CHEM 2310

Fundamentals of Analytical Chemistry

Lecture 5:
Quality Assurance and Calibration Methods
Reading: Chapter 5

Acknowledgement: Prof. Ian Williams, Jianzhen Yu 1

 Outline
 The need for quality assurance
 Basics of quality assurance
 Validation of an analytical procedure
 Limit of Detection
 Limit of Quantification
 Calibration methods for instrumental analysis
 Standard addition
 Internal standards

2
 The Need for Quality Assurance
9 national
181 Labs institutes

Results of on river water with 62.3 ±1.3 nM Pb. a) 181 analytical labs b) results from
9 national institutes close to certified range. Useful to check ‘blind’ samples of this type. 3
Note vertical bars represent error bars – maybe 2 or 3 standard deviations
 The Need for Quality Assurance

Proficiency Testing

4
 Basics of Quality Assurance (QA)

• The things we do to get the answer with

sufficient accuracy and precision to support
subsequent decisions
• The objective of QA program is to control
systematic and random sources of error.
• Two principal components of a QA program:
quality control (QC) and quality assessment.

5
 Basics of Quality Assurance (QA)

• The most important facet of quality control is a set of

written directives describing the relevant laboratory-
specific, technique-specific, sample-specific, method-
specific, and protocol-specific operations.
– Good laboratory practices (GLPs)
– Good measurement practices (GMPs)
– Standard operating procedure (SOP)
• The goals of quality assessment are: 1. to determine
when an analysis has reached a state of statistical control,
2. to detect when an analysis falls out of statistical control,
and 3. to suggest the reasons for this loss of statistical
control.
– Internal methods: Analysis of duplicate samples, blanks,
standards, Spike recoveries
– External methods: certification of a laboratory (HOKLAS)
– Evaluating QA data: control chart is the principal tool for quality
assessment.
6
 Quality Assurance Process

Several steps required, especially to specify the aim and level of assurance
Some methodologies will be outlined, but for detailed discussion see Chapter 5.

7
Standard Operation procedure

8
 Use of Control charts
A control chart is a graph showing the time-dependent change in the results of
an analysis that is used to monitor whether an analysis is in a state of statistical
control. It is a visual representation of confidence intervals for a Gaussian
distribution.
For quality assurance, n = 5 replicate
quality control samples made from
synthetic urine spiked with perchlorate
are measured every day.
The spike contains µ = 4.92 ng/mL, the
population standard deviation from
many analyses over a long time is  =
0.40 ng/mL.
Warning line, Action line

Control chart for ClO4- in urine

showing the mean value of the five
samples observed each day over a
series of days.
9
 Use of Control charts (continued)
The following conditions are considered to be so unlikely that,
if they occur, the process should be shut down for
troubleshooting:

• One single observation outside the action lines

• 2 out of 3 consecutive measurements between the warning
and the action lines
• 7 consecutive measurements all above or all below the
center line
• 6 consecutive measurements all steadily increasing or all
steadily decreasing, wherever they are located.
• 14 consecutive measurements alternating up and down,
regardless of where they are located
• An obvious nonrandom pattern.

10
 Validation of an Analytical Procedure
Need to show that a procedure meets specifications and is acceptable for its
intended purpose based on:

Selectivity ability to distinguish the analyte from other species

Accuracy closeness to true value, sufficient for purpose

Precision reproducibility of results from multiple measurements

Linearity from calibration curve – signal proportional to concentration

Range concentration interval for which above are acceptable

(note linear range can be extended to a dynamic range using polynomial curve
fitting)

Robustness unaffected by small changes in conditions

Detection limit smallest quantity of analyte significantly different from a blank

11
 Gauge accuracy and precision

• To gauge accuracy
– Calibration checks
– Fortification recoveries
– Quality control samples/blind sample
– Blanks, e.g. reagent bk, method bk, field bk

• To gauge precision
– Replicate samples
– Replicate portions of same sample

12
Spike Recovery

13
 Limits of Detection and Quantitation
Detection limit smallest quantity of analyte significantly different from a blank

1. Estimate limit and make sample ca (1-5)x the limit

2. Measure its signal n > or =7 times and compute s

3. Measure signal from n blanks

4. Minimum detectable signal is ydl = yblank + 3s

5. Detection limit = minimum detection concentration

= 3s /m where m is the slope from calibration curve

6. A signal 10x the ‘noise’ level is defined as the

Lower limit of quantitation = 10s /m

(Note a signal at 3x noise level is detectable but

Still too small to be accurately determined)

14
Steeper slope of calibration curve, bigger m, Low detection limit, high sensitivity
Limits of Detection and Quantitation

15
Example: Detection limit calculation

16
Example: Detection limit calculation

17
 Beyond the Limit..

A number of studies have indicated the danger of synthetic trans fat in the diet.
Unscrupulous companies now indicate 0g trans fat if value less than 0.5g (reporting
limit)!!! Literally these are NO SIG FIGS.
“Not detected” does not mean that analyte is not observed; it simply 18
means that analyte may be present below a prescribed level
Step 5. Detection of Analytes – Chromatographic Results

The peaks of two standard

samples of equal concentration
Theobromine and Caffeine

Experimental result on left – can know peaks are due to caffeine

and theobromine by running authentic samples (above).

Response (based on uv absorption) varies so the peak heights

must be calibrated versus known concs. for each analyte
19
20
21
 Standard Addition Calibration Curve
Calibration curves are usually used to determine relation between signal and
analyte concentration in a chemical analysis. Where these would be inappropriate
or unreliable we can use either standard addition or internal standards.

Standard addition: add a known quantity of analyte and measure increase in signal.
Standard additions are most appropriate when the sample matrix is complex and
difficult to reproduce in standard solutions.

Matrix is everything in
the sample other than
analyte.

Shall use a matrix-matched calibration curve

Matrix effect

MALDI MS spectra of a tryptic peptide digest 22

Step 5. Detection of Analytes – Chromatographic Results

YES
Its caffeine

Std addition
??

23
Standard Addition Calibration Curve:
Implementation

24
 Single Standard Addition
A complete calibration curve is time consuming; a single Standard addition: add
a known quantity of analyte and measure increase in signal.

25
Example of a Standard Addition

26
 Graphical Procedure for Standard Addition

[X]i
The use of dilution factors is complex: often
all solutions in a standard addition
experiment are made up to the same total
volume through addition of solvent.

Note standard addition useful where

samples are complex (eg blood) and
‘matrix effects’ apply. Addition of conc.
standard doesn’t affect the matrix much
27
 Internal Standards
An internal standard is a known amount of compound, different from
the analyte, that is added to the unknown. The signal of the two are
then compared to find out how much analyte is present

Internal standards are useful when instrument response

varies from run to run, e. g. in gas or liquid chromatography
or when sample loss may occur in manipulations

A calibration curve is only accurate for the one set of conditions

under which it was obtained. However, the relative response of the
detector to the analyte and standard is usually constant over a
wide range of conditions. 28
29
30
31
 AT/AIS

120000
Response
100000

80000

60000

40000 [CT]
20000
T
0
0 20 40 60 80 100 120
IS
Concentration

32
Time, min
120000
Response
100000

80000

60000

40000

20000
T
0
0 20 40 60 80 100 120

Isotope-labeled IS
Concentration

33
Time, min
 Co-enzyme Q10: Ubiquinone
Important ‘vitamin’ present in mitochondria – Has quinone, semi-quinone and
diol forms i.e. 3 oxidation states. Takes part in e- transport & cellular respiration

Tail composed of isoprene units: for Q10 this should be 10; in HPLC analysis
can analyze this coenzyme family in various organisms and their mutants,
Q6 was used as an ‘internal standard’ - Jonassen et al PNAS 2001, p421

34
Analysis of Q content in Dauer Larvae.

See Jonassen T et al. PNAS 2001;98:421-426 35

 The Worm that turned..
out to be quite intriguing ! Study of nematode worms has been intense since
they can adapt to many environments and survive severe conditions (especially in
the stage as Dauer Larvae). Understanding coenzyme Q in these species and their
mutants may be of considerable importance

http://www.scientificamerican.com/article.cfm?id=longevity-gene-caloric-restriction

Scientists are also interested in ageing and have been able to modify nematode
longevity by factor of 7x. One interesting development in particular is that a gene
linking lower caloric intake to increased longevity was examined in nematodes
36
http://www.anti-agingfirewalls.com/2010/09/26/new-extraordinary-longevity-lessons-from-the-nematode/
37