Ion Chromatography [4th, Completely Revised and Enlarged Edition] 3527320520, 9783527320523

Citation preview

James S. Fritz and Douglas T. Gjerde Ion Chromatography

Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

Further Reading Miller, JM

Chromatography - Concepts and Contrasts 2e 2006 ISBN: 978-0-471-98059-9

McMahon, Gillian

Analytical Instrumentation A Guide to Laboratory, Portable and Miniaturized Instruments 2007 ISBN: 978-0-470-02795-0

Stuart, Barbara H.

Analytical Techniques in Materials Conservation 2007 ISBN: 978-0-470-01280-2

Dean, John R.

Bioavailability, Bioaccessibility and Mobility of Environmental Contaminants 2007 ISBN: 978-0-470-31967-3

Wang, Joseph

Analytical Electrochemistry 2007 ISBN: 978-0-470-23187-6

Hahn-Deinstrop, Elke

Applied Thin-Layer Chromatography 2006 ISBN: 978-3-527-31553-6

James S. Fritz and Douglas T. Gjerde

Ion Chromatography Fourth, Completely Revised and Enlarged Edition

The Authors Prof. James S. Fritz Ames Laboratory Iowa State University 332 Wilhelm Hall Ames, IA 50011 USA Dr. Douglas T. Gjerde PhyNexus, Inc. 3670 Charter Park Dr./ Suite A San José, CA 95136 USA

&

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate. Library of Congress Card No.: applied for British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de. © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – nor transmitted or translated into machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Typesetting Kühn & Weyh, Satz und Medien, Freiburg Printing Strauss GmbH, Mörlenbach Bookbinding Litges & Dopf, Heppenheim Printed in the Federal Republic of Germany. Printed on acid-free paper.

ISBN:

978-3-527-32052-3

V

Contents Preface

13

Acknowledgments

15

1 1.1 1.2 1.2.1 1.2.2 1.2.3 1.2.4 1.2.5 1.2.6 1.3 1.3.1 1.3.2 1.3.3 1.3.4 1.3.5 1.3.6

Introduction and Overview 1 Introduction 1 Historical Development 2 Early Ion-Exchange Separations 2 Cation Separations 3 Separation of Anions 6 On-line Detection 8 The Birth of Modern Ion Chromatography 8 Non-Suppressed-Ion Chromatography 10 Principles of Ion Chromatographic Separation and Detection 13 Requirements for Separation 13 Experimental Setup 13 Performing a Separation 14 Migration of Sample Ions 15 Detection 17 Basis for Separation 17

2 2.1 2.2 2.3 2.4 2.4.1 2.5 2.6 2.6.1 2.6.2 2.6.3 2.6.4

Instrumentation 21 Components of an Ion Chromatograph (IC) Instrument General Considerations 23 Eluent 24 Pump 26 Gradient Formation 29 Sample Injector 30 Columns 31 Column Hardware 31 Column Protection 32 Column Oven 33 Two-dimensional IC 33

Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

21

VI

Contents

2.7 2.8 2.9

Suppressor 33 Detector 34 Data Acquisition and Calculation of Results 35

3 3.1 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.3 3.3.1 3.4 3.4.1 3.4.2 3.4.3 3.4.4 3.4.4.1 3.5 3.5.1 3.5.2 3.5.3 3.6 3.6.1 3.6.2 3.6.3 3.6.4

Resins and Columns 37 Introduction 37 Polymeric Resins 38 Substrate and Cross Linking 38 Microporous Resins 39 Macroporous Resins 40 Chemical Functionalization 41 Resin Capacity 42 Resins and Columns for Ion Chromatography 43 Monolith Columns 43 Anion Exchangers 45 Porous Anion Exchangers 45 Effect of Functional Group on Selectivity 47 Effect of Spacer Arm Length 52 Latex Agglomerated Ion Exchangers 54 Effect of Latex Functional Group on Selectivity 56 Cation Exchangers 57 Sulfonated Resins 57 Weak-acid Cation Exchangers 61 Other Types 63 Other Resins 63 Chelating Ion-exchange Resins 63 Metal Oxides 64 Multi-purpose Resins 64 Ion-exchange Disks 65

4 4.1 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.4 4.5

Detectors 69 Introduction 69 Conductivity Detectors 70 Conductivity Definitions and Equations 73 Principles of Cell Operation 74 Conductance Measurement 75 Conductivity Hardware and Detector Operation 75 Contactless Conductivity Detection 76 Ultraviolet-Visible (UV–Vis) Detectors 77 UV–Vis measurement 77 Direct Spectrophotometric Measurement 78 Post-column Derivatization 81 UV–Vis Hardware and Detector Operation 82 Fluorescence Detector 83 Electrochemical Detectors 85

Contents

4.5.1 4.5.2 4.5.3 4.5.4 4.5.4.1 4.5.4.2 4.5.4.3 4.5.5 4.5.6 4.6 4.7 4.7.1 4.7.2 4.7.3 4.8

Potentiometric Detection 86 Conductometric Detectors 86 Amperometric/Coulometric Detection 87 Pulsed Electrochemical Detection (PED) 89 Pulsed Amperometric Detection (PAD) 91 Integrated Pulsed Amperometric Detection (IPAD) 93 IC–PED 94 Post-column Derivatization 95 Electrochemical Hardware and Detector Operation 95 Refractive Index Detector 97 Evaporative Light Scattering Detector (ELSD) 97 Nebulizer 98 Evaporation Chamber 99 Detection Cell 99 Other Detectors 100

5 5.1 5.2 5.2.1 5.3 5.3.1 5.3.2 5.3.3 5.3.4 5.3.4.1 5.3.4.2 5.3.4.3 5.4

Principles of Ion Chromatographic Separations 105 General Considerations 105 Chromatographic Terms 105 Retention Factor 106 Selectivity 109 Selectivity Coefficients 110 Other Ion-exchange Interactions 112 Selectivity of Sulfonated Cation-exchange Resin for Metal Cations 113 Factors Affecting Selectivity 120 Polymeric Matrix Effect 121 Resin Functional Group 122 Solvation Effects 123 Chromatographic Efficiency 124

6 6.1 6.1.1 6.1.2 6.2 6.2.1 6.2.2 6.2.3 6.2.3.1 6.2.3.2 6.3 6.3.1 6.3.2 6.3.3 6.3.3.1

Anion Chromatography 131 Scope and Conditions for Separation 131 Columns 132 Separation Conditions 135 Suppressed Anion Chromatography 138 Electrolytic Suppressors 140 Solid-Phase Reagents, 1990 [7] 141 Typical Separations 142 Isocratic and Gradient Elution 144 Influence of Organic Solvents 146 Nonsuppressed Ion Chromatography 147 Principles 147 Explanation of Chromatographic Peaks 150 Eluent 150 General Considerations 150

VII

VIII

Contents

6.3.3.2 6.3.3.3 6.3.3.4 6.3.4 6.3.5 6.3.6 6.3.6.1 6.4 6.5 6.5.1 6.5.2 6.5.3 6.6 6.7 6.8 6.9 6.9.1 6.9.2

Salts of Carboxylic Acids 151 Basic Eluents 152 Carboxylic Acid Eluents 153 System Peaks 154 Scope of Anion Separations 155 Sensitivity 155 Conductance of a Sample Peak 158 Coated Columns 160 Optical Absorbance Detection 163 Introduction 163 Direct UV Absorption 163 Indirect Absorbance 164 Detection 166 Pulsed Amperometric Detector (PAD) 168 Evaporative Light Scattering Detector (ELSD) 170 Inductively Coupled Plasma Methods (ICP) 172 Atomic Emission Spectroscopy (AES) 172 Inductively Coupled Plasma Mass Spectrometry (ICP-MS)

7 7.1 7.2 7.2.1 7.2.2 7.2.3 7.2.4 7.3 7.3.1 7.3.2 7.3.3 7.4 7.4.1 7.5 7.5.1 7.5.2 7.5.3 7.5.4 7.6 7.6.1 7.6.2

Cation Chromatography 175 Introduction 175 Columns 176 Historical Development 178 Phosphonate Columns 179 Macrocycle Columns 181 Surfactant Columns 182 Separations 184 Suppressed-Conductivity Detection 184 Non-Suppressed-Conductivity Detection 187 Spectrophotometric Detection 188 Effect of Organic Solvents 191 Separation of Alkali Metal Ions 193 Separation of Metal Ions with a Complexing Eluent Principles 195 Separations 196 Use of Sample Masking Reagents 197 Weak-Acid Ion Exchangers 198 Chelating Ion-Exchange Resins and Chelation Ion Chromatography 201 Fundamentals 201 Examples of Metal-Ion Separation 204

8 8.1 8.1.1

Ion-Exclusion Chromatography Principles 207 Equipment 209

207

195

173

Contents

8.1.2 8.1.3 8.2 8.2.1 8.2.2 8.3 8.4 8.5 8.5.1 8.6 8.7 8.7.1 8.8 8.8.1 8.8.2 8.8.3 8.8.4

Eluents 209 Detectors 210 Separation of Organic Acids 211 Effect of Alcohol Modifiers 214 Separation of Carboxylic Acids on Unfunctionalized Columns 216 Simultaneous Determination of Anions and Cations 217 Conclusions 220 Determination of Carbon Dioxide and Bicarbonate 222 Enhancement Column Reactions 222 Separation of Bases 223 Determination of Water 226 Determination of Very Low Concentrations of Water by HPLC 229 Separation of Saccharides and Alcohols 230 Introduction 230 Separation Mechanism and Control of Selectivity 230 Detection 235 Contamination 235

9 9.1 9.2 9.3

Ion Pair Chromatography 239 Principles 239 Typical Separations 242 Mechanism 246

10 10.1 10.2 10.3 10.4 10.5

Zwitterion Stationary Phases 251 Introduction 251 Simultaneous Separation of Anions and Cations 253 Separation of Anions 255 Separation of Cations 256 Mechanism 259

11 11.1 11.1.1 11.1.1.1 11.1.1.2 11.1.1.3 11.1.1.4 11.2 11.2.1 11.2.2 11.2.3 11.3 11.3.1 11.3.2 11.4

Capillary Electrophoresis 263 Introduction 263 Steps in Analysis 264 Capillary Pretreatment 264 Sample Introduction 264 Sample Run 265 Detection 265 Principles 265 Terms and Relationships 265 Zone Broadening 267 Sample Injection 267 Electroosmotic Flow (EOF) 268 Effect of EOF on Separations 270 Control of EOF 271 Separation of Ions 274

IX

X

Contents

11.4.1 11.4.1.1 11.4.2 11.4.3 11.4.4 11.5 11.5.1 11.5.2 11.5.3 11.5.4 11.5.4.1 11.6

Separation of Anions 274 Separation of Isotopes 276 Separation of Cations 278 Separations at Low pH 279 Capillary Electrophoresis at High Salt Concentration 280 Capillary Electrophoretic Ion Chromatography 283 Micellar Electrokinetic Chromatography (MEKC) 284 Partial Complexation 285 Effect of Ionic Polymers 287 Effect of Alkylammonium Salts 291 Separation Mechanism 294 Summary 294

12 12.1 12.1.1 12.1.2 12.2 12.3 12.3.1 12.3.2 12.3.3 12.4 12.4.1 12.4.2 12.4.3 12.5 12.5.1 12.5.2 12.5.3 12.5.4 12.6 12.6.1 12.6.2 12.6.2.1 12.6.2.2 12.6.2.3 12.7 12.7.1 12.7.2

DNA and RNA Chromatography 299 Introduction 299 Importance of DNA and RNA Chromatography 299 Organization of this Chapter 300 DNA and RNA Chemical Structure and Properties 301 DNA and RNA Chromatography 303 Development of DNA and RNA Chromatography 303 Column Properties 305 Ion-pairing Reagent and Eluent 306 Temperature Modes of DNA and RNA Chromatography 307 Nondenaturing Mode 307 Fully Denaturing Mode 308 Partially Denaturing Mode 309 Instrumentation 310 Effect of Metal Contamination 310 The Column Oven 313 UV and Fluorescence Detection 313 Fragment Collection 314 Applications of DNA Chromatography 314 DHPLC 314 Nucleic Acid Enzymology 315 Telomerase Assays 315 Polynucleotide Kinase Assays 316 Uracil DNA Glycosylase Assays 317 Applications of RNA Chromatography 317 Separation of Messenger RNA from Ribosomal RNA 318 Analysis of Transfer RNA 319

13 13.1 13.2 13.3

Sample Pretreatment 323 Dilute and Shoot or Pre-treat the Sample? 323 Particulate and Column-contaminating Matter 324 Preconcentration 325

Contents

13.3.1 13.3.2 13.4 13.4.1 13.4.2 13.4.3 13.4.3.1 13.4.3.2 13.4.4

Collection of Ions from Air 325 Preconcentration of Ions in Water 326 Sample Pretreatment 328 Anions in Acids 328 Neutralization of Strongly Acidic or Basic Samples Dialysis Sample Preparation 329 Passive Dialysis 330 Donnan (Active) Dialysis 330 Isolation of Organic Ions 333

14 14.1 14.1.1 14.1.2 14.1.3 14.2 14.3 14.3.1 14.3.2 14.3.3 14.4 14.4.1

Method Development and Validation 335 Choosing a Method 335 Define the Problem Carefully 335 Experimental Considerations 336 Example of Method Development 338 Some Applications of Ion Chromatography Statistical Evaluation of Data 341 Common Statistical Terms 341 Distribution of Means 344 Confidence Intervals 345 Validation of Analytical Procedures 347 Analytical Control 349

15 15.1 15.2 15.3 15.4 15.5 15.5.1 15.5.2 15.5.3 15.5.4 15.5.5 15.5.6 15.5.7 15.5.8 15.5.9

Chemical Speciation 353 Introduction 353 Detection 355 Chromatography 356 Valveless Injection IC 357 Speciation of some Elements Chromium 359 Iron 360 Arsenic 361 Tellurium 362 Selenium 363 Vanadium 364 Tin 364 Mercury 365 Other Metals 365 Index

369

359

340

328

XI

XIII

Preface The first edition of this book in 1982 described the emergence of ion-exchange chromatography as a practical and rapid method of separation and analysis. The initial and somewhat restrictive definition of the subject was broadened to 'ion chromatography' in order to describe the efficient chromatographic separation of anions or cations using any form of automatic detection. In the intervening years ion chromatography, or IC as it is sometimes called, has undergone impressive development and has attracted an ever-growing number of users throughout the world. Over the years, IC has proved to be extremely rugged, sensitive and reliable. It is the technology of choice to measure ions in such diverse situations as plating baths, drinking water, nuclear power plant water, foodstuffs and so on. Although IC is a well-established technique that is widely used throughout the world, it continues to enjoy vibrant growth and development. An annual International Ion Chromatographic Symposium (IICS) is held at various locations. This 4th edition of Ion Chromatography has been expanded and extensively revised. Some new features are listed below. . With some help from William La Course (see Acknowledgements), Chapter 4 on Detectors has been expanded and completely rewritten. New features include a strong section on pulsed electrochemical detection and an extensive table of dyes for tagging for the fluorescent detection of bio ions. . Chapter 5 (Principles of IC) has been expanded to include a discussion of factors that affect selectivity and chromatographic efficiency. . Although IC is the preferred method for inorganic ions and smaller organic ions, the need for chromatographic determination of larger organic and bio ions has been growing rapidly, as for example in the pharmaceutical industry. Here, Ion Pair Chromatography (IPC) is often the preferred technique. A new chapter (Chapter 9) on IPC is now included. . Chapter 10 on Zwitterion Stationary Phases describes a fascinating variation of IC. Pure water can often be used as the eluent to separate sample ions when a zwitterion stationary phase is employed. Ions elute as cation-anion pairs. Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

XIV

Preface .

.

. .

Ions are separated on a truly micro scale on an open capillary based on differences in their electrophoretic migration under the influence of an applied electrical potential (Chapter 11). Migration rates of the analyte ions are modified by interaction with organic ions of opposite charge in the electrolyte. Analyte peaks are unusually sharp because the zone broadening due to mass transfer between two distinct phases is eliminated. The reader is introduced to the fascinating world of the chromatography of large bio ions in Chapter 12. Structural features, common terms and chromatographic techniques are explained in a manner that the non-expert can easily understand. DNA molecules are extremely large ions relative to what can normally be separated by a chromatographic process. The new Chapter 13 is devoted to Sample Pretreatment. Method development, statistical evaluation of data, and validation and control of analytical procedures are covered in a new chapter (Chapter 14).

As in the previous editions, our goal has been to describe the materials, principles and methods of ion chromatography in a clear concise style. Whenever possible the consequences of varying experimental conditions are considered. Because commercial products are constantly changing, the equipment used in ion chromatography is described in a somewhat general manner. Our approach to the literature of IC is selective rather than comprehensive. Key references are given together with the title so that the nature of the reference will be apparent. Our goal is to explain fundamentals, but also to provide more detailed information in the form of figures and tables. We have tried to write a book that is enjoyable to read as well as one that is informative. Although it has been hard work, writing this book has also been a stimulating experience. We hope that we have been able to convey this enthusiasm to our readers. James S. Fritz, Ames, Iowa Douglas T. Gjerde, San Jose, California

October 2008

XV

Acknowledgments In the preparation of our 1st Edition of Ion Chromatography, published in 1982, we received valuable help from many sources, and we have continued to benefit from the skill, knowledge and willingness of many individuals over the years. Now with this 4th Edition of the book, we are again pleased to acknowledge some valuable contributions. We thank William R. LaCourse, who wrote Section 4.5 on Electrochemical Detection. Bill, who is Professor of Analytical Chemistry at the University of Maryland, Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, [email protected], is well known for his expertise in pulsed-amperometric detection, and we are grateful for his willingness to contribute to the section which deals with this subject. We also thank Peter Scott, currently a university student, for creating the original design used on the front cover of this book. Peter also helped with index and drew several of the figures that appear in the text. Jie Li, Shelly Li and Jeff Sun provided us with a clearer insight into ion analysis as used in the pharmaceutical industry, particularly with regard to the validation of analytical methods. Wenzi Hu, a world leader in ion chromatography using zwitterion stationary phases, kindly supplied reprints of his papers. We also received figures and other information from several companies who are in the business of providing tools for Ion Chromatography. They include, in no particular order, Dionex, Hamilton, Metrohm, Zellweger Analytics, Alltech (product line of Grace David) and Transgenomic. Our assistants played a major part in preparing and proofing the manuscript. In Ames, Marilyn Kniss did most of the typing, patiently checking the manuscript for consistency and clarity of expression. She also handled the correspondence. In San Jose, Tiffany Nguyen performed additional valuable and necessary work. We would like to extend our grateful thanks for the love and support of our respective families, for they, more than anything else, make life enjoyable, worthwhile and meaningful. James S. Fritz Ames, Iowa

Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

Douglas T. Gjerde San Jose, California

1

1 Introduction and Overview 1.1 Introduction

The name ‘ion chromatography’ applies to any modern method for chromatographic separation of ions. Normally, such separations are performed on a column packed with a solid ion-exchange material, but if we define chromatography broadly as a process in which separation occurs by differences in migration, capillary electrophoresis may also be included. Ion chromatography is considered to be an indispensable tool in a modern analytical laboratory. Complex mixtures of anions or cations can usually be separated and quantitative amounts of the individual ions measured in a relatively short time. Higher concentrations of sample ions may require some dilution of the sample before introduction into the ion-chromatographic instrument. ‘Dilute and shoot’ is the motto of many analytical chemists. However, ion chromatography is also a superb way to determine ions present at concentrations down to at least the low parts per billion (lg L–1) range. Although the majority of ion-chromatographic applications have been concerned with inorganic and relatively small organic ions, larger organic anions and cations may be determined as well. As in the three previous editions, our goal has been to describe the materials, principles and methods of ion chromatography in a clear, concise style. The following résumé is intended as a kind of road map to guide the reader through the contents of this book and to highlight some of the changes made in this fourth edition. In the first chapter we recount some of the historical milestones and briefly cover the most basic principles of ion chromatography, or IC as it is often called. The various components and hardware of IC instruments are described in Chapter 2, but it is not our intention to discuss specific commercial instruments. Chapter 3 has been updated to include advances in column technology and promising new columns, such as monolithic columns. Chapter 4 on detectors has been expanded to include new material on the contactless conductivity detector (CCD) and pulsed electrochemical detectors. Chapter 5 has been completely rewritten and now includes detailed sections on the factors that influence selectivity and efficiency. An updated and detailed treatIon Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

2

1 Introduction and Overview

ment of anion chromatography and cation chromatography is presented in the next two chapters. Selection of appropriate columns, detectors and eluents is discussed and numerous examples of typical separations are given. Ion exclusion chromatography, Chapter 8, continues to be a popular and useful method for the separation of hydrophilic sample components, such as carboxylic acids, amines and carbohydrates. Ion chromatography is generally defined as an analytical method in which anions or cations are separated by differences in the rate at which they pass through a column packed with either an anion- or cation-exchange particles. However, excellent separations of ionic analytes can also be obtained by ion chromatography on a standard reversed-phase HPLC column. With this technique, separation of cations or anions is achieved by using an aqueous-organic eluent together with an ion-pairing reagent. Ion-pair chromatography, covered in the new Chapter 9, is particularly advantageous for separation of organic ions. Chapter 10 on zwitterion stationary phases is another new addition to this book. Separations are generally performed on an HPLC column coated with a zwitterion surfactant that contains both positive and negative sites. In some cases cations and anions can be separated in a single run using pure water as the eluent! Resolution of the sample peaks obtainable in either liquid or ion chromatography is limited by two factors. One is due to mechanical pumping, which gives a curved flow profile. The second limiting factor stems from a slow rate of equilibration of solutes between the mobile and stationary phases. Use of very small ionexchange particles can reduce but never eliminate this source of peak broadening. Capillary electrophoresis (CE), which is covered in Chapter 11, addresses both of these issues. No eluent is required for CE; sample ions are separated by differences in their electrophoretic flow rates through open capillary containing no packing material. The electrophoretic mobilities can be modified by selective interactions with electrolyte ions of the opposite charge. Chapter 12 on Separation of DNA/RNA highlights a trend in IC toward a greater emphasis on analytical separations of bio ions. Sample pretreatment is discussed in Chapter 13. Chapter 14 on Method Development and Validation includes tips for selecting appropriate conditions for an IC analysis, and this final chapter covers chemical speciation.

1.2 Historical Development 1.2.1 Early Ion-Exchange Separations

Modern ion chromatography is built on the solid foundation created by extensive work in classical ion-exchange chromatography. Columns containing ionexchange resins have been used for many years to separate various cations and

1.2 Historical Development

anions from one another. Cations are separated on a cation-exchange resin column, and anions on an anion-exchange resin column. The most used types are as follows:

For example, Na+ and K+ can be separated on a cation-exchange resin (Catex) column with a dilute solution of a strong acid (H+) as the eluent (mobile phase). Introduction of the sample causes Na+ and K+ to be taken up in a band (zone) near the top of the column by ion exchange: Resin–SO–H+ + Na+, K+ > Resin SO–Na, K+ + H+ Continued elution of the column with an acidic eluent (H+) introduces competition of H+, Na+ and K+ for the exchange sites (–SO3–), causing the Na+ and K+ to move down the column. K+ is more strongly retained than Na+, and thus the Na+ zone moves down the column faster than the K+ zone. As originally conceived and carried out for many years, fractions of effluent were collected from the end of the column and analyzed for Na+ and K+. Then a plot was made of concentration vs fraction number to construct a chromatogram. All this took a long time and made ion-exchange chromatography slow and awkward to use. However, it soon was realized that under a given set of conditions all of the Na+ would be in a single fraction of several milliliters and all of the K+ could be recovered in a second fraction of a certain volume. Thus, under pre-determined conditions, each ion to be separated could be collected in a single fraction and then analyzed by spectroscopy, titration, etc., to determine the amount of each sample ion. The ability to collect a single fraction that contains all of the separated sample ion permits the use of step gradients. In this mode, conditions are adjusted so that an ‘all-or-nothing’ situation prevails. A sample ion either sticks onto the ionexchange column or it passes quickly through. Conditions are selected so that only one ion type will pass through the column while the other sample ions are strongly retained and form a tight band at the top of the column. Then the eluent is changed so that a second ion is rapidly eluted, while the others remain tightly stuck. Frequently, several gradient steps can be performed to elute different sample ions at each step. 1.2.2 Cation Separations

Early studies on the separation of metal cations included separations based on affinity differences and some specific separation with complexing eluents.

3

4

1 Introduction and Overview

Strelow and his coworkers have published extensive data relating to the selectivity of a sulfonated polystyrene cation exchanger for various cations in acidic solution [1]. The equilibria of cations in hydrochloric, nitric or sulfuric acid solutions with a cation exchanger involves complexation in some cases as well as competition between H+ and the metal cation for the exchange sites. For example, mercury(II) and cadminium(II) form chloride complexes even in dilute solutions of hydrochloric acid. Selectivity data in perchloric acid probably give the best indication of true ion-exchange selectivity, because the perchlorate anion has almost no complexing properties with metal cations. In general, cations with a 3+ charge are more strongly retained by a cation exchanger than cations with a 2+ charge, and ions with a 2+ charge are retained more strongly than those with a 1+ charge. Fritz and Karraker [2] were able to separate metal cations into groups according to their charge. Most divalent metal cations were eluted with a 0.1 M solution of ethylenediammonium perchlorate. Bismuth(III) and zirconium(IV) remained quantitatively o n the cation-exchange column. The use of the 2+ ethylenediammonium ion permitted a lower concentration to be used than would have been the case with an H+ eluent. Several inorganic acids exhibit a complexing effect for metal ions. The complexing acids include HF, HCI, HBr, HI, HSCN and H2SO4. The complexed metal ions are converted into neutral or anionic complexes and are rapidly eluted, while the other cations remain on the cation-exchange column. The data for hydrochloric acid [3] indicate selective complexing between metal cations and the chloride ion. For example, cadmium(II) has a distribution coefficient of 6.5 in 0.5 M hydrochloric acid, but a D = 101 in 0.5 M perchloric acid. Calcium(II), which shows no appreciable complexing, has a distribution coefficient of 147 in 0.5 M perchloric acid and 191 in 0.5 M hydrochloric acid. Strelow, Rethemeyer and Bothma [3, 4] also reported data for nitric and sulfuric acids that showed complexation in some cases. Mercury(II), bismuth(III), cadmium(II), zinc(II), and lead(II) form bromide complexes and elute in the order given in 0.1 to 0.6 M hydrobromic acid [5]. Most other metal cations remain on the column. Aluminum(III), molybdenum(VI), niobium(V), tin(IV), tantalum(V), uranium(VI), tungsten(VI) and zirconium(IV) form anion fluoride complexes and are quickly eluted from a hydrogen-form cation-exchange column with 0.1 to 0.2 M HF [6]. An eluent containing only 1% hydrogen peroxide in dilute aqueous solution will form stable anionic complexes with several metal ions. Fritz and Abbink [7] were able to separate vanadium(IV) or (V) from 25 metal cations, including the separation of vanadium(V) from 100 times as much iron(III). Strewlow [8] used hydrogen peroxide and sulfuric acid to separate titanium(IV) from more than 20 cations by cation exchange. Fritz and Dahmer [9] separated molybdenum(VI), tungsten(VI), niobium(V) and tantalum(V) as a group from other metals by adding dilute hydrogen peroxide to the sample solution and passing it through a cation-exchange column. Most of the eluents listed above are volatile upon heating and do not interfere with colorimetric, titrimetric or other methods for chemical determination of the metal ions separated. For the most part, group separations, rather than separation

1.2 Historical Development

of individual metal ions, are obtained, and only a short ion-exchange column is needed. Another valuable ‘all-or-nothing’ group separation uses an eluent consisting of 0.1 M tartaric acid and 0.01 M nitric acid [10]. Antimony(V), molybdenum(VI), tantalum(V), tin(IV) and tungsten(VI) form tartrate complexes in this acidic medium, but lead(II) and many other metal cations are not complexed and are retained by the cation exchanger. Samples containing tin(IV) must be added to the column in the tartrate solution. In a few cases an eluent containing an organic complexing reagent has been used successfully for the chromatographic separation of several metal ions. A notable example is the separation of individual rare earth ions with a solution of 2-hydroxylisobutyric acid as the eluent [11]. However, such separations necessitate careful equilibration of the column to maintain a desired pH. Sometimes gradient elution is used, and either the pH or the eluent concentration is changed. Metal cations usually form complexes with inorganic anions much more readily in organic solvents than in water. For example, the pink cobalt(II) cation requires around 4 M or 5 M aqueous hydrochloric acid to be converted to a blue cobalt(II) chloride anion. In a predominantly acetone solution, the intensely blue cobalt(II) is formed in very dilute hydrochloric acid. Thus, the scope of ionexchange group separations is increased greatly by carrying out separations in a mixture of water and an organic solvent. Fritz and Rettig [12] showed that zinc(II), iron(III), cobalt(II), copper(II) and manganese(II) can be separated from each other on a short cation-exchange column with eluents containing a fixed, low concentration of HCl, increasing the acetone concentration from 40% to 95% in steps. Later Strelow et al. [13] published extensive lists of metal-ion distribution coefficients in water/acetone/hydrochloric acid systems. Korkisch and coworkers have studied the effect of ethanol, acetic acid, ethylene glycol and many other solvents upon the ion-exchange behavior of metal ions in systems containing hydrochloric and other complexing acids [14]. The selectivity of low-capacity cation columns for monovalent ions can be adjusted by the addition of an organic modifier to the eluent. Using a nitric acid eluent of pH 2.5, for example, the elution order for monovalent ions is Li+, Na+, NH4+, K+. Simple amines elute in the order of the carbon number, after NH4+, with the result that (CH3)NH3+ (methylammonium) can co-elute with potassium. In most cases, this co-elution is of little significance, because potassium and methylammonium are not often in the same sample. However, where the analysis of either of these species in the presence of the other is desired, the selectivity can be modified by the addition of 40% methanol to the eluent [15]. The methanol causes the potassium to elute later but does not affect the elution time of methylammonium.

5

6

1 Introduction and Overview

1.2.3 Separation of Anions

Since most metal ions are cationic, it may sound strange to discuss their separation by anion-exchange chromatography. However, Kraus and Nelson, working at Oak Ridge National Laboratory in the USA, found that in aqueous hydrochloric acid solutions a number of metal ions form anionic complexes and are strongly taken up by anion-exchange resins. For most of the metal ions, a plot of the D value of several thousand is attained. An illustration of such plots for most of the metallic elements in the periodic table was published by Kraus and Nelson in 1956 [16]. Separations are generally achieved by adding the sample to an anion-exchange column in rather concentrated hydrochloric acid and eluting the nonsorbed metal ions with the same HCl concentration. Then the sorbed metal ions are eluted one at a time by stepwise reduction of the HCl strength of the eluent. Figure 1.1 illustrates one of the many practical separations published by Kraus and his coworkers [17].

Figure 1.1 Separation of metal ions on Dowex 1 × 10 anion exchange resin. (From Ref. [18] with permission.)

In a similar manner, elements that form anionic fluoride complexes can be separated from others and from each other on an anion exchanger by eluting with eluents containing HF plus HCl [18, 19]. Extensive studies of metal ion behavior on anion-exchange columns have also been carried out with eluents containing mixed H2SO4/HF [20, 21].

1.2 Historical Development

Anion-exchange distribution coefficients for most metallic elements in sulfuric acid solution have been measured [22, 23]. Uranium(VI), thorium(IV), molybdenum(VI) and a few other elements are retained selectively by anion-exchangers from solution in approximately 6 M nitric acid [24]. Operating in a predominantly organic solvent greatly improves the ability of metal ions to form complexes with halide and pseudo-halide anions. Such complexes generally are taken up strongly by an anion-exchange resin. Korkisch, Fritz, Strelow and others have published extensively on anion-exchange separations in partly nonaqueous solutions. Korkisch and Hazan [25] describe a method to separate metal ions that form chloride complexes from those that do not. The method uses an eluent consisting of 90–95% methanol in 0.6 M hydrochloric acid and requires only a short anion-exchange column. The metal ions studied are either retained as a sharp band or quickly pass through the column. Thus, we have an ‘all-or-nothing’ situation, and excellent group separations are obtained. Chromatographic separations of individual ions are also possible, and many have been published. An example is shown in Figure 1.2. Ion exchange in nonaqueous and mixed media has been reviewed [26].

Figure 1.2 Separation of nickel(II) and manganese(II) on a 6.0 × 2.2 cm column containing Dowex 1 × 8 resin, with partly nonaqueous eluents. (From Ref. [25] with permission.)

7

8

1 Introduction and Overview

Systems containing dimethylsulfoxide, methanol and hydrochloric acid have been studied for the anion-exchange behavior of 26 elements [27]. Numerous separations of two- to four-component mixtures of metal ions were carried out with quantitative results. 1.2.4 On-line Detection

At this stage in the development of ion-exchange chromatography, separation of cations or anions was still a slow and laborious process. It was becoming apparent that widespread use of ion-exchange chromatography as an analytical tool would require a system that gave fast separations with automatic recording of chromatograms. In 1971 an apparatus for ‘forced-flow chromatography’ was described in which the eluent was pushed through the analytical column by compressed nitrogen [28]. Detection of eluted ions was by UV-Vis spectrophotometry using a 30 mm × 2 mm flow cell. Iron(III) (10–90 lg) could be separated from most other metal ions and measured quantitatively in only 6 min. Forced-flow methods were soon developed for the chromatographic separation of a number of other metal ions [29–32]. The chromatograph was modified in 1974 so that a complexing reagent such as PAR or Arsenazo could be added to the column effluent via a mixing tee [33]. This made it possible to detect virtually any metal ion that could form a highly-colored complex. A recorded chromatographic separation of all 30 rare earths was obtained in 1974 [30, 34], with this apparatus. This separation took 100 min. Five years later, Elchuk and Cassidy in Canada were able to obtain a better separation of earths in only 27 min using a similar but improved system [35]. 1.2.5 The Birth of Modern Ion Chromatography

Liquid column chromatography went ‘high performance’ around 1970 and is now commonly referred to as HPLC. Major improvements in speed and efficiency were obtained by using columns of relatively small bore packed with small spherical particles of uniform diameter, using a pump to provide constant eluent flow, and using automatic detection of the separated sample components. However, application of this technology for the separation of ions lagged. It was mainly the lack of satisfactory detectors that held up the development of high-performance ion-exchange chromatography. This situation changed dramatically with the publication of a landmark paper in 1975 by Small et al., working at the Dow Chemical Co. [36]. As the authors put it: ‘It would be desirable to employ some form of conductimetric detection as a means of monitoring ionic species in a column effluent since conductivity is a universal property of ionic species in solution and since conductance shows a simple dependence on species concentration. However, the conductivity from the species of interest is generally “swamped out” by that from

1.2 Historical Development

the much more abundant eluting electrolyte. We have solved this detection problem by using a combination of resins which strips out or neutralizes the ions of the background electrolyte leaving only the species of interest as the major conducting species in the effluent. This has enabled us to successfully apply a conductivity cell and meter as the detector system.’ This new system, which was given the name ‘ion chromatography,’ enabled the analyst to quickly separate and measure quantitatively the cations or anions at low concentrations in fairly complex samples. A diagram of the system for cation analysis is shown in Figure 1.3. The upper column, called the ‘separator column’, was packed with polystyrene–2% DVB particles, surface-sulfonated to obtain an exchange capacity of approximately 0.02 mmol g–1. The lower ‘suppressor column’ was packed with anion-exchange resin of high exchange capacity in the hydroxide form.

Figure 1.3 System for cation analysis by conductimetric chromatography. (From Ref. [36] with permission.)

A practical method for the separation of anions was described in the same paper [36]. This endeavor necessitated the development of a new low-capacity anion-exchange resin. It had been known for some time that cation- and anionexchange resins have a marked tendency to clump together. Using this principle, a satisfactory anion-exchange material of low capacity was prepared by coating surface-sulfonated cation exchanger In the original scheme for the separation of anions a mixture of sodium hydroxide and sodium phenate was used in the eluent. The suppressor column was packed with a cation-exchange resin of high capacity. The suppressor column con-

9

10

1 Introduction and Overview

verted the eluent ions to water plus phenol, while the sample anions A– were converted to the highly conducting pair H+A–. An instrument called the ‘ion chromatograph’ was offered commercially by the newly organized Dionex Co. and became an immediate success. The new technology made it possible to separate and determine both cations and most anions, but the ability to determine anions at low ppm concentrations had the greater impact. Many cations could already be determined by various spectral methods and by reasonably good chromatographic methods, but prior to the advent of ion chromatography there was no general analytical method for anions, especially at very low concentrations. Once the scientific world became aware that anions in fairly complex mixtures could be easily separated and quantified, even at low ppm concentrations, the use of ion chromatography exploded. A powerful new analytical technique had again facilitated scientific endeavors that were previously impractical. 1.2.6 Non-Suppressed-Ion Chromatography

A major disadvantage of the original Ion Chromatograph was that it required the use of a large suppressor column that contributed to peak broadening and required frequent regeneration. The eluent for anion separations had to be a base, and anions of very weak acids could not be detected because their acidic form after suppression was too weakly conducting. In 1979 a synthetic method was described for producing anion-exchange resins of very low exchange capacity [37]. A porous polymeric resin was chloromethylated under mild conditions and then alkylated with trimethylamine to form ionic quaternary ammonium groups. The exchange capacity could be varied from 0.2 to 1.5 meq g–1 by controlling the time and temperature of the chloromethylation. This drastically lower exchange capacity permitted the use of much lower eluent concentrations than had previously been possible. In 1979, Gjerde, Fritz and Schmuckler described a simple system for anion chromatography with eluents containing anions of very low conductivity such as benzoate or phthalate [38]. Anions were separated on a column containing macroporous anion-exchange particles of very low exchange capacity: 0.07, 0.04 or 0.007 mmol g–1. The eluent was an aqueous solution of the sodium or potassium salt of an organic anion that had a significantly lower equivalent conductance than the anions to be separated. In this method some of the eluent anion is replaced by a sample anion of significantly higher conductance as the sample ion is eluted from the column and passes through the conductivity detector. Because of the low resin capacity, an eluent containing only ca. 10–4 M of an organic acid salt, such as benzoate or phthalate, could be used. The eluent conductance was so low that no suppressor column was needed, and the separated sample ions could be detected with a simple conductivity detector. Numerous anion separations were demonstrated, and in some instances detection limits below 1 ppm were obtained. This method was initially called ‘single-column ion chromatography’ and later ‘non-suppressed-ion chromatography.’ An additional paper on anion chromatog-

1.2 Historical Development

raphy [39] was published in 1980. A chromatographic separation of halide ions is shown in Figure 1.4.

Figure 1.4 Separation of 4.8 ppm of fluoride, 5.1 ppm of chloride and 26.0 ppm of bromide on XAD-1, 0.04 mequiv g–1; eluent is 0.65 mM potassium benzoate, pH 4.6. (From Ref. [39] with permission.)

The non-suppressed method for anion chromatography was followed quickly in 1980 by a similar method for cations [40]. This method also introduced the concept of indirect conductivity detection. A 1 × 10–3 M solution of nitric acid was used as the eluent in conjunction with a sulfonated cation-exchange column of low exchange capacity. In this method, a baseline of relatively high conductance is established when the column is equilibrated with the acidic eluent. After introduction of a sample mixture, such as Na+ and K+, and continued elution with the eluent, the sample cations are gradually resolved into zones in which some of the highly conducting H+ (equiv. conductance = 350 S cm2 equiv–1) is exchanged for a sample cation of much lower conductance. A sharp ‘peak’ of lower conductance is obtained for each sample cation. A mixture of Li+, Na+, NH4+, K+, Rb+ and Cs+ was separated in less than 10 min with a blend of 0.17 meq g–1 and unfunctionalized cation-exchange resins with 1.25 × 10–4 nitric acid as the eluent (Figure 1.5). Although separation of divalent cations with nitric acid was not practical, a fast separation of magnesium and calcium in tap water was obtained with a 1 × 10–3 M ethylene diammonium nitrate eluent.

11

12

1 Introduction and Overview

Figure 1.5 Separation of alkali metal cations on ammonium cation exchange column with a conductivity detector. Column: 350 × 2.0 mm. Packed with 0.059 mequiv g–1 cation exchange resin.

Development of the ion-chromatographic methods that use a conductivity detector was accompanied by a significant increase in chromatographic efficiency. The ion-exchange materials were of much smaller and more uniform size and the packing efficiency of the column was also improved. The changes that occurred were not unlike those in partition chromatography when it went from ‘liquid chromatography’ to ‘high-performance liquid chromatography’ (HPLC).

1.3 Principles of Ion Chromatographic Separation and Detection

1.3 Principles of Ion Chromatographic Separation and Detection 1.3.1 Requirements for Separation

The ion-exchange resins used in modern chromatography are of smaller particle size but have a lower capacity than the older resins. Columns packed with these newer resins have more theoretical plates than the older columns. For this reason, successful separations can now be obtained even when there are only small differences in retention times of the sample ions. The major requirements of systems used in modern ion chromatography can be summarized as follows: 1. An efficient cation- or anion-exchange column with as many theoretical plates as possible. 2. An eluent that provides reasonable differences between the retention times of the sample ions. 3. A resin-eluent system that attains equilibrium quickly so that kinetic peak broadening is eliminated or minimized. 4. Elution conditions such that retention times are in a convenient range – not too short or too long. 5. An eluent and resin that are compatible with a suitable detector. 1.3.2 Experimental Setup

Anions in analytical samples are separated on a column packed with an anionexchange resin. Similarly, cations are separated on a column containing a cationexchange resin. The principles for separating anions and cations are very similar. The separation of anions will be used here to illustrate the basic concepts. A typical column used in ion chromatography might have the dimensions 150 × 4.6 mm, although columns as short as 50 mm in length or as long as 250 mm are also used. The column is carefully packed with a spherical anionexchange resin of rather low exchange capacity and with a particle diameter of 5 or 10 lm. Most anion-exchange resins are functionalized with quaternary ammonium groups, which serve as the sites for the exchange of one anion for another. The basic setup for IC is as follow. A pump is used to force the eluent through the system at a fixed rate, such as 1 mL min–1. In the FILL mode a small sample loop (typically 10–100 lL) is filled with the analytical sample. At the same time, the eluent is pumped through the rest of the system, by-passing the sample loop. In the INJECT mode a valve is turned so that the eluent sweeps the sample from the filled sample loop into the column. A detector cell is connected to a strip-chart recorder or a data-acquisition device so that a chromatogram of the separation

13

14

1 Introduction and Overview

(signal vs time) can be plotted automatically. A conductivity or UV-visible detector is most often used in ion chromatography. The eluent used in anion chromatography contains an eluent anion, E–. Usually Na+ or H+ will be the cation associated with E– The eluent anion must be compatible with the detection method used. For conductivity, the anion E– should have either a significantly lower conductivity than the sample ions or be capable of being converted to a non-ionic form by a chemical suppression system. When spectrophotometric detection is employed, E– will often be chosen for its ability to absorb strongly in the UV or visible spectral region. The concentration of E– in the eluent will depend on the properties of the ion exchanger used and on the types of anions to be separated. Factors involved in the selection of a suitable eluent are discussed later. 1.3.3 Performing a Separation

To perform a separation, the eluent is first pumped through the system until equilibrium is reached, as evidenced by a stable baseline. The time needed for this may vary from a couple of minutes to an hour or longer, depending on the type of resin and the eluent used. During this step the ion-exchange sites will be converted to the E– form: Resin–N+R3 E–. There may also be a second equilibrium in which some E– is adsorbed on the resin surface but not at specific ion-exchange sites. In such cases the adsorption is likely to occur as an ion pair, such as E–Na+ or E–H+. An analytical sample can be injected into the system as soon as a steady baseline has been obtained. A sample containing anions A1– , A2–, A3–,...., A1– undergoes ion exchange with the exchange sites near the top of the chromatography column. A1– (etc.) + Res E– > Res–A1– (etc.) + E– If the total anion concentration of the sample happens to be exactly the same as that of the eluent being pumped through the system, the total ion concentration in the solution at the top of the column will remain unchanged. However, if the total ion concentration of the sample is greater than that of the eluent, the concentration of E– will increase in the solution at the top of the column because of the exchange reaction shown above. This zone of higher E– concentration will create a ripple effect as the zone passes down the column and through the detector. This will show up as the first peak in the chromatogram, which is called the injection peak. A sample of lower total ionic concentration than that of the eluent will create a zone of lower E– concentration that will ultimately show up as a negative injection peak. The magnitude of the injection peak (either positive or negative) can be used to estimate the total ionic concentration of the sample compared with that of the eluent. Sometimes the total ionic concentration of the sample is adjusted to

1.3 Principles of Ion Chromatographic Separation and Detection

match that of the eluent in order to eliminate or reduce the size of the injection peak. Behind the zone in the column due to sample injection, the total anion concentration in the column solution again becomes constant and is equal to the E– concentration in the eluent. However, continuous ion exchange will occur as the various sample anions compete with E– for the exchange sites on the resin. As eluent containing E– continues to be pumped through the column, the sample anions will be pushed down the column. The separation is based on differences in the ion-exchange equilibrium of the various sample anions with the eluent anion, E–. Thus, if sample ion A1– has a lower affinity for the resin than ion A2–, then A1– will move at a faster rate through the column than A2–. 1.3.4 Migration of Sample Ions

The general principles for separation are perhaps best illustrated by a specific example. Suppose that chloride and bromide are to be separated on an anionexchange column. The sample contains 8 × 10–4 M sodium chloride and 8 × 10–4 M sodium bromide, and the mobile phase (eluent) contains 10 × 10–4 M sodium hydroxide. In the column equilibration step, the column packed with solid anion-exchange particles (designated as Res-Cl–) is washed continuously with the NaOH eluent to convert the ion exchanger completely to the –OH– form. Res-Cl– + OH– > Res-OH– + Cl– At the end of this equilibration step, the chloride has been entirely washed away and the liquid phase in the column contains 10 × 10–4 M Na+OH–. In the sample injection step a small volume of sample is injected into the ionexchange column. An ion-exchange equilibrium occurs in a fairly narrow zone near the top of the column. Res-OH– + Cl– > Res-Cl– + OH– Res-OH– + Br– > Res- Br– + OH– Within this zone, the solid phase consists of a mixture of Res-Cl– , Res-Br– and ResOH–. The liquid phase in this zone is a mixture of OH–, Cl– and Br– plus its accompanying Na+. The total anionic concentration is governed by that of the injected sample, which is 16 × 10–4 M (see Figure 1.6A).

15

16

1 Introduction and Overview

Figure 1.6 Anion exchange column: A, after sample injection; B, after some elution with 0.001 M NaOH.

In the elution step, pumping 10 × 10–4 M NaOH eluent through the column results in multiple ion-exchange equilibria along the column in which the sample ions (Cl– and Br–) and eluent ion (OH–) compete for ion-exchange sites next to the Q+ groups. The net result is that both Cl– and Br– move down the column (Figure 1.6B). Because bromide has a greater affinity for the Q+ sites than chloride has, the bromide moves at a slower rate. Because of their differences in rate of movement, bromide and chloride are gradually resolved into separate zones or bands. The solid phase in each of these zones contains some OH– as well as the sample ion, Cl– or Br–. Likewise, the liquid phase contains some OH– as well as Cl– or Br–. The total anionic concentration (Cl– + OH– or Br– + OH–) is equal to that of the eluent (0.0010 M) in each zone. Continued elution with Na+OH– causes the sample ions to leave the column and pass through a small detector cell. If a conductivity detector is used, the conductance of all of the anions, plus that of the cations (Na+ in this example) will contribute to the total conductance. If the total ionic concentration remains constant, how can a signal be obtained when a sample anion zone passes through the detector? The answer is that the equivalent conductance of chloride (76 ohm–1 cm2 equiv–1) and bromide (78) is much lower than that of OH– (198). The net result is a decrease in the conductance measured when the chloride and bromide zones pass through the detector. In this example, the total ionic concentration of the initial sample zone was higher than that of the eluent. This zone of higher ionic concentration will be

1.3 Principles of Ion Chromatographic Separation and Detection

displaced by continued pumping of eluent through the column until it passes through the detector. This will cause an increase in conductance and a peak in the recorded chromatogram called an injection peak. If the total ionic concentration of the injected sample is lower than that of the eluent, an injection peak of lower conductance will be observed. The injection peak can be eliminated by balancing the conduction of the injected sample with that of the eluent. Strasburg et al. studied injection peaks in some detail [41]. In suppressed-anion chromatography, the effluent from the ion-exchange column comes into contact with a cation-exchange device (Catex-H+) just before the liquid stream passes into the detector. This causes the following reactions to occur. Eluent: Chloride: Bromide:

Na+OH– + Catex-H+ → Na+Cl– + Catex-H+ → Na+Br– + Catex-H+ →

Catex-Na+ + H2O Catex-Na+ + H+Cl– Catex-Na+ + H+Br–

The background conductance of the eluent entering the detector is thus very low because virtually all ions have been removed by the suppressor unit. However, when a sample zone passes through the detector, the conductance is high due to the conductance of the chloride or bromide and the even higher conductance of the H+ associated with the anion. 1.3.5 Detection

This effect can be used to practical advantage for the indirect detection of sample anions. For example, anions with little or no absorbance in the UV spectral region can still be detected spectrophotometrically by choosing a strongly absorbing eluent anion, E–. An anion with a benzene ring (phthalate, p-hydroxybenzoate, etc.) would be a suitable choice. In this case, the baseline would be established at the high absorbance due to E–. Peaks of non-absorbing sample anions would be in the negative direction owing to a lower concentration of E– within the sample anion zones. Direct detection of anions is also possible, providing a detector is available that responds to some property of the sample ions. For example, anions that absorb in the UV spectral region can be detected spectrophotometrically. In this case, an eluent anion is selected that does not absorb (or absorbs very little). 1.3.6 Basis for Separation

The basis for separation in ion chromatography lies in differences in the exchange equilibrium between the various sample ions and the eluent ion. A more quantitative treatment of the effect of ion-exchange equilibrium on chromatographic separations is given later. Suppose the differences in the ion-exchange equilibrium are

17

18

1 Introduction and Overview

very small. This is the case for several of the transition metal cations (Fe2+, Co2+, Ni2+, Cu2+, Zn2+, etc.) and for the trivalent lanthanides. Separation of the individual ions within these groups is very difficult when it is based only on the small differences in affinities of the ions for the resin sites. Much better results are obtained by using an eluent that complexes the sample ions to different extents. An equilibrium is set up between the sample cations, C2+, and the complexing ligand, L–, in which species such as C2+, CL+, CL2 and CL3– are formed. The rate of movement through the cation-exchange column is inversely proportional to a, the fraction of the element that is present as the free cation, C2+.

References [1] F.W. E. Strelow and H. Sandrop, Distri-

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

bution coefficients and cation-exchange selectivities of elements with AG50WX8 resins in perchloric acid, Talanta, 19, 1113, 1972. J. S. Fritz and S. K. Karraker, Ion exchange separation of metal cations, Anal. Chem., 32, 957, 1960. F. W. E. Strelow, An ion exchange selectivity scale of cations based on equilibrium distribution coefficients, Anal. Chem., 32, 1185, 1960. F.W. E. Strelow, R. Rethemeyer, and C. J. C. Bothma, Ion exchange selectivity scales for cations in nitric acid and sulfuric acid with a sulfonated polystyrene resin, Anal. Chem., 37, 106, 1965. J. S. Fritz and B. B. Garralda, Cation exchange separation of metal ions with hydrobromic acid, Anal. Chem., 34, 102, 1962. J. S. Fritz, B. B. Garralda and S. K. Karraker, Cation exchange separation of metal ions by elution with hydrofluoric acid, Anal. Chem., 33, 882, 1961. J. S. Fritz and J. E. Abbink, Cation exchange separation of vanadium from metal ions, Anal. Chem., 34, 1080, 1962. F. W. E. Strelow, Separation of titanium from rare earths, beryllium, niobium, iron, aluminum, thorium, magnesium, manganese and other elements by cation exchange chromatography, Anal., Chem., 35, 1279, 1963. J. S. Fritz and L. H. Dahmer, Cation exchange separation of molybdenum, tungsten, niobium and tantalum from

[10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

other metals, Anal. Chem., 37, 1272, 1965. F. W. E. Strelow and T. N. van der Walt, Separation of lead from tin, antimony, niobium, tantalum, molybdenum and tungsten by cation exchange chromatography in tartaric-nitric acid mixtures, Anal. Chem., 47, 2272, 1975. J. N. Story and J. S. Fritz, Forced-flow chromatography of the lanthanides employing continuous in-stream detection, Talanta 21, 894, 1974. J. S. Fritz and T. A. Rettig, Cation exchange in acetone-water-hydrochloric acid, Anal. Chem., 34, 1562, 1962. F. W. E. Strelow, A. H. Victor, C. R. van Zyl, and C. Eloff, Distribution coefficients and cation exchange behavior of elements in hydrochloric acid-acetone, Anal. Chem., 43, 870, 1971. J. Korkisch, Modern methods for the separation of rarer metal ions, Pergamon, Oxford 1969. Wescan Instruments, Inc., Santa Clara, CA, ‘Wescan Ion Analyzer #6’, (1983). K. A. Kraus and F. Nelson, Proc. First U. N. Int. Conf. on Peaceful Uses of Atomic Energy, 7, 113, 1956. K. A. Kraus. G. E. Moore and F. Nelson, Anion exchange studies. XXI. Th(IV) and U(IV) in hydrochloric acid. Separation of thorium, protoactinium and uranium., J. Am. Chem. Soc., 78, 2692, 1956. F. Nelson, R. M. Rush and K. A. Kraus, Anion exchange studies. XXVII. Adsorbability of a number of elements

References

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

in HCl-HF solutions, J. Am. Chem. Soc., 82, 339, 1960. J. B. Headridge and E. J. Dixon, The analysis of complex alloys with particular reference to niobium, tantalum and tungsten, Analyst, 87, 32, 1962. E. A. Huff, Anion exchange study of a number of elements in nitric-hydrofluoric acid mixtures, Anal. Chem., 36, 1921, 1964. L. Danielsson, Adsorption of a number of elements from HNO3HF and H2SO4HF solutions by cation and anion exchange, Acta Chem. Scand., 19, 1859, 1965. L. Danielsson, Adsorption of a number of elements from sulfuric acid solutions by anion exchange, Acta Chem. Scand., 19, 670, 1965. F. W. E. Strelow and C. J. C. Bothma, Anion exchange and a selectivity scale for elements in sulfuric acid media with a strongly basic resin, Anal. Chem., 39, 595, 1967. J. S. Fritz and B. B. Garralda, Anion exchange separation of thorium using nitric acid, Anal. Chem., 34, 1387, 1962. J. Korkisch and L. Hazan, Anion exchange behavior of uranium, thorium, the rare earths and various other elements in hydrochloric acid-organic solvent media, Talanta, 11, 1157, 1964. W. R. Heumann, Ion exchange in nonaqueous and mixed media, Crit. Rev. in Anal. Chem., 2, 425, 1971. J. S. Fritz and Marcia Lehoczky Gillette, Anion-exchange separation of metal ions in dimethylsulfoxide-methanolhydrochloric acid, Talanta, 15, 287, 1968. M. D. Seymour, J. P. Sickafoose and J. S. Fritz, Application of forced-flow liquid chromatography to the determination of iron, Anal. Chem., 43, 1734, 1971. M. D. Seymour and J. S. Fritz, Rapid, selective method for lead by forced-flow liquid chromatography, Anal. Chem., 45, 1632, 1973. J. N. Story and J. S. Fritz, Forced-flow chromatography of the lanthanides

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

employing continuous in-stream detection, Talanta, 21, 892, 1974. M. D. Seymour and J. S. Fritz, Determination of metals in mixed hydrochloric and perchloric acids by forced-flow anion exchange chromatography, Anal. Chem., 45, 1394, 1973. K. Kawazu and J. S. Fritz, Rapid and continuous determination of metal ions by cation exchange chromatography, J. Chromatogr., 77, 397, 1973. J. S. Fritz and J. N. Story, Chromatographic separation of metal ions on low capacity macroreticular resins, Anal. Chem., 46, 825, 1974. J. S. Fritz and J. N. Story, Selectivity behavior of low-capacity, partially sulfonated macroporous resin beads, J. Chromatogr., 90, 267, 1974. S. Elchuk and R. M. Cassidy, Separation of the lanthanides on high-efficiency bonded phases and conventional ion exchange resin, Anal. Chem., 51, 1434, 1979. H. Small, T. S. Stevens and W. S. Bauman, Novel ion exchange chromatographic method using conductometric detection, Anal. Chem., 47, 1801, 1975. D. T. Gjerde and J. S. Fritz, Effect of capacity on the behavior of anionexchange resins, J. Chromatogr., 176, 199, 1979. D. T. Gjerde, J. S. Fritz and G. Schmuckler, Anion chromatography with low conductivity eluents, J. Chromatogr., 186, 509, 1979. D. T. Gjerde, G. Schmuckler and J. S. Fritz, Anion chromatography with low-conductivity eluents II, J. Chromatogr., 187, 35, 1980. J. S. Fritz, D. T. Gjerde and R. M. Becker, Cation chromatography with a conductivity detector, Anal. Chem., 52, 1519, 1980. R. T. Strasburg, J. S. Fritz, J. Berkowitz and G. Schmuckler, Injection peaks in anion chromatography, J. Chromatogr., 482, 343, 1989.

19

21

2 Instrumentation 2.1 Components of an Ion Chromatograph (IC) Instrument

Most ion chromatographs are integrated instruments with the various components of the instrument designed so that the plumbing and wiring match and fit together perfectly. Ion chromatography is truly global. Commercial suppliers located in the United States, Europe and Asia include a number of vendors – Dionex, Metrohm, Schimadzu, Alltech, Lachat and others. These companies offer complete instruments with a wide variety of features and options. Integrated instrumentation is desirable; this type of instrument usually performs exceptionally well giving sharp and sensitive peaks and rapid separations. One example, the Personal Ion Analyzer 1000 available from Shimadzu (Kyoto, Japan), is a portable (15 kg) nonsuppression ion chromatograph. It is operated either by AC (power supply) or DC (battery), enabling analysis to be performed on site or where there is limited laboratory space. Ion chromatographs available from other vendors are integrated but also include an autosampler and sit on a laboratory bench. Conventional high-performance liquid chromatography (HPLC) hardware [1–3] has been undergoing an evolution in recent years to be able to operate at higher pressures and with smaller-diameter columns. In some cases, the newer hardware is called Ultra HPLC or UHPLC. Regardless of what it is called, HPLC hardware has many similarities to IC hardware, and having the option of using this instrumentation can be a valuable resource when determining the best hardware for a particular ion analysis problem, or at least assembling a workable system. In this chapter, we describe the various components of an ion chromatographic instrument, their function, how the instrument is built, and how to recognize parts of the instrument in the event that maintenance is needed. An understanding of some of the available instrumental options is also helpful in achieving better separations or using other detectors different from those included with an integrated IC instrument.

Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

22

2 Instrumentation

Figure 2.1 Block diagram of an ion chromatograph.

The block diagram in Figure 2.1 shows the general arrangement of components of an IC instrument. The essential hardware requirements are as follows: 1. Eluent reservoir. Electrical generation of eluent ions is used in some cases. 2. Pump and flow regulator. An optional gradient system permits programmed changes in eluent composition. 3. Sample injection unit. Purpose is to introduce a precise volume of sample for ion analysis. 4. Column and column oven to maintain a constant, pre-set temperature. 5. Suppressor. This useful component is only needed for systems that employ suppressed-conductivity detection. 6. Detector. Function is to detect sample ions as they are separated and pass through the detector. 7. Data processor. This unit is linked to the detector output so that a digital record of the chromatographic separation is obtained. The complete chromatogram can be viewed on a computer screen and printed out if desired. Each of these components is described in the following sections. The setup typically includes the IC instrument itself and a small computer with a display screen. The entire assembly occupies only a small amount of bench space.

2.2 General Considerations

2.2 General Considerations

Everything on the high-pressure side in the system, from the pump outlet to the end of the column, must be strong enough to withstand the pressures involved. The wetted parts are usually made of PEEK and other types of plastics, although other materials, such as sapphire, ruby, or even ceramics are used in the pump heads, check valves, and injector of the system. PEEK and other high-performance plastics are the materials of choice for ion chromatography because of their ability to withstand high fluid pressures. Stainless steel, usually found in HPLC equipment, can be used provided that the system is properly conditioned to remove internal corrosion and the eluents that are used do not promote further corrosion. Almost all IC eluents are not corrosive to stainless steel provided that this has been pretreated so that surface corrosion is not present. Most acids including sulfuric and nitric acid are not corrosive; however, hydrochloric acid is extremely corrosive to stainless steel. The reader is advised to consult the instrument manufacturer for care and upkeep instructions. It has been said that having a basic knowledge of plumbing can be valuable to the ion chromatograph user. It is interesting to note that, just as household plumbing requires some skill to make kitchens and bathrooms work properly, the plumbing in ion chromatography has to be done correctly for the same reason. One of the greatest points of concern is to configure the instrument so that system dead volume is kept to a minimum. System dead volume is any empty space or volume that the fluid occupies in an IC system. This includes all of the fluid path volume from the injection volume to the detector cell including the injection valve loop and fluid path, the tubing, tubing fittings and union, column end fitting, the space between the beads in the column and the detector fluid path. Too much dead volume or un-swept volumes will lead to peak broadening or peak tailing and consequent loss in separation efficiency. The system dead volume can be controlled by using small internal diameter tubing and keeping the tubing as short as is practical. It is important to use small bore tubing (0.007 inch, 0.18 mm) in short lengths when making connections between the injection and the column and between the column and the detector. Dead volume from the pump to the injector should also be kept small to help to make possible rapid changes in the eluent composition in gradient elution. The tubing fittings must be connected properly by bottoming out the end of the tubing into the fitting before final tightening. Although all regions in the flow path are important, the most important region where peak broadening can happen is in the tubing and connections from the exit end of the column to the detector cell.

23

24

2 Instrumentation

2.3 Eluent

The prepared eluents are contained in a reservoir and pumped at a constant or programmed rate into the chromatographic system. Eluents are usually prepared by dissolving buffers, acids or bases in aqueous solvent or in an aqueous–organic solvent mixture. Eluent entering the pump should not contain any dust or other particulate matter. Particulates can interfere with pumping action and damage the seals or valves. Material can also collect on the inlet frits or on the inlet of the column, causing pressure buildup. Eluents or the water and salt solutions used to prepare the eluents are normally filtered with a 0.2 or 0.45 lm nylon filter. Various eluents that can be used with ion chromatography are described throughout this book. In this chapter, hydroxide-based eluents are described because they are generated electrolytically by the instrumentation rather than simply mixing a base with water. NaOH, KOH or other hydroxide eluents are desirable because of their low suppressed background conductivity. This leads to the ability to form eluent gradients with small shifts in the baseline. Also, low background conductivity can improve the detection limit for ions. However, hydroxide is a weak driving anion, and high concentrations are needed to elute anions that are strongly held by the column. Also, unless the column is designed to operate with hydroxide eluents, the analyte peaks may not be symmetrical. For many workers, the advantages of gradients and low detection limits offset the disadvantages of long retention times and possible asymmetrical peaks. Pure hydroxide eluents are difficult to make, because of persistent contamination by carbon dioxide that is converted to carbonate by the high pH. Carbonate is a much stronger eluting anion than hydroxide, and its presence can shift sample retention times to much shorter (and inconsistent) retention times. Carbonate will also cause baseline shifts when gradients are generated. In fact, a baseline shift during a hydroxide gradient is a good diagnostic indication that one or more of the eluent reservoirs contain bicarbonate or carbonate anion. The electrolytic generation of hydroxide eluent was first described by Dasgupta and coworkers [4, 5]. Rather than mixing reagents, the hydroxide eluent is formed electrochemically as it is being used and is introduced directly into the elution column from the generator. The system permits direct electrical control of the eluent concentration, and gradient chromatography is accomplished without mechanical proportioning. The system contains an anode and a cathode across which a DC current is passed. The reduction reaction at the cathode produces the hydroxide anion. 2 H2O + 2 e– → 2 OH– + H2 ↑ (at cathode) A counterion to hydroxide is needed to conserve electric neutrality. Also, an oxidizing reaction occurs simultaneously at the anode. OH– is electrolytically neutralized and O2 is evolved.

2.3 Eluent

H2O – 2 e– → 2 H+ + 1⁄2 O2 ↑ (at anode without NaOH) However, the feed solution for the anode also contains NaOH. 2 Na+ + 2 OH– + H2O – 2 e– → 2 Na+ + 2 H2O + 1⁄2 O2 ↑ (at anode with NaOH) The Na+ and OH– are combined to form the eluent through the use of an ionexchange membrane. A cation-exchange membrane separating the anode from eluent flow allows the Na+ to join the OH– from the cathode.

Figure 2.2 Schematic representation of EG40 electrolytic production of potassium hydroxide (KOH) eluent (courtesy of Dionex Corp.)

A variation on the concept has been introduced by Dionex as the EG40 module [6]. In this case, KOH contained in a reservoir (labeled in Figure 2.2 as K+ Electrolyte Reservoir) is used rather than an NaOH feed solution. The process is the same; however, K+ is generated from the anode, because its counterion OH– is consumed in the production of H+. K+ migrates across the cation-exchange membrane to combine with OH– formed at the cathode. Carbon dioxide is removed from the eluent stream en route to the EG40, to prevent contamination by carbonate. The electrolyte reservoir must be changed when the K+Electrolyte is depleted. An analogous system can be used to generate methanesulfonic acid (MSA) eluent for the separation of cations (Figure 2.3). In this case, the anode generates H+ for eluent production. The cathode generates OH– anion that combines with the H+ in the MSA electrolyte reservoir. MSA– anion migrates across the anionexchange membrane to combine with the H+ eluent cation (maintaining electric neutrality). Dionex offers a ‘just add water’ system for electrolyte generation and purification of commonly used eluents such as KOH and methanesulfonic acid. Eluents are generated from deionized water using an Eluent Generator (EG) cartridge and then polished of contaminants using a continuous-regeneration trap column.

25

26

2 Instrumentation

Figure 2.3 Schematic representation of EG40 electrolytic production of methanesulfonic acid (MSA) eluent (courtesy of Dionex Corp.)

While very popular, the use of basic or acidic eluents with suppressed conductivity detection is certainly not the only useful form of IC. It is often advantageous to use another chemical type of eluent in conjunction with detection by nonsuppressed conductivity, UV–Vis spectroscopy, and electrochemical methods, among others. Eluents can be prepared at almost any desired pH and are generally quite stable except for eluents that are at very high pH. A freshly prepared eluent should be filtered before use to remove any sediment.

2.4 Pump

The IC system begins with the pump, which should be completely inert, robust, and capable of high precision and accuracy needed for reproducible chromatographic results. Ideally, the pump should support a range of flow rates consistent with using columns having diameters in the range 2–9 mm. Smaller column diameters usually require lower pump flow rates. The pump should also incorporate or have available vacuum degassing of the mobile phase and provide for gradient elution through a proportioning valve. Both single-piston (SP) and dual-piston (DP) pumps are available. IC pumps are designed around an eccentric cam that is connected to a piston (Figure 2.4). The rotation of the motor is transferred into the reciprocal movement of the piston. A pair of check valves controls the direction of flow through the pump head (discussed below). A pump seal surrounding the piston body keeps the eluent form leaking out of the pump head. The pump seal will wear and must be replaced periodically. In single-headed reciprocating pumps, the eluent is delivered to the column for only half of the pumping cycle. A pulse dampener is used to soften the spike of

2.4 Pump

Figure 2.4 IC pump head, piston, and cam.

pressure at the peak of the pumping cycle and to provide an eluent flow when the pump is refilling. Use of a dual-headed pump is better because heads are operated 180° out of phase with each other. One pump head pumps while the other is filling and vice versa. The eluent flow rate is usually controlled by the pump motor speed although there are a few pumps that control flow rate by control of the piston stroke distance. Figure 2.5 shows how the check valve works. On the intake stroke, the piston is withdrawn into the pump head, causing suction. The suction causes the outlet check valve to settle onto its seat while the inlet check valve rises from its seat, allowing eluent to fill the pump head. Then the piston travels back into the pump head on the delivery stroke. The pressure increase seals the inlet check valve and opens the outlet valve, forcing the eluent to flow out of the pump head to the injection valve and through the column. Failure of either of the check valves to seal properly will cause pump head failure, and eluent will not be pumped. In most cases, this is due to air trapped in the valve so that the ball cannot sit properly. Flushing or purging the head usually takes care of this problem. Using degassed eluents is also helpful. In a few cases, particulate material can prevent sealing of the valve, and in these cases the valve must be cleaned or replaced. The pump manufacturer has instructions on how to perform this operation.

27

28

2 Instrumentation

Figure 2.5 Check valve positions during intake and delivery strokes of the pump head pistons.

2.4 Pump

2.4.1 Gradient Formation

Isocratic separations are performed with an eluent at a constant concentration of eluent buffer or salt solution. Isocratic elution is desirable because it is simpler, but it is sometimes necessary or desirable to perform separations using eluent

Figure 2.6 High-pressure mixing systems use two or more independent pumps to generate the gradient. Low-pressure mixing systems use a single pump with a proportioning valve to control composition. The advantages of high-

pressure mixing are smaller dwell volumes and faster gradient formation. The advantages of low-pressure mixing are lower costs (single pump) and more versatile gradients (four solvents).

29

30

2 Instrumentation

gradients where the eluent strength is increased gradually over a chromatographic run. This allows the separation of anions that may have a wide range of affinities for the column. Weakly adhering anions elute first, and then, as the eluent concentration is increased, more strongly adhering anions can be eluted by the stronger eluent. Figure 2.6 shows the two most popular methods for forming gradients. In the first method, flow from two high-pressure pumps is directed into a high-pressure mixing chamber. One pump contains a weak eluent while the other contains the stronger eluent. After the mixing chamber, the flow is directed to the injector and then on to the column. Control of the relative pumping rates of each pump creates the gradient. The total flow from the two pumps is constant. The gradient starts with a high flow of the weak eluent pump and a low flow of the strong eluent pump. Then, over the course of the chromatographic run, the flow rate of the strong eluent pump is increased while the flow rate of the weak eluent pump is decreased, keeping the total flow rate constant. A more popular and less expensive method of forming gradients is by using a single pump and three or four micro-proportioning valves at the inlet of the pump. Each valve controls an eluent of a given concentration. Only one valve is open at a time. The concentration of the eluent that is delivered to the pump depends on the relative time that any particular valve is open. To start a gradient, the valve controlling the low concentration of the eluent is open for longer periods of time. As a gradient increases, this valve will close off while the valve controlling higher eluent concentrations is opened. Most gradients are formed with two of the valves although it is certainly possible to use more complex gradients with 3 or even 4 valves within one run. Normally, the other valves control different types of eluents or column cleaning solutions.

2.5 Sample Injector

The sample injection hardware is designed to introduce a small (1–100 lL) and reproducible volume of sample into the ion chromatograph. The injection system may be manual or automated. An automated system (often called an autosampler) permits the storage of multiple samples for unattended injection and analysis. The injection system may be manual or automated, but both rely on the injection valve. An injection valve is designed to introduce precise amounts of sample into the sample stream with variation usually less than 0.5% volume difference from injection to injection. Figure 2.7 schematically represents a 6-port and 2-position device valve. In one position the sample is loaded and the other it is injected. In the load position, the sample from the syringe or autosampler vial is pushed into the injection loop. The loop may be partially filled (partial loop injection) or completely filled (full loop injection). Partial loop injection depends on the precision filling of the loop with small known amounts of material. If partial loop injection is used, the loop must not be filled to more than 50% of the total

2.6 Columns

loop volume or the injection may not be precise. In full loop injection, the sample is pushed completely through the loop. Normally at least a two-fold amount of sample is used to fill the loop with excess sample from the loop going to waste. Typical loop sizes are 10–200 lL.

Figure 2.7 Schematic representation of partial- and full-loop injection methods.

At the same time that the sample loop is loaded with sample, the eluent travels in the by-pass channel of the injection valve and to the column. Injection of the sample is accomplished by turning the valve and placing the injection loop into the eluent stream. Usually the flow of the eluent is opposite to that of the loading of the sample into the loop. The injected sample travels to the head of the column as a slug of fluid. The ions in the sample interact with the column and the separation process is started with the eluent pushing the sample components down the column. Injection valves require periodic maintenance and usually have to be serviced after about 5000 injections. The instrument manual should be consulted for details on service.

2.6 Columns 2.6.1 Column Hardware

IC columns are usually made of PEEK or other polymer. Even the frits at the end of the column, which hold the column packing in place, are usually made of porous PEEK. The column lengths range from about 3 to 30 cm and the inside diam-

31

32

2 Instrumentation

eters from about 1 to 7.8 mm. Figure 2.8 shows the end of a column and the type of fitting used to connect the tubing to the column. Reusable PEEK fittings are used almost exclusively to connect tubing to columns and other instrument components. As stated earlier, the tubing should be bottomed out or pushed completely into the column end before the fitting is tightened, to ensure that there is no unnecessary dead volume in the connection.

Figure 2.8 Representation of a typical IC column and end fitting. The fittings, frits and body are normally PEEK. The PEEK fittings to attach the tubing to the column are normally tightened by hand and not with a wrench. Some columns will contain a replaceable disk or frit at the top of the column to protect the column from particulates and contaminants. The trend is toward smaller-diameter columns.

2.6.2 Column Protection

Not only does column protection extend the useful life of the separation column, but proper protection of the separation column can also result in more reliable analytical results over the lifetime of the column. Scavenger columns, located between the pump and the injector, are one means of protecting the column. The scavenger removes particulate material that may be present in the eluent, but can also contain a resin to ‘polish’ the eluent of any contaminant. An example is a chelating resin to remove metal contaminants. Besides protecting the separation column, scavenger columns may also improve detection of the analytes by reducing the background signal due to residual contaminants. Most IC users do not use scavenger columns but rather prefer the use of guard columns, located directly in front of the separation column. Guard columns generally contain the same material as the separation column. Therefore, material that would be trapped and would contaminate the separation column will instead get trapped by the guard column. Guard columns are changed when the separation of a standard is no longer acceptable and the column cannot be regenerated by the recommended procedures. Several guard columns may be used for protec-

2.7 Suppressor

tion over the lifetime of the separation column. Guard columns are generally smaller than the separation columns, but can add to the retention time of the separation. Some users may prefer to use an inline filter in place of or even in addition to the guard column. 2.6.3 Column Oven

There is about a 2% change in conductance per degree C change in temperature. If the eluent has a background conductance, then temperature control is important to reduce detector noise and improve detection limits. Good quality conductivity detectors have temperature control, temperature compensation, or both. An oven can keep the temperature of the fluid constant by the time it reaches the conductivity cell, and this also helps to improve detector noise and detection limits. Some column ovens are designed to preheat the eluent to a pre-set temperature as it enters the column. Use of a column oven permits IC separations to be run at temperatures ranging from sub-ambient to around 85 °C. 2.6.4 Two-dimensional IC

Applications that employ valve switching have been successfully used in ion chromatography. Systems are available that permit sample matrix diversion or matrix elimination prior to analysis of trace components. The strategy used is as follows: In the first stage, a large-volume sample loop is used to obtain a partial separation. The objective here is to focus the partially resolved peaks of interest onto a concentrator column in the second stage. This is accomplished by valve switching. The second column is of smaller diameter and is operated at a lower flow rate relative to the first column. Resolution of the peaks of interest is completed on the second column. Improved resolution is frequently possible by using eluents with different chemical compositions in the two columns.

2.7 Suppressor

A suppressor is only needed when analyte ions are to be detected by suppressed conductivity. Hydrogen ions pass through a cation-exchange membrane and convert a basic eluent to a low-conducting weak acid. An equivalent amount of the eluent cation, K+, passes through the membrane in the opposite direction to maintain electro-neutrality. An analyte anion, such as Cl–, now has the highly conductive H+ as its counterion, which can now be detected on top of the low-background eluent signal that has now been suppressed. An acidic eluent such as methanesulfonic acid (MSA), which is used in cation chromatography, works on essentially the same principle. The H+ of the eluent is

33

34

2 Instrumentation

neutralized by OH– from the basic regenerant which passes through an anionexchange membrane. Analyte cations are now paired up with OH–, which has a high conductance. The original micromembrane suppressor, introduced by Dionex in 1991, required a chemical regenerant with a flow three to ten times faster than the eluent flow. This problem was overcome by the Self Regenerating Suppressor (SRS) [7]. In this system, H+ or OH– for suppression is generated by electrolysis of water at a Pt anode (or cathode) placed in the regenerant chambers. The current required to start the electrolysis is proportional to the eluent concentration that has to be suppressed. A continuously electrolytically regenerated packed-bed suppressor (known as the Atlas) was made available by Dionex in 2001 [8]. This instrument contains monolithic cation-exchange disks attached to flow distributor disks with small holes to allow the liquid to flow. These are placed between the ion-exchange membranes. New models of membrane suppressors work with high eluent concentrations and have very low dead volume. Another suppression device uses a packed-bed suppressor in disposable cartridges. A device with solid-phase electrochemical cells called Electrically Regenerated Ion Suppression (ERIS) is marketed by Alltech [9]. A degassing unit for removing CO2 from suppression of carbonate is included in a more recent suppressor. Metrohm markets a column suppressor module that uses two suppressor columns. One is regenerated while the other is being used. When the first is exhausted then the other is available for eluent suppression.

2.8 Detector

The various types of detectors that can be used in ion chromatography are discussed in Chapter 4. Variable wavelength UV–Vis detectors are extremely useful for detection of sample ions with sufficient absorbance at the analytical wavelength. Various electrochemical detectors, including the pulsed amperometric detector (PAD), offer excellent selectivity and sensitivity. Detection by mass spectrometry is growing in popularity. This is undoubtedly a result of the high detection sensitivity and positive identification offered by MS in association with the fast development of interfaces between column and detector that have greatly simplified the IC/MS combination. Several combinations of IC and MS detection are available. These include ICP-MS (element-specific detection) and MS with atmosphere pressure ionization (API), operated with either electrospray ionization (ESI) or atmospheric pressure chemical ionization (APCI). The use of IC-MS where the column outlet is connected directly to a mass spectrometer may impose additional requirements on the IC system. Since a vacuum must be maintained, more volatile ammonium salts rather than sodium or potassium salts are used in the eluent. Inside the mass spectrometry interface evaporation of the IC liquid, ionization of neutral species to charged species, and removal

2.9 Data Acquisition and Calculation of Results

of a very large amount of vapors from the mobile phase, must take place to maintain the required vacuum conditions. But of all of the many choices for detector, the conductivity detector remains the most popular. Conductivity detection is universal, rugged and sensitive. Adding a suppressor (discussed in Chapter 4 and elsewhere) dramatically improves the sensitivity for the majority of anions. The only exception to this is weak base anions which become low conducting in a suppressor. But the gain in detection for most anions and cations is so impressive that suppressed conductivity is the most popular form of detection, which means that a suppressor is placed after the exit of the separation column and before the conductivity cell. Membrane suppressors are marketed by Dionex. As described above, several column suppressor patents have expired and there are several companies offering this type of suppressor. Nonsuppressed conductivity, where the exit end of the column is connected directly to the conductivity cell, is effective for measuring many types of cations and weak acid anions.

2.9 Data Acquisition and Calculation of Results

Chromatographic data acquisition has come a long way from earlier days when the output of the detector was connected to a pen-and-ink recorder moving at a fixed chart speed so that the resulting chromatogram could be recorded. Now, the results of the chromatographic separation are almost always stored and displayed on a computer. The computer uses an A/D (analog to digital) board to convert the analog signal from the detector to digital. The digital information is stored and manipulated to report the results to the user. The scale of both the vertical axis (detector signal) and the horizontal axis (elution time) can be adjusted to give a record of the separation. The most useful types of information are the peak retention times and the peak areas or peak heights. Retention times are used to confirm the identity of the various peaks, and peak area or peak height is a measure of concentration. Two general methods can be used to calculate the concentration of a given peak: use of a calibration curve or use of a standard-to-unknown ratio. Generally, it is usually better to prepare a calibration curve by a plot of peak area (or height) against the concentrations of known standards. Such a plot will normally be a straight line, but a perfectly valid calibration plot may deviate from linearity, especially at lower concentrations. If the calibration curve is linear and passes through zero, a fast method is to compare an analytical peak of interest to a single standard of known concentration. This calculation can be performed by use of a simple ratio: unknown concentration known concentration ˆ unknown peak area known peak area

35

36

2 Instrumentation

therefore: unknown concentration ˆ

known concentration × unknown peak area known peak area

If the calibration curve is not linear or does not pass through zero, then it best to use the calibration curve to calculate unknown concentrations.

References [1] C. F. Poole and S. K. Poole, Chromato-

[2]

[3]

[4]

[5]

graphy Today, Elsevier Science, Amsterdam, 1991. M. W. Dong, Modern HPLC for Practicing Scientists, Wiley Interscience, New York, 2006. S. Kromidas, Ed., HPLC Made to Measure: A Practical Handbook for Optimization, Wiley VCH, Weinheim, 2006. D. L. Strong and P. K. Dasgupta, K. Friedman and J. R. Stillian, Electrodialytic eluent production and gradient generation in ion chromatography, Anal. Chem., 63, 480, 1991. D. L. Strong, C. U. Joung, P. K. Dasgupta, Electrodialytic eluent generation and suppression: ultralow background conductance suppressed

[6] [7]

[8]

[9]

anion chromatography, J. Chromatogr., 546, 159, 1991. New products brochure EG40, Dionex Corp. Sunnyvale, CA , 1999. C. Pohl, R. Slingsby, V. Barreto, K. Friedman and M. Toofan, New membrane-based electrolytic suppressor device for suppressed conductivity detection in ion chromatography, J. Chromatogr., 640, 97, 1993. H. Small, Y. Liu, J. Riviello, N. Avdalovic, K. Srinivasan. Continuous Electrolytically Regenerated Packed Bed Suppressor for Ion Chromatograph, U.S. Patent 6 325 976, 2001. L.-M. Nair and R. Saari-Nordhaus, Recent developments in surfactant analysis by ion chromatography, J. Chromatogr. A, 804, 233, 1998.

37

3 Resins and Columns 3.1 Introduction

Most ion exchangers used for ion chromatography now have a polymeric base. This is quite the opposite to high-performance liquid chromatography (HPLC) where the use of silica-based materials still predominates. For convenience, we will frequently use the word ‘resins’ to refer to both silica-based and polymeric ion exchangers. Intensive research and development has resulted in ion-exchange materials with much greater selectivity and efficiency than was possible in the earlier days of IC. Some of the new columns are simply improvements of older columns having sharper peaks and shorter analysis times. One interesting side benefit to this is that as peaks become sharper the ability to detect anions at lower levels is also improved. Other column development has been driven by specific needs such as the need to analyze drinking water disinfectants. While it has become increasingly rare for practitioners to make their own resins and pack their own columns it is still important to understand how resins for the columns are made. In this way, both the power and the limitations of the commercial resins can be recognized and used. The chemistry and properties of ion-exchange resins and columns are discussed in this chapter. Specific commercial columns and the practice of anion chromatography are covered in Chapter 5 and cation chromatography in Chapter 7. A cation exchanger is a solid particulate material with negatively charged functional groups arranged to interact with ions in the surrounding liquid phase. For convenience, we will often refer to a cation exchanger as a ‘catex’. The most common type of catex contains sulfonic acid groups. Cross-linked polystyrene particles are converted to a catex by sulfonation with concentrated sulfuric acid. (3.1)

Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

38

3 Resins and Columns

In this equation, Res denotes resin or polymer. Silica-based cation exchangers are generally prepared by reacting silica particles with an appropriate chlorosilane or methoxysilane. A common type of silica catex has the structure:

In both of these materials, the sulfonate group is chemically bonded to the solid matrix. However, the H+ is attracted electrostatically to the –SO3– and can undergo exchange reactions with other ions in solution. For example: Solid–SO3– H+ + Na+ > Solid–SO3– Na+ + H+

(3.2)

2 Solid–SO3– Na+ + Mg2+ > (Solid–SO3–)2 Mg2+ + 2Na+

(3.3)

The physical form of the catex is such that ions from the surrounding solution can readily traverse through the solid to come into contact with the interior as well as the surface sulfonate groups. The exchange reactions (Eqs. 3.2 and 3.3) are reversible and are subject to the laws of chemical equilibrium. Most monovalent metal ions are more strongly held by the catex than H+. Cations of a higher charge are usually retained more strongly than those of lower charge. Ion-exchange equilibria are treated in more detail in Chapter 5.

3.2 Polymeric Resins 3.2.1 Substrate and Cross Linking

Historically, cation- and anion-exchange resins were intended for large-scale applications such as water softening and various areas of chemical processing. These materials were designed to be durable with repeated use and to provide a high ion-exchange capacity with good pH stability. Ion exchangers used in contemporary ion chromatography have a much lower exchange capacity and are designed to emphasize chromatographic separation ability with high efficiency. However, the IC resins of today gradually evolved from earlier resin technology and synthetic methods for introduction of various functional groups. A variety of polymeric substrates can be used in ion-exchange synthesis, including polymers of esters, amides and alkyl halides. But resins based on styrene-divinylbenzene copolymers are probably the most widely used ion exchangers. The polymer is schematically represented in Figure 3.1. The resin is made up primarily of polystyrene; however, a small amount of divinylbenzene is added during the polymerization to ‘cross-link’ the resin. This cross-linking confers mechanical sta-

3.2 Polymeric Resins

bility upon the polymer bead and also dramatically decreases the solubility of the polymer by increasing the molecular weight of the average polymer chain length. Typically, 12–15 wt% of the cross-linking compound is used for microporous resins and up to 55 wt% for macroporous resins. In many cases, the resin name will indicate the cross-linking of the material. For example, a Dowex 50 x 4 cation exchanger contains 4% divinylbenzene polymer.

Figure 3.1 Schematic representation of a styrene–divinylbenzene copolymer. The divinyl-benzene ‘cross-links’ the linear chain of the styrene polymer. A high percentage of divinylbenzene produces a more rigid polymer bead.

3.2.2 Microporous Resins

The starting material for cation- and anion-exchange ‘polymer’ resins can be classified either as microporous or macroporous. Most classical work has been done with microporous ion-exchange resins. Microporous substrates are produced by a suspension polymerization in which styrene and divinylbenzene are suspended in water as droplets. The monomers are kept in suspension in the reaction vessel through rapid, uniform stirring, and the use of a surfactant. Addition of a catalyst such as benzoyl peroxide initiates the polymerization. The resulting beads are uniform and solid but are said to be microporous. The size distribution of the beads is dependent on the stirring rate, that is, faster stirring produces smaller beads. The beads swell but do not dissolve when placed in common hydrocarbon solvents. After the resin is functionalized (the ion-exchange functional groups are attached to the polymer), the bead is considerably more polar. Depending on the relative number of functional groups, polar solvents such as water will now swell the ion-exchange resin. However, nonpolar solvents will tend to dehydrate the bead and cause it to shrink.

39

40

3 Resins and Columns

The extent of ion-exchange resin hydration will also depend on the ionic form of the resin. Ion-exchange resin beads with very little cross-linking are soft and tend to swell or shrink excessively when converted from one ionic form to another. However, the amount of cross-linking used in resin synthesis is still based on a compromise of resin performance. Microporous polystyrene resins usually contain about 8% divinylbenzene. Gel-type resins with a high cross-linking tend to exclude larger ions, and the diffusion of ions of ordinary dimensions within the gel may be slower than might be desired. Resins with cross-linking lower than about 2% are too soft for most column work. 3.2.3 Macroporous Resins

Macroporous resins (sometimes called macroreticular resins) are prepared by a special suspension polymerization process. Again, as with microporous resins, the polymerization is performed while the monomers are kept as a suspension of a polar solvent. However, the suspended monomer droplets also contain an inert diluent that is a good solvent for the monomers, but not for the material that is already polymerized. Thus, resin beads are formed that contain pools of diluent distributed throughout the bead matrix. After polymerization is complete, the diluent is washed out of the beads to form the macroporous structure. The result is rigid, spherical resin beads that have a high surface area. Rather than using an inert solvent to precipitate the copolymer and form the pores, the polymerization may be carried out in the presence of an inert solid agent such as finely divided calcium carbonate to create the voids within the bead. Later, the solid is extracted from the copolymer. Both of these polymerization processes create large (although probably different) inner pores. The average pore diameter can be varied within the range of 20 Å to 500 Å. The final resin bead structure of a macroreticular resin contains many hard microspheres interspersed with pores and channels. Because each resin bead is really made up of thousands of smaller beads (something like a popcorn ball), the surface area of macroporous resins is much higher than that of microporous resins. A gel resin has a (calculated) surface area of less than 1 m2 g–1. However, macroporous resin surface areas range from 25 to as much as 800 m2 g–1. Macroporous resins are remarkably rigid because of the large amounts of crosslinking agents normally used in the synthesis. Such resins are particularly advantageous for performing ion-exchange chromatography in organic solvents since changing solvent polarity does not swell or shrink the resin bed as it might for a gel-type resin. But the high cross-linking does not inhibit the ion-exchange process as it does in gel resins because the resins have pores and channels that are easily penetrated by the ions.

3.2 Polymeric Resins

3.2.4 Chemical Functionalization

Ion exchangers are created by chemically introducing suitable functional groups into the polymeric matrix. In a few instances, monomers are functionalized first and then polymerized into beads. An attractive feature of the aromatic copolymer used in many ion exchangers is that it can be modified easily by a wide variety of chemical reactions. More recently, some ion-exchange substrates have been polymers of esters (polymethacrylate) or amides. The reaction solvent is important in ion-exchange synthesis. In many cases, gel substrates must first be swollen in the reaction solvent to achieve complete functionalization of the resin. (We shall see that complete functionalization is usually not desired in resins used in ion chromatography.) The reaction solvent does not appear to be as critical in ion-exchange synthesis of macroporous substrates. Their surface is already ‘exposed’ and ready to be converted to an ion exchanger. Weak-acid cation-exchange resins that contain carboxylic acid or phosphonic acid ion-exchange groups are often used. A popular type of cation exchanger is made by introducing a sulfonic acid functional group. Resins with the sulfonic acid group are said to be strong-acid ion exchangers. Sulfonation reactions are performed by treating the polystyrene resin with concentrated sulfuric acid. Alternatively, the beads can be reacted with chlorosulfonic acid to produce a sulfonyl chloride group, and the sulfonyl chloride group is then hydrolyzed to the acid. Presumably, the latter reaction effects a more uniform placement of the ionogenic groups because the chlorosulfonic acid reagent is dissolved in an organic solvent. The solvent swells the bead, allowing free access of the chlorosulfonic acid to the aromatic rings. Concentrated sulfuric acid is more polar. Sulfonation with this reagent occurs first on the bead surface and then moves progressively toward the center of the bead. Even though this product is not as homogeneous, resins prepared with concentrated sulfuric acid are more popular for ion chromatography. The anionic group –SO3– that is produced is chemically bound to the resin and its movement is thus severely restricted. However, the H+ counterion is free to move about and can be exchanged for another cation. When a solution of sodium chloride is brought into contact with a cation-exchange resin in the hydrogen ion form, the following exchange reaction occurs: (3.4) If this reaction goes essentially to completion, the resin is said to be in the sodium (ion) form. Traditionally, anion-exchange resins have been made by a two-stage set of reactions (although other synthesis methods are now being used). The first step is a Friedel-Crafts reaction to attach the chloromethyl group to the benzene rings of styrene-divinylbenzene copolymer. Then the anion exchanger is formed by reaction of the chloromethylated resin with an amine. The most common type of

41

42

3 Resins and Columns

strong-base anion-exchange resin contains a quaternary ammonium functional group, which is obtained by alkylation with trimethylamine.

(3.5)

In these resins only the anion is mobile and can be exchanged for another anion. Another common strong-base anion exchanger is one that contains a hydroxyethyl group in place of a methyl group on the nitrogen. Weak-base anion exchangers are synthesized by reacting the chloromethylated resin with lower substituted amines or with ammonia. Weak-base anion-exchange resins cannot function as ion exchangers unless the functional group is protonated: Resin-CH2NH2 + H+ + NO3– → Resin-CH2NH3+ NO3–

(3.6)

The protonation, of course, depends on the basicity of the functional group and the pH of the solution in which the resin beads are immersed. 3.2.5 Resin Capacity

Resin capacity is an extremely important parameter in ion chromatography. Details of the effect of capacity on the behavior of resins for ion chromatography are found in Chapter 5. Generally, resins of lower capacity will lower the eluent concentration needed to elute sample ions from the column. The capacity of a resin is usually given in milliequivalents of exchangeable ion per gram of resin. In some cases it is expressed as milliequivalents per milliliter of resin. High-capacity commercial cation resins contain approximately one functional group per benzene ring and have an exchange capacity of around 4.5 mequiv g–1. Highcapacity anion-exchange resins are typically around 3.5 to 4.0 mequiv g–1 for a geltype resin and around 2.5 mequiv g–1 for a macroporous resin. In almost all cases, the bulk resins that are available commercially are high capacity. However, ion chromatography usually employs low-capacity ion exchangers. Capacities of these resins range from 0.01 to 0.2 mequiv g–1. Many low-capacity resins are pellicular, that is, the ion-exchange groups are on or near the surface of the bead. The density of the functional groups is high on the surface so that the local capacity is high, but the overall capacity of the ion exchanger is very low. However, macroporous resin substrates are porous and have a much higher surface area. Low-capacity ion exchangers made from these substrates have the functional groups distributed throughout the bead, although the exact relationship of the performance of the functional group to its position within the resin matrix is unknown.

3.3 Resins and Columns for Ion Chromatography

Anion-exchange resins of variable but low exchange capacities are produced under mild conditions and short reaction times in the chloromethylation reaction. Conditions for the amination are chosen to convert as much of the chloromethyl group as possible to the quaternary ammonium chloride, although experience indicates that some of the chloromethyl remains unreacted. A procedure devised by Barron and Fritz [1] uses concentrated hydrochloric acid and paraformaldehyde with a Lewis acid catalyst to chloromethylate the polymer. Exceptional control of the extent of chloromethylation is possible by adjusting the concentration of reagents, the reaction temperature, and the reaction time. Chloromethyl methyl ether is not generated in situ, except for possible traces, so the reaction is relatively safe to use. Following the chloromethylation step, amination is carried out by adding a large excess of 25% trimethylamine in methanol or water and allowing the reaction to proceed overnight. Depending on the conditions chosen, this procedure can produce anion exchangers with capacities from 0.005 to 0.16 mequiv g–1 when using XAD-1 as a substrate. This range includes the capacities most useful in ion chromatography.

3.3 Resins and Columns for Ion Chromatography

A large variety of columns for IC is available. Particle size of the resin column packing has steadily decreased from about 10 lm to 5 lm, and now to 3 lm. Resins are spherical and have a narrow particle size range. Column lengths are typically 5, 10, 15 or 25 cm. Column diameter is generally 4.0–4.6 mm, although columns 2.0 mm in diameter are also available. Both anion- and cation-exchange columns are available with a choice of resin matrix (polymeric or silica) and ionic functional group. Modern IC resins often contain imbedded polar groups in addition to the ion-exchange groups. The purpose is to reduce the hydrophobic nature of the solid ion exchanger and to modify the nonionic attraction of analyte ions for the solid phase. The effects of resin composition on anion selectivity will be discussed Section 3.4. 3.3.1 Monolith Columns

Enhanced chromatographic efficiency is obtained by using a column packed with very small ion-exchange particles. However, the pressure required to push eluent through a packed column increases rapidly with decreased particle size. A new and different type of chromatographic column called a ‘monolith’ offers significant advantages to traditional packed columns. A monolith support is a single polymeric rod that can be prepared by pouring a chemical mixture into a column and carrying out the polymerization in place. The monolith structure is designed to contain an uninterrupted, interconnected network of channels of a controlled

43

44

3 Resins and Columns

size range. The back pressure is much lower than with a packed column with particles of similar size. Svec [2] has reviewed the preparation of monoliths for chromatography. Monoliths are described as separation media in a format that can be compared with a single large particle that does not contain any interparticular voids. As a result, all of the mobile phase must flow through the stationary phase. This convective flow greatly accelerates the rate of mass transfer. In contrast to diffusion, which is the driving force for mass transfer within the pores of particulate phases used in packed columns, convective flow through the pores enables a substantial increase in the speed of large biomolecules. In packed-column chromatography, interstitial voids – empty spaces between the packing particles – take up space in the column but do not aid in the separation. In an ideally packed column with equal-sized spherical particles, some 30% of the column volume is lost to voids [3]. As the mobile phase is pumped through the column during the course of a chromatographic run, the fluid flows freely through the voids but meets resistance as it permeates the interior of the porous packing materials. When a sample solution is injected into the column, differences between analyte concentrations in the void spaces and interiors of the particles cause the analytes to be transferred back and forth between the two regions. The time required for equilibration reduces chromatographic efficiency and can result in slow separations, especially for larger analytes which tend to move sluggishly because of their small diffusion coefficients. Using chemical methods to polymerize liquid precursors into a continuous porous mass of coalesced particles, two sets of parameters can be controlled simultaneously. The nature of the material, porosity, and other properties that affect separations can be optimized. The size of channels and open spaces can be controlled independently. Monoliths for chromatographic use can be described as spongy with micrometer-sized channels winding through a mass of fused particles [3]. Monoliths are based on either silica or organic polymers. Silica materials can be prepared by a sol-gel synthesis. The reaction is based on hydrolysis and polycondensation of tetramethoxysilane in the presence of polyethleneglycol, urea, and other reagents. Heating the product yields an alkaline solution, which etches tiny holes into the silica walls. One of the main concerns is ensuring that the monolithic material remains in intimate contact with the walls of the surrounding tube. One strategy for ensuring the requisite contact is to bond the material covalently to bond the growing monolith to the walls of a quartz tube via SiO2 groups. Much of the development work on application of monoliths to IC separations has been performed on reversed-phase monolithic HPLC columns, which became available well before monoliths with ion-exchange sites. These HPLC columns can be adapted for use in ion chromatography by coating with an ionic surfactant. Monolith anion- and cation-exchange columns are particularly advantageous for separation of bio ions, although small anions and cations can also be separated. Effective anion columns with high efficiency can be prepared simply by coating a nonionic column with a cationic surfactant, such as cetylpyridinium chloride (CPC) [4].

3.4 Anion Exchangers

Paull and co-workers [5, 6] used a very short (10 mm × 4 mm) coated silica monolith for the separation of inorganic anions. The back pressure was sufficiently low that a peristaltic pump could be used to pass eluent through the system. Pelletier and Lucy [7] were able to achieve rapid low-pressure chromatographic separations of anions on short monolithic columns. Reversed-phase silica columns 0.5 and 1.0 cm long were coated with a long-chain cationic surfactant to convert the column to an anion exchanger. Excellent separations of seven inorganic anions were obtained in 1–2 min. The thickness of the surfactant coating, and thus the exchange capacity of the column, can be adjusted by varying the percentage of acetonitrile in the predominately aqueous coating solution from 1 to 5%.

3.4 Anion Exchangers

The most widely used anion-exchange resins may be divided into two general types. Porous polymeric materials have quaternary ammonium functional groups throughout the resin bead although the concentration of exchange sites deeper within the bead tends to diminish. Latex agglomerated anion exchangers are pellicular materials with the exchange sites on latex particles coated in a relatively thin layer at the outer perimeter of the bead. 3.4.1 Porous Anion Exchangers

The major anion exchangers for ion chromatography are based on two substrate types: macroporous and microporous (or gel type) materials. Microporous resins were formerly popular because of their superior ion-exchange kinetics. However, microporous substrates can be ‘spongy'. Columns packed with this material may eventually undergo bed compression leading to reduced column performance. Macroporous materials are rugged, and column beds made from this substrate are stable. However, some commercial materials show poor ion-exchange kinetics with certain eluents. Gjerde [8] described a macroporous anion-exchange resin that shows good IC separations with a variety of eluents including sodium carbonate/bicarbonate and sodium hydroxide. This resin, called the Transgenomic AN1, is highly cross-linked Polystyrene Divinylbenzene with 80 Å pores and 415 m2 g–1 surface area. The commercial material has an exchange capacity of 0.05 mequiv g–1 and an average particle size of 8 mm. The resin contains a quaternary ammonium functional group: dimethylethanolamine. The substrate polymer beads used in this work are quite hard and do not swell and shrink when the solvent is changed. A measure of bead hardness is given by the swelling propensity value. This value is obtained by measuring the backpressure of a column with a tetrahydrofuran mobile phase and then with an aqueous

45

46

3 Resins and Columns

phase. After correcting the eluent backpressure for viscosity, the swelling propensity, SP, is calculated by the following equation: SP ˆ

THFpressure H2 Opressure H2 Opressure

An SP value of zero indicates nonswelling material. The substrate used in this work had an SP of 0.8. Many PS-DVB resins have SP values well above 1.0. Columns packed with the AN1 resin have given excellent separations of common anions using a variety of eluents. Polymeric resins from Hamilton have been used extensively for anion chromatography. Their PS-DVB anex resins contain trimethylammonium groups. PRP X100 and PRP X110 have exchange capacities of 0.19 and 0.11 mequiv g–1 respectively. Their RCX10 has a somewhat higher capacity: 0.35 mequiv g–1. All of these have an average pore size of 100 Å. Good separations of anions have been obtained with a variety of eluents [9]. However, these resins are quite hydrophobic. For this reason, the eluents often contain ∼7.5% methanol or ∼0.1 mM sodium thiocyanate to give better peak shapes. A resin from Alltech, sold commercially as Durasep A1, is another example of a PS-DVB anion exchanger [10]. A highly cross-linked backbone makes this material chemically and mechanically stable. It withstands organic solvents and is stable over a wide range of pH, temperature and pressure. The resin is made of polydivinylbenzene with dimethylethanolamine functional groups and was designed for use with both nonsuppressed and suppressed conductivity detectors. The material is highly cross-linked and has both micro and macro pores for efficient mass transport. It was shown that resolution and peak shape are improved by adding 5–15% methanol to the eluent. In particular, methanol reduced or eliminated the hydrophobic interaction between nitrate and the resin, resulting in a symmetrical peak shape. However, addition of acetonitrile to the eluent seemed to increase the hydrophobicity of the column toward nitrate. A wide variety of resins based on polyacrylate polymers has been produced for use in chromatography. A type known as HEMA, a macroporous copolymer of 2-hydroxyethyl methylmethacrylate and ethylene dimethacrylate, has been used extensively in ion chromatography. It is extensively cross-linked to produce a polymeric matrix with high chemical and physical stability. The structure of HEMA is shown in Figure 3.2A. The tertiary carbonyl structure of pivalic acid is one of the most stable and least hydrolyzable esters known, which allows the HEMA stationary phase to be used with a variety of eluents in the pH range 2–12. The excess hydroxyl groups on the HEMA matrix also increase the hydrophilicity of this material, which will be shown later to result in improved peak shapes for polarizable anions. The strong-base anion exchanger of HEMA, shown in Figure 3.2B, is prepared by treating the HEMA precursor with an aqueous solution of trimethylamine. The preparation procedures and the influence of different functional groups on sorbent selectivity were discussed by Vlacil and Vins [11].

3.4 Anion Exchangers

Figure 3.2 Structures of (A) HEMA and (B) strong-base anion exchanger of HEMA.

The preparation and properties of polyacrylate anion-exchange resins by Alltech has been described as a universal stationary phase for the separation of a wide variety of anions [6, 7]. The Allsep anion column contains 7-lm particles packed into columns of various lengths. The resin is methacrylate based with quaternary ammonium functional groups The A2 anion column also has methacrylate, but has a quaternary amine with alkenol, rather than alkyl, groups. It is relatively hydrophilic and resolves acetate and formate from fluoride and chloride. Both columns have a broad pH range (pH 2–11) and can be used with 0–100% of an organic eluent modifier. 3.4.2 Effect of Functional Group on Selectivity

The polymeric matrix, the chemical type and structure of the ion-exchange groups, and solvation effects all have a significant effect on determining the selectivity of anion exchangers for competing ions [12]. Variations in the chemical structure of exchange sites have been particularly effective in producing resins with greater selectivity for certain ions.

47

48

3 Resins and Columns

Okada [13] compared the effects of –NH3+ and -NR3+ groups in anion exchangers with the same polymeric matrix and almost the same exchange capacity (0.4 mmol g–1). The –NH3+ resin had a more concentrated charge and a stronger electrostatic field than the resin with –NR3+, where the positive charge was more dispersed. Going from –NH3+ to –NR3+ resulted in decreased electrostatic and hydrogen bonding interactions with sample anions and increased ion-induced dipole and London dispersion interactions. The latter two effects resulted in preferable binding of larger ions by –NR3+. As an example, the ratio of retention factors of ClO4–:Cl– was 17.4 for the resins with –NR3+ but only 1.89 for those with –NH3+ groups. Anion-exchange resins containing a benzyltrimethylammonium functional group are a widely used type for anion chromatography. It was of interest to see how changing the chemical nature of the quaternary ammonium functional group might affect the selectivity of anion exchangers toward different anions. Barron and Fritz [14] prepared 13 different resins by reacting chloromethylated XAD-1 with different tertiary amines. To ascertain the effect of functional group structure on selectivity, the various resins should have a very similar exchange capacity so that identical elution conditions could be used for each resin. To accomplish this, the relative reactivities of various amines with chloromethylated XAD-1 had to be determined. Then the degree of chloromethylation of XAD-1 could be adjusted so that the aminated resins would have similar capacities. The retention times of 17 monovalent anions on resins with different functional groups but with almost identical exchange capacities (average: 0.027 mequiv g–1) were compared with the use of a solution of a monovalent anion (sodium benzoate) as the eluent [14]. Relative retention times were calculated by dividing the measured retention times by that of chloride. Data for resins with various trialkylammonium groups are presented in Table 3.1. The data show that the relative retentions of the weak-acid anions are almost independent of the size of the alkyl groups. However, as the size of the R groups increases, large changes occur with the more polarizable anions such as nitrate, iodide, chlorate and BF4– ions. The use of anion-exchange resins with different substituents offers a useful parameter for improving the separation of some anions. For example, Figure 3.3 compares the separation of five anions on columns containing resins of approximately the same capacity but with increasingly larger alkyl groups on the quaternary nitrogen. Identical elution conditions were used. The TMA resin column gave poor resolution of bromide and nitrate. The resolution was improved on the TPA column, and a baseline separation was obtained on the THA column. Most of the monovalent anions exhibit only small changes in their relative retention on resins containing one, two, or three hydroxyethyl groups, compared to the trimethylamine resins (Table 3.1). However, when a stronger eluent was used (phthalate, with two negative charges instead of benzoate, with a single negative charge) and several divalent anions were examined, the changes resulting from hydroxyethyl groups became more apparent [15].

3.4 Anion Exchangers Table 3.1 Relative retentions of anions on trialkylammonium

resins [24]. Anion

TMA

TEtA

TPA

TBA

THA

TOA

Cl–

1.0

1.0

1.0

1.0

1.0

1.0

0.66

0.70

0.69

0.71

0.68

0.69

Br

1.20

1.19

1.25

1.34

1.32

1.41

–

2.51

2.48

3.05

3.82

5.00

>5.0

H2PO4–

0.84

0.85

0.84

0.85

0.83

0.83

NO2–

0.82

0.82

0.86

0.90

0.89

0.98

–

1.30

1.32

1.38

1.54

1.63

1.72

Acetate

0.25

0.28

0.23

0.25

0.22

0.22

Formate

0.52

0.54

0.51

0.52

0.51

0.53

Lactate

0.44

0.49

0.46

0.47

0.45

0.49

Glycolate

0.45

0.48

0.44

0.45

0.43

0.45

Nicotinate

0.31

0.32

0.31

0.31

0.31

0.33

–

1.53

1.55

1.56

1.73

1.92

2.15

–

1.05

1.06

1.03

1.08

1.08

1.14

0.36

0.39

0.34

0.37

0.35

0.37

2.70

2.58

3.41

4.34

>7.5

–

1.00

1.06

1.01

1.06

1.01

1.07

8.3

8.0

9.5

8.9

–

F

–

I

NO3

ClO3

BrO3 N3– BF4

–

CH3SO3

–

tRCl(min)

10.6

9.4

TMA = trimethylamine, TEtA = triethylamine, TPA = tripropylamine, TBA = tributylamine, THA = trihexylamine, and TOA = trioctylamine.

The data in Table 3.2 show that methyldiethanolamine (MDEA) considerably lowers the relative retention of nitrate, chlorate, iodide and thiocyanate compared to the trimethylamine (TMA) resin. It also shows that with the phthalate eluent the tributyl resin (TBA) has longer retention times for nitrate and much longer for iodide, but shorter for sulfate and thiosulfate, all compared to the TMA material.

49

50

3 Resins and Columns

Figure 3.3 Separation of five anions on three different resins of similar capacity. The resins were packed in a 500 mm × 2.0 mm i.d. column and a solution of 0.0001 M benzoic acid was used as the eluent at a flow rate of 0.93 mL min–1 (From Ref. [14], with permission).

Virtually all anion-exchange separations have been carried out with resins containing a single nitrogen atom in each exchange group, either a quaternary ammonium group (– N+R3) or a protonated amine group (– N+HR2). A novel resin has been described, containing three nitrogen atoms in each functional group [16]. A chloromethylated PS-DVB resin was reacted with diethylenetriamine to give a functional group of structure a, or b, or a mixture of the two (see below).

3.4 Anion Exchangers Table 3.2 Relative retentions of ions on three anion exchangers

of differing polarity. Phthalate eluent (0.4 mM), pH 5.0. Ion

MDEA*

TMA**

TBA***

Cl–

1.0

1.0

1.0

NO3–

1.47

1.79

2.77

ClO3–

1.78

2.42

3.18

3.81

6.16

I

–

13.9

–

7.50

14.5

–

–

9.12

–

–

SO42–

7.31

S2O32–

15.75

C2O42–

7.41

7.18

6.30

9.50

9.92

8.72

SCN

ClO4

MoO4 WO4

2–

2–

tRCI- (min)

– 3.20

7.29 16.6

– 3.80

6.36 9.23

– 3.90

Values are expressed at tRanion/ tRCI* Capacity 0.090 mequiv g–1 ** Capacity 0.092 mequiv g–1 *** Capacity 0.096 mequiv g–1

By varying the pH at which the resin is used in an IC column, one, two or three of the N atoms can be protonated giving a net charge of 1+, 2+ or 3+ for each functional group. The retention times of sample anions become longer as the operating pH becomes more acidic and the net positive charge on the ion exchanger increases. Figure 3.4 plots the retention factor as a function of eluent pH for several sample anions. Thiocyanate and molybdate are very strongly retained, even at moderately acidic pH values. Several anions were separated chromatographically at pH 7.5–7.7 with different salts in the mobile phase and with direct UV detection at 200 nm. The results in Table 3.3 show that perchlorate is a significantly better eluting anion than chloride. However, sulfate and hydrogen phosphate both give even shorter retention times by virtue of their 2– charge.

51

52

3 Resins and Columns

Figure 3.4 Eluent pH vs. anion capacity factor, k′. Conditions: 100 mm × 4.6 mm (multicharge, weak base, anion exchanger) column, 15 mM sodium perchlorate eluent, UV detection at 200 nm (from Ref. [16], with permission). Table 3.3 Retention times (min) of several anions with different eluents.

Anion

NaClO4a

NaSO4a

NaCla

Na2HPO4a

Bromide

3.45

2.15

4.51

1.98

Nitrate

3.98

2.75

5.55

2.54

Iodide

6.25

7.01

12.4

5.95

Thiocyanate

13.7

NDb

25.6

17.4

a Each eluent: 5.0 mM at pH 7.5–7.7. b ND = not detected.

3.4.3 Effect of Spacer Arm Length

Polymeric anion exchangers are normally prepared by chloromethylation of the benzene ring followed by reaction with a tertiary amine to give a quaternary ammonium group. Thus the N+ is connected to the benzene ring by a single –

3.4 Anion Exchangers

CH2 group. Suppose the N+ was connected to the benzene ring by a longer series of –CH2 groups, sometimes called the spacer arm. What effect would a longer spacer arm have on ion-exchange selectivity? Warth and Fritz synthesized a series of resins with spacer arms of varying lengths [17]. The benzene ring of a macroporous resin (Rohm & Haas XAD-1) was reacted under controlled conditions with a bromoalkene with CF3SO3H as catalyst and was then quaternized by reaction with trimethylamine. This gave a resin of the following structure:

Conditions were adjusted so that all of the resins had almost identical exchange capacities. The spacer arm length varied from one to six methylene groups. Chromatographic retention times of a number of anions were compared, with 6.0 mM nicotinic acid and 2.0 mM phthalate (pH 6) as the eluents. In many cases, the length of the spacer arm had very little effect on the relative retention times (relative to Cl–). However, the selected data in Table 3.4 show that the relative retention times of bromide, nitrate, chlorate and iodide decreased, while that of sulfate increased slightly.

Table 3.4 Adjusted retention times as a function of spacer arm

length of anion exchange resins. Spacer arm length Anion

C1

C2

C3

C4

C6

Chloride

1.0

1.0

1.0

1.0

1.0

Nitrite

1.8

1.5

1.5

1.4

1.3

Bromide

2.5

2.1

2.0

1.8

1.9

Nitrate

3.6

3.1

2.5

2.4

2.6

Chlorate

7.1

4.5

3.9

3.7

4.1

Sulfate

7.4

8.3

8.5

8.5

8.5

Iodide

17.6

12.2

8.9

8.6

11.2

Longer spacer arms would of course reduce any influence the benzene ring might have on ion-exchange retention.

53

54

3 Resins and Columns

3.4.4 Latex Agglomerated Ion Exchangers

Pellicular materials in which the stationary phase is a layer on the outside perimeter of a spherical substrate have been used frequently in liquid chromatography. The relatively thin layer ensures a rapid equilibrium between the mobile and stationary phases even when the column packing has a relatively large particle size. In their original work on ion chromatography, Small et al. [18] used a surface-sulfonated material as a pellicular cation exchanger. Such materials are easy to prepare because sulfonation of a polymer containing benzene rings proceeds from the outside in. Sulfonation for a short period under mild conditions will insert sulfonic acid groups only on or near the outside surface of a spherical resin. It was soon discovered that efficient anion-exchange resins could be prepared by coating the outside of a surface-sulfonated polymer with a layer of latex particles functionalized with quaternary ammonium groups. The first commercial anion exchangers for ion chromatography (Dionex ASI) consisted of 0.15-lm latex particles coated onto a 25-lm sulfonated substrate. A two-dimensional diagram of this coating is shown in Figure 3.5A. The positively charged latex particles are firmly held by electrostatic attraction as shown in Figure 3.5B. Each latex particle has several quaternary N+ groups, so the coated substrate will have many quaternary groups available for ion exchange. Latex agglomerated resins are very stable chemically. Even 4 M sodium hydroxide is unable to cleave the ionic bond between the substrate and the latex bead.

Figure 3.5 Latex-coated anion exchanger (Courtesy Dionex Corp.)

Several advantages have been claimed for latex-coated anion exchangers [19]: . The substrate provides mechanical stability and gives a moderate back pressure. . The small size of the latex beads and their location on the outer surface of the substrate ensure fast exchange processes and thus a high chromatographic efficiency. . Swelling and shrinkage are minimal.

3.4 Anion Exchangers

The properties of latex resins can be varied by manipulation of several parameters. Hydrophobic attraction of the exchanger for some anions can be altered by varying the type and cross-linking of the polymeric substrate. The ion-exchange capacity is determined by the substrate particle size, the size of the latex beads, and the degree of latex coverage on the substrate surface. Selectivity for various anions is governed mainly by the type of functional groups attached to the latex bead and by the degree of latex cross-linking. Over the years, Dionex has developed a wide variety of latex-agglomerated resins to meet various needs in IC. A review of these developments is given in a book by Weiss [19]. A method has been described for preparation of latex-coated anion-exchange resins that does not involve sulfonation of the substrate. A suspension of quarternized latex beads in water containing 0.01 to 0.10 M sodium chloride is used to coat an unfunctionalized polymeric substrate. Once coated, the latex sticks tightly and is not washed off by aqueous solutions. The latex can be removed by washing with pure organic solvents and may then be recoated. The exchange capacity may be varied by changing the concentration of sodium chloride or latex in the coating solution. Various resin substrates were coated with exchange capacities ranging from 5 to 400 mequiv g–1. Columns packed with these latex exchangers gave unusually efficient separations of sample anions. Extensive research effort has gone into the development of resins and columns for IC. This is perhaps best illustrated by a specific example: a stationary phase for the determination of fluoride and oxyhalides such as chlorite, chlorate and bromate [20]. The determination of fluoride has been a problem owing to its low affinity for strongly basic anion exchangers. Although carbonate–hydrogen carbonate eluents are widely used, fluoride elutes very close to the system void and its detection by suppressed conductivity at lower concentration levels is very difficult owing to interference from the negative water dip. The water dip occurs when the injected aqueous sample passes through the conductivity cell, decreasing the background conductivity. In addressing this problem, an anion-exchange column with high exchange capacity gave good resolution of fluoride and early-eluting anions when used with a dilute eluent, but the elution times for bromide and nitrate exceeded 40 min. A commercially available polymeric quaternary ammonium (quat)-coated column gave excellent resolution of fluoride and other common inorganic anions, but a rinse with acetonitrile-water (90:10) destroyed the separation ability of the column. A periodic rinse with an organic solvent is often needed to remove humic acid or other organic matter that gradually builds up on a column. The goal of the research undertaken was to develop a new solvent-compatible ion-exchange material with which fluoride is resolved from the system void volume and oxyhalides are separated from other acid anions in the same run under isocratic conditions. Because of the high degree of hydration of the fluoride ion, it was necessary to generate a solid phase with an extremely high water content. This suggested the use of very hydrophilic ion-exchange sites. Because fluoride is so highly hydrated, the only viable method of obtaining a reasonable fluoride

55

56

3 Resins and Columns

retention was to produce a latex coating with extremely low cross-linking. A crosslinking level of well under 17 leads to the generation of polymers with very high water content. The use of a functionalized latex with extremely low cross-linking creates a potential problem with regard to the exchange capacity of the coated resin. The surface area of standard microporous particles is inadequate to provide sufficient capacity when used with these low-cross-linked latexes. A resin with an acceptable capacity was finally produced by coating a super-porous polymeric substrate. The structural and physical properties of the final product (with the commercial designation ion Pac AS 12A) are given in Table 3.5.

Table 3.5 Structural and physical properties of the Ion Pac

AS12A separator. Parameter

Value

Column dimensions

200 mm × 4 mm i.d.

Particle diameter

9 lm

Substrate material

Macroporous polyethylvinylbenzene cross-linked with 55% divinylbenzene

Pore size

200 nm

Column capacity

52 l equiv.

Latex polymer

Vinylbenzyl chloride

Latex cross-linking

Very low (0.15%)

Latex diameter

14 nm

Functional group

quaternary ammonium group

pH stability

0–14

Solvent compatibility

0–100%

Two important conclusions may be drawn from these tables. One is that the hydroxide ion is a much weaker eluent for anion chromatography than carbonate. The second is that the eluting power of sodium hydroxide is enhanced considerably by using latexes with one or two hydroxyethyl groups instead of those containing only alkyl groups.

3.4.4.1 Effect of Latex Functional Group on Selectivity Slingsby and Pohl [21] investigated the effect of varying the structure of the quaternary ammonium group on the latex while keeping the percentage of cross-linking and the polymeric backbone structure constant. They estimated that each

3.5 Cation Exchangers

5-mm spherical substrate particle was coated with approximately 28 000 quaternized latex beads. Retention factors (k), corrected for column capacity, were measured for four different columns. The latex functional group in column 1 was methyldiethanolamine (MDEA), column 2 was dimethylethanolamine (DMEA), column 3 was trimethylamine (TMA), and column 4 was triethylamine (TEA). The eluent in Table 3.6 was 5 mM sodium carbonate.

Table 3.6 Retention factors (k) for different latex functional groups.

Eluent: 5 mM sodium carbonate. Column

F–

Cl–

MDEA

0.06

0.24

DMEA

0.14

TMA TEA

Br–

NO3–

ClO3–

SO42–

HPO42–

0.92

1.1

1.0

0.20

0.31

1.1

4.5

5.0

4.9

3.0

6.7

0.30

4.4

19.2

22.5

21.2

51.4

>100

0.30

5.8

26.1

55.8

24.0

24.9

>100

3.5 Cation Exchangers

The science and technology of ion chromatography is continuously evolving. Certain trends become apparent which are often pushed aside after a few years in favor of a different trend. This is particularly true in the case of resins for cation chromatography. Some of the milestones in the development of cation exchangers for IC will be traced in this section. 3.5.1 Sulfonated Resins

Most of the cation exchangers used in IC fall into two major categories: sulfonated resins, sometimes called ‘strong-acid’ exchangers, and resins with carboxylic acid groups, sometimes called ‘weak-acid’ exchangers. The ion exchangers used in IC have a much lower exchange capacity than those intended for commercial applications such as the removal of calcium and magnesium ions from hard water. Low-capacity cation-exchange resins are obtained by superficial sulfonation of styrene-divinylbenzene copolymer beads. The resin beads are treated with concentrated sulfuric acid and a thin layer of sulfonic acid groups is formed on the surface. The final capacity of the resin is related to the thickness of the layer and is dependent on the type of resin, the bead diameter, and the temperature and time of contact with the sulfuric acid. Typical capacities range from 0.005 to 0.1 mequiv g–1.

57

58

3 Resins and Columns

It can be easily appreciated that, compared to a conventional cation-exchange resin, the diffusion path length is reduced because the unreacted, hydrophobic resin core restricts analyte cations to the resin surface. This results in faster mass transfer of the cations and consequently in improved separations. Also, because of the rigidity of the resin core, there is less tendency for the bead to compress. This means that higher flow rates (at relatively low back pressures) can be used than would be possible with conventional resins. Superficially functionalized resins are stable over the pH range of 1 to 14 and swelling problems are minimal. The selectivity of the superficial cation-exchange resins for ions is similar to that observed for conventional resins. For chromatographic purposes it is necessary to control the conditions for sulfonation so that the polymer beads are sulfonated evenly and the exchange capacity is very reproducible. It is also desirable to be able to increase or decrease the resin capacity in a predictable fashion. Pepper [22] proposed that the sulfonation of neutral poly(styrene-divinylbenzene) copolymer beads proceeded in a ‘layer-by-layer’ fashion and could be stopped at any particular depth. From this concept, Pepper [23] prepared the first superficially sulfonated polymer beads. Parrish [24] also prepared surface-sulfonated beads and gave a brief example of their utility. Fritz and Story [25, 26] prepared several low-capacity sulfonated resins for the chromatographic separation of various metal cations. Macroporous resins were used and their selectivity was somewhat different from that of conventional gel resins. The surface sulfonation of poly(styrene–divinylbenzene) beads with cross-linking ranging from 0.5 to 8% divinylbenzene was studied by Small [27, 28]. Examination of the sulfonation depth of a resin bead in terms of optimum separation of several inorganic cations was subsequently studied by Stevens and Small [29]. These resins were used for the chromatographic separation of simple cations. Low-capacity resins have been prepared specifically for use in ion chromatography by Fritz et al. [30]. The resin used in the separations was a 3:2 blend of neutral resin with low-capacity sulfonated resin giving a final capacity of about 8 lequiv g–1. Very good efficiency was demonstrated for separation of the alkali cations and alkaline earths. Papanu et al. [31] fabricated a cation-exchange resin by agglomerating a sulfonated latex onto larger beads of a low-capacity anion-exchange resin. Battaerd [32, 33] prepared a superficially sulfonated bead by graft polymerization of a sulfonated olefin onto a polyolefin core. Kirkland [34] impregnated a porous fluoropolymer bead with a sulfonated fluoropolymer, giving a pellicular type of ion exchanger. Horvath et al. [35] polymerized a coating of polystyrene-DVB onto glass beads and formed a cation exchanger by sulfonation. Several authors have prepared cation exchangers by introduction of sulfonated organic group onto the surface of silica supports [36–38]. Sevenich and Fritz [39] examined the sulfonation of resins for use in IC in some detail. Spherical microporous resin beads of 4%, 6.5% and 12% cross-linking were selected for their study. Initially, sulfonation of 4% cross-linked resin

3.5 Cation Exchangers

beads under identical conditions gave poor reproducibility. This was partly due to agglomeration of the small resin beads. Dispersion of the resin beads and even initial wetting of the surface by sulfuric acid seemed to be major problems in achieving reproducible sulfonation. The procedure finally developed involved mechanical and ultrasonic dispersion of the particles in methanol, passing the resin slurry through a small sieve, then removal of as much methanol as possible by suction filtration. The resin was then sulfonated. The small amount of methanol present when the resin is added to the hot sulfuric acid immediately volatilizes and is swept out in the inert gas stream. This is confirmed by a short burst of vapor issuing from the reaction vessel when the resin is added. The resin capacity is approximately linear with reaction time and the reproducibility was ±5%. Sulfonation of gel beads of 4%, 6.5% and 12% cross-linking was compared. Data for sulfonation of each for 30, 60 and 90 min are given in Table 3.7. These results show that resins of low capacity can be obtained in all cases but that the reaction is better controlled with resin beads of higher cross-linking.

Table 3.7 Retention factors (k) for different latex functional groups.

Eluent: 5 mM sodium carbonate. Column

F–

Cl–

MDEA

0.06

0.24

DMEA

0.14

TMA TEA

Br–

NO3–

ClO3–

O42–

HPO42–

0.92

1.1

1.0

0.20

0.31

1.1

4.5

5.0

4.9

3.0

6.7

0.30

4.4

19.2

22.5

21.2

51.4

>100

0.30

5.8

26.1

55.8

24.0

24.9

>100

The location of sulfonate groups in the resin bead was visualized by completely replacing H+ with UO22+ as the counterion and obtaining transmission electron micrographs on thin slices of the resin bead. Because uranium is a very heavy metal, the uranyl ions have higher stopping power for electrons and appear as a darker area on the micrograph. A dark outline around the resin slice indicated that the uranyl ions (and hence the sulfonate groups) are located in a thin zone (approximately 200 Å) at the outer perimeter of the resin bead. Knowing the resin capacity and the estimated thickness of the sulfonated layer, a simple calculation shows nearly complete sulfonation within the sulfonated layer. By ‘complete’ we mean that there is approximately one sulfonate group for each benzene ring of the polymer. The density of sulfonate groups in this layer is similar to that in the entire bead of a typical high-capacity cation-exchange resin. Columns packed with the 12% cross-linked resin (6.1 mequiv g–1 exchange capacity) gave good separations of metal cations. Using 8 different concentrations of perchloric acid eluents, from 0.10 M to 1.00 M, retention times for 36 metal ions

59

60

3 Resins and Columns

were measured [40]. The selectivity data obtained are given in Tables 5.3 and 5.4 in Chapter 5. The Dionex Co. has developed a number of latex-coated cation-exchange materials. Their anion-exchange resins have a surface layer of quaternary ammonium latex on a surface-sulfonated substrate. Addition of a second coating layer of sulfonated latex beads provides an outer layer of resin consisting of latex beads with exposed sulfonate groups. These groups undergo cation exchange with sample and eluent cations in IC. A schematic representation for these cation exchangers is: Resin–SO3– N+R3 latex N+R3 SO3–-latex- SO3–

(3.7)

A diagram of such a resin (Ion Pac CS3) indicates that the sulfonated latex beads are of larger diameter than the quaternized beads (Figure 3.6). This configuration of the column packing permitted the use of smaller size substrate beads, provided a shorter mean free path for the analytes and greatly improved peak efficiencies for cationic analytes. Although Group I and Group II cations could be separated simultaneously, the elution time was long (about 28 min) unless some form of gradient elution was employed.

Figure 3.6 Schematic representation of IonPac CS3, a latexcoated strong-acid cation exchanger (Courtesy Dionex Corp.)

A few years later (1990), a sulfonated cation-exchange column (Ion Pac CS10) with decreased cation-exchange site density was developed. This lower exchange capacity was obtained by adding a monomer that is deactivated by sulfonation to the latex polymerization mixture. The substrate of the column packing was 50% cross-linked, instead of 4% as with previous materials, in order to obtain better compatibility with organic solvents. Six common inorganic cations could be separated isocratically in about 15 min [41].

3.5 Cation Exchangers

Several improved versions of sulfonated latex columns were developed over the next few years [41], but this type of cation-exchange column was abandoned in favor of materials with a weak-acid function. Catex columns with sulfonic acid functionalities have a relatively low selectivity for hydronium ions. A divalent cation component must be added to the eluent to efficiently elute divalent cations such as magnesium and calcium. 3.5.2 Weak-acid Cation Exchangers

Owing to their differences in selectivity, it is often difficult to find conditions for separation of cations of different positive charge on a sulfonated resin column. Eluents that provide good separation of monovalent cations are too weak to elute divalent cations in a reasonable time. There is now a trend to use weak-acid cation-exchange columns. These materials contain carboxylic acid functional groups, or in some cases mixed carboxylic acid and phosphonic acid groups. At more acidic pH values these groups are gradually converted from the ionic to the molecular form, and thus their ability to retain sample cations is diminished. By adjusting the operating pH to an appropriate value it becomes possible to separate a wider variety of cations in a single run. One company advertises their chromatographic columns of this general type as a Universal Cation Column. The properties and performance of a commercial weak-acid resin column (Dionex CS12) have been described [42]. The substrate is a highly cross-linked, macroporous ethylvinylbenzene–divinylbenzene polymer with a bead diameter of 8 lm, a pore size of 6 nm, and a specific surface area of 300 m2 g–1. In a second step, this substrate was grafted with another polymer containing carboxylate groups. The exchange capacity is listed as 2.8 mequiv per column for a 250 mm × 4 mm i.d. column. With this column, simple eluents such as hydrochloric or methanesulfonic acid can be used to separate mono- and divalent cations rapidly and efficiently under isocratic conditions. Morris and Fritz [42] described the preparation and chromatographic applications of two weak-acid resins that are easily synthesized and carry the exchange group on the cross-linking benzene ring of the resin or on a short spacer arm from the ring. The first resin (resin I) was prepared by reaction of a cross-linked polystyrene resin with succinic anhydride in a Friedel–Crafts reaction with aluminum chloride as the catalyst. The carboxyl groups are connected to the resin benzene rings by a three-carbon atom spacer arm, thus: –COCH2CH2CO2H. The second cation exchanger (resin II) was prepared by reaction of the resin with phenylchloroformate to give a phenyl ester attached to the resin benzene rings, thus: – CO–OC6H5. The ester groups were then hydrolyzed by refluxing for 1 h in a sodium hydroxide–ethanol solution to give the sodium salts of the carboxylate. The exchange capacity of resin I was 0.60 mequiv g–1 and that of resin II was 0.39 mequiv g–1. Resin II in particular gave excellent separations of divalent metal cations with a complexing eluent.

61

62

3 Resins and Columns

Most of the cation-exchange resins used today have a polymeric matrix and contain carboxylate, or both carboxylate and phosphonate functional groups. Carboxylic acid phases are used in a weakly ionized form with only a fraction of the ion-exchange sites actually available for retention of cations. Because of this, the overall functional capacity must be increased in order to get sufficient retention of inorganic cations and amines with commonly used eluent systems. In producing the Dionex CS 12A cation exchanger [41], a higher capacity was obtained by using a highly crosslinked, macroporous, spherical substrate with a high surface area. In this particular case, the substrate particle consisted of a 55% crosslinked ethylvinylbenzene–divinylbenzene core with a surface area in the region of 450 m2 g–1. A polymeric film containing a mixture of carboxylic acid and phosphonic groups was then applied to a thickness of 5–10 nm over the entire bead surface as illustrated in Figure 3.6. Because of the porous nature of the substrate, the applied coating also penetrates the bead so that exchange groups are distributed throughout the resin particle. The average particle size of the final ion exchanger was 8 lm and the pore size was 150 Å. Selectivity in cation chromatography involves more than just an electrostatic attraction of analyte cations for a negative charge at exchange sites. The entire carboxylate functional group (–CO2–) or phosphonate group (–PO32– or –PO3H–) affects cation selectivity. Selectivity can therefore be altered by introduction of phosphonic acid groups into carboxylic acid ion exchangers in varying ratios. An example is shown in Figure 3.7. The resin contained a high ratio of phosphonic acid to carboxylic acid groups. The manganese peak has moved out beyond the calcium peak in going from a 1:5 ratio to a 1:1 ratio of phosphonate to carboxylate groups. Although not shown, calcium and strontium co-elute with the 1:1 phase, which is impractical because significant amounts of strontium are commonly found in groundwaters. The final product in the CS 12A exchanger actually contains a slightly lower amount of phosphonate than the upper chromatogram in Figure 3.7.

Figure 3.7 Schematic of the Ion Pac CS12A, a weak-acid cation exchanger (Courtesy Dionex Corp.)

3.6 Other Resins

3.5.3 Other Types

Considerable interest has been shown in a novel cation exchanger first developed by Schomburg et al. [43]. The material consists of a silica substrate of very uniform particle size coated with a poly(butadiene–maleic acid) copolymer which serves as the cation-exchange moiety. Maleic acid has two acidic dissociation constants (pK1 = 2.0, pK2 = 6.3), and this retains cation-exchange properties down to a fairly low pH. Analyte cations may be eluted with acidic or complexing eluents using commercially available columns [44, 45]. As an example, lithiuim, sodium, ammonium, potassium, magnesium and calcium are readily separated in a single run with 3 mM methanesulfonic acid as the mobile phase. Eluents containing a high concentration of common organic solvents can be tolerated [44]. Macrocyclic groups have been incorporated into resins to impart a different selectivity for metal cations than is possible with ordinary cation-exchange resins. A chemical structure known as 18-crown-6 has often been used. This a doughnutlike ring with six oxygen atoms connected together in via ethylene groups to form a ring with a cavity in the center. These ligands provide novel IC separations owing to the unusual specificity with which they bind cations of various ionic radii to form charged complexes. For example, the potassium ion fits into the cavity to form a stronger complex with the macrocycle oxygen atoms than does the sodium or ammonium ion. Potassium(I) elutes later from the cation-exchange column and is easily separated from sodium or ammonium ions. Application of macrocyclic ligands to ion chromatography has been discussed [46]. A practical column was created by absorbing tetradecyl-18-crown-6, which is a macrocycle with a long hydrophobic tail, onto a polystyrene-divinylbenzene substrate. When packed into a column, the macrocycle remains adsorbed to the resin [46]. The log K binding constants to 18-crown-6 in water for monovalent inorganic cations are as follows: Li+ = 0, Na+ = 0.8, Cs+ = 0.99, NH4+ = 1.23, Rb+ = 1.56, K+ = 2.03. Exactly the same elution order was observed for these same ions with the TD18C6 column [47].

3.6 Other Resins 3.6.1 Chelating Ion-exchange Resins

The selectivity of ordinary cation-exchange resins for various metal ions is somewhat limited. However, if a suitable chelating functional group is built into a polymeric resin, it often is possible to take up only a small group of metal ions. Other chelating resins may complex a larger group of metal ions, but selectivity is attained through pH control. Chelating resins also are valuable in sorbing a

63

64

3 Resins and Columns

desired metal ion (or small group of metal ions) from solutions containing a very high concentration of a noncomplexed metal salt. Frequently the selectivity of a chelating resin is so great that a very short column can be used to retain the desired metals. Although a great many chelating resins have been described, only a few have sufficiently fast rates of metal ion equilibrium and chromatographic efficiency for practical separations. Several chelating functional groups for resins are listed in Table 7.8. Chromatographic separations using chelating resins are discussed in Section 7.5. 3.6.2 Metal Oxides

Pietrzyk and coworkers [48] showed that hydrated alumina can function as a lowcapacity anion exchanger. The isoelectric pH for alumina is about 7.5 and its ionexchange capacity (about 2 mequiv g–1 maximum) is dependent on the sample ion, the eluent pH, and the pretreatment of the alumina. The alumina must be first hydrated and then treated with either an acid or base. Anion-exchange capacity increases as the eluent pH is made more acidic. The changeover from an anion exchanger to a cation exchanger is gradual and occurs in the vicinity of eluent pH 7.5. In anion chromatography the elution order of common anions is almost the reverse of conventional ion exchangers. Thus, using a Spherisorb A54 column with a 0.10 N acetate buffer at pH 6.50, the elution order was: perchlorate, iodide, bromide, chloride and bromate. Fluoride and phosphate were so strongly held by the alumina that elution was virtually impossible. The modification of silica gel with various metals is a simple and effective way to prepare ion exchangers that often have unique selectivities for analyte ions. Ohta et al. [49] described the preparation of a cation exchanger in which silica gel was first immersed in zirconium butoxide, Zr (OC4H9)4. Then the material was calcined (heated) at temperatures up to 1000 °C to form a silica–zirconia product. An excellent separation of all the alkali metal ions plus ammonium was obtained with 10 mM tartaric acid as the eluent. Divalent metal cations were strongly retained. 3.6.3 Multi-purpose Resins

Quite logically, new columns are developed to meet a real analytical need. Proliferation in the use of surfactants is a case in point. Surfactants have both hydrophilic and hydrophobic centers and are widely used in many industries because of their ability to reduce surface tension. Anionic surfactants, classified as alkanesulfonates, alkyl sulfates and alkylbenzenesulfonates, are commonly used in detergents, cleansing agents, cosmetics and hygienic products. Cationic surfactants are quaternary ammonium compounds, which are used in cosmetics, disinfectants, foam depressants, and textile softeners. Nonionic surfactants are also present in a

3.6 Other Resins

host of commercial products. Very large quantities of surfactants are discharged into the environment. Therefore, the determination of surfactants is important for product control and for environmental monitoring [50]. Chromatographic determination of surfactants can be performed by utilizing different separation modes, such as reversed-phase, ion-exchange and size-exclusion techniques. However, a mixed-mode stationary phase has been developed that is suitable for analyzing anionic, nonanionic and cationic surfactants in a single run [51]. The packing material of the commercially available column consists of 5-lm particles with a pore size of 120 Å and a specific surface area of 300 m2 g–1. The chemical surface of the particles consists of hydrophobic alkyl chains, tertiary amino groups and polar amide functional groups. Surfactant analytes are separated by mixed-mode and dipole–dipole interactions. Retention of neutral, cationic and anionic surfactants can be controlled independently by changing ionic strength, pH and organic solvent content of the mobile phase. A higher ionic strength in the mobile phase reduces the retention time (t) of anionic surfactants but gives somewhat longer t values for cationic analytes owing to stronger ion pairing with the eluent anion. The pH is an important parameter in optimizing a separation. A higher pH reduces the net positive charge on tertiary amine groups and gives lower t values for anionic surfactants. Cationic analytes have somewhat higher t values because there is less repulsion by protonated sites as the pH is increased. The eluent pH has relatively little effect on the retention of nonionic surfactants. Each of the three types of surfactant molecule has a large alkyl chain, and there is a strong hydrophobic component in its retention by the stationary phase. A higher percentage of organic solvent in the mobile phase reduces this attraction. Amide groups embedded in the stationary phase introduce a possible dipole– dipole contribution to the retention mechanism. In addition to these mechanisms, anionic surfactants are retained by ion exchange with the protonated tertiary amine groups. Cationic analytes are repulsed to some extent by the protonated amine sites and generally elute before anionic surfactants. 3.6.4 Ion-exchange Disks

How long does an IC column need to be for an effective separation? Separation of anions on monoliths 5–10 mm in length was described in Section 3.3.1. In some cases an anion-exchange disk can function as an effective ‘column’ even though the diameter of the disk is much greater than that of the mobile phase path through the disk. The growing use of short monolithic beds in the form of one or more disks placed in a tube has been reviewed [52]. Cellulose membranes similar to filter paper can be used to concentrate ions from aqueous samples without retaining most neutral solutes. The –CH2OH groups of the cellulose are functionalized to introduce either diethylaminoethyl groups (to prepare an anion exchanger), or sulfopropyl groups to prepare a cation exchanger.

65

66

3 Resins and Columns

The 3M Co. produces membrane disks containing embedded ion-exchange particles. Each disk is < 1 mm in thickness and a small disk of any convenient diameter can be cut from a larger sheet with a cork borer. The practical use of the small disks is illustrated by a 15-s separation of chromium(III), which is a 3+ cation, and anionic chromium(VI), which exists primarily as HCrO4– [53]. Two small extraction disks are placed one on top of the other in a plastic holder. Then a syringe containing the aqueous sample is attached to the holder and the sample is pushed through the disk at a rate of about 4–5 mL min–1. All of the chromium(VI) is retained on the top anion-exchange disk and the chromium(III) is extracted by the second cation-exchange disk. The concentrations on each disk are several hundredfold higher than they were in the original sample. The amounts of chromium(III) and (VI) extracted were measured directly on the surface of the respective disks by diffuse reflectance spectroscopy (DRS).

References [1] R. E. Barron and J. S. Fritz, Reproduc-

[2]

[3]

[4]

[5]

[6]

[7]

[8]

ible preparation of low-capacity anion exchange resins, Reactive Polymers, 1, 215, 1983. F. Svec, Porous monoliths: emerging stationary phases for HPLC and emerging methods, LC-GC electronic ed., Vol. 1, No. 1, June 23, 2004. M. Jacoby, Monolithic chromatography, Chemical and Engineering News, Dec. 11, 2006, p. 14. J. Li, Yan Zhu and Y. Guo, Fast determination of anions on a short coated column, J. Chromatogr. A, 118, 46, 2006. D. Victory, P. Nesterenko and B. Paull, Low-pressure gradient micro-ion chromatography with ultra-short monolithic anion exchange columns, Analyst, 129, 700, 2004. D. Connoly, D. Victory and B. Paull, Rapid, low pressure and simultaneous ion chromatography of common inorganic anions and cations or short permanently coated monolithic columns, J. Sep. Sci., 27, 912, 2004. S. Pelletier and C. A. Lucy, Achieving rapid low-pressure ion chromatography separations on short silica-based monolithic columns, J. Chromatogr. A, 118, 12, 2006. D. T. Gjerde, New macroporous stationary phase for the separation of anions, Advances in Chromatography, vol. 2,

p 169, Century International, Medfield, MA 1990. [9] R. E. Barron and J. S. Fritz, Effect of functional group structure and exchange capacity on the selectivity of anion exchangers for divalent anions, J. Chromatogr., 316, 201, 1984. [10] L. M. Nair, B. R. Kildew and R. SaariNordhaus, Enhancing the anion separations on a polydivinylbenzene-based anion stationary phase, J. Chromatogr. A, 739, 99, 1996. [11] F. Vlacil and I. Vins, Modified hydroxyethyl methacrylate copolymers as sorbents for ion chromatography, J. Chromatogr., 391, 133. 1987 [12] J. S. Fritz, Factors affecting selectivity in ion chromatography, J. Chromatogr. A, 1085, 8, 2005. [13] T. Okada, Nonaqueous anion-exchange chromatography. I. Role of solvation in anion-exchange resin, J. Chromatogr. A, 758, 19, 1997. [14] R. E. Barron and J. S. Fritz, Effect of functional group structure on the selectivity of low-capacity anion-exchangers for monovalent anions, J. Chromatogr., 284, 13, 1984. [15] R. E. Barron and J. S. Fritz, Effect of functional group structure and exchange capacity on the selectivity of anion exchangers for divalent anions, J. Chromatogr., 316, 201, 1984.

References [16] L. Li and J. S. Fritz, Novel polymeric res-

[30] J. S. Fritz, D. T. Gjerde and R. M. Becker,

ins for anion-exchange chromatography, J. Chromatogr., 793, 231, 1998. [17] L. M. Warth and J. S. Fritz, Effect of length of alkyl linkage on selectivity of anion exchange resins, J. Chromatogr. Sci., 26, 630, 1988. [18] H. Small, T. S. Stevens and W. G. Bauman, Novel ion-exchange chromatographic method using conductimetric detection, Anal. Chem., 47, 1801, 1975 [19] J. Weiss, Ion Chromatography, 2nd Ed., p 43, VCH, Weinheim, Germany, 1995 [20] J. Weiss, S. Reinhard, C. Pohl, C. Saini and L. Narayaran, Stationary phase for the determination of fluoride and other inorganic anions, J. Chromatogr. A, 706, 81, 1995. [21] R. W. Slingsby and C. A. Pohl, Anionexchange selectivity in latex-based columns for ion chromatography, J. Chromatogr., 458, 241, 1988. [22] R. W. Pepper, Chemistry Research, 1952, p. 77, Her Majesty’s Stationary Office, London, England, 1953. [23] K. W. Pepper, Sulphonated cross-linked polystyrene: A monofunctional cationexchange resin, J. Appl. Chem., 1, 124, 1951. [24] J. R. Parrish, Superficial ion-exchange chromatography, Nature, 204, 402, 1965. [25] J. S. Fritz and J. N. Story, Selectivity behavior of low-capacity, partially sulfonated macroporous beads, J. Chromatogr., 90, 267, 1974. [26] J. S. Fritz and J. N. Story, Chromatographic separation of metal ions on lowcapacity macroreticular resins, Anal. Chem., 46, 825, 1974. [27] H. Small, Solvent extraction process for the recovery of uranium and rare earth metals from aqueous solutions, U.S. Patent 3 102 782, 1962. [28] H. Small, Gel liquid extraction. The extraction and separation of some metal salts using tri-n-butylphosphate gels, J. Inorg. Nucl. Chem., 18, 232, 1961. [29] T. S. Stevens and H. Small, Surface sulfonated styrene divinyl benzene – optimization of performance in ion chromatography, J. Liq. Chromatogr., 1, 123, 1978.

Cation chromatography with a conductivity detector, Anal. Chem., 52, 1519, 1980. S. Papanu, C. Pohl and A. Woodruff, New high speed cation exchange columns for ion chromatography, Paper presented at the Pittsburgh Conference and Exposition on Analytical Chemistry and Applied Spectroscopy, Atlantic City, NJ, U.S.A.1984. H. A. Battaerd, Core-dual shell graft copolymers with ion exchange resin shells, U.S. Patent 3 565 833, February 23, 1971. H. A. Battaerd and R. J. Siudak, Synthesis and ion-exchange properties of surface grafts, J. Macromol. Sci. Chem., A4, 1259, 1970. J. J. Kirkland, Superficially porous chromatographic packing with sulfonated fluoropolymer coating and chromatographic packing with chemically bonded organic stationary phases, U.S. Patents 3 577 266, May 4, 1971, and 3 722 181, March 27, 1973. C. G. Horvath, B. A. Preiss and S. R. Lipsky, Fast liquid chromatography: An investigation of operating parameters and the separation of nucleotides on pellicular ion exchangers, Anal. Chem., 39, 1422, 1967. C. Horvath and S. R. Lipsky, Column design in high pressure liquid chromatography, J. Chromatogr. Sci., 7, 109, 1969. D. C. Locke, J. T. Schmermund and B. Banner, Bonded stationary phases for chromatography, Anal. Chem., 44, 90, 1972. D. H. Saunders, R. A. Barford, P. Magidman, L. T. Olszewski and L. T. Rothbart, Preparation and properties of a sulfobenzylsilica cation exchanger for liquid chromatography, Anal. Chem., 46, 834, 1974. G. J. Sevenich and J. S. Fritz, Preparation of sulfonated gel resins for use in ion chromatography, Reactive Polymers, 4, 195, 1986. G. J. Sevenich and J. S. Fritz, Metal ion selectivity on sulfonated cationexchange resins of low capacity, J. Chromatogr., 371, 361, 1986.

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

67

68

3 Resins and Columns [41] D. Jensen, J. Weiss, M. A. Rey and

[42]

[43]

[44]

[45]

[46]

[47]

C. A. Pohl, Novel weak-acid cationexchange column, J. Chromatogr., 640, 65, 1993. J. Morris and J. S. Fritz, Ion chromatography of metal cations on carboxylic acid resins, J. Chromatogr., 602, 111, 1992. P. Kolla, J. Köhler and G. Schomburg, Polymer-coated cation-exchange stationary phases on the basis of silica, Chromatographia, 23, No. 7, 465, 1987. L. M. Nair, R. Saari-Nordhaus and J. M. Anderson, Jr., Simultaneous separation of alkali and alkaline-earth cations on polybutadiene-maleic acid-coated stationary phase by mineral acid eluents, J. Chromatogr., 640, 41, 1993. L. M. Nair, R. Saari-Nordhaus and J. M. Anderson, Jr., Ion chromatographic separation of transition metals on a polybutadiene maleic-acid-coated stationary phase, J. Chromatogr. A, 671, 43, 1994. R. Saari-Nordhaus, H. Pham and J. M. Anderson, Jr., A new cation stationary phase for challenging ion chromatography applications. Poster 2292, Pittcon 699, Orlando, FL. B. R. Edwards, A. P. Giauque and J. D. Lamb, Macrocycle-based column

[48]

[49]

[50]

[51]

[52]

[53]

for the separation of inorganic cations by ion chromatography, J. Chromatogr. A, 706, 69, 1995. G. L. Schmitt and D. J. Pietrzyk, Liquid chromatographic separation of inorganic anions on an alumina column, Anal. Chem., 57, 2247, 1985. K. Ohta, M. Morikawa, K. Tanaka, Y. Uwamino, M. Furikawa and M. Sando, Ion-chromatographic behavior of alkali metal cations and ammonium ion on zirconium-adsorbing silica gel, J. Chromatogr. A, 884, 123, 2000. L. M. Nair and R. Saari-Nordhaus, Recent developments in surfactants analysis by ion chromatography, J. Chromatogr. A, 804 233, 1998. X. Liu, C. A. Pohl, and J. Weiss, New polar-embedded stationary phase for surfactant analysis, J. Chromatogr. A, 1118, 29, 2006. O. W. Reif, V. Nier, U. Bahr, and R. Freitag, Use of short monolithic beds for isolation and separation of biomolecules, J. Chromatogr. A, 664, 13, 1994. A. Steiner, M. D. Porter and J. S. Fritz, Ultrafast concentration and speciation of chromium(III) and (VI), J. Chromatogr., A. 1118, 62, 2006.

69

4 Detectors 4.1 Introduction

This chapter describes several different detectors that may be used in ion chromatography. But the reader may ask why so many detectors? Why not just use the conductivity detector that came with my instrument? Too many times the methodology that is devised for a particular analytical problem is just ‘good enough’. The analysis can be performed but the method is barely adequate in terms of resolution of peaks or in terms of sensitivity. And because it is just barely adequate, the method is not rugged. Often these issues can be solved by understanding how detectors operate, how ions are detected and choosing a better detector. For example, detection of iodide or nitrate in the presence of a salt (sodium chloride) matrix can be accomplished with conductivity detection. But UV detection would be much better because iodide will absorb UV light and chloride will not. Because the detection is selective for iodide, the separation conditions can be optimized for rapid interference-free elution. In IC, the detector must be able to ‘pick out’ and measure sample ions in the presence of a background of eluent ions. There are several methods that can be employed to make this possible. One is to choose a detector that will respond only to the sample ions of interest, but not to the eluent ions. Another method is to use indirect detection (sometimes called replacement detection). This is where the eluent has a background signal and the presence of sample ions causes a decrease in eluent ions through a replacement process. The detector measures the decrease in eluent ions when the sample ion peak elutes and a decreasing signal is detected. The most widely used method for ion chromatography detection is to treat or choose the eluent prior to detection to make the eluent ions less detectable and/or make the sample ions more detectable. The most common example of this is chemical suppression used in conductometric detection. The suppressor is really a chemical reactor that reacts with the post-column eluent stream and changes the ionic counterion for the eluent and for the sample peaks. In its most common form, sodium or potassium ions are removed from the stream and hydronium ions are added in an exchange process. This makes the background signal less Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

70

4 Detectors

conducting and the sample signal more conducting. Another example of treatment prior to detection is post-column reaction with a color-forming reagent. PAR, a color-forming chelator, can be added, post-column, to a separation of metal ions to make them detectable by visible spectrophotometric detection. The eluent ions do not react with the color-forming reagent. Detectors can be classified either as general or selective. A general detector will respond to all or most of the ions that pass through the detector cell. A conductometric detector is classified as a general detector because all ions will conduct electricity (although to different degrees). UV–Vis spectrophotometric, atomic emission, atomic absorption, and electrochemical detectors can be considered to be selective detectors because they respond only to certain ions. However, any of these can be made into a general-type detector by a post-column reaction of sample ions with an appropriate reagent. General and selective detectors both have their place in ion chromatography. If only one detector were available, a general detector would be most desirable because it would be of the most general use. However, a selective detector can be extremely effective in measuring a single ion of interest (or a small group of ions) from a high eluent background or a high sample matrix ion. Another key advantage to selective detection is the ability to achieve lower detection limits. A selective detector may or may not have a higher sensitivity (signal per unit concentration) for a particular ion. However, the lower background signals produced by selective detection will translate into lower detection limits because the signal-to-noise ratio is improved.

4.2 Conductivity Detectors

Conductance is the property of a solution containing a salt to conduct electricity across two electrodes. Conductance should not be confused with conduction, which is the mechanism by which electrons flow, or with conductivity, the property of a material. The ability of a solution to conduct is directly proportional to the salt concentration and the mobility of the individual anions and cations. As the ionic character of a molecule is increased, the conductivity increases. Small, mobile ions conduct quite readily and to a much greater extent than large bulky ions. For example, the hydroxide anion is small and mobile and will conduct much better than the propionate anion, which larger and bulkier. The relative conductances of ions are shown in Table 4.1. Molecular substances such as solvents (water and methanol) and solutions of nonionized organic acids and carbohydrates do not conduct electricity and are not detected by conductivity. The portion of a weak acid that does ionize will contribute to the conductivity signal. The ionic form of the weak acid depends on pH. A weak acid with a pKa larger than about 6 cannot be detected with suppressed conductivity detection because all anions are converted to the acid form by the suppressor. So while it is possible to separate both arsenite and arsenate on the

4.2 Conductivity Detectors

anion-exchange column (with a high pH eluent), it is only possible to detect the stronger acid, arsenate, by suppressed conductivity, and is not possible to detect the weak acid, arsenite. Other anions that cannot be detected by suppressed conductivity detection include borate, silicate and cyanide. Nonsuppressed, single-column methods, selective detection, or post-column reaction methods are required for the detection of salts of weak acids.

Table 4.1 Limiting equivalent ionic conductances in aqueous

solution at 25 °C. (Units: S cm2 equiv–1) Anions

k–

Cations

k+

Hydroxide

198

H3O+

350

+

Fluoride

54

Li

39

Chloride

76

Na+

50

Bromide

78

K+

74

Iodide

77

Nitrate

71 2–

Sulfate

80

78

+

77

Rb Cs

2+

Mg

53

69

2+

Ca

60

Chlorate

72

Sr2+

59

Perchlorate

67

Ba2+

64

66

2+

71

2+

53

2+

55

2+

53

2+

Phosphate

3–

+

Thiocyanate Hydrogen carbonate Carbonate

2–

Chromate

45 72 85

Pb

Zn Cu Co

Cyanide

82

Cd

54

Formate

55

Fe2+

54

41

3+

68

3+

70

Acetate Propionate 2–

Oxalate

3–

Citrate

36 74 56

Fe La

3+

Ce

NH4

70 +

Benzoate

32

CH3NH3

Phthalate2–

38

Et4N+

Borate/gluconate

26

73 +

58 33

71

72

4 Detectors

There are some IC users who employ direct or indirect nonsuppressed conductivity detection where the separation column is connected directly to the conductivity cell. The advantage of this detection method is simplicity in instrument design and operation and the ability to detect salts of weak acids. Cation detection can be more sensitive. But because of high sensitivity and reduced background noise, the most common form of conductivity detection is with the use of suppressors (see Section 6.2). Dasgupta and coworkers have proposed a novel approach to simultaneously practicing both suppressed and nonsuppressed ion chromatography [1–3]. Several renditions were proposed (Figure 4.1). The most effective and practical (Figure 4.1c) is based on first using a conventional NaOH eluent-suppressed IC system. The effluent from detector 1 is combined with a constant concentration of NaOH. The NaOH is introduced by an in-line electrolytic NaOH generator. Then the flow is directed to a second conductivity detector. The second detection background conductance is typically maintained at a level of 20–30 lS cm–1 corresponding to about 0.1 mM NaOH. The second detector output is the same as what would be observed in a single-column mode with a low-concentration NaOH eluent. Together, the two detector outputs provide detection at the microgram per liter level across the pKa range, from fully dissociated to very weak acids up to pKa 9.5.

Figure 4.1 Three potential approaches to perform simultaneous, suppressed, and nonsuppressed detection. (a) Split stream approach. DD, dummy dispersion device; Ds, suppressed detector; Dn, nonsuppressed detector. Restrictors R1 and R2 are adjusted to provide the

same residence time and flow rate in each of the branches. (b) Single column approach. Dn precedes the suppressor. (c) NaOH introduction approach. A small constant quantity of NaOH is introduced after Ds; no restriction is placed on eluent NaOH concentration [3].

4.2 Conductivity Detectors

The two-detector system also provides qualitative information about peak purity and sample pKa. This information is gained by plotting the two detector outputs against each other. In addition to the usual ratio plots, it is suggested that multidimensional detection in any separation system is best served by plotting the raw detector outputs against each other. This type of implementation of single-column and suppressed IC provides information beyond the sum total of that obtained with either approach alone. 4.2.1 Conductivity Definitions and Equations

Electrolytic conductivity is the ability of an electrolytic solution to conduct electricity between two electrodes across which an electric field is applied. Ohm’s law, V = I R, is obeyed, and the magnitude of the current depends, in part, on the magnitude of the applied potential. The conductance, G, of a solution is expressed in terms of the solution electrolytic resistance. It is measured in reciprocal ohms (mhos) or in the Sl unit siemens (S). G = 1/R

(4.1)

Specific conductance (k) takes into account the area of the electrodes (A) in cm2 and the distance (l) between electrodes, in cm. Conductance increases with the area of the electrodes but decreases as the distance between the electrodes is increased. k = G (l/A)

(4.2)

Thus, k has the units S cm–1. The cell constant (K) is equal to l/A in Eq. (4.2) and has the units cm–1. K = l/A

(4.3)

k= G K

(4.4)

Equivalent conductance takes into account the concentration of the chemical solution and is defined by the following equation: K = 1000 k/C

(4.5)

where C is the concentration in equivalents per 1000 cm3 and K has the units S cm2 equiv–1. Combining Eqs. (4.4) and (4.5) gives the following equation, which relates equivalent conductance to measured conductance, G: G = K C/1000 K

(4.6)

73

74

4 Detectors

A conductivity detector consists of a detection cell, a readout meter, and the electronics required for measuring the conductance and varying the sensitivity setting. The readout for conductance G is given in Siemens (S), or, actually, microSiemens (lS) for solutions that are comparatively dilute. The specific conductance for a solution can be calculated from the solution conductance (G) if the cell dimensions are known [see Eq. (4.2)]. However, the usual practice is to measure the conductance of a dilute solution of known specific conductance (such as 0.00100 N KCI) and calculate the cell constant from Eq. (4.4). Once the cell constant is known, the specific conductances of other solutions can be calculated from the measurement of G. The most common cell constant for detectors is 1 or 10; the smaller value is more sensitive. With a conductivity detector with a known cell constant, the conductances of various solutions of known concentration can be calculated from a table of equivalent conductances using Eq. (4.6). The limiting equivalent conductances of some common ions are given in Table 4.1. The equivalent conductances of ions generally decrease with increasing concentration because of interionic effects. For dilute solutions (10–5 to 10–2 N) the equivalent conductances are not greatly different from the values listed in the table. Example: Calculate the expected conductance of a 1.0 × 10–4 N solution of sodium benzoate in a conductivity cell with a cell constant of 10 cm–1. From Table 4.1, the equivalent conductance of sodium benzoate will be kNa+ + kBz-. Therefore: kNa+Bz- = 50 + 32 = 82 Substituting this into Eq. (4.6): G = (82 × 1.0 × 10–4)/(1000 × 10) = 8.2 × 10–7 mhos or 0.82 lS A more typical ion to detect might be chloride. The equivalent conductance is higher so a higher signal would result from this ion, assuming the same peak width and the same sodium counterion. If hydronium ion rather than sodium is the counterion to chloride, then the signal will be multiplied by another factor of 3.4. 4.2.2 Principles of Cell Operation

When an electric field is applied to two electrodes in an electrolytic solution, anions in the solution move toward the anode electrode and cations toward the cathode electrode. The number and the velocities of the ions in the bulk electrolyte determine the resistance of the solution. The ionic mobilities, or the velocity of the ion per unit electric field, depend on the charge and size of the ion, the temperature and type of solution medium, and the ionic concentration. As the potential that is applied across the electrodes is increased, the ionic velocities increase. Thus, the detector signal is proportional to the applied potential.

4.2 Conductivity Detectors

This potential can be held to a constant value or it can oscillate to a sinusoidal or pulsed (square) wave. Cell current is easily measured; however, the cell conductance (or reciprocal resistance) is determined by knowing the potential to which the ions are reacting. This is not a trivial task. Ionic behavior can cause the effective potential that is applied to a cell to decrease as the potential is applied. Besides electrolytic resistance that is to be measured, Faradaic electrolysis impedance may occur at the cell electrodes resulting in a double layer capacitance. Formation of the double layer capacitance lowers the effective potential applied to the bulk electrolyte. 4.2.3 Conductance Measurement

Techniques involving the use of alternating electrode potentials eliminate the effects of the processes associated with the electrodes. Reversing the polarity of the applied electrode potential reverses the direction of the ion motion, changes the type of electrolysis and changes the type of capacitance formation. The relaxation time (or the ability to recovery) is different for each type of process. As the frequency is increased, effects due to electrolysis are reduced or eliminated and the bulk of current flow is through capacitance formation. An upper frequency limit for detector operation is approximately 1 MHz. At this point the ions cease to move in response to the electric field, although dipole reorientation of the ion electron structure will still occur. Capacitance effects are controlled by matching the cell capacitance in the electronic circuitry or by measuring the instantaneous current. The instantaneous current is the current that is obtained when the potential is first applied and the double layer has not formed. Some detectors apply a sinusoidal wave potential across the cell electrodes at 100 to 10 000 Hz. A typical detector of this type operates at a frequency of 1 kHz and at a potential of up to 20 V with no electrolysis occurring. This detection method is called synchronous detection. This is when the only current component measured is that current which is in phase with the applied potential frequency. In effect, the measured current flow is always due to an ‘instantaneous’ potential. Other detectors use a bipolar pulse conductance technique [4, 5]. The technique consists of the sequential application of two, short (about 100 ls) voltage pulses to the cell. The pulses are of equal magnitude and duration and opposite polarity. At exactly the end of the second pulse, the cell current is measured and the cell resistance is determined by applying Ohm’s law. Because an instantaneous cell current is measured in the bipolar pulse technique, capacitance does not affect the measurement and an accurate cell resistance measurement is made. 4.2.4 Conductivity Hardware and Detector Operation

Conventional conductivity detector cells where the electrolyte is in contact with the electrodes are likely to use electrodes made from 316 stainless steel. A new

75

76

4 Detectors

cell should be treated with 1N nitric acid for about 60 min to ‘deactivate’ or ‘passivate’ the cell and stabilize the signal. In fact, such nitric acid treatment is a good idea for all parts of a stainless steel IC system. The mobility of ions in solution varies with the solution temperature. Ionic solutions will increase in conductivity about 2% for every degree increase in temperature. Conductivity detectors usually compensate automatically for temperature change by employing a thermistor monitor and compensation circuitry in which resistance changes linearly with solution temperature. Still, the detector cell (and even the column and tubing) should be placed in an oven for the best detector performance. It is helpful to insulate other components of the ion chromatograph. If the laboratory has large daily temperature swings, temperature control of the column and cell becomes more important. Generally, the instrument temperature is set to at least 5 °C above the maximum temperature that the laboratory is likely to reach in a given day. Control of the instrument oven and detector temperature and simply keeping the instrument out of drafts are probably the most important two things that a user can do to contribute to better detector stability. The flow path may also include a compartment containing the thermistor probe for electronic feedback. Even though conductance cells tend to be low dead volume, the total dead volume in a conductance detector can be quite large, and considerable mixing of the eluent stream may take place after the peak has been measured. If two detectors are used, it may be best to place the conductivity detector cell last to avoid peak broadening. 4.2.5 Contactless Conductivity Detection

Zemann and coworkers have developed a novel contactless conductivity detector [6–12]. Contactless conductivity detection offers the advantage of avoiding detection dead volumes. This is especially important for miniaturized chromatographic and electrophoresis systems. The detector works without direct contact of the electrode with the eluent or sample. The sensor is based on two metal tubes that are placed around a fused silica capillary with a detection gap of approximately 1.5 mm (Figure 4.2). The conductivity sensor is based on two metal tubes that act as cylindrical capacitors. The electrodes may be placed around any nonconducting tubing such as fused silica, PEEK, or Teflon. Dead volume of the connecting tubing is minimized and an extremely low dead volume cell can be manufactured. A high oscillating frequency of 40–100 kHz is applied to one of the electrodes. A signal is produced on the other electrode as soon as an analyte zone with a different conductivity compared to the background passes through the detection gap. An amplifier and rectifier are connected to the second electrode to measure resistance between the two electrodes. To isolate the two capacitors associated with each electrode, a thin piece of copper is placed between the electrodes and grounded.

4.3 Ultraviolet-Visible (UV–Vis) Detectors

Figure 4.2 Schematic drawing of the contactless capacitively coupled conductivity detector [6].

4.3 Ultraviolet-Visible (UV–Vis) Detectors 4.3.1 UV–Vis measurement

A spectrophotometric UV–Vis detector is selective, yet its selectivity can be changed simply by changing the wavelength monitored by the detector. Versatility of the detector can be increased by adding a color-forming reagent to the eluent or the column effluent. The fundamental law under which ultraviolet-visible (UV– VIS) detectors operate is the Lambert–Beer law. It can be stated in the following form: A= e b C

(4.7)

A is the absorbance of a species of concentration C, and with an absorptivity e, in a cell of length b. Concentration is usually in molar concentration units and the path length is measured in cm. The term (molar) absorptivity has units that are the inverse of the C and e units. This leaves A dimensionless; it is usually described in terms of absorbance units. A detector set to a certain sensitivity, for example, 0.16, is said to be at 0.16 Absorbance Units Full Scale sensitivity (0.16 AUFS sensitivity). The Lambert–Beer equation is useful for choosing conditions for the separation and detection of ions. The eluent ions should have a low absorptivity and the sample ions should have reasonably high absorptivity. In the special case of indirect detection this should be reversed. In this case, the eluent has an absorption signal and the sample is detected by a decrease in the background signal. It is important to note that when discussing the properties of the eluent and sample ions, it makes a difference whether one is separating anions or cations.

77

78

4 Detectors

For example, if a separation of anions is being discussed, then the absorptivities of the eluent anion and sample anions are considered. But if a low background signal is needed, then of course the cation that is counter to the eluent anions must have a low absorptivity as well. Some absorbance data are given in this chapter and in some of the applications described in this book. If an ion absorptivity for a particular wavelength is unknown, it can be measured with a spectrophotometer and ion solution of known concentration. The discussion of UV–Vis detectors for use in ion chromatography is divided into two parts: (a) the direct monitoring of column effluents and (b) post-column derivatization with subsequent spectrophotometric measurement. 4.3.2 Direct Spectrophotometric Measurement

Alkali metal ions are not detected by UV. However, many anions do absorb at lower wavelengths. A list of anions and their detection wavelengths is shown in Table 4.2.

Table 4.2 Solutes for direct spectrophotometric detection in IC

after ion-exchange separation. Solute AsO43–, AsO33– Au(CN)2

–

Wavelength (nm)

Ref.

200

[13]

214

[14]

–

Br

195–214

[14]

BrO3–

195–210

[13, 16, 17]

C2O42–

205

[18]

Citrate

205

[18]

–

Cl

190

[19, 20]

CIO2–

195

[13]

CIO3–

195

[13]

CN–

200

[21]

200

[21]

365

[22]

350, 220

[23]

225

[24]

CNO– CrO4 III

2– VI

Cr , Cr -EDTA –

Cr, Pt, Au-Cl complexes

4.3 Ultraviolet-Visible (UV–Vis) Detectors Table 4.2 (Continue) Solutes for direct spectrophotometric

detection in IC after ion-exchange separation. Solute

Wavelength (nm)

Ref.

HCOO–, CH3COO–

190

[19]

210–235

[25, 26]

195–210

[13, 17]

Metal Cl complexes

210–225

[27-30]

Metal CN– complexes

210–214

[31, 32]

Metal EDTA complexes

210

[33]

200

[21]

195

[13]

200–214

[20, 34, 35]

210

[36]

190

[19]

215

–

205

–

I

–

IO3

– –

MoO4

2–

N3– –

NO2 , NO3

–

Organoarsenic acids PO4

3–

S2– 2–

2–

S2O3 , S3O6 , –

SCN

S4O62–

195–205

[37, 38]

–

SeCN

195

[13]

SeO32–

195

[13]

SeO42–

195

[13]

SO32–

200

[38]

SO42–

190

[19, 20]

The references listed show the different applications of the detection method. The 190 to 210 nm range can be used for the detection of azide, chloride, bromide, bromate, iodide, iodate, nitrite, nitrate, sulfite, sulfide, and selenite. Other work has shown that the 210 to 220 nm range is useful for detecting trithionate, tetrathionate, and pentathionate down to the low nanogram levels. UV detection is particularly useful for anions such as nitrate and iodide, which absorb at the longer wavelengths. There have been a number of methods reported for the UV detection of nitrite and nitrate in drinking water and cured meat [39]. Aromatic acids absorb well and methods for detecting these anions are powerful. The usefulness of direct UV detection can be considered to be limited because sulfate is not detected by UV, and chloride, phosphate and others are difficult to detect. Sulfate is probably the most widely analyzed anion by ion chromatography

79

80

4 Detectors

so this is a serious limitation. On the other hand, anions that are difficult to detect make ideal eluent anions. The absorbance of metal chloride complexes in the ultraviolet spectral region has been used extensively to automatically detect metal ions in liquid chromatography [27–30]. The absorption wavelength maxima of the metal chloride complexes are shown in Table 4.3. Metal–EDTA complexes also absorb quite well. Another technique is indirect detection. In this method, the eluent absorbs strongly in the visible or ultraviolet spectral region. A wavelength is selected where the (usually aromatic) eluent absorbs but the sample ions do not [41–43]. Briefly, because an ion-exchange process is involved, a sample ion can only be eluted by displacement of the eluent ion. This results in a decrease in the signal when a sample ion peak is eluted. While several authors and users have been successful with indirect UV detection, it can be difficult to get the right conditions for separation and detection. Indirect UV detection is generally used only in cases where the separation and detection conditions have been carefully worked out and where a high quality UV detector is available. Temperature control of the column is recommended to control the baseline noise. 4.3.3 Post-column Derivatization

The post-column method of derivatization of column fractions has been well established from older ion-exchange separations. An appropriate reaction is performed on each fraction to determine the metal ion concentration in that fraction. The automatic addition of a color-forming reagent to an ion-exchange column effluent and analysis by flow-through cell detection is more recent. However, many of the color-forming reagents and buffers used in ion chromatography are the same as those used in the classical fraction method determinations. The ideal color-forming reagent reacts with a large number of metal ions and has low background absorption. Sickafoose [44] studied the reagents alizarin red S, arsenazo III, chlorophosphonazo III, chrome azurol S, quinalizarin, 4-(2-pyridylazo)resorcinol (PAR), 4-(2-thiazolylazo)resorcinol (TAR), and xylenol orange. Chrome azurol S and quinalizarin are of limited value, but the other reagents each react with 20 or more metals. PAR is the most general, reacting with 34 metals. Table 4.4 shows a list compiled by Fritz and Story [45] of metals and their reaction with color-forming agents. The 0.0125% PAR solution was made in 5 M ammonium hydroxide, the 0.00375% arsenazo III solution was in 2 M ammonium hydroxide and 1 M ammonium acetate, and the arsenazo I solution was in 3 M ammonium hydroxide. Other work [46, 47] was done with more dilute PAR solutions (4 × 10–4 M PAR with 3 M ammonium hydroxide and 1 M acetic acid). For lower detection limits (because of lower background signal), the PAR concentration can be reduced even more. But care should be taken not to ‘overload’ the reagent with too high concentrations of sample ions. Imanari et al. have reported a spectrophotometric detection of many inorganic anions using a post-column reactor [48]. A stream of ferric perchlorate, which is

4.3 Ultraviolet-Visible (UV–Vis) Detectors Table 4.3 Complex formation of inorganic anions with ferric

perchlorate color-forming reagent[a] [48]. Anion

kmax(nm)

Anion

kmax(nm)

CrO42–

305, 344

SO32–

308

–

3–

310

PO3

305

H2PO2–

P+A– + Cl–

(11.11)

for which the equilibrium constant (K) is: Kˆ

‰P‡ A Š‰Cl Š ‰A Š‰P‡ Cl Š

(11:12†

11.5 Capillary Electrophoretic Ion Chromatography

At a fixed concentration of P+Cl–, a conditional constant, K′, may be written as follows: K′ = K [P+Cl–]

(11.13)

Combining Eqs.11.12 and 11.13, and rearranging gives: ‰A Š ‰Cl Š ˆ ‡ K′ ‰P A Š

(11:14†

The electrophoretic migration rate will depend primarily on the fraction of sample anion that is present as the free anion. This is true because the free anion (A–) will migrate rapidly toward the anode, while the fraction associated with the ion exchanger (P+A–) will move only very slowly in the opposite direction. The fraction present as A– will depend both on the total anion concentration ([Cl–] in Eq 11.11) and on the value of K′, which will be different for each sample anion. Incorporation of PDDAC into the capillary electrolyte also reverses the direction of EOF. Addition of 0.05% to 0.30% PDDAC to the BGE sets up a dynamic equilibrium in which PDDAC forms a thin coating on the inner walls of the capillary. This imparts a negative charge to the surface and sets up an electroosmotic flow toward the anode which is in the opposite direction to the usual cathodic EOF in uncoated capillaries. Under typical conditions the EOF in capillaries equilibrated with PDDAC was almost constant over a wide pH range (Table 11.4). A negative power supply (–10 kV) is used for anion separations. The BGE typically contains 0.05% to 0.30% PDDAC, up to 150 mM sodium chloride or lithium sulfate, and a 20 mM borate buffer. The EOF vector and the electrophoretic vectors of the sample anions are both in the anodic direction. The fraction of each sample anion that is attached to exchange sites in the PDDAC has a cathodic electrophoretic vector, although this is believed to be weak. The net electrophoretic vector for any given anion depends primarily on the fraction that exists as the free ion [Eq. (11.14)] and on the electrophoretic mobility of the free ion. Separations of common inorganic anions are quite fast, as indicated by the migration times in Table 11.4. It is also possible to obtain a baseline separation because of the greater ion-exchange affinity for iodide; the electrophoretic mobilities of bromide and iodide are almost identical. A separation of several inorganic anions is illustrated in Figure 11.15. The effect of experimental variables on the separation of anions can be summarized: 1. Increasing the BGE concentration in steps from 0.1 to 1.0% (approximately 6 to 60 mN) decreases the electrophoretic mobilities of sample anions to varying degrees. This results in longer migration times. 2. Increasing the concentration of LiSO4 in the BGE from 50 to 150 mM increases the electrophoretic mobilities of sample anions, probably by reducing the interaction between the anions and PDDAC [see (Eq. 11.11)].

289

290

11 Capillary Electrophoresis Table 11.4 Comparison of EOF and migration times at different

pH values for inorganic anions. Capillary: 40 cm × 50 mm; electrolyte: 150 mM Li2SO4, 0.05% PADDC, 20 mM borate for pH 8.5, or 20 mM sodium acetate for pH 5.0, or 20 mM HCl for pH 2.3; separation voltage: –10 kV; hydrostatic sampling: 40 s at 10 cm height; detection: UV: 214 nm; EOF marker: D.I. H2O. Migration time (min) Br– I–

PH

EOF (cm2 V–1 s–1)

2.3

–2.46 × 10–4

1.98

5.0

–2.65 × 10–4

8.5

–4

–2.74 × 10

NO3–

SCN–

2.06

2.17

2.34

1.95

2.03

2.13

2.30

1.93

2.00

2.10

2.26

Figure 11.15 Separation of organic anions. Capillary: 40 cm × 50 lm; electrolyte: 150 mM Li2SO4, 20 mM borate, 0.3% PADDC, pH 8.5; separation voltage: –10 kV; hydrostatic sampling: 40 s at 10 cm height; detection: UV, 214 nm. Peaks: 1 = benzenesulfonic acid;

2 = benzoic acid; 3 = p-toluenesulfonic acid; 4 = p-hydroxybenzoic acid; 5 = p-aminobenzoic acid; 6 = 2-naphthalenesulfonic acid; 7 = 1-naphthalenesulfonic acid; 8 = 3,5-dihydroxybenzoic acid; 9 = 2,4-dihydroxybenzoic acid (from Ref.[17] with permission).

The type of salt, as well as its concentration, can have a major effect on the migration of sample anions. In ion chromatography, sulfate is known to have a much stronger affinity for a solid quaternary ammonium anion exchanger than acetate, for example. Thus, acetate will have a much smaller inhibiting effect on the ion exchange of sample anions with PDDAC than the same concentration of sulfate. The migration times of bromide and iodide in 150 mM lithium sulfate are 5.74

11.5 Capillary Electrophoretic Ion Chromatography

and 6.88 min, respectively (a = 120). In 150 mM sodium acetate the migration times are 6.08 min for bromide and 8.77 min for iodide (a = 1.44). 11.5.4 Effect of Alkylammonium Salts

Alkylammonium cations, as well as quaternary ammonium polymers, may be used to modify the electrophoretic migration of anions. Steiner, Watson and Fritz [17] demonstrated that a 100-mM concentration of a quaternary ammonium salt provides an excellent medium for separation of anions. The electrophoretic migration of sample anions is slowed by ion association (or ion exchange) with the organic cation. The extent of anion–cation interact ion varies with the bulk and hydrophobicity of the cations as well as the cation concentration. This effect is illustrated by the electropherograms in Figure 11.16 where the BGE contained 100 mM R4N+Cl– and 5 mM buffer. The anion migration times increase in the order R = ethyl < propyl < butyl. Although excellent separations of inorganic anions and organic sulfonates were obtained, an opposing electroosmotic flow at alkaline pH values necessitated the use of an acidic pH for many of the separations. This drawback was avoided in later work [18] by precoating the capillary with PDDAC. Coating is accomplished simply by passing an aqueous solution of the soluble polymer through the capillary. The capillary acquires a semi-permanent layer of the polymer that gives a strong EOF in the desired direction (anodic) at any pH. Figure 11.17 shows the separation of eleven inorganic and organic anions with a PDDAC-coated capillary and a 100-mM aqueous solution of tetrabutylammonium acetate at pH 6.0. A negative power supply was used with co-migration of the EOF and electrophoretic vectors. In these separations, use of a moderately high concentration of an alkylammonium salt in the electrolyte again adds a chromatographic component to CE. Sample anions migrating electrophoretically toward the detector are slowed by transient association with alkylammonium cations moving in the opposite direction. This association is considered to result from a combination of ion–ion and dispersive interactions. The following equilibrium is established: R4N+ + A– > R4N+A–

(11.12)

with the association constant, KAS ˆ

‰R4 N‡ A Š ‰R4 N‡ Š‰A Š

(11:13†

This can be viewed as a dynamic equilibrium in which the anion spends a certain time fraction migrating as the free anion and another time fraction as the association complex, during which migration is much slower. The magnitude of this effect is a function of the type and concentration of the electrolyte cation, the counter-anion and of the properties of the analyte anion.

291

292

11 Capillary Electrophoresis

Figure 11.16 Effect of the alkylammonium salt on the separation of anions at pH 3. BGE contains 100 mM R4NCl + 5 mM buffer. Applied voltage, –15 kV. Column length, 50 cm, 42 cm to detector. Peak identification: 1 = bromide, 2 = iodide, 3 = thiocyanate, 4 = p-toluene-

sulfonate, 5 = 1-napththalenesulfonate, 6 = 2-naphthalenesulfonate. (A) tetraethylammonium chloride; (B) tetrapropylammonium chloride; (C) tetrabutylammonium chloride (from Ref [17] with permission).

11.5 Capillary Electrophoretic Ion Chromatography

Figure 11.17 Separation of inorganic and organic anions. Conditions: 100 mM tetrabutylammonium acetate electrolyte at pH 5.5; –15 kV. Peaks in order of increasing tm: nitrite, nitrate, iodide, 1,2-benzenedisulfonate,

thiocyanate, mandelate, p-toluenesulfonate, 1,2-naphthoquinone-4-sulfonate, salicylate, 2-naphthalenesulfonate, 1-naphthalenesulfonate (from Ref. [18]).

The association constants, KAS, were measured by the method of Keneta et al. [11]. The reciprocal of measured, or effective, electrophoretic mobility of the analyte anion (lef ) is plotted against the concentration (C) of R4N+ in the capillary electrolyte. 1 CK AS 1 ˆ ‡ lef lA lA

(11:14†

The intercept of a linear plot gives the value of 1/lA, where lA is the electrophoretic mobility of the free, uncomplexed analyte anion. Knowing this, the value of the association constant, KAS, can be calculated. Values of KAS for analyte anions in several systems are presented in Table 11.5. As might be expected, KAS is larger for tetrabutyl than for tetraethyl quaternary ammonium salts. Examination of the values for tetraethylammonium salts, which generally gave r2 values of 0.99 or better, show that KAS for sulfate was lower than for acetate. This is an indication that the process we are observing involves ion exchange in solution and not merely ion association where the anion would play no role. R4N+E– + A– > R4N+A– + E–

(11.15)

In this equation, E– is the electrolyte anion and A– is the analyte anion. Of course, the concentration of E– is several orders of magnitude higher than that of A–.

293

294

11 Capillary Electrophoresis Table 11.5 Association constants, KAS , for several anions, based

on electrophoretic mobility measurements at 0.025, 0.05 and 0.10 M concentrations of quaternary ammonium salts. The values given are estimated to be ± 2–3%. KAS Anion

Et4Nac

Et4NSO4

Bu4Nac

BDSA

1.01

–

1.46

1.06

TSA

1.49

1.34

4.94

3.69

1-NSA

3.65

2.55

11.4

Bu4NSO4

–

11.5.4.1 Separation Mechanism In the experimental system, the capillary is filled with electrolyte with each end open to a separate reservoir of electrolyte. When a voltage is applied, an electric current is established in which electrolyte anions migrate toward the positive electrode (anode) and electrolyte cations move in the opposite direction. After sample injection, sample anions are separated by differences in their electrophoretic migration rates to the detector at the anodic end of the capillary. Although the sample anions migrate against a counter flow of electrolyte cations, interaction between the cations and anions is minimal in conventional CE where the background electrolyte typically contains a fairly low concentration of sodium or ammonium cations. The situation changes drastically when a higher concentration (about 100 mM) of a larger organic cation is used in the electrolyte. Conditions are now favorable for cation–anion interactions as the sample anions migrate through a veritable sea of the larger alkylammonium cations. These interactions slow the electrophoretic migration rates of sample anions to varying degrees, thus producing a strong chromatographic component to the separation. These interactions may be said to occur on a nano scale because the effective radii of the organic cations are of the order 1 nm or less. It is possible that the organic cations may undergo some attraction for one another and form a series of nano domains. Actual micelle formation can be ruled out by the relatively low molecular weight as well as by geometrical considerations for the quaternary ammonium salts used.

11.6 Summary

Capillary electrophoresis is a technique in which ions are separated by differences in their electrophoretic migration rates when a high voltage is applied. The ‘column’ is an open tubular capillary, similar to the capillary used in gas chromatography. Addition of an ionic surfactant, polymer, or a larger organic ion also modi-

11.6 Summary

fies the migration rates of sample ions through the capillary. These electrolyte additives slow the migration rates of oppositely charged analyte ions in much the same way as a solid ion exchanger does in conventional ion chromatography. The major difference is that in the capillary system, the cation–anion interactions between the electrolyte additives and analyte ions occur between the liquid carrier and a ‘pseudo phase’ when a surfactant micelle is involved. The cation–anion interactions take place within a single liquid phase when the electrolyte additive is a soluble polymer or simply a moderately large organic ion. These conditions result in much sharper sample-ion peaks than in conventional ion chromatography where mass transfer between two distinct phases is required. Terabe [10] has proposed the name ‘ion-exchange EKC’ for separations where the electrophoretic migration of analyte ions through an open-tubular capillary is slowed to varying degrees by ion exchange with soluble polymers in the capillary electrolyte. Capillary electrophoretic ion chromatography, or CE-IC, is also an appropriate name to describe separations of this kind. Although CE methods generally provide very good separations with sharp peaks, there can be some problems. Sometimes reproducibility of migration times is relatively poor and peaks are broad and unsymmetrical. These aberrations are most likely due to interactions between the analyte ions and the capillary wall. Most difficulties can be avoided by selecting experimental conditions that: (i) give a stable, reproducible interface between the capillary electrolyte and the wall, and (ii) avoid interactions between analyte ions and the wall. Reproducible migration times require a stable EOF. The magnitude of the EOF vector is influenced by the extent to which silica silanol groups are ionized to give a negative charge, and this varies considerably with pH. Thus, a stable EOF requires a thorough equilibration of the capillary surface with a pH buffer and finally with the buffered capillary electrolyte. All solutes within the capillary can equilibrate to varying degrees between the liquid and the capillary surface. But when the ionic concentration of the carrier electrolyte is much higher than that of the sample, analyte ions are kept from the capillary surface according to the law of mass action. There will be a stable equilibrium that will hardly be affected by analytical samples of different compositions. For this reason, a fairly high concentration (100–150 mM) usually gives more reproducible times and sharper peaks than the more dilute electrolytes that were formerly recommended. Suggested conditions for capillary electrophoretic separations are summarized in Table 11.6. A larger number of experimental options are available than with conventional ion chromatography. However, the separation power of CE is so great that an adequate separation can usually be obtained even with less than optimal experimental conditions.

295

296

11 Capillary Electrophoresis Table 11.6 Suggested conditions for capillary electrophoretic separations.

General Conditions Capillary Length:

About 40 – 60 mm.

Capillary i.d.:

50 lm.

Applied potential:

–15 – 20 kV for anions; +15 – 20 kV for cations.

Format:

Counter-migration with moderate opposing EOF if possible. Co-migration is faster but gives poorer peak resolution.

Detector:

Several options are available. Direct UV-vis is most common for ions with sufficient absorption. Indirect detection with an added visualization reagent may also be used.

pH:

Appropriate for sample chemistry. Acidic pH for protonated amine cations; alkaline pH for most anions.

EOF Control:

Vary pH. Use coating for capillary surface, if necessary.

Migration time:

2–10 min range if possible. Adjust by changes in pH, capillary surface, or applied potential.

Other Conditions

Effect

1. Capillary Surface a) Pretreatment: NaOH, wash, with Hydrolyzes SiOSi groups, prepares surface. BGE buffer b) Other treatment: ionic surfactant, Adjusts or reverses EOF to a desired value. ionic polymer or R4N+ 2. Background Electrolyte a) Use a higher conc. (100 mM)

Promotes electrostacking, inhibits analyte interactions with wall.

b) Organic ions, not inorganic

Gives a lower current, improves solubility of organic analytes.

c) 5–20 mM pH buffer

Maintains desired pH.

d) Consider use of an additive to give a chromatographic effect.

Improves peak resolution.

3. Sample a) More dilute than BGE

Increases electrostacking.

b) Short injection time (~ 5 s), Gives sharper peaks. unless a larger sample is needed. c) Keep total amount of analytes low Avoids sample overloading and poorly shaped peaks.

References

References [1] J. W. Jorgenson and K. D. Lukacs, Zone

[2]

[3]

[4] [5]

[6]

[7]

[8]

[9]

electrophoresis in open-tubular glass capillaries, Anal. Chem., 53, 1298, 1981. P. Jandik and G. Bonn, Capillary electrophoresis of small molecules and ions, p. 23, VCH, New York, 1993. K. K.-C. Yeung and C. A. Lucy, Improved resolution of inorganic anions in capillary electrophoresis by modification of the reversed electroosmotic flow and the anion mobility with mixed surfactants, J. Chromatogr. A, 804, 319, 1998. K. M. Lau, PhD Thesis, University of Hong Kong, December 1997. S. A. Shamsi and N. D. Danielson, Naphthalenesulfonates as electrolytes for capillary electrophoresis of inorganic anions, organic acids, and surfactants with indirect photometric detection, Anal. Chem., 66, 3757, 1994. C. A. Lucy and T. L. McDonald, Separation of chloride isotopes by capillary electrophoresis based on the isotope effect on ion mobility, Anal. Chem., 67, 1074, 1995. M. J. Thornton and J. S. Fritz, Separation of inorganic anions in acidic solution by capillary electrophoresis, J. Chromatogr. A, 770, 301, 1997. W. Ding, M. J. Thornton and J. S. Fritz, Capillary electrophoresis of anions at high salt concentrations, Electrophoresis, 19, 2133, 1998. S. Terabe, K. Otsuka, K. Ichikawa, A. Tsuchiya and T. Andg, Electrokinetic separations with micellar solutions and open-tubular capillaries, Anal. Chem., 56, 111, 1984.

[10] S. Terabe and T. Isemura, Effect of poly-

mer ion concentrations on migration velocities in ion-exchange electrokinetic chromatography, J. Chromatogr., 515, 667, 1990. [11] T. Kaneta, S. Tanaka, M. Taga and H. Yoshida, Migration behavior of inorganic anions in micellar electrokinetic capillary chromatography using a cationic surfactant, Anal. Chem., 64, 798, 1992. [12] W. R. Jones, P. Jandik and R. Pfeifer, Capillary ion analysis, an innovative technology, Am. Lab., 5, 40, 1991. [13] Y. Shi and J. S. Fritz, Separation of metal ions by capillary electrophoresis with a complexing electrolyte, J. Chromatogr., 640, 473, 1993. [14] Y. Shi and J. S. Fritz, New electrolyte systems for the determination of metal cations by capillary electrophoresis, J. Chromatogr. A, 671, 429, 1994. [15] C. Stathakis and R. M. Cassidy, Effect of electrolyte composition in the capillary electrophoretic separation of inorganic/ organic anions in the presence of cationic polymers, J. Chromator. A, 699, 353, 1995. [16] J. Li, W. Ding and J. S. Fritz, Separation of anions by ion chromatography-capillary electrophoresis, J. Chromatogr. A, 879, 245, 2000. [17] S. A. Steiner, D. M. Watson and J. S. Fritz, Ion association with alkylammonium cations for separation of anions by capillary electrophoresis, J. Chromatogr., 1085, 170, 2005. [18] S. A. Steiner and J. S. Fritz, Unpublished work, 2006.

297

299

12 DNA and RNA Chromatography 12.1 Introduction 12.1.1 Importance of DNA and RNA Chromatography

The reader may well ask what nucleic acids, deoxyribonucleic acid (DNA) and ribonucleic acid (RNA) have to do with ion chromatography. Nucleic acids are very large biological molecules that are more like linear polymers and bear little similarity to common anions such as chloride and sulfate. Nevertheless, DNA and RNA are anionic by virtue of the phosphate groups attached to each nucleotide sugar unit. Nucleic acids can be separated by chromatography, often by ion-pair methods similar to those described in Chapter 9. For these reasons, it seems appropriate to include a chapter on DNA and RNA chromatography in this book. Advances in separation technology over the past few years have shown that chromatography can perform remarkably well for separating, measuring and purifying these biological molecules. From a technical perspective, the instrumentation and many of the methods are very similar to what is used in traditional ion chromatography. In general, an ion chromatograph instrument can be used to perform DNA and RNA separations. Although the column is specifically designed for nucleic acid separations, the eluent buffers, methods of detection and so on are similar to what is used in many types of ion chromatography. However, DNA chromatography differs in several important ways from the ion chromatography of smaller ions. Specially designed columns are required to prevent strong adsorption of the bulky nucleic acids. Temperature of the eluent and sample is a major variable in DNA chromatography. Elevated column and eluent temperatures cause structure changes that can alter the elution profile of DNA fragments. Careful precautions must be taken to avoid contamination from traces of metal in the chromatographic system. While it is understood that analytical chemists are the primary audience for this 4th edition of Ion Chromatography, it is also true that the analytical chemist and ion chromatographer are being asked more and more to provide answers to biological problems. In a sense, many of the new research challenges facing the Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

300

12 DNA and RNA Chromatography

biologist are problems that are much more analytical in nature. Because biologists are not normally trained to use advanced analytical methodology, many of them seek out the analytical chemist to help them in their research. DNA provides the general ‘blueprint’ for the cells to make proteins. After the proteins are made, they go on to do the work of the cell in performing various cellular processes. RNA provides the mechanism for translating information held by DNA into these cellular proteins. Over the past several years, DNA and RNA have become of extremely high interest because of their potential to be used as drugs to treat disease. A huge amount of academic and industrial resources are now being spent on trying to understand, manipulate and use DNA and RNA. It has also been predicted that DNA-based drugs will someday cure disease with a single treatment (rather than our current treatment methods of a daily regimen of drugs) by incorporating new genes into cells. Silencing RNA has been demonstrated to be able to turn off or control specific biological functionality. Through research that is now being pursued, it has been predicted that RNA-based drugs will prove to be extremely powerful and will someday revolutionize treatment of disease. But in order to accomplish these ambitious goals, advanced analytical tools and methods will have to be used in biological research. And these advanced analytical tools are likely to include DNA and RNA chromatography. 12.1.2 Organization of this Chapter

We start this chapter by briefly describing nucleic acid composition and structure. Nucleic acids have several different forms or structures, depending on their biological function. These forms can be controlled (or preserved if desired) through chemical and temperature control. Several column types have been used to separate nucleic acids. A discussion of the historical development of DNA and RNA chromatography is followed by a brief description of the column development that has led to the modern use of ion-pairing chromatography. There are three modes of modern nucleic acid chromatography, which are mainly dependent upon the temperature of the eluent and column under which the separations are performed. A description of specific types of separations will be discussed, along with the hardware and columns needed to perform these separations. DNA and RNA chromatography require that the instrument, column and eluent be completely free of metal contamination. The effect, source, and control of metal contamination will be discussed. A high-quality oven must be used. The instrument normally includes the option of collecting the nucleic acid in a fragment collector for further research and processing. Finally, applications of DNA and RNA chromatography will be shown. The purpose of using this type of methodology is to provide the means to answer biological questions. The power of the methodology is illustrated through its various applications.

12.2 DNA and RNA Chemical Structure and Properties

The discussions in this chapter necessarily use biological terms, some of which may be unfamiliar to the analytical chemist. These terms will be defined at the start of each section.

12.2 DNA and RNA Chemical Structure and Properties

DNA and RNA are both nucleic acids. Nucleic acids are very large molecules that are made up with a backbone of alternating sugar and phosphate molecules, bonded together in a long polymeric chain. This backbone of the nucleic molecule is represented below: –

sugar –

phosphate

– sugar –

phosphate

– sugar –

phosphate

– sugar –

The polymer is of varying lengths that, depending on its biological function, range from just a few units up to tens of thousands of units. These varying sizes are called nucleic acid fragments. For DNA, the sugar molecule in the backbone is deoxyribose. For RNA, the sugar is ribose. To complete the nucleic acid molecule, nucleotide bases are bonded to each sugar molecule. This is shown below: nucleotide base

nucleotide base

nucleotide base

nucleotide base

|

|

|

|

– sugar –

phosphate

– sugar –

phosphate

– sugar –

phosphate

– sugar –

There are four different types of nucleotide bases. In DNA these are adenine (A), thymine (T), cytosine (C), and guanine (G). A DNA nucleic acid molecule is illustrated below in which the order of nucleic acids is A, T, C and G: A

T

C

G

|

|

|

|

– sugar – phosphate – sugar – phosphate – sugar – phosphate – sugar –

In 1953, Watson and Crick published a paper in the scientific journal Nature describing the structure of DNA. Watson and Crick showed that not only is the DNA molecule double-stranded, but the two strands wrap around each other forming a coil or helix, forming complementary strands. Each nucleotide in one strand is hydrogen bonded to another nucleotide base in a strand of DNA opposite to the original. This bonding is specific. Adenine always bonds to thymine (and vice versa) and guanine always bonds to cytosine (and vice versa). This bonding occurs across the molecule, leading to a double-stranded molecule as shown next:

301

302

12 DNA and RNA Chromatography – sugar – phosphate – sugar – phosphate – sugar – phosphate – sugar – |

|

|

|

T

A

G

C

|

|

|

|

A

T

C

G

|

|

|

|

– sugar – phosphate – sugar – phosphate – sugar – phosphate – sugar –

Double-stranded DNA has the unique ability to make exact copies of itself, i.e. self-replicate. When more DNA is required by a cell (such as during reproduction or cell growth), the hydrogen bonds between the nucleotide bases break and the two single strands of DNA separate. New complementary bases are brought in by the cell and paired up with each of the two separate strands, thus forming two new, identical, double-stranded DNA molecules. A similar process called polymerase chain reaction (PCR) can be performed in the laboratory under in vitro conditions to copy and amplify specific sequences of DNA. The order in which these nucleotide bases (also called the sequence of bases) appear in a particular nucleic acid molecule constitutes the blueprint for the information carried in the molecule. In a living cell, these sequences are contained in genes. The sequence of bases found in a particular gene provides the information for the synthesis of a particular protein. In the laboratory these sequences are called nucleic fragments. They may be called a fragment of x number of base pairs if the fragment is double-stranded, or a fragment of x number of bases if the fragment is single-stranded. Ribonucleic acid, or RNA, gets its name from the sugar group in the molecule’s backbone – ribose. Several important similarities and differences exist between RNA and DNA. Like DNA, RNA has a sugar-phosphate backbone with nucleotide bases attached to it. Like DNA, RNA contains the bases adenine (A), cytosine (C), and guanine (G). However, RNA does not contain thymine (T). Instead, RNA’s fourth nucleotide is the base uracil (U). Both RNA and DNA can exist in doublestranded and single-stranded forms, although RNA is normally single-stranded. RNA is the main genetic material contained in the organisms called viruses. RNA is also important for the production of proteins in other living organisms. RNA can transport within the cells of living organisms, and thus serves as a sort of genetic messenger. Information stored in the cell’s DNA is transported by RNA from the nucleus to other parts of the cell, where it is used to help make proteins.

12.3 DNA and RNA Chromatography

12.3 DNA and RNA Chromatography 12.3.1 Development of DNA and RNA Chromatography

While ion-pairing reverse-phase chromatography is the primary separation method described in this chapter, many different types of chromatographic methods have been used to separate nucleic acid. These include gel filtration, affinity and ion exchange [1–13]. In particular, ion-exchange chromatographic separations of nucleic acids performed on nonporous anion exchangers can be extremely rapid and useful [14–17]. J. A. Thompson and coworkers published many excellent papers in the form of both review articles [18–23] and research papers [24–26]. The review articles published in 1986 and 1987 are a series of six publications, each dealing with some aspect of nucleic acid separation. Taken together, the six review articles present a comprehensive description of the different chromatographic methods for nucleic acid separations. The fundamental technology leading to modern DNA chromatography and later RNA chromatography was first described by Guenther Bonn, Christian Huber and Peter Oefner in 1993 [27–29]. Using ion-pairing reverse-phase chromatography, they obtained rapid, high-resolution separations of both double-stranded and single-stranded DNA. The separations were performed usually in less than 10 min, and in many cases resolution of fragments differing in only a single base pair in length was achieved. This form of HPLC analysis is largely (though not entirely) based upon the unique separation properties of a nonporous polystyrene–divinylbenzene polymer bead that has been functionalized with C18 alkyl groups. An alkylammonium salt, usually triethyl ammonium acetate (TEAA), is added to the eluent to form neutral ion pairs when a DNA sample is introduced into the HPLC instrument. The fragments are adsorbed by the column, and then a gradient of water and acetonitrile is used to separate the DNA fragments. The smaller fragments come off the column first, and larger fragments are then eluted from the column and detected. Figure 12.1 shows a separation of double-stranded DNA obtained by this method. Baseline resolution of fragments up to 100 base pairs long can be achieved and fragments up to 2000 and larger base pairs can also be separated on the column. Bonn, Huber and Oefner showed that the DNA separations were performed according to the size of the fragment, just as they are in gel electrophoresis. Gel electrophoresis is the classical way of separating DNA and is based on using a high electrical potential to pull the DNA fragments through a gel slab sheet. Separations are based on the size of the DNA fragment, with smaller fragments traveling faster through the gel than larger fragments. For double-stranded DNA, Bonn and coworkers showed that the sequence does not contribute to the retention of the fragment on the column; retention is based only on DNA fragment size. Figure 12.2, taken from their work, demonstrates

303

304

12 DNA and RNA Chromatography

Figure 12.1 High-performance DNA chromatography separation of double-stranded DNA using a DNASep® column 50 mm × 4.6 mm. Sample was a mixture of BR322 Hae III restriction digest and UX174 Hinc II restriction

digest. Eluent and gradient: A: 0.1 M TEAA; B: 0.1 M TEAA, 25% acetonitrile, 35 to 45% B in 2 min, 45% to 57% in 10 min., 57% to 61% in 4 min, 1.0 mL min–1 flow rate, UV detection at 254 nm (from Ref. [23] with permission).

this with a plot of retention time vs. fragment size for a number of different fragments. Various plasmids (circular double-stranded DNA of known sequence) were digested with a number of different enzymes to cut the plasmid at specific, known sequences so that mixtures of DNA fragments with different, known lengths and sequences were generated. Since the base sequence of the original plasmids is known, the effect of DNA sequence can be shown. In the plot, all of the retention times of the various fragments fall on a line drawn through the data points. This shows that the retention of DNA is dependent on fragment size, and is independent of fragment sequence. It was shown that the technology could be used to separate single-stranded DNA; later, single-stranded RNA separations were developed. These separations are very rapid with base resolution possible for most short oligo mixtures. Depending on the type of eluent used, single-stranded separations are based on differences in size, polarity and shape of the molecule. By changing the ion-pairing reagent to be more nonpolar, the separation can become mostly size based. Also, discussed later, temperature is an important parameter for single-stranded separations, especially RNA separations.

12.3 DNA and RNA Chromatography

Figure 12.2 The retention times of various plasmid digest fragments are plotted according to size. The fragments have different sequences, but are separated according to their size using a DNASep® column 50 mm × 4.6 mm.

Eluent and gradient: A: 0.1 M TEAA; B: 0.1 M TEAA, 25% acetonitrile, 0% to 100% B in 30 min, 1.0 mL min–1 flow rate, UV detection at 254 nm (from Ref. [27] with permission).

12.3.2 Column Properties

A number of modern HPLC columns have been used for DNA and RNA separations. One example is a silica-based C18 material available from Varian Corporation (Walnut Creek, CA). There have been recent reports on monolith polymeric materials providing excellent separations [30–32]. The remarkable performance of the DNA and RNA separation columns is based on a number of properties including the porosity of the packing material, its polarity, the absence of metal contamination, and the small size and narrow size distribution of the packing material. Bead pores are small relative to the nucleic acid molecule so that column interactions of the nucleic acid with the bead are on the surface of the bead. Polarity of the bead is adjusted so that nucleic acid interactions with the surface are controllable with ion pairing reagents. Most column packings are about 2-lm particles functionalized with a hydrophobic, neutral C18 alkyl group. Polymer-based column materials are used because they are rugged and can withstand extremes in eluent pH and high-temperature operation. The most popular columns, DNASep® and OligoSep™ (Transgenomic, Inc. Omaha, NE), have been cited in more than 1200 journal articles (www.Transgenomic.com). They are 2-lm, C18 surface, nonporous polymeric columns, based on the original work published by Bonn et al. [27–29].

305

306

12 DNA and RNA Chromatography

12.3.3 Ion-pairing Reagent and Eluent

The ion-pairing reagent is an amine cation salt that forms a nonpolar ion pair with the phosphate anion group of the nucleic acid. Triethylammonium acetate (TEAA) is the most common amine cation salt used for this purpose. TEAA will pair with nucleic acid fragments to form a nonpolar ion pair and adsorb to the neutral nonpolar surface of the column. Acetonitrile is gradually added to the eluent, decreasing its polarity until TEAA/nucleic acid ion-pair fragments desorb from the column. In gradient elution, the concentration of acetonitrile that is pumped through the column is increased gradually as the separation proceeds. Smaller fragments elute first, followed by fragments increasing in size as the acetonitrile concentration is increased. DNA and RNA nucleic acid fragments are extremely large molecules relative to what can normally be separated by a chromatographic process. The average molecular weight of one base pair of DNA is approximately 660 g mol–1. As an example, a double-stranded 100 base pair DNA molecule has an approximate molecular weight of 66 000. As many as 1000 and even up to 2000 base pair fragments can be separated by DNA chromatography. It is remarkable that a chromatographic system can achieve base pair resolution for fragment sizes greater than 100 base pairs, with partial resolution up to 200 base pairs. It is possible to use a very rapid gradient program of perhaps 3–5 min, although such a rapid elution program is at the expense of lower resolution of peaks. In many cases, this is adequate because the mixture being separated is not complex. A slower gradient process of 20–30 min will produce the highest resolving conditions and the greatest separation of peaks although obviously this is at the expense of a longer analysis time.

Figure 12.3 Example chromatogram using the optimized siRNA reverse-phase denaturing HPLC purification protocol described here for the WAVE Oligo System (siRNA oligonucleotide: Luciferase Antisense. n-Hexylammonium

acetate ion pairing agent separates primarily based on length rather than base composition. The product purity was less than 50% but was enriched to greater than 95%.

12.4 Temperature Modes of DNA and RNA Chromatography

It was shown that TEAA produces size-based separations for double-stranded DNA. For single-stranded DNA and RNA, separations are not solely size-based when using the TEAA eluent because the polarity of the sequence fragment is not shielded by the ion pairing reagent. However, the use of a more hydrophobic reagent such as tetrabutylammonium bromide can provide a separation that is very close to size-based. Figure 12.3 shows a separation and purification of silencing RNA synthesized material separated on a Transgenomic OligoSep Prep HC cartridge using another hydrophobic reagent, n-hexylammonium acetate as the ion-pairing reagent. Since the separation is primarily size-based, synthesis failures can be removed and the product can be purified to more than 95% purity [33].

12.4 Temperature Modes of DNA and RNA Chromatography

DNA exists in both double-stranded and single-stranded forms. To a certain extent, RNA can exist in both forms, but it is more likely to be single-stranded with some double-stranded secondary structure. The temperature of the column and the fluid entering the column can be thought of as an additional reagent in the separation of nucleic acids. Temperature controls whether the nucleic acid is separated as a single molecule, double-stranded molecule or something in between. Double stranded nucleic acids are held together by hydrogen bonding of the two strands. As the temperature is increased, these bonds are broken, making two strands of single-stranded nucleic acid. Temperature is used to achieve the three modes of operation: nondenaturing mode, fully denaturing mode, and partially denaturing mode. 12.4.1 Nondenaturing Mode

The nondenaturing mode of DNA and RNA chromatography is used to separate double-stranded nucleic acids. This is most often used for DNA, but some short sequences of RNA can also be double-stranded. Obviously, it is important to keep the double strand intact while performing these separations. The breaking of hydrogen bonds of double-stranded DNA by increasing the temperature is called melting or denaturing. The temperature at which this occurs depends on the strength of the hydrogen bonding and the environment around the DNA. Higher salt and buffer content will raise the melting temperature. Also, DNA adsorbed onto a solid surface, such as column packing material, will require a higher temperature to melt the DNA. Conversely, the presence of an organic solvent such as acetonitrile will lower the melting temperature, and the effect is increased with increasing concentrations of the solvent. In the nondenaturing mode, a sufficiently high temperature is chosen to lower eluent viscosity (and therefore column back pressure), but not so high that denaturing occurs. The normal oven temperature of operation for nondenaturing mode is 50 °C.

307

308

12 DNA and RNA Chromatography

12.4.2 Fully Denaturing Mode

Single-stranded DNA and RNA may contain secondary structure due to complementary sequences within its own fragment. If portions of sequences are complementary, the fragments may fold back on themselves through the formation of intra-molecular hydrogen bonds, giving several different possible structures for the same fragment. The presence of these secondary structures is uncertain and nonreproducible and will lead to peak broadening if they occur. Increasing the temperature can break up the hydrogen bonding and therefore reduce some or perhaps most of the secondary structure [34]. Figures 12.4a and 12.4b show the separation of an RNA ladder at 40 and 75 °C. The peaks of the RNA become sharper and more uniform with increasing oven temperature. It is not certain that temperature will always reduce secondary structure. In these cases, using several different temperatures for the same sample can give clues as to the nature of stronger secondary structure.

Figure 12.4a Chromatogram of RNA ladder at 40 °C. RNA ladder (Cat. No. 15623010, Life Technologies) has nucleotide lengths of 155, 280, 400, 530, 780, 1280, 1520, and 1770 bases. Buffer A: 1 M TEAA, pH 7.0, buffer B: 1 M

TEAA, pH 7.0 with 25% v/v acetonitrile, gradient 0.0 min 38% B, 1.0 min 40%, 16 min 60%, 22 min 66%, 22.5 min 70%, and 23 min 100%. (from reference [34] with permission).

12.4 Temperature Modes of DNA and RNA Chromatography

Figure 12.4b Same conditions as Figure 12.4a except chromatogram is at 75 °C. It is not certain that temperature will always reduce secondary structure.

12.4.3 Partially Denaturing Mode

The use of partially denaturing mode can detect mutations in DNA through a heteroduplex detection process. Remember that the hydrogen bonding holding two complementary strands of DNA together is specific. The adenine base always bonds to thymine (and vice versa) and guanine always bonds to cytosine (and vice versa). If the two strands of a DNA fragment are perfectly matched, then they are hydrogen bonded at each and every nucleic acid site. They are said to be a homoduplex nucleic acid fragment because both strands of the DNA are completely complementary to the other. A heteroduplex nucleic acid fragment arises where a genetic mutation has occurred. One of the bases has mutated so that a doublestranded fragment is not completely complementary but now contains a base that cannot hydrogen-bond to the base located on the other fragment. Under a partially denaturing mode, differences in melting of double-stranded DNA are detected because increased melting decreases the retention time of the fragment. The operating temperature of the column and eluent is chosen to partially denature or melt the fragment to enhance this mismatch. Since singlestranded DNA elutes differently from double-stranded DNA, a mixture containing both homoduplex fragments and heteroduplex fragments can be separated. Typically, oven temperatures between 54 and 72 °C are used. The method can be called a ‘difference detecting engine’ because it detects the presence of a heteroduplex regardless of the sequence being studied. This type of chromatography is sometimes called DHPLC or denaturing HPLC. More details on how DHPLC works are described later in an example of the detection of DNA mutations.

309

310

12 DNA and RNA Chromatography

12.5 Instrumentation

The components of the HPLC system are quite similar to what is used for standard ion chromatography with some important refinements in the general flow path, oven, and detector. In many cases, the nucleic acid that is separated is collected in a fragment collector to be used for further research. These instrument refinements are described in this section. 12.5.1 Effect of Metal Contamination

In the course of commercializing the DNASep® column technology, a degradation effect was discovered after the column had experienced medium to longer term use [35–38]. The DNASep® column would work for a while and then would start giving either split peaks for each fragment eluting from the column or stop giving peaks at all. At first, the packing procedure was extensively studied because split peaks mean that the chromatographic bed has become cracked or disrupted. This was not unexpected since polymer beds are prone to this kind of behavior if they are not packed properly. The packing procedure was studied to rule out the possibility of unstable packed column beds. Then, the column was put into a standard HPLC system that destroyed the separation slowly and somewhat controllably. In one set of experiments, a new column showed excellent separation of a pUC 18 Hae III digest (Figure 12.5a). As the column was used, a degradation effect was observed as a loss of resolution for base pairs greater than 200 (Figure 12.5b). As the degradation continued, increasingly shorter fragments of DNA were affected. Many of the peaks were split or doublets (Figure 12.5c). Eventually, the DNA did not elute at all from the system. Thus, the degradation or decreasing resolution appeared to be a function of the length of the polynucleotide fragment being separated. The cause of the split peaks for the larger DNA fragments was discovered when attempts were made to clean the column. Clean-up with organic solvents did not improve performance. However, subsequent clean-up of the column with injections of tetrasodium ethylenediaminetetraacetic acid (EDTA), a metal-chelating agent, largely restored chromatographic resolution. It appeared that the chelating reagent passivated the metal in the system by removing surface oxidized metals so that they could not bind with the DNA. Adding small amounts (i.e., 0.1 mM) of tetrasodium EDTA to the mobile phase can be done without significant changes to the chromatography. The most significant sources of metal ions are HPLC components containing fritted filters made of stainless steel. Fritted filter components are used in mobile phase filters, check valve filters, helium spargers, mobile phase mixers, inline filters, column frits, and other parts of the HPLC.

12.5 Instrumentation

Figure 12.5a This is the first of a series of three chromatograms showing a separation, illustrating the degradation effect of running an HPLC releasing contamination over time to the column. The separation is of doublestranded DNA, pUC18 Hae III. The degradation occurs first on larger fragments.

Smaller fragments are eventually degraded as well. This first chromatogram uses a new DNASep® column, Buffer A: 1 M TEAA, pH 7.0, buffer B: 1 M TEAA, pH 7.0 with 25% v/v acetonitrile, gradient, 0 min 35% B, 3 min 55% B, 10 min 65% B, and 12 min 100 % B.

Figure 12.5b After several minutes of use, the column shows the initial degradation of the HPLC separation with fragments greater than about 200 bp affected.

311

312

12 DNA and RNA Chromatography

Figure 12.5c After several more minutes of use, the column shows the greater degradation of the HPLC separation. Now all fragments are affected (from Ref. [35] with permission).

Metal ion contamination, such as colloidal iron, can be released from frits, travel to other parts of the HPLC, and then be trapped. These types of contaminants will interfere with DNA in solution or after it has been released and trapped on a critical component of the HPLC such as the column, an inline filter in front of the detector, or at a back-pressure device located after the detector. Iron(III) and other metals can form insoluble complexes with phosphate anion. It is likely that surface metal ions and/or colloidal iron are combining with one or more phosphate groups on the nucleic acid fragments. If this happens, the ionpair process chromatographic separation can be interrupted, causing the peaks to broaden. In extreme cases, it is possible that this metal contamination is so severe that nucleic acid fragments are completely prevented from eluting from the system and no peaks are detected. Subsequent experiments showed that even if titanium or PEEK components are used in the fluid path, then some treatment was necessary before the components could be used. Although an improvement, the use of titanium frits did not initially give consistent results. However, treatment of the frits with dilute nitric acid and then with a chelating agent did improve the performance of the instrument. Similarly, as shown in the examples, PEEK frits were not consistently suitable for DNA chromatography, but acid treatment did improve their performance. Finally, degassing the fluid before it enters the liquid chromatography system in order to remove oxygen will inhibit oxidation and hence production of metal ions in stainless steel or titanium or other tubing containing iron. All of these improvements are incorporated into commercial instruments designed for nucleic acid chromatography. It is also important to follow preventive maintenance procedures and check filters for signs of colored deposits. Since double-stranded DNA is more susceptible to contamination, the use of precautions with respect to the method and the system is much more critical than when the

12.5 Instrumentation

system is used to separate single-stranded DNA. Failure to keep the column clean is probably the most common error that the user can make. More details on procedures and recipes for instrument and column maintenance are described in a book entitled, ‘DNA Chromatography’ by Gjerde et al. [39]. 12.5.2 The Column Oven

The oven controls the temperature of the fluid entering the separation column. This includes not only the eluent but the sample that is injected into the system. The fluid entering the oven compartment is normally cooler than the oven and there is a time lag before this new fluid comes to the oven temperature. In most HPLC ovens, the fluid never does reach the set point of the oven. In DNA and RNA separation instrumentation, a pre-heat tube is used to bring the fluid to column temperature before it reaches the column. The oven temperature should be accurate, should not drift and should be precise, i.e. it should come to the same temperature each time it is directed by the run method to go to a specific temperature. The oven remains one of the most critical and difficult parameters to control in the DNA chromatograph. The reader should consult the HPLC manufacturer for information on oven use, calibration and upkeep. Selerity Technologies (Salt Lake City, UT) has developed an active eluent preheater that uses a feedback technology to maintain a set temperature for the eluent prior to entering the column. This unit appears to be compatible with different HPLC instrumentation. 12.5.3 UV and Fluorescence Detection

Nucleic acids absorb light strongly in the UV, the absorption maximum being at a wavelength of 260 nm. Variable-wavelength detectors are set at 260 nm for detection; however, single-wavelength detectors work very well at 254 nm. The ion-pairing reagent does not absorb at these wavelengths, and the nucleic acid fragments may be detected directly at the sub-nanogram level by UV automatic detection. Fluorescence detection could also be used provided that fluorescence tags are added to the DNA. DNA is tagged by using special fluorescently tagged nucleotides when the fragments are being synthesized. Use of this detection method decreases the amount of DNA that can be detected by a factor of 10 to 100 (or even 1000 in some reported cases). Common dyes used in molecular biology include FAM, TET, HEX, TAMRA, NED, Pacific Blue, and many others. A table of fluorescent tags can be found in Chapter 4.

313

314

12 DNA and RNA Chromatography

12.5.4 Fragment Collection

One of the powerful features of DNA and RNA chromatography is that material can easily be purified by collecting directly from the detector effluent. Purification can be used in biological samples, where there may be several types of nucleic acids present and a particular type is desired for study. An example of this is shown later in this chapter where several types of RNA are detected in a cell extract and can be purified. The collection can be done by hand, but is normally accomplished with an automated fragment collector and controlling software. Collection can be into single vials or large 96-well plates. Care should be taken to ensure that the actual peak has been collected. Measurement of recovery is performed by taking a small portion of the recovered peak, reinjecting and measuring the area, multiplying this by the ratio of total collected volume to reinjected volume, and comparing this value to the area of the original peak. Normal recoveries are about 80% of injected material. Of course materials such as RNA may degrade (by enzymes) after they have been collected. Higher recoveries may not be possible owing to loss of material during the concentration process through precipitation or plating of material on surfaces. An important parameter in fragment collection is careful execution of the collection times. There is a lag from the time the peak is seen in the detector cell and when the fragment is deposited from the fragment collector probe. The most reliable collection method is the timed collection. But unless the timing is correct, it is easily possible to miss some of the peak or even the entire peak. It is also important that there is as little dead volume in the tubing from the detector cell outlet to the tip of the deposition probe. Too large a dead volume will destroy the resolution of the separation and could result in cross contamination of the peak of interest with neighboring peaks. It is important that the probe is cleaned between collection of peaks. This is normally done automatically by the fragment collector.

12.6 Applications of DNA Chromatography 12.6.1 DHPLC

The use of column temperature to control separations in DNA chromatography was described in 1996 through the insights of Oefner and Underhill [40–43]. They demonstrated that DNA chromatography possessed unique properties, enabling the separation of DNA based on its relative degree of helicity. Heteroduplex DNA has a lower melting point than homoduplex DNA. The retention of singlestranded DNA is lower on the column than that of double-stranded DNA. Heteroduplex DNA melts or denatures more easily, and it elutes earlier in the separation. This technique is called denaturing HPLC (DHPLC).

12.6 Applications of DNA Chromatography

All individuals have mutations in their DNA that have spontaneously occurred in the past (due to chemical or radiation exposure) and have been passed on from generation to generation. Some of these mutations are serious and may cause disease, but most are benign. Mutations in an individual can be found by comparing specific DNA fragments of that specific individual to the same fragment sequence in the general population. Usually, DNA fragments of about 600 base pairs are examined in genes that are known to be responsible for various types of diseases. If a mutation is present, a heteroduplex is formed and can be detected. If a mutation is not present only nonmutated homoduplexes are detected. Details of the fragment selection process are beyond the scope of this chapter and can be found in the Oefner and Underhill references. An excellent flash movie showing how homoduplex and heteroduplex species are formed and then detected by DHPLC can be viewed by clicking the DHPLC icon on the web page following the link www.transgenomic.com/ap/VariationApp4.asp. 12.6.2 Nucleic Acid Enzymology

Many chemical reactions occur within biological cells. A special class of protein molecules called enzymes speed up or control these reactions. Chemical reactions need a certain amount of activation energy to take place, and enzymes can increase reaction speed by allowing a different reaction path with a lower activation energy, making it easier for the reaction to occur. Enzymes are essential for the function of cells. They are very specific to the reactions they catalyze and the chemical substrates that are involved in the reactions. Enzymes are said to fit their substrate like a key fits its lock. The control of the expression of proteins by DNA genes relies on special proteins that bind to specific DNA sequences. The way in which these proteins recognize their binding sites in the genome is an important step in understanding how these processes occur at the molecular level. Nucleic acid enzymology is the investigation of the properties of enzymes that act on or with nucleic acids. These enzymes include DNA polymerases (for replicating or copying DNA), restriction endonucleases (for base cutting at a specific known base sequence) and reverse transcriptases (for transcribing the information contained in DNA). Chromatography, coupled with fluorescence detection, can be a generic analytical platform for nucleic acid enzymology.

12.6.2.1 Telomerase Assays Telomerase is a ribonucleoprotein complex that plays a critical role in cellular mortality [44]. The vertebrate enzyme catalyzes the addition of TTAGGG sequence repeats to the ends of chromosomes. In the absence of telomerase, human telomeres undergo progressive shortening with each round of cell division, an event that may contribute to cell aging and mortality. Telomerase is known to be asso-

315

316

12 DNA and RNA Chromatography

ciated with immortalized cancer cells but is absent in most normal tissues and has therefore become a focus for diagnostic investigation. The presence of telomorase can be measured by the chromatographic full-denaturing measurement of extension products. If telomerase is present, then the DNA fragments are extended to longer fragments. The amount of telomerase present is indicated by the extent to which the fragments have been extended.

12.6.2.2 Polynucleotide Kinase Assays Polynucleotide kinase catalyzes the addition of a phosphate group to DNA and RNA. The enzymatic phosphorylation of synthetic single-stranded DNA can be applied to experimental molecular biology [45]. Radioactive (32P) phosphate groups are used to make the single-stranded probes that can be incorporated into biological reactions. Measurement of the proportion of probe that has been successfully labeled is normally difficult to determine. However, the subtle ionic difference between phosphorylated and unphosphorylated DNA is readily resolved by full denaturing DNA chromatography in fragments up to 50 bases in length. Figure 12.6 shows the effects of phosphorylation of a single-stranded oligodeoxynucleotide upon retention time following DNA chromatography. Phosphorylation of the oligodeoxynucleotide reduces the retention time of the modified DNA, because the addition of the polar phosphate group diminishes the net hydrophobicity of the DNA. These data are readily corroborated by mass spectrometry [46].

Figure 12.6 The use of DNA chromatography in the resolution of differentially phosphorylated oligodeoxynucleotides under denaturing conditions. The difference in polarity between the two species is clearly sufficient to resolve the strands, which may be up to 50 to 100

nucleotides in length. Confirmation of the products was obtained by mass spectrometry [44] which is a powerful complementary technique to DNA chromatography since the eluted material is ideal for mass spectrometry immediately post column.

12.7 Applications of RNA Chromatography

12.6.2.3 Uracil DNA Glycosylase Assays Uracil DNA glycosylase (UDGase) is an enzyme that removes uracil from both singlestranded and double-stranded DNA [47]. The normal RNA base, uracil, can arise in DNA spontaneously. Uracil in DNA is pro-mutagenic and can disrupt the binding of sequence-specific gene regulatory proteins. All DNA-containing organisms contain a specific repair pathway which removes uracil from DNA with UDGase enzyme. A novel assay was developed using full denaturing DNA chromatography [45]. After treatment with UDGase when uracil is present in the DNA strand, there is a significant shift in retention time of the modified single-stranded DNA. Removal of the uracil base from the single-stranded DNA results in elution at an earlier retention time through loss of the uracil base and subsequent decrease in hydrophobicity of the DNA.

12.7 Applications of RNA Chromatography

RNA is found in a repertoire of intra- and inter-molecular hydrogen bonding interactions. One familiar structural element in an RNA molecule is the stem-loop, in which a noncomplementary segment separates two complementary stretches of nucleotides. This structural diversity is characteristic of biological RNA. The retention of RNA molecules on a chromatography column is a function of chain length, surface chemistry, morphology and localized single- and double-stranded character. Even with thorough thermal denaturation, it is unlikely that a given population of RNA molecules is completely free from secondary and tertiary interactions. Thus, the chromatography of RNA will be different and more unpredictable compared to the chromatography of single-stranded and double-stranded DNA. Still, RNA chromatography is best performed at elevated temperatures to denature as much as possible and reduce secondary of the RNA being separated. There are several types of RNA. Silencing RNA (siRNA) is a 20–25 nucleotide long double-stranded nuclear RNA studied in biological research and has high potential in new drugs. Silencing RNA gets its name because of its ability to ’silence’ or stop a protein from being produced. The sequence of the siRNA is chosen so that a strand of the RNA molecule can hydrogen-bond with the targeted sequence of a particular gene, which produces the desired effect. Research in RNA in gene control is increasing rapidly and it is certain that many methods will be developed. Other RNA species include tRNA (transfer RNA), mRNA (messenger RNA) and rRNA (ribosomal RNA). A typical RNA chromatography separation of a total RNA extraction obtained under full denaturing conditions is shown in Fig 12.7 [34]. The earliest eluting species include the population of tRNAs (and probably includes small nuclear RNAs), the middle section of the profile is dominated by the rRNA species, and finally, underlying the entire chromatogram, is a spectrum of mRNAs with many of the fragments centered on the later retention times.

317

318

12 DNA and RNA Chromatography

Figure 12.7 RNA chromatography of a total cellular extract of RNA from tobacco plants is displayed. The large, broad peak eluting between 12 and 15 min is primarily rRNA together with mRNA.

12.7.1 Separation of Messenger RNA from Ribosomal RNA

The cellular processes that are expressed in a living cell are the direct result of the proteins contained in the cell. This in turn is closely linked to the population of mRNA molecules produced by the genes contained in a given cell type. For this reason there has been considerable interest in evaluating qualitatively and quantitatively the various types of RNA contained in a cell. One of the key experimental procedures is the systematic removal of rRNA from the mRNA fraction. Typically, mRNA has a long poly A tail that can be captured on a poly T chromatography column or batch resin. The rRNA and tRNAs pass through the column and, after a suitable washing protocol, the mRNA fraction is concentrated and stored for subsequent experimentation. The effectiveness of the purification of a typical mRNA purification scheme was followed by RNA chromatography by Hornby and coworkers [34]. It was found that at least two rounds of enrichment are usually required in order to remove the bulk of the nonpoly A RNA. It is possible to apply RNA chromatography in a preparative mode in order to produce a series of RNA fractions which can be used for subsequent research. One of the major drawbacks of poly A mRNA isolation is that some mRNAs are not polyadenylated and therefore will be excluded from any subsequent analysis. The use of RNA chromatography in a preparative mode offers the potential for isolating at least a fraction of those mRNA species that do not co-elute with rRNA. This is clearly an important area for development, since the analysis of cell-specif-

12.7 Applications of RNA Chromatography

ic RNA populations in disease is becoming increasingly important in molecular medicine. 12.7.2 Analysis of Transfer RNA

Each cell contains a population (usually referred to as a pool) of tRNAs that meet the requirements of that particular cell’s (or in the case of bacteria, that organism’s) protein synthesis machinery. The range of sizes of tRNA molecules is particularly narrow compared with mRNA, between around 60 and 100 bases, and a given cell typically contains around 100 species. While this molecular weight range is ideal for RNA chromatography, the close similarity of molecular sizes represents a problem for resolving individual components in a typical cellular pool. This is readily seen in Figure 12.8 where the total pool of tRNAs from Escherichia coli has been separated by RNA chromatography [34, 48].

Figure 12.8 RNA Chromatogram of the entire pool of tRNAs purified from E. coli. By coupling separation with RTPCR it is possible to identify and quantify the individual tRNAs. A typical population of tRNAs will contain around 50 species.

319

320

12 DNA and RNA Chromatography

References [1] H. Ellegren and T. Laas, Size-exclusion

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

chromatography of DNA restriction fragment. Fragment length determinations and a comparison with the behavior of proteins in size-exclusion chromatography, J. Chromatogr. 467, 217, 1989. G. Bernardi, Chromatography of nucleic acids on hydroxyapatite column, in: L. Grossman and K. Moldave (eds.), Methods in Enzymology, Vol. 21, Academic Press: New York, pp 92–140, 1971. M. Soyab and Y. Sen, The isolation of extrachromosomal DNA by hydroxyapatite chromatography, in Methods in Enzymology (ed. R. Wu ), Vol. 68, Academic Press: New York, pp 199–206, 1979. A. D. Kelmers, G. D. Novelli and M. P. Stulberg, Separation of transfer ribonucleic acids by reverse phase chromatography, J. Biol. Chem., 240, 3979, 1965. A. D. Kelmers, H. O. Weeren, J. F. Weiss, P. S. Pearson, M. P. Stulber and G. D. Novelli, Reversed-phase chromatography systems for transfer ribonucliec acid – preparatory scale methods, in Methods in Enzymology (eds., K. Moldave and L. Grossman) Vol. 20, Part C, Academic Press: New York, pp 9–34, 1971. R. L. Pearson, J. F.Weiss and A. D. Kelmers, Improved separation of transfer RNAs on polychlorotrifluoroethylene-supported reversed-phase chromatography columns, Biochem. Biophys. Acta, 228, 770, 1971. R. D. Wells, S. C. Hardies, G. T. Horn, B. Klein, J. E. Larson, S. K. Neuendorf, N. Panayatatos, R. K. Patient, and E. Selsign, PRC-5 column for the isolation of DNA fragments, in Methods in Enzymology (eds. L. Grossman and K. Moldave.), Vol. 65, Academic Press, New York, pp 327–347, 1980. J. A. Thompson and R. W. Blakeley, Higher performance liquid chromatography purification of nucleic acids, Fed. Proc. 42, 1953, 1983. R. D. Wells, High performance liquid chromatography of DNA, J. Chromatogr., 336, 3, 1984.

[10] R. A. Hardwick and P. R. Brown, The

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

performance of micro particle chemically-bonded anion-exchange resins in the analysis of nucleotides, J. Chromatogr., 112, 650, 1975. E. Westman, S. Eriksson, T. Laas, P. Parental and S. E. Scold, Separation of DNA restriction fragments by ionexchange chromatography on FPLC columns Mono P and Mono Q, Anal. Biochem., 166, 158, 1987. F. Flack and V. Sarantoglou, Curved DNA fragments display retarded elution upon anion exchange HPLC, Nucleic Acids Res., 19, 4181, 1991. W. Bloch, Precision and accuracy of anion-exchange separation of nucleic acids, U. S. Pat. 5 856 192, 1999. E. D. Katz and M. W. Dong, Rapid analysis and purification of polymerase chain reaction products by high-performance liquid chromatography, Biotechniques, 8, 546, 1990. E. D. Katz, Protocol: Quantitation and purification of polymerase chain reaction products by high-performance liquid chromatography, Molec. Biotechnol, 6, 79, 1996. E. D. Katz, L. A. Haff and R. Eksteen, Rapid separation, quantitation and purification of products of polymerase chain reaction by liquid chromatography, J. Chromatogr. 512, 433, 1990. J. M. Wages, X. Zhao and E. D. Katz, High-performance liquid chromatography analysis of PCR products, PCR Strategies, Academic Press: San Diego, pp 140–153, 1995. J. A. Thompson, A review of high performance liquid chromatography in nucleic acids research I. Historical perspectives, Biochromatography, 1, 16, 1986. G. Zon and J. A. Thompson, A review of high performance liquid chromatography in nucleic acids research II. Isolation, purification, and analysis of oligodeoxyribonucleotides, Biochromatography, 1, 22, 1986. J. A. Thompson, A review of high performance liquid chromatography in

References

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

nucleic acids research III. Isolation purification, and analysis of supercoiled plasmid DNA, Biochromatography, 1, 68, 1986. J. A. Thompson, A review of high performance liquid chromatography in nucleic acids research IV. Isolation, purification, and analysis of DNA restriction fragments, Biochromatography, 2, 4, 1987. J. A. Thompson, A review of high performance liquid chromatography in nucleic acids research V. Nucleic acid affinity techniques in DNA and RNA research, Biochromatography, 2, 68, 1987. J. A. Thompson, S. Garfinkel, R. B. Cohen and B. Safer, A review of high performance liquid chromatography in nucleic acids research VI. Nucleic acid affinity techniques in DNA-binding Protein research, Biochromatography, 2, 166, 1987. J. A. Thompson, R. W. Blakesley, K. Doran, C. J. Hough and R. D. Wells, Purification of nucleic acids, by RPC-5 analog chromatography: peristaltic and gravity-flow application. Methods in Enzymology, Vol. 100 Academic Press: New York, pp 368-399, 1983. A. P. Green, J. Burzynski, N. M. Helveston, G. M. Prior, W. H. Wunner and J. A. Thompson, HPLC purification of synthetic oligodeoxyribonucleotides containing baseand backbone- modified sequences, Biotechniques, 19, 836, 1995. A. P. Green, G. M. Prior, N. M. Helveston, B. E. Taittinger, X. Liu and J. A. Thompson, Preparative purification of supercoiled plasmid DNA for therapeutic applications, Biopharm., pp 52-–62, May 1997. C. G. Huber, P. J. Oefner, and G. K. Bonn, Rapid analysis of biopolymers on modified non-porous polystryrene – divinylbenzene particles. Chromatographia, 37, 653, 1993. G. Bonn, C. Huber and P. Oefner, Nucleic acid separation on alkylated nonporous polymer beads, U.S. Pat. 5 585 236, 1996. C. G. Huber, P. J. Oefner, E. Preuss and G.K. Bonn, High-resolution liquid chro-

matography of DNA fragments on highly cross-linked poly(styrene-divinylbenzene) particles, Nucleic Acids Res. 21, 1061, 1993. [30] L. Trojer, S. H. Lubbad, C.P. Bisjak, W. Wieder and G. K. Bonn, Comparison between monolithic conventional size, microbore and capillary poly (p-methylstyrene-co-1,2-bis(p-vinylphenyl)ethane) high-performance liquid chromatography columns Synthesis, application, long-term stability and reproducibility, J. Chromatogr A. 1146(2), 216–24, 2007. [31] T. A. Jakschitz, C. W. Huck, S. Lubbad and G. K. Bonn, Monolithic poly[(trimethylsilyl-4-methylstyrene)-cobis(4-vinylbenzyl)dimethylsilane] stationary phases for the fast separation of proteins and oligonucleotides, J. Chromatogr A. 1147(1), 53–58, 2007. [32] W. Wieder, C. P. Bisjak, C. W. Huck, R. Bakry and G. K. Bonn, Monolithic poly(glycidyl methacrylate-co-divinylbenzene) capillary columns functionalized to strong anion exchangers for nucleotide and oligonucleotide separation, J. Sep. Sci. 16, 2478–2484, 2006. [33] Application note AN120, Optimized purification of siRNA oligonuclotides using the WAVE® Oligo System Transgenomic Inc. Omaha, NE. [34] D. T. Gjerde, D. P. Hornby, C. P. Hanna, A. I. Kuklin, R. M. Haefele and P. D. Taylor, Method and system for RNA analysis by matched ion polynucleotide chromatography, U. S. Pat. 6 576 133, 2003. [35] D. T. Gjerde, R. M. Haefele and D. W. Togami, Method for performing polynucleotide separations using liquid chromatography, U. S. Pat. 6 017 457, 2000. [36] D. T. Gjerde, R. M. Haefele, and D. W. Togami, System and method for performing polynucleotide separations using liquid chromatography, U. S. Pat. 5 772 889, 1998. [37] D. T. Gjerde, R. M. Haefele and D. W. Togami, Apparatus for performing polynucleotide separations using liquid chromatography, U. S. Pat. 6 030 527, 2000. [38] D. T. Gjerde, R. M. Haefele and D. W. Togami, Liquid chromatography

321

322

12 DNA and RNA Chromatography

[39]

[40]

[41]

[42]

[43]

systems for performing polynucleotide separations, U. S. Pat. 5 997 742, 1999. D. T. Gjerde, C. P. Hanna and D. Hornby, DNA Chromatography, Wiley-VCH, Weinheim, 2002. P. J. Oefner and P. A. Underhill, Detection of nucleic acid heteroduplex molecules by denaturing high-performance liquid chromatography and methods for comparative sequencing, U. S. Pat. 5 795 976, 1995. P. J. Oefner and P. A. Underhill, DNA Mutation detection using denaturing high-performance liquid chromatography, in Current Protocols in Human Genetics (eds N. C. Dracopoli, J. L. Haines, B. R. Korf, D. T. Moir, C. C. Morton, C. E. Seidman, J. G. Seidman, D. R. Smith), Wiley & Sons: New York, p 7.10.1 – 7.10.12, 1998. P. A. Underhill, L. Jin, A. A. Lin, S. Q. Mehdi, T. Jenkins, D. Vollrath, R. W. Davis, L. L. Cavalli-Sforza and P. J. Oefner, Detection of numerous Y chromosome biallelic polymorphisms by denaturing high performance liquid chromatography, Genome Res. 7, 996, 1997. P. A. Underhill, L. Jin, R. Zemans, P. J. Oefner and L. L. Cavalli-Sforza, A

pre-columbian Y chromosome-specific transition and its implications for human evolutionary history, Proc. Natl. Acad. Sci. 93, 196, 1996. [44] G. B. Morin, The human telomere terminal transferase enzyme is a ribonucleoprotein that synthesizes TTAGGG repeats, Cell, 59, 521, 1989. [45] J. Sambrook, E. F. Fritsch and T. Maniatis, in Molecular Cloning: A Laboratory Manual, 2nd edn, Cold Spring Harbor Laboratory Press, New York: 1989. [46] M. J. Dickman, M. M. Matin and D. P. Hornby, High throughput analysis of nucleic acid modification reactions using IP RP HPLC, Anal. Biochem. 301, 290, 2002. [47] G. Slupphaug, I. Eftedal, B. Kavli, S. Bharati, N. M. Helle, T. Haug, D. W. Levine and H. E. Krokan, Properties of a recombinant human uracilDNA glycosylase from the ung gene and evidence that ung encodes the major uracil-DNA glycosylase, Biochemistry, 34, 28, 1995. [48] M. J. Dickman, M. Conroy, J. Grasby and D. P. Hornby, RNA footprinting analysis using ion pair reverse phase liquid chromatography, RNA, 8, 247, 2002.

323

13 Sample Pretreatment 13.1 Dilute and Shoot or Pre-treat the Sample?

One of the strengths of ion chromatography is the simplicity and ruggedness of the analytical methods developed from its technology. The chromatograms generated can be extremely reliable in that exactly the same peak profile is generated from a given sample whether it is injected once or a dozen times. In many cases very little is done to prepare the sample prior to injection. One hears the phrase ‘just dilute and shoot’. Nothing else is needed other than to dilute the sample with de-ionized water, mix and inject the sample. But sometimes more may be needed, either to protect expensive columns or to make the sample detectable. It is possible that a single injection of the wrong sample can destroy a column through contamination or plugging. Or perhaps a sample component must be enhanced to increase the concentration to a detectable level. Sometimes more is needed than ‘just dilute and shoot’. This chapter describes the various sample pretreatment methods, why they are used and where they are used. Sample pretreatment methods can be classified depending on need or effect: 1. Remove particulate (that may plug the column) (a) Pre-injection removal (syringe filter) (b) In-line column filter (before the separation column). 2. Remove column contaminating material or particulate with a guard column. 3. Enhance or enrich the concentration to improve the detection limit. 4. Make the sample detectable by a process that collects the sample or converts the sample to a detectable form. 5. Remove an interfering ion (remove an ion that co-elutes with the sample ion). Whatever is done to pre-treat or prepare the sample for analysis, it must be done while preserving the integrity of the sample. New contaminants must not be added. Sample ions must not be lost or the chemical form changed. They must Ion Chromatography, 4th Ed. James S. Fritz and Douglas T. Gjerde Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32052-3

324

13 Sample Pretreatment

not be oxidized or reduced to a different species. In that sense, just dilute and shoot is attractive because sample integrity is most likely held true in the injection process. But with proper method design, sample integrity is held with sample pretreatment as well.

13.2 Particulate and Column-contaminating Matter

The sample solutions injected into an ion chromatographic system must be free of particulate matter to avoid plugging of the capillary connecting tubing and the frits at the head of the analytical column. Even samples that appear to be clear may contain unsuspected fine particles. Disposable membrane syringe filters with a pore diameter of 0.45 lm are sufficient in most cases to clean a sample from harmful particulates. Samples with biological activity are filtered through aseptic filters with a pore diameter of 0.22 lm to avoid any change in sample composition due to bacterial oxidation or reduction. It is important to note that these filters may contain ionic contamination even if certified for use in ion chromatography applications. In general, membrane filters described below should be prerinsed with de-ionized water prior to use to avoid sample contamination. Then the sample is filtered, discarding the first portion that comes out of the filter that is diluted with the de-ionized water. Inline filters are very effective to protect columns from particulate material. These filters have low dead volume so that they can be placed anywhere after the injector and before the column. The filters are usually replaceable and cost effective to replace. The filter bodies and filter inserts can be metal or polymer. Metal filters are usually made of 316 stainless steel and can be used even if the rest of the chromatographic is plastic. It is a good idea to have an inline filter regardless of whether samples are pre-filtered or not. Solid or semi-solid samples may require extraction with an aqueous solution to isolate the ionic components in a form suitable for IC. The actual procedures vary widely, depending on the type of sample. For example, meat and sausage products to be analyzed for nitrate and nitrite are first homogenized mechanically, extracted with a 5% borax buffer solution in a hot water bath, and then subjected to a precipitation with strong solutions of potassium hexacyano ferrate and zinc sulfate. The aqueous extracts are diluted further with de-ionized water and filtered through a membrane prior to injection. Organic substances in the sample matrix may interfere with ion chromatographic separations. In some cases it is sufficient to add enough methanol or other organic solvent to completely dissolve the organic matter. But in samples that are soluble in water alone, the organic components may be adsorbed by the IC column packing and prevent reproducible results. This often occurs with dyes that are added to many commercial products. Dyes and many other types of soluble organic compounds can usually be removed by some form of solid-phase extraction (SPE) without altering the inor-

13.3 Preconcentration

ganic ion content of the sample. In simple cases an SPE cartridge or membrane disk is attached to the sample syringe. Then the liquid sample is injected through the SPE material into the ion chromatograph. SPE cartridges are attached to the sample syringe by means of a Luer tip. A solid disk can be cut from a larger SPE disk, such as the Empore disks produced by the 3M Co., and fitted tightly inside the sample syringe. A convenient semi-micro device for SPE with a syringe has been described [1]. Many methods are available for removal of organic material from aqueous samples by off-line SPE [2]. Hydrophobic organic material is best extracted by solid poly(styrene–DVB) polymers or reversed-phase silica extractants. Polyvinylpyrrolidone (PVP) is an appropriate choice for removal of humic acids, lignins and tannins from water samples. As stated before in connection with syringe filters, one should be careful of ion contamination with SPE columns. The columns should be pre-rinsed with de-ionized water. The last line of defense is the guard column, a short column that is configured directly in front of the separation columns. The guard column mimics the separation column so that whatever detrimental event would happen to the separation column happens instead to the less expensive, replaceable guard column first. In most cases, the guard column uses exactly the sample resin as the separation column (although in rare cases the particles may be a little larger). If one understands the sample and column chemistry well, it is possible to use a guard column that has a different chemistry. For example, a cation-exchange chemistry might be used in front of an anion-exchanger separation column to remove heavy metals from the sample. This might be useful to prevent certain metals from hydrolyzing and precipitating in the high-pH environment of the separation column.

13.3 Preconcentration 13.3.1 Collection of Ions from Air

In the manufacture of semiconductor integrated circuits, it is of the utmost importance to manufacture in a ‘contaminant-free’ environment. The wafers and chips containing integrated circuits will fail if a particulate or ion contaminant shorts out any one of the circuit connections. As more and more integrated circuits are placed on the silica wafer the distance between the circuit connections become less. This means contamination is becoming more of an issue so that it has been proposed to monitor the air for airborne ionic contaminants. Lue and coworkers have described a method of monitoring acidic airborne contaminants in clean rooms [3]. In this method, acidic contaminants were adsorbed on silica gel tubes by passing a known volume of air through the tubes. The adsorbed impurities were extracted by a solution of carbonate and hydrogen car-

325

326

13 Sample Pretreatment

bonate and determined by ion chromatography. The recovery of HF was 100% and that of HCl was 91–100%. 13.3.2 Preconcentration of Ions in Water

Ion chromatography is frequently used to determine anions and cations at very low concentration levels, often in the low lg L–1 (ppb) range. In the electric power industry the water used in steam generators must be almost free of Na+, Cl– and other ions to avoid stress corrosion cracking of turbines, pipes, etc. The ionic content of ultrapure water used in the electronics industry must be kept to extremely low levels. Semiconductor chip manufacturers require clean rooms with utility impurities of no more than 1 ppb for 0.35lm devices [3]. As advances in technology have been made, IC detection limits have been lowered through careful control of the column and detector temperature and improvements in pump and detector design. One way to extend ion chromatography to even lower limits of detection is to increase the injection volume of the sample. A sample loop of 100 lL or even 400 lL could be used instead of a more typical 10–50 lL loop. By injecting 10–20 times more sample, the limits of detection for sample ions should be correspondingly lowered. However, the zone occupied by the sample solution in the IC column will also be larger. Since the sample ionic strength is much lower than that of the eluent there may be a prolonged dip in the chromatographic baseline when this sample zone passes through the detector. A separate concentrator column is the method most commonly used to extend the working range of ion chromatography to significantly lower levels. A concentrator column is a short column (typically 35–50 mm in length) placed in a valve just before the analytical column. Sometimes a guard column containing identical or similar material to the separation column is used as the concentrator column. The function of a concentrator column is to strip ions from a relatively large volume of an aqueous sample of very low ionic content. A valve arrangement enables the sample to pass through the concentrator column directly to waste. Then a valve is switched and the ions taken up by the concentrator column are swept into the analytical column by the eluent stream where they are separated chromatographically. The advantage of this system is the ability to perform routine analyses for ion concentration levels at low lg L–1 (ppb). Although several valve arrangements may be used, the simplest configuration is illustrated in Figure 13.1. In the load mode the sample flows through the concentrator column and out to waste (7, 8, 4, 3 sequence). Simultaneously, the eluent bypasses the concentrator column and flows into the analytical column (1, 2, 6, 5 sequence). In the inject mode the valve is switched so that the eluent flows through the concentrator column in the opposite direction to sample loading and into the analytical column (1, 4, 8, 5 sequence). Simultaneously, the sample stream is directed to waste (7, 6, 2, 3 sequence). Sample introduction may be by a small pump or with a manual syringe.

13.3 Preconcentration

Figure 13.1 Configuration for a Dionex Low Pressure 4-Way Valve and a Concentrator Column (Courtesy Dionex Corp).

The sample breakthrough volume from the concentrator column needs to be measured in order to know how large a sample may be used. The sample must be of low ionic strength (