HPLC

A Practical Guide
RSC Chromatography Monographs
Series Editor: Roger M. Smith, Loughborough University,
Loughborough, UK
Advisory Panel: J.C. Berridge, Sandwich, UK; G.B. Cox, Illkirch, France;
I.S. Lurie, Virginia, USA; P.J. Schoenmaker, Amsterdam, The Netherlands;
C.F. Simpson, London, U E , G.G. Wallace, Wollongong, Australia.

This series is designed for the individual practising chromatographer, providing

guidance and advice on a wide range of chromatographic techniques with the
emphasis on important practical aspects of the subject.

Packed Column SFC

by T.A. Berger, Berger Instruments, Newark,
Delaware, U S A
Chromatographic Integration Methods, Second Edition
by Norman Dyson, Dyson Instruments Ltd, U K
Separation of Fullerenes by Liquid Chromatography
edited by IS.Jinno, Toyohashi University of Technology, Japan
HPLC: A Practical Guide
by T. Hanai, Health Research Foundation, Kyoto, Japan

How to obtain future titles on publication

A standing order plan is available for this series. A standing order will bring delivery of each new
volume immediately upon publication. For further information, please write to:

The Royal Society of Chemistry, Turpin Distribution Services Ltd., Blackhorse Road, Letchworth,
Hertfordshire SG6 1HN, UK.
Telephone: Letchworth +44(0) 1462 672555: Fax: + (44)O 1462 480947
 CHROMATOGRAPHY
MONOGRAPHS

HPLC
A Practical Guide

T. Hanai
Health Research Foundation, Kyoto, Japan

RSC
ROYAL SOCIETY OF CHEMISTRY
ISBN 0-85404-515-5

A catalogue record for this book is available from the British Library

0The Royal Society of Chemistry 1999

All rights reserved.

Apart from any fair dealing f o r the purposes of research or private study, or criticism or
review as permitted under the terms of the UK Copyright, Designs and Patents Act, 1988,
this publication may not be reproduced, stored or transmitted, in any form or by any means,
without the prior permission in writing of The Royal Society of Chemistry, in the case of
reprographic reproduction only in accordance with the terms of the licences issued by the
Copyright Licensing Agency in the UK, or in accordance with the terms of the licences
issued by the appropriate Reproduction Rights Organization outside the UK. Enquiries
concerning reproduction outside the terms stated here should be sent to The Royal Society
of Chemistry at the address printed on this page.

Published by The Royal Society of Chemistry,

Thomas Graham House, Science Park, Milton Road
Cambridge CB4 OWF, UK

For further information see our web site at www.rsc.org

Typeset by Paston PrePress Ltd, Beccles, Suffolk

Printed by Bookcraft (Bath) Ltd
Preface
This book answers two basic questions regarding high-performance liquid
chromatography; it focuses on how compounds can be separated and on why
particular compounds are separated by liquid chromatographic methods. It
consists of six chapters: Basic Concepts of HPLC; Instrumentation: Prepara-
tion, Testing and Selectivity of Stationary Phase Materials; Selection of the
Eluent; Separation Based on an Improved Column Efficiency; Influence of
Physical Chemistry on Separations in Liquid Chromatography. The focus is on
the basic considerations in liquid chromatography rather than the applications
but there are also sections on trouble-shooting. The book is concerned mainly
with the selection of a packing material and the preparation of the eluent.
Theoretical optimization is demonstrated using the chromatography of simple
chemicals, to aid the understanding of liquid chromatography by graduate
students. The examples given go beyond what can be found in general textbooks
on liquid chromatography. The separations can be easily understood from the
differences in properties of familiar compounds known to undergraduate
students.
The chromatographic separations described are based on solubility using
molecular properties. The selection of chromatographic mode and column is
analysed using solubility parameters. The separation factor cc is described in
detail by physical and chemical parameters. Discussion of instrumentation
focuses on high-efficiency operation and is concerned with degassing, the time
constant, flow cell design, connectors and trouble-shooting. Packing materials
and their synthesis and surface modification, liquid chromatographic analysis
of surface activity and the evaluation of packed columns, including the
measurement of void volume, are also considered.
I thank Professor Roger Smith, of Loughborough University, who reviewed
the original manuscript and offered many helpful suggestions. I also deeply
appreciate the support of The Royal Society of Chemistry for this publication.

Toshihiko Hanai, Yokohama

V
Contents
Chapter 1 Basic Concepts of High-performance Liquid
Chromatography 1
1 Physical Parameters for High-speed Separations 1
2 Physical Considerations 4
3 Chemical Influences on the Separation Factor 5
4 Basic Considerations of Liquid Chromatography 6
5 Bibliography 9

Chapter 2 Instrumentation 11
1 Chromatographic Systems 11
2 Injectors 11
3 Eluent Delivery Systems 12
Composition Gradient Delivery Systems 13
Pressure Drop Changes due to Changes in the Eluent
Composition 13
Flow Rate Gradients 15
Step-wise Elution 15
Recycle Elution 15
Column Switching Separation 17
4 Degassing Methods 17
5 Column Temperature Control 18
6 Detectors 18
Spectroscopic Detectors 18
Electrochemical Detectors 22
Mass Spectrometer 22
Other Detectors 22
Measurement of Detector Sensitivity 23
7 Recording Systems 23
Selection of Detector Time Constant 23
8 Columns and Connectors 25
9 Flow Cell Volume and Connecting Tube Dimensions for
High Efficiency Operation 25
10 Other Components in a Liquid Chromatograph 26
11 Trou ble-shooting 26
vii
 ...
Vlll Contents
12 References 29
Reference Books 30

Chapter 3 Preparation, Testing, and Selectivity of

Stationary Phase Materials 31
1 Synthesis of Stationary Phase Materials 32
Silica Gel 32
Surface Modification of Silica Gel 35
Modification Using a Halogenated Silica Surface 35
Modification Using a Chlorosilane 35
Synthesis of Ion-exchanger Stationary Phases 36
Cation-exchanger 36
Anion-exchanger 36
Organic Stationary Phase Materials 37
2 Sieving of Stationary Phase Materials 37
3 Column Packing Methods 37
4 Column Evaluation 39
Column Efficiency and Asymmetry 39
Column Test for Reversed-phase Liquid Chromatography 40
Inertness Test for Basic Compounds 40
Inertness Test for Acidic Compounds 41
Inertness Test for Heavy Metals 41
Physicochemical Tests of the Stationary Phase 43
5 Measurement of Void Volume 43
Elution Volume of Proposed Void Volume Markers 44
6 Selectivity of Stationary Phase Materials 46
Silica Gel-based Stationary Phase Materials 46
Organic Polymer-based Stationary Phase Materials 49
Other Stationary Phase Materials 54
7 References 56

Chapter 4 Selection of the Eluent 57

1 Reversed-phase Liquid Chromatography 57
Selectivity of Organic Modifiers in the Eluent 57
Comparison of the Effect of Organic Modifiers on
Analytes with logP = 3 60
Chromatographic Behaviour of log P = 3
Compounds in Aqueous Tetrahydrofuran 60
Chromatographic Behaviour of log P = 3
Compounds in Aqueous Methanol 62
Chromatographic Behaviour of log P = 3
Compounds in Aqueous Acetone 62
Chromatographic Behaviour of log P = 3
Compounds in Aqueous Dimethylformamide
(DMF) 62
Contents ix
Comparison of Effects of Modifiers on a Mixture
of Analytes 62
2 Chromatography of Ionic Compounds 65
pH Effects 65
Ion-pair Separations 70
Theory of Ion-pair Liquid Chromatography 70
Normal Phase Ion-pair Partition Liquid Chromatography 71
Reversed-phase Ion-pair Liquid Chromatography 72
pH Effects 75
Effect of the Ionic Strength of the Buffer Solution 80
Selectivity of Organic Modifier and the Concentration
Effect 80
Column Temperature Effects 80
Detection 80
Trouble-shooting in Ion-pair Liquid Chromatography 81
3 Normal-phase Liquid Chromatography 81
Classification of Solvents 85
Preparation of the Eluent 89
4 Size-exclusion Liquid Chromatography (SEC) 90
Aqueous Phase Size-exclusion 91
Organic Phase Size-exclusion 92
5 References 93
Bibliography of additional references on theoretical
approaches in ion-pair liquid chromatography 95

Chapter 5 Separation Based on an Improved Column

Efficiency 96
1 Improving Separation by Changing the Selectivity 99
2 Improving Separation by Increasing the Column Efficiency 100
Reducing the Eddy Diffusion Term 102
Reducing the Effect of Longitudinal Diffusion 103
Reducing the Effect of Resistance to Mass Transport
in the Stationary and Mobile Phases 103
3 Bibliography 108

Chapter 6 Influence of Physical Chemistry on Separations

in Liquid Chromatography 109
1 Quantitative Structure-Retention Relationships in
Reversed-phase Liquid Chromatography 109
Prediction of Retention Times from log P in Reversed-phase
Liquid Chroma tograp hy 110
Prediction of Partition Coefficient (log P) 110
Correlation between log P and log k Values 110
Prediction of Retention Times from log P 111
x Contents
Prediction of Retention Time of Ionic Compounds from
log P and pK, 113
Calculation of Dissociation Constant pK, from
Hammet’s Equation 113
2 Van der Waals Volume as the Basic Property 115
Calculation of van der Waals Volume 115
Prediction of Retention Times Based on van der Waals
Volumes 115
3 Enthalpy as a Thermodynamic Property in
Retention Studies 128
Measurement of Enthalpy by Liquid Chromatography 129
Enthalpy Related to Retention 129
4 Conclusion 131
5 References 131

Subject Index 133

CHAPTER 1

Basic Concepts of High-

performance Liquid
Chromatography

The two basic questions in high-performance liquid chromatography focus on

(a) how particular compounds can be separated, and (b) why particular
compounds were separated by the liquid chromatographic method used. The
answers can be obtained by the consideration of some simple representative
chromatograms of the separation of well-known compounds. Such separations
can be easily understood according to common principles of physics and
chemistry .
A separation is described by the following equation, which indicates the
degree of resolution between two peaks in a chromatogram, R,. A complete
separation requires R, > 1.2 units.

Rs =&+
1 a-1

The resolution can be improved by increasing the column plate number, N , and/
or the separation factor, a (a = the ratio of the retention factors of the two
compounds). N is the physical parameter and a is the chemical parameter for the
separation. Higher N a n d a values give a better separation.
The physical and chemical aspects of liquid chromatography, in addition to
mechanical aspects, are briefly described in this chapter. Theoretical approaches
are explained in detail in later chapters. The effect of stationary phase materials
on the chemical selectivity is described in Chapter 3, and the influence of the
eluent components is covered in Chapter 4. The plate number theory is
discussed in Chapter 5. Quantitative optimization is explained in Chapter 6.

1 Physical Parameters for High-speed Separations

It was thought that high-speed separations would be achieved by the develop-
ment of a physically stable pumping system and highly sensitive detectors;

1
2 Chapter 1
however, the main contribution to high-speed separation is made by small-size
stationary phase materials. A shorter separation time with complete resolution
cannot be achieved simply by increasing the flow rate or by using a small
column. The theoretical plate number of a small column must be the same as
that of a larger column to obtain the same separation.
For example, the separation of a mixture of benzene, acetophenone, toluene,
and naphthalene has been completed within 5.5 min using a 15 cm long, 4.6 mm
i.d. column, packed with 10 pm porous octadecyl-bonded silica gel, whose
',
theoretical plate number was 38 000 m- as shown in Figure 1.1A. Increasing
the flow rate 4-fold reduced the separation time to 1.5 min, because this mixture
was well separated (Figure 1.lB). The same mixture was separated within
4.5 min using a 10 cm long, 4.6 mm i.d. column packed with 3 pm octadecyl-
bonded porous silica gel with a theoretical plate number of 117000 m- (Figure
1.1C). Doubling the flow rate resulted in completion of the separation within
2 min, as shown (Figure 1.1D).
Comparison of these four chromatograms suggests that a fast separation can
be performed either using a longer column with 10 pm stationary phase material
with a high flow rate of the eluent, to give high resolution, or by a smaller

A B C

1 4

2 4

1 1 4

0 5 0 2 0 4 0 2
min
Figure 1.1 Eflect ofparticle size andflow rate on high-speedseparations. A and B column:
10 pm Cia-bonded silica gel, 15 cm x 4.6 mm i.d.; C and D column: 3 pm
C18-bonded silica gel, 10 cm x 4.6 mm i.d.; eluent, 70% aqueous acetonitrile;
flow rate A and C , 1, B, 4, and D, 2 ml min- I ; pressure drop A, 1.2; B, 5.9; C,
13.4; D, 27 MPa; detection: UV 254 nm. Peaks: 1, benzene; 2, acetophenone;
3, toluene, 4; naphthalene.
Basic Concepts of High-performance Liquid Chromatography 3
column packed with 3 pm stationary phase material. However, a high flow rate
through the 3 pm stationary phase material is limited by a high column back
pressure. The separation could also be completed within 1.2 min on the short
column packed with 3 pm stationary phase material by using a stronger eluent,
as shown in Figure 1.2. Furthermore, the sensitivity was also improved by using
the smaller-size stationary phase material because the sample is less spread out
in the eluent and is more concentrated when it reaches the detector. The actual
peak height in Figure 1.1C is 1.6 times that in Figure 1.1A. A small column
packed with small particle-size stationary phase materials promises high
performance and a high-speed separation both in theory and in practice. The
following equation describes the relationship of the column length (L) to the
column efficiency: N = L/H. The high plate number N required for good
separation is proportional to the longer column length L and small H value.
The term H i s the height equivalent to a theoretical plate (HETP), which is the
length of column needed to generate one theoretical plate. A good column has a
high plate number for its length, and, thus, a good column has a low H value.
The value of H can also be described by the following equation (which is
described in detail in Chapter 5):

i -
4 I I 1

0 2 0 1 mln
Figure 1.2 Eflect on eluent component o f f o w rate. Column: 3 pm C18-bonded silica gel,
10 cm x 4.6 mm i.d.; eluent, 80% aqueous acetonitrile; f o w rate A, 1, B,
2 ml min-'; detection, UV 254 nm. Peaks: 1, benzene; 2, acetophenone; 3,
toluene; 4,naphthalene.
(Reproduced by permission from ref. 3)
4 Chapter I
This equation indicates that the particle size, dp,is the main contributor to the H
value. The smaller the particles, the higher the theoretical plate number. The
optimum condition is obtained by the relationship between the theoretical plate
height and the flow velocity.

2 Physical Considerations
High-speed separations can be achieved with a short column packed with 3 pm
stationary phase material, as shown in Figure 1.2. The sensitivity was also
improved by the use of smaller-size stationary phase materials, due to less
sample diffusion inside the column. The following conditions are required to
obtain such a separation.

a. small-diameter, spherical stationary phase materials that have high

physical strength;
b. a high pressure pump with controlled flow rate;
c. a system that limits sample diffusion, by considering the column design,
using small inner diameter connecting tubing, and a small volume detector
flow cell; and
d. a detector and recorder capable of a high-speed response.

The theoretical plate number N of peak B can be calculated from the

chromatogram given in Figure 1.3 by the following equation:

2
N = 16($)

where VR is the retention volume and w is the peak width at the base (measured
in volume units). However, the retention volume includes the hold-up volume
V , (also called dead volume). The hold-up volume is the sum of the void
volume of the column ( V o = Y A ) , the volume of the injector ( O X ) and the
volume of the detector and connecting tubing (XY)as shown in Figure 1.3. The
actual separation efficiency is defined as the effective theoretical plate number
N e ~which
, excludes the hold-up volume:

)
2
VR - VM
Neff = 16(

Commercial instruments have a reasonable balance between the recommended

column size and the volume of the column and connecting tubing (XY).
However, the theoretical plate number of a single column may give different
values on different instruments, and even on replacement of the components
and parts of a single instrument. Such discrepancies can be understood in terms
of differences in the mechanics of the instruments and the design of their parts.
Basic Concepts of High-performance Liquid Chromatography 5

Figure 1.3 A schematic chromatogram: VR, retention volume; V r , adjusted retention

volume; V, elution volume of peak; Vo, void volume; W ,peak width; V,, hold-
up volume; O X , volume of injector; X Y , volume of detector, including volume of
tubing .

The normally acceptable extra-column dead volume (0Y in Figure 1.3) before
there is a significant effect on the efficiency of a 15 cm long, 4.6 mm i.d. column
should be less than 100 pl. This volume has to be reduced to less than 30 pl for a
5 cm long, 4.6 mm i.d. column. Replacement of the connecting tubing with
shorter lengths of narrow-bore tubing and the selection of a smaller volume
detector flow cell are necessary when using a shorter or narrower column. These
changes together with a smaller column enable reductions in the volume of
eluent required and in the separation time. This approach is economical and
environmentally friendly. However, the reduced hold-up volume becomes
technically very critical in the handling of smaller columns. Details of the basic
mechanisms and the design of instrumentation are described in Chapter 2,
which also covers the similarities and differences of various instruments.

3 Chemical Influences on the Separation Factor

One column can be used for different types of liquid chromatography by
changing the eluent components. As an example, a column packed with
octadecyl-bonded silica gel has been used for size-exclusion liquid chromato-
graphy with tetrahydrofuran (THF), normal-phase liquid chromatography with
n-hexane, and reversed-phase liquid chromatography with aqueous acetonitrile.
Examples of the chromatograms are shown in Figure 1.4.
The elution volumes of polystyrene and benzene in the size-exclusion mode
were 0.98 and 1.78 ml, respectively (Figure 1.4A). This means that separations
by molecular size can be achieved between 0.98 and 1.78 ml in this system. In the
normal phase mode the elution volumes of octylbenzene and benzene were 1.98
and 2.08 ml, respectively, in n-hexane solution (Figure 1.4B). This type of
chromatography is called adsorption or non-aqueous reversed-phase liquid
chromatography. These are adsorption liquid chromatography and non-
aqueous reversed-phase liquid chromatography. The elution order of the
alkylbenzenes in the reversed-phase mode using acetonitrile was reversed
6 Chapter I

A B C
1
5
E 4 12

h1
C
0 1 3
co
(v
>
3

I I I I I I
0 5 0 5 0 5 min
Figure 1.4 D i e r e n t modes of chromatographs using the same column. Column, 5 pm
C18-bonded silica gel, 15 cm x 4.6 mm i.d. Eluent A, tetrahydrofuran; B, n-
hexane; C, acetonitrile; flow rate, 0.5 ml min-' at ambient; detection, UV
260 nm. Peak 1, benzene; 2, ethylbenzene; 3 , butylbenzene; 4, octylbenzene;
and 5 , polystyrene.

(Figure 1.4C). The elution volumes of benzene and butylbenzene were 2.01 and
2.52 ml, respectively. The elution volumes became larger with the addition of
water to the acetonitrile eluent. In each case the elution orders are based on the
solubility of the solutes (except in size-exclusion liquid chromatography).
When separation cannot be achieved by improving the theoretical plate
number of a column, it may be achieved by the selection of an appropriate
stationary phase material and/or eluent. The degree of separation, the separa-
tion factor a, is the difference in retention volumes of analytes. The separation
factor of two compounds is given by:

where VR2 and VRl are the retention volumes of peaks 2 and 1, respectively.
These retention volumes depend on the properties of the solutes (analytes),
stationary phase materials, and eluent components. A higher a value, i.e. an
increase in the difference in retention volumes, can be achieved by using a
different stationary phase material and/or eluent. Details of the selection of
stationary phase materials and eluent are described in Chapters 3 and 4.

4 Basic Considerations of Liquid Chromatography

Identifying the most suitable separation conditions is the main objective of
separation scientists. It is easier for skilled chromatographers, but this is a
complicated subject for beginners. One approach is to find a chromatogram that
exhibits the separation of a similar mixture. However, similar mixtures may
have been separated under very different conditions; either the separation
Basic Concepts of High-performance Liquid Chromatography 7
columns and/or the components of the eluent may have been different.
Furthermore, the elution order is sometimes reversed. When an appropriate
chromatogram is found in the literature, the conditions may need to be modified
to take into account the other compounds in the mixture, any necessity for
sample pre-treatment and the purpose of the separation. The chromatographer
should be familiar with the capabilities and requirements of the following
methods.

a. Pre-treatment of samples, stationary phase materials, and elution solvents;

b. Separations based on molecular size;
(i) aqueous size-exclusion liquid chromatography;
(ii) non-aqueous size-exclusion liquid chromatography;
c. Normal-phase liquid chromatography;
d. Reversed-phase liquid chromatography;
e. Ion-exchange liquid chromatography;
f. Ion-pair liquid chromatography;
g. Chiral separation and affinity liquid chromatography.

Choosing the sample pre-treatment method is difficult. The most important

consideration is the final condition of the target compounds. What kind of
solution is obtained? The type of solvent and the concentration of the target
compounds are very important in the selection of the separation conditions.
The pre-treatment of stationary phase materials is also important for silica gel
and ion-exchangers, even when a new column is being used. Pre-treatment of the
elution solvent is also important. High-performance liquid chromatography
grade solvents from different manufacturers contain different amounts of
impurities. The purity of the water is especially critical. The specified solvent
for a special preparation stage is often not compatible with the desired
chromatography. A large amount of such a solvent should first be injected,
followed by measurement of the background of the chromatogram.
How are analytes retained on, or in, a stationary phase? This depends on the
physicochemical interaction between the analytes and the stationary phase
material. When a strong solvent, in which the solute readily dissolves, is used
for elution the solute is eluted very quickly from the column. The forces holding
an analyte on the stationary phase are similar to those responsible for
dissolution in the solvent. Eight solubility properties are recognized: van der
Waals (London dispersion) forces, dipole-dipole, ion-dipole, Coulombic and
repulsion forces, charge-transfer complexation, and hydrogen-bonding and
coordination bonds. The molecular interactions that are probably involved in
retention in liquid chromatography can be explained by these interaction
properties and are summarized in Table 1.1. The retention of a particular
molecule is not due to only one property, but rather to a combination of several
properties. The probable interactions can be estimated from the chemical
structure of the analytes and stationary phase materials.
The separation conditions employed for size-exclusion liquid chromatogra-
phy are simple. A strong solvent for analytes and a suitable stationary phase
8 Chapter I
Table 1.1 CIassiJication in liquid chromatographic methods

Size-exclusion +oo
Reversed-phase 000+n n
Reversed-phaseion-pair +
0 0 0 + n
Ion-pair partition 000 o +
Ion-exchange ooooo a + 0
Normal-phase 000 + 0 n o
Charge-transfer 0 00 0 0 +
Salting-out 0 0000 0 +
Ligand-exchange 0 00 o +
Chelation 0 00 o o +
Affinity 0 0 0 0 0 0 0 0 0 0 +
Chiral separation 0 000000 0 +
+ The most important mechanism, 0 important property, A interaction depending on stationary
phase material.

material are necessary. If the impurities have high relative molecular masses
( M J , size-exclusion chromatography can be used effectively. Size-exclusion
liquid chromatography in conjunction with a recycling system can also separate
isomers; however, it is time-consuming and the columns are usually expensive.
If a mixture of molecules with a M , of less than 2000 has to be separated and a
recycling method seems to be insufficient for the separation, the following
chromatographic technique can be carried out. If the sample concentration is
large enough for chromatographic analysis, the eluted solution obtained by a
size-exclusion chromatographic pre-treatment can be directly injected onto a
liquid chromatograph using a syringe, after membrane filtration. If a good
combination of stationary phase material and solvent cannot be found, then
methods c-g in the above list are applied.
In reversed-phase liquid chromatography, increasing the molecular size
increases the hydrophobicity of solutes and results in a greater retention
volume. This indicates that the van der Waals volume is an important property
in optimization. Increasing the number of substituents with n-electrons and
hydrogen bonding increases the solubility in water, that is they increase the
polarity of the solutes. This indicates that dipole-dipole and hydrogen-bonding
interactions contribute to hydrophobicity. Therefore, these properties are
important in controlling the retention volume in reversed-phase liquid chroma-
tography. However, the n-electrons of stationary phase materials such as
Basic Concepts of High-performance Liquid Chromatography 9

polystyrene gel and the hydrogen-bonding of non-endcapped bonded silica gels

also contribute to the retention.
Many compounds can be analysed by methods c or d, and sometimes both
methods c and d are employed. For a preparative-scale separation, method c
(normal-phase chromatography) is suitable due to the easy removal of solvent.
Identifying the best separation conditions for ion-exchange liquid chromato-
graphy (method e) is a little more difficult. Therefore, if the compounds can be
separated using methods c or d, these methods should be used. Even sacchar-
ides, organic acids, and nucleic acids are often separated by methods c and d.
The separation speed in ion-exchange liquid chromatography is also slower
than that in normal and reversed-phase liquid chromatography.
Affinity liquid chromatography and chiral separations (enantiomer separa-
tions) require similar analyte properties. The solutes may have interactions
through hydrogen-bonding, ligand formation, or Coulombic forces with the
surface of stationary phase materials or the sites of additives; however, the
selectivity is controlled by the steric effects of the structures of the analyte
molecules and the recognition molecules (chiral selectors).
The physical and chemical properties of stationary phase materials are
described in Chapter 3 (including methods for their synthesis) to clarify the
differences in similar stationary phase materials supplied from different manu-
facturers. A detailed selection guide to solvents is given in Chapter 4. The
unlimited selection of eluent components and their concentrations is a powerful
force in developing separations in liquid chromatography. Although this area
seems rather complicated, it is easy to understand the selection of a suitable
eluent when you first identify the molecular properties of the analytes and
solvents.

5 Bibliography
J.C. Giddings, ‘Dynamics of Chromatography 1. Principles and Theory’, Marcel
Dekker, New York, 1965.
M.T. Gilbert, ‘High-performance Liquid Chromatography’, Wright, Bristol, 1987.
S. Hara, S. Mori, and T. Hanai, ‘Chromatography. The Separation System’, Maruzen,
Tokyo, 1981.
H. Hatano and T. Hanai, ‘New Experimental High Performance Liquid Chromatogra-
phy’, Kagaku Dojin, Kyoto, 1988.
E. Heftmann, ‘Chromatography’, 5th edn, Journal of Chromatography Library, Vol.
5 1A, Part A ‘Fundamentals and Techniques’, Elsevier, Amsterdam, 1992.
A.M. Krstulovic and P.R. Brown, ‘Reversed-Phase High-Performance Liquid Chroma-
tography’, Wiley, New York, 1982.
S. Lindsay and D. Kealey, ‘High Performance Liquid Chromatography’, Wiley,
Chichester, 1987.
V.R. Meyer, ‘Practical High-Performance Liquid Chromatography’, Wiley, Chichester,
1994.
P.A. Sewell, B. Clarke, and D. Kealey, ‘Chromatographic Separations’, Wiley, Chiche-
ster, 1987.
R.M. Smith, ‘Retention and Selectivity in Liquid Chromatography’, Journal of Chro-
matography Library, Elsevier, Amsterdam, 1995, Vol. 57.
10 Chapter I
L.R. Snyder and J.J. Kirkland, ‘Introduction to Modern Liquid Chromatography’,
Wiley, New York, 1979.
L.R. Snyder, J.J. Kirkland, and J.L. Glajch, ‘Practical HPLC Method Development’,
2nd edn, Wiley, New York, 1997.
R.W. Yost, L.S. Ettre, and R.D. Conlon, ‘Practical Liquid Chromatography, An
Introduction’, Perkin-Elmer, Norwalk, 1980.
CHAPTER 2

Instrumentation

The basic concepts of the instrumentation for liquid chromatography are

described here, with the mechanism of their operation and their influence on
the separation of analytes.

1 Chromatographic Systems
The components of all chromatographic systems are basically the same;
however, the specifications and sizes differ between systems. A schematic
diagram of a full-scale computerized system is shown in Figure 2.1. This type
of system was originally proposed at the start of studies to design instrumenta-
tion for liquid chromatography, and the fully automated system is now
commercially available. It is powerful, providing it is trouble-free. Its operation
seems simple to the user, but trouble-shooting for this system is complicated.
The basic concept of each of the components of the system is explained in the
following sections.

2 Injectors
The maximum injection volume depends on the volume of the sample loop in
the injection valve. The reproducibility of manual injection depends on the skill
of the operator. The use of a small sample loop and an overflow injection of the
sample solution so that the loop is fully flushed with sample are basic
requirements for quantitative analysis. The highest injection reproducibility
can be obtained by an auto-injector with a fixed sample loop. The smallest
reasonable injection volume is 1 pl. A nl-scale injection valve can be con-
structed; however, the memory effect at the surface of contact parts affects
quantitative analysis compared with the use of a pl-scale injection valve. For a
semi-micro system, a low hold-up volume injection valve is desired. The
minimum injection volume is 80 nl. For a preparative-scale injection, the
sample loop can be easily replaced with a larger-volume loop, such as a 200 pl,
instead of the standard 20 pl loop.

11
12 Chapter 2

.............................................

T2

r I

Figure 2.1 Automated chromatograph. Components: solid line, liquid f l o w line; dotted
line, communication line; S1, S2, and S3, solvent reservoirs; DG, degassing
system; V1, V2, and V3, valves; Pu, pump: Pre, pressure sensor: Inj, injector;
C , column; F,fraction collector; RC, reaction bath; RR, reaction reagent;
D, detector; W, waste; T1 and T2, oven; COM, computer; Pro, solvent
programmer; Pri, printer; INT, integrator or recorder.

3 Eluent Delivery Systems

The pump is the heart of HPLC and must satisfy a number of requirements. It
must deliver a constant and pulse-free flow (stability: less than f0.5%, or
usually less than & 1 pl min- I ) which must be reproducible under high pressures
up to 50 MPa. The internal volume of the pump must be very small to enable
rapid eluent changes and enable a precise gradient to be formed. The material of
the flow lines must be compatible with all types of solvent, including organic
solvents and strongly acidic and basic solutions. Four types of eluent delivery
systems are popular. Widely used are the syringe pump and reciprocating piston
pumps. A syringe pump is used for micro-scale liquid chromatography and
Instrumentation 13
capillary liquid/supercritical fluid chromatography. The most widely used is the
double-piston reciprocating pump. This pump is suitable for a wide range of
flow rates, from semi-micro liquid chromatography (LC) to preparative-scale
LC under high pressure. The reciprocating piston pump design has been further
modified in the masterlslave reciprocating pump; one piston works to pump the
eluent and the second piston works as the flow controller. The degree of
pulsation of the flow depends on the mechanical and electrical capability of the
manufacturer. The latest models provide improved precision of the flow rate,
and can handle semi-micro-scale liquid chromatography. A precise flow rate
( f 1 pl min- ') improves the reproducibility of gradient elution on standard size
columns (6 mm > column i.d. > 3 mm). With the alternative diaphragm pump,
the maximum pressure is less than that of the reciprocating pump, but the
diaphragm resists a wider variety of solvents. A pneumatic amplifier pump is
suitable for high flow rate operations such as preparative-scale liquid chromato-
graphy, but may not deliver a constant flow rate as it operates at constant
pressure. This type of pump is also used for column packing.
Several methods are applied to reduce the separation time. The best way is the
selection of a suitable column and an eluent using isocratic elution. However,
much skill and knowledge are required to make such a system. A flow rate
gradient, step-wise elution, or eluent composition gradients are commonly
applied to reduce separation times.

Composition Gradient Delivery Systems

High and low pressure gradient systems are available, which mix two eluents to
give a defined increase in eluent strength through a separation run, effectively
reducing the separation time. A low pressure gradient system requires the pre-
mixing of two or more solvents before pumping. This system is economical, but
the internal system volume is higher than that of the high pressure gradient
system. The precision of the high pressure gradient system, in which two pumps
deliver the eluent components at changing rates to provide a total constant flow
with a changing composition, is better but the cost is higher. The change and the
reduction in elution time from an isocratic to a gradient elution are shown in
Figures 2.2 and 2.3. The maintenance of a gradient elution system is more
complicated than one using isocratic elution, but gradient elution is necessary
for the separation of complex mixtures, such as amino acids, nucleic acids, and
many biological samples.

Pressure Drop Changes due to Changes in the Eluent Composition

During a gradient elution the operational back-pressure changes due to changes
in the viscosity as the eluent mixture changes' (Figure 2.4). When methanol and
water are used as the components, the back pressure drop can become
particularly high and sometimes stops the pump because it exceeds its
maximum pressure limit. A high back pressure must be prevented, particularly
when a longer column packed with small particle size is being used.
14 Chapter 2
1

I I I I
0 10 20 30
min
Figure 2.2 Separation of aromatic compounds using isocratic elution. Conditions: column,
5 pm C18-bonded silica gel, 15 cm x 4.6 mm i.d.; eluent, 0.001 M phosphoric
acid in 55% aqueous acetonitrile; flow rate, 1 ml min-'; temperature,
ambient; detection, UV 254 nm. Peaks: 1, phenol; 2, 4-methylphenol; 3, 2,4-
dimethylphenol; 4, 2,3,5-trimethylphenol; 5 , benzene; 6, toluene; 7 , ethylben-
zene; 8, propylbenzene; and 9, butylbenzene.

1
95 %

!
A 55 %
2

3
75 %

4 5 6 7 8 9

>
3

-.JL
I
0
d I
10
I

min

Figure 2.3 Gradient elution of aromatic compounds. Chromatographic conditions as

Figure 2.2 except f o r acetonitrile concentration.
 n
Instrumentation 15

lm5

v
- Methanohater

1.o

0.5 Acetonitrilehater

-0
0 50 % 100
concentration of water in
organic modifier
Figure 2.4 Relationship of viscosity to water and organic modifier mixture ratio

Flow Rate Gradients

Increasing the flow rate also reduces the separation time. However, this method
is limited by the physical strength of the instrument and by the loss of detection
sensitivity. The chromatograms of aromatic compounds separated isocratically
and with a flow rate gradient are shown in Figures 2.2 and 2.5, respectively.
Doubling the flow rate roughly halved the separation time; however, the peak
areas became smaller. Flow programming elution is suitable for the separation
of a mixture of homologous compounds.

Step-wise Elution
If the separation by isocratic elution is perfect, the eluent can be switched during
the run to a stronger one to speed up late components (Figure 2.6). The
peak height of longer-retained compounds increases and the retention time
is shorter compared with the chromatogram of isocratic elution (Figure 2.2).
This method is not suitable for quick repeat injections, due to the requirement
for a significant column conditioning time after returning to the original
conditions.

Recycle Elution
The recycle elution method can be applied to mixtures of very similar
compounds that cannot be fully separated by a single pass through the
column. This method makes more effective use of a column. The effluent from
the column is repeatedly re-passed through the same column. The number of
cycles multiplies the total theoretical plate number of the column if the system is
16 Chapter 2
1

2 mL

I
0 10 20
min

Figure 2.5 Flow rate programming elution of aromatic compounds. Experimental con-
ditions as Figure 2.2 except the f l o w rate. Flow rate increased f r o m 1 to
2 ml min-'.

55r
1
75 %

I I
0 10 min

Figure 2.6 Step-wise elution of aromatic compounds. Experimental conditions as in

Figure 2.2 except the acetonitrile concentration was changed from 55 to 75%
in water.

carefully assembled. This method is usually applied to larger columns and

preparative-scale separations. Figure 2.7 shows the recycle chromatogram of
the purification of a mixture of anthracycline antibiotics, in which a fraction (or
heart-cut) of each separation is recycled through the column, becoming
successively more resolved on each cycle.
Instrumentation 17

* the most active fraction

1st cycle 2nd cycle 3rd cycle 10 cycle

Figure 2.7 Purlfication of a mixture of anthracycline antibiotics using heart cut recycle
liquid chromatography to yield the most active fraction. Conditions: column,
J A IGEL-3 10; eluent , chloroform-methanol-25 % N H 4 0H (200 :5 :I ) ;Jlo w
rate, 4 ml min-'; detection: refractivity index.
(Reproduced by permission from Japan Analytical Industry data)

Column Switching Separation

The selectivity of different stationary phase materials can be applied using
columns in sequence to provide high-speed isocratic separations instead of
gradient elution. An example for amino acids analysis is shown later in Figure
4.15, where the same eluent was used for all of the separations and the fraction
containing the sample components of interest was switched from one column to
another.

4 Degassing Methods
Degassing of the eluent is important for trouble-free operation and highly
sensitive detectioq2 otherwise the eluent may become supersaturated with
air that is released as bubbles in the pump check valves or the detector flow
18 Chapter 2
cell. Several methods have been proposed, and two on-line systems have
been commercialized. One method is bubbling helium gas into the eluent.
Another is an on-line vacuum. Both mechanisms are simple, but the latter
system is popular due to its compact size, low hold-up volume, and because
it does not require expensive helium gas. Even if the eluent is a premixed
mixture of water and organic solvent, pre-degassing is still important.
The eluent container should be degassed using a water pump vacuum.
If the operation is performed in an ultrasonic bath, degassing is completed
within 1 min. Longer degassing can cause loss of the more volatile eluent
components.

5 Column Temperature Control

If good reproducibility of the retention times is required, the column tem-
perature should be kept constant. A water bath or an air oven are common
systems, and thermo-tape is also effective if the column is made of steel.
When temperature control is extremely important, pre-heating of the
eluent is necessary. If the whole apparatus is placed in a temperature-
controlled box, the temperature is constant, but such a system is not usually
required.

6 Detectors
One of the first steps in the development of modern liquid chromatography
was the automation brought about by adding an on-line detector. Sub-
sequently, the development of small particle size high-performance stationary
phase materials provided more efficient separation systems, which then became
known as high-pressure or high-performance liquid chromatography (HPLC).
A variety of analytical instruments have been employed as detectors.
Originally, standard instruments were modified to monitor effluents from
a high-performance column. The highest possible sensitivity and selectivity
of detectors are required in biomedical research, and led to the introduction
of dedicated instruments with high qualitative performance. The
basic mechanism is the same as that of the original instruments, but tech-
nical improvements have been made in the mechanics and electronics.
The specifications of modern detectors for HPLC are summarized in Table
2.1.

Spectroscopic Detectors
Energy from a light beam is absorbed by molecules with a chromophore. An
absorption spectrophotometer uses this mechanism, and the energy loss
depends on the concentration and molecular absorption constant of analyte
molecules and the wavelength of the light. The most popular detector, the
Instrumen tat ion 19

Table 2.1 Specijicity of detectors

Detector Sensitivity Cell volume Remarks
Pl

Ultraviolet 1 ng ml-' 4.5-20 Common detector; selection of

(visible) 1 10-~ wavelength improves the accuracy of
spectroscopy AUFS quantitative analysis, wavelength
scanning or photodiode array permits
qualitative analysis. W-lamp or complex
lamp is suitable for visible region
detection
Fluorescence 10 pg ml-' 7-60 Sensitive for fluorescent compounds
spectroscopy 1 x 10-4FS
Refractive Index 10 ng ml-' 5-14 All purpose; temperature control for
1.6 lo-' high sensitive detection, usually used for
RIU preparative LC or size exclusion LC,
eluent must be isocratic
Polarimeter Specific for optical isomers
Infrared Suitable for carbonyl compounds; NaC1,
spectroscopy CaF2 and sapphire cells are available
Conductivity 10 ng ml- ' 1.5 Detection limit 0.001 pmho is equivalent
to 0.01 ppm NaCl, suitable for ion
detection
Polarography 1 ng ml-I, 10 Suitable for ions, fatty acids, saccharides
2 x 10-:pA
Coulometry 50 pg s-
5 x 10-8A
Amperometry 50 ng ml-', Suitable for catecholamines, phenols
10 pA
Pulsed 50 ng ml-', Suitable for saccharides
amperometry 10 pA
Flame ionization I ngml-', All purpose; suitable for saccharides and
10 pA non-volatile compounds
Electron capture 100 pg ml-' Suitable for halogenated, phosphorus
and oxygenated compounds
Mass 10 pg ml-' All purpose; qualitative and quantitative
spectrometer detector
Inductive 20 ng mi-' Element selective
coupled plasma
Atomic 20 pg ml-' Organometallic compounds
absorption
Reaction heat ng s - ' 9 All purpose
Radioactivity I4c:2 n ~ i , 50-700 Metabolite analysis
3H :5nCi
Refs: MS: Anal. Chem., 1981,53, 1603; ICP: J . Chromutogr. Sci., 1985, 23, 4;AAS, Analyst, 1981,
106, 921.
20 Chapter 2
ultraviolet absorption detector, uses a deuterium lamp as the light source. A
visible region absorption detector requires a tungsten lamp. A combined lamp is
useful for monitoring both in the ultraviolet and visible regions. Pre- or post-
column derivatization of sample molecules to introduce a strong chromophore
can be applied for highly sensitive and selective detection. Pre-column deriva-
tization is also used to aid selective extraction from the effluent for collection
purposes.
Fluorescence detection is usually more sensitive than absorption detection,
but the number of naturally fluorescent compounds is limited. Pre- or post-
column derivatization can also be applied for this type of detection. Chemi-
luminescence detection is the most sensitive method for some fluorescent
compounds.
A light beam is refracted to different extents by different compounds. This
mechanism is used for refractive index detection. This detector is not sensitive
and the selectivity differences are negligible for homologous compounds, but
any solvent with a different refractive index to the analyte can be used as the
eluent. This detector is mainly applied to size-exclusion and preparative-scale
liquid chromatography.
Infrared spectroscopy is often used for qualitative analysis, and its powerful
selectivity means that it can be used as a detector. However, the absorption of
the eluent molecules, particularly in reversed-phase separations, often interferes
with the detection of analytes. The infrared absorption detector therefore
requires mechanical assistance to eliminate the solvent or needs powerful
computer assistance to eliminate the background signal.
Atomic absorption and inductively coupled plasma spectrometers are metal-
selective spectrometers used for organic metal analysis. The connection of these
spectroscopic instruments to a liquid chromatograph is relatively simple.
Chromatograms of alkylmercury3 and aminoplatinum analytes4 are shown in
Figures 2.8 and 2.9, respectively.
A polarimeter can be used for the identification of optically active com-
p o u n d ~as
, ~shown in Figure 2.10.

I I I
0 10 min 20
Figure 2.8 Detection of alkylmercury compounds using flameless atomic absorption
detector. Conditions: column, Corasil I , 50 cm x 2.1 m m i.d.; eluent, n-
hexane; flow rate, 0.5 ml min-'; detection, flameless atomic absorption
spectrometer. Peaks: 1, benzene; 2, ethylmercury chloride; and 3, methylmer-
cury chloride.
Instrumentation 21

cis l

0 10 min
Figure 2.9 Inductively coupled plasma detection of diaminodichloroplatinum (11) . Con-
ditions: column, Shodex O H , 25 cm x 4.1 mm i.d.; eluent, 0.01 M phosphoric
acid; flow rate, 1 ml min-'; detector, emission at platinum line 265.9 nm.
Samples: cis and trans diaminodichloroplatinums. The second peaks are
considered to be their oligomers.

I I
0 15 min

(-1

Figure 2.10 Chromatograms of ( f) -trans-stilbene oxide. Conditions: column, Chiralpak

UP( +), 25 cm x 4.6 mm i.d.; eluent, methanol; column temperature,
ambient;Jow rate, 1 ml min-'; detector A, polarimeter; B, refractive index.
22 Chapter 2
Electrochemical Detectors
Electrochemical detectors, which are based on the electrochemical oxidation or
reduction of the analyte, can be applied to the analysis of selected compounds
such as phenols. It is physically simple, but is very sensitive for catecholamines.
However, the adsorption of reacted molecules on the surface of the electrodes
can reduce the conductivity. To overcome this problem a pulsed voltage is
applied, which cleans the electrode surface between measurements. This pulsed
amperometric detection is also sensitive for carbohydrates.
Conductivity detection, which involves measuring changes in the conduc-
tivity of an aqueous solution between two electrodes, is employed in ion
chromatography for the detection of ionized analytes.

Mass Spectrometer
The mass spectrometer is a very sensitive and selective instrument. However, the
introduction of the eluent into the vacuum chamber and the resulting significant
pressure drop reduces the sensitivity. The gas exhaust power of a normal
vacuum pump is some 10 ml min-' so high capacity or turbo vacuum pumps
are usually needed. The gas-phase volume corresponding to 1 ml of liquid is
176 ml for n-hexane, 384 ml for ethanol, 429 ml for acetonitrile, 554 ml for
methanol, and 1245 ml for water under standard conditions (O'C, 1 atmos-
phere). The elimination of the mobile phase solvent is therefore important,
otherwise the expanding eluent will destroy the vacuum in the detector. Several
methods to accomplish this have been developed. The commercialized inter-
faces are thermo-spray, moving-belt, electrospray ionization, ion-spray, and
atmospheric pressure ionization. The influence of the eluent is very complex,
and the modification of eluent components and the selection of an interface are
therefore important. Micro-liquid chromatography is suitable for this detector,
due to its very small flow rate (usually only 10 pl min-I).

Other Detectors
Flame ionization and electron capture detectors, which are used in gas
chromatography, have been modified for liquid chromatography but have not
been widely used. It is again necessary to eliminate the eluent to improve the
sensitivity and to broaden their application. Moving-wire and moving-belt
flame ionization detectors are useful for the detection of organic compounds,
but a difference of greater than 80°C between the boiling point of the analyte
and the eluent is required to prevent loss of the analyte. The sensitivity of the
electron capture detector is 0.12 ng ml- for aldrin; however, only volatile
organic solvents can be used as the eluent components. A radioisotope detector
has been used for radiolabelled metabolite analysis. Electron spin resonance
(ESR) has been used for the analysis of radicals. Nuclear magnetic resonance
spectroscopy (NMR) with a flow cell has also been used for qualitative analysis.
Instrumentation 23
The development of high-power NMR (800 MHz) spectrometers permits the
simple operation of LC-NMR for metabolite analysis.

Measurement of Detector Sensitivity

For ultraviolet and visible spectroscopic detectors, a standard solution of a
compound whose molar absorption constant is known must be prepared, and
placed in the flow cell. The absorbance obtained is then compared with the value
measured by a standard spectrophotometer.
For a fluorescence detector, quinine sulfate is used as the standard com-
pound. The flow cell is filled with a standard solution and the fluorescence
intensity is measured. The value is compared with that measured by a
fluorescence spectrophotometer. This standard solution is also used for fixing
the wavelength and position of the flow cell. The Raman spectrum of water can
also be used for this purpose.
For refractive index detectors, 13.33 g of sucrose is dissolved in 100 ml of
pure water and then diluted 200 times. The refractive index of the final solution
is 1 x RI, which is used to calibrate the instrument.

7 Recording Systems
The visualization of the detector signal helps to clarify the separation condi-
tions. Pen recorders were originally used, but the integrator is now popular due
to the automatic reporting of both the retention time and peak area or height.
The effective use of an integrator makes quantitative analysis easier. Computer-
based integrators are powerful for the storage and further arrangement of data,
and can also be used for column evaluation and as a system controller.

Selection of Detector Time Constant

The selection of the time constant of the detector and recording system is very
important in high-speed separations. If the peak height monitored at a high flow
rate is lower than that measured at the standard flow rate, what is the cause?
Sometimes the peak height is lower than that expected theoretically. Carrying
out a separation at a flow rate of 1 ml min-' and a recorder time constant of 1 s
gave chromatogram A in Figure 2.1 1 , and at 4 ml min-' chromatogram B was
obtained. The peak heights in chromatogram B are ca. 60% of those of A. The
peak heights in chromatogram C, using a time constant of 2 s and flow rate of
4 ml min-', are ca. 50% of those of A. When the time constant was then
reduced to 0.1 s, chromatogram D was obtained with peak heights ca. 80% of
those of A. This experiment demonstrates the importance of time constant
selection. The importance can be further observed by examining the peak height
changes of the first peak in Figure 2.1 1. These chromatograms were monitored
on a chart recorder. When the experiments were monitored by an integrator, the
area of peak 1 in Figure 2.1 1 was constant in all four runs, but the peak heights
24 Chapter 2

B C D
4

1
3 4

1 4

1 1 1 1

5 0 2' I 2 0 2 min

Figure 2.11 Eflect of time constant of detector. Conditions: column: 10 pm C18 silica gel,
15 cm x 4.6 mm i.d.; eluent, 70% aqueous acetonitrile; flow rate for A,
1 ml min- ',for B-D, 4 ml min-I; time constant A and B, 1.O s, C , 2.0 s, D,
0.1 s. Peaks: 1, benzene; 2, acetophenone; 3, toluene; and 4,naphthalene.

!
0
0

1,
.-2
+I, 0 1.0
time constant
sec 2.0

Figure 2.12 Ratio of peak area and height related to time constant of detector. Experi-
mental conditions are the same as those in Figure 2.11. Peak height and area
ratios are calculated from the data of peak nos. 1 and 4 as a % of the
maximum value. Lines: A, peak height ratio of chromatograms measured at
'
I .O ml min- pow rate; B, peak height ratio of chromatograms measured at
4.0 ml min-' flow rate; C, peak area ratio of chromatograms measured
at 1.0 ml min-' flow rate; and D, peak area ratio of chromatograms
measured at 4.0 ml min-'JEow rate.
Instrumentation 25
were affected by the time constant. The peak area ratio between peaks 1 and 4
was almost constant from 1 to 4 ml min-' (Figure 2.12, lines C and D), while
the time constant was varied from 0.1 to 2 s. A rapid time constant is required
when the flow rate exceeds 4 ml min-'. The peak height ratio of peaks 1 and 4
was affected by the time constant (Figure 2.12, lines A and B); therefore, the
selection of the time constant for quantitative analysis is important. Further-
more, the optimum peak recognition parameters of an integrator should be
determined for fast separations. When the time constant of the detector and the
peak recognition parameter are too small, the noise level becomes high and
quantitative analysis becomes difficult. Suitable conditions must be selected in
each case for quantitative analysis.

8 Columns and Connectors

The best material for column tubing is glass, but careful handling is required
due to its physical vulnerability. The most commonly used column material is
stainless steel tubing, and the inner surface is polished like a mirror. These
columns are physically stable and are compatible with a variety of compounds
and eluent components. Some biological materials interact with metals; there-
fore, organic polymer and glass-lined stainless steel tubes are preferable for
biological samples, and also for eluents with high ionic strengths which can
attack the steel. The design of the inlet and outlet connections of columns
depends on the physical strength of the materials. However, different connec-
tors and fittings from different systems should not be mixed. Different
connectors sometimes damage the efficiency of a whole system because they
can create significant extra column dead volumes. The best approach is to use
one connection system, even with instruments from different manufacturers.
Replacement of the connection system is now simple, and does not damage
individual components. Finger-tight polymer-type (PEEK) connectors are
desirable if the instrument is operated at usual pressures (up to 20 MPa).
Column systems are basically classified into two groups: conventional and
cartridge columns. A conventional column usually demonstrates better column
efficiency than does a cartridge column due to a higher column packing
pressure, but its usefulness depends on the quality control system of the
manufacturer and the skill of their technicians. A cartridge column is econom-
ical, but its theoretical performance and lifetime can be less than a conventional
column.

9 Flow Cell Volume and Connecting Tube Dimensions

for High Efficiency Operation
If the correctly sized flow cell and connecting tubing are not used, the high
efficiency of a column or high theoretical plate number columns cannot be
effectively used. The detector cell volume contributes hold-up volume. The
larger is the cell volume, the greater the peak broadening. The cell volume
26 Chapter 2
should be less than 10% of the typical volume required to elute a peak. The peak
volume is related to the retention factor k, and the longer the retention time, the
larger is the peak volume. The cell volume is thus particularly critical for
compounds with short retention times. A well-designed small volume flow cell is
also required for smaller diameter or short columns. The relationship between
the maximum desirable cell volume and column i.d. of a 10 cm long column is
given by:

Cell volume (pl) = 4.2 x 10-4x2 - 3.6 x 10-6x + 8.4 x

for a compound with k = 1, where x is the column i.d.
The connecting tubing should be as short and narrow as possible. The volume
of a 20 cm x 0.5 mm i.d. tube is 39.3 pl. That of a 20 cm x 0.25 mm i.d. tube is
9.8 pl, and that of a 20 cm x 0.125 mm i.d. tube is 2.5 pl. Some detectors are
equipped with a heat exchanger that consists of a metal block containing a
capillary tube. The volume of this tube also affects the theoretical plate number.
If highly sensitive operation is not required, the heat exchanger can be removed
or bypassed.

10 Other Components in a Liquid Chromatograph

A fraction collector and a post-column derivatization system were included
(Figure 2.1) for a comprehensive and multi-purpose instrument. However, the
fraction collector is needed only when collecting components from the effluent,
and is generally not included in an analytical system. The post-column
derivatization system is connected only when required for the selective and
sensitive detection of specially targeted compounds. Usually, most compounds
are directly detected by an on-line spectroscopic or other detector.
A computer (Figure 2.1) can handle the complicated operations of controlling
the auto-injector (injecting different volumes, adding an internal standard
compound and selecting sample vials), the pumps (controlling the flow rate,
selecting the different gradient methods, and changing the eluent composition),
the detectors (selecting wavelength, sensitivity, time constant, and the optimiza-
tion), the column temperature, the recording system, and other parts of the
system. Additional software can improve the quality of reports, including the
modification of chromatograms and the results of quantitative analysis.
Furthermore, combination with a computational chemical analysis system can
enable the production of quantitative structure-retention relationships.

11 Trouble-shooting
General problems and the appropriate maintenance procedures are summarized
in Table 2.2. The details of individual instruments are well described in the
manufacturers' manuals. The common problems are due to poor maintenance
of the instruments and poor understanding of the specificity of stationary phase
Instrumentation 27

Table 2.2 Trouble shooting and maintenance

Trouble: Diagnosis Maintenance

1. No response after switching on

A. Fuse Replace fuse
B. Electric cable Check electric cable
2. No eluent flow from pump
A. No movement of pump
a. Nocurrent See 1A and 1B
b. Lack of grease oil Oil
c. Safety pressure regulator on Fix or replace pressure gauge
d. Gears jammed Change flow rate; if this does not work,
then call the manufacturer
e. Recrystallized salt Wash with water; call manufacturer
B. Pump piston moves
a. Leak in line Fix or replace appropriate part
b. Air bubble in pump Degas the eluent, disconnect the pump
and remove air by using a high flow rate
c. Dirty check valve Disconnect pump, wash with a high flow
rate or remove check valve and wash in an
ultrasonic bath with a strong solvent;
finally, overhaul the check valve
d. Aged piston seal Fit new seal. Call manufacturer if using a
syringe or diaphragm type pump
C. Back pressure too high
a. Plug in column, precolumn, Clean inlet side of column or replace filter
guard column, or on-line filter
due to dust from injector rotor
seal or sample
b. Expansion of stationary phase Wash column using intermediate solvent
at low flow rate; change eluent; change
chromatography system
c. Solvent filter blocked Replace with a suitable one
3. Eluent flow present
A. Low flow rate
a. 2Ba-d
B. No response from detector
a. No power to detector lA, 1B
b. Aged lamp or photo cell Replace lamp or photo cell
c. Others Call manufacturer
C. No recorder movement
a. No current lA, 1B
b. Broken wire Replace
c. Malfunction of resistance wire Clean; if this fails, go to f.
d. Selection of polarity
e. Flow cells unbalanced Clean cell or replace solvent
f. Others Use another recorder, and call
manufacturer
(continued)
28 Chapter 2
Table 2.2 Continued
Trouble: Diagnosis Maintenance

4. Flow present and recorder responding

A. Irregular noise on baseline
a. Air bubble in flow cell Degas eluent; remove air bubble from flow
cell
b. Small bubble or immiscible Wash flow cell with ethanol, T H F or 6 M
solvent in flow cell HN03
c. Dirty flow cell window Apply b above; clean flow cell
d. Insufficient light energy Replace lamp; if using an unsuitable
eluent component of absorbance at
detection wavelength, change eluent
e. Malfunction of ground/earth Fix or change earth connection
B. Drifting baseline on output
a. Unstable current or voltage Use stabilizer on power line
b. Dust on light line Clean
c. Leak at flow cell Fix it, or replace seal
d. Using refractive index detector Flow rate too high; drafts; change prism
p.osition
e. Step-wise movement Clean recorder slide wire; slightly increase
gain of recorder
f. Vibration of pen recorder Decrease recorder gain a little
C. Baseline drift without gradient elution
a. Dirty flow cell, expanding air Wash flow cell
bubble
b. Using refractive index detector Control detector temperature
C. Aged reference battery of Replace battery in recorder
recorder
d. Soluble stationary phase Change chromatographic system
e. Unstable current or voltage Ba
f. Expanding air bubble inside flow Remove air from line, especially from
cell pump head; degas eluent
D. Signal out of range of baseline
a. 3Bc 3B, C
b. 4A b,c 4A b,c
c. Malfunction of light beam in Adjust flow cell position; remove flow cell
detector and fix it
d. Solvent Use HPLC grade solvent; use freshly
prepared eluent
e. Reference flow cell Fill with the same eluent
f. Fresnel type refractive index Adjust zero balance
detector
5. Bad peak reproducibility even when conditions 1-4 are satisfied
A. Increased peak retention times
a. 3A 3A
b. Column temperature reduced Use fixed oven temperature
c. Unstable column If due to previous solvent, wash column
sufficiently, if due to stationary phase,
wash column sufficiently; if stationary
phase is soluble in the eluent, change
chromatographic system
d. Chart speed Fix recorder
~~

(continued)
Instrumentation 29

Table 2.2 Continued

~~~~

Trouble: Diagnosis Maintenance

B. Decreased peak retention times

a. Aged stationary phase Replace stationary phase or column; if
stationary phase soluble, change
chromatographic system
b. Remains of previous eluent Wash column sufficiently
(especially gradient elution)
c. Increasing column temperature Fix oven temperature
C. Resolution becomes poor
a. 5Ba-c 5B a-c
b. Unbalanced density of column Repack column
stationary phase
c. Overloading of sample Resolution is poor even if repeated under
initial conditions, see 5C, a,b
D. Poor peak area reproducibility
a. Poor injection Wash micro injection syringe; fix injector
leakage
b. Trouble with detector 3B, b; 4A, c; 4C, b; 4D, f; if non-linear
range of detector used, change
concentration
c. Trouble with recorder 3C, b,c; 4C, c; fix the gain of the recorder
6. Tailing peak
A. Polluted column 5A
7. Leading peak
A. Unsuitable chromatographic Change chromatographic system
system

materials and solvents. A high-performance liquid chromatograph may be used

by many operators under different conditions. All operators should therefore
study the maintenance manuals and keep a stock of frequently required
replacement parts, such as rotor seals and pump piston seals. Many toxic
organic solvents and eluent components are used in HPLC, and operators
should therefore use safety glasses and work in areas that have adequate
ventilation.

12 References
1. H. Colin, J. C. Diez-Mesa, G. Guiochon, T. Czajkowska, and I. Miedziak, J.
Chromatogr., 1978, 167,41.
2. A. Nagai, in ‘Advances in Liquid Chromatography’, eds. T. Hanai and H. Hatano,
World Science, Singapore, 1996, pp. 151-1 67.
3. W. Funasaka, T. Hanai, and K. Fujimura, J. Chromalogr. Sci.,1974, 12, 517.
4. T. Hanai, J. Hubert, G. Luissier, M. Bellavance, and M. Lefebvre, ‘Proceedings of 4th
International Symposium on Column Liquid Chromatography’, Boston, May 1979.
5. T. Hanai, I. Suzuki, and K. Nakanishi, Anal. Sci.,1985, 1,483.
30 Chapter 2
Reference Books
J.W. Dolan and L.R. Snyder (eds.), ‘Troubleshooting LC Systems’, Humana Press,
Clifton, 1989.
E. Katz (ed.), ‘Quantitative Analysis using Chromatographic Techniques’, Wiley,
Chichester, 1987.
E.S. Yeung (ed.), ‘Detectors for Liquid Chromatography’, Wiley, New York, 1986.
H. Parvez, M. Bastart-Malsot, S. Parvez, T. Nagatsu, and G. Carpentier (eds.),
‘Electrochemical Detection in Medicine and Chemistry’, VNU Science Press,
Utrecht, 1987.
R.W. Frei and K. Zech (eds.), ‘Selective Sample Handling and Detection in High-
Performance Liquid Chromatography’, Elsevier, Amsterdam, 1988.
K. Blau and J.M. Halket (eds.), ‘Handbook of Derivatives for Chromatography’, 2nd
edn, Wiley, Chichester, 1993.
T. Hanai and H. Hatano (eds.), ‘Advances in Liquid Chromatography’, World Science,
Singapore, 1996.
CHAPTER 3

Preparation, Testing, and

Selectivity of Stationary Phase
Materials

The selection of the stationary phase material is generally not difficult when the
retention mechanism of the intended separation is understood. Stationary phase
materials have different physical and chemical properties that can be under-
stood from their chemical structure. They can be classified according to their
retention mechanism and chemical structure (Table 3.1). The symbol +
indicates a strong interaction. All stationary phase materials can be classified
in this manner, including chiral and affinity phases. For example, alkyl- or
phenyl- bonded vinyl alcohol gels are used in reversed-phase liquid chromato-
graphy, and ion-exchange group-bonded vinyl alcohol gels can be used as cation
or anion-exchangers. Non- bonded materials can be used for size-exclusion
liquid chromatography.
The fundamental behaviour of stationary phase materials is related to their
solubility-interaction properties. A hydrophobic phase acts as a partner to a
hydrophobic interaction. An ionic phase acts as a partner for ion-ion inter-
actions, and surface metal ions as a partner for ligand complex formation. A
chiral phase partners chiral recognition, and specific three-dimensional phases
partner affinity interactions.
Stationary phase materials are synthesized from different raw materials.
Those stationary phase materials synthesized from inorganic materials, such as
silica and alumina, are physically strong but chemically unstable. Conversely,
stationary phase materials synthesized from organic materials, such as poly-
styrene or poly(viny1 alcohol), are chemically stable but physically weaker.
Improvements in the chemical stability of inorganic stationary phase materials
and in the physical strength of organic stationary phase materials are required;
the marketed products do not have both and have to be used under restricted
conditions in liquid chromatography.

31
32 Chapter 3
Table 3.1 Stationary phase material selection guide

Stationary phase
material

Retention mode
Normal-phase LC
Chemisorption, chelate + a 0
Charge transfer ++OOA 00 n
Physical adsorption, hydrogen bonding b + + n
Reversed-phase LC 00000000000
Ion-exchange LC ++
Size-exclusion LC o o v vvv
Affinity, optical isomer +
Hydrogen bonding acceptor
Hydrogen bonding donor
4: strong interaction; 0: possible interaction, a: weak interaction; V: a wide variety of stationary
phase materials available; a Ferrous ion in silica; especially COOH form of ion-exchanger.

1 Synthesis of Stationary Phase Materials

The selectivity of stationary phase materials can be understood if the method of
their synthesis is understood. Differences in the same type of stationary phase
material from different manufacturers or even from the same manufacturer
depend on the synthetic methods and the quality control that has been
employed. Details of the individual synthetic processes from different manufac-
turers have not been published, but are basically the same.

Silica Gel
Silica gel is a three-dimensional polymer of silicic acid, usually synthesized from
tetrachlorosilane or sodium silicic acid. The reaction is as follows:

either
or
SiC14 + 2H20
NaSi04 + 4(H -form ion-exchanger)
+ -
+ 2Ag20 -+4AgC1 + Si(OH)4
ion-exchange

Si(OH)4 + 4(Na -form ion-exchanger)

+
Preparation, Testing, and Selectivity of Stationary Phase Materials 33
The removal of sodium and other metal ions is very important to obtain pure
silica gels. The next step in both routes is:

2Si(OH)4 + (OH)3SiOSi(OH)3 + H 2 0 + silica gel

Silica gel is also made from colloidal silica (silica sol):

Si(OH)26- -+Si(OH)4 + silica gel

The particle diameter and pore size of the silica gel materials depend on the
concentration of silicic acid, the pH, the solvent and mixing conditions, and the
reaction temperature. Technical skill is necessary for reproducible large-scale
synthesis.
The purity of silica gel made from sodium silicate depends on the ion-
exchange stage, and is usually about 99%. A purer silica gel can be produced
from organosilica compounds. In this method tetraethoxysilane is polymerized
to polyethoxysilane. The emulsion of polyethoxysilane in a water-alcohol
mixture is further dehydroxylated and polymerized with a catalyst. The
hydrolysed silica gel is sieved, dried, and sieved again before use as a stationary
phase material. The pore size and particle size depend on the concentration of
polyethoxysilane, the solvent, the temperature, the amount of catalyst, and the
mixing conditions. Strong agitation leads to a smaller particle size, and a large
amount of catalyst leads to a large pore size. The purity of this silica gel is more
than 99.99%, depending on the purity of the water used for washing and
sieving.

-.
The surface of the virgin silica gel is covered with water. Heating to 110 “C
forms a monolayer of silanol groups on the surface (8 pmol of silanol groups per
m’). Heating to over 600°C produces a surface siloxane structure, and this is
rehydrated to the silanol form under humid conditions, as shown in Figure 3.1.
Electron microscopy photographs of classic irregular (A) and spherical (B)
silica gels and high purity, spherical silica gel (C),whose purity is over 99.99%,
are shown in Figure 3.2. The surface of pure silica gel is very smooth and
physically very stable due to the homogeneous structure. Modification of the

\
Si

-600°C ).
Si

H20 K \
Si
/O

Figure 3.1 Modification of silica gel surface.

34 Chapter 3
Preparation, Testing, and Selectivity of Stationary Phase Materials 35
surface is easier (see section below on surface modification) than for ordinary
silica gels.

Surface Mod8cation of Silica Gel

Spherical porous silica gel is the easiest stationary phase material to handle;
however, although it is physically strong it is chemically unstable. Surface
modification can expand its capability for different modes of chromatography,
such as normal-phase, reversed-phase, size-exclusion, and ion-exchange liquid
chromatography. These stable modifications are performed by chemical deriva-
tization of the surface silanol groups.

ModiJication Using a Halogenated Silica Surface. Silica gel is heated in 2 M HC1,

filtered off, and washed with water; the dried silica gel is then reacted with pure
thionyl chloride. After removal of the excess of thionyl chloride and by-
products (SO2 and HCl) under vacuum, chlorinated silica gel (FSiCl) is
obtained. This is unstable in water and alcohol. The chlorinated silica gel can
then be alkylated by one of a number of different reactions:

-Si-OH + (SOCl2, Tic&, or SiC14) + -Sic1

-Sic1 + (RMgC1) -+ r S i R + MgC12 (Grignard reaction)
-Sic1 -+(RLi) -+ -SIR + LiCl
-Sic1 + (RNH2) + r S i N H R + HC1

The chlorinated silica gel suspended in diethyl ether, dimethyl sulfoxide, or

dioxaiie will also react with diamine or amino compounds. After being washed
in an excess of amine and HC1, and then dried under vacuum, the final
stationary phase materials are obtained:

=Sic1 + H2N-(CH2),-X -+ -Si-NH-(CH2),-X + HCl

X = CH3, NH2, C02H, S03H, CN, or NO2

Further modification of diamino bonded products with halogenated com-

pounds can lead to a variety of stationary phase materials:

-Si-NH-( CH2),-N H2 + Y-CH2-R-X -+ =Si-N H-( CH2),-NH-CH2-R-X

R = CH2, C6H4

Modification Using a Chlorosilane. Silica gel is heated in 2 M HCl, filtered off,

and washed with water. The dried silica gel is then boiled under reflux in toluene

Figure 3.2 Electron microscopic photos of silica gels: A, irregularly shaped; B, ordinary
spherical; and C, high purity silica.
(Contributed by Hideyuki Negishi, Kanagawa Dental College, Yokosuka,
Japan)
36 Chapter 3

Figure 3.3 Molecular model of butyl-bonded silica gel.

containing the chlorosilyl reagent, using pyridine as the catalyst. Three types of
chlorosilyl reagents are available:

3 ESi-OH + RSiC13+ =SiO3SiR + 3HC1

2 GSi-OH + R2SiC12+ rSi02SiR2 + 2HCl
-Si-OH + R3SiC1+ -SiOSiR3 + HC1
A depiction of a butyl-bonded phase is shown in Figure 3.3. Monochloro-
silane reagents produce only monomeric phases; however, trichlorosilane
reagents can produce both monomeric and polymeric phases depending upon
the concentration of silyl reagent and the surface area. It is difficult to make a
multilayer bonded phase, even when a large quantity of trichlorosilane is used
for the reaction.

Synthesis of Ion-exchanger Stationary Phases

Cation-exchanger. Phenyl-bonded silica gel is suspended in chlorosulfonic acid-
acetone and boiled under reflux. After filtration and washing with acetone and
dilute HC1, the sulfonated gel is then converted into a suitable salt form:

Anion-exchanger. Phenyl-bonded silica gel is suspended in chloromethyl methyl

ether and boiled under reflux. Then, chloromethyl methyl ether containing
anhydrous ZnC12 is added and the mixture boiled under reflux. After filtration,
Preparation, Testing, and Selectivity of Stationary Phase Materials 37
the gel is washed using 1,4-dioxane and diethyl ether. The washed silica gel is
suspended in trimethylamine- 1,4-dioxane solution, and the mixture stored at
0 "C to give the quaternary ammonium ion exchanger:

ESi-Ph + -Si-Ph-N+Me3 C1-

Organic Stationary Phase Materials

A variety of organic polymer gels are available for use as stationary phase
materials. Their physical strengths and hydrophobicity depend on the proper-
ties of the monomers. Dextran gels are polar and soft. Vinyl alcohol
copolymer gels are fairly polar. Polyacryl gels are physically hard and
relatively non-polar. Polystyrene gels (a copolymer of styrene and divinylben-
zene) are physically very strong and very hydrophobic. A higher divinylben-
zene concentration makes a physically stronger polystyrene gel. The particle
size, pore size, physical strength, and wettability in solvent depends on the
reaction solvent, temperature, mixing conditions, and catalyst. Furthermore,
ion-exchange groups can be introduced on the polystyrene surface to make
ion-exchange resins.

2 Sieving of Stationary Phase Materials

The narrower the particle size distribution, the higher in theory is the potential
theoretical plate number. A rough sieving is achieved by a water flow, air flow,
or a vibration method. A common sieving method is Hamiltonian water flow
(Figure 3.4). The particle distribution can be controlled within 1 pm by this
method. A slurry of stationary phase material is allowed to float in the
cylinder, and a solvent flows from the bottom to the top. The smaller and
lighter particles float to the top of the cylinder and the larger and heavier
particles sink to the bottom. The required particles are collected at the top of
the cylinder. The selection of suspension solvent and control of the tempera-
ture are important.

3 Column Packing Methods

Dry packing is usually only used for large particle-size stationary phase
materials, especially in preparative-scale columns. High-performance
analytical-scale columns are packed by the wet slurry method. Balanced or
viscous slurry methods are applied to reduce the sedimentation of the stationary
phase materials during the packing process. Previously, balanced density
slurries were made by mixing high-density halogenated solvents. However,
halogenated solvents are toxic, so the viscous slurry method has been
developed, which uses an ethylene glycol-alcohol suspension mixture.
First, an empty analytical column, a pre-column, and a slurry reservoir are
connected in series. The narrow-bore analytical column and pre-column are
38 Chapter 3

suspension solution

\ I /

collection

Water flow +
0 1

Figure 3.4 Hamiltonian particle sizing system.

filled with a packing solvent such as methanol. The reservoir is filled with the
stationary phase slurry, and then the packing solvent is quickly pumped
through the system under high pressure, depending on the physical strength of
the stationary phase material.
For example: for the preparation of a 15 cm long, 4.6 mm i.d. stainless tube
column, 2.5 g of octadecyl-bonded silica gel was suspended in 25 ml of hexanol-
methanol mixture, and kept in an ultrasonic bath for a few minutes to remove
air. After the reservoir was filled with the slurry, methanol was pumped in at
10 ml min- under constant pressure, 45 MPa (450 bar). After the replacement
of slurry solvent by methanol, the flow was stopped and the pressure allowed to
drop. When OMPa was reached the reservoir was removed. Then, 20ml of
water was added and methanol was again pumped in under the same conditions
as before. Again, the flow was stopped and the pressure allowed to drop until it
reached 0 MPa. The pre-column was removed and the analytical column closed.
The maximum pressure that can be applied in the filling stage is based on the
pore size, particle shape, and purity of the silica gel. This reproducible packing
procedure is performed at constant temperature by using a water bath (60-
80 "C).
Preparation, Testing, and Selectivity of Stationary Phase Materials 39

4 Column Evaluation
The performance of columns is determined by using a simple procedure for both
new and used columns. The test method and the results are usually printed on a
sheet provided by the column manufacturer. Each manufacturer uses the most
suitable method for their column to obtain the best theoretical plate number.
The column evaluation provides important information for all users of liquid
chromatography.

Column Efficiency and Asymmetry

Common standard compounds for reversed phase columns are toluene and
naphthalene, which have retention factors, k , of about 3. The eluent modifier is
methanol or acetonitrile at a concentration of 50-80%, depending on the
hydrophobicity of the stationary phase material. For other stationary phase
materials, corresponding analytes, with k = 3-5, can be used.
The efficiency of a well packed column should be about 95 000 m-' 5% for
a 5 pm octadecyl-bonded silica gel column. The equations for the calculation of
the number of theoretical plates (N)and peak asymmetry ( A s ) are as follows. A
model chromatogram is given in Figure 3.5.
The value of N is given by:

where the retention time ( t R ) is 5.87 min in Figure 3.5. Assuming the peak is
a Gaussian curve, the peak width at the base is considered to be 40 where 0 is
the uncertainty in the retention times of individual analyte molecules. The
peak width (in the figure) at half-peak-height (w+)is 0.17 min; and wb,
the peak width at the base line, is 0.30 min. Therefore, N is 6605 from w;,and
6126 from wb. The difference is due to the uncertainties in the measurement of
the peak width. For comparison between columns of different lengths the
efficiency is often expressed as plates per m [ 1OO(N/L)],where L is the length of
the column in cm.
The value of the peak asymmetry As is calculated from:

As = b/a

where a and b are, respectively, the widths of the front and back of a peak
measured at 10% peak height from the baseline. The a and b values (Figure 3.5)
are 0.12 and 0.16 min, respectively. Therefore, As is 1.33. The desired As value is
1.0-1.25 for good separations and a long column life. The lifetime of columns
with As < 1.O is usually short, due to an inhomogeneously packed bed.
40 Chapter 3

\
I I

V,O = 1.59J' 4 6 min

\ = 5.07 4

A
I
I
./ I. \ .
J

Figure 3.5 Schematic of a typical chromatogram. V,, void volume; w+,width at halfpeak
height; wb, width at base; lOh, peak height; a, width of front at l h ; b, width of
back at l h ; wi = 20.10h.

Column Test for Reversed-phase Liquid Chromatography

Inertness Test for Basic Compounds
A bonded stationary phase whose silanol groups are completely modified by
alkylation is, in theory, inert towards basic compounds. The lifetime of such
columns is long, and can be maintained for more than 2000 h in 1% trifluoro-
acetic acid solution (pH 2) or 0.05 M disodium phosphate solution (pH 8.6).
Such a column will withstand the passing of more than 10 litres of trifluoro-
acetic acid solution for continuous peptide separation. The peak asymmetry of
basic compounds is excellent. The extent of unreacted silanol groups can be
measured by the chromatographic behaviour of pyridine compared with
The relationship of the different surface activity to the number of
silanol groups is shown in Figure 3.6. If no active silanol groups are present, the
Preparation, Testing, and Selectivity of Stationary Phase Materials 41

A B C
1
! 2 2

3 3

I
0
I
10
I
0
I I
4 0
h I
6
min
Figure 3.6 Test f o r active silanols on octadecyl-bonded silica gels. Column: 5 pm
octadecyl-bonded silica gel, 15 cm x 4.6 mm i.d.; eluent, 60% aqueous
acetonitrile; flow rate, 1 ml min- ';
temperature, ambient; detection, UV
254 nm. Peak 1, pyridine; 2, phenol; and 3, toluene. A, Column with no active
silanol groups; B, some active groups; and C , numerous active groups.

pyridine peak is sharp and its symmetry is like that in chromatogram A.

Chromatogram B indicates the existence of some active silanol groups. The
pyridine is retained longer than phenol and the peak shape is poor. In
chromatogram C pyridine was completely adsorbed and could not be observed.
The lifetimes of columns B and C were short in buffered solution. These
columns are not suitable for the chromatography of nitrogen-containing basic
compounds due to a long tailing peak or the disappearance of the peak by
strong adsorption.

Inertness Test for Acidic Compounds

The retention and the peak asymmetry of benzoic acid also indicate the
inertness of the bonded phase. If basic compounds remain on the surface or
are used as reagents, the peak asymmetry of benzoic acid is poor. The peak
height is lower than that of the same quantity of o-toluic acid.3i4 This
phenomenon is observed if the basic catalyst that was used in the synthesis
process has not been completely washed off the stationary phase or if active
amino groups remain. This type of column is not suitable for the separation of
acidic compounds.

Inertness Test f o r Heavy Metals

Active compounds often form a chelate with heavy metal ions, and elution from
a column containing heavy metal ions is poor.3y4The existence of heavy metals
42 Chapter 3
on the surface can be examined by using 8-hydroxyquinoline, which is used for
heavy metal extraction from water. However, 8-hydroxyquinoline is regarded as
a human carcinogen. Therefore it has to be used under controlled conditions.
Other chelation reagents, such as pentan-2,4-dione and 2,3-dihydroxy-
naphthalene, can be used for the trace metal test, but they are not as sensitive
as 8-hydroxyquinoline. The metal contents of different grades of silica gels are
summarized in Table 3.2. Extra pure and pure silica gel contain only trace
amounts of heavy metals but contain traces of alkaline metals from the distilled
water used mainly during the sizing procedure. The effect of heavy metals in
relation to the purity of silica gel is shown in Figure 3.7. The purer the silica gel,
the sharper the peak shape, as in chromatogram A. From the peak tailing,
Chromatogram B indicates the existence of heavy metals on the surface. The

Table 3.2 Metal content of silica gels of diferent purity

Silica gel PS*" p ~ * SA*'

~ Metallppm
(pm) (A) (m2g-') Na Mg A1 Ca Ti Fe Zn

Ordinary 5 122 333 190 250 150 730 160 22 19

Purified 5 120 350 5 3 120 10 130 50 nd
Pure 10 100 350 6 nd 63 nd 70 13 6
Extra pure 5 100 450 5 1 5 9 n d 1 1

size; pore size; surface area; nd, not detected by inductive coupled plasma spectroscopy
a Particle
(ICP), Cr, Mn, and Ni were not detected.

B C

0 10 0 30 0 30
elution time min

Figure 3.7 Test for trace metal ions on the surface of octadecyl-bonded silica gel. Column
A, 5 pm octadecyl-bonded silica gel, 15 cm x 4.6 mm i d . ; By octadecyl-
bonded silica gel, 15 c m x 4.6 mm i.d.; C, octadecyl-bonded 10 pm silica gel,
30 cm x 4.0 mm i.d.; eluent, 50% aqueous acetonitrile; flow rate,
1 ml m i n - ' ; temperature, ambient; detection, UV 220 nm (the scales of A, B,
and C are not the same); sample, 8-hydroxyquinoline. Columns; A, metal ion
free; B, low proportion of metal ions; C , high level of metal ions.
Preparation, Testing, and Selectivity of Stationary Phase Materials 43
peak of 8-hydroxyquinoline could not be observed on chromatogram C. This
also indicates that the surface treatment was poor. Pure silica gels
(purity > 99.99%) are necessary for the separation of porphyrins. For the
separation of biologically active compounds, bonded phases made from pure
silica gels whose purity is known and guaranteed should be used.

Physicochemical Tests of the Stationary Phase

The carbon content of a stationary phase is measured by an elemental analyser,
as a weight balance before and after heating at 800°C. Particle size, pore size,
and surface area are measured by specific instruments, such as a particle size
analyser, nitrogen adsorption porosimeter, and mercury depression analyser,
respectively. The precision of the measurement of carbon content is high;
however, that of the other measurements is relatively poor. Therefore, it is
difficult to relate the surface area of different silica gels to analyte retention
factors.

5 Measurement of Void Volume

The retention factor, k, is the basic value in chromatography, and is related to
the void volume (dead volume). The void volume is the space inside the column,
where no retention of solutes has occurred and can be measured on a
chromatogram, as shown in Figure 1.3. The void volume is about half the total
volume of the column when it is packed with porous stationary phase materials.
In practice, the effective void experienced by the analyte is smaller because the
molecular mass of the analyte is usually much greater than that of the eluent
molecule. In a model of porous stationary phase material, the pores can be
represented as V-shape valleys (Figure 3.8), where region a is a support, such as

A B

a
b
C

Figure 3.8 A, Schematic structure of porous stationary phase; a, support; b, bonded

phase; and c, adsorbed solvent layer. B, detailed view of eluent components in
surface layer; , lower phase molecules; 0, upper phase molecules.
44 Chapter 3
silica gel or polystyrene gel, b is the bonded phase, such as octadecyl or ion-
exchange groups, and c is the solvent layer, which is the adsorbed components
of the eluent. In Figure 3.8, if a sample molecule is small it can pass to the
bottom of the valley. However, a larger sample molecule, such as a protein, is
large, and ‘jumps’ the valley if the pore size is small. The real void volume for a
particular analyte thus depends upon its shape and mass. The pore size
distribution of the stationary phase materials also affects the void volume of
individual analytes. The exact value cannot be measured. A practical com-
promise method is examined below.
In normal-phase liquid chromatography, the elution volume of the solvent is
nearly constant in eluents having different mixtures of solvent components. The
solvent peak can therefore be used to measure the void volume. In reversed-
phase liquid chromatography, the thickness of the organic solvent layer, which
is c in Figure 3.8A, depends on the concentration of organic modifier and
additives. In addition, an ion-exclusion effect is observed for ionized solutes,
and the elution volume is often less than the exclusion limit of the column.
However, measurement of the void volume is necessary to calculate k. Figure
3.8B indicates the microscopic interface of a partition. The symbols are lower-
phase molecules and o are upper-phase molecules in a solvent partition. The
distribution between the phases b and c is not clear-cut, as in a partition model.
Phase b incorporates the bonded phase and the concentration of the organic
component of the eluent is higher than in the bulk eluent; it gradually reduces
towards the mobile phase in reversed-phase liquid chromatography. The
effective void volume is therefore not constant.

Elution Volume of Proposed Void Volume Markers

Several compounds have been proposed for the measurement of the void
volume, including sodium nitrate solution, water, deuterium oxide, fructose,
acetonitrile, tetrahydrofuran (THF), meso-erythritol, gluconolactone, and 2,4-
dinitronaphthol. The elution volume of a number of these compounds has been
measured in 10-90% aqueous acetonitrile and acidic-aqueous acetonitrile. The
results are given in Figure 3.9 where the volumes in A and B were measured in
aqueous acetonitrile and in C and D were measured in aqueous acetonitrile
containing 50 mM phosphoric acid. Methanol (a) and deuterium oxide (g)
showed two peaks when monitored by a refractive index detector (Figure 3.9C).
Even the elution volume of acetonitrile was not constant in these eluents and
varied from 1.25 to 1.95 ml. The volumes of ionized sodium nitrate (e) and 2,4-
dinitronaphthol (d) were smaller than the exclusion limit (1.0 ml) of this
column. The dramatic change for 2,4-dinitronaphthol in Figure 3.9A indicates
that the volume depended strongly upon the ratio of eluent component^.^
In acidic eluents (Figure 3.9C and D), the retention of acidic compounds
becomes stronger and that of basic compounds becomes weaker. In this system,
uric acid (h) is not suitable as a void volume marker due to its longer retention
time. In neutral and basic eluents, an ionized acid can be used as the marker
because no other compounds are eluted more rapidly. Fructose (c) is a very
Preparation, Testing, and Selectivity of Stationary Phase Materials 45

mL

0.5 1 I J 0.5 I I
0 50 100 0 50 100

2.0
gi/ 2.5

1.5

mL

1.o

1.oi io 11.0
100
- 50
% conc. water in acetonitrile
100

Figure 3.9 Elution volumes ( m l ) of alternative void volume markers. Column, 5 pm

octadecyl-bonded silica gel, 15 cm x 4.5 mm i.d.; eluents A and B, 10-90%
aqueous acetonitrile, eluents C and D, 10-90% aqueous acetonitrile containing
50 mhfphosphoric acid;flow rate, 1 ml min- ’; temperature, 30 “C; detection,
UV 210 nm and refractometer. Sample: a, acetonitrile; b, methanol;
c, fructose; d, 2,4-dinitronaphthol; e, sodium nitrate; f, tetrahydrofuran;
g, deuterium oxide; and h, uric acid.

polar compound, and is not adsorbed on a hydrophobic stationary phase

surface. It is also UV-active at short wavelengths and non-ionizable under
practical conditions. Its elution volume is nearly constant in the above condi-
tions. The effect of injection volume and the necessity of a refractive index
detector are thus avoidable. Therefore, fructose can be used as the standard
void volume measuring marker. When analytes such as steroids are examined,
they are excluded from many of the pores and the molecular size of fructose is
too small to measure their void volume. Maltotriose may then be the compound
of choice for the void volume measurement.
46 Chapter 3

6 Selectivity of Stationary Phase Materials

Silica Gel-based Stationary Phase Materials
Separations can be performed on a single column by using different eluents.
Practical separations can also be achieved using column selection. In over 50%
of applications octadecyl-bonded silica gel is used as the hydrophobic stationary
phase material. As an alternative, phenyl-bonded silica gel is less hydrophobic
compared with octadecyl- or octyl-bonded silica gel. This means that a smaller
amount of organic modifier is necessary to obtain the same retentions. In
addition, aromatic compounds are relatively more highly retained on phenyl-
bonded silica gel than on octadecyl-bonded silica gel, and alkanes are more
retained on octadecyl-bonded silica gel than on phenyl-bonded silica gel as
shown in Figure 3.10. Such selectivity can be useful for the separation of
compounds having different degrees of aromatici ty.
Many manufacturers sell the same types of stationary phase materials,6 but
the most popular stationary phase materials, e.g. octadecyl-bonded silica gels,
from different sources, even from the same manufacturer, often demonstrate
different retention capacities and selectivities. Such differences are due to the
aggressive reactivity of the silylation method and the different bonding reactions
that are used, as described earlier.
The selectivity differences can be understood by the comparison of four
octadecyl-bonded silica (ODS) phases synthesized from the same silica gel. The

O+!;:
-0.5
-0.5 0
,
0.5 1
, 1
1.5

logk on octadecyl-bonded silica gels

in 70% aq. acetonitrile
Figure 3.10 Selectivity of phenyl-bonded silica gel compared with ODs silica gel.
Columns: ODs-silica, Y M C O D s , 15 cm x 6.0 mm id., phenyl-bonded
silica gel, Y M C phenyl, 10 cm x 6.0 mm i.d.; eluent, aqueous acetonitrile.
Compounds: x , polycyclic aromatic hydrocarbons; 0,alkylhenzenes;
0 ,polychlorobenzenes; + , alkanols; A,alkanes; 1, benzene; 2, benzopyrene;
3 , toluene; 4, heptylbenzene; 5 , hexachlorobenzene; 6 , hexanol; 7 , tetra-
decanol; 8, pentane; and 9, octane.
Preparation, Testing, and Selectivity of Stationary Phase Materials 47
four materials are a high-carbon (H1C)-loaded silica gel with and without a
second silylation (end-capping) (HIC-ODs-E and HIC-ODs-NE, respectively)
whose carbon content is 16 wt%, and a low-carbon (L0C)-loaded silica gel with
and without a second silylation (LOC-ODs-E and LOC-ODs-NE, respectively)
whose carbon content is 8.9 ~ t % The . ~ selectivity-related hydrophobicity
differences can be compared using the differences in the retention factors of
homologous alkylbenzenes as the van der Waals volume increases (Figure 3.1 1).
The hydrophobicity, i.e. retention capacity, of HIC-ODS was found to be about
twice that of LOC-ODs. The end-capping treatment further increased the
retention capacity. This result indicates that if the quality control of the
bonding processes is poor, the products will have different retention capacities.
This difference also depends upon the surface area of silica gels.
The relationship of the selectivity to the polarity of the analytes can be
understood from the differences in the retention factors of homologous alkanols
(Figure 3.12). The polar alkanols are relatively more retained on the non-
endcapped bonded phases (LOC-ODs-NE and HIC-ODs-NE) because
smaller-size alkanols can reach the unreacted silanol groups on the surface of
silica gels.
The relationship of the selectivity towards n-electrons can be understood
from the differences in the retention factors of polycyclic aromatic hydrocar-
bons (Figure 3.13). The difference in the retention factors on end-capped and
non-endcapped stationary phase materials is less than that of alkylbenzenes.
This is due to the water content of the stationary phase. The content may be
higher in non-endcapped bonded phases.
The best octyl- and octadecyl-bonded silica gels should be chemically stable in
highly basic solutions. The well-bonded phases give good results in the inertness

d
l T---

'1

-1 1
75
25 125 1 '5

van der Waals volume cm3/mol

Figure3.11 Comparison of the retention of toluene to decylbenzene as a measure of
hydrophobicity of different octadecyl-bonded silica gels with different carbon
loadings. Columns: 0, LOC-ODs-E; 0 ,LOC-ODs-NE; 0, HIC-ODs-E;
A,HIC-ODs-NE; eluent, 80% aqueous acetonitrile at 30 "C. Compounds:
1, toluene; 2, nonylbenzene.
48 Chapter 3

Y
8 0.5 8
0
8 0
Q n 2
rc
n o 0

0 P O O
U
0
5
c,
-0.5 0

-1 I
50 75 100 125 1 i0

van der Waals volume cm3/mol

Figure 3.12 Polar selectivity of diflerent octadecyl-bonded silica gels towards the reten-
tion of homologous alkanols. Columns: 0, LOC-ODS-E; 0 , LOC-ODS-
NE; 0, HIC-ODs-E; A,HIC-ODs-NE; eluent, 70% aqueous acetonitrile
at 30 "C. Compounds: 1, butanol; 2, dodecanol.

1
0
Y

-8 0.5- 0
A
g A0
5
c,

!i
rc
O- s
xo 0
E 0 o o 4
.-0 3
s
c,

+.r
-0.5- 0
0 2
2 1
-1 I I I I

25 50 75 100 125 1 i0

van der Waals volume cm3/mol

Figure 3.13 n-Electron selectivity of different octadecyl-bonded silica gels towards the
retention of polycyclic aromatic hydrocarbons. Columns: 0, LOC-ODs-E;
0 , LOC-ODs-NE; 0,HIC-ODs-E, A,HIC-ODs-NE; eluent, 80%
aqueous acetonitrile at 30 "C. Compounds: polycyclic aromatic hydro-
carbons; 1, benzene; 2, naphthalene; 3 , pyrene; 4,3,4-benzopyrene.

tests (see above), and can be used for continuous analysis in basic solutions, pH
9.0, over a one month period. Such stability and inertness are guaranteed by the
manufacturers. All silica-loaded materials are unstable in high pH eluents. The
amount of carbon loading can be more than 110% of the theoretical calcu-
lation; however, the accuracy of surface area measurement is less than that of
the carbon loading measurement. Non-endcapped bonded phases sometimes
Preparation, Testing, and Selectivity of Stationary Phase Materials 49
demonstrate better relative separations; however, their long-term operation is
not guaranteed. The lifetime of short-chain bonded phases, such as butyl,
cyano, and amino, is also short in buffer solutions due to their weak hydro-
phobicity. It seems that these phases do not reject the adsorption of ions from
the eluent on to the untreated silanol groups.

Organic Polymer-based Stationary Phase Materials

Surface-modified silica gels are used for a variety of separations, but organic
polymer-based stationary phase materials are more useful for long-term opera-
tions, such as for an amino acid analyser and size-exclusion liquid chromatog-
raphy.
Physically strong organic porous and non-porous polymer gels are also used
for high-performance liquid chromatography. When difficulty in achieving
separation is encountered with silica gel-based stationary phase materials,
organic gels will often solve this problem even though the theoretical plate
number may be less than that of silica gel-based stationary phase materials.
For example, polystyrene gel is very hydrophobic, and aromatic compounds
are selectively retained. Polystyrene oligomers are separated on a polystyrene
gel using isocratic elution (Figure 3.14). Vitamins, anions, and nucleic acids and
bases have been chromatographed on polystyrene gel.* These separations can
also be carried out on alkyl group-bonded silica gels, if the silanol groups are
completely ~ o v e r e dLarge
. ~ pore-size polystyrene gels are useful for the separa-
tion of proteins. Proteins with M , = 6000-66300 have been separated on a

0 min 35
Figure 3.14 Reversed-phase liquid chromatogram of polystxrene oligomers. Conditions:
columns, PLRP-S, 15 cm x 4.6 mm id., 100 A (polystyrene gel); eluent,
70% aqueous tetrahydrofuran;Jow rate, 1 ml min-'; detection, UV 254 nm.
Numbers beside peaks indicate the molecular weight.
(Reproduced by permission from Polymer Laboratories data)
50 Chapter 3
A
300 pore size polystyrene gel, and M , = 120 000 collagen and 340000
A
fibrinogen have been separated on a 1000 pore-size polystyrene gel, as
shown in Figure 3.15. Silica gel-based bonded stationary phase materials with
A
a 1000 pore size are not as physically strong as the narrow pore-size silica gels.
However, polystyrene gels have the advantages of surface inertness and
moderate physical strength.
Methacrylate polymers are as physically strong as polystyrene gels. Their
hydrophobicity is weaker than that of polystyrene gel, and the aromatic
selectivity is similar to that of octadecyl-bonded silica gel. A chromatogram of
hydantoins on a methacrylate polymer is shown in Figure 3.16. The elution
order is related to the hydrophobicity of the chemicals.
Vinyl alcohol copolymer gel is hydrophilic and has been developed for
aqueous-phase size-exclusion liquid chromatography; however, it is less polar
than the polysaccharides. Its specificity permits the direct injection of a
biological sample without deproteinization. For example, blood serum from a
patient suffering from chronic nephritis has been injected directly as a measure
of the degree of dialysis (Figure 3.17). Adenosine triphosphate, adenosine
diphosphate, and adenosine monophosphate in red blood cells have also been
separated directly (Figure 3.18). Theophylline in blood serum has been

1 2

1
0
I
I
12
4
I
0
Figure 3.15 Chromatogram of jibre-type proteins
I
min 12
J

polystyrene gels having diflerent

pore sizes. Column A, PLRP-S 300 A, 15 cm x 4.6 mm i d . ; B, PLRP-S
1000 A (polystyrene gel), 15 cm x 4.6 mm i.d.; eluent, 15 min linear
gradient from 20% of 0.25% triJuoroacetic acid to 60% of 0.25% trfluoro-
';
acetic acid in 95% aqueous acetonitrile; flow rate, 1.O ml min- detection,
UV 220 nm. Peaks: 1, collagen ( M , 120 000) and 2,jibrinogen ( M , 340 000).
(Reproduced by permission from Polymer Laboratories data)
Preparation, Testing, and Selectivity of Stationary Phase Materials 51

I I I
0 4 min 8
Figure 3.16 Chromatogram of hydantoins on polymethacrylate stationary phase. Condi-
tions: columns, Shodex RSPak DE613 (methacrylate gel); eluent, 1:1
mixture of 0.033 M disodium hydrogen phosphate and potassium dihydrogen
phosphate; flow rate, 2.0 ml min-'; detection, U V 210 nm; temperature,
50°C. Compounds: 1, R = (CH2)ZCOzH; 2, R = CH20H; 3, R = H;
4,R = CH,; and 5, R = C2H5.
(Reproduced by permission from Shodex data)

3 4
1

J
- L
I I
0 20 min 40
Figure 3.17 Chromatogram of blood serum from a patient with chronic nephritis.
Conditions: columns, Asahipak GS320 (vinyl alcohol copolymer gel),
50 cm x 7.6 mm i.d.; eluent, 0.1 M sodium phosphate containing 0.3 M
sodium chloride pH 7.0; f l o w rate, 1 ml min-'; detection, UV 250 nm;
direct injection of sample. Peaks: 1 , protein; 2, orotidine; 3, creatinine; and
4,uric acid.
(Reproduced by permission from Asahipak data)

separated from proteins (Figure 3.19). In these separations, the proteins were
eluted first and the targeted compounds eluted later; the quantitative analysis of
the target compounds is thus easy. Pesticides in blood serum have also been
directly analysed on vinyl alcohol copolymer gels, as shown in Figure 3.20.
52 Chapter 3
1

0 10 min 20
Figure 3.18 Adenosine phosphates in blood on vinyl polymer column. Conditions: column,
Asahipak GS320 (vinyl alcohol copolymer gel), 50 cm x 7.6 mm i.d.; eluent,
0.1 M sodium phos hate bufler containing 3 M sodium chloride pH 7.O;flow
rate, 1.O ml min- {detection, UV 260 nm. Peaks: 1, haemoglobin; 2, adeno-
sine triphosphate; 3, adenosine diphosphate; and 4, adenosine mono-
phosphate.
(Reproduced by permission from Asahipak data)

I I
0 min 16
'igure 3.19 Theophylline in blood serum on vinyl alcohol polymer column. Conditions:
column, Asahipak GS320 (vinyl alcohol copolymer gel), 50 cm x 7.6 mm
i.d.; eluent, 0.01 M sodium acetate bufler pH 4.0 in 10% aqueous acetonitrile;
j?ow rate, 2 ml min-'; detection, U V 280 nm. Peaks: 1, protein; 2, low M ,
impurity; and 3 , theophylline.
(Reproduced by permission from Asahipak data)

These separations can be carried out using a silica-based bonded phase;

however, the important advantage of organic polymer stationary phase
materials is their chemical stability. The columns can be washed by using an
alkaline solution after a certain number of injections. According to the
chromatograms, the proteins in serum are completely eluted and nothing
remains inside the column. However, the pressure drop in this type of analysis
- '
Yreparation, Testing, and Selectivity of Stationary Phase Materials 53
1

i I
0 min 30
Figure 3.20 Pesticide in rat blood serum on vinyl alcohol polymer column. Conditions:
column, Asahipak GS320 (vinyl alcohol copolymer gel), 50 cm x 7.6 mm
i.d.; eluent, 0.05 M phosphate bufer containing 0.2 M sodium chloride pH 7.2
in 20% aqueous methano1;JIow rate, 1.0 ml min-'; detection, UV 254 nm.
Peaks: 1, blood serum protein and 2, methomyl.
(Reproduced by permission from Asahipak data)

may increase after several injections, depending on the injection volume. This
problem is not avoidable on all types of columns. Therefore, washing using an
alkaline solution is occasionally required.
The separation of basic and metal-sensitive compounds is difficult on silica-
based stationary phase materials, but these separations can be performed on
vinyl alcohol copolymer gels. Examples are the separation of methallothionein
from dolphin kidney, a-, b-, and y-endorphin, and nucleotide and nucleoside
mixtures.8 However, an analytical-scale separation may also be performed on
surface-modified wide-pore silica gels (pore size 300 A or more), using columns
which showed a negative response in the heavy metal test described above.
Modified vinyl alcohol copolymer gels with phenyl groups improved the
hydrophobicity and selectivity for aromatic compounds. Chromatograms of
protein mixtures and enzymes (phosphoisomerase) are shown in Figures 3.2 1
and 3.22, respectively. The separation mechanism on the surface of these
stationary phase materials is basically the same as that for reversed-phase
liquid chromatography, and is called hydrophobic liquid chromatography. The
chromatograms of protein mixtures shown in Figure 3.23 were obtained on a
phenyl-bonded vinyl alcohol copolymer gel and a trimethylsilyl-bonded silica
gel using basically the same eluent. The elution order of proteins was the same,
but the resolution was a little different. Octadecyl-bonded vinyl alcohol
copolymer gels are chemically stable and overcome the weakness of octadecyl-
bonded silica gels, which are influenced by unreacted silanol groups and high
pH instability.'
The selection of inorganic or organic polymer stationary phase materials
depends on the application. When information is needed rapidly, silica-based
bonded phases have the advantage due to their physical strength and high
theoretical plate number, although their lifetime is shorter. When a longer-term
operation is necessary, such as for an amino acid analyser, organic polymer-
based stationary phase materials are appropriate. Organic polymer-based
54 Chapter 3
1 4

I I I
0 30 min 60
Figure 3.21 Separation of proteins on wide pore phenyl-bonded column. Conditions:
column, TSKgel Phenyl-5 P W , 75 mm x 7.5 mm i.d.; eluent, 60 rnin linear
gradient of ammonium sulfate concentration from 1.7 to 0 M in 0.1 M
';
phosphate bufler (pH 7.0);JIow rate, 1.O ml min- detection, UV 280 nm.
Peaks: 1, cytochrome c; 2, myoglobin; 3, ribonuclease; 4, lysozyme; 5, a-
chymotrypsinogen; and 6, a-chymotrypsin.
(Reproduced by permission from Tosoh data)

I I I
0 15 rnin 30
Figure 3.22 Chromatogram of phosphoisomerase. Conditions: column, TSKgel Phenyl-5
P W RP, 75 mm x 4.6 mm i.d.; eluent, 2 min linear gradient of acetonitrile
from 0.05 to 20% followed by 48 min linear gradient of acetonitrile from 20
to 80% in 0.05% trljluoroacetic acid;_flow rate, 1 ml min-'; detection, UV
220 nm.
(Reproduced by permission from Tosoh data)

stationary phase materials can be recycled, but silica-based stationary phase

materials are used only once; this is another selection point, which depends on
the economic considerations.
The disposal of used stationary phase materials, especially those used for
biomedical applications, should also be of concern. Silica gel-based stationary
phase materials can be heated at high temperature before disposal, and organic
polymer-based stationary phase materials can be burned. Take care.

Other Stationary Phase Materials

Hydroxyapatite and titania are also useful stationary phase materials for
biological samples. Hydroxyapatite, Ca10(P04)6(OH)2,as a hexagonal column
Preparation, Testing, and Selectivity of Stationary Phase Materials 55

I I
15 30 min 15 30
Figure 3.23 Selectivity of phenyl and alkyl bonded stationary phase materials for protein
separation. Column A, TSK gel phenyl-5PW R P , 75 mm x 4.6 mm i.d.;
B, T S K gel T M S 250, 75 mm x 4.6 mm i.d.; eluent, 60 min linear gradient
elution from 5% of 0.05% trfluoroacetic acid in 5% aqueous acetonitrile to
80% of 0.05% trfluoroacetic acid in 80% aqueous acetonitrile; flow rate,
1 ml min-'; detection, UV 220 nm. Peaks: 1, ribonuclease; 2, insulin;
3, cytochrome c; 4, lysozyme; 5, transferrin; 6, bovine serum albumin; 7,
myoglobin; and 8, ovalbumin.
(Reproduced by permission from Tosoh data)

I 1

200 E
8
E

L ! O s
I I I 5
m
0 10 min 20
Figure 3.24 Separation of plasmid D N A on a titania column. Conditions: column, H C A
column A-7610 (titania), 10 cm x 7.6 mm i.d.; eluent, step-wise gradient o
potassium phosphate bufer concentration (pH 6.8);flowrate, 1.O ml min- I!,
detection, UV 260 nm. Peaks: 1, purasmid D N A (pBR 322) and 2,
cromorimal.
(Reproduced by permission from Polymer Laboratories data)
56 Chapter 3
crystal is formed as a porous stationary phase material. The large pore size, 700-
10 000 A, is suitable for purification of bio-polymers. Bovine serum albumin,
lactic dehydrogenase, ribonucleic acid, and immunoglobulin in blood serum can
be purified. Hydroxyapatite is also used as a support for affinity liquid
chromatography .
The pore size of porous titania can be up to 2000 A.
Titania is used for the
purification of proteins and as a support for bound enzymes. The purification
of P-lactoglobulin from cheese whey, of protease from pineapple, /.I-lactamase,
and amylase can be achieved with titania. The latter two purifications are
impossible on alumina. Titania is also used as a support in peptide synthesis.
The separation of plasmid DNA is shown in Figure 3.24.

7 References
1. G.B. Alexander (ed.), ‘Silica and I,, Doubleday & Company, 1967.
2. K.K. Unger (ed.), ‘Porous Silica’, Journal of Chromatography Library, Vol. 16,
Elsevier, Amsterdam, 1979.
3. M. Ohhira, F. Ohmura, and T. Hanai, J . Liq. Chromatogr.,l989, 12, 1065.
4. T. Hanai, M. Ohhira, and T. Tamura, LC-GC, 1988,6,922.
5. L. Boksanyi, 0. Liardon, and E. sz Kovats, Adv. Colloid Interface Sci., 1976, 6, 95.
6. T. Hanai, in ‘Encyclopedia of Analytical Science’, Academic Press, London, 1995,
pp. 2558-2567.
7. J. Yamaguchi, T. Hanai, and Hong Cai, J . Chromatogr., 1988,441, 183.
8. Asahipak data from Showa Denko.
9. K. Yasukawa, Y. Tamura, T. Uchida, Y. Yanagihara, and K. Noguchi, J.
Chromatogr., 1987,410, 129.
CHAPTER 4

Selection of the Eluent

The most important parameter for the control of liquid chromatography is the
composition of the eluent. Liquid chromatography is a powerful separation
method with unlimited possibilities of eluent selection. However, it is not easy to
choose a suitable eluent within a short time without a number of trial
experiments. The crucial factor is to control the solubility of the analytes in the
eluent. Increasing the solubility of analytes in the eluent decreases their
retention times. The selection of the components of an eluent is described
below, based on the properties of the analytes to be separated. The important
properties are hydrophobicity, dipole moment, hydrogen bonding, ionization,
and steric effects.

1 Reversed-phase Liquid Chromatography

The most popular system, with over 50% of separations, is reversed-phase
liquid chromatography on octadecyl-bonded silica gel. Analytes are retained on
the stationary phase materials based on their hydrophobicity. This means that
polar compounds are eluted faster than non-polar compounds. The higher the
hydrophobicity of the stationary phase surface, the longer will be the retention
times of the analytes. Bonding of donor-electron groups onto the stationary
phase surface results in a stronger retention of compounds containing dipoles.
In some cases, interactions between untreated silanol groups on the stationary
phase and hydroxy groups on the analyte increase the selectivity; however, polar
groups on the stationary phase are generally eliminated to obtain inert
stationary phase materials, which are more stable in long-term operation.

Selectivity of Organic Modifiers in the Eluent

Three organic solvents, acetonitrile, methanol, and THF, are usually used as the
organic modifiers. Increasing the concentration of the organic modifier
decreases the overall retention times, but changes in relative retention times
depend on the properties of the analytes.

57
58 Chapter 4
The general relationship between the type of solute and its retention can be
seen by comparing the retention factors, k,of a set of standard compounds with
their octanol-water partition coefficients, i.e. the log P value (listed in Table
4.1), as a measure of their relative solubility in water. The logarithm of the
retention factor, logk, of these compounds measured in 50% aqueous aceto-
nitrile on an octadecyl-bonded silica gel column shows a close linear relation-
ship (Figure 4.1).
However, the elution behaviour of aromatic and aliphatic compounds is often
different (even though their carbon numbers and van der Waals volumes are
very similar) in eluents containing different organic modifiers. These variations
are due to differences in the solubility of analytes in the organic solvent.
Five organic solvents [acet onit rile, methanol , tetrah ydro furan (THF),
acetone, and dimethylformamide], which are homogeneously miscible with
water, have been used as modifiers to study the relationship of the selectivity of
the solvent to the molecular properties of analytes. The polar interaction

Table 4.1 Log P values of standard analytes

~

No. Chemical Log P No. Chemical Log P

1 Pentanol 1.33 11 Benzene 2.28

2 Hexanol 1.86 12 Naphthalene 3.21
3 Heptanol 2.39 13 Fluorene 3.91
4 Octanol 2.92 14 An thracene 4.38
5 Decanol 3.98 15 Pyrene 5.03
6 Dodecanol 5.04 16 Chrysene 5.79
7 Methyl benzoate 2.15 17 Toluene 2.59
8 Isopropyl benzoate 3.09 18 Ethy1benzene 3.12
9 Butyl benzoate 3.74 19 Isopropyl benzene 3.52
10 Isopentyl benzoate 4.14

o!
1
’ ,
2
I

3
I

4
I

log P
Figure 4.1 Retention factors related to log P values. Column, 5pm octadecyl-bonded silica
gel (LiChrosorb LC7) 25 cm x 4.1 mm i.d.; eluent, 50% aqueous aceto-
nitrile; j l o w rate, 1 ml min-’. Compounds: 0,alkanols; 0 , benzoates;
0, polycyclic aromatic hydrocarbons; and A,alkylbenzenes.
Selection of the Eluent 59

Table 4.2 Interaction properties of solvents for reversed-

phase liquid chromatography

Solvent Po Xe Xd Xn

Tetrahydrofuran 4.0 1.52 0.80 1.68

Acetone 5.1 1.78 1.17 2.14
Methanol 5.1 2.45 1.12 1.58
Acetonitrile 5.8 1.80 1.57 2.44
Dimethylformamide 6.4 2.50 1.34 2.56
Water 10.2 3.77 3.77 2.55

Po: polarity, Xe: proton acceptor, Xd: proton donor, Xn; dipole
moment.

properties of the solvents were calculated from reference values,' and are listed
in Table 4.2. The values of the polarity (Po)and the polar interaction properties,
proton acceptor ( X e ) , proton donor ( X d ) , and dipole moment (Xn), of the
mixed solvents can be calculated from their concentrations.
As above, the logk values of the different types of compounds [alkanols
(ROH), polycyclic aromatic hydrocarbons (PAH), alkyl benzenes (RB), and
alkyl benzoates (ROB)] each demonstrated a linear relationship with their log P
values and this relationship was observed with different ratios of acetonitrile
and water. The experiment was then performed in eluents containing different
organic modifiers and it was found that the behaviour of analytes containing
different functional groups differed and the effect depended on the l o g P of the
analy t e.
To demonstrate these differences, the experimental relationship between log P
and log k across a range of eluent compositions was determined for each group
of analytes. The results were then used to calculate the predicted logk in each
case for a theoretical model compound with log P = 5 as a hydrophobic model,
for l o g P = 1 as a hydrophilic model, and for logP = 3 as an intermediate
model.
The results in a range of aqueous acetonitrile eluents are shown in Figure 4.2.
For both alkanols and polycyclic aromatic hydrocarbons the predicted values
for each of the model compounds fall on a single line represented by the solid
points ( 0 ) (as expected from Figure 4.1). However, with T H F as the modifier,
predicted values for the model alkanols (ROH) (+) and polycyclic aromatic
hydrocarbons (PAH) (0) differ and the relative positions change with the log P
values. The log k values of log P = 5 alcohols and PAHs in the aqueous T H F
eluent were relatively smaller than those obtained for the alkanols and PAHs in
aqueous 30-70% acetonitrile, but the logk values of the l o g P = 1 model
compounds were higher. Furthermore, in aqueous T H F the logk values of
polycyclic hydrocarbons were larger than those of alkanols. These results
indicate that when a separation between aromatic and polar aliphatic com-
pounds is necessary, T H F may be an effective organic modifier.
60 Chapter 4
logPr5 log P = 3 log P = 1

2.0

x
-8 1.0

0 0

0 50 10WO 50 10010 50 100

concentration of organic modifier / water
Figure 4.2 Predicted chromatographic behaviour in aqueous acetonitrile and tetrahydro-
furan of model compounds whose log P values are 1, 3, and 5. Column, 5 pm
octadecyl-bonded silica gel (LiChrosorb LC-7), 25 cm x 4.1 mm i d . Com-
pounds and eluent: a, polycyclic aromatic hydrocarbons and alkanols in
aaueous
-. -.- - - acetonitrile:
..-- -.
. . .... .. aaueous
" alkanols in __
-, 0 -,-- - - __ - -tPtrahvdrnfuran ' --- n
- - . __ . , -~-_.-. -, and nnlv-
v )r - - j

cyclic aromatic hydrocarbons in aqueous tetrahydrofuran.

Comparison of the Efect of Organic ModiJiers on Analytes with

logP = 3
The selectivity of a number of organic modifiers was examined using the
predicted log k values of the log P = 3 models from each group in different
organic modifier-water mixtures. The composition of the eluent was adjusted so
that either the solubility parameter,' polarity (Po),proton acceptor ( X a ) ,proton
donor ( X d ) , or dipole moment ( X n ) values were kept constant to determine
which parameter affected the selectivity. The results are summarized in Table
4.3.
The compositions of the different eluents were determined using the values in
Table 4.2. The composition was adjusted in each case so that the selected
polarity term was equal in each eluent. For example, in Table 4.3 the polarity
(Po) of the 40 :60 THF-H20 is (0.4 x 4) + (0.6 x 10.2) = 7.72, when X e , 2.87;
X d , 2.58; Xn, 2.20. This has an equivalent polarity to 56:44 ACN-H20
[(0.56 x 5.8) + (0.44 x 10.2) = 7.72 but now X e , 2.66; Xd, 2.53; X n , 2.491.

Chromatographic Behaviour of log P = 3 Compounds in Aqueous Tetrahydro-

furan. For analytes whose log P = 3, in 40% aqueous T H F the predicted log k
values of the model alkanols (ROH), polycyclic hydrocarbons (PAH), alkyl-
benzenes (RB), and alkyl benzoates (ROB) are 0.64, 0.77, 0.90, and 0.79,
respectively. The polarity of 40% aqueous T H F is 7.72. The same polarity can
be obtained in 56% aqueous acetonitrile. As noted earlier, in aqueous aceto-
Selection of the Eluent 61
Table 4.3 Solubility problems of eluents

Predicted log k of log P = 3

compound

Eluent Po Xe Xd Xn ROH PAH RB ROB

40% THF-H20 7.72 2.87 2.58 2.20 0.64 0.77 0.90 0.79
56% ACN-H20 7.72 2.66 2.53 2.49 0.40
46% ACN-H20 8.19 2.87 2.76 2.50 0.63
54% ACN-H20 7.82 2.70 2.58 2.49 0.46
nc nc nc nc 2.20 nc
YOof ACN-H20 to give the same log k as 40% THF 46 40 36 43
60% MeOH-H20 7.14 2.98 2.18 1.97 0.81 0.63 0.67 0.62
69% ACN-H20 7.14 2.40 2.24 2.47 0.18
40% ACN-H20 8.44 2.98 2.89 2.50 0.78
72% ACN-HO 7.02 2.35 2.18 2.47 0.14
nc nc nc nc 1.97 nc
YOof ACN-H20 to give the same log k as 60% MeOH 39 46 44 46
50% ATN-HzO 7.65 2.78 2.47 2.35 0.35 0.33 0.41 0.31
58% ACN-H20 7.65 2.63 2.50 2.49 0.37
50% ACN-H20 7.99 2.78 2.66 2.49 0.53
59% ACN-H2O 7.60 2.61 2.47 2.48 0.36
"/o of ACN-H20 to give the same log k as 50% ATN 59 61 56 62
50% DMF-H20 8.30 3.13 2.56 2.55 1.26 0.69 0.87 0.89
43% ACN-H20 8.30 2.92 2.82 2.50 0.70
33% ACN-H20 8.77 3.13 3.06 2.51 1.02
55% ACN-H20 7.78 2.69 2.56 2.49 0.43
0% ACN-H20 10.2 3.77 3.77 2.55 nc
YOof ACN-H20 to give the same log k as 50% DMF 26 44 37 36

Po: Polarity, Xe: proton acceptor, Xd: proton donor, Xn: dipole moment, THF: tetrahydrofuran,
ACN: acetonitrile, MeOH: methanol, ATN: acetone, DMF: dimethylformamide, nc: cannot be
calculated.

nitrile the log k values for all compounds with log P = 3 are the same and thus
the predicted value for the ROH, PAH, RB, and ROB models is 0.40, unlike any
of the T H F retentions. These results indicate that chromatographic behaviour
cannot be controlled only by polarity. The proton acceptor value (Xe) of the
40% aqueous T H F eluent was 2.87. The same proton acceptor value can be
obtained in 46% aqueous acetonitrile (Table 4.3), where the predicted logk
values of each of the log P = 3 model compounds is 0.63. Thus in this case the
value of the logk of ROH is nearly equal to the value in 40% aqueous T H F
(0.64).
If the proton donor value (Xd)is fixed at 2.58, the equivalent eluent is 54%
aqueous acetonitrile and the predicted log k value for the log P = 3 compounds
should be 0.46, which is again different from any of the logk values in 40%
aqueous THF. The dipole moment ( X n ) of 40% aqueous THF was 2.20;
however, such a value cannot be obtained with any proportion of aqueous
acetonitrile. To obtain the same logk values as in 40% aqueous THF, a
62 Chapter 4
composition of 46% aqueous acetonitrile is needed for ROH, 40% aqueous
acetonitrile for PAH, 36% aqueous acetonitrile for RB and 43% aqueous
acetonitrile for ROB. This comparison demonstrates the different selectivity of
THF.

Chromatographic Behaviour of log P = 3 Compounds in Aqueous Methanol. In a

similar calculation, the logk values of ROH, PAH, RB, and ROB in 60%
aqueous methanol are 0.81, 0.63, 0.67, and 0.62, respectively (Table 4.3). The
same log k values can be obtained in 39% aqueous acetonitrile for ROH, in 46%
for PAH, 44% for RB and in 46% for ROB. The values of Po, Xe, and Xd (but
not Xn) in 60% aqueous methanol can be obtained in 69% for Po, in 40% for
Xe, and in 72% aqueous acetonitrile for Xd. Only the logk of ROH in 40%
aqueous acetonitrile demonstrated a nearly equal log k value to that obtained in
60% aqueous methanol. Thus the solvent properties do not have a specific
relationship with the log k values of these compounds.

Chromatographic Behaviour of log P = 3 Compounds in Aqueous Acetone. When

the Xe value in aqueous methanol was nearly equal to that in aqueous
acetonitrile (see above), the logk of ROH was about the same as that in
aqueous acetonitrile but the corresponding predicted value with aqueous
acetone (log k = 0.35) is very different to the equivalent aqueous acetonitrile
(logk = 0.53) (Table 4.3). However, when the Po and X d values in aqueous
acetonitrile are nearly equal to those in aqueous acetone, the predicted logk
values of OH are about the same. To obtain the same logk as in 50% aqueous
acetone it was necessary to use 59% aqueous acetonitrile for ROH, 61 YOfor
PAH, 56% for RB and 62% for ROB.

Chromatographic Behaviour of log P = 3 Compounds in Aqueous Dimethylform-

amide (DMF). DMF is a solvent with a strong dipole moment (Table 4.2). The
same logk values in 50% aqueous DMF can be obtained at a concentration of
26% aqueous acetonitrile for ROH, 44% aqueous acetonitrile for PAH, 37%
for RB, and 36% aqueous acetonitrile for ROB. When eluents with equivalent
Po, Xe, Xd, or X n values in 50% DMF were selected, different log k values were
predicted in all three cases.
The above results only demonstrate the chromatographic behaviour of
compounds with log P = 3. Their relative retention factors thus are influenced
by the properties of solvents in the eluent. The relative values cannot be
controlled by only one property of the solvents and chromatographic behaviour
therefore also depends on the properties of the analytes.

Comparison of Effects of Mod$ers on a Mixture ofAnalytes

The following results demonstrate the comparative chromatographic behaviour
of the 19 compounds listed in Table 4.2. In reversed-phase liquid chromato-
graphy, changing the concentration of organic modifier affects both the overall
retention time and the relative elution behaviour of different types of com-
Selection of the Eluent 63

pounds. The effect of the organic modifier especially influences the retention of
polycyclic aromatic hydrocarbons, as seen in Figure 4.3A (compounds 11-16).
The results in 50% aqueous acetone and aqueous acetonitrile indicate that the
retention times in 50% aqueous acetone are shorter than those in 50% aqueous
acetonitrile. A selectivity difference for alkanols ( 0 , 1-6) is not observed but the
retention time of the larger PAHs is shorter than the values expected from their
size.
From the results in 40% aqueous T H F and 50% aqueous acetonitrile (Figure
4.3B), the relative retention of PAHs (1 1-16) is weak and that of alkylbenzenes
(0, 17-19) is strong in aqueous THF. The alkyl group affects the strong
retention in aqueous THF. As found earlier (Table 4.3), a methanol-water

A B

"-
*
/ o 11
12 9 13
-7- 6

0
2

0.5 1.0 1.5

L0 0.5 1.0 1.5
log k log k
50% acetonitrildwater 50% acetonitrilehater

C D
c
1.5

1
I I I
0
0 0.5 1.0 1.5 0 0.5 1.0 1.5
log k log k
50% acetonitrildwater 50% acetonitrildwater
Figure 4.3 Relative selectivity of diflerent organic mod$ers. Numbers beside symbols,
1-6 ( 0 ) ROH; 7-10 ( 0 )ROB; 11-16 (filled hexagons) PAH; and 17-19 (0)
RB.
64 Chapter 4

mixture (Figure 4.3C) can distinguish alkanols (e, ROH) from the other analyte
groups.
Comparison of the retention behaviour in 50% aqueous D M F and acetoni-
trile (Figure 4.3D) indicated that DMF is a selective modifier for the separation
of alkanols ( 0 , 1-6) and PAHs (11-16) and of aromatic and aliphatic
compounds. Unfortunately, DMF is too aggressive as a solvent and attacks
organic polymer piston seals in the pump and, therefore, highly concentrated
solutions cannot be used as an eluent.
From the theoretical viewpoint, acetonitrile is the most suitable solvent to
study the correlation of retention times and log P values of analytes, since the
dipole moment (2.44) is nearly equal to that of water (2.55) (Figure 4.4). The
electron donor effect can therefore be eliminated, and the elution order is not
changed on modification of the acetonitrile-water mixture ratio. The first choice
of an eluent should therefore be an acetonitrile-water mixture for non-ionic
compounds in reversed-phase liquid chromatography. Methanol, acetone,
THF, or DMF can then be added to improve the resolution.

Example I : Separation of an alkylbenzene and an alkyl benzoate. The logP

values of butyl benzoate and isopropylbenzene are 3.74 and 3.52, respectively,
from Table 4.1. The separation of these compounds in 50% aqueous acetonitrile
therefore requires a high plate number column as the log k values will be nearly
identical. The most suitable eluent is 40% aqueous THF. See compounds 9 and
19 in Figure 4.3B.

Example 2: Separation of a polycyclic aromatic hydrocarbon and an alkyl

benzoate. The log P values of fluorene and butyl benzoate are 3.91 and 3.74,
respectively, from Table 4.1. The separation is poor in 50% aqueous aceto-

10

0 50 % 100

conc. Water / organic solvent

Figure 4.4 Polarity, proton acceptor and dipole properties of water and organic modifier
mixtures. -, acetonitrile; - - - , methanol; and . tetrahydrofuran.
*
Selection of the Eluent 65
nitrile; however, they can be separated in 60% aqueous methanol or 50%
aqueous DMF (compounds 13 and 9, Figures 4.3C and 4.3D).
However, in selecting a mobile phase the entire analytical system and
separation system should be considered. In many cases, a component of a
good eluent may not be suitable for research purposes. For example, the use of
water as an eluent in methods that will require further chemical reactions and
purification should be avoided due to the difficulty in concentrating the
products of interest after the separation and the stability of the products in the
process. In addition, organic solvents are flammable, combustible, and toxic.
The threshold limit concentrations of acetone, acetonitrile, DMF, methanol,
and T H F are 250, 20, 10, 200, and 200ppm, respectively. The most easily
handled solvent of these five is methanol. Acetone is the least used solvent due to
the strong UV absorption (UV cut-off = 330 nm). T H F is unstable and easily
oxidized by oxygen in air, causing a cut-off at longer wavelengths as it ages.
At present, the selection of an organic modifier is estimated from the aliphatic
or aromatic nature of analytes. However, the properties of analytes often
cannot be easily obtained. Examples of quantitative structure-retention rela-
tionships based on the lo g P and van der Waals volume of analytes are
demonstrated in Chapter 6.

2 Chromatography of Ionic Compounds

Neutral compounds are retained by hydrophobicity in reversed-phase liquid
chromatography, and by hydrogen bonding and/or n-n interactions on hydro-
philic phases in normal-phase liquid chromatography. Ionization of the analyte
reduces the retention time in reversed-phase liquid chromatography, but in
some cases increases the retention time on ion-exchangers. Retentions on ion-
exchange resins made from polystyrene gel are based both on ion-ion inter-
actions with ion-exchange groups and on hydrophobic interactions with the
polystyrene gel matrix. The addition of counter-ions can improve the resolu-
tion.

pH Effects
The retention factor of partially ionized compounds can be predicted in
reversed-phase liquid chromatography by Equation (4. 1):2

where ko is the maximum retention factor of the unionized form of the analyte.
The term ki is the retention factor of the fully ionized compound; K, is the
dissociation ionization constant; and [H '1 is the hydrogen ion concentration of
the eluent. Every compound has its own ko, ki, and K, values. When the
hydrophobicities of compounds are nearly equal, separation is difficult in the
66 Chapter 4
reversed-phase mode. However, when their dissociation constants are different,
the separation can be easily accomplished in a pH-controlled eluent by
introducing differential partial ionization. The value of ki is close to zero in a
reversed-phase system; however, the value is not negligible in the ion-exchange
mode.

Example: separation of 3-chloro and 4-ethylbenzoic acids. The log P values of

these compounds are similar, and their pKa values are 3.96 and 4.35, respec-
tively. As seen by their chromatographic behaviour in Figure 4.5, the separation
is difficult at low pH due to the similar log P values giving similar log ko values.
The separation can be easily performed in an eluent with a pH of about 4.2, due
to the difference in PKa values and different degrees of ionization.
The retention factor of an amino acid is given by the following equation, as
the amino acid has two ionizable groups (a carboxyl group and an amino group,
with Kal and Ka2,respectively):

The maximum retention factor (ko) is related to the log P value and kl and k- 1
are the retention factors of the cationic and anionic forms, respectively. The pKa
values are known, and the retention factor in a given eluent can therefore be
predicted in reversed-phase liquid chromatography using an alkyl-bonded silica
gel or polystyrene gel column. The separation conditions can be adjusted
according to their logP and pKa values by the selection of a suitable organic
modifier concentration and the pH of the eluent.374

I I I
2 4 6
PH
Figure 4.5 Eflect of p H on the retention factors of 3-chloro- and 4-ethylbenzoic acids.
Column, octadecyl-bonded vinyl alcohol copolymer gel, 10 cm x 6 mm i d . ;
eluent, 0.05 M sodium phosphate solution in 20% acetonitrile; flow rate:
1 ml min- '; temperature, 30 "C. Compounds: 3C1, 3-chiorobenzoic acid; and
4Et, 4-ethylbenzoic acid.
Selection of the Eluent 67
The separation of compounds whose pK, values are different can be carried
out easily on an ion-exchanger using ion-ion interaction^.^ For example, non-
ionized mandelic acid is eluted before benzoic acid due to weak hydrophobicity.
But mandelic acid is retained more than benzoic acid on an ion-exchanger due
to its stronger acidity. With polystyrene gel-based ion-exchangers both ion-ion
and hydrophobic interactions are important factors in the retention mechanism.
Compounds whose pK, values are similar can be separated on an ion-exchange
resin. For example, amino acids have been separated by cation-exchange liquid
chromatography (Figure 4.6) and carboxylic acids by anion-exchange liquid
chromatography (Figure 4.7). The elution order of these separations was
directly related to their dissociation constants. The high temperature of the
separation was also important, as were the acidic and ionic strength of the
eluen ts.
The separation of very hydrophobic compounds, such as proteins, is difficult
on an ion-exchange resin with a hydrophobic matrix, because organic modifiers
are required as part of the eluent. However, the separation can be performed on
an ion-exchanger with a hydrophilic matrix. A weak acidic ion-exchanger is
suitable for the separation of basic proteins, and a weak basic ion-exchanger is
suitable for the separation of acidic proteins (Figures 4.8 and 4.9, respectively).
The components of the eluents should not be harsh, to prevent the denaturation
of the proteins. A salting-out effect can be used to obtain a gradient elution. A
combination of buffer solutions and weak ion-exchangers on a polar matrix,
such as poly(viny1 alcohol), is suitable for the separation of bio-polymers,
especially for preparative-scale separations. Oligonucleotides from polyadenyl
acid and RNA of Escherichia coli have been separated on weak anion-
exchangers using ammonium acetate buffer and a sodium chloride concent ra-

1 2 12 13 1718202123 25

6
I I I I I
0 20 40 min 60 80

Figure 4.6 Separation of amino acids in body fluid. Column, M C I G E L CKIOV,

15 cm x 4.6 mm i.d.; eluent, four-step gradient of sodium citrate bufler;flow
rate, 0.5 ml min- '; temperature, 60 "C; detection, fluorescence reaction
detector of sodium hydrochloritelorthophthaldehyde, Ex. 360 nm, Em.
450 nm. Peaks: 1, S S A ; 2, T A U ; 3, GCGA; 4,HPro; 5 , Asp; 6 , Gln + Asp;
7 , Thr; 8, Ser; 9, Sar; 10, Glu; 11, Pro; 12, Gly; 13, Ala; 14, Cys; 15, Val;
16, Met; 17, Ile; 18, Leu; 19, Tyr; 20, Phe; 21, His; 22, Orn; 23, Lys; 24, Trp;
and 25, Arg.
(Reproduced by permission from Mitsubishikasei data)
68 Chapter 4
2

I I I I
0 10 20 min 30
Figure 4.7 Anion exchange separation of carboxylic acids in red wine. Column, Shodex
C811, 100 cm x 7.6 mm id.; eluent, 3 mM perchloric acid; flow rate,
';
0.9 ml min- temperature, 60 "C; detection, reaction detection using chloro-
phenol red at 430 nm. Peaks: 1, citric acid; 2, tartaric acid; 3, malic acid; 4,
succinic acid; 5, lactic acid; 6, formic acid; and 7 , acetic acid.
(Reproduced by permission from Shodex data)

Figure 4.8
lj
I
I
0
I
L

I
20 rnin 40
Cation-exchange liquid chromatography of basic proteins. Column, Asahipak
ES502C; eluent, 20 min linear gradient of sodium chloride f r o m 0 to 500 mM
in 50 mM sodium phosphate bufer pH 7.0;JEow rate, 1 ml min-'; tempera-
cure, 30 "C; detection, UV 280 nm. Peaks: l ? myoglohin f r o m horse skeletal
muscle ( M , 17 500, PI 6.8-7.3); 2, ribonuclease f r o m bovine pancreas ( M ,
13700, PI 9.5-9.6); 3, a-chymotrypsinogen A f r o m bovine pancreas ( M ,
257 000, PI 9.5); and 4,lysozyme from egg white (Mr 14 300, PI 11.O-11.4).
(Reproduced by permission from Asahikasei data)
Selection of the Eluent 69

I I I
0 20 min 40
Figure 4.9 Anion-exchange liquid chromatography of acidic proteins. Column, Asahipak
ES502N; eluent, 20 min linear gradient of sodium chloridefrom 0 to 500 PM in
50 mM bis-trisfHC1 bufer pH 7.0; flow rate, 1 ml min-'; detection, UV
280 nm. Peaks: 1, conalbumin ( M , 7700&88000, PI 6.0-6.8); 2, ovalbumin
( M , 45000, PI 4.6); 3, trypsin inhibitor ( M , 8000, PI 4.5); and 4, p-
lactoglobulin (Mr 18 400, pl 5.1).
(Reproduced by permission from Asahikasei data)

tion gradient.697Subfragments of myosin have been separated on a weak cation

exchanger using a sodium chloride concentration gradient.* The selection of
eluent components such as pH and salt concentration makes possible the direct
analysis of catecholamines in urine without deproteinization and of haemo-
globin in red blood by cation-exchange liquid chromatography, as shown in
Figures 4.10 and 4.1 1, respectively.

0
c
min
I

50

Figure 4.10 Direct analysis of catecholamines in urine sample. Column, Asahipak ES-
502C; eluent, 75 mM succinic acid + 25 mM borate bufer (pH 6.10) contain-
ing 0.5 r n M EDTA; flow rate, 1.0 min-'; detection, fluorescence reaction
detection Ex. 350 nm. Peaks: 1, adrenaline; 2, noradrenaline; and 3, dop-
amine.
(Reproduced by permission from Asahikasei data)
70 Chapter 4

L
A B

I
0 min
i
I0
,i; min
I

30
Figure 4.11 Direct injection analysis of haemoglobin in red blood cells. Column, Asahipak
ES-502C; eluent, 32 min linear gradient from 25% 30 mM sodium phosphate
bufler to 65% 30 mM sodium phosphate containing 300 m M sodium chloride
pH 5.5; $ow rate, 1.0 ml min-’; detection, 425 nm. Samples: A, normal
subject and B, diabetic patient.

Ion-pair Separations
Ionic compounds are usually separated by ion-exchange liquid chromatogra-
phy, and non-ionic compounds are separated by normal-phase or reversed-
phase liquid chromatography. The separations of mixtures of ionic and non-
ionic compounds and of different ionic and polar compounds are more difficult.
One method for extracting such compounds from aqueous solution is ion-pair
extraction where the ionic charge on the analyte is neutralized by the addition of
a counter-ion of opposite charge, and the analyte is extracted into an organic
phase as the neutral ion-pair. This ion-pair extraction method was first applied
as ion-pair partition liquid chromatography by SchilI et a1.’
The selection of the counter-ion and its concentration are important for the
separation of ionic compounds in reversed-phase and ion-exchange liquid
chromatography. The addition of hydrophobic ions is an especially powerful
method and several surfactants can be used as hydrophobic counter-ions. The
theoretical column efficiency of ion-pair liquid chromatography is much better
than that of an ion-exchange column, and the regeneration of a column is much
faster. Thus, if we can control ion-pair liquid chromatography, we can solve a
separation problem. (The important background sources in this area are listed
at the end of the chapter.)

Theory of Ion-pair Liquid Chromatography

An ionized solute A - forms a paired ion AB with a counter-ion B + in water.
The paired ion is extracted into an organic phase from the aqueous phase, as
given by the equation:
Selection of the Eluent 71

where aq is the aqueous phase and org is the organic phase. When applying this
ion-pair extraction, the following parameters must be considered.

a. Selection of stationary phase material.

b. Selection of counter-ion and the concentration.
c. pH and ionic strength of eluent.
d. Selection of organic modifier and the concentration.
e. Selection of column temperature.
f. Selection of detector.

Two types of system are used for ion-pair liquid chromatography. When
polar stationary phase materials, such as silica gel, are used an ion-pair partition
mechanism is applied. When non-polar stationary phase materials, such as
octadecyl-bonded silica gel and polystyrene gel, are employed a paired-ion
adsorption mechanism is involved. The former is called normal-phase ion-pair
partition liquid chromatography, and the latter is called reversed-phase ion-pair
liquid chromatography.
The following conditions shou€dbe satisfied to perform satisfactory ion-pair
liquid chromatography.

a. A one-to-one ion-pair must be formed.

b. The equilibrium concentration of the ion-pair in the aqueous phase must
be very low compared with that of the ion.
c. Only the paired-ion is extracted into the organic phase.
d. The counter-ion makes an ion-pair only with the target ion.
e. The paired-ion is not decomposed in the eluent.
f. The partition equilibrium is related only to the chromatographic reten-
tion.

Normal-phase Ion-pair Partition Liquid Chromatography

Analytes in the organic phase (eluent) form ion-pairs with counter-ions existing
in the hydrophilic stationary phase. The ion-pairs are then partitioned into the
organic phase. Usually, porous silica gel is used as the support, and water or
formamide is the stationary phase. The requirements are that a specific amount
of counter-ions must be contained in the stationary phase, and the pH must be
controlled to ionize the analytes. The eluent is not miscible with water, and
counter-ions must stay in the stationary phase.
The conditions were employed for the separation of sulfa drugs'' using
tetrabutylammonium hydrogen sulfate (TBAHS04).The pH of this counter-ion
is greater than 6 in water; it exists as tetrabutylammonium cation (TBAf) and
sulfate anion (SO$-). n-Pentane was used as the major component of the eluent
with a small amount of n-butanol, which enhanced the solubility of the paired-
ion. A small amount of water was added to the above mixture, which was then
left standing until equilibrium was reached. A proportion of the paired-ion
moved into the organic phase and a small amount of butanol and n-pentane
 72 Chapter 4
moved into the aqueous phase. In this equilibrated system, the p H in the
stationary phase is constant before and after equilibration. In the aqueous phase
the amount of n-butanol is ca. 4 5 % (w/w) and that of n-pentane is negligible.
The amount of water in the organic phase depends on the concentration of n-
butanol. In 25% n-butanol-n-pentane (v/v), the water concentration in the
organic phase is 1.5% (w/v), and 5.0% in the case of 50% (v/v) n-butanol-n-
pentane. These values meet the above requirements; however, it is not simple in
practice to obtain the appropriate experimental conditions.

Re versed-phase Ion-pa ir Liquid Chromat ograp hy

In this mode analytes form hydrophobic ion-pairs in the eluent, and the paired-
ions are retained on a hydrophobic stationary phase surface by their hydro-
phobicity. The hydrophobic paired ions are then eluted by reversed-phase liquid
chromatography. Gradient methods, such as organic modifier gradients,
counter-ion gradients, ionic-strength gradients, and pH gradients, can be
applied in this system. Usually less than 0.1 wt% of the counter-ion and buffer
concentration is employed. The selection of a stationary phase material is simple
and a variety of hydrophobic stationary phase materials for reversed-phase
liquid chromatography can be used. The organic modifier concentration
depends on the hydrophobicity of stationary phase materials used and the ion-
pair reagent. Increasing the size of the counter-ion increases the retention time.
The maximum retention time is reached when the counter-ion concentration
reaches the micelle condition. The relationship between counter-ions and
analytes is summarized in Table 4.4.

Example: ion-pair liquid chromatography of amino acids. Amino acids are

zwitterions. The amino group can form an ion-pair with an alkanesulfonate
ion (such as octanesulfonate), and the carboxyl group can form an ion-pair with
a tetrabutylammonium ion, depending on the pH of the solution.

' high pH

amino acids:
BYN+ -02C-CH2-NH2
systems

In reversed-phase liquid chromatography, the ionization of the solute

decreases the retention. The addition of counter-ion under these conditions
forces the formation of an ion-pair between the ionized solute and counter-ion,
and then the retention of the analyte increases as the paired-ion is retained.
Aspartic and glutamic acids were not retained in a buffer solution without a
counter-ion; however, these acids were retained by the addition of octyl sulfate
to the eluent, as seen in Figure 4.12. These amino acids can form a complex with
copper ions and will be retained on the stationary phase. The addition of both a
counter-ion and copper ions further increased the retention and made possible
Selection of the Eluent 73
Table 4.4 Relationship between analytes and counter-ions in ion-pair
liquid chromatography

Analytes Counter-ions

Strong and weak acids Quaternary amines

sulfate dyes tetrabutyl ammonium
carboxylic acids tetramethyl ammonium
sulfa drugs hexadecyl trimethylammonium
ammonia
Tertiary amines
trioctylamine
Strong and weak bases Alkyl-, allyl-sulfonic acids
catecholamines pentane, hexane, heptane, octane, decane,
dodecane sulfonic acids
Phosphoric and sulfonic acids
Alkyl sulfate, dodecyl sulfatea
Basic compounds Perchloric acid

a Use as alkyl sulfonic acids, the selectivity is slightly different.

5.0 r rn GIU
.-C Asp
E

oc c u occcu
Counter ion
Figure 4.12 Efect of counter-ions and copper on the retention of amino acids. Column,
octadecyl-bonded silica gel, 25 cm x 4.6 mm i.d.; eluent, 0.01 M sodium
acetate bufer (pH 5.6) containing 1.2 mM sodium octanesulfonate (Oc:)and/
or 0.1 mM copper acetate (Cu); flow rate, 1 ml min-'; detection, UV
220 nm. Compounds: Glu, glutamic acid; Asp, aspartic acid.

the separation of these compounds. Increasing the alkyl chain length of the
counter-ion also enhanced the retention, as seen in Figure 4.13.
Increasing the concentration of the counter-ion further increased the reten-
tion, but the retention reached a plateau, as seen in Figure 4.14. A total
separation of the amino acids by reversed-phase ion-pair liquid chromatogra-
phy could be performed." A column switching technique reduced the total
analysis time (Figure 4.15).
74 Chapter 4

.-C Ser
E
GlY

0-
C6 C7 C8
Figure4.13 Eflect of alkyl chain length of counter-ion on separation of amino acids.
Column, octadecyl-bonded silica gel, 15 cm x 4.6 mm i.d.; eluent, 0.01 M
sodium acetate bufler containing 0.4 M copper acetate and 1.0 mM sodium
alkanesulfonate (pH 5.6); flow rate, 1 ml min-'; detection, UV 230 nm.
Counter-ion: C6, sodium hexanesulfonate; C7, sodium heptanesulfonate; and
C8, sodium octanesulfonate. Compounds: Ser, serine; Gly, glycine; Glu,
glutamic acid; and Asp, aspartic acid.

I I 1
1.6 3.2 6.4 mM 12.8
counter ion concentration
Figure 4.14 Efect of counter-ion concentration on separation of amino acids. Column,
octadecyl-bonded silica gel, 15 cm x 4.6 mm i.d.; eluent, 0.01 M sodium
acetate bufler (pH 5.6) containing 0.4 mM copper acetate and 1.6-12.8 mM
sodium heptanesulfonate;Jlow rate, detection, and analytes, see Figure 4.13.

Two mechanisms for retention in reversed-phase ion-pair liquid chromato-

graphy have been considered. One is the adsorption of the hydrophobic paired
ion on the hydrophobic surface of stationary phase material. In the second
mechanism, the hydrophobic counter-ion is held on the surface of the hydro-
phobic stationary phase, and the analyte ion is retained by ion-ion interactions,
as shown in Figure 4.16. In the latter case, of a dynamic ion-exchange
Selection of the Eluent 75

I. 1 I 4.
0 10 20 30 min
Figure 4.15 Ion-pair liquid chromatography of free amino acids using a column switching
system. Column I , butyl-bonded silica gel, 50 x 4.6 mm i.d., 2, octyl-bonded
silica gel, 50 x 4.6 mm i.d., and 3, octadecyl-bonded silica gel,
250 x 4.6 mm i.d.; eluent, 0.01 M sodium acetate bufer (pH 5.6) containing
4 m M copper acetate and 0.8 m~ sodium heptanesuljionate; Jow rate,
1 ml min-'; detection, UV235 nm. Peaks: 1, Tyr; 2, Val; 3, Met; 4,His; 5,
Lys; 6, Ile, 7, Leu; 8, Phe; 9, Arg; 10, Asp; 11, Ser; 12, Glu; 13, Thr; 14, Gly;
15, Pro; 16, Cys; and 17, Ala. 1-9 were separated on column I and 10-1 7 were
separated by a combination of columns 2 and 3.

Figure 4.16 Model of reversed-phase ion-pair liquid chromatography in which counter-

ions are held on the stationary-phase surface.

mechanism, the column efficiency is much higher than in the corresponding ion-
exchange liquid chromatography using an ion-exchange resin as the stationary
phase. Examples of samples, counter-ions, and stationary phases are summar-
ized in Table 4.5.

p H Effects
The retention of acids in reversed-phase liquid chromatography can be
predicted by Equation 4.1. The same equation can be applied to predict the
Table 4.5 Applications of reversed-phase ion pair liquid chromatography
Sample Counter-ion Stationary phase Eluenl Ref-
material

Alkaloid TMA ODS silica 50 mM TMA in 20% MeOH, pH 6.1 24

Alkanesulfonic acid Cetylpyridium ODS silica 60% aq. MeOH 17
Tris(2,2’-dipyridyl)ruthenium(11)mono& dicarboxylic acids
Alkylsulfonic acid ODS silica 0.01 M Borate buffer pH 10.0 18
Amine Tetrabromophenolphthalein NH2 or OH silica Dichloromet hane/MeOH 45
Amino acid Naph t halene-2-sulfo- Phenyl silica 50 mM Phosphoric acid 14
dodecanesulfonic acid C8 silica 20% ACN + 50 mM H2S04,5 mM NaHS04 15
Dodecyl sulfate Me silica Na citrate + Na sulfate + PrOH + SSD 62
Aminoglycoside TFA C18 silica 0.2 M TFA 38
PFPa + TFA Polystyrene PFPa + TFA 39
Aminop henol Bis(2-ethylhexy1)phosphate Silicone/cellulose Citric buffer pH 3.8 12
Antidepressive amine Trimet hylnonylammonium C8 silica NaBr + trime thylnony 1 ammonium/Pentanol, 57
phosphate buffer
Apomorphine Dodecyl sulfate Ph silica 20 mM NalHP04 + 30 mM citric buffer 35
pH 3.25 + MeOH + ACN (55:36:9)
Anions TBA C18 silica TBA + salicylic acid/pH 4.62 473
Ascorbic acid Tridecylammonium C 18 silica Tridecylammonium, pH 5.0 56
Barbital TBA Bu tylnitrile/C2 0.03 M TBA pH 7.7 19
silica
Benzoates TBA Silanized silica PeOH -t- phosphate buffer 65
Benzoic acid derivatives TBA C2 silica 0.03 M TBA pH 7.4 19
Carboxylic acid 1-Phenetyl-2-pyrro- Phenyl silica Acetate buffer pH 4.6 14
liniumcetride C8 silica 35% ACN + 0.05 M ammonium acetate 15
Catecholamine Octyl sulfate C18 silica 50 mM Citric acid + 50 p~ EDTA in 50 mM Na2HP04 34 Q
(2 :1)
Cefaclor/cefradine TBA C 18 silica Phosphate buffer pH 7.0 + TBA + ACN 41,54 g5
.a,
Choline/3,5-dini tro-benzoic acid derivatives 3
Dodecyl sulfate C18 silica 5 mM Dodecyl sulfate, 0.1 % acetic acid in 50% ACN 23
Cocaine Ammonium C 18 silica 70% MeOH + 50 mM NH4C02H 16 s'
Cosmetic dyes NaHCI4 C8 silica NaC104/ACN 48 +
s
Daunorubicin Dodecyl sulfate C18 silica Dodecyl sulfate/phosphoric acid + ACN 50
Q
Diazonium salts Heptanesulfona te C18 silica Heptanesulfonate/phosphate buffer ACN 60
L-3,4-Dihydroxyp hen y lalanine metabolites trl
Heptanesulfonate C 18 silica
2
HSA + MeOH + acetic acid 66
Dipeptide Naphthalene-2-sulfonate Ph silica 50 mM Phosphoric acid 14 '
Dyes TBA C18 silica 5 mM TBA/45% MeOH 20
Enkephalins Octanesulfonate C 18 silica Octanesulfonate + citric acid/phosphate buffer, MeOH 59
Erythromycin Ammonium C 18 silica 70% MeOH + NH40H pH 4.8 16
Guanidino compounds Octanesulfonate C8 silica Octanesulfonate + ninhydrin + ACN + MeOH 37
Heroin TBA PrNH silica 5 r n TBA
~ + ACN (15:85) 33
Indoleacetic acid TBA Pentanol/C2 0.03 M TBA pH 7.4 19
silica
Iodine, Thiosulfate TBA C 18 silica TBA/MeOH + phosphate buffer 49
Neostigmine, pyridostigmin, edrohonium
Pentane sulfate C18 silica Pentanesulfonate + NaH2P04/ACN pH 3 21
Oligodeoxyribonucleotides TBA C 18 silica 2 mM TBAP + ACN/Phosphate buffer pH 7.0/gradient 36
Organic acids Iron(rr) 1,lO-phenanthroline Polystyrene Fe(phen)3S04/0.5 mM acetate buffer 46
Oxymorphone Octanesulfonate C8 silica 3 mM Phosphate buffer pH 4 + octanesulfonate/ACN 42
Penicillamine Heptanesulfona te C 18 silica 1 g Heptanesulfonate + 150 mg EDTA/1 1 water pH 4.0 32
Peptide Ammonia, TFA C18 silica 50 mM NH4COZH/SO mM TFA, pH 1.3 27
Phenylacetic acid TBA Silanized silica Phosphate buffer + PeOH 63
Phen ylenediamine Ammonium C 18 silica 30% MeOH + 20 mM N H 4 0 H pH 6.4 16
Pirlimycin TBAP C8 silica TBAP/ACN + perchloric acid 43
Prostaglandin E l TBA C l 8 silica TBA + MeOH + water + acetic acid 21
(0.15 :62 :35 :3)/gradient
(continued)

4
-4
Table 4.5 Continued
Sump Ie Counter-ion Stationary phase Eluent Ref.
material

Pyrimidine, Purine, Azapurine metabolite

TBA C 18 silica 10 mM TBA, pH 8.0 in 0.2% ACN 22
Pyrocarbine, Phyzostygmin Pentanesulfonate C18 silica 5 mM Heptanesulfonate, pH 3.6 in 4% MeOH 29
Quaternary amine 1-Phenetyl-2-pyrrolinium Phenyl silica Acetate buffer pH 4.6 14
Rebamisol Heptanesulfona te C18 silica 2% Heptanesulfonate in 0.2% acetic acid + MeOH 30
(55 + 4)
Riboxamide HTAB C 18 silica 1 mM HTAB in 10 mM phosphate buffer pH 6.0 28
Salbutamol Heptanesulfonate C8 silica 5 mM Heptanesulfonate in 5.8 mM phosphate buffer 31
pH 6.0 + MeOH (3 1: 69)
Sulfonamide TBA Butylni trile/C2 0.03 M TBA pH 7.9 19
silica
Sulfonic acids Cetyltrimethylammonium SAS silica PrOH/H20 or PrOH + CH2C12/H20 64
Sulfonates TBA Silanized silica PeOH + phosphate buffer 65
Tetracycline HC104 SC-TAS ACN 61
Tartrazine TBA C 18 silica TBA/formic acid + NaOH 59
Tetrahydrozoline Octanesulfonate C 18 silica Octanesulfonate + N,N-dimethyloctylamine/MeOH 51
Theophylline metabolites
TBA C 18 silica TBA, pH 4.75 in 50% MeOH, gradient 26
TBA C18 silica TBA + 15 mM Tris buffer + ACN + MeOH 40
Decylamine C 18 silica Decylamine/ACN 53
Tricyclic antidepressant Alkylammonium C8, C18 silica MeOH/phosphate buffer pH 2 44
Tris(2,2’-bipyridyl)Ru Heptanesulfonate C 18 silica 5 mM Heptanesulfonate buffer pH 3.5 18
N,N-Trime thy lene-bis Heptanesulfonate C18 silica Heptanesulfona te/ACN 58
(p y ridium-4-aldoxime)
dibromide
Selection of the Eluent 79
retention factor in reversed-phase ion-pair liquid chromatography. The pH and
counter-ion effects on the chromatographic behaviour of indoleacetic acid are
demonstrated in Figure 4.17.
The addition of a surfactant counter-ion reduced the retention factor at low
pH due to the surface modification of the stationary phase material. Covering
the surface of the stationary phase with the surfactant reduces the hydro-
phobicity of the stationary phase material. The addition of the tetrabutyl-
ammonium counter-ion increased the retention factor at high pH. The pK, of
the indole acetate was 5.15 without surfactant, 4.85 with octyl sulfate ion, and
5.60 with tetrabutylammonium ion. That is, the addition of a same-charged
hydrophobic ion reduced the PKa value, and the addition of the counter-ion
increased the pK, value. The difference in the PKa value on the addition of
surfactant is not constant; it is affected by the kind of ion and the concentration.
It is difficult to estimate the PKa change.
The retention factor of the molecular form of the analyte, k,, and the
retention factor of the paired-ion7kip, can be easily measured experimentally.
When the retention factor is measured at low and high pH, and the pK, value is
known, k at a given pH is roughly estimated from the following equation:

where Ka is the dissociation constant of the ion-pair complex.

15
r NaP

10

0 J
2 5 8 10
PH
Figure 4.17 Selectivity of counter-ion for the retention of indoleacetic acid. Column,
Hitachi 3011 (polystyrene gel), 25 cm x 4.1 mm i d . ; eluent, 50 mM sodium
phosphate ( N a p ) containing 25 mM tetrabutylammonium (TBA) or octyl
sulfate (C8) ion in 20% acetonitrile solution; flow rate, 1 ml min-';
temperature, 40 "C; detection, UV 254 nm.
80 Chapter 4
Effect of the Ionic Strength of the Buffer Solution
In ion-exchange liquid chromatography, an increase in the buffer concentration
or the addition of a salt is effective in improving the resolution. Increasing the
inorganic ion concentration reduces the ion-pair formation in normal-phase
ion-pair partition liquid chromatography. The retention factor values then
become smaller. In contrast, the buffer concentration does not affect the
retention of the paired-ion in reversed-phase ion-pair liquid chromatography.
However, the solubility of inorganic salts in organic solvent-rich eluents is poor,
making the maintenance of the instrument difficult. It is highly recommended
that a low-concentration buffer solution should be used if the pH is not changed
during the separation.

Selectivity of Organic Modijier and the Concentration Effect

The selectivity and the concentration effect of the organic modifier in the eluent
are the same as those explained above for reversed-phase liquid chromatogra-
phy. Increasing the modifier concentration reduces the retention factor, as does
decreasing the counter-ion concentration. The degree of change in the retention
factor is due to changes in the solubility of the solute in the organic solvent.

Column Temperature Effects

Increasing the column temperature reduces the retention factor. The ion-pair
formation is based on a chemical equilibrium; therefore, temperature control is
important to obtain reproducible results.

Detection
The molecular absorption intensity of polar compounds is usually small, but
highly sensitive detection can be obtained after pre- or post-column derivatiza-
tions. The use of ultraviolet absorption or fluorescence-active counter-ions
makes it possible to achieve highly sensitive detection of polar compounds and
enhance the capability of ion-pair liquid chromatography. For example, N,N-
dimethylprotriptyline has been used as a counter-ion for carboxylic acids12 and
picric acid for quaternary amines13 in normal-phase ion-pair partition liquid
chromatography. Phenethylammonium, cetylpyridinium, 1-phenethyl-2-pyco-
linium, and naphthalene-2-sulfonic acid have been used for sulfonic acid and
alkyl amines detection. l 4 ? l 5 Ion-pair post-column extraction was applied on-line
for fluorescence detection. l 6
Ion-pair liquid chromatography can be applied to compounds separated by
ion-exchange liquid chromatography, and mixtures of ionic and non-ionic
compounds are easily separated. The latter separation is difficult by ion-
exchange liquid chromatography. Anions can be separated by reversed-phase
ion-pair liquid chromatography (Figure 4.18).
Selection of the Eluent 81

I' I
0 min 30
Figure 4.18 Analysis of anions in water using ion-pair liquid chromatography. Column,
octadecyl-bonded silica gel, 15 cm x 4.6 mm i.d.; eluent, 2 mM tetrabutyl-
ammonium hydroxide (pH 5.3) in 3% acetonitrile-water; flow rate,
1 ml min-'; detection, UV 200 nm. Peaks: 1, Br-; 2, NO3-; and 3, I - .

Trouble-shooting in Ion-pair Liquid Chromatography

Ion-pair liquid chromatography can be applied to the separation of a wide
variety of compounds. However, the flexibility of the selection of eluent
components can confuse the operation. Trouble-shooting solutions are sum-
marized in Table 4.6.

3 Normal-phase Liquid Chromatography

Normal-phase liquid chromatography was formerly called adsorption liquid
chromatography. Pure or mixed organic solvents are used as the eluent and
stationary phase adsorbent is more polar than the eluent. Hydrogen bonding is
one of the important molecular interactions between sample molecules and the
adsorbent. The possible molecular interactions are summarized in Table 1.1.
When no molecular interaction is recognized, such chromatography is called
size-exclusion liquid chromatography.
The basic molecular interaction in normal-phase liquid chromatography is
electrostatic forces. The sample molecules are retained strongly by hydrogen
bonding when the sample molecules themselves or the adsorbent act as both a
hydrogen-bond acceptor and donor. The specificity of normal-phase liquid
chromatography is the result of the direct formation of a strong molecular
interaction between sample molecules and the adsorbent. The position of
substitution of sample molecules directly affects the separation of isomers. The
substituent effect is weak in reversed-phase liquid chromatography. Therefore,
normal-phase liquid chromatography is suitable for the separation of isomers,
such as cisltrans, ortholmetalpara, and steric isomers. The steric effect is
especially important for chiral separations on suitable chiral columns. A
Table 4.6 Trouble-shooting for ion-pair liquid chromatography

Trouble Reason Solution

No peak found Low conc. sample soln. Check sensitivity of detector & adjust conc. sample soln.
Unsuitable detector Usually UV absorption of eluent is high because of counter-ion and buffer
Select suitable wavelength or change detector
Disappeared in ghost peak (solvent, Increase conc. sample soln. & compare with previous chromatogram. Or
coun ter-ion) measure diode array chromatogram
Adsorption in column Change column
Increase polarity of column
Increase conc. organic modifier
Dilute conc. counter-ion
Leak from injector Clean and fix injector
Poor flow rate due to surfactant Clean and fix pump head
Poor detection due to surfactant Clean and fix detector flow cell
No retention of analyte Unsuitable column Change column
Unsuitable counter-ion Select further hydrophobic counter-ion
Increase conc. counter-ion
Unsuitable conc. organic modifier Decrease conc. organic modifier or don’t use organic modifier
Too high conc. buffer Decrease conc. buffer, especially inorganic buffer
Unsuitable pH Change pH, take care with pK, of analytes
Non-reproducible Unstable flow rate Clean pump and flow line
retention time Unstable temperature Use column heater or oven
Non-equilibrated column Equilibrate column, especially for initial injections
Too diluted conc. buffer Change buffer
Peak tailing Too strong adsorption on column Change column
Too dilute counter-ion Increase conc. counter-ion
Unsuitable eluent pH Change pH suitable for ionization of analytes
Non-equilibrated column Equilibrate column, usually for 30-120 min
Due to gradient Use moderate gradient
Overloaded sample Reduce amount injected
Peak broadening Low conc. organic Increase conc. organic modifier
Damaged column Change column
Unsuitable counter-ion Use more hydrophobic counter-ion
Unsuitable pH Change pH of eluent
High conc. basic counter-ion Change column, use guard column
pH > 7.5 Change column to suitable one, change eluent
Noise-like peak Unstable flow rate Clean pump head
(air, recrystallized salt, and counter-ion Clean injector
in pump head)
High pressure drop Recrystallization of salt counter-ion Change eluent, buffer components
Wash column with water then methanol
Due to injector or pump Wash and fix injector and pump
Too high conc. buffer Dilute conc. buffer and counter-ion
Low sensitivity Unsuitable wavelength Select appropriate wavelength
Dirty flow cell Wash flow cell with water and methanol
Unsuitable counter-ion Change counter-ion to one having a chromophore or fluorophore and
change the condition
Aged lamp or detector Fix detector

03
w
84 Chapter 4
variety of chiral separation phases have been synthesized; however, one phase
cannot separate many different types of optical isomers, because the steric effect
also depends on the molecular structure of the analytes.
In ligand-exchange liquid chromatography, the separation is performed by
the replacement of analytes, which form a complex with stationary phase
materials or a reagent trapped on the stationary phase surface, by components
in the eluent, which act as similar ligands to the analytes. In charge-transfer
complex liquid chromatography, analytes are retained by charge-transfer
complex formation on the surface of stationary phase material and are replaced
by other components in the eluent which form a charge-transfer complex with
the stationary phase materials or with reagents trapped on the stationary phase
surface.
Normal-phase liquid chromatography is thus a steric-selective separation
method. The molecular properties of steric isomers are not easily obtained and
the molecular properties of optical isomers estimated by computational chemi-
cal calculation are the same. Therefore, the development of prediction methods
for retention times in normal-phase liquid chromatography is difficult com-
pared with reversed-phase liquid chromatography, where the hydrophobicity of
the molecule is the predominant determinant of retention differences. When the
molecular structure is known, the separation conditions in normal-phase LC
can be estimated from Table 1 . 1 , and from the solvent selectivity. A small-scale
thin-layer liquid chromatographic separation is often a good tool to find a
suitable eluent. When a silica gel column is used, the formation of a monolayer
of water on the surface of the silica gel is an important technique. A water-
saturated very non-polar solvent should be used as the base solvent, such as
water-saturated n-hexane or isooctane.

Example I : Chromatography of phthalic acid alkyl esters. Phthalic acid alkyl

esters have been chromatographed on different types of stationary phase
materials in normal-phase liquid chromatography. The stationary phase mate-
rials used were dimethylsilylated silica gel, propyl cyano-bonded silica gel,
polystyrene gel, pellicular-type silica gel, and pellicular-type anion exchangers.
In each case the longer the alkyl chain length of the phthalate, the shorter was
the retention time. This result suggested that the elution order depended on the
alkyl chain length of the samples and was not affected by the stationary phase
materials, and that molecular interaction and solubility, not the type of
stationary phase material, were important for the ~ e p a r a t i o n . ~ ~

Example 2: Chromatography of nitroaniline isomers. The elution order of the

nitroaniline isomers was ortho, meta, and para in normal-phase liquid chroma-
tography using n-butanol-n-hexane mixtures as the eluent, when the stationary
phase material was either silica gel, alumina, an ion-exchanger, polystyrene gel,
or octadecyl-bonded silica gel. The results indicate that the separation of these
compounds can be performed on a range of different types of stationary phase
materials if the correct eluent is selected. The best separation will be achieved by
the right combination of stationary phase material and eluent.68
Selection of the Eluent 85
Classification of Solvents
The most practical method for finding a good eluent for normal-phase liquid
chromatography is a trial experiment using thin-layer chromatography (TLC).
When a rough separation is achieved by TLC and the maximum Rfvalue is less
than 0.5, the developing solvent is a good candidate to be a suitable eluent for
column liquid chromatography. The modification of components depends on
the surface activity, i.e. the water content, and the surface area of silica gel.
The selection of the solvent can be based on the solvent properties classified
by polarity (Table 4.7), hydrogen bonding (Table 4.8), and solubility para-
meters (Table 4.9). Snyder has classified solvents according to their hydrogen
bonding acceptor, hydrogen bonding donor, and dipole properties. 70 The
solvents useful for liquid chromatography are shown in Table 4.9. Water and
chloroform are classified in the same group (VIII); however, these solvents are
not miscible. The method for identifying a good developing solvent for TLC is
also applicable to liquid chromatography. First a solvent in which the analyte is
very soluble, based on the concept of 'like dissolves like' developed by Freizer,
must be selected, then a solvent in which the analyte is not soluble is selected and
these solvents have to be miscible together. The eluent is thus a mixture of
solvents for the analyte.
The selection of a solvent is based on the solvent strength' E experimentally
obtained on silica gel and alumina.69 An eluent having E' = 0.40 on alumina can
be made from a mixture of methyl acetate-pentane (19 : 8 1) or acetone-pentane
(28 : 72). An eluent having EO = 0.20 on alumina can be made from a mixture of
acetone-pentane (6 :94), chloroform-pentane (1 5 : 85), or benzene-pentane
(28:72), as shown in Figure 4.19. However, no linear relationship between

Table 4.7 ClassiJication of solvents by Hecker7'

Non-polar solvent
Hexane Tetrahydro furan
Heptane Dioxan
Isooctane Acetone
Cyclohexane Aniline
Cyclopentane Phenol
Carbon disulfide Ethanol
Carbon tetrachloride Glycol monomethyl ether
Benzene Acetic acid
Trichloroethylene Acetoni trile
Dichloroet hy lene Nitromethane
Chloroform Formic acid
Tetrachloroethylene Morpholine
Dichlorome t hane Formamide
Diethyl ether Water
Methyl ethyl ketone Inorganic acid
Ethyl acetate Salt, Buffer
Pyridine Polar solvent
86 Chapter 4
Table 4.8 Class$cation of solvent based on hydrogen bonding

Strong donor & acceptor Water, glycol, glycerol, aminoalcohol, hydroxylamine,

polyphenol, and amide
Weak donor & acceptor Alcohol, organic acid, phenol, primary and secondary
amines, oxime, nitro and nitrile compounds having
hydrogen at a-position, hydrazine, HF, and HCN
Donor Ether, ketone, aldehyde, ester, tert-amines, pyridine, and
nitro and nitrile compounds not having hydrogen at a-
position
Acceptor Chloroform, dichloroethane, CH3CHC12, CHzClCH2C1,
and CH2CICHClCH2CI
No hydrogen bonding Hydrocarbons, carbon disulfide, mercaptans, solvents
not included in above groups

5
m
A 0
A0 0
0
0
r
a A0 0 0
zUJ 0.2-* 3 .
w go0
c #D
Q)
>
I
3
0 09 I I I 'I

Figure 4.19 Solvent strength of combination of n-pentane and more polar solvents in
normal-phase liquid chromatography using alumina. Symbols: A,methyl
acetate; 0 ,acetone; 0,chloroform; and 0, benzene.

these mixed solvents is obtained; this is both a rough estimation method and a
practical method.
The order of solvent strength (6) (Table 4.9) and the adsorption parameter on
alumina (E') are not parallel. This indicates that solvent strength cannot be
directly related to solvent selection in liquid chromatography. One way of using
Table 4.9 in practice is to rearrange the solvents in terms of the values of E'.
Increasing the E' values strengthens the solubility. The retention time thus
becomes shorter. This method is not perfect, and the elution power also depends
on the structure of the analytes. Another method is the use of 8d,a, a, and a h .
These values indicate the interaction properties of the solvents. The solvent can
be selected on the basis of these values and the properties of the analytes. The
sum of these values and the steric effect may affect the chromatogram.
Table 4.9 Solvents commonly used in liquid chromatography %
2
2
Solvent 6 dd 60 6, 6h &O ul vwv M, dm RI UV BP 5.
("C) q
5
rp
For normal-phase liquid chromatography t?
Polyfluoroalkanes 6.0 6.0 1 - - - - 1.25 210 -50 i$
0 0 -0.25
Isooctane 7.0 7.0 0 0 0 0.01 0.50 101.93 114.230 0.01 5 1.3914 210 98 5
Diisopropyl ether 7.0 6.9 0.5 0.5 0 0.28 0.37 82.51 102.176 1.266 1.368 220 68
n-Pentane 7.1 7.1 0.00 0 0 0.00 0.23 67.02 72.150 0.020 1.3575 190 33
n-Hexane 7.3 7.3 0 0 0 0.0 1 0.313 77.30 86.177 0.016 1.3749 195 69
n-Heptane 7.4 7.3 0.0 0 0 0.0 1 0.41 89.70 100.203 0.006 1.3876 200 98
Diethyl ether 7.4 6.7 1.15 2 0 0.38 0.24 58.98 74.122 1.557 1.3524 218 34
Triethylamine 7.5 7.5 0 3.5 0 0.54 0.36 86.57 101.191 0.848 1.399 345 90
Cyclopentane 8.1 8.1 0.00 0 0 0.05 0.44 59.39 70.134 0.003 1.4064 200 49
Cyclohexane 8.2 8.2 0 0 0 0.04 1.00 70.89 84.161 0.000 1.4262 200 80
Propylchloride 8.3 7.3 3 0 0 0.30 0.35 51.18 78.541 1.725 1.460 1 263 76
Carbon tetrachloride 8.6 8.6 0 0.5 0 0.18 0.97 53.86 153.823 0.001 1.460 1 263 76
Ethyl acetate 8.6 7.0 1.88 2 0 0.58 0.45 57.15 88.106 1.769 1.3724 256 76
Propylamine 8.7 7.3 4 6.5 0.5 - 0.35 50.14 59.111 1.444 1.385 - 48
Chloroform 9.1 8.1 1.15 0.5 0 0.40 0.57 44.50 119.378 1.156 1.4458
1.362 245 60
Methyl acetate 9.2 6.8 4.5 2 0 0.60 0.37 45.29 74.079 1.725 260 56
Benzene 9.2 9.2 0 0.5 0 0.32 0.65 54.47 78.113 1.305 1.5011 278 80
Acetone 9.4 6.8 2.69 2.5 0 0.56 0.36 40.60 58.080 2.91 1 1.3587 330 56
Dichloromethane 9.6 6.4 1.14 0.5 0 0.42 0.44 36.14 84.933 1.503 1.424 1 233 40
Propanol 10.2 7.2 3.09 4 4 0.82 2.3 46.13 60.096 1.534 1.3856 210 98
Pyridine 10.4 9.0 2.37 5 0 0.71 0.95 50.39 79.101 1.973 1.5102 330 115
Ethanol 11.2 6.8 4.0 5 5 0.88 1.20 34.24 46.069 1.551 1.361 210 78.5
Dimethylformamide 11.5 7.9 3.86 - - - 0.92 49.42 73.094 3.550 1.4305 268 153
Acetic acid 12.4 7.0 - - - 1.o 1.26 32.87 60.052 1.885 1.372 230 118

(continued)
00
4
 00
00
Table 4.9 Continued
Solvent

For re versed-phuse liquid chromatography

Triethylamine 7.5 7.5 0 3.5 0 0.54 0.36 86.57 101.191 0.848 1.399 345 90
Propylamine 8.7 7.3 4 6.5 0.5 - 0.35 50.14 59.1 11 1.444 1.385 - 48
Tetrahydrofuran 9.1 7.5 1.75 3 0 0.45 0.55 40.87 68.075 0.492 1.4072 212 190
Acetone 9.4 6.8 2.69 2.5 0 0.56 0.36 40.60 58.080 2.91 1 1.3587 330 56
Dioxane 9.8 7.8 0.45 3 0 0.56 1.37 56.16 88.106 0,010 1.4224 215 101
Pro panol 10.2 7.2 3.09 4 4 0.82 2.3 46.13 60.096 1.534 1.3856 210 98
Isopropanol - - - - - 0.82 1.9 45.66 60.096 1.618 1.384 205 82
Ethanol 11.2 6.8 4.0 5 5 0.88 1.20 34.24 46.069 1.551 1.361 210 78.5
Dimethyl formamide 11.5 7.9 - - - 1.o 1.26 49.42 73.094 3.550 1.372 230 118
Acetoni trile 11.8 6.5 3.44 2.5 0 0.65 0.38 27.68 41.052 2.894 1.3441 190 82
Acetic acid 12.4 7.0 - - - 1 .o 1.26 32.87 60.052 1.885 1.372 230 118
Trifluoroacetic acid - - - - - - - 37.75 114.024 1.879 1.480 260 72.4
Dimethyl sulfoxide 12.8 8.4 3.9 5 0 0.6 2.24 48.43 78.129 3.948 1.483 260 190
Methanol 12.9 6.2 2.87 7.5 7.5 0.95 0.55 22.40 32.042 1.621 1.3284 205 65
Ethanolamine 13.5 8.3 L L L - - 41a.57 6 1.083 1.166 1.4540 - 170
Formamide 17.9 8.3 L L L - - 24.42 45.041 3.699 1.4470 275 210
Water 21.0 6.3 L L L - 1.oo 9.38 18.015 1.860 1.333 190 100
~ ~~

6: Total solubility parameter indicates solvent strength and polarity. Larger 6 means polar, smaller means non-polar (hydrophobic), Sd: the ability of a solvent to
participate in dispersive interactions, which indicates degree of solubility of aromatic compounds having halogen and sulfur substituents. Larger values mean strong
dispersive interaction. 6,: orientation interaction; indicates degree of solubility of dipole compounds. 6,: proton donor property; indicates degree of solubility of
alcohol, phenol, and carboxylic acid. &: proton acceptor property, indicates selective degree of sqlubility of basic compounds. 8': solvent strength in normal-phase
liquid chromatography using alumina. q: viscosity (cp at 20 "C), VWV*: van der Waals volume A3, M,:relative molecular mass, d,: dipole moment in debye, RI:
refractive index, UV: UV cut-off (nm), BP: boiling point ("C), L: too large, *calculated using CACheTMprogram from Sony-Tektronix. This table is based on
references by L.R. Snyder (ed.), 'Principles of adsorption chromatography', Marcel Dekker, New York, 1968, p. 174, and J.J. Kirkland (ed.), 'Modern practice of
liquid chromatography', Wiley-Interscience,New York, 1971, p. 136. s29
A
Selection of the Eluent 89
For example, if the separation of a steroid mixture is difficult in one
chromatographic system, the selection of a different system is needed to separate
the mixtures. The substituents with a hydrogen-bonding effect due to a hydroxy
group, a hydrogen-bonding donor effect of a carbonyl group and a hydrophobic
effect due to an ester suggest the elution order and solvent selection. That is, the
solvent selection is based on &, So, a, and Sh of analytes. Subsequently, another
solvent that is miscible with the first solvent is used for dilution. The final eluent
is a mixture of these two s ~ l v e n t s . ~ l - ~ ~

Preparation of the Eluent

‘Like dissolves like’ is the basic concept for the selection of solvents in the eluent
for liquid chromatography. Controlling the solubility of analytes is the key to
success. If the selected solvent or mixture of solvents does not interfere with
detection, it is a good eluent. The selection of a suitable solvent for low-
wavelength absorption detection and post-column derivatization detection is
important to obtain highly sensitive detection. The selection of a volatile solvent
is the key for preparative-scale liquid chromatography and for mass spectro-
metric detection.
The selection of the solvent is based on the retention mechanism. The
retention of analytes on stationary phase material is based on the physicochem-
ical interactions. The molecular interactions in thin-layer chromatography have
been extensively discussed, and are related to the solubility of solutes in the
solvent. The solubility is explained as the sum of the London dispersion (van der
Waals force for non-polar molecules), repulsion, Coulombic forces (compounds
form a complex by ion-ion interaction, e.g. ionic crystals dissolve in solvents
with a strong conductivity), dipole-dipole interactions, inductive effects,
charge-transfer interactions, covalent bonding, hydrogen bonding, and ion-
dipole interactions. The steric effect should be included in the above interactions
in liquid chromatographic separation.
One example is the separation of tricarboxylic acid cycle organic acids. These
organic acids were originally separated on sulfonic acid-treated silica gel as the
stationary phase material, and a chloroform and isopentyl alcohol mixture was
used as the eluent. However, this eluent is not suitable for UV detection. A
mixture of n-hexane, THF, and tert-butanol was therefore selected for the same
~eparation.~~
One component of the eluent should have properties similar to those of
the analytes, and this solvent is diluted by another solvent to control the
retention time. The basic idea can be understood from the chromatographic
behaviour of phthalic acid esters and polycyclic aromatic hydrocarbons (PAH).
This approach can be applied to the separation procedure for a variety of
stationary phase materials, including silica gel, polystyrene gel, and ion-
exchangers.
The elution order of phthalic esters is related to the carbon chain length. The
longer the chain length, the shorter the retention time in normal-phase liquid
chromatography, and the elution order is reversed in reversed-phase liquid
90 Chapter 4
chromatography. The elution order of PAHs is based on the number of double
bonds, an increased number of double bonds leads to longer retention time in
both normal- and reversed-phase liquid chromatographies. These elution orders
are always observed on a variety of stationary phase materials, including ion-
exchangers. The retention time can be modified by changing the eluent
components or the ratio of the solvent mixture. Increasing the solubility in the
eluent shortens the retention time in both normal and reversed-phase liquid
chromatographies. Increasing the alkyl chain length of analytes makes them
more similar to alkanes. The analyte becomes more hydrophobic (lipophilic)
and hence soluble in n-hexane, but not in water. The solubility of both short and
long alkyl chain compounds is good in ethanol, which is miscible with n-hexane
and water. The addition of ethanol to the eluent makes the retention time
shorter in both normal and reversed-phase liquid chromatographies. This
phenomenon can be observed in the chromatographic behaviour of aliphatic
acids. However, the solubility of PAHs is poor even in ethanol; therefore, the
elution order is always the same in both normal- and reversed-phase liquid
chroma tographies.
In normal-phase liquid chromatography, the elution order of benzene
derivatives on silica gel is alkylbenzenes with a long alkyl chain < alkylbenzenes
with a shorter alkyl chain < benzene < fluorobenzene < chlorobenzene <
cyanobenzene < nitrobenzene << aniline < phenol < benzoic acid. The elution
order of aniline derivatives is dinitroaniline < benzene < anisidine < chloro-
aniline < nitroaniline < phenylenediamine < aminophenol. This elution order
is also observed for anion-exchangers. Alkanes are almost unretained. Alkenes
are retained according to the number of double bonds. The elution order for cis
and trans isomers is cis,cis < cis,trans < trans,trans. This also suggests that the
eluent component should be selected according to the 6, of dipole moment, 6, of
hydrogen bonding acceptor, and 6 h of hydrogen bonding donor, but not the
polarity 6.
In other interactions the surface silanol groups of silica gel can form hydrogen
bonds, and an alumina surface can form hydrogen bonds and a charge-transfer
complex. However, such molecular interactions are caused by positive and
negative charge sites, as in Lewis acid-base interactions. The chloro-ion of
chloride-form anion-exchangers may form hydrogen bonds with aniline and
phenol. Silver ions of silver-form cation-exchangers form charge-transfer
complexes with the n-electrons of alkenes and aromatic compounds.

Size-exclusion Liquid Chromatography (SEC)

This separation method is based on the molecular size of analytes. Analytes pass
through porous stationary phase materials having different pore sizes, and
molecular interactions between analytes and the stationary phase surface must
be eliminated. A very strong solvent is therefore required in this system. This
system is also called gel filtration liquid chromatography, gel-permeation liquid
chromatography, or molecular sieve chromatography. This system is used to
Selection of the Eluent 91
measure the average molecular mass and for the purification of large- from
small-size molecules.

Aqueous Phase Size-exclusion

The elution time is related to the logarithmic relative molecular mass (log M,)
within a certain range of M , for a particular column, as seen in Figure 4.20. The
relationship for proteins is not as simple as that for dextran and polyethylene
glycol. This is because the molecular shape of proteins differ even though 0.2 M
sodium phosphate buffer, pH 6.8, was used as the solvent. The molecular shape
effect can be understood from the different calibration curves in Figure 4.20.
The elution time of polyethylene glycol, whose structure is like a straight chain,
was longer than that of dextran. Dextran and polyethylene glycol can be eluted
with only water as the solvent. The hydrolysates of dextran have been separated
by SEC in pure water.75
The non-uniform behaviour of proteins makes the measurement of their
molecular mass difficult; however, their rough estimation and purification are
very important for biochemical research. Generally, phosphate buffer and tris-
HCl buffer (pH 7) are used for SEC for biological polymers, and some additives
are required for further separation. Only 0.2 M sodium phosphate buffer was

105 -
E0
s
.I

I
& lo4 -
3
0
I
Q)

103 -

20 30 40
Elution time min
Figure 4.20 Calibration curves for size-exclusion liquid chromatography. Column, TSK
GEL G3000SW, 120 cm x 7.5 mm i.d.; eluent, 0.2 M sodium phosphate
bufler pH 6.8; flow rate, 1.0 ml min-'. Standards: 1, protein; 2, dextran;
and 3 , polyethylene glycol.
(Reproduced by permission from Tosoh data)
92 Chapter 4
used for the analysis of gelatin. However, 0.1 M sodium phosphate buffer (pH
7.0) containing 0.1 M sodium chloride was required for the analysis of ribonu-
cleic acid from E. ~ o l iand
, ~ 0.1 M sodium phosphate buffer (pH 7.0) containing
0.3 M sodium chloride was required for that of serum from bone marrow.* Tris-
HCl buffer (pH 7.5) containing 0.1% sodium dodecylsulfate was used for the
analysis of cellular membrane protein from rat micro some^,^ and 50 mM Tris-
HCl buffer (pH 7.2) containing 2 mM EDTA and 300 mM NaCl for the analysis
of plasmid DNA fragments as the elution solvent.8
SEC is a useful tool for monitoring enzyme reactions, as seen in Figure 4.21,
where the speed of decomposition of P-lactoglobulin by a-chymotrypsin is
shown. SEC is widely used for purification of proteins, but the separation is due
only to the difference in molecular mass. Therefore, ion-exchange liquid
chromatography is combined with SEC to improve the selectivity.

Organic Phase Size-exclusion

The average molecular mass of synthesized polymers can be measured by SEC.
This system has been called gel-permeation liquid chromatography, using
porous polystyrene gel as the stationary phase material. Polymers are usually
dissolved in THF or chloroform for separation by SEC. The molecular density

I I I
10 20 min 30
Figure 4.21 Monitoring of an enzyme reaction using size-exclusion liquid chromatogra-
phy. Column, TSK GEL G3000SW, 60 cm x 7.5 mm i.d.; eluent, 0.07 M
potassium phosphate bufler containing 0.1 M potassium chloride; flow rate,
1 rnl min-'; detection, UV 280 nm. Peaks: 1, J-lactoglobulin; 2, a-chymo-
trypsin; and 3, decomposedproducts.
(Reproduced by permission from Tosoh data)
Selection of the Eluent 93

0 5 10 15 min 20 25
Figure 4.22 High temperature size-exclusion liquid chromatography of an engineering
plastic, poly(pheny1 sulfate). Column, SSC GPS-3506, 50 cm x 8 mm i.d.;
';
eluent, I-chloronaphthalene; $ow rate, 1 .O ml min- column temperature,
2 10 "C; detector, refractive index detector.
(Reproduced by permission from Senshu Kagaku data)

of analytes is measured by comparison of chromatograms obtained from light

scattering and refractive index detectors.
Newly synthesized organic polymers having high temperature resistance and
physical strength are not soluble in THF or chloroform. High-temperature SEC
is therefore being developed, and the pressure-drop measurement detector and
flame ionization detector are also used for monitoring the high temperature
effluent. A chromatogram of an engineering plastic, poly(pheny1 sulfate), is
shown in Figure 4.22. The column temperature was 210°C and the eluent was
1-~hloronaphthalene.~~

5 References
1. L.R. Snyder, J . Chromatogr. Sci., 1978, 16, 223.
2. D.J. Pietrzyk and C.-H. Chu, Anal. Chem., 1977,49,860.
3. T. Hanai, K.C. Tran, and J. Hubert, J . Chromatogr., 1982,239, 385.
4. T. Hanai and J. Hubert, HRC & CC, 1984,7,524.
5. Y. Arai, M. Hirukawa, and T. Hanai, Nihon Kagaku Kaishi, 1986,969.
6. From the Mitsubishi Kasei catalogue.
7. From the Tosoh catalogue.
8. From the Asahipak catalogue (Showa Denkko).
9. G. Schill, R. Modin, and B.-A. Person, Acta Pharm. Seuc., 1965,2, 119.
10. B.L. Karger, S.C. Su, S. Maichese, and B.-A. Person, J. Chromatogr. Sci., 1974,12,
678.
11. M. Hirukawa, M. Maeda, A. Tsuji, and T. Hanai, J. Chromatogr., 1990,507,95.
12. S . Eksborg, P.-0. Lagerstrom, R. Modin, and G. Schill, J. Chromatogr., 1993,83,99.
13. E. Eksborg and G. Schill, Anal. Chem., 1973,45,2029.
14. M. Denkert, L. Hackzell, G. Schill, and E. Sjogren, J . Chromatogr., 1981,218, 31.
15. L.E. Vera-Ada, M. Caude, and R. Rosset, Analusis, 1982, 10,43.
16. K. Sugden and G.B. Loscombe, J . Chromatogr., 1978,149,377.
17. B.A. Bildingmeyer and F.V. Warren, Anal. Chem., 1984,54,2351 .
18. S.J. Valenty and P.E. Behnken, Anal. Chem., 1978,50, 834.
19. B. Franssod, K.-G. Wahlund, I.M. Johansson, and G. Schill, J . Chrornatogr., 1976,
125, 327.
94 Chapter 4
20. J.F. Lawrence, F.E. Lancaster, and H.B. Conacher, J. Chromatogr., 1981,210,168.
21. M.G.M. de Ryter, R. Cronnelly, and N. Castagnoli, Jr., J. Chromatogr., 1980, 183,
193.
22. W. Voelter, K. Zech, P. Arnold, and G. Ludwig, J. Chromatogr., 1980,199, 345.
23. D.N. Buchanan, F.R. Fucek, and E.F. Domino, J. Chromatogr., 1980,181,329.
24. R.G. Peterson, B.H. Rumack, J.B. Sullivan, and A. Makowski, J. Chromatogr.,
1980,188,420.
25. M. Ghias-Ud-Din, E.B. Olson, Jr., and J. Rankin, J . Chromatogr., 1980, 192,463.
26. K.T. Muir, J.H.G. Jonkman, D.-S. Tang, M. Kunitani, and S. Riegelman, J.
Chromatogr., 1980,221,85.
27. M. Schoneshofer and A. Fenner, J. Chromatogr., 1981,224,472.
28. C.M. Riley, L.A. Sternson, and A.J. Repta, J. Chromatogr., 1983,276,93.
29. M. Kneczke, J. Chromatogr., 1980,198,529.
30. M. Alvinerie, P. Galtier, and G. Escoula, J. Chromatogr., 1981,223,445.
31. N. Kurosawa, S. Morishima, E. Owada, and K. Ito, J. Chromatogr., 1984,305,485.
32. D. Beales, R. Finch, and A.E. McLean, J. Cllromatogr., 1981 226,498.
33. P.B. Baker and T.A. Gough, J . Chromatogr. Sci., 1981,19,483.
34. J. Wagner, M. Palpeyman, and M. Zraika, J . Chromatogr., 1979,164,41.
35. R.V. Smith, J.C. Glade, and D.W. Humphrey, J . Chromatogr., 1979,172,520.
36. K. Makino, H. Ozaki, T. Matsumoto, H. Imaishi, and T. Takeuchi, J. Chromatogr.,
1987,400,271.
37. Y. Kobayashi, H. Kubo, and T. Kinoshita, J. Chromatogr., 1987,400, 113.
38. G. Inchauspe, P. Delrieu, P. Dupin, M. Laurent, and D. Samain, J. Chromatogr.,
1987,404, 53.
39. I. de Miguel, E. Puech-Costes, and D. Samain, J. Chromatogr., 1987,407, 109.
40. J.J. Lauff, J. Chromatogr., 1987,417,99.
41. K. Lindgren, J. Chromatogr., 1987,413, 351.
42. G. Lam, R.M. Williams, and C.C. Whitney, J. Chromalogr., 1987,413,309.
43. D.L. Theis, J. Chromatogr., 1987,402,335.
44. A. Sokolowski and K.-G. Wahlund, J. Chromatogr., 1980, 189,299.
45. Y. Yoshitoku, M. Moriyasu, and Y. Hashimoto, Chem. Lett., 1982, 193-194.
46. P.G. Rigas and D.J. Pietrzyk, Anal. Chem., 1987,59, 1388.
47. B.A. Bildingmeyer, C.T. Santasania, and F.V. Warren, Jr., Anal. Chem., 1987, 59,
1843.
48 L. Gagliardi, G. Cavazzutti, A. Amato, A. Basili, and D. Tonelli, J. Chromatogr.,
1987,394,345.
49 M.L. Abaugh, 1,s.Krull, and W.R. Lacourse, J. Chromatogr., 1987,387, 301.
50. .J.H. McB. Miller, C. Pascal, and M. Tissieres, J . Chromatogr., 1987, 392, 361.
51. .J.A. De Shutter, W. Van Den Bossche, and P. De Moerloose, J. Chromatogr., 1987,
391, 303.
52 H. Lentzen and R. Simon, J . Chromatogr., 1987,389,444.
53 E. Naline, B. Flouvat, C. Advenier, and M. Pays, J . Chromatogr., 1987,419, 177.
54. J.H.
# Kennedy and B.A. Olsen, J. Chromatogr., 1987,389, 369.
55 M. Dreux, M. Lafosse, P. Agbo-Hazoume, B. Chaabane-Doumandji, M. Gibert,
and Y. Levi, J. Chromatogr., 1986,354, 116.
56 S.P. Sood, L.E. Sartori, D.P. Wittmer, and W.G. Haney, Anal. Chem., 1976,48,796.
57 K.-G. Wahlund and A. Sokolowski, J. Chromatogr., 1978, 151, 299.
58. N.D. Brown and H.K. Sleeman, J. Chromatogr., 1977, 138,449.
59. D.P. Wittmer, N.O. Nuessle, and W.G. Haney, Jr., Anal. Chem., 1975,47, 1422.
60. E. Fitzgerald, Anal. Chem., 1976,48, 1734.
61. J.H. Knox and J. Jurand, J. Chromatogr., 1975, 110, 103.
62. J.C. Kraak, K.M. Jonker, and J.F.K. Huber, J . Chromatogr., 1977,142,671.
63. K.-G. Wahlund and U. Lund, J. Chromatogr., 1976,122,269.
64. J.H. Knox and G.R. Laird, J . Chromatogr., 1976, 122, 17.
65. K.-G. Wahlund, J. Chromatogr., 1973, 115,411.
Selection of the Eluent 95
66. J. Mitchell and C.J. Coscia, J. Chromatogr., 1978, 145, 295.
67. W. Funasaka, T. Hanai, and K. Fujimura, J. Chromatogr. Sci., 1974,12,517.
68. T. Hanai and K. Fujimura, J. Chromatogr. Sci., 1976, 14, 140.
69. R. Neher, in ‘Thin-layer chromatography’, ed. G.B. Marini-Bettola, Elsevier,
Amsterdam, 1964, p. 75.
70. L.R. Snyder, J. Chromatogr., 1974,92,223.
71. S. Hara, J. Chromatogr., 1977, 137, 199.
72. S. Hara, Y. Fujii, M. Hirasawa, and S. Miyamoto, J. Chromatogr., 1978,149, 143.
73. S. Hara and A. Ohsawa, J . Chromatogr., 1980,200,85.
74. H. Hyakutake and T. Hanai, J . Chromatogr., 1975,108,385.
75. From Shodex data.
76. From Senshukagaku data.

Bibliography of additional references on theoretical approaches in

ion-pair liquid chromatography
A. Bartha, H.A.H. Billiet, L.D. Galan, and G. Vigh, ‘Studies in reversed-phase ion-pair
chromatography 111. The effect of counter-ion concentration’, J. Chromatogr., 1984,
291, 91.
A. Bartha, G . Vigh, H. Billiet, and L. de Galan, ‘Effect of the type of ion-pairing reagent
in reversed-phase ion-pair chromatography’, Chromatographia, 1985,10, 587.
B.A. Bidlingmeyer, S.N. Deming, W.P. Price, Jr., B. Sachok, and M. Petrusek,
‘Retention mechanism for reversed-phase ion-pair liquid chromatography’, J.
Chromatogr., 1979,186,419.
J.B. Green, ‘Liquid chromatography on silica using mobile phases containing tetraalkyl-
ammonium hydroxides’, J. Chromatogr., 1986,358, 53.
M.T.W. Hearn (ed.), ‘Ion-pair Chromatography’, Marcel Dekker, New York, 1985.
S.O. Jansson, ‘Effect of counter-ions in ion-pair liquid chromatography of hydrophobic
amines on non-polar bonded phases’, J . Liq. Chromatogr., 1982,5,677.
J.C. Kraak, H.H. van Rooij, and J.L.G. Thus, ‘Reversed-phase ion-pair systems for the
prediction of n-octanollwater partition coefficients of basic compounds by high-
performance liquid chromatography’, J. Chromatogr., 1986,352,455.
W.R. Melander and Cs. Horvath, ‘Mechanistic study of ion-pair reversed-phase
chromatography’, J. Chromatogr., 1980, 201, 21 1.
M. Puttemans, L. Dryon, and D.L. Massart, ‘Extraction of organic acids by ion-pair
formation with tri-N-octylamine’, Anal. Chim. Acta, 1984, 161,221.
W. Santi, J.M. Huen, and R.W. Frei, ‘High-speed ion-pair partition chromatography in
pharmaceutical analysis’, J . Chromatogr., 1975,115,423.
G. Schill, ‘Selective extraction of organic compounds as ion-pairs and adducts’, Talanta,
1975,22, 1017.
M. Tang and S.N. Deming, ‘Interfacial tension effects of nonionic surfactants in
reversed-phase liquid chromatography’, Anal. Chem., 1983,55,425.
O.A.G.J. van der Houwen, R.H.A. Sore], A. Hulshoff, J. Teeuwsen, and A.W.M.
Indemans, ‘Ion-exchange phenomena and concomitant pH shifts on the equilibration
of reversed-phase stationary phase materials with ion-pairing reagents’, J. Chroma-
togr., 1981, 209, 393.
CHAPTER 5

Separation Based on an Improved

Column Eficiency

High-speed separation can be achieved by reducing the column length, column

internal diameter, the size of the stationary phase particles, and by increasing
the eluent flow velocity, as explained in Chapter 1. In addition, the development
and optimization of pumps, detectors, injectors, and the recording systems
contribute to improved separations. The influence of diffusion theory on the
experimental separation should also be considered. If the unresolved chromato-
gram shown in Figure 5.1A has been obtained for two compounds, two
methods can be applied to achieve a complete separation. Figure 5.1B
demonstrates that achieved by improving selectivity (relative retentions), and
Figure 5.1C demonstrates a separation achieved by improving the column
efficiency.
The resolution, R,, is used to describe the separation conditions of two
compounds. The R, value is given by the ratio of the distance between

I B

elution volume -+
Figure 5.1 Improvement of separation. A, unresolved peaks; B, increase in peak separa-
tion; C , improved peak eficiency.

96
Separation Based on an Improved Column Efficiency 97
, the mean of the peak widths,
the maxima of two peaks, A ~ R and wb, at their
base:

where tR1and tR2 are the retention times of the peaks, and wbl and wb2 are the
base widths of the two peaks.
Theoretically, when R, is 1, the two peaks are 98% separated. An R, value
> 1 indicates a good separation and an R, value c 1 indicates a poor
separation. In Figure 5.2, the resolution between peaks 3 and 4 was 1.1. The
resolution between peaks 4 and 5 was 1.5, and that between peaks 5 and 6
was 2.4. The resolutions do not exactly match with their respective peak
shapes because of their different intensities. If two symmetric peaks have a
separation that results in a resolution of R, = 1, when the peaks have equal

1
5

I
0 4 8 min
0 1 2 mL
-~

peak 3 4 5 6

plate number N 3,364 2.116 3.330 1.129

Rs 1.1 1.5 2.4

Figure 5.2 Normal phase separation of phthalates. Column, Yanapak C N ,

25 cm x 2 mm id.; eluent, n-hexane-n-butanol (200.1); flow rate,
0.25 ml min-'; pressure, 3 MPa. Peaks: 1, lauryl phthalate; 2, heptyl
phthalate; 3, butyl phthalate; 4, propyl phthalate; 5 , ethyl phthalate; and
6 , methylphthalate.
98 Chapter 5
areas (Figure 5.3A), the resolution will deteriorate if there is a change in the
ratio of the areas of the peaks, even if the retention times remain constant. An
example is shown in Figure 5.3B;as the ratio increases the experimental
resolution decreases. In designing a procedure it is therefore desirable to
obtain equal peak areas for overlapping peaks. When the wavelength of the
isosbestic point of the spectra of two compounds is used for photometric
detection, the peak areas of a 1:l mixture will be equal. Alternatively, the
selection of the wavelength can be used to generate equal peak areas even if
the mixture ratio is very different. When pure standard chemicals are used,
their spectra should first be measured, and the most suitable wavelength
selected to obtain good resolution and the correct peak area ratio. In practice,
the selectivity depends on the structure of the analytes and the mixture ratio,
and it is therefore often difficult to obtain an appropriate relationship between
the R, value and the mixture ratio. When one large peak is observed on a
chromatogram, the purity of the peak should be studied by using a longer
column or changing the detector wavelength. Conversely, poor resolution on
a chromatogram can be improved by appropriate selection of the detector
wavelength.
An improvement in the resolution can also be achieved by increasing the
separation of the peak (AtR) or narrowing the peak width (increased N ) . An
increase in AtR depends on the selection of stationary phase materials and
eluent; a reduction of the peak width depends on improving the column

(A) theoretical (B) experimental

volume resolution volume resolution
ratio 0.8 1.0 1.25 ratio

1:l A A 1:l 1.46

2:l

4:1
h 2:l 1.46

8:1 Jisi 4:l

8:l
1.32

0.84
32:l

128:l 16:l

Figure 5.3 Resolution related to peak area ratio. (A) theoretical and (B) experimental
changes in resolution with volume ratio. Column, 10 pm octadecyl-bonded
silica gel, 15 cm x 4.6 mm id.; eluent, 90% aqueous acetonitri1e;jow rate,
1 ml min-’; detection, UV 254 nm. Sample: diflerent ratios of 10 pl benzene
and toluene in 10 ml ethanol.
Separation Based on an Improved Column Efficiency 99
efficiency. The enhancement of high-performance liquid chromatography
depends mainly on improving the column efficiency.

1 Improving Separation by Changing the Selectivity

The selectivity of a separation is based on differences in the physicochemical
interactions of the analytes with the stationary and mobile phases. This
difference is measured as AtR on a chromatogram.
The retention volume of each component, V,, is given by:

where V Mis the volume of mobile phase inside the column, Vs is the volume of
the stationary phase in reversed-phase chromatography, and K is the distribu-
tion constant of the analyte between the mobile and stationary phases. In
normal-phase liquid chromatography, V , is related to the surface area of the
stationary phase material, and in ion-exchange liquid chromatography to the
ion-exchange capacity. Differences in K values between analytes in one system
reflect differences in the properties of the compounds. In any one column the
distribution between the stationary and mobile phases is described by the mass
distribution coefficient. To change values of AK between analytes, either the
phases must be changed (changes in stationary and mobile phase selectivity, see
earlier) or the temperature must be changed.

amount of solute in stationary phase

Mass distribution constant =
amount of solute in mobile phase

where Vs and V Mare constant for all analytes in a particular column. This value
is equal to the retention factor k . The stronger is the interaction between the
sample molecule and the stationary phase the larger the retention of solute.
Thus the larger the difference in K between analytes (AK), the better will be their
separation. To improve the separation, the difference between the distribution
constants ( K values) of compounds 1 and 2 must be increased, or the volume of
the stationary phase ( V s )must be increased.

To increase Vs, the chromatographer can increase the surface area of the
stationary phase materials in normal-phase liquid chromatography, increase the
stationary phase volume in reversed-phase or partition liquid chromatography,
or increase the ion-exchange capacity in ion-exchange liquid chromatography.
In general, if the internal diameter of a column is constant, the retention time
100 Chapter 5
difference (AtR) is related to the column length (L). Therefore, doubling the
column length doubles the difference A t R . However, the resolution of the
analyte (R,)is only related to ,/L due to the diffusion of the samples inside the
column.
There is no molecular interaction in size-exclusion liquid chromatography,
and therefore the resolution can only be improved by increasing the column
length.

2 Improving Separation by Increasing the Column

Efficiency
The peak broadening due to diffusion is related to the square root of L as
follows:

where H i s the column efficiency, i.e. the height equivalent to a theoretical plate
(HETP). A smaller H value gives better resolution.
From Equation 5.3:

If a peak diffuses as a theoretical Gaussian distribution, wb is equal to 40. One

peak width is 40 standard deviation, from Equation 5.4:

H cc 02/L (5-5)

The separation factor a of two compounds is given from Figure 3.4:

a=
tR2 - tRO -
- v R 2 - VO -
--K2
tR1 - tRO V R l - VO K1

The number of theoretical plates is given by:

(2)2- (2)2=
N = 16 - - 16 5 . 5 4 5 ( 3 2 = (y

The effective number of theoretical plates is the number of usable theoretical

plates and is calculated after elimination of the void volume:

If the peak widths of peaks 1 and 2 are assumed to be equal, the following
equation is derived from Equation 5.1:
Separation Based on an Improved Column Efficiency 101

From Equation 5.7,

*wb2
=0 . 2 5 a (5.10)

From Equation 5.2, ( V R ~- VM) = K2 Vs, ( VRI- VM) = K1 Vs

Hence:

From Equation 5.6, a = K 2 / K l ,therefore

The mass distribution coefficient k = K,V,/VM = (VR - VM)/VM, and

Therefore, from Equation 5.9:

a
R ~ = - x - a - 1 ~ k x m =1 ~ ( - 1- - a- -&1 - & ) f i (5.11)
4 a

Combination of Equations 5.8 and 5.1 1 gives:

(5.12)

The relationships between R,,a, and N are given in Figure 5.4. When a is
small, a large N is required for a good separation. However, the use of a longer
column should be avoided for fast analysis even if a longer column provides
excellent plate numbers. An improvement in the value of a can be estimated for
a good separation.
A column having a smaller HETP value is a good column because diffusion
inside the column is small, resulting in better separation. The HETP value is
given by the Van Deemter equation, which describes the peak broadening of
packed columns through which a non-compressible solvent is moving.
102 Chapter 5

105

L
0,
P 104
5
E

1 r I

1.01 1.10 1.20 1.30

a
Figure 5.4 Relationship between separation factor a, resolution R,, and plate number N .
Line: 1, R, = 1.5; 2, R, = 1.2.5; 3 , R, = 1.0; and4, R, = 0.8.

H = H F + H L + Hs + H M .The H value can be expressed as the sum of (a) eddy

diffusion ( H F ) , (b) longitudinal diffusion (HL), and (c) resistance to mass
transport in stationary and mobile phases (Hs and H M ,respectively).

Reducing the Eddy Diffusion Term

Eddy diffusion is a result of the presence of particles of stationary phase
material in a column, and depends on the stationary phase conditions, shape
of the column, and the structure of the stationary phase material. The influence
of the stationary phase material can be divided into the particle size (d,,), the
shape of the particles, and the porosity of the particles. The standard deviation
for peak broadening due to the particles is described by: oF2= 2AdpL, and
therefore from Equation 5.5:

HF = 2Adp (5.13)

1 depends on the irregularity of the particles (particle shape) and on the column
material (steel gives more disturbance than glass); the relative effect will always
increase as the column diameter decreases. The above-mentioned effects have
led to an optimum internal diameter of 2-5 mm for LC columns. In a column
with this internal diameter, which is uniformly packed with small spherical
particles, the eddy diffusion will be limited. The A value is about 1 for spherical
particles.
Separation Based on an Improved Column Efficiency 103
Reducing the Effect of Longitudinal Diffusion
Molecules tend to diffuse randomly, in no particular direction, within any fluid,
independently of the flow rate of the mobile phase. Their diffusion rate is
determined by the type of molecule, the nature of the mobile phase, and the
temperature, and is expressed quantitatively by their diffusion constants.
Molecular diffusion is related to time, as shown by the Einstein equation:

o2 = 2Dt (5.14)

where D is the diffusion coefficient. The standard deviation due to longitudinal

diffusion is oL, and the time during which diffusion can occur is t, = L/u;
therefore, oL2= 2 y ~ , t , = (2 7 4 . &)/# , and from Equation 5.5, H = 0 2 / ~

where y is the factor of anti-diffusion, and is less than 1; DM is the diffusion

constant of sample molecules in the mobile phase; and u is the linear velocity of
the mobile phase. The term HL is important in gas chromatography because D M
is high. The DM value in liquids is ca. low5,and the flow rate is greater than the
diffusion speed; therefore, HL is not as important.
The HL value is reduced by an increase in the viscosity of the solvent or by a
decrease in the temperature. Longitudinal diffusion can thus be reduced by
decreasing the diffusion coefficient and increasing the flow rate; however, these
two actions are counter-effective in liquid chromatography because of the mass
transport term.

Reducing the Effect of Resistance to Mass Transport in the

Stationary and Mobile Phases
This dispersion effect results from the mass transfer of the analyte between the
stationary and mobile phases, and is an important phenomenon in high-
performance liquid chromatography. The injected analyte molecules are first
present in the mobile phase. They then transfer back and forth from the
stationary phase in order to interact. This process is repeated along the
column, from the inlet to outlet. Such a phenomenon occurs in all kinds of
chromatographies: normal-phase, reversed-phase, partition and ion-exchange
liquid chromatography. (In size-exclusion liquid chromatography, such move-
ment of sample molecules occurs as a diffusion process between the inside and
outside of the stationary phase material.) Under ideal conditions, the sample
molecules move along the column at the same speed in both the mobile and
stationary phases (dotted lines in Figure 5.5). In practice, the movement of the
mobile phase drags the analyte along faster than their movement through the
stationary phase because equilibration at the interface is restricted by the
diffusion rates within the phases. Therefore peak broadening occurs according
to the total diffusion profile, as shown in Figure 5.5. The total diffusion effect
104 Chapter 5
Total diffusion

Mobile
phase
,-
Stationary
phase \ “t,, / ”
;, Flow direction
’ u - - . ’ i d e a l diffusion
Diffusion in stationary
Rhase

Figure 5.5 Diffusion of analyte in mobile and stationary phases while moving along the
column.

can be divided into diffusion in the stationary phase ( H s ) and that in the mobile
phase (Ifhl). Within a certain time ( t a ) , the sample molecules in the mobile phase
move a particular distance (Uta) at a linear velocity (u). During this period, the
movement of the sample molecules in the stationary phase is zero. If the number
of sample molecules transferred between the mobile and stationary phases is
defined as n,, and the distance the centre of the band has moved is defined as dR;
n, can be related to 2dR/uta. The time the analyte spends in the stationary phase
is ?d. The ratio R of the moving speed of the band and the mobile phase is
defined as ta/(ta + t d ) . The following equation is then obtained:

The distance (I,) that the band centre has moved is related to R and Utd:

The standard deviation caused by molecular diffusion is:

o2 = 12ns (5.16)

Therefore, the standard deviation due to molecular diffusion inside the

stationary phase (as) is:

as2 = 2qR(1 - R ) u L t d (5.17)

where q is a parameter that depends on the type of stationary phase used, e.g.
spherical, irregular, fiber or porous. The q value is 213 for pellicular-type
stationary phase materials and 1/30 for porous-type stationary phase materials.
The average residence time ( t d ) of the sample molecules in the stationary
phase is related to the thickness of the stationary phase (ds) and the diffusion
constant ( D s ) of the sample molecules in the stationary phase. Thus 0’ = 2Dt
(Equation 5.14) and ?d = ds2/2Ds is obtained.
Separation Based on an Improved Column Efficiency 105

R = VM(VM + KVs) and DM = KVs/ V,, therefore R = 1/(1 + DM). Then,

Equation 5.17 is converted into:

Therefore, the band broadening due to diffusion of sample molecules in the

stationary phase (Hs)is given by:

From this equation, the Hs value is small in a high-efficiency column. Such a

column has a stationary phase with a thin film or small ds value. That is, a high-
performance separation can be achieved by using a stationary phase in which
the depth into which the sample molecules can move is diminished. This means
that a thinner or shallower stationary phase gives higher performance in
reversed-phase liquid chromatography, and a smaller particle size gives higher
performance in ion-exchange and size-exclusion liquid chromatography.
However, speeding up the diffusion counteracts lowering the flow rate. For
diffusion in the mobile phase, we can consider the eddy diffusion (HF) of
Equation 5.13, longitudinal diffusion (HL)of Equation 5.15, and the diffusion
toward the wall from the centre of a column (or space between particles) (HD).
The time taken for the sample molecules to move from the column side wall or
particle surface is given by Equation 5.14, where the distance is given as wadp:

where w, is the ratio between the distance moved to the wall and the particle
diameter (dp). When cob is defined as the ratio between limited partial linear
velocity and the average linear velocity, the distance for sample molecules
movement ( I ) is opt. Therefore, the standard deviation, cr, is:

where w = oa2wp2/2.Then the diffusion effect due to the mobile phase, HD, is:

(5.19)

where co is the column constant, which depends on the shape of the column and
the packing conditions.
As measured, the Hs term means that a smaller particle size reduces the
distance between particles, and thus the spreading due to diffusion of sample
molecules is minimized. From Equations 5.18 and 5.19, a decrease in H value is
achieved by increasing the diffusion speed (elevating the column temperature,
106 Chapter 5
using a low-density, low-viscosity solvent), decreasing the flow rate, reducing
the thickness of the stationary phase, and decreasing the particle size.
The overall effects of peak diffusion can be summarized as follows:

Overall, the most effective factor in Equation 5.20 is the particle size. The
smaller the particle size, the higher the column efficiency. Equations 5.13, 5.15,
and 5.18 are depicted in Figure 5.6 against flow velocity as A, B, and C ,
respectively. The band spreading is thus influenced by Equation 5.15 at a low
flow rate. The band spreading is influenced by Equations 5.18 and 5.19 at a high
flow rate. For gas chromatography curve D is obtained.
In liquid chromatography, the diffusion rates are slower than that in gas
chromatography, and the values of DM and Ds are very small; therefore, the
minimum H value is obtained at a low flow rate, as shown by curve E in Figure
5.6. The value of H increases slowly at higher flow rates in liquid chromato-
graphy. An experimental result is shown in Figure 5.7. The HETP was minimal
at a certain flow rate, and the measured optimum value was less than 10 pm for
',
this column. The optimum flow rate was about 0.9 ml min- corresponding to
a linear flow velocity of about 55 mm min-'.
Experimentally, high-performance separations can be performed at lower
flow rates using small particle-size stationary phase materials as shown in
Figure 1.1. A fast analysis can be achieved at a higher flow rate when the H

n
c
w
I

linear velocity cdmin

Figure 5.6 Van Deemter curves relating H to linearflow velocity. A, eddy difusion term;
B, molecular difusion term; C , resistance to mass transfer term; D, summation
for gas chromatography; and E, summation for liquid chromatography.
Separation Based on an Improved Column Efficiency 107
flow rate 1.0 mL I min

A benzene k = 0.86

b toluene k = 1.18
x ethylbenzene k = 1.55
0 propylbenzene k = 2.18

+ butylbenzene k = 3.18

A pentylbenzene k = 4.34

N = 100,000 / m

0 20 40 60 80 100
Linear flow velocity mdmin
Figure 5.7 HETP of compounds with diflerent k values related to linear flow velocity.
Column: 5 pm ODs-bonded silica gel, 15 cm x 4.6 mm i.d.; eluent: 80%
aqueous acetonitrile.

value is smaller; however, Equations 5.18 and 5.19 indicate that a high flow rate
operation has negative effect. Curve E in Figure 5.6 indicates that a high flow
rate operation does not significantly increase the H value in liquid chromato-
graphy. This means that a fast, efficient analysis can be performed by using a
short column packed with small particle size stationary phase materials.
The pressure drop, Ap, inside a column is given by Equation 5.21:

where tj is the viscosity of the eluent, L is the column length, u is the linear flow
velocity,fis the sum of the void volume of a column, and K O is the parameter
indicating the column permeability. When the void volume between the
stationary phase materials is E , K O is given by Equation 5.22:
dp2 E~
KO=-X- (5.22)
180 (1 - E ) 2

From tR = L(DM + l)/u and Equation 5.21, the following equation is obtained:

(5.23)

Thus when q , f , and L are constant, shortening the analysis time can be achieved
by increasing the pressure drop and the K O value. Increasing K O can be done by
decreasing E and/or increasing the particle size, dp, of the stationary phase
material. Increasing the particle size decreases the resolution, R,;therefore,
small particle-size stationary phase materials must be densely packed. As a
result, a high-pressure operation is unavoidable.
108 Chapter 5
In conclusion, the smaller is the particle size the better the separation;
however, a high pressure drop requires the optimum design of columns and
instruments. A column well packed with spherical small particles has a high
plate number, and is suitable for the separation of a homologous series of
compounds. However, the separation of isomers requires a high selectivity of
stationary and mobile phases as described in Chapters 3 and 4.

3 Bibliography
See Chapter 1 for details.
CHAPTER 6

Influence of Physical Chemistry

on Separations in Liquid
Chromatography

1 Quantitative Structure-Retention Relationships in

Reversed-phase Liquid Chromatography
Various methods have been employed for the prediction of retention times in
reversed-phase liquid chromatography.
As the alkyl chain length of fatty acids increases, their solubility in water
becomes poorer and that in n-hexane becomes higher. Such a phenomenon is
easily understood from a comparison of the properties of acetic acid and butter
fat. In reversed-phase liquid chromatography, the retention time of fatty acids is
related to the number of methylene units, i.e. the chain length. This relationship
is a basic concept used to study retention times in reversed-phase liquid
chromatography. In early work the number of carbon atoms of the aliphatic
chain was related to the retention of alcohol^'^ and Even the
polymerization of very polar compounds, such as amino acids and monosac-
charides, makes them less soluble in water than their monomers. This means
that polymerized molecules become more hydrophobic and their retention times
become longer in reversed-phase liquid chromatography. Therefore, the reten-
tion time of a homologous series of compounds is easily predicted from the
number of monomer units, and sometimes by the carbon numbers. It can also
be related to molecular size, van der Waals volume, and surface area.
A quantitative analysis of the structure-retention relationship can be derived
by using the relative solubility of solutes in water. One parameter is the partition
coefficient, log P, of the analyte measured as the octanol-water partition
distribution. In early work, reversed-phase liquid chromatography was used to
measure log P values for drug design. Log P values were later used to predict the
retention times in reversed-phase liquid chromatography.The calculation of the
molecular properties can be performed with the aid of computational chemical
calculations. In this chapter, examples of these quantitative structure-retention
relationships are described.

109
110 Chapter 6
Prediction of Retention Times from log P in Reversed-phase Liquid
Chromatography
Prediction of Partition Coeficient (log P )
The measurement of the solubility of drugs in polar and non-polar media is very
important in the pharmaceutical field. One method proposed to describe this
solubility is the partition coefficient between octanol and water. The mathema-
tical calculation of an octanol-water partition coefficient from values for
functional groups was first proposed by Hansch et al. as Hansch's n constants,'
and was later developed by Rekker as hydrophobic fragmental constants
(log P).2 This method was further improved by the use of molecular connectiv-
ities.17 The prediction of logP values can be performed by either a computer
program or by manual calculation. For example, approximate partition
coefficients (log P)have been calculated by Rekker's method:

A part of his fragmental constant tables is given in Table 6.1. (The predicted
log P values are summarized in Table 6.4.)
Attempts have also been made to determine octanol-water partition coeffi-
cients empirically using liquid chromatography by comparison of retention
times with those of compounds of known log P.3-'2

Correlation between log P and log k Values

It seemed possible to correlate the elution order in reversed-phase liquid
chromatography with the octanol-water partition coefficient, log P. For
example, the partition coefficients calculated by Rekker's method showed a

Table 6.1 Rekker 's hydrophobic fragment constants

Fragment A log P Fragment A log P Fragment A log P

1.886 (a1)COOH -0.954 (al)OCH2COOH - 1.155

1.688 (ar)COOH - 0.093 (ar)OCH2COOH -0.581
1.431 (al)C00 - 1.292
0.702 (ar)COO - 0.43 1
0.530 (a1)CO - 1.703
0.235 (ar)CO -0.842
0.36 (a1P - 1.581
0.15 (am -0.433
0.935 (a1)OH - 1.491
0.73 (ar)OH -0.343
0.175 (ar)COH -0.38

ar: substituent on aromatic compound, al: substituent on aliphatic compound; hydrophobic

fragment constants in octanol-water from Rekker2.
Influence of Physical Chemistry on Separations in Liquid Chromatography 111
linear relation to log k values measured in reversed-phase liquid chromatog-
raphy: l 8
logk=y x logP+m (6.2)

where y and rn are constants in a given system. An example is given in Figure

6.1. The values of log P obtained from Equation 6.1 using the measured log k of
58 compounds, including alkanols, alkylphenones, alkyl benzoates, polycyclic
aromatic hydrocarbons, halogenated benzenes, etc., were related to the log P
values calculated according to Rekker’s method. The average difference was
’’
within 0.16 log P unit. The correlation coefficient was 0.975. However, the
peak for aniline showed tailing. An organic modifier effect on logP was
described, and a method for the qualitative analysis of free fatty acids was
’’
proposed. 39 8920

Prediction of Retention Timesfrom log P

The prediction of retention times in a given eluent from log P has been proposed
for aromatic hydrocarbons.’’ The log k values of phenols2’ and nitrogen-
containing compounds22 were also related to their logP, and the calculated
log P was used for the qualitative analysis of urinary aromatic acids, i.e. for the
identification of metabolites in urine from the differences of l o g P in reversed-
phase liquid chromatography.23724
A good correlation was obtained in 2&80% acetonitrile-water mixtures. The
standard non-ionic compounds used to evaluate the columns were 2-hydroxy-
acetophenone, coumarin, acetophenone, indole, propiophenone, butyro-
phenone, isopropyl benzoate, butyl benzoate, and isopentyl benzoate. The
plotted lines for the linear relationship measured in five different proportions

-2 0 2 4 6 1

logP by Rekker’s
Figure 6.1 Relationship of Rekker’s log P values to log k values on octadecyl-bonded silica
gel column at 30°C. Column, Chromosorb LC7; eluent, 60% aqueous
acetonitrile. Number beside symbols: see Table 6.4.
112 Chapter 6
of acetonitrile-water mixtures merged at a single point. These slopes ( y ) have a
fourth-degree polynomial relation with the acetonitrile concentration ( x )
(Equation 6.3, where ai is a contant):
i=n
y= Ca;x xi n = 2-8
i=O

Further, an optimization of the organic modifier concentration for the

separation of known compounds was proposed. When the value of the actual
column plate number, the resolution, and log P of pairs of compounds a and b
were known, the percentage concentration (x)required for their separation was
calculated from Equations 6.4 to 6.6:'9720

R, = [(a- l)/a]O.25& (6-4)

The difference between the logk values measured and those predicted by
Equations 6.1 and 6.2 from log P values calculated by Rekker's hydrophobic
fragmental constant was within 5%. An example of logk values obtained
experimentally compared with those predicted from Rekker's log P values is
shown in Figure 6.2, where the predicted and measured retention factors of
aromatic acids are shown. The correlation coefficient was 0.993 (n = 32), and
the slope was 1.014. The precision is very high.

50
L
0
c,
0 40
(P
rc
C
.-0
c, 30
c
Q)
c,
Q)
L
20
3
c,

.-
U
0
10
!!
n
0
0 10 20 30 40 !D

Measured retention factor

Figure 6.2 Comparison of measured and predicted retention factors of aromatic acids
from Rekker's log P values. Column, 10 pm polystyrene gel, Hitachi 3011,
15 cm x 4.6 mm i.d.; eluent, 30% aqueous acetonitrile containing 50 mM
phosphoric acid at 55 "C. Numbers beside symbols: see Table 6.4.
Influence of Physical Chemistry on Separations in Liquid Chromatography 113

Prediction of Retention Time of Ionic Compoundsfrom log P and

PK,
The retention times of neutral non-ionized compounds can be predicted by the
above calculation method, but those of ionized compounds are also very
important. The inclusion of the dissociation constant in the calculation made it
possible to predict the retention times of ionized aromatic The
dissociation constant was calculated by the method proposed by Perrin et
When a compound is ionized, the retention becomes shorter than that of the
molecular form, and the retention factor of the ionized form is obtained from
Equation 6.7:

where km is the retention factor of the non-ionized acid and can be obtained
from log P as above and ki is the retention factor of the 100% ionized acid. The
ki value cannot be predicted mathematically at present, but the value is close to
zero in many cases.
Equation 6.7 was further modified to improve the precision in low pH
solutions where the strong acid used for pH control reduced the retention of
weak acids by excluding them from the hydrophobic surface of the stationary
phase.26 A modified version of Equation 6.7 is given in the following form (a
slight modification of constant A improved the precision):

An example of the relationship between the predicted and observed log k values
at pH 4.00 is shown in Figure 6.3.
The agreement between the observed and predicted k values of aromatic acids
was within 10%. The correlation coefficient was 0.954 ( n = 32). An error of
greater than 10% for 3-hydroxy-2-naphthoic acid and 2-hydroxybenzoic acid
was attributed mainly to an error in their Ka values.25The partition coefficient,
logP, and dissociation constant, pKa, of analytes can be obtained by simple
calculations and by computational chemical calculations, and thus the retention
time can be predicted in reversed-phase liquid chromatography.

Calculation of Dissociation Constant p K , from Hammet’s Equation

The dissociation constant of an analyte can be calculated mathematically from
Hammet’s equation.27 The organic solvent effect on the pKa has also been
examined:26
114 Chapter 6

20 3

measured retention factor at pH 4.00

Figure 6.3 Comparison of measured and predicted retention factors of aromatic acids
from Rekker's log P values. Column, 5 pm polystyrene gel, Hitachi 3013,
15 cm x 4.6 mm i.d.; eluent, 20% aqueous acetonitrile containing 50 mM
sodium phosphate at 55 "C. Numbers beside symbols: see Table 6.4.

where A and B are constants for individual groups of compounds, and Q is

Hammet's Q constant. However, it is difficult to predict the pK, values of ortho-
substituted compounds. Examples of pK,s from modified Hammet's equations
are:25

Benzoic acids: 4.20-1 .OO C Q

2-Hydroxybenzoic acids: 4.20-1.13 C Q
Phenylacetic acids: 4.30-0.35 Q c
Cinnamic acids: 4.38-0.66 C CT
Mandelic acids: 3.38-0.454 C CT
The original equations for phenolic and nitrogen-containing compounds are:27

Phenolic Compounds: 9.92-2.23 C G

Anilines 4.58-2.28 Qc
Pyridines (1): 5.25-5.90 C CT
Pyridines (2): 5.39-5.70 C Q
Quinolines (2-substituted) 5.12-9.04 Y 0
Quinolines (8-substituted) 4.64-3.1 1 0
a-Naph t h ylamines: 3.85-2.81 C Q
P-Naph t h y lamines: 4.29-2.8 1 C Q
Benzylamines (ring substituted): 9.39-1.05 Q
N-Substituted anilines 5.06-3.46 C CT
1-Aminoanthracene: aniline + 0.17
Quinoline: pyridine (1) + 0.06

where Q = Hammet's sigma constant. Examples of G constants are given in

Influence of Physical Chemistry on Separations in Liquid Chromatography 115
Table 6.2 Hammet’s substituent constants

Substituent nmeta Upara Substituent urnera npara

Br 0.39 0.22 CH3 -0.06 -0.14

c1 0.37 0.24 OCH3 0.11 -0.28
H 0.00 0.00 NH2 0.00 -0.57
I 0.35 0.15 COzH 0.35 0.44
OH 0.13 - 0.38 C2H5 -0.07 -0.15

From Ref. 27.

Table 6.2. The calculated values are summarized in Table 6.4 (along with
reference values).

2 Van der Waals Volume as the Basic Property

In a reversed-phase liquid chromatography system, if the hydrogen-bonding
and Coulombic forces are negligible, the retention of molecules depends upon
their size, and the presence of 71 electrons enhances the selectivity. The molecular
size, i.e. the van der Waals volume, can be calculated by hand, by Bondi’s
method,29 or by certain computer programs.
The possibility for predicting the retention time of polycyclic aromatic
hydrocarbons from a combination of the calculated van der Waals volume and
the reference values of the delocalization energy has been d e m ~ n s t r a t e dThe
.~~
precision of a method for predicting the retention times of polycyclic aromatic
hydrocarbons in reversed-phase liquid chromatography has been improved by a
combination of the van der Waals volume calculated by the computer program
MOPAC-BlogP and the kinetic energy calculated by molecular dynamic^.^'
However, the precision for polychlorinated benzenes could not be improved by
this method.31 The prediction of retention times of phenols was further
improved by using the ionization potential calculated by the computational
chemical program CACheTM.

Calculation of van der Waals Volume

Bondi’s calculation method is simple. The van der Waals volumes are the sum of
the van der Waals volumes of fragments, as given in Table 6.3. The calculated
van der Waals volumes are summarized in Table 6.4. However, these volumes
are different from those calculated using the MOPAC-BlogP program, even
though the correlation is excellent.

Prediction of Retention Times Based on van der Waals Volumes

The relationship between the log k and the van der Waals volumes of alkanes,
alkylbenzenes, and polycyclic aromatic hydrocarbons measured in 70%
aqueous acetonitrile on octadecyl-bonded silica gels3’ is shown in Figure 6.4.
116 Chapter 6
Table 6.3 Bondi's group contributions to the van der Waals
volume of hydrocarbons

Groups VWV" Groups VWV"

Alkane
C 3.33 6.96
CH 6.78
CH2 10.23 k=C/ 10.02
13.67 / \
CH3
CH4 17.12 =CH- 8.47
=CH2 11.94

Aromatic \C=CH2 16.95

/

\-(cond.)
/ 4.74 \C=CH- 13.49
/

\-R 5.54 Acetylenic

/

\-H 8.05 8.05

/
Benzene 48.36 rC-H 11.55
Phenyl 45.84 EC-( diacet y lene) 7.82

" Van der Waals volume (cm3 mol- ') from Ref. 28.

1.5 -
I 4 0 6 2 :x 3 9 g7
v 0 2SA
O O X
1 OA
0 0

Y A
93
- 0.5
A
I Id
3 32
A

-0.5 ! 1 1

25 75 125 175

van der Waals volume by Bondi cm3/mole

Figure 6.4 Relationship of log k values measured on ODs-bonded silica gel to Bondi's van
der Waals volumes. Column, Develosil O D s , 15 cm x 4.6 mm i.d.; eluent,
70% aqueous acetonitrile at 30 "C. Numbers beside symbols: see Table 6.4:
0 , Polycyclic aromatic hydrocarbons; x , alkylbenzenes; 0, halogenated
benzenes; A,alkanols; and 0, alkanes.
Table 6.4 Molecular properties of non-ionic compounds %ff
No. Chemical Mr VWV" SAb IF Slog@ -AHe VWVf logp -AHh 5
1 Pentane 72.150 95.970 136.780 11.060 3.395 5175 58.03 2.994 2.301 d
2 Hexane 86.177 112.787 158.420 11.060 3.972 6048 68.26 3.524 2.367 $
3 Heptane 100.203 129.552 180.714 11.060 4.570 6919 78.49 4.054 2.957 fi'
4 Octane 114.230 146.408 202.272 11.060 5.139 7794 88.72 4.584 3.372
5 Nonane 128.257 163.232 224.205 11.064 5.698 8668 - - -
6 Decane 142.284 180.007 246.467 11.062 6.267 9539 109.18 -
5.644
i3r
- - - g
7 Undecane 156.311 196.800 267.965 11.063 6.738 10411
8 Dodecane 170.337 21 3.653 289.660 11.063 7.209 11286 129.64 6.704 - <
0
9 Tridecane 184.364 230.555 3 11.976 11.032 7.648 12164 - - -
10 Benzene 78.1 13 83.789 109.752 9.652 2.432 3402 43.86 2.280 1.484 3
11 Naphthalene 128.173 127.598 156.999 8.71 1 3.729 4897 73.96 3.208 1.926 5
12 Acenaphthylene 152.195 144.969 173.766 8.943 4.274 5359 - - - a
13 Acenaphthene 154.211 150.675 181.904 8.492 4.297 5668 - - -
g.
2i
@
14 Biphenyl 154.211 155.227 189.990 8.950 4.316 6237 90.08 3.772 2.034
15 Fluorene 166.222 161.639 194.923 8.717 4.545 6145 93.22 3.906 2.235 ''
16 Phenanthrene 178.233 171.286 20 1.409 8.617 4.799 65 19 99.56 4.378 2.476 &
17 Anthracene 178.233 171.486 204.540 8.123 4.808 6539 99.56 4.378 2.543 5
18 Pyrene 202.255 187.603 216.623 8.130 5.246 6847 109.04 5.030 2.885 0
19 Fluorancene 202.255 188.723 220.619 8.629 5.248 699 1 - - - F
20 Chrysene 228.293 214.558 244.822 8.371 5.581 820 1 125.16 5.545 3.207
21 Tetracene 228.293 215.352 252.509 7.747 5.634 8733 125.16 5.790 3.442 5
22 Benzo[a]anthracene 228.293 215.841 25 1.956 7.938 5.554 8470 - - -
23 p-Terphenyl 230.309 226.657 273.024 8.746 5.710 9205 - I -
5
%
24 Benzo[e]pyrene 252.3 15 230.588 256.615 8.214 5.899 8595 - - - c
25 Benzopyrene 252.3 15 23 1.048 260.079 7.918 5.912 8571 134.64 6.200 3.700
26 Benzo[k]fluoranthene 252.3 15 232.418 268.140 8.302 5.962 8669 - - -
27 Benzo[b]fluorancene 252.315 232.462 267.348 8.559 5.941 8663 - - -
c
(continued) Z
 c
w
Table 6.4 Continued 00
~~~ ~ ~

No. Chemical Mr VWY" SAb Ip" Slog@ -AHe VWVf 1ogP -AHh

28 Benzo[ghi]perylene 276.337 247.198 272.98 1 8.032 6.188 8900

278.353 259.435 296.969 8.030 6.06 1 10099 - -
29 Dibenzo[ah]anthracene
30 Pentacene 278.353 259.277 299.886 7.49 1 6.197 9885 150.76 6.960
276.337 248.683 279.958 8.138 6.197 904 1 - -
3 1 Indeno[ 1,2,3-cd]pyrene
32 Toluene 92.140 100.456 131.549 9.333 2.778 4510 59.51 2.588 1.658
33 Ethylbenzene 106.167 1 17.067 152.079 9.298 3.189 5217 69.74 3.1 18 1.794
34 Isopropylbenzene 120.194 134.032 172.293 9.388 3.614 6159 77.02 3.525 1.906
35 Propylbenzene 120.194 133.897 173.691 9.297 3.613 6107 79.97 3.648 2.053
36 Butylbenzene 134.221 150.713 195.232 9.298 4.052 7002 90.20 4.178 2.331
148.247 167.564 217.077 9.301 4.496 7847 - - -
37 Pentylbenzene
38 Hexylbenzene 162.274 184.315 239.063 9.317 4.933 8715 110.66 5.238 2.797
39 Heptylbenzene 176.301 201.171 260.952 9.300 5.357 9580 120.89 5.758 3.200
40 Octylbenzene 190.328 217.884 283.590 9.299 5.764 10458 131.12 6.297 3.673
41 Nonylbenzene 204.355 234.706 305.2 18 9.299 6.1 19 11305 141.35 3.453 6.828
42 Decylbenzene 218.381 25 1.540 326.927 9.301 6.433 12116 151.58 3.655
43 Acetophenone 120.151 119.403 153.029 9.934 1.936 5546 - 1.75
44 Propiophenone 134.177 136.200 174.031 9.925 2.346 6453 2.28
45 Butyrophenone 148.204 153.052 196.058 9.925 2.768 7313 2.81
46 Pentanophenone 162.231 169.817 2 17.968 9.922 3.197 8153
47 Hexanophenone 176.258 186.582 240.137 9.926 3.625 9035
48 Heptanophenone 190.285 203.4 15 261.523 9.924 4.009 9915
49 Octanophenone 204.311 220.180 283.668 9.924 4.393 10772
50 Nonanophenone 218.388 236.945 305.856 9.924 4.75 1 11640
5 1 Decanophenone 232.365 253.710 327.255 9.924 5.035 12507
246.392 270.484 349.826 9.925 5.344 13371 - -
52 Undecanophenone
53 Chlorobenzene 112.559 97.300 126.293 9.560 2.826 4047 57.84 2.808 1.656 0
54 1,2-Dichlorobenzene 147.004 110.741 141.224 9.602 3.295 4744 67.32 3.532 1.884
55 1,3-Dichlorobenzene 147.004 110.911 142.517 9.686 3.324 4744 67.32 3.532 1.950 3
56 1,4-Dichlorobenzene 147.004 110.844 142.627 9.523 3.393 4745 67.32 3.532 1.944 q
57 1,2,3-Trichlorobenzene 181.449 124.174 156.074 9.785 3.771 545 1 76.80 4.197
58 1,2,4-TrichIorobenzene 181.449 124.307 157.548 9.653 3.828 5466 76.80 4.157
59 1,3,5-Trichlorobenzene 181.449 124.505 158.784 9.921 3.889 5480 76.80 4.197
60 1,2,3,4-Tetrachlorobenzene 2 15.894 137.542 170.685 9.734 4.256 6180 86.28 4.944
6 1 1,2,3,5-Tetrachlorobenzene 21 5.894 137.770 172.135 9.763 4.305 6208 86.28 4.944
62 1,2,4,5-Tetrachlorobenzene 215.894 137.730 172.368 9.656 4.352 6209 86.28 4.944
63 Pentachlorobenzene 250.339 150.870 185.175 9.786 4.736 6938 95.76 5.691
64 Hexachlorobenzene 284.784 164.037 197.770 9.91 1 5.176 7667 105.24 6.438
65 2-Chlorotoluene 126.585 114.010 146.978 9.415 3.236 5079 - 3.40
66 3-Chlorotoluene 126.585 114.119 148.096 9.427 3.21 1 5217 - 3.40
67 4-Chlorotoluene 126.585 114.140 148.082 9.297 3.206 5217 - 3.40
68 2-Bromotoluene 171.036 121.894 153.533 9.473 3.606 522 1 - 3.61
69 4-Iodotoluene 218.037 129.514 162.180 9.410 3.918 5432 - 3.93
70 1-Chloro-4-iodobenzene 238.455 126.392 156.666 9.621 4.110 5005 -
4.06
7 1 1-Bromo-3-chlorobenzene 191.455 118.912 149.925 9.714 3.693 4915 - 3.74
72 2-Bromo-4-chloroacetophenone 233.492 154.618 194.224 9.914 3.232 7367 - 2.57
73 Bromobenzene 157.010 105.336 133.720 9.604 3.217 4208 - 3.017
74 Iodobenzene 204.0 10 112.716 140.247 9.645 3.590 4292 - 3.334
75 Methyl benzoate 136.150 128.120 164.986 10.020 1.969 6205 - 2.15
76 Isopropyl benzoate 164.204 161.823 206.896 9.977 2.776 7976 - 3.09
77 Butyl benzoate 178.230 78.537 229.265 9.997 3.228 8719 - 3.74
78 Isopentyl benzoate 192.257 95.268 248.769 9.991 3.561 9603 - 4.15
79 p-Xylene 106.167 17.182 153.244 9.059 3.224 5684 - 3.10
80 1,4-Dimethoxybenzene 138.170 34.452 173.357 8.869 1.804 6645 - 2.23
8 1 Phenethyl acetate 150.177 44.920 187.287 9.493 2.377 7293 - 2.36
82 Benzyl ethyl ether 122.166 25.900 163.363 8.961 2.190 5910 - 2.64
83 2-Methylnaphthalene 142.200 44.195 178.866 8.617 4.068 6013 - 3.70
84 2,4-Dimethylnaphthalene 156.227 160.782 198.828 8.503 4.429 7056 - 4.20
85 1-Chloronaphthalene 162.618 140.961 171.558 8.781 4.105 5619 - 3.80
86 Propanol 60.096 70.540 104.025 10.847 0.362 4222 - 0.27 1
87 Butanol 74.122 87.399 125.838 10.845 0.922 5042 52.40 0.801
 c
h,
Table 6.4 Continued 0

No. Chemical Mr VWV" SAh IP' BlogP" -AHe VWVf logP AH^
88 Pentanol 88.149 104.174 147.423 10.851 1.497 591 1 62.63 1.331
89 Hexanol 102.176 121.084 169.7 12 10.846 2.093 6786 72.86 1.861
90 Heptanol 116.203 137.862 191.712 10.851 2.670 766 1 83.09 2.391
91 Octanol 130.230 154.627 213.447 10.849 3.259 8535 93.32 2.921
92 Nonanol 144.256 171.395 234.930 10.851 3.791 9406 - -
93 Decanol 158.283 188.472 257.55 1 10.845 4.347 10279 113.78 3.98 1
94 Undecanol 172.310 204.994 278.645 10.852 4.800 11157 - -
95 Dodecanol 186.337 222.059 301.238 10.847 5.280 12020 134.24 5.041
96 Tridecanol 200.364 238.703 322.333 10.853 5.634 12880 - -
97 Tetradecanol 214.390 255.670 345.123 10.848 6.046 13760 154.70 6.701
98 Pentadecanol 228.417 272.009 365.820 10.852 6.294 14636 - -
99 Hexadecanol 242.444 289.335 389.083 10.846 6.603 15515 175.16 7.101
100 Benzyl alcohol 108.140 108.505 141.102 9.362 1.848 5145 - 0.93
153.475 -
101 Cinnamanol 120.15 1 119.356 8.588 2.038 5769 1.49
102 Indole 117.150 112.880 142.179 8.405 1.602 4705 - 2.06
103 Coumarin 146.145 127.931 158.467 9.461 1.665 5403 - 1.57
9.948 1.212 5462 -
104 Benzamide 121.138 114.245 148.148 0.78
105 Salicylamide 137.138 121.914 155.779 9.509 0.718 6029 - 1.28
-
106 Phenacetin 179.2 18 173.529 222.040 8.348 1.566 5403 0.88
121.421 9.972 - -
107 Uracil 112.088 90.666 - 0.588 4565
108 Caffeine 194.193 167.556 210.514 8.964 0.326 9068 - - 0.07
109 Nitrobenzene 123.111 106.298 137.585 10.557 1.987 4876 - 1.81
1 10 1-Fluoro-2,4-dinitrobenzene 186.099 132.899 169.714 11.264 1.685 6945 - 1.67
11 1 Benzyl methyl ketone 120.151 119.549 153.022 9.938 1.938 5518 - 1.75

"van der Waals volume (A3 mol-'); "surface area $A2) and 'partition coefficient are calculated by MOPAC-BlogP from CACheTM;'ionization potential (eV) and 0
'enthalpy are calculated by MOPAC from CAChe M;-fvander Waals volume calculated by Bondi's method; partition coefficient calculated by Rekker's method;
" enthalpy measured by liquid chromatography.
Table 6.4B Molecular properties of acidic compounds
rb
. 3
No. Chemical w VWV" SAb IP' BlogP" -AH' VWVJ logF -AHh pK,' 2
112 Phenol 94.1 13 91.738 120.577 9.1 14 1.295 4115 53.88 1.54 1.980 10.02 $
113 2-Methylphenol 108.140 108.327 140.162 8.959 1.699 5157 65.03 2.05 1.683 10.32 3
114 3-Methylphenol 108.140 108.432 142.265 9.041 1.706 5238 65.03 2.05 - 10.09 ff'
115 4-Methylphenol 108.140 108.447 142.192 8.88 1 1.703 5304 65.03 2.05 1.910 10.27
116 2,3-Dimethylphenol 122.166 124.810 160.190 8.924 2.163 6237 76.18 2.49 1.926 10.54 2'
117 2,4-Dimethylphenol 122.166 125.003 162.149 8.78 1 2.162 6298 76.18 2.49 1.872 10.60 3
118 2,5-Dimethylphenol 122.166 125.035 162.359 8.896 2.141 6312 76.18 2.49 1.880 10.41
119 2,6-Dimethylphenol 122.166 124.8 16 159.997 8.887 2.138 6316 76.18 2.49 1.941 10.63
120 3,4-Dimethylphenol 122.166 125.229 161.752 8.803 2.146 9186 76.18 2.49 1.906 10.36 9
121 3,5-Dimethylphenol 122.166 125.339 163.956 8.975 2.139 6442 76.18 2.49 1.889 10.19 3
122 2,3,4-Trime thy lphenol 136.193 141.221 178.530 8.726 2.562 7158 - - - -
123 2,3,5-Trimethylphenol -
3
136.193 141.587 181.784 8.853 2.578 7307 87.33 3.02 1.947 - 2.
124 2,3,6-Trimethylphenol 136.930 141.094 179.555 8.785 2.568 7222 87.33 3.02 1.940
125 2,4,6-Trirnethylphenol 136.193 141.518 182.671 8.783 2.597 7433 87.22 3.02 1.898 10.88 .:
126 2,3,4,5-Tetramethylphenol 150.220 157.430 197.433 8.633 2.921 8220 - - - -
127 2,3,5,6-Te tramethylphenol - t-
150.220 157.741 197.874 8.763 2.922 8330 98.48 3.55 1.960 $'
128 Pen tamet hy lphenol 164.247 173.376 2 12.220 8.609 3.205 9126 - - - - 5
129 2-Ethylphenol 122.166 124.897 160.104 8.998 2.099 5922 75.26 2.58 1.877 10.2
130 3-Ethylphenol 122.166 124.979 162.093 9.017 2.096 5984 75.26 2.58 1.838 9.9
131 4-Ethylphenol 122.166 125.084 161.793 8.862 2.118 5975 75.26 2.58 1.886 10.0 $
132 2-Chlorophenol 128.558 105.235 136.109 9.259 1.743 4775 63.03 2.27 1.899 8.48 2
133 3-Chlorophenol 128.558 105.249 136.77 1 9.335 1.731 4808 63.03 2.27 2.118 9.02
134 4-Chlorophenol 128.558 105.220 137.032 9.125 1.773 4810 63.03 2.27 2.085 9.38
135 2,3-Dichlorophenol 163.003 118.717 150.720 9.430 2.203 5560 72.51 2.93 2.06 1 7.45 5
136 2,4-Dichlorophenol 163.003 118.954 152.192 9.230 2.252 5521 72.51 2.93 2.029 7.89
137 2,5-Dichlorophenol 163.003 118.947 152.222 9.303 2.304 5512 72.51 2.93 2.03 1 7.50
138 2,6-Dichlorophenol 163.003 118.595 149.937 9.374 2.278 5548 72.51 2.93 2.032 6.79

(continued) !2
-
 w
h)
Table 6.4B Continued w

No. Chemical Mr VWV" SAh IP" Slog@ -AHe VWVf logp -AHh pK:

139 3,4-Dichlorophenol 163.003 118.826 152.197 9.280 2.221 5514 72.51 2.93 2.133 8.39
140 3,5-Dichlorophenol 163.003 118.965 153.141 9.538 2.276 5548 72.51 2.93 2.209 8.18
141 2,3,4-Trichlorophenol 197.448 132.121 165.377 9.380 2.684 6244 81.99 3.68 2.083 7.59
142 2,3,5-TrichlorophenoI 197.448 133.301 167.060 9.484 2.740 6251 81.99 3.68 2.374 7.23
143 2,3,6-Trichlorophenol 197.448 131.961 164.831 9.535 2.735 6293 81.99 3.68 2.093 6.12
144 2,4,5-TrichlorophenoI 197.448 132.324 167.370 9.322 2.775 6282 81.99 3.68 2.157 7.33
145 2,4,6-Trichlorophenol 197.448 132.193 166.307 9.390 2.802 6286 81.99 3.68 - 6.42
146 3,4,5-Trichlorophenol 197.448 132.273 166.399 9.422 2.722 6243 81.99 3.68 2.335 7.74
147 2,3,4,5-Tetrachlorophenol 23 1.893 145.518 179.888 9.457 3.170 6974 91.47 4.43 2.470 6.96
148 2,3,5,6-Tetrachlorophenol 23 1.893 145.328 179.604 9.629 3.238 6978 91.47 4.43 2.25 1 5.44
149 Pentachlorophenol 266.338 158.497 192.012 9.574 3.640 7748 100.95 5.17 2.439 5.26
150 2-Chloro-5-methylphenol 142.585 121.556 154.549 9.127 2.140 5860 74.51 2.71 - -
151 4-Chloro-2-methylphenol 142.585 121.654 155.526 9.023 2.176 5891 74.51 2.71 - -
152 4-Chloro-3-methylphenol 142.585 122.025 157.213 9.03 1 2.189 5852 74.51 2.71 - -
153 2-Bromophenol 173.009 112.895 141.882 9.301 2.197 4938 66.48 2.48 - 8.44
154 3-Bromophenol 173.009 113.321 144.179 9.348 2.229 4971 66.48 2.48 2.039 9.03
155 4-Bromophenol 173.009 113.358 144.363 9.190 2.148 4954 66.48 2.48 2.024 9.36
156 2,4-Dibromophenol 25 1.905 133.715 162.464 9.461 3.001 5852 79.08 3.35 2.021 7.8
157 2,6-Dibromophenol 25 1.905 134.337 164.099 9.444 2.956 5831 79.08 3.35 - 6.6
158 2-Nitrophenol 139.110 113.337 146.167 9.952 1.644 5640 67.11 1.92 - 7.23
159 3-Nitrophenol 139.110 114.417 148.348 9.948 1.571 5655 68.16 1.27 - 8.40
160 4-Nitrophenol 139.110 114.403 147.925 10.072 1.723 5584 68.16 1.27 - 7.15
161 2,4-Dinitrophenol 184.108 136.132 172.035 10.757 1.221 7105 81.39 1.58 - 4.09
162 2,5-Dinitrophenol 184.108 136.159 171.927 10.613 1.420 7105 81.39 1.58 - 5.22
163 2,6-Dinitrophenol 184.108 135.574 169.758 10.656 1.167 7172 81.39 1.58 - 3.71
164 3,4-Dinitrophenol 184.108 137.036 175.184 10.718 0.849 7167 82.44 0.93 - 5.43 9
165 2-Hydroxyacetophenone 136.150 126.970 161.167 9.319 1.461 6265 - 1.21 - -
166 4-t-Butylphenol 150.220 158.057 198.996 8.893 2.887 7599 - 3.60 - - ez
167 Prowl 4-hvdroxvbenzoate 180.203 169.667 219.136 9.5 16 2.488 8719 - 2.67 - - rn
168 Butyl4-hydroxybenzoate 194.230 186.310 240.792 9.520 2.930 958 1 3.20
169 4-Chloro-3,5-dimethylphenol 156.612 138.449 177.464 8.892 2.624 689 1 3.24
170 1,2-Dihydroxybenzene 110.112 99.793 130.523 8.796 0.878 5004 1.oo
171 1,3-Dihydroxybenzene 110.112 99.888 131.163 9.044 0.853 4836 1.oo
172 1,4-Dihydroxybenzene 110.112 99.852 131.027 8.736 0.878 4915 1.oo
173 1-Hydroxynaphthalene 144.173 135.452 166.326 8.465 2.618 5637 2.86
174 2-Hydroxynaphthalene 144.173 135.586 167.716 8.643 2.612 5613 2.86
175 1-Hydroxy-2,4-dinitronaphthalene 234.168 178.767 212.8 16 9.792 2.173 8717 - 2.45 -
176 Benzoic acid 122.123 110.724 142.853 10.083 1.612 5159 65.36 1.94 2.378 4.200 9
177 3,4,5-Trihydroxybenzoic acid 170.121 134.292 171.289 9.372 0.756 7595 0.72 - 4.320 2
178 3,4-Dihydroxybenzoic acid 154.122 126.668 163.122 9.230 0.992 6773 0.99 4.450 5
179 3,5-Dihydroxybenzoic acid 154.122 126.699 164.064 9.407 0.968 6681 0.94 3.940 .2
180 2,4-Dihydroxybenzoic acid 154.122 126.388 161.686 9.505 0.765 6652 1.30 3.250 g
181 4-Hydroxybenzoic acid 138.123 118.751 153.441 9.605 1.260 5874 1.28 4.580
182 3-Hydroxybenzoic acid 138.123 118.740 153.334 9.519 1.361 5916 1.37 4.070 %
183 2-Hydroxybenzoic acid 138.123 118.396 151.204 9.506 1.146 5967 2.18 2.821
184 4-Hydroxy-3-methoxybenzoic acid 168.149 143.959 184.413 9.100 1.501 7643 1.31 4.470 2.
185 3-Methoxybenzoic acid 152.149 135.989 174.473 9.391 1.825 6799 1.99 4.090 E
186 4-Methoxybenzoic acid 152.149 136.002 174.655 9.479 1.688 6790 - 1.86 - 4.480 8'
187 2-Methylbenzoic acid 136.150 126.868 160.859 9.728 2.038 6095 76.51 2.297 2.571 - ?t
188 3-Methylbenzoic acid 136.150 127.432 164.572 9.747 2.020 6343 76.51 2.297 2.465 4.270 2:
189 4-Methylbenzoic acid 136.150 127.336 164.514 9.824 2.014 6325 76.51 2.297 2.428 4.340 Q
190 2-Ethylbenzoic acid 150.177 - - 0
143.994 182.946 9.78 1 2.390 71 12 86.74
191 3-Ethylbenzoic acid 150.177 143.930 184.475 9.707 2.422 7042 86.74 - - - T
192 4-Ethylbenzoic acid 150.177 143.963 184.244 9.784 2.413 7042 86.74 2.827 2.539 4.350 2
193 2,4-Dimethylbenzoic acid 150.177 143.412 181.614 9.636 2.423 7255 87.66 2.623 - s
m
194 2,5-Dimethylbenzoic acid 150.177 143.428 181.896 9.446 2.448 7254 87.66 2.623 -
195 2,6-Dimethylbenzoic acid 150.177 143.334 179.269 9.520 2.429 71 15 87.66 2.273 - 4
196 3,4-Dimethylbenzoic acid 150.177 144.135 186.656 9.636 2.472 7294 87.66 2.486 4.440
.1"
197 3,5-Dimethylbenzoic acid 150.177 144.135 186.056 9.581 2.472 7518 87.66 2.513 4.340
198 2,4,6-Trimethylbenzoic acid 164.204 159.835 200.682 9.488 2.808 8298 98.81 2.393

(continued) E
 c.
h)
Table 6.4B Continued P

No. Chemical Mr VWV" SAb IF BlogP' -AHe VWVf logP -AHh pKai

199 2-Chlorobenzoic acid 156.568 124.325 159.776 9.93 1 1.991 5930 74.84 - 2.566 3.380
200 3-Chlorobenzoic acid 156.568 124.254 159.033 9.946 2.135 5870 74.84 - 2.704 3.960
201 4-Chlorobenzoic acid 156.568 124.250 159.022 10.023 2.143 5864 74.84 2.5 17 2.718 3.960
202 2,4-Dichlorobenzoic acid 191.013 137.827 175.267 10.066 2.538 6647 84.32 - 2.862 -
203 2,5-Dichlorobenzoic acid 191.013 137.947 176.071 9.847 2.495 6666 84.32 - 2.918 -
204 2,6-Dichlorobenzoic acid 191.013 137.710 174.850 9.97 1 2.398 6674 84.32 - 2.619 -
205 3,4-Dichlorobenzoic acid 191.013 137.721 173.886 9.989 2.655 6591 84.32 - 3.008 3.600
206 3,5-Dichlorobenzoic acid 191.013 137.913 175.409 10.010 2.607 6609 84.32 - 2.964 3.460
207 2-Bromobenzoic acid 201.019 132.145 166.909 9.952 2.300 6122 77.96 - 2.677 -
208 3-Bromobenzoic acid 201.019 132.23 166.492 9.969 2.491 6045 77.96 - 2.797 3.810
209 4-Bromobenzoic acid 136.150 132.209 166.737 9.824 2.488 6047 77.96 2.726 2.828 3.980
2 10 Phenylacetic acid 136.150 127.527 163.937 9.682 2.037 6010 79.55 1.94 2.428 4.300
21 1 2-Methylphenylacetic acid 150.177 143.730 182.111 9.476 2.413 7073 86.74 - 2.441 -
2 12 3-Methylphenylacetic acid 150.177 144.013 184.65 9.443 2.443 7188 86.74 - 2.486 4.325
2 13 4-Methylphenylacetic acid 150.177 144.143 185.405 9.383 2.470 7184 86.74 - - -
2 14 2-Ethylphenylacetic acid 164.204 160.255 201.187 9.436 2.776 7776 - - - -
2 15 3-Ethylphenylacetic acid 164.204 160.509 204.385 9.361 2.846 8390 - - - -
2 16 4-Ethylphenylacetic acid 164.204 160.492 204.073 9.310 2.838 7885 - - - -
217 2-Chlorophenylacetic acid 170.595 140.569 177.038 9.643 2.564 6751 85.07 - 2.599 -
218 3-Chlorophenylacetic acid 170.595 140.759 178.739 9.622 2.568 7124 85.07 - - -
219 4-Chlorophenylacetic acid 170.595 140.631 178.664 9.518 2.495 7124 85.07 2.473 2.777 4,216
220 3,4-Dihydroxyphenylaceticacid 168.149 143.030 181.183 8.890 1.498 7667 - 0.98 - 4.387
221 2,5-Dihydroxyphenylaceticacid 168.149 143.039 182.608 8.810 1.360 7620 - 0.80 - 3.964
222 2-Hydroxyphenylacetic acid 152.149 134.959 171.653 9.195 1.726 6783 - 1.47 - 4.225
223 4-Hydroxyphenylacetic acid 152.149 135.192 173.179 9.106 1.807 6788 - 1.26 - 4.433
224 4-Hydroxy-3-methoxyphenylacetic acid 182.176 160.492 203.920 8.752 1.872 8539 - 1.18 - 4.394 n
225 3-Methoxyphenylacetic acid 166.176 152.512 194.122 9.1 11 2.111 7675 - 1.99 - 4.261
- -
226 4-Methoxyphenylacetic acid 166.176 152.489 193.786 8.992 2.253 7709 1.95 4.394 3
227 3,4-Dihydroxymandelic acid 184.148 151.882 195.496 8.937 1.145 8638 - 0.144 - 3.493
228 Mandelic acid 152.149 135.783 173.022 9.991 1.752 6650 80.18 - 1.650 3.380 3
229 4-Hydroxy-3-methoxymandelic acid 198.175 168.774 2 13.274 8.784 1.551 9369 - 0.84 - 3.507
230 3-Methoxymandelic acid 182.176 160.809 203.41 1 9.088 1.892 8558 - 1.33 - 3.330 $
23 1 3,4-Dihydroxycinnamic acid 180.160 154.457 197.932 8.947 1.762 8310 - 1.15 - 4.545
232 4-Hydroxycinnamic acid 164.160 146.507 188.227 9.1 12 2.016 7416 - 1.46 - 4.630 %
233 trans-Cinnamic acid 148.161 138.435 177.359 9.472 2.333 6676 82.32 2.32 2.710 4.380
234 3-Methoxycinnamic acid 178.187 163.810 209.030 9.2 16 2.539 8326 - 2.37 - 4.307
235 4-Methyl( trans)cinnamic acid 162.188 155.116 198.854 9.259 2.751 7826 - - - -
s.s
236 4-Hydroxy-3-methoxycinnamic acid 194.187 171.801 219.02 1 8.854 2.230 9159 - 1.51 - 4.558 9
237 4-Methylcinnamic acid 162.188 155.207 199.577 9.253 2.765 7851 93.47 - 2.789 4.492 2
238 4-Methoxy-3-methoxycinnamic acid 208.213 189.048 239.78 1 8.755 2.616 10039 - 2.063 - 4.673 5
239 Phenylpropionic acid 150.177 144.183 185.755 9.520 2.454 6948 86.04 2.250 2.601 4.579 2
240 Phenylbutyric acid 164.204 160.924 207.898 9.434 2.890 7853 96.05 2.522 2.609 4.719 9
24 1 Pheny lpen tanoic acid 178.230 177.872 229.559 9.41 1 3.327 8727 - - - -
242 3-Indoleacetic acid 161.160 139.965 176.493 8.825 0.864 6419 91.65 1.92 3.033 4.590 $
243 3-Indolepropionic acid 175.187 156.451 195.771 8.445 1.380 7416 114.11 - 3.28 1 - e
244 3-Indolebutyric acid 189.213 175.527 219.961 8.431 1.604 8407 121.34 - 3.388 - 2.
245 Hippuric acid 179.175 158.769 203.669 10.020 - - 3:
1.573 8184 96.15 1.518
246 2-Hydroxyhippuric acid 195.174 166.550 212.370 9.569 1.038 8802 - 1.55 - 3.506 S
247 Uric acid 168.112 123.569 159.147 9.165 - 1.064 6557 - - - -
b
9
248 2-Naphthoic acid 172.183 154.456 190.262 9.063 2.820 6766 - - - -
249 3-Hydroxy-2-naphthoic acid 188.182 162.354 199.135 8.903 2.311 7575 - - - - ;i:
A

2
“van der Waals volume (A’ mol-’); ’surface area $A2) and “partition coefficient are calculated by MOPAC-BlogP from CACheTM;‘ionization potential (eV) and 2
’enthalpy are calculated by MOPAC from CAChe M;.fvander Waals volume calculated by Bondi’s method; partition coefficient calculated by Rekker’s method;
“enthalpy measured by liquid chromatography; ‘ dissociation constant.
’2-
126 Chapter 6
The ionization potential calculated by MOPAC seemed to be capable of
correcting the log k values of polychlorobenzenes when alkylbenzenes were
used as standards, as shown in Figure 6.5 where the ionization potentials of
alkyl and polychlorobenzenes are shown.
The relationship between the van der Waals volume ( V ) and the logk values
of the alkylbenzenes is given by:

logk=Ax V + B (6.9)

The constants A and B are related to the percentage concentration of

acetonitrile (x), and their values are:

A = 1.1503 x + 4.9525 x 10-5x - 1 .Of325 x 10-6x2 (r2 = 1 .OOO)

B = 1.305 - 3.2936 x lO-'x + 1.2900 x 10-4x2 (r2 = 1.000)

For hypothetical alkylbenzenes having the same van der Waals volume as
alkanes, polycyclic aromatic hydrocarbons, and polychlorobenzenes, the calcu-
lated logk values were determined by means of Equation 6.9 with constants A
and B obtained from the alkylbenzene results. The difference (A logk) between
the measured and the calculated values of logk in the same eluent was
determined.
The ionization potential energy, IP, of alkylbenzenes calculated by MOPAC
was used as the standard, and it was almost constant for a series of alkyl
compounds. The difference in the ionization potential, AIP, of other com-
pounds from that of alkylbenzenes was calculated. The relationship between
A log k and AIP is given by:
Alogk = CAIP +D (6.10)

10.0 1 1

J."' ' I . I . I . 1

50 100 150 200 250

van der Waals volume A3/ mole

Figure 6.5 Relationship of ionization potential to van der Waals volume. Compounds:
W , alkylbenzenes; and , polychlorobenzenes. Number beside symbols: see
Table 6.4.
Influence of Physical Chemistry on Separations in Liquid Chromatography 127
where C and D are constants in the same eluent. The relationship between C, D,
and the acetonitrile concentration (x)was obtained from

C = 0.266 + 2.826 x 10-3x - 3;640 x 10-5x2 (? = 0.999) (6.1 1)

D = -0.1 13 + 2.177 x 10A3x- 8.798 x 10A6x2 (r2 = 0.989) (6.12)

The log k of alkanes, polycyclic aromatic hydrocarbons, and polychloro-

benzenes was therefore estimated from Equations 6.9 and 6.10 in 60-90%
aqueous acetonitrile. The calculated equation is:
log kcalc = A X + B + CAIP + D (6.13)

where A , B, C, and D are constants. The correlation coefficient between the

measured and calculated retention factors was -0.922 (n = 27), 0.946 (n = 32),
0.966 (n = 3 9 , and 0.971 (n = 37) in 60, 70, 80, and 90% aqueous acetonitrile,
respectively.
This method was applied to the prediction of the retention factor of phenols.
First, this method requires a relationship between logk and the van der Waals
volume of a homologous series of alkyl compounds such as alkylbenzenes or
alkylphenones. When alkylbenzenes are used as the standard, the value of B in
Equation 6.8 should be altered to give a parallel relationship at a suitable
position ( E )on the y-axis for a different group of compounds. When B is moved
to E for phenols, the logk values of para-alkylated phenols 01) are simply
predicted from their van der Waals volume by the following equation:

logkcalc = Ax +E (6.14)

For other substituted phenols, the difference in their ionization potential from
that of phenol was calculated, and then their Alogk was obtained from
Equation 6.10. The logk of substituted phenols was then estimated
from Equations 6.9 and 6.13.
The retention time of phenols was predicted in 70 and 60% acidic aqueous
acetonitrile on an ODS silica gel column. The constants A , B, C, and D were
obtained from the above equations. The result in 60% aqueous acetonitrile is
shown in Figure 6.6. The correlation coefficients between the measured and
predicted retention factors of substituted phenols in 60 and 70% acidic aqueous
acetonitrile were 0.974 (n = 36) and 0.967 (n = 36), respectively. In this system,
the values of the slopes, which indicate the relationship between the measured
and predicted retention factors, were 0.81 and 0.94 in 60 and 70% acidic
aqueous acetonitrile, respective~y.~~
This system of using the van der Waals volume and ionization potential is
much simpler than previous systems. It requires only one homologous series of
alkyl compounds and one standard such as phenol for substituted phenols. The
precision could be improved for polar compounds if para-alkylphenols and
para-alkylbenzoic acids are used; however, only a few such compounds are
128 Chapter 6

149
0
148
0
Old7

112
O! I I I

0 2 4 6 8
measured retention factor
Figure 6.6 Comparison of measured and predicted retention factors of substitute phenols.
Column, 5 pm ODs-bonded silica gel, 15 cm x 4.6 m m i.d.; eluent, 60%
aqueous acetonitrile containing 0.05 M phosphoric acid. Number beside
symbols: see Table 6.4.

available at present. This system seems to be practical for estimating the

approximate retention time.

3 Enthalpy as a Thermodynamic Property in Retention

Studies
The chromatographic retention parameters used in correlation studies are
normally assumed to be proportional to the free energy change.
Adsorption is accompanied by a loss of surface free energy. Free energy
change is related to enthalpy, AH, and entropy, AS, changes by the Gibbs
equation:

AG = A H - TAS (6.15)

The retention factor is related to the thermodynamic equilibrium constant K for

solute binding by k = PK, where P is the phase ratio of the column. The free-
energy change for the chromatographic process is expressed by

AG = -RTlnK = -RTln(k/P) (6.16)

where R is a gas constant (8.314 J K-' mol-') (1 cal is 4.184 J). From
Equations 5.38 and 5.39:

Ink = -(dH/RT) + ( d S / R ) + In /? (6.17)

A van't Hoff plot of Ink against 1/T yields a straight line if the stationary phase
is not deformed. The relation Ink and - A H / R T is used to study the
tempera ture-dependen t selectivity of the stationary phase.33-3
Influence of Physical Chemistry on Separations in Liquid Chromatography 129
Measurement of Enthalpy by Liquid Chromatography
Enthalpy can be measured by liquid chromatography where enthalpy is a slope
of the relationship between Ink and the inverse value of the absolute tempera-
ture. A schematic diagram is shown in Figure 6.7. The slope depends upon the
solutes being retained by the same liquid chromatographic mechanism. An
example is given in Table 6.4. The results, measured on an octadecyl-bonded
vinyl alcohol copolymer gel, did not show a simple linear relationship. This is
due to a conformation change of the octadecyl-bonded vinyl alcohol copolymer
gel stationary phase material, which has a phase transition point at about 33 "C.

Enthalpy Related to Retention

The higher the enthalpy change ( A H ) , the longer the retention, and log k shows
a linear correlation to e n t h a l ~ yAs ~ in Figure 6.8, a number of groups of
. ~ seen
compounds exhibit such a linear relationship between enthalpy and log k,
although it is not perfect. The AH values were independent of their van der
Waals volumes, as seen in Figure 6.9.
The AAH value of a methylene unit of alkanols is fairly constant.36 However,
the AAH value of a methylene unit of alkylbenzenes was not constant and the
larger molecules had a higher AAH value than the average AAH value of a
methylene unit. This larger value may be due to dimerization between solutes as
described ~chematically,~~ and/or to a strong direct interaction between alkyl-
benzenes and the octadecyl group of the stationary phase material, because
alkylbenzenes with a longer alkyl group must lose their aromaticity.
The isomeric effect of the AAH value of substituted chlorobenzenes was very
small. This means that the molecular size of the isomers is almost the same;
however, the actual size is not the same under these chromatographic condi-
tion~.~~

AH = slope = b / a

1/ T
r e c i p r o c a l of a b s o l u t e t e m p e r a t u r e
Figure 6.7 Measurement of enthalpy using chromatography for the relationship between
absolute temperature and retention factor.
130 Chapter 6
4000
2 5 ~ 40

-Q)
0
9 5 B
0
ox4
-Ip
3000
64
Y
>I
-P
Ip 2000
5C
a 0 32
87 A 10
1OW
AA I I I

5 0 0.5 1 1.5

log k
Figure 6.8 Relationship of enthalpy to log k values. Column, O D s silica, ERC-ODs,
15 cm x 6.0 mm i.d.; eluent, 80% aqueous acetonitrile at 30°C. Numbers
beside symbols: see Table 6.4. 0 ,Polycyclic aromatic hydrocarbons; x ,alkyl-
benzenes; 0, halogenated benzenes; A, alkanols; and 0, alkanes.

4000 ,
40 @ 2 5
Q)
I
4 o <Ag5
2i
:
Q 3000-
Y

->ra **-
(II
5E
Q)
100 A
1000-
87AA
25 75 125 1 5

van der Waals volume cm3/mole

Figure 6.9 Relationship of enthalpy to Bondi's van der Waals volumes. Column, ODs
silica, ERC-ODs, 15 cm x 4.6 mm i.d.; eluent, 80% aqueous acetonitrile.
Numbers beside symbols: see Table 6.4. 0 ,Polycyclic aromatic hydrocarbons;
x ,alkylbenzenes; 0, halogenated benzenes; A,alkanols; and 1, alkanes.

The enthalpy of methylphenols was about 2.0 kcal mol-', and that of
'
chlorophenols varied from 2.0 to 2.4 kcal mol- in the case of pentachlorophe-
nol, indicating that the retention difference depended not upon the size but on
the n-electron density.39 A similar result was obtained for alkylated and
halogenated aromatic acids, whose enthalpies were nearly equal, but whose
retention factors were different.40 The AH values may depend on the type of
stationary phase used and the water content of the eluent.41
Influence of Physical Chemistry on Separations in Liquid Chromatography 131

The highest enthalpy was found on non-endcapped high carbon octadecyl-

bonded silica gels, and was slightly higher on high-carbon endcapped gels than
on low-carbon endcapped gels. There was a selectivity difference in the retention
between alkanols and other hydrophobic compounds that depended on the end-
capping treatment applied.42 The order of enthalpies of phenols showed no
direct relationship to the retention factor or hydrophobicity of octadecyl-
bonded silica gels with different carbon loadings.43 Further, enthalpy was not
directly related to the log k values nor to the alkyl chain length of bonded-silica
gels, although the retention behaviour of individual groups of compounds such
as polycyclic aromatic hydrocarbons, and especially the retention of smaller-
size polar compounds, appeared to indicate this.44

4 Conclusion
The qualitative analysis of retention behaviour in liquid chromatography has
now become possible. Quantitative retention-prediction is, however, still
difficult; the prediction of retention time and the optimization of separation
conditions based on physicochemical properties have not yet been completely
successful. One reason is the lack of an ideal stationary phase material. The
stationary phase material has to be stable as part of an instrument, and this is
very difficult to achieve in normal-phase liquid chromatography because the
moisture in organic solvents ages the silica gel.
Several improved stationary phase materials have been synthesized for
reversed-phase liquid chromatography. One material is vinyl alcohol copolymer
gel. This stationary phase is quite polar and chemically very stable; however, it
demonstrated a strong retention capacity for polycyclic aromatic hydrocar-
b o n ~ . Although
~ ’ ~ ~ stable octadecyl- and octyl-bonded silica gels have been
synthesized from pure silica and are now commercially available, such
an optimization system has not yet been built. Further experiments are required
to elucidate the retention mechanism, and to systematize it within the context of
instrumentation.

5 References
1. A. Leo, C. Hansch, and D. Elkins, Chern. Rev., 1971,71, 525.
2. R.F. Rekker, ‘The Hydrophobic Fragmental Constant’, Elsevier Scientific, Amster-
dam, 1977.
3. W.J. Haggerty and E.A. Murrill, Res. Dev., 1974, 25, 30.
4. R.M. Calson, J. Chromatogr., 1975, 107, 219.
5. J.M. McCall, J. Med. Chem., 1975, 18, 549.
6. M.S. Mirrless, S.T. Moulton, C.T. Murphy, and P.J. Taylor, J. Med. Chem., 1976,
19, 615.
7. K. Miyake and H. Terada, J. Chromatogr., 1978, 157, 386.
8. M.J. O’Hare and E.C. Nice, J. Chromatogr., 1979, 171, 209.
9. J. Grunenberg and R. Herges, J. Chem. If. Comput. Sci., 1995,35,905.
10. C. Yamagami, M. Yokora, and N. Takao, J. Chrornatogr., 1994,662,49.
11. A. Kakoulidou and R.F. Rekker, J . Chromatogr., 1984,295, 341.
12. M.J.M. Wells, C.R. Clark, and R.M. Patterson, J. Chromatogr. Sci., 1981, 19, 573.
132 Chapter 6
13. M.D. Amboise and T. Hanai, J. Liq. Chromatogr., 1982,5,229.
14. K. Calmmer, L.-E. Edholm, and B.E.F. Smith, J. Chromatogr., 1977,136,45.
15. J.F. Schabron, R.J. Hurtubise, and H.F. Silver, Anal. Chem., 1980,50, 19 1 1 .
16. N. Nomura, Toyama Univ. Kyoikugakubu-Kiyo, 1979,27, 1.
17. B.L. Karger, J.R. Gant, A. Harkopt, and P.H. Weiner, J. Chromatogr., 1976, 128,
65.
18. T. Hanai, Chromatographia, 1979, 12,77.
19. T. Hanai, K.C. Tran, and J. Hubert, HRC & CC, 1981,4,454.
20. T. Hanai, S. Hara, S. Mori, and T. Hanai (eds), ‘Chromatography, the Separation
System’, Maruzen, Tokyo, 1981, p. 138 (in Japanese).
21. T. Hanai and J. Hubert, HRC & CC, 1983,6,20.
22. T. Hanai and J. Hubert, J. Liq. Chromatogr., 1985,8,2463.
23. T. Hanai and J. Hubert, J. Chromatogr., 1982,239, 527.
24. T. Hanai and J. Hubert, HRC & CC, 1981,4,500.
25. T. Hanai, K.C. Tran, and J. Hubert, J . Chromatogr., 1982,239,385.
26. T. Hanai and J. Hubert, HRC & CC, 1984,7,524.
27. D.D. Perrin, B. Dempsey, and E.P. Serjeant, ‘pK, Prediction for Organic Acids and
Bases’, Chapman & Hall, London, 198 1 .
28. A. Bondi, J. Phys. Chem., 1964,68,441.
29. T . Hanai, J. Chromatogr., 1985,332, 189.
30. T. Hanai, H. Hatano, N. Nimura, and T. Kinoshita, J. Liq. Chromatogr., 1993, 16,
1453.
31. J. Yamaguchi, T. Hanai, and H. Cai, J. Chromatogr., 1988,441, 183.
32. T. Hanai, H. Hatano, N. Nimura, and T. Kinoshita, Analyst, 1994, 119, 1167.
33. W. Melander, D.E. Campbell, and C. Horvath, J. Chromatogr., 1978,158,215.
34. T. Paryjczak, ‘Gas Chromatography in Adsorption and Catalysis’, Ellis Horwood
Limited, New York, 1986, p. 231.
3 5. R. Kaliszan, ‘Quantitative Structure-C hromatograp hic Reten tion Rela tionships’,
Wiley, New York, 1987.
36. T. Hanai and J. Hubert, J. Chromatogr., 1984, 291, 81.
37. A. Ben-Neim, ‘Hydrophobic Interaction’, Plenum Press, New York, 1980.
38. T. Hanai, A. Jukurogi, and J. Hubert, Chromatographia, 1984,19,266.
39. Y. Arai, M. Hirukawa, and T. Hanai, J . Chromatogr., 1987,384,279.
40. Y. Arai, J. Yamaguchi, and T. Hanai, J. Chromatogr., 1987,400,21.
41. H. Colin, J.C. Diez-Mesa, G. Guiochon, T. Czaikowska, and I. Miedziak, J .
Chromatogr., 1978, 167,41.
42. J. Yamaguchi, T. Hanai, and H. Cai, J. Chromatogr., 1988,441, 183.
43. J. Yamaguchi and T. Hanai, J. Chromatogr. Sci., 1989,27, 710.
44. J. Yamaguchi and T. Hanai, Chromatographia, 1989,27, 371.
45. T. Hanai, Y. Arai, M. Hirukawa, K. Noguchi, and Y. Yanagihara, J. Chromatogr.,
1985,349,323.
46. Y. Arai, M. Hirukawa, and T. Hanai, Nihon Kagaku Kaishi, 1986,969.
47. Y. Arai, M. Hirukawa, and T. Hanai, J. Liq. Chromatogr., 1987, 10,635.
48. M. Hirukawa, Y. Arai, and T. Hanai, J. Chromatogr., 1987,395,481.
49. Y. Arai, M. Hirukawa, and T. Hanai, J. Chromatogr., 1987,400,27.
50. T. Hanai, M. Ohhira, and T. Tamura, L C & GC, 1988,6,922.
51. M. Ohhira, F. Ohmura, and T. Hanai, J . Liq. Chromatogr., 1989, 12, 1065.
Subject Index
Affinity liquid chromatography, 7,8 Heavy metal, 41
Amino acid, 67,74 Height equivalent to a theoretical plate
Asymmetry (As), 39 ( H ) , 3,100
Hold-up volume ( VM), 4
Blood serum, 51, 52, 53 Hydantoin, 5 1
Bondi, 116
Injector, 11
Carboxylic acid, 68 Ion-exchange, 7,8, 36
Catecholamine, 69 Ion-pair liquid chromatography, 7, 70,
Cell volume (flow cell), 25 76
Charge-transfer, 8 Ionic strength, 80
Chiral separation, 7, 8 Ionization potential, 117, 126
Chiration, 8
Chlorosilane, 35 Ligand-exchange, 8
Chromatographic system, 11 Longitudinal diffusion, 103
Column efficiency, 96, 100 log P,58, 110, 117
Column length (L),3, 100
Column packing, 37 Mobile phase, 103
Column permeability (KO), 107 Molecular mass (Mr),8, 117
Column switching, 17 Molecular property, 117
Column temperature, 18,80
Column test, 40 Normal-phase liquid chromatography, 7,
Connecting tubing (volume), 26 8, 81
Counter ion, 73, 79
Organic modifier, 57,60, 80
Dead volume ( VM), 4 Organic polymer, 49
Degassing, 17 Organic solvent, 57, 59,6 1, 87
Detector (detection), 19, 80
Detector sensitivity, 23 Particle size (dp),4, 102
Dissociation constant (K), 113 pH, 65,75
pK,, 66,79,113,121
Eddy diffusion, 102 Properties of solvent, 59
Effective theoretical plate number (N,K), Protein, 50, 54, 55, 68,69
4, 100
Eluent, 89 Quantitative structure retention
Enthalpy, 117, 128 relationship, 109
Entropy, 128
Rekker, 111
Gradient, 13 Recycle, 15
Resolution (Rs), 1,96, 101
Haemoglobin (hemoglobin), 70 Retention volume (V,), 4
Hamiltonian, 38 Reversed-phase ion-pair, 8
Hammet, 113 Reversed-phase, 7, 8, 57

133
134 Subject Index
Salting-out, 8 Theoretical plate number (N), 1,4,39,97,
Separation factor (a), 1,6, 100 100
Sieving, 37 Time constant, 23
Silica gel, 32,46 Trouble shooting and maintenance, 27, 82
Size-exclusion liquid chromatography, 7,
8,90 Van Deemter, 106
Solubility, 7, 89 Van der Waals, 115, 117
Solvent, 57, 59,61,85, 87 Van’t Hoff plot, 128
Stationary phase, 32,46, 103 Viscosity, 15
Void volume, 43