Microchip Capillary Electrophoresis. Methods and Protocols

METHODS IN MOLECULAR BIOLOGY ™ 339 Microchip Capillary Electrophoresis Methods and Protocols Edited by Charles S. Hen...

Author: Charles S. Henry

27 downloads 821 Views 3MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

METHODS IN MOLECULAR BIOLOGY ™

339

Microchip Capillary Electrophoresis Methods and Protocols Edited by

Charles S. Henry

Microchip Capillary Electrophoresis

M E T H O D S I N M O L E C U L A R B I O L O G Y™

John M. Walker, SERIES EDITOR 352. Protein Engineering Protocols, edited by Kristian 352 Müller and Katja Arndt, 2006 351. C. elegans: Methods and Applications, edited by 351 Kevin Strange, 2006 350. Protein Folding Protocols, edited by Yawen Bai 350 and Ruth Nussinov 2006 349. YAC Protocols, Second Edition, edited by Alasdair 349 MacKenzie, 2006 348. Nuclear Transfer Protocols: Cell Reprogramming 348 and Transgenesis, edited by Paul J. Verma and Alan Trounson, 2006 347. Glycobiology Protocols, edited by Inka 347 Brockhausen-Schutzbach, 2006 346. Dictyostelium discoideum Protocols, edited by 346 Ludwig Eichinger and Francisco Rivero-Crespo, 2006 345. Diagnostic Bacteriology Protocols, Second Edition, 345 edited by Louise O'Connor, 2006 344. Agrobacterium Protocols, Second Edition: 344 Volume 2, edited by Kan Wang, 2006 343. 343 Agrobacterium Protocols, Second Edition: Volume 1, edited by Kan Wang, 2006 342. 342 MicroRNA Protocols, edited by Shao-Yao Ying, 2006 341. 341 Cell–Cell Interactions: Methods and Protocols, edited by Sean P. Colgan, 2006 340. 340 Protein Design: Methods and Applications, edited by Raphael Guerois and Manuela López de la Paz, 2006 339. 339 Microchip Capillary Electrophoresis: Methods and Protocols, edited by Charles S. Henry, 2006 338. 338 Gene Mapping, Discovery, and Expression: Methods and Protocols, edited by M. Bina, 2006 337. 337 Ion Channels: Methods and Protocols, edited by James D. Stockand and Mark S. Shapiro, 2006 336. 336 Clinical Applications of PCR, Second Edition, edited by Y. M. Dennis Lo, Rossa W. K. Chiu, and K. C. Allen Chan, 2006 335. 335 Fluorescent Energy Transfer Nucleic Acid Probes: Designs and Protocols, edited by Vladimir V. Didenko, 2006 334. 334 PRINS and In Situ PCR Protocols, Second Edition, edited by Franck Pellestor, 2006 333. 333 Transplantation Immunology: Methods and Protocols, edited by Philip Hornick and Marlene Rose, 2006 332. 332 Transmembrane Signaling Protocols, Second Edition, edited by Hydar Ali and Bodduluri Haribabu, 2006 331. 331 Human Embryonic Stem Cell Protocols, edited by Kursad Turksen, 2006 330. 330 Embryonic Stem Cell Protocols, Second Edition, Vol. II: Differentiation Models, edited by Kursad Turksen, 2006 329. 329 Embryonic Stem Cell Protocols, Second Edition, Vol. I: Isolation and Characterization, edited by Kursad Turksen, 2006

328. 328 New and Emerging Proteomic Techniques, edited by Dobrin Nedelkov and Randall W. Nelson, 2006 327 Epidermal Growth Factor: Methods and Protocols, 327. edited by Tarun B. Patel and Paul J. Bertics, 2006 326 In Situ Hybridization Protocols, Third Edition, 326. edited by Ian A. Darby and Tim D. Hewitson, 2006 325 Nuclear Reprogramming: Methods and Protocols, 325. edited by Steve Pells, 2006 324 Hormone Assays in Biological Fluids, edited by 324. Michael J. Wheeler and J. S. Morley Hutchinson, 2006 323 Arabidopsis Protocols, Second Edition, edited by 323. Julio Salinas and Jose J. Sanchez-Serrano, 2006 322 Xenopus Protocols: Cell Biology and Signal 322. Transduction, edited by X. Johné Liu, 2006 321. Microfluidic Techniques: Reviews and Protocols, 321 edited by Shelley D. Minteer, 2006 320. Cytochrome P450 Protocols, Second Edition, edited 320 by Ian R. Phillips and Elizabeth A. Shephard, 2006 319. 319 Cell Imaging Techniques: Methods and Protocols, edited by Douglas J. Taatjes and Brooke T. Mossman, 2006 318 Plant Cell Culture Protocols, Second Edition, edited 318. by Victor M. Loyola-Vargas and Felipe Vázquez-Flota, 2005 317 317. Differential Display Methods and Protocols, Second Edition, edited by Peng Liang, Jonathan Meade, and Arthur B. Pardee, 2005 316. Bioinformatics and Drug Discovery, edited by 316 Richard S. Larson, 2005 315. Mast Cells: Methods and Protocols, edited by Guha 315 Krishnaswamy and David S. Chi, 2005 314 314. DNA Repair Protocols: Mammalian Systems, Second Edition, edited by Daryl S. Henderson, 2006 313. Yeast Protocols, Second Edition, edited by Wei 313 Xiao, 2005 312. Calcium Signaling Protocols, Second Edition, 312 edited by David G. Lambert, 2005 311. Pharmacogenomics: Methods and Protocols, 311 edited by Federico Innocenti, 2005 310. 310 Chemical Genomics: Reviews and Protocols, edited by Edward D. Zanders, 2005 309. 309 RNA Silencing: Methods and Protocols, edited by Gordon Carmichael, 2005 308. 308 Therapeutic Proteins: Methods and Protocols, edited by C. Mark Smales and David C. James, 2005 307. 307 Phosphodiesterase Methods and Protocols, edited by Claire Lugnier, 2005 306. 306 Receptor Binding Techniques, Second Edition, edited by Anthony P. Davenport, 2005 305. 305 Protein–Ligand Interactions: Methods and Applications, edited by G. Ulrich Nienhaus, 2005

M E T H O D S I N M O L E C U L A R B I O L O G Y™

Microchip Capillary Electrophoresis Methods and Protocols

Edited by

Charles S. Henry Colorado State University, Fort Collins, CO

© 2006 Humana Press Inc. 999 Riverview Drive, Suite 208 Totowa, New Jersey 07512 www.humanapress.com All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise without written permission from the Publisher. Methods in Molecular BiologyTM is a trademark of The Humana Press Inc. All papers, comments, opinions, conclusions, or recommendations are those of the author(s), and do not necessarily reflect the views of the publisher. This publication is printed on acid-free paper. ∞ ANSI Z39.48-1984 (American Standards Institute) Permanence of Paper for Printed Library Materials. Cover illustration:Figure 1 from Chapter 4, "Fabrication of Polymer Microfluidic Systems by Hot Embossing and Laser Ablation," by Laurie E. Locascio, David J. Ross, Peter B. Howell, and Michael Gaitan. Production Editor: Erika J. Wasenda Cover design by Patricia F. Cleary For additional copies, pricing for bulk purchases, and/or information about other Humana titles, contact Humana at the above address or at any of the following numbers: Tel.: 973-256-1699; Fax: 973-256-8341; E-mail: [email protected]; or visit our website at www.humanapress.com. Photocopy Authorization Policy: Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Humana Press Inc., provided that the base fee of US $30.00 per copy is paid directly to the Copyright Clearance Center at 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license from the CCC, a separate system of payment has been arranged and is acceptable to Humana Press Inc. The fee code for users of the Transactional Reporting Service is: [1-58829-293-2/06 $30.00 ]. Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1 E-ISBN 1-59745-076-6 ISSN 1064-3745 Library of Congress Cataloging-in-Publication Data Microchip capillary electrophoresis : methods and protocols / edited by Charles S. Henry. p. ; cm. -- (Methods in molecular biology ; 339) Includes bibliographical references and index. ISBN 1-58829-293-2 (alk. paper) 1. Capillary electrophoresis. 2. Integrated circuits. I. Henry, Charles S. (Charles Sherman), 1972- . II. Series: Methods in molecular biology ; v. 339. [DNLM: 1. Electrophoresis, Microchip. W1 ME9196J v.339 2006 / QU 25 M626 2006] QP519.9.C36M53 2006 572.8028--dc22 2005028755

Preface Microchip capillary electrophoresis emerged as an important new analytical technique in the early 1990s out of the pioneering work of a group of dedicated scientists at Ciba Geigy. This new technique was a result of the marriage of the ability of conventional capillary electrophoresis to analyze ultrasmall volumes (nL) and microfabrication techniques perfected in the semiconductor industry to produce very small structures in silicon. The resulting technology holds significant promise to improve our ability for analysis of biological systems because it can handle objects as small or smaller than a single cell, integrate such sample processing steps as filtration and the polymerase chain reaction, and generate answers in seconds or minutes compared with the hours used for many traditional techniques. The goal of this volume of Humana’s series Methods in Molecular BiologyTM is to provide the reader an overview of the methods currently in place for microchip capillary electrophoresis, as well as to provide useful practical information on how to get started in the field. The text of Microchip Capillary Electrophoresis is divided into four sections. Part I deals with fabrication methods for the production of microchips because this is fundamental to the ability to use the technology. The chapters are divided based on the substrate material and include glass (Chapter 2), poly(dimethylsiloxane) (Chapter 3), and other polymers including polymethylmethacrylate (Chapter 4). The information provided in these chapters should be suitable for even the novice to produce simple microchips for standard separations. Part II discusses methods to control the surface chemistry and measure the resulting alterations in microfluidic devices. Surface chemistry plays an important role in systems at this scale and must be carefully considered for optimal operation. Chapter 5 provides a general overview of several methods of both adsorbed and covalent surface modification. Chapter 6 provides more detail on a simple, yet effective, adsorbed coating system. Part III describes different detection modes for microchip capillary electrophoresis with detail provided for mass spectrometry (Chapter 7), electrochemistry (Chapter 8), and finally conductivity (Chapter 9). The last section of this book outlines applications of microchip capillary electrophoresis for biological analysis. Chapters 10–12 deal with the analysis of DNA, proteins, and peptides, respectively. Chapter 13 discusses techniques for measuring the impact of surface modification on flow in microfluidic channels. Chapter 14 discusses single cell analysis. The last chapter (Chapter 15) is a forward-looking review of the integration of the

v

vi

Preface

polymerase chain reaction into a capillary electrophoresis microchip, part of the next generation of a technology that is focused on the eventual integration of all laboratory functions in a single device. Microchip Capillary Electrophoresis is intended to be both a practical guide for those interested in using this exciting new technology in their own research, as well as an important source of fundamental information detailing how the technique works at the molecular scale. As such, our book is not meant to be an all inclusive, exhaustive text on every aspect of microchip capillary electrophoresis. Readers are encouraged to use the book as the reference guide it was intended to be, and then move on to seek current literature in this rapidly evolving field for applications more specific to their work. Finally, the editor would like to thank Professor John Walker for his significant help in completing this work. Suggestions for content from Drs. Scott Martin and Steve Soper are also acknowledged. Finally, the continued support of the publisher throughout this lengthy process was also greatly appreciated.

Charles S. Henry

Contents Preface .............................................................................................................. v Contributors .....................................................................................................ix 1 Microchip Capillary Electrophoresis: An Introduction Charles S. Henry ................................................................................... 1

PART I MICROCHIP FABRICATION METHODS 2 Fabrication of a Glass Capillary Electrophoresis Microchip With Integrated Electrodes Mark M. Crain, Robert S. Keynton, Kevin M. Walsh, Thomas J. Roussel, Jr., Richard P. Baldwin, John F. Naber, and Douglas J. Jackson ................................................................... 13 3 Micro-Molding for Poly(dimethylsiloxane) Microchips Carlos D. García and Charles S. Henry .............................................. 27 4 Fabrication of Polymer Microfluidic Systems by Hot Embossing and Laser Ablation Laurie E. Locascio, David J. Ross, Peter B. Howell, and Michael Gaitan ........................................................................ 37

PART II SURFACE MODIFICATION METHODS 5 Surface Modification Methods for Enhanced Device Efficacy and Function Barbara J. Jones and Mark A. Hayes ................................................... 49 6 Polyelectrolyte Coatings for Microchip Capillary Electrophoresis Yan Liu and Charles S. Henry ............................................................. 57

PART III DETECTION METHODS FOR MICROCHIP CAPILLARY ELECTROPHORESIS 7 Interfacing Microchip Capillary Electrophoresis With Electrospray Ionization Mass Spectrometry Trust Razunguzwa and Aaron T. Timperman ..................................... 67 8 Interfacing Amperometric Detection With Microchip Capillary Electrophoresis R. Scott Martin .................................................................................... 85 9 Conductivity Detection on Microchips Roland Hergenröder and Benedikt Graß .......................................... 113

vii

viii

PART IV APPLICATIONS

Contents OF

MICROCHIP CAPILLARY ELECTROPHORESIS

10 DNA Separations Andrea W. Chow .............................................................................. 11 Protein Separations Andrea W. Chow .............................................................................. 12 Microchip Capillary Electrophoresis: Application to Peptide Analysis Barbara A. Fogarty, Nathan A. Lacher, and Susan M. Lunte ............ 13 Measuring Electroosmotic Flow in Microchips and Capillaries S. Douglass Gilman and Peter J. Chapman ....................................... 14 Single Cell Analysis on Microfluidic Devices Christopher T. Culbertson ................................................................ 15 Rapid DNA Amplification in Glass Microdevices Christopher J. Easley, Lindsay A. Legendre, James P. Landers, and Jerome P. Ferrance ................................................................ Index ............................................................................................................

129 145

159 187 203

217 233

Contributors RICHARD P. BALDWIN • Department of Chemistry, University of Louisville, Louisville, KY PETER J. CHAPMAN • Department of Chemistry, University of Tennessee, Knoxville, TN ANDREA W. CHOW • Microfluidics Research and Development, Caliper Life Sciences, Mountain View, CA MARK M. CRAIN • Lutz Micro/Nanotechnology Cleanroom, University of Louisville, Louisville, KY CHRISTOPHER T. CULBERTSON • Department of Chemistry, Kansas State University, Manhattan, KS CHRISTOPHER J. EASLEY • Department of Chemistry, University of Virginia, Charlottesville, VA JEROME P. FERRANCE • Department of Chemistry, University of Virginia, Charlottesville, VA BARBARA A. FOGARTY • Life Sciences Interface, Tyndall Institute, Cork, Ireland MICHAEL GAITAN • Analytical Chemistry Division, National Institute of Standards and Technology, Gaithersburg, MD CARLOS D. GARCÍA • Department of Chemistry, University of Texas– San Antonio, San Antonio, TX S. DOUGLASS GILMAN • Department of Chemistry, Louisiana State University, Baton Rouge, LA BENEDIKT GRAß • ISAS-Institute for Analytical Sciences, Dortmund, Germany MARK A. HAYES • Department of Chemistry, Arizona State University, Tempe, AZ CHARLES S. HENRY • Department of Chemistry, Colorado State University, Fort Collins, CO ROLAND HERGENRÖDER • ISAS-Institute for Analytical Sciences, Dortmund, Germany PETER B. HOWELL • Analytical Chemistry Division, National Institute of Standards and Technology, Gaithersburg, MD DOUGLAS J. JACKSON • Department of Electrical and Computer Engineering, University of Louisville, Louisville, KY BARBARA J. JONES • Analytical Chemistry Division, National Institute of Standards and Technology, Gaithersburg, MD

ix

x

Contributors

ROBERT S. KEYNTON • Department of Bioengineering, University of Louisville, Louisville, KY NATHAN A. LACHER • Analytical Research and Development, Pfizer Global Biologics, St. Louis, MO JAMES P. LANDERS • Department of Chemistry, University of Virginia, Charlottesville, VA LINDSAY A. LEGENDRE • Department of Chemistry, University of Virginia, Charlottesville, VA YAN LIU • Department of Chemistry, Colorado State University, Fort Collins, CO LAURIE E. LOCASCIO • Analytical Chemistry Division, National Institute of Standards and Technology, Gaithersburg, MD SUSAN M. LUNTE • Department of Pharmaceutical Chemistry, University of Kansas, Lawrence, KS R. SCOTT MARTIN • Department of Chemistry, Saint Louis University, St. Louis, MO JOHN F. NABER • Department of Electrical and Computer Engineering, University of Louisville, Louisville, KY TRUST RAZUNGUZWA • Department of Chemistry, West Virginia University, Morgantown, WV DAVID J. ROSS • Analytical Chemistry Division, National Institute of Standards and Technology, Gaithersburg, MD THOMAS J. ROUSSEL, JR. • Department of Bioengineering, University of Louisville, Louisville, KY AARON T. TIMPERMAN • Department of Chemistry, West Virginia University, Morgantown, WV KEVIN M. WALSH • Department of Electrical and Computer Engineering, University of Louisville, Louisville, KY

1 Microchip Capillary Electrophoresis An Introduction Charles S. Henry Summary Microchip capillary electrophoresis emerged in the early 1990s as an intresting and novel approach to the high-speed separation of biological compounds, including DNA and proteins. Since the early development in this area, growth in the research field has exploded and now includes chemists and engineers focused on developing new and better microchips, as well as biologists and biochemists who have begun to apply this exciting and still relatively new methodology to real-world problems. This chapter seeks to outline the historical development of microchip, the key elements of microchip capillary electrophoresis, and finally some of the important applications beign develop that utilize microchip capillary electrophoresis. Key Words: Microchip capillary electrophoresis; capillary electrophoresis; microfabrication; bioanalytical chemistry.

1. Introduction Microchip capillary electrophoresis (CE) has appeared over the decades as a result of the marriage of chemical analysis and microfabrication techniques from the integrated circuit world. The concept of microchip separations is not new, however, with the first report of a microfabricated gas chromatography column appearing in 1979 (1) and the first example of a microfabricated liquid chromatography system appearing in 1990 (2). Little excitement was generated over these early developments, however, because of the complexity of the operational systems and generally poor performance of the devices. Active development of microchip separation technology did not begin until the early 1990s with the seminal work of Manz, Harrison, Verpoorte, and Widmer in microchip CE (3). CE proved to be an excellent match for microchip technologies because it easily manipulates volumes at the nanoliter scale, requires no moving parts, and provides fast, high-resolution separations. From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

1

2

Henry

Since the early reports of microchip CE, the field has grown exponentially in both the number of investigators and the areas of application and fundamental development. One general theme now driving the field is the integration of function, with the ultimate goal being the development of miniaturized total analysis systems (also referred to as Lab-on-a-Chip) that integrate all functions of the modern analytical laboratory in a single device. Unlike traditional CE instrumentation, which consists of essentially a single capillary (or an array of capillaries in parallel), many different capillaries and fluidic channels can be patterned on a microfluidic device, providing the potential for high-throughput, massively parallel analysis. Microchip CE has also garnered significant attention because of the potential applications and overall device performance. Although the primary early focus in the field was on DNA analysis (4), microchip CE has been rapidly adapted to many biological, environmental, and industrial applications (5–12). Microchip CE also has the added benefits of low cost, small size, and fast analysis times, which are general goals of many chemical analysis methods. Functionality, such as polymerase chain reaction (PCR), enzymatic digestion, and solid phase extraction, can be incorporated into the microchip to provide pre- and postsample processing. This capability is important because it allows a raw sample to be added to the microchip and a final quantitative answer provided. 2. Theory and Mechanisms of Action in Microchip CE The theory and mechanisms of action for microchip CE are based on fundamental knowledge gained in the development of conventional CE. The first report of conventional CE is generally attributed to Lukacs and Jorgenson in the early 1980s (13). In this work, mixtures of proteins and peptides were separated in glass capillaries. The two major outcomes of this technology were (1) the demonstration of open tubular electrophoresis for high-speed separations of biomolecules and (2) the presence of electroosmotic flow (EOF). EOF is a bulk solution flow phenomenon that occurs in capillaries filled with mild ionic solutions (typically <100 mM) when a voltage is established across the capillary. EOF will be discussed in more detail later in this chapter and in the chapter dealing specifically with its measurement (Chapter 13). Since the initial report of CE, a number of new submethods have appeared that provide enhancements in separation selectivity (14). These submethods include micellar electrokinetic chromatography (MEKC), capillary electrochromatography (CEC), capillary gel electrophoresis (CGE), and capillary isoelectric focusing (CIEF). Some of these methods are truly new methods (MEKC and CEC), whereas others are adaptations of traditional slab gel electrophoresis (CGE and CIEF). For CE separations, the velocity of a given analyte species is determined by the mobility of the analyte plus the EOF as shown in Eq. 1.

An Introduction to Microchip CE

3

v = ( µ e + µ eof ) E

(1)

Where v is the velocity of the analyte, µe is the electrophoretic mobility of the analyte, µeof is the EOF, and E is the field strength. Therefore, the electroosmotic flow is a constant for a specific set of run conditions for all molecules being separated. The electrophoretic mobility of the analyte molecules, therefore, is the most important factor for separation. The two most important factors influencing electrophoretic separations in zone electrophoresis are the analyte mass and net charge, as shown in Eq. 2. µ=

qE 6 πηr

(2)

Where q is the charge on the molecule, E is the applied field strength (V/cm), η is the viscosity of the mobile phase, and r is the radius of the molecule, which is related to its mass. Because the mass of a molecule is constant, the easiest way to influence electrophoretic separations is through changes in the mobile phase pH, which affect the charge on the molecules being separated. Equation 2 shows the importance of molecular charge on the analyte to achieve separations in capillary zone electrophoresis. However, there are times when either the analyte is neutral or several compounds have very similar massto-charge ratios and cannot be resolved by pH changes alone. In these cases, another CE mode (either MEKC or CEC), which has been developed for separation of neutrals, can be applied. In MEKC, surfactants are added in sufficient amount to form micelles with polar, charged exteriors and hydrophobic interiors (15). The micelles have their own unique mobility in the mobile phase. As analyte molecules move through the mobile phase, they are influenced by several factors. When they are in free solution, the analyte’s velocity (both magnitude and direction) is equal to the sum of the EOF and their own electrophoretic mobility. However, owing to partitioning between the aqueous and micelle phases, the analytes spend part of their time inside the micelle. While in the micelle, their velocity is equal to the velocity of the micelle. The time the analyte molecules spend inside the micelles is dependent on their ability to partition between free solution and the micelle. Thus, the migration time increases with the hydrophobicity of the compound for micelles made from anionic surfactants, such as sodium dodecyl sulfate. CEC is a second alternative to improve separation resolution. CEC is a hybrid separation technique combining electrophoretic pumping from CE with the stationary phases of liquid chromatography to provide high-resolution separations. CEC is not only useful in the separation of neutral compounds, but also situations where species have similar mass-to-charge ratios. Separation in CEC involves the complex interaction between the analyte in the mobile phase (net electrophoretic mobility) plus

4

Henry

the time spent interacting with the stationary phase packed in the capillary. Traditional CEC stationary phases (3–5 µm silica particles modified with the selective phase of interest) have been used in microchip CEC. In addition, newer monolithic column technology, where the stationary phase is made from a single continuous polymer monolith, has also been used in microchip CEC (16–18). Monolithic columns have the advantage of being able to be patterned directly with ultraviolet radiation, confining them in the columns without the use of packing frits. EOF also plays an important role in CE separations. In traditional CE, EOF serves to speed up (or delay) the separation of analytes and allow simultaneous separation of both anions and cations in a single run. In microchip CE and microfluidics in general, EOF plays an even more important and useful role. In addition to decreasing separation time, EOF can control the direction and magnitude of flow in connecting channels for injection and other operations such as sample preparation. The two key factors that control EOF are the applied field strength (E) and the thickness of the double layer at the capillary–solution interface. The double layer forms as a result of the ions from solution trying to offset the charge on the surface of the capillary. In fused silica capillaries and glass microchips, the surface charge results from the deprotonation of silanol groups and is pH dependent. In bare polymer microchips, a variety of functional groups can give rise to a charge; however, all reported to date have been anionic. Of the two factors that control EOF, E is the easiest to control. The double layer thickness is more difficult to control but in general is reduced as the ionic strength increases or as the pH decreases. In principle, the thinner the double layer, the slower the flow rate. 3. Microchip Construction and Operation 3.1. Microchip Construction The first microchip CE systems were constructed using glass as a substrate material (3,19). Glass, being a silicon compound, has well-known and well-developed microfabrication methods that were adapted directly from the electrical engineering field. In addition, glass (or quartz) has excellent optical properties with optical transparency throughout all useful wavelength ranges. Finally, glass has the same surface chemistry as the fused silica capillaries used for conventional CE and many of the same developments made in that field can be applied directly to glass microchips, particularly for surface modification. Despite the advantages of glass, there are also disadvantages. First, the fabrication process yields one device, resulting in a relatively high manufacturing cost per microchip. Second, glass is fragile and easily broken if not handled with extreme care. Third, high-quality glass is expensive. As a result of these issues, a significant amount of attention has

An Introduction to Microchip CE

5

focused on the fabrication of microchips using polymer substrates. Polymer substrates are less expensive and easier to fabricate than glass. Polymers are fabricated by either a molding procedure or through the use of direct write protocols using ultraviolet lasers, electron beams, or X-rays. Either fabrication method provides high-throughput and uses lower cost materials. The major limitation of polymers for microchip CE is the poorly understood and poorly controlled surface chemistry of these materials. Understanding and controlling the material properties of polymer devices is an important goal that is yet to be adequately addressed. 3.2. Microchip Operation 3.2.1. Injection Standard analysis protocols for microchip CE involve injection, separation, and detection. Injection in conventional CE is accomplished by placing one end of the capillary in the sample solution and applying either pressure or voltage for a short period of time to introduce a finite volume of sample into the capillary. Microfluidic channels cannot be physically manipulated in the same way and, therefore, new injection protocols had to be developed for microchip CE. There are two common forms of injection in microchip CE, gated injection and cross injection. In gated injection, a flow boundary is established between two solutions, the mobile phase and the sample solution, at the intersection of several channels. No mixing of solutions occurs at this intersection because the interaction time is small, minimizing diffusion and the flow regime is laminar. At a specific point in time, the voltage is turned off to the mobile phase channel and the sample solution is injected into the separation channel. After the specified time length, the voltage to the mobile phase reservoir is re-established and the injection plug is defined. Gated injection is simple but results in the biasing of sample injection toward cations over anions owing to their higher mobility under normal flow conditions. Cross injection is more complicated but does not give a biased injection. In cross injection, sample solution is flowed across the separation channel using two side channels. After a fixed time period, the voltage is switched to direct flow down the separation channel. The volume of the injected sample plug is defined by the volume at the intersection of the channels. Small “push back” voltages are applied to the side channels during separation to prevent leakage into the separation channel during the analysis. An added advantage to crossinjection is the ability to physically define the volume of the injection plug through the use of double-T injectors. In double-T injectors, the two side channels are not directly opposite each other, but are instead offset by a fixed distance. The result is that during injection, the larger volume intersection is filled with sample, giving an increase in injected volume.

6

Henry

3.2.2. Separation Microchip CE separations occur in much the same way as conventional CE separations. Capillary zone electrophoresis, MEKC, CEC, CGE, and CIEF have all been used to effect the desired separation. In some cases, the operations are made easier using microchip approaches. For example, it is easier to fabricate a monolithic column with open injectors for microchip CEC than in conventional CEC. In other cases, operation is made more difficult. For example, it is generally harder to pack microchip columns with gel than conventional columns. 3.2.3. Detection Several modes of detection have been demonstrated with microchip CE with the most common being laser-induced fluorescence (LIF). LIF was first adapted to microchip CE because it is relatively easy to couple the two techniques. In addition, LIF is very sensitive for detection of the small mass quantities present in microchip separations. LIF has been used for many applications with microchip CE including DNA analysis, protein analysis, and measurement of small molecules (20). The biggest advantage of LIF is the ability to detect very small quantities of analyte, with single molecule detection being reported. LIF also has several disadvantages that have led to research in other detection methods. LIF normally requires relatively large and expensive lasers and optics systems. Although there is promise for the integration of microchip lasers and detectors, this is still some way off. Furthermore, most analytes are not fluorescent at usable wavelengths, requiring derivatization. Two additional techniques that have received attention are mass spectrometry (MS) and electrochemistry (EC). MS is the ultimate identification tool in chemical analysis as it can provide molecular weight information and fragmentation patterns. MS has been coupled to microchip CE for a variety of applications but most reports have focused on protein sequencing (proteomics). The disadvantages of MS include high-instrument cost, a high level of technical expertise needed for operation, and large instrument size. EC is another method of detection that has been successfully coupled with microchip CE. EC relies on selective electron transfer reactions at an electrode surface to achieve detection. Selectivity can be achieved by control of the detection potential. Many more molecules are naturally electrochemically active than are fluorescent, especially when pulsed potential methods are used. EC detection systems are also inexpensive and easy to miniaturize, making this method an ideal match for creating portable analysis systems. One disadvantage of microchip CE-EC, however, is sensitivity. On average, reported detection limits for microchip CE-EC are around 1 µM. Reducing detection limits to approx 1 nM or less has become a major focus in this research area. Another form of EC detection that has significant potential is conductivity. Conductivity detection is a universal mode of detection, which

An Introduction to Microchip CE

7

makes it widely applicable to a variety of analysis situations. One potential problem with conductivity, however, is a lack of selectivity. 4. Microchip CE Applications 4.1. DNA Separations The separation of DNA for both sequencing and fragment analysis purposes has taken on an important role in biological research, clinical measurements, and forensics. Arguably, one of the most successful application areas for microchip CE has been DNA separation. DNA separations are performed in a manner very similar to traditional capillary gel electrophoresis experiments. The same gel matrixes are often used and detection is achieved using LIF. The short columns of microchip CE decrease the separation time relative to conventional CE and should also decrease the resolution. However, the reduced injection volumes used in microchip CE (pL) help maintain resolution. In addition to DNA separation, the PCR has been demonstrated onchip both independently and directly coupled to microchip CE. This high level of integration has led several independent investigators to suggest the possibility of a personal genetic analyzer and/or the development of crime-scene analyzers. 4.2. Protein Analysis A second major area of application is protein analysis. There is a great need to develop new analytical methodologies for the measurement of proteins for clinical research, pharmaceutical development, and to better understand fundamental biological processes. With the sequencing of the human genome, it is anticipated that thousands of new proteins will be discovered and/or synthesized. Traditional methods are often too slow to be able to handle the necessary high-throughput analysis that will be required by this number of analytes. There are a variety of protein analysis methods that can be developed on-chip. Direct separation of fluorescently labeled proteins has been demonstrated numerous times for both characterization of new microchip CE devices and for measuring the purity of protein samples. Capillary electrophoretic immunoassays (CEIA) have also been performed at the microchip scale. Microchip CEIA allows for the direct measurement of protein analytes from complex samples such as blood and urine, making it a very useful tool clinically. A second useful analysis method being developed focuses on measuring enzyme kinetics and activities. Enzyme assays have several important applications, including determining functional pathways in cellular systems, as well as amplifying signals to improve the sensitivity of detection for routine analytical measurements.

8

Henry

4.3. Small Molecule Analysis The last major area of development in microchip CE has been the analysis of small molecules. Small molecule analysis was a problem early on because most small molecules do not fluoresce. Amino acid separations were carried out after derivatization but many other molecules were not readily measured. The use of indirect LIF, MS, and EC has more recently led to an increase in the number of articles appearing on the topic of small molecule analysis. Specific applications range from detection of explosive components to the measurement of clinical analytes in urine and serum. 5. Conclusion The future is promising for microchip CE and its application in biological sciences. The number of investigators and applications continues to grow exponentially. Despite the growth, a number of important issues remain unresolved. These issues include better control of the surface chemistry of microchip CE devices, especially for polymer microchips, better interfacing with the real world, and finally, a higher degree of integration of function. With improvements in these areas, the use of microchip CE will certainly increase and the technique will become a significant contributor to modern biochemical analysis. References 1. Terry, S. C., Jerman, J. H., and Angell, J. B. (1979) A gas chromatograph air analyzer fabricated on a silicon wafer. IEEE Trans. Electron. Devices ED-26, 1880. 2. Manz, A., Graber, N., and Widmer, H. M. (1990) Miniaturized total chemical analysis systems: a novel concept for chemical sensing. Sensors and Actuators, B: Chemical B1, 244–248. 3. Manz, A., Harrison, D. J., Verpoorte, E. M. J., et al. (1992) Planar chips technology for miniaturization and integration of separation techniques into monitoring systems. Capillary electrophoresis on a chip. J. Chromatogr. 593, 253–258. 4. Woolley, A. T. and Mathies, R. A. (1994) Ultra-high-speed DNA fragment separations using microfabricated capillary array electrophoresis chips. Proc. Natl. Acad. Sci. USA 91, 11,348–11,352. 5. Verpoorte, E. (2002) Microfluidic chips for clinical and forensic analysis. Electrophoresis 23, 677–712. 6. Wang, J. (2002) On-chip enzymatic assays. Electrophoresis 23, 713–718. 7. Liu, Y., Garcia, C. D., and Henry, C. S. (2003) Recent progress in the development of mu TAS for clinical analysis. Analyst 128, 1002–1008. 8. Tokeshi, M., Kikutani, Y., Hibara, A., Sato, K., Hisamoto, H., and Kitamori, T. (2003) Chemical processing on microchips for analysis, synthesis, and bioassay. Electrophoresis 24, 3583–3594. 9. Chang, H. -T., Huang, Y. -F., Chiou, S. -H., Chiu, T. -C., and Hsieh, M. -M. (2004) Advanced capillary and microchip electrophoretic techniques for proteomics. Current Proteomics 1, 325–347.

An Introduction to Microchip CE

9

10. Erickson, D. and Li, D. Q. (2004) Integrated microfluidic devices. Analytica Chimica Acta 507, 11–26. 11. Guttman, A., Varoglu, M., and Khandurina, J. (2004) Multidimensional separations in the pharmaceutical arena. Drug Discov. Today 9, 136–144. 12. Vilkner, T., Janasek, D., and Manz, A. (2004) Micro total analysis systems. Recent developments. Anal. Chem. 76, 3373–3385. 13. Jorgenson, J. W. and Lukacs, K. D. (1981) Zone electrophoresis in open-tubular glass-capillaries. Anal. Chem. 53, 1298–1302. 14. Landers, J. P. (1997) Handbook of Capillary Electrophoresis, CRC Press, Boston, MA. 15. Moore, A. W., Jr., Jacobson, S. C., and Ramsey, J. M. (1995) Microchip separations of neutral species via micellar electrokinetic capillary chromatography. Anal. Chem. 67, 4184–4189. 16. Breadmore, M. C., Shrinivasan, S., Wolfe, K. A., et al. (2002) Towards a microchip-based chromatographic platform. Part 1: evaluation of sol-gel phases for capillary electrochromatography. Electrophoresis 23, 3487–3495. 17. Yu, C., Svec, F., and Frechet, J. M. (2000) Towards stationary phases for chromatography on a microchip: molded porous polymer monoliths prepared in capillaries by photoinitiated in situ polymerization as separation media for electrochromatography. Electrophoresis 21, 120–127. 18. Yu, C., Davey, M. H., Svec, F., and Frechet, J. M. (2001) Monolithic porous polymer for on-chip solid-phase extraction and preconcentration prepared by photoinitiated in situ polymerization within a microfluidic device. Anal. Chem. 73, 5088–5096. 19. Jacobson, S. C., Hergenroder, R., Koutny, L. B., and Ramsey, J. M. (1994) Highspeed separations on a microchip. Anal. Chem. 66, 1114–1118. 20. Johnson, M. E. and Landers, J. P. (2004) Fundamentals and practice for ultrasensitive laser-induced fluorescence detection in microanalytical systems. Electrophoresis 25, 3513–3527.

I MICROCHIP FABRICATION METHODS

2 Fabrication of a Glass Capillary Electrophoresis Microchip With Integrated Electrodes Mark M. Crain, Robert S. Keynton, Kevin M. Walsh, Thomas J. Roussel, Jr., Richard P. Baldwin, John F. Naber, and Douglas J. Jackson Summary In this chapter, a detailed outline delineating the processing steps for microfabricating capillary electrophoresis (CE) with integrated electrochemical detection (ECD) platforms for performing analyte separation and detection is presented to enable persons familiar with microfabrication to enter a cleanroom and fabricate a fully functional Lab-on-a-Chip (LOC) microdevice. The processing steps outlined are appropriate for the production of LOC prototypes using easily obtained glass substrates and common microfabrication techniques. Microfabrication provides a major advantage over existing macro-scale systems by enabling precise control over electrode placement, and integration of all required CE and ECD electrodes directly onto a single substrate with a small footprint. In the processing sequences presented, top and bottom glass substrates are photolithographically patterned and etched using wet chemical processing techniques. The bottom substrate contains seven electrodes required for CE/ECD operation, whereas the top substrate contains the microchannel network. The flush planar electrodes are created using sputter deposition and lift-off processing techniques. Finally, the two glass substrates are thermally bonded to create the final LOC device. Key Words: Capillary electrophoresis (CE); electrochemical detection (ECD); Lab-on-aChip (LOC); microfluidics; microelectrodes; microfabrication.

1. Introduction This chapter presents a detailed description of the fabrication steps involved in constructing a microfluidic platform for analyte separation via capillary electrophoresis (CE) including electrochemical detection (ECD) electrodes integrated “on-chip.” The objective of this chapter is to provide a detailed processing roadmap that will provide persons familiar with microfabrication technology the necessary information to enter a cleanroom and fabricate a functioning Lab-on-a-Chip (LOC) microdevice. The process steps outlined are From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

13

14

Crain et al.

appropriate for the production of LOC prototypes using traditional glass substrates and general microfabrication manufacturing techniques. Although viable alternative microfabrication techniques exist for the mass production of these devices (e.g., micromechanical machining, micro-molding, micro-embossing), traditional fabrication techniques will be described because these processes leverage 40 yr of success by the integrated circuit industry and, thus, are well characterized and readily available. 2. Device Design Prior to fabrication, the microfluidic device was designed and modeled using a finite element (FEA) software package (CoventorWare®, Coventor, Raleigh, NC) to determine the appropriate geometry for the channels and the detection electrodes. Previous microcapillary investigations performed by Jacobson and Ramsey (1) demonstrated that by “focusing” the analyte stream at the intersection of the sample loading and separation microchannel, unwanted diffusion of the analyte from the injection stream into the separation microchannel arm could be eliminated. Unfortunately, accomplishing this with an unbalanced (i.e., unequal arm lengths) microchannel system geometry requires multiple high voltage power supplies, typically one for each terminating capillary arm. To solve this problem, a “balanced-cross” microcapillary pattern was developed and modeled, consisting of four equal length capillary arms (10,000 µm). This greatly simplified the electrical requirements for driving the electrokinetic flow by allowing operation of the system with a single power supply for both injection and separation modes. The details of the supporting electronics is beyond the scope of this chapter and, therefore, will not be addressed here, but a description of the design and development of the power supply and electrochemical detection circuitry has been previously reported (2). To reduce the “footprint” required by this geometry, each of the two arms that establish the sample loading microchannel were designed with 90° bends (Fig. 1). It has been previously demonstrated that bends or turns in a microfluidic channel will distort an analyte plug such that detection can be negatively affected (3). In this case, the channels that incorporate 90° bends are utilized only during sample loading (also referred to as sample injection); and, therefore, the plug integrity at the intersection is not compromised by either of these bends. 3. Microfabrication 3.1. Photomask Development After the microchip design process has been completed, the desired channel and electrode configurations are created in a pattern generation software Fig. 2. (Opposite page) L-edit mask pattern for the (A) microchannels (channel linewidth = 50 µm) and (B) electrodes (working electrode linewidth = 40 µm). The final device design following superposition of the negative of the two masks is shown in (C). In Fig. 2A, the circles represent the location of the reservoirs and the square represents

Fabrication of a Glass CE Microchip With Integrated ECD

15

Fig. 1. Schematic of the capillary electrophoresis system developed using finite element modeling software. Highlighted are the four reservoirs (sample, waste, buffer, and detection) and seven patterned electrodes required for operation (four high-voltage electrodes for CE and three amperometric electrodes for ECD). The three amperometric electrodes (working, reference, and counter) are positioned in the detection reservoir such that the working electrode is closest to the microchannel exit (4).

the location of the larger reservoir required to accommodate the EC detection and CE driving electrodes. In Fig. 2B, R = the reference electrode; C = the counter electrode; W = the working/detection electrode; and D1, D2, D3, and D4 are the driving electrodes for the detection, sample, buffer, and waste reservoirs, respectively.

16

Crain et al.

(L-EDIT; Tanner Research, Inc., Pasadena, CA) and saved as a CIF (CalTech Intermediate Form), GDS (graphic design system), DXF (drawing exchange format), or PS (postscript) file. Examples of the patterns developed for the microchannels and electrodes to be discussed are shown in Fig. 2. The L-EDIT file is used to generate the master set of photomasks required for the micromanufacturing of the prototype devices. Depending upon the resolution desired for the final device, the photomask set can be fabricated either on glass via an optical or laser pattern generator or on transparency sheets (such as Mylar®) using a high resolution image setter, such as those found at a local print shop (see Note 1). The major advantage of glass photomasks fabricated using an optical or laser pattern generator is the quality in the line resolution, where linewidths of 1–2 µm can generally be achieved. However, photomasks fabricated with optical and laser pattern generators are expensive because these systems typically cost between $150,000 and $300,000, and the sensitized chrome-covered photomask glass blanks are also relatively expensive. Several organizations provide mask generation services, which typically charge $300–500 per mask, as compared with the lower resolution Mylar photomasks, which typically cost between $15 and $50 per sheet. The two most significant advantages of the Mylar photomask approach are low cost and short turnaround time, both of which are important issues for research and development and rapid prototyping applications. However, the advantages of using the transparency solution are often outweighed by the reduction in the quality of the line resolution, which at best is approx 10 µm. 3.2. Photolithography Soda-lime glass was chosen as the primary substrate material because it is commercially available, can be processed using traditional microfabrication techniques, and naturally possesses the surface characteristics needed for the requisite wall–buffer charge interface (ζ-potential), thereby eliminating the necessity for any chemical modification of the capillary wall surfaces. This

Fig. 3. Photomask blank prior to patterning.

Fabrication of a Glass CE Microchip With Integrated ECD

17

Fig. 4. Schematic of bottom substrate photoresist layer upon exposure to ultraviolet light.

native potential is a fundamental requirement for establishing the electroosmotic bulk flow of the buffer fluid in response to an applied electric field. One of the easiest and cheapest methods for etching a detailed pattern into glass is to simply use an unexposed photomask blank as the starting substrate material. The photomasks used in this process are 60 mil (1524 µm) thick ultraflat, soda-lime glass precoated with a low reflective chrome and positive resist (supplied by Nanofilm, Inc.; Fig. 3). These photomask blanks become the actual glass substrates, and the ultra-flat surface of the photomask allows the resolution and integrity of the fabricated features to be maintained. The combination of these two properties is critical in the final bonding process. 3.3. Bottom Substrate Processing The bottom substrate (5 × 3.5 cm) contains the required seven electrodes for CE/ECD operation. These electrodes are formed by a photolithographic liftoff process, bulk micromachining, and sputtering as follows (see Note 2): 1. The electrode pattern is first transferred onto the precoated positive resist layer using the appropriate darkfield photomask (Fig. 2B) and contact lithography (Fig. 4). The photomask blank is positioned with the patterned mask in direct contact with the photoresist. The photoresist is exposed to G-line ultraviolet (UV) light from a Mercury arc lamp with a dose of 150 mJ/cm2 to allow the resist to be patterned as described in step 2. 2. The exposed photoresist is then removed with MF-319 developer (Shipley Co., LLC, Marlborough, MA), which is a dilute tetramethylammonium hydroxide (TMAH) solution, to reveal the underlying chrome layer (Fig. 5). It is essential

18

Crain et al.

Fig. 5. Schematic of the patterned photoresist layer on the bottom substrate.

Fig. 6. Schematic of bottom substrate following removal of chrome layer.

Fig. 7. Schematic of etched recessions in glass substrate before electrode deposition. that the unexposed photoresist remains on this piece of glass for the liftoff patterning of the Ti/Pt electrode layer described next. 3. The exposed chrome layer is then patterned by a short etch in CEP-200 Micro Chrome etchant (Microchem, Inc., Newton, MA). The final product is a soda-lime glass substrate with a patterned layer of photoresist and chrome, revealing the glass substrate in the desired electrodes regions only (Fig. 6). 4. After lithography and chrome etching, the exposed glass is etched for 30 s in a 6:1 buffered oxide etch (BOE; J.T. Baker, Phillipsburg, NJ) to form approx 0.3-µm recessions (Fig. 7). This short etch produces shallow recesses that allow the subsequent sputter-deposited metal layer to lie flush or just below the original glass surface, which is critical for keeping the final surface of the bottom substrate planar; otherwise, electrodes protruding above the surface would hinder the

Fabrication of a Glass CE Microchip With Integrated ECD

19

Fig. 8. Schematic of Ti deposited layer.

Fig. 9. Schematic of Pt deposited layer.

Fig. 10. Schematic of bottom substrate after lift-off process. bonding process. BOE is a solution of ammonium fluoride and hydrofluoric acid that etches the exposed soda-lime glass at a rate of approx 600 nm/min at room temperature. Following this step, the sample is removed from the solution, rinsed in deionized (DI) water, and dried with nitrogen. 5. A Technics sputtering system with both DC and RF sputtering heads is used to deposit a 10-nm adhesion layer of titanium via RF sputtering (pressure = 10 mTorr; power = 300 W; t = 0.5 min) followed by DC sputtering of 300 nm of platinum (pressure = 10 mTorr; power = 300 W; t = 13 min) over the entire glass substrate (Figs. 8 and 9). 6. Patterning of the electrode material is accomplished by soaking the substrate in acetone, which causes the underlying positive photoresist layer to dissolve (Fig. 10). As this protective resist layer is removed in the solvent, the Ti/Pt electrode material located above it is “lifted off.” The electrode material remains

20

Crain et al.

anchored to the glass substrate only in the exposed recessed patterns in the glass substrate. This liftoff process allows the dual-composition electrodes to be patterned in a single processing step as opposed to wet etching, which would require multiple selective etching steps. 7. The final process is the removal of the remaining chrome layer in the nonelectrode regions. The original chrome layer which was once protected by the liftoff resist is removed by a second etch in CEP 200 Micro Chrome etchant (Fig. 11). The exposed Pt electrodes are unaffected by the selective chrome etch. This substrate only requires final cleaning before it is ready to be bonded to its companion top glass substrate that will contain the necessary etched channels and reservoirs (described in Subheading 3.4.).

3.4. Top Substrate Processing Fabrication of the top glass substrate (5 × 2.5 cm) is slightly more straightforward. This substrate contains the microcapillaries and reservoir openings

Fig. 11. Schematic of completed bottom substrate with recessed electrodes.

Fig. 12. Schematic of top substrate patterning with ultraviolet light.

Fabrication of a Glass CE Microchip With Integrated ECD

21

Fig. 13. Schematic of top substrate following removal of the photoresist layer.

Fig. 14. Schematic of top substrate following removal of chrome layer.

Fig. 15. Schematic of top substrate following removal of the final photoresist layer.

and is formed by a combination of photolithography, bulk micromachining, and conventional glass-drilling. 1. The microchannel pattern is transferred onto the precoated positive resist layer of a second ultra-flat soda-lime photomask blank (Nanofilm, Inc., Valley View, OH) using the appropriate darkfield photomask (Fig. 2A) and contact lithography (Fig. 12). The photomask blank is positioned with the patterned mask in direct contact with the photoresist. The photoresist is exposed to G-line UV light from a Mercury arc lamp with a dose of 150 mJ/cm2 to allow the resist to be patterned. 2. The exposed photoresist is then removed with MF-319 developer (Shipley Co.) to reveal the underlying chrome layer (Fig. 13).

22

Crain et al.

Fig. 16. Schematic of top substrate following glass etching.

Fig. 17. Schematic of top substrate following reservoir drilling. 3. The exposed chrome layer is then etched using CEP-200 Micro Chrome etchant, providing access to the glass substrate only in the desired microchannel regions (Fig. 14). 4. Although it is useful to keep the photoresist on the bottom glass substrate during electrode formation, it is best to remove the resist prior to etching the top microchannel glass substrate. Therefore, the top substrate is placed in acetone to remove most of the photoresist; and the remaining organic residue is removed by a short dip in NanoStrip, which is a mixture of sulfuric acid and hydrogen peroxide (Cyantek, Fremont, CA). It is important to rinse the Nano-Strip coated substrate by first placing it in a beaker of DI water and then rinsing under running DI water for 1 min. Rinsing the Nano-Strip coated substrate directly under running DI water with a layer of NanoStrip remaining on the substrate may cause a loss of the patterned chrome layer, which is the only layer remaining to act as a mask during the glass etching step (Fig. 15). 5. Once dried with nitrogen, the 20-µm deep channels are etched in the top glass substrate by placing the substrate for 1 min in 6:1 BOE (Fig. 16). Once etched, this glass must be rinsed thoroughly in DI water, dried with nitrogen, and inspected under the microscope. If a crazed formation is seen in the etched glass regions, the glass substrate should be discarded and the procedure started over since the crazes in the glass will potentially result in irregular fluid flow presumably owing to interruption of the ζ-potential at the fluid–wall interface. If the glass looks smooth but has many small, black fragments around the perimeter of the channels, then the etching process has begun appropriately. The dry glass substrate should then be dipped in Nano-Strip and rinsed in a beaker of DI water, followed by a rinse under running DI water and again dried with nitrogen. Finally, the top glass substrate is etched in BOE for 29 min, followed by the same rinsing, nitrogen drying,

Fabrication of a Glass CE Microchip With Integrated ECD

23

Fig. 18. Schematic of top substrate following removal of the chrome layer. dipping in Nano-Strip, rinsing, and nitrogen drying process previously described (see Note 3). 6. With the final etch width criteria of the channel met (80 µm for the top of the channel; 50 µm for the bottom of the channel) and depth (20 µm), the four reservoir openings (d = 5 mm) are then drilled into the top glass substrate at the positions of the terminal ends of each capillary arm (Fig. 17) using a Dremmel® tool fitted with a diamond core drill bit. These predrilled holes serve as the buffer, sample, waste, and detection reservoirs. An extra hole is drilled approximately one radius away from the center of the original detection reservoir hole to accommodate a slightly larger detection reservoir. This can be done as a possible preventive measure to minimize the gradual background current increase during amperometric detection because of the buildup of unreacted analyte after successive CE separations. 7. The final processing step involves the removal of the remaining chrome layer using CEP 200 Micro Chrome etchant (Fig. 18). The top glass substrate requires only a final cleaning before it is bonded to the bottom glass substrate containing the electrode patterns.

3.5. Glass-to-Glass Bonding Prior to bonding, the bottom glass substrate containing the electrodes and the top glass substrate containing the microchannels are dipped in Nano-Strip, rinsed with DI water in a beaker and then rinsed under running DI water, and finally dried with a nitrogen gun. An RCA1 base clean is prepared by heating five parts DI water and one part 40% ammonium hydroxide in a beaker to 70°C. Once at 70°C, the heat is turned off, and one part 30% hydrogen peroxide is stirred into the solution. The microchannel glass substrate is submerged in the solution for a minimum of 5 min. The electrode glass substrate is added to the solution for a short soak of no more than 1 min (to minimize adhesion damage to the electrodes). Both glass pieces are rinsed under running DI water for 2 min; the contacting faces of the electrode substrate and the channel substrate are brought together under running DI water. Although water is allowed to remain between the two joined pieces of glass, the outer regions of the glass are dried with nitrogen. A well-joined microchip will have just enough water between the two glass substrates so the substrates will not slide freely (see Note 4). The electrodes and channels are aligned manually under a stereo zoom optical microscope with the adhesion of the water maintaining the alignment during transit to the furnace (see Note 5).

24

Crain et al.

Fig. 19. Schematic of bonding “stack” prior to insertion in furnace for thermal bonding.

Fig. 20. Images of the (A) final microfabricated capillary electrophoresis/electrochemical detection device with integrated Pt electrodes and thermally bonded micromachined glass substrates and (B) ×40 magnification of the detection reservoir containing the electrochemical detection electrodes.

The assembled microchip is placed on an alumina tile, and the top of the microchip is covered with a second alumina tile. An additional steel block (weight) is added to the top alumina tile to provide a 50-g/cm2 pressure on the two glass plates (see Note 6). The “stack” assembly is carefully placed in a well-controlled tube furnace for thermal bonding (Fig. 19). The bonding sequence involves heating the substrates to a maximum temperature of 625°C at a ramp rate of 3°C/min. The stack is allowed to dwell at this temperature for 30 min in an air environment and is then removed using furnace gloves. With

Fabrication of a Glass CE Microchip With Integrated ECD

25

Fig. 21. Electropherogram from the capillary electrophoresis–electrochemical detection Lab-on-a-Chip device for the separation and detection of dopamine (1.1 mM) and catechol (2.3 mM). Conditions: CE voltage, 200 V; EC potential, +0.75 V vs Pt wire.

the proper safety clothing and face shield, the alumina tiles and microchip are pried apart. The microchip and alumina blocks are returned to the furnace without the steel block (load mass) and allowed to cool at a ramp down rate of 3°C/min. Careful ramping of the furnace is required in order to keep the bonded glass assembly from fracturing because of induced thermal stresses. The chip is removed when the furnace reaches 100°C. 3.6. Capillary Electrophoresis Microchip With Integrated CE/ECD Electrodes Using the previously described process, a functioning CE microchip with integrated CE and EC detection electrodes is manufactured (Fig. 20). Such devices have been utilized for hundreds of electrochemical detection experiments typically involving the detection of the neurotransmitters dopamine and catechol (Fig. 21), as well as tagged DNA-related compounds. The device has been successfully characterized and has been found to yield reproducible results over a 4-mo time period with little-to-no decrease in performance (4). It takes approx 1 wk to fabricate a device once the master photomask set has been generated. Cleanliness is critical during all stages of the manufacturing process, and it is

26

Crain et al.

especially important to minimize surface contamination during the thermal bonding process. Our research group has produced dozens of LOC prototypes in this fashion, and the manufacturing process has proven to be both reliable and reproducible. 4. Notes 1. Print shops typically require the PS format. 2. Figures 3–19 are not drawn to scale. 3. The process of etching, drying, and nanostrip cleaning may need to be repeated a number of times to maintain etch quality. 4. Occasionally there is so little water that the chips cannot be moved for alignment. If this occurs, the chips can be placed under running DI water so they can be separated and placed together again for an additional alignment attempt. 5. A final press of the top chip to the bottom chip will keep the two halves from moving with respect to each other during transit. 6. Important to use a steel block and/or weight with an overall surface area equal to or greater than the overall footprint of your chip.

References 1. Jacobsen, S. C. and Ramsey, J. M. (1997) Electrokinetic Focusing in Microfabricated Channel Structures. Anal. Chem. 69, 3212–3217. 2. Jackson, D. J., Naber, J. F., Roussel, T. J., Jr., et al. (2003) Portable high-voltage power supply and electrochemical detection circuits for microchip capillary electrophoresis. Anal. Chem. 75, 3311–3317. 3. Paegel, B. M., Hutt, L. D., Simpson, P. C., and Mathies, R. A. (2000) Turn geometry for minimizing band broadening in microfabricated capillary electrophoresis channels. Anal. Chem. 72, 3030–3037. 4. Baldwin, R. P., Roussel, T. J., Jr., Crain, M. M., et al. (2002) Fully integrated onchip electrochemical detection for capillary electrophoresis in a microfabricated device. Anal. Chem. 74, 3690–3697.

3 Micro-Molding for Poly(dimethylsiloxane) Microchips Carlos D. García and Charles S. Henry Summary In the present chapter, some basic definitions about the photolithography process are explained and then the standard preparation of the silicon wafer, the fabrication of the mold, and the preparation and assembly of poly(dimethylsiloxane) (PDMS)-based microchips are discussed. The purpose of this chapter is to describe the most used techniques for preparation of PDMS microchips. A list of tips is included in order to provide a troubleshooting guide for the most common difficulties found during the fabrication process. Some recent alternative approaches to microfabrication are also discussed. Key Words: Photolithography; photoresist; mold; substrate; PDMS; microchip; reversible sealing; irreversible sealing.

1. Introduction Microfluidic systems have found many applications in biochemical analysis, chemical reactions, and cell-based assays. These microfluidic devices have many advantages over conventional bench-top systems. Some of these advantages may include better flexibility of design, reduced use of reagents and samples, lower waste generation, increased speed, and portability (1). These characteristics may take into account the compatibility of the material used to fabricate the chip. It is well known that soft-polymeric systems have many properties that are desirable for use in microfluidic devices. Polymeric systems are less expensive than silicon and glass, and involve simpler and less expensive manufacturing processes (2). Some of the polymers used to fabricate chips are polyurethane, polycarbonate, polymethyl methacrylate (PMMA), polystyrene, polyethyleneterephthalate glycol (PETG), polyvinylchloride, polyethylene, and poly(dimethylsiloxane) (PDMS) (3). In particular, PDMS-based devices can easily and inexpensively be fabricated by casting the polymer against a mold prior to cross-linking. Because the casting step does not require From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

27

28

García and Henry

access to a clean room, this methodology is accessible to a larger number of investigators than many other methods. In the present chapter some basic definitions are presented and then, the standard preparation of the silicon wafer, the fabrication of the mold, and the preparation and assembling of PDMS-based chips will be discussed. Some recent alternative approaches to microfabrication will be also mentioned. 2. Materials 1. Silicon wafers (Silicon Inc.). 2. Piranha solution (1:1, H2SO4:H2O2). Make fresh. This is a very strong oxidizing agent, so it must be handled with extreme precaution. 3. SU-8 2035 Positive Photoresist (MicroChem Inc.). Light and temperature sensitive, stable at room temperature for at least 1 yr. 4. Developer (propylene glycol methyl ether acetate [PGMEA]) (Aldrich). 5. Methanol (Fisher). 6. Isopropyl Alcohol (Fisher). 7. Sylgard 184 and curing agent (Dow Corning).

3. Methods 3.1. Photolithography Photolithography is the basic operation that removes specific portions of the top layer on a wafer surface. In this process a required pattern is first formed in a mask and transferred to the wafer surface. In addition, it can also build up layers on the surface for molding. The transfer takes two steps. First, the required pattern is transferred to a layer of photoresist. Photoresist is a material that, after being exposed to light, changes its structure and properties owing to a polymerization reaction. The second step consists of the development of the photoresist. During this step, the unpolymerized regions of the photoresist are removed, leaving the desired pattern attached to the wafer. There are two types of photoresists. A typical positive photoresist are the phenol-formaldehyde resins, also called novolaks. Within the resist, the polymer is relatively insoluble and after the exposure to ultraviolet (UV) light, the resists converts to a more soluble state. Negative photoresists, on other hand, change from an unpolymerizated to a polymerizated state after the exposure to UV light step. Because negative photoresists are the most used for microchip CE applications, the term photoresist will be associated to negative photoresists for this chapter. 3.1.1. Making the Mask Designs of network of channels can be generated using computer drafting software (CorelDraw, FreeHand, or AutoCAD) with a wide variety of shapes and sizes. The negative of the desired patterns are printed on a high-resolution

Micro-Molding for PDMS Microchips

29

printer using a transparency slide or on a microfilm printer. The smallest features (channel widths) that can be achieved are 25 µm using this technique. These masks are very accessible (low time and cost), however standard lithography glass masks can also be used if high resolution or smaller features are needed. 3.1.2. Wafer Surface Preparation The wafers may have particulate or contamination that must be removed to ensure that the photoresist will stick to the surface. For this, a wet chemical cleaning/etching is recommended prior to applying the photoresist (see Note 1). 1. Rinse the wafer with solvent (e.g., methanol). 2. Immerse the wafer in piranha solution for 10 min. 3. Rinse with deionized (DI) water. A clean wafer can be easily identified because the surface should be hydrophobic so, the rinsing should form small water droplets. 4. Dry the surface using N2. 5. Dehydrate the surface: bake the wafer at 200°C for 15 min over a hot plate or 30 min in a convection oven.

3.1.3. Photoresist Deposition The goal of this step is to obtain a thin, uniform, and defect-free film of photoresist. The most common methodology used is spin coating. An amount of the liquid photoresist is deposited in the center of the wafer and allowed to spread while the wafer is spinning (see Note 2). The amount of resist deposited is critical only in the extremes. Too small an amount will result in incomplete resist coverage, and too much will cause a build up of a resist rim, or result in resist on the back of the wafer as seen in Fig. 1. The thickness of the layer (which later will be translated to the PDMS channel deep) is determined by the speed of the second spin cycle. In Table 1, the recommended speeds to coat a wafer with SU-8 photoresists are listed. 1. Pour around 1 mL per inch of wafer. 2. Spin the wafer at low speed of approx 500 rpm for 30 s to disperse the resin. 3. Accelerate to a higher speed to complete the spread, remove the excess, and achieve a uniform layer.

A second approach to coat the wafers with photoresist was recently proposed by Chang’s group (4). The novel process for the fabrication of SU-8 structures uses a constant-volume injection to develop films up to 1.5-mm thick by a single coating step. Without using a spin coater and a modified baking process, less resin is used and uniform, highly reproducible films (97%) were obtained (see Note 3).

30

García and Henry

Fig. 1. Effect of different amounts of photoresist poured on the wafer resulting in either incomplete coverage (A), optimum (B), or overloaded (C). Table 1 Thickness as a Function of Spin Speed for Different SU-8 Resists Product

Viscosity (cst)

Thickness (µm)

Spin speed (rpm)

SU-8 2025 SU-8 2035

4800 7000

SU-8 2050 SU-8 2075

17,000 32,000

25 35 55 110 165 225

3000 3000 2000 1000 1000 1000

3.1.4. Soft Bake After spreading the photoresist a soft-baking step is required to evaporate the solvent (cyclopentanone), increase the density of the film, and promote the adhesion of the resist to the wafer. Over-baking may cause the nonspecific polymerization on the photoresist. To perform the soft baking convection ovens, furnaces, hair dryers, or heating lamps can be used but hot plates are the most commonly used device (see Notes 4 and 5). The different temperatures and times suggested for soft baking of SU-8 2000 resists are summarized in Table 2. 1. Place the coated wafer on the hot plate at 65°C. 2. Ramp to 95°C. 3. Cool down to room temperature.

3.1.5. Alignment and Exposure If multiple layers of photoresist are required, an alignment step (positioning of a mask in a certain part of the wafer) may be included before the exposure. If only one layer of photoresist will be used, no alignment is required. However, even when a single channel will be pattered on the surface, it is convenient to inspect the wafer before placing the mask. In case of finding imperfections (bubbles, holes, and so on), the channel can be placed in a zone free of defects. In this way, the coated wafer can be used without the need of reprocessing. The conditions of the exposure step usually depend on the photoresist type and film thickness. As an example, the recommended exposure doses for SU-8 photoresists are summarized in Table 3.

Micro-Molding for PDMS Microchips

31

Table 2 Temperatures and Times Required to Perform the Soft-Bake Steps for SU-8 Photoresists on a Hot Plate Product

Thickness (µm)

Step 1 (65°C)

Step 2 (95°C)

25 35 55 110 165 225

1 2 3 5 5 5

3 5 6 20 30 45

SU-8 2025 SU-8 2035

SU-8 2050 SU-8 2075

Table 3 Recommended Exposure Doses for SU-8 2035 as Function of the Film Thickness Product

Thickness (µm)

SU-8 2025 SU-8 2035

SU-8 2050 SU-8 2075

25 35 55 110 165 225

Exposure dose (mJ/cm2) 220–240 250–350 400–500 500–650 600–650 625–675

3.1.6. Postexposure Baking After exposure, a second baking step should be performed to cross-link the exposed resin. As with the pre-exposure step, two heating ramps are recommended to decrease the heating stress and minimize cracking on the resin. The different temperatures and times suggested for soft baking of some SU-8 resists are summarized in Table 4. 3.1.7. Development There are several methods to develop a resist film. For immersion development, the wafer is placed in a Petri dish with the solvent for a specific time to dissolve the unexposed photoresist. For SU-8 photoresist, commonly used solvents are PGMEA, diacetone alcohol, and ethyl lactate. The time also depends on the thickness of the film and will be optimized by the manufacturer. Between 5 and 15 min is enough but generally, however, there are no problems associated with the over-development of the wafer. However, if the wafer is not completely developed, the channel will be poorly defined (see Note 6). A variation of the immersion procedure can be done by placing the wafer in the spin-coater and applying the development with a washing bottle while the

32

García and Henry

Table 4 Temperatures and Times Required to Perform the Postexposure Bake Steps for SU-8 Photoresists on a Hot Plate Product SU-8 2025 SU-8 2035

SU-8 2050 SU-8 2075

Thickness (µm)

Step 1 (65°C)

Step 2 (95°C)

25 35 55 110 165 225

1 2 3 5 5 5

3 5 6 20 30 45

wafer is rotated. A lower usage of solvent (and lower residue generation), better image definition, and the removal of the small pieces of polymerized resin from the wafer, make this technique very attractive (see Note 7). Nevertheless, the compatibility of the solvent and spinner material should be taken into consideration. 3.1.8. Hard Bake Hard baking is the third heat treatment performed after development of the mold. The procedure is basically performed as with the other baking steps with the purpose to harden the photoresist and improve the adhesion. The molds used to make PDMS chips usually have good mechanical properties and can be reused many times; however, if some specific application of the photoresist requires a hard bake, it can be performed by applying 150–200°C for several minutes (see Notes 8 and 9). 3.2. Characteristics and Preparation of the Substrate This elastomeric material is supplied in matched kits consisting of the base PDMS material and a curing agent in separate containers. 1. Mix a w/w ratio of 10:1 (base and curing agent, Sylgard 184). 2. Vacuum degassing is highly recommended to reduce the number of bubbles. A residual pressure of 10–20 mmHg applied for 30 min will sufficiently de-gas the material. The mixture is stable at room temperature for several hours. 3. Cure in 2 h at 65°C. The same results can be obtained by exposing the material for 24 h at 23°C, 2 h at 65°C, 1 h at 100°C, or 15 min at 150°C.

3.3. Assembling the Chip The PDMS channel can be sealed to a variety of surfaces including glass, PDMS, and hard plastics by a variety of methods. Compared with glass or other polymers, sealing PDMS channels is easier, faster, and high temperatures

Micro-Molding for PDMS Microchips

33

or voltages are not required. Two main sealing methods can be performed: reversible or irreversible. 3.3.1. Reversible Sealing This methodology has several advantages, such as the possibility of making a precise alignment or the chance of dismounting the chip if occasional clogging occurs. 1. Rinse both surfaces with methanol. 2. Make the conformal contact. 3. Dry the remaining solvent in an oven at 50–65°C. The two layers will be held in place by van der Waal’s forces.

3.3.2. Irreversible Sealing 1. Rinse both surfaces with methanol. 2. Expose both pieces to an air plasma source for 20–60 s at medium or high power to oxidize the exposed surfaces. 3. Put the two pieces together. The resulting silanol-rich surfaces will react covalently when brought in contact with each other, resulting in a permanent bond. An advantage of plasma sealing is that the chips will have higher EOF values, compared with reversible sealing (3).

Another way to permanently seal two layers of PDMS is to add an excess of monomer to one layer and an excess of curing agent to the other. When the two parts are put together the interface will react to form a material indistinguishable from the bulk (5). 3.4. Process Checklist 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Design and print the mask. Wafer cleaning: oxidation and dehydration. Coat the wafer with the photoresist. Prebaking at 65°C/95°C. Alignment and exposure to UV. Postexposure baking at 65°C/95°C. Develop. Mix the elastomer with the curing agent and de-air. Pour the mixture on the wafer and bake for 2 h at 60°C. Peel off and seal.

3.5. Alternative Technologies 3.5.1. Fast Prototyping A new method for the ultra-rapid prototyping of microfluidic systems using liquid phase photopolymerization, requiring less than 5 min from

34

García and Henry

design to prototype was recently presented (6). The method consists of the following steps: introduction of a liquid prepolymer into a plastic cartridge, UV exposure through a mask to define the channel geometry, removal of unpolymerized prepolymer, and a final rinse. Rapidly fabricated masters for PDMS micromolding were also demonstrated and compared with SU-8 50 photoresist processes. This is a novel approach to the fabrication of either molds for PDMS chips or chips, however, no capillary electrophoresis applications were presented. 3.5.2. Direct-Write There are a large number of different technologies that allow the specific deposition of materials with differences in resolution, speed, capability of switching materials, and applications. Direct-write technologies do not compete with photolithography for size and scale, but rather complement it for specific applications such as electronic components, sensors, and micromachining (7). Some advantages of this emerging technology may include low cost and higher speed. 4. Notes The described procedure is generally reproducible and by following the instructions no problems should appear. However there are some tips that may help to improve the results. 1. The hydrophobic clean wafer surface is highly reactive and will tend to come back to the hydrophilic state. To avoid that, keep the room humidity as low as possible and coat the wafer with the photoresist as soon as possible after cleaning. 2. Instead of using two hot plates (at two different temperatures) a temperature ramp is recommended. By this, a more uniform heating process is achieved, decreasing the heating stress and improving the adhesion of the photoresist to the wafer. 3. To decrease the number of air bubbles in the photoresist, use a syringe with a larger diameter opening to dispense the resist. If this does not work, the resin can be poured straight from the bottle to the wafer. Another possibility is to warm up the resist at 60°C for 30 min in order to decrease the viscosity and allow the air bubbles to escape from the liquid. Remember to cool down the resist to room temperature before the spinning step. 4. If little holes appear on the resist after the spinning and during the prebaking step, the resin may be contaminated with dry-resin particulates. To solve this problem, the resin can be baked at 60°C for 30 min and filtered. 5. To know if the developing process is complete, the wafer can be rinsed with isopropyl alcohol. If white spots appear, the development is not complete. The wafer can be returned to the developer to complete the development. 6. To obtain cleaner molds, the wafer can be rinsed with fresh developer (PGMEA) using a washing bottle. Residual photoresist particles will be removed by this method.

Micro-Molding for PDMS Microchips

35

7. Once the photoresist is exposed, baked, and developed it is resistant to almost every solvent and can be rinsed with organic solvents (methanol, ethanol, isopropanol, acetone, PGMEA) or water without changing the structure. 8. There are some ways to remove developed photoresist from the wafer. If this becomes necessary, the easiest way is by lifting it off with a razor blade. Another possibility to remove poorly cross-linked resist is by the use of a commercially available remover (PG Remover, from MicroChem Inc.). 9. For higher aspect ratios, try longer baking and exposure steps. If really high aspect ratios (>100) and resolution are needed, the use of other radiation sources like X-rays, electron beams, or excimer lasers is recommended.

References 1. Ng, J. M., Gitlin, I., Stroock, A. D., and Whitesides, G. M. (2002) Components for integrated poly(dimethylsiloxane) microfluidic systems. Electrophoresis 23, 3461–3473. 2. Becker, H. and Locascio, L. (2002) Polymer microfluidic devices. Talanta 56, 267–287. 3. Duffy, D. C., McDonald, J. C., Schueller, O. J. A., and Whitesides, G. M. (1998) Rapid prototyping of microfluidic systems in poly(dimethylsiloxane). Anal. Chem. 70, 4974–4984. 4. Lin, C. -H., Lee, G. -B., Chang, B. -W., and Chang, G. -L. (2002) A new fabrication process for ultra-thick microfluidic microstructures utilizing SU-8 photoresist. J. Micromech. Microeng. 12, 590–597. 5. Unger, M. A., Chou, H. -P., Thorsen, T., Scherer, A., and Quake, S. R. (2000) Monolithic microfabricated valves and pumps by multilayer soft lithography. Science 288, 113–116. 6. Khoury, C., Mensing, G. A., and Beebe, D. J. (2002) Ultra rapid prototyping of microfluidic systems using liquid phase photopolymerization. Lab on a Chip 2, 50–55. 7. Chrisey, D. B. and Pique, A. (2002). Introduction to Direct-Write technologies for rapid prototyping, in Direct-Write Technologies for Rapid Prototyping Applications: Sensors, Electronics, and Integrated Power Sources (Pique, A. and Chrisey, D. B., eds.), Academic Press, San Diego, CA, pp. 1–16.

4 Fabrication of Polymer Microfluidic Systems by Hot Embossing and Laser Ablation Laurie E. Locascio, David J. Ross, Peter B. Howell, and Michael Gaitan Summary Fabrication of microfluidic channels in common commercially available thermoplastic materials can be easily accomplished using hot embossing or ultraviolet (UV) laser ablation. Hot embossing involves replication of a microfluidic network in a polymer substrate from a stamp (or template) fabricated in silicon or metal. UV laser ablation is performed by either exposing the polymer substrate through a mask or by using a laser direct-write process. The resulting polymer microfluidic channels are most often sealed with another polymer piece using thermal bonding or solvent bonding to complete the fabrication procedure. Unlike their silicon and glass counterparts, polymer microfluidic systems can be fabricated by these methods in less than 1 h, making the materials attractive for both research prototyping and commercialization. Key Words: Hot embossing; laser ablation; polymer; microfluidics; thermoplastic; bonding; excimer laser.

1. Introduction In the late 1990s, several new techniques were introduced for the fabrication of polymer microchannels including hot embossing (1), laser ablation (2), X-ray photolithography (3), injection molding (4), and soft embossing (5). Polymers have distinct advantages over other substrates for many applications because of their low cost and the ease with which they can be micromachined. In fact, techniques such as polymer embossing and laser ablation made microfluidics technology widely accessible, thus, greatly expanding the field of microfluidics in a short time. Our group was one of the first to introduce methods for hot embossing and room temperature embossing (stamping, imprinting), and we have demonstrated their usefulness for fabricating microchannel networks in a variety of thermoplastic materials (1). During the hot embossing process, a polymer is pressed against a stamp (template) while heating to a temperature greater than From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

37

38

Locascio et al.

the glass transition temperature allowing the polymer to flow and, thus, leaving an imprint of the stamp feature. In the room temperature embossing process, a polymer is pressed against a stamp under higher pressure, thus transferring the feature without allowing the polymer to flow and then reform. Both methods of embossing permit rapid prototyping of microchannels that can be accomplished in minutes. The limiting step in this process is the fabrication of the micromachined stamp or template. Once fabricated, however, the stamp can be used repeatedly to emboss hundreds of polymer microchannel networks. For rapid device prototyping, laser ablation (2) is an attractive alternative to embossing because it does not require the fabrication of an embossing template. Potential drawbacks to laser ablation are that the ablated channel surfaces are typically much rougher than embossed surfaces, and the surface chemistry (hydrophobicity, surface charge) of ablated channels is very different from that of the native sheet plastic. The modification of surface chemistry by laser ablation can be used to advantage, however, as it allows for the potential machining and chemical modification of microfluidic channels in one step. The surface chemistry resulting from ablation can even be manipulated through variation of the atmosphere surrounding the plastic during ablation (6). 2. Materials (see Note 1) 1. For hot embossing, commercial thermoplastic sheet such as polycarbonate (PC), polystyrene (PS), or polymethylmethacrylate (PMMA) is used. Also required for this procedure are a micromachined template or stamp, a hydraulic press, and two polished temperature-controlled heating blocks. The materials and methods are described in detail next for the silicon micromachining; however, this requires a clean room and microfabrication capabilities. Alternatively, the template can be fabricated at fully equipped microfabrication facilities that provide contract services. 2. Materials required for laser micromachining include polymer or glass substrates (there is much more flexibility in the choice of material when using laser ablation). Also needed are a laser ablation system with a pulsed UV excimer laser, a motorized translation stage, optics to shape and direct the laser output to the polymer surface, and a computer to control laser firing and stage movement. 3. For polymer channel sealing, materials include a drill, microscope slides, binder clips, and a circulating air oven.

3. Methods 3.1. Polymer Microchannel Fabrication by Hot Embossing Microfluidic channels are fabricated by hot embossing using a micromachined silicon wafer (1) as a stamp or template. The fabrication procedure for the micromachined silicon template is described in detail in Subheading 3.1.1., followed by the procedure for making polymer microfluidic systems from these templates.

Fabrication of Polymer Microfluidic Systems

39

3.1.1. Silicon Template Micromachining 1. Low-doped (100) silicon wafers are placed in a wet oxidation furnace at 980°C for 20 min to create a 100-nm thick oxide film. 2. Photoresist is spun-on the wafer and the wafer is soft baked. Hexamethyldisilazane is spun-on before application of the photoresist as an adhesion promoter. In addition, photoresist is spun-on to the backside of the wafer. Finally, the wafers are rinsed in deionized (DI) water and spin dried. 3. A photomask is prepared for patterning the photoresist. For this work, our designs generally require a 20-µm minimum feature size. We create the photomasks by high-resolution printout onto clear transparency film. Alternatively, chromium photomasks are fabricated for us at the University of California, Berkeley Microfabrication Laboratory (http://microlab.berkeley.edu/) when smaller features are required. The mask contains the drawings of the fluid channel elements and an alignment mark for rotational alignment to the flat of the silicon wafer. All features are drawn in Manhattan format, i.e., lines or boxes parallel or perpendicular to the alignment mark, so that they will be aligned to the (111) planes of the silicon wafer. This alignment is required for anisotropic etching of the silicon wafer. 4. The mask is placed over the front side of the silicon wafer and aligned to the flat using the alignment mark. The photoresist is exposed using an UV light source and patterned using a developer, then rinsed and dried. 5. Once the photoresist is patterned, it is used to pattern the silicon dioxide film by a 2 min wet etch in a 6% (volume fraction) buffered hydrogen fluoride (HF) oxide etch solution, then rinsed and dried. Following this, the photoresist is stripped in acetone, and then the wafer is again rinsed and dried. 6. The silicon wafers are then anisotropically etched in tetramethylammonium hydroxide (TMAH) solution. We prepare a TMAH etch solution, based on the work reported by Tabata et al. (7), consisting of 450 mL of 25% by weight electronic grade TMAH in aqueous solution mixed with 900 mL of DI water (one part 25% TMAH to two parts DI water). The solution is placed in a reflux etch container and heated to 80°C on a hotplate. The wafers are placed in a Teflon holder and immersed in the solution. Following the work by Klaassen et al. (8), 1 g of ammonium peroxydisulfate (APODS) in powder form is added every 10 min during etching in order to eliminate the formation of hillocks. The etch rate for this solution in the (100) direction was measured as 0.9 µm/min. 7. Finally, the silicon dioxide mask is stripped by a second 2 min wet etch in 6% buffered HF, then rinsed and dried. The cross-section of the raised features on the template are trapezoidal-shaped as depicted in Fig. 1, with a width at the top surface ranging between 20 and 100 µm depending on our application. Smaller template features can be made using the Cr masks. The sides of the trapezoid are aligned to the (111) planes of the silicon crystal forming a 54.74° angle from the wafer surface plane (9). 8. The silicon template may be used to emboss channels, or the micromachined silicon template may be used as a master to fabricate a more robust template in metal (4). In the second case, a metal electroform is made from the silicon master that is

40

Locascio et al.

Fig. 1. Micromachined silicon template for channel embossing. the mirror image of the silicon template. Then, a second metal electroform is created from the first electroform that is an exact replica of the original silicon template. Thus, micrometer features are transferred to a more robust metal substrate that can be used to fabricate microchannels in plastic substrates by hot embossing or injection molding.

3.1.2. Polymer Embossing We have used the etched silicon template to emboss microchannels in plastic materials at room temperature (10), or at elevated temperatures (hot embossing) (11). The hot embossing procedure is as follows: 1. Many types of commercially available sheet plastic, such as PS, PC, PMMA, polyethylenetetraphthalate glycol (PETG), and polyvinylchloride (PVC), have been used to fabricate polymer microchannels (1,10,12–15). In our lab, the sheet plastic is cut into rectangular pieces that are approx 2.0 × 6.0 cm. A cut plastic piece is washed with ethanol and then dried under nitrogen, and placed on top of a clean silicon template in a laminar flow hood. 2. The silicon template and plastic piece are placed between two polished aluminum blocks, one on the top and one on the bottom, each with its own embedded heater connected to a common temperature controller. The aluminum blocks are heated to a temperature either equal to or slightly greater than the glass transition temperature of the plastic material, but below the melt temperature. 3. The assembly composed of the aluminum blocks sandwiching the silicon template and plastic piece is placed in a hydraulic press. Pressure (2.8 × 106 Pa to 2.1 × 107 Pa, 400–3000 psi) is applied for a time that ranges typically from 5 to 60 min depending

Fabrication of Polymer Microfluidic Systems

41

Table 1 Hot Embossing Parameters for Various Polymer Substrates Polymer substrate PMMA PETG PC

Temperature (°C)

Time (min)

Pressure (Pa)

110 80 155

60 20 90

5.1 × 106 (740 psi) 3.4 × 106 (500 psi) 1.4 × 107 (2000 psi)

Fig. 2. Hot embossed microfluidic channel in polymethylmethacrylate. on the plastic used and the embossing temperature. After the required time, the pressure is released, and the top aluminum block is removed. Typical conditions for imprinting polymer microchannels are described in Table 1. 4. The silicon template and the plastic piece are removed together with tweezers from the bottom aluminum block while at elevated temperature, and are placed on the bench top for cooling. Upon cooling, the plastic piece releases rapidly from the silicon template owing to a mismatch in the thermal expansion coefficients of the two materials. The resulting plastic piece contains microchannels that are the exact mirror of the raised features on the silicon template as shown in Fig. 2. 5. If the plastic piece warps significantly upon rapid cooling and release, the assembly (aluminum blocks, silicon template, plastic piece) can be cooled to approx 10°C below the glass transition temperature while still under pressure. Then pressure is released and the top aluminum block is removed. The silicon template and the plastic piece are removed from the bottom block and cooled to room temperature on the bench top as stated previously.

Alternatively, plastic devices can be imprinted at room temperature, rather than at elevated temperatures, and at higher pressures. When devices are imprinted at room temperature, the microchannel depth is dependent on imprinting pressure,

42

Locascio et al.

imprinting time, and properties of the plastic; therefore, the resulting channel may not be an exact inverted replica of the silicon template. An advantage of room temperature imprinting, however, is that fabrication time is significantly reduced as compared with hot embossing. Reproducible imprints can be made at room temperature in less than 2 min (9). 3.2. Polymer Microchannel Fabrication by Laser Ablation A typical direct-write laser ablation system similar to the one in our laboratory consists of a pulsed UV excimer laser, a motorized translation stage that holds the substrate to be ablated, the optics to shape and direct the laser output to the surface to be ablated, and a computer to coordinate and control the translation stage and the firing of the laser. During ablation, the laser is fired with a repetition rate of 100 Hz, and the translation stage is used to move the plastic substrate beneath the laser to form microchannels. The basic procedure for direct-write laser ablation of a microfluidic device using a system from Potomac Photonics (see Note 1) (Lanham, MD) is as follows: 1. The laser system is first turned on and warmed up, and the laser cavity gas is changed, if necessary. Changing the gas and firing the laser for a few minutes at 100 Hz helps to ensure a more constant laser power and hence more uniform channel depths. We use either ArF or KrF gas to ablate at 193 or 248 nm, respectively. 2. The desired beam-defining aperture is installed, and the laser power is adjusted for the desired depth of cut. The aperture, along with the system optics, is used to set the shape and width of the ablated channels. The depth of the channels is determined by the material, laser pulse power (and wavelength), the repetition rate, and the feed rate of the translation stage. In most of our work, 50–60 µm wide by 50–100 µm deep channels are cut in commercially available sheet plastics using a round aperture, a power of 65 µJ/pulse, a repetition rate of 100–200 Hz, and a feed rate of 50 µm/s. Figure 3 shows examples of polymer microchannels ablated using these parameters. 3. The geometry of the microchannel network for the device to be fabricated is determined by a simple coded program that specifies sequences of translation stage movements and laser firings. New device designs can be created or old designs modified with as little as a few minutes of programming. 4. The plastic substrate is affixed to the translation stage, and the program is started to cut the device. Using the settings given in ref. 2, a device can be ablated in between 5 min and 2 h depending on the complexity. 5. The ablation process can create a great deal of particulate matter that can be redeposited onto the substrate and in the newly formed microchannels. After ablation, much of this debris is removed using a combination of rinsing and sonication in water, buffer, ethanol, or a mixture of the three. Once the particulate matter is removed, the device is ready for sealing using the same techniques applicable to embossed channels.

Fabrication of Polymer Microfluidic Systems

43

Fig. 3. Microchannel cross-sections for laser ablation in various sheet polymers. All of the channels were cut using the same laser settings: a round aperture to define a 60-µm spot at the surface of the polymer, a laser power of 65 µJ/pulse, a laser repetition rate of 200 Hz, and a translation rate of 50 µm/s.

3.3. Sealing Polymer Microchannels by Thermal Bonding Polymer devices can be sealed by a number of techniques that include thermal bonding and solvent bonding. In our laboratories, most devices are sealed by thermal bonding processes that are described next. 1. A polymer piece that serves as the lid of the device is cut from sheet plastic to be approximately the same dimensions as the substrate containing the embossed or ablated channels. The polymer lid can be cut from the same substrate material that was used to form the microchannels. Alternatively, for easier sealing, the polymer lid can be cut from a different material with a lower glass transition temperature than that used to create the microchannels. 2. Holes are drilled into the lid to provide fluidic access. After drilling, the lid substrate is thoroughly rinsed in water, then ethanol, and dried under a compressed nitrogen jet to remove particulates. 3. The lid is placed over the microchannel substrate aligning the holes in the lid to the ends of the microchannels. The two substrates, lid and microchannel, are clamped together between four microscope glass slides, two on each side, using binder clips. The device is then thermally bonded in a circulating air oven. It is important to keep the time and temperature as low as possible during the sealing process to avoid physical alteration of the microchannel feature. Typical sealing times and temperatures for three common polymers are given in Table 2.

3.4. Polymer Microchannel Fabrication by Hot Embossing 1. When hot embossing, the polymer substrate should fit entirely on the silicon substrate. If the polymer extends over the edge of the silicon template during heating,

44

Locascio et al. Table 2 Sealing Parameters for Various Polymer Substrates Polymer substrate PMMA PETG PC

Sealing temperature (°C)

Time (min)

103 75 135

12 10 60

the polymer can deform around the edges of the silicon template, thus preventing its removal once the process is complete. 2. When hot embossing, pressure should not be applied in the hydraulic press until the desired temperature is reached. Applying pressure at cooler temperatures may result in silicon template fracture. Careful handling of the silicon template during hot embossing will permit repeated usage of the template for the fabrication of hundreds of plastic microfluidic devices. 3. When room temperature imprinting, the polymer piece should be larger than the silicon template. If the polymer substrate does not cover the template completely, the template will crack when pressure is applied. When embossing at room temperature, the lifetime of a silicon template is much shorter than with hot embossing and the template is subject to fractures, particularly with harder plastics such as PMMA and PC. Much longer template lifetimes can be achieved when embossing softer plastics, such as PVC or PETG.

3.5. Polymer Microchannel Fabrication by Laser Ablation 1. During laser ablation, it is important to keep the lens-to-polymer distance constant and to maintain constant laser power. 2. When ablating deep channels, the laser will defocus, thus changing the geometry of the ablated feature. 3. When ablating cross intersections, the point in the middle of the intersection will be approximately two times as deep as the surrounding microchannels unless great care is taken to prevent this by modifying laser parameters when ablating the intersection.

3.6. Sealing Polymer Microchannels by Thermal Bonding 1. When sealing polymer microchannels using thermal bonding, binder clips should be aligned directly over the channels so that air is not entrapped in the vicinity of the microchannel. 2. If the microchannel sealing is difficult to accomplish without melting the embossed feature, a polymer with a lower glass transition temperature can be used to seal the device. In this case, the sealing temperature is approximately the glass transition temperature of the lid material and not the channel material. 3. Cooling to room temperature while the template and plastic piece are under pressure will break the silicon template. Always release the pressure when cooling to a temperature that is less than 10°C below the polymer glass transition temperature.

Fabrication of Polymer Microfluidic Systems

45

4. Good air circulation is required to maintain good temperature control during sealing. The best results are obtained in a circulating air oven such as a gas chromatography oven. 5. Room temperature imprinted channels cannot generally be sealed using thermal bonding. The polymer will often flow during the bonding process and destroy the feature during this type of sealing process. 6. After drilling the fluid access holes in the device lid, it is important to remove all particulates that are on the edge of the hole as a result of drilling. If the edge is not completely clear of particulates, or if the edge is rough, the end of the microchannel will become occluded during sealing.

4. Notes 1. Certain commercial equipment, instruments, or materials are identified in this report to adequately specify the experimental procedure. Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.

References 1. Martynova, L., Locascio, L. E., Gaitan, M., Kramer, G. W., Christensen, R. G., and MacCrehan, W. A. (1997) Fabrication of plastic microfluid channels by imprinting methods. Anal. Chem. 69, 4783–4789. 2. Roberts, M. A., Rossier, J. S., Bercier, P., and Girault, H. (1997) UV laser machined polymer substrates for the development of microdiagnostic systems. Anal. Chem. 69, 2035–2042. 3. Soper, S. A., Ford, S. M., Qi, S., McCarley, R. L., Kelly, K., and Murphy, M. C. (2000) Polymeric microelectromechanical systems. Anal. Chem. 72, 642A–651A. 4. McCormick, R. M., Nelson, R. J., AlonsoAmigo, M. G., Benvegnu, J., and Hooper, H. H. (1997) Microchannel electrophoretic separations of DNA in injectionmolded plastic substrates. Anal. Chem. 69, 2626–2630. 5. Duffy, D. C., McDonald, J. C., Schueller, O. J. A., and Whitesides, G. M. (1998) Rapid prototyping of microfluidic systems in poly(dimethylsiloxane). Anal. Chem. 70, 4974–4984. 6. Pugmire, D. L., Waddell, E. A., Haasch, R., Tarlov, M. J., and Locascio, L. E. (2002) Surface characterization of laser-ablated polymers used for microfluidics. Anal. Chem. 74, 871–878. 7. Tabata, O., Asahi, R., Funabashi, H., Shimaoka, K., and Sugiyama, S. (1992) Anisoptropic etching of silicon in TMAH solutions. Sens. Actuators A Phys. 34, 51–57. 8. Klassen, E. H., Reay, R. J., Storment, C., et al. (1996) Micromachined Thermally Isolated Circuits. Proc. Solid-State Sensor and Actuator Workshop, 127–131. 9. Seidel, H. (1987) The Mechanism of Anisotropic Silicon Etching and its Relevance for Micromachining. Proc. Transducers 87, 120–125. 10. Xu, J. D., Locascio, L., Gaitan, M., and Lee, C. S. (2000) Room-temperature imprinting method for plastic microchannel fabrication. Anal. Chem. 72, 1930–1933.

46

Locascio et al.

11. Johnson, T. J., Waddell, E. A., Kramer, G. W., and Locascio, L. E. (2001) Chemical mapping of hot embossed and UV laser ablated microchannels in poly(methyl methacrylate) using carboxylate specific fluorescent probes. Appl. Surf. Sci. 181, 149–159. 12. Henry, A. C., Waddell, E. A., Shreiner, R., and Locascio, L. E. (2002) Control of electroosmotic flow in laser-ablated and chemically modified hot imprinted poly(ethylene terephthalate glycol) microchannels. Electrophoresis 23, 791–798. 13. Becker, H. and Heim, U. (2000) Hot embossing as a method for the fabrication of polymer high aspect ratio structures. Sens. Actuators A Phys. 83, 130–135. 14. Ueno, K., Kitagawa, F., Kim, H. B., et al. (2000) Fabrication and characteristic responses of integrated microelectrodes in polymer channel chip. Chem. Letters 8, 858, 859. 15. Uchiyama, K., Xu, W., Yamamoto, M., Shimosaka, T., and Hobo, T. (1999) Development of imprinted polymer microchannel capillary chip for capillary electrochromatography. Anal. Sciences 15, 825, 826.

II SURFACE MODIFICATION METHODS

5 Surface Modification Methods for Enhanced Device Efficacy and Function Barbara J. Jones and Mark A. Hayes Summary Currently available microfluidic devices can accomplish a variety of tasks useful in molecular biology. When moving analytical processes to a microenvironment, the properties of the device surface play a larger role in the functioning of the device. Surface modification may become necessary or advantageous for the purpose of control of the functional mechanics of the device, keeping cell components from adsorbing, attaching antibodies to the surface for detection of biological components, and attaching a functional bonding complex. Modification of the surface of microfluidic devices for the control of flow and device function, or for funtionalization of the surface to tailor the device to a specific use, can be accomplished in numerous bench-top, postfabrication procedures. The use of polyelectrolyte multilayers, ultraviolet grafting of polymers, and polydimethylsiloxane/surfactant coating to control flow and mitigate adsorption is discussed. In addition, the funtionalization of devices through amine termination of surfaces, and immobilization of biotin within a phosphotidylcholine bilayer is detailed. Key Words: Surface modification; biotinylation; microfluidics; polyelectrolyte multilayer.

1. Introduction Microfluidic devices will be used by every molecular biologist at some point in their career. Is that a wild and overstated claim? Probably not. There are a myriad of reasons why microfluidic devices will be commonplace. They reduce reagent needs, waste production, and time; the financial savings alone make the adaptation of devices to your particular needs worth exploring. The commercial devices already available can accomplish a variety of tasks, from protein separation and DNA sequencing, to combinatorial chemistry. One key aspect of microfluidic devices is that the samples are necessarily small and the ratio of the volume to the surface area shrinks significantly. As a consequence of this, the surface properties become important and certain enhancements may become necessary. Surface modification plays four overlapFrom: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

49

50

Jones and Hayes

ping roles. Control of the surfaces offers the best means of control for the functional mechanics of the device. Modifications such as dynamic and covalently bound coatings will keep proteins and “gooey” cell components from sticking (1–7). Attachment of antibodies to the surface will allow you to better detect biological components (8,9). And finally, attaching a functional bonding complex to the surface makes individualized attachments possible (4). All of this can be done on your bench-top specific to your desired application. In addition to using small samples, microchips are fabricated in a wide array of materials, and specific surface modification may indeed be necessary for the optimization of your method. There are many modification methods, some of which are detailed here, that will help you to effect better separation of biological components and also provide the ability to modify your device to perform your specific analysis. The purposes for modifying the surface of microfluidic devices fall into two main categories: improvement of the surface properties for the normal function of the device, and funtionalization of the surface to tailor the device to a specific use. This chapter will provide several techniques for each. 1.1. Improvement of Surface Properties of Microfluidic Devices Improvement of surface properties may be necessary in any of the devices that you choose, whether you purchase commercially available devices or fabricate them yourself. Devices produced in polymers, such as poly(dimethylsiloxane) (PDMS) or polycarbonate, tend to be hydrophobic in character and, thus, loading of the channels can prove difficult. These polymer surfaces can also have poorly controlled electroosmotic flow (EOF; a property that greatly affects the separation efficiency and precision of a device; see Chapter 13). In addition, adsorption of biological components to the surface and even migration into the polymer matrix makes surface modification necessary. A promising technique that will be detailed here is the application of polyelectrolyte multilayers (PEMs) to these polymer device channel surfaces (2,7). This technique provides a simple and reproducible way to modify the surface of any polymer-based microfluidic device to provide a wettable channel with reproducible and stable EOF. An increased benefit to surface modification using this technique is the reproducibility of the EOF across devices fabricated from different polymers. The application of PEMs is easily achieved. The technique consists merely of filling the channel successively with alternating solutions of positive and negative polyelectrolytes allowing for the multilayers to form electrostatic bonds. Though these layers are not covalently bound to the surface, the multilayers provide complete coverage and remain robust even after long-term storage. A second technique for imposing a hydrophilic character on PDMS surfaces involves the ultraviolet (UV) grafting of polymers to the surface of the

Surface Modification Methods

51

channels (5). This grafting is achieved by first creating grafting sites, radicals, at the surface by exposing the surface to UV irradiation while simultaneously exposing the device to a monomer solution. The monomers quickly react to form a polymer covalently bound at the reactive site. Devices fabricated in glass are highly charged on the surface and proteins often adsorb to these surfaces, making separations difficult. The manipulation of this surface charge and thus the EOF through changes in buffer pH can be useful; however, this variability in EOF may be a hindrance in many applications. To control EOF and decrease adsorption of proteins and other biological components, a PDMS/surfactant coating of glass microdevices can be utilized (1). 1.2. Funtionalization of Surfaces Funtionalization of the surfaces describes the immobilization of a component to the surface of a device, which then provides an available functional group for further bonding. This type of surface preparation is useful when developing assays or combinatorial reactions within the channels of the microdevice. There are many ways to functionalize devices; however, this discussion will be limited to funtionalization that is easily produced in postfabricated devices. The modification of poly(methylmethacrylate) (PMMA) devices to provide an amine-terminated surface has been demonstrated (4). These surfaces are then available for attachment of any variety of functional groups or targets. The procedure for amine termination can be accomplished on benchtop with little procedural complexity. The PMMA device pieces are first exposed to N-lithiodiaminopropane (N-LDAP). The reaction is then simply quenched with water. The amineterminated surface can additionally be exposed to octadecane chains to produce a highly ordered monolayer on the channel surfaces. Immobilization of biotin within a phosphotidylcholine bilayer on the surface of either PDMS or glass microdevices allows the streptavidin–biotin conjugation to be exploited in any desired utility. In this method, vesicles that contain a small percentage of biotinylated lipids are infused into the channels of either PDMS or glass microdevices where the vesicles adsorb to the surface and fuse to form a uniform bilayer coating (9). The biotin headgroups are then available at a number of surface sites. Any enzyme or reagent of interest that can be linked to streptavidin can thus be flowed over the biotinylated surface and immobilized at these sites. 2. Materials 2.1. Polyelectrolyte Multilayer (Option 1) 1. Dextran sulfate (MWav 5000, DS), 3% solution in ultrapurified water. 2. Hexadimethrine bromide (Polybrene, PB), 5% solution in ultrapurified water. 3. 0.1 M NaOH.

52

Jones and Hayes

2.2. Polyelectrolyte Multilayer (Option 2) 1. Polystyrene microdevice top section with imprinted channels and a PDMS cover film. 2. 60 mM poly(styrene sulfonate) sodium salt (PSS) (MWav 500,000) in 0.5 M NaCl adjusted to pH 9.0 with NaOH. 3. 20 mM poly(allylamine hydrochloride) (PAH) (MWav 70,000) in 0.5 M NaCl adjusted to pH 9.0 with NaOH. 4. 1 M NaOH. 5. All water used is ultrapurified (18 MΩ cm).

2.3. UV Graft Polymerization 1. Microfluidic device fabricated in PDMS. 2. Any of the following monomers: a. Acrylic acid (AA). b. Acrylamide (AM). c. Dimethylacrylamide (DMA). d. 2-hydroxylethyl acrylate (HEA). 3. Benzyl alcohol. 4. Sodium periodate (NaIO4).

2.4. PDMS Coating on Glass Microchips 1. 2. 3. 4. 5. 6.

Glass microchip with channels. 0.1 M NaOH. Dichloromethane (dried and distilled). PDMS 200. 0.01% Tween-20. Dry N2 gas.

2.5. Amine-Terminated PMMA 1. 2. 3. 4. 5. 6. 7.

PMMA microdevice, top and bottom separated. N-LDAP. n-Octadecane-1-isocynate (99%). Hexanes. Toluene. Acetone. Dry N2 gas.

2.6. Immobilization of Biotin Within a Phosphotidylcholine Bilayer 1. PDMS microdevice. 2. 1,2-dilauroyl-sn-glycero-3-phosphocholine (DLPC) (Avanti Polar Lipids, Inc., Alabaster, AL). 3. N-Biotinyl-CAP-PE (Avanti Polar Lipids). 4. Chloroform. 5. Phosphate buffer (pH 7.4, ionic strength 150 mM).

Surface Modification Methods

53

6. 5 mg/mL bovine serum albumin (BSA) in pH 3.8 phosphate buffer solution. 7. NaCO3/NaHCO3 buffer (pH 9.8, ionic strength 150 mM).

3. Methods 3.1. Polyelectrolyte Multilayer (Option 1) 1. Rinse PDMS channels with 0.1 M NaOH for 4 min, then water for 4 min by filling inlet reservoirs with the rinsing solution and applying a vacuum to the waste reservoir (see Note 1). 2. Fill the channel with 5% PB solution to form the cationic layer and let stand 2 min, then remove with vacuum. 3. Let stand 15 min empty (see Note 2). 4. Fill the channel with 3% DS solution to form the anionic layer and let stand 2 min, then remove with vacuum. 5. Let stand 15 min empty. 6. A subsequent cationic layer can be applied if desired by repeating steps 2 and 3 (see Note 3).

3.2. Polyelectrolyte Multilayer (Option 2) 1. 2. 3. 4. 5. 6. 7. 8. 9.

Separate channel section of device from cover film. Rinse channel section with NaOH at 55°C for 15 min (see Note 4). Rinse with water thoroughly and dry with nitrogen. Pipet PAH solution (cationic) onto device, completely covering channels. Allow to stand for 30 min. Remove the PAH solution by rinsing thoroughly with water. Pipet PSS solution (anionic) onto device, completely covering channels. Allow to stand for 30 min. Continue to alternate the PAH and PSS solutions for 5 min with water rinses between each deposition, 14–15 layers should be present depending on the desired nature of the top layer (cationic or anionic) (see Note 5). 10. Cover with the PDMS film and cut inlet and outlet wells using a cork bore.

3.3. UV Graft Polymerization 1. Clean and thoroughly dry PDMS microfluidic device segments (channels and cover film). 2. Prepare an aqueous solution of 0.5 mM NaIO4, benzyl alcohol (0.5% wt), and one of the monomers listed in Subheading 2.3. (10% wt). 3. Immerse the device segments in the monomer solution. 4. Place the immersed segments under a 200-W mercury lamp, with the distance between the sample and the lamp at 5 cm. 5. Irradiate the segments while immersed in the monomer for 3.5 h, rotating the container to ensure even UV irradiation. 6. Remove the segments to distilled water at 80°C and wash under constant stirring for 24 h to remove excess monomer and polymer. 7. Dry under vacuum at room temperature.

54

Jones and Hayes

3.4. PDMS Coating on Glass Microchips 1. Prepare a 1% (v/v) PDMS solution in dried and distilled dichloromethane. 2. Load 0.1 M NaOH solution into the inlet well of the microchip. Draw the NaOH solution into the channel using a vacuum. Let solution stand in the channel for 30 min. 3. Rinse repeatedly with purified water by loading water in the inlet well and drawing through the channel with vacuum. 4. Thoroughly dry the channel with dry N2 gas. 5. Rinse the channel with dichloromethane repeatedly using vacuum. 6. Apply the PDMS coating by loading the 1% PDMS solution into the channel and allowing it to stand in the channel for 10 min. 7. Remove the solution using a vacuum. 8. Flush the remaining excess PDMS using dry N2 gas. 9. Place the coated microchip in an oven, increasing the temperature 10°C/min until the temperature reaches 400°C. Allow the chip to remain in the oven at 400°C for 0.5 h. 10. Cool to room temperature in the oven (see Note 6).

3.5. Amine-Terminated PMMA 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Clean the PMMA surface by first soaking in 2-propanol for 15 min. Rinse thoroughly with 18 MΩ⋅cm water. Dry the PMMA pieces under dry N2 gas. Place PMMA pieces in a sealed vessel and purge for 20 min with dry N2 gas. Using a syringe, transfer N-LDAP onto the PMMA pieces without corrupting the N2 environment. Allow the N-LDAP to react with the PMMA for 10 min. Add enough water to completely cover the PMMA pieces to quench the reaction between the PMMA and the N-LDAP. Remove from the reaction flask and rinse the aminated surface thoroughly with purified water. Dry with N2 gas. Place freshly amine-terminated PMMA in an airtight vessel and purge with nitrogen for 20 min. For subsequent attachment of the octadecane: a. Dispense n-octadecane-1-isocynate onto the PMMA without corrupting the N2 environment. b. Allow the PMMA to remain exposed to the n-octadecane-1-isocynate for 10 min. c. Remove the PMMA pieces and quickly rinse with copious amounts of hexanes, toluene, and then acetone. d. Dry with N2 gas. e. Bond the top and bottom PMMA sections thermally (see Note 7).

3.6. Immobilization of Biotin Within a Phosphotidylcholine Bilayer 1. Thoroughly clean and dry PDMS channels and the glass or PDMS cover sheet. 2. Render all PDMS surfaces hydrophilic by exposing to oxygen plasma treatment for 15 s (PDC-32G plasma cleaner, Harrick Scientific, Ossining, NY) (see Note 8).

Surface Modification Methods

55

3. Prepare vesicles by dissolving the lipids in chloroform and combining them in the mole ration of 0.1% mol biotinylated lipids to DLPC. 4. Evaporate the chloroform in a stream of dry nitrogen, then vacuum evaporate for 4 h. 5. Reconstitute the dried lipids in pH 7.4 phosphate buffer, then sonicate for 5 min using a titanium tip. 6. Centrifuge the solution at 94,500g for 30 min. 7. Centrifuge the supernatant at 176,900g for an additional 3 h. 8. Using a syringe pump (Harvard PHD 2000) infuse the vesicles into the channels and allow to fuse for 30 min. 9. Flush excess vesicles from the channels using phosphate buffer (pH 7.4). 10. Before introduction of streptavidin conjugates, inject BSA solution and incubate for at least 1 h to ensure passivation of any defect sites in the bilayer coating. 11. Wash excess BSA from the channels using pH 9.8 buffer solution.

4. Notes 1. Filling and extraction of PDMS channels is best performed with vacuum, as positive pressure may overwhelm the adsorptive forces bonding the PDMS to a support substrate. For channels that are 50 µm or less, a disposable syringe can be used to create this pressure. Cut a small piece of tubing to fit the end of a syringe (with no needle attached). Heat the end of a metal bolt or nail that is slightly larger than the tubing. Work the end of the tubing onto the end of the bolt creating a flange at the tubing end that is slightly larger than the inlet or outlet well opening. To create the necessary pressure, seal the flanged end of the syringe and tubing over the outlet well and draw the fluid through the channel. The fluid can be removed from the channel using a glass capillary or the end of a filter paper. 2. Though this procedure does not call for a water rinse between PEM coatings, it may be necessary to perform a rinse at this juncture to improve uniformity of the coatings. 3. Atomic force microscopy is useful to verify the PEM coating. For an immediately tangible test, the wettability of the channel is also useful in determining if the coating has adsorbed properly. 4. Do not rinse the PDMS cover film with NaOH. 5. Before the cover film is applied, you can check the efficacy of the coating using contact angle measurements. 6. Addition of 0.10% Tween-20 to your separation buffer will provide well-defined peaks for protein samples. 7. Thermal bonding in this procedure was accomplished by heating each surface of the PMMA on a hotplate to 150°C for 5–10 min. Do not heat the entire PMMA part in a furnace, as outgassing and bubbling are possible. Align the two device pieces, preferably in an alignment holder, then place a 50 lb weight on the PMMA pieces and allow to cool to room temperature slowly in a programmable oven over 2 h. 8. Oxygen plasma treatment of the surfaces for 1 min should be sufficient. Once treated, the PDMS will bond more strongly adsorptively and can form covalent bonds with the glass substrate. Because of this, the PDMS channels cannot easily be removed after this step.

56

Jones and Hayes

References 1. Badal, M. Y., Wong, M., Chiem, N., Salimi-Moosavi, H., and Harrison, D. J. (2002) Protein separation and surfactant control of electroosmotic flow in poly(dimethylsiloxane)-coated capillaries and microchips. J. Chromatogr. A 947, 277–286. 2. Barker, S. L. R., Tarlov, M. J., Canavan, H., Hickman, J. J., and Locascio, L. E. (2000) Plastic microfluidic devices modified with polyelectrolyte multilayers. Anal. Chem. 72, 4899–4903. 3. Gottschlich, N., Jacobson, S. C., Culbertson, C. T., and Ramsey, J. M. (2001) Twodimensional electrochromatography/capillary electrophoresis on a microchip. Anal. Chem. 73, 2669–2674. 4. Henry, A. C., Tutt, T. J., Galloway, M., et al. (2000) Surface modification of poly(methyl methacrylate) used in the fabrication of microanalytical devices. Anal. Chem. 72, 5331–5337. 5. Hu, S. W., Ren, X. Q., Bachman, M., Sims, C. E., Li, G. P., and Allbritton, N. (2002) Surface modification of poly(dimethylsiloxane) microfluidic devices by ultraviolet polymer grafting. Anal. Chem. 74, 4117–4123. 6. Katayama, H., Ishihama, Y., and Asakawa, N. (1998) Stable cationic capillary coating with successive multiple ionic polymer layers for capillary electrophoresis. Anal. Chem. 70, 5272–5277. 7. Liu, Y., Fanguy, J. C., Bledsoe, J. M., and Henry, C. S. (2000) Dynamic coating using polyelectrolyte multilayers for chemical control of electroosmotic flow in capillary electrophoresis microchips. Anal. Chem. 72, 5939–5944. 8. Eteshola, E. and Leckband, D. (2001) Development and characterization of an ELISA assay in PDMS microfluidic channels. Sens. Actuators B Chem. 72, 129–133. 9. Mao, H. B., Yang, T. L., and Cremer, P. S. (2002) Design and characterization of immobilized enzymes in microfluidic systems. Anal. Chem. 74, 379–385.

6 Polyelectrolyte Coatings for Microchip Capillary Electrophoresis Yan Liu and Charles S. Henry Summary In chip-based electrophoretic analysis of biomolecules, chemical modification of the microchannel is widely employed to reduce or eliminate the analyte–wall interactions and alter electroosmotic flow (EOF) in the microchannel. A stable polyelectrolyte multilayer coating is one common way to regulate or eliminate EOF and prevent analyte adsorption for the rapid, efficient separation of biomolecules within microchannels. A wide variety of polyelectrolytes have been used as coatings. This chapter deals with how to coat microchips with polyelectrolytes and the expected results using polybrene and dextran sulfate as models. The technique presented here is generally applicable to any polyelectrolyte. Key Words: Polyelectrolyte; coating; microchip; capillary electrophoresis; electroosmotic flow; poly(dimethylsiloxane); polybrene; dextran sulfate.

1. Introduction Recently, microchip capillary electrophoresis (CE) has become a powerful tool to analyze biomolecules (1,2). Initially, work focused on microfabrication of glass microchips because of the mature micromachining technology (3,4). However, the time- and labor-consuming fabrication process of glass microchips and the requirement of clean-room facilities have led to the investigation of alternative substrate materials for the construction of CE microchips (5,6). Poly(dimethylsiloxane) (PDMS) attracted significant attentions as a microchip substrate because of its potential for mass production, rapid protyping, and good optical properties. However, peak tailing owing to the analyte absorption into PDMS microchip has been well documented for nonpolar hydrophobic species (7,8). In addition, one significant problem for PDMS devices is poorly defined electroosmotic flow (EOF) (9). Under CE conditions, the EOF dominates the flow velocity of both the run buffer and the analytes being separated. From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

57

58

Liu and Henry

In PDMS microchips, the nature of EOF is dependent on the process used for sealing chips, which generates different surface charge densities (9,10). In addition, the use of multiple substrate materials in a single device may cause inconsistency of the flow velocity and diminish separation efficiency owing to nonuniform flow in the capillary columns resulting from the different ζ-potentials of different substrate materials. Generally, these discrepancies in PDMS substrates are the result of minimal characterization of surface ionizable groups under typical CE conditions. Moreover, the EOF decreases with pH, making rapid separations of mixtures of anions and cations difficult at low pH. To overcome these problems, several surface modification techniques, including dynamic, covalent, and noncovalent coating, have been used to control the EOF (11–13). Dynamic coating is typically prepared by rinsing the capillary with a solution containing a coating agent that is either a polymer or a small molecular-mass inorganic ion. A small amount of coating agent is also added to the run buffer to keep the coating on the capillary wall surface. The life time of dynamic-coated capillaries can be extended by using an occasional, simple regeneration process. The resulting coating is used for both EOF suppression and prevention of protein adsorption. Covalent coating is the most prevalent, and perhaps the most effective strategy to control EOF and prevent biomolecule adsorption. Covalent coating involves the covalent immobilization of molecules onto the capillary wall. However, lengthy derivatization procedures, unstable coating layers beyond a limited pH range, and poor reproducibility limit the wide use of covalent coatings. To overcome the drawbacks associated with covalent derivatization, charged polyelectrolytes, such as polybrene and polyethyleneimine, have been used for adsorbed noncovalent capillary coatings. Owing to the strong electrostatic attraction between these polycations and the anionic silanols on the capillary inner surface, these polymers adsorb strongly onto the capillary inner wall. The capillary can be coated, regenerated, and then recoated, making them more cost and time effective than covalent coatings. 2. Materials 2.1. Equipment 1. 2. 3. 4.

Harrick plasma cleaner/sterilizer PDC-32G. Fisher FS20 sonicator. Voltameter. High voltage power supply (Standford PS350/5000V-25W, or Spellman CZE 1000R or similar). 5. Whatman 0.2 µm-syringe filter. 6. 1 kΩ resistor. 7. Two platinum wire electrodes (diameter: 0.5 mm).

Polyelectrolyte Coatings for Microchip CE

59

2.2. Chemicals and Reagents 1. 2. 3. 4. 5. 6.

Sylgard 184 silicone elastomer and curing agent. Polybrene. Dextran sulfate. Phosphate buffers: 20 mM pH 3.0–10.0. Sodium hydroxide. Methanol.

3. Methods 3.1. PDMS Device Fabrication A brief summary of the fabrication of PDMS microchips is provided here. A more detailed discussion of PDMS fabrication is presented in the appropriate chapter in this text (Chapter 3). A degassed mixture of sylgard 184 silicone elastomer and curing agent (10:1) was poured onto a 3-in. silicon mold that was patterned by lithographic technique and had been cleaned sequentially with deionized water (DI) and methanol and dried with a stream of nitrogen gas. After at least 2 h of curing at 65°C, the PDMS replica was peeled from the mold, resulting in a pattern of negative relief channels and reservoirs in the PDMS. Buffer reservoirs were then opened with a circular punch and the PDMS was trimmed to size with a razor blade. Bare PDMS replicas were formed by casting the PDMS mixture on a dry, clean nonpatterned silicon wafer. 3.2. Microchip Sealing Modifications of previously published reversible and irreversible sealing methods can be used to assemble the microchips (9,10). Reversible sealing involved thoroughly rinsing a PDMS replica and a glass plate (or a second piece of PDMS) with methanol and bringing the two surfaces into contact with one another prior to drying. The assembled microchip is then dried in an oven at 65°C for 10 min. The air bubbles between the two layers are driven out by pressure. This method of reversible sealing gives the most consistent sealing and does not require the use of clean room/hood facilities. Irreversible sealing is accomplished by first thoroughly rinsing a PDMS replica and a glass plate with methanol, and then drying them separately under a stream of nitrogen. The two pieces are then placed in an air plasma cleaner and oxidized at high power for 45 s. The substrates are brought into conformal contact immediately after removal from the plasma cleaner and an irreversible seal forms spontaneously. This seal is sufficiently strong that the two surfaces can not be separated without destroying the assembled microchip. 3.3. Noncovalent Coating Although channels can be coated with different polyelectrolyte layers, the whole coating procedure is similar to the one that was developed by Katayama

60

Liu and Henry

Fig. 1. Schematic for multilayer coating. (A) Preconditioned native poly(dimethylsiloxane) microchip, (B) first layer coating with a 5% polybrene water solution, and (C) second layer coating with 3% dextran sulfate water solution.

and coworkers for the conventional CE (14,15) and is applicable to microchips constructed from a variety of materials. Briefly, the channel is preconditioned for a few minutes then sequentially flushed with the desired polyelectrolyte solutions. A schematic of the whole procedure is shown in Fig. 1, with polybrene (PB) and dextran sulfate (DS) coating the channel as an example. Briefly, the separation channel is rinsed with 0.1 M NaOH for 4 min and flushed with DI for 4 min, respectively. Once preconditioned, the channel is sequentially filled with 5% PB solution and 3% DS solution (both in water) for 2 min each with a 15-min waiting period after each rinse. This procedure of successive coating results in a bilayer of PB/DS on the channel walls. The polymer rinsing steps can be repeated multiple times to build up additional layers. Finally the channel is flushed with the run buffer solution.

Polyelectrolyte Coatings for Microchip CE

61

3.4. EOF Measurements It is important to measure the EOF to ensure that an appropriate coating has been generated. For more detailed information on measuring EOF consult the appropriate chapter (see Chapter 13). Briefly, the running electrolyte for electrophoresis experiments is 20 mM pH 3.0–10.0 phosphate buffer. The pH was established by titrating a solution of either o-phosphoric acid or sodium dihydrogen phosphate with sodium hydroxide. All buffers are prepared in DI, passed through a 0.20-µm pore size syringe filter, and degassed for 5 min in a sonicator before use. A modification of a previously published current monitoring method is used to determine the EOF (16). Both reservoirs are filled with dilute buffer (2:1 buffer: water), and the channels are subsequently conditioned under an electric field of 1200 V for 15 min. The increased dilution factor as compared with standard protocol (19:1 buffer: water) is used to ease end point detection. No statistical differences in the absolute values are noted between the two protocols. A 1 kΩ resistor is placed in line between the waste reservoir and electrical ground to follow the separation current. A voltameter records the potential changes across the resistor, which correlates to the current through Ohm’s Law. The sample reservoir is then filled with concentrated buffer, and the potential is reapplied. The time required for the current plateau is measured for each run and is indicative of the concentrated buffer’s filling the separation channel. The sample reservoir is then filled with dilute buffer and the above procedure repeats. The time required for the current to reach this plateau was used as the migration rate of a neutral marker, and the EOF is determined by µEOF = L2 / Vt

(1)

where L is the length of the separation channel, V is the total applied voltage, and t is the time in s required to reach the new current plateau. The typical voltage profile across the resistor is shown in Fig. 2, and the time to reach a current plateau is 53 s for this microchip. This is a modification of the traditional mobility equation that takes into account that the total and effective capillary lengths are identical. 3.5. Stability and Reproducibility of the Coating Layer One concern with the noncovalent coating is the stability of the coating. A pH of 3.0 is chosen for evaluation of the PB-coating lifetime because no EOF is detected at this pH point for native PDMS microchips. As the coating becomes detached from the channel wall, the EOF will approach zero until the coating becomes completely detached, at which point the EOF direction will be reversed and no current change will be detected. Another concern with the noncovalent coating is the reproducibility of the coated layer from chip-to-chip. The EOF is detected in six PB/DS-coated

62

Liu and Henry

Fig. 2. Typical voltage profile for electroosmotic flow measurement. Table 1 Comparison of Electroosmotic Flow for Uncoated, PB/DS-Coated, and PB-Coated PDMS/Glass Microchipa pH Uncoated chip PB/DS-coated chip PB-coated chip aUnit:

3.0 0 2.47 –4.29

4.0 2.73 2.93 –3.34

5.0 3.06 3.29 –2.40

6.0 3.01 3.63 –2.95

7.0 4.04 3.65 –2.21

8.0 3.89 3.63 –1.93

9.0 4.43 3.63 –1.97

10.0 4.89 3.69 –1.95

× 10–4 cm2⋅V–1⋅s–1.

PDMS/glass chips at three different pH values in different days. The relative standard deviation (RSD) of the EOF for six PB/DS-coated PDMS/glass chips is considered as an indicative parameter of the complete and effective covering of the channel walls by the PB and DS layers. 4. Notes 1. Although the EOF differs significantly between uncoated oxidized and native PDMS/glass microchips, the PB/DS coating layer can compensate for the differences in type and density of anionic groups on the surface, as well as the difference between chip-sealing techniques, and generates a constant EOF regardless of substrate and sealing technique. The coating is solely responsible for the generation and control of EOF in the microchannels. The EOF results from native-, PB-, and PB/DS-coated microchips are shown in Table 1. 2. If electrochemical detection is involved in the microchip CE, one concern is the electrode fouling resulting from either analyte or polymer adsorption on the work-

Polyelectrolyte Coatings for Microchip CE

63

ing electrode surface. This is of particular interest when using a coated microchip in which any detached coating could potentially adsorb onto the electrode surface, resulting in the decrease of detection signal. Separations of dopamine and hydroquinone in both PB/DS-coated and uncoated microchips can give the demonstration whether the detached coating adsorb on the working electrode surface. 3. The noncovalent-coating technique is applicable to numerous polyelectrolytes, among which polybrene is the most popular and is sold commercially as a capillary treatment. In addition, polyamine, poly(dimethyl diallyl ammonium chloride), polyethyleneimine, and polyarginine are widely used to prevent biomolecule adsorption on the capillary surface.

References 1. Barta, C., Ronai, Z., Nemoda, Z., et al. (2001) Analysis of dopamine D4 receptor gene polymorphism using microchip electrophoresis. J. Chromatogr. A 924, 285–290. 2. Fanguy, J. C. and Henry, C. S. (2002) Pulsed amperometric detection of carbohydrates on an electrophoretic microchip. Analyst 127, 1021–1023. 3. Culbertson, C. T., Jacobson, S. C., and Ramsey, J. M. (2000) Microchip devices for high-efficiency separations. Anal. Chem. 72, 5814–5819. 4. Kopp, M. U., Mello, A. J., and Manz, A. (1998) Chemical amplification: continuousflow PCR on a chip. Science 280, 1046–1048. 5. Martin, R. S., Gawron, A. J., and Lunte, S. M. (2000) Dual-electrode electrochemical detection for poly(dimethylsiloxane)-fabricated capillary electrophoresis microchips. Anal. Chem. 72, 3196–3202. 6. McClain, M. A., Culbertson, C. T., Jacobson, S. C., and Ramsey, J. M. (2001) Flow cytometry of Escherichia coli on microfluidic devices. Anal. Chem. 73, 5334–5338. 7. McDonald, J. C., Duffy, D. C., Anderson, J. R., et al. (2000) Fabrication of a configurable, single-use microfluidic device. Electrophoresis 21, 27–40. 8. Effenhauser, C. S., Bruin, G. J., and Paulus, A. (1987) Integrated chip-based capillary electrophoresis. Electrophoresis 18, 2203–2213. 9. Ocvirk, G., Munroe, M., Tang, T., Oleschuk, R., Westra, K., and Harrison, D. J. (2000) Electrokinetic control of fluid flow in native poly(dimethylsiloxane) capillary electrophoresis devices. Electrophoresis 21, 107–115. 10. Duffy, D. C., McDonald, J. C., Schueller, O. J. A., and Whitesides, G. M. (1998) Rapid prototyping of microfluidic systems in poly(dimethylsiloxane). Anal. Chem. 70, 4874–4884. 11. Giordano, B. C., Copeland, E. R., and Landers, J. P. (2001) Towards dynamic coating of glass microchip chambers for amplifying DNA via the polymerase chain reaction. Electrophoresis 22, 334–340. 12. Badal, M. Y., Wong, M., Chiem, N., Salimi-Moosavi, H., and Harrison, D. J. (2002) Protein separation and surfactant control of electroosmotic flow in poly(dimethylsiloxane)-coated capillaries and microchips. J. Chromatogr. A 947, 277–286. 13. Horvath, J. and Dolnik, V. (2001) Polymer wall coatings for capillary electrophoresis. Electrophoresis 22, 644–655.

64

Liu and Henry

14. Katayama, H., Ishihama, Y., and Asakawa, N. (1998) Stable cationic capillary coating with successive multiple ionic polymer layers for capillary electrophoresis. Anal. Chem. 70, 5272–5277. 15. Katayama, H., Ishihama, Y., and Asakawa, N. (1998) Stable capillary coating with successive multiple ionic polymer layers. Anal. Chem. 70, 2254–2260. 16. Huang, X., Gordon, M. J., and Zare, R. N. (1988) Current-monitoring method for measuring the electroosmotic flow rate in capillary zone electrophoresis. Anal. Chem. 60, 1837, 1838.

III DETECTION METHODS FOR MICROCHIP CAPILLARY ELECTROPHORESIS

7 Interfacing Microchip Capillary Electrophoresis With Electrospray Ionization Mass Spectrometry Trust Razunguzwa and Aaron T. Timperman Summary Microfluidic devices are a unique enabling technology for chemical separations, modification, and synthesis that are ideally suited for the manipulation of low volume samples on the order of a few nanoliters in volume. Complex patterns of capillary-sized channels with zero dead volume connections are the distinguishing features of many microfluidic devices. Concurrently, mass spectrometry has undergone further development, and is now arguably the method of choice for structural characterization of mass- and volume-limited samples. The production of ions in the gas phase from the solution phase is critical for direct coupling of fluidic devices with the mass spectrometer, and the electrospray ionization (ESI) sources are well suited for this application. Micro- and nanoflow ESI interfaces are ideal for these applications as they cover flow rate ranges from the hundreds to a few nanoliters per minute, which are the same as the flow rates used by most microfluidic devices. Herein, the assembly and operation of a simple ESI interface for coupling a microfluidic device and mass spectrometer is described. Key Words: Capillary electrophoresis; microfluidic device; mass spectrometry; electrospray; interface.

1. Introduction Microfluidic devices have a promising future for chemical analysis and synthesis at low volume and mass, while being amenable to massive parallelism for increasing sample throughput. Functionalities, which have been developed on the chip, include sample preparation, separation of complex mixtures, preconcentration of analytes, and tryptic protein digestions (1–4). The ability of microfluidic devices to handle and manipulate extremely small volumes of solutions, in the nanoliter regime, make them especially attractive for biological samples whose amounts are frequently limited. Fluidic networks are characterized by intersections with zero dead volume enabling efficient processing of complex samples. In general, microchips for chemical analysis allow parallel From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

67

68

Razunguzwa and Timperman

sample analysis, shorter analysis times, increased separation efficiencies, lower detection limits, and reduced reagent consumption and waste generation (2,5). The mass spectrometer is a powerful analytical tool owing to its high resolving power, ion isolation capability, and ability to measure the mass to charge ratio of both parent and fragment ions. Electrospray ionization mass spectrometry (ESI-MS) is concentration (not mass) sensitive. Therefore, miniaturizing the sample introduction system achieves the highest sample concentrations and the lowest detection limits. Microfluidic devices interfaced with ESI-MS provide a convenient, miniaturized sample preparation/introduction system. Solutions are commonly transported through microfluidic channels by application of an electric potential (electrokinetically) or by applying a pressure (hydrodynamically). In addition, flow control on microfluidic devices can also be governed electrokinetically, although the development of reliable mechanical valves is desirable. Most systems interfaced with mass spectrometers are electrically driven systems, and these will be our primary focus (6–8). Early attempts to interface microchips to MS involved spraying fluid directly from an exposed channel on the microchip (9,10). Although this design was attractive, in that the design did not require complex machining, it resulted in large dead volumes owing to the formation of large Taylor cones from the solution exiting out of the open end of the microchannel, leading to sample dilution (lower sensitivities) and band broadening. Also, because the surface at the edge of the microchip was flat, these devices required an impractically high voltage to overcome the liquid surface tension and initiate electrospray. Other researchers attempted using hydrophobic coatings (to minimize surface tension) at the edge of the microchip and on-chip nebulizers, but this has been met with limited success (11). More recently, spraying from a transfer capillary spray tip attached to a microchip has been found to be an effective way of sample delivery into a mass spectrometer, as it provides a spray tip from which electrospray can be more easily generated (12–14). For these devices, application of high voltage can be through a platinum electrode inserted in solution, liquid junction, or by a conductive coating at the outlet of the capillary. Most recent advancements are utilizing microfabrication techniques to integrate nanospray tips directly onto the microfluidic device, which eliminates band broadening associated with the dead volume at the chip–capillary interface and the extra column volume associated with capillary spray tip-coupled microfluidic devices. An example is a device developed by the Henion group and Advion Biosciences that is based on a silicon substrate with etched nozzles in the planar surface of the silicon wafer, used directly as microspray emitters perpendicular to the chip (15). Smith’s group also recently reported a polycarbonate-based device with an ESI tip constructed on polycarbonate plates by laser micromachining for isoelectric focusing-ESI-MS (16).

Interfacing Microchip Capillary Electrophoresis With ESI-MS

69

Considering the different ways in which a microchip–MS interface can be configured, it is of paramount importance to identify the parameters that affect electrospray generation at the outlet of the microchannel or capillary to obtain a functional interface. The first requirement for a chip–MS interface is maintaining bulk flow of solution to the spray tip, which is needed to sustain droplet formation in the electrospray source (10). For an electrokinetically driven system, this means that electroosmotic flow (EOF) must be present in the microchannel, requiring a high density of charged sites on the channel surface. A basic pH is usually sufficient to meet this requirement given that many microfluidic devices are fabricated on glass with silanol groups that deprotonate as pH increases. Stable spray is dependent on the diameter of the tip and the spray capillary inner diameter (I.D), which generate more resistance to flow as they get smaller. At the low flow rates afforded by microfluidic devices, small spray tip sizes and spray capillary I.Ds, small droplets with high surface-to-volume ratio form at the tip, giving rise to better ionization and desolvation efficiencies because of rapid droplet solvent evaporation. Mann and Wilm estimated the droplet size formed at a 1- to 2-µm tip to be less than 200 nm in diameter, corresponding to an average concentration of one analyte molecule per droplet for a 1-pmol/µL solution (17). By separating molecules into different droplets, cluster formation is minimized and improves analysis of sample with high-salt concentrations. McLafferty et al. demonstrated this by further reducing the flow rate of 25 nL/min used by Mann and Wilm to 1 nL/min using a 5-µm I.D, 2-µm spray tip, and achieved attomole sensitivity detection of large biomolecules (18). It is, therefore, beneficial to use spray capillaries with the smallest capillary I.Ds and spray tip sizes to achieve efficient electrospray and consequently high sensitivities. Other factors need to be considered for the operation of a microchip–MS interface. The lowest detection limits in ESI-MS are usually realized when the ionization source is operated in the positive ion mode (i.e., a positive voltage is applied at the spray tip). This requirement dictates a low pH spray solution. However, in order to support sufficient EOF under acidic conditions, the native glass surface of the microchannel must be coated with a material that provides a high surface charge at acidic pH. A make-up or sheath flow solution may be necessary in some cases to modify the electrophoresis buffer to achieve the necessary conditions (pH, concentration, or flow) for electrospray. Finally, the chip substrate (usually glass) must be compatible with organic solvents and acids while producing minimal chemical background. Solvents must be of highest obtainable purity and volatile without ion suppressing agents such as triflouroacetic acid (TFA). The assembly of the chip capillary electrophoresis (CE)–ESI-MS interface starts with fabrication of the microfluidic device using standard photolithography and wet chemical etching procedures to introduce predesigned

70

Razunguzwa and Timperman

microchannels onto the glass surfaces (19). Compared to the use of polymers, glass is preferable because of its excellent optical properties and its stability toward organic solvents. Bonding to enclose the channels can be achieved with direct bonding at low and high temperature without adhesives. This is particularly important for chip ESI-MS detection because fillers, unreacted monomer, and plastic can increase the background signal. This process is followed by surface coating of the channels with a positively charged coating to increase EOF and minimize analyte adsorption. Reservoirs are attached to the chip, and the chip is conditioned with buffer followed by spray tip attachment. Finally, the microchip is brought in front of the MS for sample introduction. 2. Materials 2.1. Equipment 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Sawed/scribed wafer cleaner (Ultra and Equipment Co., Fremont, CA). Airclean 600 workstation (Airclean Systems, Raleigh, NC). Blue tape and mounting rings for holding glass plates in wafer washer. DC power supply (Agilent, Palo Alto, CA). High voltage power supply (Matsusada Precision Inc., San Francisco, CA). Stereomicroscope GZ6 (Leica Microsystems, Bannockburn, IL). Multimeter (Fluke, Everett, WA). Electrospray ionization mass spectrometer (Thermo Finnigan, San Jose, CA). Soda lime glass plates with chromium layer and photoresist (Telic Co., Santa Monica, CA). Ultraviolet (UV) lamp and aligner (Advanced Radiation Corporation, Santa Clara, CA). Stylus instrument (Tencor Instruments, Milpitas, CA). Fused silica capillary (Polymicro Technologies, Phoenix, AZ). High precision drill press (Tralmike’s-Tool-A-Rama, Plainfield, NJ). Carbide microdrills (Kyocera Tycom, Owego, NY). Nanoport assemblies (Upchurch Scientific, Oak Harbor, WA). Three-dimensional translational stage (Newport). Laser puller (Sutter Instruments, Phoenix, AZ). Syringe pump (Harvard Apparatus, Holliston, MA).

2.2. Reagents 1. Glass etching solution: 50% hydrofluoric acid (HF)/60% nitric acid/water, (2/1/7 v/v/v). 2. Piranha solution: concentrated sulphuric acid/30% hydrogen peroxide, (3/1 v/v). 3. Ammonium hydroxide reagent: concentrated ammonium hydroxide/30% hydrogen peroxide/water, (1/1/5 v/v/v). 4. Acetone. 5. 10% Tetramethylammonium hydroxide developer solution (Clariant).

Interfacing Microchip Capillary Electrophoresis With ESI-MS

71

Fig. 1. An outline of the photolithography, wet chemical etching, and bonding processes of glass microchips. 6. Soap solution: one scoop of powdered soap in 1 L of water. 7. Make-up solution and separation buffer: 0.1% acetic acid and 5% acetonitrile in water. 8. N-trimethoxypropyl-N,N,N-trimethyl ammonium chloride (TMAC) (Gellest). TMAC is toxic and has a very strong odor.

3. Methods 3.1. Chip Fabrication Fabrication of glass microchips is carried out using standard photolithography and wet chemical etching (20) on a 10 × 5 cm soda-lime glass (see Note 1), followed by drilling of access holes on the glass substrate, and finally, lowtemperature and high-temperature bonding of the etched glass plate to a cover plate (21). An outline of the standard fabrication process from photolithography to bonding is shown in Fig. 1. Glass plates coated with chromium and photoresist layers are commercially available. 3.1.1. Photolithography 1. Channel patterns are designed and drawn using Macromedia Freehand 10. Negative transparencies of these patterns are printed on an Afga Accuset 1000 printer with a resolution of 3000 dpi and used as photomasks in the photolithography process.

72

Razunguzwa and Timperman

2. A soda-lime glass chip with photoresist is placed on the aligner and held in place by vacuum. The photomask with the channel pattern is aligned and placed over the glass chip. 3. The chip and mask are positioned under the UV lamp and exposed for 15–40 s at 365 nm depending on the smallest channel size on the mask (see Note 2). 4. After exposure to UV light, the photoresist is developed. The photoresist polymer that was exposed to UV light is washed away by the developer solution AZ 312 MIF/H2O (1:1), leaving the pattern of the photomask on the photoresist layer. 5. The channel pattern on the photoresist layer is inspected using a microscope to check for poorly developed channels.

3.1.2. Wet Chemical Etching 1. A chromium etchant solution is used to remove the chromium within the channel pattern leaving an exposed glass layer. The chips are placed in the chromium etchant solution for 2 min. The glass chips are then rinsed with deionized (DI) water and blown dry with nitrogen. 2. The photoresist layer is stripped using acetone and the channels are inspected for completeness of the chromium etching using the microscope. 3. The glass substrates are etched with 50% HF/60% HNO3/H20, (2/1/7 v/v/v), rinsed with DI water, and blown dry with acetone. Caution: HF should be handled with extreme care as it is very hazardous. Exposure to vapor and skin contact is extremely dangerous and, therefore, the accepted methods for handling and use should be followed. The chromium layer is removed using the chromium etchant solution after glass etching and DI water rinsing of the glass plates (heat to 50°C if necessary) (see Note 3). 4. The chip is rinsed with DI water and blown dry with nitrogen gas soon after the channels are inspected for completeness using a microscope. Channel profiles and dimensions are then obtained using the stylus instrument (see Note 4).

3.1.3. Glass Bonding The glass bonding procedure is outlined next: 1. Clean the glass plates (etched plate and cover plate) with acetone using a lint free swab. 2. Clean the glass plates with detergent solution (one scoop of powdered soap in 1 L of DI water) and rinse with DI water. 3. Place the glass plates in piranha solution at 100°C for 45 min to 1 h and rinse with DI water. 4. Place the glass plates in ammonium hydroxide reagent at 80°C for 45 min to 1 h and rinse with DI water. 5. Wipe down the Airclean workstation with a lint-free cloth containing a few squirts of isopropanol. Clean again with mitten if necessary. Be sure to wipe down all the parts to the washer. Wipe down mounting rings and place blue tape onto two mounting rings.

Interfacing Microchip Capillary Electrophoresis With ESI-MS

73

Fig. 2. (A) Illustration of the vertical access hole drilled at the end of a channel on the microchip. (B) Shows a minimal dead volume connection of a fused silica capillary spray tip to the microchip. The 360- and 650-µm holes are drilled using carbide microdrills. A silica seal-tight sleeve is inserted into the 650-µm hole and holds a 360-µm O.D. capillary spray-tip, which extends into the 360-µm hole. There is a tight seal between the sleeve and the 650-µm hole, and between the 360-µm spray-tip and the 360-µm hole (see Note 8). 6. Load the glass plates on the blue tape and place them in the high-pressure washer and perform one washing and one drying cycle. 7. Align the glass plates using a spacer ring in between the mounting rings and bond them using a rubber roller to press them together. Eliminate all Newton rings by squeezing them out from in between the glass plates. Be sure to bond the etched side to the plain glass plate (see Note 5). 8. Permanent bonding is achieved by placing the glass plates in the furnace and applying the temperature program: 25–100°C at 40°C/min, 100°C for 15 min, 100–600°C at 10°C/min, 600°C for 15 min. The furnace was allowed to cool naturally to ambient temperature.

3.1.4. Access Hole Drilling The access holes to the channels of the microchips are drilled directly on the etched glass plate before bonding using a high-precision drilling press and carbide microdrills ranging from 200 to 650 µm. As shown in Fig. 2A, the etched glass plate is first drilled half way (0.5 mm) through the glass from the open channel side using a 200-µm microdrill, and then from the other side using a 360-µm microdrill until it meets the 200-µm hole drilled from the other side of the glass (see Note 6). The bottom of the 360-µm hole is flattened using a flat endmill microdrill. This drilling method of the access holes allows for the tight fit connection of a 360-µm outer diameter (O.D) fused silica capillary to the

74

Razunguzwa and Timperman

microchip with its end lying flat at the bottom of the 360-µm hole and the capillary bore emptying into the 200-µm hole. The drilling process is carried out with the aid of a microscope to align the microdrill to the end of a channel as well, as aligning the microdrill to the hole drilled from one side halfway through the glass, when drilling from the other side. A microscope is also used to follow the progress of the drilling process. The connection of the spray tip to the microchip is achieved by drilling microholes through the junction between the cover plate and the etched glass plate, at the edge of the bonded microchip. A 360-µm hole is first drilled 8 mm into the glass chip through a channel running off the edge of the microchip. A 650-µm hole is then drilled through the 360-µm hole 5 mm into the glass plate before an end mill microdrill is used to flatten the bottom (see Note 7). Figure 2B shows the schematic cross-section of the end result of this drilling process. 3.1.5. Microchip Fittings Polyetheretherketone (PEEK™) polymer Upchurch nanoport assemblies are attached to the microchips via preformed adhesive rings to act as reservoirs leading to the access holes and connecting fused silica capillaries to the microchip. Two types of nanoport assemblies are used to attach a fused silica capillary column directly to the microchip and these are the 6.32 flat-bottom and 6.32-coned assemblies (Upchurch, F.123S and F.124S, respectively). The Upchurch 1/4.28 FB nanoports are used as reservoirs. 3.2. Microchip Coating A positively charged quaternary amine reagent that provides a positive permanent coating is used to modify the inner surface of the microchip and the fused silica capillary spray tip to give a positively charged surface (see Note 9). The coating generates a strong anodic EOF at low pH, in addition to minimizing adsorption of the positively charged sample components to the channel walls by electrostatic repulsions between the positively charged channel walls and positively charged ions of the sample. The coating procedure is outlined next: 1. 2. 3. 4. 5. 6.

Rinse all the channels with 1 M NaOH for 5 min. Flush DI water through the channels for 3 min. Flush all channels with 10% TMAC and fill all the channels with 10% TMAC. Fill the reservoirs with pure methanol. Allow the chip to stand for 2 h while replenishing evaporated methanol. Wash out unreacted TMAC and the solvent methanol from the channels using DI water followed by the separation buffer.

The coating procedure should be carried out in a fume hood because TMAC is toxic and has a strong offensive odor.

Interfacing Microchip Capillary Electrophoresis With ESI-MS

75

3.3. Making Spray Tips A fused silica capillary emitter (6 cm, 50 µm I.D., and 360 µm O.D) is attached to the microchip as previously described in Subheading 3.1.4. The spray tips can be made in house with a capillary puller or purchased directly from New Objective (Woburn, MA). Using a laser puller, the emitter exit end is tapered to approx 5 µm I.D. A 15-cm fused silica capillary column is cut and the polyimide coating burnt off the center of the capillary. The section without the polyimide coating is wiped clean using methanol and a Kimwipe. The capillary is placed in the laser puller instrument with the exposed silica portion in the path of the laser. Once the capillary is aligned, the laser is activated and the tip pulled. Two spray emitters are obtained and are cut to the desired length (see Note 10). The laser puller instrument is programmed according to the tip size required and the capillary size. It has been demonstrated, in literature, that emitters with small tip sizes afford better sensitivities than larger ones because they form smaller droplets with high charge densities, which result in better desolvation and ionization efficiencies (15). The spray tips can be etched with 50% HF solution for 30–60 s to further reduce the tip size and thickness of tip wall. 3.4. Interfacing of Microchip to ESI-MS Once all the reservoirs and fittings have been attached to the microchip, the channels are filled with the separation buffer solution by applying a vacuum sequentially to all the reservoirs. It is imperative that no bubbles are introduced into the channels because these can introduce void volumes and current breakdown problems. All the buffers have to be degassed beforehand by sonication or vacuum degassing. The channel resistances are then determined indirectly by measuring the voltage drop across a 10-kΩ resistor to obtain an ohm plot. A high voltage power supply is used to apply voltages to the buffer solutions in the reservoirs. Caution: exercise caution when applying dangerous high voltages to avoid electrocution owing to the large charging capacitance of the power supplies. Good electrical connectivity is tested for all the channels on the microchip by establishing whether a current flows through the electrolyte in the channels when the high voltage is applied. The microchip is mounted on a three-dimensional translational stage and brought in front of the mass spectrometer where the spray tip is aligned to inlet of the MS orifice using the x-y-z manipulator. A distance of approx 3–6 mm is left between the spray tip and the MS inlet orifice. 3.4.1. Microchip ESI-MS The design of the microchip used to interface to ESI-MS is shown in Fig. 3. Reservoirs A, B, and C correspond to the separation buffer, sample, and

76

Razunguzwa and Timperman

Fig. 3. Schematic representation of the microchip-electrospray ionization mass spectrometry assembly. The microchannels have a positive charge and the device is operated at low pH. Ports A–D are the buffer, sample, waste, and make-up solution reservoirs, respectively. For an electrospray compatible separation buffer, side channel D can be excluded from the design, because in this case no make-up solution is required. Loading of a sample plug is performed by applying a potential difference between buffer reservoirs B and C, with the potential at B being more negative relative to the potential at C. The sample plug is then injected and separated by applying a negative voltage to the separation buffer reservoir A with reservoirs B and C floating. The sample migrates down the separation channel until it reaches the spray tip for the onset of electrospray.

waste reservoirs, respectively. Reservoir D is used to connect to a syringe pump for delivery of the make-up flow solution in case where it is necessary. A sample is dried down and redissolved in the separation buffer followed by the replacement of the buffer solution in reservoir B with the sample solution. The operation of the microchip–ESI/MS interface is outlined as follows: 1. The sample is loaded between the tees by applying a negative voltage to the sample reservoir, ground to the waste reservoir, and leaving reservoir (A) floating (see Note 11). The sample is then injected into the separation channel by applying a negative voltage to the separation buffer reservoir (A) with reservoirs B and C floating and the mass spectrometer at ground (see Note 12). The sample plug migrates down the separation channel with the concomitant separation of the different sample species based on their charge, size, and shape. The channels are coated with a permanent positive coating that generates a strong EOF toward the anode (toward the mass spectrometer) at low pH. The EOF in the microchannel results from the motion of mobile counter-ions in the diffuse region of the electric double layer, in response to the applied electric field. Because the mobile counter-ions in this case are negative ions (for a positively charged wall), the negative ions reach the spray tip first, followed by the neutrals,

Interfacing Microchip Capillary Electrophoresis With ESI-MS

77

and then the positive ions. The separation of the positive ions is therefore better because they have more time to separate, and these are the ions detected by positive ion mode MS. EOF is usually preferred over pressure-driven flow for onchip separations because of the plug-like flow profile of EOF, which results in better separation efficiencies. However, EOF is not very reproducible because of several factors that can induce velocity gradients in the bulk liquid and these should be minimized. These include partial blockage or narrowing anywhere in the channel, a nonuniform wall surface charge distribution, slight height differences between free surfaces in the reservoirs, capillary forces resulting in curvatures of the free surfaces in the reservoirs, and contact angle hysteresis of a bubble entrapped in the channel. 2. For an electrophoresis buffer compatible with electrospray, the sample bands are delivered to the spray tip by EOF where the electrospray process occurs. Small and charged buffer droplets form at the spray tip and the electric field at the tip generates an electrostatic force that is sufficient to pull the droplets out of the tip toward the ground plate of the mass spectrometer. This phenomenon occurs when the electrostatic force has become equal to the surface tension of the liquid and at that point the droplet changes shape to an elliptically shaped Taylor cone that is drawn out toward the mass spectrometer orifice along the axis of the fluid flow. The gap distance between the spray tip and MS orifice, as well as the alignment of the spray tip is therefore critical, and so the gap should be kept between 3 and 6 mm. The initial fine droplets sprayed from the tip shrink by solvent evaporation, and ions, which are involatile, are retained in the shrinking droplet. The increase in the repulsive forces between the excess charges in the droplet eventually overcome cohesive forces in the droplet to cause disintegration into smaller droplets or columbic explosion (22). A small orifice in the counter electrode then allows some of the ions from solution to enter the vacuum chamber of the mass spectrometer for mass analysis. 3. In cases where the electrophoresis buffer is incompatible with electrospray, a sheath liquid or make-up solution can be introduced through a side arm microchannel as shown in Fig. 3 (see Note 13). The sample meets the make-up solution flow from a syringe pump at the intersection of the channel from D and the electrophoresis main channel. A 75-µm I.D and 360-µm O.D fused silica capillary column is connected at one end to reservoir D on the microchip through one of the Upchurch nanoport assemblies, and the other end to the mass spectrometer syringe pump. A 250-µL syringe is filled with the make-up or sheath solution, and delivered on to the chip at a rate lower than the EOF stream, which is typically less than 100 nL/min. A flow splitter is connected between the microchip and the syringe pump to achieve the required low flow rates. If the flow rate of the make-up solution becomes too high, the solution can migrate backwards into the separation channel toward the separation buffer reservoir and the separated sample components never reach the spray tip. The flow rate should be kept between 50 and 100 µL/min for electrospray conditions described in this chapter. If the separation buffer is ESI compatible, then the use

78

Razunguzwa and Timperman

of this make-up solution can be eliminated (see Note 14). The make-up solution is used to modify the electrophoresis buffer to meet the requirements of electrospray conditions. In cases where the electrophoresis buffer is too concentrated (which is usually the case because electrophoresis buffer needs to be highly conductive for significant EOF), the make-up solution is used to dilute the buffer for the ESI-MS, which requires a lower concentration of buffer to minimize background noise. 4. The EOF, together with the hydrodynamic flow stream of the make-up solution, deliver the sample to the spray tip of the fused silica capillary emitter where electrospray occurs. The CE voltage applied at reservoir A can be decoupled from the electrospray process by applying the ESI voltage at reservoir D (see Note 15). 5. Separated sample components are detected using positive ion mode MS because separation is at low pH and the sample ions are predominantly positive ions.

4. Notes 1. In photolithography, alternative borosilicate glass types such as D263, Corning 0211, or Schott Borofloat can be used. 2. For smaller channel features, more UV exposure time is required. Thus, UV irradiation time is dependant upon the time needed to create preferable dimensions of the smallest width of the channels. 3. Etching time varies with types of glass and the desired etching depth. Soda-lime glass etches faster than Corning 0211, D263, and the Schott Borofloat. 4. It should be realized that the anisotropic etching of the glass substrates by HF normally produces trapezoidal-shaped channels. The stylus instrument is used to measure the width at half the height and depth of the channels. 5. In the event that the bonding process yields interfering Newton rings or a weak bond, place the chip in water for several hours, depending on the strength of the bond that would have been formed, and insert a wedge between the glass plates to separate the two. Perform the washing and drying cycle again in the wafer washer, after which bonding steps 6–7 (Subheading 3.1.3.) are repeated. If there are a few Newton rings that do not cross the channel network, proceed on to the hightemperature bonding of the microchip. 6. When drilling the vertical access holes do not use any liquid to collect glass chippings or aid the drilling process, instead simply fan accumulating particulates away. 7. Water is filled in all the channels before the drilling of horizontal dead volume connection because the movement of water denotes the success of a connection and the water prevents the glass particles from the drilling process from getting lodged in the channel during the course of the drilling process. 8. Ensure that the capillary tip end is flattened by the use of a sandpaper or emery cloth to avoid use of jagged edge capillaries. View the tip with a microscope to verify the flatness.

Interfacing Microchip Capillary Electrophoresis With ESI-MS

79

9. A variety of coatings can be used to modify the surface of the microchip to enhance or suppress EOF and to prevent protein adsorption (23). Examples are ([acryloylamino]propyl)trimethylammonium chloride (BCQ) and (3-aminopropyl) silane, which have been used to impart a positive charge on the surface of the microchip channels for EOF enhancement through the microchannel at low pH (24,25). Linear polyacrylamide and polyvinyl alcohol (PVA) have also been used for EOF suppression and minimization of sample adsorption to channel surface for protein and peptide separations (26). 10. The length of the spray capillary should be kept at a minimum because of the electrokinetically induced pressures that develop at the microchip–capillary junction. The pressure term introduced by the capillary reduces the EOF through the tip (27). 11. Sample leakage into the separation channel can be avoided by applying small biasing voltages to reservoirs B and C. 12. The injection plug length can be varied by increasing the distance between the side channels of the double-tee injector in the fabrication process. The pinched injection strategy can also be used for injection using a cross or single-tee instead of a double-tee (28). Continuous infusion of the sample into the MS can be achieved by applying a negative voltage to the sample reservoir (B) and leaving A and C floating. 13. Although a number of researchers have made use of a side arm microchannel to introduce make-up solution (12,13,29), other groups have used on-chip porous microjunctions (27). Make-up solutions can also be introduced through external liquid junctions at the outlet of the CE microchannel. 14. There are a few reports in literature of direct coupling of microchip CE and ESIMS with direct generation of electrospray from EOF through the electrophoresis channel, without the use of a make-up solution. An example is a report by Gobry et al., where a polymer-based microchip was directly coupled to the MS for detection of small ions, but the performance of the interface was hampered by the large dead volume from the droplet that formed at the outlet of the microchannel (30). Lazar et al. also demonstrated electrospray from a glass microchip that made use of extremely low EOF rates (20–30 nL/min) to achieve subattomole sensitivity for proteins and peptides. The device consisted of a nanospray tip perpendicularly attached to the microchip surface (31). However, most interfaces in literature, even those that use an attached capillary, normally make use of make-up solution to modify the electrophoresis buffer to achieve the best possible conditions for stable electrospray. 15. The separation voltage can be decoupled from the high voltage used for electrospray. Other than the method described in this chapter, the ESI voltage can be applied and controlled at the spray tip using either a direct electrical connection through a gold-coated capillary used by Harrison and co-workers (12,13,29,32,33), a liquid junction used by Karger et al. (11,26,34,35), or a sheath flow junction (12). The gold-coated capillary and liquid junction configurations are shown in Fig. 4.

80

Razunguzwa and Timperman

Fig. 4. (A) Schematic diagram of a microchip-capillary electrophoresis mass spectrometry configuration using a gold-coated spray tip. The electrospray ionization (ESI) voltage is applied via an electrode attached to the gold coating. (B) A schematic representation of a liquid junction. In this case the ESI voltage and the make-up solution is applied at the liquid junction.

References 1. Waters, L. C., Jacobson, S. C., Kroutchinina, N., Khandurina, J., Foote, R. S., and Ramsey, J. M. (1998) Microchip device for cell lysis, multiplex PCR amplification, and electrophoretic sizing. Anal. Chem. 70, 158–162. 2. Manz, A., Harrison, D. J., Verpoorte, E. M. J., et al. (1992) Planar chips technology for miniaturization and integration of separation techniques into monitoring systems. Capillary electrophoresis on a chip. J. Chromatogr. 593, 253–258. 3. Ross, D. and Locascio, L. E. (2002) Microfluidic temperature gradient focusing. Anal. Chem. 74, 2556–2564.

Interfacing Microchip Capillary Electrophoresis With ESI-MS

81

4. Wang, C., Oleschuk, R., Ouchen, F., Li, J., Thibault, P., and Harrison, D. J. (2000) Integration of immobilized trypsin bead beds for protein digestion within a microfluidic chip incorporating capillary electrophoresis separations and an electrospray mass spectrometry interface. Rapid Commun. Mass Spectrom. 14, 1377–1383. 5. Manz, A. and Becker, H. (1997) Parallel capillaries for high throughput in electrophoretic separations and electroosmotic drug discovery systems. Transducers 97, International Conference on Solid-State Sensors and Actuators, Chicago, June 16–19, 2, 915–918. 6. Liu, H., Felten, C., Xue, Q., et al. (2000) Development of multichannel devices with an array of electrospray tips for high-throughput mass spectrometry. Anal. Chem. 72, 3303–3310. 7. Tang, K., Lin, Y., Matson, D. W., Kim, T., and Smith, R. D. (2001) Generation of multiple electrosprays using microfabricated emitter arrays for improved mass spectrometric sensitivity. Anal. Chem. 73, 1658–1663. 8. Gelpi, E. (2002) Interfaces for coupled liquid-phase separation/mass spectrometry techniques. An update on recent developments. J. Mass. Spectrom. 37, 241–253. 9. Xue, Q., Foret, F., Dunayevskiy, Y. M., Zavracky, P. M., McGruer, N. E., and Karger, B. L. (1997) Multichannel microchip electrospray mass spectrometry. Anal. Chem. 69, 426–430. 10. Ramsey, R. S. and Ramsey, J. M. (1997) Generating electrospray from microchip devices using electroosmotic pumping. Anal. Chem. 69, 1174–1178. 11. Zhang, B., Liu, H., Karger, B. L., and Foret, F. (1999) Microfabricated devices for capillary electrophoresis-electrospray mass spectrometry. Anal. Chem. 71, 3258–3264. 12. Li, J., Thibault, P., Bings, N. H., et al. (1999) Integration of microfabricated devices to capillary electrophoresis-electrospray mass spectrometry using a low dead volume connection: application to rapid analyses of proteolytic digests. Anal. Chem. 71, 3036–3045. 13. Li, J., Kelly, J. F., Chernushevich, I., Harrison, D. J., and Thibault, P. (2000) Separation and identification of peptides from gel-isolated membrane proteins using a microfabricated device for combined capillary electrophoresis/nanoelectrospray mass spectrometry. Anal. Chem. 72, 599–609. 14. Figeys, D., Ning, Y., and Aebersold, R. (1997) A microfabricated device for rapid protein identification by microelectrospray ion trap mass spectrometry. Anal. Chem. 69, 3153–3160. 15. Schultz, G. A., Corso, T. N., Prosser, S. J., and Zhang, S. (2000) A fully integrated monolithic microchip electrospray device for mass spectrometry. Anal. Chem. 72, 4058–4063. 16. Wen, J., Lin, Y., Xiang, F., Matson, D. W., Udseth, H. R., and Smith, R. D. (2000) Microfabricated isoelectric focusing device for direct electrospray ionizationmass spectrometry. Electrophoresis 21, 191–197. 17. Wilm, M. and Mann, M. (1996) Analytical properties of the nanoelectrospray ion source. Anal. Chem. 68, 1–8.

82

Razunguzwa and Timperman

18. Valaskovic, G. A., Kelleher, N. L., Little, D. P., Aaserud, D. J., and McLafferty, F. W. (1995) Attomole-sensitivity electrospray source for large-molecule mass spectrometry. Anal. Chem. 67, 3802–3805. 19. Jacobson, S. C., Hergenroder, R., Koutny, L. B., Warmack, R. J., and Ramsey, J. M. (1994) Effects of injection schemes and column geometry on the performance of microchip electrophoresis devices. Anal. Chem. 66, 1107–1113. 20. Khaledi, M. G. (1998) High-Performance Capillary Electrophoresis: Theory, Techniques, and Applications, Wiley, New York, NY. 21. Chiem, N., Lockyear-Shultz, L., Andersson, P., Skinner, C., and Harrison, D. J. (2000) Room temperature bonding of micromachined glass devices for capillary electrophoresis. Sens. Actuators B Chem. B63, 147–152. 22. Gaskell, S. J., Bolgar, M. S., Riba, I., and Summerfield, S. G. (1997) Electrospray ionization: theory and application. NATO Adv. Stud. Inst. Ser C: Mathematical and Physical Sciences 504, 3–16. 23. Belder, D. and Ludwig, M. (2003) Surface modification in microchip electrophoresis. Electrophoresis 24, 3595–3606. 24. Li, J., LeRiche, T., Tremblay, T. -L., et al. (2002) Application of microfluidic devices to proteomics research: identification of trace-level protein digests and affinity capture of target peptides. Mol. Cell Proteomics 1, 157–168. 25. Figeys, D. and Aebersold, R. (1998) Nanoflow solvent gradient delivery from a microfabricated device for protein identifications by electrospray ionization mass spectrometry. Anal. Chem. 70, 3721–3727. 26. Zhang, B., Foret, F., and Karger, B. L. (2000) A microdevice with integrated liquid junction for facile peptide and protein analysis by capillary electrophoresis/electrospray mass spectrometry. Anal. Chem. 72, 1015–1022. 27. Lazar, I. M., Ramsey, R. S., Jacobson, S. C., Foote, R. S., and Ramsey, J. M. (2000) Novel microfabricated device for electrokinetically induced pressure flow and electrospray ionization mass spectrometry. J. Chromatogr. 892, 195–201. 28. Alarie, J. P., Jacobson, S. C., and Ramsey, J. M. (2001) Electrophoretic injection bias in a microchip valving scheme. Electrophoresis 22, 312–317. 29. Li, J., Wang, C., Kelly, J. F., Harrison, D. J., and Thibault, P. (2000) Rapid and sensitive separation of trace level protein digests using microfabricated devices coupled to a quadrupole—time-of-flight mass spectrometer. Electrophoresis 21, 198–210. 30. Gobry, V., Van Oostrum, J., Martinelli, M., et al. (2002) Microfabricated polymer injector for direct mass spectrometry coupling. Proteomics 2, 405–412. 31. Lazar, I. M., Ramsey, R. S., Sundberg, S., and Ramsey, J. M. (1999) Subattomolesensitivity microchip nanoelectrospray source with time-of-flight mass spectrometry detection. Anal. Chem. 71, 3627–3631. 32. Deng, Y., Henion, J., Li, J., Thibault, P., Wang, C., and Harrison, D. J. (2001) Chip-based capillary electrophoresis/mass spectrometry determination of carnitines in human urine. Anal. Chem. 73, 639–646. 33. Li, J., Tremblay, T. -L., Thibault, P., Wang, C., Attiya, S., and Harrison, D. J. (2001) Integrated system for high-throughput protein identification using a micro-

Interfacing Microchip Capillary Electrophoresis With ESI-MS

83

fabricated device coupled to capillary electrophoresis/nanoelectrospray mass spectrometry. Eur. J. Mass. Spectrom. 7, 143–155. 34. Zhang, B., Foret, F., and Karger, B. L. (2001) High-throughput microfabricated CE/ESI-MS: Automated sampling from a microwell plate. Anal. Chem. 73, 2675–2681. 35. Foret, F., Zhou, H., Gangl, E., and Karger, B. L. (2000) Subatmospheric electrospray interface for coupling of microcolumn separations with mass spectrometry. Electrophoresis 21, 1363–1371.

8 Interfacing Amperometric Detection With Microchip Capillary Electrophoresis R. Scott Martin Summary Amperometric detection is a sensitive and selective way to monitor separations in microchip capillary electrophoresis (CE). This review contains 78 references and will educate the reader of the issues that are involved with interfacing amperometric detection and microchip CE. These issues include special injection protocols, separation mechanisms, and ways to integrate the working electrode with the separation channel. Some useful biological applications of the technique will also be described. Key Words: Amperometric; electrochemistry; capillary electrophoresis; microchip; micrototal analysis system; portable analysis system; microfluidics.

1. Introduction Capillary electrophoresis (CE) in the microchip format has come a long way over the past decade. Initial demonstrations of the technique (1–4) showed the numerous benefits of the microchip format that include fast analysis times, the use of high separation field strengths, minute consumption of solvents, and the possibilities for disposable/portable devices. These numerous advantages have lead to many exciting applications of microchip CE including fully integrated multichannel separation-based immunoassays (5), complex two-dimensional separations (6), and high-throughput 384-channel DNA separations (7). The small volume (<500 pL) of sample that is injected into the separation component of the chip and the micron-sized separation bands that are produced dictate that a sensitive detection technique be employed to monitor separations in these devices. Initially, most studies utilized laser-induced fluorescence (LIF) detection, primarily because of the simplicity of constructing these systems, the ease of focusing the laser beam in the channels, and the low limits of detection (LOD) that are achievable (8). The disadvantages of the LIF approach From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

85

86

Martin

include the requirement for either pre- or postcolumn derivatization with a fluorophore and the fact that only a select number of wavelengths can be used for excitation. Mass spectrometry (MS) has been used more recently (9). MS can provide a massive amount of chemical information, but commercially available systems are not inherently portable and are more costly and less sensitive than LIF detection. Other techniques have been demonstrated but not widely utilized. These include absorbance (10), chemiluminscence (11), electrochemiluminescence (12), refractive index (13), and Shah convolution Fourier transform detection (14). Obviously, as the interest and use of microchip CE grows, other detection techniques that can sensitively monitor analytes are needed. Additionally, an ideal detection technique is amenable to miniaturization, so that the detection system can be integrated onto the microchip. Electrochemical (EC) detection is a technique that fits these criteria. Many compounds can be detected sensitively and selectively without derivatization, with typical LODs in the mid-to-low nanomolar range. It is possible to fabricate the microelectrodes with many of the same photolithographic procedures that are used to construct the separation component of the microchips. In addition, the electrode can be fabricated directly on the chip, leading to a fully integrated system. Although the small channels used in microchip CE compromise the sensitivity of optical detection modes, the sensitivity of EC detection is not, as the increased flux toward the microelectrode surface (15) and the reduced background current of microelectrodes (16) leads to an increased signal-to-noise ratio. Electrochemical detection is a broad term that is used to describe three different detection modes, conductimetry, potentiometry, and amperometry. This review will focus on interfacing amperometric detection with microchip CE. Conductimetric and potentiometric detection for conventional and microchip CE has been recently reviewed (17) and issues involved with interfacing conductimetric detection with microchip CE is the subject of a chapter in this book by Hergenröder. The reader is referred to these works for further information into other modes of EC detection. For simplicity, future reference to EC detection will be made with respect to amperometric detection. To date, amperometric detection is the most widely reported mode of EC detection for microchip CE, although conductimetric detection is gaining in popularity. Amperometric detection involves using a potentiostat to apply a constant potential to the working electrode, relative to a reference electrode, and measuring the current that is produced from redox reactions (oxidation or reduction of electroactive compounds) that occur at the working electrode surface. With a conventional three-electrode setup, an auxiliary electrode is also used to ensure that no current passes through the reference electrode and to minimize ohmic potential drops (18). The current that is produced at the working

Interfacing Amperometric Detection With Microchip CE

87

electrode is directly proportional to the number of moles of analyte oxidized or reduced at the electrode surface. This can be described by Faraday’s law: Q = nFN

(1)

where the number of Coulombs (Q) is directly proportional to the number of moles (N) of product converted at the electrode surface, the number of electrons (n) transferred in the redox reaction, and Faraday’s constant (96485 C/mol) (19). In a separation (where current is monitored as a function of time), Q can be obtained from the peak area for a particular analyte and is directly proportional to concentration. Instantaneous current (such as peak height in a separation) is also directly proportional to concentration and can be derived from Faraday’s law by differentiating Eq. 1 with respect to time: it =

dQ dN = nF dt dt

(2)

where t is time (19). The potential of the working electrode can be adjusted to achieve selectivity. This potential usually corresponds to the limiting current plateau region for the electroactive compounds of interest, which can be determined by hydrodynamic voltammetry. The purpose of this review is to give the reader insight into the issues involved with integrating amperometric detection and microchip CE. Although some applications will be described, this review will focus on performing the technique and is not intended to be a comprehensive review of the applications of microchip CEEC. The reader is referred to other reviews for a comprehensive listing of microchip CEEC applications (8,15,20,21). 2. Materials 2.1. Microchip Fabrication The two most common substrates used in the fabrication of devices for microchip CE with amperometric detection are glass and poly(dimethylsiloxane) (PDMS). Typically, glass devices are fabricated from soda lime glass coated with chrome and photoresist. Such glass can be coated in-house or purchased precoated from commercial sources (Telic Company, Santa Monica, CA; www.telic2000.com). Description of glass chip fabrication can be found throughout the literature and in Chapter 2 (22,23). PDMS-based devices are produced by casting PDMS prepolymer against a positive-relief structure as described previously and highlighted in Chapter 3 (24). PDMS is commercially available from Dow Corning (Midland, MI; www.dowcorning.com) as a Sylgard 184 kit with base and curing agent. PDMS layers can be sealed against another layer of PDMS or to other substrates (such as glass) in a reversible (conformal contact) or irreversible (after plasma oxidation) fashion (24). The latter is typically accomplished

88

Martin

with a low-cost plasma oxidizer (such as those sold by Harrick Scientific, Ossining, NY; www.harricksci.com) using air as the oxygen source. 2.2. Amperometric Detection Electrodes for microchip CE can be obtained from many different sources depending upon the type of alignment/integration strategy that is being utilized (see Subheading 3.). In particular, integrated thin-layer electrodes are usually deposited onto a substrate (typically glass) by sputtering or evaporation, followed by patterning with photolithography (25–27). Starting metal materials for most evaporation/sputtering systems are available from Kurt J. Lesker Co. (Clairton, PA; www.lesker.com). Screen-printed electrodes are produced by large screen-printers (28) and the carbon ink is available from both Acheson Industries (Port Huron, MI; www.achesonindustries.com) and Ercon Inc. (Wareham, MA; www.erconinc.com). Conventional electrode materials, such as carbon fibers and metal electrodes, are commonly employed in microchip CE and are available from various suppliers including Union Carbide (source of carbon fiber; Danbury, CT; www.unioncarbide.com) and Goodfellow (source of metal electrodes; Berwyn, PA; www.goodfellow.com). Amperometric detection is typically carried out in a three-electrode format using commercially available potentiostats. The two most common sources of potentiostats are CH Instruments (Austin, TX; www.chinstruments.com) and Bioanalytical Systems (West Lafayette, IN; www.bioanalytical.com). These companies also provide conventional reference and auxiliary electrodes. Since the channel dimensions and IR drops are small in these devices, two electrode configurations can be utilized using custom (29) or commercially available potentiostats (Keithley Instruments, Cleveland, OH; www.keithley.com) (30). 3. Methods The issues involved with interfacing microchip CE and amperometric detection can be broken down into three logical parts: injection, separation, and detection. Although the injection and separation aspects of microchip CEEC are not radically different from microchip CE with LIF detection, there are issues with these components that are unique to amperometric detection. 3.1. Injection Injection of discrete volumes of analyte from a sample reservoir into the separation channel is an important component of a microchip CE device. Injection protocols are unique to microchip CEEC because in most cases the detection reservoir, which contains the working, auxiliary, and reference electrodes connected to a potentiostat (Fig. 1), must be held at constant ground (see Note 1). This is not a requirement of microchip CE with optical detection modes, as

Interfacing Amperometric Detection With Microchip CE

89

Fig. 1. A typical microchip CEEC device with a potential being applied across the separation channel.

voltages can be applied to the detection reservoir without affecting the detector response. Four different injection protocols have been used with microchip CEEC devices. These are unpinched (25,31,32), pinched (26,33), gated (34,35), and hydrodynamic (36). Each of these approaches holds the detection reservoir at constant ground. The simplest injection methodology is the unpinched approach, which only requires a single power supply (25,31,32). A high voltage is applied to the sample reservoir for 1–5 s with the detection reservoir held at ground (Fig. 1). During this time, the sample is introduced directly into the separation channel by electrokinetic forces. After the injection is complete, the high voltage is switched back to the buffer reservoir, and the separation is initiated. Unpinched injections do not use pushback voltages to prohibit sample leaking from the side channels into the main separation channel. In addition, the unpinched approach can result in irreproducible injections and larger than normal sample plugs. A more reproducible way to introduce discrete sample plugs is termed pinched injection. This approach requires voltage control of each reservoir, meaning that one needs a separate voltage source for the sample, sample waste, and buffer reservoirs (Fig. 1) (37–39). To carry out a pinched injection sequence for microchip CEEC, the tee (Fig. 1) or twin-tee (Fig. 2B) is first filled for a specified period of time by applying a negative potential to the sample waste reservoir while keeping the sample, buffer, and detection reservoirs at ground (26,33). After the tee is filled, the separation is initiated by applying a high voltage to the

90 Fig. 2. Two injection schemes for microchip CEEC. (A) Gated injection: I. Chip layout and fluorescent micrographs showing the injection sequence; II. Plot showing the effect of the injection to separation voltage ratio on the response of the detector. (B) Hydrodynamic injection: I. Chip layout; II. Fluorescent micrographs showing the injection sequence. (Reprinted with permission from refs. 35 and 36, respectively.)

Interfacing Amperometric Detection With Microchip CE

91

buffer reservoir and a fraction of this high voltage to the sample and sample waste reservoirs. The application of a partial voltage to the sample and sample waste reservoirs maintains a small pushback flow in these channels and keeps analyte from leaking into the separation channel, while maintaining the desired plug shape. By taking into account the design geometry and the resistivity of the buffers, the voltage at the tee junction can be calculated to adequately ensure proper voltage settings and plug formation (40). A third method that can be used to inject discrete samples is a gated injection approach (34,35,41). This type of injection requires voltage control over each reservoir. The steps involved in this injection sequence are depicted in Fig. 2A (35). This type of approach utilizes two continuously flowing streams. A high voltage is applied to the buffer reservoir (B) and a fraction of that high voltage is applied to the sample reservoir (S) with the sample waste (SW) and detection reservoir (D) remaining at ground (the reservoir denoted as SR in Fig. 2A is not normally present in simple gated injection chip designs and can be ignored for this discussion). This results in a flow of sample toward the sample waste reservoir and a separate flow stream from the buffer to detection reservoir (micrograph a in Fig. 2A). Because of the low Reynolds numbers that are achieved in microchannels, mixing of the two streams does not occur. To inject a plug from the sample reservoir, the high voltage that is applied to the buffer reservoir is floated for a short period of time (1 s or less, micrograph b). This sweeps sample into the separation channel and the separation is initiated by resuming the high voltage to the buffer reservoir (micrograph c). For each new chip design, the ratio of injection (S to SW) to separation (B to D) voltages should be determined. With the chip design shown in Fig. 2A, a plot of peak height vs the ratio (Fig. 2A) shows that for this particular design, a ratio around 0.9 was sufficient, but this will vary depending upon the chip layout (35). A newly described injection methodology for microchip CEEC is depicted in Fig. 2B (36). This hydrodynamic method uses a twin-tee configuration, two voltage sources, and a sample reservoir that, as compared with the other reservoirs, has a slightly higher level of liquid. A gated injection-like approach is used, as separate voltages are applied from the buffer reservoir (B) to the detection reservoir (D), and from the sample reservoir (S) to the sample waste reservoir (SW). As shown in Fig. 2B (micrograph a) this leads to two separately flowing streams. Injection is carried out (micrograph b) by turning both supplies off for 3 s. During this time, the combination of the slightly higher fluid in the sample reservoir and the shorter length of the sample channel create a hydrodynamic pressure large enough to force sample solution into the separation channel (36). Turning both power supplies back on (micrograph c) results

92

Martin

in the sample plug being swept toward the detection reservoir and the flow streams are again separated. The particular injection methodology that one should utilize is dependent upon the type of analysis, the sample matrix, the amount of precision needed, and the number of voltage sources available. Unpinched injections are the simplest to implement but suffer from an electrokinetic bias and the precision of the injection can be problematic (33). In comparison, pinched injections offer superior injections in terms of precision, but when the ionic strengths of the buffer and sample are mismatched, the actual amount injected can be inconsistent with the appearance of the plug formed in the injector (38). Initially, pinched injections were thought to not suffer electrokinetic bias (37). A later study has shown that neutral species are injected preferentially to anionic species (39), but the extent of this bias can be controlled by appropriate selection of the electric field strengths. Gated injections are also more precise than the unpinched approach and can be run in a continuous sampling mode, however, this method also leads to an electrokinetic bias (42) and the ionic strength of the sample and buffer must be similar (35). Finally, the hydrodynamic method does not suffer any sort of bias, as it is a pressure-based injection, but the approach described here is sensitive to evaporation processes. 3.2. Separation Issues involved in performing CE-based separations with amperometric detection are not substantially different than those in other microchip CE devices (see Note 2). As previously noted, the detection reservoir must be held at constant ground, thus, no type of separation voltages can be applied to the detection reservoir. Buffers that are typically used in other microchip CE devices can also be used with microchip CEEC. However, some buffers such as phosphate and sodium hydroxide lead to very large background detector currents in CE (43). Zwitterionic buffers (such as TES) (43) and boric acid lead to much reduced background detector currents, and thus, lower limits of detection. Some experiments require buffer additives to enhance the separation or alter the electroosmotic flow (EOF). With microchip CEEC, care must be taken to ensure that the additives do not interact with the electrode surface. This can lead to electrode passivation or increased background detector currents. There are several examples of using buffer additives with microchip CEEC. Schwarz et al. (44) demonstrated chiral separations with microchip CEEC using carboxymethyl-β-cyclodextrin to separate enantiomers of adrenaline, noradrenaline, ephedrine, and pseudoephedrine. In some cases 18-crown-6 ether was included to enhance the separation. Baseline resolution of these enantiomers was possible and the background detector current was not affected by the additive. However, the signal for each analyte was significantly decreased in the pres-

Interfacing Amperometric Detection With Microchip CE

93

ence of the cyclodextrin because the complexed fraction of the analyte is rendered nonelectroactive (44). Often, additives are used to control the EOF or modify the surface of the separation channel. Dou et al. utilized 2-morpholinoethanesulfonic acid (MES), a surfactant, to interact with the hydrophobic channel walls in chips made from (PDMS) (45). The addition of MES to the sodium hydroxide buffer system enhanced the separation efficiency and peak shape of arginine and glucose, which were detected with a copper electrode. This was owing to interaction of the apolar section of MES with the hydrophobic PDMS surface, leaving the polar head group oriented in the separation channel, making the overall surface more hydrophilic (45). Another dynamic coating approach was described by Henry’s group (34). In this work, a polymer bilayer consisting of a cationic layer of polybrene and an anionic layer of dextran sulfate was formed in PDMS channels. The coating exhibited pH-independent EOF in the pH range from 5.0 to 10.0 and was stable for over 100 runs. In addition, when comparing coated and uncoated channels, it was found that there was no significant variation in the dopamine and hydroquinone amperometric signals, signifying that the electrode surface was not passivated (34). Finally Susan Lunte’s group was able to demonstrate reversal of the EOF with use of dodecyl trimethyl ammonium bromide (DTAB) (35). In this work, the target analyte, nitrite, is small in size, negatively charged, and elutes very late under normal EOF conditions. Reversal of the EOF with DTAB and a negative separation polarity reversed the elution order, with small anions eluting first. This allowed the separation and EC detection of nitrite in 40 s, with the EOF modifier not adversely affecting the carbon fiber electrode (35). 3.3. Amperometric Detection A very important issue that arises when using amperometric detection with microchip CE is isolation of the detector from the separation voltage. Commercially available potentiostats are grounded and if some method is not used to isolate the potentiostat from the separation voltage the potentiostat provides a path to ground, exposing the detector electronics to the high separation fields. Many different approaches to this interface have been described for microchip CEEC and these approaches can be divided into many different classifications. First, the chip/detector interface can be classified in terms of how the electrode is integrated with the chip. The working electrode can be integrated directly on the chip or positioned external to the chip and positioned carefully at the end of the separation channel. The chip/detector interface can be further classified in terms of how the working electrode is isolated from the EC detector. The most prevalent method, end-channel detection, involves placing the working electrode tens of microns from the exit of the separation channel.

94

Martin

The second method is termed in-channel detection and it entails using an electrically isolated potentiostat to place the working electrode directly in the separation channel. The third method, decoupled detection, involves grounding the separation voltage slightly before it reaches the working electrode. All of these methods are described next. 3.4. Integrated On-Chip Electrode Alignment This electrode alignment method involves integrating the working electrode directly on the microchip and assembling the chip so that the working electrode is aligned properly with the separation channel (Fig. 1). Some type of photolithographic procedure usually defines the working electrode. Gold (25,34,46), platinum (23,26,30,47–49), palladium (50), copper (51), gold wires (52), carbon fiber (33,35,53), carbon paste (54), and carbon ink (55–57) have been patterned by microfabrication procedures and used in one of the alignment schemes described next. Once the electrode is properly aligned at the end of the separation channel, the electrode layer is bonded to the separation channel layer, where the bonding method is dependent upon the type of substrate(s) (22,24,58). With this approach, the reference electrode should be placed reproducibly close to the working electrode, as changing this distance effects the effective working electrode potential (59) and can change the analyte response (26) (see Note 3). Advantages to integrating the electrode directly on the chip include the fact that the electrodes are reproducibly made with the same photolithographic procedures that are used to make the separation component of the chip, meaning the chip is amenable to mass production. Furthermore, the working electrode can be reproducibly aligned with the separation channel, leading to increased reproducibility (see Note 4). Finally, the other pertinent electrodes (high voltage, auxiliary, and reference) can be integrated on the chip as well, minimizing the complexity of the external inputs (23). A disadvantage to this approach is the fact that the electrode-containing layer must be bonded with the separation channel layer. While this is usually not a troublesome issue with plastic (58) and PDMS devices (25), bonding of glass devices requires thermal methods. The stability of metal electrodes may be effected during this bonding, as the lifetime of bilayer metal electrodes (glass adhesion layer such as titanium and metal layer such as platinum or gold) is affected by grain boundary diffusion of the metal adhesion layer through the overlying metal layer, and the rate of this diffusion process is increased at higher temperatures (27,60). An approach where the metal electrodes are evaporated into preformed glass channels seems to alleviate this problem, as the electrode is surrounded on three sides by glass; this also eliminates incomplete bonding around the electrode surface (23). The integrated electrode approach can be further classified by how the working electrode is isolated from the separation voltage. End-channel, in-channel,

Interfacing Amperometric Detection With Microchip CE

95

and decoupled detection have all been described for the integrated electrode approach. 3.4.1. End-Channel Detection End-channel detection is the most popular approach to integrating on-chip electrodes. This approach has been used with devices made in glass (23,26, 30,47,48) and PDMS (33,35,54,57), as well as PDMS/glass hybrid devices (25,34,46,51). The key part to this process is alignment of the working electrode with respect to the end of the separation channel. The small gap between the electrode and channel form the basis for decoupling of the separation voltage, as the detection reservoir, where the electrode is located, provides a path to ground, thus protecting the potentiostat from high voltage. An example of end-channel alignment is shown in Fig. 3A (23). In this example, the platinum working electrode is placed 50-microns from the end of the separation channel. This all-glass device also utilized a “shelf” structure, where the channel exit widens horizontally but not vertically, similar to an approach used in another glass device by Wooley et al. (26). This helps to limit the amount of diffusion away from the electrode surface while still providing sufficient grounding. Obviously, this type of alignment must be made reproducibly. This can be achieved using microscopes with calibrated reticules and placing alignment marks on the prebonded chip surfaces (see Note 4). Henry’s group recently described a PDMS-based chip where channels for electrodes were integrated on the same layer as the separation channel, which fixes the end-channel alignment to an exact and reproducible distance (52). A general problem with an endchannel approach is the band broadening that occurs as a result of the gap between the separation channel and the working electrode. This gap leads to band dispersion and loss of efficiency, and is the primary reason that the separation performance of EC detection has been inferior to LIF detection (53). Whereas LIF detects analytes directly in the separation channel, end-channel EC configurations detect analytes as they exit the separation channel. The following two methods have been used to place the working electrode directly in the separation channel to eliminate band broadening. 3.4.2. In-Channel Detection In-channel detection involves placement of the working electrode directly in the separation channel (Fig. 3B) (53). This was made possible by the development of a miniaturized, electrically isolated potentiostat. This “floating” potentiostat draws its power from a 9V battery and transmits data through optical isolators. Since there is not a path to ground, the potentiostat electronics are protected and the electrode is not exposed to the electric field. Although other

96

Martin

Fig. 3. Types of integrated on-chip electrode alignment schemes. (A) End-channel detection; (B) in-channel detection; (C) decoupled detection. (Reprinted with permission from refs. 23, 53, and 50, respectively.)

studies have shown that glass devices with metal-based electrodes exhibit electrolysis (and subsequent bubble formation) at electrode surface when they are exposed to electric fields (12), this PDMS device with carbon working electrodes did not, as PDMS is permeable to gases (24). This approach was used to study the effect of the working electrode position on the separation performance (in terms of plate height and peak skew) by comparing an end-channel configuration with this new in-channel approach. Using catechol as the test analyte, it was found that in-channel EC detection decreased the total plate height by a factor of 4.6 and lowered the peak skew by a factor of 1.3. A similar trend was observed for the small, inorganic ion nitrite. However, as with end-column detection in conventional CE (59), a potential shift of the working electrode occurs. This shift is a function of the separation voltage and the distance of the electrode from the end of the separation channel. Therefore, hydrodynamic voltammetry must be used to ensure optimum detection potential. Studies using a fluorescent and electrochemically active amino acid derivative to compare the separation performance of in-channel EC detection to that of a widely used

Interfacing Amperometric Detection With Microchip CE

97

laser-induced fluorescence (LIF) detection scheme were also performed. In this case, it was found that the plate height and peak skew for both detection schemes were essentially equal, and the separation performance of in-channel EC detection is comparable to LIF detection. In addition, the isolated potentiostat was compact (4 × 9 × 2 cm), which holds promise for integrating the EC detector directly on the chip. 3.4.3. Decoupled Detection In this method, the EC detector is isolated from the separation voltage by means of a “decoupler,” which provides a current path to ground just before the working electrode (Fig. 3C) (50). The decoupler shunts the separation voltage to ground, creating a field-free region where analytes are pushed past the working electrode by the EOF generated prior to the decoupler. This approach is very similar to the previously described in-channel alignment, but the working electrode is minimally influenced by the separation voltage. Several different groups have described decouplers for microchip CEEC. Rossier et al. utilized a microchip device composed of two different polymers (55,56). Separation and injection channels were made in a layer of polyethyleneterephthalate (PET) and several 10-micron holes were made in a layer of polyethylene (PE) The two different substrates were laminated together and a decoupling reservoir was made on the PE layer over the 10-micron holes. These holes provided a path to ground while the inherent hydrophobicity of PE helped to prevent bulk for some of the fluid flow through the holes. The working electrode, made of carbon ink, was placed just downstream from the decoupler. One disadvantage of this report is the relatively high limits of detection (5 µM for aminophenol), possibly owing to incomplete decoupling or analyte leakage (56). A palladium metal film decoupler was first described by Chen and co-workers (50). Metals of the platinum group, such as palladium, have the ability to absorb hydrogen that is produced by the electrolysis of water at the electrophoretic ground. In this work, palladium metal was evaporated onto a plastic microchip and used for both the decoupler and working electrodes (Fig. 3C). The palladium decoupler did not produce sample leakage or discernible dead volumes and a detection limit of 290 nM for dopamine was reported. These authors did not have access to any fabrication facilities; therefore, the methods they used for producing the chips generated very large separation channels and working electrodes (200 µm and 1 mm, respectively). This and the large spacing between the decoupler and working electrode (1 mm) may help to explain the low separation efficiencies that were achieved (<10,000 plates). Nevertheless, this approach holds great promise for the continued development of microchip CEEC devices.

98

Martin

Lacher et al. recently described a hybrid PDMS-glass device that employed a palladium decoupler and working electrode (61). The effect of the decoupler size on the ability to remove hydrogen was evaluated with regard to reproducibility/ longevity. A 500 µm decoupler was found to be the optimum decoupler size, with effective voltage isolation lasting for approx 6 h at a constant field strength of 600 V/cm. The effect of distance between the decoupler and working electrode on noise and resolution for the separation of dopamine and epinephrine was also investigated. It was found that 250 µm was the optimum spacing between the decoupler and working electrode. At this spacing, laser-induced fluorescence detection at various points around the decoupler established that the band broadening resulting from pressure-induced flow that occurs after the decoupler did not significantly affect the separation efficiency of fluorescein. The limit of detection for dopamine using the optimized design was found to be 500 nM (61). Finally, a platinum decoupler was described for microchip CEEC, with fully integrated working, reference, and auxiliary electrodes (49). In this approach, the authors fabricated gold electrodes and electrically deposited platinum nanoparticles on the ground and reference electrode. Although platinum is commonly used as a grounding electrode material, as previously described, electrolysis of water at the cathode does produce H2 gas and platinum does not absorb H2 as effectively as palladium. If the gas is not effectively absorbed, H2 bubbles form and block the separation channel. In this work, while using a MES buffer, a field strength of 90 V/cm was able to be applied before this bubble formation took place. While this is a low field strength for CE, the baseline was extremely low in noise (<0.05 pA) and an LOD of 125 nM for dopamine was achieved. Although the low field strengths that are required limit the separation efficiencies, as in CE where separation efficiencies are directly proportional to field strength (with typical field strengths in microchip CE applications being 300 V/cm or greater), the authors were able to show that the pseudo platinum reference electrode offers a stable potential for EC detection, both with and without the presence of the electric field. 3.5. Externally Positioned Off-Chip Electrode Alignment Although an integrated on-chip electrode alignment results in a complete device that is amenable to mass production and has a fixed electrode alignment, it is not always possible to utilize this approach. For example, Micralyne (www.micralyne.com) offers commercially available chips that are already bonded and contain separation channels. The fact that they are bonded prohibits any on-chip integration of electrodes, but the use of externally positioned electrodes allows groups who do not have fabrication facilities to utilize microchip CEEC in their studies. Furthermore, for laboratory studies it is often easier to externally position the electrode off the chip and align the electrode at the end

Interfacing Amperometric Detection With Microchip CE

99

of the separation channel with some type of manipulator. An attractive feature of this approach is that if something goes wrong with the electrochemical part of the device, the separation component of the chip can still be used, and vice versa. This is in contrast to the integrated on-chip approach, where if any part of the chip fails, the entire chip must be discarded. This section will describe devices that have used externally positioned electrodes with microchip CEEC. The only type of alignment scheme that can be used with this approach is endchannel, as in all of these examples the chip is completely bonded before the electrode is interfaced with the chip. 3.5.1. Screen-Printed Approach Work from Joe Wang’s lab at New Mexico State University has focused on the use of screen-printed electrodes for microchip CEEC. The key component to this approach is the screen-printed electrode, which is made by using a screen printer and a stencil to pattern 100-micron thick structures on alumina ceramic plates (62). Up to 30 electrode strips can be made per printing, leading to reproducible mass production of the EC component of the chip. Several different inks can be utilized (32,62) and the inks can be used in microchip CEEC directly (32,63) or after electrochemical-based coating with gold (64–69) or palladium (70). The screen-printed electrode is interfaced with the microchip part of the device by the approach demonstrated in Fig. 4A (23). The screen-printed electrode (L, where the electrode material, M, is insulated, O, and connected to the potentiostat via silver ink, N) is interfaced to the chip by a specially made Plexiglas® chip holder. This holder has a plastic screw (U) that, along with a tape spacer on the electrode (P), is used to position the working electrode (M) with the end of the separation channel (Q). The Plexiglas holder is precisely made so that the middle of the working electrode is centered with the separation channel. The rest of the device houses reservoirs for buffer solutions, platinum electrodes for the separation (T) and the counter electrode (R), and a silver wire for the reference electrode (S). The advantage of this approach is the ease in which either the chip or the electrode can be interchanged. However, the distance between the electrode and chip has to be made manually and this distance must be made reproducibly, or a shift in the half-wave potential can occur (53,59). Regardless, this approach has been shown to be a powerful approach to interfacing amperometric detection with microchip CE. 3.5.2. Micromanipulated Electrode Conventional CE with EC detection is most commonly performed with microelectrodes that are pulled in glass capillaries (71). The same approach has been used with microchip CE, with the microelectrodes being interfaced to the microchip with commercially available micromaniupluators. An advantage of

100 Fig. 4. Types of externally positioned off-chip electrode alignment scheme. (A) Screen-printed electrode approach. (B) Micromanipulated electrode approach. Labels are explained in the text. (Reprinted with permission from refs. 32 and 73, respectively.)

Interfacing Amperometric Detection With Microchip CE

101

this approach is the wide variety of electrode materials that can be utilized, as wires or fibers, and are commercially available from a wide variety of sources. Electrode materials that have been used with this approach include platinum (72,73), gold (74), copper (45), carbon fiber (36,75), platinum coated with gold (29,44,76), and copper (29). An example of how these microelectrodes are interfaced with microchip CE is shown in Fig. 4B (73). In this example, the microchip was a hybrid PDMS/glass device. The glass part of the chip (J) and the detection surface of the working electrode (C) were beveled with a sander to aid the electrode alignment. A commercially available translational stage (A) was used to micromanipulate the working electrode at the end of the separation channel (see inset of Fig. 4B). The auxiliary (D) and reference (E) electrodes were manually inserted into the detection reservoir (R4). Different variants of the micromanipulator approach have been described (29,36,44,45,72–76), but all of these methods use a similar methodology. A disadvantage of this approach is the manual nature of aligning the working electrode and the ease with which this alignment can be altered (see Note 4). As stated earlier, this alignment is crucial to the reproducibility of the method (53,59). Hauser’s group has described an approach that simplifies the number of electrodes needed to perform the separation and EC detection (29). In this work, a micromanipulated electrode was used in conjugation with a specially designed potentiostat. The system utilizes only a working electrode and the electrophoretic ground to complete the EC cell. The working electrode is aligned at the end of the separation channel as previously described and the ground electrode serves as both a pseudo-reference and the counter electrode (29). This setup is feasible because the electrolysis of water occurs at the ground electrode, providing a constant reference potential that is unique for a given set of separation conditions. However, a shift in the reference potential can take place if the current passing through the ground electrode changes significantly during a separation. Therefore, with this approach, a hydrodynamic voltammogram should be obtained for each analyte under any new separation conditions. 3.6. Selected Results Microchip CE with amperometric detection has been used for many different applications ranging from environmental monitoring, to clinical assays, to DNA analysis. Although reviewing all of these applications is beyond the scope of this review, several examples for each of the alignment schemes previously described are given. The reader is referred to other reviews for a comprehensive listing of microchip CEEC applications (8,15,20,21). 3.6.1. Integrated On-Chip Amperometric Detection

102

Martin

Fig. 5. DNA analysis using microchip CEEC with an integrated on-chip platinum electrode in an end-channel alignment. (A) Separation of a ΦX174 HaeIII restriction digest. (B) Separation of a Salmonella PCR product (shaded) and ΦX174 HaeIII restriction digest. (Reprinted with permission from ref. 26.)

One of the first reports of microchip CE with amperometric detection utilized a completely integrated EC detector, with a platinum-working electrode aligned in an end-channel configuration (26). The advantages of utilizing microchip CEEC were demonstrated by performing DNA analysis. Although capillary array electrophoresis with LIF detection was used to sequence the entire genome, the sequence is currently a compilation of only a few individuals; to map the genomes of additional individuals and other species, improvements in technology, in terms of speed, size, and portability, are needed. The advantages of microchip CEEC fit many of those needed improvements and Mathies’ group demonstrated the ability to monitor DNA restriction fragments and PCR products in a microchip CEEC format (26). In this work, an indirect EC detection approach was used with the electrochemically active intercalation reagent

Interfacing Amperometric Detection With Microchip CE

103

Fe(phen)3+2. A constant background results from free Fe(phen)3+2; when complexes containing Fe(phen)3+2 migrate past the working electrode, a decrease in the background current results. This approach was used for the sizing of a ΦX174 HaeIII restriction digest (Fig. 5A) and to size PCR products from Salmonella (Fig. 5B) in less than 4 min. This study showed that CEEC microchips can be used for rapid, highly sensitive biochemical assays and the performance is competitive with traditional fluorescence methods. An approach that uses microchip CEEC to indirectly measure nitric oxide (NO) production by monitoring of nitrate and nitrite has recently been described using a carbon fiber-working electrode in an end-channel configuration (35). NO is an important signaling compound in many physiological events, including neurotransmission, vasodilatation, and inflammation. The regulation of NO production is a goal of therapeutic approaches to a number of different pathophysiological conditions, including autoimmune disorders, inflammation, hypertension, and neuropathic abnormalities. Because of its short half-life, direct measurement of NO is difficult; therefore, monitoring of nitrate and nitrite has been employed as a useful indicator of in vivo NO production. The approach described by Kikura-Hanajiri et al. (35) combines determination of nitrite by direct amperometric detection with separation by microchip CE. In a separate step, nitrate is converted to nitrite on-chip, by chemical reduction using copper-coated cadmium particles; once the reaction is complete the resulting nitrite is separated and detected by microchip CEEC. The amount of nitrate is quantified by calculating the difference in the amount of nitrite in the sample before and after the reduction of nitrate. As can be seen in Fig. 6, separation, injection, detection, and the reduction reaction were all successfully integrated onto one microchip device. Figure 6A shows the nitrite signal from a nitrite standard, whereas Fig. 6B shows the nitrite signal that results from the on-chip reduction of a nitrate sample. This approach was used to monitor the production of nitrate and nitrite from 3-morpholinosydnonimine, a metabolite of the vasodilator molsidomine and a NO-releasing compound (Fig. 6C). This study demonstrated the usefulness of microchip CEEC for monitoring of NO, as well as the ability to integrate multiple processes (reactions, injections, separation, and detection) onto one microchip device. 3.6.2. Externally Positioned Off-Chip Amperometric Detection Microchip CEEC with screen-printed electrodes positioned off-chip has been used for many applications including explosives analysis (32), monitoring of phenolics in the environment (64), amino acid analysis (64), and enzyme bioassays (66,68,77). Another interesting application of these devices is the development of microchip-based immunoassays (63,69). There are many advantages to using the microchip format to perform immunoassays; these include the use of

104

Interfacing Amperometric Detection With Microchip CE

105

small amounts of often precious reagents, the ability to separate reagents from the complexes by CE, and, with electrochemical detection, the ability to sensitively measure the complexes. Both direct (noncompetitive) (63,69) and competitive (69) formats have been demonstrated. An example of a competitive electrochemical immunoassay using microchip CEEC with an externally position screen-printed electrode is shown in Fig. 7 (69). In this study, ferrocene-labeled antigen (Ag*, where the antigen is 3,3′,5-triiodo-L-thyronine, or T3) and antigen analyte (Ag) are mixed on-chip with the antibody (Ab, where the antibody is mouse immunoglobulin G, or IgG), and a separation of the free and bound labeled antigen is carried out (Fig. 7A). The antigen (T3) cannot be measured under direction immunoassay, where the antibody is labeled with ferrocene, owing to the small difference in the mass/charge ratio between the free antibody and the immunocomplex. Figure 7B shows the calibration data for the competitive assay of T3, with the inset showing an electropherogram of the ferrocenelabeled T3 complex along with an aminophenol internal standard. As compared with a direct assay, the competitive assay exhibited higher sensitivity. These studies show the usefulness of microchip CEEC to sensitively and selectively monitor biologically important analytes via an immunoassay approach. Finally, a micromanipulated off-chip electrode approach was used to develop a clinical assay for uric acid (73). Clinical studies have shown that monitoring of uric acid levels in urine and blood serum can be used to diagnosis several disorders including hyperuricemia and hypouricemia. Miniaturizing all of the components of microchip CEEC has the possibility of resulting in a portable analysis system that can be used in remote settings, such as small hospitals with limited laboratory facilities. In this study, it was demonstrated that urine could be directly injected after dilution with buffer, and the uric acid content determined in the range of 15 to 110 µM. The limit of detection was found to be 1 µM and the separation was accomplished in less than 30 s (Fig. 8) (73). The measured uric acid concentration was verified with the established uricase reaction. In addition, six urine samples were evaluated with this device and the uric acid concentration for each sample was found to be in the expected clinical concentration range. Overall, this study showed the ability of microchip CEEC to perform sensitive assays on clinical samples with minimal sample pretreatment. Fig. 6. (Opposite page) Indirect measurement of nitric oxide production by monitoring nitrate and nitrite using microchip CEEC with an integrated on-chip carbon electrode in an end-channel alignment. (A) standard nitrite solution; (B) nitrite produced by the on-chip reduction of nitrate; (C) measurement of the time course production of nitrite and nitrate from 3-morpholinosydnonimine, a nitric oxide releasing compound. Symbols: (•) nitrite and (♦) reduction reaction mixture (nitrite and nitrite produced from nitrate). (Reprinted with permission from ref. 35.)

106

Martin

Fig. 7. Competitive electrochemical immunoassay using microchip CEEC with an externally positioned off-chip electrode alignment using a screen printed electrode. (A) Schematic of the on-chip competitive immunoassay. (B) Calibration data for the T3 antigen competitive immunoassay with inset showing the electropherogram for one of the calibration points. See text for details. (Reprinted with permission from ref. 69.)

4. Notes 1. When a potential is first applied to the working electrode, a charging current is exhibited. This current will quickly subside after a few minutes, and stable baseline will be seen. When switching from injection to separation voltages with alignments that result in the working electrode still being exposed to some portion of the electric field, a charging current will be seen at the very beginning of the electropherogram. However, this current will subside within the first few seconds of the separation.

Interfacing Amperometric Detection With Microchip CE

107

Fig. 8. Uric acid clinical assay using microchip CEEC with an externally positioned offchip electrode alignment using a micromanipulated platinum electrode. Electropherogram of a diluted urine sample. (Reprinted with permission from ref. 73.)

2. As with all microchip CE systems, siphoning or pressure-induced flow can be brought about by reservoirs that have different fluid levels (78). Special attention has to be paid with microchip CEEC, as the detection reservoir usually has a much higher fluid capacity than the other reservoirs. Therefore, care should be taken to ensure that all fluid levels are equal. 3. The position of the reference electrode relative to the working electrode affects the working electrode potential (59) and can change the analyte response (26). This can be accomplished by fabricating the reference electrode close to the working electrode (23) or using some type of holder to ensure that this distance is as reproducible as possible. 4. The alignment of the working electrode is very crucial. This distance needs to be made reproducible. This can be done with the aid of a mask aligner, a microscope with calibrated reticule, or using alignment marks on the prebonded chip layers.

References 1. Harrison, D. J., Manz, A., Fan, Z., Luedi, H., and Widmer, H. M. (1992) Capillary electrophoresis and sample injection systems integrated on a planar glass chip. Anal. Chem. 64, 1926–1932.

108

Martin

2. Effenhauser, C. S., Manz, A., and Widmer, H. M. (1993) Glass chips for highspeed capillary electrophoresis separations with submicrometer plate heights. Anal. Chem. 65, 2637–2642. 3. Jacobson, S. C., Hergenroder, R., Koutny, L. B., and Ramsey, J. M. (1994) High speed separations on a chip. Anal. Chem. 66, 1114–1118. 4. Woolley, A. T. and Mathies, R. A. (1994) Ultra-high-speed DNA fragment separations using microfabricated capillary array electrophoresis chips. Proc. Natl. Acad. Sci. USA 91, 11,348–11,352. 5. Cheng, S. B., Skinner, C. D., Taylor, J., et al. (2001) Development of a multichannel microfluidic analysis system employing affinity capillary electrophoresis for immunoassay. Anal. Chem. 73, 1472–1479. 6. Gottschlich, N., Jacobson, S. C., Culbertson, C. T., and Ramsey, J. M. (2001) Twodimensional dlectrochromatography/capillary electrophoresis on a microchip. Anal. Chem. 73, 2669–2674. 7. Emrich, C. A., Tian, H., Medintz, I. L., and Mathies, R. A. (2002) Microfabricated 384-lane capillary array electrophoresis bioanalyzer for ultrahigh-throughput genetic analysis. Anal. Chem. 74, 5076–5083. 8. Schwarz, M. A. and Hauser, P. C. (2001) Recent developments in detection methods for microfabricated devices. Lab Chip 1, 1–6. 9. Oleschuk, R. D. and Harrison, D. J. (2000) Analytical microdevices for mass spectrometry. TrAC 19, 379–388. 10. Salimi-Moosavi, H., Jiang, Y., Lester, L., McKinnon, G., and Harrison, D. J. (2000) A multireflection cell for enhanced absorbance detection in microchipbased capillary electrophoresis devices. Electrophoresis 21, 1291–1299. 11. Mangru, S. D. and Harrison, D. J. (1998) Chemiluminescence detection in integrated post-separation reactors for microchip-based capillary electrophoresis and affinity electrophoresis. Electrophoresis 19, 2301–2307. 12. Arora, A., Eijkel, J. C. T., Morf, W. E., and Manz, A. (2001) A wireless electrochemiluminescence detector applied to direct and indirect detection for electrophoresis on a microfabricated glass device. Anal. Chem. 73, 3282–3288. 13. Swinney, K., Markov, D., and Bornhop, D. J. (2000) Chip-scale universal detection based on backscatter interferometry. Anal. Chem. 72, 2690–2695. 14. Crabtree, H. J., Kopp, M. U., and Manz, A. (1999) Shah convolution fourier transform detection. Anal. Chem. 71, 2130–2138. 15. Wang, J. (2002) Electrochemical detection for microscale analytical systems: a review. Talanta 56, 223–231. 16. Wightman, R. M. (1981) Microvoltammetric electrodes. Anal. Chem. 53, 1125A–1134A. 17. Tanyanyiwa, J., Leuthardt, S., and Hauser, P. C. (2002) Conductimetric and potentiometric detection in conventional and microchip capillary electrophoresis. Electrophoresis 23, 3659–3666. 18. Wang, J. (2000) Analytical Electrochemistry, Wiley-VCH, New York, NY.

Interfacing Amperometric Detection With Microchip CE

109

19. Lunte, S. M., Lunte, C. E., and Kissinger, P. T. (1996) Electrochemical Detection in Liquid Chromatography and Capillary Electrophoresis. In: Laboratory Techniques in Electroanalytical Chemistry, 2nd ed.; (Kissinger, P. T. and Heineman, W. R., eds.), Marcel Dekker, New York, NY, pp. 813–853. 20. Vandaveer IV, W. R., Pasas, S. A., Martin, R. S., and Lunte, S. M. (2002) Recent developments in amperometric detection for microchip capillary electrophoresis. Electrophoresis 23, 3667–3677. 21. Lacher, N. A., Garrison, K. E., Martin, R. S., and Lunte, S. M. (2001) Microchip capillary electrophoresis/electrochemistry. Electrophoresis 22, 2526–2536. 22. Dolnik, V., Liu, S., and Jovanovich, S. (2000) Capillary electrophoresis on microchip. Electrophoresis 21, 41–54. 23. Baldwin, R. P., Roussel, T. J., Crain, M. M., et al. (2002) Fully integrated on-chip electrochemical detection for capillary electrophoresis in a microfabricated device. Anal. Chem. 74, 3690–3697. 24. McDonald, J. C., Duffy, D. C., Anderson, J. R., et al. (2000) Fabrication of microfluidic systems in poly(dimethylsiloxane). Electrophoresis 21, 27–40. 25. Martin, R. S., Gawron, A. J., Lunte, S. M., and Henry, C. S. (2000) Dual-electrode electrochemical detection for poly(dimethylsiloxane)-fabricated capillary electrophoresis microchips. Anal. Chem. 72, 3196–3202. 26. Woolley, A. T., Lao, K., Glazer, A. N., and Mathies, R. A. (1998) Capillary electrophoresis chips with integrated electrochemical detection. Anal. Chem. 70, 684–688. 27. Anderson, J. L., and Winograd, N. (1996) Film Electrodes. In: Laboratory Techniques in Electroanalytical Chemistry, 2nd ed.; (Kissinger, P. T. and Heineman, W. R., eds.), Marcel Dekker, New York, NY, pp. 333–366. 28. Wring, S. A. and Hart, J. P. (1992) Chemically modified, screen-printed carbon electrodes. Analyst 117, 1281–1286. 29. Schwarz, M. A., Galliker, B., Fluri, K., Kappes, T., and Hauser, P. C. (2001) A two-electrode configuration for simplified amperometric detection in a microfabricated electrophoretic separation device. Analyst 126, 147–151. 30. Lapos, J. A., Manica, D. P., and Ewing, A. E. (2002) Dual fluorescence and electrochemical detection on a electrophoresis microchip. Anal. Chem. 74, 3348–3353. 31. Wang, J., Tian, B., and Sahlin, E. (1999) Integrated electrophoresis chips/ amperometric detection with sputtered gold working electrodes. Anal. Chem. 71, 3901–3904. 32. Wang, J., Tian, B., and Sahlin, E. (1999) Micromachined electrophoresis chips with thick-film electrochemical detectors. Anal. Chem. 71, 5436–5440. 33. Gawron, A. J., Martin, R. S., and Lunte, S. M. (2001) Fabrication and evaluation of a carbon-based dual-electrode detector for poly(dimethylsiloxane)electrophoresis. Electrophoresis 22, 242–248. 34. Liu, Y., Fanguy, J. C., Bledsoe, J. M., and Henry, C. S. (2000) Dynamic coating using polyelectrolyte multilayers for chemical control of electroosmotic flow in capillary electrophoresis microchips. Anal. Chem. 72, 5939–5944. 35. Kikura-Hanajiri, R., Martin, R. S., and Lunte, S. M. (2002) Indirect measurement of nitric oxide production by monitoring nitrate and nitrite using microchip electrophoresis with electrochemical detection. Anal. Chem. 74, 6370–6377.

110

Martin

36. Backofen, U., Matysik, F. M., and Lunte, C. E. (2002) A chip-based electrophoresis system with electrochemical detection and hydrodyanmic injection. Anal. Chem. 74, 4054–4059. 37. Jacobson, S. C., Koutny, L. B., Hergenroder, R., Moore, A. W., and Ramsey, J. M. (1994) Effects of injection schemes and column geometry on the performance of microchip electrophoresis devices. Anal. Chem. 66, 1107–1113. 38. Shultz-Lockyear, L. L., Colyer, C. L., Fan, Z. H., Roy, K. I., and Harrison, D. J. (1999) Effects of injector geometry and sample matrix on injection and sample loading in integrated capillary electrophoresis devices. Electrophoresis 20, 529–538. 39. Alarie, J. P., Jacobson, S. C., and Ramsey, J. M. (2001) Electrophoretic injection bias in a microchip valving scheme. Electrophoresis 22, 312–317. 40. Seller, K., Fan, Z. H., Fluri, K., and Harrision, D. J. (1994) Electroosmotic pumping and valveless control of fluid flow within a manifold of capillaries on a glass chip. Anal. Chem. 66, 3485–3491. 41. Jacboson, S. C., Hergenroder, R., Moore, A. W., and Ramsey, J. M. (1994) Precolumn reactions with electrophoretic analysis integrated on a microchip. Anal. Chem. 66, 4127–4132. 42. Jacobson, S. C., Koutny, L. B., Hergenroder, R., Moore, A. W., and Ramsey, J. M. (1994) Microchip capillary electrophoresis with an integrated postcolumn reactor. Anal. Chem. 66, 3472–3476. 43. Zhou, J. and Lunte, S. M. (1995) Direct determination of amino acids by capillary electrophoresis/electrochemistry using a copper microelectrode and zwitterionic buffers. Electrophoresis 16, 498–503. 44. Schwarz, M. A. and Hauser, P. C. (2001) Rapild chiral on-chip separation with simplified amperometric detection. J. Chromatogr. A. 928, 225–232. 45. Dou, Y. H., Bao, N., Xu, J. J., and Chen, H. Y. (2002) A dynamically modified microfluidic poly(dimethylsiloxane) chip with electrochemical detection for biological analysis. Electrophoresis 23, 3558–3566. 46. Manica, D. P. and Ewing, A. E. (2002) Prototyping disposable electrophoresis microchips with electrochemical detection using rapid marker masking and laminar flow etching. Electrophoresis 23, 3735–3743. 47. Gavin, P. F. and Ewing, A. G. (1996) Continuous separations with microfabricated electrophoresis-electrochemical array detection. J. Am. Chem. Soc. 118, 8932–8936. 48. Gavin, P. F. and Ewing, A. G. (1997) Characterization of electrochemical array detection for continuous channel electrophoretic separations in micrometer and submicrometer channels. Anal. Chem. 69, 3838–3845. 49. Wu, C. C., Wu, R. G., Huang, J. G., Lin, Y. C., and Chang, H. C. (2003) Threeelectrode electrochemical detector and platinum film decoupler integrated with a capillary electrophoresis microchip for amperometric detection. Anal. Chem. 75, 947–952. 50. Chen, D. -C., Hsu, F. -L., Zhan, D. -Z., and Chen, C. -H. (2001) Palladium film decoupler for amperometric detection in electrophoresis chips. Anal. Chem. 73, 758–762.

Interfacing Amperometric Detection With Microchip CE

111

51. Hebert, N. E., Kuhr, W. G., and Brazill, S. A. (2002) Microchip capillary electrophoresis coupled to sinusoidal voltammetry for the detection of native carbohydrates. Electrophoresis 23, 3750–3759. 52. Garcia, C. D. and Henry, C. S. (2003) Direct determination of carbohydrates, amino acids, and antibiotics by microchip electrophoresis with pulsed amperometric detection. Anal. Chem. 75, 4778–4783. 53. Martin, R. S., Ratzlaff, K. L., Huynh, B. H., and Lunte, S. M. (2002) In-channel electrochemical detection for microchip capillary electrophoresis using an electrically isolated potentiostat. Anal. Chem. 74, 1136–1143. 54. Martin, R. S., Gawron, A. J., Fogarty, B. A., Regan, F. B., Dempsey, E., and Lunte, S. M. (2001) Carbon paste-based electrochemical detectors for microchip capillary electrophoresis/electrochemistry. Analyst 126, 277–280. 55. Rossier, J. S., Schwarz, A., Reymond, F., Ferrigno, R., Bianchi, F., and Girault, H. H. (1999) Microchannel networks for electrophoretic separations. Electrophoresis 20, 727–731. 56. Rossier, J. S., Ferrigno, R., and Girault, H. H. (2000) Electrophoresis with electrochemical detection in a polymer microdevice. J. Electroanal. Chem. 492, 15–22. 57. Wang, J., Pumera, M., Chatrathi, M. P., et al. (2002) Thick-film electrochemical detectors for poly(dimethylsiloxane)-based microchip capillary electrophoresis. Electroanal. in press. 58. Becker, H. and Gartner, C. (2000) Polymer microfabrication methods for microfluidic analytical applications. Electrophoresis 21, 12–26. 59. Wallenborg, S. R., Nyholm, L., and Lunte, C. E. (1999) End-column amperometric detection in carpillary electrophoresis: influence of the separation-related parameters on the observed half-wave potential for dopamine and catechol. Anal. Chem. 71, 544–549. 60. Goss, C. A., Charych, D. H., and Majda, M. (1991) Application of (3-mercaptopropyl)trimethoxysilane as a molecular adhesive in the fabrication of vapordeposited gold electrodes on glass substrates. Anal. Chem. 63, 85–88. 61. Lacher, N. A., Lunte, S. M., and Martin, R. S. (2004) Development of a microfabricated palladium decoupler/electrochemical detector for microchip capillary electrophoresis using a hybrid glass/poly(dimethylsiloxane) device. Anal. Chem. 76, 2482–2491. 62. Wang, J., Tian, B., Nascimento, V. B., and Angnes, L. (1998) Performance of screenprinted carbon electrodes fabricated from different carbon inks. Electrochimica Acta 43, 3459–3465. 63. Wang, J., Ibanez, A., Chatrathi, M. P., and Escarpa, A. (2001) Electrochemical enzyme immunoassays on microchip platforms. Anal. Chem. 73, 5323–5327. 64. Wang, J., Chatrathi, M. P., and Tian, B. (2000) Capillary electrophoresis microchips with thick-film amperometric detectors: separation and detection of phenolic compounds. Anal. Chim. Acta 416, 9–14. 65. Wang, J., Chatrathi, M. P., and Tian, B. (2000) Micromachined separation chips with a precolumn reactor and end-column electrochemical detector. Anal. Chem. 72, 5774–5778.

112

Martin

66. Wang, J., Chatrathi, M. P., Tian, B., and Polsky, R. (2000) Microfabricated electrophoresis chips for simultaneous bioassays of glucose, uric acid, ascorbic acid, and acetaminophen. Anal. Chem. 72, 2514–2518. 67. Wang, J. and Chatrathi, M. P. (2003) Microfabricated electrophoresis chip for bioassay of renal markers. Anal. Chem. 75, 525–529. 68. Wang, J., Chatrathi, M. P., Ibanez, A., and Escarpa, A. (2002) Micromachined separation chips with post-column enzymatic reactions of “class” enzymes and end-column electrochemical detection: assays of amino assays. Electroanal. 14, 400–404. 69. Wang, J., Ibanez, A., and Chatrathi, M. P. (2002) Microchip-based amperometric immunoassays using redox tracers. Electrophoresis 23, 3744–3749. 70. Wang, J., Chatrathi, M. P., Tian, B., and Polsky, R. (2000) Capillary electrophoresis chips with thick-film amperometric detectors: separation and detection of hydrazine compounds. Electroanal. 12, 691–694. 71. Lunte, S. M., Martin, R. S., and Lunte, C. E. (2002) Capillary Electrophoresis/ Electrochemistry. In: Electroanalytical Methods for Biological Materials, (Brajter-Toth, A. and Chambers, J. Q., eds.), Marcel Dekker: New York, NY, pp. 461–490. 72. Fanguy, J. C. and Henry, C. S. (2002) Pulsed amperometric detection of carbohydrates on a electrophoretic microchip. Analyst 127, 1021–1023. 73. Fanguy, J. C. and Henry, C. S. (2002) The analysis of uric acid in urine using microchip capillary electrophoresis with electrochemical detection. Electrophoresis 23, 767–773. 74. Hilmi, A. and Luong, J. H. T. (2000) Micromachined electrophoresis chips with electrochemical detectors for analysis of explosive compounds in soil and groundwater. Environ. Sci. Technol. 34, 3046–3050. 75. Zeng, Y., Chen, H., Pang, D. W., Wang, Z. L., and Cheng, J. K. (2002) Microchip capillary electrophoresis with electrochemical detection. Anal. Chem. 74, 2441–2445. 76. Hilmi, A. and Luong, J. H. T. (2000) Electrochemical detectors prepared by electroless deposition for microfabricated electrophoresis chips. Anal. Chem. 72, 4677–4682. 77. Wang, J., Chatrathi, M. P., and Ibanez, A. (2001) Microseparation chips for performing multi-enzymatic dehydrogenase/oxidase assays. Anal. Chem. 73, 1296–1300. 78. Crabtree, H. J., Cheong, E. C. S., Tilroe, D. A., and Backhuse, C. J. (2001) Microchip injection and separation anomalies due to pressure effects. Anal. Chem. 73, 4079–4086.

9 Conductivity Detection on Microchips Roland Hergenröder and Benedikt Graß Summary Conductivity detection as a versatile detection technique for chip-based electrophoretic separation methods is described and the basic principles are discussed. The necessary electronic equipment and technologies to implement the detection electrodes on a microchip are presented. A difference between contact and contactless detection is made and the resulting advantages and problems are shown. Different analytical applications are listed and discussed that show the basic applicability of conductivity detection in capillary electrophoresis and in isotachophoresis, which may serve as a guideline for more specific developments in the future. Central to all applications is the choice of the buffer system. It determines not only the separation capability but also puts limits on the achievable dynamic detection range. Key Words: Isotachophoresis; capillary electrophoresis; conductivity; microchip; thin film; electrodes.

1. Introduction Conventional capillary electrophoresis (CE) is dominated by optical detection methods such as ultraviolet (UV) absorption or laser-induced fluorescence. For different reasons, recent development of on-chip CE has brought renewed interest in conductivity detection. This is owing to the fact that the reduction of the dimensions of the separation-channel on microdevices considerably shortens the optical light path. Under these circumstances conductivity detection as an alternative to UV-absorbance becomes increasingly interesting. A second reason may be found in difficulty to implement electrochemical methods in conventional CE. With the advent of microchips this point has changed considerably. Now microstructuring technologies and thin film techniques enable totally new approaches to the problem. Compared with conventional conductivity detection in a miniaturized electrode system, higher sensitivity, reduced noise, and faster response times can be expected. This is the result of an increased current density and better control possibilities over parasitic capacities. From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

113

114

Hergenröder and Graß

But, it should also be noted that the use of microchip technology does not change anything about the principal problems conductivity detection faces. Still, a low conductivity buffer system is necessary to achieve maximum dynamic range, which is an additional optimization and restriction parameter that has to be accounted for. This is totally different for isotachophoresis (ITP), which be discussed in Subheading 3.3. Therefore, in this case conductivity detection is a natural choice. The following chapters are dedicated to different aspects of conductivity detection on microchips. Currently, no commercial instrumentation is available in this area. Therefore, a brief description of a typical electrode manufacturing process, together with some downscaling aspects will be given (Subheading 3.1.). A short discussion of electronic details will conclude the section (Subheading 3.2.). Conductivity detection has not yet been studied systematically as a detection method for microchips. Nonetheless, the broad range of analytes discussed in the second section of the chapter that have been separated and detected on microchips with conductivity detection demonstrate the potential behind it and will serve as a guideline for further applications. 2. Materials 2.1. Typical Low Conductivity Electrolytes in CE 1. Separation of cations: 2-(N-Morpholino)-ethanesulfonic acid (MES) in combination with L-Histidine, or 2-hydroxyisobutyric acid (HIBA). 2. Separation of anions: MES. 3. 2-(N-cyclohexylamino)-ethanesulfonic acid (CHES). 4. CHES/arginine (CHES/Arg). 5. 2-(N-cyclohexylamino)-1-propanesulfonic acid/arginine (CAPS/Arg). 6. 3-(N-morpholino)-2-hydroxypropanesulfonic acid/arginine (MOPSO)/Arg. 7. 3-(N-tris[hydroxymethyl]methylamino)-2-hydroxypropanesulfonic acid/arginine (TAPSO/Arg).

2.2. Electroosmotic Flow Modifiers 1. Cetyltrimethylammonium bromide (CTAB). 2. Tetradecyl trimethylammonium bromide (TTAB). 3. Sodium dodecyl sulfate (SDS).

3. Methods 3.1. Electrode Manufacturing The detection principle is based on the electrical properties of ions. A common detector consists of two metal electrodes, which test the conductivity of the analyte solution. An alternating potential is applied across the electrodes. Under the presence of the electric field, the cations migrate during one cycle toward one electrode, and the anions toward the other. The corresponding

Conductivity Detection on Microchips

115

Fig. 1. A typical electrode and fluidic structure before (A) and after the bonding (B). Besides electrodes for the electrochemical detection, there are also electrodes for separation voltage. After bonding, these electrodes are directly under the reservoirs.

current is measured and related to the conductivity of the solution through the value of the equivalent measured resistance, R. Conductivity detection can be classified according to different criteria. From the standpoint of manufacturing, a classification into electrodes that are in galvanic contact and electrodes without galvanic contact to the analyte seems to be appropriate. Later, it will become clear that this classification naturally corresponds in most cases to a two- and four-electrode classification scheme. Metal deposition, and thus microelectrode fabrication, is well known in microfabrication technology. As a result, electrodes can easily be incorporated onto a microdevice along with the fluidic structure (Fig. 1). A typical standard metal deposition process consists of three different steps. In a first step a photolithographic structure has to be created on a substrate. Therefore, the substrate is coated with a photoresist and exposed with an UV-lamp through a photomask. Typically, this is an e-beam written mask (see Note 1). After exposure and

116

Hergenröder and Graß

development the metallization starts (see Note 2). Because of the higher kinetic ion energy the adhesion of a sputtered metal film is much higher compared with a vapor-deposited layer. Therefore, sputtering should always be preferred over vapor deposition. After stripping of the photoresist, the result will be a thin layer electrode (~200 nm) on the substrate. But even the adhesion of platinum or gold films, which are usually used because of their electrochemical inertness, on glass or plastic is unsatisfactory and an extra adhesion layer is necessary. A well known and characterized material for an adhesion layer in thin-film technology is chromium (layer thickness: ~10 nm). However, pure chromium is prone to be dissolved in electrolyte solutions especially in addition with high voltages. Under typical separation conditions for CE this can result in severe damage of the electrodes. A way to prevent this is to oxidize the chromium during the sputter process (see Note 3). Under optimized conditions, the adhesion of the chromium dioxide will be adequate and the damage of the electrodes will be impeded. It is clear that the bump an electrode constitutes on the surface causes problems during the sealing process when the cover slip with the electrodes is adjusted to the structured part of the chip that contains the fluidic manifold. Up to now, only polymeric material has been shown to be flexible enough for a tight sealing. For glass chips, either end-on detectors have been constructed or a relatively complicated procedure of embedded electrodes has been developed. In this procedure, two trenches are etched into the glass wafer, filled with aluminium, and sealed with silicon nitride. Through isotropic etching in an inductively coupled plasma etcher, a planar surface that enables leak-free bonding is produced. In this way a four-electrode system for nongalvanic contact conductivity detection has been produced and tested (1). The completed electrode substrate has to be bonded to the chip. Here an alignment problem appears. A typical separation channel is approx 50 µm. The electrodes have to be positioned relative to it in a reproducible manner. Two electrode designs are feasible in this situation (Fig. 2). The face-to-face design (Fig. 2A,B) has the advantage that both electrodes see the same potential and no spurious extra peaks appear if any switching occurs (e.g., for sample injection, the end electrodes of the fluidic structure changes the electric potential in the detection channel). On the other hand, this electrode configuration is rather sensitive to misalignments. A small tilting exposes the electrodes to different positions along the channel and the advantage and chip-to-chip reproducibility is lost. An electrode arrangement where the electrodes are along the channel is much easier to align (Fig. 2C), but now the electrodes always see a potential difference and will detect changes in it that can give spurious extra peaks and modulations in the base line, but in a chip-to-chip reproducible manner. As an alternative to the previously described advanced thin-film technologies, there are also attempts to produce pure laboratory equipment. It usually

Conductivity Detection on Microchips

117

Fig. 2. Three different electrode designs for conductivity detection. (A) The electrode width is 12 µm and the spacing is 76 µm. Alignment is rather difficult in this case. Resolution and sensitivity is good. (B) The electrode width is 49 µm and the spacing is 77 µm. Obviously, resolution is lower as in case A. But alignment is much easier and fairly reproducible. (C) Electrode width is 22 µm and spacing between the electrodes is 24 µm. Alignment is no problem, resolution and sensitivity are comparable to case B.

118

Hergenröder and Graß

consists of two micrometer-sized wires that are brought in close contact with the chip. Usually the measurement cell is easily produced but has deficits in (chip-to-chip) reproducibility and sensitivity. A very simple fabrication procedure for contactless conductivity detection was presented from Pumera et al. (2). Pumera and colleagues put thin aluminum foil strips on the outside of a poly(methylmethacrylate) (PMMA) microchip and used them for the measurement of the impedance of the solution in the channels. With this device Pumera et al. have analyzed various inorganic cations and anions in 20 mM MES/histidine (2,3). The detection limits were found to be 2.8 and 6.4 µM for potassium and chloride, respectively. Linear response between 20 µM and 7 mM could be achieved for lithium and fluoride. 3.2. Detector Electronics One of the difficulties in on-chip conductivity detection is that the measurement is conducted at relatively high electrical potentials. Therefore, the conductivity measurement has to be decoupled from ground. In the case of galvanic contact, this is usually achieved by an inductive decoupler. A typical electric circuit is shown in Fig. 3. The modulated current is 1:1 transformed over an inductive coil. This is straight-forward and standard. Because thin-film electrodes are embedded, some caution has to be taken to match the impedance. In the nongalvanic contact case the decoupling is achieved by a dielectric layer that separates the analyte from the electrode. The direct faradic current is not measured, but the impedance change is monitored. This extra capacitor that is created through this multi-layer system deteriorates the detector response if the conductivity is measured with only two electrodes. Therefore, a four-electrode system may be preferable. It has been shown that the linearity, accuracy, and sensitivity are increased compared with a noncontact two-electrode system (4). But two aspects have to be considered. The electronics are more sophisticated and have to be optimized for a special electrode system and the dimensions of a fourelectrode system are inherently larger than a two-electrode system, which adds to the band broadening. In some high speed, high-resolution separations this may be adversarial. 3.3. Applications 3.3.1. Buffer Solutions To obtain good resolution and the best sensitivity when conductivity detection is employed for electrophoretic separations, the composition of the buffer solutions has to be chosen very carefully. Best results can be achieved with a maximum difference in conductance between carrier electrolytes and analytes. Therefore for CE, background electrolytes with low equivalent conductivity at considerably high concentrations are preferable. A high conductivity of the carrier electrolyte would

119 Fig. 3. Typical electronic circuit to decouple the conductivity measurement from high separation potentials.

120

Hergenröder and Graß

have a negative influence on the detection, because very small changes in conductivity would have to be measured on a high conductivity background. Typical low conductivity background electrolytes for the separation of cations are MES in combination with L-Histidine or HIBA. For the separation of anions a variety of carrier electrolytes can be employed, as there are MES, CHES, CHES/Arg, CAPS/Ar, MOPSO, TAPSO/Arg. The surfactants, CTAB or TTAB, are used as electroosmotic flow modifiers (5). In ITP the buffer system has to be chosen in dependence of the sample composition. The difference in conductance between leading and terminating buffer has to be as large as possible, with the conductance of the analytes lying in between. Because of its high effective mobility, the typical leading ion for the analysis of anions is the chloride ion in concentrations between 5 and 10 mM. Co-ions, like the amino acids β-alanine and histidine, are used for adjustment of the pH. If chloride has to be analyzed, the dithionate ion can be employed as a leading ion (6). Co-counter ions like bis-trispropane (BTP) or the magnesium cation Mg2+ can be added as complexing reagents. To suppress the EOF, during the ITP separation methylhydroxyethylcellulose (MHEC) is used as additive (7,8). As terminating ion with the lowest mobility in the system, weak organic acids like glutamate, capronate, aspartate, or citrate are used depending on the analytes. The pH of the terminating buffer is adjusted with counterions like histidine or β-alanine. For ITP separations of cations with ITP buffer systems with potassium acetate or acetic acid as the leading electrolyte, γ–butyric acid, acetic acid, or β-alanine as the terminating electrolytes are employed (8). 3.3.2. Analytes For ITP, conductivity detection is the method of choice because the separation results in zones with different conductivity. In conventional capillary zone electrophoresis (CZE) conductivity detection has only seldom been used (9,10). Therefore, it is not surprising that the first microchips in combination with conductivity detection were developed for ITP separations (11–13). Because of their high equivalent conductivity, inorganic ions are well suited for the determination with conductivity detection, and they are very often used to characterize different detector designs on miniaturized devices and to demonstrate the performance of direct and contactless conductivity detection. Typical analytes are anions such as chloride, fluoride, phosphate, nitrate, and nitrite or the alkali cations sodium, lithium, and potassium, and transition metal and heavy metal cations. Other compounds often analyzed with conductivity detection are small organic acids like tartaric, lactic, malic, citric, or fumaric acid. Prest et al. (13) separated mixtures of sodium and potassium on a polymer microchip made from poly-(dimethylsiloxane) (PDMS). The detector consisted of a single 25-µm diameter platinum-iridium wire that was placed between two

Conductivity Detection on Microchips

121

polymer pieces before they were bonded. The signal derives from changes in the potential measured at one point (potential gradient detector) (see Note 4). The buffer system for the separation consists of chloride as the leading ion with hydroxyethyl cellulose (HEC) to suppress any electroosmotic flow (EOF) and carnitine hydrochloride as the terminating ion. The same detector was used for the separation of several lanthanide cations (14). Here, the leading electrolyte was sodium hydroxide with HIBA as the complexing agent and the terminating ion was again carnitine hydrochloride. The concentrations of the samples were in the low millimolar range. The same group showed the separation of small inorganic cations and anions (ammonium, sodium, magnesium, calcium, lithium, nitrate, chlorate, sulfate, fluoride) at the same time with bidirectional ITP (15). The concentration of the analytes was 0.01 M each. The anolyte consists of 10 mM chloride with 3 mM BTP as the complexing agent and HEC as EOF suppressor. The pH was adjusted to 3.6 with 40 mM glycylglycine. The caesium cation was used as catholyte (25 mM Cs+) with 7 mM 18-crown-6 and 4 mM N-(2-acetamido)iminodiacetic acid (ADA) as complexing agents and HEC as additive. The pH of the catholyte was adjusted to 4.7 with 50 mM acetic acid. The detector used for this separation was built by inserting a platinum wire into a PMMA substrate before the separation channel was milled. During the milling the wire was cut and the result were two opposed detector electrodes. The same detector has been used for the speciation of arsenic compounds (16). Good linearity could be achieved for arsenic (V) between 0.5 and 10 mg/L and arsenic (III) between 2.5 and 50 mg/L, respectively. The limits of detection were 1.8 mg/L for arsenic (V) and 4.8 mg/L for arsenic (III). This corresponds to an absolute amount of 330 pmol arsenic (III) and 130 pmol arsenic (V) that could be analyzed. With the employed buffer system consisting of 8 mM chloride, 10 mM cyclodextrin (complexing agent), adjusted with tris(hydroxymethyl)aminomethane (TRIS) to pH 9.0 and HEC (EOF suppressor) as the leading electrolyte and 10 mM glycine adjusted with barium hydroxide to pH 9.5 as the terminating electrolyte, the simultaneous separation of arsenic (V) and arsenic (III), antimony (III), molybdenum (VI), and tellurium (IV) could be demonstrated. Another approach was introduced from Baldock et al. (17) with the use of molded polymer electrodes, integrated into molded polystyrene and Zeonor substrates. The capability of the different microchips and the performance of the detector electrodes were demonstrated with three anionic dyes, inorganic anions, and different metal cations. A mixture of eight alkaline, earth, transition, and lanthanide metal cations (magnesium [1 mM], calcium, manganese, nickel, zinc, lanthanum, neodymium, gadolinium [each 0.8 mM]), could be separated. The leading electrolyte was 20 mM sodium hydroxide, with 15 mM HIBA and HEC adjusted with propionic acid to pH 4.95 and the terminating electrolyte

122

Hergenröder and Graß

was 10 mM carnitine hydrochloride. Grass et al. and Kaniansky and co-workers presented PMMA microchips for ITP with sputtered thin-film platinum electrodes for conductivity detection (18,19). These chips were used for one and two-dimensional ITP separations, as well as for the coupling of ITP and CZE. Several applications have been published for this miniaturized system for different inorganic anions and cations in water samples (18,20,21), organic acids in water, wine, fruit juice, and urine (18,19,22–25), enantiomeric separations of amino acids (26), anions in food additives (27), and the separation of seleno amino acids (28). In general, the buffer systems consist of chloride as leading ion, adjusted to the required pH with β-alanine and histidine. Terminating ions were glutamate, capronate, aspartate, or citrate. The concentration of the analyzed samples varied between 10 µM and 1 mM. For some compounds, e.g., bromate in drinking water (21), limits of detection down to 20 nM could be detected with the thin-film detection electrodes. One of the first microchips for CE equipped with conductivity detection was presented from Weber et al. (29). They demonstrated the determination of inorganic cations (potassium, sodium, lithium), anions (chloride, fluoride), and organic acids (oxalic, tartaric, succinic, acetic, lactic, glutamic acid) on a PMMA microchip with sputtered thin-film electrodes for direct and contactless conductivity detection. The carrier electrolyte was 20 mM MOPSO/20 mM histidine at pH 6.4. For organic acids they achieved absolute detection limits of 12 fmol for direct detection and 25 fmol for contactless detection. Since then, several microchips for CE employing direct or contactless conductivity detection were introduced from different groups. The capability of the detectors was in general demonstrated with separations of the inorganic cations potassium, sodium, and lithium. Lichtenberg et al. (30) manufactured a glass microchip with in-plane electrodes for contactless conductivity detection. Tanyanyiwa et al. (31) achieved detection limits under 1 µM for the determination of potassium, sodium, and magnesium and between 2 and 7 µM for manganese, zinc, and chromium (III) with a high-voltage capacitively coupled contactless conductivity detector. The carrier electrolyte consisted of 10 mM MES/10 mM histidine with 18-crown-6 or HIBA. Though the main focus lies on the determination of small inorganic species, biomolecules have also been determined with conductivity detection in various background electrolytes. The integrated conductivity detector described by Galloway et al. was used to detect separations of organic mono- and polyanionic biomolecules (32,33). A pair of polished platinum wires inserted into guide channels embossed in the substrate served as detection electrodes. The wires were situated in the fluid channel with an end-to-end spacing of approx 20 µm. Amino acids, peptides, proteins, and oligonucleotides could be

Conductivity Detection on Microchips

123

detected with this detector after having been separated with free solution zone electrophoresis, micellar electrokinetic chromatography (MEKC), and capillary electrochromatography (CEC) (32). The amino acids were separated with zone electrophoresis in a carrier electrolyte consisting of 10 mM triethylammonium acetate (TEAA). For alanine the calibration plot was linear in a concentration range from 10 to 100 nM. The concentration detection limit was 8 nM, which corresponds to an absolute amount of 3.4 amol alanine. With MEKC, baseline resolution could be achieved for a mixture of eight different proteins. Tris-HCl at pH 9.2 containing SDS was used as background electrolyte. The separation of DNA fragments was achieved by means of CEC in a C18-modified PMMA channel. The mobile phase consisted of 25% acetonitrile, 75% aqueous phase, and 50 mM TEAA as ion pairing agent. The same detector was also used for the detection of polymerase chain reaction products separated with ion-pair microcapillary electrochromatography (33). The mobile phase was again acetonitrile/water (15:85) with 25 mM TEAA, pH 7.4. The mass detection sensitivities in this investigation were found to be in the range of 10–21 mol. Recently, Zuborova at al. (34) reported the separation of a standard protein mixture (cytochrome c, avidin, conalbumin, human hemoglobin, and trypsin inhibitor) on a PMMA chip with sputtered platinum thin-film electrodes for conductivity detection. The carrier electrolyte consists of 100 mM acetic acid and methylhydroxethylcellulose as EOF suppressor. Guijt et al. (35) reported the determination of short chain peptides with an on-chip four-electrode capacitively coupled conductivity detector. Two peptides, each with a concentration of 1 mM, were separated within 60 s in 50 mM phosphate buffer (pH 2.5) with 2 mM SDS as surfactant. With a similar background electrolyte (SDS-modified phosphate buffer), Deyl et al. (36) showed the separation and detection of SDS-complexes of a model set of proteins (cytochrome c, albumin, catalase, transferrin, chymotrypsinogen A) on corundum-based microchips. Besides these applications for biomolecules, conductivity detection recently has also been employed for the determination of drugs. Tanyanyiwa et al. (31) have shown the determination of the anti-inflammatory nonsteroid drugs 4-acetamidophenol, ibuprofen, and salicylic acid each with a concentration of 100 µM with high-voltage capacitively coupled contactless conductivity detection. The buffer system 10 mM CAPS/arginine at pH 10.0 was used as carrier electrolyte. In conclusion, it can be said that for microcapillary separations conductivity detection is mainly employed for the determination of inorganic species. Low conductivity background electrolytes like MES/His are used for the separations. These applications are well investigated and can be used for the characterization of new channel geometries and new detector designs. However, a lot

124

Hergenröder and Graß

of research has been done, especially for the contactless conductivity detection, to open the way for other more sophisticated applications like the determination of biomolecules. 4. Notes 1. If the electrode dimensions are not too small (>50 µm) a transparency printed with a high resolution laser printer may do it. It is much cheaper and turn-around-times are considerably shorter. 2. To improve the metal/substrate adhesion, the whole process should start with a cleaning step. Ion etching or back sputtering would be typical processes which clean the surface and increase the roughness and, therefore, improve the adhesion. 3. This can be done by adding some percentage of oxygen to the sputter gas flux. It is difficult to tell precisely how much oxygen will be necessary because it depends on the actual sputter system and an optimization for the sputter system under consideration will be necessary. 4. This detection method utilizes the driving current of the separation for measuring changes of the conductivity in the solution. If the conductivity in the detection region varies owing to different sample zones passing by, the voltage drop across the separation channel is not uniform. The advantage is that these changes can be monitored without additional application of a measuring signal to the detector electrode. In ITP this detector design leads to signals with different gradients for each zone instead of to the characteristic flat steps.

References 1. Guijt, R. M., Baltussen, E., van der Steen, G., et al. (2001) Capillary electrophoresis with on-chip four-electrode capacitively coupled conductivity detection for application in bioanalysis. Electrophoresis 22, 2537–2541. 2. Pumera, M., Wang, J., Opekar, F., et al. (2002) Contactless conductivity detector for microchip capillary electrophoresis. Anal. Chem. 74, 1968–1971. 3. Wang, J., Pumera, M., Collins, G., Opekar, F., and Jelinek, I. (2002) A chip-based capillary electrophoresis-contactless conductivity microsystem for fast measurements of low-explosive ionic components. Analyst 127, 719–723. 4. Laugere, F., Lubking, G. W., Bastemeijer, J., Bossche, A., and Vellekoop, M. J. (2002) Design of an electronic interface for capacitively coupled four-electrode conductivity detection in capillary electrophoresis microchip. Sens. Actuators B Chem. 83, 104–108. 5. Engelhardt, H., Beck, W., and Schmitt, Th. (1994) Kapillarelektrophorese, Vieweg, Braunschweig/Wiesbaden. 6. Meissner, Th., Eisenbeiss, F., and Jastorff, B. (1999) New leading electrolyte for the direct determination of chloride and other anions in analytical isotachophoresis. J. Chromatogr. A 838, 81–88. 7. Everaerts, F. M., Beckers, J. L., and Verheggen, Th. P. E. M. (1976) Isotachophoresis: Theory, Instrumentation and Applications, Elsevier, Amsterdam. 8. Bocek, P., Deml, M., and Gebauer, P. (1988) Analytical Isotachophoresis, VCH, Weinheim.

Conductivity Detection on Microchips

125

9. Tanyanyiwa, J., Leuthardt, S., and Hauser, P. C. (2002) Conductimetric and potentiometric detection in conventional and microchip capillary electrophoresis. Electrophoresis 23, 3659–3666. 10. Zemann, A. J. (2001) Conductivity detection in capillary electrophoresis. Trends Anal. Chem. 20, 346–354. 11. Fielden, P. R., Baldock, S. J., Goddard, N. J., et al. (1998) A miniaturized planar isotachophoresis separation device for transition metals with integrated conductivity detection. In: Micro Total Analysis Systems (Harrison, D. J. and van den Berg, A., eds.), Kluwer Academic Publishers, Dordrecht, pp. 323–326. 12. Baldock, S. J., Bektas, N., Fielden, P. R., et al. (1998) Isotachophoresis on planar polymeric substrates. In: Micro Total Analysis Systems (Harrison, D. J. and van den Berg, A., eds.), Kluwer Academic Publishers, Dordrecht, pp. 359–362. 13. Prest, J. E., Baldock, S. J., Bektas, N., Fielden, P. R., and Brown, B. J. T. (1999) Single electrode conductivity detection for electrophoretic separation systems. J. Chromatogr. A 836, 59–65. 14. Prest, J. E., Baldock, S. J., Fielden, P. R., and Brown, B. J. T. (2001) Determination of metal cations on miniaturized planar polymeric separation devices using isotachophoresis with integrated conductivity detection. Analyst 126, 433–437. 15. Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., and Treves Brown, B. J. (2002) Bidirectional isotachophoresis on a planar chip with integrated conductivity detection. Analyst 127, 1413–1419. 16. Prest, J. E., Baldock, S. J., Fielden, P. R., Goddard, N. J., and Treves Brown, B. J. (2003) Miniaturised isotachophoretic analysis of inorganic arsenic speciation using a planar polymer chip with integrated conductivity detection. J. Chromatogr. A 990, 325–334. 17. Baldock, S. J., Fielden, P. R., Goddard, N. J., Prest, J. E., and Treves Brown, B. J. (2003) Integrated moulded polymer electrodes for performing conductivity detection on isotachophoresis microchips. J. Chromatogr. A 990, 11–22. 18. Kaniansky, D., Masár, M., Bielcikova, J., et al. (2000) Capillary electrophoresis separationson a planar chip with the column-coupling configuration of the separation channels. Anal. Chem. 72, 3596–3604. 19. Grass, B., Neyer, A., Jöhnck, M., et al. (2001) A new PMMA-microchip device for isotachophoresiswith integrated conductivity detection. Sens. Actuators B Chem. 72, 249–258. 20. Bodor, R., Madajová, V., Kaniansky, D., Masár, M., Jöhnck, M., and Stanislawski, B. (2001) Isotachophoresis and isotachophoresis: zone electrophoresis separations of inorganic anions present in water samples on a planar chip with column-coupling separation channels and conductivity detection. J. Chromatogr. A 916, 155–165. 21. Bodor, R., Kaniansky, D., Masár, M., Silleova, K., and Stanislawski, B. (2002) Determination of bromate in drinking water by zone electrophoresis-isotachophoresis on a column-coupling chip with conductivity detection. Electrophoresis 23, 3630–3637. 22. Grass, B., Siepe, D., Neyer, A., and Hergenröder, R. (2001) Comparison of different conductivity detector geometries on an isotachophoresis PMMA-microchip. Fresenius J. Anal. Chem. 371, 228–233.

126

Hergenröder and Graß

23. Masár, M., Zuborova, M., Bielcikova, J., Kaniansky, D., Joehnck, M., and Stanislawski, B. (2001) Conductivity detection and quantitation of isotachophoretic analytes on a planar chip with on-line coupled separation channels. J. Chromatogr. A 916, 101–111. 24. Masár, M., Kaniansky, D., Bodor, R., Jöhnck, M., and Stanislawski, B. (2001) Determination of organic acids and inorganic anions in wine by isotachophoresis on a planar chip. J. Chromatogr. A 916, 167–174. 25. Zuborova, M., Masár, M., Kaniansky, D., Joehnck, M., and Stanislawski, B. (2002) Determination of oxalate in urine by zone electrophoresis on a chip with conductivity detection. Electrophoresis 23, 774–781. 26. Ölvecka, E., Masár, M., Kaniansky, D., Joehnck, M., and Stanislawski, B. (2001) Isotachophoresis separations of enantiomers on a planar chip with coupled separation channels. Electrophoresis 22, 3347–3353. 27. Bodor, R., Zuborova, M., Olvecka, E., et al. (2001) Isotachophoresis and isotachophoresis-zone electrophoresis of food additives on a chip with columncoupling separation channels. J. Sep. Sci. 24, 802–809. 28. Grass, B., Hergenröder, R., Neyer, A., and Siepe, D. (2002) Determination of seleno amino acids by coupling of isotachophoresis/capillary electrophoresis on a PMMA microchip. J. Sep. Sci. 25, 135–140. 29. Weber, G., Jöhnck, M., Siepe, D., Neyer, A., and Hergenröder, R. (2000) Capillary electrophoresis with direct and contactless conductivity detection on a polymer microchip. In: Micro Total Analysis Systems (Harrison, D. J. and van den Berg, A., eds.), Kluwer Academic Publishers, Dordrecht, pp. 383–386. 30. Lichtenberg, J., de Rooij, N. F., and Verpoorte, E. (2002) A microchip electrophoresis system with integrated in-plane electrodes for contactless conductivity detection. Electrophoresis 23, 3769–3780. 31. Tanyanyiwa, J. and Hauser, P. C. (2002) High-voltage capacitively coupled contactless conductivity detection for microchip capillary electrophoresis. Anal. Chem. 74, 6378–6382. 32. Galloway, M., Stryjewski, W., Henry, A., et al. (2002) Contact conductivity detection in poly(methylmethacrylate)-based microfluidic devices for analysis of mono- and polyanionic molecules. Anal. Chem. 74, 2407–2415. 33. Galloway, M. and Soper, St. A. (2002) Contact conductivity detection of polymerase chain reaction products analyzed by reverse-phase ion pair microcapillary electrochromatography. Electrophoresis 23, 3760–3768. 34. Zuborova, M., Demianova, Z., Kaniansky, D., Masár, M., and Stanislawski, B. (2003) Zone electrophoresis of protein on a poly(methyl methacrylate) chip with conductivity detection. J. Chromatogr. A 990, 179–188. 35. Guijt, R. M., Baltussen, E., van der Steen, G., et al. (2001) Capillary electrophoresis with on-chip four-electrode capacitively coupled conductivity detection for application in bioanalysis. Electrophoresis 22, 2537–2541. 36. Deyl, Z., Miksik, I., and Eckhardt, A. (2003) Comparison of standard capillary and chip separations of sodium dodecylsulfate-protein complexes. J. Chromatogr. A 990, 153–158.

IV APPLICATIONS OF MICROCHIP CAPILLARY ELECTROPHORESIS

10 DNA Separations Andrea W. Chow Summary The use of two types of commercialized microfluidic chips for separation of double-stranded DNA (dsDNA), suitable for personal scale and high throughput use, is described. Compared with conventional approaches such as slab-gel and capillary electrophoresis (CE), these devices offer the advantages of faster separation times, better data reproducibility, greater ease of use, labor savings in quatitative analysis, and ease in data archiving and data sharing owing to the digital data format. With some simple precautions taken in keeping bubbles and particulates out of the microchannels, Lab-on-a-Chip devices have been adopted by many researchers in molecular biology and genomics laboratories to increase their productivity. Key Words: DNA separation; DNA sizing; microfluidics; Lab-on-a-Chip; LabChip® devices; sipper chip; gel electrophoresis; genomics.

1. Introduction Double-stranded DNA (dsDNA) was one of the first analytes used to demonstrate the feasibility of fast, high-resolution separations in microfabricated devices (1). It was also the first application introduced commercially in a separation-based microfluidics platform in 1999. The methods and materials focused on in this chapter will be based on recent chip-based commercial products that can be accessible to many laboratories and used for common molecular biology applications, such as analysis of polymerase chain reaction (PCR) products, DNA restriction digest fingerprinting, and plasmid clone identification. However, many more methods have been developed and reported in the literature for other applications including DNA sequencing (2) and DNA separation by microlithographic arrays (3) in microfabricated devices. It is out of the scope of this chapter to discuss them in details since the devices and experimentation methods are only available in very few research laboratories with prototype devices and instrumentation. From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

129

130

Chow

Electrophoretic separation of DNA has been a subject of intensive studies for over half a century because DNA is the remarkable biomolecule that stores all the genetic information for living organisms. Separation of DNA by size has played a key role in genome mapping, and will likely be important in many more genomic applications, such as in molecular diagnostics. Because DNA can be made readily available in large quantities with monodisperse length and exact same sequence using PCR technology, it is also an ideal model for studying the fundamentals of migration mechanisms in gel electrophoresis in analytical chemistry and polymer physics. In physiological pH, DNA is negatively charged. In free solution electrophoresis, dsDNA molecules in sizes greater than approx 200 bp have similar electrophoretic mobility and therefore no separation by size can be accomplished (4). In a polymer matrix such as a chemically cross-linked gel or a polymer solution, however, DNA molecules of different lengths interact hydrodynamically with the matrix differentially as they travel under the influence of an electric field, resulting in smaller DNA fragments migrating faster than larger fragments and, thus, yielding separation by size. Several excellent reviews of experimental results and theory of DNA migration in cross-linked and entangled polymer matrices have been published recently (5–7). The evolution of laboratory instrumentation for gel electrophoresis has been relatively gradual over many decades. The use of slab gels has been a standard laboratory tool for DNA separations since the 1960s. The invention of capillary electrophoresis (CE) in the late 1970s greatly accelerated the use of electrophoretic separations for research in biological science and biotechnology, especially after the introduction of automated, high-throughput capillary-array sequencers in the 1990s to replace slab-gel sequencers, which enabled the completion of the human genome project a few years ahead of schedule. In CE, fused silica capillaries with a typical inner diameter of 50–100 µm and a length of 10–50 cm are used. In a representative DNA separation experiment, a sample band of about a millimeter or more in width is injected into the capillary, and it usually takes tens of minutes or longer to complete the run. Microchips for DNA separations can potentially further accelerate the research and technology development in molecular biology. In these devices, gel electrophoresis can be performed in a much shorter time scale, typically by more than an order of magnitude improvement than in capillaries with equivalent performance metrics, such as the number of theoretical plates and resolution. The main reason for the faster separation is the ability to inject a much narrower band of sample in a microfluidic device, typically on the order of 10–100 µm. In electrophoretic separation, the number of theoretical

DNA Separations

131

plates (N) increases as the square of column length (L) for small sample plug: N = 12 (L/w)2/(1+L/L0); L0 = µEw2/24D

where µ is the electrophoretic mobility of the analyte, E is the electric field, D is the diffusion coefficient of the analyte, and w is the root mean square of the width of the sample band and the width of the detector. In microfluidic experiments, the detector is typically focused down to 10–50 µm where it becomes comparable in length scale to the sample width. The time needed to achieve a certain degree of resolution between peaks (for L = L0) scales with w: t0 = w2/24D

As a result of the faster separation time, the requirement for column length L has also been reduced from tens of centimeters in capillaries to merely a few centimeters in a microchip. 2. Materials Currently there are two commercially available instrument platforms based on microfluidics technology for DNA separations. One is a bench-top research instrument called the 2100 Bioanalyzer (see Fig. 1) commercialized by Agilent Technologies, Inc., and the other is a higher throughput instrument called LabChip® 90 (see Fig. 2) commercialized by Caliper Life Sciences, Inc. The microchips used in the 2100 Bioanalyzer are “planar” microfluidic devices consisting of a soda-lime glass sandwich structure with microchannels micromachined onto the glass substrate using photolithography and chemical wet etching techniques (8). The microchannels are sealed by thermal bonding with a top glass plate containing access holes exposing the ends of the microchannels for reagent and sample introduction into the chip. A photograph illustrating examples of planar chips is shown in Fig. 3. The channel dimensions inside the chip are typically 10–25 µm deep and 20–100 µm wide. The microchannel design for the DNA chip is illustrated in Fig. 4. In this design, up to 12 samples can be analyzed in one chip using a single separation microchannel column. The samples are manually pipeted into the reservoirs of the chip after the microchannels are first filled with the polymer sieving matrix by a step called “chip priming.” The matrix is a solution containing a high molecular weight polymer, poly(dimethylmethacrylate) or PDMA, in a TAPScontaining buffer mixed with a DNA intercalating fluorescent dye. The fluorescent dye is light sensitive, therefore the dye stock and solutions containing dye should be stored in the dark when not in use. The as-received reagents and buffers are to be stored at –20°C. After the initial use, they should be stored

132

Chow

Fig. 1. The Agilent 2100 Bioanalyzer.

Fig. 2. The Caliper LabChip® 90.

DNA Separations

133

Fig. 3. Examples of planar LabChip® devices for separations.

Fig. 4. Microchannel design layout for planar DNA chip.

at 4°C to avoid freeze–thaw cycles (see Note 1). The polymer concentration and molecular weight distribution in the sieving matrix vary depending on the range of DNA fragment lengths to be separated and the specified resolution. Other reagents provided with the chips include a sample buffer and a DNA ladder with known fragment sizes and concentration as a calibration standard for sizing and quantitation. These reagents, supplied with 25 single-use, disposable

134

Chow

Fig. 5. A photograph of a sipper chip.

chips in a kit, should also be stored at 4°C after opening and between use. The details of the chip preparation method and the processes occurring inside the microchip are described in Subheading 3.1. In laboratories processing a large quantity of DNA samples, parallel sample preparation and experimentation are performed in a standard format called the microtiter plate. All microtiter plates have a standardized footprint, but the number of sample wells inside this footprint varies, with the most common ones being 96 (8 × 12) and 384 (16 × 24) wells. The LabChip 90 instrument is designed to interface with both the 96- and 384-well plate formats for high-throughput DNA analysis as it offers automated sampling performed by the chip to eliminate the need for manual sample pipeting. The chip contains a “sipper,” which is a fused silica capillary attached to the planar chip structure (also made of fused silica) as shown in Fig. 5. The fused silica planar structure is made using similar photolithography and chemical wet etching process as described previously for sodalime glass chips. The microchannel layout for the DNA chip is illustrated in Fig. 6. The channel dimensions are typically 5–25 µm deep and 20–500 µm wide. The sipper diameter is typically 20–50 µm. After the polymer matrix is properly introduced into the electrophoresis channel network of the chip, the sipper dips into each of the microtiter wells sequentially and samples a small quantity, about 100 nL, of DNA sample onto the chip for sizing and quantitation analysis. Sipping of sample is actuated by a small vacuum applied in the waste well, and subsequent injection and separation are performed electrokinetically on chip. A single sipper chip can usually be used to analyze hundreds to thousands of samples. The details

DNA Separations

135

Fig. 6. The microchannel design layout of the sipper DNA chip.

of chip preparation and operations are described in Subheading 3.2. Reagents provided with each chip kit include the sample buffer, marker DNA, ladder DNA, a PDMA polymer solution, and a DNA intercalating fluorescent dye concentrate. The dye concentrate should be kept in the dark to avoid degradation. All reagents and buffers should be stored at –20°C upon arrival, and stored at 4°C after opening and between use. The sipper chips should be stored at 4°C before use, and at room temperature after the first gel priming (see Note 1). After analyzing each 384-plate of samples, the chip should be reprimed with fresh gel-dye mix prior to analyzing another batch of samples. Typically, each chip can be used for up to 1800 samples if handled with good laboratory practices and care. 3. Methods 3.1. Planar Microfluidic Chips The procedures to perform DNA separations on planar chips consist of chip preparation, sample loading, and placing the chip onto the instrument. Chip preparation requires the complete filling of the channels of a dry chip with the separation polymer matrix mixed with the intercalating dye as follows (see Note 1): 1. First, 9 µL of the polymer solution mixed with dye is pipeted into well C4 (see Fig. 4 for chip layout). 2. A priming station consisting of a syringe is used to fill the microchannels with the polymer solution by applying pressure of 3 atm at well C4 for 60 s. 3. Once the channels are filled, 9 µL each of gel-dye mix is pipeted into the wells A4 and B4, and 5 µL in well D4.

136

Chow

Fig. 7. The chip-to-instrument interface of the 2100 Bioanalyzer. 4. Next, 5 µL each of sample buffer containing a lower and upper DNA marker is pipeted in all 12 sample wells. 5. The chip is now ready for sample loading. The steps for sample loading are: a. 1 µL each of DNA samples (see Note 2) is added into one of the 12 sample wells. b. 1 µL of DNA ladder is added into well D4. c. The entire chip is then placed on a vortex mixer for 60 s at a 2400 rpm setting to ensure that the sample and buffer are well mixed in each well.

The chip is now ready for loading onto the instrument as shown in Fig. 7. Once placed inside the instrument, the chip sits on a heater block and there is an electrode block located on the lid that matches the well pattern on the chip. When the lid is closed, each electrode pin makes contact with the liquid in each of the 16 wells, and voltage and current can be applied to the channels through the reagent wells. The optics is located below the chip to excite and record the fluorescence of the intercalating dye. The experimental run commences by activating the start button on screen. The following procedures have been programmed into the software script controlling the voltage and current, so they are transparent to the users. The script starts with a warm and focus step, which allows the chip to come to temperature equilibration with the heater block at 30°C and the auto focus function to locate and focus the light source onto the separation channel. During this time, the instrument also checks for electrical connections in all the wells before starting the separation analysis. The DNA ladder is first loaded by applying an electric field between the source “ladder” well D4 toward well B4. After the slowest moving DNA reaches the injection intersection,

DNA Separations

137

Fig. 8. Data output of the 2100 Bioanalyzer after sizing12 DNA samples in one chip.

the voltage is switched to wells A4 and C4. A small aliquot of ladder DNA sample with a sample band on the order of 50 µm is then injected into the separation channel, and the DNA fragments are separated by size through the polymer-dye matrix. As the separation of the ladder occurs, sample DNA from well A1 is simultaneously loaded by applying a voltage gradient from A1 to D4, with D4 now serving as the loading waste well. This overlapping injection and separation procedure follows until all six samples (A1 through B3) on the right have been analyzed. Then the DNA samples on the left, C1 through D3, are serially injected using B4 as the loading waste well and separated through the same separation channel. All the steps after the start of the experiment take approx 30 min to complete. The results are digital recordings of fluorescence intensity as a function of transit time for the ladder and samples as shown in Fig. 8. The electrophoregram can also be represented in a virtual gel image as shown in the lower left-hand corner of Fig. 8, and as a result table summarizing size and quantitation of identifiable fragments in each sample. The DNA size of each fragment is computed from the transit time as compared with the transit time of the ladder after the upper and lower markers in the samples and ladder are aligned by the software. DNA concentration

138

Chow

is computed from the area under each peak as compared with the known concentrations of DNA in the ladder. 3.2. Sipper Microfluidic Chips Unlike the planar chip, which is shipped dry, a sipper chip comes in a wet container in which the microchannels, the sipper, and the wells of the chip are filled with water. After the chip is removed from the container, it should be prepared using the following steps (see Note 1): 1. All the wells are first rinsed with filtered deionized water. 2. The fluid in wells 5, 6, and 8 are replaced with 75 µL of polymer matrix mixed with the dye concentrate (see Fig. 6 for chip layout). 3. The chip is next placed in a priming station where the electrophoresis channels are filled with the polymer-dye matrix solution by applying a pressure of 3 atm to wells 5, 6, and 8 for 4 min. 4. The fluid in well 3 is replaced with 75 µL of the polymer matrix. 5. The fluid in well 4 is replaced with 130 µL of marker dye solution containing both the upper and lower DNA markers.

The chip is now ready to be loaded in the instrument for high-throughput analysis. When the sipper chip is placed into instrument at the bottom of the optics module as shown in Fig. 9, the sipper capillary hangs out from the enclosed module where it can interface with the microtiter plate placed on the plate holder. The plate holder sits on a XYZ robotic arm and can accommodate the microtiter plate containing DNA samples (see Note 2) and two well strips, one for the ladder and one for sample buffer. The buffer well is used to rinse the capillary between samples to eliminate cross-contamination. Inside the instrument, the sipper chip rests on a heater block, which maintains the chip at a constant temperature of 30°C throughout the experiment. The initial chip warm up and focus step consists of sipping sample buffer by applying a vacuum in the waste well followed by multiple injection and separation steps. The excitation laser is focused on the separation channel using an auto-focus mechanism. Figure 10 schematically illustrates the automated operations inside the sipper chip. At the start of an experiment, the sipper visits the ladder well to bring onto the main channel about 100 nL of DNA ladder in sample buffer. As shown in Fig. 10B, the DNA sample from the sipper is mixed with the markers on the chip by applying a vacuum in the waste well, thereby causing flow toward the vacuum well. The mixture of DNA and markers are loaded onto the separation channel network electrokinetically using a voltage gradient between wells 3 and 6 (Fig. 10C). The voltage is then switched to wells 5 and 8 in order to inject a small aliquot of DNA into the separation channel, which separates the fragments by size through the polymer matrix (Fig. 10D). The excitation laser and

DNA Separations

139

Fig. 9. The sipper chip-to-instrument interface of the LabChip® 90.

detector are located near the end of the separation channel to record the fluorescence intensity of the DNA fragments. During electrophoretic separation of the first sample, a second sample is simultaneously sipped from the plate to be ready for sample injection. The sipping operation of one sample does not affect the separation resolution of another because of the high matrix viscosity and shallower channel depth in the separation channel, which prevent significant pressure-driven flow to occur in the separation channel. Such overlap sipping and separation processes are repeated sequentially until all the samples on the plate are analyzed. The DNA ladder is sipped repeatedly after every 12 samples. A 384-well plate full of samples takes approx 3 h to analyze. The sizing of DNA fragments is computed after the lower and upper markers in each of the 12 samples are aligned by software with those in the ladder sipped immediately prior to the samples, and the fragment transit times in each sample are compared with a calibration curve determined by the ladder transit time vs known DNA size. The DNA concentration is computed using the relative

140

Chow

Fig. 10. Schematic illustration of automated DNA sampling and separation in a sipper chip.

area under the peak as discussed before. Results can be presented as electrophoregrams, a virtual gel view, and tables summarizing computed sizes and relative concentration as shown in Fig. 11. The results from the tables can be exported for further postprocessing analysis (see Note 3). 4. Notes 1. Chip and reagent handling: as in CE, bubbles and debris in microchannels are usually detrimental to the separation performance and longevity of a microfluidic device. Before introducing any fluid onto the chip, it is highly recommended that it should be filtered with a micron- or submicron-pore filter first. During manual pipeting of solutions into chip wells, it is important to touch the tip of the pipet to the bottom of the small opening of the glass plate inside the well, not on top of the glass plate or to the side wall of the well, to avoid introducing or trapping bubbles in the well. Samples placed in tubes and microplates should be centrifuged in order to settle particulates to the bottom. Subsequent manual pipeting or automated sipping should be taken from the top half of the tube or microplate wells to minimize introducing unwanted particulates into the channels of the chip. The planar chips should always be run with all the sample wells filled with the recommended volume of fluid. If there are fewer than 12 samples to be used

DNA Separations

141

Fig. 11. Data output of a LabChip® 90 DNA run. in one device, the remaining wells should be filled with an equivalent amount of buffer to balance the hydrodynamics. The chip will not run properly if any wells are left empty. It is extremely important that the buffer and sample in the wells are thoroughly mixed using the vortex mixer at the conditions specified in Subheading 3.1. before loading the chip onto the instrument, otherwise DNA concentration determination will not be reproducible. The speed of the vortex mixer should be set to the indicated setting at 2400 rpm. To avoid DNA crosscontamination, the electrodes in the instrument should be cleaned periodically as recommended in the user manuals. As mentioned in Subheading 2., the intercalating dye is light sensitive. Therefore, it is important to keep solutions containing the dye in the dark when not in use. Otherwise, when the dye degrades significantly, signal intensity and, thus, detection sensitivity could be affected. If the dye concentration falls below a critical point owing to degradation, the electrophoregram may show a small dip following each DNA peak. This phenomenon should not affect the sizing accuracy of the analysis, but could degrade the quantitation accuracy since the baseline determination of each peak by the analysis software could be affected. Because new sipper chips and many reagents are refrigerated before use, adequate time should be allowed for the chip and reagents to reach room temperature before starting the chip preparation steps outlined in Subheading 3.2. If cold

142

Chow

reagents are introduced into the chip, warming up of an aqueous solution to room temperature in an enclosed microchannel could cause outgassing, especially under a slight vacuum condition. 2. Effects of sample buffers and other components: in both planar and sipper chip applications, the DNA samples are mixed with a sufficient amount of sample buffer such that the conductivity and main ionic species of all samples are primarily dictated by those in the sample buffer. Consequently, electrokinetic injection of all samples is similar without bias. However, if the samples are unusually salty, with ionic content larger than twice that of a typical PCR buffer, or have a pH outside the recommended range of 6.0 to 9.0, sample injection may not be representative, and the sensitivity limit of detection and peak resolution could be degraded. It is not recommended that DNA samples containing any significant quantity of plasmid DNA be analyzed in the polymer matrix formulated for separating linear dsDNA because the migration behavior of plasmid DNA is very different from that of linear DNA. Plasmid DNA has a ring structure and tends to migrate much more slowly through the polymer matrix for linear DNA. In samples of linear DNA containing some plasmid, the observed behavior during the analysis is that the first few samples may appear normal, but the fluorescent baseline rises dramatically in later samples as dye-stained plasmid arrives at the detector. Analysis of plasmid DNA samples requires the use of a polymer solution with a different pore size range to be effective (9). 3. Concluding remarks: with some proper care and good laboratory practices, these separation-based microfluidics systems have been proven to be robust and laborsaving compared with the traditional slab-gel technology. Perhaps more importantly, the output data is of high quality and available in a digital form which can be easily archived and shared with other laboratories across the company and around the world. Microfluidics technology has the potential to transform the productivity of molecular biologists in this era of an increasingly more global community with lightning-speed demand for information sharing.

References 1. Woolley, A. T. and Mathies, R. A. (1994) Ultra-high-speed DNA fragment separations using microfabricated capillary array electrophoresis chips. Proc. Natl. Acad. Sci. USA 91, 11,348–11,352. 2. Woolley, A.T. and Mathies, R. A. (1995) Ultra-high-speed DNA fragment separations using microfabricated capillary array electrophoresis chips. Anal. Chem. 67, 3676–3680. 3. Volkmuth, W. and Austin, R. H. (1992) DNA electrophoresis in microlithographic arrays. Nature (London) 358, 600–602. 4. Mohanty, U. and Stellwagen, N. C. (2002) Free solution mobility of oligomeric DNA. Biopolymer 49, 209–214. 5. Viovy, J. -L. (2000) Electrophoresis of DNA and other polyelectrolytes: physical mechanisms. Rev. Modern Phys. 72, 813–872.

DNA Separations

143

6. Albarghonthi, M. N. and Barron, A. E. (2000) Polymeric matrices for DNA sequencing by capillary electrophoresis. Electrophoresis 21, 4096–4111. 7. Slater, G. W., Kenward, M., McCormick, L. C., and Gauthier, M. G. (2003) The theory of DNA separation by capillary electrophoresis. Curr. Op. Biotech. 14, 58–64. 8. Bousse, L., Dubrow, B., and Ulfelder, K. (1998) High-performance DNA separations in microchip electrophoresis systems. In: Micro total Analysis Systems ‘98 (Harrison D. J. and van den Berg, A. eds.), Kluwer Academic Publishers, Dordrecht, pp. 271–275. 9. Ding, L., Williams, K., Ausserer, W., Bousse, L., and Dubrow, R. (2003) Analysis of plasmid samples on a microchip. Anal. Biochem. 316, 92–102.

11 Protein Separations Andrea W. Chow Summary This chapter describes the use of two types of commercialized microfluidic chips for protein separation, suitable for personal scale and high-throughput use. Compared with conventional approaches, such as sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and capillary electrophoresis (CE), these devices offer the advantages of faster separation times, better data reproducibility, greater ease of use, labor savings in quantitative analysis, and ease in data archiving and data sharing owing to the digital data format. With some simple precautions taken to keep bubbles and particulates out of the microchannels, Lab-on-a-Chip devices have been adopted by many researchers in protein processing, protein engineering, and proteomics research laboratories to increase their productivity. Key Words: Protein separation; protein sizing; microfluidics; Lab-on-a-Chip; LabChip® devices; sipper chip; SDS-PAGE; proteomics.

1. Introduction Identifying protein structure, function, and expression level under varying environmental conditions in cells is an important activity in most life science research and pharmaceutical discovery. Determining the molecular weight and purity of protein samples is routinely performed in laboratories using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in the last few decades (1,2). This process is highly manual and takes many hours to prepare the gel, run the samples, stain and destain the protein fractions, capture the gel image, and analyze the image to determine the protein sizes against a ladder. With commercially available precast gel, the first step of the process is becoming simpler and less messy, but overall it is still a relatively slow and tedious procedure. In large-scale protein processing operations and proteomics research laboratories, protein analysis often is a bottleneck in the workflow. The benefits in integration and automation offered by microfluidics have proven to be a faster and easier alternative to the traditional SDS-PAGE. From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

145

146

Chow

When a protein sample is denatured in the presence of a high concentration of SDS greater than its critical micellar concentration (CMC), the protein molecules are coated by SDS micelles, in general in the ratio of about 1.4 g of SDS per gram of protein (3). This binding ratio, somewhat surprisingly, is relatively independent of the protein sequence for proteins greater than 15 kDa. As a result of this interaction, the native charge of the protein is mostly masked by the charge of SDS, and the overall charge of the protein–SDS complex is primarily determined by the number of SDS micelles associating with the protein molecule. In the presence of an electric field without a polymer sizing matrix such as a polyacrylamide gel, all protein–SDS complex species migrate electrophoretically at roughly the same speed regardless of protein size, because the charge of the complex increases proportionately with the protein molecular weight. In the presence of a sizing matrix, on the other hand, the differential hydrodynamic interaction of the complex with the matrix causes the larger complexes to migrate slower than the smaller ones, resulting in an effective sizing mechanism by electrophoresis. The implementation of sizing protein–SDS complexes in capillary electrophoresis (CE) has not been widely adopted, unlike in sizing of DNA using CE. The difficulty comes from the lack of a convenient and sensitive means to detect the separated protein species. For DNA, a selection of fluorescent intercalation dyes binding to the double-stranded DNA (dsDNA) structure is available to detect the separated DNA fragments in CE with high sensitivity. For proteins, however, fluorescent dyes are available to stain the SDS micelles bound to the protein molecules, but they also stain the free SDS micelles in the gel, resulting in a low signal-to-noise protein measurement owing to the high fluorescent background caused by the signal from the free micelles. This high fluorescent background problem has been cleverly solved in a microfluidic chip using an integrated on-chip “destaining” step (4). After the protein fractions are sufficiently resolved in the separation column (see Chapter 10, Subheading 1. for a discussion of the time and length scaling laws for electrophoretic separation on chip), an intersecting channel structure is introduced as shown in Fig. 1. The side channels are filled with the same polymersizing matrix as in the main separation channel, but the buffer solution contains no SDS. By applying an electric current from the side channels into the main channel, two focusing streams containing no SDS are induced, creating a concentration gradient in SDS transverse to the flow direction downstream of the intersection. By Brownian diffusion, free SDS micelles in the center of the channel diffuse laterally outward into the zones of low SDS concentration, causing a breakup of the micellar structure when the local concentration of SDS drops below CMC. As free micelles break up, fluorescent dyes are released. Using dyes that do not fluoresce when they are dissociated from the micelles, the high background drops to a low level. This destaining process inside

Protein Separations

147

Fig. 1. Illustration of the destain intersection in a microfluidic chip for protein separation.

the microchannel network is photographically captured in the right-hand side in Fig. 1. Bound SDS micelles, on the other hand, cannot migrate outward as readily owing to the large protein size and they tend to remain stable because of the association with the proteins. Consequently, dyes captured in the protein–SDS complexes remain highly fluorescent after destaining. Figure 2 illustrates the phenomenon in a plot of measured fluorescent intensity as a function of protein migration time when a detector is placed downstream of the destaining channel intersection. When the destain ratio, a ratio of flow from the destain channels to that from the separation channel, is high enough to break down the free micelles, the background signal drops significantly and the protein peak intensity goes up simultaneously, most likely resulting from the availability of more dyes for the bound micelles after they are released from the free micelles. The time required for destaining of the free micelles is dictated by Brownian diffusion as described by the Einstein relationship: t = x2/2D

For free SDS micelles in the center streamline pinched laterally by two destaining solution streams as shown in Fig. 1, the diffusion distance into the SDS-free zone is approx 15 µm, and the characteristic diffusion time is on the order of 200 ms. As a result, the location of the detector can be designed to be appropriately placed downstream of the destaining intersection to optimize the benefits of the on-chip operation. This destaining step in protein sizing is one good example

148

Chow

Fig. 2. A plot of fluorescence vs protein transit time at a range of destain ratios.

of an integrated function easily performed on chip by exploiting the scaling laws of microfluidics that cannot be readily achieved in a conventional format. 2. Materials Two types of microfluidic chips are available for protein sizing and relative quantitation. One is suitable for analyzing a few protein samples (up to 10) at a time using the Agilent 2100 Bioanalyzer. The samples are pipeted into designated wells of the planar chips manually. The second type of device is designed for higher-throughput analysis, tens to hundreds of samples at a time, using a sipper chip and the Caliper LabChip® 90 instrument. The sipper is a capillary attached to the microfluidic chip to enable automated introduction of protein samples from a standard 96-microtiter plate format onto the microchannel network of the chip. A brief description on how these chips are made can be found in Subheading 2. of Chapter 10. Figure 3 illustrates the microchannel design layout of the planar chip for protein separation. The samples are manually pipeted into the reservoirs of the chip after the microchannels are filled with the polymer-sieving matrix. The matrix is

Protein Separations

149

Fig. 3. The microchannel design layout of the planar protein chip.

a solution containing a high molecular weight polymer, poly(dimethylmethacrylate) (PDMA), in a Tris-Tricine containing buffer (the gel-matrix) mixed with a noncovalent fluorescent dye and SDS (the dye concentrate). Other reagents provided with the chips include a sample buffer and a protein ladder with known molecular weight and concentration as a calibration standard for sizing and quantitation. The as-received reagents and buffers should be stored at –20°C. These reagents are supplied with 25 single-use, disposable chips in a kit, and should be stored at 4°C after opening and between use in order to avoid freeze–thaw cycles. The dye concentrate, as well as the gel-dye mix, should be kept in the dark owing to the light sensitivity of the dye (see Note 1). The details of the chip preparation method and the processes occurring inside the microchip are described in Subheading 3.2. The microchannel design layout for the protein sipper chip is illustrated in Fig. 4. After the polymer solution with and without SDS-dye is properly introduced into the electrophoresis and destaining channels of the chip, the sipper dips into each of the microtiter wells sequentially and samples a small quantity, approx 100 nL, of protein sample onto the chip by applying a small vacuum in the waste well. Subsequent injection and separation of samples are performed electrokinetically on chip. A single sipper chip can usually be used to analyze hundreds of samples. The details of the chip operations are provided in Subheading 3.3. Reagents provided with each chip kit include a sample buffer, a denaturing buffer, a marker dye, a protein ladder, a destaining solution (PDMA

150

Chow

Fig. 4. The microchannel design layout of the protein sipper chip.

polymer solution), and a dye concentrate (SDS mixed with an intercalating fluorescent dye). The dye concentrate and marker dye should be kept in the dark to avoid degradation. All reagents and buffers should be stored at –20°C upon arrival, and stored at 4°C after opening and between use. The sipper chips should also be stored at 4°C before use, and at room temperature after the first gel priming (see Note 1). Once the chip is primed, it can sit in the instrument for use during that day. Multiple batches of samples can be analyzed with the same chip without further chip preparation. After one 8-h day use, the chip should be reprimed with fresh gel-dye mix prior to use the next day. Typically, each chip can be used for up to 300 samples if handled with good laboratory practices and care. 3. Methods 3.1. Denaturing Protein Samples Before sizing analysis in microfluidic chips, the protein samples and protein ladder are denatured in high concentration of SDS or LDS (lithium dodecyl sulphate), at greater than 0.5%, either under reducing or nonreducing conditions. For reducing conditions, β-mercaptoethanol or dithiothreitol is typically added to the denaturing solution. The following steps are recommended for denaturing protein samples: 1. The protein samples, mixed in the denaturing solution as instructed in the user manuals, are heat-denatured at 90–100°C for 3–5 min. 2. After cooling down, the samples are further diluted with deionized water and/or sample buffer as directed in the user manuals, usually by approx 10- to 15-fold before use.

Protein Separations

151

The protein samples should be analyzed in the same day after denaturization. 3.2. Planar Microfluidic Chips The procedures to perform protein separations on planar chips consist of chip preparation, sample loading, and placing the chip onto the instrument. Chip preparation requires the complete filling of the separation channels of a dry chip with the polymer matrix mixed with SDS and dye (gel-dye mix) as follows (see Note 1): 1. First, 12 µL of the gel-dye mix is pipeted into the A4 well in the lower rightcorner of the well pattern of the chip (see Fig. 3 for chip layout). 2. A priming station consisting of a syringe is used to fill the channels with the polymer solution by applying pressure of 3 atm at well A4 for 60 s. 3. Once the channels are filled, 12 µL of gel-dye mix is pipeted into wells B4, C4 and D3. 4. Next, 12 µL of the destaining solution (gel without SDS-dye) is pipeted into well D4. 5. 6 µL each of sample (see Note 2) diluted into the sample buffer containing lower and upper protein markers is pipeted in all 10 sample wells. 6. 6 µL of the diluted ladder is pipeted into the ladder well D2.

The chip is now ready for loading onto the instrument. Inside the instrument, the chip rests on a heater block and there is an electrode block located on the lid that matches the well pattern on the chip. After the lid is closed, each electrode pin makes contact with the liquid in each of the 16 wells, and voltage and current can be applied to the channels through the reagent wells. The optics is located below the chip to excite and record the fluorescence of the intercalating dye. The sizing experiment commences by activating the start button on screen. The following procedures have been programmed into the software script controlling the voltage and current, so they are transparent to the users. The script starts with a warm and focus step, which allows the chip to come to temperature equilibration with the heater block at 30°C and the auto focus function to locate and focus the light source onto the separation channel. During this time, the instrument also checks for electrical connections in all the wells before starting the separation analysis. The protein ladder is first loaded by applying an electric field between the source “ladder” well (D2) toward well B4. After the upper protein marker reaches the injection intersection, the voltage is switched from well A4 to wells C4 and D4 for separation. A small aliquot of ladder protein with a band on the order of 50 µm is injected into the separation channel, and the protein species are separated by size through the SDS-dye containing polymer matrix followed by destaining and detection as described in Subheading 1. Overlap loading and separation of subsequent samples is performed to minimize the total analysis time. All the steps after the start of the experiment take approx

152

Chow

Fig. 5. Data output of the 2100 Bioanalyzer protein run.

30 min to complete. The results are digital recordings of fluorescence intensity as a function of transit time for the ladder and samples as shown in Fig. 5. The electropherograms can also be displayed as a virtual gel image by software and a result table. The protein size is computed from the transit time as compared with a calibration curve determined by the transit time of the ladder after the upper and lower markers in the samples and ladder are aligned by the software (see Note 3). Relative protein concentration is computed from the area under each peak as compared with the known concentrations of protein in the ladder. 3.3. Sipper Microfluidic Chips Unlike the planar chip, which is shipped dry, a sipper chip comes in a wet container in which the microchannels, the sipper, and the wells of the chip are filled with water. Before starting an experiment, the chip should be prepared using the following steps (see Note 1): 1. After the chip is removed from the container, all the wells are rinsed with filtered deionized water. 2. Contents in wells 5, 6, and 8 are replaced with 75 µL of polymer matrix containing SDS and dye, and contents in wells 2 and 7 are replaced with 75 µL of polymer matrix without SDS (see Fig. 4 for chip layout).

Protein Separations

153

3. The chip is next placed in a priming station where the electrophoresis channels are filled with the polymer-dye matrix solution by applying a pressure of 3 atm to wells 2, 5, 6, 7, and 8 for 10 min. 4. The fluid in well 3 is then replaced with 75 µL of the polymer matrix. 5. The fluid in well 4 is replaced with 120 µL of marker dye solution.

The chip is now ready to be loaded in the instrument for high-throughput analysis. When the sipper chip is placed onto the instrument at the bottom of the optics module, the sipper capillary hangs out from the enclosed module where it can interface with the microtiter plate placed on the plate holder. The plate holder sits on a XYZ robotic arm and can accommodate the microtiter plate containing denatured protein samples (see Notes 2 and 4) and two well strips, one for the ladder and one for sample buffer. The buffer well is used to rinse the capillary between samples to eliminate cross-contamination. Inside the instrument, the sipper chip rests on a heater block, which maintains the chip at a constant temperature of 30°C throughout the experiment. The initial chip warm up and focus step consists of sipping sample buffer by applying a vacuum in the waste well followed by multiple injection and separation. The excitation laser is focused onto the separation channel using an auto-focus mechanism. At the start of an experiment, the sipper visits the ladder well to bring onto the main channel approx 100 nL of protein ladder by applying a small vacuum in the waste well. Once arriving the planar channel on chip, the protein sample from the sipper is mixed with the marker on the chip and flows toward the waste well. The mixture of protein and marker is next loaded onto the separation channel network electrokinetically by applying a voltage gradient between wells 3 and 6. Then the voltage is switched to wells 5 and 8 to inject a small aliquot of the mixture into the separation channel, which separates the protein species by size through the polymer matrix. Voltage is also applied to wells 2 and 7 for destaining as described in Subheading 1. The excitation laser and detector are located downstream of the destain intersection to record the fluorescence intensity of the protein fractions. During electrophoretic separation of the first sample, a second sample is simultaneously sipped from the plate to be ready for sample injection. The sipping operation of one sample does not affect the separation resolution of another because of the high matrix viscosity and shallower channel depth in the separation channel, which prevent significant pressure-driven flow to occur in the separation channel. Such overlap sipping and separation processes are repeated sequentially until all the samples on the plate are analyzed. The protein ladder is sipped repeatedly after every 12 samples. A 96-well plate full of samples takes approx 1 h and 20 min to analyze. The system can be used to analyze full or partially filled plates. It is also possible to do continuous or intermittent analysis throughout the day using a single chip with one chip priming.

154

Chow

Fig. 6. Data output of a LabChip® 90 protein run.

The sizing of protein species in each sample is computed after the lower marker is aligned by software with that in the ladder sipped prior to the sample, and the transit times are compared with a calibration curve of ladder transit time vs known protein ladder size (see Note 2). The protein concentration for each peak is computed using the relative area under the peak as previously discussed after the peak areas are normalized by the area of the marker dye peak. Results can be presented as electrophoregrams, a virtual gel view, and tables summarizing computed sizes and relative concentration as shown in Fig. 6. The software has an “expected protein” feature to mark protein peaks detected at a predetermined size range. Automated calculation of protein purity can also be enabled. The results from the tables can be exported for further postprocessing analysis (see Note 5). 4. Notes 1. Chip and reagent handling: in microfluidics chips, keeping bubbles and debris away from the channels is important for good device performance and lifetime. Much of the general good laboratory practices on handling DNA chips and reagents discussed in Note 1, Chapter 10 are applicable to protein chips and reagents as well. In the following sections, only issues specific to protein sizing will be described.

Protein Separations

155

2. Effects of sample buffers and compositions: thus far, the detection sensitivity of proteins in commercial microfluidic chips is about the same as Coomassie Blue to mid-range of Colloidal Coomassie. It is advisable not to decrease protein sample dilution from the recommended ratio to try to increase sensitivity. Because protein buffers vary a lot in composition and conductivity owing to different protein processing conditions, decreasing denatured protein sample dilution from the recommended value would result in more ionic species associated with the protein sample being injected into the separation column. It has been found that in certain components of protein buffers, such as some detergents, relatively high concentrations in the injected sample band can affect the baseline flatness of the electrophoregram or the migration rate of the lower marker, potentially causing degradation in accuracy of sizing and relative concentration determination. 3. Data comparison with SDS-PAGE: as researchers using SDS-PAGE for protein sizing are switching over to microfluidic devices, one obvious question is how well the measured molecular weight and relative concentration are compared between the two methods. So far, experimental data show that these two approaches provide similar answers for most proteins, although a few proteins seem to be stained differently by the fluorescent dye and to migrate a little differently relative to other proteins. It has been accepted by the SDS-PAGE user community that migration rates and dye staining of protein–SDS complexes are somewhat dependent on the protein sequence and the specific reagents used. However, complete understanding of how these parameters depend on sequence and reagents does not yet exist. Unlike dsDNA, which can be described successfully by a wormlike chain model with a measured persistence length (5), there has been no satisfactory conformational description of a protein–SDS complex structure that has received wide acceptance. It was first thought that a protein–SDS complex adopts short rod-like segments connected by flexible regions (6). The “necklace” model, describing the protein–SDS complex as consisting of a series of spherical micelles surrounded by protein segment and connected by protein strands (7,8), is receiving increasing attention. With a lack of good understanding of protein–SDS conformation, however, it is difficult to ascertain the fundamental electrophoretic migration behavior. So far, phenomenological observation shows a linear relationship of electrophoretic mobility plotted against logarithmic molecular weight for protein sizing data generated by SDS-PAGE (1). Data obtained from the LabChip 90 sipper devices also show such a relationship in the molecular weight range of 10 to 250 kD in a variety of PDMA polymer solutions as illustrated in Fig. 7. In contrast, an equivalent plot for dsDNA migration is not linear (Fig. 8), and the changes in slope represent changes in the DNA conformation during migration through the polymer matrix (9). Thus, it is reasonable to propose that protein–SDS complexes migrate very similarly in the conventional and microfluidic equivalent of SDS-PAGE for protein sizing. The small discrepancies observed in a gel image generated by a SDSPAGE run vs the virtual gel image of a microfluidic sizing run are potentially because of the differences in the polymer matrix and dye used. Because the gel

156

Chow

Fig. 7. A plot of electrophoretic mobility vs log protein size.

Fig. 8. A plot of electrophoretic mobility vs log DNA size.

Protein Separations

157

Fig. 9. A comparison of a LabChip® 90 virtual gel view and sodium dodecyl sulfatepolyacryamide gel electrophoresis for a lysate sample. (Data provided as the courtesy of Structural GenomiX of San Diego.) images of two SDS-PAGE runs generated by reagents made by two different manufacturers do not necessarily match, one should not expect a perfect match of separation resolution and relative staining intensities between a conventional gel image and one generated by a microfluidic chip on the same protein sample. 4. Lysates and glycosylated proteins: analysis of whole cell lysates is essential for protein expression profiling studies. Figure 9 shows a comparison of a lysate sample analyzed by a sipper chip and by SDS-PAGE. The over-expressed protein around 48 kD is well detected by both methods, although the data reproducibility appears better in the chip-based method. In analyzing lysates, it is important that the sample concentration be within the upper limit of the recommended range, approx 2000 ng/µL, in order to avoid the potential problem of sipper clogging by potential unsolubilized protein aggregates. It is well known in SDS-PAGE analysis that glycosylated proteins migrate slower than their expected molecular weights. Qualitatively similar phenomenon has been observed in microfluidic chip separation of glycosylated proteins. Therefore, similar precautions in data interpretation should be taken in chip-based analysis as in SDS-PAGE. 5. Concluding remarks: with some proper care and good laboratory practices, these microfluidics systems have been proven to be robust and labor-saving compared with the traditional SDS-PAGE. Perhaps more importantly, the output data is of high quality and is available in a digital form, which can be easily archived and shared with other laboratories across the company and around the world. Protein detection sensitivity down to silver-stain equivalence on microfluidic chips is not yet commercial, but approaches explored so far look promising (10). Moreover, many research studies have also been reported on interfacing microfluidic chips with electrospray and MADLI mass spectrometry (11), and a polyimide LC/MS chip has recently been commercialized by Agilent. Microfluidics-based protein

158

Chow separation technologies have the potential to revolutionize the productivity of proteomics research in this era of an increasingly more global community with lightning-speed demand for information sharing.

References 1. Shapiro, A. L., Vinuela, E., and Maizel, J. V. Jr. (1967) Molecular weight estimation of polypeptide chains by electrophoresis in SDS-polyacrylamide gels. Biochem. Biophys. Res. Commun. 28, 815–820. 2. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 3. Weber, K. and Osborn, M. J. (1969) The reliability of molecular weight determinations by dodecyl sulfate-polyacrylamide gel electrophoresis. J. Biol. Chem. 244, 4406–4412. 4. Bousse, L., Mouradian, S., Minella, A., Lee, H., William, K., and Dubrow, R. (2001) Protein sizing on a microchip. Anal. Chem. 73, 1207–1212. 5. Smith, D. E., Perkins, T. T., and Chu, S. (1996) Dynamical scaling of DNA diffusion coefficients. Macromolecules 29, 1372, 1373. 6. Reynolds, J. A. and Tanford, C. (1970) The gross conformation of protein sodium sodecyl sulfate complex. J. Biol. Chem. 254, 5161–5165. 7. Guo, X. H. and Chen, S. H. (1990) The structure and thermodynamics of proteinSDS complexes in solution and the mechanism of their transport in gel electrophoresis proecess. Chem. Phys. 149, 129–139. 8. Narenberg, R., Kliger, J., and Horn, D. (1999) Angew. Chem., Int. Ed. Engl. 38, 1626. 9. Viovy, J. -L. (2000) Electrophoresis of DNA and other polyelectrolytes: physical mechanisms. Rev. Modern Phys. 72, 813–872. 10. Molho, J. I., Park, C., Price, K., et al. (2004) Ultrasensitive protein sizing using integrated isotachophoresis: gel electrophoresis. In: Proceedings of the 8th International Conference on Miniaturized Systems for Chemistry and Life Sciences (Northrup, M. A., ed.), Royal Society of Chemistry, Cambridge, UK, pp. 387–389. 11. Mouradian, S. (2002) Lab-on-a-chip: applications in proteomics. Curr. Opin. Chem. Biol. 6, 51–56.

12 Microchip Capillary Electrophoresis Application to Peptide Analysis Barbara A. Fogarty, Nathan A. Lacher, and Susan M. Lunte Summary The development of analytical methodologies to elucidate mechanisms of peptide transport and metabolism is important for the understanding of disease states and the design of effective drug therapies. Interest in the use of microchip capillary electrophoresis (CE) devices for peptide analysis stems from the ability to perform fast, highly efficient separations combined with small sample volume requirements. Many of the separation modes developed on conventional systems, including electrochromatography, isoelectric focusing, and electrophoretic bioaffinity assays, have been demonstrated on microchip devices. Steps that include sample preparation and labeling can also be integrated onto the microchip platform. This chapter will discuss considerations for peptide analysis using microchip CE and will focus on different approaches to sample preparation, separation, and detection. Key Words: Peptide analysis; microchip electrophoresis; electrochromatography; isoelectric focusing; electrophoretic bioaffinity assays.

1. Introduction Peptides are a large class of molecules that differ in size, charge, conformation, hydrophobicity, and the ability to form biospecific complexes. Peptides play an important role in many physiological processes, including the regulation of pain, blood pressure, and immune response, and can act as antibiotics, coenzymes/enzyme inhibitors, drugs, growth stimulators, hormones, neurotransmitters, and toxins (1). Since the identification of leu-enkephalin in the 1970s, the role of neuropeptides in the regulation of the central nervous system and signal transduction has been extensively investigated (1). Identification of specific peptides and their functions may permit the design of pharmacologically active synthetic analogs that could be used for the treatment of neurological diseases such as From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

159

160

Fogarty, Lacher, and Lunte

Parkinson’s and Alzheimers. Peptide analysis is also important for proteomics research. Digestion of a protein yields distinct peptide mixtures, which are separated and mapped in order to characterize the parent molecule. In some cases, multi-dimensional separations are needed to achieve resolution and identification of all peptide components. 1.1. Considerations for Peptide Analysis The determination of peptide concentrations in complex biological samples, protein digests, and drug formulations demands the use of highly efficient, sensitive, and selective analytical separations. Electrophoretic separation methods are routinely employed because of their fast analysis times, high separation efficiencies, and compatibility with sensitive detection techniques (2). Electrophoreticbased separation modes have been used for the determination of peptide mobility, isoelectric point, relative mass and charge, dissociation/association constants of complexes, diffusion coefficients, and hydrophobicity (3). For enzyme and immunoassays where biological activity must be maintained, buffers of physiological pH and temperature can be used. For hydrophobic peptides, the addition of the surfactants to the run buffer can aid peptide solubility. Coating the capillary wall is also used to minimize adsorption of analyte to the capillary surface (4). Lastly, because most peptides are amphoteric, analyte charge can be altered by adjusting the pH. This can change mobility, as well as enhance solubility and reduce adsorption to the negative wall. Many of the methods developed for peptide analysis on conventional capillary electrophoresis (CE) systems are now being transferred to microchip platforms owing to the many additional advantages of miniaturized analytical systems (5). This chapter will discuss general considerations for peptide analysis using microchip CE, highlight different techniques that can be used, and look at some specific applications of this technology. 1.2. Microchip Devices In microchip electrophoresis, separation is typically achieved through a combination of electrophoresis (µe) and electroosmotic flow (EOF) (µEOF). Charged analytes migrate under the influence of an electric field and are separated owing to differences in their mass/charge ratios (Fig. 1). Generation of EOF (a bulk flow that causes all ions regardless of charge to migrate to the cathode) is achieved through the ionization of silanol groups on the wall of the separation channel. The net negative charge on the channel wall attracts positively charged solvated ions from the buffer solution, forming an electric double layer. Upon generation of the electric field, the migration of solvated cations to the cathode creates a bulk flow of solution within the channel toward the cathode, where the detection region is located. As long as the magnitude of the EOF is greater than

Application of Microchip CE to Peptide Analysis

161

Fig. 1. Microchip capillary electrophoresis.

the electrophoretic mobility of the analytes (µEOF > µe), all species in solution, regardless of charge, will eventually pass the detector. Separations on microchips are faster and more efficient, and a higher sample throughput can be realized than with conventional systems. Microchip devices have minimal sample and reagent volume requirements, and the small footprints (device dimensions) allow portability. Microchip devices have been fabricated in a number of different substrates. Glass devices fabricated in soda lime, borosilicate, and quartz substrates have surface chemistries similar to that of the fused silica capillaries used in conventional systems. This allows the direct transfer of conventional methods to microchip systems. However, glass microchips have limitations, including the time required for fabrication, high cost, and the fragile nature of devices. In addition, integration of functional elements, such as separation and detection electrodes, is challenging owing to the high temperatures needed for the glass bonding process. Inexpensive polymer substrates including poly(methylmethacrylate) (PMMA) and poly(dimethylsiloxane) (PDMS) have been employed for the production of low-cost microchips (6). These disposable devices eliminate issues

162

Fogarty, Lacher, and Lunte

of sample carryover and cross-contamination. However, adsorption of peptides to hydrophobic polymers such as PDMS can be an issue (7). In addition, PDMS generates a reduced EOF relative to that of glass, which can influence the efficiency of separation (see Note 1). Approaches to minimize peptide adsorption to channel walls include the use of dynamic and permanent modifications. Dynamic coatings have been used to minimize adsorption of hydrophobic peptides to channel surfaces. Polymers such as polyvinylpyrrolidine have been used to modify channels in a glass microchip, resulting in improved resolution of small test peptides (8). PDMS devices have also been modified with different monomers using ultraviolet (UV) grafting techniques (9). Reduced peak adsorption and improved peak efficiencies were observed for two model peptides, F-PKC and F-src. A major advantage of using planar microchip devices for peptide analysis includes the ability to integrate sample preparation, injection, separation, and detection steps onto a single platform. Approaches to each of these stages of peptide analysis with respect to microchip technology will be examined in further detail. 2. Methods 2.1. Sample Preparation Preparation of peptide samples prior to analysis can involve dialysis, filtering, or extraction techniques to ensure that the sample is as clean as possible, preferably free from proteins and salts. Techniques, such as solid-phase extraction, can be used for preconcentration of peptides present at low concentrations in complex mixtures. Although sample preparation is generally carried out offline, the integration of extraction, dialysis, and digestion steps for the development of a true miniaturized total analysis system has been implemented with some success. However, much of the work in this area has been proof-of-concept with limited application to real-world samples thus far (10). 2.1.1. Sample Preconcentration and Desalting Extraction and preconcentration of peptides for electrophoretic analysis is traditionally achieved using conventional chromatographic stationary-phase materials. To use conventional stationary-phase particles in microfluidic devices, a mechanism for effective particle containment is needed to prevent clogging of microchannels. This is typically achieved using tapered reaction chambers to promote keystone packing of particles or the use of weirlike structures. Packed C18 particles have been used to preconcentrate and to remove salt and urea from peptide mixtures (11,12). Limitations of particle-based systems include the generation of high backpressures and the potential for clogging. One solution to these problems involves

Application of Microchip CE to Peptide Analysis

163

Fig. 2. Microchip capillary electrophoresis separation of two-peptide mixture sampled on-line through 2-mm brain probe. (Parameters: 1.5 s fill time, 2200 V). Analytes: FITCGly-His, 15 µM; FITC-Gly-Gly-His, 10 µM. (Reprinted with permission from ref. 14.)

the use of polymer monolith materials. Advantages of these continuous bed systems for sample extraction and preconcentration include low back-pressures, tailored chemistries, and easy control of pore size. For microfluidic applications, monoliths can be formed in situ and polymerized using heat or UV light. Methacrylate-based monolith systems have been employed for preconcentration of the hydrophobic tetrapeptide Phe-Gly-Phe-Gly by a factor of 1000 (13). 2.1.2. Microdialysis Microdialysis is an in vivo sampling technique used for continuous monitoring applications. Probes are implanted in biological tissues and fluids to monitor biological events such as neurotransmitter release. A membrane with a predetermined molecular weight cutoff is used to exclude large molecules such as proteins, while allowing smaller molecules such as amino acids and peptides to pass. This technique has recently been coupled to microchip electrophoresis (14). Figure 2 shows a microchip CE separation of two fluorescently labeled peptides following on-line recovery through a microdialysis probe (15).

164

Fogarty, Lacher, and Lunte

2.1.3. Protein Digestion: Proteomic Applications Proteolytic enzymes, such as trypsin, are used for the cleavage and digestion of protein molecules into their respective peptide fragments prior to identification/sequencing. Trypsin-coated beads have been employed for the on-chip digestion of proteins in microchip CE devices. High bead-to-peptide mass ratios and efficient mass transfer can allow rapid digestions of proteins, such as melitten, cytochrome c, bovine serum albumin, and β-casein (16). Pressure or EOF can be used to pump proteins through trypsin reactors (17). Localization of trypsin by enzyme adsorption to integrated membranes such as poly(vinylidene fluorine) has been used for the on-chip digestion of cytochrome c (18,19). Trypsin has also been immobilized in monolithic supports and applied to the digestion of myoglobin (20). More recently, sol-gel encapsulation has been investigated and used for the digestion of arginine ethyl ester (ArgOEt) and bradykinin (21,22) (see Note 2). Figure 3 shows a microchip device with integrated digestion, separation, and postcolumn derivatization for the detection of labeled peptide fragments of insulin B-chain (23). 2.2. Injection Once sample cleanup has been performed there are a number of different options for loading into the separation channel. Generally, this necessitates voltage control over multiple reservoirs (24). Popular methods include pinched and gated injections. Pinched injections demand simultaneous voltage control over all microchip reservoirs. A double T injector configuration reproducibly defines the volume of the sample plug. During separation, sample leakage into the channel is prevented by the application of a “pinching” or pushback voltage. Limitations of this approach include the need for multiple power supplies and the inability to change the volume injected using a double T design. In gated injection schemes, the sample plug is not volume-defined and can be altered. A voltage is applied to the sample reservoir with the sample waste reservoir held at ground. A second voltage is applied at the buffer reservoir with the buffer waste reservoir held at ground. Two separate flowing streams are established within the device. For injection, the sample reservoir is floated and a small plug of sample enters the separation channel. Both flows are then re-established for separation. This approach uses a simple T injector configuration and requires two power supplies. Although injection volumes can be varied, sample and buffer streams must be of similar ionic strength. Further information on injection methods is included in a recent review (25).

Application of Microchip CE to Peptide Analysis

165

Fig. 3. (A) Cross-sectional view of a capillary electrophoresis microchip, heating element, and the thermocouples. (B) Schematic of the microchip used for on-chip reactions, incorporating separation and postcolumn labeling. The fluid reservoirs are: (1) substrate, (2) enzyme or DTT, (3) buffer, (4) sample waste, (5) NDA, and (6) waste. (Reprinted with permission from ref. 23.)

2.3. Separation Methods 2.3.1. Capillary Zone Electrophoresis Capillary zone electrophoresis (CZE) has been investigated for peptide separations on microchip systems owing to the simplicity of the buffer system (see Note 3). The separation of Gly-Leu and Gly-Gly on a glass microchip was demonstrated using a 20-mM borate buffer, pH 9.5 (26). A mixture of bradykinin, substance P, luteinizing hormone, bombasin, oxytocin, and enkephalins was separated on a PMMA device in 100 µM phosphate at pH 5.0 (27). A simple CZE buffer consisting of 15 mM boric acid, pH 9.2, was used for the separation of neuropeptides on PDMS microchip devices (28). Angiotensin peptides have been separated on both PDMS and glass microchip devices using

166

Fogarty, Lacher, and Lunte

Fig. 4. Separation and laser-induced fluorescence detection of fluorescein-5isothiocyanate-labeled angiotensin peptides (angiotensin I, angiotensin II, and angiotensin III) at a concentration of 500 nM using an injection voltage of 2000 V. The buffer consisted of boric acid, Tris (20 mM, 100 mM, pH 9.0). (A) Separation obtained using a Pyrex microchip. The separation voltage used is 3200 V with an anti-leak voltage of 1600 V. (B) Separation obtained using a PDMS microchip. The separation voltage used is 4000 V with an anti-leak voltage of 2000 V. (Reprinted with permission from ref. 7.)

20 mM boric acid and 100 mM tris(hydroxymethyl)aminomethane, pH 9.0 (7). Figure 4 shows separations of these peptides performed on both glass and PDMS microchips with the polymer devices displaying significantly lower separation efficiencies. This is most likely owing to peptide adsorption to the channel surface (see Note 4). Methods to improve detection limits for microchip CZE of peptides include the use of polarity switching to initiate sample stacking (29). 2.3.2. Micellar Electrokinetic Chromatography In micellar electrokinetic chromatography (MEKC), separation is based on differences in the extent to which analytes partition between surfactant micelles and the surrounding aqueous buffer solution. When the concentration of a charged surfactant added to the buffer exceeds a value known as the critical micelle concentration (CMC), an aggregate, termed a micelle, is formed (30). The hydrophobic tails of the surfactant orient toward the center of the aggregate, and the polar headgroups orient toward the aqueous solution (31). A hydrophobic core is formed into which neutral hydrophobic analytes can partition (32). This method, initially developed for the separation of neutral molecules, is especially

Application of Microchip CE to Peptide Analysis

167

useful for peptides with a blocked N- or C-terminus. MEKC can also be used to separate charged peptides with similar electrophoretic mobilities but different hydrophobicites. A microchip MEKC separation of lysozyme, trypsin inhibitor, and carbonic anyhydrase was accomplished on a PMMA microchip with 1% SDS in the Tris-HCl run buffer (27). MEKC has also been used in conjunction with 2D separation schemes, which will be discussed later. 2.3.3. Isotachophoresis Isotachophoresis (ITP) is a separation technique based on electrophoresis occurring at a uniform speed where analyte migration is independent of mobility. It has been termed a “moving boundary” electrophoretic technique (33). ITP employs a leading and a terminating electrolyte with the sample zone sandwiched between them. For successful implementation of ITP, careful control of sample conductivity is needed. This can be difficult to obtain for biological samples containing high salt concentrations. In contrast to CZE, ITP will not separate anions and cations in the same run because EOF is generally suppressed. For successful implementation of ITP, the leading electrolyte must have a greater mobility than any of the analytes in the sample, whereas the terminating electrolyte must have a lower mobility. When the electric field is generated, the ions of interest will migrate toward the electrode of opposite charge in zones determined by their electrophoretic mobility. Zone width adjusts so that all zones have the same conductivity. After the initial zone formation, equilibrium is reached inside the channel, and the analyte ions then migrate past the detector at the same velocity. Narrowing of the analyte band can be accomplished by adding a high concentration of both the leading and terminating electrolyte. Microchip-ITP has been used as a preconcentration technique to improve detection limits for the determination of a fluorescently labeled peptide substrates following enzyme assay (34). Other applications have included the preconcentration of cytochrome c following on-line protein digestion (18) and characterization of peptide phosphorylation by protein kinase (35). 2.3.4. Isoelectric Focusing Isoelectric focusing (IEF) separates peptides and proteins based on differences in their isoelectric point (pI). IEF can be used to separate proteins differing in pI by 0.0005 U or less (33,36–39). An advantage of IEF is that the whole channel can be filled with sample, improving detection sensitivity. The sample is dissolved in a mixture of carrier ampholytes prepared in different pH ranges, and the channel is filled (40). Loading is followed by focusing and mobilization steps (Fig. 5). Once the sample has been loaded, a voltage is applied and the carrier ampholytes separate into individual bands, establishing a pH gradient inside of the microchip. There are three different types of behavior that the peptide can exhibit once the

168

Fogarty, Lacher, and Lunte

Fig. 5. Principle of capillary isoelectric focusing by simultaneous pressure/voltage mobilization. (A) Catholyte is backflushed past the detection point and a sample plug is introduced into the coated capillary (no high voltage). (B) Focusing of sample is complete and the sample components are driven toward the detector by a low-pressure rinse. High voltage is applied during this step. (Reprinted with permission from ref. 40.)

voltage is applied. If the pH of the band is lower than the pI, the peptide will be positively charged and will migrate toward the cathode. If the pH is higher than the pI, the peptide will be negatively charged and will migrate toward the anode, and if the pI is equal to the pH the peptide is neutral and will not migrate. All peptides will migrate until the pH is equal to the pI. At this point the current will approach zero, the resistance in the channel will be high, and focusing will cease. Eventually, all of the peptides will be focused into bands where their pI is equal to the pH and they will stop migrating. One drawback of IEF is that, because there is a pH gradient formed along the channel, the extent of surface ionization varies. This can produce an uneven generation of EOF and cause band broadening. Therefore, the channel is usually modified to eliminate the EOF and to ensure that focusing is achieved prior to detection. This is accomplished by coating the microchip walls with a neutral coating, which also limits peptide adsorption (41). For single point detection systems, analytes must be mobilized past the detector region following the focusing step. The two most common approaches to analyte mobilization are hydraulic or chemical (31). Hydraulic mobilization of the focused zones occurs by applying either pressure or a vacuum to one end of a channel. Chemical mobilization is accomplished through the addition of a salt, such as NaCl, to one of the buffer reservoirs, followed by application of a

Application of Microchip CE to Peptide Analysis

169

high voltage. The actual mobilization in this case occurs owing to the presence of a competing ion such as Cl– that competes with OH–. This lowers the pH, and the peptides become charged and migrate toward the cathode. The Na+ competes with H+, causing an increase in pH and making more negative analytes migrate toward the anode. This pH shift occurs across the entire length of the microchip channel (31). The end result of this pH shift is the mobilization of the previously focused analytes toward the cathode. An alternative approach termed “one-step IEF” achieves focusing and mobilization simultaneously. As the pH gradient is not stationary, the EOF mobilizes the sample zones during focusing. One-step IEF can be accomplished using either coated or uncoated devices (42). EOF-driven mobilization is useful for microchip systems because of the speed of analysis and minimal instrumentation requirements. This approach was used for the separation of Cy5-labeled peptides; however, higher separation efficiencies and reproducibilities were achieved using chemical and hydraulic mobilization strategies (42). The requirement for sample mobilization past a single detection point can be obviated through the use of whole-column imaging. This technique has been applied to the separation of test peptides using an UV absorbance detector (43). Because of reduced detection sensitivity, high concentrations of peptide were required, and glycerol was added to maintain peptide solubility. A more sensitive approach used fluorescence detection for microchip IEF of rhodamine green-labeled peptides on a PMMA device (44). Real-time imaging of the channel was performed using a scanning detection system that facilitated the determination of peptide migration times along with the estimation of EOF and pressure-driven flows. A scanning detection system based on acousto–optical deflection has also been developed and used for imaging of proteins during IEF (45). Advantages of this technology include fast response times, the absence of mechanical parts, and minimal background noise. 2.3.5. Electrochromatography The development of capillary electrochromatography (CEC), a hybrid of CE and liquid chromatography (LC), has allowed highly efficient separations of peptide mixtures (46). In CEC the channel is packed with a stationary phase for chromatographic analyte retention similar to LC, but EOF is also generated. The main advantage of this technique over LC is improved peak efficiencies owing to the flat plug flow profile generated by the EOF (41). The use of conventional bead-based retention systems in microfluidic devices requires methods (weirs, channels) to localize the beads and prevent device clogging. A 200-µm channel packed with octadecylsilyl stationaryphase particles was used to separate a fluorescently labeled angiotensin peptide from excess labeling agent (47).

170

Fogarty, Lacher, and Lunte

Fig. 6. Scheme of a collocated monolithic support structures separation column. (Reprinted with permission from ref. 49.)

In chip-based electrochromatography, the stationary phase, the channel, inlets, and outlets can all be produced in one device using lithographic patterning techniques that are already employed for the fabrication of microfluidic devices. Collocated monolithic support structures (COMOSS) that are molded directly into the channel have been created in both quartz and PDMS (48,49). Advantages of this approach include a uniform distribution of support structures defined during fabrication and homogenous channel dimensions. Issues involved with packing and retention of stationary-phase particles are also eliminated as the reversed phase stationary phase is bonded directly to the structures, and there is no need for the production of frits. Figure 6 shows a COMOSS fabricated in PDMS. Applications of COMOSS modified with C18 phases include the separation of tryptic digests of ovalbumin (48) and fluoroisothiocyanate-labeled bovine serum albumin (50).

Application of Microchip CE to Peptide Analysis

171

Fig. 7. Schematic of the microchip used for electrochromatography. B, S, BW, and SW denote reservoirs containing buffer, sample, buffer waste, and sample waste, respectively. The inset shows a scanning electron micrograph (SEM) of a channel cross section filled with photoinitiated acrylate polymer monolith. The mean pore diameter is 1 mm. (Reprinted with permission from ref. 52.)

Monolithic polymer stationary phase materials are proving useful for microchip CEC applications (Fig. 7). Advantages of these phases include in situ polymerization, high flow rate tolerance, and functionalities that can be tailored to suit a specific application (51). In situ photopolymerization can be achieved by selective exposure to UV light. The location of the polymer can be accurately defined using a photomask. Acrylate-based porous polymers have been cast in glass channels for the separation of fluorescently labeled peptides, including angiotensins (52). The glass microchips could be reused following thermal incineration of the monolith at 550°C for 2 h and overnight incubation in 0.2 N NaOH. Sol-gels modified with polyelectrolyte multilayers are also being investigated for potential application as cation-exchange materials for microchip CEC separations of peptides, including enkephalins (53). 2.3.6. Electrophoretic Bioaffinity Assays Electrophoretic bioaffinity assays combine the specificity of an immunological or enzymatic response with the separation power of electrophoresis. Separation and quantification of enzymatic conversion or antibody/antigen binding can be achieved using microchip CE. Potential applications of affinity separation technologies include biochemical, clinical, and drug development operations (54). 2.3.6.1. IMMUNOASSAYS

Immunoassay techniques are employed for the determination of binding constants for antibodies and antigens in competitive and noncompetitive

172

Fogarty, Lacher, and Lunte

studies. The separation of human immunoglobulin M (IgM) and anti-human IgM has been achieved using a PMMA microchip device with conductivity detection (55). Tween surfactant was added to the run buffer to reduce the adsorption of the proteins to the channel wall. A PMMA device was also used for the separation of fluorescently labeled human anti-goat antibody and unlabeled goat IgG. In this case no significant adsorption of IgG protein was observed (56). MEKC has been employed for an immunoassay monitoring serum levels of theophylline (57). A competitive immunoassay to monitor theophylline in serum samples and a direct assay for the determination of monoclonal mouse IgG in mouse ascites fluid were demonstrated on-chip using a tricine buffer containing Tween-20 and NaCl (58). A subsequent microchip system incorporating mixing, reaction, separation, and detection steps was developed for the same application (59). A microchip CE device that performed electrokinetic mixing of antibody with bovine serum albumin (BSA) and a diluting buffer was demonstrated for the on-chip preparation of calibration standards. These standards were then reacted with flourescein-labeled BSA, allowing the on-line generation of a standard curve. The device was used to assay the BSA antibody prepared from diluted mouse ascites fluid (60). A competitive chip-based immunoassay has been developed for the determination of insulin secretion from single islet cells (61). Online mixing of reagents allowed continuous sampling, and the simple device design needed only one voltage power supply for operation. A competitive immunoassay using rabbit polyclonal anticortisol antiserum and a fluorescently labeled cortisol derivative was demonstrated on a glass microchip device (62). The working range of the device was determined to be within clinical range (1–60 µg/dL cortisol). A heterogeneous competitive immunoassay of human IgG was demonstrated on a hybrid PDMS and glass device (63). Cy5-labeled human IgG was used as a tracer and Cy3-mouse IgG was used as internal standard. The microchip system was used for the determination of IgG levels in human serum with fluorescence detection. 2.3.6.2. ENZYME ASSAYS

The determination of endogenous extracellular signal-regulated protein kinase was achieved using microchip CE by the separation of the substrate and product of the enzyme assay from an internal standard within 20 s. The device was used to measure the activity of endogenous levels of the enzyme in cell lysates (64). Figure 8 shows a separation of unphosphorylated and phosphorylated fluorescein-labeled “Kemptide” following on-chip enymatic conversion

Application of Microchip CE to Peptide Analysis

173

Fig. 8. Time course of the protein kinase A reaction on a 12A chip. Electrophoretic separation of unphosphorylated and phosphorylated fluorescein labeled Kemptide upon enzymatic conversion by PA in a reagent well of the 12A microchip. Conditions: 13.3 mM Fl-Kemptide, 24.5 nM protein kinase A in 100 mM Hepes, pH 7.5, 1 M NDSB195, 5 mM MgCl2, 100 mM ATP, 50 mM cAMP, 0.1% Triton X-100, 10 mM dithiothreitol. (Reprinted with permission from ref. 65.)

in a reagent well by protein kinase (65). The Km for “Kemptide” was determined using Lineweaver-Burk plots. A capillary array was used to demonstrate a multiplexed approach to the analysis of kinases (66). The assay was performed by the addition of multiple enzymes to a reaction tube. Simultaneous resolution of four product and three substrate peaks was demonstrated within 30 s. The multiplexed approach was also tested using a phospholipase and another kinase, demonstrating the ability to screen different types of enzymes and potential inhibitors in one analysis. Another advantage of microchip systems is the ability to fabricate multiple channels for high-throughput analysis. Microchip array electrophoresis has also been used to monitor the kinetics of a phosphate substrate and alkaline phosphatase conjugate as a function of reaction time (67). The enzymatic products were detected using laser-induced fluorescence (LIF) with a miniature semiconductor laser. An electrochemical enzyme immunoassay was developed incorporating postcolumn reactions of alkaline phosphatase-labeled antibody (68). Amperometric detection was then used to identify the free antibody and the antibody–antigen complex. Microchip CE has been used for the separation of four different inhibitors of acetylcholinesterase following on-chip mixing of enzyme, substrate, inhibitor, and derivatizing reagents (69). Acetylthiocholine is hydrolyzed to thiocholine

174

Fogarty, Lacher, and Lunte

by acetylcholinesterase. The reaction of thiocholine and coumarinylphenylmaleimide forms a thioether that can be detected using fluorescence. Modulation of enzyme activity was identified by a decrease in fluorescence intensity. 2.3.7. Multi-Dimensional Separations Chromatographic and electrophoretic separation systems have a maximum peak capacity that is dependent on column length and separation efficiency. With LC the peak capacity typically ranges from 20 to 100, whereas for CE, values of 50–100 can be obtained (70). For the analysis of complex biological samples, a one-dimensional system will have a limited peak capacity; therefore, the use of two-dimensional separations can extend the applicability of the technique. For 2D separations, the output of the first separation technique serves as the input to the second. Therefore, to successfully analyze each peak, the separation in the second dimension must be faster than in the first. A conventional capillary high pressure (HP)LC system has been interfaced to microchip electrophoresis for the separation of fluorescently labeled peptides from tryptic digests of BSA (71). Effluent from the HPLC was injected into the CE channel every 20 s. On-chip coupling of 2D separation systems is facilitated by the planar nature of the microchip device. A combination of MEKC (10 mM sodium dodecyl sulfate [SDS]) and CZE was demonstrated on a microchip platform for the separation of tryptic peptides (70). Output from the first dimension was introduced to the second using a gated injection scheme. A similar approach was adopted for the coupling of open channel electrochromatography and CE on a glass microchip (72). Figure 9A shows the device used with the spiral channel used for the CEC, which was coated with a C18 phase, connected to the straight channel used for CZE in the second dimension. Figure 9B shows the application of the device to a 2D separation of fluorescently labeled tryptic peptides of β-casein. Acrylic devices have been employed for the on-chip coupling of IEF and CZE (73). The acrylic microchip supported a reduced EOF relative to that of glass, and this property was exploited as a slow mobilization method. When the output of the first dimension (IEF channel) is introduced to the second (CZE channel), the voltage is switched and applied to the CE channel only. As a result, the analysis time of the entire system is dominated by that of the second dimension (<1 min). The device was applied to the analysis of fluorescently labeled dextran and ovalbumin. On-chip tryptic digestion followed by 2D electrochromatography has been demonstrated for histidine-containing peptides on a PDMS device (74). Digestion of BSA was followed by separation of the peptides by metal affinity chromatography and reversed-phase electrochromatography.

Application of Microchip CE to Peptide Analysis

175

Fig. 9. Capillary electrochromatography-capillary electrophoresis (CEC-CE) chip. (A) The CEC serpentine channel extends from V1 to V2 (25 cm), the CE channel from V2 to the detection point y (0.8 cm), (B) 2-D contour plot of a β-casein tryptic digest. (Reprinted with permission from ref. 72.)

2.4. Detection Modes Once a separation has been achieved, the most popular modes for peptide detection include fluorescence, electrochemistry (EC), or mass spectrometry (MS). Because of enhanced sensitivity, fluorescence and MS are the more popular modes of detection for peptide analysis with microchip CE. 2.4.1. Fluorescence Detection Fluorescence detection is the most commonly reported detection mode for microchip CE owing to the high sensitivity that can be obtained and the ability to focus collimated light onto micron-sized separation channels with relative ease. Excitation sources include lamps and LEDs; however, despite their high cost, lasers are also popular because of their ability to provide high-intensity radiation. Some peptides and proteins contain fluorescent amino acids, such as tyrosine and tryptophan, and exhibit native fluorescence (75). However, most peptides/ proteins must be labeled with an extrinsic fluorophore prior to analysis in order to be detected. Fluorescein-5-isothiocyanate (FITC) is the most popular reagent used for labeling of peptides; it is compatible with the 488-nm line of the argon-ion laser. Other fluorophores that have been employed include fluorescein, o-phthaldialdehyde (OPA), naphthalene-2,3-dicarbozaldehyde (NDA), BODIPY®

176

Fogarty, Lacher, and Lunte

TR, rhodamine red, dansyl chloride, 5- (and 6-) carboxyfluorescein, and NanoOrange. Cy5 label has been used to label proteins through the primary amines. Cy5 is excited at 633 nm, with collection of resulting emission at 680 nm. The reaction rate, fluorescent product stability, quantum yield, excitation wavelength, and emission wavelength should all be considered when selecting a reagent for a specific application. On-chip derivatization can take place prior to analysis (precolumn) or after the separation (postcolumn). The on-line postcolumn labeling of peptide fragments following tryptic digestion of oxidized insulin B-chain has been demonstrated using NDA/CN (23). 2.4.2. Electrochemical Detection Electrochemistry is currently the detection mode most amenable to the production of a truly portable device. As the response at the electrode is not pathlength dependent, electrochemical detectors can be miniaturized without a loss in sensitivity. Electrodes can be microfabricated using the same techniques employed for chip manufacture and integrated onto the microchip platform. Although EC detection encompasses conductimetry, potentiometry, and amperometry, the last is the most commonly reported mode of EC detection for microchip-based peptide separations (76–83). Electroactive amino acids include tyrosine (Tyr), tryptophan (Trp), and cysteine (Cys). If the peptide does not contain one of these electroactive amino acids, it must then be derivatized or complexed with an electroactive substance in order to enable detection. Derivatization agents that label the primary amine of the peptide have been used to enable electrochemical detection of aminecontaining compounds; these include o-phthaldialdehyde/β-mercaptoethanol (OPA/βΜΕ) and naphthalene-2,3-dicarboxaldehyde/cyanide (NDA/CN). Both are oxidized at 750 mV vs Ag/AgCl at a carbon electrode (81,84). Precolumn derivatization of small peptides with NDA/CN for electrochemical detection was demonstrated on a glass microchip device using a reaction chamber (26). An end-column detection alignment was used in conjunction with an external screen-printed carbon electrode (see Note 5). Most schemes require that the analyte possess a primary amine functionality for derivatization; however, it is possible to detect peptides with blocked N-termini, peptides lacking lysine, and cyclic peptides by the formation of a peptide-copper (II) complex (85,86). As these complexes can undergo reversible redox reactions, dual-electrode detection can be used. This approach typically involves oxidation at the first electrode and reduction at the second. Figure 10 shows a dual-electrode electropherogram of the Des-Tyr leuenkephalin copper complex obtained using a microchip CE-EC device (83). A mixture of 50 mM boric acid, 3 mM tartaric acid, and 1 mM CuSO4 buffered at pH 9.8 was used for separation (see Note 6).

Application of Microchip CE to Peptide Analysis

177

Fig. 10. (A) Microchip capillary electrophoresis with end-column dual electrode detection. (B) Separation and dual-electrode detection in a series configuration with carbon paste electrodes. Detection of precomplexed copper complexes of: TyrGlyGly (340 µM) and des-Tyr leu-enkephalin (450 µM). Separation conditions: 50 mM boric acid, 3 mM tartaric acid, 1 mM copper sulfate, pH 9.8; applied voltage = 1170 V (350 V/cm). Injection: 1 s (S to SW) at 1170 V. E1 = 900 mV vs Ag/AgCl; E2 = 300 mV vs Ag/AgCl. (Reprinted with permission from ref. 83.)

Although amperometry has proven to be the most popular mode of electrochemical detection for peptides, several examples of the use of conductivity detection have also been described. Conductivity detection is a universal detection technique that can be used in conjunction with microchip CE. The only requirement is that the analyte conductivity is different from that of the carrier electrolyte. An integrated conductivity detector for microchip CE has been described (27). The detector was constructed from platinum wires aligned using guide channels and integrated into a PMMA microchip device by the thermal annealing of a PMMA cover plate. A CZE separation with indirect detection of a mixture of nine peptides, including bradykinin, substance P, and leu-enkephalin, was demonstrated. The same device and configuration were applied to the separation of a protein mixture by MEKC. A limitation of contact conductivity detection is the lower field strengths that must be applied (generally on the order of 50 V/cm) to minimize gas generation and bubble formation at the electrodes. This can influence separation efficiency but can be overcome through the use of a contactless approach. As there is no direct contact with the solution, electrode fouling is no longer an issue. This approach was used in conjunction with glass microchips for the detection of two small test peptides. Capacitive coupling with the separation electrolyte was achieved by coating electrodes with a

178

Fogarty, Lacher, and Lunte

thin layer of silicon carbide, which acted as the insulating layer (87). A contactless configuration was also employed for the detection of human immunoglobulin M in glass and PMMA microchip devices (55). Tween surfactant, added to the buffer to minimize adsorption to the walls of the device, did not appear to interfere with the detection. 2.4.3. Mass Spectrometric Detection Because of the ability of MS to provide molecular weight and structural information without the need for derivatization of analytes, it has become a very desirable detection mode for peptides. MS is also useful for high-throughput analysis. The ability to make disposable separation devices obviates many of the issues of cross-contamination, a concern for high-throughput applications. Often, microfluidic chips are used as sample introduction methods for MS as the low flow rates used are compatible with sample delivery. The ability to integrate sample cleanup prior to analysis is also an advantage. Electrospray ionization (ESI) is the most popular ionization technique used in conjunction with microfabricated devices, although matrix-assisted laser desorption/ionization is also compatible. Ionization can occur on- or off-chip, with some emitters fabricated directly onto the chip device. For successful ionization, analytes must be charged and in solution. Ions are transferred to the gas phase by generation of an electric field between two electrodes forming a spray of sample solution. ESI is compatible with a number of mass analyzers; however, time-of-flight (TOF) and quadrupole-TOF mass analyzers are perhaps best suited to handling high-throughput analysis and have found greater application. Considerations for coupling microchip to MS include minimization of deadvolume, establishing a steady flow rate for sample introduction, and minimizing sample adsorption to the microfluidic device and solvent compatibility of the device. In addition, careful selection of the pH of the analyte solution is required, as this will determine analyte charge (this relates to the detection mode setting for positive or negative ion monitoring), and the magnitude and direction of EOF generated on the microchip device (this will determine whether the peptide mobility is sufficient to reach the emitter) (88). A microfabricated chip was interfaced to ESI-MS and applied to the analysis of a tryptic digest of β-casein (89). Coating the microchannel and the transfer capillary with a polybrene solution improved both delivery and response. The separation and identification of peptides were achieved using a microfabricated CE device with nanoelectrospray MS (90). Initial protein extraction was followed by SDS-polyacrylamide gel electrophoresis (PAGE) and tryptic digestion. The microfluidic device was then used to perform sample cleanup prior to MS analysis. A microfluidic device has also been coupled

Application of Microchip CE to Peptide Analysis

179

to a quadrupole TOF-MS for trace analysis of protein digests isolated on a gel. Sample stacking was used to improve detection limits for hydrophilic peptides, whereas solid-phase extraction techniques were used to preconcentrate hydrophobic peptides (29). Sample preparation procedures, such as filtration, solid-phase extraction (SPE), preconcentration, and protein digestion, have been incorporated onto the chip platform prior to the introduction of the sample into the mass spectrometer, representing some of the highest degrees of integration (91–94). Analysis of both proteins and peptides has been achieved using this approach (95). Further optimization of chip-to-MS interfaces combining ionization and desalting steps is needed before this technology will be used on a routine basis. 3. Summary Microchip electrophoresis has made great strides since the initial concept was described by Manz and coworkers in 1990 (5). Thus far, CZE has been commonly employed as separation method on-chip, but other CE modes such as MEKC, CEC, ITP, IEF, and 2D separations are being applied to peptide analysis. Many of the detection methods used with conventional systems such as LIF, EC, and MS have been successfully adapted for use with microchip systems. The development of truly integrated miniaturized analysis systems that can perform sample isolation, digestion, separation, and detection will significantly improve the field of peptide analysis. The resulting devices could then be used for screening, cell culture, and proteomic studies. 4. Notes 1. Selection of the microchip substrate will ultimately depend on compatibility with sample and buffer solutions and fabrication facilities available. Glass microchips are fabricated using chemical etch procedures, whereas polymer devices are typically fabricated by hot embossing or casting techniques. Because of the hydrophobic nature of the polymer PDMS, reversible sealing of the separate layers can be achieved through contact bonding. If a permanent seal is required, PDMS may be plasma oxidized, which also improves the magnitude of EOF over that of native devices. 2. For on-chip protein digestion, immobilization of trypsin in sol-gel materials can increase stability compared with that in free solution. 3. For simple CZE separations, microchip channels should be conditioned at the beginning of the day by sequentially flushing for 10 min with degassed and filtered solutions of 0.1 M NaOH, H2O, and buffer. At the end of the day, channels should be rinsed with water to prevent buffer crystal precipitation. 4. The use of acidic or basic buffer pHs can aid charge repulsion from channel surfaces minimizing peptide adsorption. Neutral hydrophilic coatings can help to

180

Fogarty, Lacher, and Lunte

reduce issues of peptide adsorption to channel walls. For CEC applications, using an acidic elution buffer can reduce peptide adsorption and tailing effects, commonly seen with silica-based phases. 5. A major consideration for EC detection with any CE application is the isolation of the separation field from the detection in order to protect the potentiostat. For amperometric detection, electrodes can be aligned at the end of the channel to protect the potentiostat from the high field strengths needed for the electrophoretic separation. However, limitations of this approach include analyte diffusion, leading to reduced peak efficiencies. Alternative approaches to the effective isolation of the detector from the separation voltage include off-channel detection using decouplers designed to shunt away separation voltage prior to detection or inchannel detection requiring the development of electronically isolated (floating) potentiostats. 6. Copper(II)-peptide complexes oxidize at a carbon electrode (750 mV) vs an Ag/AgCl reference. For dual detection, the Cu(II) complex is oxidized at the first electrode and the resulting Cu(III) complex is reduced to a Cu(II) complex at the second electrode. The reduction of the complex at the second electrode takes place at a mild reduction potential (100–200 mV), which results in a better S/N ratio, lower LOD, and better selectivity.

References 1. Strand, F. L. (1999) Neuropeptides. Regulators of Physiological Processes, MIT Press, Cambridge, MA. 2. Issaq, H. J. (2001) The role of separation science in proteomics research. Electrophoresis 22, 3629–3683. 3. Kasicka, V. (2003) Recent advances in capillary electrophoresis and capillary electrochromatography of peptides. Electrophoresis 24, 4013–4046. 4. Rodriguez, I. and Li, S. F. Y. (1999) Surface deactivation in protein and peptide analysis by capillary electrophoresis. Anal. Chim. Acta 383, 1–26. 5. Manz, A., Graber, N., and Widmer, H. M. (1990) Miniaturized Total Chemical Analysis Systems: a Novel Concept for Chemical Sensing. Sens. Actuators B 1, 244–248. 6. Soper, S. A., Ford, S. M., Qi, S., McCarley, R. L., Kelly, K., and Murphy, M. C. (2000) Polymeric microelectromechanical systems. Anal. Chem. 72, 643A–651A. 7. Lacher, N. A., de Rooij, N. F., Verpoorte, E., and Lunte, S. M. (2003) Comparison of the performance characteristics of poly(dimethylsiloxane) and Pyrex microchip electrophoresis devices for peptide separations. J. Chrom. A 1004, 225–235. 8. Chen, H., Yong, Z., Bi-Feng, L., Dai-Wen, P., and Jie-Ke, C. (2002) Influence of soluble polymer polyvinylpyrrolidone on separation of small peptides and amino acids by microchip-based capillary electrophoresis. Anal. Bioanal. Chem. 373, 314–317. 9. Hu, S., Ren, X., Bachman, M., Sims, C. E., Li, G. P., and Albritton, N. (2002) Surface modification of poly(dimethylsiloxane) microfluidic devices by ultraviolet polymer grafting. Anal. Chem. 74, 4177–4123.

Application of Microchip CE to Peptide Analysis

181

10. Lichtenberg, J., de Rooij, N. F., and Verpoorte, E. (2002) Sample pretreatment on microfabricated devices. Talanta 56, 233–266. 11. Ekström, S., Malmström, J., Wallman, L., et al. (2002) On-chip microextraction for proteomic sample preparation of in-gel digests. Proteomics 2, 413–421. 12. Bergkvist, J., Ekström, S., Wallman, L., et al. (2002) Improved chip design for integrated solid-phase microextraction in on-line proteomic sample preparation. Proteomics 2, 422–429. 13. Yu, C., Davey, M. H., Svec, F., and Frechet, J. M. J. (2001) Monolithic porous polymer for on-chip solid-phase extraction and preconcentration prepared by photoinitiated in situ polymerization within a microfluidic device. Anal. Chem. 73, 5088–5096. 14. Huynh, B., Fogarty, B., Martin, R. S., and Lunte, S. (2004) On-line coupling of microdialysis sampling with microchip-based capillary electrophoresis. Anal. Chem. 76, 6440–6447. 15. Huynh, B., Fogarty, B., Martin, R. S., and Lunte, S. (2004) Interfacing microdialysis to CE for near real-time monitoring of biological events. In: 17th International Symposium on Microscale Separations and Capillary Electrophoresis. Salzburg, Austria. 16. Wang, C., Oleschuk, R., Ouchen, F., Li, J., Thibault, P., and Harrison, D. J. (2000) Integration of immobilized trypsin bead beds for protein digestion within a microfluidic chip incorporating capillary electrophoresis separations and an electrospray mass spectrometry interface. Rapid Commun. Mass Spectrom. 14, 1377–1383. 17. Jin, L. J., Ferrance, J., Sanders, J. C., and Landers, J. P. (2003) A microchip-based proteolytic digestion system driven by electroosmotic pumping. Lab Chip 3, 11–18. 18. Gao, J., Xu, J. D., Locascio, L. E., and Lee, C. S. (2001) Integrated microfluidic system enabling protein digestion, peptide separation, and protein identification. Anal. Chem. 73, 2648–2655. 19. Jiang, Y. and Lee, C. S. (2001) On-line coupling of micro-enzyme reactor with micro-membrane chromatography for protein digestion, peptide separation, and protein identification using electrospray ionization mass spectrometry. J. Chrom. A 924, 315–322. 20. Peterson, D. S., Rohr, T., Svec, F., and Frechet, J. M. J. (2002) Enzymatic microreactor-on-a-chip: protein mapping using trypsin immobilized on porous polymer monoliths molded in channels of microfluidic devices. Anal. Chem. 74, 4081–4088. 21. Sakai-Kato, K., Kato, M., and Toyo’oka, T. (2003) Creation of an on-chip enzyme reactor by encapsulating trypsin in sol-gel on a plastic microchip. Anal. Chem. 75, 388–393. 22. Kim, Y. D., Park, C. B., and Clark, D. S. (2001) Stable sol-gel microstructured and microfluidic networks for protein patterning. Biotechnol. Bioeng. 73, 331–337. 23. Gottschlich, N. (2000) Integrated microchip-device for the digestion, separation and postcolumn labeling of proteins and peptides. J. Chrom. B 745, 243–249.

182

Fogarty, Lacher, and Lunte

24. Bruin, G. J. (2000) Recent developments in electrokinetically driven analysis on microfabricated devices. Electrophoresis 21, 3981–3951. 25. Gawron, A. J., Martin, R. S., and Lunte, S. M. (2001) Microchip electrophoretic separation systems for biomedical and pharmaceutical analysis. Eur. J. Pharm. Sci. 14, 1–12. 26. Wang, J., Chen, G., and Pumera, M. (2003) Microchip separation and electrochemical detection of amino acids and peptides following precolumn derivatization with naphthalene-2,3-dicarboxyaldehyde. Electroanalysis 15, 862–865. 27. Galloway, M., Stryjewski, W., Henry, A., et al. (2002) Contact conductivity detection in poly(methylmethacylate)-based microfluidic devices for analysis of monoand polyanionic molecules. Anal. Chem. 74, 2407–2415. 28. Gawron, A. J. and Lunte, S. M. (2000) Detection of neuropeptides using on-capillary copper complexation and capillary electrophoresis with electrochemical detection. Electrophoresis 21, 3205–3211. 29. Li, J., Wang, C., Kelly, J. F., Harrison, D. J., and Thibault, P. (2000) Rapid and sensitive separation of trace level protein digests using microfabricated devices coupled to a quadrupole—time-of-flight mass spectrometer. Electrophoresis 21, 198–210. 30. Skoog, D. A., Holler, F. J., and Nieman, T. A. (1998) Capillary electrophoresis and capillary electrochromatograpy. In: Principles of Instrumental Analysis, (Skoog, D. A., Holler, F. J., and Nieman, T. A., eds.) Harcourt Brace College Publishers, Philadelphia, PA, pp. 778–795. 31. Weinberger, R. (1993) Practical Capillary Electrophoresis, 1st edition, Academic Press, New York, NY, pp. 312. 32. Terabe, S. T., Otsuka, K., and Ando, T. (1985) Electrokinetic chromatography with miceller solution and open tubular capillary. Anal. Chem. 57, 834–841. 33. Heiger, D. N. (1992) High Performance Capillary Electrophoresis: An Introduction. Hewlett-Packard Company Paris, France, pp. 136. 34. Wainright, A., Williams, S. J., Ciambrone, G., Xue, Q., Wei, J., and Harris, D. (2002) Sample pre-concentration by isotachophoresis in microfluidic devices. J. Chrom. A 979, 69–80. 35. Kurnik, R. T., Boone, T. D., Nguyen, U., Ricco, A. J., and Williams, S. J. (2003) Use of floating electrodes in transient isotachophoresis to increase the sensitivity of detection. Lab Chip. 3, 86–92. 36. Hjerten, S. and Zhu, M. D. (1985) Adaptation of the equipment for high-performance electrophoresis to isoelectric focusing. J. Chrom. 346, 265. 37. Hjerten, S., Elenbring, K., Kilar, F., et al. (1987) Carrier-free zone electrophoresis, displacement electrophoresis and isoelectric focusing in a high-performance electrophoresis apparatus. J. Chrom. 403, 47–61. 38. Kilar, F. and Hjerten, S. (1989) Separation of the human transferrin isoforms by carrier-free high-performance zone electrophoresis and isoelectric focusing. J. Chrom. 480, 351–357. 39. Kilar, F. and Hjerten, S. (1989) Fast and high resolution analysis of human serum transferrin by high performance isoelectric focusing in capillaries. Electrophoresis 10, 23–29.

Application of Microchip CE to Peptide Analysis

183

40. Schwartz, H. and Pritchett, T. (1994) Separation of Proteins and Peptides by Capillary Electrophoresis: Application to Analytical Biotechnology, Vol. V, Beckman Instruments, Fullerton, CA. 41. Landers, J., ed. (1997) Handbook of Capillary Electrophoresis, 2nd edition, CRC Press, New York, NY, pp. 894. 42. Hofmann, O., Che, D., Cruickshank, K. A., and Muller, U. R. (1999) Adaptation of capillary isoelectric focusing to microchannels on a glass chip. Anal. Chem. 71, 678–686. 43. Mao, Q. and Pawliszyn, J. (1999) Demonstration of isoelectric focusing on an etched quartz chip with UV absorption imaging detection. Analyst 124, 637–641. 44. Raisi, F., Belgrader, P., Borkholder, D. A., et al. (2001) Microchip isoelectric focusing using miniature scanning detection system. Electrophoresis 22, 2291–2295. 45. Sanders, J. C., Huang, Z., and Landers, J. P. (2001) Acousto-optical deflectionbased whole channel scanning for microchip isoelectric focusing with laser-induced fluorescence detection. Lab Chip 1, 167–172. 46. Walhagen, K., Unger, K. K., and Hearn, M. T. W. (2000) Capillary electroosmotic chromatography of peptides. J. Chrom. A 887, 165–185. 47. Jemere, A. B., Oleschuck, R. D., Ouchen, F., Fajuyigbe, F., and Harrison, D. J. (2002) An integrated SPE system for sub-picomolar detection. Electrophoresis 23, 3537. 48. He, B., Ji, J., and Regnier, F. E. (1999) Capillary chromatography of peptides in a microfabricated system. J. Chrom. A 853, 257–262. 49. Slentz, B. E., Penner, N. A., Lugowska, E., and Regnier, F. (2001) Nanoliter capillary electrochromatography columns based on collocated monolithic supposrt structures molded in poly(dimethylsiloxane). Electrophoresis 22, 3736–3743. 50. Slentz, B. E., Penner, N. A., and Regnier, F. (2002) Sampling BIAS at channel junctions in gated flow injection on chips. Anal. Chem. 74, 4835–4840. 51. Svec, F. (2003) Porous monoliths: the newest generation of stationary phases for HPLC and related methods. LC GC Europe June, 2–6. 52. Throckmorton, D. J., Shepodd, T. J., and Singh, A. K. (2002) Electrochromatography in microchips: reversed-phase separation of peptides and amino acids using photopatterned rigid polymer monoloths. Anal. Chem. 74, 784–789. 53. Breadmore, M. C. (2003) Towards a microchip-based chromatographic platform. Part 2: sol-gel phases modified with polyelectrolyte multilayers for capillary electrochromatography. Electrophoresis 24, 1261–1270. 54. Guijt, R. M., Baltussen, E., and van Dedem, G. W. (2002) Use of bioaffinity interactions in electrokinetically controlled assays on microfabricated devices. Electrophoresis 23, 823–835. 55. Abad-Villar, E. M., Tanyanyiwa, J., Fernandez-Abedul, M. T., Costa-Garcia, A., and Hauser, P. C. (2004) Detection of human immunoglobulin in microchip and conventional capillary electrophoresis with contactless conductivity measurements. Anal. Chem. 76, 1282–1288. 56. Martynova L, Locascio L. E., Gaitan, M., Kramer, G. W., Christensen, R. G., and MacCrehan, W. A. (1997) Fabrication of plastic microfluid channels by imprinting methods. Anal. Chem. 69, 4783–4789.

184

Fogarty, Lacher, and Lunte

57. von Heeren, F., Verpoorte, E., Manz, A., and Thormann, W. (1996) Micellar electrokinetic chromatography separations and analyses of biological samples on a cyclic planar microstructure. Anal. Chem. 68, 2044–2053. 58. Chiem, N. and Harrison, D. J. (1997) Microchip-based capillary electrophoresis for immunoassays: analysis of monoclonal antibodies and theophylline. Anal. Chem. 69, 373–378. 59. Chiem, N. H. and Harrison, D. J. (1998) Microchip systems for immunoassay: an integrated immunoreactor with electrophoretic separation for serum theophylline determination. Clin. Chem. 44, 591–598. 60. Qiu, C. X. and Harrison, D. J. (2001) Integrated self-calibration via electrokinetic solvent proportioning for microfluidic immunoassays. Electrophoresis 18, 3949–3958. 61. Roper, M. G., Shackman, J. G., Dahlgren, G. M., and Kennedy, R. T. (2003) Microfluidic chip for continuous monitoring of hormone secretion from live cells using an electrophoresis-based immunoassay. Anal. Chem. 75, 4711–4717. 62. Koutny, L. B., Schmalzing, D., Taylor, T. A., and Fuchs, M. (1996) Microchip electrophoretic immunoassay for serum cortisol. Anal. Chem. 68, 18–22. 63. Linder, V., Verpoorte, E., de Rooij, N. F., Sigrist, H., and Thormann, W. (2002) Application of surface biopassivated disposable poly(dimethylsiloxane)/glass chips to a heterogeneous competitive human serum immunoglobulin G immunoassay with incorporated internal standard. Electrophoresis. 23, 740–749. 64. Starkey, D. E., Abdelaziez, Y., Ahn, C. H., et al. (2003) Determination of endogenous extracellular signal-regulated protein kinase by microchip capillary electrophoresis. Anal. Biochem. 316, 181–191. 65. Cohen, C. B., Chin-Dixon, E., Jeong, S., and Nikiforov, T. T. (1999) A microchipbased enzyme assay for protein kinase A. Anal. Biochem. 273, 89–97. 66. Xue, Q., Wainright, A., Gangakhedkar, S., and Gibbons, I. (2001) Multiplexed enzyme assays in capillary electrophoretic single-use microfluidic devices. Electrophoresis 18, 4000–4007. 67. Song, J. M., Guy, D., and Vo-Dinh, T. (2003) Application of an integrated microchip system with capillary array electrophoresis to optimization of enzymatic reactions. Anal. Chim. Acta 487, 75–82. 68. Wang, J., Ibanez, A., Chatrathi, M. P., and Escarpa, A. (2001) Electrochemical enzyme immunoassays on microchip platforms. Anal. Chem. 73, 5323–5327. 69. Hadd, A. G., Jacobson, S. C., and Ramsey, J. M. (1999) Microfluidic assays of acetylcholinesterase inhibitors. Anal. Chem. 71, 5206–5212. 70. Rocklin, R. D., Ramsey, R. S., and Ramsey, J. M. (2000) A microfabricated fluidic device for performing two-dimensional liquid-phase separations. Anal. Chem. 72, 5244–5249. 71. Yang, X., Zhang, X., Li, A., Zhu, S., and Huang, Y. (2003) Comprehensive twodimensional separations based on capillary high-performance liquid chromatography and microchip electrophoresis. Electrophoresis 24, 1451–1457. 72. Gottschlich, N., Jacobson, S. C., Ramsey, R. S., and Ramsey, J. M. (2001) Twodimensional electrochromatography/capillary electrophoresis on a microchip. Anal. Chem. 73, 2669–2674.

Application of Microchip CE to Peptide Analysis

185

73. Herr, A. E., Molho, J. I., Drouvalakis, K. A., et al. (2003) On-chip coupling of isoelectric focusing and free solution electrophoresis for multidimensional separations. Anal. Chem. 75, 1180–1187. 74. Slentz, B. E., Penner, N. A., and Regnier, F. E. (2003) Protein proteolysis and the multi-dimensional electrochromatographic separation of histidine-containing peptide fragments on a chip. J. Chrom. A 984, 97–107. 75. Lakowicz, J. R. (1983) Principles of Fluorescent Spectroscopy. Plenum Press, New York. 76. Gavin, P. F. and Ewing, A. G. (1997) Characterization of electrochemical array detection for continuous channel electrophoretic separations in micrometer and submicrometer channels. Anal. Chem. 69, 3838–3845. 77. Martin, R. S., Gawron, A. J., and Lunte, S. M. (2000) Dual-electrode electrochemical detection for poly(dimethylsiloxane)-fabricated capillary electrophoresis microchips. Anal. Chem. 72, 3196–3202. 78. Schwarz, M. A. and Hauser, P. C. (2001) Recent developments in detection methods for microfabricated analytical devices. Lab Chip 1, 1–6. 79. Martin, R. S., Gawron, A. J., Fogarty, B. A., Regan, F. B., Dempsey, E., and Lunte, S. M. (2001) Carbon paste-based electrochemical detectors for microchip capillary electrophoresis/electrochemistry. Analyst 126, 277–280. 80. Schwarz, M. A., Galliker, B., Fluri, K., Kappes, T., and Hauser, P. C. (2001) A two-electrode configuration for simplified amperometric detection in a microfabricated electrophoretic separation device. Analyst 126, 147–151. 81. Martin, R. S., Ratzlaff, K. L., Huynh, B. H., and Lunte, S. M. (2002) In-channel electrochemical detection for microchip capillary electrophoresis using an electrically isolated potentiostat. Anal. Chem. 74, 1136–1143. 82. Wang, J., Pumera, M., Chatrathi, M. P., et al. (2002) Towards disposable lab-ona-chip: poly(methylmethacrylate) microchip electrophoresis device with electrochemical detection. Electrophoresis 23, 596–601. 83. Gawron, A. J., Martin, R. S., and Lunte, S. M. (2001) Fabrication and evaluation of a carbon-based dual-electrode detector for poly(dimethylsiloxane) electrophoresis chips. Electrophoresis 22, 242–248. 84. Allison, A. L., Mayer, G. S., and Shoup, R. E. (1984) The o-phthalaldehyde derivatives of amines for high-speed liquid chromatography/electrochemistry. Anal. Chem. 56, 1089–1096. 85. Woltman, S. J., Chen, J. G., Weber, S. G., and Tolley, J. O. (1995) Determination of the pharmaceutical peptide TP9201 by post-column reaction with copper(II) followed by electrochemical detection. Pharm. Biomed. Anal. 14, 155–164. 86. Chen, J. -G., Woltman, S. J., and Weber, S. G. (1996) Electrochemical detection of biomolecules in liquid chromatography and capillary electrophoresis. In: Advances in Chromatography, (Brown, P. R. and Grushka, E., eds.) Marcel Dekker, New York, NY, pp. 273–314. 87. Guijt, R. M., Baltussen, E., van der Steen, G., et al. (2001) Capillary electrophoresis with on-chip four-electrode capacitively coupled conductivity detection for application in bioanalysis. Electrophoresis 22, 2537–2541.

186

Fogarty, Lacher, and Lunte

88. Limbach, P. A. and Meng, Z. (2002) Integrating micromachined devices with modern mass spectroscopy. Analyst 127, 693–700. 89. Pinto, D. M., Ning, Y., and Figeys, D. (2000) An enhanced microfluidic chip coupled to an electrospray Qstar mass spectrometer for protein identification. Electrophoresis 21, 181–190. 90. Li, J., Kelly, J. F., Chernushevich, I., Harrison, D. J., and Thibault, P. (2000) Separation and identification of peptides from gel-isolated membrane proteins using a microfabricated device for combined capillary electrophoresis/nanoelectrospray mass spectrometry. Anal. Chem. 72, 599–609. 91. Broyles, B. S., Jacobson, S. C., and Ramsey, J. M. (2003) Sample filtration, concentration, and separation integrated on microfluidic devices. Anal. Chem. 75, 2761–2767. 92. Breadmore, M. C., Wolfe, K. A., Arcibal, I. G., et al. (2003) Microchip-based purification of DNA from biological samples. Anal. Chem. 75, 1880–1886. 93. Lazar, I. M., Ramsey, R. S., and Ramsey, J. M. (2001) On-chip proteolytic digestion and analysis using “wrong-way-round” electrospray time-of-flight mass spectrometry. Anal. Chem. 73, 1733–1739. 94. Wolfe, K. A., Breadmore, M. C., Ferrance, J. P., et al. (2002) Toward a microchipbased solid-phase extraction method for isolation of nucleic acids. Electrophoresis 23, 727–733. 95. Tachibana, Y., Otsuka, K., Terabe, S., Arai, A., Suzuki, K., and Nakamura, S. (2004) Effects of the length and modification of the separation channel on microchip electrophoresis-mass spectrometry for analysis of bioactive compounds. J. Chrom. A 1025, 287–296.

13 Measuring Electroosmotic Flow in Microchips and Capillaries S. Douglass Gilman and Peter J. Chapman Summary Electrophoretic migration and electroosmotic flow (EOF) combine to determine the migration rate of charged compounds in capillary electrophoresis (CE) and microchip capillary electrophoresis (MCE). Uncontrolled and unmeasured changes in EOF will lead to irreproducible peak migration times and poor peak quantitation. The two most common methods for measuring EOF for CE and MCE are detailed. Experimental results for application of the neutral marker method and the current monitoring method to EC are presented, and related calculations of EOF rates and electroosmotic mobility are described. The strengths and shortcomings of these two EOF measurement techniques are discussed. Additional approaches for studying and measuring EOF and for improving the reproducibility of migration times for CE and MCE are summarized. Key Words: Electroosmotic flow; microchip capillary electrophoresis; capillary electrophoresis; neutral marker; current monitoring.

1. Introduction Capillary electrophoresis (CE) and microchip capillary electrophoresis (MCE) are attractive options for carrying out rapid analytical separations of complex samples. These techniques offer high resolution and large peak capacities. The selectivity of these separation techniques can be altered dramatically by making simple changes to the separation solution (pH, ionic strength, buffer salts, surfactants, organic solvents, and others). The lure of CE and MCE for carrying out difficult separations is countered by two significant limitations—poor reproducibility, particularly relative to high performance liquid chromatography (HPLC), and the small scale of the techniques, which is both an advantage and a major limitation of CE and MCE. Many authors have examined CE reproducibility and its underlying limits in detail (1–3). Electroosmotic flow (EOF) is a primary characteristic of CE and MCE, and this phenomenon is intimately linked to both the major advantages and From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

187

188

Gilman and Chapman

disadvantages of these separation techniques. The velocity of an analyte zone in a CE/MCE separation is described by the following equation: vnet = unet E = (uep + ueof ) E

(1)

where vnet is the net velocity of a compound owing to electrophoresis and EOF, unet is the net electrophoretic mobility, E is the applied field strength (V/cm), uep is the electrophoretic mobility of the analyte molecule, and ueof is the electroosmotic mobility. Equation 1 shows that analyte migration and consequently, peak retention times, are directly affected by EOF. It is well known that the magnitude of EOF is often irreproducible and unstable because of changes to the surface and solution chemistry in a capillary or microchannel (discussed next) (1–4). EOF is generated at the capillary or microchannel surface in the presence on an applied potential by electrophoretic migration of solvated ions near the surface. If the inner surface of a capillary or microchannel includes charged functional groups (e.g., Si-O– for fused silica), a double-layer structure is formed in solution at the surface (typically nm in scale), which contains an excess of ions opposite in charge to the bound surface charge on the inner wall. In the presence of an applied potential, these solvated counter ions in the diffuse part of the double layer will migrate toward one electrode, generating a bulk fluid flow. In effect, the entire inner surface acts as a fluid pump for CE and MCE. For a typical experiment with a fused silica capillary filled with an aqueous solution near neutral pH, EOF is in the direction from the anode (positive polarity) to the cathode (negative polarity) and is proportional to the applied potential. Equation 2 describes the electroosmotic mobility in a capillary: ueof =

εζ 4 πη

(2)

The structure of the double layer defines the zeta potential, ζ. The dielectric constant of the buffer filling the capillary is ε, and the viscosity of the solution near the capillary surface is η. All three of these terms contribute to the irreproducibility of EOF. Problematic changes in EOF are often observed between consecutive runs, and EOF can change significantly during the course of a single run (5,6). Changes to the pH of the solution can alter the surface charge, which will affect ζ. Adsorption to the capillary or microchannel surface (particularly of sample components) similarly impacts ζ. The ionic strength of the solution affects the double layer structure and therefore, ζ, as well as ε. Temperature changes and solution composition changes will alter η. Despite the reproducibility problems associated with EOF, it provides significant advantages for CE and MCE. Without EOF it is only possible to separate and detect either anionic molecules or cationic molecules in a single experiment.

Measuring EOF

189

Strong EOF carries most analytes from the injection end to the detection end of a column (|ueof | > |uep |). Because EOF is generated at the capillary or microchannel surface, its flow profile is very flat compared with pressure-driven flow, and it generally does not contribute significantly to peak broadening, preserving the high separation efficiencies and excellent resolution obtained with CE and MCE. An additional advantage of EOF is that it is generated naturally during CE and MCE, and does not require external pumps or high-pressure interfaces. It can be generated and adjusted simply through the applied potential. This advantage is especially important to the concept of portable Lab-ona-Chip devices (7–9). There are two primary solutions to the reproducibility problems associated with EOF. One solution is to suppress EOF by modifying the capillary or microchannel surface. This approach is often used for protein separations, where EOF changes owing to surface adsorption are particularly common (10–12). This is also used commonly for DNA separations, where all analytes are anions and simultaneous analysis of cationic species is not an important consideration (13). A second approach that maintains the advantages of EOF is measuring flow to correct for changes in EOF and to indicate to the user that it is unstable so that steps can be taken to make EOF, and the experiment, more reproducible. The measurement of EOF is experimentally challenging because of the small scale of CE and MCE. Practical commercial devices to continuously monitor EOF and maintain constant flow at the scale of CE and MCE are simply not available today. In this chapter we will present the two most common methods used to measure EOF in CE and MCE. 1.1. Neutral Marker Method The neutral marker method is the most common technique used for measuring EOF in CE, and it is also applied frequently to MCE. This technique was developed in the early days of CE (14,15). In the neutral marker method, a short zone of a neutral compound is injected and detected using the analyte detector. This can be carried out in an experiment separate from the separation, but it is most useful if the neutral marker is added to the analytical sample or injected with it. The primary advantage of this approach is simplicity. No additional equipment or complicated methodology is required relative to carrying out an analytical separation by CE or MCE. Data analysis and calculations (see Subheading 3.1.3.) are straightforward. There are two important disadvantages of the neutral marker method. First, it only provides an average EOF value from the time the neutral marker is injected until the time it is detected. Secondly, the average EOF value obtained does not include the time after neutral marker detection, but that time window often contains most of the analyte peaks (anionic species when the injection end is the anode). Despite

190

Gilman and Chapman

these limitations, the neutral marker method is the most common and most reliable method for measuring EOF. 1.2. Current Monitoring Method The current monitoring method was developed as an alternative to the neutral marker method to measure EOF for CE (16), and it is still applied commonly to CE today. This technique is, however, the dominant method for measuring EOF in MCE. The current monitoring method measures the time required for a dilute buffer to displace a more concentrated buffer (typically) in the entire capillary or microchannel. The dilute buffer is less conductive than the concentrated buffer, so the electrophoretic current decreases as the dilute buffer replaces the concentrated buffer. When the dilute buffer has filled the entire capillary or microchannel, the current stops changing at a new, lower value. Ideally, the dilute buffer fills the capillary or microchannel at the rate of EOF for the undiluted separation buffer. The current monitoring method provides essentially the same information as the neutral marker method, and it has the same primary advantages and disadvantages. One disadvantage of the current monitoring method relative to the neutral marker method is that it is not normally performed during the same experiment as the separation of the sample of interest. In addition, the current monitoring method itself can cause flow changes in a microchip device, and cross channels can complicate these EOF measurements (5). For MCE, current monitoring is the most common EOF measuring method because it is even simpler experimentally than the neutral marker method. The neutral marker method requires injection of a narrow sample zone containing the neutral compound, and the neutral marker must be detected. In addition, the neutral marker must not interact with the microchannel surface, and this is particularly problematic with microchip devices made from some polymers (5). The current monitoring technique does not require a noninteracting neutral molecule, sample injection, or analyte detection. EOF can be measured in microchip devices without mastering analyte injection techniques and developing an analyte detection method. These injection and detection procedures can be more difficult operations for MCE compared with CE. 1.3. Other Techniques for Studying EOF EOF in capillaries and microchips has been studied using a number of other approaches, but today these methods are not yet practical for routine measurement of EOF, and they are not considered further in this chapter. Flow has been imaged in capillaries and microchannels using a variety of techniques (17–26). One theme of these studies has been imaging the flow profile of EOF in CE and MCE (17,18,22–25). Flow imaging is used regularly in MCE to examine the fluid flow through increasingly complicated microchip devices and to better

Measuring EOF

191

characterize more routine operations like sample injection or migration through turns in channels (19,20,24,26). Methods to continuously monitor EOF in CE and MCE have been under development since the 1980s, and work in this area continues today (5,6,27–33). Several groups have recently reported new methods to continuously monitor EOF, but today these techniques are only being applied as basic research tools to study EOF (5,6,29,32). A long-range goal of this work is to develop routine devices and methods to overcome the limitations of the neutral marker and current monitoring methods for the typical CE and MCE user. 2. Materials Measuring average EOF using either the neutral marker method or the current monitoring method does not require highly specialized instrumentation. The same detector used to detect analytes for CE/MCE is typically used to detect a neutral marker compound. Most commercial instruments routinely record the electrophoretic current during a CE separation. With the exception of the neutral marker compound, the same chemicals and solutions used to perform CE/MCE separations are appropriate for measuring EOF using the neutral marker and current monitoring methods. Experimental simplicity is the primary advantage of these two techniques. 2.1. Neutral Marker Method The experiments presented in this chapter were carried out with a laboratoryconstructed CE instrument equipped with a commercial capillary ultraviolet (UV) absorbance detector (Acutect 500, Fisher Scientific; Pittsburgh PA). These same experiments can be carried out with any commercial CE instrument. The same fused silica capillaries were used for the neutral marker method and the current monitoring method. The capillaries had an inner diameter of 50 µm and an outer diameter of 220 µm (SGE; Austin, TX) with a total length of 65.0 cm. A window for neutral marker and analyte detection was created by burning the polyimide coating from approx 1 cm of the capillary with a low gas flame at a distance of 40.0 cm from the injection end of the capillary. The detection wavelength was 210 nm, and the rise time for the absorbance detector was 0.1 s. The injection end of the capillary was housed in a Plexiglass enclosure designed to prevent exposure of the operator to the applied separation potential. The electrophoretic potential was applied using a CZE1000R high-voltage power supply (Spellman; Hauppauge, NY). Platinum wire electrodes were used at both the injection and detection reservoirs. The injection end electrode was connected to the high-voltage power supply, and the detection end electrode was grounded. Standard glass vials or 1.5-mL plastic microcentrifuge tubes were used as injection and detection buffer reservoirs. The analog signal from the

192

Gilman and Chapman

absorbance detector was filtered with a 50-Hz low-pass filer and then converted to a digital signal at 10 Hz and saved to a file using an analog-to-digital board (Lab-PC-1200/AI; National Instruments; Austin, TX) and a LabView program (National Instruments) written in-house. Data were analyzed and plotted using Excel (Microsoft; Redmond, WA). The electrophoretic current was also recorded during these experiments as described in Subheadings 2.2. and 3.2. Dibasic sodium phosphate was obtained from Mallinckrodt (Phillipsburg, NJ), and mesityl oxide was purchased from Sigma (St. Louis, MO). All solutions were prepared from doubly distilled water. Phosphate buffer (10.00 mM at pH 7.04) was prepared by addition of dibasic sodium phosphate to water. The pH was adjusted to 7.04 by addition of 2 M HCl while monitoring the pH with a pH electrode. All buffers were filtered with a 0.45-µm filter before use to remove particulates and bacterial contamination. A neutral marker solution of 9 × 10–4 M mesityl oxide in the separation buffer (10.00 mM phosphate, pH 7.04) was prepared by adding mesityl oxide to the phosphate buffer. 2.2. Current Monitoring Method Current monitoring experiments were conducted using the same instrumentation, data collection hardware and software, and chemicals described in Subheading 2.1. For measurement of the electrophoretic current, the wire from the platinum electrode (cathode, grounded) first passed through a 100 kΩ resistor. The potential across this resistor was detected to monitor the electrophoretic current (V = IR, 1.00 V ⇒ 10.0 µA). The electrophoretic current data were collected and analyzed using the same hardware, software, and conditions used to collect absorbance data as described in Subheading 2.1. All solutions were prepared from doubly distilled water and filtered as described in Subheading 2.1. Separation buffer solutions were diluted for current monitoring experiments with doubly distilled water by either 10% (9.00 mM phosphate buffer at pH 7.04) or 50% (5.00 mM phosphate buffer at pH 7.04). 3. Methods The experimental steps required for the neutral marker and the current monitoring methods are relatively simple, and related data analysis and calculations are straightforward. These methods deviate little from the procedures for carrying out CE and MCE separations. It is important, however, for the user to understand the limitations of these techniques and the potential pitfalls one can encounter when using these methods. Simple demonstrations of each method for CE are presented in this section, as well as a discussion of the data analysis.

Measuring EOF

193

Fig. 1. Electropherogram for a neutral marker experiment to measure electroosmotic flow. Mesityl oxide was injected for 3.0 s at 25.0 kV and detected by absorbance at 210 nm. The electrophoretic potential was 25.0 kV. Table 1 Electroosmotic Flow Measurements Method Neutral marker Current monitoring (10 mM, 9 mM) Current monitoring (10 mM, 5 mM)

veof (cm/s)

µeof (×10–4 cm2/Vs)

195.9 ± 0.2a 317 ± 11b

0.2042 ± 0.0002 0.205 ± 0.007

5.308 ± 0.006 5.3 ± 0.2

270 ± 11b

0.241 ± 0.010

6.3 ± 0.3

t (s)

at . nm bt . cm

3.1. Neutral Marker Method Figure 1 shows an electropherogram of a neutral marker measurement of EOF. This experiment was repeated three times to generate the data reported in Table 1. The following procedure was used for these experiments: 1. Mesityl oxide (9 × 10–4 M) was used as the neutral marker and was prepared in the separation buffer (10.00 mM phosphate, pH 7.04) (see Notes 1 and 2). All injections were electrokinetic (see Note 3) for 3.0 s at 25.0 kV. A single injection of the neutral marker solution was performed for each EOF measurement (see Notes 4 and 5).

194

Gilman and Chapman

2. A separation potential of 25.0 kV was applied (see Note 6), and the neutral marker peak was detected by UV absorbance at 210 nm. 3. The experiment was repeated three times to determine an average migration time for the neutral marker peak, tnm. The EOF velocity, Veof , was calculated using Eq. 3: veof =

Ld tnm

(3)

Here Ld is the distance from the injection end of the capillary to the point of UV absorbance detection on the capillary (40.0 cm here). Equation 4 was used to calculate the electroosmotic mobility, ueof : ueof =

veof Lt V

(4)

In this case, Lt is the total length of the capillary or microchannel (65.0 cm here), and V is the applied voltage (25.0 kV here). The resulting values for the experiments (represented by Fig. 1) are presented in Table 1 (see Notes 7–9). In principle, there are no significant differences between applying the neutral marker method to capillaries and microchip devices, but in practice the current monitoring method is used more commonly for MCE (see Note 10). 3.2. Current Monitoring Method Figure 2 shows a plot of current vs time for a current monitoring measurement of EOF. This experiment was repeated four times to generate the data reported in Table 1. The following procedure was used for these experiments: 1. The capillary was initially filled with the same buffer used in Subheading 3.1. (10.0 mM phosphate buffer at pH 7.04). 2. The buffer in the injection reservoir was removed and replaced with the 10% diluted buffer (9.00 mM phosphate buffer at pH 7.04). The more concentrated buffer (10.00 mM) remained in the capillary. The electrophoretic potential of 25.0 kV was applied, and the electrophoretic current was recorded (see Notes 4 and 5). The current was measured by simply placing a resistor between the detection electrode and ground and measuring the voltage across this electrode (Subheading 2.2.) (see Note 11). 3. The endpoint for this experiment, tcm, is the point at which the current stops decreasing, indicating that the more dilute replacement buffer has completely filled the capillary (see Note 12). Tangential lines as shown in Fig. 2 were used to estimate this point for the data shown in Table 1 (see Note 13). 4. The electroosmotic velocity is calculated using Eq. 5: veof =

Lt tcm

(5)

Measuring EOF

195

Fig. 2. Electrophoretic current vs time for a current monitoring measurement of electroosmotic flow. The separation buffer, 10 mM phosphate (pH 7.04) was replaced by 9 mM phosphate (10% dilution). The electrophoretic potential was 25.0 kV. The dashed lines illustrate the method used to determine the endpoint for this experiment, tcm (drawn below the data points for clarity). Here tcm is the time that the electrophoretic current stops changing as previously described. The total length of the capillary is used because the electrophoretic current is determined by the resistance of the entire capillary (from injection end to detection end). The electroosmotic mobility, ueof , is calculated from Eq. 4 (see Note 7). 5. Identical experiments also were carried out by replacing the 10.00 mM buffer with 50% diluted buffer (5.00 mM phosphate buffer at pH 7.04). Data from these experiments are presented in Fig. 3 and Table 1 (see Note 14).

The current monitoring method is the most common technique used to measure EOF in microchips because of its experimental simplicity. The experiment is initiated by simply replacing the buffer in the injection reservoir for both CE and MCE experiments. However, flow changes owing to the difference of the initial solution and replacement solution have been shown to be a serious concern for current monitoring experiments in MCE (5). Because of the potential for more complicated flow dynamics in microchips with intersecting channels, the recommendations of Zare and coworkers (16) about dilutions for current monitoring are more important for MCE; however, the literature indicates that the MCE community has been less careful about following this advice. The recent work by Pittman et al. (5) indicates that there is a real risk of obtaining

196

Gilman and Chapman

Fig. 3. Electrophoretic current vs time for a current monitoring measurement of electroosmotic flow. The separation buffer, 10 mM phosphate (pH 7.04) was replaced by 5 mM phosphate (50% dilution). The electrophoretic potential was 25.0 kV.

erroneous results when using the current monitoring method for MCE and CE, and application of the neutral marker method is recommended for MCE whenever possible. 4. Notes 1. The exact concentration of the neutral marker is not critical. What is important is that a concentration is used that provides a sufficient signal-to-noise ratio so that the peak is readily detected, but that is low enough that the peak is not off scale. It is also advisable to use a concentration on the lower end of this range to reduce the possibility of obscuring peaks for slightly charged species that might elute close to the neutral peak. Addition of a neutral marker at a very high concentration could also change the local buffer conductivity, resulting in sample stacking behavior (34). 2. Neutral marker selection is both straightforward and critical. A good neutral marker should have three characteristics: (1) it must have no net charge at the pH of the experiment (zwitterionic, net neutral species are acceptable). (2) It should be readily detectable using the same detector used for analytes. (3) It must not interact with the capillary or microchannel surface, and it must not interact with charged compounds in the separation buffer and sample (for this reason the neutral marker method is normally not applicable to separations using buffer additives such as micelles, cyclodextrins, or polymer solutions). 3. Electrokinetic injection was used for these experiments for convenience, but pressure injection, gravity injection, or one of the many injection techniques devel-

Measuring EOF

4.

5.

6.

7.

197

oped for MCE are equally applicable (2,7,35,36). The main requirement is that the injection method provides a narrow zone of the neutral marker at a reproducible distance from the analyte detector so that an accurate and precise migration time can be measured for EOF determination. It is advisable to equilibrate the capillary or microchannel with the running buffer if the capillary is new or if it is being used for the first time during a day of experiments. In the authors’ laboratory this is typically accomplished by applying the separation potential for 15–30 min after initially filling the capillary. Another common approach is to flush the capillary with the running buffer by pressure. Many different procedures for conditioning a capillary or microchannel prior to experiments and between experiments have been published. The goal of all of these is to establish a relatively stable EOF and electrophoretic current before performing measurements. It is also advisable to periodically replace the buffer in the injection and detection reservoirs with fresh buffer because it can change over time owing to electrochemical reactions during electrophoresis and evaporation. There is no simple formula to estimate how frequently the buffer should be refreshed, but the time period at which this should be done will decrease if smaller buffer reservoirs are used and will decrease with increased electrophoretic current. It is important to keep both ends of a capillary at the same height for CE and fluid levels in the injection reservoirs and detection reservoirs at the same height for CE and MCE. If this precaution is not taken, gravity flow will take place in addition to EOF. This will reduce the accuracy of EOF measurements and will reduce separation efficiencies owing to the parabolic flow profile of gravity flow. Ideally, the neutral marker can be added directly to the analytical sample (or injected just before or after the sample). It is important to be certain that the neutral marker does not overlap with any important peaks in the sample or interact with any sample components. If the neutral marker EOF measurement experiments are run separately, the conditions for the EOF measurement should be as close as possible to those of the relevant analytical separation. Identical buffers should be used for the separation and EOF measurement. If the injected sample is dissolved in a buffer different than the separation buffer, the neutral marker compound should also be dissolved in this different buffer. The same separation potential should be used for the analytical separation and EOF measurement because EOF is not always linearly related to the applied potential owing to Joule heating (37,38). Ultimately the electroosmotic mobility (Eq. 4) is more useful than the EOF rate (Eq. 3) because ueof can be used to compare EOF for different columns and channels at different applied fields. If the neutral marker is injected with the analytical sample, the determined value of ueof and the measured migration time of the analyte peak can be used to calculate the analyte electrophoretic mobility, uep, which typically is more reproducible than the analyte migration time. The analyte migration time can be used to calculate unet from Eq. 6. unet =

vnet Lt V

(6)

198

Gilman and Chapman Here vnet is the net migration velocity of the analyte zone and is calculated from Eq. 7: vnet =

8.

9.

10.

11. 12.

13.

14.

Ld tnet

(7)

In this case tnet is the migration time of the analyte zone to the detector. Equation 1 can be used to calculate uep. A limitation of the neutral marker method is that the EOF value determined does not include times after detection of the neutral marker (after tnm). This is problematic if EOF continues to change during the separation after this time. One solution is to use electrophoretic mobility markers in addition to the neutral marker (3,39,40). These markers are compounds with known electrophoretic mobilities for the same experimental conditions. They serve the same function as the neutral marker, but provide information after the elution time of the neutral species. It is often desirable to characterize residual EOF for capillaries or microchannels where EOF has been suppressed by modification of the capillary surface. The basic neutral marker and current monitoring methods are not convenient approaches for this. For example, if the EOF has been suppressed to 1% of its value with a native surface, the time required for an EOF measurement (tnm or tcm) will be 100 times longer than the time to make an EOF measurement with the native surface. This is usually impractical. Several researchers have devised alternative methods for measuring residual EOF that are related to the neutral marker method (41–43). One reason for the predominance of the current monitoring method for MCE is that laser-induced fluorescence is the dominant detection technique for MCE, but fewer appropriate neutral fluorescent molecules are available compared with compounds that can be used for CE with UV detection. A second reason is that although many proven methods exist for making extremely short MCE injections (7,36), making this work properly for a given device, particularly in an inexperienced laboratory, can be more difficult than making a CE injection. Conductivity detection has also been used to measure EOF in MCE using an approach analogous to the current monitoring method (44). The current monitoring method can also be carried out by using a more concentrated buffer to replace the more dilute buffer, but this approach is used less frequently in the literature. In this case, the current will increase as the replacement buffer fills the capillary before reaching a higher, stable level. In principle, a derivative plot of the current monitoring data could also be used to determine this point though this approach appears to be rarely used based on literature reports. For the experiments presented in Fig. 3, the concentration of the replacement buffer (5.00 mM) is 50% lower than the initial buffer (10.00 mM). It is much easier to detect the point where the current stops changing, tcm, when using a more dilute replacement buffer, but it is not advisable to use such a large difference in buffer concentration for current monitoring. As pointed out by Zare and coworkers in

Measuring EOF

199

their 1988 paper, the ideal concentration change for performing the current monitoring method is the smallest change in concentration for which the end point can be detected accurately and precisely (16). They used a 5% change in their 1988 paper, and a 10% change was used for the data in Fig. 2. Of course, the experimental conditions for EOF measurement should be as close as possible to the conditions for the analytical separation as discussed for the neutral marker method in Note 6. The data in Table 1 shows that the current monitoring method produces two different EOF values depending on the concentration of the replacement buffer. This result and the recommendation in the original publication describing this technique are not surprising when one considers Eq. 2. EOF increases with decreasing ionic strength owing to an increase in ζ, so the average EOF determined will increase if a more dilute (50% vs 10%) replacement buffer is used. The data in Table 1 are consistent with this prediction.

References 1. Wielgos, T., Turner, P., and Havel, K. (1997) Validation of analytical capillary electrophoresis methods for use in a regulated environment. J. Cap. Elec. 4, 273–278. 2. Schaeper, J. P. and Sepaniak, M. J. (2000) Parameters affecting reproducibility in capillary electrophoresis. Electrophoresis 21, 1421–1429. 3. Mayer, B. X. (2001) How to increase precision in capillary electrophoresis. J. Chromatogr. A 907, 21–37. 4. Guiochon, G. (1998) Reflections on analytical separations. Amer. Lab. 30, 14, 15. 5. Pittman, J. L., Henry, C. S., and Gilman, S. D. (2003) Experimental studies of electroosmotic flow dynamics in microfabricated devices during current monitoring experiments. Anal. Chem. 75, 361–370. 6. Pittman, J. L., Gessner, H. J., Frederick, K. A., Raby, E. M., Batts, J. B., and Gilman, S. D. (2003) Experimental studies of electroosmotic flow dynamics during sample stacking for capillary electrophoresis. Anal. Chem. 75, 3531–3538. 7. Polson, N. A. and Hayes, M. A. (2001) Microfluidics controlling fluids in small places. Anal. Chem. 73, 312A–319A. 8. Reyes, D. R., Iossifidis, D., Auroux, P. -A., and Manz, A. (2002) Micro total analysis systems. 1. Introduction, theory, and technology. Anal. Chem. 74, 2623–2636. 9. Auroux, P. -A., Iossifidis, D., Reyes, D. R., and Manz, A. (2002) Micro total analysis systems. 2. Analytical standard operations and applications. Anal. Chem. 74, 2637–2652. 10. Corradini, D. (1997) Buffer additives other than the surfactant sodium dodecyl sulfate for protein separations by capillary electrophoresis. J. Chromatogr. B 699, 221–256. 11. Rodriguez, I. and Li, S. F. Y. (1999) Surface deactivation in protein and peptide analysis by capillary electrophoresis. Anal. Chim. Acta 383, 1–26. 12. Doherty, E. A. S., Meagher, R. J., Albarghouthi, M. N., and Barron, A. E. (2003) Microchannel wall coatings for protein separations by capillary and chip electrophoresis. Electrophoresis 24, 34–54.

200

Gilman and Chapman

13. Mitchelson, K. R. and Cheng, J., eds. (2001). Capillary Electrophoresis of Nucleic Acids Volumes I and II. Vol. 162–163. Humana Press, Totowa, NJ. 14. Terabe, S., Otsuka, K., Ichikawa, K., Tsuchiya, A., and Ando, T. (1984) Electrokinetic separations with micellar solutions and open-tubular capillaries. Anal. Chem. 56, 111–113. 15. Lukacs, K. D. and Jorgenson, J. W. (1985) Capillary zone electrophoresis: Effect of physical parameters on separation efficiency and quantitation. J. High Res. Chromatogr. Commun. 8, 407–411. 16. Huang, X., Gordon, M. J., and Zare, R. N. (1988) Current-monitoring method for measuring the electroosmotic flow rate in capillary zone electrophoresis. Anal. Chem. 60, 1837–1838. 17. Taylor, J. A. and Yeung, E. S. (1993) Imaging of hydrodynamic and electrokinetic flow profiles in capillaries. Anal. Chem. 65, 2928–2932. 18. Tsuda, T., Ikedo, M., Jones, G., Dadoo, R., and Zare, R. N. (1993) Observation of flow profiles in electroosmosis in a rectangular capillary. J. Chromatogr. 632, 201–207. 19. Harrison, D. J., Fluri, K., Seiler, K., Fan, Z., Effenhauser, C. S., and Manz, A. (1993) Micromachining a miniaturized capillary electrophoresis-based chemical analysis system on a chip. Science 261, 895–897. 20. Jacobson, S. C., Hergenroder, R., Koutny, L. B., Warmack, R. J., and Ramsey, J. M. (1994) Effects of injection schemes and column geometry on the performance of microchip electrophoresis devices. Anal. Chem. 66, 1107–1113. 21. Preisler, J. and Yeung, E. S. (1996) Characterization of nonbonded poly(ethylene oxide) coating for capillary electrophoresis via continuous monitoring of electroosmotic flow. Anal. Chem. 68, 2885–2889. 22. Paul, P. H., Garguilo, M. G., and Rakestraw, D. J. (1998) Imaging of pressureand electrokinetically driven flows through open capillaries. Anal. Chem. 70, 2459–2467. 23. Herr, A. E., Molho, J. I., Santiago, J. G., Mungal, M. G., Kenny, T. W., and Garguilo, M. G. (2000) Electroosmotic capillary flow with nonuniform zeta potential. Anal. Chem. 72, 1053–1057. 24. Barker, S. L. R., Ross, D., Tarlov, M. J., Gaitan, M., and Locascio, L. E. (2000) Control of flow direction in microfluidic devices with polyelectrolyte multilayers. Anal. Chem. 72, 5925–5929. 25. Tallarek, U., Rapp, E., Scheenen, T., Bayer, E., and Van As, H. (2000) Electroosmotic and pressure-driven flow in open and packed capillaries: velocity distributions and fluid dispersion. Anal. Chem. 72, 2292–2301. 26. Molho, J. I., Herr, A. E., Mosier, B. P., et al. (2001) Optimization of turn geometries for microchip electrophoresis. Anal. Chem. 73, 1350–1360. 27. Altria, K. D. and Simpson, C. F. (1987) High voltage capillary zone electrophoresis: Operating parameters effects on electroendosmotic flows and electrophoretic mobilities. Chromatographia 24, 527–532. 28. Wanders, B. J., van de Goor, T. A. A. M., and Everaerts, F. M. (1993) On-line measurement of electroosmosis in capillary electrophoresis using a conductivity cell. J. Chromatogr. A 652, 291–294.

Measuring EOF

201

29. Lee, T. T., Dadoo, R., and Zare, R. N. (1994) Real-time measurement of electroosmotic flow in capillary zone electrophoresis. Anal. Chem. 66, 2694–2700. 30. St. Claire, J. C. and Hayes, M. A. (2000) Heat index flow monitoring in capillaries with interferometric backscatter detection. Anal. Chem. 72, 4726–4730. 31. Schrum, K. F., Lancaster, J. M., III, Johnston, S. E., and Gilman, S. D. (2000) Monitoring electroosmotic flow by periodic photobleaching of a dilute, neutral fluorophore. Anal. Chem. 72, 4317–4321. 32. Pittman, J. L., Schrum, K. F., and Gilman, S. D. (2001) On-line monitoring of electroosmotic flow for capillary electrophoretic separations. Analyst 126, 1240–1247. 33. Markov, D. A. and Bornhop, D. J. (2001) Nanoliter-scale non-invasive flow-rate quantification using micro-interferometric back-scatter and phase detection. Fresenius J. Anal. Chem. 371, 234–237. 34. Chien, R. -L. and Burgi, D. S. (1992) On-column sample concentration using field amplification in CZE. Anal. Chem. 64, 489A–496A. 35. Rose, D. J. Jr. and Jorgenson, J. W. (1988) Characterization and automation of sample introduction methods for capillary zone electrophoresis. Anal. Chem. 60, 642–648. 36. Greenwood, P. A. and Greenway, G. M. (2002) Sample manipulation in micro total analytical systems. Tr. Anal. Chem. 21, 726–740. 37. Jorgenson, J. W. and Lukacs, K. D. (1981) Zone electrophoresis in open-tubular glass capillaries. Anal. Chem. 53, 1298–1302. 38. Knox, J. H. and McCormack, K. A. (1994) Temperature effects in capillary electrophoresis. 1: Internal capillary temperature and effect upon performance. Chromatographia 38, 207–214. 39. Lee, T. T. and Yeung, E. S. (1991) Facilitating data transfer and improving precision in capillary zone electrophoresis with migration indices. Anal. Chem. 63, 2842–2848. 40. Jumppanen, J. H. and Riekkola, M. -L. (1995) Marker techniques for high-accuracy identification in CZE. Anal. Chem. 67, 1060–1066. 41. Williams, B. A. and Vigh, G. (1996) Fast, accurate mobility determination method for capillary electrophoresis. Anal. Chem. 68, 1174–1180. 42. Sandoval, J. E. and Chen, S. -M. (1996) Method for the accelerated measurement of electroosmosis in chemically modified tubes for capillary electrophoresis. Anal. Chem. 68, 2771–2775. 43. Ermakov, S. V., Capelli, L., and Righetti, P. G. (1996) Method for measuring very weak, residual electroosmotic flow in coated capillaries. J. Chromatogr. A 744, 55–61. 44. Liu, Y., Wipf, D. O., and Henry, C. S. (2001) Conductivity detection for monitoring mixing reactions in microfluidic devices. Analyst 126, 1248–1251.

14 Single Cell Analysis on Microfluidic Devices Christopher T. Culbertson Summary There is significant variability among cells of the same type at the single cell level. This variability may be because of external stimuli that vary temporally or spatially among a population of cells. It may also be owing to the nonsynchronized responses of cells to various stimuli. In addition, differences in otherwise similar cells may be generated by genetic mutations acquired by one or more of the cells. Often times multiple biochemical pathways and molecules are involved in such differences. In order to better understand these differences and to detect those rare cells in a large population that may be indicative of early disease states, methods that are capable of rapidly quantifying multiple molecular species in single cells are desired. Microfluidic devices may provide the optimal platform upon which to develop such methods. Microfluidics has the capability of combining the high-throughput manipulation and transport of cells with rapid, high-efficiency separations and high-sensitivity detection. This chapter describes how to fabricate microfluidic devices for the high-throughput manipulation and rapid electrical lysis of single, nonadherent (suspension) cells followed by the injection and separation of the fluorescently labeled cell contents. Key Words: Microchips; microfluidic devices; single cell analysis; laser induced fluorescence detection; lymphocytes; Jurkat cells.

1. Introduction The misregulation of cellular processes, which are normally tightly controlled, leads to the manifestation of many diseases including cancers (1). To understand the etiology of such diseases, a fundamental understanding of cell biochemistry, therefore, is necessary. Cells, however, are small and only contain minute quantities of the analytes of interest. Most conventional analytical methods cannot handle volumes on the order of a cell, i.e., 1–3 pL, and so the analytes in the cells generally have to be diluted a thousand- to a millionfold prior to analysis. This often results in analyte concentrations below the detection limit of the method. Much of what is known about cellular biochemistry, therefore, has been gained From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

203

204

Culbertson

by pooling large numbers of cells, lysing them, and then analyzing the lysate to determine the presence/absence and/or concentrations of the analytes of interest (2). This approach has worked well for deciphering many fundamental cellular processes that respond slowly to external stimuli. Cell survival, however, also depends upon a cell’s ability to quickly respond to certain external conditions. The processes involved in such rapid responses can be affected (activated/ deactivated) during the previously mentioned pooling and lysing procedures, thereby giving misleading results. In addition, rapid cellular response mechanisms are often not synchronized among a population of cells, so the results obtained in a pooled population may not be representative of the actual process itself (2–4). Signal transduction pathways that are regulated by kinases are good examples of cellular processes that change rapidly because of external stimulation (3). The concentrations of such molecules can change by an order of magnitude in less than 1 s (4). These processes, therefore, are best studied at the single cell level where the cell can be kept under tightly controlled environmental conditions prior to being rapidly lysed and analyzed. Several techniques have been developed to analyze single cells. These include flow cytometry, fluorescence microscopy, electrochemical methods using microelectrodes (5–10), high performance liquid chromatography (HPLC) (5–8), and capillary electrophoresis (CE) (2,5,9,11). Flow cytometry and fluorescence microscopy are both limited in the number of analytes that they can examine simultaneously to approximately four. This limitation is owing to the wide spectral bandwidths of the fluorescent dyes used. CE does not suffer from the spectral bandwidth limitations of flow cytometry and fluorescence microscopy because analytes can be separated prior to detection. CE throughput is, however, lower than that of flow cytometry. Often, only 8–10 cells can be analyzed in a day. The reasons for this low throughput have to do with the manual manipulations that have to be performed and the long CE run times. This also makes the technique both tedious and expensive in terms of time and equipment needed (3,12–14). Microfluidic devices potentially have several advantages over conventional CE when it comes to performing single cell analysis (15). First, these devices are able to precisely move and control cells throughout a complex channel manifold. Second, a large number of cells can be loaded into a reservoir and automatically introduced into the separation channel in a sequential fashion. Third, cell processing and the loading of cells with test analytes can be integrated automatically with cell handling, lysing, the injection of the lysate into a separation channel, and the separation of the lysate contents. Fourth, lysate separations can be carried out much more quickly because of the shorter injection plug lengths and the higher field strengths available on microfluidic devices. Fifth, because of the flat interfaces, sensitive optical

Single Cell Analysis on Microfluidic Devices

205

detection can be more easily performed. Finally, heating elements can be integrated into the microchips to keep the cells at the appropriate temperature. Recently, several papers demonstrating assays performed on microfluidic devices using both prokaryotic and eukaryotic cells have been reported (15–26). The single cell analysis protocols discussed next will specifically deal with the fabrication of microfluidic devices and detection systems for the analysis of nonspecific esterase activity in nonadherent mammalian cells (15). 2. Materials (see Note 1) 2.1. Reagents and Consumables 1. Photomask blanks are a convenient substrate from which to fabricate microfluidic devices. They can be ordered precoated with a thin layer of chrome and photoresist. A 4 × 4 in. square substrate is a good size. From this substrate eight 2 × 1-in. chips can be fabricated. The photomask blanks can be obtained from a variety of sources including Telic Company (Santa Monica, CA; www.telic2000.com), Nanofilm (Westlake Village, CA; www.Nanofilm.com), and Hoya Corporation (Shelton, CT). They come in a variety of glass types (soda lime, white crown, borofloat, and quartz) and thicknesses. The master grade soda-lime glasses that are 0.060-in. thick are usually the easiest to use. The developing procedure described next is for the general Shipley AZ type photoresists that are usually applied to these substrate blanks. 2. Photoresist developer: Microposit 453 developing solution (Shipley, Inc., Marlboro, MA) (see Note 2). 3. Detergent solution: Fisherbrand Versa-Clean liquid concentrate diluted 30:1 in ultrapure (18 MΩ 0.45-µm filtered) water. 4. Chrome etchant (Transene Co., Inc., Danvers, MA) (see Note 3). 5. Glass etching solution: Transene 1:10 Buffered oxide etch (BOE) (see Note 4). 6. Glass hydrolysis solution: 1 part NH4OH, 4 parts H2O, 1 part H2O2 (see Note 5). 7. Channel Coating Reagents: poly(dimethylsiloxane) (Sylgard 184, Dow Corning), pluronic F-127 (BASF, Mt. Olive, NJ), and hexane. 8. Cells: Jurkat cells ATCC no. TIB-152 (American Type Culture Collection, Rockville, MD). 9. Cell media: RMPI 1640 medium supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 µg/mL penicillin, 100 µg/mL streptomycin. These can be obtained from a wide variety of sources. 10. Separation buffer: 10 mM sodium phosphate, pH 7.4, 50 mM NaCl, 5%(w/v) polyethylene glycol, and 3% pluronic P84 (BASF). 11. Extracellular buffer (ECB): phosphate buffered saline (PBS) with 18% (v/v) Optiprep (Fisher Scientific) and 5 mM tissue grade glucose (see Note 6). 12. Labeling agents (esterase activity reporters): Oregon green 488 carboxylic acid diacetate 6-isomer, carboxy fluorescein diacetate, and Calcein AM (Molecular probes; Eugene OR).

206

Culbertson

2.2. Equipment 1. Flood exposure system: Oriel 87435 (Thermo-Oriel, Long Beach, CA). 2. Laminar flow hood: a Labconco Clean Bench or similar type of hood is important to have access to when fabricating microfluidic devices. 3. Laser: 488-nm Argon Ion lasers can be obtained from several sources. 4. Microscope objective: an inexpensive ×40 extra-long working distance (ELWD) microscope objective can be obtained from Creative Devices (Neshanic Station, NJ). 5. Photomultiplier tube: Hamamatsu R928 (Hamamatsu; Bridgewater, NJ). 6. Low noise current amplifier: SR570 (Stanford Research Systems, Inc., Sunnyvale, CA). 7. Filters: bandpass filters (530DF30) and dichroic mirrors (505DRLP) can be obtained from Omega Optical (Brattleboro, VT). A 488-nm notch filter can be obtained from Kaiser Optical (Ann Arbor, MI). 8. Multifunction I/O card: 6036E (National Instruments, Austin, TX). 9. Syringe pump: a Harvard 22 syringe pump (Harvard Apparatus, Holliston, MA) or other syringe pump capable of working in the withdrawal mode. 10. Waveform generator: Wavetek 164 (Wavetek, San Diego, CA). 11. AC Amplifier: Trek 609-A (Trek, Inc., Medina, NY).

3. Methods 3.1. Microchip Design and Operation Overview Figure 1 shows the microchip design used for the experiments outlined next. The cells are brought to the cross intersection by a syringe pump, which is applied in the withdrawal mode at the waste reservoir (see Note 7). One hundred and thirty microns prior to the cross intersection, an emulsification reagent is added through a side channel to keep the lipids, which are released by cell lysis, in solution. As the cells pass through the cross intersection they are electrically lysed and their contents are released. These contents are then electrophoretically injected into one of the side channels and separated. The analyte injection into and migration down the side channel are owing to electrophoresis only, as the channel wall coatings that are used to minimize cell adsorption also significantly reduce electroosmotic flow. Application of a subambient pressure to the waste reservoir generates fluid flow in all of the channels of the chip. This fluid flow in the separation channel runs counter to the migration direction of the analytes, and its flow profile is parabolic, which can generate excess band broadening during the separation. To minimize this counter flow and its associated band broadening, the separation channel is designed so that its resistance to flow is about an order of magnitude greater than the flow resistance in the other channels. Such a design also improves the injection efficiency, as the analytes have to be injected and then migrate against this pressure-generated flow.

Single Cell Analysis on Microfluidic Devices

207

Fig. 1. Schematic of single cell analysis chip. A subambient pressure is applied to the waste reservoir (W) to pull cells into the cell channel (C). The cells pass the emulsification (E) channel entrance (IE) where they are pushed to the side of the channel nearest the separation channel. One-hundred and thirty micrometers after passing the emulsification channel entrance they come to the lysis intersection (IL) where they are exposed to an electric field (see Subheading 3.1. for details) and lysed. The lysate is then injected into the separation channel (S) and separated.

The linear fluid velocity in an approximately rectangular channel is given by the following equation: vp =

∆Pd 2 ⎡ 1 16d ⎛ πw ⎞ ⎤ − 5 tanh ⎜ ⎢ ⎝ 2 d ⎟⎠ ⎥⎦ l η ⎣ 12 π w

where ∆P is the pressure drop, d is the channel depth, l is the channel length, η is the solution viscosity, and w is the width of the channel (27). For channels where w > 3d the tanh function is approx 1 and can be dropped. For channels where w > 10d the entire last term in the equation is approx 1. This equation can be used to design the appropriate flow resistances into the channel manifold on the microfluidic device.

208

Culbertson

3.2. Microchip Fabrication 3.2.1. Glass Etching (28–30) 1. Photomasks with the appropriate channel designs can be drawn using AutoCAD LT2000 or another CAD drawing program and saved as .dwg files. Eight 2 × 1-in. channel designs are generally drawn for each 4 × 4 in. mask. The lines on the mask should be 10–20 µm wide. Glass etches isotropically, so when etching channels made from masks with a line width of lm, the width of the channels will be approx 2d + lm wide at the top and lm wide at the bottom where d is the depth of the channel. The .dwg files containing the mask designs can be electronically sent to the variety of photomask manufacturers including HTA Photomask (San Jose, CA, www.htaphotomask.com). 2. Align the photomask with a photomask blank on a flood exposure system with the chrome-plated side touching the resist of the blank substrate and expose for 5 s using a flood exposure system (see Note 8). The exposure time may vary with the photoresist used and also with the age of the lamp. 3. Develop the exposed masks in Microposit 453 for 60 s, rinse with ultrapure (18 MΩ) water for 1 min, and dry under argon or some other dry, inert gas. 4. Etch the exposed chrome on the masks using the chrome etchant for 3–4 min, and then rinse with ultrapure water. 5. Remove undeveloped resist with acetone. Rinse with water and dry using argon. Repeat resist removal procedure if necessary. 6. Place the substrate in a stirred 1:10 BOE solution in an appropriate plastic container. The etch rate is very sensitive to temperature and will change with the age of the solution. Average etch rates are approx 0.25–0.50 µm/min. Remove the substrate periodically from the solution (see Note 9), rinse with water, blow dry with argon, and measure the depth of the channels with a stylus-based profiler (TencorKLA Alphastep IQ or similar instrument). When the channels have been etched to a depth of approx 18 µm, remove the substrate, rinse with water and then 95% ethanol, and dry (see Note 10). 7. Remove the chrome on the substrate by agitating in chrome etchant for approx 10 min. Rinse thoroughly with water. Blow dry. 8. Cut substrate into eight 2 × 1-in. chips using a glass dicing saw. Also cut a blank substrate (i.e., uncoated clean glass) into 2 × 1-in. rectangles for use as the cover plates. On the cover plate, drill access holes to the channels using diamond-coated drill bits or an ultrasonic mill.

3.2.2. Bonding (28–30) 1. Sonicate substrate and cover plate in a detergent solution for 10–20 min, rinse with ultrapure water, and dry. 2. Sonicate in acetone for 10 min and then blow dry. 3. Etch in 1:10 BOE solution for 1 min, rinse with water, and place directly into the heated hydrolysis solution. 4. The hydrolysis solution should be heated to approx 50°C prior to the immersion of the substrates, and the substrates should soak in this solution for 20 min (see Notes 11 and 12).

Single Cell Analysis on Microfluidic Devices

209

5. Remove the substrates from the hydrolysis solution and sonicate in running ultrapure water for 15 min. 6. Remove the substrate and glass while the water is still running and join together. This step should be performed in a laminar flow hood or cleanroom. 7. Let sit at room temperature for 5 min and then place in a 95°C oven for 30 min. After 30 min remove from oven to see if bonding has occurred. If bonding looks good then replace in oven for another 6 h. Nonbonded chips easily come apart. Partially bonded chips usually show a series of Newton’s rings (diffraction patterns) in the areas on the chip where bonding has failed. 8. To permanently anneal the chips, place them in an oven and raise the temperature to the annealing temperature for that type of glass. For soda-lime glasses that temperature is generally 525–550°C, for borofloat glasses it is 625–675°C, and for quartz it is approx 1100°C. The following procedure can be used for soda-lime glasses. Raise the temperature of the oven from 95 to 200°C at a rate of 0.2–0.5°C/min. From 200 to 525°C the temperature can be ramped at 5–10°C/min (see Note 13). Let sit at 525°C for 6 h and then passively cool to room temperature (see Note 14). 9. Upchurch Nanoport® reservoirs (Upchurch Scientific, Oak Harbor, WA) can be added to the chip to increase the fluid volume of the reservoir. These reservoirs also provide a handy way to couple the waste channel to the syringe pump.

3.3. Fluorescence Detection Setup Several fluorescent detection schemes have be reported in the literature (29,31–33). Below is an outline for setting up a basic single point detection system. The basic support structure for an epi-illumination setup (Fig. 2) can be built from equipment obtained from Thorlabs (Newton, NJ; www.thorlabs.com) and Linos Photonics (Milford, MA; www.linos-photonics.com). For Calcein, Oregon green, and carboxyfluorescein, the 488-nm line on an argon ion laser is used to excite the fluorescence. The laser beam is reflected off of a dichroic mirror and focused through an inexpensive ×40 ELWD microscope objective. The same objective is used to collect the fluorescent emission from the dyes. The fluorescent emission is passed through the dichroic and imaged onto a spatial filter (1-mm pinhole). The spatially filtered emission is then passed through a 530-nm bandpass filter and finally detected at a PMT. A 488-nm notch filter can also be placed into the light path for better noise rejection. The signal from the PMT is amplified by the low noise current amplifier and digitized by the multifunction I/O card. To detect the intact cells just prior to lysis, in addition to the lysate in the separation channel, a beam splitter is used to split the light from the laser. The split light is then focused into the microchip channel approx 10–20 µm upstream of the lysis/injection cross. The fluorescence from the intact cells as they pass through the focused laser beam is collected by a second microscope objective, passed through a spatial and spectral filter, and detected by a PMT.

210

Culbertson

Fig. 2. Single-point detection schematic (see Subheading 3.3. for details).

3.4. Channel Wall Coating To prevent cell and cell lysate adhesion to the channel walls, the following channel-coating procedure is used. This procedure also significantly reduces the electroosmotic flow in the separation channel (see Note 15). 1. Prepare the channels in the chip to be coated by flushing 1 M NaOH through the channels for 5 min. Follow this by flushing water through the channels for 5 min. Dry the channels out by pulling air through them. Finally put the chip in a 110°C oven for 2 h. 2. Mix 1 g of Sylgard 184 part A with 0.1 g of Sylgard 184 part B. 3. Dilute this mixture to 20% (v/v) in hexane. 4. Place mixture in all of the reservoirs and apply a vacuum to the waste reservoir for 5 min.

Single Cell Analysis on Microfluidic Devices

211

5. Replace the diluted Sylgard mixture with neat hexane and aspirate through the chip for 10 min to remove excess Sylgard mixture. 6. Dry the chip. 7. Place in a 65°C oven overnight to cure (see Note 16). 8. Just prior to filling the chip with the separation buffer for the cell analysis, aspirate a solution of 30% (v/v) Pluronic F-127 in water through the chip for 20 min.

3.5. Cell Culture (34) 1. A supplemented RMPI 1640 medium is used to culture the Jurkat cells as reported in the Subheading 2.1. 2. An inoculate of cells is added to 10 mL of the supplemented media in 25-mL polystyrene culture flasks. 3. The cultures should be maintained at 37°C with a CO2 concentration of 5%. 4. The cells should be passed when the cell density reaches 106/mL. Cell density can be determined using a hemacytometer (34). 5. Cells should not be used until after the third passage.

3.6. Cell Loading 1. Dyes: Oregon green 488 carboxylic acid diacetate (OGCA-D), carboxyfluorescein diacetate (CF-D), and Calcein AM (C-AM) are all fluorogenic and cell membrane permeable. 2. Take a 1-mL aliquot of cells in media and spin the cells down in a centrifuge. 3. Decant the media and replace with extracellular buffer (see Subheading 2.1.) at room temperature. 4. Gently resuspend the cells at a concentration of about 1 × 106/mL. 5. OGCA-D should be constituted at a concentration of 20 µM in the extracellular buffer containing the cells and incubated at room temperature for 10 min. CF-D and C-AM should be constituted at a concentration of 1 µM in the extracellular buffer containing the cells and incubated at room temperature for 10 min (12,15). At these concentrations most of the dye diffuses into the cell and there is no need to replace the buffer. This step can be performed in the cell reservoir on the chip (see Subheading 3.7., step 2).

3.7. Cell Transport and Lysis 1. Fill all of the reservoirs with the separation buffer and flush through the chip to fill all of the channels. 2. Remove separation buffer from the cell reservoir and add either the loaded or unloaded cells in extracellular buffer. If the cells are unloaded, then add a sufficient amount of dye to bring the final dye concentration in the reservoir to that specified in Subheading 3.6., step 5. Unloaded cells are spun down in a centrifuge and resuspended in extracellular buffer at a concentration of approx 1 × 106/mL (see Note 17). 3. Generate a subambient pressure at the cell waste reservoir using a syringe pump.

212

Culbertson

Fig. 3. Dual point detection example. The bottom trace is from the single point detection system 3 mm down the separation channel from the lysis intersection. The peaks resulting from each individual cell are underlined and numbered. The top trace is the fluorescent signal from the intact cells just prior to lysis. The arrows from the top trace point to the peak envelopes in the bottom trace to which they correspond. The trace has been inverted to more easily see the correspondence between the signals. (Reprinted with permission from ref. 15 © 2003 American Chemical Society.) 4. Increase the subambient pressure until the cells are flowing at a rate of approx 1 mm/s. To measure the cell velocities, they can be imaged as discussed in Subheading 3.8. 5. Apply an electric field across the separation channel using the output from a waveform generator amplified by an AC amplifier (see Subheading 2.2. and Note 18). The electric potential should be applied as a square wave at a frequency of approx 75 Hz. The field strength should vary from approx 450 to 900 V/cm at the nadir and apex of the wave. The time-dependent electric field is used to reduce Joule heating problems. 6. With the first part of the single point detection system located 3–5 mm downstream of the lysis intersection in the separation channel and the second part just above the lysis intersection, a successful experiment results in data which can be seen in Fig. 3 (see Subheading 3.3. and Note 19). The separation should be completed in less than 5 s with migration time reproducibilities for all peaks of less than 1% (15).

3.8. Chip and Cell Imaging To ascertain if the cells are moving properly through the chip and to optimize the lysis and injection efficiencies, the cross intersection of the microfluidic

Single Cell Analysis on Microfluidic Devices

213

device can be imaged. This is performed most conveniently using an inverted microscope equipped with an extra long distance condenser lens and an epifluorescence attachment. Such a setup allows plenty of room above the microscope stage for the microchip and its accessory components. The microchip is simply taped to the glass stage ring and imaged. 3.9. Lysate Injection Efficiency Measurements Lysate injection efficiency can be quantitated by looking for fluorescence in the waste channel when the electric field is on. When the electric field is off, one should be able to detect the fluorescence from intact cells migrating down the waste channel; however, once the voltage has been applied to the chip, then the fluorescence should disappear. If some fluorescence is seen, then the cell flow rate needs to be decreased to allow more time for the lysate to migrate into the injection channel once the cell is lysed. 3.10. Channel Wall Coating Durability The channel wall coating has a limited lifetime. When the analyte peak migration times and peak shapes begin to degrade, then the chip coating needs to be replaced. The coating may last several days before needing to be replaced. To recoat a chip, the old coating is first pyrolyzed at 500°C in an oven for 8 h. The channel is then rinsed with water and dried. The coating procedure in Subheading 3.4. is then repeated. 3.11. Data Collection and Analysis Simple data collection and analysis routines can be written using LabVIEW (National Instruments, Austin, TX) and Igor Pro (Wavemetrics, Lake Oswego, OR). 4. Notes 1. Many of the materials used for the experiments described below entail the use of potentially hazardous substances, so care should be taken in their use. One should be familiar with the Material Safety Data Sheet sheets of each component, use the proper personal protective gear, and perform the experiments in a fume hoods where necessary. 2. The photoresist developer is a strong base, so care should be exercised when using this solution along with wearing the proper personal protection. 3. The glass etching solution contains strong acids, so care should be exercised when using this solution along with wearing the proper personal protection. 4. This solution contains hydrofluoric acid, so extreme care should be exercised when using this solution along with wearing the proper personal protection. 5. This solution contains both a strong base and oxidizer, so care should be exercised when using this solution along with wearing the proper personal protection. The

214

6.

7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17.

18.

19.

Culbertson exothermic nature of the hydrolysis reaction can further raise the temperature upon immersion of the glass slides. The optiprep makes the extracellular buffer isopycnic so that the cells do not settle out in the cell reservoir over time. This helps to keep that rate of cell entry into the main chip channel constant over the course of the entire experiment. The Jurkat cells used for these experiments are very sensitive to electric fields and, therefore, cannot be transported via electrokinesis. The chrome on the photomask is only on the order of 50- to 100-nm thick, so care should be exercised in handling the masks to prevent scratching. Place substrate in a plastic container to prevent dripping of hydrofluoric acid. If etch rates are uneven from one edge of the substrate to another, the substrate can be rotated by 90° at periodic intervals. The hydrolysis of the surface of the glass slide is exothermic. Depending on the volume of the hydrolysis solution and the number of substrates immersed, a considerable temperature rise in the solution can be seen. The hydrogen peroxide in the solution has a limited lifetime of only a few hours, so the solution needs to be regularly replaced. The annealing temperature varies slightly with different types of glasses, so the final temperature may need to be adjusted slightly. A weight may also be placed on the chip to help assure good bonding. An alternative procedure developed by Harrison’s group for pyrolyzing poly(dimethylsiloxane) on the channel wall surface works equally well (35). The curing should be performed in an explosion proof oven. The cells are viable in the extracellular buffer at room temperature for several hours; however, the dyes do begin to bleed out of the cells approx 1 h after loading. Experiments, therefore, should be carried out within 1 h of the loading process. The high voltages used to lyse the cells and then to separate the lysate are potentially dangerous. For this reason it is prudent to include an interlock system in the experimental setup to prevent operators from accidentally shocking themselves during an experiment. Cells enter the main channel of the microchip at random time intervals; so the average distance between cells in the main channel and the average time between lysis events also varies randomly. For the analytes released from one cell to be completely separated from the cell prior to it and after it, a certain average distance between cells must be maintained. For detection distances of 3–5 mm in the system described here, the optimal cell concentration was 1 × 106 cells/mL.

Acknowledgments This work was supported by the National Institutes of Health by Grant RO1GM067905. References 1. Lodish, H., Baltimore, D., Berk, A., Zipursky, S. L., Matsudaira, P., and Darnell, J. (1995) Molecular Cell Biology, 3rd Edition, Scientific American Books, New York, NY.

Single Cell Analysis on Microfluidic Devices

215

2. Zabzbyr, J. L. and Lillard, S. J. (2001) New Approaches to Single-cell Analysis by Capillary Electrophoresis. Trends Analyt. Chem. 20, 467–476. 3. Sims, C. E., Meredith, G. D., Krasieva, T. B., Berns, M. W., Tromberg, B. J., and Allbritton, N. L. (1998) Laser-micropipet combination for single-cell analysis. Anal. Chem. 70, 4570–4577. 4. Berridge, M. J. (1993) Inositol trisphosphate and calcium signalling. Nature 361, 312–325. 5. Chen, G. and Ewing, A. G. (1997) Chemical analysis of single cells and exocytosis. Crit. Rev. Neurobiol. 11, 59–90. 6. Hsieh, S., Dreisewerd, K., van der Schors, R. C., et al. (1998) Separation and identification of peptides in single neurons by microcolumn liquid chromatographymatrix-assisted laser desorption/ionization time-of-flight mass spectrometry and postsource decay analysis. Anal. Chem. 70, 1847–1852. 7. Hsieh, S. and Jorgenson, J. W. (1997) Determination of enzyme activity in single bovine adrenal medullary cells by separation of isotopically labeled catecholamines. Anal. Chem. 69, 3907–3914. 8. Pihel, K., Hsieh, S., Jorgenson, J. W., and Wightman, R. M. (1995) Electrochemical detection of histamine and 5-hydroxytryptamine at isolated mast cells. Anal. Chem. 67, 4514–4521. 9. Swanek, F. D., Ferris, S. S., and Ewing, A. G. (1997) Capillary Electrophoresis for the Analysis of Single Cells: Electrochemical, Mass Spectrometric, and Radiochemical Detection. In: Handbook of Capillary Electrophoresis, (Khaledi, M. G., ed.), CRC Press, Inc., Boca Raton, FL, pp. 495–521. 10. Ewing, A. G., Chen, T. -K., and Chen, G. (1995) Voltammetric and Amperometric Probes for Single-Cell Analysis. In: Voltammetric Methods in Brain Systems, (Boulton, A., Baker, G., and Adams, R. N., eds.), Humana Press, Totowa, NJ, pp. 269–304. 11. Lillard, S. J. and Yeung, E. S. (1997) Capillary Electrophoresis for the Analysis of Single Cells: Laser-Induced Fluorescence Detection. In: Handbook of Capillary Electrophoresis (Khaledi, M. G., ed.), CRC Press, Inc., Boca Raton, FL, pp. 523–544. 12. Han, F., Wang, Y., Sims, C. E., et al. (2003) Fast electrical lysis of cells for capillary electrophoresis. Anal. Chem. 75, 3688–3696. 13. Meredith, G. D., Sims, C. E., Soughayer, J. S., and Allbritton, N. L. (2000) Measurement of kinase activation in single mammalian cells. Nature Biotech. 18, 309–312. 14. Lee, C. L., Linton, J., Soughayer, J. S., Sims, C. E., and Allbritton, N. L. (1999) Localized measurement of kinase activation in oocytes of Xenopus laevis. Nat. Biotechnol. 17, 759–762. 15. McClain, M. A., Culbertson, C. T., Jacobson, S. C., Allbritton, N. L., Sims, C. E., and Ramsey, J. M. (2003) Microfluidic devices for the high-throughput chemical analysis of cells. Anal. Chem. 75, 5646–5655. 16. Fu, A. Y., Chou, H. -P., Spence, C., Arnold, F. H., and Quake, S. R. (2002) An integrated microfabricated cell sorter. Anal. Chem. 74, 2451–2457. 17. Fu, A. Y., Spence, C., Scherer, A., Arnold, F. H., and Quake, S. R. (1999) A microfabricated fluorescence-activated cell sorter. Nature Biotech. 17, 1109–1111.

216

Culbertson

18. McClain, M. A., Culbertson, C. T., Jacobson, S. C., and Ramsey, J. M. (2001) Flow cytometry of Escherichia coli on microfluidic devices. Anal. Chem. 73, 5334–5338. 19. Roper, M. G, Shackman, J. G., Dahlgren, G. M., and Kennedy, R. T. (2003) Microfluidic chip for continuous monitoring of hormone secretion from live cells using an electrophoresis-based immunoassay. Anal. Chem. 75, 4711–4717. 20. Wheeler, A. R., Throndset, W. R., Whelan, R. J., et al. (2003) Microfluidic device for single-cell analysis. Anal. Chem. 75, 3581–3586. 21. Fuhr, G. R. and Reichle, C. (2000) Living cells in opto-electrical cages. Trends Analyt. Chem. 19, 402–409. 22. Yang, J., Huang, Y., Wang, X. -B., Becker, F. F., and Gascoyne, P. R. C. (1999) Cell separation on microfabricated electrodes using dielectrophoretic/gravitational field-flow fractionation. Anal. Chem. 71, 911–918. 23. Yang, M., Li, C. -W., and Yang, J. (2002) Cell docking and on-chip monitoring of cellular reactions with a controlled concentration gradient on a microfluidic device. Anal. Chem. 74, 3991–4001. 24. Schilling, E. A., Kamholz, A. E., and Yager, P. (2002) Cell lysis and protein extraction in a microfluidic device with detection by a fluorogenic enzyme assay. Anal. Chem. 74, 1798–1804. 25. Muller, T., Gradl, G., Howitz, S., Shirley, S., Schnelle, T., and Fuhr, G. (1999) A 3-D microelectrode system for handling and caging single cells and particles. Biosens. Bioelectron. 14, 247–256. 26. Li, P. C. H. and Harrison, D. J. (1997) Transport, manipulation, and reaction of biological cells on-chip using electrokinetic effects. Anal. Chem. 69, 1564–1568. 27. White, F. M. (1991) Viscous Fluid Flow, Second Edition, McGraw-Hill, New York, NY. 28. Fortina, P., Cheng, J., Kricka, L. J., et al. (2001) DOP-PCR amplification of whole genomic DNA and microchip-based capillary electrophoresis. In: Capillary Electrophoresis of Nucleic Acids, Volume 2, (Mitchelson, K. R. and Cheng, J., eds.), Humana Press, Totowa, NJ, pp. 211–219. 29. Jacobson, S. C., Hergenröder, R., Koutny, L. B., Warmack, R. J., and Ramsey, J. M. (1994) Effects of injection schemes and column geometry on the performance of microchip electrophoresis devices. Anal. Chem. 66, 1107–1113. 30. Stjernstrom, M. and Roeraade, J. (1998) Method for fabrication of icrofluidic systems in glass. J. Micromech. Microeng. 8, 33–38. 31. Mets, U. and Rigler, R. (1994) Submillisecond detection of single rhodamine molecules in water. J. Fluoresc. 4, 259–264. 32. Nie, S., Chiu, D. T., and Zare, R. N. (1994) Probing individual molecules with confocal fluorescence microscopy. Science 266, 1018–1021. 33. Schrum, D. P., Culbertson, C. T., Jacobson, S. C., and Ramsey, J. M. (1999) Microchip flow cytometry using electrokinetic focusing. Anal. Chem. 71, 4173–4177. 34. McAteer, J. A. and Davis, J. (1994) Basic cell culture technique and the maintenance of cell lines. In: Basic Cell Culture (Davis, J., ed.), Oxford University Press, New York, NY, pp. 93–148. 35. Badal, M. Y., Wong, M., Chiem, N., Salimi-Moosavi, H., Harrison, D. J. (2000) J. Chromatogr. A 947, 277–286.

15 Rapid DNA Amplification in Glass Microdevices Christopher J. Easley, Lindsay A. Legendre, James P. Landers, and Jerome P. Ferrance Summary The polymerase chain reaction (PCR) for amplification of DNA has become a very useful tool in scientific research and analytical laboratories, yet conventional techniques are time-consuming, and the reagents are expensive. Miniaturization of this technique has the potential to drastically reduce amplification time and reagent consumption while simultaneously improving the efficiency of the reaction. Increasing the surface area-to-volume ratio using microfluidic reaction chambers allows homogeneous solution temperatures to be achieved much more rapidly than in conventional heating blocks. Employing infrared radiation to selectively heat the reaction solution can additionally reduce the time and energy needed for thermocycling; the reaction container is not heated and can even serve as a heat sink for enhancement of cooling. Microchip systems also provide the potential for fabrication of structures for additional processing steps directly in line with the PCR chamber. Not only can amplification be integrated with product separation and analysis, but sample preparation steps can also be incorporated prior to amplification. The ultimate goal is a miniature total-analysis-system with seamlessly coupled sample-in/answer-out capabilities that consumes very low volumes of reagents and drastically reduces the time for analysis. This chapter will focus on the materials and methods involved in simple straight-channel microchip PCR on glass substrates using non-contact thermocycling. Key Words: PCR; microchip; infrared; non-contact.

1. Introduction Since its first reported use in 1985 (1), and particularly since the advent of thermally stable polymerase enzymes, the polymerase chain reaction (PCR) has rapidly become a basic and essential tool in biochemical research and analytical laboratories. PCR has developed into an invaluable technique in clinical diagnostics for detecting infectious agents, and is being applied for detection of genetic changes associated with specific diseases. The reaction has also been widely developed in the forensic community for identification of suspects or victims in criminal investigations and can be used for identification of victims From: Methods in Molecular Biology, vol. 339: Microchip Capillary Electrophoresis: Methods and Protocols Edited by: C. S. Henry © Humana Press Inc., Totowa, NJ

217

218

Easley et al.

of terrorist attacks. More recent motivation for the application of PCR has been found in the rapid detection of agents of bioterrorism for accurate and focused response to attacks. PCR is an in vitro technique allowing amplification of a specific DNA fragment of known length. Short oligonucleotides, called primers, which are complementary to the 3′ sequences at the ends of the specific DNA fragment of interest, are mixed with nucleotides, a small amount of template DNA, and Taq DNA polymerase enzyme in the appropriate buffer. The temperature of the reaction mixture is then cycled to denature the double stranded DNA, allow annealing of primers at the ends of the DNA sequence to be amplified (typically 50–1000 bp in length), then extend the new DNA strands. This cycle is repeated to ideally double the number of fragments present in the mixture with each thermocycle. The final amount of amplified DNA fragment produced is dependent on the number of cycles, the number of starting copies of DNA template, and the efficiency of the reaction. Each step of the PCR is temperature-dependent, thus, accurate control of the solution temperature is very important. In the denaturing step, the temperature is raised to approx 95°C where the double-stranded template DNA denatures into single strands. If a high enough temperature is not reached, the DNA will not denature, but too high of a temperature in this step can inactivate the polymerase enzyme; even though it is more heat stable than normal DNA polymerases, Taq polymerase will still degrade in less than 1 h if held at 94°C. Typically, the first denaturing step is maintained for a longer time than subsequent cycles to ensure complete melting of any supercoiled DNA. The temperature is then lowered to within the range of 48 to 74°C depending on the sequence and length of the primers selected; this allows the single-stranded primers to anneal to the single strands of DNA at the appropriate locations. The annealing temperature of a particular target must be optimized empirically using the melting temperature of the primers as a starting point. The optimal annealing temperature is a compromise between the high specificity achieved at higher temperatures and the high percentage of annealing achieved at lower temperatures. If the optimal annealing temperature is not achieved, the specificity of amplification is compromised, where either undesired fragments will be amplified (annealing temperature too low) or not enough products will be formed (annealing temperature too high). After primer annealing, the oligonucleotides are extended through addition of deoxynucleotides by the DNA polymerase using the target DNA sequence as the template. This step is normally performed at the optimal enzymatic temperature (~72°C), but the enzyme is active over a range of temperatures. The dwell time at this step can be short because of the speed of the polymerase enzyme, but it can be extended if long amplicons are being amplified. To achieve sufficient amplification, the whole three-step

Rapid DNA Amplification in Glass Microdevices

219

Fig.1. A single thermal cycle for a conventional polymerase chain reaction (PCR) carried out in a heating block. The block temperature (dotted trace) is compared with the actual solution temperature (solid trace). Note that the solution temperature lags behind the block because of heat transfer through the polypropylene tubes. (Reprinted in part with permission from ref. 13. © 2000 American Chemical Society.)

cycle is normally repeated for 20–45 cycles. Often the extension time of the final cycle is held longer to ensure full extension of any incomplete fragments. 1.1. Conventional PCR PCR is typically performed in a commercially available thermocycler that consists of a heating block with up to 96 wells. The PCR mixture is placed in thin wall polypropylene tubes that fit securely into the wells in the block; more recent instruments are also designed for use of microtiter plates in place of the tubes. In the traditional cyclers, the denaturing and annealing steps use dwell times in the range of 5 to 180 s, whereas the extension dwell times are normally tens of seconds to minutes. These times are not limited by the biochemistry of the reaction, but rather by the physics of the process. The required temperature of the mixture for each step is attained by changing the temperature of the heating block in the instrument. Limited by the volume of the solution and heat transfer through the tube wall, the temperature of the solution takes longer to reach the same temperature as the block. Figure 1 shows a typical temperature cycle experienced by the block of a conventional PCR thermocycler where a lag in the reaction solution temperature is observed. The speed of the reaction is, therefore, limited both by the time it takes to change the temperature of the block and the dwell time required for the solution to achieve the proper temperature.

220

Easley et al.

Decreased thermocycling times have been reported in traditional thermocyclers (2). Though actual solution temperatures might not have been achieved in this system, the authors showed successful amplification for the particular DNA targets chosen. Denaturation of DNA was shown to be completed at temperatures as low as 88°C (2), thus long dwell times at 94°C were not necessary. In the same way, driven by the large excess of primer, DNA annealing was reduced to less than 1 s; as a further benefit, short annealing times have been shown to reduce the amount of mispriming, which leads to nonspecific product (3). The extension step is limited by the enzymatic rate of the polymerase, but at 35–100 nucleotides/s (3), only a few seconds of reaction time are needed near the extension temperature. In addition to time concerns, the cost of the PCR is also an issue, particularly when high-throughput processing is desired. Reaction volumes in polypropylene tubes of traditional thermocyclers are normally in the 25 to 50 µL range; this uses significant amounts of the primers and the Taq polymerase enzyme to achieve the necessary concentrations. Smaller volume (~5 µL) PCR is now possible in traditional cyclers, but significant cost savings could be achieved by reducing the volumes even further. 1.2. Microchip PCR Wilding et al. (4) first demonstrated translation of PCR to a chamber in a microchip device, using silicon/glass hybrid devices that held 5–10 µL of reaction mixture. Though the silicon provided rapid heat transfer, each cycle of PCR was approx 3 min long. The additional thermal mass of the copper block heater and Peltier stage was largely responsible for these slower cycle times. One other issue that arose in the development of microchip PCR was the necessity to coat the chamber walls to prevent adsorption of either the enzyme or the DNA to the silicon or silica surface. This initial work was followed by a subsequent study that investigated the effect of surface passivation on microchip amplification (5). It was found that silanization of the microchamber surface using a covalent coating method provided the best results. A number of other microchip PCR devices and designs have been investigated since this initial work, most notable of which is the flow-through method reported by Kopp et al. in Science (6). All of the currently available microchip PCR methods are detailed in a recent review by Krika and Wilding (7). These authors also report on the types of surface coatings, both dynamic and static, that have been utilized in microchip PCR. One of the benefits of the microchip amplification procedure is the ability to integrate additional processing or analysis steps directly into the same device. Waters et al. (8) performed cell lysis, multiplex PCR amplification, and electrophoretic sizing on a single microchip device. The entire microchip was placed

Rapid DNA Amplification in Glass Microdevices

221

into a commercial thermocycler with individual steps of 94°C for 2 min, 37°C for 3 min, and 72°C for 4 min (24 total cycles). The benefits of rapid microscale PCR were unrealized, for the temperature transition times were limited by the thermal mass of the heating block. Burns et al. (9) have approached this problem from an engineering perspective, creating a sophisticated device for amplification and analysis. A resistive heating region with temperature sensors, a sample loading region, and a gel-based separation region were fabricated into a single microdevice; a 106-bp fragment of DNA was successfully amplified and separated. Lagally et al. (10) have developed a device with microfabricated resistance heaters, resistance temperature sensors, and PCR chambers seamlessly connected to electrophoretic separation channels. Valves and hydrophobic vents are used to prevent evaporation of the PCR solution and for fluidic control. Successful sex determination using a multiplex PCR reaction from human genomic DNA was demonstrated in less than 15 min. These two devices are complex and costly to fabricate, however, making them less amenable for large-scale manufacturing or disposability. Development of possible disposable microdevices that contain integrated functionality has been reported by Koh et al. (11) who designed poly (cyclic olefin) devices that incorporated valves and a separation channel, but manufacturing of these devices was still a complex endeavor. 1.3. Infrared Heating One method for reducing the cost of PCR microdevices is to eliminate the need for on-chip heating components. Remote heating of small volume PCR solutions in rectangular glass capillaries using an inexpensive tungsten lamp was first shown by Oda et al. (12). The method relied on direct heating of the solution by excitation of the vibrational bands of the water by adsorption of infrared (IR) radiation from the lamp. This drastically reduced the heat transfer problem associated with heating because the reaction solution was selectively heated. Cooling was performed by flowing a stream of nitrogen past the microcontainers to enhance the rate of heat removal from the container, and thus from the solution. This methodology reduced the total time for one cycle to 17 s, with individual dwell times of 2 s at 94°C, 2 s at 54°C, and 4 s at 72°C. Further efforts showed that decreasing the volume of the reaction to 160 nL increased the efficiency of the PCR (13) and also allowed faster cycling times (Fig. 2). Using a two-temperature protocol recommended by the manufacturer for specific amplification of a 500 bp fragment from a λ-phage DNA template, as few as 10 cycles produced an adequate mass of DNA in less than 12 min for analysis using capillary electrophoresis (CE) with laser-induced fluorescence (LIF) detection. Giordano, et al. (14) applied the IR-mediated heating technique to microchips, performing PCR amplifications in chambers in polyimide microdevices. Because

222

Easley et al.

Fig. 2. In only 10 thermal cycles, PCR amplification of a 500-bp fragment of λ-DNA was achieved in less than 12 min with only 593 starting copies. This amplification, when compared with conventional thermocycling, illustrates the high efficiency possible with the use of nanoliter scale volumes and exclusive heating of the solution. (Reprinted in part with permission from ref. 13. © 2000 American Chemical Society.)

polyimide is transparent in the 600- to 3000-nm range, the IR radiation again selectively heated the solution, and the low thermal mass of the chips provided rapid cooling even without forced convection. Heating and cooling rates as high as 10°C/s and short dwell times resulted in cycles that required only 12 s to complete. Adequate amounts of PCR product were observed using LIF detection after 15 cycles, a process taking only 240 s. Surface passivation was also required with these devices, with polyethylene glycol included in the PCR mixture to dynamically coat the chip surface. Polyimide does not make a good substrate for integration of additional processing steps owing to incompatibilities with existing separation and detection techniques however, so the IR method was transferred to glass microdevices. With the much greater thermal mass of the glass chips and the ability of glass to dissipate heat, the heating and cooling rates were reduced, but amplification could still be performed in approx 20 min. Figure 3 shows a thermocycling profile for a glass microdevice using IR-mediated heating along with a microchip electrophoretic analysis of PCR product for amplification of the 275bp invasive A (invA) gene from Salmonella typhimurium (primers designed inhouse). In this example, separation and detection were performed on separate microdevices, but PCR product could also be directly analyzed in an integrated

Rapid DNA Amplification in Glass Microdevices

223

Fig. 3. PCR amplification of the target 275-bp invasive A (invA) gene of Salmonella typhimurium DNA using primers designed in-house. (A) Three-temperature thermocycling was carried out as follows: 35 cycles at 95°C for 3 s, 64°C for 8 s, and 75°C for 3 s, with a final extension at 75°C for 60 s (total time ~21 min). (B) Separation was carried out by microchip gel electrophoresis with sizing markers of 15 and 1500 bp to confirm the presence of the 275-bp amplicon.

electrophoretic separation channel on the same device (15). Surface passivation in these devices relied on an absorbed epoxy poly(dimethylacrylamide) (EPDMA) coating that was validated for glass devices (16) with bovine serum albumin (BSA) added to the PCR mixture for additional passivation. This chapter details the IR-mediated amplification method utilized in these publications, but focuses on simple glass devices containing only the PCR chamber and a thermocouple reference chamber.

224

Easley et al.

2. Materials 2.1. Fabrication of Microdevices 1. Borofloat glass plates 1.1-mm thick, coated with chrome and photoresist (Nanofilms, Westlake, CA). 2. Borofloat glass for cover plates 1.1-mm thick (S.I. Howard Glass, Worcester, MA) (see Note 1). 3. Flood ultraviolet (UV) light source (OAI Associates, Milipitas, CA). 4. Mask with image of device (Fig. 4) (see Note 2). 5. Photoresist developer AZ 400K (Clariant Corp, Sunnyvale, CA). 6. Chromium etchant CR-7S (Cyantek Corp, Fremont, CA). 7. Photoresist stripper AZ 300T (Clariant Corp). 8. Hydrofluoric acid (HF). 9. Nitric acid (HNO3). 10. Diamond-tipped drill bits (Crystalite Corp, Lewis Center, OH). 11. Ceramic plates. 12. Colloidal graphite (Renite S-24, Columbus, OH).

2.2. Microchip Reservoirs 1. Sylgard 184 silicone elastomer base and curing agent (Dow Corning Corp, Midland, MI), both stored below 32°C. 2. Vacuum desiccation chamber.

2.3. Coating Polymer Preparation 1. 2. 3. 4. 5.

N,N-dimethylacrylamide (Aldrich). Allylglycidyl ether (Aldrich). TEMED (Aldrich). Ammonium persulfate (APS) solution (40% w/v) in distilled water. Dialysis tubing.

2.4. Preparation of the Microchip 1. 1 M NaOH. 2. Autoclaved Nanopure water (Barnstead International, Dubuque, IA). 3. 2 mg/mL BSA (Aldrich).

2.5. Temperature Detection 1. Miniature type-T copper-constantan thermocouple (T240C, Physitemp Instruments, Clifton, NJ). 2. Thermocouple amplifier (TAC-386-TC; Omega Engineering, Inc., Stamford, CT) with output of 1 mV/°C.

2.6. IR-Mediated Thermocycler 1. 5-V/12-V DC power supply (Power-One, Camarillo, CA). 2. 8-V, 50-W Tungsten filament lamp (General Electric, Cleveland, OH). 3. Solenoid valve with a room-temperature compressed air source (such as a nitrogen tank) and appropriate tubing.

Rapid DNA Amplification in Glass Microdevices

225

Fig. 4. Positive mask design for a PCR microchip including (A) a reference chamber for the thermocouple and (B) a PCR chamber for DNA amplification. Void regions were also etched in proximity to the chambers to reduce thermal losses to the glass. The mask negative should be used for ultraviolet exposure. 4. Circuitry for active control of lamp and solenoid (see Note 3). 5. LabVIEW application to control the switching circuitry (see Note 4). 6. Gold-mirrored surface to enhance heating (Edmund Industrial Optics, Barrington, NJ).

2.7. Preparation of the PCR Mixture 1. 2. 3. 4.

25 mM MgCl2, stored at –20°C. 10X PCR buffer: 100 mM Tris, 500 mM KCl, stored at –20°C (see Note 5). 100 mM dNTPs: dATP, dGTP, dCTP, dTTP, store at –20°C. 20 µM of forward and reverse primers complimentary to the end regions of the DNA fragment being amplified. 5. 5000 U/mL Taq polymerase. 6. 2 mg/mL BSA. 7. Template DNA containing the fragment to be amplified. The specific DNA fragment will dictate the amount of starting material that will be required.

2.8. Loading the Microchip and Thermocycling 1. Molecular biology grade mineral oil.

3. Methods 3.1. Fabrication of Microdevice 1. The borofloat glass with chrome and photoresist are exposed to the UV source through the mask negative for 5 s (see Note 6). 2. The exposed photoresist is removed using developer then the resist hard-baked at 110°C for 30 min. 3. The exposed chrome is removed using chromium etchant.

226

Easley et al.

Fig. 5. Image of a completed microdevice for PCR amplification. The channels were filled with blue dye for visualization. 4. The glass is etched using a solution of HF:HNO3 :H2O (50:14:36), etching at a rate of 3 µm/min. 5. The glass is etched to 150-µm deep. 6. The remaining photoresist is removed using stripper. 7. The chrome is removed using chromium etchant. 8. Top plates are fabricated by drilling reservoir holes that will line up with the ends of the PCR chamber using a diamond-tipped drill bit (1.1-mm diameter). 9. Bottom and top plates are cut to size then cleaned with an ammonia-based window cleaner (Windex® window cleaner). 10. The glass plates are pressed together, placed between graphite-coated ceramic plates, and placed in a high-temperature furnace for bonding. 11. The furnace temperature is ramped to 550°C at 8°C/min, then at 3°C/min to 670°C where it is held for 3.5 h before naturally cooling to room temperature to avoid cracking. Figure 5 shows a completed device (see Note 7).

3.2. Microchip Reservoirs 1. Combine silicone elastomer base and curing agent in a 15:1 ratio, mixing well (see Note 8).

Rapid DNA Amplification in Glass Microdevices

227

2. Degas the mixure in a vacuum desiccator for 15 min or until no air bubbles remain. 3. Pour degassed mixture into a Petri dish to a depth of approx 5 mm. 4. Bake at 60°C for 2 h, making sure that the dish is level. 5. With a razor blade, cut a piece of the cured poly(dimethylsiloxane) (PDMS) to match the shape of the microchip, and produce the appropriate reservoirs with a hole punch (see Note 9). 6. Thoroughly clean the PDMS and glass microchip surfaces with soap and water, and dry them well with a stream of nitrogen to ensure a tight seal to the glass. 7. Align the PDMS reservoir with the etched reservoirs of the microchip, bring them into contact, and apply light pressure until no air bubbles exist between the two layers. Although unnecessary, heating the chip at this point (~60°C for ~10 min) has been found to enhance the seal. This reservoir should be capable of containing any aqueous solution, and the seal is readily reversible. 8. If the seal is ineffective, simply peel away and repeat steps 6 and 7.

3.3. Preparation of Coating Polymer (17) 1. A 10 mL solution of 0.4 M N,N-dimethylacrylamide, 0.008 M allylglycidyl ether in water is degassed for 10 min. 2. The reaction is initiated using 10 µL of TEMED and 10 µL of APS solution added below the surface, and allowed to react for 24 h at room temperature. 3. The reaction mixture is dialyzed for 4 h against three changes of distilled water to remove any unreacted monomer. 4. The mixture is lyophylized to obtain a solid, then resuspended in water at a concentration of 0.2% (w/v).

3.4. Preparation of the Microchip 1. 2. 3. 4. 5.

Rinse the chamber with 1.0 M NaOH for 10 min to prepare for passivation. Flush for 10–20 min with 0.2% EPDMA to passivate the chamber (see Note 10). Rinse with distilled, autoclaved water for 20 min. Rinse with BSA for 5 min to enhance the EPDMA passivation. Rinse with distilled, autoclaved water for 5 min.

3.5. Temperature Detection 1. If needed, carefully sand the end of the miniature thermocouple with fine sandpaper until it will fit into the reference chamber. 2. Gently thread the thermocouple into the central region of the reference chamber where the lamp will be focused (see Note 11). 3. Secure the thermocouple in place by taping to the chip.

3.6. IR-Mediated Thermocycler 1. Build the appropriate electrical circuit box to allow computer control of the heating lamp and solenoid air valve (see Note 12).

228

Easley et al.

2. Build a chip holder that positions the microchip reaction chamber in the focus of the lamp and in the stream of the compressed air source (see Note 13).

3.7. Preparation of the PCR Mixture (25 µL total volume) (see Note 14) 1. Combine the following reagents to make the “PCR mixture” (see Note 15): 16.6 µL distilled, autoclaved water. 3 µL 25 mM MgCl2 (see Note 5). 2.5 µL 10X PCR buffer. 0.05 µL 100 mM dATP. 0.05 µL 100 mM dTTP. 0.05 µL 100 mM dGTP. 0.05 µL 100 mM dCTP. 0.2 µL 20 µM forward primer. 0.2 µL 20 µM reverse primer. 0.3 µL 2 mg/mL BSA. 2 µL DNA at the appropriate concentration for the application (water is used as a negative control) (see Note 16). 2. To each 25-µL aliquot of the previously mentioned mixture, add 0.6 µL of Taq polymerase. a. b. c. d. e. f. g. h. i. j. k.

3.8. Loading the Microchip and Thermocycling 1. Making sure the passivated PCR chamber is empty, use a pipet to add about 20 µL of the PCR mixture to one of the reservoirs and allow capillary action to pull it through the chamber and into the opposing reservoir (see Note 17). 2. Fill the reference chamber in the same manner with PCR buffer. 3. Cover the solution in all reservoirs with just enough mineral oil to prevent evaporation. 4. Place the microchip into the holder above the focal point of the IR heating source in a position to promote equal heating of the entire PCR chamber and the reference chamber 5. Place the mirrored surface above the heated region to increase the heating rate. 6. The denaturing/annealing/extension temperatures appropriate for the gene to be amplified are entered in the program (see Notes 18 and 19). Figure 3 shows an example of thermocycling along with the conditions used.

4. Notes 1. Borofloat glass for etching and cover plates can also be purchased in 0.7-mm thickness, which can be used to fabricate these devices. The thinner glass allows an increased speed of thermocycling, but results in more fragile microchips. 2. Photographic negative masks can be prepared from AutoCAD files of the microdevice design by commercial printers at approx 6 dpi resolution using lithographic printing. Metal masks can also be designed for this purpose. 3. Only beginner-level knowledge of electronics is necessary for this purpose. The lamp and solenoid should be connected to solid-state relays, which can feed the

Rapid DNA Amplification in Glass Microdevices

4.

5.

6. 7.

8.

9.

10.

11.

12. 13.

14.

15.

229

appropriate voltages from the power supply. The relays should be connected to a PC outfitted with an analog-to-digital (A/D) converter board to allow computer control. A proportional integral differential algorithm is best suited for thermal control of PCR. This type of algorithm will help the system avoid overshooting the temperatures needed for PCR. The LabVIEW application should use the input from the thermocouple to actively control the solenoid and solid-state relays, and maintain the temperature of the system at the input values. Note that the 10X PCR buffer can be purchased both with and without MgCl2. If the correct concentration of MgCl2 is included with the 10X PCR buffer, then adding MgCl2 is not necessary. Figure 4 shows a positive image of an example mask pattern for microchip PCR. The negative image should be used for UV exposure. There are multiple chip geometries possible, and many are likely suitable for use with the protocol. This particular straight-channel geometry was chosen simply because the authors have been successful with its use. A 15:1 ratio of elastomer:curing agent has been found to enhance the seal between the PDMS reservoir and the glass microchip when compared with the 10:1 ratio recommended by the manufacturer. To ensure accurate reservoir fabrication, use a printout of the microchip design as a guide. Also, be sure to create a larger reservoir at the reference chamber to ease the installation of the thermocouple. It is important not to rinse with water before the EPDMA passivation. The 1 M NaOH rinse before the coating is used to prepare the surface for polymer adsorption, and rinsing with water will adversely alter the surface. This technique takes time to master, and use of a microscope is helpful. Tweezers can be used to thread the thermocouple into the chamber, but they may cause abrasions to the coating and foul the sensor. Pipet tips were found to be useful for this purpose, with sensor fouling much less likely. Also, miniature thermocouples are very delicate. Caution must be taken when transporting the microchip to different stations if the thermocouple is installed. The fast response time of solid-state relays make them ideal for this application, but any sort of switching circuitry may be used. The chip holder should be constructed from an insulating material to avoid fast heat dissipation. If the holder is too good of a heat sink, the solution in the chamber may not be able to reach the high temperatures or may take too long to reach them. Less than 1 µL is needed to fill the PCR chamber, but the reservoirs contain more to help prevent evaporation. The volume of the solution can be reduced by decreasing the size of the reservoirs to save expensive reagents while still achieving amplification. Allow all PCR reagents except the Taq polymerase to thaw to room temperature. Because Taq is thermally sensitive, it should be kept at –20°C and should be the last reagent added to the reaction. The thermocycling should be initiated within 5 min of enzyme addition.

230

Easley et al.

16. Because the surface area-to-volume ratio is drastically increased in microdevices, some DNA and enzyme may still adsorb to the glass surface, even with the passivation. Two to five nanograms of genomic DNA is a typical amount included in a single reaction mixture (25 µL). 17. Be sure to allow room in the reservoirs for the mineral oil when filling with the PCR mixture. Overflowing the reservoirs with mineral oil can provide a leakage path while thermocycling. If the mineral oil leaks away, the reaction mixture will evaporate. 18. The annealing temperature is dependent on the primer sequence; primers chosen for PCR are dependent on the DNA fragment being amplified. 19. Because the overall reaction environment is different in glass microchips than in traditional thermocyclers, some adjustments to the typical temperature protocols may be required for optimal amplification.

References 1. Saiki, R. K., Scharf, S., Faloona, F., et al. (1985) Enzymatic amplification of betaglobin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230, 1350–1354. 2. Mai, M., Grabs, R., Barnes, R. D., Vafiadis, P., and Polychronakos, C. (1998) Shortened PCR cycles in a conventional thermal cycler. Biotechniques 25, 208–210. 3. Wittwer, C. T. and Garling, D. J. (1991) Rapid cycle DNA amplification: time and temperature optimization. Biotechniques 10, 76–83. 4. Wilding, P., Shoffner, M. A., and Kricka, L. J. (1994) PCR in a silicon microstructure. Clin. Chem. 40, 1815–1818. 5. Wilding, P., Shoffner, M. A., Cheng, J., Huichia, G., and Kricka, L. J. (1995) Thermal cycling and surface passivation of micromachined devices for PCR. Clin. Chem. 41, 1367, 1368. 6. Kopp, M. U., Mello, A. J., and Manz, A. (1998) Chemical amplification: continuousflow PCR on a chip. Science 280, 1046–1048. 7. Kricka, L. J. and Wilding, P. (2003) Microchip PCR. Anal. Bioanal. Chem. 377, 820–825. 8. Waters, L. C., Jacobson, S. C., Kroutchinina, N., Khandurina, J., Foote, R. S., and Ramsey, J. M. (1998) Microchip device for cell lysis, multiplex PCR amplification, and electrophoretic sizing. Anal. Chem. 70, 158–162. 9. Burns, M. A., Mastrangelo, C. H., Sammarco, T. S., et al. (1996) Microfabricated structures for integrated DNA analysis. Proc. Natl. Acad. Sci. USA 93, 5556–5561. 10. Lagally, E. T., Emrich, C. A., and Mathies, R. A. (2001) Fully integrated PCRcapillary electrophoresis microsystem for DNA analysis. Lab on a Chip 1, 102–107. 11. Koh, C. G., Tan, W., Zhao, M. Q., Ricco, A. J., and Fan, Z. H. (2003) Integrating polymerase chain reaction, valving, and electrophoresis in a plastic device for bacterial detection. Anal. Chem. 75, 4591–4598.

Rapid DNA Amplification in Glass Microdevices

231

12. Oda, R. P., Strausbauch, M. A., Huhmer, A. F., et al. (1998) Infrared-mediated thermocycling for ultrafast polymerase chain reaction amplification of DNA. Anal. Chem. 70, 4361–4368. 13. Huhmer, A. F. and Landers, J. P. (2000) Noncontact infrared-mediated thermocycling for effective polymerase chain reaction amplification of DNA in nanoliter volumes. Anal. Chem. 72, 5507–5512. 14. Giordano, B. C., Ferrance, J., Swedberg, S., Huhmer, A. F., and Landers, J. P. (2001) Polymerase chain reaction in polymeric microchips: DNA amplification in less than 240 seconds. Anal. Biochem. 291, 124–132. 15. Ferrance, J. P., Wu, Q., Giordano, B., et al. (2003) Developments toward a complete micro-total analysis system for Duchenne muscular dystrophy diagnosis. Analytica Chimica Acta 500, 223–236. 16. Giordano, B. C., Copeland, E. R., and Landers, J. P. (2001) Towards dynamic coating of glass microchip chambers for amplifying DNA via the polymerase chain reaction. Electrophoresis 22, 334–340. 17. Chiari, M., Cretich, M., Damin, F., Ceriotti, L., and Consonni, R. (2000) New adsorbed coatings for capillary electrophoresis. Electrophoresis 21, 909–916.

Index

233

Index A Amperometric detection, capillary electrophoresis, 92, 93 externally positioned off-chip electrode alignment, advantages, 98, 99 applications, 103, 105 micromanipulated electrode, 99, 101, 107 screen-printed approach, 99 injection, 88, 89, 91, 92, 105, 106 integrated on-chip electrode alignment, applications, 101–103 decoupled detection, 97, 98 end-channel detection, 95, 107 in-channel detection, 95–97 working electrode potential and alignment, 94, 107 materials, 87, 88 popularity, 86 principles, 86, 87 voltage separation, 93, 94

analytes and applications, 120–124 buffer systems, 118, 120 electrode manufacturing, 114–118, 124 electronics, 118 materials, 114 Current monitoring, see Electroosmotic flow D Destaining, microchips for protein separation, 146–148 Direct-write, see Laser ablation DNA amplification, see Polymerase chain reaction DNA separation, historical perspective, 129, 130 lab-on-a-chip device separation, materials, 131, 133–135, 140–142 planar microfluidic chips, 135–138 principles, 130, 131 sipper microfluidic chips, 138–142

B, C Biotinylation, surface modification, 54, 55 Bonding, see Glass bonding Capillary electrochromatography (CEC), principles, 2–4 Capillary zone electrophoresis, peptide analysis, 165, 166, 179, 180 CEC, see Capillary electrochromatography Cell analysis, see Single cell analysis Conductivity detection, advantages and limitations, 113, 114

E Electrochemical detection, see also Amperometric detection; Labon-a-chip, modes, 86 peptide detection, 176–178, 180 Electrochromatography, peptide analysis, 169–171 Electroosmotic flow (EOF), measurement, current monitoring method, materials, 192 overview, 190 technique, 194–199

233

234 flow imaging, 190, 191 neutral marker method, materials, 191, 192 overview, 189, 190 technique, 193, 194, 196–198 polyelectrolyte multilayers, 61 principles, 2–4, 187–189 reproducibility, 188 Electrospray coupling, see Mass spectrometry Enzyme assay, peptide analysis, 172–174 EOF, see Electroosmotic flow F, G Faraday’s law, 87 Fast prototyping, see Hot embossing Glass bonding, DNA separation microchips, 23–25, 34 electrospray-coupled microchips, 72, 73, 78 H Hot embossing, materials, 38, 45 microchannel fabrication, 43, 44 poly(dimethylsiloxane) microchip, 33, 34 polymer embossing, 40–42 principles, 37, 38 silicon template micromatching, 39, 40 I, J Immunoassay, peptide analysis, 171, 172 Infrared heating, microchip polymerase chain reaction, 221–223, 227, 229 Instantaneous current, equation, 87 Isoelectric focusing, peptide analysis, 167–169 Isotachophoresis, peptide analysis, 167 Jurkat cell, see Single cell analysis L Lab-on-a-chip (LOC), device design, 14

Index DNA separation, materials, 131, 133–135, 140–142 planar microfluidic chips, 135–138 principles, 130, 131 sipper microfluidic chips, 138–142 microfabrication, bottom substrate processing, 17–20, 34 electrode integration, 25, 26 glass-to-glass bonding, 23–25, 34 photolithography, 16, 17 photomask development, 14, 16, 34 top substrate processing, 20–23 protein separation, advantages, 157, 158 denaturation of samples, 150, 151 destaining, 146–148 materials, 148–150, 154 planar microfluidic chips, 151, 152, 154, 155, 157 sipper microfluidic chips, 152–155, 157 Laser ablation, materials, 38, 45 poly(dimethylsiloxane) microchip, 34 principles, 38 technique, 42–44 Laser-induced fluorescence (LIF), overview of detection, 85, 86 single cell analysis, data collection and analysis, 213 setup, 209 LIF, see Laser-induced fluorescence LOC, see Lab-on-a-chip M Mass spectrometry (MS), advantages in capillary electrophoresis detection, 68 electrospray coupling of microchip capillary electrophoresis, chip fabrication, access hole drilling, 73, 74, 78 fittings, 74

Index glass bonding, 72, 73, 78 photolithography, 71, 72, 78 wet chemical etching, 72, 78 coating of microchips, 74, 79 interfacing and operation, 75–79 materials, 70, 71 overview, 68–70 spray tip preparation, 75 peptide detection, 178, 179 Mass, calculation for analyte, 3 MEKC, see Micellar electrokinetic chromatography Micellar electrokinetic chromatography (MEKC), peptide analysis, 166, 167 principles, 2, 3 Microchip capillary electrophoresis, applications, DNA separation, 7 overview, 2 protein analysis, 7 small molecule analysis, 8 detection, 6, 7 historical perspective, 1, 2 injection, 5 microchip construction, 4, 5 separation, 6 theory, 2–4 Microdialysis, peptide samples, 163 MS, see Mass spectrometry N–P Net charge, calculation for analyte, 3 Neutral marker, see Electroosmotic flow PCR, see Polymerase chain reaction PDMS microchip, see Poly(dimethylsiloxane) microchip Peptide analysis, applications, 159, 160 considerations, 160 microchip electrophoresis, capillary zone electrophoresis, 165, 166, 179, 180 detection,

235 electrochemical detection, 176–178, 180 fluorescence, 175, 176 mass spectrometry, 178, 179 electrochromatography, 169–171 enzyme assay, 172–174 immunoassays, 171, 172 injection, 164 isoelectric focusing, 167–169 isotachophoresis, 167 micellar electrokinetic chromatography, 166, 167 multi-dimensional separations, 174 principles, 160–162 sample preparation, digestion of proteins, 164, 179 microdialysis, 163 preconcentration and desalting, 162, 163 substrate selection, 161, 162, 179 Photolithography, electrospray-coupling chip fabrication, 71, 72, 78 lab-on-a-chip, 16, 17 poly(dimethylsiloxane) microchip, alignment and exposure, 30 development, 31, 32, 34, 35 hard bake, 32, 35 mask preparation, 28, 29 photoresist deposition, 29, 34 postexposure baking, 31 soft bake, 30 water surface preparation, 29, 34 Planar microfluidic chip, DNA separation, 135–138 protein separation, 151, 152, 154, 155, 157 Poly(dimethylsiloxane) (PDMS) microchip, advantages, 27, 57 coating of glass microchips, 51, 54, 55 fabrication, assembly, irreversible sealing, 33, 59 reversible sealing, 33, 59

236 checklist, 33 direct-write, 34 fast prototyping, 33, 34 materials, 28 overview, 27, 28 photolithography, alignment and exposure, 30 development, 31, 32, 34, 35 hard bake, 32, 35 mask preparation, 28, 29 photoresist deposition, 29, 34 postexposure baking, 31 soft bake, 30 water surface preparation, 29, 34 substrate characteristics and preparation, 32 polyelectrolyte multilayer, see Polyelectrolyte multilayer Polyelectrolyte multilayer (PEM), electroosmotic flow measurements, 61 materials, 58, 59 noncovalent coating, 59, 60 polyelectrolyte types, 63 rationale, 57, 58 stability and reproducibility of coating, 61, 62 surface modification, 50, 53, 55 Polymerase chain reaction (PCR), conventional thermocycling, 219, 220 historical perspective, 217, 218 microchip polymerase chain reaction, coating polymer preparation, 227 infrared heating, 221–223, 227, 229 loading and thermocycling, 228, 230 materials, 224, 225, 228, 229 microchip fabrication, 225, 226, 227, 229 microchip reservoirs, 226, 227, 229

Index overview, 220, 221 reaction mixture preparation, 228, 229 temperature detection, 227, 229 principles, 218, 219 Poly(methylmethacrylate) devices, amine termination, 51, 54, 55 Protein separation, gel electrophoresis overview, 145, 146 lab-on-a-chip device separation, advantages, 157, 158 denaturation of samples, 150, 151 destaining, 146–148 materials, 148–150, 154 planar microfluidic chips, 151, 152, 154, 155, 157 sipper microfluidic chips, 152–155, 157 R, S Reservoir, microchip reservoirs for polymerase chain reaction, 226, 227, 229 Sealing, poly(dimethylsiloxane) microchip assembly, irreversible sealing, 33, 59 reversible sealing, 33, 59 thermal bonding, 43, 44, 45 Single cell analysis, microfluidic devices, advantages, 204, 205 channel wall coating and durability, 210, 211, 213, 214 imaging of chip and cell, 212, 213 injection efficiency measurement, 213 Jurkat cell culture, 211 laser-induced fluorescence detection, data collection and analysis, 213 setup, 209 loading of cells, 211

Index materials, 205, 206, 213, 214 microchip, bonding, 208, 209, 214 design and operation, 206, 207 glass etching, 208, 214 transport and lysis of cells, 211, 212, 214 overview of techniques, 203, 204 Sipper chip, DNA separation, 138–142 protein separation, 152–155, 157 Surface modification, amine termination of poly(methylmethacrylate) devices, 51, 54, 55 biotinylation, 54, 55 materials, 51–53

237 poly(dimethylsiloxane) coating of glass microchips, 51, 54, 55 polyelectrolyte multilayer, 50, 53, 55 rationale, functionalization of surfaces, 51 practical effects, 49, 50 surface property improvement, 50, 51 ultraviolet graft polymerization, 53 T–V Temperature detection, microchip polymerase chain reaction, 227, 229 Ultraviolet graft polymerization, surface modification, 53 Velocity, formula, 3

Microchip Capillary Electrophoresis. Methods and Protocols

Capillary Electrophoresis. Methods and Protocols

Capillary Electrophoresis: Methods and Protocols