Citation
Exploration of cze experimental parameters as a function of capillary length for the use of microfluidic device

Material Information

Title:
Exploration of cze experimental parameters as a function of capillary length for the use of microfluidic device
Creator:
Fiala, Rachel Erin
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
xv, 91 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Separation (Technology) ( lcsh )
Capillary electrophoresis ( lcsh )
Microfluidic devices ( lcsh )
Capillary electrophoresis ( fast )
Microfluidic devices ( fast )
Separation (Technology) ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (M.S.)--University of Colorado Denver, 2010. Department of Chemistry
Bibliography:
Includes bibliographical references (leaves 88-91).
Thesis:
Department of Chemistry
Statement of Responsibility:
by Rachel Erin Fiala.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
657066355 ( OCLC )
ocn657066355

Downloads

This item has the following downloads:


Full Text

PAGE 1

EXPLORATION OF CZE EXPERIMENTAL PARAMETERS AS A FUNCTION OF CAPILLARY LENGTH FOR THE USE OF A MICROFLUIDIC DEVICE by Rachel Erin Fiala B.S. Chemistry, University of Colorado Denver, 2008 A thesis submitted to the University of Colorado Denver in partial fulfillment of the requirements for the degree of Master of Science Chemistry 2010

PAGE 2

This thesis for the Master of Science degree by Rachel Erin Fiala has been approved by Larry Anderson Scott Reed 20. 2010 Date

PAGE 3

Fiala, Rachel, E. (Master of Science, Chemistry) Exploration of CZE Experimental Parameters as a Function of Capillary Length for the use of a Microfluidic Device Thesis directed by Dr. Mark Anderson ABSTRACT Reductions in both separation efficiency and resolution are associated with the minimization of capillary length. By understanding how the experimental parameters could be adjusted for more optimal separation conditions, it was possible to limit such reductions. The separation potential was decreased, while the current was adjusted so that there were minimal baseline fluctuations, as the capillary length was shortened. As expected, capillary modification of various length capillaries did not drastically influence the electroosmotic flow, separation efficiency, and resolution. Separation of a complex mixture containing (-)-epinephrine, 3-hydroxytyramine hydrochloride, tyramine, and tyrosine was accomplished using unmodified, SO um i.d. capillaries having lengths of 40and 12-cm. A simple mixture containing tyramine and ascorbic acid was subsequently separated within a microfluidic device using similar applied electric field strengths. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. Signed / Mark Anderson

PAGE 4

ACKNOWLEDGMENT I would like to thank Professor Mark Anderson first and foremost for helping me to complete research and compose this document. His help has been truly invaluable. The National Science Foundation deserves acknowledgement for both the financial aid and academic support they provided through the GK12 Grant (NSF DGE-0742434). My family has also been crucial in providing support and advice throughout my college career, and especially while I've been a graduate student. The faculty members in the chemistry department at the University of Colorado Denver also deserve special thanks, as I have grown as a person through their wisdom and guidance. Finally, I would like to thank Professor Larry Anderson and Professor Scott Reed for reviewing this document and considering my application for a Master of Science Degree in analytical chemistry.

PAGE 5

TABLE OF CONTENTS LIST OF FIGURES ................................................................................................................................. VI LIST OF TABLES .................................................................................................................................... x LIST OF EQUATIONS ........................................................................................................................ XIII CHAPTER 1: INTRODUCTION ......................................................................................................... 1 Method ................................................................................................................................................. 2 Detection ............................................................................................................................................. 5 Origin of Electroosmotic Flow (EOF) ................................................................................... 10 Electroosmotic Flow and Separation ................................................................................... 14 Flow Profile ..................................................................................................................................... 15 Efficiency, Capacity and Resolution ..................................................................................... 17 Solute-Zone Dispersion ............................................................................................................. 20 Lab-on-a-chip Concept ............................................................................................................... 26 Capillary Modification ................................................................................................................ 28 Objectives ........................................................................................................................................ 30 CHAPTER 2: EXPERIMENTAL ...................................................................................................... 31 CZE Setup ......................................................................................................................................... 31 Wavelength of Maximum Absorbance and Acquisition Channels .......................... 32 Capillary Preparation-Part I (unmodified capillaries) ................................................ 33 Capillary Preparation-Part II (PDDA modified capillaries) ...................................... 34 Capillary Preparation-Part II (PSS modified capillaries) ........................................... 35 Reagents and Sample Preparation ........................................................................................ 36 Setting Voltage and Current .................................................................................................... 37 Detection-Setting the Baseline .............................................................................................. 38 Data Acquisition Parameters .................................................................................................. 38 Introduction Method .................................................................................................................. 38 Microfluidic Device ...................................................................................................................... 39 Experimental Tests and Separations ................................................................................... 40 CHAPTER 3: RESULTS AND DISCUSSION ............................................................................... 43 Method Development (Part 1) ................................................................................................. 43 Method Development (Part II) ............................................................................................... 45 Elution Time and Capillary Length ....................................................................................... 4 7 Setting Parameters ...................................................................................................................... 49 Separations Continued ............................................................................................................... 52 Capillary Modification ................................................................................................................ 54 Changing Length and PSS Concentration of Modified Capillaries ........................... 57 Efficiency and Resolution ......................................................................................................... 64 Complex Mixture (Part I) .......................................................................................................... 70 Complex Mixture (Part II) ........................................................................................................ 76 Microfluidic Device ...................................................................................................................... 79 CHAPTER 4: CONCLUSIONS ......................................................................................................... 84 Future Applications ..................................................................................................................... 85 BIBLIOGRAPHY .................................................................................................................................. 88 v

PAGE 6

LIST OF FIGURES Figure 1. Schematic diagram of CZE setup ...................................................................................... 4 2. Electropherogram depicting the elution of four compounds (in order of elution: tyramine, D-tyrosine, 3-hydroxytyramine hydrochloride, and epinephrine) from an unmodified 12 em length, 50 um i.d. capillary (separation conditions: 0.5 kV, 6 uA, ( +) polarity mode). ......................................... 5 3. Depiction of the eletrophoretic motion of charged and neutral ions, with separation of charged ions being a function of their attraction to the anode/cathode, and motion of neutral ions being a function of the electroosmotic flow of buffer solution ............................................................................ 10 4. Diagram depicting the attraction of buffer cations to the negative oxygen atoms of the capillary wall to create the fixed layer, and the buffer cations that become mobile as the zeta potential of the wall decreases to zero .......... 12 5. Depiction ofthe charge density at the capillary wall being lowered as a function of the pH of buffer solution (e.g. pH < 7), and the zeta-potential that decreases to zero within the mobile layer .................................................................... 14 6. Illustration of the laminar flow profile in which there is an observed velocity gradient, and turbulent flow profile in which chaotic fluid motion generates a flat-flow profile ................................................................................................ 17 7. Depiction of the solute-zone dispersion from a discrete concentration source, where each curve represents varying products of Dt, where Dis the diffusion coefficient and tis the time. The products of Dt are as follows: (A) Dt = -0.00, (B) Dt = 0.01, (C) Dt = 1.00, (D) Dt = oo ..................................................... 23 8. Illustration of the components of the microfluidic device used in this work .......................................................................................................................................................... 28 vi

PAGE 7

9. Illustration of the lightsource and detector that are situated at a 180 angle about the capillary window, using at-piece component. ....................................... 32 10. Diagram showing how the wavelength of maximum absorbance is determined .................................................................................................................................. 33 11. Depiction of the electrostatic interaction between the SiOgroups at the capillary wall and the cationic polymer, poly(diallyldimethylammonium chloride) ....................................................................................................................................... 35 12. Depiction of the electrostatic interaction between the cationic polymer, poly(diallyldimethylammonium chloride) and the anionic polymer, poly( sodium 4-styrenesulfonate) ..................................................................................... 36 13. Depiction of the elution time ofhydroquinone as a function of capillary length, for unmodified capillaries having lengths of 50-, 25-, and 12-em ....... 48 14. Elution of ascorbic acid and hydroquinone, in negative polarity mode, from a SO um i.d., 50 em length capillary modified with 0.01 M PDDA ............. 56 15. Diagram depicting the elution ofhydroquinone from a 50 um i.d., 60 em length capillary modified with 0.01 M PSS .................................................................... 61 16. Diagram depicting the hydroquinone elution from an unmodified 50 um i.d., 50 em length capillary. The separation efficiency is calculated from the peak width at the baseline and the resolution is calculated from the full width at half the maximum peak height. ..................................................................................... 65 17. Diagram depicting the hydroquinone elution from an unmodified 50 um i.d., 25 em length capillary. The separation efficiency is calculated from the peak width at the baseline and the resolution is calculated from the full width at half the maximum peak height. ..................................................................................... 66 18. Diagram depicting the hydroquinone elution from an unmodified 50 um i.d., 12 em length capillary. The separation efficiency is calculated from the vii

PAGE 8

peak width at the baseline and the resolution is calculated from the width at half the maximum peak height. .......................................................................................... 66 19. Diagram depicting the elution (from left to right) of: tyramine, tyrosine, 3-hydroxytyramine HCl (dopamine), and (-)-epinephrine, when 27S nm is monitored for absorbances ................................................................................................. 72 20. Diagram depicting the elution (from left to right) of: tyramine, tyrosine, 3-hydroxytyramine HCl (dopamine), and (-)-epinephrine, when 278 nm is monitored for absorbances ................................................................................................. 72 21. Diagram depicting the elution (from left to right) of: tyramine, tyrosine, 3-hydroxytyramine HCl (dopamine), and (-)-epinephrine, when 280 nm is monitored for absorbances ................................................................................................. 73 22. Screen snapshot of the (-)-epinephrine absorbance peak in the scope mode of the OOIBASE32 program .................................................................................... 73 23. Screen snapshot of the 3-hydroxytyramine HCl absorbance peak in the scope mode of the OOIBASE32 program ....................................................................... 74 24. Screen snapshot of the tyramine absorbance peak in the scope mode of the OOIBASE32 program ...................................................................................................... 74 2S. Screen snapshot of the D-tyrosine absorbance peak in the scope mode of the OOIBASE32 program ...................................................................................................... 7S 26. Electropherogram of trial 1 depicting the elution of four compounds (in order of elution: tyramine, D-tyrosine, 3-hydroxytyramine hydrochloride, and epinephrine) from an unmodified 12 em length, SO um i.d. capillary ...... 78 27. Electropherogram of trial 2 depicting the elution of four compounds (in order of elution: tyramine, D-tyrosine, 3-hydroxytyramine hydrochloride, and epinephrine) from an unmodified 12 em length, SO um i.d. capillary ...... 78 viii

PAGE 9

28. Trial 1: elution of tyramine and ascorbic acid from the microfluidic device ............................................................................................................................................ 81 29. Trial 2: elution of tyramine and ascorbic acid from the microfluidic device ............................................................................................................................................ 82 30. Trial 3: elution of tyramine and ascorbic acid from the microfluidic device ............................................................................................................................................ 82 ix

PAGE 10

LIST OF TABLES Table 1. The length from capillary introduction to the detection cell, for various length capillaries ...................................................................................................................... 33 2. Experimental parameters associated with the separations conducted with the microfluidic device .......................................................................................................... 40 3. Elution of hydroquinone and ascorbic acid from a 75 um i.d., 41.5 em length capillary, conditioned with 1.054 M NaOH and 0.05 M sodium di-jmono-basic buffer solution ............................................................................................................... 45 4. Elution ofhydroquinone and ascorbic acid from a 50 um i.d., 50 em length capillary, conditioned with 1.054 M NaOH and 0.05 M sodium di-/mono-basic buffer solution ............................................................................................................... 46 5. Elution of hydroquinone and ascorbic acid from 50 um i.d. capillaries having lengths of 50-, 25, and 12-em. Capillaries were conditioned with 1.054 M NaOH and 0.05 M sodium di-/mono-basic buffer solution prior to sample introduction ................................................................................................................................ 48 6. Elution of hydroquinone from an unmodified 50 um i.d., 50 em length capillary when the voltage and current were adjusted to various settings .... 50 7. Average voltage and current readings associated with unmodified capillaries of various lengths, in positive polarity mode ........................................ 51 8. Documented fluctuations in voltage and current readings associated with a 50 um i.d., 50 em length capillary modified with 0.005 M PSS ............................. 51 9. Separation of ascorbic acid and hydroquinone via unmodified SO um i.d. capillaries having lengths of 60-,50-,40-,25-, and 12-em ........................................ 53 X

PAGE 11

10. Elution ofhydroquinone and ascorbic acid from a 50 um i.d., 50 em length capillary modified with 0.01 M PDDA. ............................................................................ 55 11. Elution ofhydroquinone and ascorbic acid from a simple mixture that was introduced to a 50 um i.d., 50 em length capillary modified with 0.005 M PDDA .............................................................................................................................................. 57 12. Comparison of elution times associated with an unmodified 50 um i.d., 25 em length capillary versus a 50 um i.d., 25 em length capillary modified with 0.01 M PSS ................................................................................................................................... 59 13. Comparison of hydroquinone elution times associated with an unmodified 50 um i.d., 50 em length capillary versus a 50 um i.d., 50 em length capillary modified with 0.01 M PSS .................................................................... 59 14. Comparison ofhydroquinone elution from 50 um i.d., 60 em length capillaries modified with 0.005-, 0.01-, 0.015, or 0.02-M PSS .............................. 61 15. Comparison ofhydroquinone elution from 50 um i.d., 25 em length capillaries modified with 0.005-, 0.01-, 0.015, or 0.02-M PSS .............................. 63 16. Calculated efficiency and resolution associated with unmodified 50 um i.d. capillaries having lengths of 50-, 25-, and 12-em. Note that the voltage setting for the 25 em length capillary is in excess of the value mentioned in the Setting Parameters section ........................................................................................... 6 7 17. Calculated efficiency and resolution associated with unmodified 50 um i.d. capillaries having lengths of 50-, 25-, and 12-em. Note that the voltage setting for the 25 em length capillary is 5 kV, which is in agreement with the optimal setting mentioned in the Setting Parameters section .............................. 67 18. Average hydroquinone elution time, separation efficiency, and resolution associated with 50 um i.d., 60 em length capillaries modified with 0.005-, 0.01-, 0.015-, or 0.02-M PSS ................................................................................................. 68 xi

PAGE 12

19. Average hydroquinone elution time, separation efficiency, and resolution associated with 50 um i.d., 25 em length capillaries modified with 0.005-, 0.01-, 0.015-, or 0.02-M PSS ................................................................................................. 69 20. Average hydroquinone elution time, efficiency and resolution associated with 50 um i.d. capillaries having lengths of 60-, 50-, 40-, 25-, and 12-em. For each capillary length, an average was taken of the voltage, current, hydroquinone elution time, efficiency and resolution associated with the four capillaries modified with 0.005-, 0.01-, 0.015-, and 0.02-M PSS ......................... 70 21. The pKa value(s) and wavelength of maximum absorbance associated with the components of the complex and simple mixtures ................................... 71 22. Elution of the four compounds contained in the complex mixture from an unmodified 50 um i.d., 40 em length capillary ............................................................. 76 23. Elution of the four compounds contained in the complex mixture from an unmodified 50 um i.d., 12 em length capillary ............................................................. 77 24. Experimental parameters and average hydroquinone elution time from the microfluidic device .......................................................................................................... 80 25. Experimental parameters and elution times associated with the tyramine/ascorbic acid mixture from the microfluidic device ............................ 81 xii

PAGE 13

LIST OF EQUATIONS Equation 1. Beer's law where E is the wavelength dependent molar absorptivity coefficient, b is the path length of the detection cell, cis the analyte concentration (M), and A is the absorbance ................................................................................... 6 2.J.lp is the electrophoretic mobility (in m2 /V js ), E is the electric field strength (in V /m), and Up is the electrophoretic migration velocity (in m/s) ................ 7 3. z is the ionic charge, 'YJ is the viscosity of the medium, r is the Stokes radius for the analyte, and J.lp is the electrophoretic mobility .............................................................. 7 4. ks is the Boltzmann constant, Tis temperature, D is the diffusion coefficient, and r is the Stokes radius ................................................................................................ 7 5. k is a constant, q is the charge of the ion, M is the mass of the ion, and J.lp is the electrophoretic mobility .................................................................................................................. B 6. E is the relative permittivity of the buffer solution, E is the applied electric potential, Cis the zeta potential of the capillary wall, 'YJ is the viscosity of the medium, and llF.OF is the electroosmotic mobility ....................................................................... 9 7.J.lp is the electrophoretic mobility of the charged analyte, flo is the electroosmotic mobility, E is the electric field strength, up is the electrophoretic migration velocity, and Uo is the electroosmotic migration velocity ......................................................................................................................................................... lO 8. v(r) is the linear velocity of the bulk solution at radial position r, Pis the uniform applied pressure gradient, E is the applied electric field, 'YJ is the viscosity coefficient, E is the dielectric constant, k is the multiplicative inverse (reciprocal) of the double-layer thickness, tp'" is the potential at the capillary wall and lc is the zero-order modified Bessel function of the first kind ........................ 13 xiii

PAGE 14

9. p. is the analyte's apparent mobility, Dis the diffusion coefficient, E is the electric field strength, and Ns is the number of theoretical plates ..................................... 19 10. lR is the analyte retention time, w is the peak width at the baseline, and N is the number of theoretical plates .................................................................................................. 19 11. J.lp is the electrophoretic mobility, N is the separation efficiency, J.lo is the electroosmotic mobility, and Rs is the resolution ..................................................................... 20 12. lR is the analyte retention time, FWHM is the full width at half the maximum peak height, and R is the resolution .......................................................................... 20 dC 13. Dis the diffusion coefficient, dx is the derivative of concentration with respect to distance, and] is the mass flux of solute per unit area .................................... 21 d:C 14. D is the diffusion coefficient, dxis the second derivative of dC concentration with respect to distance, and dl is the derivative of concentration with respect to time ................................................................................................. 2 2 15. Dis the diffusion coefficient, tis time, xis the distance, and Cis the concentration gradient ......................................................................................................................... 23 16. a is the variance, xis the distance, and Cis the concentration gradient.. ............... 23 17. pis the density ofthe fluid, u is the linear velocity, dis capillary diameter for the open-tubular capillary, and TJ is the absolute viscosity of the fluid ................... 25 18. Re is the Reynolds number, TJ is the viscosity of the fluid, pis the density, and r is the radius of the tube ............................................................................................................ 26 19. LT is the length of the capillary tube in meters, Lois the distance from introduction end ofthe capillary to the detection window (in meters), Vis the xiv

PAGE 15

potential difference in volts across the tube, to is the migration time of an electrically neutral analyte (in seconds), and f.lo is the electroosmotic mobility (in m2/V js) .............................................................................................................................. 44 XV

PAGE 16

CHAPTER 1: INTRODUCTION The use of capillary electrophoresis dates back to 1937, when Arne Tiselius developed moving boundary electrophoresis.1 The setup for this separation technique includes a U-shaped cell that is filled with buffer solution, and electrodes that are placed in each opening to transmit power from the power supply. When sample is introduced, the positively charged analytes migrate through the cell to the cathode, while the negatively charged analytes migrate toward the anode.1 Stellan Hjerten expanded on this work in 1967 by utilizing a high voltage power supply to separate analytes along a rotating 3-mm i.d. capillary.2 In 1981, Jorgenson and Luckacs conducted capillary electrophoresis experiments using 75-um i.d. capillaries.2 The application of capillary electrophoresis was then extended to the separation of DNA samples, between 1988 and 1992.2 Since capillary electrophoresis is utilized to perform separations with high efficiency and resolution, work is being done in this field to minimize the CE setup size without reducing the efficiency and resolution of the separation. The first microchip CE device was developed in the early 1990s, and since this time a great deal of research has been conducted in this field.3 Like the diameter of typical CE capillaries, the diameter of the channel in 1

PAGE 17

microfluidic devices is typically 10's of micrometers. Because the behavior of fluids at the microscale level often differs from that at the macroscopic level, the ultimate goal of this research is to understand the fluid dynamics to compensate for, or control the factors that dominate at the microscopic level. Surface tension, energy dissipation and fluidic resistance are just a few of the factors that become more important at the microscale level. In fact, the Reynolds number (which relates momentum with the velocity of the fluid) is typically very low for microfluidic devices, and mixing within the fluid is therefore limited to diffusion. By controlling the temperature, pH and concentration, it is possible to minimize certain effects or at least provide more uniform separation conditions.4 Method Capillary Zone Electrophoresis Many separation techniques utilize methods similar to capillary electrophoresis to separate mixtures along the length of a capillary having a very small inner diameter (i.e. in the micrometer range).5 These techniques include capillary zone electrophoresis, capillary gel electrophoresis, capillary isoelectric focusing, isotachophoresis, and capillary electrochromatography, to name a few. Of these techniques, capillary zone electrophoresis is considered the most basic. Capillary zone electrophoresis takes advantage of 2

PAGE 18

analyte mass-to-charge ratio to separate analytes in the presence of a strong electric field. Some of the basic requirements for the CZE separation include a homogeneous buffer solution and constant electric field strength.s The overall CZE setup is shown schematically in figure 1.6 The high voltage from the power supply is applied to the anode end of the capillary, and is isolated from the operator using a plexiglass interlock box. The introduction and exit ends of the capillary are placed in vials that contain aqueous buffer solution having a pH of typically -7.0, a pH value experimentally found to keep the capillary walls deprotonated. A large potential is applied to initiate motion of buffer by electroosmotic flow through the capillary, and sample introduction is achieved by placing the introduction end of the capillary in the vial that contains the sample and momentarily reapplying voltage. Following this process, the capillary end is repositioned in the source vial (that contains only the background buffer) where it will remain until the experiment has reached completion.6 3

PAGE 19

.ktode Source Vll S8mple Vll ....---------, High Volt.ge Po-Suppy Figure 1. Schematic diagram of CZE setup +-+----CMhode Because of the electroosmotic flow, cationic, anionic and neutral analytes migrate toward the destination vial from the bulk flow of the buffer solution.6 The separation of analytes within this medium is therefore related to the differences in the electrophoretic mobility of these ions. The figure below depicts an electropherogram that is generated as analytes within the capillary pass the detector. Each peak in the electropherogram corresponds to different chemical species having a specific retention time.6 Note that the complex mixture contains four compounds: (-)-epinephrine, 3-hydroxytyramine hydrochloride, tyramine, and D-tyrosine. The order of elution (from left to right) based on pKa values is as follows: tyramine, 0-tyrosine, 3hydroxytyramine hydrochloride, and epinephrine. 4

PAGE 20

Absorbance vs TimeComplex Mix (12 em) 0.06 0.05 \.,j 0.04 c !0 .. & A IV\ ..11.1. .... --...1' ......,.. ,Q 0.03 J. Q Ill ,Q 0.02 < --.,... .L-c.oua / 0.01 ..... i 0 0 100 200 300 400 500 Elapsed Time (s) Figure 2. Electropherogram depicting the elution of four compounds (in order of elution: tyramine, D-tyrosine, 3-hydroxytyramine hydrochloride, and epinephrine) from an unmodified 12 em length, 50 um i.d. capillary (separation conditions: 0.5 kV, 6 uA, (+) polarity mode). Detection Ultra-violet/visible absorption spectroscopy is often used for detection of biological species, provided that they have a suitable chromophore to absorb light. When combined with capillary electrophoresis, the hv detection is on-column, and therefore does not add to the analyte dispersion.7 On-column detection is accomplished by removing the polymer coating from a small section of the capillary to ultimately make it optically transparent. The primary advantage of using an on-column detection system is that there is no loss to resolution. This type of CE detection system, however, has a detection cell with a very small path length.7 5 I I

PAGE 21

The traditional UV cell has a path length of approximately 1-cm while the CE cell used in this work has a cross-section of approximately 50-um.7 Given the proportionality between absorbance and path length, as described by Beer's Law (equation 1), ultraviolet detection CE systems suffer in their sensitivity when compared to separations that use other types of detection. The sensitivity of the UV detection system can be increased by expanding a portion of the capillary to create a bubble cell, or by adding additional tubing to create a z-cell. These methods effectively increase sensitivity by increasing the path length, but not without a loss in resolution.7 (1) Equation 1. Beer's law where E is the wavelength dependent molar absorptivity coefficient, b is the path length of the detection cell, cis the analyte concentration (M), and A is the absorbance Separation Principles Electrophoretic migration velocity (uP) is a function of the electrophoretic mobility and strength of the electric field; which is calculated by dividing the applied potential by the total capillary length.8 Since the electrophoretic mobility (,..11) describes the forces acting on each individual analyte ion, the electrophoretic mobility is directly proportional to the ionic 6

PAGE 22

charge and inversely proportional to counteracting frictional forces that slow motion of the analyte ion. Frictional forces are related to the viscosity of the medium (i.e. buffer solution) and the size/shape ofthe ion, as described by Stokes law for spherical particles. The electrophoretic mobility is therefore described by equation 3.8 II_.-p._.E (2) Equation 2, J.lp is the electrophoretic mobility (in m2 fV fs), E is the electric field strength (in V fm), and Up is the electrophoretic migration velocity (in m/s) l H IJ,. ---at a given p 61rrJT (3) Equation 3, z is the ionic charge, TJ is the viscosity of the medium, r is the Stokes radius for the analyte, and J.lp is the electrophoretic mobility A:.T 6:rrJD (4) Equation 4, kB is the Boltzmann constant, Tis temperature, Dis the diffusion coefficient, and r is the Stokes radius Depending on the net charge of the ion ( l ), the viscosity of the medium ( rJ) and the Stokes radius ( T), ions will migrate faster /slower through the capillary due to the applied electric field.7 Equation 4 describes the Stokes radius for a given analyte, which depends on the Boltzmann 7

PAGE 23

constant, temperature, and diffusion coefficient. The migration rate is therefore directly proportional to the charge of the ion ( q) and inversely proportional to the mass (M) of the ion, as described by equation 5, wherek is a constant. This facilitates separation of complex mixtures, as no two molecules will likely have the same charge-to-mass ratio at a given pH.9 (5) Equation 5, k is a constant, q is the charge of the ion, M is the mass of the ion, and J.lp is the electrophoretic mobility Because the migration velocity is a function of the analyte's net charge, neutral species are difficult to separate from the mixture since they are not influenced by the applied electric field.8 The migration of an uncharged species (whereq-0) is therefore related to the rate of electroosmotic flow of the buffer solution. In positive polarity mode, the anode resides at the sample introduction and the cathode resides at the exit end of the capillary. The cations that make up the buffer solution migrate from the positively charged anode to the negatively charged cathode. This movement of the buffer cations drags along the solvent, establishing the bulk flow due to electroosmosis. Electroosmosis creates conditions where neutral species migrate toward the destination vial. The electroosmotic mobility, which depends on the relative permittivity of the buffer solution, the applied 8

PAGE 24

electrical field strength, the zeta potential at the capillary wall, and the viscosity of the medium, is given by equation 6.a (6) Equation 6, E is the relative permittivity of the buffer solution, !; is the zeta i-lEM potential of the capillary wall, TJ is the viscosity of the medium, and is the electroosmotic mobility In most circumstances the electroosmotic mobility can be related to the retention time of a neutral species.7 Figure 3 schematically illustrates that the small more-positively charged ions are attracted to the cathode and travel faster than large less-positively charged ions, neutral species, and negatively charged ions that are attracted instead to the anode, but are carried toward the cathode by the electroosmotic flow. This means that the retention time (tit) associated with EOF should lie somewhere between the retention time of a positively charged ion and that of the negatively charged analyte. Equation 7 can therefore be used to relate the velocity of a charged analyte to the velocity of a neutral analyte. Note that the electrophoretic mobility of any negatively charged analyte is less than the electroosmotic flow of buffer solution at pH 7, since this transport medium is positively charged at neutral pH.7 9

PAGE 25

8 Al-A Al D A2-Al-Al+ A2+ A. 'lode A cathode Electroosmotic Flow Figure 3. Depiction of the eletrophoretic motion of charged and neutral ions, with separation of charged ions being a function of their attraction to the anode/cathode, and motion of neutral ions being a function of the electroosmotic flow of buffer solution (7) Equation 7, 1-'P is the electrophoretic mobility of the charged analyte, /-'o is the electroosmotic mobility, E is the electric field strength, Up is the electrophoretic migration velocity, and Uo is the electroosmotic migration velocity Origin of Electroosmotic Flow (EOF) Helmholtz first observed the phenomenon of EOF in 1877 when he applied potential to the ends of a glass capillary filled with an aqueous solution containing ions.1o He found that a negative charge acquired along the interior wall, while the solvent and the surface of charged particles in close proximity to the wall acquired a positive charge. The term electroosmosis specifically refers to the bulk motion of liquid near the interior wall toward the cathode in positive polarity, or anode in negative polarity mode. In accordance with his observations, it was found that an electric 10

PAGE 26

charge is acquired on almost all solid surfaces in which there is a solid/ electrolyte interface. to The most common material used forCE capillaries is bare silica; therefore further explanation of electroosmosis will be given with this material in mind. Silica contains silanol groups (Si-OH) that reside along the interior wall of the capillary.11 When exposed to solutions with pH > 4, the silanol groups become deprotonated, leaving negatively charged silonate groups (Si-O-) at the interface. Basic electrolyte solutions can be used to ionize the interface. The attraction between the negative atom charge of the interface and cations in the electrolyte solution ultimately creates the double layer.11 Scheme 1. The double layer, which has a total thickness of approximately 100 A, (but is a function of the ionic strength) is composed of both a fixed and mobile portion.12 The fixed layer results from the strong electrostatic interactions between negatively charged silonate groups and cations (from the adjacent buffer solution) that are in close proximity to the interface. The mobile layer results from the electric field that attracts solvated buffer cations towards the capillary wall. The electric potential drops linearly as the distance from the wall increases within the fixed layer, and it decreases exponentially to zero further into the mobile layer. The perceived line that 11

PAGE 27

exists between the fixed and mobile layer is termed surface of the shear (defined as a stress applied parallel to a surface; results from the loss in fluid velocitydue to friction experienced along immobile positive charge of capillary wall) and the potential at the shear layer is called the zeta potentia/.12 The mobile layer is composed of a high concentration of cations and these cations are attracted to the cathode end of the capillary, when potential is applied. Figure 4 depicts their electrophoretic motion.13 Motion of these cations drags along the bulk solvent and creates the electroosmotic flow (EOF). The EOF rate is dependent on both the field strength and charge density at the capillary wall; the latter being a function of pH increases EOF to a maximum wherein all silanol groups are ionized.13 'J( 'J( 'J( \V ' cr o o o cr o cr cr q cr q cr GJ Electroosmotic Flow [J Anode Cathode o-ao o-aao-gqQ-aa Si Si Si Si .i Si Si Si Si Si Si /1\ /I\ /1\ /1\ )k II\ A\ /I\ /1\ /1\ /1\ /1\ Figure 4. Diagram depicting the attraction of buffer cations to the negative oxygen atoms of the capillary wall to create the fixed layer, and the buffer cations that become mobile as the zeta potential of the wall decreases to zero. 12

PAGE 28

Given the differences in electrical and frictional forces experienced by ions in the double layer, the local velocity is a function of distance from the capillary wall.12 The velocity is zero at the capillary wall, as the buffer ions are adsorbed onto the surface, and increases within the layer to a maximum speed at a very small distance away. Smoluchowski was the first to propose the aforementioned theory in which the bulk of solution migrates at maximum velocity, but in order to specifically address motion of liquid in narrow cylindrical capillaries, Rice and Whitehead incorporated the axial pressure and axial electrical potential gradients.12 Equation 8 is an extension of the electroosmotic flow equation previously discussed, and the expression: :1 lc.(.tr) f I h d "b h h d h ) IS use u m t at It escn est e net c arge ens1ty m t e 4Jr Ia{ka double layer.12 (8) Equation 8, v(r) is the linear velocity of the bulk solution at radial position r, Pis the uniform applied pressure gradient, E is the applied electric field, 11 is the viscosity coefficient, E is the dielectric constant, k is the multiplicative inverse 'Pr. (reciprocal) of the double-layer thickness, is the potential at the capillary wall I c. and is the zero-order modified Bessel function of the first kind 13

PAGE 29

I <:) 0 0 f @. 0 0 <:) 0 .!!! 0 0 c: .S! 0 0 a.. IE >IE Diffuse Layer '\ Stem Layer Figure 5. Depiction of the charge density at the capillary wall being lowered as a function of the pH of buffer solution (e.g. pH < 7), and the zeta-potential that decreases to zero within the mobile layer Electroosmotic Flow and Separation The advantages of EOF are most notably associated with the separation of anions and cations for simultaneous detection, and the flow profile that is much narrower for electroosmotic flow versus pressure-driven flow.13 Some of the disadvantages of electroosmotic flow were recognized through empirical studies and additional analysis. Since the contents of the capillary are drained during the separation process, the sample capacity of the CE system is limited by electroosmotic flow. The reproducibility of elution times is also a concern, as there must be consistency in the 14

PAGE 30

electroosmotic flow rate from run to run. Another disadvantage is related to the electrophoretic mobility of the analyte ions. When the eletrophoretic mobility of an ion approaches the electroosmotic mobility, the analyte may not be separated from the mixture. Peak resolution is also impacted by the electroosmotic flow rate. Because analyte ions may have very similar electrophoretic mobilities, the capillary may be drained of its contents before separation has been achieved. To achieve separation of closely eluting analytes, the separation potential is often lowered so that species migrate at a much slower rate through the capillary.13 Flow Profile One factor that enhances the overall efficiency of CE systems is the flat flow profile associated with EOF versus the rounded, laminar flow profile that is associated with chromatography systems having pressure-driven flow.14 Laminar flow is best described as a smooth, streamline flow wherein a series of parallel layers flow in the same direction without any disruption. The flow is laminar when the diameter of the pipe and flow velocity are both small. This type of flow profile is represented by a series of moving cylindrical layers that travel fastest toward the middle of the pipe, and slowest toward the wall. In fact the cylinder physically touching the wall is immobilized as a result of frictional forces. Laminar flow fundamentally differs from turbulent 15

PAGE 31

flow because the pushing force (or applied pressure) causes friction along the walls usually in slower flow close to the capillary walls, and the path of least resistance is through the middle of the pipe's opening.u Unlike pressure-driven flow that creates a velocity gradient across the capillary diameter, EOF is a function of the double layer and the strength of the applied electric field.7 Since the EOF originates from the forces that exist at the capillary walls, the flow velocity is highest at the wall and slower with increasing distance from the wall. This creates a turbulent flow profile. Turbulent flow is best characterized as having a flat flow profile. Because electroosmotic flow is created near the capillary wall with motion of buffer in the mobile layer, and the electric field (i.e. electrostatic interaction between buffer cations and negative charge of the capillary wall) drops linearly to zero in the mobile layer, the flow rate at some small fixed distance from the capillary wall is essentially equal to the flow rate at the middle of the capillary opening. The amount of band-broadening that results from laminar flow of pressure-driven systems is therefore greater than that amount resulting from turbulent flow of CE systems.7 16

PAGE 32

Laminar -------Turbulent Figure 6. Illustration of the laminar flow profile in which there is an observed velocity gradient, and turbulent flow profile in which chaotic fluid motion generates a flat-flow profile. Efficiency, Capacity and Resolution Separation efficiency is related to the ability of the separation to distinguish between closely eluting components of a mixture.15 This is described quantitatively by the number of theoretical plates of the separation. The term "theoretical plates" is historical and comes from descriptions of the equilibrations found in a distillation column (e.g. zones where equilibrium is established between two phases of a single substance). As the number of equilibration zones increases, the efficiency of the distillation also increases. This analogy is applied to other separation systems as well, such as capillary zone electrophoresis.15 17

PAGE 33

The separation efficiency in a typical high performance liquid chromatography (HPLC) experiment is directly related to the interaction that analyte molecules have between the stationary phase (i.e. material coating the inner wall in an unpacked column) and the mobile phase (i.e. solvent) of the separation.15 In the ideal situation, analytes will partition between both phases so that discrete solute zones will be distributed along the length of the column. Although there is no obvious physical analog to the equilibration zone in capillary electrophoresis, the term persists as a way of quantifying separation efficiency.ts Unlike HPLC experiments, in which there is a mass transfer between the two phases, separation efficiency of capillary electrophoresis is dependent on the apparent mobility (/-l) of the analytes within the separation medium, the diffusion coefficient of the analyte (DJ, and the electric field strength (E) of the separation, as given by equation 9.16 Since there is no mass transfer between phases in CZE, and CZE is subject to a flat flow profile, CE separation efficiency is generally much greater than HPLC separation efficiency.16 Note that for simplification, equation 10 was used to calculate the CE separation efficiency in this work.17 18

PAGE 34

N _IJE I 2D (9) Equation 9, Jl is the analyte's apparent mobility, Dis the diffusion coefficient, E is the electric field strength, and Ns is the number of theoretical plates. Traditionally, N is calculated from the shape of the analyte zone, eq. 10 (10) Equation 10, tR is the analyte retention time, w is the peak width at the baseline, and N is the number of theoretical plates. For equation 9 and 10, we can use the electropherogram to determine either, the analyte mobility or the diffusion coefficient. The number of theoretical plates can reach up to several hundred thousand for a typical CE separation, although this depends on the sample capacity.' The sample capacity is defined as the amount of sample that can be reliably injected without risk of overload. The overload limit refers to the injected sample mass that will decrease the separation efficiency to 90% of its normal value.18 Like separation efficiency and sample capacity, resolution is considered a figure of merit when evaluating data generated from CE separations. This factor is influenced by the magnitude of electrophoretic and electroosmotic mobilities. Although opposite in sign, the greater the electrophoretic and electroosmotic mobilities, the higher the resolution. The 19

PAGE 35

following equation (equation 11) describes the resolution (Rs) ofCE separations and suggests that lowering the velocity and therefore increasing the analysis time, ultimately increases the resolution.16 Note that for simplification, equation 12 was used to calculate the resolution in this work.19 R,.!_(fliJ, {N) 4 ll,. + IJ .. (11) Equation 11, J.l.p is the electrophoretic mobility, N is the separation efficiency, J.l.o is the electroosmotic mobility, and Rs is the resolution R-( rl! ) FWHM (12) Equation 12, tR is the analyte retention time, FWHM is the full width at half the maximum peak height, and R is the resolution Solute-Zone Dispersion Given that resolution is a direct product of separation efficiency, the concept of solute-zone dispersion will be discussed in more detail using rate theory versus plate theory.20 The plate theory was previously used to discuss the separation efficiency in both HPLC and CE applications. The CE separation efficiency was said to be dependent on the apparent mobility of 20

PAGE 36

the analyte within the separation medium, the diffusion coefficient of the analyte, and the electric field strength, as given by equation 9.20 The diffusion phenomenon arises from the random Brownian motion of molecules, which leads to the migration of molecules from an area of higher concentration to an area of lower concentration.20 This concentration gradient is described by Fick's first law (equation 13), where Dis the diffusion coefficient, dC. is the derivative of concentration with respect to tb: distance, and] is the mass flux of solute per unit area. The rate at which the molecules migrate from an area of higher concentration to that of lower concentration is given by Fick's second law (equation 14), where the change in concentration with respect to time is defined as being proportional to the rate at which the concentration gradient changes.20 dC J--D --b (13) dC Equation 13, Dis the diffusion coefficient, tb: is the derivative of concentration with respect to distance, and J is the mass flux of solute per unit area 21

PAGE 37

(14) Equation 14, Dis the diffusion coefficient, dx: is the second derivative of dC concentration with respect to distance, and dt is the derivative of concentration with respect to time The following solution to Fick's second law (equation 15) can be applied to systems that have a discrete concentration source.20 In capillary electrophoresis, the sample vial is the concentration source. The analytes that are introduced to the capillary migrate from the anode to the cathode end of the capillary, and separation is a function of the differences in electrophoretic mobility of these analyte ions. The concentration of each analyte is contained within a sphere, in which the core of the sphere has the highest concentration and the outermost portion has the lowest concentration. The concentration gradient for a given analyte is therefore normally distributed, as depicted in the following diagram, where the solutezone dispersion increases with the time/dispersion coefficient. This means the variance can be found by comparing equation 15 to the normalized Gaussian function described by equation 16, where the variance is 2Dt in equation 15.2 0 22

PAGE 38

cex -1 2(11Dr}"'1 4Dr (15) Equation 15, Dis the diffusion coefficient, tis time, xis the distance, and Cis the concentration gradient c Q ::: Concentration Gradient .. c Iii Cj c "CS Iii .!::! = z -6 -4 -2 0 Distance 2 4 6 Figure 7. Depiction of the solute-zone dispersion from a discrete concentration source, where each curve represents varying products of Dt, where Dis the diffusion coefficient and tis the time. The products of Dt are as follows: (A) Dt = -0.00, (B) Dt = 0.01, (C) Dt = 1.00, (D) Dt = CC, 1 c-( ), .. ex ---. 2 :rol .. 2o-(16) Equation 16, a is the variance, xis the distance, and Cis the concentration gradient The other contribution to solute-zone dispersion is related to the flow phenomenon.20 Because the mobile phase of the CE system behaves like a 23

PAGE 39

Newtonian fluid, any stress applied to the mobile phase will cause a proportional change in the fluid velocity or position.zo.z1 This is not entirely true when the flow is first initiated, since the fluid behaves like a non Newtonian fluid until the steady state of the system has been reached.zo,21 Any applied stress to the fluid within this time frame will only compress the molecules until their densities become a function of the current temperature and pressure. Although this non-Newtonian behavior leads to deviations from the Newtonian model, such deviations typically become negligible when the steady state has been reached. The contributions to solute-zone dispersion will therefore be discussed as a function of the Reynolds number associated with the turbulent flow-profile of the CE system. 2 o The Reynolds number is given by equation 17, where pis the density of the fluid, u is the linear velocity, dis capillary diameter for the opentubular capillary, and 1J is the absolute viscosity of the fluid.20 The numerator of equation 17 is representative of the inertial forces (pu2 ) in the fluid, while the denominator is representative of the viscous forces ( 1J ujd) that are present. The flow profile is governed by the interaction between inertial and viscous forces, since inertial forces contribute to the continuous motion of the fluid and viscous forces have the propensity to restrict this motion.20 24

PAGE 40

, (17) ,., Equation 17, pis the density of the fluid, u is the linear velocity, dis capillary diameter for the open-tubular capillary, and 17 is the absolute viscosity of the fluid The flow profile of the CE system is described as being turbulent because the inertial forces outweigh the viscous forces, and therefore the Reynolds number is higher than it would be if viscous forces dominated.ZZ The contributions from inertial forces tend to cause random fluid motion, typically observed in the form of eddies, vortices, etc. The Reynolds number is indicative of this chaotic fluid motion, although it is not always possible to determine at which Reynolds number the flow becomes turbulent versus laminar for the open-tubular capillary. Experimentally it has been shown that a Reynolds number above 4000 is most likely associated with turbulent flow, whereas those less than 2100 are most likely associated with laminar flow. The region between 2100 and 4000 is therefore the transition region.23 The laminar-turbulent transition arises from the boundary layer effect. The boundary layer is described as being the fluid layer that is near the surface of the capillary wall.22 The layer physically extends from the wall of the capillary to the field region, where external forces can disrupt the flow pattern. In the transition from laminar to turbulent flow, viscous forces cause a disruption in the non-viscous flow of the field region. This causes more 25

PAGE 41

random fluid motion characteristic of turbulent flow,22 The critical velocity (equation 18) is described as being the point at which eddies and chaotic motion cause turbulent flow, without an increase in volume flow rate. The critical velocity depends on the viscosity of the medium, the density of the fluid, and the radius of the tube. Because there is an increase in resistance associated with turbulence, a large amount of pressure would be needed to increase the volume flow rate.24 &!1. vm,2pr (18) Equation 18, Re is the Reynolds number, TJ is the viscosity of the fluid, pis the density, and r is the radius of the tube Lab-on-a-chip Concept The microfluidic device can often perform multiple laboratory functions, while only being millimeters to a few square centimeters in size.25 Since sample volumes within and slightly below the picoliter range are typically analyzed, this device offers several advantages in terms of sample waste generation, and amount of both reagents and crude sample that are required.zs,26 Other advantages include faster analysis and response times, the ability to achieve rapid heating due to increased surface to volume ratios and small heat capacities, and the opportunity for high-throughput analyses. 26

PAGE 42

This type of device is also relatively inexpensive to fabricate, which makes the mass production of disposable chips a reasonable possibility.26,27,28 Although CZE applications are being explored, the lab-on-a-chip technology is still not fully developed.28 Another disadvantage is that small scale analyses often coincide with more intense physical and chemical effects, such as forces within the capillary and surface roughness. The ability to obtain reasonable signal-to-noise ratios may also reflect how well the manufacturer was able to scale-down detection parameters to fit the particular device. An additional concern is that manufacturers may not be able to fabricate devices with high accuracy and precision since they are very small. 26.za The lab-on-a-chip microfluidic device, like any analytical device, has several advantages and disadvantages that must be evaluated in terms of the specific laboratory application. In this work, a simple microfluidic device was used to separate a simple mixture containing tyramine and ascorbic acid. The actual dimensions of the device are as follows: 5.1 em length, 2.6 em width, and 3.9 em channel length (approximately 200 11m i.d.). The diagram below shows the components of the device when it is in full operational use, including buffer and sample wells, electrodes, a high voltage power supply, light source, and light detector. A potential is applied between the ends of the 3.9 em length channel by placing an electrode and background buffer solution in each well. 27

PAGE 43

Sample introduction is then achieved by 1) momentarily applying potential between the ends of the channel that runs perpendicular to the separation channel, or by 2) removing buffer from the well at the anode end, filling with sample and momentarily applying potential, then removing the sample and refilling with buffer solution. Ughtsource Light Detector Figure 8. Illustration of the components of the microfluidic device used in this work Capillary Modification Depending on the CZE application, capillary modification can be used for several purposes.29 Some advantages of modification include the ability to 1) change the charge density at the capillary wall and 2) reverse the direction of flow so that anions may be detected before cations.29 Because the charge density at the capillary wall and the zeta potential are proportional, the 28

PAGE 44

increase/decrease in charge density resulting from modification effectively changes the EOF mobility.12 The ability to reverse the direction of electroosmotic flow through the adsorption of cations onto the capillary wall is useful in that anionic analytes can be detected before neutral and cationic analytes, in time-sensitive applications.29 The over-arching goal of this project was to minimize the capillary electrophoresis separation time and still be able to perform separations with high efficiency and resolution. The focus was placed on determining if and how the polymer concentration during capillary modification and the capillary length influence the electroosmotic flow rate, separation efficiency, and resolution for the separation of anions and cations. As a result, capillary modification was accomplished by 1) using hydrostatic pressure to flow a 1 M sodium hydroxide solution through the capillary to deprotonate the silanol groups, and leave a permanent negative charge along the inner walls at pH 7, 2) then to use electrostatic deposition to layer a cationic polymer to create a permanent positive charge, and add an anionic polymer to restore the original polarity. Each polymer layer has an estimated total thickness of approximately 1-nm, which suggests that it would be possible to continue layering without observing a significant or even noticeable difference in elution time. 29

PAGE 45

Objectives This work explores how changing the length of the capillary influences the capillary electrophoresis parameters, without sacrificing separation resolution. By modulating the resistance (V /1 = R), the experimental parameters may be scaled-down from moderate-to minimal length capillaries to keep the electroosmotic flow rate essentially constant. The reductions in resolution and separation efficiency can therefore be minimized. Because capillary modification increases the experimental options for the separation of analytes, modification was also investigated. The goal was to determine whether modification influences the separation efficiency, resolution, and electroosmotic flow rate (via the elution time of hydroquinone). 30

PAGE 46

CHAPTER 2: EXPERIMENTAL CZE Setup The components of the capillary electrophoresis setup include the Spellman CZE1000R power supply, electrodes placed at both ends of the capillary (in buffer solution), a plexiglass interlock box, a Mikropack DH2000-BAL UV-VIS-NIR lightsource, an Ocean Optics USB2000 light detection system, and a DELL computer monitoring system equipped with the OOIBASE32 software program for absorbance, wavelength, and time measurements. Both ends of the capillary are placed in vials that contain aqueous buffer solution, and alligator clips connect the electrodes in each vial to the high voltage power source. The capillary is fed through at-piece, such that light enters the capillary in one direction and detection is accomplished at a 180 angle. 31

PAGE 47

Ocean Optics Light Detector Mikropack Lightsource Figure 9. Illustration of the lightsource and detector that are situated at a 180 angle about the capillary window, using at-piece component. Wavelength of Maximum Absorbance and Acquisition Channels The wavelength of maximum absorbance was established for each chemical species by placing a concentrated sample in a cuvette and connecting the lightsource and light detector to the platform, as shown in figure 10. In the OOIBASE32 program, a screen snapshot was taken of the absorbance versus wavelength plot, and the midpoint of the absorbance peak was used as the wavelength of maximum absorbance. The OOIBASE32 program has acquisition channels A thru F that can be set to different wavelengths, and therefore several wavelengths can be monitored simultaneously depending on the components of the sample being analyzed. 32

PAGE 48

Figure 10. Diagram showing how the wavelength of maximum absorbance is determined Capillary Preparation-Part I (unmodified capillaries) The length of the capillary was carefully measured and the ends were cut at 60 angles to ensure that solution would easily flow in and out of the capillary. Nitrogen gas was used as a pressure source to flow 1.054 M NaOH solution through the capillary for 15 min, followed by 0.05 M sodium phosphate mono-/di-basic buffer solution (15 min). To prepare the capillary for the CZE experiment, the polymer coating was removed from a small section of the capillary, and the capillary was fed through the t-piece so that this optically transparent portion was aligned with the light source. The approximate distance from the introduction end of the capillary to the optically transparent portion is summarized in table 1 for the five different length capillaries that were evaluated in this work. Table 1. The length from capillary introduction to the detection cell, for various length capillaries Capillary Length (em) 60 so 40 25 12 Length to Detector (em) 45 35 25 15 6 33

PAGE 49

Capillary PreparationPart II (PDDA modified capillaries) After being conditioned with sodium hydroxide and phosphate buffer solution, hydrostatic pressure was used to flow a 0.01 M poly(diallyldimethyl-ammonium chloride) solution through the capillary for 15 min. The addition of PDDA to the unmodified capillary creates a permanent positive charge along the inner wall of the capillary. With the charge at the wall being positive, the direction of electroosmotic flow is reversed so that flow is initiated at the cathode and is directed to the anode. By changing the polarity mode from positive to negative, the cathode resides at the capillary introduction end while the anode resides at the exit end of the capillary. Sample introduction can then be accomplished in the typical manner, by injecting at the capillary introduction end since the charge along the wall is positive and the electroosmotic flow is directed from the cathode to the anode. Figure 11 illustrates the permanent charge, and the polarity reversal so that detection can be accomplished at the same relative position on the capillary. 34

PAGE 50

Ground End Bectrode Capillary Wall Bectroosmotlc Flow HV Bectrode + Figure 11. Depiction of the electrostatic interaction between the Siogroups at the capillary wall and the cationic polymer, poly(diallyldimethylammonium chloride). Capillary PreparationPart II (PSS modified capillaries) After electrostatically depositing the cationic polymer (PDDA) onto the capillary wall, hydrostatic pressure was used to flow a solution of poly( sodium 4-styrenesulfonate) having a concentration of 0.005-, 0.0 1-, 0.015-, or 0.02-M through the capillary. The addition of PDDA creates a permanent positive charge along the capillary wall, and the electrostatic deposition of PSS onto this layer creates a permanent negative charge at the wall. Each polymer layer changes the charge density at the wall, and the final process of modifying with PSS restores the original polarity so that the direction of electroosmotic flow is directed in the typical manner, from anode to cathode with the charge at the wall being negative. Figure 12 depicts each polymer layer and the direction of electroosmotic flow through the capillary. 35

PAGE 51

HV Bectrode + Bectrooamotlc Row ------. Ground End Bectrode Figure 12. Depiction of the electrostatic interaction between the cationic polymer, poly(diallyldimethylammonium chloride) and the anionic polymer, poly(sodium 4-styrenesu lfonate). Reagents and Sample Preparation To prepare the 0.05 M phosphate buffer solution, sodium phosphate monobasic (Lot No. 954703; assay 99.0%) from Fisher Scientific, and anhydrous sodium phosphate dibasic (Lot No. 53H0264; reagent grade) from Sigma Chemical Company, were combined in solution using deionized water. The 1.054 M sodium hydroxide solution was prepared by dissolving reagent grade sodium hydroxide pellets in deionized water. The poly(diallyldimethylamonium chloride) solutions were prepared in concentrations of 0.005and 0.0 1-M (concentration with respect to the monomer formula weight), using 0.05 M phosphate buffer solution and poly( diallyldimethylamonium chloride) from Sigma Aldrich (Lot No. 36

PAGE 52

10930PH; 20 wt.% in water). The poly(sodium 4-styrenesulfonate) powder from Sigma Aldrich was used to make 0.005-, 0.01-, 0.015-, and 0.02-M PSS solutions (concentration with respect to the monomer formula weight), using 0.05 M phosphate buffer solution. Simple mixtures were prepared from hydroquinone crystals (G92R-1; photo purified) from Mallinckrodt Chemical Works and dried L-ascorbic acid powder (Lot No. 213601; assay 99.0%) from J.T. Baker Chemical Company. Complex mixtures were prepared from(-)epinephrine (Lot No. 1416374) from Sigma Aldrich Company, 3hydroxytyramine hydrochloride (Lot No. 1417536) from Sigma Aldrich Company, tyramine (Lot No. S47658-488; assay 99.0%) from Sigma Aldrich Company, and D-tyrosine (Lot No. 1327221; assay 99.0%) from Sigma Aldrich Company. Setting Voltage and Current Analog dials were used to set the voltage and current readings from the high voltage power supply. The voltage was set at different values for capillaries having lengths of 60-, 50-, 40-, 25-, and 12-cm. In some instances, the current may also be adjusted relative to the voltage setting so that minimal or no fluctuations are observed in the baseline readings from the OOIBASE32 program, in the absorbance mode. 37

PAGE 53

DetectionSetting the Baseline Quantitative absorbance measurements using the Ocean Optics spectrometer were calibrated by setting the system response in the presence and absence of source radiation. Briefly, (in scope mode) the dark signal was stored with the light source powered off; and, the light signal was stored with the light source powered on. The mode was then switched from scope to absorbance, and the wavelength scale was monitored from 220to 550-nm. Data Acquisition Parameters The integration time was set to 10 msec, with an average of 5, a boxcar average of 1, and flash delay of 100 msec during acquisition of the absorbance signal. Introduction Method The data acquisition was initiated by sequentially pressing the timer and play button in the OOIBASE32 program. Following an initial 60 sec interval, the pause button was pressed, the lid of the plexiglass interlock box was opened (relieving the system of voltage from the power supply), the capillary introduction end and electrode were taken from the buffer vial and positioned in the sample vial, the lid on the interlock box was set back in 38

PAGE 54

place, and sample was introduced for 5 sec. The lid of the box was then removed again, the capillary end and electrode were repositioned in the buffer vial, and the lid was set back onto the lever that initiates the high voltage power supply. The play button was then pressed and data was acquired until all of the compounds eluted from the capillary. At the conclusion of the experiment, the stop button was pressed and data acquisition was ceased. Microfluidic Device The channel was aligned with the light source so that detection was accomplished at approximately 2.6 em from the introduction end. A potential was applied between the 3.9 em length channel by placing an electrode in the introduction and exit wells, and filling each well with the 0.05 M phosphate buffer solution. The baseline was set according to the "DetectionSetting the Baseline" section and the data acquisition parameters were set according to the "Data Acquisition Parameters" section. Sample introduction was accomplished after an initial 60 sec interval by removing buffer from the well at the anode end, filling with sample and applying potential for 5 sec, then removing the sample and refilling with buffer solution. Table 2 summarizes the experimental parameters. 39

PAGE 55

Table 2. Experimental parameters associated with the separations conducted with the microfluidic device. Length to Detector (em) Polarity Mode Voltage (kV) Current (uA) 2.6 (+) 0.5 30 Experimental Tests and SeparationsThe following CZE experiments were conducted to evaluate the necessary experimental conditions to optimize the measurement system. 1) Capillary electrophoresis methods were used to determine the elution times associated with pure samples of hydroquinone and ascorbic acid, using an unmodified 75 um i.d., 41.5 em length capillary. The voltage was set to 10 kV and the current was 150 uA. 2) A simple mixture containing hydroquinone and ascorbic acid was separated using an unmodified 50 um i.d., 50 em length capillary. Voltage and current readings were 10 kV and 42 uA, respectively. 3) A hydroquinonejascorbic acid mixture was introduced to, and separated within 50 um i.d. capillaries having lengths of 50-, 25-, and 12-em. The voltage and current readings for the 50-, 25-, and 12-em length capillaries were 10 kV and 42 uA, 10 kV and 78 uA, and 3 kV and 65 uA, respectively. 4) Pure samples of hydroquinone were introduced to an unmodified 50 um i.d., 50 em length capillary. The average elution times were documented at 7-, 10-, 13-, and 15-kV. 40

PAGE 56

5) By method of trial and error, the optimal voltage and current readings were established for unmodified 50 um i.d. capillaries having lengths of 60-, 50-,40-, 25-, and 12-cm. These average readings were used as a reference for setting voltage and current in subsequent experiments. 6) The variations in voltage and current were documented (for a total of two days) for a 50 um i.d., 50 em length capillary modified with 0.005 MPSS. 7) Simple hydroquinonejascorbic acid mixtures were introduced to, and separated within unmodified 50 um i.d. capillaries having lengths of 60-, 50-, 40-, 25-, and 12-cm, respectively. The voltage and current readings are summarized in table 8 of the Results and Discussion section. 8) A pure hydroquinone sample and simple hydroquinonejascorbic acid mixture were introduced to a 50 um i.d., 50 em length capillary modified with 0.01 M PDDA. Detection was accomplished in negative polarity mode, with voltage and current readings of 10 kY and 24 uA. 9) A simple hydroquinonejascorbic acid mixture was separated within a 50 um i.d., 50 em length capillary modified with a 0.005 M PDDA solution. The compounds were detected in negative polarity mode, with voltage and current readings of 10 kY and 33 uA. 41

PAGE 57

10) Pure samples ofhydroquinone were introduced to 50 um i.d. capillaries having lengths of 60-, 50-, 40-, 25-, and 12-em, which were modified with 0.005-, 0.01-, 0.015-, and 0.02-M solutions of PSS. 11) A complex mixture (containing epinephrine, 3-hydroxytyramine hydrochloride, tyramine and tyrosine) was introduced to, and separated within an unmodified 50 um i.d., 40 em length capillary. The voltage and current readings were 8 kV and 36 uA, respectively. 12) The complex mixture containing epinephrine, 3-hydroxytyramine, tyramine, and tyrosine was separated within an unmodified 50 um i.d., 12 em length capillary, with voltage and current readings of 0.5 kV and 6 uA. 13) A pure sample of hydroquinone was introduced to the microfluidic device having the following dimensions: 5.1 em length, 2.6 em width, and 3.9 em channel length. The voltage and current readings were 0.5 kV and 30 uA, respectively. 14) A mixture containing tyramine and ascorbic acid was introduced to the microfluidic device (having an inner diameter of approximately 200 11m). The separation potential and current were 0.5 kV and 30 uA, respectively. 42

PAGE 58

Chapter 3: Results and Discussion Method DevelopmentPure samples of hydroquinone and ascorbic acid were introduced to a 75 urn i.d., 41.5 em length capillary. The elution times associated with pure samples of hydroquinone and ascorbic acid are summarized in table 3 for a 75 um i.d., 41.5 em length capillary. Since the pKa of hydroquinone is 10.35 and the pKa of ascorbic acid is 4.10, hydroquinone is neutral at pH 7.0 and ascorbic acid is negatively charged. This means the retention time associated with hydroquinone is shorter than that for ascorbic acid at neutral pH (e.g. 361 3 sec compared to 931 21 sec), in the positive polarity mode. The elution time of hydroquinone is associated with the electroosmotic flow rate, and can therefore be used to evaluate the EOF. The electroosmotic mobility can be calculated from equation 19, where to is the elution time of an electrically neutral analyte. For the 75 um i.d., 41.5 em length capillary, the electroosmotic mobility is calculated from the following: 0.415 m capillary tube length, 0.27 m length from introduction to detection, 10,000 V potential across the tube, and 361 sec migration time associated with hydroquinone. The calculated electoosmotic mobility is therefore 3.7x10-B m2/Vfs. By decreasing the length of the capillary, the 43

PAGE 59

applied potential decreases and the electroosmotic mobility should effectively remain the same. The relationship between EOF and the current through the capillary is evident when considering the proportionality between voltage and current, and the change in the elution time of hydroquinone based on the simultaneous increase/decrease in voltage and current. (19) Equation 19, Lr is the length of the capillary tube in meters, Lois the distance from introduction end of the capillary to the detection window (in meters), Vis the potential difference in volts across the tube, to is the migration time of an electrically neutral analyte (in seconds), and J.lo is the electroosmotic mobility (in m2/V /s) Sample calculation using equation 19: -3.7.rlO ... (m: IV l.s) (IO.OOOVX299.726.s) 44

PAGE 60

Table 3. Elution of hydroquinone and ascorbic acid from a 7S um i.d., 41.S em length capillary, conditioned with l.OS4 M NaOH and O.OS M sodium di-/monobasic buffer solution. -E > u ... ra..c: ::r., a. c: "' Cll U...l 41.5 41.5 > .. -Cll ... "0 ...!l! 0 (+) (+) > =-CII till "' .. 0 > 10 10 .:. .. c: Cll ... ... ::II u 150 150 *HQ-abbreviation for hydroquinone *AAabbreviation for ascorbic acid Cll ii E "' 11'1 HQ AA Cll 7 Cll 7 -E E 0 O'i= .;!!. i= .;!!. ... c:t Cll ::1: c: > c: > .a Cll c:t 0 Cll 0 "0 "0 aj:t:l aj:t:l ::II-"0 "0 .. .. w...; 11'1 w..; 11'1 5 361 :t::3 --5 931 :t::21 This preliminary data suggests that separation of simple mixtures can be reliably accomplished using moderate length capillaries that have an inner diameter of 75 um. Method Development-By testing the electrophoretic system, it was found that separation of a simple mixture could be achieved using a SO urn i.d., SO em length capillary. Elution times associated with pure and mixed hydroquinonejascorbic acid samples are documented in table 4 for a 50 um i.d., 50 em length capillary conditioned with 0.05 M phosphate buffer. For a series of measurements taken with the same experimental parameters, the elution times are consistent. The reproducibility of the data helps to confirm that 1) 45

PAGE 61

the current method generates results which fall into a narrow time interval for a given data set (i.e. low standard deviation relative to elution time), and 2) the elution time for a pure sample of hydroquinonejascorbic acid is similar to that obtained from a mixture of hydroquinone and ascorbic acid (e.g. hydro elution: 485 3 sec for pure sample, and 494 2 sec for mixture). For these measurements, the voltage and current readings remained constant. Table 4. Elution of hydroquinone and ascorbic acid from a SO um i.d., SO em length capillary, conditioned with 1.0S4 M NaOH and O.OS M sodium di-/monobasic buffer solution. -E > u ... .. =Ill) a. c ra Gl U....l so so so (+) (+) (+) > =-QI Ill) ra .. 0 > 10 10 10 42 42 42 .!! a. E ra VI HQ AA HQ/AA Mix *HQ-abbreviation for hydroquinone *AAabbreviation for ascorbic acid 0 ... Ql ..D Ill E;; ::Is s s 48S ::1::3 --1143 494 ::1::2 1098 This data suggests that separation of simple mixtures can be reliably accomplished using moderate length capillaries that have a smaller inner diameter of 50 um. 46 ::1::37 ::t15

PAGE 62

Elution Time and Capillary Length-By changing the capillary length, it was found that data becomes less reproducible due to experimental parameters. The ultimate goal of this portion of the project was to establish an acceptable method for separating chemical species on a smaller scale using a microfluidic device. The elution of hydroquinone from capillaries of various lengths will therefore be discussed, as the elution time of any chemical species (whether charged or neutral) is directly related to capillary length. Since the velocity of electroosmotic flow is decreased as the capillary length is shortened, it is expected that a proportional decrease in EOF will be associated with a proportional decrease in capillary length. Figure 13 depicts the average elution times ofhydroquinone from a simple hydroquinonejascorbic acid mixture. The separation conditions are summarized in table 5. The elution times clearly differ with capillary length, and are internally consistent within each data set. However, this data was taken before a reliable method was established for setting the experimental parameters for different length capillaries. So even though this data is useful, the Setting Parameters section discusses the importance of establishing a method that works and can be used for essentially every experiment of this type to collect reproducible data. Further discussion of this data set is also referenced in the Separation Efficiency and Resolution section. 47

PAGE 63

Absorbance vs Time 2.5 2 1.5 Ill ... c "' n ..c 1 ... 0 Ill -5o em -25cm ..c c:c 0.5 12 em 0 100 150 200 250 300 350 400 450 500 550 600 -0.5 Time(s) Figure 13. Depiction of the elution time of hydroquinone as a function of capillary length, for unmodified capillaries having lengths of 50-, 25-, and 12-cm Table 5. Elution of hydroquinone and ascorbic acid from 50 um i.d. capillaries having lengths of 50-, 25, and 12-cm. Capillaries were conditioned with 1.054 M NaOH and 0.05 M sodium di-/mono-basic buffer solution prior to sample introduction. > < Ql -E -E Ill > 1.1 2. 0 ai= .:!:!. .. -> Ql .. .. > Ql Ql .. 110 c ii ..D :::1: c Ql =ta -Ql Ill Ql 0 .. '1:1 '1:1 ..!!! 0 .. .. E a. c 0 .. :I ->:I-'1:1 Ill Ql .f::E :I Ill .. > :I-c:c-"' ..... 1098 270 214 iii .:!:!. > Ql '1:1 '1:1 .. 11'1

PAGE 64

This data suggests that separation of simple mixtures can be reliably accomplished using shorter length capillaries that have an inner diameter of SO um. Because the potential is expected to decrease with capillary length, the goal is to set the voltage so that the current readings remain about the same regardless of the capillary length (e.g. set the voltage so that the current reads 30 uA). Setting Parameters-By examining the reproducibility of elution times from different length capillaries, it became evident that some method needed to be developed for reliably setting voltage and current. The potential was kept at 10 kV for the separations outlined in table 4 for a SO um i.d., SO em length capillary. This is because preliminary tests showed that 1) as the separation potential dropped much below 10 kV for a SO um i.d., SO em length capillary, the current subsequently dropped to a point where chemical species were eluting at a much slower rate or no longer eluting from the capillary, and 2) as the separation potential increased much above 10 kV, the solution began to boil and the capillary eventually shorted-out. Random and frequent baseline fluctuations were another indication that the potential was in excess of 10 kV. Table 6 summarizes the results obtained for an unmodified SO um i.d., SO em length capillary when the separation potential was varied between 7-and 1S-kV. 49

PAGE 65

Table 6. Elution of hydroquinone from an unmodified 50 um i.d., 50 em length capillary when the voltage and current were adjusted to various settings Ave. HQ Capillary Polarity Voltage Current Number Elution Std dev Additional Length Mode (kV) (uA) of Trials Time/EOF (s) Comments (em) (s) 50 (+) 36 3 850 N/A 50 (+) 54 3 475 N/A Poor 50 (+) 78 3 329 Resolution Current 50 (+) 102 1 269 N/A dropped after Triall *HQ-abbreviation for hydroquinone Intuitively, both voltage and current were experimentally found to be dependent on, or a function of capillary length. To achieve consistent results, either the voltage was adjusted to an appropriate setting relative to the capillary length (i.e. a maximum voltage of 10 kV was employed for 50and 60-cm length capillaries), or the current was set to be consistent. By the method of trial and error, the voltage was set and adjusted for unmodified 50 urn i.d. capillaries having lengths of 60-, 50-, 40-, 25-, and 12-cm, respectively. Although there were slight fluctuations in regards to the potential and current readings that would give rise to consistent separations from one run to the next, the most common readings are summarized in table 7 below. A table summarizing more considerable fluctuations in the current (before equilibration was established) for the modified 50 em length, 0.005 M PSS capillary can also be found in table 8. Since the capillary was tested for so

PAGE 66

two days, the Set Number refers to data acquisition files that were saved during periods of little fluctuation. During periods in which there was considerable variability in current, data acquisition was terminated due to rapid and unpredictable baseline fluctuations. Table 7. Average voltage and current readings associated with unmodified capillaries of various lengths, in positive polarity mode. Capillary Length (em) Polarity Mode Ave. Voltage (kV) Ave. Current (uA) 60 (+) 10 25 50 (+) 10 38 40 (+) 9 29 25 (+) 5 36 12 (+) 3 42 Table 8. Documented fluctuations in voltage and current readings associated with a SO um i.d., SO em length capillary modified with O.OOS M PSS. Capillary Day Trial PSS cone. Polarity Voltage Current Number Mode 50 1 (Set 1) 1 0.005 (+) 50 1 (Set 1) 2-3 0.005 (+) 50 1 (N/A) 1 0.005 (+) 50 1 (Set 2) 1 0.005 (+) 50 2 (Set 1) 1-5 0.005 (+) This data suggests that the voltage can be set to a specific value based on the capillary length so that the current is around 30to 40-uA. This method generates reproducible data for 50 urn i.d. capillaries. 51

PAGE 67

Separations ContinuedThe Setting Parameters section can be used as a reference for how experimental parameters were set for all separations from this point forward, unless otherwise specified. Different length capillaries were conditioned with 0.05 M phosphate buffer, and a simple mixture containing hydroquinone and ascorbic acid was introduced to, and separated within each capillary. The elution times summarized in table 9 were obtained from unmodified 50 um i.d. capillaries having lengths of 60-, 50-, 40-, 25-, and 12-em, respectively. Note that because the voltage was set to 10 kV for the 50 em capillary and 3 kV for the 12 em capillary in table 5, this data has been entered for comparison purposes into table 9. Although in the following section there is a discussion of modified capillaries, it is expected that elution times should be similar for both modified and unmodified capillaries of the same length. This is because the final modification process of adding PSS to the modified capillaries should restore the negative charge along the interior wall to being like that of the unmodified capillaries (as discussed in the introduction section). 52

PAGE 68

Table 9. Separation of ascorbic acid and hydroquinone via unmodified 50 um i.d. capillaries having lengths of 60-,50-,40-,25-, and 12-cm. > < Cl.l Cl.l E -E Ill E ..:. 0 .:!!. >u ai= c(j:: ... -> Cl.l .. ... > Cl.l Cl.l III.S: .. til) c a. .a ::z::: c Cl.l c( 5 .. -Cl.l Ill Cl.l 0 =til) ... "D .. ... E E"* ai:t:i "D ai:t:i a. c .!!! 0 ... "D Ill Cl.l :::s Ill :::s -.. U....l u II) II) 60 (+) 10 30 HQ/AA 5 798 1983 Mix so (+) 10 42 HQ/AA 5 494 1098 Mix 40 (+) 8 42 HQ/AA 5 356 718 Mix 25 (+) 5 36 HQ/AA 5 297 520 Mix 12 (+) 3 65 HQ/AA 5 124 214 Mix *HQ-abbreviation for hydroquinone *AAabbreviation for ascorbic acid This data confirms that separation of a simple mixture can be accomplished on moderate and shorter length capillaries having an inner u; .:!!. > Cl.l "D "D .. II) diameter of SO um. Because the data generated from 12 em length capillaries is reproducible, the experimental parameters are set so that separations could be accomplished with shorter length capillaries or a microfluidic device. 53

PAGE 69

Capillary ModificationModifying the inner wall of the capillary is advantageous for experimental reasons, which were discussed in the introduction and experimental sections. In this work, capillary modification is important in two ways. The first being that modification with a cationic polymer (i.e. PDDA) allows for the detection of negatively charged species before positively charged species, by reversing the polarity from positive to negative. Modification is also important through additional modification with an anionic polymer (i.e. PSS), which can alter the charge density at the wall and restore the negative polarity. To ensure that the PDDA-cationic polymer was in fact coating the inner walls of the capillary, an unmodified 50 um i.d., 50 em length capillary was conditioned with 0.05 M sodium phosphate buffer solution, followed by the 0.01 M PDDA solution. In positive polarity mode the anode resides at the capillary introduction end, while the cathode resides at the exit end, as previously mentioned. An unmodified capillary conditioned with sodium hydroxide will have a permanent negative charge residing along the interior wall, which means the buffer cations will migrate through the capillary toward the negatively charged electrode at the exiting end. When the capillary is modified with the PDDA-cationic polymer, a permanent positive charge then resides along the interior wall, and the direction of 54

PAGE 70

electroosmotic flow is reversed by switching from positive to negative polarity mode. A pure hydroquinone sample and simple mixture containing hydroquinone and ascorbic acid were introduced to the PDDA coated capillary. For the simple mixture, the elution of ascorbic acid followed by hydroquinone in negative polarity mode verifies that modification has been successfully accomplished. The negatively charged ions of ascorbic acid are attracted to the positively charged anode that now resides at the exiting end, while the neutral hydroquinone molecules continue to elute with the electroosomotic flow of buffer solution. The separation conditions and elution times for hydroquinone and ascorbic acid are summarized below in table 10. Note that the retention time of hydroquinone is almost twice that of ascorbic acid (e.g. 959 15 sec compared to 486 9 sec). Table 10. Elution of hydroquinone and ascorbic acid from a SO um i.d., SO em length capillary modified with 0.01 M PDDA. -E >u ... n:J..C =bo CL C "' CLJ U....l 50 so CLJ ... '0 .!!! 0 (-) (-) > =CLJ 1111 "' .... 0 > 10 10 24 24 CL E "' II) HQ HQ/AA Mix *HQ-abbreviation for hydroquinone *AAabbreviation for ascorbic acid 55 0 ... CLJ .a Ill Eiii ::II5 5 931 :t:S 959 486

PAGE 71

Absorbance vs Time O.OlM PDDA 0.1 : i 0.075 -j--------------I Q) u 0.05 c "' .a i i 0 Ill 0.025 .a
PAGE 72

Table 11. Elution of hydroquinone and ascorbic acid from a simple mixture that was introduced to a SO um i.d., SO em length capillary modified with O.OOS M PDDA. -E ni.C =ro a. c "' cu U-1 50 (-) > =. cu 1111 "' 0 > 10 < ..=. ... c cu ... ... ::II u 33 cu a. E "' 11'1 HQ/AA Mix *HQ-abbreviation for hydroquinone *AAabbreviation for ascorbic acid 0 ... cu .Q Ill Eiii ::II5 1249 547 This data confirms that capillary modification with PDDA allows for the reversal of electroosmotic flow, and further suggests that modification of shorter length capillaries would provide more information about modification of the channel in a microfluidic device. Changing Length and PSS Concentration of Modified CapillariesThe main goal was to minimize the entire CE setup, however, elution times were also studied as a function of PSS concentration. Table 12 shows the elution times ofhydroquinone and ascorbic acid from SO urn i.d., 25 em length unmodified and modified capillaries. The hydroquinone elution times from simple mixtures are similar for the unmodified capillary and 0.01 M PSS modified capillary (133 2 sec versus 146 1 sec). The ascorbic acid elution times are also similar, although they differ by nearly 100 sec (270 2 sec versus 366 6 sec). Table 13 includes 57

PAGE 73

comparable results for the elution of hydroquinone from 50 urn i.d., 50 em length unmodified and modified capillaries. The elution time of hydroquinone from the unmodified capillary was found to be 485 3 sec, while that from the 0.01 M PSS capillary was 385 17 sec. Arguably, the elution times could differ because of discrepancies in voltage and current readings, and perhaps even from small differences in the distance from the introduction end of the capillary to the detection window. The discrepancies in elution times may also be due to the sample introduction. Sample introduction is accomplished by pausing the data acquisition to: remove the lid on the plexiglass box, place the capillary end and electrode in the sample vial, replace the lid on the plexiglass box to apply potential for the 5 sec introduction, remove the lid and transfer the capillary and electrode to the background buffer, and finally place the lid back in place and press the play button for data acquistion. The computer program documents the exact times that the pause and play buttons are pressed so that this interval can be subtracted from the "elution time" of each compound. There may however be some discrepancies in the amount of time that elapses between the final steps in which 1) the potential is applied once the capillary is back in the buffer solution, and 2) the play button is pressed. Although the response time could undoubtedly vary from run to run, the elution times are fairly consistent. 58

PAGE 74

Table 12. Comparison of elution times associated with an unmodified 50 um i.d., 25 em length capillary versus a 50 um i.d., 25 em length capillary modified with 0.01 M PSS. -E > u ... 111..C =ro Q. c "' cu U....l 25 25 u c 0 u 11'1 11'1 11. N/A 0.01 (+) (+) > cu !Ill "' .. 0 > 10 10 78 90 *HQ-abbreviation for hydroquinone *AAabbreviation for ascorbic acid cu -a. E "' 11'1 HQ/ AAMix HQ/ AAMix 0 ... cu .a Ill Eiii ::II5 5 133 146 Table 13. Comparison of hydroquinone elution times associated with an unmodified 50 um i.d., 50 em length capillary versus a 50 um i.d., 50 em length capillary modified with 0.01 M PSS. -E 111..C =tO Q. c "' cu U....l 50 50 u c 0 u 11'1 11'1 11. N/A 0.01 > .. -cu ... "CI ..!!! 0 (+) (+) > =-cu !Ill "' .. 0 > 10 10 < ::II .. c cu ... ... ::II u 42 48 *HQabbrev1at10n for hydroqumone *AAabbreviation for ascorbic acid .!! Q. E "' 11'1 HQ HQ 0 ... cu .a Ill Eiii ::II5 5 485 385 Because of the similarities between unmodified and modified capillaries, additional experiments were done to evaluate the elution times associated with 50 um i.d. capillaries having various lengths, which were 59 270 366

PAGE 75

conditioned with varying concentrations of PSS. Figure 15 was generated from time and absorbance data acquired from a SO um i.d., 60 em length capillary modified with a 0.01 M solution of PSS. The average elution time of hydroquinone associated with the 60 em length capillaries modified with varying concentrations of PSS are documented in table 14. When comparing the elution times summarized in this table, it becomes apparent that the elution of hydroquinone is not significantly dependent on the PSS concentration. The separation parameters were kept relatively constant from one capillary to the next, given the average separation voltage and current readings of 10 kV and 24 uA, respectively for all 60 em length capillaries. This means a significant variation in hydroquinone elution time relative to an increase in PSS concentration was not clearly observed, as the documented variability could be (at least in part) attributed to other factors such as: equilibration (i.e. not allowing enough time for the voltage and current to establish steady values before starting data acquisition), random fluctuations in voltage/current readings, exact distance from capillary introduction to detection cell. The latter would understandably result if the capillary window was aligned with the light source but some transparent section of the window extended towards the introduction end of the capillary. The physical measurement from introduction to detection cell would therefore only be 60

PAGE 76

considered a crude estimate, which again could explain (in part) why the elution times differ by nearly 60 sec. Absorbance versus Time 60cm, O.OlM PSS 0.15 0.125 0.1 G.l 1.1 1: 0.075 Ill ..c ... 0 0.05 Ill ..c < 0.025 0 -... ..... .. 100 200 300 400 500 600 700 -0.025 Elapsed time (s) Figure 15. Diagram depicting the elution of hydroquinone from a 50 um i.d., 60 em length capillary modified with 0.01 M PSS. Table 14. Comparison of hydroquinone elution from SO um i.d., 60 em length capillaries modified with 0.005-, 0.01-, 0.015, or 0.02-M PSS. 8 Capillary PSS Voltage Current Number of Ave. HQ Std dev Length (em) cone. (kV) (uA) Trials Elution Time (s) (M) (s) 60 0.005 10 24 5 647 60 0.01 10 24 5 669 60 0.015 10 24 5 735 60 0.02 10 24 5 732 *HQ-abbreviation for hydroquinone 61 00

PAGE 77

This is not meant to definitively state that the hydroquinone elution times or electroosmotic flow rates should be identical for equal length capillaries that have been modified with differing concentrations of PSS. However, such findings do raise interesting questions about what should be expected when differing concentrations of PSS are introduced to capillaries that have the same length. The volume of electrolyte contained in a 50 urn i.d., 60 em length capillary is approximately 1.7x10-3 mL which is equivalent to about one-twentieth of a single drop of water. When considering how the polymer concentration might influence the elution time of both neutral and charged species, it seems reasonable that even the lowest concentration of polymer would fully saturate the inner wall of the capillary if introduced for a long enough period of time. The 15 min introduction would presumably allow adequate time for the polymer to completely coat the inner wall, and in this case the elution time would be more a function of introduction time versus polymer concentration. Table 15 was generated from the elution data associated with 50 urn i.d., 25 em length capillaries modified with varying concentrations of PSS. Like the 50 urn i.d., 60 em length capillaries previously discussed, the average hydroquinone elution times are similar regardless of the PSS concentration. The differences in average elution times are less than 50 sec, which is reasonable given the overall average elution time of 228 sec. Because the 62

PAGE 78

overall average elution time is 696 sec for the 50 um i.d., 60 em length capillaries modified with PSS, and the deviation from the average elution time is less than 100 sec, the results are similar for capillaries that have different lengths. There is a proportional difference in the elution times and standard deviations associated with the 60 em length and 25 em length capillaries. Table 15. Comparison of hydroquinone elution from SO um i.d., 25 em length capillaries modified with 0.005-, 0.01-, 0.015, or 0.02-M PSS. Capillary PSS Voltage Current Number of Ave. HQ Std dev Length (em) cone. (kV) (uA) Trials Elution Time (::t::s) (M) (s) 25 0.005 5 54 5 207 ::1::1 25 0.01 5 30 5 238 ::1::2 25 0.015 5 30 5 216 ::1::6 25 0.02 5 30 5 251 ::t::2 *HQ-abbreviation for hydroquinone This data suggests that capillary modification with PSS does not dramatically influence the elution time of hydroquinone or the electroosmotic mobility. Capillary modification is advantageous for experimental purposes, and the fact that it does not alter the elution times is an interesting result. 63

PAGE 79

Efficiency and ResolutionBy minimizing capillary length there is an observed loss in efficiency and resolution. For this reason, it is important to understand 1) how dramatic such losses are and 2) whether the concentration of PSS has any influence on efficiency and resolution. To evaluate the separation achieved within unmodified capillaries (having lengths of 50-, 25-, and 12-cm), the separation efficiency was found using equation 10. Additionally, separations were evaluated in terms of resolution, which is given by equation 12. Note that the values reported in table 16 do not reflect an average of the five runs that were completed, but instead reflect the separation efficiency and resolution associated with a single run, having a retention time closest to the average tR for that particular capillary length. The following diagrams (figures 16, 17 and 18) illustrate the loss in both resolution and separation efficiency as the capillary length is decreased from 50to 25and finally to 12-cm. The Setting Parameters section, however, discussed the fact that voltage should be set according to table 7 and the current should be set to some maximum position relative to this voltage setting. Since the data associated with the 25and 12-cm capillaries was obtained before the aforementioned method was established to set experimental parameters, the resolution and separation efficiency associated with the 25and 12-cm capillaries are N=5625.17, R=22.1 and N=690.219, R=17.7, respectively. As expected, the values associated with the 25 em 64

PAGE 80

length capillary differ from those referenced in table 17 for an unmodified capillary of similar length, since the first capillary was evaluated with a separation potential of 10 kV and the second capillary was evaluated with separation potential of 5 kV. 2.16 2.14 2.12 fl 2.1 c l5 2.08 "' ..a C( 2.06 2.04 2.02 2 Absorbance vs Time --50 em 2.024 2.024 485 487 489 491 493 495 497 499 501 503 505 Time (s) Figure 16. Diagram depicting the hydroquinone elution from an unmodified 50 um i.d., 50 em length capillary. The separation efficiency is calculated from the peak width at the baseline and the resolution is calculated from the full width at half the maximum peak height. 65

PAGE 81

Absorbance vs Time ---25 em 1.14 1.12 1.1 Gl 1.08 u c I'll .a 1.04 "' .a C( 1.02 1 0.98 0.96 125 127 129 131 133 135 137 139 141 Time (s) Figure 17. Diagram depicting the hydroquinone elution from an unmodified 50 um i.d., 25 em length capillary. The separation efficiency is calculated from the peak width at the baseline and the resolution is calculated from the full width at half the maximum peak height. 0.12 0.1 0.08 Gl u c 0.06 I'll .a "' 0.04 .a C( 0.02 0 100 .02 Absorbance vs Time --12 em -0.001 105 110 115 120 125 Time(s) -0.001 -. 130 135 140 145 150 Figure 18. Diagram depicting the hydroquinone elution from an unmodified 50 um i.d., 12 em length capillary. The separation efficiency is calculated from the peak width at the baseline and the resolution is calculated from the width at half the maximum peak height. 66

PAGE 82

Table 16. Calculated efficiency and resolution associated with unmodified SO um i.d. capillaries having lengths of 50-, 25-, and 12-em. Note that the voltage setting for the 25 em length capillary is in excess of the value mentioned in the Setting Parameters section. Capillary Length Ave. Voltage (kV) Efficiency Resolution (em) 50 10 30876 99.0 25 10 5625 22.1 12 3 690 17.7 Table 17. Calculated efficiency and resolution associated with unmodified SO um i.d. capillaries having lengths of 50-, 25-, and 12-em. Note that the voltage setting for the 25 em length capillary is 5 kV, which is in agreement with the optimal setting mentioned in the Setting Parameters section. Capillary Length Ave. Voltage (kV) Efficiency Resolution (em) 50 10 30876 99.0 25 5 6853 46.4 12 3 690 17.7 Because the established method for setting parameters generates reproducible data that can be evaluated in terms of separation efficiency and resolution, it is possible to determine exactly how much loss in efficiency and resolution there should be for minimal length capillaries. Table 17 summarizes data collected from capillaries that were evaluated under the appropriate experimental conditions, and since the losses are not significant enough to dramatically interfere with separation and detection, minimal length capillaries seem to be an acceptable option. With that being said, it is also necessary to understand whether or not efficiency and resolution are impacted by the concentration of PSS. 67

PAGE 83

Table 18 summarizes the average efficiency and resolution associated with 60 em length capillaries modified with varying concentrations of PSS. The average voltage and current readings remained the same for each capillary, regardless of PSS concentration; however, the average efficiency and resolution values clearly differ. The fact that such differences are not orders of magnitude less than or greater than any one value, suggests that the average efficiency and resolution for any 60 em length capillary modified with some concentration of PSS is somewhere around 28000 and 104, respectively. Table 19 summarizes the average efficiency and resolution associated with SO um i.d., 25 em length capillaries modified with varying concentrations of PSS. Notably, the average values are similar regardless of the PSS concentration, like that documented for the 60 em length capillaries. This means the average efficiency and resolution associated with 25 em length capillaries modified with some concentrations of PSS should lie somewhere around 4600 and 46, respectively. Table 18. Average hydroquinone elution time, separation efficiency, and resolution associated with SO um i.d., 60 em length capillaries modified with 0.005-, 0.01-, 0.015-, or 0.02-M PSS. Ave. HQ Capillary PSS cone. Voltage (kV) Current Elution Ave. Ave. Length (em) (M) (uA) Time (s) Efficiency Resolution 60 0.005 10 24 647 29000 100 60 0.01 10 24 669 32000 117 60 0.015 10 24 735 27000 108 60 0.02 10 24 732 25000 92 68

PAGE 84

Table 19. Average hydroquinone elution time, separation efficiency, and resolution associated with 50 um i.d., 25 em length capillaries modified with 0.005-, 0.01-, 0.015-, or 0.02-M PSS. Ave. HQ Capillary PSS cone. Voltage (kV) Current Elution Ave. Ave. Length (em) (M) (uA) Time (s) Efficiency Resolution 25 0.005 5 54 207 3900 37 25 0.01 5 30 238 5400 so 25 0.015 5 30 216 5200 55 25 0.02 5 30 251 3800 40 Table 20 summarizes the average voltage, current, hydroquinone elution time, efficiency, and resolution for 50 um i.d. capillaries having various lengths, modified with concentrations of PSS ranging from 0.005 M to 0.02 M. Tables 18 and 19 illustrate the similarities between the average efficiency and resolution for 60 em and 25 em length capillaries, thus an overall average provides more information about the expected values for 50 um i.d. capillaries having various lengths, modified with varying concentrations of PSS. 69

PAGE 85

Table 20. Average hydroquinone elution time, efficiency and resolution associated with 50 um i.d. capillaries having lengths of 60-, 50-, 40-, 25-, and 12-cm. For each capillary length, an average was taken of the voltage, current, hydroquinone elution time, efficiency and resolution associated with the four capillaries modified with 0.005-, 0.01-, 0.015-, and 0.02-M PSS. Number of Number of Capillary Capillaries Trials Ave. Ave. Ave. Ave. Ave. Length Considered Completed Voltage Current HQ Efficiency Resolution (em) in Average w/each (kV} (uA) Elution Capillary Time (s) 60 4 5 10 24 696 28000 104 50 4 5 10 38 485 14000 76 40 4 5 9 44 343 7200 62 25 4 5 5 36 228 4600 45 12 4 5 2.5 30 164 2200 23 This data suggests that the concentration of PSS is not important when the polymer solution is introduced to the capillary for 15 min. Because the hydroquinone elution times, separation efficiency, and resolution remained similar for capillaries having the same length, which were modified with varying concentrations of PSS, an average was taken of the experimental parameters and results for each capillary length. Complex Mixture (Part 1)A mixture containing four components with similar chemical properties were separated using an unmodified SO urn i.d., 40 em length capillary. The pKa values are similar for (-)-epinephrine, 3-hydroxytyramine hydrochloride, tyramine and tyrosine.3031 Note that table 21 includes the 70

PAGE 86

pKa and wavelength of maximum absorbance for each of these compounds, in addition to those values for hydroquinone and ascorbic acid for comparison purposes. Based on the pKa values, the order of elution is as follows: tyramine, tyrosine, 3-hydroxytyramine HCI (dopamine), and epinephrine.32 The absorbances of the four compounds were analyzed at 275 nm (channel D). The plots of absorbance versus elapsed time are summarized in figures 19, 20, and 21 for the three acquisition channels (channel D: 275 nm, channel E: 278 nm, and channel F: 280 nm, respectively), and based on the similarities in absorbance at each wavelength, a single acquisition channel was chosen for the overall analysis. Figures 22, 23, 24, and 25 illustrate the fact that each compound absorbs at 275 nm, even though this is not the wavelength of maximum absorbance for all four of the compounds. Table 21. The pKa value(s) and wavelength of maximum absorbance associated with the components of the complex and simple mixtures. Compound Temp. (C)/pKal Temp. Temp. Amax (nm) (C)/pKa2 (C)/pKa3 (-)-Epinephrine 280 3-Hydroxytyramine HCI 25/11 25/10.6 280 Tyramine -278 D-Tyrosine 25/2.20 25/10.46 275 Hydroquinone 20!1111 290 Ascorbic Acid 16/11.79 -265 71

PAGE 87

Channel D (2 7 5 nm) 0.08 0.06 QJ 0.04 u c I"'S -e 0.02 0 Ill 0 -0.02 -0.04 Elapsed Time (s) Figure 19. Diagram depicting the elution (from left to right) of: tyramine, tyrosine, 3-hydroxytyramine HCI (dopamine), and (-)-epinephrine, when 275 nm is monitored for absorbances Channel E (2 78 nm) 0.08 0.06 QJ u 0.04 c I"'S -e 0.02 0 Ill 0 -0.02 -0.04 Elapsed Time (s) Figure 20. Diagram depicting the elution (from left to right) of: tyramine, tyrosine, 3-hydroxytyramine HCI (dopamine), and (-)-epinephrine, when 278 nm is monitored for absorbances 72 600

PAGE 88

Channel F (2 80 nm) 0.08 0.06 llJ 0.04 u c -e 0.02 0 "' ..0
PAGE 89

Figure 23. Screen snapshot of the 3-hydroxytyramine HCI absorbance peak in the scope mode of the OOIBASE32 program. Figure 24. Screen snapshot of the tyramine absorbance peak in the scope mode of the OOIBASE32 program. 74

PAGE 90

Figure 2 5. Screen snapshot of the D-tyrosine absorbance peak in the scope mode of the OOIBASE32 program. The average elution times associated with the four compounds are summarized below in table 22, in addition to relevant experimental parameters. Because the standard deviations are low relative to the average elution times, the data is consistent. The absorbance of each compound was between approximately 0.1 and 0.3, which means the signal-to-noise ratios are low, but with the consistency in the elution times (of the eight trials taken), there is not a question of whether the experimental method generates reliable data. With the primary goal of this portion of the experiment being the separation of the four components using an unmodified capillary of moderate length, the next section expands on this work by evaluating the separation of such components using a shorter length capillary modified with PDDA and PSS. 75

PAGE 91

Table 22. Elution of the four compounds contained in the complex mixture from an unmodified SO um i.d., 40 em length capillary. 40 40 (+) (+) Ill 1111 "' .. 8 8 2. .. c Ill ... ... ::I u 36 36 "' iii ;: .... 0 ... Ill ..a E ::I z 5 8 c 0 ;; ::I iii Eiii Ill Ill E 322 322 *Tm abbreviation for tyramine *Ts-abbreviation for tyrosine ::t:3 340 ::t:4 ::t:2 340 ::t:3 *HTm-abbreviation for 3-hydroxytyramine HCI (dopamine) *EN abbreviation for epinephrine c 0 ;; 2 w E Ill Ill E 356 356 ::t:4 ::t:3 This data suggests that separation of a complex mixture can be accomplished reliably using moderate length capillaries by setting the experimental parameters according to those outlined in the Setting Parameters section. Complex Mixture (Part II) The complex mixture containing epinephrine, 3-hydroxytyramine, c 0 ;; 2 w Ziii w Ill Ill E 485 485 tyramine, and tyrosine was separated within an unmodified 50 um i.d., 12 em length capillary. Table 23 summarizes the experimental parameters and the average elution time associated with each compound. There were only two 76 ::t:4 ::t:3

PAGE 92

trials that were taken due to inconsistencies in the voltage and current readings before and after the two separations. In terms of reproducibility, the two trials are consistent with one another based on the relative standard deviations. Figures 26 and 27 show the electrophorogram associated with triall and trial 2. To expand on this work, additional research was done with the microfluidic device discussed in the introduction section. Table 23. Elution of the four compounds contained in the complex mixture from an unmodified 50 um i.d., 12 em length capillary. ..!!! E: .E: E: 0 .. "' 0 E: ... 110 cu ;: E: "'C 1-... 0 ..:! cu 0 > ;(" .... ::I iii ... iii w iii ... ::?! ::I 0 w .:!!. ..:! .:!!. .:!!. > -w ... Eiii ... > cu .. cu > "' > > .!!! .. 1111 E: .D 1-cu 1cu cu ;: "' cu E cu "'C "'C cu "'C a.-e "' .. ... ai ai ai 0 ... E "'C "'C E "'C "' u ::I ::I .. > .. .. U-a. u z Ill c( Ill Ill 12 (+) 0.5 6 2 348 :1::2 365 :1::3 378 :1::4 *Tm abbreviation for tyramine *Ts -abbreviation for tyrosine *HTm-abbreviation for 3-hydroxytyramine HCI (dopamine) *EN -abbreviation for epinephrine 77 E: 0 ... ::I w Ziii wcu ai E 462 iii .:!!. > cu "'C "'C .. Ill :1:19

PAGE 93

Absorbance vs Time Complex Mix (12 em) 0.03 0.02 0.01 c: ra -e 0 0 Ill ..c < -0.01 400 500 -0.02 -------0.03 Elapsed Time (s) Figure 26. Electropherogram of trial 1 depicting the elution of four compounds (in order of elution: tyramine, D-tyrosine, 3-hydroxytyramine hydrochloride, and epinephrine) from an unmodified 12 em length, 50 um i.d. capillary. Absorbance vs Time-Complex Mix (12 em) 1 I j 0.06 ------------0.05 Q,j 1.1 0.04 c: ra ..c 0.03 '"' 0 Ill ..c 0.02 < 0.01 0 0 100 200 300 400 500 Elapsed Time (s) Figure 27. Electropherogram of trial 2 depicting the elution of four compounds (in order of elution: tyramine, D-tyrosine, 3-hydroxytyramine hydrochloride, and epinephrine) from an unmodified 12 em length, 50 um i.d. capillary. 78 I

PAGE 94

This data confirms that separation of a complex mixture can be reliably accomplished with a 12 em length capillary by appropriately setting the experimental parameters. The information within this section provides useful information about what separations should be expected for the microfluidic device. Microfluidic Device A pure sample of hydroquinone was introduced to the microfluidic device having the following dimensions: 5.1 em length x 2.6 em width. The microfluidic channel length is 3.9 em, which is considerably less than the 12 em length capillaries discussed in the previous sections. The elution time associated with hydroquinone is 79 3 sec. Because the standard deviation is low relative to the average elution time and the voltage and current readings remained constant from run to run, a mixture containing tyramine and ascorbic acid was introduced to the microfluidic device, and absorbances were monitored at 280and 265-nm, respectively. Table 24 summarizes the elution times and experimental parameters associated with the pure hydroquinone sample, and table 25 summarizes those for the tyramine/ascorbic acid mixture. Because there was some variability in the tyramine and ascorbic acid elution times, the electropherograms associated with the three separations 79

PAGE 95

are depicted in figures 28, 29 and 30. It was expected that the elution times would differ, but the time differences between the tyramine and ascorbic acid peaks would be similar. Table 25 indicates that as the elution times increase, there is an observed increase in the time that separates each peak. While this was not the predicted outcome, it does make sense that the time interval between two peaks will increase as the retention times increase. The applied potential will have a longer time to influence the electrophoretic mobility of ions in a longer length capillary versus a shorter length capillary, and therefore more separation will be observed. This point is clearly illustrated when comparing, for example, the elution times of the 50 em length capillaries to the elution times of the 25 em length capillaries. Table 24. Experimental parameters and average hydroquinone elution time from the microfluidic device. Microfluidic Channel Polarity Voltage Current Number of Ave. HQ Std Length (em) Mode (kV) (uA) Trials Elution Time dev (s) (s) 3.9 (+) 0.5 30 5 79 *HQ-abbreviation for hydroquinone 80

PAGE 96

Table 25. Experimental parameters and elution times associated with the tyramine/ascorbic acid mixture from the microfluidic device. Microfluidic Polarity Voltage Current Trial Tm AA Diff. Channel Mode (kV) (uA) Number Elution Elution Between Length (em) Time (s) Time (s) Peaks (s) 3.9 (+) 0.5 30 1 75 83 8 3.9 (+) 0.5 30 2 95 119 24 3.9 (+) 0.5 30 3 105 145 40 *Tm -abbreviation for tyramine *AA-abbreviation for ascorbic acid Absorbance vs Time-Microfluidic 0.02 ....-----------------------0.01 ., u r:: AI Jl 0 ... 0 Ill 200 Jl <( -0.01 -0.02 ..1---------------------Time(s) Figure 28. Trial 1: elution of tyramine and ascorbic acid from the microfluidic device. 81

PAGE 97

Absorbance vs TimeMicrofluidic 0.02 -r-----------------------0.03 ......._ ____________________ Time(s) Figure 29. Trial 2: elution of tyramine and ascorbic acid from the microfluidic device. Absorbance vs TimeMicrofluidic 0.03 -r---------------------u 0.01 II ...D ... 0 0 c:( 300 -0.01 +-----------------------0.02 ......_ _____________________ Time (s) Figure 30. Trial 3: elution of tyramine and ascorbic acid from the microfluidic device. 82

PAGE 98

This data confirms that separation of a simple mixture can be accomplished reliably using the microfluidic device. Although baseline fluctuations were clearly observed, this section provides insight about the necessary conditions or modifications that could be made to accomplish separations of more complex mixtures. 83

PAGE 99

CHAPTER 4: CONCLUSIONS The research conducted in this work demonstrates the importance of evaluating experimental parameters for moderate and small length capillaries so that separation conditions may be optimized for the use of microfluidic devices. By establishing a method for setting voltage and current, reproducible data was obtained from unmodified 50 um i.d. capillaries having lengths of 60-, 50-, 40-, 25-, and 12-cm. The separation efficiency and resolution were evaluated for unmodified capillaries having lengths of 50-, 25-, and 12-cm. To expand on this work, capillaries of various lengths were modified with varying concentrations of PSS, and the separation efficiency and resolution were evaluated. It was found that the hydroquinone elution time, efficiency, and resolution remained relatively constant for capillaries of the same length that were modified with different concentrations of PSS. Because the elution times, efficiency, and resolution are not dramatically influenced by the concentration of PSS used for modification, low concentration polymer solutions may be used to limit waste generation. Additionally, since there is not a dramatic loss in efficiency and resolution associated with the 12 em 84

PAGE 100

length capillary versus the 25-, 40-, 50-, or 60-cm length capillaries, the data obtained from the 12 em length capillary is reliable. To determine whether separation of closely eluting compounds could be accomplished with small length capillaries, the complex mixture was introduced to 40and 12-cm length capillaries. Because the four compounds were reliably detected from both the 40and 12-cm length capillaries, the microfluidic device was tested to determine whether separations could be accomplished. Separation of tyramine and ascorbic acid was achieved using the microfluidic device. Given the reasonable separation efficiency and resolution associated with the separation, the microfluidic device can be used to perform separations of simple mixtures. Separation of more complex mixtures may also be accomplished, although this would likely require modification of the buffer solution pH or adjustment in the voltage and current settings (e.g. decrease the potential to a value closer to 0 kV). Future Applications Microfluidic devices utilize capillary electrophoresis to perform rapid separations on picoliter sample volumes. Because of the reasonable separation efficiency and resolution, there is a strong possibility that microfluidic devices will be used to perform diagnostic tests in countries that do not have access to expensive medical equipment. The devices would 85

PAGE 101

ideally be inexpensive to make and disposable. This may however be a challenge with the current manufacturing considerations.3 The microfluidic devices would need to be manufactured with high accuracy and precision, be engineered so that there are minimal chemical/physical effects, and have reasonable signal-to-noise ratios. As a direct extension of this work, it may be useful to test the microfluidic device using a peristaltic pump instead of the high voltage power supply. Random and frequent baseline fluctuations were associated with the separation of the tyramine/ascorbic acid mixture, so it would be expected that pressure-driven flow through the channel would result in less baseline variability. The downside is that separation of multiple analytes would not be possible without electrophoresis. As a comparison to the CZE separations, the peak efficiency and resolution could be calculated from the introduction of a single component. Because the flow of the pressure-driven system would be laminar, this would lead to more solute-zone dispersion. By adjusting the injection time to minimize peak broadening, the efficiency and resolution may be increased without drastically reducing peak absorbance. Although the primary focus of this paper was the use of the microfluidic device, it may be useful to further investigate capillary modification by 1) expanding the range of PSS concentrations from 0.001 M to 0.02 M, and 2) varying the introduction period from 3 min to 15 min. To 86

PAGE 102

begin, the lowest concentration (0.001 M) and highest concentration (0.02 M) PSS solutions could be introduced to capillaries in time intervals of 3-, 6-, 9-, 12-, and 15-min. Without observing a definitive trend in the hydroquinone elution time relative to the introduction time of 0.02 M PSS solution, it might be useful to investigate whether there are trends associated with the inroduction of lower concentration PSS solutions (e.g from 0.001 M to 0.005 M). It is plausible that at higher PSS concentrations the capillary wall is fully saturated after the 3-m in introduction. Depending on the experimental application, the concentration of the PSS solution or introduction time could then (theoretically) be adjusted to produce the desired result in the electroosmotic flow rate. 87

PAGE 103

BIBLIOGRAPHY 1. Westermeier, R.; Gronau, S. Electrophoresis in Practice, 3rd ed.; Wiley, 2005; p 9. 2. Butler, J. M. Houston DNA Training Workshop. [Online] April 3, 2007, p 2. http:/ jwww.cstl.nist.gov /biotechjstrbasejtraining.htm. 3. The Industrial Physicist. A New Wave of Microfluidic Devices. http:/ jwww.aip.org/tip/INPHFA/vol-9 jiss-4jp14.html (accessed Jan 15, 2010) 4. Cornell University College of Enginerring Kirby Research Group. Kirby Lab Microfluidic/Nanofluidics: Microfluidics. http:/ jwww.kirbyresearch.comjindex.cfmjpagejrijufluids.htmhttp:/ I www.kirbyresearch.comjindex.cfmjpagejrijufluids.htm (accessed Jan 15, 2010) 5. Chankvetadze, B. Capillary Electrophoresis in Chiral Analysis, 1st ed.; Wiley, 1997; pp 40-61. 6. Cazes, J. Encyclopedia of Chromatography, 2nd ed.; CRC Press: New York, 2005; Vol. 1; pp 281-282. 7. Skoog, D.A.; Holler, F.J.; Crouch, S.R. Principles of Instrumental Analysis, 6th ed.; Thomson Brooks/Cole Publishing: Belmont, 2007; pp 867-873. 88

PAGE 104

8. Frazier, R. A.; Ames, J.M.; Nursten, H.E. Background Theory and Principles of Capillary Electrophoresis, 1st ed.; RSC Publishing, 2000; pp 1-9. 9. Chankvetadze, B.; Blaschke, G. Enantioseparations in Capillary Electromigration Techniques: Recent Developments and Future Trends. journal of Chromatography A. 2001, 906, 309-363. 10. Landers, J.P. Handbook of Capillary Electrophoresis, znd ed.; CRC Press: New York, 1997; p 9. 11. Altria, K.D. Capillary Electrophoresis Guidebook: Principles, Operation, and Applications; Humana Press, 1996; Vol. 52; pp S-6. 12. Chankvetadze, B. Capillary Electrophoresis in Chiral Analysis, 1st ed.; Wiley, 1997; pp 11-12. 13. Target Discovery, inc. Target Discovery: From Omnics to Knowmics. http:/ Jwww.targetdiscovery.com/index.php?topic=pro2.tech.flow (accessed Jan 15, 2010). 14. The Engineering ToolBox. Laminar, Transitional or Turbulent Flow. http:/ jwww.engineeringtoolbox.comjlaminar-transitional-turbulentflow-d_S77.html (accessed Jan 16, 2010). 15. Sheffield Hallam University. Chromatography-Introductory Theory. http:/ jteaching.shu.ac.ukjhwbjchemistry jtutorialsjchromjchroml.ht m (accessed Jan 20, 2010). 89

PAGE 105

16. Wehr, T.; Rodriguez-Diaz, R.; Zhu, M. Capillary Electrophoresis of Proteins; CRC Press: New York, 1999; Vol. 80; pp 9-10. 17. Poole, C.F. The Essence of Chromatography; Elsevier Ltd, 2003; pp 24-26. 18. Sample Capacity. hplc.chem.shu.edu/NEW /HPLC_Book/glossary /df_s.html (accessed Jan 22, 2010). 19. Reference for resolution equation 20. Heftmann, E. Chromatography: Fundamentals and Techniques, 6th ed.; Elsevier Ltd, 2004; Vol. 69A; pp 9-14. 21. Tay, F.E. Microfluidics and BioMEMS Applications; Springer, 2002; p 281. 22. National Aeronautics and Space Administration. Boundary Layer. http:/ jwww.grc.nasa.gov /WWW /K-12/airplanejboundlay.html (accessed Jan 29, 2010). 23. Watson, A.Y.; Bates, R.R.; Kennedy, D. Air Pollution, the Automobile, and Public Health; The National Academies Press, 1988; p 326. 24. HyperPhysics: Georgia State University. Pressure. http:/ /hyperphysics.phy-astr.gsu.edu/Hbase/pturb.html (accessed Feb 2, 2010). 25. Technovelgy.com: Where Science Meets Fiction. Programmable Lab-On A-Chip: Science Fiction in the News. http:/ jwww.technovelgy.com/ct/Science-Fiction News.asp?NewsNum=2268 (accessed Feb 2, 2010). 90

PAGE 106

26. Bhushan, B. Springer Handbook of Nanotechnology, 2nct ed; Springer, 2006; pp 539-540. 27. Ramachandran, G. Lab-on-a-chip-Cutting Edge Technology in the Field of Bioscience. [Online] 2007. http:/ jwww.frost.comjprodfservletfmarket-insighttop.pag?docid=102591934 (accessed Feb 2, 2010). 28. AllExperts. Lab-on-a-chip (LOC). http:/ fen.allexperts.comfe/1/lajlabon-a-chip.htm (accessed Feb 3, 2010). 29. Horvath, J.; Dolnik, V. Polymer Wall Coatings for Capillary Electrophoresis. Electrophoresis. 2001, 22, 644-655. 30. Dissociation Constants of Organic Acids and Bases. http:/ fwww.zirchrom.com/organic.htm (accessed Jan 2, 2010). 31. Hajratwala, B.R. Kinetics of Sulfite-Induced Anaerobic Degradation of Epinephrine. journal of Pharmaceutical Sciences. 2006, 64, 45-48. 32. Tseng, H.; Barrett, D. A. Micellar Electrokinetic Biofluid Analysis of Biogenic Amines Using On-line Sample Concentration and UV Laser Induced Native Fluorescence Detection. journal of Chromatography A. 2009, 1216, 3387-3391. 91

PAGE 107

VITAE I was born in Lakewood, Colorado and attended Lakewood High School, graduating in 2004. In the summer of 2004, I enrolled in a cultural anthropology course at Red Rocks Community College and traveled to Dublin, Ireland. I earned my prerequisites from Red Rocks Community College in 2006, and traveled to Edinburgh, Scotland to complete a humanities course in the summer of 2006. I then transferred to the University of Colorado Denver to study chemistry. I graduated from the University of Colorado Denver in 2008, earning my Bachelor of Science Degree in chemistry. As an undergraduate, I completed a graduate level course in mass spectrometry and began doing research with Professor Mark Anderson in capillary electrophoresis. In the summer of 2008, I was awarded a grant through the National Science Foundation and worked as a resident scientist at the Englewood Leadership Academy. In the summer of 2009, I enrolled in a renewable energy course offered through the University of Colorado Denver, and attended the World Renewable Energy Conference in Bangkok, Thailand. I was a general chemistry I teaching assistant in the fall of 2009, and completed my coursework r_equirements for the graduate program during this semester. Throughout the spring semester of 2010, I worked diligently with Professor Mark Anderson to complete research and compile this document.