Capillary electrochromatography : fundamentals and applications · CAPILLARY ELECTROCHROMATOGRAPHY; FUNDAMENTALS AND APPLICATIONS PROEFSCHRIFT ter verkrijging van de graad van doctor

Capillary electrochromatography : fundamentals andapplicationsCitation for published version (APA):Jiskra, J. (2002). Capillary electrochromatography : fundamentals and applications. Eindhoven: TechnischeUniversiteit Eindhoven. https://doi.org/10.6100/IR558066

DOI:10.6100/IR558066

Document status and date:Published: 01/01/2002

Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne

Take down policyIf you believe that this document breaches copyright please contact us at:[email protected] details and we will investigate your claim.

Download date: 23. Apr. 2020

https://doi.org/10.6100/IR558066

https://doi.org/10.6100/IR558066

https://research.tue.nl/en/publications/capillary-electrochromatography--fundamentals-and-applications(b8d50d8e-ef64-4d78-8e11-b65ed6126c91).html

CAPILLARY ELECTROCHROMATOGRAPHY; FUNDAMENTALS AND APPLICATIONS

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de

Technische Universiteit Eindhoven, op gezag van de Rector

Magnificus, prof.dr. R.A. van Santen, voor een commissie

aangewezen door het College voor Promoties in het openbaar te

verdedigen op dinsdag 1 oktober 2002 om 16.00 uur

door

Jan Jiskra

geboren te Turnov, Tsjechië

Dit proefschrift is goedgekeurd door de promotoren: prof.dr.ir. C.A.M.G. Cramers en prof.dr. G.J. de Jong Copromotor: dr. H.A. Claessens

Jaroslav Seifert

Píseň

Bílým šátkem mává,

kdo se loučí,

každého dne se něco končí,

něco překrásného končí.

Poštovní holub křídly o vzduch

bije,

vraceje se domů;

s nadějí i bez naděje,

věčně se vracíme domů.

Šetři si slzy

a usměj se uplakanýma očima,

každého dne se něco počíná,

něco překrásného se počíná.

CIP-DATA LIBRARY TECHNISCHE UNIVERSITEIT EINDHOVEN Jiskra, Jan Capillary electrochromatography : fundamentals and applications / by Jan Jiskra. – Eindhoven : Technische Universiteit Eindhoven, 2002. Proefschrift. – ISBN 90-386-2584-7 NUR 967 Subject headings: capillary electrochromatography / CEC / reversed phase stationary phases / reversed phase stationary phases ; testing / CEC ; retention mechanism Bibliotheek Werktuigbouwkunde en Scheikundige Technologie Postbus 513, 5600 MB Eindhoven W-hal 0.01, tel. 040-2472555 © Copyright 2002, J. Jiskra Omslagontwerp: Jan-Willem Luiten, JWL Production, Eindhoven Druk: Universiteitsdrukkerij, TU Eindhoven

Contents - i -

CONTENTS

1 INTRODUCTION AND SCOPE ................................................................................................... 1

2 STATIONARY AND MOBILE PHASES IN CAPILLARY ELECTROCHROMATOGRAPHY ................................................................................................ 5

2.1 INTRODUCTION ..................................................................................................................................5 2.2 ELECTROOSMOTIC FLOW .................................................................................................................6 2.3 STATIONARY PHASES IN CEC ..........................................................................................................7 2.3.1 NORMAL PHASES...............................................................................................................................8 2.3.2 REVERSED PHASES ..........................................................................................................................13 2.3.3 PHASES WITH ENHANCED EOF .......................................................................................................26 2.3.4 PHASES WITH CHARGED GROUPS ....................................................................................................27 2.3.5 CHIRAL AND SPECIAL STATIONARY PHASES .................................................................................36 2.3.6 ORGANIC POLYMER BASED .............................................................................................................39 2.4 MOBILE PHASES...............................................................................................................................43 2.4.1 NON-AQUEOUS MOBILE PHASES .....................................................................................................46 2.5 CONCLUSIONS ..................................................................................................................................47

3 CHROMATOGRAPHIC PROPERTIES OF REVERSED PHASE STATIONARY PHASES UNDER PRESSURE AND ELECTRO DRIVEN CONDITIONS; EFFECT OF ORGANIC MODIFIER ..................................................................................................................................... 65

3.1 INTRODUCTION ................................................................................................................................65 3.2 EXPERIMENTAL ...............................................................................................................................67 3.2.1 COLUMNS.........................................................................................................................................67 3.2.2 INSTRUMENTATION .........................................................................................................................68 3.2.3 CHEMICALS......................................................................................................................................69 3.2.4 TEST PROCEDURE ............................................................................................................................69 3.3 RESULTS AND DISCUSSION ..............................................................................................................71 3.3.1 COLUMN HYDROPHOBICITY AND HYDROPHOBIC SELECTIVITY.....................................................71 3.3.2 SILANOL ACTIVITY ..........................................................................................................................78 3.4 CONCLUSIONS ..................................................................................................................................83

4 CHROMATOGRAPHIC PROPERTIES OF REVERSED PHASE STATIONARY PHASES UNDER PRESSURE AND ELECTRO DRIVEN CONDITIONS; EFFECT OF BUFFER COMPOSITION............................................................................................................................. 87

4.1 INTRODUCTION ................................................................................................................................88 4.2 EXPERIMENTAL ...............................................................................................................................89 4.2.1 COLUMNS.........................................................................................................................................89 4.2.2 INSTRUMENTATION .........................................................................................................................90 4.2.3 CHEMICALS......................................................................................................................................90 4.2.4 TEST PROCEDURE ............................................................................................................................92 4.3 RESULTS AND DISCUSSION ..............................................................................................................92 4.3.1 POLAR COMPOUNDS ........................................................................................................................92 4.3.2 APOLAR COMPOUNDS......................................................................................................................96 4.4 CONCLUSIONS ............................................................................................................................... 100

- ii - Contents

5 PREPARATION AND CHARACTERIZATION OF MONOLITHIC POLYMER COLUMNS FOR CAPILLARY ELECTROCHROMATOGRAPHY.............................................................105

5.1 INTRODUCTION ............................................................................................................................. 106 5.2 EXPERIMENTAL ............................................................................................................................ 107 5.2.1 CHEMICALS................................................................................................................................... 107 5.2.2 COLUMN PREPARATION................................................................................................................ 107 5.2.3 INSTRUMENTATION ...................................................................................................................... 109 5.3 RESULTS AND DISCUSSION ........................................................................................................... 110 5.3.1 COLUMN EFFICIENCY IN CEC ...................................................................................................... 110 5.3.2 SELECTIVITY AND RETENTION IN CEC........................................................................................ 112 5.3.3 COMPARISON BETWEEN HPLC AND CEC................................................................................... 120 5.3.4 POROSITY...................................................................................................................................... 124 5.3.5 REPRODUCIBILITY AND STABILITY.............................................................................................. 125 5.4 CONCLUSION................................................................................................................................. 127

6 QUANTITATIVE STRUCTURE RETENTION RELATIONSHIPS IN COMPARATIVE STUDIES OF BEHAVIOUR OF STATIONARY PHASES UNDER HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY AND CAPILLARY ELECTROCHROMATOGRAPHY CONDITIONS............................................................................................................................... 131

6.1 INTRODUCTION ............................................................................................................................. 132 6.2 EXPERIMENTAL ............................................................................................................................ 133 6.2.1 COLUMNS...................................................................................................................................... 133 6.2.2 INSTRUMENTATION ...................................................................................................................... 135 6.2.3 CHEMICALS................................................................................................................................... 135 6.2.4 TEST PROCEDURE ......................................................................................................................... 136 6.3 RESULTS AND DISCUSSION ........................................................................................................... 137 6.4 CONCLUSIONS ............................................................................................................................... 151

7 THERMODYNAMIC BEHAVIOUR IN CAPILLARY ELECTROCHROMATOGRAPHY .............................................................................................155

7.1 INTRODUCTION ............................................................................................................................. 155 7.2 EXPERIMENTAL ............................................................................................................................ 158 7.2.1 CHEMICALS................................................................................................................................... 158 7.2.2 COLUMNS...................................................................................................................................... 158 7.2.3 INSTRUMENTATION ...................................................................................................................... 159 7.2.4 TEST PROCEDURE ......................................................................................................................... 160 7.3 RESULTS AND DISCUSSION ........................................................................................................... 161 7.3.1 EFFECT OF TEMPERATURE ON THE ELECTROOSMOTIC FLOW ..................................................... 161 7.3.2 VAN ‘T HOFF PLOTS...................................................................................................................... 164 7.4 CONCLUSIONS ............................................................................................................................... 169

8 METHOD DEVELOPMENT FOR THE SEPARATION OF STEROIDS BY CAPILLARY ELECTROCHROMATOGRAPHY .............................................................................................173

8.1 INTRODUCTION ............................................................................................................................. 173 8.2 EXPERIMENTAL ............................................................................................................................ 175 8.2.1 CHEMICALS................................................................................................................................... 175 8.2.2 COLUMNS...................................................................................................................................... 176 8.2.3 INSTRUMENTATION ...................................................................................................................... 177 8.2.4 PREDICTION SOFTWARE ............................................................................................................... 177 8.3 RESULTS AND DISCUSSION ........................................................................................................... 178

Contents - iii -

8.3.1 PRELIMINARY EXPERIMENTS ....................................................................................................... 178 8.3.2 SELECTIVITIES .............................................................................................................................. 179 8.3.3 EFFECT OF ACETONITRILE COMPOSITION .................................................................................... 182 8.3.4 EFFECT OF PH OF TRIS BUFFER.................................................................................................... 185 8.3.5 EFFECT OF TRIS CONCENTRATION ............................................................................................... 185 8.3.6 EFFECT OF TEMPERATURE............................................................................................................ 188 8.3.7 EFFECT OF INJECTED PLUG WIDTH............................................................................................... 188 8.3.8 REPEATABILITY ............................................................................................................................ 192 8.3.9 DETECTION LIMITS ....................................................................................................................... 192 8.4 CONCLUSIONS ............................................................................................................................... 193

9 SEPARATION OF BASIC CENTRAL NERVOUS SYSTEM DRUGS BY CAPILLARY ELECTROCHROMATOGRAPHY .............................................................................................195

9.1 INTRODUCTION ............................................................................................................................. 195 9.2 EXPERIMENTAL ............................................................................................................................ 197 9.2.1 CHEMICALS................................................................................................................................... 197 9.2.2 COLUMNS...................................................................................................................................... 198 9.2.3 INSTRUMENTATION ...................................................................................................................... 199 9.3 RESULTS AND DISCUSSION ........................................................................................................... 199 9.3.1 PRELIMINARY EXPERIMENTS ....................................................................................................... 199 9.3.2 COLUMN AND MOBILE PHASE MODIFIER CHOICE ........................................................................ 201 9.3.3 HYPERSIL C8 MOS....................................................................................................................... 201 9.3.4 HYPERSIL PHENYL........................................................................................................................ 205 9.3.5 REPEATABILITY, INFLUENCE OF VARIABLES AND DETECTION LIMITS ...................................... 208 9.4 CONCLUSIONS ............................................................................................................................... 211

SUMMARY ..........................................................................................................................................213

SAMENVATTING..............................................................................................................................217

DANKWOORD...................................................................................................................................221

CURRICULUM VITAE......................................................................................................................223

BIBLIOGRAPHY................................................................................................................................225

- 1 -

CHAPTER 1 1 INTRODUCTION AND SCOPE

Capillary electrochromatography (CEC) is a separation technique in which the flow of the

mobile phase or buffer is driven through a chromatographic column by an electric field,

rather than by an applied pressure. As a consequence, it is a technique that combines the

separation and selectivity potential of high-performance liquid chromatography (HPLC)

and the high efficiency of capillary electrophoresis (CE), originating from the plug-like flow

profile that is inherent of this latter technique. The origin of the use of an electroosmotic

flow (EOF) in chromatography was already suggested in 1939 by Strain [1] using a

combination of electrophoretic and chromatographic adsorption methods for the

separation of a number of organic dyes on Tswett adsorption columns [2]. However, this

suggestion has not developed any further until 1974 when Pretorius and co-workers [3]

applied this concept for the use of an electroosmosic flow in liquid chromatography. At the

beginning of 1980s, technical developments in the field of the manufacturing capillaries

pushed electroseparation techniques forward for practical laboratory use. In spite of the

promising and interesting perspectives of CEC, until now a number of fundamental

questions have only been partly answered. In addition to that, especially the behaviour of

both stationary and mobile phases, as well as solutes under applied high-voltage conditions,

are major fundamental questions in CEC. Obviously, a knowledge of such behaviour is

important for the prediction and/or optimization of separations, method development and

also the transfer of separation protocols from related, well-established techniques such as

HPLC, to CEC. Moreover, too few convincing applications have been shown to justify the

wide spread introduction of CEC as a sound routine analysis technique. This Ph.D. thesis is

devoted to the study of the chromatographic / electrophoretic behaviour of mainly

reversed-phase (RP) stationary phases and eluents under pressure-driven (viz. HPLC) and

- 2 - Chapter 1

electro-driven (viz. CEC) conditions with special attention on the chromatographic

element.

Chapter 2 reviews the current status of CEC with a focus on the behaviour of stationary

and mobile phases under CEC conditions compared to the behaviour under HPLC

conditions. In this chapter, a number of applications is presented, too.

Chapter 3 studies the chromatographic properties such as the retention factor, silanol

activity and hydrophobicity of seven different RP stationary phases using different portions

of the organic modifier, acetonitrile or methanol in an aqueous solvent buffer. Applying a

number of testing methods, the differences in behaviour of the stationary phases under

CEC and HPLC conditions are demonstrated.

Chapter 4 further explores the chromatographic properties of RP stationary phases under

CEC conditions using different buffers and/or buffer compositions. Depending on the

type of buffer applied, substantial differences in the behaviour of the stationary phases

under CEC and HPLC conditions were observed.

Chapter 5 concentrates on the characterization of a new type of stationary phases viz.

monolithic polymers under CEC conditions. In this study it is demonstrated that under

CEC conditions this new type of stationary phase possesses attractive properties such as

high polarity and the presence of unique pores.

Chapter 6 applies quantitative structure retention relationship (QSRR) methods as a tool for

learning more about the separation mechanism under CEC conditions. It is shown that

QSRRs are capable of distinguishing between CEC and HPLC separation mechanisms.

Chapter 7 is devoted to the thermodynamic background of the separation of specific

analytes under CEC and HPLC conditions. Analyzing Van’t Hoff plots, differences in

separation mechanisms between CEC and HPLC modes are shown.

Chapters 8 and 9 demonstrate the systematic development of methods in CEC. In these

two studies the separations of steroid hormones and central nervous system drugs were

optimized.

- 3 -

REFERENCES

1. H.H. Strain, J. Am. Chem. Soc., 61 (1939) 1292.

2. M. Tswett, Ber. Dtsch. Bot. Ges., 24 (1906) 384.

3. V. Pretorius, B.J. Hopkins, J.D. Schieke, J. Chromatogr., 99 (1974) 23.

- 4 - Chapter 1

- 5 -

CHAPTER 2 2 STATIONARY AND MOBILE PHASES IN CAPILLARY

ELECTROCHROMATOGRAPHY

Summary

This review paper describes the state-of-the-art of capillary electrochromatography (CEC).

Properties of and interactions between stationary and mobile phases applied in CEC are

described and discussed; developments in stationary phases are reviewed. Special attention

is paid to the comparison of the behaviour of stationary and/or mobile phases under CEC

versus HPLC conditions with respect to variables such as particle and pore size of the

stationary phase, mobile phase composition, and temperature. These issues are discussed

throughout the paper. A number of applications in CEC is presented as well.

2.1 Introduction Capillary electrochromatography (CEC) is a chromatographic technique in which the

mobile phase is driven through the chromatographic bed by electroosmosis rather than by

pressure as applied in liquid chromatography (LC). In 1974 Pretorius [1] reported the use of

electroosmosis as a new concept for high-speed LC, making many research groups focus

their attention on CEC. The role of stationary phases, mobile phases and solutes

determining chrom atographic properties of CEC systems has been intensively investigated

over the past two decades. These efforts have resulted in new perspectives for this

This chapter has been submitted for publication in Journal of Separation Science.

- 6 - Chapter 2

technique and brought challenging solutions for many application problems. This review

provides a thorough overview of stationary and mobile phases for CEC, as well as in a view

on its relationship to the pressure-driven chromatographic technique. The influence of the

type of the stationary phase, mobile phase composition, applied voltage, and temperature

are discussed, together with questions about and possible answers to the chromatographic

relationship between CEC and HPLC. The chromatographic parameters of solutes such as

retention factor, selectivity, and efficiency, and the behaviour of the electroosmotic flow

and its influence on the separation properties are emphasized, too. A number of reports of

a fundamental nature [2-25], related to the status of CEC columns, column technology and

stationary phases [26-39] as well as application oriented articles [39-50, 415-416] have

shown a rapid development of this technique. Finally, approaches to new CEC stationary

phases, their future prospects and characteristics are discussed too.

2.2 Electroosmotic flow Electroosmotic flow (EOF) is the bulk flow of liquid in a capillary and is a consequence of

the surface charge on the interior capillary wall [51]. Obviously, stationary phases in CEC

possessing a much higher surface area than an interior capillary wall, are major contributors

to EOF in CEC too [52-56]. Such a charge originates from the ionization of the surface

silanols or any other ionizable groups, or from the adsorption of ionic species on the

surface. A double layer of electric charge is then formed by counter ions built-up on the

surface, which maintain the balance between the solution and surface charge. The voltage

drop between the charged surface and the counter ions at the plane of shear is known as

the zeta potential (ζ). Upon the application of voltage along a capillary, solvated cations or

anions in the diffuse layer migrate toward the cathode or anode (depending on the surface

charge), dragging the solvent molecules with them. The velocity of the movement (EOF) is

described by the equation below [57]:

η

ζεε Ev rEOF

0= (2.1)

where ε0 is the permittivity in vacuum, εr the relative permittivity, ζ the zeta potential, E the

applied electric field and η the solvent viscosity. The flow profile differs from the parabolic

Stationary and Mobile Phases in CEC - 7 -

flow in HPLC and was thoroughly studied already in 1965 by Rice et al. [58]. These authors

predicted that vEOF is nearly plug-like if the capillary (channel) diameter d>>δ is where δ is

the double layer thickness. Further calculations on this subject were performed by Knox

and Grant [59, 60] resulting in acceptable prediction values for the EOF in capillary tubes.

The influence of the thickness and the channel (pore) diameter on the chromatographic

performance will be discussed in detail in Section 2.3. Generally, the zeta potential, double

layer thickness and viscosity are parameters strongly influenced by a number of factors,

such as the nature of the surface and surface charge, mobile phase composition and

temperature.

2.3 Stationary Phases in CEC A schematic overview of stationary phases used in CEC is shown in Figure 2.1.

Normal Phases

Example: C8, C18, C30, polymeric phases

Conventional

Example: CEC Hypersil C18

Phases with Enhacened EOF

Example: Cation- or Anion-exchangers

Phases with Charged Groups

Example: fluorocarbon coated; chiral stationary phases: β-CD bonded, cholesteryl, vancomycin bonded

Special Phases

Reversed Phases

Silica Based

Acrylamide Gels

Polymethacrylates

Cellulose based

PEEK based

Organic Polymer Based

Stationary Phases in CEC

Figure 2.1

Overview of stationary phases used in CEC.

- 8 - Chapter 2

As can be seen, a variety of stationary phases is applied in CEC to investigate fundamental

problems in CEC and solve application problems. At the start of CEC, an attempt was

made to use conventional HPLC phases as stationary phases in CEC. However, this

approach failed in a number of cases because particular stationary phases were unable to

generate a substantial or stable EOF under the applied mobile phase composition. This

resulted in the development of new generations of stationary phases for CEC, where

advanced silanization procedures substantially reduced the number of free silanols needed

to generate EOF. A number of manufacturers and scientists attempted to develop

stationary phases that are more suitable for CEC. This included higher and more stable

EOF, column performance and column preparation. The following sections provide a

detailed overview of the stationary phases used in CEC.

2.3.1 Normal Phases Normal phases are by definition stationary phases that are more polar compared to the

apolar mobile phase. As in HPLC, also in CEC the number of applications under normal

phases conditions is limited. Lai and Dabek-Zlotorzynska [61] used normal phase CEC

mode to separate caffeine, theophylline and theobromine on silica using acetonitrile/Tris,

isopropanol/hexane/Tris and acetonitrile/isopropanol/hexane/ammonium acetate mobile

phases with an efficiency up to 63,000 plates/m. Bare silica has also been successfully used

in the separation of basic compounds.

Wei et al. [62] separated strong bases such as berberine and jatrorrhizine using ACN/Tris

mobile phase (Figure 2.2). The authors discussed the contribution of high silanol density of

bare silica, adsorption of counterions and influence of organic modifier properties (ε/η) to

the EOF. The effect of ionic strength (k vs. 1/csalt) and the higher retention of basics at

higher pH suggested an ion-exchange mechanism of separation. Using the same mobile

phase Gillott et al. [63] separated pharmaceutical bases. In spite of high efficiency,

irreproducible retention and splitting peaks were found. The use of competing bases such

as triethanolamine (TEA)-phosphate and TEA-trifluoroacetate may overcome these

problems. Under such conditions, acids, bases and neutrals can be simultaneously

separated.


Maruška et al. [65] used columns packed with Polygosil 100-10 silica (Macherey-Nagel

GmbH) using pure ACN, MeOH and MeOH / EtOH / hexane mobile phases to separate

non-polar to very polar compounds.

Figure 2.2

Electrochromatogram of the separation of seven basic drugs on a Micra silica stationary

phase. Experimental conditions: packed column, 27 cm (20 cm effective length) × 75 µm

I.D. (internal diameter); packing: silica dp=3 µm; mobile phase: CH3CN-10 mM TRIS-HCl

buffer (pH 8.29) (80/20, V/V). Solutes: 1, aniline; 2, cocaine hydrochloride; 3, berberine

hydrochloride; 4, thebaine; 5, jatrorrhizine hydrochloride; 6, ephedrine hydrochloride; 7,

codeine phosphate.

The authors used octadecylated cellulose, revealing normal and reversed-phase properties

depending on the solvents used. Furthermore, Grom-Sil silica (Grom Analytik+HPLC

GmbH) has used in fritless CEC experiments [66] or in experiments with on-line coupled

NMR [67]. Nucleosil silicas (Macherey-Nagel GmbH) with pore diameters of different

porosities were extensively investigated by a number of groups [68, 69, 400] to study the

fundamentals of CEC. The group of Venema [68, 400] studied pore-flow effects by size

exclusion electrochromatography using predominantly DMF (dimethylformamide)/LiBr

(lithium bromide) as a mobile phase. This group found significant pore flow under CEC

conditions (Figure 2.3).

- 10 - Chapter 2

Figure 2.3

Retention of polystyrene standards with A) Nucleosil 5 µm, 500 Å and B) Nucleosil 5 µm,

100 Å. Signs: A) Q: Pressure drive; ×: electro drive, mobile phase: DMF/0.1 mmol.L-1 LiBr;

∆: electro drive, mobile phase: DMF/0.5 mmol.L-1 LiBr; ◊: electro drive, mobile phase:

DMF/1 mmol.L-1 LiBr; field strength: 10 kV. Signs B) Q: Pressure drive; ×: electro drive,

mobile phase DMF/1 mmol.L-1 LiBr; ∆: electro drive, mobile phase: DMF/10 mmol.L-1

LiBr; ◊: electro drive, mobile phase: DMF/20 mmol.L-1 LiBr; field strength: 10 kV; τ,

exclusion coefficient; Log M, logarithm of polymer concentration.

Such pore flow increases intraparticle mass transfer and thus produces higher efficiencies

than can been achieved in LC. However, the disadvantage of a large pore flow is the

resulting smaller retention window under the size-exclusion condition. This can be

overcome by using stationary phases with smaller pores and is also more easily controlled

A)

B)


via the ionic strength of the eluent. Increasing the ionic strength causes larger pore flow,

and at low ionic strengths double layer overlap results in a decrease in the pore flow. A

comparison of through-pore flows for the Nucleosil silicas of 100 Å and 500 Å and mobile

phase strengths ranging from 0.1 mmol.L-1 of LiBr up to 20 mmol.L-1 of LiBr is given in

Figure 2.3.

In the same way Stol et al. [69, 70] characterized size exclusion systems in CEC on

LiChrosorb Si silica and Nucleosil silica using DMF/LiCl as the mobile phase. For the

prediction of the separation results the authors used two models describing the channel

system, parallel cylindrical channel with different diameters and cylindrical channels with

different diameters in series. They successfully predicted pore flow and exclusion limits

using these two models (Figure 2.4). Pressure and electro-driven flow has also been studied

by Witowski and Kennedy [71] with respect to fast chromatography. It has been shown for

both non-porous ODS and bare silica that efficiencies up to 370,000 plates/m can be

achieved in pressure-driven mode and from 670,000 to 1,000,000 plates/m in electro-driven

mode. Ye et al. [73] used bare silica dynamically coated with cetyltrimethylammonium

bromide (CTAB). A successful separation of anilines and peptides could be achieved. It is

also worth mentioning that due to a change in the amounts of adsorbed CTAB, non-linear

log k versus percentage of methanol was found in this particular case. Wei et al. [64] used

bare silica dynamically coated with hydroxypropyl-β-cyclodextrin (hydroxypropyl-β-CD). In

that paper, where CE was compared with CEC, it was found out that hydroxypropyl-β-CD

is adsorbed on bare silica resulting in a cation-exchange mechanism. This was obvious from

the resolution dependence on pH, ionic strength and an organic modifier (MeOH).

- 12 - Chapter 2

Figure 2.4

Relative retention of polystyrene standards in an electrically driven system on Lichrosorb

Si-100 silica stationary phase. (●): experimental values; (─): predictions with the parallel (a)

and series (b) models; (---): prediction for a pressure driven system. Mobile phase: (A)

DMF/1 mM LiCl; (B) DMF / 10 mM LiCl; τ, exclusion coefficient.


2.3.2 Reversed Phases

2.3.2.1 Conventional Reversed-phase Stationary Phases Due to the availability of many HPLC reversed-phase stationary phases, an obvious trend in

using these phases can be seen throughout the history of CEC. Moreover, the use of

conventional, non-endcapped stationary phases with high silanol activity is advantageous

since such phases provide a sufficient and stable EOF. A complete list of stationary phases

and references are listed in Table 2.1.

Table 2.1 Reversed-phase stationary phases applied in CEC.

Manufacturer Name/type of the reversed

phase

References

Hypersil C18

(incl. CEC phases)

52, 59, 60, 66, 74-132, 406-408, 417

Hypersil C8 78, 80-81, 83-85, 90, 93, 101, 133-

136

Hypersil ThermoQuest

Hypersil Phenyl 78, 81, 93, 95, 101, 133, 137

Spherisorb ODS (type I, II) 54, 81, 85, 98, 100-101, 117, 124,

130, 133, 138-161, 390, 407, 409-

410, 417

Waters

SymmetryShield 144

Macherey-Nagel Nucleosil (C18 and C8) 124, 126, 162-179. 398 (entrapped

in polymethacrylate)

NPS 61, 71, 101, 143-148, 180-189, 390 Micra Scientific

Synchropack 208

Merck LiChrosper, Purospher,

Monospher, Chromspher

60, 74, 95, 124, 163, 190-194, 239-

240, 411

Grom GromSil ODS 66-67, 106, 195-203

- 14 - Chapter 2

Table 2.1 continued Manufacturer Name/type of the reversed

phase

References

Nomura Chemical Co. Devosil 144, 153, 204, 205, 401

Jones Chromatography Apex ODS 209

SynChrom SynChrom 206-207

Hamilton Hamilton 164

Agilent Technologies Zorbax (ODS, C8) 124, 132, 150, 210-219

Vydac Vydac 220-223

C18 131, 226-235

C8 131

Unimicro

Phenyl 131, 236

Whatman Partisil 72, 124, 187

Yamamura Chemical

Company

YMC (ODS, C30) 60 (home made ODS), 173, 175,

237-238

VDS Optilab OptiLabSpher 163

Rainin Rainin ODS 265, 392

Shisedo Capcell 120SG ODS 385, 397

Organosilica Organosilica 224-225

Home made or non-

specified

- 60, 114, 180, 241-244, 264, 395

Continuous bed - 245-255, 379, 386

Open tubular (OT) and

etched

- 256-263, 378, 396, 399


This broad spectrum of reversed-phase stationary phases has been applied in numerous

applications and fundamental studies. Several important application areas can be

distinguished:

A) neutrals, acids, bases – standards

B) pharmaceuticals – antibiotics, barbiturates, steroids

C) environmentals

D) biomolecules – amino acids, peptides and proteins, saccharides.

In more fundamentally oriented research projects, basic questions about the behaviour of

stationary and mobile phases, and separation mechanisms, have been the focus of intensive

studies. Extensive studies on chromatographic behaviour in terms of retention, selectivity

and column stability were performed by Dittman et al. [52, 82-83, 315]. Typically, as in

HPLC, these authors found different selectivities for different reversed-phase stationary

phases under CEC conditions; similar observations were made for different organic

modifiers. Within the tested set of columns, a different EOF is generated. The highest EOF

observed by these authors was for the CEC Hypersil C18 stationary phase (Figure 2.5) and

Spherisorb C6/SCX (strong cation exchanger) stationary phase. This group also studied the

ratio of ε/η and its relationship to EOF. The inconsistency in this relationship is caused by

a change in the surface charge density and the adsorption of the ions on the surface,

resulting in changes in the double layer properties. Zimina et al. [124] suggested that the

EOF velocity is proportional to the surface area, so that stationary phases with larger

surface areas generate higher EOF unless deactivated (Figure 2.6). An example of a

deactivated RP LC stationary phase is BDS Hypersil ODS stationary phase. The

immobilization of the stationary phase in a column by heating techniques, for example, is

another major concern in CEC. Different selectivities and EOF velocities were observed by

Adam et al. [100] after the immobilization of a specific stationary phase in a similar set of

columns using a heating technique. Here, the stationary phase is immobilized by a heated

wire moving along the column, resulting in a satisfactory mechanical stability of the column.

The authors are aware of possible extensive heating of the column and an eventual loss of

hydrocarbonaceous chains from the RPLC-phase. This problem also is characteristic of the

majority of frit (in and/or outlet) preparation techniques. In addition, loss of

- 16 - Chapter 2

hydrocarbonaceous chains from frits prepared by fusing the stationary phase may cause

extra peak broadening and/or tailing due to adsorption. After frit preparation by fusing, the

groups of Carney [146], Behnke [200] and Chen [265] redeactivated frits again using

different agents such as chloro-dimethyloctadecylsilane or diphenyltetramethyldisilazane. As

a result, an improved baseline, less spikes and a significant reduction of adsorption of

dansyl-leucine derivative were observed. Obviously, this is related to the exposure of

silanols to the analytes causing unwanted interactions.

Figure 2.5

Separation of PAHs on five reversed-phase C18 stationary phases. Column 250(335) mm ×

0.1 mm, 3 µm, mobile phase: acetonitrile-50 mM Tris-HCl, pH 8 (80/20, V/V), voltage:

20 kV, temperature: 20°C, 10 bar pressure applied to both ends of capillary, 20°C. Samples

injected were not identical for depicted stationary phases but all contained thiourea (1),

naphthalene (2), and fluoranthene (3).

Lurie et al. [135] analyzed basic compounds on the CEC Hypersil C18 stationary phase

using 0.2% hexylamine as a mobile phase additive. Such additives adsorb on surface silanols

and consequently unwanted interactions between components and the silanol groups are

limited. A similar approach was also applied to Hypersil and Waters stationary phases by

Dittmann et al. [85] and Hilhorst et al. [417].


Figure 2.6

Relationship between electroosmotic flow and reported surface areas of stationary phases.

Stationary phases: Nucleosil 5 C18, LiChrospher RP-18, Spherisorb Diol, Zorbax BP ODS,

Spherisorb S5 ODS2, Hypersil ODS, mobile phase: acetonitrile-50 mM CAPSO buffer

pH 9.53 (70/30, V/V), EOF marker: thiourea.

Rue et al. [226] used pressurized gradient CEC for the analysis of eighteen amino-acid

derivatives on a Unimicro C18 stationary phase. With increasing voltage the resolution

increased, however, peaks disappeared at high voltage. The authors suggested that this was

due to adsorption of the components at the stationary phase. The same group investigated

the behaviour of the Unimicro C18 stationary phase under pressurized CEC using forward

and reversed pressure. In both cases, they found that the EOF was the dominant mobile

phase driving force. Eimer et al. [136] found that retention factors of hydrophobic analytes

were 36-40% lower in pressurized CEC compared to HPLC. They explained this by the

higher polarity of the stationary phase under the applied voltage (Figure 2.7).

- 18 - Chapter 2

Figure 2.7

Change of selectivity in capillary LC due to the voltage applied before a run (15 kV for 20

min). (1) Ethosuccinimide, (2) phenytoin, (3) pyrimidone, (4) carbamazepine-10,11-diol,

(5) carbamazepine-10,11-epoxid, (6) carbamazepine. Column: 100 mm × 0.1 mm

Spherisorb ODS-1, total length 260 mm, mobile phase: methanol-5 mM tetraborate buffer

pH 8.5 (60/40, V/V), voltage: -12 kV.

Furthermore, Ishizuka et al. [404] found lower k values of alkylbenzenes and aromatic

hydrocarbons on octadecylated silica rods (Figure 2.8) under CEC condition compared to


HPLC. The same authors found a smaller effect of k-values on the plate height in CEC

compared to HPLC.

Figure 2.8

Chromatograms obtained for alkylbenzenes (C6H5(CH2)nH, n = 0-6 (a, c)) and polyaromatic

hydrocarbons (b, d)) in pressure-driven (HPLC (a, b)) and electro-driven (CEC (c, d))

elution. Stationary phase: continuous macroporous silica gel (reversed-phase), mobile phase:

(a, b) 80% acetonitrile; (c, d) acetonitrile-Tris-HCl, 50 mM, pH 8 (80/20, V/V). Pressure:

(a, b) 0.9 kg/cm2. Applied voltage: (c, d) 750 V/cm.

The differences in the velocity along the streamlines of the EOF in the various parts of the

through-pores due to the plug-type flow profiles are claimed to be responsible for that

phenomenon. In addition, Jiskra et al. [131] also observed that the chromatographic

characteristics are dependent on whether a column is operated under HPLC or CEC

conditions. In that paper, linear relationships of log k versus percentage of methanol and

acetonitrile on a Unimicro C8 stationary phase for benzene as the test component were

presented and discussed (see Figure 3.4 in Chapter 3). Moreover, slopes of log k vs.

percentage of acetonitrile for both separation modes, HPLC and CEC, were significantly

different (Figure 3.4 A in Chapter 3). In addition to that, for methanol as the organic

modifier, different values of log kw (retention factor extrapolated to pure water as the

mobile phase) were found too for these two separation modes (Figure 3.4B in Chapter 3).

- 20 - Chapter 2

Moffat et al. [105] the studied analysis of pesticides by CEC and HPLC and found a non-

linear relationship of ln k vs. percentage of acetonitrile. However, CEC and HPLC analysis

profiles and retention factors were almost identical. Asiaie et al. [255] found linear

relationships for log k vs. percentage of acetonitrile for benzyl alcohol and benzaldehyde,

and the differences between CEC and HPLC were negligible. Jinno et al. [220] used

cholesteryl stationary phase and a Vydac C18 stationary phase in the analysis of

benzodiazepines; the authors assume the same chromatographic properties for both

stationary phases. In that study, plots of kCEC vs. kLC proved to be linear with the exception

of two compounds, cloxazolam and medazepam, for which a nonlinear kCEC vs. kLC

relationship was observed. Wen et al. [210] studied the dynamics of CEC and found a linear

correlation of retention factors under CEC and HPLC conditions on a Zorbax ODS, a

Spherisorb stationary phase (ODS and SCX) and also on a gigaporous polystyrene-

divinylbenzene (PS-DVB) column. These authors confirmed the existence of an

intraparticle flow under CEC conditions. In addition, these authors also studied and

compared parameters for the Van Deemter equation under pressure and electro-driven

conditions. Figure 2.9A outlines the A-term, representing the eddy diffusion under HPLC

and CEC conditions. Three phases were compared, a Spherisorb ODS 300 Å, a Spherisorb

SCX 300 Å and a polystyrene-divinylbenzene stationary phase PL-SCX 1000 Å (Polymer

Labs, Church Stretton, UK). In all these cases, A-terms were on average two to four times

lower under CEC conditions than for micro-HPLC. In Figure 2.9B, the intraparticle mass

transfer is shown (determining the magnitude of the C-term) under HPLC and CEC

condition. The authors also found linear van ‘t Hoff plots for the components, e.g.

acrylamide, benzaldehyde, naphthalene, biphenyl, fluorene and m-terphenyl, which is typical

for the majority of RP separations. Tang et al. [245] studied CEC monolithic columns where

particles of specific reversed-phase stationary phases (Spherisorb ODS1 and Nucleosil C18

1400 Å stationary phases) were entrapped in a continuous bed of silica. These columns

were used for the separation of corticosteroids, alkaloids and aromatic amines. As could be

expected, the high-pore Nucleosil C18 stationary phase exhibited significantly higher

efficiency than the Spherisorb ODS1 (80 Å) stationary phase. Furthermore, Chirica et al.

[169] found that the surface charge on a monolithic (entrapped) column is close to the same

as a conventionally packed (i.e. non-entrapped) column. The increase in the efficiency due

to pore flow was extensively studied by Stol et al. [162, 170] using Nucleosil C18 stationary

phases with different porosities.


The influence of the pore size and ionic strength on the experimentally obtained

efficiencies are depicted in Figures 2.10 and 2.11. As already mentioned in Section 2.3.1,

higher through-pore flow causes higher mass transfer and thus higher efficiencies. From

Figure 2.10 and 2.11 it can be seen that highest efficiencies occur in large pore stationary

phases and higher ionic strength mobile phases. A similar study was performed by Li et al.

[167] and Seifar et al. [74] on Nucleosil C18 stationary phases with different particle sizes.

Although the efficiency decreased from 3 µm > 5 µm > 7 µm particle sizes, no relationship

between EOF and particle size was found in that study. Banholzer et al. [165] employed the

Smoluchowski equation (Eq. 2.1) on a Nucleosil C18 stationary phase in the study of the

influence of mobile phase composition on behaviour in CEC. They found out that the

maximum velocity of EOF can be reached using the buffer (sodium phosphate)

concentration of 0.4 - 4 mmol.L-1, however, no correlation of the plate number on buffer

concentration could be found. It was concluded that the dependence of the velocity of

EOF on the buffer concentration could not be due to double-layer overlap effects. In

addition, no correlation of ε/η of the mobile phase with the velocity of EOF was found.

- 22 - Chapter 2

Figure 2.9

(A) Plots of parameter A against the buffer concentration in the µ-HPLC mode (∆) and the

CEC mode (○). Columns, (a) 21/29 cm×50 µm capillaries packed with 5 µm Spherisorb

ODS 300 Å; (b) 26/34 cm×50 µm capillaries packed with 5 µm Spherisorb SCX 300 Å;

(c) 34/42 cm×75 µm capillaries packed with 8 µm PL-SCX 1000 Å; eluents, (a) sodium

phosphate in water-acetonitrile mixture (1:1, V/V), (b-c) sodium-phosphate in water,

pH 7.0. (B) Artist's rendition of intraparticle mass transfer with, (a) viscous flow,

(b) electroosmotic flow. The slowness of mass transfer determines the magnitude of C-

term. In HPLC, transport of solutes is by diffusion only while in CEC, intraparticle EOF

augments transport between the interstitial fluid and the binding sites inside the porous

particles by convection. The circulating patterns inside particle symbolize that even in dead-

end pores EOF can enhance intraparticulate mass transport by convective mixing.

A)

B)


Figure 2.10

The effect of pore size of the three different stationary phases on the theoretical plate

height (H) of the fluorine peak against linear velocities (U) of the mobile phase. Columns

were approximately 33 cm long (25 cm effective). Mobile phase: acetonitrile-water (80:20,

V/V) containing 10 mmol.L-1 tetraborate, pH 8.3. Stationary phase: Nucleosil C18 7 µm

with porosities: (▲)=500 Å, (■)=1000 Å, (♦)=4000 Å.

Figure 2.11

The effect of tetraborate buffer concentration on the separation efficiency on Nucleosil

4000 7 µm C18. Mobile phase: acetonitrile-water (80/20, V/V) and buffer,

(♦)=0.1 mmol.L-1 tetraborate pH 8.3; (■)=1.0 mmol.L-1; (▲)=10 mmol.L-1.

- 24 - Chapter 2

Tallarek et al. [238, 316] extensively studied flow-field dynamics in pressure and electro-

driven systems and found a significant performance advantage of the electro-driven mode

for both open-tubular and packed capillary systems. These authors found that the dynamic

displacement time in electro-driven systems is significantly shorter than that in pressure-

driven systems (Figures 2.12 A, B).

Figure 2.12

(A) Axial displacement probability distribution, Pav(R,∆), of the fluid molecules near the

surface in a 0.65 m × 250 µm I.D. (360 µm O.D.) fused-silica capillary. (a) Electroosmotic

flow (E=23.1 kV/m, I=49 µA). (b) Pressure-driven flow. Stationary phase: 40 µm

rehydroxylated silica particles, mobile phase, borate buffer (2 × 10-3 M, pH 9.0); observation

time, ∆=14.2 ms; ambient temperature, 26±0.5°C.

(B) Reduced axial plate height (ha=Ha/dp) versus the reduced flow velocity (ν=dpuav/Dm) for

the CHPLC (capillary HPLC) and CEC modes.

A)

B)


Using two different stationary phases, YMC C18 and a Nucleosil C18, Pyell et al. [173]

studied band broadening in CEC using on-column injection and on-column detection. The

authors applied a mathematical peak shape and theoretical peak width mode to study their

relationship to injection plug lengths. A maximum tolerated injection plug length was

predicted:

NU

LLtIeo

T

µ×= 7.0max (2.2)

where L is the column length to the detection window, LT the total column length, µeo the

electroosmotic mobility, UI the injection voltage and N the plate number. Moreover, these

authors found a clear relationship between peak width and sample composition with

respect to the water content therein (Figure 2.13). Such zone sharpening is well known

from capillary electrophoresis. Stevens et al. [72] in 1983 studied flow profiles in CEC on

normal and reversed-phase stationary phases. The accurate flow profile from this reference

is depicted in Figure 2.14.

Extensive studies on retention mechanisms have also been performed by Wei [233] on a

Unimicro C18 stationary phase. The authors evaluated the retention behaviour of solutes

with solvatochromic parameters (LSER) and found the molecular volume (V),

dipolarity/polarizability (π), hydrogen bond acidity (α) and hydrogen bond basicity (β) of

the solute of equal importance in CEC. In contrast, V and β were found to be the most

significant parameters for the retention of solutes in HPLC. In that paper, the possible

distortion of the double layer by a strong electric field in CEC resulting in a different kind

of retention behaviour was discussed as well.

- 26 - Chapter 2

Figure 2.13

Dependence of the peak width at half height on the volume fraction of water in the sample

solution (injection: electrokinetic, 23 kV, 7 s; column: 345 mm (395 mm) ×180 µm; packing

octadecylsilica gel, dp=3.0 µm; mobile phase: ACN-phosphate buffer pH 7.3 (80/20, V/V);

in-column photometric detection, 230 nm; assignment: ▲=methyl benzoate, +=ethyl

benzoate, *=propyl benzoate, □=benzyl benzoate, ×=butyl benzoate, ◊=isopropyl

benzoate)

2.3.3 Phases with enhanced EOF In this group, specially designed stationary phases are adjusted for optimal performance

under CEC conditions. These phases are derived from conventional phases and may

provide a higher and more stable EOF due to a larger silanol activity. A typical example of

such a stationary phase is CEC Hypersil C18. The increased silanol activity and its

dependency on experimental conditions have been described by Jiskra et al. [131, 132].


Figure 2.14

Accurate representation of the electroosmotic flow velocity profile; δ, thickness of double

layer.

2.3.4 Phases with charged groups This group consists of reversed phases stationary phases that are additionally modified with

charged functional groups either directly on the silica surface or on the hydrocarbonaceous

chains.

2.3.4.1 Strong Cation Exchangers The great advantage of charged stationary phases (both cation and anion exchangers) is the

limited dependency of the EOF on the buffer pH as shown in Figure 2.15 [195]. Cikalo et

al. [268, 269] studied behaviour of the EOF and the field strength in open tubular and

packed capillaries.

- 28 - Chapter 2

Table 2.2 Ion-exchangers, silica based stationary phases applied in CEC.

A)

Manufacturer Name/type of the cation

exchanger (CX) phase

References

Spherisorb SCX (SCX, C3/SCX,

C6/SCX, C18/SCX)

52, 54, 80-81, 83, 100, 101, 139,

147, 149, 151, 158, 267-273,

408

Waters

Symmetry 149

Hypersil ThermoQuest Hypersil SCX, Duet 52, 80, 87, 97, 276

AllTech AllTech SCX 101

BioRad BioRad SCX 104

Tosoh - 389

Xtec Consultants - 161, 274-275

Home made Sulfonated and octadecylated 55, 277, 278, 317, 413

B)

Manufacturer Name/type of anion

exchanger (AX) phase

References

Waters Spherisorb SAX 280, 319

Hypersil ThermoQuest Hypersil SAX 87, 97

AllTech AllTech SAX 281

Xtec Consultants - 161, 275, 388

Home made +monolithic - 56, 282-286


Figure 2.15

Plots of electroosmotic flow versus mobile phase pH for sol-gel bonded continuous bed

columns. Conditions - column: 25/34 cm × 75 µm I.D. continuous bed columns

containing sol-gel bonded (□) 3 µm, 80 Å ODS1, and (∆) 3 µm, 80 Å ODS/SCX; mobile

phase: ACN/H2O/50 mM phosphate buffer (70/25:5, V/V/V); injection: 5 kV × 2 s;

applied voltage: 30 kV; EOF marker: 0.3 mM thiourea.

The contribution of the packing to the EOF is shown in Figure 2.16. The authors also

studied the field strength in the packed and open section of the capillary using different

portions of packed sections of a Spherisorb SCX stationary phase. They found that the field

strength remained similar at moderate eluent pH. For extreme eluent pH-value, however,

field strengths were larger. Differences in EOF along the capillary may explain bubble

formation in some CEC systems. Smith et al. [149] studied the contribution of charged

packing to the EOF and found that the contribution of walls to the EOF to be minor. This

finding is in agreement with results of Dittmann et al. [52]. Hilder et al. [87] prepared frits in

open tubular (OT) capillaries from ODS, SCX and SAX packing materials and found

increased EOF values for OT+frit compared to OT columns alone. The highest EOF was

achieved for SCX packing materials.

- 30 - Chapter 2

Figure 2.16

Effect of length of SCX packed bed on the EOF at pH 10.5 (●), 7.5 (□) and 2.9 (▲). µEOF

values calculated using the assumption that the voltage drop is (a) over the total length of

capillary and (b) over the packed section only. Conditions – column: 25/33 effective/total

length, 100 µm I.D. packed with Spherisorb SCX dp=3 µm; mobile phase: acetonitrile- 10

mM buffer (carbonate, phosphate or KCl-HCl) different pH (80/20, V/V); EOF marker:

thiourea; applied voltage: 10 kV.

Zhang et al. [55, 277, 278, 317] investigated sulfonated octadecylated silica. Strong and

constant EOF values over a wide range of pH were observed (Figure 2.17).


Figure 2.17

Effect of the pH of the mobile phase on the EOF. Conditions – column: 20.5 cm/27 cm

effective/total length, 100 µm I.D. packed with 10 µm particles of octadecyl, ODS (1),

octadecylsulfonated, ODSS (2) and sulfonated-ODSS (3) silica stationary phase; mobile

phase: ACN/1.25 mM sodium phosphate (75/25, V/V); applied voltage: 20 kV; EOF

marker: thiourea.

In addition to that, different elution patterns due to the permanently charged sublayer of

sulfonated hydrocarbonaceous chains were found (Figure 2.18) too. The authors discussed

three retention models:

1) Ion-pair

2) Dynamic ion exchange

3) Dynamic complex exchange

The authors concluded that the last model could best reproduce experimental observations.

Similarly, a strong cation exchanger on a silica support with polymeric layers carrying

sulfonic moieties was used by Wei [279] for the analysis of compounds containing nitrogen

(berberine, palmitine and jatrorrhizine). Good efficiencies and reproducibilities were

obtained but some analyte peaks showed significant tailing.

- 32 - Chapter 2

Figure 2.18

Electrochromatograms of a mixture of benzene and alkylbenzene homologous series

obtained on (a) ODSS, (b) sulfonated-ODSS, and (c) ODS. Capillary column, 20.5/27 cm ×

100 µm I.D., packed with 10 µm particles of ODSS, ODS or sulfonated ODSS. Mobile

phase: ACN-1.25 mM sodium phosphate (75/25, V/V); applied voltage: 20 kV; solutes:

1, benzene; 2, toluene; 3, ethylbenzene; 4, propylbenzene; 5, butylbenzene; 6, amylbenzene.

Very high efficiencies up to 6 milions plates/m have been observed by Smith and Evans

[147] for the analysis of steroids. The reproducibility of such efficient systems, however,

was complicated and strongly depended on the experimental conditions [406]. Ye et al. [318]

observed very good resolution using a Spherisorb SCX stationary phase, dynamically

modified with cetyl-trimethylammonium bromide (CTAB) rather than silica gel. The

authors presented the successful separation of acids, bases and neutrals.


2.3.4.2 Anion Exchangers A list of commonly applied anion exchangers can be found is in Table 2.2. Strong anion

exchanger stationary phases usually show a stable EOF from pH 2 to pH 6-8. As already

mentioned in the previous section, strong anion exchangers are used in most of the

investigations.

Figure 2.19

(A) The effect of buffer pH on the linear velocity of the EOF in 3 µm Waters Spherisorb

SAX packed columns, measured using thiourea and uracil. Conditions: Duplex column (●).

Mobile phase: acetonitrile-20 mM buffer (pH variable) (50/50, V/V). Applied voltage: -

20 kV. Column dimensions: 220 mm ×100 µm I.D. (total length 305 mm). Fully packed

column (■). Mobile phase: acetonitrile-20 mM ammonium acetate buffer (pH variable)

(50/50, V/V). Applied voltage: -20 kV. Column dimensions: 220 mm × 100 µm I.D. (total

column length 220 mm). (B) The effect of pH on linear velocity in untreated fused-silica,

PVA and amine coated capillary packed with 3 µm Waters Spherisorb SAX material.

Conditions: mobile phase: acetonitrile-20 mM sodium dihydrogenphosphate pH 2.5

(50/50, V/V). Applied voltage –25 kV. Injection –5 kV for 5 s. Column dimensions:

220 mm × 100 µm I.D. (total column length 320 mm)

A) B)

- 34 - Chapter 2

Byrne et al. [319] did a detailed study of the contribution of the capillary wall and stationary

phase to the EOF using a Spherisorb SAX stationary phase. The columns under study were

either fully or partially (viz. duplex) packed. The results are shown in Figure 2.19A. From

their data the influence of the open part of the column on the EOF is obvious. The wall

contribution was also compared for columns with the internal wall chemically derived by

amino groups or covered with polyvinylalcohol (PVA). For the sake of comparison, in

Figure 2.19B the results for a normal capillary is shown as a reference. Clearly, in this

experimental framework, the contribution of the capillary wall to the EOF is substantial.

The capillary coated with PVA showed nearly no EOF dependency on pH. The reverse is

true for the capillary coated with amine or for the normal, untreated capillary. However,

Scherer et al. [282] found that negative charges in an open tube part do not significantly

influence the EOF (Figure 2.20).

Figure 2.20

Influence of the capillary wall charges on the mobility of the EOF. Conditions – columns:

100 µm I.D. packed with TAM2 and TAMS3 connected to 50 µm I.D. detection capillaries

of different lengths; mobile phase: ACN-water-20 mM Tris pH 7 (70/20/10, V/V/V);

voltage: -25 kV; injection: -3 kV, 3 s; detection: UV, 210 nm; inert marker: thiourea.

Similarly, as also discussed in the previous section, anion exchange materials can be

dynamically coated. Ye et al. [320] used sulfonated β-cyclodextrins for the separation of

enantiomers of alkaloids and some important pharmaceuticals (Figure 2.21).


Figure 2.21

Chiral separation of tryptophan, atropine and verapamil in a single run by strong anion

exchanger-CEC dynamically modified with sulfated-β-cyclodextrin. Conditions: stationary

phase: Spherisorb SAX, dp=5 µm; applied voltage, 15 kV; mobile phase: methanol-2

mg/mL sulfated-β-cyclodextrin in 20 mM acetic acid-triethanolamine buffer (pH 4.0)

(30/70, V/V). Solutes: (1) L-tryptophan, (2) D-tryptophan, (3) and (4) atropine, (5) and (6)

verapamil.

- 36 - Chapter 2

2.3.5 Chiral and Special Stationary Phases Table 2.3 Chiral and other, silica based stationary phases applied in CEC.

Type Name/type of the phase References

Vancomycin bonded 287-290, 377

β-Cyclodextrin bonded 293-300, 391, 417

Quinine bonded 179, 292

Teicoplanin bonded 301

Cellulose derivative 302-304, 314

Amino acid bonded 304-305

Chiral polymer bonded 306-307

Naproxen bonded, Whelk-O 308, 381

Chiral stationary

phases (CSP)

Cholesteryl 220

Antibody 380

Fluorinated 309

Polymer coated 310-312

Special Purpose

Ionenene coated 313

Obviously, many tailor-made stationary phases include chiral stationary phases with a

physically or chemically bonded chiral selector. The combination of high efficiency and

enantioselectivity in CEC is very promising for this type of separations. However, the

preparation of a chiral stationary phase, the experimental conditions and the equilibration

of a column are difficult in many cases. Therefore, a number of authors use an

enantioselector in the mobile phase, rather than covalently bonded on the stationary phase.

Moreover, the most common solvent, acetonitrile, cannot be used in enantioseparation due

to the loss in enantioselectivity. Often methanol is used instead. However, Krause et al.

[304] reported better resolution of enantiomers on cellulose CSP using acetonitrile as an

organic modifier (Figure 2.22). In addition, the poor coverage of that stationary phase


results in high adsorption behaviour, which is close to that for normal phases. In some

cases, however, LC provides better separation.

Figure 2.22

CEC enantioseparation of bendroflumethazide in a Chiraspher®-packed capillary using (a)

methanol/50 mM NaH2PO4, pH 8.0 (60/40,V/V) and (b) acetonitrile/50 mM NaH2PO4,

pH 8.0 (40:60, V/V). Applied voltage: (a) 20 kV, (b) 12.5 kV; applied pressure: 10 bar on

inlet and outlet vial.

Otsuka et al. [303] found differences in the enantioseparation of propranolol under CEC

and HPLC modes using silica gel coated with cellulose-tris(3,5-dimethylphenylcarbamate).

These changes, however, can be associated to differences in the silica supports used.

Meyring et al. [418] used different polysaccharide types of stationary phases in non-aqueous

µ-LC and CEC in the separation of thalidomide and its hydroxylated derivatives. In CEC,

however, the baseline separation of six components could not be achieved. Lämmerhoffer

[179, 292] immobilized quinine derivative on different silica supports (Kromasil, Hypersil,

- 38 - Chapter 2

Micra NPS) and found different EOF and its dependency on the pH (Figure 2.23).

Figure 2.23

Influence of mobile phase pH (apparent pH, pHa) on electroosmotic mobility (µeo) on weak

anion echanger (WAX) type CSPs A-C. For sake of comparison also the EOF

characteristics of a reversed-phase type CEC capillary column are depicted in the plot. CEC

conditions - mobile phase: (a) MeOH-10 mM CH3COOH (pHa adjusted with

triethylamine, NEt3) (80:20, V/V); (b) ACN-100 mM MES (pHa adjusted with NEt3)

(80:20, V/V); EOF marker: thiourea; temperature: 20°C; voltage: ±15 kV; injection: ±5

kV/5 s; detection: UV at 250 nm.

Reynolds et al. [225] prepared non-porous organo-silica spheres with different sizes and

surface areas and found no correlation between the surface area of the stationary phase and

the velocity of the EOF. This contrasts with the findings of Zimina et al. [124] on

conventional reversed phase stationary phases as already mentioned in Section 2.3.2.1. The

group of Chaiyasut et al. [309] studied fluorinated bonded silica and observed changes in the

elution profiles of neutral compounds at different voltages (Figure 2.24). The authors

suggested that this must be attributed to the alteration of the stationary phase under the

applied voltage. In addition, the retention mechanism of benzene and o-cresol appeared to

be the same. The retention mechanisms of aniline and o-nitrophenol, however, were

different, probably due to dipole-dipole interaction with fluorine atoms of the bonded

phase. Nonlinearity in the EOF vs. voltage also suggests changes in the surface charge


density. Matyska et al. [257] studying the temperature influences in OT-CEC with bonded

cholesteryl phase found irregularities due to the changes in surface morphology.

Figure 2.24

Relationship between apllied voltage and k’/ '0k value of neutral compounds. Solutes are

aniline (diamond), o-nitrophenol (triangle), o-cresol (circle) and benzene (rectangular).

Conditions – column: 18.7 cm/33.3 cm effective/total length, I.D. 150 µm, packed with

fluorinated bonded silica stationary phase; mobile phase: 0.1% aqueous

tetrabutylammonium chloride containing 0.1% acetic acid (pH 3.35).

2.3.6 Organic polymer based The major advantage of polymer-based columns is their flexibility and ease of preparation.

The fritless technology resulting from in-situ preparation and the possibilities offered by

tuning the retention and (enantio-)selectivity of components by building in different

retentive groups, as well as its potential high efficiency, shows a promising future for

polymer-based CEC columns. Intensive studies on the preparation and modification of

polymeric monolithic columns were performed by Peters et al. [326-328, 350]. This group

- 40 - Chapter 2

studied the electrochromatographic performance of monoliths based on a cross-linked

polymethacrylate.

Table 2.4 Polymer based stationary phases applied in CEC.

Type Type of the phase References

Positively Charged 224, 321-343, 383

Negatively Charged 282, 332, 343-348, 412

Chiral with in-built chiral

selector

349-351, 345-347

Polacrylate,

polyacrylamide,

polymethacrylate based

Chiral, molecular imprinted 352-357, 393-394

PS-DVB - 358-364

Cellulose based - 365

PEEK, ECTFE 366-367 Other

other 285, 311, 368-376, 382, 384, 387

The properties and behaviour of these columns could easily be tuned by the ratio of

monomers, the amount and composition of the porogenic solvent, and by the amount of

aminopropanesulfonic acid yielding in sufficient charge on the polymeric surface. Polymeric

back-bones together with butyl side chains from butyl methacrylate monomer are

responsible for the reversed-phase behaviour of these stationary phases. Examples of

separation of benzene derivatives are shown in Figures 2.25 and 2.26. Further studies have

also been performed by Jiang et al. [342]. They studied the performance of the monolithic

columns prepared from ethyleneglycol dimethacrylate and butyl methacrylate under HPLC

and CEC conditions. Apart from the high efficiency under CEC conditions, higher polarity

(viz. tailing broad peak of benzylamine) and broad peaks for small bulky compounds like

1,3,5-triisopropylbenzene were found (see Figures 5.7 and 5.12 in Chapter 5). It has been

suggested that a substantial broadening of these bulky components compared to other

small molecules is caused by the micropores in the organic monolith. Furthermore, as also

suggested by other authors, part of these micropores may have a dead end. These so-called


ink pot pores strongly limit the unhindered back-diffusion of components out of such

pores. This effort of course is more dramatic for bulky types of molecules. It appears that

the broadening of peaks is a more common problem. In addition, other authors like e.g.

Dulay et al. [224] and Tan et al. [242] worked on polymethacrylate monoliths, Jinno et al.

[365] on cellulose acetate fibres, Fujimoto et al. [367] on derivatized PEEK observed such a

phenomenon.

Figure 2.25

Effect of porosity of monolithic capillaries on their electrochromatographic properties.

Conditions: capillary column, 150 µm I.D. × 30 cm active length; stationary phase:

polymethacrylate monolithic column with 0.6 mol % AMPS in monomer mixture; mode

pore size of 4000 (a), 1230 (b), and 670 nm (c); mobile phase: ACN-5 mmol/L phosphate

buffer pH 7 (80/20, V/V); UV detection at 215 nm; voltage 25 kV; pressure in vials,

0.2 MPa; sample concentration, 2 mg/mL of each compound; injection, 5 kV for 3 s.

Peaks: thiourea (1), benzyl alcohol (2), benzaldehyde (3), benzene (4), toluene

(5), ethylbenzene (6), propylbenzene (7), butylbenzene (8), and amylbenzene (9).

Some of these authors attributed the differences in the adsorption of the aromatic part of

the component molecule to the sorptive interaction of π-electrons with the polymeric

monolith. By determining desorption-sorption cycles of organic anions such as benzoic

acid, phtalic acid or benzenesulfonic acid Kitagawa et al. [368, 370] proved that organic

- 42 - Chapter 2

anion-exchange gel matrix is able to recognize the direction of the applied electric field.

This has, of course, a direct effect on the distribution coefficient. In spite of these

observations, most authors report linear dependencies of log k versus the percentage of

organic modifier in the mobile phase similar as in pressure driven reversed-phase

chromatography. Analyzing polar compounds, deviations from the linearity have also been

observed on some highly charged polymers [342].

Figure 2.26

Separation of benzene derivatives on monolithic capillary column using mobile phases

containing different percentages of acetonitrile. Conditions: column: polymethacrylate

monolith; column diameters: 100 µm I.D. × 30 cm active length; mobile phase, 60/40

(V/V) (a), 70/30 (V/V) (b), and 80/20 (V/V) (c) mixtures of acetonitrile and 5 mM

phosphate buffer, pH 7.

Successful chiral separations using organic based monoliths were demonstrated too. Two

types of chiral stationary phases have been prepared:


i) with a built-in chiral selector

ii) based on a molecular imprinting technology.

So far, type i) proved to be more successful in enantioseparations of such types of columns.

Furthermore, type ii) monoliths have lower demands on preparation.

2.4 Mobile phases A schematic overview of commonly used phases in CEC is shown in Figure 2.27.

Used organic solvents: acetonitrile, methanol, ethanol, tetrahydrofuran

Aqueous

Used organic solvents: n-hexane, dimethylformamide, acetonitrile, methanol

Non-aqueous

Mobile Phases

Examples: phosphates, tetraborates, chlorides

Inorganic

Examples: Tris, MES, HEPES, CTAB, SDS

Organic

Buffers

Mobile Phases and Buffers

Figure 2.27

Schematic overview of mobile phases used in CEC.

Typically, the majority of mobile phases in CEC consist of a mixture of an aqueous buffer

mixed with one or more organic modifiers, like acetonitrile for example. In some cases non-

aqueous buffers in organic solvents are applied too. Though several authors [79] have used

- 44 - Chapter 2

non-buffered mobile phases, there is some concern about the stability of the EOF and thus

on the reproducibility of the obtained data (see Figure 4.1 in Chapter 4).

As discussed earlier, acetonitrile as the organic modifier has the most optimal ε/η ratio.

EOF-values generated using ACN containing systems are about two times higher than for

MeOH [52, 402] and about three times higher than for THF [52]. It is also interesting to

study the EOF dependency on the percentage of organic modifier in the mobile phase. The

relationship between ε/η and the percentage of organic solvent is described in Figure 2.28.

Figure 2.28

Variation of the ratio of the dielectric constant and the viscosity, ε/η, with solvent

composition at 25°C. Conditions - capillary: fused-silica capillary 43 cm/96 cm

effective/total length, 100 µm I.D., mobile phase: organic solvent-10 mM KCl + 1 mM

phosphoric acid + KOH; solvent code: meoh, methanol; etoh, ethanol; proh, 2-propanol;

acn, acetonitrile; dmso, dimethylsulfoxide; acet, acetone.


The results for acetonitrile as an organic modifier reported in the literature, however, are

contradictory. Several authors report both an increase [e.g. 52, 62, 82, 103, 105, 113, 116,

165, 190, 215, 223, 255, 360] and a decrease [79, 95, 98, 343, 361, 405] of the EOF with an

increasing concentration of acetonitrile as the organic modifier in the buffer. Asiaie et al.

[255] compared the effect of acetonitrile concentration on the EOF in unmodified,

octadecylated fused-silica capillaries and on sintered and reoctadecylated Zorbax ODS

stationary phases. The results are presented in Figure 2.28.

Figure 2.28

Plots of electrosmotic mobility as a function of acetonitrile concentration. (●) 75 µm I.D. ×

23 effective/33 cm total length fused-silica capillary with monolithic packing of sintered

and reoctadecylated 6-µm Zorbax-ODS (80 Å), (○) 50 µm I.D. × 20 cm effective/27 cm

total length open fused-silica capillary with octadecylated innerwall and (□) 50 µm I.D. × 20

cm effective/27 cm total length raw open fused-silica capillary. Mobile phase: ACN/10 mM

sodium tetraborate, pH 8.0; detection, 214 nm; electrokinetic injection, 1 s, 5 kV; EOF

marker, 2 µl/mL formamide in the mobile phase.

In open unmodified capillaries the EOF velocity decreases with increasing organic modifier

(ACN) concentration. In contrast, reversed observations were obtained for ODS open and

packed capillary. It should be noted that in the reported cases the ionic strength is not kept

constant while changing the content of the organic modifier. In addition, miniaturized

techniques are also favourable for separations carried out with expensive solvents. As an

example, deuterated solvents have been used [67, 158, 197, 403] in CEC/NMR. Particularly

- 46 - Chapter 2

for non-porous stationary phases, while aiming to achieve a constant and stable EOF,

several authors report the addition of SDS to the buffer in a sub-critical micellar

concentration [74, 89, 181, 189, 239, 340, 392, 400].

2.4.1 Non-aqueous mobile phases Though applications on non-aqueous CEC are not

commonly applied yet, a number of them have been

developed. A precondition here is that salts to be used in

non-aqueous systems must be compatible and soluble in

the solvent. Among them lithium and ammonium salts

are the most commonly used. The major use of non-

aqueous CEC has been applied on samples with

potentially high retention properties, such as fats [109-

111, 123] or fullerenes [419] and also for more

fundamentally oriented studies, i.e. through-pore flow in

particle beads (see sections 2.3.1 and 2.3.2). Wright et al.

[79] investigated the electrochromatographic behaviour

of polyaromatic hydrocarbons using non-aqueous and

aqueous non-buffered mobile phases. The authors

discussed possibilities for different association structures

of the molecules of the mobile phase during fluctuations

in the zeta potential and thus EOF. Furthermore, the

authors showed an obvious relationship between vEOF

and the ε/η ratio of the buffer (Figure 2.29).

Figure 2. 29

Relationship of electroosmotic mobility (µeo) to (A) permittivity (ε), (B) viscosity (η), and

(C) permittivity/viscosity ratio (ε/η). Conditions – column: fused-silica capillary 45 cm/75

cm effective/total length, 50 µm I.D.; mobile phase: pure water, methanol (MeOH),

acetonitrile (ACN), dimethylformamide (DMF) and dimethylsulfoxide (DMSO); EOF

marker: acetone.


2.5 Conclusions The capillary electrochromatography, CEC, developed in the past years has become a

highly efficient separation technique. The present lack of sufficient robustness of this

technique particularly in terms of column preparation, column and EOF stability, however,

gave CEC a position of complimentarity rather than making it competing technique

compared to HPLC for example. Presently, the state of the art of CEC can be characterized

as follows:

a) There is a strong tendency to manufacture CEC columns from silica and organic

monoliths. It can be expected that the use of conventional HPLC-like columns will

decrease.

b) In physico-chemical terms the role of organic modifier is not very clear yet. From an

experimental point of view in CEC acetonitrile is preferred over methanol.

To increase the role of CEC, more attention must be paid to the separation mechanism,

behaviour of the stationary phases, mobile phases and solutes under electro-driven

conditions. Still there exits substantial contradiction in literature on the interpretation of

data obtained under CEC conditions. Increasing the knowledge of the fundamentals is one

of the necessary conditions to turn CEC into a reliable separation technique well positioned

between the techniques like HPLC for example. Furthermore, with the successful

development of new stationary phases for CEC, this technique may become an interesting

alternative technique in many application protocols.

REFERENCES

1. A.S. Rathore, Cs. Horváth, J. Chromatogr. A, 781 (1997) 185.

2. J.S. Kowalczyk, Chem. Anal. (Warsaw), 41 (1996) 157.

3. L.A. Colón, G. Burgos, T.D. Maloney, J.M. Cintrón, R.L. Rodríguez, Electrophoresis,

21 (2000) 3965.

4. M.G. Cikalo, K.D. Bartle, M.M. Robson, P. Myers, M.R. Euerby, Analyst, 123 (1998)

87R.

5. M.M. Robson, M.G. Cikalo, P. Myers, M.R. Euerby, K.D. Bartle, J. Microcolumn Sep.,

9 (1997) 357.

- 48 - Chapter 2

6. K.D. Bartle, R.A. Carney, A. Cavazza, M.G. Cikalo, P. Myers, M.M. Robson, S.C.P.

Roulin, K. Sealey, J. Chromatogr. A, 892 (2000) 279.

7. M.M. Dittmann, K. Wienand, F. Bek, G.P. Rozing, LC-GC, 13 (1995) 800.

8. A.L. Crego, A. González, M.L. Marina, Crit. Rev. Anal. Chem., 26 (1996) 261.

9. L.A. Colón, K.K. Reynolds, R. Alicea-Maldonando, A.M. Fermier, Electrophoresis, 18

(1997) 2162.

10. R. Stevenson, K. Mistry, I.S. Krull, Intl. Laboratory, 11 (1998) 11C.

11. T. Tsuda, LC-GC Eur., 5 (1992) 26.

12. R. Weinberger, Am. Lab (Shelton, Conn.), 29 (1997) 24FF.

13. N.W. Smith, A.S. Carter-Finch, J. Chromatogr. A, 892 (2000) 219.

14. K.D. Bartle, P. Myers, J. Chromatogr. A, 916 (2001) 3.

15. T. de Boer, R.A. de Zeeuw, G.J. de Jong, K. Ensing, Electrophoresis, 20 (1999) 2989.

16. L. Schweitz, M. Patersson, T. Johansson, S. Nilsson, J. Chromatogr. A, 892 (2000) 203.

17. J.H. Knox, R. Boughtflower, TrAC, Trends Anal. Chem., 19 (2000) 643.

18. F. Steiner, B. Scherer, J. Chromatogr. A, 887 (2000) 55.

19. G. Choudhary, A. Apffel, H. Yin, W. Hancock, J. Chromatogr. A, 887 (2000) 85.

20. C.A. Rimmer, S.M. Piraino, J.G. Dorsey, J. Chromatogr. A, 887 (2000) 115.

21. R.E. Majors, LC-GC, 16 (1998) 96.

22. J.H. Miyawa, M.S. Alasandro, LC-GC, 16 (1998) 36.

23. K.D. Altria, J. Chromatogr. A, 856 (1999) 443.

24. K.D. Altria, N.W. Smith, C.H. Turnbull, Chromatographia, 46 (1997) 664.

25. J.J. Pesek, M.T. Matyska, Electrophoresis, 18 (1997) 2228.

26. L. Schweitz, L.I. Andersson, S. Nilsson, J. Chromatogr. A, 817 (1998) 5.

27. M. Pursch, L.C. Sander, J. Chromatogr. A, 887 (2000) 313.

28. J.J. Pesek, M.T. Matyska, J. Chromatogr. A, 887 (2000) 31.

29. H. Zou, M. Ye, Electrophoresis, 21 (2000) 4073.

30. G.M.J. Bruin, Electrophoresis, 21 (2000) 3931.

31. K. Jinno, H. Sawada, TrAC, Trends Anal. Chem., 19 (2000) 664.

32. U. Pyell, J. Chromatogr. A, 892 (2000) 257.

33. L.A. Colón, T.D. Maloney, A.M. Fermier, J. Chromatogr. A, 887 (2000) 43.

34. K.K. Unger, D. Kumar, M. Grün, G. Büchel, S. Lüdtke, T. Adam, K. Schumacher, S.

Renker, J. Chromatogr. A, 892 (2000) 47.

35. L.I. Andersson, J. Chromatogr. B, 745 (2000) 3.


36. F. Svec, E.C. Peters, D. Sýkora, C. Yu, J.M.J. Fréchet, J. High Resolut. Chromatogr., 23

(2000) 3.

37. F. Svec, E.C. Peters, D. Sýkora, J.M.J. Fréchet, J. Chromatogr. A, (2000) 3.

38. Q. Tang, M.L. Lee, TrAc, Trends Anal. Chem., 19 (2000) 648.

39. V. Schurig, D. Wistuba, Electrophoresis, 20 (1999) 2313.

40. D. Wistuba, V. Schurig, J. Chromatogr. A, 875 (2000) 255.

41. I.S. Krull, A. Sebag, R. Stevenson, J. Chromatogr. A, 887 (2000) 137.

42. D. Wistuba, V. Schurig, Electrophoresis, 21 (2000) 4136.

43. A. Karcher, Z. El Rassi, Electrophoresis, 20 (1999) 3280.

44. Z. El Rassi, Electrophoresis, 20 (1999) 3134.

45. G.W. Sovocool, W.C. Brumley, J.R. Donelly, Electrophoresis, 20 (1999) 3297.

46. P.T. Vallano, V.T. Remcho, J. Chromatogr. A, 887 (2000) 125.

47. K. Walhagen, K.K. Unger, M.T.W. Hearn, J. Chromatogr. A, 887 (2000) 165.

48. T.D. Hatajik, P.R. Brown, J. Capillary Electrophor., 5 (1998) 143.

49. K.D. Altria, N.W. Smith, C.H. Turnbull, J. Chromatogr. B, (1998) 341.

50. G. Blaschke, B. Chankvetadze, J. Chromatogr. A, 875 (2000) 3.

51. HP booklet

52. M.M. Dittmann, G.P. Rozing, J. Microcolumn Sep., 9 (1997) 399.

53. A.S. Rathore, Cs. Horváth, Anal. Chem., 70 (1998) 3271.

54. A.S. Rathore, E. Wen, Cs. Horváth, Anal. Chem., 71 (1999) 2633.

55. M. Zhang, Z. El Rassi, Electrophoresis, 19 (1998) 2068.

56. C. Yang, Z. El Rassi, Electrophoresis, 21 (2000) 1977.

57. J. Jiskra, H.A. Claessens, C.A. Cramers, J. Sep. Sci., 25 (2002) 1.

58. C. Rice, R. Whitehead, J. Phys. Chem., 69 (1965) 4017.

59. J.H. Knox, I.H. Grant, Chromatographia, 24 (1987) 135.


61. E.P.C. Lai, E. Dabek-Zlotorzynska, Electrophoresis, 20 (1999) 2366.

62. W. Wei, G.A. Luo, G.Y. Hua, C. Yan, J. Chromatogr. A, 817 (1998), 65.

63. N.C. Gillott, M.R. Euerby, C.M. Johnson, D.A. Barrett, P.N. Shaw, Chromatographia,

51 (2000), 167.

64. W. Wei, G.A. Luo, R. Xiang, C. Yan, J. Microcolumn Sep., 11 (1999) 263.

65. A. Maruška, U. Pyell, J. Chromatogr. A, 782 (1997) 167.

66. M. Mayer, E. Rapp, C. Marck, G.J.M. Bruin, Electrophoresis, 20 (1999) 43.

- 50 - Chapter 2

67. K. Pusecker, J. Schewitz, P. Gfrörer, L.-H. Tseng, K. Albert, E. Bayer, Anal. Chem., 70

(1998) 3280.

68. E. Venema, J.C. Kraak, H. Poppe, R. Tijssen, J. Chromatogr. A, 837 (1999) 3.

69. R. Stol, W.Th. Kok, H. Poppe, J. Chromatogr. A, 914 (2001) 201.

70. R. Stol, H. Poppe, W.Th. Kok, J. Chromatogr. A, 887 (2000) 199.

71. S.R. Witowski, R.T. Kennedy, J. Microcolumn Sep., 11 (1999) 723.

72. T.S. Stevens, H.J. Cortes, Anal. Chem., 55 (1983) 1365.

73. M. Ye, H. Zou, Z. Liu, J. Ni, Y. Zhang, J. Chromatogr. A, 855 (1999) 137.

74. R.M. Seifar, S. Heemstra, W. Th. Kok, J.C. Kraak, Biomed. Chromatogr. 12 (1998)

140.

75. S. Suzuki, M. Yamamoto, Y. Kuwahara, K. Makiura, S. Honda, Electrophoresis, 19

(1998) 2682.

76. L. Roed, E. Lundanes, T. Greibrokk, J. Microcolumn Sep., 11 (1999) 421.

77. J. Helboe, S.H. Hansen, J. Chromatogr. A, 836 (1999) 315.

78. N.C. Gillot, M. R. Euerby, C.M. Johnson, D.A. Barret, P.N. Shaw, Anal. Commun., 35

(1998) 217.

79. P.B. Wright, A.S. Lister, J.G. Dorsey, Anal. Chem., 69 (1997) 3251.

80. K. Walhagen, K.K. Unger, A.M. Olsson, M.T.W. Hearn, J. Chromatogr. A, 853 (1999)

263.

81. M.R. Euerby, C.M. Johnson, K.D. Bartle, LC-GC, (1998) 386.

82. M.M. Dittmann, G.P. Rozing, J. Chromatogr. A, 744 (1996) 63.

83. M.M. Dittmann, G.P. Rozing, Biomed. Chromatogr., 12 (1998) 136.


85. M.M. Dittmann, K. Masuch, G.P. Rozing, J. Chromatogr. A, 887 (2000) 209.

86. P.A. Cooper, K.M. Jessop, F. Moffatt, Electrophoresis, 21 (2000) 1574.

87. E.F. Hilder, C.W. Klampfl, P.R. Haddad, P. Myers, Analyst, 215 (2000) 1.

88. L.Y. Fang, E.A. Huang, M.M. Safarpour, J. Capillary Electrophor., 5 (1998) 115.

89. C.W. Henry III, M.E. McCarroll, I.M. Warner, J. Chromatogr. A, 905 (2001) 319.

90. I.S. Lurie, R.P. Meyers, T.S. Conver, Anal. Chem., 70 (1998) 3255.

91. M.R. Euerby, D. Gilligan, C.M. Johnson, K.D. Bartle, Analyst (Cambridge, U.K.),

(1997) 1087.

92. M. R Euerby, C.M. Johnson, S.F. Smyth, N.C. Gillot, D.A. Barrett, P.N. Shaw, J.

Microcolumn Sep., 11 (1999) 305.


93. L.A. Frame, M.L. Robinson, W.J. Lough, J. Chromatogr. A, 798 (1998) 243.

94. H. Yamamoto, J. Bauman, F. Erni, J. Chromatogr., 593 (1992) 313.

95. S.L. Abidi, K.A. Rennick, J. Chromatogr. A, 913 (2001) 379.

96. M. Lämmerhofer, W. Lindner, J. Chromatogr. A, 839 (1999) 167.

97. E.F. Hilder, M. Macka, P.R. Haddad, Anal. Commun., 36 (1999) 299.

98. C. Yan, D. Schaufelberger, F. Erni, J. Chromatogr. A, 670 (1994) 15.

99. R.M. Seifar, S. Heemstra, W.Th. Kok, J.C. Kraak, H. Poppe, J. Microcolumn Sep., 10

(1998) 41.

100. T. Adam, K.K. Unger, M.M. Dittmann, G.P. Rozing, J. Chromatogr. A, 887 (2000)

327.

101. P.D.A. Angus, C.W. Demarest, T. Catalano, J. F. Stobaugh, J. Chromatogr. A, 887

(2000) 347.

102. A. Hilmi, J.H.T. Luong, Electrophoresis, 21 (2000) 1395.

103. D.A. Stead, R.G. Reid, R.B. Taylor, J. Chromatogr. A, 798 (1998) 259.

104. J. Wang, D. Schaufelberger, N.A. Guzman, J. Chromatogr. Sci., 36 (1998) 155.

105. F. Moffat, P.A. Cooper, K.M. Jessop, J. Chromatogr. A, 855 (2000) 215.

106. N.M. Djordjevic, P.W.J. Fowler, F. Houdiere, G. Lerch, J. Liq. Chromatogr. Relat.

Technol., 21 (1998) 2219.


108. S.E. Van Den Bosch, S. Heemstra, J.C. Kraak, H. Poppe, J. Chromatogr. A, 755 (1996)

165.

109. A. Dermaux, A. Medvedovici, M. Ksir, E. Van Hove, M. Talbi, P. Sandra, J.


110. A. Dermaux, P. Sandra, V. Ferraz, Electrophoresis, 20 (1999) 74.

111. A. Dermaux, P. Sandra, M. Ksir, K. Fellat, F. Zarrouck, J. High Resolut. Chromatogr.,

21 (1998) 545.

112. A.S. Lister, J.G. Dorsey, D.E. Burton, J. High Resolut. Chromatogr., 20 (1997) 523.

113. A.S. Lister, C.A. Rimmer, J.G. Dorsey, J. Chromatogr. A, 828 (1998) 105.

114. G.A. Lord, D.B. Gordon, L.W. Tetler, C.M. Carr, J. Chromatogr. A, 700 (1995) 27.

115. J.M. Miyawa, D.K. Lloyd, M.S. Alasandro, J. High Resolut. Chromatogr., 21 (1998)

161.

116. F. Moffatt, P.A. Cooper, K.M. Jessop, J. Chromatogr. A, 855 (2000) 215.

- 52 - Chapter 2

117. F. Moffatt, P. Chamberlain, P.A. Cooper, K.M. Jessop, Chromatographia, 48 (1998)

481.

118. F. Moffatt, P.A. Cooper, K.M. Jessop, Anal. Chem., 71 (1999) 1119.

119. P.K. Owens, J. Johansson, Anal. Chem., 72 (2000) 740.

120. J. Reilly, M. Saeed, J. Chromatogr. A, 829 (1998) 175.

121. K.J. Reynolds, T.D. Maloney, A.M. Fermier, L.A. Colón, Analyst, 123 (1998) 1493.

122. M. Saeed, M. Depala, D.H. Craston, I.G.M. Anderson, Chromatographia, 49 (1999)

391.

123. P. Sandra, A. Dermaux, V. Ferraz, M.M. Dittmann, G.P. Rozing, J. Microcolumn Sep.,

9 (1997) 409.

124. T.M. Zimina, R.M. Smith, P. Myers, J. Chromatogr. A, 758 (1997) 191.

125. J. Seavels, M. Wuyts, A. Van Schepdael, E. Roets, J. Hoogmartnes, J. Pharm Biomed.

Anal., 20 (1999) 513.

126. J. Ding, P. Vouros, Anal. Chem., 69 (1997) 379.

127. D. Jianmei, J. Szeliga, A. Dipple, P. Vouros, J. Chromatogr. A, 781 (1997) 327.

128. J. Ding, T. Barlow, A. Dipple, P. Vouros, J. Am. Soc. Mass Spectrom., 9 (1998) 823.

129. S.R. Sirimanne, J.R. Barr, D.G. Patterson, Jr., J. Microcolumn Sep., 11 (1999) 109.

130. R.J. Boughtflower, T. Underwood, C.J. Paterson, Chromatographia, 40 (1995) 329.

131. J. Jiskra, C.A. Cramers, M. Byelik, H.A. Claessens, J. Chromatogr. A, 862 (1999) 121.

132. J. Jiskra, T. Jiang, H.A. Claessens, C.A. Cramers, J Microcolumn Sep., 12 (2000) 530.

133. P.D.A Angus, E. Victorino, K.M. Payne, C.W. Demerest, T. Catalano, J.F. Stobaugh,

Electrophoresis, 19 (1998) 2073.

134. R. Stol, M. Mazareeuw, U.R. Tjaden, J. Van Der Graaf, J. Chromatogr. A, 873 (2000)

293.

135. I.S. Lurie, T.S. Conver, V.L. Ford, Anal. Chem., 70 (1998) 4563.

136. T. Eimer, K.K. Unger, J. Van Der Greef, TrAC, Trends Anal. Chem., 15 (1996) 463.

137. X. Cahours, P. Morin, L. Agrofoglio, M. Dreux, J. High Resolut. Chromatogr., 23

(2000) 138.

138. M. Qui, X.-F. Li, C. Stathakis, N.J. Dovichi, J. Chromatogr. A, 853 (1999) 131.

139. V. Spikmans, S.J. Lane, U.R. Tjaden, J. Van Der Graaf, Rapid Commun. Mass

Spectrom., 13 (1999) 141.

140. N.M. Djordjevic, F. Fitzpatrick, F. Houdiere, G. Lerch, G.P. Rozing, J. Chromatogr. A,

887 (2000) 245.


141. M.R. Euerby, C.M. Johnson, K.D. Bartle, P. Myers, S.C.P. Roulin, Anal. Commun., 33

(1996) 403.

142. M.G. Cikalo, K.D. Bartle, P. Myers, J. Chromatogr. A, 836 (1999) 35.

143. B. Xin, M. Lee, J. Microcolumn Sep., 11 (1999) 271.

144. N.W. Smith, J. Chromatogr. A, 887 (2000) 233.

145. S.C.P. Roulin, R. Dmoch, R.A. Carney, K.D. Bartle, P. Myers, M.R. Euerby, C.M.

Johnson, J. Chromatogr. A, 887 (2000) 307.

146. R.A. Carney, M.M. Robson, K.D. Bartle, P. Myers, J. High Resolut. Chromatogr., 22

(1999) 29.

147. N.W. Smith, M.B. Evans, Chromatographia, 41 (1995) 197.

148. L. Zhang, W. Shi, H. Zou, J.Y. Ni, Y. Zhang, J. Liq. Chromatogr. Relat. Technol., 22

(1999) 2715.

149. N. Smith, M.B. Evans, J. Chromatogr. A, 832 (1999) 41.


151. M.G. Cikalo, K.D. Bartle, P. Myers, Anal. Chem., 71 (1999) 1820.

152. G.A. Lord, D.B. Gordon, P. Myers, B.W. King, J. Chromatogr. A, 768 (1997) 9.

153. N.W. Smith, CAST (1999) 10.

154. N.P. Cassells, C.S. Lane, M. Depala, M. Saeed, D.H. Craston, Chemospher,e 40 (2000)

609.

155. R.D. Oleschuk, L.L. Shultz-Lockyear, Y. Ning, D.J. Harrison, Anal. Chem., 72 (2000)

585.

156. L. Zhang, H. Zou, W. Shi, J.Y. Ni, Y. Zahng, J. Capillary Electrophor., 5 (1998) 97.

157. Y. Zhang, W. Shi, L. Zhang, H. Zou, J. Chromatogr. A, 802 (1998) 59.

158. D. Hindocha, N.W. Smith, J. Chromatogr. A, 904 (2000) 99.

159. M.M. Robson, K.D. Bartle, P. Myers, Chromatographia, 50 (1999) 711.

160. M.M. Robson, S.C.P. Roulin, S.M. Shariff, M.W. Raynor, K.D. Bartle, A.A. Clifford, P.

Myers, M.R. Euerby, C.M. Johnson, Chromatographia, 43 (1996) 313.

161. C.W. Klampfl, W. Buchberger, P.R. Haddad, J. Chromatogr. A, 911 (2001) 277.

162. R. Stol, W.Th. Kok, H. Poppe, J. Chromatogr. A, 853 (1999) 45.

163. M. Schmid, F. Bäuml, A.P. Köhne, T. Welsch, J. High Resolut. Chromatogr., 22 (1999)

438.

164. M. Hugener, A.P. Tinke, W.M.A. Niessen, U.R. Tjaden, J. Van Der Greef, J.

Chromatogr., 647 (1993) 375.

- 54 - Chapter 2

165. A. Banholczer, U. Pyell, J. Chromatogr. A, 869 (2000) 363.

166. L. Roed, E. Lundanes, T. Greibrokk, J. Chromatogr. A, 890 (2000) 347.

167. D. Li, V.T. Remcho, J. Microcolumn Sep., 9 (1997) 389.

168. G. Chirica, V.T. Remcho, Electrophoresis, 21 (2000) 3093.


170. R. Stol, H. Poppe, W.Th. Kok, Anal. Chem., 73 (2001) 3332.

171. M. Mazareeuw, V. Spikmans, U.R. Tjaden, J. Van Der Greef, J. Chromatogr. 879

(2000) 219.

172. N.M. Djordjevic, F. Fitzpatrick, F. Houdiere, G. Lerch, J. High Resolut. Chromatogr.,

22 (1999) 599.

173. U. Pyell, H. Rebscher, A. Banholczer, J. Chromatogr. A, 779 (1997) 155.

174. P.T. Vallano, V.T. Remcho, Anal. Chem., 72 (2000) 4255.


176. A. Banholczer, U. Pyell, J. Microcolumn Sep., 10 (1998) 321.

177. H. Rebscher, U. Pyell, Chromatographia, 42 (1996) 171.

178. T. Tegeler, Z. El Rassi, Anal. Chem., 73 (2001) 3365.

179. E. Tobler, M. Lämmerhofer, W. Lindner, J. Chromatogr. A, 875 (2000) 341.

180. S. Luedtke, T. Adam, N. Von Doehren, K.K. Unger, J. Chromatogr. A, 887 (2000)

339.

181. J.-T. Lim, R.N. Zare, C.G. Bailey, D.J. Rakestraw, C. Yan, Electrophoresis, 21 (2000)

737.

182. I.S. Lurie, C.G. Bailey, D.S. Anex, M.J. Bethea, T.D. McKibben, J.F. Casale, J.

Chromatogr. A, 870 (2000) 53.

183. P.D.A. Angus, C.W. Demarest, T. Catalano, J.F. Stobaugh, Electrophoresis, 20 (1999)

2349.

184. H. Engelhardt, S. Lamotte, F.-T. Hafner, Am. Lab., 30 (1998) 40.

185. C.G. Bailey, C. Yan, Anal. Chem., 70 (1998) 3275.

186. R. Dadoo, R.N. Zare, C. Yan, D.S. Anex, Electrophoresis, 70 (1998) 4787.

187. I.S. Lurie, C.G. Bailey, D.S. Anex, M.J. Bethea, T.D. McKibben, J.F. Casale, J.

Chromatogr. A, 870 (2000) 53.

188. J.-R. Chen, R.N. Zare, E.C. Peters, F. Svec, J.J. Frechét, Anal. Chem., 73 (2001) 1987.

189. C.G. Bailey, S.R. Wallenborg, Electrophoresis, 21 (2000) 3081.



191. C. Desiderio, L. Ossicini, S. Fanali, J. Chromatogr. A, 887 (2000) 489.

192. D.B. Strickmann, G. Blaschke, J. Chromatogr. B, 748 (2000) 213.

193. C. Desiderio, S. Fanali, J. Chromatogr. A, 895 (2000) 123.

194. D. Nurok, M.C. Frost, D.M. Chenoweth, J. Chromatogr. A, 903 (2000) 211.

195. E. Rapp, E. Bayer, J. Chromatogr. A, 887 (2000) 367.

196. K. Schmeer, B. Behnke, E. Bayer, Anal. Chem., 67 (1995) 3656.

197. P. Gfrörer, J. Schewitz, K. Pusecker, L.-H. Tseng, K. Albert, E. Bayer, Electrophoresis,

20 (1999) 3.

198. B. Behnke, J. Johansson, E. Bayer, S. Nilsson, Electrophoresis, 21 (2000) 3102.

199. B. Behnke, E. Bayer, J. Chromatogr. A, 680 (1994) 93.

200. B. Behnke, J. Johansson, S. Zhang, E. Bayerm S. Nilsson, J. Chromatogr. A, 818 (1998)

257.

201. B. Behnke, J.W. Metzger, Electrophoresis, 20 (1999) 80.

202. C. Rentel, P. Gfrörer, E. Bayer, Electrophoresis, 20 (1999) 2329.

203. P. Gfrörer, L.-H. Tseng, E. Rapp, K. Albert, E. Bayer, Anal. Chem., 73 (2001) 3234.

204. T. Tsuda, Anal. Chem., 59 (1987) 521.

205. T. Tsuda, Analusis, 26 (1998) M48.

206. C. Yan, R. Dadoo, R.N. Zare, D.J. Rakestraw, D.S. Anex, Anal. Chem., 68 (1996) 2726.

207. C. Yan, R. Dadoo, H. Zhao, R.N. Zare, D.J. Rakestraw, Anal. Chem., 67 (1995) 2026.

208. A.M. Fermier, L.A. Colón, J. Microcolumn Sep., 10 (1998) 439.

209. M.R. Taylor, P. Teale, S.A. Westwood, D. Perret, Anal.Chem. 69 (1997) 2554.

210. E. Wen, R. Asiaie, Cs. Horváth, J. Chromatogr. A, 855 (1999) 349.

211. A.S. Rathore, Cs. Horváth, Anal. Chem., 70 (1998) 3069.

212. Q.-H. Wan, J. Chromatogr. A, 782 (1997) 181.

213. C.G. Huber, G. Choudhary, Cs. Horváth, Anal. Chem., 69 (1997) 4429.


215. R.M. Seifar, J.C. Kraak, W.Th. Kok, H. Poppe, J. Chromatogr. A, 808 (1998) 71.

216. G. Choudhary, Cs. Horváth, J. Chromatogr. A, 781 (1997) 161.

217. G. Choudhary, Cs. Horváth, J.F. Banks, J. Chromatogr. A, 828 (1998) 469.



220. K. Jinno, H. Sawada, A.P. Catabay, H. Watanabe, N.B.H. Sabli, J.J. Pesek, M.T.

Matyska, J. Chromatogr. A, 887 (2000) 479.

- 56 - Chapter 2

221. P. Huang, J.-T. Wu, D.M. Lubman, Anal. Chem., 70 (1998) 3003.

222. K.W. Whitaker, M.J. Sepaniak, Electrophoresis, 15 (1994) 1341.

223. Q. Tang, B. Xin, M. Lee, J. Chromatogr. A, 837 (1999) 35.

224. M. T. Dulay, J.P. Quirino, B.D. Bennett, M. Kato, R.N. Zare, Anal. Chem., 73 (2001),

3921.

225. K.J. Reynolds, L.A. Colón, J. Liq. Chromatogr. Relat. Technol., 23 (2000) 161.

226. Q.-H. Ru, J. Yao, G.A. Luo, Y.-X. Zhang, C. Yan, J. Chromatogr. A, 894 (2000) 337.

227. S. Thiam, S.A. Shamsi, C.W. Henry III, J.W. Robinson, I.M. Warner, Anal. Chem., 72

(2000) 2541.

228. V. Lopez-Avila, J. Bedenedicto, C. Yan, J. High Resolut. Chromatogr., 20 (1997) 615.

229. E. Dabek-Zlotorzynska, E.P.C. Lai, J. Chromatogr. A, 853 (1999) 487.

230. W. Guo, J.A. Koropchak, C. Yan, J. Chromatogr. A, 849 (1999) 587.

231. Q.-H. Ru, G.-A. Luo, Y.-R. Fu, J. Chromatogr. A, 924 (2001) 331.

232. Y.S. Fung, Y.H. Long, J. Chromatogr. A, 907 (2001) 301.

233. W. Wei, Y.M. Wang, G.A. Luo, R.J. Wang, Y.H. Guan, C. Yan, J. Liq. Chromatogr.

Relat. Technol., 21 (1998) 1433.

234. R. Dadoo, C. Yan, R.N. Zare, D.S. Anex, D.J. Rakestraw, G.A. Hux, LC-GC, 15

(1997) 630.

235. V. Lopez-Avila, J. Benedicto, C. Yan, J. Chromatogr. Sci., 37 (1999) 165.

236. X. Cahours, P. Morin, M. Dreux, J. Chromatogr. A, 845 (1999) 203.

237. L. Roed, E. Lundanes, T. Greibrokk, Electrophoresis, 20 (1999) 2373.

238. U. Tallarek, D. Van Dusschoten, H. Van As, G. Guiochon, E. Bayer, Angew. Chem.,

Int. Ed., 37 (1998) 1882.

239. R.M. Seifar, J.C. Kraak, H. Poppe, W. Th. Kok, J. Chromatogr. A, 832 (1999) 133.

240. R.M. Seifar, W.Th. Kok, J.C. Kraak, H. Poppe, Chromatographia, 46 (1997) 131.

241. S. Suzuki, Y. Kuwahara, K. Makiura, S. Honda, J. Chromatogr. A, 873 (2000) 247.

242. T. Tanigawa, T. Nakagawa, K. Kimata, H. Nagayama, K. Hosoya, N. Tanaka, J.

Chromatogr. A, 887 (2000) 299.

243. M.T. Dulay, C. Yan, D.J. Rakestraw, R.N. Zare, J. Chromatogr. A, 725 (1996) 361.

244. J. Dickson, M. Odom, F. Ducheneaux, J. Murray, R.E. Milofsky, Anal. Chem., 72

(2000) 3038.

245. Q. Tang, M. Lee, J. High Resolut. Chromatogr., 23 (2000) 73.


246. N. Tanaka, N. Nagayama, H. Kobayashi, T. Ikegami, K. Hosoya, N. Ishizuka, H.

Minakuchi, K. Nakanishi, K. Cabrera, D. Lubda, J. High Resolut. Chromatogr., 23

(2000) 111.

247. C.K. Ratnayake, C.S. Oh, M.P. Henry, J. Chromatogr. A, 887 (2000) 277.

248. C. Fujimoto, J. High Resolut. Chromatogr., 23 (2000) 89.

249. C.K. Ratnayake, C.S. Oh, M.P. Henry, J. High Resolut. Chromatogr., 23 (2000) 81.

250. A.L. Crego, J. Martínez, M.L. Marina, J. Chromatogr. A, 869 (2000) 329.

251. M.C. Breadmore, M.Macka, N. Avdalovic, P.R. Haddad, Anal. Chem., 73 (2001) 820.

252. J.J. Pesek, M.T. Matyska, J.E. Sandoval, E.J. Williamsen, J. Liq. Chromatogr. Relat.

Technol., 19 (1996) 2843.

253. B. He, J. Ji, F. Regnier, J. Chromatogr. A, 853 (1999) 257.

254. M.T. Dulay, R.P. Kulkarni, R.N. Zare, Anal. Chem., 70 (1998) 5103.

255. R. Asiaie, X. Huang, D. Farnan, Cs. Horváth, J. Chromatogr. A, 806 (1998) 251.

256. S.C. Jacobsen, R. Hergenroeder, L.B. Koutny, J.M. Ramsey, Anal. Chem., 66 (1994)

2369.

257. M.T. Matyska, J.J. Pesek, R.I. Boysen, M.T.W. Hearn J. Chromatogr. A, 924 (2001)

211.

258. J.J. Pesek, M.T. Matyska, L. Mauskar, J. Chromatogr. A, 763 (1997) 307.

259. J.J. Pesek, M.T. Matyska, S. Cho, J. Chromatogr. A, 845 (1999) 237.

260. J.J. Pesek, M.T. Matyska, J. Capillary Electrophor,. 4 (1997) 213.

261. M.T. Matyska, J.J. Pesek, L. Yang, J. Chromatogr. A, 887 (2000) 497.

262. M.T. Matyska, Chem. Anal. (Warsaw), 43 (1998) 637.

263. M.J. Hoffman, A. Dovletoglou, J. High Resolut. Chromatogr., 22 (1999) 309.

264. L.C. Sander, M. Pursch, B. Maerker, S.A. Wise, Anal. Chem., 71 (1999) 3477.

265. Y. Chen, G. Gerhardt, R.M. Cassidy, Anal. Chem., 72 (2000) 610.

266. Y. Deng, J. Zhang, T. Tsuda, P.H. Yu, A.A. Boulton, R.M. Cassidy, Anal. Chem., 70

(1998) 4586.

267. R.N. Warriner, A.S. Craze, D.E. Games, S.J. Lane, Rapid Commun. Mass Spectrom.,

12 (1998) 1143.



270. S.J. Lane, A. Pipe, Rapid Commun. Mass Spectrom., 12 (1998) 667.

271. M. Ye, H. Zou, Z. Liu, J.Y. Ni, Y. Zhang, Anal. Chem., 72 (2000) 616.

- 58 - Chapter 2

272. M. Ye, H. Zou, Z. Liu, J.Y. Ni, J. Chromatogr. A, 869 (2000) 385.

273. S. Constantin, R. Freitag, J. Chromatogr. A, 887 (2000) 253.

274. C.W. Klampfl, P.R. Haddad, J. Chromatogr. A, 884 (2000) 277.

275. C.W. Klampfl, E.F. Hilder, P.R. Haddad, J. Chromatogr. A, 888 (2000) 267.

276. V. Spikmans, S.J. Lane, N.W. Smith, Chromatographia, 51 (2000) 18.


278. M. Zhang, C. Yang, Z. El Rassi, Anal. Chem., 71 (1999) 3277.

279. W. Wei, G.A. Luo, C. Yan, Am. Lab. (Shelton, Conn.), 30 (1998) 20C.

280. M. Ye, H. Zou, Z. Liu, J.Y. Ni, J. Chromatogr. A, 887 (2000) 223.

281. P. Huang, X. Jin, Y. Chen, J.R. Srinivasan, D.M. Lubman, Anal. Chem., 71 (1999)

1786.

282. B. Scherer, F. Steiner, J. Chromatogr. A, 924 (2001) 197.

283. J. Zhang, X. Huang, S. Zhang, Cs. Horváth, Anal. Chem., 72 (2000) 3022.

284. J. Hayes, A. Malik, Anal. Chem., 72 (2000) 4090.

285. N. Guan, Z. Zeng, Y. Wang, E. Fu, J. Cheng, Anal. Chim. Acta, 418 (2000) 145.

286. J.D. Hayes, A. Malik, Anal. Chem., 73 (2001) 987.

287. C. Desiderio, Z. Atukuri, S. Fanali, Electrophoresis, 22 (2001) 535.

288. H. Wikström, L.A. Svensson, A. Torstensson, P.K. Owens, J. Chromatogr. A, 869

(2000) 395.

289. A. Dermaux, F. Lynen, P. Sandra, J. High Resolut. Chromatogr., 21 (1998) 575.

290. M. Kato, M.T. Dulay, B. Bennett, J.-R. Chen, R.N. Zare, Electrophoresis, 21 (2000)

3145.

291. M. Lämmerhofer, E. Tobler, W. Lindner, J. Chromatogr. A, 887 (2000) 421.


293. D. Wistuba, H. Czesla, M. Roeder, V. Schurig, J. Chromatogr. A, 815 (1998) 183.



296. M. Meyring, B. Chankvetadze, J. Chromatogr. A, 876 (2000) 157.

297. M. Girod, B. Chankvetadze, G. Blaschke, J. Chromatogr. A, 887 (2000) 439.

298. F. Lelièvre, C. Yan, R.N. Zare, P. Gareil, J. Chromatogr. A, 723 (1996) 145.

299. S. Mayer, V. Schurig, J. High Resolut. Chromatogr., 15 (1992) 129.

300. D. Wistuba, V. Schurig, Electrophoresis, 21 (2000) 3152.

301. A.S. Carter-Finch, N.W. Smith, J. Chromatogr. A, 848 (1999) 375.


302. K. Kawamura, K. Otsuka, S. Terabe, J. Chromatogr. A, 924 (2001) 251.

303. K. Otsuka, C. Mikami, S. Terabe, J. Chromatogr. A, 887 (2000) 457.

304. K. Krause, M. Girod, B. Chankvetadze, G. Blaschke, J. Chromatogr. A, 837 (1999) 51.

305. M. Kato, M.T. Dulay, B.D. Bennett, J.P. Quirino, R.N. Zare, J. Chromatogr. A, 924

(2001) 187.

306. K. Krause, B. Chankvetadze, Y. Okamoto, G. Blaschke, J. Microcolumn Sep., 12

(2000) 398.

307. K. Krause, B. Chankvetadze, Y. Okamoto, G. Balschke, Electrophoresis, 20 (1999)

2772.

308. C. Wolf, P.L. Spence, W.H. Pirkle, E.M. Derrico, D.M. Cavender, G.P. Rozing, J.

Chromatogr. A, 782 (1997) 175.

309. C. Chaiyasut, T. Tsuda, S. Kitagawa, H. Walda, T. Monde, Y. Nakabeya, J.


310. J.J. Pesek, M.T. Matyska, J. Liq. Chromatogr. Relat. Technol. 21 (1998) 2923-2934

311. W. Xu, F. Regnier, J. Chromatography A 853 (1999) 243.

312. G. Zhang, B. Xu, J. Chromatogr. A, 912 (2001) 335.

313. A.V. Pirogov, W. Buchberger, J. Chromatogr. A, 916 (2001) 51.

314. S. Mayer, X. Briand, E. Francotte, J. Chromatogr. A, 875 (2000) 331.

315. M.M. Dittmann, G.P. Rozing, G. Ross, T. Adam, K.K. Unger, J. Capillary

Electrophor., 4 (1997) 201.

316. U. Tallarek, E. Rapp, T. Scheenen, E. Bayer, H. Van As, Anal. Chem., 72 (2000) 2292.

317. M. Zhang, G.K. Ostrander, Z. El Rassi, J. Chromatogr. A, 887 (2000) 287.

318. M. Ye, H. Zou, Z. Liu, J. Ni, Y. Zhang, J. Chromatogr. A, 855 (1999) 137.

319. C.D. Byrne, N.W. Smith, H.S. Dearie, F. Moffatt, S.A.C. Wren, K.P. Evans, J.

Chromatogr. A, 927 (2001), 169.

320. M. Ye, H. Zou, Z. Lei, R. Wu, Z. Liu, J.Y. Ni, Electrophoresis, 22 (2001) 518.

321. R. Wu, H. Zou, M. Ye, Z. Lei, J.Y. Ni, Electrophoresis, 22 (2001) 544.

322. Z.J. Tan, V.T. Remcho, J. Microcolumn Sep., 10 (1998) 99.

323. Z.J. Tan, V.T. Remcho, Anal. Chem., 69 (1997) 581.

324. C. Fujimoto, Anal. Chem., 67 (1995) 2050.

325. C. Fujimoto, Y. Fujise, E. Matsuzawa, Anal. Chem., 68 (1996) 2753.

326. E.C. Peters, M. Petro, F. Svec, J.M.J. Fréchet, Anal. Chem., 69 (1997) 3646.


- 60 - Chapter 2


329. A.M. Enlund, C. Ericson, S. Hjerten, D. Westerlund, Electrophoresis, 22 (2001) 511.

330. J.-L. Liao, N. Chen, C. Ericson, S. Hjertén, Anal. Chem., 68 (1996) 3468.

331. T. Koide, K. Ueno, Anal. Sci., 15 (1999) 791.

332. R. Shediac, S.M. Ngola, D.J. Throckmorton, D.S. Anex, T.J. Shepodd, A.K. Singh, J.

Chromatogr. A, 925 (2001) 251.

333. C. Ericson, J.-L. Liao, K. Nakazato, S. Hjertén, J. Chromatogr. A, 767 (1997) 33.

334. A.H. Que, A. Palm, A.GF. Baker, M. Novotny, J. Chromatogr. A, 887 (2000) 379.

335. C. Ericson, J. Holm, T. Ericson, S. Hjertén, Anal. Chem., 72 (2000) 81.

336. C. Fujimoto, J. Kino, H. Sawada, J. Chromatogr. A, 716 (1995) 107.

337. J.J. Pesek, M.T. Matyska, S. Menezes, J. Chromatogr. A, 853 (1999) 151.

338. A. Palm, M. Novotny, Anal. Chem., 69 (1997) 4499.

339. D. Hoegger, R. Freitag, J. Chromatogr. A, 914 (2001) 211.

340. S. Hjertén, Ind. Eng. Chem. Res., 38 (1999) 1205.

341. A.H. Que, V. Kahle, M. Novotny, J. Microcolumn Sep., 12 (2000) 1.

342. T. Jiang, J. Jiskra, H.A. Claessens, C.A. Cramers, J. Chromatogr. A, 923 (2000) 215.

343. S. Zhang, X. Huang, J. Zhang, Cs. Horváth, J. Chromatogr. A, 887 (2000) 465.

344. S. Zhang, J. Zhang, Cs. Horváth, J. Chromatogr. A, 914 (2001) 189.

345. M. Lämmerhofer, F. Svec, J.M.J. Frechét, W. Lindner, Anal. Chem., 72 (2000), 4623.

346. T. Koide, K. Ueno, J. High Resolut. Chromatogr., 23 (2000) 59-.

347. Á. Végvári, A. Földesi, C. Hetényi, O. Kocnegarova, M. Schmid, V. Kudirkaite, S.

Hjertén, Electrophoresis, 21 (2000) 3116.

348. M. Lämmerhofer, F. Svec, J.M.J. Fréchet, W. Lindner, J. Chromatogr. A, 925 (2001)

265.

349. T. Koide, K. Ueno, J. Chromatogr. A, 909 (2001) 305.

350. E.C Peters, K. Lewandowski, M. Petro, F. Svec, J.M.J. Fréchet, Anal. Commun., 35

(1998) 83.

351. M. Schmid, N. Grobuschek, C. Tuscher, G. Guebitz, Á. Végvári, A. Maruska, S.

Hjertén, Electrophoresis, 21 (2000) 3141.


353. L. Schweitz, L.I. Andersson, S. Nilsson, Anal. Chem., 69 (1997) 1179.

354. J.-M. Lin, T. Nakagama, K. Uchiyama, T. Hobo, Biomed. Chromatogr., 11 (1997) 298.

355. Z.J. Tan, V.T. Remcho, Electrophoresis, 19 (1998) 2055.


356. J.-M. Lin, T. Nakagama, K. Uchiyama, T. Hobo, J. Pharm. Biomed. Anal., 15 (1997)

1351.

357. J.-M. Lin, T. Nakagama, K. Uchiyama, T. Hobo, J. Liq. Chromatogr. Relat. Technol.,

20 (1997) 1489.

358. J. Saevels, M. Wuyts, A. Van Schepdael, J. Hoogmartens, Biomed. Chromatogr., 12

(1998) 149.

359. B. Xiong, L. Zhang, Y. Zhang, H. Zou, J. Wang, J. High Resolut. Chromatogr., 23

(2000) 67.

360. Y. Liu, D.J. Pietrzyk, J. Chromatogr. A, 920 (2001) 367.

361. X. Huang, J. Zhang, Cs. Horváth, J. Chromatogr. A, 858 (1999) 91.

362. I. Gusev, X. Huang, Cs. Horváth, J. Chromatogr. A, 855 (1999) 273.

363. K. Hosoya, H. Ohta, K. Yoshizako, K. Kimata, T. Ikegami, N. Tanaka, J. Chromatogr.

A, 853 (1999) 11.

364. M.C. Breadmore, M. Macka, P.R. Haddad, Electrophoresis, 20 (1999) 1987.

365. K. Jinno, J. Wu, H. Sawada, Y. Kiso, J. High Resolut. Chromatogr., 21 (1998) 617.

366. R. Alicea-Maldonando, L.A. Colón, Electrophoresis, 20 (1999) 37.

367. C. Fujimoto, M. Sakurai, Y. Muranaka, J. Microcolumn Sep., 11 (1999) 693.

368. S. Kitagawa, T. Tsuda, Anal. Sciences, 14 (1998) 571.

369. S. Kitagawa, A. Tsuji, H. Watanabe, M. Nakashima, T. Tsuda, J. Microcolumn Sep., 9

(1997) 347.

370. S. Kitagawa, H. Watanabe, T. Tsuda, Electrophoresis, 20 (1999) 9.

371. W.-H. Chen, S.-Y. Lin, C.-Y. Liu, Anal. Chim. Acta, 410 (2000) 25.

372. K. D. Cole, H. Cabezas, Jr., J. Chromatogr. A, 760 (1997), 259.

373. M.C. Boyce, M.C. Breadmore, M. Macka, P. Doble, P.R. Haddad, Electrophoresis, 21

(2000) 3073.

374. M.C. Breadmore, M.C. Boyce, M. Macka, N. Avdalovic, P.R. Haddad, J. Chromatogr.

A, 892 (2000) 303.

375. E.F. Hilder, C.W. Klampfl, P.R. Haddad, J. Chromatogr. A, 890 (2000) 337.

376. K. Krause, B. Chankvetadze, Y. Okamoto, G. Blaschke, Electrophoresis, 20 (1999)

2772.

377. C. Karlsson, L. Karlsson, D.W. Armstrong, P.K. Owens, Anal. Chem., 72 (2000) 4394.

378. Y. Guo, L.A. Colón, Anal. Chem., 67 (1995) 2511.

379. Q. Tang, N. Wu, M. Lee, J. Microcolumn Sep., 11 (1999) 550.

- 62 - Chapter 2

380. D.H. Thomas, D.J. Rakestraw, J.S. Schoeniger, V. Lopez-Avila, J. Van Emon,


381. C. Wolf, P.L. Spence, W.H. Pirkle, D.M. Cavender, E.M. Derrico, Electrophoresis, 21

(2000) 917.

382. M.C. Breadmore, M. Macka, N. Avdalovic, P.R. Haddad, Analyst, 125 (2000) 1235.

383. H. Sawada, K. Jinno, Electrophoresis, 20 (1999) 24.

384. M.R. Schure, R.E. Murphy, W.L. Klotz, W. Lau, Anal. Chem., 70 (1998) 4985.

385. S. Kitagawa, T. Tsuda, J. Microcolumn Sep., 6 (1994) 91.

386. N. Ishizuka, H. Minakuchi, K. Nakanishi, N. Soga, H. Nagayama, K. Hosoya, N.

Tanaka, Anal. Chem., 72 (2000) 1275.

387. M. Gilar, A. Belenky, A.S. Cohen, Electrophoresis, 21 (2000) 2999-3009

388. M. C. Breadmore, E.F. Hilder, M. Macka, N. Avdalovic, P.R. Haddad, Electrophoresis,

22 (2001) 503.

389. T. Sakaki, S. Kitagawa, T. Tsuda, Electrophoresis, 21 (2000) 3088.

390. B. Xin, M. Lee, Electrophoresis, 20 (1999) 67.

391. H. Jakubetz, H. Czesla, V. Schurig, J. Microcolumn Sep., 9 (1997) 421.

392. J.-R. Chen, M.T. Dulay, R.N. Zare, F. Svec, E.C. Peters, Anal. Chem., 72 (2000) 1224.

393. J.-M. Lin, T. Nakagama, X.-Z. Wu, U. Katsami, T. Hobo, Fresenius’ J. Anal. Chem.,

357 (1997) 130.

394. J.-M. Lin, K. Uchiyama, T, Hobo, Chromatographia, 47 (1998) 625.

395. S. Lüdtke, T. Adam, K.K. Unger, J. Chromatogr. A, 786 (1997) 229.

396. X. Zhang, F. Regnier, J. Chromatogr. A, 869 (2000) 319.

397. T. Tsuda, Y. Muramatsu, J. Chromatogr., 515 (1990) 645.

398. G. Chirica, V.T. Remcho, Anal. Chem., 72 (2000) 3605.

399. N. Gottschlich, S.C. Jacobson, C.T. Culbertson, J.M. Ramsey, Anal. Chem., 73 (2001)

2669.

400. E. Venema, J.C. Kraak, H. Poppe, R. Tijssen, Chromatographia, 48 (1998) 347.

401. T. Tsuda, Anal. Chem., 60 (1988) 1677.

402. C. Schwer, E. Kenndler, Anal. Chem., 63 (1991) 1801.

403. P. Gfrörer, J. Schewitz, K. Pusecker, E. Bayer, Anal. Chem., (1999) 315A.


Tanaka, Anal. Chem., 72 (2000) 1275.



406. Y. Li, H. Liu, X. Ji, J. Li, Electrophoresis, 21 (2000) 3109.

407. M.R. Euerby, D. Gilligan, C.M. Johnson, S.C.P. Roulin, P. Myers, K.D. Bartle, J.


408. T. Adam, K.K. Unger, J. Chromatogr. A, 894 (2000) 241.

409. A. Cavazza, K.D. Bartle, P. Dugo, L. Mondello, Chromatographia, 53 (2001) 57.

410. S. Dube, R.M. Smith, Chromatographia, 53 (2001) 51.

411. D.B. Strickmann, B. Chankvetadze, G. Blaschke, C. Desiderio, S. Fanali, J.

Chromatogr. A, 887 (2000) 393.

412. A.H. Que, T. Konse, A.G. Baker, M.V. Novotny, Anal. Chem., 72 (2000) 2703.

413. Q. Tang, M.L. Lee, J. Chromatogr. A, 887 (2000) 265.

414. J.P.C. Vissers, H.A. Claessens, P. Coufal, C.A. Cramers, J. High Resolut. Chromatogr.,,

18 (1995) 540.

415. J. Ding, P. Vouros, J. Chromatogr. A., 887 (2000) 103.

416. M. Lämmerhofer, F. Svec, J.M.J. Fréchet, W. Lindner, TrAC, Trends Anal. Chem., 19

(2000) 676.

417. M.J. Hilhorst, G.W. Somsen, G.J. de Jong, J. Chromatogr. A, 872 (2000) 315.

418. M. Meyring, B. Chankvetadze, G. Blaschke, J. Chromatogr. A, 876 (2000) 157.

419. K.W. Whitaker, M.J. Sepaniak, Electrophoresis, 15 (1994) 1341.

- 64 - Chapter 2

- 65 -

CHAPTER 3 3 CHROMATOGRAPHIC PROPERTIES OF REVERSED

PHASE STATIONARY PHASES UNDER PRESSURE AND

ELECTRO DRIVEN CONDITIONS; EFFECT OF

ORGANIC MODIFIER

Summary

Seven different reversed-phase (RP) stationary phases were examined under pressure- viz.

liquid chromatographic and electro-driven viz. capillary electrochromatographic

conditions. Characterization of the stationary phases was performed following well-

established test procedures providing a number of distinct column descriptors –

hydrophobicity, hydrophobic selectivity and silanol activity. These parameters were used

to describe the behaviour of the RP-columns under both pressure- and electro-driven

conditions. It is shown that chromatographic characteristics of porous RP-phases

substantially depend on the mode of operation. In contrast column descriptors of a non-

porous viz. solid RP-phase material hardly differed under pressure- and electro-driven

conditions.

This chapter has been published: J. Jiskra, C.A. Cramers, M. Byelik, H.A. Claessens, J. Chromatogr. A, 862 (1999) 121.

3.1 Introduction Capillary electrochromatography (CEC) is a recently developed separation technique

combining the excellent efficiency usually achieved in electrophoretic separation techniques

and the high selectivity which is characteristic for high-performance liquid chromatography

- 66 - Chapter 3

(HPLC) [1, 2]. At present CEC is considered as a potential alternative technique for micro-

HPLC which latter technique has been already established for longer time [3-7]. Many

papers have been published recently on the theory and practical aspects in

electrochromatography [8-22]. A number of these reports start from the theoretical

concepts originating from capillary electrophoresis (CE) explaining the high efficiencies

from the plug-like velocity profiles obtained in these techniques [20-22]. Yan et al. [23]

compared the efficiencies that can be achieved in CEC and in micro-HPLC of 50 µm I.D.

columns packed with 3 µm Hypersil ODS. They found significantly higher efficiency for

the column under electro-driven conditions. As discussed in [24], in CEC the flow velocity

and the plug-like velocity profile are not dependent on the particle size down to

approximately 0.5 µm or even lower as long as no double-layer overlapping occurs [24-26].

Opposite to HPLC where the use of smaller particles is seriously limited by pressure drop

limitations, in CEC such small particles can be easily used. This makes CEC potentially a

highly efficient technique compared to HPLC. Furthermore, based on detection

developments in the field of micro-HPLC, on-column fluorescence detection was recently

introduced into CEC, too and provides additional increase in efficiency [27-28]. At present

the introduction of CEC is hampered by two major problems. First the technical difficulties

encountered in the manufacturing of suitable and reliable CEC columns are still substantial.

In addition the achievement of a sufficient and stable electroosmotic flow (EOF) and the

mechanical stability of the packed bed in CEC-columns are still problematic. Secondly there

remains a number of major questions on the backgrounds of retention and selectivity in

CEC. In that framework the eventual changes in physico-chemical properties of HPLC

stationary phases under electrically driven conditions is an issue of great interest. For

instance Vissers et al. [29] earlier showed that using the same stationary phase the retention

for neutral compounds in CEC is about 20% higher than under micro-HPLC mode. In

contrast Eimer et al. [30] found that for more hydrophobic analytes the retention factors

were 36-40% lower in pressurized electrochromatography (PEC) than in capillary LC. They

concluded that this might be attributed to the higher polarity of the stationary phase under

electric field conditions. Using RP columns Wei et al. [31] used solvatochromic parameters

to study retention in CEC and found that the retention behaviour under pressure- and

electro-driven conditions is very dissimilar for these columns. It appeared that the hydrogen

bond acidity and dipolarity/polarizibity of solutes play a more dominant role in CEC than

Chromatographic Properties of Reversed Phase … - 67 -

in HPLC. Furthermore they conclude that the effects of solute size and hydrogen bond

basicity on retention are similar in both separation modes. Djordjevic et al. [32] compared

the retention mechanism of neutral solutes under CEC and HPLC conditions and observed

lower retention factors for the former mode on a column packed with CEC Hypersil C18 3

µm. They attributed the differences to the heat generation in CEC which causes significant

differences between the set and the actual column temperature. Opposite to the findings

reported above Zhang et al. [33] found the retention behaviour in CEC comparable to that

in HPLC and obtained similar linear energy equations in CEC, PEC and HPLC using linear

solvatation energy relationship analysis. In another report it has also been shown that

HPLC methods for neutral compounds can be easily transferred to CEC [34]. Applying

further identical conditions the comparison of stationary phases in the pressure- and

electro-driven mode may reveal an answer for these partly contradictory findings.

Following this approach this paper seeks to characterize and to compare a number of

reversed-phase stationary phases under pressure versus electrically driven conditions. The

characterization of the columns in both modes was performed using a well-defined

standard test mixture and test procedure described by Galushko [35]. In addition also other

tests were performed for column evaluation.

3.2 Experimental 3.2.1 Columns The columns used in this study are listed in Table 3.1 together with relevant data provided

by the manufacturer. The column packed bed was 25 cm, and 33.5 cm total length. Prior

to use in CEC, the columns were conditioned. This was accomplished by applying 10 bar

pressure on both sides of the column and increasing the voltage from 0-25 kV in 5 kV

steps per 10 min. After that the pressure was increased to 12 bar and a 30 kV voltage was

applied for 10 min. For the micro-HPLC experiments, the columns were conditioned

until the column pressure was stabilized (approx. 1 h). Note that in these experiments the

same columns were tested under pressure- and electro-driven conditions using the same

batches of eluents. All columns were tested preferably in order CEC, HPLC to ensure the

same flow velocity size.

- 68 - Chapter 3

Table 3.1 List of investigated columns.

Column/Stationary phase

Column Diameter

Average Particle size

Nominal Pore Size

Å

Hypersil ODS 75 µm 3 µm 120

CEC Hypersil C18 75 µm 3 µm 120

CEC Hypersil C18 (1) 100 µm 2.5 µm 120

CEC Hypersil C18 (2) 100 µm 2.5 µm 120

Unimicro C18 75 µm 3 µm 300

Unimicro C8 75 µm 3 µm 300

Unimicro Phenyl 75 µm 3 µm 300

Micra NPS ODS 75 µm 3 µm Non-porous

(1), (2) = home-packed columns.

3.2.2 Instrumentation All CEC chromatograms were obtained on an Agilent Technologies 3DCE (Agilent

Technologies GmbH, Waldbronn, Germany) instrument equipped with a pressure facility

of up to 12 bar at the outlet and/or inlet vial. This pressurization option of the instrument

was used to prevent bubble formation in the capillaries. Samples were injected

electrokinetically (5 kV for 2-15 s). For each run a voltage of 20 kV (600 V.cm-1 electric

field strength) was applied with 10 bar pressure on both ends of a capillary. The detection

wavelengths were 210 and/or 254 nm. High voltage was applied as 6-s time ramp to avoid

column stress. The column cassette temperature was maintained at 20ºC.

Micro HPLC separations were carried out on a system consisting of a Phoenix 20 CU

syringe pump (Carlo Erba Instruments, Milan, Italy), a microUVIS20 ultraviolet/visible

absorbance detector (Carlo Erba Instruments, Milan, Italy) operating at 210 or 254 nm, and

an injector with a 200 nL loop (VICI-AG Valco Europe, Schenkon, Switzerland). The flow-

rate was approx. 0.2-0.3 µL/min using a 1/100-flow splitter. The experiments were

performed at air-conditioned laboratory temperature (±21ºC) without additional

thermostating.


3.2.3 Chemicals The buffer consisted of di-sodium tetraborate decahydrate (Merck, Darmstadt, Germany),

dissolved in deionized water and adjusted to pH=8.0 using concentration hydrochloric acid

(Merck, Darmstadt, Germany). Acetonitrile (ACN) and methanol (MeOH) were used as

organic modifiers and were of HPLC supra gradient-grade purity (both from Biosolve,

Valkenswaard, Netherlands). The eluents were prepared by mixing the tetraborate buffer

with an appropriate amount of the organic modifier and degassed (5 min) with helium prior

to use. The same batch of eluent was used to test a specific column in both separation

modes. Ionic strength in the eluent was kept constant at a tetraborate concentration of

1.5 mmol.L-1. The test sample comprised the following compounds: thiourea (t0), phenol,

aniline, benzene, toluene, p-ethylaniline, N,N-dimethylaniline, ethylbenzoate, ethylbenzene,

biphenyl, naphthalene, fluorene, anthracene (all from Merck, Darmstadt, Germany).

Samples were prepared by dissolving these compounds in the mobile phase or in the pure

organic modifier and then diluted with the tetraborate buffer.

3.2.4 Test procedure For the characterization of the RPLC stationary phases under CEC and HPLC conditions

a test procedure described by Galushko, was applied [35]. Opposite to the aqueous

methanol eluents used in the original test in our experiments we applied a tetraborate

buffer pH=8.0 (to guarantee a sufficient electroosmotic flow velocity for all tested

columns together with minimal packing degradation [50]) instead of water. Furthermore,

besides methanol also acetonitrile was used as another organic modifier in these column

tests. Both modifiers were used at various concentrations in the eluents. Unless otherwise

noted the standard test conditions were the following:

Eluent: methanol/aqueous tetraborate buffer pH=8.0 60/40 V/V

Temperature: 20ºC

Test compounds: thiourea (t0), aniline, phenol, benzene, and toluene.

Column descriptors were obtained using the software ChromLife (Merck, Darmstadt,

Germany) and comprised of following parameters:

1. Hydrophobicity (H) = (kbenzene + ktoluene)/2

- 70 - Chapter 3

2. Hydrophobic (methylene) selectivity (HS); retention data of benzene, toluene and

phenol are used to calculate capacity factors of ethylbenzene and propylbenzene

as described [35].

(HS)=kpropylbenzene/kethylbenzene

3. Silanol activity (NI) =1+3×[kaniline/kphenol - 1]

k = retention factor

To preserve a sufficient wetted state of the stationary phase ligands column tests were

performed not below 20% organic modifier. Table 3.2 further summarizes retention times

of thiourea in CEC as a t0 marker for all columns and all mobile phase compositions. Flow

velocity in HPLC mode for a given condition was adjusted to that in CEC mode.

Table 3. 2 Retention times of thiourea in CEC as a t0 marker for all columns measured at several

percentages of methanol and acetonitrile as organic modifier; test compound: thiourea (t0); eluent:

mixtures of methanol or acetonitrile and tetraborate buffer (1.5 mM in total); for other

experimental conditions see text; (-) = data not available ACN (%) MeOH (%) Column

30 40 50 60 70 80 50 60 65 70 75 80

Hypersil ODS

3 µµµµm

4.049

±0.024

3.186

±0.013

3.504

±0.010

3.010

±0.010

2.714

±0.022

- 5.676

±0.029

5.965

±0.023

5.699

±0.014

5.714

±0.010

5.023

±0.017

-

CEC Hypersil C18

3 µµµµm

- 2.048

±0.021

2.085

±0.011

2.340

±0.017

2.318

±0.020

2.400

±0.004

- - - - - -

CEC Hypersil C18

2.5 µµµµm (1)

4.467

±0.025

4.118

±0.024

3.915

±0.015

3.875

±0.013

3.579

±0.021

3.510

±0.015

7.721

±0.027

8.475

±0.039

8.396

±0.033

8.249

±0.036

8.043

±0.035

-

CEC Hypersil C18

2.5 µµµµm (2)

4.065

±0.027

3.582

±0.020

3.604

±0.009

3.603

±0.001

3.444

±0.010

3.285

±0.003

- - - - - -

Unimicro C18

3 µµµµm

- 5.379

±0.013

5.183

±0.006

4.955

±0.018

4.744

±0.024

4.616

±0.012

- 13.014

±0.009

12.880

±0.008

13.078

±0.012

15.588

±0.011

12.106

±0.010

Unimicro C8

3 µµµµm

- 3.374

±0.004

2.269

±0.002

3.149±

0.007

3.119

±0.007

3.047

±0.005

- 15.601

±0.007

15.371

±0.012

15.125

±0.019

14.086

±0.016

13.694

±0.025

Unimicro Phenyl

3 µµµµm

- 4.029

±0.014

4.058

±0.018

3.871

±0.009

3.758

±0.015

3.410

±0.012

9.292

±0.013

9.308

±0.012

9.098

±0.009

8.899

±0.023

8.633

±0.019

8.472

±0.022

Micra NPS ODS

3 µµµµm

5.310

±0.001

5.162

±0.003

5.133

±0.002

- - - 11.075

±0.009

13.004

±0.006

14.149

±0.008

- - -


For details about this test method the reader is referred to paper [35] and references

therein. The other test compounds (PAHs, p-ethylaniline, N,N-dimethylaniline) were used

for additional tests of the RP stationary phases under study. Under all conditions all the

solutes are supposed to behave as neutral compounds. None of them is subjected to

electrophoretic mobility, which has been proved by capillary zone electrophoresis

experiments.

3.3 Results and discussion 3.3.1 Column hydrophobicity and hydrophobic selectivity As an example in Fig. 3.1a and Fig. 3.1b hydrophobic selectivities (HS) obtained on the

Unimicro C18 3 µm and Unimicro Phenyl 3 µm columns under pressure-driven (HPLC)

and electro-driven (CEC) conditions are plotted. Under HPLC conditions for RP-phases of

similar ligand length generally the selectivity of specific increments (e.g. CH2-group) is fairly

constant under constant experimental conditions and decreases with increasing portions of

organic modifier in the eluent [36-38].

1.1

1.2

1.3

1.4

1.5

1.6

1.7

0 10 20 30 40 50 60 70 80 90 100

ACN (%)

HS

(hyd

roph

obic

sel

ectiv

ity)

CEC

HPLC

Figure 3.1a

Hydrophobic selectivity (HS) values for the Unimicro C18 3 µm column under pressure-

and electro-driven conditions; eluent: tetraborate buffer (1.5 mM in total), pH=8.0 and

acetonitrile, detection: 254 nm; test compounds: thiourea (t0), phenol, benzene, toluene; for

other experimental conditions see section 3.2.2 Instrumentation.

- 72 - Chapter 3

In addition, ideally under further similar conditions stationary phases behaving identical

under both HPLC and CEC eluent-drive conditions will show equal to one ratios of

specific chromatographic properties e.g. hydrophobicity or hydrophobic selectivity.

1.1

1.2

1.3

1.4

1.5

1.6

1.7

0 10 20 30 40 50 60 70 80 90 100

ACN (%)

HS

(hyd

roph

obic

sel

ectiv

ity)

CEC

HPLC

Figure 3.1b

Hydrophobic selectivity (HS) values for the Unimicro Phenyl 3 µm column under pressure-

and electro-driven conditions; eluent: tetraborate buffer (1.5 mM in total), pH=8.0 and

acetonitrile, detection: 254 nm; test compounds: thiourea (t0), phenol, benzene, toluene; for

other experimental conditions see section 3.2.2 Instrumentation.

The HS-values on the Unimicro C18 3 µm column (Fig. 3.1a) are in good agreement with

data usually obtained under HPLC conditions on RP-columns. Furthermore, the HS-values

obtained under HPLC and CEC conditions on this column differ not much if at all,

suggesting a similar behaviour of this stationary phase for both separation modes. In

contrast the HS-values obtained at 70 and 80% organic modifier on the Unimicro Phenyl

3 µm column differ significantly up to 23%.

In Fig. 3.2 the CH2-selectivity ratios HSHPLC/HSCEC of all columns are presented together

with the ideal line HSHPLC/HSCEC=1; HSHPLC and HSCEC represent hydrophobic selectivity

obtained in the HPLC and CEC modes, respectively, under further similar experimental

conditions.


0.8

0.85

0.9

0.95

1

1.05

1.1

1.15

0 10 20 30 40 50 60 70 80 90 100

Concentration of org. modifier (%)

HS H

PLC/H

S CEC

Hypersil ODS 3 µm (ACN) Hypersil ODS 3 µm (MeOH) Hypersil CEC ODS 3 µm (ACN)Hypersil CEC ODS 2.5 µm (ACN) -1 Hypersil CEC ODS 2.5 µm (MeOH) -1 Hypersil CEC ODS 2.5 µm (ACN)- 2Unimicro C18 3 µm (ACN) Unimicro C18 3 µm (MeOH) Unimicro C8 3 µm (ACN)Unimicro C8 3 µm (MeOH) Unimicro Phenyl 3 µm (ACN) Unimicro Phenyl 3 µm (MeOH)NPS ODS II 3 µm (ACN) NPS ODS II 3 µm (MeOH)

Figure 3.2 Ratios of CH2-selectivities under pressure- and electro-driven conditions HSHPLC/HSCEC

for all columns for methanol and acetonitrile as organic modifier; eluent: tetraborate buffer

(1.5 mM in total) pH=8.0 and organic modifier (for each column indicated in brackets);

detection: 254 nm; test compounds: thiourea (t0), phenol, benzene, toluene; for other

experimental conditions see section 3.2.2 Instrumentation. ()=HSHPLC/HSCEC=1.

In addition somewhat lower deviations in HS-ratios of up to –8% for the Hypersil ODS

3 µm (ACN), more polar phases such as the CEC Hypersil C18 2.5 µm -2 (ACN),

Unimicro Phenyl 3 µm (MeOH), Unimicro C8 3 µm (ACN and MeOH), and one of the

self-packed CEC Hypersil C18 2.5 µm – No. 2 (ACN), showed deviations in HSHPLC/HS-

CEC-ratios of up to ±10%. Unimicro C8 3 µm column for both acetonitrile and methanol as

organic modifier, and ±6% for the Hypersil ODS 3 µm (ACN) columns were obtained. An

unexplained exception is the 5% and the 23% deviation found for the Unimicro Phenyl

3 µm column using 70 and 80% of acetonitrile, respectively. At all other modifier

concentrations for this column the HS-ratio deviation is maximally 5%. All other columns

showed only moderate deviations of a few percents in their HSHPLC/HSCEC ratios. More

specifically the changes were up to 1.7%, 4% and -2% for the CEC Hypersil C18 3 µm

under ACN-conditions, the NPS ODS 3 µm and the Hypersil ODS 3 µm, respectively,

- 74 - Chapter 3

used with methanol as the organic modifier. Hydrophobicity and hydrophobic selectivity

are related to length, ordering and orientation of the ligands on a substrate’s surface [39,40].

More particularly orientation and ordering will also depend on ligand coverage density and

the nature and concentration of the organic modifier [41]. Since the column tests only

differed in the mode of application (CEC vs. HPLC), we speculate that the observed

deviations in HSHPLC/HSCEC ratios must be attributed to stationary phase changes under

electrical field conditions. This is in agreement with findings of others like Eimer et al. [30],

Wei et al. [31], and Angus et al. [42], who also observed dissimilarities in stationary phase

properties under HPLC and CEC conditions. Obviously, even not all nominally identical

stationary phases respond in a similar way to the application of an electrical field. For

instance, the HS-values of the CEC Hypersil C18 2.5 µm and the CEC Hypersil C18

2.5 µm (No. 2) both under acetonitrile conditions are 10 and 2%, respectively.

In contrast to these relatively small deviations in HS-ratios, much larger differences in

hydrophobicity (H) properties were found. As an example in Fig. 3.3 the hydrophobicity

(H) values obtained under HPLC and CEC conditions for CEC Hypersil C18 3 µm for

acetonitrile as organic modifier are presented. Obviously, for this column hydrophobicity

significantly differs under both these conditions. In addition, column hydrophobicity in the

CEC mode is in all cases smaller than in the HPLC mode under further similar

experimental conditions. This finding is in accordance with earlier results of Eimer et al. [30]

who reported an average decrease of approximately up to 40% in hydrophobicity, when

applying an RP-column under PCEC (pressure-driven CEC) conditions. Furthermore, for

this column the hydrophobicity under HPLC conditions increases relatively more

compared to CEC conditions at decreasing fraction of acetonitrile in the eluent.


0

1

2

3

4

5

6

7

8

9

0 10 20 30 40 50 60 70 80 90

ACN (%)

H (h

ydro

phob

icity

) HPLC

CEC

Figure 3.3

Hydrophobicity (H) values for the CEC Hypersil C18 3 µm column under pressure

(HPLC) and electro-driven (CEC) conditions; eluent: tetraborate buffer (1.5 mM in total)

pH=8.0 and acetonitrile; test compounds: thiourea (t0), benzene, toluene; further

experimental conditions see section 3.2.2 Instrumentation.

In addition, to compare stationary phase behaviour, log k values of benzene vs. percentage

of organic modifier (ACN and MeOH) are plotted for the Unimicro C8 3 µm column in

Figures 3.4a and 3.4b respectively. In this particular case for both separation modes and

both modifiers a linear relationship has been found with regression coefficients r from

0.9988 to 0.9999.

As statistically determined (on the level of significance of 0.05) slopes log k vs. percentage of

acetonitrile are not identical. Figure 3.4b further shows that log kw values of benzene

(capacity factor of benzene extrapolated to pure water as the mobile phase) differ in

methanol as organic modifier. This implicates the differences between both separation

modes and modifiers used.

- 76 - Chapter 3

Figure 3.4

Log k values of benzene for the Unimicro C8 3 µm column under pressure (HPLC) and

electro-driven (CEC) conditions; eluent: tetraborate buffer (1.5 mM in total) pH=8.0 and a)

acetonitrile or b) methanol; test compound: benzene; further experimental conditions see

section 3.2.2 Instrumentation.

In Fig. 3.5 the hydrophobicity ratios HHPLC/HCEC for all columns are presented together

with the ideal line HHPLC/HCEC=1. From these results it is obvious that significant

differences in hydrophobicity occur depending on whether a column is used under HPLC

or CEC conditions. Besides the specific properties of the stationary phase, this also appears

to depend on the nature and concentration of the applied organic modifier. For instance,

deviations in column HHPLC/HCEC ratios are up to 47% for Hypersil ODS 3 µm and -59%

for CEC Hypersil C18 3 µm for acetonitrile as the organic modifier in the buffer. In

addition, deviations in HHPLC/HCEC ratios of up to –25% in ACN and also in MeOH for

the Unimicro C18 3 µm, 31% for the Unimicro C8 3 µm column and up to 22% for CEC

Hypersil C18 2.5 µm (No. 2) both for acetonitrile were observed. Note that these Hypersil

and Unimicro packings have nominal pore sizes of 120 Å and 300 Å, respectively. Within

this limited number of packing materials no clear influence of pore size on HHPLC/HCEC

ratios could be observed. The smallest deviations were found for the CEC Hypersil C18

2.5 µm column (No. 1) (up to 5%) and NPS ODS 3 µm (± 5%for both ACN and MeOH).

log k = -0.0279φ + 1.5843 r = -0.9998

log k = -0.0279φ + 1.6414 r = -0.9999

-1.3

-1.1

-0.9

-0.7

-0.5

-0.3

-0.1

0.1

0.3

0.5

0 10 20 30 40 50 60 70 80 90 100

Concentration of methanol (%)

Log

k(be

nzen

e)

HPLC

CEC

log k = -0.0223φ + 1.1149 r = -0.9988

log k = -0.0237φ + 1.1058 r = -0.9994

-1.5

-1.3

-1.1

-0.9

-0.7

-0.5

-0.3

-0.1

0.1

0.3

0.5

0 10 20 30 40 50 60 70 80 90 100

Concentration of acetonitrile (%)

Log

k(be

nzen

e)

HPLC

CEC

a) b)


0

0.5

1

1.5

2

2.5

0 10 20 30 40 50 60 70 80 90 100


HHPLC/HCEC

Hypersil ODS 3 µm (ACN) Hypersil ODS 3 µm (MeOH) Hypersil CEC ODS 3 µm (ACN)Hypersil CEC ODS 2.5 µm (ACN) -1 Hypersil CEC ODS 2.5 µm (MeOH) -1 Hypersil CEC ODS 2.5 µm (ACN)- 2Unimicro C18 3 µm (ACN) Unimicro C18 3 µm (MeOH) Unimicro C8 3 µm (ACN)Unimicro C8 3 µm (MeOH) Unimicro Phenyl 3 µm (ACN) Unimicro Phenyl 3 µm (MeOH)NPS ODS II 3 µm (ACN) NPS ODS II 3 µm (MeOH)

Figure 3.5 Ratios of hydrophobicities HHPLC/HCEC of the columns under pressure- and electro-driven

conditions using methanol and acetonitrile as organic modifier (for each column indicated

in brackets); ()=HHPLC/HCEC=1; further experimental conditions see section 3.2.2

Instrumentation.

With an exception of CEC Hypersil ODS 2.5 µm (No. 2) HHPLC/HCEC-ratios of the porous

packing materials generally are much closer to one for methanol than acetonitrile as the

organic modifier. These modifiers significantly differ in their hydrogen bond donor capacity

(MeOH; 0.43 vs. 0.15 for ACN) and polarity/polarizability (0.60 for ACN and 0.28 for

methanol) [43]; (normalized values).

These differences in the physico-chemical properties between both these modifiers are

responsible for different state of wettability and interphase layer of the ligands under

normal HPLC conditions [41,44-45]. We assume that in addition to these effects the

application of an electrical field may cause different ligand orientation in the interphase

resulting in the observed differences in HHPLC/HCEC-ratios. An alternative explanation

might be found in the phase ratios of CEC columns under HPLC and CEC conditions. As

- 78 - Chapter 3

earlier discussed by Antle and Ying and co-workers [39, 40] the various phase ratios of RP-

columns often count significantly for the major part of the observed differences in

hydrophobicities between columns rather than differences in distribution coefficients.

Rathore and Horváth discussed the phenomenon of “electroosmotic whirlwind” around

particles and pores [46] of packings under CEC-conditions. It is worthwhile mentioning

that measurining actual phase ratio of columns under different eluent driven conditions in

rather impossible.

These effects may cause limited access of test probes to the stationary phase internal pore

volume contributing to virtually different phase ratios.

It seems likely to assume that this effect will a.o. depends on the mean pore size and more

particularly, on the pore size distribution (fraction micropores versus macropores) in a

specific packing material. Further support for this assumption is found in the close to one

HHPLC/HCEC values observed for the non-porous (NPS) stationary phase in this study. For

this latter type of packing and for both organic modifiers these ratios are 0.95-1.02 over the

entire range of the modifier concentration. We assume that the nearly complete absence of

pores of this NPS phase prevents the occurrence of different phase ratio values, in turn

causing the constant HHPLC/HCEC ratios found.

3.3.2 Silanol activity Silanol activity of RP-columns is a rather empirical term and may include a number of van

der Waals and ion exchange solute to stationary phase interactions [47]. In the Galushko

test applied here, silanol activity is based on the measurement of the ratio of retention

factors of aniline and phenol.

Silanol activity data were calculated as a function of the nature and percentage of organic

modifier and results are summarized in Table 3.3.


Table 3.3 Silanol activity results for all columns measured at several percentages of methanol and

acetonitrile as organic modifier; test compounds: thiourea (t0), aniline, phenol; eluent: mixtures of

methanol or acetonitrile and tetraborate buffer (1.5 mM in total); for other experimental conditions

see text; (-) = data not available ACN (%) MeOH (%) Column Mode

30 40 50 60 70 80 50 60 65 70 75 80

HPLC 0.64 1.02 1.30 1.33 1.63 - 0.04 0.01 0.07 0.24 - - Hypersil ODS

3 µµµµm CEC 0.65 0.75 0.97 1.00 1.00 - 0.02 0.08 0.07 0.26 - -

HPLC - 1.47 1.52 1.97 2.19 2.48 - - - - - - CEC Hypersil C18

3 µµµµm CEC - 2.05 2.46 1.59 2.27 2.87 - - - - - -

HPLC - 1.51 1.76 1.96 2.25 3.00 - 1.26 1.24 - 1.51 - CEC Hypersil C18

2.5 µµµµm (1) CEC - 1.72 1.87 2.12 4.28 3.06 - 2.24 1.96 1.83 1.87 -

HPLC 1.38 1.60 1.82 2.09 2.41 2.94 - - - - - CEC Hypersil C18

2.5 µµµµm (2) CEC 1.16 1.71 1.73 1.67 2.73 2.41 - - - - - -

HPLC - 1.00 1.19 1.53 1.68 1.98 2.18 1.00 1.00 1.00 1.00 1.00 Unimicro C18

3 µµµµm CEC - 1.05 1.32 1.52 1.83 2.30 2.07 0.67 0.83 0.34 1.00 -0.82

HPLC - 1.23 1.54 2.09 2.27 2.73 - 1.00 1.00 -0.06 1.00 1.00 Unimicro C8

3 µµµµm CEC - 1.00 1.37 1.66 2.17 6.87 - 0.53 0.71 1.00 1.00 1.00

HPLC - 1.50 1.81 2.56 5.35 1.00 1.51 1.93 2.54 2.94 3.39 4.42 Unimicro Phenyl

3 µµµµm CEC - 1.59 1.93 2.79 10.27 10.39 5.81 6.90 7.98 15.70 16.18 19.57

HPLC 1.00 1.00 1.00 - - - 1.00 1.00 1.00 - - - Micra NPS ODS

3 µµµµm CEC 1.00 1.00 1.00 - - - 1.00 1.00 1.00 - - -

In Fig. 3.6 the ratios of silanol activities NI(HPLC)/NI(CEC) measured on each column

are plotted together with the ideal line NI(HPLC/NI(CEC) = 1. From these results it can

immediately be seen that for some columns the silanol activity ratios vary substantially as

a function of the nature and fraction of the organic modifier in the buffer. For instance,

the NIHPLC/NICEC-ratios of Hypersil ODS 3 µm and Unimicro C18 and C8 3 µm with

methanol vary substantially over the investigated concentration range. In contrast for

- 80 - Chapter 3

these columns under acetonitrile conditions these ratios are smoother and less

pronounced. Note that from these columns under acetonitrile conditions the Hypersil

ODS 3 µm shows an NIHPLC/NICEC-ratio > 1. In all other cases depending on nature and

concentration of the modifier NI-ratios larger or smaller than one were found. In contrast

to the findings mentioned above other columns in this set showed for both modifiers

smoother and much less pronounced NIHPLC/NICEC-ratios over the investigated modifier

concentration range. E.g. for Unimicro phenyl 3 µm and Hypersil CEC ODS 2.5 µm (1)

for both modifiers rather smooth and constant ratios were found. This with an exception

for Unimicro Phenyl 3 µm in the 60-80% acetonitrile range. Note that the NIHPLC/NICEC-

ratio is < 1 under all conditions for the latter column.

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

0 10 20 30 40 50 60 70 80 90 100


NI HPLC/NI CEC

Hypersil ODS 3 µm (ACN) Hypersil ODS 3 µm (MeOH) Hypersil CEC ODS 3 µm (ACN)Hypersil CEC ODS 2.5 µm (ACN) -1 Hypersil CEC ODS 2.5 µm (MeOH) -1 Hypersil CEC ODS 2.5 µm (ACN)- 2Unimicro C18 3 µm (ACN) Unimicro C18 3 µm (MeOH) Unimicro C8 3 µm (ACN)Unimicro C8 3 µm (MeOH) Unimicro Phenyl 3 µm (ACN) Unimicro Phenyl 3 µm (MeOH)NPS OD II 3 µm (ACN) NPS ODS II 3 µm (MeOH)

Figure 3.6 Ratios of silanol activity values NI(HPLC)/NI(CEC) of the columns under pressure- and

electro-driven conditions using methanol and acetonitrile as organic modifier (for each

column indicated in brackets); ()=NI(HPLC)/NI(CEC)=1, further experimental

conditions see section 3.2.2 Instrumentation.

For NPS ODS no difference in retention of aniline and phenol was found for both

modifiers. Consequently NIHPLC/NICEC-ratios were one over the entire modifier

concentration range. Obviously the silanol activity of this latter column type is rather


independent from the modifier nature (acetonitrile vs. methanol) and operating

conditions (HPLC vs. CEC).

From the results reported in Table 3.3 and Fig. 3.6 it is clear that for porous RP-packing

materials silanol activity generally is not independent of the mode of operation and the

applied modifier. For example, for the Unimicro Phenyl column under 50% methanol

conditions silanol activity is 1.51 and 5.81 under HPLC and CEC conditions, respectively.

In addition, for the same column substantially different silanol activities of 1.81 and 1.51

for 50% acetonitrile and methanol, respectively, under the same HPLC-mode can be

observed.

Earlier studies have shown that under pressure-driven conditions silanol activity of

porous RP columns may depend substantially on the nature of the modifier, e.g. whether

methanol or acetonitrile is used in the eluent [48,49 and refs. citated therein]. Our results

from the present study are in agreement with these earlier findings. In addition the same

appears to be true for porous packing materials under CEC-conditions.

Similar as for hydrophobicity in the Galushko test silanol activity is determined from

retention factors of two different compounds (see section 3.2.4). Referring to the

previous section on column hydrophobicity, we believe that the observed differences in

silanol activity for porous packings must also be attributed to:

i. Apparent differences in phase ratios under HPLC and CEC-conditions caused by

electroosmotic whirlwind effects.

ii. Different ligand orientation and thus silanol accessibility under both HPLC and

CEC modes.

Furthermore note that for porous packings from the results in Fig. 3.6 and Table 3.3 it

can be concluded that the differences and changes in silanol activity under CEC and

HPLC conditions are much more pronounced for methanol rather than acetonitrile.

In contrast to the observations for porous stationary phases the non-porous NPS ODS

3 µm packing showed remarkably different silanol activity behaviour. Irrespective of the

modifier (methanol or acetonitrile) or the applied mode (HPLC vs. CEC) silanol activity

values of one were observed for all experiments. Again this might be taken as evidence

that the absence of electroosmotic whirlwind effects in these solid packings are

responsible for the more similar behaviour under different conditions (modifier and mode

of operation) compared to porous packing materials. In Table 3.4 two parameters, USP

tailing factor and plate number of aniline, are given both separation modes and eluent

- 82 - Chapter 3

containing 70% of acetonitrile as the organic modifier (40% of acetonitrile for the NPS

ODS 3 µm column).

Table 3.4 USP tailing factor and plate numbers of aniline in CEC and HPLC for all columns

measured at 70 percent (40 percent for the NPS ODS 3 µm column) acetonitrile as organic

modifier; test compound: aniline; eluent: 70 percent of acetonitrile and tetraborate buffer (1.5 mM

in total); for other experimental conditions see text.

Column Mode USP Tailing factor

Plate Number /column (half-width method)

HPLC 1.206 13115 Hypersil ODS

3 µµµµm CEC 1.208 19800

HPLC 1.225 14602 CEC Hypersil C18

3 µµµµm CEC 1.257 18878


2.5 µµµµm (1) CEC 1.329 22281


2.5 µµµµm (2) CEC 1.311 22239

HPLC 1.258 16052 Unimicro C18

3 µµµµm CEC 1.205 36532

HPLC 1.229 10571 Unimicro C8

3 µµµµm CEC 1.075 26323

HPLC 1.092 17776 Unimicro Phenyl

3 µµµµm CEC 1.390 30157

HPLC 1.067 23205 Micra NPS ODS

3 µµµµm CEC 1.087 28590

Special attention has been made to the comparison of the electroosmotic flow (EOF)

(Table 3.2) and the chromatographic behaviour of the silanol sensitive compound

(aniline) as long as remaining silanol groups of stationary phase packing are responsible

for the EOF. It can be concluded that columns under this particular condition generating

lower EOF gave higher efficiencies (the Unimicro C18 3 µm, the Unimicro Phenyl 3 µm

and NPS ODS 3 µm columns). No direct relationship between EOF and USP tailing

factor of aniline can be drawn, which is with the exception of the Unimicro C18 3 µm and


the Unimicro C8 3 µm columns higher in CEC mode. Finally, no clear relationship

between silanol activity, average pore diameter and EOF [51] of the packing could be

derived from the present results.

3.4 Conclusions Under pH = 8 condition and using methanol and acetonitrile as modifiers eight columns

packed with seven different RP-phases have been tested under electro- and pressure driven

conditions. With an exception for the Unimicro Phenyl column methylene (hydrophobic)

selectivity differed not substantially between both modes: maximally 10% for the CEC

Hypersil C18 2.5 µm (No. 2) column. The limited ligand chain length can easily explain the

strongly deviating results for the former column.

In contrast for porous stationary phases substantial differences in the major column

descriptors, hydrophobicity and silanol activity, were found between the HPLC- and CEC-

modes. In addition, these differences were also a function and for some cases strongly

dependent on the nature and concentration of the applied modifier. These observations can

probably be explained from different ligand orientations caused by the conditions of both

eluent-driven modes. Alternative explanation of these findings may be found in the

occurrence of electroosmotic whirlwind effects in porous packing causing different phase

ratios under batch mode conditions.

Contrasting for the non-porous stationary phase for the HPLC and CEC eluent driven

mode nearly similar hydrophobicity and silanol activity data were measured. This has also

been taken as additional evidence for the electroosmotic whirlwind effect in porous

packings. The results of the present study confirmed the differences usually found in silanol

activity between methanol and acetonitrile as the organic modifier under HPLC condition.

These effects, however, appeared to be more manifest under CEC conditions. It was also

found that generally the use of acetonitrile delivered smoother HPLC/CEC ratio curves

versus percentage of modifier than methanol.

Finally, the results of this study clearly show that at least for porous stationary phases the

transfer of existing HPLC methods to CEC analysis protocols is not straightforward.

- 84 - Chapter 3

REFERENCES 1. J.H. Knox, I.H. Grant, Chromatographia, 32 (1991) 317.


3. V.L. McGuffin, M. Novotný, J. Chromatogr., 255 (1983) 381.

4. A. Berloni, A. Cappiello, G. Famiglini, P. Palma, Chromatographia, 39 (1994) 279.

5. J.P. Chervet, M. Ursem, J.P. Salzmann, R.W. Vannoort, J. High. Resolut. Chromatogr.,

12 (1989) 278.

6. J. Vissers, J.P. Chervet, J.P. Salzmann, Int. Lab., 26 (1996) 12A.

7. J.P.C. Vissers, H.A. Claessens, C.A. Cramers, J. Chromatogr. A, 779 (1997) 1.

8. M.R. Euerby, C.M. Johnson, K.D. Bartle, LC-GC Int., 11 (1998) 39.

9. I.H. Grant, in: Methods Molecular Biology, Vol. 52, K. Altria (ed), Humana Press Inc.,

Totowa, NJ 1996, p. 197-209.

10. R.J. Boughtflower, T. Underwood, C.J. Peterson, Chromatographia, 40 (1995) 329.


12. A.S. Lister, J.G. Dorsey, D.E. Burton, J. High. Resolut. Chromatogr., 20 (1997) 523.

13. T. Hanai, H. Hatano, N. Nimura, T. Kinoshita, J. High. Resolut. Chromatogr., 14

(1991) 481.

14. J.W. Jorgenson, K.D. Lukacs, J. Chromatogr., 218 (1981) 209.




18. C. Yan, R. Dadoo, R.N. Zare, D.J. Rakestraw, D.S. Anex, Anal Chem., 68 (1996) 2726.


(1996) 403.


21. T.S. Stevens, H.J. Cortes, Anal. Chem., 55 (1983) 1365.


23. C. Yan, D. Schaufelberger, F. Erni, J. Chromatogr. A, 670 (1994) 15.

24. G. Ross, M. Dittmann, F. Bek, G. Rozing, Am. Lab. (Shelton, Conn.), 28 (1996) 34.

25. S. Lüdtke, T. Adam, K.K. Unger, J. Chromatogr. A, 786 (1997) 229.

26. C.L. Rice, R. Whithead, J. Phys. Chem., 69 (1965) 4017.

27. R. Dadoo, C. Yan, R.N. Zare, D.S. Anex, D.J. Rabestraw, G.A. Hux, LC-GC, 15 (1997)

630.



29. J.P.C. Vissers, H.A. Claessens, P. Coufal, J. High Resolut. Chromatogr., 18 (1995) 540.

30. T. Eimer, K.K. Unger, J. van der Greef, TrAC, Trends Anal. Chem., 15 (1996) 463.


Relat. Technol, 21 (1998) 1433.

32. N.M. Djordjevic, P.W.J. Fowler F. Houdiere, G. Lerch, J. Liq. Chromatogr. Relat.

Technol., 21 (1998) 2219.


34. G. Ross, M.M. Dittmann, G.P. Rozing, publication number 5965-9031E 1997,

Hewlett-Packard, Waldbronn, Germany.

35. S.V. Galushko, Chromatographia, 36 (1993) 39.

36. R.M. Smith (Ed.), Retention and Selectivity in Liquid Chromatography; Prediction,

Standardization and Phase Comparison, Chapter 8, J. Chromatogr. Libr., 57, Elsevier,

Amsterdam, 1995.

37. H. Figge, A. Deege, J. Köhler, G. Schomburg, J. Chromatogr., 351 (1986) 393.

38. Cs. Horváth, High Performance Liquid Chromatography, Advances and Perspectives,

Vol. 2, Academic Press, New York, 1980.

39. P. Antle, A.P. Goldberg, L.R. Snyder, J. Chromatogr., 321 (1985) 1.

40. P.T. Ying, J.G. Dorsey, Talanta, 38 (1991) 237.

41. K.B. Sentell, J.G. Dorsay, Anal. Chem., 61 (1989) 930.

42. P.D.A. Angus, E. Victorino, K.M. Payne, C.W. Demarest, T. Catalano, J.F. Stobaugh,


43. L.R. Snyder, P.W. Carr, S.C. Rutan, J. Chromatogr. A, 656 (1993) 537.

44. U.D. Neue, HPLC columns, theory, technology and practice, Wiley-VCH, Inc., New

York 1997.

45. J.G. Dorsey, K.A. Dill, Chem. Rev., 89 (1989) 331.


47. J. Nawrocki, J. Chromatogr. A, 779 (1997) 29.

48. H.A. Claessens, E.A. Vermeer, C.A. Cramers, LC-GC Eur., 6 (1993) 692.

49. D.V. McCalley, R.G. Brereton, J. Chromatogr. A, 828 (1998) 407.

50. H.A. Claessens, M.A. van Straten, J.J. Kirkland, J. Chromatogr. A, 728 (1996) 259.


- 86 - Chapter 3

- 87 -

CHAPTER 4 4 CHROMATOGRAPHIC PROPERTIES OF REVERSED

PHASE STATIONARY PHASES UNDER PRESSURE AND

ELECTRO DRIVEN CONDITIONS; EFFECT OF BUFFER

COMPOSITION

Summary

Four different reversed-phase (RP) stationary phases (CEC Hypersil C18, Zorbax Rx SIL

C18, Zorbax 300 Rx SIL C18 and Zorbax PSM1000/C18) were examined under high-

performance liquid chromatographic (pressure-driven, HPLC), and capillary

electrochromatographic (electro-driven, CEC) conditions using an acetonitrile mobile

phase combined with twenty different buffer systems (different cations, anions, pH

and/or ionic strengths). Chromatographic performance tests under HPLC and CEC

conditions were carried out using acidic, basic and neutral polar/non-polar compounds.

Parameters such as plate number, retention factor and asymmetry were used to describe

the behaviour of the RP-columns under both HPLC and CEC conditions. The buffer

systems differently influence chromatographic characteristics of porous RP-phases under

CEC and HPLC conditions. Thus the choice of an appropriate buffer can be critical for

an application applied on an entire system.

This chapter has been published: T. Jiang, J. Jiskra, H.A. Claessens, C.A. Cramers, J. Chromatogr. A, 923 (2001) 215.

- 88 - Chapter 4 4.1 Introduction Occurrence and stability of the electroosmotic flow (EOF) strongly depend on the nature

and composition of the applied buffer. Moreover, in both modes capillary

electrochromatography (CEC) as well as in high-performance liquid chromatography

(HPLC) chromatographic properties and the analysis results are also strongly and in a

different way determined by the buffer properties. Therefore, apart from the organic

modifier [1], also the buffer composition must be critically considered, e.g. its nature, ionic

strength, buffering capacity and stability (depletion). However, several additional problems

encountering in CEC systems should be considered, too:

i. most RP stationary phases are not able to generate sufficient and/or stable EOF

since the residual amount of free silanol groups is limited

ii. effect of Joule heating limits the use of more concentrated buffer solutions

iii. use of low concentration buffers causes problems with buffering capacity, double

layer overlap and buffer depletion.

These problems mentioned above can be solved in principle by (i) development of new

stationary phases [2-21] or (ii) by use of organic buffers that have low ionic mobilities

(lower current and/or lower Joule heating) compared to inorganic buffers. This makes

especially zwitterionic buffers as e.g. 2-(N-morpholino)ethanesulonic acid (MES) attractive

for CEC eluents [19,22-31]. However, tris(hydroxymethyl) aminomethane (Tris), one of the

most used buffers, should not be included into the group of zwitterionic buffers because of

its insignificant dissociation constant of CH2-OH group in water containing systems. Tris

has been reported as a base [32,33] and has been misinterpreted as zwitterionic in e.g.

[24,34]. In case of mass spectrometry detection volatile organic buffers as ammonium

acetate, triethylammonium (TEA) acetate, acetic and formic acid are usually applied [35-46].

Ammonium acetate, lithium acetate together with MES are particularly used in non-

aqueous CEC as long as they are sufficiently soluble in non-aqueous media [25-26,47-48].

Furthermore, to prevent and to suppress tailing effects, nitrogen-containing buffers with

their ability of shielding silanol groups are used in separation of nitrogen containing

compounds e.g. pharmaceuticals [49-51], amino acids, peptides and proteins [22,35,52-57]

and nucleosides [58]. Among all inorganic buffers, tetraborate and phosphate buffers are

most often used [e.g. 28-29,59-72,78]. They are widely applied at least for the reason that

the buffers are often based on HPLC protocols and used for CE methods. They suffer

from problems pointed out in (i), (ii) and (iii); phosphate is also known as a factor


promoting degradation of silica supports substantially. However, formation of the complex

of phosphate with silanols is taken also as an advantage in CE separations [73]. Other

additives in mobile phases in CEC comprise aminoacids (β-alanine, γ-aminobutyric acid),

chiral selectors (cyclodextrins such as HPβCD) [74] and EOF modifiers/stabilizers (sodium

dodecylsulfate, SDS) [75-78].


by the manufacturer. The column packed bed was 25 cm, and 33.5 cm total length.

Prior to use in the CEC mode, the columns were conditioned. This was accomplished by

applying 10 bar pressure on both sides of the column and increasing the voltage from 0 to

25 kV in 5 kV steps per 10 min. Next to that the pressure was increased to 12 bar and a

30 kV voltage was applied for 10 min. For the micro-HPLC experiments, the columns

were conditioned until the column pressure was stabilized (approximately 1 h). Note that

in these experiments the columns were tested under pressure- and electro-driven

conditions using the same batches of eluents.

Table 4.1 List of investigated columns; each column diameter was 100 µm

Column Average Particle

Size

Pore Size

Pore Volume

Surface Area

Carbon Load

CEC Hypersil C18

3 µm 130 Å 0.65 cm3.g-1 170 m2.g-1 8.5%

Zorbax Rx SIL C18

5 µm 80 Å 0.45 cm3.g-1 180 m2.g-1 12.4%

Zorbax 300 Rx SIL C18

5 µm 300 Å 0.42 cm3.g-1 45 m2.g-1 3.6%

Zorbax PSM1000/C18

5 µm 800 Å 0.38 cm3.g-1 15 m2.g-1 1.2%

- 90 - Chapter 4 All columns were supplied in double (same batch) and tested simultaneously in CEC and

HPLC. HPLC experiments were adjusted to similar flow velocities as obtained in CEC. As

a consequence the HPLC experiments are not optimized with respect to plate height.





electrokinetically (5 kV for 2-15 s). For each run a voltage of 20 kV (600 V.cm-1 electrical

field strength) was applied with 10-bar pressure on both ends of a capillary. The detection

wavelength was 210 nm. High voltage was applied as 6-s time ramp to avoid column stress.

The column cassette temperature was maintained at 20ºC.


syringe pump (Carlo Erba Instruments, Milan, Italy), a microUVIS20 ultraviolet/visible

absorbance detector (Carlo Erba Instruments, Milan, Italy) operating at 210 nm, and an

injector with a 200 nL loop (VICI-AG Valco Europe, Schenkon, Switzerland). The flow-

rate was approx. 0.2-0.3 µL/min using a 1/100-flow splitter (VICI-AG Valco Europe,

Schenkon, Switzerland). The experiments were performed at air-conditioned laboratory

temperature (about 2 ºC) without additional thermostating.

4.2.3 Chemicals Buffers used in the experiments are listed in Table 4.2 providing data of the manufacturer

and concentrations and/or pHs used. All compounds were of analytical purity grade.

Acetonitrile (ACN) of HPLC supra gradient-grade purity was used as organic modifier

(Biosolve, Valkenswaard, Netherlands). The eluents were prepared by mixing

corresponding buffer system with an appropriate amount of the organic modifier and

degassed ultrasonically for 15 min prior to use. The same batch of eluent was used to test a

specific column in both separation modes. The test sample comprised the following

compounds: thiourea (t0), phenol, aniline, benzene, toluene, dimethyl phthalate, diethyl

phthalate, biphenyl and o-terphenyl (Merck, Darmstadt, Germany and Aldrich, Steinheim,


Germany). Samples were prepared by dissolving these compounds in the mobile phase or

in the pure organic modifier and then diluted with water.

Table 4.2 List of buffers used.

Inorganic buffers Organic buffers

Name Specifications Name Specifications

Ammonium tetraborate (Sigma-Aldrich, Milwaukee, USA)

0.5 mM, pH=8.0 Ammonium acetate (Merck, Darmstadt, Germany)

1 mM, pH=7.0

Lithium tetraborate (Merck, Darmstadt, Germany)

0.5 mM, pH=8.0 Citric acid, sodium salt (Merck, Darmstadt, Germany)

1 mM, pH=7.0

Potassium tetraborate (Sigma-Aldrich, Milwaukee, USA)

0.5 mM, pH=8.0 Glycine, sodium salt (Merck, Darmstadt, Germany)

1 mM, pH=8.5

Sodium carbonate (Merck, Darmstadt, Germany)

1 mM, pH=8.0 2-(N-Morpholino)-ethanesulfonic acid (MES) (Serva, Heidelberg, Germany)

1 mM, pH=6.0

Sodium phosphate (Merck, Darmstadt, Germany)

1 mM, pH=8.0,

7.0, 6.0

Triethylamine acetate (Fluka, Buchs, Switzerland)

1 mM, pH=7.0

Sodium tetraborate (Merck, Darmstadt, Germany)

0.1 mM, 0.2 mM,

0.5 mM and

1 mM, pH=8.0

Tris(hydroxymethyl)- aminomethane (TRIS) (Merck, Darmstadt, Germany)

0.5 mM, 1 mM,

5 mM and 10

mM, pH=8.0

- 92 - Chapter 4 4.2.4 Test procedure Characterization of RP-stationary phases was carried out using acetonitrile as the organic

modifier combined with twenty different buffer systems (different type, pH, ionic strength).

Chromatographic performance under HPLC and CEC conditions was measured using

acidic, basic and polar/non-polar compounds as test samples. Parameters such as retention

factor, plate number or peak asymmetry were recorded in order to compare the behaviour

of RP-columns under both HPLC and CEC conditions.

Under all conditions all solutes are supposed to behave as non-charged compounds.

Thus, none of them is subjected to electrophoretic mobility, which has been confirmed

by capillary zone electrophoresis experiments.

4.3 Results and discussion 4.3.1 Polar compounds Silica based RP-stationary phases are the most used phases in HPLC analysis. Depending

on the manufacturing procedure, the remaining silanol groups can provide sufficient

electroosmotic flow for CEC. As known from HPLC, undesired interactions of the silica

support with polar compounds such as organic bases (pharmaceuticals etc.) could also be a

great drawback in capillary electrochromatography. It has been found out that isolated

silanol groups are mainly responsible for strong interactions with basic compounds. In

order to suppress such interferences e.g. the influence of buffer system has been thoroughly

studied in HPLC analysis of basic compounds [79]. In contrast, most of the residual silanol

groups are believed to be hydrogen bonded and claimed to be not effective in peak

deterioration viz. tailing processes. However, the behaviour of these silanols under electrical

field conditions may be different compared to their properties under pressure-driven

conditions. Under CEC conditions the choice of an appropriate buffer system, its

concentration and/or pH are critical parameters in the development of reliable analysis

protocols for CEC. In Table 4.3 the silanol activities of the columns measured in different

concentrations of sodium tetraborate buffer under CEC and HPLC conditions are

presented.


Table 4.3 Values of silanol activities (NI =1+3×[kaniline/kphenol - 1]) [82] measured for the

columns in different concentrations of sodium tetraborate buffer (values given are total

concentrations) and acetonitrile as organic modifier (30/70 V/V).

Columns Mode 0.1 mM Na2B4O7

0.2 mM Na2B4O7

0.5 mM Na2B4O7

1 mM Na2B4O7

Silanol activity value

CEC 14.17 3.58 3.06 2.74 CEC Hypersil C18 HPLC 2.20 2.23 2.29 2.37

CEC 1.73 1.65 1.56 1.44 Zorbax Rx SIL C18 HPLC 1.64 1.58 1.61 1.64

CEC 1.33 1.39 1.63 1.13 Zorbax 300 Rx SIL C18

HPLC 1.43 1.46 1.39 1.42

CEC 1.07 1.57 1.49 1.05 Zorbax PSM1000/C18

HPLC 1.07 1.23 1.17 1.38

The implications of the buffer systems are demonstrated in Figure 4.1.

For the 0.5 mM buffer a substantial shift of the aniline peak in CEC can be observed. In

addition, severe tailing of the peak of aniline occurs at the lower buffer concentration, too,

suggesting that the active places on the silica surface are not fully shielded by the buffer. In

contrast, Joule heating strongly influences efficiency of apolar compounds in the CEC

mode. The plate number for the benzene peak is 32 000 plates/ column in 0.5 mM sodium

tetraborate buffer but only 23 000 plates/column in 1 mM tetraborate concentration.

Under HPLC conditions the observed drop in plate number going from the 1 mM to

0.5 mM buffer concentration was less than 1% (15 500 plates/column for 1 mM sodium

tetraborate and 17 500 plates/column for 0.5 mM sodium tetraborate, respectively).

- 94 - Chapter 4

0

2

4

6

8

10

12

0 2 4 6 8 10 12 14

Time (min)

Res

pons

e

CEC

HPLC

Thiourea

Phenol

Aniline

Benzene

Toluene

a)

0

2

4

6

8

10

12

14

16

0 2 4 6 8 10 12

Time (min)

Res

pons

e

CEC

HPLC

Thiourea

Phenol

Aniline

Benzene

Toluene

b)

Figure 4.1 Chromatograms of test mixture containing thiourea, aniline, phenol, benzene and toluene

under HPLC and CEC mode using the CEC Hypersil C18 stationary phase and acetonitrile

mobile phase with different ionic strength, a) 1 mM and b) 0.5 mM sodium tetraborate

buffer, respectively.

Similarly to ionic strength also the selection of the buffer cation may influence the activity

of silanols. To illustrate the effect, in Figure 4.2 the ratios of retention factors in HPLC vs.

CEC for several cations are presented.


Li+Na+

K+NH4+

CEC Hypersil C18

Zorbax PSM1000/C18

Zorbax 300 Rx SIL

Zorbax Rx SIL C18

0.6

0.7

0.8

0.9

1

1.1

1.2

k(HPLC)/k(CEC)

0.5 mM tetraborate

Figure 4.2 Ratios of retention factors of aniline kHPLC/kCEC of the columns under HPLC and CEC

conditions using acetonitrile as organic modifier and 0.5 mM tetraborate buffer (70/30,

V/V) with different cations.

The retention decreasing power of cations in CEC is in the order Li+<Na+<K+< +4NH .

From the same data it can be further concluded that ammonium salts generally show the

best silanol shielding potential power. This because ammonium ions can easily interact with

silanols either due to ion exchange, hydrogen-bonding or hydrophobic forces (for primary,

secondary, tertiary or quaternary ammonium salts) [80]. However, not all nitrogen-

containing buffers show the same shielding effects. For example as presented in Figure 4.3,

the adsorption of zwitterionic buffers such as glycine onto the stationary phase can even

cause dissimilar effects.

The data suggest that this behaviour might originate from the adsorption on the polarized

silica surface and the exposure of the acidic part of the buffer molecule to the analyte as

schematically depicted in Figure 4.4.

- 96 - Chapter 4

-2

0

2

4

6

8

10

12

0 2 4 6 8 10 12

Time (min)

Res

pon

se

Thiourea

Phenol

Aniline

Benzene Toluene

Figure 4.3 Capillary electrochromatography of mixture containing thiourea (t0), phenol, aniline,

benzene and toluene. Mobile phase: acetonitrile/1 mM (total conc.) glycine buffer pH=8.5

(70/30, V/V).

Figure 4.4

Possible adsorption of the zwitterionic

buffers on the silica surface and their

influence on the chromatographic

performance of the basic compounds, a)

C8 modified silica surface, b) C8

modified silica surface with adsorbed

glycine, c) C8 modified silica surface

with adsorbed 2-(N-morpholino)-

ethanesulfonic acid.

4.3.2 Apolar compounds The chromatographic behaviour of apolar compounds under CEC conditions with no

acidic or basic centers in the molecule was also investigated. The partitioning and/or

adsorption mechanisms on bonded phases in CEC are still poorly understood. At a first

SiOO

O

OSi

Si

Si

H3N+ COO-

SiOO

O

OSi

Si

Si

SiOO

O

OSi

Si

Si

N+

O

H

SO3-

a) b)

c)


glance, we may assume similar separation mechanisms as in HPLC. On the other hand the

occurrence of electroosmosis causing through-pore flows may suggest that on particular

RP-phases different separation mechanisms are present. In addition, another consideration

may be the consequent differences in the phase ratios occurring under pressure- and

electro-driven conditions. As an example, Figure 4.5 illustrates the influence of eight buffer

systems (organic and inorganic) on the chromatographic behaviour of o-terphenyl. Under

CEC and HPLC conditions the ratios of retention factors are plotted against the buffer

systems for all four columns under investigation.

1 mM N

a tetr

abora

te

1 mM N

a carb

onate

1 mM N

a citra

te

1 mM TRIS

1 mM N

H4 ace

tate

1 mM TEAC

1 mM M

ES

1 mM gl

ycine

CEC Hypersil C18 Zorbax PSM1000/C18

Zorbax 300 Rx SIL Zorbax Rx SIL C18 0

0.2

0.4

0.6

0.8

1

1.2

1.4

k(HPLC)/k(CEC)

Figure 4.5 Ratios of retention factors of o-terphenyl kHPLC/kCEC of the columns under HPLC and

CEC conditions using acetonitrile as organic modifier and different organic and inorganic

buffers mixed in ratio 70/30 (V/V). Total concentration of the buffers is given in the plot.

For each column, much smaller variations of kHPLC/kCEC values for non-polar than for

polar compounds within the column are observed (relative standard deviation or RSD for

the CEC Hypersil C18, Zorbax Rx SIL C18 and Zorbax 300 Rx SIL columns is up to 1.5%

and for the Zorbax PSM1000/C18 column is 5%, respectively).

- 98 - Chapter 4 In contrast to polar compounds, zwitterionic buffers have nearly no influence on the

chromatographic behaviour of apolar compounds in electrochromatography. Similar

findings were also observed for other apolar compounds as benzene and toluene present in

the test mixture. However, the kHPLC/kCEC values differ significantly between the columns.

We assume that this may be attributed to phase ratios differences due to CEC effects in the

pores [81].

Based on the discussion in the section 4.3.1 on ammonium salts it might be expected that

all three factors (ion exchange, hydrogen-bonding or hydrophobic forces) may contribute

to the adsorption on to the stationary phase. If we presume that the stationary phase

becomes more polar in CEC due to the alteration in silanols, even more triethylammonium

can be loaded to the stationary phase. As a consequence this might further contribute to the

changes in partitioning of apolar compounds in the CEC mode. The same is true if the

concentration of nitrogen containing buffers in the mobile phase is increased. Figure 4.6

illustrates for several concentrations of the Tris buffer and toluene as the test compound.

0.5 mM1 mM

5 mM10 mM

CEC Hypersil C18

Zorbax PSM1000/C18

Zorbax 300 Rx SIL

Zorbax Rx SIL C18

0.9

0.95

1

1.05

1.1

1.15

1.2

k(HPLC)/k(CEC)

c(Tris)

Figure 4.6 Ratios of retention factors of toluene kHPLC/kCEC of the columns under HPLC and CEC

conditions using acetonitrile as organic modifier and Tris buffer (70/30, V/V) in different

total concentration.

Obviously, a general trend of increasing ratios of retention factor in HPLC vs. CEC as a

function of higher buffer concentration can be observed. The greatest change is observed


on the high-porous Zorbax PSM 1000/C18 stationary phase where the ratio of retention

factors (kHPLC/kCEC) increases from 1.0 for 0.5 mM Tris buffer concentration to 1.1 for

10 mM Tris, respectively.

The typical influence of organic buffers on the efficiency of the biphenyl peak is given in

Table 4.4.

Table 4.4 Reduced plate heights calculated from biphenyl peak. Columns Mode 1 mM

Na2B4O7 1 mM

phosphate, pH=7.0

1 mM Tris 5 mM Tris 1 mM sodium citrate

Reduced plate height

CEC 3.5 3.8 3.5 3.1 3.6 CEC Hypersil C18 HPLC 4.5 4.3 3.9 4.3 4.1

CEC 2.8 3.0 2.8 2.6 2.9 Zorbax Rx SIL C18 HPLC 5.1 5.7 5.6 5.6 5.6

CEC 5.0 4.4 3.9 4.0 4.1 Zorbax 300 Rx SIL C18 HPLC 4.5 4.6 4.1 4.0 4.5

CEC 2.9 2.2 3.8 2.5 3.2 Zorbax PSM1000/C18 HPLC 4.0 3.6 4.2 4.0 3.5

The mean values of efficiency (as reduced plate heights) of three consecutive injections are

presented (RSD less then 4%). Generally, similar or higher efficiencies under the CEC

mode have been found for organic buffers, this particularly on stationary phases with lower

porosity (the CEC Hypersil C18 and Zorbax Rx SIL C18 stationary phases). This finding

was observed also for other organic buffers and apolar solutes used. However, the high

porous Zorbax PSM1000/C18 showed better efficiency under the CEC mode when

inorganic buffers were applied.

- 100 - Chapter 4

pH=8.0pH=7.0

pH=6.0

CEC Hypersil C18

Zorbax PSM1000/C18

Zorbax 300 Rx SIL

Zorbax Rx SIL C18

0

0.2

0.4

0.6

0.8

1

1.2

k(HPLC)/k(CEC)

sodium phosphate

Figure 4.7 Ratios of retention factors of toluene kHPLC/kCEC of the columns under HPLC and CEC

conditions using acetonitrile as organic modifier and sodium phopshate buffer (70/30,

V/V) at different pH (pH of the phosphate buffer measured prior mixing with acetonitrile).

The influence of inorganic buffers on the ratios of retention factors of apolar compounds is

less than 0.5% RSD. Figure 4.7 demonstrates that also differences of pH using sodium

phosphate as the inorganic buffer and toluene as the test analyte has minor influence on

partitioning of apolar compounds in general. However, similar to Fig. 4.5 differences in

kHPLC/kCEC ratios between the columns are observed, too.

4.4 Conclusions The influence of commonly used organic and inorganic buffers on the chromatographic

behaviour of reversed-phase stationary phases under pressure-driven (HPLC) and electro-

driven (CEC) conditions has been investigated in this study. Four RP stationary phases

were tested under both conditions using acetonitrile as organic modifier mixed with

appropriate buffer systems.

As a result, inorganic buffers were found to have greater impact on the chromatographic

behaviour compared to organic buffers. Concentration viz. ionic strength of inorganic

buffers generally influences retention behaviour of polar compounds. Again, changes in

cation type and/or size have an impact on retention behaviour of polar compounds

particularly on mixed-mode stationary phases (the CEC Hypersil C18 stationary phase) and


phases with medium porosity. Using sodium phosphate buffer systems adjusted to different

pHs, the well-known influence on electroosmotic flow has been observed. Moreover, peak

asymmetries of polar compounds were closer to one with a phosphate buffer of pH=7.0.

In the field of organic buffers, zwitterionic buffers showed exceptional behaviour. Their

acido-basic moieties have a great effect on behaviour of electron-donor compounds as

amines. Further, the efficiencies were found better in organic buffer systems compared with

inorganic buffers of the same concentration on stationary phases with lower porosity (the

CEC Hypersil C18 and Zorbax Rx SIL C18 stationary phases). In contrast, the high porous

Zorbax PSM1000/C18 stationary phase showed much higher efficiencies with inorganic

buffers and organic buffers of higher ionic strength.

Finally, as a consequence of these above-mentioned findings a proper choice of buffer

systems for capillary electrochromatography is a critical step in CEC analysis.

Acknowledgement The authors gratefully acknowledge Dr. G. Rozing from Agilent Technologies, Germany

for providing the Zorbax columns.

REFERENCES 1. A. Dermaux, P. Sandra, Electrophoresis, 20 (1999) 3027.

2. I. Gusev, X. Huang, Cs. Horváth, J. Chromatogr. A, 855 (1999) 273.

3. J.J. Pesek, M.T. Matyska, C.A. Kein, Spec. Publ. - R. Soc. Chem., 235 (1999) 97.

4. J. Liu, H. Zou, J.Y. Ni, Y. Zhang, Anal. Chim. Acta, 378 (1999) 73.

5. C. Fujimoto, Anal. Chem., 67 (1995) 2050.

6. J.-L. Liao, N. Chen, C. Ericson, S. Hjertén, Anal. Chem., 68 (1996) 3468.




10. M. Zhang, C. Yang, Z. El Rassi, Anal. Chem., 71 (1999) 3277.

11. M.T. Matyska, Chem. Anal. (Warsaw), 43 (1998) 637.




- 102 - Chapter 4 15. R. Alicea-Maldonando, L.A. Colón, L.A. Electrophoresis, 20 (1999) 37.

16. K. Jinno, J. Wu, H. Sawada, Y. Kiso, J. High Resolut. Chromatogr., 21 (1998) 617.

17. Z.J. Tan, V.T. Remcho, J. Microcolumn Sep., 10 (1998) 99.

18. Q. Tang, N. Wu, M. Lee, J. Microcolumn Sep., 11 (1999) 550.

19. J.J. Pesek, M.T. Matyska, S. Cho, J. Chromatogr. A, 845 (1999) 237.

20. Q. Tang, B. Xin, M. Lee, J. Chromatogr. A, 837 (1999) 35.

21. K. Hosoya, H. Ohta, K. Yoshizako, K. Kimata, T. Ikegami, N. Tanaka, J. Chromatogr.

A, 853 (1999) 11.

22. K.D. Altria, N.W. Smith, C.H. Turnbull, J. Chromatogr. B, 717 (1998) 341.

23. C.G. Bailey, C. Yan, Anal. Chem., 70 (1998) 3275.

24. R.J. Boughtflower, T. Underwood, C.J. Paterson, Chromatographia, 40 (1995) 329.

25. A. Dermaux, P. Sandra, V. Ferraz, Electrophoresis, 20 (1999) 74.

26. A. Dermaux, P. Sandra, M. Ksir, K. Fellat, F. Zarrouck, J. High Resolut. Chromatogr.,

21 (1998) 545.


28. M.M. Dittmann, G.P Rozing, J. Microcolumn Sep., 9 (1997) 399.

29. M.R. Euerby, C.M. Johnson, S.F. Smyth, N.C. Gillot, D.A Barrett, P.N. Shaw, J.



31. C. Wolf, P.L. Spence, W.H. Pirkle, E.M. Derrico, D.M. Cavender, G.P. Rozing, J.

Chromatogr. A, 782 (1997) 175.

32. R.G. Bates, H.B. Hetzer, J. Phys. Chem., 65 (1961) 667.

33. J.H. Fossum, P.C. Markunas, J.A. Riddick, Anal. Chem., 23 (1951) 491.

34. K.D. Altria, I.H. Grant, LC-GC Intl., 1998, 588.

35. G. Choudhary, Cs. Horváth, J.F. Banks, J. Chromatogr. A, 828 (1998) 469.

36. J. Ding, P. Vouros, Anal. Chem., 69 (1997) 379.

37. J. Ding, T. Barlow, A. Dipple, P. Vouros, J. Am. Soc. Mass Spectrom., 9 (1998) 823.

38. M. Hugener, A.P. Tinke, W.M.A. Niessen, U.R. Tjaden, J. van der Greef, J.

Chromatogr., 647 (1993) 375.

39. S.J. Lane, A. Pipe, Rapid Commun. Mass Spectrom., 12 (1998) 667.


41. G.A Lord, D.B Gordon, L.W. Tetler, C.M.J. Carr, J. Chromatogr. A, 700 (1995) 27.


42. C.J. Paterson, R.J. Boughtflower, D. Highton, E. Palmer, Chromatographia, 46 (1997)

599.

43. C. Rentel, P. Gfrörer; E. Bayer, Electrophoresis, 20 (1999) 2329.


45. V. Spikmans, S.J. Lane, U.R. Tajden, J. van der Greef, Rapid Commun. Mass

Spectrom., 13 (1999) 141.

46. R.N. Warriner, A.S. Craze, D.E. Games, S.J. Lane, Rapid Commun. Mass Spectrom.,

12 (1998) 1143.

47. P. Sandra, A. Dermaux, V. Ferraz, M.M. Dittmann, G.P. Rozing, J. Microcolumn Sep.,

9 (1997) 409.

48. L. Roed, E. Lundanes, T. Greibrokk, Electrophoresis, 20 (1999) 2373.

49. N.C. Gillot, M.R. Euerby, C.M. Johnson, D.A Barrett, P.N. Shaw, Anal. Commun., 35

(1998) 217.


51. J. Reilly, M. Saees, J. Chromatogr. A, 829 (1998) 175.

52. P. Huang, J.-T. Wu, D.M. Lubman, Anal. Chem., 70 (1998) 3003.



55. J.-M. Lin, T. Nakagama, K. Uchiyama, T. Hobo, J. Pharm. Biomed. Anal., 15 (1997)

1351.

56. J.J. Pesek, M.T. Matyska, L. Mauskar, J. Chromatogr. A, 763 (1997) 307.

57. J.J. Pesek, M.T. Matyska, S. Swedberg, S. Udivar, Electrophoresis, 20 (1999) 2343.

58. T. Helboe, S.H. Hansen, J. Chromatogr. A, 836 (1999) 315.

59. R. Asiaie, X. Huang, D. Farnan, Cs. Horváth, J. Chromatogr. A, 806 (1998) 251.


61. H. Engelhardt, S. Lamotte, F.-T. Hafner, Am. Lab. (Shelton, Conn.), 30 (1998) 40.

62. C.G. Huber, G. Choudhary, Cs. Horváth, Anal. Chem., 69 (1997) 4429.

63. H. Jakubetz, H. Czesla, V. Schurig, J. Microcolumn. Sep., 9 (1997) 421.

64. D. Li, V.T. Remcho, J. Microcolumn. Sep., 9 (1997) 389.

65. V. Lopez-Avila, J. Bedenedicto, C. Yan, J. High Resolut. Chromatogr., 20 (1997) 615.

66. S. Mayer, V. Schurig, J. High Resolut. Chromatogr., 15 (1992) 129.


- 104 - Chapter 4 68. M.M. Robson, S.C.P. Roulin, S.M. Shariff, M.W. Raynor, K.D. Bartle, A.A. Clifford, P.


69. M.R. Schure, R.E. Murphy, W.L. Klotz, W. Lau, Anal. Chem., 70 (1998) 4985.


71. Q.-H. Wan, J. Chromatogr. A, 782 (1997) 181.

72. C. Yan, R. Dadoo, H. Zhao, R.N. Zare, D.J. Rakestraw, Anal. Chem., 67 (1995) 2026.

73. P. Jandik, G. Bonn, Capillary Electophoresis of Small Molecules and Ions, VCH

Publishers, 1993.

74. W. Wei, G.A Luo, R. Xiang, C. Yan, J. Microcolumn Sep., 11 (1999) 263.

75. R.M. Seifar, S. Heemstra, W.T. Kok, J.C. Kraak, H. Poppe, Biomed. Chromatogr., 12

(1998) 140.

76. R.M. Seifar, W.T. Kok, J.C. Kraak, H. Poppe, Chromatographia, 46 (1997) 131.

77. R.M. Seifar, J.C. Kraak, H. Poppe, W.T. Kok, J. Chromatogr. A, 832 (1999) 133.

78. R.M. Seifar, J.C. Kraak, W.T. Kok, H. Poppe, J. Chromatogr. A, 808 (1998) 71.

79. S.H. Hansen, P. Helboe, M. Thomsen, J. Chromatogr., 544 (1991) 53.

80. J. Nawrocki, J. Chromatogr. A, 799 (1997) 29.

81. U. Tallarek, E. Rapp, T. Scheenen, E. Byer, H. van As, Anal. Chem., 72 (2000) 2292.


- 105 -

CHAPTER 5 5 PREPARATION AND CHARACTERIZATION OF

MONOLITHIC POLYMER COLUMNS FOR CAPILLARY


Summary

A series of micro-monolithic columns with different porosities were prepared for capillary

electrochromatography (CEC) by in-situ copolymerization of butyl methacrylate, ethylene

glycol dimethacrylate, and 2-acrylamido-2-methyl-1-propanesulfonic acid in the presence

of a porogen in fused silica capillaries of 100 µm I.D. Different column porosities were

obtained by changing the ratios of monomers to porogenic solvents. Columns were

investigated and evaluated under both pressure-driven (high-performance liquid

chromatography, HPLC) and electro-driven (capillary electrochromatography, CEC)

conditions. Each column exhibited different efficiency and dependency on flow velocity

under electro-driven conditions. Abnormally broad peaks for some relatively bulky

molecules were observed. Possible explanations are discussed. The differences in column

efficiency and retention behaviour between the two eluent-driven modes were studied in

detail. In addition, other column properties, such as morphology, porosity, stability and

reproducibility, were extensively tested.

This chapter has been published: J. Jiskra, T. Jiang, H.A. Claessens, C.A. Cramers, J. Microcolumn Sep., 12 (2000) 530.

- 106 - Chapter 5 5.1 Introduction Capillary electrochromatography (CEC) has emerged as a promising micro-separation

technique that combines chromatographic selectivity with high efficiency and

miniaturization potential of capillary electrophoresis (CE) [1-6]. To date, a number of

papers have been published on the theoretical and practical aspects of CEC [3, 7-30].

However, some technical problems including column preparation have slowed down its

development and widespread implementation [31-32]. Most columns used nowadays in

CEC are packed in a similar way as in micro-HPLC. High experimental skill and

experience is required to reproducibly pack micron-sized particles into a narrow-bore

tube and to prepare a solid and stable frit at both ends of the column. Additionally, it is

extremely difficult to avoid side effects arising from frits [17], i.e., bubble formation and

heterogeneity of the electrical field across the column. Furthermore, there are only a few

commercially available stationary phases dedicated for CEC which are expected to

provide sites for the required interactions as well as the charged group for generation of

the electroosmotic flow, EOF [30]. All these technical problems have stimulated the

development of various alternative approaches [33-48]. Among these, one competing

strategy is to prepare monolithic media by in-situ polymerization within the confine of a

mold. This approach has recently attracted substantial attention in view of its capability to

eliminate the difficulties described above for making packed columns [37-49]. Presently,

there exist two main types of monolithic materials, silica sol-gels [37-40] and porous

organic polymers. In principle, the polymeric approach exhibits more potential

advantages and a more promising future compared to its silica based counterparts. This is

due to the simpler preparation process, higher efficiency, easier pore size control, and

more facile adaptability to adjust column selectivity, as well as the direction and the speed

of the electro-osmotic flow (EOF) offered by polymer rods. Several groups have spent

considerable effort in this field [41-49]. However, this technique is still virtually in its

infancy. Until now, most applications have concentrated on simple small neutral

compounds. Moreover, the physico-chemical properties of these porous polymers under

electric field conditions are still not straightforward to be explained.

In this study, we prepared a series of poly-alkylmethacrylate based monolithic capillary

columns with different permeabilities. The column performances under both CEC

(electro-driven, ED) and HPLC (pressure-driven, PD) were evaluated and compared in

detail. In addition, the micropore size distributions in these polymeric monoliths

Preparation and Characterization of Monolithic … - 107 -

were investigated, and some practical and theoretical problems associated with these types

of rods are also discussed.

5.2 Experimental 5.2.1 Chemicals Butyl methacrylate (BMA), ethylene glycol dimethacrylate (EGDMA), 2-acrylamido-2-

methyl-1-propanesulfonic acid (AMPS) and 3-(trimethoxysilyl)propyl methacrylate were

obtained from Aldrich. Acetonitrile (ACN) and tetrahydrofuran (THF) with HPLC grade

purity were from Biosolve (Valkenswaard, the Netherlands). Polystyrene standards were

obtained from DuPont (Wilmington, DE) and all the analytes and other chemicals were

from Merck (Darmstadt, Germany). Fused-silica tubing (100 µm I.D.) was purchased from

Polymicro Technologies (Phoenix, AZ, USA). The mobile phase was composed of 20%

(V/V) sodium phosphate buffer (5 mM, pH 7) and 80% (V/V) ACN. It was degassed by

ultrasonication and filtered through a filter (pore size = 0.45 µm) before use.

5.2.2 Column preparation The monomers (60.0 wt% BMA, 39.7 wt% EGDMA and 0.3 wt% AMPS),

azobisisobutyronitrile (AIBN) used as a polymerization initiator (1 wt% with respect to the

total monomer amount) and porogen (1-propanol, 1,4-butanediol and 10 wt% water) were

mixed ultrasonically into a homogenous solution [50-52]. For different columns, different

porogen concentrations and volume fractions were used, as detailed in Table 5.1.

Subsequently, the reactant solution was purged with nitrogen for 3 min before a small part

of the reactant mixture was introduced into a capillary (unmodified or silanized) by a 10 µL

syringe. The capillary was either filled completely or to a certain distance from one end.

After both ends of the capillary were sealed in a micro-connector, it was kept at 60°C in an

oven for 24 h. All columns were conditioned by mobile phase using a syringe pump prior

to HPLC and CEC experiments.

- 108 - Chapter 5 Table 5.1 Characteristics of different columns.

Column Porogen

fraction (vol%)1)

Conc. of 1-propanol in porogen

(wt%)

Capillary Globule size

(nm)2)

Porosity

(εεεεT)3)

C%4) H in CEC

(µµµµm) 5)

a 47 62 unmodified 150 0.50 0.0 6.7

b 52 62 unmodified 250 0.56 0.0 8.3

c 57 62 unmodified 300 0.61 0.3 10.0

d 62 62 unmodified 800 0.67 2.0 11.0

e 67 62 unmodified 1200 0.72 3.2 33.3

f 62 60 unmodified 1200 NM NM 30.0

g 62 63 unmodified 450 NM NM 12.0

h 57 62 modified 300 NM 0.1 10.0

1) Porogen volume fraction in reactant solution. 2) Microglobule size estimated from SEM pictures. 3) NM: not measured 4) Compressibility, C%= 100× (L0-L1)/L0. L0: initial column length; L1: column

length after high pressure was applied. 5) Plate height (H) under CEC conditions. Column efficiency was calculated by half

peak-width method.

To investigate eventual wall effects, the inner wall of some capillaries was modified before

the introduction of the reactant mixture [53]. The procedure was as follows: the

unmodified fused silica capillary was first washed then filled with 1M sodium hydroxide

solution, both ends were put in a sealed vial filled with sodium hydroxide solution, and the

capillary was then kept at 95°C for 2 h in an oven. Thereafter, the capillary was washed with

ca. 60 column volumes of deionized water and then with the same volume of acetone. The

capillary was dried at 60°C under purging nitrogen for 1 h, followed by rinsing with ten

column-volumes of a silanizing solution, a solution of 50% (V/V) of 3-

(trimethoxysilyl)propyl methacrylate in N,N-dimethylformamide (DMF) containing 0.02%

(w/V) of hydroquinone (an inhibitor). After both ends of the capillary were sealed, it was


heated in an oven at 100°C for about 8 h, and then washed with DMF and acetone. Finally,

the capillary was again dried with a nitrogen stream.

5.2.3 Instrumentation

CEC experiments were performed on a 3DCE system (Agilent Technologies, Waldbronn,

Germany) equipped with a DAD 1050 UV detector and an external pressure device for

CEC. Control of the chromatographic system and data acquisition were carried out by a

ChemStation system. Samples were injected electrokinetically. The cassette temperature was

set at either 21 or 30°C. The detection wavelength was 205 nm. During each run 8 bar

pressure was applied at two ends of the column.

The µ-HPLC system was composed of an ISCO model 100 DM syringe pump (Isco, Inc.,

Lincoln, NE, USA), a microUV-Vis SSI 500 detector (Scientific Systems, Inc., State

College, PA, USA) and a Rheodyne injector (VICI AG, Valco Europe, Schenkon,

Switzerland) with a 100 nL internal loop. A TEE-piece after injector with a 1 m × 25 µm

(I.D.) capillary was used as a splitter. The split ratio was typically 100:1. All the

experiments were performed at room temperature (ca. 21°C).

The column morphology was studied by a Model JSM-840A scanning electron microscope

(SEM) (JEOL, Inc., Tokyo, Japan). Samples of 5-mm long rod pieces were cut from the

columns, placed on an aluminium stub via a double-sided carbon tape, and sputter-coated

with a gold/palladium alloy using SPI Sputter for 4 min at 30 mA. The measurements were

carried out at 10 kV at a filament current of 40 mA. According to SEM pictures the

macropore sizes in the capillary rods were estimated. The micropore size distribution of the

polymer was measured by BET nitrogen sorption method on an ASAP 2010 instrument

(Micromeritis, Morcross, GA, USA). The equilibrium interval was 30 s, and the low-

pressure dose was 0.5 g/cm3 STP. The calculation for the micropore size is based on the

slit pore model. The samples were polymers made in a 2 mL bottle by the same

polymerization process as used for the corresponding columns. Before measurement, the

bulk polymer was cut into small pieces, Soxhlet-extracted with methanol for 12 h, vacuum-

dried for 2 days and degassed for 5 h at 65°C.

- 110 - Chapter 5 5.3 Results and discussion 5.3.1 Column efficiency in CEC The lowest plate height (H) at the test alkyl benzenes for each column is presented in Table

5.1. From comparing the efficiencies of columns d, f and g with different 1-propanol

content but the same porogen volume fraction, it is clearly shown that column efficiency is

very sensitive to the amount of 1-propanol. This is consistent with earlier reported results

[50-52]. The column with 62 wt% 1-propanol provided the highest efficiency. Therefore,

for all other columns in this study, the porogen with 62 wt% of 1-propanol was chosen.

Figure 5.1 shows the plots of plate height as a function of eluent velocity for alkyl benzene

compounds on columns a~e prepared by using different porogen volume fractions. As

porogen volume fraction decreases from 67% (column e) to 47% (column a), the column

efficiency increases from about 30,000 to 150,000 plates/m (Table 5.1), the steepness of all

plots decreases, and the optimal flow velocity increases from less than 1 to about 6

cm/min. In addition, it is interesting to note that, for columns a and b with a lower porogen

content, the column efficiency remains almost constant, regardless of the different retention

factors (k) for all of the test compounds with retention factor (k) up to 6.5. This suggests

that with such columns rapid separations can be obtained without loss in resolution.

In order to understand the differences in column efficiency in columns a~e, the polymer

morphology in the capillaries was examined by scanning electron microscope (SEM). As an

example, some SEM photos are shown in Figure 5.2. According to the SEM pictures, the

connecting microglobule size of the polymer were estimated, and presented in Table 5.1.

The structures of the various monolithic columns differ significantly, and strongly depend

on the porogen content in the reactant solutions. With the decrease of the porogen content,

the microglobules become smaller (from about 1 µm for column e down to 150 nm for

column a), and the globule stacking and the channel distribution become more uniform.

Obviously, for the columns with lower porogen concentration, the Eddy diffusion is

smaller and the mass transfer is faster, resulting in higher efficiencies and less steep Van

Deemter curves.


0 2 4 6 8 10 12

100

200

300

400

500

600

700

800

900

plat

e he

ight

(µ

m)

Flow velocity (cm/min)

0 2 4 6 8

10

20

30

40

50

Pla

te h

eigh

t (µ

m)


0 2 4 6 8

10

20

30

40

50

Pla

te h

eigh

t (µ

m)


0 1 2 3 4 5 6 7

10

20

30

40

50

Pla

te h

eigh

t (µ

m


0 1 2 3 4 5 6

10

20

30

40

50

Pla

te h

eigh

t (µ

m)


d

b

c

e 1 2 3 4

a

1 - toluene 2 - propyl benzene 3 - amylbenzene 4 - octylbenzene

Figure 5.1 Plots of plate height as a function of flow velocity in CEC mode for columns a~e.

Conditions: the active lengths of column a~e are 40, 40, 40, 25, 21 cm, respectively; sample:

thiourea (t0 marker), toluene, propylbenzene, amylbenzene and octylbenzene; injection:

5 kV for 3 s; applied voltages: for column a~d, 30, 25, 20, 15, 10, 5 and 2.5 kV, for column

e, 20, 15, 10, 5, 2.5 and 1.5 kV; cassette temperature, 30°C; eluent: 20% (V/V) sodium

phosphate buffer (5 mM, pH 7) and 80% (V/V) ACN.

- 112 - Chapter 5

Figure 5.2 Scanning electron microscope (SEM) pictures of the columns a, d and e.

5.3.2 Selectivity and retention in CEC Apart from column efficiency, column selectivity and the retention mechanism for columns

a~e were investigated as well. The retention mechanism of reversed-phase liquid

chromatography (RPLC) of benzene derivatives for these polymer monoliths has been

described in literature [50-52]. In the first instance, reversed-phase retention of neutral

compounds is determined by the hydrophobicity of the stationary phase, which can be

estimated by the intercept of the linear plot of carbon numbers of alkyl substitutes of

alkylbenzenes as their log k values under similar eluent conditions. For example from Figure

5.3 column a clearly is more hydrophobic compared to column b. Furthermore, in Figure

5.3, the plots for all columns have the same slope, suggesting that all the columns made

from different ratio of porogen to monomers show similar CH2- selectivity [54].

a d e


0 2 4 6 8 10-0.4

0.0

0.4

0.8

abcde

log

k

Carbon number of alkyl in alkylbenzene

Figure 5.3 Plots of the carbon numbers of the alkyl substitutes of alkylbenzene vs. their log k-values

for the columns a-e. Applied separation voltage is 25 kV. Other conditions same as in

Figure 5.1.

Figure 5.4 presents the effect of porogen content on the distribution constant (K) of each

test compound on the different columns. K-values were calculated by using the

experimentally obtained retention factors and the phase ratios (β) of the columns (K = kβ,

β-values were calculated from column porosity, εT). Obviously, the tested mono-substituted

benzene derivatives have similar Ks on the different columns. As a result, the retention

mechanism for these mono-substituted benzene compounds is the same. In contrast, for

multi-substituted benzenes, especially with bulky groups, e.g., 1,3,5-triisopropylbenzene

(TIPB, compound No. 7), big difference in retention between column e and columns a~d is

observed. On column e, TIPB is eluted extremely fast, even earlier than benzene; 1,3-

diisopropylbenzene (DIPB) also is earlier eluted in comparison with the other columns. It

appears that, besides the reversed-phase retention, column e simultaneously exhibits size

exclusion retention for relatively bulky compounds.

- 114 - Chapter 5

0

1

2

3

4

45 50 55 60 65

Porogen volume fraction (%)

Dis

tribu

tion

cons

tant

(K)

1 2 3 4 5 6 7

a b c d e

Compounds:

Figure 5.4. Effect of porogen content on the distribution constant (K) of each test compound. Solutes

are thiourea, benzylalcohol (1), benzene (2), propylbenzene (3), butylbenzene (4), 1,3-

diisopropylbenzene (5), amylbenzene (6) and 1,3,5-triisopropylbenzene (7). Phase ratios

were calculated from: β = εT/(1-εT) where εT is porosity, see Table 5.1. Other conditions

same as in Figure 5.1.

The micropore size distribution of the polymers was investigated by BET nitrogen sorption

method as described in the experimental section. As shown in Figure 5.5, the size range of

micropores is similar for columns a~d, around 10 Å.


0

0,0004

0,0008

0,0012

0,0016

0 5 10 15 20 25 30 35 40 45 50

Pore diameter (Å)

Diff

eren

tial p

ore

volu

me

(cm

3 .g-1

.Å-1

)abcde

Figure 5.5 Micropore size distribution for columns a~e measured by BET nitrogen sorption; slit pore

model is used for micropore calculation.

Column e, however, has a wider range of pore size distribution ranging from 8 to 30 Å.

Furthermore, the total amount of micropores in columns a~e is significantly different.

Generally, at higher porogen contents a less amount of micropores is formed. For example,

from Figure 5.5, column a contains the highest amount of micropores while column e has

almost no micropores and its surface area is only 4.6 m2/g (column a 41.3 m2/g; column b

33.9 m2/g; column c 13.6 m2/g column d 5.4 m2/g). Theoretically, micropores in monolithic

column have the same function as the inner pores in conventional porous packing particles,

viz. they are the main areas where chromatographic retention occurs. However, if the

analytes are relatively large compared to the micropores, such as TIPB with sizes of

9×9×4.5 Å, the probability of these compounds to enter micropores becomes smaller due

to steric hindrance and thus bulky analytes are prone to partial exclusion, which is more

obvious on columns with lower amount of micropores. To illustrate this, on column e,

TIPB is eluted much earlier than on the other more micropores containing columns (Fig.

5.4). Considering that the polymers in the measurements of BET nitrogen sorption were in

the dry state, size exclusion liquid chromatography (SEC) was used to further investigate

the size exclusion behaviours of compounds such as TIPB and DIPB by using

- 116 - Chapter 5 tetrahydrofuran (THF) as the mobile phase. In Figure 5.6 ∆t-values of various analytes are

studied where ∆t equals the value of t-te where t is the retention time of a specific analyte

and te is the retention time of polystyrene with a molecular weight of 2,700,000. Apparently,

due to size exclusion, the ∆t values of TIPB and DIPB are smaller than these of

comparable small molecules. Obviously from Figure 5.6, on columns with decreasing

porogen content (columns e>d>c>b>a), ∆t becomes larger and the resolution for the test

compounds is somewhat better.

0

1

2

3

4

5

6

7

8

0 2 4 6 8 10

∆t (min)

Com

poun

d nu

mbe

r

b c d eColumns:

Figure 5.6

∆t-values (retention time of analyte minus the retention time of polystyrene MW 2.700.000,

t-te) for different analytes under SEC conditions on columns b~e. Samples: 1. polystyrene

(MW, 540); 2. TIPB; 3. o-terphenyl; 4. DIPB; 5. triphenylene; 6. biphenyl; 7. 1,3,5-

trimethylbenzene. Mobile phase: THF.

This results from the gradually larger number of micropores of these columns, which also is

in good agreement with the BET results. Additionally, from the larger ∆t value of TIPB

(9×9×4.5 Å) compared to triphenylene (9.2×9.2×2 Å), it can be concluded that the


micropores in the polymer rods are of a slit rather than a cylinder type of shape. Therefore,

the slit model was chosen to calculate micropore size of our rod columns.

As discussed above, columns a~d primarily show a reversed-phase retention mechanism.

As an example, a chromatogram of alkylbenzenes on column c is given in Figure 5.7.

Interestingly, abnormally broad peaks for DIPB and especially TIPB were obtained; similar

broad peaks were also observed on columns a, b and d (plate height of TIPB: column a, 24

µm; column b, 39 µm and column d, 11500 µm).

-5

5

15

25

35

45

55

65

75

0 5 10 15 20 min

8

1

2 3

4

5

6

7

Figure 5.7 CEC chromatogram on column c. Sample is composed of thiourea (1), benzene (2), toluene

(3), ethylbenzene (4), propylbenzene (5), DIPB (6), TIPB (7) and heptylbenzene (8).

Applied voltage, 30 kV. Other conditions are the same as in Figure 5.1.

To illustrate this in more detail, in Figure 5.8 the plate height vs. flow velocity for TIPB,

DIPB and heptylbenzene are plotted. The Van Deemter C-term values for these

compounds are 109.1, 6.4 and 2.4×10-4 min, respectively, which were calculated from the

fitted curve. TIPB shows an extremely slow mass transfer, which leads to a substantial

dependence of H on flow velocity of these compounds. On the contrary, small molecule

heptylbenzene shows very low velocity dependence. As a consequence, H values are

satisfactory and between 21 and 35 µm in the optimal velocity span. Similar plate height

function was also observed on the other columns. We assume that the slow mass transfer

of TPIB results from the small micropores (ca. 1 nm) in the polymer rod that are too small

for TIPB to easily enter and exit.

- 118 - Chapter 5

0 2 4 6 80

200

400

600

800

1000

1,3,5-Triisopropylbenzene 1,3-Diisopropylbenzene Heptylbenzene

Plat

e he

ight

(µm

)

Flow velocity (cm.min-1)

Figure 5.8. Plots of plate height as a function of flow velocity for column c. Sample consists of TIPB

(1), DIPB (2) and heptylbenzene (3). Other conditions are the same as in Figure 5.1.

Further evidence can be obtained from Figure 5.9 of the plots of plate height vs.

temperature. For TIPB, there is a strong dependency of the plate height (H) on

temperature: H decreases dramatically with the increasing temperature, and the plate height

of DIPB also shows similar tendency. However, the temperature dependency of H, for the

small compounds, is very small. It must be emphasized here that the linear relationship

between voltage and current suggests that Joule heating is negligible (Fig. 5.10). We believe

that Joule heating affects both bulky and small compounds to the same extend. Therefore,

in our opinion Joule heating did not cause the abnormal broadening peaks of bulky

molecules.


0

100

200

300

400

500

15 25 35 45 55

Temperature ( ºC)

Plat

e he

ight

(µm

)

4

3

2 1

Figure 5.9 Effect of temperature on plate height for column c. Solutes: 1. thiourea; 2. propylbenzene;

3. DIPB; 4. TIPB. Other conditions are the same as in Figure 5.1.

As a result, it is obvious that for bulky molecules mass transfer is the predominant factor to

plate height. Under high temperature, the mass transfer can be greatly speeded up by the

increase of their diffusion coefficients, and consequently the column efficiency can be

enhanced substantially.

- 120 - Chapter 5

0

5

10

15

20

25

30

0 1 2 3 4 5 6 7 8

Current (µµµµA)

Volta

ge (k

V)

Figure 5.10 Plot of current versus voltage (absolute values) for column d. Conditions are the same as in

Figure 5.1.

5.3.3 Comparison between HPLC and CEC Columns were also investigated under pressure-driven mode. Figure 5.11 shows the Van

Deemter curves for thiourea under both CEC and µ-HPLC conditions on column c. The

calculated Van Deemter parameters A and C are 6.0 µm, 5.7×10-4 min under pressure- and

5.6 µm, 1.7×10-4 min under electro-driven conditions, respectively. As expected, the plate

height under electro-driven conditions is much lower than that in the pressure mode.

Furthermore, the plate height differences between two modes become larger with the

increase of the flow rate due to the flat flow profile under electro-driven mode.


0 1 2 3 4 5 6 7 80

10

20

30

40

50

60

CEC

HPLC

Plat

e he

ight

(µm

)

Linear flow velocity (cm.min-1)

Figure 5.11 Comparison of Van Deemter curves under electro- and pressure-driven conditions on

column c. Column c: 41.5 cm (8.5 cm from detection window to outlet); sample, thiourea;

temperature, 21 °C; other conditions are the same as in Figure 5.1.

As an example, a comparison of chromatograms on column c for various test compounds

under both modes is shown in Figure 5.12. From the peak widths, the difference in

column efficiency obtained under two eluent driven modes can be easily seen. The plate

number under electro-driven conditions in Figure 5.12A is more than twice of that in the

pressure-driven mode.

- 122 - Chapter 5

0 5 10 15 20 25 30 35

HPLC

CEC1 2

3

4

56

7

8 9 10 1211

A

0 5 10 15 20 25

1

13

HPLC

CEC

B

min Figure 5.12 Comparison of chromatograms on column c under electro- and pressure driven conditions.

Column c, 41.5 cm (8.5 cm from detection window to outlet). Sample: 1. thiourea; 2.

pyridine; 3. benzylalcohol; 4. phenol; 5. aniline; 6. p-ethylaniline; 7. nitrobenzene; 8.

benzene; 9 hexanophenone; 10. chlorobenzene; 11. propylbenzene; 12. biphenyl; 13.

benzylamine. Separation conditions: temperature, 21 °C; µµµµ-HPLC, pressure drop: A, 150

bar; B, 185 bar. CEC, applied voltage: A, +11.5 kV; B, +15 kV.


Typical values for the plate numbers (plates per meter) of peaks 2 to 12 are 40,000 under

HPLC and 105,000 under CEC conditions. The retention factors of each compound in

CEC (kCEC) and µ-HPLC (kHPLC) and their ratios (kHPLC/kCEC) are listed in Table 5.2.

Table 5.2 Values of kHPLC, kCEC and kHPLC/kCEC of the test compounds in Figure 5.12.

Compounds 2 3 4 5 6 7 8 9 10 11 12 13

kHPLC 0.14 0.20 0.32 0.38 0.50 0.62 0.87 1.08 1.26 1.54 1.79 0.11

kCEC 0.13 0.19 0.30 0.37 0.48 0.60 0.84 1.04 1.22 1.48 1.72 0.20

kHPLC/kCEC 1.05 1.04 1.04 1.03 1.04 1.04 1.04 1.04 1.03 1.04 1.04 0.55

For the apolar and polar compounds in the test (Figure 5.12A), the kHPLC/kCEC and the

peak symmetry values (not shown) are close to one. These compounds show similar

retention behaviours under both HPLC and CEC conditions. For compounds, however,

with higher polarity, e.g. benzylamine in Figure 5.12B, the columns show completely

different performance under the two eluent driven modes. Under CEC conditions, the

benzylamine peak is extremely broad and tailing and it elutes later than under HPLC

conditions. It is noteworthy to see that under the same conditions the other basic

compounds such as pyridine and p-ethylaniline show high-efficient and symmetric peaks

(Figure 5.12A). Clearly, this must be attributed to the pKa values of these compounds. The

pKa values of p-ethylaniline and pyridine are approximately 5; the pKa value of benzylamine

is significantly larger (9.3). In the mobile phase of 80% acetonitrile and 20% phosphate

buffer of pH 7 benzylamine is partly positively charged, which has been verified by capillary

zone electrophoresis (CZE) experiments. Due to its electrophoretic mobility

superimposition on the liquid chromatographic elution, benzylamine was expected to leave

the column earlier under CEC condition compared to HPLC mode. However, this was not

observed in our experiments. We speculate that charged groups on the polymer were

polarized and orientated in the electrical field. Under such a condition, the polymer may

become more polar, and the interaction between benzylamine and polymeric stationary

phase may be much stronger.

- 124 - Chapter 5 Table 5.3 shows the retention factor values of benzylamine on different columns under

CEC and µ-HPLC.

Table 5.3 Retention factors kHPLC and kCEC of benzylamine in CEC and micro-HPLC of

some columns b, c, d and e from Table 5.1.

type b c d e Column

phase ratio 1.27 1.56 2.03 2.57

kHPLC 0.14 0.11 0.08 0.05

kCEC 0.30 0.20 elutes before t0

For columns d and e with higher phase ratios, the retention factor values are negative under

CEC conditions; the higher the phase ratio, the smaller the retention factor. In contrast, on

columns b and c with lower phase ratios, benzylamine has a positive retention factor and

kCEC is larger for the column with a smaller phase ratio. As discussed above, under the

conditions used in this study benzylamine exists predominantly as a cation, so its retention

is expected to be determined by both the electrophoretic migration and the interaction with

the chromatographic stationary phase. On columns with a higher phase ratio, the retention

of benzylamine is controlled by the electrophoretic effect but not by chromatographic

influences; cations elute faster. In contrast, on columns with lower phase ratios, the

retention of benzylamine is mainly attributed to reversed-phase chromatographic effects.

Obviously, the strong interaction of benzylamine with polymer is responsible for the longer

retention of benzylamine on columns b and c under CEC conditions.

5.3.4 Porosity

Total porosity (εT) is another important parameter for column evaluation. Various methods

are available to measure εT, such as the flow method [55], applying chromatographic

pressure-driven conditions, the conductivity method using electro-driven conditions [56],

and the gravimetric method [57]. In this study, the flow method was used to measure εT

values. The calculation of εT was based on a following equation: εT = 4Fto/(d2π L), where F

is the volumetric flow rate, to the retention time of an unretained marker (thiourea), d the


column inner diameter, and L the column length. The average εT values at different flow

rates are given in Table 5.1. Considering the possible compressibility of some columns (see

next section), a flow rate lower than 0.5 µL/min was applied for εT-measurements. The

relative standard deviation (RSD) of the measurement was less than 1.5%. Ideally, the

porosity of a column is equal to the porogen volume fraction in the reaction solution.

However, it is possible that some monomers may remain unreacted, some small polymer

pieces may be soluble, and normally after polymerization the volume of highly cross-linked

polymer may be smaller than the volume of the starting monomers. As a result, the real εT

is larger than the porogen volume fraction.

5.3.5 Reproducibility and stability Reproducibility of various column parameters is a critical consideration in the field of

preparation and application of rod columns. Therefore, a number of important parameters

such as EOF, efficiency and retention factors were determined to te t column-to-column as

well as batch-to-batch reproducibility. The results of the reproducib

by the retention time of thiourea, the retention factor (k), and the

summarized in Table 5.4.

Table 5.4 Reproducibility of electrochromatographic properties of

RSD%

Variable n EOF k (0.8-2.1

Run-to-run 10 0.26 < 0.35 Single column Day-to-day 6 0.78 < 0.85

One batch 6 0.76 < 3.90 Different columns

Different

batches 4 1.28 < 1.84

The day-to-day reproducibility of one column and the reproducibili

were averaged from the results of 10 continuously repeating injec

Table 5.4, the reproducibility for both the EOF and the retention

addition, the RSD values for column efficiency are acceptable, e

different batches. However, to apply these columns in daily r

improvements to decrease variation in plate number should be made

s

ility of EOF evaluated

column efficiency are

column c.

) Column efficiency

< 1.52

< 3.54

< 7.87

< 4.07

ty of different columns

tions. Obviously, from

factor is satisfactory. In

ither for one batch or

outine practice further

.

- 126 - Chapter 5 The mechanical stability of the columns was also studied. After using the columns for

longer than one week, about 1 mm long empty parts at both ends of the column were

observed. This was more obvious for columns prepared by using high porogen content.

Possible factors, such as EOF, pressure, and shrinkage that may affect the length of the

polymer rod were systematically examined. It was found that the shrinkage of a polymer

itself and the pressure might cause the polymer bed to become shorter. Under high

pressure (up to 300 bar), columns with a porogen volume fraction of higher than 57% are

compressible (see Table 5.1). As can be concluded from the linear relationship between

pressure drop and flow rate, the polymer bed remains stable after column equilibration

under high pressure. The column efficiency, however, decreases drastically after a column is

compressed (Table 5.1). For columns d and e, the ratios of column efficiency after and

before compression are 0.78 and 0.45, respectively. Hence, it is worthwhile to notice that a

low flow rate should be applied for conditioning these CEC columns if a pump is used.

However, when a modified fused silica capillary pre-treated with acrylic double bond on the

inner wall is used, both shrinkage and compressibility of the columns can be overcome. As

seen in the SEM photo in Figure 5.13, the polymer monolith in this column is covalently

bounded to the inner wall of the capillary. There is no cleft between the polymer rod and

the inner wall, which significantly differs from the SEM photos in Figure 5.2. Therefore,

modified capillaries may be a better alternative than unmodified fused silica capillaries in

preparation of monolithic columns.


Figure 5.13 SEM photos of a polymer rod column (h) with an inner wall modified fused silica capillary.

5.4 Conclusion Capillary monolithic CEC columns based on poly(alkylmethacrylate) with different phase

ratios were prepared by using different ratios of monomers to porogen. Small uniformly

linked polymer microglobules were obtained for the columns at low porogen content. The

high efficiency up to 150,000 plates/m and the high optimal flow velocity rates (up to

~6 cm/min) were achieved for all the test compounds with a wide range of retention

factors. These columns are promising for fast analysis if a high voltage is used.

From the comparison of the column behaviour under both electro- and pressure-driven

modes, the higher efficiency and the flatter Van Deemter curve in CEC were observed as

expected. The polymeric monolith in CEC demonstrated higher polarity possibly because

the charged groups on polymer are polarized and become oriented under the electric field.

All the columns exhibit similar reversed phase chromatographic retention mechanism for

most of the tested neutral compounds. For charged analytes under electro-driven

conditions, a competition between chromatographic and electrophoretic retention was

clearly seen. Moreover, for relatively bulky molecules columns show simultaneously size

- 128 - Chapter 5 exclusion retention and chromatographic retention. In such a case, as long as the

chromatographic partitioning is the leading force in the chromatographic process, the peaks

are abnormally broad because of the slow mass transfer of bulky molecules, resulting from

the steric resistance of micropores (< 2 nm). This type of monolith is suitable for the

analysis of very small molecules or macromolecules. In order to expand application of these

columns, it seems critical to enlarge the micropores to mesopore size similar to the inner

pores of normal silica based packing particles. Applications for the analysis of

macromolecules and the preparation of new monoliths with mesopores are under

development.

Acknowledgements

The authors are grateful to Mr. N. Lousberg for SEM experiments and to Mr. A.P.B.

Sommen for his assistance with micropore size measurements. We also acknowledge Dr.

X.W. Lou and Dr. W. H. Ming for their beneficial discussions.

REFERENCES 1. J.H. Knox, I.H. Grant, Chromatographia, 24 (1987) 135.





6. C. Yan, D. Schaufelberger, F. Erni, J. Chromatogr., A 670 (1994) 15.

7. R. Stol, H. Poppe, W. Th. Kok, J. Chromatogr. A, 887 (2000) 199.


9. G. Choudhary, Cs. Horváth, J. Chromatogr. A, 781 (1997) 161

10. A.S. Rathore, E. Wen, Cs. Horváth, Anal. Chem., 71 (1999) 2633.


12. S.R. Witowski, R.T. Kennedy, J. Microcolumn Sep., 11 (1999) 723.

13. R. Stol, W. Th. Kok, H. Poppe, J. Chromatogr. A, 853 (1999) 45.


15. D. Li, V.T. Remcho, J. Microcol. Sep., 9 (1997) 389.

16. U. Tallarek, E. Rapp, T. Scheenen, E. Bayer, H. van As, Anal. Chem., 72 (2000) 2292.


17. R.A. Carney, M.M. Robson, K.D. Bartle, P. Myers, J. High Resolut. Chromatogr., 22

(1999) 29.

18. M.G. Cikalo, K.D. Bartle, M.M. Robson, P. Myers, M.R. Euerby, Analyst, 123 (1998)

87R.

19. J.S. Kowalczyk, Chem. Anal. (Warsaw), 41 (1996) 157.

20. R. Stevenson, K. Mistry, I.S. Krull, Int. Lab., 11 (1998) 11C.

21. L.A. Colón, Y. Guo, A. Fermier, Anal. Chem., 8 (1997) 461A.

22. T. Tsuda, Analusis, 26 (1998) M48.

23. L.A. Colón, K.J. Reynolds, R.A. Maldonado, A.M. Fermier, Electrophoresis, 18 (1997)

2162.

24. K.D. Altria, N.W. Smith, C.H. Turnbull, Chromatographia, 46 (1997) 664.

25. C. Fujimoto, TrAC, Trend Anal. Chem., 18 (1999) 291.

26. S. Krull, A. Sebag, R. Stevenson, J. Chromatogr. A, 887 (2000) 137.

27. U. Pyell, J. Chromatogr. A, 892 (2000) 257.

28. F. Steiner, B. Scherer, J. Chromatogr. A, 887 (2000) 55.

29. C.A. Rimmer, S.M. Piraino, J.G. Dorsey, J. Chromatogr. A, 887 (2000) 115.

30. M. Pursch, C. Sander, J. Chromatogr. A, 887 (2000) 313.


32. B. Boughtflower, T. Underwood, Chromatographia, 38 (1995) 329.

33. M.M. Dittmann, G.P. Rozing, G. Ross, T. Adam, K.K. Unger, J. Capillary

Electrophor., 4 (1997) 201.



36. O.L. Tang, B.M. Xin, M.L. Lee, J. Chromatogr. A, 837 (1999) 35.


Tanaka, Anal. Chem., 72 (2000) 1275.

38. N. Minakuchi, K. Nakanishi, K.N. Soga, N. Ishizuka, N. Tanaka, Anal. Chem., 68

(1996) 3498.

39. S.M. Fields, Anal. Chem., 68 (1996) 2709.

40. N. Ishizuka, H. Minakuchi, K. Nakanishi, N. Soga, K. Hosoya, N. Tanaka, J. High

Resolut. Chromatogr., 21 (1998) 477.

41. F. Svec, E.C. Peters, D. Sýkora, C. Yu, J.M.J. Fréchet, J. High Resolut. Chromatogr., 23

(2000) 3.

- 130 - Chapter 5 42. Gusev, X. Huang, Cs. Horváth, J. Chromatogr., 855 (1999) 273.

43. Palm, M.V. Novotný, Anal. Chem., 69 (1997) 4499.


45. Maruška, C. Ericson, A. Vegvari, S. Hjertén, J. Chromatogr. A, 837 (1999) 25.

46. Ericson, S. Hjertén, Anal. Chem., 71 (1999) 1621.

47. Ericson, J. Holm, T. Ericson, S. Hjertén, Anal. Chem., 72 (2000) 81.

48. Ericson, J.-L. Liao, K. Nakazato, S. Hjertén, J. Chromatogr. A, 767 (1997) 33.

49. C. Fujimoto, J. Kino, H. Sawada, J. Chromatogr. A, 716 (1995) 107.




53. X. Huang, Cs. Horváth, J. Chromatogr. A, 788 (1997) 155.

54. H.A. Claessens, Characterization of Stationary Phases for Reversed-Phase Liquid

Chromatography: Column Testing, Classification and Chemical Stability, Ph.D. Thesis,

Eindhoven University of Technology, The Netherlands, 1999, p. 151.

55. Cs. Horváth, H.J. Lin, J. Chromatogr., 126 (1976) 401.

56. J. Bear, Dynamics of Fluids in Porous Media, Dover Publications, New York, 1988,

p.113.

57. K.K. Unger, Porous Silica, Its properties and Use as Support in Columns Liquid

Chromatography, Elsevier, Amsterdam, 1979, p. 171.

- 131 -

CHAPTER 6

6 QUANTITATIVE STRUCTURE RETENTION

RELATIONSHIPS IN COMPARATIVE STUDIES OF

BEHAVIOUR OF STATIONARY PHASES UNDER HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY AND

CAPILLARY ELECTROCHROMATOGRAPHY

CONDITIONS

Summary

Quantitative structure-retention relationships (QSRR) have been employed to study

molecular mechanism of chromatographic separations under pressure- (HPLC) and

electro-driven (CEC) conditions. Logarithms of retention factors corresponding to zero

percent of organic modifier in aqueous eluent, log kw, were determined on eight reversed-

phase stationary phases under both HPLC and CEC conditions at similar eluent flow

velocities. QSRR equations describing log kw in terms of linear solvation energy

relationship (LSER) parameters of analytes, in terms of simple structural descriptors

acquired by calculation chemistry, and in terms of logarithms of n-octanol-water partition

coefficients, were derived. Parameters of corresponding QSRR equations for individual

stationary phases were compared for both HPLC and CEC modes and the resulting

similarities and differences in retention mechanisms were discussed. It has been

This chapter has been submitted for publication in Journal of Chromatography A.

- 132 - Chapter 6 concluded that at least in the case of regular neutral analytes the specific inputs to

separation mechanism due to the electric field in CEC are of secondary importance.

6.1 Introduction Over two decades of development of capillary electrochromatography (CEC) many articles

have been published on the underlaying theory [1-10]. As a highly efficient separation

technique CEC was proposed mainly for analysis of steroids, peptides and proteins [11-16].

Much effort was devoted to column technology, development of stationary phases and to

specific column design.

CEC, combining the theory and practice of capillary electrophoresis (CE) and capillary

high-performance liquid chromatography (HPLC), attracted special attention from the

point of view of the molecular mechanism of retention. The retention behavior of test

analytes has been studied by a number of authors, who usually compared retention factors

or plate numbers of neutral compounds under both HPLC and CEC conditions. Recently,

model-based methods have been employed, in these comparative studies. Among them the

Galushko model [17,18] or models based on structural descriptors of analytes from

molecular modeling or from linear solvation energy relationships (LSER) [19-25] are

probably best known. The latter models have successfully been used in HPLC to

differentiate reversed-phase (RP) stationary phases and to predict retention of analytes.

The simplest quantitative structure-retention relationship (QSRR) model used in

comparative studies of stationary phases relates log kw to logarithm of n-octanol- water

partition coefficient, log P:

Pkkkw loglog 21 += (6.1)

where log kw is retention factor extrapolated to a pure water (buffer) mobile phase, the

coefficients k1 and k2 are characteristics of the systems representing differences in the

individual properties between the mobile and the stationary phase.

The following model relates log kw to structural descriptors of analytes provided by

molecular modeling:

SASkkkkkw'4

2'3min

'2

'1log +++= µδ (6.2)

where minδ is the largest atomic excess of electrons, 2µ is square of total dipole moment and

SAS is van der Waals surface area of a molecule that is accessible to a molecule of water,

Quantitative Structure Retention Relationships … - 133 -

the coefficients '1k , '

2k , '3k and '

4k are characteristics of the systems representing differences

in the individual properties between the mobile and the stationary phase.

The LSER model of QSRR is characterized by the following general equation:

xHHH

w VkkkkRkkk ''62

''52

''42

''32

''2

''1log +++++= βαπ (6.3)

where R2 is excess molar refraction, H2π is dipolarity/polarizability, H

2α is hydrogen-bond

acidity, H2β is hydrogen-bond basicity and Vx is characteristic volume of McGowan, the

coefficients ''1k , ''

2k , ''3k , ''

4k , ''5k and ''

6k are characteristics of the systems representing

differences in the individual properties between the mobile and the stationary phase.

Wei et al. [26] applied LSER in a comparative study of 3 µm ODS particles. They used 70%

ACN:30% aqueous buffer (2 mM Tris/HCl) as a mobile phase and found out that QSRR

parameters referring to CEC conditions differed from those obtained at RP-HPLC

conditions. Unlike in RP-HPLC, hydrogen-bond basicity of a solute ( H2β ) was statistically

significant in CEC. The parameters Vx, H2π , H

2α and H2β were all of similar significance in

CEC. On the other hand, in HPLC the most significant were Vx and H2β parameters;

actually, H2π was statistically significant at 95% significance level but it was of lesser

importance for retention description.

Employing LSER Liu et al. [27] discussed the role of organic modifier in RP-HPLC and in

CEC. The same group [28] studied also behavior of Spherisorb ODS II stationary phase in

CEC, pressurized electrochromatography (PEC) and HPLC. For acetonitrile as an organic

modifier, the reported LSER equations obtained under CEC, PEC and HPLC conditions

were closely similar. In the view of limited and rather fragmental actual knowledge it

appeared worthwhile to systematically study the molecular mechanism of separations in

analogous chromatographic systems operated at CEC and HPLC conditions employing the

QSRR approach. For that purpose eight RP stationary phases were subjected to a study

under both HPLC and CEC conditions using acetonitrile as organic modifier.


by the manufacturer. The column packed bed and the total length was 25 cm and 33.5

cm, respectively.

- 134 - Chapter 6 Prior to use in the CEC mode, the columns were conditioned. This was accomplished by

applying 10 bar pressure on both sides of the column and increasing the voltage from 0 to

25 kV in 5 kV steps for 10 min. Next, the pressure was increased to 12 bar and a 30 kV

voltage was applied for 10 min. For the micro-HPLC experiments, the columns were

conditioned until the column pressure was stabilized (about 1 h).

Table 6.1 List of investigated columns; each column diameter: 100 µm; average particle

size, 3 µm.

Column Pore Size

[Å] Pore

Volume [cm3/g]

Surface Area

[m2/g]

Carbon Load [%]

CEC Hypersil C18

130 0.65 170 8.5

Hypersil C8 120 0.65 170 6.5

Hypersil Phenyl 120 0.65 170 5

Spherisorb ODS 80 0.50 200 6.2

Spherisorb C8 80 0.50 200 5.8

Unimicro C18 Data not available

Unimicro C8 Data not available

Unimicro Phenyl

Data not available

The columns were tested under pressure- and electro-driven conditions using the same

batches of eluents. All the columns were supplied in duplicate (the same batch with

maximum 2% RSD in retention factor under HPLC conditions). The requirement was

met for analysis of up to 1% RSD in retention factor under CEC conditions and of up to

0.5% RSD under HPLC conditions; each for six consecutive injections. Hold-up time (t0)

was measured using thiourea added to the solutions of the analytes and varied between 3-7

minutes depending on column and percentage of organic modifier. t0 time in HPLC was adjusted


to the obtained t0 time in CEC for particular column and particular mobile phase composition.

HPLC conditions were adjusted to similar flow velocities as obtained in CEC. As a

consequence, the HPLC experiments were not optimized with respect to the plate height.

6.2.2 Instrumentation All the CEC chromatograms were obtained on a 3DCE instrument (Agilent Technologies

GmbH, Waldbronn, Germany) equipped with a pressure facility of up to 12 bar at the

outlet and/or inlet vial. This pressurization option of the instrument was used to prevent

bubble formation in the capillaries. Samples were injected electrokinetically (5 kV for 2-

15 s). For each run a voltage of 20 kV (600 Vcm-1 electric field strength) was applied with

10 bar pressure at both ends of a capillary. The detection wavelength was 210 nm. High

voltage was applied as a 6 s time ramp to avoid column stress. The column cassette

temperature was maintained at 20ºC.


syringe pump and microUVIS20 ultraviolet/visible absorbance detector operated at 210 nm

both from (Carlo Erba Instruments, Milan, Italy), and an injector with a 200 nL loop

(VICI-AG Valco Europe, Schenkon, Switzerland). The flow-rate was adjusted to that in

CEC experiments (approx. 0.2-0.3 µL/min) using a VICI-AG 1/100-flow splitter. The

experiments were performed at air-conditioned laboratory conditions (temperature about

21ºC) without additional thermostating.

6.2.3 Chemicals Acetonitrile (ACN) of HPLC supra gradient-grade purity (Biosolve, Valkenswaard,

Netherlands) was used as the organic modifier in various concentrations. The eluents were

prepared by mixing phosphate buffer (pH 7.0, final concentration 1 mmol/L) with an

appropriate amount of the organic modifier and degassed ultrasonically for 15 min prior to

use. The same batch of eluent was used to test a given column at both separation modes.

The set of test analytes is listed in Table 6.2 together with their structural descriptors. The

series of analytes was taken as previously designed [29] with the well-defined hydrogen-

bond capacity descriptors derived from the complexation scale of Abraham [25,26].

Samples were prepared by dissolving the analytes in the mobile phase or in the pure organic

modifier and then diluting with phosphate buffer.

- 136 - Chapter 6 Table 6.2 Structural descriptors of test analytes used in QSRR equations. No. Solute log P R2 H

2π H2α H

2β Vx minδ 2µ SAS

1 n-Hexylbenzene 5.52 0.591 0.50 0.00 0.15 1.562 -0.2104 0.03880 415.40

2 1,3,5-Triisopropylbenzene 6.36 0.627 0.40 0.00 0.22 1.985 -0.2057 0.00624 478.27

3 1,4-Dinitrobenzene 1.47 1.130 1.63 0.00 0.41 1.065 -0.3418 0.00012 312.07

4 3-Trifluoromethylphenol 2.95 0.425 0.87 0.72 0.09 0.969 0.2454 4.39321 302.54

5 3,5-Dichlorophenol 3.62 1.020 1.10 0.83 0.00 1.020 0.2434 1.98246 306.77

6 4-Cyanophenol 1.60 0.940 1.63 0.79 0.29 0.930 -0.2440 10.9693 290.61

7 4-Iodophenol 2.91 1.380 1.22 0.68 0.20 1.033 -0.3021 2.51856 301.47

8 Anisole 2.11 0.708 0.75 0.00 0.29 0.916 -0.2116 1.56000 288.13

9 Benzamide 0.64 0.990 1.50 0.49 0.67 0.973 -0.4334 12.8450 293.30

10 Benzene 2.13 0.610 0.52 0.00 0.14 0.716 -0.1301 0.00000 244.95

11 Chlorobenzene 2.89 0.718 0.65 0.00 0.07 0.839 -0.1295 1.70824 269.49

12 Cyclohexanone 0.81 0.403 0.86 0.00 0.56 0.861 -0.2944 8.83278 269.31

13 Dibenzothiophene 4.38 1.959 1.31 0.00 0.18 1.379 -0.2709 0.27457 364.54

14 Phenol 1.47 0.805 0.89 0.60 0.30 0.775 -0.2526 1.52028 256.72

15 Hexachlorobutadiene 4.78 1.019 0.85 0.00 0.00 1.321 -0.0750 0.06708 352.14

16 Indazole 1.77 1.180 1.25 0.54 0.34 0.905 -0.2034 2.39011 285.46

17 Caffeine -0.07 1.500 1.60 0.00 1.35 1.363 -0.3620 13.3298 367.02

18 4-Nitrobenzoic acid 1.89 0.990 1.07 0.62 0.54 1.106 -0.3495 11.7786 321.77

19 N-Methyl-2- pyrrolidinone -0.38 0.491 1.50 0.00 0.95 0.820 -0.3532 12.9168 270.53

20 Naphthalene 3.30 1.340 0.92 0.00 0.20 1.085 -0.1277 0.00000 313.25

21 4-Chlorophenol 2.39 0.915 1.08 0.67 0.20 0.898 -0.2482 2.18448 280.38

22 Toluene 2.73 0.601 0.52 0.00 0.14 0.716 -0.1792 0.06916 274.50

23 Benzonitrile 1.56 0.742 1.11 0.00 0.33 0.871 -0.1349 11.1222 277.91

24 Benzoic acid 1.87 0.730 0.90 0.59 0.40 0.932 -0.3651 5.85156 288.00

25 1,3-Diisopropylbenzene 4.90 0.605 0.46 0.00 0.20 1.562 -0.2055 0.08820 399.79

log P = logarithm of n-octanol-water partition coefficient; R2 = excess molar refraction; H2π = dipolarity/polarizability; H

2α = hydrogen-bond acidity; H2β = hydrogen-bond

basicity; Vx = characteristic volume of McGowan; minδ = highest electron excess charge on

an atom in the analyte molecule (in electrons); 2µ = square of total dipole moment (in

Debyes); SAS = solvent (water)-accessible molecular surface area (in Å2).

6.2.4 Test procedure Analytes were chromatographed with mobile phases being mixtures of organic modifier

with an aqueous buffer of composition ranging from 90/10 (V/V) to 40/60 (V/V). Based

on the linear relationship between the logarithm of retention factor (log k) and the

percentage of organic modifier in the mobile phase, the values of log kw corresponding to


100% aqueous eluent were obtained by extrapolation. The data are summarized in Table

6.3.

6.3 Results and discussion Table 6.3 summarizes values of log kw (retention factor extrapolated to 100% aqueous

mobile phase) for all the columns under both HPLC and CEC conditions. It is evident that

both Spherisorb stationary phases (ODS and C8) show retention patterns different than the

remaining phases. Values of log kw of polar compounds, such as 1,4-dinitrobenzene, 3,5-

dichlorophenol, 4-cyanophenol or cyclohexanone, are much higher than the corresponding

data determined on Hypersil or Unimicro stationary phases. Principal component analysis

(PCA) clearly distinguishes the Spherisorb C18 stationary phase under both HPLC and

CEC conditions as an outlier as regards the retention mechanism (Fig. 6.1a). PCA done for

the columns remaining after excluding Spherisorb C18 evidences that behaviour of the

Spherisorb C8 stationary phase also differs from the other phases though the difference is

not as pronounced as for the Spherisorb C18 stationary phase (Fig. 6.1b).

Comparing the data on the stationary phases given in Table 6.1 one may notice that, for

example, Hypersil columns have a higher carbon load and a lower surface area than

Spherisorb. The differences in log kw on those phases are remarkable. Most probably, they

are to some extent also due to the differences in column dipolarity/polarizibility. Spherisorb

stationary phases are based on a type of silica substrate that is apparently different than in

the case of other phases studied as the selectivity differences for the phases based on similar

substrate are usually minor [30].

- 138 - Chapter 6

25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 N

o.

1,3-Diisopropylbenzene

Benzoic acid Benzonitrile Toluene 4-Chlorophenol N

aphthalene n-M

ethyl-2- pyrrolidinone 4-N

itrobenzoic acid Caffeine Indazole H

exachlorobutadiene Phenol D

ibenzothiophene Cyclohexanone Chlorobenzene Benzene Benzam

ide Anisole 4-Iodophenol 4-Cyanophenol 3,5-D

ichlorophenol 3-T

rifluoromethylphenol

1,4-Dinitrobenzene

1,3,5-Triisopropylbenzene n-H

exylbenzene

Analyte

6.2297 N

/A

2.4781 3.6731 2.8017 4.4222 -1.5731 N

/A

-1.117 1.4766 5.8905 1.8713 5.288 0.578 3.7446 2.9563 -0.0636 2.9818 3.1399 2.4421 3.7453 3.3498 3.3307 7.5969 6.8896 H

PLC

6.1735 N

/A

2.5528 3.7153 2.801 4.4403 -1.5032 N

/A

-1.0608 1.5621 5.807 1.9034 5.2275 0.9771 3.8149 3.0568 0.0284 3.031 3.1695 2.2699 3.82

3.4113 3.328 7.4694 6.7875 CE

C H

ypersil C18

6.4277 N

/A

2.7396 4.1474 2.9211 4.883

-2.1933 N

/A

-1.5547 1.586 5.9595 1.8566 6.0893 0.9234 4.2018 3.3792 0.1706 3.3711 3.4818 1.9323 4.1998 3.8029 3.5104 7.8589 6.8952 H

PLC

6.308 N

/A

2.7173 4.1025 2.8923 4.7714 -2.1923 N

/A

-1.4163 1.5356 5.8911 1.8303 5.921 0.9817 4.1603 3.3502 0.1214 3.0874 3.3882 1.7582 4.116 3.7493 3.4699 7.634 6.7579 CE

C H

ypersil C8 MO

S

5.7691 N

/A

2.2436 3.2389 2.5838 3.9157 -1.5955 N

/A

-0.9153 1.4586 5.3608 1.8956 4.8632 0.6586 3.3607 2.6257 0.3093 2.839 2.9769 2.1608 3.617 3.5055 3.1718 7.0597 6.1838 H

PLC

5.7907 N

/A

2.3027 3.298 2.5305 3.9475 -1.3497 N

/A

-1.3497 1.3001 5.4501 1.6201 4.9205 0.6516 3.3741 2.7196 0.056 2.8745 2.9368 1.8683 3.0982 3.2564 3.0677 6.4122 6.2189 CE

C H

ypersil Phenyl

5.798 N

/A

3.2451 4.0619 4.1911 4.4095 -2.0614 N

/A

-1.8807 3.3165 5.6157 7.336 5.1093 3.2235 3.7416 3.4653 1.4421 3.8934 4.8794 8.0989 5.4174 6.3703 6.3164 7.0826 6.4547 H

PLC

6.0112 N

/A

3.6883 4.0697 4.274 4.4376 -1.7704 N

/A

-1.3489 1.9024 5.6734 4.5947 5.1538 1.7929 4.0986 3.7403 0.3557 3.7837 4.1149 6.8314 4.3205 5.2672 5.0714 7.0962 6.4859 CE

C Spherisorb O

DS

log kw

Table 6.3 Logarithm

s of retention factors extrapolated to 100% aqueous eluent in individual chrom

atographic systems.


CEC

6.24

9 7.

0719

3.

2227

3.

5613

3.

6621

2.

0727

3.

1233

2.

9127

0.

4501

2.

8514

3.

4757

1.

1013

4.

8745

1.

8713

5.

43

1.65

64

-0.7

478

N/A

-1

.547

1 4.

0112

2.

7058

3.

4061

2.

5092

N

/A

5.86

01

Uni

mic

ro P

heny

l H

PLC

6.39

47

7.22

21

3.61

97

3.85

37

3.94

54

2.85

3.

4974

3.

1488

0.

6348

3.

1787

3.

7288

1.

3878

5.

0753

2.

2888

5.

5756

2.

0131

-0

.949

6 N

/A

-1.0

976

4.22

94

3.11

5 3.

6116

2.

8172

N

/A

5.99

79

CEC

6.85

42

7.75

46

3.83

64

4.06

52

3.94

84

2.09

86

3.68

96

3.60

88

0.57

27

3.64

99

4.39

96

1.53

65

6.01

55

2.22

12

6.36

57

1.90

67

-0.9

13

N/A

-1

.616

1 4.

9618

3.

2502

4.

3017

3.

0946

N

/A

6.46

77

Uni

mic

ro C

8 H

PLC

6.86

34

7.81

04

3.81

15

4.09

09

4.11

8 2.

4349

3.

7421

3.

5924

0.

434

3.60

33

4.37

73

1.44

93

5.59

09

2.18

19

6.39

09

1.89

58

-1.0

928

N/A

-1

.785

2 4.

9962

3.

2379

4.

2957

3.

0572

N

/A

6.44

94

CEC

6.12

51

6.83

85

2.59

97

2.86

41

3.28

12

1.09

09

2.59

82

2.50

52

-0.6

775

2.54

78

3.35

22

0.28

06

4.82

39

1.00

19

5.06

92

0.76

94

-1.9

089

N/A

-2

.513

2 4.

035

2.09

93

3.29

24

1.87

47

N/A

5.

461

Uni

mic

ro C

18

HPL

C 6.

2938

7.

0644

2.

6063

2.

9729

3.

2747

1.

362

2.66

73

2.54

16

-0.5

647

2.56

75

3.39

84

0.38

47

4.93

06

1.06

6 5.

2663

0.

8925

-1

.998

8 N

/A

-2.6

003

4.10

2 2.

121

3.32

56

1.91

4 N

/A

6.03

03

CEC

6.40

09

7.32

78

4.14

11

4.31

5 3.

7131

2.

8712

3.

5546

3.

1648

0.

264

3.15

17

3.83

96

1.14

57

5.37

76

2.29

26

5.66

2 1.

6479

-1

.174

5 N

/A

-1.6

597

4.39

66

3.35

69

3.79

65

2.76

94

N/A

6.

2742

log

k w

Sphe

risor

b C8

H

PLC

6.85

24

7.98

65

4.66

3 4.

1665

4.

4597

2.

1452

3.

8838

3.

3855

0.

567

3.25

08

3.93

58

1.53

05

5.68

67

2.60

05

6.01

37

1.97

49

-1.2

51

N/A

-1

.686

7 4.

642

3.15

46

4.00

61

2.71

84

N/A

6.

2177

Ana

lyte

n-H

exyl

benz

ene

1,3,5

-Trii

sopr

opyl

benz

ene

1,4-D

initr

oben

zene

3-

Trif

luor

omet

hylp

heno

l 3,

5-D

ichl

orop

heno

l 4-

Cyan

ophe

nol

4-Io

doph

enol

An

isole

Be

nzam

ide

Benz

ene

Chlo

robe

nzen

e Cy

cloh

exan

one

Dib

enzo

thio

phen

e Ph

enol

H

exac

hlor

obut

adie

ne

Inda

zole

Ca

ffein

e 4-

Nitr

oben

zoic

aci

d n-

Met

hyl-2

- pyr

rolid

inon

e N

apht

hale

ne

4-Ch

loro

phen

ol

Tolu

ene

Benz

onitr

ile

Benz

oic

acid

1,3

-Diis

opro

pylb

enze

ne

No.

1 2 3 4 5 6 7 8 9 10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Tabl

e 6.

3 (c

ontin

ued)

- 140 - Chapter 6

Plot of Component Weights

Component 1

Com

pone

nt 2

SpheriODSCEC

SpheriODSHPLC

0.19 0.21 0.23 0.25 0.27-0.14

0.06

0.26

0.46

0.66

0.86

HypC8CEC UniC8CEC

Plot of Component Weights

Component 1

Com

pone

nt 2

HypC18CEC

HypC18HPLC

HypC8HPLC HypPhenCEC

HypPhenHPLC

SpheriC8CEC

SpheriC8HPLC

UniC18CEC

UniC18HPLC

UniC8HPLC

UniPhenCEC

UniPhenHPLC

266 266.3 266.6 266.9 267.2 267.5 267.8(x 0.001)

-0.35

-0.15

0.05

0.25

0.45

0.65

UniC8CEC

HypC8CEC

Figure 6.1 Plot of first two component weights resulting from principal component analysis of log kw

data determined in a) all the separation systems studied b) with Spherisorb ODS stationary

phases excluded. SpheriODS=Spherisorb ODS, SpheriC8=Spherisorb C8,

HypC18=Hypersil C18, HypC8=Hypersil C8, HypPhen=Hypersil Phenyl,

UniC18=Unimicro C18, UniC8=Unimicro C8, UniPhen=Unimicro Phenyl; symbols CEC

and HPLC after the name of the column indicate performed under CEC conditions or

HPLC conditions, respectively.

Results of QSRR analysis of retention data for test series of solutes are collected in Tables

6.4-6.6.


Table 6.4 Regression coefficients (± standard deviation), numbers of data points used to

derive regression (n), correlation coefficient (R), standard errors of estimate (s) and F-test

values (F) of regression equations Pkkkw loglog 21 += . The values are statistically

significant on 99% confidence level.

Column Mode k1 k2 n R s F

HPLC -0.2886 (±0.2436) 1.3294 (±0.0888) 20 0.9601 0.6047 224 CEC Hypersil C18

CEC -0.1722 (±0.2374) 1.2961 (±0.0866) 20 0.9601 0.5892 224

HPLC -0.3769 (±0.2881) 1.4502 (±0.1050) 20 0.9536 0.7150 191 Hypersil C8

CEC -0.3710 (±0.2816) 1.4206 (±0.1027) 20 0.9538 0.6988 192

HPLC -0.1667 (±0.2144) 1.1970 (±0.0782) 20 0.9618 0.5322 235 Hypersil Phenyl

CEC -0.2934 (±0.2357) 1.2176 (±0.0860) 20 0.9558 0.5849 201

HPLC 1.7140 (±0.8298) 1.0461 (±0.3025) 20 0.6215 2.0596 12 Spherisorb ODS

CEC 1.0111 (±0.6089) 1.1422 (±0.2220) 20 0.7630 1.1511 27

HPLC 0.1181 (±0.3054) 1.3222 (±0.1114) 20 0.9387 0.7580 141 Spherisorb C8

CEC 0.1546 (±0.3181) 1.2404 (±0.1160) 20 0.9260 0.7895 114

HPLC -1.0248 (±0.2529) 1.4124 (±0.0922) 20 0.9618 0.6276 235 Unimicro C18

CEC -1.0359 (±0.2527) 1.3894 (±0.0921) 20 0.9607 0.6272 227

HPLC 0.0782 (±0.2912) 1.3556 (±0.1062) 20 0.9464 0.7228 163 Unimicro C8

CEC 0.1363 (±0.2853) 1.3421 (±0.1040) 20 0.9474 0.7080 167

HPLC 0.3204 (±0.2429) 1.1605 (±0.0886) 20 0.9489 0.6029 172 Unimicro Phenyl

CEC -0.0070 (±0.2131) 1.1772 (±0.0777) 20 0.9610 0.5289 230

Table 6.4 summarizes coefficients k1 and k2 of the regression equations relating log kw data

to log P (Eqn. 6.1). There are no statistically significant differences (t-test, 95% confidence

level) in neither k1 nor k2 between HPLC and CEC modes for none of the stationary

phases tested. Therefore, for this particular case in terms of analyte partition between

mobile and stationary zone, the nature of eluent driving force is not important (see

explanation futher). Also, the differences of k1 and k2 among individual phases studied are

insignificant for either HPLC or CEC conditions. Perhaps that does not concern the pair of

- 142 - Chapter 6 phases: Hypersil C8 and Spherisorb ODS. An evidently lower regression coefficient k2 at

log P term in Eqn. 6.1, observed for Spherisorb ODS, indicates a lower lipophilicity of that

phase which may arise from the specific properties of the silica substrate.

The observed similarity of partition properties of the stationary phases studied is not

surprising because all of them are modern bonded-silica reversed-phase materials designed

to maximally reduce specific, hard to control contributions to retention. It can be

concluded that the log P parameter of analytes is not sensitive enough to clearly distinguish

possible differences in retention properties of modern materials.

Lipophilicity (or hydrophobicity) parameters, like log P, are complex net measures of

various intermolecular interactions between analyte, on one hand, and components of a

given partition system, on the other hand. Two main types of intermolecular interactions

are distinguished as governing both slow-equilibrium and chromatographic separations:

non-specific ones, i.e., molecular-bulkiness-related, dispersive, London’s interactions and

structurally specific, polar interactions including dipole-dipole, dipole-induced dipole,

hydrogen bonding and electron pair donor-electron pair acceptor interactions [19].


Table 6.5 Regression coefficients (± standard deviation), numbers of data points used

to derive regression (n), correlation coefficient (R), standard errors of estimate (s) and F-

test values (F) of regression equations SASkkkkkw'4

2'3min

'2

'1log +++= µδ . The values are

significant on 99% confidence level; the values (-) are statistically not significant on 99%

confidence level.

Column Mode '1k

'2k

'3k

'4k n R s F

HPLC -0.5592

(±1.1657)

8.3868

(±2.5836)

-0.2273

(±0.0471)

0.0209

(±0.0034)

22 0.9388 0.8691 47 CEC Hypersil C18

CEC -0.2529

(±1.1329)

8.1938

(±2.5108)

-0.2252

(±0.0458)

0.0198

(±0.0033)

22 0.9389 0.8497 47

HPLC 0.1180

(±1.3471)

8.3014

(±2.9857)

-0.2717

(±0.0544)

0.0198

(±0.0039)

22 0.9293 1.0045 40 Hypersil C8

CEC 0.1451

(±1.3015)

8.2231

(±2.8846)

-0.2653

(±0.0526)

0.0194

(±0.0038)

22 0.9307 0.9705 41

HPLC -0.6564

(±1.1067)

6.6522

(±2.4528)

-0.2135

(±0.0447)

0.0190

(±0.0032)

22 0.9325 0.8252 42 Hypersil Phenyl

CEC -0.1301

(±1.1256)

7.3085

(±2.4947)

-0.2163

(±0.0455)

0.0176

(±0.0033)

22 0.9298 0.8393 40

HPLC 5.5476

(±0.5476)

- -0.3160

(±0.0886)

- 22 0.6144 2.0510 13 Spherisorb ODS

CEC 5.1352

(±0.4690)

- -0.3205

(±0.0758)

- 22 0.6779 1.7566 18

HPLC -0.5645

(±1.3180)

- -0.3378

(±0.0460)

0.0172

(±0.0040)

22 0.9084 1.0316 47 Spherisorb C8

CEC 0.4576

(±1.3565)

6.6421

(±3.0065)

-0.2417

(±0.0548)

0.0172

(±0.0039)

22 0.9072 1.0115 29

HPLC -0.8682

(±1,2422)

9.0435

(±2.7532)

-0.2461

(±0.0502)

0.0209

(±0.0036)

22 0.9374 0.9262 46 Unimicro C18

CEC -0.6761

(±1.1792)

8.7647

(±2.6134)

-0.2456

(±0.0476)

0.0198

(±0.0034)

22 0.9403 0.8792 48

HPLC 0.7885

(±1.2607)

8.7656

(±2.7941)

-0.2459

(±0.0509)

0.0184

(±0.0035)

22 0.9284 0.9400 41 Unimicro C8

CEC 0.6595

(±1.2235)

8.1854

(±2.7118)

-0.2485

(±0.0494)

0.0184

(±0.0035)

22 0.9325 0.9123 42

HPLC 0.4209

(±1.1894)

6.8211

(±2.6361)

-0.2102

(±0.0481)

0.0169

(±0.0034)

22 0.9169 0.8869 33 Unimicro Phenyl

CEC -0.3641

(±1.0831)

6.8911

(±2.4005)

-0.2076

(±0.0438)

0.0186

(±0.0031)

22 0.9332 0.8076 43

- 144 - Chapter 6 Unlike the rather crude analyte property descriptor, log P, in Eqn. 6.1, in QSRR equations

of the form of Eqns. 6.2 and 6.3, the terms are present which should account for

differences in specific intermolecular interactions if such were to manifest themselves in

CEC with respect to HPLC or among the individual stationary phases operated in a given

separation mode. Table 6.5 summarizes parameters characterizing QSRR equations

describing log kw in terms of structural descriptors of analytes that are easily acquired by

standard computational chemistry programs (Eqn. 6.2). Again, for none of the eight

stationary phases under study any statistically significant difference in regression coefficients

k k1'

4'− (t-test, 95% confidence level) was found between the HPLC and the CEC modes.

On the other hand, when comparing respective QSRR equations for individual stationary

phase materials one can distinguish Spherisorb C18 (both in HPLC and CEC mode). In

QSRR equations in Table 6.5 for Spherisorb C18 the terms related to the highest electron

excess on an atom in analyte molecule, δmin, and to a water accessible van der Waals surface,

SAS, are insignificant. Instead, significant are the square of the total dipole moment, µ2,

and the free term k1’ which is very large.


Table 6.6 Regression coefficients (± standard deviation), numbers of data points used to

derive regression (n), correlation coefficient (R), standard errors of estimate (s) and F-test

values (F) of regression equations xHHH

w VkkkkRkkk ''62

''52

''42

''32

''2

''1log +++++= βαπ . The

values are significant on 99% confidence level, the values (-) are statistically not significant

on 99% confidence level.

Column Mode ''1k

''4k ''

5k ''6k n R s F

HPLC 0.9181 (±0.3237)

-1.2948 (±0.2504)

-5.8859 (±0.2536)

4.1017 (±0.2243)

22 0.9792 0.3640 298 CEC Hypersil C18

CEC 1.15623 (±0.2840)

-1.3581 (±0.2197)

-5.7990 (±0.2225)

3.89315 (±0.2243)

22 0.9831 0.3194 368

HPLC 1.5178 (±0.3423)

-1.5598 (±0.2648)

-6.6188 (±0.2681)

3.9963 (±0.2703)

22 0.9799 0.3849 309 Hypersil C8

CEC 1.5081 (±0.3458)

-1.5662 (±0.2675)

-6.4638 (±0.2709)

3.8934 (±0.2731)

22 0.9785 0.3889 289

HPLC 0.6970 (±0.2769)

-0.9110 (±0.2142)

-5.2908 (±0.2169)

3.8519 (±0.2187)

22 0.9814 0.3114 335 Hypersil Phenyl

CEC 1.1717 (±0.2877)

-1.3617 (±0.2226)

-5.4672 (±0.2254)

3.4641 (±0.2272)

22 0.9799 0.3235 309

HPLC 6.3256 (±0.4942)

- -6.3134 (±1.1162)

- 22 0.6037 1.6364 32 Spherisorb ODS

CEC 2.9499 (±0.8434)

-- -6.0670 (±0.7376)

2.6854 (±0.7248)

22 0.8050 0.6920 41

HPLC 1.5344 (±0.4041)

-1.0892 (±0.3126)

-6.0795 (±0.3166)

3.8866 (±0.3191)

22 0.9678 0.4544 190 Spherisorb C8

CEC 1.5686 (±0.4287)

-0.9786 (±0.3316)

-5.8186 (±0.3358)

3.5986 (±0.3385)

22 0.9598 0.4821 151

HPLC 0.6178 (±0.2806)

-1.5986 (±0.2171)

-6.3602 (±0.2198)

4.0938 (±0.2216)

22 0.9859 0.3156 444 Unimicro C18

CEC 0.7257 (±0.2881)

-1.6406 (±0.2229)

-6.2226 (±0.2257)

3.8737 (±0.2275)

22 0.9843 0.3240 396

HPLC 1.9426 (±0.3179)

-1.4767 (±0.2459)

-6.2899 (±0.2490)

3.6846 (±0.2510)

22 0.9804 0.3575 317 Unimicro C8

CEC 1.9832 (±0.3126)

-1.6537 (±0.2418)

-6.1486 (±0.2449)

3.6645 (±0.2468)

22 0.9807 0.3515 321

HPLC 1.4823 (±0.2669)

-0.8489 (±0.2065)

-5.3637 (±0.2091)

3.4852 (±0.2107)

22

0.9818 0.3001 341 Unimicro Phenyl

CEC 0.9686 (±0.2620)

-1.0362 (±0.2026)

-5.2434 (±0.2052)

3.7126 (±0.2068)

22 0.9828 0.2946 363

Table 6.6 summarizes statistical parameters of QSRR equations based an analyte descriptors

from linear solvation energy relationship theory (Eqn. 6.3). With the series of test analytes

- 146 - Chapter 6 employed, the LSER-based analyte descriptors R2 and π2 appeared insignificant in case of

each stationary phase and the separation mode studied.

QSRR equations based on LSER descriptors also do not prove actual difference in

molecular mechanism of separation between the two modes compared, i.e., between HPLC

and CEC. There are no statistically significant differences (t-test, 95% confidence level)

between the k4'' , k5

'' and k6" coefficients in HPLC and CEC. The lack of the significance of

term corresponding to McGowan volume, Vx, in QSRR for Spherisorb C18 operated at

HPLC conditions, whereas it is significant at CEC conditions, may not be conclusive.

Furthermore, there has been no explanation found on low correlation coefficient for this particular

stationary phase for all three QSRR methods. On the other hand, there seems to be a systematic

trend in coefficients collected in Table 6.6 when comparing analogous QSRR equations for HPLC

and CEC. Namely, k6" and k5

" (negative sign) tend to be higher whereas k4" (negative sign)

tends to be lower in case of CEC. Physical meaning of that observation, if any, may better

be checked if data given in Table 6.6 will be related to those Table 6.5. That can be done

because the QSRR equations of general form Eqns. 6.2 and 6.3 are mutually related. If one

compares the ordering of separation systems on the plot of k6" in Eqn. 6.3 vs. k4

" in Eqn.

6.2 (Figure 6.2), i.e., according to the regression coefficients at the volume of the analyte

(Vx) and at its van der Waals surface area that is accessible to water (SAS), one will notice a

clear trend. Namely, the higher coefficients stand at the C18 compared to the C8 stationary

phases. It is rational because the C18 phases have a larger surface area of the hydrocarbon

ligand that is accessible to the analyte. The same phases under CEC conditions have lower

values of both k6'' and k4

' coefficients than under HPLC conditions. That finding seems to

be reasonable in view of a previous study report by Jiskra et al. [31]. Those authors

suggested that generating electroosmotic flow on the stationary phase under CEC

conditions causes reordering of hydrocarbon chains of the ligand. That reordering may lead

to a decrease of the overall contact of the solute with the hydrocarbonaceous stationary

phase. Euerby et al. [34] used the CEC Hypersil C18, Hypersil C8 and Hypersil Phenyl for

separation of barbiturates. The authors observed increase in retention on the Hypersil C8

stationary phase compared to the CEC Hypersil C18 stationary phase under CEC

conditions while only minimum increase has been observed under HPLC conditions [35].

However, the separation order remained the same. This confirms further findings of this

group [e.g. 36] and others [e.g. 37-38]. Wen et al. [39] found linear relationship between


kHPLC and kCEC for neutral small molecules on the Spherisorb ODS (300 Å) and Zorbax

ODS (80 and 300 Å). However, the slope of this relationship differed from one (namely

1.12). In this paper, the authors further focused on the Van Deemter parameters A (eddy

diffusion term) and C (mass transfer resistance). It has been found that the value of both

parameters was by a factor of two to four lower in HPLC compared to CEC due to the

peculiarities of the EOF flow profile in the interstitial space and the generation of

intraparticle EOF inside the porous particles of the column packing.

3.4

3.5

3.6

3.7

3.8

3.9

4

4.1

4.2

0.015 0.016 0.017 0.018 0.019 0.02 0.021 0.022 0.023

k4'

k 6''

CEC Hypersil C18-HPLC

CEC Hypersil C18-CEC

Hypersil C8-HPLC

Hypersil C8-CEC

Hypersil Phenyl-HPLC

Hypersil Phenyl-CEC

Spherisorb C8-HPLC

Shperisorb C8-CEC

Unimicro C18-HPLC

Unimicro C18-CEC

Unimicro C8-HPLC

Unimicro C8-CEC

Unimicro Phenyl-HPLC

Unimicro Phenyl-CEC

Stationary phases with most developed surface

Stationary phases with least developed surface

Spherisorb C8

Spherisorb C8 CEC Hypersil C18 (CEC) + Unimicro C18 (CEC)

CEC Hypersil C18 (HPLC) + Unimicro C18 (HPLC)

Unimicro C8 (HPLC + CEC)

Hypersil C8 (HPLC + CEC)

Figure 6.2 Ordering of stationary phases according to their non-specific retentivity due to dispersion

interaction characterized by the coefficient ''6k for the Vx variable in Eq. 6.4 and the

coefficient '4k for the SAS variable in Eq. 6.3.

Coefficients k5" in Eqn. 6.3 and k3

' in Eqn.6.2 (Tables 6.5 and 6.6), may both be related to

the amount and activity of free silanol groups which are accessible to analytes. That would

explain a correlation between the coefficients (Fig. 6.3). Similarly as in Fig. 6.2, the

stationary phases having a less negative value are the phases with higher amount/activity of

free silanols (therefore, these phases compete more effectively for analytes with strongly

polar eluents). Typically, the phenyl stationary phases show higher silanol activity whereas

the C8 and C18 phases lower values. The exception is CEC Hypersil C18 stationary phase

- 148 - Chapter 6 under both CEC and HPLC conditions. This is not surprising, as this particular stationary

phase has been designed for use in CEC as possessing higher amount of free silanols. The

outliers are Spherisorb C18, Spherisorb C8 and Unimicro Phenyl stationary phases. In

general, there is a trend that the stationary phases exhibit a higher silanol activity under

CEC conditions (open symbols) than under HPLC conditions. That confirms previous

reports [31-33].

-6,8

-6,6

-6,4

-6,2

-6

-5,8

-5,6

-5,4

-5,2

-5-0,35 -0,33 -0,31 -0,29 -0,27 -0,25 -0,23 -0,21 -0,19 -0,17 -0,15

k3'

k 5''

CEC Hypersil C18-HPLCCEC Hypersil C18-CECHypersil C8-HPLC

Hypersil C8-CEC

Hypersil Phenyl-HPLCHypersil Phenyl-CEC

Spherisorb ODS-HPLCSpherisorb ODS-CECSpherisorb C8-HPLC

Spherisorb C8-CEC

Unimicro C18-HPLC

Unimicro C18-CEC

Unimicro C8-HPLC

Unimicro C8-CEC

Unimicro Phenyl-HPLCUnimicro Phenyl-CEC

Phenyl stationary phases

CEC –C18 stationary phase

C8 + C18 stationary phases

Outliers

Figure 6.3 Ordering of stationary phases according to their hydrogen-bond donor activity

characterized by the coefficient ''5k for the H

2β variable in Eq. 6.4 and the coefficient '3k for

the 2µ variable in Eq. 6.3.

Figure 6.4 depicts a plot of stationary phase hydrogen-bond basicity, k5" in Table 6.6, under

CEC conditions vs. that under HPLC conditions. Ideal line (tag α = 1) and the regression

line for all the stationary phases are given. Similarly as in Fig. 6.3 one can see separate

clusters of the phenyl, the C8 (except Hypersil C8) and the C18 stationary phases. Except

the Hypersil Phenyl stationary phase, all the other stationary phases find themselves above

the ideal line. In other words, under CEC conditions most of the tested stationary phases

exhibit a higher activity of free silanols that at HPLC conditions. In the linear CEC-HPLC

relationship the intercept is statistically different from zero according to the t-test value.


That means that the activity of free silanol groups under CEC conditions is different from

that under HPLC conditions.

-7

-6,8

-6,6

-6,4

-6,2

-6

-5,8

-5,6

-5,4

-5,2

-5

-7 -6,8 -6,6 -6,4 -6,2 -6 -5,8 -5,6 -5,4 -5,2 -5

HPLC

CEC

Hypersil C8

Unimicro C18

Unimicro C8

Spherisorb C18

Spherisorb C8Hypersil C8

Unimicro Phenyl

Hypersil Phenyl

Hydrogen bond basicityy=-1.16+0.78xConfidence interval for the slope:L1,2=0.78±0.19Test of the intercept:t=1.16/0.59=1.95 > tcritical

CEC=HPLC, tg αααα

= 1

Figure 6.4

Plot of regression coefficients at the hydrogen-bond basicity variable ( H2β ) in QSRR

equations derived for HPLC and CEC modes. Error bars and the ideal line tag α = 1 are

given. The critical t-value is a value corresponding to 90% confidence level.

In the same way as for hydrogen-bond basicity one can test regression coefficients at the

hydrogen-bond acidity parameters of analytes, k4" in Table 6.6 (Fig. 6.5). That term

describes the ability of an analyte to donate a proton to form a solute-solvent and/or

solute-stationary phase hydrogen bond. In Fig. 6.5 the majority of stationary phases are

below the ideal line of tag α = 1. The exception is the Spherisorb C8 stationary phase. The

Spherisorb C18 stationary phase could not be included in the plot because for this phase

the values of the coefficient of the solute hydrogen-bond acidity were not statistically

significant. The clustering of stationary phases (especially C8 and C18) is not that evident as

in the case of the coefficients of the solute hydrogen-bond basicity. The intercept in Fig. 6.5

differs significantly from zero implying differences in behaviour under HPLC and CEC

conditions.

- 150 - Chapter 6

-1,9

-1,7

-1,5

-1,3

-1,1

-0,9

-0,7

-1,9 -1,7 -1,5 -1,3 -1,1 -0,9 -0,7

HPLC

CEC

Unimicro C18

Unimicro C8

Hypersil C8

Hypersil C18 Hypersil Phenyl

Unimicro Phenyl

Spherisorb C8

Hydrogen bond acidityy=-0.52+0.69xConfidence interval for the slope:L1,2=0.69±0.38Test of the intercept:t=0.52/0.25=2.08 > tcritical


= 1

Figure 6.5

Plot of regression coefficients at the hydrogen-bond acidity variable ( H2α ) in QSRR

equations derived for HPLC and CEC modes. Error bars and the ideal line tag α = 1 are

given. The critical t-value is a value corresponding to 90% confidence level.

Positive values of the coefficient of the McGowan parameter of analytes, k6" in Table 6.6,

means that the dispersive interactions of the analyte with the hydrocarbonaceous stationary

phase are stronger than analogous interactions with the mobile phase. This explains the

higher values of k6'' observed for the C8 and C18 stationary phases than for the phenyl

stationary phases as the former phases contain more hydrocarbon ligand. When comparing

k6’’ values for the same phase under both HPLC and CEC mode one notes that, with

exception of the Unimicro Phenyl stationary phase, the values obtained in HPLC are

generally higher than those found under CEC conditions.


3,2

3,4

3,6

3,8

4

4,2

4,4

3,2 3,4 3,6 3,8 4 4,2 4,4HPLC

CEC Unimicro C18Hypersil C8

Hypersil C18

Spherisorb C8

Hypersil Phenyl

Unimicro C8

Unimicro Phenyl

McGowan volumey=2.20+0.39xConfidence interval for the slope:L1,2=0.39±0.60Test of the intercept:t=2.20/1.17=1.88 < tcritical


= 1

Figure 6.6 Plot of regression coefficients at the McGowan volume variable (Vx) in QSRR equations

derived for HPLC and CEC modes. Error bars and the ideal line tag α = 1 are given. The

critical t-value is a value corresponding to 90% confidence level.

It means that either dispersive interactions between analyte and the hydrocarbonaceous

phase are stronger under HPLC mode than under CEC mode or the interaction between

the analyte and the mobile phase is stronger under the applied electric field (CEC), or a

combination of both. As discussed in the previous paper [31], it may be that due to the

different orientation of hydrocarbonaceous chains under CEC conditions the interaction of

an analyte with the stationary phase is weaker than under HPLC conditions. The t-test

analysis demonstrated that t-value is lower than critical and the intercept in Fig. 6.6 is

therefore statistically not significantly different from zero.

6.4 Conclusions The QSRR models provide rational interpretation of differences and/or similarities in the molecular

mechanism of chromatographic separations between HPLC and CEC reversed-phase systems. The

models can be of help in objective comparison of separation properties of modern stationary

phases.

- 152 - Chapter 6 Three models of QSRR relating standardized retention parameters as obtained on eight

modern reversed-phase materials, demonstrated the lack of substantial differences in

molecular mechanism of separation which would depend on the nature of the eluent

driving force, i.e., high pressure in the HPLC mode or electroosmotic flow in the CEC

mode. Neither the partition coefficient of analytes, nor their molecular size or polarity

related structural descriptors from molecular modelling or from LSER theory clearly

distinguished separation patterns on the same phase at HPLC and CEC conditions.

Detailed comparative QSRR analysis supplied evidences of stronger non-specific dispersive

interactions attracting analytes to the hydrocarbonaceous stationary phase in the HPLC

mode as related to the CEC mode and a higher activity of free silanols under CEC

conditions with respect to HPLC. These differences do not manifest themselves strongly

enough to substantially change the mechanism of retention in the two modes, however. On

the other hand, these differences are significant enough to distinguish some reversed-phase

materials from the other.

In view of this work there is a rather limited chance that replacing high pressure with

electroosmotic force will result in a dramatic improvement of separations. There are

advantages of CEC over HPLC, like high peak capacity or different selectivity for complex

analytes. On the other hand, CEC implies rather sophisticated technical solutions.

Therefore, a question remains to be answered whether further development of CEC may

result in a cost/effectiveness ratio that will be acceptable from the point of view of practical

analytical applications.

REFERENCES 1. J.W. Jorgenson, K.D. Lukacs, J. Chromatogr., 218 (1981) 209.





6. B.A. Grines, J.J. Meyers, A.I. Liapis, J. Chromatogr. A, 890 (2000) 61.

7. R. Stol, H. Poppe, W.Th. Kok, J. Chromatogr. A, 887 (2000) 199.

8. J. Ståhlberg, J. Chromatogr. A, 892 (2000) 291.



10. K.D. Bartle, R.A. Carney, A. Cavazza, M. Cikalo, P. Myers, M.M. Robson, S.C.P.

Roulin, K. Sealey, J. Chromatogr. A, 892 (2000) 279.

11. S.J. Lane, R. Boughtflower, C. Paterson, M. Morris, Rapid Commun. Mass Spectr., 10

(1996) 733.

12. M.R. Euerby, C.M. Johnson, K.D. Bartle, P. Myers, S.C. Roulin, Anal. Commun., 33

(1996) 403.

13. K. Walhagen, K.K. Unger, M.T. Hearn, J. Chromatogr. A, 893 (2000) 401.

14. P. Huang, X. Jin, Y. Chen, J.R. Srinivasan, D.M. Lubman, Anal. Chem., 71 (1999)

1786.

15. J. Zhang, X. Huang, S. Zhang, Cs. Horváth, Anal. Chem., 72 (2000) 3022.

16. X. Huang, J. Zhang, Cs. Horváth, J. Chromatogr. A, 858 (1999) 91.


18. S.V. Galushko, A.A. Kamenchuk, G.L. Pit, J. Chromatogr. A, 660 (1994) 47.

19. R. Kaliszan, Structure and Retention in Chromatography. A Chemometric Approach,

Harwood Academic Publishers, Amsterdam, 1997.

20. R. Kaliszan, Anal. Chem., 64 (1992) 619A.

21. P.C. Sadek, P.W. Carr, R.M. Doherty, M.J. Kamlet, R.W. Taft, M.H. Abraham, Anal.

Chem., 57 (1985) 2971.

22. P.W. Carr, Microchem. J., 48 (1993) 4.

23. P.W. Carr, R.M. Doherty, M.J. Kamlet, R.W. Taft, W. Melander, Cs. Horváth, Anal.

Chem., 58 (1986) 2674.

24. L.C. Tan, P.W. Carr, M.H. Abraham, J. Chromatogr. A, 752 (1996) 1.

25. M.H. Abraham, M. Ross, C.F. Poole, S.K. Poole, J. Phys. Org. Chem., 10 (1997) 358.


Relat. Technol., 21 (1998) 1433.

27. Z. Liu, H. Zou, M. Ye, J. Ni, Y. Zhang, J. Chromatogr. A, 863 (1999) 69.


29. R. Kaliszan, M.A van Straten, M. Markuszewski, C.A. Cramers, H.A. Claessens, J.

Chromatogr. A, 855 (1999) 455.

30. L.C. Sander, S.A. Wise, Adv. Chromatogr., 25 (1986) 139.

31. J. Jiskra, M. Byelik, C.A. Cramers, H.A. Claessens, J. Chromatogr. A, 862 (1999) 121.

32. J. Jiskra, T. Jiang, H.A. Claessens, C.A. Cramers, J. Microcolumn Sep., 12 (2000) 530.

33. T. Eimer, K. Unger, J. van der Greef, TrAC, Trends Anal. Chem., 15 (1996) 463.

- 154 - Chapter 6 34. M.R. Euerby, C.M. Johnson, S.F. Smyth, N. Gillot, D.A. Barrett, P.N. Shaw, J.


35. D.A. Barrett, V.A. Brown, P.N. Shaw, M.C. Davies, H.J. Ritchie, P.J. Ross, J.

Chromatogr. Sci., 34 (1996) 146.




38. F. Moffatt, P.A. Cooper, K.M. Jessop, J. Chromatogr. A, 855 (1999) 215.


- 155 -

CHAPTER 7

7 THERMODYNAMIC BEHAVIOUR IN CAPILLARY


Summary

The thermodynamic behaviour of solutes was examined on four reversed-phase (RP)

stationary phases under pressure-driven (viz. high performance liquid chromatography,

HPLC) and electro-driven (viz. capillary electrochromatography, CEC) conditions.

Thermodynamic constants as standard enthalpy, entropy and Gibbs free energy were

compared for both separation modes. Differences and/or similarities observed were used

as a tool in better understanding of the separation mechanism in CEC compared to HPLC.

In addition, temperature effects on the magnitude of electroosmotic flow (EOF) are

explained in detail.

7.1 Introduction Optimization of separation in both pressure-driven (viz. high-performance liquid

chromatography, HPLC) and electro-driven chromatography (viz. capillary

electrochromatography, CEC) is a challenging task. In HPLC and CEC, solvent (mobile

phase) strength programming (isocratic or gradient) is usually the best choice to arrive at

optimal separation condition because solvent programming offers large possibilities to

adjust selectivity, analysis time and efficiency [1]. Besides that, most of

This chapter has been accepted for publication in Journal of Separation Science.

- 156 - Chapter 7 HPLC and CEC separations, however, are also carried out at ambient or at constant

elevated temperatures. Wider use of temperature programming is restricted in practice by

the following reasons:

a) many samples e.g. biological samples do not allow analysis at higher temperatures

b) higher temperatures may negatively influence column lifetime.

The major effects of increasing temperature are decrease of retention times and improved

kinetics of the separation [2]. These advantages have also been introduced into CEC.

Jakubetz et al. [3] studied the influence of temperature on resolution and efficiency in

enantiomeric separation by open-tubular (OT) LC and CEC and found that the

temperature effects were less pronounced in OTLC compared to OTCEC. Thiam et al. [4]

applied elevated temperatures in the separation of cholesterol and its derivatives. Analysis

time could be reduced by 50% at 60ºC, however, the resolution deteriorated at

temperatures higher than 40ºC. Dabek-Zlotorzynska et al. [5] reached 45% reduction in

analysis time in the separation of carbonyl 2,4-dinitrophenylhydrazones in CEC using

higher temperature. Djordjevic et al. [6] investigated separations of mixtures of cortisones

and benzenes exhibiting a wide range of retention factor (0.41<k<11.65). Optimised

separation was performed using temperature programming from 25ºC to 60ºC with a rate

of 3ºC/min. Wolf et al. [7] investigated the resolution of enantiomers on brush-type chiral

stationary phases. They studied the dependence of retention factor on temperature. Lin et

al. [8] studied the temperature effects on chiral recognition of some amino acids with

molecular imprinted polymer as stationary phase. Higher temperatures for enantiomeric

separations by CEC using a macrocyclic antibiotic chiral stationary phase was employed

by group of Carter-Finch [9]. Extensive studies of temperature effects on the separation

of retinyl esters in non-aqueous CEC were performed by Roed et al. [10-12] on the

Hypersil ODS and home made C30 stationary phase. Recently from the same group,

Greibrokk and Andersen [13] summarized temperature programming in liquid

chromatography including electrochromatography.

In many chromatographic systems, a linear relationship (Van ‘t Hoff plots) is found

between logarithm of retention factor (k) and reciprocal value of (absolute) temperature

according to the following equation:

φlnln +∆+∆−=RS

RTHk

��

(7.1)

Thermodynamic Behaviour in CEC - 157 -

where �H∆ is the standard enthalpy of transfer of a solute from the mobile phase to the

stationary phase, �S∆ is the standard entropy of transfer of a solute from the mobile

phase to the stationary phase, φ is the phase ratio, R is the gas constant and T is the

absolute temperature. Discontinuities in linearity are typically caused by:

i) retention by a mixed mechanism

ii) change in solute conformation

iii) existence of more forms of the solute having different retention properties

iv) conformation changes in hydrocarbonaceous phase of the stationary phase [14]

Roulin et al. [15] studied Van ‘t Hoff plots for capsaicin and found above described linear

relationship. Wen et al. [16] extensively studied dynamics in capillary

electrochromatography and found linear Van ‘t Hoff plot for the Spherisorb ODS

column; electroosmotic flow (EOF) increased with temperature as a result of the decrease

of viscosity of the mobile phase. Cahours et al. [17] studied the performance of

benzodiazepines in CEC on phenyl silica stationary phases and also found linear

relationship as described by the Eqn. 7.1. Zhang et al. [18] studied rapid separation of

peptides and proteins by isocratic CEC at elevated temperatures and found linear

relationships of logarithm k vs. 1/T and velocity of EOF vs. 1/T. On the contrary,

Walhagen et al. [19] found discontinuation in the linear relationship of Van ‘t Hoff plot

between 30 and 40ºC by studying temperature influence on the behaviour of small

peptides (enkephalins) in CEC. These latter authors attribute the change to the

reorganization of n-octadecyl chains bonded to the silica surface. Furthermore, Walhagen

et al. also found a linear relationship of the velocity of EOF vs. T . Djordjevic et al. [20]

used thermodynamic data to compare retention mechanisms of neutral solutes between

CEC and HPLC. The authors found retention factors on the Hypersil C18 column to be

lower under CEC conditions than under HPLC conditions. Moreover, retention

enthalpies were also lower in CEC than in HPLC and in the CEC mode �G∆ (free energy

of transfer of a solute from the mobile phase to the stationary phase) values were positive

while in HPLC mode negative values were obtained. This shows that solute transfer from

the mobile phase to the stationary phase in CEC is less favourable than in HPLC. These

authors attribute these differences, in part, to the self-heating (Joule heating) which occurs

in electro-driven systems. Joule heating and influence of temperature on retention and

- 158 - Chapter 7 resolution in columns packed with a supercritical fluid carrier was also reviewed by

Robson et al. [21].

This study evaluates temperature effects on retention, standard enthalpy, entropy, Gibbs

energy and electroosmotic flow under pressure and electro-driven conditions on four

reversed-phase stationary phases. Findings in solute transfer differences help in deeper

understanding of molecular retention mechanism in pressure- and electro-driven

chromatography.

7.2 Experimental 7.2.1 Chemicals Acetonitrile (ACN) with HPLC grade purity was from Biosolve (Valkenswaard, the

Netherlands); tris(hydroxymethyl)aminomethane (Tris) and hydrochloric acid (fuming,

37%) of analytical purity was from Merck (Darmstadt, Germany). Thiourea, 4-cyanophenol

(4-hydroxybenzonitrile), caffeine, N,N-dimethylaniline, N-methyl-2-pyrrolidinone (1-

methylpyrrolidin-2-one), 4-iodophenol, hexachlorobutadiene, dibenzothiophene, n-

hexylbenzene, 1,3,5-triisopropylbenzene and naphthalene were obtained from recognized

laboratory chemicals suppliers. Samples were prepared by dissolving these compounds in

the mobile phase or in the pure organic modifier and then diluted with water. The mobile

phase was composed of 70% acetonitrile:30% aqueous Tris buffer, 5 mM total

concentration (V/V), filtered through a filter (pore size = 0.45 µm) and degassed by

ultrasonication before use.

7.2.2 Columns The columns used in this study are listed in Table 7.1 together with relevant data provided

by the manufacturer. The column packed bed was 25 cm, and 33.5 cm total length. Prior

to use in the CEC mode, the columns were conditioned. This was accomplished by

applying 10 bar pressure on both sides of the column and increasing the voltage from 0-

25 kV in 5 kV steps per 10 min. Next to that the pressure was increased to 12 bar and a

30 kV voltage was applied for 10 min.


Table 7.1 List of columns; each column diameter, 100 µm; average particle size, 3 µm;

note that the Unimicro columns contains 10% of pure silica to provide stable EOF [36]

Column Pore Size Pore Volume

Surface Area

Carbon Load

CEC Hypersil C18

130 Å 0.65 cm3/g 170 m2/g 8.5%

Hypersil C8

120 Å 0.65 cm3/g 170 m2/g 6.5%

Unimicro C18

Data not available

Unimicro C8

Data not available

For the micro-HPLC experiments, the columns were conditioned until the column pressure

was stabilized (approx. 1 h). Note that in these experiments the columns were tested under

pressure- and electro-driven conditions using the same batches of eluents. HPLC

experiments were adjusted to similar flow velocities as obtained in CEC. As a consequence

the HPLC experiments are not optimized with respect to plate height.





electrokinetically (5 kV for 2-15 s). For each run a voltage of 20 kV (600 V.cm-1 electrical

field strength) was applied with 10-bar pressure on both ends of a capillary. The detection

wavelength was 210 nm. High voltage was applied as 6-s time ramp to avoid column stress.

The column cassette temperature was maintained at various temperatures from 7.5 to 60ºC.


- 160 - Chapter 7 syringe pump (Carlo Erba Instruments, Milan, Italy), a microUVIS20 ultraviolet/visible

absorbance detector (Carlo Erba Instruments, Milan, Italy) operating at 210 nm, and an

injector with a 200 nL loop (VICI-AG Valco Europe, Schenkon, Switzerland). The flow-

rate was approximately 0.2-0.3 µL/min using a 1/100-flow splitter (VICI-AG Valco

Europe, Schenkon, Switzerland). The column temperature was maintained at various

temperatures from 7.5 to 60ºC using a water bath (Thermo NESLAB, Portsmouth, NH).

7.2.4 Test procedure Columns were tested under both CEC and HPLC conditions using the same batch of

eluent and the sample mixture containing thiourea, caffeine, 4-cyanophenol, N-methyl-3-

pyrrolidinone, 4-iodophenol, N,N-dimethylaniline, naphthalene, dibenzothiophene,

hexachlorobutadiene, n-hexylbenzene and 1,3,5-triisopropylbenzene. The temperature of

the capillaries was controlled by air circulation (CEC experiments, 0.1ºC accuracy) and

water circulation (HPLC experiments, 0.1ºC accuracy). Please, note that the same portions

of the capillary have been controlled under both separation modes; here, 80% of the

packed section (with respect to the commercial instrumentation). The flow under HPLC

conditions was measured by a 10 µL syringe.


Table 7.2 Structural descriptors and constants of test analytes.

No. Solute M.W. log P pKa Vx SAS

1 4-Cyanophenol 119.12 1.60 7.97 0.930 290.61

2 Caffeine 194.19 -0.07 14.00 1.363 367.02

3 N-Methyl-2-pyrrolidinone 99.13 -0.38 - 0.820 270.53

4 N,N-Dimethylaniline 121.18 2.31 5.15 - -

5 4-Iodophenol 220.01 2.91 9.21 1.033 301.47

6 Naphthalene 128.18 3.30 - 1.085 313.25

7 Dibenzothiophene 184.26 4.38 - 1.379 364.54

8 Hexachlorobutadiene 260.76 4.78 - 1.321 352.14

9 n-Hexylbenzene 162.28 5.52 - 1.562 415.40

10 1,3,5-Triisopropylbenzene 204.36 6.36 - 1.985 478.27

M.W.=molecular weight (g/mol); log P = logarithm of n-octanol-water partition coefficient;

Vx = characteristic volume of McGowan; pKa=dissociation constant; SAS = solvent

(water)-accessible molecular surface area (in Å2). (-) – data not available

7.3 Results and discussion 7.3.1 Effect of temperature on the electroosmotic flow The relationship between the temperature and the velocity of the electroosmotic flow can

be described as follows [22]:

ηζεε Ev r

EOF0= (7.2)

while for a monovalent electrolyte the zeta potential (ζ) can be expressed by the following

equation:

21

202

=

cFRT

rεεσζ (7.3)

where vEOF is the velocity of the electroosmotic flow, σ the superficial excess charge

density, ε0 the permittivity of vacuum, εr the relative permittivity, R the gas constant, T the

absolute temperature, c the molar concentration, F the Faraday constant, E the electric field

- 162 - Chapter 7 strength and η the viscosity. The viscosity is also temperature dependent and can be

described by the equation [23]: 2

10log DTCTTBAL +++=η (7.4)

where A, B, C and D are constants and T is the absolute temperature. The temperature

effect on permittivity can be expressed by the following equation (abbreviated form) [24]: 2)( cTbTaTr ++=ε (7.5)

where a, b and c are constants for given temperature range and T is the absolute

temperature. Walhagen et al. [19] plotted the velocity of EOF against T and found a

linear relationship with regression coefficient up to 0.9988. The authors claim that decrease

in viscosity with increasing temperature has little effect on EOF compared to the impact of

temperature on zeta potential resulting in a linear relationship of the EOF value vs. T . It

is of interest to note that changes in temperature from 10 to 60ºC results in a change of

T from 16.83 to 18.25 (theoretical increase of veof by 8%) while viscosity (e.g. for ACN)

changes from 0.40 to 0.26 mPa.s (theoretical increase of veof by 54%) and relative

permittivity (e.g. for ACN) from 39.2 to 31.3 (theoretical decrease of veof by 11%). In this

particular case, the increase of veof was 60%; note that values of viscosity and relative

permittivity vary a lot for acetonitrile-water mixtures [e.g. 25]. Zhang et al. [18] and

Djordjevic et al. [6] found linear relationship for 1/T as a result of viscosity change. As can

be seen from Figures 7.1a), 7.1b) and 7.1c), electroosmotic flow dependence for the CEC

Hypersil C18 stationary phase is linearly related to both T (regression coefficient 0.9998),

1/T (regression coefficient 0.9994) as well as logarithm of the velocity of EOF against 1/T

(regression coefficient 0.9989); in fact also dependency of veof on T is linear (regression

coefficient of 0.9996; plot not shown). This is rather surprising considering the fact that the

dependency of the velocity of EOF on temperature is rather complicated (see Eqns. 7.2-

7.5).


Figure 7.1

Relationship of velocity of electroosmotic flow on a) T , b) 1/T, and c) the relationship of

logarithm of the velocity of electroosmotic flow on 1/T, where T is temperature in K.

Experimental conditions; mobile phase: 70% ACN/30% Tris buffer (5 mM as a total

concentration), column: CEC Hypersil C18, voltage: 20 kV, EOF marker: thiourea.

y = 0.433x - 6.2662

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

16.5 16.7 16.9 17.1 17.3 17.5 17.7 17.9 18.1 18.3 18.5

T1/2 (K)

vEO

F (

mm

/s)

a)

b)

y = -1152.3x + 5.083

0.5

0.7

0.9

1.1

1.3

1.5

1.7

1.9

0.0028 0.0029 0.003 0.0031 0.0032 0.0033 0.0034 0.0035 0.0036 0.0037

1/T (K-1)

vEO

F (m

m/s

)

c)

y = -901.65x + 3.2086

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.0029 0.003 0.0031 0.0032 0.0033 0.0034 0.0035 0.0036 0.0037

1/T (K-1)

ln v

EOF

- 164 - Chapter 7 7.3.2 Van ‘t Hoff plots The proper understanding of retention mechanism in capillary electrochromatography is a

fundamental problem for this recently developed separation technique. At first, it is

believed that partitioning of neutral solutes on the reversed-phase CEC is identical or at

least very similar to reversed-phase HPLC. However, many articles have been published

revealing dissimilarities in retention [26-34]. Retention of neutral solutes (excluding

electrophoretic mobility from retention) in both HPLC and CEC is related to the

concentration of a solute in the stationary (s) and mobile (m) phase:

φKk = (7.6)

where K is a distribution coefficient characterizing portion (concentration, c) of a solute in

the mobile and stationary phase ( sm ccK = ) and φ is phase ratio ( ms VV ). Solute

concentration in each phase (K) is determined thermodynamically by the Gibbs (free)

energy of transfer of a solute from the mobile phase to the stationary phase:

KRTG ln−=∆ � (7.7) �� STHG ∆−∆=∆ (7.8)

The combination of Eqns. 7.6, 7.7 and 7.8 leads to Eqn. 7.1. It is obvious that the study of

the thermodynamic behaviour ( �H∆ and �S∆ of solute transfer) is a very useful tool for a

deeper understanding of the retention mechanism (mechanism of solute interaction with

bonded stationary phase) in CEC and HPLC. As known from previous studies [26-27] and

other literature [28-34], retention behaviour of uncharged polar and non-polar compounds

may differ from that in HPLC. For the non-polar compound the retention on the same

column is lower under CEC conditions than under HPLC conditions. The reverse is true

for the polar compound. Furthermore, both slopes and intercepts (viz. �H∆ and �S∆ of

solute transfer) clearly differ for both compounds and both eluent-driven modes. Tables 7.3

a), b) summarize �H∆ and �S∆ values for all columns and compounds; correlations

coefficients of the relationship o ln k vs. 1/T are given, too; the phase ratio needed for �S∆

was calculated using flow measurements and capillary parameters. The values, varying from

0.327 for the Hypersil C8 columns to 0.404 for the Hypersil C18 column, are in agreement

with other literature values [34].


Figure 7.2

Relationship between ln k and reciprocal temperature for a) dibenzothiophene on the CEC

Hypersil C18 stationary phase and b) N-methyl-2-pyrrolidinone on the Hypersil C8

stationary phase. Experimental conditions: see text.

ln k= 902.07/T - 2.4482

ln k = 1050.2/T - 3.0023

-0.1

0.1

0.3

0.5

0.7

0.9

0.0029 0.003 0.0031 0.0032 0.0033 0.0034 0.0035 0.0036

1/T (K-1)

ln k

CEC

HPLC

a)

b)

ln k= 612.28/T - 3.4233

ln k= 653.41/T - 3.4945

-2

-1.9

-1.8

-1.7

-1.6

-1.5

-1.4

-1.3

-1.2

-1.1

-1

0.0029 0.003 0.0031 0.0032 0.0033 0.0034 0.0035 0.0036

1/T (K-1)

ln k

CEC

HPLC

- 166 - Chapter 7 Table 7.3 a) Standard enthalpies and entropies (together with their 95% confidence

limits) of the tested solutes on the Hypersil columns ∆H°/kJ.mol-1 ∆S°/J.mol-1.K-1

Column Compound HPLC CEC

∆rel

CEC-HPLC/HPLC

(%)

HPLC CEC

∆rel

CEC-HPLC/HPLC

(%)

4-cyanophenol -7.86±0.69 -7.39±0.39 -5.92 -37.32±2.29 -29.28±1.30 -21.53

caffeine -6.09±0.38 -5.91±0.16 -2.88 -27.99±1.25 -26.35±0.53 -5.85

N-methyl-2-pyrrolidinone -5.09±0.35 -5.43±0.15 6.71 -20.93±1.17 -21.52±0.50 2.83

4-iodophenol -6.19±0.14 -5.98±0.11 -3.29 -22.68±0.48 -22.21±0.36 -2.08

N,N-dimethylaniline -5.19±0.08 -5.99±0.09 15.42 -12.55±0.25 -15.52±0.31 23.71

naphthalene -5.31±0.09 -6.01±0.13 13.10 -11.05±0.31 -13.73±0.44 24.20

dibenzothiophene -7.50±0,07 -8.73±0.09 16.42 -12.82±0.24 -17.43±0.31 35.94

hexachlorobutadiene -4.55±0.23 -5.08±0,30 11.83 -0.74±0.78 -2.90±1.00 291.27

n-hexylbenzene -6.82±0.16 -7.85±0.26 15.11 -4.81±0.53 -8.81±0.87 83.40

CEC Hypersil C18

1,3,5-triisopropylbenzene -5.46±0.28 -6.10±0.36 11.76 1.55±0.92 -1.20±1.20 -177.66

4-cyanophenol -7.60±0.55 -7.89±0.66 3.72 -31.93±1.85 -30.21±2.20 -5.38

caffeine -0.22±0.37 1.18±0.12 -631.38 -16.47±1.25 -12.25±0.40 -25.59

N-methyl-2-pyrrolidinone -2.05±0.75 -0.14±0.14 -93.12 -19.23±2.48 -13.09±0.45 -31.89

4-iodophenol -5.86±0.23 -6.41±0.11 9.29 -18.40±0.77 -20.47±0.36 11.23


naphthalene -6.33±0.16 -6.90±0.15 9.07 -12.67±0.55 -14.68±0.49 15.89

dibenzothiophene -7.18±0.16 -8.07±0.18 12.35 -11.75±0.53 -14.95±0.61 27.18

hexachlorobutadiene -6.54±0.21 -7.37±0.27 12.72 -5.07±0.72 -7.99±0.89 57.43

n-hexylbenzene -7.66±0.22 -8.80±0.29 14.94 -5.91±0.74 -10.09±0.97 70.69

Hypersil C8

1,3,5-triisopropylbenzene -7.21±0,29 -8.26±0.36 14.52 -0.91±0.95 -4.91±1.19 441.32


Table 7.3 b) Standard enthalpies and entropies (together with their 95% confidence

limits) of the tested solutes on the Unimicro columns ∆H°/kJ.mol-1 ∆S°/J.mol-1.K-1

Column Compound HPLC CEC

∆rel

CEC-HPLC/HPLC

(%)

HPLC CEC

∆rel

CEC-HPLC/HPLC

(%)

4-cyanophenol - -15.34±2.15 - -61.09±7.19 -

caffeine - -2.08±0.23 - - -25.25±0.77 -

N-methyl-2-pyrrolidinone - 0.41±0.84 - - 13.85±19.90 -

4-iodophenol -8.65±1.04 -5.86±0.11 -32.31 -39.67±3.49 -26.92±0.38 -32.14


naphthalene -5.04±0.14 -5.44±0.06 7.87 -16.36±0.48 -16.74±0.20 2.38



n-hexylbenzene -7.44±0.11 -7.82±0.14 5.09 -10.84±0.36 -12.30±0.46 13.40

Unimicro C18

1,3,5-triisopropylbenzene -5.32±0,24 -6.09±0.33 14.62 -1.62±0.81 -4.31±1.10 166.73

4-cyanophenol - -4.67±0.66 - - -19.45±2.19 -

caffeine -5.00±0.69 -0.06±0.21 -98.74 -24.19±2.29 -10.59±0.69 -56.23

N-methyl-2-pyrrolidinone -3.05±1.05 -1.81±0.14 -40.74 -13.06±3.51 -12.14±0.48 -7.07

4-iodophenol -4.34±0.63 -4.88±0.11 12.56 -11.43±2.11 -14.75±0.36 29.03


naphthalene -4.85±0.39 -5.29±0.13 8.99 -8.09±1.28 -8.86±0.43 9.63



n-hexylbenzene -5.97±0.34 -6.54±0.25 9.57 -0.73±1.11 -3.18±0.82 338.09

Unimicro C8

1,3,5-triisopropylbenzene -5.71±0.39 -6.03±0.31 5.53 3.55±1.28 1.85±1.04 -47.78

Obviously, substantial differences in �H∆ and �S∆ values are observed. Analyzing data for

non-polar compounds, it can be calculated that �H∆ in CEC has always a more negative

value compared to HPLC. It means that differences in bonding energy between solute-

stationary phase and solute-mobile phase are greater under CEC conditions. The difference

between HPLC and CEC modes in terms of �H∆ for tested compounds fluctuate from

5.09% for n-hexylbenzene on the Unimicro C18 columns up to 16.42% for

dibenzothiophene on the CEC Hypersil C18 column. Clearly, from Fig. 7.2a) lower

retention of these compounds under CEC conditions and higher negative value of standard

enthalpy causes higher differences of retention at higher temperatures than at lower

temperatures. The situation is different for polar compounds. N,N-Dimethylaniline

- 168 - Chapter 7 typically follows retention trends of non-polar compounds as known from HPLC [35].

Other polar compounds tested reveal higher values of �H∆ under CEC conditions with the

exception of 4-cyanophenol on the Hypersil C8 columns, 4-iodophenol on the Hypersil C8

and the Unimicro C8 column and N-methyl-2-pyrolidinone on the CEC Hypersil C18

column. Caffeine on the Hypersil C8 column and N-methyl-2-pyrrolidinone on the

Unimicro C18 column have positive values of standard enthalpy of transfer under CEC

condition. Apparently, their bonding energy with the mobile phase is stronger than with the

stationary phase. There is a clear correlation of the �H∆ values of caffeine and N-methyl-2-

pyrrolidinone to their negative log P (partition coefficient between n-octanol and water)

values (Table 7.2). Other factors such as entropy contributes to the retention of theses

solutes on reversed-phase stationary phases. It is also interesting to analyze �S∆ values.

Negative values of �S∆ means that a solute is more ordered on the stationary phase than in

the mobile phase. In CEC (Tables 7.3a) and b)), with the exception of some polar

compounds, most of the �S∆ values are more negative compared with HPLC. This can be

explained by two assumptions:

1) origin of the electroosmotic flow on the silica support increases ordering of the

hydrocarbonaceous chains of the stationary phase and thus increases ordering of the

solutes on the stationary phase

2) mobile phase structure (uniformity) gets more organized under applied electric field

and thus more disturbed by solute cavity.

The second assumption explains the higher �S∆ values in CEC for polar compounds as

they can easier penetrate into the mobile phase structure with e.g. hydrogen bonds.

Increased organization of hydrocarbonaceous chains on the stationary phase also confirms

more negative values of enthalpy under CEC conditions and thus better contact of solutes

with the stationary phase. Interestingly, for 1,3,5-triisopropylbenzene the �S∆ values on the

Unimicro C8 column are positive for both HPLC and CEC. This indicates superior

ordering of the solute in the mobile phase than on the stationary phase. Comparing �H∆

and �S∆ (or T �S∆ as in Eqn. 7.6) values, we can see that �S∆ values play more important

role in retention of polar rather than non-polar compounds. The �G∆ values are

summarized in Table 7.4 and valid for the temperature of 293.15 K (20°C). The values

reveal that for the majority of solutes on the Hypersil C18, Hypersil C8 and Unimicro C8


stationary phases the solute transfer from the mobile phase to the stationary phase is

thermodynamically less favourable under CEC conditions that under HPLC conditions.

The exception is the Unimicro C18 column where this conclusion is valid only for two

most hydrophobic compounds (n-hexylbenzene and 1,3,5-triisopropylbenzene). This

finding is in agreement with finding of the group of Djordjevic et al. [20]. The explanation

can be found in differences of separation mechanism, which implicates that the overall

thermodynamic behaviour of compounds is dependent of the type of eluent driven mode.

Table 7.4 Gibbs energy of transfer at 293.15 K.

∆G°/kJ.mol-1 (at 293.15 K)

Column

CEC Hypersil

C18 Hypersil C8 Unimicro C18 Unimicro C8

Compound

HPLC CEC HPLC CEC HPLC CEC HPLC CEC

4-cyanophenol 3.08 1.19 1.76 0.97 - 2.57 - 1.03

caffeine 2.11 1.81 4.61 4.77 - 5.32 2.10 3.04

N-methyl-2-pyrrolidinone 1.04 0.88 3.58 3.70 - 4.47 0.78 1.75

4-iodophenol 0.46 0.53 -0.47 -0.41 2.98 2.04 -0.99 -0.56

N,N-dimethylaniline -1.51 -1.44 -1.90 -1.88 0.58 0.19 -2.37 -2.12

naphthalene -2.07 -1.98 -2.61 -2.60 -0.25 -0.53 -2.48 -2.69

dibenzothiophene -3.74 -3.62 -3.73 -3.68 -2.11 -2.15 -3.84 -3.63

hexachlorobutadiene -4.33 -4.23 -5.05 -5.02 -2.94 -3.11 -5.06 -4.95

n-hexylbenzene -5.41 -5.26 -5.92 -5.84 -4.26 -4.21 -5.76 -5.61

1,3,5-triisopropylbenzene -5.91 -5.74 -6.94 -6.82 -4.84 -4.83 -6.75 -6.57

7.4 Conclusions Thermodynamic studies in capillary electrochromatography reveal that thermodynamic

constants of solutes obtained under CEC conditions significantly differ from those

obtained under HPLC conditions. Results on temperature dependence of electroosmotic

flow do not confirm that Joule heating causes differences in thermodynamic behaviour of

solutes in electrochromatographic systems. As a consequence, structural changes within the

- 170 - Chapter 7 stationary phase and the mobile phase such as increased organization of

hydrocarbonaceous chains on the stationary phase and increased mobile phase structure are

the most probable influence in CEC. Moreover, there exist statistically significant

dissimilarities in thermodynamic behaviour between polar and non-polar solutes when

compared under both eluent-driven modes. The results show that for the enthalpy values,

non-polar compounds show more negative values under CEC conditions rather than under

HPLC conditions. The reverse is true for polar compounds with the exception of 4-

cyanophenol on the Hypersil C8 columns, 4-iodophenol on the Hypersil C8 and the

Unimicro C8 column and N-methyl-2-pyrolidinone on the CEC Hypersil C18 column.

Similarly, increased organization of the stationary phase possibly causes better organization

of non-polar solutes in the stationary phase under CEC conditions (more negative values of

entropy) than under HPLC conditions. While for polar compounds higher organization

structure of the mobile phase increases organization of the solutes in the mobile phase

under CEC conditions. Based on these findings, we conclude that for neutral, uncharged

solutes the retention mechanism under electro-driven conditions statistically significant

differ from the retention mechanism under pressure-driven conditions.

REFERENCES

1. L.R. Snyder, J.W Dolan, Chem. Anal. (Warsaw), 43 (1998) 495.

2. J.H. Knox, Chromatographia, 26 (1988) 329.

3. H. Jakubetz, H. Czesla, V. Schurig, J. Microcolumn Sep., 11 (1999) 421.


(2000) 2541.

5. E. Dabek-Zlotorzynska, E.P.C. Lai, J. Chromatogr. A, 853 (1999) 487.

6. N.M. Djordjevic, F. Fitzpatrick, F. Houdiere, G. Lerch, G. Rozing, J. Chromatogr. A,

887 (2000) 245.

7. C. Wolf, P.L. Spence, W.H. Pirkle, D.M. Cavender, E.M. Derrico, Electrophoresis, 21

(2000) 917.

8. J.-M. Lin, T. Nakagama, K. Uchiyama, T. Hobo, Biomed. Chromatogr., 11 (1997) 298.

9. A.S. Carter-Finch, N.W. Smith, J. Chromatogr. A, 848 (1999) 375.




12. L. Roed, E. Lundanes, T. Greibrokk, J. Sep. Sci., 24 (2001) 435.

13. T. Greibrokk, T. Andersen, J. Sep. Sci., 24 (2001) 899.

14. R.K. Gilpin, P. Kasturi, J. Microcolumn Sep., 12 (2000) 236.

15. S.P. Roulin, K.D. Bartle, P. Myers, M.R. Euerby, unpublished results.


17. X. Cahours, Ph. Morin, M. Dreux, J. Chromatogr. A, 845 (1999) 203.

18. S. Zhang, J. Zhang, Cs. Horváth, J. Chromatogr. A, 914 (2001) 189.

19. K. Walhagen, K.K. Unger, M.T.W Hearn, J. Chromatogr. A, 893 (2000) 401.

20. N.M. Djordjevic, P.W.J. Fowler, F. Houdiere, G. Lerch, J. Liq. Chromatogr. Relat.

Technol., 21 (1998) 2219.

21. M.M Robson, S. Roulin, S.M. Shariff, M.W. Raynor, K.D. Bartle, A.A. Clifford, P.



23. C.L. Yaws, Handbook of Viscosity, Gulf Publishing Company, Houston, Texas, 1995.

24. D.L. Ride, CRC Handbook of Chemistry and Physics, 82nd Edition, CRC Press, Boca

Raton, 2001.

25. C. Moreau, G. Douhéret, Thermochim. Acta, 13 (1975) 385.


27. J. Jiskra, T. Jiang, H.A. Claessens, C.A. Cramers, J. Microcolumn Sep., 12 (2000) 530.

28. T. Eimer, K. Unger, J. van der Greef, TrAC, Trends Anal. Chem., 15 (1996) 463.



Tanaka, Anal. Chem., 72 (2000) 1275.


Relat. Technol., 21 (1998) 1433.

32. C. Chaiyasut, T. Tsuda, S. Kitagawa, H. Walda, T. Monde, Y. Nakabeya, J.


33. T. Jiang, J. Jiskra, H.A. Claessens, C.A. Cramers, J. Chromatogr. A 923 (2000) 215.

34. L.C. Sander, L.R. Field, Anal. Chem., 52 (1980) 2009.

35. S.J. Schmitz, R. Zwanziger, H. Engelhardt, J. Chromatogr., 544 (1991) 381.

36. M.T. Dulay, C. Yan, D.J. Rakestraw, R.N. Zare, J. Chromatogr. A, 725 (1996) 361.

- 172 - Chapter 7

- 173 -

CHAPTER 8 8 METHOD DEVELOPMENT FOR THE SEPARATION OF

STEROIDS BY CAPILLARY


Summary

A rapid capillary electrochromatography (CEC) method was developed to separate five

structurally related steroid compounds from the production line of steroid hormones. The

separation was performed on a Hypersil C8 MOS and Unimicro C18 stationary phases

using acetonitrile (ACN), methanol (MeOH) and tetrahydrofuran (THF) as organic

modifiers and tris(hydroxymethyl)aminomethane (Tris) as buffer additive. The Hypersil

C8 MOS stationary phases performed best together with ACN as organic modifier and

Tris buffer. The method was extensively tested on ruggedness with respect of sensitivity

to temperature, ACN composition, pH change, concentration of Tris buffer, injected plug

length and run-to-run and day-to-day repeatability. The minimal detectable concentration

and amount were investigated for quantification purposes. The developed CEC method

was shown to be fast, rugged and well suited for quantification of the steroids under

study.


8.1 Introduction Steroids are compounds possessing the skeleton of cyclopenta[a]phenanthrene or a

skeleton therefrom by one or more bond scissions or ring expansion or contraction [1].

Steroids play an important role in human life – cholesterol is one of the main components

- 174 - Chapter 8 of cell membranes; testosterone, progesterone, cortisole, estradiol are important steroid

hormones, cholic acids assure absorption of fats e.g. during digestion; vitamin D is

important for proper body formation; many synthetic steroids found their use in

medicine. Capillary electrochromatography (CEC) as a separation technique offers high

efficient, fast and very often baseline separations in pharmaceutical analysis such as the

analyses of steroids also providing high peak capacity. Steroids are in many cases neutral,

lipophilic or very lipophilic compounds whereas the high-performance liquid

chromatography (HPLC) protocols may be easily transferred to CEC.

Thiam et al. [2] employed CEC in separation of cholesterol and its ester derivatives. The

effects of acid modifier, buffer concentration, mobile phase composition, applied voltage,

temperature and pseudostationary phase were thoroughly studied. Stead et al. [3], Seifar et

al. [4,5] studied the CEC analyses of endogenous steroids such as testosterone and

progesterone; Que et al. [6] used a macroporous acrylic monolithic stationary phase for

isocratic and gradient analysis of endogenous steroids and their derivatives in CEC with

laser-induced fluorescence (LIF) and electrospray ionization (ESI) mass spectrometry

(MS) detection; Taylor et al. [7] showed application of CEC in the analysis of

corticosteroids. Wang et al. [8] applied CEC in the analysis of norgestimate and its

potential degradation products and reached detection limits as low as 0.01% for

degradation impurities. Lord et al. [9] investigated the use of tapered and narrow restrictor

capillaries for use in CEC and demonstrated their use in on-line CEC-MS separation of

steroids such as bufalin and digitoxigenin. Euerby et al. [10-12] extensively studied use of

CEC in pharmaceutical analysis e.g. analysis of tipredane and related substances or

analysis of tipredane and its C-17 diastereoisomer or very fast analysis of three

pharmaceutically active steroids using short-end injection technique. Smith et al. [13] used

CEC for the separation of Fluticasone propionate from a steroid test mixture.

This article deals with the development of a method and the establishment of the

protocol for the separation of five closely related pharmaceutically active steroids by

capillary electrochromatography.

Method Development for the Separation … - 175 -

8.2 Experimental 8.2.1 Chemicals Acetonitrile (ACN), methanol (MeOH) and tetrahydrofuran (THF) with HPLC grade

purity were obtained from Biosolve (Valkenswaard, the Netherlands);

tris(hydroxymethyl)aminomethane (Tris) and hydrochloric acid (37% w/w) of analytical

purity were purchased from Merck (Darmstadt, Germany). Steroid samples were kindly

supplied by Organon (Oss, the Netherlands). The stock concentration of the steroids was 1

mg/mL of the mobile phase, for the limit of detection the concentration range varied from

1 mg/mL down to 0.01 mg/mL of the mobile phase. Structures of the steroids are depicted

in Figure 8.1. The mobile phases were composed of different amounts of ACN, MeOH or

THF and buffered with Tris buffer (5 mmol.L-1) at pH 8.0. Tris concentration and pH were

varied during the validation procedure. All mobile phases were filtered through a filter (pore

size = 0.45 µm) and degassed by ultrasonication before use. As an EOF marker (t0),

thiourea (Sigma-Aldrich, Steinheim, Germany) has been used in concentration of 1 mg/mL

in water.

- 176 - Chapter 8

O

H H

OH

OH

H HO

H H

OH

H

O

H H

OH

H HO

OO

H

H

OH

O

H H

OH

H H

#1 = Org 38585 #2 = Org 41423

#3 = Org 2761 #4 = Org 34517

#5 = Org 41634 Figure 8.1 Structures of the tested steroids.

8.2.2 Columns Two columns were chosen for the investigation: a Unimicro C18 (Unimicro Technologies,

Pleasanton, CA), column dimensions: 100 µm I.D. × 40 cm total length (31.5 cm effective)

and a Hypersil MOS C8 (Thermo Hypersil-Keystone, Runcorn, UK), column dimensions:

100 µm I.D. 33.5 cm total length (25 cm effective). All columns were conditioned by

mobile phase using a syringe pump and then in the CEC mode by applying 10 bar pressure

on both sides of the column and increasing the voltage from 0-25 kV in 5 kV steps per 10

min.






electrokinetically (5 kV for 2-15 s). For each run a voltage of 20 kV was applied with 10-bar

pressure on both ends of the capillary. The detection wavelength was 210 nm. High voltage

was applied as 6-s time ramp to avoid column stress. The column cassette temperature was

maintained at 20ºC.

8.2.4 Prediction software Based on the relationship between CEC and HPLC, Chromsword® HPLC optimization

software (Merck GmbH, Darmstadt, Germany) was employed for preliminary experiments

to estimate starting CEC conditions. This program software contains a set of stationary

phases tested in acetonitrile (ACN) and methanolic (MeOH) mobile phases and also allows

drawing a structure of analytes to be tested. The program further calculates certain

structural parameters such as molecular volume and predicts the best separation on the set

of the columns or on a specific one. The result on the closest possible stationary phase, the

Hypersil C18 stationary phase, using 70% aqueous acetonitrile as mobile phase can be seen

on Figure 8.2.

#1

#2

#3 #4 #5

Figure 8.2 Simulated chromatogram of the tested steroids on the Hypersil ODS stationary phase using

ChromSword® software, mobile phase: acetonitrile/water 70/30 (V/V).

- 178 - Chapter 8 Table 8.1 Steroid parameters generated by Chromsword® software.

Steroid Parameter

#1 #2 #3 #4 #5

Molecular volume (cm3.mol-1)

220.5 209.4 231.2 320.3 299.50

Molecular surface (cm2.mol-1)

176.5 170.5 182.2 226.4 216.5

Molecular mass (g.mol-1)

324.42 310.39 310.44 430.54 388.55

∆G (kJ.mol-1) -205.0 -179.6 -147.9 -189.6 -171.6

The data of the steroids calculated by Chromsword® are summarized in Table 8.1. The

prediction accuracy of this program is limited [14], however, it remains a useful tool that

safes time in method development.

8.3 Results and discussion 8.3.1 Preliminary experiments Two columns were chosen for CEC experiments – the Unimicro C18 stationary phase and

the Hypersil C8 MOS stationary phase – and mobile phase containing 70% acetonitrile and

30% Tris buffer (5 mmol.L-1 as total concentration). The initial chromatograms are shown

in Figure 8.3. The Unimicro C18 stationary phase offered similar selectivity as the Hypersil

C8 MOS stationary phase, however, the latter one exhibits higher silanol activity providing

a higher and more stable electroosmotic flow (EOF). Therefore, the Hypersil C8 MOS was

further used in the method development.


-5

10

25

40

55

0 2 4 6 8 10 12 14 16 18 20Time (min)

Resp

onse

(m

AU)

t0

O

H H

OH

OH

H H

O

H H

OH

H H

O

H H

OH

H H1

2

3

1

2

3

Conditions:70 / 30 ACN / 5 mM Tris

column: Unimicro C18

length: 40 cm

Voltage: 25 kV

O

H H

OH

H

O

OO

H

H

OH

-5

10

25

40

55

0 2 4 6 8 10Time (min)

Resp

onse

(m

AU

)

t0

O

H H

OH

OH

H H

O

H H

OH

H H

O

H H

OH

H H

1

2

31

2

3

Conditions:70 / 30 ACN / 5 mM Tris

column: Hypersil C8 MOS

length: 25 cm

Voltage: 20 kV

O

H H

OH

H

O

OO

H

H

OHB)

A)

Figure 8.3 CEC chromatograms of tested steroids on A) the Unimicro C18 stationary phase, B) the

Hypersil C8 MOS stationary phase using 70% ACN/5mM Tris buffer (V/V) as the mobile

phase.

8.3.2 Selectivities

The selectivity (as separation factor α) and overall separation performance on the

Hypersil C8 MOS stationary phase was compared for three organic modifiers –

acetonitrile, methanol and tetrahydrofuran. The starting mobile phase used contained

30% Tris buffer and 70% ACN or isoeluotropic amount of other organic modifiers

resulting in the same solvent strength {86% MeOH (V/V) or 50% THF (V/V) in the

mobile phase} as shown in Figures 8.4 A and B.

- 180 - Chapter 8

-2

13

28

0 2 4 6 8 10 12 14 16Time (min)

Resp

onse

(m

AU)

t0

O

H H

OH

OH

H H

O

H H

OH

H H

O

H H

OH

H H

1

2

3

1

2

3

Conditions:86 / 14 MeOH / 5 mM Tris


length: 25 cm

Voltage: 20 kV

O

H H

OH

H

O

OO

H

H

OH

-3

12

27

42

57

0 5 10 15 20 25 30Time (min)

Resp

onse

(m

AU)

t0

O

H H

OH

OH

H H

O

H H

OH

H H

O

H H

OH

H H

1

2

1

2

Conditions:50 / 50 THF / 5 mM Tris


length: 25 cm

Voltage: 20 kV

O

H H

OH

H

O

OO

H

H

OH

A)

B)

Figure 8.4 CEC chromatograms of tested steroids on the Hypersil C8 MOS stationary phase using

A) 86% MeOH/5 mM Tris buffer (V/V), B) 50% THF/5 mM Tris buffer (V/V) as the

mobile phase

Table 8.2 summarizes retention factors and selectivity data. It is well known that

acetonitrile containing mobile phases exhibit higher ratio of ε/η where ε is the

permittivity and η the solvent viscosity. This ratio influences electroosmotic flow in the

electrodriven system according to the following equation:

ηεζEvEOF = (8.1)

where ε is the permittivity, ζ the zeta potential, E the electric field strength and η the

solvent viscosity. It is obvious that the separation of the steroids is faster in mobile phases

containing acetonitrile.


Table 8.2 Retention factors of tested steroids in different mobile phases with the same

solvent strength on the Hypersil C8 MOS stationary phase.

Steroid

#1 #2 #3 #4 #5

Organic modifier

t0

min

k α k α k α k α k α

ACN (70%)

3.8 0.11 - 0.43 3.91 0.52 1.21 0.72 1.38 1.50 2.08

MeOH (86%)

8.7 0.05 - 0.20 4.00 0.24 1.20 0.35 1.46 0.60 1.71

THF (50%)

11.9 0.30 - 0.60 2.00 0.67 1.12 0.84 1.25 1.63 1.94

k=retention factor; α=separation factor

The separation time for ACN-containing phase is about 10 min, however, under the same

conditions the separation time is about 15 min for the MeOH-containing mobile phase and

32 min for THF-containing mobile phase (Table 8.2). The separation performance was

further compared for 3 concentrations of acetonitrile in the mobile phase (60, 70 and 80%,

Table 8.3). The electroosmotic flow decreases with decreasing portion of ACN in the

mobile phase. Even greater effects are seen with respect of the retention factor (k). The

retention factor of steroid #5 increases from 0.73 at 80% ACN to 3.09 at 60% ACN. The

retention and selectivity data are summarized in Table 8.3.

- 182 - Chapter 8 Table 8.3 Retention factors (k) of tested steroids in different ACN-compositions of

mobile phase on the Hypersil C8 MOS stationary phase.

Steroid

#1 #2 #3 #4 #5

ACN composition

t0

min

k α k α k α k α k α

60% 4.6 0.20 - 0.74 3.70 0.89 1.20 1.38 1.55 3.09 2.24

70% 3.8 0.11 - 0.43 3.91 0.52 1.21 0.72 1.38 1.50 2.08

80% 3.5 0.06 - 0.25 4.17 0.31 1.24 0.38 1.23 0.73 1.92

It was found that the separation time at 80% ACN is 7 min with complete baseline

resolution. Therefore, acetonitrile at a concentration of 80% was used in the subsequent

validation of the method.

8.3.3 Effect of acetonitrile composition Retention factors are influenced by the change of acetonitrile compositions according to

the linear or fairly linear dependence of retention factor on the concentration of organic

modifier in the mobile phase:

ϕSkk += 0lnln (8.2)

where k is the retention factor, k0 the retention factor extrapolated to 100% aqueous

mobile phase, S the slope of the dependence function and φ the volume fraction of organic

modifier in the mobile phase. From Figure 8.5A it can be calculated that a change of 1% in

the ACN-composition causes a change in retention of factor of up to 5%. The effect on the

plate number is smaller, overall change in concentration from 76% (V/V) to 80% (V/V) as

depicted in Figure 8.5B doesn’t cause significant changes in the plate number of the peaks,

a further increase causes a light drop in the plate number. In a similar way as retention

factors, also the resolution is affected by a change of acetonitrile composition. This

according to the following equation:


+

−=2

2

11

4 kkNRS α

α (8.3)

where N is the plate number, α the separation factor and k2 the retention factor of the

peak. The change in resolution is thus dependent on the plate number and retention factors

of the compounds; as an example the difference in resolution is up to +/- 6% for steroid

#5 for a change of +/- 1% of ACN composition (Fig. 8.5C).

- 184 - Chapter 8

ACN composition

-25-20-15-10-505

1015202530

74 76 78 80 82 84 86

ACN (%)

∆ k

(%)

#1#2#3#4#5

ACN composition

0

5000

10000

15000

20000

25000

30000

35000

40000

74 76 78 80 82 84 86

ACN (%)

Plat

e nu

mbe

r (pe

r col

umn)

#1#2#3#4#5

ACN composition

-40

-30

-20

-10

0

10

20

30

40

74 76 78 80 82 84 86

ACN (%)

∆ R

(%)

#1#2#3#4#5

A)

B)

C)

Figure 8.5 Influence of the ACN composition on A) retention factor (k), B) plate number (N) and C)

resolution of the tested steroids (R, calculated as resolution of steroid #n and steroid #n-1

or thiourea for steroid #1).


8.3.4 Effect of pH of Tris buffer It is well known that pH has a major influence on the electroosmotic flow in CEC. Here,

we focus on the retention parameters of the separation process. Fig. 8.6A shows influence

of pH on the change of retention factor. It is interesting to see that there’s small influence

of pH on retention factor of neutral compounds (up to 0.2% for steroid #5 by change of

pH by 0.1 unit). It has been discussed in previous articles [e.g. 15] that EOF originating on

the stationary phase may influence ordering of hydrocarbonaceous chains and therefore the

separation process. The influence of pH on plate number is minimal (Fig. 8.6B). However,

the resolution change by change of pH by 0.1 unit is remarkable, e.g. for steroid #3 is up to

4% (Fig. 8.6C); this due to added effects of plate number and retention factor differences

(see Eq. 8.3, and conclusions further).

8.3.5 Effect of Tris concentration A higher concentration of Tris buffer in the mobile phase decreases retention. It seems

that adsorption of Tris buffer as a low hydrophilic buffer decreases slightly

hydrophobicity of the stationary phase. The overall change in retention factor is up to 2%

for the change by 2 mM (Fig. 8.7A), for change by 1 mM from the starting conditions

(5 mM Tris) the change is from +0.4% to –1% for steroid #1. Other steroids show

similar behaviour. Also, with increasing concentration of Tris buffer the plate numbers

increase (Fig. 8.7B). This is in agreement with findings of other authors [e.g. 16].

Summarizing as an example, with an increase of Tris buffer concentration by 1 mM we

obtained a gain in resolution up to 13% for steroid #5 (Fig. 8.7C).

- 186 - Chapter 8

pH

-0,6

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

1,2

7,4 7,6 7,8 8 8,2 8,4 8,6

pH

∆ k

(%)

#1#2#3#4#5

pH

0

5000

10000

15000

20000

25000

30000

35000

40000

7,4 7,6 7,8 8 8,2 8,4 8,6

pH

Plat

e nu

mbe

r (pe

r col

umn)

#1#2#3#4#5

pH

-25

-20

-15

-10

-5

0

5

7,4 7,6 7,8 8 8,2 8,4 8,6

pH

∆R (%

)

#1#2#3#4#5

B)

A)

C)

Figure 8.6 Influence of pH of Tris buffer on A) retention factor (k), B) plate number (N) and C)




TRIS concentration

-2

-1,5

-1

-0,5

0

0,5

1

1,5

2

0 2 4 6 8

c(TRIS)/mM

∆ k

(%)

#1#2#3#4#5

TRIS concentration

05000

1000015000200002500030000350004000045000

0 2 4 6 8

c(TRIS)/mM

Palte

num

ber #1

#2#3#4#5

TRIS concentration

-20

-15

-10

-5

0

5

10

15

20

25

0 2 4 6 8

c(TRIS)/mM

∆ R

(%)

#1#2#3#4#5

B)

A)

C)

Figure 8.7 Influence of concentration of Tris buffer composition on A) retention factor (k), B) plate

number (N) and C) resolution of the tested steroids (R, calculated as resolution of steroid

#n and steroid #n-1 or thiourea for steroid #1).

- 188 - Chapter 8 8.3.6 Effect of temperature

Temperature has large effect on both, the chromatographic process and capillary

electrophoretic parameters. The influence on retention factor can be seen in Figure 8.8A.

Generally, the Agilent 3DCE instrument is able to maintain temperature of the cassette

within a range of +/- 0.1°C. Under these conditions, such fluctuation can cause changes up

to 0.15% in retention factor or a change in resolution of about 1%. However, the influence

on the plate number exhibits no specific trend within the temperature range tested (Fig.

8.8B).

8.3.7 Effect of injected plug width

Overall increase and/or decrease of retention factor was found up to 2%, caused by change

in injection time from 10 to 20 s or from 10 to 2.5 s, respectively (Fig. 8.9A). Steroid #1

with low retention exhibits changes up to 1.1% upon a change of 1 s from the starting

condition (injection time=10 s). There is no obvious influence of the plug length on plate

number as well as resolution of the steroids (Fig. 8.9B and C) within the investigated

injection range. Pyell et al. [17] investigated the influence of sample plug width on band

broadening in CEC in detail. According to the theory, the optimum injection time can be

calculated as 3.2 s for our conditions. Higher injection times negatively influence peak

broadening. However, we found no obvious influence of the plug length on the plate

number in the studied injection range. Resolution was affected up to +/- 2% when

changing the injection time by 1 s.


Temperature

-8

-6

-4

-2

0

2

4

6

8

10 15 20 25 30

Temperature (ºC)

? k

(%

)

#1#2#3#4#5

Temperature

05000

1000015000200002500030000350004000045000

10 15 20 25 30

Temperature (ºC)

Plat

e nu

mbe

r (pe

r col

umn)

#1#2#3#4#5

A)

B)

C) Temperature

-15

-10

-5

0

5

10

15

10 15 20 25 30

Temperature (ºC)

? R

(%)

#1#2#3#4#5

Figure 8.8

Influence of temperature on A) retention factor (k), B) plate number (N) and C) resolution

of the tested steroids (R, calculated as resolution of steroid #n and steroid #n-1 or thiourea

for steroid #1).

- 190 - Chapter 8

Plug length

-8

-6

-4

-2

0

2

4

6

0 5 10 15 20 25

Injection (s)

∆ k

(%)

#1#2#3#4#5

Plug length

0

5000

10000

15000

20000

25000

30000

35000

40000

0 5 10 15 20 25

Injection (s)

Plat

e nu

mbe

r (pe

r col

umn)

#1#2#3#4#5

B)

A)

C) Plug length

-20

-15

-10

-5

0

5

10

0 5 10 15 20 25

Injection (s)

∆ R

(%)

#1#2#3#4#5

Figure 8.9

Influence of injected plug legth on A) retention factor (k), B) plate number (N) and C)




R (#5)

9.

77

1.93

3

9.51

1.75

9.05

0.78

10.1

2

2.22

9.84

2.26

9.46

3

11.3

4

4.18

9.87

7.4

N

(#5)

30

512

3.58

2925

3

3.54

2534

3

3

3201

9

4.89

3074

4

5.28

2802

1

6.85

4090

3

10.4

9

3097

1

15.7

3

k

(#5)

0.

731

0.26

0.73

2

0.16

0.73

9

0.34

0.73

4

0.21

0.73

7

0.17

0.73

3

0.19

0.71

7

0.62

0.73

2

0.97

t (#

5)

6.00

2

0.47

6.05

3

0.33

6.12

2

0.46

6.05

8

0.24

6.08

2

0.44

6.06

8

0.23

5.89

1

1.01

6.04

1.23

R

(#4)

2.

47

1.96

2.42

3.26

2.3 0 2.57

2.63

2.49

2.28

2.4

2.78

2.92

3.14

2.51

7.93

N

(#4)

30

033

4.55

2847

5

4.27

2614

4

0.85

3261

2

3.65

3064

4

5.32

2851

5

5.09

4335

7

4.88

3139

7

17.9

9

k

(#4)

0.

381

0.24

0.38

3

0.15

0.38

6

0.43

0.38

3

0.16

0.38

6

0.25

0.38

3

0.28

0.37

4

0.62

0.38

2

1.01

t (#

4)

4.79

1

0.44

4.83

2

0.33

4.87

3

0.43

4.83

3

0.3

4.85

4

0.46

4.84

1

0.27

4.71

2

0.95

4.82

1

1.1

R

(#3)

2.

1 4

2.41

2.1

3.17

2 0 2.2

2.14

2.14

2.41

2.04

4.13

2.46

2.1

2.15

6.98

N

(#3)

33

930

4.39

3294

9

6.52

2971

3

3.49

3653

8

4.98

3410

8

4.51

3070

9

4.24

4816

9

5.51

3515

9

17.5

4

k

(#3)

0.

307

0.24

0.30

8

0.19

0.31

1

0.43

0.30

9

0.16

0.31

1

0.3

0.30

8

0.25

0.30

1

0.59

0.30

8

1.09

t (#

3)

4.53

2

0.43

4.57

3

0.33

4.61

6

0.42

4.57

3

0.31

4.59

1

0.45

4.57

9

0.26

4.46

4

0.92

4.56

1

1.08

R

(#2)

7.

8 4

1.37

7.53

2.26

7.19

1.53

8.04

3 7.8

3.2

7.53

1.78

9.1

1.87

7.86

14

7.77

N

(#2)

35

016

1.9

3329

0

4.81

2984

2

4.99

3650

1

7.11

3428

5

5.82

3241

3

6.15

4926

3

3.58

3580

1

17.5

9

k

(#2)

0.

249

0.23

0.25

0

0.19

0.25

0.37

0.25

0

0.16

0.25

2

0.33

0.24

9

0.29

0.24

5

0.51

0.24

9

0.93

t (#

2)

4.33

0

0.42

4.36

7

0.32

4.40

6

0.39

4.36

6

0.31

4.38

3

0.44

4.37

4

0.26

4.27

1

0.89

4.35

7

1.01

R

(#1)

2.

2

3.03

2.12

4.33

2.02

2.09

2.3

2.05

2.24

4.80

2.19

4.54

2.64

4.44

2.24

8.73

N

(#1)

39

082

4.6

3487

1

5.11

3102

8

4.1

4099

5

5.53

3902

4

9.35

3595

0

3.11

5525

6

5.58

3945

8

19.5

4

k

(#1)

0.

060

0.46

0.06

1

0.61

0.06

1

0.34

0.06

1

0.61

0.06

2

0.93

0.06

0

0.84

0.06

0

0.56

0.06

1

0.93

t (#

1)

3.67

7

0.39

3.70

7

0.32

3.73

4

0.33

3.70

5

0.34

3.71

9

0.43

3.71

3

0.25

3.63

8

0.84

3.69

9

0.86

t 0

3.46

8

0.4

3.49

47

0.33

3.52

1

0.32

3.49

3

0.31

3.50

2

0.42

3.50

1

0.24

3.43

1

0.83

3.48

7

0.84

mea

n

RSD

(%)

mea

n

RSD

(%)

mea

n

RSD

(%)

mea

n

RSD

(%)

mea

n

RSD

(%)

mea

n

RSD

(%)

mea

n

RSD

(%)

mea

n

RSD

(%)

Day

1 2 3 4 5 6 7 Day

-to-d

ay

Tabl

e 8.

4 R

epea

tabi

lity

studi

es o

f tes

ted

stero

ids.*

* t 0=

dead

tim

e (m

in),

t=re

tent

ion

time

(min

), k

=re

tent

ion

fact

or, N

=pl

ate

num

ber,

R=re

solu

tion,

RSD

=re

lativ

e sta

ndar

d de

viat

ion

- 192 - Chapter 8 8.3.8 Repeatability Table 8.4 summarizes repeatability studies performed over 7 days. This table contains

results on retention times, retention factors, plate numbers and resolution of all the steroids

for 10 consecutive injections for 7 days. Relative standards deviation for retention factors of

steroids #2 - #5 within 10 consecutive injections (run-to-run) is below 0.5%, when

including the low-retained steroid #1 the relative standard deviation (RSD) is below 1%.

The day-to-day repeatability of the retention factor is below 1.1%. Run-to-run relative

standard deviation for plate numbers for the tested steroids is below 10%, day-to-day is

below 19.5%.

8.3.9 Detection limits

In the frame of the work it is important to test a system for detection limits [18]. The

production process always involves non-reacted products and by-products. Either of these

may be transferred in small or very small amounts to the final products or intermediates. In

order to detect these impurities, detection limits must be determined. Table 8.5 summarizes

detection limits – lowest detectable concentration (Cm) and lowest detectable amount (w0)

for all 5 steroids. Detection limits were determined by injecting sequentially diluted standard

mixtures of steroids of 1 mg/mL. Steroids of concentration as low as 10 µg/mL were easily

detectable (at least 5 times higher than detector noise level). Detection limits as low as

39.8 pg (steroid #1) to 119 pg (steroid #3) were found.


Table 8.5 Detection limits of the tested steroids.*

Detection limits Steroid

Cm (mg/mL) w0 (g)

#1 2.13×10-3 3.98×10-11

#2 1.96×10-3 4.64×10-11

#3 4.79×10-3 1.19×10-10

#4 1.91×10-3 5.39×10-11

#5 3.79×10-3 1.12×10-10

* Cm=lowest detectable concentration, w0=lowest detectable amount

8.4 Conclusions A fast and simple capillary electrochromatography method of separation of five

structurally related steroids was developed. The method was developed for a Hypersil C8

MOS stationary phase using acetonitrile/aqueous Tris buffer mobile phase. Ruggedness

of the method in terms of acetonitrile composition, Tris buffer concentration, Tris buffer

pH, temperature and sample plug width was extensively studied. The separation protocol

is as follows: stationary phase, Hypersil C8 MOS 3 µm; mobile phase, acetonitrile:Tris

buffer (5mM as total concentration, pH 8.0) 80/20 (V/V), injection: 10 s at 5 kV,

temperature 20°C. It was also concluded that the experimental conditions must be strictly

kept constant. This due to the dependency of retention and resolution on the

experimental parameters. The retention factor repeatability was found below 1% RSD

under these conditions. Finally, detection limits as low as 39.8 pg of steroid were found.

Acknowledgement The authors gratefully acknowledge Dr. J.R.M. Vervoort from AKZO Nobel, NV

Organon, The Netherlands for providing the steroid samples.

- 194 - Chapter 8 REFERENCES 1. International Union of Pure and Applied Chemistry, Nomenclature of Organic

Chemistry, Sections A, B, C, D, E, F and H, 1979 Edition, Pergamon Press, Oxford,

1979.


(2000) 2541.

3. D.A Stead, R.G. Reid, R.B. Taylor, J. Chromatogr. A, 798 (1998) 259.

4. R.M. Seifar, W. Th. Kok, J.C. Kraak, H. Poppe, Chromatographia, 46 (1997) 131.

5. R.M. Seifar, S. Heemstra, W. Th. Kok, J.C. Kraak, H. Poppe, J. Microcolumn Sep., 10

(1998) 41.

6. A.H. Que, A. Palm, A.G. Baker, M.V. Novotny, J. Chromatrogr. A, 887 (2000) 379.

7. M. R. Taylor, P. Teale, S.A. Westwood, D. Perrett, Anal. Chem., 69 (1997) 2554.

8. J. Wang, D.E. Schaufelberger, N.A. Guzman, J. Chromatogr. Sci., 36 (1998) 155.




11. M. R. Euerby, C.M. Johnson, K.D. Bartle, LC-GC Eur., 11 (1998) 39.


(1996) 403.


14. T. Baczek, R. Kaliszan, H.A. Claessens, M.A. van Straten, LC-GC Int., 14 (2001) 304.


16. A. Banholczer, U. Pyell, J. Chromatogr. A, 869 (2000) 363.


18. C.F. Poole, S.K. Poole, Chromatography Today, Elsevier Science, Amsterdam, 1991.

- 195 -

CHAPTER 9 9 SEPARATION OF BASIC CENTRAL NERVOUS SYSTEM

DRUGS BY CAPILLARY ELECTROCHROMATOGRAPHY

Summary

This article describes the sample preparation and separation of three strongly basic central

nervous system (CNS) drugs by capillary electrochromatography. The separation was

developed on Hypersil C8 MOS and Hypersil Phenyl stationary phases using acetonitrile

(ACN) as organic modifier together with mobile phase additives, ammonia,

ethylenediamine or 1,3-diaminopropane. A successful, fast separation after a simple

derivatization procedure of the compound was achieved on the Hypersil Phenyl

stationary phase using ethylenediamine as mobile phase additive. The general approach to

method development for complex mixtures is discussed.


9.1 Introduction Development of drugs for central nervous system (CNS) disorders is one of the most

progressive areas of the pharmaceutical industry [1]. At present, Alzheimer’s disease drugs

account for 32% of total peak sales of CNS products. Furthermore, migraine, neurogenic

and opioid-resistant pain drugs have chance to contribute significantly to the expansion of

the CNS area. In current practice, high-performance liquid chromatography (HPLC) is

the preferred technique in analyses of pharmaceuticals. Capillary electrochromatography

(CEC), however, offers new separation potentials such as high peak capacity and different

selectivity. Since the introduction of capillary electrochromatography (CEC) in practice at

- 196 - Chapter 9 the end on 1980’s, only few applications were shown and particularly at the beginning

CEC analyses were limited to the analyses of polyaromatic hydrocarbons (PAHs). In spite

of the recent development in CEC, analyses of certain groups of solutes such as basic

compounds remain a problematic task. In this particular case, two contradictory items

arise:

i. the charged stationary phase (e.g from ionized silanol groups) is responsible for

electroosmotic flow (EOF) under CEC conditions

ii. as known from high-performance liquid chromatography, interactions between

the charged surface of the stationary phase (e.g. from ionized silanol groups)

cause deterioration in chromatographic performance of these solutes e.g. tailing

peaks [2]

At presence, three groups of basic analytes found their application field in the

electrochromatography of pharmaceuticals (incl. opiates) and other nitrogen-containing

compounds (e.g. primary amines). Gillott et al. [3-4] showed a successful separation of

basic pharmaceuticals such as procainamide and nortriptyline on a C18 stationary phase

and bare silica using competing bases, triethylamine (TEA) and triethanolamine (TEOA),

as mobile phase additives. The effects of mobile phase pH, concentration of competing

bases and acetonitrile (ACN) concentration were examined. Cahours et al. [5] used a

phenyl stationary phase for the analysis of benzodiazepines using ACN/aqueous Tris

buffered mobile phase. Lurie et al. [6] presented the separation of strongly, moderately

and weakly basic, acid and neutral solutes in a single run using a CEC Hypersil C18

stationary phase with a ACN/aqueous phosphate mobile phase containing n-hexylamine

as mobile phase additive (competing base) at low pH. Dittmann et al. [7] separated basic

pharmaceuticals as procaine, ambroxol and antipyrine on a variety of stationary phases,

Spherisorb ODS I, CEC Hypersil C18 and CEC Hypersil C8 using a “Lurie” [5] mobile

phase. The influence of several parameters including the concentration of the competing

base was thoroughly studied. Smith et al. [8] analyzed amitriptyline, imiprazine on

Spherisorb ODS I and Spherisorb Silica stationary phases using non-aqueous mobile

phase (acetonitrile/methanol) containing Tris buffer. These authors showed the

successful performance of the SymmetryShield RP-8 stationary phase using an

ACN/aqueous Tris buffer mobile phase; the possible mechanism of interactions of

solutes with the stationary phase is given. Strickmann et al. [9] applied LiChrospher 100

Separation of Basic CNS drugs … - 197 -

RP-18 stationary phase with ACN/aqueous ammonium formate mobile phase for analysis

of the drug etodolac using ESI-MS detection. Special purpose stationary phases such as

cholesteryl silica bonded phase (Jinno et al. [10]), continuous beds based on acrylamide

polymers (Enlund et al. [11]) or polysaccharide-type stationary phases (Meyring et al. [12])

used for the separation of basic pharmaceuticals- benzodiazepines [10], nortriptyline,

amitriptyline [11] or thalidomide and its hydroxylated metabolites [12]. Wei et al. [13]

separated cocaine, codeine and thebaine on bare silica stationary phase using

ACN/aqueous Tris mobile phase. Basic opiates (morphine, codeine, diacetylmorphine)

were separated by Lim et al. [14] on a non-porous ODS stationary phase with

ACN/aquesous Tris {tris(hydroxymethyl)aminomethane} mobile phase with sodium

dodecylsulfate (SDS) as mobile phase additive. Similarly, Wu at al [15] separated basics,

acids and neutrals on a monolithic capillary column using modifiers such as SDS and

CTAB (cetyl-trimethylammonium bromide). Koide et al. [16] analyzed twelve primary

amines on monolithic chiral stationary phases using ACN/aqueous boric acid mobile

phase with TEA as mobile phase additive. Lopez-Avila et al. [17] studied the

determination of selected heterocyclic compounds containing nitrogen, oxygen or sulfur

on a C18 bonded silica phase using ACN/aqueous sodium tetraborate mobile phase.

Klampfl and co-workers [18-20] extensively studied separations of aromatic nitrogen-

containing compounds (anilines, pyridines, pyrimidines) on phases that exhibit strong

cation-exchange (SCX) or strong anion-exchange (SAX) characteristics.

In this paper, we describe a method for the separation of three basic CNS drugs on two

stationary phases in CEC using basic mobile phase additives as silanol shielding agents

together with a simple active compound derivatization procedure. This paper discusses

the method development for the analysis of basic solutes under capillary

electrochromatography conditions.

9.2 Experimental 9.2.1 Chemicals Acetonitrile (ACN) with HPLC grade purity was obtained from Biosolve (Valkenswaard,

the Netherlands); ammonia solution (25%) of analytical purity was supplied by Merck

(Darmstadt, Germany), ethylenediamine, redistilled, 99.5+% and 1,3-diaminopropane

99+% were purchased from Aldrich (Steinheim, Germany), 1,5-diaminopentane (≥97%),

- 198 - Chapter 9 boron trifluoride in methanol (~10% in MeOH) was obtained from Fluka (Buchs,

Switzerland). CNS drug samples were kindly supplied by Organon (Oss, the Netherlands).

The structures of the drugs are depicted in Figure 9.1; drug #1 is the main component,

chemically 1-[6-chloro-5-(trifluoromethyl)pyridin-2-yl]piperazine; drugs #2 and #3 are

related substances, chemically 2-chloro-6-piperazine-1-ylnicotinic acid and 1-(6-

chloropyridine-2-yl)piperazine, respectively. Eluents: acetonitrile/water (70/30, V/V) with

added ammonia, ethylenediamine or 1,3-diaminopropane without pH additional

adjustment. The concentration of ammonia, ethylenediamine and 1,3-diaminopropane

varied throughout the procedure. All mobile phases were filtered through a filter (pore size

= 0.45 µm) and degassed by ultrasonication before use.

NNHN

CF3

Cl

NNHN

Cl

COOH

NNHN

Cl

. HCl

#1 = Org 12962

.CF3 COOH

#2 = Org 14229

. HCl

#3 = Org 14191 Figure 9.1 Structures of the tested CNS drugs.

9.2.2 Columns Two columns were investigated, Hypersil MOS C8 and Hypersil Phenyl (Thermo Hypersil-

Keystone, Runcorn, UK); column dimensions: 100 µm I.D. × 33.5 cm total length (25 cm

effective); particle size 3 µm; detection wavelength: UV, 254 nm. These columns differ

substantially in their hydrophobicity and silanol activity [24]. Both columns were

conditioned by the mobile phase using a syringe pump and then in the CEC mode by

applying 10 bar pressure on both sides of the column and increasing the voltage from 0-

25 kV in 5 kV steps per 10 min.






electrokinetically (5 kV for 2-15 s). For each run a voltage of 20 kV (Hypersil C8 MOS) or

15 kV (Hypersil Phenyl) was applied with 10-bar pressure on both ends of the capillary.

The detection wavelength was 254 nm. High voltage was applied as a 6-s time ramp to

avoid column stress. The column cassette temperature was maintained at 20ºC.

9.3 Results and discussion 9.3.1 Preliminary experiments Based on the relationship of CEC and HPLC (high-performance liquid chromatography),

Chromsword® HPLC optimization software (Merck GmbH, Darmstadt, Germany) was

used for preliminary experiments to establish starting CEC conditions. The elution order of

the CNS drugs on the Hypersil C18 stationary phase using 70% acetonitrile/30% water

(V/V) was found to be as follows: #2, #3, #1. It is obvious that all drug compounds will

show partial charge under most CEC conditions and particularly drug #2 with zwitterionic

characteristic will thus deviate from the HPLC software generated chromatogram. For

charged compounds, retention in CEC may be expressed as follows [21]:

totid

eff

eof

r

LLV

t

kt µ+

+=1

)1( (9.1)

where tr is the retention time, k the retention factor, teof the dead time, µeff the electroosmotic

mobility, V the column volume, Lid the injector-to-detector length of the column and Ltot

the total length of the column. Figure 9.2 shows the CEC separation of all three drugs on

the Hypersil Phenyl stationary phase using 70% ACN and 30% aqueous ethylenediamine

adjusted with acetic acid to pH 7.0. In this case, the bases #1 and #3 elute before the t0

marker, proving that their electrophoretic mobilities contribute substantially to their

retention. Drug #2, which should have zero or negative charge under these conditions,

elutes after t0. Therefore it migrates after t0 due its lower mobility.

- 200 - Chapter 9

-5

10

25

40

55

70

85

100

0 5 10 15 20 25Time (min)

Resp

onse

(m

AU

)

t0

Conditions:70 / 30 ACN / en(250 µl/100 ml mob. phase, pH 7.0)column: Hypersil Phenyllength: 25 cmVoltage: 20 kV

NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.188

NNHN

Cl

COOH

Figure 9.2 Chromatograms of all CNS drugs. Conditions: columns - Hypersil Phenyl, mobile phase -

acetonitrile/aqueous ethylenediamine (250 µL per 100 mL of mobile phase, pH 7.0

adjusted with acetic acid) 70/30 (V/V), voltage-20 kV.

Based on previous literature [22] and the experience with capillary electrochromatography,

the drugs will be analyzed in mobile phases adjusted to higher pH using competing strong

bases such as ammonia (pKa 9.25), ethylenediamine (pKa1 10.71, pKa2 7.56), 1,3-

diaminopropane (pKa1 10.94, pKa2 9.03) and 1,5-diaminopentane. The zwitterionic drug

(#2) will be derivatized prior analysis using a simple procedure in order to convert it to a

simple basic compound. This procedure [23] transforms the carboxylic group into esters,

known from gas-chromatographic sample pre-treatment. The esterification is performed by

dissolving the sample in methanol containing boron trifluoride as a catalyst at 60°C as

depicted in Figure 9.3.


NNHN

Cl

COOH NNHN

Cl

COOCH3

CH3OH

BF3

Figure 9.3 Derivatization procedure of drug #2.

It is important to emphasize that a slower reaction occurs derivatizing the nitrogen of

piperazine. Using this derivatization procedure we can again simulate chromatograms using

the Chromsword® HPLC optimization software to revealing in the elution order: #2-

methyl ester, #3, #1.

9.3.2 Column and mobile phase modifier choice The selection the optimal column and mobile phase modifier will be demonstrated on the

separation of drugs #1 and #3.

9.3.3 Hypersil C8 MOS Ammonia was chosen for the comparative study of basic competing agents (mobile phase

modifiers) as the simplest one. Three experiments (Fig. 9.4 A, B and C) show the

improvement in the peak shape of both basic analytes upon increasing concentration of

ammonia in the mobile phase.

- 202 - Chapter 9

-10

5

20

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU

)

t0

Conditions:70 / 30 ACN / ammonia(1 ml NH3/100 ml mob. phase)column: Hypersil C8 MOSlength: 25 cmVoltage: 20 kV

NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.091

-15

0

15

30

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU

) t0


NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.126

-20

-5

10

25

40

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU

)

t0


NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.255

A)

B)

C)

Figure 9.4 Chromatograms of all CNS drugs. Conditions: column - Hypersil C8 MOS, mobile phase

–70 % acetonitrile/30% (V/V) aqueous ammonia, A) 1 mL per 100 mL of mobile phase,

B) 2 mL per 100 mL of mobile phase, C) 5 mL per 100 mL of mobile phase, voltage-20

kV.

Three concentrations were chosen, 70% aqueous acetonitrile with 1 mL 25% ammonia

per 100 mL mobile phase (Fig. 9.4A), with 2 mL 25% ammonia (Fig. 9.4B) and 5 mL of


25% ammonia (Fig. 9.4C). The symmetry of the first peak (drug #3) improved from

0.091 (experiment A) to 0.126 (experiment B) and 0.255 (experiment C). Symmetry of the

peak is calculated using following equation:

43

21

mmmmsymmetry

++= (9.2)

where m1, m2, m3, m4 are the moments of the peak calculated by the integrator (peak

inflections and time slices on the baseline are calculated) [25]. Approximately, peak

asymmetry is the reciprocal value of above-mentioned symmetry. Further, much stronger

and more hydrophobic bases, ethylenediamine and 1, 3-diaminopropane, were compared

with respect to performance. Ethylenediamine in concentrations of 100 µL per 100 mL of

the mobile phase (70% aqueous ACN) (Fig. 9.5A), 250 µL per 100 mL of the mobile phase

(Fig. 9.5B) and 500 µL per 100 mL of the mobile phase (Fig. 5C) improved the peak

symmetry substantially. The peak symmetry of drug #3 improved from a value of 0.167

(experiment A) up to 0.285 (experiment C). Addition of 250 µL of the more hydrophobic

1,3-diaminopentane to the mobile phase further improved performance; a symmetry value

of 0.286 for drug #3 was reached. It appeared that improvement to more symmetric peaks

(close to the value=1) might be difficult even with more hydrophobic amines. The reason

may be found in accessibility of free silanols of the stationary phase by the shielding agents

(mobile phase additives).

- 204 - Chapter 9

-5

10

25

40

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU

)

t0

Conditions:70 / 30 ACN / en(100 µl/100 ml mob. phase)column: Hypersil C8 MOSlength: 25 cmVoltage: 20 kV

NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.167

-5

10

25

40

55

70

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU

)

t0


NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.220

-5

10

25

40

55

70

85

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU

)

t0


NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.285

A)

B)

C)

Figure 9.5 Chromatograms of all CNS drugs. Conditions: column - Hypersil C8 MOS, mobile phase –

70 % acetonitrile/30% (V/V) aqueous ethylenediamine, A) 100 µL per 100 mL of mobile

phase, B) 250 µL per 100 mL of mobile phase, C) 500 µL per 100 mL of mobile phase,

voltage-20 kV.


9.3.4 Hypersil Phenyl

Ammonia as the mobile phase additive in concentrations of 100 µL per 100 mL 70%

aqueous acetonitrile was used for studying the initial performance (Fig. 9.6)

-5

10

25

40

0 5 10 15 20 25 30Time (min)

Resp

onse

(m

AU

)

t0

Conditions:70 / 30 ACN / ammonia(100 µl/100 ml mob. phase)column: Hypersil Phenyllength: 25 cmVoltage: 15 kV

NNHN

CF3

ClNNH

NCl

Figure 9.6 Chromatogram of CNS drugs #1 and #2. Conditions: columns - Hypersil Phenyl, mobile

phase - acetonitrile/aqueous ammonia (100 µL per 100 mL of mobile phase) 70/30 (V/V),

voltage - 15 kV.

Non separated peaks with poor peak shape were achieved. As expected, addition of 100 mL

of ethylenediamine improved peak shape, however, the analytes were not baseline separated

(Fig. 9.7A). This goal was finally achieved by addition of 250 µL of ethylenediamine to the

mobile phase. A peak symmetry of drug #3 of 0.580 and a resolution of 2.1 was reached

(Fig. 9.7B).

- 206 - Chapter 9

-5

10

25

40

55

70

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU)

t0

Conditions:70 / 30 ACN / ethylenediamine(100 µl/100 ml mob. phase)column: Hypersil Phenyllength: 25 cmVoltage: 15 kV

NNHN

CF3

Cl

NNHN

Cl

-2

13

28

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU)

t0


NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.580

A)

B)

Figure 9.7 Chromatograms of CNS drugs #1 and #3. Conditions: column - Hypersil Phenyl, mobile

phase –70 % acetonitrile/30% aqueous ethylenediamine (no pH adjustment), A) 100 µL

per 100 mL of mobile phase, B) 250 µL per 100 mL of mobile phase, voltage - 15 kV.

When comparing the results with more hydrophobic amines as mobile phase additives, 1,3-

diaminopropane and 1,5-diaminopentane, more symmetric peaks of basic solutes were

achieved (symmetry of 0.692 and 0.760, respectively), however, the resolution decreased

(Fig. 9.8 A and B). Obviously, there is a link between amount of remaining silanols,

hydrophobicity of the stationary phase (due to adsorbed hydrophobic amines) and

resolution of basic analytes.


-5

10

25

40

55

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU)

t0

Conditions:70 / 30 ACN / propylenediamine(250 µl/100 ml mob. phase)column: Hypersil Phenyllength: 25 cmVoltage: 15 kV

NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.760

-5

10

25

40

55

70

85

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU)

t0

Conditions:70 / 30 ACN / 1,5-diaminopentane(250 µl/100 ml mob. phase)column: Hypersil Phenyllength: 25 cmVoltage: 15 kV

NNHN

CF3

Cl

NNHN

Cl

Symmetry: 0.692

A)

B)

Figure 9.8 Chromatograms of CNS drugs #1 and #3. Conditions: column - Hypersil Phenyl, mobile

phase –70 % acetonitrile/30% aqueous A) 1,3-diaminopropane, 250 µL per 100 mL of

mobile phase, B) 1,5-diaminopentane, 250 µL per 100 mL of mobile phase, voltage - 15 kV.

Successful separation at 70% aqueous ACN containing 250 µL ethylenediamine in the

mobile phase was used to monitor the derivatization of drug #2 (Fig. 9.9), a small later

eluting peak from the extensive derivatization is supposed to be corresponding to the N-

methyl analog. For further purpose, the derivatization has been optimized to minimize

further methylation of the steroid #2. All three substances were subsequently analyzed

under the same conditions resulting in a satisfactory baseline separation with symmetrical

peaks of the drugs as demonstrated in Fig. 9.10.

- 208 - Chapter 9

-2

13

28

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU

)

t0Conditions:70 / 30 ACN / ethylenediamine(250 µl/100 ml mob. phase)column: Hypersil Phenyllength: 25 cmVoltage: 15 kV

NNHN

Cl

COOH NNHN

Cl

COOCH3

CH3OH

BF3

Before derivatization

After derivatization

Figure 9.9 Monitoring of derivatization of CNS drug #2. Conditions same as in Fig. 9.7B

-2

3

8

13

0 2 4 6 8 10 12 14Time (min)

Resp

onse

(m

AU

)

t0


NNHN

CF3

Cl

NNHN

Cl

NNHN

Cl

COOCH3

Figure 9.10

Chromatograms of CNS drugs #1 and #3 and derivatized #2. Conditions: column -

Hypersil Phenyl, mobile phase –70 % acetonitrile/30% aqueous ethylenediamine, 250 µL

per 100 mL of mobile phase, voltage - 15 kV.

9.3.5 Repeatability, Influence of variables and Detection limits

Table 9.1 lists values of a repeatability test for all three basic drugs over three days. Run-to-

run repeatability in terms of retention times varied within 0.26% for ten consecutive


injections, for the retention factor the results were within 1.26%. Day-to-day repeatability

results were within 3.03% for retention times and 5.22% for retention factors.

Table 9.1 Repeatability of retention and retention factors.

Compound

#1 #2

(methyl ester)

#3

Day t0

(min)

tR

(min)

k tR

(min)

k tR

(min)

k

mean 6.08 9.00 0.48 7.69 0.26 8.23 0.35 1

RSD (%) 0.26 0.24 1.26 0.16 1.03 0.16 0.89

mean 6.16 9.29 0.51 7.89 0.28 8.48 0.38 2

RSD (%) 0.20 0.26 0.74 0.17 0.70 0.22 0.60

mean 6.27 9.56 0.53 8.07 0.29 8.72 0.39 3

RSD (%) 0.15 0.20 0.27 0.23 0.99 0.19 0.45

mean 6.17 9.29 0.50 7.88 0.28 8.48 0.37 Day-to-day RSD (%) 1.51 3.03 4.60 2.40 4.23 2.91 5.22

The effects of variables as temperature, acetonitrile composition and ethylenediamine

concentration in the mobile phase are summarized in Table 9.2. Evidently, the

ethylenediamine concentration has the greatest effect; retention factors vary up to 2%

(estimated) for drug #3 upon a change of 10 µL of ethylenediamine in the mobile phase.

Acetonitrile concentration ranging from 69-71% change the retention factor of drug #1 of

+/- 5%. For the same compound the change in temperature of 0.1°C (precision of the

instrument) causes a change in the retention factor of up to 0.2%.

- 210 - Chapter 9 Table 9.2 Influence of variables, temperature, ACN concentration and concentration of

ethylenediamine in the mobile phase.

Compound Variable

#1 #2-methyl ester #3

Temperature (ºC) ∆kvalue-mean (%)

15 4.62 2.68 4.21

17 2.19 1.48 2.03

18 1.49 0.93 1.31

19 0.47 0.30 0.42

20 (mean) 0 0 0

21 -1.91 -0.70 -1.57

22 -2.05 -0.42 -1.50

23 -2.85 -0.73 -2.17

25 -4.75 -1.85 -3.95

ACN composition (%)

66 23.17 17.02 17.07

68 9.78 7.53 7.20

70 (mean) 0 0 0

72 -9.67 -5.46 -6.82

74 -19.11 -8.34 -13.95

Ethylenediamine conc. (µµµµL)

100 33.04 27.72 45.20

200 6.75 10.12 8.35

250 (mean) 0 0 0

300 -5.79 -6.80 -6.65

Detection limits, lowest detectable concentration (Cm) and lowest detectable amount (w0) [2]

for all three drugs are summarized in Table 9.3. The detection limits for the zwitterionic


drug are approximate with a 50% derivatization yield. Detection limits as low as 670 pg

(drug #3) and 730 pg (drug #1) were found.

Table 9.3 Detection limits for tested CNS drugs.

Detection limits Compound

Cm (mg/mL) w0 (g)

#1 8.1×10-3 7.3×10-10

#2 4×10-2 5×10-9

#3 8.6×10-3 6.7×10-10

* Cm=lowest detectable concentration, w0=lowest detectable amount

9.4 Conclusions A capillary electrochromatography method was developed for the separation of three basic

drugs on reversed-phase stationary phases after a simple derivatization procedure of one of

the compounds. The best performance with respect of separation time and resolution was

achieved on the Hypersil Phenyl stationary phase using 70% acetonitrile/30% water with

250 µL of ethylenediamine as a mobile phase additive (per 100 mL of mobile phase). The

experimental conditions should be strictly kept constant due to the dependency of

chromatographic performance on the experimental parameters. Run-to-run retention factor

repeatability was below 1.26% RSD, day-to-day repeatability within 5.22% RSD. The

influence of acetonitrile composition, temperature and ethylenediamine concentration was

extensively studied. Detection limits as low as 670 pg of drug were found.

Acknowledgement The authors gratefully acknowledge Dr. J.R.M. Vervoort from AKZO Nobel, NV

Organon, The Netherlands for providing the basic analyte samples.

REFERENCES 1. D. Hughes, Pharm. Sci. Technol. Today, 1 (1998) 186.

2. C.F. Poole, S.K. Poole, Chromatography Today, Elsevier Science, 1991, p. 366.

- 212 - Chapter 9 3. N.C. Gillott, M.R. Euerby, C.M. Johnson, D.A. Barrett, P.N. Shaw, Anal. Commun.,

35 (1998) 217.

4. N.C. Gillott, M.R. Euerby, C.M. Johnson, D.A. Barrett, P.N. Shaw, Chromatographia,

51 (2000) 167.

5. X. Cahours, Ph. Morin, M. Dreux, J. Chromatogr. A, 845 (1999) 203.


7. M.M. Dittmann, K. Masuch, G.P. Rozing, J. Chromatogr. A, 887 (2000) 209.

8. N.W. Smith, J. Chromatogr. A, 887 (2000) 233.

9. D.B. Strickmann, G. Blaschke, J. Chromatogr. B, 748 (2000) 213.

10. K. Jinno, H. Sawada, A.P. Catabay, H. Watanabe, N.B.H. Sabli, J.J. Pesek, M.T.

Matyska, J. Chromatogr. A, 887 (2000) 479.

11. A.M. Enlund, C. Ericson, S. Hjertén, D. Westerlund, Electrophoresis, 22 (2001) 511.

12. M. Meyring, B. Chankvetadze, G. Blaschke, J. Chromatogr. A, 876 (2000) 157.

13. W. Wei, G.A. Luo, G.Y. Hua, C. Yan, J. Chromatogr. A, 817 (1998) 65.

14. J.-T. Lim, R.N. Zare, C.B. Bailey, D.J. Rakestraw, C. Yan, Electrophoresis, 21 (2000)

737.

15. R. Wu, H. Zou, M. Ye, Z. Lei, J. Ni, Electrophoresis, 22 (2001) 544.

16. T. Koide, K. Ueno, J. Chromatogr. A, 909 (2001) 305.

17. V. Lopez-Avila, J. Benedicto, C. Yan, J. High Resolut. Chromatogr., 20 (1997) 615.

18. C.W. Klampfl, E.F. Hilder, P.R. Haddad, J. Chromatogr. A, 888 (2000) 267.

19. C.W. Klampfl, P.R. Haddad, J. Chromatogr. A, 884 (2000) 277.

20. C.W. Klampfl, W. Buchberger, P.R. Haddad, J. Chromatogr. A, 911 (2001) 277.


22. M. J. Hilhorst, G. W. Somsen, G. J. de Jong, J. Chromatogr. A, 872 (2000) 315.

23. L.D. Metcalfe, A.A. Schmitz, Anal. Chem., 33 (1951) 363.

24. M.R. Euerby, C.M. Johnson, S.F. Smith, N. Gillot, D.A. Barrett, P.N. Shaw, J.

Microcolumn Sep. 11 (1999) 305.

25. Understanding Your ChemStation, Publication Number G2070-91114, Agilent

Technologies GmbH, Waldbronn, 2001.

- 213 -

SUMMARY

Analytical techniques such as electrophoresis and chromatography already have a long

tradition in science. The roots of electrophoresis originate from the end of the 18th century

with the formulation of Michael Faraday’s laws of electrophoresis. The first

chromatographic experiments were performed at the beginning of 20th century by Mikhail

Semenovich Tswett. Both separation techniques rapidly developed during the second half

of the 20th century and turned into high-performance separation techniques. Thus, capillary

electrochromatography (CEC) as a hybrid technique of high-performance liquid

chromatography and electrophoresis seemed to be a break-through in separation science.

The use of submicron particles to achieve very high efficiencies compared to high

performance liquid chromatography (HPLC) was only one of the most promising

properties of CEC. However, since the very beginning CEC has struggled with a number of

technical difficulties, which until now have hampered its widespread application in routine

analysis.

Chapters 1 and 2 provide a thorough overview of the present status of CEC, theoretical

considerations on electroosmosis, presently applied stationary and mobile phases and

developed applications. Furthermore, these chapters outline a number of related questions

and answers concerning the fundamentals in CEC and its relationship to HPLC and CE. It

can be concluded that CEC is a highly efficient separation technique. The use of

conventional stationary phases is limited due to their electroosmotic flow (EOF)

capabilities. New CEC stationary phases, however, have been developed to satisfy users’

demands. Submicron particles (< 1 µm) have been successfully used yielding high

efficiencies without loss of EOF. Until now, however, only very short columns have been

packed due to inherent packing problems. A variety of mobile phases can be applied in

CEC, though practically only acetonitrile and partly methanol as organic modifier in mobile

phases facilitate the generation of a substantial EOF. A number of applications has been

developed in CEC, though the requested robustness and ruggedness in comparison to

HPLC protocols have not been achieved, yet.

- 214 - Summary A number of fundamental aspects of CEC are discussed in Chapters 3 to 7. As described in

Chapter 3, column descriptors such as hydrophobicity and silanol activity may differ

substantially for the same stationary phase depending on whether the column is operated

under HPLC or CEC conditions. Moreover, these differences also are a function of the

nature and concentration of the organic modifier. It has been suggested that these

observations can be explained from different ligand orientations caused by the nature of the

eluent-driven modes, being either pressure or electro-driven. In addition, the observation

may also be due to the occurrence of electroosmotic whirlwind effects in porous packings

causing different phase ratios.

Furthermore, in Chapter 4 it is found that inorganic buffers have a great influence on the

chromatographic behaviour, while the impact of organic buffers is much lower. Generally,

cation/anion type and/or size have a specific impact on the retention behaviour of polar

compounds in CEC.

Chapter 5 describes the preparation of polymer monolithic CEC columns based on

poly(alkylmethacrylate). These columns were tested under HPLC and CEC conditions. All

columns showed reversed-phase characteristics of retention. Bulky molecules showed an

additional size-exclusion retention effect. Furthermore, using the basic compound

benzylamine as a test component, stationary phases showed higher silanol activity under

CEC conditions compared to HPLC.

Based on QSRR models, Chapter 6 compares the retention behaviour and the molecular

mechanism behind it under HPLC and CEC conditions in more detail. Detailed

comparative QSRR analysis revealed evidence for stronger non-specific dispersive

interactions between analytes and hydrocarbonaceous stationary phases in HPLC mode

compared to CEC conditions. In addition, higher silanol activity was observed under CEC

conditions compared to results obtained under HPLC conditions. Thermodynamic studies

in CEC in Chapter 7 reveal that structural changes within the stationary and the mobile

phase such as an increased organization of hydrocarbonaceous chains on the stationary

phase are imaginable under CEC conditions. Moreover, an increased mobile phase

organization in the CEC mode compared to the HPLC mode might be possible, too. This

was confirmed by the results of thermodynamic data obtained from non-polar and polar

solutes.

Summary - 215 -

Summarizing the retention mechanism in CEC, with respect to hydrophobicity, for the

same stationary and mobile phase combination in all cases this parameter is significantly

different under CEC compared to HPLC conditions. Concerning silanol activity, this

parameter also in most cases differs for a specific column and mobile phase combination

under CEC compared to HPLC conditions. These observations are confirmed by the

QSRR and thermodynamic results. In addition, different size exclusion effects for silica

based as well as organic monolithic columns have been observed between CEC and HPLC

conditions.

Chapters 8 and 9 focus on method development in CEC for two different groups of

compounds of different polarity viz. steroid hormones and basic central nervous system

drugs. Successful separations were developed and method protocols accessed. For steroids,

retention factor repeatabilities were found lower than 1% R.S.D. with detection limits as

low as 39.8 pg. For CNS drugs, run-to-run retention factor repeatabilities were found below

1.26% R.S.D with detection limits as low as 670 pg.

Finally, whether or not CEC will develop to a generally accepted routine analysis technique

strongly depends also on the development of a sufficient number of convinving

applications.

- 216 -

- 217 -

SAMENVATTING

Analytische technieken als elektroforese en chromatografie hebben reeds een lange traditie

in de wetenschap. De basis van de elektroforese vindt zijn oorsprong aan het eind van de

18e eeuw met de formulering van Michael Faraday's wetten van elektroforese. Daarnaast

werden de eerste chromatografische experimenten uitgevoerd aan het begin van de 20e

eeuw door Mikhail Semenowitch Tswett. Beide scheidingstechnieken ontwikkelden zich

snel gedurende de tweede helft van de 20e eeuw en hebben zich een plaats verworven als

hoogwaardige scheidingstechnieken. Daarom leek capillaire electrochromatografie (CEC)

als een hybride techniek van High Performance Liquid Chromatography (HPLC) en

Capillary Electrophoresis (CE) een doorbraak in de scheidingswetenschap te zijn. Het

gebruik van submicrondeeltjes om zeer hoge efficiencies te bereiken was in vergelijking tot

HPLC slechts één van de veelbelovende eigenschappen van CEC. Desondanks heeft de

toepassing van CEC vanaf het begin een aantal technische moeilijkheden opgeleverd, die

grootschalige toepassing in de routineanalyse belemmeren.

In hoofdstuk 1 en 2 wordt een grondig overzicht gegeven van de huidige status van CEC,

zoals theoretische overwegingen bij electro-osmose, huidige toepassing van stationaire en

mobiele fasen en ontwikkelde toepassingen. Verder worden in deze hoofdstukken een

aantal verwante vragen en antwoorden behandeld met betrekking tot de grondbeginselen

van CEC en de relatie hiervan met HPLC en CE. Geconcludeerd kan worden dat CEC een

hoog efficiënte scheidingstechniek is. Echter, het gebruik van conventionele stationaire

fasen in CEC is vanwege hun electro-osmotisch debiet (EOF) eigenschappen beperkt. De

nieuwe generatie stationaire fasen voor CEC is daarentegen ontwikkeld om meer tegemoet

te komen aan de huidige eisen. Submicrondeeltjes (< 1 µm) zijn met succes toegepast voor

het realiseren van hoge efficiencies zonder verlies van EOF. Wegens problemen, die

inherent zijn aan het pakkingproces, zijn tot nu toe echter slechts zeer korte kolommen

gepakt. In CEC kan een verscheidenheid aan mobiele fasen worden toegepast, hoewel tot

nu toe in de praktijk mobiele fasen, welke als organische modifier acetonitril en soms ook

methanol bevatten, een bevredigende EOF kunnen genereren. Er zijn inmiddels een aantal

- 218 - Samenvatting toepassingen in CEC ontwikkeld, echter de “robustness” en “ruggedness” van de

methoden blijven achter in vergelijking met HPLC methoden. Een aantal fundamentele

aspecten van CEC wordt besproken in hoofdstuk 3 tot 7.

Zoals beschreven in hoofdstuk 3 kunnen de karakteristieken van een kolom, zoals

hydrophobiciteit en silanolactiviteit, significant verschillen voor eenzelfde stationaire fase

afhankelijk of deze onder HPLC of CEC condities wordt gebruikt. Bovendien zijn deze

verschillen ook een functie van de aard en concentratie van de gebruikte organische

modifier. Mogelijke verschillende oriëntaties van de liganden aan de stationaire fase,

afhankelijk van het feit of de mobiele fase hydraulisch dan wel elektrisch voortbewogen

wordt, zijn aangevoerd als verklaring hiervoor. Deze waarnemingen kunnen ook verband

houden met het optreden van electro-osmotische wervelstromingen in poreuze stationaire

fasen, hetgeen eveneens de fasenverhouding kan beïnvloeden.

In hoofdstuk 4 wordt geconcludeerd dat anorganische buffers een grote invloed op het

chromatografische gedrag van CEC kolommen hebben, terwijl het effect van organische

buffers beduidend lager is. Over het algemeen hebben aard en/of afmetingen van de

kation/anion combinatie in de gebufferde mobiele fase een specifiek effect op het

retentiegedrag van polaire componenten in CEC.

In hoofdstuk 5 wordt de synthese van polymere monolithische CEC-kolommen op basis

van poly(alkylmethacrylaat) beschreven. Deze kolommen werden getest onder HPLC en

CEC condities. Alle kolommen vertoonden een significant reversed-phase retentiegedrag.

Grotere moleculen vertoonden bovendien ook een additioneel “size-exclusion” retentie-

effect. De silanolactiviteit gemeten voor de basische testcomponent benzylamine voor

stationaire fasen onder CEC condities is significant groter dan onder HPLC condities.

In hoofdstuk 6 worden retentiegedrag en het daarachter liggende moleculaire mechanisme

in CEC en HPLC met behulp van een QSRR-model beschreven. Gedetailleerde

vergelijkende QSRR-resultaten laten zien dat onder HPLC condities de niet-specifieke

interacties tussen componenten en apolaire stationaire fasen groter zijn onder HPLC dan

onder CEC condities. Bovendien blijkt de silanolactiviteit van de kolommen onder CEC

condities significant groter dan onder HPLC condities.

Thermodynamische studies, beschreven in hoofdstuk 7, laten zien dat er aanwijzigen voor

veranderingen in de ordering van stationaire en mobiele fasen onder CEC condities zijn. Er

zijn eveneens indicaties dat onder CEC condities een verhoogde ordering in de mobiele

Samenvatting - 219 -

fase kan optreden in vergelijking met HPLC. Resultaten van deze thermodynamische

studies bevestigen dat voor niet-polaire en polaire componenten.

De situatie met betrekking tot retentiemechanismen in CEC kan als volgt worden

samengevat: voor elke combinatie van stationaire en mobiele fasen is de hydrofobiciteit van

een kolom gemeten onder CEC-condities verschillend vergeleken met die bepaald onder

HPLC-condities. In vrijwel alle gevallen is dit ook van toepassing voor de silanolactiviteit

van deze combinaties van stationaire en mobiele fasen gemeten onder CEC- en HPLC-

condities. Deze waarnemingen worden bevestigd door de resultaten verkregen uit QSRR en

thermodynamische studies. Bovendien vertonen conventionele silica en ook organische

monolithische kolommen een verschillende sterische selectiviteit onder CEC- en HPLC-

condities.

De hoofdstukken 8 en 9 beschrijven de ontwikkeling van CEC methoden van de scheiding

van twee verschillende groepen componenten, steroïd hormonen en basische CNS

geneesmiddelen. Voor beide groepen componenten zijn methoden ontwikkeld en

protocollen vastgelegd. De herhaalbaarheid in de retentiefactor voor de steroïden bedraagt

minder dan 1% RSD en de detectielimiet 39,8 pg. Voor de CNS geneesmiddelen bedroeg

de herhaalbaarheid minder dan 1,26% RSD en de detectielimiet 670 pg.

De vraag tenslotte of CEC zich zal ontwikkelen tot een algemeen aanvaarde routinematige

analysetechniek zal mede sterk afhangen van het beschikbaar komen van een voldoende

aantal overtuigende toepassingen.

- 220 -

- 221 -

DANKWOORD

Graag breng ik een woord van dank uit aan prof.dr.ir. Carel A. Cramers voor de

gelegenheid die hij geboden heeft om mijn promotieonderzoek uit te voeren aan het

Laboratorium van Instrumentale Analyse aan de TU te Eindhoven. Mijn speciale dank

komt toe aan dr. Henk A. Claessens. Allereerst voor zijn stimulerende wijze van leiding

geven aan de afdeling waar ik mijn onderzoek verricht heb en de ondersteuning bij mijn

wetenschappelijk werk. Op de tweede plaats dank ik Henk graag voor zijn bereidwillige

ondersteuning aan het begin van de periode van mijn dagelijks leven in Nederland.

Tevens dank ik graag mijn collega's: Marion van Straten voor haar ondersteuning, zowel

binnen als buiten het laboratorium; Denise Tjallema-Dekker, speciaal voor haar

medewerking bij de druk van de artikelen en het proefschrift; en verder aan allen aan de

universiteiten van Tsjechië, Polen, Roemenië, Oekraïne en Frankrijk met wie ik een

vruchtbare en plezierige tijd heb doorgebracht.

Herewith I would like to thank Prof.dr.ir. Carel A. Cramers for giving me the opportunity

to perform post-graduate research at the Laboratory of Instrumental Analysis in

Eindhoven. My special thanks belong to dr. Henk A. Claessens for his leadership

throughout the years at TU Eindhoven and the good atmosphere I met there. I would like

to thank him particularly for kind help with my daily life in the Netherlands, with my

scientific work, with publication of scientific articles, for his good ideas that went further

than science only.

I would also like to thank my colleagues: Marion van Straten, for the great help in (and also

outside) the laboratory. I would like to thank Denise Tjallema-Dekker, especially for the

help with printing of the articles and the thesis and further all the colleagues from

universities from the Czech Republic, Poland, Rumania, Ukraine and France, I spent nice

moments with.

- 222 -

- 223 -

CURRICULUM VITAE

Jan Jiskra was born on January 1, 1973 in Turnov in the Czech Republic. In 1996 he

completed his Master degree study in analytical chemistry at the Charles University in

Prague, Czech Republic, with specialisation in high-performance liquid chromatography

and electrophoretic techniques. In the same year, he joined the group of Instrumental

Analysis of the Eindhoven University of Technology, The Netherlands.

At present, he works as a stability control officer at the pharmaceutical company Synthon in

Nijmegen, The Netherlands.

- 224 -

- 225 -

BIBLIOGRAPHY

1. V. Pacáková, K. Štulík, J. Jiskra, High-performance separations in the determination of

triazine herbicides and their residues, Journal of Chromatography A, 754 (1996) 17.

2. J. Jiskra, V. Pacáková, M. Tichá, K. Štulík, T. Barth, Use of capillary electrophoresis

and high-performance liquid chromatography for monitoring of glycosylation of the

peptides dalargin and desmopressin, Journal of Chromatography A, 761 (1997) 285.

3. M. Tichá, T. Trnka, V. Pacáková, J. Jiskra, L. Hauzerová, K. Ubik, T. Barth, Saccharide

derivatives of dalargin: physicochemical and biological evaluation of glycoconjugates,

Pept. 1996, Proc. Eur. Pept. Symp., 24th (1998), Meeting Date 1996, 827.

4. J. Jiskra, C. A. Cramers, M. Byelik and H. A. Claessens, Chromatographic properties of

reversed-phase stationary phases under pressure- and electro-driven conditions, Journal

of Chromatography A, 862 (1999) 121.

5. J. Jiskra, T. Jiang, H.A. Claessens, C.A. Cramers, Chromatographic properties of

reversed phase stationary phases under pressure and electro driven conditions. Effect

of buffer composition, Journal of Microcolumn Separations, 12 (2000) 530.

6. T. Jiang, J. Jiskra, H.A. Claessens, C.A. Cramers, Preparation and characterization of

monolithic polymer columns for capillary electrochromatography, Journal of

Chromatography A, 931 (2001) 215.

7. J. Jiskra, H.A. Claessens, C.A. Cramers, Thermodynamic behaviour in capillary

electrochromatography, accepted for publication in Journal of Separation Science.

8. J. Jiskra, H.A. Claessens, C.A. Cramers, Method development for the separation of

steroids by capillary electrochromatography, accepted for publication in Journal of Separation

Science.

9. J. Jiskra, H.A. Claessens, C.A. Cramers, Separation of basic central nervous system

drugs by capillary electrochromatography, accepted for publication in Journal of Separation

Science.

10. J. Jiskra, H.A. Claessens, C.A. Cramers, R. Kaliszan, Quantitative structure retention

relationships in comparative studies of behaviour of stationary phases under high-

performance liquid chromatography and capillary electrochromatography conditions,

submitted for publication in Journal of Chromatography A.

- 226 - 11. J. Jiskra, H.A. Claessens, C.A. Cramers, Stationary and Mobile Phases in Capillary

Electrochromatography, submitted for publication in Journal of Separation Science.

Documents

Capillary electrochromatography : fundamentals and applications · CAPILLARY ELECTROCHROMATOGRAPHY; FUNDAMENTALS AND APPLICATIONS PROEFSCHRIFT ter verkrijging van de graad van doctor