms in a molecule as being rigidly fixed in space relative to each other, the individual atoms are
in a constant state of motion: bond lengths increase and decrease by stretching and compressing, and bond angles change as the
result of the bending of the bonds relative to each other. Figure 16.1.3 shows two different types of stretching (symmetric and
asymmetric) and four different types of bending (in-plane rocking, in-plane scissoring, out-of-plane wagging, and out-of-plane
twisting).

Figure 16.1.3 : The six types of stretching and bending motions. For out-of-plane wagging and out-of-plane twisting, the plus sign
(+) indicates that the atom is moving toward the viewer and the minus sign (− ) indicate that the atom is moving away from the

viewer.
Even a simple molecule can have many vibrational modes that give rise to a peak in the IR spectrum, as is the case for ethanol
(Figure 16.1.1). The number of possible normal vibrational modes for a linear molecule is 3N − 5 , where N is the number of
atoms, and 3N − 6 for a non-linear molecule. Ethanol, for example, has 3 × 9 − 6 = 21 possible vibrational modes. As we will

16.1.2 https://chem.libretexts.org/@go/page/366599
see later in this section, some of these modes may not lead to a change in dipole moment, decreasing the number of peaks in the IR
spectrum.

 Note

Why does a non-linear molecule have 3N − 6 vibrational modes? Consider a molecule of methane, CH4. Each of methane’s
five atoms can move in one of three directions (x, y, and z) for a total of 3 × 5 = 15 different ways in which the molecule’s
atoms can move. A molecule can move in three ways: it can move from one place to another, which we call translational
motion; it can rotate around an axis, which we call rotational motion; and its bonds can stretch and bend, which we call
vibrational motion. Because the entire molecule can move in the x, y, and z directions, three of methane’s 15 different ways of
moving are translational. In addition, the molecule can rotate about its x, y, and z axes, accounting for three additional forms of
motion. This leaves 15 − 3 − 3 = 9 vibrational modes. A linear molecule, such as CO2, has 3N − 5 vibrational modes
because it can rotate around only two axes.

Mechanical Model of Stretching in a Diatomic Molecule

The simplest model system for the the stretching and compressing of a bond is a weight with a mass, m, attached to an ideal spring
that hangs from the ceiling as shown in Figure 16.1.4a. If we pull on the mass and then release it, we initiate a simple oscillating
harmonic motion that we can model using Hooke's law. If we displace the weight by a distance, y, then the force, F, that acts on the
weight is

F = −ky (16.1.1)

where k is the spring's force constant—a measure of the spring's springiness. The negative sign in Equation 16.1.1 indicates that
this is the force needed to restore the spring to its original position; that is, the force is in the direction opposite to our action of
pulling down on the weight.

Figure 16.1.4 : The model for a simple harmonic oscillator is shown in (a) and its potential energy as a function of displacement is
shown in (b).

Potential Energy of a Harmonic Oscillator

Let's take the potential energy, E, of the spring and weight as 0 when they are at rest (y = 0). If we pull down on the weight by a
distance of dy , then the change in the system's potential energy, dE , must increase by the product of force and distance
dE = −F × dy = −ky × dy (16.1.2)

Integrating Equation 16.1.2 from E = 0 to E = E and from y = 0 to y = y

E y

∫ dE = −k ∫ ydy (16.1.3)
0 0

gives the energy as

1 2
E = ky (16.1.4)
2

16.1.3 https://chem.libretexts.org/@go/page/366599
Figure 16.1.4b shows the resulting potential energy curve, for which the maximum potential energy is 1

2
kA
2
when the weight is at
its maximum displacement. Note that the potential energy curve is a parabola.

Vibrational Frequency
The simple harmonic oscillator described above and shown in Figure 16.1.4 vibrates with a frequency, ν , given by the equation
0

−−
−
1 k
ν0 = √ (16.1.5)
2π m

where k is the spring's force constant and m is the weight's mass. We can extend this to a spring that connects two weights to each
other by substituting for the mass, m, the system's reduced mass, μ
m1 × m2
μ = (16.1.6)
m1 + m2

where m and m are the masses of the two weights. Substituting Equation 16.1.6 into Equation 16.1.5 gives
1 2

−
− −−−−−−−−−−
1 k 1 k(m1 + m2 )
ν0 = √ = √ (16.1.7)
2π μ 2π m1 × m2

If we make the assumption that Equation 16.1.7 applies to simple diatomic molecules, then we can estimate the bond's force
constant, k , by measuring its vibrational frequency.

Quantum Treatment of Vibrations

Equations 16.1.4\ and 16.1.7 are based on a classical mechanics treatment of the simple harmonic oscillator in which any
displacement, and, thus, any energy is possible. Molecular vibrations, however, are quantized; thus
−
−
1 1 k 1
E = (v + ) ×h × √ = (v + ) h ν0 (16.1.8)
2 2π μ 2

where v is the vibrational quantum number, which has allowed values of 0, 1, 2, …. The difference in energy, ΔE, between any
two consecutive vibrational energy levels is hν . As allowed transitions in quantum mechanics are limited to Δν = ±1 and as the
0

difference in energy is limited to ΔE = hν , any particular mode of vibration should give rise to a single peak.
0

Anharmonic Behavior
The ideal behavior described in the last section, in which each vibrational motion that produces a change in dipole moment results
in a single peak, does not hold due to a variety of reasons, including the coulombic interactions between the atoms as they move
toward and away from each other. One result of these non-ideal behaviors is that the value ΔE does not remain constant for all
values of the vibrational quantum number v . For larger values of v , the value of ΔE becomes smaller and transitions where
Δv = ±2 or Δv = ±3 become possible giving rise to what are called overtone lines at frequencies that are 2× or 3× that for ν . 0

Why Do We See More or Fewer Vibrational Peaks Than Expected?

Figure 16.1.5 shows the IR spectrum for carbon dioxide, CO2, which consists of three clusters of peaks located at approximately
670 cm–1, 2350 cm–1, and 3700 cm–1. As carbon dioxide is a linear molecule that consists of two carbon-oxygen double bonds
(O=C=O), it has 3 × 3 − 5 = 9 − 5 = 4 vibrational modes. So why do we see just three clusters of peaks?

16.1.4 https://chem.libretexts.org/@go/page/366599
 Figure 16.1.5 : IR spectrum for CO2. The raw data is from NIST.
One of the requirements for the absorption of infrared radiation, is that the vibrational motion must result in a change in dipole
moment. Figure 16.1.6 shows the four vibrational modes for CO2. Of these four vibrational modes, the symmetric stretch does not
result in a change in dipole moment. Although this appears to explain why we see just three clusters of peaks, a close examination
of the two bending motions in Figure 16.1.6 should convince you that they are identical and, therefore, will appear as a single peak.

Figure 16.1.6 : The four vibrational modes for CO2. The symmetric stretch is not IR active as it does not involve a change in the
molecules dipole moment. The asymmetric stretch occurs at 2349 cm–1. The two bending motions are identical (or degenerate to
each other) and appear at 667 cm–1.
So what is the source of the cluster of peaks around 3700 cm–1? Sometimes the absorption of a single photon excites two or more
vibrational modes. In this case, the wavenumber for this absorption band is equivalent to the sum of the wavenumbers for the
asymmetric stretch and the two degenerate bending modes (2349 + 667 = 3016 cm–1, and 2349 + 667 + 667 = 3683 cm–1). These
are called combination bands.
Another source of additional peaks are overtone bands in which Δv = ±2 or Δv = ±3 . Figure 16.1.7 shows the IR spectrum for
carbonyl sulfide, OCS, which is analagous to CO2 in which one of the oxygens is replaced with sulfur. The peak at 520 cm–1 is for
its two degenerate bending motions and is labeled ν . The asymmetric stretch at 2062 cm–1 (ν ) and the symmetric stretch at 859
2 3

cm–1 (ν ) are the other two fundamental absorption bands. The remaining peaks are overtones, such as the peak labeled 2ν at
1 2

1040 cm–1, or combination bands, such as the peak labeled ν + ν at 2921 cm–1. Many of the peaks appear as two peaks; this is
3 1

the result of changes in rotational energy as well.

16.1.5 https://chem.libretexts.org/@go/page/366599
 Figure 16.1.7 : IR spectrum for OCS. The raw data is form NIST. The peak labeled as I (in red) is an impurity.

This page titled 16.1: Theory of Infrared Absorption Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

16.1.6 https://chem.libretexts.org/@go/page/366599
16.2: Infrared Sources and Transducers
Instrumentation for IR spectroscopy requires a source of infrared radiation and a transducer for detecting the radiation after it
passes through the sample.

Infrared Sources
Most IR sources consist of a solid material that emits radiation when heated by passing a current through the device. The intensity
of emitted light typically is greatest at 5900–5000 cm–1 and then decreases steadily to 500 cm–1. Common examples of IR sources
include the Nernst glower (a ceramic rod heated to 1200–2200 K), a globar (a silicon carbide rod heated to 1300–1500 K), an
incandescent wire (a nichrome wire heated to approximately 1100 K).

Infrared Transducers
Because IR radiation has a much lower energy than visible and ultraviolet light, the types of detectors used in UV/Vis spectroscopy
are not suitable for recording IR spectra. Most IR detectors measure heat either directly or by a temperature-dependent change in
one of its properties.
When two different metals, M1 and M2, are connected to each other in a closed loop, forming two M1–M2 junctions, a potential
difference exists between the two junctions. The magnitude of this difference in potential depends on the difference in the
temperatures of the two junctions. If the temperature of one junction is held constant or, if the source radiation is chopped—see
Chapter 9.2 for a discussion of chopping—then the change in temperature of the other junction can be measured. The active
junction is usually coated with a dark material to enhance the absorbance of thermal energy, and is small in size. A high-quality
thermocouple is sensitive temperature differences as small as 10–6 K.
A bolometer is fashioned from materials for which the resistance is temperature dependent. As is true for a thermocouple, the
active part of the detector is coated with a dark material and kept small in size.
Triglycerine sulfate, (NH2CH2COOH)3 • H2SO4, TGS, is a crystalline pyroelectric material. It usually is partially deuterated
(DTGS) and, perhaps, doped with L-alanine (DLaTGS). The pyroelectric material is placed between two electrodes, one of which
is optically transparent to infrared radiation. The absorption of infrared radiation results in a change in temperature and a resulting
change in the detector's capacitance and, therefore, the current that flows.

This page titled 16.2: Infrared Sources and Transducers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

16.2.1 https://chem.libretexts.org/@go/page/366600
16.3: Infrared Instruments
Instrumentation for infrared spectroscopy use one of three common optical benches: non-dispersive instruments, dispersive
instruments, and Fourier transform instruments. As we have already examined non-dispersive and dispersive instruments in
Chapter 13, and because they are no longer as common as they once were, we give them only a brief consideration here. Fourier
transform instruments, which dominate the current marketplace, receive a more detailed treatment.

Non-Dispersive Instruments
The simplest instrument for IR absorption spectroscopy is a filter photometer similar to that shown earlier in Figure 13.4.1 for
UV/Vis absorption. These instruments have the advantage of portability, which makes the useful in the field, and typically are used
as dedicated analyzers for gases such as HCN and CO.

Dispersive Instruments
Infrared instruments using a monochromator for wavelength selection use double-beam optics similar to that shown earlier in
Figure 13.4.3. Double-beam optics are preferred over single-beam optics because the sources and detectors for infrared radiation
are less stable than those for UV/Vis radiation. In addition, it is easier to correct for the absorption of infrared radiation by
atmospheric CO2 and H2O vapor when using double-beam optics. Resolutions of 1–3 cm–1 are typical for most instruments.

Fourier Transform Instruments

We covered the basic concepts of the Fourier transform in Chapter 7, which you may wish to review. In this section we take a more
detailed look at the application of Fourier transforms to infrared instrumentation.

Components of a FT-IR
In a Fourier transform infrared spectrometer, or FT–IR, the monochromator is replaced with an interferometer (Figure 16.3.1).
There are four key components that make up the interfometer: the drive mechanism that moves the moving mirror, the beam
splitter, the light source, and the detector.

Figure 16.3.1 : Schematic diagram of an interferometer. This is the same figure as Figure 7.7.1. The sample is positioned between
the beam splitter and the detector.
Drive Mechanism

As we learned in Chapter 7, the Fourier transform encodes information about the wavelength or frequency of source radiation
absorbed by the sample by observing how the signal reaching the detector varies with time. As the moving mirror is displaced in
space, some frequencies of light experience complete constructive interference, some frequencies of light experience complete
destructive interference, and other frequencies fall somewhere in between giving rise to a time domain spectrum. As the signal is
monitored as function of time and the moving mirror is traversing a variable distance, the drive mechanism must allow for a precise
and accurate relationship between the two. The mechanism of the moving mirror must be capable of moving the mirror through a
distance of up to 20 cm at a scan rate as fast as 10 cm/s; it must also accomplish this while maintaining the mirror's orientation
relative to the axis of its movement. To maintain accuracy, a HeNe laser, which emits visible light with a wavelength of 632.8 nm,
is aligned with the light source so that they follow the same optical path.

16.3.1 https://chem.libretexts.org/@go/page/366603
Beam Splitter
The beam splitter is designed to reflect 50% of the source radiation to the fixed mirror and to pass the remaining 50% of the source
radiation to the moving mirror. The materials used to construct the beam splitter depends on the range of wavelengths being used.
The most common range of wavelengths, which is called mid-IR, runs from approximately 670 cm–1 to 4000 cm–1. Instruments for
mid-IR use a beam splitter that consists of silicon or germanium coated onto a substrate of KBr or NaCl.
Sources and Transducers
The most common sources for FT-IR are those discussed in the previous section, such as a Nernst glower. The most common
transducer for FT-IR is pyroelectric triglycine sulfate.

This page titled 16.3: Infrared Instruments is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

16.3.2 https://chem.libretexts.org/@go/page/366603
 CHAPTER OVERVIEW
17: Applications of Infrared Spectrometry
Infrared spectroscopy finds wide use for both qualitative and quantitative analysis. Our organization of IR applications follows that
traditionally used by others by dividing the broad range of infrared radiation into three distinct units: the near-IR (4000 cm–1 to
14,000 cm–1, or 2,500 nm to 700 nm), the mid-IR (670 cm–1 to 4000 cm–1, or 15 µm to 2.5 µm), and the far-IR (670 cm–1 to 10
cm–1, or 1000 µm to 15 µm). Note that the near in near-IR means that it is nearest to the visible range of light. Of these, the most
important in terms of the breadth of applications is the mid-IR.
17.1: Mid-Infrared Absorption Spectometry
17.2: Mid-Infrared Reflection Spectrometry
17.3: Near-Infrared and Far-Infrared Spectroscopy

This page titled 17: Applications of Infrared Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

1
17.1: Mid-Infrared Absorption Spectometry
Mid-infrared spectrometry is used for the routine qualitative analysis and, to a lesser extent, the quantitative analysis of organic
molecules. In this section we consider absorption spectrometry in which we measure the absorbance of IR light as it passes through
a gas, solution, liquid, or solid sample. In Section 17.2 we consider reflectance spectrometry in which we measure the absorbance
of IR light as it reflects off the surface of a solid sample or a thin film of a liquid sample.

Sample Handling
Infrared spectroscopy is routinely used to analyze gas, liquid, and solid samples. We know from Beer's law, A = ϵbC , that
absorbance is a linear function of the analyte's concentration, C , and the distance, b , the light travels through the sample. The
challenge with obtaining an IR spectrum, is rarely the analyte's concentration or path length; instead it is finding materials and
solvents that are transparent to IR radiation. The optical windows in IR cells are made from materials, such as NaCl and KBr, that
are transparent to infrared radiation.

Gas Phase Samples

The cell for analyzing a sample in the gas phase generally is a 5–10 cm glass cylinder fitted with optically transparent windows.
For an analyte with a particularly small concentration, the sample cell is designed with reflective surfaces that allow the infrared
radiation to make several passes through the cell before it exits the sample cell, increasing the pathlength and, therefore, the
absorbance.

Solution Solutions
The analysis of a sample in solution is limited by the solvent’s IR absorbing properties, with carbon tetrachloride, CCl4, carbon
disulfide, CS2, and chloroform, CHCl3, being common solvents. A typical solution cell is shown in Figure 17.1.1. It is fashioned
with two NaCl windows separated by a spacer. By changing the spacer, pathlengths from 0.015–1.0 mm are obtained. The sample
is introduced into the cell using a syringe and the sample inlet port.

Figure 17.1.1 : IR cell for analytes that are in solution. (a) View from above showing the sample inlet, the sample outlet, and the
NaCl window. (b) View from the side showing the two NaCl plates.

Liquid Phase Samples

A sample that is a volatile liquid may be analyzed using the solution cell in Figure 17.1.1. For a non-volatile liquid sample,
however, a suitable sample for qualitative work can be prepared by placing a drop of the liquid between the two NaCl plates shown
in Figure 17.1.2a, forming a thin film that typically is less than 0.01 mm thick. An alternative approach is to place a drop of the
sample on a disposable card equipped with a polyethylene "window" that IR transparent with the exception of strong absorption
bands at 2918 cm–1 and 2849 cm–1 (Figure 17.1.2b).

17.1.1 https://chem.libretexts.org/@go/page/379840
 Figure 17.1.2 : Two examples of IR sample cells for liquid samples: (a) NaCl salts plates; (b) disposable card with a polyethylene
window that is IR transparent with the exception of strong absorption bands at 2918 cm–1 and 2849 cm–1.

Solid Phase Samples

Transparent solid samples are analyzed by placing them directly in the IR beam. Most solid samples, however, are opaque, and are
first dispersed in a more transparent medium before recording the IR spectrum. If a suitable solvent is available, then the solid is
analyzed by preparing a solution and analyzing as described above. When a suitable solvent is not available, solid samples are
analyzed by preparing a mull of the finely powdered sample with a suitable oil and then smearing it on a NaCl salt plate or a
disposable IR card (Figure 17.1.2). Alternatively, the powdered sample is mixed with KBr and pressed, under high pressure, into a
thin, optically transparent pellet, as shown in Figure 17.1.3.

Figure 17.1.3 : KBr press for preparing a solid sample for an IR analysis. The photo on the left shows the press. A threaded bolt is
screwed into the back of the press (hidden from view). A portion of the sample, mixed with KBr, is added to the press and the
second threaded bolt is screwed into the press. The two bolts are tightened by turning against each other, forming an optically
transparent pellet, as shown by the photo on the right that looks through the pellet. With the threaded bolts removed, the press sits
in a holder and placed in the spectrometer's optical path.

Qualitative Analysis
The most important application of mid-infrared spectroscopy is in the qualitative identification of organic molecules. Figure 17.1.4
shows mid-IR solution spectra for four simple alcohols: methanol, CH3OH, ethanol, CH3CH2OH, propanol, CH3CH2CH2OH, and
isopropanol, (CH3)2CHOH. Clearly there are similarities and differences in these four spectra: similarities that might lead us to
expect that each molecule contains the same functional groups and differences that appear as features unique to a particular
molecule. The similarities in these four spectra appear at the higher wavenumber end of the x-axis scale; we call the peaks we find
there group frequencies. The differences in these four spectra occur below approximately 1500 cm–1 in what we call the fingerprint
region.

 Note

The fingerprint region is defined here as beginning at 1500 cm–1, extending to the lowest wavenumber shown on the x-axis. If
you do some searching on the fingerprint region you will see that there is no broad agreement on where it begins. In my
searching, I found sources that place the beginning of the fingerprint region as 1500 cm–1, 1450 cm–1, 1300 cm–1, 1200 cm–1,
and 1000 cm–1.

17.1.2 https://chem.libretexts.org/@go/page/379840
 Figure 17.1.4 : IR spectra for four simple alcohols: methanol, CH3OH, ethanol, CH3CH2OH, propanol, CH3CH2CH2OH, and iso-
propanol, (CH3)2CHOH. Samples were prepared using carbon tetrachloride as a solvent. The vertical dashed red line at 1500 cm–1
marks the beginning of the fingerprint region. The original data for these spectra are from NIST.

Group Frequencies
All four of the spectra in Figure 17.1.4 share a small intensity, sharp peak at approximately 3650 cm–1, a strong intensity, broad
peak at approximately 3350 cm–1, and two medium intensity, sharp peaks at 2950 cm–1 and 3850 cm–1. By comparing spectra for
these and other compounds, we know that the presence of a broad peak between approximately 3200 cm–1 and 3600 cm–1 is good
evidence that the compound contains a hydrogen-bonded –OH functional group. The sharp peak at approximately 3650 cm–1 also is
evidence of an –OH functional group, but one that is not hydrogen-bonded. The two sharp peaks at 2950 cm–1 and 3850 cm–1 are
consistent with C–H bonds. All four of these peaks are for stretching vibrations. Tables of group frequencies are routinely
available.

The "Fingerprint" Region

Figure 17.1.5 shows a close-up of the fingerprint region for the alcohol samples in Figure 17.1.4. Of particular interest with this set
of samples is the increasing complexity of the spectra as we move from the simplest of these alcohols (methanol), to the most
complex of these alcohols (propanol and isopropanol). Also of interest is that each spectrum is unique in a way that allows us to
confirm a sample by matching it against a library of recorded spectra. There are a number of accessible collections of spectra that
are available for this purpose. One such collection of spectra is the NIST Webbook—NIST is the National Institute of Standards
and Technology—which is the source of the data used to display the spectra included in this section's figures and which includes
spectra for over 16,000 compounds.

17.1.3 https://chem.libretexts.org/@go/page/379840
 Figure 17.1.5 : IR spectra in the fingerprint region for four simple alcohols: methanol, CH3OH, ethanol, CH3CH2OH, propanol,
CH3CH2CH2OH, and iso-propanol, (CH3)2CHOH. Samples were prepared using carbon tetrachloride as a solvent. The original
data for these spectra are from NIST.

Computer Search Systems

With the availability of computerized data acquisition and storage it is possible to build digital libraries of standard reference
spectra. The identity of an a unknown compound often can be determined by comparing its spectrum against a library of reference
spectra, a process known as spectral searching. Comparisons are made using an algorithm that calculates the cumulative difference
between the sample’s spectrum and a reference spectrum. For example, one simple algorithm uses the following equation
n

D = ∑ |(Asample )i − (Aref erence )i | (17.1.1)

i=1

where D is the cumulative difference, Asample is the sample’s absorbance at wavelength or wavenumber i, Areference is the
absorbance of the reference compound at the same wavelength or wavenumber, and n is the number of digitized points in the
spectra. Note that the spectra are defined here by absobrance instead of transmittance as absorbance is directly proportional to
concentration. The cumulative absolute difference is calculated for each reference spectrum. The reference compound with the
smallest value of D is the closest match to the unknown compound. The accuracy of spectral searching is limited by the number
and type of compounds included in the library, and by the effect of the sample’s matrix on the spectrum.
Another advantage of computerized data acquisition is the ability to subtract one spectrum from another. When coupled with
spectral searching it is possible to determine the identity of several components in a sample without the need of a prior separation
step by repeatedly searching and subtracting reference spectra. An example is shown in Figure 17.1.6 in which the composition of
a two-component mixture is determined by successive searching and subtraction. Figure 17.1.6a shows the spectrum of the
mixture. A search of the spectral library selects cocaine•HCl (Figure 17.1.6b) as a likely component of the mixture. Subtracting the
reference spectrum for cocaine•HCl from the mixture’s spectrum leaves a result (Figure 17.1.6c) that closely matches mannitol’s
reference spectrum (Figure 17.1.6d). Subtracting the reference spectrum for mannitol leaves a small residual signal (Figure
17.1.6e).

17.1.4 https://chem.libretexts.org/@go/page/379840
 Figure 17.1.6 : Identifying the components of a mixture by spectral searching and subtracting. (a) IR spectrum of the mixture; (b)
Reference IR spectrum of cocaine•HCl; (c) Result of subtracting the spectrum of cocaine•HCl from the mixture’s spectrum; (d)
Reference IR spectrum of mannitol; and (e) The residual spectrum after removing mannitol’s contribution to the mixture’s
spectrum. IR spectra traditionally are displayed using percent transmittance, %T, along the y-axis. Because absorbance—not
percent transmittance—is a linear function of concentration, spectral searching and spectral subtraction, is easier to do when
displaying absorbance on the y-axis.

Quantitative Applications
A quantitative analysis based on the absorption of infrared radiation, although important, is encountered less frequently than with
UV/Vis absorption, primarily due to the three issues raised here.

Deviations from Beer's Law

One challenge for quantitative IR is the greater tendency for instrumental deviations from Beer’s law when using infrared radiation.
Because an infrared absorption band is relatively narrow, any deviation due to the lack of monochromatic radiation is more
pronounced. In addition, infrared sources are less intense than UV/Vis sources, which makes stray radiation more of a problem.
Differences between the path lengths for samples and for standards when using thin liquid films or KBr pellets are a problem,
although an internal standard can correct for any difference in pathlength; alternatively, we can use the cell shown in Figure 17.1.1
to maintain a constant path length.

Background Correction
The water and carbon dioxide in air have strong absorbances in the mid-IR. A double-beam dispersive instrument corrects for the
contributions of CO2 and H2O vapor because they are present in both pathways through the instrument. An FT-IR, however,
includes only a single optical path, so it is necessary to collect a separate spectrum to compensate for the absorbance of
atmospheric CO2 and H2O vapor. This is done by collecting a background spectrum without the sample and storing the result in the
instrument’s computer memory. The background spectrum is removed from the sample’s spectrum by taking the ratio the two
signals. Another approach is to flush the sample compartment with nitrogren.

17.1.5 https://chem.libretexts.org/@go/page/379840
Measuring Absorbance
Another challenge for quantitative IR is that establishing a 100% T (A = 0) baseline often is difficult because the optical properties
of NaCl sample cells may change significantly with wavelength due to contamination and degradation. We can minimize this
problem by measuring absorbance relative to a baseline established for the absorption band. Figure 17.1.7 shows how this is
accomplished.

Figure 17.1.7 : Method for determining absorbance from an IR spectrum.

Typical Applications
A recent review paper [Fahelelbom, K. M.; Saleh, A.; Al-Tabakha, M. A.; Ashames, A. A. Rev. Anal. Chem. 2022, 41, 21–33]
summarizes the rich literature in quantitative mid-infrared spectrometry. Among the areas covered are the analysis of
pharmaceuticals, including antibiotics, antihypertensives, antivirals, and counterfeit drugs. Mid-infrared spectrometry also finds use
for the analysis of environmentally significant gases, such as methane, CH4, hydrogen chloride, HCl, sulfur dioxide, SO2, and nitric
oxide, NO.

This page titled 17.1: Mid-Infrared Absorption Spectometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

17.1.6 https://chem.libretexts.org/@go/page/379840
17.2: Mid-Infrared Reflection Spectrometry
The first section of this chapter considered mid-IR absorption spectrometry in which we measure the amount of light that is
transmitted by the sample, which we can convert, if we wish, into absorbance values. In the process, we examined both
transmittance (Figure 7.1.4 and Figure 7.1.5) and absorbance (Figure 7.1.6) spectra. In this section, we consider experiments in
which we measure the reflection of infrared radiation by a sample.

Types of Reflections
There are two broad classes of reflection: internal and external. As shown in Figure 17.2.1, internal reflection occurs when light
encounters an interface between two media—here identified as the sample and the support—that have different refractive indicies,
n. When the refractive index of the support is greater than the refractive index of the sample, then some of the light reflects off the
interface. Attenuated total reflectance spectrometry is one example of an instrumental method that relies on internal reflection.

Figure 17.2.1 : Internal reflection occurs at the interface between a support that has a greater refractive index than an overlying
sample. For small angles of incidence (dashed lines), the light experiences refraction and crosses the interface. For a sufficiently
large angle of incidence, the light experiences total reflection, as shown by the solid line.
External reflectance occurs when light reflects off of the sample's surface. As shown in Figure 17.2.2, the way in which light
reflects depends on the nature of the sample's surface. In specular reflectance, the angle of reflection is the same at all locations
because the sample's surface is smooth; in diffuse reflectance, the angle of reflection varies between locations due to the roughness
of the sample's surface. Diffuse reflectance spectrometry is one example of an instrumental method that relies on external
reflection.

Figure 17.2.2 : (a) Specular reflectance at a smooth surface and (b) diffuse reflectance from a rough surface.

Attenuated Total Reflectance Spectrometry

The analysis of an aqueous sample is complicated by the solubility of the NaCl cell window in water. One approach to obtaining an
infrared spectrum of an aqueous solution is to use attenuated total reflectance instead of transmission. Figure 17.2.3 shows a
diagram of a typical attenuated total reflectance (ATR) FT–IR instrument. The ATR cell consists of a high refractive index material,
such as ZnSe or diamond, sandwiched between a low refractive index substrate and a lower refractive index sample. Radiation
from the source enters the ATR crystal where it undergoes a series of internal reflections before exiting the crystal. During each
reflection the radiation penetrates a short distance into the sample. This depth of penetration, d , depends on the wavelength of the
p

light, λ , the refractive index of the ATR crystal, n , the refractive index of the sample, n , and the angle of the incident radiation,
1 2

θ.

λ
dp = −−−−−−−−−−− (17.2.1)
2 2 2
2π √ n sin θ−n
1 2

17.2.1 https://chem.libretexts.org/@go/page/379842
For example, when using ZnSe as the ATR crystal (n = 2.4 ) and an angle of incidence of 45 , light of 1000 cm–1 penetrates to a
1
∘

depth of 2.0 µm in a sample with a refractive index similar to that for KBr (n   = 1.5 ).
2

Figure 17.2.3 .
FT-IR spectrometer equipped with a diamond ATR sample cell. The inserts show a
close-up photo of the sample platform, a sketch of the ATR’s sample slot, and a schematic
showing how the source’s radiation interacts with the sample. The pressure tower is used to
ensure proper contact of a solid sample with the ATR crystal.
Solid samples also can be analyzed using an ATR sample cell. After placing the solid in the sample slot, a compression tip ensures
that it is in contact with the ATR crystal. Examples of solids analyzed by ATR include polymers, fibers, fabrics, powders, and
biological tissue samples. ATR spectra are similar, but not identical, to those obtained by measuring transmission. An important
contribution to this is the wavelength-dependent depth of penetration of the infrared radiation where a decrease in wavenumber
(longer wavelength) results in a greater depth of penetration, which changes the intensity and width of absorption bands.

Diffuse Reflectance Spectrometry

Another reflectance method is diffuse reflectance, in which radiation is reflected from a rough surface, such as a powder. Powdered
samples are mixed with a non-absorbing material, such as powdered KBr, and the reflected light is collected and analyzed. As with
ATR, the resulting spectrum is similar to that obtained by conventional transmission methods. Figure 17.2.4 shows the IR spectrum
for urea obtained using transmission and diffuse reflectance (both collected using an FT-IR). Both spectra show similar features
between 1000 cm–1 and 2000 cm–1, although there are differences in relative peak heights and background absorption.

17.2.2 https://chem.libretexts.org/@go/page/379842
 Figure 17.2.4 : Diffuse reflection spectra for urea recorded using a FT-IR operating in transmission sample cell (top) or using a
diffuse reflectance sample cell (bottom). The y-axis for the transmission spectrum is –log(T), where T is the transmittance, and the
y-axis for the diffuse reflectance spectrum is –log(R), where R is the reflectance.

This page titled 17.2: Mid-Infrared Reflection Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

17.2.3 https://chem.libretexts.org/@go/page/379842
17.3: Near-Infrared and Far-Infrared Spectroscopy
At the beginning of this chapter we divided infrared radiation into three areas: the near-IR, the mid-IR, and the far-IR. The mid-IR,
which runs from 4000 cm–1 to 670 cm–1 (2.5 µm to 15 µm) is the most analytical useful region and was the subject of the previous
two sections. Here we briefly turn our attention to applications using the near-IR and the far-IR.

Near-Infrared (NIR) Spectroscopy

The near-IR extends from approximately 13,000 cm–1 (a wavelength of 770 nm or 0.77 µm, the upper wavelength limit of visible
light) to 4000 cm–1 (a wavelength of 2,500 nm or 2.5 µm). Earlier we noted that absorption bands in the region that extends from
1500 cm–1 to 4000 cm–1 are called group frequencies. The absorption bands in the near-infrared often are overtones and
combination bands of these group frequencies. Of particular importance are functional groups that include hydrogen: OH, CH, and
NH are examples. Absorption bands generally are less intense and less broad. Compared to mid-IR, the NIR is more useful for a
quantitative analysis of aqueous samples because the OH absorption bands are much weaker. The instrumentation for NIR
spectroscopy, both in transmission mode and in reflectance mode, is similar to that for UV/visible spectrometers and for mid-IR
spectrometry.

Far-Infrared (FIR) Spectroscopy

The far-IR extends from approximately 670 cm–1 (a wavelength of 15 µm) to 10 cm–1 (a wavelength of 1000 µm or 1 mm). FIR
spectroscopy finds applications in the analysis of materials that include metals, including metal oxides, metal sulfides, and metal-
ligand complexes. FIR spectroscopy has also been applied to the analysis of polyamides, peptides, and proteins. Because the FIR
merges into the microwave region, it also finds use in the analysis of the rotational energies of gases.

This page titled 17.3: Near-Infrared and Far-Infrared Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

17.3.1 https://chem.libretexts.org/@go/page/379844
 CHAPTER OVERVIEW
18: Raman Spectroscopy
18.1: Theory of Raman Spectroscopy
18.2: Instrumentation
18.3: Applications of Raman Spectroscopy
18.4: Other Types of Raman Spectroscopy

This page titled 18: Raman Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1
18.1: Theory of Raman Spectroscopy
The blue color of the sky during the day and the red color of the sun at sunset are the result of the scattering of light by small
particles of dust, by molecules of water vapor, and by other gases in the atmosphere. The efficiency of a photon's scattering
depends on its wavelength. We see the sky as blue during the day because violet and blue light scatter to a greater extent than other,
longer wavelengths of light. For the same reason, the sun appears red at sunset because red light scatters less efficiently and is more
likely to pass through the atmosphere than other wavelengths of light. If we send a focused, monochromatic beam of radiation with
a wavelength of λ through a medium of particles—whether solid particulates or individual molecules—that have dimensions
< 1.5λ, then the radiation scatters in all directions. For example, infrared light in the near-IR with a wavelength of 700 nm will

scatter from any particle whose longest dimension is less than 1,300 nm. Even in an otherwise transparent sample, scattering from
molecules occurs.

Raman Spectra
There are two general classes of scattering: elastic scattering and inelastic scattering. In elastic scattering, a photon is first absorbed
by a particle and then emitted without a change in its energy (ΔE = 0 ); this is called Rayleigh scattering. With inelastic scattering,
a photon is first absorbed by a particle and then emitted with a change in its energy (ΔE ≠ 0 ); this is called Raman scattering. A
plot that shows the intensity of scattered radiation as a function of the scattered photon's energy, expressed as a change in the
wavenumber, Δν̄ , is called a Raman spectrum and values of Δν̄ are called Raman shifts.
¯
¯ ¯
¯

Figure 18.1.1 shows a portion of the Raman spectrum for carbon tetrachloride and illustrates several important features. First,
Rayleigh scattering produces an intense peak at Δν̄ = 0 . Although the peak is intense, it carries no useful information as the
¯
¯

absolute energy is just that for the source. Second, Raman scattering has two components—Stokes lines and the anti-Stokes lines—
that have identical absolute shifts relative to the line for Rayleigh scattering, but that have different signs. The Stokes lines have
positive values for Δν̄ and the anti-Stokes lines have negative values for Δν̄ . Third, each of the Stokes lines is more intense than
¯
¯ ¯
¯

the corresponding anti-Stokes line. Fourth, because we measure the shift in a peak's wavenumber relative to the source radiation,
the spectrum is independent of the source radiation.

Figure 18.1.1 : The Raman spectrum for carbon tetrachloride, CCl4, showing the intense peak for Rayleigh scattering at Δν̄ = 0
¯
¯

and five lines each for Stokes and anti-Stokes scattering with Δν̄ = ±217.0 cm , Δν̄ = ±313.5 cm , Δν̄ = ±458.7 cm ,
¯
¯ −1 ¯
¯ −1 ¯
¯ −1

¯
¯
Δν̄ = ±761.7 cm , and Δν̄ = ±790.4 cm . The values for these lines are from NIST; the spectrum itself is simulated.
−1 ¯
¯ −1

 Note
The energy—and, thus, the wavenumber—of a photon that experiences Stokes scattering is less than the energy—and, thus, the
wavenumber—of the source radiation, which begs the question of why a Stokes shift is reported as a positive value instead of a
negative value. Although you will find most Raman spectra with positive values for the Stokes shift, you also will find
examples where Stokes shifts are reported with negative values. Because the Stokes lines are more intense than the anti-Stokes
lines, and, therefore, more useful, and because their respective shifts result from the same changes in vibrational energy states

18.1.1 https://chem.libretexts.org/@go/page/386418
 that we find in IR spectroscopy, it is convenient to report the Stokes lines as positive values so that we can align a species's
Raman and IR spectra. See the next two sections for additional details.

Mechanism of Raman and Rayleigh Scattering

In Chapter 6 we examined the mechanism by which absorption and emission occur. In subsequent chapters we explored atomic
absorption and atomic emission spectrometry, ultraviolet and visible molecular absorption spectrometry, molecular luminescence
spectrometry, and infrared molecular absorption spectrometry. In each case we began by considering an energy level diagram that
explains the origin of absorption and emission. Figure 18.1.2 provides an energy diagram that we can use to explain the origin of
the lines that make up a Raman spectrum, such as the spectrum for carbon tetrachloride in Figure 18.1.1.

Figure 18.1.2 : Energy level diagram that shows the origin of Rayleigh, Stokes, and anti-Stokes scattering. The thickness of the
individual arrows provides a sense of the relative number of photons absorbed or emitted—the thicker the arrow, the greater the
number of photons—although the individual thicknesses are not to scale.
The first thing to note about the energy level diagram in Figure 18.1.2 is that, in addition to showing the ground electronic state and
the first excited electronic state—each with three vibrational energy levels—it also shows a virtual electronic state, something we
did not encounter with other methods (see, for example, the energy diagram for UV and IR molecular absorption spectrometry in
Figure 6.4.2). The ground and the first excited electronic states are quantized, which means that absorption cannot happen if the
source's energy does not match exactly the change in energy between the two electronic states. The energy of an emitted photon
also is fixed by the difference in the energy of the two electronic states. A virtual electronic state, however, is not quantized and is
determined by the energy of the source radiation. The source of radiation, therefore, does not need to match a particular change in
energy.
Absorption of a photon of source radiation moves the analyte from the ground electronic state to a virtual electronic state without a
change in vibrational energy state, as seen by the two arrows at the far left of the diagram. Because the ground vibrational energy
state, ν , is more populated than the vibrational energy state, ν , more of the analyte ends up in a virtual electronic state's lowest
0 1

vibrational energy than in a higher vibrational energy state, which is shown here by the relative thickness of the two arrows.
Once in a virtual electronic state, the analyte can return to the ground excited state in one of three ways. It can do so without a
change in the vibrational energy level. In this case, the energy of absorption and the energy of emission are the same and ΔE = 0
and Δν̄ = 0 . This is Rayleigh scattering and, as suggested by the combined thickness of the two arrows in Figure 18.1.2, it is the
¯
¯

most important mechanism of relaxation.

When relaxation includes a change in the vibrational energy level, the result is an absolute change in energy equivalent to the
difference in energy, ΔE, between adjacent vibrational energy levels. For Stokes scattering, relaxation is to a higher vibrational
energy level, such as ν → ν and, for anti-Stokes scattering, relaxation to a lower vibrational energy level, such as ν → ν . As
0 1 1 0

suggested by the thickness of the lines for Stokes and anti-Stokes scattering in Figure 18.1.2, the Stokes lines are more intense than
the anti-Stokes lines because they begin in a more heavily populated excited state.

18.1.2 https://chem.libretexts.org/@go/page/386418
Relationship Between IR and Raman Spectra
One important feature of Figure 18.1.2 is that the transition that gives rise to a particular Stokes line or anti-Stokes line is the same
transition that will give rise to a corresponding IR band. If the selection rules for these transitions are the same for a particular
species, then we expect that its IR spectrum and its Raman spectrum will have peaks at the same (or similar) values of ν̄ and Δν̄ ¯
¯ ¯
¯

for its fundamental vibrations; however, as we see in Table 18.1.1 for carbon tetrachloride, CCl4, there are five fundamental
vibrations in its Raman spectrum, but just three in its IR spectrum.
Table 18.1.1 : Fundamental Vibrational Energies for CCl4 (values from NIST).
Infrared (ν , cm–1)
¯
¯¯
Raman (Δν , cm–1)
¯
¯¯

— 217.0 (l)

309.9 (l) weak 313.5 (l)

— 458.7 (l)

768 (g); very strong 761.7 (l)

789 (g); very strong 790.4 (l)

The designations (g) and (l) indicate the sample's phase where (g) is gas phase and (l) is liquid phase. The designations of weak and very strong
for the IR peaks indicates the relative extent of absorption: very strong means a relatively small %T (and a strong absorbance) and weak means a
relatively large %T (and a weak absorbance). See Figure 18.1.1 for the relative amount of scattering in the Raman spectrum for CCl4.

In Chapter 16 we learned that in IR spectroscopy a compound's fundamental vibrational energy is active—that is, we see a peak in
its IR spectrum—only if the corresponding stretch or bend results in a change in the compound's dipole moment. For Raman
spectroscopy, a compound's fundamental vibrational energy is active only if the corresponding stretch or bend results in a change in
the polarizability of its electrons. Polarizability essentially is a measure of how easy it is to distort a compound's electron cloud by
applying an external electric field, such as when a photon from the source is absorbed; in general, polarizability increases when a
stretching or bending motion increases the compound's volume as the electrons are then spread over a greater amount of space.
Figure 18.1.3 shows the four stretching and bending modes for CCl4. The stretching motion in (a), in which all four C–Cl bond
lengths increase and decrease together, means the molecule's volume increases and decreases; thus, this vibrational mode is Raman
active. The symmetry of the stretching motion, however, means there is no change in the molecule's dipole moment and the
vibrational mode is IR inactive. The asymmetric stretch in (b), on the other hand, is both IR and Raman active. The bending motion
in (c) results in the molecule becoming more or less compact in size, and is Raman active; the symmetry of the scissoring motions,
however, means that the vibrational mode is IR inactive. The bending motions in (d) are both IR and Raman active.

18.1.3 https://chem.libretexts.org/@go/page/386418
 Figure 18.1.3 . The four fundamental types of vibrational motions for CCl4. The molecule has a tetrahedral shape. The black arrows
show the direction of movement of the individual atoms. A molecule of CCl4 has 3 × 5 − 6 = 15 − 6 = 9 vibrational modes. The
bending motion in (d) and the stretching motion in (b) are triply degenerate, the bending motion in (c) is doubly degenerate, and the
stretching motion in (a) is singly degenerate; this accounts for the nine possible vibrational modes. The asymmetric stretching
motion in (b) gives rise to a split peak for reasons we will not consider here.
In general, symmetric stretching and bending modes result in relatively strong Raman scattering peaks, but no absorption in the IR,
while symmetric stretching and bending modes result in both IR and Raman peaks. As a result, IR and Raman are complementary
techniques.

Raman Depolarization Ratios

If the source of electromagnetic radiation is plane-polarized, then it is possible to collect a Raman spectrum using light scattered in
a plane that is parallel to the source and, separately, in a plane that is perpendicular to the source. The ratio of a line's intensity of
scattering in the perpendicular spectrum, I , to the intensity of scattering in the parallel spectrum, I , is called the depolarization
⊥ ||

ratio, p.
I⊥
p = (18.1.1)
I||

A Raman line that originates from a vibrational mode that does not change the molecules shape will result in a depolarization ratio
close to zero and an absence of the line in the perpendicular spectrum. Figure 18.1.4 shows the Raman spectrum when collecting
data parallel (top) and perpendicular (bottom) to the light source. The absence of the peak at 458.7 cm–1 in the perpendicular
spectrum confirms that this is the symmetric stretch illustrated in Figure 18.1.3a.

18.1.4 https://chem.libretexts.org/@go/page/386418
 Figure 18.1.4 : Raman spectrum for CCl4 measured (top) parallel to the plane-polarized light source and (bottom) perpendicular to
the plane-polarized source. The dashed red line shows the position of the symmetric stretch at 458.7 cm–1; the absence of this peak
in the perpendicular spectrum—the depolarization ratio, p is 0.005—confirms that this is the symmetric stretch. The remaining
peaks have depolarization ratios of 0.75. The slope of increasing intensity below 200 cm–1 is from the Rayleigh scattering peak.
The values for these lines are from NIST; the spectrum itself is simulated.

This page titled 18.1: Theory of Raman Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

18.1.5 https://chem.libretexts.org/@go/page/386418
18.2: Instrumentation
The basic instrumentation for Raman spectroscopy is similar to that for other spectroscopic techniques: a source or radiation, an
optical bench for bringing the source to the sample, and a suitable detector.

Sources
One of the notable features of the Raman spectrum for CCl4 (see Figure 18.1.1) is the low intensity of the Stokes lines and the anti-
Stokes lines relative to the line for Rayleigh scattering. The low intensity of these lines requires that we use a high intensity source
so that there are a sufficient number of scattered photons to collect. For this reason, a laser is the most common source for Raman
spectroscopy, providing a high intensity, monochromatic source. Table 18.2.1 summarizes some of the more common lasers.
Table 18.2.1 . Examples of Laser Sources for Raman Spectroscopy
type of laser wavelengths (nm)

Ar ion 488.0 or 514.5

Kr ion 530.9 or 647.1

He/Ne 632.8

Near Infrared (NIR) Diode Laser 785 or 830

Nd/YAG 532 or 1064

The intensity of Raman scattering is proportional to 1

4
, where λ is the wavelength of the source radiation; thus, the smaller the
λ

wavelength, the more intense the intensity of scattered light. For example, the intensity of scattering using an Ar ion laser at 488.0
nm is almost 23× greater than the intensity of scattering using a Nd/YAG laser at 1064 nm
4
(1/480)
= 22.6
4
(1/1064)

The increased scattering when using a smaller wavelength laser comes at a cost, however, of an interference from fluorescence
from species that are promoted into excited electronic states by the source. The NIR diode laser and Nd/YAG laser, when operated
at 1064 cm–1, discriminate against fluorescence and are useful, therefore, for samples where fluorescence is a problem.

Samples
Raman spectroscopy has several advantages over infrared spectroscopy. Because water does not exhibit much Raman scattering it
is possible to analyze aqueous samples; this is a serious limitation for IR spectroscopy where water absorbs strongly. The ability to
focus a laser onto a small area makes it possible to analyze very small samples. A liquid sample, for example, can be held in the tip
of a 1-mm inner diameter capillary tube, such as that used for measuring melting points. Solid samples and gaseous samples can be
sampled using the same types of cells used in IR and FT-IR (see Chapter 17). Fiber optic probes make it possible to collect samples
remotely. Figure 18.2.1 shows the basic set-up. A small bundle of fibers (shown in blue) brings light from the source to the sample
where a second bundle of fibers (shown in green) brings the scattered light to the slit that passes light onto the detector.

18.2.1 https://chem.libretexts.org/@go/page/386420
 Figure 18.2.1 : Illustration of a fiber optic probe for Raman spectroscopy. The darker blue circles are cross-sections of individual
fibers that bring light from the source to the sample, and the lighter blue circles are cross-sections of individual fibers that collect
scattered light from the sample and bring it to the detector. The circular arrangement of the collection fibers are gathered into a
linear array and focused onto the entrance slit of the instrument's detector.

Optical Bench
Raman spectrometers use optical benches similar to those for UV/Vis or IR spectroscopy, which were covered in Chapter 7.
Dispersive instruments place the laser source and the detector at 90° to each other so that any unscattered high intensity emission
from the laser source is not collected by the detector. A filter is used to remove the Rayleigh scattering. To record a spectrum one
either uses a scanning monochromator or a multichannel detector. Fourier transform instruments are similar to those used in FT-IR
and include a filter to isolate the Stokes lines.

This page titled 18.2: Instrumentation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

18.2.2 https://chem.libretexts.org/@go/page/386420
18.3: Applications of Raman Spectroscopy
Raman spectroscopy is useful for both qualitative and quantitative analyses, examples of which are provided in this section.

Qualitative Applications
There are numerous databases that provide reference spectra for inorganic compounds, for minerals, for synthetic organic pigments,
for natural and synthetic inorganic and organic pigments, and for carbohydrates. Such data bases are often searchable by not only
name and formula, but by the prominent Raman scattering lines. Examples of spectra are included here using data from the
databases linked to above.

Figure 18.3.1 : Examples of Raman spectra for the mineral calcite (CaCO3), the organic pigment alizarine red (C14H7NaO7S), the
inorganic pigment azurite (Cu3(CO3)2(OH)2), and the carbohydrate fructose (C6H12O6). The data for these spectra were obtained
from the databases linked to at the beginning of this section.

Quantitative Applications
The intensity of Raman scattering, I (ν ) , is directly proportional to the intensity of the source radiation, I , and the concentration
R l

of the scattering species, C . The direct proportionality between I (ν ) and I is important given that each photon experiencing
R l

Raman scattering requires approximately 10 excitation photons. Using a laser as a source of radiation and increasing its power
8

leads to an improvement in sensitivity. The direct proportionality between I (ν ) and the concentration of the scattering species
R

means that a calibration curve of band intensity (or band area) is a linear function of concentration, allowing for a quantitative
analysis.

This page titled 18.3: Applications of Raman Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

18.3.1 https://chem.libretexts.org/@go/page/386422
18.4: Other Types of Raman Spectroscopy
Traditional Raman spectroscopy has several limitations, perhaps the most important of which is that the probability of Raman
scattering is much less than that for Rayleigh scattering, which leads to low sensitivity with detection limits often as large as 0.1 M.
Here we briefly describe two forms of Raman spectroscopy that allow for significant improvements in detection limits.

Resonance Raman Spectroscopy (RRS)

If the wavelength of the source is similar to the wavelength needed to move the species from its ground electronic state to its first
electronic excited state (not the virtual excited state shown in Figure 18.1.2), then the lines associated with the symmetric
fundamental vibrations increase in intensity by a factor of 10 to 10 . The improvement in sensitivity results in a substantial
2 6

reduction in detection limits as low as 10  M . The use of a tunable laser makes it possible to adjust the wavelength of light
−8

emitted by the source to maximize the intensity of scattering.

Surface-Enhanced Raman Spectroscopy (SERS)

For reasons that are poorly understood, the intensity of Raman scattering lines is enhanced when the scattering species is absorbed
to the surface of colloidal particles of metals such as Ag, Au, or Cu, or to the surface of etched metals. The phenomenon is not
limited to just a few lines—as is the case for RRS—and results in a 10 to 10 ) improvement in the intensity of scattering. If a
3 6

tunable laser is used for the source, allowing for both RRS and SERS, detection limits of 10  M to 10  M are possible.
−9 −12

This page titled 18.4: Other Types of Raman Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

18.4.1 https://chem.libretexts.org/@go/page/386424
 CHAPTER OVERVIEW
19: Nuclear Magnetic Resonance Spectroscopy
19.1: Theory of Nuclear Magnetic Resonance
19.2: Environmental Effects on NMR Spectra
19.3: NMR Spectrometers
19.4: Applications of Proton NMR
19.5: Carbon-13 NMR
19.6: Two-Dimensional Fourier Transform NMR

This page titled 19: Nuclear Magnetic Resonance Spectroscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

1
19.1: Theory of Nuclear Magnetic Resonance
As is the case with other forms of optical spectroscopy, the signal in nuclear magnetic resonance (NMR) spectroscopy arises from a
difference in the energy levels occupied by the nuclei in the analyte. In this section we develop a general theory of nuclear
magnetic resonance spectroscopy that draws on quantum mechanics and on classical mechanics to explain these energy levels.

Quantum Mechanical Description of NMR

The quantum mechanical description of an electron is given by four quantum numbers: the principal quantum number, n , the
angular momentum quantum number, l, the magnetic quantum number, m , and the spin quantum number, m . The first three of
l s

these quantum numbers tell us something about where the electron is relative to the nucleus and something about the electron's
energy. The last of these four quantum numbers, the spin quantum number, tells us something about the ability of an electron to
interact with an applied magnetic field. An electron has possible spins of +1/2 or of –1/2, which we often refer to as spin up, using
an upwards arrow, ↑, to represent it, or as spin down, using a downwards arrow, ↓, to represent it.
A nucleus, like an electron, carries a charge and has a spin quantum number. The overall spin, I , of a nucleus is a function of the
number of protons and neutrons that make up the nucleus. Here are three simple rules for nuclear spin states:
If the number of neutrons and the number of protons are both even numbers, then the nucleus does not have a spin; thus, 12C,
with six protons and six neutrons, has no overall spin and I = 0 .
If the number of neutrons plus the number of protons is an odd number, then the nucleus has a half-integer spin, such as 1/2 or
3/2; thus, 13C, with six protons and seven neutrons, has an overall spin of I = 1/2 ; this also is true for 1H.
If the number of neutrons and the number of protons are both odd numbers, then the nucleus has an integer spin, such as 1 or 2;
thus, 2H, with one proton and one neutron, has an overall spin of I = 1 .

 Note
Predicting that 13C has a spin of I = 1/2 , but that 127I has a spin of I = 3/2 and that 17O has a spin of I = 5/2 is not trivial.
A periodic table that provides spin states for elements is available here.

The total number of spin states—that is, the total number of possible orientations of the spin—is equal to (2 × I ) + 1 . To be NMR
active, a nucleus must have at least two spin states so that a change in spin states, and, therefore, a change in energy, is possible;
thus, 12C, for which there are (2 × 0) + 1 = 1 spin states, is NMR inactive, but 13C, for which there are (2 × 1/2) + 1 = 2 spin
states with values of m = +1/2 and of m = −1/2 , is NMR active, as is 2H for which there are (2 × 1) + 1 = 3 spin states with
values of m = +1/2 , m = 0 , and m = −1/2 . As our interest in this chapter is in the NMR spectra for 1H and for 13C, we will
limit ourselves to considering I = 1/2 and spin states of m = +1/2 and of m = −1/2 .

Energy Levels in an Applied Magnetic Field

Suppose we have a large population of 1H atoms. In the absence of an applied magnetic field the atoms are divided equally between
their possible spin states: 50% of the atoms have a spin of +1/2 and 50% of the atoms have a spin of –1/2. Both spin states have the
same energy, as is the case on the left side of Figure 19.1.1, and neither absorption nor emission occurs.

Figure 19.1.1 . Energy level diagram for a proton, 1H, in the absence (left) and in the presence (right) of an applied magnetic field.
In the presence of an applied magnetic field, as on the right side of Figure 19.1.1, the nuclei are either aligned with the magnetic
field with spins of m = +1/2 , or aligned against the magnetic field with spins of m = −1/2 . The energies in these two spin

19.1.1 https://chem.libretexts.org/@go/page/397284
states, Elower and E
upper , are given by the equations
γh
Elower = − B0 (19.1.1)
4π

γh
Eupper = + B0 (19.1.2)
4π

where γ is the magnetogyric ratio for the nucleus, h is Planck's constant, and B is the strength of the applied magnetic field. The 0

difference in energy, ΔE, between the two states is

γh γh γh
ΔE = Eupper − Elower = + B0 − (− B0 ) = B0 (19.1.3)
4π 4π 2π

Substituting Equation 19.1.3 into the more familiar equation ΔE = hν gives the frequency, ν , of electromagnetic radiation needed
to effect a change in spin state as
γB0
ν = (19.1.4)
2π

This is called the Larmor frequency for the nucleus. For example, if the magnet has a field strength of 11.74 Tesla, then the
frequency needed to effect a change in spin state for 1H, for which γ is 2.68 × 10  rad T s , is 8 −1 −1

8 −1 −1
(2.68 × 10 rad T s )(11.74 T)
8 −1
ν = = 5.01 × 10  s
2π

or 500 MHz, which is in the radio frequency (RF) range of the electromagnetic spectrum. This is the Larmor frequency for 1H.

Population of Spin States

The relative population of the upper spin state, N upper , and of the lower spin state, N lower , is given by the Boltzmann equation
Nupper
−ΔE/kT
=e (19.1.5)
Nlower

where k is Boltzmann's constant (1.38066 × 10 −23

 J/K ) and T is the temperature in Kelvin. Substituting in Equation 19.1.3 for
ΔE gives this ratio as

Nupper
−γhB0 /2πkT
=e (19.1.6)
Nlower

IF we place a population of 1H atoms in a magnetic field with a strength of 11.74 Tesla, the ratio
Nupper

Nlower
at 298 K is
8 −1 −34
( 2.68×10   r ad   s ) ( 6.626×10   Js ) ( 11.74  T)
Nupper −
−23 −1
=e ( 2π) ( 1.38×10   JK ) ( 298  K)
= 0.99992
Nlower

If this ratio is 1:1, then the probability of absorption and emission are equal and there is no net signal. In this case, the difference in
the populations is on the order of 8 per 100,000, or 80 per 1,000,000, or 80 ppm. The small difference in the two populations means
that NMR is less sensitive than many other spectroscopic methods.

Classical Description of NMR

To understand the classical description of an NMR experiment we draw upon Figure 19.1.2. For simplicity, let's assume that in the
population of nuclei available to us, there is an excess of just one nucleus with a spin state of +1/2. In Figure 19.1.2a, we see that
the spin of this nucleus is not perfectly aligned with the applied magnetic field, B , which is aligned with the z-axis; instead the 0

nucleus precesses around the z-axis at an angle of theta, Θ. As a result, the net magnetic moment along the z-axis, μ , is less than z

the magnetic moment, µ, of the nucleus. The precession occurs with an angular velocity, ω , of γB . 0 0

19.1.2 https://chem.libretexts.org/@go/page/397284
 Figure 19.1.2 . Illustration that shows in (a) the magnetic moment, µ, of a nucleus whose spin is aligned with the applied magnetic
field, B . Application of the magnetic field, B flips the spin of the nucleus, as shown in (b).
0 1

If we apply a source of radio frequency (RF) electromagnetic radiation along the x-axis such that its magnetic field component, B , 1

is perpendicular to B , then it will generate its own angular velocity in the xy-plane. When the angular velocity of the precessing
0

nucleus matches the angular velocity of B , absorption takes place and the spin flips, as seen in Figure 19.1.2b.
1

Relaxation
When the magnetic field B is removed, the nucleus returns to its original state, as seen in Figure 19.1.2a, a process called
1

relaxation. In the absence of relaxation, the system is saturated with equal populations of the two spin states and absorption
approaches zero. This process of relaxation has two separate mechanisms: spin-lattice relaxation and spin-spin relaxation.
In spin-lattice relaxation the nucleus in its higher energy spin state, FIgure 19.1.2b, returns to its lower energy state spin state,
Figure 19.1.2a, by transferring energy to other species present in the sample (the lattice in spin-lattice). Spin-lattice relaxation is
characterized by first-order exponential decay with a characteristic relaxation time of T that is a measure of the average time the
1

nucleus remains in its higher energy spin state. Smaller values for T result in more efficient relaxation.
1

If two nuclei of the same type, but in different spin states, are in close proximity to each other, they can trades places in which the
nucleus in the higher energy spin state gives up its energy to the nucleus in the lower energy spin state. The result is a decrease in
the average life-time of an excited state. This is called spin-spin relaxation and it is characterized by a relaxation time of T . 2

Continuous Wave NMR vs. Fourier Transform NMR

In Chapter 16 we learned that we can record an infrared spectrum by using a scanning monochromator to pass, sequentially,
different wavelengths of IR radiation through a sample, obtaining a spectrum of absorbance as a function of wavelength. We also
learned that we can obtain the same spectrum by passing all wavelengths of IR radiation through the sample at the same time using
an interferometer, and then use a Fourier transform to convert the resulting interferogram into a spectrum of absorbance as a
function of wavelength. Here we consider their equivalents for NMR spectroscopy.

Continuous Wave NMR

If we scan B while holding B constant—or scan B while holding B constant—then we can identify the Larmor frequencies
1 0 0 1

where a particular nucleus absorbs. The result is an NMR spectrum that shows the intensity of absorption as a function of the
frequency at which that absorption takes place. Because we record the spectrum by scanning through a continuum of frequencies,
the method is known as continuous wave NMR. Figure 19.1.2 provides a useful visualization for this experiment.

Fourier Transform NMR

In Fourier transform NMR, the magnetic field B is applied as a brief pulse of radio frequency (RF) electromagnetic radiation
1

centered at a frequency appropriate for the nucleus of interest and for the strength of the primary magnetic field, B . The pulse 0

typically is 1-10 µs in length and applied in the xy-plane. From the Heisenberg uncertainty principle, a short pulse of Δt results in a
broad range of frequencies as Δf = 1/Δt ; this ensures that the pulse spans a sufficient range of frequencies such that the nucleus
of interest to us will absorb energy and enter into an excited state.

19.1.3 https://chem.libretexts.org/@go/page/397284
Before we apply the pulse, the population of nuclei are aligned parallel to the applied magnetic field, B , some with a spin of +1/2
0

and others with a spin of –1/2. As we learned above, there is a slight excess of nuclei with spins of +1/2, which we can represent as
a single vector that shows their combined magnetic moments along the z-axis, μ , as shown in Figure 19.1.3a. When we apply a
z

pulse of RF electromagnetic radiation with a magnetic field strength of B , the spin states of the nuclei tip away from the z-axis by
1

an angle that depends on the nucleus's magnetogyric ratio, γ, the value of B , and the length of the pulse. If, for example, a pulse
1

of 5 µs tips the the magnetic vector by 45° (Figure 19.1.3b), then a pulse of 10 µs will tip the magnetic vector by 90° degrees
(Figure 19.1.3c), so that it now lies completely within the xy-plane.

Figure 19.1.3 . Illustration that shows (a) the net magnetic moment on the z-axis for a population of nuclei in an applied magnetic
field with a field strength of B . Application of a pulsed RF field, B , tips the magnetic moment (b) away from the z-axis. The
0 1

longer the pulse of RF, the greater the angle of tipping, as shown in (c).
At the end of the pulse, the nuclei begin to relax back to their original state. Figure 19.1.4 shows that this relaxation occurs both in
the xy-plane (spin-spin relaxation) and along the z-axis (spin-lattice relaxation). If we were to trace the path of the magnetic vector
with time, we would see that it follows a spiral-like motion as its contribution in the xy-plane decreases and its contribution along
the z-axis increases. We measure this signal—called the free induction decay, or FID—during this period of relaxation.

Figure 19.1.4 . Illustration showing the process of relaxation during an NMR experiment. The bottom row shows the net magnetic
moment aligned with the z-axis, μ , as a function of time. The top row shows the net magnetic moment in the xy-plane with the
z

figure on the top right showing the overall change in the net magnetic moment in three dimensions.
The FID for a system that consists of only one type of nucleus is the simple exponentially damped oscillating signal in Figure
19.1.5a. The Fourier transform of this simple FID gives the spectrum in Figure 19.1.5b that has a single peak. A sample with a

more than one type of nucleus yields a more complex FID pattern, such as that in Figure 19.1.5c, and a more complex spectrum,
such as the two peaks in Figure 19.1.5d. Note that, as we learned in an earlier treatment of the Fourier transform in Chapter 7, a
broader peak in the frequency domain results in a faster decay in the time domain.

19.1.4 https://chem.libretexts.org/@go/page/397284
 Figure 19.1.5 . Panels (a) and (b) show the FID signal for a sample that yields single peak and the result of taking the FT of the
FID. Panels (c) and (d) show the FID signal for a sample that consists of two peaks and the result of taking the FT of the FID.
Figure 19.1.6 shows a typical pulse sequence highlighting the total cycle time and its component parts: the pulse width, the
acquisition time during which the FID is recorded, and a recycle delay before applying the next pulse and beginning the next cycle.

Figure 19.1.1 : Example of a pulse sequence.

This page titled 19.1: Theory of Nuclear Magnetic Resonance is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

19.1.5 https://chem.libretexts.org/@go/page/397284
19.2: Environmental Effects on NMR Spectra
In the previous section we showed that there is a relationship between the Larmor frequency for a nucleus, ν , its magnetogyric
ratio, γ, and the primary applied magnetic field strength, B 0

γB0
ν = (19.2.1)
2π

and we used this equation to show that the Larmor frequency for the 1H nucleus in a magnetic field of B = 11.74 T is 500 MHz.
0

If this is the only thing that determines the frequency where absorption takes place, then all compounds that contain hydrogens will
yield a 1H NMR spectrum with a single peak at the same frequency. If all spectra are identical, then NMR provides little in the way
of useful information. The NMR spectrum for propane (CH3–CH2—CH3) in Figure 19.2.1 shows two clusters of peaks that give us
confidence in the utility of NMR. In this case, it seems likely that the cluster of peaks between 250 Hz and 300 Hz, which have a
greater total intensity, are for the six hydrogens in the two methyl groups (–CH3) and that the cluster of peaks around 400 Hz are
due to the methylene group (–CH2–). In this section we will consider why the location of a nucleus within a molecule—what we
call its environment—might affect the frequency at which it absorbs and why a particular absorption line might appear as a cluster
of individual peaks instead of as a single peak.

Figure 19.2.1 : 1H NMR spectrum for propane. The original data used to construct this spectrum is here and was obtained on a 300
MHz instrument.

The NMR Spectrum's Scale

Before we consider how a nucleus's environment affects the frequencies at which it absorbs, let's take a moment to become familiar
with the scale used to plot a NMR spectrum. The label on the x-axis of the NMR spectrum for propane in Figure 19.2.1 raises
several questions that we will answer here.

Why is the Scale Relative?

From Equation 19.2.1 we see that the frequency at which a nucleus absorbs is a function of the magnet's field strength, B . This 0

means the frequency of a peak in an NMR spectrum depends on the value of B . One complication is that instruments with
0

identical nominal values for B likely will have slightly different actual values, which leads to small variations in the frequency at
0

which a particular hydrogen absorbs on different instruments. We can overcome this problem by referencing a hydrogen's measured
frequency to a reference compound that is set to a frequency of 0. The difference between the two frequencies should be the same
on different instruments. For example, the most intense peak in the NMR spectrum for propane, Figure 19.2.1, has a frequency of
269.57 Hz when measured on an NMR with a nominal field strength of 300 MHz, which means that its frequency is 269.57 Hz
greater than the reference, which is identified as TMS.

What is TMS?
The reference compound is tetramethylsilane, TMS, which has the chemical formula of (CH3)4Si in which four methyl groups are
in a tetrahedral arrangement about the central silicon. TMS has the advantage of having all of its hydrogens in the same
environment, which yields a single peak. Its hydrogen atoms also absorbs at a low frequency that is well removed from the
frequency at which most other hydrogen atoms absorb, which makes it easy to identify its peak in the NMR spectrum.

19.2.1 https://chem.libretexts.org/@go/page/397286
How Can We Create a Universal Scale?
An additional complication with the spectrum in Figure 19.2.1 is that the frequency at which a particular hydrogen absorbs is
different when using a 60 MHz NMR than it is when using a 300 MHz NMR, a consequence of Equation 19.2.1. To create a single
scale that is independent of B we divide the peak's frequency, relative to TMS, by B , expressing both in Hz, and then report the
0 0

result on a part-per-million scale by multiplying by 106. For example, the most intense peak in the NMR spectrum for propane,
Figure 19.2.1, has a frequency of 269.57 Hz; the NMR on which the spectrum was recorded had a field strength of 300 MHz. On a
parts-per-million scale, which we identify as delta, δ , the peak appears at
269.57 Hz
6
δ = × 10 = 0.899 ppm
6
300 × 10  Hz

If we record the spectrum of propane on a 60 MHz instrument, then we expect that this peak to appear at 0.899 ppm, or a frequency
of
6
0.899 ppm × (60 × 10  Hz
ν = = 53.9 Hz
6
10

relative to TMS.
Most hydrogens have values of δ between 1 and 13. Figure 19.2.2 shows the 1H NMR for propane using a ppm scale. The right
side of the ppm scale is described as being upfield, with absorption occurring at a lower frequency, and with a smaller difference in
energy, ΔE, between the ground state and the excited state. The left side of the ppm scale is described as being downfield, with
absorption occurring at a higher frequency, and with a greater difference in energy, ΔE, between the ground state and the excited
state.

Figure 19.2.2 : 1H NMR spectrum for propane expressed in ppm. The original data used to construct this spectrum is here and was
obtained on a 300 MHz instrument. The peak for TMS is not shown.

Types of Environmental Effects

The NMR spectrum for propane in Figure 19.2.2 shows two important features: the peaks for the two types of hydrogen in propane
are shifted downfield relative to the reference and the methylene hydrogens are shifted further downfield than the methyl
hydrogens. Both groups appear as clusters of peaks instead of as single peaks. In this section we consider the source of these two
phenomena.

Chemical Shifts
In the presence of a magnetic field, the electrons in a molecule circulate, generating a secondary magnetic field, B , that usually,
e

but not always, opposes the primary applied magnetic field, B . The result is that the nucleus is partially shielded by the
appl

electrons such that the field it experiences, B , usually is smaller than the applied field and
0

B0 = Bappl − Be (19.2.2)

The greater the shielding, the smaller the value of B and the further to the right the peak appears in the NMR spectrum. For
0

example, in the NMR spectrum for propane in Figure 19.2.2 the cluster of peaks for the –CH3 hydrogens centered at 0.899 ppm
shows greater shielding than the cluster of peaks for the –CH2– hydogens that is centered at 1.337 ppm.

19.2.2 https://chem.libretexts.org/@go/page/397286
Chemical shifts are useful for determining structural information for molecules. A few examples are listed in the following table
and more extensive tables here. Note that the range of chemical shifts for the methyl and the methylene groups encompass the
values for propane in Figure 19.2.2.
Table 19.2.1 . 1
H Shifts in ppm
type of hydrogen example range of chemical shifts (ppm)

primary alkyl R−CH

3
0.7 – 1.3

secondary alkyl R−CH – R

2
1.2 – 1.6

tertiary alkyl R CH
3
1.4 – 1.8

methyl ketone R– C(=O)– CH

3
2.0 – 2.4

aromatic methyl C H – CH
6 5 3
2.4 – 2.7

alkynyl R– C≡C– H 2.5 – 3.0

alkyl halide (X = F, Cl, Br, I) R X– CH

2
2.5 – 4.0

alcohol R – C– OH
3
2.5 – 5.0

vinylic R – C=C(– R)– H

2
4.5 – 6.5

aryl C H –H
6 5
6.5 – 8.0

aldehyde R– C(=O)– H 9.7 – 10.0

carboxylic acid R– C(=O)– OH 11.0 – 12.0

Spin-Spin Coupling
Chemical shifts are the result of shielding from the magnetic field associated with a molecule's circulating electrons. The splitting
of a peak into a multiplet of peaks is the result of the shielding of one nucleus by the nuclei on adjacent atoms, and is called spin-
spin coupling. Consider the NMR for propane in Figure 19.2.2, which consists of two clusters of peaks. The six hydrogens in the
two methyl groups are sufficiently close to the two hydrogens in the methylene group that the spins of the methylene hydrogens
can affect the frequency at which the methyl hydrogens absorb. Figure 19.2.3a shows how this works. Each of the two methylene
hydrogens has a spin and those spins can both be aligned with the magnetic field, B , both be aligned against B , or two
0 0

configurations in which one is aligned with B and one is aligned against B , as seen by the arrows. When the two spins are
0 0

aligned with B , the frequency at which the methyl hydrogens absorb is shifted downfield, and when the two spins are aligned
0

against B , the frequency at which the methyl hydrogens absorb is shifted upfield; in the remaining two cases, there is no change in
0

the ferquency at with the methyl hydrogens absorb. The result, as seen in Figure 19.2.3a is a triplet of peaks in a 1:2:1 intensity
ratio.
The analysis for the effect of the six methyl hydrogens on the two methylene hydrogens is a bit more complex, but works in the
same way. Figure 19.2.3b, for example, shows that there are 15 ways to arrange the spins of the six methyl hydrogens such that two
are spin down and four are spin up. Figure 19.2.3c show the resulting NMR spectrum, which is a set of seven peaks in a
1:6:15:20:15:6:1 intensity ratio.

19.2.3 https://chem.libretexts.org/@go/page/397286
 Figure 19.2.3 . Illustration that shows the origin of spin-spin splitting: (a) The methyl hydrogens in propone split into three peaks
by the spins of the two hydrogens on the adjacent methylene carbon; (b) The 15 different ways in which the six methyl hydrogens
in propane have two nuclei that are spin up and four that are spin down; (c) The methylene hydrogens in propane are split into
seven peaks by the six methyl hydrogens.
Figure 19.2.4 provides the splitting pattern observed for nuclei with I = +1/2, such as 1H. The pattern is defined by the
coefficients of a binomial distribution—asking how many different ways you can get X outcomes in Y attempts is at the heart of a
binomial distribution—this is easy to represent using Pascal's triangle—which shows us that for six equivalent nuclei we expect to
find seven peaks with relative peak areas (or other measure of the signal) of 1:6:15:20:15:6:1. Note that the first and the last entry
in any row is 1 and that all other entries in a row, as illustrated for the third entry in the seventh row, are the sum of the two entries
in the row immediately above. The pattern also is know as the N + 1 rule as the N equivalent hydrogens will split the peak for an
adjacent hydrogen into N + 1 peaks.

Figure 19.2.4 . Pascal's triangle defines the splitting patterns in 1H NMR. The annotation shows how the values in any row of
Pascal's triangle give are derived from the previous row.
Figure 19.2.5 compares the experimental NMR for propane with its simulatd spectrum based on spin-spin splitting and the 2:6 ratio
of methylene hydrogens relative to methyl hydrogens. The overall agreement between the two spectra is pretty good. The splitting
of the individual peaks is designated by the coupling constant, J, which is shown in Figure 19.2.5 for both the experimental and the
calculated spectra. Note that the coupling constant is the same whether we are considering the effect of the methyl hydrogens on
the methylene hydrogens, or the effect of the methylene hydrogens on the methyl hydrogens. Values of the coupling constant
become smaller the greater the distance between the nuclei.

19.2.4 https://chem.libretexts.org/@go/page/397286
 Figure 19.2.5 : Experimental and calculated 1H NMR spectra for propane. The original data used to construct this spectrum is here
and was obtained on a 300 MHz instrument. The peak for TMS is not shown.
The treatment of spin-spin coupling above works well if the difference in the chemical shifts for the two nuclei is significantly
greater than the magnitude of their coupling constant. When this is not true, the splitting patterns can become much more complex
and often are difficult to interpret. There are a variety of to simplify spectra, one of which, decoupling, is outlined in Figure 19.2.6.
The original spectrum (top) shows two doublets, suggesting that we have two individual nuclei that are coupled to each other. If we
irradiate the nucleus on the right at its frequency, we can saturate its ground and excited states such that it ceases to absorb. As a
result, the nucleus on the left no longer shows evidence of spin-spin coupling to the nucleus on the right (middle) and appears as a
singlet. When we turn off the decoupler (bottom) the spin-spin coupling between the two nuclei returns more quickly than
relaxation returns the signal for the nucleus on the right.

Figure 19.2.6 . Illustration showing the effect of decoupling on an NMR spectrum. See the text for additional details.

This page titled 19.2: Environmental Effects on NMR Spectra is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

19.2.5 https://chem.libretexts.org/@go/page/397286
19.3: NMR Spectrometers
Earlier in this chapter, we noted that there are two basic experimental designs for recording a NMR spectrum. One is a continuous-
wave instrument in which the range of frequencies over which the nucleus of interest absorb is scanned linearly, exciting the
different nuclei sequentially. Most instruments, however, use pulses of RF radiation to excite all nuclei at the same time and then
use a Fourier transform to recover the signals from the individual nuclei. Our attention in this chapter is limited to instruments for
FT-NMR.

Components of Fourier Transform Spectrometers

Figure 19.3.1 includes a photograph of 400 MHz NMR and a cut-away illustration of the instrument; together, these show the key
components of a FT-NMR: a magnet that provides the applied magnetic field, B , a nucleus-dependent probe that provides the
0

radio-frequency signal that yields the magnetic field, B , and a way to insert the sample into the instrument. The NMR in the
1

photograph also is equipped with a sample changer that allows the user to load 30 or more samples that are analyzed sequentially.

Figure 19.3.1 . Photograph of a 400 MHz NMR and an illustration of the magnet's housing and the location of the sample.

Magnets
The NMR in Figure 19.3.1 is described as having a frequency, ν , of 400 MHz. The relationship between frequency and the
magnet's field strength, B , is given by the equation
0

γB0
ν = (19.3.1)
2π

where γ is the magnetogyric ratio for the nucleus. A NMR's frequency is defined in terms of a nucleus of 1H; thus, a 400 MHz
NMR has a magnet with a field strength of
6 −1
(2π) × ν (2π) × (400 × 10  s )
B0 = = = 9.4 T
8 −1 −1
γ 2.86 × 10  rad T  s

Early instruments used a permanent magnet and were limited to field strengths of 0.7, 1.4, and 2.1 T, or 30, 60, and 90 MHz. As
higher frequencies provide for greater sensitivity and resolution, modern instruments use a tightly wrapped coil of wire—typically
a niobium/tin alloy or a niobium/titanium wire—that becomes superconducting when cooled to the temperature of liquid He (4.2
K). The result is a magnetic field strength of as much as 21 T or 900 MHz for 1H NMR. The magnetic coil is held within a
reservoir of liquid He, which, itself, is held within a reservoir of liquid N2.
To be useful, the magnetic field must remain stable—that is, it must not drift—and it must be homogeneous throughout the sample.
These are accomplished by using a reference to lock the magnetic field in place and by shimming.

Locking the Magnetic Field

Samples for NMR are prepared using a solvent in which the protons are replaced with deuterium. For example, instead of using
chloroform, CHCl3, as a solvent, we use deuterated chloroform, CDCl3, where D is equivalent to 2H. This has the benefit of
providing a solvent that will not contribute to the signals in the NMR spectrum. It also has the benefit that 2H has a spin of I = 1 ,

19.3.1 https://chem.libretexts.org/@go/page/397288
and a corresponding Larmor frequency. By monitoring the frequency at which 2H absorbs, the instrument can use a feed-back loop
to maintain its value by adjusting the magnet's field strength.

Shimming
A magnetic field that is not homogeneous is like a table with four legs, one of which is just a bit shorter than the others. To balance
the table, we place a small wedge, or shim, under the shorter leg. When a magnetic field is not homogeneous, small, localized
adjustments are made to the magnetic field using a set of shimming coils arranged around the sample. Shimming can be
accomplished by the operator by monitoring the quality of the signal for a particular nucleus, however, most instruments use an
algorithm that allows the instrument to shim itself.

The Sample Inlet and the Sample Probe

The center of the instrument, which runs from the sample input at the top to the sample probe at the bottom, is open to the
laboratory environment and is at room temperature. The sample is placed in a cylindrical tube (Figure 19.3.2a), that is made from
thin-walled borosilicate glass and is 180 mm long and 5 mm in diameter. The tube is then inserted into a teflon sleeve—called a
spinner—as shown in Figure 19.3.2b, which is designed to both situate the sample at the proper depth within the sample probe, and
to spin the sample about its long axis. This spinning is used to ensure that the sample averages out any inhomogeneities in the
magnetic field not resovled by shimming.

Figure 19.3.2 . Photographs showing an NMR sample tube (left) and the spinner assembly (right).
The sample probe contains the coils needed to excite the sample and to detect the NMR signal as the excited states undergo
relaxation. Figure 19.3.3 shows two configurations for this; in both configurations, the same coil is used for both excitation and
detection. In the design on the left, which uses a permanent magnet, the applied magnetic field, B , is oriented horizontally across
0

the sample's diameter and the radio frequency electromagnetic radiation and its field, B is oriented vertically using a spiral coil. In
1

the design on the right, which is used with a superconducting magnet, the applied magnetic field, B , is oriented vertically and the
0

pulse of radio frequency electromagnetic radiation and its field, B , is oriented horizontally using a saddle coil.
1

Figure 19.3.3 . Two configurations for applying the magnetic fields B and B . The configuration on the left is used with a
0 1

permanent, such as a continuous wave instrument or a benchtop FT-NMR. The configuration on the right is used with a
superconducting magnet, such as the instrument in Figure 19.3.1 .

19.3.2 https://chem.libretexts.org/@go/page/397288
Data Processing
In Chapter 19.1 we used the following figure to describe a pulse NMR experiment. Following a pulse that is applied for 1–10 µs,
the free-induction decay, FID, is recorded for a period of time that may range from as little as 0.1 seconds to as long as 10 seconds,
depending on the nucleus being probed.

Figure 19.3.4 . An example of a pulse sequence for FT-NMR. See the text for a description of the terms shown here.
The FID is an analog signal in the form of a voltage, typically in the µV range. This analog signal must be converted into a digital
signal for data processing, which is called an analog-to-digital conversion, ADC. Two important considerations are needed here:
how to ensure that the signal—more specifically, the location of the peaks in the NMR spectrum—is not distorted, and how to
accomplish the ADC when the frequencies are on the order of hundreds of MHz.

Analog-to-Digital Conversion
An analog-to-digital converter maps the signal onto a limited number of possible values—expressed in binary notation—and are
characterized by the number of available bits. A 2-bit ADC convertor, for example, is limited to 2 = 4 possible binary values of
2

00, 01, 10, and 11 that correspond to the decimal numbers 1, 2, 3, and 4. Having only four possible values, of course, would distort
the FID pattern in Figure 19.3.4 from a smoothly varying oscillating signal into a series of steps. Using an ADC convertor with 16
bits allows for 65,536 unique digital values, a significant improvement. Another form of distortion occurs if we do not sample the
FID with sufficient frequency. Consider, for example, the simple sine wave in Figure 19.3.5a that is shown as a solid line. If we
sample this signal only five times over a period of less than four complete cycles, as shown by the five equally-spaced dots in
Figure 19.3.5a, then the apparent signal is that shown by the dadshed line.

Figure 19.3.5 : Effect of sampling frequency when monitoring a periodic signal. Individual samples are shown by the red dots (•).
In (a) the sampling frequency is approximately 1.5 samples per period. The dashed red line shows the apparent signal based on five
samples and the solid blue line shows the true signal. In (b) a sampling frequency of six samples per period accurately reproduces
the true signal.
According to the Nyquist theorem, to determine accurately the frequency of a periodic signal, we must sample the signal at least
twice during each cycle or period. Given a sampling rate of Δ, the following equation

19.3.3 https://chem.libretexts.org/@go/page/397288
 1
Δ = (19.3.2)
2νmax

defines the highest frequency, ν max , that we can monitor accurately. A sampling rate of six samples per period is more than
sufficient to reproduce the real signal in Figure 19.3.5.
A peak with a frequency that is greater than ν is not absent from the spectrum; instead, it simply appears at a different location.
max

For example, suppose we can monitor accurately any frequency within the window shown in Figure 19.3.6 and that we only
measure frequencies within this window. A peak with a frequency that is greater than what we can measure accurately by Δν
appears at an apparent frequency that is Δν greater than the frequency window's lower limit. This is called folding.

Figure 19.3.6 . Illustration showing the origin of folding in which a peak in the NMR spectrum appears at a value of δ that is
different from its actual value.

Managing MHz Signals

The instrument in Figure 19.3.1 is a 400 MHz NMR. This is a range of frequencies that is too large for an analog-to-digital
convertor to handle with accuracy. The frequency window of interest to us, however, is typically 10 ppm for 1H NMR (see Chapter
19.2 to review the NMR scale). For a 400 MHz NMR this corresponds to just 4000 Hz, with the useful range running from 400.000
MHz to 400.004 MHz. Subtracting the instrument's frequency of 400 MHz from the signal's frequency limits the latter to the range
of 0–4000 Hz, a range that is easy for an ADC to handle.

Signal Integrators
Integrating to determine the area under the peaks provides a way to gain some quantitative information about the sample. Figure
19.3.7 shows the integration of the NMR of propane first seen in Chapter 19.2. Integration of the peak for the two methyl groups

gives a result of 1766 and integration of the peak for the methylene group gives a result of 710. The ratio of the two is
1766
= 2.5
710

which is somewhat smaller than the expected 3:1 ratio.

Figure 19.3.1 . Illustration of integration. The original data used to construct this spectrum is here and was obtained on a 300 MHz
instrument.

19.3.4 https://chem.libretexts.org/@go/page/397288
This page titled 19.3: NMR Spectrometers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

19.3.5 https://chem.libretexts.org/@go/page/397288
19.4: Applications of Proton NMR
Proton (1H) NMR finds use for both qualitative analyses and quantitative analyses; in this section we briefly consider each of these
areas.

Identification of Compounds
Proton NMR is an essential tool for the qualitative analysis of organic, inorganic, and biochemical compounds. Figure 19.4.1
provides a simple example that shows the relationship between structure and a 1H NMR's peaks. The spectra in this figure are for a
set of four simple organic molecules, each of which has a chain of three carbons and an oxygen: 1-propanol, CH3CH2CH2OH, 2-
propanol, CH3CH(OH)CH3, propanal, CH3CH2CHO, and propanoic acid, CH3CH2COOH. The first two of these molecules are
alcohols, the third is an aldehyde, and the last is an acid. The main spectrum runs from 0–14 ppm, with insets showing the spectra
over a narrower range of 0–5 ppm.
Each of these molecules has a terminal –CH3 group that is the most upfield peak in its spectrum, appearing between 0.94 – 1.20
ppm. Each of these molecules has a hydrogen that either is bonded to an oxygen or a hydrogen bonded to the same carbon as the
oxygen. The hydrogens in the –OH groups of the two alcohols have similar shifts of 2.16 ppm and 2.26 ppm, but the aldehyde
hydrogen in the –CHO group and the acid hydrogen in –COOH are shifted further downfield appearing at 9.793 ppm and 11.73
ppm, respectively. The hydrogens in the two –CH2– groups of 1-propanol have very different shifts, with the one adjacent to the –
OH group appearing more downfield at 3.582 ppm than the one next to the –CH3 group at 1.57 ppm. Not surprisingly, the –CH–
hydrogen in 2-proponal, which is adjacent to the –OH group appears at 4.008 ppm.
Comparisons such of this make it possible to build tables of chemical shifts—see Table 19.2.1 in Chapter 19.2 for an example—
that can help in determining the identify of the molecule giving rise to a particular NMR spectrum. As this receives extensive
coverage in other courses, particularly courses in organic chemistry, we will not provide a more extensive coverage here.

Figure 19.4.1 . 1H NMR spectra for 1-propanol, 2-propanol, propanal, and propanoic acid. The full spectrum are shown using the
scale at the bottom of each figure. The insets show close-ups of the NMR spectra from 0 – 5 ppm. The original data used to
construct these spectra are found here. The spectrum for propanal was recorded on a 300 MHz instrument; the other three spectra
were recorded on a 90 MHz instrument.

Quantitative Analysis
A quantitative analysis requires a method of standardization, which for NMR usually makes use of an internal standard. A good
internal standard should have high purity and should have a relatively simple NMR spectrum with peaks that do not overlap with
the analyte or other species present in the sample. If we are interested in only the relative concentrations of the analyte and the
internal standard, then we can use the following formula
Ma Ia Nis
= × (19.4.1)
Mis Iis Na

where M is the molar concentration of the analyte or internal standard, I is the intensity of the NMR peak for the analyte or
internal standard, and N is the number of nuclei giving rise to the NMR peak for the analyte and the internal standard. Even if we
don't know the exact concentration of the internal standard, if we know that its concentration is the same in all samples, then we
can determine the relative concentration of analyte in a collection of samples.

19.4.1 https://chem.libretexts.org/@go/page/397290
If we are interested in determining the absolute concentration of analyte in a sample, then we must know the absolute concentration
of the internal standared; when true, then Equation 19.4.1 becomes
Ia Nis
Ma = × × Mis (19.4.2)
Iis Na

Determining the purity of an analyte, P , in a sample, we can use the equation

a

Ia Nis Ma Wis
Pa = × × × × Pis (19.4.3)
Iis Na Mis Wa

where W is the weight of the internal standard or the sample that contains our analyte.

This page titled 19.4: Applications of Proton NMR is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

19.4.2 https://chem.libretexts.org/@go/page/397290
19.5: Carbon-13 NMR
The relatively slow development of instrumentation for 13C NMR spectra is the result of its limited sensitivity compared to 1H
NMR. This difference in sensitivity is due to two key differences between the nuclei 1H and 13C: their relative abundances and their
relative magnetogyric ratios. While 1H comprises 99% of all hydrogen, 13C accounts for just 1% of all carbon. The strength of an
NMR signal also depends on the difference in energy, ΔE, between the ground state and the excited state, which is a function of
the magnetogyric ratio, γ
γB0
ΔE = hν = (19.5.1)
2π

The greater the difference in energy, the greater the difference in the population between the ground and the excited states, and the
greater the signal. The magnetogyric ratio, γ, for 1H is 4× greater than that for 13C. As a result of these two factors, 1H NMR is
approximately 6400× more senstitive than 13C. The development of magnets with higher field strengths and the capabilities of
signal averaging (see Chapter 5 on signals and noise) when using Fourier transforms to gather and analyze data, make 13C feasible.
Figure 19.5.1 shows 13C NMR spectrum for three related molecules: p-nitrophenol, o-nitrophenol, and m-nitrophenol. There are
three things to make note of from this figure. First, each spectrum consists of a set of peaks, each of which is a singlet, suggesting
that no spin-spin coupling is taking place. Second, the number of peaks in each spectrum is the same as the number of unique types
of carbon—four unique carbons for p-nitrophenol and six each for m-nitrophenol and o-nitrophenol—which suggests that chemical
shifts in 13C provide useful information about the environment of the carbon atoms and, therefore, the molecule's structure. And,
third, unlike 1H, there is no relationship between the intensity of a 13C peak and the number of carbon atoms. This is particularly
evident when comparing the intensity of the peaks for the carbon bonded to the –NO2 group and the carbon bonded to the –OH
group, which are significantly less intense than the peaks for other carbons. We will consider each of these observations in the
remainder of this section.

Figure 19.5.1 . 13C NMR spectra for three nitrophenols. The original data used to construct these spectra are found here. The
spectra were recorded on a 15 MHz instrument (with respect to 13C, or 60 MHz with respect to 1H).

Proton Decoupling
In 13C NMR there is no coupling between adjacent carbon atoms because it is unlikely that both are 13C, the only isotope of carbon
that is NMR active (the odds that two adjacent carbons are both 13C is 0.01 × 0.01, or 0.0001 or 0.010%). Coupling does take
place between 13C and 1H when the hydrogen atoms are attached to the carbon atom. Such coupling follows the same N+1 rule as
in 1H NMR; thus, a quartenary carbon (R4C) appears as a singlet, a methine carbon (R3CH) appears as a doublet, a methylene
carbon (R2CH2) appears as a triplet, and a methyl carbon (RCH3) appears as a quartet. Even with the extensive range of ppm values
over which 13C peaks appear—chemical shifts for 13C spectra run from 250 – 0 ppm instead of 14 – 0 ppm for 1H spectra—a
compound with many different types of carbon atoms, each with 1 – 3 hydrogen atoms results in a complex spectrum. For this
reason, 13C NMR spectra are acquired in a way that prevents coupling between 13C and 1H. This is called proton decoupling.
The most common method of proton decoupling is to use a second RF generator to irradiate the sample with a broad-band of RF
signals that spans the range of frequencies for the protons. As described earlier in Section 19.3, the effect is to saturate the proton's
ground and excited states, which prevents the protons from absorbing energy and from coupling with each other and with the
carbons atoms. The 13C spectra in Figure 19.5.1 are examples of decoupled spectra.

Qualitative Applications of 13C NMR

Just as with 1H NMR spectra, tables of chemical shifts for 13C peaks aids in determining a molecules structure. Table 19.5.1

provides ranges of chemical shifts for different types of carbon atom. A set of tables is available here.

19.5.1 https://chem.libretexts.org/@go/page/397292
 Table 19.5.1 . 13C Shifts in ppm
type of carbon atom example range of chemical shifts (ppm)

primary alkyl R−CH

3
10 – 30

secondary alkyl R−CH – R

2
15 – 55

tertiary alkyl R CH
3
20 – 60

quaternary alkyl R C
4
30 – 40

alkynl R– C≡C– H 65 – 90

alkenyl R– C=C– H 100 – 150

aromatic C H
6 6
110 – 170

ester R−C(=O)−O−R 165 – 175

amide R−C(=O)−N−R
2
165 – 175

carboxylic acid R– C(=O)– OH 175 – 185

aldehyde R– C(=O)– H 190 – 220

ketone R−C(=O)−R 205 – 220

attached to iodine C−I 0 – 40

attached to bromine C−Br 25 – 65

attached to chlorine C−Cl 35 – 80

attached to oxygen C−O 40 – 80

attached to nitrogen C−N 40 – 60

Nuclear Overhauser Enhancement

The most intense peak in the 13C NMR spectrum for m-nitrophenol (see Figure 19.5.1 above) is the carbon in the benzene ring
labeled as position 4, with an intensity of 1000 (the intensity scale is normalized here to a maximum value of 1000, but for this
section, let's take it as an absolute value). Because the spectrum was acquired with proton decoupling turned on, the peak for this
carbon appears as a singlet. If we turn proton decoupling off, then we expect the peak to appear as a doublet as this carbon has one
hydrogen attached to it. We might reasonably expect to find that each peak has an intensity of 500, giving a total intensity of 1000.
The actual intensities of the peaks for this carbon, however, are smaller than expected. Put another way, when we turn proton
decoupling on, the intensity of a 13C line increases more than expected and the more hydrogens, the greater the effect. This is called
nuclear overhauser enhancement (NOE).
NOE is the result of the relative populations of the ground and excited states. The technical details are more than we will consider
here, but the extent of the total enhancement of the peak intensities is proportional to the ratio of the magnetogyric ratios of the
irradiated nucleus (1H) and the observed nucleus (13C), which for a 1H decoupled 13C NMR results in a total enhancement of the
intensity of approximately 200%. As magnetogyric ratios can be negative, as is the case for 15N, a decoupled spectrum can result in
less intense peaks. One important consequence of NOE, is that integrated peak areas are not proportional to the number of identical
carbon atoms, which is a loss of information.

 Note

Although our focus in this chapter is on 1H and 13C NMR, other nuclei, such as 31P, 19F, and 15N are useful for the study of
chemically and biochemically important molecules.

This page titled 19.5: Carbon-13 NMR is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

19.5.2 https://chem.libretexts.org/@go/page/397292
19.6: Two-Dimensional Fourier Transform NMR
The 1H and 13C spectra up to this point are shown in one dimension (1D), which is the frequency absorbed by the analyte's nuclei
expressed in ppm. These spectra were acquired by applying a brief RF pulse to the sample, recording the resulting FID, and then
using a Fourier transform to obtain the NMR spectrum. In addition to 1D experiments, there are a host of 2D experiments in which
we apply a sequence of two or more pulses, recording the resulting FID after applying the last pulse. In this section we will
consider one example of a 2D NMR experiment in some detail: 1H – 1H correlation spectroscopy, or 1H – 1H COSY.
Figure 19.6.1 shows the basic experimental details for the COSY experiment. The pulse train is shown in (a) and consists of a first
pulse that is followed by a delay time, t , in which the nuclear spins are allowed to precess and relax. This is followed by the
1

application of a second pulse and the measurement of the resulting FID during the acquisition period, t , that consists of n
2 t2

individual data points. The COSY experiment consists of a sequence of n individual pulse trains, each with a different t . The
t1 1

result (b) is a matrix with n rows and n columns. To process the data, each row in this matrix is Fourier transformed, the
t1 t2

resulting n × n matrix is transposed to a n × n matrix and then Fourier transformed again to give (c) a n × n matrix that
t1 t2 t2 t1 t2 t2

shows the intensity of the signal for all possible combinations of applied frequencies.

Figure 19.6.1 . Details of a COSY experiment: (a) the pulse sequence; (b) the data matrix; and (c) the COSY spectrum. See the text
for additional details.
Figure 19.6.2 shows the 1H – 1H COSY spectrum for ethyl acetate. Instead of just annotating the two axes with numerical values of
the frequencies, they are displayed by superimposing the 1D 1H NMR spectrum for ethyl acetate in ppm. The points that fall on the
diagonal line are just the three frequencies where ethyl acetate has 1H peaks. Of more interest are the points that fall on either side
of the diagonal line—these are called cross peaks—as these show pairs of hydrogens that are coupled (or correlated) to each other.
The cross peaks in the COSY spectrum are symmetrical about the diagonal line and show the correlation between the hydrogens on
the methyl carbon and and the methylene carbon that are adjacent to each other (the ethyl part of ethyl acetate). The remaining
methyl group is not coupled to the other hydrogen's in the ethyl acetate, so there is no cross peak at the intersection of
δ = 2.038 ppm and δ = 1.260 ppm . The information about coupling from cross peaks assists in interpreting complex H NMR
1

spectra.

19.6.1 https://chem.libretexts.org/@go/page/397296
 Figure 19.6.2 . The COSY spectrum for ethyl acetate. The original data used to create the two 1H NMR spectra is available here
and was obtained using a 300 MHz NMR. The COSY spectrum is simulated.
COSY is one example of a homonuclear 2D NMR experiment because it examines coupling between identical nuclei, such as 1H in
Figure 19.6.2. There are many other 2D NMR experiments, each of which uses a sequence of pulses—some that use two pulses, as
in COSY, and others that use three or more pulses—and a data analysis algorithm. Table 19.6.1 provides details on some of these
methods.
Table 19.6.1 . Selected examples of 2D NMR experiments.
method information obtained from cross peaks

coupling between two protons (1H – 1H) that are within three chemical
correlation spectroscopy (COSY)
bonds of each other

total correlation spectroscopy (TOCSY) coupling between all protons (1H) in the molecule

coupling between a proton (1H) and another nucleus, such as carbon (1H
heteronuclear correlation spectroscopy (HETCOR)
– 13C) or nitrogen (1H – 15N)
coupling between two protons (1H – 1H) that are within approximately 5
nuclear overhauser and exchange spectroscopy (NOSEY)
Å of each other
coupling between a proton (1H) and another nucleus, such as carbon (1H
heteronuclear single quantum correlation (HSQC)
– 13C) or nitrogen (1H – 15N)
coupling between a proton and a carbon (1H – 13C) that are two or three
heteronuclear multiple bond coherence spectroscopy (HMBC)
bonds apart

incredible natural abundance double-quantum transfer

coupling between adjacent carbon atoms (13C – 13C)
(INADEQUATE)

double quantum filtered correlation spectroscopy (DQF–COSY) suppresses signals from water

This page titled 19.6: Two-Dimensional Fourier Transform NMR is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

19.6.2 https://chem.libretexts.org/@go/page/397296
 CHAPTER OVERVIEW
20: Molecular Mass Spectrometry
20.1: Molecular Mass Spectra
20.2: Ion Sources
20.3: Mass Spectrometers
20.4: Applications of Molecular Mass Spectrometry

This page titled 20: Molecular Mass Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1
20.1: Molecular Mass Spectra
Figure 20.1.1 shows a mass spectrum for p-nitrophenol, C6H5NO3, which has a nominal (integer) mass of 139 daltons. If we send a
beam of energetic electrons through a gas phase sample of p-nitrophenol, it loses an electron, which we write as the reaction
− +∙ −
C H NO +e → C H NO3 + 2e (20.1.1)
6 5 3 6 5

where the product is a radical cation that has a charge of +1 and that retains the nominal mass of 139 daltons. We call this the
molecular ion—highlighted here in green—and it has a mass-to-charge ratio (m/z) of 139.

 Note
Some of the terminology in this chapter was covered earlier in Chapter 11 on atomic mass spectrometry. See the first section of
that chapter for a discussion of atomic mass units (amu) and daltons (Da), and of mass-to-charge ratios.

If reaction 20.1.1 is all that happens when p-nitrophenol interacts with an energetic electron, then it would not provide much in the
way of useful information. The radical cation C H NO , however, retains sufficient excess energy from the initial electron-
6 5
+∙
3

molecule collision that it is in an excited state. In returning to its ground state, the molecular ion undergoes a series of
fragmentations that result in the formation of ions—called daughter ions—with different mass-to-charge ratios. A plot that shows
the relative intensity of these ions as a function of their mass-to-charge ratios is called a mass spectrum. The most abundant
fragment in the spectrum—shown here in red and which is called the base peak—is assigned a relative intensity of 100; the
intensity of all other ions is reported relative to the base perk.

Figure 20.1.1 : The mass spectrum for p-nitrophenol showing the molecular ion at a m/z of 139 (in green) and the daughter ions (in
blue) formed through fragmentation. The most abundant fragment is the base peak at a m/z of 65 (in red). The original data is
available here. This particular spectrum was collected using an electron impact source; more on that in Section 20.2.
A molecule's fragmentation patterns provides rich information about its structure. Figure 20.1.2 compares the mass spectra for o-
nitrophenol, m-nitrophenol, and p-nitrophenol. All three molecules have clusters of fragment ions at similar mass-to-charge ratios,
but the relative abundance of the ions in these clusters varies quite a bit from molecule-to-molecule. For example, the pink
rectangle in each spectrum highlights peaks with mass-to-charge ratios from approximately 104 m/z to 115 m/z. All three molecules
share the property of producing fragment ions with these mass-to-charge ratios; the relative abundance of the fragment ions,
however, varies substantially between the three molecules with o-nitrophenol and p-nitrophenol having a major peak at 109 m/z,
but m-nitrophenol showing no more than a trace peak at 109 m/z. We will return to the use of mass spectrometry for determining
structure information later in this chapter.

20.1.1 https://chem.libretexts.org/@go/page/391117
 Figure 20.1.2 : The mass spectra for o-nitrophenol (top), m-nitrophenol (middle), and p-nitrophenol (bottom). All three molecules
show the same pattern of the mass-to-charge ratios where clusters of fragment ions appear; however, the relative abundance of each
cluster varies from molecule-to-molecule. The pink rectangle highlights the m/z values of 104–115. The original data for these
spectra are here, here, and here from top-to-bottom.

This page titled 20.1: Molecular Mass Spectra is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

20.1.2 https://chem.libretexts.org/@go/page/391117
20.2: Ion Sources
Since a mass spectrum shows the relative abundance of ions with different mass-to-charge ratios, a mass spectrometer must include
a way to generate ions. More specifically, it needs a method that generates the initial ion as it, once formed, will undergo
fragmentation without additional help from the analyst (which does not mean the analyst cannot assist in that fragmentation; see
discussion of tandem mass spectrometry in Section 20.4). In this section we consider several common ion sources. We can describe
these sources using two characteristic properties: (a) the physical state of the species that is initially ionized (gas, liquid, or solid
phase), and (b) whether ionization favors the formation of fragment ions or the formation of molecular ions (hard sources or soft
sources).

Electron Ionization Sources (gas phase/hard source)

The electron ionization (EI) source, also known as an electron impact source, uses a beam of energetic electrons to ionize the
analyte. As shown in 20.2.1, the sample is volatilized prior to entering the ion source as gas phase molecules, M(g). A heated
tungsten filament is used to generate electrons, which are pulled toward a positively charged anode. This electron beam intersects
with the gas phase molecules at 90° where ionization occurs
− +∙ −
M(g) + e → M (g) + 2 e (20.2.1)

The molecular ions, M +∙

(g) , are then swept into the mass analyzer using a set of accelerating plates (not shown here).

Figure 20.2.1 : Illustration of an electron ionization source. The sample is volatilized and brought into the ion source where it
intersects with an electron beam, forming the molecular ion, M , which is carried on into the mass analyzer.
+∙

The electron beam in Figure 20.2.1 has a lot of energy due to the significant potential difference between the cathode and the
anode, which may be as much as 70 V. The kinetic energy of these electrons is equivalent to the product of the electron's charge in
Coulombs, the applied potential in volts, and Avogadro's number
−19 −23 6
e × V × NA = 1.6 × 10 × 6.022 × 10 = 6.7 × 10  J/mol

or 6,700 kJ/mol. This energy is much greater than typical bond energies, which range from approximately 150–600 kJ/mol for
single bonds, from approximately 500–750 kJ/mol for double bonds, and from approximately 800–1100 kJ/mol for triple bonds.
The significant difference between the energy of the electrons and bond energies explains why electron ionization spectra are rich
in fragment ions, as we saw earlier in Section 20.1 for o-nitrophenol, m-nitrophenol, and p-nitrophenol. This extensive
fragmentation is useful in determining an analyte's structure—which is an advantage of a hard ionization method—but at the
possible cost of the loss of the molecular ion peak for some analytes. For example, Figure 20.2.2 shows the electron ionization
mass spectrum from 1-decanol, C10H22O, which has a nominal mass of 158 daltons. The small peak at m/z = 157 is for the
fragment ion C10H21O+; the molecular ion is not observed in this spectrum.

20.2.1 https://chem.libretexts.org/@go/page/391119
 Figure 20.2.2 : Mass spectrum for 1-decanol, C10H22O, using electron ionization. The peak furthest to the right (m/z = 157 amu) is
one atomic mass unit less than the expected peak for the molecular ion, which does not appear in the spectrum. The original data is
available here.

Chemical Ionization Sources (gas phase/soft source)

Electron ionization is a hard source because the electron beam's energy results in easy fragmentation. In chemical ionization, we
introduced a reagent molecule, such as methane, into the electron ionization (CI) source so that it is present at level that is 1000×
to 10, 000× greater than the analyte. At this higher concentration, it is the reagent molecule that is ionized; for example, when
using CH4 as the reagent gas, ions such as CH and CH form. These ions then react with additional methane molecules
+
4
+
3

+ +
CH (g) + CH (g) → CH (g) + CH (g) (20.2.2)
4 4 5 3

+ +
CH (g) + CH (g) → C H (g) + H (g) (20.2.3)
3 4 2 5 2

to form CH and C
+
5 2
+
H
5
, species that are sufficiently reactive that they easily transfer a hydrogen to a molecule of the analyte, MH
+ +
CH (g) + MH(g) → MH (g) + CH (g) (20.2.4)
5 2 4

to give a molecular ion that we identify as [M + H]+ and that has a mass that is one amu unit greater than that for M. Alternatively,
they can easily remove a hydrogen from a molecule of the analyte, MH
+ +
C H (g) + MH(g) → M (g) + C H (g) (20.2.5)
2 5 2 6

to give a molecular ion that we identify as [M – H]– and that has a mass that is one amu less than that for M. Because formation of
the molecular ion occurs indirectly and less energetically, fragmentation is suppressed, leading to a mass spectrum with a molecular
ion peak and with only a small number of other ions. Figure 20.2.2 shows the mass spectrum for 1-decanol when using chemical
ionization with CH as the reagent gas.
4

Figure 20.2.3 : Mass spectrum for 1-decanol, C10H22O, using chemical ionization. The peak highlighted in red is the molecular ion,
which has a m/z ratio of 158 − 1 = 157 amu . The original data is available here.

20.2.2 https://chem.libretexts.org/@go/page/391119
Electrospray Ionization Sources (liquid phase/soft source)
Electron impact and chemical ionization are gas phase sources because the sample is volatilized before it enters the mass
spectrometer's inlet. In electrospray ionization (ESI), the sample is a liquid and ions desorb from that matrix in the mass
spectrometer's inlet system. The liquid sample is pulled into the spectrometer's inlet through a capillary needle, forming a mist of
droplets. The application of a large potential across this inlet assures that the droplets carry positive charges. These charged
droplets then enter a chamber where they undergo desolvation, which decreases the size of the droplet and increases their charge
density (see Figure 20.2.4). As this charge density increases, the droplets eventually become unstable, for reasons that are not fully
understood, and ionized gas-phase ions desorb from the droplets and enter into the mass analyzer.

Figure 20.2.4 : Illustration showing the change in the size of droplets and the charge density of droplets during electrospray
ionization.
A typical electrospray ionization mass spectrum for a small molecule is shown in Figure 20.2.5 for the compound (4-
aminophenyl)arsonic acid. As we saw in Figure 20.2.3 for chemical ionization, a soft ionization source results in a limited amount
of fragmentation and a strong peak for the molecular ion, which here includes a proton transfer to give the [M + H]+ peak at a m/z
of 218 amu.

Figure 20.2.5 : Mass spectrum for (4-aminophenyl)arsonic acid obtained using electrospray ionization. The spectrum contains only
two majors peak: the molecular ion at m/z = 218 amu and a fragment at m/z = 109 amu. The original data is available here.
Electrospray ionization is particularly useful for biological molecules, such as peptides and proteins, because the soft ionization
ensures that molecular weight information is retained. Because these molecules are large, they readily pick up multiple protons,
forming multiply charged ions of the general form [M + zH]z+ where z is the number of protons added. Figure 20.2.6 shows a
hypothetical spectrum for the molecule M and Table 20.2.1 provides the corresponding m/z values for the mass spectrum's peaks.

20.2.3 https://chem.libretexts.org/@go/page/391119
 Figure 20.2.6 : Simulated electrospray ionization mass spectrum for a molecule, M, that shows eight peaks with mass-to-charge
ratios between 600 amu and 1200 amu. The subscripts in the symbols M serve to identify the peaks; they are not an indication of
i

charge.
Table 20.2.1 . Mass-to-charge ratios for the peaks in Figure 20.2.6 .
peak m/z

M1 1112

M2 1001

M3 910

M4 834

M5 770

M6 715

M7 667

M8 626

If we take the mass-to-charge ratios for any two adjacent peaks, M and M , where the peak M has the greater value for m/z, and
i j i

if we assume that M has one additional hydrogen atom, giving it a charge that is one unit higher, then
j

Mj − 1
Zi = (20.2.6)
Mi − Mj

where Z is the charge on the ion M . Table 20.2.2 shows the calculated charges for the ions M to M .
i i 1 7

 Note

Here is a derivation for Equation 20.2.6. Suppose the molecule of interest has a molecular weight of m. If the charge on the
ion responsible for peak M is Z , then it must be the case that the peak's mass is equal to m + Z as it has Z extra hydrogens and
i

it must be the case that its mass-to-charge ratio is

m +Z
Mi =
Z

and its molecular weight, m, is

m = (Mi × Z) − Z

In the same way, the peak M has a charge of Z + 1 and

j

m +Z +1
Mj =
Z +1

m = (Mj × Z) + Mj − Z − 1

Setting the two equations for m equal to each other and solving for Z gives

20.2.4 https://chem.libretexts.org/@go/page/391119
 (Mi × Z) − Z = (Mj × Z) + Mj − Z − 1

(Mi × Z) = (Mj × Z) + Mj − 1

(Mi × Z) − (Mj × Z) = Mj − 1

Mj − 1
Z =
Mi − Mj

Table 20.2.2 . Calculated charges on the ions for peaks M to M

1 7

peak m/z Z

M1 1112 9

M2 1001 10

M3 910 11

M4 834 12

M5 770 13

M6 715 14

M7 667 15

M8 626 —

The molecular weight, m, is given by the equation

m = (Mi × Z) − Z (20.2.7)

Table 20.2.3 shows the molecular weights for the ions M to M and their average value. The simulated mass spectrum was
1 7

created by setting the molecular weight to 10,000 amu and with charges ranging from +9 to +16.
Table 20.2.3 . Calculated molecular weights for the ions M to M .
1 7

peak m/z Z m

M1 1112 9 9,999

M2 1001 10 10,000

M3 910 11 9,999

M4 834 12 9,996

M5 770 13 9,997

M6 715 14 9,996

M7 667 15 10,005

M8 626 — —

average molecular weight 9,999

Matrix-Assisted Laser Desorption/Ionization Sources (solid phase/soft source)

Matrix-assisted laser desorption ionization (MALDI) is a soft ionization source for obtaining the mass spectrum for biologically
important molecules, such as proteins and peptides. Figure 20.2.7 illustrates the basic steps in obtaining a MALDI spectrum. The
sample is first mixed with a small molecule—which is called the matrix—to form a solution; the matrix usually is present in a 10:1
ratio. A drop of this mixture is placed on a sample probe and allowed to dry, leaving the sample in a solid form. A pulsed laser
beam (λ = 237 nm is typical) is focused on the solid sample–matrix mixture. The matrix absorbs the laser pulse and the absorption
of the laser's energy volatilizes both the matrix and the sample. Ionization of the sample forms molecular ions, usually [M + H]+
ions, which are then swept into the mass analyzer.

20.2.5 https://chem.libretexts.org/@go/page/391119
 Figure 20.2.7 : Illustration of matrix-assisted laser desorption ionization. The sample is dispersed in a liquid matrix and then
allowed to dry. A pulsed laser beam is used to vaporize and ionize the sample.
When the sample is a digestion of a protein, then it is a mixture of peptides, each of which appeara as a [M + H]+ peak in the
resulting mass spectrum. For example, a peptide with the sequence AWSVAR (alanine–tryptophan–serine–valine–alanine–arginine)
will appear as a peak with a mass of 689.8 daltons. To find this value we add together the molecular weights of the amino acids,
account for the loss of a molecule of water for each peptide bond that forms, and then account for the hydrogen that gives the [M +
H]+ ion. In this case we have
+
[M + H] = 89.1 + 204.2 + 105.1 + 117.1 + 89.1 + 174.2 − (5 × 18.0) + 1 = 689.8 amu

where the term 5 × 18.0 accounts for the loss of five molecules of H 2
O when forming the five peptide bonds.

Fast Atom Bombardment Sources (liquid phase/soft source)

Fast atom bombardment (FAB) bears some similarity to MALDI: the sample is mixed with a liquid matrix (often glycerol) and
bombarded with a beam of xenon or argon atoms (instead of a laser). Desorption of the sample from its matrix forms gas phase ions
that are swept into the mass analyzer. Spectra usually contain both a molecular ion and fragmentation patterns.

This page titled 20.2: Ion Sources is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

20.2.6 https://chem.libretexts.org/@go/page/391119
20.3: Mass Spectrometers
A mass spectrometer has four essential elements: a means for introducing the sample to the instrument, a means for generating a
mixture of ions, a means for separating the ions, and a means for counting the ions. In Chapter 20.2 we introduced some of the
most important ways to generate ions. In this section we turn our attention to sample inlet systems and to separating and counting
ions. You may wish to review Chapter 11 where we considered these topics in the context of atomic mass spectrometry. As we
noted in Chapter 11, a mass spectrometer must operate under a vacuum to ensure that ions can travel long distances without
undergoing undesired collisions that affect their ion energy.

Sample Inlet Systems

When the sample is a gas or a volatile liquid, it is easy to transfer a portion of the sample into a reservoir as a gas maintained at a
relatively small pressure. The sample is then allowed to enter into the mass spectrometer's ion source through a diaphragm that
contains a pin-hole, drawn in by holding the ion source at lower pressure.
Solid and non-volatile liquids are sampled by inserting them directly into the ion source through a vacuum lock that allows the
mass spectrometer to remain under vacuum except for the ion source where the sample is inserted. The sample is placed in a
capillary tube or a small cup at the end of a sample probe, and then moved into the ion source. The sample probe includes a heating
coil that is used, along with the instrument's vacuum, to help volatilize the sample.
Of particular importance are inlet systems that couple a chromatographic or an electrophoretic instrument to a mass spectrometer,
providing a way to separate a complex mixture into its individual components and then using the mass spectrometer to determine
the composition of those components. The interface between a gas chromatograph and a mass spectrometer (GC-MS) must account
for the significant drop in pressure from atmospheric pressure to a pressure of 10–8 torr; the interface for LC-MS and for EC-MS
must provide a way to remove the liquid eluent, to volatilize the samples, and to account for the drop in pressure. See Chapters 27,
28, and 30 for more details.

Mass Analyzers
The purpose of the mass analyzer is to separate the ions by their mass-to-charge ratios. Ideally we want the mass analyzer to allow
us to distinguish between small differences in mass and to do so with a strong signal-to-noise ratio. As we learned in Chapter 7
when introducing optical spectroscopy, these two desires usually are in tension with each other, with improvements in resolution
often coming with an increase in noise.

Resolution
The resolution between two peaks, R , in mass spectrometry is defined as the ratio of their average mass to the difference in their
masses
¯
¯¯¯
¯
m
R = (20.3.1)
Δm

The following table shows how resolution varies as a function of the average mass and the difference in mass. A resolution of
1,000, for example is sufficient to resolve two ions with an average mass of 100 amu that differ by 0.1 amu, or two ions that have
an average mass of 1,000 amu that differ by 1 amu.
Table 20.3.1 . Resolution for stated values of m and Δm.
¯
¯¯¯
¯

¯
¯¯¯
¯
m →
100 amu 1000 amu 10,000 amu
Δm ↓

0.1 amu 1,000 10,000 100,000

1 amu 100 1,000 10,000

10 amu 10 100 1,000

Magnetic Sector Mass Analyzers

When a beam of ions passes through a magnetic field, its path is altered, as we see in Figure 20.3.1. The ions experience an
acceleration as they exit the ion source and enter the mass analyzer with a kinetic energy that is given by the equations

20.3.1 https://chem.libretexts.org/@go/page/391121
 KE = zeV (20.3.2)

1 2
KE = mv (20.3.3)
2

where z is the ion's charge (usually +1), e is the electronic charge in Coulombs, V is the applied voltage responsible for the
acceleration, m is the ion's mass, and v is the ion's velocity after acceleration. Equation 20.3.2 shows us that all ions with the same
charge have the same kinetic energy. Equation 20.3.3, then, tells us that ions with a greater mass will move more slowly.

Figure 20.3.1 : Illustration of a magnetic sector analyzer for mass spectrometry. Ions from the ion source enter into the mass
analyzer. The magnets in the mass analyzer bend the path followed by the ions. For any combination of magnetic field strength and
accelerating voltage, only ions with a compatible mass-to-charge ratio are able to leave the mass analyzer through the exit slit; ions
that are too heavy or too light, fail to exit the mass analyzer.
An ion's path through the magnetic field is determined by two forces. The first of these forces is the magnetic force, F , that acts M

on the ion, which is

FM = Bzev (20.3.4)

where B is the magnetic field strength. The second of these forces is the centripetal force, FC , that acts on the ion as it moves
along its curved path, which is
2
mv
FC = (20.3.5)
r

where r is the magnet's radius of curvature. An ion can only navigate these opposing forces if F M and F are equal to each other.
C

This requires that

2
mv
Bzev = (20.3.6)
r

Solving for v gives

Bzer
v= (20.3.7)
m

Substituting back into Equation 20.3.3 and solving for the mass-to-charge ratio gives
2 2
m B r e
= (20.3.8)
z 2V

Equation 20.3.8 tells us that for any combinaton of magnetic field strength, B , and accelerating voltage, V , only one mass-to-
charge ratio has the correct value of r to reach the director. Ions that are too heavy or ions that are too light, will collide with the
sides of the mass analyzer before they reach the detector. The mass spectrum is recorded by holding V and r constant and varying
the magnetic field strength, B . The resolution of a magnetic sector instrument is usually less than 2000.

20.3.2 https://chem.libretexts.org/@go/page/391121
Double-Focusing Mass Analyzers
The resolution of a magnetic sector instrument suffers from limitations that affect its ability to narrow the range of kinetic energies
—and, thus, velocities—possessed by the ions when they exit the ion source and enter the mass analyzer. The double-focusing
mass analyzer in Figure 20.3.2 compensates for this by placing an electrostatic analyzer before the magnetic analyzer, separating
the two by a slit. The electrostatic analyzer consists of two curved metal plates, one of which is held at a positive potential and one
at a negative potential. As ions pass betweeen the plates, those ions that have too much energy and those that have too little energy
fail to pass through the slit that separates the electrostatic analyzer from the magnetic analyzer. In this way, the distribution of
energies—and, thus, velocities—is tightened, improving the resolution achieved by the magnetic sector analyzer. Depending on its
design, a double-focusing analyzer can achieve a resolution as large as 100,000.

Figure 20.3.2 : Illustration of a double-focusing mass analyzer that consists of an electrostatic mass analyzer and a magnetic mass
analyzer. Ions that make it through the slit that separates the two analyzers have a smaller range of kinetic energies than the ions in
a magnetic sector mass analyzer, which improves resolution.

Quadrupole Mass Analyzers

The quadrupole mass analyzer was introduced in Chapter 11 and the treatment here is largely the same. A quadupole mass analyzer
is compact in size, low in cost, easy to use, and easy to maintain. As shown in Figure 20.3.3, it consists of four cylindrical rods,
two of which are connected to the positive terminal of a variable direct current (dc) power supply and two of which are connected
to the power supply's negative terminal; the two positive rods are positioned opposite of each other and the two negative rods are
positioned opposite of each other. Each pair of rods is also connected to a variable alternating current (ac) source operated such that
the alternating currents are 180° out-of-phase with each other. An ion beam from the source is drawn into the channel between the
quadrupoles and, depending on the applied dc and ac voltages, ions with only one mass-to-charge ratio successfully travel the
length of the mass analyzer and reach the transducer; all other ions collide with one of the four rods and are destroyed.

Figure 20.3.3 : Basic arrangement of the quadrupole mass analyzer. The plus and the minus signs on each rod indicate which are
connected to the positive terminal of the dc power supply. Not shown here is the circuitry for applying the alternating current. The
ion beam enters the channel between the quadrupoles. Depending on the applied dc and ac voltages, some of the ions emerge from
this channel and reach the transducer.
To understand how a quadrupole mass analyzer achieves this separation of ions, it helps to consider the movement of an ion
relative to just two of the four rods, as shown in Figure 20.3.4 for the poles that carry a positive dc voltage. When the ion beam
enters the channel between the rods, the ac voltage causes the ion to begin to oscillate. If, as in the top diagram, the ion is able to
maintain a stable oscillation, it will pass through the mass analyzer and reach the transducer. If, as in the middle diagram, the ion is
unable to maintain a stable oscillation, then the ion eventually collides with one of the rods and is destroyed. When the rods have a
positive dc voltage, as they do here, ions with larger mass-to-charge ratios will be slow to respond to the alternating ac voltage and
will pass through the transducer. The result is shown in the figure at the bottom (and repeated in Figure 20.3.5a) where we see that

20.3.3 https://chem.libretexts.org/@go/page/391121
ions with a sufficiently large mass-to-charge ratios successfully pass through the transducer; ions with smaller mass-to-charge
ratios do not. In this case, the quadrupole mass analyzer acts as a high-pass filter.

Figure 20.3.4 : Illustration of how a quadrupole mass analyzer achieves separation of ions by their mass-to-charge ratio. See the
text for additional details.
We can extend this to the behavior of the ions when they interact with rods that carry a negative dc voltage. In this case, the ions
are attracted to the rods, but those ions that have a sufficiently small mass-to-charge ratio are able to respond to the alternating
current's voltage and remain in the channel between the rods. The ions with larger mass-to-charge ratios move more sluggishly and
eventually collide with one of the rods. As shown in Figure 20.3.5b, in this case, the quadrupole mass analyzer acts as a low-pass
filter. Together, as we see in Figure 20.3.5c, a quadrupole mass analyzer operates as both a high-pass and a low-pass filter, allowing
a narrow band of mass-to-charge ratios to pass through the transducer. By varying the applied dc voltage and the applied ac
voltage, we can obtain a full mass spectrum.

Figure 20.3.5 : Illustration that shows how a quadrupole mass analyzer acts as both (a) a high-pass filter and (b) a low pass filter,
with the result that it passes only ions with a narrow range of mass-to-charge ratios.
Quadrupole mass analyzers provide a modest mass-to-charge resolution of about 1 amu and extend to m/z ratios of approximately
2000.

Time-Of-Flight Mass Analyzers

In a time-of-flight mass analyzers, Figure 20.3.6, ions are created in small clusters by applying a periodic pulse of energy to the
sample using a laser beam or a beam of energetic particles to ionize the sample. The small cluster of ions are then drawn into a tube
by applying an electric field and then allowed to drift through the tube in the absence of any additional applied field; the tube, for
obvious reasons, is called a drift tube. All of the ions in the cluster enter the drift tube with the same kinetic energy, KE, which we
define as
1 2
KE = mv = zeV (20.3.9)
2

The time, T , that it takes the ion to travel the distance, L, to the detector is

20.3.4 https://chem.libretexts.org/@go/page/391121
 L
T = (20.3.10)
v

Substituting Equation 20.3.10 into Equation 20.3.9

−
−− −−− −
m 1
T =√ ×√ (20.3.11)
z 2eV

shows us that the time it takes an ion to travel through the drift tube is proportional to the square rate of its mass-to-charge ratio. As
a result, lighter ions move more quickly than heavier ions. Flight times are typically less than 30 µs. A time-of-flight mass analyzer
provide better resolution than a quadrupole mass analyzer, but is limited to sources that can be pulsed. A linear time-of-flight
analyzer, such as that in Figure 20.3.6, provide a resolution of approximately 4,000; other configurations can achieve resolutions of
10,000 or better. The time-of-flight analyzer is well-suited for MALDI ionization as the time between pulses of the laser provides
the time needed for detection to occur.

Figure 20.3.6 : Illustration of a linear time-of-flight mass analyzer showing three views of the relative positions of three ions with a
small (green), medium (blue), and large (red) mass-to-charge ratios as they migrate through the drift tube.

Ion Trap Mass Analyzers

Figure 20.3.7 provides an illustration of an ion trap mass analyzer, which consists of three electrodes—a central ring electrode and
two conical end cap electrodes—that create a cavity into which ions are drawn. The ions in the cavity experience stabilizing and
destabilizing forces that affect their movement within the cavity. Ions that adopt stable orbits remain in the cavity. By varying the
potentials applied to the electrodes, ions with different mass-to-charge ratios enter into destabilizing orbits and exit through a small
hole at the bottom of the trap. An ion trap typcially provides a resolution of 1,000.

20.3.5 https://chem.libretexts.org/@go/page/391121
 Figure 20.3.7 : Illustration
of the ion trap mass analyzer showing the cavity defined by the top end
cap, the ring electrode, and the bottom end cap. The photographs on the bottom show the
conical-shaped top end cap with small holes (inside the yellow circle) that allows ions to enter
into the cavity, the ring electrode, and the conical-shaped bottom end cap, which has a small
hole (inside the yellow circle) that allows ions to exit the cavity and reach the detector.
Ion Cyclotron Resonance Mass Analyzer
The ion cyclotron resonance (ICR) analyzer is a form of an ion trap but operates in a way that retains all ions within the trap. When
a gas phase ion is placed within an applied magnetic field, the ions move in a circular orbit that is perpendicular to the applied field
(Figure 20.3.8). In discussing the magnetric sector analyzer, we showed that the velocity, v , of an ion in an applied magnetic field
with a strength of B is a function of the radius of the ion's motion, r, and its charge
Bzer
v= (20.3.12)
m

Solving for the ratio v/r gives the ion's cyclotron frequency, w , as
c

v zeB
wc = = (20.3.13)
r m

When an ion moving in a circular orbit, as shown by the smaller of the two circular orbits in Figure 20.3.8a, absorbs energy equal
to its cyclotron frequency, w , its velocity, v , and the radius of its orbit, r both increase to maintain a constant value for w ; the
c c

result is an ion that moves in a circular orbit of greater radius. As w depends on the mass-to-charge ratio, all ions of equal m/z
c

experience the same change in their orbit, while ions with other mass-to-charge ratios are unaffected. Ions in the larger orbits
eventually return to their original circular orbit as a result of collisions in which they lose energy.

20.3.6 https://chem.libretexts.org/@go/page/391121
 Figure 20.3.8 : Illustration of how an ICR mass analyzer works. In (a) an ion originally in a circular orbit with a small radius,
shown in blue moves into a circular orbit with a larger radius, shown in green when it absorbs energy equal to its cyclotron
frequency, w , following the path shown in pink. The ICR cell in (b) consists of two transmitter plates that apply the potential that
c

alters each ion's orbit, and two receiver plates that measure the current generated by the ions; there are two additional plates, one in
front and one in back, that provide an entrance for bringing ions into the ICR cell from the ion source and for removing the ions in
preparation for the next sample. The signal in (c) decays with time as the ions lose energy through collisions.

The trap itself, as seen in Figure 20.3.8b, is defined by two pairs of plates (four in all). The transmitter plates are used to apply the
potential that alters the orbits of the ions. Movement of the ions generates a current in the receiver plates that serves as the signal,
as seen in Figure 20.3.8c, that is positive when the ion is closer to one receiver plate and negative when it is closer to the other
receiver plate. The initial magnitude of the current is proportional to the number of ions with the mass-to-charge ratio.
The ion cyclotron resonance analyzer is usually operated by applying a short pulse of energy that varies linearly in its frequency.
This sets all ions into motion, with each mass-to-charge ratio yielding a current response similar to that in Figure 20.3.8c.
Collectively, these individual current-time curves gives a time domain spectrum that we can covert into a frequency domain
spectrum by taking the Fourier transform. The frequency domain spectrum yields the mass spectrum through Equation 20.3.13. FT-
ICR instruments are capable of achieving resolutions of 1,000,000.

This page titled 20.3: Mass Spectrometers is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

20.3.7 https://chem.libretexts.org/@go/page/391121
20.4: Applications of Molecular Mass Spectrometry
Qualitative Applications
In a qualitative analysis our interest is in determining the identity of a substance of interest to us. By itself, mass spectrometry is a
powerful tool for determining the identity of pure compounds. The analysis of mixtures, however, is possible if we use a mass
spectrometer as the detector for a separation technique, such as gas chromatography, or if we string together two or more mass
analyzers in sequence.

Identification of Pure Compounds

There are several ways to use a mass spectrum to identify a compound, including identifying its molecular weight, using isotopic
ratios, examining fragmentation patterns, and by searching through data bases.
Using Molecular Weight Information. A molecular ion peak, M+•, when it is present, a [M + H]+ peak or a [M – H]+ peak,
provides information about the compound's molecular weight. When using a low resolution mass analyzer, this may be sufficient to
distinguish between molecular ions with, for example, a nominal mass of 95 amu and a nominal mass of 96 amu, but insufficient to
distinguish between molecular ions with a more precise mass of 96.0399 amu and 96.0575 amu. When using a high resolution mass
analyzer, the difference between the last pair of molecular ions may be feasible.
Using Isotopic Ratios. The molecule cycloheptene has the formula C7H12 and a nominal mass of 96 amu, and the molecule
cyclohexenone has the formula C6H8O and a nominal mass of 96 amu. Although both molecules will produce a molecular ion with
the same nominal mass-to-charge ratio, each will also have a peak with a nominal mass of M + 1 due to the presence of isotopes of
carbon, hydrogen, and oxygen. Because cycloheptene and cyclohexenone have different chemical formulas, the relative heights of
their M + 1 peaks are different. Here is how we can work this out.
For every 100 atoms of 12C there are 1.08 atoms of 13C (that is, 1.08% of the carbon atoms are 13C), for every 100 atoms of 1H
there are 0.015 atoms of 2H, and for every 100 atoms of 16O there are 0.04 atoms of 17O. For cycloheptene, this means that the
relative height of its M + 1 peak to its M peak is
(7 × 1.08) + (12 × 0.015) = 7.74

and for cyclohexenone we have

(6 × 1.08) + (8 × 0.015) + (1 × 0.04) = 6.64

Here we see that a careful examination of the relative height of the M + 1 peak provides a way to distinguish between C7H12 and
C6H8O even though they have the same nominal masses. On-line calculators are available—this link provides one example—that
you can use to calculate the full isotopic abundance patterns, including M + 2, M + 3, and other peaks. Isotopic patterns are
particularly useful for identifying the presence of chlorine and bromine in a molecule because each has one isotope with a
significant abundance: for chlorine, 37Cl has an abundance of 32.5% relative to 35Cl, and for bromine, 81Br has an abundance of
98.0% of 79Br.
Using Fragmentation Patterns. Figure 20.4.1 shows the mass spectrum of p-nitrophenol, which we first considered in Section
20.1. A molecule's mass spectrum is unique and contains information that we can use to deduce its structure. Interpretation of a
mass spectrum relies on identifying possible sources for the loss of mass, such as the a Δm of 30 amu corresponding to the loss of
NO, or a Δm of 46 amu corresponding to the loss of NO2. Some mass-to-charge ratios are recognized as evidence for a particular
ion, such as C5H5+ at a mass-to-charge ratio of 65. The interpretation of fragmentation patterns is covered elsewhere in the
curriculum, particularly in organic chemistry, and is not given more consideration here.

20.4.1 https://chem.libretexts.org/@go/page/391123
 Figure 20.4.1 : Mass spectrum of p-nitrophenol.
Using Computer Searching. Large databases of mass spectra are available (see here for a source from NIST). A peak table of
mass-to-charge ratios and peak intensities for a sample is entered into an algorithm that searches the database and identifies the
most likely matches.

Analysis of Mixtures Using MS as a Detector for a GC or LC Separation

Mass spectrometry is a powerful analytical technique when the sample we are analyzing is pure (or if impurities are of sufficiently
low concentration that they have little effect on the mass spectrum). For a mixture of two or more analytes, the interpretation of the
mass spectrum is difficult, if not impossible. To analyze such a mixture, we need a means of separating the analytes from each
other. One approach is to interface a mass spectrometer to a gas chromatograph or a liquid chromatograph. The GC, or LC
separates the mixture into its component parts with the mass spectrometer serving as the detector. See Chapter 27 and Chapter 28
for further details about GC-MS and LC-MS.

Analysis of Mixtures Using Tandem Mass Spectrometry

Another approach to working with a complex sample is to use two or more mass analyzers in what is called tandem mass
spectrometry. For example, if we place three quadrupole mass analyzers in a sequence, we can use a soft ionization source to
generate mostly molecular ions of the form [M + H]+ for each of the sample's analytes, and then let the first quadrupole separate
these molecular ions by the differences in their mass-to-charge ratio. The [M + H]+ molecular ions for one of the analytes is then
selectively passed into the second quadrupole where it is allowed to undergo fragmentation by collision with a gas, such as He.
Finally, these fragment ions are passed along to the third quadrupole where the mass spectrum is obtained. By sequentially passing
each of the molecular ions from the first quadrupole through the second and third quadrupoles, we are able to obtain mass spectra
for each molecule in the mixture.

Quantitative Applications
As a detector for other instrumental methods, such as gas chromatography and liquid chromatography, mass spectrometry provides
for a quantitative analysis by monitoring either the total ion count or by monitoring ions of a single mass-to-charge ratio, which is
known as selective ion monitoring. As an Independent method for determining an analyte's concentration, mass spectrometry is less
attractive due to the difficulty of controlling the amount of sample or standard introduced into the instrument and the affect of the
sample's matrix on fragmentation. The use of an internal standard improves precision and accuracy.

This page titled 20.4: Applications of Molecular Mass Spectrometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

20.4.2 https://chem.libretexts.org/@go/page/391123
 CHAPTER OVERVIEW
21: Surface Characterization by Spectroscopy and Microscopy
21.1: Introduction to the Study of Surfaces
21.2: Spectroscopic Surface Methods
21.3: Scanning Electron Microscopy
21.4: Scanning Probe Microscopes

This page titled 21: Surface Characterization by Spectroscopy and Microscopy is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

1
21.1: Introduction to the Study of Surfaces
Thus far we have considered methods for analyzing the bulk properties of samples, such as determining the identity or
concentration of an ion in a solution, of a molecule in a gas, or of several elements in a solid. In doing so, we did not concern
ourselves with the sample's homogeneity or heterogeneity. In this chapter we give consideration to how we can gather information
about the composition of a sample's surface and how it differs from the sample's bulk composition. But first, let's consider several
important questions.

What Is A Surface?
A surface is a boundary, or interface, between two phases, such as a solid and a gas (the type of interface of particular interest to us
in this chapter). This is a helpful, but not a sufficient description. Also of interest is the question of depth. Is a surface just the
outermost layer of atoms, ions, or molecules, or does it extend several layers into the sample? In what ways might the composition
of a sample at the surface differ from its composition in the sample's bulk interior? And, what about variations in composition
across a surface? Is the surface itself homogeneous or heterogeneous in its composition? Different analytical methods will sample
the surface to different depths and with different surface areas, which means the volume of sample analyzed will vary from
method-to-method. For this reason, we usually define a sample's surface as what is analyzed by the analytical method we are using.

Why Are Surfaces of interest?

Figure 21.1.1 shows the crystal structure of AgCl(s), which consists of a repeating pattern of Ag+ ions and Cl– ions. If you look at
the ions in interior of the structure, you will see that each Ag+ ion is surrounded by six Cl– ions, and each Cl– ion is surrounded by
six Ag+ ions. On the surface, however, we see that Cl– ions and Ag+ ions no longer are surrounded by six ions of opposite charge.
As a result, the Ag+ ions and Cl– ions on the surface are more chemically reactive than those in the interior and can serve as sites
for interesting chemistry. The chemical and physical properties of a sample's surface are likely to be very different than the
sample's bulk properties.

Figure 21.1.1 : Ball-and-stick model showing the lattice structure of AgCl. Each silver ion in the lattice’s interior binds with six
chloride ions and each chloride ion in the interior binds with six silver ions. Those ions on the lattice’s surface or edges bind to
fewer than six ions and carry a partial charge. A silver ion on the surface, for example, carries a partial positive charge. These
charges, for example, make the surface of a precipitate of AgCl an active site for chemical and physical interactions.

What Challenges Does a Surface Present?

Suppose we are interested in studying the surface of a piece of zinc metal using a probe that samples just the outermost layer of
atoms and that samples a circular surface area that is 1 µm2. How many atoms of Zn might we expect our probe to encounter? Here
is some useful information about zinc: it has a molar mass of 65.38 g/mol, it has a density of 7.14 g/cm3, and it has an atomic
radius of approximately 0.13 nm. From this we calculate the atoms per unit volume as

21.1.1 https://chem.libretexts.org/@go/page/391924
 23 15
7.14 g 100 cm 1m 1 mol 6.002 × 10  atoms 6.6 × 10  atoms
× × × × =
3 9 2
cm m 10  nm 65.38 g mol cm  nm

The units in the denominator may look odd to you, but writing them this way emphasizes that we are interested both in the depth
from which information is received (given here in nanometers, nm) and in the surface area from which information is received
(given here in square centimeters, cm2). Multiplying this value by the thickness of an atomic layer of zinc, which is twice its atomic
radius, suggests we are analyzing approximately
15
6.6 × 10  atoms 15
atoms
× 0.26 nm = 1.7 × 10
2 2
cm  nm cm

If we multiply this value by the surface area that we are sampling from, then we are interacting with approximately
2 2 2
atoms 100 cm 1m
15 7
1.7 × 10 ×( ) ×( ) = 1.7 × 10  atoms of Zn
cm
2
m 10
6
µm

Although 17 million may seem like a large number, it is not a particularly large number of atoms on which to carry out an analysis.
Now, suppose the surface has a 10 ppm impurity of copper atoms; that is, there are 10 copper atoms for every 106 zinc atoms. In
this case, our probe of the sample involves just

7
10 atoms of Cu
1.7 × 10  atoms of Zn × = 170 atoms of Cu
6
10  atoms of Zn

As a comparison, if we analyze a sample in which the analyte is present at a concentration that is 1 × 10

−6
 mol/L using an
analytical method that gathers information from a volume that is just 1 mm3, then we are sampling
−6 3
1 × 10  mol 1L 1cm 3 23 −1 11
× ×( ) × 1  mm × 6.022 × 10  mol = 6.0 × 10  particles of analyte
L 1000 cm 3 10 mm

An additional challenge when we attempt to analyze a surface is that a freshly exposed surface becomes contaminated with an
absorbed layer of gas molecules almost instantly when sitting on a laboratory bench, and in a few seconds to a few minutes at
pressures in the range of 10–6 torr to 10–8 torr. Analysis of a surface requires careful attention to how the surface is prepared.

What Opportunities Does a Surface Present?

Compared to many of the methods in Chapters 6–20 and in Chapters 22–34, the use of a probe that samples from a small area
allows for moving the probe across the surface—this is called rastering—developing a two-dimensional image of the surface.
When using an energetic beam that can etch a hole in the sample, we can obtain information at depth—a process called depth
profilling—that provides information in a third dimension. These are particularly important strengths of surface analytical methods.

How Can We Probe the Surface?

To study a surface, we put energy into it in the form of a beam of photons, electrons, or ions and then we measure the energy that
exits the surface in the form of a beam of photons, electrons, or ions. Table 21.1.1 shows some of the possibilities. Also included in
this table are methods in which an applied field generates a response from the surface. Entries in bold receive attention in this
chapter. Surface enhanced Raman spectroscopy received a brief mention in Chapter 18. Note that Auger electron spectroscopy
appears twice as the emission of electrons can follow the input of X-ray photons or electrons.
Table 21.1.1 . Classifying surface analysis methods based on the input energy and the output energy.
energy out →
photon electron ion field
energy in ↓

X-ray photoelectron
surface enhanced Raman spectroscopy (XPS)
spectroscopy (SERS) Auger electron laser-microprobe mass
photon —
extended X-ray absorption spectroscopy (AES) spectrometry (LAMMA)
fine structure (EXAFS) UV-photoelectron
spectroscopy (UPS)

21.1.2 https://chem.libretexts.org/@go/page/391924
 energy out →
photon electron ion field
energy in ↓

Auger electron
spectroscopy (AES)
energy dispersive X-ray
scanning electron
electron spectroscopy (EDS) — —
microscopy (SEM)
electron microprobe (EM)
low energy electron
diffraction (LEED)

Rutherford back scattering

(RBS)
ion — —
secondary ion mass
spectrometry (SIMS)

scanning tunneling atomic force microscopy

field — —
microscopy (STM) (AFM)

 Note

There are other ways to probe a surface by putting energy into it, including the application of thermal energy and the use of
neutral species. See the text Methods of Surface Analysis, Czanderna, A. Editor, Elsevier: Amsterdam (1975) and the article
"Analytical Chemistry of Surfaces" by D. M. Hercules and S. H. Hercules, J. Chem. Educ. 1984, 61, 402–409 for detailed
reviews. Although neither is a recent publication, both provide an excellent introduction to surface analysis.

This page titled 21.1: Introduction to the Study of Surfaces is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

21.1.3 https://chem.libretexts.org/@go/page/391924
21.2: Spectroscopic Surface Methods
In this section we consider three representative surface analytical methods: X-ray photoelectron spectroscopy, in which the input is
a beam of X-ray photons and the output is electrons; Auger electron spectroscopy, in which the input is either a beam of electrons
or of X-ray photons and the output is electrons; and secondary-ion mass spectrometry, in which the input is a beam of ions and the
output is ions.

X-Ray Photoelectron Spectroscopy (XPS)

In X-ray photoelectron spectroscopy, which is also known as electron spectroscopy for chemical analysis (ESCA), we measure the
kinetic energy of electrons ejected from a sample following the absorption of X-ray photons. The resulting spectrum is a count of
these emitted electrons as a function of their energy.

Principles of XPS
We can explain the origin of X-ray photoelectron spectroscopy using the photoelectric effect. Figure 21.2.1a shows the energy
level diagram for an element's 1s, 2s, and 2p core-level electrons along with their KLM designations (see Chapter 12.1 for a
previous discussion of this way of designating electrons). A nearly monoenergetic X-ray beam of known energy is focused on the
sample, which results in the ejection of a core-level photoelectron, as shown in Figure 21.2.1b. The kinetic energy of this emitted
electron, E , is related to its binding energy to the nucleus, E , by the equation
KE BE

EKE = hν − EEB − Φw (21.2.1)

where hν is the energy of the X-ray photon and Φ is the spectrometer's work function (the energy needed to remove the electron
w

from the surface and into the vacuum). The most common sources of X-rays are the Mg K line with an energy of 1253.6 eV or
α

the Al Kα line with an energy of 1486.6 eV.

The core-level vacancy created by the photoelectron leaves the atom with an unstable electron configuration. Relaxation to the
ground state occurs when this vacancy is filled by an electron from a higher energy shell, with the excess energy released as either
the emission of a second electron or the fluorescent emission of a characteristic X-ray, as seen in Figure 21.2.1c. The secondary
electron in Figure 21.2.1c is called an Auger electron.

Figure 21.2.1 : Energy level diagrams for X-ray photoelectron spectroscopy. In (a) we see the 1s, 2s, and 2p electrons along with
their KLM designations. When a photon of sufficient energy is absorbed by the sample (b), a core-level electron is ejected as a
photoelectron leaving an unstable electron configuration in which there is a vacancy in the 1s orbital. In (c) an electron from a
higher energy orbital moves into the vacancy created by the initial photon with the remaining energy released as a secondary
electron or as an X-ray photon.
Figure 21.2.2 provides an example of an XPS survey spectrum for aluminum oxide, Al O , using the K line for aluminum as the
2 3 α

source of X-rays. The peak table gives the binding energies of the peaks for aluminum and for oxygen using the K for Al and, for
α

comparison, the K line for Mg. Note the difference in how the major peaks are labeled. The photoelectrons ejected in the process
α

shown in Figure 21.2.1b are designated by the element and the ns notation that specifies the orbital from which the electron was
ejected. Auger electrons are designated using the KLM notation, specifying the initial vacancy created by the absorbed photon, the
source of the electron that fills that vacancy, and the source of the ejected Auger electron. The OKLL peak in this spectrum is
consistent with the scheme shown in Figure 21.2.1c. When the source of the second and third electrons is from the valance shell,
then notation is sometimes written as KVV; the OKLL peak here could be designated as OKVV.

21.2.1 https://chem.libretexts.org/@go/page/391926
 Figure 21.2.2 : The XPS spectrum for Al O using the Al K line, with the major peaks labeled. A spectrum such as this, which
2 3 α

spans a wide range of energies, is called a survey spectrum. The source of each peak is shown above or next to the spectrum's
peaks. The peak table provides the binding energies for aluminum and for oxygen using both the K for Al and the K for Mg.
α α

The original data used to create this figure is from the xpslibrary.
There are a few additional interesting features to note in the survey spectrum for Al2O3. One is the presence of a peak for carbon
even though the sample, Al2O3, does not contain carbon. The prevalence of carbon in the atmosphere means that trace levels of
carbon appear in almost all XPS spectra.
A second feature is the increase in the signal on the high binding energy (low kinetic energy) side of peaks, which is particularly
visible here for the O1s peak the C1s peak, and the Al2s peak. The source of this background is electrons that fail to escape the
sample without undergoing inelastic collisions that result in a loss of kinetic energy and, given Equation 21.2.1, are recorded as if
they have a larger than expected binding energy. Because X-rays penetrate more deeply into the sample than the depth from which
electrons can travel without undergoing an inelastic collision, this background is unavoidable. Note that the background is more
significant at higher binding energies (smaller kinetic energies).
A third feature is that the binding energy of an XPS peak is independent of the X-ray source, but the binding energy for an Auger
peak varies with the X-ray source's energy (see table in Figure 21.2.2). The kinetic energy of the O1s, Al2s, and Al2p photoelectrons
is the difference between the energy of the X-ray photon, hν , and each electron's binding energy, BE; if we change the X-ray
source, then hν and KE increase in value, but the BE remains fixed. For the OKLL Auger electron, the kinetic energy depends on
the difference in the binding energies of the three electrons involved
KE ≈ BE K − BE L − BE L (21.2.2)

and is, therefore, independent of the energy of the X-ray source. Given Equation 21.2.1, if the KE remains constant, then an
increase in the energy of the X-ray photon, hν , means that the apparent BE must increase. This shift in binding energy when using
a different source is one way to identify a peak as resulting from Auger electrons.

Instrumentation
The basic instrumentation for XPS is shown in Figure 21.2.3. The most common X-ray sources, as noted above, are Mg (1253.6
eV) and Al (1486.6 ev), which have the advantage of relatively narrow line-widths (0.7 and 0.9 eV, respectively) and, therefore, a
relatively narrow range of energies. Higher energy sources are available, such as Ag (2984.4 ev), but at the cost of a wider line-
width (2.6 eV). A system of electron lenses collects and focuses the ejected electrons onto the entrance slit of a hemispherical
analyzer. The path of an electron through the analyzer depends upon its kinetic energy. By varying the potentials applied to the
hemispherical analyzer's inner and outer plates, electrons of different kinetic energies reach the detector. A sputtering gun is an
optional feature that can be used to clean the surface of the sample or to remove successive layers of the sample, allowing for the
gathering of spectra at various depths within the sample. Calibration of the spectrometer's binding energy scale, which accounts for

21.2.2 https://chem.libretexts.org/@go/page/391926
the spectrometer's work function, is made using specific lines for one or more conductive metals; examples include Au 4f7/2 at
83.95 eV and Cu 2p3/2 at 932.63 ev. The peak for carbon that appears in almost all XPS spectra provides an additional way to check
the calibration of the binding energy scale.

Figure : Schematic diagram of the basic instrumentation for X-ray photoelectron

21.2.3

spectroscopy: an X-ray source for generating photoelectrons; a sputtering gun for cleaning the
sample's surface or for removing layers of the sample; a set of electron lenses for collecting
and focusing the photoelectrons; a hemispherical analyzer for isolating electrons by their
kinetic energies; and a transducer to count electrons. The entire instrument is held under a
vacuum.
Applications
X-ray photoelectron spectroscopy is a particularly useful tool for determining the composition and structure of a sample. XPS also
can provide information about how the composition of a sample varies with depth and quantitative information about a sample's
components.
Qualitative Analysis. One of the strengths of X-ray photoelectron spectroscopy is the ability to determine the elements that make
up a sample's surface. Figure 21.2.4 shows a survey scan from 0 eV to 1100 eV of a ceramic material. In addition to the O1s peak,
we see strong peaks for Si and Al—probably an aluminosilicate ceramic—and small peaks for a variety of elements: La, Ba, Mn,
Sn, Ca, Cl, P, and Mg. NIST maintains an extensive, and searchable, database of XPS peaks that help in identifying the elements in
a sample.

Figure 21.2.4 : X-ray photoelectron survey spectrum for an unidentified ceramic material showing the peak assignments, which
include major elements that we might reasonably expect from a ceramic (oxygen, magnesium, aluminum, calcium, and silicon) and
a variety of trace elements (lanthinum, barium, manganese, tin, chlorine, and phosphorous). The original data used to create this
figure is from the xpslibrary.

21.2.3 https://chem.libretexts.org/@go/page/391926
Chemical Shifts. An element's binding energy is sensitive to its chemical environment, particularly with respect to oxidation states
and structure. For example, Table 21.2.1 provides the binding energy for chlorine's 2p line in three potassium salts drawn from the
NIST database with all values from the same literature source. Using KCl, in which chlorine has an oxidation state of –1, as a
baseline, the Cl 2p peak for KClO3 is shifted by +8.4 eV and the Cl 2p peak for KClO4 is shifted by +10.3 eV. The direction of the
shift makes sense, as we expect that it will require more energy to remove an electron from an element that has a more positive
oxidation state.
Table 21.2.1 . Binding energies for the Cl 2p XPS peak as a function of chlorine's oxidation state.
compound oxidation state for chlorine binding energy (eV) for Cl 2p line ΔBE relative to KCl

KCl –1 198.1 —

KClO3 +5 206.5 +8.4

KClO4 +7 208.4 +10.3

Chemical shifts also reflect changes in chemical structure. Figure 21.2.5, for example, shows a high resolution scan for the oxygen
2p peak for same sample of aluminum oxide, Al2O3, in Figure 21.2.1. The surface of a metal oxide often has three distinct sources
of oxygen: oxides, which make up the bulk of the sample, hydroxides that form at the surface following the chemisorption of water,
and water that is physically absorbed to the surface. Curve-fitting of the raw data shows the contribution of each type of oxygen to
the raw data and, through the peak areas, their relative abundance.

Figure 21.2.5 : The results of curve-fitting to determine the relative contribution of different sources of oxygen—oxides,
hydroxides, and water—to the O2p peak for aluminum oxide. The original data used to create this figure is from the xpslibrary.
Wagner Plots. Both the binding energy of an X-ray photoelectron and the kinetic energy of an Auger electron convey information
about the element from which the electrons were emitted. A Wagner plot shows both the binding energy for a photoelectron that
leaves a particular core-level vacancy and the kinetic energy of the Auger electron whose origin arises from the filling of this core-
level vacancy. Figure 21.2.6 shows an example of a Wagner plot for copper based on its 2p3/2 X-ray photoelectron and its LMM
Auger electron. The diagonal lines are called the modified Auger parameter, which is defined as the sum of the XPS binding energy
and the AES kinetic energy. Values for 20 compounds are included in this plot. Of interest here is the clustering of the individual
compounds. For example, all of the samples for which copper has an oxidation state of +1 (shown as magenta squares) have similar
binding energies between 932 eV and 933 eV, but with more variable kinetic energies, which range from 914 eV to 917 eV. Most of
the compounds in which copper has an oxidation state of +2 (shown as blue diamonds) have modified Auger parameters between
approximately 1850 eV and 1851 eV, although there is signifiant variation in their individual binding energies and kinetic energies.
The two metals (shown as green circles) and the commpounds CuS and CuSe occupy a similar space within the Wagner plot.
Interestingly, both CuS and CuSe are transition metal chalcogenides and have semiconducting properties.

21.2.4 https://chem.libretexts.org/@go/page/391926
 Figure 21.2.6 : Wagner plot for 20 copper compounds. The raw data is from https://srdata.nist.gov/xps/Default.aspx and includes
140 measurements of binding energies and 99 measurements of kinetic energies. Compounds are identified by the oxidation state of
copper.
Information at Depth. X-rays penetrate into a sample to a depth that is greater than the distance the photoelectron can travel
without losing energy to inelastic collisions. We can take advantage of this to vary the depth from which we gather information.
Figure 21.2.7 shows how this is accomplished by changing the angle from which we collect and analyze photoelectrons. The
length of the solid black line is the distance an electron can travel without losing energy in an inelastic collision. When the detector
is placed at 90° to the sample's surface, the sampling depth is at its greatest. Adjusting the detector so that it is at 30° to the surface,
results in its detecting electrons from a depth that is just half of that when the detector is at 90°.

Figure 21.2.7 : Illustration showing the measurement of the photoelectron's angle of emission. See text for explanation for how this
affects sampling depth.
Quantitative Analysis. The intensity of an XPS peak—either is peak height or its peak area—is proportional to the number of
atoms of the element responsible for the peak. This allows for determining the relative concentration, C , of an element in a
x

sample; thus
Ix / Sx
Cx = (21.2.3)
∑ (Ii / Si )

where I is the peak intensity for the element, S is the sensitivity factor for the element, and I and S are the peak intensities and
x x i i

sensitivity factors for all other elements in the sample. Sensitivity factors account for differences in the ease with which
photoelectrons are produced and escape from the sample. Published tables of sensitivity factors are available, although they may
vary some from instrument-to-instrument. Sensitivity factors are referenced to a standard line, typically C1s, which is assigned a
sensitivity factor of 1.00.
In many cases we are interested only in the relative concentration of just two elements. In this case, we write Equation 21.2.3 as

21.2.5 https://chem.libretexts.org/@go/page/391926
 Cx Ix / Sx
= (21.2.4)
Cy Iy / Sy

where x and y are the two elements. For example, using the data in Figure 21.2.4, the Si2p peak has a height of 30 mm and the Al2p
peak has a peak height of 20 mm. Their sensitivity factors are, respectively 0.817 and 0.737. Using these values, we find that

CSi 30/0.817
= = 1.4
CAl 20/0.737

there are approximately 1.4× as many atoms of silicon as there are atoms of aluminum.

Auger Electron Spectroscopy (AES)

In Figure 21.2.2 we learned that following the ejection of a photoelectron from an atom's core, the now unstable atom releases
energy by either emitting a secondary electron from a higher energy orbital or by releasing a photon. In XPS we measure the
intensity of the ejected photoejected electrons as a function of their binding energy; in Auger spectroscopy we measure the
intensity of these secondary electrons as a function of their kinetic energies.
The instrumentation for AES can be coupled with an XPS spectrometer or be a stand-alone instrument. In either case, the basic
instrumentation is similar to that shown in Figure 21.2.3, although AES spectra are usually initiated using an electron gun as a
source instead of an X-ray source. One advantage to using an electron gun is that it can be focused into a smaller beam and can
then easily rastered across a surface, allowing for imaging of the sample's surface. Depth profiling, using an ion beam to remove
layers of the sample, is another common use of AES.
Figure 21.2.8 provides an example of a typical AES spectrum, in this case for the mineral calcite, CaCO3. The raw data, on the left,
shows the intensity of the signal as a count of electrons with a particular kinetic energy. The large background signal is from
electrons that lose kinetic energy as the result of inelastic collisions. The two broad features between approximately 100 eV and
600 eV are the Auger peaks for calcium and for oxygen. The peaks in this case are easy to see because the two analytes are present
in bulk. For a trace-level analyte, the Auger peaks in a normal plot may be difficult to see. For this reason, Auger spectra are
usually presented by plotting the derivative of the raw data giving the spectrum on the right.

Figure 21.2.8 : AES spectrum for the mineral calcite, CaCO3. The figure on the left shows the spectrum as initially recorded and the
figure on the right shows the data after taking the derivative of counts as a function of energy. The original data used to create thid
figure is from the xpslibrary.
Figure 21.2.9 provides an example of how AES is used to study changes in the composition of a single crystal of CaCO3 that was
allowed to equilibrate with a solution containing Mg2+ ions. The sample was mounted on a sample probe and the AES spectrum
recorded. An Ar+ ion beam was used to remove layers of the sample while the spectrometer was used to record spectra of the
sample. As expected, the surface is enriched in Mg2+ ions that diffused into the crystal with the relative concentration of Mg2+
decreasing. The relative abundance of Ca2+ increases with depth; the relative concentration of oxygen remains more or less
constant with depth.

21.2.6 https://chem.libretexts.org/@go/page/391926
 Figure 21.2.9 : Depth profiles for calcium, oxygen, and magnesium in a sample of CaCO3 that was suspended in a solution of
Mg2+ ion. The shading for each element shows the relative abundance of the element with depth. See Mucci, A.; Morse, J. W.;
Kaminsky, M. S. "Auger Spectroscopy Analysis of Magnesian Calcite Overgrowths Precipitated From Seawater and Solutions of
Similar Composition," Am. J. Sci. 1985, 289-305 for the paper suggesting this figure.

Secondary-Ion Mass Spectrometry (SIMS)

In SIMS we bombard the surface of a sample with an energetic primary beam—typically 5-20 keV—of an ion, such as Ar+, O2+, or
Cs+. The primary ion penetrates the sample's surface, ejecting a variety of particles, including neutral atoms, electrons, and, more
importantly, secondary ions (singly-charged or multiply charged cations and anions, and clusters of ions). These secondary ions are
characterized by their mass-to-charge ratios and by their kinetic energies. The use of a primary ion beam of Cs+ favors the
formation of secondary ions that are anions, and a primary beam of O2+ favors the formation of secondary ions that are cations.
The instrumentation for SIMS includes an ion gun for generating the primary beam and a mass spectrometer for analyzing the
secondary ions. In static mode, the primary ion beam is run using a low current density that minimizes the extent to which the
sample's outermost layers are removed. In dynamic mode, the primary beam is run at a higher current density that removes more of
the sample's surface. Dynamic SIMS is well suited to depth profiling. High mass resolution is obtained using a time-of-flight mass
analyzer or a double-focusing mass analyzer; see Chapter 20 for a discussion of different types of mass analyzers.
SIMS is well suited for imaging as the positively charged primary ion can be rastered across the sample's surface. Figure 21.2.10
provides an example of imaging a surface using SIMS by measuring the yield of 14N12C ions while rastering the primary ion beam
across a 10 µm × 10 µm section of the sample.

Figure 21.2.10: Image created using the R package lans2r ["lans2r: Work with Look at NanoSims Data in R", Sebastian Kopf, R
package version 1.1.0, 2020]. The image shows the analysis of a 10 µm by 10 µm portion of a surface for the 14N12C ion, the two
major isotopes of carbon and nitrogen. The bright areas have larger total ion counts.

This page titled 21.2: Spectroscopic Surface Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

21.2.7 https://chem.libretexts.org/@go/page/391926
21.3: Scanning Electron Microscopy
In optical microscopy we use photons to provide images of a sample. Although extraordinarily useful and powerful, the ability to
resolve features in optical microscopy is limited by the source of light; in general, we can distinguish between two objects if they
are separated by a distance that is greater than the wavelength of the photons being used. The maximum resolution for an optical
microscope is about 0.2 µm (200 nm), which means we can use an optical microscope to view a human hair (20-200 µm), an
eukaryotic cell (10-100 µm), a chloroplast (5-8 µm), and a mitochondrion (1-3 µm), but not a ribosome (0.01 µm-0.02 µm). In this
section we will consider the electron microscope, which has a resolution limit of approximately 0.2 nm, or approximately 1000×
more than an optical microscope. In Section 21.4, we will examine two additional types of non-optical microscopy.

Instrumentation
In scanning electron microscopy we raster a beam of high-energy electrons over a surface using a two-dimensional grid. Figure
21.3.1 shows the basic instrumental needs. The electron gun usually is just a simple tungsten wire that releases electrons when it is

heated resistively. Other sources include solid-state crystals of lanthanum hexaboride (LaB6) or cerium hexaboride (CeB6) and the
field emission gun, which uses a tungsten wire with a tip that has a radius of about 100 µm. Regardless of their source, these
electrons are accelerated to an energy of 1-40 keV and passed through a series of lens that narrow and focus them into a beam with
a diameter that falls within a range of 1 nm to 1000 nm (0.001 µm to 1 µm). A set of coiled scan controls deflects the electron
beam in a raster pattern across the sample's surface (see inset at the bottom left of Figure 21.3.1). An electron detector monitors the
electrons that scatter back from the sample; the type of detector used varies with the type of emission from the sample that we
choose to monitor—see the next sub-heading for types of emission—but typically are scintillation devices when monitoring
electrons and energy-dispersive detectors when monitoring X-rays.

Figure 21.3.1 : Schematic diagram of an instrument for scanning electron microscopy, which consists of an electron gun, a system
of magnetic lenses, a scan control, and a detector. The inset on the lower left shows a raster pattern in which the electron beam is
scanned back-and-forth across the sample's surface.

Interaction of Electron Beams With Solids

Figure 21.3.1 suggests that the only type of signal is the measurement of electrons that are scattered back toward the detector. The
interaction between the electron beam and the sample, however creates a variety of signals, including both electrons and X-rays.
Figure 21.3.2 illustrates the types of emission that follows from the interaction of the electron beam with the sample. The electron
beam penetrates approximately 1-2 µm into the sample. As you might expect from the previous section on electron spectroscopy,
the interaction of an electron beam with a sample results in the emission of some Auger electrons; these electrons come from a
volume near the vacuum-sample interface. Of more importance are secondary electrons and backscattered electrons.
As the electron beam penetrates into the sample, the electrons undergo collisions with the sample's atoms. Some of these collisions
are elastic in which the electron changes its direction, but retains its kinetic energy. With sufficient time, these electrons eventually
undergo a collision in which they cross the sample-vacuum interface and exit the solid. These backscattered electrons are collected
and passed along to the detector. Other electrons undergo inelastic collisions, losing kinetic energy and, eventually, become
embedded in the sample. Backscattered electrons come from a depth as great as 50% of the depth to which the electron beam
penetrates. Another source of electrons comes from a process in which the electron beam induces the ejection of electrons from the

21.3.1 https://chem.libretexts.org/@go/page/391929
sample's conduction band. These secondary electrons are less numerous than backscattered electrons and they also come from a
much shallower depth, typically 5-50 nm.
The electron beam also stimulates the release of X-rays, including the characteristic X-rays of the sample's elements, a broad
continuum, and fluorescent X-ray emission. See Chapter 12 for more details about atomic X-ray emission.

Figure 21.3.2 : Illustration showing the types of emission that results from the interaction of an electron beam with a sample. The
electron beam penetrates to a depth of approximately 1-2 µm, with the interaction volume increasing with depth. X-rays are emitted
from deeper in the sample than are electrons.
As the electron beam is rastered across the sample, the intensity of the backscattered electrons from a specific position on the
sample that reaches the detector is stored in the corresponding pixel on the instrument's monitor. The image created in this way is
not an optical picture, but a digitized electronic reproduction of the sample's surface. The extent of magnification depends on the
length of the detector's monitor relative to the length of a single scan across the sample; scanning a shorter distance results in a
greater magnification. An optical microscope usually provides a maximum magnification of 1000×; an SEM can achieve a
magnification of 1, 000, 000×.

Applications
Figure 21.3.3 shows four examples of applications of scanning electron microscopy for the measurement of particle size (upper
left), for the evaluation of nanowires (upper right), for characterizing the channels in a microfluidic device (lower left), and for
examining the tip of a cantilever and tip used for atomic force microscopy. Other applications include biological samples, films and
coatings, fibers, and powders, to name a few.

21.3.2 https://chem.libretexts.org/@go/page/391929
 Figure 21.3.3 : Scanning electron microscope images. The image in (a) shows six particulate spheres with diameters of about 2 µm
each. The image in (b) shows a cluster of nanowires, each with a diameter of approximately 200 nm. The image in (c) shows the
intersection of channels in a microfluidic device. The channels have widths of approximately 10 µm, although the upper channel
narrows to about 2-4 µm at the intersection of the channels. The image in (d) is the tip of a cantilever used for atomic force
microscopy, which we will explore in more detail in Section 21.4. All images are released under a CC-BY license and are available
here.

This page titled 21.3: Scanning Electron Microscopy is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

21.3.3 https://chem.libretexts.org/@go/page/391929
21.4: Scanning Probe Microscopes
in the last section we considered how we can image a surface using an electron beam. In this section we consider a very different
approach to developing an image of a surface, one in which we bring a probe close to the surface and examine how the probe
interacts with the surface. One advantage of this approach is that the interaction between the probe and the surface can include
attraction and repulsion, which opens up a third dimension to the image.

Scanning Tunneling Microscope (STM)

In the scanning tunneling microscope we take advantage of the ability of a current to pass through the gap between the tip of a
conducting probe and a conducting sample when the probe and the sample are held at different potentials. Figure 21.4.1 shows the
basic arrangement in which the probe has, ideally, a single atom at its tip. The tunneling current, I is given by
t

−C d
It = V e (21.4.1)

where V is the applied voltage, d is the distance between the probe's tip and the sample, and C is a constant whose value depends
upon the composition of the probe and the sample. The exponential decrease in the tunneling current with distance means that a
small change in the position of probe's tip relative to the sample results in a significant change in the signal, providing vertical
resolution on the order of 0.1 nm. Probes are fashioned using tungsten wires or platinum-iridium wires.

Figure 21.4.1 : Illustration showing the relationship between the sample and the tip of the probe in scanning tunneling microscopy.
Scanning tunneling microscopy images are created by moving the probe back-and-forth across the sample while measuring the
current. The signal is acquired in one of two modes: constant current or constant height. In constant current mode, the probe's tip is
brought near the surface and the current measured, which establishes a setpoint. As the probe moves across the sample, it is raised
or lowered to maintain the setpoint current. The result is measure of the distance, d , between the probe's tip and the sample along
the z-axis as a function of the xy position of the probe's tip. In constant height mode, the distance, d , between the probe's tip and the
sample is held constant, and the current, I , is measured as a function of the xy position of the probe's tip. Constant height mode
t

allows for faster data acquisition, but is limited to samples that have flat surfaces.
Positioning of the sample and the probe's tip relative to each other is accomplished by either moving the probe or moving the
sample. In either case, the control of movement in accomplished using a piezoelectric scanner. A piezoelectric material, as shown
in Figure 21.4.2 experiences a change in its length when a dc potential is applied across its sides, either extending its length or
contracting its length.

Figure 21.4.2 : Illustration showing the behavior of a piezoelectric material when a dc potential is applied to it. The extension and
contraction shown here are, of course, exaggerated.

21.4.1 https://chem.libretexts.org/@go/page/391931
Figure 21.4.3 shows a configuration of cylindrical piezoelectric scanner in which the cylinder's upper half controls movement
along the z-axis, and the cylinder's lower half is used to control movement along the x-axis, the y-axis, or both.

Figure 21.4.3 : Illustration that shows the design of a piezoelectric scanner. The top half of the scanner controls movement along the
z-axis and the lower half of the scanner controls movement along the x-axis and the y-axis. The piezoelectric material is shown in
black, the positive electrodes in blue, and the ground in red.
One limitation of STM is that the sample must be conductive. It is possible to image a non-conducting sample if it first coated with
a conductive material, such as gold, although such coatings can mask surface features.

Atomic Force Microscope (AFM)

Unlike the scanning tunneling microscope, the atomic force microscope does not require a conducting sample and imaging is
achieved without a current flowing between the sample and the tip of the probe. Instead, as shown in Figure 21.4.4, the probe is
attached to the end of a flexible cantilever. The tip of the probe (see photograph in Figure 21.4.3) is pyramidal in shape and extends
about 10 µm from its base on the cantilever. The tip of the probe has a diameter on the order of 10 nm and is made of silicon, Si, or
silicon nitride, Si3N4. The cantilever typically is 100-500 µm in length. The probe is scanned across the sample's surface and the
position of the probe relative to the surface is determined by reflecting the beam from a diode laser off the probe-end of the
cantilever to a detector.

Figure 21.4.4 : The figure on the left illustrates how the position of the cantilever and the probe's tip is monitored in atomic force
microscopy. The photo on the right shows an SEM image of a typical cantilever and probe tip. The photograph is released under a
CC-BY license and is available here.
The force in atomic force is the interaction between the probe's tip and the sample, which may be a force of attraction or a force of
repulsion. When the probe's tip is in contact with the sample—known as contact mode—there is a force of repulsion between them.
Because the cantilever has a smaller force constant than the atoms in the probe's tip, the cantilever bends. Moving the sample stage
to maintain a constant deflection of the laser off of the cantilever provides an image of the sample's surface. Contact mode allows
for rapid scanning and work well for samples with rough surfaces, although it may damage samples with softer surfaces.
In non-contact mode, the probe's tip is brought close to the sample's surface, but not allowed to come into contact with it. The
cantilever is place into an oscillatory motion. The amplitude of this oscillation is proportional to the force of attraction between the
probe's tip and the sample, which varies with the distance between the probe's tip and the sample. Moving the sample stage to

21.4.2 https://chem.libretexts.org/@go/page/391931
maintain a constant oscillation provides an image of the sample's surface. Non-contact mode AFM generally provides lower
resolution images, but is less damaging to the sample.
A third mode for collecting data is called intermittent or tapping mode. In this mode the cantilever is set to oscillate at its resonant
frequency with the probe's tip coming into contact with the sample's surface when it reaches the bottom of the cantilever's
oscillation. The frequency of the oscillation is sensitive to the distance between the probe's tip and the sample. Moving the sample
stage to maintain the resonant frequency provides an image of the sample's surface.
You can view a gallery of scanning tunneling microscopy images here, and a gallery of atomic force microscopy images here.

This page titled 21.4: Scanning Probe Microscopes is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

21.4.3 https://chem.libretexts.org/@go/page/391931
 CHAPTER OVERVIEW
22: An Introduction to Electroanalytical Chemistry
In Chapters 6–21 we examined a wide range of spectroscopic techniques that take advantage of the interaction between
electromagnetic radiation and matter. In this chapter we turn our attention to electrochemical techniques in which the potential,
current, or charge in an electrochemical cell serves as the analytical signal.
Although there are only three fundamental electrochemical signals, there are many possible experimental designs—too many, in
fact, to cover adequately in an introductory textbook. The simplest division of electrochemical techniques is between bulk
techniques, in which we measure a property of the solution in the electrochemical cell, and interfacial techniques, in which the
potential, current, or charge depends on the species present at the interface between an electrode and the solution in which it sits.
The measurement of a solution’s conductivity, which is proportional to the total concentration of dissolved ions, is one example of
a bulk electrochemical technique. A determination of pH using a pH electrode is an example of an interfacial electrochemical
technique. Only interfacial electrochemical methods receive further consideration in this textbook. In this chapter we provide an
introduction to electrochemistry, introducing ideas relevant to understanding the specific electroanalytical methods introduced in
Chapters 23–25.
22.1: Electrochemical Cells
22.2: Potentials in Electroanalytical Cells
22.3: Electrode Potentials
22.4: Calculation of Cell Potentials from Electrode Potentials
22.5: Currents in Electrochemical Cells
22.6: Types of Electroanalytical Methods

This page titled 22: An Introduction to Electroanalytical Chemistry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

1
22.1: Electrochemical Cells
A schematic diagram of a typical electrochemical cell is shown in Figure 22.1.1. The electrochemical cell consists of two half-
cells, each of which contains an electrode immersed in a solution of ions whose activities determine the electrode’s potential. A salt
bridge that contains an inert electrolyte, such as KCl, connects the two half-cells. The ends of the salt bridge are fixed with porous
frits, which allow the electrolyte’s ions to move freely between the half-cells and the salt bridge. This movement of ions in the salt
bridge completes the electrical circuit, allowing us to measure the potential using a potentiometer.

Figure 22.1.1 . Example of a potentiometric electrochemical cell. The activities of Zn2+ and Ag+ are shown below the two half-
cells.
The reason for separating the electrodes is to prevent the oxidation reaction and the reduction reaction from occurring at the same
electrode. For example, if we place a strip of Zn metal in a solution of AgNO3, the reduction of Ag+ to Ag occurs on the surface of
the Zn at the same time as a portion of the Zn metal oxidizes to Zn2+. Because the transfer of electrons from Zn to Ag+ occurs at the
electrode’s surface, we can not pass them through the potentiometer.

Conduction in a Cell
Current moves through the cell in Figure 22.1.1 as a result of the movement of two types of charged particles: electrons and ions.
First, when zinc, Zn(s) underoges an oxidation reaction
2+ −
Zn(s) ⇌  Zn (aq) + 2 e (22.1.1)

it releases two electrons. These electrons move through the circuit that connects the metallic Zn electrode in the left half-cell to the
metallic Ag electrode in the right half-cell, where it effects the reduction of Ag+(aq).
+ −
Ag (aq) + e ⇌ Ag(s) (22.1.2)

If this is all that happens, then the half-cell on the left will develop an excess of positive charge as Zn2+(aq) ions accumulate and
the half-cell on the right will develop an excess of negative charge due to the loss of Ag+(aq). The salt bridge provides a way to
continue the movement of charge, and thus the current, with the K+ ions moving toward the right half-cell and Cl– ions moving
toward the left half-cell.

Galvanic and Electrolytic Cells

The net reaction for the electrochemical cell in Figure 22.1.1 is
+ 2+
Zn(s) + 2 Ag (aq) ⇌ 2Ag(s) + Zn (aq) (22.1.3)

which simply is the result of adding together the reactions in the two half-cells after adjusting for the difference in electrons. As
shown by the arrows in the figure, when we connect the electrodes to the potentiometer, current spontaneously flows from the left
half-cell to the right half-cell. We call this a galvanic cell. If we apply a potential sufficient to reverse the direction of the current
flow, resulting in a net reaction of
2+ +
2Ag(s) + Zn (aq) ⇌ Zn(s) + 2 Ag (aq)

then we call the system an electrolytic cell. A galvanic cell produces electrical energy and an electrolytic cell consumes electrical
energy.

22.1.1 https://chem.libretexts.org/@go/page/333388
Anodes and Cathodes
The half-cell where oxidation takes place is called the anode and, by convention, it is shown on the left for a galvanic cell. The
half-cell where reduction takes place is called the cathode and, by convention, it is shown on the right for a galvanic cell.

Faradaic and Non-Faradaic Currents

When we oxidize or reduce an analyte at the electrode in one half-cell, the electrons pass through the potentiometer to the electrode
in the other half-cell where a corresponding reduction or oxidation reaction takes place. In either case, the current from the redox
reactions at the two electrodes is called a faradaic current. A faradaic current due to the reduction of an analyte is called a cathodic
current and carries a positive sign. An anodic current results from the analyte’s oxidation and carries a negative sign.
In addition to the faradaic current from a redox reaction, the current in an electrochemical cell includes non-faradaic sources.
Suppose the charge on an electrode is zero and we suddenly change its potential so that the electrode’s surface acquires a positive
charge. Cations near the electrode’s surface will respond to this positive charge by migrating away from the electrode; anions, on
the other hand, will migrate toward the electrode. This migration of ions occurs until the electrode’s positive surface charge and the
excess negative charge of the solution near the electrode's surface are equal. Because the movement of ions and the movement of
electrons are indistinguishable, the result is a small, short-lived non-faradaic current that we call the charging current. Every time
we change the electrode’s potential, a short-lived charging current flows.
Even in the absence of analyte, a small, measurable current flows through an electrochemical cell. This residual current has two
components: a faradaic current due to the oxidation or reduction of trace impurities and a non-faradaic charging current. Methods
for discriminating between the analyte’s faradaic current and the residual current are discussed later in this chapter.

The Electrical Double Layer

As noted in the previous section, when we apply a potential to an electrode it develops a positive or negative surface charge, the
magnitude of which is a function of the metal and the applied potential. Because the surface carries a charge, the composition of the
layer of solution immediately adjacent to the electrode changes with, for example, the concentration of cations increasing and the
concentration of anions decreasing if the electrode's surface carries a negative charge. As we move away from the electrode's
surface, the net potential first decreases in a linear manner, due to the imbalance of the cations and anions, and then in an
exponential manner until it reaches zero. This structured surface is called the electrical double layer and consists of an inner layer
and a diffuse layer. Anytime we change the potential applied to the electrode, the structure of the electrical double layer changes
and a small charging current flows.

Figure 22.1.2 : The electrical double layer. When the electrode carries a negative surface charge, the layer of solution immediately
in contact with the electrode, the inner layer, becomes enriched with cations and deficient in anions. At greater distances from the
electrode's surface, the concentration of cations and anions become more balanced—this is called the diffuse layer—eventually
reaching the bulk solution where the net charge are in balance. The potential decreases linearly through the inner layer and
exponentially through the diffuse layer.

Mass Transfer in Cells with the Passage of Current

The magnitude of a faradaic current is determined by the rate at which the analyte is oxidized at the anode or reduced at the
cathode. Two factors contribute to the rate of an electrochemical reaction: the rate at which the reactants and products are

22.1.2 https://chem.libretexts.org/@go/page/333388
transported to and from the electrode—what we call mass transport—and the rate at which electrons pass between the electrode and
the reactants and products in solution.
There are three modes of mass transport that affect the rate at which reactants and products move toward or away from the
electrode surface: diffusion, migration, and convection. Diffusion occurs whenever the concentration of an ion or a molecule at the
surface of the electrode is different from that in bulk solution. For example, if we apply a potential sufficient to completely reduce
Ag
+
at the electrode surface, the result is a concentration gradient similar to that shown in Figure 22.1.3. The region of solution
over which diffusion occurs is the diffusion layer. In the absence of other modes of mass transport, the width of the diffusion layer,
δ , increases with time as the Ag must diffuse from an increasingly greater distance.
+

Figure 22.1.3 . Concentration gradients (in red) for Ag following the application of a potential that completely reduces it to Ag(s).
+

Before we apply the potential (t = 0) the concentration of Ag is the same at all distances from the electrode’s surface. After we
+

apply the potential, its concentration at the electrode’s surface decreases to zero and Ag diffuses to the electrode from bulk
+

solution. The longer we apply the potential, the greater the distance over which diffusion occurs. The dashed red line shows the
extent of the diffusion layer at time t3. These profiles assume that convection and migration do not contribute significantly to the
mass transport of Ag .
+

Convection occurs when we mix the solution, which carries reactants toward the electrode and removes products from the
electrode. The most common form of convection is stirring the solution with a stir bar; other methods include rotating the electrode
and incorporating the electrode into a flow-cell.
The final mode of mass transport is migration, which occurs when a charged particle in solution is attracted to or repelled from an
electrode that carries a surface charge. If the electrode carries a positive charge, for example, an anion will move toward the
electrode and a cation will move toward the bulk solution. Unlike diffusion and convection, migration affects only the mass
transport of charged particles.

Schematic Representations of Cells

Although Figure 22.1.1 provides a useful picture of an electrochemical cell, it is not a convenient way to represent it. Imagine
having to draw a picture of each electrochemical cell you are using! A more useful way to describe an electrochemical cell is a
shorthand notation that uses symbols to identify different phases and that lists the composition of each phase. We use a vertical
slash (|) to identify a boundary between two phases where a potential develops, and a comma (,) to separate species in the same
phase or to identify a boundary between two phases where no potential develops. Shorthand cell notations begin with the anode and
continue to the cathode. For example, we describe the electrochemical cell in Figure 22.1.1 using the following shorthand notation.

Zn(s)| ZnCl2 (aq, a 2+ = 0.0167)|| AgNO (aq, a + = 0.100)|Ag(s)

Zn 3 Ag

The double vertical slash (||) represents the salt bridge, the contents of which we usually do not list. Note that a double vertical
slash implies that there is a potential difference between the salt bridge and each half-cell.

22.1.3 https://chem.libretexts.org/@go/page/333388
  Example 22.1.1
What are the anodic, the cathodic, and the overall reactions responsible for the potential of the electrochemical cell in Figure
22.1.4? Write the shorthand notation for the electrochemical cell.

Solution
The oxidation of Ag to Ag+ occurs at the anode, which is the left half-cell. Because the solution contains a source of Cl–, the
anodic reaction is
− −
Ag(s) + Cl (aq) ⇌  AgCl(s) + e

The cathodic reaction, which is the right half-cell, is the reduction of Fe3+ to Fe2+.
3+ − 2+
Fe (aq) + e ⇌  Fe (aq)

The overall cell reaction, therefore, is

3+ − 2+
Ag(s) +  Fe (aq) +  Cl (aq) ⇌ AgCl(s) +  Fe (aq)

The electrochemical cell’s shorthand notation is

3+
Ag(s)|HCl(aq, aCl− = 0.100), AgCl(sat’d)|| FeCl2 (aq, aFe2+ = 0.0100),  Fe (aq, aFe3+ = 0.0500)|Pt(s)

Note that the Pt cathode is an inert electrode that carries electrons to the reduction half-reaction. The electrode itself does not
undergo reduction.

Figure 22.1.4 . Potentiometric electrochemical cell for Example 22.1.1 .

This page titled 22.1: Electrochemical Cells is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

22.1.4 https://chem.libretexts.org/@go/page/333388
22.2: Potentials in Electroanalytical Cells
If a Zn(s) electrode in a solution of Zn2+(aq) in an electrochemical cell is at equilibrium, the current is zero and the potential is
fixed in value. If we change the potential from its equilibrium value, current will flow as the system moves to its new equilibrium
position. Although the initial current is quite large, it decreases over time, reaching zero when the reaction reaches equilibrium. The
current, therefore, changes in response to the applied potential. Alternatively, we can pass a fixed current through the
electrochemical cell, forcing the oxidation of Zn(s) to Zn2+(aq), effecting a change in the potential. In short, if we choose to control
the potential, then we must accept the resulting current, and we must accept the resulting potential if we choose to control the
current.

The Thermodynamics of Cell Potentials

Because a redox reaction involves a transfer of electrons from a reducing agent to an oxidizing agent, it is convenient to consider
the reaction’s thermodynamics in terms of the electron. For a reaction in which one mole of a reactant undergoes oxidation or
reduction, the net transfer of charge, q, in coulombs is
q = nF (22.2.1)

where n is the moles of electrons per mole of reactant, and F is Faraday’s constant (96485 C/mol). The free energy, ∆G, to move
this charge given an applied potential, E, is
ΔG = Eq (22.2.2)

The change in free energy (in kJ/mole) for a redox reaction, therefore, is
ΔG = −nF E (22.2.3)

where ∆G has units of kJ/mol. The minus sign in Equation 22.2.3 is the result of a different convention for assigning a reaction’s
favorable direction. In thermodynamics, a reaction is favored when ∆G is negative, but a redox reaction is favored when E is
positive. Substituting Equation 22.2.3 into the thermodynamic equation that relates the free energy to its standard state value
∘
ΔG = ΔG + RT ln Qr (22.2.4)

gives
∘
−nF E = −nF E + RT ln Qr (22.2.5)

Dividing by –nF leads to the Nernst equation

RT
∘
E =E − ln Qr (22.2.6)
nF

where Eo is the potential under standard‐state conditions (more on this in Section 22.3). Substituting appropriate values for R and
F, assuming a temperature of 25 oC (298 K), and switching from the natural logarithm (ln) to the base 10 logarithm (log) gives the
potential in volts as

o
0.05916
E =E − log Qr (22.2.7)
n

The term Q in the previous equations is the reaction quotient, which has the same mathematical form as the reaction's equilibrium
r

constant expression, but uses the instantaneous amounts of reactants and products in place of their equilibrium values. For the cell
in Figure 22.1.1, for example, the overall reaction is
+ 2+
Zn(s) + 2 Ag (aq) ⇌ 2Ag(s) + Zn (aq) (22.2.8)

and Equation 22.2.7 becomes

2 +
0.05916 [ Zn ]
∘
E =E − log (22.2.9)
+ 2
2
[ Ag ]

Equation 22.2.9 shows us how the potential changes as the concentrations of Zn2+ and Ag+ change.

22.2.1 https://chem.libretexts.org/@go/page/333390
 As we will see in 22.3, 22.2.9 actually is expressed in terms of activities instead of concentrations. The appendix in Chapter
35.7 explains what activity is, why it is important to make a distinction between activity and concentration, and when it is
reasonable to use concentrations in place of activities.

Liquid Junction Potentials

A junction potential develops at the interface between two ionic solutions if there is a difference in the concentration and the
mobility of the ions. Consider, for example, a porous membrane that separates a solution of 0.1 M HCl from a solution of 0.01 M
HCl (Figure 22.2.1a). Because the concentration of HCl on the membrane’s left side is greater than that on the right side of the
membrane, H+ and Cl– will diffuse in the direction of the arrows. The mobility of H+, however, is greater than that for Cl–, as
shown by the difference in the lengths of their respective arrows. Because of this difference in mobility, the solution on the right
side of the membrane develops an excess concentration of H+ and a positive charge (Figure 22.2.1b). Simultaneously, the solution
on the membrane’s left side develops a negative charge because there is an excess concentration of Cl–. We call this difference in
potential across the membrane a junction potential and represent it as Ej.

Figure 22.2.1 . Origin of the junction potential between a solution of 0.1 M HCl and a solution of 0.01 M HCl.
The magnitude of a junction potential depends upon the difference in the concentration of ions on the two sides of the interface,
and may be as large as 30–40 mV. For example, a junction potential of 33.09 mV has been measured at the interface between
solutions of 0.1 M HCl and 0.1 M NaCl [Sawyer, D. T.; Roberts, J. L., Jr. Experimental Electrochemistry for Chemists, Wiley-
Interscience: New York, 1974, p. 22]. A salt bridge’s junction potential is minimized by using a salt, such as KCl, for which the
mobilities of the cation and anion are approximately equal. We also can minimize the junction potential by incorporating a high
concentration of the salt in the salt bridge. For this reason salt bridges frequently are constructed using solutions that are saturated
with KCl. Nevertheless, a small junction potential, generally of unknown magnitude, is always present.

This page titled 22.2: Potentials in Electroanalytical Cells is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

22.2.2 https://chem.libretexts.org/@go/page/333390
22.3: Electrode Potentials
We began this chapter by examining the electrochemical cell in Figure 22.1.1 where Zn(s) is oxidized to Zn 2 +
(aq) and Ag
+
(aq)

is reduced to Ag(s), as shown by the following reaction.

2+ +
2Ag(s) + Zn (aq) ⇌ Zn(s) + 2 Ag (aq)

The reaction proceeds as written because the reduction of Ag+(aq) to Ag(s)

+ −
Ag (aq) + e ⇌ Ag(s) (22.3.1)

is more thermodynamically favorable than the reduction of Zn 2 +

(aq) to Zn(s)
2+ −
 Zn (aq) + 2 e ⇌ Zn(s) (22.3.2)

But, how do we know this is true? In this section we answer this question by taking a close look at electrode potentials.

Nature of Electrode Potentials

The potential of an electrochemical cell is the difference between the potential at the cathode, E , and the potential at the
cathode

anode, E anode, where both potentials are defined in terms of a reduction reaction (and are called reduction potentials); thus
Ecell = Ecathode − Eanode (22.3.3)

+ −
1
H (aq) + e = H2 (g) (22.3.4)
2

which is the reaction that defines the standard hydrogen electrode, or SHE.

The Standard Hydrogen Electrode (SHE)

The SHE consists of a Pt electrode immersed in a solution in which the activity of hydrogen ion is 1.00 and in which the partial
pressure of H2(g) is 1.00 atm (Figure 22.3.1). A conventional salt bridge connects the SHE to the indicator half-cell. The short
hand notation for the standard hydrogen electrode is
+
Pt(s),  H 2 (g, fH2 = 1.00) |  H (aq, aH+ = 1.00) ∥ (22.3.5)

and the standard-state potential for the reaction 22.3.4 is, by definition, 0.000 V at all temperatures.

Figure 22.3.1 . Schematic diagram showing the standard hydrogen electrode.

Practical Reference Electrodes

Although the standard hydrogen electrode is the standard against which all other potentials are referenced, it is not practical for
routine use as it is difficult to prepare and maintain. Instead, we use one of several other reference electrodes. The two most
common of these alternative reference electrodes are the calomel, or Hg/Hg2Cl2 electrode, which is based on the following redox
couple between Hg2Cl2 and Hg (calomel is the common name for Hg2Cl2)
− −
Hg Cl2 (s) + 2 e ⇌ 2Hg(l) + 2 Cl (aq)
2

and the Ag/AgCl reference electrode, which is based on the reduction of AgCl to Ag
− −
AgCl(s) + e ⇌ Ag(s) + Cl (aq)

22.3.1 https://chem.libretexts.org/@go/page/333392
A more detailed examination of these two reference electrodes is found in Chapter 23.1.

Definition of Electrode Potential

To determine the potential for the reduction of Zn2+(aq) to Zn(s) we make it the cathode in the following electrochemical cell
+ 2 +
Pt(s),  H 2 (g, PH2 = 1.00) |  H (aq, aH+ = 1.00) ∥ Zn (aq, a 2+ = x) |Zn(s) (22.3.6)
Zn

where x is the activity of Zn2+ in its half-cell. For example, when a 2+

Zn
= 1.00 , the potential of the electrochemical cell is
−0.763V. If we find that the potential for the electrochemical cell

2 + +
Zn(s)| Zn (aq, aZn2+ = 1.00)∥ Ag (aq, aAg+ = 1.00)|Ag(s)

is +1.562 V, then knowing that

Ecell = EAg+ /Ag − EZn2 + /Zn = EAg+ /Ag − (−0.763V)

gives E Ag
+
/Ag
= +0.799 . In this way, we can build tables of potentials for individual half-reactions.

Sign Convention for Electrode Potentials

In Section 22.2 we noted the following relationship between an electrochemical potential, E , and the Gibbs free energy, ΔG

ΔG = −nF E (22.3.7)

which tells us that a positive potential corresponds to a thermodynamically favorable reaction. Knowing that the potential for the
electrochemical cell in Equation 22.3.6 is −0.763V tells us that the reduction of Zn2+(aq) to Zn(s) is not thermodynamically
favorable relative to the reduction of H+(aq) to H2(g); that is, we do not expect the reaction
2 + +
Zn (aq) + H (g) ⇌ 2 H (aq) + Zn(s) (22.3.8)
2

to occur; however, with a potential of +0.799 V, we do expect the reaction

+ +
2 Ag (aq) + H (g) ⇌ 2 H (aq) + 2Ag(s) (22.3.9)
2

to occur. Or, looking at this another way, we expect that Zn(s), but not Ag(s), will dissolve in acid.

Effect of Activity on Electrode Potentials

In Chapter 22.2 we wrote the Nernst equation for the reaction
+ 2+
Zn(s) + 2 Ag (aq) ⇌ 2Ag(s) + Zn (aq) (22.3.10)

in terms of the concentrations of Zn2+(aq) and Ag+(aq)

2 +
0.05916 [ Zn ]
∘
E =E − log (22.3.11)
2
2 +
[ Ag ]

Although there are times when we will write the Nernst equation in terms of concentrations, thermodynamic functions are more
correctly written in terms of the activities of ions. Under ideal conditions, individual ions and molecules of gases behave as
independent particles. When this is true, then an ion's activity and concentration are equal and we can write the Nernst equation
using concentrations; under other conditions, then the Nernst equation is more correctly written in terms of activities
0.05916 a 2 +
∘ Zn
E =E − log (22.3.12)
2
2
(aAg+ )

where a Zn
and a
2 + are the activities of Zn2+ and Ag+. Equation 22.3.12 shows us how the potential changes as the activities of
Ag
+

Zn2+ and Ag+ change.

If you are not familiar with activity, or need a reminder on the relationship between activity and concentration, then see the
appendix in Chapter 35.7, which explains what activity is, why it is important to make a distinction between activity and
concentration, and when it is reasonable to use concentrations in place of activities.

22.3.2 https://chem.libretexts.org/@go/page/333392
The Standard Electrode Potential E ∘

The standard electrode potential, E , for a half-reaction is the potential when all species are present at unit activity or, for gases,
∘

unit fugacity. Its value is independent of how we choose to write the half-reaction; that is, the standard state potential for the
reduction of Ag+(aq) to Ag(s), which is the cathode in the electrochemical cell in Figure 22.1.1, is +0.799 V whether we write the
half-reacation as
+ −
Ag (aq) + e ⇌ Ag(s) (22.3.13)

or as
+ −
2 Ag (aq) + e ⇌ 2Ag(s) (22.3.14)

At first glance, this seems counterintuitive; however, if we calculate the potential when the activity of Ag+ is 0.50 we get
0.05916 1 0.05916 1
∘
E =E − log = 0.799 − log = 0.781V
1 a + 1 0.50
Ag

when using reaction 22.3.13, and

∘
0.05916 1 0.05916 1
E =E − log = 0.799 − log = 0.781V
2 2
2 (aAg + ) 2 0.50

The appendix in Chapter 35.8 provides a table of standard state reduction potentials for a wide variety of half-reactions at 298 K.

Some Limitations to the Use of Standard Electrode Potentials

Although standard electrode potentials are valuable, they are several important limitations to their use, which we outline here.

Substitution of Concentration for Activities

One important limitation is that that the Nernst equation is defined in terms of the activity of ions instead of their concentrations.
Although it is easy to prepare a solution for which the concentration of Na+ is 0.100 M using NaCl—just weigh out 5.844 g of
NaCl and dissolve in 1.00 L of water—it is much more challenging to prepare a solution for which the activity of Na+ is 0.100. For
this reason, in calculations we usually substitute concentrations for activities when using the Nernst equation. This simplification
generally is okay for dilute solutions where the difference between activities and concentrations are small.

Effect of Other Equilibrium Reactions

A standard state potential tells us about the equilibrium position of a redox half-reaction reaction under standard state conditions. If
one or more of the species in the half-reaction are involved in other equilibrium reactions, then these reactions will affect the value
of the standard potential. For example, Fe2+ and Fe3+ form a variety of metal-ligand complexes with Cl– which explains why
E
∘

Fe
3 +
is 0.771 in the absence of chloride ion, but is 0.70 in 1 M HCl.
/ Fe
2 +

Formal Potentials
One way to compensate for using concentrations and partial pressures in place of activities and fugacities, and to compensate for
other equilibrium reactions, is to replace the standard state potentials, E with a formal potential, E , that is measured using
∘ ∘′

concentrations of 1.00 for ions, partial pressures of 1.00 for gases, and for a specific concentration of other reagents. The table
below, which is adapted from the appendix in Chapter 35.8, provides formal potentials for Fe3+/Fe2+ half-reaction in five different
solvents.

iron E
∘
(V) E
∘′
(V)

0.70 in 1 M HCl
0.767 in 1 M HClO 4

Fe
3 +
+e
−
⇌ Fe
2 +
0.771 0.746 in 1 M HNO 3

0.68 in 1 M H SO 2 4

0.44 in 0.3 M H PO 3 4

22.3.3 https://chem.libretexts.org/@go/page/333392
Reaction Rates
The reduction of Fe3+ to Fe2+ consumes an electron, which is drawn from the electrode. The oxidation of another species, perhaps
the solvent, at a second electrode is the source of this electron. Because the reduction of Fe3+ to Fe2+ consumes one electron, the
flow of electrons between the electrodes—in other words, the current—is a measure of the rate at which Fe3+ is reduced. One
important consequence of this observation is that the current is zero when the reaction Fe (aq) ⇌  Fe (aq) + e is at
3+ 2+ −

equilibrium. If redox half-reaction cannot maintain an equilibrium because the reaction in one direction is too slow, then we cannot
measure a meaningful standard state potential.

This page titled 22.3: Electrode Potentials is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

22.3.4 https://chem.libretexts.org/@go/page/333392
22.4: Calculation of Cell Potentials from Electrode Potentials
The potential of an electrochemical cell is the difference between the electrode potentials of the cathode and the anode
Ecell = Ecathode − Eanode (22.4.1)

where E cathode and Eanode are both reduction potentials. Given a set of conditions, we can use the Nernst equation to calculate the
cell potential, as shown by the following example.

 Example 22.4.1

Calculate (a) the standard state potential and (b) the potential when [Ag+] = 0.0020 M and [Cd2+] = 0.0050 M, for the
following reaction at 25oC.
+ 2+
Cd(s) + 2 Ag (aq) ⇌ 2Ag(s) + Cd (aq)

For part (b), calculate the potential twice, once using concentrations and once using activities assuming that the solution's ionic
strength is 0.100.

Solution
(a) In this reaction Cd is oxidized at the anode and Ag+ is reduced at the cathode. Using standard state electrode potentials from
Appendix 3, we find that the standard state potential is
∘ ∘ ∘
E =E +
−E 2+
= 0.7996 − (−0.4030) = 1.2026 V
Ag /Ag Cd /Cd

(b) To calculate the potential when [Ag+] is 0.0020 M and [Cd2+] is 0.0050 M, we use the appropriate relationship for the
reaction quotient, Qr, when writing the Nernst equation
2+
0.05916 V [ Cd ]
∘
E =E − log
+ 2
n
[ Ag ]

0.05916 V 0.050
E = 1.2026 V − log = 1.14 V
2
2 (0.020)

To calculate the potential using activities, we first calculate the activity coefficients for Cd2+ and Ag+. Following the approach
outlined in the appendix in Chapter 35.7 gives
2
−−−−
−0.51 × (+2 ) × √0.100
log γ 2 + = −−−− = −0.2078
Cd
1 + 3.3 × 0.50 × √0.100

2
−−−−
−0.51 × (+1 ) × √0.100
]logγ + = −−−− = −0.1279
Ag
1 + 3.3 × 0.25 × √0.100

2+
a 2 + =γ 2 + × [ Cd ] = 0.6197 × 0.0050 = 0.003098
Cd Cd

+
aAg+ = γAg+ × [ Ag ] = 7449 × 0.0020 = 0.00149

Finally, we substitute activities for concentrations in the Nernst equation to arrive at a potential of
0.05916 V 0.003098
E = 1.2026 V − log = 1.11 V
2 (0.00149)2

This page titled 22.4: Calculation of Cell Potentials from Electrode Potentials is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

22.4.1 https://chem.libretexts.org/@go/page/333394
22.5: Currents in Electrochemical Cells
Most electrochemical techniques rely on either controlling the current and measuring the resulting potential, or controlling the
potential and measuring the resulting current; only potentiometry (see Chapter 23) measures a potential under conditions where
there is essentially no current. Understanding the relationship between current, i, and potential, E, is important. Although we
learned in Sections 22.3 and 22.4 how to calculated electrode potentials and cell potentials using the Nernst equation, the
experimentally measured potentials may differ from their thermodynamic values for a variety of reasons that we outline here.

iR drop
The movement of an electrical charge in an electrochemical cell generates a potential, E defined by Ohm's law
ir

Eir = iR (22.5.1)

where i is the current and R is the solution's resistance. To account for this, we can include an additional term to the equation for the
electrochemical cell's potential
Ecell = Ecathode − Eanode − Eir = ENernst − iR (22.5.2)

where E Nernst is the potential from the Nernst equation. The resulting decease in the potential from its idealized value is called the
iR drop.

Polarization
Equation 22.5.2 indicates that we expect a linear relationship between an electrochemical cell's potential, E . When this is not
cell

the case, the electrochemical cell is said to be polarized. There are several sources that contribute to polarization, which we
consider in this section; first, however, we define ideal polarized and nonpolarized electrodes.

Ideal Polarized and Nonpolarized Electrodes and Electrochemical Cells

An ideal polarized electrode is one in which a change in potential over a fairly wide range has no effect on the current that flows
through the electrode, as we see in Figure 22.5.1a for the range of potentials defined by the solid green line. Such electrodes are
useful because they do not themselves undergo oxidation or reduction—they are electrochemically inert—which makes them a
good choice for studying the electrochemical behavior of other species.
An ideal nonpolarized electrode is one in which a change in current has no effect on the electrode's potential, as we see in Figure
22.5.1b between the limits defined by the solid red line with deviations shown by the dashed red line. Such electrodes are useful

because the provided a stable potential against which we can reference the redox potential of other species.

22.5.1 https://chem.libretexts.org/@go/page/333396
 Figure 22.5.1 : Current-potential curves for (a) an ideal polarized electrode and for (b) an ideal nonpolarized electrode. For both, the
solid line shows the range of (a) potentials or (b) currents where the electrode shows ideal behavior.

Overpotential
The magnitude of polarization when drawing a current is called the overpotential, η and expressed as the difference between the
applied potential, E, and the potential from the Nernst equation.

η = E − ENernst (22.5.3)

The overpotential can be subdivided into a variety of sources, a few of which are discussed below.

Concentration Polarization
The reduction of Fe3+ to Fe2+ in an electrochemical cell consumes an electron, which is drawn from the electrode. The oxidation of
another species, perhaps the solvent, at a second electrode is the source of this electron. Because the reduction of Fe3+ to Fe2+
consumes one electron, the flow of electrons between the electrodes—in other words, the current—is a measure of the rate at which
Fe3+ is reduced.

The rate of the reaction Fe 3+

(aq) ⇌  Fe
2+
(aq) + e
−
is the change in the concentration of Fe3+ as a function of time.

In order for the reduction of Fe3+ to Fe2+ to take place, Fe3+ must move from the bulk solution into the layer of solution
immediately adjacent to the electrode and then diffuse to the electrode's surface; this is called the diffusion layer. Once the
reduction takes place, the Fe2+ produced must diffuse away from the electrode's surface and enter into the bulk solution. These two
processes are called mass transfer and if we try to change the electrode's potential too quickly, mass transfer may result in
concentrations of Fe3+ to Fe2+ at the electrode's surface that are different from that in bulk solution, resulting in concentration
polarization.
Let's use the reduction of Fe3+ to Fe2+ at the cathode of a galvanic cell to think though how concentration polarization affects the
potential we measure. From the Nernst equation we know that
2 + 2 +
0.05916 [ Fe ] 0.05916 [ Fe ]
∘
E =E 3 + 2 +
− log = +0.771 − log (22.5.4)
Fe / Fe 3 + 3 +
1 [ Fe ] 1 [ Fe ]

If the mass transfer of Fe3+ from bulk solution to the electrode's surface is slow and if mass transfer of Fe2+ from the electrode's
surface to bulk solution is slow, then the concentration of Fe3+ at the electrode's surface is smaller than in bulk solution and the
2 +

concentration of Fe2+ at the electrode's surface is greater than in the bulk solution. As a result, the ratio
[ Fe ]

3 +
is greater than that
[ Fe ]

22.5.2 https://chem.libretexts.org/@go/page/333396
predicted by the bulk concentrations of Fe3+ and Fe2+ and the potential of the cathode is smaller (less positive) than the value
+0.771 V predicted by the bulk concentrations of Fe3+ and Fe2+. The resulting potential of the electrochemical cell
Ecell = Ecathode − Eanode (22.5.5)

is less positive than that predicted by the bulk concentrations of Fe3+ and Fe2+ due to this concentration polarization.
Other kinetic processes can contribute to polarization, including the rate of chemical reactions that take place within the layer of
solution near the electrode's surface, the kinetics of reactions in which the electroactive species absorb or desorb from the
electrode's surface, and the kinetics of the electron transfer process itself. More details on these are included in later chapters
covering specific electrochemical techniques.

This page titled 22.5: Currents in Electrochemical Cells is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

22.5.3 https://chem.libretexts.org/@go/page/333396
22.6: Types of Electroanalytical Methods
In the next three chapters we will consider a variety of different interfacial electrochemical experiments; that is, experiments in
which the redox reaction takes place at the surface of an electrode. Because electrochemistry is such a broad field, let’s use Figure
22.6.1 to organize these techniques by the experimental conditions we choose to use (Do we control the potential or the current?

How do we change the applied potential or applied current? Do we stir the solution?) and the analytical signal we decide to
measure (Current? Potential?).
At the first level, we divide electrochemical techniques into static techniques and dynamic techniques. In a static technique we do
not allow current to pass through the electrochemical cell and, as a result, the concentrations of all species remain constant.
Potentiometry, in which we measure the potential of an electrochemical cell under static conditions, is one of the most important
quantitative electrochemical methods and is discussed in Chapter 23.
Dynamic techniques, in which we allow current to flow and force a change in the concentration of species in the electrochemical
cell, comprise the largest group of interfacial electrochemical techniques. Coulometry, in which we measure current as a function of
time, is covered Chapter 24. Voltammetry and amperometry, in which we measure current as a function of a fixed or variable
potential, are the subjects of Chapter 25.

Figure 22.6.1 : Family tree that highlights the similarities and differences between a number of interfacial electrochemical
techniques. The specific instrumental methods are shown in red, the experimental conditions are shown in blue, and the analytical
signals are in green.

This page titled 22.6: Types of Electroanalytical Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

22.6.1 https://chem.libretexts.org/@go/page/333398
 CHAPTER OVERVIEW
23: Potentiometry
In potentiometry we measure the potential of an electrochemical cell under static conditions. Because no current—or only a
negligible current—flows through the electrochemical cell, its composition remains unchanged. For this reason, potentiometry is a
useful quantitative method of analysis. The first quantitative potentiometric applications appeared soon after the formulation, in
1889, of the Nernst equation, which relates an electrochemical cell’s potential to the concentration of electroactive species in the
cell [Stork, J. T. Anal. Chem. 1993, 65, 344A–351A].
Potentiometry initially was restricted to redox equilibria at metallic electrodes, which limited its application to a few ions. In 1906,
Cremer discovered that the potential difference across a thin glass membrane is a function of pH when opposite sides of the
membrane are in contact with solutions that have different concentrations of H3O+. This discovery led to the development of the
glass pH electrode in 1909. Other types of membranes also yield useful potentials. For example, in 1937 Kolthoff and Sanders
showed that a pellet of AgCl can be used to determine the concentration of Ag+. Electrodes based on membrane potentials are
called ion-selective electrodes, and their continued development extends potentiometry to a diverse array of analytes.
23.1: Reference Electrodes
23.2: Metallic Indicator Electrodes
23.3: Membrane Ion-Selective Electrodes
23.4: Molecular-Selective Electrode Systems
23.5: Instruments for Measuring Cell Potentials
23.6: Quantitative Potentiometry

This page titled 23: Potentiometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

1
23.1: Reference Electrodes
In potentiometry we measure the difference between the potential of two electrodes. The potential of one electrode—the working
or indicator electrode—responds to the analyte’s activity and the other electrode—the counter or reference electrode—has a known,
fixed potential. By convention, the reference electrode is the anode; thus, the short hand notation for a potentiometric
electrochemical cell is
reference electrode || indicator electrode
and the cell potential is

Ecell = Eind − Eref

The ideal reference electrode provides a stable, known potential so that we can attribute any change in Ecell to the analyte’s effect
on the indicator electrode’s potential. In addition, the reference electrode should be easy to make and easy to use. Although the
standard hydrogen electrode is the reference electrode used to define electrode potentials, it use is not common. Instead, the two
reference electrodes discussed in this section find the most applications.

Calomel Electrodes
A calomel reference electrode is based on the following redox couple between Hg2Cl2 and Hg (calomel is the common name for
Hg2Cl2)
− −
Hg2 Cl2 (s) + 2 e ⇌ 2Hg(l) + 2 Cl (aq)

for which the potential is

0.05916 2
0.05916 2
o
E =E − log (aCl− ) = +0.2682V − log (aCl− )
Hg2 Cl2 /Hg
2 2

The potential of a calomel electrode, therefore, depends on the activity of Cl– in equilibrium with Hg and Hg2Cl2.
As shown in Figure 23.1.1, in a saturated calomel electrode (SCE) the concentration of Cl– is determined by the solubility of KCl.
The electrode consists of an inner tube packed with a paste of Hg, Hg2Cl2, and KCl, situated within a second tube that contains a
saturated solution of KCl. A small hole connects the two tubes and a porous wick serves as a salt bridge to the solution in which the
SCE is immersed. A stopper in the outer tube provides an opening for adding addition saturated KCl. The short hand notation for
this cell is

Hg(l)| Hg2 Cl2 (s), KCl(aq,  sat'd )∥

Because the concentration of Cl– is fixed by the solubility of KCl, the potential of an SCE remains constant even if we lose some of
the inner solution to evaporation. A significant disadvantage of the SCE is that the solubility of KCl is sensitive to a change in
temperature. At higher temperatures the solubility of KCl increases and the electrode’s potential decreases. For example, the
potential of the SCE is +0.2444 V at 25oC and +0.2376 V at 35oC. The potential of a calomel electrode that contains an unsaturated
solution of KCl is less dependent on the temperature, but its potential changes if the concentration, and thus the activity of Cl–,
increases due to evaporation.

For example, the potential of a calomel electrode is +0.280 V when the concentration of KCl is 1.00 M and +0.336 V when the
concentration of KCl is 0.100 M. If the activity of Cl– is 1.00, the potential is +0.2682 V.

23.1.1 https://chem.libretexts.org/@go/page/333586
 Figure 23.1.1 . Schematic diagram showing the saturated calomel electrode.

Silver/Silver Chloride Electrodes

Another common reference electrode is the silver/silver chloride electrode, which is based on the reduction of AgCl to Ag.
− −
AgCl(s) + e ⇌ Ag(s) + Cl (aq)

As is the case for the calomel electrode, the activity of Cl– determines the potential of the Ag/AgCl electrode; thus
∘
E =E − 0.05916 log aCl− = 0.2223 V − 0.05916 log aCl−
AgCl/Ag

When prepared using a saturated solution of KCl, the electrode's potential is +0.197 V at 25oC. Another common Ag/AgCl
electrode uses a solution of 3.5 M KCl and has a potential of +0.205 V at 25oC. As you might expect, the potential of a Ag/AgCl
electrode using a saturated solution of KCl is more sensitive to a change in temperature than an electrode that uses an unsaturated
solution of KCl.
A typical Ag/AgCl electrode is shown in Figure 23.1.2 and consists of a silver wire, the end of which is coated with a thin film of
AgCl, immersed in a solution that contains the desired concentration of KCl. A porous plug serves as the salt bridge. The
electrode’s short hand notation is

Ag(s)| Ag Cl(s), KCl (aq, aCl− = x) ∥

Figure 23.1.2 . Schematic diagram showing a Ag/AgCl electrode. Because the electrode does not contain solid KCl, this is an
example of an unsaturated Ag/AgCl electrode.

Converting Potentials Between Reference Electrodes

The standard state reduction potentials in most tables are reported relative to the standard hydrogen electrode’s potential of +0.00
V. Because we rarely use the SHE as a reference electrode, we need to convert an indicator electrode’s potential to its equivalent
value when using a different reference electrode. As shown in the following example, this is easy to do.

23.1.2 https://chem.libretexts.org/@go/page/333586
  Example 23.1.1

The potential for an Fe3+/Fe2+ half-cell is +0.750 V relative to the standard hydrogen electrode. What is its potential if we use a
saturated calomel electrode or a saturated silver/silver chloride electrode?

Solution
When we use a standard hydrogen electrode the potential of the electrochemical cell is

Ecell = E 3+ 2+ − ESHE = 0.750 V − 0.000 V = 0.750 V

Fe / Fe

We can use the same equation to calculate the potential if we use a saturated calomel electrode

Ecell = E 3+ 2+ − ESHE = 0.750 V − 0.2444 V = 0.506 V

Fe / Fe

or a saturated silver/silver chloride electrode

Ecell = EFe3+ / Fe2+ − ESHE = 0.750 V − 0.197 V = 0.553 V

Figure 23.1.3 provides a pictorial representation of the relationship between these different potentials.

Figure 23.1.3 . Relationship between the potential of an Fe3+/Fe2+ half-cell relative to the reference electrodes in Example 23.1.1 .
The potential relative to a standard hydrogen electrode is shown in blue, the potential relative to a saturated silver/silver chloride
electrode is shown in red, and the potential relative to a saturated calomel electrode is shown in green.

This page titled 23.1: Reference Electrodes is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

23.1.3 https://chem.libretexts.org/@go/page/333586
23.2: Metallic Indicator Electrodes
In potentiometry, the potential of the indicator electrode is proportional to the analyte’s activity. Two classes of indicator electrodes
are used to make potentiometric measurements: metallic electrodes, which are the subject of this section, and ion-selective
electrodes, which are covered in the next section.

Electrodes of the First Kind

If we place a copper electrode in a solution that contains Cu2+, the electrode’s potential due to the reaction
2+ −
Cu (aq) + 2 e ⇌ Cu(s)

is determined by the activity of Cu2+.

0.05916 1 0.05916 1
o
E =E 2+
− log = +0.3419V − log
Cu /Cu 2 a 2+ 2 a 2+
Cu Cu

If copper is the indicator electrode in a potentiometric electrochemical cell that also includes a saturated calomel reference
electrode
2+
SCE∥ Cu (aq, aCu2+ = x) |Cu(s)

then we can use the cell potential to determine an unknown activity of Cu2+ in the indicator electrode’s half-cell
0.05916 1
Ecell = E ind  − ESCE  = +0.3419V − log − 0.2224V
2 a 2+
Cu

An indicator electrode in which the metal is in contact with a solution containing its ion is called an electrode of the first kind. In
general, if a metal, M, is in a solution of Mn+, the cell potential is
0.05916 1 0.05916
Ecall = K − log =K+ log aM n+
n aM n+ n

where K is a constant that includes the standard-state potential for the Mn+/M redox couple and the potential of the reference
electrode.
For a variety of reasons—including the slow kinetics of electron transfer at the metal–solution interface, the formation of metal
oxides on the electrode’s surface, and interfering reactions—electrodes of the first kind are limited to the following metals: Ag, Bi,
Cd, Cu, Hg, Pb, Sn, Tl, and Zn.

Many of these electrodes, such as Zn, cannot be used in acidic solutions because they are easily oxidized by H+.
+ 2+
Zn(s) + 2 H (aq) ⇌  H 2 (g) + Zn (aq)

Electrodes of the Second Kind

The potential of an electrode of the first kind responds to the activity of Mn+. We also can use this electrode to determine the
activity of another species if it is in equilibrium with Mn+. For example, the potential of a Ag electrode in a solution of Ag+ is
E = 0.7996V + 0.05916 log aAg+ (23.2.1)

If we saturate the indicator electrode’s half-cell with AgI, the solubility reaction
+ −
Agl(s) ⇌ Ag (aq) + I (aq) (23.2.2)

determines the concentration of Ag+; thus

Ksp,Agl
aAg+ = (23.2.3)
aI−

where Ksp,AgI is the solubility product for AgI. Substituting Equation 23.2.3 into Equation 23.2.1

23.2.1 https://chem.libretexts.org/@go/page/333588
 Ksp, Agl
E = 0.7996 V + 0.05916 log (23.2.4)
a −
I

shows that the potential of the silver electrode is a function of the activity of I–. If we incorporate this electrode into a
potentiometric electrochemical cell with a saturated calomel electrode
−
SCE∥AgI(s),  I (aq, aI− = x) |Ag(s) (23.2.5)

then the cell potential is

Ecell = K − 0.05916 log a − (23.2.6)
I

+
where K is a constant that includes the standard-state potential for the Ag /Ag redox couple, the solubility product for AgI, and the
reference electrode’s potential.
If an electrode of the first kind responds to the activity of an ion in equilibrium with Mn+, we call it an electrode of the second kind.
Two common electrodes of the second kind are the calomel and the silver/silver chloride reference electrodes.

In an electrode of the second kind we link together a redox reaction and another reaction, such as a solubility reaction. You
might wonder if we can link together more than two reactions. The short answer is yes. An electrode of the third kind, for
example, links together a redox reaction and two other reactions. Such electrodes are less common and we will not consider
them in this text.

Metallic Redox Electrodes

An electrode of the first kind or the second kind develops a potential as the result of a redox reaction that involves the metallic
electrode. An electrode also can serve as a source of electrons or as a sink for electrons in an unrelated redox reaction, in which
case we call it a redox electrode. The Pt cathode in 23.2.1 is a redox electrode because its potential is determined by the activity of
Fe2+ and Fe3+ in the indicator half-cell. Note that a redox electrode’s potential often responds to the activity of more than one ion,
which limits its usefulness for direct potentiometry.

Figure 23.2.1. Potentiometric electrochemical cell in which the anode is a metallic electrode of the first kind (Ag) and the cathode
is a metallic redox electrode (Pt).

This page titled 23.2: Metallic Indicator Electrodes is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

23.2.2 https://chem.libretexts.org/@go/page/333588
23.3: Membrane Ion-Selective Electrodes
If metals were the only useful materials for constructing indicator electrodes, then there would be few useful applications of
potentiometry. In 1906, Cremer discovered that the potential difference across a thin glass membrane is a function of pH when
opposite sides of the membrane are in contact with solutions that have different concentrations of H3O+. The existence of this
membrane potential led to the development of a new class of indicator electrodes, which we call ion-selective electrodes (ISEs). In
addition to the glass pH electrode, ion-selective electrodes are available for a wide range of ions. It also is possible to construct a
membrane electrode for a neutral analyte by using a chemical reaction to generate an ion that is monitored with an ion-selective
electrode. The development of new ion-selective membrane electrodes continues to be an active area of research.

Classification of Ions-Selective Membranes

There are two broad classes of materials that are used as membranes: crystalline solid-state membranes and non-crystalline
membranes. Examples of crystalline solid-state membranes are single crystals of LaF3 and polycrystalline AgS. Examples non-
crystalline membranes are glass and hydrophobic membranes that hold a liquid ion-exchanger. Each of these are considered below.

Properties of Ion-Selective Membranes

To be useful, an ion-selective membrane must be structurally stable (a crystalline membrane that is soluble, for example, is not
structurally stable), capable of being machined to a suitable size and shape that can be incorporated into the indicator electrode in a
potentiometric electrochemical cell, electrically conductive so it is possible to measure the electrochemical cell's potential, and
selective toward the analyte.

The Membrane's Boundary Potential

Figure 23.3.1 shows a typical potentiometric electrochemical cell equipped with an ion-selective electrode. The short hand notation
for this cell is

 ref (sample)  ∥
∥A samp  (aq, aA = x) |membrane| A int  (aq, aA = y)∥
∥  ref (internal) 
s am p int

where the ion-selective membrane separates the two solutions that contain analyte with activities of x and y: the sample solution
and the ion-selective electrode’s internal solution. The potential of this electrochemical cell
Ecell = Eref(int) − Eref(samp) + Emem (23.3.1)

includes the potential of each reference electrode and the difference in potential across the membrane, Emem, which is the
membrane's boundary potential. The notations ref(sample) and ref(internal) represent a reference electrode immersed in the sample
and a reference electrode immersed in the ion-selective electrode's internal solution. Because the potential of the two reference
electrodes are constant, any change in Ecell reflects a change in the membrane’s boundary potential.

Figure 23.3.1 . Schematic diagram that shows a typical potentiometric cell with an ion-selective electrode. The ion-selective
electrode’s membrane separates the sample, which contains the analyte at an activity of (aA)samp, from an internal solution that
contains the analyte with an activity of (aA)int.
The analyte’s interaction with the membrane generates a boundary potential if there is a difference in its activity on the membrane’s
two sides. Current is carried through the membrane by the movement of either the analyte or an ion already present in the
membrane’s matrix. The membrane potential is given by the following Nernst-like equation

23.3.1 https://chem.libretexts.org/@go/page/333590
 RT (aA )
int
Emem = Easym − ln (23.3.2)
zF (aA )samp

where (aA)samp is the analyte’s activity in the sample, (aA)int is the analyte’s activity in the ion-selective electrode’s internal solution,
and z is the analyte’s charge. Ideally, Emem is zero when (aA)int = (aA)samp. The term Easym, which is an asymmetry potential,
accounts for the fact that Emem usually is not zero under these conditions.

For now we simply note that a difference in the analyte’s activity results in the membrane's boundary potential. As we consider
different types of ion-selective electrodes, we will explore more specifically the source of the membrane potential.

Substituting Equation 23.3.2 into Equation 23.3.1, assuming a temperature of 25oC, and rearranging gives
0.05916
Ecell = K + log (aA ) (23.3.3)
samp
z

where K is a constant that includes the potentials of the two reference electrodes, the asymmetry potential, and the analyte's activity
in the internal solution. Equation 23.3.3 is a general equation and applies to all types of ion-selective electrodes.

Membrane Selectivity
The membrane's boundary potential results from a chemical interaction between the analyte and active sites on the membrane’s
surface. Because the signal depends on a chemical process, most membranes are not selective toward a single analyte. Instead, the
membrane's boundary potential is proportional to the concentration of each ion that interacts with the membrane’s active sites. We
can rewrite Equation 23.3.3 to include the contribution to the potential of an interferent, I
0.05916
zA / zI
Ecell = K + log{aA + KA,I (aI ) } (23.3.4)
zA

where zA and zI are the charges of the analyte and the interferent, and KA,I is a selectivity coefficient that accounts for the relative
response of the interferent. The selectivity coefficient is defined as
(aA )e
KA,I = (23.3.5)
zA / zI
(aI )e

where (aA)e and (aI)e are the activities of analyte and the interferent that yield identical cell potentials. When the selectivity
coefficient is 1.00, the membrane responds equally to the analyte and the interferent. A membrane shows good selectivity for the
analyte when KA,I is significantly less than 1.00.
Selectivity coefficients for most commercially available ion-selective electrodes are provided by the manufacturer. If the selectivity
coefficient is not known, it is easy to determine its value experimentally by preparing a series of solutions, each of which contains
the same activity of interferent, (aI)add, but a different activity of analyte. As shown in Figure 23.3.2, a plot of cell potential versus
the log of the analyte’s activity has two distinct linear regions. When the analyte’s activity is significantly larger than KA,I × (aI)add,
the potential is a linear function of log(aA), as given by Equation 23.3.3. If KA,I × (aI)add is significantly larger than the analyte’s
activity, however, the cell’s potential remains constant. The activity of analyte and interferent at the intersection of these two linear
regions is used to calculate KA,I.

Figure 23.3.2 . Diagram showing the experimental determination of an ion-selective electrode’s selectivity for an analyte. The
activity of analyte that corresponds to the intersection of the two linear portions of the curve, (aA)inter, produces a cell potential
identical to that of the interferent. The equation for the selectivity coefficient, KA,I, is shown in red.

23.3.2 https://chem.libretexts.org/@go/page/333590
  Example 23.3.1

Sokalski and co-workers described a method for preparing ion-selective electrodes with significantly improved selectivities
[Sokalski, T.; Ceresa, A.; Zwicki, T.; Pretsch, E. J. Am. Chem. Soc. 1997, 119, 11347–11348]. For example, a conventional
Pb2+ ISE has a log K Pb
2+ of –3.6. If the potential for a solution in which the activity of Pb2+ is 4.1 × 10
/ Mg
2+ is identical to −12

that for a solution in which the activity of Mg2+ is 0.01025, what is the value of log K for their ISE? Pb
2+ 2+
/ Mg

Solution
Making appropriate substitutions into Equation 23.3.5, we find that
−12
(a 2+ )e 4.1 × 10
Pb −10
KPb 2+ / Mg2+ = = = 4.0 × 10
z /z +2/+2
Pb
2+
Mg
2+
(0.01025)
(a 2+ )e
Mg

The value of log K Pb

2+
/ Mg
2+ , therefore, is –9.40.

The Glass Electrode for pH Measurements

The earliest ion-selective electrodes were based on the observation that a thin glass membrane separating two solutions with
different levels of acidity develops a measurable difference in potential on opposite sides of the membrane. Incorporating the glass
electrode into a potentiometer along with a reference electrode provides a way to measure the potential. Commercial glass
membrane pH electrodes often are available in a combination form that includes both the indicator electrode and the reference
electrode. The use of a single electrode greatly simplifies the measurement of pH. An example of a typical combination electrode is
shown in Figure 23.3.3.

Figure 23.3.3 . Schematic diagram showing a combination glass electrode for measuring pH. The indicator electrode consists of a
pH-sensitive glass membrane and an internal Ag/AgCl reference electrode in a solution of 0.1 M HCl. The sample’s reference
electrode is a Ag/AgCl electrode in a solution of KCl (which may be saturated with KCl or contain a fixed concentration of KCl).
A porous wick serves as a salt bridge between the sample and its reference electrode.

The Composition and Structure of Glass Membranes

The first commercial glass electrodes were manufactured using Corning 015, a glass with a composition that is approximately 22%
Na2O, 6% CaO, and 72% SiO2. Membranes fashioned from Corning 015 have an excellent selectivity for hydrogen ions, H+, below
a pH of 9; above this pH the membrane becomes more selective for other cations and the measured pH value deviates from its
actual value. Replacing Na2O and CaO with Li2O and BaO extends the useful pH range of glass membranes to pH levels greater
than 12.

Origin of the Boundary Potential for a Glass Membrane

When immersed in an aqueous solution for several hours, the outer approximately 10 nm of the glass membrane’s surface becomes
hydrated, resulting in the formation of negatively charged sites, —SiO–. Sodium ions, Na+, serve as counter ions. Because H+ binds
more strongly to —SiO– than does Na+, they displace the sodium ions on both sides of the membrane.
+ − + − + +
H + −SiO Na ⇌ −SiO H + Na (23.3.6)

23.3.3 https://chem.libretexts.org/@go/page/333590
explaining the membrane’s selectivity for H+. The transport of charge across the membrane is carried by the Na+ ions within the
glass membrane. The potential of a glass electrode obeys the equation
Ecell = K + 0.05916 log aH+ (23.3.7)

Alkaline and Acid Errors

As noted above, at sufficiently basic pH values a glass electrode no longer provides an accurate measure of a sample's pH as the
membrane becomes more selective for other monovalent cations, such as Na+ and K+.

 Example 23.3.2

For a Corning 015 glass membrane, the selectivity coefficient KH+/Na+ is ≈ 10 . What is the expected error if we measure the
−11

pH of a solution in which the activity of H+ is 2 × 10 and the activity of Na+ is 0.05?

−13

Solution
A solution in which the activity of H+, (aH+)act, is 2 × 10 −13
has a pH of 12.7. Because the electrode responds to both H+ and
Na+, the apparent activity of H+, (aH+)app, is
−13 −11 −13
(a + )app = (a + )act + (K + + ×a + ) = 2 × 10 + (10 × 0.05) = 7 × 10
H H H / Na Na

The apparent activity of H+ is equivalent to a pH of 12.2, an error of –0.5 pH units.

Glass pH electrodes also show deviations from ideal behavior at pH levels less than 0.5, although the reasons for this are not clear.
Still, a glass electrode has a wide dynamic range for measuring pH.

Other Limitations to Glass Electrodes

Because an ion-selective electrode’s glass membrane is very thin—it is only about 50 μm thick—they must be handled with care to
avoid cracks or breakage. Glass electrodes usually are stored in a storage buffer recommended by the manufacturer, which ensures
that the membrane’s outer surface remains hydrated. If a glass electrode dries out, it is reconditioned by soaking for several hours
in a solution that contains the analyte. The composition of a glass membrane will change over time, which affects the electrode’s
performance. The average lifetime for a typical glass electrode is several years.

Glass Electrodes for Other Cations

The observation that the Corning 015 glass membrane responds to ions other than H+ led to the development of glass membranes
with a greater selectivity for other cations. For example, a glass membrane with a composition of 11% Na2O, 18% Al2O3, and 71%
SiO2 is used as an ion-selective electrode for Na+. Other glass ion-selective electrodes have been developed for the analysis of Li+,
K+, Rb+, Cs+, NH , Ag+, and Tl+. Table 23.3.1 provides several examples.
+

Table 23.3.1 . Representative Examples of Glass Membrane Ion-Selective Electrodes for Analytes Other Than H+
analyte membrane composition selectivity coefficients

K + + = 1000
Na /H

Na+ 11% Na2O, 18% Al2O3, 71% SiO2 K

Na
+
/K
+ = 0.001

KNa+ /Li+ = 0.001

K + + = 0.3
Li+
Li /Na
15% Li2O, 25% Al2O3, 60% SiO2
K + + = 0.001
Li /K

K+ 27% Na2O, 5% Al2O3, 68% SiO2 K

K
+
/Na
+ = 0.05

Selectivity coefficients are approximate; values found experimentally may vary substantially from the listed values. See Cammann, K. Working
With Ion-Selective Electrodes, Springer-Verlag: Berlin, 1977.

Crystalline Membrane Electrodes

A solid-state ion-selective electrode has a membrane that consists of either a polycrystalline inorganic salt or a single crystal of an
inorganic salt. We can fashion a polycrystalline solid-state ion-selective electrode by sealing a 1–2 mm thick pellet of AgS—or a

23.3.4 https://chem.libretexts.org/@go/page/333590
mixture of AgS and a second silver salt or another metal sulfide—into the end of a nonconducting plastic cylinder, filling the
cylinder with an internal solution that contains the analyte, and placing a reference electrode into the internal solution. Figure
23.3.4 shows a typical design.

The NaCl in a salt shaker is an example of polycrystalline material because it consists of many small crystals of sodium
chloride. The NaCl salt plates used in IR spectroscopy, on the other hand, are an example of a single crystal of sodium
chloride.

Figure 23.3.4 . Schematic diagram of a solid-state electrode. The internal solution contains a solution of analyte of fixed activity.
The membrane potential for a Ag2S pellet develops as the result of a difference in the extent of the solubility reaction
+ 2−
Ag2 S(s) ⇌ 2 Ag (aq) + S (aq) (23.3.8)

on the membrane’s two sides, with charge carried across the membrane by Ag+ ions. When we use the electrode to monitor the
activity of Ag+, the cell potential is
Ecell  = K + 0.05916 log a + (23.3.9)
Ag

The membrane also responds to the activity of S 2−

, with a cell potential of
0.05916
Ecell = K − log a 2− (23.3.10)
S
2

If we combine an insoluble silver salt, such as AgCl, with the Ag2S, then the membrane potential also responds to the concentration
of Cl–, with a cell potential of
Ecell  = K − 0.05916 log aCl− (23.3.11)

By mixing Ag2S with CdS, CuS, or PbS, we can make an ion-selective electrode that responds to the activity of Cd2+, Cu2+, or
Pb2+. In this case the cell potential is
0.05916
Ecell = K + ln a 2+ (23.3.12)
M
2

where aM2+ is the activity of the metal ion.

Table 23.3.2 provides examples of polycrystalline, Ag2S-based solid-state ion-selective electrodes. The selectivity of these ion-
selective electrodes depends on the relative solubility of the compounds. A Cl– ISE using a Ag2S/AgCl membrane is more selective
for Br– (KCl–/Br– = 102) and for I– (KCl–/I– = 106) because AgBr and AgI are less soluble than AgCl. If the activity of Br– is
sufficiently high, AgCl at the membrane/solution interface is replaced by AgBr and the electrode’s response to Cl– decreases
substantially. Most of the polycrystalline ion-selective electrodes listed in Table 23.3.2 operate over an extended range of pH
levels. The equilibrium between S2– and HS– limits the analysis for S2– to a pH range of 13–14.
23.3.2 . Representative Examples of Polycrystalline Solid-State Ion-Selective Electrodes
analyte membrane composition selectivity coefficients

23.3.5 https://chem.libretexts.org/@go/page/333590
 analyte membrane composition selectivity coefficients
−6
K + 2+ = 10
Ag /Cu

Ag+ Ag2S K
Ag
+
/Pb
2+ = 10
−10

Hg2+ interferes
K 2+ 2+ = 200
Cd /Fe

Cd2+ CdS/Ag2S K
Cd
2+
/Pb
2+ = 6

Ag+, Hg2+, and Cu2+ must be absent

K 2+ 3+ = 10
Cu /Fe

Cu2+ CuS/Ag2S K
Cu
2+
/Cu
+ = 10
−6

Ag+ and Hg2+ must be absent

K 2+ 3+ = 1
Pb /Fe

Pb2+ PbS/Ag2S K
Pb
2+
/Cd
2+ = 1

Ag+, Hg2+, and Cu2+ must be absent

K − − = 5000
Br /I

KBr− /Cl− = 0.005

Br– AgBr/Ag2S −5
K − − = 10
Br /OH

S2– must be absent

6
K − − = 10
Cl /I

K − − = 100
Cl– AgCl/Ag2S Cl /Br

K − − = 0.01
Cl /OH

S2– must be absent

K − 2− = 30
I /S

−4
K = 10
I–
− −

AgI/Ag2S I /Br

−6
KI− /Cl− = 10

−7
K − − = 10
I /OH

3
K − − = 10
SCN /I

K − − = 100
SCN– AgSCN/Ag2S SCN /Br

K − − = 0.1 K − − = 0.01
SCN /Cl SCN /OH

S2– must be absent

S2– Ag2S Hg2+ must be absent

Selectivity coefficients are approximate; values found experimentally may vary substantially from the listed values. See Cammann, K. Working
With Ion-Selective Electrodes, Springer-Verlag: Berlin, 1977.

The membrane of a F– ion-selective electrode is fashioned from a single crystal of LaF3, which usually is doped with a small
amount of EuF2 to enhance the membrane’s conductivity. Because EuF2 provides only two F– ions—compared to the three F– ions
in LaF3—each EuF2 produces a vacancy in the crystal’s lattice. Fluoride ions pass through the membrane by moving into adjacent
vacancies. As shown in Figure 23.3.4, the LaF3 membrane is sealed into the end of a non-conducting plastic cylinder, which
contains a standard solution of F–, typically 0.1 M NaF, and a Ag/AgCl reference electrode.
The membrane potential for a F– ISE results from a difference in the solubility of LaF3 on opposite sides of the membrane, with the
potential given by
Ecell = K − 0.05916 log a − (23.3.13)
F

– –
One advantage of the F ion-selective electrode is its freedom from interference. The only significant exception is OH (KF–/OH– =
0.1), which imposes a maximum pH limit for a successful analysis. Below a pH of 4 the predominate form of fluoride in solution is
HF, which does not contribute to the membrane potential. For this reason, an analysis for fluoride is carried out at a pH greater than
4.

23.3.6 https://chem.libretexts.org/@go/page/333590
  Example 23.3.3

What is the maximum pH that we can tolerate if we need to analyze a solution in which the activity of F– is 1 × 10 −5
with an
error of less than 1%?

Solution
In the presence of OH– the cell potential is

Ecell = K − 0.05916 {aF− + KF− / OH − × aOH − }

To achieve an error of less than 1%, the term K F

−
/ OH
− × aOH − must be less than 1% of aF–; thus

K − − ×a − ≤ 0.01 × aF−
F / OH OH

−5
0.10 × a − ≤ 0.01 × (1.0 × 10 )
OH

Solving for aOH– gives the maximum allowable activity for OH– as 1 × 10 −6
, which corresponds to a pH of less than 8.

Unlike a glass membrane ion-selective electrode, a solid-state ISE does not need to be conditioned before it is used, and it may be
stored dry. The surface of the electrode is subject to poisoning, as described above for a Cl– ISE in contact with an excessive
concentration of Br–. If an electrode is poisoned, it can be returned to its original condition by sanding and polishing the crystalline
membrane.

Poisoning simply means that the surface has been chemically modified, such as AgBr forming on the surface of a AgCl
membrane.

Liquid Membrane Electrodes

Another class of ion-selective electrodes uses a hydrophobic membrane that contains a liquid organic complexing agent that reacts
selectively with the analyte. Three types of organic complexing agents have been used: cation exchangers, anion exchangers, and
neutral ionophores. A membrane potential exists if the analyte’s activity is different on the two sides of the membrane. Current is
carried through the membrane by the analyte.

An ionophore is a ligand whose exterior is hydrophobic and whose interior is hydrophilic. The crown ether shown here is one
example of a neutral ionophore.

One example of a liquid-based ion-selective electrode is that for Ca2+, which uses a porous plastic membrane saturated with the
cation exchanger di-(n-decyl) phosphate. As shown in Figure 23.3.5, the membrane is placed at the end of a non-conducting
cylindrical tube and is in contact with two reservoirs. The outer reservoir contains di-(n-decyl) phosphate in di-n-
octylphenylphosphonate, which soaks into the porous membrane. The inner reservoir contains a standard aqueous solution of Ca2+
and a Ag/AgCl reference electrode. Calcium ion-selective electrodes also are available in which the di-(n-decyl) phosphate is
immobilized in a polyvinyl chloride (PVC) membrane that eliminates the need for the outer reservoir.

23.3.7 https://chem.libretexts.org/@go/page/333590
 Figure 23.3.5 . Schematic diagram showing a liquid-based ion-selective electrode for Ca2+. The structure of the cation exchanger,
di-(n-decyl) phosphate, is shown in red.
The membrane potential for the Ca2+ ISE develops as the result of a difference in the extent of the complexation reaction
2+ −
Ca (aq) + 2 (C10 H21 O) PO (mem) ⇌ Ca [ (C10 H21 O) PO2 ] (mem) (23.3.14)
2 2 2 2

on the two sides of the membrane, where (mem) indicates a species that is present in the membrane. The cell potential for the Ca2+
ion-selective electrode is
0.05916
Ecell = K + log aca2+ (23.3.15)
2

The selectivity of this electrode for Ca2+ is very good, with only Zn2+ showing greater selectivity.
Table 23.3.3 lists the properties of several liquid-based ion-selective electrodes. An electrode using a liquid reservoir can be stored
in a dilute solution of analyte and needs no additional conditioning before use. The lifetime of an electrode with a PVC membrane,
however, is proportional to its exposure to aqueous solutions. For this reason these electrodes are best stored by covering the
membrane with a cap along with a small amount of wetted gauze to maintain a humid environment. Before using the electrode it is
conditioned in a solution of analyte for 30–60 minutes.
Table 23.3.3 . Representative Examples of Liquid-Based Ion-Selective Electrodes
analyte membrane composition selectivity coefficients

K 2+ 2+ = 1 −5
Ca /Zn

K 2+ 3+ = 0.90
Ca /Al

Ca2+ di-(n-decyl) phosphate in PVC K

Ca
2+
/Mn
2+ = 0.38

K 2+ 2+ = 0.070
Ca /Cu

K 2+ 2+ = 0.032
Ca /Mg

K + + = 1.9
K /Rb

K+ valinomycin in PVC K
K
+
/Cs
+ = 0.38

−4
KK+ /Li+ = 10

KLi+ /H+ = 1

Li+ ETH 149 in PVC K

Li
+
/Na
+ = 0.03

K + + = 0.007
Li /K

KNH+ /K+ = 0.12

4

KNH+ /H+ = 0.016

NH
+
4
nonactin and monactin in PVC 4

K + + = 0.0042
NH /Li
4

K + + = 0.002
NH4 /Na

K − − = 1
ClO4 /OH

KClO− /I− = 0.012

− Fe(o-phen)
3+
3
in p-nitrocymene with porous 4

ClO K − − = 0.0015
ClO /NO3
3
membrane 4

−4
K − − = 5.6 × 10
ClO4 /Br

−4
K − − = 2.2 × 10
ClO4 /Cl

K − − = 0.006
NO3 /Cl
NO
−
3
tetradodecyl ammonium nitrate in pVC −4
K − − = 9 × 10
NO3 /F

23.3.8 https://chem.libretexts.org/@go/page/333590
 analyte membrane composition selectivity coefficients

Selectivity coefficients are approximate; values found experimentally may vary substantially from the listed values. See Cammann, K. Working
With Ion-Selective Electrodes, Springer-Verlag: Berlin, 1977.

This page titled 23.3: Membrane Ion-Selective Electrodes is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

23.3.9 https://chem.libretexts.org/@go/page/333590
23.4: Molecular-Selective Electrode Systems
The electrodes in Chapter 23.3 are selective toward ions. In this section we consider how we can incorporate an ion-selective
electrode into an electrode that responds to neutral species, such as volatile analytes, such as CO2 and NH3, and biochemically
important compounds, such as amino acids and urea.

Gas-Sensing Membrane Electrodes

A number of membrane electrodes respond to the concentration of a dissolved gas. The basic design of a gas-sensing electrode, as
shown in Figure 23.4.1, consists of a thin membrane that separates the sample from an inner solution that contains an ion-selective
electrode. The membrane is permeable to the gaseous analyte, but impermeable to nonvolatile components in the sample’s matrix.
The gaseous analyte passes through the membrane where it reacts with the inner solution, producing a species whose concentration
is monitored by the ion-selective electrode. For example, in a CO2 electrode, CO2 diffuses across the membrane where it reacts in
the inner solution to produce H3O+.
− +
CO2 (aq) + 2 H2 O(l) ⇌  HCO (aq) +  H 3 O (aq) (23.4.1)
3

The change in the activity of H3O+ in the inner solution is monitored with a pH electrode, for which the cell potential, from Chapter
23.3, is
Ecell = K + 0.09516 log aH+ (23.4.2)

To find the relationship between the activity of H3O+ in the inner solution and the activity of CO2 in the inner solution we rearrange
the equilibrium constant expression for reaction 23.4.1; thus
aCO
2
aH +
O
= Ka × (23.4.3)
3
a −
HCO3

where Ka is the equilibrium constant. If the activity of HCO in the internal solution is sufficiently large, then its activity is not
−
3

affected by the small amount of CO2 that passes through the membrane. Substituting Equation 23.4.3 into Equation 23.4.2 gives
′
Ecell = K + 0.05916 log aco (23.4.4)
2

where K′ is a constant that includes the constant for the pH electrode, the equilibrium constant for reaction 23.4.1 and the activity
of HCO in the inner solution.
−
3

Figure 23.4.1 . Schematic diagram of a gas-sensing membrane electrode.

Table 23.4.1 lists the properties of several gas-sensing electrodes. The composition of the inner solution changes with use, and both
the inner solution and the membrane must be replaced periodically. Gas-sensing electrodes are stored in a solution similar to the
internal solution to minimize their exposure to atmospheric gases.
Table 23.4.1 . Representative Examples of Gas-Sensing Electrodes
analyte inner solution reaction in inner solution ion-selective electrode

10 mM NaHCO3 CO 2 (aq) + 2 H2 O(l) ⇌

CO2 glass pH ISE
10 mM NaCl  HCO
−
3
(aq) +  H3 O
+
(aq)

23.4.1 https://chem.libretexts.org/@go/page/333594
 analyte inner solution reaction in inner solution ion-selective electrode

HCN(aq) + H2 O(l) ⇌
HCN 10 mM KAg(CN)2 − +
Ag2S solid-state ISE
CN (aq) +  H3 O (aq)

HF(aq) + H2 O(l) ⇌
HF 1 M H3O+ − +
F– solid-state ISE
F (aq) +  H3 O (aq)

H2 S(aq) +  H2 O(l) ⇌

H2S pH 5 citrate buffer − +
Ag2S solid state ISE
HS (aq) +  H3 O (aq)

10 mM NH4Cl NH3 (aq) +  H2 O(l) ⇌

NH3 glass pH ISE
0.1 M KNO3 NH
+
4
(aq) +  OH
−
(aq)

20 mM NaNO2 2 NO2 (aq) + 3 H2 O(l) ⇌

NO2 glass pH ISE
0.1 M KNO3 NO
−
3
(aq) +  NO
−
2
(aq) + 2 H3 O
+
(aq)

1 mM NaHSO3 SO 2 (aq) + 2 H2 O(l) ⇌

SO2 glass pH ISE
pH 5 HSO
−
3
(aq) +  H3 O
+
(aq)

Source: Cammann, K. Working With Ion-Selective Electrodes, Springer-Verlag: Berlin, 1977.

Biocatalytic Membrane Electrodes

The approach for developing gas-sensing electrodes can be modified to create potentiometric electrodes that respond to a
biochemically important species. The most common class of potentiometric biosensors are enzyme electrodes, in which we trap or
immobilize an enzyme at the surface of a potentiometric electrode. The analyte’s reaction with the enzyme produces a product
whose concentration is monitored by the potentiometric electrode. Potentiometric biosensors also have been designed around other
biologically active species, including antibodies, bacterial particles, tissues, and hormone receptors.
One example of an enzyme electrode is the urea electrode, which is based on the catalytic hydrolysis of urea by urease
+ −
CO(NH2 ) (aq) + 2 H2 O(l) ⇌ 2 NH (aq) +  CO (aq) (23.4.5)
2 4 3

Figure 23.4.2 shows one version of the urea electrode, which modifies a gas-sensing NH3 electrode by adding a dialysis membrane
that traps a pH 7.0 buffered solution of urease between the dialysis membrane and the gas permeable membrane [(a)
Papastathopoulos, D. S.; Rechnitz, G. A. Anal. Chim. Acta 1975, 79, 17–26; (b) Riechel, T. L. J. Chem. Educ. 1984, 61, 640–642].
An NH3 electrode, as shown in Table 23.4.1, uses a gas-permeable membrane and a glass pH electrode. The NH3 diffuses across
the membrane where it changes the pH of the internal solution.

Figure 23.4.2 . Schematic diagram showing an enzyme-based potentiometric biosensor for urea. A solution of the enzyme urease is
trapped between a dialysis membrane and a gas permeable membrane. Urea diffuses across the dialysis membrane and reacts with
urease, producing NH3 that diffuses across the gas permeable membrane. The resulting change in the internal solution’s pH is
measured with the pH electrode.
When immersed in the sample, urea diffuses through the dialysis membrane where it reacts with the enzyme urease to form the
ammonium ion, NH , which is in equilibrium with NH3.
+

+ +
NH (aq) + H2 O(l) ⇌  H 3 O (aq) +  NH 3 (aq) (23.4.6)
4

23.4.2 https://chem.libretexts.org/@go/page/333594
The NH3, in turn, diffuses through the gas permeable membrane where a pH electrode measures the resulting change in pH. The
electrode’s response to the concentration of urea is
Ecell  = K − 0.05916 log aurea  (23.4.7)

Another version of the urea electrode (Figure 23.4.3) immobilizes the enzyme urease in a polymer membrane formed directly on
the tip of a glass pH electrode [Tor, R.; Freeman, A. Anal. Chem. 1986, 58, 1042–1046]. In this case the response of the electrode is
pH = Kaurea (23.4.8)

Figure 23.4.3 . Schematic diagram of an enzyme-based potentiometric biosensor for urea in which urease is immobilized in a
polymer membrane coated onto the pH-sensitive glass membrane of a pH electrode.
Few potentiometric biosensors are available commercially. As shown in Figure 23.4.2 and Figure 23.4.3, however, it is possible to
convert an ion-selective electrode or a gas-sensing electrode into a biosensor. Several representative examples are described in
Table 23.4.2, and additional examples can be found in this chapter’s additional resources.
Table 23.4.2 . Representative Examples of Potentiometric Biosensors
analyte biologically active phase substance determined

5
′
-AMP AMP-deaminase (E) NH3

L-arginine arginine and urease (E) NH3

asparagine asparaginase (E) NH

+

L-cysteine Proteus morganii (B) H2S

L-glutamate yellow squash (T) CO2

L-glutamine Sarcina flava (B) NH3

oxalate oxalate decarboxylase (E) CO2

penicillin penicllinase (E) H3O+

L-amino acid oxidase/horseradish peroxidase

L-phenylalanine I–
(E)

sugars bacteria from dental plaque (B) H3O+

urea urease (E) NH3 or H3O+

Source: Complied from Cammann, K. Working With Ion-Selective Electrodes, Springer-Verlag: Berlin, 1977 and Lunte, C. E.; Heineman, W. R.
“Electrochemical techniques in Bioanalysis,” in Steckham, E. ed. Topics in Current Chemistry, Vol. 143, Springer-Verlag: Berlin, 1988, p.8.
Abbreviations for biologically active phase: E = enzyme; B = bacterial particle; T = tissue.

This page titled 23.4: Molecular-Selective Electrode Systems is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

23.4.3 https://chem.libretexts.org/@go/page/333594
23.5: Instruments for Measuring Cell Potentials
To measure the potential of an electrochemical cell in a way that draws essentially no current we use a potentiometer. To help us
understand how a potentiometer accomplishes this, we will describe the instrument as if the analyst is operating it manually. To do
so the analyst observes a change in the current or the potential and manually adjusts the instrument’s settings to maintain the
desired experimental conditions. It is important to understand that modern electrochemical instruments provide an automated,
electronic means for controlling and measuring current and potential, and that they do so by using very different electronic circuitry
than that described here.

Figure 23.5.1 . Schematic diagram of a manual potentiometer: C is the reference electrode; W is the working electrode; SW is a
slide-wire resistor; T is a tap key and i is an ammeter for measuring current.
Figure 23.5.1 shows a schematic diagram for a manual potentiometer that consists of a power supply, an electrochemical cell with
a working electrode and a counter electrode, an ammeter to measure the current that passes through the electrochemical cell, an
adjustable, slide-wire resistor, and a tap key for closing the circuit through the electrochemical cell. Using Ohm’s law, the current in
the upper half of the circuit is
EPS
iupper = (23.5.1)
Rab

where EPS is the power supply’s potential, and Rab is the resistance between points a and b of the slide-wire resistor. In a similar
manner, the current in the lower half of the circuit is
Ecell
ilower = (23.5.2)
Rcb

where Ecell is the potential difference between the working electrode and the counter electrode, and Rcb is the resistance between
the points c and b of the slide-wire resistor. When iupper = ilower = 0, no current flows through the ammeter and the potential of the
electrochemical cell is
Rcb
Ecoll = × EPS (23.5.3)
Rab

To determine Ecell we briefly press the tap key and observe the current at the ammeter. If the current is not zero, then we adjust the
slide wire resistor and remeasure the current, continuing this process until the current is zero. When the current is zero, we use
Equation 23.5.3 to calculate Ecell.
Using the tap key to briefly close the circuit through the electrochemical cell minimizes the current that passes through the cell and
limits the change in the electrochemical cell’s composition. For example, passing a current of 10–9 A through the electrochemical
cell for 1 s changes the concentrations of species in the cell by approximately 10–14 moles.
−9 −9
10  A = 10  C/s (23.5.4)

−9
1 mol −14
10  C/s × 1 s × = 1.0 × 10  mol (23.5.5)
96485C

Of course, trying to measure a potential in this way is tedious. Modern potentiometers use operational amplifiers to create a high-
impedance voltmeter that measures the potential while drawing a current of less than 10 A. The relative error, E , in the
–9
r

measured potential

23.5.1 https://chem.libretexts.org/@go/page/333596
 Rcell
Er = (23.5.6)
Rmeter + Rcell

where R is the resistance of the solution in the electrochemical cell and R

cell is the resistance of the meter. For a solution with
meter

a resistance of 10MΩ to achieve a relative error of −0.1%) or −0.001 requires an R of

meter

−10 MΩ
−0.001 =
Rmeter + 10 MΩ

−0.001 × Rmeter − 0.01 = −10 MΩ

−0.001 × Rmeter = −9.990 MΩ

Rmeter = 9990 MΩ

This page titled 23.5: Instruments for Measuring Cell Potentials is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

23.5.2 https://chem.libretexts.org/@go/page/333596
23.6: Quantitative Potentiometry
The most important application of potentiometry is determining the concentration of an analyte in solution. Most potentiometric
electrodes are selective toward the free, uncomplexed form of the analyte, and do not respond to any of the analyte’s complexed
forms. This selectivity provides potentiometric electrodes with a significant advantage over other quantitative methods of analysis
if we need to determine the concentration of free ions. For example, calcium is present in urine both as free Ca2+ ions and as
protein-bound Ca2+ ions. If we analyze a urine sample using atomic absorption spectroscopy, the signal is propor- tional to the total
concentration of Ca2+ because both free and bound calcium are atomized. Analyzing urine with a Ca2+ ISE, however, gives a signal
that is a function of only free Ca2+ ions because the protein-bound Ca2+ can not interact with the electrode’s membrane. In this
section, we consider several important aspects of quantiative potentiometry.

The Relationship Between Concentration and Potential

In Chapter 23.3, we showed that the potential of an ion-selective electrode for an ion with a charge of z is
0.05916
Ecell = K + log (aA )samp (23.6.1)
z

where K is a constant that includes the potentials of the ion-selective electrode's internal and external reference electrodes, any
asymmetry potential associated with the ion-selective electrode's membrane, and the analyte's activity in the ion-selective
electrode's internal solution. Equation 23.6.1 is a general equation and applies to all types of ion-selective electrodes. Note that
when the analyte is a cation, an increase in the analyte's activity results in results in an increase in the potential; when the analyte is
an anion, which makes z a negative number, an increase in the analyte's activity results in a decrease in the potential.
As the concentrations of ions in solution often are reported as pX values, where

pX = − log aX (23.6.2)

it is convenient to substitute Equation 23.6.2 into Equation 23.6.1

0.05916
Ecell = K −  pA (23.6.3)
z

Note that for a cation, an increase in pA results in a decrease in the potential; when the analyte is an anion, an increase in pA results
in an increase in the potential.

Calibrating Potentiometric Electrodes

To use Equation 23.6.3 we need to determine the value of K, which we can do using one or more external standards or by the
method of standard addition, both of which were covered in Chapter 1.5. One complication, of course, is that potential is a function
of the analyte's activity instead of its concentration.

Activity and Concentration

Equation 23.6.1 is written in terms of the analyte's activity. When we use a potentiometric electrode, however, our goal is to
determine the analyte’s concentration. As we learned in Chapter 22, an ion’s activity is the product of its concentration, [Mn+], and
a matrix-dependent activity coefficient, γ .
Mnn+

n+
aM n+ = [ M ] γM n+ (23.6.4)

Substituting Equation 23.6.4 into Equation 23.6.1 and rearranging, gives

0.05916 0.05916 n+
Ecell = K + log γM n+ + log[ M ] (23.6.5)
n n

We can solve Equation 23.6.5 for the metal ion’s concentration if we know the value for its activity coefficient. Unfortunately, if
we do not know the exact ionic composition of the sample’s matrix—which is the usual situation—then we cannot calculate the
value of γMn
n+ . There is a solution to this dilemma. If we design our system so that the standards and the samples have an identical
matrix, then the value of γMn
n+ remains constant and Equation 23.6.5 simplifies to

′
0.05916 n+
Ecell = K + log[ M ] (23.6.6)
n

23.6.1 https://chem.libretexts.org/@go/page/333598
where K includes the activity coefficient.
′

Calibration Using External Standards

In the absence of interferents, a calibration curve of Ecell versus logaA, where A is the analyte, is a straight-line. A plot of Ecell
versus log[A], however, may show curvature at higher concentrations of analyte as a result of a matrix-dependent change in the
analyte’s activity coefficient. To maintain a consistent matrix we add a high concentration of an inert electrolyte to all samples and
standards. If the concentration of added electrolyte is sufficient, then the difference between the sample’s matrix and the matrix of
the standards will not affect the ionic strength and the activity coefficient essentially remains constant. The inert electrolyte added
to the sample and the standards is called a total ionic strength adjustment buffer (TISAB).

 Example 23.6.1

The concentration of Ca2+ in a water sample is determined using the method of external standards. The ionic strength of the
samples and the standards is maintained at a nearly constant level by making each solution 0.5 M in KNO3. The measured cell
potentials for the external standards are shown in the following table.

[Ca2+] (M) Ecell (V)

1.00 × 10
−5
–0.125

5.00 × 10
−5
–0.103

1.00 × 10
−4
–0.093

5.00 × 10
−4
–0.072

1.00 × 10
−3
–0.063

5.00 × 10
−3
–0.043

1.00 × 10
−2
–0.033

What is the concentration of Ca2+ in a water sample if its cell potential is found to be –0.084 V?

Solution
Linear regression gives the calibration curve in Figure 23.6.1, with an equation of
2+
Ecell = 0.027 + 0.0303 log[ Ca ]

Substituting the sample’s cell potential gives the concentration of Ca2+ as 2.17 × 10 M. Note that the slope of the calibration
−4

curve, which is 0.0303, is slightly larger than its ideal value of 0.05916/2 = 0.02958; this is not unusual and is one reason for
using multiple standards. One reason that it is not unusual to find that the experimental slope deviates from its ideal value of
0.05916/n is that this ideal value assumes that the temperature is 25°C.

Figure 23.6.1 . Calibration curve for the data in Example 23.6.1 .

23.6.2 https://chem.libretexts.org/@go/page/333598
Calibration Using Standard Additions
Another approach to calibrating a potentiometric electrode is the method of standard additions, which was introduced in Chapter
1.5. First, we transfer a sample with a volume of Vsamp and an analyte concentration of Csamp into a beaker and measure the
potential, (Ecell)samp. Next, we make a standard addition by adding to the sample a small volume, Vstd, of a standard that contains a
known concentration of analyte, Cstd, and measure the potential, (Ecell)std. If Vstd is significantly smaller than Vsamp, then we can
safely ignore the change in the sample’s matrix and assume that the analyte’s activity coefficient is constant. Example 23.6.9
demonstrates how we can use a one-point standard addition to determine the concentration of analyte in a sample.

 Example 23.6.2

The concentration of Ca2+ in a sample of sea water is determined using a Ca ion-selective electrode and a one-point standard
addition. A 10.00-mL sample is transferred to a 100-mL volumetric flask and diluted to volume. A 50.00-mL aliquot of the
sample is placed in a beaker with the Ca ISE and a reference electrode, and the potential is measured as –0.05290 V. After
adding a 1.00-mL aliquot of a 5.00 × 10 M standard solution of Ca2+ the potential is –0.04417 V. What is the concentration
−2

of Ca2+ in the sample of sea water?

Solution
To begin, we write the Nernst equation before and after adding the standard addition. The cell potential for the sample is
0.05916
(Ecell ) =K+ log Csamp
samp
2

and that following the standard addition is

0.05916 Vsamp Vstd
(Ecell )std = K + log{ Csamp + Cstd }
2 Vtot Vtot

where Vtot is the total volume (Vsamp + Vstd) after the standard addition. Subtracting the first equation from the second equation
gives
0.05916 Vsamp Vstd 0.05916
ΔE = (Ecell ) − (Ecell ) = log{ Csamp + Cstd } − log Csamp
std samp
2 Vtot Vtot 2

Rearranging this equation leaves us with

2ΔEcell Vsamp Vstd Cstd
= log{ + }
0.05916 Vtot Vtot Csamp

Substituting known values for ΔE, Vsamp, Vstd, Vtot and Cstd,
2×{−0.04417−(−0.05290)}
=
0.05916

−2
(1.00 mL)(5.00×10 M)
50.00 mL
log{ + }
51.00 mL (51.00 mL)Cs am p

−4
9.804×10
0.2951 = log{0.9804 + }
Cs am p

and taking the inverse log of both sides gives

−4
9.804 × 10
1.973 = 0.9804 +
Csamp 

Finally, solving for Csamp gives the concentration of Ca2+ as 9.88 × 10 M. Because we diluted the original sample of
−4

seawater by a factor of 10, the concentration of Ca2+ in the seawater sample is 9.88 × 10 M. −3

The Operational Definition of pH

23.6.3 https://chem.libretexts.org/@go/page/333598
With the availability of inexpensive glass pH electrodes and pH meters, the determination of pH is one of the most common
quantitative analytical measurements. The potentiometric determination of pH, however, is not without complications, several of
which we discuss in this section.
One complication is confusion over the meaning of pH [Kristensen, H. B.; Saloman, A.; Kokholm, G. Anal. Chem. 1991, 63,
885A–891A]. The conventional definition of pH in most general chemistry textbooks is given in terms of the concentration of H+
+
pH = − log[ H ] (23.6.7)

As we now know, when we measure pH it actually is a measure of the activity of H+.

pH = − log aH+ (23.6.8)

Try this experiment—find several general chemistry textbooks and look up pH in each textbook’s index. Turn to the
appropriate pages and see how it is defined. Next, look up activity or activity coefficient in each textbook’s index and see if
these terms are indexed.

Equation 23.6.7 only approximates the true pH. If we calculate the pH of 0.1 M HCl using Equation 23.6.7, we obtain a value of
1.00; the solution’s actual pH, as defined by Equation 23.6.8, is 1.1 [Hawkes, S. J. J. Chem. Educ. 1994, 71, 747–749]. The activity
and the concentration of H+ are not the same in 0.1 M HCl because the activity coefficient for H+ is not 1.00 in this matrix. Figure
23.6.2 shows a more colorful demonstration of the difference between activity and concentration.

Figure 23.6.2 . A demonstration of the difference between activity and concentration using the indicator methyl green. The
indicator is pale yellow in its acid form (beaker a: 1.0 M HCl) and is blue in its base form (beaker d: H2O). In 10 mM HCl the
indicator is in its base form (beaker b: 20 mL of 10 mM HCl with 3 drops of methyl green). Adding 20 mL of 5 M LiCl to this
solution shifts the indicator's color to green (beaker c); although the concentration of HCl is cut in half to 5 mM, the activity of H+
has increased as evidenced by the green color that is intermediate between the indicator’s pale yellow, acid form and its blue, base
form. The demonstration shown here is adapted from McCarty, C. G.; Vitz, E. “pH Paradoxes: Demonstrating That It Is Not True
That pH ≡ –log[H+],” J. Chem. Educ. 2006, 83, 752–757. This paper provides several additional demonstrations that illustrate the
difference between concentration and activity.
A second complication in measuring pH is the uncertainty in the relationship between potential and activity. For a glass membrane
electrode, the cell potential, (Ecell)samp, for a sample of unknown pH is
RT 1 2.303RT
(Ecell )samp = K − ln =K− pH samp (23.6.9)
F aH+ F

where K includes the potential of the reference electrode, the asymmetry potential of the glass membrane, and any junction
potentials in the electrochemical cell. All the contributions to K are subject to uncertainty, and may change from day-to-day, as well
as from electrode-to-electrode. For this reason, before using a pH electrode we calibrate it using a standard buffer of known pH.
The cell potential for the standard, (Ecell)std, is
2.303RT
(Eccll )std = K − pHstd (23.6.10)
F

where pHstd is the standard’s pH. Subtracting Equation ??? from Equation 23.6.9 and solving for pHsamp gives

{(Ecell ) − (Ecell ) }F
samp std

pH samp = pH std − (23.6.11)

2.303RT

which is the operational definition of pH adopted by the International Union of Pure and Applied Chemistry [Covington, A. K.;
Bates, R. B.; Durst, R. A. Pure & Appl. Chem. 1985, 57, 531–542].
Calibrating a pH electrode presents a third complication because we need a standard with an accurately known activity for H+.
Table 23.6.1 provides pH values for several primary standard buffer solutions accepted by the National Institute of Standards and

23.6.4 https://chem.libretexts.org/@go/page/333598
Technology.
Table 23.6.1 . pH Values for Selected NIST Primary Standard Buffers
saturated 0.008695 m
(at 25oC) 0.05 m 0.05 m 0.025 m KH2PO4, 0.025 m
KHC4H4O7 KH2C6H5O7 KHC8H4O4 KH2PO4, 0.025 0.03043 m 0.01 m NaHCO3, 0.025
temp (oC) (tartrate) (citrate) (phthlate) m NaHPO4 Na2HPO4 Na4B4O7 m Na2CO3

0 — 3.863 4.003 6.984 7.534 9.464 10.317

5 — 3.840 3.999 6.951 7.500 9.395 10.245

10 — 3.820 3.998 6.923 7.472 9.332 10.179

15 — 3.802 3.999 6.900 7.448 9.276 10.118

20 — 3.788 4.002 6.881 7.429 9.225 10.062

25 3.557 3.776 4.008 6.865 7.413 9.180 10.012

30 3.552 3.766 4.015 6.854 7.400 9.139 9.966

35 3.549 3.759 4.024 6.844 7.389 9.012 9.925

40 3.547 3.753 4.035 6.838 7.380 9.068 9.889

45 3.547 3.750 4.047 6.834 7.373 9.038 9.856

50 3.549 3.749 4.060 6.833 7.367 9.011 9.828

Source: Values taken from Bates, R. G. Determination of pH: Theory and Practice, 2nd ed. Wiley: New York, 1973. See also Buck, R. P., et.
al.“Measurement of pH. Definition, Standards, and Procedures,” Pure. Appl. Chem. 2002, 74, 2169–2200. All concentrations are molal (m).

To standardize a pH electrode using two buffers, choose one near a pH of 7 and one that is more acidic or basic depending on your
sample’s expected pH. Rinse your pH electrode in deionized water, blot it dry with a laboratory wipe, and place it in the buffer with
the pH closest to 7. Swirl the pH electrode and allow it to equilibrate until you obtain a stable reading. Adjust the “Standardize” or
“Calibrate” knob until the meter displays the correct pH. Rinse and dry the electrode, and place it in the second buffer. After the
electrode equilibrates, adjust the “Slope” or “Temperature” knob until the meter displays the correct pH.
Some pH meters can compensate for a change in temperature. To use this feature, place a temperature probe in the sample and
connect it to the pH meter. Adjust the “Temperature” knob to the solution’s temperature and calibrate the pH meter using the
“Calibrate” and “Slope” controls. As you are using the pH electrode, the pH meter compensates for any change in the sample’s
temperature by adjusting the slope of the calibration curve using a Nernstian response of 2.303RT/F.

This page titled 23.6: Quantitative Potentiometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

23.6.5 https://chem.libretexts.org/@go/page/333598
 CHAPTER OVERVIEW
24: Coulometry
In a potentiometric method of analysis we determine an analyte’s concentration by measuring the potential of an electrochemical
cell under static conditions in which no current flows and the concentrations of species in the electrochemical cell remain fixed.
Dynamic techniques, in which current passes through the electrochemical cell and concentrations change, also are important
electrochemical methods of analysis. In this Chapter we consider coulometry. Voltammetry and amperometry are covered in
Chapter 25.
24.1: Introduction to Coulometry
24.2: Controlled-Potential Coulometry
24.3: Controlled-Current Coulometry

This page titled 24: Coulometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

1
24.1: Introduction to Coulometry
Coulometry is based on an exhaustive electrolysis of the analyte. By exhaustive we mean the analyte is oxidized or reduced
completely at the working electrode, or reacts completely with a reagent generated at the working electrode. There are two forms of
coulometry: controlled-potential coulometry, in which we apply a constant potential to the electrochemical cell, and controlled-
current coulometry, in which we pass a constant current through the electrochemical cell.
During an electrolysis, the total charge, Q, in coulombs, that passes through the electrochemical cell is proportional to the absolute
amount of analyte by Faraday’s law
Q = nF NA (24.1.1)

where n is the number of electrons per mole of analyte, F is Faraday’s constant (96 487 C mol–1), and NA is the moles of analyte. A
coulomb is equivalent to an A•sec; thus, for a constant current, i, the total charge is
Q = ite (24.1.2)

where te is the electrolysis time. If the current varies with time, as it does in controlled-potential coulometry, then the total charge is
te

Q =∫ i(t)dt (24.1.3)
0

In coulometry, we monitor current as a function of time and use either Equation 24.1.2 or Equation 24.1.3 to calculate Q. Knowing
the total charge, we then use Equation 24.1.1 to determine the moles of analyte. To obtain an accurate value for NA, all the current
must oxidize or reduce the analyte; that is, coulometry requires 100% current efficiency or an accurate measurement of the current
efficiency using a standard.

Current efficiency is the percentage of current that actually leads to the analyte’s oxidation or reduction.

This page titled 24.1: Introduction to Coulometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

24.1.1 https://chem.libretexts.org/@go/page/333839
24.2: Controlled-Potential Coulometry
The easiest way to ensure 100% current efficiency is to hold the working electrode at a constant potential where the analyte is
oxidized or reduced completely and where no potential interfering species are oxidized or reduced. As electrolysis progresses, the
analyte’s concentration and the current decrease. The resulting current-versus-time profile for controlled-potential coulometry is
shown in Figure 24.2.1. Integrating the area under the curve from t = 0 to t = te gives the total charge. In this section we consider
the experimental parameters and instrumentation needed to develop a controlled-potential coulometric method of analysis and its
applications.

Figure 24.2.1 . Current versus time for a controlled-potential coulometric analysis. The measured current is shown by the red curve.
The integrated area under the curve, shown in blue, is the total charge.

Selecting a Constant Potential

To understand how an appropriate potential for the working electrode is selected, let’s develop a constant-potential coulometric
method for Cu2+ based on its reduction to copper metal at a Pt working electrode.
2+ −
Cu (aq) + 2 e ⇌ Cu(s) (24.2.1)

Figure 24.2.2 shows the three reduction reactions that can take place in an aqueous solution of Cu2+ and their standard state
reduction potentials: the reduction of O2 to H2O, the reduction of Cu2+ to Cu, and the reduction of H3O+ to H2. From the diagram
we know that reaction 24.2.1 is favored when the working electrode’s potential is more negative than +0.342 V versus the standard
hydrogen electrode. To ensure a 100% current efficiency, however, the potential must be sufficiently more positive than +0.000 V
so that the reduction of H3O+ to H2 does not contribute significantly to the total current flowing through the electrochemical cell.

Figure 24.2.2 . Standard state reduction potentials for an aqueous solution of Cu2+ showing for the reductions of O2 to H2O, of Cu2+
to Cu, and of H3O+ to H2. For each reaction, the oxidized species is in blue and the reduced species is in red.
We can use the Nernst equation for reaction 24.2.1 to estimate the minimum potential for quantitatively reducing Cu2+.
0.05916 1
o
E =E 2+
− log (24.2.2)
Cu /Cu 2+
2 [ Cu ]

24.2.1 https://chem.libretexts.org/@go/page/333854
 So why are we using the concentration of Cu2+ in Equation 24.2.2 instead of its activity as we did in Chapter 23 when we
considered potentiometry? In potentiometry we used activity because we used Ecell to determine the analyte’s concentration.
Here we use the Nernst equation to help us select an appropriate potential. Once we identify a potential, we can adjust its value
as needed to ensure a quantitative reduction of Cu2+. In addition, in coulometry the analyte’s concentration is given by the total
charge, not the applied potential.

If we define a quantitative electrolysis as one in which we reduce 99.99% of Cu2+ to Cu, then the concentration of Cu2+ at te is
2+ 2+
[ Cu ] = 0.0001 × [ Cu ] (24.2.3)
te 0

where [Cu2+]0 is the initial concentration of Cu2+ in the sample. Substituting Equation 24.2.3 into Equation 24.2.2 allows us to
calculate the desired potential.
0.05916 1
∘
E =E 2+
− log (24.2.4)
Cu /Cu 2+
2 0.0001 × [ Cu ]

If the initial concentration of Cu2+ is 1.00 × 10 M, for example, then the working electrode’s potential must be more negative
−4

than +0.105 V to quantitatively reduce Cu2+ to Cu. Note that at this potential H3O+ is not reduced to H2, maintaining 100% current
efficiency.

Many controlled-potential coulometric methods for Cu2+ use a potential that is negative relative to the standard hydrogen
electrode—see, for example, Rechnitz, G. A. Controlled-Potential Analysis, Macmillan: New York, 1963, p.49. Based on
Figure 24.2.2 you might expect that applying a potential <0.000 V will partially reduce H3O+ to H2, resulting in a current
efficiency that is less than 100%. The reason we can use such a negative potential is that the reaction rate for the reduction of
H3O+ to H2 is very slow at a Pt electrode. This results in a significant overpotential—the need to apply a potential more
positive or a more negative than that predicted by thermodynamics—which shifts Eo for the H3O+/H2 redox couple to a more
negative value.

Minimizing Electrolysis Time

In controlled-potential coulometry, as shown in Figure 24.2.1, the current decreases over time. As a result, the rate of electrolysis—
recall from Chapter 22 that current is a measure of rate—becomes slower and an exhaustive electrolysis of the analyte may require
a long time. Because time is an important consideration when designing an analytical method, we need to consider the factors that
affect the analysis time.
We can approximate how the current changes as a function of time (Figure 24.2.1 ) as an exponential decay; thus, the current at
time t is
−kt
it = i0 e (24.2.5)

where i0 is the current at t = 0 and k is a rate constant that is directly proportional to the area of the working electrode and the rate
of stirring, and that is inversely proportional to the volume of solution. For an exhaustive electrolysis in which we oxidize or reduce
99.99% of the analyte, the current at the end of the analysis, te, is

ite ≤ 0.0001 × i0 (24.2.6)

Substituting Equation 24.2.6 into Equation 24.2.5 and solving for te gives the minimum time for an exhaustive electrolysis as
1 9.21
te = − × ln(0.0001) = (24.2.7)
k k

From this equation we see that a larger value for k reduces the analysis time. For this reason we usually carry out a controlled-
potential coulometric analysis in a small volume electrochemical cell, using an electrode with a large surface area, and with a high
stirring rate. A quantitative electrolysis typically requires approximately 30–60 min, although shorter or longer times are possible.

24.2.2 https://chem.libretexts.org/@go/page/333854
Instrumentation
We can use the three-electrode potentiostat in Figure (24.2.3) to set and control the potential in controlled-potential coulometry .
The potential of the working electrode is measured relative to a constant-potential reference electrode that is connected to the
working electrode through a high-impedance potentiometer. To set the working electrode’s potential we adjust the slide wire
resistor that is connected to the auxiliary electrode. If the working electrode’s potential begins to drift, we adjust the slide wire
resistor to return the potential to its initial value. The current flowing between the auxiliary electrode and the working electrode is
measured with an ammeter. Of course, a modern potentionstat uses operational amplifiers to maintain the constant potential without
our intervention.

Figure 24.2.3 . Schematic diagram for a manual potentiostat: W is the working electrode; A is the auxiliary electrode; R is the
reference electrode; SW is a slide-wire resistor, E is a high-impendance potentiometer; and i is an ammeter.
The working electrode is usually one of two types: a cylindrical Pt electrode manufactured from platinum-gauze (Figure 24.2.4), or
a Hg pool electrode. The large overpotential for the reduction of H3O+ at Hg makes it the electrode of choice for an analyte that
requires a negative potential. For example, a potential more negative than –1 V versus the SHE is feasible at a Hg electrode—but
not at a Pt electrode—even in a very acidic solution. Because mercury is easy to oxidize, it is less useful if we need to maintain a
potential that is positive with respect to the SHE. Platinum is the working electrode of choice when we need to apply a positive
potential.

Figure 24.2.4 . Example of a cylindrical Pt-gauze electrode used in controlled-potential coulometry. The electrode shown here has a
diameter of 13 mm and a height of 48 mm, and was fashioned from Pt wire with a diameter of approximately 0.15 mm. The
electrode’s surface has 360 openings/cm2 and a total surface area of approximately 40 cm2.
The auxiliary electrode, which often is a Pt wire, is separated by a salt bridge from the analytical solution. This is necessary to
prevent the electrolysis products generated at the auxiliary electrode from reacting with the analyte and interfering in the analysis.
A saturated calomel or Ag/AgCl electrode serves as the reference electrode.
The other essential need for controlled-potential coulometry is a means for determining the total charge. One method is to monitor
the current as a function of time and determine the area under the curve, as shown in Figure 24.2.1. Modern instruments use
electronic integration to monitor charge as a function of time. The total charge at the end of the electrolysis is read directly from a
digital readout.

24.2.3 https://chem.libretexts.org/@go/page/333854
Electrogravimetry
If the product of controlled-potential coulometry forms a deposit on the working electrode, then we can use the change in the
electrode’s mass as the analytical signal. For example, if we apply a potential that reduces Cu2+ to Cu at a Pt working electrode, the
difference in the electrode’s mass before and after electrolysis is a direct measurement of the amount of copper in the sample. An
analytical technique that uses mass as a signal a gravimetric technique; thus, we call this electrogravimetry.

Quantitative Applications
The majority of controlled-potential coulometric analyses involve the determination of inorganic cations and anions, including
trace metals and halides ions. Table 24.2.1 summarizes several of these methods.
Table 24.2.1 . Representative Controlled-Potential Coulometric Analyses for Inorganic Ions
analyte electrolytic reaction electrode

antimony Sb(III) + 3 e
−
⇌ Sb Pt

arsenic As(III) ⇌ As(V) + 2e

−
Pt

cadmium Cd(II) + 2 e
−
⇌ Cd Pt or Hg

cobalt Co(II) + 2 e
−
⇌ Co Pt or Hg

copper Cu(II) + 2 e
−
⇌ Cu Pt or Hg

halides (X–) Ag + X
−
⇌ AgX + e
−
Ag

iron Fe(II) ⇌ Fe(III) + e

−
Pt

lead Pb(II) + 2 e
−
⇌ Pb Pt or Hg

nickel Ni(II) + 2 e
−
⇌ Ni Pt or Hg

plutonium Pu(III) ⇌ Pu(IV) + e

−
Pt

silver Ag(I) + 1 e
−
⇌ Ag Pt

tin Sn(II) + 2 e
−
⇌ Sn Pt

uranium U(VI) + 2 e
−
⇌ U(IV) Pt or Hg

zinc Zn(II) + 2 e
−
⇌ Zn Pt or Hg

Source: Rechnitz, G. A. Controlled-Potential Analysis, Macmillan: New York, 1963.

Electrolytic reactions are written in terms of the change in the analyte’s oxidation state. The actual species in solution depends on the analyte.

The ability to control selectivity by adjusting the working electrode’s potential makes controlled-potential coulometry particularly
useful for the analysis of alloys. For example, we can determine the composition of an alloy that contains Ag, Bi, Cd, and Sb by
dissolving the sample and placing it in a matrix of 0.2 M H2SO4 along with a Pt working electrode and a Pt counter electrode. If we
apply a constant potential of +0.40 V versus the SCE, Ag(I) deposits on the electrode as Ag and the other metal ions remain in
solution. When electrolysis is complete, we use the total charge to determine the amount of silver in the alloy. Next, we shift the
working electrode’s potential to –0.08 V versus the SCE, depositing Bi on the working electrode. When the coulometric analysis
for bismuth is complete, we determine antimony by shifting the working electrode’s potential to –0.33 V versus the SCE,
depositing Sb. Finally, we determine cadmium following its electrodeposition on the working electrode at a potential of –0.80 V
versus the SCE.
We also can use controlled-potential coulometry for the quantitative analysis of organic compounds, although the number of
applications is significantly less than that for inorganic analytes. One example is the six-electron reduction of a nitro group, –NO2,
to a primary amine, –NH2, at a mercury electrode. Solutions of picric acid—also known as 2,4,6-trinitrophenol, or TNP, a close
relative of TNT—is analyzed by reducing it to triaminophenol.

Another example is the successive reduction of trichloroacetate to dichloroacetate, and of dichloroacetate to monochloroacetate

24.2.4 https://chem.libretexts.org/@go/page/333854
 − + − − −
Cl3 CCOO (aq) + H3 O (aq) + 2 e ⇌ Cl2 HCCOO (aq) + Cl (aq) + H2 O(l)

− + − − −
Cl2 HCCOO (aq) +  H 3 O (aq) + 2 e ⇌  ClH2 CCOO (aq) +  Cl (aq) + H2 O(l)

We can analyze a mixture of trichloroacetate and dichloroacetate by selecting an initial potential where only the more easily
reduced trichloroacetate reacts. When its electrolysis is complete, we can reduce dichloroacetate by adjusting the potential to a
more negative potential. The total charge for the first electrolysis gives the amount of trichloroacetate, and the difference in total
charge between the first electrolysis and the second electrolysis gives the amount of dichloroacetate.

 Example 24.2.1

One useful application of controlled-potential coulometry is determining the number of electrons involved in a redox reaction.
To make the determination, we complete a controlled-potential coulometric analysis using a known amount of a pure
compound. The total charge at the end of the electrolysis is used to determine the value of n using Faraday’s law. A 0.3619-g
sample of tetrachloropicolinic acid, C6HNO2Cl4, is dissolved in distilled water, transferred to a 1000-mL volumetric flask, and
diluted to volume. An exhaustive controlled-potential electrolysis of a 10.00-mL portion of this solution at a spongy silver
cathode requires 5.374 C of charge. What is the value of n for this reduction reaction?

Solution
The 10.00-mL portion of sample contains 3.619 mg, or 1.39 × 10 −5
mol of tetrachloropicolinic acid. Solving for n gives
Q 5.374 C −
n = = = 4.01  mol e /mol C6 HNO2 Cl4
− −5
F NA (96478 C/mol e ) (1.39 × 10  mol C6 HNO2 Cl4 )

Thus, reducing a molecule of tetrachloropicolinic acid requires four electrons. The overall reaction, which results in the
selective formation of 3,6-dichloropicolinic acid, is

This page titled 24.2: Controlled-Potential Coulometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

24.2.5 https://chem.libretexts.org/@go/page/333854
24.3: Controlled-Current Coulometry
A second approach to coulometry is to use a constant current in place of a constant potential, which results in the current-versus-time
profile shown in Figure 24.3.1. Controlled-current coulometry has two advantages over controlled-potential coulometry. First, the
analysis time is shorter because the current does not decrease over time. A typical analysis time for controlled-current coulometry is less
than 10 min, compared to approximately 30–60 min for controlled-potential coulometry. Second, because the total charge is simply the
product of current and time, there is no need to integrate the current-time curve in Figure 24.3.1.

Figure 24.3.1 . Current versus time for a controlled-current coulometric analysis. The measured current is shown by the red curve. The
integrated area under the curve, shown in blue, is the total charge.
Using a constant current presents us with two important experimental problems. First, during electrolysis the analyte’s concentration—
and, therefore, the current that results from its oxidation or reduction—decreases continuously. To maintain a constant current we must
allow the potential to change until another oxidation reaction or reduction reaction occurs at the working electrode. Unless we design the
system carefully, this secondary reaction results in a current efficiency that is less than 100%. The second problem is that we need a
method to determine when the analyte's electrolysis is complete. In a controlled-potential coulometric analysis we know that electrolysis
is complete when the current reaches zero, or when it reaches a constant background or residual current. In a controlled-current
coulometric analysis, however, current continues to flow even when the analyte’s electrolysis is complete. A suitable method for
determining the reaction’s endpoint, te, is needed.

Maintaining Current Efficiency

To illustrate why a change in the working electrode’s potential may result in a current efficiency of less than 100%, let’s consider the
coulometric analysis for Fe2+ based on its oxidation to Fe3+ at a Pt working electrode in 1 M H2SO4.
2+ 3+ −
Fe (aq) ⇌  Fe (aq) + e (24.3.1)

Figure 24.3.2 shows the relevant potentials for this system. At the beginning of the analysis, the potential of the working electrode
remains nearly constant at a level near its initial value.

Figure 24.3.2 . Potential for the constant-current coulometric analysis of Fe2+. The red arrow and text shows how the potential drifts to
more positive values, decreasing the current efficiency.
As the concentration of Fe2+ decreases and the concentration of Fe3+ increases, the working electrode’s potential shifts toward more
positive values until the oxidation of H2O begins.

24.3.1 https://chem.libretexts.org/@go/page/333856
 + −
2 H2 O(l) ⇌  O 2 (g) + 4 H (aq) + 4 e (24.3.2)

Because a portion of the total current comes from the oxidation of H2O, the current efficiency for the analysis is less than 100% and we
cannot use the equation Q = it to determine the amount of Fe2+ in the sample.
Although we cannot prevent the potential from drifting until another species undergoes oxidation, we can maintain a 100% current
efficiency if the product of that secondary oxidation reaction both rapidly and quantitatively reacts with the remaining Fe2+. To
accomplish this we add an excess of Ce3+ to the analytical solution. As shown in Figure 24.3.3, when the potential of the working
electrode shifts to a more positive potential, Ce3+ begins to oxidize to Ce4+
3+ 4+ −
Ce (aq) ⇌  Ce (aq) + e (24.3.3)

Figure 24.3.3 . Standard state reduction potentials for the constant-current coulometric analysis of Fe2+ in the presence of a Ce3+
mediator. As the potential drifts to more positive values, we eventually reach a potential where Ce3+ undergoes oxidation. Because Ce4+,
the product of the oxidation of Ce3+, reacts with Fe2+, we maintain current efficiency.
The Ce4+ that forms at the working electrode rapidly mixes with the solution where it reacts with any available Fe2+.
4+ 2+ 3+ 3+
Ce (aq) +  Fe (aq) ⇌  Ce (aq) +  Fe (aq) (24.3.4)

2+ 3+
Combining reaction 24.3.3 and reaction 24.3.4 shows that the net reaction is the oxidation of Fe to Fe
2+ 3+ −
Fe (aq) ⇌  Fe (aq) + e (24.3.5)

which maintains a current efficiency of 100%. A species used to maintain 100% current efficiency is called a mediator.

Endpoint Determination
Adding a mediator solves the problem of maintaining 100% current efficiency, but it does not solve the problem of determining when
the analyte's electrolysis is complete. Using the analysis for Fe2+ in Figure 24.3.3, when the oxidation of Fe2+ is complete current
continues to flow from the oxidation of Ce3+, and, eventually, the oxidation of H2O. What we need is a signal that tells us when no more
Fe2+ is present in the solution.
For our purposes, it is convenient to treat a controlled-current coulometric analysis as a reaction between the analyte, Fe2+, and the
mediator, Ce3+, as shown by reaction 24.3.4. This reaction is identical to a redox titration; thus, we can use the end points for a redox
titration—visual indicators and potentiometric or conductometric measurements—to signal the end of a controlled-current coulometric
analysis. For example, ferroin provides a useful visual endpoint for the Ce3+ mediated coulometric analysis for Fe2+, changing color
from red to blue when the electrolysis of Fe2+ is complete.

Instrumentation
We can carry out controlled-current coulometry using the two-electrode galvanostat shown in Figure 24.3.4, which consists of a working
electrode and a counter electrode. The working electrode—often a simple Pt electrode—also is called the generator electrode since it is
where the mediator reacts to generate the species that reacts with the analyte. If necessary, the counter electrode is isolated from the
analytical solution by a salt bridge or a porous frit to prevent its electrolysis products from reacting with the analyte. The current from
the power supply through the working electrode is
EPS
i = (24.3.6)
R + Rcell

24.3.2 https://chem.libretexts.org/@go/page/333856
where EPS is the potential of the power supply, R is the resistance of the resistor, and Rcell is the resistance of the electrochemical cell. If
R >> Rcell, then the current between the auxiliary and working electrodes
EPS
i = ≈ constant (24.3.7)
R

maintains a constant value. To monitor the working electrode’s potential, which changes as the composition of the electrochemical cell
changes, we can include an optional reference electrode and a high-impedance potentiometer.

Figure 24.3.4 . Schematic diagram of a galvanostat: A is the auxiliary, or counter electrode; W is the working electrode; R is an optional
reference electrode, E is a high-impedance potentiometer, and i is an ammeter. The working electrode and the optional reference
electrode are connected to a ground. Although a modern galvanostat uses very different circuitry, we can use this figure and the
accompanying discussion to understand how we can use the working electrode and the counter electrode to control the current. The
galvanostat here includes an optional reference electrode, but its presence or absence is not important if we are not interested in
monitoring the working electrode’s potential.
Alternatively, we can generate the oxidizing agent or the reducing agent externally, and allow it to flow into the analytical solution.
Figure 24.3.5 shows one simple method for accomplishing this. A solution that contains the mediator flows into a small-volume
electrochemical cell with the products exiting through separate tubes. Depending upon the analyte, the oxidizing agent or the reducing
reagent is delivered to the analytical solution. For example, we can generate Ce4+ using an aqueous solution of Ce3+, directing the Ce4+
that forms at the anode to our sample.

Figure 24.3.5 . One example of a device for the external generation of oxidizing agents and reducing agents for controlled-current
coulometry. A solution containing the mediator flows into a small-volume electrochemical cell. The resulting oxidation products, which
form at the anode, flow to the right and serve as an oxidizing agent. Reduction at the cathode generates a reducing agent.
There are two other crucial needs for controlled-current coulometry: an accurate clock for measuring the electrolysis time, te, and a
switch for starting and stopping the electrolysis. An analog clock can record time to the nearest ±0.01 s, but the need to stop and start the
electrolysis as we approach the endpoint may result in an overall uncertainty of ±0.1 s. A digital clock allows for a more accurate
measurement of time, with an overall uncertainty of ±1 ms. The switch must control both the current and the clock so that we can make
an accurate determination of the electrolysis time.

24.3.3 https://chem.libretexts.org/@go/page/333856
Coulometric Titrations
A controlled-current coulometric method sometimes is called a coulometric titration because of its similarity to a conventional titration.
For example, in the controlled-current coulometric analysis for Fe2+ using a Ce3+ mediator, the oxidation of Fe2+ by Ce4+ (reaction
24.3.4) is identical to the reaction in a redox titration.

There are other similarities between controlled-current coulometry and titrimetry. If we combine the equation Q = nF Na and the
equation Q = it and solve for the moles of analyte, NA, we obtain the following equation.
e

i
NA = × te (24.3.8)
nF

Compare Equation 24.3.8 to the relationship between the moles of analyte, NA, and the moles of titrant, NT, in a titration
NA = NT = MT × VT (24.3.9)

where MT and VT are the titrant’s molarity and the volume of titrant at the end point. In constant-current coulometry, the current source is
equivalent to the titrant and the value of that current is analogous to the titrant’s molarity. Electrolysis time is analogous to the volume of
titrant, and te is equivalent to the a titration’s end point. Finally, the switch for starting and stopping the electrolysis serves the same
function as a buret’s stopcock.

For simplicity, we assumed above that the stoichiometry between the analyte and titrant is 1:1. The assumption, however, is not
important and does not effect our observation of the similarity between controlled-current coulometry and a titration.

Quantitative Applications
The use of a mediator makes a coulometric titration a more versatile analytical technique than controlled-potential coulometry. For
example, the direct oxidation or reduction of a protein at a working electrode is difficult if the protein’s active redox site lies deep within
its structure. A coulometric titration of the protein is possible, however, if we use the oxidation or reduction of a mediator to produce a
solution species that reacts with the protein. Table 24.3.1 summarizes several controlled-current coulometric methods based on a redox
reaction using a mediator.
Table 24.3.1 . Representative Examples of Coulometric Redox Titrations
mediator electrochemically generated reagent and reaction representative application
2+
H2 C2 O4 (aq) + 2 Ag (aq) + 2 H2 O(l) ⇌
Ag+ Ag
+
⇌ Ag
2+
+e
−

+ +
2 CO 2 (g) + 2 Ag (aq) + 2 H3 O (aq)

H2 S(aq) +  Br2 (aq) + 2 H2 O(l) ⇌

Br– 2 Br
−
⇌ Br 2 + 2 e
−

− +
S(s) + 2 Br (aq) + 2 H3 O (aq)

4– 4+
Fe(CN) (aq) +  Ce (aq) ⇌
Ce3+ Ce
3+
⇌ Ce
4+
+e
− 6

3+
3−
Fe(CN ) (aq) +  Ce (aq)
6

Cl– 2 Cl
−
⇌ Cl 2 + 2 e
−
Ti(I)(aq) +  Cl2 (aq) ⇌ Ti(III)(aq) + 2 Cl
−
(aq)

2− 2+ +
Cr 2 O (aq) + 6 Fe (aq) + 14 H3 O (aq) ⇌
Fe3+ 3+ − 2+ 7
Fe +e ⇌ Fe
3+ 3+
2 Cr (aq) + 6 Fe (aq) + 21 H2 O(l)

I– 3I
−
⇌ I
–
3
+ 2e
−
2 S2 O
2−
3
(aq) + I
−
3
(aq) ⇌ S4 O
2−

6
(aq) + 3 I
−
(aq)

Mn2+ Mn
2+
⇌ Mn
3+
+e
−
As(III)(aq) + 2 Mn
3+
(aq) ⇌ As(V)(aq) + 2 Mn
2+
(a

Note: The electrochemically generated reagent and the analyte are shown in bold.

For an analyte that is not easy to oxidize or reduce, we can complete a coulometric titration by coupling a mediator’s oxidation or
reduction to an acid–base, precipitation, or complexation reaction that involves the analyte. For example, if we use H2O as a mediator,
we can generate H3O+at the anode
+ −
6 H2 O(l) ⇌ 4 H3 O (aq) +  O 2 (g) + 4 e

and generate OH– at the cathode.

− −
2 H2 O(l) + 2 e ⇌ 2 OH (aq) +  H 2 (g)

24.3.4 https://chem.libretexts.org/@go/page/333856
If we carry out the oxidation or reduction of H2O using the generator cell in Figure 24.3.5, then we can selectively dispense H3O+ or
OH– into a solution that contains the analyte. The resulting reaction is identical to that in an acid–base titration. Coulometric acid–base
titrations have been used for the analysis of strong and weak acids and bases, in both aqueous and non-aqueous matrices. Table 24.3.2
summarizes several examples of coulometric titrations that involve acid–base, complexation, and precipitation reactions.
Table 24.3.2 . Representative Coulometric Titrations Using Acid–Base, Complexation, and Precipitation Reactions
electrochemically generated reagent
type of reaction mediator and reaction representative application

acid-base H2O 6 H2 O ⇌ 4 H3 O
+
+  O2 + e
−
OH
−
(aq) +  H3 O
+
(aq) ⇌ 2 H2 O(l)

acid-base H2O 2 H2 O + 2 e
−
⇌ 2 OH
−
+  H2 H3 O
+
(aq) +  OH
−
(aq) ⇌ 2 H2 O(l)

2− + − 2+ 3−
HgNH Y +  NH + 2e ⇌ Ca (aq) +  HY (aq) +  H2 O(l) ⇌
complexation HgNH3Y2– (Y = EDTA) 3 4
2− +
3−
HY +  Hg + 2 NH3 CaY (aq) +  H3 O (aq)

precipitation Ag Ag ⇌  Ag
+
+e
−
I
−
(aq) +  Ag
+
(aq) ⇌ Ag I(s)

precipitation Hg 2Hg ⇌ Hg
2+
2
+ 2e
−
2 Cl
−
(aq) +  Hg
2+
2
(aq) ⇌  Hg Cl 2 (s)
2

2+ + 4−
3− 4− 3 Zn (aq) + K (aq) + 2 Fe(CN) (aq)
precipitation Fe(CN)
6
Fe(CN )
3−
6
+e
−
⇌  Fe(CN)
6
6

K2 Zn3 [Fe(CN) ] (s)

6 2

Note: The electrochemically generated reagent and the analyte are shown in bold.

In comparison to a conventional titration, a coulometric titration has two important advantages. The first advantage is that
electrochemically generating a titrant allows us to use a reagent that is unstable. Although we cannot prepare and store a solution of a
highly reactive reagent, such as Ag2+ or Mn3+, we can generate them electrochemically and use them in a coulometric titration. Second,
because it is relatively easy to measure a small quantity of charge, we can use a coulometric titration to determine an analyte whose
concentration is too small for a conventional titration.
The following example shows the calculations for a typical coulometric analysis.

 Example 24.3.1

To determine the purity of a sample of Na2S2O3, a sample is titrated coulometrically using I– as a mediator and I as the titrant. A −

sample weighing 0.1342 g is transferred to a 100-mL volumetric flask and diluted to volume with distilled water. A 10.00-mL
portion is transferred to an electrochemical cell along with 25 mL of 1 M KI, 75 mL of a pH 7.0 phosphate buffer, and several drops
of a starch indicator solution. Electrolysis at a constant current of 36.45 mA requires 221.8 s to reach the starch indicator endpoint.
Determine the sample’s purity.

Solution
As shown in Table 24.3.1, the coulometric titration of S 2−
2 O3 with I is −

2− − 2− −
2 S2 O (aq) +  I (aq) ⇌  S 4 O (aq) + 3 I (aq)
3 3 6

The oxidation of S O to S O requires one electron per

2
2−

3 4
2−

6
S2 O
2−

3
(n = 1). Combining the equations Q = nF NA and Q = ite ,
and solving for the moles and grams of Na2S2O3 gives
ite (0.03645 A)(221.8 s)
−5
NA = = = 8.379 × 10  mol Na2 S2 O3
nF 1 mol e
−
96487 C
( )( −
)
mol Na2 S2 O3 mol e

This is the amount of Na2S2O3 in a 10.00-mL portion of a 100-mL sample; thus, there are 0.1325 grams of Na2S2O3 in the original
sample. The sample’s purity, therefore, is
0.1325 g Na 2 S2 O3
× 100 = 98.73% w/w Na2 S2 O3
0.1342 g sample 

Note that for this calcuation, it does not matter whether S 2 O3

2−
is oxidized at the working electrode or is oxidized by I . −
3

This page titled 24.3: Controlled-Current Coulometry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

24.3.5 https://chem.libretexts.org/@go/page/333856
 CHAPTER OVERVIEW
25: Voltammetry
In voltammetry we apply a time-dependent potential to an electrochemical cell and measure the resulting current as a function of
that potential. We call the resulting plot of current versus applied potential a voltammogram, and it is the electrochemical
equivalent of a spectrum in spectroscopy, providing quantitative and qualitative information about the species involved in the
oxidation or reduction reaction [Maloy, J. T. J. Chem. Educ. 1983, 60, 285–289]. The earliest voltammetric technique is
polarography, developed by Jaroslav Heyrovsky in the early 1920s—an achievement for which he was awarded the Nobel Prize in
Chemistry in 1959. Since then, many different forms of voltammetry have been developed. Before we examine some these
techniques and their applications in more detail, we must first consider the basic experimental design for voltammetry and the
factors influencing the shape of the resulting voltammogram.
25.1: Potential Excitation Signals and Currents in Voltammetry
25.2: Voltammetric Instrumentation
25.3: Linear Sweep Voltammetry
25.4: Cyclic Voltammetry
25.5: Polarography
25.6: Stripping Methods
25.7: Applications of Voltammetry

This page titled 25: Voltammetry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

1
25.1: Potential Excitation Signals and Currents in Voltammetry
In voltammetry we apply a time-dependent potential to an electrochemical cell and measure the resulting current as a function of
that potential.

Potential Excitation Signals

As shown in Figure 25.1.1, the potential may consist of (a) a linear scan or (b) a series of pulses. For the linear scan in (a), the
direction of the scan can be reversed and repeated for additional cycles. The series of pulses in (b) shows just one of several
different pulsed potential excitation signals; we will consider other pulse trains in the section on polarography.

Figure 25.1.1 : Two examples of potential excitation signals and three examples of current responses. In (a) the potential varies
linearly with time and, as shown by the dashed line, the direction of the scan may be changed and repeated. In (b) the potential is
changed by applying a series of pulses. The nature of the current response depends on both the potential excitation signal and other
experimental conditions, such as if the solution is stirred.

Current
The current responses in Figure 25.1.1 show the three common types of signals. In (c) and (d) the current is monitored directly as
the potential is changed. In (e) a change in current is recorded using the current immediately before and after the application of a
potential pulse. The current itself has three components: faradic current from the oxidation or reduction of the analyte, a charging
current, and residual currents.

Faradaic Current
Faradic current is the result of oxidation or reduction of the analyte at the working electrode. The ease with which electrons move
between the electrode and the species that reacts at the electrode affects the faradiac current. When electron transfer kinetics are
fast, the redox reaction is at equilibrium. Under these conditions the redox reaction is electrochemically reversible and the Nernst
equation applies. If the electron transfer kinetics are sufficiently slow, the concentration of reactants and products at the electrode
surface—and thus the magnitude of the faradaic current—are not what is predicted by the Nernst equation. In this case the system
is electrochemically irreversible.

Charging Currents
In addition to the faradaic current from a redox reaction, the current in an electrochemical cell includes other, nonfaradaic sources.
Suppose the charge on an electrode is zero and we suddenly change its potential so that the electrode’s surface acquires a positive
charge. Cations near the electrode’s surface will respond to this positive charge by migrating away from the electrode; anions, on
the other hand, will migrate toward the electrode. This migration of ions occurs until the electrode’s positive surface charge and the
negative charge of the solution near the electrode are equal. Because the movement of ions and the movement of electrons are

25.1.1 https://chem.libretexts.org/@go/page/335293
indistinguishable, the result is a small, short-lived nonfaradaic current that we call the charging current. Every time we change the
electrode’s potential, a transient charging current flows.
The migration of ions in response to the electrode’s surface charge leads to the formation of a structured electrode-solution
interface that we call the electrical double layer, or EDL. When we change an electrode’s potential, the charging current is the result
of a restructuring of the EDL. The exact structure of the electrical double layer is not important in the context of this text, but you
can consult this chapter’s additional resources for additional information. See Chapter 22.1 for additional details.

Residual Current
Even in the absence of analyte, a small, measurable current flows through an electrochemical cell. In addition to the charging
current discussed above, the residual current includes a faradaic current from the oxidation or reduction of trace impurities in the
sample. Methods for discriminating between the analyte’s faradaic current and the residual current are discussed later in this
chapter.

This page titled 25.1: Potential Excitation Signals and Currents in Voltammetry is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

25.1.2 https://chem.libretexts.org/@go/page/335293
25.2: Voltammetric Instrumentation
Although early voltammetric methods used only two electrodes, a modern voltammeter makes use of a three-electrode potentiostat,
such as that shown in Figure 25.2.1. The potential of the working electrode is measured relative to a constant-potential reference
electrode that is connected to the working electrode through a high-impedance potentiometer. The auxiliary electrode generally is a
platinum wire and the reference electrode usually is a SCE or a Ag/AgCl electrode. We apply a time-dependent potential excitation
signal to the working electrode—changing its potential relative to the fixed potential of the reference electrode—and measure the
current that flows between the working electrode and the auxiliary electrode. Modern potentiostats include waveform generators
that allow us to apply a time-dependent potential profile, such as a series of potential pulses, to the working electrode.

Figure 25.2.1 . Schematic diagram for a manual potentiostat: W is the working electrode; A is the auxiliary electrode; R is the
reference electrode; SW is a slide-wire resistor, E is a high-impendance potentiometer; and i is an ammeter.

Working Electrodes
For the working electrode we can choose among several different materials, including mercury, platinum, gold, silver, and carbon.
The earliest voltammetric techniques used a mercury working electrode. Because mercury is a liquid, the working electrode usually
is a drop suspended from the end of a capillary tube. In the hanging mercury drop electrode, or HMDE, we extrude the drop of Hg
by rotating a micrometer screw that pushes the mercury from a reservoir through a narrow capillary tube (Figure 25.2.2a).

Figure 25.2.2 . Three examples of mercury electrodes: (a) hanging mercury drop electrode, or HMDE; (b) dropping mercury
electrode, or DME; and (c) static mercury drop electrode, or SMDE.
In the dropping mercury electrode, or DME, mercury drops form at the end of the capillary tube as a result of gravity (Figure
25.2.2b). Unlike the HMDE, the mercury drop of a DME grows continuously—as mercury flows from the reservoir under the

influence of gravity—and has a finite lifetime of several seconds. At the end of its lifetime the mercury drop is dislodged, either
manually or on its own, and is replaced by a new drop. The static mercury drop electrode, or SMDE, uses a solenoid driven plunger
to control the flow of mercury (Figure 25.2.2c). Activation of the solenoid momentarily lifts the plunger, allowing mercury to flow

25.2.1 https://chem.libretexts.org/@go/page/335295
through the capillary, forming a single, hanging Hg drop. Repeated activation of the solenoid produces a series of Hg drops. In this
way the SMDE may be used as either a HMDE or a DME. There is one additional type of mercury electrode: the mercury film
electrode. A solid electrode—typically carbon, platinum, or gold—is placed in a solution of Hg2+ and held at a potential where the
reduction of Hg2+ to Hg is favorable, depositing a thin film of mercury on the solid electrode’s surface.
Mercury has several advantages as a working electrode. Perhaps its most important advantage is its high overpotential for the
reduction of H3O+ to H2, which makes accessible potentials as negative as –1 V versus the SCE in acidic solutions and –2 V versus
the SCE in basic solutions (Figure 25.2.3). A species such as Zn2+, which is difficult to reduce at other electrodes without
simultaneously reducing H3O+, is easy to reduce at a mercury working electrode. Other advantages include the ability of metals to
dissolve in mercury—which results in the formation of an amalgam—and the ability to renew the surface of the electrode by
extruding a new drop. One limitation to mercury as a working electrode is the ease with which it is oxidized. Depending on the
solvent, a mercury electrode can not be used at potentials more positive than approximately –0.3 V to +0.4 V versus the SCE.

Figure 25.2.3 . Approximate potential windows for mercury, platinum, and carbon (graphite) electrodes in acidic, neutral, and basic
aqueous solvents. The useful potential windows are shown in green; potentials in red result in the oxidation or the reduction of the
solvent or the electrode. Complied from Adams, R. N. Electrochemistry at Solid Electrodes, Marcel Dekker, Inc.: New York, 1969
and Bard, A. J.; Faulkner, L. R. Electro- chemical Methods, John Wiley & Sons: New York, 1980.
Solid electrodes constructed using platinum, gold, silver, or carbon may be used over a range of potentials, including potentials that
are negative and positive with respect to the SCE (Figure 25.2.3). For example, the potential window for a Pt electrode extends
from approximately +1.2 V to –0.2 V versus the SCE in acidic solutions, and from +0.7 V to –1 V versus the SCE in basic
solutions. A solid electrode can replace a mercury electrode for many voltammetric analyses that require negative potentials, and is
the electrode of choice at more positive potentials. Except for the carbon paste electrode, a solid electrode is fashioned into a disk
and sealed into the end of an inert support with an electrical lead (Figure 25.2.4). The carbon paste electrode is made by filling the
cavity at the end of the inert support with a paste that consists of carbon particles and a viscous oil. Solid electrodes are not without
problems, the most important of which is the ease with which the electrode’s surface is altered by the adsorption of a solution
species or by the formation of an oxide layer. For this reason a solid electrode needs frequent reconditioning, either by applying an
appropriate potential or by polishing.

Figure 25.2.4 . Schematic showing a solid electrode. The electrode is fashioned into a disk and sealed in the end of an inert polymer
support along with an electrical lead.

Electrochemical Cells
A typical arrangement for a voltammetric electrochemical cell is shown in Figure 25.2.5. In addition to the working electrode, the
reference electrode, and the auxiliary electrode, the cell also includes a N2-purge line for removing dissolved O2, and an optional

25.2.2 https://chem.libretexts.org/@go/page/335295
stir bar. Electrochemical cells are available in a variety of sizes, allowing the analysis of solution volumes ranging from more than
100 mL to as small as 50 μL.

Figure 25.2.5 . Typical electrochemical cell for voltammetry.

This page titled 25.2: Voltammetric Instrumentation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

25.2.3 https://chem.libretexts.org/@go/page/335295
25.3: Linear Sweep Voltammetry
In the simplest voltammetric experiment we apply a linear potential ramp as an excitation signal and record the current that flows
in response to the change in potential. Among the experimental variables under our control are the initial potential, the final
potential, the scan rate, and whether we choose to stir the solution or leave it unstirred. We call this linear sweep voltammetry.
To illustrate how linear sweep voltammetry works, let's consider what happens when we reduce Fe(CN) to Fe(CN) at the 3−

6
4−

working electrode. The relationship between the concentrations of Fe(CN) , the concentration of Fe(CN) , and the potential is
3−

6
4−

given by the Nernst equation

4−
[Fe(CN) ]
6 x=0
E = +0.356 V − 0.05916 log (25.3.1)
3−
[Fe(CN) ]
6 x=0

where +0.356V is the standard-state potential for the Fe(CN) /Fe(CN) 6

redox couple, and x = 0 indicates that the
3− 4−

6
3− 4−
concentrations of Fe(CN) and Fe(CN) are those at the surface of the working electrode. We use surface concentrations
6 6

instead of bulk concentrations because the equilibrium position for the redox reaction
3− − 4−
Fe(CN) (aq) + e ⇌ Fe(CN) (aq) (25.3.2)
6 6

is established at the electrode’s surface.

Let’s assume we have a solution for which the initial concentration of Fe(CN) is 1.0 mM and that Fe(CN) is absent. Figure
3−

6
4−

25.3.1 shows the relationship between the applied potential and the species that are stable at the electrode's surface.

Figure 25.3.1 . Potential diagram for the Fe(CN) /Fe(CN) redox half-reaction showing the potentials where Fe(CN) and
3− 4− 3−

6 6 6

where Fe(CN) are the dominate species. At a potential of +0.30 V, only Fe(CN) is stable at the electrode's surface.
4− 3−

6 6

If we apply a potential of +0.530 V to the working electrode, the concentrations of Fe(CN) and Fe(CN) at the surface of the 3−

6
4−

6
3−
electrode are unaffected, and no faradaic current is observed. If we switch the potential to +0.356 V some of the Fe(CN) at the 6

electrode’s surface is reduced to Fe(CN) until we reach a condition where

4−

3− 4−
[Fe(CN) ] = [Fe(CN) ] = 0.50 mM (25.3.3)
6 x=0 6 x=0

If this is all that happens after we apply the potential, then there would be a brief surge of faradaic current that quickly returns to
zero, which is not the most interesting of results (although this is the basis for chronoamperometry, an electrochemical method we
will not consider in this text). Although the concentrations of Fe(CN) and Fe(CN) at the electrode surface are 0.50 mM,
3−

6
4−

their concentrations in bulk solution remains unchanged.

Because of this difference in concentration, there is a concentration gradient between the electrode’s surface and the bulk solution.
This concentration gradient creates a driving force that transports Fe(CN) away from the electrode and that transports
4−

6
3− 3− 4−
Fe(CN)
6
to the electrode (Figure 25.3.2). As the Fe(CN) arrives at the electrode it, too, is reduced to Fe(CN) . A faradaic
6 6

current continues to flow until there is no difference between the concentrations of Fe(CN) and Fe(CN) at the electrode and 3−

6
4−

their concentrations in bulk solution (although this might take a long time!).

25.3.1 https://chem.libretexts.org/@go/page/335297
 Figure . Schematic diagram showing the transport of Fe(CN) away from the electrode’s surface and the transport of
4−
25.3.2
6

toward the electrode’s surface following the reduction of Fe(CN) to Fe(CN) .

3− 3− 4−
Fe(CN)
6 6 6

Although the potential at the working electrode determines if a faradaic current flows, the magnitude of the current is determined
by the rate of the resulting oxidation or reduction reaction. Two factors contribute to the rate of the electrochemical reaction: the
rate at which the reactants and products are transported to and from the electrode—what we call mass transport—and the rate at
which electrons pass between the electrode and the reactants and products in solution.

Concentration Profiles at the Working Electrode

There are three modes of mass transport that affect the rate at which reactants and products move toward or away from the
electrode surface: diffusion, migration, and convection. Diffusion occurs whenever the concentration of an ion or a molecule at the
surface of the electrode is different from that in bulk solution. If we apply a potential sufficient to completely reduce Fe(CN) at 3−

the electrode surface, the result is a concentration gradient similar to that shown in Figure 25.3.3. The region of solution over
which diffusion occurs is the diffusion layer. In the absence of other modes of mass transport, the width of the diffusion layer, δ ,
3−
increases with time as the Fe(CN) must diffuse from an increasingly greater distance.
6

Figure 25.3.3 . Concentration gradients (in red) for Fe(CN) following the application of a potential that completely reduces it to
3−

. Before we apply the potential (t = 0) the concentration of Fe(CN) is the same at all distances from the electrode’s
4− 3−
Fe(CN)
6 6

surface. After we apply the potential, its concentration at the electrode’s surface decreases to zero and Fe(CN) diffuses to the
3−

electrode from bulk solution. The longer we apply the potential, the greater the distance over which diffusion occurs. The dashed
red line shows the extent of the diffusion layer at time t3. These profiles assume that convection and migration do not contribute
significantly to the mass transport of Fe(CN) .
3−

Convection occurs when we mix the solution, which carries reactants toward the electrode and removes products from the
electrode. The most common form of convection is stirring the solution with a stir bar; other methods include rotating the electrode
and incorporating the electrode into a flow-cell.
The final mode of mass transport is migration, which occurs when a charged particle in solution is attracted to or repelled from an
electrode that carries a surface charge. If the electrode carries a positive charge, for example, an anion will move toward the
electrode and a cation will move toward the bulk solution. Unlike diffusion and convection, migration affects only the mass
transport of charged particles.
The movement of material to and from the electrode surface is a complex function of all three modes of mass transport. In the limit
where diffusion is the only significant form of mass transport, the current, i, in a voltammetric cell is proportional to the slope of

25.3.2 https://chem.libretexts.org/@go/page/335297
the concentration profile in Figure 25.3.3
∂C
i ∝ (25.3.4)
∂x

where C is the concentration of Fe(CN) 3−

6
and x is distance.
For Equation 25.3.4 to be valid, convection and migration must not interfere with the formation of a diffusion layer. We can
eliminate migration by adding a high concentration of an inert supporting electrolyte. Because ions of similar charge are equally
attracted to or repelled from the surface of the electrode, each has an equal probability of undergoing migration. A large excess of
an inert electrolyte ensures that few reactants or products experience migration. Although it is easy to eliminate convection by not
stirring the solution, there are experimental designs where we cannot avoid convection, either because we must stir the solution or
because we are using an electrochemical flow cell. Fortunately, as shown in Figure 25.3.4, the dynamics of a fluid moving past an
electrode results in a small diffusion layer—typically 1–10 μm in thickness—in which the rate of mass transport by convection
drops to zero.

Figure 25.3.4 . Concentration gradient for Fe(CN) when stirring the solution. Diffusion is the only significant form of mass
3−

transport close to the electrode’s surface. At distances greater than δ, convection is the only significant form of mass transport,
maintaining a homogeneous solution in which the concentration of Fe(CN) at δ is the same as its concentration in bulk solution.
3−

Concentration Profiles in an Unstirred Solution

Figure 25.3.5 shows the linear sweep voltammogram (the center image, which shows the current as a function of time) and eight
3− 4−
snapshots of the concentration profiles for the reduction of Fe(CN) to Fe(CN) in an unstirred solution. The initial potential
6 6

was set to +0.530 V and the final potential was set to +0.182 V with a scan rate of 0.050 V/s.
At the initial potential, only Fe(CN) is stable at the electrode surface, and no current flows. After 0.696 s the potential is 0.495
3−

6
3−
V (image to the left of the linear sweep voltammogram) and, because Fe(CN) remains stable at the electrode surface, no current
6

flows. Moving clockwise around the linear sweep voltammogram, the applied potential becomes smaller and the concentration of
Fe(CN)
3−

6
at the electrode surface decreases and the concentration of Fe(CN) increases. Initially the slope of the concentration
4−

gradient, and, therefore, the current increases; as the concentration of Fe(CN) at the electrode surface approaches zero, however,
3−

the concentration gradient becomes less steep and the current decreases. The result is the linear sweep voltammogram in the center
of the diagram.

25.3.3 https://chem.libretexts.org/@go/page/335297
 Figure 25.3.5 : The central image shows the linear sweep voltammogram in an unstirred solution for the reduction of Fe(CN) to
3−

. The remaining images show the concentration profiles for Fe(CN) (in blue) and for Fe(CN) (in red) as the
4− 3− 4−
Fe(CN)
6 6 6

potential is swept from +0.530 V to +0.182 V (arranged clockwise, beginning immediately left of the linear sweep voltammogram.

Concentration Profiles in a Stirred Solution

If we run the same experiment as in Figure 25.3.5, but stir the solution, the resulting linear sweep voltammogram and
concentration profiles are those in Figure 25.3.6. Stirring the solution, as we saw in Figure 25.3.4 creates a diffusion layer whose
thickness is independent of time.

Figure : The central image shows the linear sweep voltammogram in a stirred solution for the reduction of Fe(CN) to
3−
25.3.6
6

. The remaining images show the concentration profiles for Fe(CN) (in blue) and for Fe(CN) (in red) as the
4− 3− 4−
Fe(CN)
6 6 6

potential is swept from +0.530 V to +0.182 V (arranged clockwise, beginning immediately left of the linear sweep voltammogram.

25.3.4 https://chem.libretexts.org/@go/page/335297
As a result, instead of the peak current in Figure 25.3.5, the current reaches a steady-state value, which we call the limiting current,
i . The linear sweep voltammogram also has a characteristic half-wave potential, E
l , when the current is 50% of the limiting1/2

current. Figure 25.3.7 shows how the limiting current and half-wave potential are measured.

Figure 25.3.7 : The limiting current and half-wave potential for a linear sweep voltammetry experiment in a stirred solution.

Voltammetric Currents
Earlier we noted, in Equation 25.3.4, that the current in linear sweep voltammetry is proportional to the slope of the concentration
3− 4−
profile. The current is also a function of other variables, as shown here for the reduction of Fe(CN) to Fe(CN) 6 6

3 − 3 −
nF AD ([Fe(CN) ] − [Fe(CN) ] )
6 6
bulk x =0
i = (25.3.5)
δ

where n the number of electrons in the redox reaction, F is Faraday’s constant, A is the area of the electrode, D is the diffusion
3− 3 − 3 −
coefficient for Fe(CN) , δ is the thickness of the diffusion layer, and ([Fe(CN)
6 6
] − [Fe(CN)
6
] ) is the difference
bulk x =0

in the concentration of Fe(CN) 3 −

6
between the bulk solution and the electrode's surface.
Because n , F , A , and D are constants, and because δ is a constant if we stir the solution, we can write Equation 25.3.5 as
3 − 3 −
i =K 3 − ([Fe(CN) ] − [Fe(CN) ] ) (25.3.6)
Fe(CN) 6 6
6
bulk x =0

where K 3 −
Fe(CN)6
is a constant. If we use the limiting current, then [Fe(CN) 3 −

6
] is zero, and Equation 25.3.6 becomes
x =0

3 −
il = K 3 − [Fe(CN) ] (25.3.7)
Fe(CN)6 6
bulk

Current/Voltage Relationships for Reversible Reactions

A reversible electrochemical reaction is one in which the concentration of the oxidized and reduced species at the electrode surface
remain in thermodynamic equilibrium with each other. When this is true, the Nernst equation explains the relationship between the
applied potential, their concentration, and the standard state potential.
Equation 25.3.7 shows us that the limiting current is a measure of the concentration of Fe(CN) in bulk solution, which means 3−

we can use the limiting current for quantitative work. Figure 25.3.7 also shows that there is a qualitative relationship between the
half-wave potential, E , and the limiting current; however, it is not yet clear what the half-wave potential represents.
1/2

If we solve Equation 25.3.7 for [Fe(CN) 3 −

6
] and substitute into Equation 25.3.6 and rearrange, we have
bulk

3 −
il − i
[Fe(CN) ] = (25.3.8)
6
x =0 K 3 −
Fe(CN)6

If we take the same approach with Fe(CN) , which forms at the electrode solution, then we have
4−

4 − 4 −
nF AD ([Fe(CN) ] − [Fe(CN) ] )
6 6
bulk x =0 4 −
i =− =K 4 − [Fe(CN) ] (25.3.9)
Fe(CN) 6
δ 6
x =0

4 − −i
[Fe(CN)6 ] = (25.3.10)
x =0 K 4 −
Fe(CN)
6

where the minus sign accounts for the concentration profile having a negative slope. Substituting Equation 25.3.9 and Equation
25.3.10 into Equation 25.3.1, which is the Nersnt equation, gives

25.3.5 https://chem.libretexts.org/@go/page/335297
 −i/K 4 −
Fe(CN)
∘ 6
E =E − 0.05916 log (25.3.11)
(il − i)/ K 3 −
Fe(CN)
6

K 3 −
Fe(CN) i
∘ 6
E =E + 0.05916 log − 0.05916 log (25.3.12)
K 4 − il − i
Fe(CN)
6

il −i
When i = 2
, which is the definition of E 1/2 , Equation 25.3.12 simplifies to
K 3 −
Fe(CN)6
∘
E1/2 = E + 0.05916 log (25.3.13)
K 4 −
Fe(CN)
6

The only difference between K

Fe(CN)
3 − and K Fe(CN)
4 − are the diffusion coefficients, D , for Fe(CN)
3 −

6
and for Fe(CN)
4 −

6
. As
6 6

these values should be similar, we have

∘
E1/2 ≈ E (25.3.14)

and E 1/2 provides an estimate for the standard state reduction potential.

Current/Voltage Relationships for Irreversible Reactions

When an electrochemical reaction is not reversible, the Nernst equation no longer applies, which means we can no longer assume
that the half-wave potential provides an estimate for the standard state reduction potential. The relationship between the limiting
current and the concentration of the electroactive species in bulk solution still holds true and quantitative work remains possible.

Oxygen Waves
The presence of dissolved oxygen creates a complication as it is capable of undergoing reduction reactions at the electrode's surface
that may interfere with the determination of the analyte's limiting current or half-wave potential. For example, O2 is reduced to
H2O2 with a standard state potential of +0.695 V
+ −
O (g) + 2 H (aq) + 2 e ⇌ H O (aq) (25.3.15)
2 2 2

and H2O2 subsequently is reduced to H2O at a standard state potential of +1.763 V.

+ −
H O (aq) + 2 H (aq) + 2 e ⇌ 2 H O(aq) (25.3.16)
2 2 2

This is the reason that a typical cell for voltammetry (see Figure 25.2.5) includes the ability to pass N2 through the solution to
remove dissolved O2. Once the solution is deaerated, N2 is allowed to flow over the solution to prevent O2 from reentering the
solution.

Applications of Linear Sweep Voltammetry

As we learned in the previous section, the limiting current in linear sweep voltammetry is proportional to the concentration of the
species undergoing oxidation or reduction at the electrode surface, which makes it a useful tool for a quantitative analysis. Because
we are interested only in the limiting current, most quantitative methods simply hold the potential of the working electrode at a
fixed value and measure the limiting current. Because we are measuring the current as a function of time instead of potential, these
are called amperometric methods (where ampere is the unit for current). Several examples of amperometic methods are gathered
here.

Amperometrics Detectors in Chromatography and Flow-Injection Analysis

One important detector for high-performance liquid chromatography (HPLC) is one in which the mobile phase eluting from the
column passes through a small volume electrochemical cell in which the working electrode is held at a potential that will oxidize or
reduce the analytes. The resulting current is plotted as function of time to yield the chromatogram. A similar arrangement is used in
flow-injection analysis (FIA). See Chapter 28 (HPLC) and Chapter 33 (FIA) for further details.

Amperometric Senors
One important application of amperometry is in the construction of chemical sensors. One of the first amperometric sensors was
developed in 1956 by L. C. Clark to measure dissolved O2 in blood. Figure 25.3.9 shows the sensor’s design, which is similar to a
potentiometric membrane electrode. A thin, gas-permeable membrane is stretched across the end of the sensor and is separated

25.3.6 https://chem.libretexts.org/@go/page/335297
from the working electrode and the counter electrode by a thin solution of KCl. The working electrode is a Pt disk cathode, and a
Ag ring anode serves as the counter electrode. Although several gases can diffuse across the membrane, including O2, N2, and CO2,
only oxygen undergoes reduction at the cathode
+ −
O2 (g) + 4 H3 O (aq) + 4 e ⇌ 6 H2 O(l) (25.3.17)

with its concentration at the electrode’s surface quickly reaching zero. The concentration of O2 at the membrane’s inner surface is
fixed by its diffusion through the membrane, which creates a limiting current. The result is a steady-state current that is
proportional to the concentration of dissolved oxygen. Because the electrode consumes oxygen, the sample is stirred to prevent the
depletion of O2 at the membrane’s outer surface.

The oxidation of the Ag anode is the other half-reaction.

− −
Ag(s) +  Cl (aq) ⇌ AgCl(s) + e

Figure 25.3.9 . Clark amperometric sensor for determining dissolved O2. The diagram on the right is a cross-section through the
electrode, which shows the Ag ring electrode and the Pt disk electrode.
Another example of an amperometric sensor is a glucose sensor. In this sensor the single membrane in Figure 25.3.10 is replaced
with three membranes. The outermost membrane of polycarbonate is permeable to glucose and O2. The second membrane contains
an immobilized preparation of glucose oxidase that catalyzes the oxidation of glucose to gluconolactone and hydrogen peroxide.

β − D − glucose (aq) +  O 2 (aq) + H2 O(l) ⇌ gluconolactone (aq) +  H 2 O2 (aq) (25.3.18)

The hydrogen peroxide diffuses through the innermost membrane of cellulose acetate where it undergoes oxidation at a Pt anode.
− −
H2 O2 (aq) + 2 OH (aq) ⇌  O 2 (aq) + 2 H2 O(l) + 2 e (25.3.19)

Figure 25.3.10 summarizes the reactions that take place in this amperometric sensor. FAD is the oxidized form of flavin adenine
nucleotide—the active site of the enzyme glucose oxidase—and FADH2 is the active site’s reduced form. Note that O2 serves a
mediator, carrying electrons to the electrode.

Figure 25.3.10. Schematic showing the reactions by which an amperometric biosensor responds to glucose.
By changing the enzyme and mediator, it is easy to extend to the amperometric sensor in Figure 25.3.10 to the analysis of other
analytes. For example, a CO2 sensor has been developed using an amperometric O2 sensor with a two-layer membrane, one of

25.3.7 https://chem.libretexts.org/@go/page/335297
which contains an immobilized preparation of autotrophic bacteria [Karube, I.; Nomura, Y.; Arikawa, Y. Trends in Anal. Chem.
1995, 14, 295–299]. As CO2 diffuses through the membranes it is converted to O2 by the bacteria, increasing the concentration of
O2 at the Pt cathode.

This page titled 25.3: Linear Sweep Voltammetry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

25.3.8 https://chem.libretexts.org/@go/page/335297
25.4: Cyclic Voltammetry
In linear sweep voltammetry we scan the potential in one direction, either to more positive potentials or to more negative potentials.
In cyclic voltammetry we complete a scan in both directions. Figure 25.4.1a shows a typical potential-excitation signal. In this
example, we first scan the potential to more positive values, resulting in the following oxidation reaction for the species R.
−
R ⇌ O + ne (25.4.1)

When the potential reaches a predetermined switching potential, we reverse the direction of the scan toward more negative
potentials. Because we generated the species O on the forward scan, during the reverse scan it reduces back to R.
−
O + ne ⇌ R (25.4.2)

Cyclic voltammetry is carried out in an unstirred solution, which, as shown in Figure 25.4.1b, results in peak currents instead of
limiting currents. The voltammogram has separate peaks for the oxidation reaction and for the reduction reaction, each
characterized by a peak potential and a peak current.

Figure 25.4.1 . Details for cyclic voltammetry. (a) One cycle of the triangular potential-excitation signal showing the initial
potential and the switching potential. A cyclic voltammetry experiment can consist of one cycle or many cycles. Although the
initial potential in this example is the negative switching potential, the cycle can begin with an intermediate initial potential and
cycle between two limits. (b) The resulting cyclic voltammogram showing the measurement of the peak currents and peak
potentials.
The peak current in cyclic voltammetry is given by the Randles-Sevcik equation
5 3/2 1/2 1/2
ip = (2.69 × 10 ) n AD ν CA (25.4.3)

where n is the number of electrons in the redox reaction, A is the area of the working electrode, D is the diffusion coefficient for the
electroactive species, ν is the scan rate, and CA is the concentration of the electroactive species at the electrode. For a well-behaved
system, the anodic and the cathodic peak currents are equal, and the ratio ip,a/ip,c is 1.00. The half-wave potential, E1/2, is midway
between the anodic and cathodic peak potentials.
Ep,a + Ep,c
E1/2 = (25.4.4)
2

Scanning the potential in both directions provides an opportunity to explore the electrochemical behavior of species generated at
the electrode. This is a distinct advantage of cyclic voltammetry over other voltammetric techniques. Figure 25.4.2 shows the
cyclic voltammogram for the same redox couple at both a faster and a slower scan rate. At the faster scan rate, 25.4.2a, we see two
peaks. At the slower scan rate in Figure 25.4.2b, however, the peak on the reverse scan disappears. One explanation for this is that
the products from the reduction of R on the forward scan have sufficient time to participate in a chemical reaction whose products
are not electroactive.

25.4.1 https://chem.libretexts.org/@go/page/335299
 Figure 25.4.2 . Cyclic voltammograms for R obtained at (a) a faster scan rate and at (b) a slower scan rate. One of the principal uses
of cyclic voltammetry is to study the chemical and electrochemical behavior of compounds. See this chapter’s additional resources
for further information.

This page titled 25.4: Cyclic Voltammetry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

25.4.2 https://chem.libretexts.org/@go/page/335299
25.5: Polarography
The first important voltammetric technique to be developed—polarography—uses the dropping mercury (DME) electrode as the
working electrode (see Figure 25.2.2 for a schematic diagram of the DME as well as two other types of Hg electrodes). In
polarography, as in linear sweep voltammetry, we vary the potential and measure the current. The change in potential can be in the
form of a linear ramp, as was the case for linear sweep voltammetry, or it can involve a series of pulses.

Normal Polarography
As shown in Figure 25.5.1, the current is measured while applying a linear potential ramp.

Figure 25.5.1 . Details of normal polarography: (a) the linear potential-excitation signal, and (b) the resulting voltammogram.
Although polarography takes place in an unstirred solution, we obtain a limiting current instead of a peak current. When a Hg drop
separates from the glass capillary and falls to the bottom of the electrochemical cell, it mixes the solution. Each new Hg drop,
therefore, grows into a solution whose composition is identical to the bulk solution. The oscillations in the current are a result of the
Hg drop’s growth, which leads to a time-dependent change in the area of the working electrode. The limiting current—which also
is called the diffusion current—is measured using either the maximum current, imax, or from the average current, iavg. The
relationship between the analyte’s concentration, CA, and the limiting current is given by the Ilkovic equations
1/2 2/3 1/6
imax = 706nD m t CA = Kmax CA (25.5.1)

1/2 2/3 1/6

iavg = 607nD m t CA = Kavg CA (25.5.2)

where n is the number of electrons in the redox reaction, D is the analyte’s diffusion coefficient, m is the flow rate of Hg, t is the
drop’s lifetime and Kmax and Kavg are constants. The half-wave potential, E1/2, provides qualitative information about the redox
reaction.

Pulse Polarography
Normal polarography has been replaced by various forms of pulse polarography, several examples of which are shown in Figure
25.5.2 [see Osteryoung, J. J. Chem. Educ. 1983, 60, 296–298 for a comprehensive review]. Normal pulse polarography (Figure

25.5.2a), for example, uses a series of potential pulses characterized by a cycle of time τ , a pulse-time of tp, a pulse potential of

ΔE , and a change in potential per cycle of ΔE . Typical experimental conditions for normal pulse polarography are τ ≈ 1 s , tp ≈
p s

50 ms, and ΔE ≈ 2 mV . The initial value of ΔE ≈ 2 mV , and it increases by ≈ 2 mV with each pulse. The current is sampled
s p

at the end of each potential pulse for approximately 17 ms before returning the potential to its initial value. The shape of the
resulting voltammogram is similar to Figure 25.5.1, but without the current oscillations. Because we apply the potential for only a
small portion of the drop’s lifetime, there is less time for the analyte to undergo oxidation or reduction and a smaller diffusion layer.
As a result, the faradaic current in normal pulse polarography is greater than in the polarography, resulting in better sensitivity and
smaller detection limits.

25.5.1 https://chem.libretexts.org/@go/page/335301
 Figure 25.5.2 . Potential-excitation signals and voltammograms for (a) normal pulse polarography, (b) differential pulse
polarography, (c) staircase polarography, and (d) square-wave polarography. The current is sampled at the time intervals shown by
the black rectangles. When measuring a change in current, Δi , the current at point 1 is subtracted from the current at point 2. The
symbols in the diagrams are as follows: τ is the cycle time; ΔE is a fixed or variable pulse potential; ΔE is the fixed change in
p s

potential per cycle, and tp is the pulse time.

In differential pulse polarography (Figure 25.5.2b) the current is measured twice per cycle: for approximately 17 ms before
applying the pulse and for approximately 17 ms at the end of the cycle. The difference in the two currents gives rise to the peak-
shaped voltammogram. Typical experimental conditions for differential pulse polarography are τ ≈ 1 s , tp ≈ 50 ms, ΔE ≈ 50 mV, p

and ΔE ≈ 2 mV.
s

The voltammogram for differential pulse polarography is approximately the first derivative of the voltammogram for normal
pulse polarography. To see why this is the case, note that the change in current over a fixed change in potential, Δi/ΔE,
approximates the slope of the voltammogram for normal pulse polarography. You may recall that the first derivative of a
function returns the slope of the function at each point. The first derivative of a sigmoidal function is a peak-shaped function.

Other forms of pulse polarography include staircase polarography (Figure 25.5.2c) and square-wave polarography (Figure
25.5.2d). One advantage of square-wave polarography is that we can make τ very small—perhaps as small as 5 ms, compared to 1

s for other forms of pulse polarography—which significantly decreases analysis time. For example, suppose we need to scan a
potential range of 400 mV. If we use normal pulse polarography with a ΔE of 2 mV/cycle and a τ of 1 s/cycle, then we need 200
s

s to complete the scan. If we use square-wave polarography with a ΔE of 2 mV/cycle and a τ of 5 ms/cycle, we can complete the
s

scan in 1 s. At this rate, we can acquire a complete voltammogram using a single drop of Hg!

25.5.2 https://chem.libretexts.org/@go/page/335301
Applications
Polarography is used extensively for the analysis of metal ions and inorganic anions, such as IO and NO . We also can use
−
3
−
3

polarography to study organic compounds with easily reducible or oxidizable functional groups, such as carbonyls, carboxylic
acids, and carbon-carbon double bonds.

This page titled 25.5: Polarography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

25.5.3 https://chem.libretexts.org/@go/page/335301
25.6: Stripping Methods
Another important voltammetric technique is stripping voltammetry, which consists of three related techniques: anodic stripping
voltammetry, cathodic stripping voltammetry, and adsorptive stripping voltammetry. Because anodic stripping voltammetry is the
more widely used of these techniques, we will consider it in greatest detail.
Anodic stripping voltammetry consists of two steps (Figure 25.6.1). The first step is a controlled potential electrolysis in which we
hold the working electrode—usually a hanging mercury drop or a mercury film electrode—at a cathodic potential sufficient to
deposit the metal ion on the electrode. For example, when analyzing Cu2+ the deposition reaction is
2+ −
Cu + 2e ⇌ Cu(Hg) (25.6.1)

where Cu(Hg) indicates that the copper is amalgamated with the mercury. This step serves as a means of concentrating the analyte
by transferring it from the larger volume of the solution to the smaller volume of the electrode. During most of the electrolysis we
stir the solution to increase the rate of deposition. Near the end of the deposition time we stop the stirring—eliminating convection
as a mode of mass transport—and allow the solution to become quiescent. Typical deposition times of 1–30 min are common, with
analytes at lower concentrations requiring longer times.

Figure 25.6.1 . Potential-excitation signal and voltammogram for anodic stripping voltammetry at a hanging mercury drop electrode
or a mercury film electrode.
In the second step, we scan the potential anodically—that is, toward a more positive potential. When the working electrode’s
potential is sufficiently positive, the analyte is stripped from the electrode, returning to solution in its oxidized form.
2+ −
Cu(Hg) ⇌  Cu + 2e (25.6.2)

Monitoring the current during the stripping step gives a peak-shaped voltammogram, as shown in Figure 25.6.1. The peak current
is proportional to the analyte’s concentration in the solution. Because we are concentrating the analyte in the electrode, detection
limits are much smaller than other electrochemical techniques. An improvement of three orders of magnitude—the equivalent of
parts per billion instead of parts per million—is routine.

Applications
Anodic stripping voltammetry is very sensitive to experimental conditions, which we must carefully control to obtain results that
are accurate and precise. Key variables include the area of the mercury film or the size of the hanging Hg drop, the deposition time,
the rest time, the rate of stirring, and the scan rate during the stripping step. Anodic stripping voltammetry is particularly useful for
metals that form amalgams with mercury, several examples of which are listed in Table 25.6.1.
Table 25.6.1 . Representative Examples of Analytes Determined by Stripping Voltammetry
anodic stripping voltammetry cathodic stripping voltammetry adsorptive stripping voltammetry

25.6.1 https://chem.libretexts.org/@go/page/335303
 anodic stripping voltammetry cathodic stripping voltammetry adsorptive stripping voltammetry

Bi3+ Br– bilirubin

Cd2+ Cl– codeine

Cu2+ I– cocaine

Ga3+ mercaptans (RSH) digitoxin

In3+ S2– dopamine

Pb2+ SCN– heme

Tl+ monesin

Sn2+ testosterone

Zn2+

Source: Compiled from Peterson, W. M.; Wong, R. V. Am. Lab. November 1981, 116–128; Wang, J. Am. Lab. May 1985, 41–50.

The experimental design for cathodic stripping voltammetry is similar to anodic stripping voltammetry with two exceptions. First,
the deposition step involves the oxidation of the Hg electrode to Hg , which then reacts with the analyte to form an insoluble film
2+
2

at the surface of the electrode. For example, when Cl– is the analyte the deposition step is
− −
2Hg(l) + 2 Cl (aq) ⇌  Hg 2 Cl2 (s) + 2 e (25.6.3)

Second, stripping is accomplished by scanning cathodically toward a more negative potential, reducing Hg
2+

2
back to Hg and
returning the analyte to solution.
− −
Hg2 Cl2 (s) + 2 e ⇌ 2Hg(l) + 2 Cl (aq) (25.6.4)

Table 25.6.1 lists several analytes analyzed successfully by cathodic stripping voltammetry.
In adsorptive stripping voltammetry, the deposition step occurs without electrolysis. Instead, the analyte adsorbs to the electrode’s
surface. During deposition we maintain the electrode at a potential that enhances adsorption. For example, we can adsorb a neutral
molecule on a Hg drop if we apply a potential of –0.4 V versus the SCE, a potential where the surface charge of mercury is
approximately zero. When deposition is complete, we scan the potential in an anodic or a cathodic direction, depending on whether
we are oxidizing or reducing the analyte. Examples of compounds that have been analyzed by absorptive stripping voltammetry
also are listed in Table 25.6.1.

This page titled 25.6: Stripping Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

25.6.2 https://chem.libretexts.org/@go/page/335303
25.7: Applications of Voltammetry
Voltammetry finds use for both quantitative analyses and characterization analyses. Examples of each are highlighted in this
section.

Quantitative Applications
Voltammetry has been used for the quantitative analysis of a wide variety of samples, including environmental samples, clinical
samples, pharmaceutical formulations, steels, gasoline, and oil.

Selecting the Voltammetric Technique

The choice of which voltammetric technique to use depends on the sample’s characteristics, including the analyte’s expected
concentration and the sample’s location. For example, amperometry is ideally suited for detecting analytes in flow systems,
including the in vivo analysis of a patient’s blood or as a selective sensor for the rapid analysis of a single analyte. The portability
of amperometric sensors, which are similar to potentiometric sensors, also make them ideal for field studies. Although cyclic
voltammetry is used to determine an analyte’s concentration, other methods described in this chapter are better suited for
quantitative work.
Pulse polarography and stripping voltammetry frequently are interchangeable. The choice of which technique to use often depends
on the analyte’s concentration and the desired accuracy and precision. Detection limits for normal pulse polarography generally are
on the order of 10–6 M to 10–7 M, and those for differential pulse polarography, staircase, and square wave polarography are
between 10–7 M and 10–9 M. Because we concentrate the analyte in stripping voltammetry, the detection limit for many analytes is
as little as 10–10 M to 10–12 M. On the other hand, the current in stripping voltammetry is much more sensitive than pulse
polarography to changes in experimental conditions, which may lead to poorer precision and accuracy. We also can use pulse
polarography to analyze a wider range of inorganic and organic analytes because there is no need to first deposit the analyte at the
electrode surface.
Stripping voltammetry also suffers from occasional interferences when two metals, such as Cu and Zn, combine to form an
intermetallic compound in the mercury amalgam. The deposition potential for Zn is sufficiently negative that any Cu2+ in the
sample also deposits into the mercury drop or film, leading to the formation of intermetallic compounds such as CuZn and CuZn2.
During the stripping step, zinc in the intermetallic compounds strips at potentials near that of copper, decreasing the current for zinc
at its usual potential and increasing the apparent current for copper. It is possible to overcome this problem by adding an element
that forms a stronger intermetallic compound with the interfering metal. Thus, adding Ga3+ minimizes the interference of Cu when
analyzing for Zn by forming an intermetallic compound of Cu and Ga.

Correcting for the Residual Current

In any quantitative analysis we must correct the analyte’s signal for signals that arise from other sources. The total current, itot, in
voltammetry consists of two parts: the current from the analyte’s oxidation or reduction, iA, and a background or residual current, ir.

itot = iA + ir (25.7.1)

The residual current, in turn, has two sources. One source is a faradaic current from the oxidation or reduction of trace interferents
in the sample, iint. The other source is the charging current, ich, that accompanies a change in the working electrode’s potential.

ir = iint + ich (25.7.2)

We can minimize the faradaic current due to impurities by carefully preparing the sample. For example, one important impurity is
dissolved O2, which undergoes a two-step reduction: first to H2O2 at a potential of –0.1 V versus the SCE, and then to H2O at a
potential of –0.9 V versus the SCE. Removing dissolved O2 by bubbling an inert gas such as N2 through the sample eliminates this
interference. After removing the dissolved O2, maintaining a blanket of N2 over the top of the solution prevents O2 from reentering
the solution.
There are two methods to compensate for the residual current. One method is to measure the total current at potentials where the
analyte’s faradaic current is zero and extrapolate it to other potentials. This is the method shown in Figure 25.3.7. One advantage of
extrapolating is that we do not need to acquire additional data. An important disadvantage is that an extrapolation assumes that any
change in the residual current with potential is predictable, which may not be the case. A second, and more rigorous approach, is to
obtain a voltammogram for an appropriate blank. The blank’s residual current is then subtracted from the sample’s total current.

25.7.1 https://chem.libretexts.org/@go/page/335305
Analysis for Single Components
The analysis of a sample with a single analyte is straightforward using any of the standardization methods discussed in Chapter 1.

 Example 25.7.1

The concentration of As(III) in water is determined by differential pulse polarography in 1 M HCl. The initial potential is set to
–0.1 V versus the SCE and is scanned toward more negative potentials at a rate of 5 mV/s. Reduction of As(III) to As(0)
occurs at a potential of approximately –0.44 V versus the SCE. The peak currents for a set of standard solutions, corrected for
the residual current, are shown in the following table.

[As(III)] (µM) ip (µM)

1.00 0.298

3.00 0.947

6.00 1.83

9.00 2.72

What is the concentration of As(III) in a sample of water if its peak current is 1.37 μA?

Solution
Linear regression gives the calibration curve shown in Figure 25.7.1, with an equation of

ip = 0.0176 + 3.01 × [As(III)]

Substituting the sample’s peak current into the regression equation gives the concentration of As(III) as 4.49 μM.

Figure 25.7.1 . Calibration curve for the data in Example 25.7.1 .

Multicomponent Analysis
Voltammetry is a particularly attractive technique for the analysis of samples that contain two or more analytes. Provided that the
analytes behave independently, the voltammogram of a multicomponent mixture is a summation of each analyte’s individual
voltammograms. As shown in Figure 25.7.2, if the separation between the half-wave potentials or between the peak potentials is
sufficient, we can determine the presence of each analyte as if it is the only analyte in the sample. The minimum separation
between the half-wave potentials or peak potentials for two analytes depends on several factors, including the type of electrode and
the potential-excitation signal. For normal polarography the separation is at least ±0.2–0.3 V, and differential pulse voltammetry
requires a minimum separation of ±0.04–0.05 V.

25.7.2 https://chem.libretexts.org/@go/page/335305
 Figure 25.7.2 . Voltammograms for a sample that contains two analytes showing the measurement of (a) limiting currents, and (b)
peak currents.
If the voltammograms for two analytes are not sufficiently separated, a simultaneous analysis may be possible. An example of this
approach is outlined the following example.

 Example 25.7.2

The differential pulse polarographic analysis of a mixture of indium and cadmium in 0.1 M HCl is complicated by the overlap
of their respective voltammograms [Lanza P. J. Chem. Educ. 1990, 67, 704–705]. The peak potential for indium is at –0.557 V
and that for cadmium is at –0.597 V. When a 0.800-ppm indium standard is analyzed, Δi (in arbitrary units) is 200.5 at –
p

0.557 V and 87.5 at –0.597 V relative to a saturated Ag/AgCl reference electorde. A standard solution of 0.793 ppm cadmium
has a Δi of 58.5 at –0.557 V and 128.5 at –0.597 V. What is the concentration of indium and cadmium in a sample if Δi is
p p

167.0 at a potential of –0.557 V and 99.5 at a potential of –0.597V.

Solution
The change in current, Δi , in differential pulse polarography is a linear function of the analyte’s concentration
p

Δip = kA CA

where kA is a constant that depends on the analyte and the applied potential, and CA is the analyte’s concentration. To determine
the concentrations of indium and cadmium in the sample we must first find the value of kA for each analyte at each potential.
For simplicity we will identify the potential of –0.557 V as E1, and that for –0.597 V as E2. The values of kA are
200.5 −1
kIn,E1 = = 250.6 ppm
0.800 ppm

87.5 −1
kIn,E2 = = 109.4 ppm
0.800 ppm

58.5
−1
kCdE = = 73.8 ppm
1
0.793 ppm

128.5 −1
kCdE2 = = 162.0 ppm
0.793 ppm

Next, we write simultaneous equations for the current at the two potentials.
−1 −1
ΔiE1 = 167.0 = 250.6 ppm × CIn + 73.8 ppm × CCd

−1 −1
△iE2 = 99.5 = 109.4 ppm × CIn + 162.0 ppm × CCd

Solving the simultaneous equations, which is left as an exercise, gives the concentration of indium as 0.606 ppm and the
concentration of cadmium as 0.205 ppm.

Environmental Samples
Voltammetry is one of several important analytical techniques for the analysis of trace metals in environmental samples, including
groundwater, lakes, rivers and streams, seawater, rain, and snow. Detection limits at the parts-per-billion level are routine for many

25.7.3 https://chem.libretexts.org/@go/page/335305
trace metals using differential pulse polarography, with anodic stripping voltammetry providing parts-per-trillion detection limits
for some trace metals.
One interesting environmental application of anodic stripping voltammetry is the determination of a trace metal’s chemical form
within a water sample. Speciation is important because a trace metal’s bioavailability, toxicity, and ease of transport through the
environment often depends on its chemical form. For example, a trace metal that is strongly bound to colloidal particles generally is
not toxic because it is not available to aquatic lifeforms. Unfortunately, anodic stripping voltammetry can not distinguish a trace
metal’s exact chemical form because closely related species, such as Pb2+ and PbCl+, produce a single stripping peak. Instead, trace
metals are divided into “operationally defined” categories that have environmental significance.

Operationally defined means that an analyte is divided into categories by the specific methods used to isolate it from the
sample. There are many examples of operational definitions in the environmental literature. The distribution of trace metals in
soils and sediments, for example, often is defined in terms of the reagents used to extract them; thus, you might find an
operational definition for Zn2+ in a lake sediment as that extracted using 1.0 M sodium acetate, or that extracted using 1.0 M
HCl.

Although there are many speciation schemes in the environmental literature, we will consider one proposed by Batley and Florence
[see (a) Batley, G. E.; Florence, T. M. Anal. Lett. 1976, 9, 379–388; (b) Batley, G. E.; Florence, T. M. Talanta 1977, 24, 151–158;
(c) Batley, G. E.; Florence, T. M. Anal. Chem. 1980, 52, 1962–1963; (d) Florence, T. M., Batley, G. E.; CRC Crit. Rev. Anal. Chem.
1980, 9, 219–296]. This scheme, which is outlined in Table 25.7.2, combines anodic stripping voltammetry with ion-exchange and
UV irradiation, dividing soluble trace metals into seven groups. In the first step, anodic stripping voltammetry in a pH 4.8 acetic
acid buffer differentiates between labile metals and nonlabile metals. Only labile metals—those present as hydrated ions, weakly
bound complexes, or weakly adsorbed on colloidal surfaces—deposit at the electrode and give rise to a signal. Total metal
concentration are determined by ASV after digesting the sample in 2 M HNO3 for 5 min, which converts all metals into an ASV-
labile form.
Table 25.7.1 . Operational Speciation of Soluble Trace Metals
method speciation of soluble metals

ASV labile metals nonlabile or bound metals

Ion-Exchange removed not removed removed not removed

UV Irradiation released not released released not released released not released

Groups I II III IV V VI VII

Group I: free metal ions; weaker labile organic complexes and inorganic complexes
Group II: stronger labile organic complexes; labile metals absorbed on organic solids
Group III: stronger labile inorganic complexes; labile metals absorbed on inorganic solids
Group IV: weaker nonlabile organic complexes
Group V: weaker nonlabile inorganic complexes
Group VI: stronger nonlabile organic complexes; nonlabile metals absorbed on organic solids
Group VII: stronger nonlabile inorganic complexes; nonlabile metals absorbed on inorganic solids
Operational definitions of speciation from (a)Batley,G.E.;Florence,T.M.Anal.Lett.1976,9,379–388;
(b)Batley,G.E.;Florence,T.M.Talanta1977,24,151–158; (c) Batley, G. E.; Florence, T. M. Anal. Chem. 1980, 52, 1962–1963; (d) Florence, T. M.,
Batley, G. E.; CRC Crit. Rev. Anal. Chem. 1980, 9, 219–296.

A Chelex-100 ion-exchange resin further differentiates between strongly bound metals—usually metals bound to inorganic and
organic solids, but also those tightly bound to chelating ligands—and more loosely bound metals. Finally, UV radiation
differentiates between metals bound to organic phases and inorganic phases. The analysis of seawater samples, for example,
suggests that cadmium, copper, and lead are present primarily as labile organic complexes or as labile adsorbates on organic
colloids (Group II in Table 25.7.1).
Differential pulse polarography and stripping voltammetry are used to determine trace metals in airborne particulates, incinerator
fly ash, rocks, minerals, and sediments. The trace metals, of course, are first brought into solution using a digestion or an

25.7.4 https://chem.libretexts.org/@go/page/335305
extraction.
Amperometric sensors also are used to analyze environmental samples. For example, the dissolved O2 sensor described earlier is
used to determine the level of dissolved oxygen and the biochemical oxygen demand, or BOD, of waters and wastewaters. The
latter test—which is a measure of the amount of oxygen required by aquatic bacteria as they decompose organic matter—is
important when evaluating the efficiency of a wastewater treatment plant and for monitoring organic pollution in natural waters. A
high BOD suggests that the water has a high concentration of organic matter. Decomposition of this organic matter may seriously
deplete the level of dissolved oxygen in the water, adversely affecting aquatic life. Other amperometric sensors are available to
monitor anionic surfactants in water, and CO2, H2SO4, and NH3 in atmospheric gases.

Clinical Samples
Differential pulse polarography and stripping voltammetry are used to determine the concentration of trace metals in a variety of
clinical samples, including blood, urine, and tissue. The determination of lead in blood is of considerable interest due to concerns
about lead poisoning. Because the concentration of lead in blood is so small, anodic stripping voltammetry frequently is the more
appropriate technique. The analysis is complicated, however, by the presence of proteins that may adsorb to the mercury electrode,
inhibiting either the deposition or stripping of lead. In addition, proteins may prevent the electrodeposition of lead through the
formation of stable, nonlabile complexes. Digesting and ashing the blood sample mini- mizes this problem. Differential pulse
polarography is useful for the routine quantitative analysis of drugs in biological fluids, at concentrations of less than 10–6 M
[Brooks, M. A. “Application of Electrochemistry to Pharmaceutical Analysis,” Chapter 21 in Kissinger, P. T.; Heinemann, W. R.,
eds. Laboratory Techniques in Electroanalytical Chemistry, Marcel Dekker, Inc.: New York, 1984, pp 539–568.]. Amperometric
sensors using enzyme catalysts also have many clinical uses, several examples of which are shown in Table 25.7.2.
Table 25.7.2 . Representative Amperometric Biosensors
analyte enzyme species detected

choline choline oxidase H2O2

ethanol alcohol oxidase H2O2

formaldehyde formaldehyde dehydrogenase NADH

glucose glucose oxidase H2O2

glutamine glutaminase, glutamine oxidase H2O2

glycerol glycerol dehydrogenase NADH, O2

lactate lactate oxidase H2O2

phenol polyphenol oxidase quinone

inorganic phosphorous nucleoside phosphoylase O2

Source: Cammann, K.; Lemke, U.; Rohen, A.; Sander, J.; Wilken, H.; Winter, B. Angew. Chem. Int. Ed. Engl. 1991, 30, 516–539.

Miscellaneous Samples
In addition to environmental samples and clinical samples, differential pulse polarography and stripping voltammetry are used for
the analysis of trace metals in other sample, including food, steels and other alloys, gasoline, gunpowder residues, and
pharmaceuticals. Voltammetry is an important technique for the quantitative analysis of organics, particularly in the pharmaceutical
industry where it is used to determine the concentration of drugs and vitamins in formulations. For example, voltammetric methods
are available for the quantitative analysis of vitamin A, niacinamide, and riboflavin. When the compound of interest is not
electroactive, it often can be derivatized to an electroactive form. One example is the differential pulse polarographic determination
of sulfanilamide, which is converted into an electroactive azo dye by coupling with sulfamic acid and 1-napthol.
In the previous section we learned how to use voltammetry to determine an analyte’s concentration in a variety of different
samples. We also can use voltammetry to characterize an analyte’s properties, including verifying its electrochemical reversibility,
determining the number of electrons transferred during its oxidation or reduction, and determining its equilibrium constant in a
coupled chemical reaction.

25.7.5 https://chem.libretexts.org/@go/page/335305
Characterization Applications
In a characterization application we study the properties of a system. Three examples are described here: determining if a redox
reaction is electrochemically reversible, determining the number of electrons involved in the redox reaction, and studying metal-
ligand complexation.

Electrochemical Reversibility and Determination of n

Earlier in this chapter we derived a relationship between E1/2 and the standard-state potential for a redox couple using the Nernst
equation, noting that a redox reaction must be electrochemically reversible. How can we tell if a redox reaction is reversible by
looking at its voltammogram? As we learned in Chapter 25.3, for a reversible redox reaction the relationship between potential and
current is
0.05916 i
E = E½ − log (25.7.3)
n il − i

If a reaction is electrochemically reversible, a plot of E versus log i

il −i
is a straight line with a slope of –0.05916/n. In addition, the
slope should yield an integer value for n.

 Example 25.7.3

The following data were obtained from a linear scan hydrodynamic voltammogram of a reversible reduction reaction.

E (V vs. SCE) current (μA)

–0.358 0.37

–0.372 0.95

–0.382 1.71

–0.400 3.48

–0.410 4.20

–0.435 4.97

The limiting current is 5.15 μA. Show that the reduction reaction is reversible, and determine values for n and for E1/2.

Solution
Figure 25.7.3 shows a plot of E versus log . Because the result is a straight-line, we know the reaction is electrochemically
i

il −i

reversible under the conditions of the experiment. A linear regression analysis gives the equation for the straight line as
i
E = −0.391V − 0.0300 log
il − i

From Equation 25.7.3, the slope is equivalent to –0.05916/n; solving for n gives a value of 1.97, or 2 electrons. From Equation
25.7.3 we also know that E1/2 is the y-intercept for a plot of E versus log ; thus, E1/2 for the data in this example is –0.391
i

il −i

V versus the SCE.

25.7.6 https://chem.libretexts.org/@go/page/335305
 Figure 25.7.3 . Determination of electrochemical reversibility for the data in Example 25.7.3 .
We also can use cyclic voltammetry to evaluate electrochemical reversibility by looking at the difference between the peak
potentials for the anodic and the cathodic scans. For an electrochemically reversible reaction, the following equation holds true.
0.05916 V
ΔEp = Ep,a − Ep,c = (25.7.4)
n

As an example, for a two-electron reduction we expect a ΔEp of approximately 29.6 mV. For an electrochemically irreversible
reaction the value of ΔE is larger than expected.
p

Determining Equilibrium Constants for Coupled Chemical Reactions

Another important application of voltammetry is determining the equilibrium constant for a solution reaction that is coupled to a
redox reaction. The presence of the solution reaction affects the ease of electron transfer in the redox reaction, shifting E1/2 to a
more negative or to a more positive potential. Consider, for example, the reduction of O to R
−
O + ne ⇌ R (25.7.5)

the voltammogram for which is shown in Figure 25.7.4. If we introduce a ligand, L, that forms a strong complex with O, then we
also must consider the reaction
O + pL ⇌ OLp (25.7.6)

In the presence of the ligand, the overall redox reaction is

−
OLp + ne ⇌ R + pL (25.7.7)

Because of its stability, the reduction of the OLp complex is less favorable than the reduction of O. As shown in Figure 25.7.4, the
resulting voltammogram shifts to a potential that is more negative than that for O. Furthermore, the shift in the voltammogram
increases as we increase the ligand’s concentration.

Figure 25.7.4 . Effect of a metal-ligand complexation reaction on a voltammogram. The voltammogram in blue is for the reduction
of O in the absence of ligand. Adding the ligand shifts the potentials to more negative potentials, as shown by the voltammograms
in red.
We can use this shift in the value of E1/2 to determine both the stoichiometry and the formation constant for a metal-ligand
complex. To derive a relationship between the relevant variables we begin with two equations: the Nernst equation for the
reduction of O

25.7.7 https://chem.libretexts.org/@go/page/335305
 0.05916 [R]x=0
∘
E =E − log (25.7.8)
O/R
n [O]x=0

and the stability constant, β for the metal-ligand complex at the electrode surface.
p

[OLp ]
x=0
βp = (25.7.9)
p
[O]x=0 [L]
x=0

In the absence of ligand the half-wave potential occurs when [R]x = 0 and [O]x = 0 are equal; thus, from the Nernst equation we have
∘
(E1/2 ) =E (25.7.10)
nc O/R

where the subscript “nc” signifies that the complex is not present. When ligand is present we must account for its effect on the
concentration of O. Solving Equation 25.7.9 for [O]x = 0 and substituting into the Equation 25.7.8 gives
p
0.05916 [R]x=0 [L] βp
∘ x=0
E =E − log (25.7.11)
O/R
n [OLp ]
x=0

If the formation constant is sufficiently large, such that essentially all O is present as the complex OLp, then [R] x=0 and [OL p ]x=0

are equal at the half-wave potential, and Equation 25.7.11 simplifies to

0.05916 p
∘
(E1/2 ) =E − log [L] βp (25.7.12)
c O/R x=0
n

where the subscript “c” indicates that the complex is present. Defining ΔE 1/2
as

△E1/2 = (E1/2 ) − (E1/2 ) (25.7.13)

c nc

and substituting Equation 25.7.10 and Equation 25.7.12 and expanding the log term leaves us with the following equation.
0.05916 0.05916p
ΔE1/2 = − log βp − log [L] (25.7.14)
n n

A plot of ΔE 1/2 versus log[L] is a straight-line, with a slope that is a function of the metal-ligand complex’s stoichiometric
coefficient, p, and a y-intercept that is a function of its formation constant β . p

 Example 25.7.4

A voltammogram for the two-electron reduction (n = 2) of a metal, M, has a half-wave potential of –0.226 V versus the SCE.
In the presence of an excess of ligand, L, the following half-wave potentials are recorded.

[L] (M) (E1/2)c (V vs. SCE)

0.020 –0.494

0.040 –0.512

0.060 –0.523

0.080 –0.530

0.100 –0.536

Determine the stoichiometry of the metal-ligand complex and its formation constant.

Solution
We begin by calculating values of ΔE 1/2 using Equation 25.7.13, obtaining the values in the following table.

[L] (M) ΔE1/2 (V vs. SCE)

0.020 –0.268

0.040 –0.286

25.7.8 https://chem.libretexts.org/@go/page/335305
 [L] (M)
0.060 ΔE1/2 (V vs. SCE)
–0.297

0.020
0.080 –0.268
–0.304

0.040
0.100 –0.286
–0.310

0.060plot of ΔE
Figure 25.7.5 shows the resulting as a function of log[L]. A linear regression–0.297
analysis gives the equation for the
1/2

straight line as 0.080 –0.304

0.100 △E1/2 = −0.370V − 0.0601 log [L] –0.310

From Equation 25.7.14 we know that the slope is equal to –0.05916p/n. Using the slope and n = 2, we solve for p obtaining a
value of 2.03 ≈ 2. The complex’s stoichiometry, therefore, is ML2. We also know, from Equation 25.7.14, that the y-intercept is
equivalent to –(0.05916/n)logβ . Solving for β gives a formation constant of 3.2 × 10 .
p 2
12

Figure 25.7.5 . Determination of the stoichiometry and formation constant for a metal-ligand complex using the data in Example
25.7.4 .

Cyclic voltammetry is one of the most powerful electrochemical techniques for exploring the mechanism of coupled
electrochemical and chemical reactions. The treatment of this aspect of cyclic voltammetry is beyond the level of this text, although
you can consult this chapter’s additional resources for additional information.

This page titled 25.7: Applications of Voltammetry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

25.7.9 https://chem.libretexts.org/@go/page/335305
 CHAPTER OVERVIEW
26: Introduction to Chromatographic Separations
In previous chapters we explored the application of spectroscopy and electroanalytical chemistry to the quantitative analysis of an
analyte in a sample. Despite the power of these instrumental methods of analysis, their use is often limited if the sample contains
species that will interfere with the analysis. A UV/Vis analysis, for example, is easy to complete if the analyte is the only species
present that absorbs light at the analytical wavelength. If two species contribute to the overall absorbance, a quantitative analysis is
still possible if we can measure the sample's absorbance at two wavelengths. Things become more complex, however, as the
number of analytes or interferents increase or if we do not know the identity of an intereferent.
Chromatography provides a solution to the analysis of complex samples by providing a way to separate the individual species in a
sample prior to their analysis by a spectroscopic or electroanalytical method of analysis. In this chapter we provide a general
introduction to chromatographic separations. In the four chapters that follow, we will consider specific chromatographic methods.
26.1: A General Description of Chromatography
26.2: Migration Rates of Solutes
26.3: Zone Broadening and Column Efficiency
26.4: Optimization and Column Performance
26.5: Summary of Important Relationships for Chromatography
26.6: Applications of Chromatography

This page titled 26: Introduction to Chromatographic Separations is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

1
26.1: A General Description of Chromatography
In chromatography we pass a sample-free phase, which we call the mobile phase, over a second sample-free stationary phase that
remains fixed in space (Figure 26.1.1). We inject or place the sample into the mobile phase. As the sample moves with the mobile
phase, its components partition between the mobile phase and the stationary phase. A component whose distribution ratio favors
the stationary phase requires more time to pass through the system. Given sufficient time and sufficient stationary and mobile
phase, we can separate solutes even if they have similar distribution ratios.

Figure 26.1.1 . In chromatography we pass a mobile phase over a stationary phase. When we inject a sample into the mobile phase,
the sample’s components both move with the mobile phase and partition into the stationary phase. The solute that spends the most
time in the stationary phase takes the longest time to move through the system.

Classification of Chromatographic Methods

There are many ways in which we can identify a chromatographic separation: by describing the physical state of the mobile phase
and the stationary phase; by describing how we bring the stationary phase and the mobile phase into contact with each other; or by
describing the chemical or physical interactions between the solute and the stationary phase. Let’s briefly consider how we might
use each of these classifications.

We can trace the history of chromatography to the turn of the century when the Russian botanist Mikhail Tswett used a column
packed with calcium carbonate and a mobile phase of petroleum ether to separate colored pigments from plant extracts. As the
sample moved through the column, the plant’s pigments separated into individual colored bands. After effecting the separation,
the calcium carbonate was removed from the column, sectioned, and the pigments recovered. Tswett named the technique
chromatography, combining the Greek words for “color” and “to write.” There was little interest in Tswett’s technique until
Martin and Synge’s pioneering development of a theory of chromatography (see Martin, A. J. P.; Synge, R. L. M. “A New
Form of Chromatogram Employing Two Liquid Phases,” Biochem. J. 1941, 35, 1358–1366). Martin and Synge were awarded
the 1952 Nobel Prize in Chemistry for this work.

Types of Mobile Phases and Stationary Phases

The mobile phase is a liquid or a gas, and the stationary phase is a solid or a liquid film coated on a solid substrate. We often name
chromatographic techniques by listing the type of mobile phase followed by the type of stationary phase. In gas–liquid
chromatography, for example, the mobile phase is a gas and the stationary phase is a liquid film coated on a solid substrate. If a
technique’s name includes only one phase, as in gas chromatography, it is the mobile phase.

Contact Between the Mobile Phase and the Stationary Phase

There are two common methods for bringing the mobile phase and the stationary phase into contact. In column chromatography we
pack the stationary phase into a narrow column and pass the mobile phase through the column using gravity or by applying
pressure. The stationary phase is a solid particle or a thin liquid film coated on either a solid particulate packing material or on the
column’s walls.
In planar chromatography the stationary phase is coated on a flat surface—typically, a glass, metal, or plastic plate. One end of the
plate is placed in a reservoir that contains the mobile phase, which moves through the stationary phase by capillary action. In paper
chromatography, for example, paper is the stationary phase.

Interaction Between the Solute and the Stationary Phase

The interaction between the solute and the stationary phase provides a third method for describing a separation (Figure 26.1.2). In
adsorption chromatography, solutes separate based on their ability to adsorb to a solid stationary phase. In partition
chromatography, the stationary phase is a thin liquid film on a solid support. Separation occurs because there is a difference in the

26.1.1 https://chem.libretexts.org/@go/page/349937
equilibrium partitioning of solutes between the stationary phase and the mobile phase. A stationary phase that consists of a solid
support with covalently attached anionic (e.g., −SO ) or cationic (e.g., −N(CH ) ) functional groups is the basis for ion-
−

3 3
+
3

exchange chromatography in which ionic solutes are attracted to the stationary phase by electrostatic forces. In size-exclusion
chromatography the stationary phase is a porous particle or gel, with separation based on the size of the solutes. Larger solutes are
unable to penetrate as deeply into the porous stationary phase and pass more quickly through the column.

Figure 26.1.2 . Four examples of interactions between a solute and the stationary phase: (a) adsorption on a solid surface, (b)
partitioning into a liquid phase, (c) ion-exchange, and (d) size exclusion. For each example, the smaller, green solute is more
strongly retained than the larger, red solute.

There are other interactions that can serve as the basis of a separation. In affinity chromatography the interaction between an
antigen and an antibody, between an enzyme and a substrate, or between a receptor and a ligand forms the basis of a
separation.

Elution Chromatography on Columns

Of the two methods for bringing the stationary phase and the mobile phases into contact, the most important is column
chromatography. In this section we develop a general theory that we may apply to any form of column chromatography.
Figure 26.1.3 provides a simple view of a liquid–solid column chromatography experiment. The sample is introduced as a narrow
band at the top of the column. Ideally, the solute’s initial concentration profile is rectangular (Figure 26.1.4a). As the sample moves
down the column, the solutes begin to separate (Figure 26.1.3b,c) and the individual solute bands begin to broaden and develop a
Gaussian profile (Figure 26.1.4b,c). If the strength of each solute’s interaction with the stationary phase is sufficiently different,
then the solutes separate into individual bands (Figure 26.1.3d and Figure 26.1.4d).

Figure 26.1.3. Progress of a column chromatographic separation of a

two-component mixture. In (a) the sample is layered on top of the Figure 26.1.4 . An alternative view of the separation in Figure 26.1.1
stationary phase. As mobile phase passes through the column, the showing the concentration of each solute as a function of distance down
sample separates into two solute bands (b–d). In (e) and (f), we collect the column.
each solute as it elutes from the column.

26.1.2 https://chem.libretexts.org/@go/page/349937
We can follow the progress of the separation by collecting fractions as they elute from the column (Figure 26.1.3e,f), or by placing
a suitable detector at the end of the column. A plot of the detector’s response as a function of elution time, or as a function of the
volume of mobile phase, is known as a chromatogram (Figure 26.1.5), and consists of a peak for each solute.

Figure 26.1.5 . Chromatogram for the separation shown in Figure 26.1.3 and Figure 26.1.4 , showing the detector’s response as a
function of the elution time.

There are many possible detectors that we can use to monitor the separation. Later sections of this chapter describe some of the
most popular.

This page titled 26.1: A General Description of Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

26.1.3 https://chem.libretexts.org/@go/page/349937
26.2: Migration Rates of Solutes
Our ability to separate two solutes depends on the equilibrium interactions of the solute with the stationary phase and the mobile
phase, which effects both the time it takes a solute to travel through the column and how the width of the solute's elution profile. In
this section we consider the rate at which the solute moves through the column.

Distribution Constants
Let’s assume we can describe a solute’s distribution between the mobile phase and stationary phase using the following equilibrium
reaction
Sm ⇌ Ss (26.2.1)

where Sm is the solute in the mobile phase and Ss is the solute in the stationary phase. The equilibrium constant for this reaction is
an equilibrium partition coefficient, KD.
[ Ss ]
KD = (26.2.2)
[ Sm ]

This is not a trivial assumption. In this section we are, in effect, treating the solute’s equilibrium between the mobile phase and
the stationary phase as if it is identical to the equilibrium in a liquid–liquid extraction. You might question whether this is a
reasonable assumption. There is an important difference between the two experiments that we need to consider. In a liquid–
liquid extraction, which takes place in a separatory funnel, the two phases remain in contact with each other at all times,
allowing for a true equilibrium. In chromatography, however, the mobile phase is in constant motion. A solute that moves into
the stationary phase from the mobile phase will equilibrate back into a different portion of the mobile phase; this does not
describe a true equilibrium.
So, we ask again: Can we treat a solute’s distribution between the mobile phase and the stationary phase as an equilibrium
process? The answer is yes, if the mobile phase velocity is slow relative to the kinetics of the solute’s movement back and
forth between the two phase. In general, this is a reasonable assumption.

Retention Time
We can characterize a chromatographic peak’s properties in several ways, two of which are shown in Figure 26.2.1. Retention time,
tr, is the time between the sample’s injection and the maximum response for the solute’s peak. A chromatographic peak’s baseline
width, w, as shown in Figure 26.2.1, is determined by extending tangent lines from the inflection points on either side of the peak
through the baseline. Although usually we report tr and w using units of time, we can report them using units of volume by
multiplying each by the mobile phase’s velocity, or report them in linear units by measuring distances with a ruler.

For example, a solute’s retention volume,Vr, is t r ×u where u is the mobile phase’s velocity through the column.

Figure 26.2.1 . Chromatogram showing a solute’s retention time, tr, and baseline width, w, and the column’s void time, tm, for
nonretained solutes. In this section we focus on retention time; we will consider the width of the solute's peak in Section 26.3.

26.2.1 https://chem.libretexts.org/@go/page/349939
In addition to the solute’s peak, Figure 26.2.1 also shows a small peak that elutes shortly after the sample is injected into the
mobile phase. This peak contains all nonretained solutes, which move through the column at the same rate as the mobile phase. The
time required to elute the nonretained solutes is called the column’s void time, tm.

The Rate of Solute Migration: The Retention Factor

In the absence of any additional equilibrium reactions in the mobile phase or the stationary phase, KD is equivalent to the
distribution ratio, D,
[ S0 ] (mol S)s / Vs
D = = = KD (26.2.3)
[ Sm ] (mol S)m / Vm

where Vs and Vm are the volumes of the stationary phase and the mobile phase, respectively.
A conservation of mass requires that the total moles of solute remain constant throughout the separation; thus, we know that the
following equation is true.
(mol S)tot = (mol S)m + (mol S)s (26.2.4)

Solving Equation 26.2.4 for the moles of solute in the stationary phase and substituting into Equation 26.2.2 leaves us with
{(mol S)tot − (mol S)m } / Vs
D = (26.2.5)
(mol S)m / Vm

Rearranging this equation and solving for the fraction of solute in the mobile phase, fm, gives
(mol S)m Vm
fm = = (26.2.6)
(mol S)tot DVs + Vm

Because we may not know the exact volumes of the stationary phase and the mobile phase, we simplify Equation 26.2.6 by
dividing both the numerator and the denominator by Vm; thus
Vm / Vm 1 1
fm = = = (26.2.7)
DVs / Vm + Vm / Vm DVs / Vm + 1 1 +k

where k
Vs
k = D× (26.2.8)
Vm

is the solute’s retention factor. Note that the larger the retention factor, the more the distribution ratio favors the stationary phase,
leading to a more strongly retained solute and a longer retention time.

Other (older) names for the retention factor are capacity factor, capacity ratio, and partition ratio, and it sometimes is given the
symbol k . Keep this in mind if you are using other resources. Retention factor is the approved name from the IUPAC Gold
′

Book.

We can determine a solute’s retention factor from a chromatogram by measuring the column’s void time, tm, and the solute’s
retention time, tr (see Figure 26.2.1). Solving Equation 26.2.7 for k, we find that
1 − fm
k = (26.2.9)
fm

Earlier we defined fm as the fraction of solute in the mobile phase. Assuming a constant mobile phase velocity, we also can define
fm as
 time spent in the mobile phase  tm
fm = = (26.2.10)
 time spent in the stationary phase  tr

Substituting back into Equation 26.2.9 and rearranging leaves us with

26.2.2 https://chem.libretexts.org/@go/page/349939
 tm
1− tr − tm
′
tr
tr
k = = = (26.2.11)
tm
tm tm
tr

where t is the adjusted retention time.

′
r

 Example 26.2.1

In a chromatographic analysis of low molecular weight acids, butyric acid elutes with a retention time of 7.63 min. The
column’s void time is 0.31 min. Calculate the retention factor for butyric acid.

Solution
tr − tm 7.63 min − 0.31 min
kbut = = = 23.6
tm 0.31 min

Figure 26.2.2 . Chromatogram for Exercise 26.2.1 .

Relative Migration Rates: The Selectivity Factor

Selectivity is a relative measure of the retention of two solutes, which we define using a selectivity factor, α
kB tr,B − tm
α = = (26.2.12)
kA tr,A − tm

where solute A has the smaller retention time. When two solutes elute with identical retention time, α = 1.00 ; for all other
conditions α > 1.00.

 Example 26.2.2

In the chromatographic analysis for low molecular weight acids described in Example 26.2.1, the retention time for isobutyric
acid is 5.98 min. What is the selectivity factor for isobutyric acid and butyric acid?

Solution
First we must calculate the retention factor for isobutyric acid. Using the void time from Example 26.2.1 we have
tr − tm 5.98 min − 0.31 min
kiso = = = 18.3
tm 0.31 min

The selectivity factor, therefore, is

kbut  23.6
α = = = 1.29
kiso  18.3

26.2.3 https://chem.libretexts.org/@go/page/349939
This page titled 26.2: Migration Rates of Solutes is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

26.2.4 https://chem.libretexts.org/@go/page/349939
26.3: Zone Broadening and Column Efficiency
Suppose we inject a sample that has a single component. At the moment we inject the sample it is a narrow band of finite width. As
the sample passes through the column, the width of this band continually increases in a process we call band broadening. Column
efficiency is a quantitative measure of the extent of band broadening.

The Shape of Chromatographic Peaks

When we inject a sample onto a column it has a uniform, or rectangular concentration profile with respect to distance down the
column. As the sample passes through the column, the individual solute particles move in and out of stationary phase, remaining in
place when in the stationary phase and moving down the column when in the mobile phase. Because some solute particles will, on
average, spend more time in the mobile phase and some will, on average, spend more time in the stationary phase, the original
rectangular band increases in width and takes on a Gaussian shape, as we see in Figure 26.3.1.

Figure 26.3.1 . The initial injection of a sample results in the rectangular band of solutes seen in (a). As the sample migrates
through the column, the individual components of the sample take on a Gaussian-shaped band profile, as seen in (b-d).
Our treatment of chromatography in this section assumes that a solute elutes as a symmetrical Gaussian peak, such as that shown in
Figure 26.3.1. This ideal behavior occurs when the solute’s partition coefficient, KD
[ Ss ]
KD = (26.3.1)
[ Sm ]

is the same for all concentrations of solute. If this is not the case, then the chromatographic peak has an asymmetric peak shape
similar to those shown in Figure 26.3.2. The chromatographic peak in Figure 26.3.2a is an example of peak tailing, which occurs
when some sites on the stationary phase retain the solute more strongly than other sites. Figure 26.3.2b, which is an example of
peak fronting most often is the result of overloading the column with sample.

Figure 26.3.2 . Examples of asymmetric chromatographic peaks showing (a) peak tailing and (b) peak fronting. For both (a) and (b)
the green chromatogram is the asymmetric peak and the red dashed chromatogram shows the ideal, Gaussian peak shape. The
insets show the relationship between the concentration of solute in the stationary phase, [S]s, and its concentration in the mobile
phase, [S]m. The dashed red lines show ideal behavior (KD is constant for all conditions) and the green lines show nonideal
behavior (KD decreases or increases for higher total concentrations of solute). A quantitative measure of peak tailing, T, is shown in
(a).

26.3.1 https://chem.libretexts.org/@go/page/349941
As shown in Figure 26.3.2a, we can report a peak’s asymmetry by drawing a horizontal line at 10% of the peak’s maximum height
and measuring the distance from each side of the peak to a line drawn vertically through the peak’s maximum. The asymmetry
factor, T, is defined as
b
T = (26.3.2)
a

Methods for Describing Column Efficiency

In their original theoretical model of chromatography, Martin and Synge divided the chromatographic column into discrete
sections, which they called theoretical plates. Within each theoretical plate there is an equilibrium between the solute present in the
stationary phase and the solute present in the mobile phase [Martin, A. J. P.; Synge, R. L. M. Biochem. J. 1941, 35, 1358–1366].
They described column efficiency in terms of the number of theoretical plates, N,
L
N = (26.3.3)
H

where L is the column’s length and H is the height of a theoretical plate. For any given column, the column efficiency improves—
and chromatographic peaks become narrower—when there are more theoretical plates.
If we assume that a chromatographic peak has a Gaussian profile, then the extent of band broadening is given by the peak’s
variance or standard deviation. The height of a theoretical plate is the peak’s variance per unit length of the column
2
σ
H = (26.3.4)
L

where the standard deviation, σ, has units of distance. Because retention times and peak widths usually are measured in seconds or
minutes, it is more convenient to express the standard deviation in units of time, τ , by dividing σ by the solute’s average linear
velocity, u, which is equivalent to dividing the distance it travels, L, by its retention time, tr.
¯¯
¯

σ σtr
τ = = (26.3.5)
¯¯
¯
u L

For a Gaussian peak shape, the width at the baseline, w, is four times its standard deviation, τ .

w = 4τ (26.3.6)

Combining Equation 26.3.4, Equation 26.3.5, and Equation 26.3.6 defines the height of a theoretical plate in terms of the easily
measured chromatographic parameters tr and w.
2
Lw
H = (26.3.7)
2
16tr

Combing Equation 26.3.7 and Equation 26.3.3 gives the number of theoretical plates.
2 2
tr tr
N = 16 = 16 ( ) (26.3.8)
2
w w

 Example 26.3.4

A chromatographic analysis for the chlorinated pesticide Dieldrin gives a peak with a retention time of 8.68 min and a baseline
width of 0.29 min. Calculate the number of theoretical plates? Given that the column is 2.0 m long, what is the height of a
theoretical plate in mm?

Solution
Using Equation 26.3.8, the number of theoretical plates is
2 2
tr (8.68 min)
N = 16 = 16 × = 14300 plates
2 2
w (0.29 min)

Solving Equation 26.3.3 for H gives the average height of a theoretical plate as

26.3.2 https://chem.libretexts.org/@go/page/349941
 L 2.00 m 1000 mm
H = = × = 0.14 mm/plate
N 14300 plates m

It is important to remember that a theoretical plate is an artificial construct and that a chromatographic column does not contain
physical plates. In fact, the number of theoretical plates depends on both the properties of the column and the solute. As a result,
the number of theoretical plates for a column may vary from solute to solute.
The number of theoretical plates for an asymmetric peak shape is approximately
2 2
t t
r r
41.7 × 2
41.7 × 2
( w0.1 ) (a+b)
N ≈ = (26.3.9)
T + 1.25 T + 1.25

where w0.1 is the width at 10% of the peak’s height [Foley, J. P.; Dorsey, J. G. Anal. Chem. 1983, 55, 730–737].

Asymmetric peaks have fewer theoretical plates, and the more asymmetric the peak the smaller the number of theoretical
plates. For example, the following table gives values for N for a solute eluting with a retention time of 10.0 min and a peak
width of 1.00 min.

b a T N

0.5 0.5 1.00 1850

0.6 0.4 1.50 1520

0.7 0.3 2.33 1160

0.8 0.2 4.00 790

Kinetic Variables Affecting Zone Broadening

Another approach to understanding the broadening as a solute band passes through a column is to consider the factors that affect
the rate at which a solute moves through the column and how that is affected by the velocity with which the mobile phase moves
through the column. We will consider one approach that considers four contributions: variations in path lengths, longitudinal
diffusion, mass transfer in the stationary phase, and mass transfer in the mobile phase.

Multiple Paths: Variations in Path Length

As solute molecules pass through the column they travel paths that differ in length. Because of this difference in path length, two
solute molecules that enter the column at the same time will exit the column at different times. The result, as shown in Figure
26.3.3, is a broadening of the solute’s profile on the column. The contribution of multiple paths to the height of a theoretical plate,

Hp, is

Hp = 2λ dp (26.3.10)

where dp is the average diameter of the particulate packing material and λ is a constant that accounts for the consistency of the
packing. A smaller range of particle sizes and a more consistent packing produce a smaller value for λ . For a column without
packing material, Hp is zero and there is no contribution to band broadening from multiple paths.

26.3.3 https://chem.libretexts.org/@go/page/349941
 Figure 26.3.3 . The effect of multiple paths on a solute’s band broadening. The solute’s initial band profile is rectangular. As this
band travels through the column, individual solute molecules travel different paths, three of which are shown by the meandering
colored paths (the actual lengths of these paths are shown by the straight arrows at the bottom of the figure). Most solute molecules
travel paths with lengths similar to that shown in blue, with a few traveling much shorter paths (green) or much longer paths (red).
As a result, the solute’s band profile at the end of the column is broader and Gaussian in shape.

An inconsistent packing creates channels that allow some solute molecules to travel quickly through the column. It also can
creates pockets that temporarily trap some solute molecules, slowing their progress through the column. A more uniform
packing minimizes these problems.

Longitudinal Diffusion
The second contribution to band broadening is the result of the solute’s longitudinal diffusion in the mobile phase. Solute molecules
are in constant motion, diffusing from regions of higher solute concentration to regions where the concentration of solute is smaller.
The result is an increase in the solute’s band width (Figure 26.3.4). The contribution of longitudinal diffusion to the height of a
theoretical plate, Hd, is
2γDm
Hd = (26.3.11)
u

where Dm is the solute’s diffusion coefficient in the mobile phase, u is the mobile phase’s velocity, and γ is a constant related to the
efficiency of column packing. Note that the effect of Hd on band broadening is inversely proportional to the mobile phase velocity:
a higher velocity provides less time for longitudinal diffusion. Because a solute’s diffusion coefficient is larger in the gas phase
than in a liquid phase, longitudinal diffusion is a more serious problem in gas chromatography.

Figure 26.3.4 . The effect of longitudinal diffusion on a solute’s band broadening. Two horizontal cross-sections through the
column and the corresponding concentration versus distance profiles are shown, with (a) being earlier in time. The red arrow shows
the direction in which the mobile phase is moving.

Mass Transfer
As the solute passes through the column it moves between the mobile phase and the stationary phase. We call this movement
between phases mass transfer. As shown in Figure 26.3.5, band broadening occurs if the solute’s movement within the mobile
phase or within the stationary phase is not fast enough to maintain an equilibrium in its concentration between the two phases. On
average, a solute molecule in the mobile phase moves down the column before it passes into the stationary phase. A solute
molecule in the stationary phase, on the other hand, takes longer than expected to move back into the mobile phase. The
contributions of mass transfer in the stationary phase, Hs, and mass transfer in the mobile phase, Hm, are given by the following
equations

26.3.4 https://chem.libretexts.org/@go/page/349941
 2
qkd
f
Hs = u (26.3.12)
2
(1 + k) Ds

2 2
f n (dp , dc )
Hm = u (26.3.13)
Dm

where df is the thickness of the stationary phase, dc is the diameter of the column, Ds and Dm are the diffusion coefficients for the
solute in the stationary phase and the mobile phase, k is the solute’s retention factor, and q is a constant related to the column
packing material. Although the exact form of Hm is not known, it is a function of particle size and column diameter. Note that the
effect of Hs and Hm on band broadening is directly proportional to the mobile phase velocity because a smaller velocity provides
more time for mass transfer.

The abbreviation fn in Equation 26.3.13 means “is a function of.”

Figure 26.3.5 . Effect of mass transfer on band broadening: (a) Ideal equilibrium Gaussian profiles for the solute in the mobile
phase and in the stationary phase. (b, c) If we allow the solute’s band to move a small distance down the column, an equilibrium
between the two phases no longer exits. The red arrows show the movement of solute—what we call the mass transfer of solute—
from the stationary phase to the mobile phase, and from the mobile phase to the stationary phase. (d) Once equilibrium is
reestablished, the solute’s band is now broader.

Putting It All Together

The height of a theoretical plate is a summation of the contributions from each of the terms affecting band broadening.
H = Hp + Hd + Hs + Hm (26.3.14)

An alternative form of this equation is the van Deemter equation

B
H = A+ + Cu (26.3.15)
u

which emphasizes the importance of the mobile phase’s velocity. In the van Deemter equation, A accounts for the contribution of
multiple paths (Hp), B/u accounts for the contribution of longitudinal diffusion (Hd), and Cu accounts for the combined contribution
of mass transfer in the stationary phase and in the mobile phase (Hs and Hm).
There is some disagreement on the best equation for describing the relationship between plate height and mobile phase velocity
[Hawkes, S. J. J. Chem. Educ. 1983, 60, 393–398]. In addition to the van Deemter equation, other equations include
B
H = + (Cs + Cm ) u (26.3.16)
u

where Cs and Cm are the mass transfer terms for the stationary phase and the mobile phase and
B
1/3
H = Au + + Cu (26.3.17)
u

All three equations, and others, have been used to characterize chromatographic systems, with no single equation providing the best
explanation in every case [Kennedy, R. T.; Jorgenson, J. W. Anal. Chem. 1989, 61, 1128–1135].
To increase the number of theoretical plates without increasing the length of the column, we need to decrease one or more of the
terms in Equation 26.3.14. The easiest way to decrease H is to adjust the velocity of the mobile phase. For smaller mobile phase
velocities, column efficiency is limited by longitudinal diffusion, and for higher mobile phase velocities efficiency is limited by the

26.3.5 https://chem.libretexts.org/@go/page/349941
two mass transfer terms. As shown in Figure 26.3.6 —which uses the van Deemter equation—the optimum mobile phase velocity
is the minimum in a plot of H as a function of u.

Figure 26.3.6 . Plot showing the relationship between the height of a theoretical plate, H, and the mobile phase’s velocity, u, based
on the van Deemter equation.
The remaining parameters that affect the terms in Equation 26.3.14 are functions of the column’s properties and suggest other
possible approaches to improving column efficiency. For example, both Hp and Hm are a function of the size of the particles used to
pack the column. Decreasing particle size, therefore, is another useful method for improving efficiency.

For a more detailed discussion of ways to assess the quality of a column, see Desmet, G.; Caooter, D.; Broeckhaven, K.
“Graphical Data Represenation Methods to Assess the Quality of LC Columns,” Anal. Chem. 2015, 87, 8593–8602.

Perhaps the most important advancement in chromatography columns is the development of open-tubular, or capillary columns.
These columns have very small diameters (dc ≈ 50–500 μm) and contain no packing material (dp = 0). Instead, the capillary
column’s interior wall is coated with a thin film of the stationary phase. Plate height is reduced because the contribution to H from
Hp (Equation 26.3.10) disappears and the contribution from Hm (Equation 26.3.13) becomes smaller. Because the column does not
contain any solid packing material, it takes less pressure to move the mobile phase through the column, which allows for longer
columns. The combination of a longer column and a smaller height for a theoretical plate increases the number of theoretical plates
by approximately 100×. Capillary columns are not without disadvantages. Because they are much narrower than packed columns,
they require a significantly smaller amount of sample, which may be difficult to inject reproducibly. Another approach to
improving resolution is to use thin films of stationary phase, which decreases the contribution to H from Hs (Equation 26.3.12).

The smaller the particles, the more pressure is needed to push the mobile phase through the column. As a result, for any form
of chromatography there is a practical limit to particle size.

This page titled 26.3: Zone Broadening and Column Efficiency is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

26.3.6 https://chem.libretexts.org/@go/page/349941
26.4: Optimization and Column Performance
The goal of a chromatographic separation is to take a sample with more than one solute and to separate the solutes such that each
solute elutes by itself. Our ability to separate two solutes from each other—to resolve them—is affected by a number of variables;
how we can optimize the separation of two solutes, is the subject of this section.

Column Resolution
The goal of chromatography is to separate a mixture into a series of chromatographic peaks, each of which constitutes a single
component of the mixture. The resolution between two chromatographic peaks, RAB, is a quantitative measure of their separation,
and is defined as
tt,B − tt,A 2Δtr
RAB = = (26.4.1)
0.5 (wB + wA ) wB + wA

where B is the later eluting of the two solutes. As shown in Figure 26.4.1, the separation of two chromatographic peaks improves
with an increase in RAB. If the areas under the two peaks are identical—as is the case in Figure 26.4.1—then a resolution of 1.50
corresponds to an overlap of only 0.13% for the two elution profiles. Because resolution is a quantitative measure of a separation’s
success, it is a useful way to determine if a change in experimental conditions leads to a better separation.

Figure 26.4.1 . Three examples that show the relationship between resolution and the separation of a two component mixture. The
green peak and the red peak are the elution profiles for the two components. The chromatographic peak—which is the sum of the
two elution profiles—is shown by the solid black line.

 Example 26.4.1

In a chromatographic analysis of lemon oil a peak for limonene has a retention time of 8.36 min with a baseline width of 0.96
min. γ-Terpinene elutes at 9.54 min with a baseline width of 0.64 min. What is the resolution between the two peaks?

Solution
Using Equation 26.4.1 we find that the resolution is
2Δtr 2(9.54 min − 8.36 min)
RAB = = = 1.48
wB + wA 0.64 min + 0.96 min

The Effect of Retention and Selectivity Factors on Resolution

Now that we have defined the solute retention factor, selectivity, and column efficiency we are able to consider how they affect the
resolution of two closely eluting peaks. Because the two peaks have similar retention times, it is reasonable to assume that their
peak widths are nearly identical. If the number of theoretical plates is the same for all solutes—not strictly true, but not a bad
assumption—then the ratio tr/w is a constant. If two solutes have similar retention times, then their peak widths must be similar.
Equation 26.4.1, therefore, becomes
tr,B − tr,A tr,B − tr,A tr,B − tr,A
RAB = ≈ = (26.4.2)
0.5 (wB + wA ) 0.5 (2 wB ) wB

26.4.1 https://chem.libretexts.org/@go/page/349943
where B is the later eluting of the two solutes. Solving equation 26.3.8 for wB and substituting into Equation 26.4.2 leaves us with
the following result.
−−
−
√NB tr,B − tr,A
RAB = × (26.4.3)
4 tr,B

Rearranging equation 26.2.11 provides us with the following equations for the retention times of solutes A and B.
tr,A = kA tm + tm (26.4.4)

tr,B = kB tm + tm (26.4.5)

After substituting these equations into Equation 26.4.3 and simplifying, we have
−−
−
√NB kB − kA
RAB = × (26.4.6)
4 1 + kB

Finally, we can eliminate solute A’s retention factor by substituting in equation 26.2.12. After rearranging, we end up with the
following equation for the resolution between the chromatographic peaks for solutes A and B.
−−
−
√NB α −1 kB
RAB = × × (26.4.7)
4 α 1 + kB

Although Equation 26.4.7 is useful for considering how a change in N, α , or k qualitatively affects resolution—which suits
our purpose here—it is less useful for making accurate quantitative predictions of resolution, particularly for smaller values of
N and for larger values of R. For more accurate predictions use the equation
−
−
√N kB
RAB = × (α − 1) ×
4 1 + kavg

where kavg is (kA + kB)/2. For a derivation of this equation and for a deeper discussion of resolution in column chromatography,
see Foley, J. P. “Resolution Equations for Column Chromatography,” Analyst, 1991, 116, 1275-1279.

Equation 26.4.7 contains terms that correspond to column efficiency, selectivity, and the solute retention factor. We can vary these
terms, more or less independently, to improve resolution and analysis time. The first term, which is a function of the number of
theoretical plates (for Equation 26.4.7), accounts for the effect of column efficiency. The second term is a function of α and
accounts for the influence of column selectivity. Finally, the third term in both equations is a function of kB and accounts for the
effect of solute B’s retention factor. A discussion of how we can use these parameters to improve resolution is the subject of the
remainder of this section.

The Effect of Resolution on Retention Time

In addition to resolution, another important factor in chromatography is the amount of time needed to elute a pair of solutes, which
we can approximate using the retention time for solute B.
2 2 3
16 R H α (1 + kB )
AB
tr,B = ×( ) × (26.4.8)
2
u α −1 k
B

where u is the mobile phase’s velocity.

Variables That Affect Column Performance

Using the Retention Factor to Optimize Resolution
One of the simplest ways to improve resolution is to adjust the retention factor for solute B. If all other terms in Equation 26.4.7
remain constant, an increase in kB will improve resolution. As shown by the green curve in Figure 26.4.2 , however, the
improvement is greatest if the initial value of kB is small. Once kB exceeds a value of approximately 10, a further increase produces
only a marginal improvement in resolution. For example, if the original value of kB is 1, increasing its value to 10 gives an 82%
improvement in resolution; a further increase to 15 provides a net improvement in resolution of only 87.5%.

26.4.2 https://chem.libretexts.org/@go/page/349943
 Figure 26.4.2 . Effect of kB on the resolution for a pair of solutes, RAB, and the retention time for the later eluting solute, tr,B. The y-
axes display the resolution and retention time relative to their respective values when kB is 1.00.
Any improvement in resolution from increasing the value of kB generally comes at the cost of a longer analysis time. The red curve
in Figure 26.4.2 shows the relative change in the retention time for solute B as a function of its retention factor. Note that the
minimum retention time is for kB = 2. Increasing kB from 2 to 10, for example, approximately doubles solute B’s retention time.

The relationship between retention factor and analysis time in Figure 26.4.2 works to our advantage if a separation produces
an acceptable resolution with a large kB. In this case we may be able to decrease kB with little loss in resolution and with a
significantly shorter analysis time.

To increase kB without changing selectivity, α , any change to the chromatographic conditions must result in a general, nonselective
increase in the retention factor for both solutes. In gas chromatography, we can accomplish this by decreasing the column’s
temperature. Because a solute’s vapor pressure is smaller at lower temperatures, it spends more time in the stationary phase and
takes longer to elute. In liquid chromatography, the easiest way to increase a solute’s retention factor is to use a mobile phase that is
a weaker solvent. When the mobile phase has a lower solvent strength, solutes spend proportionally more time in the stationary
phase and take longer to elute.

Using Selectivity to Optimize Resolution

A second approach to improving resolution is to adjust the selectivity, α . In fact, for α ≈ 1 usually it is not possible to improve
resolution by adjusting the solute retention factor, kB, or the column efficiency, N. A change in α often has a more dramatic effect
on resolution than a change in kB. For example, changing α from 1.1 to 1.5, while holding constant all other terms, improves
resolution by 267%. In gas chromatography, we adjust α by changing the stationary phase; in liquid chromatography, we change
the composition of the mobile phase to adjust α .
To change α we need to selectively adjust individual solute retention factors. Figure 26.4.3 shows one possible approach for the
liquid chromatographic separation of a mixture of substituted benzoic acids. Because the retention time of a compound’s weak acid
form and its weak base form are different, its retention time will vary with the pH of the mobile phase, as shown in Figure 26.4.3 a.
The intersections of the curves in Figure 26.4.3 a show pH values where two solutes co-elute. For example, at a pH of 3.8
terephthalic acid and p-hydroxybenzoic acid elute as a single chromatographic peak.

26.4.3 https://chem.libretexts.org/@go/page/349943
 Figure 26.4.3 . Example showing how the mobile phase pH in liquid chromatography affects selectivity: (a) retention times for four
substituted benzoic acids as a function of the mobile phase’s pH; (b) alpha values for three pairs of solutes that are difficult to
separate. See text for details. The mobile phase is an acetic acid/sodium acetate buffer and the stationary phase is a nonpolar
hydrocarbon. Data from Harvey, D. T.; Byerly, S.; Bowman, A.; Tomlin, J. “Optimization of HPLC and GC Separations Using
Response Surfaces,” J. Chem. Educ. 1991, 68, 162–168.
Figure 26.4.3 a shows that there are many pH values where some separation is possible. To find the optimum separation, we plot a
for each pair of solutes. The red, green, and orange curves in Figure 26.4.3 b show the variation in a with pH for the three pairs of
solutes that are hardest to separate (for all other pairs of solutes, α > 2 at all pH levels). The blue shading shows windows of pH
values in which at least a partial separation is possible—this figure is sometimes called a window diagram—and the highest point
in each window gives the optimum pH within that range. The best overall separation is the highest point in any window, which, for
this example, is a pH of 3.5. Because the analysis time at this pH is more than 40 min (Figure 26.4.3 a), choosing a pH between
4.1–4.4 might produce an acceptable separation with a much shorter analysis time.

Let’s use benzoic acid, C6H5COOH, to explain why pH can affect a solute’s retention time. The separation uses an aqueous
mobile phase and a nonpolar stationary phase. At lower pHs, benzoic acid predominately is in its weak acid form,
C6H5COOH, and partitions easily into the nonpolar stationary phase. At more basic pHs, however, benzoic acid is in its weak
base form, C6H5COO–. Because it now carries a charge, its solubility in the mobile phase increases and its solubility in the
nonpolar stationary phase decreases. As a result, it spends more time in the mobile phase and has a shorter retention time.

Although the usual way to adjust pH is to change the concentration of buffering agents, it also is possible to adjust pH by
changing the column’s temperature because a solute’s pKa value is pH-dependent; for a review, see Gagliardi, L. G.; Tascon,
M.; Castells, C. B. “Effect of Temperature on Acid–Base Equilibria in Separation Techniques: A Review,” Anal. Chim. Acta,
2015, 889, 35–57.

Using Column Efficiency to Optimize Resolution

A third approach to improving resolution is to adjust the column’s efficiency by increasing the number of theoretical plates, N. If
we have values for kB and α , then we can use Equation 26.4.7 to calculate the number of theoretical plates for any resolution. Table
26.4.1 provides some representative values. For example, if α = 1.05 and kB = 2.0, a resolution of 1.25 requires approximately 24
800 theoretical plates. If our column provides only 12 400 plates, half of what is needed, then a separation is not possible. How can
we double the number of theoretical plates? The easiest way is to double the length of the column, although this also doubles the
analysis time. A better approach is to cut the height of a theoretical plate, H, in half, providing the desired resolution without
changing the analysis time. Even better, if we can decrease H by more than 50%, it may be possible to achieve the desired
resolution with an even shorter analysis time by also decreasing kB or α .
Table 26.4.1 . Minimum Number of Theoretical Plates to Achieve Desired Resolution for Selected Values of kB and α
RAB = 1.00 RAB = 1.25 RAB = 1.50

kB α = 1.05 α = 1.10 α = 1.05 α = 1.10 α = 1.05 α = 1.10

0.5 63500 17400 99200 27200 143000 39200

1.0 28200 7740 44100 12100 63500 17400

1.5 19600 5380 30600 8400 44100 12100

26.4.4 https://chem.libretexts.org/@go/page/349943
 RAB = 1.00 RAB = 1.25 RAB = 1.50

kB α = 1.05 α = 1.10 α = 1.05 α = 1.10 α = 1.05 α = 1.10

2.0 15900 4360 24800 6810 35700 9800

3.0 12500 3440 19600 5380 28200 7740

5.0 10200 2790 15900 4360 22900 6270

10.0 8540 2340 13300 3660 19200 5270

The General Elution Problem

Adjusting the retention factor to improve the resolution between one pair of solutes may lead to unacceptably long retention times
for other solutes. For example, suppose we need to analyze a four-component mixture with baseline resolution and with a run-time
of less than 20 min. Our initial choice of conditions gives the chromatogram in Figure 26.4.4 a. Although we successfully separate
components 3 and 4 within 15 min, we fail to separate components 1 and 2. Adjusting conditions to improve the resolution for the
first two components by increasing k2 provides a good separation of all four components, but the run-time is too long (Figure
26.4.4 b). This problem of finding a single set of acceptable operating conditions is known as the general elution problem.

Figure 26.4.4 . Example showing the general elution problem in chromatography. See text for details.
One solution to the general elution problem is to make incremental adjustments to the retention factor as the separation takes place.
At the beginning of the separation we set the initial chromatographic conditions to optimize the resolution for early eluting solutes.
As the separation progresses, we adjust the chromatographic conditions to decrease the retention factor—and, therefore, to decrease
the retention time—for each of the later eluting solutes (Figure 26.4.4 c). In gas chromatography this is accomplished by
temperature programming. The column’s initial temperature is selected such that the first solutes to elute are resolved fully. The
temperature is then increased, either continuously or in steps, to bring off later eluting components with both an acceptable
resolution and a reasonable analysis time. In liquid chromatography the same effect is obtained by increasing the solvent’s eluting
strength. This is known as a gradient elution. We will have more to say about each of these in later sections of this chapter.

This page titled 26.4: Optimization and Column Performance is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

26.4.5 https://chem.libretexts.org/@go/page/349943
26.5: Summary of Important Relationships for Chromatography
In this chapter we have introduced many chromatographic variables, some directly measured from the chromatogram, provided by
the manufacturer, or from the operating conditions, and some derived from these variables. The following two tables summarize
these variables.
Table 26.5.1 . Chromatographic Variables Directly Measured From Chromatogram, Provided by Manufacturer, or From Operating Conditions
variable name source

tr retention time for solute chromatogram

tm retention time for non-retained solute chromatogram

w peak width chromatogram

u mobile phase flow rate operating conditions

L length of column's stationary phase manufacturer

dc diameter of column manufacturer

dp diameter of packing material manufacturer

df thickness of stationary phase manufacturer

Vs volume of stationary phase operating conditions

Table 26.5.2 . Derived Chromatographic Variables

variable name equation to derive value

Vm volume of mobile phase Vm = tm u

tr −tm
k retention factor k =
tm

′
tr adjusted retention time ′
tr = tr − tm

Vs
D distribution ratio D = k×
Vm

kB
α selectivity factor α =
kA

√NB
RAB resolution RAB = ×
α−1
×
kB

4 α 1+kB

2
N number of theoretical plates N = 16(
tr

w
)

2
Lw L
H height of theoretical plates H =
16tr
2
=
N

Hp height due to multiple paths Hp = 2λ dp

Hd height due to longitudinal diffusion Hd =

2γDm

2
q kdf
Hs height due to mass transfer in stationary phase Hs =
2
u
(1+k) Ds

2 2
fn(dp ,dc )
Hm height due to mass transfer in mobile phase Hm = u
Dm

This page titled 26.5: Summary of Important Relationships for Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

26.5.1 https://chem.libretexts.org/@go/page/349945
26.6: Applications of Chromatography
Although the primary purpose of chromatography is the separation of a complex mixture into its component parts, as outlined here,
a chromatographic separation also provides qualitative and quantitative information about our samples. More detailed examples of
qualitative and quantitative applications are found in the chapters that follow.

Qualitative Analysis
As we learned in Section 26.2, solutes migrate through a chromatographic system at rate that is a function of the properties of the
mobile phase and the stationary phase. This means that a particular solute will elute with a consistent retention time. If we expect a
solute to elute with a retention time of 5.0 min, the presence of a peak at 5.0 min is suggestive of, but not definitive evidence of the
solute's presence in our sample; however, the absence of a peak at 5.0 min is strong evidence that the solute is not present in our
sample. For a complex mixture, this sort of screening technique is a useful qualitative application of chromatography. As we will
see in the chapters that follow, the type of detector used to monitor a chromatographic separation may provide useful qualitative
information.

Quantitative Analysis
A chromatographic separation yields a sequence of peaks that, ideally, represent a single solute. These peak's are characterized by
an area that is proportional to the amount of analyte injected into the mobile phase. By injecting a series of standards, a calibration
curve of peak area as a function of the analyte's concentration provides a way to determine the analyte's concentration in a sample.
Any of the calibration strategies discussed in Chapter 1.5—external standards, standard additions, and internal standards—find use
in a quantitative chromatographic analysis. Determining a solute's peak area is relatively straight-forward when using a computer-
interfaced instrument with appropriate software. Alternatively, peak height, which is easier to measure, can be used as a substitute,
although care must be taken to ensure that the peaks are symmetrical and that peak widths are consistent for the standards and the
samples.

This page titled 26.6: Applications of Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

26.6.1 https://chem.libretexts.org/@go/page/362558
 CHAPTER OVERVIEW
27: Gas Chromatography
In gas chromatography (GC) we inject the sample, which may be a gas or a liquid, into a gaseous mobile phase (often called the
carrier gas). The mobile phase carries the sample through a packed or capillary column that separates the sample's components
based on their ability to partition between the mobile phase and the stationary phase. Because it combines separation with analysis,
gas chromatography provides excellent selectivity. By adjusting conditions it usually is possible to design a separation so that
analytes elute by themselves, even when the mixture is complex. Additional selectivity is possible by using a detector that does not
respond to all analytes.
27.1: Principles of Gas Chromatography
27.2: Instruments for Gas Chromatography
27.3: Gas Chromatographic Columns and Stationary Phases
27.4: Applications of Gas Chromatography

This page titled 27: Gas Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1
27.1: Principles of Gas Chromatography
In Chapter 26 we covered several important elements of chromatography, including the factors that affect the migration of solutes,
the factors that contribute to band broadening, and the factors under our control that we can use to optimize the separation of a
mixture. Here we consider two topics that apply to a gas chromatographic separation, both of which are a function of the properties
of gases.

Retention Times and Retention Volumes

Many of the chromatographic variables in gathered in the tables included in Chapter 26.5—both those that are measured directly,
provided by the manufacturer, or given by the operating conditions, and those derived from these variables—are given in terms or
retention times for the solutes, t , and for the mobile phase, t . The product of time and flow rate is a volume
r m

Vr = tr × u

Vm = tr × u

where V and V are the volume of mobile phase needed to elute a solute and the volume of mobile phase needed to elute a non-
r m

retained solute, which allows us to describe the retention in terms of volumes instead of times.
Because the volume of gas is a function of pressure, and the pressure drops across the column from an inlet pressure of P to an i

outlet pressure of P , the retention times are particularly sensitive to operating condition. We can, however, correct the retention
o

volumes by accounting for the compressibility of the gas

o
Vr = jtr u

o
Vm = jtm u

where j is a correction factor that accounts for the drop in pressure

2
3 × (Pi / Po ) −1
j=
3
2 × (Pi / Po ) −1

and where V and V are the corrected retention volumes for the solute and the non-retained solutes, respectively. The solute's
r
o o
m

corrected retention volume can be further normalized by dividing the adjusted retention volume, V − V , by the mass of the
r
o o
m

stationary phase, w, and by adjusting for the column's temperature, T , relative to 273 K
c

o o
Vr − Vm 273
Vg = ×
w Tc

yielding the solute's specific retention volume, Vg . This value is reasonably insensitive to operating conditions, which makes it
useful for qualitative purposes.

Effect of Diffusion in the Gas Phase on Band Broadening

In Chapter 26 we considered three factors that affect band broadening—multiple paths, longitudinal diffusion, and mass transfer—
expressing the relationship between the height of a theoretical plate, H , as a function of the mobile phase's velocity, u, using the
van Deemter equation
B
H = A+ + Cu
u

where A is the contribution from multiple paths, B is the contribution from longitudinal diffusion, and C is the contribution from
mass transfer. Because solutes have large diffusion coefficients in the gas phase, the term B/u is often the limiting factor in gas
chromatography.

This page titled 27.1: Principles of Gas Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

27.1.1 https://chem.libretexts.org/@go/page/350896
27.2: Instruments for Gas Chromatography
In gas chromatography (GC) we inject the sample, which may be a gas or a liquid, into an gaseous mobile phase (often called the
carrier gas). The mobile phase carries the sample through a packed or a capillary column that separates the sample’s components
based on their ability to partition between the mobile phase and the stationary phase. Figure 27.2.1 shows an example of a typical
gas chromatograph, which consists of several key components: a supply of compressed gas for the mobile phase; a heated injector,
which rapidly volatilizes the components in a liquid sample; a column, which is placed within an oven whose temperature we can
control during the separation; and a detector to monitor the eluent as it comes off the column. Let’s consider each of these
components.

Figure 27.2.1 . Example of a typical gas chromatograph with insets showing the heated injection ports—note the symbol indicating
that it is hot—and the oven that houses the column. This particular instrument is equipped with an autosampler for injecting
samples, a capillary column, and a mass spectrometer (MS) as the detector. Note that the carrier gas is supplied by a tank of
compressed gas.

Mobile Phase
The most common mobile phases for gas chromatography are He, Ar, and N2, which have the advantage of being chemically inert
toward both the sample and the stationary phase. The choice of carrier gas often is determined by the needs of instrument’s
detector. For a packed column the mobile phase velocity usually is 25–150 mL/min. The typical flow rate for a capillary column is
1–25 mL/min.

Sample Introduction
Three factors determine how we introduce a sample to the gas chromatograph. First, all of the sample’s constituents must be
volatile. Second, the analytes must be present at an appropriate concentration. Finally, the physical process of injecting the sample
must not degrade the separation. Each of these needs is considered in this section.

Preparing a Volatile Sample

Not every sample can be injected directly into a gas chromatograph. To move through the column, the sample’s constituents must
be sufficiently volatile. A solute of low volatility, for example, may be retained by the column and continue to elute during the
analysis of subsequent samples. A nonvolatile solute will condense at the top of the column, degrading the column’s performance.
An attractive approach to isolating analytes is a solid-phase microextraction (SPME). In one approach, which is illustrated in
Figure 27.2.2 , a fused-silica fiber is placed inside a syringe needle. The fiber, which is coated with a thin film of an adsorbent
material, such as polydimethyl siloxane, is lowered into the sample by depressing a plunger and is exposed to the sample for a
predetermined time. After withdrawing the fiber into the needle, it is transferred to the gas chromatograph for analysis.

27.2.1 https://chem.libretexts.org/@go/page/350898
 Figure 27.2.2 . Schematic diagram of a solid-phase microextraction device. The absorbent is shown in red.
Two additional methods for isolating volatile analytes are a purge-and-trap and headspace sampling. In a purge-and-trap, we bubble
an inert gas, such as He or N2, through the sample, releasing—or purging—the volatile compounds. These compounds are carried
by the purge gas through a trap that contains an absorbent material, such as Tenax, where they are retained. Heating the trap and
back-flushing with carrier gas transfers the volatile compounds to the gas chromatograph. In headspace sampling we place the
sample in a closed vial with an overlying air space. After allowing time for the volatile analytes to equilibrate between the sample
and the overlying air, we use a syringe to extract a portion of the vapor phase and inject it into the gas chromatograph.
Alternatively, we can sample the headspace with an SPME.
Thermal desorption is a useful method for releasing volatile analytes from solids. We place a portion of the solid in a glass-lined,
stainless steel tube. After purging with carrier gas to remove any O2 that might be present, we heat the sample. Volatile analytes are
swept from the tube by an inert gas and carried to the GC. Because volatilization is not a rapid process, the volatile analytes often
are concentrated at the top of the column by cooling the column inlet below room temperature, a process known as cryogenic
focusing. Once volatilization is complete, the column inlet is heated rapidly, releasing the analytes to travel through the column.

The reason for removing O2 is to prevent the sample from undergoing an oxidation reaction when it is heated.

To analyze a nonvolatile analyte we must convert it to a volatile form. For example, amino acids are not sufficiently volatile to
analyze directly by gas chromatography. Reacting an amino acid, such as valine, with 1-butanol and acetyl chloride produces an
esterified amino acid. Subsequent treatment with trifluoroacetic acid gives the amino acid’s volatile N-trifluoroacetyl-n-butyl ester
derivative.

Adjusting the Analyte's Concentration

In an analyte’s concentration is too small to give an adequate signal, then we must concentrate the analyte before we inject the
sample into the gas chromatograph. A side benefit of many extraction methods is that they often concentrate the analytes. Volatile
organic materials isolated from an aqueous sample by a purge-and-trap, for example, are concentrated by as much as 1000×.
If an analyte is too concentrated, it is easy to overload the column, resulting in peak fronting and a poor separation. In addition, the
analyte’s concentration may exceed the detector’s linear response. Injecting less sample or diluting the sample with a volatile
solvent, such as methylene chloride, are two possible solutions to this problem.

27.2.2 https://chem.libretexts.org/@go/page/350898
Injecting the Sample
In Chapter 26 we examined several explanations for why a solute’s band increases in width as it passes through the column, a
process we called band broadening. We also introduce an additional source of band broadening if we fail to inject the sample into
the minimum possible volume of mobile phase. There are two principal sources of this precolumn band broadening: injecting the
sample into a moving stream of mobile phase and injecting a liquid sample instead of a gaseous sample. The design of a gas
chromatograph’s injector helps minimize these problems.
An example of a simple injection port for a packed column is shown in Figure 27.2.3 . The top of the column fits within a heated
injector block, with carrier gas entering from the bottom. The sample is injected through a rubber septum using a microliter
syringe, such as the one shown in in Figure 27.2.4 . Injecting the sample directly into the column minimizes band broadening
because it mixes the sample with the smallest possible amount of carrier gas. The injector block is heated to a temperature at least
50oC above the boiling point of the least volatile solute, which ensures a rapid vaporization of the sample’s components.

Figure 27.2.3 . Schematic diagram of a heated GC injector port for use with packed columns. The needle pierces a rubber septum
and enters into the top of the column, which is located within a heater block.

Figure 27.2.4 . Example of a syringe for injecting samples into a gas chromatograph. This syringe has a maximum capacity of 10
μL with graduations every 0.1 μL.
Because a capillary column’s volume is significantly smaller than that for a packed column, it requires a different style of injector
to avoid overloading the column with sample. Figure 27.2.5 shows a schematic diagram of a typical split/splitless injector for use
with a capillary column.

27.2.3 https://chem.libretexts.org/@go/page/350898
 Figure 27.2.5 . Schematic diagram of a split/splitless injection port for use with capillary columns. The needle pierces a rubber
septum and enters into a glass liner, which is located within a heater block. In a split injection the split vent is open; the split vent is
closed for a splitless injection.
In a split injection we inject the sample through a rubber septum using a microliter syringe. Instead of injecting the sample directly
into the column, it is injected into a glass liner where it mixes with the carrier gas. At the split point, a small fraction of the carrier
gas and sample enters the capillary column with the remainder exiting through the split vent. By controlling the flow rate of the
carrier gas as it enters the injector, and its flow rate through the septum purge and the split vent, we can control the fraction of
sample that enters the capillary column, typically 0.1–10%.

For example, if the carrier gas flow rate is 50 mL/min, and the flow rates for the septum purge and the split vent are 2 mL/min
and 47 mL/min, respectively, then the flow rate through the column is 1 mL/min (= 50 – 2 – 47). The ratio of sample entering
the column is 1/50, or 2%.

In a splitless injection, which is useful for trace analysis, we close the split vent and allow all the carrier gas that passes through the
glass liner to enter the column—this allows virtually all the sample to enters the column. Because the flow rate through the injector
is low, significant precolumn band broadening is a problem. Holding the column’s temperature approximately 20–25oC below the
solvent’s boiling point allows the solvent to condense at the entry to the capillary column, forming a barrier that traps the solutes.
After allowing the solutes to concentrate, the column’s temperature is increased and the separation begins.
For samples that decompose easily, an on-column injection may be necessary. In this method the sample is injected directly into the
column without heating. The column temperature is then increased, volatilizing the sample with as low a temperature as is
practical.

Temperature Control
Control of the column’s temperature is critical to attaining a good separation when using gas chromatography. For this reason the
column is placed inside a thermostated oven (see Figure 27.2.1 ). In an isothermal separation we maintain the column at a constant
temperature. To increase the interaction between the solutes and the stationary phase, the temperature usually is set slightly below
that of the lowest-boiling solute.
One difficulty with an isothermal separation is that a temperature that favors the separation of a low-boiling solute may lead to an
unacceptably long retention time for a higher-boiling solute. Temperature programming provides a solution to this problem. At the
beginning of the analysis we set the column’s initial temperature below that for the lowest-boiling solute. As the separation
progresses, we slowly increase the temperature at either a uniform rate or in a series of steps.

Detectors for Gas Chromatography

The final part of a gas chromatograph is the detector. The ideal detector has several desirable features: a low detection limit, a
linear response over a wide range of solute concentrations (which makes quantitative work easier), sensitivity for all solutes or
selectivity for a specific class of solutes, and an insensitivity to a change in flow rate or temperature.

27.2.4 https://chem.libretexts.org/@go/page/350898
Thermal Conductivity Detector (TCD)
One of the earliest gas chromatography detectors takes advantage of the mobile phase’s thermal conductivity. As the mobile phase
exits the column it passes over a tungsten-rhenium wire filament (see Figure 27.2.6 ). The filament’s electrical resistance depends
on its temperature, which, in turn, depends on the thermal conductivity of the mobile phase. Because of its high thermal
conductivity, helium is the mobile phase of choice when using a thermal conductivity detector (TCD).

Figure 27.2.6 . Schematic diagram of a thermal conductivity detector showing one cell of a matched pair. The sample cell takes the
carrier gas as it elutes from the column. A source of carrier gas that bypasses the column passes through a reference cell.

Thermal conductivity, as the name suggests, is a measure of how easily a substance conducts heat. A gas with a high thermal
conductivity moves heat away from the filament—and, thus, cools the filament—more quickly than does a gas with a low
thermal conductivity.

When a solute elutes from the column, the thermal conductivity of the mobile phase in the TCD cell decreases and the temperature
of the wire filament, and thus it resistance, increases. A reference cell, through which only the mobile phase passes, corrects for any
time-dependent variations in flow rate, pressure, or electrical power, all of which affect the filament’s resistance.
Because all solutes affect the mobile phase’s thermal conductivity, the thermal conductivity detector is a universal detector. Another
advantage is the TCD’s linear response over a concentration range spanning 104–105 orders of magnitude. The detector also is non-
destructive, which allows us to recover analytes using a postdetector cold trap. One significant disadvantage of the TCD detector is
its poor detection limit for most analytes.

Flame Ionization Detector (FID)

The combustion of an organic compound in an H2/air flame results in a flame that contains electrons and organic cations,
presumably CHO+. Applying a potential of approximately 300 volts across the flame creates a small current of roughly 10–9 to 10–
12
amps. When amplified, this current provides a useful analytical signal. This is the basis of the popular flame ionization detector,
a schematic diagram of which is shown in Figure 27.2.7 .

Figure 27.2.7 . Schematic diagram of a flame ionization detector. The eluent from the column mixes with H2 and is burned in the
presence of excess air. Combustion produces a flame that contains electrons and the cation CHO+. Applying a potential between the
flame’s tip and the collector gives a current that is proportional to the concentration of cations in the flame.
Most carbon atoms—except those in carbonyl and carboxylic groups—generate a signal, which makes the FID an almost universal
detector for organic compounds. Most inorganic compounds and many gases, such as H2O and CO2, are not detected, which makes
the FID detector a useful detector for the analysis of organic analytes in atmospheric and aqueous environmental samples.

27.2.5 https://chem.libretexts.org/@go/page/350898
Advantages of the FID include a detection limit that is approximately two to three orders of magnitude smaller than that for a
thermal conductivity detector, and a linear response over 106–107 orders of magnitude in the amount of analyte injected. The
sample, of course, is destroyed when using a flame ionization detector.

Electron Capture Detector (ECD)

The electron capture detector is an example of a selective detector. As shown in Figure 27.2.8 , the detector consists of a β-emitter,
such as 63Ni. The emitted electrons ionize the mobile phase, usually N2, generating a standing current between a pair of electrodes.
When a solute with a high affinity for capturing electrons elutes from the column, the current decreases, which serves as the signal.
The ECD is highly selective toward solutes with electronegative functional groups, such as halogens and nitro groups, and is
relatively insensitive to amines, alcohols, and hydrocarbons. Although its detection limit is excellent, its linear range extends over
only about two orders of magnitude.

A β-particle is an electron.

Figure 27.2.8 . Schematic diagram showing an electron capture detector.

Mass Spectrometer (MS)

A mass spectrometer is an instrument that ionizes a gaseous molecule using sufficient energy that the resulting ion breaks apart into
smaller ions. Because these ions have different mass-to-charge ratios, it is possible to separate them using a magnetic field or an
electrical field. The resulting mass spectrum contains both quantitative and qualitative information about the analyte. Figure 27.2.9
shows a mass spectrum for toluene.

Figure 27.2.9 . Mass spectrum for toluene highlighting the molecular ion in green (m/z=92), and two fragment ions in blue
(m/z=91) and in red (m/z= 65). A mass spectrum provides both quantitative and qualitative information: the height of any peak is
proportional to the amount of toluene in the mass spectrometer and the fragmentation pattern is unique to toluene.
Figure 27.2.10 shows a block diagram of a typical gas chromatography-mass spectrometer (GC–MS) instrument. The effluent from
the column enters the mass spectrometer’s ion source in a manner that eliminates the majority of the carrier gas. In the ionization
chamber the remaining molecules—a mixture of carrier gas, solvent, and solutes—undergo ionization and fragmentation. The mass
spectrometer’s mass analyzer separates the ions by their mass-to-charge ratio and a detector counts the ions and displays the mass
spectrum.

27.2.6 https://chem.libretexts.org/@go/page/350898
 Figure 27.2.10 . Block diagram of GC– MS. A three component mixture enters the GC. When component A elutes from the
column, it enters the MS ion source and ionizes to form the parent ion and several fragment ions. The ions enter the mass analyzer,
which separates them by their mass-to-charge ratio, providing the mass spectrum shown at the detector.
There are several options for monitoring a chromatogram when using a mass spectrometer as the detector. The most common
method is to continuously scan the entire mass spectrum and report the total signal for all ions that reach the detector during each
scan. This total ion scan provides universal detection for all analytes. We can achieve some degree of selectivity by monitoring one
or more specific mass-to-charge ratios, a process called selective-ion monitoring. A mass spectrometer provides excellent detection
limits, typically 25 fg to 100 pg, with a linear range of 105 orders of magnitude. Because we continuously record the mass spectrum
of the column’s eluent, we can go back and examine the mass spectrum for any time increment. This is a distinct advantage for
GC–MS because we can use the mass spectrum to help identify a mixture’s components.

Other Detectors
Two additional detectors are similar in design to a flame ionization detector. In the flame photometric detector, optical emission
from phosphorous and sulfur provides a detector selective for compounds that contain these elements. The thermionic detector
responds to compounds that contain nitrogen or phosphorous.
A Fourier transform infrared spectrophotometer (FT–IR) also can serve as a detector. In GC–FT–IR, effluent from the column
flows through an optical cell constructed from a 10–40 cm Pyrex tube with an internal diameter of 1–3 mm. The cell’s interior
surface is coated with a reflecting layer of gold. Multiple reflections of the source radiation as it is transmit- ted through the cell
increase the optical path length through the sample. As is the case with GC–MS, an FT–IR detector continuously records the
column eluent’s spectrum, which allows us to examine the IR spectrum for any time increment.

This page titled 27.2: Instruments for Gas Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

27.2.7 https://chem.libretexts.org/@go/page/350898
27.3: Gas Chromatographic Columns and Stationary Phases
There are two broad classes of chromatographic columns: packed columns and capillary columns. In general, a packed column can
handle larger samples and a capillary column can separate more complex mixtures.

Packed Columns
Packed columns are constructed from glass, stainless steel, copper, or aluminum, and typically are 2–6 m in length with internal
diameters of 2–4 mm. The column is filled with a particulate solid support, with particle diameters ranging from 37–44 μm to 250–
354 μm. Figure 27.3.1 shows a typical example of a packed column.

Figure 27.3.1 . Typical example of a packed column for gas chromatography. This column is made from stainless steel and is 2 m
long with an internal diameter of 3.2 mm. The packing material in this column has a particle diameter of 149–177 μm. To put this
in perspective, beach sand has a typical diameter of 700 μm and the diameter of fine grained sand is 250 μm.
The most widely used particulate support is diatomaceous earth, which is composed of the silica skeletons of diatoms. These
particles are very porous, with surface areas ranging from 0.5–7.5 m2/g, which provides ample contact between the mobile phase
and the stationary phase. When hydrolyzed, the surface of a diatomaceous earth contains silanol groups (–SiOH), that serve as
active sites for absorbing solute molecules in gas-solid chromatography (GSC).
In gas-liquid chromatography (GLC), we coat the packing material with a liquid mobile phase. To prevent uncoated packing
material from adsorbing solutes, which degrades the quality of the separation, surface silanols are deactivated by reacting them
with dimethyldichlorosilane and rinsing with an alcohol—typically methanol—before coating the particles with stationary phase.

Figure 27.3.2 , for example, has approximately 1800 plates/m, or a total of approximately 3600 theoretical plates.

Capillary Columns
A capillary, or open tubular column is constructed from fused silica and is coated with a protective polymer coating. Columns
range from 15–100 m in length with an internal diameter of approximately 150–300 μm. Figure 27.3.2 shows an example of a
typical capillary column.

Figure 27.3.2 . Typical example of a capillary column for gas chromatography. This column is 30 m long with an internal diameter
of 247 μm. The interior surface of the capillary has a 0.25 μm coating of the liquid phase.

27.3.1 https://chem.libretexts.org/@go/page/350900
Capillary columns are of three principal types. In a wall-coated open tubular column (WCOT) a thin layer of stationary phase,
typically 0.25 nm thick, is coated on the capillary’s inner wall. In a porous-layer open tubular column (PLOT), a porous solid
support—alumina, silica gel, and molecular sieves are typical examples—is attached to the capillary’s inner wall. A support-coated
open tubular column (SCOT) is a PLOT column that includes a liquid stationary phase. Figure 27.3.3 shows the differences
between these types of capillary columns.

Figure 27.3.3 . Cross sections through the three types of capillary columns.
A capillary column provides a significant improvement in separation efficiency because it has more theoretical plates per meter and
is longer than a packed column. For example, the capillary column in Figure 27.3.2 has almost 4300 plates/m, or a total of 129 000
theoretical plates. On the other hand, a packed column can handle a larger sample. Because of its smaller diameter, a capillary
column requires a smaller sample, typically less than 10–2 μL.

Stationary Phases for Gas-Liquid Chromatography

Elution order in gas–liquid chromatography depends on two factors: the boiling point of the solutes, and the interaction between the
solutes and the stationary phase. If a mixture’s components have significantly different boiling points, then the choice of stationary
phase is less critical. If two solutes have similar boiling points, then a separation is possible only if the stationary phase selectively
interacts with one of the solutes. As a general rule, nonpolar solutes are separated more easily when using a nonpolar stationary
phase, and polar solutes are easier to separate when using a polar stationary phase.
There are several important criteria for choosing a stationary phase: it must not react with the solutes, it must be thermally stable, it
must have a low volatility, and it must have a polarity that is appropriate for the sample’s components. Table 27.3.1 summarizes the
properties of several popular stationary phases.
Table 27.3.1 . Selected Examples of Stationary Phases for Gas-Liquid Chromatography
stationary phase polarity trade name temperature limit (oC) representative applications

low-boiling aliphatics
squalane nonpolar Squalane 150
hydrocarbons
amides, fatty acid methyl
Apezion L nonpolar Apezion L 300
esters, terpenoids
alkaloids, amino acid
polydimethyl siloxane slightly polar SE-30 300–350 derivatives, drugs,
pesticides, phenols, steroids
alkaloids, drugs, pesticides,
phenylmethyl polysiloxane polyaromatic
moderately polar OV-17 375
(50% phenyl, 50% methyl) hydrocarbons,
polychlorinated biphenyls

trifluoropropylmethyl alkaloids, amino acid

polysiloxane derivatives, drugs,
moderately polar OV-210 275
(50% trifluoropropyl, 50% halogenated compounds,
methyl) ketones

cyanopropylphenylmethyl
polysiloxane
polar OV-225 275 nitriles, pesticides, steroids
(50%cyanopropyl, 50%
phenylmethyl)

27.3.2 https://chem.libretexts.org/@go/page/350900
 stationary phase polarity trade name temperature limit (oC) representative applications

aldehydes, esters, ethers,

polyethylene glycol polar Carbowax 20M 225
phenols

Many stationary phases have the general structure shown in Figure 27.3.4 a. A stationary phase of polydimethyl siloxane, in which
all the –R groups are methyl groups, –CH3, is nonpolar and often makes a good first choice for a new separation. The order of
elution when using polydimethyl siloxane usually follows the boiling points of the solutes, with lower boiling solutes eluting first.
Replacing some of the methyl groups with other substituents increases the stationary phase’s polarity and provides greater
selectivity. For example, replacing 50% of the –CH3 groups with phenyl groups, –C6H5, produces a slightly polar stationary phase.
Increasing polarity is provided by substituting trifluoropropyl, –C3H6CF, and cyanopropyl, –C3H6CN, functional groups, or by
using a stationary phase of polyethylene glycol (Figure 27.3.4 b).

Figure 27.3.4 . General structures of common stationary phases: (a) substituted polysiloxane; (b) polyethylene glycol.
An important problem with all liquid stationary phases is their tendency to elute, or bleed from the column when it is heated. The
temperature limits in Table 27.3.1 minimize this loss of stationary phase. Capillary columns with bonded or cross-linked stationary
phases provide superior stability. A bonded stationary phase is attached chemically to the capillary’s silica surface. Cross-linking,
which is done after the stationary phase is in the capillary column, links together separate polymer chains to provide greater
stability.
Another important consideration is the thickness of the stationary phase with thinner films of stationary phase improving separation
efficiency, as we learned in Chapter 26.4. The most common thickness is 0.25 μm, although a thicker films is useful for highly
volatile solutes, such as gases, because it has a greater capacity for retaining such solutes. Thinner films are used when separating
low volatility solutes, such as steroids.
A few stationary phases take advantage of chemical selectivity. The most notable are stationary phases that contain chiral
functional groups, which are used to separate enantiomers [Hinshaw, J. V. LC .GC 1993, 11, 644–648].

This page titled 27.3: Gas Chromatographic Columns and Stationary Phases is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by David Harvey.

27.3.3 https://chem.libretexts.org/@go/page/350900
27.4: Applications of Gas Chromatography
Quantitative Applications
Gas chromatography is widely used for the analysis of a diverse array of samples in environmental, clinical, pharmaceutical,
biochemical, forensic, food science and petrochemical laboratories. Table 27.4.1 provides some representative examples of
applications.
Table 27.4.1 . Representative Applications of Gas Chromatography
area applications

green house gases (CO2, CH4, NOx) in air

pesticides in water, wastewater, and soil
environmental analysis
vehicle emissions
trihalomethanes in drinking water
drugs
clinical analysis
blood alcohols
analysis of arson accelerants
forensic analysis
detection of explosives
volatile organics in spices and fragrances
consumer products trace organics in whiskey
monomers in latex paint
purity of solvents
petrochemical and chemical industry refinery gas
composition of gasoline

Quantitative Calculations
In a GC analysis the area under the peak is proportional to the amount of analyte injected onto the column. A peak’s area is
determined by integration, which usually is handled by the instrument’s computer or by an electronic integrating recorder. If two
peak are resolved fully, the determination of their respective areas is straightforward.

Before electronic integrating recorders and computers, two methods were used to find the area under a curve. One method
used a manual planimeter; as you use the planimeter to trace an object’s perimeter, it records the area. A second approach for
finding a peak’s area is the cut-and-weigh method. The chromatogram is recorded on a piece of paper and each peak of interest
is cut out and weighed. Assuming the paper is uniform in thickness and density of fibers, the ratio of weights for two peaks is
the same as the ratio of areas. Of course, this approach destroys your chromatogram.

Overlapping peaks, however, require a choice between one of several options for dividing up the area shared by the two peaks
(Figure 27.4.1 ). Which method we use depends on the relative size of the two peaks and their resolution. In some cases, the use of
peak heights provides more accurate results [(a) Bicking, M. K. L. Chromatography Online, April 2006; (b) Bicking, M. K. L.
Chromatography Online, June 2006].

27.4.1 https://chem.libretexts.org/@go/page/350901
 Figure 27.4.1 . Four methods for determining the areas under two overlapping chromatographic peaks: (a) the drop method; (b) the
valley method; (c) the exponential skim method; and (d) the Gaussian skim method. Other methods for determining areas also are
available.
For quantitative work we need to establish a calibration curve that relates the detector’s response to the analyte’s concentration. If
the injection volume is identical for every standard and sample, then an external standardization provides both accurate and precise
results. Unfortunately,even under the best conditions the relative precision for replicate injections may differ by 5%; often it is
substantially worse. For quantitative work that requires high accuracy and precision, the use of internal standards is recommended.

 Example 27.4.1

Marriott and Carpenter report the following data for five replicate injections of a mixture that contains 1% v/v methyl isobutyl
ketone and 1% v/v p-xylene in dichloromethane [Marriott, P. J.; Carpenter, P. D. J. Chem. Educ. 1996, 73, 96–99].

injection peak peak area (arb. units)

I 1 48075

2 78112

II 1 85829

2 135404

III 1 84136

2 132332

IV 1 71681

2 112889

V 1 58054

2 91287

Assume that p-xylene (peak 2) is the analyte, and that methyl isobutyl ketone (peak 1) is the internal standard. Determine the
95% confidence interval for a single-point standardization with and without using the internal standard.

Solution
For a single-point external standardization we ignore the internal standard and determine the relationship between the peak
area for p-xylene, A2, and the concentration, C2, of p-xylene.

A2 = kC2

Substituting the known concentration for p-xylene (1% v/v) and the appropriate peak areas, gives the following values for the
constant k.

27.4.2 https://chem.libretexts.org/@go/page/350901
 78112 135404 132332 112889 91287

The average value for k is 110 000 with a standard deviation of 25 100 (a relative standard deviation of 22.8%). The 95%
confidence interval is

¯¯¯
¯
ts (2.78)(25100)
μ =X ± = 111000 ± = 111000 ± 31200
− –
√n √5

For an internal standardization, the relationship between the analyte’s peak area, A2, the internal standard’s peak area, A1, and
their respective concentrations, C2 and C1, is
A2 C2
=k
A1 C1

Substituting in the known concentrations and the appropriate peak areas gives the following values for the constant k.

1.5917 1.5776 1.5728 1.5749 1.5724

The average value for k is 1.5779 with a standard deviation of 0.0080 (a relative standard deviation of 0.507%). The 95%
confidence interval is

¯¯¯
¯
ts (2.78)(0.0080)
μ =X ± = 1.5779 ± = 1.5779 ± 0.0099
− –
√n √5

Although there is a substantial variation in the individual peak areas for this set of replicate injections, the internal standard
compensates for these variations, providing a more accurate and precise calibration.

 Exercise 27.4.1

Figure 27.4.2 shows chromatograms for five standards and for one sample. Each standard and sample contains the same
concentration of an internal standard, which is 2.50 mg/mL. For the five standards, the concentrations of analyte are 0.20
mg/mL, 0.40 mg/mL, 0.60 mg/mL, 0.80 mg/mL, and 1.00 mg/mL, respectively. Determine the concentration of analyte in the
sample by (a) ignoring the internal standards and creating an external standards calibration curve, and by (b) creating an
internal standard calibration curve. For each approach, report the analyte’s concentration and the 95% confidence interval. Use
peak heights instead of peak areas.

Answer
The following table summarizes my measurements of the peak heights for each standard and the sample, and their ratio
(although your absolute values for peak heights will differ from mine, depending on the size of your monitor or printout,
your relative peak height ratios should be similar to mine).

[standard] (mg/mL) peak height of standard (mm) peak height of analyte (mm) peak height ratio

0.20 35 7 0.20

0.40 41 16 0.39

0.60 44 27 0.61

0.80 48 39 0.81

1.00 41 41 1.00

sample 39 21 0.54

Figure (a) shows the calibration curve and the calibration equation when we ignore the internal standard. Substituting the
sample’s peak height into the calibration equation gives the analyte’s concentration in the sample as 0.49 mg/mL. The 95%
confidence interval is ±0.24 mg/mL. The calibration curve shows quite a bit of scatter in the data because of uncertainty in
the injection volumes.
Figure (b) shows the calibration curve and the calibration equation when we include the internal standard. Substituting the
sample’s peak height ratio into the calibration equation gives the analyte’s concentration in the sample as 0.54 mg/mL. The

27.4.3 https://chem.libretexts.org/@go/page/350901
 95% confidence interval is ±0.04 mg/mL.

The data for this exercise were created so that the analyte’s actual concentration is 0.55 mg/mL. Given the resolution of my
ruler’s scale, my answer is pretty reasonable. Your measurements may be slightly different, but your answers should be
close to the actual values.

Figure 27.4.2 . Chromatograms for Exercise 27.4.1.

Qualitative Applications
In addition to a quantitative analysis, we also can use chromatography to identify the components of a mixture. As noted earlier,
when using an FT–IR or a mass spectrometer as the detector we have access to the eluent’s full spectrum for any retention time. By
interpreting the spectrum or by searching against a library of spectra, we can identify the analyte responsible for each
chromatographic peak.

In addition to identifying the component responsible for a particular chromatographic peak, we also can use the saved spectra
to evaluate peak purity. If only one component is responsible for a chromatographic peak, then the spectra should be identical
throughout the peak’s elution. If a spectrum at the beginning of the peak’s elution is different from a spectrum taken near the
end of the peak’s elution, then at least two components are co-eluting.

When using a nonspectroscopic detector, such as a flame ionization detector, we must find another approach if we wish to identify
the components of a mixture. One approach is to spike a sample with the suspected compound and look for an increase in peak
height. We also can compare a peak’s retention time to the retention time for a known compound if we use identical operating
conditions.
Because a compound’s retention times on two identical columns are not likely to be the same—differences in packing efficiency,
for example, will affect a solute’s retention time on a packed column—creating a table of standard retention times is not possible.
Kovat’s retention index provides one solution to the problem of matching retention times. Under isothermal conditions, the

27.4.4 https://chem.libretexts.org/@go/page/350901
adjusted retention times for normal alkanes increase logarithmically. Kovat defined the retention index, I, for a normal alkane as
100 times the number of carbon atoms. For example, the retention index is 400 for butane, C4H10, and 500 for pentane, C5H12. To
determine the a compound’s retention index, Icpd, we use the following formula
′ ′
log t − log tr,x
r,cpd
Icpd = 100 × + Ix (27.4.1)
′ ′
log t − log tr,x
r,x+1

where t′
r,cpd
is the compound’s adjusted retention time, t and t
′
r,x are the adjusted retention times for the normal alkanes that
′
r,x+1

elute immediately before the compound and immediately after the compound, respectively, and Ix is the retention index for the
normal alkane that elutes immediately before the compound. A compound’s retention index for a particular set of chromatographic
conditions—stationary phase, mobile phase, column type, column length, temperature, etc.—is reasonably consistent from day- to-
day and between different columns and instruments.

Tables of Kovat’s retention indices are available; see, for example, the NIST Chemistry Webbook. A search for toluene returns
341 values of I for over 20 different stationary phases, and for both packed columns and capillary columns.

 Example 27.4.2

In a separation of a mixture of hydrocarbons the following adjusted retention times are measured: 2.23 min for propane, 5.71
min for isobutane, and 6.67 min for butane. What is the Kovat’s retention index for each of these hydrocarbons?

Solution
Kovat’s retention index for a normal alkane is 100 times the number of carbons; thus, for propane, I = 300 and for butane, I =
400. To find Kovat’s retention index for isobutane we use Equation 27.4.1.
log(5.71) − log(2.23)
Iisobutane = 100 × + 300 = 386
log(6.67) − log(2.23)

 Exercise 27.4.2

When using a column with the same stationary phase as in Example 27.4.2 , you find that the retention times for propane and
butane are 4.78 min and 6.86 min, respectively. What is the expected retention time for isobutane?

Answer
Because we are using the same column we can assume that isobutane’s retention index of 386 remains unchanged. Using
Equation 27.4.1, we have
log x − log(4.78)
386 = 100 × + 300
log(6.86) − log(4.78)

where x is the retention time for isobutane. Solving for x, we find that
log x − log(4.78)
0.86 =
log(6.86) − log(4.78)

0.135 = log x − 0.679

0.814 = log x

x = 6.52

the retention time for isobutane is 6.5 min.

27.4.5 https://chem.libretexts.org/@go/page/350901
This page titled 27.4: Applications of Gas Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

27.4.6 https://chem.libretexts.org/@go/page/350901
 CHAPTER OVERVIEW
28: High-Performance Liquid Chromatography
In high-performance liquid chromatography (HPLC) we inject the sample, which is in solution form, into a liquid mobile phase.
The mobile phase carries the sample through a packed or capillary column that separates the sample’s components based on their
ability to partition between the mobile phase and the stationary phase. Because it combines separation with analysis, HPLC
provides excellent selectivity. By adjusting conditions it usually is possible to design a separation so that analytes elute by
themselves, even when the mixture is complex. Additional selectivity is possible by using a detector that does not respond to all
analytes.
28.1: Scope of HPLC
28.2: Column Efficiency in Liquid Chromatography
28.3: Instruments for Liquid Chromatography
28.4: Partition Chromatography
28.5: Adsorption Chromatography
28.6: Ion-Exchange Chromatography
28.7: Size-Exclusion Chromatography

This page titled 28: High-Performance Liquid Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

1
28.1: Scope of HPLC
Gas chromatography consists largely of two specific types of interactions, both of which involve the stationary phase: the
partitioning of the solute into a polar or a non-polar stationary phase, or the adsorption of the solute onto a solid packing material.
The separation of a complex mixture into its component parts is determined primarily by the boiling points of the solutes and
differences in the solubility of the solutes in the stationary phase. The properties of the mobile phase, on the other hand, are less
important. It is not surprising that there is not much variety in the basic types of gas chromatography.
High-performance liquid chromatography consists of much richer group of techniques, both because the separation depends on the
ability of the solutes to partition into the stationary phase and to partition into the mobile phase. The range of the possible types of
interactions between the solutes and the stationary phase also is greater in HPLC than in GC. In addition to separations based on
differences in solubility of the solutes in the stationary phase and the mobile phase (normal and reverse phase partition
chromatography) and Separations based on the adsorption of solutes on a solid substrate (adsorption chromatography), the
separation of ions is possible using ion-exchange resins as stationary phases (ion-exchange chromatography) and the separation of
ions by size (size-exclusion chromatography).

This page titled 28.1: Scope of HPLC is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

28.1.1 https://chem.libretexts.org/@go/page/362581
28.2: Column Efficiency in Liquid Chromatography
In Chapter 26 we considered three factors that affect band broadening—multiple paths, longitudinal diffusion, and mass transfer—
expressing the relationship between the height of a theoretical plate, H , as a function of the mobile phase's velocity, u, using the
van Deemter equation
B
H = A+ + Cu
u

where A is the contribution from multiple paths, B is the contribution from longitudinal diffusion, and C is the contribution from
mass transfer. Unlike gas chromatography, where there is little distance between the point of injection and the column, and little
distance between the column and the detector, an HPLC instrument must often include additional tubing to connect together the
sample injection port and the column, and the column and the detector. Solutes moving through this tubing, which does not include
stationary phase, travel with a velocity that is slower at the walls of the tubing and faster at the center of the tubing; the result is
additional band broadening. The magnitude of this contribution to band broadening is minimized by keeping the length of
connecting tubing as short as possible. by using tubing with a smaller internal diameter, and by using lower flow rates.

This page titled 28.2: Column Efficiency in Liquid Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

28.2.1 https://chem.libretexts.org/@go/page/362583
28.3: Instruments for Liquid Chromatography
In high-performance liquid chromatography (HPLC) we inject the sample, which is in solution form, into a liquid mobile phase.
The mobile phase carries the sample through a packed or capillary column that separates the sample’s components based on their
ability to partition between the mobile phase and the stationary phase. Figure 28.3.1 shows an example of a typical HPLC
instrument, which has several key components: reservoirs that store the mobile phase; a pump for pushing the mobile phase
through the system; an injector for introducing the sample; a column for separating the sample into its component parts; and a
detector for monitoring the eluent as it comes off the column. Let’s consider each of these components.

Figure 28.3.1 . Example of a typical high-performance liquid chromatograph with insets showing the pumps that move the mobile
phase through the system and the plumbing used to inject the sample into the mobile phase. This particular instrument includes an
autosampler. An instrument in which samples are injected manually does not include the features shown in the two left-most insets,
and has a different style of loop injection valve.

HPLC Columns
An HPLC typically includes two columns: an analytical column, which is responsible for the separation, and a guard column that is
placed before the analytical column to protect it from contamination.

Analytical Columns
The most common type of HPLC column is a stainless steel tube with an internal diameter between 2.1 mm and 4.6 mm and a
length between 30 mm and 300 mm (Figure 28.3.2 ). The column is packed with 3–10 µm porous silica particles with either an
irregular or a spherical shape. Typical column efficiencies are 40 000–60 000 theoretical plates/m. A 25-cm column with 50 000
plates/m has 12 500 theoretical plates.

Figure 28.3.2 . Typical packed column for HPLC. This particular column has an internal diameter of 4.6 mm and a length of 150
mm, and is packed with 5 μm particles coated with stationary phase.
Capillary columns use less solvent and, because the sample is diluted to a lesser extent, produce larger signals at the detector. These
columns are made from fused silica capillaries with internal diameters from 44–200 μm and lengths of 50–250 mm. Capillary
columns packed with 3–5 μm particles have been prepared with column efficiencies of up to 250 000 theoretical plates [Novotony,
M. Science, 1989, 246, 51–57].
One limitation to a packed capillary column is the back pressure that develops when pumping the mobile phase through the small
interstitial spaces between the particulate micron-sized packing material (Figure 28.3.3 ). Because the tubing and the fittings that
carry the mobile phase have pressure limits, a higher back pressure requires a lower flow rate and a longer analysis time.
Monolithic columns, in which the solid support is a single, porous rod, offer column efficiencies equivalent to a packed capillary
column while allowing for faster flow rates. A monolithic column—which usually is similar in size to a conventional packed
column, although smaller, capillary columns also are available—is prepared by forming the mono- lithic rod in a mold and
covering it with PTFE tubing or a polymer resin. Monolithic rods made of a silica-gel polymer typically have macropores with

28.3.1 https://chem.libretexts.org/@go/page/362585
diameters of approximately 2 μm and mesopores—pores within the macropores—with diameters of approximately 13 nm [Cabrera,
K. Chromatography Online, April 1, 2008].

Figure 28.3.3 . The packing of smaller particles creates smaller interstitial spaces than the packing of larger particles. Although
reducing particle size by 2× increases efficiency by a factor of 1.4, it also produces a 4-fold increase in back pressure.

Guard Columns
Two problems tend to shorten the lifetime of an analytical column. First, solutes that bind irreversibly to the stationary phase
degrade the column’s performance by decreasing the amount of stationary phase available for effecting a separation. Second,
particulate material injected with the sample may clog the analytical column. To minimize these problems we place a guard column
before the analytical column. A Guard column usually contains the same particulate packing material and stationary phase as the
analytical column, but is significantly shorter and less expensive—a length of 7.5 mm and a cost one-tenth of that for the
corresponding analytical column is typical. Because they are intended to be sacrificial, guard columns are replaced regularly. If you
look closely at Figure 28.3.1 , you will see the small guard column just above the analytical column.

HPLC Plumbing
In a gas chromatograph the pressure from a compressed gas cylinder is sufficient to push the mobile phase through the column.
Pushing a liquid mobile phase through a column, however, takes a great deal more effort, generating pressures in excess of several
hundred atmospheres. In this section we consider the basic plumbing needed to move the mobile phase through the column and to
inject the sample into the mobile phase.

Moving the Mobile Phase

A typical HPLC includes between 1–4 reservoirs for storing mobile phase solvents. The instrument in Figure 28.3.1 , for example,
has two mobile phase reservoirs that are used for an isocratic elution or a gradient elution by drawing solvents from one or both
reservoirs.
Before using a mobile phase solvent we must remove dissolved gases, such as N2 and O2, and small particulate matter, such as
dust. Because there is a large drop in pressure across the column—the pressure at the column’s entrance is as much as several
hundred atmospheres, but it is atmospheric pressure at the column’s exit—gases dissolved in the mobile phase are released as gas
bubbles that may interfere with the detector’s response. Degassing is accomplished in several ways, but the most common are the
use of a vacuum pump or sparging with an inert gas, such as He, which has a low solubility in the mobile phase. Particulate
materials, which may clog the HPLC tubing or column, are removed by filtering the solvents.

Bubbling an inert gas through the mobile phase releases volatile dissolved gases. This process is called sparging.

The mobile phase solvents are pulled from their reservoirs by the action of one or more pumps. Figure 28.3.4 shows a close-up
view of the pumps for the instrument in Figure 28.3.1 . The working pump and the equilibrating pump each have a piston whose
back and forth movement maintains a constant flow rate of up to several mL/min and provides the high output pressure needed to
push the mobile phase through the chromatographic column. In this particular instrument, each pump sends its mobile phase to a
mixing chamber where they combine to form the final mobile phase. The relative speed of the two pumps determines the mobile
phase’s final composition.

28.3.2 https://chem.libretexts.org/@go/page/362585
 Figure 28.3.4 . Close-up view of the pumps for the instrument shown in Figure 28.3.1 . The working cylinder and the equilibrating
cylinder for the pump on the left take solvent from reservoir A and send it to the mixing chamber. The pump on the right moves
solvent from reservoir B to the mixing chamber. The mobile phase’s flow rate is determined by the combined speeds of the two
pumps. By changing the relative speeds of the two pumps, different binary mobile phases can be prepared.
The back and forth movement of a reciprocating pump creates a pulsed flow that contributes noise to the chromatogram. To
minimize these pulses, each pump in Figure 28.3.4 has two cylinders. During the working cylinder’s forward stoke it fills the
equilibrating cylinder and establishes flow through the column. When the working cylinder is on its reverse stroke, the flow is
maintained by the piston in the equilibrating cylinder. The result is a pulse-free flow.

There are other ways to control the mobile phase’s composition and flow rate. For example, instead of the two pumps in
Figure 28.3.4 , we can place a solvent proportioning valve before a single pump. The solvent proportioning value connects two
or more solvent reservoirs to the pump and determines how much of each solvent is pulled during each of the pump’s cycles.
Another approach for eliminating a pulsed flow is to include a pulse damper between the pump and the column. A pulse
damper is a chamber filled with an easily compressed fluid and a flexible diaphragm. During the piston’s forward stroke the
fluid in the pulse damper is compressed. When the piston withdraws to refill the pump, pressure from the expanding fluid in
the pulse damper maintains the flow rate.

Injecting the Sample

The operating pressure within an HPLC is sufficiently high that we cannot inject the sample into the mobile phase by inserting a
syringe through a septum, as is possible in gas chromatography. Instead, we inject the sample using a loop injector, a diagram of
which is shown in Figure 28.3.5 . In the load position a sample loop—which is available in a variety of sizes ranging from 0.5 μL
to 5 mL—is isolated from the mobile phase and open to the atmosphere. The sample loop is filled using a syringe with a capacity
several times that of the sample loop, with excess sample exiting through the waste line. After loading the sample, the injector is
turned to the inject position, which redirects the mobile phase through the sample loop and onto the column.

Figure 28.3.5 . Schematic diagram of a manual loop injector. In the load position the flow of mobile phase from the pump to the
column (shown in green) is isolated from the sample loop, which is filled using a syringe (shown in blue). Rotating the inner valve
(shown in red) to the inject position directs the mobile phase through the sample loop and onto the column.

The instrument in Figure 28.3.1 uses an autosampler to inject samples. Instead of using a syringe to push the sample into the
sample loop, the syringe draws sample into the sample loop.

28.3.3 https://chem.libretexts.org/@go/page/362585
Detectors for HPLC
Many different types of detectors have been use to monitor HPLC separations, most of which use spectroscopy or electrochemistry
to generate a measurable signal.

Spectroscopic Detectors
The most popular HPLC detectors take advantage of an analyte’s UV/Vis absorption spectrum. These detectors range from simple
designs, in which the analytical wavelength is selected using appropriate filters, to a modified spectrophotometer in which the
sample compartment includes a flow cell. Figure 28.3.6 shows the design of a typical flow cell when using a diode array
spectrometer as the detector. The flow cell has a volume of 1–10 μL and a path length of 0.2–1 cm.

Figure 28.3.6 . Schematic diagram of a flow cell for a detector equipped with a diode array spectrometer.
When using a UV/Vis detector the resulting chromatogram is a plot of absorbance as a function of elution time. If the detector is a
diode array spectrometer, then we also can display the result as a three-dimensional chromatogram that shows absorbance as a
function of wavelength and elution time. One limitation to using absorbance is that the mobile phase cannot absorb at the
wavelengths we wish to monitor. Absorbance detectors provide detection limits of as little as 100 pg–1 ng of injected analyte. If an
analyte is fluorescent, we can place the flow cell in a spectrofluorimeter. Detection limits are as little as 1–10 pg of injected analyte.

Electrochemical Detectors
Another common group of HPLC detectors are those based on electrochemical measurements such as amperometry, voltammetry,
coulometry, and conductivity. Figure 28.3.7 , for example, shows an amperometric flow cell. Effluent from the column passes over
the working electrode—held at a constant potential relative to a downstream reference electrode—that completely oxidizes or
reduces the analytes. The current flowing between the working electrode and the auxiliary electrode serves as the analytical signal.
Detection limits for amperometric electrochemical detection are from 10 pg–1 ng of injected analyte.

Figure 28.3.7 . Schematic diagram showing a flow cell for an amperometric electrochemical detector.

Other Detectors
Several other detectors have been used in HPLC. Measuring a change in the mobile phase’s refractive index is analogous to
monitoring the mobile phase’s thermal conductivity in gas chromatography. A refractive index detector is nearly universal,
responding to almost all compounds, but has a relatively poor detection limit of 0.1–1 μg of injected analyte. An additional
limitation of a refractive index detector is that it cannot be used for a gradient elution unless the mobile phase components have
identical refractive indexes.

28.3.4 https://chem.libretexts.org/@go/page/362585
Another useful detector is a mass spectrometer. Figure 28.3.8 shows a block diagram of a typical HPLC–MS instrument. The
effluent from the column enters the mass spectrometer’s ion source using an interface the removes most of the mobile phase, an
essential need because of the incompatibility between the liquid mobile phase and the mass spectrometer’s high vacuum
environment. In the ionization chamber the remaining molecules—a mixture of the mobile phase components and solutes—
undergo ionization and fragmentation. The mass spectrometer’s mass analyzer separates the ions by their mass-to-charge ratio
(m/z). A detector counts the ions and displays the mass spectrum.

Figure 28.3.8 . Block diagram of an HPLC–MS. A three component mixture enters the HPLC. When component A elutes from the
column, it enters the MS ion source and ionizes to form the parent ion and several fragment ions. The ions enter the mass analyzer,
which separates them by their mass-to-charge ratio, providing the mass spectrum shown at the detector.
There are several options for monitoring the chromatogram when using a mass spectrometer as the detector. The most common
method is to continuously scan the entire mass spectrum and report the total signal for all ions reaching the detector during each
scan. This total ion scan provides universal detection for all analytes. We can achieve some degree of selectivity by monitoring
only specific mass-to-charge ratios, a process called selective-ion monitoring. The advantages of using a mass spectrometer in
HPLC are the same as for gas chromatography. Detection limits are very good, typically 0.1–1 ng of injected analyte, with values
as low as 1–10 pg for some samples. In addition, a mass spectrometer provides qualitative, structural information that can help to
identify the analytes. The interface between the HPLC and the mass spectrometer is technically more difficult than that in a GC–
MS because of the incompatibility of a liquid mobile phase with the mass spectrometer’s high vacuum requirement.

This page titled 28.3: Instruments for Liquid Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

28.3.5 https://chem.libretexts.org/@go/page/362585
28.4: Partition Chromatography
Of the many forms of liquid chromatography, partition chromatography is the most common. In partition chromatography, a
solute's retention time is determined by the extent to which it moves from the mobile phase into the stationary phase, and from the
stationary phase back into the mobile phase. The extent of this equilibrium partitioning is determined by the polarity of the solutes,
the stationary phase, and the mobile phase. In normal-phase partition chromatography, the stationary phase is polar and the mobile
phase is non-polar (or of low polarity), with more polar solutes taking longer to elute as they are more strongly retained by the
polar stationary phase. In reverse-phase partition chromatography, the stationary phase is non-polar and the mobile phase is polar,
with more polar solutes eluting more quickly as they are less strongly retained by the stationary phase. Of the two modes, reverse-
phase partition chromatography is the more common.

Stationary Phases for Partition Chromatography

In partition chromatography the stationary phase is a liquid film coated on a packing material, typically 3–10 μm porous silica
particles. Because the stationary phase may be partially soluble in the mobile phase, it may elute, or bleed from the column over
time. To prevent the loss of stationary phase, which shortens the column’s lifetime, it is bound covalently to the silica particles.
Bonded stationary phases are created by reacting the silica particles with an organochlorosilane of the general form Si(CH3)2RCl,
where R is an alkyl or substituted alkyl group.

To prevent unwanted interactions between the solutes and any remaining –SiOH groups, Si(CH3)3Cl is used to convert unreacted
sites to – SiOSi(CH ) ; such columns are designated as end-capped.
3 3

The properties of a stationary phase depend on the organosilane’s alkyl group. If R is a polar functional group, then the stationary
phase is polar. Examples of polar stationary phases include those where R contains a cyano (–C2H4CN), a diol (–
C3H6OCH2CHOHCH2OH), or an amino (–C3H6NH2) functional group. The most common nonpolar stationary phases use an
organochlorosilane where the R group is an n-octyl (C8) or n-octyldecyl (C18) hydrocarbon chain. Most reversed-phase separations
are carried out using a buffered aqueous solution as a polar mobile phase, or using other polar solvents, such as methanol and
acetonitrile. Because the silica substrate may undergo hydrolysis in basic solutions, the pH of the mobile phase must be less than
7.5.

It seems odd that the more common form of liquid chromatography is identified as reverse-phase instead of normal phase. One
of the earliest examples of chromatography was Mikhail Tswett’s separation of plant pigments, which used a polar column of
calcium carbonate and a nonpolar mobile phase of petroleum ether. The assignment of normal and reversed, therefore, is all
about precedence.

Mobile Phases for Partition Chromatography

The elution order of solutes in HPLC is governed by polarity. For a normal-phase separation, a solute of lower polarity spends
proportionally less time in the polar stationary phase and elutes before a solute that is more polar. Given a particular stationary
phase, retention times in normal-phase HPLC are controlled by adjusting the mobile phase’s properties. For example, if the
resolution between two solutes is poor, switching to a less polar mobile phase keeps the solutes on the column for a longer time and
provides more opportunity for their separation. In reversed-phase HPLC the order of elution is the opposite that in a normal-phase
separation, with more polar solutes eluting first. Increasing the polarity of the mobile phase leads to longer retention times. Shorter
retention times require a mobile phase of lower polarity.

Choosing a Mobile Phase: Using the Polarity Index

There are several indices that help in selecting a mobile phase, one of which is the polarity index [Snyder, L. R.; Glajch, J. L.;
Kirkland, J. J. Practical HPLC Method Development, Wiley-Inter- science: New York, 1988]. Table 28.4.1 provides values of the
polarity index, P , for several common mobile phases, where larger values of P correspond to more polar solvents. Mixing
′ ′

28.4.1 https://chem.libretexts.org/@go/page/362587
together two or more mobile phases—assuming they are miscible—creates a mobile phase of intermediate polarity. For example, a
binary mobile phase made by combining solvent A and solvent B has a polarity index, P , of ′
AB

′ ′ ′
P = ΦA P + ΦB P (28.4.1)
AB A B

where P and P are the polarity indices for solvents A and B, and Φ and Φ are the volume fractions for the two solvents.
′
A
′
B A B

Table 28.4.1 . Properties of HPLC Mobile Phases

mobile phase polarity index (P ) ′
UV cutoff (nm)

cyclohexane 0.04 210

n-hexane 0.1 210

carbon tetrachloride 1.6 265

i-propyl ether 2.4 220

toluene 2.4 286

diethyl ether 2.8 218

tetrahydrofuran 4.0 220

ethanol 4.3 210

ethyl acetate 4.4 255

dioxane 4.8 215

methanol 5.1 210

acetonitrile 5.8 190

water 10.2 —

 Example 28.4.1

A reversed-phase HPLC separation is carried out using a mobile phase of 60% v/v water and 40% v/v methanol. What is the
mobile phase’s polarity index?

Solution
Using Equation 28.4.1 and the values in Table 28.4.1 , the polarity index for a 60:40 water–methanol mixture is
′ ′ ′
P = Φwater Pwater + Φmethanol P
AB methanol

′
P = 0.60 × 10.2 + 0.40 × 5.1 = 8.2
AB

 Exercise 28.4.1

Suppose you need a mobile phase with a polarity index of 7.5. Explain how you can prepare this mobile phase using methanol
and water.

Answer
If we let x be the fraction of water in the mobile phase, then 1 – x is the fraction of methanol. Substituting these values into
Equation 28.4.1and solving for x

7.5 = 10.2x + 5.1(1 − x)

7.5 = 10.2x + 5.1 − 5.1x

2.4 = 5.1x

gives x as 0.47. The mobile phase is 47% v/v water and 53% v/v methanol.

28.4.2 https://chem.libretexts.org/@go/page/362587
As a general rule, a two unit change in the polarity index corresponds to an approximately 10-fold change in a solute’s retention
factor. Here is a simple example. If a solute’s retention factor, k, is 22 when using water as a mobile phase (P = 10.2), then ′

switching to a mobile phase of 60:40 water–methanol (P = 8.2) decreases k to approximately 2.2. Note that the retention factor
′

becomes smaller because we are switching from a more polar mobile phase to a less polar mobile phase in a reversed-phase
separation.

Choosing a Mobile Phase: Adjusting Selectivity

Changing the mobile phase’s polarity index changes a solute’s retention factor. As we learned in Chapter 26.4, however, a change
in k is not an effective way to improve resolution when the initial value of k is greater than 10. To effect a better separation between
two solutes we must improve the selectivity factor, α . There are two common methods for increasing α : adding a reagent to the
mobile phase that reacts with the solutes in a secondary equilibrium reaction or switching to a different mobile phase.
Taking advantage of a secondary equilibrium reaction is a useful strategy for improving a separation [(a) Foley, J. P.
Chromatography, 1987, 7, 118–128; (b) Foley, J. P.; May, W. E. Anal. Chem. 1987, 59, 102–109; (c) Foley, J. P.; May, W. E. Anal.
Chem. 1987, 59, 110–115]. Figure 28.4.1 shows the reversed-phase separation of four weak acids—benzoic acid, terephthalic acid,
p-aminobenzoic acid, and p-hydroxybenzoic acid—on a nonpolar C18 column using an aqueous buffer of acetic acid and sodium
acetate as the mobile phase. The retention times for these weak acids are shorter when using a less acidic mobile phase because
each solute is present in an anionic, weak base form that is less soluble in the nonpolar stationary phase. If the mobile phase’s pH is
sufficiently acidic, the solutes are present as neutral weak acids that are more soluble in the stationary phase and take longer to
elute. Because the weak acid solutes do not have identical pKa values, the pH of the mobile phase has a different effect on each
solute’s retention time, allowing us to find the optimum pH for effecting a complete separation of the four solutes.

Figure 28.4.1 . Example showing how the mobile phase pH in liquid chromatography affects selectivity: (a) retention times for four
substituted benzoic acids as a function of the mobile phase’s pH; (b) alpha values for three pairs of solutes that are difficult to
separate. The mobile phase is an acetic acid/sodium acetate buffer and the stationary phase is a nonpolar hydrocarbon. Data from
Harvey, D. T.; Byerly, S.; Bowman, A.; Tomlin, J. “Optimization of HPLC and GC Separations Using Response Surfaces,” J.
Chem. Educ. 1991, 68, 162–168.
In Example 28.4.1 we learned how to adjust the mobile phase’s polarity by blending together two solvents. A polarity index,
however, is just a guide, and binary mobile phase mixtures with identical polarity indices may not resolve equally a pair of solutes.
Table 28.4.2 , for example, shows retention times for four weak acids in two mobile phases with nearly identical values for P . ′

Although the order of elution is the same for both mobile phases, each solute’s retention time is affected differently by the choice of
organic solvent. If we switch from using acetonitrile to tetrahydrofuran, for example, we find that benzoic acid elutes more quickly
and that p-hydroxybenzoic acid elutes more slowly. Although we can resolve fully these two solutes using mobile phase that is
16% v/v acetonitrile, we cannot resolve them if the mobile phase is 10% tetrahydrofuran.
Table 28.4.2 . Retention Times for Four Weak Acids in Mobile Phases With Similar Polarity Indexes
16% acetonitrile (CH3CN) 10% tetrahydrofuran (THF)
retention time (min) 84% pH 4.11 aqueous buffer (P = 9.5)
′
90% pH 4.11 aqueous buffer (P = 9.6)
′

tr, BA 5.18 4.01

tr, PH 1.67 2.91

tr, PA 1.21 1.05

tr, TP 0.23 0.54

28.4.3 https://chem.libretexts.org/@go/page/362587
 16% acetonitrile (CH3CN) 10% tetrahydrofuran (THF)
retention time (min) 84% pH 4.11 aqueous buffer (P = 9.5)
′
90% pH 4.11 aqueous buffer (P = 9.6)
′

Key: BA is benzoic acid; PH is p-hydroxybenzoic acid; PA is p-aminobenzoic acid; TP is terephthalic acid

Source: Harvey, D. T.; Byerly, S.; Bowman, A.; Tomlin, J. “Optimization of HPLC and GC Separations Using Response Surfaces,” J. Chem.
Educ. 1991, 68, 162–168.

Figure 28.4.2 . Solvent triangle for optimizing a reversed-phase HPLC separation. The three blue circles show mobile phases
consisting of an organic solvent and water. The three red circles are binary mobile phases created by combining equal volumes of
the pure mobile phases. The ternary mobile phase shown by the purple circle contains all three of the pure mobile phases.
One strategy for finding the best mobile phase is to use the solvent triangle shown in Figure 28.4.2 , which allows us to explore a
broad range of mobile phases with only seven experiments. We begin by adjusting the amount of acetonitrile in the mobile phase to
produce the best possible separation within the desired analysis time. Next, we use Table 28.4.3 to estimate the composition of
methanol/H2O and tetrahydrofuran/H2O mobile phases that will produce similar analysis times. Four additional mobile phases are
prepared using the binary and ternary mobile phases shown in Figure 28.4.2 . When we examine the chromatograms from these
seven mobile phases we may find that one or more provides an adequate separation, or we may identify a region within the solvent
triangle where a separation is feasible. Figure 28.4.3 shows a resolution map for the reversed-phase separation of benzoic acid,
terephthalic acid, p-aminobenzoic acid, and p-hydroxybenzoic acid on a nonpolar C18 column in which the maximum desired
analysis time is set to 6 min [Harvey, D. T.; Byerly, S.; Bowman, A.; Tomlin, J. J. Chem. Educ. 1991, 68, 162–168]. The areas in
blue, green, and red show mobile phase compositions that do not provide baseline resolution. The unshaded area represents mobile
phase compositions where a separation is possible.

The choice to start with acetonitrile is arbitrary—we can just as easily choose to begin with methanol or with tetrahydrofuran.

Table 28.4.3 . Composition of Mobile Phases With Approximately Equal Solvent Strengths
%v/v CH3OH % v/v CH3CN %v/v THF

0 0 0

10 6 4

20 14 10

30 22 16

40 32 24

50 40 30

6 50 36

70 60 44

80 72 52

90 87 62

100 99 71

28.4.4 https://chem.libretexts.org/@go/page/362587
 Figure 28.4.3. Resolution map for the separation of benzoic acid (BA), terephthalic acid (TP), p-
aminobenzoic acid (PA), and p-hydroxybenzoic acid (PH) on a nonpolar C18 column subject to a
maximum analysis time of 6 min. The shaded areas represent regions where a separation is not possible,
with the unresolved solutes identified. A separation is possible in the unshaded area. See Harvey, D. T.;
Byerly, S.; Bowman, A.; Tomlin, J. “Optimization of HPLC and GC Separations Using Response
Surfaces,” J. Chem. Educ. 1991, 68, 162–168 for details on the mathematical model used to generate the
resolution map.

Choosing a Mobile Phase: Isocratic and Gradient Elutions

A separation using a mobile phase that has a fixed composition is an isocratic elution. One difficulty with an isocratic elution is that
an appropriate mobile phase strength for resolving early-eluting solutes may lead to unacceptably long retention times for late-
eluting solutes. Optimizing the mobile phase for late-eluting solutes, on the other hand, may provide an inadequate separation of
early-eluting solutes. Changing the mobile phase’s composition as the separation progresses is one solution to this problem. For a
reversed-phase separation we use an initial mobile phase that is more polar. As the separation progresses, we adjust the
composition of mobile phase so that it becomes less polar (see Figure 28.4.4 ). Such separations are called gradient elutions.

Figure 28.4.4 . Gradient elution separation of a mixture of flavonoids. Mobile phase A is an aqueous solution of 0.1% formic acid
and mobile phase B is 0.1% formic acid in acetonitrile. The initial mobile phase is 98% A and 2% B. The percentage of mobile
phase B increases in four steps: from 2% to 5% over 5 min, beginning at 0.5 min; from 5% to 12% over 1 min, beginning at 5.5
min; from 12% to 25% over 15 min, beginning at 6.5 min; and from 25% to 60% over 20 min, beginning at 21.5 min. Data
provided by Christopher Schardon, Kyle Meinhardt, and Michelle Bushey, Department of Chemistry, Trinty University.

Choosing a Detector
The availability of different types detectors provides another way to build selectivity into an analysis. Figure 28.4.50 , for example,
shows the reverse-phase separation of a mixture of flavonoids using UV/Vis detection at two different wavelengths. In this case, a
wavelength of 260 nm increases the method's sensitivity for rutin relative to that for taxifolin.

28.4.5 https://chem.libretexts.org/@go/page/362587
 Figure 28.4.5 . HPLC separation of a mixture of flavonoids with UV/Vis detection at 360 nm and, in the inset, at 260 nm. The
choice of wavelength affects each analyte’s signal. By carefully choosing the wavelength, we can enhance the signal for the
analytes of greatest interest. Data provided by Christopher Schardon, Kyle Meinhardt, and Michelle Bushey, Department of
Chemistry, Trinty University.

As shown in Figure 28.4.6 , a fluorescence detector provides additional selectivity because only a few of a sample’s components
are fluorescent.

Figure 28.4.6 . HPLC chromatogram for the determination of riboflavin in urine using fluorescence detection with exci-tation at a
wavelength of 340 nm and detection at 450 nm. The peak corresponding to riboflavin is marked with a red asterisk (*). The inset
shows the same chromatogram when using a less-selective UV/Vis detector at a wavelength of 450 nm. Data provided by Jason
Schultz, Jonna Berry, Kaelene Lundstrom, and Dwight Stoll, Department of Chemistry, Gustavus Adolphus College.
With a mass spectrometer as a detector, there are several options for monitoring the chromatogram. The most common method is to
continuously scan the entire mass spectrum and report the total signal for all ions reaching the detector during each scan. This total
ion scan provides universal detection for all analytes. As seen in Figure 28.4.7 , we can achieve some degree of selectivity by
monitoring only specific mass-to-charge ratios, a process called selective-ion monitoring.

28.4.6 https://chem.libretexts.org/@go/page/362587
 Figure 28.4.7 . HPLC–MS/MS chromatogram for the determination of riboflavin in urine. An initial parent ion with an m/z ratio of
377 enters a second mass spectrometer where it undergoes additional 20 ionization; the fragment ion with an m/z ratio of 243
provides the signal. The selectivity of this detector is evident when you compare this chromatogram to the one in Figure 28.4.7 ,
which uses fluoresence deterction. Data provided by Jason Schultz, Jonna Berry, Kaelene Lundstrom, and Dwight Stoll,
Department of Chemistry, Gustavus Adolphus College.

Quantitative Applications of Partition Chromatography

Partition chromatography is used routinely for both qualitative and quantitative analyses of environmental, pharmaceutical,
industrial, forensic, clinical, and consumer product samples.

Preparing Samples for Analysis

Samples in liquid form are injected into the HPLC after a suitable clean-up to remove any particulate materials, or after a suitable
extraction to remove matrix interferents. In determining polyaromatic hydrocarbons (PAH) in wastewater, for example, an
extraction with CH2Cl2 serves the dual purpose of concentrating the analytes and isolating them from matrix interferents. Solid
samples are first dissolved in a suitable solvent or the analytes of interest brought into solution by extraction. For example, an
HPLC analysis for the active ingredients and the degradation products in a pharmaceutical tablet often begins by extracting the
powdered tablet with a portion of mobile phase. Gas samples are collected by bubbling them through a trap that contains a suitable
solvent. Organic isocyanates in industrial atmospheres are collected by bubbling the air through a solution of 1-(2-
methoxyphenyl)piperazine in toluene. The reaction between the isocyanates and 1-(2-methoxyphenyl)piperazine both stabilizes
them against degradation before the HPLC analysis and converts them to a chemical form that can be monitored by UV absorption.

Quantitative Calculations
A quantitative HPLC analysis is often easier than a quantitative GC analysis because a fixed volume sample loop provides a more
precise and accurate injection. As a result, most quantitative HPLC methods do not need an internal standard and, instead, use
external standards and a normal calibration curve.

An internal standard is necessary when using HPLC–MS because the interface between the HPLC and the mass spectrometer
does not allow for a reproducible transfer of the column’s eluent into the MS’s ionization chamber.

 Example 28.4.2

The concentration of polynuclear aromatic hydrocarbons (PAH) in soil is determined by first extracting the PAHs with
methylene chloride. The extract is diluted, if necessary, and the PAHs separated by HPLC using a UV/Vis or fluorescence
detector. Calibration is achieved using one or more external standards. In a typical analysis a 2.013-g sample of dried soil is
extracted with 20.00 mL of methylene chloride. After filtering to remove the soil, a 1.00-mL portion of the extract is removed
and diluted to 10.00 mL with acetonitrile. Injecting 5 μL of the diluted extract into an HPLC gives a signal of 0.217 (arbitrary
units) for the PAH fluoranthene. When 5 μL of a 20.0-ppm fluoranthene standard is analyzed using the same conditions, a
signal of 0.258 is measured. Report the parts per million of fluoranthene in the soil.

Solution
For a single-point external standard, the relationship between the signal, S, and the concentration, C, of fluoranthene is

S = kC

28.4.7 https://chem.libretexts.org/@go/page/362587
 Substituting in values for the standard’s signal and concentration gives the value of k as
S 0.258
−1
k = = = 0.0129  ppm
C 20.0 ppm

Using this value for k and the sample’s HPLC signal gives a fluoranthene concentration of
S 0.217
C = = = 16.8 ppm
k 0.0129 ppm −1

for the extracted and diluted soil sample. The concentration of fluoranthene in the soil is
10.00 mL
16.8 g/mL × × 20.00 mL
1.00 mL
= 1670 ppm fluoranthene 
2.013 g sample 

 Exercise 28.4.2

The concentration of caffeine in beverages is determined by a reversed-phase HPLC separation using a mobile phase of 20%
acetonitrile and 80% water, and using a nonpolar C8 column. Results for a series of 10-μL injections of caffeine standards are
in the following table.

[caffeine] (mg/L) peak area (arb. units)

50.0 226724

100.0 453762

125.0 559443

250.0 1093637

What is the concentration of caffeine in a sample if a 10-μL injection gives a peak area of 424195? The data in this problem
comes from Kusch, P.; Knupp, G. “Simultaneous Determination of Caffeine in Cola Drinks and Other Beverages by Reversed-
Phase HPTLC and Reversed-Phase HPLC,” Chem. Educator, 2003, 8, 201–205.

Answer
The figure below shows the calibration curve and calibration equation for the set of external standards. Substituting the
sample’s peak area into the calibration equation gives the concentration of caffeine in the sample as 94.4 mg/L.

This page titled 28.4: Partition Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

28.4.8 https://chem.libretexts.org/@go/page/362587
28.5: Adsorption Chromatography
In adsorption chromatography (or liquid-solid chromatography, LSC) the column packing also serves as the stationary phase. In
Tswett’s original work the stationary phase was finely divided CaCO3, but modern columns employ porous 3–10 μm particles of
silica or alumina. Because the stationary phase is polar, the mobile phase usually is a nonpolar or a moderately polar solvent.
Typical mobile phases include hexane, isooctane, and methylene chloride. The usual order of elution—from shorter to longer
retention times—is
olefins < aromatic hydrocarbons < ethers < esters, aldehydes, ketones < alcohols, amines < amide < carboxylic acids
Nonpolar stationary phases, such as charcoal-based absorbents, also are used. For most samples, liquid–solid chromatography does
not offer any special advantages over liquid–liquid chromatography. One exception is the analysis of isomers, where LSC excels.

This page titled 28.5: Adsorption Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

28.5.1 https://chem.libretexts.org/@go/page/362589
28.6: Ion-Exchange Chromatography
In ion-exchange chromatography (IEC) the stationary phase is a cross-linked polymer resin, usually divinylbenzene cross-linked
polystyrene, with covalently attached ionic functional groups (see Figure 28.6.1 and Table 28.6.1 ). The counterions to these fixed
charges are mobile and are displaced by ions that compete more favorably for the exchange sites. Ion-exchange resins are divided
into four categories: strong acid cation exchangers; weak acid cation exchangers; strong base anion exchangers; and weak base
anion exchangers.

Figure 28.6.1 . Structures of styrene, divinylbenzene, and a styrene–divinylbenzene co-polymer modified for use as an ion-
exchange resin are shown on the left. The ion-exchange sites, indicated by R and shown in blue, are mostly in the para position and
are not necessarily bound to all styrene units. The cross-linking is shown in red. The photo on the right shows an example of the
polymer beads. These beads are approximately 0.30–0.85 mm in diameter. Resins for use in ion-exchange chromatography
typically are 5–11 μm in diameter.
Table 28.6.1 . Examples of Common Ion-Exchange Resins
type functional group examples
−
−SO
strong acid cation exchanger sulfonic acid 3
−
−CH 2 CH 2 SO
3

−
−COO
weak acid cation exchanger carboxylic acid −
−CH 2 COO

+
−CH 2 N(CH )
strong base anion exchanger quaternary amine 3 3
+
−CH 2 CH 2 N(CH CH 3 )
2 3

+
−NH
weak base anion exchanger amine 4
+
−CH 2 CH 2 NH(CH CH 3 )
2 3

Strong acid cation exchangers include a sulfonic acid functional group that retains it anionic form—and thus its capacity for ion-
exchange—in strongly acidic solutions. The functional groups for a weak acid cation exchanger, on the other hand, are fully
protonated at pH levels less then 4 and lose their exchange capacity. The strong base anion exchangers include a quaternary amine,
which retains a positive charge even in strongly basic solutions. Weak base anion exchangers remain protonated only at pH levels
that are moderately basic. Under more basic conditions a weak base anion exchanger loses a proton and its exchange capacity.
The ion-exchange reaction of a monovalent cation, M+, exchange site is
− + + − + +
−SO 3 H (s) + M (aq) ⇌ −SO 3 M (s) + H (aq)

The equilibrium constant for this ion-exchange reaction, which we call the selectivity coefficient, K, is
− + +
{−SO 3 M } [H ]
K = (28.6.1)
− + +
{−SO 3 H } [M ]

where we use curly brackets, { }, to indicate a surface concentration instead of a solution concentration.

28.6.1 https://chem.libretexts.org/@go/page/362591
 We don’t usually think about a solid’s concentration. There is a good reason for this. In most cases, a solid’s concentration is a
constant. If you break a piece of chalk into two parts, for example, the mass and the volume of each piece retains the same
proportional relationship as in the original piece of chalk. When we consider an ion binding to a reactive site on the solid’s
surface, however, the fraction of sites that are bound, and thus the concentration of bound sites, can take on any value between
0 and some maximum value that is proportional to the density of reactive sites.

Rearranging Equation 28.6.1 shows us that the distribution ratio, D, for the exchange reaction
+
 amount of M  in the stationary phase 
D =
+
 amount of M  in the mobile phase 

− + − +
{−SO 3 M } {−SO 3 H }
D = =K× (28.6.2)
+ +
[M ] [H ]

is a function of the concentration of H+ and, therefore, the pH of the mobile phase.

An ion-exchange resin’s selectivity is somewhat dependent on whether it includes strong or weak exchange sites and on the extent
of cross-linking. The latter is particularly important as it controls the resin’s permeability, and, therefore, the accessibility of
exchange sites. An approximate order of selectivity for a typical strong acid cation exchange resin, in order of decreasing D, is
Al3+ > Ba2+ > Pb2+ > Ca2+ > Ni2+ > Cd2+ > Cu2+ > Co2+ > Zn2+ > Mg2+ > Ag+ > K+ > NH > Na+ > H+ > Li+ +
4

Note that highly charged cations bind more strongly than cations of lower charge, and that for cations of similar charge, those with
a smaller hydrated radius, or that are more polarizable, bind more strongly. For a strong base anion exchanger the general elution
order is
SO
2−

4
> I– > HSO > NO > Br– > NO > Cl– > HCO > CH3COO– > OH– > F–
−

4
−

3
−

2
−

Anions of higher charge and of smaller hydrated radius bind more strongly than anions with a lower charge and a larger hydrated
radius.
The mobile phase in IEC usually is an aqueous buffer, the pH and ionic composition of which determines a solute’s retention time.
Gradient elutions are possible in which the mobile phase’s ionic strength or pH is changed with time. For example, an IEC
separation of cations might use a dilute solution of HCl as the mobile phase. Increasing the concentration of HCl speeds the elution
rate for more strongly retained cations because the higher concentration of H+ allows it to compete more successfully for the ion-
exchange sites.

From Equation 28.6.2 , a cation’s distribution ratio, D, becomes smaller when the concentration of H+ in the mobile phase
increases.

An ion-exchange resin is incorporated into an HPLC column either as 5–11 μm porous polymer beads or by coating the resin on
porous silica particles. Columns typically are 250 mm in length with internal diameters ranging from 2–5 mm.
Measuring the conductivity of the mobile phase as it elutes from the column serves as a universal detector for cationic and anionic
analytes. Because the mobile phase contains a high concentration of ions—a mobile phase of dilute HCl, for example, contains
significant concentrations of H+ and Cl– ions—we need a method for detecting the analytes in the presence of a significant
background conductivity.
To minimize the mobile phase’s contribution to conductivity, an ion-suppressor column is placed between the analytical column
and the detector. This column selectively removes mobile phase ions without removing solute ions. For example, in cation-
exchange chromatography using a dilute solution of HCl as the mobile phase, the suppressor column contains a strong base anion-
exchange resin. The exchange reaction
+ − + − + −
H (aq) + Cl (aq) + Resin OH (s) ⇌ Resin Cl (s) + H2 O(l)

replaces the mobile phase ions H+ and Cl– with H2O. A similar process is used in anion-exchange chromatography where the
suppressor column contains a cation-exchange resin. If the mobile phase is a solution of Na2CO3, the exchange reaction

28.6.2 https://chem.libretexts.org/@go/page/362591
 + 2− − + − +
2 Na (aq) + CO3 (aq) + 2 Resin H (s) ⇌ 2 Resin Na (s) + H2 CO3 (aq)

replaces a strong electrolyte, Na2CO3, with a weak electrolyte, H2CO3.

Ion-suppression is necessary when the mobile phase contains a high concentration of ions. Single-column ion chromatography, in
which an ion-suppressor column is not needed, is possible if the concentration of ions in the mobile phase is small. Typically the
stationary phase is a resin with a low capacity for ion-exchange and the mobile phase is a very dilute solution of methane sulfonic
acid for cationic analytes, or potassium benzoate or potassium hydrogen phthalate for anionic analytes. Because the background
conductivity is sufficiently small, it is possible to monitor a change in conductivity as the analytes elute from the column.
A UV/Vis absorbance detector can be used if the analytes absorb ultraviolet or visible radiation. Alternatively, we can detect
indirectly analytes that do not absorb in the UV/Vis if the mobile phase contains a UV/Vis absorbing species. In this case, when a
solute band passes through the detector, a decrease in absorbance is measured at the detector.
Ion-exchange chromatography is an important technique for the analysis of anions and cations in water. For example, an ion-
exchange chromatographic analysis for the anions F–, Cl–, Br–, NO , NO , PO , and SO takes approximately 15 minutes
−
2
−
3
3−
4
2−
4

(Figure 28.6.2 ). A complete analysis of the same set of anions by a combination of potentiometry and spectrophotometry requires
1–2 days. Ion-exchange chromatography also is used for the analysis of proteins, amino acids, sugars, nucleotides, pharmaceuticals,
consumer products, and clinical samples.

Figure 28.6.2 . Ion-exchange chromatographic analysis of water from Big Walnut Creek in Putnam County, Indiana. Data provided
by Jeanette Pope, Department of Geosciences, DePauw University.

This page titled 28.6: Ion-Exchange Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

28.6.3 https://chem.libretexts.org/@go/page/362591
28.7: Size-Exclusion Chromatography
We have considered two classes of micron-sized stationary phases in this chapter: silica particles and cross-linked polymer resin
beads. Both materials are porous, with pore sizes ranging from approximately 5–400 nm for silica particles, and from 5 nm to 100
μm for divinylbenzene cross-linked polystyrene resins. In size-exclusion chromatography—which also is known by the terms
molecular-exclusion or gel permeation chromatography—the separation of solutes depends upon their ability to enter into the pores
of the stationary phase. Smaller solutes spend proportionally more time within the pores and take longer to elute from the column.
A stationary phase’s size selectivity extends over a finite range. All solutes significantly smaller than the pores move through the
column’s entire volume and elute simultaneously, with a retention volume, Vr, of
Vr = Vi + Vo (28.7.1)

where Vi is the volume of mobile phase occupying the stationary phase’s pore space and Vo is volume of mobile phase in the
remainder of the column. The largest solute for which Equation 28.7.1 holds is the column’s inclusion limit, or permeation limit.
Those solutes too large to enter the pores elute simultaneously with an retention volume of
Vr = Vo (28.7.2)

Equation 28.7.2 defines the column’s exclusion limit.

For a solute whose size is between the inclusion limit and the exclusion limit, the amount of time it spends in the stationary phase’s
pores is proportional to its size. The retention volume for these solutes is

Vr = DVi + Vo (28.7.3)

where D is the solute’s distribution ratio, which ranges from 0 at the exclusion limit to 1 at the inclusion limit. Equation 28.7.3
assumes that size-exclusion is the only interaction between the solute and the stationary phase that affects the separation. For this
reason, stationary phases using silica particles are deactivated as described earlier, and polymer resins are synthesized without
exchange sites.
Size-exclusion chromatography provides a rapid means for separating larger molecules, including polymers and biomolecules. A
stationary phase for proteins that consists of particles with 30 nm pores has an inclusion limit of 7500 g/mol and an exclusion limit
of 1.2 × 10 g/mol. Mixtures of proteins that span a wider range of molecular weights are separated by joining together in series
6

several columns with different inclusion and exclusion limits.

Another important application of size-exclusion chromatography is the estimation of a solute’s molecular weight (MW).
Calibration curves are prepared using a series of standards of known molecular weight and measuring each standard’s retention
volume. As shown in Figure 28.7.1 , a plot of log(MW) versus Vr is roughly linear between the exclusion limit and the inclusion
limit. Because a solute’s retention volume is influenced by both its size and its shape, a reasonably accurate estimation of molecular
weight is possible only if the standards are chosen carefully to minimize the effect of shape.

Figure 28.7.1 . Calibration curve for the determination of molecular weight by size-exclusion chromatography. The data shown
here are adapted from Rouessac, F.; Rouessac, A. Chemical Analysis: Modern Instrumentation Methods and Techniques, Wiley:
Chichester, England, 2004, p 141.

28.7.1 https://chem.libretexts.org/@go/page/362593
Size-exclusion chromatography is carried out using conventional HPLC instrumentation, replacing the HPLC column with an
appropriate size-exclusion column. A UV/Vis detector is the most common means for obtaining the chromatogram.

This page titled 28.7: Size-Exclusion Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

28.7.2 https://chem.libretexts.org/@go/page/362593
 CHAPTER OVERVIEW
29: Supercritical Fluid Chromatography
Although there are many analytical applications of gas chromatography and liquid chromatography, they can not separate and
analyze all types of samples. Capillary column GC separates complex mixtures with excellent resolution and short analysis times.
Its application is limited, however, to volatile analytes or to analytes made volatile by a suitable derivatization reaction. Liquid
chromatography separates a wider range of solutes than GC, but the most common detectors—UV, fluorescence, and
electrochemical— have poorer detection limits and smaller linear ranges than GC detectors, and are not as universal in their
selectivity. For some applications, supercritical fluids provide an attractive solution to these limitations.
29.1: Properties of Supercritical Fluids
29.2: Supercritical Fluid Chromatography

This page titled 29: Supercritical Fluid Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

1
29.1: Properties of Supercritical Fluids
As shown in Figure 29.1.1 , a supercritical fluid is a species held at a temperature and a pressure that exceeds its critical point.
Under these conditions the species is neither a gas nor a liquid. Instead, it is a supercritical fluid.

Figure 29.1.1 . Phase diagram showing the combinations of temperature and pressure for which a compound is in its solid state, its
liquid state, and its gas state. For pressures and temperatures above the critical point, the compound is a supercritical fluid with
properties intermediate between a gas and a liquid.
Some properties of a supercritical fluid, as shown in Table 29.1.1 , are similar to a gas; other properties, however, are similar to a
liquid. The viscosity of a supercritical fluid, for example, is similar to a gas, which means we can move a supercritical fluid through
a capillary column or a packed column without the need for high pressures. The density of a supercritical fluid, on the other hand,
is much closer to that of a liquid, which explains why supercritical fluids are good solvents.
Table 29.1.1 . Typical Properties of Gases, Liquids, and Supercritical Fluids
phase density (g/cm3) viscosity (g cm-1 s-1) diffusion coefficient (cm2 s-1)

gas ≈ 10
−3
≈ 10
−4
≈ 0.1

supercritical fluid ≈ 0.1 − 1 ≈ 10

−4
− 10
−3
≈ 10
−4
− 10
−3

liquid ≈ 1 ≈ 10
−2
≈ 10
−3

The most commonly used supercritical fluid is CO2. Its low critical temperature of 31.1oC and its low critical pressure of 72.9 atm
are relatively easy to achieve and maintain. Although supercritical CO2 is a good solvent for nonpolar organics, it is less useful for
polar solutes. The addition of an organic modifier, such as methanol, improves the mobile phase’s elution strength. Other common
mobile phases and their critical temperatures and pressures are listed in Table 29.1.2 .
Table 29.1.2 . Critical Points for Selected Supercritical Fluids
compound critical temperature (oC) critical pressure (atm)

carbon dioxide 31.3 72.9

ethane 32.4 48.3

nitrous oxide 36.5 71.4

ammonia 132.3 111.3

diethyl ether 193.6 36.3

isopropanol 235.3 47.0

methanol 240.5 78.9

ethanol 243.4 63.0

water 374.4 226.8

This page titled 29.1: Properties of Supercritical Fluids is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

29.1.1 https://chem.libretexts.org/@go/page/362611
29.2: Supercritical Fluid Chromatography
The instrumentation for supercritical fluid chromatography essentially is the same as that for a standard HPLC. The only important
additions are a heated oven for the column and a pressure restrictor downstream from the column to maintain the critical pressure.
Gradient elutions are accomplished by changing the applied pressure over time. The resulting change in the mobile phase’s density
affects its solvent strength. Detection is accomplished using standard GC detectors or HPLC detectors. Analysis time and
resolution, although not as good as in GC, usually are better than in conventional HPLC. Supercritical fluid chromatography has
many applications in the analysis of polymers, fossil fuels, waxes, drugs, and food products.

This page titled 29.2: Supercritical Fluid Chromatography is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

29.2.1 https://chem.libretexts.org/@go/page/362613
 CHAPTER OVERVIEW
30: Capillary Electrophoresis and Capillary Electrochromatography
Electrophoresis is a class of separation techniques in which we separate analytes by their ability to move through a conductive
medium—usually an aqueous buffer—in response to an applied electric field. In the absence of other effects, cations migrate
toward the electric field’s negatively charged cathode. Cations with larger charge-to-size ratios—which favors ions of greater
charge and of smaller size—migrate at a faster rate than larger cations with smaller charges. Anions migrate toward the positively
charged anode and neutral species do not experience the electrical field and remain stationary. In this chapter we focus on the most
important of these techniques, capillary electrophoresis and a few related variants.
30.1: An Overview of Electrophoresis
30.2: Capillary Electrophoresis
30.3: Applications of Capillary Electrophoresis

This page titled 30: Capillary Electrophoresis and Capillary Electrochromatography is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by David Harvey.

1
30.1: An Overview of Electrophoresis
Electrophoresis is a class of separation techniques in which we separate analytes by their ability to move through a conductive
medium—usually an aqueous buffer—in response to an applied electric field. In the absence of other effects, cations migrate
toward the electric field’s negatively charged cathode. Cations with larger charge-to-size ratios—which favors ions of greater
charge and of smaller size—migrate at a faster rate than larger cations with smaller charges. Anions migrate toward the positively
charged anode and neutral species do not experience the electrical field and remain stationary.

As we will see shortly, under normal conditions even neutral species and anions migrate toward the cathode.

There are several forms of electrophoresis. In slab gel electrophoresis the conducting buffer is retained within a porous gel of
agarose or polyacrylamide. Slabs are formed by pouring the gel between two glass plates separated by spacers. Typical thicknesses
are 0.25–1 mm. Gel electrophoresis is an important technique in biochemistry where it frequently is used to separate DNA
fragments and proteins. Although it is a powerful tool for the qualitative analysis of complex mixtures, it is less useful for
quantitative work.
In capillary electrophoresis the conducting buffer is retained within a capillary tube with an inner diameter that typically is 25–75
μm. The sample is injected into one end of the capillary tube, and as it migrates through the capillary the sample’s components
separate and elute from the column at different times. The resulting electropherogram looks similar to a GC or an HPLC
chromatogram, and provides both qualitative and quantitative information.

This page titled 30.1: An Overview of Electrophoresis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

30.1.1 https://chem.libretexts.org/@go/page/362618
30.2: Capillary Electrophoresis
Theory of Electrophoresis
In capillary electrophoresis we inject the sample into a buffered solution retained within a capillary tube. When an electric field is
applied across the capillary tube, the sample’s components migrate as the result of two types of actions: electrophoretic mobility
and electroosmotic mobility. Electrophoretic mobility is the solute’s response to the applied electrical field in which cations move
toward the negatively charged cathode, anions move toward the positively charged anode, and neutral species remain stationary.
The other contribution to a solute’s migration is electroosmotic flow, which occurs when the buffer moves through the capillary in
response to the applied electrical field. Under normal conditions the buffer moves toward the cathode, sweeping most solutes,
including the anions and neutral species, toward the negatively charged cathode.

Electrophoretic Mobility
The velocity with which a solute moves in response to the applied electric field is called its electrophoretic velocity, νep ; it is
defined as

νep = μep E (30.2.1)

where μ is the solute’s electrophoretic mobility, and E is the magnitude of the applied electrical field. A solute’s electrophoretic
ep

mobility is defined as
q
μep = (30.2.2)
6πηr

where q is the solute’s charge, η is the buffer’s viscosity, and r is the solute’s radius. Using Equation 30.2.1 and Equation 30.2.2
we can make several important conclusions about a solute’s electrophoretic velocity. Electrophoretic mobility and, therefore,
electrophoretic velocity, increases for more highly charged solutes and for solutes of smaller size. Because q is positive for a cation
and negative for an anion, these species migrate in opposite directions. A neutral species, for which q is zero, has an electrophoretic
velocity of zero.

Electroosmotic Mobility
When an electric field is applied to a capillary filled with an aqueous buffer we expect the buffer’s ions to migrate in response to
their electrophoretic mobility. Because the solvent, H2O, is neutral we might reasonably expect it to remain stationary. What we
observe under normal conditions, however, is that the buffer moves toward the cathode. This phenomenon is called the
electroosmotic flow.
Electroosmotic flow occurs because the walls of the capillary tubing carry a charge. The surface of a silica capillary contains large
numbers of silanol groups (–SiOH). At a pH level greater than approximately 2 or 3, the silanol groups ionize to form negatively
charged silanate ions (–SiO–). Cations from the buffer are attracted to the silanate ions. As shown in Figure 30.2.1 , some of these
cations bind tightly to the silanate ions, forming a fixed layer. Because the cations in the fixed layer only partially neutralize the
negative charge on the capillary walls, the solution adjacent to the fixed layer—which is called the diffuse layer—contains more
cations than anions. Together these two layers are known as the double layer. Cations in the diffuse layer migrate toward the
cathode. Because these cations are solvated, the solution also is pulled along, producing the electroosmotic flow.

30.2.1 https://chem.libretexts.org/@go/page/362621
 Figure 30.2.1 . Schematic diagram showing the origin of the double layer within a capillary tube. Although the net charge within
the capillary is zero, the distribution of charge is not. The walls of the capillary have an excess of negative charge, which decreases
across the fixed layer and the diffuse layer, reaching a value of zero in bulk solution.

The anions in the diffuse layer, which also are solvated, try to move toward the anode. Because there are more cations than
anions, however, the cations win out and the electroosmotic flow moves in the direction of the cathode.

The rate at which the buffer moves through the capillary, what we call its electroosmotic flow velocity, νeof , is a function of the
applied electric field, E, and the buffer’s electroosmotic mobility, μ . eof

νeof = μeof E (30.2.3)

Electroosmotic mobility is defined as

εζ
μeof = (30.2.4)
4πη

where ϵ is the buffer dielectric constant, ζ is the zeta potential, and η is the buffer’s viscosity.
The zeta potential—the potential of the diffuse layer at a finite distance from the capillary wall—plays an important role in
determining the electroosmotic flow velocity. Two factors determine the zeta potential’s value. First, the zeta potential is directly
proportional to the charge on the capillary walls, with a greater density of silanate ions corresponding to a larger zeta potential.
Below a pH of 2 there are few silanate ions and the zeta potential and the electroosmotic flow velocity approach zero. As the pH
increases, both the zeta potential and the electroosmotic flow velocity increase. Second, the zeta potential is directly proportional to
the thickness of the double layer. Increasing the buffer’s ionic strength provides a higher concentration of cations, which decreases
the thickness of the double layer and decreases the electroosmotic flow.

The definition of zeta potential given here admittedly is a bit fuzzy. For a more detailed explanation see Delgado, A. V.;
González-Caballero, F.; Hunter, R. J.; Koopal, L. K.; Lyklema, J. “Measurement and Interpretation of Electrokinetic
Phenomena,” Pure. Appl. Chem. 2005, 77, 1753–1805. Although this is a very technical report, Sections 1.3–1.5 provide a
good introduction to the difficulty of defining the zeta potential and of measuring its value.

The electroosmotic flow profile is very different from that of a fluid moving under forced pressure. Figure 30.2.2 compares the
electroosmotic flow profile with the hydrodynamic flow profile in gas chromatography and liquid chromatography. The uniform,
flat profile for electroosmosis helps minimize band broadening in capillary electrophoresis, improving separation efficiency.

30.2.2 https://chem.libretexts.org/@go/page/362621
 Figure 30.2.2 . Comparison of hydrodynamic flow and electroosmotic flow. The nearly uniform electroosmotic flow profile means
that the electroosmotic flow velocity is nearly constant across the capillary.

Total Mobility
A solute’s total velocity, ν , as it moves through the capillary is the sum of its electrophoretic velocity and the electroosmotic flow
tot

velocity.

νtot = νep + νeof

As shown in Figure 30.2.3 , under normal conditions the following general relationships hold true.

(νtot )cations > νeof

(νtot )neutrals = νeof

(νtot )anions < νeof

Cations elute first in an order that corresponds to their electrophoretic mobilities, with small, highly charged cations eluting before
larger cations of lower charge. Neutral species elute as a single band with an elution rate equal to the electroosmotic flow velocity.
Finally, anions are the last components to elute, with smaller, highly charged anions having the longest elution time.

Figure 30.2.3 . Visual explanation for the general elution order in capillary electrophoresis. Each species has the same
electroosmotic flow, ν . Cations elute first because they have a positive electrophoretic velocity, ν . Anions elute last because
eof ep

their negative electrophoretic velocity partially offsets the electroosmotic flow velocity. Neutrals elute with a velocity equal to the
electroosmotic flow.

Migration Time
Another way to express a solute’s velocity is to divide the distance it travels by the elapsed time
l
νtot = (30.2.5)
tm

where l is the distance between the point of injection and the detector, and tm is the solute’s migration time. To understand the
experimental variables that affect migration time, we begin by noting that

νtot = μtot E = (μep + μeof )E (30.2.6)

Combining Equation 30.2.5 and Equation 30.2.6 and solving for tm leaves us with
l
tm = (30.2.7)
(μep + μeof ) E

30.2.3 https://chem.libretexts.org/@go/page/362621
The magnitude of the electrical field is
V
E = (30.2.8)
L

where V is the applied potential and L is the length of the capillary tube. Finally, substituting Equation 30.2.8 into Equation 30.2.7
leaves us with the following equation for a solute’s migration time.
lL
tm = (30.2.9)
(μep + μeof ) V

To decrease a solute’s migration time—which shortens the analysis time—we can apply a higher voltage or use a shorter capillary
tube. We can also shorten the migration time by increasing the electroosmotic flow, although this decreases resolution.

Efficiency
As we learned in Chapter 26.3, the efficiency of a separation is given by the number of theoretical plates, N. In capillary
electrophoresis the number of theoretic plates is
2
l (μep + μeof ) El
N = = (30.2.10)
2Dtm 2DL

where D is the solute’s diffusion coefficient. From Equation 30.2.10, the efficiency of a capillary electrophoretic separation
increases with higher voltages. Increasing the electroosmotic flow velocity improves efficiency, but at the expense of resolution.
Two additional observations deserve comment. First, solutes with larger electrophoretic mobilities—in the same direction as the
electroosmotic flow—have greater efficiencies; thus, smaller, more highly charged cations are not only the first solutes to elute, but
do so with greater efficiency. Second, efficiency in capillary electrophoresis is independent of the capillary’s length. Theoretical
plate counts of approximately 100 000–200 000 are not unusual.

It is possible to design an electrophoretic experiment so that anions elute before cations—more about this later—in which
smaller, more highly charged anions elute with greater efficiencies.

Selectivity
In chromatography we defined the selectivity between two solutes as the ratio of their retention factors. In capillary electrophoresis
the analogous expression for selectivity is
μep,1
α =
μep,2

where μ ep,1 and μ ep,2 are the electrophoretic mobilities for the two solutes, chosen such that α ≥ 1 . We can often improve
selectivity by adjusting the pH of the buffer solution. For example, NH is a weak acid with a pKa of 9.75. At a pH of 9.75 the
+

concentrations of NH and NH3 are equal. Decreasing the pH below 9.75 increases its electrophoretic mobility because a greater
+

fraction of the solute is present as the cation NH . On the other hand, raising the pH above 9.75 increases the proportion of neutral
+

NH3, decreasing its electrophoretic mobility.

Resolution
The resolution between two solutes is
−
−
0.177(μep,2 − μep,1 )√V
R = −−−−−−−−−−−− (30.2.11)

√ D(μavg + μeof )

where μ avgis the average electrophoretic mobility for the two solutes. Increasing the applied voltage and decreasing the
electroosmotic flow velocity improves resolution. The latter effect is particularly important. Although increasing electroosmotic
flow improves analysis time and efficiency, it decreases resolution.

30.2.4 https://chem.libretexts.org/@go/page/362621
Instrumentation
The basic instrumentation for capillary electrophoresis is shown in Figure 30.2.4 and includes a power supply for applying the
electric field, anode and cathode compartments that contain reservoirs of the buffer solution, a sample vial that contains the sample,
the capillary tube, and a detector. Each part of the instrument receives further consideration in this section.

Figure 30.2.4 . Schematic diagram of the basic instrumentation for capillary electrophoresis. The sample and the source reservoir
are switched when making injections.

Capillary Tubes
Figure 30.2.5 shows a cross-section of a typical capillary tube. Most capillary tubes are made from fused silica coated with a 15–35
μm layer of polyimide to give it mechanical strength. The inner diameter is typically 25–75 μm, which is smaller than the internal
diameter of a capillary GC column, with an outer diameter of 200–375 μm.

Figure 30.2.5 . Cross section of a capillary column for capillary electrophoresis. The dimensions shown here are typical and are
scaled proportionally in this figure.
The capillary column’s narrow opening and the thickness of its walls are important. When an electric field is applied to the buffer
solution, current flows through the capillary. This current leads to the release of heat, which we call Joule heating. The amount of
heat released is proportional to the capillary’s radius and to the magnitude of the electrical field. Joule heating is a problem because
it changes the buffer’s viscosity, with the solution at the center of the capillary being less viscous than that near the capillary walls.
Because a solute’s electrophoretic mobility depends on its viscosity (see Equation 30.2.2), solute species in the center of the
capillary migrate at a faster rate than those near the capillary walls. The result is an additional source of band broadening that
degrades the separation. Capillaries with smaller inner diameters generate less Joule heating, and capillaries with larger outer
diameters are more effective at dissipating the heat. Placing the capillary tube inside a thermostated jacket is another method for
minimizing the effect of Joule heating; in this case a smaller outer diameter allows for a more rapid dissipation of thermal energy.

Injecting the Sample

There are two common methods for injecting a sample into a capillary electrophoresis column: hydrodynamic injection and
electrokinetic injection. In both methods the capillary tube is filled with the buffer solution. One end of the capillary tube is placed
in the destination reservoir and the other end is placed in the sample vial.
Hydrodynamic injection uses pressure to force a small portion of sample into the capillary tubing. A difference in pressure is
applied across the capillary either by pressurizing the sample vial or by applying a vacuum to the destination reservoir. The volume

30.2.5 https://chem.libretexts.org/@go/page/362621
of sample injected, in liters, is given by the following equation
4
ΔP d πt 3
Vinj = × 10 (30.2.12)
128ηL

where ΔP is the difference in pressure across the capillary in pascals, d is the capillary’s inner diameter in meters, t is the amount
of time the pressure is applied in seconds, η is the buffer’s viscosity in kg m–1 s–1, and L is the length of the capillary tubing in
meters. The factor of 103 changes the units from cubic meters to liters.

For a hydrodynamic injection we move the capillary from the source reservoir to the sample. The anode remains in the source
reservoir. A hydrodynamic injection also is possible if we raise the sample vial above the destination reservoir and briefly
insert the filled capillary.

 Example 30.2.1

In a hydrodynamic injection we apply a pressure difference of 2.5 × 10 Pa (a ΔP ≈ 0.02 atm) for 2 s to a 75-cm long
3

capillary tube with an internal diameter of 50 μm. Assuming the buffer’s viscosity is 10–3 kg m–1 s–1, what volume and length
of sample did we inject?

Solution
Making appropriate substitutions into Equation 30.2.12 gives the sample’s volume as
3 −6 4
−1 −2
(2.5 × 10  kg  m  s ) (50 × 10  m) (3.14)(2 s)
3 3
Vinj = × 10 L/ m
(128) (0.001 kg  m −1  s −1 ) (0.75 m)

−9
Vinj = 1 × 10  L = 1 nL

Because the interior of the capillary is cylindrical, the length of the sample, l, is easy to calculate using the equation for the
volume of a cylinder; thus
−9 −3 3
Vinj (1 × 10  L) (10  m /L)
−4
l = = = 5 × 10  m = 0.5 mm
2 2
πr −6
(3.14) (25 × 10  m)

 Exercise 30.2.1

Suppose you need to limit your injection to less than 0.20% of the capillary’s length. Using the information from Example
30.2.1 , what is the maximum injection time for a hydrodynamic injection?

Answer
The capillary is 75 cm long, which means that 0.20% of that sample’s maximum length is 0.15 cm. To convert this to the
maximum volume of sample we use the equation for the volume of a cylinder.
2
2 −4 −6 3
Vinj = lπ r = (0.15 cm)(3.14) (25 × 10  cm) = 2.94 × 10  cm

Given that 1 cm3 is equivalent to 1 mL, the maximum volume is 2.94 × 10 −6

mL or 2.94 × 10 −9
L. To find the maximum
injection time, we first solve Equation 30.2.12for t
128 Vinj ηL
−3 3
t = × 10  m /L
4
Pd π

and then make appropriate substitutions.

−9 −1 −1
(128) (2.94 × 10  L) (0.001 kg  m  s ) (0.75 m) −3 3
10  m
t = × = 5.8 s
4
3 −1 −2 −6 L
(2.5 × 10  kg m  s ) (50 × 10  m) (3.14)

30.2.6 https://chem.libretexts.org/@go/page/362621
 The maximum injection time, therefore, is 5.8 s.

In an electrokinetic injection we place both the capillary and the anode into the sample and briefly apply an potential. The volume
of injected sample is the product of the capillary’s cross sectional area and the length of the capillary occupied by the sample. In
turn, this length is the product of the solute’s velocity (see Equation 30.2.6) and time; thus
2 2 ′
Vinj = π r L = π r (μep + μeof )E t (30.2.13)

where r is the capillary’s radius, L is the capillary’s length, and E is the effective electric field in the sample. An important
′

consequence of Equation 30.2.13 is that an electrokinetic injection is biased toward solutes with larger electrophoretic mobilities. If
two solutes have equal concentrations in a sample, we inject a larger volume—and thus more moles—of the solute with the larger
μep.

The electric field in the sample is different that the electric field in the rest of the capillary because the sample and the buffer
have different ionic compositions. In general, the sample’s ionic strength is smaller, which makes its conductivity smaller. The
effective electric field is
χbuffer
′
E =E×
χsample

where χ buffer and χ

sample are the conductivities of the buffer and the sample, respectively.

When an analyte’s concentration is too small to detect reliably, it maybe possible to inject it in a manner that increases its
concentration. This method of injection is called stacking. Stacking is accomplished by placing the sample in a solution whose
ionic strength is significantly less than that of the buffer in the capillary tube. Because the sample plug has a lower concentration of
buffer ions, the effective field strength across the sample plug, E , is larger than that in the rest of the capillary.
′

We know from Equation 30.2.1 that electrophoretic velocity is directly proportional to the electrical field. As a result, the cations in
the sample plug migrate toward the cathode with a greater velocity, and the anions migrate more slowly—neutral species are
unaffected and move with the electroosmotic flow. When the ions reach their respective boundaries between the sample plug and
the buffer, the electrical field decreases and the electrophoretic velocity of the cations decreases and that for the anions increases.
As shown in Figure 30.2.6 , the result is a stacking of cations and anions into separate, smaller sampling zones. Over time, the
buffer within the capillary becomes more homogeneous and the separation proceeds without additional stacking.

Figure 30.2.6 . The stacking of cations and anions. The top diagram shows the initial sample plug and the bottom diagram shows
how the cations and anions are concentrated at opposite sides of the sample plug.

Applying the Electrical Field

Migration in electrophoresis occurs in response to an applied electric field. The ability to apply a large electric field is important
because higher voltages lead to shorter analysis times (Equation 30.2.9), more efficient separations (Equation 30.2.10), and better
resolution (Equation 30.2.11). Because narrow bored capillary tubes dissipate Joule heating so efficiently, voltages of up to 40 kV
are possible.

Because of the high voltages, be sure to follow your instrument’s safety guidelines.

30.2.7 https://chem.libretexts.org/@go/page/362621
Detectors
Most of the detectors used in HPLC also find use in capillary electrophoresis. Among the more common detectors are those based
on the absorption of UV/Vis radiation, fluorescence, conductivity, amperometry, and mass spectrometry. Whenever possible,
detection is done “on-column” before the solutes elute from the capillary tube and additional band broadening occurs.
UV/Vis detectors are among the most popular. Because absorbance is directly proportional to path length, the capillary tubing’s
small diameter leads to signals that are smaller than those obtained in HPLC. Several approaches have been used to increase the
pathlength, including a Z-shaped sample cell and multiple reflections (see Figure 30.2.7 ). Detection limits are about 10–7 M.

Figure 30.2.7 . Two approaches to on-column detection in capillary electrophoresis using a UV/Vis diode array spectrometer: (a) Z-
shaped bend in capillary, and (b) multiple reflections.
Better detection limits are obtained using fluorescence, particularly when using a laser as an excitation source. When using
fluorescence detection a small portion of the capillary’s protective coating is removed and the laser beam is focused on the inner
portion of the capillary tubing. Emission is measured at an angle of 90o to the laser. Because the laser provides an intense source of
radiation that can be focused to a narrow spot, detection limits are as low as 10–16 M.
Solutes that do not absorb UV/Vis radiation or that do not undergo fluorescence can be detected by other detectors. Table 30.2.1
provides a list of detectors for capillary electrophoresis along with some of their important characteristics.
Table 30.2.1 . Characteristics of Detectors for Capillary Electrophoresis
selectivity (universal or detection limited (moles
detector detection limit (molarity) on-column detection?
analyte must ...) injected)

have a UV/Vis
UV/Vis absorbance 10
−13 −16
− 10 10
−5
− 10
−7
yes
chromophore

indirect absorbancd universal 10

−12 −15
− 10 10
−4
− 10
−6
yes

have a favorable quantum

fluoresence 10
−13 −17
− 10 10
−7
− 10
−9
yes
yield

have a favorable quantum

laser fluorescence 10
−18 −20
− 10 10
−13
− 10
−16
yes
yield

universal (total ion)

mass spectrometer 10
−16 −17
− 10 10
−8 −10
− 10 no
selective (single ion)

undergo oxidation or
amperometry 10
−18 −19
− 10 10
−7 −10
− 10 no
reduction

conductivity universal 10
−15 −16
− 10 10
−7
− 10
−9
no

radiometric be radioactive 10
−17 −19
− 10 10
−10
− 10
−12
yes

30.2.8 https://chem.libretexts.org/@go/page/362621
This page titled 30.2: Capillary Electrophoresis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

30.2.9 https://chem.libretexts.org/@go/page/362621
30.3: Applications of Capillary Electrophoresis
There are several different forms of capillary electrophoresis, each of which has its particular advantages. Several of these methods
are described briefly in this section.

Capillary Zone Electrophoresis (CZE)

The simplest form of capillary electrophoresis is capillary zone electrophoresis. In CZE we fill the capillary tube with a buffer and,
after loading the sample, place the ends of the capillary tube in reservoirs that contain additional buffer. Usually the end of the
capillary containing the sample is the anode and solutes migrate toward the cathode at a velocity determined by their respective
electrophoretic mobilities and the electroosmotic flow. Cations elute first, with smaller, more highly charged cations eluting before
larger cations with smaller charges. Neutral species elute as a single band. Anions are the last species to elute, with smaller, more
negatively charged anions being the last to elute.
We can reverse the direction of electroosmotic flow by adding an alkylammonium salt to the buffer solution. As shown in Figure
30.3.1 , the positively charged end of the alkyl ammonium ions bind to the negatively charged silanate ions on the capillary’s walls.
The tail of the alkyl ammonium ion is hydrophobic and associates with the tail of another alkyl ammonium ion. The result is a layer
of positive charges that attract anions in the buffer. The migration of these solvated anions toward the anode reverses the
electroosmotic flow’s direction. The order of elution is exactly opposite that observed under normal conditions.

Figure 30.3.1 . Two modes of capillary zone electrophoresis showing (a) normal migration with electroosmotic flow toward the
cathode and (b) reversed migration in which the electroosmotic flow is toward the anode.
Coating the capillary’s walls with a nonionic reagent eliminates the electroosmotic flow. In this form of CZE the cations migrate
from the anode to the cathode. Anions elute into the source reservoir and neutral species remain stationary.
Capillary zone electrophoresis provides effective separations of charged species, including inorganic anions and cations, organic
acids and amines, and large biomolecules such as proteins. For example, CZE was used to separate a mixture of 36 inorganic and
organic ions in less than three minutes [Jones, W. R.; Jandik, P. J. Chromatog. 1992, 608, 385–393]. A mixture of neutral species,
of course, can not be resolved.

Micellar Electrokinetic Capillary Chromatography (MEKC)

One limitation to CZE is its inability to separate neutral species. Micellar electrokinetic capillary chromatography overcomes this
limitation by adding a surfactant, such as sodium dodecylsulfate (Figure 30.3.2 a) to the buffer solution. Sodium dodecylsulfate, or
SDS, consists of a long-chain hydrophobic tail and a negatively charged ionic functional group at its head. When the concentration
of SDS is sufficiently large a micelle forms. A micelle consists of a spherical agglomeration of 40–100 surfactant molecules in
which the hydrocarbon tails point inward and the negatively charged heads point outward (Figure 30.3.2 b).

30.3.1 https://chem.libretexts.org/@go/page/362623
 Figure 30.3.2 . (a) Structure of sodium dodecylsulfate and (b) cross section through a micelle showing its hydrophobic interior and
its hydrophilic exterior.
Because micelles have a negative charge, they migrate toward the cathode with a velocity less than the electroosmotic flow
velocity. Neutral species partition themselves between the micelles and the buffer solution in a manner similar to the partitioning of
solutes between the two liquid phases in HPLC. Because there is a partitioning between two phases, we include the descriptive
term chromatography in the techniques name. Note that in MEKC both phases are mobile.
The elution order for neutral species in MEKC depends on the extent to which each species partitions into the micelles.
Hydrophilic neutrals are insoluble in the micelle’s hydrophobic inner environment and elute as a single band, as they would in
CZE. Neutral solutes that are extremely hy- drophobic are completely soluble in the micelle, eluting with the micelles as a single
band. Those neutral species that exist in a partition equilibrium between the buffer and the micelles elute between the completely
hydro- philic and completely hydrophobic neutral species. Those neutral species that favor the buffer elute before those favoring
the micelles. Micellar electrokinetic chromatography is used to separate a wide variety of samples, including mixtures of
pharmaceutical compounds, vitamins, and explosives.

Capillary Gel Electrophoresis (CGE)

In capillary gel electrophoresis the capillary tubing is filled with a polymeric gel. Because the gel is porous, a solute migrates
through the gel with a velocity determined both by its electrophoretic mobility and by its size. The ability to effect a separation
using size is helpful when the solutes have similar electrophoretic mobilities. For example, fragments of DNA of varying length
have similar charge-to-size ratios, making their separation by CZE difficult. Because the DNA fragments are of different size, a
CGE separation is possible.
The capillary used for CGE usually is treated to eliminate electroosmotic flow to prevent the gel from extruding from the capillary
tubing. Samples are injected electrokinetically because the gel provides too much resistance for hydrodynamic sampling. The
primary application of CGE is the separation of large biomolecules, including DNA fragments, proteins, and oligonucleotides.

Capillary Electrochromatography (CEC)

Another approach to separating neutral species is capillary electrochromatography. In CEC the capillary tubing is packed with 1.5–
3 μm particles coated with a bonded stationary phase. Neutral species separate based on their ability to partition between the
stationary phase and the buffer, which is moving as a result of the electroosmotic flow; Figure 30.3.3 provides a representative
example for the separation of a mixture of hydrocarbons. A CEC separation is similar to the analogous HPLC separation, but
without the need for high pressure pumps. Efficiency in CEC is better than in HPLC, and analysis times are shorter.

30.3.2 https://chem.libretexts.org/@go/page/362623
 Figure 30.3.3 . Capillary electrochromatographic separation of a mixture of hydrocarbons in DMSO. The column contains a porous
polymer of butyl methacrylate and lauryl acrylate (25%:75% mol:mol) with butane dioldacrylate as a crosslinker. Data provided by
Zoe LaPier and Michelle Bushey, Department of Chemistry, Trinity University.

This page titled 30.3: Applications of Capillary Electrophoresis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

30.3.3 https://chem.libretexts.org/@go/page/362623
 CHAPTER OVERVIEW
31: Thermal Methods
A thermal method of analysis is a technique in which measure a physical property of a material as we subject it to a change in
temperature. In this chapter we consider three examples of thermal methods: thermogravimetry, differential thermal analysis, and
differential scanning calorimetry.
31.1: Thermogravimetry
31.2: Differential Thermal Analysis and Differential Scanning Calorimetry

This page titled 31: Thermal Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

1
31.1: Thermogravimetry
One method for determining the products of a thermal decomposition is to monitor the sample’s mass as a function of temperature,
a process called a thermogravimetric analysis (TGA) or thermogravimetry. Figure 31.1.1 shows a typical thermogram in which
each change in mass—each “step” in the thermogram—represents the loss of a volatile product. As the following example
illustrates, we can use a thermogram to identify a compound’s decomposition reactions.

Figure 31.1.1 . Thermogram for CaC2O4•H2O obtained by heating a sample from room temperature to 1000oC at a rate of
20oC/min. Each change in mass results from the loss of a volatile product. The sample’s initial mass and its mass after each loss are
shown by the dotted lines. See Example 31.1.1 for information on interpreting this thermogram.

 Example 31.1.1

The thermogram in Figure 31.1.1 shows the mass of a sample of calcium oxalate monohydrate, CaC2O4•H2O, as a function of
temperature. The original sample of 17.61 mg was heated from room temperature to 1000oC at a rate of 20oC per minute. For
each step in the thermogram, identify the volatilization product and the solid residue that remains.

Solution
From 100–250oC the sample loses 17.61 mg – 15.44 mg, or 2.17 mg, which is
2.17 mg
× 100 = 12.3%
17.61 mg

of the sample’s original mass. In terms of CaC2O4•H2O, this corresponds to a decrease in the molar mass of

0.123 × 146.11 g/mol = 18.0 g/mol

The product’s molar mass and the temperature range for the decomposition, suggest that this is a loss of H2O(g), leaving a
residue of CaC2O4.
The loss of 3.38 mg from 350–550oC is a 19.2% decrease in the sample’s original mass, or a decrease in the molar mass of

0.192 × 146.11 g/mol = 28.1 g/mol

which is consistent with the loss of CO(g) and a residue of CaCO3.

31.1.1 https://chem.libretexts.org/@go/page/363138
 Finally, the loss of 5.30 mg from 600-800oC is a 30.1% decrease in the sample’s original mass, or a decrease in molar mass of

0.301 × 146.11 g/mol = 44.0 g/mol

This loss in molar mass is consistent with the release of CO2(g), leaving a final residue of CaO. The three decomposition
reactions are
CaC2 O4 ⋅ H2 O(s) →  CaC2 O4 (s) + 2 H2 O(l)

CaC2 O4 (s) →  CaCO3 (s) + CO(g)

CaCO3 (s) →  CaO(s) + CO2 (g)

Identifying the products of a thermal decomposition provides information that we can use to develop an analytical procedure. For
example, the thermogram in Figure 31.1.1 shows that we must heat a precipitate of CaC2O4•H2O to a temperature between 250 and
400oC if we wish to isolate and weigh CaC2O4. Alternatively, heating the sample to 1000oC allows us to isolate and weigh CaO.

 Exercise 31.1.1

Under the same conditions as Figure 31.1.1 , the thermogram for a 22.16 mg sample of MgC2O4•H2O shows two steps: a loss
of 3.06 mg from 100–250oC and a loss of 12.24 mg from 350–550oC. For each step, identify the volatilization product and the
solid residue that remains. Using your results from this exercise and the results from Example 31.1.1 , explain how you can use
thermogravimetry to analyze a mixture that contains CaC2O4•H2O and MgC2O4•H2O. You may assume that other components
in the sample are inert and thermally stable below 1000oC.

Answer
From 100–250oC the sample loses 13.8% of its mass, or a loss of

0.138 × 130.34 g/mol = 18.0 g/mol

which is consistent with the loss of H2O(g) and a residue of MgC2O4.

From 350–550oC the sample loses 55.23% of its original mass, or a loss of

0.5523 × 130.34 g/mol = 71.99 g/mol

This weight loss is consistent with the simultaneous loss of CO(g) and CO2(g), leaving a residue of MgO.
We can analyze the mixture by heating a portion of the sample to 300oC, 600oC, and 1000oC, recording the mass at each
temperature. The loss of mass between 600oC and 1000oC, Δm , is due to the loss of CO2(g) from the decomposition of
2

CaCO3 to CaO, and is proportional to the mass of CaC2O4•H2O in the sample.

1 mol CO2 146.11 g CaC2 O4 ⋅ H2 O
g CaC2 O4 ⋅ H2 O = Δm2 × ×
44.01 g CO2 mol CO2

o o
The change in mass between 300 C and 600 C, Δm , is due to the loss of CO(g) from CaC2O4•H2O and the loss of CO(g)
1

and CO2(g) from MgC2O4•H2O. Because we already know the amount of CaC2O4•H2O in the sample, we can calculate its
contribution to Δm .1

1 mol CO 28.01 g CO
(Δm1 ) = g CaC2 O4 ⋅ H2 O = Δm2 × ×
Ca
146.11 g CaC2 O4 ⋅ H2 O mol CO

o o
The change in mass between 300 C and 600 C due to the decomposition of MgC2O4•H2O

(m1 ) = Δm1 − (Δm1 )

Mg Ca

provides the mass of MgC2O4•H2O in the sample.

1 mol (CO  +  CO2 ) 78.02 g  (CO  +  CO2 )
g MgC O4 ⋅ H2 O = (Δm1 ) × ×
2 Mg
130.35 g MgC2 O4 ⋅ H2 O mol  (CO  +  CO2 )

31.1.2 https://chem.libretexts.org/@go/page/363138
Instrumentation
In a thermogravimetric analysis, the sample is placed on a small balance pan attached to one arm of an electromagnetic balance
(Figure 31.1.2 ). The sample is lowered into an electric furnace and the furnace’s temperature is increased at a fixed rate of a few
degrees per minute while monitoring continuously the sample’s weight. The instrument usually includes a gas line for purging the
volatile decomposition products out of the furnace, and a heat exchanger to dissipate the heat emitted by the furnace.

Figure 31.1.2 . (a) Instrumentation for conducting a thermogravimetric analysis. The balance sits on the top of the instrument with
the sample suspended below. A gas line supplies an inert gas that sweeps the volatile decomposition products out of the furnace.
The heat exchanger dissipates the heat from the furnace to a reservoir of water. (b) Close-up showing the balance pan, which sits on
a moving platform, the thermocouple for monitoring temperature, a hook for lowering the sample pan into the furnace, and the
opening to the furnace. After placing a small portion of the sample on the balance pan, the platform rotates over the furnace and
transfers the balance pan to a hook that is suspended from the balance. Once the balance pan is in place, the platform rotates back
to its initial position. The balance pan and the thermocouple are then lowered into the furnace.

Applications
Perhaps the most important application gravimetry is exploring a compound's thermal stability, as illustrated in Figure 31.1.1 and
Exercise 31.1.1 for calcium oxalate hydrate. TGA is particularly useful for studying the thermal stability of polymers.

This page titled 31.1: Thermogravimetry is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

31.1.3 https://chem.libretexts.org/@go/page/363138
31.2: Differential Thermal Analysis and Differential Scanning Calorimetry
Differential thermal analysis (DTA) and differential scanning calorimetry (DSC) are similar methods in which the response of a
sample and a reference to a change in temperature. In DTA the temperature applied to the sample is increased linearly and the
difference between the temperature of the reference material, T , and the temperature of the sample, T
ref , is recorded as
samp

function of the sample's temperature

ΔT = Tref − Tsamp

When the sample undergoes an exothermic process, such as a crystallization or a chemical reaction, the temperature of the sample
increases more than does the temperature of the reference, resulting in a more negative value for ΔT . For an endothermic process,
such as melting of a crystalline material or the loss of waters of hydration, the sample's temperature lags behind that for the
reference materials, resulting in a more positive value for ΔT . Figure 31.2.1 shows the general shape of DTA curve with negative
peaks signaling an endothermic process and positive peaks signaling an exothermic process. Changes in ΔT that are not peaks, but
shifts in the baseline—as seen at the far left of the curve in Figure 31.2.1—are the result of a simple phase transition for which
ΔH = 0 .In DSC the temperature applied to the sample is increased linearly and the relative amount of heat needed to maintain the

sample and the reference at the same temperature is measured. For an endothermic process, more heat flows into the sample and for
an exothermic process, less heat flows into the sample. The result is a DSC curve that looks similar to that for DTA (see Figure
31.2.1).

Figure 31.2.1 : Illustration showing typical experimental curves for a differential thermal analysis (DTA) or a differential scanning
calorimetry analysis (DSC) for a polymeric material. For a DTA analysis, the difference in temperature, ΔT , is measured and for a
DSC analysis, the flow of heat is measured. The three responses seen here are, from left-to-right, a phase transition, an exothermic
crystallization, and an endothermic melting.

Instrumentation
Figure 31.2.2 shows the basic components of a heat-flux differential scanning calorimeter. The sample and the reference materials
are sealed within small aluminum pans and placed on separate platforms within the sample chamber. The two platforms are
connected by a metal disk that provides a low resistance path for moving heat between the sample and the reference to maintain a
ΔT of zero between the two. Another instrumental design for differential scanning calorimetry, which is called power

compensation DSC, places the sample and the reference in separated heating chambers and measures the difference in the power
applied to the two chambers needed to maintain a ΔT of zero.

31.2.1 https://chem.libretexts.org/@go/page/363140
 Figure 31.2.2 : Components of instrument for a differential scanning calorimetry (DSC) experiment. The sample and the reference
material are (a) placed in small aluminum pans, each with a separate lid, and then (b) the lid and the pan are crimped together. The
sample chamber in (c) allows for applying heat to the sample and the reference, which are place in the center of the chamber. A
close-up of the sample chamber (d) shows two—well-used—platforms on which the sample and the reference are placed. An
insulated cap sits on top of the sample chamber. Although the basic set-up shown here is for DSC, it is similar to the
instrumentation for a differential thermal analysis. For a sense of scale, the pan in (a) is approximately 5 mm across and has a depth
of 1 mm, and the platforms in (d) also are 5 mm across.

Applications
Integrating a peak in DSC or DTA to determine its its area, A , gives a signal that is proportional to ΔH

ΔH = K × A

where the calibration constant, k , is determined using an established reference material. Both DSC and DTA find applications in
the study of polymers, liquid crystals, and pharmaceutical compounds.

This page titled 31.2: Differential Thermal Analysis and Differential Scanning Calorimetry is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by David Harvey.

31.2.2 https://chem.libretexts.org/@go/page/363140
 CHAPTER OVERVIEW
32: Radiochemical Methods
Radiochemical methods of analysis take advantage of the instability of some elemental isotopes, which decay through the release
of alpha particles, beta particles, gamma rays, and/or X-rays, often provide for a selective analysis for one analyte in a complex
mixture of other species without the need for a prior separation. In this chapter we review the basics of radioactive decay and its
direct application to samples, and two additional methods of importance: neutron activation and isotope dilution.
32.1: Radioactive Isotopes
32.2: Instrumentation
32.3: Neutron Activation Methods
32.4: Isotope Dilution Methods

Thumbnail: Visual representation of alpha decay. (Public Domain; Inductiveload via Wikipedia)

This page titled 32: Radiochemical Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1
32.1: Radioactive Isotopes
Atoms that have the same number of protons but a different number of neutrons are isotopes. To identify an isotope we use the
notation E , where E is the element’s atomic symbol, Z is the element’s atomic number, and A is the element’s atomic mass
A
Z

number. Although an element’s different isotopes have the same chemical properties, their nuclear properties are not identical. The
most important difference between isotopes is their stability. The nuclear configuration of a stable isotope remains constant with
time. Unstable isotopes, however, disintegrate spontaneously, emitting radioactive decay particles as they transform into a more
stable form.

An element’s atomic number, Z, is equal to the number of protons and its atomic mass, A, is equal to the sum of the number of
protons and neutrons. We represent an isotope of carbon-13 as C because carbon has six protons and seven neutrons.
13
6

Sometimes we omit Z from this notation—identifying the element and the atomic number is repetitive because all isotopes of
carbon have six protons and any atom that has six protons is an isotope of carbon. Thus, 13C and C–13 are alternative notations
for this isotope of carbon.

Types of Radioactive Decay Particles

The most important types of radioactive particles are alpha particles, beta particles, gamma rays, and X-rays. An alpha particle, α ,
is equivalent to a helium nucleus, He. When an atom emits an alpha particle, the product is a new atom whose atomic number and
4
2

atomic mass number are, respectively, 2 and 4 less than its unstable parent. The decay of uranium to thorium is one example of
alpha emission.
238 234
U ⟶ Th + α
92 90

A beta particle, β, comes in one of two forms. A negatron, 0

−1
, is produced when a neutron changes into a proton, increasing the
β

atomic number by one, as shown here for lead.

214 214 0
Pb ⟶ Bi + β
82 83 −1

The conversion of a proton to a neutron results in the emission of a positron, 0

1
β .
30 30 0
P ⟶ Si + β
15 14 1

A negatron, which is the more common type of beta particle, is equivalent to an electron.

The emission of an alpha or a beta particle often produces an isotope in an unstable, high energy state. This excess energy is
released as a gamma ray, γ, or as an X-ray. Gamma ray and X-ray emission may also occur without the release of an alpha particle
or a beta particle.

Radioactive Decay Rates

A radioactive isotope’s rate of decay, or activity, follows first-order kinetics
dN
A−t = − = λN (32.1.1)
dt

where A is the isotope’s activity, N is the number of radioactive atoms present in the sample at time t, and λ is the isotope’s decay
constant. Activity is expressed as the number of disintegrations per unit time.
As with any first-order process, we can rewrite Equation 32.1.1 in an integrated form.
−λt
Nt = N0 e (32.1.2)

Substituting Equation 32.1.2 into Equation 32.1.1 gives

−λt −λt
At = λ N0 e = A0 e (32.1.3)

32.1.1 https://chem.libretexts.org/@go/page/364128
If we measure a sample’s activity at time t we can determine the sample’s initial activity, A0, or the number of radioactive atoms
originally present in the sample, N0.
An important characteristic property of a radioactive isotope is its half-life, t1/2, which is the amount of time required for half of the
radioactive atoms to disintegrate. For first-order kinetics the half-life is
0.693
t1/2 = (32.1.4)
λ

Because the half-life is independent of the number of radioactive atoms, it remains constant throughout the decay process. For
example, if 50% of the radioactive atoms remain after one half-life, then 25% remain after two half-lives, and 12.5% remain after
three half-lives.

Suppose we begin with an N0 of 1200 atoms During the first half-life, 600 atoms disintegrate and 600 remain. During the
second half-life, 300 of the 600 remaining atoms disintegrate, leaving 300 atoms or 25% of the original 1200 atoms. Of the
300 remaining atoms, only 150 remain after the third half-life, or 12.5% of the original 1200 atoms.

Kinetic information about a radioactive isotope usually is given in terms of its half-life because it provides a more intuitive sense of
the isotope’s stability. Knowing, for example, that the decay constant for Sr is 0.0247 yr–1 does not give an immediate sense of
90
38

how fast it disintegrates. On the other hand, knowing that its half-life is 28.1 yr makes it clear that the concentration of Sr in a 90
38

sample remains essentially constant over a short period of time.

Counting Statistics
Radioactivity does not follow a normal distribution because the possible outcomes are not continuous; that is, a sample can emit 1
or 2 or 3 alpha particles (or some other integer value) in a fixed intervale, but it cannot emit 2.59 alpha particles during that same
interval. A Poisson distribution, however, provides the probability that a given number of events will occur in a fixed interval in
time or space if the event has a known average rate and if each new event is independent of the preceding event. Mathematically a
Poisson distribution is defined by the equation
−λ X
e λ
P (X, λ) =
X!

where P (X, λ) is the probability that an event happens X times given the event’s average rate, λ . The Poisson distribution has a
theoretical mean, μ , and a theoretical variance, σ , that are each equal to λ .
2

 Note

For a more detailed discussion of the distribution of data, including normal distributions and Poisson distributions, see
Appendix 1.

The accuracy and precision of radiochemical methods generally are within the range of 1–5%. We can improve the precision—
which is limited by the random nature of radioactive decay—by counting the emission of radioactive particles for as long a time as
is practical. If the number of counts, M, is reasonably large (M ≥ 100), and the counting period is significantly less than the
isotope’s half-life, then the percent relative standard deviation for the activity, (σ ) , is approximately
A rel

1
(σA ) = − − × 100
rel
√M

For example, if we determine the activity by counting 10 000 radioactive particles, then the relative standard deviation is 1%. A
radiochemical method’s sensitivity is inversely proportional to (σ ) , which means we can improve the sensitivity by counting
A rel

more particles.

Analysis of Radioactive Analytes

The concentration of a long-lived radioactive isotope remains essentially constant during the period of analysis. As shown in
Example 32.1.1 , we can use the sample’s activity to calculate the number of radioactive particles in the sample.

32.1.2 https://chem.libretexts.org/@go/page/364128
 Example 32.1.1

The activity in a 10.00-mL sample of wastewater that contains 90

38
Sr is 9.07 × 10
6
disintegrations/s. What is the molar
concentration of Sr in the sample? The half-life for Sr is 28.1 yr.
90
38
90
38

Solution
Solving Equation 32.1.4 for λ , substituting into Equation 32.1.1, and solving for N gives
A × t1/2
N =
0.693

Before we can determine the number of atoms of Sr in the sample we must express its activity and its half-life using the
90
38

same units. Converting the half-life to seconds gives t1/2 as 8.86 × 10 s; thus, there are 8

6 8
(9.07 × 10  disintegrations/s ) (8.86 × 10  s)
16 90
= 1.16 × 10  atoms Sr
38
0.693

The concentration of 90
38
Sr in the sample is
16 90
1.16 × 10  atoms  Sr
38 −6 90
= 1.93 × 10  M  Sr
23 38
(6.022 × 10  atoms/mol ) (0.01000L)

The direct analysis of a short-lived radioactive isotope using the method outlined in Example 32.1.1 is less useful because it
provides only a transient measure of the isotope’s concentration. Instead, we can measure its activity after an elapsed time, t, and
use Equation 32.1.3 to calculate N0.
One example of a characterization application is the determination of a sample’s age based on the decay of a radioactive isotope
naturally present in the sample. The most common example is carbon-14 dating, which is used to determine the age of natural
organic materials. As cosmic rays pass through the upper atmosphere, some N atoms in the atmosphere capture high energy 14
7

neutrons, converting them into C. The C then migrates into the lower atmosphere where it oxidizes to form C-14 labeled CO2.
14
6
14
6

Animals and plants subsequently incorporate this labeled CO2 into their tissues. Because this is a steady-state process, all plants
and animals have the same ratio of C to C in their tissues. When an organism dies, the radioactive decay of C to N by β
14
6
12
6
14
6
14
7
0
−1

emission (t = 5730 years) leads to predictable reduction in the C to C ratio. We can use the change in this ratio to date samples
14
6
12
6

that are as much as 30000 years old, although the precision of the analysis is best when the sample’s age is less than 7000 years.
The accuracy of carbon-14 dating depends upon our assumption that the natural C to C ratio in the atmosphere is constant over 14
6
12
6

time. Some variation in the ratio has occurred as the result of the increased consumption of fossil fuels and the production of C 14
6

during the testing of nuclear weapons. A calibration curve prepared using samples of known age—examples of samples include
tree rings, deep ocean sediments, coral samples, and cave deposits—limits this source of uncertainty.

There is no need to prepare a calibration curve for each analysis. Instead, there is a universal calibration curve known as
IntCal. The most recent such curve, IntCal13 is described in the following paper: Reimer, P. J., et. al. “IntCal13 and Marine 13
Radiocarbon Age Calibration Curve 0–50,000 Years Cal BP,” Radiocarbon 2013, 55, 1869–1887. This calibration spans 50
000 years before the present (BP).

 Example 32.1.2

To determine the age of a fabric sample, the relative ratio of 14

6
C to 12
6
C was measured yielding a result of 80.9% of that found
in modern fibers. How old is the fabric?

Solution
Equation 32.1.3 and Equation 32.1.4 provide us with a method to convert a change in the ratio of 14
6
C to 12
6
C to the fabric’s
age. Letting A0 be the ratio of 14
6
C to C in modern fibers, we assign it a value of 1.00. The ratio of
12
6
14
6
C to
12
6
C in the sample,

32.1.3 https://chem.libretexts.org/@go/page/364128
 A, is 0.809. Solving gives
A0 t1/2 1.00 5730 yr 
t = ln × = ln × = 1750 yr 
A 0.693 0.809 0.693

Other isotopes can be used to determine a sample’s age. The age of rocks, for example, has been determined from the ratio of the
number of U to the number of stable Pb atoms produced by radioactive decay. For rocks that do not contain uranium, dating
238
92
206
82

is accomplished by comparing the ratio of radioactive K to the stable Ar. Another example is the dating of sediments collected
40
19
40
18

from lakes by measuring the amount of Pb that is present.

210
82

This page titled 32.1: Radioactive Isotopes is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

32.1.4 https://chem.libretexts.org/@go/page/364128
32.2: Instrumentation
Alpha particles, beta particles, gamma rays, and X-rays are measured by using the particle’s energy to produce an amplified pulse
of electrical current in a detector. These pulses are counted to give the rate of disintegration. There are three common types of
detectors: gas-filled detectors, scintillation counters, and semiconductor detectors. A gas-filled detector consists of a tube that
contains an inert gas, such as Ar. When a radioactive particle enters the tube it ionizes the inert gas, producing an Ar+/e– ion-pair.
Movement of the electron toward the anode and of the Ar+ toward the cathode generates a measurable electrical current. A Geiger
counter is one example of a gas-filled detector. A scintillation counter uses a fluorescent material to convert radioactive particles
into easy to measure photons. For example, one solid-state scintillation counter consists of a NaI crystal that contains 0.2% TlI,
which produces several thousand photons for each radioactive particle. Finally, in a semiconductor detector, adsorption of a single
radioactive particle promotes thousands of electrons to the semiconductor’s conduction band, increasing conductivity.

This page titled 32.2: Instrumentation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

32.2.1 https://chem.libretexts.org/@go/page/364130
32.3: Neutron Activation Methods
Few analytes are naturally radioactive. For many analytes, however, we can induce radioactivity by irradiating the sample with
neutrons in a process called neutron activation analysis (NAA). The radioactive element formed by neutron activation decays to a
stable isotope by emitting a gamma ray, and, possibly, other nuclear particles. The rate of gamma-ray emission is proportional to
the analyte’s initial concentration in the sample. For example, if we place a sample containing non-radioactive Al in a nuclear 27
13

reactor and irradiate it with neutrons, the following nuclear reaction takes place.
27 1 28
Al + n⟶ Al
13 0 13

The radioactive isotope of 28Al has a characteristic decay process that includes the release of a beta particle and a gamma ray.
28 28 0
Al ⟶ Al + β +γ
13 14 −1

When irradiation is complete, we remove the sample from the nuclear reactor, allow any short-lived radioactive interferences to
decay into the background, and measure the rate of gamma-ray emission.
The initial activity at the end of irradiation depends on the number of atoms that are present. This, in turn, is a equal to the
difference between the rate of formation for Al and its rate of disintegration
28
13

dN28
13
Al
= Φσ N27 Al − λ N28 Al (32.3.1)
dt 13 13

where Φ is the neutron flux and σ is the reaction cross-section, or probability that a Al nucleus captures a neutron. Integrating
27
13

Equation 32.3.1 over the time of irradiation, ti, and multiplying by λ gives the initial activity, A0, at the end of irradiation as
−kt
A0 = λ N28 Al = Φσ N27 Al (1 − e )
13 13

If we know the values for A0, Φ, σ, λ , and ti, then we can calculate the number of atoms of 27
13
Al initially present in the sample.
A simpler approach is to use one or more external standards. Letting (A ) and (A ) represent the analyte’s initial activity in an
0 x 0 s

unknown and in an external standard, and letting w and w represent the analyte’s weight in the unknown and in the external
x s

standard, we obtain the following pair of equations

(A0 ) = k wx (32.3.2)
x

(A0 ) = k ws (32.3.3)
s

that we can solve to determine the analyte’s mass in the sample.

As noted earlier, gamma ray emission is measured following a period during which we allow short-lived interferents to decay into
the background. As shown in Figure 32.3.1 , we determine the sample’s or the standard’s initial activity by extrapolating a curve of
activity versus time back to t = 0. Alternatively, if we irradiate the sample and the standard at the same time, and if we measure
their activities at the same time, then we can substitute these activities for (A0)x and (A0)s. This is the strategy used in the following
example.

Figure 32.3.1 . Plot of gamma-ray emission as a function of time showing how the analyte’s initial activity is determined.

 Example 32.3.1

32.3.1 https://chem.libretexts.org/@go/page/364132
 The concentration of Mn in steel is determined by a neutron activation analysis using the method of external standards. A
1.000-g sample of an unknown steel sample and a 0.950-g sample of a standard steel known to contain 0.463% w/w Mn are
irradiated with neutrons for 10 h in a nuclear reactor. After a 40-min delay the gamma ray emission is 2542 cpm (counts per
minute) for the unknown and 1984 cpm for the external standard. What is the %w/w Mn in the unknown steel sample?

Solution
Combining equations 32.3.2 and 32.3.3 gives
Ax
wx = × ws
As

The weight of Mn in the external standard is

0.00463 g Mn
ws = × 0.950 g steel  = 0.00440 g Mn
 g  steel 

Substituting into the above equation gives

2542 cpm
wx = × 0.00440 g Mn = 0.00564 g Mn
1984 cpm

Because the original mass of steel is 1.000 g, the %w/w Mn is 0.564%.

Among the advantages of neutron activation are its applicability to almost all elements in the periodic table and that it is
nondestructive to the sample. Consequently, NAA is an important technique for analyzing archeological and forensic samples, as
well as works of art.

This page titled 32.3: Neutron Activation Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

32.3.2 https://chem.libretexts.org/@go/page/364132
32.4: Isotope Dilution Methods
Another important radiochemical method for the analysis of nonradioactive analytes is isotope dilution. An external source of
analyte is prepared in a radioactive form with a known activity, A , for its radioactive decay—we call this form of the analyte a
T

tracer. To prepare a sample for analysis we add a known mass of the tracer, wT, to a portion of sample that contains an unknown
mass, wx , of analyte. After homogenizing the sample and tracer, we isolate wA grams of analyte by using a series of appropriate
chemical and physical treatments. Because these chemical and physical treatments cannot distinguish between radioactive and
nonradioactive forms of the analyte, the isolated material contains both. Finally, we measure the activity of the isolated sample, AA.
If we recover all the analyte—both the radioactive tracer and the nonradioactive analyte—then AA and A are equal and wx = wA –
T

wT. Normally, we fail to recover all the analyte. In this case A is less than A , and
A T

wA
AA = AT × (32.4.1)
wx + wT

The ratio of weights in Equation 32.4.1 accounts for any loss of activity that results from our failure to recover all the analyte.
Solving Equation 32.4.1 for wx gives
AT
wx = wA − wT (32.4.2)
AA

How we process the sample depends on the analyte and the sample’s matrix. We might, for example, digest the sample to bring
the analyte into solution. After filtering the sample to remove the residual solids, we might precipitate the analyte, isolate it by
filtration, dry it in an oven, and obtain its weight.
Given that the goal of an analysis is to determine the amount of nonradioactive analyte in our sample, the realization that we
might not recover all the analyte might strike you as unsettling. A single liquid–liquid extraction rarely has an extraction
efficiency of 100%. One advantage of isotope dilution is that the extraction efficiency for the nonradioactive analyte and for
the tracer are the same. If we recover 50% of the tracer, then we also recover 50% of the nonradioactive analyte. Because we
know how much tracer we added to the sample, we can determine how much of the nonradioactive analyte is in the sample.

 Example 32.4.1
The concentration of insulin in a production vat is determined by isotope dilution. A 1.00-mg sample of insulin labeled with
14C having an activity of 549 cpm is added to a 10.0-mL sample taken from the production vat. After homogenizing the

sample, a portion of the insulin is separated and purified, yielding 18.3 mg of pure insulin. The activity for the isolated insulin
is measured at 148 cpm. How many mg of insulin are in the original sample?

Solution
Substituting known values into Equation 32.4.1 gives
549 cpm
wx = × 18.3 mg − 1.00 mg = 66.9 mg insulin 
148 cpm

Equation 32.4.1 and Equation 32.4.2 are valid only if the tracer’s half-life is considerably longer than the time it takes to conduct
the analysis. If this is not the case, then the decrease in activity is due both to the incomplete recovery and the natural decrease in
the tracer’s activity. Table 32.4.1 provides a list of several common tracers for isotope dilution.
Table 32.4.1 . Common Tracers for Isotope Dilution
isotope half-life
3H 12.5 years
14C 5730 years
32P 14.3 days

32.4.1 https://chem.libretexts.org/@go/page/364134
 isotope half-life
35S 87.1 days
45Ca 152 days
55Fe 2.91 years
60Co 5.3 years
131I 8 days

An important feature of isotope dilution is that it is not necessary to recover all the analyte to determine the amount of analyte
present in the original sample. Isotope dilution, therefore, is useful for the analysis of samples with complex matrices, where a
complete recovery of the analyte is difficult.

This page titled 32.4: Isotope Dilution Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

32.4.2 https://chem.libretexts.org/@go/page/364134
 CHAPTER OVERVIEW
33: Automated Methods of Analysis
33.1: Overview of Automated Methods of Analysis
33.2: Flow-Injection Analysis
33.3: Other Automated Methods of Analysis

This page titled 33: Automated Methods of Analysis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

1
33.1: Overview of Automated Methods of Analysis
An automated method of analysis is one in which the one or more steps in an analysis are completed without the direct action of the
analyst. Instead, the instrument itself completes these actions. Some of these actions are carried out discretely, such as an
autosampler that can complete all facets of sample preparation, from collecting discrete samples, adding reagents, and diluting the
mixture to a desired volume, prior to the analysts analyzing the samples. Another example of a discrete instrument is an automated
titrator (see Figure 33.1.1) that relieves the analyst from manually operating a buret. Instead, the analyst introduces the sample into
the automated titrator and lets the instrument complete the titration.

Figure 33.1.1 . Typical instrumentation for an automated acid–base titration showing the titrant, the pump, and the titrand. The pH
electrode in the titrand’s solution is used to monitor the titration’s progress. You can see the titration curve in the lower-left
quadrant of the computer’s display. Modified from: Datamax (commons. Wikipedia.org).
Other automated instruments are continuous in nature, in which samples are injected, either manually or with an autosampler, into a
flowing stream of reagents that can serve to transport the samples to a detector and can serve as a source of reagents that convert
the analyte into a form suitable for analysis.
Both discrete and continuous automated methods of analysis have the advantage of allowing for a high throughput of samples and
providing for greater reproducibility in results by relieving the analyst of the tedium associated with completing repetitive tasks. In
general, continuous automated methods can handle more samples per unit time than can a discrete method.

This page titled 33.1: Overview of Automated Methods of Analysis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

33.1.1 https://chem.libretexts.org/@go/page/364254
33.2: Flow-Injection Analysis
In this section we consider the technique of flow injection analysis—a continuous automated method—in which we inject the
sample into a flowing carrier stream that gives rise to a transient signal at the detector. The shape of this transient signal depends on
the physical and chemical kinetic processes that take place in the carrier stream during the time between injection and detection.

Theory and Practice

Flow injection analysis (FIA) was developed in the mid-1970s as a highly efficient technique for the automated analyses of
samples [see, for example, (a) Ruzicka, J.; Hansen, E. H. Anal. Chim. Acta 1975, 78, 145–157; (b) Stewart, K. K.; Beecher, G. R.;
Hare, P. E. Anal. Biochem. 1976, 70, 167–173; (c) Valcárcel, M.; Luque de Castro, M. D. Flow Injection Analysis: Principles and
Applications, Ellis Horwood: Chichester, England, 1987]. Unlike the centrifugal analyzer described later in this chapter, in which
the number of samples is limited by the transfer disk’s size, FIA allows for the rapid, sequential analysis of an unlimited number of
samples. FIA is one example of a continuous-flow analyzer, in which we sequentially introduce samples at regular intervals into a
liquid carrier stream that transports them to the detector.
A schematic diagram detailing the basic components of a flow injection analyzer is shown in Figure 33.2.1 . The reagent that serves
as the carrier is stored in a reservoir, and a propelling unit maintains a constant flow of the carrier through a system of tubing that
comprises the transport system. We inject the sample directly into the flowing carrier stream, where it travels through one or more
mixing and reaction zones before it reaches the detector’s flow-cell. Figure 33.2.1 is the simplest design for a flow injection
analyzer, which consists of a single channel and a single reagent reservoir. Multiple channel instruments that merge together
separate channels, each of which introduces a new reagent into the carrier stream, also are possible.

Figure 33.2.1 . Schematic diagram of a simple flow injection analyzer showing its basic components. After its injection into the
carrier stream the samples mixes and reacts with the carrier stream’s reagents before reaching the detector.
When we first inject a sample into the carrier stream it has the rectangular flow profile of width w shown in Figure 33.2.2 a. As the
sample moves through the mixing zone and the reaction zone, the width of its flow profile increases as the sample disperses into
the carrier stream. Dispersion results from two processes: convection due to the flow of the carrier stream and diffusion due to the
concentration gradient between the sample and the carrier stream. Convection occurs by laminar flow. The linear velocity of the
sample at the tube’s walls is zero, but the sample at the center of the tube moves with a linear velocity twice that of the carrier
stream. The result is the parabolic flow profile shown in Figure 33.2.2 b. Convection is the primary means of dispersion in the first
100 ms following the sample’s injection.

Figure 33.2.2 . Effect of dispersion on the shape of a sample’s flow profile, shown in blue, at different times during a flow injection
analysis: (a) at injection; (b) when convection dominates dispersion; (c) when convection and diffusion contribute to dispersion;
and (d) when diffusion dominates dispersion. The red line shows the width, w, of the samples flow profile.

33.2.1 https://chem.libretexts.org/@go/page/364256
The second contribution to the sample’s dispersion is diffusion due to the concentration gradient that exists between the sample and
the carrier stream. As shown in Figure 33.2.2 , diffusion occurs parallel (axially) and perpendicular (radially) to the direction in
which the carrier stream is moving. Only radial diffusion is important in a flow injection analysis. Radial diffusion decreases the
sample’s linear velocity at the center of the tubing, while the sample at the edge of the tubing experiences an increase in its linear
velocity. Diffusion helps to maintain the integrity of the sample’s flow profile (Figure 33.2.2 c) and prevents adjacent samples in
the carrier stream from dispersing into one another. Both convection and diffusion make significant contributions to dispersion
from approximately 3–20 s after the sample’s injection. This is the normal time scale for a flow injection analysis. After
approximately 25 s, diffusion is the only significant contributor to dispersion, resulting in a flow profile similar to that shown in
Figure 33.2.2 d.

Figure 33.2.3 . Illustration showing axial and radial diffusion. The blue band is the sample’s flow profile and the red arrows
indicate the direction of diffusion.
An FIA curve, or fiagram, is a plot of the detector’s signal as a function of time. Figure 33.2.4 shows a typical fiagram for
conditions in which both convection and diffusion contribute to the sample’s dispersion. Also shown on the figure are several
parameters that characterize a sample’s fiagram. Two parameters define the time for a sample to move from the injector to the
detector. Travel time, ta, is the time between the sample’s injection and the arrival of its leading edge at the detector. Residence
time, T, on the other hand, is the time required to obtain the maximum signal. The difference between the residence time and the
travel time is t , which approaches zero when convection is the primary means of dispersion, and increases in value as the
′

contribution from diffusion becomes more important.

Figure 33.2.4 . Typical fiagram for flow injection analysis showing the detector's response as a function of time. See the text for an
explanation of the parameters, t , t , Δt, T , and h .
a
′ ′

The time required for the sample to pass through the detector’s flow cell—and for the signal to return to the baseline—is also
described by two parameters. The baseline-to-baseline time, Δt, is the time between the arrival of the sample’s leading edge to the
departure of its trailing edge. The elapsed time between the maximum signal and its return to the baseline is the return time, T . ′

The final characteristic parameter of a fiagram is the sample’s peak height, h.

Of the six parameters shown in Figure 33.2.4 , the most important are peak height and the return time. Peak height is important
because it is directly or indirectly related to the analyte’s concentration. The sensitivity of an FIA method, therefore, is determined
by the peak height. The return time is important because it determines the frequency with which we may inject samples. Figure
33.2.5 shows that if we inject a second sample at a time T after we inject the first sample, there is little overlap of the two FIA
′

curves. By injecting samples at intervals of T , we obtain the maximum possible sampling rate.
′

33.2.2 https://chem.libretexts.org/@go/page/364256
 Figure 33.2.5 . Effect of return time, T , on sampling frequency.
′

Peak heights and return times are influenced by the dispersion of the sample’s flow profile and by the physical and chemical
properties of the flow injection system. Physical parameters that affect h and T include the volume of sample we inject, the flow
′

rate, the length, diameter and geometry of the mixing zone and the reaction zone, and the presence of junctions where separate
channels merge together. The kinetics of any chemical reactions between the sample and the reagents in the carrier stream also
influence the peak height and return time.
Unfortunately, there is no good theory that we can use to consistently predict the peak height and the return time for a given set of
physical and chemical parameters. The design of a flow injection analyzer for a particular analytical problem still occurs largely by
a process of experimentation. Nevertheless, we can make some general observations about the effects of physical and chemical
parameters. In the absence of chemical effects, we can improve sensitivity—that is, obtain larger peak heights—by injecting larger
samples, by increasing the flow rate, by decreasing the length and diameter of the tubing in the mixing zone and the reaction zone,
and by merging separate channels before the point where the sample is injected. With the exception of sample volume, we can
increase the sampling rate—that is, decrease the return time—by using the same combination of physical parameters. Larger
sample volumes, however, lead to longer return times and a decrease in sample throughput. The effect of chemical reactivity
depends on whether the species we are monitoring is a reactant or a product. For example, if we are monitoring a reactant, we can
improve sensitivity by choosing conditions that decrease the residence time, T, or by adjusting the carrier stream’s composition so
that the reaction occurs more slowly.

Instrumentation
The basic components of a flow injection analyzer are shown in Figure 33.2.6 and include a pump to propel the carrier stream and
the reagent streams, a means to inject the sample into the carrier stream, and a detector to monitor the composition of the carrier
stream. Connecting these units is a transport system that brings together separate channels and provides time for the sample to mix
with the carrier stream and to react with the reagent streams. We also can incorporate separation modules into the transport system.
Each of these components is considered in greater detail in this section.

Figure 33.2.6 . Example of a typical flow injection analyzer that shows the pump, the injector, the transport system, which consists
of mixing/reaction coils and junctions, and the detector (minus the spectrophotometer). This particular configuration has two
channels: the carrier stream and a reagent line.

Propelling Unit
The propelling unit moves the carrier stream through the flow injection analyzer. Although several different propelling units have
been used, the most common is a peristaltic pump, which, as shown in Figure 33.2.7 , consists of a set of rollers attached to the

33.2.3 https://chem.libretexts.org/@go/page/364256
outside of a rotating drum. Tubing from the reagent reservoirs fits between the rollers and a fixed plate. As the drum rotates the
rollers squeeze the tubing, forcing the contents of the tubing to move in the direction of the rotation. Peristaltic pumps provide a
constant flow rate, which is controlled by the drum’s speed of rotation and the inner diameter of the tubing. Flow rates from
0.0005–40 mL/min are possible, which is more than adequate to meet the needs of FIA where flow rates of 0.5–2.5 mL/min are
common. One limitation to a peristaltic pump is that it produces a pulsed flow—particularly at higher flow rates—that may lead to
oscillations in the signal.

Figure 33.2.7 . Schematic diagram of a peristaltic pump.

Injector
The sample, typically 5–200 μL, is injected into the carrier stream. Although syringe injections through a rubber septum are
possible, the more common method—as seen in Figure 33.2.6 —is to use a rotary, or loop injector similar to that used in an HPLC.
This type of injector provides for a reproducible sample volume and is easily adaptable to automation, an important feature when
high sampling rates are needed.

Detector
The most common detectors for flow injection analysis are the electrochemical and optical detectors used in HPLC. These
detectors are discussed in Chapter 28 and are not considered further in this section. FIA detectors also have been designed around
the use of ion selective electrodes and atomic absorption spectroscopy.

Transport System
The heart of a flow injection analyzer is the transport system that brings together the carrier stream, the sample, and any reagents
that react with the sample. Each reagent stream is considered a separate channel, and all channels must merge before the carrier
stream reaches the detector. The complete transport system is called a manifold.
The simplest manifold has a single channel, the basic outline of which is shown in Figure 33.2.8 . This type of manifold is used for
direct analysis of analyte that does not require a chemical reaction. In this case the carrier stream serves only as a means for rapidly
and reproducibly transporting the sample to the detector. For example, this manifold design has been used for sample introduction
in atomic absorption spectroscopy, achieving sampling rates as high as 700 samples/h. A single-channel manifold also is used for
determining a sample’s pH or determining the concentration of metal ions using an ion selective electrode.

Figure 33.2.8 . Example of a single-channel manifold in which the reagent serves as the carrier stream and as a species that reacts
with the sample. The mixing/reaction coil is wrapped around a plastic cylinder.
We can also use the single-channel manifold in Figure 33.2.8 for an analysis in which we monitor the product of a chemical
reaction between the sample and a reactant. In this case the carrier stream both transports the sample to the detector and reacts with

33.2.4 https://chem.libretexts.org/@go/page/364256
the sample. Because the sample must mix with the carrier stream, a lower flow rate is used. One example is the determination of
chloride in water, which is based on the following sequence of reactions.
− −
Hg(SCN)2 (aq) + 2 Cl (aq) ⇌ HgCl2 (aq) + 2 SCN (aq)

3+ − 2+
Fe (aq) + SCN (aq) ⇌ Fe(SCN) (aq)

The carrier stream consists of an acidic solution of Hg(SCN)2 and Fe3+. Injecting a sample that contains chloride into the carrier
stream displaces thiocyanate from Hg(SCN)2. The displaced thiocyanate then reacts with Fe3+ to form the red-colored Fe(SCN)2+
complex, the absorbance of which is monitored at a wavelength of 480 nm. Sampling rates of approximately 120 samples per hour
have been achieved with this system [Hansen, E. H.; Ruzicka, J. J. Chem. Educ. 1979, 56, 677–680].
Most flow injection analyses that include a chemical reaction use a manifold with two or more channels. Including additional
channels provides more control over the mixing of reagents and the interaction between the reagents and the sample. Two
configurations are possible for a dual-channel system. A dual-channel manifold, such as the one shown in Figure 33.2.9 a, is used
when the reagents cannot be premixed because of their reactivity. For example, in acidic solutions phosphate reacts with molybdate
to form the heteropoly acid H3P(Mo12O40). In the presence of ascorbic acid the molybdenum in the heteropoly acid is reduced from
Mo(VI) to Mo(V), forming a blue-colored complex that is monitored spectrophotometrically at 660 nm [Hansen, E. H.; Ruzicka, J.
J. Chem. Educ. 1979, 56, 677–680]. Because ascorbic acid reduces molybdate, the two reagents are placed in separate channels that
merge just before the loop injector.

Figure 33.2.9 . Two examples of a dual-channel manifold for flow injection analysis. In (a) the two channels merge before the loop
injector, and in (b) the two channels merge after the loop injector.
A dual-channel manifold also is used to add a second reagent after injecting the sample into a carrier stream, as shown in Figure
33.2.9 b. This style of manifold is used for the quantitative analysis of many analytes, including the determination of a
wastewater’s chemical oxygen demand (COD) [Korenaga, T.; Ikatsu, H. Anal. Chim. Acta 1982, 141, 301–309]. Chemical oxygen
demand is a measure of the amount organic matter in the wastewater sample. In the conventional method of analysis, COD is
determined by refluxing the sample for 2 h in the presence of acid and a strong oxidizing agent, such as K2Cr2O7 or KMnO4. When
refluxing is complete, the amount of oxidant consumed in the reaction is determined by a redox titration. In the flow injection
version of this analysis, the sample is injected into a carrier stream of aqueous H2SO4, which merges with a solution of the oxidant
from a secondary channel. The oxidation reaction is kinetically slow and, as a result, the mixing coil and the reaction coil are very
long—typically 40 m—and submerged in a thermostated bath. The sampling rate is lower than that for most flow injection
analyses, but at 10–30 samples/h it is substantially greater than the redox titrimetric method.
More complex manifolds involving three or more channels are common, but the possible combination of designs is too numerous
to discuss. One example of a four-channel manifold is shown in Figure 33.2.10 .

33.2.5 https://chem.libretexts.org/@go/page/364256
 Figure 33.2.10 . Example of a four-channel manifold for a flow injection analysis.

Separation Modules
By incorporating a separation module into the flow injection manifold we can include a separation—dialysis, gaseous diffusion and
liquid-liquid extractions are examples—in a flow injection analysis. Although these separations are never complete, they are
reproducible if we carefully control the experimental conditions.
Dialysis and gaseous diffusion are accomplished by placing a semipermeable membrane between the carrier stream containing the
sample and an acceptor stream, as shown in Figure 33.2.11 . As the sample stream passes through the separation module, a portion
of those species that can cross the semipermeable membrane do so, entering the acceptor stream. This type of separation module is
common for the analysis of clinical samples, such as serum and urine, where a dialysis membrane separates the analyte from its
complex matrix. Semipermeable gaseous diffusion membranes are used for the determination of ammonia and carbon dioxide in
blood. For example, ammonia is determined by injecting the sample into a carrier stream of aqueous NaOH. Ammonia diffuses
across the semipermeable membrane into an acceptor stream that contains an acid–base indicator. The resulting acid–base reaction
between ammonia and the indicator is monitored spectrophotometrically.

Figure 33.2.11 . Separation module for a flow injection analysis using a semipermeable membrane. The smaller green solutes can
pass through the semipermeable membrane and enter the acceptor stream, but the larger blue solutes cannot. Although the
separation is not complete—note that some of the green solute remains in the sample stream and exits as waste—it is reproducible
if we do not change the experimental conditions.
Liquid–liquid extractions are accomplished by merging together two immiscible fluids, each carried in a separate channel. The
result is a segmented flow through the separation module, consisting of alternating portions of the two phases. At the outlet of the
separation module the two fluids are separated by taking advantage of the difference in their densities. Figure 33.2.12 shows a
typical configuration for a separation module in which the sample is injected into an aqueous phase and extracted into a less dense
organic phase that passes through the detector.

Figure 33.2.12 . Separation module for flow injection analysis using a liquid–liquid extraction. The inset shows the equilibrium
reaction. As the sample moves through the equilibration zone, the analyte extracts into the organic phase.

33.2.6 https://chem.libretexts.org/@go/page/364256
Quantitative Applications
In a quantitative flow injection method a calibration curve is determined by injecting a series of external standards that contain
known concentrations of analyte. The calibration curve’s format—examples include plots of absorbance versus concentration and
of potential versus concentration—depends on the method of detection.
Flow injection analysis has been used to analyze a wide variety of samples, including environmental, clinical, agricultural,
industrial, and pharmaceutical samples. The majority of analyses involve environmental and clinical samples, which is the focus of
this section.
Quantitative flow injection methods have been developed for cationic, anionic, and molecular pollutants in wastewater,
freshwaters, groundwaters, and marine waters, three examples of which were described in the previous section. Table 33.2.1
provides a partial listing of other analytes that have been determined using FIA, many of which are modifications of standard
spectrophotometric and potentiometric methods. An additional advantage of FIA for environmental analysis is the ability to
provide for the continuous, in situ monitoring of pollutants in the field [Andrew, K. N.; Blundell, N. J.; Price, D.; Worsfold, P. J.
Anal. Chem. 1994, 66, 916A–922A].
Table 33.2.1 . Selected Flow Injection Analysis Methods for Environmental Samples
analyte sample sample volume (µL) concentration range sampling frequency (h–1)

Ca2+ freshwater 20 0.8–7.2 ppm 80

Cu2+ groundwater 70–700 100–400 ppb 20

2+
Pb groundwater 70–700 0–40 ppb 20
2+
Zn seawater 1000 1–100 ppb 30–60

NH
+
4
seawater 60 0.18–18.1 ppb 288

NO
−
3
rainwater 1000 1–10 ppb 40

SO
2−
4
freshwater 400 4–140 ppb 180
–
CN industrial 10 0.3–100 ppm 40

Source: Adapted from Valcárcel, M.; Luque de Castro, M. D. Flow-Injection Analysis: Principles and Practice, Ellis Horwood: Chichester,
England, 1987.

Several standard methods for the analysis of water involve an acid–base, complexation, or redox titration. It is easy to adapt these
titrations to FIA using a single-channel manifold similar to that shown in Figure 33.2.8 [Ramsing, A. U.; Ruzicka, J.; Hansen, E. H.
Anal. Chim. Acta 1981, 129, 1–17]. The titrant—whose concentration must be stoichiometrically less than that of the analyte—and
a visual indicator are placed in the reagent reservoir and pumped continuously through the manifold. When we inject the sample it
mixes thoroughly with the titrant in the carrier stream. The reaction between the analyte, which is in excess, and the titrant
produces a relatively broad rectangular flow profile for the sample. As the sample moves toward the detector, additional mixing
occurs and the width of the sample’s flow profile decreases. When the sample passes through the detector, we determine the width
of its flow profile, ΔT , by monitoring the indicator’s absorbance. A calibration curve of ΔT versus log[analyte] is prepared using
standard solutions of analyte.
Flow injection analysis has also found numerous applications in the analysis of clinical samples, using both enzymatic and
nonenzymatic methods. Table 33.2.2 summarizes several examples.
Table 33.2.2 . Selected Flow Injection Analysis Methods for Clinical Samples
analyte sample sample volume (µL) concentration range sampling frequency (h–1)

nonenzymatic methods

Cu2+ serum 20 0.7–1.5 ppm 70

Cl– serum 60 50–150 meq/L 125

PO
3−
4
serum 200 10–60 ppm 130

total CO2 serum 50 10–50 mM 70

chlorpromazine blood plasma 200 1.5–9 μM 24

33.2.7 https://chem.libretexts.org/@go/page/364256
 analyte sample sample volume (µL) concentration range sampling frequency (h–1)

enzymatic methods

glucose blood serum 26.5 0.5–15 mM 60

urea blood serum 30 4–20 mM 60

ethanol blood 30 5–30 ppm 50

Source: Adapted from Valcárcel, M.; Luque de Castro, M. D. Flow-Injection Analysis: Principles and Practice, Ellis Horwood: Chichester,
England, 1987.

Evaluation
The majority of flow injection analysis applications are modifications of conventional titrimetric, spectrophotometric, and
electrochemical methods of analysis; thus, it is appropriate to compare FIA methods to these conventional methods. The scale of
operations for FIA allows for the routine analysis of minor and trace analytes, and for macro, meso, and micro samples. The ability
to work with microliter injection volumes is useful when the sample is scarce. Conventional methods of analysis usually have
smaller detection limits.
The accuracy and precision of FIA methods are comparable to conventional methods of analysis; however, the precision of FIA is
influenced by several variables that do not affect conventional methods, including the stability of the flow rate and the
reproducibility of the sample’s injection. In addition, results from FIA are more susceptible to temperature variations.
In general, the sensitivity of FIA is less than that for conventional methods of analysis for at least two reasons. First, as with
chemical kinetic methods, measurements in FIA are made under nonequilibrium conditions when the signal has yet to reach its
maximum value. Second, dispersion dilutes the sample as it moves through the manifold. Because the variables that affect
sensitivity are known, we can design the FIA manifold to optimize the method’s sensitivity.
Selectivity for an FIA method often is better than that for the corresponding conventional method of analysis. In many cases this is
due to the kinetic nature of the measurement process, in which potential interferents may react more slowly than the analyte.
Contamination from external sources also is less of a problem because reagents are stored in closed reservoirs and are pumped
through a system of transport tubing that is closed to the environment.
Finally, FIA is an attractive technique when considering time, cost, and equipment. When using an autosampler, a flow injection
method can achieve very high sampling rates. A sampling rate of 20–120 samples/h is not unusual and sampling rates as high as
1700 samples/h are possible. Because the volume of the flow injection manifold is small, typically less than 2 mL, the consumption
of reagents is substantially smaller than that for a conventional method. This can lead to a significant decrease in the cost per
analysis. Flow injection analysis does require the need for additional equipment—a pump, a loop injector, and a manifold—which
adds to the cost of an analysis.

For a review of the importance of flow injection analysis, see Hansen, E. H.; Miró, M. “How Flow-Injection Analysis (FIA)
Over the Past 25 Years has Changed Our Way of Performing Chemical Analyses,” TRAC, Trends Anal. Chem. 2007, 26, 18–
26.

This page titled 33.2: Flow-Injection Analysis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

33.2.8 https://chem.libretexts.org/@go/page/364256
33.3: Other Automated Methods of Analysis
In the last two sections we introduced two examples of automated methods of analysis: a brief mention of automated titrators and a
more extensive coverage of flow-injection analysis. In this section we consider three additional examples of automated methods of
analysis: the stopped-flow analyzer, the centrifugal analyzer, and disposable single-test analyzers based on thin films, screen-
printing, and paper.

Stopped-Flow Analyzer
A variety of instruments have been developed to automate the kinetic analysis of fast reactions. One example, which is shown in
Figure 33.3.1 , is the stopped-flow analyzer. The sample and the reagents are loaded into separate syringes and precisely measured
volumes are dispensed into a mixing chamber by the action of a syringe drive. The continued action of the syringe drive pushes the
mixture through an observation cell and into a stopping syringe. The back pressure generated when the stopping syringe hits the
stopping block completes the mixing, after which the reaction’s progress is monitored spectrophotometrically. With a stopped-flow
analyzer it is possible to complete the mixing of sample and reagent, and initiate the kinetic measurements in approximately 0.5
ms. By attaching an autosampler to the sample syringe it is possible to analyze up to several hundred samples per hour.

Figure 33.3.7 . Schematic diagram of a stopped-flow analyzer. The blue arrows show the direction in which the syringes are
moving.

Centrifugal Analyzer
Another instrument for kinetic measurements is the centrifugal analyzer, a partial cross section of which is shown in Figure 33.3.2 .
The sample and the reagents are placed in separate wells, which are oriented radially around a circular transfer disk. As the
centrifuge spins, the centrifugal force pulls the sample and the reagents into the cuvette where mixing occurs. A single optical
source and detector, located below and above the transfer disk’s outer edge, measures the absorbance each time the cuvette passes
through the optical beam. When using a transfer disk with 30 cuvettes and rotating at 600 rpm, we can collect 10 data points per
second for each sample.

33.3.1 https://chem.libretexts.org/@go/page/364273
 Figure 33.3.2 . Cross sections through a centrifugal analyzer showing (a) the wells that hold the sample and the reagents, (b) the
mixing of the sample and the reagents, and (c) the configuration of the spectrophotometric detector.

The ability to collect lots of data and to collect it quickly requires appropriate hardware and software. Not surprisingly,
automated kinetic analyzers developed in parallel with advances in analog and digital circuitry—the hardware—and computer
software for smoothing, integrating, and differentiating the analytical signal. For an early discussion of the importance of
hardware and software, see Malmstadt, H. V.; Delaney, C. J.; Cordos, E. A. “Instruments for Rate Determinations,” Anal.
Chem. 1972, 44(12), 79A–89A.

Disposable, Single-Test Analyzers

In comparison to other techniques, potentiometry provides a rapid, relatively low-cost means for analyzing samples. The limiting
factor when analyzing a large number of samples is the need to rinse the electrode between samples. The use of inexpensive,
disposable ion-selective electrodes can increase a lab’s sample throughput. Figure 33.3.3 shows one example of a disposable ISE
for Ag+ [Tymecki, L.; Zwierkowska, E.; Głąb, S.; Koncki, R. Sens. Actuators B 2003, 96, 482–488]. Commercial instruments for
measuring pH or potential are available in a variety of price ranges, and includes portable models for use in the field.

Figure 33.3.3 . Schematic diagram of a disposable ion-selective electrode created by screen-printing. In (a) a thin film of
conducting silver is printed on a polyester substrate and a film of Ag2S overlaid near the bottom. In (b) an insulation layer with a
small opening is layered on top exposes a portion of the Ag2S membrane that is immersed in the sample. The top of the polyester
substrate remains uncoated, which allows us to connect the electrode to a potentiometer through the Ag film. The small inset shows
the electrode’s actual size.

This page titled 33.3: Other Automated Methods of Analysis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

33.3.2 https://chem.libretexts.org/@go/page/364273
 CHAPTER OVERVIEW
34: Particle Size Determination
In our coverage of high-performance liquid chromatography (see Chapter 28) we noted that a typical packed column used porous
silica particles with a mean diameter between 3-10 µm in size. A column that a manufacturer advertises as using 5 µm particles,
however, will contain particles that are smaller and particles that are larger. As the size of the particles has an effect on the pressure
it takes to move the mobile phase through the column, knowledge of the distribution of particle sizes is of interest. In this chapter
we consider different methods for determining particle size.
34.1: Overview
34.2: Measuring Particle Size Using Sieves
34.3: Measuring Particle Size by Sedimentation
34.4: Measuring Particle Size Using Image Analysis
34.5: Measuring Particle Size Using Light Scattering

This page titled 34: Particle Size Determination is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

1
34.1: Overview
What Do We Mean By a Particle?
Particles come in many forms. Some are very small, such as nanoparticles with dimensions of 1-100 nm and that might consist of
just a few hundred atoms, and some are much larger, as in the beads of ion-exchange resin shown in Figure 34.1.1, which range in
size from approximately 300 µm to 850 µm. Or, consider soils, which generally are subdivided into four types of particles: clay,
which has particles with diameters smaller than 2 µm, silt, which has particle with diameters that range from 2 µm to 50 µm, sand,
which has particle with diameters from 50 µm to 2000 µm, and gravel, which has particles with diameters larger than 2000 µm in
size.

Figure 34.1.1 : Example of an ion-exchange resin. The individual beads seen here range in size from approximately 300 µm to 850
µm in diameter.
We often hold as an image that particles are spherical in shape, which means we can characterize them by reporting a single
number: the particle's diameter. Many particulate materials, however, are not uniform in shape. Although many of the resin beads in
Figure 34.1.1 appear spherical—the largest bead in the small cluster at the left certainly looks spherical—others of the resin beads
are distorted in shape, often appearing somewhat flattened. Still, it is not unusual to treat particles as if they are spheres. There are a
number of reasons for this. If the method we use to determine size is not based on a static image (as is the case in Figure 34.1.1),
but on a suspension of particles that are rotating rapidly on the timescale of our measurement, then the particle's shape averages out
to a sphere even if the particle itself is not a sphere. The size we report, in this case, is called an equivalent spherical diameter
(ESD), which may vary from method-to-method.

How do We Report Particle Size?

Suppose we use a method to determine the size of 10000 particles. A simple way to display the data is to use a histogram that
reports the frequency of particles in different size ranges, as we see in Figure 34.1.2. We can characterize this distribution by
reporting one or more measures of its central tendency and a measure of its spread.
Typical measures of central tendency are the mode, which is the most common result, the median, which is the result that falls
exactly in the middle of all recorded values, and the mean, which is the numerical average. For the data in Figure 34.1.2, the mode
is 0.255 µm (the center of the bin that begins at 0.200 µm and ends at 0.250 µm), the median is 0.265 µm, and the mean is 0.287
µm. If the distribution was symmetrical, then the mode, median, and mean would be identical; here, the distribution has a long tail
to the right, which increases the mean relative to the median, and increases the mean and the median relative to the mode.
A common way to report the spread is to use the width of the distribution at a frequency that is half of the maximum frequency;
this is called the full-width-at-half-maximum (FWHM). For the data in Figure 34.1.2, the maximum frequency is 1230 counts. The
FWHM is at a frequency of 615 and runs from a diameter of 0.050 µm to 0.450 µm, or FWHM = 0.450 – 0.050 = 0.400 µm.

34.1.1 https://chem.libretexts.org/@go/page/364364
 Figure 34.1.2 : Histogram showing the distribution of particle sizes for a sample consisting of 10000 particles. Each bin has a size
of 0.05 µm. The mode, median, and mean diameters are shown by the vertical dashed lines. The horizontal dashed line show the
data's full-width-at-half-maximum.
Another way to characterize the distribution of particle sizes is to plot the cumulative frequency (as a percent) as a function of the
diameter of the particles. Figure 34.1.3 shows this for the data in Figure 34.1.2 using both the binned data for the histogram
(shown as the circular black points), and using all of the underlying data (shown as the dashed blue line). The red, purple, and
green lines show the particle diameters that include 10% (D10), 50% (D50), and 90% (D90) of all particles. The value of 0.264 µm
indicates that half of the particles have diameters less than 0.264 µm and that half have diameters greater than 0.264 µm. One
measure of the distribution's relative width is the span, which is defined as
D90 − D10 0.511 − 0.093
span = = = 1.59
D50 0.264

Figure 34.1.3 : The cumulative distribution of the data from Figure 34.1.2 . The dashed lines show the D10, D50, and D90 values.

How Can We Measure Particle Size?

There are a variety of methods that we can use to determine the distribution in the sizes of a particulate material, more than we can
cover in a single chapter. Instead, we will consider four common methods: sieving, sedimentation, imaging, and light scattering.
When choosing a method, the size and form of the particles are important factors. Sieving, for example, is a practical choice when
working with solid particulates that have diameters as small as 20 µm and as large as 125 mm (Note the change from µm to mm!).
Sedimentation is useful for particles with diameters of 1 µm, which we can extend to diameters as small as 1 nm by using a

34.1.2 https://chem.libretexts.org/@go/page/364364
centrifuge. Image analysis is useful for particles between 0.5 µm and 1500 µm. Finally, light scattering is useful for particles as
small as 0.8 nm.

This page titled 34.1: Overview is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

34.1.3 https://chem.libretexts.org/@go/page/364364
34.2: Measuring Particle Size Using Sieves
The particulates in a solid matrix are separated by size using one or more sieves (Figure 34.2.1). Sieves are available in a variety of
mesh sizes, ranging from approximately 25 mm to 40 μm. By stacking together sieves of different mesh size—with the largest
mesh at the top and the smallest mesh at the bottom—we can isolate particulates into several narrow size ranges. Using the sieves
in Figure 34.2.1, for example, we can separate a solid into particles with diameters >1700 μm, with diameters between 1700 μm
and 500 μm, with diameters between 500 μm and 250 μm, and those with a diameter <250 μm. The sample is place in the
uppermost sieve and mechanical shaking used to effect the separation.

Figure 34.2.1 . Three sieves with, from left to right, mesh sizes of 1700 μm, 500 μm, and 250 μm. Source: BMK
(commons.wikimedia.com).
Because we cannot use more than a limited number of sieves in a single stack, the methods for analyzing the particle size data
presented in Chapter 34.1 will be discrete in nature instead of continuous; thus, histograms will have a relatively small number of
bins and a cumulative distribution will consist of a discrete number of points. One limitation to a sieve is that irregularly shaped
particles are sized based on their two smallest dimensions.

This page titled 34.2: Measuring Particle Size Using Sieves is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

34.2.1 https://chem.libretexts.org/@go/page/364519
34.3: Measuring Particle Size by Sedimentation
When a particle that is larger than 5 µm is placed in suspension it will slowly settle toward the bottom of its container due to the
force of gravity, a process called sedimentation. The time it takes for a particle to move a fixed distance is inversely proportional to
the difference in the density of the particle and the density of the fluid in which the particles are suspended, and inversely
proportional to the square of the particle's diameter. Larger (and denser) particles, therefore settle out more quickly than do smaller
particles, as we see in Figure 34.3.1.

Figure 34.3.1 : Schematic illustration of how sedimentation is used to determine particle size. The four rectangular sample cells
show the state of sedimentation as a function of time, with the large red particles falling to the bottom of the sample cell more
quickly than the medium green particles and the small blue particles. A light source is passed through a narrow cross-section of the
sample cell, shown here by the arrow. The light reaching the detector serves as the analytical signal.
To follow the process of sedimentation, a light source is passed through a narrow portion of the sample and the amount of light
passing through the sample monitored as a function of time. Once the largest particles pass through the sampling zone, the
transmittance of light increases. Standards with well characterized particle sizes are used to calibrate the instrument.
For smaller particles, which may remain suspended due to Brownian motion, sedimentation can be carried out using a centrifuge, a
technique known as differential centrifigual separation (DCS). As shown in Figure 34.3.2, the sample is introduced in the center of
a disk that contains the fluid through which the particles will move. As the disk spins, larger particles move more quickly,
eventually reaching the detector located at the outer edge of the disk.

Figure 34.3.2 : Schematic diagram of the disc used to separate particles by size in differential centrifugal sedimentation. The sample
is injected into the center of the disc, which contains the fluid through which particles will move. As the centrifuge spins, the
particles form bands that move toward the edge of the disc. The detector is located near the disc's edge.

This page titled 34.3: Measuring Particle Size by Sedimentation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by David Harvey.

34.3.1 https://chem.libretexts.org/@go/page/364530
34.4: Measuring Particle Size Using Image Analysis
The chapter overview includes a photograph of an ion-exchange resin's beads. The photograph includes a scale and, in principle,
we could use the photograph and scale to estimate the size of the resin's beads. Although the estimates likely are pretty crude, this
still serves as an example of image analysis in which we equip an optical microscope or electron microscope with a digital camera
that can capture images of the microscope's field of view. Software is used to differentiate the particles from the background, to
establish the particle's boundaries, and to determine the particle's size. As shown in Figure 34.4.1a, the sample is dispersed on an
optical platform and light is passed through the optical platform where it is magnified and focused before capturing the image using
a camera. The optical platform can be manually or automatically moved in the xy-plane to capture more images, as in Figure
34.4.1b. The software then sorts the particles into groups based on size and reports a count of particles in each group, as in Figure

34.4.1c. Because the particles remained immobile, this is called a static image analysis.

Figure 34.4.1 : Illustration of a static image analysis. The basic instrument is shown in (a) and consists of a light source, an movable
optical platform, lens for magnifing and focusing the image, and a camera for capturing the image, an example of which appears in
(b). The instrument's software returns, in (c), information about the distribution of particles sizes.
One limitation to static imaging analysis is that it generally samples a small number of particles as they must be sufficiently
dispersed on the optical platform to allow the individual particles to be imaged, analyzed, and counted. In dynamic imaging
analysis, the sample is placed in a flow cell set perpendicular to the camera and the light source. Images are collected by using a
high-speed flash and shutter speed to capture a sequence of images that are analyzed. By essentially creating an infinite optical
platform, dynamic imaging analysis can achieve analysis rates of 10000 particles per minute.

This page titled 34.4: Measuring Particle Size Using Image Analysis is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by David Harvey.

34.4.1 https://chem.libretexts.org/@go/page/364541
34.5: Measuring Particle Size Using Light Scattering
The blue color of the sky during the day and the red color of the sun at sunset are the result of light scattered by small particles of
dust, molecules of water, and other gases in the atmosphere. The efficiency of a photon’s scattering depends on its wavelength. We
see the sky as blue during the day because violet and blue light scatter to a greater extent than other, longer wavelengths of light.
For the same reason, the sun appears red at sunset because red light is less efficiently scattered and is more likely to pass through
the atmosphere than other wavelengths of light. The scattering of radiation has been studied since the late 1800s, with applications
beginning soon thereafter. The earliest quantitative applications of scattering, which date from the early 1900s, used the elastic
scattering of light by colloidal suspensions to determine the concentration of colloidal particles.

Origin of Scattering
If we send a focused, monochromatic beam of radiation with a wavelength λ through a medium of particles with dimensions
< 1.5λ, the radiation scatters in all directions. For example, visible radiation of 500 nm is scattered by particles as large as 750 nm

in the longest dimension. Two general categories of scattering are recognized. In elastic scattering, radiation is first absorbed by the
particles and then emitted without undergoing a change in the radiation’s energy. When the radiation emerges with a change in
energy, the scattering is inelastic. Only elastic scattering is considered in this chapter.
Elastic scattering is divided into two types: Rayleigh, or small-particle scattering, and large-particle scattering. Rayleigh scattering
occurs when the scattering particle’s largest dimension is less than 5% of the radiation’s wavelength. The intensity of the scattered
radiation is proportional to its frequency to the fourth power, ν —which accounts for the greater scattering of blue light than red
4

light—and is distributed symmetrically (Figure 34.5.1 a). For larger particles, scattering increases in the forward direction and
decreases in the backward direction as the result of constructive and destructive interferences (Figure 34.5.1 b).

Figure 34.5.1 . Distribution of radiation for (a) Rayleigh, or small-particle scattering, and (b) large-particle scattering.

Rayleigh Scattering
Small particle, or Rayleigh scattering, measured at an angle of θ is the ratio of the intensity of the scattered light, I , to the intensity
of the light source, I , and is expressed as
o

I 6
Rθ = = Kr
Io

where r is the radius of the particle and K is a constant that is a function of the angle of scattering, the wavelength of light used,
0

the refractive index of the particle, and the distance to the particle, R .

Dynamic Light Scattering

In dynamic light scattering (DLS), we use a laser as a light source (see Figure 34.5.2 for an illustration). When the light from the
source reaches the sample, which is in a sample cell, it scatters in all directions, as shown in Figure 34.5.1.

34.5.1 https://chem.libretexts.org/@go/page/364543
 Figure 34.5.2 . Basic set-up for a DLS experiment. To avoid saturating the detector, the light from the laser is attenuated before
entering the sample cell. (CC BY-NC; Ümit Kaya via LibreTexts).
A detector is placed at a fixed angle to collect the light that scatters at that angle. The resulting intensity of scattered light is
measure as a function of time. Because the particles in the sample are moving due to Brownian motion, the intensity of light varies
with time yielding a noisy signal. Smaller particles diffuse more rapidly than larger particles, which means that fluctuations in
intensity with a small particle occurs more rapidly than for a large particle, as seen in Figure 34.5.3.

Figure 34.5.3 : The time-dependent intensity of scattered light for (top) large particles and for (bottom) small particles.
To process the data in DLS, we examine the correlation of the signal with itself over small increments of time. This is
accomplished by shifting the signal by a small amount (we call this the delay time, τ ) and computing the correlation between the
original signal and the delayed signal. For short delay times, the correlation in intensities is close to 1 because the particles have not
had time to move, and for longer delay times the correlation in intensities is close to 0 because the particles have moved
significantly; in between these limits, the correlation undergoes an exponential decay. Figure 34.5.4 shows examples of the
resulting correlograms for large particles and for small particles. The correlation function, G(τ ), is defined as
−2Γτ
G(τ ) = A[1 + Be ] (34.5.1)

The terms A and B are, respectively, the baseline and the intercept of the correlation function, and Γ = Dq , where D is the 2

translational diffusion coefficient and where q is equivalent to (4πn/λ )sin(θ/2) where n is the refractive index, λ is the
0 0

wavelength of the laser, and θ is the angle at which scattered light is collected. The relationship between the size of the particles
and the translational diffusion coefficient is give by the Stokes-Einstein equation

34.5.2 https://chem.libretexts.org/@go/page/364543
 kT
d =
3πηD

where k is Boltzmann's constant, T is the absolute temperature, and η is the viscosity. Fitting one or more equations for G(τ ) to
the correlogram yields the distribution of particle sizes.

Figure 34.5.4 : Correlograms for (a) large and for (b) small particles.

Static Light Scattering

In dynamic light scattering we are interested in how the intensity of scattering changes with time; in static light scattering, we are
interested in how the average intensity of scattered light varies with the concentration of particles, c , and the angle, θ , at which
scattering is measured. The extent of scattering, R , for each combination of c and θ is plotted as Kc/R , where K is a constant
θ θ

that is a function of the solvent's refractive index, the change in refracative index with concentration, and Avogadro's number as a
function of the angle; the value of S for the x-axis is chosen to maintain a separation between the data. A typical plot, which is
known as a Zimm plot, is shown in Figure 34.5.5.

Figure 34.5.5 . Zimm plot used to determine the size and molecular weight of particles. Data is collected using concentrations and
angles shown in brown. Extrapolating the data to a concentration of zero (dashed green lines) and a scattering angle of zero (dashed
blue lines) yields the solid blue and solid green lines that, in turn, yield values for the molecular weight, M , and particle size, R .
g

Each of the solid brown points gives the value of Kc/R for a combination of concentration and angle. For each angle, the change
θ

in Kc/R is extrapolated back to a concentration of zero (the dashed green lines) and for each concentration, the change in Kc/R
θ θ

is extrapolated back to an angle of zero (the dashed blue lines). The resulting extrapolation of c to θ = 0 gives an intercept that is
0

34.5.3 https://chem.libretexts.org/@go/page/364543
the inverse of the particles molecular weight, M , and the slope of the extrapolations of θ to a concentration of zero gives
0 Rg ,
which is the particle's radius of gyration (it is moving, after all), which is its effective particle size.

34.5: Measuring Particle Size Using Light Scattering is shared under a not declared license and was authored, remixed, and/or curated by
LibreTexts.

34.5.4 https://chem.libretexts.org/@go/page/364543
 CHAPTER OVERVIEW
35: Appendicies
The appendices gathered here provide a lengthy introduction to the analysis of data, a variety of tables that contain critical values
for the statistical analysis of data and standard oxidation-reduction potential, a discussion of activity, and a list of acronyms and
abbreviations used in this textbook and in other resources for analytical chemistry.
35.1: Evaluation of Analytical Data
35.2: Single-Sided Normal Distribution
35.3: Critical Values for t-Test
35.4: Critical Values for F-Test
35.5: Critical Values for Dixon's Q-Test
35.6: Critical Values for Grubb's Test
35.7: Activity Coefficients
35.8: Standard Reduction Potentials & Polarographic Half-wave Potentials
35.9: Recommended Primary Standards
35.10: Acronyms and Abbreviations

This page titled 35: Appendicies is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David Harvey.

1
35.1: Evaluation of Analytical Data
The material in this appendix is adapted from the textbook Chemometrics Using R, which is available through LibreTexts using
this link. In addition to the material here, the textbook contains instructions on how to use the statistical programming language R
to carry out the calculations.

Types of Data
At the heart of any analysis is data. Sometimes our data describes a category and sometimes it is numerical; sometimes our data
conveys order and sometimes it does not; sometimes our data has an absolute reference and sometimes it has an arbitrary reference;
and sometimes our data takes on discrete values and sometimes it takes on continuous values. Whatever its form, when we gather
data our intent is to extract from it information that can help us solve a problem.

Ways to Describe Data

If we are to consider how to describe data, then we need some data with which we can work. Ideally, we want data that is easy to
gather and easy to understand. It also is helpful if you can gather similar data on your own so you can repeat what we cover here. A
simple system that meets these criteria is to analyze the contents of bags of M&Ms. Although this system may seem trivial, keep in
mind that reporting the percentage of yellow M&Ms in a bag is analogous to reporting the concentration of Cu2+ in a sample of an
ore or water: both express the amount of an analyte present in a unit of its matrix.
At the beginning of this chapter we identified four contrasting ways to describe data: categorical vs. numerical, ordered vs.
unordered, absolute reference vs. arbitrary reference, and discrete vs. continuous. To give meaning to these descriptive terms, let’s
consider the data in Table 35.1.1, which includes the year the bag was purchased and analyzed, the weight listed on the package,
the type of M&Ms, the number of yellow M&Ms in the bag, the percentage of the M&Ms that were red, the total number of
M&Ms in the bag and their corresponding ranks.
Table 35.1.1 . Distribution of Yellow and Red M&Ms in Bags of M&Ms.
bag id year weight (oz) type number yellow % red total M&Ms rank (for total)

a 2006 1.74 peanut 2 27.8 18 sixth

b 2006 1.74 peanut 3 4.35 23 fourth

c 2000 0.80 plain 1 22.7 22 fifth

d 2000 0.80 plain 5 20.8 24 third

e 1994 10.0 plain 56 23.0 331 second

f 1994 10.0 plain 63 21.9 333 first

The entries in Table 35.1.1 are organized by column and by row. The first row—sometimes called the header row—identifies the
variables that make up the data. Each additional row is the record for one sample and each entry in a sample’s record provides
information about one of its variables; thus, the data in the table lists the result for each variable and for each sample.

Categorical vs. Numerical Data

Of the variables included in Table 35.1.1, some are categorical and some are numerical. A categorical variable provides qualitative
information that we can use to describe the samples relative to each other, or that we can use to organize the samples into groups
(or categories). For the data in Table 35.1.1, bag id, type, and rank are categorical variables.
A numerical variable provides quantitative information that we can use in a meaningful calculation; for example, we can use the
number of yellow M&Ms and the total number of M&Ms to calculate a new variable that reports the percentage of M&Ms that are
yellow. For the data in Table 35.1.1, year, weight (oz), number yellow, % red M&Ms, and total M&Ms are numerical variables.
We can also use a numerical variable to assign samples to groups. For example, we can divide the plain M&Ms in Table 35.1.1 into
two groups based on the sample’s weight. What makes a numerical variable more interesting, however, is that we can use it to
make quantitative comparisons between samples; thus, we can report that there are 14.4× as many plain M&Ms in a 10-oz. bag as
there are in a 0.8-oz. bag.

35.1.1 https://chem.libretexts.org/@go/page/363066
 333 + 331 664
= = 14.4
24 + 22 46

Although we could classify year as a categorical variable—not an unreasonable choice as it could serve as a useful way to group
samples—we list it here as a numerical variable because it can serve as a useful predictive variable in a regression analysis. On the
other hand rank is not a numerical variable—even if we rewrite the ranks as numerals—as there are no meaningful calculations we
can complete using this variable.

Nominal vs. Ordinal Data

Categorical variables are described as nominal or ordinal. A nominal categorical variable does not imply a particular order; an
ordinal categorical variable, on the other hand, coveys a meaningful sense of order. For the categorical variables in Table 35.1.1,
bag id and type are nominal variables, and rank is an ordinal variable.

Ratio vs. Interval Data

A numerical variable is described as either ratio or interval depending on whether it has (ratio) or does not have (interval) an
absolute reference. Although we can complete meaningful calculations using any numerical variable, the type of calculation we can
perform depends on whether or not the variable’s values have an absolute reference.
A numerical variable has an absolute reference if it has a meaningful zero—that is, a zero that means a measured quantity of none
—against which we reference all other measurements of that variable. For the numerical variables in Table 35.1.1, weight (oz),
number yellow, % red, and total M&Ms are ratio variables because each has a meaningful zero; year is an interval variable because
its scale is referenced to an arbitrary point in time, 1 BCE, and not to the beginning of time.
For a ratio variable, we can make meaningful absolute and relative comparisons between two results, but only meaningful absolute
comparisons for an interval variable. For example, consider sample e, which was collected in 1994 and has 331 M&Ms, and sample
d, which was collected in 2000 and has 24 M&Ms. We can report a meaningful absolute comparison for both variables: sample e is
six years older than sample d and sample e has 307 more M&Ms than sample d. We also can report a meaningful relative
comparison for the total number of M&Ms—there are
331
= 13.8×
24

as many M&Ms in sample e as in sample d—but we cannot report a meaningful relative comparison for year because a sample
collected in 2000 is not
2000
= 1.003×
1994

older than a sample collected in 1994.

Discrete vs. Continuous Data

Finally, the granularity of a numerical variable provides one more way to describe our data. For example, we can describe a
numerical variable as discrete or continuous. A numerical variable is discrete if it can take on only specific values—typically, but
not always, an integer value—between its limits; a continuous variable can take on any possible value within its limits. For the
numerical data in Table 35.1.1, year, number yellow, and total M&Ms are discrete in that each is limited to integer values. The
numerical variables weight (oz) and % red, on the other hand, are continuous variables. Note that weight is a continuous variable
even if the device we use to measure weight yields discrete values.

Visualizing Data
The old saying that "a picture is worth a 1000 words" may not be universally true, but it true when it comes to the analysis of data.
A good visualization of data, for example, allows us to see patterns and relationships that are less evident when we look at data
arranged in a table, and it provides a powerful way to tell our data's story. Suppose we want to study the composition of 1.69-oz
(47.9-g) packages of plain M&Ms. We obtain 30 bags of M&Ms (ten from each of three stores) and remove the M&Ms from each
bag one-by-one, recording the number of blue, brown, green, orange, red, and yellow M&Ms. We also record the number of yellow
M&Ms in the first five candies drawn from each bag, and record the actual net weight of the M&Ms in each bag. Table 35.1.2
summarizes the data collected on these samples. The bag id identifies the order in which the bags were opened and analyzed.
Table 35.1.2 . Analysis of Plain M&Ms in 47.9 g Bags.

35.1.2 https://chem.libretexts.org/@go/page/363066
 yellow_first
bag store blue brown green orange red yellow net_weight
_five

1 CVS 3 18 1 5 7 23 2 49.287

2 CVS 3 14 9 7 8 15 0 48.870

3 Target 4 14 5 10 10 16 1 51.250

4 Kroger 3 13 5 4 15 16 0 48.692

5 Kroger 3 16 5 7 8 18 1 48.777

6 Kroger 2 12 6 10 17 7 1 46.405

7 CVS 13 11 2 8 6 17 1 49.693

8 CVS 13 12 7 10 7 8 2 49.391

9 Kroger 6 17 5 4 8 16 1 48.196

10 Kroger 8 13 2 5 10 17 1 47.326

11 Target 9 20 1 4 12 13 3 50.974

12 Target 11 12 0 8 4 23 0 50.081

13 CVS 3 15 4 6 14 13 2 47.841

14 Kroger 4 17 5 6 14 10 2 48.377

15 Kroger 9 13 3 8 14 8 0 47.004

16 CVS 8 15 1 10 9 15 1 50.037

17 CVS 10 11 5 10 7 13 2 48.599

18 Kroger 1 17 6 7 11 14 1 48.625

19 Target 7 17 2 8 4 18 1 48.395

20 Kroger 9 13 1 8 7 22 1 51.730

21 Target 7 17 0 15 4 15 3 50.405

22 CVS 12 14 4 11 9 5 2 47.305

23 Target 9 19 0 5 12 12 0 49.477

24 Target 5 13 3 4 15 16 0 48.027

25 CVS 7 13 0 4 15 16 2 48.212

26 Target 6 15 1 13 10 14 1 51.682

27 CVS 5 17 6 4 8 19 1 50.802

28 Kroger 1 21 6 5 10 14 0 49.055

29 Target 4 12 6 5 13 14 2 46.577

30 Target 15 8 9 6 10 8 1 48.317

Having collected our data, we next examine it for possible problems, such as missing values (Did we forget to record the number of
brown M&Ms in any of our samples?), for errors introduced when we recorded the data (Is the decimal point recorded incorrectly
for any of the net weights?), or for unusual results (Is it really the case that this bag has only yellow M&M?). We also examine our
data to identify interesting observations that we may wish to explore (It appears that most net weights are greater than the net
weight listed on the individual packages. Why might this be? Is the difference significant?) When our data set is small we usually
can identify possible problems and interesting observations without much difficulty; however, for a large data set, this becomes a
challenge. Instead of trying to examine individual values, we can look at our results visually. While it may be difficult to find a
single, odd data point when we have to individually review 1000 samples, it often jumps out when we look at the data using one or
more of the approaches we will explore in this chapter.

35.1.3 https://chem.libretexts.org/@go/page/363066
Dot Plots
A dot plot displays data for one variable, with each sample’s value plotted on the x-axis. The individual points are organized along
the y-axis with the first sample at the bottom and the last sample at the top. Figure 35.1.1 shows a dot plot for the number of brown
M&Ms in the 30 bags of M&Ms from Table 35.1.2. The distribution of points appears random as there is no correlation between
the sample id and the number of brown M&Ms. We would be surprised if we discovered that the points were arranged from the
lower-left to the upper-right as this implies that the order in which we open the bags determines whether they have many or a few
brown M&Ms.

Figure 35.1.1 : Dot plot for the brown M&Ms in each of the 30 bags included in Table 35.1.2 .

Stripcharts
A dot plot provides a quick way to give us confidence that our data are free from unusual patterns, but at the cost of space because
we use the y-axis to include the sample id as a variable. A stripchart uses the same x-axis as a dot plot, but does not use the y-axis to
distinguish between samples. Because all samples with the same number of brown M&Ms will appear in the same place—making
it impossible to distinguish them from each other—we stack the points vertically to spread them out, as shown in Figure 35.1.2.

35.1.4 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.2 : Stripchart for the brown M&Ms in each of the 30 bags included in Table 35.1.2 .
Both the dot plot in Figure 35.1.1 and the stripchart in Figure 35.1.2 suggest that there is a smaller density of points at the lower
limit and the upper limit of our results. We see, for example, that there is just one bag each with 8, 16, 18, 19, 20, and 21 brown
M&Ms, but there are six bags each with 13 and 17 brown M&Ms.
Because a stripchart does not use the y-axis to provide meaningful categorical information, we can easily display several stripcharts
at once. Figure 35.1.3 shows this for the data in Table 35.1.2. Instead of stacking the individual points, we jitter them by applying a
small, random offset to each point. Among the things we learn from this stripchart are that only brown and yellow M&Ms have
counts of greater than 20 and that only blue and green M&Ms have counts of three or fewer M&Ms.

35.1.5 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.3 : Stripcharts for each color of M&Ms in each of the 30 bags included in Table 35.1.2 .

Box and Whisker Plots

The stripchart in Figure 35.1.3 is easy for us to examine because the number of samples, 30 bags, and the number of M&Ms per
bag is sufficiently small that we can see the individual points. As the density of points becomes greater, a stripchart becomes less
useful. A box and whisker plot provides a similar view but focuses on the data in terms of the range of values that encompass the
middle 50% of the data.
Figure 35.1.4 shows the box and whisker plot for brown M&Ms using the data in Table 35.1.2. The 30 individual samples are
superimposed as a stripchart. The central box divides the x-axis into three regions: bags with fewer than 13 brown M&Ms (seven
samples), bags with between 13 and 17 brown M&Ms (19 samples), and bags with more than 17 brown M&Ms (four samples).
The box's limits are set so that it includes at least the middle 50% of our data. In this case, the box contains 19 of the 30 samples
(63%) of the bags, because moving either end of the box toward the middle results in a box that includes less than 50% of the
samples. The difference between the box's upper limit (19) and its lower limit (13) is called the interquartile range (IQR). The thick
line in the box is the median, or middle value (more on this and the IQR in the next chapter). The dashed lines at either end of the
box are called whiskers, and they extend to the largest or the smallest result that is within ±1.5 × IQR of the box's right or left
edge, respectively.

35.1.6 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.4 : Box-and-whisker plot for the brown M&Ms in each of the 30 bags included in Table 35.1.2 showing individual
samples as a jittered stripchart.
Because a box and whisker plot does not use the y-axis to provide meaningful categorical information, we can easily display
several plots in the same frame. Figure 35.1.5 shows this for the data in Table 35.1.2. Note that when a value falls outside of a
whisker, as is the case here for yellow M&Ms, it is flagged by displaying it as an open circle.

Figure 35.1.5 : Box-and-whisker plots for each of the 30 bags included in Table 35.1.2 organized by color.
One use of a box and whisker plot is to examine the distribution of the individual samples, particularly with respect to symmetry.
With the exception of the single sample that falls outside of the whiskers, the distribution of yellow M&Ms appears symmetrical:
the median is near the center of the box and the whiskers extend equally in both directions. The distribution of the orange M&Ms is
asymmetrical: half of the samples have 4–7 M&Ms (just four possible outcomes) and half have 7–15 M&Ms (nine possible
outcomes), suggesting that the distribution is skewed toward higher numbers of orange M&Ms (see Chapter 5 for more information
about the distribution of samples).

35.1.7 https://chem.libretexts.org/@go/page/363066
Figure 35.1.6 shows box-and-whisker plots for yellow M&Ms grouped according to the store where the bags of M&Ms were
purchased. Although the box and whisker plots are quite different in terms of the relative sizes of the boxes and the relative length
of the whiskers, the dot plots suggest that the distribution of the underlying data is relatively similar in that most bags contain 12–
18 yellow M&Ms and just a few bags deviate from these limits. These observations are reassuring because we do not expect the
choice of store to affect the composition of bags of M&Ms. If we saw evidence that the choice of store affected our results, then we
would look more closely at the bags themselves for evidence of a poorly controlled variable, such as type (Did we accidentally
purchase bags of peanut butter M&Ms from one store?) or the product’s lot number (Did the manufacturer change the composition
of colors between lots?).

Figure 35.1.6 : Box-and-whisker plots for yellow M&Ms for each of the 30 bag in Table 35.1.2 organized by the store where the
bags were purchased.

Bar Plots
Although a dot plot, a stripchart and a box-and-whisker plot provide some qualitative evidence of how a variable’s values are
distributed—we will have more to say about the distribution of data in Chapter 5—they are less useful when we need a more
quantitative picture of the distribution. For this we can use a bar plot that displays a count of each discrete outcome. Figure 35.1.7
shows bar plots for orange and for yellow M&Ms using the data in Table 35.1.2.

35.1.8 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.7 : Bar plots for orange M&Ms and yellow M&Ms using the data in Table 35.1.2 .
Here we see that the most common number of orange M&Ms per bag is four, which is also the smallest number of orange M&Ms
per bag, and that there is a general decrease in the number of bags as the number of orange M&M per bag increases. For the yellow
M&Ms, the most common number of M&Ms per bag is 16, which falls near the middle of the range of yellow M&Ms.

Histograms
A bar plot is a useful way to look at the distribution of discrete results, such as the counts of orange or yellow M&Ms, but it is not
useful for continuous data where each result is unique. A histogram, in which we display the number of results that fall within a
sequence of equally spaced bins, provides a view that is similar to that of a bar plot but that works with continuous data. Figure
35.1.8, for example, shows a histogram for the net weights of the 30 bags of M&Ms in Table 35.1.2. Individual values are shown

by the vertical hash marks at the bottom of the histogram.

Figure 35.1.8 : Histogram of net weights for the data in Table 35.1.2 . There are, for example, four bags of M&Ms with net weights
between 47 g and 48 g.

35.1.9 https://chem.libretexts.org/@go/page/363066
Summarizing Data
In the last section we used data collected from 30 bags of M&Ms to explore different ways to visualize data. In this section we
consider several ways to summarize data using the net weights of the same bags of M&Ms. Here is the raw data.
Table 35.1.3 : Net Weights for 30 Bags of M&Ms.
49.287 48.870 51.250 48.692 48.777 46.405

49.693 49.391 48.196 47.326 50.974 50.081

47.841 48.377 47.004 50.037 48.599 48.625

48.395 51.730 50.405 47.305 49.477 48.027

48.212 51.682 50.802 49.055 46.577 48.317

Without completing any calculations, what conclusions can we make by just looking at this data? Here are a few:
All net weights are greater than 46 g and less than 52 g.
As we see in Figure 35.1.9, a box-and-whisker plot (overlaid with a stripchart) and a histogram suggest that the distribution of
the net weights is reasonably symmetric.
The absence of any points beyond the whiskers of the box-and-whisker plot suggests that there are no unusually large or
unsually small net weights.

Figure 35.1.9 : Two visualizations of the net weights of packages of M&Ms.

Both visualizations provide a good qualitative picture of the data, suggesting that the individual results are scattered around some
central value with more results closer to that central value that at distance from it. Neither visualization, however, describes the data
quantitatively. What we need is a convenient way to summarize the data by reporting where the data is centered and how varied the
individual results are around that center.

Where is the Center?

There are two common ways to report the center of a data set: the mean and the median.
The mean, Y , is the numerical average obtained by adding together the results for all n observations and dividing by the number of
¯
¯¯¯

observations
n

¯
¯¯¯
∑i=1 Yi 49.287 + 48.870 + ⋯ + 48.317
Y = = = 48.980 g
n 30

35.1.10 https://chem.libretexts.org/@go/page/363066
The median, Y˜ , is the middle value after we order our observations from smallest-to-largest, as we show here for our data.
Table 35.1.4 : The data from Table 35.1.3 Sorted From Smallest-to-Largest in Value.
46.405 46.577 47.004 47.305 47.326 47.841

48.027 48.196 48.212 48.317 48.377 48.395

48.599 48.625 48.692 48.777 48.870 49.055

49.287 49.391 49.477 49.693 50.037 50.081

50.405 50.802 50.974 51.250 51.682 51.730

If we have an odd number of samples, then the median is simply the middle value, or
˜ =Y
Y n+1

where n is the number of samples. If, as is the case here, n is even, then
Y n
+Y n
+1
2 2
48.692 + 48.777
˜ =
Y = = 48.734 g
2 2

When our data has a symmetrical distribution, as we believe is the case here, then the mean and the median will have similar
values.

What is the Variation of the Data About the Center?

There are five common measures of the variation of data about its center: the variance, the standard deviation, the range, the
interquartile range, and the median average difference.
The variance, s2, is an average squared deviation of the individual observations relative to the mean
n ¯
¯¯¯ 2 2 2
∑ (Yi − Y ) (49.287 − 48.980 ) + ⋯ + (48.317 − 48.980 )
2 i=1
s = = = 2.052
n−1 30 − 1

and the standard deviation, s, is the square root of the variance, which gives it the same units as the mean.
−−−−−−−−−−−−− −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−
n ¯
¯¯¯ 2 2 2
∑ (Yi − Y ) (49.287 − 48.980 ) + ⋯ + (48.317 − 48.980 )
i=1
√ √
s = = = 1.432
n−1 30 − 1

The range, w, is the difference between the largest and the smallest value in our data set.

w = 51.730 g − 46.405 g = 5.325 g

The interquartile range, IQR, is the difference between the median of the bottom 25% of observations and the median of the top
25% of observations; that is, it provides a measure of the range of values that spans the middle 50% of observations. There is no
single, standard formula for calculating the IQR, and different algorithms yield slightly different results. We will adopt the
algorithm described here:
1. Divide the sorted data set in half; if there is an odd number of values, then remove the median for the complete data set. For our
data, the lower half is
Table 35.1.5 : The Lower Half of the Data in Table 35.1.4 .
46.405 46.577 47.004 47.305 47.326

47.841 48.027 48.196 48.212 48.317

48.377 48.395 48.599 48.625 48.692

and the upper half is

Table 35.1.6 : The Upper Half of the Data in Table 35.1.4 .
48.777 48.870 49.055 49.287 49.391

49.477 49.693 50.037 50.081 50.405

35.1.11 https://chem.libretexts.org/@go/page/363066
 50.802 50.974 51.250 51.682 51.730

2. Find FL, the median for the lower half of the data, which for our data is 48.196 g.
3. Find FU , the median for the upper half of the data, which for our data is 50.037 g.
4. The IQR is the difference between FU and FL.

FU − FL = 50.037 g − 48.196 g = 1.841 g

The median absolute deviation, MAD, is the median of the absolute deviations of each observation from the median of all
observations. To find the MAD for our set of 30 net weights, we first subtract the median from each sample in Table 35.1.3.
Table 35.1.7 : The Results of Subtracting the Median From Each Value in Table 35.1.3 .
0.5525 0.1355 2.5155 -0.0425 0.0425 -2.3295

0.9585 0.6565 -0.5385 -1.4085 2.2395 1.3465

-0.8935 -0.3575 -1.7305 1.3025 -0.1355 -0.1095

-0.3395 2.9955 1.6705 -1.4295 0.7425 -0.7075

-0.5225 2.9475 2.0675 0.3205 -2.1575 -0.4175

Next we take the absolute value of each difference and sort them from smallest-to-largest.
Table 35.1.8 : The Data in Table 35.1.7 After Taking the Absolute Value.
0.0425 0.0425 0.1095 0.1355 0.1355 0.3205

0.3395 0.3575 0.4175 0.5225 0.5385 0.5525

0.6565 0.7075 0.7425 0.8935 0.9585 1.3025

1.3465 1.4085 1.4295 1.6705 1.7305 2.0675

2.1575 2.2395 2.3295 2.5155 2.9475 2.9955

Finally, we report the median for these sorted values as

0.7425 + 0.8935
= 0.818
2

Robust vs. Non-Robust Measures of The Center and Variation About the Center
A good question to ask is why we might desire more than one way to report the center of our data and the variation in our data
about the center. Suppose that the result for the last of our 30 samples was reported as 483.17 instead of 48.317. Whether this is an
accidental shifting of the decimal point or a true result is not relevant to us here; what matters is its effect on what we report. Here
is a summary of the effect of this one value on each of our ways of summarizing our data.
Table 35.1.9 : Effect on Summary Statistics of Changing Last Value in Table 35.1.3 From 48.317 g to 483.17 g.
statistic original data new data

mean 48.980 63.475

median 48.734 48.824

variance 2.052 6285.938

standard deviation 1.433 79.280

range 5.325 436.765

IQR 1.841 1.885

MAD 0.818 0.926

Note that the mean, the variance, the standard deviation, and the range are very sensitive to the change in the last result, but the
median, the IQR, and the MAD are not. The median, the IQR, and the MAD are considered robust statistics because they are less

35.1.12 https://chem.libretexts.org/@go/page/363066
sensitive to an unusual result; the others are, of course, non-robust statistics. Both types of statistics have value to us, a point we
will return to from time-to-time.

The Distribution of Data

When we measure something, such as the percentage of yellow M&Ms in a bag of M&Ms, we expect two things:
that there is an underlying “true” value that our measurements should approximate, and
that the results of individual measurements will show some variation about that "true" value
Visualizations of data—such as dot plots, stripcharts, boxplot-and-whisker plots, bar plots, histograms, and scatterplots—often
suggest there is an underlying structure to our data. For example, we have seen that the distribution of yellow M&Ms in bags of
M&Ms is more or less symmetrical around its median, while the distribution of orange M&Ms was skewed toward higher values.
This underlying structure, or distribution, of our data as it effects how we choose to analyze our data. In this chapter we will take a
closer look at several ways in which data are distributed.

Terminology
Before we consider different types of distributions, let's define some key terms. You may wish, as well, to review the discussion of
different types of data in Chapter 2.
Populations and Samples
A population includes every possible measurement we could make on a system, while a sample is the subset of a population on
which we actually make measurements. These definitions are fluid. A single bag of M&Ms is a population if we are interested only
in that specific bag, but it is but one sample from a box that contains a gross (144) of individual bags. That box, itself, can be a
population, or it can be one sample from a much larger production lot. And so on.
Discrete Distributions and Continuous Distributions
In a discrete distribution the possible results take on a limited set of specific values that are independent of how we make our
measurements. When we determine the number of yellow M&Ms in a bag, the results are limited to integer values. We may find 13
yellow M&Ms or 24 yellow M&Ms, but we cannot obtain a result of 15.43 yellow M&Ms.
For a continuous distribution the result of a measurement can take on any possible value between a lower limit and an upper limit,
even though our measuring device has a limited precision; thus, when we weigh a bag of M&Ms on a three-digit balance and
obtain a result of 49.287 g we know that its true mass is greater than 49.2865... g and less than 49.2875... g.

Theoretical Models for the Distribution of Data

There are four important types of distributions that we will consider in this chapter: the uniform distribution, the binomial
distribution, the Poisson distribution, and the normal, or Gaussian, distribution. In the previous sections we used the analysis of
bags of M&Ms to explore ways to visualize data and to summarize data. Here we will use the same data set to explore the
distribution of data.
Uniform Distribution

In a uniform distribution, all outcomes are equally probable. Suppose the population of M&Ms has a uniform distribution. If this is
the case, then, with six colors, we expect each color to appear with a probability of 1/6 or 16.7%. Figure 35.1.10 shows a
comparison of the theoretical results if we draw 1699 M&Ms—the total number of M&Ms in our sample of 30 bags—from a
population with a uniform distribution (on the left) to the actual distribution of the 1699 M&Ms in our sample (on the right). It
seems unlikely that the population of M&Ms has a uniform distribution of colors!

35.1.13 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.10: Comparison of (on the left) a uniform distribution of 1699 M&Ms with (on the right) the actual distribution from
the data that make up the sample in Table P ageIndex2 .
Binomial Distribution

A binomial distribution shows the probability of obtaining a particular result in a fixed number of trials, where the odds of that
result happening in a single trial are known. Mathematically, a binomial distribution is defined by the equation
N!
X N −X
P (X, N ) = ×p × (1 − p )
X!(N − X)!

where P(X,N) is the probability that the event happens X times in N trials, and where p is the probability that the event happens in a
single trial. The binomial distribution has a theoretical mean, μ , and a theoretical variance, σ , of
2

2
μ = Np σ = N p(1 − p)

Figure 35.1.11 compares the expected binomial distribution for drawing 0, 1, 2, 3, 4, or 5 yellow M&Ms in the first five M&Ms—
assuming that the probability of drawing a yellow M&M is 435/1699, the ratio of the number of yellow M&Ms and the total
number of M&Ms—to the actual distribution of results. The similarity between the theoretical and the actual results seems evident;
in a later section we will consider ways to test this claim.

35.1.14 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.11: Comparison of (on the left) the theoretical binomial distribution of yellow M&Ms in the first five selected from a
bag of M&Ms and (on the right) the actual distribution of M&Ms.
Poisson Distribution

The binomial distribution is useful if we wish to model the probability of finding a fixed number of yellow M&Ms in a sample of
M&Ms of fixed size—such as the first five M&Ms that we draw from a bag—but not the probability of finding a fixed number of
yellow M&Ms in a single bag because there is some variability in the total number of M&Ms per bag.
A Poisson distribution gives the probability that a given number of events will occur in a fixed interval in time or space if the event
has a known average rate and if each new event is independent of the preceding event. Mathematically a Poisson distribution is
defined by the equation
−λ X
e λ
P (X, λ) =
X!

where P (X, λ) is the probability that an event happens X times given the event’s average rate, λ . The Poisson distribution has a
theoretical mean, μ , and a theoretical variance, σ , that are each equal to λ .
2

The bar plot in Figure 35.1.12 shows the actual distribution of green M&Ms in 35 small bags of M&Ms (as reported by M. A. Xu-
Friedman “Illustrating concepts of quantal analysis with an intuitive classroom model,” Adv. Physiol. Educ. 2013, 37, 112–116).
Superimposed on the bar plot is the theoretical Poisson distribution based on their reported average rate of 3.4 green M&Ms per
bag. The similarity between the theoretical and the actual results seems evident; in Chapter 6 we will consider ways to test this
claim.

35.1.15 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.12: Comparison of a Poisson distribution for green M&Ms (dots and line) to experimental results (bars). The data are
from M. A. Xu-Friedman, “Illustrating concepts of quantal analysis with an intuitive classroom model,” Adv. Physiol. Educ. 2013,
37, 112–116.
Normal Distribution

A uniform distribution, a binomial distribution, and a Poisson distribution predict the probability of a discrete event, such as the
probability of finding exactly two green M&Ms in the next bag of M&Ms that we open. Not all of the data we collect is discrete.
The net weights of bags of M&Ms is an example of continuous data as the mass of an individual bag is not restricted to a discrete
set of allowed values. In many cases we can model continuous data using a normal (or Gaussian) distribution, which gives the
probability of obtaining a particular outcome, P(x), from a population with a known mean, μ , and a known variance, σ . 2

Mathematically a normal distribution is defined by the equation

1 2 2
−(x−μ ) /(2 σ )
P (x) = e
−−−−
√2πσ 2

Figure 35.1.13 shows the expected normal distribution for the net weights of our sample of 30 bags of M&Ms if we assume that
their mean, X , of 48.98 g and standard deviation, s, of 1.433 g are good predictors of the population’s mean, μ , and standard
¯¯¯
¯

deviation, σ. Given the small sample of 30 bags, the agreement between the model and the data seems reasonable.

35.1.16 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.13: Comparison of a normal distribution for the net weights of M&Ms (line) to the experimental results (bars).

The Central Limit Theorem

Suppose we have a population for which one of its properties has a uniform distribution where every result between 0 and 1 is
equally probable. If we analyze 10,000 samples we should not be surprised to find that the distribution of these 10000 results looks
uniform, as shown by the histogram on the left side of Figure 35.1.14. If we collect 1000 pooled samples—each of which consists
of 10 individual samples for a total of 10,000 individual samples—and report the average results for these 1000 pooled samples, we
see something interesting as their distribution, as shown by the histogram on the right, looks remarkably like a normal distribution.
When we draw single samples from a uniform distribution, each possible outcome is equally likely, which is why we see the
distribution on the left. When we draw a pooled sample that consists of 10 individual samples, however, the average values are
more likely to be near the middle of the distribution’s range, as we see on the right, because the pooled sample likely includes
values drawn from both the lower half and the upper half of the uniform distribution.

35.1.17 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.14: Distribution of results when analyzing samples of size n = 1 (left) and samples of size n = 10 (right) drawn from a
uniform distribution.
This tendency for a normal distribution to emerge when we pool samples is known as the central limit theorem. As shown in Figure
35.1.15, we see a similar effect with populations that follow a binomial distribution or a Poisson distribution.

Figure 35.1.15: Distribution of results when analyzing samples of size n = 1 (left) and samples of size n = 10 (right) drawn from
a binomial distribution with p = 0.167 (top) and a Poisson distribution with λ = 4 (bottom).
You might reasonably ask whether the central limit theorem is important as it is unlikely that we will complete 1000 analyses, each
of which is the average of 10 individual trials. This is deceiving. When we acquire a sample of soil, for example, it consists of
many individual particles each of which is an individual sample of the soil. Our analysis of this sample, therefore, is the mean for a
large number of individual soil particles. Because of this, the central limit theorem is relevant.

35.1.18 https://chem.libretexts.org/@go/page/363066
Uncertainty of Data
In the last section we examined four ways in which the individual samples we collect and analyze are distributed about a central
value: a uniform distribution, a binomial distribution, a Poisson distribution, and a normal distribution. We also learned that
regardless of how individual samples are distributed, the distribution of averages for multiple samples often follows a normal
distribution. This tendency for a normal distribution to emerge when we report averages for multiple samples is known as the
central limit theorem. In this chapter we look more closely at the normal distribution—examining some of its properties—and
consider how we can use these properties to say something more meaningful about our data than simply reporting a mean and a
standard deviation.

Properties of a Normal Distribution

Mathematically a normal distribution is defined by the equation
1 2 2
−(x−μ ) /(2 σ )
P (x) = −− −−e
√2πσ 2

where P (x) is the probability of obtaining a result, x, from a population with a known mean, μ , and a known standard deviation,
σ. Figure 35.1.16 shows the normal distribution curves for μ = 0 with standard deviations of 5, 10, and 20.

Figure 35.1.16: Three examples of normal distribution curves. Although the height and width are affected by σ , the area under
each curve is the same.
Because the equation for a normal distribution depends solely on the population’s mean, μ , and its standard deviation, σ, the
probability that a sample drawn from a population has a value between any two arbitrary limits is the same for all populations. For
example, Figure 35.1.17 shows that 68.26% of all samples drawn from a normally distributed population have values within the
range μ ± 1σ , and only 0.14% have values greater than μ + 3σ .

35.1.19 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.17: Normal distribution curve for μ = 0 and σ = 1 showing area under the curve for various values of z in μ ± zσ .
This feature of a normal distribution—that the area under the curve is the same for all values of σ—allows us to create a probability
table (see Appendix 2) based on the relative deviation, z , between a limit, x, and the mean, μ .
x −μ
z =
σ

The value of z gives the area under the curve between that limit and the distribution’s closest tail, as shown in Figure 35.1.18.

Figure 35.1.18: Normal distribution curve for μ = 0 and σ = 1 showing (on the left) the area under the curve for z = −1.5 and
(on the right for z = +0.5 .

35.1.20 https://chem.libretexts.org/@go/page/363066
 Example 35.1.1

Suppose we know that μ is 5.5833 ppb Pb and that σ is 0.0558 ppb Pb for a particular standard reference material (SRM).
What is the probability that we will obtain a result that is greater than 5.650 ppb if we analyze a single, random sample drawn
from the SRM?

Solution
Figure 35.1.19 shows the normal distribution curve given values of 5.5833 ppb Pb for μ and of 0.0558 ppb Pb σ. The shaded
area in the figures is the probability of obtaining a sample with a concentration of Pb greater than 5.650 ppm. To determine the
probability, we first calculate z
x −μ 5.650 − 5.5833
z = = = 1.195
σ 0.0558

Next, we look up the probability in Appendix 2 for this value of z , which is the average of 0.1170 (for z = 1.19 ) and 0.1151
(for z = 1.20 ), or a probability of 0.1160; thus, we expect that 11.60% of samples will provide a result greater than 5.650 ppb
Pb.

Figure 35.1.19: Normal distribution curve for the amount of lead in a standard reference with μ = 5.5833 ppb and σ = 0.0558

ppb. The shaded area shows those results for which the concentration of lead exceeds 5.650 ppb.

 Example 35.1.2

Example 35.1.1 considers a single limit—the probability that a result exceeds a single value. But what if we want to determine
the probability that a sample has between 5.580 g Pb and 5.625 g Pb?

Solution
In this case we are interested in the shaded area shown in Figure 35.1.20. First, we calculate z for the upper limit
5.625 − 5.5833
z = = 0.747
0.0558

and then we calculate z for the lower limit

35.1.21 https://chem.libretexts.org/@go/page/363066
 5.580 − 5.5833
z = = −0.059
0.0558

Then, we look up the probability in Appendix 2 that a result will exceed our upper limit of 5.625, which is 0.2275, or 22.75%,
and the probability that a result will be less than our lower limit of 5.580, which is 0.4765, or 47.65%. The total unshaded area
is 71.4% of the total area, so the shaded area corresponds to a probability of

100.00 − 22.75 − 47.65 = 100.00 − 71.40 = 29.6%

Figure 35.1.20: Normal distribution curve for the amount of lead in a standard reference with μ = 5.5833 ppb and σ = 0.0558
ppb. The shaded area shows those results for which the concentration of lead is more than 5.580 ppb and less than 5.625 ppb.

Confidence Intervals
In the previous section, we learned how to predict the probability of obtaining a particular outcome if our data are normally
distributed with a known μ and a known σ. For example, we estimated that 11.60% of samples drawn at random from a standard
reference material will have a concentration of Pb greater than 5.650 ppb given a μ of 5.5833 ppb and a σ of 0.0558 ppb. In
essence, we determined how many standard deviations 5.650 is from μ and used this to define the probability given the standard
area under a normal distribution curve.
We can look at this in a different way by asking the following question: If we collect a single sample at random from a population
with a known μ and a known σ, within what range of values might we reasonably expect to find the sample’s result 95% of the
time? Rearranging the equation
x −μ
z =
σ

and solving for x gives

x = μ ± zσ = 5.5833 ± (1.96)(0.0558) = 5.5833 ± 0.1094

where a z of 1.96 corresponds to 95% of the area under the curve; we call this a 95% confidence interval for a single sample.
It generally is a poor idea to draw a conclusion from the result of a single experiment; instead, we usually collect several samples
and ask the question this way: If we collect n random samples from a population with a known μ and a known σ, within what
range of values might we reasonably expect to find the mean of these samples 95% of the time?
We might reasonably expect that the standard deviation for the mean of several samples is smaller than the standard deviation for a
set of individual samples; indeed it is and it is given as

35.1.22 https://chem.libretexts.org/@go/page/363066
 σ
σ¯ =
x −
√n

where σ
is called the standard error of the mean. For example, if we collect three samples from the standard reference material
√n

described above, then we expect that the mean for these three samples will fall within a range

zσ (1.96)(0.0558)
x̄ = μ ± zσ ¯ = μ± − = 5.5833 ± – = 5.5833 ± 0.0631
X
√n √3

that is ±0.0631 ppb around μ , a range that is smaller than that of ±0.1094 ppb when we analyze individual samples. Note that the
relative value to us of increasing the sample’s size diminishes as n increases because of the square root term, as shown in Figure
35.1.21.

Figure 35.1.21: Plot showing how the standard error of the mean varies with the size of the sample. The value for σ is 10.
Our treatment thus far assumes we know μ and σ for the parent population, but we rarely know these values; instead, we examine
samples drawn from the parent population and ask the following question: Given the sample’s mean, x̄, and its standard deviation,
s , what is our best estimate of the population’s mean, μ , and its standard deviation, σ.

To make this estimate, we replace the population’s standard deviation, σ, with the standard deviation, s , for our samples, replace
the population’s mean, μ , with the mean, x̄, for our samples, replace z with t , where the value of t depends on the number of
samples, n
ts
x̄ = μ ± −
√n

and then rearrange the equation to solve for μ .

ts
μ = x̄ ± −
√n

We call this a confidence interval. Values for t are available in tables (see Appendix 3) and depend on the probability level, α ,
where (1 − α) × 100 is the confidence level, and the degrees of freedom, n − 1 ; note that for any probability level, t ⟶ z as
n ⟶ ∞.

We need to give special attention to what this confidence interval means and to what it does not mean:
It does not mean that there is a 95% probability that the population’s mean is in the range μ = x̄ ± ts because our
measurements may be biased or the normal distribution may be inappropriate for our system.

35.1.23 https://chem.libretexts.org/@go/page/363066
 It does provide our best estimate of the population’s mean, μ given our analysis of n samples drawn at random from the parent
population; a different sample, however, will give a different confidence interval and, therefore, a different estimate for μ .

Testing the Significance of Data

A confidence interval is a useful way to report the result of an analysis because it sets limits on the expected result. In the absence
of determinate error, or bias, a confidence interval based on a sample’s mean indicates the range of values in which we expect to
find the population’s mean. When we report a 95% confidence interval for the mass of a penny as 3.117 g ± 0.047 g, for example,
we are stating that there is only a 5% probability that the penny’s expected mass is less than 3.070 g or more than 3.164 g.
Because a confidence interval is a statement of probability, it allows us to consider comparative questions, such as these:
“Are the results for a newly developed method to determine cholesterol in blood significantly different from those obtained using a
standard method?”
“Is there a significant variation in the composition of rainwater collected at different sites downwind from a coal-burning utility
plant?”
In this chapter we introduce a general approach that uses experimental data to ask and answer such questions, an approach we call
significance testing.
The reliability of significance testing recently has received much attention—see Nuzzo, R. “Scientific Method: Statistical Errors,”
Nature, 2014, 506, 150–152 for a general discussion of the issues—so it is appropriate to begin this chapter by noting the need to
ensure that our data and our research question are compatible so that we do not read more into a statistical analysis than our data
allows; see Leek, J. T.; Peng, R. D. “What is the Question? Science, 2015, 347, 1314-1315 for a useful discussion of six common
research questions.
In the context of analytical chemistry, significance testing often accompanies an exploratory data analysis
"Is there a reason to suspect that there is a difference between these two analytical methods when applied to a common sample?"
or an inferential data analysis.
"Is there a reason to suspect that there is a relationship between these two independent measurements?"
A statistically significant result for these types of analytical research questions generally leads to the design of additional
experiments that are better suited to making predictions or to explaining an underlying causal relationship. A significance test is the
first step toward building a greater understanding of an analytical problem, not the final answer to that problem!

Significance Testing
Let’s consider the following problem. To determine if a medication is effective in lowering blood glucose concentrations, we
collect two sets of blood samples from a patient. We collect one set of samples immediately before we administer the medication,
and we collect the second set of samples several hours later. After we analyze the samples, we report their respective means and
variances. How do we decide if the medication was successful in lowering the patient’s concentration of blood glucose?
One way to answer this question is to construct a normal distribution curve for each sample, and to compare the two curves to each
other. Three possible outcomes are shown in Figure 35.1.22. In Figure 35.1.22a, there is a complete separation of the two normal
distribution curves, which suggests the two samples are significantly different from each other. In Figure 35.1.22b, the normal
distribution curves for the two samples almost completely overlap each other, which suggests the difference between the samples is
insignificant. Figure 35.1.22c, however, presents us with a dilemma. Although the means for the two samples seem different, the
overlap of their normal distribution curves suggests that a significant number of possible outcomes could belong to either
distribution. In this case the best we can do is to make a statement about the probability that the samples are significantly different
from each other.

35.1.24 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.22: Three examples of the possible relationships between the normal distribution curves for two samples. In (a) the
curves do not overlap, which suggests the samples are significantly different from each other. In (b) the two curves are almost
identical, suggesting the samples are indistinguishable. The partial overlap of the curves in (c) means that the best we can do is
evaluate the probability that there is a difference between the samples.
The process by which we determine the probability that there is a significant difference between two samples is called significance
testing or hypothesis testing. Before we discuss specific examples let's first establish a general approach to conducting and
interpreting a significance test.
Constructing a Significance Test

The purpose of a significance test is to determine whether the difference between two or more results is sufficiently large that we
are comfortable stating that the difference cannot be explained by indeterminate errors. The first step in constructing a significance
test is to state the problem as a yes or no question, such as
“Is this medication effective at lowering a patient’s blood glucose levels?”
A null hypothesis and an alternative hypothesis define the two possible answers to our yes or no question. The null hypothesis, H0,
is that indeterminate errors are sufficient to explain any differences between our results. The alternative hypothesis, HA, is that the
differences in our results are too great to be explained by random error and that they must be determinate in nature. We test the null
hypothesis, which we either retain or reject. If we reject the null hypothesis, then we must accept the alternative hypothesis and
conclude that the difference is significant.
Failing to reject a null hypothesis is not the same as accepting it. We retain a null hypothesis because we have insufficient evidence
to prove it incorrect. It is impossible to prove that a null hypothesis is true. This is an important point and one that is easy to forget.
To appreciate this point let’s use this data for the mass of 100 circulating United States pennies.
Table 35.1.10. Masses for a Sample of 100 Circulating U. S. Pennies
Penny Weight (g) Penny Weight (g) Penny Weight (g) Penny Weight (g)

1 3.126 26 3.073 51 3.101 76 3.086

2 3.140 27 3.084 52 3.049 77 3.123

3 3.092 28 3.148 53 3.082 78 3.115

4 3.095 29 3.047 54 3.142 79 3.055

5 3.080 30 3.121 55 3.082 80 3.057

6 3.065 31 3.116 56 3.066 81 3.097

7 3.117 32 3.005 57 3.128 82 3.066

8 3.034 33 3.115 58 3.112 83 3.113

9 3.126 34 3.103 59 3.085 84 3.102

10 3.057 35 3.086 60 3.086 85 3.033

11 3.053 36 3.103 61 3.084 86 3.112

12 3.099 37 3.049 62 3.104 87 3.103

13 3.065 38 2.998 63 3.107 88 3.198

14 3.059 39 3.063 64 3.093 89 3.103

15 3.068 40 3.055 65 3.126 90 3.126

35.1.25 https://chem.libretexts.org/@go/page/363066
 16 3.060 41 3.181 66 3.138 91 3.111

17 3.078 42 3.108 67 3.131 92 3.126

18 3.125 43 3.114 68 3.120 93 3.052

19 3.090 44 3.121 69 3.100 94 3.113

20 3.100 45 3.105 70 3.099 95 3.085

21 3.055 46 3.078 71 3.097 96 3.117

22 3.105 47 3.147 72 3.091 97 3.142

23 3.063 48 3.104 73 3.077 98 3.031

24 3.083 49 3.146 74 3.178 99 3.083

25 3.065 50 3.095 75 3.054 100 3.104

After looking at the data we might propose the following null and alternative hypotheses.
H0: The mass of a circulating U.S. penny is between 2.900 g and 3.200 g
HA: The mass of a circulating U.S. penny may be less than 2.900 g or more than 3.200 g
To test the null hypothesis we find a penny and determine its mass. If the penny’s mass is 2.512 g then we can reject the null
hypothesis and accept the alternative hypothesis. Suppose that the penny’s mass is 3.162 g. Although this result increases our
confidence in the null hypothesis, it does not prove that the null hypothesis is correct because the next penny we sample might
weigh less than 2.900 g or more than 3.200 g.
After we state the null and the alternative hypotheses, the second step is to choose a confidence level for the analysis. The
confidence level defines the probability that we will incorrectly reject the null hypothesis when it is, in fact, true. We can express
this as our confidence that we are correct in rejecting the null hypothesis (e.g. 95%), or as the probability that we are incorrect in
rejecting the null hypothesis. For the latter, the confidence level is given as α , where
confidence interval (%)
α =1−
100

For a 95% confidence level, α is 0.05.

The third step is to calculate an appropriate test statistic and to compare it to a critical value. The test statistic’s critical value
defines a breakpoint between values that lead us to reject or to retain the null hypothesis, which is the fourth, and final, step of a
significance test. As we will see in the sections that follow, how we calculate the test statistic depends on what we are comparing.

The four steps for a statistical analysis of data using a significance test:
1. Pose a question, and state the null hypothesis, H0, and the alternative hypothesis, HA.
2. Choose a confidence level for the statistical analysis.
3. Calculate an appropriate test statistic and compare it to a critical value.
4. Either retain the null hypothesis, or reject it and accept the alternative hypothesis.

One-Tailed and Two-tailed Significance Tests

Suppose we want to evaluate the accuracy of a new analytical method. We might use the method to analyze a Standard Reference
Material that contains a known concentration of analyte, μ . We analyze the standard several times, obtaining a mean value, X , for
¯¯¯
¯

¯¯¯
¯
the analyte’s concentration. Our null hypothesis is that there is no difference between X and μ
¯¯¯
¯
H0 : X = μ

¯¯¯
¯
If we conduct the significance test at α = 0.05, then we retain the null hypothesis if a 95% confidence interval around X contains
μ . If the alternative hypothesis is

¯¯¯
¯
HA : X ≠ μ

35.1.26 https://chem.libretexts.org/@go/page/363066
then we reject the null hypothesis and accept the alternative hypothesis if μ lies in the shaded areas at either end of the sample’s
probability distribution curve (Figure 35.1.23a). Each of the shaded areas accounts for 2.5% of the area under the probability
distribution curve, for a total of 5%. This is a two-tailed significance test because we reject the null hypothesis for values of μ at
either extreme of the sample’s probability distribution curve.

Figure 35.1.23: Examples of (a) two-tailed, and (b, c) one-tailed, significance test of X and μ . The probability distribution curves,
¯¯¯
¯

which are normal distributions, are based on the sample’s mean and standard deviation. For α = 0.05, the blue areas account for 5%
of the area under the curve. If the value of μ falls within the blue areas, then we reject the null hypothesis and accept the alternative
hypothesis. We retain the null hypothesis if the value of μ falls within the unshaded area of the curve.
We can write the alternative hypothesis in two additional ways
¯¯¯
¯
HA : X > μ

¯¯¯
¯
HA : X < μ

rejecting the null hypothesis if μ falls within the shaded areas shown in Figure 35.1.23b or Figure 35.1.23c, respectively. In each
case the shaded area represents 5% of the area under the probability distribution curve. These are examples of a one-tailed
significance test.
For a fixed confidence level, a two-tailed significance test is the more conservative test because rejecting the null hypothesis
requires a larger difference between the results we are comparing. In most situations we have no particular reason to expect that
one result must be larger (or must be smaller) than the other result. This is the case, for example, when we evaluate the accuracy of
a new analytical method. A two-tailed significance test, therefore, usually is the appropriate choice.
We reserve a one-tailed significance test for a situation where we specifically are interested in whether one result is larger (or
smaller) than the other result. For example, a one-tailed significance test is appropriate if we are evaluating a medication’s ability to
lower blood glucose levels. In this case we are interested only in whether the glucose levels after we administer the medication are
less than the glucose levels before we initiated treatment. If a patient’s blood glucose level is greater after we administer the
medication, then we know the answer—the medication did not work—and we do not need to conduct a statistical analysis.
Errors in Significance Testing

Because a significance test relies on probability, its interpretation is subject to error. In a significance test, α defines the probability
of rejecting a null hypothesis that is true. When we conduct a significance test at α = 0.05, there is a 5% probability that we will
incorrectly reject the null hypothesis. This is known as a type 1 error, and its risk is always equivalent to α . A type 1 error in a two-
tailed or a one-tailed significance tests corresponds to the shaded areas under the probability distribution curves in Figure 35.1.23.
A second type of error occurs when we retain a null hypothesis even though it is false. This is a type 2 error, and the probability of
its occurrence is β. Unfortunately, in most cases we cannot calculate or estimate the value for β. The probability of a type 2 error,
however, is inversely proportional to the probability of a type 1 error.
Minimizing a type 1 error by decreasing α increases the likelihood of a type 2 error. When we choose a value for α we must
compromise between these two types of error. Most of the examples in this text use a 95% confidence level (α = 0.05) because
this usually is a reasonable compromise between type 1 and type 2 errors for analytical work. It is not unusual, however, to use a
more stringent (e.g. α = 0.01) or a more lenient (e.g. α = 0.10) confidence level when the situation calls for it.

Significance Tests for Normal Distributions

A normal distribution is the most common distribution for the data we collect. Because the area between any two limits of a normal
distribution curve is well defined, it is straightforward to construct and evaluate significance tests.

Comparing X to μ
¯¯¯
¯

One way to validate a new analytical method is to analyze a sample that contains a known amount of analyte, μ . To judge the
method’s accuracy we analyze several portions of the sample, determine the average amount of analyte in the sample, X , and use a ¯¯¯
¯

35.1.27 https://chem.libretexts.org/@go/page/363066
significance test to compare ¯¯¯
X
¯
to μ . The null hypothesis is that the difference between ¯¯¯
X
¯
and μ is explained by indeterminate
¯¯¯
¯ ¯¯¯
¯
errors that affect our determination of X . The alternative hypothesis is that the difference between X and μ is too large to be
explained by indeterminate error.
¯¯¯
¯
H0 : X = μ

¯¯¯
¯
HA : X ≠ μ

The test statistic is texp, which we substitute into the confidence interval for μ

¯¯¯
¯
texp s
μ =X ± −
√n

Rearranging this equation and solving for t exp

¯¯¯
¯ −
|μ − X | √n
texp =
s

gives the value for t exp when μ is at either the right edge or the left edge of the sample's confidence interval (Figure 35.1.24a).

Figure 35.1.24: Relationship between a confidence interval and the result of a significance test. (a) The shaded area under the
normal distribution curve shows the sample’s confidence interval for μ based on texp. The solid bars in (b) and (c) show the
expected confidence intervals for μ explained by indeterminate error given the choice of α and the available degrees of freedom, ν .
For (b) we reject the null hypothesis because portions of the sample’s confidence interval fall outside the confidence interval
explained by indeterminate error. In the case of (c) we retain the null hypothesis because the confidence interval explained by
indeterminate error completely encompasses the sample’s confidence interval.
To determine if we should retain or reject the null hypothesis, we compare the value of texp to a critical value, t(α, ν ), where α is
the confidence level and ν is the degrees of freedom for the sample. The critical value t(α, ν ) defines the largest confidence
interval explained by indeterminate error. If t > t(α, ν ) , then our sample’s confidence interval is greater than that explained by
exp

indeterminate errors (Figure 35.1.24b). In this case, we reject the null hypothesis and accept the alternative hypothesis. If
texp ≤ t(α, ν ) , then our sample’s confidence interval is smaller than that explained by indeterminate error, and we retain the null

hypothesis (Figure 35.1.24c). Example 35.1.24 provides a typical application of this significance test, which is known as a t-test of
X to μ . You will find values for t(α, ν ) in Appendix 3.
¯¯¯
¯

 Example 35.1.3

Before determining the amount of Na2CO3 in a sample, you decide to check your procedure by analyzing a standard sample
that is 98.76% w/w Na2CO3. Five replicate determinations of the %w/w Na2CO3 in the standard gives the following results
98.71% 98.59% 98.62% 98.44% 98.58%

Using α = 0.05, is there any evidence that the analysis is giving inaccurate results?

Solution
The mean and standard deviation for the five trials are

35.1.28 https://chem.libretexts.org/@go/page/363066
 ¯¯¯
¯
X = 98.59 s = 0.0973

Because there is no reason to believe that the results for the standard must be larger or smaller than μ , a two-tailed t-test is
appropriate. The null hypothesis and alternative hypothesis are
¯¯¯
¯ ¯¯¯
¯
H0 : X = μ HA : X ≠ μ

The test statistic, texp, is

¯¯¯
¯ − –
|μ − X | √n |98.76 − 98.59| √5
texp = = = 3.91
2 0.0973

The critical value for t(0.05, 4) from Appendix 3 is 2.78. Since texp is greater than t(0.05, 4), we reject the null hypothesis and
¯¯¯
¯
accept the alternative hypothesis. At the 95% confidence level the difference between X and μ is too large to be explained by
indeterminate sources of error, which suggests there is a determinate source of error that affects the analysis.

 Note
There is another way to interpret the result of this t-test. Knowing that texp is 3.91 and that there are 4 degrees of freedom, we
use Appendix 3 to estimate the value of α that corresponds to a t(α , 4) of 3.91. From Appendix 3, t(0.02, 4) is 3.75 and t(0.01,
4) is 4.60. Although we can reject the null hypothesis at the 98% confidence level, we cannot reject it at the 99% confidence
level. For a discussion of the advantages of this approach, see J. A. C. Sterne and G. D. Smith “Sifting the evidence—what’s
wrong with significance tests?” BMJ 2001, 322, 226–231.

Earlier we made the point that we must exercise caution when we interpret the result of a statistical analysis. We will keep returning
to this point because it is an important one. Having determined that a result is inaccurate, as we did in Example 35.1.3, the next
step is to identify and to correct the error. Before we expend time and money on this, however, we first should critically examine
our data. For example, the smaller the value of s, the larger the value of texp. If the standard deviation for our analysis is
unrealistically small, then the probability of a type 2 error increases. Including a few additional replicate analyses of the standard
and reevaluating the t-test may strengthen our evidence for a determinate error, or it may show us that there is no evidence for a
determinate error.
Comparing s to σ
2 2

If we analyze regularly a particular sample, we may be able to establish an expected variance, σ , for the analysis. This often is the
2

case, for example, in a clinical lab that analyzes hundreds of blood samples each day. A few replicate analyses of a single sample
gives a sample variance, s2, whose value may or may not differ significantly from σ . 2

We can use an F-test to evaluate whether a difference between s2 and σ is significant. The null hypothesis is H : s = σ and the
2
0
2 2

alternative hypothesis is H : s ≠ σ . The test statistic for evaluating the null hypothesis is Fexp, which is given as either
A
2 2

2 2
s 2 2
σ 2 2
Fexp =  if s > σ  or Fexp =  if σ >s
2 2
σ s

depending on whether s2 is larger or smaller than σ

2
. This way of defining Fexp ensures that its value is always greater than or
equal to one.
If the null hypothesis is true, then Fexp should equal one; however, because of indeterminate errors, Fexp, usually is greater than
one. A critical value, F (α, ν ,ν ), is the largest value of Fexp that we can attribute to indeterminate error given the specified
num den

significance level, α , and the degrees of freedom for the variance in the numerator, ν , and the variance in the denominator,
num

νden . The degrees of freedom for s2 is n – 1, where n is the number of replicates used to determine the sample’s variance, and the
degrees of freedom for σ is defined as infinity, ∞. Critical values of F for α = 0.05 are listed in Appendix 4 for both one-tailed
2

and two-tailed F-tests.

 Example 35.1.4

A manufacturer’s process for analyzing aspirin tablets has a known variance of 25. A sample of 10 aspirin tablets is selected
and analyzed for the amount of aspirin, yielding the following results in mg aspirin/tablet.

35.1.29 https://chem.libretexts.org/@go/page/363066
 254 249 252 252 249 249 250 247 251 252

Determine whether there is evidence of a significant difference between the sample’s variance and the expected variance at
α = 0.05.

Solution
The variance for the sample of 10 tablets is 4.3. The null hypothesis and alternative hypotheses are
2 2 2 2
H0 : s =σ HA : s ≠σ

and the value for Fexp is

2
σ 25
Fexp = = = 5.8
2
s 4.3

The critical value for F(0.05, ∞, 9) from Appendix 4 is 3.333. Since Fexp is greater than F(0.05, ∞, 9), we reject the null
hypothesis and accept the alternative hypothesis that there is a significant difference between the sample’s variance and the
expected variance. One explanation for the difference might be that the aspirin tablets were not selected randomly.

Comparing Variances for Two Samples

We can extend the F-test to compare the variances for two samples, A and B, by rewriting our equation for Fexp as
2
s
A
Fexp =
2
s
B

defining A and B so that the value of Fexp is greater than or equal to 1.

 Example 35.1.5

The table below shows results for two experiments to determine the mass of a circulating U.S. penny. Determine whether there
is a difference in the variances of these analyses at α = 0.05.

First Experiment Second Experiment

Penny Mass (g) Penny Mass (g)

1 3.080 1 3.052

2 3.094 2 3.141

3 3.107 3 3.083

4 3.056 4 3.083

5 3.112 5 3.048

6 3.174

7 3.198

Solution
The standard deviations for the two experiments are 0.051 for the first experiment (A) and 0.037 for the second experiment (B).
The null and alternative hypotheses are
2 2 2 2
H0 : s =s HA : s ≠s
A B A B

and the value of Fexp is

2 2
s (0.051) 0.00260
A
Fexp = = = = 1.90
2 2
s (0.037) 0.00137
B

35.1.30 https://chem.libretexts.org/@go/page/363066
 From Appendix 4 the critical value for F(0.05, 6, 4) is 9.197. Because Fexp < F(0.05, 6, 4), we retain the null hypothesis. There
is no evidence at α = 0.05 to suggest that the difference in variances is significant.

Comparing Means for Two Samples

Three factors influence the result of an analysis: the method, the sample, and the analyst. We can study the influence of these
factors by conducting experiments in which we change one factor while holding constant the other factors. For example, to
compare two analytical methods we can have the same analyst apply each method to the same sample and then examine the
resulting means. In a similar fashion, we can design experiments to compare two analysts or to compare two samples.
Before we consider the significance tests for comparing the means of two samples, we need to understand the difference between
unpaired data and paired data. This is a critical distinction and learning to distinguish between these two types of data is important.
Here are two simple examples that highlight the difference between unpaired data and paired data. In each example the goal is to
compare two balances by weighing pennies.
Example 1: We collect 10 pennies and weigh each penny on each balance. This is an example of paired data because we use the
same 10 pennies to evaluate each balance.
Example 2: We collect 10 pennies and divide them into two groups of five pennies each. We weigh the pennies in the first group
on one balance and we weigh the second group of pennies on the other balance. Note that no penny is weighed on both
balances. This is an example of unpaired data because we evaluate each balance using a different sample of pennies.
In both examples the samples of 10 pennies were drawn from the same population; the difference is how we sampled that
population. We will learn why this distinction is important when we review the significance test for paired data; first, however, we
present the significance test for unpaired data.

 Note

One simple test for determining whether data are paired or unpaired is to look at the size of each sample. If the samples are of
different size, then the data must be unpaired. The converse is not true. If two samples are of equal size, they may be paired or
unpaired.

Unpaired Data
¯¯¯
¯ ¯¯¯
¯
Consider two analyses, A and B, with means of X and X , and standard deviations of sA and sB. The confidence intervals for μ
A B A

and for μ are

B

¯¯¯
¯ tsA
μA = X A ±
−
−−
√nA

¯¯¯
¯ tsB
μB = X B ±
−
−−
√nB

where nA and nB are the sample sizes for A and for B. Our null hypothesis, H : μ = μ , is that any difference between μ and
0 A B A

μB is the result of indeterminate errors that affect the analyses. The alternative hypothesis, H : μ ≠ μ , is that the difference
A A B

between μ and μ is too large to be explained by indeterminate error.

A B

To derive an equation for texp, we assume that μ equals μ , and combine the equations for the two confidence intervals
A B

texp sA texp sB
¯¯¯
¯ ¯¯¯
¯
XA ± = XB ±
−
−− −
−−
√nA √nB

Solving for |X ¯¯¯

¯
A
¯¯¯
− XB |
¯
and using a propagation of uncertainty, gives
−−−−−−−−
2 2
s s
¯¯¯
¯ ¯¯¯
¯ A B
| X A − X B | = texp × √ +
nA nB

Finally, we solve for texp

35.1.31 https://chem.libretexts.org/@go/page/363066
 ¯¯¯
¯ ¯¯¯
¯
|X A − X B |
texp = −−−−−−−
2 2
s s
√ A
+
B

nA nB

and compare it to a critical value, t(α, ν ), where α is the probability of a type 1 error, and ν is the degrees of freedom.
¯¯¯
¯
Thus far our development of this t-test is similar to that for comparing X to μ , and yet we do not have enough information to
evaluate the t-test. Do you see the problem? With two independent sets of data it is unclear how many degrees of freedom we have.
Suppose that the variances s and s provide estimates of the same σ . In this case we can replace
2
A
2
B
2
s
2
A
and s
2
B
with a pooled
variance, s , that is a better estimate for the variance. Thus, our equation for t becomes
2
pool exp

¯¯¯
¯ ¯¯¯
¯ ¯¯¯
¯ ¯¯¯
¯ −−−−− −−−
|X A − X B | |X A − X B | nA nB
texp = = ×√
−−−−−−−
1 1 spool nA + nB
spool × √ +
nA nB

where spool, the pooled standard deviation, is

−−−−−−−−−−−−−−−−−−−−
2 2
(nA − 1)s + (nB − 1)s
A B
spool = √
nA + nB − 2

The denominator of this equation shows us that the degrees of freedom for a pooled standard deviation is n + n − 2 , which also A B

is the degrees of freedom for the t-test. Note that we lose two degrees of freedom because the calculations for s and s require 2
A
2
B
¯¯¯
¯ ¯¯¯
¯
the prior calculation of X amd X .
A B

 Note

So how do you determine if it is okay to pool the variances? Use an F-test.

If s and s are significantly different, then we calculate texp using the following equation. In this case, we find the degrees of
2
A
2
B

freedom using the following imposing equation.

2 2 2
s s
A B
( + )
nA nB

ν = −2
2 2
2 2
s s
A B
( ) ( )
n n
A B

+
nA +1 nB +1

Because the degrees of freedom must be an integer, we round to the nearest integer the value of ν obtained from this equation.

 Note

The equation above for the degrees of freedom is from Miller, J.C.; Miller, J.N. Statistics for Analytical Chemistry, 2nd Ed.,
Ellis-Horward: Chichester, UK, 1988. In the 6th Edition, the authors note that several different equations have been suggested
for the number of degrees of freedom for t when sA and sB differ, reflecting the fact that the determination of degrees of
freedom an approximation. An alternative equation—which is used by statistical software packages, such as R, Minitab, Excel
—is
2 2 2 2 2 2
s s s s
A B A B
( + ) ( + )
nA nB nA nB

ν = =
2 2 4 4
s
2
s
2 s s
A B
A B
(
n
) (
n
)
2
+ 2
A B n ( nA −1) n ( nB −1)
A B
+
nA −1 nB −1

For typical problems in analytical chemistry, the calculated degrees of freedom is reasonably insensitive to the choice of
equation.

Regardless of whether how we calculate texp, we reject the null hypothesis if texp is greater than t(α, ν ) and retain the null
hypothesis if texp is less than or equal to t(α, ν ).

35.1.32 https://chem.libretexts.org/@go/page/363066
 Example 35.1.6

Example 35.1.3 provides results for two experiments to determine the mass of a circulating U.S. penny. Determine whether
there is a difference in the means of these analyses at α = 0.05.

Solution
First we use an F-test to determine whether we can pool the variances. We completed this analysis in Example 35.1.5, finding
no evidence of a significant difference, which means we can pool the standard deviations, obtaining
−−−−−−−−−−−−−−−−−−−−−−−−−−
2 2
(7 − 1)(0.051 ) + (5 − 1)(0.037 )
spool = √ = 0.0459
7 +5 −2

with 10 degrees of freedom. To compare the means we use the following null hypothesis and alternative hypotheses

H0 : μA = μB HA : μA ≠ μB

Because we are using the pooled standard deviation, we calculate texp as

−−−−−
|3.117 − 3.081| 7 ×5
texp = ×√ = 1.34
0.0459 7 +5

The critical value for t(0.05, 10), from Appendix 3, is 2.23. Because texp is less than t(0.05, 10) we retain the null hypothesis.
For α = 0.05 we do not have evidence that the two sets of pennies are significantly different.

 Example 35.1.7

One method for determining the %w/w Na2CO3 in soda ash is to use an acid–base titration. When two analysts analyze the
same sample of soda ash they obtain the results shown here.
Analyst A: 86.82% 87.04% 86.93% 87.01% 86.20% 87.00%

Analyst B: 81.01% 86.15% 81.73% 83.19% 80.27% 83.93%

Determine whether the difference in the mean values is significant at α = 0.05.

Solution
We begin by reporting the mean and standard deviation for each analyst.
¯¯¯
¯
X A = 86.83% sA = 0.32%

¯¯¯
¯
X B = 82.71% sB = 2.16%

To determine whether we can use a pooled standard deviation, we first complete an F-test using the following null and
alternative hypotheses.
2 2 2 2
H0 : s =s HA : s ≠s
A B A B

Calculating Fexp, we obtain a value of

2
(2.16)
Fexp = = 45.6
2
(0.32)

Because Fexp is larger than the critical value of 7.15 for F(0.05, 5, 5) from Appendix 4, we reject the null hypothesis and accept
the alternative hypothesis that there is a significant difference between the variances; thus, we cannot calculate a pooled
standard deviation.
To compare the means for the two analysts we use the following null and alternative hypotheses.

35.1.33 https://chem.libretexts.org/@go/page/363066
 ¯¯¯
¯ ¯¯¯
¯ ¯¯¯
¯ ¯¯¯
¯
H0 : X A = X B HA : X A ≠ X B

Because we cannot pool the standard deviations, we calculate texp as

|86.83 − 82.71|
texp = −−−−−−−−−−− = 4.62
2 2
(0.32) (2.16)
√ +
6 6

and calculate the degrees of freedom as

2
2 2
(0.32) (2.16)
( + )
6 6

ν = − 2 = 5.3 ≈ 5
2 2
2 2
( 0.32) ( 2.16)
( ) ( )
6 6

+
6+1 6+1

From Appendix 3, the critical value for t(0.05, 5) is 2.57. Because texp is greater than t(0.05, 5) we reject the null hypothesis
and accept the alternative hypothesis that the means for the two analysts are significantly different at α = 0.05.

Paired Data
Suppose we are evaluating a new method for monitoring blood glucose concentrations in patients. An important part of evaluating
a new method is to compare it to an established method. What is the best way to gather data for this study? Because the variation in
the blood glucose levels amongst patients is large we may be unable to detect a small, but significant difference between the
methods if we use different patients to gather data for each method. Using paired data, in which the we analyze each patient’s blood
using both methods, prevents a large variance within a population from adversely affecting a t-test of means.

 Note

Typical blood glucose levels for most non-diabetic individuals ranges between 80–120 mg/dL (4.4–6.7 mM), rising to as high
as 140 mg/dL (7.8 mM) shortly after eating. Higher levels are common for individuals who are pre-diabetic or diabetic.

When we use paired data we first calculate the individual differences, di, between each sample's paired resykts. Using these
¯
¯¯
individual differences, we then calculate the average difference, d , and the standard deviation of the differences, sd. The null
hypothesis, H : d = 0 , is that there is no difference between the two samples, and the alternative hypothesis, H : d ≠ 0 , is that
0 A

the difference between the two samples is significant.

The test statistic, texp, is derived from a confidence interval around d ¯
¯¯

¯
¯¯ −
| d | √n
texp =
sd

where n is the number of paired samples. As is true for other forms of the t-test, we compare texp to t(α, ν ), where the degrees of
freedom, ν , is n – 1. If texp is greater than t(α, ν ), then we reject the null hypothesis and accept the alternative hypothesis. We
retain the null hypothesis if texp is less than or equal to t(a, o). This is known as a paired t-test.

 Example 35.1.8

Marecek et. al. developed a new electrochemical method for the rapid determination of the concentration of the antibiotic
monensin in fermentation vats [Marecek, V.; Janchenova, H.; Brezina, M.; Betti, M. Anal. Chim. Acta 1991, 244, 15–19]. The
standard method for the analysis is a test for microbiological activity, which is both difficult to complete and time-consuming.
Samples were collected from the fermentation vats at various times during production and analyzed for the concentration of
monensin using both methods. The results, in parts per thousand (ppt), are reported in the following table.

Sample Microbiological Electrochemical

1 129.5 132.3

2 89.6 91.0

35.1.34 https://chem.libretexts.org/@go/page/363066
 3 76.6 73.6

4 52.2 58.2

5 110.8 104.2

6 50.4 49.9

7 72.4 82.1

8 141.4 154.1

9 75.0 73.4

10 34.1 38.1

11 60.3 60.1

Is there a significant difference between the methods at α = 0.05?

Solution
Acquiring samples over an extended period of time introduces a substantial time-dependent change in the concentration of
monensin. Because the variation in concentration between samples is so large, we use a paired t-test with the following null
and alternative hypotheses.
¯
¯¯ ¯
¯¯
H0 : d = 0 HA : d ≠ 0

Defining the difference between the methods as

di = (Xelect )i − (Xmicro )i

we calculate the difference for each sample.

sample 1 2 3 4 5 6 7 8 9 10 11

di 2.8 1.4 –3.0 6.0 –6.6 –0.5 9.7 12.7 –1.6 4.0 –0.2

The mean and the standard deviation for the differences are, respectively, 2.25 ppt and 5.63 ppt. The value of texp is
−−
|2.25| √11
texp = = 1.33
5.63

which is smaller than the critical value of 2.23 for t(0.05, 10) from Appendix 3. We retain the null hypothesis and find no
evidence for a significant difference in the methods at α = 0.05.

One important requirement for a paired t-test is that the determinate and the indeterminate errors that affect the analysis must be
independent of the analyte’s concentration. If this is not the case, then a sample with an unusually high concentration of analyte
¯
¯¯
will have an unusually large di. Including this sample in the calculation of d and sd gives a biased estimate for the expected mean
and standard deviation. This rarely is a problem for samples that span a limited range of analyte concentrations, such as those in
Example 35.1.6 or Exercise 35.1.8. When paired data span a wide range of concentrations, however, the magnitude of the
determinate and indeterminate sources of error may not be independent of the analyte’s concentration; when true, a paired t-test
may give misleading results because the paired data with the largest absolute determinate and indeterminate errors will dominate d . ¯
¯¯

In this situation a regression analysis, which is the subject of the next chapter, is more appropriate method for comparing the data.

 Note
The importance of distinguishing between paired and unpaired data is worth examining more closely. The following is data
from some work I completed with a colleague in which we were looking at concentration of Zn in Lake Erie at the air-water
interface and the sediment-water interface.

35.1.35 https://chem.libretexts.org/@go/page/363066
 sample site ppm Zn at air-water interface ppm Zn at the sediment-water interface

1 0.430 0.415

2 0.266 0.238

3 0.457 0.390

4 0.531 0.410

5 0.707 0.605

6 0.716 0.609

The mean and the standard deviation for the ppm Zn at the air-water interface are 0.5178 ppm and 0.01732 ppm, and the mean
and the standard deviation for the ppm Zn at the sediment-water interface are 0.4445 ppm and 0.1418 ppm. We can use these
values to draw normal distributions for both by letting the means and the standard deviations for the samples, X and s , serve
¯¯¯
¯

as estimates for the means and the standard deviations for the population, μ and σ. As we see in the following figure

the two distributions overlap strongly, suggesting that a t-test of their means is not likely to find evidence of a difference. And
yet, we also see that for each site, the concentration of Zn at the sediment-water interface is less than that at the air-water
interface. In this case, the difference between the concentration of Zn at individual sites is sufficiently large that it masks our
ability to see the difference between the two interfaces.
If we take the differences between the air-water and sediment-water interfaces, we have values of 0.015, 0.028, 0.067, 0.121,
0.102, and 0.107 ppm Zn, with a mean of 0.07333 ppm Zn and a standard deviation of 0.04410 ppm Zn. Superimposing all
three normal distributions

35.1.36 https://chem.libretexts.org/@go/page/363066
 shows clearly that most of the normal distribution for the differences lies above zero, suggesting that a t-test might show
evidence that the difference is significant.

Outliers
Table 35.1.11 provides one more data set giving the masses for a sample of pennies. Do you notice anything unusual in this data?
Of the 100 pennies included in our earlier table, no penny has a mass of less than 3 g. In this table, however, the mass of one penny
is less than 3 g. We might ask whether this penny’s mass is so different from the other pennies that it is in error.
Table 35.1.11. Mass (g) for Additional Sample of Circulating U. S. Penniese
3.067 2.514 3.094

3.049 3.048 3.109

3.039 3.079 3.102

A measurement that is not consistent with other measurements is called an outlier. An outlier might exist for many reasons: the
outlier might belong to a different population
Is this a Canadian penny?
or the outlier might be a contaminated or an otherwise altered sample
Is the penny damaged or unusually dirty?
or the outlier may result from an error in the analysis
Did we forget to tare the balance?
Regardless of its source, the presence of an outlier compromises any meaningful analysis of our data. There are many significance
tests that we can use to identify a potential outlier, three of which we present here.
Dixon's Q-Test

One of the most common significance tests for identifying an outlier is Dixon’s Q-test. The null hypothesis is that there are no
outliers, and the alternative hypothesis is that there is an outlier. The Q-test compares the gap between the suspected outlier and its
nearest numerical neighbor to the range of the entire data set (Figure 35.1.25).

35.1.37 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.25: Dotplots showing the distribution of two data sets containing a possible outlier. In (a) the possible outlier’s value is
larger than the remaining data, and in (b) the possible outlier’s value is smaller than the remaining data.
The test statistic, Qexp, is
gap |outlier's value − nearest value|
Qexp = =
range largest value − smallest value

This equation is appropriate for evaluating a single outlier. Other forms of Dixon’s Q-test allow its extension to detecting multiple
outliers [Rorabacher, D. B. Anal. Chem. 1991, 63, 139–146].
The value of Qexp is compared to a critical value, Q(α, n), where α is the probability that we will reject a valid data point (a type 1
error) and n is the total number of data points. To protect against rejecting a valid data point, usually we apply the more
conservative two-tailed Q-test, even though the possible outlier is the smallest or the largest value in the data set. If Qexp is greater
than Q(α, n), then we reject the null hypothesis and may exclude the outlier. We retain the possible outlier when Qexp is less than
or equal to Q(α, n). Table 35.1.12 provides values for Q(α, n) for a data set that has 3–10 values. A more extensive table is in
Appendix 5. Values for Q(α, n) assume an underlying normal distribution.
Table 35.1.12: Dixon's Q-Test
n Q(0.05, n)

3 0.970

4 0.829

5 0.710

6 0.625

7 0.568

8 0.526

9 0.493

10 0.466

Grubb's Test
Although Dixon’s Q-test is a common method for evaluating outliers, it is no longer favored by the International Standards
Organization (ISO), which recommends the Grubb’s test. There are several versions of Grubb’s test depending on the number of
potential outliers. Here we will consider the case where there is a single suspected outlier.

 Note

For details on this recommendation, see International Standards ISO Guide 5752-2 “Accuracy (trueness and precision) of
measurement methods and results–Part 2: basic methods for the determination of repeatability and reproducibility of a standard
measurement method,” 1994.

The test statistic for Grubb’s test, Gexp, is the distance between the sample’s mean, ¯¯¯
X
¯
, and the potential outlier, Xout , in terms of
the sample’s standard deviation, s.
¯¯¯
¯
| Xout − X |
Gexp =
s

35.1.38 https://chem.libretexts.org/@go/page/363066
We compare the value of Gexp to a critical value G(α, n), where α is the probability that we will reject a valid data point and n is
the number of data points in the sample. If Gexp is greater than G(α, n), then we may reject the data point as an outlier, otherwise
we retain the data point as part of the sample. Table 35.1.13 provides values for G(0.05, n) for a sample containing 3–10 values. A
more extensive table is in Appendix 6. Values for G(α, n) assume an underlying normal distribution.
Table 35.1.13: Grubb's Test
n G(0.05, n)

3 1.115

4 1.481

5 1.715

6 1.887

7 2.020

8 2.126

9 2.215

10 2.290

Chauvenet's Criterion
Our final method for identifying an outlier is Chauvenet’s criterion. Unlike Dixon’s Q-Test and Grubb’s test, you can apply this
method to any distribution as long as you know how to calculate the probability for a particular outcome. Chauvenet’s criterion
states that we can reject a data point if the probability of obtaining the data point’s value is less than (2n ), where n is the size of
−1

the sample. For example, if n = 10, a result with a probability of less than (2 × 10) , or 0.05, is considered an outlier.
−1

To calculate a potential outlier’s probability we first calculate its standardized deviation, z

¯¯¯
¯
| Xout − X |
z =
s

¯¯¯
¯
where X out is the potential outlier, X is the sample’s mean and s is the sample’s standard deviation. Note that this equation is
identical to the equation for Gexp in the Grubb’s test. For a normal distribution, we can find the probability of obtaining a value of z
using the probability table in Appendix 2.

 Example 35.1.9

Table 35.1.11 contains the masses for nine circulating United States pennies. One entry, 2.514 g, appears to be an outlier.
Determine if this penny is an outlier using a Q-test, Grubb’s test, and Chauvenet’s criterion. For the Q-test and Grubb’s test, let
α = 0.05.

Solution
For the Q-test the value for Q exp is
|2.514 − 3.039|
Qexp = = 0.882
3.109 − 2.514

From Table 35.1.12, the critical value for Q(0.05, 9) is 0.493. Because Qexp is greater than Q(0.05, 9), we can assume the
penny with a mass of 2.514 g likely is an outlier.
For Grubb’s test we first need the mean and the standard deviation, which are 3.011 g and 0.188 g, respectively. The value for
Gexp is
|2.514 − 3.011|
Gexp = = 2.64
0.188

Using Table 35.1.13, we find that the critical value for G(0.05, 9) is 2.215. Because Gexp is greater than G(0.05, 9), we can
assume that the penny with a mass of 2.514 g likely is an outlier.

35.1.39 https://chem.libretexts.org/@go/page/363066
 For Chauvenet’s criterion, the critical probability is (2 × 9) , or 0.0556. The value of z is the same as Gexp, or 2.64. Using
−1

Appendix 1, the probability for z = 2.64 is 0.00415. Because the probability of obtaining a mass of 0.2514 g is less than the
critical probability, we can assume the penny with a mass of 2.514 g likely is an outlier.

You should exercise caution when using a significance test for outliers because there is a chance you will reject a valid result. In
addition, you should avoid rejecting an outlier if it leads to a precision that is much better than expected based on a propagation of
uncertainty. Given these concerns it is not surprising that some statisticians caution against the removal of outliers [Deming, W. E.
Statistical Analysis of Data; Wiley: New York, 1943 (republished by Dover: New York, 1961); p. 171].

 Note

You also can adopt a more stringent requirement for rejecting data. When using the Grubb’s test, for example, the ISO 5752
guidelines suggest retaining a value if the probability for rejecting it is greater than α = 0.05, and flagging a value as a
“straggler” if the probability for rejecting it is between α = 0.05 and α = 0.01. A “straggler” is retained unless there is
compelling reason for its rejection. The guidelines recommend using α = 0.01 as the minimum criterion for rejecting a
possible outlier.

On the other hand, testing for outliers can provide useful information if we try to understand the source of the suspected outlier. For
example, the outlier in Table 35.1.11 represents a significant change in the mass of a penny (an approximately 17% decrease in
mass), which is the result of a change in the composition of the U.S. penny. In 1982 the composition of a U.S. penny changed from
a brass alloy that was 95% w/w Cu and 5% w/w Zn (with a nominal mass of 3.1 g), to a pure zinc core covered with copper (with a
nominal mass of 2.5 g) [Richardson, T. H. J. Chem. Educ. 1991, 68, 310–311]. The pennies in Table 35.1.11, therefore, were drawn
from different populations.

Calibrating Data
A calibration curve is one of the most important tools in analytical chemistry as it allows us to determine the concentration of an
analyte in a sample by measuring the signal it generates when placed in an instrument, such as a spectrophotometer. To determine
the analyte's concentration we must know the relationship between the signal we measure , S , and the analyte's concentration, C , A

which we can write as

S = kA CA + Sblank

where k is the calibration curve's sensitivity and S

A blank is the signal in the absence of analyte.
How do we find the best estimate for this relationship between the signal and the concentration of analyte? When a calibration
curve is a straight-line, we represent it using the following mathematical model

y = β0 + β1 x

where y is the analyte’s measured signal, S, and x is the analyte’s known concentration, C , in a series of standard solutions. The
A

constants β and β are, respectively, the calibration curve’s expected y-intercept and its expected slope. Because of uncertainty in
0 1

our measurements, the best we can do is to estimate values for β and β , which we represent as b0 and b1. The goal of a linear
0 1

regression analysis is to determine the best estimates for b0 and b1.

Unweighted Linear Regression With Errors in y

The most common method for completing a linear regression makes three assumptions:
1. the difference between our experimental data and the calculated regression line is the result of indeterminate errors that affect y
2. any indeterminate errors that affect y are normally distributed
3. that indeterminate errors in y are independent of the value of x
Because we assume that the indeterminate errors are the same for all standards, each standard contributes equally in our estimate of
the slope and the y-intercept. For this reason the result is considered an unweighted linear regression.
The second assumption generally is true because of the central limit theorem, which we considered earlier. The validity of the two
remaining assumptions is less obvious and you should evaluate them before you accept the results of a linear regression. In
particular the first assumption is always suspect because there certainly is some indeterminate error in the measurement of x. When

35.1.40 https://chem.libretexts.org/@go/page/363066
we prepare a calibration curve, however, it is not unusual to find that the uncertainty in the signal, S, is significantly greater than the
uncertainty in the analyte’s concentration, C . In such circumstances the first assumption usually is reasonable.
A

How a Linear Regression Works

To understand the logic of a linear regression consider the example in Figure 35.1.26, which shows three data points and two
possible straight-lines that might reasonably explain the data. How do we decide how well these straight-lines fit the data, and how
do we determine which, if either, is the best straight-line?

Figure 35.1.26: Illustration showing three data points and two possible straight-lines that might explain the data. The goal of a
linear regression is to find the one mathematical model, in this case a straight-line, that best explains the data.
Let’s focus on the solid line in Figure 35.1.26. The equation for this line is
^ = b0 + b1 x
y

where b0 and b1 are estimates for the y-intercept and the slope, and y^ is the predicted value of y for any value of x. Because we
assume that all uncertainty is the result of indeterminate errors in y, the difference between y and y^ for each value of x is the
residual error, r, in our mathematical model.
^ )
ri = (yi − y i

Figure 35.1.27 shows the residual errors for the three data points. The smaller the total residual error, R, which we define as
n

2
^ )
R = ∑(yi − y i

i=1

the better the fit between the straight-line and the data. In a linear regression analysis, we seek values of b0 and b1 that give the
smallest total residual error.

 Note

The reason for squaring the individual residual errors is to prevent a positive residual error from canceling out a negative
residual error. You have seen this before in the equations for the sample and population standard deviations introduced in
Chapter 4. You also can see from this equation why a linear regression is sometimes called the method of least squares.

Figure 35.1.27: Illustration that shows the evaluation of a linear regression in which we assume that all uncertainty is the result of
indeterminate errors in y. The points in blue, y , are the original data and the points in red, y^ , are the predicted values from the
i

regression equation, y^ = b + b x .The smaller the total residual error, the better the fit of the straight-line to the data.
0 1

Finding the Slope and y-Intercept for the Regression Model

Although we will not formally develop the mathematical equations for a linear regression analysis, you can find the derivations in
many standard statistical texts [ See, for example, Draper, N. R.; Smith, H. Applied Regression Analysis, 3rd ed.; Wiley: New

35.1.41 https://chem.libretexts.org/@go/page/363066
York, 1998]. The resulting equation for the slope, b1, is
n n n
n∑ xi yi − ∑ xi ∑ yi
i=1 i=1 i=1
b1 =
n n 2
2
n∑ x − (∑ xi )
i=1 i i=1

and the equation for the y-intercept, b0, is

n n
∑ yi − b1 ∑ xi
i=1 i=1
b0 =
n

Although these equations appear formidable, it is necessary only to evaluate the following four summations
n n n n

2
∑ xi ∑ yi ∑ xi yi ∑x
i

i=1 i=1 i=1 i=1

Many calculators, spreadsheets, and other statistical software packages are capable of performing a linear regression analysis based
on this model; see Section 8.5 for details on completing a linear regression analysis using R. For illustrative purposes the necessary
calculations are shown in detail in the following example.

 Example 35.1.10

Using the calibration data in the following table, determine the relationship between the signal, yi , and the analyte's
concentration, x , using an unweighted linear regression.
i

Solution
We begin by setting up a table to help us organize the calculation.
2
xi yi xi yi x
i

0.000 0.00 0.000 0.000

0.100 12.36 1.236 0.010

0.200 24.83 4.966 0.040

0.300 35.91 10.773 0.090

0.400 48.79 19.516 0.160

0.500 60.42 30.210 0.250

Adding the values in each column gives

n n n n

2
∑ xi = 1.500 ∑ yi = 182.31 ∑ xi yi = 66.701 ∑x = 0.550
i

i=1 i=1 i=1 i=1

Substituting these values into the equations for the slope and the y-intercept gives
(6 × 66.701) − (1.500 × 182.31)
b1 = = 120.706 ≈ 120.71
2
(6 × 0.550) − (1.500)

182.31 − (120.706 × 1.500)

b0 = = 0.209 ≈ 0.21
6

The relationship between the signal, S , and the analyte's concentration, C , therefore, is A

S = 120.71 × CA + 0.21

For now we keep two decimal places to match the number of decimal places in the signal. The resulting calibration curve is
shown in Figure 35.1.28.

35.1.42 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.28: Calibration curve for the data in Example 35.1.10.
Uncertainty in the Regression Model
As we see in Figure 35.1.28, because of indeterminate errors in the signal, the regression line does not pass through the exact
center of each data point. The cumulative deviation of our data from the regression line—the total residual error—is proportional to
the uncertainty in the regression. We call this uncertainty the standard deviation about the regression, sr, which is equal to
−−−−−−−−−−−−−
n 2
∑ ^ )
(yi − y
i=1 i
sr = √
n−2

where yi is the ith experimental value, and y^ is the corresponding value predicted by the regression equation y^ = b + b x . Note
i 0 1

that the denominator indicates that our regression analysis has n – 2 degrees of freedom—we lose two degree of freedom because
we use two parameters, the slope and the y-intercept, to calculate y^ . i

A more useful representation of the uncertainty in our regression analysis is to consider the effect of indeterminate errors on the
slope, b1, and the y-intercept, b0, which we express as standard deviations.
−−−−−−−−−−−−−−−−−−− −−−−−−−−−−−−
2 2
nsr sr
sb =√ =√
1
n n 2 n 2
2 ¯¯
n∑ x − (∑ xi ) ∑ (xi − x̄)
i=1 i i=1 i=1

−−−−−−−−−−−−−−−−−−− −−−−−−−−−−−−− −
 2 n 2
 2 n 2
 sr ∑ x  sr ∑ x
i=1 i i=1 i
sb0 = =
2 n n 2 n 2
⎷ n∑ x − (∑i=1 xi ) ⎷n∑ ¯¯
(xi − x̄)
i=1 i i=1

We use these standard deviations to establish confidence intervals for the expected slope, β , and the expected y-intercept, β 1 0

β1 = b1 ± tsb1

β0 = b0 ± tsb0

where we select t for a significance level of α and for n – 2 degrees of freedom. Note that these equations do not contain the factor
of (√−
n)
−1
seen in the confidence intervals for μ because the confidence interval here is based on a single regression line.

 Example 35.1.11

Calculate the 95% confidence intervals for the slope and y-intercept from Example 35.1.10.
Solution
We begin by calculating the standard deviation about the regression. To do this we must calculate the predicted signals, y^ , i

using the slope and the y-intercept from Example 35.1.10, and the squares of the residual error, (y − y^ ) . Using the last i i
2

standard as an example, we find that the predicted signal is

^ = b0 + b1 x6 = 0.209 + (120.706 × 0.500) = 60.562
y 6

and that the square of the residual error is

2 2
(yi − y
^ ) = (60.42 − 60.562 ) = 0.2016 ≈ 0.202
i

The following table displays the results for all six solutions.
2
xi yi ^
y (yi − y
^ )
i i

35.1.43 https://chem.libretexts.org/@go/page/363066
 2
xi yi ^
y ^ )
(yi − y
i i

0.000 0.00 0.209 0.0437

0.100 12.36 12.280 0.0064

0.200 24.83 24.350 0.2304

0.300 35.91 36.421 0.2611

0.400 48.79 48.491 0.0894

0.500 60.42 60.562 0.0202

Adding together the data in the last column gives the numerator in the equation for the standard deviation about the regression;
thus
−−−−−−
0.6512
sr = √ = 0.4035
6 −2

Next we calculate the standard deviations for the slope and the y-intercept. The values for the summation terms are from
Example 35.1.10.
−−−−−−−−−−−−−−−−−−
2
6 × (0.4035)
sb1 = √ = 0.965
(6 × 0.550) − (1.500)2

−−−−−−−−−−−−−−−−−−
2
(0.4035 ) × 0.550
sb =√ = 0.292
0
2
(6 × 0.550) − (1.500)

Finally, the 95% confidence intervals (α = 0.05, 4 degrees of freedom) for the slope and y-intercept are

β1 = b1 ± tsb1 = 120.706 ± (2.78 × 0.965) = 120.7 ± 2.7

β0 = b0 ± tsb0 = 0.209 ± (2.78 × 0.292) = 0.2 ± 0.80

where t(0.05, 4) from Appendix 3 is 2.78. The standard deviation about the regression, sr, suggests that the signal, Sstd, is
precise to one decimal place. For this reason we report the slope and the y-intercept to a single decimal place.

Using the Regression Model to Determine a Value for x Given a Value for y

Once we have our regression equation, it is easy to determine the concentration of analyte in a sample. When we use a normal
calibration curve, for example, we measure the signal for our sample, Ssamp, and calculate the analyte’s concentration, CA, using the
regression equation.
Ssamp − b0
CA =
b1

What is less obvious is how to report a confidence interval for CA that expresses the uncertainty in our analysis. To calculate a
confidence interval we need to know the standard deviation in the analyte’s concentration, s , which is given by the following
CA

equation
−−−−−−−−−−−−−−−−−−−−−−−−−−−−− −
 2
¯¯
¯ ¯¯
¯

(S samp − S std )
sr  1 1
sCA =  + +
 2
b1 m n n ¯
¯¯¯
2
⎷ (b1 ) ∑ (Cstd − C std )
i=1 i

where m is the number of replicates we use to establish the sample’s average signal, Ssamp, n is the number of calibration standards,
¯
¯¯¯
Sstd is the average signal for the calibration standards, and C and C
stdi are the individual and the mean concentrations for the
std

calibration standards. Knowing the value of s , the confidence interval for the analyte’s concentration is
CA

μCA = CA ± tsCA

35.1.44 https://chem.libretexts.org/@go/page/363066
where μ is the expected value of CA in the absence of determinate errors, and with the value of t is based on the desired level of
CA

confidence and n – 2 degrees of freedom.

A close examination of these equations should convince you that we can decrease the uncertainty in the predicted concentration of
analyte, C if we increase the number of standards, n , increase the number of replicate samples that we analyze, m, and if the
A

¯¯
¯ ¯¯
¯
sample’s average signal, S , is equal to the average signal for the standards, S . When practical, you should plan your
samp std

calibration curve so that Ssamp falls in the middle of the calibration curve. For more information about these regression equations
see (a) Miller, J. N. Analyst 1991, 116, 3–14; (b) Sharaf, M. A.; Illman, D. L.; Kowalski, B. R. Chemometrics, Wiley-Interscience:
New York, 1986, pp. 126-127; (c) Analytical Methods Committee “Uncertainties in concentrations estimated from calibration
experiments,” AMC Technical Brief, March 2006.

 Note

The equation for the standard deviation in the analyte's concentration is written in terms of a calibration experiment. A more
general form of the equation, written in terms of x and y, is given here.
−−−−−−−−−−−−−−−−−−−−−−−− −
 2
 ¯
¯¯¯
¯¯
 1 ( Y − ȳ )
sr 1

sx = + +
n 2
b1 ⎷ m n 2 ¯¯
¯
(b1 ) ∑ (xi − x)
i=1

 Example 35.1.12

Three replicate analyses for a sample that contains an unknown concentration of analyte, yields values for Ssamp of 29.32, 29.16
and 29.51 (arbitrary units). Using the results from Example 35.1.10 and Example 35.1.11, determine the analyte’s
concentration, CA, and its 95% confidence interval.
Solution
¯¯
¯
The average signal, S samp , is 29.33, which, using the slope and the y-intercept from Example 35.1.10 , gives the analyte’s
concentration as
¯¯
¯
S samp − b0 29.33 − 0.209
CA = = = 0.241
b1 120.706

¯¯
¯
To calculate the standard deviation for the analyte’s concentration we must determine the values for S std and for
2 ¯
¯¯¯
∑i=1 (Cstdi − C std ) . The former is just the average signal for the calibration standards, which, using the data in Table
2

¯
¯¯¯
35.1.10, is 30.385. Calculating ∑ looks formidable, but we can simplify its calculation by recognizing that
2 2
(C −C ) stdi std
i=1

this sum-of-squares is the numerator in a standard deviation equation; thus,

n
¯
¯¯¯ 2 2
∑(Cstdi − C std ) = (sCstd ) × (n − 1)

i=1

where sC
std
is the standard deviation for the concentration of analyte in the calibration standards. Using the data in Table
35.1.10 we find that s is 0.1871 and
Cstd

n
¯
¯¯¯ 2 2
∑(Cstdi − C std ) = (0.1872 ) × (6 − 1) = 0.175

i=1

Substituting known values into the equation for s CA gives

−−−−−−−−−−−−−−−−−−−−−− −
0.4035 1 1 (29.33 − 30.385)2
sC = √ + + = 0.0024
A
2
120.706 3 6 (120.706 ) × 0.175

Finally, the 95% confidence interval for 4 degrees of freedom is

μCA = CA ± tsCA = 0.241 ± (2.78 × 0.0024) = 0.241 ± 0.007

Figure 35.1.29 shows the calibration curve with curves showing the 95% confidence interval for CA.

35.1.45 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.29: Example of a normal calibration curve with a superimposed confidence interval for the analyte’s concentration.
The points in blue are the original data from Table 35.1.10. The black line is the normal calibration curve as determined in
Example 35.1.1 . The red lines show the 95% confidence interval for CA assuming a single determination of Ssamp.
Evaluating a Regression Model

You should never accept the result of a linear regression analysis without evaluating the validity of the model. Perhaps the simplest
way to evaluate a regression analysis is to examine the residual errors. As we saw earlier, the residual error for a single calibration
standard, ri, is
^ )
ri = (yi − y i

If the regression model is valid, then the residual errors should be distributed randomly about an average residual error of zero,
with no apparent trend toward either smaller or larger residual errors (Figure 35.1.30a). Trends such as those in Figure 35.1.30b
and Figure 35.1.30c provide evidence that at least one of the model’s assumptions is incorrect. For example, a trend toward larger
residual errors at higher concentrations, Figure 35.1.30b, suggests that the indeterminate errors affecting the signal are not
independent of the analyte’s concentration. In Figure 35.1.30c, the residual errors are not random, which suggests we cannot model
the data using a straight-line relationship. Regression methods for the latter two cases are discussed in the following sections.

Figure 35.1.30: Plots of the residual error in the signal, Sstd, as a function of the concentration of analyte, Cstd, for an unweighted
straight-line regression model. The red line shows a residual error of zero. The distribution of the residual errors in (a) indicates
that the unweighted linear regression model is appropriate. The increase in the residual errors in (b) for higher concentrations of
analyte, suggests that a weighted straight-line regression is more appropriate. For (c), the curved pattern to the residuals suggests
that a straight-line model is inappropriate; linear regression using a quadratic model might produce a better fit.

 Example 35.1.13

Use your results from Exercise 35.1.10 to construct a residual plot and explain its significance.

Solution
To create a residual plot, we need to calculate the residual error for each standard. The following table contains the relevant
information.

xi yi ^
y ^
yi − y
i i

0.000 0.000 0.0015 –0.0015

1.55 × 10
−3
0.050 0.0473 0.0027

35.1.46 https://chem.libretexts.org/@go/page/363066
 xi 10
3.16 ×
−3
0.093
y
i 0.0949
^
y
i –0.0019
y −y
i
^
i

0.000
4.74 × 10
−3
0.143
0.000 0.1417
0.0015 –0.0015
0.0013

1.55 × 10
6.34
−3
0.188
0.050 0.1890
0.0473 –0.0010
0.0027

7.92 × 10
−3
0.236 0.2357 0.0003
3.16 × 10
−3
0.093 0.0949 –0.0019
The figure below shows a plot of the resulting residual errors. The residual errors appear random, although they do alternate in
4.74 × 10
−3
0.143 0.1417 0.0013
sign, and they do not show any significant dependence on the analyte’s concentration. Taken together, these observations
suggest that 6.34
−3
our regression
× 10 0.188
model is appropriate. 0.1890 –0.0010

7.92 × 10
−3
0.236 0.2357 0.0003

Weighted Linear Regression With Errors in y

Our treatment of linear regression to this point assumes that any indeterminate errors that affect y are independent of the value of x.
If this assumption is false, then we must include the variance for each value of y in our determination of the y-intercept, b0, and the
slope, b1; thus
n n
∑ wi yi − b1 ∑ wi xi
i=1 i=1
b0 =
n

n n n
n∑ wi xi yi − ∑ wi xi ∑ wi yi
i=1 i=1 i=1
b1 =
n 2 n 2
n∑ wi x − (∑ wi xi )
i=1 i i=1

where wi is a weighting factor that accounts for the variance in yi

−2
n(sy )
i

wi =
n −2
∑ (sy )
i=1 i

and s is the standard deviation for yi. In a weighted linear regression, each xy-pair’s contribution to the regression line is inversely
yi

proportional to the precision of yi; that is, the more precise the value of y, the greater its contribution to the regression.

 Example 35.1.14

Shown here are data for an external standardization in which sstd is the standard deviation for three replicate determination of
the signal.

Cstd (arbitrary units) Sstd (arbitrary units) sstd

0.000 0.00 0.02

0.100 12.36 0.02

0.200 24.83 0.07

0.300 35.91 0.13

0.400 48.79 0.22

0.500 60.42 0.33

Determine the calibration curve’s equation using a weighted linear regression. As you work through this example, remember
that x corresponds to Cstd, and that y corresponds to Sstd.

35.1.47 https://chem.libretexts.org/@go/page/363066
Solution
We begin by setting up a table to aid in calculating the weighting factors.

Cstd (arbitrary units) Sstd (arbitrary units) sstd (sy )

i
−2
wi

0.000 0.00 0.02 2500.00 2.8339

0.100 12.36 0.02 250.00 2.8339

0.200 24.83 0.07 204.08 0.2313

0.300 35.91 0.13 59.17 0.0671

0.400 48.79 0.22 20.66 0.0234

0.500 60.42 0.33 9.18 0.0104

Adding together the values in the fourth column gives

n

−2
∑(sy )
i

i=1

which we use to calculate the individual weights in the last column. As a check on your calculations, the sum of the individual
weights must equal the number of calibration standards, n. The sum of the entries in the last column is 6.0000, so all is well.
After we calculate the individual weights, we use a second table to aid in calculating the four summation terms in the equations
for the slope, b , and the y-intercept, b .
1 0

2
xi yi wi wi xi wi yi wi x wi xi yi
i

0.000 0.00 2.8339 0.0000 0.0000 0.0000 0.0000

0.100 12.36 2.8339 0.2834 35.0270 0.0283 3.5027

0.200 24.83 0.2313 0.0463 5.7432 0.0093 1.1486

0.300 35.91 0.0671 0.0201 2.4096 0.0060 0.7229

0.400 48.79 0.0234 0.0094 1.1417 0.0037 0.4567

0.500 60.42 0.0104 0.0052 0.6284 0.0026 0.3142

Adding the values in the last four columns gives

n n n n

2
∑ wi xi = 0.3644 ∑ wi yi = 44.9499 ∑ wi x = 0.0499 ∑ wi xi yi = 6.1451
i

i=1 i=1 i=1 i=1

which gives the estimated slope and the estimated y-intercept as

(6 × 6.1451) − (0.3644 × 44.9499)
b1 = = 122.985
(6 × 0.0499) − (0.3644)2

44.9499 − (122.985 × 0.3644)

b0 = = 0.0224
6

The calibration equation is

Sstd = 122.98 × Cstd + 0.2

Figure 35.1.31 shows the calibration curve for the weighted regression determined here and the calibration curve for the
unweighted regression. Although the two calibration curves are very similar, there are slight differences in the slope and in the
y-intercept. Most notably, the y-intercept for the weighted linear regression is closer to the expected value of zero. Because the
standard deviation for the signal, Sstd, is smaller for smaller concentrations of analyte, Cstd, a weighted linear regression gives
more emphasis to these standards, allowing for a better estimate of the y-intercept.

35.1.48 https://chem.libretexts.org/@go/page/363066
 Figure 35.1.31: A comparison of the unweighted and the weighted normal calibration curves. See Example 35.1.11 for details of
the unweighted linear regression and Example 35.1.14 for details of the weighted linear regression.

Equations for calculating confidence intervals for the slope, the y-intercept, and the concentration of analyte when using a weighted
linear regression are not as easy to define as for an unweighted linear regression [Bonate, P. J. Anal. Chem. 1993, 65, 1367–1372].
The confidence interval for the analyte’s concentration, however, is at its optimum value when the analyte’s signal is near the
weighted centroid, yc , of the calibration curve.
n
1
yc = ∑ wi xi
n
i=1

Weighted Linear Regression With Errors in x and y

If we remove our assumption that indeterminate errors affecting a calibration curve are present only in the signal (y), then we also
must factor into the regression model the indeterminate errors that affect the analyte’s concentration in the calibration standards (x).
The solution for the resulting regression line is computationally more involved than that for either the unweighted or weighted
regression lines. Although we will not consider the details in this textbook, you should be aware that neglecting the presence of
indeterminate errors in x can bias the results of a linear regression.

 Note

See, for example, Analytical Methods Committee, “Fitting a linear functional relationship to data with error on both variable,”
AMC Technical Brief, March, 2002), as well as this chapter’s Additional Resources.

Curvilinear, Multivariable, and Multivariate Regression

A straight-line regression model, despite its apparent complexity, is the simplest functional relationship between two variables.
What do we do if our calibration curve is curvilinear—that is, if it is a curved-line instead of a straight-line? One approach is to try
transforming the data into a straight-line. Logarithms, exponentials, reciprocals, square roots, and trigonometric functions have
been used in this way. A plot of log(y) versus x is a typical example. Such transformations are not without complications, of which
the most obvious is that data with a uniform variance in y will not maintain that uniform variance after it is transformed.

 Note
It is worth noting here that the term “linear” does not mean a straight-line. A linear function may contain more than one
additive term, but each such term has one and only one adjustable multiplicative parameter. The function
2
y = ax + bx

is an example of a linear function because the terms x and x2 each include a single multiplicative parameter, a and b,
respectively. The function
b
y =x

is nonlinear because b is not a multiplicative parameter; it is, instead, a power. This is why you can use linear regression to fit a
polynomial equation to your data.
Sometimes it is possible to transform a nonlinear function into a linear function. For example, taking the log of both sides of
the nonlinear function above gives a linear function.

35.1.49 https://chem.libretexts.org/@go/page/363066
 log(y) = b log(x)

Another approach to developing a linear regression model is to fit a polynomial equation to the data, such as y = a + bx + cx . 2

You can use linear regression to calculate the parameters a, b, and c, although the equations are different than those for the linear
regression of a straight-line. If you cannot fit your data using a single polynomial equation, it may be possible to fit separate
polynomial equations to short segments of the calibration curve. The result is a single continuous calibration curve known as a
spline function. The use of R for curvilinear regression is included in Chapter 8.5.

 Note

For details about curvilinear regression, see (a) Sharaf, M. A.; Illman, D. L.; Kowalski, B. R. Chemometrics, Wiley-
Interscience: New York, 1986; (b) Deming, S. N.; Morgan, S. L. Experimental Design: A Chemometric Approach, Elsevier:
Amsterdam, 1987.

The regression models in this chapter apply only to functions that contain a single dependent variable and a single independent
variable. One example is the simplest form of Beer's law in which the absorbance, A , of a sample at a single wavelength, λ ,
depends upon the concentration of a single analyte, C A

Aλ = ϵλ,A b CA

where ϵ λ,A is the analyte's molar absorptivity at the selected wavelength and b is the pathlength through the sample. In the presence
of an interferent, I , however, the signal may depend on the concentrations of both the analyte and the interferent

Aλ = ϵλ,A b CA + ϵλ,I b CI

where ϵ is the interferent’s molar absorptivity and CI is the interferent’s concentration. This is an example of multivariable
λ,I

regression, which is covered in more detail in Chapter 9 when we consider the optimization of experiments where there is a single
dependent variable and two or more independent variables.
In multivariate regression we have both multiple dependent variables, such as the absorbance of samples at two or more
wavelengths, and multiple independent variables, such as the concentrations of two or more analytes in the samples. As discussed
in Chapter 0.2, we can represent this using matrix notation

⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯
⎡ ⎤ ⎡ ⎤ ⎡ ⎤

⎢ ⎥ = ⎢ ⎥ ×⎢ ⎥
⎢ ⋮ A ⋮ ⎥ ⎢ ⋮ ϵb ⋮ ⎥ ⎢ ⋮ C ⋮ ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦
⋯ ⋯ ⋯ r×c
⋯ ⋯ ⋯ r×n
⋯ ⋯ ⋯ n×c

where there are r wavelengths, c samples, and n analytes. Each column in the ϵb matrix, for example, holds the ϵb value for a
different analyte at one of r wavelengths, and each row in the C matrix is the concentration of one of the n analytes in one of the c
samples. We will consider this approach in more detail in Chapter 11.

 Note
For a nice discussion of the difference between multivariable regression and multivariate regression, see Hidalgo, B.;
Goodman, M. "Multivariate or Multivariable Regression," Am. J. Public Health, 2013, 103, 39-40.

This page titled 35.1: Evaluation of Analytical Data is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

35.1.50 https://chem.libretexts.org/@go/page/363066
35.2: Single-Sided Normal Distribution
Table 35.2.1, at the bottom of this appendix, gives the proportion, P, of the area under a normal distribution curve that lies to the
right of a deviation, z
X −μ
z =
σ

where X is the value for which the deviation is defined, μ is the distribution’s mean value and σ is the distribution’s standard
deviation. For example, the proportion of the area under a normal distribution to the right of a deviation of 0.04 is 0.4840 (see entry
in red in the table), or 48.40% of the total area (see the area shaded blue in Figure 35.2.1). The proportion of the area to the left of
the deviation is 1 – P. For a deviation of 0.04, this is 1 – 0.4840, or 51.60%.

Figure 35.2.1. Normal distribution curve showing the area under a curve greater than a deviation of +0.04 (blue) and with a
deviation less than –0.04 (green).
When the deviation is negative—that is, when X is smaller than μ —the value of z is negative. In this case, the values in the table
give the area to the left of z. For example, if z is –0.04, then 48.40% of the area lies to the left of the deviation (see area shaded
green in Figure 35.2.1.
To use the single-sided normal distribution table, sketch the normal distribution curve for your problem and shade the area that
corresponds to your answer (for example, see Figure 35.2.2, which is for Example 4.4.2).

Figure 35.2.2 . Normal distribution for the population of aspirin tablets in Example 4.4.2. The population’s mean and standard
deviation are 250 mg and 5 mg, respectively. The shaded area shows the percentage of tablets containing between 243 mg and 262
mg of aspirin.
This divides the normal distribution curve into three regions: the area that corresponds to our answer (shown in blue), the area to
the right of this, and the area to the left of this. Calculate the values of z for the limits of the area that corresponds to your answer.
Use the table to find the areas to the right and to the left of these deviations. Subtract these values from 100% and, voilà, you have
your answer.
Table 35.2.1 : Values for a Single-Sided Normal Distribution
z 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

35.2.1 https://chem.libretexts.org/@go/page/363063
0.0 0.5000 0.4960 0.4920 0.4880 0.4840 0.4801 0.4761 0.4721 0.4681 0.4641

0.1 0.4602 0.4562 0.4522 0.4483 0.4443 0.4404 0.4365 0.4325 0.4286 0.4247

0.2 0.4207 0.4168 0.4129 0.4090 0.4502 0.4013 0.3974 0.3396 0.3897 0.3859

0.3 0.3821 0.3783 0.3745 0.3707 0.3669 0.3632 0.3594 0.3557 0.3520 0.3483

0.4 0.3446 0.3409 0.3372 0.3336 0.3300 0.3264 0.3228 0.3192 0.3156 0.3121

0.5 0.3085 0.3050 0.3015 0.2981 0.2946 0.2912 0.2877 0.2843 0.2810 0.2776

0.6 0.2743 0.2709 0.2676 0.2643 0.2611 0.2578 0.2546 0.2514 0.2483 0.2451

0.7 0.2420 0.2389 0.2358 0.2327 0.2296 0.2266 0.2236 0.2206 0.2177 0.2148

0.8 0.2119 0.2090 0.2061 0.2033 0.2005 0.1977 0.1949 0.1922 0.1894 0.1867

0.9 0.1841 0.1814 0.1788 0.1762 0.1736 0.1711 0.1685 0.1660 0.1635 0.1611

1.0 0.1587 0.1562 0.1539 0.1515 0.1492 0.1469 0.1446 0.1423 0.1401 0.1379

1.1 0.1357 0.1335 0.1314 0.1292 0.1271 0.1251 0.1230 0.1210 0.1190 0.1170

1.2 0.1151 0.1131 0.1112 0.1093 0.1075 0.1056 0.1038 0.1020 0.1003 0.0985

1.3 0.0968 0.0951 0.0934 0.0918 0.0901 0.0885 0.0869 0.0853 0.0838 0.0823

1.4 0.0808 0.0793 0.0778 0.0764 0.0749 0.0735 0.0721 0.0708 0.0694 0.0681

1.5 0.0668 0.0655 0.0643 0.0630 0.0618 0.0606 0.0594 0.0582 0.0571 0.0559

1.6 0.0548 0.0537 0.0526 0.0516 0.0505 0.0495 0.0485 0.0475 0.0465 0.0455

1.7 0.0466 0.0436 0.0427 0.0418 0.0409 0.0401 0.0392 0.0384 0.0375 0.0367

1.8 0.0359 0.0351 0.0344 0.0336 0.0329 0.0322 0.0314 0.0307 0.0301 0.0294

1.9 0.0287 0.0281 0.0274 0.0268 0.0262 0.0256 0.0250 0.0244 0.0239 0.0233

2.0 0.0228 0.0222 0.0217 0.0212 0.0207 0.0202 0.0197 0.0192 0.0188 0.0183

2.1 0.0179 0.0174 0.0170 0.0166 0.0162 0.0158 0.0154 0.0150 0.0146 0.0143

2.2 0.0139 0.0136 0.0132 0.0129 0.0125 0.0122 0.0119 0.0116 0.0113 0.0110

2.3 0.0107 0.0104 0.0102 0.00964 0.00914 0.00866

2.4 0.00820 0.00776 0.00734 0.00695 0.00657

2.5 0.00621 0.00587 0.00554 0.00523 0.00494

2.6 0.00466 0.00440 0.00415 0.00391 0.00368

2.7 0.00347 0.00326 0.00307 0.00289 0.00272

2.8 0.00256 0.00240 0.00226 0.00212 0.00199

2.9 0.00187 0.00175 0.00164 0.00154 0.00144

3.0 0.00135

3.1 0.000968

3.2 0.000687

3.3 0.000483

3.4 0.000337

3.5 0.000233

3.6 0.000159

3.7 0.000108

3.8 0.0000723

35.2.2 https://chem.libretexts.org/@go/page/363063
 3.9 0.0000481

4.0 0.0000317

This page titled 35.2: Single-Sided Normal Distribution is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

35.2.3 https://chem.libretexts.org/@go/page/363063
35.3: Critical Values for t-Test
Assuming we have calculated texp, there are two approaches to interpreting a t-test. In the first approach we choose a value of α for
rejecting the null hypothesis and read the value of t(α, ν ) from the table below. If t > t(α, ν ) , we reject the null hypothesis and
exp

accept the alternative hypothesis. In the second approach, we find the row in the table below that corresponds to the available
degrees of freedom and move across the row to find (or estimate) the a that corresponds to t = t(α, ν ) ; this establishes largest
exp

value of α for which we can retain the null hypothesis. Finding, for example, that α is 0.10 means that we retain the null
hypothesis at the 90% confidence level, but reject it at the 89% confidence level. The examples in this textbook use the first
approach.
Table 35.3.1 : Critical Values of t for the t-Test
Values of t for…

…a confidence interval of: 90% 95% 98% 99%

…an α value of: 0.10 0.05 0.02 0.01

Degrees of Freedom

1 6.314 12.706 31.821 63.657

2 2.920 4.303 6.965 9.925

3 2.353 3.182 4.541 5.841

4 2.132 2.776 3.747 4.604

5 2.015 2.571 3.365 4.032

6 1.943 2.447 3.143 3.707

7 1.895 2.365 2.998 3.499

8 1.860 2.306 2.896 3.255

9 1.833 2.262 2.821 3.250

10 1.812 2.228 2.764 3.169

12 1.782 2.179 2.681 3.055

14 1.761 2.145 2.624 2.977

16 1.746 2.120 2.583 2.921

18 1.734 2.101 2.552 2.878

20 1.725 2.086 2.528 2.845

30 1.697 2.042 2.457 2.750

50 1.676 2.009 2.311 2.678

∞ 1.645 1.960 2.326 2.576

The values in this table are for a two-tailed t-test. For a one-tailed test, divide the α values by 2. For example, the last column has
an α value of 0.005 and a confidence interval of 99.5% when conducting a one-tailed t-test.

This page titled 35.3: Critical Values for t-Test is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

35.3.1 https://chem.libretexts.org/@go/page/363085
35.4: Critical Values for F-Test
The following tables provide values for F (0.05, ν ,ν
num ) for one-tailed and for two-tailed F-tests. To use these tables, we first
denom

decide whether the situation calls for a one-tailed or a two-tailed analysis and calculate Fexp
2
s
A
Fexp =
2
s
B

where S
A
2
is greater than s . Next, we compare Fexp to F (0.05, ν
2
B
,ν num ) and reject the null hypothesis if
denom

Fexp > F (0.05, νnum , νdenom ) . You may replace s with σ if you know the population’s standard deviation.
Table 35.4.1 : Critical Values of F for a One-Tailed F-Test
νnum ⟶

νden om↓
1 2 3 4 5 6 7 8 9 10 15 20 ∞

1 161.4 199.5 215.7 224.6 230.2 234.0 236.8 238.9 240.5 241.9 245.9 248.0 254.3

2 18.51 19.00 19.16 19.25 19.30 19.33 19.35 19.37 19.38 19.40 19.43 19.45 19.50

3 10.13 9.552 9.277 9.117 9.013 8.941 8.887 8.845 8.812 8.786 8.703 8.660 8.526

4 7.709 6.994 6.591 6.388 6.256 6.163 6.094 6.041 5.999 5.964 5.858 5.803 5.628

5 6.608 5.786 5.409 5.192 5.050 4.950 4.876 4.818 4.722 4.753 4.619 4.558 4.365

6 5.987 5.143 4.757 4.534 4.387 4.284 4.207 4.147 4.099 4.060 3.938 3.874 3.669

7 5.591 4.737 4.347 4.120 3.972 3.866 3.787 3.726 3.677 3.637 3.511 3.445 3.230

8 5.318 4.459 4.066 3.838 3.687 3.581 3.500 3.438 3.388 3.347 3.218 3.150 2.928

9 5.117 4.256 3.863 3.633 3.482 3.374 3.293 3.230 3.179 3.137 3.006 2.936 2.707

10 4.965 4.103 3.708 3.478 3.326 3.217 3.135 3.072 3.020 2.978 2.845 2.774 2.538

11 4.844 3.982 3.587 3.257 3.204 3.095 3.012 2.948 2.896 2.854 2.719 2.646 2.404

12 4.747 3.885 3.490 3.259 3.106 2.996 2.913 2.849 2.796 2.753 2.617 2.544 2.296

13 4.667 3.806 3.411 3.179 3.025 2.915 2.832 2.767 2.714 2.671 2.533 2.459 2.206

14 4.600 3.739 3.344 3.112 2.958 2.848 2.764 2.699 2.646 2.602 2.463 2.388 2.131

15 4.534 3.682 3.287 3.056 2.901 2.790 2.707 2.641 2.588 2.544 2.403 2.328 2.066

16 4.494 3.634 3.239 3.007 2.852 2.741 2.657 2.591 2.538 2.494 2.352 2.276 2.010

17 4.451 3.592 3.197 2.965 2.810 2.699 2.614 2.548 2.494 2.450 2.308 2.230 1.960

18 4.414 3.555 3.160 2.928 2.773 2.661 2.577 2.510 2.456 2.412 2.269 2.191 1.917

19 4.381 3.552 3.127 2.895 2.740 2.628 2.544 2.477 2.423 2.378 2.234 2.155 1.878

20 4,351 3.493 3.098 2.866 2.711 2.599 2.514 2.447 2.393 2.348 2.203 2.124 1.843

∞ 3.842 2.996 2.605 2.372 2.214 2.099 2.010 1.938 1.880 1.831 1.666 1.570 1.000

Table 35.4.2 : Critical Values of F for a Two-Tailed F-Test

νnum ⟶

νden om↓
1 2 3 4 5 6 7 8 9 10 15 20 ∞

1 647.8 799.5 864.2 899.6 921.8 937.1 948.2 956.7 963.3 968.6 984.9 993.1 1018

2 38.51 39.00 39.17 39.25 39.30 39.33 39.36 39.37 39.39 39.40 39.43 39.45 39.50

3 17.44 16.04 15.44 15.10 14.88 14.73 14.62 14.54 14.47 14.42 14.25 14.17 13.90

4 12.22 10.65 9.979 9.605 9.364 9.197 9.074 8.980 8.905 8.444 8.657 8.560 8.257

5 10.01 8.434 7.764 7.388 7.146 6.978 6.853 6.757 6.681 6.619 6.428 6.329 6.015

6 8.813 7.260 6.599 6.227 5.988 5.820 5.695 5.600 5.523 5.461 5.269 5.168 4.894

35.4.1 https://chem.libretexts.org/@go/page/363087
 7 8.073 6.542 5.890 5.523 5.285 5.119 4.995 4.899 4.823 4.761 4.568 4.467 4.142

8 7.571 6.059 5.416 5.053 4.817 4.652 4.529 4.433 4.357 4.259 4.101 3.999 3.670

9 7.209 5.715 5.078 4.718 4.484 4.320 4.197 4.102 4.026 3.964 3.769 3.667 3.333

10 6.937 5.456 4.826 4.468 4.236 4.072 3.950 3.855 3.779 3.717 3.522 3.419 3.080

11 6.724 5.256 4.630 4.275 4.044 3.881 3.759 3.644 3.588 3.526 3.330 3.226 2.883

12 6.544 5.096 4.474 4.121 3.891 3.728 3.607 3.512 3.436 3.374 3.177 3.073 2.725

13 6.414 4.965 4.347 3.996 3.767 3.604 3.483 3.388 3.312 3.250 3.053 2.948 2.596

14 6.298 4.857 4.242 3.892 3.663 3.501 3.380 3.285 3.209 3.147 2.949 2.844 2.487

15 6.200 4.765 4.153 3.804 3.576 3.415 3.293 3.199 3.123 3.060 2.862 2.756 2.395

16 6.115 4.687 4.077 3.729 3.502 3.341 3.219 3.125 3.049 2.986 2.788 2.681 2.316

17 6.042 4.619 4.011 3.665 3.438 3.277 3.156 3.061 2.985 2.922 2.723 2.616 2.247

18 5.978 4.560 3.954 3.608 3.382 3.221 3.100 3.005 2.929 2.866 2.667 2.559 2.187

19 5.922 4.508 3.903 3.559 3.333 3.172 3.051 2.956 2.880 2.817 2.617 2.509 2.133

20 5.871 4.461 3.859 3.515 3.289 3.128 3.007 2.913 2.837 2.774 2.573 2.464 2.085

∞ 5.024 3.689 3.116 2.786 2.567 2.408 2.288 2.192 2.114 2.048 1.833 1.708 1.000

This page titled 35.4: Critical Values for F-Test is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

35.4.2 https://chem.libretexts.org/@go/page/363087
35.5: Critical Values for Dixon's Q-Test
The following table provides critical values for Q(α, n), where α is the probability of incorrectly rejecting the suspected outlier
and n is the number of samples in the data set. There are several versions of Dixon’s Q-Test, each of which calculates a value for
Qij where i is the number of suspected outliers on one end of the data set and j is the number of suspected outliers on the opposite
end of the data set. The critical values for Q here are for a single outlier, Q10, where
|outlier's value − nearest value|
Qexp = Q10 =
largest value − smallest value

The suspected outlier is rejected if Qexp is greater than Q(α, n). For additional information consult Rorabacher, D. B. “Statistical
Treatment for Rejection of Deviant Values: Critical Values of Dixon’s ‘Q’ Parameter and Related Subrange Ratios at the 95%
confidence Level,” Anal. Chem. 1991, 63, 139–146.
Table 35.5.1 : Critical Values for Dixon's Q-Test
α⟶

n↓
0.1 0.05 0.04 0.02 0.01

3 0.941 0.970 0.976 0.988 0.994

4 0.765 0.829 0.846 0.889 0.926

5 0.642 0.710 0.729 0.780 0.821

6 0.560 0.625 0.644 0.698 0.740

7 0.507 0.568 0.586 0.637 0.680

8 0.468 0.526 0.543 0.590 0.634

9 0.437 0.493 0.510 0.555 0.598

10 0.412 0.466 0.483 0.527 0.568

This page titled 35.5: Critical Values for Dixon's Q-Test is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

35.5.1 https://chem.libretexts.org/@go/page/363089
35.6: Critical Values for Grubb's Test
The following table provides critical values for G(α, n), where α is the probability of incorrectly rejecting the suspected outlier
and n is the number of samples in the data set. There are several versions of Grubb’s Test, each of which calculates a value for Gij
where i is the number of suspected outliers on one end of the data set and j is the number of suspected outliers on the opposite end
of the data set. The critical values for G given here are for a single outlier, G10, where
¯¯¯
¯
| Xout − X |
Gexp = G10 =
s

The suspected outlier is rejected if Gexp is greater than G(α, n).

Table 35.6.1 : Critical Values for the Grubb's Test
α⟶

n↓
0.05 0.01

3 1.155 1.155

4 1.481 1.496

5 1.715 1.764

6 1.887 1.973

7 2.020 2.139

8 2.126 2.274

9 2.215 2.387

10 2.290 2.482

11 2.355 2.564

12 2.412 2.636

13 2.462 2.699

14 2.507 2.755

15 2.549 2.755

This page titled 35.6: Critical Values for Grubb's Test is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

35.6.1 https://chem.libretexts.org/@go/page/363093
35.7: Activity Coefficients
Careful measurements on the metal–ligand complex Fe(SCN)2+ suggest its stability decreases in the presence of inert ions [Lister,
M. W.; Rivington, D. E. Can. J. Chem. 1995, 33, 1572–1590]. We can demonstrate this by adding an inert salt to an equilibrium
mixture of Fe3+ and SCN–. Figure 35.7.1 a shows the result of mixing together equal volumes of 1.0 mM FeCl3 and 1.5 mM
KSCN, both of which are colorless. The solution’s reddish–orange color is due to the formation of Fe(SCN)2+.
3+ − 2+
Fe (aq) + SCN (aq) ⇌ Fe(SCN) (aq) (35.7.1)

Figure 35.7.1 . The effect of a inert salt on a reaction’s equilibrium position is shown by the solutions in these two beakers. The
beaker on the left contains equal volumes of 1.0 mM FeCl3 and 1.5 mM KSCN. The solution’s color is due to the formation of the
metal–ligand complex Fe(SCN)2+. Adding 10 g of KNO3 to the beaker on the left produces the result shown on the right. The
lighter color suggests that there is less Fe(SCN)2+ as a result of the equilibrium in reaction 35.7.1 shifting to the left.
Adding 10 g of KNO3 to the solution and stirring to dissolve the solid, produces the result shown in Figure 35.7.1 b. The solution’s
lighter color suggests that adding KNO3 shifts reaction 35.7.1 to the left, decreasing the concentration of Fe(SCN)2+ and increasing
the concentrations of Fe3+ and SCN–. The result is a decrease in the complex’s formation constant, K1.
2+
[Fe(SCN) ]
K1 = (35.7.2)
3+ −
[ Fe ] [ SCN ]

Why should adding an inert electrolyte affect a reaction’s equilibrium position? We can explain the effect of KNO3 on the
formation of Fe(SCN)2+ if we consider the reaction on a microscopic scale. The solution in Figure 35.7.1 b contains a variety of
cations and anions: Fe3+, SCN–, K+, NO , H3O+, and OH–. Although the solution is homogeneous, on average, there are slightly
−

more anions in regions near the Fe3+ ions, and slightly more cations in regions near the SCN– ions. As shown in Figure 35.7.2 ,
each Fe3+ ion and each SCN– ion is surrounded by an ionic atmosphere of opposite charge (δ and δ ) that partially screen the ions
– +

from each other. Because each ion’s apparent charge at the edge of its ionic atmosphere is less than its actual charge, the force of
attraction between the two ions is smaller. As a result, the formation of Fe(SCN)2+ is slightly less favorable and the formation
constant in Equation 35.7.2 is slightly smaller. Higher concentrations of KNO3 increase δ and δ , resulting in even smaller values
– +

for the formation constant.

Figure 35.7.2 . Ions of Fe3+ and SCN– are surrounded by ionic atmospheres with net charges of δ and δ . Because of these ionic
– +

atmospheres, each ion’s apparent charge at the edge of its ionic atmosphere is less than the ion’s actual charge.

35.7.1 https://chem.libretexts.org/@go/page/362969
Ionic Strength
To factor the concentration of ions into the formation constant for Fe(SCN)2+, we need a way to express that concentration in a
meaningful way. Because both an ion’s concentration and its charge are important, we define the solution’s ionic strength, μ as
n
1
2
μ = ∑ ci z
i
2
i=1

where ci and zi are the concentration and charge of the ith ion.

 Example 35.7.1

Calculate the ionic strength of a solution of 0.10 M NaCl. Repeat the calculation for a solution of 0.10 M Na2SO4.

Solution
The ionic strength for 0.10 M NaCl is
1 + 2 − 2
μ = {[ Na ] × (+1 ) + [ Cl ] × (−1 ) }
2

1 2 2
μ = {(0.10) × (+1 ) + (0.10) × (−1 ) } = 0.10 M
2

For 0.10 M Na2SO4 the ionic strength is

1 + 2 2− 2
μ = {[ Na ] × (+1 ) + [ SO ] × (−2 ) }
2 4

1 2 2
μ = {(0.20) × (+1 ) + (0.10) × (−2 ) } = 0.30 M
2

In calculating the ionic strengths of these solutions we are ignoring the presence of H3O+ and OH–, and, in the case of Na2SO4,
the presence of HSO from the base dissociation reaction of SO . In the case of 0.10 M NaCl, the concentrations of H3O+
−

4
2−

and OH– are 1.0 × 10 , which is significantly smaller than the concentrations of Na+ and Cl–. Because SO is a very weak
−7 2−
4

base (Kb = 1.0 × 10 ), the solution is only slightly basic (pH = 7.5), and the concentrations of H3O+, OH–, and HSO are
−12 −

negligible. Although we can ignore the presence of H3O+, OH–, and HSO when we calculate the ionic strength of these two
−

solutions, be aware that an equilibrium reaction can generate ions that might affect the solution’s ionic strength.

Note that the unit for ionic strength is molarity, but that a salt’s ionic strength need not match its molar concentration. For a 1:1 salt,
such as NaCl, ionic strength and molar concentration are identical. The ionic strength of a 2:1 electrolyte, such as Na2SO4, is three
times larger than the electrolyte’s molar concentration.

Activity and Activity Coefficients

Figure 35.7.1 shows that adding KNO3 to a mixture of Fe3+ and SCN– decreases the formation constant for Fe(SCN)2+. This
creates a contradiction. Earlier in this chapter we showed that there is a relationship between a reaction’s standard‐state free energy,
∆Go, and its equilibrium constant, K.
∘
△G = −RT ln K

Because a reaction has only one standard‐state, its equilibrium constant must be independent of solution conditions. Although ionic
strength affects the apparent formation constant for Fe(SCN)2+, reaction 35.7.1 must have an underlying thermodynamic formation
constant that is independent of ionic strength.
The apparent formation constant for Fe(SCN)2+, as shown in Equation 35.7.2, is a function of concentrations. In place of
concentrations, we define the true thermodynamic equilibrium constant using activities. The activity of species A, aA, is the product
of its concentration, [A], and a solution‐dependent activity coefficient, γ A

aA = [A] γA

The true thermodynamic formation constant for Fe(SCN)2+, therefore, is

35.7.2 https://chem.libretexts.org/@go/page/362969
 2+
a 2+ [Fe(SCN) ]γ 2+
Fe(SCN) Fe(SCN)
K1 = =
3+ −
a 3+ ×a −
[ Fe ]γ 3+ [ SCN ]γ −
Fe SCN Fe SCN

Unless otherwise specified, the equilibrium constants in the appendices are thermodynamic equilibrium constants.

A species’ activity coefficient corrects for any deviation between its physical concentration and its ideal value. For a gas, a pure
solid, a pure liquid, or a non‐ionic solute, the activity coefficient is approximately one under most reasonable experimental
conditions.

For a gas the proper terms are fugacity and fugacity coefficient, instead of activity and activity coefficient.

For a reaction that involves only these species, the difference between activity and concentration is negligible. The activity
coefficient for an ion, however, depends on the solution’s ionic strength, the ion’s charge, and the ion’s size. It is possible to
estimate activity coefficients using the extended Debye‐Hückel equation
2 −
−0.51 × z × √μ
A
log γA = (35.7.3)
−
1 + 3.3 × αA × √μ

where zA is the ion’s charge, α is the hydrated ion’s effective diameter in nanometers (Table 6.2), μ is the solution’s ionic
A

strength, and 0.51 and 3.3 are constants appropriate for an aqueous solution at 25oC. A hydrated ion’s effective radius is the radius
of the ion plus those water molecules closely bound to the ion. The effective radius is greater for smaller, more highly charged ions
than it is for larger, less highly charged ions.
Table 35.7.1 . Effective Diameters (α ) for Selected Ions
ion effective diameter (nm)

H3O+ 0.9

Li+ 0.6

Na+, IO , HSO , HCO , H

−
3
−
3
−
3 2 PO4
−
0.45

OH–, F–, SCN–, HS–, ClO , ClO , MnO −

3
−
4
−
4
0.35

K+, Cl–, Br–, I–, CN–, NO

−
2
, NO −
3
0.3

Cs+, Tl+, Ag+, NH +

4
0.25

Mg2+, Be2+ 0.8

Ca2+, Cu2+, Zn2+, Sn2+, Mn2+, Fe2+, Ni2+, Co2+ 0.6

Sr2+, Ba2+, Cd2+, Hg2+, S2– 0.5

Pb2+, SO 2−
4
, SO 2−
3
0.45

Hg
2+
2
, SO 2−
4
,S 2 2 O3
2−
, CrO 2−
4
, HPO 2−
4
0.40

Al3+, Fe3+, Cr3+ 0.9

PO
3−
4
, Fe(CN) 3−
6
0.4

Zr4+, Ce4+, Sn4+ 1.1

Fe(CN)
4−
6
0.5

Source: Kielland, J. J. Am. Chem. Soc. 1937, 59, 1675–1678.

Several features of Equation 35.7.3 deserve our attention. First, as the ionic strength approaches zero an ion’s activity coefficient
approaches a value of one. In a solution where μ = 0 , an ion’s activity and its concentration are identical. We can take advantage of
this fact to determine a reaction’s thermodynamic equilibrium constant by measuring the apparent equilibrium constant for several
increasingly smaller ionic strengths and extrapolating back to an ionic strength of zero. Second, an activity coefficient is smaller,

35.7.3 https://chem.libretexts.org/@go/page/362969
and the effect of activity is more important, for an ion with a higher charge and a smaller effective radius. Finally, the extended
Debye‐Hückel equation provides a reasonable estimate of an ion’s activity coefficient when the ionic strength is less than 0.1.
Modifications to Equation 35.7.3 extend the calculation of activity coefficients to higher ionic strengths [Davies, C. W. Ion
Association, Butterworth: London, 1962].

Including Activity Coefficients When Solving Equilibrium Problems

Earlier in this chapter we calculated the solubility of Pb(IO3)2 in deionized water, obtaining a result of 4.0 × 10 mol/L. Because −5

the only significant source of ions is from the solubility reaction, the ionic strength is very low and we can assume that γ ≈ 1 for
both Pb2+ and IO . In calculating the solubility of Pb(IO3)2 in deionized water, we do not need to account for ionic strength. But
−
3

what if we need to know the solubility of Pb(IO3)2 in a solution that contains other, inert ions? In this case we need to include
activity coefficients in our calculation.

 Example 35.7.2

Calculate the solubility of Pb(IO3)2 in a matrix of 0.020 M Mg(NO3)2.

Solution
We begin by calculating the solution’s ionic strength. Since Pb(IO3)2 is only sparingly soluble, we will assume we can ignore
its contribution to the ionic strength; thus
1 2 2
μ = {(0.020)(+2 ) + (0.040)(−1 ) } = 0.060 M
2

Next, we use Equation 35.7.3 to calculate the activity coefficients for Pb2+ and IO . −
3

2
−−−−
−0.51 × (+2 ) × √0.060
log γ 2+ = −−−− = −0.366
Pb
1 + 3.3 × 0.45 × √0.060

γ 2+ = 0.431
Pb

2
−−−−
−0.51 × (−1 ) × √0.060
log γ − = = −0.0916
IO 3 −−−−
1 + 3.3 × 0.45 × √0.060

γ − = 0.810
IO 3

Defining the equilibrium concentrations of Pb2+ and IO in terms of the variable x

−

Concentrations Pb(IO3)2 (s) ⇌ Pb2+ (aq) + 2IO (aq)

−
3

initial solid 0 0

change solid +x +2x

equilibrium solid x 2x

and substituting into the thermodynamic solubility product for Pb(IO3)2 leaves us with
2+ − 2 −13
2 2
Ksp = aPb 2+ × a − = γPb 2+ [ Pb ] ×γ − [ IO3 ] = 2.5 × 10
IO IO
3 3

2 2 −13
Ksp = (0.431)(x)(0.810 ) (2x ) = 2.5 × 10

3 −13
Ksp = 1.131 x = 2.5 × 10

Solving for x gives 6.0 × 10 and a molar solubility of 6.0 × 10 mol/L for Pb(IO3)2. If we ignore activity, as we did in our
−5 −5

earlier calculation, we report the molar solubility as 4.0 × 10 mol/L. Failing to account for activity in this case
−5

underestimates the molar solubility of Pb(IO3)2 by 33%.

The solution’s equilibrium composition is

35.7.4 https://chem.libretexts.org/@go/page/362969
 2+ −5
[ Pb ] = 6.0 × 10  M

− −4
[ IO ] = 1.2 × 10  M
3

2+
[ Mg ] = 0.020 M

−
[ NO ] = 0.040 M
3

Because the concentrations of both Pb2+ and IO are much smaller than the concentrations of Mg2+ and NO our decision to
−
3
−
3

ignore the contribution of Pb2+ and IO to the ionic strength is reasonable.

−
3

How do we handle the calculation if we can not ignore the concentrations of Pb2+ and IO when calculating the ionic −
3

strength. One approach is to use the method of successive approximations. First, we recalculate the ionic strength using the
concentrations of all ions, including Pb2+ and IO . Next, we recalculate the activity coefficients for Pb2+ and IO using this
−
3
−
3

new ionic strength and then recalculate the molar solubility. We continue this cycle until two successive calculations yield the
same molar solubility within an acceptable margin of error.

 Exercise 35.7.1

Calculate the molar solubility of Hg2Cl2 in 0.10 M NaCl, taking into account the effect of ionic strength. Compare your answer
to that from Exercise 6.7.2 in which you ignored the effect of ionic strength.

Answer
We begin by calculating the solution’s ionic strength. Because NaCl is a 1:1 ionic salt, the ionic strength is the same as the
concentration of NaCl; thus μ = 0.10 M. This assumes, of course, that we can ignore the contributions of Hg and Cl– 2+
2

from the solubility of Hg2Cl2.

Next we use Equation 35.7.3to calculate the activity coefficients for Hg 2+
2
and Cl–.
2 −−−
−
−0.51 × (+2 ) × √0.10
log γ 2+ = = −0.455
Hg −−−
−
2
1 + 3.3 × 0.40 × √0.10

γ 2+ = 0.351
Hg
2

2
−−−
−
−0.51 × (−1 ) × √0.10
log γCl− = = −0.12
−−−
−
1 + 3.3 × 0.3 × √0.10

γ − = 0.75
Cl

Defining the equilibrium concentrations of Hg 2+

2
and Cl– in terms of the variable x

concentrations Hg2Cl2 (s) ⇌ Hg

2+

2
(aq) + 2Cl– (aq)

initial solid 0 0.10

change solid +x +2x

equilibrium solid x 0.10 + 2x

and substituting into the thermodynamic solubility product for Hg2Cl2, leave us with
2 2+ 2 − 2 −18
Ksp = a 2+ (a − ) =γ 2+ [ Hg ] (γ − ) [ Cl ] = 1.2 × 10
Hg Cl Hg 2 Cl
2 2

Because the value of x likely is small, let’s simplify this equation to

2 2 −18
(0.351)(x)(0.75 ) (0.1 ) = 1.2 × 10

Solving for x gives its value as 6.1 × 10 . Because x is the concentration of Hg and 2x is the concentration of Cl–, our
−16 2+

decision to ignore their contributions to the ionic strength is reasonable. The molar solubility of Hg2Cl2 in 0.10 M NaCl is

35.7.5 https://chem.libretexts.org/@go/page/362969
 6.1 × 10
−16
mol/L. In Exercise 6.7.2, where we ignored ionic strength, we determined that the molar solubility of Hg2Cl2
is 1.2 × 10 −16
mol/L, a result that is 5× smaller than the its actual value.

As Example 35.7.2 and Exercise 35.7.1 show, failing to correct for the effect of ionic strength can lead to a significant error in an
equilibrium calculation. Nevertheless, it is not unusual to ignore activities and to assume that the equilibrium constant is expressed
in terms of concentrations. There is a practical reason for this—in an analysis we rarely know the exact composition, much less the
ionic strength of aqueous samples or of solid samples brought into solution. Equilibrium calculations are a useful guide when we
develop an analytical method; however, it only is when we complete an analysis and evaluate the results that can we judge whether
our theory matches reality. In the end, work in the laboratory is the most critical step in developing a reliable analytical method.

This is a good place to revisit the meaning of pH. In Chapter 2 we defined pH as

+
pH = − log[ H3 O ]

Now we see that the correct definition is

pH = − log aH +
O
3

+
pH = − log γH + [ H3 O ]
3O

Failing to account for the effect of ionic strength can lead to a significant error in the reported concentration of H3O+. For
example, if the pH of a solution is 7.00 and the activity coefficient for H3O+ is 0.90, then the concentration of H3O+ is
1.11 × 10
−7
M, not 1.00 × 10 M, an error of +11%. Fortunately, when we develop and carry out an analytical method, we
−7

are more interested in controlling pH than in calculating [H3O+]. As a result, the difference between the two definitions of pH
rarely is of significant concern.

This page titled 35.7: Activity Coefficients is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by David
Harvey.

35.7.6 https://chem.libretexts.org/@go/page/362969
35.8: Standard Reduction Potentials & Polarographic Half-wave Potentials
Standard/Formal Reduction Potentials
The following table provides Eo and Eo ́ values for selected reduction reactions. Values are from the following sources (primarily
the first two):
Bard, A. J.; Parsons, B.; Jordon, J., eds. Standard Potentials in Aqueous Solutions, Dekker: New York, 1985
Milazzo, G.; Caroli, S.; Sharma, V. K. Tables of Standard Electrode Potentials, Wiley: London, 1978;
Swift, E. H.; Butler, E. A. Quantitative Measurements and Chemical Equilibria, Freeman: New York, 1972.
Bratsch, S. G. "Standard Electrode Potentials and Temperature Coefficients in Water at 298.15K, J. Phys. Chem. Ref. Data,
1989, 18, 1–21.
Latimer, W. M. Oxidation Potentials, 2nd. Ed., Prentice-Hall: Englewood Cliffs, NJ, 1952
Solids, gases, and liquids are identified; all other species are aqueous. Reduction reactions in acidic solution are written using H+ in
place of H3O+. You may rewrite a reaction by replacing H+ with H3O+ and adding to the opposite side of the reaction one molecule
of H2O per H+; thus
H3AsO4 + 2H+ +2e– ⇌ HAsO2 +2H2O
becomes
H3AsO4 + 2H3O+ +2e– ⇌ HAsO2 +4H2O
Conditions for formal potentials (Eo ́) are listed next to the potential.
For most of the reduction half-reactions gathered here, there are minor differences in values provided by the references above. In
most cases, these differences are small and will not affect calculations. In a few cases the differences are not insignificant and the
user may find discrepancies in calculations. For example, Bard, Parsons, and Jordon report an Eo value of –1.285 V for
2− − −
Zn(OH) + 2e ⇌ Zn(s) + 4 OH
4

while Milazzo, Caroli, and Sharma report the value as –1.214 V, Swift reports the value as –1.22, Bratsch reports the value as –
1.199 V, and Latimer reports the value as –1.216 V.

Aluminum E (V) Eo ́ (V)

Al
3+
+ 3e
−
⇌ Al(s) –1.676

Al(OH)
−

4
+ 3e
−
⇌ Al(s) + 4 OH
−
–2.310

AlF
3−
6
+ 3e
−
⇌ Al(s) + 6 F
−
–2.07

Antimony E (V) Eo ́ (V)

Sb + 3 H
+
+ 3e
−
⇌ SbH 3 (g) –0.510

Sb 2 O5 + 6 H
+
+ 4e
−
⇌ 2 SbO
+
+ 3 H2 O(l) 0.605

SbO
+
+ 2H
+
+ 3e
−
⇌ Sb(s) + H2 O(l) 0.212

Arsenic E (V) Eo ́ (V)

As(s) + 3 H
+
+ 3e
−
⇌ AsH3 (g) –0.225

H3 AsO4 + 2 H
+
+ 2e
−
⇌ HAsO2 + 2 H2 O(l) 0.560

HAsO2 + 3 H
+
+ 3e
−
⇌ As(s) + 2 H2 O(l) 0.240

Barium E (V) Eo ́ (V)

Ba
2+
+ 2e
−
⇌ Ba(s) –2.92

BaO(s) + 2 H
+
+ 2e
−
⇌ Ba(s) + H2 O(l) –2.166

35.8.1 https://chem.libretexts.org/@go/page/362971
 Beryllium E (V) Eo ́ (V)

Be
2+
2+
+ 2e
−
⇌ Be(s) –1.99

Bismuth E (V) Eo ́ (V)

Bi
3+
+ 3e
−
⇌ Bi(s) 0.317

BiCl
−
4
+ 3e
−
⇌ Bi(s) + 4 Cl
−
0.199

Boron E (V) Eo ́ (V)

B(OH)
3
+ 3H
+
+ 3e
−
⇌ B(s) + 3 H2 O(l) –0.890

B(OH)
−

4
+ 3e
−
⇌ B(s) + 4 OH
−
–1.811

Bromine E (V) Eo ́ (V)

Br2 (l) + 2 e
−
⇌ 2 Br
−
1.087

HOBr + H
+
+ 2e
−
⇌ Br
−
+ H2 O(l) 1.341

HOBr + H
+
+e
−
⇌
1

2
Br2 + H2 O(l) 1.604

BrO
−
+ H2 O(l) + 2 e
−
⇌ Br
−
+ 2 OH
−
0.76 in 1 M NaOH

BrO
−
3
+ 6H
+
+ 5e
−
⇌
1

2
Br2 (l) + 3 H2 O(l) 1.5

BrO
−
3
+ 6H
+
+ 6e
−
⇌ Br
−
+ 3 H2 O(l) 1.478

Cadmium E (V) Eo ́ (V)

Cd
2+
+ 2e
−
⇌ Cd(s) –0.4030

Cd(CN)
2−
4
+ 2e
−
⇌ Cd(s) + 4 CN
−
–0.943

Cd(NH )
3
2+
4
+ 2e
−
⇌ Cd(s) + 4 NH3 –0.622

Calcium E (V) Eo ́ (V)

Ca
2+
+ 2e
−
⇌ Ca(s) –2.84

Carbon E (V) Eo ́ (V)

CO 2 (g) + 2 H
+
+ 2e
−
⇌ CO(g) + H2 O(l) –0.106

CO 2 (g) + 2 H
+
+ 2e
−
⇌ HCO2 H –0.20

2 CO 2 (g) + 2 H
+
+ 2e
−
⇌ H2 C2 O4 –0.481

HCHO + 2 H
+
+ 2e
−
⇌ CH 3 OH 0.2323

Cerium E (V) Eo ́ (V)

Ce
3+
+ 3e
−
⇌ Ce(s) –2.336

1.70 in 1 M HClO4
1.44 in 1 M H2SO4
Ce
4+
+e
−
⇌ Ce
3+
1.72
1.61 in 1 M HNO3
1.28 in 1 M HCl

Chlorine E (V) Eo ́ (V)

Cl 2 (g) + 2 e
−
⇌ 2 Cl
−
1.396

ClO
−
+ H2 O(l) + e
−
⇌
1

2
Cl 2 (g) + 2 OH
−
0.421 in 1 M NaOH

35.8.2 https://chem.libretexts.org/@go/page/362971
 o
ClO
−
+ H2 O(l) + 2 e Chlorine−
⇌ Cl
−
+ 2 OH
−
E (V) 0.890 in 1 M NaOH E ́ (V)

Cl 2 (g)2+
HClO +22
eH ⇌+22
Cle
−+ −
−
⇌ HOCl + H2 O(l) 1.396
1.64

ClO
−
+ H2 O(l) + e
−
⇌
1

2
Cl 2 (g) + 2 OH
−
0.421 in 1 M NaOH
Chlorine E (V) Eo ́ (V)

ClO
−
ClO − + H2 O(l)
+ 2H
+
+e
+−2 e ⇌ Cl + 2 OH
⇌ ClO 2 (g) + H2 O(l)
− − −

1.175 0.890 in 1 M NaOH

3

HClO
ClO
− 2 + 2H
+ 3H
+ + 2−
+ 2e
e
+
⇌ HOCl + H2 O(l)
⇌ HClO2 + H2 O(l)
−
1.64
1.181
3

ClO
−
4
+ 2H
+
+ 2e
−
⇌ ClO
−
3
+ H2 O(l) 1.201

Chromium E (V) Eo ́ (V)

Cr
3+
+ 3e
−
⇌ Cr(s) –0.424

Cr
2+
+ 2e
−
⇌ Cr(s) –0.90

Cr 2 O
2−
7
+ 14 H
+
+ 6e
−
⇌ 2 Cr
3+
+ 7 H2 O(l) 1.36
−
CrO
2−
4
+ 4 H2 O(l) + 3 e
−
⇌ Cr(OH)
4
+ 4 OH
−
–0.13 in 1 M NaOH

Cobalt E (V) Eo ́ (V)

Co
2+
+ 2e
−
⇌ Co(s) –0.277

Co
3+
+ 3e
−
⇌ Co(s) 1.92

Co(NH )
3
3+

6
+e
−
⇌ Co(NH )
3
2+

6
0.1

Co(OH) (s) + e
3
−
⇌ Co(OH) (s) + OH
2
−
0.17

Co(OH) (s) + 2 e
2
−
⇌ Co(s) + 2 OH
−
–0.746

Copper E (V) Eo ́ (V)

Cu
+
+e
−
⇌ Cu(s) 0.520

Cu
2+
+e
−
⇌ Cu
+
0.159

Cu
2+
+ 2e
−
⇌ Cu(s) 0.3419

Cu
2+
+I
−
+e
−
⇌ CuI(s) 0.86

Cu
2+
+ Cl
−
+e
−
⇌ CuCl(s) 0.559

Fluorine E (V) Eo ́ (V)

F2 (g) + 2 H
+
+ 2e
−
⇌ 2HF(g) 3.053

F2 (g) + 2 e
−
⇌ 2F
−
2.87

Gallium E (V) Eo ́ (V)

Ga
3+
+ 3e
−
⇌ Ga(s) –0.529

Gold E (V) Eo ́ (V)

Au
+
+e
−
⇌ Au(s) 1.83

Au
3+
+ 2e
−
⇌ Au
+
1.36

Au
3+
+ 3e
−
⇌ Au(s) 1.52

AuCl
−
4
+ 3e
−
⇌ Au(s) + 4 Cl
−
1.002

Hydrogen E (V) Eo ́ (V)

35.8.3 https://chem.libretexts.org/@go/page/362971
2H
+
+ 2e
−
⇌ H2 (g) Hydrogen 0.00000 E (V) Eo ́ (V)

H2 O(l) +e ⇌
−
H2 (g) + OH
1 −
–0.828
2H
+
+ 2e
− 2
⇌ H2 (g) 0.00000

H2 O(l) + e
−
⇌
1
H2 (g) + OH
−
–0.828
2 Iodine E (V) Eo ́ (V)

I2 (s) + 2 e
−
⇌ 2I
−
0.5355

Iodine E (V) Eo ́ (V)

I
−
3
+ 2e
−
⇌ 3I
−
0.536

HIO + H
+
+ 2e
−
⇌ I
−
+ H2 O(l) 0.985

IO
−
3
+ 6H
+
+ 5e
−
⇌
1

2
I2 (s) + 3 H2 O(l) 1.195

IO
−
3
+ 3 H2 O(l) + 6 e
−
⇌ I
−
+ 6 OH
−
0.257

Iron E (V) Eo ́ (V)

Fe
2+
+ 2e
−
⇌ Fe(s) –0.44

Fe
3+
+ 3e
−
⇌ Fe(s) –0.037

0.70 in 1 M HCl
0.767 in 1 M HClO4
Fe
3+
+e
−
⇌ Fe
2+
0.771 0.746 in 1 M HNO3
0.68 in 1 M H2SO4
0.44 in 0.3 M H3PO4

Fe(CN)
3−

6
+e
−
⇌ Fe(CN)
4−

6
0.356

Fe(phen)
3+

3
+e
−
⇌ Fe(phen)
2+

3
1.147

Lanthanum E (V) Eo ́ (V)

La
3+
+ 3e
−
⇌ La(s) –2.38

Lead E (V) Eo ́ (V)

Pb
2+
+ 2e
−
⇌ Pb(s) –0.126

PbO2 (s) + 4 H
+
+ 2e
−
⇌ Pb
2+
+ 2 H2 O(l) 1.46

PbO2 (s) + SO
2−
4
+ 4H
+
+ 2e
−
1.690
⇌ PbSO 4 (s) + 2 H2 O(l)

PbSO 4 (s) + 2 e
−
⇌ Pb(s) + SO
2−
4
–0.356

Lithium E (V) Eo ́ (V)

Li
+
+e
−
⇌ Li(s) –3.040

Magnesium E (V) Eo ́ (V)

Mg
2+
+ 2e
−
⇌ Mg(s) –2.356

Mg(OH) (s) + 2 e
2
−
⇌ Mg(s) + 2 OH
−
–2.687

Manganese E (V) Eo ́ (V)

Mn
2+
+ 2e
−
⇌ Mn(s) –1.17

Mn
3+
+e
−
⇌ Mn
2+
1.5

MnO 2 (s) + 4 H
+
+ 2e
−
⇌ Mn
2+
+ 2 H2 O(l) 1.23

35.8.4 https://chem.libretexts.org/@go/page/362971
MnO
−
4
+ 4H
+
+ 3eManganese
−
⇌ MnO 2 (s) + 2 H2 O(l) 1.70 E (V) Eo ́ (V)

Mn
MnO +
2+−
+28
eH ⇌
4
+Mn(s)
5e ⇌ Mn
− + − 2+
+ 4 H2 O(l) –1.17
1.51

Mn
MnO +
3+−
+e2 H⇌
4
Mn + 3 e
2 O(l)
− 2+ −
⇌ MnO 2 (s) + 4 OH 1.5
0.60
−

MnO 2 (s) + 4 H
+
+ 2e
−
⇌ Mn
2+
+ 2 H2 O(l) 1.23
Mercury E (V) Eo ́ (V)

Hg
2+ −
MnO ++
2e4 H⇌+
4
Hg(l)
3e
− +
⇌ MnO 2 (s) + 2 H2 O(l)
−
0.8535
1.70

2 Hg
MnO
2+
−
4
+
+28e
H ⇌
+ Hg
5e
−+ 2+
−
2
⇌ Mn
2+
+ 4 H2 O(l) 0.911
1.51

MnO
−
4
+ 2 H2 O(l) + 3 e
−
⇌ MnO 2 (s) + 4 OH 0.60
−

Mercury E (V) Eo ́ (V)

Hg
2+
2
+ 2e
−
⇌ 2Hg(l) 0.7960

Hg Cl 2 (s) + 2 e
2
−
⇌ 2Hg(l) + 2 Cl
−
0.2682

HgO(s) + 2 H
+
+ 2e
−
⇌ Hg(l) + H2 O(l) 0.926

Hg Br2 (s) + 2 e
2
−
⇌ 2Hg(l) + 2 Br
−
1.392

Hg I2 (s) + 2 e
2
−
⇌ 2Hg(l) + 2 I
−
–0.0405

Molybdenum E (V) Eo ́ (V)

Mo
3+
+ 3e
−
⇌ Mo(s) –0.2

MoO2 (s) + 4 H
+
+ 4e
−
⇌ Mo(s) + 2 H2 O(l) –0.152

MoO
2−

4
+ 4 H2 O(l) + 6 e
−
⇌ Mo(s) + 8 OH
−
–0.913

Nickel E (V) Eo ́ (V)

Ni
2+
+ 2e
−
⇌ Ni(s) –0.257

Ni(OH) (s) + 2 e
2
−
⇌ Ni(s) + 2 OH
−
–0.72

Ni(NH )
3
2+
6
+ 2e
−
⇌ Ni(s) + 6 NH3 –0.49

Nitrogen E (V) Eo ́ (V)

N2 (g) + 5 H
+
+ 4e
−
⇌ N2 H
+
5
–0.23

N2 O(g) + 2 H
+
+ 2e
−
⇌ N2 (g) + H2 O(l) 1.77

2NO(g) + 2 H
+
+ 2e
−
⇌ N2 O(g) + H2 O(l) 1.59

HNO2 + H
+
+e
−
⇌ NO(g) + H2 O(l) 0.996

2 HNO2 + 4 H
+
+ 4e
−
⇌ N2 O(g) + 3 H2 O(l) 1.297

NO
−

3
+ 3H
+
+ 2e
−
⇌ HNO2 + H2 O(l) 0.94

Oxygen E (V) Eo ́ (V)

O2 (g) + 2 H
+
+ 2e
−
⇌ H2 O2 0.695

O2 (g) + 4 H
+
+ 4e
−
⇌ 2 H2 O(l) 1.229

H2 O2 + 2 H
+
+ 2e
−
⇌ 2 H2 O(l) 1.763

O2 (g) + 2 H2 O(l) + 4 e
−
⇌ 4 OH
−
0.401

O3 (g) + 2 H
+
+ 2e
−
⇌ O2 (g) + H2 O(l) 2.07

Phosphorous E (V) Eo ́ (V)

35.8.5 https://chem.libretexts.org/@go/page/362971
P(s, white) + 3 H Phosphorous
+
+ 3e ⇌ PH
−
3 (g) –0.063 E (V) Eo ́ (V)

H3 PO3 + 2 H
+
+ 2e
−
⇌ H3 PO2 + H2 O(l) –0.499
P(s,
H white) ++3+
3 PO4 + 2 H
H 2e
+− 3e ⇌PO
⇌ H3
PH3+
3
(g)
H2 O(l)
+ −
–0.063
–0.276
H3 PO3 + 2 H
+
+ 2e
−
⇌ H3 PO2 + H2 O(l) –0.499
Platinum E (V) Eo ́ (V)
H3 PO4 + 2 H
+
+ 2e
−
⇌ H3 PO3 + H2 O(l) –0.276
Pt
2+
+ 2e
−
⇌ Pt(s) 1.188

PtCl
2−
4
+ 2e
−
⇌ Pt(s) + 4 Cl
−
0.758

Potasium E (V) Eo ́ (V)

K
+
+e
−
⇌ K(s) –2.924

Ruthenium E (V) Eo ́ (V)

Ru
3+
+ 3e
−
⇌ Ru(s) 0.249

RuO 2 (s) + 4 H
+
+ 4e
−
⇌ Ru(s) + 2 H2 O(l) 0.68

Ru(NH )
3
3+
6
+e
−
⇌ Ru(NH )
3
2+
6
0.10

Ru(CN)
3−

6
+e
−
⇌ Ru(CN)
4−

6
0.86

Selenium E (V) Eo ́ (V)

Se(s) + 2 e
−
⇌ Se
2−
–0.67 in 1 M NaOH

Se(s) + 2 H
+
+ 2e
−
⇌ H2 Se(g) –0.115

H2 SeO 3 + 4 H
+
+ 4e
−
⇌ Se(s) + 3 H2 O(l) 0.74

SeO
3−
4
+ 4H
+
+e
−
⇌ H2 SeO 3 + H2 O(l) 1.151

Silicon E (V) Eo ́ (V)

SiF
2−
6
+ 4e
−
⇌ Si(s) + 6 F
−
–1.37

SiO 2 (s) + 4 H
+
+ 4e
−
⇌ Si(s) + 2 H2 O(l) –0.909

SiO 2 (s) + 8 H
+
+ 8e
−
⇌ SiH 4 (g) + 2 H2 O(l) –0.516

Silver E (V) Eo ́ (V)

Ag
+
+e
−
⇌ Ag(s) 0.7996

AgBr(s) + e
−
⇌ Ag(s) + Br
−
0.071

Ag C2 O4 (s) + 2 e
2
−
⇌ 2Ag(s) + C2 O
2−
4
0.47

AgCl(s) + e
−
⇌ Ag(s) + Cl
−
0.2223

AgI(s) + e
−
⇌ Ag(s) + I
−
–0.152

Ag S(s) + 2 e
2
−
⇌ 2Ag(s) + S
2−
–0.71

Ag(NH )
3
+
2
+e
−
⇌ Ag(s) + 2 NH3 –0.373

Sodium E (V) Eo ́ (V)

Na
+
+e
−
⇌ Na(s) –2.713

Strontium E (V) Eo ́ (V)

Sr
2+
+ 2e
−
⇌ Sr(s) –2.89

35.8.6 https://chem.libretexts.org/@go/page/362971
 Sulfur E (V) Eo ́ (V)

S(s) + 2 e
−
⇌ S
2−
–0.407

S(s) + 2 H
+
+ 2e
−
⇌ H2 S(g) 0.144

S2 O
2−
6
+ 4H
+
+ 2e
−
⇌ 2 H2 SO 3 0.569

S2 O
2−
8
+ 2e
−
⇌ 2 SO
2−
4
1.96

S4 O
2−
6
+ 2e
−
⇌ 2 S2 O
2−
3
0.080

2 SO
2−
3
+ 2 H2 O(l) + 2 e
−
⇌ S2 O
2−
4
+ 4 OH
−
–1.13

2 SO
2−
3
+ 3 H2 O(l) + 4 e
−
⇌ S2 O
2−
3
+ 6 OH
−
–0.576 in 1 M NaOH

2 SO
2−
4
+ 4H
+
+ 2e
−
⇌ S2 O
2−
6
+ 2 H2 O(l) –0.25

SO
2−
4
+ H2 O(l) + 2 e
−
⇌ SO
2−
3
+ 2 OH
−
–0.936

SO
2−
4
+ 4H
+
+ 2e
−
⇌ H2 SO 3 + H2 O(l) 0.172

Thallium E (V) Eo ́ (V)

3+ − + 1.25 in 1 M HClO4
Tl + 2e ⇌ Tl
0.77 in 1 M HCl

Tl
3+
+ 3e
−
⇌ Tl(s) 0.742

Tin E (V) Eo ́ (V)

Sn
2+
+ 2e
−
⇌ Sn(s) –0.19 in 1 M HCl

Sn
4+
+ 2e
−
⇌ Sn
2+
0.154 0.139 in 1 M HCl

Titanium E (V) Eo ́ (V)

Ti
2+
+ 2e
−
⇌ Ti(s) –0.163

Ti
3+
+e
−
⇌ Ti
2+
–0.37

Tungsten E (V) Eo ́ (V)

WO2 (s) + 4 H
+
+ 4e
−
⇌ W(s) + 2 H2 O(l) –0.119

WO3 (s) + 6 H
+
+ 6e
−
⇌ W(s) + 3 H2 O(l) –0.090

Uranium E (V) Eo ́ (V)

U
3+
+ 3e
−
⇌ U(s) –1.66

U
4+
+e
−
⇌ U
3+
–0.52

UO
+
2
+ 4H
+
+e
−
⇌ U
4+
+ 2 H2 O(l) 0.27

UO
2+
2
+e
−
⇌ UO
+
2
0.16

UO
2+
2
+ 4H
+
+ 2e
−
⇌ U
4+
+ 2 H2 O(l) 0.327

Vanadium E (V) Eo ́ (V)

V
2+
+ 2e
−
⇌ V(s) –1.13

V
3+
+e
−
⇌ V
2+
–0.255

VO
2+
+ 2H
+
+e
−
⇌ V
3+
+ H2 O(l) 0.337

VO
+
2
+ 2H
+
+e
−
⇌ VO
2+
+ H2 O(l) 1.000

35.8.7 https://chem.libretexts.org/@go/page/362971
 Zinc E (V) Eo ́ (V)

Zn
2+
2+
+ 2e
−
⇌ Zn(s) –0.7618

Zn(OH)
2−
2−
4
+ 2e
−
−
⇌ Zn(s) + 4 OH
−
–1.285

Zn(NH 3 )
3
2+
2+
4
+ 2e
−
⇌ Zn(s) + 4 NH3 –1.04

Zn(CN)
2−
2−
4
+ 2e
−
−
⇌ Zn(s) + 4 CN
−
–1.34

Polarographic Half-Wave Potentials

The following table provides E1/2 values for selected reduction reactions. Values are from Dean, J. A. Analytical Chemistry
Handbook, McGraw-Hill: New York, 1995.

Element E
1/2
(volts vs. SCE) Matrix

Al
3 +
(aq) + 3 e
− −
↽⇀
− Al(s) –0.5 0.2 M acetate (pH 4.5–4.7)

0.1 M KCl
Cd
2 +
(aq) + 2 e
− −
↽⇀
− Cd(s) –0.6 0.050 M H2SO4
1 M HNO3

3 + − −⇀
–0.35 (+3 ⟶ +2) 1 M NH4Cl plus 1 M NH3
Cr (aq) + 3 e ↽− Cr(s)
–1.70 (+2 ⟶ 0) 1 M NH4+/NH3 buffer (pH 8–9)

–0.5 (+3 ⟶ +2)

Co
3 +
(aq) + 3 e
− −
↽⇀
− Co(s) 1 M NH4Cl plus 1 M NH3
–1.3 (+2 ⟶ 0)

Co
2 +
(aq) + 2 e
− −
↽⇀
− Co(s) –1.03 1 M KSCN

0.1 M KSCN
2 + − −⇀
0.04 0.1 M NH4ClO4
Cu (aq) + 2 e ↽− Cu(s)
–0.22 1 M Na2SO4
0.5 M potassium citrate (pH 7.5)

–0.17 (+3 ⟶ +2)

Fe
3 +
(aq) + 3 e
− −
↽⇀
− Fe(s) 0.5 M sodium tartrate (pH 5.8)
–1.52 (+2 ⟶ 0)

Fe
3 +
(aq) + e
− −
↽⇀
− Fe
2 +
(aq) –0.27 0.2 M Na2C2O4 (pH < 7.9)

2 + − −⇀
–0.405 1 M HNO3
Pb (aq) + 2 e ↽− Pb(s)
–0.435 1 M KCl

Mn
2 +
(aq) + 2 e
− −
↽⇀
− Mn(s) –1.65 1 M NH4Cl plus 1 M NH3

2 + − −⇀
–0.70 1 M KSCN
Ni (aq) + 2 e ↽− Ni(s)
–1.09 1 M NH4Cl plus 1 M NH3

2 + − −⇀
–0.995 0.1 M KCl
Zn (aq) + 2 e ↽− Zn(s)
–1.33 1 M NH4Cl plus 1 M NH3

This page titled 35.8: Standard Reduction Potentials & Polarographic Half-wave Potentials is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by David Harvey.

35.8.8 https://chem.libretexts.org/@go/page/362971
35.9: Recommended Primary Standards
All compounds are of the highest available purity. Metals are cleaned with dilute acid to remove any surface impurities and rinsed
with distilled water. Unless otherwise indicated, compounds are dried to a constant weight at 110 oC. Most of these compounds are
soluble in dilute acid (1:1 HCl or 1:1 HNO3), with gentle heating if necessary; some of the compounds are water soluble.

Element Compound FW (g/mol) Comments

aluminum Al metal 26.982

antimony Sb metal 121.760

prepared by drying
KSbOC H O
4 4 6
324.92 KSbC H O
4 4
⋅
6
H O at 100 °C
1

2 2

and storing in a desicator

arsenic As metal 74.922

As O
2 3
197.84 toxic

barium BaCO
3
197.84 dry at 200 oC for 4 h

bismuth Bi metal 208.98

boron H BO
3 3
61.83 do not dry

bromine KBr 119.01

cadmium Cd metal 112.411

CdO 128.40

calcium CaCO
3
100.09

cerium Ce metal 140.116

(NH ) Ce(NO )
4 2 3 4
548.23

cesium Cs CO
2 3
325.82

Cs SO
2 4
361.87

chlorine NaCl 58.44

chromium Cr metal 51.996

K Cr O
2 2 7
294.19

cobalt Co metal 58.933

copper Cu metal 63.546

CuO 79.54

do not store solutions in glass

fluorine NaF 41.99
containers

iodine KI 166.00

KIO
3
214.00

iron Fe metal 55.845

lead Pb metal 207.2

lithium Li CO
2 3
73.89

magnesium Mg metal 24.305

manganese Mn metal 54.938

mercury Hg metal 200.59

molybdenum Mo metal 95.94

35.9.1 https://chem.libretexts.org/@go/page/362973
 Element Compound FW (g/mol) Comments

nickel Ni metal 58.693

phosphorous KH PO
2 4
136.09

P O
2 5
141.94

potassium KCl 74.56

K CO
2 3
138.21

K Cr O
2 2 7
294.19

KHC H O
8 4 2
204.23

silicon Si metal 28.085

SiO
2
60.08

silver Ag metal 107.868

AgNO
3
169.87

sodium NaCl 58.44

Na CO
2 3
106.00

Na C O
2 2 4
134.00

strontium SrCO
3
147.63

sulfur elemental S 32.066

K SO
2 4
174.27

Na SO
2 4
142.04

tin Sn metal 118.710

titanium Ti metal 47.867

tungsten W metal 183.84

uranium U metal 238.029

U O
3 8
842.09

vanadium V metal 50.942

zinc Zn metal 81.37

Sources:
Smith, B. W.; Parsons, M. L. J. Chem. Educ. 1973, 50, 679–681
Moody, J. R.; Greenburg, P. R.; Pratt, K. W.; Rains, T. C. Anal. Chem. 1988, 60, 1203A–1218A.

This page titled 35.9: Recommended Primary Standards is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by David Harvey.

35.9.2 https://chem.libretexts.org/@go/page/362973
35.10: Acronyms and Abbreviations
acronym or abbreviation name

AC alternating current

ADC analog-to-digital convertor

atomic emission spectroscopy

AES
auger electron spectroscopy

AFM atomic force microscopy

AFS atomic fluorescence spectroscopy

ASV anodic stripping voltammetry

ATF attenuated total reflectance

BPC binary pulse counter

CCD charge-coupled device

CE capillary electrophoresis

CEC capillary electrochromatography

CGE capillary gel electrophoresis

CI chemical ionization

CID charge injection device

COSY correlation spectroscopy

CZE capillary zone electrophoresis

CSV cathodic stripping voltammetry

CV cyclic voltammetry

DAC digital-to-analog convertor

DAS diode array spectrometer

DC direct current

DCP direct current plasma

DCS differential centrifugal separation

DL detection limit

DME dropping mercury electrode

DPP differential pulse polarography

DRIFT diffuse reflectance infrared Fourier transform

DSC differential scanning calorimetry

DTA differential thermal analysis

electron capture detector

ECD
electrochemical detector

EDL electrical double layer

EM electromagnetic

ESR electron spin resonance

FAAS flame atomic absorption spectrometry

FIA flow-injection analysis

35.10.1 https://chem.libretexts.org/@go/page/362975
 acronym or abbreviation name

flame ionization detector

FID
free induction decay

FIR far infrared

FT Fourier transform

GC gas chromatography

GC-MS gas chromatography-mass spectrometry

GDMS glow-discharge mass spectrometry

HCL hollow cathode lamp

HETCOR heteronuclear correlation spectroscopy

HETP height-equivalent of theoretical plate

HMDE hanging mercury drop electrode

HPLC high performance liquid chromatography

ICP inductively coupled plasma

IEC ion-exchange chromatography

IR infrared

IS internal standard

ISE ion-selective electrode

LC liquid chromatography

LOI limit-of-identification

LOQ limit-of-quantification

LSV linear sweep voltammetry

MALDI matrix-assisted laser desorption ionization

MEKC micellar electrokinetic capillary chromatography

MS mass spectrometry

MW molecular weight

M/Z mass-to-charge ratio

NAA neutron activation analysis

NHE normal hydrogen electrode

NIR near infrared

NMR nuclear magnetic resonance

NOSEY nuclear overhauser and exchange spectroscopy

NP normal polarography

NPP normal pulse polarography

ORD optical rotary dispersion

PDA photodiode array

PLOT porous-layer open tubular column

PP pulse polarography

RPC reverse phase chromatography

35.10.2 https://chem.libretexts.org/@go/page/362975
 acronym or abbreviation name

RRM resonance Raman spectrosocopy

SCE saturated calomel electrode

SCOT support-coated open tubular column

SEC size-exclusion chromatography

SEM scanning electron microscopy

SERS surface-enhanced Raman spectroscopy

SFC supercritical fluid chromatography

SHE standard hydrogen electrode

SIMS secondary ion mass spectrometry

SMDE static mercury drop electrode

S/R (or SNR) signal-to-noise ratio

SSMS spark-source mass spectrometry

STM scanning tunneling microscopy

TCD thermal conductivity detector

TGA thermal gravimetric analysis

TOCOSY total correlation spectroscopy

TOF time-of-flight

UV ultraviolet

UV/Vis ultraviolet/visible

Vis visible

WCOT wall-coated open tubular column

XPS x-ray photoelectron spectroscopy

XRF x-ray fluorescence

This page titled 35.10: Acronyms and Abbreviations is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
David Harvey.

35.10.3 https://chem.libretexts.org/@go/page/362975
Index
A E Liquid Junction Potentials
absorbance Effective Nuclear Charge 22.2: Potentials in Electroanalytical Cells
13.1: Transmittance and Absorbance 2.4: Semiconductors
attenuated total reflectance spectrometry Electroosmotic Mobility M
17.2: Mid-Infrared Reflection Spectrometry 30.2: Capillary Electrophoresis MEKC
electrophoresis 30.3: Applications of Capillary Electrophoresis
B 30.1: An Overview of Electrophoresis Micellar electrokinetic capillary
Beer’s Law Electrophoretic Mobility chromatography
13.2: Beer's Law 30.2: Capillary Electrophoresis 30.3: Applications of Capillary Electrophoresis
binary numbers emission spectra modulation
4.2: Working With Binary Numbers 6.4: Emission and Absorbance Spectra 5.3: Signal-to-Noise Enhancement
energy states molar absorptivity
C 6.4: Emission and Absorbance Spectra 14.1: What is Molar Absorptivity?
calomel electrodes environmental noise monochromator
23.1: Reference Electrodes 5.2: Sources of Instrumental Noise 7.3: Wavelength Selectors
capillary electrochromatography
30: Capillary Electrophoresis and Capillary F N
Electrochromatography fiber optics NAA
30.3: Applications of Capillary Electrophoresis 7.6: Fiber Optics 32.3: Neutron Activation Methods
Capillary Electrophoresis filtering Neutron Activation Analysis
30: Capillary Electrophoresis and Capillary 5.3: Signal-to-Noise Enhancement 32.3: Neutron Activation Methods
Electrochromatography
30.2: Capillary Electrophoresis flicker noise noise
capillary gel electrophoresis 5.2: Sources of Instrumental Noise 5.1: The Signal-to-Noise Ratio
30.3: Applications of Capillary Electrophoresis flow injection analysis nuclear magnetic resonance spectroscopy
Capillary Zone Electrophoresis 33.2: Flow-Injection Analysis 19.1: Theory of Nuclear Magnetic Resonance
30.3: Applications of Capillary Electrophoresis fluorescence
cell potential 15.1: Theory of Fluorescence and Phosphorescence O
22.4: Calculation of Cell Potentials from Electrode fluorescence quantum yield operational amplifier
Potentials 15.1: Theory of Fluorescence and Phosphorescence 3.1: Operational Amplifiers
CGE fluorometer optical filter
30.3: Applications of Capillary Electrophoresis 15.2: Instruments for Measuring Fluorescence and 7.3: Wavelength Selectors
chemiluminscence Phosphorescence
optrodes
15.4: Chemiluminscence 7.6: Fiber Optics
column efficiency G
26.3: Zone Broadening and Column Efficiency glow discharge mass spectrometry P
coulometry 11.4: Other Forms of Atomic Mass Spectrometry
partition chromatography
24: Coulometry 28.4: Partition Chromatography
24.1: Introduction to Coulometry I phosphorescence
Cyclic Voltammetry infrared spectromters 15.1: Theory of Fluorescence and Phosphorescence
25.4: Cyclic Voltammetry 16.3: Infrared Instruments phosphorescence quantum yield
infrared transducer 15.1: Theory of Fluorescence and Phosphorescence
D 16.2: Infrared Sources and Transducers photoelectric effect
differential amplifier interference filter 6.3: Quantum Mechanical Properties of
5.3: Signal-to-Noise Enhancement 7.3: Wavelength Selectors Electromagnetic Radiation
Differential Scanning Calorimetry interferometer polarography
31.2: Differential Thermal Analysis and Differential 7.7: Fourier Transform Optical Spectroscopy 25.5: Polarography
Scanning Calorimetry internal reflection potentiometer
differential thermal analysis 17.2: Mid-Infrared Reflection Spectrometry 23.5: Instruments for Measuring Cell Potentials
31.2: Differential Thermal Analysis and Differential iR drop
Scanning Calorimetry
pulse polarography
22.5: Currents in Electrochemical Cells 25.5: Polarography
diffuse reflectance spectrometry
isotope dilution methods
17.2: Mid-Infrared Reflection Spectrometry
digital data
32.4: Isotope Dilution Methods Q
4.1: Analog and Digital Data quantitative potentiometry
distribution constant
L 23.6: Quantitative Potentiometry
26.2: Migration Rates of Solutes
laser
Doppler broadening 7.2: Sources of Radiation R
8.1: Optical Atomic Spectra
line spectra radioactive decay
6.4: Emission and Absorbance Spectra
Dynamic Light Scattering 32.1: Radioactive Isotopes
34.5: Measuring Particle Size Using Light Scattering
Linear Sweep Voltammetry Radioactive Isotopes
25.3: Linear Sweep Voltammetry 32.1: Radioactive Isotopes

1 https://chem.libretexts.org/@go/page/332632
Raman Spectroscopy S T
18.1: Theory of Raman Spectroscopy sedimentation thermal noise
Rayleigh scattering 34.3: Measuring Particle Size by Sedimentation 5.2: Sources of Instrumental Noise
34.5: Measuring Particle Size Using Light Scattering selectivity factor thermogravimetry
Reference Electrodes 26.2: Migration Rates of Solutes 31.1: Thermogravimetry
23.1: Reference Electrodes shot noise transmission
retention factor 5.2: Sources of Instrumental Noise 13.1: Transmittance and Absorbance
26.2: Migration Rates of Solutes slit width
Retention time 7.3: Wavelength Selectors V
26.2: Migration Rates of Solutes standard hydrogen electrode
27.1: Principles of Gas Chromatography
van Deemter equation
22.3: Electrode Potentials 27.1: Principles of Gas Chromatography
retention volume
supercritical fluid chromatography 28.2: Column Efficiency in Liquid Chromatography
27.1: Principles of Gas Chromatography
29: Supercritical Fluid Chromatography
X
XPS
21.2: Spectroscopic Surface Methods

2 https://chem.libretexts.org/@go/page/332632
Detailed Licensing
Overview
Title: Instrumental Analysis (LibreTexts)
Webpages: 199
Applicable Restrictions: Noncommercial
All licenses found:
CC BY-NC-SA 4.0: 97.5% (194 pages)
Undeclared: 2.5% (5 pages)

By Page
Instrumental Analysis (LibreTexts) - CC BY-NC-SA 4.0 4.3: Cleaning Up Signals and Counting Events - CC
Front Matter - CC BY-NC-SA 4.0 BY-NC-SA 4.0
TitlePage - CC BY-NC-SA 4.0 5: Signals and Noise - CC BY-NC-SA 4.0
InfoPage - CC BY-NC-SA 4.0 5.1: The Signal-to-Noise Ratio - CC BY-NC-SA 4.0
Table of Contents - Undeclared 5.2: Sources of Instrumental Noise - CC BY-NC-SA
Licensing - Undeclared 4.0
1: Introduction - CC BY-NC-SA 4.0 5.3: Signal-to-Noise Enhancement - CC BY-NC-SA
4.0
1.1: Classification of Analytical Methods - CC BY-
NC-SA 4.0 6: An Introduction to Spectrophotometric Methods - CC
1.2: Types of Instrumental Methods - CC BY-NC-SA BY-NC-SA 4.0
4.0 6.1: General Properties of Electromagnetic Radiation
1.3: Instruments For Analysis - CC BY-NC-SA 4.0 - CC BY-NC-SA 4.0
1.4: Selecting an Analytical Method - CC BY-NC-SA 6.2: Wave Properties of Electromagnetic Radiation -
4.0 CC BY-NC-SA 4.0
1.5: Calibration of Instrumental Methods - CC BY- 6.3: Quantum Mechanical Properties of
NC-SA 4.0 Electromagnetic Radiation - CC BY-NC-SA 4.0
2: Electrical Components and Circuits - CC BY-NC-SA 6.4: Emission and Absorbance Spectra - CC BY-NC-
4.0 SA 4.0
2.1: Basic Terminology and Laws of Electricity - CC 6.5: Quantitative Considerations - CC BY-NC-SA 4.0
BY-NC-SA 4.0 7: Components of Optical Instruments - CC BY-NC-SA
2.2: Direct Current (DC) Circuits - CC BY-NC-SA 4.0 4.0
2.3: Alternating Current Circuits - CC BY-NC-SA 4.0 7.1: General Design of Optical Instruments - CC BY-
2.4: Semiconductors - CC BY-NC-SA 4.0 NC-SA 4.0
3: Operational Amplifiers - CC BY-NC-SA 4.0 7.2: Sources of Radiation - CC BY-NC-SA 4.0
3.1: Operational Amplifiers - CC BY-NC-SA 4.0 7.3: Wavelength Selectors - CC BY-NC-SA 4.0
3.2: Operational Amplifier Circuits - CC BY-NC-SA 7.4: Sample Containers - CC BY-NC-SA 4.0
4.0 7.5: Radiation Transducers - CC BY-NC-SA 4.0
3.3: Amplification and Measurement of Signals - CC 7.6: Fiber Optics - CC BY-NC-SA 4.0
BY-NC-SA 4.0 7.7: Fourier Transform Optical Spectroscopy - CC
3.4: Mathematical Operations Using Operational BY-NC-SA 4.0
Amplifiers - CC BY-NC-SA 4.0 8: An Introduction to Optical Atomic Spectroscopy - CC
BY-NC-SA 4.0
4: Analog and Digital Data - CC BY-NC-SA 4.0
8.1: Optical Atomic Spectra - CC BY-NC-SA 4.0
4.1: Analog and Digital Data - CC BY-NC-SA 4.0
8.2: Atomization Methods - CC BY-NC-SA 4.0
4.2: Working With Binary Numbers - CC BY-NC-SA
8.3: Sample Introduction Methods - CC BY-NC-SA
4.0
4.0
9: Atomic Absorption Spectrometry - CC BY-NC-SA 4.0

1 https://chem.libretexts.org/@go/page/417082
 9.1: Sample Atomization Techniques - CC BY-NC-SA 15.4: Chemiluminscence - CC BY-NC-SA 4.0
4.0 15.5: Evaluation of Molecular Luminescence - CC
9.2: Atomic Absorption Instrumentation - CC BY-NC- BY-NC-SA 4.0
SA 4.0 16: An Introduction to Infrared Spectrometry - CC BY-
9.3: Interferences in Absorption Spectroscopy - CC NC-SA 4.0
BY-NC-SA 4.0 16.1: Theory of Infrared Absorption Spectrometry -
9.4: Atomic Absorption Techniques - CC BY-NC-SA CC BY-NC-SA 4.0
4.0 16.2: Infrared Sources and Transducers - CC BY-NC-
10: Atomic Emission Spectrometry - CC BY-NC-SA 4.0 SA 4.0
10.1: Emission Spectroscopy Based on Flame and 16.3: Infrared Instruments - CC BY-NC-SA 4.0
Plasma Sources - CC BY-NC-SA 4.0 17: Applications of Infrared Spectrometry - CC BY-NC-
10.2: Emission Spectroscopy Based on Arc and Spark SA 4.0
Sources - CC BY-NC-SA 4.0
17.1: Mid-Infrared Absorption Spectometry - CC BY-
11: Atomic Mass Spectrometry - CC BY-NC-SA 4.0 NC-SA 4.0
11.1: General Features of Atomic Mass Spectrometry 17.2: Mid-Infrared Reflection Spectrometry - CC BY-
- CC BY-NC-SA 4.0 NC-SA 4.0
11.2: Mass Spectrometers - CC BY-NC-SA 4.0 17.3: Near-Infrared and Far-Infrared Spectroscopy -
11.3: Inductively Coupled Plasma Mass Spectrometer CC BY-NC-SA 4.0
- CC BY-NC-SA 4.0 18: Raman Spectroscopy - CC BY-NC-SA 4.0
11.4: Other Forms of Atomic Mass Spectrometry - 18.1: Theory of Raman Spectroscopy - CC BY-NC-SA
CC BY-NC-SA 4.0 4.0
12: Atomic X-Ray Spectrometry - CC BY-NC-SA 4.0 18.2: Instrumentation - CC BY-NC-SA 4.0
12.1: Fundamental Principles - CC BY-NC-SA 4.0 18.3: Applications of Raman Spectroscopy - CC BY-
12.2: Instrument Components - CC BY-NC-SA 4.0 NC-SA 4.0
12.3: Atomic X-Ray Fluorescence Methods - CC BY- 18.4: Other Types of Raman Spectroscopy - CC BY-
NC-SA 4.0 NC-SA 4.0
12.4: Other X-Ray Methods - CC BY-NC-SA 4.0 19: Nuclear Magnetic Resonance Spectroscopy - CC BY-
13: Introduction to Ultraviolet/Visible Absorption NC-SA 4.0
Spectrometry - CC BY-NC-SA 4.0 19.1: Theory of Nuclear Magnetic Resonance - CC
13.1: Transmittance and Absorbance - CC BY-NC-SA BY-NC-SA 4.0
4.0 19.2: Environmental Effects on NMR Spectra - CC
13.2: Beer's Law - CC BY-NC-SA 4.0 BY-NC-SA 4.0
13.3: Effect of Noise on Transmittance and 19.3: NMR Spectrometers - CC BY-NC-SA 4.0
Absorbance Measurements - CC BY-NC-SA 4.0 19.4: Applications of Proton NMR - CC BY-NC-SA
13.4: Instrumentation - CC BY-NC-SA 4.0 4.0
14: Applications of Ultraviolet/Visible Molecular 19.5: Carbon-13 NMR - CC BY-NC-SA 4.0
Absorption Spectrometry - CC BY-NC-SA 4.0 19.6: Two-Dimensional Fourier Transform NMR -
14.1: What is Molar Absorptivity? - Undeclared CC BY-NC-SA 4.0
14.2: Absorbing Species - CC BY-NC-SA 4.0 20: Molecular Mass Spectrometry - CC BY-NC-SA 4.0
14.3: Qualitative and Characterization Applications - 20.1: Molecular Mass Spectra - CC BY-NC-SA 4.0
CC BY-NC-SA 4.0 20.2: Ion Sources - CC BY-NC-SA 4.0
14.4: Quantitative Applications - CC BY-NC-SA 4.0 20.3: Mass Spectrometers - CC BY-NC-SA 4.0
14.5: Photometric Titrations - CC BY-NC-SA 4.0 20.4: Applications of Molecular Mass Spectrometry -
15: Molecular Luminescence - CC BY-NC-SA 4.0 CC BY-NC-SA 4.0
15.1: Theory of Fluorescence and Phosphorescence - 21: Surface Characterization by Spectroscopy and
CC BY-NC-SA 4.0 Microscopy - CC BY-NC-SA 4.0
15.2: Instruments for Measuring Fluorescence and 21.1: Introduction to the Study of Surfaces - CC BY-
Phosphorescence - CC BY-NC-SA 4.0 NC-SA 4.0
15.3: Applications and Photoluminescence methods - 21.2: Spectroscopic Surface Methods - CC BY-NC-SA
CC BY-NC-SA 4.0 4.0

2 https://chem.libretexts.org/@go/page/417082
 21.3: Scanning Electron Microscopy - CC BY-NC-SA 26.3: Zone Broadening and Column Efficiency - CC
4.0 BY-NC-SA 4.0
21.4: Scanning Probe Microscopes - CC BY-NC-SA 26.4: Optimization and Column Performance - CC
4.0 BY-NC-SA 4.0
22: An Introduction to Electroanalytical Chemistry - CC 26.5: Summary of Important Relationships for
BY-NC-SA 4.0 Chromatography - CC BY-NC-SA 4.0
22.1: Electrochemical Cells - CC BY-NC-SA 4.0 26.6: Applications of Chromatography - CC BY-NC-
22.2: Potentials in Electroanalytical Cells - CC BY- SA 4.0
NC-SA 4.0 27: Gas Chromatography - CC BY-NC-SA 4.0
22.3: Electrode Potentials - CC BY-NC-SA 4.0 27.1: Principles of Gas Chromatography - CC BY-
22.4: Calculation of Cell Potentials from Electrode NC-SA 4.0
Potentials - CC BY-NC-SA 4.0 27.2: Instruments for Gas Chromatography - CC BY-
22.5: Currents in Electrochemical Cells - CC BY-NC- NC-SA 4.0
SA 4.0 27.3: Gas Chromatographic Columns and Stationary
22.6: Types of Electroanalytical Methods - CC BY- Phases - CC BY-NC-SA 4.0
NC-SA 4.0 27.4: Applications of Gas Chromatography - CC BY-
23: Potentiometry - CC BY-NC-SA 4.0 NC-SA 4.0
23.1: Reference Electrodes - CC BY-NC-SA 4.0 28: High-Performance Liquid Chromatography - CC BY-
23.2: Metallic Indicator Electrodes - CC BY-NC-SA NC-SA 4.0
4.0 28.1: Scope of HPLC - CC BY-NC-SA 4.0
23.3: Membrane Ion-Selective Electrodes - CC BY- 28.2: Column Efficiency in Liquid Chromatography -
NC-SA 4.0 CC BY-NC-SA 4.0
23.4: Molecular-Selective Electrode Systems - CC 28.3: Instruments for Liquid Chromatography - CC
BY-NC-SA 4.0 BY-NC-SA 4.0
23.5: Instruments for Measuring Cell Potentials - CC 28.4: Partition Chromatography - CC BY-NC-SA 4.0
BY-NC-SA 4.0 28.5: Adsorption Chromatography - CC BY-NC-SA
23.6: Quantitative Potentiometry - CC BY-NC-SA 4.0 4.0
24: Coulometry - CC BY-NC-SA 4.0 28.6: Ion-Exchange Chromatography - CC BY-NC-SA
24.1: Introduction to Coulometry - CC BY-NC-SA 4.0 4.0
24.2: Controlled-Potential Coulometry - CC BY-NC- 28.7: Size-Exclusion Chromatography - CC BY-NC-
SA 4.0 SA 4.0
24.3: Controlled-Current Coulometry - CC BY-NC-SA 29: Supercritical Fluid Chromatography - CC BY-NC-SA
4.0 4.0
25: Voltammetry - CC BY-NC-SA 4.0 29.1: Properties of Supercritical Fluids - CC BY-NC-
SA 4.0
25.1: Potential Excitation Signals and Currents in
29.2: Supercritical Fluid Chromatography - CC BY-
Voltammetry - CC BY-NC-SA 4.0
NC-SA 4.0
25.2: Voltammetric Instrumentation - CC BY-NC-SA
4.0 30: Capillary Electrophoresis and Capillary
25.3: Linear Sweep Voltammetry - CC BY-NC-SA 4.0 Electrochromatography - CC BY-NC-SA 4.0
25.4: Cyclic Voltammetry - CC BY-NC-SA 4.0 30.1: An Overview of Electrophoresis - CC BY-NC-
25.5: Polarography - CC BY-NC-SA 4.0 SA 4.0
25.6: Stripping Methods - CC BY-NC-SA 4.0 30.2: Capillary Electrophoresis - CC BY-NC-SA 4.0
25.7: Applications of Voltammetry - CC BY-NC-SA 30.3: Applications of Capillary Electrophoresis - CC
4.0 BY-NC-SA 4.0
26: Introduction to Chromatographic Separations - CC 31: Thermal Methods - CC BY-NC-SA 4.0
BY-NC-SA 4.0 31.1: Thermogravimetry - CC BY-NC-SA 4.0
26.1: A General Description of Chromatography - CC 31.2: Differential Thermal Analysis and Differential
BY-NC-SA 4.0 Scanning Calorimetry - CC BY-NC-SA 4.0
26.2: Migration Rates of Solutes - CC BY-NC-SA 4.0 32: Radiochemical Methods - CC BY-NC-SA 4.0
32.1: Radioactive Isotopes - CC BY-NC-SA 4.0

3 https://chem.libretexts.org/@go/page/417082
 32.2: Instrumentation - CC BY-NC-SA 4.0 35.1: Evaluation of Analytical Data - CC BY-NC-SA
32.3: Neutron Activation Methods - CC BY-NC-SA 4.0
4.0 35.2: Single-Sided Normal Distribution - CC BY-NC-
32.4: Isotope Dilution Methods - CC BY-NC-SA 4.0 SA 4.0
33: Automated Methods of Analysis - CC BY-NC-SA 4.0 35.3: Critical Values for t-Test - CC BY-NC-SA 4.0
33.1: Overview of Automated Methods of Analysis - 35.4: Critical Values for F-Test - CC BY-NC-SA 4.0
CC BY-NC-SA 4.0 35.5: Critical Values for Dixon's Q-Test - CC BY-NC-
33.2: Flow-Injection Analysis - CC BY-NC-SA 4.0 SA 4.0
33.3: Other Automated Methods of Analysis - CC BY- 35.6: Critical Values for Grubb's Test - CC BY-NC-SA
NC-SA 4.0 4.0
35.7: Activity Coefficients - CC BY-NC-SA 4.0
34: Particle Size Determination - CC BY-NC-SA 4.0
35.8: Standard Reduction Potentials & Polarographic
34.1: Overview - CC BY-NC-SA 4.0 Half-wave Potentials - CC BY-NC-SA 4.0
34.2: Measuring Particle Size Using Sieves - CC BY- 35.9: Recommended Primary Standards - CC BY-NC-
NC-SA 4.0 SA 4.0
34.3: Measuring Particle Size by Sedimentation - CC 35.10: Acronyms and Abbreviations - CC BY-NC-SA
BY-NC-SA 4.0 4.0
34.4: Measuring Particle Size Using Image Analysis -
Back Matter - CC BY-NC-SA 4.0
CC BY-NC-SA 4.0
Index - CC BY-NC-SA 4.0
34.5: Measuring Particle Size Using Light Scattering
Glossary - CC BY-NC-SA 4.0
- Undeclared
Detailed Licensing - Undeclared
35: Appendicies - CC BY-NC-SA 4.0

4 https://chem.libretexts.org/@go/page/417082