repository.arizona.edu · 2020. 4. 2. · THE UNIVERSITY OF ARIZONA ® GRADUATE COLLEGE As members of the Final Examination Ccnnnittee, we certify that we have read the dissertation

STRUCTURE-FUNCTION RELATIONSHIPS IN CHROMATOGRAPHIC

STATIONARY PHASES; CHARACTERIZATION OF EXISTING PHASES AND

IMPROVED STRATEGIES FOR MATERIALS FABRICATION

by

Christopher Jay Orendorff

A Dissertation submitted to the Faculty of the

DEPARTMENT OF CHEMISTRY

In partial fulfillment of the requirements For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

2 0 0 3

UMI Number: 3108940

INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy

submitted. Broken or indistinct print, colored or poor quality illustrations and

photographs, print bleed-through, substandard margins, and improper

alignment can adversely affect reproduction.

In the unlikely event that the author did not send a complete manuscript

and there are missing pages, these will be noted. Also, if unauthorized

copyright material had to be removed, a note will indicate the deletion.

UMI UMI Microform 3108940

Copyright 2004 by ProQuest Information and Learning Company.

All rights reserved. This microform edition is protected against

unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company 300 North Zeeb Road

P.O. Box 1346 Ann Arbor, Ml 48106-1346

THE UNIVERSITY OF ARIZONA ® GRADUATE COLLEGE

As members of the Final Examination Ccnnnittee, we certify that we have

read the dissertation prepared by Christopher Jay Orendorff

entitled Structure - Function Relationships in Chromatographic

Stationary Phases: Characterization of Existing Phases

and Improved Strategies for Materials Fabrication

and recomicend that it fae accepted as fulfilling the dissertation

requirement for the Degree of Doctor of Philosophy

Dat Z-z 63

Dr. Nâl R. Arm^r

q "2-

Dr. S. Scott Saavedra

Dr. Robert Bates

Dr. Indraneel Ghosh

Dats

Datfe ~

Dat/ ' f

Date/^ ^

Final approval and acceptance of this dissertation is contingent upon

the candidate's submission of the final copy of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

Disser

I I Date

lo 03

3

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation arc allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED;

4

ACKNOWLEDGMENTS

I would like to start by thanking my family. My parents, Connie and John Orendorff, unselfishly gave me every opportunity to succeed while growing up. For that 1 will always be grateful. I would like to thank my sister, Steph, my brother-in-law, Steve, and my grandmother, Bernice Brighton. I also want to thank Judi for making me so very happy and for sticking with me for the past several years.

1 would also like to thank the people I have had the pleasure of working with over the past few years. Dr. Jeanne Pemberton allowed me to work in her research group and put me in a position to be successful. I greatly appreciate her patience and support over the past few years. Thanks Jeanne. Dr. Mike Ducey helped me get started with my research and Dr. Dom Tiani, Dr. Adam Hawkridge, Dr. Chris Zangmeister, Dr. Joey Robertson, and Zhaohui Liao have been instrumental in helping me along the way both scientifically and personally. Kim Menezes has been invaluable throughout this entire process. Thanks Kim.

To all of my friends I've met in Tucson, here's one more knuckle bash from me. Thanks for everything. It has been a rockin' good time.

To Mom, Dad, and my Grandfathers,

Bob OrendortT

and

Leland Brighton

6

TABLE OF CONTENTS

LIST OF FIGURES 16

LIST OF TABLES 50

ABSTRACT 53

CHAFER 1: TNTRODUCTION TO ALKYLSRANE-BASED STATIONARY PHASES FOR REVERSED-PHASE LIQUID CHROMATOGRAPHY; INTERFACIAL IMPORTANCE AND RETENTION MECHANISMS 56

Alkylsilane-Based Stationary Phases in Reversed-Phase Liquid Chromatography 56

Retention Models 58 Solvophobic Theory 59 Partitioning Model 60

Importance of Alkylsilane Structure 61

Goals of this Research 66

CHAPTER 2; EXPERIMENTAL 68

Instrumentation 68 Raman Spectroscopic Instrumentation 68 High Performance Liquid

Chromatography Instrumentation 71 Scanning Electron Microscopy 73 Elemental Analysis 73

Materials 73 Isolute SPE phases 75 Kromasil Silicas and Bulk Ci8 Stationary Phases 78 NIST Ci8 Stationary Phases 84

Methodology 88

7

TABLE OF CONTENTS - Continued

HPLC Column Packing 88 Surface Coverage Determination 90 RPLC Column Selectivity Determination 91 Spectral Analysis 94

Molecular Pictures 96

CHAPTER 3; QUANTITATIVE CORRELATION OF RAMAN SPECTRAL INDICATORS IN DETERMINING CONFORMATIONAL ORDER IN ALKYL CHAINS 99

Introduction 99

Goals of This Work 103

Methodology 104

General Overview of Alkane Conformational Order for Octadecane and Polyethylene from Raman Spectroscopy 105

Raman Spectroscopy of Octadecane 115

Raman Spectroscopy of Polyethylene 128

Comparison of Common Raman and IR-Active Indicators of Conformational Order 142

Summary 146

Conclusions and Relevance to Chromatographic Studies 149

Future Directions 150

Appendix 3.1 150

8


CHAPTER 4; SOLVENT EFFECTS ON THE CONFORMATIONAL ORDER OF COMMERCIAL SOLID PHASE EXTRACTION ALKYLSILANE STATIONARY PHASES 160

Introduction 160


Methodology 162

Review of Previous Studies of Isolute C18MF SPE Phases 163

Raman Spectroscopy of Isolute C18MF in Solvent 164

Effect of Solvent on Alkylsilane Order for Isolute C18MF 169

Conclusions and Future Directions 173

CHAPTER 5: STRUCTURE-FUNCTION RELATIONSHIPS IN OCTADECYLSILANE STATIONARY PHASES BY RAMAN SPECTROSCOPY: EFFECTS OF TEMPERATURE, SURFACE COVERAGE, AND PREPARATION PROCEDURE 175

Introduction 175


Methodology 180

Raman Spectroscopy ofNIST Cig Stationary Phases 182

Effect of Temperature on the Conformational Order of Alklysilane Stationary Phases 186

First-Order Phase Transition Treatment of Alkylsilane Stationary Phases 192

9



Appendix 5 .1 210

CHAPTER 6; STRUCTURE-FUNCTION RELATIONSHIPS IN OCTADECYLSILANE STATIONARY PHASES BY RAMAN SPECTROSCOPY: EFFECT OF COMMON MOBILE PHASE SOLVENTS 214

Introduction 214


Methodology 219

Raman Spectroscopy of High-Density Alkylsilane Stationary Phases in Solvent 220

Effect of Solvent on Conformational Order 226

Effect of Temperature on Conformation Order in the Presence of Solvent 234

Subtle Differences in Temperature Effects Between Differently Solvated Stationary Phases 241

Peak Frequencies of v(C-H) Modes as Indicators of Solvent Effects on Conformational Order 245


Appendix 6.1 250

CHAPTER 7; STRUCTURE-FUNCTION RELATIONSHIPS IN OCTADECYLSILANE STATIONARY PHASES BY RAMAN SPECTROSCOPY: EFFECTS OF SELF-ASSOCIATING SOLVENTS 278

Introduction 278

10



Methodology 280


Review of Solvent Effects on High-Density Stationary Phase Conformational Order 284

Effect of Self-Associating Solvent 287

Effect of Temperature in Self-Associating Solvents 300

Conclusions 308

Appendix 7.1 309

CHAPTER 8: STRUCTURE-FUNCTION RELATIONSHIPS IN OCTADECYLSILANE STATIONARY PHASES BY RAMAN SPECTROSCOPY: EFFECTS OF NEUTRAL AND BASIC AROMATIC COMPOUNDS 319

Introduction 3 19


Methodology 321


Effect of Aromatic Compounds on the Order of These High-Density Stationary Phases 325

Effects of Temperature 340

Conclusions 347

11


Appendix 8.1 348

CHAPTER 9: PRELIMINARY INVESTIGATIONS OF SALTING-OUT AND pH EFFECTS ON THE CONFORMATIONAL ORDER OF OCTADECYLSILANE STATIONARY PHASES BY RAMAN SPECTROSCOPY 361

Introduction 361


Methodology 363

Salting-Out Alkylsilane Stationary Phases 364

Effects of pH of Aqueous Acidic and Basic Aromatic Compounds 371

Benzoic Acid 371 Aniline 377

Benzylamine 379


CHAPTER 10: SOLUTE, SOLVENT, AND STATIONARY PHASE INTERACTIONS OF REVERSED-PHASE LIQUID CHROMATOGRAPHIC SYSTEMS DETERMINED BY RAMAN SPECTROSCOPY 385

Introduction 385

Goals of This Research 386

Methodology 387

Thermodynamic Description of Retention inRPLC 388

Kinetics and Band Broadening in RPLC 390

12


Separation of Monosubstitutecl Aromatic Compounds using MFC18 393

Raman Spectroscopy of MFC 18 in the Presence of Solute and Mobile Phase 405

Aniline Interactions 412 Anisole Interactions 414 Toluene Interactions 416

Molecular Pictures of the MFC) 8 (Chromatographic Interface 419

Separation of Monosubstituted Aromatic Compounds using TFC18SF 428

Raman Spectroscopy of TFC18SF in the Presence of Solute and Mobile Phase 437

Effects of Solutes 437

Molecular Pictures of the TFC18SF Chromatographic Interface 446

Concentration Effects 451 Toluene in Methanol 451 T oluene in Acetonitrile 454

Conclusions 458

Future Directions 460

CHAPTER 11: SELECTIVITY AND STRUCTURAL PROPERTIES OF ANION AND CATION EXCHANGE CHROMATOGRAPHY SYSTEMS DETERMINED BY RAMAN SPEC TROSCOPY 462

Introduction 462


Methodology 468

13


Isolute SAX 471 Raman Spectroscopy of Isolute SAX 471 Raman Spectroscopy of Isolute SAX

in Aqueous Sodium Halide and Oxyanion Solutions 472

Raman Spectroscopy of Model Tetramethylammonium Salts and Aqueous Soluitions 496

Isolute SCX-2 503 Raman Spectroscopy of Isolute SCX-2 503 Alkali-metal Interactions with

SCX-2 507 Hydration Effects 510 Counter Ion Effects 511 Raman Spectroscopy of Model

Methanesulfonic Acid Solutions 511 Tetraalkylammonium Interactions

with SCX-2 515

Comparison of SAX and SCX-2 Raman Studies 528


CHAPTER 12: PRELIMARY STUDIES ON THE CHARACTERIZATION OF OTHER COMMERCIAL STATIONARY PHASES FOR REVERSED-PHASE LIQUID CHROMATOGRAPHY 532

Introduction 532


Methodology 534

Raman Spectroscopy of Isolute CN SPE phases 536

Interactions Between Solvent and Isolute CN 536

Raman Spectroscopy of Isolute Phenyl SPE Phases 546

14


Interactions Between Solvent and Isolute Phenyl 549


CHAPTER 13; ALKYLSILANE-BASED STATIONARY PHASES VIA CONTROLLED SOLUTION MODIFIER TECHNIQUES: SYNTHETIC APPROACH, CHARACTERIZATION, AND CHROMATOGRAPHIC PERFORMANCE 554

Introduction 554


Methodology 556 Stationary' Phase Synthesis 556 Characterization of Stationary Phases 559 Silica Substrates 560 RPLC Column Preparation and

Performance 561

Physical Characterization of Kromasil Cig 565

Effects of Substrate 566

Modifier Approach to Controlling Surface Coverage 5 75

Geometric Shape Selectivity of High Surface Coverage Phases 592


CHAPTER 14; CONCLUSIONS AND FUTURE DIRECTIONS 602

Overview of the Problem 602 Understanding Retention Mechanism 602 Stationary Phase Fabrication 602

Summary of Research 603

15


Future Studies 606

REFERENCES 611

16

LIST OF FIGURES

FIGURE 1.1 Schematic of the hydrolysis and condensation of n-alkylchlorosilanes on silica substrates 57

FIGURE 2.1 Schematic of Raman spectroscopic instrumentation 69

FIGURE 2.2 Schematic of the BAS 200B liquid chromatograph 72

FIGURE 2.3 Dimethyloctadecylchlorosilane and a condensed monolayer of the monofunctional silane 77

FIGURE 2.4 Cyanopropyltrichlorosilane and a condensed monolayer of cyanopropyltrichlorosilane on the 1ST CN surface 79

FIGURE 2.5 Phenyltrichlorosilane and a condensed monolayer of phenyltrichlorosilane on the 1ST Phenyl surface 80

FIGURE 2.6 Trichlorosilylpropyltrimethylamomium chloride and a simple illustration of the SAX surface 81

FIGURE 2.7 Trichlorosilylpropyl-sulfonic acid and a simple illustration of the SCX-2 surface 82

FIGURE 2.8 Octadecyltrichlorosilane and a schematic of the condensed trifimctional silane monolayer 86

FIGURE 2.9 Methyloctadecyldichlorosilane and a schematic of the condensed difunctional monolayer 87

FIGURE 2.10 Illustration of the RPLC column packer and a cut-away view of the reducing union fitted with a porous frit 89

FIGURE 2.11 Chemical structures and geometric shapes of PAHs used to determine RPLC column shape selectivity 92

FIGURE 2.12 Chromatograms of PAHs used to determine shape selectivity adapted from reference 2.8 93

17 LIST OF FIGURES - Continued

FIGURE 2.13 (a) calibrated Raman spectra of DFC18SL, (b) illustration of the 5-point fit used to baseline correct Raman spectra in the v(C-H) region, and (c) background corrected Raman spectra of DFC18SL in anisole at 15 "C 95

FIGURE 3.1 Raman spectra in the v(C-C), 8(C-H) and v(C-H) regions for octadecane at (a) 0 °C and (b) 60 °C. Integration time 2 min 106

FIGURE 3.2 Schematic of alkyl chain trans and gauche conformers 109

FIGURE 3.3 Schematic of a subtle alkyl chain rotation 110

FIGURE 3.4 Schematic of alkyl chain decoupling Ill

FIGURE 3.5 Raman spectra in the v(C-CX 6(C-H) and v(C-H) regions for polyethylene at (a) 0 °C, (b) 74 "C and (c) 100 °C. Integration time 2 min 114

FIGURE 3.6 Intensity ratio of the V(C-C)G to the V(C-C)T as a function of the intensity ratio of the VaCCHz) to the VS(CH2) for octadecane. Error bars are included for each point, but may be smaller than the size of the points 116

FIGURE 3.7 Intensity ratio of the Vs(CH3)fr to the Vs(CH2) as a function of the intensity ratio of the Va(CH2) to the VsCCHa) for octadecane 118

FIGURE 3.8 Molar volume (mL/moIe) of octadecane as a function of the intensity ratio of the Va(CH2) to the VS(CH2) 119

FIGURE 3.9 VS(CH2) frequency as a function of the intensity ratio of the Va(CH2) to the VsCCHi) for octadecane 120

18

LIST OF FIGURES - Continued

FIGURE 3.10 Va(CH2) frequency as a function of the intensity ratio of the VafCHi) to the VsCCHj) for octadecane 121

FIGURE 3.11 x(CH2) frequency as a function of the intensity ratio of the Va(CH2) to the Vs(CH2) for octadecane 122

FIGURE 3.12 t(CH2) asymmetry factor as a function of the intensity ratio of the Va(CH2) to the VsCCHa) for octadecane 125

FIGURE 3.13 t(CH2) full-width half maximum (FWHM) as a fiinction of the intensity ratio of the Va(CH2)

to the VsCCHi) for octadecane 126

FIGURE 3.14 x(CH2) FWHM as a function of the intensity ratio of the V(C-C)G to the V(C-C)T for octadecane 127

FIGURE 3.15 The peak area ratio of the [5(CH2) + 5(CH3)] to the 5(CH2) as a function of the intensity ratio of the Va{CH2) to the VsCCHi) for octadecane 129

FIGURE 3.16 Representative peak fits to the Raman spectra of octadecane in the 5(CH-H) region at (a) 0 °C and (b) 60 "C 130

FIGURE 3.17 The intensity ratio of the V(C-C)G to the v(C-C)R as a function of the intensity ratio of the Va(CH2) to the VS(CH2) for polyethylene 132

FIGURE 3.18 VS(CH2) frequency as a function of the intensity ratio of the Va(CH2) to the Vs(CH2) for polyethylene 133

FIGURE 3.19 Va(CH2) frequency as a function of the intensity ratio of the Va(CH2) to the Vs(CH2) for polyethylene 135

19

LIST OF TABLES - Continued

FIGURE 3.20 t(CH2) frequency as a function of the intensity ratio of the Va(CH2) to the VsfCHi) for polyethylene 136

FIGURE 3.21 Intensity ratio of the VS(CH3)FR to the VS(CH2) as a function of the intensity ratio of the Va(CH2) to the Vs(CH2) for polyethylene 137

FIGURE 3.22 t(CH2) asymmetry factor as a function of the intensity ratio of the VaCCHj) to the VS(CH2) for polyethylene 138

FIGURE 3.23 t(CH2) FWHM as a flinction of the intensity ratio of the Va(CH2) to the Vs(CH2) for polyethylene 139

FIGURE 3.24 t(CH2) FWHM as a function of the intensity ratio of the V(C-C)G to the V(C-C)T for polyethylene 141

FIGURE 3.25 The peak area ratio of the [6(CH3) + 5(CH2)] to the 5(CH2)ORTHO + 6(CH2) as a function of the intensity ratio of the Va(CH2) to the VS(CH2) for polyethylene 143

FIGURE 3.26 The peak area ratio of the 5(CH2)ORTHO to the {[5(CH2) + 5(CH3)] + S(CH2)} as a function of the intensity ratio of the Va(CH2) to the VS(CH2) for polyethylene 144

FIGURE 3.27 Raman spectra of the deformation modes (1390-1510 cm"') peak fit using a Gaussian distribution at (a) 0 °C and (b) 89 "C 145

FIGURE A3.1 Raman spectra in the v(C-C), 5(C-H) and v(C-H) regions for octadecane at (a) -15 "C, (b) -10 "C, and (c) 10 °C. Integration time 2 min 151

20


FIGURE A3.2 Raman spectra in the v(C-C), 6(C-H) and v(C-H) regions for octadecane at (a) 20 "C and (b and c) through the phases transition. Integration time 2 min 152

FIGURE A3 .3 Raman spectra in the v(C-C), 5(C-H) and v(C-H) regions for octadecane (a, b, and c) through the phases transition. Integration time 2 min 153

FIGURE A3.4 Raman spectra in the v(C-C), 8(C-H) and v(C-H) regions for octadecane (a) through the phases transition and at (b) 30 "C and (c) 40 "C. Integration time 2 min 154

FIGURE A3.5 Raman spectra in the v(C-C), 6(C-H) and v(C-H) regions for octadecane at (a) 50 "C, (b) 70 "C, and (c) 80 "C. Integration time 2 min 155

FIGURE A3.6 Raman spectra in the v(C-C), 6(C-H) and v(C-H) regions for octadecane at (a) 90 "C and (b) 100 "C. Integration time 2 min 156

FIGURE A3.7 Raman spectra in the v(C-C), 5(C-H) and v(C-H) regions for polyethylene at (a) 20 "C, (b) 43 °C and (c) 81 "C. Integration time 2 min 157

FIGURE A3.8 Raman spectra in the v(C-C), 6(C-H) and v(C-H) regions for polyethylene at (a) 85 °C, (b) 89 "C and (c) 91 "C. Integration time 2 min 158

FIGURE A3.9 Raman spectra in the v(C-C), 6(C-H) and v(C-H) regions for polyethylene at (a) 94 "C, (b) 106 "C and (c) 150 "C. Integration time 2 min 159

21


FIGURE 4.1 Schematic of the fractal pore structure of I solute silica substrates showing the exclusion of octadecylsilane from micropores (left), giving overall less available surface area for octadecylsilane modification relative to propyl si lane modification (right) 165

FIGURE 4.2 Raman spectra in the v(C-H) region for C18MF in (a) air, (b) methanol, (c) acetonitrile, (d) water, (e) isopropanol, (f) acetone, (g) chloroform, (h) benzene, and (i) toluene. Integration times for each spectra are (a) 40 sec, (b) 1 min, (c) 40 sec, (d) 50 sec, (e) 30 sec, (t) 40 sec, (g) 30 sec, (h) 25 sec, and (i) 30 sec .167

FIGURE 4.3 Raman spectra in the v(C-H) region for C18MF in (a) air, (b) methanol, (c) acetonitrile, (d) water, (e) isopropanol, (f) acetone, (g) chloroform, (h) benzene, and (i) toluene .171

FIGURE 5,1 Raman spectra in the v(C-H) region of TFC18 SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in air at (a) 263 K and (b) 333 K. Integration times for all spectra are 10-min. .183

FIGURE 5.2 I[Va(CH2)]/I[Vs(CH2)] as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC18 (•). The error bars represent one standard deviation .187

FIGURE 5 .3 Schematic of phase diagrams for a homologous series of materials with increasing surface coverage, F. .194

FIGURE 5 .4 R as a function of temperature for TFC18SF (•), TFCIBSL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•). The error bars represent one standard deviation. Best fits for R = a In T + b shown as solid lines. ,196

22


FIGURE 5.5 (a) Clapeyron plots of dR/dT from best fit lines in Figure 5 .4 as a function of 1/T. TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•). (b) Slope of Clapeyron plots as a function of alkylsilane surface coverage 198

FIGURE 5.6 I[Va(CH2)]/I[Vs(CH2)] as a function of surface coverage at 258 (•), 293 (•), and 333 K (•) 201

FIGURE 5.7 VsCCHi) frequency as a function of temperature for (a) TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•). The error bars represent one standard deviation 202

FIGURE 5.8 Proposed temperature-dependent disordering process for TFC18SF 206

FIGURE 5.9 Proposed temperature-dependent disordering process for MFC 18 207

FIGURE 5.10 Proposed temperature-dependent disordering process for DFC18SF 208

FIGURE A5.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in air at (a) 258 K, (b) 273 K, and (c) 283 K. Integration times for all spectra are 10-min 211

FIGURE A5.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in air at (a) 293 K, (b) 303 K, and (c) 313 K, Integration times for all spectra are 10-min 212

FIGURE A5.3 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in air at (a) 323 K and (b) 343 K. Integration times for all spectra are 10-min 213


FIGURE 6.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in (a) methanol, (b) acetonitrile, and (c) water at 20 "C, Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min

FIGURE 6.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in (a) acetone, (b) hexane, and (c) THF at 20 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-niin, for MFC 18 in all solvents are 5-min

FIGURE 6.3 Raman spectra in the •(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in (a) benzene, (b) toluene, and (c) chloroform at 20 °C Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min

FIGURE 6.4 I[Va(CH2)]/l[Vs(CH2)] as a function of solvent solvatochromic parameter, TZ*, for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•) in the polar solvents methanol, acetonitrile, and water at 20 "C. The error bars represent one standard deviation

FIGURE 6.5 I[Va(CH2)]/I[Vs(CH2)] as a function of solvent solvatochromic parameter, n*, for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•) in the nonpolar solvents hexane, toluene, chloroform, THF, benzene, and acetone at 20 °C. (a) All solvents and (b) expanded view of the region containing data for toluene, chloroform, THF, benzene, and acetone. The error bars represent one standard deviation

24


FIGURE 6.6 I[Va(CH3)]/I[Vs(CH2)] as a function of temperature for TFC18SF in (a) polar and (b) nonpolar solvents, methanol (•), water (•), acetonitrile (O), toluene {•), benzene (A), hexane (•), chloroform (•), acetone (•), and THF (^). The error bars represent one standard deviation 235

FIGURE 6.7 l[Va(CH2)]/l[Vs(CH2)] as a ftinction of temperature for MFC 18 in (a) polar and (b) nonpolar solvents, methanol (•), water (•), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•), acetone (•), and THF (•^). The error bars represent one standard deviation 236

FI GURE 6 .8 I[Va(CH2)]/I[Vs(CH2)] as a function of temperature for TFC I8SL in (a) polar and (b) nonpolar solvents, methanol (•), water (•), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•). acetone (•), and THF (•). The error bars represent one standard deviation 238

FIGURE 6.9 I[Va(CH2)]/I[Vs(CH2)] as a function of temperature for DFC18SF in (a) polar and (b) nonpolar solvents, methanol (•), water (•), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•), acetone (•), and THF (^). The error bars represent one standard deviation 239

FIGURE 6.10 I[Va(CH2)]/I[Vs(CH2)] as a function of temperature for DFC18SL in (a) polar and (b) nonpolar solvents, methanol (•), water (•), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•), acetone (•), and THF (*). The error bars represent one standard deviation 240


25

FIGURE6.il

FIGURE 6 12

FIGURE 6 13

FIGURE 6.14

FIGURE A6.1

FIGURE A6.2

I [ Va( C H2)]/1 [ VsCCHi) ] as a fiinction of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•). DFC18SL (Oi and MFC 18 (•) in (a) methanol and (b) acetonitrile. The error bars represent one standard deviation 242

I[Va(CH2)]/I[Vs(CH2)] as a tunction of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•) in (a) chloroform and (b) toluene. The error bars represent one standard deviation 243

VS(CH2) frequency as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•) in (a) methanol and (b) acetonitrile. The error bars represent one standard deviation 246

VS(CH2) frequency as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (OX and MFC 18 (•) in (a) chloroform and (b) toluene. The error bars represent one standard deviation 247

Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in methanol at (a) -15 °C, (b) -10 °C, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC18 are 5-min 251

Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in methanol at (a) 10 °C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 252

26


FIGURE A6.3 Raman spectra in the v(C-H) region of TFCl8SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in methanol at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF, TFCl SSL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 253

FIGURE A6.4 Raman spectra in the v(C-H) region of TFC18SF, TFCl SSL, DFC18SF, DFC18SL, and MFC 18 in acetonitrile at (a) -15 °C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFCl SSL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 254

FIGURE A6.5 Raman spectra in the v(C-H) region of TFC 18SF, TFCl SSL, DFC18SF, DFC18SL, and MFC18 in acetonitrile at (a) 10 "C and (b) 30 °C. Integration times for TFC18SF, TFCl SSL, DFC18SF, and DFCISSL are 2-min, for MFC 1S are 5-min 255

FIGURE A6.6 Raman spectra in the v(C-H) region of TFC 18SF, TFCl SSL, DFCISSF, DFCISSL, and MFC 18 in acetonitrile at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFCISSF, and DFC1 SSL are 2-min, for MFC 18 are 5-min 256

FIGURE A6.7 Raman spectra in the v(C-H) region of TFC 18 SF, TFCl SSL, DFCISSF, DFCISSL, and MFC IS in water at (a) -15 °C, (b) -10 "C, and (c) 0 "C. Integration times for TFCISSF, TFCl SSL, DFCISSF, and DFCISSL are 2-min, for MFC 1S are 5-min 257

FIGURE A6.8 Raman spectra in the v(C-H) region of TFC 18SF, TFCl SSL, DFCISSF, DFCISSL, and MFC IS in water at (a) 10 "C and (b) 30 "C. Integration times for TFCISSF, TFCl SSL, DFCISSF, and DFCISSL in all solvents are 2-min, for MFC 18 in all solvents are 5-min 258

27


FIGURE A6.9 Raman spectra in the v(C-H) region of TFC18 SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in water at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 259

FIGURE A6.10 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetone at (a) -15 "C, (b) -10 "C, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 260

FIGURE A6.11 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFCI8SF, DFC18SL, and MFC 18 in acetone at (a) 10 °C and (b) 30 "C. Integration times for TFC18SF, TFC 1 SSL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 261

FIGURE A6.12 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetone at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 262

FIGURE A6.13 Raman spectra in the v(C-H) region of TFCl8SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in hexane at (a) -15 "C, (b) -10 "C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 263

FIGURE A6.14 Raman spectra in the v(C-H) region of TFC 18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in hexane at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 264

28


FIGURE A6.15 Raman spectra in the v(C-H) region of TFC18 SF, TFC18SL, DFC18SF, DFC I8SL, and MFC 18 in hexane at (a) 40 °C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 265

FIGURE A6.16 Raman spectra in the v(C-H) region of TFC18SF, TFC18SU DFC18SF, DFC18SL, and MFC 18 in chloroform at (a) -15 "C, (b) -10 °C, and (c) 0 ''C. Integration times for TFCISSF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 266

FIGURE A6.17 Raman spectra in the v(C-H) region of TFCISSF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in chloroform at (a) 10 "C and (b) 30 "C. Integration times for TFCISSF, TFCISSL, DFCISSF, and DFC18SL are 2-min, for MFC 18 are 5-min 267

FIGURE A6.18 Raman spectra in the v(C-H) region of TFCISSF, TFCISSL, DFCISSF, DFC18SL, and MFC IS in chloroform at (a) 40 °C and (b) 50 "C. Integration times for TFCISSF, TFCISSL, DFCISSF, and DFC18SL are 2-min, for MFC 18 are 5-min 268

FIGURE A6.19 Raman spectra in the v(C-H) region of TFC18SF, TFCISSL, DFCISSF, DFCISSL, and MFC 18 in benzene at (a) -15 °C, (b) -10 °C, and (c) 0 °C. Integration times for TFCISSF, TFCISSL, DFCISSF, and DFCISSL are 2-min, for MFC IS are 5-min 269

FIGURE A6.20 Raman spectra in the v(C-H) region of TFCISSF, TFCISSL, DFCISSF, DFCISSL, and MFCIS in benzene at (a) 10 "C and (b) 30 T. Integration times for TFCISSF, TFCISSL, DFCISSF, and DFCISSL are 2-min, for MFCIS are 5-min 270

29


FIGURE A6.21 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in benzene at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 271

FIGURE A6.22 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in toluene at (a) -15 "C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC18 are 5-min 272

FIGURE A6.23 Raman spectra in the v(C-H) region of TFC18SF, TFC18SU DFC18SF, DFC18SL, and MFC 18 in toluene at (a) 10 °C and (b) 30 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 273

FIGURE A6.24 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in toluenel at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 274

FIGURE A6.25 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SU and MFC 18 in THF at (a)-15 X, (b)-10 T, and (c) 0 X. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 275

FIGURE A6.26 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in THF at (a) 10 X and (b) 30 X. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 276

30


FIGURE A6.27 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in THF at (a) 40 T and (b) 50 T. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFCI8SL are 2-min, for MFC 18 are 5-min 277

FIGURE 7.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in (a) methanol, (b) acetic acid, and (c) 1-butanol. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min 282

FIGURE 7.2 (a) I[Va(CH2)]/I[Vs(CH2)] and (b) v,(CH2) peak frequency as a function of solvent solvatochromic parameter n* for TFC18SF (•) and DFC18SF (O) in air, methanol, acetonitrile, water, acetic acid, ethanol, and 1-butanol. The error bars represent one standaid deviation 288

FIGURE 7.3 Molecular picture of the interactions between TFC18SF and methanol at the chromatographic interface 291

FIGURE 7.4 Molecular picture of the interactions between TFC18SF and acetonitrile at the chromatographic interface 292

FIGURE 7.5 Molecular picture of the interactions between TFC18SF and water at the chromatographic interface 293

FIGURE 7.6 Molecular picture of the interactions between TFC18SF and ethanol at the chromatographic interface 294

FIGURE 7.7 Molecular picture of the interactions between TFC18SF and acetic acid at the chromatographic interface 295

FIGURE 7.8 Molecular picture of the interactions between TFC18SF and 1 -butanol at the chromatographic interface 296

31


FIGURE 7.9 (a) I[Va(CH2)]/I[Vs(CH2)] and (b) v.s(CH2) frequency as a function of solvent solvatochromic parameter k* for TFC18SL (A), DFC18SL (•), and MFC 18 (V) in air, methanol, acetonitrile, water, acetic acid, ethanol, and 1-butanol. The error bars represent one standard deviation .298

FIGURE 7.10 (a) I[Va(CH2)]/I[Vs(CH2)] and (b) v.CCHj) frequency as a function of temperature for TFC18SF in methanol (•), acetonitrile (O), water acetic acid (V), ethanol (A), and 1 -butanol (•), The error bars represent one standard deviation .301

FIGURE 7.11 (a) l[Va(CH2)]/l[Vs(CH2)] and (b) v.CCHs) frequency as a function of temperature for DFC18SF in methanol (•), acetonitrile (O), water (¥), acetic acid (V), ethanol (A), and 1 -butanol (•). The error bars represent one standard deviation .302

FIGURE 7.12 (a) I[Va(CH2)]/I[Vs(CH2)] and (b) v.CCHj) frequency as a function of temperature for TFC18SL in methanol (•), acetonitrile (O), water (¥), acetic acid (V), ethanol (A), and 1 -butanol (•). The error bars represent one standard deviation .304

FIGURE 7.13 l[Va(CH2)]/I[Vs(CH2)] as a function of temperature for DFC18SL in methanol (•), acetonitrile (O), water (*), acetic acid (V), ethanol (A), and 1-butanol (•). The error bars represent one standard deviation .305

FIGURE 7.14 I[Va(CH2)]/I[vs(CH2)] as a function of temperature for MFC 18 in methanol (•), acetonitrile (O), water (¥), acetic acid (V), ethanol (A), and 1 -butanol (•). The error bars represent one standard deviation .306

32


FIGURE A7.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in ethanol at (a) -15 °C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 310

FIGURE A7.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in ethanol at (a) 10 "C and (b) 30 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 311

FIGURE A7.3 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in ethanol at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 312

FIGURE A7.4 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetic acid at (a) -15 "C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 313

FIGURE A7.5 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetic acid at (a) 10 "C and (b) 30 "C. Integration times tor TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 314

FIGURE A7.6 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetic acid at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 315

33


FIGURE A7.7 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in 1 -butanol at (a) -15 "C, (b) -10 "C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFCISSF, and DFC18SL are 2-min, for MFC 18 are 5-min 316

FIGURE A7.8 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in 1-butanol at (a) 10 "C and (b) 30 T. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 317

FIGURE A7,9 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFCISSF, DFC18SL, and MFC 18 in 1 -butanol at (a) 40 °C and (b) 50 T. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 318

FIGURE 8.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in (a) anisole and (b) p-xylene. Integration times for TFC18SF, TFC18SL, DFCISSF, and DFC18SL are 2-min; for MFC 18 are 5-min in all solvents 323

FIGURE 8.2 Raman spectra in the v(C-H) region of TFC 18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in (b) aniline and (b) pyridine. Integration times for TFC18SF, TFC18SL, DFCISSF, and DFC18SL are 2-min; for MFC 18 are 5-min in all solvents 324

FIGURE 8 .3 l[Va(CH2)]/I[Vs(CH2)] as a function of log Kow for TFC18SF (•), TFC 1 SSL (A), DFCISSF (O), DFCISSL (•), and MFC 18 ( V) in p-xylene, toluene, benzene, anisole, aniline, and pyridine. The error bars represent one standard deviation 327

34


FIGURE 8.4 VsCCHs) frequency as a function of log Kow for TFC18SF (•), TFC18SL (•), DFC18SF (O), DFC18SL (•), and MFC 18 (V) in p-xylene, toluene, benzene, anisole, aniline, and pyridine. The error bars represent one standard deviation 333

FIGURE 8.5 Molecular picture of the interactions between TFC18SF and anisole at the chromatographic interface 335

FIGURE 8.6 Molecular picture of the interactions between TFC18SF and toluene at the chromatographic interface 336

FIGURE 8 .7 Molecular picture of the interactions between MFC 18 and anisole at the chromatographic interface 337

FIGURE 8.8 Molecular picture of the interactions between MFC 18 and aniline at the chromatographic interface 338

FIGURE 8 .9 Molecular picture of the interactions between MFC 18 and toluene at the chromatographic interface 339

FIGURE 8.10 (a) l[Va(CH2)]/I[Vs(CH2)] and (b) v.s(CH2) frequency as a function of temperature for TFC18SF in benzene (•), aniline (A), toluene (O), p-xylene (•), anisole (V), and pyridine (*). The error bars represent one standard deviation 342

FIGURE 8.11 (a) I[va(CH2)]/I[v.(CH2)] and (b) Vs(CH2) frequency as a function of temperature for MFC 18 in benzene (•), aniline (•), toluene (O), p-xylene (•), anisole (V), and pyridine (*). The error bars represent one standard deviation 343

FIGURE 8.12 1 [ Va(C H2)]/I [ Vs(C H2)] as a function of temperature for TFC18SL in benzene (•), aniline (•), toluene (O), p-xylene (•), anisole (V), and pyridine (*). The error bars represent one standard deviation 344

35


FIGURE 8.13 I[Va(CH2)]/I[Vs(CH2)] as a function of temperature for DFC18SF in benzene (•), aniline (•), toluene (O), p-xylene (•), anisole (V), and pyridine (*). The error bars represent one standard deviation .345

FIGURE 8.14 I[Va(CH2)]/I[v.s(CH2)| as a function of temperature for DFC18SL in benzene (•), aniline (•), toluene (O), p-xylene (•), anisole (V), and pyridine (¥). The error bars represent one standard deviation .346

FIGURE A8 1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in anisole at (a) -15 "C, (b) -10 °C, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC 1 SSL are 2-min, for MFC 18 are 5-min .349

FIGURE A8.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in anisole at (a) 10 °C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFCI8SL are 2-min, for MFC 18 are 5-min .350

FIGURE A8.3 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFCI8SL, and MFC 18 in anisole at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFCl 8SI. are 2-min, for MFC 18 are 5-min .351

FIGURE A8.4 Raman spectra in the v(C-H) region of TFC18SF, TFCI8SL, DFC18SF, DFC18SL, and MFC 18 in p-xylene at (a) -15 °C, fb) -10 T, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFCl SSL are 2-min, for MFC 18 are 5-min .352

FIGURE A8.5 Raman spectra in the v{C-H) region of TFCISSF, TFC18SL, DFC18SF, DFCl SSL, and MFC 18 in p-xylene at (a) 10 "C and (b) 30 "C. Integration times for TFCISSF, TFCISSL, DFC18SF, and DFCl SSL are 2-min, for MFC 18 are 5-min .353

36


FIGURE A8.6 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in p-xylene at (a) 40 "C and (b) 50 "C, Integration times for TFC18SF, TFC18SL, DFC18SF, and DFCI8SL are 2-min, for MFC 18 are 5-min 354

FIGURE A8.7 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in aniline at (a) -15 °C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC18 are 5-min 355

FIGURE A8.8 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in aniline at (a) 10 "C and (b) 30 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 356

FIGURE A8.9 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in aniline at (a) 40 °C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 357

FIGURE A8.10 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in pyridine at (a) -15 "C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 358

FIGURE A8.11 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in pyridine at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 359

37

FIGURE A8.12

FIGURE 9.1

FIGURE 9.2

FIGURE 9.3

FIGURE 9.4

FIGURE 9.5


Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC!8 in pyridine at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min 360

Raman spectra of TFC18SF in (a and f) 5 mM, (b and g) 15 mM, (c and h) 55 mM, (d and i) 100 mM, and (e and j) 200 mM aqueous NH4CI (left panel) and NaCl (right panel). Acquisition times for all spectra are 2 min 365

Raman spectra of MFC 18 in (a and d) 5 mM, (b and e) 15 mM, and (c and t) 55 mM aqueous NH4CI (left panel) and NaCl (right panel).

Acquisition times for all spectra are 5 min 366

Raman spectra of TFC18SF (left panel) and MFC 18 (right panel) in pH-adjusted water at pH (a and f) 2.29, (b) 3.93, (c and g) 6.16, (d and h) 7.08, and (e and I) 9.0. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 spectra are 5 min 367

I[Va(CH2)]/I[Vs(CH2)] as a function of (a) salt concentration and (b) pH for TFCISSF in aqueous NaCl (V), NH4C1 (•) and in pH-adjusted water (•) and for MFC 18 in aqueous NaCl (O), NH4CI (•) and in pH-adjusted water (•). Error bars represent one standard deviation 369

Raman spectra of TFCISSF (left panel) and MFC 18 (right panel) in 37 mM aqueous benzoic acid solutions at pH (a) 2.30, (b) 3.38, (c) 4.23, (d) 7.86, (e) 9.33, (f) 2.30, (g) 3.41, (h) 4.12, (i) 5.25, and (j) 9.00. Acquisition times for TFCISSF Spectra are 2 min and for MFC IS spectra are 5 min 372

38


FIGURE 9.6 Raman spectra of TFC18SF (left panel) and MFC 18 (right panel) in 37 mM aqueous benzylamine solutions at pH (a and e) 1.93, (b and f) 4.10, (c and g) 9.40, (d and h) 10.8. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 are 5 min 373

FIGURE 9.7 Raman spectra of TFC18SF in 37 mM aqueous aniline solutions at pH (a) 1.83, (b) 3.86, (c) 4.60, (d) 4.86, (e) 5.95, and (f) 7.95. Acquisition times for all spectra are 2 min 374

FIGURE 9.8 l[Va(CH2)]/I[Vs(CH2)] as a function of pH for (a) TFC18SF in 37 mM aqueous benzoic acid (•) and pH-adjusted water (•) and (b) MFC 18 in 37 mM aqueous benzoic acid (•) and pH-adjusted water (O). Error bars represent one standard deviation 375

FIGUHE 9.9 l[Va(CH2)]/I[Vs(CH2)] as a function of pH for TFC18SF in 37 mM aqueous aniline (A) and pH-adjusted water (•). Error bars represent one standard deviation 378

FIGURE 9.10 I[Va(CH2)]/I[vs(CH2)] as a function of pH for (a) TFC18SF in 37 mM aqueous benzylamine (A) and pH-adjusted water (•) and (b) MFC 18 in 37 mM aqueous benzylamine (*) and pH-adjusted water (O). Error bars represent one standard deviation. .380

FlGURE9.il I[Va(CH2)]/I[Vs(CH2)3 as a function of log Kow for TFC18SF (•) and MFC 18 (•) in air, neat aniline, and neat benzylamine. Error bars represent one standard deviation .382

FIGURE 10.1 Overlaid chromatograms showing the separation of toluene, anisole and aniline on MFC! 8. Experiments were done in a 100% methanol mobile phase with a constant flow rate of 1.0 mL/min .395

39


FIGURE 10.2 Chromatograms showing increased aniline retention on MFC 18 with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water 398

FIGURE 10.3 Chromatograms showing increased anisole retention on MFC 18 with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water 399

FIGURE 10.4 Chromatograms showing increased toluene retention on MFC18 with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water 400

FIGURE 10.5 Raman spectra of MFC 18 in the v(C-H) region in 5 mM aniline (left panel), 5 mM anisole (center panel), and 5 mM toluene (right panel) solutions in (a, d, and g) 100% methanol, (b, e, and h) 75:25 methanol/water, and (c, f and i) 50:50 methanol/water. Acquisition times all spectra are 5 min 406

FIGURE 10.6 Plot of I[Va(CH2)]/I[Vs(CH2)] as a function of solute capacity factor for toluene, anisole, and aniline on MFC 18 (•) columns 408

FIGURE 10.7 Plot of I[Va(CH2)]/I[Vs(CH2)] and capacity factor as a function of percent water in methanol for MFC 18 in aniline/methanol/water solutions. Horizontal lines represent the value of I[Va(CH2)]/I[Vs(CH2)] for MFC 18 in neat methanol 413

FIGURE 10.8 Plot of I[Va(CH2)]/I[Vs(CH2)] and capacity factor as a function of percent water in methanol for MFC 18 in anisole/methanol/water solutions. Horizontal lines represent the value of I [Va(CH2)]/I [Vs(CH2)] for MFC 18 in neat methanol 415

40


FIGURE 10.9 Plot of l[Va(CH2)]/I[v.s(CH2)] and capacity factor, as a function of percent water in methanol for MFC 18 in toluene/methanol/water solutions. Horizontal lines represent the value of I[Va(CH2)]/I[Vs(CH2)] for MFC 18 in neat methanol 417

FIGURE 10.10 Molecular picture of the methanol mobile phase-MFCIS interface 420

FIGURE 10.11 Molecular pictures of the aniline/MFC 18 interface for (a) MFC 18 in neat anilne and (b) MFC 18 in 5 mM aniline in methanol 422

FIGURE 10.12 Molecular pictures of the anisole/MF C18 interface for (a) MFC 18 in neat anisole and (b) MFC 18 in 5 mM anisole in methanol 424

FIGURE 10.13 Molecular pictures of the toluene/MFC 18 interface for (a) MFC 18 in neat toluene and (b) MFC 18 in 5 mM toluene in methanol 425

FIGURE 10.14 Molecular pictures of (a) MFC 18 in 5 mM anisole in 50:50 methanol/water and (b) MFC 18 in 5 mM toluene in 50:50 methanol/water 427

FIGURE 10.15 Overlaid chromatograms showing the separation of toluene, anisole and aniline on TFC18SF. Experiments were done in a 100% methanol mobile phase with a constant flow rate of 1.0 mL/min 429

FIGURE 10.16 Chromatograms showing increased aniline retention with increasing aqueous component of the mobile phase from (a) 100% methanol to fb) 75:25 methanol/water to (c) 50:50 methanol/water 431

FIGURE 10.17 Chromatograms showing increased anisole retention with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water 432

41


FIGURE 10.18 Chromatograms showing increased toluene retention with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water 433

FIGURE 10.19 Schematic of the free volume differences between MFC 18 and TFC18SF which lead to differences in H values for each phase 436

FIGURE 10.20 Raman spectra of TFC18SF in the v{C-H) region in 5 niM aniline (left panel), 5 mM anisole (center panel), and 5 mM toluene (right panel) solutions in (a, d, and g) 100% methanol, (b, e, and h) 75:25 methanol/water, and (c, f, and i) 50:50 methanol/water. Acquisition times all spectra are 2 min 438

FIGURE 10.21 Plot of I[Va(CH2)J/I[v.s(CH2)] as a function of solute capacity factor for toluene, anisole, and aniline on the TFC18SF column. Horizontal lines represent the value of I[Va(CH2)]/I[Vs(CH2)] for

TFC18SF in neat methanol, aniline, anisole, and toluene 439

FIGURE 10.22 Plot of I[Va(CH2)]/I[Vs(CH2)] and capacity factor as a function of percent water in methanol for TFC18SF in aniline/methanol/water solutions. Horizontal lines represent the value of I[Va(CH2)]/I[Vs(CH2)] for TFC18SF in neat methanol 442

FIGURE 10.23 Plot of I[Va(CH2)]/I[Vs(CH2)] and capacity factor as a function of percent water in methanol for TFC18SF in anisole/methanol/water solutions. Horizontal lines represent the value of I[Va(CH2)]/I[Vs(CH2)] for TFC18SF in neat methanol 443

42


FIGURE 10.24

FIGURE 10.25

FIGURE 10.26

FIGURE 10.27

Plot of I[Va(CH2)]/I[Vs(CH2)] and capacity factor as a function of percent water in methanol for TFC18SF in toluene/methanol/water solutions. Horizontal lines represent the value of I[Va(CH2)]/I[Vs(CH2)] for TFC18SF in neat methanol 444

Molecular pictures of the aniline/TFC18SF interface for (a) TFC18SF in neat anilne and (b) TFC18SF in 5 mM aniline in methanol 447

Molecular pictures of the anisole/TFClBSF interface for (a) TFC18SF in neat anisole and (b) TFC18SF in 5 mM aniline in methanol 448

Molecular pictures of the toluene/TFC18SF interface for (a) TFC18SF in neat toluene and (b) TFC18SF in 5 mM aniline in methanol 450

FIGUE 10.28 Raman spectra of TFC18SF (left panel) and MFC 18 (right panel) in (a and c) 10 mM and (b and d) 25 mM toluene in 100% methanol solutions. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 spectra are 5 min. .452

FIGURE 10.29 Plot of I[Va(CH2)]/I[Vs(CH2)] as a function of toluene concentration for TFC18SF (•) and MFC 18 (A) in toluene/methanol solutions. Horizontal lines represent the value of I[Va(CH2)]/I[vs(CH2)] for TFC18SF and MFC 18 in neat methanol .453

FIGURE 10.30 Raman spectra of TFC18SF (left panel) and MFC 18 (right panel) in (a and d) 5 mM, (b and e) 10 mM, and (c and f) 25 mM toluene in 100% acetonitrile solutions. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 spectra are 5 min .456

43


FIGURE 10.31 Plot of I[Va(CH2)]/I[Vs(CH2)] as a function of toluene concentration for TFC18SF (•) and MFC 18 (A) in toluene/acetonitrile solutions. Horizontal lines represent the value of I[Va(CH2)]/I[Vs(CH2)] for TFC18SF and MFC 18 in neat acetonitrile 457

FIGURE 11.1 Trichlorosilylpropyltrimethylamomium chloride and a simple illustration of the SAX surface 469

FIGURE 11.2 Trichlorosilylpropylsulfonic acid and a simple illustration of the SCX-2 surface 470

FIGURE 11.3 Raman spectra of (a) dry SAX (as received) and (b) SAX in water at 20 °C. All spectra are acquired with 30-min integrations 473

FIGURE 11,4 Raman spectra of SAX in 0.3 M aqueous solutions of (a) NaCl, (b) NaBr, and (c) Nal. All spectra are acquired with 30-min integrations 476

FIGURE 11.5 Raman spectra of SAX in 0.3 M aqueous solutions of (a) NaC104, (b) NaNOs, and (c) NalO, All spectra are acquired with 30-min integrations 478

FIGURE 11.6 Raman spectra of SAX in 0.3 M aqueous solutions of (a) Na2S04, (b) Na3P04, (c) NaHS04, and (d) NaH2P04. All spectra are acquired with 30-min integrations 479

FIGURE 11.7 Plot of Va(CH3)K frequency as a function of anion selectivity for SAX in 0.3 M aqueous solutions of sodium halides (O ) and sodium oxyanions (•). Error bars represent one standard deviation 487

44

FIGURE 11.8

FIGURE 11.9

FIGURE 11.10

FIGURE 11.11

FIGURE 11.12

FIGURE 11.13

FIGURE 11.14


Plot of I[Va(C-N-C)]/I[Vs(C-N-C)] as a function of anion selectivity for SAX in 0.3 M aqueous solutions of sodium halides (O) and sodium oxyanions (•). Error bars represent one standard deviation 488

Raman spectra of (a) TMACl, (b) TMABr, (c) TMAI, (d) TMAHSO4, (e) TMANO3, and (1) TMACIO4. Integration times are 5-min for all spectra 498

Raman spectra of 0.3 M aqueous solutions of (a) TMACl, (b) TMABr, (c) TMAI, (d) TMAHSO4, (e) TMANO3, and (f) saturated TMACIO4. Integration times are 30-min for all spectra 499

Va(CH3)N frequency as a function of anion selectivity for (a) tetramethylammonium (TMA) halide salts (O) and TMA oxyanion salts (•) and (b) 0.3 M aqueous solutions of TM A halide salts (O) and TMA oxyanion salts (•) . Error bars represent one standard deviation 501

Raman spectra of SCX-2 (a) dry and (b) in water at 20 °C. Integrations are 20-min in the low frequency region and 30-min in the high frequency region 505

Raman spectra of SCX-2 in 0.2 M aqueous solutions of (a) NaCl (b) KCl and (c) CsCl at 20 "C. Integrations are 20-min in the low frequency region and 15-min in the high frequency region 508

(a) v(S03) frequency and (b) v(S03) FWHM as a function of cation selectivity for dry SCX-2 (•), SCX-2 in water (•) and in 0.2 M aqueous solutions of NaCl (A), KCl (•), CsCl (•). Error bars represent one standard deviation 509

45

FIGURE 11.15

FIGURE 11.16

FIGURE 11.17

FIGURE 11.18

FIGURE 11.19

FIGURE 11.20

FIGURE 12.1


Raman spectra of (a) methanesulfonic acid and (b) 0.2 M aqueous methanesulfonate. Integrations are 5-min in the low frequency region and 2-min in the high frequency region for methanesulfonic acid and 15-min for 0.2 M aqueous methanesulfonate 514

Raman spectra of 0.2 M aqueous solutions of methanesulfonate and (a) NaCl,(b) KCl, and (c) CsCl. Integrations are 15-min for all spectra 518

(a) v(S03) frequency and (b) vfSOs) FWHM as a function of cation selectivity for 0.2 M aqueous methanesulfonate (•) and with 0.2 M aqueous solutions of NaCl (A), KCl (•), CsCl (•). Error bars represent one standard deviation 519

Raman spectra of (a) crystalline TMABr, (b) 0.2 M aqueous TMABr, and (c) 0.2 M aqueous TMABr + SCX-2 at 20 "C. Integrations are (a) 2-min low frequency and 3-min high frequency, (b) 15-min low frequency and 10-min high frequency, and (c) 30-min 521

Raman spectra of (a) crystalline TEABr, (b) 0.2 M aqueous TEABr, and (c) 0.2 M aqueous TEABr + SCX-2 at 20 °C. Integrations are (a) 2-min low frequency and 3-min high frequency, (b) 15-min low frequency and 10-min high frequency, and (c) 30-min 522

Raman spectra 0.2 M methanesulfonate with (a) 0.2 M aqueous TMABr and (b) 0.2 M aqueous TEABr at 20 °C. Integrations are 15-min for all spectra 526

Raman spectra of 1 solute CN in (a) air, (b) acetonitrile, (c) methanol, (d) water, (e) toluene, (f) benzene, (g) chloroform, (h) heptane. Acquisition times all spectra are 2 min 537

46


FIGURE 12.2 v(CN) frequency as a function of solvatochromic parameter k* for I solute CN in air, heptane, toluene, chloroform, benzene, methanol, acetonitrile, and water. Error bars represent one standard deviation 540

FIGURE 12.3 v.sCCH:) + Vs(CH2)c:n frequency as a function of solvatochromic parameter 7t* for Isolute CN in air, heptane, toluene, chloroform, benzene, methanol, acetonitrile, and water. Error bars represent one standard deviation 543

FIGURE 12.4 Va(CH2)c .v frequency as a function of solvatochromic parameter 7i* for Isolute CN in air, heptane, toluene, chloroform, benzene, methanol, acetonitrile, and water. Error bars represent one standard deviation .544

FIGURE 12.5

FIGURE 12.6

VaCCHs) frequency as a function of solvatochromic parameter k* for Isolute CN in air, heptane, toluene, chloroform, benzene, methanol, acetonitrile, and water. Error bars represent one standard deviation .545

Raman spectra of Isolute Phenyl in (a) air, (b) acetonitrile, (c) water, (d) benzene, (e) heptane, and (f) chloroform. Acquisition times are 1 niin in the low frequency region and 2 min in the high frequency region .547

FIGURE 12.7 Va(CH2) frequency as a function of solvatochromic parameter n* for Isolute CN in air, heptane, chloroform, benzene, acetonitrile, and water. Error bars represent one standard deviation 550

FIGURE 12.8 Vs(CH2) frequency as a function of solvatochromic parameter n* for Isolute CN in air, heptane, chloroform, benzene, acetonitrile, and water. Error bars represent one standard deviation 551

47


FIGURE 13.1 Chromatograms of benzene, ethylbenzene, and cumene using (a) Krom-Test A (b) Krom-Test B, and (c) Krom-Test C columns (Kromasil Cig commercially available stationary phase) in 80:20 methanol/water at a flow rate of 80 fiL/min 562

FIGURE 13.2 Chromatograms of benzene, ethylbenzene, and cumene using (a) BAS-1 and (b) BAS-2 commercial Cix columns in 80:20 methanol/water at a flow rate of 80 [iL/min 563

FIGURE 13.3 Raman spectra in the v(C-C), 8(C-H), and v(C-H) regions for Kromasil Cig. Acquisition times are 8 min in the low frequency region

and 4 min in the high frequency region 567

FIGURE 13.4 SEM images at 3000 x magnification of (a) Kromasil Cig and (b) unmodified Kromasil silica 568

FIGURE 13 .5 Raman spectra in the v(C-C), 6(CH-H), and v(C-H) regions for (a) CJ017-1, (b) CJ017-2, (c) CJ017-3, and (d) CJ0225. Acquisition times are 8 min in the low frequency region and 4 min in the high frequency region for all spectra 571

FIGURE 13.6 SEM images at 3000 x magnification of (a) Kromasil Cig, (b) CJ0225, and two regions ofCJOI7-2 (c and d) 572

FIGURE 13.7 Self-assembly of hydrolyzed octadecylsilane into multilayer structures adapted from reference 13.28 574

FIGURE 13.8 Chromatograms benzene, ethylbenzene, and cumene on (a) CJ017-2 and (b) CJ0225 in 80:20 methanol/water at a flow rate of 80 jiL/min 576

48


FIGURE 13.9 Raman spectra in the v(C-C), 5(C-H), and v(C-H) regions for (a) CJ03-1, (b) CJ03 -2, and (c) CJ03 -3. Acquisition times are (a and b) 8 min in the low frequency region and 2 min in the high frequency region for and (c) 12 min in the low frequency region and 5 min in the high frequency region .579

FIGURE 13.10 Raman spectra in the v(C-C), 5(C-H), and v(C-H) regions for (a) CJ04-1, (b) CJ04 -2, and (c) CJ04 -3. Acquisition times are (a and b) 8 min in the low frequency region and 2 min in the high frequency region for and (c) 12 min in the low frequency region and 5 min in the high frequency region .580

FIGURE 13.11 Raman spectra in the v(C-C), 8(C-H), and v(C-H) regions for (a) CJ05-1, (b) CJ05 -2, (c) CJ05 -3, and (d) CJ05 -4. Acquisition times are 8 min in the low frequency region and 3 min in the high frequency region for all spectra 583

FIGURE 13 .12 I[Va(CH2)]/I[Vs(CH2)] as a function of surface coverage (fimol/m^) for CJ04-1 and CJ05 phases (•) and NIST Cig phases (O) 584

FIGURE 13.13 Chromatograms of benzene, ethylbenzene, and cumene using (a) CJ03-2, (b) CJ04-1, and (c) CJ04-3 in 80:20 methanol/water at a flow rate of 80 (.iL/min 586

FIGURE 13.14 Chromatograms of benzene, ethylbenzene, and cumene using (a) CJ05-1, (b) CJ05 -2, (c) CJ05 -3, and (d) CJ05 -4 in 80:20 methanol/water at a flow rate of 80 jiL/min 587

FIGURE 13.15 Schematic of octanol displacement by alkylsilanes at the silica surface 590

49


FIGURE 13.16 Schematic of octanol incorporation in self-assembled, hydrolyzed alkylsilane 591

FIGURE 13.17 Chemical structures and geometric shapes of PAHs used to determine RPLC column shape selectivity 593

FIGURE 13.18 Chromatograms of PAHs used to determine shape selectivity adapted from reference 2.8 594

FIGURE 13.19 Chromatograms of PhPh, BaP, and TBN (SRM 869a) using (a) BAS-1, (b) Krom-Test C, and (c) CJ04-3 columns in a mobile phase of 85:15 acetonitrile/water at a flow rate of 170 [xL/min 596

FIGURE 13.20 Chromatograms of PhPh, BaP, and TBN (SRM 869a) using (a) CJ05-1, (b) CJ04-1, (c) CJ05-2, (d) CJ05-3, and (e) CJ05-4 columns in a mobile phase of 85:15 acetonitrile/water at a flow rate of 170 (iL/min 597

FIGURE 13.21 ajBN/BaP as a function of surface coverage for CJO phases (•), commercial Kromasil Cig and BAS phases (*), and NIST Cis phases (O) 599

50

LIST OF TABLES

TABLE 2.1; Properties of Isolute stationary phases 76

TABLE 2.2: Physical properties of Kromasil silicas and stationary phases 83

TABLE 2.3; Physical properties ofNIST Cis stationary phases 85

TABLE 3.1; Raman peak frequencies (cm"') and assignments for octadecane and polyethylene 107

TABLE 4.1: Raman peak frequencies and assignments for Isolute C18MF at 24 T 168

TABLE 4,2; Solvatochromic parameters" K*, a, and (3 for solvents 170

TABLE 5.1: Stationary phase properties 181

TABLE 5.2; Peak frequencies and assignments for NIST Cix stationary phases 184

TABLE 5.3; Break point temperatures and slopes from temperature-dependent I[Va(CH2)]/I[Vs(CH2)] measurements 190

TABLE 6.1; Solvatochromic parameters", %*, a, and P for solvents 228

TABLE 7.1; Solvatochromic parameters", a, and P for solvents 285

TABLE 8.1; Octanol-water partition coefficients for aromatic compounds 326

TABLE 10.1; Solution free energies for solutes in methanol, water, and hexadecane 391

TABLE 10.2; Predicted AGRETN values from solution free energy values 392

TABLE 10.3; Capacity factor, k', values for analytes on MFC18 401

TABLE 10.4; Numbers of theoretical plates and plate heights for solutes on MFC 18 403

51


TABLE 10.5: Capacity factor, k\ values for analytes on TFC18SF 434

TABLE 10.6; Numbers of theoretical plates and plate heights for solutes on TFC18SF 435

TABLE 11.1; Peak frequencies and assignments for Isolute SAX (TMAP^) 474

TABLE 11.2; Peak frequencies and assignments for NaNOs 480

TABLE 11.3; Peak frequencies and assignments for NaHSO-j 481

TABLE 11.4; Peak frequencies and assignments for NaH2P04 482

TABLE 11.5; Peak frequencies and assignments for NaC104 483

TABLE 11.6; Peak frequencies and assignments for NalOi 484

TABLE 11.7; Peak frequencies and assignments for Na2S04 485

TABLE 11.8; Peak frequencies and assignments for Na3P04 486

TABLE 11.9; Hydration enthalpies, ionic radii, and hydrated radii of anions 490

TABLE 11.10; Approximate charge-separation distance between TMAP^ and halides 493

TABLE 11.11; Peak frequencies and assignments for TMACl, TMABr, and TMAl 500

TABLE 11.12; Peak frequencies and assignments for SCX-2 506

TABLE 11.13; Hydration enthalpies, ionic radii, and hydrated radii of cations 512

TABLE 11.14; Approximate charge-separation distance between alkali-metal ions and surface-bound sulfonate 513

52


TABLE 11.15: Peak frequencies and assignments for methane sulfonic acid 516

TABLE 11,16: Peak frequencies and assignments for methane sulfonate 519

TABLE 11.17: Peak Frequencies and Assignments for TEABr 523

TABLE 11.18: Peak frequencies of Va(CH3)N and VaCCHs) modes of TMA" and TEA (cm"') 524

TABLE 11.19: Peak frequencies of key modes for tetraalkylammonium methanesulfonate solutions 527

TABLE 12.1: Properties of 1 solute stationary phases 535

TABLE 12.2: Peak frequencies and assignments for Isoiute CN 538

TABLE 12.3: Solvatochromic parameters'", %*, a, and P for solvents 540

TABLE 12.4: Peak frequencies and assignments for Isoiute Phenyl 548

TABLE 13 .1: Packing pressures and number of theoretical plates (N) for Kromasil Cig and HAS Cig columns 564

TABLE 13.2: CJOl and CJ0225 phases and properties 570

TABLE 13.3: CJ03 and CJ04 phases and properties 578

TABLE 13.4: CJ05 phases and properties 582

TABLE 13.5: Number of theoretical plates (N) for CJ03, CJ04, and CJ05 columns 588

TABLE 13.6: Shape selectivities (axBN/Bap) for Kromasil Cig, BAS, and CJO columns 598

ABSTRACT

The use of Raman spectroscopy to determine structure-function relationships in

octadecylsilane stationary phases for reversed-phase liquid chromatography is presented.

Raman spectroscopy is used to determine rotational and conformational order of a series

of high-density octadecylsilane stationary phases that vary in surface coverage from 3.09

to 6 .45 p,mol/m^ as a function of numerous chromatographic parameters. The effect of

these conditions on the conformational order of the alkylsilanes offers information about

molecular interactions at the chromatographic interface, which can be used to gain insight

to the mechanisms of solute retention in reversed-phase liquid chromatography.

Rotational and conformational order of these unique phases is studied under a

plethora of chromatographic conditions including temperature, exposure to polar and

nonpolar solvent, and in the presence of multi-component solutions including mobile

phase/solute systems. Temperature-induced surface phase changes for this homologous

series of materials are observed that are demonstrated to adhere to the Clapeyron

equation for a simple first-order transition. Phase changes are discussed in terms of

molar volume, molar enthalpy, and molar entropy, all of which are highly dependent on

the surface coverage of the stationary phase

Solvent interactions with these phases are dependent not only on solvent

properties (size, shape, polarity, hydrogen-bonding ability, etc.), but also stationary phase

properties including surface coverage and homogeneity. In general, nonpolar

hydrophobic solvents induce conformational disorder to the low coverage stationary

54

phases. Conversely, polar solvents have a subtle ordering effect on the higher density,

homogeneous alkylsilane phases, attributed to solvent self-associating at the stationary

phase/solvent interface. In the presence of mobile phase/solute systems, both high and

low coverage phases are very well ordered, because of the combination of strong

interactions of solutes with both the stationary phase and the mobile phase and mobile

phase solvent self-association at the interface. Spectral data are correlated with the

chromatographic behavior of the solutes to gain insights into retention mechanisms for

these compounds. Results suggest solute partitioning into the stationary phase takes

place, but partitioning in some cases is very subtle, restricted to the distal end of the

alkylsialnes.

Raman spectroscopy is also used to investigate fimdamental interactions of ion

exchange chromatographic systems. For anion exchange systems, the structure of

surface-bound trimethylammonium ions is affected by hydrated counter anions of varying

charge-density in aqueous solutions, changing systematically with chromatographic

retention of the halide series. However, the structure of surface-confined sulfonate ions

is unaffected by the presence of counter cations of different size or charge-density for

cation exchange systems. The difference in behavior of these systems is attributed to

poor hydration of trimethylammonium and anions relative to sulfonate ions and cations.

In addition to the characterization of existing phases, alkylsilane-based stationary

phases are fabricated using a novel solution modifier approach. This simple approach

allows for the synthesis of materials that vary in surface coverage including the

55

fabrication of high surface coverage phases (>5 lamol/m') using the same reaction

chemistry. Phases have aJkylsilane structure and architecture similar to existing phases

of comparable coverage. In addition, the phases give comparable chromatographic

performance to commercial phases including a high degree of selectivity of high-density

phases for planar polycyclic aromatic hydrocarbons.

56

Chapter 1

introduction to Alkylsilane-Based Stationary Phases for Reversed-Phase

Liquid Chromatography; Interfacia! Importance and Retention Mechanisms

Aikylsilane-Based Stationary Phases in Reversed-Phase Liquid Chromatography

The most common stationary phases used in reversed-phase liquid

chromatography (RPLC) are alkylsilane-modified silicas in packed-bed columns. While

other stationary phases and column architectures have been proven to be useful for RPLC

separations,'^"'^ they represent only a small traction of reversed-phase separations.

Commercially available RPLC silica-based phases are the stationary phases of choice

because they can be manufactured precisely, offer excellent mechanical stability, and can

be packed into columns reproducibly, resulting in good column-to-column separation

reproducibility.

The use of alkylsilanes to modify particulate supports for liquid chromatography

was first described by Abel et al. in 1966.' This relatively simple modification

procedure quickly gained popularity and became commonplace in analytical separations

in the few years following its conception.'"'"' The hydrolysis and condensation of n-

alkyltrichlorosilanes on silica substrates is shown schematically in Figure 1.1,

Chlorosilanes are hydrolyzed to silicic acid in the presence of water followed by the

57

/ CI

CH3-(CHA-Si-Cl

CI

Hydrolysis

/ OH

CH3-(CH2)„-Si - OH

\)H

Condensation

/O-

CH3-(CH2)„-Si-O-l

/O

CH3-(CH2)„-Si-O-

OH

o •

03

Figure 1.1 Schematic of the hydrolysis and condensation of n-alkylchlorosilanes on silica substrates.

condensation of silicic acid with surface silanols to form covalent Si-O-Si bonds. The

result is an organically-modified, hydrophobic surface.

Alkylsilane phases are prepared using precursors of varying functionality in order

to synthesize phases with difterent bonded architectures. Alkyltrichlorosilanes and

alkyldichlorosilanes are used in the preparation of what are called "polymeric" phases,

because these silanes have multiple hydrolyzable groups that can undergo not only

condensation on the silica surface, but can cross-link with neighboring silanes. Phases

prepared using alkylchlorosilanes or monofunctional silanes are termed monomeric

phases because these silanes only have one attachment point. These different silanes

allow for different bonding densities where the theoretical maximum surface coverage of

a monofunctional silane is - 4 [imol/m^, compared to ~ 8 [.imol/m^ for a trifunctional

silane, which leads to variations in chromatographic performance.^ While n-alkyl

phases are the most common RPLC phases and will be the focus of this discussion,

phases prepared with silanes with different functional groups at the distal end of the alkyl

chain are also used including phenyl-, cyano-, alcohol-, and amine-terminated

alkylsilanes.

Retention Models

While the basis for analyte separation in RPLC is often described in terms of

polarity differences between analytes, the mechanism of analyte retention is an issue of

great debate. It is understood that different physical properties of analytes are ultimately

59

responsible for their separation in RPLC. However, a detailed molecular understanding

of analyte interactions with mobile and stationary phases under chromatographic

conditions, in other words, retention mechanisms, remain elusive.

Numerous models have been proposed over the past twenty-five years to explain

and predict retention in RPLC. Nearly all variations and derivatives of these models can

be classified into one of two general categories, the solvophobic and partitioning models.

Each of these models will be described in terms of their treatment of the stationary phase

and the consequences on interfacial chromatographic phenomena.

Solvophobic Theory

The solvophobic (also referred to as hydrophobic or adsorptive) model of

chromatographic retention was first described by Horvath in 1976.' This model treats

the stationary phase as simply a hydrophobic, passive entity.' The driving force for

retention in this model is dominated by the free energy of the mobile phase. Initially,

nonpolar solute solvated in the polar mobile phase is entropically unfavorable as solvent

molecules in contact with the solute are thought to form a highly ordered network. Then,

when the solute adsorbs to the nonpolar stationary phase, the mobile phase "relaxes" to

an entropically more favorable energetic state. Here, the energetics of the mobile phase

dictate the extent of adsorption and desorption of solutes to the stationary phase

regardless of the conformation of the stationary phase. By this model, the adopted

60

conformation of the stationary phase equilibrated in mobile phase should be unchanged in

the presence of solute.

Partitioning Model

The partitioning model (also referred to as the interphase model), indroduced by

Dill et al.,^'^^ treats the stationary phase as an active unit that can compete with the

mobile phase in the solvation process/'^"^ The energetics for retention are governed by

opposing solvation energies of the solute in the stationary phase and in the mobile phase.

Solutes can become imbedded in the stationary phase and the stationary phase can act as

an immobilized solvent. In order to house solutes within the stationary phase, the phase

must undergo conformational changes to accommodate solute in cavities. In this model,

the conformation of the stationary phase should be considerably different in the presence

and absence of solute.

While these models offer possible molecular interactions tiiat could govern solute

retention, neither model may accurately describe retention for all stationary phase, solute,

and mobile phases systems. Several pieces of evidence suggest that solute retention

cannot be described completely by either model, including the following; (1) wetting

studies suggest that solute or solvent molecules are not always completely imbedded

within the stationary phase,' (2) observations that entropic contributions dominate

solute retention with increasing bonding density of alkylsilanes,'^' '^'''' and (3) no

single equation has been derived to describe solute retention following either model.

61

Solute retention in RPLC is a complex process which may require pieces of each of

these two models or additional information to accurately describe retention for all

possible chromatographic systems.

Importance of Alkvlsilane Structure

The inability to accurately determine a mechanism for retention stems from the

difficulty of studying the interactions known to effect solute retention, including solvent-

stationary phase, solute-stationary phase, and stationary phase inter- and intramolecular

interactions. Thus, understanding the role of the stationary phase is crucial for

determining retention mechanisms in RPLC, specifically (1) understanding the structure

of the stationary phase and how structure influences solvent and solute interactions with

the phase and (2) understanding how solvent and solute interactions with the stationary'

phase affect its structure.

A number of models were proposed to describe the structure of surface-confined

alkylsilanes at the onset of the popularity of RPLC in the late 1970s, determined

exclusively from chromatographic experiments. Karch was the first to describe the

structure of the alkylsilane surface as hydrophobic "bristles", much like a bristle brush,

where the alkyl chains are fully extended away from the silica surface.'^" Hemetsberger

presented a different model which considers the hydrophobic or solvophobic interactions

between alkyl chains, in which the alkyl chains are folded on to one another, minimizing

exposed surface area and maximizing hydrophobic interactions.' An alternative model

was presented by Lochmuller, where the alkylsilanes are arranged in isolated patches on

the silica surface as condensed liquid droplets. Perhaps the most usefol model

proposed for alkylsilane structure is one proposed by Gilpin/ In this model, the

stationary phase is treated as a dynamic entity and transitions between folded or collapsed

states and bristle-like structures take place depending on the chemical environment of the

phase.

In more recent years, there has been a great deal of experimental and

computational work done to determine the role of alkylsilane structure in retention

mechanisms in RPLC. Chromatographic methods,^ tluorescence

s p e c t r o s c o p y , ' ^ ' s m a l l a n g l e n e u t r o n s c a t t e r i n g , ' a n d d i f f e r e n t i a l s c a n n i n g

calorimetry^'^^'^'^^ have been employed to study retention in RPLC. However, utilizing

molecular dynamic simulations' " and experimental techniques sensitive to

alkylsilane structure and order including nuclear magnetic resonance (NMR)

spectroscopy,' înfrared (IR) spectroscopy,' and Raman spectroscopy

have perhaps offered the most insight into molecular retention in RPLC.

Molecular dynamic simulations have offered a significant amount of alkyl

structure information as well as information about solvent and analyte interactions at the

chromatographic interface. Klatte and Beck'^^' have shown the thickness of

alkysilane monolayers to increase with increasing surface coverage from 2.5 to 4.0

p.mol/m*, suggesting that a lower coverages the alkyl chains are tilted and form a layered

structure in order to maximize chain-chain coupling interactions and the tilt angle

63

approaches zero to the surface normal as surface coverage increases, maintaining

considerable chain-chain interactions. These observations are confirmed by neutron

152 1 54 • 1 57 scattering experiments ' ' ' and other molecular dynamics work. ' Solvent

interactions with alkylsilane monolayers have also been studied by molecular dynamics.

Slusher and Mountain'" have shown the existence of a sharp interface between a Cg

alkylsilane monolayer of relatively high surface coverage, 5.1 jimol/m^, and pure water,

^ O with some water penetration at the distal end of the alkyl chains by ~5 A. In the presence

of mixed solvent systems, they determined a high degree of segregated organic

component at the interface.

^^Si and ' 'C cross-polarization/magic angle spinning (CP/MAS) NMR have been

used to characterize alkylsilane structure of stationary phases. ^^Si CP/MAS NMR gives

surface specific information about the chemical environment of Si atoms on both the

alkylsilanes and the substrate including degree of crosslinking of alkylsilanes and

differentiation between bonded and free surface silanols.' ^ CP/MAS NMR data

give alkyl chain mobility information. ' • "' Pursh and coworkers have shown chain

rigidity to increase with increasing alkylsilane surface coverage for a series of C22-

stationary phase' and with increasing alkyl chain length for C18-C34 phases. While

these data can be used to infer alkyl conformational order, it is important to note that

NMR experiments do not directly measure alkyl chain conformation. These solid-state

NMR techniques are quite useful in determining stationary phase composition and

64

mobility, but cannot adequately describe specific analyte interactions with the stationary

phase.

In addition to NMR spectroscopy, FT-IR spectroscopy has been used to study

alkylsilane stationary phases, but to a lesser extent. Sander and coworkers used FT-IR

spectroscopy to determine alkylsilane conformation of bonded RPLC phases.^''"* The IR

spectra of alkanes contain a great deal of conformational order including vibrational

modes that indicate end-gauche (1341 cm"'), kink (1367 cm"'), and double-gauche (1354

cm"') conformers. Results suggest a significant degree of disorder in CIK and C22 bonded

phases at 44 "C, where ~ 40% of the end methyl groups arc in the end-gauche

conformation, and ~ 30% of the internal C-C bonds are in the gauche conformation. The

degree of disorder in the alkyl chains decreases with decreasing temperature. However,

the alkyl chains do not achieve exclusively an all trans conformation, even at - 30 T. In

addition, the effect of solvent on the conformational order of the Cig phase is assessed.

While there is a great deal of conformational order information available in the IR

spectrum, much of this information is lost in the presence of solvent due to severe

spectral interference. Despite the interference, information available in the v(C-C) region

suggests that the alkyl chains become considerably more ordered in the presence of

methanol. This ordering effect is attributed to preferential methanol orientation with the

methyl group intercalated between the alkyl chains, allowing the hydroxyls to hydrogen-

bond with neighboring intercalated molecules and with those in bulk solution.

65

Perhaps the most useM experimental technique for determining the structure of

alkanes is Raman spectroscopy because of its sensitivity to alkyl chain conformation,

large number of Raman-active modes for alkanes, and, most importantly, spectral data

provide a direct measure of alkyl chain conformational order.These characteristics

of the technique are elaborated in Chapter 3 and throughout this Dissertation. Despite

these well-known characteristics, Raman spectroscopy has only been applied to the study

of alkylsilane-bonded stationary phases in a limited number of reports. Work done by Ho

et shows the conformational order of two low coverage Cig solid phase

extraction (SPE) phases, a polymeric and a monomeric phase, at room temperature to be

the same as liquid octadecyltrichlorosilane (OTS) and dimethyloctadecylchlorosilane

(DOS), respectively. Furthermore, it is determined that these phases are more ordered at

liquid Nj temperature than at ambient temperature, but still contain a significant fraction

of gauche conformers. While neat OTS and DOS have abrupt solid-liquid phase

transitions, when condensed on silica substrates these alkylsilanes have much smoother

phase transitions, but never achieve the extreme levels of order and disorder found in the

true crystalline and liquid states. These results are in good agreement with the

temperature-dependent order behavior of alkylsilane stationary phases from the IR work

discussed above.Similar results were also reported by Doyle et al,"^ for 3.52 and

2.43 (.imol/m' Cis phases studied by Raman spectroscopy.

In addition to temperature studies, Doyle et al.'^^ investigated the effects of

solvents on the order of the 3 .52 jimol/m^ Macroporasil phase. The conformational order

66

of this phase appeared to be unaffected by the presence of chloroform, methanol and

acetonitrile in these experiments. However, these experiments were done in the presence

of hydrogenated solvents which lead to significant spectral interferences in the v(C-H),

v(C-C), and S(CH-H) regions. The conclusion about solvent effects is made based on

qualitative observations that Raman spectral features do not change in the presence of

different solvents, for solvent-subtracted Raman spectra. Given the sensitivity of the

order indicators in the Raman spectra and the subtle differences in conformational order

that can be measured very precisely,' it is likely that conformational order

differences exist for phases in solvents studied by Doyle, but these spectral differences

are lost in the background subtraction step.

Goals of this Research

The goals of this work were focused on studying interactions at the

chromatographic interface in order to gain some insight into a molecular understanding of

retention in RPLC Interfacial interactions were studied using Raman spectroscopy to

probe the structure of the stationary phase. The conformational order of the stationary

phase was determined as a function of numerous chromatographic parameters and were

used as an indicator of interactions within the stationary phase or with solvent or solute

molecules. The work presented in this Dissertation is centered primarily around the

investigation of alkylsilane-based reversed-phase systems, but preliminary work done on

67

the characterization of interactions of other chromatographic systems will also be

presented.

Specific goals of this work included:

1.1 Develop a detailed understanding of how Raman spectral indicators of order

relate to specific physical changes in the structure of alkanes

1.2 Use alkyl chain conformational order information in Raman spectral data to

determine the effect of alkylsilane surface coverage on the order and

chromatographic behavior of high-density alkylsilane stationary phases

1.3 Determine the temperature-dependent phase behavior for a series of high-density

alkylsilane stationary phases using Raman spectral indicators

1.4 Elucidate the effects of solvent on the order and structure of alkylsilane phases

and relate these effects to chromatographic behavior

1.5 Determine the effect of binary solute/mobile phase systems on the conformational

order of alkylsilane phases and correlate this molecular information to the

retention behavior of specific analjês

1.6 Extend the use of Raman spectroscopy toward the understanding of fundamental

interactions of other chromatographic systems including anion and cation

chromatographic supports

1.7 Develop improved strategies for stationary phase fabrication using relatively

simple synthetic approaches.

68

Chapter 2

Experimental

This chapter provides a detailed description of the instrumentation, materials and

methodology used in the work presented in this Dissertation. Additional experimental

details are outlined in each chapter as deemed appropriate.

Instrumentation

Raman Spectroscopic Instrumentation

Figure 2.1 shows the block diagram of the system used to acquire the Raman

spectral data reported in this Dissertation. All Raman spectral data were acquired with a

Spex 1877 Triplemate Spectrograph with a 600 gr/mm grating in the filter stage and a

1200 gr/mm grating in the spectrograph stage.

Raman scattered photons are generated using 100 mW of 532-nm radiation from a

Coherent Verdi Nd:YVO4 laser (Coherent, Inc.) Laser power at the sample was

measured using a model 1815-C optical power meter (Newport Corporation). Laser light

was polarized peipendicular with respect to the plane of incidence using a Fresnel Rhomb

model FR-2C425 (CVI Laser Corp.) Low laser powers were used to prevent sample

heating.

f X

Coherent Verdi Nd:YV04 Laser

( X „ 5 3 2 n m )

Polarizer

CCD Camera

Spex Triplemate 1877

Spectrometer

Sample Stage

Minolta' Polarization Collection Lens Scrambler

Figure 2.1 Schematic of Raman spectroscopic instrumentation. ON \o

Samples sealed in 5.0-mm-dia NMR tubes were positioned in front of the laser

using a copper sample mount. A temperature-controlled medium (50:50 ethylene

glycol/water [v/v]) was flowed through the copper mount using a model RTE-110

temperature controller (Neslab Corp.)

Raman scattered photons were collected perpendicular to the incident beam and

focused onto the entrance slit of the monochromator using a Minolta f/1.2 cameral lens.

A polarization scrambler was used to correct for the different transmission efficiencies of

parallel and perpendicularly polarized light through the spectrometer.

The Triplemate spectrometer has three slits. The entrance slit was set to 0.5 mm

and the filter stage slit was 7.0 mm for all experiments. The spectrograph slit of the

Triplemate was either 150 or 25 t4,m, corresponding to a spectral band-pass of 5.3 and

0.97 cm"\ respectively, when the spectrometer was centered at 2900 cm"\ Settings of the

spectrograph slit are given in each chapter.

The detector in these experiments was a charged-coupied device (CCD) system

based on a thinned, back-illuminated, antireflection-coated RTEl 10-PB CCD of pixel

format 1100 x 330 (Princeton Instruments), cooled with liquid Nj to -90 °C. Images fi"om

these detectors were processed first in WinSpec32 software (Roper Scientific) and were

imported into Grams32 (Galactic Industries, Inc.) for further data manipulation, including

calibration, background subtraction and curve-fitting. Raw spectral data were calibrated

externally using Ne lines.

71

High Pert'ormance Liquid Chromatography Instrumentation

Chromatograms shown in Chapter 13 were acquired using a BAS model 200B

liquid chromatograph with a BAS 116a IJV-vis detector at 254 nm (Bioanalytical

Systems, inc.), shown in Figure 2.2. The BAS 200B was converted to a microbore LC

system to accommodate low flow rates by using narrow tubing (0.007 to 0.025 in) for

connections after the injection port and by using a splitter to direct only 10% of the

mobile phase flow into the analytical microbore column. The remaining 90% of the

mobile phase flowed through a secondaiy back-pressure Cis column, allowing for

moderate flow rates to be used (0.5-1.0 mL/min) at reasonable column pressure (2000-

3000 psi). The injection volume for all experiments was 10 jiL.

The BAS 200B liquid chromatograph was controlled using the BAS

Chromatgraphy™ Control Software (Bioanalytical Systems). All chromatograms were

analyzed using the BAS Chromatgraphy™ Report Software (Bioanalytical Systems). No

baseline corrections were made to any of the chromatograms shown in this Dissertation.

Chromatograms shown for the NIST-CIH stationary phases in Chapter 10 were

acquired by Dr. Lane C. Sander at the Nation Institute of Standards and Technology

(NIST, Gaithersburg, MD). The liquid chromatograph consisted of a reciprocating piston

pump with an autosampler and a multi-wavelength ultraviolet detector. The injection

volume for these experiments was 5 |iL.

Splitter

Back-Pressure Column \ To waste

Rheodyne

Injection Port

Analytical C jg

Microbore Column

Solvent

Reservoirs

Dual Reciprocating j

Piston Pump Detector (>..„ = 254 nm .

Figure 2.2 Schematic of the BAS 200B Liquid Chromatograph.

to

73

Scanning Electron Microscopy

SEM images presented in this Dissertation were acquired with a Hitachi S-2460N

variable pressure scanning electron microscope. All images were acquired using an

electron accelerating voltage of 18 kV and a condenser lens setting of 9. Samples were

mounted on sample stubs and were sputter coated with Au/Pd using the Hummer IV

Sputtering System (Anatech, Ltd.) to minimize sample charging. Images were digitized

using Emispec Digital Imaging software, but are otherwise unmodified.

Elemental Analysis

The total carbon content of organically-modified silica stationary phases was

determined using a Perkin Elmer 2400 Series IICHNS Elemental Analyzer (Perkin

Elmer); experiments were conducted by Desert Analytics (Tucson, AZ).

Materials

All materials were used as received unless stated otherwise. Perdeuterated

solvents water-da (99.9%), 1-butanol-dio (>99%), and anisole-ds-methyl (99%) were

purchased from Aldrich. Perdeuterated solvents methanol-d4 (99.8%), acetonitrile-da

(99.8%), ethanol-dc, (99%), acetic acid-d4 (99.5%), chloroform-d (99.8%), hexane-dw

(99%), acetone-dg (99.9%), tetrahydro&ran-dg (THF) (99.5%), benzene-de (99.5%),

toluene-ds (99.8%), />xylene-di() (98%), aniline-dy (98%), and pyridine-ds (99.5%) were

74

purchased from Cambridge Isotopes. Perdeuterated benzylamine-d? (98%) was

purchased from CDN Isotopes.

Ethanol (200 proof, absolute) was purchased from Aaper Alcohol and Chemical

Company. Methanol (100,0%) was purchased from J.T. Baker. Pentane (99.5%) was

purchased from Burdick and Jackson. Cyclohexane (99.7%) was purchased from Fischer

Chemical. Acetonitrile (99.8%) and tetrahydrofuran (THF) (>99.9%) were purchased

from EM Scientific. Toluene (100.0%), acetone (99.7%) and ethylene glycol (99%) were

purchased from Mallinckrodt. Octadecane (99%), benzoic acid (99.5%), aniline (99%),

anisole (99%), benzene (>99.9%), ethylbenzene (99.8%), isopropylbenzene (cumene)

(99%), 1 -octanol (99%), octadecyltrimethoxysilane (C18-TM0S, >90%), and

tetramethoxysilane (TMOS, 99%) were purchased from Aldrich. Low density, linear

polyethylene (weight-average molecular weight, Mw - 4000, number-average molecular

weight, Mn = 17000, d = 0.920 g/mL) was purchased from Aldrich,

Methanesulfonic acid (>90%) and KOH (85%) were purchased from Aldrich,

HCl, NH4OH, and H2SO4 were purchased from EM Science, HNO3 was purchased from

Malinckrodt Inc,

CsCl (99,9%) and (CH3)4NC104 (99%) were purchased from Alfa Aesar,

NaH2P04*H20 (99%) was purchased from Curtin Matheson Scientific, Inc, NaNOj

(>99%), NaCl (99.8%), and Nal (>99%) were purchased from Malinckrodt Inc, NaBr

(>99%), KCl (>99%), NaHS04®4H20 (98%), NaC104 (98%), (CH3)4NBr (>99%),

75

(CH3)4N1 (>99%), (CH3)4NC1 (>98%), (CH2CH3)4NBr (98%), (CH3)4NHS04 (>99%),

(CH3)4NN03 (>99%), and CH3SO4H (99%) were purchased from Aldrich.

Standard Reference Material (SRM) 869a (Column Selectivity Test Mixture for

Liquid Chromatography) was purchased from the National Institute of Standards and

Technology (NIST, Gaithersburg, MD).

Water was purified with a reverse osmosis (Milli-RO 10 Plus) system and then

further purified with a Milli-Q UV Plus system (Millipore Corp.) Final resistivity of the

ultrapure water was 18.2 MQ/cm, and the total organic content was measured to be less

than 6 ppb with a Millipore AlO total organic carbon (TOC) detector.

[solute SPE Stationary^ Phases

Isolute SPE phases were purchased as bulk packings from International Sorbent

Technologies. These materials were fabricated on an unusual silica substrate of large

O particle size (-50-60 jim), small average pore diameter (—60 A) and high surface area

(-300-500 m^/g measured by BET analysis) relative to silica substrates most commonly

used as supports for stationary phases. The properties of these phases are given in

Table 2.1.

Isolute C18MF (referred to as C18MF) is a monomeric stationary' phase formed

from bonding dimethyloctadecylchlorosilanes to the silica substrate; a simple schematic

of this is shown in Figure 2.3. Isolute CM is based on a cyanopropyltrichlorosilane

76

Table 2.1. Properties of Isolute stationary phases

Isolute Phase"

Particle Diameter

(|im)

Surface Area

(m^/g)

Pore Diameter

(A)

Surface Coverage (jimol/m^)

Capacity (mmol/g)

C18MF 63 300 60 1.9 NA'

CN 40-70 NR^ 60 1,2 NA

Phenyl 40-70 NR 60 1.0 NA

SAX 40 521 60 NA 0.6

SCX-2 62 521 54 NA 0.4

"MF = monofiuictional silane precursor; CN = cyano-terminated silane precursor; Phenyl = phenyl-terminated silane precursor; SAX = strong cation exchanger; SCX-2 = strong cation exchanger. ^ôt Reported. °Not Applicable.

H j C y a

Si. -CH^ \

IT

Figure 2.3 Dimethyloctadecylchlorosilane and a condensed monolayer of the monotunctional silane..

•-J <1

bonded to the silica substrate, as shown in Figure 2.4. Isolate Phenyl is made by reacting

phenyltrichlorosilane with the silica substrate, as shown in Figure 2.5.

Isolate SAX (referred to as SAX) is a strong anion exchange stationary phase

based on trichlorosilylpropyl-trimethylammonium chloride-modified irregular silica

particles, shown in Figure 2.6. Isolute SCX-2 (referred to SCX-2) is a strong cation

exchange stationary phase fabricated using trichlorosilylpropylsulfonic acid to

functionalize the silica substrate, shown in Figure 2.7.

Despite the relatively high surface area of this substrate, previous research has

shown a large fraction of the pore volume to be inaccessible to solvent or analyte

molecules.^^ Therefore, the reported surface coverage and capacity values of these

phases, determined by using the total surface area, are likely to be underestimates of the

effective surface coverage and capacity.

Kromasil Silica and Bulk CIR Stationary Phase

Kromasil silicas and stationary phases were purchased in bulk from Eka

Chemicals (Akzo Nobel). The physical properties of these materials, including particle

size, surface area, porosity, and surface coverage are given in Table 2.2.

Kromasil silica particles, SIL-5-100 (referred to as Krom-100) and SIL-5-200

(referred to as Krom-200), were used as substrates for alkylsilane stationary phases

fabricated by procedures described in Chapter 13. These particles are spherical and

regular in shape, with a narrow size dispersity. C18-5-100 (referred to as Kromasil Cn)

Si—CI CI / CI

OH

Si—OH

Figure 2.4 Cyanopropyltrichlorosilane and a condensed monolayer of cyanopropyltrichlorosilane on the 1ST CN surface.

-J

CI Si

CI

CI

.OH

Si-O I O I HO' I "O / o q O \

/

Figure 2.5 PhenyltricMorosilane and a condensed monolayer of phenyltrichlorosilan on the 1ST

Phenyl surface.

00 O

Me^ Me CI.

Me

Cl-Si^ / n

CI

Me^

N + / -Me

HO-Si^

Me Me

Cl^ N + Cl^

Me

HO-Si

O / ^ OH

Trichlorosilylpropyltrimethylammonium chloride and a simple illustration of the SAX surface.

p- «* O" H 0= H

Cl-Si HO-Si^ °/'-a o' ° o' > o'

Trichlorosilylpropylsulfonic acid and a simple illustration of the SCX-2 surface

83

Table 2.2. Physical properties of Kromasil silicas and stationary phases

Kromasil Material Particle Diameter (|j,m)

Surface Area (m^/g)

Pore Diameter (A)

Surface Coverage (jj.mol/m^)

Kromasil SEv-5-100 5 297 100 NA'

Kromasil SIL-5-200 5 213 200 NA

Kxomasil C18-5-100 5 300 100 3.66

®Not Applicable.

84

is a monofunctional stationary phase modified using a dimethyloctadecylchlorosilane

precursor. Kromasil Cig was used as a standard material to test column packing

procedures and efficiency.

NIST Cis Stationary Phases

NIST Ci8 stationary phases were a gift from Dr. Lane C. Sander at the National

Institute of Standards and Technology (NIST). A variety of synthetic approaches are

used to fabricate this series of high-density phases, described by Sander and Wise,^'^

whose physical properties are given in Table 2.3.

All stationary phases were prepared using the same lot of YMC S1L-200-S3

spherical silica (YMC, Inc.) This substrate has a nominal particle size of 3 |im, average

pore diameter of 200 A, and surface area of 200 mVg determined by BET analysis.

The trifunctional phases were synthesized by bonding octadecyltrichlorosilanes to

the silica substrate as shown by the schematic in Figure 2.8. The difunctional phases

were prepared using methyloctadecyldichlorosilane precursors, given in Figure 2.9.

Solution polymerized phases were made by adding water to the silica slurry containing

di- or trifunctional silanes. The hydrolysis and condensation (i.e. polymerization) of

alkylsilanes began in solution followed by deposition onto the silica substrate. Surface

polymerized phases were prepared by adding water to dried silica, then adding the "wet"

silica to a solution containing di- or trifunctional silanes. Polymerization in this case is

thought to occur only at the silica surface. The monofunctional phase was prepared by

85

Table 2.3 Physical properties of NIST Cig stationary phases"

Stationary Phase Previous Phase Preparation'' % Carbon Surface Coverage Designation" [wt/wt] (limol/m^)

TFC18SF SI Surface Polymerization 19.84 6.45

TFC18SL PI Solution Polymerization 17.07 5.26

DFC18SF S3 Surface Polymerization 17.37 5.00

DFC18SL P2 Solution Polymerization 15.10 4.17

MFC18 Ml 12.45 3.09

^From references 2.3 and 2.4,

''Surface polymerization and solution polymerization described in the text; TF = trifunctional alkylsilane precursor, DF == di&nctional alkylsilane precursor, MF = monofimctional alkylsilane precursor.

Figure 2.8 Octadecyltrichlorosilane and a schematic of the condensed trifunctional silane monolayer.

CO C3\

CKVCH, H3C Si\

O '

HgC-Si^ /Ô

o I \

Figure 2.9 Methyloctadecyldichlorosilane and a schematic of the condensed difunctional monolayer.

00

88

reacting dimethyloctadecylchlorosilane with silica in the presence of 2,6-lutidine to

initiate the hydrolysis.

Methodology

Specific procedures that are related to a single type of experiment are given in the

appropriate chapter. The following list consists of the more commonly used

experimental procedures.

RPLC Column Packing

Stationary' phases were packed into 1.0 mm ID. x 1/16 in O.D. 316 stainless steel

columns cut to 10 mm in length. Columns were slurry packed in the downward direction

using a 316 stainless steel column packer, diagramed in Figure 2.10. The slurry reservoir

was a V2 in I.D. Swagelok T-fitting with V2 in fittings on the ends and a Va in fitting in the

center. One V2 in end was reduced to V* in using a series of tubing, filters and a reducing

union and was connected to the BAS 200B HPLC pump. The other Vi in end was capped

The 10 cm column tubing was connected to the % in end of the reservoir with a reducing

union. The end of the 10 cm tubing was capped with a 1/16 in dia stainless steel 0.2 |i,m

porous fiit allowing solvent to flow through the tubing while keeping stationary phase

particles in the column.

Columns were prepared using a modified packing protocol provided by Akzo

Nobel Chemical Co., and can be found on the web at www.kromasil.com. Columns were

http://www.kromasil.com

Methanol push

solvent from

HPLC pump

Sluny reservoir

1 mm ID. X 1/16 in O.D

stainless steel tube

(RPLC Column)

1/16 in. dia. stainless

steel 0.2 }im frit 1/16 in. union w/ 0.2 jim frit

To solvent

waste

Cutaway view of the union

Figure 2.10 Illustration of the RPLC column packer and a cut-away view of the reducing union fitted with a porous frit.

90

packed at 3200 psi (+200 psi) for 15 to 30 min by pushing methanol from the HPLC

solvent reservoir. The flow rate was monitored and adjusted continuously for the

duration of the packing procedure in order to maintain a relatively constant packing

pressure. The slurry concentrations used were ca. 45-55 mg/mL of stationary' phase in

90:10 THF/isopropanol (IPA, v/v). The slurry was constantly stirred throughout the

packing procedure to prevent the stationary phase from settling out of suspension.

After the columns were packed, the solvent flow stopped, and the pressure

decreased to 0 psi, columns were disconnected from the packer and capped with a 1/16 in

dia stainless steel 0.2 p,m porous frit. Columns were equilibrated in 80:20

methanol/water for ca. 3 hrs immediately following column packing.

Column eflBciency was determined by calculating the number of theoretical plates

for a test mixture of monosubstituted aromatic analjês, benzene, ethylbenzene, and

cumene. Analytes were separated using an 80:20 methanol/water mobile phase and a 0.7

p.L/min flow rate.

Surface Coverage Determination

Surface coverage calculations for stationary phases were made based on the

carbon content of each phase by Desert Anahlics Inc. The weight percentage of carbon

(%C) of these phases was corrected for the carbonaceous contamination of the

unmodified silica substrates by substraction. With the corrected values of %C, values for

alkylsilane surface coverages ([.imol/m^) were calculated using equation 2.1, developed

by Berendsen and de Galan^

r (limol/m^) = (l/SA) * 10''%C / [1200no - Po(MW - 1)] 2.1

where SA is the surface area of the silica substrate in units of mVg, nc in the number of

carbons in an individual bonded alkylsilane molecule, and MW is the molecular weight

of the bonded alkylsilane. For all stationary phases fabricated as described in Chapter 13,

the alkylsilane precursor used was octadecyltrimethoxysilane. Since the extent of

polymerization was unknown for these phases, the molecular weight used in this surface

coverage calculation was that of the surface-bound hydrolyzed precursor

CHj(CH2)i7Si(0H)20- (MW = 331), described by Sander and Wise.^ ^'

RPLC Column Selectivitv Determination

SEM 869a contains a mixture of three polycyclic aromatic hydrocarbons (PAHs),

benzo[a]pyrene (BaP, planar), phenanthro[3,4-c]phenanthrene (PhPh, nonplanar), and

1,2:3,4:5,6:7,8-tetrabenzonaphthalene (TEN, nonplanar) in 85:15 acetomtrile/water; the

chemical structures and geometric shapes of these PAHs are shown in Figure 2.11. The

elution order of these PAHs had been shown to correlate with stationary phase shape

recognition ability and allowed phases to be categorized as polymeric (highly shape

selective) or monomeric (poor shape selectivity) as given in Figure 2.12.^''^'^'^ The

Benzo[a]pyrene (BaP)

/ %

w

Phenanthro[3,4-c]phenanthrene (PhPh)

1,2:3,4:5,6:7,8-Tetrabenzonaphthalene (TBN)

Figure 2.11 Chemical structures and geometric shapes o f PAHs used to determine RPLC column shape selectivity.

93

TBN PhPh

BaP Monomeric a = 2.08

TBN PhPh + BaP

a = L80 Monomeric

TBN PhPh

I BaP a - L27 Intermediate

o tzi

TBN + BaP

PhPh Polymeric

TBN

PhPh BaP a = 0.48 Polymeric

Time

Figure 2.12 Chromatograms of PAHs used to determine shape selectivity adapted from reference 2 .8.

94

selectivity coefficient, AXSN/BAP (ratio of FTBN/ A'&IP), has been used as a numeric

descriptor of shape selectivity. For axBN/BaP < 1, Cig columns are considered to be

polymeric and for axBN/BaP >1.7, Cig columns are considered to be monomeric. For the

monomeric columns, BaP elutes before TBN, the retention of BaP increases relative to

TBN as the columns become more polymeric, and the elution order is reversed for

polymeric columns.

Spectral Analvsis

All Raman spectra were analyzed and manipulated using Grams32 software

(Galactic Industries, Inc.) Raw spectral data were in units of pixels along the x-axis and

were converted to wavenumbers (cm"^) using Ne emission lines as a calibration source.

All Raman spectra were superimposed on a relatively large background that is a

complex convolution of spectrometer response, detector response, sample fluorescence,

scattering characteristics of the sample, and shot noise. These backgrounds were

corrected using a multi-point linear fit over the spectral region of interest. An example of

a typical background correction sequence is given in Figure 2.13. A raw spectrum of

DFC18SL in anisole-d? at -15 °C in the C-H stretching region between 2400 and 3400

cm"' is shown in Figure 2.13a. In this case, a 5-point-linear fit was used to correct the

background with points chosen such that they did not interfere with the envelope of

vibrational modes, shown in Figure 2.13b. The result was a linear, horizontal

95

200000

2500 2700 2900 3100

Wavenumbers (cm"')

3300

20000

2900 2500 2700 3100 3300

Wavenumbers (cm'^)

10000 c

•

2800 3000

Wavenumbers (cm'' )

Figure 2.13 (a) calibrated Raman spectra of DFC18SL, (b) illustration of the 5-point fit used to baseline correct Raman spectra in the v(C-H) region, and (c) background corrected Raman spectra of DFC18SL in anisole at 15 "C.

96

background of the C-H stretching region for DFC18SL in anisole-ds at-15 °C presented

in only the information-containing region of 2750 to 3050 cm'\ Figure 2.13c.

Analysis of Raman spectral data including measuring peak intensities, peak

frequencies, full width at half maximum (FWHM), peak asymmetry factors, and curve

fitting were done on the calibrated, background corrected spectra using Grams 32.

Molecular Models

Models of the solvent/stationar>' phase interface were constructed using

ChemDraw 3D (CambridgeSoft). Alkyl chains of the stationary phase were arranged in

either 5 x 2 or 5 x 3 matrices; the spacing of the chains was determined from the

relationship between the average Si-O-Si spacing in the silica matrix (~ 3.2 and the

theoretical bonding density of close-packed alkyl chains on a silica substrate (~ 8

2 2 3 2 1 1 2 }imol/m ). • ' For example, a 4.00 nmol/m coverage material is one-half the coverage

of the theoretical limit. Therefore, the alkylsilane spacing of a homogeneously-

distributed 4.00 [imol/m" material should be twice the silica Si-O-Si spacing at the

theoretical limit, or 6.4 A. For the spacing between di- and monofunctional alkylsilanes

containing Si-bound methyl groups, the minimum distance between alkyl chains was

assigned as the diameter of the methyl groups in both dimensions in order to account for

all possible orientations of the methyl groups. The silica surface was drawn uneven and

nonlinear to mimic the uneven surface of silica particles; however, the alkylsilanes in the

97

model do not conform to this surface structure, but are modeled on essentially flat

surfaces.

Placement of solvent molecules at the stationary phase interface in these pictures

depends on solvent properties, stationary phase packing density, and the overall quality of

the final energy-minimized picture. Strongly hydrophobic solvent molecules are

generally placed deeper (vertically) into the alkyl matrix than highly polar solvent

molecules for materials where there is sufficient distance between alkyl chains. If the

initial placement of solvent molecules is too far away from the alkyl chains, the resulting

minimum energy structure is one where the solvent molecules are several molecular

dimensions removed from the aJkylsilane in either the vertical or horizontal direction. In

this case, where the chemical feasibility of the final picture was unrealistic, the image

was discarded and other choices of solvent placement were made until the chemical

credibility of the final picture improved. Molecular models were energy-minimized

using the modified MM2 force field computation provided by ChemDraw 3D to a

precision of 0.1 kcal/mol. This computation employed numerous chemical parameters

including bond stretch energy, angle bend energy, torsion and nonbonding constraints, n-

system calculations, charge and dipole terms, and cut-off parameters for van der Waals

and electrostatic interactions. During the energy minimization computation, the silicon

atoms, oxygen atoms, and the five proximal methylene groups on each alkyl chain were

fixed in space, simulating a surface-confined, densely-packed array of alkyl moieties. It

is important to note that while the models presented here were energy-minimized by the

98

modified MM2 computation, they were used only to develop a less arbitrary picture of

the chromatographic interface and to provide a framework for visualizing the relevant

intermolecular interactions reported by the Raman spectral data.

99

Chapter 3

Quantitative Correlation of Raman Spectral Indicators In Determining

Conformational Order in Alkyl Chains

Introduction

Materials based on alkane-containing molecules are used in countless

technological applications including self-assembled monolayers on metal and oxide

surfaces (both two- and three-dimensional), ' "' biological membranes, ' '

3 7 3 8 3 9 3 1 4 biosensors, ' ' ' Langmuir-Blodgett films, ' ' ' chemically-modified silicas for

chromatography,^'""and polymeric alkane materials.^ The desired attributes of

alkane chain conformational order in these systems vary substantially Irom highly

ordered (crystalline) to disordered (liquid-like) and are application-dependent. Alkane

chain structure and interfacial properties dictate the function and utility of these

materials. For example, biological membranes allow the transport of ions, proteins and

other species from one side of the membrane to the other, as well as the lateral difilision

of membrane components within the bilayer. Thus, the alkane chains in the lipids of

biological membranes must be partially liquid-like to allow this transport and lateral

difilision to occur. Alkane chains of a more rigid, crystalline character would restrict the

transport of molecules across the membrane, while a membrane consisting of completely

liquid alkane chains would lack the stability to remain intact.

100

Of primary importance to the majority of the work presented in this Dissertation,

was the effect of conformational order of alkylsilane stationary phases on separation

"X '7A "Z 'y'i selectivity and performance in RPLC. " "' For example, highly ordered, high-density

stationary phase materials have been shown to have shape selectivity for planar

polynuclear aromatic hydrocarbons, while less ordered stationary phase materials exhibit

no geometric selectivity. ''^"^ Clearly, describing alkane chain conformational order on the

molecular level is crucial for understanding how such materials function in their

respective applications.

Numerous experimental techniques have been used to study alkane-chain

structure and conformation including nuclear magnetic resonance (NMR)

spectroscopy;'' Fourier transform-infrared spectroscopy (FTIR),and

Raman spectroscopy.^^'' "'""" NMR experiments specifically probe alkane

chain mobility allowing information about conformational order to be inferred. Although

a distinct correlation between chain mobility and conformational order exists, a direct

measure of conformational order is necessary for understanding the subtleties of alkane

structure and function. FTIR spectroscopy does offer direct conformational order

information for alkane-based systems, an advantage over NMR spectroscopy where alkyl

chain mobility is inferred from spectral data. However, FTIR suffers from major

limitations in spectral interference from solvents, resolution, activity of alkane vibrational

101

modes, and experimental sensitivity, and proves to be inadequate for studying subtle

molecular changes in conformational order.^

In contrast, Raman spectroscopy has been shown to be a powerlul experimental

tool for determining subtle conformational changes in a wide variety of alkane-based

systems.Such conformational order information is usefiil in

understanding the chemical environment of alkane chains and provides insight into the

molecular interactions between alkane chains and their surroundings. Raman

spectroscopy has the unique ability to provide direct information about conformational

order, non-destructively and tree irom spectral interferences that plague FTIR

spectroscopy. ' ^ In addition, the plethora of spectral bands in the v(C-H), v(C-C) and

6(C-H) regions of the Raman spectrum provide a great deal of information about

conformational order including the extent of alkane chain coupling, intramolecular

motion, relative numbers of gauche and trans conformers, and chain twisting and

bending.

Larsson and Rand first described the utility of Raman spectroscopy for

determining alkane chain conformational order in lipids and lipid-protein complexes.

They noted that changes in spectral bands near 1100 cm"' (later assigned to the V(C-C)G

and V(C-C)T modes) correlate to relative numbers of gauche and trans conformers in

alkane chains and that the intensity ratio of the 2885 cm' VaCCHi) band to the 2850 cm"'

Vs(CH2) band, I [ v'a(CH2)]/! [ VsCCHi) j, is a measure of lateral packing density of alkane

chains, and is itself an indication of order. Upon examining the relationship between

102

I[VA(CH2)]/I[VS(CH2)] and the intensity ratio of the V(C-C)G to the v(C-C) i modes near

1100 cm"', I[V(C-C)G]/I[V(C-C)T], they concluded that the former is more sensitive than

the latter to subtle changes in order both for disordered molecules and for molecules in an

all-trans configuration.

It is important to note that the ratio I[Va(CH2)]/I[Vs(CH2)] measures different

quantities for all-trans and disordered alkanes.^'"^^ In the crystalline phase, the

I[ Va(CH2)]/I [ VSCCHj)] is a measure of rotational disorder in all-trans alkane chains, and in

disordered, liquid-like alkanes, this ratio is additionally sensitive to the addition of

gauche con formers. Thus, the I[Va(CH2)]/I[Vs(CH2)] is a subtle and sensitive indicator of

order for ciystalline alkane chains.

Larsson and Rand also noted an increase in intensity of the 2930 cm"^ VS(CHOIR

mode for increasingly disordered alkane assemblies. The intensity ratio of this mode to

the 2850 cm"' Vs(CH2) mode was later correlated to intermolecular coupling of alkane

chains by Snyder et al.' '^

In addition to the work done by Larsson and Rand on lipid-protein complexes,

Raman spectral indicators were also used extensively in the late 1960's and early 1970's

to examine the conformational order of alkane chains in model systems including

'3E ^ *2 /B "2 /FLIC '3 ^*2 polymers and biological membranes. ' ' ' ' • ' ' More recently, these indicators have

been used to describe the conformational order of more complex alkane assemblies

similar to those mentioned above including self-assembled monolayers on a variety of

substrates,^''^^'^'"^^'^''^^'^'^^ Langmuir-Blodgett layersând chemically modified-silicas for

103

reversed-phase liquid chromatography ' Even though discrete differences between

bulk or solution phase alkane systems and the surface-bound systems exist,

intermolecular hydrophobic interactions between alkane chains are similar. Therefore,

the conformational order behavior of surface-confined alkane systems is complementary

to that of bulk or solution phase materials when probed by Raman spectroscopy.

Goals of This Work

Despite the valuable work that had been done in an effort to understand the

relationships between these Raman spectral indicators of conformational order, very little

had been done to quantitatively correlate these indicators with the goal of understanding

at what degree of disorder each parameter changes. The goal of this work was to acquire

Raman spectral data for the temperature-induced disordering of octadecane and a low

molecular weight polyethylene. Detailed analysis of these data would allow for the

quantitative correlation of numerous conformational order indicators in the v(C-C), 6(C-

H) and v(C-H) spectral regions. These correlations would give a unique understanding of

how conformational changes in alkanes are manifested throughout the entire Raman

spectrum. Information obtained from this work will be used throughout this Dissertation

to describe the order and structure of alkylsilane-based stationary phases under

chroniatographically relevant conditions.

104

Methodology

Octadecane and low molecular weight polyethylene (average Mw ~ 4,000; average

Mn -1,700) were obtained from Aldrich and used as received.

Raman spectra were collected using 100 mW of 532 nm radiation from a

Coherent Verdi Nd; YVO4 laser on a Spex Triplemate spectrograph, described in Chapter

2. Slit settings were 0.5/7.0/0.150 mm for all experiments. Samples were sealed in 0.5-

mm-dia NMR tubes and positioned in the laser beam using a copper sample mount. The

temperature of the copper mount was regulated by circulating 50:50 ethylene

glycol: water through it and around the sample tube using a Neslab NTE-110 temperature

controller, allowing temperature control from -15 to 110 "C. Temperatures greater than

110 "C were regulated by resistively heating the copper sample holder. Samples were

allowed to equilibrate at each temperature for a minimum of 30 min prior to analysis. A

minimum of three measurements were made at each temperature on samples

independently prepared from the same batch of material.

Raman spectral decomposition for polyethylene in the 6(CH-H) region was done

using the curve-fitting program in the Grams32 software. Decompositions were done

using a 100% Gaussian peak shape.

105

General Overview of Alkane Conformational Order for Octadecane and Polyethylene

from Raman Spectroscopy

Raman spectra of neat octadecane at 0 and 60 °C are shown in Figure 3.1 for two

frequency regions, 1000 to 1500 cm"' and 2750 to 3050 cm"'. Raman spectra of neat

octadecane at other temperatures from 10 to 106 "C are shown in Appendix 3.1. The

corresponding peak frequencies are reported in Table 3.1. Ten unique Raman spectral

indicators had been identified in the v(C-H), 5(CH-H) and v(C-C) regions that reflect

structural order in alkane chains: the peak intensity ratios I[Va(CH2)]/I[vs(CH2)],

I[VS(CH2)FR]/I[VS(CH2)], and I[V(C-C)G]/I[V(C-C)T]; the integrated peak area ratios

A[5(CH2) + 5(CH3)]/A[5(CH2)] and A[5(CH2)ORTHO]/{A[5(CH2) + OCCH:,)] +

A[§(CH2)]}; the peak frequencies of the Va(CH2), Vs(CH2), and X(CH2) modes; the

asymmetry of the x(CH2) mode; and the full-width-at-half-maximum (FWHM) of the

X(CH2) mode. All of these were investigated here to elucidate the subtleties of alkane

structure upon melting to provide a definitive molecular picture of this phenomenon.

Although the v(C-H) region (2750 to 3050 cm"^) is complex for alkanes as a result

of numerous spectral bands and Fermi resonance splittings, conformational order

information can be obtained empirically from the peak intensity ratio

I[Va(CH2)]/I[Vs(CH2)]. This ratio spans a range from -0.6 to 2,0 for normal alkanes in the

liquid (0.6-0.9) and crystalline (1.6-2.0) states. ' This ratio is sensitive to subtle

changes in conformational order from rotations, kinks, twists and bends of the alkane

chains; essentially, this ratio changes for any deviation from a perfect all-trans

v(C-C). v(C-C)

400000 100000 HH

v(C~C),

2900 1000 1100 1200 1300 1400

Wavenumbers (cm"^)

2800 3000

Figure 3.1 Raman spectra in the v(C-C) and v(C-H) regions for octadecane at (a) 0 "C and (b) 60 "C. Integration time 2 min.

107

Table 3.1. Raman peak frequencies (cm"') and assignments for octadecane and polyethylene.

Octadecane Liquid

Octadecane Crystalline

Polyethylene Liquid

Polyethylene Crystalline

Assignment"

1065

1081

1124

1303

1369

1440

1450

2852

2888

2925

2935

2965

1062

1128

1296

1371

1441

1445

2842

2876

2925

2935

2959

1065

1081

1124

1302

1369

1440

1450

2852

2886

2925

2935

2965

1062

1128

1295

1371

1420

1441

1445

2847

2881

2925

2935

2960

V(C-C)T

V(C-C)G

v(C-C),

TCCHJ)

5(CH:0

8(CH2)ORTHO

5(CH2)

SaCCHj)

VS(CH2)

Va(CH2)

VS(CH2)IR

VS(CH3)FR

VACCHS)

V = stretch; x = twist, 8 = bend and/or scissor.

108

configuration. Schematic illustrations of alkane chain rotations, chain conformers,

and chain coupling are shown in Figures 3.2-3.4. Despite the extensive previous

investigation and use of this intensity ratio as a measure of alkane order/disorder, no

reports have appeared in which this ratio is quantitatively correlated to other chain

coupling, twisting, bending or gauche conformer spectral indicators of order.

Other spectral indicators of alkane order are also useful. The frequencies at which

the v'sCCHs) and VafCHi) modes are observed reflect conformational order and interchain

coupling; these modes are observed at -2853 and 2888 cm"', respectively, for octadecane

i n t h e l i q u i d s t a t e , a n d d e c r e a s e b y 6 t o 8 c m " ' i n t h e c r y s t a l l i n e s t a t e . T h e

frequency of the x(CH2) mode at -1300 cm"' has also been related to the degree of

coupling between alkane chains, with decoupling correlated with an increase in frequency

of this mode. Chain coupling information can also be obtained from the intensity ratio of

the VS(CHOi r at -2930 cm"^ to the VsCCHj) at -2850. ' "^® As the chains decouple

(intermolecular interactions decrease), the terminal methyl groups experience increased

rotational and vibrational freedom, and the intensity ratio of the VS(CH3)FR to the Vs(CH2)

increases.

Conformational order information for alkanes can be obtained from vibrational

modes in the v(C-C) and 6(C-H) regions as well. The relative number of gauche and trans

conformers in alkane chains is indicated by the intensity ratio of the V(C-C)G (-1080 cm'

') to the V(C-C)T (-1062 cm"'). For octadecane, this intensity ratio spans a range from

zero in the crystalline state to 1.1-1.3 in the liquid state. Unlike modes in the v(C-H)

Trans

V

Gauche

c ft

' H 1

/-\ \

/ S» "i

A A >

/ /

Side View

.» Trans

i

i < Gauche 1 -> if

j«t îflJ

Top View

Figure 3.2 Schematic of alkyl chain trans and gauche conformers.

o

Trans Rotated

( > V

Side View

& AV

i

Trans

Rotated

Top View

Figure 3.3 Schematic of a subtle alkyi chain rotation.

o

Coupled Decoupled

- , \ \

—^ \ \

/ - , / /

\ . •„ "X • "A ••"'% \

f \ ' y

\ >

Figure 3.4 Schematic of alkyl chain decoupling.

112

region, the v(C-C) modes are only sensitive to true gauche and trans conformations and

3 5 5 not to methylene rotations or other bond deformations. '

Two other spectral envelopes of bands contain conformational order information,

those centered around 1300 and 1450 cm'\ The tCCHi) (-1300 cm"') is superimposed on

a number of modes including, but not limited to, the methylene wag, the H-C-H

deformation and other methylene deformation modes.The intensities of these modes

change from liquid to crystalline states of alkanes and are primarily sensitive to

conformer changes between gauche and trans. The intensity changes of the underlying

modes cause visible changes in the x(CH2), and can be quantified by changes in width,

asymmetry and frequency of this mode.

The spectral region near 1450 cm"' contains a number of overlapping deformation

modes including the symmetric methylene bend (SCCHj) at -1440 cm"') and the overlap

of the methylene scissor and the symmetric methyl bend (5(CH2) + 6(CH3) at -1460 cm"

An increase in the integrated peak area ratio of these two envelopes, designated as

A[S(CH2) + 6(CH3)]/A[5(CH2) ], is an indication of the increased population of methyl

and methylene groups free to undergo the scissor or bending motion, respectively.

Structurally, this denotes increased methyl and methylene intramolecular motion that

increases with alkane chain decoupling.

The phase transition for neat octadecane (melting point 29 °C) is abrupt and

occurs over an -1 "C range, typical for small molecules. As a result, the conformational

order information obtained near the phase transition is limited by a combination of the

113

kinetics o f the transition and the temperature resolution (0.5 "C ) of the experimental

system. Due to the difficulty in controlling the melting event at the phase transition,

spectra obtained during the phase transition reflect those for a biphasic mixture; thus, the

spectral response represents the average conformational state of liquid and solid phases

present in the biphasic mixture.

In order to facilitate greater understanding of more subtle changes in

conformational order throughout the phase transition, a low molecular weight

polyethylene (melting point ~98 °C) was studied as well. The phase transition for

polyethylene is more gradual, and conformational order is observed to change over a

temperature range of several degrees. Raman spectra of polyethylene at 0, 60, and 100

°C are shown for the v(C-C) and v(C-H)/6(CH-H) regions in Figure 3.5 with the

corresponding peak frequencies in Table 3.1. Additional Raman spectra of

p o l y t h e y t h y l e n e f r o m 1 0 t o 1 5 0 " C a r e s h o w n i n A p p e n d i x 3 . 1 .

Comparing these spectral data to those of octadecane, the most obvious difterence

is the presence of the 5(CH2)ORTHO mode at -1420 cm"' and the loss of intensity of the

6(CH2) mode at 1440 cm"' at low temperatures. The 5(CH2)ORTHO mode signifies an

orthorhombic crystal structure of polyethylene and results from crystal field splitting of

the 5(CH2) mode at 1440 cm"'."' The intensity of the S(CH2)ORTHO decreases with

increasing temperature as the volume unit cell of the crystal expands and finally

disappears in the liquid phase. As a result of this crystal field splitting, quantification of

ORTHO

1000 1100 1200 1300 1400 2800

Wavenumbers (cni'^)

Figure 3.5 Raman spectra in the v(C-C)) and v(C-H) regions for polyethylene at (a) 0 "C, (b) 74 "C and (c) 100 "C. Integration time 2 min.

115

the peaks in this region for polyethylene from curve fitting are presented as the sum of

the integrated peak areas of the 5(CH2)ORTHO and 5(CH2) bands.

Raman Spectroscopy of Octadecane

The correlation between I[va(CH2)]/I[Vs(CH2)] and I [v(C-C)G]/I[ v(C-C)T] is

shown in Figure 3.6. The plot is divided into liquid, biphasic and crystalline regions by

dashed lines. As noted above, the difficulty in temperature control in the phase transition

region renders these lines only an approximation of the phase boundary; thus, they are

intended only as a guide to the reader. The crystalline region spans a temperature range

from -15 to 28 "C, the biphasic region is at a temperature of ~29 "C, and the liquid region

is between 30 and 100 "C.

In the crystalline state, no gauche conformers exist in alkanes; however, some

aspect of conformational order changes with increasing temperature, as indicated by a

decrease in I[Va(CH2)]/I[Vs(CH2)] from 2.1 to 1.5. These changes are thought to be due to

methyl and methylene rotational disorder, predominantly near the ends of the alkane

chains. Further into the biphasic region, the average number of gauche conformers

increases, starting at a value of I[Va(CH2)]/I[Vs(CH2)] of-1.2 and continuing for

decreasing values of I[Va(CH2)]/I[Vg(CH2)] to 0.75. In the liquid phase, the extent of

deviation from the all-trans state remains constant (i.e. l[Va(CH2)]/I[Vs(CH2)] is constant),

while the relative number of gauche conformers increases with increasing thermal

energy. In this region, the rotational disorder responsible for the variance of

116

>30

_ H O I

u

u 6 X

0.6-

0.4

-0.2

Temperature ("C)

-29 <28

- Liquid

1

li 1 1 ii—B—1 1

Biphasic *

' Crystalline 1 1 1 1 I 1 1 1

1

1

1 1

I 1

1

H

• . • . T—'—1—^—1 r '

1 1 1 1 1 1 1

1 ! 1 • 1 ' I

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v(CH)]/I[v(CH)]

Figure 3.6 Intensity ratio of the V(C-C)G to the V(C-C)T as a function of the intensity ratio of the vJCH^) to the v/CHj) for octadecane. Error bars are included for each point, but may be smaller than the size of the points.

117

I[Va(CH2)]/I[Vs(CH2)] in the biphasic region is converted to true gauche conformers with

increasing temperature.

Alkane intermolecular interactions, specifically interchain coupling, for

octadecane as embodied in 1[VS(CH3)FR]/I[VS(CH2)] is correlated to I[Va(CH2)]/I[Vs(CH2)]

in Figure 3,7. Alkane chain coupling decreases with increasing rotational disorder

(I[Va(CH2)]/I[Vs(CH2)] decreases from 2.1 to 1.3). Here, alkane chains spread out

sufficiently that rotational disorder becomes extensive. At temperatures in the vicinity of

the melting temperature where I[Va(CH2)]/I[Vs(CH2)] attains values of 1.2 to 0.8, the

chains are sufficiently decoupled to allow gauche conformers to be formed. Decoupling

continues in the liquid state with increased thermal energy, although the rotational

disorder is at its maximum, and hence, the I[Va(CH2)]/I[Vs(CH2)] remains constant. These

changes are consistent with a change in molar volume of the liquid alkane as a function

of increasing temperature. That such a change is indeed occurring can be seen from the

plot in Figure 3.8 of the molar volume for octadecane in the liquid state^'*^ as a function

of the spectral indicator for interchain decoupling, I[VS(CH3)FR]/1[VS(CH2)]. In this plot,

the molar volume increases only slightly over the solid/liquid phase boundary, but then

increases dramatically in the liquid phase.

Alkane chain coupling information can also be obtained from the peak frequency

of VS(CH2); this parameter is correlated to I[Va(CH2)]/I[Vs(CH2)] in Figure 3.9. The

Va(CH2) and x(CH2) peak frequencies are also indicators of interchain coupling and are

correlated to 1[Va(CH2)]/I[Vs(CH2)] in Figures 3.10 and 3.11. Although the v(CH2) peak

118

Temperature (°C) >30 -29 <28

0 . 8 -

u

0.7-

RYI

0.6-

U 0.5-C<9

0.4

Biphasic Crystalline

0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 I[v(CH)]/I[v(CHJ]

Figure 3.7 Intensity ratio of the to the v.CCHj) as a function of the intensity ratio of the v/CHj) to the v/CHj) for octadecane.

119

Temperature ("C) 30 100

o 340

0.5 0.6 0.7

I[vXCH3)J/I[V/CH^)]

Figure 3.8 Molar volume (mL/mole) of octadecane as a function of the intensity ratio of the v^CCHj) to the v^fCHj).

120

2854 >30

J-

Liquid

2852 •;

Temperature ("C) -29

I

Biphasic

2850

g 2848 cr

£ 2846-

2844

<28 i_

Crystalline

P. 2842-

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v.(CH,)]/I[v.(CH,)]

Figure 3.9 The v/CHj) peak frequency as a function of the intensity ratio of the v.,(CH2) to the Vs(CH2) for octadecane.

2890

2888

2886

2884 0

1 2882 o"

^ 2880

g 2878 PH

2876 U

2874

2872 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v/CH^)]/I[v/CH^)]

Temperature ( C) >30 -29 <28 ___J 1 L

-Liquid

|i

Biphasic Crystalline

- • :

"i** 1

1

i_p_i

1

1

1 . j . 1 1 1 I 1 1 1 y 1 1

•hh

1 1 , 1

Figure 3.10 v/CHj) peak frequency as a function of the intensity ratio of the v^CCHj) to the v^tCHj) for octadecane.

1304-

1302

1300

1298

1296

1294'

>30 1

Temperature (°C) -29

1 <28

1

Liquid Biphasic jCrystalline

. i: 1

1 1

_ • ' 1

• 1 1 1

1 i

* H i 1

1 1 1

1

!

1 1

t i l l ' '"I » 1 • 1 «

i

1 1 1

1 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v(CH)]/I[v(CHJ]

Figure 3.11 peak frequency as a function of the intensity ratio of the Va(CH2) to the v.CCHj) for octadecane.

123

frequencies and I[vs(CH3)fr]/I[Vs(CH2)] are indicators of interchain coupling, it is

important to note that they represent measures of different parameters. The peak

frequencies are sensitive to subtle chain interactions that may dampen or otherwise affect

V(CH2) vibrations. In contrast, the I[VS(CH3)FR]/I[VS(CH2)] is sensitive to larger coupling

interactions that affect the terminal methyl group of the alkane chains and molar volume

changes of the material. The v(CH2) vibrations are damped by highly coupled alkane

chains in the crystalline material and arc shifted to lower frequency. As the thermal

energy of the system increases in the solid, the chains decouple slightly and the

frequencies increase. Through the phase transition, the peak frequencies increase slightly

and the chains lliither decouple.

The VS(CH2) peak frequency is a popular indicator of order not only in bulk alkane

systems, but in surface-bound alkane systems as probed by surface vibrational methods

' '7 A f \ 'J such as surface FTIR and vibrational sum-frequency generation. ' '' ' ' This indicator

should be used with some caution, however, in light of the fact that the Vs(CH2)

frequency shifts by 6-8 cm"' for neat octadecane in the crystalline state. This shift is a

clear indication that this indicator alone does not adequately describe the conformational

state of these systems. While the lowest frequency of this mode is indeed an indication

of a highly-ordered crystalline state, and the frequency of the Vs(CH2) is sensitive to chain

decoupling in the crystal, information regarding the disorder imparted to the alkane

chains during a phase transition and in the liquid phase is not available from this

parameter.

124

Alkane chain intramolecular motion and chain coupling as indicated by the

xCCHj) peak asymmetry and width (FWllM) are correlated with I[Va(CH2)]/I[Vs(CH2)] in

Figures 3.12 and 3.13, respectively. The peak asymmetry is quantified by calculating an

asymmetry factor equal to the ratio of the frequency distance from the left and right edges

(measured at the FWHM) to the peak frequency. As shown in Figure 3.12, the x(CH2)

remains symmetric (asymmetry factor =1.0) for I[Va(CH2)]/I[Vs(CH2)] values between

2.1 and 1,5. In the melting region, the peak becomes slightly asymmetric as the

rotational order decreases (I[Va(CH2)]/I[Vs(CH2)] values between 1.2 and 0.8). Once in

the liquid state, this peak continues to increase in asymmetry as the rotational disorder

remains at its maximum. Temperature regions in which the t(CH2) peak becomes

asymmetric coincide with regions in which the number of gauche conformers increases as

indicated by the I[V(C-C)G]/I[V(C-C)T]. A similar trend is observed for the increase in

the FWHM of the t(CH2) mode with both increasing rotational disorder and with

increasing gauche conformers as shown in Figure 3.14. Clearly the development of

asymmetry in this band results from changes in the population of gauche conformers. The

fact that a discrete band is not observed for the t(CH2) mode of gauche conformers

suggests that a range of alkane chain twisting motions are accessible to gauche

conformers whereas a much narrow range is accessible to trans conformers. In light of

the considerable increased rotational disorder that accompanies the formation of gauche

conformers, this observation is not surprising.

>30 Temperature (°C)

-29

Crystalline Liquid

I

Biphasic

—1—I—1—I—I—I—I—I—I—I—I—I— 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v(CH)]/I[v(CH)]

Figure 3.12 xCCHj) peak asymmetry factor as a function of the the intensity ratio of the v/CHj) to the v^(CH2) for octadecane.

126

>30 Temperature fC)

-29 <28

Liquid Biphasic

I! Crystalline

i + 1 ' m

I 1 j 1 ip-jH

26

24

22-

20

1 8 -

1 6 -

14-

12 H

10

8

6

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v(CH)]/I[v(CH)]

Figure 3.13 T(CH2) full-width half maximum (FWHM) as a function of the intensity ratio of the Va(CH2) to the v/CHj) for octadecane.

127

Temperature (^^C)

26-

24-

22-20-

18-CJ 16-

14-u 12-

10-8-

£ 6-

4-

2-0--0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

I[v(C-C)J/I[v(C~C)J

Figure 3.14 x(CH2) FWHM as a fimction of the the intensity ratio of the v(C-C)(; to the v(C-C)-|. for octadecane.

128

Similar behavior is observed for the correlation between the A[5(Cll2) + 5(CH3)]/

A[6(CH2)1 and I[VaCCHj)J/I[VsCCHi)] as shown in Figure 3.15. The complexity and

general uncertainty of the peak assignments in this region led to the use of integrated

peak area ratios instead of peak intensity ratios as used for the other indicators. The peak

areas of these two modes were determined using a Gaussian lineshape analysis of the

spectra. These fits are shown in Figure 3.16. The A[5(CH2) + 8(CH})]/A[6(CH2)1

remains constant until gauche conformers are introduced into the alkane structure. This

value increases in the biphasic region and increases with increased thermal energy in the

liquid state. This trend is again a result of the increasing population of gauche

conformers and the increased intramolecular motion allowed as octadecane melts.

Raman Spectroscopy of Polyethylene

The rapidity of the phase transition for octadecane makes acquisition of quality

spectroscopic information near the phase transition difficult. Moreover, the phase

transition for octadecane results in a biphasic mixture of species and not a system

containing systematically varying disorder. For these reasons, the Raman spectral

response for polyethylene was also probed on both sides of the phase transition. In light

of its fewer degrees of freedom and a less abrupt phase transition, this system provided a

more appropriate model for surface-confined alkane systems. For this investigation, a

medium density (p ~ 0.92 g/cm^), linear polyethylene was used to spectroscopically

examine alkane melting. Due to differences in the thermal behavior of this system as

129

>30 Temperature (°C)

-29 <28

Liquid

1 Diphasic Crystalline

1 1 •h •-H

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 I[v/CH^)]/I[v/CH^)]

Figure 3.15 The peak area ratio of the [SCCHj) + SCCHj)] to the SfCHj) as a function of the intensity ratio of the Va(CH2) to the v^CCHj) for octadecane.

1400 1420 1440 1460 1480

Waveniunbers (crn'^)

1500

1500 I 1440 1460

Wavenumbers (cm"\)

1480 1400 1420

Figure 3.16 Representative peak fits to the Raman spectra of octadecane in the S(CH-H) region at (a) 0 <'C and (b) 60 "C.

131

compared to octadecane, the polyethylene data are presented in a slightly different format

in that an expanded temperature range is considered and the three regions identified are

the termed the crystalline, amorphous solid and liquid regions.

The correlation between I[V(C-C)G]/I[V(C-C)T] and I[VA(CH2)]/I[VS{CH2)] for

polyethylene is shown in Figure 3.17. The behavior for polyethylene is similar to that

observed for octadecane (Figure 3 .6) at high temperatures in that the number of gauche

conformers increases significantly between I[Va(CH2)]/I[Vs(CH2)] values of 1.2 and 0.8,

and continues to increase in the liquid phase while I[va(CH2)]/I[Vs(CH2)] remains

constant. The most significant differences in order behavior between the two materials

occur in the low temperature region. Polyethylene is not as inherently ordered as

octadecane at these temperatures due to a small number of gauche defects in the polymer

chains that still exist even when I[Va(CH2)]/I[Vs(CH2)] has values of 1.6 to 1.3. This

inherent disorder is due to the orthorhombic crystal structure of the polymer (in contrast

to the hexagonal crystal structure of octadecane) in which the polymer chains contain

hairpin bends.The more restrictive orthorhombic crystal structure of the

polyethylene results in fewer degrees of freedom for the chains to undergo decoupling

and rotational disorder.

Insight into the extent of interchain coupling through the Vs(CH2) peak frequency

is correlated with I[VA(CH2)]/I[VS(CH2)] in Figure 3.18. In solid polyethylene, the

VsCCHi) peak frequency never gets as low as it does for solid octadecane, indicating an

inherently more decoupled crystalline system. During melting, this peak frequency

132

1.4

1.2

^ 1.0 u u 0.8

0.6

O 0.4 U

S 0.2 i-H

0.0

-0.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v/CH^)]/I[v/CH^)]

Temperature ( C ) 150 100 90 50 0

J I t I . I § 1

Liquid " Amorphous ; Crystalline ai I Solid

V

Figure 3.17 The intensity ratio of the V{C-C)q to the v(C-C)-, as a function of the intensity ratio of the v^CCHj) to the v/CHj) for polyethylene.

2854

2852

I, 2850

O § 2848 gn

£ 2846

S 2844

B 2842 &0

>

2840 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v/CH^)]/I[v/CH^)]

Temperature (C) 150 100 90 50 0 _j I I I I 1

. Liqui^ 1 1

Amorphous Solid

1 Crystalline

• 1

1 1 Amorphous

Solid

1 —1 1 T r 1

•

I 1 I 1 > 1 1 1

Figure 3.18 v/CHj) peak frequency as a function of the the intensity ratio of the v^CCH j) to the v/CHj) for polyethylene.

134

increases with increasing temperature in the solid as the orthorhombic unit cell expands

and the chains decouple. Beyond the melting temperature, the frequency continues to

increase in the liquid phase as the thermal energy of the system increases and chain

decoupling increases, similar to the behavior observed for octadecane.

The correlations of the Va(CH2) and t(CH2) peak frequencies and the

I[VS(CH3)FR]/I[VS(CH2)] with T[Va(CH2)]/I[Vs(CHi)] are shown in Figures 3.19, 3,20, and

3.21, respectively. 1 [Vs(CH<)i.R]/1 [Vs(CH2)] correlates similarly with

I[Va(CH2)]/I[Vs(CH2)] for polyethylene as for octadecane: increased chain decoupling

(indicated by an increase in I[VS(CH3)FR]/I[VS(CH2)]) occurs with increasing disorder.

Polyethylene chain decoupling can also be correlated to an increase in specific volume

(mL/g) in the solid and liquid forms of the polymer as described for a similar

polyethylene by Sibukawa et al.'^ *'^ For polyethylenes of similar physical characteristics,

the specific volume increases slightly (<10%) in going from the crystalline to the

amorphous solid state (20 °C to the m.p. ), and then increases -25% for increasing

temperatures throughout the liquid phase. These observations are similar to those made

for the molar volume changes in octadecane (Figure 3.8), although the ditTerences in

magnitude of the changes in molar and specific volumes for octadecane and

polyethylene, respectively, stem from the unique physical and stmctural properties of

these two systems.

As shown by the data in Figures 3.22 and 3.23, the correlation of the t(CH2) peak

asymmetry and the T(CH2) FWHM with the I[Va(CH2)]/I[Vs(CH2)] provides insight into

135

o

g 3 cr

<u CU

u

2890

2888

2886

2884-

2882-

2880

2878 H

2876

2874

Temperature ("C) 150 100 90 50 0

2872

i Liquid ' Amorphous Crystalline

1 Solid 1

-* i

1 • : M 1 1 1 1 1 1 1

1 1 1

•

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 I[vrCHJ]/I[v(CHJ]

Figure 3.19 v ,(CH2) peak frequency as a function of the the intensity ratio of the v^CCHj) to the v/CHj) for polyethylene.

136

1304'

S o

¥ c <u 3 a"

u.

0) D-

o

1298

!

1296

1294'

150 100 Temperature ("C) 90 50 0

1 1 1

Liquid ' ^

Amorphous ' Crystalline 1 Solid 1

1

; i 1

1 1 1

1 1

1 1

* 1

1 1 1

1

1

L 1

< 1 1 ' 1 1 • I < I < I < I < I < I

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[vrCHJ]/I[v(CHJ]

Figure 3.20 t(CH2) peak frequency as a function of the the intensity ratio of the vJCHj) to the v/CHj) for polyethylene.

137

Temperature (°C) 150 100 90 50 0

U i/j

' Dd

u X

0.85

0.80

0.75'

0.70-1

0.65

0.60-

0.55

0.50

0.45

0.40

0.35

Liquid " Amorphous

t! Solid I

Crystalline

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 I[v(CH)]/I[v(CHJ]

Figure 3.21 Intensity ratio of the to the v/CHj) as a function of the intensity ratio of the Vj,(CH2) to the v^(CH2) for polyethylene.

138

U

i-A B

<D

•M

2.8

2.6

2.4

2.2

2.0

1.8'

1 . 6 -

1.4-

1 . 2 -

1.0-

0.8'

Temperature (°C) 150 100 90 50 0

J I L

Liquid • Amorphous ^ Solid

Crystalline

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v/CH^)]/I[v/CH^)]

Figure 3.22 x(CH2) peak asymmetry factor as a function of the intensity ratio of the v/CHj) to the v/CHj) for polyethylene.

139

26-

S o

U

20-

16-

10-

Temperature ("C) 150 100 90 50 0

1

Liquid 1 1

Amorphous

1

Crystalline

* • 1 Solid

i • i

i 1 m •

( m

—• 1 • ' 1 ' 1 • 1 • 1 1 1 1 1 > 1 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

I[v(CH)]/I[v(CH)]

Figure 3.23 xCCHj) FWHM as a function of the intensity ratio of the v^(CH2) to the vJCHj) for polyethylene.

140

intramolecular motion of the alkane chains. In contrast to octadecane, the xCCHa) mode

for polyethylene becomes extremely asymmetric at higher temperatures, reaching a

distinct maximum value of asymmetry at the melting point. Interestingly, at temperatures

above the solid/liquid phase transition, the T(CH2) peak asymmetry decreases sharply and

becomes comparable in value to that for liquid octadecane. Other researchers reported

similar qualitative observations of this increasing asymmetry of the ^(CHs) mode for

polyethylene and attributed it to an increase in population of amorphous states of the

polymer.' For octadecane, the amorphous component (i.e. the liquid phase) is only

present at temperatures above the transition temperature. No amorphous solid exists for

solid octadecane. Conversely, for polyethylene, an amorphous component develops in

the solid material as temperature increases, but stays below the transition temperature.

The increase in the x(CH2) peak asymmetry is a quantitative measure of this amorphous

character of solid polyethylene, since this vibration is sensitive to changes in the

crystaUine, amorphous solid and amorphous liquid phases.

The x(CH2) FWHM increases with decreasing conformational order and increases

with the number of gauche conformers as shown in Figure 3.24. This behavior is similar

to that observed for octadecane. While the asymmetry of the xlCHo) is dependent on

rotational disorder, the population of gauche conformers and the crystallinity of the

material (crystalline, amorphous solid, or liquid), the x(CH2) FWHM depends only on the

population of gauche conformers and rotational disorder.

141

26-

24-

22-

1 20-B -

o 18--

16-U

14-14-

i 12-

1 10-

8-

6-

Temperature ("C) 150 100 90 50 0

Liquid Amorphous Solid

#

1 ' Crystalline

f 1

• i i • •

1 1 1 1 I

1 1 1 1

1—'—I—f 1 r-

-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

I[v(C-C)J/I[v(C-.C)g

Figure 3.24 xCCHj) FWHM as a function of the intensity ratio of the V(C-C)Q to the v(C-C)[. for polyethylene.

Similar behavior occurs for the A[6(CH2) + 5(CHJ)]/A[5(CH2) + 5(CH2)ORTHO]

when correlated to the I[Va(CH2)]/I[Vs(CH2)] for polyethylene as shown in Figure 3.25,

with representative peak fits of these data shown in Figure 3.26 for polyethylene at 0 and

89 "C. In this case, A[5(CH2) + S(CH3)]/A[8(CH2) + S(CH2)ORTHO] increases with the

number of gauche confonners, indicating increased intramolecular motion of the alkane

chains. This behavior is similar to that observed for octadecane throughout the biphasic

region.

The degree of orthorhombic crystalline character of the polyetheylene can be

determined by monitoring changes in peak area of the 8(CH2)ORTHO as a function of

temperature. In quantifying a change in crystallinity, it is important to account for the

total area of this deformation mode, since it comes from the crystal field splitting of the

S(CH2) (1440 cm"'). Therefore, the area of the 6(CH2)ORTHO normalized to the summed

area of the 5(CH2) and 5(CH2) + SCCHu) modes is correlated to I[Va(CH2)]/I[Vs(CH2)] in

Figure 3.27 using an area ratio similar to that described by Abbate et al.^'^^ The

crystalline polymer loses its orthorhombic character with increasing temperature as the

unit cell expands. Once the polymer undergoes the solid/liquid phase transition, the

5(CH2)ORTHO mode disappears completely.

Comparison of Common Raman and IR-Active Indicators of Conformational Order

Conformational order is not as easily assessed trom infrared spectroscopy due to

limited resolution, inactivity and weakness of major alkane vibrational modes, and poor

143

150 100 Temperature (°C) 90 50 0

Liquid T

:

Amorphous Solid

Crystalline

1

1 •I K

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 I[v(CH)]/I[v(CH)]

Figure 3.25 The peak area ratio of the SCCH,) + 5(CH2) to the 5(CH2)ortho + SCCHj) as a function of the intensity ratio of the v/CHj) to the v/CHj) for polyethylene.

144

0-20 •

U

< 0.16

0.12'

0.08

0.04

Temperature ("C) 150 100 90 50 0

I I I I I

Liquid ' Amorphous J _Crystalline 1 Solid

i

I X

0.00-

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 I[v(CHJ]/I[v(CH)]

Figure 3.26 The peak area ratio of the 5(CH2)ORTHO to the {[Sf CHj) + 6( CH , )] + 5(CH2)} as a function of the intensity ratio of the VgCCHj) to the Vg(CH2) for polyethylene.

145

1400 1420 1440 1460 1480 1500

Wavenumbers (cm"^)

!—!

1400 1420 1440 1460 1480 1500

Wavenumbers (crti'^)

Figure 3.27 Raman spectra of the deformation modes (1390-1510 cm"') peak fit using a Gaussian distribution at (a) 0 "C and (b) 89 "C.

146

experimental sensitivity when compared to the results presented here for Raman

spectroscopy. Nonetheless, the infrared spectra of alkanes also contain modes that are

sensitive to alkane conformational order. Specifically, those modes cited most frequently

as containing information about conformational order include the Vs(CH2) and Va(CH2)

peak frequencies (2850 and 2880 cm"') and the v(C-C) gauche, trans and kink modes

(~1000-1300 and -620-700 cm"'),^^^"^ '^'^ These indicators are important in elucidating

alkane structure in materials; however, the chain coupling information and numbers of

gauche and trans conformers provided by these indicators are only a small portion of the

information necessary for a complete molecular understanding of the disorder in these

materials. More direct comparison between Raman and IR spectral data for elucidation

of conformational order is difficult, because the majority of intense Raman-active modes

are either weak in the IR or IR-inactive and vise versa.^^' Thus, the completeness of the

conformational order indicators provided by the Raman spectra provide a compelling

case for the better utility of Raman spectroscopy relative to infrared spectroscopy for the

determination of conformational order in alkane-based systems whenever possible.

Summary

The work described here demonstrates the utility of Raman spectroscopy in

determining the general conformational order, specifically subtle changes in

intermolecular interactions, intramolecular motion and chain conformations, for alkane-

based systems. For octadecane, thermally-induced alkane-disordering is initiated by

147

decoupling of the alkane chains with concomitant rotational disorder in the hexagonal

crystal as indicated by increasing values of I[VS(CH3)fr]/I[VS(CH2)] (0.4 to 0.65), VsCCHj)

peak frequency (2841 to 2849 cm"') and decreasing values of I[ Va(Ci l2)]/l[Vs(CH2)] (2.1

to 1.4). Chain decoupling is accompanied by an increase in intramolecular motion as

indicated by an increase in the values of A[5(CH2) + 6(CH})]/A[5(CH2)] (0.8 to 4.9), the

t(CH2) FWHM (7 to 19 cm"') and the t(CH2) peak frequency (1295 to 1300 cm"'). Once

adequate thermal energy has been added to the system to sufficiently decouple the alkane

chains, gauche conformers form and the increase in decoupling ceases while the

octadecane become more liquid-like. The increase in gauche conformers is indicated by

the increase in the value of I[V(C-C)G]/I[V(C-C)T] (0 to 0.8) while the

I[VS(CH3)FR]/I[VS(CH2)] (-0.65) and the Vs(CH2) peak frequency (-2849 cm"') remain

constant. Once the phase transition from crystalline to liquid is complete, significant

increases in the molar volume of the liquid occur with increased temperature due to the

conversion of methylene trans and partially-rotated conformers to true gauche

conformers. These changes are reflected as increases in I[VS(CH3)FR]/I[VS(CH2)] (0.65 to

0.75), the VS(CH2) peak frequency (2849 to 2852 cm"') and I[V(C-C)G]/I[V(C-C)T] (0.8 to

1.3), while the value of I[Va(CH2)]/I[Vs(CH2)] (-0.75) remains constant.

Similar behavior occurs in the thermally-induced disordering of polyethylene,

with the exception of the unique structural differences between the hexagonal crystal

structure of octadecane and the orthorhombic structure of crystalline polyethylene,

indicated by the presence of the 5(CH2)ORTHO mode and the highly asymmetric T(CH2)

148

mode of melting crystalline polyethylene. At temperatures below the melting

temperature, the onset of alkane chain disordering stems from decoupling of the chains

and increased rotational motion of the chains in the orthorhombic crystal as indicated by

an increase in the values of 1 [VS(CH?)I.R]/1[VS(CH2)] (0.4 to 0.65) and the VsCCHi) peak

frequency (2846 to 2850 cm"'), a decrease in the value of I[Va(CH2)]/I[Vs(CH2)] (1.6 to

1.1), and the presence of the 8(CH2)ORTHO mode (1420 cm"'). Increased chain decoupling

is accompanied by an increase in intramolecular motion as indicated by an increase in the

values of A[5(CH2) + 6(CH3)]/A[5(CH2) + 5(CH2)ORTHO] (0.8 to 4.5), the T(CH2) FWHM

(7 to 19 cm"'), and the x(CH2) peak frequency (1295 to 1300 cm"'). Once adequate

thermal energy has been added to the system to sufficiently decouple the alkane chains,

additional gauche conformers form and the increase in decoupling ceases. In

contradistinction to octadecane, these changes begin while polyethylene is still in the

solid state and function to add more amorphous character to the solid prior to true

melting. The increase in gauche conformers is indicated by the increase in the value of

I[V(C-C)G]/I[V(C-C)T] (0.2 to 0.8) while the 1[VS(CH3)FR]/I[VS(CH2)] (-0.65) and the

VS(CH2) peak frequency (—2851 cm"') remain constant. Once the phase transition from

crystalline to liquid is complete, significant increases in the molar volume of the liquid

occur with increased temperature due to the conversion of methylene trans and partially-

rotated conformers to true gauche conformers. These changes are reflected as increases

in I[VS(CH3)FR]/I[VS(CH2)] (0.65 to 0.75), the Vs(CH2) peak frequency (2851 to 2853 cm"')

and I[V(C-C)G]/I[V(C-C)T] (0.8 to 1.3), while the value of I[VA(CH2)]/I[VS(CH2)] (-0.75)

149

remains constant. In addition, the conversion of the orthorhombic crystal to an

amorphous solid, and finally, to the amorphous liquid is indicated by an increase in

x(CH2) peak asymmetry (1.0 to 2 .6) and a decrease in the A[8(CH2)ORTHO]/[A[5(CH2)] +

A[5(CH2) + 5(CH3)]] (0.18 to 0).

Conclusions and Relevance to Chromatographic Studies

Raman spectral indicators of conformational order for alkanes were successfully

correlated in terms of alkyl chain conformations (gauche and trans), chain coupling, and

intramolecular rotation and motion for octadecane and a low molecular weight

polyethylene. These data can be directly related to understanding Raman spectral

indicators of order in other alkyl-based systems including alkylsilane stationary phases

investigated throughout this Dissertation. With these correlation data, it is possible to

probe only one spectral indicator, such as the I[Va(CH2)]/I[Vs(CH2)], and be able to obtain

a complete understanding of alkyl chain coupling, relative number of gauche conformers

and the degree of alkyl chain rotation. This will be particularly useful in investigating

stationary phase order in the presence of mobile phase and/or solutes were spectral

interference from solvent and solute modes could make much of the conformational order

information in the Raman spectra inaccessible.

150

Future Directions

Future directions for this work include extending the correlation of Raman

spectra] indicators not only to each other, but also to alkane indicators of conformational

order and mobility from IR and NMR spectroscopic data. These types of correlations

would also allow the information obtained from these complementary techniques to be

utilized more effectively and to bring a new appreciation for how closely related these

data from different techniques really are to one another. In addition, repeating these

types of experiments for deuterated alkane systems would also be beneficial. By

developing similar spectral indicator correlations for deuterated alkanes one could

independently distinguish conformational changes in binary alkane systems. This

information would be valuable to researchers in numerous interest areas including self-

assembled monolayers, l.angmuir-Blodgett films, chromatography, and alkyl-protected

nanoparticles.

Appendix 3.1

This appendix contains Raman spectra in the v(C-C), 5(C-H), and v(C-H) for

octadecane and polyethylene at temperatures between 10 and 150 "C.

2900 3000 1000 1100 1200 1300 1400

Waveniimbers (cm"\)

2800

Figure A3.1 Raman spectra in the v(C-C) and v(C-H) regions for octadecane at (a) -15 "C, (b) -10 "C and (c) 10 "C. Integration time 2 min.

3000 2900 1400 2800

Wavenumbers (cm'^)

1000

Figure A3.2 Raman spectra in the v(C-C) and v(C-H) regions for octadecane at (a) 20 "'C and (b and c) through the phase transition. Integration time 2 min.

1000 1100 1200 1300 1400 2800

Wavenumbers (cm"^)

Figure A3.3 Raman spectra in the v(C-C) and v(C-H) regions for octadecane through the phase transition. Integration time 2 min.

2900 3000 2800 1000 1100 1200 1300 1400

Wavenumbers (cm"\)

Figure A3.4 Raman spectra in the v(C-C) and v(C-H) regions for octadecane (a) through the phase transition and at (b) 30 "C and (c) 40 "C. Integration time 2 min.

3000 2800 2900 1000 1100 1200 1300 1400

Wavenumbers (cm"\)

Figure A3.5 Raman spectra in the v(C-C) and v(C-H) regions for octadecane (a) 50 °C, (b) 70 "C and (c) 80 "C. Integration time 2 min.

1000 1100 1200 1300 1400 2800

Waveiiumbers (cm"^)

3000 2900

Figure A3.6 Raman spectra in the v(C-C) and v(C-H) regions for octadecane (a) 90 "C and (b) 100 "C. Integration time 2 min.

ON

1000 1100 1200 1300 1400 2800

Wavenumbers (cm*^)

2900 3000

Figure A3,7 Raman spectra in the v(C-C) and v(C-H) regions for polyethylene (a) 20 "C, (b) 43 °C and (c) 81 "C, Integration time 2 min. ->3

3000 2800 2900 1000 1100 1200 1300 1400

Wavenumbers (cm"\)

Figure A3.8 Raman spectra in the v(C-C) and v(C-H) regions for polyethylene (a) 85 "C, (b) 89 °C and (c) 91 "'C. Integration time 2 min.

2900 2800 3000 1000 1100 1200 1300 1400

Wavenumbers (cni"^)

Figure A3.9 Raman spectra in the v(C-C) and v(C-H) regions for polyethylene (a) 94 "C, (b) 106 "C and (c) 150 °C. Integration time 2 min.

160

Chapter 4

Solvent Effects on the Confonnational Order of Commercial Solid Phase

Extraction Alkylsilane Stationary Phases

Introduction

Despite efforts toward understanding solute retention in reversed-phase liquid

chromatography and solid-phase extraction, including experimental and computational

research, a molecular mechanism remains elusive. This inability to accurately determine

a mechanism for retention stems from the difficulty of studying the interactions known to

affect solute retention, including solvent-stationary phase, solute-stationary phase, and

stationary phase inter- and intramolecular interactions. In ail cases, the architecture and

structure of the alkyl chains will greatly affect interactions at the chromatographic

interface.

Solvent effects in reversed-phase separations systems have been the study of

numerous reports. While much of this work has been chromatographic,"^'from which

solvent-stationary phase interactions can only be infen ed, a significant amount of the

research has been spectroscopic,''which allows for direct measurement of these

interactions. Independent nuclear magnetic resonance (NMR) spectroscopy,''" ^

Fourier-transform infrared (FT-IR) spectroscopy,''^' and molecular dynamics

simulations'" ' show that polar solvents including methanol, water, and acetonitrile, have

161

very little alfect on the mobility and conformational order of surface-bound alkylsilanes.

Results from additional NMR experiments by Kelusky'*" indicate that nonpolar solvents,

such as benzene, have a strong ability to solvate the terminal portion of alkyl chains,

increasing the mobility of the terminal end of the stationary phase relative to these phases

in air.

Doyle and coworkers'^ used Raman spectroscopy to determine solvent effects on

the conformational order of a low surface coverage CIH-RPLC phase (3.52 fimol/m ).

The conformational order of this phase appeared to be unaffected by the presence of

chloroform, methanol and acetonitrile in these experiments. However, these experiments

were done in the presence of hydrogenated solvents which leads to significant spectral

interferences in the v(C-H), v(C-C), and 6(C-H) regions. The conclusion about solvent

effects was made based on qualitative observations that Raman spectral features do not

change in the presence of different solvents, for solvent-subtracted Raman spectra.

Results presented in Chapter 3 and other reports from this laboratory"*^" ''^' show that

quantitative indicators of conformational order for surface-bound alkylsilanes can be

measured very precisely, but changes in alkyl chain conformational order determined by

these indicators can be extremely subtle. Thus, if conformational order differences exist

for phases in solvent studied by Doyle et al, it is likely that subtle spectral information

would not be observed using spectral subtraction methods.

162

Goals of this Work

The goal of this work was to acquire Raman spectra of a commercial Cig-

stationary phase in the presence of solvent and to use these spectral data to determine the

effect of solvent on the conformational order of the surface-bound alkylsilanes. We were

interested in determining solvation differences of surface-bound alkylsilanes in polar and

nonpolar solvents, as well as evaluating solvation trends with changing solvent strength.

These data can be used to develop an understanding of how solvent-stationary phase

interactions affect solute retention in reversed-phase separation experiments.

Methodology

I solute C18MF (refered to as C18MF) was a gift from International Sorbent

Technologies (Hengoed, UK). Properties and characteristics of this phase are described in

Chapter 2. Briefly, C18MF is based on a monofunctional precursor,

dimethyloctadecylchlorosilane, bound to the silica substrate. Perdeuterated water-d2

(99.9%) was purchased from Aldrich. Perdueterated solvents, methanoi-d4 (99.8%),

acetonitrile-ds (99.8%), chloroform-d (99.8%), acetone-de (99,9%), benzene-de (99.5%),

toluene-dg, and isopropanol-dg (99%) were purchased from Cambridge Isotopes.

Samples were prepared by adding 200 |iL of solvent to 25-50 mg C18MF in 5.0-

mm-dia NMR tubes. Tubes were sealed and samples are sonicated for 15 min followed

by equilibration at room temperature for >18 hrs. Raman spectra were collected using

100 mW of 532-nm radiation from a Coherent Verdi Nd: YVO4 laser on a Spex

163

T riplemate spectrograph as described in Chapter 2. Slit settings of the Triplemate were

0.5/7.0/0.150 mm for all experiments corresponding to a spectral bandpass of 5.3 cm"'.

Samples were allowed to equilibrate at 24 °C for a minimum of 30 min prior to spectral

analysis, A minimum of three measurements were made on each sample. Integration

times for each spectrum are provided in the figure captions.

Raman spectra were processed and analyzed as described in Chapter 2. In

addition, spectra were smoothed using the Grams32 software (Roper Scientific) with a

Savitsky-Golay third order polynomial, 11 -point convolution function.

Review of Previous Studies of 1 solute C18MF SPE Phases

C18MF and other Isolute SPE phases have been studied previously by

A ly/y < ^ Q . . . A O'J

chromatography, ' Si cross-polarization magic angle spinning (CP/MAS) NMR, '

and by Raman spectroscopy'^^' to determine the nature of the silica surface, bonding

structure of various fiinctional silanes and to determine the conformational order of the

alkyl-modified phases. The silica substrates used for these SPE phases are quite unique

because of their large particle size (50 |j.m), large surface area (300-500 m^/g),

determined by BET analysis, and small pore size (~60 A.) Piccoli determined that a large

fraction of the surface area of these substrates is contained in micropores and is

inaccessible to large silane ligands, including octadecylsilanes.'*^^ Thus, the available

surface area for silane binding and solute interaction is less than the reported value

determined by BET analysis. For C18MF, the reported surface coverage is 1.9 jimol/m^;

164

however, because of this unique pore structure, the effective surface coverage is likely to

be -3-3.5 (imol/m^. This exclusion of octadecylsilane from micropores is shown

schematically in Figure 4.1, relative to available surface area for a propylsilane.

Previous Raman spectroscopy investigated the conformational order of C18MF as

a function of temperature in air.'' ^ These results show that at liquid N2 temperature, the

fraction of gauche conformers is <10%. In addition, it was also determined that from-15

to 95 °C, the number of gauche conformers of C18MF only increases to -48-55%.

Considerable disorder exists in these phases, even at low temperatures, and increasing

temperature only slightly increases the conformational disorder of these phases. This

small temperature effect is attributed to the fact that significant disorder exists in these

phases inherently due to low surface coverage such that the alkyl chains are forced to tilt

toward the silica surface and toward each other in order to maximize van der Waals

interactions with neighboring alkyl chains.

Raman Spectroscopy of Isolute C18MF in Solvent

The effect of stationary phase solvation on the alkyl chain conformational order

was determined by exposing C18MF to a range of solvents. In order to minimize spectral

interference from intense solvent modes in the v(C-H) region (2750 to 3050 cm"'),

C18MF was exposed to perdeuterated solvents. However, the use of perdeuterated

solvents does not eliminate the majority of spectral interferences in the v(C-C) and 8(C-

H) regions (1000 to 1500 cm"'). Therefore, interrogation of conformational order for

Figure 4.1 Schematic of the fractal pore structure of Isolute silica substrates showing the exclusion of octadecylsilane from micropores (left), giving overall less available surface area for octadecylsilane modification relative to propylsilane modification (right). Adapted from reference 4.22.

166

C18MF in solvent is restricted to the v(C-H) spectral region. Figure 4.2 shows Raman

spectra of C18MF in air and in a range of solvents in the v(C-H) region with

corresponding vibrational mode assignments listed in Table 4.1.

Spectra are free from solvent spectral interference; however, because these are

monofunctional phases, significant interference of the VaCCHs) mode (2880 cm"') arises

from the v(CH3)si band (2905 cm') of the two methyl groups attached to the silicon

atom. Despite this interference, conformational order can be assessed using the intensity

ratio of the VaCCHj) mode (2880 cm"') to the Vs(CH2) mode (2852 cm"'),

I[Va(CH2)]/I[Vs(CH2)]. While there is important conformational order information in the

low frequency regions of the Raman spectra of alkanes, conformational order can be

assessed adequately using the value of I[Va(CHi)]/![Vs(CH2)] in the v(C-H) region based

on previous work done to correlate the spectral indicators in each region as described in

Chapter 3. Discussion of the conformational order of C18MF determined by the value of

1 [V:,(CHi)]/!fVs(CH2)] will be correlated to alky! chain order information in all regions of

the Raman spectrum.

Raman spectra for C18MF in air and in all solvents are similar. From qualitative

observation of the value of I[Va(CH2)]/I[Vs(CH2)] from the Raman spectra, it is clear that

the effect of solvent on the order of C18MF is very small. However, the value of

I[Va(CH2)]/I[Vs(CH2)] is calculated precisely and can be used to discern subtle differences

in solvation behavior of the monofunctional phase.

167

CO C u

2800 2900 3000

2800 2900 3000 Wavenumbers (cm-^ )

Wavenumbers (cm-^)

Figure 4.2 Raman spectra in the v(C-H) region for C18MF in (a) air, (b) methanol, (c) acetonitrile. (d) water, (e) isopropanol, (f) acetone, (g) chloroform, (h) benzene, and (i) toluene. Integration times for each spectra are (a) 40 sec, (b) 1 min, (c) 40 sec, (d) 50 sec, (e) 30 sec, (f) 40 sec, (g) 30 sec, (h) 25 sec, and (i) 30 sec.

168

Table 4.1. Raman peak frequencies and assignments for 1 solute C18MF at 24 "C

Peak Frequency (cm ') Assignment"

28M " v,(CH2)

2887 Va(CH2)

2903 VsCCHOsi

2925 Vs(CH2)"'

2935 Vs(CH3)fr

2962 Va(CH3)

v = stretch

169

Effect of Solvent on Alkvlsilane Order for Isolate C18MF

The solvents examined in this study include water, acetonitrile, methanol,

isopropanol, acetone, chloroform, benzene, and toluene. These solvents exhibit a range

of characteristics including polarity, dipole moment, polarizability, hydrogen bond

accepting/donating ability, hydrophobicity, shape and size that have been quantified

according to the solvatochromic parameter, n*, developed by Kamlet et al.'^ " These

parameters are listed in Table 4.2. n* discriminates solvents by their dipole moment

normalized to the solvent polarizability; in other words, it is a measure of electron

mobility within a molecule. Solvatochromic parameters were developed to describe

linear solvation energy relationships or general solvent strength and have been used to

describe solvation effects in liquid chromatography.

Figure 4.3 shows a plot of 1[Va(CH2)]/l[VsCCHs)] as a function of 7C* for C18MF in

air and in solvent. In the more polar solvents water, methanol, and acetonitrile, the value

of l[va(CH2)]/I[Vs(CH2)] for C18MF is the same as for this phase in air. In the less polar

solvents isopropanol, acetone, and chloroform, this value for C18MF is only slightly less

than it is in air. The value of I[Va(CH2)]/I[Vs(CH2)] for C18MF is less in nonpolar toluene

and benzene than in air.

In air, the value of I[Va(CH2)]/I[Vs(CH2)] for C18MF is ~ 0,98. From work

presented in the previous chapter for octadecane and polyethylene, a value of

I[Va(CH2)]/I[Vs(CH2)] of-1.0 corresponds to a very disordered alkyl system with a large

population of gauche conformers (-30-35% gauche), significant alkyl chain decoupling

170

Table 4.2. Solvatochromic parameters'* 7u*, a, and j3 for solvents

Solvent 71* a P

"™™water L09 "OAE ll?

acetonitrile 0.75 0.31 0.19

acetone 0.71 0.48 0.06

methanol 0.60 0.62 0.93

benzene 0.59 0.10 0

chloroform 0.58 0 0.44

toluene 0.54 0 0

"As described by Kamlet et al.'^ "

171

1.00

0.98-

d 0.96

r-«^

O

0.94-

0.92

1 Methanol Acetonitrile

Air •

Isopropanol

Chloroform

Acetone

Water

Toluene TT Benzene

0.90 [ 1 1 ! 1 1 -J ! 1 1 1 1 1 0.0 0.2 0.4 0.6 0.8 1.0 1.2

n*

Figure 4.3 Raman spectra in the v(C-H) region for C18MF in (a) air, (b) methanol, (c) acetonitrile, (d) water, (e) isopropanol, (f) acetone, (g) chloroform, (h) benzene, and (i) toluene.

172

and rotational disorder, and substantial intemiolecular motion within the chains The

range of I[Va(CH2)]/I[Vs(CH2)] values for C18MF extends from ~ 0.98 for air to ~ 0.92

for toluene, indicating that C18MF is at least as ordered or less disordered in all of these

solvents than in air. However, this range of alkyl chain disordering in solvent is

extremely subtle. At a I[Va(CH2)]/I[Vs(CH2)] value of ~ 0.92, increased disorder for

these alkyl systems comes from a slight increase in the population of gauche conformers

(~ 40-45% gauche). Alkyl chain decoupling and intermolecular motion within the chains

remain unchanged for I[Va(CH2)]/I[Vs(CH2)] values between 0.9 and 1.0,

Even though the disordering observed for C18MF in solvent was very slight, the

trends between classes of solvents are interesting. The most polar solvents methanol,

acetonitrile, and water, had no effect on the conformational order of C18MF as indicated

by the value of I[va(CH2)]/I [VsCCHi)] in each of these solvents; these observations are in

good agreement with previous reports.'^ These solvents are common mobile

phase components in RPLC and SPE experiments, and the relative amounts of each are

known to greatly influence solute retention. The fact that these solvents have no effect on

the conformational order of C18MF is an indication that the solvent-stationary' phase

interactions are not of principal importance in this case, but that solute-solvent

interactions are the most critical in detennining retention behavior. Isopropanol, acetone

and chloroform have a similarly minimal effect on C18MF.

173

The conformational order of C18MF in the most nonpoiar solvents benzene and

toluene, decreased compared to C18MF in air. This solvent-induced disordering, albeit

very slight, is consistent with the expectation that nonpoiar molecules should act as better

solvents for a nonpoiar stationary phase than polar solvents. Benzene and toluene

molecules can fully solvate the alkyl chains, imparting conformational disorder. This

result is in good agreement with previous work that suggests that nonpoiar molecules can

penetrate between alkyl chains and can solvate at least the terminal end of the chains."^'

Conclusions and Future Directions

The results presented here demonstrate the ability of Raman spectroscopy to

determine subtle differences in alkylsilane general conformational order for a commercial

SPE phase in a range of solvents. Despite only slight changes in conformational order

observed for C18MF in solvent, the differences found in solvation ability between polar

and nonpoiar solvents are reasonable and consistent with previous reports.'*

Polar solvents are relatively poor solvators of the alkyl chains and have very little effect

on their conformational order, while the nonpoiar solvents can better solvate the terminal

end of alkylsilanes and as a result induce more conformational disorder. Significant

disorder exists in these phases under ambient conditions, because the surface coverage is

low, forcing alkyl chains into disordered conformations in order to maximize van der

Waals interactions.'*^' As a result, it is reasonable that solvating these disordered alkyl

174

chains in either poor or strong solvents does not have a dramatic effect on alkyl chain

order.

Future directions for this work include the investigation of solvent effects on the

conformational order of higher coverage stationary phase materials. For C18MF, a low

coverage, disordered phase in air, solvent effects are very small. Perhaps by investigating

higher coverage, more ordered phases in the ambient, more pronounced solvent effects

can be observed. In addition, studying the effect of binary or ternary solvent systems

important as mobile phases in SPE and RPLC experiments is also necessary since

mixtures of two or three components are often used.

175

Chapter 5

Structure-Function Relationships in Octadecylsilane Stationary Phases by

Raman Spectroscopy: Effects of Temperature, Surface Coverage, and

Preparation Procedure

Introduction

Although the use of reversed-phase liquid chromatography (RPLC) is widespread,

significant debate remains regarding the principles and mechanisms of solute retention.

This is due, in part, to a general lack of understanding of the molecular processes giving

rise to retention. Such molecular processes include the interactions of solute and mobile

phase solvent with the stationary' phase and intermolecular interactions of the stationary

phase itself Such knowledge is critical, since each of these parameters is known to affect

the selectivity and efficiency of chromatographic separations.^ Numerous

experimental techniques including chromatographic methods,^ ' fluorescence

spectroscopy,"""''^^' nuclear magnetic resonance (NMR) spectroscopy,"^ small angle

C C f j is C Of: C I Q neutron scattering, ' "' ditYerential scanning calorimetry, ' "' infrared (IR)

5 30 5 33 5 34 5 39 spectroscopy, ' "' and Raman spectroscopy ' " ' have been employed in the

investigation of retention on alkylsilane-modified stationary phases. While the results of

these investigations have provided valuable insight into the retention process, a complete

176

molecular picture remains elusive. Among the many variables known to affect the

retention process, the structure of the alkyl chains and the effect of chromatographic

variables such as temperature and alkyl chain density on the resulting conformational

order of these alkyl chains may be the least understood. This information is best

provided by techniques sensitive to alkyl chain conformation and chemical environment

such as NMR, IR, and Raman spectroscopies.

Several models have been advanced to describe the surface structure of

alkylsilane stationary phases. The alkyl component has been envisioned as brush

bristles,"a folded or collapsed structure,' "" condensed liquid drops,' "^^ and as a phase

in which transitions between the liquid drop and bristle structures are possible with

changes in temperature.' In addition, the surface distribution of alkylsilanes has

sometimes been described as heterogeneous with the alkylsilane existing as island-like

patches.'^

Sander and coworkers reported shape selectivity in chromatographic retention for

a large number of octadecylsilane (Cig) stationary phases prepared using several

alkylsilane precursors and grafting procedures.^'® In these studies, the selectivity factor of

each stationary phase for 1,2;3,4:5,6:7,8-tetrabenzonaphthalene (a nonplanar aromatic)

relative to benzo|a]pyrene was assessed. Results indicate a generally linear correlation

between alkylsilane surface coverage and the presence of shape selectivity; improved

separation of isomers and other structurally-related solute classes are usually possible

with increasing surface coverage.

177

In earlier work, Sander and coworkers examined a series of dimethylalkylsilane

stationary phases, hereafter referred to as monofunctional phases, of differing alkyl chain

lengths using Fourier transform IR (FTIR) spectroscopy.^ These phases possessed

similar bonding densities (2 to 2.5 |umolW) and were chosen to elucidate differences in

alkyl chain conformational order as a function of length and temperature. Order within

the alkyl chain was assessed by examining the spectra for the presence of bands

corresponding to gauche carbon-carbon bonds and kink defects in the alkyl chains. A

significant fraction of gauche defects was observed in the absence of solvent where the

degree of disorder was comparable to that observed in the corresponding n-alkane

liquids. These experimental results suggest that surface-confined alkylsilane components

in such systems exist in a state with liquid-like properties. In addition to the large degree

of conformational disorder, no evidence for temperature-dependent phase transitions was

observed for these phases, although large continuous changes in conformational order

were found with changes in temperature.

More recently, Pursch and coworkers employed 'H magic angle spinning (MAS)

NMR and ' 'C cross-polarization magic angle spinning (CP/MAS) NMR to assess alkyl

chain order in a series of high-density Cig-stationary phases with varying surface

coverage."^ A more rigid alkyl chain environment was indicated for these higher

surface coverage stationary phases. In addition, a larger fraction of alkyl chains in trans

conformations was observed on high surface coverage stationary phase materials.

178

In addition to experimental investigations of alkylsilane-modified silicas,

properties of these materials have also been modeled using molecular dynamics

simulations/''*^ Employing this approach to examine monofunctional Cig-stationaiy

phases of different chain densities, Klatte and Beck observed that at surface coverages

from 1.5 to 4 fimol/m^, the alkyl chains form layered structures collapsed onto the

exposed silica surface. The alkyl chains in these model phases were predicted to possess

tilt angles exceeding 55°, supporting a picture in which the chains lie on the silica

surface. The authors observed increasing order and decreasing tilt angle with surface

coverages exceeding 4 |imol/m^. A larger degree of conformational disorder was

observed in portions of the chains more distal to the surface pendant group, with

increasing order as the surface is approached. Such behavior is attributed to an alkyl

chain environment with more degrees of freedom in methyl and methylene units a greater

distance from the surface. It is of interest to note that, for surface coverages ranging from

1.5 to 4 |Limol/m^, the alkyl chain order in these calculations exhibited no temperature

dependence from 230 to 340 K.

Perhaps the most powerful experimental tool for characterizing conformational

order changes in alkyl chains is Raman spectroscopy, since it provides direct information

about conformational order.'^ Additionally, such measurements are relatively

free from spectral interferences from the silica substrate and adsorbed water that plague

5 30 5 35 5 37 IR studies of such systems. • In previous reports, ' ' ' the feasibility and utility of

Raman spectroscopy for elucidating information about conformational order of

179

alkylsilane stationary phases have been demonstrated. As part of an ongoing effort to

C 'J/1 •4! < T;*? elucidate retention mechanisms in reversed phase liquid chromatography, ' '' ' ' these

initial studies are expanded in this report to include the effects of temperature, surface

coverage, stationary phase preparation (surface polymerized or solution polymerized) and

the nature of the silane precursor (octadecyltrichlorosilane,

methyloctadecyldichlorosilane or dimethyloctadecylchlorosilane).

Goals of this Work

The synthesis and properties/'^ chromatographic behavior,^'^'^'"^ and NMR

spectroscopy^''*'^ of the high-density Cig stationary phases considered here have been

previously described. As noted above, these phases show selectivity toward planar

polycyclic aromatic hydrocarbons (PAHs) that is highly dependent on surface

coverage. Such results have been interpreted as being indicative of increasing alkyl

chain order with increasing surface coverage. The goal of this work was to examine the

conformational order of these stationary- phases as a function of temperature to elucidate

trends in behavior that may be attributable to differences in surface distribution, or

intermolecular interactions. The resulting phase change behavior of these phases is

described using the Clapeyron relationship for first-order transitions.

180

Methodology

Octadecylsilane-modified silica-based stationary phases were prepared and

characterized as previously described"" and briefly discussed in Chapter 2.

Specifications and properties of these phases are provided in Table 5.1. Raman spectra

were collected using 100 mW of 532 nm radiation from a Coherent Verdi Nd:YV04 laser

on a Spex Triplemate spectrograph as described in detail in Chapter 2. The majority of

the spectral data were acquired by Dr. Michael Ducey. The spectral band pass of the

Triplemate was 5 cm"' for the slit settings of (0.5/7.0/0.150) mm used for all experiments.

Samples were sealed in 5-mm-dia NMR tubes and positioned in the laser beam using a

copper sample mount. The temperature of the copper mount was regulated by circulating

a temperature controlled medium (50;50 ethylene glycol;water) through it (and around

the sample tube) using a NESLAB NTE-110 temperature controller, allowing

temperature control from 258 to 343 K. Samples were allowed to equiUbrate at each

temperature for a minimum of 30 min prior to analysis to ensure that any temperature-

induced changes in conformational order were complete. A minimum of three

measurements were made on each sample at each temperature. Integration times for each

spectrum are provided in the figure captions.

Raw spectral data in the v(CH) region were first corrected for background using a

5-point linear fit of the data in the region between 2750 and 3050 cm"'. Peak intensities

of the Va(CH2) and Vs(CH2) bands were determined as peak heights from the baseline on

the corrected spectra. In addition, Raman spectra of the monofunctional phase were

181

Table 5.1. Stationary phase properties

Stationary Previous Preparation Method % Carbon Surface Coverage, T Phase" Designation'' (nmolW)

TFC18SF SI Surface Polymerization 19.84 6.45

TFC18SL PI Surface Polymerization 17.37 5.26

DFC18SF S3 Solution Polymerization 17.07 5.00

DFC18SL P2 Solution Polymerization 15.10 4.17

MFC 18 Ml 12.45 3.09

"TF, trifunctional alkylsilane precursor; DF, difunctional alkylsilane precursor; MF, monofunctional alkylsilane presursor. ''Designation of stationary- phases from references 5.8 and 5.44. "Preparation method as described in the text and in reference 5.8.

182

smoothed using the Grams32 software (Roper Scientific) with a Sav-Golay 3"* order

polynomial, 11 -point convolution function.

Raman Spectroscopy of N 1ST Cis Stationarv Phases

The effect of temperature on the conformational order of the series of high-

density stationary phases described above was examined by determination of the

magnitudes of two Raman spectral conformational indicators in the v(C-H) region

between 2750 and 3050 cm"^ over a temperature range from 258 to 343 K. Within this

region, bands due to symmetric and antisymmetric v(CH2) and vCCH?) vibrations as well

as bands due to the Fermi resonance of these modes with overtones of the 6(CH) modes

were observed. As a result of the large number CHi groups in the alkyl component of the

stationary phase, this region is spectrally complex.'Figure 5.1 represents typical

Raman spectra in this region for stationary phases prepared from trifunctional (TFC18SF

and TFC18SL), difunctionaJ (DFC18SF and DFC18SL), and monofunctional (MFC 18)

Cig alkylsilane precursors with surface coverages of 6.45, 5.26, 5.00, 4.17 and 3.09

|imol/m", respectively, at 263 K (Figure 5.1a) and 333 K (Figure 5.1b). Additional

spectra of these stationary phases in air at temperatures between 258 and 343 K are

provided in Appendix 5.1. Vibrational mode assignments for each of these stationary'

phase materials are provided in Table 5.2. In general, the prominent bands represented in

each spectrum correspond to the Vs(CH2) at 2852 cm"', the Va(CH2) at 2885 cm"', the

VS(CH2)FR at 2924 cm"', and the Va(CH3) at 2955 cm"'; the Vs(CHO was observed as a

183

TFC18SF

TFC18SL

DFC18SF

DFC18SL

60000 MFC18

2800 2900 3000

FC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

2800 2900 3000 Wavenumbers (cnr') Wavenumbers (cm-\)

Figure 5.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC I8SL, and MFC 18 in air at (a) 263 K and (b) 333 K. Integration times for all spectra are 10-min.

184

Table 5.2. Peak frequencies and assignments for NIST Ci8 stationary phases

Peak Frequency (cm"') Assignment

TFC18SF TFC18SL DFC18SF DFC18SL MFC 18

2846 2851 2851 2851 2853 V.(CH2)

2880 2880 2882 2884 2884 Va(CH2)

2995 2997 2998sh 2996 2900 VS(CH3)

2905sh 2905sh 2905sh v(CH3)si

2932 2929 2925 2926 2920 Vs(CH3)fR

2954 2954 2954 2955 2954 Va(CH3)

sh = shoulder V = stretch

185

shoulder at 2898 cm"' in most spectra. Assignments for these systems are well-

established and are made based on those previously reported for alkyl chain systems.' '*^'"

5.58

While these spectral features are relatively distinct, several less-obvious features

were identified within each spectrum. Of particular significance is the v(CH3)si mode for

the methyl groups attached to the proximal silicon atom, present in both the difunctional

and monofunctional stationary phase materials. This mode is located at 2905 cm"V'^^

and is superimposed on the VS(CH2)FR at 2924 cm"' and the VsCCH}) at 2898 cm"'. As a

consequence of its position, it may also contribute slightly to the intensity of the VaCCHj)

band, as described in Chapter 4.

It had been previously demonstrated that Raman spectra can provide a wealth of

detailed information regarding alkyl chain conformation,"^ especially in

the v(C-H) region. While this region exhibits considerable complexity as a result of

Fermi resonance modes and a plethora of both symmetric and antisymmetric v(C-H)

modes, previous work has shown^that conformational order information can be

empirically derived from the peak intensity ratio of the antisymmetric (2885 cm"') to

symmetric (2852 cm"') vCCHi) bands, l[Va(CH2)]/I[Vs(CH2)]. The frequencies at which

the VsCCHa) and Va(CH2) modes are observed also reflect conformational order; these are

observed at 2856 and 2888 cm"', respectively, for alkanes in the liquid state, and decrease

by 6 to 8 cm"' for crystalline alkanes.^'^^"^'^® The peak frequency of the Vs(CH2) band has

been related to the extent of coupling between alkyl chains, where increased chain

186

decoupling results in an increase in the peak frequency.'The behavior of these and

many other spectral indicators of alkyl chain order were described in detail in Chapter 2.

A qualitative examination of the spectra in Figure 5 .1a reveals a wide range of

conformational order between these stationary phases as reflected by the values of

I[Va(CH2)]/I[Vs(CH2)]. Particularly pronounced is the difference in conformational order

between TFC18SF (6.45 fimol/m^) and MFC 18 (3.09 fimol/m^). Furthermore, when the

temperature dependence of the Raman spectral response between 263 K (Figure 5. la) and

333 K (Figure 5.1b) is considered, a significant difference in alkyl chain order is

observed and is especially apparent for the highest surface coverage phase (6.45

jimol/m^). When considering these data, it is important to recall that the intensity of the

Va(CH2) mode in MFC 18 may be slightly elevated by the presence of the v(CH3)si mode

resulting in a slightly higher than expected value of I[Va(CH2)]/I[Vs(CH2)]. Thus,

stationary phase materials containing Si-CH? groups may appear slightly more ordered

than they truly are.

Effect of Temperature on the Conformational Order of Alklysilane Stationary Phases

When the 1 [Va(CH2)]/I[v,s(CH2)] order indicator was systematically examined

from 258 to 343 K, the dependence on temperature became even more apparent for each

of the stationary phase materials examined. Figure 5.2 shows plots of

I[ Va(CH2)]/I[Vs(CH2)] as a function of temperature for the five high-density phases

examined here. Although the magnitudes of the changes observed in this parameter are

187

CO ^

K 1.4 3

i

t 250 275 300 325

Temperature (K)

Figure 5.2 I [ v^CCHj)]/! [ Vj,{C Hj) ] as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•). The error bars represent one standard deviation.

188

small, the data exhibit some scatter, and the temperature dependence is in some cases

complex, general trends in the behavior of this parameter with coverage and temperature

are unmistakably observed that allow important conclusions to be drawn regarding

stationary phase conformational order as a function of these two variables. First, at any

given temperature, the magnitude of I[Va(CH2)]/I[Vs(CH2)] increases with surface

coverage indicating an increase in order with increasing surface coverage. Also, for any

given surface coverage, I[Va(CH2)]/I[vs(CH2)] decreases with increasing temperature

indicating that disorder is introduced with increasing temperature.

The three important regions of conformational order indicated by the magnitude

of the I[Va(CH2)]/I[Vs(CH2)] values is indicated by different shading in Figure 5.2. The

two highest surface coverage stationary' phases exhibit values of I[Va(CH2)]/I[Vs(CH2)]

that range from almost crystalline-like order at the lowest temperatures to a state in which

the onset of gauche conformers occurs at the higher temperatures. In contrast, the

stationary phases with the lowest three surface coverages exhibit values of

I[Va(CH2)]/I[Vs(CH2)] that start in the region in which the onset of gauche conformers

occurs below room temperature. In other words, these stationary phases exhibit more

liquid-like disorder even at the lowest temperatures investigated here. The

I[Va(CH2)]/l[VsfCH2)] values of these stationary phases extend down to the region where

the rapid addition of gauche conformers occurs as the temperature approaches and

exceeds room temperature. Thus, for each stationary phase, systematic trends in the

189

I[Va(CH2)]/I[Vs(CH2)] values as a function of temperature are indicative of significant

differences in conformational order that might reflect phase changes.

The temperature dependence of each stationary phase can be fit to two

approximately linear portions, albeit with scatter. Nonetheless, each stationary phase

exhibits a change in the rate at which I[va(CH2)]/I[Vs(CH2)] varies with temperature. The

intersection of the two best-fit lines for each stationary phase material is the break point

temperature (TB) for that particular material. The temperature at which this break point is

observed is generally higher as the stationary phase coverage decreases (Table 5.3). This

break point is thought to represent a phase change for the material. Moreover, the

magnitude of I[Va(CH2)]/I[Vs(CH2)] at which this break point occurs for stationary phases

of different surface coverage reveals important differences in stationary phase

organization.

The break point in the rate of change of I[Va(CH2)]/I[Vs(CH2)] for the two highest

surface coverage stationary phases (both fabricated from trifunctional precursors) is

observed at lower temperatures than those for the lower surface coverage materials;

moreover, this break point is observed in the region in which the onset of gauche

conformer formation occurs. In light of the relatively high surface coverage of the alkyl

chains in these two stationary phases, and the relatively large volume that

accommodation of a gauche conformer requires, the break point clearly occurs when the

stationary phase cannot easily accommodate additional gauche conformers in its existing

190

Table 5.3. Break point temperatures and slopes from temperature-dependent I[Va(CH2)]/I[Vs(CH2)] measurements

Stationary Surface Coverage TB (°C) LOW Temperature High Temperature Phase Slope ("C') Slope ("C')

TFC18SF 6^5 5 -8.0 x 10"' -3.0 x lO"'

TFC18SL 5.26 -2 -1.2 x 10"' -3.4x10'^

DFC18SF 5.00 22 -5.2 x 10"' -2.4 x 10"'

DFC18SL 4.17 38 -3.5x 10"^ -1.2 x 10"^

MFC18 3.09 37 -2.7 x 10"' -4.0 x 10"^

191

free volume and requires considerable additional thermal energy to do so. Thus, the rate

at which these gauche conformers are added decreases beyond this break point.

For the lowest three surface coverage stationary phases, which inherently possess

greater free volume, many more gauche conformers can be accommodated with a smaller

amount of additional thermal energy. Thus, the break point in the rate of change of

I[va(CH2)]/l[Vs(CH2)] with temperature is observed at higher temperatures, and most

likely corresponds to the point at which the addition of more gauche conformers is

statistically less likely.

A less obvious feature of the confonnational order-temperature relationships for

the four stationary phases of lowest surface coverage is that the slopes of the two portions

of the best linear fits around the break point are related to surface coverage (T able 5.3),

with higher surface coverage generally resulting in a steeper slope in each temperature

region. The slopes for the stationary- phase with 6.45 fimol/cm^ are similar to those for

the 5.26 |imol/cm^ stationary phase, although slightly smaller, resulting in a variance with

the trend just identified. It is imperative the reader bear in mind that the slopes of these

lines represent only the rate of change of the gauche population whereas the magnitude of

I[Va(CH2)]/I [VsCCHj)] reflects the total population of gauche conformers and rotational

disorder in the alkyl chain environment. Thus, the apparently counter-intuitive decrease

in slope with decreasing surface coverage for the lowest four surface coverage stationary

phase materials reflects the fact that the rate at which these systems add additional gauche

conformers with increasing temperature decreases even though the lower surface

192

coverage materials are inherently more disordered and possess an absolute greater

number of gauche conformers as indicated by the magnitude of I [ Va(CH2)]/I[ Vs(CHj)]).

First-Order Phase Transition Treatment of Alkylsilane Stationary Phases

In assessing whether these temperature-induced changes in l [Va{CH2)]/l[Vs(CH2)]

are equivalent to phase changes, it is important to review first-order phase transitions.

For the first-order melting of a solid to a liquid, where the density of the solid is greater

than that of the liquid, typical for alkanes, the phase transition line in the phase diagram is

described by the following relationship

where P is pressure, T is temperature, Ppef and TRd- represent an arbitrary starting

reference pressure and temperature, respectively, AH&s is the molar enthalpy change of

fusion, and AV&s is the molar volume change of fusion. Taking into account that PKOT and

TRef are arbitrary constants on the phase line, this expression can be rearranged to

P - Puef + (AHfus/AVfus) In (T/TRef) (1)

P = PRrf + (AHWAV&s) In T - (AHWAVFES) In TR^ (2)

At any point on the phase transition line, the slope of its tangent is given by

dP 1 — =(AH^ , /AV^ , ) -

193

(3)

Therefore, a plot of dP/dT as a function of 1/T should be linear with a slope of

(AH&s/AVfus) for a simple first-order phase transition from a solid to a liquid.

The stationary phases studied here are considered a homologous series of similar

materials, each of which exhibits its own first-order solid-to-liquid phase transition, based

on the conformational order behavior shown in Figure 5.2 and the fact that they are

prepared from difierent silane precursor molecules. Thus, considering this series from

the lowest to highest alky 1 silane surface coverage, the phase transition line is expected to

systematically shift to the right on the phase diagram, reflecting the greater temperature

required to cause "melting" at a given pressure. A schematic of the expected phase

transition lines is shown in Figure 5.3 for five components of such a homologous series.

Within a given temperature range, AT, for this series, the slope of the phase transition line

at any given temperature, dP/dT, systematically decreases with alkylsilane surface

coverage according to eq 3.

If the inverse of I[Va(CH2)]/I[Vs(CH2)] is plotted as a function of temperature for

each stationary phase material, the shapes and relative positions of the resulting plots

resemble those of the phase diagrams in Figure 5.3. This suggests a proportionality

between pressure and I[Va(CH2)]/I[vs(CH2)]. Defining R as follows.

0<

H

H

a

•• • •• • •• • •• • I I I

•• •

•• • + •• •

•• •

Temperature

Figure 5.3 Schematic of phase diagrams for a homologous series of materials with increasing surface

coverage, F. 4i.

R = I[Vs(CH2)]/I[Va(CH2)]

195

(4)

Then a proportionality between pressure and R is indicated:

R « p = PRef + (AHfUs/AVfUs) In T - (AH&s/AV&s) In iRef (5)

R = A'PRe,- + X(AHfos/AVfas) In T - ^(AH&s/AV&O In TR^- (6)

where K is the unknown proportionality constant. Thus,

If the temperature-induced changes observed in the Raman spectral response of these

stationary phases reflect phase changes, their behavior should follow eqs 6 and 7.

Figure 5.4 shows plots of i? as a function of temperature for these stationary

phases (symbols) along with the best (i? = a In T + b) fits of the data (lines), where a =

(K AHfus/AVfbs)) and b = {AT>Ref - AHWAVius) In Tudj are the variable parameters.

Although there is some scatter in the data, the logarithmic fits are acceptable based on the

calculated R^ values of 0.9 or greater in all cases.

LOS

LOO

0.95

0.90

0.85

^ 0.80

0.75

0.70

0.65

0.60 250 275 300 325 350

Temperature (K)

Figure 5,4 R as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•). The error bars represent one standard deviation. Best fits for /? = a In T + b shown as solid lines.

197

The slopes of the best fit lines, dR/dT, were determined as a function of

temperature and are plotted as a function of inverse temperature in Figure 5.5a. As

predicted by eq 7, these Clapeyron plots are linear with near-zero intercepts. Figure 5.5b

shows the slopes of these Clapeyron plots, which reflect the magnitude of {K AHtus/AV&s)

for a phase change, as a function of alkylsilane surface coverage. For a simple first-order

phase transition, the slope of the Clapeyron plot represents the tangent to the solid/liquid

phase line at a particular temperature; as shown by the phase diagrams in Figure 5.3, this

slope is expected to decrease with increasing temperature or pressure for a phase

transition for a bulk material. However, for a surface-confined system with a fixed

surface density such as the stationary phases described here, AHtm for the phase change is

expected to increase to a certain point then level oflF with increasing surface density, or

surface coverage. In contrast, AVfUs is expected to decrease and then level off with

increasing surface density. As a result of these trends, the slope of the Clapeyron plot

should increase and then level off with increasing surface coverage, assuming K is the

same for all five materials. The slopes of the Clapyeron plots for the four lowest

coverage stationary phases follow these trends: increasing with increasing surface

coverage, then leveling off.

However, a deviation from this trend is observed for the highest coverage

stationary' phase, which is attributed to a breakdown of the assumptions used to treat

these data. Specifically, the assumption that TFC18SF is a part of the homologous series

of similar materials is incorrect. Examining Figure 5.2 it is observed that at all

0.0036

0.0033

0.0030

0.0027

^ 0.0024

§ 0.0021

0.0018

0.0015

0.0012

0.0025 0.0030 0.0035 0.0040 0.0045

1/T (K"')

AA TFC18SL

DFC18SF

TFC18SF

DFC18SL

MFC 18

T ' 1 ' r

0.85

0.80

^ 0.75 o K a 0.70 0

g, 0.65 H

^ 0.60

1 0.55-

0.50

0.45

2 Surface Coverage (|amol/m )

Figure 5.5 (a) Clapeyron plots of dR/dT from best fit lines in Figure 5.4 as a function of 1/T. TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•). (b) Slope of Clapeyron plots as a fiinction of alkylsilane surface coverage.

199

temperatures, the minimum value of I[Va(CH2)]/I(Vs(CH2)] is ~ 1.2, where there is no

significant fraction of gauche conformers in the alkyl chains. Thus, over the temperature

range examined, the alkyl chains on the TFC18SF surface are solid-like and do not

undergo a solid-liquid phase transition.

The simple observation of temperature-dependent changes in alkyl chain order is

critical to describing the surface structure and extent or even presence of short- and long-

range order in the alkyl component. Chromatographic manifestations of phase changes

are nonlinear van't HotT relationsliips. Many previous examinations of chromatographic

behavior and stationary phase structure based on Van't HofFplots have resulted in the

observation of either no distinct phase change, or a very subtle transition^^" '

similar to those observed here. When such nonlinearities have been observed, they

were attributed to changes in the nature of the alkyl chain condensed phase at a particular

temperature under specific solvent conditions.^ Phase changes are

thus indirectly reflected in chromatographic retention data. The vast majority of these

studies have been carried out on stationary phases with surface coverages below 3 .0

pmol/m^ (surface coverages typically employed for analytical separations). In light of

the data presented here and the observation that the temperature dependence is generally

described by the Clapeyron relationship, it is not surprising that phase transitions were

not observed or were very shallow in systems with low alkylsilane surface coverages.

In examining Figure 5.2, it appears that the differences in conformational order

between stationary phases decrease with increasing temperature. This effect is more

200

evident when 1 [Va(CH2)]/[{ Vs( CH2)] is examined as a function of surface coverage at

several temperatures (Figure 5.6). This plot suggests that the conformational order

approaches some asymptotic value of I [ Va(CH2)]/I [ VsCCHj)] dictated by the temperature

as opposed to the surface coverage. This value is indicative of the maximum extent to

which the surface-bound alkyl chains can be disordered. It is interesting to note that this

asymptotic value of I[Va(CH2)]/I[Vs(CH2)] is larger than that observed for liquid alkanes.

In other words, the alkyl chains in these stationary phases cannot become as disordered as

in a bulk liquid. This result is not surprising considering that alkyl chains used as

stationary' phase materials are bound to the surface, thereby restricting the mobility of one

end.

Up to this point in the discussion, only I[Va(CH2)]/I(Vs(CH2)] has been used to

describe changes in conformational order in these systems. As noted above, however, the

frequency at which these bands are observed is also reflective of alkyl chain order. The

temperature dependence of the VsCCHj) frequencies is shown in Figure 5.7. In general, for

temperatures below 313 K, the trend in peak frequency mimics the trend observed in the

inverse of the intensity ratio of these bands shown in Figure 5.2. Furthermore, the peak

frequency decreases with increasing surface coverage reflecting a greater extent of

interchain coupling (or crystallinity). Differences in this parameter at low temperatures

most likely result from increased chain spacing with decreasing surface coverage. At

high temperatures (>323 K), very little difference in interchain coupling is observed in

each phase as reflected by similar Vs(CH2) peak frequencies. This similarity indicates

201

1.6-

1

2^ 1.4-U

^ 1.3-T

5" 1.2^ I i

1.0-

T • A

• 1

X

«

4

3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5

Surface Coverage (|imol/m^)

Figure 5.6 ^v/CHj)]/ I[v/CH2)] as a function of surface coverage at 258 (•), 293 (•), and 333 K (•).

202

* t cr 2850

260 280 300 320 Temperature (K)

340

a O.

<u £

Hi a,

S' u

2854

2852-

2850

2848

2846

r ' . j • i t 4 . ^ o i l

§ I ^

> t * f

260 280 300 320

Temperature (K) 340

r *£. ol

i

CD

O

S-P5

P a>

fp

Figure 5.7 v/CHj) peak frequency as a function of temperature for (a) TFC18SF mi TFC18SL (A), DFC18SF (•), (b) DFC18SL (O), and MFC18 (•). The error bars represent one standard deviation.

203

some minimum value of interchain coupling that must result from either a physical

separation of the alkyl chains, or for higher surface coverage materials, from thermal

agitation of the chains. These effects are most likely convoluted in the data presented

here due to the use of different precursor molecules and grafting protocols and two

different types of silica supports.

The manner in which the peak frequency changes as a fimction of temperature

differs between these phases as well. The highest surface coverage phase, TFC18SF,

exhibits sigmoidal behavior in the Vs(CH2) frequency with temperature revealing an

apparent phase change at -313 K. The remaining phases (except for MFC 18) exhibit

VS(CH2) frequencies that level off at temperatures below 273 K, then increase in a linear

fashion with increasing temperature. In contrast, the frequency of the Vs(CH2) mode in

MFC18 is relatively temperature-independent.

It is interesting to note that the inflection points in the Vs(CH2) frequency-

temperature behavior for the two stationary phases prepared with trifiinctional precursors

(TFC18SF and TFC18SL) do not correlate with the transition temperature indicated by

the I[Va(CH2)]/I[Vs(CH2)] values shown in Figure 5.2. One must recall the exact nature of

conformational order indicated by each of these parameters to obtain a clear molecular

interpretation of this apparent discrepancy. In contrast to the value of

I[Va(CH2)]/IVs(CH2)], which reflects any deviation from the all-trans arrangement, the

frequency of the VsCCHa) depends only on interchain coupling. It is possible to envision a

situation in which alkyl chains add kinks or twist around the long axis of the chain

204

(resulting in a decrease in I[Va(CH2)]/I[Vg(CH2)]) with the chains remaining in close

proximity to one another, or in other words, having insufficient thermal energy to

overcome interchain coupling. An inflection point on the Vs(CH2) frequency-temperature

curve would therefore indicate the temperature at which interchain coupling begins to

change. A closer examination of the data in light of this picture does indeed reveal

differences in interchain coupling between these phases. The data also indicate only

slight differences between the two stationary phases with similar surface coverages

(TFC18SL, 5.26 |nmol/m^ and DFC18SF, 5.00 fimol/m^). This response is expected if the

alkyl chains are distributed relatively homogeneously over the surfacc.

Combining the results of Figure 5.2 with those of Figure 5.7, a more complete

picture of the temperature-dependent behavior is revealed. Beginning with the highest

surface coverage system, the molecular picture indicated is one in which the alkyl chains

become more disordered with temperature, although in the confined area within which

the each chain can move (25.7 compared with 20.8 for the maximum theoretical

surface coverage of 8 jumol/m^), a significant population of gauche bonds can not

develop. Near 273 K, the alkyl chains undergo a phase change without altering

significantly the extent of alkyl chain coupling. Such a phase change may result from a

change in alkyl chain tilt angle relative to the surface. At temperatures >323 K, the

chains possess significant thermal agitation to overcome interchain coupling, as reflected

by the increase in Vs(CH2). While interchain coupling appears to be minimized at

temperatures above 323 K, large changes in I[ Va(CH2)]/![Vs(CH2)] are not observed.

205

indicating that the maximum amount of deviation from all-trans behavior that can be

sustained in these chains has been achieved. The resulting molecular picture would

therefore be one in which the alkyl chains undergo rapid but sterically-restricted motion

that does not result in the development of large populations of gauche bonds (Figure 5.8).

As the surface coverage decreases, the aJkyl component adopts a more disordered

state. The monofunctional material examined here (MFC 18) has larger interchain

distances (58 A^/chain) compared to the higher surface coverage materials. This

conclusion is consistent with the lack of any temperature-dependence of the Vs(CH2)

frequency. Considering the length of the Cig chain (23 A), homogeneously distributed

alkyl chains at this bonding density must deviate significantly fi"om the all-trans

configuration in order to interact (Figure 5.9). Figure 5.2 reveals that the alkyl chains in

this material are more disordered than in the higher surface coverage phases, but the

increased distance between chains allows for such disorder through the formation of

gauche defects in the chains.

Finally, when considering possible differences in architecture of the alkyl chains

in the two stationary phases with similar surface coverages, the situation shown in Figure

5.10 is proposed. Although the Clapeyron plot suggests that the alkyl component is

distributed homogeneously across the surface, it is possible to have local heterogeneity in

the difunctional materials that results from asymmetry in the head group. A ramification

of this asymmetry is the possibility of alkyl chains separated by two distances, one

dictated by the methyl side chains on the head group and the second dictated by the

Heat

Figure 5.8 Proposed temperature-dependent disordering process for

TFC18SF.

1 Heat

Proposed temperature-dependent disordering process

MFC 18.

process for

209

Si-O-Si bond. Such local heterogeneity could alter the transition temperature through a

decrease in overall interchain coupling.


A detailed examination of the temperature-dependent behavior of a series of high-

density alkylsilane stationary phases with differing surface coverages is presented.

Results indicate that conformational order and interchain coupling of the alkyl

component of these phases are sensitive to temperature and surface coverage.

Preparation procedure and nature of the alkylsilane precursor do not appear to

significantly affect the chain conformational order or interchain interactions. Each of the

stationary phases exhibited subtle temperature-dependent phase changes that are

generally described by the Clapeyron relationship. The temperature at which the phase

change occurs is postulated to be reflective of the condensed phase adopted by the alkyl

component or may be reflective of the grafting procedure employed in the preparation of

the phase. The results presented here complement the chromatographic observations

made by Sander and coworkers in the correlation between surface coverage and

selectively of the phase for planar molecules.These results also complement the

observations made by Pursch and coworkers based on NMR spectroscopy that the

fraction of gauche defects decreases and interchain coupling increases with increasing

surface coverage.

Future directions for this work include the temperature-dependent phase behavior

of other homologous series of stationary phases of varying chain lengths, to better

understand phase behavior of shorter and longer alkylsilanes. In addition, deriving a

quantitative relationship between R, I[Vs(CH2)]/I[Va(CH2)], and surface pressure is also of

interest such that quantitative thermodynamic values, AHfUs and AV&s, for these stationary

phases can be determined from the values of Raman conformational order indicators.

This could be accomplished by using the Clapeyron treatments of R vs. temperature data

for bulk alkane materials, such as polyethylene, for which AH&s and AVtus are known.

Thus, a proportionality constant between R and pressure could be determined; however,

its relevance to surface confined alkylsilanes would still remain an unanswered question.

Appendix 5.1

This appendix contains Raman spectra of TFC18SF, TFC18SL, DFC18SF,

DFC18SL, and MFC 18 in air at 258, 273, 283, 293, 303, 313, 323, and 343 K.

TFC18SF

TFC18SL'

DFCI8SF

DFC18SL

MFC18

2900 3000 2800

TFC18SF

C I SSL

DFC18SF

DFC18SL

MFC 18

2800 2900 3000

JFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

3000 2800 2900

Wavenumbers (cm"0 Wavenumbers (cm"') Wavenumbers (cm"')

Figure A5.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 at (a) 258 K, (b) 273 K, and (c) 283 K. Integration times are 10-min for all spectra. w

TFC18SF TFC18SF

TFC18SL TFC18SL

DFC18SF DFC18SF

DFC18SL DFC18SL

MFC18 MFC 18

3000 2900 2800 2900 3000 2800

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC 18

2900 3000 2800 Wavenumbers (cnr\) Wavenuinbers (cm-\) Wavenumbers (cnrO

Figure A5.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 at (a) 293 K, (b) 303 K, and (c) 313 K. Integration times are 10-min for all spectra. w

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

TFC18SL

^FCISSF

DFC18SL

MFC18

3000 2900 2800

Wavenumbers (cm"^) Wavenumbers (cm-^)

Figure A5.3 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 at (a) 323 K and (b) 343 K. Integration times are 10-min for all spectra. M

214

Chapter 6

Stnicture-Function Relationships in Ocladecylsilane Stationary Phases by

Raman Spectroscopy; Effect of Common Mobile Phase Solvents

Introduction

The use of alkyl-modified silica particles for the separation of molecules by

reversed-phase liquid chromatography has evolved into a popular standard analytical

technique. The least understood aspects of retention are intermolecular interactions

between each chromatographic component (bonded alkyl chain, solvent, and solute) over

the course of a separation. Such interactions include solvent-solute, solute-stationary

phase, solvent-stationary phase, and the intermolecular interactions of the alkyl chains

within the stationary phase. The importance of each of these interactions is obvious in

the context of solute retention. Of equal importance are the effects that changes in

chromatographic parameters (solvent, temperature, nature of the alkylsilane, alkylsilane

bonding density, and architecture) have on these interactions.

In Chapter 5, the effect of temperature, surface grafting procedure, surface

coverage, and identity of the alkylsilane precursor on conformational order of several

high density alkylsilane stationary phases'*'"^'' was examined. Here, the effect of mobile

phase is examined in the context of each of these chromatographic parameters in an effort

to clarify the role of each parameter in solute retention on such phases.

215

The effect of solvent on the intemiolecular interactions of alkyl chains has been

examined in numerous studies.^'""'^^ While many of these have been

chromatographic/'resulting in inferred solvent effects on the structure and order of

the stationary phase, several have been spectroscopic/'' attempting to

directly observe the effects of stationary phase solvation. From these investigations, two

models of solute retention have been proposed. The solvophobic (also called the

hydrophobic) modeftreats the stationary phase as a passive entity. In this model

the stationary phase plays no role in the separation other than to provide a sorption site

for the solute. The solvophobic model treats the separation as an adsorptive, rather than

partitioning, process in which the solvation characteristics of the solvent for a particular

solute dictate the extent of solute adsorption. In the partitioning model (also called the

'J F\ '2 "3 interphase model) ' ' the solute can become foUy embedded within the chains of the

stationary phase, thereby acting to compete with the mobile phase solvent to "solvate" the

solute. Thus, the stationary phase in this model plays an active role in retention with such

solute partitioning inducing changes in conformational order of the alkyl chains of the

stationary phase. Although these models specifically address the interaction between

solute and stationary phase, solvent-stationary phase interactions should be similarly

described.

The two models of solute retention represent clear pictures of the required

intermolecular interactions during a separation; however, neither model may accurately

describe the retention of every solute under all conditions. For example, several pieces of

216

evidence suggest that solute retention can not be completely described by either the

solvophobic or partitioning models including: 1) the observation that no single equation

h a s b e e n d e r i v e d t o a c c o u n t f o r t h e r e t e n t i o n b e h a v i o r o f a s o l u t e , ^ 2 )

obser\'ations'''" that solute retention becomes increasingly entropic with increasing

density of the alkyl component, and 3) wetting studies that suggest that solute or solvent

molecules are not always completely embedded within the stationary phase.''

In order to observe changes in stationary phase architecture or conformational

order, as well as to examine changes in the chemical environment within the stationary

phase, techniques sensitive to such changes must be employed. Spectroscopic methods

such as NMR, infrared (IR), and Raman spectroscopies are ideal for such investigations.

While each technique has its own unique advantages, limitations of IR and NMR have

precluded their widespread use in the examination of alkylsilane stationary- phases. The

use of NMR spectroscopy requires a soHd-state approach, while TR spectroscopy suffers

from spectral interferences from surface silanol groups of the stationary phase and water,

either sorbed on the surface or employed in the mobile phase.

As previously demonstrated, Raman spectroscopy is an ideal technique for

examining the intermolecular interactions of the alkyl chains in reversed-phase

chromatographic materials^ due to its ability to provide a direct indication of the

conformational order of the alkane component. Raman spectroscopy measurements are

also free from spectral interferences from water and silanols that adversely affect IR

measurements. Finally, Raman spectroscopy measurements can be carried out in the

217

presence of solvent if care is taken to choose solvents that do not have interfering

vibrational modes in spectral regions of interest. Thus, as part of an ongoing effort to

elucidate retention mechanisms in reversed-phase liquid chromatography/'^'"*'our

studies were expanded to include the effects of solvent in addition to those parameters

previously investigated.

When stationary phase solvation is examined according to simple solvent

classifications, including nonpolar solvents (possessing a weak dipole moment), polar

solvents (possessing a strong dipole moment), and water, some trends are observed. In

many cases, nonpolar solvents have been reported to disorder the alkyl chains of the

stationary phase.®'^"^'^ Such disordering has been suggested to be the result of deep

intercalation or partitioning of the solvent into the stationary phase layer (fully solvating

the alkyl chains). In contrast, polar solvents have been reported to have little to no effect

on stationary phase order^ suggesting an adsorptive interaction of the solvent

with the stationary phase surface. In fact, several reports have indicated that polar

solvents such as methanol increase alkyl component order relative to water.^'^'^'^""^'^^ In

water, the stationary phase has been described as collapsed in which the alkyl chains act

to exclude water from the interior of the phase.

A few direct observations of alkyl chain conformation in the presence of solvent

have been made by 6.12 Raman spectroscopies.^'

An early examination of octadecylsilane (Cis) stationary phases in the presence of

methanol and methanol/water mobile phases found that the alkyl chains of the stationary

218

phase were ordered by these polar mobile phase solvents.^ This was identified through

the observation of a decrease in the number of carbon-carbon gauche bonds relative to the

stationary phase material in the absence of solvent. The molecular picture presented to

explain these results was one in which the methanol molecules intercalate between the

alkyl chains with the nonpolar methyl group embedded in the stationary phase material

and the polar hydroxyl group extending out of the surface.

While the effect of solvent on intermolecular interactions of the stationary phase

is not yet clearly defined, the additional effect of temperature may be even less

understood. Temperature effects on solute retention have generally been investigated

using a van't Hoff analysis of the dependence of the capacity factor on temperature.

While this approach has been useful in identifying temperatures at which a change in the

retention mechanism is observed (resulting from a phase change within the stationary

phase), the resulting changes in alkyl chain architecture must be inferred from the data.

Goals of this Work

The main goal of this work was to acquire Raman spectra of Cig stationary phases

in the presence of solvent without spectral interference. These spectral data were used to

examine the role of solvent and temperature as well as the interplay between them in

dictating the conformational order of several previously described^'octadecylsilane

stationary phases. In addition, the effect of surface grafting method, silane precursor

identity, and surface coverage was also examined as a fiinction of solvent and

219

temperature. Results were compared to previous chromatographic and spectroscopic

investigations of stationary phase order in the presence of different classes of solvents.

Methodology


characterized as previously described in published work'' and in Chapter 2 of this

Dissertation. Perdeuterated water (99.9%) was purchased from Aldrich. Additional

perdeuterated solvents used in these experiments including methanol (99.8%),

acetonitrile (99.8%), acetone (99.9%), tetrahydrofuran (THF) (99.5%), chloroform

(99.8%), benzene (99.5%), toluene (99.8%), and hexane (99%) were obtained from

Cambridge Isotope Laboratories, Inc. All solvents were used as received.

Samples were prepared by placing between 25 and 100 mg of stationary phase

material into a 5-mm dia NMR tube; 200 jiL of solvent was then added. Samples were

sonicated for 10 min and equilibrated at 20 °C for a minimum of 12 h prior to spectral

acquisition. Raman spectra were collected using 100 mW of 532 nm radiation from a

Coherent Verdi Nd:YV04 laser on a Spex Triplemate spectrograph as described in

Chapter 2. Slit settings of the Triplemate were (0.5/7.0/0.150) mm for all experiments.

Some of the spectral data presented in this chapter were collected by Dr. Michael W.

Ducey. All samples were equilibrated at the appropriate temperature for at least 30 min

prior to spectral acquisition. Integration times for each spectrum are provided in the

220

figure captions. Raw spectral data in the v(C-H) region were analyzed and processed as

described in Chapter 2.

Raman Spectroscopy of High-Density Alkylsilane Stationary Phases in Solvent

Absolute retention in RPLC is governed by numerous factors, but is most

influenced by mobile-phase composition. Understanding mobile phase solvent-stationary

phase interactions at the molecular level is key to elucidating the molecular basis of

retention. Given the small values of the partition coefficient for analytes in a single

theoretical plate (K ~ 2 to 100), with the associated small change in free energy of

partitioning (AG < 5 kcal/mol), the concomitant changes in molecular structure of the

stationary phase are expected to be extremely subtle. Thus, a technique extraordinarily

sensitive to such small variances in molecular structure is required in order to probe the

changes that form the basis of retention. Previous reports from this laboratory

demonstrate that Raman spectroscopy is one method sensitive to small variations in

alkylsilane stationary phase order and structure in the presence of solvent as a measure of

these intermolecular interactions,^'Raman spectroscopy is a powerful tool for the

investigation of alkylsilane stationary phase structure due to the plethora of spectral

indicators of conformational order, especially in the v(C-H) region between 2750 and

3050 cm"'. Spectral interrogation of conformational order in this region is achieved in

the presence of solvent without significant spectral interference from the silica substrate

and surface-bound water. Although the spectral region between approximately 900 and

1500 cm"' containing the v(C-C) and v(C-H) modes also contains a number of significant

conformational order indicators for alkane systems, these high-density stationary phase

materials were prepared on silica particles that possess considerable inherent fluorescence

that interferes with acquisition of Raman vibrational information in this region.

Therefore, for these systems, conformational order assessment was restricted to the high-

quality spectral data that can be obtained in the v(C-H) region.

Raman spectra for each of the five stationary phase materials at room temperature

in nine solvents, methanol, acetonitrile, water, acetone, THF, hexane, benzene, toluene,

and chloroform, are shown in Figures 6.1-6.3. Spectra for the five stationary phases in

these nine solvents at temperatures from -15 to 50 "C are shown in Appendix 6.1.

Information about conformational order of these alkane systems is available from

multiple indicators in their Raman spectra. These conformational indicators were

described in detail in Chapter 3. Several of these indicators come from the v(C-H) region

of alkane systems between 2750 and 3050 cm"'. Although this region is complex,^'"^"^'®'^^

containing up to 14 possible vibrational modes, several indicators of conformational

order are contained in this region. One very sensitive conformational indicator is the

peak intensity ratio of the Va(CH2) (2885 cm"') to the Vs(CH2) (2850 cm"'),

I[Va(CH2)]/I[Vs(CH2)]. The behavior of this indicator was described in detail for alkyl-

containing materials and stationary phases in Chapters 2-5. Briefly, it is sensitive to

changes in conformational order resulting both from rotational disorder in small portions

of the alkyl chain for alkanes in a more ordered state (and hence, its sensitivity) as well as

v/CH,) v/CH,)

TFC18SF

v,(CH3),^

TFC18SL

DFC18SF

DFC18SL

MFC 18

2800 2900 3000 Wavenumbers (cm"^)

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

2800 2900 3000

Wavenumbers (cm"')

K V^FC18SF

\TFC18SL

^N^FC18SF

f \ M F C 1 8

2800 2900 3000

Wavenumbers (cm*')

Figure 6.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in (a) methanol, (b) acetonitrile, and (c) water at 20 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

K) bJ to

VaCCH^)

^sCCH^),

TFC18SF

^s(^^3)fR

TFC18SL

DFC18SF

DFC18SL

v(CH,),^

MFC 18

2800 2900 3000 Waveniimbers (cm"')

/V ^->yTFC18SF

A f\ \TFC18SL

I'"-'

^yDFCI8SF

/ /X Wii-cis

2800 2900 3000

Waveniimbers (cm"')

V ^^TFC18SF

\TFC18SL

^\DFC18SF

\DFC18SL

\^C18

2800 2900 3000 Wavenumbers (cm"')

Figure 6.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in (a) acetone, (b) hexane, and (c) THF at 20 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC18 in all solvents are 5-min.

to U)

TFC18SF

TFC18SL

DFC18SF

DFC18SL

'C18

3000 2800 2900 Waveniimbers (cm"0

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC 18

3000 2900 2800

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

2900 3000 2800

Wavenumbers (cm'^) Wavenumbers (cm-')

Figure 6.3 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in (a) benzene, (b) toluene, and (c) chloroform at 20 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC18 in all solvents are 5-min,

225

from the development of true C-C gauche conformers for alkanes in a more liquid-like

state. Given the sensitivity of this indicator to small structural changes in alkyl chains,

this ratio spans a range from approximately 0.6 to 0.9 for aJkanes in the liquid state and

1.6 to 2.0 for alkanes in the crystalline state.^'^^'^'^'* It is important to note, however, that

while this ratio contains information about subtle attributes of alkyl chain order, it is not

linearly correlated with the number of gauche conformers in the alkyl chains, as

discussed in Chapter 3. Rather, this indicator reflects the total population of C-C bonds

that deviates in any way from the all-trans configuration, and is therefore sensitive to

gauche defects as well as rotations in the alkyl chain. While the perceptible number of

gauche defects (based on the intensity of the V(C-C)G at 1080 cm-1) remains negligible

for values of 1.4 to 1.6 for this intensity ratio, an increasing number of gauche defects is

observed for values of this ratio between 0.9 and 1.1 (Chapter 3). The existence of a

range of values for this indicator for a given chemical state suggests that small changes in

alkyl chain rotational order or number of gauche conformers in the alkyl chains are

precisely reflected by this parameter for the solid and liquid states, respectively, even

though these changes do not impart macroscopic changes in chemical state on the system.

The implication of this extreme sensitivity is that very small measured ditYerences in this

ratio for a given chemical system contain meaningful information about alkyl chain

behavior.

Acquiring Raman spectra from these systems in the presence of alkyl-containing

or aromatic solvents is difficult due to spectral interference from solvent Va(CH2) and

226

VS(CH2) modes, as discussed in Chapter 4. Solvent spectral interference can be

eliminated through use of perdeuterated solvents. However, additional spectral

interference arises from the v(CH3)si band (2905 cm"') in stationary phases prepared with

difiinctional and monofonctional alkylsilane precursors. Since this mode is superimposed

on the Va(CH2) band, Va(CH2) intensity values that are slightly high can result; therefore,

care must be exercised when comparing values of the I[Va(CH2)]/I[Vs(CH2)] indicator

between stationary phases prepared from different precursors.

In addition to changes in peak intensities, the peak frequencies of the Va(CH2) and

VsCCHi) stretching modes contain conformational order information as they shift to lower

frequency by 6 to 8 cm"' with increasing conformational order. This frequency shift is

the result of the symmetric and antisymmetric stretching vibrations dampened by

neighboring alkyl chains as they order in a more close-packed arrangement.

Effect of Solvent on Conformational Order

To examine solvation effects on conformational order of alkyl moieties relevant

to chromatographic separations, each stationary phase was exposed to a range of common

chromatographic solvents. The solvents examined included water, methanol, acetonitrile,

acetone, THF, chloroform, benzene, toluene, and hexane. These solvents exhibit wide

variability in their characteristics including polarity, dipole moment, polarizability,

hydrogen bond accepting/donating ability, hydrophobicity, shape, and size that have been

quantified according to the solvatochromic parameter, k*, developed by Kamlet et

227

and previously described in Chapter 4. These parameters are listed in Table 6,1. It is

important to note throughout this discussion the close relationship between n* and the

other solvatochromic parameters including hydrogen-bonding donating and accepting

ability (a and P, respectively), the Hildenbrand solubility parameter (5H), and the solvent

self-association parameter Figures 6.4 and 6.5 show the

I[Va(CH2)]/I[Vs(CH2)] for all five stationary phases as a function of the solvatochromic

parameter, n*, for polar (Figure 6.4) and nonpolar (Figures 6.5a and b) solvents at room

temperature (20 "C). The horizontal dashed lines in each figure are at the values of

I[Va(CH2)]/I[Vs(CH2)] for each stationary phase in air as a reference.

In the polar solvents methanol, acetonitrile, and water, stationary phase

conformational order is highly dependent on surface coverage, consistent with the

observations made for each phase in air, the subject of Chapter 5. For all stationary

phases except that with the lowest alkylsilane surface coverage (MFC 18, 3.09 pmol/m^),

conformational order decreases slightly in polar solvents with increasing n*. In addition

to these solvent trends, it is also important to note changes in conformational order in

solvents relative to air. For the phases prepared by surface polymerization, TFC18SF

(6.45 jimol/m^) and DFC18SF (5.00 (.imol/m^), the conformational order increases in

polar solvents relative to that in air. In contrast, the phases prepared by solution

polymerization, TFC18SL (5.26 |imol/m^) and DFC1 SSL (4.17 j.imol/m^), are more

disordered in polar solvents than in air. The lowest surface coverage material, C18MF, is

also more disordered in polar solvents than in air.

228

Table 6.1, Solvatochromic parameters", TC*, a, and P for solvents

Solvent 7t* a p

Water 1.09 0.18 1.17

Acetonitrile 0.75 0.31 0.19

Acetone 0.71 0.48 0.06

Methanol 0.60 0.62 0.93

Benzene 0.59 0.10 0

THF 0.58 0.55 0

Chloroform 0.58^ 0 0.44

Hexane -0.08 0 0

Toluene 0.54 0.11 0

"As described by Kamlet et al/' '^^ ''Actual value of n* for chloroform, but plotted as 0.57 in the figures to discriminate it from THF.

L6~

1.5-

^ L4-Air Air

B X/i ::t: ^ L2-r—1

J. 1 1 1

g 1-1-

1 1 i

11

f 1 1

1

^ 1.0-K-H i 1 ! i 1

0.9"-1 i 1 1

0.8-1 i 1

Methanol^

? T ¥

I

Acetonitrile

0,0 0.2 0.4 0.6 n*

.Water

OJ 1.0 L2

Figure 6.4 I[v.,(CH2)]/l[Vj.(CH2)] as a function of solvent solvatochromatic parameter, %*, for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC18 (•) in the polar solvents methanol, acetonitrile, and water at 20 "C. The error bars represent one standard deviation. to

Benzene Twc • Acetone

CHCJJ

Toluene

4 p. 1.1 -

f

230

-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

n*

1.6

X 1-3

S 1.2

1.0

OA

THF CHCl •

Toluene ; (Benzene

Acetone

f 1 ' '

1 t

1

u

1 z z i z & m z z z z z z z z z z z z z z z - _ L. 1

• -Tif

m f

T : Y —^ ' ' 1

1 1 • 1

1 1 1 1 I « 1 '

0.50 0.55 0.60 0.65

n* 0.70 0.75

Figure 6.5 I[Va(CH2)]/I[Vg(CH2)] as a function of solvent solvatochromatic parameter, n*, for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•) in the nonpolar solvents hexane, toluene, chloroform, THF, benzene, and acetone at 20 °C. (a) All solvents and (b) expanded view of the region containing data for toluene, chloroform, THF, benzene, and acetone. The error bars represent one standard deviation.

231

These observations are consistent with interaction of the polar solvents with the

surface-polymerized materials (TFC18SF and DFC18SF) primarily, according to an

adsorptive model, as a result of high surface coverage, solvent self-association ability,

and more importantly, surface homogeneity (as determined by NMR experiments

performed on these materials)/'' In other words, values of this order indicator

comparable to those observed in air for polar solvents with surface-polymerized

stationary phases suggest that these solvents do not intercalate into these stationary

phases to any significant extent. The stationary phase only presents a sorption site for

solvent interaction at the distal methyl end of alkyl chains extended away from the

surface. This picture of solvation for polar solvents indicates that solvent-alkyl chain

interaction is weaker than alkyl chain-alkyl chain interaction in these systems.

Polar solvent interaction with the solution-polymerized materials (TFC18SL and

DFC18SL) also appears to be primarily adsorptive; however, conformational order

differences relative to the surface-polymerized phases are a consequence of stationary

phase properties. These solution-polymerized materials arc lower in surface coverage

(comparing the surface-polymerized and solution-polymerized trifunctional systems and

the surface-polymerized and solution-polymerized difunctional systems, respectively),

but more importantly, have been proposed to be two-dimensionally heterogeneous with

isolated patches of densely-populated alkyl chains and regions with reduced alkyl surface

coverage. As a result of this heterogeneity in surface coverage, regions exist in which

the average chain spacing is greater and solvent molecules can interact with not only the

232

distal methyl groups, but with the underlying terminal methylene groups as well,

inducing slightly greater conformational disorder (and hence lower values of

I[Va(CH2)]/I[v.(CH2)]).

For C18MF in polar solvents, although possessing homogeneous alkylsilane

coverage, the relatively low surface coverage and steric hindrance between silane-bound

methyl groups give rise to large interchain spacings. Small angle neutron scattering

experiments on monofunctional stationary phases of comparable surface coverage

suggest that the aJkyl chains in such systems are bent and disordered on the silica surface,

thereby maximizing hydrophobic interactions/ " '' '^'^ In this case, solvent interaction with

the stationary phase is primarily partitioning, where the more hydrophobic solvent,

methanol, solvates the stationary phase more effectively (i.e., induces more disorder) than

the less hydrophobic solvent (water). As a result, the conformational order for MFC 18

increases with n*.

The conformational order of these materials in nonpolar solvents (Figure 6.5) is

also dependent on surface coverage, but is generally more disordered than in polar

solvents. The trend in conformational order with n* observed for the polar solvents is

absent for these materials in nonpolar solvents. When comparing the conformational

order of these materials in nonpolar solvents to those in air, the results are quite different

from those in polar solvents. An increase in conformational order relative to air is only

observed for the highest surface coverage material, TFC18SF, in hexane, benzene, THF,

and acetone. In this case, the high surface coverage dictates that solvent interaction can

only occur via adsorplive sites at distal methyl groups of the alkyl chains. The other

surface-polymerized material, DFC18SF, is disordered in all nonpolar solvents relative to

air. According to the adsorption solvation model for polar solvents described above,

solvation, and its concomitant impact on conformational order, is primarily a function of

surface homogeneity and less dependent on surface coverage. In nonpolar solvents,

solvation occurs according to a partitioning model and is more dependent on surface

coverage. Nonpolar solvents intercalate and solvate the alkyl chains to a greater extent

even on homogeneously covered surfaces, suggesting that solvent-alkyl chain interactions

rival alkyl chain-alkyl chain interactions in strength in these systems. Similar disordering

is observed for the solution-polymerized and monofunctional materials indicating that

nonpolar solvent interaction is partitioning in these systems, causing considerable

conformational disorder in the alkyl chains.

The complexity of the solvation mechanisms dictated by the molecular

interactions between solvents and the alkyl chains is better examined by considering (1)

similar conformational behavior of these materials in dissimilar solvents and (2)

distinctly different conformational order of the stationary phases in similar solvents. As

shown in Figure 6.5, the stationary phases TFC18SF, DFC18SF, and C18MF each have

statistically the same conformational order in toluene and chloroform at 20 "C despite

differences in solvent polarizability, dipole moment, hydrophobicity, dielectric constant,

size and shape. Conversely, the conformational order of the same stationary' phases is

different in toluene and benzene, two solvents with similar physical properties. The

234

disorder induced by chloroform is the result of its relatively small size and ability to

penetrate deep into the alkyl chains, while the planar toluene molecule is envisioned to

intercalate between chains and induce disorder primarily due to the steric bulk of the

methyl group. The disorder caused by the methyl group of toluene is rationalized to be

the difference in conformational order of the stationary phases in toluene and benzene.

The subtle differences in molecular interactions between these and other monosubstituted

aromatics and these high-density stationary phases will be considered in more detail in

Chapter 8.

Effect of Temperature on Conformation Order in the Presence of Solvent

Examination of the effects of temperature and solvent on the conformational order

of these stationary phase materials extends our understanding of the complex solvation of

these systems. Figures 6.6 and 6.7 show plots of I[va(CH2)]/I[Vs(CH2)] as a function of

temperature for TFC18SF in polar (Figures 6.6a) and nonpolar (6.6b) solvents and

C18MF in polar (Figures 6.7a) and nonpolar (Figures 6.7b) solvents. For TFC18SF in

polar solvents (Figure 6.6a), temperature-induced disordering of the alkyl chains was

observed with a convergence at higher temperatures to a minimum value of

conformational order that is independent of solvent. For TFC18SF in nonpolar solvents

(Figure 6.6b) and MFC 18 (Figure 6.7a) in polar solvents, disorder was primarily

temperature-induced with the retention of solvated conformational differences over the

entire temperature range. The conformational order for MFC18 in nonpolar solvents

I 5 U 1.4

1.2

235

-20 -10 0 10 20 30 Temperature ("C)

I

-20 -10 0 10 Temperature ("C)

Figure 6.6 l[v^(CH2)]/I[v^(CH2)] as a function of temperature for TFC18SF in (a) polar and (b) nonpolar solvents, methanol (•), water (A), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•), acetone (•), and THF (^). The error bars represent one standard deviation.

k ® 1 . 2 -

I

236

-20 -10 0 10 20 30 40 50 60 Temperature (°C)

1 .7 -

1.6-

1 .5-

5" 1 .4-93

> 1 .3-

o. X

1 ,2-

u 1 .1-> 1—1 1 ,0-

0 ,9-

0,8-.

! T—'—r

-20 -10 10 20 30 40 50 60 Temperature ("C)

Figure 6.7 as a function of temperature for MFCl 8 in (a) polar and (b) nonpolar solvents, methanol (•), water (A), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•), acetone (•), and THF (^). The error bars represent one standard deviation.

237

(Figure 6.7b) depended more on solvent than temperature. For hexane, acetone, and

chloroform, no changes in conformational order with temperature were observed. For

benzene, toluene, and THF, only a weak temperature dependence of conformational order

was observed, with order decreasing very slightly with increasing temperature. Although

difficult to interpret more definitively due to the fact that I[Va(CH2)]/I[Vs(CH2)] did not

change linearly with order, these results support a subtle interplay between temperature

and solvation in disordering of the alkyl chains of these stationary phases. This interplay

for different classes of solvents was observed not only for the highest and lowest surface

coverage materials examined here, but for the stationary phases of intermediate coverage

as well. Similar plots of I[Va(CH2)]/I[Vs(CH2)] as a function of temperature for

DFC18SF, TFC18SL, and DFC18SL are shown in Figures 6.8, 6.9, and 6.10,

respectively, where the interplay between solvation ability and surface coverage is not as

pronounced for the intermediate coverage materials, but remains apparent in the

I[Va(CH2)]/I[Vs(CH2)] results.

These results for the effect of temperature on alkyl silane stationary phases are

consistent with observations made by McGuffin et al. in which changes in molar enthalpy

and volume were observed for high and low surface coverage stationary phases in solvent

as a function of temperature.^'®*^ For high-density stationary phase materials, the molar

volume and enthalpy increase with temperature and plateau at some maximum value. In

contrast, the molar volume and enthalpy remain unchanged by increasing temperature for

238 1.7-

1.6-

1.5- -r 1.4-1 si U si

> 1-3- % 1.2-

S ' U 1 1 -Mi !=^ 1.0-

0.9-

0.8- 8" i !

1.7-

1.6-

1.5-

x ' 1.4-u

1.3-

1.2-

u 1.1-33

H 1.0-

0.9-

0.8-

a

i 1 5 I I T—'—i—I—I—I—r

-20 -10 0 10 20 30 40 50 60 Temperature (°C)

-10 0 10 20 30 Temperature (°C)

T—i—r 40 50 60

Figure 6.8 I[v.,( CHj)]/![v^{C)] as a function of temperature for TFC18SL in (a) polar and (b) nonpolar solvents, methanol (•), water (A), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•), acetone (•), and THF (^). The error bars represent one standard deviation.

1.7-

1.6-

1.5-

X 1.4-u -

X/l

> 1.3 -

§ 1.2--

u 1.1 -

H-! 1.0-

0.9-

0.8-

1.7-

1.6-

1,5-

X 1.4-P, •

us > 1.3-

-

1.2--

1.1 -eS > -

1—1 1.0-

0.9-

0.8-

239

0 10 20 30 40 50 60 Temperature (°C)

I

-20 -10 0 10 20 30 40 50 60 Temperature ("C)

Figure 6.9 l[vJCH2)]/l[Vs(CH2)] as a function of temperature for DFC18SF in (a) polar and (b) nonpolar solvents, methanol (•), water (A), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•), acetone (T), and THF (^). The error bars represent one standard deviation.

1.7-

1.6-

1.5-r—1

rs 1.4-

u 1.3-> 1.3-

1.2--

u 1,1 -1,1 -> •

1.0-

0.9-

0.8-

1,7-

1 .6-

1.5-r—1

-fS X 1,4-,u. -

1.3 --

1,2-x' -

u 1.1 -> -

HH 1,0-

0.9-

0 8 -

240

-20 -10 0 10 20 30 40 50 60

Temperature ("C)

40 50 1—i—r -20 -10 0 10 20 30

Temperature (°C)

Figure 6,10 Ifv,(CH2)]/I[v/ClT2)] as a function of temperature for DFC18SL in (a) polar and (b) nonpolar solvents, methanol (•), water (A), acetonitrile (O), toluene (•), benzene (A), hexane (•), chloroform (•), acetone (•), and THF (^). The error bars represent one standard deviation.

low-density materials. McGuffin's work again illustrates the interplay between

temperature- and solvent-induced disorder in chromatographic stationary phases.

241

Subtle DitYerences in Temperature Effects Between Differently Solvated Stationary

Phases

The convolution of temperature and solvent in dictating conformational order in

these stationary phases and the distinct differences each of these variables has on the

order of a specific stationary phase material is shown more clearly in the plots in Figures

6.11 and 6.12. In these plots, I[Va(CH2)]/l[Vs(CH2)] is shown as a function of temperature

for each of the five stationary phases in four solvents, methanol (Figure 6.1 la),

acetonitrile (Figure 6.1 lb), chloroform (Figure 6.12a), and toluene (Figure 6.12b), chosen

to represent the broad range of temperature- and solvent-induced effects on order. For

these materials in air, changes in slope of I[VA(CH2)]/I[VS(CH2)] with temperature are

observed that are sensitive to the nature of the alkylsilane precursor (Chapter 5). These

discontinuities have been interpreted to reflect changes in alkyl phase, and have been

described by the Clapeyron relationship.

In the presence of solvent, the trend of increasing conformational order with

increasing surface coverage is generally maintained. However, the phase changes

observed in these materials in air are muted by interactions of the alkyl chains with

solvent. In some cases, different, more subtle phase changes are observed that depend on

the extent of solvation of the stationary phase and alkylsilane surface coverage. In

1.6H

1.5

^ 1 4 -(N

U 1.3-

^ 1.2-

i as

^ 1.0-1

0.9

0.8

a A • m

m A

s ? A f T

• A * A

• • • t • I ¥ A

^ $ § W

•

-20 -10 0 10 20 30

Temperature (°C)

40

242

50

X 1 . 1 - S

Temperature (°C)

Figure 6.11 l[v^(CH2)]/l[v,(CH2)] as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (#), DFC18SL (O), and MFC 18 (•) in (a) methanol and (b) acetonitrile. The error bars represent one standard deviation.

243 1.6

1.5-

^ 1.4-

U 1.3-

# 1.2-1

5 11

^ i.oH

0.9-

0 . 8 -

i i

-20 -10

i . • i ^

0 0 5

T 10 20 30 40

Temperature ("C)

a

T 50

1 . 6 -

1.5-

1.4-

U 1.3

# 1.2-1 r—

5 03

1.0 H

0.9-

0.8

i

O

-20 -10

i

o T

Y f Q

f I 0 10 20 30

Temperature ("C )

40 50

Figure 6.12 l[v.,(CH2)]/l[Vj.(CH2)] as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•) in (a) chloroform and (b) toluene. The error bars represent one standard deviation

244

methanol, all of the tri- and difunctional materials (TFC18SF, TFC18SL, DFC18SF, and

DFC18SL) have plateau regions of relatively temperature-independent order below 0 "C,

Above 0 °C, the order of these phases decreases monotonically with increasing

temperature. The monofunctional stationary phase exhibits a distinct change in order at

approximately 10 °C in methanol, with relatively constant values below and above this

temperature.

in acetonitrile, plateau regions exist for the surface-polymerized materials

(TFC18SF and DFC18SF) below 0 "C with a monotonic decrease in order above this

temperature. TFC18SL exhibits a monotonic change in order with temperature without

any indication of a phase change. However, DFC18SL shows a monotonic decrease in

order with temperature but with a phase change (i.e., change in slope) at approximately

10 "C. Finally, the monofunctional stationary phase, MFC 18, shows two regions of

approximately constant order with a change at approximately 0 °C.

In chloroform and toluene, all stationary phases are more disordered than in

methanol or acetonitrile, as discussed for the nonpolar solvents in general above. In

chloroform, the conformational order behavior exploits differences in alkylsilane

precursor and surface coverage, rather than polymerization method. The two highest

surface coverage materials, both of which are trifunctional, exhibit changes in slope at 30

"C, with plateau regions of strictly chloroform-induced disorder at high temperatures.

The other three materials display a conformational order that is relatively independent of

temperature. In toluene, although TFC18SF shows a low temperature plateau region, all

245

stationary phases generally exhibit a monotonic decrease in this order indicator with

temperature.

Although one must interpret these changes carefully in light of the non-linear

dependence of I[Va(CH2)]/I[Vs(CH2)] on overall conformational order, these results

further support the subtle interplay between temperature- and solvent-controlled order in

these systems. Furthermore, these results show the broad range of solvent interactions

with alkylsilane stationary phase materials that depend not only on solvent

characteristics, but on stationary phase properties as well.

Peak Frequencies of v(C-H) Modes as Indicators of Solvent Effects on Conformational

Order

In addition to I[Va(CH2)]/I[Vs(CH2)] as an indicator of conformational order, the

frequency of either the Va(CH2) or Vs(CH2) can be used to ascertain interchain coupling

information for the solvated alkyl chains of these stationary phases. The precision of

these measurements is generally not as high as that for the I[Va(CH2)]/I[Vs(CH2)]

indicator. Nonetheless, these indicators serve as alternate independent checks of the

order of these alkylsilane systems, since they respond to chain coupling and not chain

rotational disorder or traction of C-C bonds in gauche conformations as does the

I[Va(CH2)]/I[Vs(CH2)] indicator.

The frequency of the VsCCHi) mode is plotted as a function of temperature in

Figures 6.13 and 6.14 for the same four representative solvents, methanol (Figure 6.13a),

's

2852

2851

2850

S 2849

^ 2848

^ 2847

p. 2846

2845

t t J T

246

-20 -10 0 10 20 30 40 50 60

Temperature (°C)

2851 -

^ 2848-

I t ffi P, 2846

-20 -10 0 10 20 30 40

Temperature (°C)

50 60

Figure 6 .13 v/CHj) peak frequency as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC18 (T) in (a) methanol and (b) acetonitrile. The error bars represent one standard deviation.

2849-

2848-

2847-

2846-

247

2845-1—^—r -20 -10

2852-

T i 1 ' 1 1 r

10 20 30 40 Temperature (°C)

50 60

II I 2849

^ 2848 <u ^ 2847

B 2846

0 10 20 30 40 50 60 Temperature ("C)

Figure 6.14 v/CH^) peak frequency as a function of temperature for TFC18SF (•), TFC18SL (A), DFC18SF (•), DFC18SL (O), and MFC 18 (•) in (a) chloroform and (b) toluene. The error bars represent one standard deviation.

248

acetonitrile (Figure 6.13b), chloroform (Figure 6.14a), and toluene (Figure 6.14b), for all

five stationary phases. In all solvents, the Vs(CH2) frequency is temperature- and surface

coverage-dependent, generally increasing (i.e., less chain coupling) with increasing

temperature and decreasing coverage, as observed for these materials in air. However,

perhaps not surprisingly, the behavior of this indicator does not exactly track that of the

I[Va(CH2)]/I[Vs(CH2)] indicator, since it reflects changes in interchain coupling.

Nonetheless, these two indicators of conformational order do change with the same

general trends with solvent and temperature. Moreover, the subtle interplay between

temperature- and solvent-induced disorder observed in the I[Va(CH2)]/I[Vs(CH2)]

indicator behavior is also observed in the Vs(CH2) peak frequency results. For all

stationary phases except the highest surface coverage material, significant chain

decoupling occurs for temperatures above 0 °C. Indeed, the two lowest surface coverage

materials, MFC 18 and DFC18SL, are completely decoupled (Vs(CH2) ~ 2851 cm"') or

significantly decoupled (vs(CH2) ~ 2850 cm"'), respectively, in all solvents above 0 °C.

Moreover, the Vs(CH2) frequency for MFC 18 in chloroform and toluene exhibits behavior

that is temperature-independent and strictly solvent-dependent. For DFC18SL, the

behavior is only slightly temperature-dependent. For TFC18SF, the Vs(CH2) frequency

does exhibit a dependence on temperature, with a similar low temperature plateau region

(i.e., temperature-independent) and a monotonic change at higher temperatures as was

observed for the I[Va(CH2)]/l[v,s(CH2)] indicator.

249

Given that these two indicators reflect ditTerent types of alkyl chain disorder, it is

interesting to consider how they respond to changes in temperature relative to each other.

In general, the temperature dependence of these two indicators for the high-density

stationary phase systems is similar to the behavior observed for the melting of bulk

alkanes, where chain decoupling (as indicated by the frequency of the Vs(CH2) mode)

requires more thermal energy to change than to induce subtle changes in C-C bond

conformation (as indicated by the I[va(CH2)]/I[vs(CH2)] indicator).


A detailed examination of solvent-stationary phase interactions probed by Raman

spectroscopy for a series of high-density Cig materials is presented. Results indicate that

rotational and conformational order information is dependent on solvent parameters,

stationary phase grafting procedure, and surface coverage, and can be used to describe

specific solvent-stationary phase interactions. In general, polar solvents increase the

rotational order of homogeneously distributed high surface-coverage materials, fitting an

absorption model of interaction at the distal methyl group. Nonpolar solvents generally

intercalate into the alkyl chains and fit a partitioning model of interaction;

heterogeneously distributed and lower coverage materials are more disordered by this

partitioning than higher coverage materials. In addition to the solvent-stationary phase

interactions, an interplay between solvent- and temperature-dependent ordering of these

materials is observed, more solvent-dependent order is observed for lower coverage

250

materials in nonpolar solvents and more temperature-dependent order is observed for

higher coverage materials in polar solvents.

Future directions for this work include investigating the conformational order of

these phases in more complex binary or ternary solvent systems, more customary of real

chromatographic mobile phases. In addition, examining alkylsilane conformational order

of a series of phases prepared from the same precursor and polymerization method with

variable surface coverage would be of interest such that the effect of solvation and

surface coverage can be determined without being convoluted with precursor or

preparation method effects.

Appendix 6.1

This appendix contains Raman spectra in the v(C-H) for the five high-density

stationary phases in methanol, acetonitrile, water, acetone, hexane, chloroform, benzene,

toluene, and THF at -15, 10, 0, 10, 30, 40, and 50 "C.

TFC18SF

FCISSL

•FC18SF

DFC18SL

MFC 18

2900 2800 3000

TFC18SF

DFC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

'C18SL

<u FC18SF

DFC18SL

MFC 18

2900 3000 2800 Wavenumbers (cnrO Wavenumbers (cm"0 Wavenumbers (cm-')

Figure A6.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFCIBSL, and MFC] 8 in methanol at (a) -15 "C, (b) -10 "C, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, m DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min. ^

TFC18SF TFC18SF

'C18SL TFC18SL

m C <L) »FC18SF LFC18SF

DFC18SL PFC18SL

MFC18 MFC18

2900 3000 2800 3000 2800 2900 Wavenumbers (cm"^) Wavenumbers (cm"^)

Figure A6.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in methanol at (a) 10 °C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min. ^

TFC18SF TFC18SF T"'

•FC18SL TFC18SL

tFC18SF •FC18SF

DFC18SL DFC18SL

MFC18 ,MFC18

2900 3000 2800 2900 3000 2800 Wavenumbers (cm-') Wavenumbers (cm-')

Figure A6.3 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in methanol at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min. ^

TFC18SF TFC18SF

TFC18SL TFC18SL

DFC18SF DFC18SF

DFC18SL DFC18SL

MFC18 MFC 18

—f— 3000 2900 3000 2800 2900 2800

TFC18SF

TFC18SL:

DFC18SF

DFC18SL

MFC) 8

3000 2900 2800

Wavenumbers (cm"^) Wavenumbers (cnr') Wavenumbers (cm"')

Figure A6.4 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetonitrile at (a) -15 °C, (b) -10 "C, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, m DFC18SF, and DFCl SSL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

m C o

K FCISSF

J ^\TFC18SL

\ \DFC18SF

^\DFC18SL

J' VMFCIS

2800 2900 3000

Wavenumbers (cm'\)

TFC18SF

TFC18SL

DFC18SF

1^

DFC18SL

MFC18

2800 2900 3000

Wavenumbers (cnr\)

Figure A6.5 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetonitrile at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFCI8SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

2900 3000 2800

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

3000 2800 2900

Wavenumbers (cm"^) Wavenumbers (cm"^)

Figure A6.6 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in acetonitrile at (a) 40 "C and (b) 50 °C, Integration times for TFC18SF, TFC I8SU DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF TFC18SF

TFC18SL JFC18SL

OJ DFC18SF DFC18SF

DFC18SL DFC18SL

MFC18 MFC18

3000 2900 2800 3000 2900 2800

Wavenumbers (cm-') Wavenumbers (cm"') Wavenumbers (cm-')

Figure A6.7 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in water at (a) -15 "C, (b) -10 "C, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, m DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

TFC18SF TFC18SF

'FC18SL TFC18SL

DFC18SF DFC18SF

DFC18SL DFC18SL

MFC18 MFC18

3000 2900 2800 Wavenumbers (cm'^) Wavenumbers (cm"^)

Figure A6.8 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in water at (a) 10 "C and (b) 30 °C, Integration times for TFC18SF, TFC18SU DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

TFC18SF TFC18SF

TFC18SL TFC18SL

<D DFC18SF DFC18SF

DFC18SL DFC18SL

MFC18 MFC18

3000 2900 2800 3000 2800 2900 Wavenumbers (cm"^) Wavenumbers (cm*0

Figure A6.9 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in water at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC 18

2900 3000 2800

TFC18SF TFC18SF

TFC18SL TFC18SL

DFC18SF DFC18SF

DFC18SL DFC18SL

MFC18 MFC18

2800 2900 3000

Wavenumbers (cm"^) Wavenunibers (cm"') Wavenunibers (cm"')

Figure A6.10 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetone at (a) -15 °C, (b) -10 "C, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, m DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

JFC18SL

a> DFC18SF

DFC18SL

MFC18

2900 3000 2800

Wavenumbers (cm-\) Wavenumbers (cnr^

Figure A6.11 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetone at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

c

^\TFC18SF

N \TFC18SL

r ^FC18SF

^DFC18SL

J- \MFC18

2800 2900 3000 Wavenumbers (cm"^)

N \TFC18SL

^ DFC18SF

A

\DFC18SL

/A

X.MFC18

aZ. 2800 2900 3000 Wavenumbers (cm"\)

Figure A6.12 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in acetone at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

ON to

TFC18SF

TFC18SL

DFC18SF

DFC18SL

.MFC 18

3000 2800 2900

TFC18SF TFC18SF

TFC18SL 'FC18SL

.

DFC18SF DFC18SF

DFC18SL DFC18SL

MFC 18 MFC18

3000 2900 2800 3000 2800 2900

Wavenumbers (cnr\) Wavenumbers (cm-') Wavenumbers (cm"')

Figure A6.13 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in hexane at (a) -15 "C, (b)-10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, w DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min. «

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

3000 2800 2900

Wavenumbers (cm-') Wavenumbers (cnr')

Figure A6.14 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in hexane at (a) 10 "C and (b) 30 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFCl 8 in all solvents are 5-niin.

TFC18SF

TFC18SL

DFC18SF

DFG18SL

MFC18

3000 2900 2800

TFC18SF

TFC18SL

cn s DFC18SF

DFC18SL

MFC 18

2800 2900 3000

Wavenumbers (cm-\) Wavenumbers (cm-^)

Figure A6.15 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in hexane at (a) 40 °C and (b) 50 "C. integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC18 in all solvents are 5-min. ^

TFC18SF

TFC18SL

DFC18SF

IFC18SL

MFC 18

3000 2900 2800

TFC18SF

TFC18SL

•FC18SF

PFC18SL

MFC18

2800 2900 3000

TFC18SF

'FC18SL

(U DFC18SF

DFC18SL

MFC 18

2800 2900 3000

Wavenumbers (cm"') Wavenumbers (cm*^) Wavenumbers (cm'^)

Figure A6.16 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in chloroform at (a) -15 "C, (b) -10 "C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min,

TFC18SF

JFC18SL

DFC18SF

IFC18SL

MFC18

3000 2900 2800

TFC18SF

JFC18SL

<u •FC18SF

DFC18SL

MFC18

3000 2900 2800

Wavenimibers (cm-') Wavenumbers (cnr')

Figure A6,17 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFCISSL, and MFC! 8 in chloroform at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFCISSL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

to -J

TFC18SF

'FC18SL

iFClSSF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

3000 2900 2800 Wavenumbers (cm-\) Waveniimbers (cm-\)

Figure A6.18 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in chloroform at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC18 in all solvents are 5-min.

TFC18SF

TFCISSL

DFC18SF

DFC18SL

€18

2800 2900 3000

TFC18SF TFC18SF

'FC18SL TFC18SL

§ iDFC18SF DFC18SF

DFC18SL DFC18SL

MFC 18 MFC 18

3000 2800 2900 3000 2800 2900

Wavenumbers (cnr') Wavenumbers (cm'O Wavenumbers (cm"')

Figure A6.19 Raman spectra in the v(C-H) region of TFC18SF, TFCISSL, DFC18SF, DFC18SL, and MFC 18 in benzene at (a)-15 "C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFCISSL, w DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min. ^

TFC18SF TFC18SF

TFC18SL TFC18SL

DFC18SF DFC18SF

DFC18SL DFC18SL

MFC18 MFC18

3000 2900 2800 3000 2800 2900

Wavenumbers (cm-') Wavenumbers (cm-')

Figure A6.20 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in benzene at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min. •=>

TFC18SF TFC18SF

TFC18SL TFC18SL

m g DFC18SF DFC18SF

DFC18SL DFC18SL

MFC18 MFC18

3000 2900 2800 3000 2800 2900

Wavenumbers (cm-\) Wavenumbers (cm"^)

Figure A6.21 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in benzene at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF. TFC18SL, DFC18SF, ^ and DFC18SL in all solvents are 2-min, for MFC18 in all solvents are 5-min.

TFC18SF

FC18SL

DFC18SF

DFC18SL

MFC 18

2900 3000 2800

TFC18SF TFC18SF

TFC18SL FC18SL

DFC18SF DFC18SF

DFC18SL IFC18SL

MFC18 MFC18

2900 3000 2800 2900 3000 2800 Wavenimibers (cm"0 Wavenumbers (cnr ') Wavenumbers (cm"^)

Figure A6.22 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in toluene at (a) -15 "C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, m DFC18SF, and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min, ^

TFC18SF

JFCISSL

PFC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

JFCISSL

S J3FC18SF

DFC18SL

MFC 18

3000 2800 2900 Wavenumbers (cm*' ) Wavenumbers (cm"')

Figure A6.23 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in toluene at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min. ^

TFC18SF

TFC18SL

'FC18SF

PFC18SL

MFC18

3000 2900 2800

TFC18SF

JFC18SL

>FC18SF

DFC18SL

MFC18

2900 3000 2800

Wavenumbers (cm-\) Wavenumbers (cm-')

Figure A6.24 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in toluene at (a) 40 °C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, w and DFC18SL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

•t I

K v-^TC18SF

A

^\TFC18SL

A

ADFC18SF

\DFC18SL

J" \MFC18

b / TFC18SF

\K TFC18SL

N^C18SF

VN \DFC18SL

V.MFC18

2800 2900 3000 Wavenumbers (cm"^)

2800 2900 3000 Wavenumbers (cm'\)

/ ^-^TFC18SF

/

Av ^NJFC18SL

' V DFC18SF / Av

^NJFC18SL

' V DFC18SF

/ \DFC18SL

A /Av \^FC18

2800 2900 3000 Wavenumbers (cm"')

Figure A6.25 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in THF at (a)-15 "C, (b) -10 "C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 in are 5-min.

o

TFC18SF

TFC18SL

•FC18SF

•FC18SL

MFC 18

3000 2800 2900

TFC18SF

TFC18SL

s cu •FC18SF

DFC18SL

MFC 18

3000 2800 2900

Wavenumbers (cm"^) Wavenumbers (cnr')

Figure A6.26 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in THF at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL are 2-min, for MFC 18 are 5-min.

TFC18SF TFC18SF

TFC18SL TFC18SL

•FC18SF IFC18SF

DFC18SL ,DFC18SL

MFC 18 MFC 18

3000 2900 2800 3000 2900 2800 Wavenumbers (cm"^) Wavenumbers (cm"^)

Figure A6,27 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in THF at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, m and DFC18SL are 2-min, for MFC 18 are 5-min.

278

Chapter 7


Raman Spectroscopy: Effects of Self-Associating Solvents

Introduction

Of the thousands of reversed-phase liquid chromatographic (RPLC) separations

performed daily on mixtures of small organic compounds, a large number are performed

with mobile phases composed primarily of a polar organic solvent or a mixture of a polar

solvent with water. The most common mobile phases for RPLC separations of this type

include mixtures of either methanol or acetonitrile and water. The utility and popularity

of these mobile phases stem from their solvation strengths for elution strengths) for a

large number of small organic analytes and their ability to solvate or interact with the

stationary phase; indeed, both characteristics are crucial for chromatographic success.

In Chapter 6, the effect of pure mobile phase solvent on the conformational order

of several high-density alkylsilane stationary phase materials was examined in the

context of several chromatographic parameters including temperature, surface grafting

procedure, surface coverage, and nature of the alkylsilane precursor. In the polar solvents,

methanol, acetonitrile, and water, the two most homogeneous, high-density stationary

phase materials, TFC18SF (6.45 [.imolW) and DFC18SF (5.00 fimolW), were observed

279

to be more ordered than in air (Chapter 5). These three solvents all have the ability to

strongly self-associate^' and have appreciable hydrogen bond donating (HBD) and

accepting (HBA) abilities/ These results are in good agreement with those obtained

from FT-IR experiments on C\x phases in solvent by Sander et al7 ' In this work, the

investigators observed a decrease in the relative numbers of gauche conformers for

alkylsilane phases in 70 to 100% methanol/vvater solutions relative to in air. They

attribute this solvent-induced ordering to specific adsorption of the methanol molecules

to the alkyl chains, orientating with the methyl group intercalating between the distal

ends of the alkylsilanes and the polar hydroxyl group extending into the bulk mobile

phase.

Goals of this Work

As a follow-up to work presented in Chapter 6, the main goal of this work was to

examine whether additional solvents with self-associating characteristics induce ordering

of these high-density stationary phases. In addition, determining the effect of surface

coverage on self-associating solvent interaction was of interest, as surface coverage was

observed to be a crucial factor in solvent interaction with common mobile phase solvents

(Chapter 6). From these results, molecular pictures of the chromatographic interface

were constructed to show the proposed mechanism of solvent interaction with these

alkylsilane phases.

280

Methodology


characterized as previously described by Sander and Wise"^'^'^'' and in Chapter 2 of this

Dissertation. Perdeuterated water (99.9%) and 1-butanol (> 99%) were purchased from

Aldrich. Additional perdeuterated solvents used in these experiments including methanol

(99.8%), acetonitrile (99.8%), ethanol (99%), and acetic acid (99.5%) were obtained from

Cambridge Isotope Laboratories, Inc. All solvents were used as received.


material into a 5-mm-dia NMR tube; 200 |iL of solvent was then added. Samples were

sonicated for 10 min and equilibrated at 20 "C for a minimum of 12 h prior to spectral


Coherent Verdi Nd: YVO4 laser on a Spex Triplemate spectograph as described in

Chapter 2. Slit settings of the Triplemate were (0.5/7.0/0.150) mm for all experiments,

corresponding to a spectral bandpass of 5 cm"'. All samples were equilibrated at the

appropriate temperature for at least 30 min prior to spectral acquisition. Integration times

for each spectrum are provided in the figure captions. Low laser power and short

acquisition times were used to avoid unnecessary sample heating.

Raw spectral data in the v(C-H) region were processed and analyzed as described in

Chapter 2. Pictures of the solvent/stationary phase interface were constmcted and energy

minimized using ChemDraw 3D (CambridgeSoft) as described in Chapter 2,

281

Raman Spectroscopy of High-Density Alkylsilane Stationary Phases in Solvent

The utility of Raman spectroscopy for this purpose stems from the abundance of

spectral indicators of conformational order, especially in the v(C-H) region between 2750

and 3050 cm"\ Moreover, spectral interrogation of conformational order in this region

can be achieved in the presence of solvent without significant spectral interference fiôm

the silica substrate or surface-adsorbed water. It should be noted that although significant

conformational order information exists for alkane systems in the v(C-C) and 8(C-H)

regions between 900 and 1500 cm'^, the stationary phase materials studied here were

prepared on silica particles that exhibit innate fluorescence that interferes with the

acquisition of quality Raman spectra in these regions. Therefore, for this study,

conformational order assessment is limited to spectral data obtained in the v(C-H) region.

Raman spectra for all five high-density stationary phase materials at room

temperature (20 X) in polar, self-associating solvents, perdeuterated ethanol, acetic acid

and 1 -butanol, are shown in Figure 7.1. Additional spectra of these materials in these

self-associating solvents at temperatures from-15 to 50 °C are given in Appendix 7.1.

Structure and order information about alkane-based systems can be obtained from

multiple indicators in their Raman spectra; these have been described in detail

elsewhere." Several of these indicators come from the v(C-H) region, even though

this region is quite complex, • ' • "' containing up to 14 possible vibrational modes.

One important empirical indicator of conformational order is the peak intensity ratio of

the VaCCHj) (2885 cm"') to the Vs(CH2) (2850 cm"'), I[Va(CH2)]/I[Vs(CH2)]."'^ ""-'-^ ''

Va(CH,)

Vs(CH,)i

FC18SF

^S(?^3)FR

TFC18SL

DFC18SF

DFCISSL

v(CH3)3,

MFC18

2800 2900 3000 Wavenumbers (cm"')

TFC18SF

TFC18SL

DFC18SF

DFCISSL

MFC18

2800 2900 3000 Wavenumbers (cm"')

c j ^grcissF

\]JC18SL

\pFC18SF

\ DFCISSL

Vwcis

2800 2900 3000 Wavenumbers (cm"\)

Figure 7.1 Raman spectra in the v(C-Er) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in (a) ethanol, (b) acetic acid, and (c) 1-butanol. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFCISSL in all solvents are 2-min, for MFC 18 in all solvents are 5-min.

to 00 to

283

This indicator is extremely sensitive to small changes in conformational order that stem

from subtle rotational disorder in short segments of the alkyl chains and from the addition

of true C-C gauche conformers. This ratio spans a range from approximately 0.6 to 0.9

for bulk alkanes in the liquid state and 1.6 to 2.0 for alkanes in the crystalline state/'^"^'"

although changes as small as 0.02 in this value can be reproducibly measured.

Spectral interference from the v(CH3)si mode at 2903 cm"^ occurs for stationary

phases prepared with methyloctadecyldichlorosilane and dimethyloctadecylchlorosilane

precursors. Since the VaCCHi) is superimposed on the v(CH3)si band for mono- and

difunctional materials, a slight increase in its intensity can result. One must be cognizant

of this subtle difference when comparing I[va(CH2)]/I[Vs(CH2)] values between stationary-

phases prepared with different alkylsilane precursors.

In addition to the intensities of the Va(CH2) and Vs(CH2) modes, the peak

frequencies of these bands are also a function of alkyl structure and order, decreasing by

6 to 8 cm"' from the liquid state to the crystalline state. This frequency decreases as a

result of damping of these stretching modes by neighboring alkyl chains as they order in

a close-packed arrangement. Thus, this indicator is a direct measure of alkyl chain

coupling. Detailed correlations between these and other indicators of conformational

order are described in Chapter 2. These correlations are used to interpret the structure of

these alkyl-modified stationary phases in detail.

284

Review of Solvent Effects on High-Density Stationary Phase Conformational Order

In Chapter 6, we examined the effect of common mobile-phase solvents (benzene,

toluene, tetrahydrofuran, acetone, chloroform, hexane, acetonitrile, methanol, and water)

on the conformational order of these high-density stationary phases. These solvents

exhibit a wide range of polarity, dipole moment, polarizability, shape, size,

hydrophobicity, and hydrogen bond donating/accepting ability that have been quantified

by the solvatochromic parameters, TI*, a, and P as described by Kamlet et al.^ ' The

values of these parameters for these solvents and the additional solvents used in this study

(perdeuterated 1-butanol, water, methanol, acetic acid, ethanol, acetonitrile) are listed in

Table 7.1. The n* parameter is a measure of the solvent dipole normalized to its

polarizability, and a and P are measures of the solvent hydrogen bond donating (HBD)

and accepting (HBA) abilities, respectively. It is important to note throughout this

discussion the close relationship between n* and other solvatochromic parameters

including the Hildenbrand solubility parameter (6H), and the solvent self-association

parameter (5SA)-'

In the polar solvents, methanol, acetonitrile, and water, the two stationary phases

that have the most homogeneously distributed alkylsilanes with the highest surface

coverages (TFC18SF and DFC18SF) were found to exhibit a statistically-significant

degree of order greater than that in air, albeit small (Figure 6.4). This solvent-induced

ordering appears to be a fiinction of solvent self-association ability, high surface

coverage, and alkyl chain homogeneity of these two materials (as determined by NMR

285

Table 7.1. Solvatochromic parameters'*, %*, a, and P for solvents

Solvent TT* a p

Water 1.09 0.18 1.17


Acetone 0.71 0.48 0.06

Acetic acid 0.64 1.12

Methanol 0,60 0.62 0.93

Benzene 0.59 0.10 0

THF 0.58 0.55 0

Chloroform 0.58*^ 0 0.44

Ethanol 0.54 0.77 0.83

Toluene 0.54 0.11 0

1-Butanoi 0.47 0.88 0.79

Hexane -0.08 0 0

"As described by Kamlet et al7'^ ''Actual value of TI* for chloroform, but plotted as 0.57 in the figures to discriminate it from THF, '''Not Reported.

286

7 7 7 19 spectroscopy " ), since the other three stationary' phases do not exhibit this effect. The

solvents that cause such ordering all have significant, non-zero values of both a and P

indicating their ability to both donate and accept hydrogen bonds, respectively. As a

result of these characteristics, these solvents are strongly self-associating.

The interaction of these self-associating solvents with TFC18SF and DFC18SF

can be described primarily by an adsorptive model in which the stationary phase presents

a sorption site for solvent at the distal methyl group of alkyl chains extended away from

the surface In this interaction, solvent molecules self-associate through hydrogen

bonding and dipole interactions resulting in lateral ordering of solvent at the interface.

This lateral ordering of solvent at the interface is translated to increased order of the alkyl

moieties. To determine whether solvents with similar characteristics (i.e., significant,

non-zero values of a and f5) induce similar conformational ordering of these high-density,

homogeneous materials, additional investigations in ethanol, acetic acid, and 1 -butanol

were undertaken.

While these organic solvents, and the other primary alcohols used in these

experiments, can contain considerable amounts of water, the interactions between the

hydrophobic stationary phases and trace water at the interface is negligible compared to

contributions from organic solvent. Thus, descriptions of the molecular interface between

alkylsilanes and organic solvents do not consider the effects of trace water.

287

Effect of Self-Associating Solvent

The effect of self-associating solvents on the conformational order of the

TFC18SF and DFC18SF stationary phases is shown in Figure 7.2 as plots of

I[V;,(CH2)]/I[ VS(CH2)] (Figure 7.2a) and VsCCHa) peak frequency (Figure 7.2b) as a

function of the solvatochi omic parameter, %* for water, acetonitrile, methanol, acetic

acid, ethanol, and 1-butanol. In all solvents except 1-butanol, the value of

l[va(CH2)]/l[Vs(CH2)] is greater than in air for both TFC18SF and DFC18SF. In contrast,

in 1 -butanol, the value of I[Va(CH2)]/I[Vs(CH2)] is less than in air. In addition,

I[Va(CH2)]/I[Vs(CH2)] generally decreases with increasing n* from ethanol to water. In

all solvents except water, the Vs(CH2) peak frequency for TFC18SF and DFC18SF is less

than in air. This frequency is slightly greater than in air for both materials in water. No

systematic trend of changing Vs(CH2) peak frequency with ji* for these solvent/stationary

phase systems is apparent.

The interactions of the polar solvents methanol, acetonitrile, and water with

TFC18SF and DFC18SF were previously described in terms of an adsorptive model of

interaction at the distal end of alkyl chains extending away from the surface. The

solvent-induced ordering of these materials is a result of solvent self-association ability,

high surface coverage, and most importantly, surface homogeneity of the alkyl chains."^

Based on an identical effect of ethanol, acetic acid, and 1 -butanol on the

I[Va(CH2)]/I[vs(CH2)] value, the interactions of ethanol and acetic acid with TFC18SF

and DFC18SF can be described using a similar picture.

1.7

1 . 6 -

-.1.5-X B 1-4-1

_ J 3

m 1.2 u ^"1.1 H

1.0

0.9

0.8

Air

-t-

Methanol Ethanol|| Acetic Acid

l-Butanol ! ^ Âcetonitrile

a

Water

0.0 !

0.2

.(D_.

288

0.4 0.6

TI*

0.8 1.0 1,2

2854-

>. 2852 a Water Acetic Acid

Methanol o 2850 LIRI

Ethanol

1 -Butanol O ^ 2848

Acetonitrile

0

2844

0.8 1.0 1.2

n*

Figure 7.2 (a) IfVaiCHj)]/![v,(CH2)] and (b) v/CHj) peak frequency as a function of solvent solvatochromatic parameter ti* for TFC18SF (•) and DFC18SF (O) in air, methanol, acetonitrile, water, acetic acid, ethanol, and 1-butanol The error bars represent one standard deviation.

289

At the sorption site of the stationary phase, these solvent molecules can self-

associate through hydrogen-bonding to neighboring interfacial solvent molecules and to

bulk solvent. It is proposed that this lateral ordering of self-associated solvent induces

slight conformational ordering of the alkyl moieties of the stationary phase. Thus, in all

solvents investigated here except 1-butanol, the conformational order of TFC18SF and

DFC18SF is greater than in air. Moreover, as shown in Figure 7.2a, the value of

I[va(CH2)]/I[Vs(CH2)]generally decreases with increasing n*. This latter trend can best be

understood by one of two perspectives of the meaning of re*. First, as %* increases, so

generally does the solvent self-association ability.^ ' Thus, as self-association ability

increases, the tendency for interaction with or solvation of the alkyl chains decreases

leading to a decrease in ordering relative to less self-associated solvents. An alternate

view is that as decreases, the hydrophobic nature of the solvent increases and hence,

so does its tendency to solvate the alkyl chains. Either explanation adequately describes

the general trends observed for these stationary phases.

For TFC18SF and DFC18SF in water, solvent self-association is stronger than

solvation of the alkyl stationary phase. Therefore, relatively little interaction of water

molecules with the stationary phase occurs with only a slight increase in conformational

order relative to in air. In acetonitrile, acetic acid, methanol and ethanol. the

conformational order of TFC18SF and DFC18SF is greater than in water and increases

with decreasing solvent self-association ability. Although at first glance, this trend might

seem counter-intuitive, in fact, decreasing solvent self-association leads to slightly greater

290

interaction of these solvents with the distal ends of the stationary phase alkyl chains

leading to their greater order. Increased chain coupling as indicated by the decreased

values of the Vs(CH2) peak frequency (Figure 7.2b) further supports this picture.

As interaction of the solvent with the stationary phase alkyl chains increases

further, a reversal in the ordering effect of these solvents is observed. Slight disordering

relative to air is observed due to increased alkyl chain solvation by 1 -butanol, for

example. Due to the more alkane-like characteristics of 1 -butanol relative to the other

primary alcohols (methanol and ethanol) studied, it solvates these stationary phases more

like hexane. As a result, the interaction of 1 -butanol with these high-density,

homogeneous materials is more consistent with partitioning in which the solvent

molecules intercalate between alkyl chains of the stationary phase. This solvation

imparts conformational disorder as indicated by a decrease in I[Va(CH2)]/I[Vs(CH2)]

(Figure 7.2a), while maintaining alkyl chain coupling interactions between the stationary

phase and the 1-butanol as indicated by low frequencies of the VsCCHs) mode (Figure

7.2b).

The conformational order and chain coupling information obtained for TFC18SF

and DFC18SF in self-associating solvents leads to the development of detailed molecular

pictures of the solvent/stationary phase interface with the aid of the energy-minimized

molecular mechanics (MM2) computations described above. Proposed pictures of these

interfacial interactions are shown in Figures 7.3-7.8 for TFC18SF in methanol (Figure

7.3), acetonitrile (Figure 7.4), water (Figure 7.5), ethanol (Figure 7.6), acetic acid (Figure

291

Figure 7.3 Molecular picturc of the interactions between TFC18SF and methanol at the chromatographic interface.

Figure 7.4 Molecular picture of the interactions between TFC18SF and acetonitrile at thp chromatographic interface.

293

I

1 r:

Figure 7.5 Molecular picture of the interactions between TFC18SF and water at the chromatographic interface.

294

i-: •I i-4^

i %

Figure 7.6 Molecular piclure of the interactions between TFC18SF and ethane 1 at the chromatographic interface.

295

C*

i

Figure 7.7 Molecular picture of the interactions between TFC18SF and acetic acid at the chromatographic interface.

296

i

V

Figure 7.8 Molecular picture of the interactions between TFC18SF and 1 -butanol at the chromatographic interface.

297

7.7), and 1 -butanol (Figure 7.8), respectively, to illustrate the relevant intermolecular

interactions of these solvents with these high-density, homogeneous coverage stationary

phase materials. The initial models for MM2 computations that resulted in these pictures

were developed to best reflect the conformational changes reported by the Raman

spectral data as described above. Given that these models were not developed on the

basis of a priori calculations or dynamic simulations, specific attributes, such as the exact

alkyl chain tilt with respect to the surface, may be somewhat in error. Nonetheless, the

solvent-alkyl chain interactions represented in Figures 7.3-7.8 are believed to accurately

represent the alkyl chain conformational order behavior revealed in the Raman spectral

data.

For these systems, lateral hydrogen-bonding (methanol, acetic acid, and ethanol)

and dipole-dipole interactions (acetonitrile) with neighboring interfacial solvent

molecules are proposed to cause the observed increase in conformational order within the

stationary phase. In 1 -butanol (Figure 7.8), the alkyl chains accommodate partitioning of

the 1 -butanol solvent molecules by the formation of cavity regions between chains,

contributing to conformational disorder while maintaining significant chain coupling.

In contrast to the behavior observed for TFC18SF and DFC18SF in self-

associating solvents, the behavior of the other three stationary phase materials, TFC18SL,

DFC18SL, and MFC 18, is quite different, because of the difference in surface

homogeneity between the surface polymerized and solution polymerized or lower surface

coverage materials. As shown in Figure 7.9a, the values of I[Va(CH2)]/I[Vs(CH2)] for

298

1 .7-

1 .6-

i—^ 1,5-

u 1.4-

1 .3--

1.2-X ' . u

CQ 1.1 -

1 .0-

0 .9-

0 .8-

a

s

(U

IX

U

2854-

2852-f

2850-

2848-

2846 H

2844-

Air

-t-

Methanol

Ethanol Acetonitrile

•f' Water 1 -Butanol

Acetic Acid Water T

Acetonitrile Ethanol

1 -Butanol

Figure 7.9 (a) I[v3(CH2)]A[v,(CH2)] and (b) v/CHj) peak frequency as a function of solvent solvatochromatic parameter tc* for TFC18SL (A), DFC18SL (•), and MFC 18 (V) in air, methanol, acetonitrile, water, acetic acid, ethanol, and 1-butanol. The error bars represent one standard deviation.

299

these materials in water, acetonitrile, acetic acid, methanol, ethanol, and 1-butanol

suggest that the stationary phase alkyl chains are more disordered in all cases than in air.

Moreover, the extent of conformational order in these solvents depends on alkyl chain

surface coverage for these three materials, and generally increases with increased surface

coverage as indicated by higher values of I[Va(CH2)]/I[vs(CH2)] (Figure 7.9a) and by

decreased values of Vs(CH2) peak frequency (Figure 7.9b).

This difference between solvent interaction with homogeneous and heterogeneous

alkyl surfaces is shown explicitly when the 1[Va(CH2)]/l[VsCCHz)] values of TFC18SL and

DFC18SF are compared. Even though TFC18SL (5.26 |.imol/m^) has a slightly higher

surface coverage than DFC18SF (5.26 [.imol/m^), alkyl conformational order in solvent is

greater than in air for the homogeneously covered DFC18SF, but conformational order

decreases in solvent relative to in air for the more heterogeneous TFC18SL surface.

These variations in I[va(CH2)]/I[vs(CH2)] and v»(CH2) peak frequency with

surface coverage are consistent with previous observations made for these stationary

phases in air, and in polar and nonpolar solvents from Chapter 6. In addition to

adsorption of these self-associating solvents at the distal end of the alkyl chains, for lower

surface coverage materials, intercalation (i.e., partitioning) of these solvents into the alkyl

chains is also a possible mechanism of interaction.

300

Effect of Temperature in Self-Associating Solvents

Examination of the elfects of temperature and solvent on the conformational

order, specifically the solvent-induced ordering of these materials, extends our

understanding of the complex solvation of these stationary phases. Figures 7.10 and 7.11

show plots of I[Va(CH2)]/I[Vs(CH2)] (Figures 7.10a and 7.1 la) and Vs(CH2) peak

frequency (Figures 7.10b and 7,1 lb) as a function of temperature for TFC18SF (Figure

7.10) and DFC18SF (Figure 7.11) in water, acetonitrile, methanol, acetic acid, ethanol,

and 1-butanol. For TFC18SF in all solvents except 1-butanol (Figure 7.10a),

temperature-induced disordering of the alkyl chains is observed with the range of

I[Va(CH2)]/I[Vs(CH2)] values observed at high temperatures considerably less dependent

on solvent than at low temperatures. The extent of alkyl chain coupling, as measured by

the VS(CH2) peak frequency (Figure 7.10b), decreases slightly with increasing

temperature for TFC18SF in these solvents as expected based on the behavior of the

I[Va(CH2)]/l[Vs(CH2)] values.

In 1-butanol, TFC18SF is more disordered at all temperatures than in the other

self-associating solvents based on its lower values of I[Va(CH2)]/I[Vs(CH2)]. Interestingly,

however, the extent of alkyl chain coupling as indicated by the Vs(CH2) peak frequency is

the highest of all solvents at the lower temperatures, and remains relatively high at all

temperatures. These observations are both consistent with 1 -butanol intercalation into the

alkyl chains.

1.7-

1.6-

1.5--

f f i " ' 1.4-u,

1.4-

> 1.3--

r~i 1.2-X -

u 1.1 -

1—1 1.0-

0.9-

0.8-

i • k

-20 -10 T—'—r

10 20 30

Temperature ("C)

T—'—r 40 50 60

2850

I -20 -10 0 10 20 30 50 60

Temperature (°C)

Figure 7.10 (a) Ifv^{CH2)]/I[Vj,(CH2)] and (b) v/CHj) peak frequency as a function of temperature for TFC18SF in methanol (•), acetonitrile (O), water (^), acetic acid (V), ethanol (A), and 1-butanol (•). The error bars represent one standard deviation.

302

1.7-

1 . 6

1 .5H

5 2° 1.3-

S 1.2^ 5 u.

^ 1.0

0.9-

0.8

V

t •

a

I I

—1 1 1 1 1 1 1 1 1 . 1 1 1 r 1 r--20 -10 0 10 20 30 40 50 60

Temperature ("C)

2854-

0 2852 -g

1 2850-

I £ 2848-

3' U

2846-

2844-

-20 -10 0 10 20 30 40 50 60 Temperature ("C)

P -]—I—I—I—I—,—I—1—I—I—I—I—I—i—r

Figure 7.11 (a) I[v^(CH2)]/T[v/CH2)] and (b) v/Ctlj) peak frequency as a function of temperature for DFC18SF in methanol (•), acetonitrile (O), water (^), acetic acid (V), ethanol (A), and 1-butanol (•). The error bars represent one standard deviation.

303

Essentially identical observations were made for the behavior of the

1 [Va(CH2)]/l[VS(CH2)] conformational order indicator and the VsCCHj) peak frequency for

DFC18SF in all solvents (Figures 7.11). T emperature-induced disordering of the alkyl

chains is predominant with a convergence to a minimum value of conformation order at

high temperatures in all solvents except 1-butanol. The disordering effects of these

solvents, as manifested in both spectral indicators, are greater for DFC18SF than for

TFC18SF, consistent with the lower surface coverage of the former.

This interplay between temperature-induced and solvent-induced disorder is

observed not only for the most homogeneous of these high-density materials (the surface-

polymerized TFC18SF and DFC18SF), but also for the solution-polymerized materials of

intermediate coverage (TFC18SL and DFC18SL) and for the low coverage

monofunctional phase (MFC 18) as well. Figures 7.12-7.14 show plots of

I[va(CH2)]/I[Vs(CH2)] as a function of temperature for TFC18SL (Figure 7.12), DFC18SL

(Figure 7.13) and MFC 18 (Figure 7.14). As surface coverage decreases from 5.26

[j-mol/m^ (TFC18SL) to 3.09 pmol/m^ (MFC 18), the inherent disorder becomes so large

that alkyl chain disordering becomes primarily a function of solvent and less dependent

on temperature over the range of temperatures studied. For DFC18SL and MFC 18 in all

solvents except 1 -butanol, solvent-induced changes in stationary phase order are greater

than temperature-induced changes for temperatures greater than 20 "C In 1-butanol,

order is largely temperature-independent over the entire range of temperatures

investigated, especially for MFC 18. Collectively, the results for temperature effects on

304

ffi 1.4

-20 -10 0 10 20 30 40 50 60 Temperature ("C)

Figure 7.12 I[Va(CH2)]/I[v/CH2)] as a function of temperature for TFC18SL in methanol (•), acetonitrilc (O), water (^), acetic acid (V), ethanol (•), and 1-butanol (•). The error bars represent one standard deviation.

305

ffi 1.4

II

-20 -10 0 10 20 30 40 50 60 Temperature (°C)

Figure 7.13 ][v^(CH2)]/f[v^(CH2)] as a function of temperature for DFC18SL in methanol (•), acetonitrile (O), water (^), acetic acid (V), ethanol (•), and 1 -butanol (•). The error bars represent one standard deviation.

306

V • -20 -10 0 10 20 30 40

Temperature (°C)

Figure 7.14 I[Va(CH2)]/I[Vg(CH2)] as a function of temperature for MFC18 in methanol (•), acetonitrile (O), water (*), acetic acid (V), ethanol (•), and 1 -butanol (•). The error bars represent one standard deviation.

307

alkyl chain order for these high-density stationary phases in polar, self-associating

solvents are similar to those in other polar and nonpolar solvents.

In addition to the general interplay between solvent- and temperature-dependent

order of these phases, subtle differences in response to temperature changes were

observed for different solvents. The molecular basis of these subtle differences has been

discussed in detail previously for this class of materials in polar and nonpolar solvents, in

Chapter 6. Similar arguments are adequate to describe the effect of temperature on these

materials in these self-associating solvents. As noted previously, the distinct temperature-

induced phase changes observed for these materials in air are muted by solvation of the

alkyl chains. In some cases, more subtle phase changes were observed that were

dependent on the extent of solvation and the surface coverage of these materials. For

example, the values of I[Va(CH2)]/I[Vs(CH2)] for all five stationary phase materials

(Figures 7.10a, 7.1 la, 7.12-7.14) in acetic acid, methanol, acetonitrile, water, and 1-

butanol exhibited low temperature plateaus where order is independent of temperature.

At higher temperatures in these solvents and in ethanol, the order of these materials

decreased with increasing temperature. As surface coverage decreased, the slope of this

decrease in I[Va(CH2)]/I[Vs(CH2)] with temperature became smaller as well, becoming

essentially zero for MFC 18 in 1 -butanol, water, and acetonitrile. Although the

interpretation of these results must be made with the understanding that the values of this

spectral order indicator do not change proportionately with absolute conformational

308

order, they support the subtle interplay between temperature- and solvent-dependent

ordering in these systems.

Conclusions

A detailed examination of solvent-stationary phase interactions probed by Raman

spectroscopy for a series of high-density alkylsilane materials in polar, self-associating

solvents is presented. Results indicate that conformational ordering of these stationary

phases in solvent is dependent on solvent self-association ability, alkylsilane surface

coverage, and most importantly, surface homogeneity of the alkyl chains. Solvent

interactions with the two homogeneous stationary phases for all self-associating solvents

except 1-butanol can best be described by an adsorptive model of interaction in which

alkyl chain ordering is observed due to solvent self-association at the distal methyl end of

the alkyl chains. In contrast, 1 -butanol is more alkane-like in character and can

intercalate into the alkyl chains of these homogeneous materials according to a

partitioning model of interaction. The low surface coverage and heterogeneous

distribution of alkyl chains in the two solution polymerized and mono&nctional

stationary phases leads to partitioning of these self-associating solvents into these

materials.

In addition to this solvent-induced ordering of these homogeneous, high-density

materials, an interplay between solvent- and temperature-dependent ordering of these

materials was observed, where more solvent-dependent order was observed for lower

309

coverage materials in more nonpolar solvents (1-butanol) and more temperature-

dependent order was observed for higher coverage materials in more polar, self-

associating solvents.

Appendix 7.1


stationary phases in perdeuterated ethanol, acetic acid, and 1 -butanol at-15, -10, 0, 10,

30, 40, and 50 "C.

TFC18SF TFC18SF TFC18SF

TFC18SL TFCI8SL 'C18SL

3 DFC18SF DFC18SF DFCISSF

DFC18SL DFC18SL DFC18SL

MFC18 CIS €18

2900 3000 2800 2800 2900 3000 2800 2900 3000

Wavenumbers (cm"') Wavenumbers (cnr^) Wavenumbers (ciirO

Figure A7.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFCISSF, DFC18SL, and MFC 18 in ethanol at (a) -15 "C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFCISSF, and DFC18SL are 2-min. for MFC 18 are 5-min.

TFC18SF TFC18SF

FC18SL

PFC18SF »FC18SF

DFC18SL 1FCI8SL

MFC 18 MFC 18

2900 3000 2800 3000 2800 2900

Wavenumbers (cm-') Wavenumbers (cm-^)

Figure A7.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in ethane I at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min.

TFC18SF

TFC18SL

FC18SF

FC18SL

MFC18

2800 2900 3000

Wavenumbers (cm'^ )

TFC18SF

TFC18SL

DFC18SF

FC18SL

MFC18

2800 2900 3000

Wavenumbers (cm'\)

Figure A7.3 Raman spectra in the v(C-B[) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in ethanol at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min.

U)

.6" s

±J~ X^FCISSF

^VTFCISSL

*

XmFClSSF

^PFC18SL

\MFC18

2800 2900 3000

Wavenumbers (cm'O

TFC18SF

TFC18SL

FC18SF

DFC18SL

MFC18

2800 2900 3000

Wavenumbers (cm"' )

c k FC18SF

k

A / FCMSL

/Hw DFC18SF

-J" / \DFC18SL

t \MFC18

2800 2900 3000

Wavenumbers (cm'O

Figure A7.4 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetic acid at (a) -15 °C, (b) -10 "C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min.

U)

TFC18SF TFC18SF

TFC18SL TFC18SL

PFC18SF IFC18SF

DFC18SL 'FC18SL

MFC18 .MFC 18

3000 2900 2800 2900 3000 2800

Wavenumbers (cm"') Wavenumbers (cm"0

Figure A7.5 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in acetic acid at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, w and DFC18SL are 2-min, for MFC18 are 5-niin.

TFC18SF TFC18SF

JFC18SL C18SL

PFC18SF •FC18SF

DFC18SL DFC18SL

MFC18 MFC18

3000 2900 2800 3000 2800 2900

Wavenumbers (cm"0 Wavenumbers (cm"^)

Figure A7.6 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFCI8SL, and MFC18 in acetic acid at (a) 40 °C and (b) 50 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, w and DFC18SL are 2-min, for MFC 18 are 5-min. ^

CD

s

Xw TFC18SF

^'•\rFC18SL

^-\DFC18SF

\DFC18SL

J" V^FC18

2800 2900 3000 Wavenumbers (cm"')

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC 18

2800 2900 3000

Wavenumbers (cm"' )

TFC18SL

DFC18SF

DFC18SL

/VA\ / V. MFC 18

2800 2900 3000 Wavenumbers (cm"')

Figure A7.7 Raman spectra in the v(C-H) region of TFC I8SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in 1 -butanol at (a) -15 "C, (b) -10 °C, and (c) 0 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min. ON

•t I

Nw TFC18SF

r >JFCI8SL

N ^\OTC18SF

\DFC18SL

MFC18

2800 2900 3000

Wavenumbers (cm"^)

IJ- \^TFC18SF

\TFC18SL

J" N \OTC18SF

\PFC18SL

H

As \MF^

2800 2900 3000

Wavenumbers (cm'\)

Figure A7.8 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFCISSL, and MFC18 in 1 -butanol at (a) ICC and (b) 30 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFCISSL are 2-min, for MFC 18 are 5-niin.

s

TFC18SF

/ \lTC18SL

DFC18SL

/X \ MFC18

2800 2900 3000

Wavenumbers (cm*^)

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC 18

2800 2900 3000

Wavenumbers (cm'^)

Figure A7.9 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in 1 -butanol at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFCI8SF, and DFC18SL are 2-min, for MFC 18 are 5-min. ©0

319

Chapter 8


Raman Spectroscopy; Effects of Neutral and Basic Aromatic Compounds

Introduction

Separation in reversed-phase liquid chromatography (RPLC) is often based on

small differences in polarity of analytes within a class of compounds. This is observed

by the separation of countless classes of small molecules, including the popular

separations of substituted aromatic compounds as standards to determine column

performance. Substituted aromatics are structural components of environmentally,

pharmacologically and biologically interesting molecules including polycyclic aromatic

hydrocarbons (PAHs), polychlorinated biphenyl congeners (PCBs), and carotenoids.

Furthermore, substituted aromatics of varying polarity are used to probe fundamental

interactions in RPLC. Thus, investigation of solute-stationary phase and solvent-solute

interactions involving these core aromatic systems under chromatographic conditions

would lead to a greater understanding of the molecular basis of their retention as well as

the retention of more complex aromatic-containing analytes.

As part of previous Raman spectral studies on the effects of polar and nonpolar

solvents on the conformational order of high-density alkylsilane stationary phases in

Chapter 6, spectral data were acquired for two nonpolar, aromatic solvents, benzene and

320

toluene. Interestingly, the addition of the methyl group in toluene results in significant

changes in interaction with these high-density stationary phases relative to benzene. In

general, these stationary phases are at least as disordered, if not more disordered, in

toluene compared to benzene at all temperatures, although this behavior varies with

surface coverage and alkylsilane surface homogeneity. In this report, the effects on

conformational order of these high-density stationary phase materials of the neutral and

basic aromatic compounds benzene, toluene, p-sylene, anisole, aniline, and pyridine were

investigated using Raman spectroscopy. Although only benzene and toluene are

prominent in RPLC as mobile phase components and mobile phase additives, all five of

these solutes are common analytes in RPLC. In particular, the behavior of these aromatic

compounds is important because of the analytical problems produced by substantial peak

tailing in reversed-phase separation of basic analytes of pharmaceutical or forensic

interest.

Goals of this Work

This work is part of a continuing effort toward understanding molecular retention

at the chromatographic interface, specifically complex analyte-solute-stationary phase

interactions. The previous three chapters in this Dissertation describe the inter- and

intramolecular interactions within the stationary phase (Chapter 5) and solvent-stationary

phase interactions (Chapters 6 and 7). This report focuses on analyte-stationary phase

interactions, albeit not in an experimental chromatographically relevant system. These

321

experiments probe interactions of neat analytes in the absence of solvent, but are crucial

for understanding all the individual components at the PRLC interface, before these

analytes can be investigated at chromatographic concentrations in the presence of solvent.

Methodology


characterized as previously described.''Perdeuterated anisole (99%) was purchased

from Aldrich. Perdeuterated benzene (99.5%), toluene (99.8%), p-xylene (98%), aniline

(98%), and pyridine (99.5%) were obtained from Cambridge Isotope Laboratories. All

solvents are used as received.


material into a 5-mm dia NMR tube; 200 piL of solvent was then added. Samples were


acquisition. Raman spectra were collected, processed, and analyzed using the same

instrument parameters and conditions described in Chapter 2. Models of the

sol vent/stationary phase interface were constructed and energy-minimized using the same

software and mathematical parameters as described in Chapter 2.

Raman Spectroscopy of High-Density Alkvlsilane Stationarv Phases in Solvent

Absolute retention in RPLC is governed by numerous factors, but is most

influenced by mobile-phase composition. Understanding solvent-stationary phase

322

interactions at the molecular level is key to elucidating retention processes as a whole.

The aromatics of interest here can be viewed as both mobile phase components and

solutes in chromatographic systems. Chapters 4-7 demonstrate the utility of Raman

spectroscopy to probe alkylsilane stationary phase order and structure in the presence of

solvent as a measure of these molecular interactions. This utility stems from the

abundance of spectral indicators of conformational order, especially in the v(C-H) region

between 2750 and 3050 cm"'. Furthermore, spectral interrogation of conformational

order in this region is achieved in the presence of solvent without significant spectral

interference from the silica substrate or surface-adsorbed water. Although there is

significant conformational order information for alkane systems in the v(C-C) and 8(C-

H) regions between 900 and 1500 cm'\ these stationary phase materials were prepared on

silica particles that possess innate fluorescence that interferes with the acquisition of

quality Raman spectra in this region. Therefore, for these materials, assessment of

conformational order is limited to spectral data obtained in the v(C-H) region.

Typical spectra in the v(C-H) region (2750 to 3050 cm"') for all five high-density

stationary phase materials at room temperature in pyridine, aniline, anisole, and p-xylene

are shown in Figures 8.1 and 8.2. Raman spectra of these phases in anisole, p-xylene,

aniline, and pyridine at temperatures from -15 to 50 °C are given in Appendix 8.1.

Raman spectra in the v(C-H) region for these stationary phases in benzene and toluene

are shown in Figure 5.3 (Chapter 5).

323

TFC18SF

TFC18SL

<u DFC18SF

DFC18SL

MFC18

2800 2900 3000

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

2800 2900 3000 Wavenumbers (cm-') Wavenumbers (cm-')

Figure 8.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC I8SF, DFC18SL, and MFC 18 in (a) anisoic and (b) p-xylene. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min; for MFC!8 are 5-min in all solvents.

324

Va(CH,

TFC18SF

TFC18SL

<D DFCISSF

DFC18SL

MFC18

2800 2900 3000

TFC18SF

TFC18SL

DFCISSF

DFC18SL

MFC18

2800 3000 2900 Wavenumbers (cnr' ) Wavenumbers (cm-^

Figure 8.2 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFCISSF, DFC18SL, and MFC 18 in (a) aniline and (b) pyridine. Integration times for TFC18SF, TFC18SL, DFCISSF, and DFC18SL are 2-min; for MFC 18 are 5-min in all solvents.

325

The basis for interpretation of the Raman spectral results for surface-bound

alkylsilanes in the v(C-H) region using the intensity ratio of the VaCCHj) to the VsCCHi)

{I[Va(CH2)]/I[Vs(CH2)]}, the VaCCHs) and Vs(CH2) peak frequencies, and the intensity

ratio of the VS(CH3)IR to the VS(CH2) {I[VS(CH3)FR]/I[VS(CH2)]} was discussed in Chapters

2 and 5 in this Dissertation.

Effects of Aromatic Compounds on the Order of These High-Density Stationary Phases

In previous chapters, solvent-stationary phase interactions for these high-density

materials were quantified by the solvatochromic parameter, TC*, as described by Kamlet et

al.^^ In this report, however, because these aromatic compounds can be treated as either

solvents or solutes in chromatographic systems, we have chosen not to limit the

description of these systems to solvents, but rather to describe them in terms of octanol-

water partition coefficients (log Kow) which are measures of hydrophobicity applicable to

both solvents and solutes. Kow values for the aromatic systems studied here are tabulated

"L-l O 1 8,7-8,8 in Table 8.1.

In Figure 8 .3,1[va(CH2)]/I[Vs(CH2)] is plotted as a function of log K,w for all five

stationary phases in neat solutions of each aromatic compound at 20 "C. In general,

I[Va(CH2)]/I[Vs(CH2)] increases with increasing alkylsilane surface coverage in these

neat aromatic compounds. This behavior is identical to that observed for these materials

in air, polar solvents, and nonpolar solvents, shown in Chapters 5-7. In addition,

I[Va(CH2)]/I[Vs(CH2)] varies with hydrophobicity and polarity of these aromatic

326

Table 8.1. Octanol-water partition coefficients for aromatic compounds"

Compound log K^w

Pyridine 0.65

Aniline 0.90

Anisole 2.11

Benzene 2.13

Toluene 2.73

p-Xylene 3.15

"From references 8.7 and 8.8,

327

U

K

1.7 -

1. 6 -

1.5 -

1.4- Air

1.3-% 1.2 - 1

1 1

1.1-

1.0 -! 1 1

0.9 -1 1 1 1

08-J I 1

" i

Anisole

•*1 : Benzene

Toluene p-Xylene

Aniline

Pyridine

ow

Figure 8.3 I[Vg(CH2)]/I[v/CH2)] as a function of log K,,^, for TFC18SF (•), TFC18SL (A) , DFC18SF (O), DFC18SL (•), and MFC18 (V) in p-xylene, toluene, benzene, anisole, aniline, and pyridine. The error bars represent one standard deviation.

328

compounds. With the exception of anisole, which clearly exhibits anomalous behavior as

discussed below, the I[Va(CH2)]/I[Vs(CH2)l values for the high-coverage, homogeneous

stationary' phases (TFC18SF and DFC18SF) are fairly constant with increasing

hydrophobicity (log K<M) for air, pyridine, and benzene, and then decrease slightly with

increasing log Kow from benzene to p-xylene. For TFC18SL, l[Va(CH2)]/I[Vs(CH2)]

decreases relative to air for all aroniatics, especially p-xy!ene. For DFC18SL and

MFC 18,1 [Va(CH2)]/l[VsCCHi)] decreases relative to air for all aromatics, with the

disorder generally increasing with log Kow

In the most hydrophobic, nonpolar aromatic compounds toluene and p-xylene, the

values of I[Va(CH2)]/I[ Vs(CH2)] for all five stationary phases were the lowest of all of the

aromatic compounds investigated, suggesting the least conformational order of the alkyl

chains. These observations are rationalized in terms of greater interaction of these

hydrophobic compounds with the alkyl chains. Indeed, the significant values of the

dissolution enthalpies for toluene and p-xylene in hexadecane®'^ of-8.58 and -9.89

kcal/mol, respectively, support the interpretation of this behavior in terms of deep

penetration (i.e. partitioning in the chromatographic sense) of these aromatics into the

alkylsilanes. The methyl substituents on each of these compounds confers additional

hydrophobicity relative to benzene, for example, that facilitates this deeper penetration.

The effect of benzene on the order of these materials was more dependent on

surface coverage than the effect of either toluene or p-xylene. MFC18 and DFC18SL

were significantly more disordered in benzene than in air, TFC18SL and DFC18SF were

329

slightly more disordered in benzene than in air, but TFC18SF was relatively unchanged

in benzene relative to air.

Interestingly, all stationary phases except TFC18SL were significantly more

ordered in anisole than in air at 20 °C. Similar ordering had previously been observed for

TFC18SF and DFC18SF in self-associating solvents (e.g., methanol, ethanol, acetic acid

and acetonitrile) discussed in Chapter 7, and had been attributed to self-association of

solvent molecules at the distal methyl end of the alkylsilane chains in a manner that

actually increases chain order. This ordering phenomenon had been observed exclusively

for stationary phases composed of alkyl chains that are both high in surface coverage and

homogeneously distributed across the silica surface (as determined by NMR

spectroscopy^^).

In anisole, not only were the two high-coverage, homogeneous materials ordered

relative to air, but the materials with lower surface coverage were ordered relative to air

as well. Anisole is known to strongly self-associate through hydrogen-bonding

interactions at the oxygen atom.^Thus, a picture similar to that proposed earlier for

other systems in which self-association of the solvent at the distal methyl end of the alkyl

chains holds the chains together is adequate to account for ordering of the high-coverage,

homogeneous stationary phases. However, the observation of anisole-induced ordering

of the heterogeneous, low-coverage materials MFC 18 and DFC18SL was initially

puzzling, since other self-associating solvents such as ethanol do not order these

stationary phases. If one considers differences in solubility behavior and structure

330

between anisole and ethanol, for example, the difference in ordering of these phases in

these two media becomes clear. Based on the considerably larger octanol-water partition

coefficient for anisole (log Kow = 2.11) relative to ethanol (log Ko„ - -0.30) and the

larger enthalpy of dissolution in hexadecane of anisole (Al ls = -9.90 kcal/mol) relative to

ethanol (AHs = - 3 .90 kcal/mol),^^"^^ one would predict that the interaction of anisole

with the stationary phase should be greater than that of ethanol. In part, the difference in

enthalpy of dissolution is due to the much larger size of anisole (Vn, = 0.108 nm') relative

3 8 118 12 to ethanol (Vm = 0.053 nm ). ' • However, even when corrected for these molecular

volume differences, the dissolution of anisole in hexadecane (AHg/Vm = -92

kcal/mol-nm') is favored over the dissolution of ethanol in hexadecane (AH/Vm = -74

kcal/mol-nm'). This strong interaction, a result of solubility and structural characteristics,

coupled with the ability of anisole to self-associate at the stationary phase-solvent

interface, provide favorable energetics for both the high- and low-coverage stationary

phases to order in anisole relative to air.

Although aniline and pyridine are significantly more hydrophilic than the other

four aromatic compounds as indicated by their log Kow values, the values of

I[Va(CI l2)]/I[VS(CH2)] in these solvents are low suggesting that they penetrate deeply into

the alkyls chains and impart significant disorder to these alkylsilanes. For aniline and

pyridine, this partitioning into the alkyl chains should be thermodynamically comparable

to that of the other hydrophobic aromatics based on their enthalpies of dissolution in

hexadecane (AHs (aniline) = -9.99 kcal/mol, AHs (pyridine) = -7.80 kcal/mol).Indeed,

one might predict slightly less interaction of pyridine with the alkyl chains based on its

slightly less favorable enthalpy. In contrast to this prediction, however, these stationary

phases are generally slightly more disordered in pyridine than aniline. This behavior is

proposed to be due to the additional driving force for partitioning of pyridine due to

hydrogen-bonding with unreacted surface silanols (also called silanophilic interactions),

which results in peak tailing in chromatographic experiments.*^ ' Pyridine is a

slightly stronger base than aniline [Ka (pyridine) = 5.23, Ka (aniline) = 4.60] which

increases its hydrogen-bonding to surface silanols.

The effect of these additional Bronsted acid-base interactions is expected to be a

tunction of the number of unreacted surface silanols which scales inversely with

alkylsilane surface coverage. The increased conformational disorder of TFC18SF, in

which the number of unreacted surface silanols is low, in pyridine relative to aniline is

attributed to its slightly greater base strength and smaller molecular size. For the other

stationary phases, lower alkylsilane surface coverage allows substantial access to surface

silanols, and the conformational order of these materials in aniline and pyridine is

approximately equivalent. Unfortunately, the expected shifts due to hydrogen-bonding in

the silanol v(O-H) and the perdeuterated aniline v(N-D) modes cannot be observed as a

result of spectral interference from the bulk solvent and silica. Nonetheless, even without

direct Raman spectral evidence to support hydrogen-bonding of these species, the

plethora of chromatographic data that identify silanophilic interactions for basic analytes

332

support this interpretation of the conformational order of alkylsilanes in these

systems.

In addition to the confonnational order information obtained from

1 [ Va(CH2 )]/'T [ VS(CH2)] values, the effect of aromatic compounds on alkyl chain coupling

in these high-density stationary phases can be determined from the VsCCHa) peak

frequency. Figure 8.4 is a plot of Vs(CH2) peak frequency as a function of log Kow for

these stationary phases in perdeuterated p-xylene, toluene, benzene, anisole, aniline and

pyridine. As seen for these materials in air and in other solvents, the VsCCHi) peak

frequency depends on surface coverage in these aromatic compounds, generally

increasing with decreasing surface coverage. For the highest coverage materials, the

VsCCHj) peak frequency is generally constant with log K^w However, for MFC 18, the

VsCCHi) peak frequency increases with increasing log Kow from pyridine to anisole, and

then decreases from anisole to p-xylene.

For TFC18SF, aromatic compounds have very little effect on alkyl chain

coupling. For the intermediate coverage materials, TFC18SL and DFC18SF, appreciable

disorder in p-xylene relative to air is suggested by the value of I[va(CH2)]/I[Vs(CH2);

however, the VsCCHa) peak frequencies for these systems in p-xylene suggests greater

alkyl chain coupling in p-xylene relative to air, despite the apparent conformational

disorder. The contradiction between these two spectral indicators, albeit atypical, may

not be too surprising considering that they are measures of two separate parameters

(Chapter 3). Nonetheless, this apparent discrepancy indicates the need for a more

333

2854-

2852-

2850-

2848-

Benzene Anisole p-Xylene

^ Toluene Pyridine ^ Aniline

0.0 0.5 1.0 1.5 2.0

logK c\\\r

z

Figure 8.4 v/CHj) peak frequency as a fiinction of log for TFC18SF (•), TFC18SL (A), DFC18SF (O), DFC18SL (•), and MFC 18 (V) in p-xylene, toluene, benzene, anisole, aniline, and pyridine. The error bars represent one standard deviation.

334

systematic study of the relationship between chain coupling and conformational order in

these alkylsilane systems as would be accessible by access to a larger number of spectral

indicators, especially those in the frequency region containing information about true

gauche C-C bonds. Given that the inherent fluorescence of the silica substrates on which

these materials are based precludes access to this region of the spectrum, the resolution of

this discrepancy awaits the fabrication of a similar series of systems on non-fluorescent

silica. We hope to undertake the fabrication of such systems and will report our spectral

findings on these systems at a later date.

For the lowest coverage materials, DFC18SL and MFC 18, the alkyl chains are

more coupled in aromatic compounds than in air. In anisole, the increase in chain

coupling is not as great in the disordering aromatic solvents. For stationary phases with

intermediate and low surface coverages (3.09 to 5.26 [imol/m^), alkyl chains are more

coupled in these aromatic compounds than in air at room temperature. For the high

coverage materials, interaction at the distal end of the alkyl chains decreases coupling at

the end of the alkyl chains.

The conformational order and chain coupling information obtained from the

Raman spectroscopy of these high-density stationary phases in aromatic compounds

allow development of detailed molecular pictures of the solvent/stationary phase

interface. These pictures were created with Chem3D using energy minimized MM2

computations on a small matrix of alkylsilanes. The proposed pictures of these interfacial

interactions are shown in Figures 8.5-8.9 for TFCl 8SF in anisole (Figure 8.5), and

Molecular picture of the interactions between TFC18SF and anisole at the chromatographic interface.

336

i

f

Figure 8.6 Molecular picture of the interactions between TFC18SF and toluene at the chromatographic interface.

337

•

I r

Figure 8.7 Molecular picture of the interactions between MFC 18 and anisolc at the chromatographic interface.

338

Figure 8.8 Molecular picture of the interactions between MFC] 8 and aniline at the chromatographic interface.

j.r i i ri ,i''' ft v', F -i 1^ • yV*

ir /f # -X" VU V« V

;

X > '

^ A

. 4'

T ^

r 9v,

.. -?

SA *} X

W-'".: '-r

Figure 8.9 Molecular picture of the interactions between MFC 18 and toluene at the chromatographic interface.

340

toluene (Figure 8.6), and MFC 18 in anisole (Figure 8.7), aniline (Figure 8.8), and toluene

(Figure 8.9), respectively, as representative of the various interactions proposed for high

and low coverage materials in these aromatic compounds. The interaction of anisole at

the distal methyl group of the alky! chains of TFC18SF and MFC 18 allows lateral

ordering of the anisole molecules through hydrogen-bonding. Moreover, this interaction

induces conformational ordering of the alkyl chains with slight decoupling of the distal

portion for the close-packed TFC18SF but slight coupling of the distal end of the chains

for MFC 18. For TFC18SF in toluene, the hydrophobic methyl groups can interact with

the distal methyl groups of the alkyl chains and induce rotational disorder in the ends of

the chains. For MFC 18 in aniline, it is proposed that the high fraction of exposed surface

silanols causes penetration of aniline molecules to the silica surface to establish

hydrogen-bonds. As a result, the alkyl chains are slightly more disordered than in air

because of the cavity formed to accommodate the aniline molecules at the surface, even

though these chains couple at the distal end away from the silica surface. For MFC 18 in

toluene, significant disorder is imparted to the alkyl chains with the retention of large

portions of coupled alkyl chains.

F-ffects of Temperature

Examination of the effects of temperature and solvent on the conformational order

of these high-density stationary phases extends our understanding of the complex

interactions of these materials with aromatic compounds as both solvents and analytes in

341

chromatographic systems. Figures 8.10 and 8.11 shows plots ofI[va(CH2)]/I[Vs(CH2)]

and VsCCHs) peak frequency as a function of temperature for TFC18SF (Figures 8.10a

and b) and MFC 18 (Figures 8.1 la and b) in p-xylene, toluene, benzene, anisole, aniline,

and pyridine. For TFC18SF (Figure 8.10a), temperature-induced disordering of the alkyl

chains was generally observed with a convergence at high temperature to a minimum

value of I [Va(CH2)]/I[Vs(CH2)]. The exception to this behavior for TFC18SF was anisole

in which the alkyl chains remained more ordered at high temperatures than in the other

aromatics. In addition, a small degree of temperature-induced chain decoupling was

observed for TFC18SF in all compounds with convergence to a maximum decoupled

state at high temperatures (Figure 8 1 Ob).

For MFC 18, conformational order was much less temperature-dependent for

toluene and benzene and was essentially independent of temperature for anisole, pyridine,

and aniline. Similarly, alkyl chain decoupling for MFC 18 was exclusively solvent-

dependent for all aromatic compounds with the exception of p-xylene. This interplay

between solvent- and temperature-induced disordering and chain decoupling as a function

of surface coverage is consistent with observations made for these stationary phases in

polar and nonpolar solvents. Similar trends were observed for the intermediate surface

coverage materials, TFC18SL, DFC18SF, and DFC18SL, as indicated by the plots of

I[Va(CH2)]/I[Vs(CH2)] as a function of temperature in Figures 8.12-8.14.

At low temperatures, all stationary phases are more ordered in p-xylene than in

any other aromatic compound or indeed any solvent from previous investigations from

¥

• I

342

-20 -10 0 10 20 30 40 50 60 Temperature ("C)

2854-

! I i y. 2846

0 10 20 30 40 Temperature ("C)

50 60

Figure 8.10 (a) I[v^(CH2)] /Ifv^CCHj)] and (b) v/CHj) peak frequency as a function of temperature for TFCl 8SF in benzene (•), aniline (•), toluene (O), p-xylene (•), anisole (V), and pyridine (*). The error bars represent one standard deviation.

S 1.6

S 1.4

-20 -10 0 10 20 30 40 50 60 Ternperature (°C)

2852-

H

2848-

y, 2846

2844-

-20 -10 0 10 20 30 40

Temperature ("O

50 60

Figure 8.11 (a) l^v/CHj)] and (b) v^CCHj) peak frequency as a function of temperature for MFC 18 in benzene (•), aniline (A), toluene (O), p-xylene (•), anisole (V), and pyridine (*). The error bars represent one standard deviation.

344

* i

I f

-20 -10 0 10 20 30 40 50 60

Temperature ("C)

Figure 8.12 ^v^CCHj)] /I[v/CH2)] as a function of temperature for TFC18SL in benzene (•), aniline (•), toluene (O), p-xylene (•), anisole (V), and pyridine (^). The error bars represent one standard deviation.

345

® 1.6

n

•20 40 0 10 20 30 40 50 60 Temperature (""C)

Figure 8.13 (a) l[Vj,(CH2)] as a function of temperature for DFC18SF in benzene (•), aniline (•), toluene (O), p-xylene (•), anisole (V), and pyridine (*). The error bars represent one standard deviation.

346

1.6-

$ II I

-20 -10 0 10 20 30 40 50 60 Temperature ("C)

Figure 8.14 I[v3(CH2)] /IjVj^CCHj)] as a ftinction of temperature for DFC18SL in benzene (•), aniline (•), toluene (O), p-xylene (•), anisole (V), and pyridine (•). The error bars represent one standard deviation.

347

Chapters 5-7. Neat p-xylene freezes at -13 °C and the sample visually appears to be a

crystalline solid at temperatures < 5 "C. Lowering the temperature after the room

temperature addition of p-xylene to the stationary phase freezes the p-xylene and the

solvated stationary phase resulting in a highly ordered state that exhibited slightly less

chain coupling than in neat crystalline hydrocarbons (e.g., octadecane, polyethylene),

shown in Chapter 3. Once the temperature is sufficiently elevated to melt the p-xylene-

stationary phase system, p-.xylene-alkyl chain interactions predominate and significant

disorder and chain decoupling is imparted to the stationary phase.

Conclusions

A detailed examination of interactions between aromatic compounds and high-

density stationary' phases probed by Raman spectroscopy is presented. Results indicate

that conformational order of these materials is a function of the degree and nature of the

aromatic substituents and how these substituents impact the properties of the aromatic

compound (e.g., polarity, hydrophobicity, steric bulk, hydrogen-bonding ability).

Interaction of the hydrophobic aromatic compounds benzene, toluene, and p-xylene with

these stationary phases is best described in terms of a partitioning model in which

significant disorder is imparted to these stationary phase alkyl chains. The interaction of

the basic aromatic compounds aniline and pyridine with these phases is proposed to

reflect a combination of hydrophobic interactions with the alkyl chains and silanophilic

interactions with silanols on the silica surface. Unfortunately, direct Raman spectral

evidence for these silanophilic interactions is unavailable due to the small number of

surface hydrogen-bonded molecules relative to those in the bulk, although such

interactions are clearly supported by chromatographic data for basic compounds. Anisole

interacts with the distal portions of these phases in a manner that allows lateral hydrogen-

bonding at the methoxy oxygen atom between anisole molecules and induces

conformational ordering of the alky! chains.

This work proposed interactions of neat aromatic compounds, commonly used as

solutes in RPLC, with alkylsilane stationary phases of varying surface coverage. While

these systems do not mimic real chromatographic systems, information obtained from

these data provides additional information about molecular interactions at the alkylsilane-

solvent interface. In addition, this work lays the foundation for future work including

studies of mixed mobile phase-solute systems with these high-density alkylsilane

stationary phases.

Appendix 8.1


stationary phases in perdeuterated anisole, p-xylene, aniline, and pyridine at -15, -10, 0,

10, 30, 40, and 50 T.

TFC18SF TFC18SF

TFC18SL TFC18SL

'FC18SF PFC18SF

DFC18SL •FC18SL

MFC 18 €18

3000 2800 2900 2900 3000 2800

TFC18SF

'FC18SL

DFC18SF

DFC18SL

MFC18

3000 2800 2900 Wavenimibers (cnr ') Wavenumbers (cnr ) Wavenumbers (cnr' )

Figure A8.1 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in anisole at (a)-15 "C, (b) -10 "C, and (c) 0 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC18 are 5-min.

<u

AgTC18SF

A \ \TFC18SL

A

\^C18SF

\PFC18SL

V MFC 18

2800 2900 3000 Wavenumbers (cm'^)

TFC18SF

:FCI8SL

FC18SF

DFC18SL

MFC18

2800 2900 3000

Wavenumbers (cm"' )

Figure A8.2 Raman spectra in the v(C~H) region of TFC18SF, TFC1 SSL, DFC18SF, DFC1 SSL, and MFC 18 in anisole at (a) 10 °C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min.

U1 o

TFC18SF

'FC18SL

>FC18SF

•FC18SL

MFC18

3000 2900 2800

TFC18SF

TFC18SL

DFC18SF

IFC18SL

MFC18

3000 2800 2900 Wavenumbers (cm'^) Wavenumbers (cm"0

Figure A8,3 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in anisole at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, w and DFC18SL are 2-min, for MFC 18 are 5-min,

TFC18SF

TFCISSL

DFC18SF

DFC18SL

MFC 18

2800 2900 3000

TFC18SF

TFC18SL

PFC18SF

DFC18SL

MFC18

3000 2800 2900

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC 18

2800 2900 3000

Wavenumbers (cnr^) Wavenumbers (cm"') Wavenumbers (cnr^)

Figure A8.4 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in p-xylene at (a) -15 "C, (b) -10 "C, and (c) 0 °C. Integration times for TFC18SF, TFCISSL, ^ DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min.

TFCI8SF

'FC18SL

DFC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC 18

2800 2900 3000

Wavenumbers (cm-^ Wavenimibers (cm'^)

Figure A8.5 Raman spectra in the v(C-H) region of TFC18 SF, TFC18 SL, DFC18 SF, DFC18 SL, and MFC18 in p-xylene at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, w and DFC18SL are 2-min, for MFC 18 are 5-min.

TFC18SF TFC18SF

TFC18SL TFC18SL

o DFC18SF •FC18SF

DFC18SL DFC18SL

MFC18 MFC 18

3000 2900 2800 3000 2900 2800

Wavenumbers (cm"^) Waveniimbers (cm"')

Figure A8.6 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in p-xylene at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min.

TFC18SF TFC18SF

TFC18SL TFC18SL

DFC18SF •FC18SF

DFC18SL DFC18SL

MFC 18 €18

2800 2900 3000 3000 2800 2900

TFC18SF

TFC18SL

^3 c 4) FC18SF

DFC18SL

MFC 18

2800 2900 3000

Wavenumbers (cnr^) Wavenumbers (cm-') Wavenumbers (cm-\)

Figure A8,7 Raman spectra in the v(C-H) region of TFC18SF, TFC1 SSL, DFC18SF, DFC1 SSL, and MFC 18 in aniline at (a) -15 "C, (b) -10 °C, and (c) 0 °C. Integration times for TFCIBSF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFCIS are 5-min.

TFC18SF TFCIBSF

JFC18SL TFC18SL

'FC18SF DFC18SF

DFC18SL DFC18SL

MFC 18 MFC 18

3000 2900 2800 2900 3000 2800 Wavenumbers (cnr^ Wavenumbers (cm"')

Figure A8.8 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in aniline at (a) 10 "C and (b) 30 "C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min.

TFC18SF

TFC18SL

•FC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

TFC18SL

CD DFC18SF

DFC18SL

MFC 18

3000 2800 2900

Wavenumbers (cm"\) Wavenumbers (cm"\)

Figure A8.9 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in aniline at (a) 40 "C and (b) 50 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min. for MFC 18 are 5-min,

TFC18SF

TFCISSL

DFC18SF

DFC18SL

MFC18

3000 2900 2800

TFC18SF

TFC18SL

DFC18SF

DFC18SL

MFC18

2900 3000 2800

TFC18SF

TFC18SL

m

s >FC18SF

DFC18SL

'C18

2800 2900 3000

Wavenumbers (cm"0 Wavenumbers (cm'\) Wav en umbers (cm"')

Figure A8.10 Raman spectra in the v(C-H) region of TFC18SF, TFCISSL, DFC18SF, DFC18SL, and MFC 18 in pyridine at (a)-15 °C, (b) -10 "C, and (c) 0 "C, Integration times for TFC18SF, TFCISSL, DFCISSF, and DFCISSL are 2-min, for MFC 18 are 5-min. O)

TFC18SF

TFC18SL

FC18SF

DFC18SL

MFC 18

2800 2900 3000 Wavenumbers (cm"')

TFC18SF

TFC18SL

DFC18SF

DFC18SL

/ VJVIFC18

2800 2900 3000

Wavenumbers (cm"')

Figure A8.11 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC 18 in pyridine at (a) 10 "C and (b) 30 °C. Integration times for TFC18SF, TFC18SL, DFC18SF, and DFC18SL are 2-min, for MFC 18 are 5-min, \o

C CD

\ V TFC18SF

'X \tFC18SL

t 'X

\DFC18SF

i APFC18SL

r V. MFC18

2800 2900 3000 Wavenumbers (cm"^)

\{TC18SL

r •

\DFC18SF

\PFC18SL

J a VVMFC18

2800 2900 3000 Wavenumbers (cm°^)

Figure A8.12 Raman spectra in the v(C-H) region of TFC18SF, TFC18SL, DFC18SF, DFC18SL, and MFC18 in pyridine at (a) 40 "C and (b) 50 "C. Integration times for TFC18SF, TFC18SL, DFCI8SF, and DFC18SL are 2-min, for MFC 18 are 5-min.

u> ON o

361

Chapter 9

Preliminary Investigations of Salting-Oiit and pH Effects on the

Conformational Order of Octadecylsilane Stationary Phases by Raman

Spectroscopy

Introduction

Salts effects and the solubility of hydrophobic solutes have been studied in

numerous reports, primarily with regard to "salting-in" or "salting-out" biologically-

relevant solutes from aqueous solution.'' Increasing the solubility of a solute with the

addition of electrolyte to aqueous solution is called "salting-in" while a decrease in

solubility with the addition of electrolyte is called "salting-out ." However, while

electrolyte-containing mobile phases are used routinely in RPLC separations, salt effects

have not been studied exclusively. This is somewhat surprising since computational and

experimental work has shown salts to greatly affect the solvent structure surrounding

hydrophobic solutes, and therefore, could significantly affect the RPLC separation.^'^

McDevit has shown that water molecules that solvate benzene in aqueous solution are

highly ordered and entropically unfavorable.'^^ Upon the addition of salt to the aqueous

benzene solution, the water molecules solvating the benzene solute participate in

solvation of the ions in solution and become less ordered, are thermodynamically more

stable, resulting in "salting-out" of the hydrophobic solute. An additional consideration

for RPLC systems is the effect of salt solutions on structure of the alkyl chains of the

stationary phase. Although the effect of salt solutions on the conformational order of

alkyl-based systems has not been reported in the literature, previous solvent studies

presented in Chapters 4-8 of this Dissertation point to the importance of considering

alkyl-salt solution interactions.

Introducing salt solutions or buffers to mobile phases in RPLC is commonplace

for the separation of compounds that are pH sensitive. The use of pH-controlled mobile

phases has been demonstrated for RPLC separations of numerous compounds including

biologically-relevant amino acids'^ *'"''''and nucleoside bases/ "'^'" forensic

compounds/ quantitation of small acidic molecules in food,^ and acids and

bases of environmental importance.^'^^"^'^^ Thus, it is relevant to study the pH-dependent

interactions of solute molecules with alkylsilane stationary phases from aqueous solution

in addition to the salt-dependent structure of these phases.

Goals of this Work

The main goal of this work was to determine the effects of aqueous salt solutions

on the conformational order of high-density octadecylsilane stationary phases. An

additional goal was to determine the effects of pH on the interactions between acidic and

basic monosubstituted aromatic compounds in aqueous solutions with alkylsilane

stationary phases. In Chapters 6-8, solvent-induced ordering of alkylsilane stationary

363

phases was observed in the presence of neat self-associating solvents, attributed to

interfacial ordering of these molecules. It was hypothesized that aqueous solutions of

self-associating acidic compounds, such as benzoic acid, would have a similar effect on

the conformational order of these phases, but that this effect would be pH-dependent.

Methodology

Aqueous benzoic acid solutions, 37 mM, were prepared in distilled, de-ionized

(DDI) water and were pH adjusted using conc. NH4OH to give five solutions ranging

from ~ pH 2 to 8. Aqueous 37 mM aniline and 37 mM benzylamine solutions were also

prepared in DDI water but were pH-adjusted using conc. HCl to give four solutions

ranging from ~ pH 2 to 11. Water control samples were prepared by adjusting the water

to ~ pH 2 using conc. HCl followed by conc. NH4OH resulting in higher pH solutions.

Concentrated acids and bases were used to adjust the pH in order to avoid significant

sample volume changes.


material into a 5.0-mm-dia NMR tube; 200 jiL aqueous solutions were then added.

Samples were sonicated for 10 min and equilibrated at 20 "C for a minimum of 12 h prior

to spectral acquisition. Raman spectra were collected using 100 mW of 532 nm radiation

from a Coherent (Verdi - 2W) Nd: YVO4 laser on a Spex Triplemate spectograph as

described in Chapter 2. Slit settings of the Triplemate were 0.5/7.0/0.150 mm for all

experiments. Integration times for each spectrum are provided in the figure captions.

364

Low laser power and short acquisition times were used to avoid unnecessary sample

heating.

The data presented in this chapter are somewhat incomplete relative to the

previous studies because of limited quantities of stationary phases available for these

experiments. However, there is sufficient information in the data presented such that

reasonable interactions between these solutions and stationary phases can be determined.

Salting-Out of Alkvlsilane Stationary Phases

While aqueous salt effects on the conformational order of alkylsilane stationary

phases had not been reported previously in the literature, the likelihood of significant

effects on the order and structure of these phases was high because of what was known

about the salting-out of hydrophobic solutes from aqueous solutions.As such

solutes salt out of aqueous solutions, they interacted more with each other than with

water molecules, it is likely that increased solute-solute interaction would have a

significant effect on the order and structure of these solutes. Figures 9.1 and 9.2 show

Raman spectra of TFC18SF (6.45 [.imol/m^. Figure 9.1) and MFC 18 (3.09 jimol/m^,

Figure 9.2) in the presence of aqueous solutions ofNaCl and NH4CI and Figure 9.3

shows Raman spectra of these phases in water pH-adjusted with conc. HCl and conc.

NH4OH. Raman spectra were only acquired for MFC 18 in 5, 15, and 55 mM aqueous

solutions ofNaCl and NH4CI because of the limited quantity of material available.

Spectra were originally acquired in pH-adjusted water as control experiments for the

365

NH^Cl NaCl

C/3 c <U

2800 2900 3000 3000 2800 2900 Wavenumbers (cm"^) Wavenumbers (cm"')

Figure 9,1 Raman spectra of TFC18SF in (a and f) 5 mM, (b and g) 15 mM, (c and h) 55 mM, (d and i) 100 mM, and (e and j) 200 mM aqueous NH4CI (left panel) and NaCl (right panel). Acquisition times for all spectra are 2 min.

366

NH4CI NaCl

a

b

c

3000 2800 2800 2900 2900 3000 Wavenumbers (cni'^) Wavenumbers (cm'^)

Figure 9 .2 Raman spectra of MFC 18 in (a and d) 5 mM, (b and e) 15 mM, and (c and f) 55 mM aqueous NH4CI (left panel) and NaCl (right panel). Acquisition times for all spectra are 5 min.

367

TFC18SF

!U

MFC 18 vfCH,)

vfCH3U

2800 2900 3000

Wavenumbers (cm"^)

2800 2900 3000

Wavenumbers (cm"^)

Figure 9.3 Raman spectra of TFC18SF (left panel) and MFC18 (right panel) in pH-adjusted water at pH (a and 1) 2.29, (b) 3.93, (c and g) 6.16, (d and h) 7.08, and (e and I) 9.0. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 spectra are 5 min.

368

pH/solute studies presented in the later half of this chapter; however, because of the

equivalent observations made between these experiments and the aqueous salt solution

experiments, they are presented together.

Structure and order information about alkane-based systems can be obtained from

multiple indicators in their Raman spectra, as described in Chapter 3. However, the

discussion of order for t hese phases is restricted to the peak intensity ratio of the VaCCHj)

(2885 cm"') to the VsCCHi) (2850 cm"'), I[Va(CH2)]/I[Vs(CH2)], as this indicator is

extremely sensitive to small changes in conformational order that stem from both true C-

C gauche conformers and subtle rotational disorder in short segments of alkyl chains. As

observed for these materials in neat solvents (Chapters 4, 6-8), in aqueous salt solutions

and pH-adjusted water, Raman spectra in the v(C-H) region are free from spectral

interference.

It is difficult to discern differences in the Raman spectra for these phases in

different aqueous salt solutions; however, the Raman spectra for MFC 18 in aqueous salt

solutions appear to have a more intense Va(CH2) mode relative to the Vs(CH2) intensity,

compared to the Raman spectra of MFC 18 in DDI water, shown in Chapter 6 (Figure

6.1). Subtle differences in the value of I[Va(CH2)]/l[Vs(CH2)] will be used to describe

differences in conformational order of these phases in the presence of different aqueous

solutions.

Figure 9.4 shows a plot of l[Va(CH2)]/l[Vs(CH2)] as a function of salt

concentration (Figure 9.4a) and as a function of pH (Figure 9.4b) for TFC18SF and

1.6 369

1.5-

g 1.4.

S 1.3-

g 1 , 2 -

1 . 1

1.0

1,5-

f f i " 1 4 -o

g 1.3-1

X O

1 . 2 -

1.1

1 . 0 -

ft ^

-

TFC18SF/Water

f a

MFCIS/Water

50 100 150

Salt Concentration (mM)

200

• •

TXFCISSF/Water

"1

•

MFC 18/Water "I ' I ' r

2 3 4 5 6

pH

i

7 T

8 T

9 10

Figure 9.4 I[Vg(CH2)]/I[v,(CH2)] as a function of (a) salt concentration and (b) pH for TFC18SF in aqueous NaCl (V), NH4CI (•) and in pH-adjusted water (•) and for MFC 18 in aqueous NaCl (O), NH4CI (•) and in pH-adjusted water (• ). Error bars represent one standard deviation.

370

MFC 18 in aqueous solutions of NaCl and NI-I4CI (Figure 9.4a) and water pH-adjusted

with HCl and NH4OH (Figure 9.4b). The horizontal lines represent the values of

1 [va(CH2)]/l[Vs(CH2)] for TFC18SF and MFC 18 in DDI water as a reference. The value

of l[Va(CH2)]/l[Vs(CH2)] for TFC18SF increases slightly from 5 mM aqueous NaCl and

NH4CI to 50 mM NaCl and NH4CI, then remains constant with increasing salt

concentration. For MFC18 similar observations are made; however, no data for salt

concentrations greater than 50 mM are presented. In pH-adjusted water, the value of

I[Va(CH2)]/I[Vs(CH2)] for TFC18SF is constant over the pH range 2-9 and is the same for

both materials in aqueous salt solutions > 50 mM. In both aqueous solutions, the value of

I[va(CH2)]/l[Vs(CH2)] is greater than in DDI water alone for both materials. For MFC 18,

the value of I[Va(CH2)]/l[Vs(CH2)] is constant from pH 2-7.5, then decreases significantly

at pH 9.0. Disordering the stationary phase at high pH could be due to a large fraction of

ammonium ions interacting with exposed deprotonated silanols. Over the pH range 2-9

the value of I[Va(CH2)]/I[Vs(CH2)] is greater than that for MFC 18 in DDI water.

These observations indicate that TFC18SF and MFC 18 are more ordered in the

presence of aqueous salt solutions, or solutions of appreciable ionic strength, than in DDI

water and that this ordering effect is slightly concentration dependent. As discussed in

Chapter 6, for TFC18SF and MFC!8 in DDI water, there is evidence for solvation of

these phases by interfacial water molecules, albeit minimal. As described in the

literature, increasing ionic strength or salt concentration decreases the solubility of

hydrophobic alkyl chains in aqueous solution and increases the hydrophobic

371

Thus, interactions between alkyl chains are increased, as they are no

longer solvated, and the conformational order of the alkyl chains increases. While this

effect has not been explicitly documented previously for hydrophobic solute/aqueous

solution systems, this behavior is in good agreement with the previous studies on these

alkylsilane stationary phases presented in Chapter 5, where increasing alkyl chain

interactions by decreasing temperature leads to increasing conformational order.

Eftects of pH of Aqueous Acidic and Basic Aromatic Compounds

Raman spectra of TFC18SF and MFC 18 in 37 mM aqueous benzoic acid,

aqueous benzylamine, TFC18SF in aqueous aniline between pH 2.0 and 9.0 are shown in

Figures 9.5-9.7, respectively. Spectra for MFC 18 in aqueous aniline could not be

acquired because of significant background spectral interference. No observable

differences in the Raman spectra of these phases in aqueous solutions as a function of pH

or with different analytes are observed; however, the value of I[Va(CH2)]/I[Vs(CH2)] can

be used to determine subtle changes in conformational order of these phases in different

solution conditions.

Benzoic Acid

Figure 9.8 shows a plot of I[Va(CH2)]/I[Vs(CH2)] as a function of pH for TFC18SF

(Figure 9.8a) and MFC 18 (Figure 9.8b) in 37 mM aqueous benzoic acid and in pH

adjusted water, as a control. The value of I[Va(CH2)]/I[Vs(CH2)] for both phases in

372

TFC18SF MFC 18

m a <u

2800 2900 3000 2800 2900 3000 Wavenumbers (cm-i) Wavenumbers (cm-')

Figure 9.5 Raman spectra of TFC18SF (left panel) and MFC 18 (right panel) in 37 mM aqueous benzoic acid solutions at pH (a) 2.30, (b) 3.38, (c) 4.23, (d) 7.86, (e) 9.33, (f) 2.30, (g) 3.41, (h) 4.12, (i) 5.25, and (j) 9.00. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 spectra are 5 min.

373

TFC18SF MFC 18

3000 2800 2900 2800 2900 3000

Wavenumbers (cm"') Wavenumbers (cm'' )

Figure 9.6 Raman spectra of TFC18SF (left panel) and MFC 18 (right panel) in 37 mM aqueous benzylamine solutions at pH (a and e) 1.93, (b and f) 4.10, (c and g) 9.40, (d and h) 10.8. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 are 5 min.

374

m G a>

2800 2900 3000 Wavenumbers (cm"' )

Figure 9.7 Raman spectra of TFC18SF in 37 mM aqueous aniline solutions at pH (a) 1.83, (b) 3.86. (c) 4.60, (d) 4.86, (e) 5.95, and (f) 7.95, Acquisition times for all spectra are 2 min.

1.6 375

1.5

ffi"

U

S" 1-3 •

I-c« > ^ 1.1-

1.0-

TFClSSFAVater

2 I '—I——r 3 4 5

"T—I—!—I—r 6 7 8

pH

1.5-

u S" 1-3-r—I

CN ffi l.2-\ O

as

^ i.H

6

MFClS/Water 1 . 0 -

T ! j-

6 7

pH

a -T • 1 r-9 10 11

I T

8 9 10 11

Figure 9.8 I[Vj,(CH2)]/I[Vg(CH2)] as a function of pH for (a) TFC18SF in 37 mM aqueous benzoic acid (•) and pH-adjusted water (•) and (b) MFC 18 in 37 mM aqueous benzoic acid (•) and pH-adjusted water (O). Error bars represent one standard deviation.

376

37 mM aqueous benzoic acid is greater than in DDI water over the pH range examined.

For TFC18SF, I[Va(CH2)]/I[Vs(CH2)] decreases from pH 2.0 to ~ 4.5 and then remains

constant with increasing pH, while I[Va(CH2)]/T[Vs(CH2)] for MFC 18 is constant at all pH

values.

At pH values lower than its pKa, benzoic acid does not impact the conformational

order of TFC18SF. Benzoic acid has an octanol-water partition coefficient greater than

one (log Kow = 1-87), suggesting that it should have significant interaction with the

hydrophobic stationary phase. In addition, benzoic acid is known to be stongly self-

Q «c Q associating m solid and solution phases. ' '' Thus, the combination of strong benzoic

acid interaction with the stationary phase, which could contribute to disorder in the phase,

coupled with benzoic acid self-association, which could promote ordering of the alkyl

chains as described for self-associating compounds in Chapter 7, leads to no net effect on

the conformational order of the phase. As pH increases and is made greater than its pKa,

the self-association ability of benzoic acid decreases due to electrostatic repulsion of

deprotonated benzoic acid molecules and the stationary phase is slightly disordered as a

result of benzoic interaction relative to pH-adjusted water. Benzoic acid interaction with

TFC18SF is likely to occur primarily at the distal ends of the stationary phase because the

high surface coverage restricts significant solute penetration deep between the alkyl

chains.

For MFC18, no pH-dependent interaction between benzoic acid and the stationary

phase is indicated. The conformational order of MFC 18 was unaffected by the presence

377

of benzoic acidrelative to pH-adjusted water at all pH values due to the greater free

volume of MFC 18 relative to TFC18SF that accommodates the benzoic acid without

structural consequences.

Aniline

Figure 9.9 shows a plot of I[Va(CH2)]/![Vs(CH2)J as a function of pH for TFC18SF

in 37 mM aqueous aniline solutions and in pH-adjusted water (control). Within the

scatter of the I[va(CH2)]/I[Vs(CH2)] data, no difference in the I[Va(CH2)]/I[Vs(CH2)] values

is observed in these two systems. This lack of structural effect is proposed to be due to

more favorable aniline interaction with water than with the stationary phase as indicated

by its log Kow of 0.90. Assuming that this log Kow value doe not change significantly

with pH/'^^ one would predict little effect of the presence of aniline on the

conformational order of the stationary phase.

As stated above, spectral data for MFC 18 in aqueous aniline solutions were not

acquired due to severe spectral interference; thus, the effect on anilne on the

conformational order of MFC 18 in aqueous aniline cannot be assessed. This would be an

interesting experiment, considering the effects of neat aniline and pyridine on the

conformational order of MFC 18 as discussed in the previous chapter. Despite log Kow

values less than 1.0 for both compounds, disordering of MFC 18 was observed in the

presence of neat aniline and pyridine and attributed to the strong driving force for acid-

base interactions with exposed surface silanols on this low coverage stationary phase.

378

1.6-1

1.5-

1.4-

u -

tZJ 1.3-

1.2-u

1.2-cs

> 1.1-

1.0-

TFClSSFAVater

^—'—I—^ 12 3 4

-1—,—1—I—1—r

5 6 7 8

pH

9 10 11

Figure 9.9 Uv.CCH^jl/llv/CHj)] as a Amction ofpH for TFC18SF in 37 niM aqueous aniline (A) and pH-adjusted water (•). Error bars represent one standard deviation.

379

While the effect of aqueous bases on the conforaiational order of MFC 18 was not studied

for aniline, it was investigated for benzylamine.

Benzylamine

Figure 9.10 shows plots of I[Va(CH2)]/I[Vs(CH2)] as a function of pH for

TFC18SF (Figure 9.10a) and MFC 18 (Figure 9.10b) in 37 mM aqueous benzylamine and

in pH-adjusted water. The value of I[Va(CH2)]/I[Vs(CH2)] is not pH-dependent for either

TFC18SF or MFC 18 in aqueous benzylamine. However, for both stationary phases,

I [Va(CH2)]/T [Vs(CH2)] is significantly less in aqueous benzylamine solutions than in pH-

adjusted water from pH 2 to 11. These observations contrast considerably with those for

aqueous aniline and benzoic acid solutions, in which these values are not significantly

different than those in pH-adjusted water.

In the presence of benzylamine solutions, TFC18SF is more disordered than in

pH-adjusted water, although the conformational order is independent of pH. This

behavior suggests that benzylamine has appreciable interaction with the stationary phase,

consistent with its octanol-water partition coefficient (log Knv) of 1.09.^'^^ In addition,

benzylamine self-association is weak relative to benzoic acid such that benzylamine

interaction at the distal ends of the alkyl chains provides no driving force for chain

ordering.

MFC 18 is also more disordered in the presence of aqueous benzylamine than in

pH-adjusted water or in aqueous benzoic acid. On MFC 18 a driving force for interaction

1.6 380

S u

1.5

1.4-

1 3H

1.1

1 . 0 -

5

1 • TFCISSF/Water

a •T—I—'—1—'—r ~T—'—1—'—I—1—1—'—T"""—r~^

5 6 7 8 9 10 11

pH

1.5-

O 1-4 • X u 2° 1-3-

cs

m >

l=f 1.1

t . . M .

Q

O

MFCIS/Water

1 . 0 -

pH

Figure 9.10 I[v,(CH2)]/I[v,(CH2)] as a function of pH for (a) TFC18SF in 37 mM aqueous benzylamine (A) and pH-adjusted water (•) and (b) MFC 18 in 37 mM aqueous benzylamine (^) and pH-adjusted water (O). Error bars represent one standard deviation.

381

with the stationary phase exists in addition to the hydrophobic effect exists through acid-

base interactions with exposed surface silanois as described in Chapter 8. Even though

the log Kow for benzylamine (1.09) is less than that for benzoic acid (1.87)/ ^

benzylamine penetrates deeper into the alkyl chains because of favorable interactions

with the silica surface leading to more disorder in the system.

Additional evidence for strong benzylamine interactions with these stationary

phases is presented in Figure 9.11 in a plot of I[Va(CH2)]/I[Vs(CH2)] as a function of log

Kow for TFC18SF and MFC 18 in neat benzylamine and aniline (data from Chapter 8). In

benzylamine, the value of I[va(CH2)]/I[v,(CH2)] for each phase is less than in air or in

aniline. This benzylamine-induced disordering of these phases is attributed to a

combination of a favorable log Kow value and a high pKa of 9 .5, leading to favorable

stationary phase solvation and strong silanophilic interactions.


A detailed examination of the effects of aqueous salt solutions and the effects of

acidic and basic compounds, and pH on the rotational and conformational order of two

stationary phases is presented. Results indicate that rotational and conformational order

of these materials in 5 to 200 mM aqueous salt solutions is increased relative to the order

of these phases in deionized water. These stationary phases are effectively "salted-out"

of aqueous solution causing alkyl chain intermolecular interactions to increase. The

result is more ordered stationary phases in the presence of aqueous salt solutions than in

382

1.6

1.5

1.3

U

i .H

1.0

0.9

Air Aniline

Air

s S

Benzylamine

• •

• •

0.0 0.2 0.4 0.6 0.8 1.0 1.2

LogK ow

Figure 9.11 I[v,(CH2)]/I[v,(CH2)] as a function of log for TFC18SF (•) and MFC 18 (•) in air, neat aniline, and neat benzylamine . Error bars represent one standard deviation.

383

deionized water. In the presence of aqueous acidic and basic compounds, the

conformational order of both stationary phases is largely independent of pH, with the

exception of TFC18SF in aqueous benzoic acid. In this case, at low pH benzoic acid self-

association keeps the molecules from perturbing the conformational order of TFC18SF.

As pH increases, benzoic acid interacts slightly more with the stationary phase, resulting

in a slight pH-depend disordering of TFC18SF.

Considerable work could be done to follow up on these preliminary studies.

These possible future studies include;

1. Examining stationary phase order over a wider range of salt concentrations to see

if high salt concentrations (e.g. 1-5 M) solutions have a similar salting-out effect.

2. Investigating the effect of ion size on this "salting-out." The Hofmeister ion

series shows that ion size affects salting-in and salting-out of hydrophobic solutes

and proteins from aqueous solution. This effect should be observed for these

stationary phases and can be monitored by changes in conformational order.

3. Determining the effect of salt or buffer solutions on the retention of analytes in

chromatographic experiments. Observations of changing conformational order of

two stationary phases in salt solutions have been made, but whether or not this

affects analyte retention is unknown.

4. Studying the effects of a wide range of pH-sensitive analytes on the

conformational order of these stationary phases. Possible analytes include those

with multiple acidic protons or basic sites, more hydrophobic solutes, and

biologically-relevant compounds (nucleotides, amino acids, and peptides).

385

Chapter 10

Solute, Solvent, and Stationary Phase Interactions of Reversed-Phase Liquid

Cliromatographic Systems Determined by Raman Spectroscopy

Introduction

While the basis of separation in reversed-phase liquid chromatography (RPLC) is

generally accepted as a difference in polarity between a series of solutes, the mechanism

of solute retention is an issue that has been debated for nearly 40 years. There is an

increasing awareness in the field of the importance of interfacial phenomena. In fact,

peak tailing of solutes,'" separation quality dependence on alkyl chain length,'" ""'"

solute selectivity,'" and silica substrate cleanliness,'" are all evidence for

interfacial interactions influencing chromatographic performance. The complexity of the

chromatographic interface has been a major impediment to adequately describing

molecular retention, however, as pointed out throughout this Dissertation.

Despite valuable information obtained from numerous reports aimed at

elucidating retention mechanisms in relatively few have studied retention

as it takes place in chromatographic systems. Fluorescence spectroscopy'"'^^"'"'^^ and

vibrational sum frequency generation (vSFG) spectroscopy'" '*""'" "^^ generally have been

carried out using model planar quartz or fused silica substrates. While these may be

386

appropriate chemical models of particulate silica substrates used as RPLC supports, their

macroscale structures are significantly different. In addition, although nuclear magnetic

resonance (NMR) spectroscopy'" '^'"'" '**^' and Fourier-transform infrared (FT-IR)

spectroscopy '" *'^"'"have been used successfully to study the structure and mobility of

alkylsilane-based stationary phases as a function of temperature in air and in solvents, no

reports have appeared in which these techniques have been used to study stationary phase

structure and order in the presence of solvent and solute.

Performing such experiments under more chromatographically-relevant

conditions than those mentioned above should allow a better understanding of retention

mechanisms in RPLC. In previous work, two models, the solvophobic (also referred to

as the adsorption model) and partitioning models, have been proposed to describe the

retention process in RPLC; the details of each were discussed in Chapter 1. It is our hope

that these experiments will offer new information that can be related to these two

descriptors of solute retention.

Goals of this Research

The goal of the work reported here was to use Raman spectroscopy in the v(C-H)

region on two high-density alkylsilane stationary phases, TFC18SF and MFC 18, in the

presence of mobile phase and solute to determine their conformational order, and hence,

the mechanism of solute interaction at the stationary phase/mobile phase interface. From

the work presented in Chapters 6-8, a clear understanding of the interactions between

387

neat solvents and these stationary phases, and how these interactions relate to

conformational order of these phases, have been achieved. Here, this approach is

extended to understanding relevant chromatographic systems with stationaiy phase,

mobile phase and solute components. The Raman spectral data are correlated to the

chromatographic behavior of the solutes to provide insight into the mechanism of

retention.

Methodology

The synthesis and properties of TFC18SF and MFC 18 are described by Sander

and Wise'"''''^'" '' and are summarized in Chapter 2. Toluene (100%), anisole (99%),

aniline (99%), and DjO (99.9%) were purchased from Aldrich Chemical Co.

Perdeuterated methanol-d-i (99.8%) and acetonitrile-ds (99.8) were purchased from

Cambridge Isotopes. All chemicals were used as received.

Samples were prepared by adding 200 |.iL of analyte solution to between 25 and

50 mg of stationary phase in a 5.0-mm dia NMR tube. Samples were sonicated for 10

min and equilibrated at 20 "C for a minimum of 48 h prior to spectral analysis. Raman

spectra were collected using 100 mW of 532 nni radiation from a Coherent Verdi

Nd: YVO4 laser on a Spex Triplemate spectograph as described in Chapter 2. Slit settings

of the Triplemate were 0.5/7.0/0.150 mm for all experiments, corresponding to a spectral

bandpass of 5.3 cm"' at 2900 cm"'. Raw spectral data in the v(C-H) region were first

corrected for background using a 5-point linear fit of the data in the region between 2750

388

and 3050 cm"'. Peak intensities of the VaCCHs) and VsCCHa) bands were determined as

peak heights from the baseline of the corrected spectra. Total acquisition times for each

spectrum are given in the figure captions.

Chromatographic experiments presented in this chapter were carried out by Dr.

Lane Sander at NIST using a liquid chromatograph with a reciprocating piston pump,

autosampler, and ultraviolet absorbance detector (lex = 254 nm). Columns were slurry

packed at 9000 psi using pentane, as previously described." Solute concentrations

used are ~ 1 |iL/mL in methanol with an injection volume of 5 îL,

Pictures of the solvent/stationary phase interface were constructed using

ChemDraw 3D (CambridgeSoft) following the protocol described in Chapter 2.

Thermodynamic Description of Retention in RPLC

Ranatunga and Carr'"'"^^ state that favorable free energy of retention, AGRETN,

results primarily from favorable interaction of a solute molecule with the stationary

phase, AGSTAT, and not from favorable interactions with the mobile phase, AGMOBILE-

They are careful to point out, however, that the greater dependence on favorable

interactions with the stationary phase is only slightly more important than those with the

mobile phase, indicating that the free energies in both phases play crucial roles in

retention.

This view of retention free energy takes into account the sum of two processes in

a standard thermodynamic cycle (Scheme 1): solute transfer from the mobile phase to an

389

ideal gas phase followed by solute transfer from the ideal gas phase to the stationary

phase. Incorporating the ideal gas phase concept as an intermediate between the two

condensed phase states allows solution free energy values, defined as the free energy to

transfer a solute from the ideal gas phase to a given liquid phase, AGs (v-->Iiq),'" "' "' ~'' to

be used for both AGSTAT and AGMOBILE- Thus, the free energy of retention, AGRETN, is

the sum of these two free energy contributions as shown in Scheme 1:

SolutecAs

-AGmobiu^^^^ \^^GSTAT

SoluteMORIT F. —— • SolutesTAT

AGRETN = E steps - - AGMOBILE + AGSTAT

Scheme 1

AGMOBILE is of the same magnitude as the solution free energy for a given solute in the

mobile phase solvent or mixture of solvents, but is opposite in sign since the solute is

removed from the mobile phase upon transfer to the gas phase. AGSTAT for a typical CIG

stationary phase is approximated using the solution free energy value for the solute in n-

hexadecane.'" "" Although this approximation does not account for entropic

390

considerations of the surface-confined alkylsilanes, it is a reasonable approximation of

bulk solvation properties.

These solution free energies can be used to predict AGRETN values for systems of

interest, and were used to not only describe the retention behavior of these solutes

(capacity factor, k\ and relative retention, a), but to rationalize structural changes in

alkylsilane stationary phases with the retention of different solutes. Free energy values

from Abraham and coworkers'"^'^'"'^^' were used to calculate AGRETN, AGSTAT, and

AGMOBILE values and are tabulated in Tables 10.1 and 10.2. Solution free energy data for

solutes in acetonitrile are not readily available in the literature and are not presented in

the discussion of these data. Free energy values for mixed solvent systems are not

available in the literature, but are approximated as a weighted average of the sum of the

components in the mixture.

Kinetics and Band Broadening in RPLC

Band broadening in BPLC is a result of finite mass transport processes of solutes

as they travel down the column. These kinetic constraints are reflected in the separation

efficiency, quantified as the number of theoretical plates, N, for a given column and the

height equivalent to a theoretical plate, H. The magnitude of H is dependent on A, the

longitudinal ditTusion coefficient of a solute, B, the multiple path coefficient for a solute,

C, the mass transport coefficient for the solute in the stationary phase and in the mobile

39!

Table 10.1 Solution tree energies for solutes in methanol, water, and hexadecane

Solute Solvent Solution Free Energy, AGs (v ->liq) (kcal/mol)®

Toluene

Anisole

Aniline

Water

Methanol

Hexadecane

Water

Methanol

Hexadecane

Water

Methanol

Hexadecane

3.48

- 1 . 8 8

-1.94

1.82

-2.44

• 2.74

-1.23

-3.25

-2,83

'Values taken from references 10.53 and 10.54.

392

Table 10.2 Predicted AGRETN values from solution free energy values

Solute Mobile Phase AGSTAT -AGMOBILE Predicted [AGs (v^HD)]" [-AGS (v- ->MP)]'' AGRETN

Toluene 100% Methanol

75:25 Methanol/ Water

50;50 Methanol/ Water

Anisole 100% Methanol



Aniline 100% Methanol



-1.94

-1.94

-1,94

-2.74

-2.74

•••-2.74

2.83

-2.83

-2.83

+1.88

+ 1.32

-0.80

+2.44

+1.38

+0.31

+3.25

+2.75

2.24

-0.06

-0.62

-2.74

-0.30

-1.36

-2.43

+0.42

-0 .08

-0.59

"HD = hexadecane = mobile phase

phase, and p., the linear velocity of the mobile phase, described by the van Deemter

equation;

393

H — A + B/|i + (Cs + Cm)|.t (1)

where

Uk')dl

In these expressions, CsM' and Cm|i are both functions of solute capacity factor, fs(k') and

fm(k') and slower mass transport kinetics or larger values of H can arise from greater

stationary phase film thickness, df, slower mass transport of the solutes in either the

stationary phase, Dg, or mobile phase. Dm, or larger stationary phase particle diameter, dp.

Chromatographic data presented here are interpreted in terms of these kinetic parameters

in addition to the thermodynamic models discussed above.

SeparatioD of Monosubstituted Aromatic Compounds using MFC18

In Chapters 4-8 of this Dissertation, only solvent effects on the conformational

order of alkylsilane stationary phases have been considered. The spectral data have been

interpreted with conventional retention models, and molecular pictures of the

394

chromatographic interface have been presented to aid in the explanation of the Raman

spectral data. However, the one outstanding question yet to be addressed is whether the

Raman spectral data presented in this work correlate with the chromatographic behavior

of these solutes. While using the spectroscopic data to come up with a molecular picture

of the chromatographic interface is quite informative, correlating these data to real

chromatography experiments is essential for elucidating retention mechanisms for these

systems.

Figure 10.1 shows overlays of chromatograms of toluene, anisole, and aniline

separated in a mobile phase of 100% methanol using columns of MFC 18, a stationary

phase comparable in surface coverage (3.09 jimol/m^) to most conventional,

commercially-available RPLC phases. As expected in this mobile phase, these solutes

are poorly retained on this column with capacity factor, k\ values < 0.20. Despite this

relatively poor retention, these solutes were used in this series of experiments, because

they had been used as neat solvents in previous experiments (Chapters 6 and 8). The

unique ability of these compounds to be used as neat solvents and as dilute solutes in

common mobile phase solvents facilitates the comparison of their interactions with Cig

stationary phases as solvents and solutes.

The thermodynamic retention model ofRanatunga and Carr'*' "''^ allows prediction

of the retention order of these solutes. For aniline, AGSTAT [AGs(An-HD)] is -2.83

kJ/mol and -AGMOBOJ; is +3.25 kcal/mol,'""''''*^^'' resulting in a small, but

thermodynamically unfavorable predicted AGKETN value of +0.42 kcal/mol. Thus, in a

395

o

'm

1

Time (min)

Figure 10.1 Overlaid chromatograms showing the separation of toluene, anisole and aniline on MFC18. Experiments were done in a 100% methanol mobile phase with a constant flow rate of 1.0 mL/min.

396

100% methanol mobile phase, aniline is predicted to be only weakly retained, consistent

with the observation of aniline elution just after the void volume in the actual

chromatographic experiment on this stationary phase (Figure 10.1).

For anisole, AGSTAT is -2.74 kcal/mol and -AGMOBILE is +2.44

kcal/mol. '"^^ '"^'^hus, AGRETN is predicted to be -0.30 kcal/mol for this system.

Although both interactions are favorable, the interaction of anisole with the stationary

phase is predicted to be greater. This prediction is confirmed by the increased retention

of anisole relative to aniline. Similarly, for toluene, the predicted AGRETN is -0 .06

kcal/mol and interactions of toluene with the stationary phase are slightly more favorable

than those with the methanol mobile phase

A discrepancy is noted, however, between the predicted AGRETN values for anisole

and toluene and their retention order; toluene is retained slightly longer than anisole

despite a slightly less favorable predicted AGRETN- The failure of this thermodynamic

retention model is most likely due to modeling of the stationary phase using liquid n-

hexadecane, which does not adequately represent the free volume restrictions that exist

for surface-confined alkyl systems. It is reasonable that anisole interactions with the

stationary phase are entropically less favorable than for toluene, because the methoxy

substituent has a larger volume than the methyl group (V.OCH3 ~ 0.030 NM'; V-CID ~ 0.022

nm').'" " Moreover, this larger methoxy substituent is oriented out of the plane of the

phenyl ring,'"^'^""'"^" rendering the effective molar volume of anisole (V,„ ~ 0.120 nm^)

almost 20% larger than that of toluene (V,n ~ 0.102 nm^).'""''^ Furthermore, anisole

397

solvation in methanol (AGs(Anisole-MeOH) = - 2.44 kcaJ/mo!)" ' is more favorable

than toluene solvation in methanol (AGs(Toluene-MeOH) = -1.88 kcal/mol)'' farther

decreasing the retention of anisole. Thus, shortcomings associated with hexadecane as

the stationary phase model rationalize greater toluene interaction with the stationary

phase and longer retention than anisole.

Retention for these solutes increases with increasing aqueous component of the

mobile phase, as shown in chromatograms in Figure 10.2 for aniline. Figure 10.3 for

anisole. Figure 10.4 for toluene, and tabulated capacity factors for these analytes in all

three mobile phases in Table 10.3. This trend follows the same thermodynamic

description of retention discussed previously. For each of the solutes, the predicted

AGRETN values become more negative with increasing amount of water in the mobile

phase and retention becomes more favorable. For aniline, AGRETN is predicted to only

decrease from +0.42 kcal/mol in 100% methanol to -0.59 kcal/mol in 50:50

methanol/water, because aniline is quite soluble in both methanol and water. As a

consequence, aniline retention increases slightly, but W values in all mobile phases are

<1.0. For anisole and toluene, AGRETN is predicted to decrease significantly from -0.30

kcal/mol in methanol to - 2.43 kcal/mol in 50:50 methanol/water for anisole and from

-0.06 kcal/mol in methanol to - 2.74 kcal/mol in 50:50 methanol/water for toluene.

Retention increases significantly for both solutes as a result. Toluene retention is

predicted to increase more than anisole retention with increasing water in the mobile

phase as toluene is far less soluble in water (AGs(Tol-Water) = +3.48 kcal/mol; AGs(An-

398

Aniline b

tm

i

Aniline ^

1

tm

I

c

0 5 10 15

Time (min)

Figure 10.2 Chromatograms showing increased aniline retention on MFC 18 with increasing aqueous component of the mobile phase from (a) 100% methanol to fb) 75:25 methanol/water to (c) 50:50 methanol/water.

399

i

1 Anisole C

Anisole 3-

Anisole b

' m

i

0 5 10 15

Time (min)

Figure 10.3 Chromatograms showing increased anisole retention on MFC18 with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water.

400

Toluene

X:, 5™, O QO

tm

i

Toluene b

Toluene

0 5 10 15

Time (min)

Figure 10.4 Chromatograms showing increased toluene retention on MFC 18 with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water.

401

Table 10.3. Capacity factor, k\ and relative retention, a, values for solutes on MFC]8

le a

Mobile Phase Aniline Anisole Toluene 0^;\n.To ^'As.To

100% 0.045 0.127 0.174 2.82 3.87 1.37 methanol

75:25 0.217 0.624 1.056 2.88 4.87 1.69 methanol/water

50;50 0.721 3.076 6.453 4.27 8.95 2.10 methanol/water

402

Water) = +1.82 kcal/mol). As observed in Table 10.3, A' tor anisole increases from 0.13

to 3.0 and A' for toluene increases from 0.17 to 6.5 as the water component of the mobile

phase increases from 0 to 50%.

In general, the thermodynamic retention model is a good predictor of retention

order of these monosubstituted aromatics. Despite failure to predict the elution order of

anisole and toluene under conditions of very weak retention, because of neglected

entropic differences between a surface-confined alkyl-based stationary phase and the

hexadecane model, the thermodynamic retention model is generally useful for predicting

elution order of these three solutes.

Of equal importance to the thermodynamic description of retention for these

solutes, however, is the mass transport kinetics of these solutes, which contribute to the

peak widths of these chromatographic bands. Table 10.4 gives values for the number of

theoretical plates (N) and the plate height (H) for the retention of aniline, anisole, and

toluene on MFC 18 in all three mobile phases. In order to eliminate redundancies and

simplify the discussion, only H values will be discussed in detail. For aniline in 100%

methanol, H increases with greater solute retention, F in contrast to the behavior

observed for anisole and toluene. This difference suggests slower mass transfer for

aniline into and out of the stationary phase than for anisole or toluene. Slower mass

transport of aniline could be due to a number of factors including a smaller diffusion

coefficient in the stationary phase (Ds) or in the mobile phase (Dm) or a larger effective

stationary phase thickness (dr) than for toluene or anisole since H is proportional to both

Table 10.4. Numbers of theoretical plates (N) and plate heights (H) for solutes on MFC 18

N H (10- m)

Mobile Phase Aniline Anisole Toluene Aniline Anisole Toluene

100% 5088 5832 6446 1.96 1.70 1.55 methanol

75:25 2605 6538 8036 3.84 1.53 1.24 methanol/water

50:50 745 8218 7443 13.4 1.22 1.35 methanol/water

404

d»/Ds and 1/D„,. The significant peak tailing observed for aniline suggests substantial

hydrogen-bonding interactions with exposed surface silanols,'"^"'" "^ which would

contribute to a decreased value od Ds and an increased value of df for aniline relative to

anisole and toluene. In other words, the diffusion of aniline within the stationary phase is

slow because of strong interactions with the silica support surface, and the effective

stationary phase thickness is maximal as aniline must penetrat the entire depth of the

alkyl chains in order to interact with surface silanols. Additional evidence for slower

aniline transport kinetics in more polar mobile phases is evident in the chromatograms of

these solutes. Figures 10.2-10.4, where the aniline peaks are wider (0.22 and 0.66 min)

than those for anisole (0.16 and 0.35 min) and toluene peaks (0 18 and 0.51 min) in 75:25

and 50:50 methanol/water mobile phases, respectively. As aniline retention becomes

more thermodynamically favorable, opportunity for silanol interaction is farther

promoted and contributes to the slower kinetics of aniline in more polar mobile phases.

For anisole and toluene, as retention increases with increased mobile phase

polarity, mass transport time within the stationary phase becomes faster. While these two

solutes spend more time on average in the stationary phase, giving rise to longer retention

times, anisole and toluene do not have a strong affinity for silanol groups and the

mobility of these solutes in this liquid-like stationary phase is greater than for aniline.

Thus, the frequency of transport into and out of the stationary phase for anisole and

toluene is greater with increased polarity of the mobile phase resulting in longer retention

times and narrower peak widths. The differences in H values for toluene and anisole with

405

varying mobile phase polarity are likely due to solute diffusion in the stationary phase

because the addition of water is likely to have the same effect on anisole and toluene

diffusion in the mobile phase.

Such interaction subtleties are well known to affect peak shape in chromatograms,

but their effect on the stmcture of the stationary phase has not been considered

previously. Both kinetic and thermodynamic factors that influence the retention behavior

of these solutes will be considered to rationalize the structure of the alkylsilane stationary

phase under chromatographic conditions,

Raman Spectroscopv of MFC 18 in the Presence of Solute and Mobile Phase

Up to this point in the discussion, only retention behavior and thermodynamic

descriptions of retention have been considered for this series of solutes. In order to

ascertain structure-function relationships for these stationary phases, the effects of these

solutes and mobile phases on the structure and conformational order of these alkylsilane

stationary phases must be correlated with the retention behavior discussed in the previous

section.

Raman spectra in the v(C-H) stretching region from 2750 to 3050 cm"^ were

acquired for MFC 18 at 20 "C in methanol/water solutions containing aniline, anisole, and

toluene. These are shown in Figure 10.5, Rotational and conformational order of these

phases was assessed using the intensity ratio of the VaCCHs) to the VsCCHz),

I[Va(CH2)]/I[Vs(CH2)], as described in Chapters 3-9. For these experiments, a

Aniline Anisole Toluene

d a

g e b

f c

—T— 3000 2900 3000 2800 2900 2800 3000 2800 2900

Wavenumbers (cnr') Wavenumbers (cnr') Wavenumbers (cm'^)

Figure 10.5 Raman spectra of MFC 18 in the v(C-H) region in 5 mM aniline (left panel), 5 mM anisole (center panel), and 5 mM toluene (right panel) solutions in (a, d, and g) 100% methanol, (b, e, and h) 75:25 methanol/water, and (c, f, and i) 50:50 methanol/water. Integration times for all spectra are 5 min.

O Os

407

concentration of 5 mM was used for all solutes. In the chromatographic experiments,

concentrations of- 1 pL/mL were used, corresponding to a solute concentration within a

chromatographic peak of between lO"'^ and 10'^ M. While 5 mM solutions are the upper

limit of chromatographic solute concentrations, they are a reasonable analogue to the

chromatographic experiment.

In the discussion of these data, it is important to remember that the values of

I[Va(CH2)]/I[Vs(CH2)] for MFC 18 are determined with interference to the Va(CH2) (-2885

cm"') mode from the v(CH3)si (-2900 cm ") associated with the proximal end of the

silane. As a result, direct comparison of I[Va(CH2)]/I[Vs(CH2)] values between

monofunctional and trifunctional phases is problematic. However, comparison of values

between different environments of a monofunctional phase system is appropriate, and

general trends between mono- and trifunctional phases can be made with this important

caveat in mind.

Qualitatively, the effect of these solute/mobile phase systems on MFC 18 is very

small. However, subtle differences in the value of I[Va(CH2)]/I[Vs(CH2)] lend insight into

the relevant interactions at the chromatographic interface. Figure 10.6 shows a plot of

I[Va(CH2)]/I[vs(CH2)] versus chromatographic capacity factor, k\ for MFC 18 in 5 mM

toluene, anisole, and aniline in 100% methanol. The horizontal lines represent the values

of I[Va(CH2)]/I[Vs(CH2)] for MFC 18 in the presence of neat methanol, toluene, aniline,

and anisole The value of I[Va(CH2)]/I[Vs(CH2)] for MFC 18 in 5 mM aniline/'methanol is

approximately the same as in neat methanol and neat aniline. However, in 5 mM

408

U

o

1 . 7

1.6

1.5

1.4'

1.3

1.2

1.1

1.0

0.9'

0.8 0

-

-Anisole ^ 5 Toluene

MFC 18/Anisole

Aniline „ MFC 18/Aniline

" 1 —' 1 1—

MF C18/Methanol

MFCIS/Toluene • r « 1

.00 0.05 0.10 0.15

Capacity Factor, A*

0.20

Figure 10.6 I[v,,(CH2)]/ l[v^(CH2)] as a function of solute capacity factor for toluene, anisole, and aniline in 100 % methanol on MFC 18 (•) columns.Dashed lines show values for MFC 18 in neat solvent environments.

409

anisole/methanol, the value of I[Va(CH2)]/I[Vs(CH2)] for MFC18 is greater than in neat

methanol and slightly greater than in neat anisole. In 5 mM toluene/methanol, the value

is significantly greater than in neat toluene or in neat methanol.

The presence of 5 mM aniline in methanol has very little effect on the

conformational order of the alkylsilane stationary phase. Aniline does interact with the

stationary phase (k' is nonzero); however, the interaction is weak as indicated by the poor

retention behavior (F = 0.045) and positive predicted AGRETN value (+0.42 kcal/mol).

When aniline does interact with the stationary phase, it interacts primarily with exposed

surface silanols through hydrogen-bonding, indicated by the peak tailing, slow mass

transport kinetics (Dg), and a large effective stationary phase thickness (df). Thus, when

aniline does partition into the stationary phases, which is a relatively rare occurrence (k'

is small), it penetrates through the entire depth of the stationary phase down to the silica

surface and only slightly alters the stmcture of the stationary phase such that it remains

generally indistinguishable from solvation by neat methanol or neat aniline.

The conformational order of MFC 18 is significantly greater in 5 mM

anisole/methanol than in neat methanol or in neat anisole. This observation is consistent

with previous explanations of solute partitioning that implied greater order of the

stationary' phase in the presence of solute. The Raman spectral data indicate that

even a low coverage stationary phase is ordered in the presence this solute/mobile phase

combination despite weak retention of the solute. In Chapter 8, the ordering of this phase

in neat anisole was discussed and attributed to anisole interaction with the alkyl chains

410

coupled with strong anisole self-association at the solvent/stationary phase interface. For

5 mM anisole in methanol, the conformational order of MFC 18 is even greater than in

neat anisole despite the much smaller anisole number density; thus a different mechanism

of stationary phase ordering than anisole self-association must be operative.

To rationalize this ordering, contributions from both sides of the chromatographic

interface must be considered. For anisole, AGSTAT [AGS(AS-HD)] is -2.74 kcal/mol and

(AGMOBEJE [-AGs(As-MeOH)] is +2.44 kcal/mol, resulting in a small, but favorable,

predicted AGRETN of-0.30 kcal/mol. These nearly identical free energies of solvation

indicate that anisole is similarly solvated by both phases, although it is slightly better

solvated by the stationary phase. Given the approximately equal solvation of anisole by

both phases, the average position for anisole during a chromatographic experiment is

proposed to be in the interfaciai region between the stationary and mobile phases where it

can interact with both the distal end of the alkylsilanes and the mobile phase. This

proposal is consistent with the relatively fast mass transfer kinetics of anisole into and out

of the stationary phase. In this position, interfaciai anisole molecules could hydrogen-

bond with methanol increasing conformational and rotational order of the MFC 18 alkyl

chains. This molecular picture is consistent with a smaller effective stationary phase film

thickness for anisole, which contributes to smaller H values for anisole relative to aniline.

Surprisingly, MFC 18 is as well ordered in 5 mM toluene/methanol as in 5 mM

anisole/methanol despite the fact that MFC 18 in neat toluene is one of the most

disordered stationary phase/solvent systems examined in these studies. As was observed

4 1 1

for anisole, partitioning and retention of toluene increase stationary phase order. Using

the thermodynamics of the chromatographic interface and the kinetics of toluene mass

transfer, this ordering can be rationalized. For toluene, the free energies of solvation in

hexadecane and methanol are -1.94 and -1.88 kcal/mol, respectively, resulting in a

predicted AGRETN of 0.06 kcal/mol. Although this represents a slight driving force for

toluene interaction with the stationary phase (i.e. AGRETN is negative), toluene is well

solvated by interfacial methanol. Thus, the average position of toluene during a

separation is likely to be at the interface where it can interact with the distal end of the

alkyl chains and interfacial solvent molecules simultaneously. This molecular

arrangement must promote unique methanol self-association that supports increased

conformational order in the stationary phase. This picture is also consistent with the

relatively rapid mass transport kinetics indicated by small H values for toluene.

The chromatography results indicate almost no interaction of any solute with the

stationary phase based on the values of k\ although some interaction is indicated in the

kinetic parameters N and H. Nonetheless, Raman spectral data clearly show evidence of

solute/stationary phase interaction highlighting the sensitivity of Raman spectroscopy in

detecting subtle changes in stationary phase environment and the utility of this technique

in providing insight into retention mechanisms in RPLC.

One question that remains, however, is how the conformational order of these

materials changes in the presence of solutes that are more significantly retained on this

412

column. This question can be addressed through the use of mobile phases that are more

polar, thereby resulting in longer retention of these relatively nonpolar solutes.

Aniline Interactions

Chromatograms are shown for aniline elution from the MFC 18 column in Figure

10.2 at 100% methanol, 75:25 methanol/watcr, and 50:50 methanol/water. .Aniline

retention increases with mobile phase polarity, but only very slightly. Even with a 50:50

methanol/water mobile phase, the retention time of aniline the MFC 18 column is only 1.5

min for a K of 0.72. This chromatographic behavior is consistent with predictions of the

thermodynamic model for aniline retention, that predicts an unfavorable AGRETN of+0.42

kcal/mol for a 100% methanol mobile phase and becoming only slightly more

energetically favorable in 50:50 methanol/water (AGRETN of -0.59 kcal/mol). The peak

widths and peak tailing of aniline increase significantly with increasing mobile phase

polarity, resulting in decreasing values of N and increasing plate height H. This trend

suggests that increased aniline retention also increases aniline-silanol interactions,

leading to slower mass transport kinetics.

Just as aniline retention remains relatively unchanged by mobile phase polarity,

the Raman spectra of MFC 18 are unchanged by mobile phase polarity in solutions of

aniline (Figure 10.5). Figure 10.7 shows a plot of I[Va(CH2)]/I[Vs(CH2)] and JC as a

function of water composition in the mobile phase with 5 mM aniline for MFC 18. The

value of 1 [va(CH2)]/I[VS(CH2) ] is unatYected by mobile phase polarity for aniline solutions

413

a u

u

o

p o

Tl p

&

10 20 30 40 % Water in Methanol [v/v j

Figure 10.7 I[Vj,(CH2)]/ I[v/CH2)] (A) and capacity factor, k\ (•) as a function of percent water in methanol for MFC18 in aniline/methanol/water solutions. The horizontal line represents the value of [[v^XCH^)]/ I[v,(CH2)] for MFC18 in 100% methanol.

indicating that aniline interactions with the stationary phase do not change significantly.

More favorable aniline retention and increased peak tailing suggest aniline continues to

penetrate to the silica surface and a greater traction of its time is spent hydrogen-bonded

to surface silanols. Thus, the conformational order of MFC 18 is unchanged in mobile

phases of increased polarity, as aniline interactions with the stationary phase and silica

surface are relatively unchanged.

Anisole Interactions

Figure 10.3 shows chromatograms indicating increased anisole retention with

increasing water in the mobile phase on MFC 18. As discussed above, increased anisole

retention with increasing polarity of the mobile phase is predicted by the thermodynamic

retention model. However, the effects of mobile phase polarity and increased anisole

retention on the conformational order of MFC 18 were unexpected. Figure 10.8 shows a

plot of I[VA(CH2)]/I[VS(CH2)] and capacity factor as a function of water composition in

the mobile phase for MFC 18 in 5 mM anisole. l[Va(CH2)]/I[Vs(CH2)] decreases

significantly from 100% methanol to 75:25 methanol/water, while anisole retention only

increases very slightly. Additional water in the mobile phase increases the retention of

anisole significantly, but the value of 11 Va(CH2)]/f [VsCCHj)] remains constant from 75:25

methanol/water to 50:50 methanol/water.

The difference in conformational order of MFC 18 between anisole solutions of

100% methanol and 75:25 methanol/water is attributed to more anisole partitioning into

415

1 7-

1.6-

1.5-

1.4-u -

i2LI 1.3--

1.2-V

X -

u 1.1 --

1.0-

0.9-

0 8 -

o

iTl 03 a o H-S VJ

10 20 30 40

% Water in Methanol [v/v]

Figure 10.8 I[Vj,(CH2)]/ [[v^jfCHj)] (A) and capacity factor, k\ (•) as a function of percent water in methanol for MFC 18 in anisole/methanol/water solutions. The horizontal line represents the value of ^v/CHj)]/ I[v^(CH2)] for MFC 18 in neat methanol.

416

the stationary phase in the latter mobile phase. At 100% methanol, anisole partitioning

into the alkyl chains of MFC18 is energetically unfavorable because of the large free

volume in light of the relatively small free energy required retention (-0.30 kcaJ/mol).

The predicted value of AGRETN is-1.36 kcal/'mol in 75:25 methanol/water, more

favorable than in 100% methanol, allowing the energy barrier to anisole partitioning to be

overcome. Thus, anisole partitioning into the stationary phase disrupts the order of the

system by inducing a small number of gauche defects as it penetrates deeper between the

alkyl chains. In addition, the decrease in H with increasing water content in the mobile

phase indicates faster mass transport of anisole in the stationary phase, probably due to

the fact that it is less restricted to the interfacial region as it is in 100%o methanol,

allowing it to sample more of the liquid-like alkyl region of the stationary phase in the

mixed methanol/water mobile phases. The increase in disorder of the stationary phase

with increasing water content of the mobile phase is consistent with this picture.

Toluene Interactions

Toluene retention increases with increasing polarity of the mobile phase in RPLC.

This is shown in Figure 10.4 in chromatograms of toluene acquired with mobile phases of

100% methanol, 75:25 methanol/water, and 50:50 methanol/water on the MFC 18

column. Figure 10.9 shows a plot of 1 [Va(CHj)]/![Vs(CH2)] and capacity factor as a

function of water composition in the mobile phase for MFC 18 in 5 mM toluene.

417

f f i 1 4

0 10 20 30 40


O

§

T1 p a o \J

Figure 10.9 I[v^(CB[2)]/ Ifv/CHj)] (A) and capacity factor, k\ (•) as a function of percent water in methanol for MFC! 8 in toluene/methanol/water solutions. The horizontal line represents the value of Il vJCHj)]/ ILv/CHj)] for MFC 18 in 100% methanol.

418

Adding water to the mobile phase decreases the value of I[Va(CH2)]/I[Vs(CH2)]

slightly for MFC 18 relative to in 100% methanol, but it remains significantly greater than

the value of 1[Va(CH2)]/1 [VsCCHi)] in neat methanol neat toluene or in either 75:25

methanol/water or 50:50 methanol/water. This behavior is considerably different from

that of the other two solutes that both induce more significant disordering of the

stationary phase in the more polar mobile phases. The predicted values of AGRK TK

become increasingly more favorable upon increasing water composition of the mobile

phase: 0,62 kcal/mol in 75:25 methanol/water and - 2.74 kcal/mol in 50:50

methanol/water. This trend is due to the limited solubility of toluene in water [AGs (Tol-

Water) = +3.48 kcal/mol]. In addition, toluene interaction with the stationary phase

clearly increases with increasing water content, since F increases to 1.1 in 75:25

methanol/water and to 6.4 in 50:50 methanol/water. Increased toluene retention is

accompanied by a decrease in H, suggesting that the mass transfer kinetics for toluene in

MFC 18 in 75:25 and 50:50 methanol/water are faster than in 100% methanol. In the

more polar mobile phases, the stationary phase is slightly more disordered (i.e. more

liquid-like) than in 100% methanol and the mobility of toluene is likely to be greater in

the more liquid-like stationary phase. The result is faster mass transport of toluene into

and out of the stationary phase in more polar mobile phases. Increased toluene

partitioning into the stationary phase is responsible for subtle rotational disordering of

MFC 18 with increasing water in the mobile phase. It is also possible that the addition of

water to the mobile phase disrupts the ordered methanol solvent structure at the

4 1 9

alkylsilane-mobile phase interface, contributing to a slight decrease in rotational order of

the phase.

In summary, the retention of aniline, anisole, and toluene on MFC 18 generally

follows the predictions made using a strictly thermodynamic model of retention.

However, free volume restrictions for surface-bound alkylsilanes exist that are not

considered in the thermodynamic model. These are especially important in rationalizing

anisole retention in pure methanol. Both thermodynamic and kinetic contributions to

retention are necessary to Mly describe differences in conformational order of MFC 18 in

each solute/mobile phase system. The kinetic contributions, in particular, provide insight

into how these solutes are retained on MFC18, From the Raman spectral data, molecular

pictures of the chromatographic interface can be developed to illustrate interactions of

solutes and mobile phases with the stationary phase and their effect on conformational

order of the phase.

Molecular Pictures of the MFC 18 Chromatographic Interface

Figure 10.10 shows a picture of MFC 18 in 100% methanol mobile phase. In

methanol, MFC 18 is more disordered than in air, with a larger fraction of gauche

conforniers and greater rotational disorder in the distal ends of the alkyl chains. While

methanol solvation of MFC 18 does induce disorder within the stationary phase, it is not

as disordering as solvents like toluene and chloroform, described in Chapter 6. Based on

the value of I[va(CH2)]/I[Vs(CH2)] for MFC 18 in more polar mobile phases of mixtures of

f

t %

Figure 10.10 Molecular picture of the methanol mobile phase-M FC 18 interface.

421

methanol and water, the picture of the interface is believed to be very similar, where

methanol solvates the alkyl chains at the interface inducing the same degree of disorder

and water molecules primarily remain in the bulk mobile phase.

Figure 10.11 shows a picture of the aniline/MFC 18 interface in neat aniline

(Figure 10.1 la) and in 5 mM aniline in methanol (Figure 10.1 lb). In neat aniline, the

conformational order of MFC 18 is the same as in 5 mM aniline in methanol based on the

value of I [ Va(CH2)]/![ VS(CH2)] for both systems. In neat aniline, molecules have

appreciable hydrophobic interaction with the stationary phase as well as hydrogen-

bonding interactions with exposed surface silanols, both contributing to the disorder of

the system. In 5 mM aniline in methanol, however, the frequency of aniline interactions

with the stationary phase is small (A' is small), but when aniline does interact with the

stationary phase, its interactions are isolated primarily at the surface silanols (large H

values) due to excellent aniline solvation by methanol. The result is the possibility for

two mechanisms of aniline interaction with the stationary phase, one in which aniline

interacts only weakly with the stationary phase at the outer edge because of favorable

solubility in methanol and the other in which anisole penetrates the stationary phase to

exposed surface silanols. In both cases, the conformational order of MFC18 is the same

and indistinguishable from that in neat methanol. With the addition of water to the

mobile phase, the conformational order of MFC18 is unchanged, but the frequency of

aniline interactions with surface silanols increases (H increases and AGRETN becomes

more favorable), but the picture of aniline interaction with the stationary phase in 75:25

aniline

aniline

Figure 10.11 Molecular pictures of the aniline/MFCIS interface for (a) MFC18 in neat anilne and (b) MFC18 ^ in 5 mM aniline in 100% methanol. to

423

and 50; 50 methanol/water is predicted to be much the same as that for aniline in 100%

methanol as shown in Figure 10,1 lb.

Figure 10.12 shows molecular pictures of the anisole/MFClS interface for neat

anisole (Figure 10.12a) and for 5 mM anisole in methanol (Figure 10.12b). In neat

anisole, MFC 18 is more ordered than in air (Chapter 8), attributed to appreciable anisole

self-association at the interface. Surprisingly, in 5 mM anisole in methanol, MFC 18 is

more ordered than in neat anisole, suggesting that anisole partitions only part way into

the stationary' phase providing an environment that allows both for methanol solvation

and stationary phase interaction. InterfaciaJ methanol molecules laterally order at the

interface to form the ordered solvent/analyte/stationary phase structure shown in Figure

10.12b. This molecular picture is flirther supported by the relatively fast mass transport

kinetics of anisole into and out of the stationary phase indicated by the small H values

observed

Molecular pictures of the toluene/MFC 18 interface are shown in Figure 10.13 for

Ml'C18 in neat toluene (Figure 10.13a) and in 5 mM toluene in methanol (Figure

10.13b). Most striking about the toluene/MFCl 8 system is the difference in

conformational order, indicated by value of I[Va(CHi)]/!! Vs(CHj)], between MFC 18 in

neat toluene, neat methanol, and in 5 mM toluene in methanol. It is quite surprising that

two disordering solvents like toluene and methanol can induce such significant ordering

of the stationary phase when combined. This difference is attributed to the strong

solvation of toluene by methanol at the interface. Similar to the molecular picture of

anisole self-association

i .

H anisole

J

J - „ / r

t « *

^y%

Figure 10.12 Molecular pictures ofthe MFC18/anisole interface for (a) MFC18 in neat anisole and (b) MFC18 in 5 mM anisole in 100% methanol.

4s. tsJ Ji.

m I

toluene

Figure 10.13 Molecular pictures of the MFC18/toluene interface for (a) MFC18 in neat toluene and (b) MFC18 in 5 mM toluene in 100% methanol.

4^ to Ul

anisole/methanol at the MFC 18 interface shown in Figure 10.13b, it is proposed that

toluene partitions only slightly into the stationary phase allowing for a reasonably ordered

methanol structure to form around toluene and the distal end of the stationary phase,

increasing order in the overall phase.

While the molecular pictures of the anisole/MFC18 and toluene/MFC 18 are

similar for 5 mM solute solutions in methanol, the pictures are quite different for anisole

and toluene in methanol/water solutions. Figure 10.14 shows molecular pictures of

MFC 18 in 5 mM anisole (Figure 10.14a) and 5 mM toluene (Figure 10.14b) in 50:50

methanol/water. In these pictures, more methanol than water molecules are shown to

represent interfacial segregation of the methanol component that is known to occur at

alkylsilane stationary phase surfaces.For MFC 18 in anisole/methanol/water, the

order of the phase is disrupted considerably relative to the picture presented in Figure

10.12 for anisole in neat methanol. In 50:50 methanol/water, the free energy for anisole

interaction with the stationary phase is favorable enough to overcome the entropic barrier

to anisole partitioning that exists in 100% methanol. Thus, anisole partitioning into the

stationary phase induces appreciable disorder in the phase. However, mass transport

kinetics for anisole are faster because the mobility of anisole in the more liquid-like phase

is greater. For toluene in 50:50 methanol/water, the well-ordered

toluene/methanol/MFC 18 interfacial structure is still intact, but slightly more toluene

partitioning into the stationary phase results in a slight decrease in conformational order

of MFC 18 relative to a 100% methanol mobile phase.

^ r

anisole :r s V> K-^ »•"-* e At; vv.'-

>-<>- • »> i\y I \ u << *>

->/- '4

•-./• :!> 'x.'

toluene

Figure 10.14 Molecular pictures of (a) MFCl 8 in 5 mM anisole in 50:50 methano I/water and (b) MFC! 8 in 5 mM toluene in 50:50 methano 1/water.

41. to

428

Separation of Monosubstituted Aromatic Compounds using TFC18SF

In addition to studying the effect of chromatographic retention on the order of low

surface coverage, more conventional stationary phases (e.g. MFC 18, 3 .09 |imol/m^),

similar studies have been performed on a high surface coverage stationary phase,

TFC18SF (6,45 jimol/m^). Previous work has shown the chromatographic behavior of

analytes on these two stationary phases to be quite different as a result of the difference in

alkylsilane surface coverage, in particular for the separation of polycylic aromatic

hydrocarbons (PAH).""^ TFC18SF has been shown to be highly selective for planar

PAHs over non-planar PAHs, while MFC 18 is not selective for molecular shape.

Furthermore, work presented in Chapters 6-8 has shown that solvent interactions with

TFC18SF are quite different than with MFC 18. Thus, analyte retention and the effect of

analyte/solvent interactions on the conformational order of TFC18SF are important to

study.

Figure 10.15 shows overlays of chromatograms of aniline, anisole, and toluene

using a mobile phase of 100% methanol on a TFC18SF column. The elution order of the

analytes is the same on the TFC18SF column as it is on the MFC 18 column (Figure

10.1). Thus, the same thermodynamic treatment of retention can be used to predict the

elution order of these solutes. It is important to note, as previously discussed, that this

model does not account for free volume constraints of surface-confmed alkylsilane

stationary phases that are a greater influence for the high coverage TFC18SF phase than

for MFC 18. While the ciution order of these solutes is the same on TFC18SF as MFC 18,

429

<u o

o 'm

1

Time (min)

Figure 10.15 Overlaid chromatograms showing the separation of toluene, anisole and aniline on TFC18SF. Experiments were done in a 100% methanol mobile phase with a constant flow rate of 1.0 mL/min.

430

the retention times for each of the solutes are shorter due to the inability of these solutes

to significantly interact with the stationary phase because of high surface coverage. This

free volume eflfect on retention is not predicted using the thermodynamic model.

As was observed for separation of these solutes using MFC! 8, retention on

TFC18S increases with increasing mobile phase polarity. This is shown in

chromatograms of aniline, anisole, and toluene in mobile phases of 100% methanol,

75:25 methanol/water, and 50:50 methanol/water using TFCl 8SF given in Figures 10.16-

10.18. Capacity factors, relative retention values, numbers of plates, and plate heights are

tabulated in Tables 10.5 and 10.6. The retention order of these solutes (F value) is

accurately predicted by the thermodynamic model as was observed for MFC 18.

However, the k' values are smaller and the H values are approximately twice as large on

TFC18SF as on MFC18. This behavior is attributed to the lower free volume in

TFC18SF that restricts the extent of solute interaction regardless of thermodynamic

favorability, resulting in less retention and smaller capacity factors. Mass transport into

and out of the higher coverage phase is also slower than on MFC18 and the number of

interaction sites is smaller and the mobility of solutes in this more crystalline-like phase

is slower, contributing to larger values of H for a given solute. These effects are shown

schematically in Figure 10.19 by fewer sites on TFC18SF, all at the distal ends of the

chains compared with MFC 18.

431

«r> w QJ o

O en

Anilme

Anilme

i

^ iline C

0 5 10 15

Time (min)

Figure 10.16 Chromatograms showing increased aniline retention on TFC18SF with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water.

432

I

. Anisole r

Anisole

n

tm

i

Anisole a

0 5 10 15

Time (min)

Figure 10.17 Chromatograms showing increased anisole retention on TFC18SF with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water.

433

Toluene

Toluene

Toluene

0 10 5

Time (min)

Figure 10.18 Chromatograms showing increased toluene retention on TFC18SF with increasing aqueous component of the mobile phase from (a) 100% methanol to (b) 75:25 methanol/water to (c) 50:50 methanol/water.

434

Table 10.5. Capacity factor, k\ and relative retention, a, values for anaJytes on TFC18SF

K a

Mobile Phase Aniline Anisole Toluene OtAn,As OÂn,To CtAs,To

100% 0.036 0.095 0.135 2.64 3.75 1.42 methanol

75:25 0.184 0.479 0.804 2.60 4.37 1.68 methanol/water

50:50 0.834 2.80 5.874 3.36 7.04 2.10 methanol/water

Table 10.6. Numbers of theoretical plates (N) and plate heights (H) for solutes on TFC18SF

N n (10 "' m)

Mobile Phase Aniline Anisole Toluene Aniline Anisole Toluene

100% 2980 3685 3876 3.35 2.71 2.58 methanol

75:25 893 3510 4479 11.2 2.85 2.23 methanol/water

50:50 74 5835 6607 135 1.72 1.5 methanol/water

436

MFC 18 Denotes ability for solute intercalation

between chains

TFC18SF

Figure 10.19 Schematic of the free volume differences between MFC 18 and TFC18SF which lead to differences in H values for each phase.

437

Raman Spectroscopy of TFC18S.F in the Presence of Solute and Mobile Phase

Raman spectra in the v(C-H) stretching region from 2750 to 3050 cm ' are shown

in Figure 10.20 for TFC18SF at 20 "C in solute/mobile phase solutions including

aniline/methanol/vvater, anisole/methanol/water, and toluene/methanol/water.

Qualitatively, it is difficult to discern differences between these spectra, but the value of

I[Va(CH2)]/l[Vs(CH2)] can be used to quantitatively determine differences in order of

TFC18SF in these solute/mobile phases solutions.

Effects of Solutes

The effect of solute/methanol solutions on the structure of TFC18SF is shown in

Figure 10.21 in a plot of .I[Va(CH2)]/I[Vs(CH2)] as a function of k' for aniline, anisole, and

toluene in 100% methanol. The horizontal lines represent the value of

I[Va(CH2)]/I[Vs(CH2)] for TFC18SF in neat methanol, aniline, anisole, and toluene for

reference. The value of I[Va(CH2)]/l[v,s(CH2)j for TFC18SF is the same in

aniline/methanol and anisole/methanol solutions and in neat methanol, but is significantly

larger in the presence of toluene/methanol.

The order trends of TFC18SF are the same in aniline and toluene as observed for

MFC 18, but slightly different in anisole/methanol solutions. The rotational order of

TFC18SF is the same in the presence of aniline as in neat methanol. The stationary phase

is primarily solvated by methanol, which induces slight conformational ordering of the

phase as described in Chapter 6, interspersed with weak aniline interaction at the distal

Aniline Anisole Toluene

d a

e b

f c

2900 3000 2900 3000 2800 2800 2800 2900 3000

Wavenumbers (cm"^) Wavenumbers (cm' ) Wavenumbers (cm"^)

Figure 10.20 Raman spectra of TFC18SF in the v(C-H) region in 5 niM aniline (left panel), 5 mM anisole (center panel), and 5 mM toluene (right panel) solutions in (a, d, and g) 100% methanol, (b, e, and h) 75:25 methanol/water, and (c, f, and i) 50:50 methanol/water. Integration times for all spectra are 2 min.

u> (X»

439

U sy

, > ,

r—1

u

1.7

1 .6 -

1 .5

1 . 4

1 .3

1.2

1.1

1.0

0.9

0.8 0.

- Toluene ^

Aniline Anisole

.#

TFCIBSF/Anisole

TFC18 SF/Methanol

TFCISSF/Aniline

" 1 » 1 '

TFC18SF/T oluene

1 • 1 00 0.05 0.10 0.15

Capacity Factor, K

0.20

Figure 10.21 Plot of v/CHj)]/ v/CHj)] as a function of solute capacity factor for toluene, anisole, and aniline on the TFC18SF column. Horizontal lines represent the value of 1[V3(CH2)]/ ^v/CHj)] for TFC18SF in neat methanol, aniline, anisole, and toluene.

440

end of the alkyl chains. In this system, significant aniline partitioning into the stationary

phase is unfavorable due to a combination of limited free stationary phase volume and

favorable solubility of aniline in methanol, resulting in a small value of k\ However, the

broad peak shape and significant peak tailing of aniline suggests an infrequent interaction

between aniline and surface silanols. At sites in the stationary phase in which there is

sufficient volume to accommodate aniline partitioning, it must penetrate deep into the

stationary phase in order to find an exposed silanol. In the case of either weak interaction

with the phase or intermittent deep intercalation, the rotational order of the stationary

phase is unaffected by 5 mM aniline.

The rotational order of TFC18SF is the same in anisole/methanol as in neat

methanol. Thus, the influence of anisole on the ordering of TFC18SF in methanol is

quite different from its effect on MFC 18 in methanol. This difference is presumably due

to the differences in surface coverage and free volume within these two phases. The

limited free volume within TFC18SF is most likely to impact its interaction with anisole

because of the volume and orientation of the methoxy substituent, as discussed above.

Thus, anisole is prohibited from significant partitioning into the TFC18SF phase leading

to little change in conformational order of the phase relative to that in neat methanol.

Anisole exclusion is further consistent with the high degree of selectivity of TFC18SF for

planar molecules.""^ In separations of PAH isomers on TFC18SF, non-planar isomers

elute before planar isomers, due to exclusion of the non-planar isomers from the high-

density phase. Similarly the retention of anisole in 100% methanol is less on TFC18SF

441

than on MFC 18 due to unfavorable entropic effects of solute interaction with the high-

density phase, decreasing the availability of surface sites for anisole interaction. In

addition, based on the magnitude of H, the kinetics of anisole mass transport are slower

in TFC18SF than in MFC 18, indicating some anisole interaction with the TFC18SF

phase, albeit more infrequent than on MFC18.

In toluene/methanol, the rotational order of TFC18SF is greater than in neat

methanol. Toluene interaction with TFC18SF is proposed to be similar to its interaction

with MFC 18 in that the toluene is envisioned to partition into the distal end of the alkyl

chains where it remains solvated by methanol at the interface. This combined interaction

induces slight ordering of the stationary phase.

There are significant rotational order differences of TFC18SF in the presence of

aniline, anisole, and toluene even when the phenyl ring of each solute is likely to interact

with the stationary phase. The difference could be due to an electronic structure effect of

the substituen. In this case, interactions of electron donating -NH2 and -OCH3 groups

are different than the electron withdrawing -CH3 group with methanol and the alkyl

chains.

The conformational order of TFC18SF is only slightly affected by changes in

mobile phase polarity for these solute solutions, shown in Figures 10.22-10.24. For

aniline (Figure 10.22), increasing the polarity of the mobile phase has no effect on the

order of TFC18SF as indicated by the invariant value of I[ VafCHa)]/!(Vs(CH2)]. Although

the frequency of aniline interaction with the stationary phase increases {k' increases and

442

1.7n

1.6-

1.5- -

K" 1.4-u

TJl 1.3--

<N 1.2-- -

U Ci 1.1- -

-

I"—1' 1.0-- • "

0.9- • • —•

0.8- 1 1 1 1 1 t { 9 —1—I—1—1-

0 10 20 30 40


50

7

6

5

4

3

2

1

O

p o

•n

v»

Figure 10.22 I[v/CH2)]/ I[v^(CH2)] (A) and capacity factor, k\ (•) as a function of percent water in methanol for TFC18SF in aniline/methanol/water solutions. The horizontal line represents the value of UvJCHj)]/ I[v/CH^)] for TFC18SF in 100% methanol.

443

1.7-1

1.6-

1.5- •

1.4- .

1.3-

w" o

CS

1.2-

1.1-

• •

1—H 1.0-

0.9- • m—^ •

0.8- 1 > - """r-"'- -T s " 8 « 1 > 1 0 10 20 30 40


50

7

6

5

4

3

2

1

n

I o

Figure 10.23 I[v^(CH2)]/ I[Vg(CH2)] (A) and capacity factor, k\ (•) as a function of percent water in methanol for TFC18SF in anisole/methanol/water solutions. The horizontal line represents the value of I[v,(CH2)]/ Ifv/CHj)] for TFC18SF in 100% methanol.

444

U XJ

u

10 20 30 40


0 1 o

hn p

& •I \p

Figure 10. 24 I[v/CH2)]/ I[Vj.(CH2)] (A) and capacity factor, k\ (•) as a function of percent water in methanol for TFC18SF in toluene/methanol/water solutions. The horizontal line represents the value of^vJCHj)]/ ILv/CHj)] for TFC18SF in 100 % methanol.

445

H increases), the mechanism of interaction is proposed to be the same as in 100%

methanol based on the unchanged order of TFC18SF.

Similarly, in anisole (Figure 10.23) the order of TFC18SF decreases only slightly

with increasing polarity of the mobile phase. As for anisole on MFC 18, the predicted

free energy of retention is more favorable in methanol/water mobile phases than in 100%

methanol. Despite this apparently favorable driving force for partitioning, the limited

free volume within the TFC18SF phase allows only very slight anisole partitioning,

resulting in only a very small decrease in order. However, the interaction between

anisole and TFC18SF is sufficiently strong that the mass transport kinetics of anisole into

and out of TFC18SF are slower than on MFC 18, but much faster than aniline on

TFC18SF.

The conformational order of TFC18SF decreases slightly with increasing mobile

phase polarity in toluene solutions (Figure 10.24). This disordering indicates a slightly

deeper partitioning of toluene into the alkyl chains. Such partitioning is a consequence of

the lower solubility in more polar mobile phases and the more planar nature of toluene

relative to anisole. This toluene planarity results in a slightly more ordered alkyl region

than in neat methanol. Nonetheless, this partitioning is not envisioned to be significantly

deeper than three of four carbons on the distal end of the alkyl chains based on the fast

mass transport kinetics for toluene relative to aniline and anisole.

446

Molecular Pictures of the TFC18SF Chromatographic I nterface

Molecular pictures of the TFClSSF/solute/mobile phase interface are proposed to

reflect the effect of solute and mobile phase on the structure of TFC18SF. Figure 10.25

shows a picture of the interface at a TFC18SF phase for neat aniline (Figure 10.25a) and

for 5 mM aniline in methanol (Figure 10.25b). In neat aniline, the stmcture of TFC IBSF

is the same as in air, suggesting that aniline has very little interaction with the alkyl

chains. In 5 mM aniline in methanol, the order of TFC18SF is greater than in air and

comparable to that in neat methanol. Aniline has little interaction with the stationary

phase resulting in a degree of order comparable to that in pure methanol. Where there are

surface sites containing sufficient free volume, aniline can intercalate between the alkyl

chains and hydrogen-bond with surface silanols, leading to slow mass transport.

However, very few of these sites exist on this high-density phase, rendering these

interactions infrequent.

Molecular pictures of the anisole/TFClSSF interface are shown in Figure 10.26.

Figure 10.26a shows a picture of the neat anisole/TFClSSF interface and Figure 10.26b

shows a picture of the 5 mM anisole in methanol/TFC 18SF interface. In neat anisoie,

TFC18SF is more ordered than in air because of the self-association ability of anisole. In

5 mM anisole in methanol, however, TFC18SF is less ordered than in neat anisole but

more ordered than in air. Anisole is prohibited from significant intercalation between the

alkyl chains because of limited tree volume in the phase, its relatively large molecular

volume, and favorable solvation in methanol. Thus, the stationary phase is primarily

A f .-A 1

't,"' •^«^.

.' %;"" "

^•v

t i t ? : f f 'f • $ # fcx ,* I J.

Figure 10.25 Molecular pictures of the TFCl SSF/aniline interface for (a) TFCl 8SF in neat aniline and (b) TFC18SF in 5 niM aniline in methanol. 4:

Figure 10.26 Molecular pictures of the TFClSSF/anisole interface for (a) I'FCl 8SF in neat anisole and (b) TFC18SF in 5 mM anisole in methanol.

00

449

solvated by methanol, resulting in a degree of order comparable to that in neat methanol.

The addition of water to the mobile phase does not significantly change the interaction of

anisole with the stationary phase or the conformational order of TFC 18SF.

Figure 10.27 shows molecular pictures of the neat toluene/TFClSSF interface

(Figure 10.27a) and the 5 mM toluene in methanol/TFC 18SF interface (Figure 10.27b).

In neat toluene TFC18SF, is more disordered than in any other solvent. This disorder is

due to toluene solvation of the distal ends of the stationary phase, inducing significant

rotational disorder in the phase but no gauche conformers. In contrast to the effect of

neat toluene, in 5 mM toluene in methanol, TFC18SF is very well-ordered; in fact, the

phase is more ordered than under any other experimental condition studied during the

course of this work. This highly ordered state is attributed to strong interactions of

toluene with both the stationary phase and methanol and the self-associating ability of

methanol. This interaction of toluene with the distal end of the alkyl chains provides fast

mass transfer of toluene into and out of the stationary phase, resulting in small H values.

While the elution order of these solutes is the same on TFCl 8SF and MFC 18

stationary phases, the subtle molecular interactions that lead to solute retention are

different, given by the structural information for each phase under the same

chromatographic conditions. MFC 18 allows deeper solute partitioning into the phase

than TFC18SF in which partitioning is restricted to the distal end of the alkyl chains.

Solute interaction with TFC I8SF is restricted to intercalation between the distal ends of

the stationary phase, while the phase remains very well ordered.

1 - A . t ' 1 - '

Figure 10.27 Molecular pictures of the TFClSSF/toluene interface for (a) TFC18SF in neat toluene and (b) TFC18SF in 5 mM toluene in methanol.

451

Mass transport of these solutes into and out of MFC18 and TFC18SF are also

quite different. Slow mass transport kinetics of solutes in TFC18SF relative to MFC18

are due to the crystaJlinity of the phase and limited free volume. MFC 18 is a liquid-like

phase, while TFC18SF more closely resembles a liquid-crystalline phase. Solute mass

transport differences between the two stationary phases are described as mobility

diff erences of solutes within the phase. It is important to note, however, that solute

movement within the phase is limited by the mobility of the alkyl chains. In TFC18SF

longer times are required for alkyl chain reorganization upon introduction of solute.

Thus, alkyl chain reorganization is directly responsible for the slow mass transport of

solutes within the stationary phase.

Concentration Effects

Toluene in Methanol

In the work presented above, solute concentrations were 5 mM, the approximate

solute concentration for the solvent volume in a chromatographic peak. However, the

effect of solute concentration on the order of these phases was also investigated. Raman

spectra of MFC 18 and TFC18SF in 10 and 25 mM toluene in methanol are given in

Figure 10.28. I[VaCCHj)]/![VsCCHz)] values are plotted versus toluene concentration in

Figure 10.29. I[ VA(CHs)J/1 [VS(CH2)] values for TFC18SF and MFC 18 in neat methanol

(from Chapter 5) are shown as dashed horizontal lines for reference. TFC18SF and

MFC 18 are both more ordered in toluene/methanol solutions for all toluene

452

TFC18SF MFC18

a

b

3000 2800 2900 3000 2800 2900 Wavenumbers (cm"^) Wavenumbers (cm-\)

Figure 10.28 Raman spectra of TFC18SF (left panel) and MFC 18 (right panel) in (a and c) 10 mM and (b and d) 25 mM toluene in 100% methanol solutions. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 spectra are 5 min.

453

1.7

1.6

1 . 5 '

1 . 4 '

1.3-

1.2-

B 1.1 as

0.9

0.8 0

ffi u

a 1-i

TFCISSF/Methanol

MFC 18/Methanol

5 10 15 20 25

Toluene Concentration (mM)

30

Figure 10.29 Plot of I[V3(CH2)]/ l[Vg(CH2)] as a function of toluene concentration for TFC18SF (•) and MFC 18 (A) in toluene/methanol solutions. Horizontal lines represent the value oflLv/CHj)]/ v^CHj)] for TFC18SF and MFC18 in neat methanol.

454

concentrations investigated than in neat methanol. The value of I[v.i(CH2)]/![ Vs(CH2)] for

TFC18SF decreases as the toluene concentration increases from 5 to 10 mM, but then

remains constant from 10 to 25 mM. This trend suggests a slight decrease in order with

an increase in toluene concentration for TFC18SF. In contrast, the value of

I[Va(CH2)]/I[Vs(CH2)] increases with concentration for MFC 18, indicating that the phase

becomes slightly more ordered.

As the toluene concentration increases from 5 to 25 mM, TFC18SF becomes

slightly more disordered, while MFC 18 orders. For TFC18SF, increasing the amount of

toluene at the chromatographic interface disrupts the order of the phase, as more toluene

molecules partition into the limited free volume of the phase. The disorder induced by 10

and 25 mM toluene is far less than that induced by neat toluene, because of strong

solvation of toluene by methanol at the interface. However, the disorder in these

concentrations is slightly more than in 5 mM tolueme. For MFC18, the opposite trend is

observed, due to greater free volume within the stationary phase. Thus, an increase in

toluene at the interface results in a greater number of toluene/methanol/MFCl 8

interaction of the type that result in the highly ordered toluene/methanol structure at the

distal end of the alkyl chains discussed above.

Toluene in Acetonitrile

In addition to examining the order of TFC18SF and MFC 18 in methanol

solutions, the order of these phases was also examined in toluene/ aceto nitrile solutions.

455

The order of these phases in neat acetonitrile was discussed in Chapter 5 and the value of

I[Va(CH;)|/I[Vs(CH2)] for TFCl 8SF in neat acetonitrile is 1.36 and I 05 for MFCl 8.

Raman spectra of MFC 18 SF and TFC18SF in 5, 10, and 25 mM toluene in acetonitrile

are shown in Figure 10.30. Figure 10.31 shows plots of 1 [VA{CH2)]/I[VS(CH2)] as a

function of toluene concentration for TFC18SF and MFCl 8 in 5, 10, and 25 mM toluene

in acetonitrile. Analogous to the behavior of TFCl 8 SF and MFCl 8 in toluene/methanol

solutions, the values of I[Va(CH2)]/I[Vs(CH2)] for both phases are greater in

toluene/acetonitrile solutions than in neat acetonitrile or neat toluene. However, no

dependence of conformational order on toluene concentration is observed over the

concentration range investigated.

In all toluene/acetonitrile solutions, the conformational order of both TFC18SF

and MFC 18 is greater than in neat acetonitrile. This behavior is consistent with solute-

induced ordering of these phases in toluene/methanol solutions. It is reasonable to expcct

the interactions of toluene/methanol and toluene/acetonitrile with these phases to be

similar because the solvation of these phases by neat methanol and neat acetonitrile are

largely the same. Unfortunately, data on the relative solubility of toluene in acetonitrile

are not readily available in the literature. The fact that the ordering of both phases in

toluene/acetonitrile solutions is independent of toluene concentration and that the order of

these phases in toluene/acetonitrile solutions is not as great as in methanol/toluene

solutions suggests that the interactions between acetonitrile and toluene are different from

those between methanol and toluene. In the absence of solubility data it is difficult to

456

TFC18SF MFCl

a

b

c

2800 2900 2800 2900 3000 3000 Wavenumbers (cm^^) Wavenumbers (cm"^)

Figure 10.30 Raman spectra of TFCI8SF (left panel) and MFC18 (right panel) in (a and d) 5 mM, (b and e) 10 mM, and (c and f) 25 mM toluene in 100% acetonitrile solutions. Acquisition times for TFC18SF spectra are 2 min and for MFC 18 spectra are 5 min.

457

1.6H

1.5

1.4

1 . 3 '

g 1.2-

5 1 . 1 -

i . o .

0.9 H

0.8 0

i

TFC18SF/ Acetonitrile

MFC 18/Acetonitrile

1 ^ 5 10 15 20 25

Toluene Concentration (mM)

30

Figure 10.31 Plot of I[Vg(CH2)]/ I[Vj,(CH2)] as a function of toluene concentration for TFC18SF (•) and MFC 18 (A) in toluene/acetonitrile solutions. Horizontal lines represent the value of I[v^(CH2)]/I[v/CH2)] for TFC18SF and MFC 18 in neat acetonitrile.

458

tUlly interpret the differences between acetonitrile and methanol. More organic mobile

phases should be investigated before meaning&l conclusions can be drawn about the

effect of mobile phase solvent on the interactions of toluene with these phases.

Conclusions

Combined chromatographic and spectral data offer unique insight into retention

mechanisms for a series of solutes separated on a high-density and low-density stationary

phases. For these weakly retained solutes, W values are smaller and H values are larger

on TFC18SF relative to MFC 18. Smaller values of k\ i.e. poorer solute retention, are a

result of fewer interaction sites available and limited free volume within the high surface

coverage stationary phase. Likewise, larger H values are a result of free volume

restrictions within the phase where solute mobility is hindered by crystalline-like alkyl

chains of TFC18SF, relative to the more liquid-like alkyl chains of MFC 18.

Spectral data provide stinctural information for these stationary phases under

chromatographic conditions, which can be used to ascertain molecular interaction

information for these systems. On MFC 18, poor aniline retention is a result of

unfavorable interactions with the stationary phase relative to its solubility in these three

mobile phases. Slow aniline mass transport is a result of strong hydrogen-bonding with

available surface silanols. Thus, the conformational order of the alkyl chains is largely

unaffected by the presence of aniline. Anisole partitions into the MFC 18 stationary phase

inducing similar disorder as aniline in mobile phases of 75:25 methanol/water and 50:50

methanol/water. In 100% methanol, anisole partitioning into the stationary phase is

energetically unfavorable due to strong solubility of anisole in methanol and volume

restrictions to anisole partitioning. Toluene only partitions very slightly into the

stationary phase, remaining accessible to interfacial solvent molecules. Toluene is

incorporated in the highly ordered interfacial methanol structure, resulting in rotational

ordering of the stationary phase.

Solute interactions with TFC18SF are somewhat different than with MFC 18, in

general, because of the high density of alkyl chains on TFC18SF, restricting significant

analyte partitioning into the phase. However, intercalation or partitioning of solutes

between the distal ends of the alkyl chains is believed to occur from the subtle disorder

imparted to these phases in aniline and anisole in methanol/water mobile phases. While

this view of partitioning is different from the idea of cavity formation deep within alkyl

chains, it is a more partitioning-like interaction than classical adsorption or solvophobic

interaction.

Results from these Raman spectroscopic and chromatographic studies lend

credence to the idea that solute/stationary phase interactions occur predominantly at the

stationary phase/mobile phase interface and support the idea of solute partitioning as a

retention mechanism. These experiments prove that the stationary phase has an active

role in solute retention, where interactions depend heavily on the free volume of the

stationary phase, and cannot be classified only as solvophobic. However, partitioning in

these systems is very subtle and these data suggest that even for hydrophobic solutes on

460

low coverage phases, intercalation does not take place deep into cavities between alkyl

chains.

Future Directions

There are a plethora of experiments that could be done in the fiiture as extensions

of the work presented in this Chapter, as studies were only done using two stationary

phases, three solutes and four different mobile phases. Two experiments of primary

importance are: (I) investigating the etfect of a homologous series of alkylbenzenes on

the order of alkylsilane phases and (2) repeating these experiments on similar high and

low coverage phases with polycyclic aromatic hydrocarbon solutes (PAHs).

Study of a homologous series of alkylbenzenes is important because these data

can be correlated to separation behavior of alkylbenzenes,'"a long time

favorite of chromatographers, used to determine retention thermodynamics in a multitude

of mobile phases at varying temperatures. This would allow for the Raman spectral data

to be correlated with actual retention free energy values, instead of the approximations

used in this report.

One unique characteristic of the high-density NIST Cig-phases is their geometric

shape selectivity for planar PAHs, determined by Sander and Wise."""''"" Thus using

PAHs as solutes in our spectroscopic experiments would show a relationship to these

chromatographic experiments. It is hypothesized that the conformational order of high

surface coverage phases such as TFC18SF would be different for planar PAH solutes

461

than for nonplanar PAHs, because of their different interactions with high bonding

density phases. In general, these experiments would give some structural information

about densely-packed alkylsilanes during chromatographic separations that could be used

to better understand their functions as shape selective phases.

There is also a multitude of additional experiments of lesser priority that could be

performed in the near future. These experiments include investigating the order of

similar phases with the same monosubstituted aromatic solutes over a wider range of

mobile phase polarity (0-100% water) with different organic modifiers (acetonitrile and

THF), studying the order of phases with varying alkyl chain length (C4-C30), and

exploring the effects of different solutes relevant in areas of environmental (chlorinated

aromatics), biological (amino acids or nucleotide bases), or forensic (drugs of abuse)

interest where RPLC is a commonly employed analytical technique.

462

Chapter 11

Selectivity and Structural Properties of Anion and Cation Exchange

Chromatographic Systems Determined by Raman Spectroscopy

Introduction

Ion chromatography (IC) is a class of experimental techniques used to separate

ions. Included in this class of methods are ion exchange chromatography, ion-pair

chromatography, ion exclusion chromatography, extraction chromatography,

complexation chromatography among others. These techniques allow separation and

quantitation of ionic species that cannot be accomplished with normal-phase or reversed-

phase liquid chromatographic techniques. Ion chromatography has been proven useful

for the separation of ionic species in biological systems, such as proteins, peptides and

nucleotide bases,^'^"'''^ transition metal cations, radioactive ions of environmental and

nuclear importance,"^""'"' and in the extraction of narcotics from biological fluids for

forensic applications."^^'

The earliest and most commonly used stationary phases or exchange resins used

in ion chromatography were based on functionalized polymers (styrene-divinylbenzene

copolymers (PS-DVB) or acrylate based polymers). More recently, modified-silica or

other modified-inorganic particles have been used as supports for ion separations. While

both particulate and resin materials are widely used for all types of ion chromatography.

463

there are some important distinctions to be made between them. The polymeric resins

generally have lower capacities (0.01 to 0.5 mmol/g) and can be used over a wide pH

range (pH 1 to 14). However, separations using these materials suffer from the

susceptibility of these polymers to free volume changes, swelling and shrinking in

solvent. Modified-silica substrates offer better chromatographic performance and

excellent mechanical stability; however, the use of these materials is limited to a

narrower pH range of 2 to 8." " The most common functional groups for strong and

weak cation exchange materials are sulfonic acid and carboxylic acid groups,

respectively, while quaternary amines, primarily trimethylalkylammonium ions, are used

for strong anion exchange materials. Primary or secondary amines are used as strong and

weak anion exchange supports. Quaternary amine strong anion exchange materials are

the focus of this work.

Retention in ion chromatographic systems is governed by electrostatic attraction

between an analyte ion in solution and a surface-confined ion of the exchange resin or

modified particle. Retention in ion chromatography has been shown to be dependent not

only on the exchange resin properties, but also analyte ion charge, size, hydration and

solution activity. For example, anion-exchange selectivity for the halide ions follows the

order CI" < Bf < F. Since these anions are poorly hydrated in solution, this selectivity

order is solely dependent on ion size in the crystal lattice and not on the hydrated

radius." For example, I" is more strongly held than CF because F has a larger

crystal radius (Ri- = 0.216 nm; Rq- = 0.181 nm), a lower charge density, and is less

464

well hydrated than CP even though both have approximately that same hydrated radius

(Ri- (hydrated) ~ 0.330 nm; Rci- (hydrated) ~ 0.330 nm)."'"

Selectivity for cations in cation chromatography follows the order Li" < Na " < K'

< Cs \ However, unlike anions, cation selectivity is dependent on the charge-density of

the hydrated ions, as cations are generally more hydrated than anions." For the series

of alkali metals, cesium is the least hydrated ion (Rcs+ = 0.169 nm, Rcs+ (hydrated) =

11 13 0.329 nm), ' has the highest charge-density of the ions and is held most strongly.

Conversely, lithium has the lowest charge-density in the series (Ry- = 0.060 nm; Ru

(hydrated = 0.382 nm)" ' ' and is held weakly.

The trend of increased selectivity with increasing charge-density for both anions

and cations follows the Hotmeister ion series.This series was developed to

explain the observations that smaller ions salt out egg lysozyme at higher concentrations

than large ions in aqueous solutions and has been used to explain similar behavior for

numerous proteins and hydrophobic solutes. Smaller ions have less interaction with

solutes because they are more hydrated than larger ions, effectively excluding the solute

from solution. Counter ion hydration (charge-density) and its effect on interactions with

other solution species are key issues in understanding selectivity in ion chromatographic

systems, as well.

While interactions between surface-confined trimethylalkylammonium ions or

sulfonates and solution counter ions are predominantly electrostatic, additional factors

can influence ion retention in ion chromatography because one ion of the pair is tethered

465

to a surface. Additional considerations for surface-bound ionic interactions include

surface heterogeneity of charged sites, accessibility of charged sites to solvent and

counter ions, hydration characteristics of surface-bound ions, solvation effects of

substrate, as well as hydrophobic or other nonspecific adsorption of analyte to the

substrate. " These issues add to the complexity of describing interactions in these

systems and affect the structure of both surface-bound and solution counter ions.

The tetramethylammonium ion (TMA ) has been used to model the

trimethylalkylammonium ions of ion exchange materials. The structures of TMA" in

crystal lattices, melts and aqueous solutions has been investigated using numerous

11 22 11 23 approaches including Fourier transform infrared (FT-IR) spectroscopy, ' ' ' nuclear

magnetic resonance molecular dynamics simulations/' and

Raman spectroscopy " ^latter is tremendously powerfol because of its

sensitivity to structural changes and chemical environment, superior resolution and

activity of spectral bands that are indicators of ionic structure. For example. Pal and

coworkers^"' have described Raman spectral indicators of structural changes of TMA*

in crystalline halide salts including the Va(CH3)N mode (-3020 cm"'), which shifts in

frequency by up to 12 cm"' for different halides (Cf, Br , I") as well as changes in

intensity of the Vs(C4N) mode (-747 cm"') with temperature. However, to the best of our

knowledge, structural changes of TMA ' in solution or surface-confined

trimethylalkylammonium analogues of TMA had not been previously reported.

466

The structure and conformation of sulfonate ions and sulfonic acid species have

been investigated by FT-IR spectroscopy, ' ' ' nuclear magnetic resonance

(NMR)," 41 as well as by utilizing molecular dynamics simulations/' but

R a m a n s p e c t r o s c o p y h a s b e e n p e r h a p s m o s t e f f e c t i v e l y u s e d f o r s u c h s t u d i e s . 4 -

For example, Edwards et al." described Raman spectral changes including frequency

shifts of the v(S03) (-1040 cm"') and v(CS) (-770 cm'^) modes of up to 7 and 8 cm"',

respectively, and intensity changes of the v(S02) (1035 and 1125 cm"') bands for

methanesulfonate ions in hydrated and dehydrated environments and when paired with

different counter ions.

The ionic structure of surface-confined sulfonates/sulfonic acids has also been

investigated using Raman spectroscopy for sulfonated polymer systems in battery, fiiel

ceil or biomedical membrane applications." For example, Edwards et al.'' ^ have

reported shifts in v(S03) peak frequency and full-width at half maximum (FWHM) with

hydration and deprotonation for sulfonic acid-modified polystyrene (PS) resins. In

addition, Mattsson and coworkers" have observed similar fi-equency shifts in the

v(S03) band upon hydration and additional changes in the frequency and FWHM of the

v(S03) band with exchanging counter ion, H and Na", for dry polyethylene oxide (PEO)

and propylene oxide (PPO) polymer resins.

FT-IR spectroscopy has been applied to the determination of hydration effects and

counter ion on sulfonate structure in cation exchange resins. In separate reports,

Gordon" '4 and Lowry et al." " studied the structure of sulfonated polystyrene-

467

divinylbenzene and perfluorinated resins as a function of hydration, polymer swelling,

and counter ion exchange for the dehydrated resin, by monitoring the peak frequency and

peak width of the v(S03) mode. Results indicate that the sulfonate group is sensitive to

changes in hydration and association with counter ions of varying size in the dehydrated

resin. However, the use of Raman spectroscopy to investigate structural changes of

surface-confined polyatomic ions in the context of cation or anion chromatographic

systems either on polymeric resin-based or particulate, alumina or silica substrates had

not been reported.

Goals of this work

The goal of this work was to acquire Raman spectra of anion and cation exchange

materials in aqueous solutions of different counter ions. We wanted to identify structural

changes to surface-confined trimethylaminopropylsilane (TMAP ) and

propylsilylsulfonate (-SO3) groups of the exchange materials by monitoring certain

Raman spectral indicators, similar to the FT-IR spectroscopy experiments previously

11 \ A 11 c reported. ' ' However, unlike the previous FT-IR work, we study these structural

changes of ion chromatographic materials under aqueous conditions, mimicing real ion

separation experimental conditions. Observed structural changes can be correlated to the

selectivity of these ion exchange materials for different counter ions. This information

can be used to determine whether the strength of counter ion interaction influences the

structure of the surface-bound ions. In addition, we model these interactions using solid

468

and aqueous tetramethylammonium salts and aqueous methanesulfonic acid to determine

if interactions at the surface mimic aqueous or solid state conditions.

Methodology

Commercial solid-phase extraction stationary phases, I solute SAX (referred to as

SAX) and Isolute SCX-2 (referred to as SCX-2) were purchased from International

Sorbent Technologies (Hengoed, UK). Properties and characteristics of these phases were

described in Chapter 2. Briefly, SAX is a strong anion exchange stationary phase based

on trichlorosilylpropyltrimethylammonium chloride-modified irregular silica particles,

shown in Figure 11.1. SCX-2 is a strong cation exchange stationary phase fabricated

using trichlorosilylpropylsulfonic acid, shown in Figure 11.2. CsCl (99.9%) and

(CH3)4NC104 (99%) were purchased from Alfa Aesar. NaH2P04*H20 (99%) was

purchased from Curtin Matheson Scientific, Inc. NaNOs (>99%), NaCl (99,8%), and Nal

(>99%) were purchased from Malinckrodt Inc. NaBr (>99%), KCl (>99%),

NaHS04»4H20 (98%), NaC104 (98%), NalO, (99%), Na3P04»12H20 (99%), Na2S04

(>99%), (CH3)4NBr (>99%), (CH3).»N1 (>99%), (CH3)4NC1 (>98%), (CH2CH3)4NBr

(98%), (CH3)4NHS04 (>99%), (CH3)4NN03 (>99%), and CH3SO3H (99%) were

purchased from Aldrich. Samples were prepared by adding 100 |iL of 0.3 M sodium salt

anion solutions (CF, Br", I", NO3, HSO4", CIO4" and H2PO4") and 100 |iL of 0.2 M

chloride or bromide cation solutions (Na\ K', Cs , (CH3)4 , (C2H5)4^) to 50 mg of

stationary phase in a 5.0 mm-dia NMR tube, such that the number of moles of solution

Me^ Me

v.- CI-Me

Me

Me / Me

Me^ Me

N+ Cl-Me

CI—Si^ / CI

CI

HO~Si^ c,-

o HO"Si^

d' O"

!

Trichlorosilylpropyltrimethylammonium chloride and a simple illustration of the SAX surface.

O- H +

Cl-Si

O- H +

HO-Si /

H +

HO-Si^

o' OH I

Figure 11.2 Trichlorosilylpropylsulfonic acid and a simple illustration of the SCX-2 surface.

•-4 O

471

ions and surface-confined ions were approximately equivalent based on the nominal

reported capacities of these materials. For the anion system, model solutions (100 jiL)

were 0.3 M aqueous solutions of the tetramethylammonium salts listed above. For the

cation system, model solutions were prepared by mixing 50 |iL of 0.4 M aqueous

methanesulfonic acid (pKa ~2) with 50 |iL 0.4 M aqueous cation chloride solutions pH

—1.04, giving a solution that was 100 {jL of 0.2 M cations and aqueous methanesulfonate.

Raman spectra were collected using 100 mW of 532-nm radiation from a

Coherent Verdi Nd:YV04 laser on a Spex Triplemate spectrograph, described in Chapter

2. Slit settings of the Triplemate were 0.5/7.0/0.150 mm for the SAX/halide ion

experiments and 0.5/7.0/0.025 mm for all other experiments, corresponding to a spectral

bandpass of 5.3 and 0.90 cm"', respectively. Samples were allowed to equilibrate at 20

"C for a minimum of 30 main prior to spectral analysis to negate temperature effects on the

vibrations of these surface-bound and solution ions. A minimum of three measurements

were made on each sample. Integration times for each spectrum are provided in the figure

captions.

1 solute SAX

Raman Spectroscopy of Isolute SAX

Although Raman spectroscopy has been used to determine structural changes of

tetramethylammonium ions in various chemical environments,'' no previous

reports describe the use of Raman spectroscopy to study the structure of surface-confined

472

quaternary amines on any substrate. As demonstrated in previous reports from this

laboratory, high quality Raman sepctra from -400 to 3000 cm"' can be obtained for

weakly scattering surface-confined alkylsilanes on the identical silica substrate used for

the 1 solute SAX material." Spectra collected for surface-confined molecules on

these particulate substrates are free from spectral aberrations. Figure 11.3a shows the

Raman spectrum of dry SAX in the spectral regions from 400 to 1500 cm"' and 2750 to

3300 cm"' and Figure 11.3b shows the Raman spectrum of SAX in water for the same

frequency regions. The corresponding peak frequencies and assignments are given in

Table 11.1. The spectum of SAX is dominated by modes from the trimethylammonium

portion of the silica modifier as opposed to the underlying propyl chain. In the low

frequency region, very little effect of the solubilization of SAX in water is readily

apparent in the Raman spectrum. In the high frequency region, only the presence of the

water envelope and the shift in peak frequency of the Va(CH3)N mode from -3030 to

-3042 cm"' for aquated SAX are qualitative differences that are observed. More subtle

differences in these Raman spectra that are related to structural changes of these surface-

confined ions as a function of changing chemical and physical environment are discussed

in detail below.

Raman Spectroscopy of Isolute SAX in Aqueous Sodium Halide and Oxyanion Solutions

From previous reports on the structure of tetramethylammonium salts, it was

hypothesized that the structure of the surface-confined trimethylaminopropylsilane

® FH m c CD

12500 V/CH3) 5(CN+) 5,(CH3) v/CH3),^ I

. , / n x T X in V,(CH3)N Va(CH3)s.

viCN-) 25000

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm"0

Figure 11.3 Raman spectra of (a) dry SAX (as received) and (b) SAX in water at 20 °C. All spectra are acquired with 30-min integrations.

•--J

Table 11.1. Peak frequencies and assignments for Isolute SAX (TMAP )

Peak Frequency (cm"\) Assignment" ''

SAX (Dry) SAX (Hydrated)

440 460 5(CN')

752 755 V.(C4N0

912 915 v.(CN")

952 953 Va(CN )

970 971 V (CN*)

1055 1056 SsCCH:,)

1185 1180 r(CH3)

1250 1250 r(CH3)

1419 1420 6(CH2)

1450 1451 5a(CH3)

2902 2903 Vs(CH3)si

2933 2933 VS(CH3)FR

2975 2976 VS(CH3)

3030 3042 Va(CH3)\

"Assignments from references 11.22, 11.23, 11.31. ''5 = bend or deformation; v = stretch; r = rock, s = symmetric; a = antisymmetric.

475

derivative (TMAP ) should be sensitive to its chemical and physical environment."^'

Modifying the stationary phase particle hydration state and characteristics of the counter

ion in solution are ways to change the environment of surface ions that mimic ion

chromatography separation or exchange conditions. The interactions between stationary

phase materials and counter ions in solution are the fundamental basis for retention and

selectivity for these systems; therefore, examining structural changes in TMAP' with

counter ions of varying size and charge is crucial for understanding retention in ion

chromatography. The effect of counter ion size and charge on the structure of TMA' has

only been studied by Raman spectroscopy for the solid-state crystalline salts. These

conditions are not directly applicable to understanding ion chromatographic experimental

11 23 1131 11 32 conditions, ' ' ' ' ' and hence, the justification for the studies reported here.

For SAX, the structure of the surface-confined quaternary amine TMAP' was first

investigated by Raman spectroscopy as a function of hydration and with the exchange of

the halide ions, CI", Br, and T. Raman spectra of SAX in 0.3 M solutions of the sodium

salts of these anions are shown in Figure 11.4. These spectra are qualitatively similar to

those of dry and hydrated SAX (Figure 11.3) with two notable exceptions. Two Raman

spectral indicators have been identified that reflect differences in counter ion interactions:

(1) the peak frequency of the antisymmetric C-H stretch for the methyl groups of the

quaternary amine (VaCCHs)^, -3040 cm"') and (2) the intensity ratio of the antisymmetric

C-N-C stretch of the amine (va(C-N-C), -950 cm"') to the symmetric carbon-nitrogen

stretch (Vs(C-N-C), -912 cm"'), represented as I[Va(C-N-C)]/l[Vs(C-N-C)].

7500 25000

m

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm-^)

Figure 11.4 Raman spectra of SAX in 0.3 M aqueous solutions of (a) NaCl, (b) NaBr, and (c) Nal. All spectra are acquired with 30-min integrations. o^

477

In addition to studying structural changes in SAX with the monoatomic halides, it

was also important to investigate the structure of SAX in the presence of more complex,

polyatomic counter ions. Figures 11.5 and 11.6 show Raman spectra of SAX in 0.3 M

aqueous solutions ofNaNOi, NaC104, and NalO? (Figure 11.5) and Na2S04, NaHS04,

Na3P04 and NaH2P04 (Figure 11.6). In addition to the spectral bands from SAX,

additional spectral bands from each of the oxyanions are observed in the region between

400 and 1500 cm"'. The corresponding peak frequency assignments for these modes are

given in Tables 11.2-11.8. In the high frequency region, no spectral interference from

oxyanion modes, and the Va(CH3)\ frequency can be precisely measured. Some spectral

overlap of modes from the oxyanions with the surface-bound TMAP^ in the low

frequency region occurs, including interference with the Vs(C-N-C) mode from the

Vs(C104), VS[P(0H)2], and Vs(P04) bands from CIO4", H2PO4", and PO4 ' respectively. For

these oxyanion/SAX systems, a value of I [ Va(C-N-C)]/l [ Vs(C-N-C)] cannot be

determined.

Figures 11.7 shows a plot of Va(CH3)\ frequency as a fimction of ion

chromatographic selectivity of CI", Br", I", NO3', H2PO4', HSO4', CIO4", S04'^, IO3", and

P04'^. Figure 11.8 shows a plot of I[Va(C-N-C)]/I[Vs(C-N-C)] as a function of anion

selectivity for CI", Br", I", NO3", HSO4", S04"^, and IO3". The oxyanions P04"^, CIO4", and

H2PO4' are not shown in Figure 11.8 because of spectral interference with the Vs(C-N-C)

modes. Because the use of modified-silica substrates in ion chromatography is relatively

new compared to the use of polymeric resins, selectivity values for silica-based

7500 25000

m C

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm-^)

Figure 11.5 Raman spectra of SAX in 0.3 M aqueous solutions of (a) NaC104, (b) NaNO,, and (c) NalOj. All spectra are acquired with 30-min integrations. So

25000 7500

C (D

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cnr^)

Figure 11,6 Raman spectra of SAX in 0.3 M aqueous solutions of (a) Na2S04, (b) Na^PO.,, (c) NaHSO.,, and (d) NaH2P04. All spectra are acquired with 30-min integrations

480

Table 11.2. Peak frequencies and assignments for NaNOs

Peak Frequency (cm" ) Assignment"*

NaNOs 0.3MNaN03(aq)

725 SipCNOj)

1068 1048 VsCNOa)

1385 VafNOa)

"Assignments from references 11.60 and 11.61.

481

Table 11.3. Peak frequencies and assignments for NaHS04

Peak Frequency (cm" ) Assignment^

NaHS04»4H20 0.3 M NaHS04 (aq)

415 5(S0)

438 8(S0)

576 Ss(S03)

605 597 5S(S03)

859 Vs(S-O-H)

884 891 Vs(S-O-H)

981 Vs(S-O-H)

1040 1052 Vs(S04)

"Assignments from references 11.62 and 11.63.

482

Table 11.4. Peak frequencies and assignments for NaH2P04

Peak Frequency (cm"') "Assignment

NaH2P04*Hr0 03 M NaH2P04 (aq)

383 5(P0)

512 6(P0)

526 S(PO)

910 878 Vs[P(0H2)]

974 Vh[P(0H2)]

1075 VsCPOj)

1172 VaCPOs)

"Assignments from reference 11.64.

483

Table. 11.5. Peak frequencies and assignments for NaCl04

Peak Frequency (cm"') Assignment

NaCioI 0.3 M NaC104 (aq)

443 6(Cl-0)

482 6(Cl-0)

619 5(Cl-0)

628 5a(C104)

656 5a(C104)

952 935 Vs(C104)

1087 Va(C104)

1097 Va(C104)

1147 Va(C104)

Âssignments from reference 11.65.

484

Table 11.6. Peak frequencies and assignments for NalO^

Peak Frequency (cm" ) "Assignment

Naiol 0.3 M NalO:, (aq)

757 799 v,(I03)

776 Va(I03)

788 VadOj)

Assignments from reference 11.66.

485

Table 11.7. Peak frequencies and assignments for Na2S04

Peak Frequency (cm"') "Assignment

NalsO^ 0.3 MNa2S04 (aq)

451 ^ " 5(S-0)

466 5(S-0)

620 6(S-0)

631 5a(S04)

647 8a(S04)

992 981 Vs(S04)

1101 Va(S04)

1131 Va(S04)

1152 Va(S04)

"Assignments from references 11.67.

486

Table 11.8. Peak frequencies and assignments for Na3P04

Peak Frequency (cm" ) Âssignment

Na3F04 0.3MNa3P04(aq)

420 5s(P04)

548 8a(P04)

940 938 Vs(P04)

1008 991 Va(P04)

1018 Va(P04)

1027 Va(P04)

1068 1067 VsCPOs)

''Assignments from reference 11.68.

487

3044-,

3042- H^PO;

- ^ Br 3040- -• 0

• IO3 ^ 3038- NO3-

PO.

a o

<1> ^ 3036^

I ^ 3034

5 3032-

3030-

I • S O ,

\ Hso;

-2

§Dry

CIO.

-2 0 2 4 6 8 Anion Selectivity, K

10 12 14

AVCi-

Figure 11.7 Plot of Vj,(CH3)^. peak frequency as a function of anion selectivity for SAX in 0.3 M aqueous solutions of sodium halides (O) and sodium oxyanions (•). Error bars represent one standard deviation.

488

Q I

U QT

uL

U s

a a!

>

•

2 4 6 8 10

Anion Selectivity, K A-/a-

Figure 11.8 Plot of I[Vj(C-N-C)]/I[Vj.(C-N-C)] as a function of anion selectivity for SAX in 0.3 M aqueous solutions of sodium halides (O) and sodium oxyanions (•). Error bars represent one standard deviation.

489

chromatographic materials are not readily available in the literature. Selectivity values

used for anions in this study were taken from those measured on polymeric

tetramethylamnionium-functionalized materials by Peterson et al.^' and others," " '"^

the properties of which have been characterized in detail and for which the retention

behavior of ions is similar to that on silica-based materials."

For the halides, the Va(CH3)N peak frequency (Figure 11.7) of the surface-bound

quaternary amine chloride increases as SAX is hydrated, and then systematically

decreases as the counter ion is changed in solution from CI" to Br' to Y. These results

reflect the sensitivity of the Va(CH3)N peak frequency to chemical environment of the

quaternary amine. For SAX, the chloride counter ion makes a contact ion-pair with the

amine, both ions are without a hydration sphere, and the Va(CH3)N peak frequency

decreases, probably due to an increase in reduce mass. In aqueous solution, both the

amine and the chloride ion are hydrated and exist as a solvent-separated ion pair, shifting

the peak to higher frequency. Upon changing the counter ion to larger, less well-hydrated

ions such as Br" and I" (i.e. greater charge-density of the hydrated ion), the separation

between the ions decreases, thereby increasing the effective reduced mass of the amine,

and resulting in a decrease in the Va(CH3)N peak frequency. In addition, the hydration

enthalpy (AHHYD) for the halides decreases from CI' (AHHYD (CI") = -367 kJ/mol) to Br"

(AHHYD (Br ) = -336 kJ/mol) to I' (AHHYD (I ) = -291 kJ/mol);"'^^ a complete listing of

hydration enthalpy, ionic and hydrated radii for all anions in this study are given in Table

11.9. As a result of poorer hydration, interaction with the hydrophobic amine surface

490

Table 11.9. Hydration enthalpies, ionic radii, and hydrated radii of anions

Anion ^Hydration Enthalpy (AHuvd) in kJ/mol

^lonic Radius (ri) in nm ^Hydrated Radius (rn) in nm

cr -367 0.181 0.332

Br" -336 0.195 0.330

r -291 0.216 0.331

NOj- -325 0.264 0.335

HSO4" -360 0.260'* 0.344''

IO3" -450 0.330 0.374

CIO4" -227 0.292 0.338

H2PO4' -522 ~ 0.200 0,425''

S04^ -1035 0.290 0.379

P04'^ -2879 0.238'-' ~ 0.450*'

"Values from reference 11.56. ''Ionic and hydrated radii from reference 11.13, """Hydrated radius for H2PO4" from reference 11.69. ''ionic and hydrated radii for HSO4" from reference 11.70. Ionic radii for P04"^ from reference 11.71. 'Hydrated radii for P04"^ approximated from reference 11.72.

491

perturbs the hydration sphere of I" more easily than that of CI", making the interaction of

r with TMAP^ a closer resemblance to that of a contact ion-pair, Counter ion

interactions with TMAP \ which result in perturbation of the TMAP" structure as

reflected by a change in VaCCH^)^ frequency, depends on charge-density of the counter

ion as well as hydration enthalpy of the ion. These observations are consistent with the

Hofmeister series in which larger ions with lower hydration enthalpies have more

interaction with hydrophobic solutes, in this case, TMAP', and are poorer at salting-out

solutes than smaller ions with greater hydration enthalpies."" "^"

For the halides, the value of I[Va(C-N-C)]/I[Vs(C-N-C)] (Figure 11.8) shows a

similar dependence on ion selectivity as the Va(CH3)N frequency. This ratio increases

with increasing selectivity of the counter ion as a result of similar changes in ion

separation and anion charge-density as described above to explain the frequency shift in

the Va(CH3)N mode. For dry SAX, the Va(C-N-C) mode is more susceptible to a change in

ion polarizability than the Vs(C-N-C) mode, because the Vs(C-N-C) mode is more

physically restricted by the proximate chloride ion. When the ions are hydrated, they are

further apart, the Vs(C-N-C) vibration is no longer restricted, and I[Va(C-N-C)]/I[Vs(C-N-

C)] decreases. Upon exchanging chloride for bromide or iodide, which are less well

hydrated, the ions are closer together, restricting the polarizability of the Vs(C-N-C) mode

relative to the Va(C-N-C) mode, and the value of I[Va(C-N-C)]/I[Vs(C-N-C)] increases.

The sensitivity of the TMAP of SAX to structural changes induced by counter

ion hydration and charge-density is a function of its architecture. The positive charge on

492

TMAP' is isolated on the central and stericaJly-protected nitrogen atom. When counter

ions are brought in contact or close proximity for charge pairing, the tetrahedraJ stmcture

of TMAP' deforms. The approximate charge separation distance for these solvent-

separated ion pairs were calculated by taking the difference between the ionic and

hydrated radii for TMA' and each halide and summing them. The resulting values are

given in Table 11.10. For T and perhaps Br", it is important to note that the charge

separation distance in Table 11.10 is an overestimate of the distance, as the hydration

potential of these ions is less than that of CI", resulting in perturbation of the hydration

sphere. Since TMAP' is sensitive to counter ion proximity, C-H and C-N vibrations are

most affected by counter ion interactions, analogous to observations made for TMA'

salts. Interestingly, however, although the intensities of the Va(C-N-C) and Vs(C-N-C)

modes are affected by counter ion interactions, the peak frequencies of these modes and

that of the v.,(C4N') at -755 cm"' are insensitive to counter ion interactions at this surface.

In contrast to the behavior of SAX observed with halides, for the oxyanions, the

VaCCHj)^ peak frequency does not change systematically with the chromatographic

selectivity. This mode is observed at a relatively constant frequency of-3040 cm"' for all

oxyanions except H2PO4" and P04"^. For these two anions the Va(CH3)N frequency is

observed at -3041.5 cm"'. In addition to the insensitivity of VaCCHj)^ frequency to

oxy anion selectivity it is also unaffected by anion charge. The Va(CH3)N mode for

TMAP " in the presence of aqueous H2PO4", 864"^ or P04"^ is observed at identical

frequencies.

493

Table. 11.10. Estimated charge-separation distance between TMAP^ and halides

Ion Hydrated Ionic Radius" Hydration Sphere Estimated Distance Radius'' (nm) (nm) Thickness (nm) Between Ion Pairs (nm)

TMAP' 0.367 0.347 0.020 -

cr 0.332 0.181 0.151 0.171

Bf 0.330 0.195 0.135 0.155

r 0.331 0.216 0.115 0.132

"Radius value for TMAP^ approximated by tetramethylammonium (TMA ) as given in reference 11.13.

494

For the halides, the Va(CH3)N frequency is affected by counter ion charge-density

and hydration. For the oxyanions, the Va(CH3)N frequency appears to be dependent on

anion hydrated radii. Most of the oxyanions have hydrated radii (rn) between 0.335 and

0.379 nm, with the exception of the considerably larger hydrated radii of H2PO4" and

PO4 ra (H2PO4') = 0.425 nm and rn (P04'^) ~ 0.450 nm. In the presence of aqueous

solutions of these phosphate ions, the Va(CH3)N is greater than in solutions of the other

anions.

One interesting fact about trimethylammonium anion exchange materials is their

high selectivity for CIO4", while the ionic potential, = z/r, for CIO4" is less than that for

lower selectivity anions such as T and CI". This is counter-intuitive considering

electrostatic interactions are thought to govern interactions in these ion chromatographic

systems. It must be that hydrophobic interactions dominate the selectivity for C104'over

r as C104' is known to be one of the most hydrophobic anions."^' In addition, these ions

are proposed to exist as solvent-separated ion pairs in the ion chromatographic case.

Since the hydrated radii of these anions are similar, there is little difference in ionic

potential between the hydrated anions suggesting that ion pair strength does not dominate

the selectivity behavior of these phases.

Moreover, the SAX Va(CH3)N frequency is influenced by factors other than ionic

potential or ion pair strength that can be elucidated from comparison of the effects of

oxyanions and halides for anions of comparable chromatographic selectivity. For the

high selectivity ions, I' and CIO4, the hydration enthalpy of and hydrated radii of each

495

are similar; however, the Va(CH3)x frequency is significantly lower for 1' than for CIO4".

The significant difference is that the charge of I" is isolated on one atom, whereas the

charge on CIO4' is diffused and shared between multiple atoms. This difliise nature of

the CIO4' charge, minimizes its ability to act like a contact ion pair with TMAP", and the

Va(CH3)N peak frequency is not reduced as it is in the presence of aqueous I". For the mid

selective anions, Br", NO3, HSO4", the hydrated radii and hydration enthalpies are

comparable and the VaCCH Ox peak frequency of SAX is the same in aqueous solutions of

these ions. However, for these anions, the hydration potentials are greater than those of

CIO4' and r, masking the effects of the diffuse charges on the oxyanions.

The observation that the Va(CH3)x peak frequency is independent of oxyanion

charge is rationalized by examining the surface charge-density of the SAX surface and

treating it as a charged planar substrate. Based on the capacity and surface area of SAX

particles, 0.6 meq/g and 521 mVg, respectively, the charge-density is estimated to be ~11

[iC/cm^. In order to reach electroneutrality, the counter ion charge-density on the

solution side of the surface must also be 11 }iC/cm^. For NO3', the number density is

only 24% of the total SAX surface area to reach electroneutrality, based on the hydrated

surface area of NO3'. For multiply charged anions, the surface area of anions required to

reach 11 jiC/cm^ decreases to —12% for S04"^ and ~8% for P04'^. All oxyanions in the

interface need for charge balance can be contained in first layer and the maximum

number density of anions is too small to observe an effect of anion charge state.

496

It is difficult to judge whether the value of l[Va(C-N-C)]/I[Vs(C-N-C)] changes

systematically with oxyanion selectivity, because only a limited subset of oxyanions for

which these value can be determined exists, due to spectral interferences. For the data

presented in Figure 11.8, no observable trends with selectivity are noted, consistent with

the observations made for the VafCHj)^ peak frequency. It is likely that hydration

enthalpy, hydrated radii, and diffuse charge of the oxyanions are important in giving rise

to changes in the value of Va(CH3)N peak frequency, but with the large standard deviation

in these data, it is difficult to discern these subtleties.

Raman Spectroscopy of Model Tetramethylammonium Salts and Aqueous Soluitions

As described previously, analyte ion interactions at these ion chromatographic

substrates are not exclusively electrostatic. Other considerations that influence these ion-

pairing interactions include the heterogeneity of charged sites, accessibility of charged

sites to solvent and counter ions, hydration characteristics of the surface-bound ions, and

hydrophobic or other nonspecific interaction of the analyte with the surface. To

understand the magnitude of these additional considerations, interactions between SAX

and the two classes of anions were modeled using TMA salt solutions. Unfortunately,

TMAH2PO4 was unavailable from any commercial sources, so it was excluded from this

series of model solutions. Therefore, investigation of interactions between TMA and

oxyanions are limited to TMANO3, TMAHSO4 and TMACIO4 salts and solutions.

497

Figures 11.9 and 11.10 show Raman spectra of tetramethylammonium halide and

oxyanion salts (Figure 11.9) and 0.3 M aqueous solutions of TMA' halides and

oxyanions (Figure 11.10). The corresponding peak frequencies in Table 11.11 are

slightly different from those for SAX (Table 11.1). For the TMA' salts, significant

spectral changes in TMA" modes were observed for the salts with different counter ions

that can be related to different TMA environments. These spectral changes included

shifts in the Vs(C4N') (-750 cm"'), Va(C4N^) (-950 cm"') and Va(CH3)N (-3030 cm"')

frequencies as well as changes in the intensity ratio of the Va(C4N ') to the Vs(C4N ' ) mode,

J[Va(C4N')]/I[Vs(C4N')]. These observations are consistent with those made by Pal et

al." for TMA' halide salts; however, such structural changes in TMA' had not been

previously reported for oxy anion salts. In contrast to the behavior of the TMA salts,

aqueous solutions of these salts exhibit no differences in TMA" peak frequencies or

intensities with anion as shown in Figures 11.9 and 11.10. It is important to note that the

low frequency modes in the nitrate and bisulfate anion spectra do not significantly

interfere with key TMA' bands for either the salts or solutions, but the Vs(C104) mode

does overlap with the Vs(CN4) mode.

Insight into the structural changes of TMA " when paired with different counter

ions compared to those observed for SAX can be realized by plotting the analogous

spectral indicators for TMA' as a function of anion selectivity. As an example. Figure

11.11 shows a plot of Va(CH3)N peak frequency as a function of anion selectivity for the

TMA" salts (Figure 11.11a) and 0.3 M aqueous solutions (Figure 11.1 lb). For the

£

3100 3300 2900 600 1400 1000

Waven umbers (cm"^)

Figure 11.9 Raman spectra of (a) TM ACI, (b) TMABr, (c) TMAI, (d) TMAHSO4, (e) TMANO3, and (f) TMACIO4. Integration times are 5-min for all spectra.

499

G <D +-> C

KH

2900 3100 3300 600 1000 1400

Wavenumbers (cm"^)

Figure 11.10 Raman spectra of 0.3 M aqueous solutions of (a) TMACl, (b) TMABr, (c) TMAI, (d) TMAHSO^, (e) TMANO3, and (f) saturated TMACIO4. Integration times are 30-min for all spectra.

500

Table 11.11. Peak frequencies and assignments for TMACl, TMABr, and TMAI

Peak Frequency (cm" ) Assignment"

TMACl TMACl (aq) TMABr TMABr (aq) TMAI TMAI (aq)

457 454 453 454 452 452 6a(CN^)

760 753 757 753 748 753 VS(C4N')

948 952 945 952 942 952 VA(C4N^)

1181 1173 1178 1174 1173 1174 r(CH3)

1192 1188 1181 r(CH3)

1287 1285 1282 r(CH3)

1402 1399 1398 6s(CH3)

1455 1455 1451 1455 1448 1455 SaCCHs)

1477 1472 1465 NR^

1483 1477 1469 5a(CH3)

2788 2783 2782 VS(CH3)FR

2841 2824 2835 2825 2832 2824 NR"

2875 2870 2867 VA(CH3)FR

2950 2990 2944 2989 2937 2989 VS(CH3)N

3018 3014 3010 VA(CH3)x

3026 3043 3021 3043 3013 3044 VA(CH3)N

"Assignments from reference 11.31. Vot reported.

3052

3048

"Q 3044

^ 3040

§ 3036

a 3032 fii

3028

3024

g" 3020

T" 3016

3012 0 2 4 6 8 10 12 14

Anion Selectivity. K_^

•

HI, Hso; cio;

CI ©

0Bf

©r T ' 1 1 1 ' 1 ' 1 ' r

3044 -

Anion Selectivity. K

Figure 11.11 frequency as a ftinction of anion selectivity for (a) tetramethylammonium (TMA') halide salts (O) and TMA" oxyanion salts (•) and (b) 0.3 M aqueous solutions of TMA^ halide salts (O) and TMA' oxyanion salts (•) . Error bars represent one standard deviation.

502

TMA halide salts, these results are analogous to those observed for S AX in halide

solutions where the Va(CH3)x peak frequency decreases with increasing anion size (i.e.

increasing selectivity). This behavior is due to the increased reduced mass of the

quaternary amine with a larger counter ion leading to a decrease in frequency of the

Va(CH3)K mode. In contrast, the VaCCHi)^ frequency of the TMA" oxyanion salts does

not follow this same trend of decreasing frequency with increasing anion selectivity, just

as was observed for SAX in aqueous solutions of oxyanions.

Figure 11.11b shows no difference in Va(CH3)N peak frequency paired with

different counter ions in solution, as well as no difference in frequency for the halide or

oxyanion pairs. For these ion pairs in solution, the susceptibility of TMA molecular

vibrations to size of the counter ion is muted by the hydration sphere and free space

between pairs.

Comparing the Raman spectral results of the model TMA' salts and solutions to

the spectra of SAX with different counter ions gives a greater understanding of

interactions at the SAX surface. It was initially hypothesized that the interaction of

anions at the SAX surface would mimic those of aqueous solutions of the TMA' salts.

However, based on the results from the TMA" salts and aqueous solutions, the

interactions of anions at the SAX surface are better modeled by the crystalline salts than

by the aqueous salt solutions. The ion-dependent behavior of the VaCCH ,)^ peak

frequency and I[Va(C-N-C)]/I[Vs(C-N-C)] for SAX in solution is analogous to the

behavior of these indicators for crystalline TMA halide salts, while the anion-invariant

frequency behavior of the Vs(C4N'), Vs(CN') and Va(CN') modes of SAX in aqueous

solutions is similar to the frequency behavior of the Vs(C4N') and Va(C4N') modes in

aqueous solutions of the TMA salts.

The crystalline salt-like behavior of SAX in solution is hypothesized to be a

manifestation of the surface-confined nature of TMAP^ as it impacts interfacial

characteristics such as surface hydration, surface heterogeneity and accessibility of

TMAP ' sites, and tree space between sites. Since TMAP sites are tethered to the silica

surface, less free space between each site exists than would exist in solution. In addition,

possible heterogeneity in the distribution of TMAP sites could also contribute to regions

of high surface charge-density that are more indicative of the crystalline salt environment

than a solution environment. Also, these silica substrates are known to have fractal

porosity,'^ which might render a significant portion of the pores inaccessible to solvent

or counter ions.

Isolute SCX-2

Raman Spectroscopy of Isolute SCX-2

Only a limited number of reports have been published on the structural

characterization of cation chromatography materials by FT-IR or NMR spectroscopy"^'"

" and none of these reports utilize Raman spectroscopy. Most of these reports are

simple experimental proof-of-concept descriptions showing that IR spectra of modified

polymer resins can be acquired, with the exception of work done by Gordon" and

504

Lowry" as mentioned earlier. Moreover, all of the spectroscopic characterization of

these materials had been done for sulfonated polymeric resins; no reports on the

structural behavior of other cation chromatographic particulate-based materials have

appeared.

Figure 11.12 shows the Raman spectra of Isolute SCX-2 (SCX-2) in the spectral

regions from 400 to 1500 cm"' and 2700 to 3200 cm"' for dry SCX-2 (Figure 11.12a) and

for SCX-2 in deionized water (Figure 11.12b) with the corresponding peak assignments

in Table 11.12. In both spectra, the low frequency regions are dominated by modes from

the ionic portion of the silica modifier, sulfonate and trimethylammonium bands, opposed

to the underlying propyl chain. In addition, very little difference is observed in the

Raman spectra of either material in the presence or absence of bulk water in the low

frequency region. In the high frequency region, the presence of the water envelope

starting at —3200 cm"' is the only qualitative difference observed in the spectra of this

material.

The Raman spectrum of dry SCX-2 contain only sulfonate modes and no bands

indicative of sulfonic acid species." These materials have been handled in ambient

conditions, and because the pKa of organic sulfonic acids is small (typically < 2), the

possibility of sulfonic acid deprotonation and ion exchange from moisture in the

atmosphere is likely. Thus, the initial state of the SCX-2 stationary phase is actually

sulfonate salts and not sulfonic acids.

• ?=-(

C o s

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm-')

Figure 11.12 Raman spectra of SCX-2 (a) dry and (b) in water at 20 °C. Integrations are 20-min in the low frequency region and 30-min in the high frequency region.

o

25000

506

Table 11.12. Peak frequencies and assignments for SCX-2

Peak Frequency (cm' ) Assignment

SCX-2 (Dry) SCX-2 (Hydrated)

488 488 5a(S03)

820 821 v(CS)

1045 1045 Vs(S03)

1125 1123 v(S02)

1314 1312 v(S02)

1417 1417 5(CH2)

"v = stretch and 5 = bend and/or scissor. Âssignments taken from references 11.33 and 11.44.

507

Figure 11.13 shows Raman spectra of SCX-2 in 0.2 M aqueous solutions of NaCl,

KCl, and CsCl. Raman spectra of SCX-2 in aqueous solutions appear to be identical to

each other, to dry SCX-2 and to SCX-2 in water. These observations are different from

those made previously for that dehydrated sulfonate-resins in which v(SO^) frequency

shifts were observed for polystyrene and fluoropolymer-based cation exchange phases.

These observations suggest that hydration of the SCX-2 material and the counter ions

plays an important role in their ability to interact at the SCX-2 surface. This behavior is

in contrast to the distinct VaCCHsV frequency shifts observed for SAX in going from the

dry state to the fully hydrated state and upon exchange of solution counter ions.

Alkali-Metal Interactions with SCX-2

To understand the effect of hydration and counter ion pairing on the structure of

surface-confined sulfonates, the vC SO?) peak frequency and v(S03) full-width at half

maximum (FWHM) were monitored in water and in aqueous solutions of alkali ions.

Figure 11.14 show plots of v(S03) frequency and FWHM as a function of the selectivity

for alkali metal cations. As was observed for SAX, selectivity values for particulate-

based ion exchange materials are not readily available in the literature, so values for

Dowex 50 (sulfonated polystyrene resin) were used."'^^ As alluded to above, the vCSOi)

frequency of SCX-2 is constant at —1045 cm"' under all experimental conditions. The

vCSOs) FWHM decreases by ~5 cm"' in water but, then remains constant upon changing

cation in aqueous solution.

10000 5000

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm-^)

Figure 11.13 Raman spectra of SCX-2 in 0.2 M aqueous solutions of (a) NaCl (b) KCl and (c) CsCl at 20 °C. Integrations are 20-min in the low frequency region and 15-min in the high

frequency region. KJ\ o 00

1047-

1046-

509

1045-§ g-(U

1044-! D

Ph O, 1043-1 O 00

1042-

21-

20--

'B 19-o

§ 18-

1 17-

16-O

15->

14-

13-

a

0.0 0.5 1.0 1.5 Relative Selectivity, K

"A,B

Relative Selectivity, K

2.0

AB

Figure 11.14 (a) v(S03) frequency and (b) v(S03)FWHM as a function of cation selectivity for dry SCX-2 (•), SCX-2 in water (•) and in 0.2 M aqueous solutions ofNaCl (A), KCl (•), CsCl (•). Error bars represent one standard deviation.

510

Hydration Effects

Others have reported vCSOj) frequency shifts of up to 15 cm' for organic

sulfonate salts in going from the crystalline solid into aqueous solution.'' ^ "

However, for SCX-2, the v(S03) frequency appears to be insensitive to changes in

hydration state. It is likely that despite attempts to completely dehydrate SCX-2 prior to

spectral analysis, this material has retained a significant amount of water, due to the silica

substrates. " Thus, if sufficient water is present to solvate the sulfonate, additional

water should not change the sulfonate structure. Additional attempts were made to

remove adsorbed water from these substrates; unfortunately, even slight heating of these

phases to 30-40 °C under vacuum either degrades or contaminates SCX-2 such that no

vibrational modes of TMAP" are observed in the Raman spectra.

Although the vCSO?) frequency does not change with additional water, the v(S03)

FWHM decreases from ~20 to 15 cm"' upon full hydration of SCX-2 as shown in Figure

11.14b. This behavior is consistent with results reported by Lowery for the hydration of

Nafion resins." The larger vfSO?) FWHM for SCX-2 in ambient indicates that,

although the majority of sulfonate groups are hydrated the surface still retains

considerable heterogeneity. Adding excess water to the material fully solvates the

sulfonate moieties, thereby causing the vfSOa) mode to narrow, even though v(S03)

frequency remains unchanged.

511

Counter Ion Effects

The structure of surface-confined propylsulfonate of SCX-2 does not change with

counter ion in aqueous solution as shown by the constant vCSO?) frequency and FWHM

in solutions of different cations (Figure 11.14.) While this behavior differs from that of

dehydrated Nation cation exchange resins," these results are consistent with

observations made previously for the effect of hydration on the vibrational behavior of

SCX-2. Thus, for cation exchange systems, hydration effects appear to dominate the

structure of surface-confined sulfonates. This situation is considerably different from that

for anion exchange systems in which surface TMAP " sites and their corresponding

solution counter anions are poorly hydrated compared to the comparable aqueous salts.

The hydrated radii and hydration enthalpies are significantly greater for cations (Table

11.13) than for anions producing greater charge-separation distances between the solution

cations and surface-bound sulfonates (Table 11.14.)

Raman Spectroscopy of Model Methanesulfonic Acid Solutions

Evidence from these experiments suggests that counter ion interactions at the

SCX-2 surface are insensitive to perturbation by surface effects such as heterogeneity and

accessibility of sulfonate sites, and free space between sulfonates. Thus, modeling these

interactions with aqueous solutions of the corresponding salts should be appropriate.

Figure 11.15 shows Raman spectra of methanesulfonic acid and 0.2 M aqueous

methanesulfonate. The corresponding vibrational mode frequencies and assignments are

512

Table 11.13. Hydration enthalpies, ionic radii, and hydrated radii of cations

Anion "Hydration Enthalpy "Tronic Radius (n) in "Hydrated Radius (rn) in nm (AHiivd) in kJ/mol nm

H" -1103 m 0.282

Na' -416 0.095 0.358

K' -334 0.133 0.331

Cs" -283 0.169 0.329

(CH3)4N^ -208 0,347 0,367

(C2H5)4N" -73 0.400 0.400

"Values from reference 11.56. ^lonic and hydrated radii from reference 11.13. °Not reported.

513

Table 11.14, Approximate charge-separation distance between alkali-metal ions and surface-bound sulfonate

Ion "Hydrated Ionic Radius Hydration Sphere Distance between Radius (nm) (nm) Thickness (nm) ion pairs (nm)

•U-SOa" 0.379 0.290 0.089 -

H' 0.282 - < 0.282 < 0.371

Na' 0.358 0.095 0.263 0,352

r 0.331 0.133 0.198 0.287

Cs 0.329 0.169 0.160 0,249

"Ionic and hydrated radii from reference 11.13. ^Radii for surface-bound sulfonate are approximated using radii for S04"^ from reference 11.13,

75000 v(CS) 40000 viCH,

Va(CH,) v(SO-H)

viSO,) 25000

5.(SO,)

I t I I I I I I I 400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cnr')

Figure 11.15 Raman spectra of (a) methanesulfonic acid and (b) 0.2 M aqueous methanesulfonate. Integrations are 5-min in the low frequency region and 2-min in the high frequency region for methanesulfonic acid and 15-min for 0.2 M aqueous methanesulfonate.

515

given in Tables 11.15 and 11.16, respectively. Figure 11.16 shows Raman spectra of 0.2

M aqueous solutions of methanesulfonate with 0.2 M aqueous NaCl, KCl, and CsCl.

Since the pKa of methanesulfonic acid is small (~2), it is mostly ionized at a

concentration of 0.2 M, and should be paired with alkali cations. Raman spectra for

methanesulfonic acid and 0.2 M aqueous methanesulfonate are consistent with those

reported in the literature." " For methanesulfonic acid, the presence of v(SO-H)

(-901 cm') and 5(0H) (-1122 cm"') modes, indicative of the protonated form, are

distinctly different than modes for aqueous solutions of methanesulfonate and any

spectrum of SCX-2 presented here. This observation further supports the contention that

SCX-2 is not modified with propylsilylsulfonic acid upon receipt from the manufacture,

but is modified with a propylsilylsulfonate salt.

In a manner analogous to the approach used for SCX-2, the vCSOs) frequency and

FWHM are monitored as a function of aqueous counter ion as shown in Figure 11.17. As

is observed for SCX-2, the v(S03) frequency for aqueous methanesulfonate is constant at

-1050 cm"' and independent of counter ion size or hydration. In addition, the v(S03)

FWHM is -8 cm"' for aqueous methanesulfonate regardless of counter ion, making these

solutions adequate models for SCX-2 in solution.

Tetraalkylammonium Interactions with SCX-2

In addition to alkali counter ions, the effect of tetraalkylammonium ions (TAA")

on the structure of SCX-2 was investigated for two reasons; (1) to see if poorly-hydrated

Table 11.15. Peak frequencies and assignments for methanesulfonic acid

Peak Frequency (cm"') "''Assignment

O

00

r(S-OH)

504 r(S02)

535 6(S02)

• 772 v(CS)

901 v(S-OH)

985 NR'

1122 5(0H)

1170sh Vs(S02)

1344 VaCSOa)

1418 8a(CH3)

2945 VS(CH2)

3033 Va(CH2)

= stretch, r = rock, and 8 = bend and/or scissor, Âssignments taken from reference 11.44. ^^^Fot reported. ''sh = shoulder.

517

Table 11.16. Peak frequencies and assignments for aqueous methanesulfonate at pH 1.04

Peak Frequency {cr^} ^•''Assignment

555 6U(S03)

786 v(CS)

968 r(CH3)

1050 Vs(S03)

1197 Va(S03)

1424 5a(CH3)

"v = stretch, r = rock, and 6 = bend and/or scissor. ''Assignments taken from reference 11.44.

CO

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm-^)

Figure 11.16 Raman spectra of 0.2 M aqueous solutions of methanesulfonate and (a) NaCl,(b) KCI, and (c) CsCl. Integrations are 15-min for all spectra.

©o

519 1053

0.5 1.0 1.5 Relative Selectivity, K

Relative Selectivity, K XB

Figure 11.17 (a) v(S03) frequency and (b) vCSOj) FWHM as a function of cation selectivity for 0.2 M aqueous methanesulfonate (•) and with 0.2 M aqueous solutions ofNaCl (•), KCl (•), CsCl (•). Error bars represent one standard deviation.

520

cations have an effect on the structure of surface sulfonates and (2) to see if the structure

of the TAA' is affected by interactions with SCX-2. Figure 11.18 shows Raman spectra

of crystalline TMABr (tetramethylammonium bromide), 0.2 M aqueous solutions of

TMABr, and SCX-2 + 0.2 M aqueous solutions of TMABr in two spectral regions from

400 to 1500 cm"' and 2750 to 3300 cm"'. A similar series of Raman spectra were

acquired for crystalline TEABr (tetraethylammonium bromide), 0.2 M aqueous solutions

of TEABr and SCX-2 + 0.2 M aqueous solutions of TEABr and shown in Figure 11.19.

Corresponding vibrational modes and assignments for both TEABr and aqueous TEABr

are given in Table 11.17.

While considerable spectral overlap occurs between SCX-2 and both TA.A*

solutions, information from key spectral indicators including the v(S03) frequency of

SCX-2, the Va(CH3)N frequency of TMA", and the Va(CH3) frequency of TEA is readily

available from these spectra. The v(S03) frequency of SCX-2 in 0.2 M aqueous TMABr

is 1044.8 ± 0.25 and in 0.2 M aqueous TEABr is 1044.8 ± 0.15. Thus, even when

interacting with TAA that are poorly hydrated, the v(S03) frequency is identical to that

observed in aqueous solutions of alkali cations. This observation supports the previous

conclusion that sulfonate hydration dominates the surface structure of SCX-2.

An additional interesting consideration is the effect of interaction with SCX-2 on

the structure of TAA' in aqueous solution, since the studies described above on SAX

have shown the quaternary amine structure to be sensitive to counter ion interactions in

aqueous solution. Table 11.18 shows the Va(CH3)N frequency (-3040 cm"') for TMABr,

40000 50000 Va(CH3)N viQN)

8,(CH0 viQN v r C K

SiQN)

i

25000

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm"^)

Figure 11.18 Raman spectra of (a) crystalline TMABr, (b) 0.2 M aqueous TMABr and 0.2 M aqueous TMABr + SCX-2 at 20 °C. Integrations are (a) 2-min low frequency and 3-min high frequency, (b) 15-min low frequency and 10-min high frequency, and (c) 30-min.

Ui to

20000 v(C-C)

r(CH,) 5(CH,)N

T(CH,)

v.CCH,) 25000

v/CHj) I v.CCH,)

• fsH Xf l G

30000

400 600~~80^ lo'oo 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm"^)

Figure 11.19 Raman spectra of (a) crystalline TEABr, (b) 0.2 M aqueous TEABr and 0.2 M aqueous TEABr + SCX-2 at 20 "C. Integrations are (a) 2-min low freq, and 3-min high frequency, (b) 15-min low frequency and lO-inin high frequency, and (c) 30-min.

Table 11.17. Peak frequencies and assignments for TEABr

Peak Frequency (cm" ) Assignment®'

TEABr TEABr (aq)

421 420 Sa(C4N^)

666 Vs(-C-N)

678 675 Vs(-C-N)

787 790 r(CH2)

908 900 v(C-C)

1003 1004 r(CH3)

1069 1072 Va(C-C)

1114 1120 r(CH:,)

1176 1178 r(CH3)

1264 mc

1302 1303 X(CH2)

1392 1394 8(CH2)N

1440 1449 5S(CH2)

1464 1465 5a(CH3)

1498 5s(CH3)

2901 2903 VS(CH3)

2944 2954 VsCCHa)

2971 Va(CH3)

2997 3002 VaCCHs)

"v = stretch, r = rock, x = twist, and S = bend and/or scissor. Âssignments taken from reference 11.59. •""Not reported.

524

Table 11.18. Peak frequencies of Va(CH3)N and VaCCH?) modes of IMA and TEA ' (cm"')

Sample condition Va(CH3)N (TMA ) VaCCHs) (TEA"*")

Crystalline TAABr Salts 3021.73 + 0.05 2997.64 ± 0.28

200 mM Aqueous TAA" 3044.16 + 0.09 3002.19 ± 0.20 Solutions

SCX-2 + 200 mM 3043.06 ± 0.11 3000.55 ± 0.26 Aqueous TAA' Solutions

525

and the VaCCHs) frequency (-3000 cm ') for TEABr in the crystalline salt form, for 0.2 M

aqueous solutions, and for 0.2 M aqueous solutions + SCX-2. For TMA", the

frequency decreases by only 1 cm"' in 0.2 M aqueous solution in the presence of SCX-2,

while the VaCCH?) frequency of TEA' decreases by 2 cm"' in aqueous solutions with

SCX-2. These decreases in peak frequency, albeit small, may indicate slightly stronger

interaction with the surface-confined sulfonates of SCX-2 than with solution Br".

However, this frequency shift is very close to the resolution of the spectrometer and

cannot be discerned with statistical certainty.

Interactions of TAA* with SCX-2 were modeled using aqueous solutions of

methanesulfonate. Raman spectra of 0.2 M aqueous solutions of TMABr and TEABr

with methanesulfonate are shown in Figure 11.20. While there is significant spectral

overlap of TAA " and methanesulfonate modes, bands that can be used as indicators of

structure, including v(S03), Va(CH3)x, (TMA'), and VaCCHs) (TEA ), are accessible.

Table 11.19 gives the fi-equencies of these modes and the v(S03) FWHM for

TAA '/methanesulfonate aqueous solutions. The v(S03) frequency and FWHM in both

TMA and TEA" solutions are -1050 and ~8 cm"', respectively. These values are

identical to those for methanesulfonate in alkali-metal ion solutions. The Va(CH3)N and

Va(CH3) frequencies for TMA' and TEA ' in aqueous methanesulfonate solutions,

respectively, are the same as those for 0.2 M TMABr and TEABr solutions and 1-2 cm"'

higher than for TAA* with SCX-2 (Table 11.19). While these values are statistically

250000

400 600 800 1000 1200 1400 2800 2900 3000 3100 3200 3300

Wavenumbers (cm-')

Figure 11.20 Raman spectra 0.2 M methanesulfonate with (a) 0.2 M aqueous TMABr and (b) 0.2 M aqueous TEABr at 20 "C. Integrations are 15-min for all spectra.

Ni ON

527

Table 11.19. Peak frequencies (in cm"^) of key modes for tetraalkylammonium methanesulfonate solutions

Condition v(S03) vCSOa) FWHM Va(CH3)N (TMA ) VaCCHj) (TEA~)

0.2 M TMABr 1049.50 ±0.03 8.23 ± 0.05 KA NA

0.2 M TEABr 1049.43 +0.03 8.47 ± 0.03 NA NA

0.2 M NA" NA 3043.8 ±0.1 3002.5 ± 0.25 Methanesulfonate

"Not applicable.

528

different, the fact that these frequency shifts are near the resolution of the spectrometer

make them difficult to interpret them with any confidence.

Comparison of SAX and SCX-2 Raman Studies

The differences in sensitivity of surface-bound sulfonate and TMAP ' to solution

counter ion interactions are a result of differences in charge distribution characteristics of

these ions, the hydration differences between SAX and SCX-2 surfaces, and hydration of

anions and cations in general. The positive charge of TMAP' is isolated on the nitrogen

atom in the center of the ion, surrounded by three methyl goups, while the negative

charge on the deprotonated sulfonate ion is diffuse on the periphery of the ion, shared

equally by three equivalent oxygen atoms. In light of these charge distributions it is

reasonable that the TMAP' should be more sensitive to counter ion interactions,

regardless of hydration characteristics, because of the diffuse nature of the sulfonate

charge.

In addition to charge distribution differences, hydration variations between the

anion and cation exchange systems also contribute to the disparity in behavior. For the

anion exchange system, the surface-confined TMAP" and aqueous halides and oxyanions

are poorly hydrated, with generally smaller hydration enthalpies and hydrated radii than

alkali-metal cations. As a result, the charge separation of ion pairs in the cation exchange

system is greater than that in the anion exchange system, as shown in Tables 11.10 and

529

11.14. In fact, for SCX-2, the charge separation between all the alkali ions and the

sulfonates is greater than the separation for all the anions and TMAP" for SAX.

It was shown for the SCX-2 phase that surface water had a significant effect on

the structure of what is referred to as "dry" SCX-2, to the point where dry SCX-2 is

spectroscopically indistinguishable from fully hydrated SCX-2. However, surface water

does not appear to have any effect on the structure of TMAP ', despite that these phases

were prepared on the same silica substrates. Again, this discrepancy between the phases

is a result of the hydration differences between TMAP and sulfonate. It is reasonable

that the structure of hydrophobic TM AP would be completely unaffected by adsorbed

surface water on the silica substrates, while the sulfonate structure would be greatly

impacted by surface water.


Interactions between monoatomic and polyatomic anions and cations at the

surface of anion and cation exchange silica materials are presented. Structural changes in

surface-confined TMAP' ions upon interaction with different counter ions are manifested

by changes in the Va(CH3)N frequency and the I[Va(C-N-C)]/I[Vs(C-N-C)]. These spectral

indicators are markers of anion selectivity. The VaCCHa)^ frequency increases and the

I[Va(C-N-C)]/I[Vs(C-N-C)] decreases systematically with anion selectivity for a series of

halides and oxyanions. Modeling these interactions with tetramethylammonium salts and

solutions shows that both electrostatic interactions and the convolution of surface effects

530

greatly influence the structure of surface-bound TMAP ions. In addition, the changes in

structure and hydration of solution oxyanions changes upon interaction with the SAX

surface and are also largely dependent on surface effects.

Structural changes in surface-confined sulfonates do not occur under hydrating

conditions or in aqueous solutions of alkali or tetraalkylammonium ions. The invariable

sulfonate structure is a combination of the facts that the SCX-2 surface is hydrophilic and

that inorganic cations are strongly hydrated in aqueous solutions. The interactions at the

SCX-2 surface are adequately modeled using aqueous solutions of methanesulfonate and

are not influenced by the surface-confined nature of the sulfonate sites.

Future directions for these studies include doing similar experiments with other

ions of environmental or biological interest such as lanthanide and actinide oxides, amino

acids and small peptides. In addition, investigating the structures of these surface-bound

ions in solvents such as methanol, acetonitrile, and THF or mixtures of these with

aqueous solutions is also important, as these solvents are common modifiers in ion

chromatographic experiments. Experiments could also be done on other ion exchange

phases including polystyrene or perfluorinated polymer resins, weak anion exchange

phases (primary and secondary amines), or weak cation exchange phases (carboxylic

acids).

Specifically for the SCX-2 phase, undertaking experiments in which cations are

first exchanged in aqueous solutions and then spectroscopicaily interrogated after drying

these phases would be of great benefit to determine if the surface sulfonate structure is

531

affected by counter ion exchange under dry conditions, analogous to previous IR

spectroscopy work."

532

Chapter 12

Preliininary Studies on the Characterization of Other Commercial Stationaiy

Phases for Reversed-Phase Liquid Chromatography

Introduction

While n-alkylsilane stationary phases in reversed-phase liquid chromatography

(RPLC) and solid phase extraction (SPE) are the most common, numerous additional

commercial phases, modified with different functional silanes, have also proven to be

quite useful for certain separations applications. These phases have a variety of terminal

functional groups including phenyl, cyano, alcohol, and primary or secondary- amines and

offer improved selectivity and separation efficiency for certainsolutes relative to alkyl

phases. ' ' ' For example, Kim and coworkers ' determined the selectivity for

hydrogen-bonding acids and bases to be higher on cyano- and amino-phases than on n-

alkyl phases in mobile phases of high water content. This behavior was proposed to be

due to the ability of the nitrogen atom of the cyano group to hydrogen-bond with analytes

and retain them more strongly than n-alkyl phases in which only weaker dispersion

interactions are possible. Another example is the improved selectivity of aromatic

compounds using phenyl stationary phases over aliphatic phases. Gelsema and De Ligny

reported slightly better selectivities for aromatic carboxylic acids over aliphatic

carboxylic acid homologues on

533

12 1 phenyl phases. ' Furthermore, Zhao and Carr have reported greater selectivity toward

large molecular weight polycyclic aromatic hydrocarbons (PAHs) than alkylbenzene

homologues on phenyl-terminated stationary phases.'^' In both of these examples,

improved selectivities for aromatic compounds on phenyl phases was attributed to strong

Tt-% interactions between these aromatic solutes and the stationary phase.

Throughout this Dissertation, the importance of characterizing the structure of

stationary phases under chromatographic parameters has been emphasized for

understanding retention mechanisms in chromatography. In the work presented in this

Chapter, Raman spectroscopy was used to determine solvent interactions with stationary

phases of functionality other than n-alkylsilane phases.

Goals of this Work

The goal of this work was to use Raman spectroscopy to characterize the

interactions between cyano- and phenyl-based stationary phases and solvents used in

RPLC. While solvent interactions with aliphatic phases are restricted to dispersion

interactions, a greater variety of chemical interactions are possible with cyano- and

phenyl-moditied phases including hydrogen-bonding, acid-base, and %-% interactions.

The role of each of these types of interactions which were investigated for each system.

Much like the work presented in Chapter 6, this work will act as the foundation for more

in-depth investigations of these phases under other chromatographic conditions.

534

Methodology

I solute Phenyl and CN were gifts from International Sorbent Technologies (UK).

The properties of these phases are given in Table 12.1 and were discussed in Chapter 2.

Perdeuterated water (99.9%) was purchased from Aldrich. Additional perdeuterated

solvents used in these experiments including methanol (99.8%), acetonitrile (99.8%),

acetone (99.9%), tetrahydrofuran (TIIF) (99.5%), chloroform (99.8%), benzene (99.5%),

toluene (99.8%), and heptane (99%) were obtained from Cambridge Isotope

Laboratories, Inc. "C-labeled acetonitrile (CHi-'^C^N) was purchased from Cambridge

Isotopes, All solvents were used as received.


material into a 5-mm dia NMR tube; 200 pL of solvent was then added. Samples were



Coherent Verdi Nd: YV04 laser on a Spex Triplemate spectrograph as described in

Chapter 2. Slit settings of the Triplemate were (0.5/7.0/0.150) mm for all experiments.

All samples were equilibrated at the appropriate temperature for at least 30 min prior to

spectral acquisition. Integration times for each spectrum are provided in the figure

captions. Raw spectral data were analyzed and processed as described in Chapter 2.

535

Table 12.1. Properties of Isolute stationary phases

Isolute Phase"

Particle Diameter

(|im)

Surface Area

(m^/g)

Pore Diameter O

(A)

Surface Coverage (}imol/m^)

CN 40-70 mC 60 1,2

Phenyl 40-70 NR 60 1.0

"CN = cyano-terminated silane precursor. Phenyl = phenyl-terminated silane precursor. ^Not Reported.

536

Raman Spectroscopy of Isolute CN SPE phases

Raman spectra of Isolute CN (propylcyano-terminated phase, referred to as IST-

CN) were acquired in the v(C=N) region between 2150 and 2350 cm"' the and v(C-H)

regions between 2750 and 3050 cm"' for dry IST-CN and in the presence of perdeuterated

solvents. These spectra are shown in Figure 12.1, with the corresponding peak

frequencies and vibrational mode assignments provided in Table 12.2. As shown in

Chapters 4 and 11 and in previous reports,'^ Raman spectra can be acquired for

modified Isolute silica substrates without significant spectral interference from silica

impurities. Additionally, these spectra are free from solvent spectral interference in both

regions, since the '^C-labeled isotope of acetonitrile (CHj-'^C'^N) was used to acquire the

spectrum of IST-CN in acetonitrile in the v(C=N) region to avoid interference between

the two v(CN) modes (Figure 12. lb).

Subtle changes are observed in the Raman spectrum for IST-CN in the presence

of solvent. The v(CN) frequency shifts in the presence of different solvents and becomes

slightly asymmetric indicating multiple chemical environments. In addition, subtle

changes are observed in peak trequencies of the v(CH2) modes of the propyl chain

suggesting its solvation.

Interactions Between Solvent and Isolute CN

Previous reports have shown that the frequency of the v(CN) mode of nitriles is

12 10 12 11 12 12 sensitive to chemical environment. Swanson ' ' ' and Shriver ' have shown

537

V,(CH2) + â(CH2)csN

Vg(CH2)c=N v(CN)

c 0)

2150 2200 2250 2300 3000 2800 2900

Wavenumbers (cm'')

Figure 12.1 Raman spectra of Isolate CN in (a) air, (b) acetonitrile, (c) methanol, (d) water, (e) toluene, (f) benzene, (g) chloroform, (h) heptane. Acquisition times for all spectra are 2 min.

538

Table 12.2. Peak frequencies and assignments for Isolute CN

Peak Frequency (cm"' ) Assignment'

2251 — v(CN)

2899 V.(CH2), Vs(CH2)C N

2932 Va(CH2)c^N

2962 VaCCHs)

"Assignments taken from reference 12.7.

539

the v(CN) ifrequency to increase upon interaction with electopMlic Lewis acids. Electron

density is donated from the lone pair on the nitrogen atom of the cyano group to the

Lewis acid resulting in lengthening and weakening of the C=N bond, and increasing its

vibrational frequency. Conversely, a decrease in v(CN) frequency is observed upon

interactions of nitriles with Lewis bases.Electron density from the base is

donated to the electron-deficient carbon atom of the cyano group, decreasing its

vibrational frequency.

For IST-CN, the v(CN) frequency is monitored in the presence of a variety of

solvents. As described in Chapters 4 and 6, these solvents vary in size, shape,

polarizability, dipole moment, hydrogen-bonding accepting/donating ability, and

hydrophobicity, and have been quantified by Kamlet'^ by defining solvent

solvatochromic parameters, Solvatochromic parameter TC*, which is the dipolarity

normalized to polarizability, a, hydrogen-bond accepting ability, and |3, hydrogen-bond

donating ability are listed in Table 12,3, Figure 12.2 shows a plot of v(CN) frequency as

a function of solvent solvatochromic parameter n*. The horizontal dashed line indicates

the v{CN) frequency for dry IST-CN as a guide. The v(CN) peak frequency of IST-CN

1 * * 1 shifts from 2251 cm' m air to 2253 cm' in water, is approximately the same in air as in

hexane, methanol, and chloroform, decreases slightly in benzene and toluene, and

decreases by 3 cm"' in acetonitrile. While shifts in the v(CN) frequency are subtle and

approach the spectral bandpass of the spectrometer, they are significant and can be

rationalized by solvent properties.

540

Table 12.3. Solvatochromic parameters®, %*, a, and j3 for solvents

Solvent K* a (3

Water ~ ~ 1.09 ~ ols ~ TaT


Acetone 0.71 0.48 0.06

Methanol 0.60 0.62 0.93

Benzene 0.59 0.10 0

Chloroform 0.58 0 0.44

Heptane -0.08 0 0

Toluene 0.54 0.11 0

'As described by Kamlet et al.'^^'

541

B o

¥ g

2254-

2253

2252'

g- 2251 -2

Heptane

UH

2250-

} (U

PH

g 2249' W

2248-1

Water H

r! Chloroform

Methanol

Toluene • I#

Benzene

Âcetonitrile

0.0 0.2 0.4 0.6

n*

0.8 1.0 1.2

Figure 12.2 v(CN) frequency as a function of solvatochromic parameter %* for Isolute CN in air, heptane, toluene, chloroform, benzene, methanol, acetonitrile, and water. Error bars represent one standard deviation.

542

The increase in the v(CN) frequency in water relative to in air is a result of

hydrogen-bonding of water molecules with the surface-bound cyano group. Hydrogen-

bonding through the nitrogen atom of the cyano group weakens the C=N bond, increasing

its vibrational frequency. In poorer hydrogen-bonding solvents, methanol, hexane, and

chloroform, no change in v(CN) frequency is observed. In the presence of benzene and

toluene, the slight decrease in v(CN) frequency relative to JST-CN in air is attributed to

Ti -TC interactions between the aromatic rings and the cyano group where excess electron

density of the aromatic ring is partially donated to the cyano group.

In acetonitrile, the v(CN) frequency of IST-CN is shifted to lower frequency

relative to air. It is hypothesized that the most likely interaction between acetonitrile and

IST-CN would be through dipole-dipole dimerization of the cyano groups, because of the

substantial evidence for such interactions in the literature.'' However, the

fonnation of such acetonitrile dimers has been shown to lead to an increase in vibrational

frequency. ' ' '' Thus, interactions between IST-CN and bulk acetonitrile must be

different from this end-on dimer formation. The shift in v(CN) to lower frequency

suggests Lewis base coordination with acetonitrile between the lone pair on the nitrogen

atom of acetonitrile and the carbon atom of the surface-bound cyano group.

The frequencies of the v(C-H) modes are also affected by solvation of IST-CN.

Figures 12.3, 12.4, and 12.5 show plots of VH(CH2) + Vs(CH2)c N, Va(CH2)c N, and

VA(CH2) peak frequency, respectively, as a function of TT*. All modes are shifted to lower

frequency in the presence of all solvents, with the exception of water and acetonitrile. In

543

^ 2901.0

I f; 2900.5 0

1 2900.0 cr

^ 2899.5

2899.0

^ 2898.5

54 2898.0 >

i 2897.5

H; 2897.0 ^ 0.0 0.2 0.4 0.6 0.8 1.0 1.2

n*

Acetonitrile

Air Water _||. _B—

Chloroform • Methanol

Sm± Benzene Heptane Toluene

Figure 12.3 Vg(CH2) + v^(CH2)£.=j^ frequency as a function of solvatochromic parameter TT* for Isolute CN in air, heptane, toluene, chloroform, benzene, methanol, acetonitrile, and water. Error bars represent one standard deviation.

544

S w

o c 3 CT" <D

(D PH

a

U

2933

2932

2931

2930-1

2929

2928-

2927-

2926

2925

Air

Heptane

Acetonitrile

i Water

Methanol

Chloroform

Toluene Benzene

n*

Figure 12.4 frequency as a function of solvatochromic parameter %* for Isolute CN in air, heptane, toluene, chloroform, benzene, methanol, acetonitrile, and water. Error bars represent one standard deviation.

545

2962

"C 2961 B

^ 2960 I 1 2959

2 2958 JJH

"i 2957 Cu ^ 2956

2955 >

2954

0.0 0.2 0.4 0.6 0.8 1.0 1.2

II*

Figure 12.5 VgCCHj) frequency as a function of solvatochromic parameter TT* for I solute CN in air, heptane, toluene, chloroform, benzene, methanol, acetonitrile, and water. Error bars represent one standard deviation.

Air..__

Acetonitrile

S Methanol

^ Chloroform

j| a I Benzene Toluene

1 1 1 1 1 1 1 1 1 1 1 r-

546

water, the frequencies of these modes are the same as they are in air. In acetonitrile, the

V/CH2) + VS(CH2)C=N and V3(CH2)ceN modes are shifted to higher frequency than in air.

In nonpolar solvents, the decrease in frequency observed for these modes indicates

solvation of the propyl chain, which puts greater numbers of molecules in close proximity

relative to IST-CN in air increasing the C=N reduced mass. Water molecules do not

undergo strong interactions with the hydrophobic propyl chain, and thus, the frequencies

of these modes are unaffected. Interactions between water molecules are IST-CN are

restricted to hydrogen-bonding through the cyano-group.

In acetonitrile, the Vs(CH2) + v.s(CH2)c x and Va(CH2)c m modes are shifted to

higher frequency. This observation suggests that solvation of IST-CN by acetonitrile is

restricted to the cyano group, resulting in an observable impact on only the frequency of

the methylene group directly attached to the cyano group but not the ValCH^) modes of

the methylene groups at the proximal end of the propyl chain.

Raman Spectroscopy of Isolute Phenyl SPE Phases

Raman spectra for Isolute Phenyl (referred to as IST-PH) are acquired in the ring-

breathing region between 900 and 1250 cm"' and in the v(C-H) region between 2850 and

3100 cm"' and are shown in Figure 12.6, with the corresponding peak frequencies and

vibrational mode assignments provided in Table 12.4. Some spectral interference from

solvents occurs in the low frequency region for IST-PH in hexane, benzene, and

547

V,(CH)

1000 1100 1200

Wavenumbers (cm"\)

900 2900 3000 3100

Figure 12.16 Raman spectra of Isolute Phenyl in (a) air, (b) acetonitrile, (c) water, (d) benzene, (e) heptane, and (f) chloroform. Acquisition times are 1 min in the low frequency region and 2 min in the high frequency region.

Table 12.4. Peak frequencies and assignments for Isolute Phenyl

Peak Frequency (cm"^) Assignment"

1000 v(C-C)ring

1032 6.p(CH), v(C-C)

1134 v(C-Si)

1158 5s(CH)

1194 5s(CH)

2906 Va(CH)

2965 Va(CH)

2976sh Va(CH)

3060 Vs(CH)

"Assignments taken from reference 12.20-12.22.

549

acetonitrile; however, the predominant v(C-C) ring breathing mode (1000 cm'^) and other

v(C-C), 5(CH), and ai! v(CH) modes are observable without interference.

Interactions Between Soivent and I solute Phenyl

Measurable frequency shifts in the V;,(CH2) (2905 cm'' ) and Vs(CH2) (3059 cm ')

modes for IST-PH in the presence of solvent. Figure 12.7 shows a plot of Va(CH)

frequency as a function of n* and Figure 12.8 shows a plot of Vs(CH) frequency as a

function of n* for IST-PH in air and in solvent. The horizontal dashed lines on each

graph represent the Va(CH) and Vs(CH) frequencies in air as a reference. In general, the

Va(CH) and Vs(CH) frequencies are shifted to lower frequency in nonpolar solvents

relative to in air. Polar solvents have less of an effect on the frequencies of these modes

than nonpolar solvents, with the exception of methanol, which causes a significant shift

of the Vg(CH) mode to lower frequency.

IST-PH is a low surface coverage phase (T = 1.0 [.imol/m^), and assuming a

relatively homogeneous distribution of functional groups on the surface, the phenyl rings

are largely decoupled in air, having little or no interaction with neighboring molecules.

In the presence of nonpolar solvents such as heptane, benzene, toluene, and chloroform,

the hydrophobic IST-PH surface is well-solvated. Increased solvation of the surface puts

a large number of molecules around each surface-bound phenyl ring. Thus, increasing

550

2906'

8 2905 -

g

& <D

a> 0-.

Air - - A

Water

2904-

£ 2903 U I

i

Acetonitrile

Chloroform IT A Benzene T.

2902 0.0 0.2 0.4 0.6

n*

0.8 1.0 1.2

Figure 12.7 Va(CH) frequency as a function of solvatochromic parameter k* for Isolute Phenyl in air, heptane, chloroform, benzene, acetonitrile, and water. Error bars represent one standard deviation.

551

s o

3060

3059-

1 3058

£ B 3057 PH

u

">" 3056

T Air i

T Air Water 1 " ~"" iiiHliii — — — — — — — — — — — — — — — — — — — —

1

T Acetonitrile

i

Chloroform]

J Heptane J ^ Benzene

J.

0.0 0.2 0.4 0.6 0.8 1.0 1.2

n*

Figure 12.8 v,(CH) frequency as a function of solvatochromic parameter n* for Isolute Phenyl in air, heptane, chloroform, benzene, acetonitrile, and water. Error bars represent one standard deviation.

552

the number of molecules interacting with the phenyl groups increases the reduced mass

and the Va(CH) and Vs{CH) modes shift to lower frequency.

In the presence of polar solvents such as water, acetonitrile, and methanol, IST-

PH is generally less well-solvated than in nonpolar solvents. Water molecules have very

little interaction with surface-bound phenyl groups and have no impact on the frequency

of either the Va(CH) and Vs(CH) modes relative to air. However, methanol and

acetonitrile solvate the phenyl rings better than water, which leads to a slight decrease in

the Va(CH) frequency, and a larger shift of the Vs(CH) mode to lower frequency.

For IST-PH, no observable shift in the v(C-C) ring breathing mode is detected in

the presence of these solvents. This is due to the fact that the phenyl ring is directly

attached to the silica substrate and not as flexible or susceptible to orientational changes

as if it were connected to the substrate via an aliphatic linker. In addition, orientation

changes are likely to be too subtle, leading to shifts in v(C-C) frequency that are below

the resolution of the spectrometer.


Solvent interactions with cyano- and phenyl-stationary phases are determined by

Raman spectroscopy. Shifts in vibrational mode frequencies of surface-bound cyano and

phenyl groups in the presence of solvent are attributed to Lewis acid, Lewis base, or van

der Waals interactions with solvent molecules. In the previous investigations of solvent

effects on n-alkyi stationary phases described in Chapters 4 and 6 through 10, the only

553

information available in the spectral data were iedicators of conformationa} order of the

al!c}'l chains to determine solvation elfects. For stationary phases with more functionaiity

such as cyano- and phenyl-terminated ones, more spectral indicators of solvent

iriteraction exist.

Future directions for this work include investigating solvent interactions with

stationary phases containing different terminal functional groups such as alcohols and

amines. In addition, studying these phases along with the cyano- and phenyl phase in

binary solvent systems and in the presence of analytes are important experiments since

these better conditions mimic real chromatographic or extraction conditions than single

solvents.

554

Chapter 13

Alkylsilane-Based Stationary Phases via Controlled Solution Modifier

Techniques: Synthetic Approach, Characterization, and Chromatographic

Performance

Introduction

The use of alkylsilanes to modify particulate supports for liquid chromatography

was first described by Abel et al. in 1960,^ '' and quickly became the industry standard

for analytical separations with contributions from Unger'^ ^ '^' and Kirkland." ^"'^ The

scheme for the reaction between hexadecylsilane and Celite 545 (a silica-based mineral)

used by Abel is as follows:

1) Dried petroleum spirit

CeUte 545 + CisHjjSiClj (b p. 60-80 C) C,6-modlfied (y 2) Shake 5 h stahonaiy phase

While the synthesis of modem stationaiy phases is more sophisticated, using dry solvents

and high purity reagents, the basic hydrolysis and condensation chemistry has not

changed in 40 years. Scheme 2 shows the generalized synthesis of common alkylsilane

stationary phases that has appeared in the literature since 1970:

555

1) Dried toluene.

Silica particles + n-alkylchlorosilane — ^—p. n-alkyl-modified ^2)

2) Reflux 4-100 h

However, it has been shown that in order to specifically tailor the properties of the

stationary phases, this simple reaction scheme must be modified. For example, high

surface coverage polymeric stationary phases have been shown to be highly selective and

quite useful for the separation of shape-constrained isomers including polycyclic

aromatic hydrocarbons (PAHs) and carotenoids.^'^'^"^'^^ This was first described by

Hennion et al.^'" in a study showing that the separation of anthracene and naphthalene

improves using higher coverage stationary phases. Since this work, the majority of

investigations on shape selectivity of stationary phases has been reported by Sander and

Wise. 13-18-13.21

Preparation of stationary phases that vary in surface coverage, especially the high

surface coverage phases (>5 jimol/m^), takes a considerable amount of experimental

effort. In the preparation procedures described by Sander,dry solvent, dry silica

substrates, high purity reagents, refluxing conditions, multiple silanes, precise control of

water and humidity, and in some cases, multiple condensation steps, have all been used to

prepare phases with a range of surface coverages. While these methods work quite well,

they can be arduous and time consuming. The ideal situation would be to control

stationary phase characteristics, such as surface coverage, without sacrificing the

simplicity of the reaction chemistry shown in Schemes 1 and 2.

556

Goals of this Work

The goal of this work was to develop improved methods for the synthesis of

alkylsilane-based stationary phases. In particular, one goal was to develop a generic

procedure for creating stationary phases of different surface coverage based on a single

reaction scheme. Phases were characterized by various analytical techniques to

determine their physical properties. In addition, Raman spectroscopy was used to

determine the conformational order of these phases, and the results are compared to the

previous investigations ofNIST Cig phases. Chromatographic performance of these

phases was determined for a series of hydrophobic solutes, and their geometric selectivity

is tested using a series of PAHs.

Methodology

Stationary Phase Synthesis

Stationary' phases were all prepared using either Kromasil 100 (referred to as

Krom-100) or Kromasil 200 (referred to as Krom-200) silica substrates. The properties

of these substrates were discussed in Chapter 2. Briefly, both substrates are 5 }j,m

spherical silica particles with 100 A (Krom-100) and 200 A (Krom-200) pore sizes. The

557

surface area for Krom-100 is 297 m^/g and for Krom-200 is 213 mVg. In all synthetic

procedures, the substrates were used as received.

Stationary phases were all synthesized via the same general procedure in which

surface coverage is controlled by varying the volume ratio of alkylsilane precursor,

octadecyltrimethoxysilane (Cig-TMOS), to organic modifier, octanol. All reactions were

done under ambient temperatures in sealed polyethylene vials (50 mL). After completion

of the condensation reactions, all stationary phases were washed and dried by sonication

in -20 mL solvent for 30 min, followed by vacuum filtration to remove the solvent. A

typical washing sequence usually followed the order cylcohexane, THF, toluene, pentane,

acetone, and ethanol. After the final washing step, stationary phases were dried on the

vacuum filter for ~12 h. Stationary phases were collected in a 5 mL round bottom flask

and dried under vacuum at ambient temperature for a minimum of three days.

Sample CJ017-1 was prepared by adding 300 mg of Krom-100 silica to a mixture

containing 5 mL cyclohexane, 1 mL THF (cosolvent), 300 nL of octanol (modifier), and

30 fiL of 0.1 M HCl catalyst with stirring. To this mixture, a 110 }iL aliquot of Ci8-

TMOS was added dropwise. After 24 and 48 h, additional 110 pL aliquots of Cis-TMOS

were added dropwise, respectively, for a total Cis-TMOS volume of 330 fiL. Samples

CJ017-2 and CJ017-3 were prepared under the same conditions with the exception of

the amounts of solvents and modifier used. For sample CJ017-2, 6 mL of cyclohexane

and 300 fiL of octanol were used with no THF. For sample CJ017-3, 5 mL of

cyclohexane and 1 mL of THF were used with no octanol modifier. The reactions were

558

stirred at ambient temperature for a total of six days, after which the resulting modified

silica stationary phases were washed and dried as described above.

Sample CJ0225 was prepared without the octanoi modifier by adding 440 mg of

Krom-200 silica to 20 mL cyclohexane, 1 mL THF (cosolvent), and 30 nL of 0.1 M HCl

catalyst with stirring. To this mixture, a 100 îL aliquot of Cis-TMOS was added

dropwise. After 24 and 48 h, respectively, additional 100 },iL aliquots of Cig-TMOS were

added dropwise for a total volume of 300 fiL. The reaction mixture was stirred for a total

of 4 days, after which the stationary phase was washed and dried.

Sample CJ03-1 was prepared by adding 440 mg of Krom-200 silica to 20 mL

cyclohexane, 150 mL octanoi, and 30 pL of 0.1 M HCl catalyst with stirring. To this

mixture, a 100 p-L aliquot of Cis-TMOS was added dropwise. After 24 and 48 h,

respectively, additional 100 }jL aliquots of Cig-TMOS were added dropwise for a total

volume of 300 |iL. For sample CJ03-2, 1 mL octanoi was used with only one 100 ul.

aliquot of Cis-TMOS. For sample CJ03-3, 1 mL octanoi was used with only 50 pL Cig-

TMOS. These reaction mixtures were stirred for a total of 4 days, after which the

resulting stationary phases were washed and dried.


cyclohexane, 300 jiL octanoi, and 30 |i,L of 0.1 M HCl catalyst with stirring. To this

mixture, a 100 |iL aliquot of Cig-TMOS was added dropwise. Every 8 to 12 h, an

additional 100 |iL aliquot of Cig-TMOS was added over the next 3 days such that the

559

total volume of Cig-TMOS was 600 [iL. For sample CJ04-2, the same procedures were

used except that nine 100 }iL aliquots of Cig-TMOS were added, giving a total Cig-

TMOS volume of 900 jiL. For sample CJ04-3, 150 jjL octanol was used with nine 100

|j,L aliquiots of Cig-TMOS. These reactions were stirred for a total of 8 days, after which

the stationary phases were washed and dried.


cyclohexane, 600 |iL octanol, and 30 |iL of 0.1 M HQ catalyst with stirring. To this

mixture, a 100 fjL aliquot of Cig-TMOS was added dropwise. Every 8 to 12 h, an

additional 100 }iL aliquot of Cig-TMOS was added over the next 3 days such that the

total volume of Cis-TMOS was 600 îL. For samples CJ05-2, CJ05-3, and CJ05-4, the

same volume of Cig-TMOS was used, but the volume of octanol modifier was varied

from 150 (JL (CJ05-2) to 50 |iL (CJ05-3) to zero (CJ05-4). These reaction mixtures

were stirred for a total of 8 days, after which the stationary phases were washed and

dried.

Characterization of Stationary Phases

The total carbon content of organically-modified silica stationary phases was

determined using a Perkin Elmer 2400 Series II CHNS Elemental Analyzer (Perkin

Elmer, experiments conducted by Desert Analytics, Tucson, AZ). From the total carbon

content, %C by mass, the alkylsilane surface coverage was detennined as described by

fy C Berendsen and de Galan ' as discussed in Chapter 2.

560

Raman spectra were collected using 100 mW of 532 nm radiation from a

Coherent Verdi Nd: YVO4 laser on a Spex Triplemate spectrograph. Slit settings were

0.5/7.0/0.150 mm for all experiments. SEM images presented in this Dissertation were

acquired with a Hitachi S-2460N variable pressure scanning electron microscope.

Chromatograms were acquired using a BAS model 200B liquid chromatograph,

retrofitted with a microbore-LC system, and a BAS 116a UV-vis detector (Bioanalytical

Systems, Inc.). All instrumemtation and experimental techniques are described in greater

detail in Chapter 2.

Silica Substrate

The presence of significant impurities, primarily metallic impurities, in silica

substrates used in RPLC has been shown to significantly hinder chromatographic

13 22 13 25 performance. ' " ' As discussed in Chapter 5, metal impurities in the YMC silica

substrates of the NIST C18 phases lead to fluorescent backgrounds of considerable

intensity in Raman spectra that prohibit the acquisition of data in the low frequency

region (<1600 cm"'). Thus, it is critical that clean silica substrates be used in the

fabrication of stationary phases for which Raman spectral characterization is undertaken.

Numerous substrates have been organically modified by procedures similar to

those described above including Zorbax SIL (Agilent Technologies), Zorbax- Rx SDL

(Agilent Technologies), Apex I silica (Jones Chromatography), Econosphere (Altech

Associates, Inc.), Adsorbosphere (Altech Associates, Inc.), Luna (Phenomenix), YMC

561

silica (YMC, Inc.), Genesis (Jones Chormatography), and Kromasil (Akzo Nobel/Eka

Chemicals). Of these substrates, only Apex I and Kromasil silica are sufficiently clean to

produce fluorescence-free Raman spectra. Kromasil silica was chosen as the substrate to

use in these experiments because it is readily available.

RPLC Column Preparation and Performance

In order to use the chromatographic performance of these materials as an indicator

of stationary phase properties and quality, multiple variables must be considered. The

most influential variable is column packing efficiency. Poor separation in a

chromatography experiment might result from column packing inefficiency, but this

variable must be distinguished from a failure of the stationary phase material. Thus, the

column packing procedure used here was first tested using commercially-available

Kromasil Cig stationary phase, a material known to give good chromatographic behavior.

Three test columns of Kromasil Cig were prepared using the slurry packing procedure

described in Chapter 2 and their performance was compared to that of two commercial

Ci8 microbore columns of the same size (1 mm I.D. x 10 cm length).

Column efficiency was determined from the number of theoretical plates (N ) for

each column for the separation of substituted aromatics, benzene, ethylbenzene, and

cumene, in a mobile phase of 80:20 methanol/water. Representative chromatograms for

each of the five columns are presented in Figures 13.1 and 13.2 with the calculated

number of plates and packing pressures provided in Table 13.1. The absolute number of

562

w a> o

§

I 10 20 30 40 50 60

Time (min)

Figure 13.1 Chromatograms of benzene, ethylbenzene, and cumene using (a) Krom-Test A (b) Krom-Test B , and (c) Krom-Test C columns packed with Kromasil Cjg stationary phase) in 80:20 methanol/water at a flow rate of 80 jiL/min.

563

Bz EtBz

Cu

EtBz

Cu Bz

20 30 4(

Time (min)

Figure 13.2 Chromatograms of benzene, ethylbenzene, and cumene using (a) BAS-1 and (b) BAS-2 commercial Cjg microbore columns in 80:20 methanol/water at a flow rate of 80 liL/min.

T able 13.1. Packing pressures and number of theoretical plates (N) for Kromasil C i g BAS Ci8 columns

Column Packing pressure (psi) N

Krom Test-A 3000 375

Krom Test-B 3200 302

Krom Test-C 5300 460

BAS-1 NA" 560

BAS-2 NA 396

"Not applicable.

565

plates for each of the three Kromasil columns are small but comparable to literature

values for columns packed at relatively low pressure. ' " ' Test column C, packed at

5300 psi, was more efficient than the other two columns packed at lower pressures, also

13 26 13 2T consistent with reported trends of efficiency with packing pressure. ' ' ' However,

packing test column C at 5300 psi damaged the seals inside the pressure transducer of the

chromatograph; therefore all other columns were packed at -3000 psi to avoid additional

damage to the system in spite of the decreased efficiency. It is important to note,

however, that the efficiencies of the Kromasil test columns are comparable to those of the

two commercial Cig columns. Given the comparable efficiencies of the commercial and

in-house-packed columns, the column packing procedure was deemed acceptable and

used to prepare columns of all other stationary phases.

Physical Characterization of Kromasil Cjs

In addition to using the Kromasil Cig phase as a chromatographic standard, the

physical properties including surface coverage, conformational order, and physical

appearance of the phase were compared to those synthesized in this work using the

octanol modifier to determine if this synthetic approach leads to useful stationary phase

materials. Alkylsilane surface coverage for Kromasil Cis (3 .66 pmol/m^) was determined

by the manufacturer from %C values.

Raman spectra of Kromasil Cig were acquired at 20 "C in the v(C-C), S(C-H), and

v(C-H) regions to determine the conformational order of the phase. A representative

566

spectrum is shown in Figure 13.3. Unlike spectra acquired for the NIST Cig phases

described in Chapters 5-10, data in the low frequency region can be successfully acquired

without spectral interference from the silica substrate. Based on several spectral

indicators, this Kromasil CIG phase is fairly disordered at 20 °C as expected for a

relatively low surface coverage. Specifically, the presence of a large fraction of gauche

conformers as indicated by the intensity of the V(C-C)G at -1080 cm"', the high frequency

position and broad shape of the t(CH2) mode at -1302 cm"' (FWHM -13 cm"'), and an

intensity ratio of 1.12 for the antisymmetric to symmetric methylene stretching modes,

I[Va(CH2)]/I[Vs(CH2)] indicate this disorder. Although numerous spectral indicators of

conformational order exist throughout the spectrum, I[Va(CH2)]/I[Vs(CH2)] is primarily

used here to facilitate direct comparison with the order ofNIST Cig phases.

SEM images were also acquired for the Kromasil Cig phase and for the

unmodified silica substrate, Krom-100, as shown in Figure 13.4. At relatively low

magnification (x 3000), no difference in shape or regularity is noted between the

modified and unmodified silica particles.

Effects of Substrate

Before the stationary phase properties can be tailored using this new modifier-

based synthetic approach, this technique first had to be validated as a method for

preparing usefiil chromatographic materials. Numerous other phases were prepared prior

to the CJ017 phases; all were physically similar and gave comparable chromatographic

567

* T—4 zn a D

v^CH^) ÂCCHJ)

T(CH,) V(C-C)T N \ V(C-C) \

1000 1100 1200 1300 1400

Wavenumbers (cm-^)

2800 2900 3000

Figure 13.3 Raman spectra in the v(C-C), 5(C-H), and v(C-H) regions for Kromasil Cjg. Acquisition times are 8 min in the low frequency region and 4 min in the high frequency region.

SEM images at 3000x magnification of (a) Kromasil Cjg and (b) unmodified Kromasil silica.

569

performance to that of the CJ017 phases. Therefore, reported results are limited to

discussion of the CJ017 and CJ0225 samples for brevity.

The physical properties of samples CJ017-1, CJ017-2, CJ017-3, and CJ0225 are

listed in Table 13.2, including J[Va(CH2)]/I[Vs(CH2)] values for each phase as determined

from their Raman spectra shown in Figure 13.5. Although the surface coverage of all

CJ017 phases is greater than that of CJ0225, the most striking diflference between the

phases is their values of I[va(CH2)]/I[Vs(CH2)]. The values of I[va(CH2)]/I[Vs(CH2)] for

all CJ017 phases are close to 2 and independent of surface coverage, suggesting that

these phases are approaching the order of crystalline octadecane (Chapter 3).

Significantly, these values of I[Va(CH2)]/T[Vs(CH2)] are even greater than those for the

TFC18SF NIST Cig phase (6.45 [.imol/m^) that has a I[Va(CH2)]/I[Vs(CH2)] value of 1.38

(Chapter 5). In contrast, for CJ0225, the value of I[Va(CH2)]/I[Vs(CH2)] is 1.00,

indicating a considerably more disordered phase that is comparable to other phases with

surface coverages on the order of 3-3.5 jimol/m^ described in Chapters 4 and 5.

SEM images of CJ017-2 and CJ0225 shown in Figure 13.6 reveal the reason for

the observed differences in order. (Images of CJO 17-1 and CJ017-3 are not shown but

are identical to those shown for CJ017-2.) The images of CJ017-2 exhibit large

aggregates of condensed alkylsilane between the silica particles and the appearance of

particles that have been polymerized together in some cases (Figure 13.6d). A large

portion of the alkylsilane used to modify these particles is in the form of condensed, solid

alkylsilane, and the Raman spectra reflect the highly crystalline nature of this crystalline

570

Table 13.2. CJ017 and CJ0225 phases and properties

Phase Modifier (M-L)

Cig-TMOS

(HL) Substrate %

Carbon r

(timol/m^)

R"

CJ017-1 300 110 Krom 100 18.74 4.08 1.97 + 0.007

CJ017-2 300 110 Krom 100 23.41 5.66 1.88 ± 0.010

CJ017-3 0 110 Krom 100 22.01 5.15 1.96 + 0.005

CJ0225 0 100 Krom 200 11.86 3.15 1.00 + 0.001

"R = I[va(CH2)]/I[Vs(CH2)] as calculated from the spectral data in Figure 13.5.

571

v(C-C>

1000 1100 1200 1300 1400 3000 2800

Waven umbers (cm-^

2900

Figure 13.5 Raman spectra in the v(C-C), 6(C-H), and v(C-H) regions for samples (a) CJ017-1, (b) CJ017-2, (c) CJ017-3, and (d) CJ0225. Acquisition times are 8 min in the low frequency region and 4 min in the high frequency region for all spectra.

572

Figure 13.6 SEM images at 3000x magnification of (a) Kronmsil Cjg, (b) CJ0225, and two regions of CJ017-2 (c and d).

573

phase. Images of C J0225 show only scarce areas of isolated aggregate formation, and

these particles are largely indistinguishable from the unmodified silica substrates (Figure

13 .4). In this case, most of the alkylsilane used to modify the substrate either condenses

on the substrate or is washed away, resulting in a relatively disordered alkylsilane-

modified material, consistent with the Raman spectrum of Figure 13 .4.

Although the solvent and modifier conditions in the synthesis of these phases are

different (Table 13.2), it seems unlikely that these differences would result in such large

differences in the condensation behavior of Cig-TMOS. The other variable in these

syntheses is the nature of the substrate. CJ017 phases were prepared on Krom-100 silica

and CJ0225 was prepared on Krom-200. Thus, it is possible that pore size of the silica

particles results in the different stationary phase properties observed.

Parikh et al."'^^ have described the self-assembly of hydrolyzed ocatdecylsilane

particles as an extended bilayer structure, similar to what is shown in Figure 13.7. The

height of these structures is ~50 A and the width is on the order of hundreds of

angstroms. Thus, it is quite reasonable that the pore size of the silica substrate could

affect the ability of octadecylsilane to condense on the surface. For Krom-100,

hydrolyzed, self-assembled octadecylsilane particles of these dimensions would be

excluded from much of the surface area of the silica particles that is contained within

pores resulting in condensed, solid octadecylsilane aggregates that form separately and

leave much of the silica surface is unmodified. For Krom-200, the pores are large

Si _^Si Si ^Si

HO. OH OH OH OH

Water

Lin 9H OH OH OH "Os IyOv I ^ Os I Ôn I ÔH

Si Si Si Si

Figure 13 7 Self-assembly of hydrolyzed octadecylsilane into multilayer structures adapted from reference 13 .28.

575

enough to accommodate self-assembled octadecylsilane particles allowing alkylsilane

modification of the silica surfaces inside the pores.

Attempts to pack the CJ017 phases into columns failed because of the difficulties

inherent in keeping the large aggregates of crystalline hydrolyzed octadecylsilane

material in suspension, with the exception of one column prepared with CJ017-2. The

chromatographic performance of CJ017-2 and CJ0225 was assessed through the

separation of the akylbenzenes benzene, ethylbenzene, and cumene in 80:20

methanol/water as shown in Figure 13,8. The CJ017-2 column gives very poor

separation of these solutes with only 78 theoretical plates. In contrast, the CJ0225

column yields a much better separation of these solutes with baseline resolution of all

peaks and 517 theoretical plates. This efficiency is comparable to that observed on both

the Kromasil Cis column and the commercial BAS columns.

Modifier Approach to Controlling Surface Coverage

The chromatographic results suggest that fabrication of stationary phases using

the general procedure described for CJ0225 results in materials whose performance is

comparable to that of commercially-available phases. However, CJ0225 was prepared

with no modifier. Thus, the question that still remains to be answered is whether

stationary phase materials of controlled surface coverage can be successfully fabricated

through control of the relative amount of octanol modifier present.

576

EtBz Cu

Bz

(D O EtBz Cu

Bz

0 20 30 40 50

Time (min)

Figure 13 .8 Chromatograms of benzene, ethylbenzene, and cumene on (a) CJ017-2 and (b) CJ0225 in 80:20 methanol/water at a flow rate of 80 |iL/min.

577

In the first of two series of materials made using octanol as a modifier, all solution

conditions were varied from previous values to determine and an appropriate protocol for

phases with reasonably high alkylsilane surface coverages by this route. The properties

of the CJ03 and CJ04 phases are listed in Table 13.3. The I[Va(CH2)]/Ilv,(CH2)] values

listed in this table were calculated from the Raman spectra of these phases shown in

Figures 13.9 and 13.10. In this series of materials the volume of Cig-TMOS and the

volume of octanol are varied making it difficult to quantitatively determine the effect of

modifier amount on alkylsilane surface coverage. However, in general, as the volume

ratio of Cig-TMOS;octanol increases, alkylsilane surface coverage increases. Similarily,

as the Cis-TMOS: octanol ratio increases, the value of I[Va{CH2)]/I[VsfCHi)] also

increases, indicating that these phases become more ordered with increasing surface

coverage. This trend is identical to that observed for the NIST Cig phases as described in

Chapter 5. Moreover, the values of I[Va(CH2)]/I[Vs(CH2)] for these phases are

comparable to those for the NIST phases of similar surface coverage. For example, the

value of I[Va(CH2)]/I[Vs(CH2)] for the NIST phase DFC18SL (4.17 |imol/m^) is 1.12 and

the comparable values for the two phases synthesized using the octanol modifier

approach yielding surface coverages of 4.10 |imol/m^ are 1.09 and 1.14, respectively.

One interesting observation about these phases is that even though CJ04-1 and

CJ04-2 were prepared with different Cig-TMOS;octanol volume ratios of 2; 1 and 3:1,

respectively, the resulting stationary phases have approximately the same surface

coverage and comparable conformational order. This observation would be easy to

578

Table 13.3. CJ03 and CJ04 phases and properties

Phase Modifier C18-TMOS Substrate % r (jiL) (pL) Carbon (jimol/m^)

CJ03-1 150 300 Krom-200 12.38 3.32 1.03 ± 0.003

CJ03-2 1000 100 Krom-200 7.23 1.77 0.96 ± 0.004

CJ03-3 1000 50 Krom-200 4.30 1.00 0.94 ± 0.002

CJ04-1 300 600 Krom-200 14.66 4.10 1.14 ± 0.002

CJ04-2 300 900 Krom-200 14.66 4.10 1.09 + 0.002

CJ04-3 150 900 Krom-200 17.06 5.02 1.26 1- 0.003

"R = I[Va(CH2)]/I[Vs(CH2)] as calculated firom the spectral data in Figures 13.9 and 13.10.

579

Vs(CH,)^ v,(CH,)

y(C~C).

1000 1100 1200 1300 1400 2800

Wavenumbers (cm"^)

2900 3000

Figure 13.9 Raman spectra in the v(C-C), 8(C-H), and v(C-H) regions for (a) CJ03-1, (b) CJ03 -2, and (c) CJ03 -3. Acquisition times are (a and b) 8 min in the low frequency region and 2 min in the high frequency region and (c) 12 min in the low frequency region and 5 min in the high frequency region.

580

TCCH )

1000 1100 1200 1300 1400 2800 2900 3000

Wavenumbers (cm"^)

Figure 13.10 Raman spectra in the v(C-C), 5(C-H), and v(C-H) regions for (a) CJ04-1, (b) CJ04 -2, and (c) CJ04 -3. Acquisition times are (a and b) 8 min in the low frequency region and 2 min in the high frequency region and (c) 12 min in the low frequency region and 5 min in the high frequency region.

581

rationalize if the volumes of Cig-TMOS used were the same for each preparation, but,

although the octanol volume was 300 pL in both preparations, the Cig-TMOS varied

from 600 to 900 p,L, While this preliminary trend has been made only on the basis of two

preparations, it is an indication that the presence of the octanol modifier has a significant

impact on the resulting Cig surface coverage and that the octanol modifier might have

more of an influence on surface coverage than the amount of alkylsilane precursor.

Based on these observations, a second series of stationary phases was prepared

using a Cig-TMOS volume of 600 (iL and variable amounts of octanol. The properties of

these phases are listed in Table 13.4, including values of I[Va(CH2)]/I[Vs(CH2)]

determined from the Raman spectra of these materials shown in Figure 13.11. In addition

to the four new materials, CJ04-1 is also listed in Table 13.4 as it was also prepared

using the same volume of alkylsilane. With the same amount of Cig-TMOS used in each

preparation, it is clear that the octanol modifier has a significant influence on the

resulting Ci8 surface coverage. Decreasing the amount of octanol increases the Cis

surface coverage, and the phase prepared in the absence of octanol has the highest Cig

surface coverage. Moreover, the conformational order of these phases systematically

increases with increasing Ci8 surface coverage, as indicated by the values of

I[Va(CH2)]/I[Vs(CH2)]. In fact, the conformational order of these phases compares well to

those of the NIST Cig phases for similar surface coverages. This trend is shown more

clearly in a plot of I[Va(CH2)]/I[Vs(CH2)] as a fiinction of Cig surface coverage in Figure

13.12.

582

Table 13.4. CJ05 phases and properties

Phase Modifier Cig-TMOS Substrate % r R" (fiL) (|iL) Carbon (jimol/m^)

CJ05-1 600 600 Krom-200 12.71 3.43 1.02 ± 0.007

CJ04-1 300 600 Krom-200 14,66 4.10 1.14 ± 0.002

CJ05-2 150 600 Krom-200 16.07 4.63 1.17 + 0.002

CJ05-3 50 600 Krom-200 17.01 5.00 1.22 ± 0.005

CJ05-4 0 600 Krom-200 19.13 5.88 1.31 + 0.004

"R = I[va(CH2)]/I[vs(CH2)] as calculated from the spectral data in Figures 13.10 and 13.11.

583

TCCHJ)

v(C-C)

v(C-C),

3000 1000 1100 1200 1300 1400 2800

Wavenumbers (cm"^)

2900

Figure 13.11 Raman spectra in the v(C-C), 8(C-H), and v(C-H) regions for (a) CJ05-1, (b) CJ05 -2, (c) CJ05 -3, and (d) CJ05 -4. Acquisition times are 8 min in the low frequency region and 3 min in the high frequency region for all spectra.

584

Surface Coverage (}imol/m)

Figure 13.12 I[v^(CH2)]/I[Vs(CH2)] as a function of C,jj surface coverage (fimol/m^) for CJ04-1 and CJ05 phases (•) and NIST Cjg phases (O).

585

It is interesting to note the difference in the value of If Va(CH2)]/I[Vs(CH2)] for the

two lowest surface coverage phases, the 3.09 |amol/m^ NIST phase (MFC 18) and CJ05-1

(3.43 (imol/m ). CJ05-1 is considerably more disordered than MFCIS, even though it

has a slightly higher Cis surface coverage. This could be due to differences in

homogeneity between the phases, but it is difficult to draw meaningful conclusions with

such a limited data set.

The utility of these phases as chromatographic supports was tested by the

separation of the hydrophobic solutes benzene, ethylbenzene, and cumene in the

chromatograms shown in Figures 13.13 and 13 .14, with column efficiencies listed in

Table 13.5. All columns gave separation efficiencies that are comparable to those of the

Kromasil Cig and commercial BAS columns, 300-600 theoretical plates, with the

exception of CJ03-2 and CJ05-4. Separations on these columns are much less efficient.

For CJ03-2 (1.77 jimol/m^the alkylsilane surface coverage is simply too low to give a

quality separation of these solutes under these experimental conditions (mobile phase,

flow rate, etc.). For CJ05-4, the opposite effect is observed. The surface coverage of

this phase (5.88 jimol/m^) is sufficiently high that mass transport and free volume

restrictions of the phase negatively impact the separation efficiency, as was observed for

TFC18SF (6.45 ^mol/m^) in the separation of toluene, anisole, and aniline in Chapter 10.

While it is clear that the use of the octanol modifier allows control of alkylsilane

surface coverage, the mechanism by which this control is afforded is not clear. It is likely

that the condensation is controlled by a combination of two events: (1) solvation of or

586

EtBz Cu

Bz

EtBz Cu

Bz

o m

EtBz Cu

Bz

0 10 20 30 50 60

Time (min)

Figure 13.13 Chromatograms of benzene, ethylbenzene, and cumene using (a) CJ03-2, (b) CJ04-1, and (c) CJ04-3 in 80:20 methanol/water at a flow rate of 80 jjL/min.

EtBz

Cu Bz

EtBz

Cu Bz

AAJ V,

EtBz (U o

o Cfl

Cu Bz

EtBz

Cu Bz

0 20 30 50 60

Time (min)

Figure 13.14 Chromatograms of benzene, ethylbenzene, and cumene using (a) CJ05-1, (b) CJ05-2, (c) CJ05-3, and (d) CJ05-4 in 80:20 methanol/water at a flow rate of 80 jiL/min.

588

Table 13.5. Number of theoretical plates (N) for CJ03, CJ04, and CJ05 columns

Column r (|imol/m^) N

CJ03-2 TTT W

C J04-1 4.10 629

CJ04-3 5.02 416

CJ05-1 3.43 423

CJ05-2 4.63 508

CJ05-3 5.00 320

CJ05-4 5 88 134

589

adsorption on the silica substrate by octanol prior to alkylsilane condensation and (2)

solvation of self-assembled, hydrolyzed alkylsilane by octanol prior to condensation. If

the silica surface is solvated by octanol then displacement of octanol at the surface could

act as an energetic or kinetic barrier to alkylsilane condensation on the surface because

octanol and other primary alcohols are known to adsorb strongly to silica surfaces.

" More octanol on the surface would increase the barrier leading to a lower surface

coverage. This is shown schematically in Figure 13 .15 where hydrolyzed alkylsilanes

must displace octanol molecules at the silica surface prior to surface condensation.

Increasing the amount of octanol increases the energetic or kinetic barrier for alkylsilanes

to displace sufficient quantities of octanol to reach the silica surface.

Another possible mechanism for octanol to influence the bonding structure of

alkylsilanes is one in which octanol is incorporated into the self-assembled, multilayer

structure of octadecylsilane shown in Figure 13.7. Solvation of hydrolyzed alkylsilanes

by octanol would increase the spacing between alkyl chains. In this case, when these

structures approach the silica surface, they might form bonds with the surface while

retaining this expanded configuration. This is shown in Figure 13.16 where octanol is

incorporated between self-associated alkylsilane increasing the space between them in

solution. Once bonding occurs, the octanol "filler" could be washed away with no

structural consequences. Thus, varying the octanol:alkylsilane ratio results in various

alkylsilane spacings on the surface, i.e. surface coverage. These mechanisms are just two

preliminary hypotheses describing the effect of octanol on the formation of such phases.

Si HO" ' OH tiU Qjj Si

ho" ' OH MU Qjj

OH ojj OH OH OH OH

OH

Si OH

OH HO

OH Si

* DH HO ' OH OH OH

SiO- SiO,

Figure 13.15 Schematic of octanol displacement by alkylsilanes at the silica surface.

U1 O

OH

OH OH ^ ^ OH > OH

,Si Si > Si HO' ^ HO ^ Ô^ ' ""OH ^ HO'^ ^

OH OH OH OH OH OH

Si ./ I \

Water

OH

HO,P?O.P? J Si Si

OH

OH

OH

OH HOsîyOH OH

OH ^

OH Si Si

OH OH OH Si

HOîÔÔH OH HO'AÔH OH

SiO.

Figure 13.16 Schematic of octanol incorporation in self-assembled, hydrolyzed alkylsilane. CA VO

592

Further experimental work, including microscopy on modified model planar substrates or

more detailed spectroscopy on these stationary phases to obtain information about surface

homogeneity, must be undertaken in order to obtain a better understanding of these

effects.

Geometric Shape Selectivity of High Surface Coverage Phases

While it is significant that the presence of the octanol modifier in these

preparations still allows functional chromatographic phases to be fabricated, of greater

interest is whether these phases exhibit the shape selectivity of the high surface coverage

phases prepared by conventional methods. To test the shape selectivity of these phases, a

1 Th*? 1 "X "XA standard reference mixture of analytes developed by Sander and Wise was used. • ' "

SRM 869a contains a mixture of three polycyclic aromatic hydrocarbons (PAHs),

benzo[a]pyrene (BaP, planar shape), phenanthro[3,4-c]phenanthrene (PhPh, nonplanar

shape), and 1,2:3,4:5,6:7,8-tetrabenzonaphthalene (TBN, nonplanar shape) in 85:15

acetonitrile/water; the chemical structures and geometric shapes of these are shown in

Figure 13.17. The elution order of these PAHs has been shown to correlate with

stationary phase shape recognition ability and allows phases to be categorized as

polymeric (highly shape selective) or monomeric (poor shape selectivity) as shown in

Figure 13.18.The selectivity coefficient, ctTBN/BaPhas been used

as a numerical descriptor of shape selectivity. For values of axBN/BaP < 1, Cis columns

are considered to be polymeric whereas for values of ajeN/BaP > 1.7, Cig columns are

593

Benzo[a]pyrene (BaP)

Phenanthro[3,4-c]phenanthrene (PhPh)

1,2:3,4:5,6:7,8-Tetrabenzonaphthalene (TBN)

Figure 13.17 Chemical structures and geometric shapes of PAHs used to determine RPLC column shape selectivity.

594

BaP PhPh

.TBN

i a = 2.08

PhPh + BaP .TBN

PhPh TBN

BaP

TBN + BaP

TBN

Time

a - L80

a - 1.27

a = LOO

a = 0.48

o >-s fti P QO

(K? m c

W o ft)

O o < CD T-S

(n a

Figure 13.18 Chromatograms of PAHs used to determine shape selectivity. Adapted from reference 13.30.

595

considered to be monomeric. With monomeric columns, BaP elutes before TBN. The

retention of BaP relative to TBN increases as the column becomes more polymeric and

the elution order is reversed for polymeric columns.

Figures 13.19 and 13 .20 show chromatograms of BaP, PhPh, and TBN for

Kromasil Ci8, BAS, and selected CJO phases, with their calculated values of axBN/BaP

listed in Table 13.6. The low coverage (<4 umol/m^) commercial and CJO phases

exhibit poor shape selectivity with aiBN/BaP values >1. However, the higher coverage

phases (>4.5 |.imol/m^) exhibit a high degree of shape selectivity with aiBN/BaP values <1.

The shape selectivity of these phases as a function of Cig surface coverage compares well

with those of the Cig phases prepared by Sander and Wise.'"'"' The adherence of the CJO

phases to previous trends is made clear by the plot of arBN/eaP as a function of C18

surface coverage shown in Figure 13.21. The axBN/BaP values for the CJO phases are in

line with those of the NIST phase reported earlier. The arBN/BaP values of the two

commercial phases are included in this plot and further establish the relationship between

otTBN/Bap and CI8 surface coverage. Thus, these results collectively support the

contention that stationary phases with systematically varying characteristics can be

rationally designed using the approach as described here.


In summary, the preparation of stationary phases via solution modifier techniques

results in useful chromatographic supports of varying bonding densities that can be

596

TBN

BaP PhPh

TBN BaP o

o c«

PhPh

TBN

PhPh BaP

Time (min)

Figure 13.19 Chromatograms of PhPh, BaP, and TBN (SRM 869a) using (a) BAS-1, (b) Krom-T est C, and (c) CJ04-3 columns in a mobile phase of 85:15 acetonitrile/water at a flow rate of 170 fiL/min.

TBN

BaP PhPh

TBN

BaP PhPh

TBN

<u o PhPh § o

TBN PhPh

BaP

TBN

PhPh BaP

20 30 40

Time (min)

0

Figure 3.20 Chromatograms of PhPh, BaP, and TBN (SRM 869a) using (a) CJ05-1, (b) CJ04-1, (c) CJ05-2, (d) CJ05-3, and (e) CJ05-4 columns in a mobile phase of 85:15 acetonitrile/water at a flow-rate of 170 |iL/min.

598

Table 13.6. Shape selectivities (aiBN/Bap) for Kroniasil Cis, BAS, CJO, and NIST Ci8 columns

Column r (jj,mol/ni^) otTBN/Bap

BAS-1 320 1.63

Krom Test-3 3 .66 1.60

CJ05-1 3.43 1.13

CJ04-1 4.10 1.06

CJ05-2 4.63 0.71

CJ05-3 5.00 0.68

CJ04-3 5.02 0.84

CJ05-4 5.88 0.51

M-l" 3.51 1.82

¥-2" 4.48 1.48

P-r 5.26 0.73

S-l" 6.45 0.31

"NIST Ci8 columns from reference 13 .19.

599

2.0 n

1.8- <P

1.6-

1.4- o ^

% 1.2- • 1 1.0- • <» a

0.8- • 0.8-• #o

0.6-- •

0.4-

0.2- i ' B « 1 ' U i 1 . 1 . . i 1 1 , "T ' I ' I I ' [ ' 1 ' ~

2 3 4 5 6 7 8

Surface Coverage (lumol/m^)

Figure 13.21 a^BN/BaP 3-s a function of surface coverage for CJO phases (•), commercial Kromasil C,g and BAS phases (^), and NIST Cjg phases (O).

600

controlled using only one set of relatively simple reaction conditions. All phases were

prepared without any special pretreatment or purification of reagents or silica substrates

under ambient conditions. Phases of conventional surface coverage (--3-4 jimol/m^) gave

separation efficiencies that are comparable to commercially available columns for a

series of substituted aromatic solutes. High surface coverage phases exhibited a high

degree of shape selectivity, analogous to phases of comparable coverage prepared by

Sander and Wise."''*^

Only a limited set of phases prepared by these methods have been characterized

here with a great deal of work yet to be done. More phases should be prepared under the

same experimental conditions to determine not only reproducibility of the technique, but

to better determine the precision with which surface coverage can be controlled using this

approach. Several fundamental reaction conditions must be optimized as well, including

reaction time, amounts of reagents, and scalability of these procedures.

In addition to the stationary phase fabrication, understanding how the solution

modifiers affect the condensation of alkylsilanes is another area for future studies.

Different modifiers including longer and shorter chain primary alcohols, aliphatic acids,

and amines could be used which could have more interaction with either the alkylsilanes

or the silica surface and could ultimately affect the condensation. Also, performing this

modification procedure on model planar substrates would allow condensed silanes to be

studied by a variety of spectroscopic and microscopic techniques to gain more

information about the surfaces of these modified stationary phases.

601

Other future studies include the fabrication of stationary' phases with controlled

surface coverage using sol-gel techniques. While organically-modified sol-gels have

been used previously as chromatographic stationary support coatings for open-tubular

capillary electrophoresis (OTCE)'^ and gas chromatography (GC)'^ " and as

1 52 1 AA

porous monolithic supports in capillary electrochromatography (CEC), ' " ' sol-gels

have not yet been used to modify silica particles for packed-bed liquid chromatography.

Although control of the uniformity and accessibility of the organic portion of these

coatings could be challenging, if successful, this approach would offer an attractive

ahemative to traditional stationary phase preparation procedures that could result in

unique properties.

602

Chapter 14


Overview of the Problems

Understanding Retention Mechanisms

A molecular description of solute retention in reversed-phase liquid

chromatography (RPLC) has eluded the chromatography community for years, despite a

considerable amount of experimental and computational work. Numerous models have

been proposed to describe molecular retention, but these models fail to sufficiently

describe retention for the countless possibilities of chromatographic stationary phases,

mobile phases, and solute combinations. The inability to adequately describe the

mechanism of solute retention on the molecular level is a result of the difficulty in

studying fundamental molecular interactions known to sffect chromatographic

separations. Such interactions include solvent-stationary phase, solute-stationary phase

and stationary phase inter- and intramolecular interactions.

Stationary Phase Fabrication

The synthesis of conventional monomeric, low surface coverage alkylsiiane

stationary phases for RPLC is relatively simple, an attribute that has contributed to

making these phases the standard in analytical RPLC separations. However, for certain

603

applications including the separation of geometrically constrained solutes and polycyclic

aromatic hydrocarbons (PAHs), polymeric phases of higher surface coverage have been

shown to give profoundly better separations than monomeric phases. Unfortunately,

controlling stationary phase surface coverage and fabricating high-density phases are

considerably more difficult than the fabrication of an individual, monomeric phase.

Thus, the development of a synthetic approach to control alkylsilane surface coverage

without sacrificing the simpHcity of the monomeric phase modification procedures would

be ideal.

Summary of Research

Raman spectroscopy was used to determine that conformational order and

interchain coupling of the alky! component of a series of high-density phases are

sensitive to temperature and surface coverage. Each of the stationary phases exhibited

subtle temperature-dependent phase changes that are generally described by the

Clapeyron relationship. The temperature at which the phase change occurs is postulated

to be reflective of the condensed phase adopted by the alkyl component or may be

reflective of the grafting procedure employed in the preparation of the phase.

In addition, alkylsilane rotational and conformational order of this same series of

stationary phases is dependent on not only solvent parameters, but on stationary phase

grafting procedure, and surface coverage, and can be used to describe specific solvent-

stationary phase interactions. In general, polar solvents increase the rotational order of

604

homogeneously distributed high surface-coverage materials, fitting an absorption model

of interaction at the distal methyl group. The increase in rotational order observed for

these phases is attributed to solvent self-association at the interface, where lateral

ordering of solvent molecules induces ordering of the stationary phase. Nonpolar

solvents generally intercalate into the alkyl chains and fit a partitioning model of

interaction; heterogeneously distributed and lower coverage materials are more

disordered by this partitioning than higher coverage materials. In addition to the solvent-

stationary phase interactions, an interplay between solvent- and temperature-dependent

ordering of these materials was observed; more solvent-dependent order was observed for

lower coverage materials in nonpolar solvents and more temperature-dependent order

was observed for higher coverage materials in polar solvents.

Raman spectroscopy wa used to examine the order of the highest and lowest

coverage phases in this series in the presence of solute/mobile phase systems. In the

presence of aniline, anisole, and toluene in methanol/water mobile phases, each of these

stationary phases exhibited a high degree of conformational order. Results provide direct

molecular interaction information for these systems and were correlated to the

chromatographic behavior of these solutes, giving insight into their retention

mechanisms. Each of these solutes interact with these phases via a partitioning

mechanism, where the solutes intercalate between the alkyl chains; however, this

intercalation in some cases is very slight, restricted to the distal methyl and methylene

groups of the stationary phase. This work represents to first experimental evidence of

605

stationary phase structural changes in response to solute retention and the correlation of

structural information with chromatographic behavior to elucidate a retention mechanism.

Raman spectroscopy was also used to investigate fundamental interactions of ion

exchange chromatographic systems. For anion exchange systems in a series of aqueous

halide solutions, the structure of surface-bound trimethylammonium ions is more

significantly affected by hydrated halides of increasing charge-density. However, the

structure of surface-confined sulfonate ions is unaffected by the presence of alkali metal

ions of different size or charge-density for cation exchange systems. The difference in

behavior of these systems is attributed to poor hydration of trimethylammonium and

halides relative to sulfonate and alkali metal ions.

A novel approach to controlling surface coverage in the preparation of alkylsilane

stationary phases was also presented. The preparation of stationary phases via solution

modifier techniques gave useful chromatographic supports of controllable bonding

densities including high-density phases (>5 ^mol/m^). All phases were prepared without

any special pretreatment or purification of reagents or silica substrates under ambient

conditions. Phases of conventional surface coverage (-3-4 lamolW) gave separation

efficiencies that are comparable to commercially available columns for a series of

substituted aromatic solutes. The high-density phases exhibit a high degree of shape

selectivity, analogous to existing NIST Cig phases of comparable coverage studied in this

Dissertation

606

Future Studies

The work presented in this Dissertation suggests a wide range of future studies in

which Raman spectroscopy can be used to characterize interfaciaJ chromatographic

phenomena and additional strategies for alkylsilane stationary' phase fabrication.

Temperature-Induced Phase Transitions in Alkylsilane Stationary Phases

Future directions for this work include the temperature-dependent phase behavior

of other homologous series of stationary phases of varying chain lengths, to better

understand phase behavior of shorter and longer surface-confined alkylsilanes. Perhaps

the most useful study would be to derive a quantitative relationship between R,

I[Vs(CH2)]/I[Va(CH2)], and surface pressure, such that quantitative thermodynamic values,

AHfiis and AVfUs, for these stationary phases can be determined from the values of Raman

conformational order indicators. One strategy would be to fit R vs. temperature data for

bulk alkane materials such as polyethylene to the Clapeyron equation, materials for

which AHfts and AVfUs are known. Thus, a proportionality constant between R and

surface pressure could be determined.

Solute-Mobile Phase Effects on the Conformational Order of Alkylsilane Stationary

Phases

Investigations of the solute/mobile phase/stationary phase interface presented in

this Dissertation are the first reports of structural changes of the stationary phase in the

607

presence of solute under chromatographic conditions which provide insight into

molecular retention of specific solutes. Two future experiments of primary importance

are: (1) investigating the effect of a homologous series of alkylbenzenes on the order of

alkylsilane phases and (2) repeating these experiments on similar high and low coverage

phases with polycyclic aromatic hydrocarbon solutes (PAHs). The chromatographic

behavior of the homologous series of alkylbenzenes has been studied extensively'^

and correlation of the chromatographic behavior to the conformational order information

would provide alkylsilane structural information that can be used to further explain and

quantify the retention behavior of these compounds.

One unique characteristic of the high-density NIST Cis-phases is their geometric

shape selectivity for planar PAHs, determined by Sander and Wise.''^ *''''''^ Thus, using

PAHs as solutes in our spectroscopic experiments will show a relationship to these

chromatographic experiments. It is hypothesized that the conformational order of high

surface coverage phases, such as TFC18SF, would be different for planar PAH solutes

than for nonplanar PAHs, because of their different interactions with high bonding

density phases, in general, these experiments would give some structural information

about densely-packed alkylsilanes during chromatographic separations that can be used to

better understand their functions as shape selective phases.

There is also a multitude of additional experiments of lesser priority that could

also be performed in the near future. These experiments include investigating the order

608

of similar phases with the same monosubstituted aromatic solutes over a wider range of

mobile phase polarity (0-100% water) with different organic modifiers (acetonitrile and

THF), studying the order of phases with varying alkyl chain length {C4-C30), and

exploring the effects of different solutes relevant to areas of environmental (chlorinate

aromatics), biological (amino acids or nucleotide bases), or forensic (drugs of abuse)

interest where RPLC is a commonly employed analytical technique.

Ion-Exchange Chromatographic Systems

Future directions for these studies include doing similar experiments with other

ions of environmental or biological interest, such as lanthanide and actinide oxides,

amino acids and small peptides. In addition, investigating the structures of these surface-

bound ions in solvents such as methanol, acetonitrile, and THF or mixtures of these with

aqueous solutions are also important, as these solvents are common modifiers in ion

chromatographic experiments. Experiments could also be done on other ion exchange

phases including polystyrene or perfluorinated polymer resins, weak anion exchange

phases (primary and secondary amines), and weak cation exchange phases (carboxylic

acids).

Specifically for the cation exchange studies, no structural changes are observed

for the materials for exchanging cations in aqueous solutions, because of the well-

hydrated sulfonate surface and solution cations. Repeating these experiments, but

dehydrating the exchange resin prior to spectral analysis, would be useful to determine if

609

surface sulfonate structure is affected by counter ion exchange, analogous to previous IR

14 g 14 9 spectroscopy work. '' '

Additional Strategies for Stationary Phase Fabrication

Work presented on the fabrication of stationary phases is only preliminary, with a

great deal of work yet to be done. In general, more phases need to be prepared under the

same experimental conditions to determine not only reproducibility of the technique, but

how precisely surface coverage can actually be controlled using this technique. There are

a number of flindamental reaction condition issues that need to be optimized as well,

including reaction time, amounts of reagents, and determining how this procedure

behaves for scaled up reactions (multiple grams).

In addition to the stationary phase fabrication, understanding how the solution

modifiers affect the condensation of alkylsilanes is another area for future studies. The

use of different modifiers including longer chain primary alcohols, aliphatic acids, and

amines could be used which could have more interaction with either the alkylsilanes or

the silica surface and could ultimately affect the condensation. Also, performing this

modification procedure on model planar substrates would allow for the condensed silanes

to be studied by a variety of spectroscopic and microscopy techniques in order gain more

information about the surface of these modified stationary phases.

Other future studies include the fabrication of stationary phases with controlled

surface coverage using sol-gel techniques. While organically-modified sol-gels have

610

been used previously as chromatographic stationary supports coatings for open-tubular

capillary electrophoresis (OTCE)'"' and as porous monolithic supports in capillary

electrochromatography sol-gels have never been used to modify silica

particles for packed-bed liquid chromatography. This approach could offer an additional

experimental procedure to control alkylsilane surface coverage using a single generic

method.

611

References

Chapter I

1.1 Laurent, C.; Billet, H. A. H. De Galan, L. Chromatographia, 1983, 17, 253-258.

1.2 Pesek, J. J.; Tang, V. H. Chromatographia, 1994, 39, 649-654.

1.3 Weber, T. P.; Carr, P. W Anal. Chem. 1990, 62, 2620-2625,

1.4 Weber, T. P.; Jackson, P. T.; Carr, P. W. Anal. Chem. 1995, 67, 3042-3050.

1.5 Quo, Y.; Colon, L. A. Anal. Chem. 1995, 67, 2511-2516.

1.6 Guo, Y.; Colon, L. A. J. Microcolumn Sep. 1995, 7, 485-491.

1.7 Hayes, J. D.; Malik, A. Anal. Chem. 2000, 72, 4090-4099.

1.8 Wang, D.; Chong, S. L.; Malik, A. Anal. Chem. 1997, 69, 4566-4576.

1.9 Abel, E. W ; Pollard, F. H., Uden, P. C.; Nickless, G. J. Chromatogr. 1966, 22, 23-28.

1.10 linger, K. K., Becker, N.; Rouneliotis, P. J. Chromatogr. 1976, 125, 115-127.

1.11 Kirkland, J. J. Anal. Chem. 1971, 43, 36A-48A.

1.12 Berendsen, G. E.; Pikaart, K. A.; de Galan, L. J. Liq. Chromatogr. 1980, 3, 1437-1464.

1.13 Kirkland, J. J., DeStefano, J. J. J, Chromatogr. Sci. 1970, 8, 309-314.

1.14 Majors, R. E. J. Chromatogr. Sci. 1974, 12, 767-778.

1.15 Kirkland, J. J. J. Chromatogr. Sci. 1971, 9, 206-214.

1.16 Sander, L. C.; Wise, S. A. Anal. Chem. 1984, 56, 504-510.


1.18 Horvath, C; Melander, W.; Molnar, I. J. Chromatogr. 1976, 125, 129-156.

6 1 2

1.19 Melander, W.; Campbell, D, E.; Horvath, C. J. Chromatogr. 1978, 158, 215-225.

1.20 Karger, B L.; Gant, J. R.; Hartkopf, A.; Weiner, P. H. J. Chromatogr. 1976, 128, 65-78.

1.21 Vailaya, A.; Horvath, C. J. Phys. Chem. B 1997, 101, 5875-5888.

1.22 Vailaya, A.; Horvath, C J. Phys. Chem. B 1998, 102, 701-718,

1.23 Dill, K. A, J. Phys. Chem. 1987, 91, 1980-1988.

1.24 Dorsey, J. G.; Dill, K. A. Chem. Rev 1989, 89, 331-346.

1.25 Montgomery M. E.; Greeen, M. A.; Wirth, M. J. Anal. Chem. 1992, 64, 1170-1175.

1.26 Burbage, J. D.; Wirth, M. J. J. Phys. Chem. 1992, 96, 5943-5948.

1.27 Montogmery, M. E.; Wirth, M. J. Anal. Chem. 1994, 66 680-684.

1.28 Wong, A. L.; Hunnicutt, M. L.; Harris, J. M. Anal. Chem. 1991, 63, 1076-1081.

1.29 Marqusee, J. A., Dill, K. A. J. Chem. Phys. 1986, 85, 434.

1.30 Dill, K. A; Naghizadeh, J., Marquesee, J. A. Ann. Rev. Phys. Chem. 1988, 39, 425-461.

1.31 Lochmuller, C. H.; Reese, C.; Aschman, A. J.; Breiner, S. J. J. Chromatogr. 1993, 656, 3-18.

1.32 Snyder, L. R., Kirkland, J. J.; Glajch, J. L. Practical HPLC Method Development, 1997, Wiley, New York.

1.33 Synder, L. R , Dolan, J. W.; Rigney, M, P. LC-GC, 1986, 4, 921-929.

1.34 Snyder, L. R.; Quarry, M. A. J. Liq, Chromatogr. 1987, 10, 1789-1820.

1.35 Stuart, J. D.; Lisi, D. D.; Snyder, L. R. J. Chromatogr. 1989, 485, 657-672.

1.36 Abraham, M. H., Roses, M.; Poole, C. F.; Poole, S. K. J. Phys. Org. Chem. 1997, 10,358-368.

613

1.37 Wang, A.; Tan, L. C.; Carr, P. W. J. Chromatogr., A 1999, 848, 21-37,

1.38 Karch, K.; Sebestian, I., Halasz, I. J. Chromatogr. 1976, 125, 3-16.

1.39 Hemetsberger, H.; Behrensmeyer, P.; Henning, J.; Ricken, H. Chromatographia 1979, 12, 71-76.

1.40 Lochmuller, C. H.; Wilder, D. R. J. Chromatogr. Sci. 1979, 17, 574-579.

1.41 Gilpin, R. K.; Squires, J. A. J. Chromatogr. Sci. 1981, 19, 195-199.

1.42 Gilpin, R. K.; Gangoda, M. E.; Krishen, A. E. J. Chromatogr. Sci. 1982, 20, 345-348.

1.43 Tchapla, A.; Heron, S.; Lesellier, E. J. Chromatogr. A 1993, 656, 81-112.

1.44 Schunk, T.C.; Burke, M.F. J. Chromatogr. A 1993, 656, 289-316.

1.45 Schunk, T.C.; Burke, M.F. Int. J. Environ. Anal. Chem. 1986, 25, 81-103,

1.46 Cole, L A,; Dorsey, J.G. Anal. Chem. 1992, 64, 1317-1323.

1.47 Lochmuller, C.H., Colbom, A,S,; Hunnicutt, M L,; Harris, J.M, J, Am, Chem, Soc. 1984, 106, 4077-4082.

1.48 Lochmuller, C.H.; Marshall, D.B.; Wilder, D R. Anal. Chem. Acta 1981, 130, 31-43,

1.49 Carr, J.W.; Harris, J.M. Anal. Chem. 1987, 59, 2546-2550.

1.50 Montgomery. M.E.; Greeen, M.A., Wirth, M.J, Anal. Chem. 1992, 64, 1170-1175.

1.51 Zulli, S.L.; Kovaleski, J.M.; Zhu, X R,; Harris, J,M,; Wirth, M,J, Anal. Chem. 1994, 66,1708-1712.

1.52 Beaufils, J.P.; Hennion, M.C.; Rosset, R. Anal. Chem. 1985, 57, 2593-2596.

1.53 Glinhka, C,J,; Sander, L,C,; Wise, S.A.; Berk, N.F. Mater, Res, Soc, Symp, Proc, 1990, 166,415-420,

1.54 Sander, L,C ; Glinka, C,J,; Wise, S.A, Anal, Chem, 1990, 62, 1099-1101,

614

1.55 Klatte, S, J.; Beck, T. L. J. Phys. Chem. 1995, 99, 16024-16032.

1.56 Klatte, S. J.; Beck, T. L. J. Phys. Chem. 1996, 100, 5931-5934,

1.57 Slusher, J. T.; Mountain, R. D. J. Phys. Chem. B 1999, 103, 1354-1362.

1.58 Sindorf, D.W.; Maciel, G.E. J. Am. Chem. Soc. 1983, 105, 1848-1851.

1.59 Kelusky, E C.; Fyfe, C.A. J. Am. Chem. Soc. 1986, 108, 1746-1749.

1.60 Shah, P.; Rogers, L.B.; Fetaer, J.C. J. Chromatogr. 1987, 388, 411-419.

1.61 Bayer, E.; Paulus, A.; Peters, B.; Laupp. G.; Reiners, J.; Albert, K. J. Chromatogr. 1986, 364, 25-37.

1.62 Albert, K.; Bayer, E. J. Chromatogr. 1991, 544, 345-370.

1.63 Albert, K.; Brindle, R.; Martin, P.; Wilson, ID, J. Chromatogr. A 1994, 665, 253-258.

1.64 Sander, L C ; Callis, J.B.; Field, L.R. Anal. Chem. 1983, 55, 1068-1075.

1.65 Ohtake, T ; Mino, N.; Ogawa, K. Langmuir 1992, 8, 2081-2083.

1.66 Tripp, C P.; Hair, M.L. Langmuir 1992, 8, 1120-1126.

1.67 Jinno, K.; Wu, J.; Ichikawa, M.; Takata, I. Chromatographia 1993, 37, 627-634.

1.68 Thompson, W.R.; Pemberton, IE. Anal. Chem. 1994, 66, 3362-3370.

1.69 Ho, M.; Cai, M.; Pemberton, IE. Anal. Chem. 1997, 69, 2613-2616.

1.70 Doyle, C.A.; Vickers, T.J.; Mann, C.K.; Dorsey, J.G. J. Chromatogr. A 1997, 779, 91-112.

1.71 Ho, M.; Pemberton, IE. Anal. Chem. 1998, 70, 4915-4920.



6 1 5

1.74 Pursch, M.; Sander, L.C.; Albert, K. Anal. Chem. 1996, 68, 4107-4113.

1.75 Brindle, R ; Albert, K. J. Chromatogr., A 1997, 757, 3-20.

1.76 Ellwanger, A., Brindle, R.; Albert, K. J. High Resolut. Chromatogr. 1997, 20, 39-45.

1.77 Pursch, M.; Sander, L. C.; Egelhaaf, H-J.; Raitza, M.; Wise, S. A.; Oelkrug, D.; Albert, K. J Am. Chem. See. 1999, 121, 3201-3213.

1.78 Gaber, B. P.; Peticolas, W. L. Biochim. Biophys. Acta 1977, 465, 260.

1.79 Larsson, K.; Rand, R. P. Biochim. Biophys. Acta 1973, 326, 245.

1.80 Snyder, R. G.; Hsu, S. L.; Krimm, S. Spectrochim. Acta 1978, 34A, 395.

1.81 Snyder, R. G.; Strauss, H. L.; Elliger, C. A. J. Phys. Chem. 1982, 86, 5145.

Chapter 2

2.1 Unger, K. Porous Silica: Its Properties and Use as Support in Column Liquid Chromatography, Elsevier; New York, 1979.

2.2 Piccoli, R. F. Ph. D. Dissertation, University of Arizona, 1998.

2.3 Sander, L. C ; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292.

2.4 Pursch, M.; Sander, L C ; Albert, K. Anal. Chem. 1996, 68, 4107-4113.

2.5 Berendsen, G. E ; de Galan, L. J. Liq. Chromatogr. 1978, 1, 561-586.

2.6 Sander, L. C.; Wise, S. A, Anal. Chem. 1984, 56, 504-510.

2.7 Sander, L. C.; Wise, S. A. LC-GC 1990, 8, 378-390.

2.8 Sander, L. C.; Wise, S. A. certificate of analysis, SRM 869a, Standard Reference Materials Program, NIST, Gaithersburg, MD, 1990.

2.9 Sander, L. C ; Wise, S. A. HRC CC, J. High Resolut. Chromatogr. Chromatogr. Commun. 1988, 11, 383-387.

616

2.10 Stevens, M. J. Langmuir 1999, 15, 2773-2778.

2.11 Wirth, M. 1; Fatunmbi, H. O. Anal. Chem, 1992, 64, 2783-2788.

Chapter 3

3.1 Templeton, A. C ; Wuelfmg, W. P.; Murray, R. W. Acc. Chem. Res. 2000, 33, 27.

3.2 Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir-Blodgett to Self-Assembly; Academic Press, Inc.; San Diego, 1991.

3.3 Ulman, A. Chem. Rev. 1996, 96, 1533.

3.4 Yellin, N.; Levin, I. W. Biochim. et Biophys. Acta 1977, 489, 177.

3.5 Gaber, B. P.; Peticolas, W. L. Biochim. Biophys. Acta 1977, 465, 260.

3.6 Erwin, J. A. Lipid and Biomembranes of Eukaryotic Microorganisms; Academic Press: New York, 1973.

3.7 Haupt, K.; Mosbach, K. Chem. Rev. 2000, 100, 2495.

3.8 Crooks, R M.; Ricco, A. J. Acc. Chem. Res. 1998, 31, 219.

3.9 Mendelsohn, R.; Moore, D. J. Chem. Phys. Lipids 1998, 96, 141.

3.10 Muniz-Miranda, M.; Puggelli, M.; Ricceri, R.; Gabrielli, G. Langmuir 1996, 12, 4417.

3 .11 Roberts, G. Langmuir-Blodgett Films; Plenum Press: New York, 1990.

3.12 Selle, C., Pohle, W.; Fritzsche, H. J. Mol. Struct. 1999, 480-481, 401.

3.13 Silvestro, L.; Axel sen, P. H. Chem. Phys. Lipids 1998, 96, 69.

3.14 Lewis, R. N. A. H.; McElhaney, R. N. Chem. Phys. Lipids 1998, 96, 9.

3.15 Ho, M.; Cai, M.; Pemberton, J. E. Anal. Chem. 1997, 69, 2613.

3.16 Ho, M., Pemberton, J. E. Anal. Chem. 1998, 70, 4915,

3 .17 Boenig, H V. Polyolefins; Structure and Properties; Elsevier Publishing Co. . Amsterdam, 1966.

3.18 Clauss, J.; Schmidt-Rohr, K.; Adam, A.; BoefFel, C.; Spiess, H. W. Macromolecules 1992, 25, 5208.

3.19 O'Brien, D. F.; Armitage, B.; Benedicto, A.; D E. Bennett; Lamparski, H. G.; Lee, Y.; Srisiri, W.; Sisson, T. M. Acc. Chem. Res. 1998, 31, 861.

3.20 Dorsey, J. G.; Dill, K. A. Chem. Rev. 1989, 89, 331.

3.21 Dorsey, J. G.; Cooper, W. T. Anal. Chem. 1994, 66, 857A.

3.22 Tchapla. A.; Heron, S.; Lesellier, E. J. Chromatogr. A 1993, 656, 81.

3.23 Schunk, T. C.; Burke, M. F. J. Chromatogr. A 1993, 656, 289.


3.25 Ellwanger, A.; Sander, L. C ; Bell, C. M.; Handel, H.; Albert, K. Solid State Nuc. Magn. Reson. 1997, 9, 191.

3.26 Pursch, M.; Brindle, R.; Ellwanger, A., Sander, L. C.; Bell, C. M.; Handel, H.; Albert, K. Solid State Nucl. Magn. Reson. 1997, 9, 191.

3.27 Sander, L. C ; Callis, J. B.; Field, L. R Anal. Chem 1983, 55, 1068.

3.28 Snyder, R. G.; Maroncelli, M.; Qi, S. P.; Strauss, H. L. Science 1981, 214, 188.

3.29 Hostetler, M. J., Stokes, J. J.; Murray, R. W. Langmuir 1996, 12, 3604.

3.30 Gericke, A.; Mendelsohn, R. Langmuir 1996, 12, 758.

3.31 Osman, M. A.; Seyfang, G.; Suter, U. W. J. Phys. Chem B 2000, 104, 4433.

3.32 Casal, H. L.; Yang, P. W.; Mantsch. H. H. Can, J, Chem, 1986, 64, 1544.

3.33 Hoffmann, H.; Mayer, U.; Brunner, H.; Krischanitz, A. Vibrational Spectroscopy 1995, 8, 151.

3.34 Riou, S, A.; Chien, B. T.; Hsu, S. L.; Stidham, H. D, J. Poly. Sci, Poly. Phys. 1997, 35, 2843.

618

3.35 Zhu, X.; Boiadjiev, V.; Mulder, J. A.; Hsung, R. P.; Major, R. C. Langmuir 2000, 16, 6766.

3.36 Allara, D. L.; Parikh, A. N.; Rondelez, R. Langmuir 1995, 11, 2357.

3.37 Grange, J. D. L.; Markham, J. L.; Kurkjian, C. R. Langmuir 1993, 9, 1749.

3.38 Harder, P.; Grunze, M.; Dahint, R.; Whitesides, G. M.; Laibinis, P. E. J. Phys. Cham. B 1998, 102, 426.

3.39 Valiokas, R.; Ostblom, M.; Svedhem, S.; Svensson, S. C. T.; Liedberg, B. J. Phys. Chem. B 2000, 104, 7565.

3.40 Shimada, H.; Grutzner, J. B.; Kozlowski, J. F.; McLaughlin, J. L. Biochemistry 1998, 37, 854.

3.41 Bryant, M. A.; Pemberton, J. E. J. Am. Chem. Soc. 1991, 113, 3629.

3.42 Bryant, M. A.; Pemberton, J. E. J Am. Chem. Soc. 1991, 113, 8284.

3.43 Kim, Y.; Strauss, H. L.; Snyder, R. G. J. Phys. Chem. 1989, 93, 7520.

3.44 Pemberton, J. E.; Cai, M. Fres. Anal. Chem. 2001, 369, 328.

3.45 Snyder, R. G.; Hsu, S. L.; Krimm, S. Spectrochim. Acta 1978, 34A, 395.

3.46 Snyder, R G.; Strauss, H. L.; Elliger, C. A. J. Phys. Chem. 1982, 86, 5145.

3.47 Strobl, G. R.; Hagedom, W. J. Polym. Sci. Polym. Phys. Ed. 1978, 16, 1181.

3.48 Thompson, W. R.; Pemberton, J. E. Anal. Chem, 1994, 66, 3362.

3.49 Wallace, B. A; Blount, E. R. Proc. Natl. Acad. Sci. USA 1979, 76, 1775.

3.50 Wunder, S. L. Macromolecules 1981, 14, 1024.

3.51 Lin-Vien, J. G.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press: New York, 1991.

3.52 Harrand, M. J. Chem. Phys. 1983, 79, 5639.

619

3.53 MacPhail, R. A.; Snyder, R. G ; Strauss, H. L. J. Chem. Phys. 1982, 77, 1118.

3.54 Wallach, D. F, H.; Varma, S. P.; Fookson, J. Biochim. Biophys. Acta 1979, 559, 153.

3.55 Larsson, K.; Rand, R P. Biochim. Biophys. Acta 1973, 326, 245.

3.56 Schoenfisch, M. H.; Pemberton, J. E. Langmuir 1999, 15, 509.

3.57 Bower, D I.; Maddams, W. F. The vibrational spectroscopy of polymers; Cambridge University Press: New York, 1989.

3.58 Abbate, S., Zerbi, G.; Wunder, S. L. J. Phys. Chem. 1982, 86, 3140.

3.59 Miranda, P. B.; Pflumio, V.; Saijo, H.; Shen, Y. R. J. Am. Chem. Soc. 1998, 120, 12092.

3.60 Keller, A. Philos. Mag. 1957, 2, 1171.

3.61 Niegisch, W. D.; Swan, P. R. J. Appl. Phys. I960, 31, 1906.

3.62 Shibukawa, T.; Gupta, V. D., Turner, R.; Dillon, J. H.; Tobolsky, A. V. Textile Res. J. 1962, 810.

Chapter 4

4.1 Scott, R. P. W.; Simpson, C. R. J. Chromatogr. 1980, 197, 11-20.

4.2 Morel, D.; Serpinet, J. J. Chromatogr. 1981, 214, 202-208.

4.3 Hennion, M. C., Picard, C.; Combellas, C.; Caude, M.; Rosset, R. J. Chromatogr. 1981,210,211-228.


4.5 Gilpin, R. K.; Gangoda, M E.; Krisen, A. E. J. Chromatogr. Sci. 1982, 20, 345-348.

4.6 Morel, D.; Serpinet, J.; Letoffe, J. M.; Claudy, P. Chromatographia 1986, 22, 103-108.

620

4.7 Cole, L. A.; Dorsey, J, G.; Dill, K, A. Anal. Chem. 1992, 64, 1324-1327.

4.8 Tchapla, A.; Heron, S, J. Chromatogr., A 1994, 684, 175-188.

4.9 Montgomery, M. E.; Green, M. A.; Wirth, M. J. Anal. Chem. 1992, 64, 1170-1175.

4.10 Bayer, E.; Paulus, A.; Peters, B.; Laupp. G; Reiners, J.; Albert, K. J. Chromatogr. 1986, 364, 25-37.

4.11 Sander, L. C.; Callis, J. B.; Field, L. R. Anal. Chem. 1985, 57, 2593-2596.

4.12 Doyle, C A.; Vickers, T. J.; Mann, C K.; Dorsey, J. G. J. Chromatogr., A 2000, 877, 25-39.

4.13 Bliesner, D. M.; Sentell, K. B. Anal. Chem. 1993, 65, 1819-1826.


4.15 Kelusky, E. C.; Fyfe, C. A. J, Am. Chem. Soc. 1986, 108, 1746-1749.

4.16 Sentell, K. B ; Bliesner, D M.; Shearer, S. T. Chem. Modif. Surf. 1994, 139, 190-202.

4.17 Shah, P.; Rogers, L. B.; Ffetzer, J. C. J. Chromatogr. 1987, 388, 411-419.

4.18 Ludes, M. D.; Wirth, M. J. Anal. Chem. 2002, 74, 386-393.

4.19 Klatte, S. J.; Beck, T. L. J. Phys. Chem. 1996, 100, 5931-5934.

4.20 Thompson, W. R., Pemberton, J. E. Anal. Chem. 1994, 66, 3362-3370.

4.21 Ho, M.; Pemberton, J, E. Anal. Chem. 1998, 70, 4915-4920.

4.22 Piccoli, R F. PhD Dissertation, 1998, University of Arizona.

4.23 Kamlet, M. J.; Abbound, J. M.; Abraham, M. H.; Taft, R. W. J. Org. Chem. 1983, 48, 2877-2887.

4.24 Sadek, P. C ; Carr, P. W.; Doherty, R. M.; Kamlet, M. J.; Taft, R. W.; Abraham, M. H. Anal. Chem. 1985, 57, 2971-2978.

621

4.25 Carr, P. W.; Doherty, R. M.; Kamlet, M. J.; Taft, R. W.; Melander, W.; Horvath, C. Anal, Chem. 1986, 58, 2674-2680.

4.26 Helburn, R. S.; Rutan, S. C.; Pompano, J. Mitchem, D.; Patterson, W. T. Anal. Chem. 1994, 66, 610-618.

4.27 Tan, L. C.; Carr. P. W. J. Chromatogr., A 1998, 799, 1-19.

4.28 Gritti, F., Felix, G,; Achard, M. F.; Hardouin, F. J. Chromatogr., A 2001, 922, 51-61.

4.29 Wang, A.; Carr, P. W. J. Chromatogr., A 2002, 965, 3-23.

Chapter 5

5.1 Dorsey, J.G.; Cooper, W.T. Anal. Chem. 1994, 66, 857A-867A.

5.2 Dorsey, J.G.; Dill, K.A. Chem. Rev 1989, 89, 331-346.

5.3 Tchapla. A.; Heron, S.; Lesellier, E. J. Chromatogr. A 1993, 656, 81-112.

5.4 Schunk, T.C.; Burke, M.F. J. Chromatogr. A 1993, 656, 289-316.

5.5 Gilpin, R.K.; Gangoda, M.E.; Krishen, A.E. J. Chromatogr. Sci. 1982, 20, 345-348.

5.6 Schunk, T.C.; Burke, M.F. Int. J. Environ. Anal. Chem. 1986, 25, 81-103.

5.7 Cole, L.A.; Dorsey, J.G. Anal. Chem. 1992, 64, 1317-1323,

5.8 Sander, L.C,; Wise, S,A, Anal, Chem, 1995, 67, 3284-3292,

5.9 Sander, L.C,; Pursch, M.; Wise, S,A, Anal, Chem. 1999, 71, 4821-4830,

5.10 Lochmuller, C.H.; Colbom, A.S.; Hunnicutt, M.L.; Harris, J.M. J. Am. Chem. Soc. 1984, 106, 4077-4082.

5.11 Lochmuller, C.H.; Marshall, D.B.; Wilder, D.R. Anal. Chem. Acta 1981, 130, 31-43.

5.12 Carr, J.W,; Harris, J,M, Anal, Chem, 1987, 59, 2546-2550,

622

5.13 Montgomery', M.E.; Greeen, M.A.; Wirth, M.J. Anal. Chem. 1992, 64, 1170-1175.

5.14 Burbage, J.D.; Wirth, M.J. J. Phys. Chem. 1992, 96, 5943-5948.

5.15 Montogmery, M.E.; Wirth, M J. Anal. Chem. 1994, 66 680-684.

5.16 Zulli, S.L.; Kovaleski, J.M.; Zhu, X.R ; Harris, J.M.; Wirth, M.J. Anal. Chem. 1994, 66,1708-1712.

5.17 Sindorf, D.W.; Maciel, G.E, J, Am. Chem. Soc. 1983, 105, 1848-1851.

5.18 Kelusky, E C , Fyfe, C.A. J. Am. Chem. Soc. 1986, 108, 1746-1749.

5.19 Shah, P.; Rogers, L.B.; Fetaer, J.C. J. Chromatogr. 1987, 388, 411-419.

5.20 Bayer, E.; Paulus, A.; Peters, B.; Laupp. G; Reiners, J.; Albert, K. J. Chromatogr. 1986, 364, 25-37.


5.22 Albert, K.; Brindle, R.; Martin, P.; Wilson, I.D. J. Chromatogr. A 1994, 665, 253-258.

5.23 Beaufils, J.P.; Hennion, M.C.; Rosset, R. Anal. Chem. 1985, 57, 2593-2596.

5.24 Glinhka, C.J.; Sander, L.C.; Wise, S.A.; Berk, N.F. Mater. Res Soc. Symp. Proc. 1990, 166,415-420.

5.25 Sander, L.C.; Glinka, C.J.; Wise, S.A. Anal. Chem. 1990, 62, 1099-1101.

5.26 Claudy, P.; Letoffe, J.M.; Gaget, C.; Morel, D.; Serpinet, J. J. Chromatogr. 1985, 329, 331-349.

5.27 Gonnet, C.; Morel, D.; Ramamnojinirina, E.; Serpinet, J., Claudy, P. J. Chromatogr, 1985, 330, 227-241.

5.28 Claudy, P.; LetofFe, J.M.; Gaget, C.; Morel, D.; Ramamnojinirina, E.; Serpinet, J. Bull. Soc. Chim. Fr. 1985, 1098-1102.

5.29 Morel, D.; Taber, K.; Serpinet, J., Claudy, P.; LetofFe, J.M. J. Chromatogr. 1987, 395, 73-84.

623

5.30 Sander, L.C.; Callis, IB.; Field, L.R. Anal. Chem. 1983, 55, 1068-1075.

5.31 Ohtake, T.; Mino, N.; Ogawa, K. Langmuir 1992, 8, 2081-2083.

5.32 Tripp, C.P ; Hair, M.L. Langmuir 1992, 8, 1120-1126.

5.33 Jinno, K.; Wu, J.; Ichikawa, M., Takata, I. Chromatographia 1993, 37, 627-634.

5.34 Thompson, W.R.; Pemberton, J.E. Anal. Chem. 1994, 66, 3362-3370.

5.35 Ho, M.; Cai, M.; Pemberton, J.E. Anal. Chem. 1997, 69, 2613-2616.

5.36 Doyle, C.A.; Vickers, T.J.; Mann, C.K.; Dorsey, J G. J. Chromatogr. A 1997, 779, 91-112.

5.37 Ho, M.; Pemberton, J.E. Anal. Chem. 1998, 70, 4915-4920.



5.40 Karch, K.; Sebestian, I; Halasz, I. J. Chromatogr. 1976, 122, 3-16.

5.41 Hemetsberger, H.; Behrensmeyer, P.; Henning, J.; Ricken, H. Chromatographia 1979, 12, 71-76,

5.42 Lochmuller, C.H.; Wilder, D R. J. Chromatogr. 1979, 17, 574-579,

5.43 Gilpin, R.K.; Squires, J.A. J, Chromatogr, Sci. 1981, 1, 561-586,


5.45 Klatte, S.J.; Beck, T.L. J. Phys. Chem. 1995, 99, 16024-16032.

5.46 Snyder, R.G.; Schachtschneider, J.H. Spectrochim. Acta 1963, 19, 85-116.

5.47 Snyder, R.G.; Schachtschneider, J.H. Spectrochim. Acta 1963, 19, 117-168.

5.48 Levin, W.W.; Hill, I.R. J. Chem. Phys. 1979, 70, 842-851.

624

5.49 Snyder, R.G.; Hsu, S.L.; Krimm, S. Spectrochim. Acta 1978, 34A, 395-406.

5.50 Snyder, R.G., Scherer, J R. J. Chem. Phys. 1979, 71, 3221-3228.

5.51 Synder, R.G.; Strauss, H.L.; Elliger, C A. J. Phys. Chem. 1982. 86, 5145-5150.

5.52 MacPhail, R.A.; Synder, R.G.; Strauss, H.L. J. Chem. Phys. 1982, 77, 1118-1137.

5.53 Young, C.W.; Koehler, J.S.; McKinney, D.S. J. Am. Chem. Soc. 1947, 69, 1410-1415.

5.54 Lin-Vien, J.G.; Colthup, N.B.; Fateley, W.G.; Grasselli, J.G.; In The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules, Academic Press; New York, 1991.

5.55 Wallach, D.F.H.; Varma, S.P.; Fookson, J. Biochim. Biophys. Acta 1979, 559, 153-208.

5.56 Gaber, B.P.; Peticolas, W.L. Biochim. Biophys. Acta 1977, 465, 260-274.

5.57 Harrand, M. J. Chem. Phys. 1983, 79, 5639-5651.

5.58 Yellin, N.; Levin, l.W. Biochim. Biophys. Acta 1977, 489, 177-190.

5.59 Wheeler, J.F., Beck, T.L.; Klatte, S.J.; Cole, L.A.; Dorsey, J.G. J. Chromatogr. A 1993,656,317-333.

5.60 Bell, C.M.; Sander, L.C.; Wise, S.A. J. Chromatogr. A 1997, 757, 29-39.

5.61 Jinno, K.; Lin, Y. Chromatographia 1995, 41, 311.

5.62 Sander, L.C.; Wise, S.A. Anal. Chem. 1989, 61, 1749-1754.


5.64 Morel, D.; Serpinet, J. J. Chromatogr, 1981, 214, 202-208.

5.65 Morel, D.; Serpinet, J. J. Chromatogr. 1982, 248, 231 -240.

5.66 Hansen, S.J., Callis, J.B. J. Chromatogr. Sci. 1981, 19, 195-199.

625

5.67 Pursch, M.; Strohschein, S.; Handel, H.; Albert, K. Anal. Chem. 1996, 68, 386-393.

5.68 Kelusky, E.G.; Fyfe, C.A. J. Am. Chem. Soc. 1986, 108, 1746-1749.

5.69 Gangoda, M E.; Gilpin, R.K. J. Magn. Reson. 1983 53, 140-143.

5.70 Sentell, K.B.; Henderson, A.N. Anal. Chim Acta 1991, 246, 139-149.

5.71 Nahum, A.; Horvath, C. J. Chromatogr. 1981, 203, 53-63.

5.72 Jinno, K.; Nagoshi, T ; Tanaka, N.; Okamoto, M, Fetzer, J.C.; Briggs, W.R. J. Chromatogr. 1988,436, 1-10.

5.73 Cole, L.A.; Dorsey, J.G.; Dill, K.A, Anal. Chem. 1992, 64, 1324-1327.

5.74 Hammers, W E.; Verschoor, P.B.A. J. Chromatogr. 1983, 282, 41-58.

Chapter 6


6.2 Pursch, M.; Sander, L. C.; Albert, K. Anal. Chem. 1996, 68, 4107-4113.

6.3 Pursch, M.; Vanderhart, D. L.; Sander, L. C.; Gu, X.; Nguyen, T.; Wise, S. A.; Gajewski, D. A. J. Am. Chem. Soc. 2000, 122, 6997-7011.

6.4 Montgomery. M. E.; Greeen, M. A.; Wirth, M. J. Anal. Chem. 1992, 64, 1170-1175.

6.5 Burbage, J. D., Wirth, M. J. J. Phys. Chem. 1992, 96, 5943-5948.

6.6 Montogmery, M. E.; Wirth, M. J. Anal. Chem. 1994, 66 680-684.

6.7 Zulli, S. L.; Kovaleski, J.M.; Zhu, X. R.; Harris, J. M.; Wirth, M. J. Anal. Chem. 1994, 66,1708-1712.

6.8 Kelusky, E. C.; Fyfe, C. A. J. Am. Chem. Soc. 1986, 108, 1746-1749.

6.9 Shah, P.; Rogers, L. B.; Fetzer, J. C. J. Chromatogr. 1987, 388, 411-419.

626

6.10 Bayer, E.; Paulus, A.; Peters, B.; Laupp. G.; Reiners, J.; Albert, K. J, Chromatogr. 1986, 364, 25-37.

6.11 Beaufils, J. P.; Hennion, M. C.; Rosset, R. Anal. Chem. 1985, 57, 2593-2596.

6.12 Sander, L. C.; Callis, J. B.; Field, L. R. Anal. Chem. 1983, 55, 1068-1075.

6.13 Doyle, C. A.; Vickers, T. J.; Mann, C. K.; Dorsey, J. G. J. Chromatogr. A 2000, 877, 25-39.

6.14 Sentell, K. B.; Bliesner, D. M.; Shearer, S. T. Chem. Mod. Surf. 1994, 139, 190-202.

6.15 Scott, R. P. W.; Simpson, C. F. J. Chromatogr. 1980, 197, 11-20.


6.17 Hennion, M. C.; Picard, C.; Combellas, C.; Caude, M.; Rosset, R. J. Chromatogr. 1981,210,211-228.


6.19 Gilpin, R. K.; Gangoda, M. E.; Krishen, A. E. J. Chrom. Sci. 1982, 20, 345-348.

6.20 Morel, D., Serpinet, J.; LetofFe, J. M.; Claudy, P. Chromatographia 1986, 22, 103-108.

6.21 Cole, L. A., Dorsey, J. G.; Dill, K. A. Anal. Chem. 1992, 64, 1324-1327.

6.22 Tchapla, A.; Heron, S. J. Chromatogr., A 1994, 684, 175-188.

6.23 Bliesner, D. M.; Sentell, K. B. Anal. Chem. 1993, 65, 1819-1826.

6.24 Klatte, S. J.; Beck, T L. J. Phys. Chem. 1996, 100, 5931-5934.


6.26 Horvath, C.; Melander, W.; Molnar, I. J. Chromatogr. 1976, 125, 129-156.

6.27 Karger, B. L.; Gant, J. R.; Hartkopf, A.; Weiner, P. H. J. Chromatogr. 1976, 128, 65-78.

627

6.28 Vailaya, A.; Horvath, C. J. Phys. Chem. B 1997, 101, 5875-5888.

6.29 Vailaya, A.; Horvath, C, J. Phys. Chem. B 1998, 102, 701-718.

6.30 Marqusee, J. A.; Dill, K. A. J. Chem. Phys. 1986, 85, 434.

6.31 Dill, K. A. J. Phys. Chem, 1987, 91, 1980-1988.

6.32 Dill, K. A.; Naghizadeh, J.; Marquesee, J. A. Ann. Rev. Phys. Chem. 1988, 39, 425-461.

6.33 Dorsey, J. G.; Dill, K. A. Chem. Rev. 1989, 89, 331-346.

6.34 Lochmuller, C. H.; Reese, C.; Aschman, A. J.; Breiner, S. J. J. Chromatogr. 1993, 656, 3-18.

6.35 Snyder, L, R ; Kirkland, J. J.; Glajch, J. L. Practical HPLC Method Development, 1997, Wiley, New York.

6.36 Synder, L. R.; Dolan, J. W.; Rigney, M. P. LC-GC, 1986, 4, 921-929.

6.37 Snyder, L. R.; Quarry, M. A. J. Liq. Chromatogr. 1987, 10, 1789-1820.

6.38 Stuart, J. D.; Li si, D. D.; Snyder, L. R. J, Chromatogr, 1989, 485, 657-672.

6.39 Abraham, M. H.; Roses, M.; Poole, C. F.; Poole, S, K. J, Phys, Org. Chem. 1997, 10, 358-368.

6.40 Wang, A.; Tan, L. C.; Carr, P. W, J. Chromatogr., A 1999, 848, 21-37.

6.41 Thompson, W. R.; Pemberton, J. E, Anal. Chem. 1994, 66, 3362-3370.

6.42 Ho, M.; Cai, M.; Pemberton, J. E. Anal. Chem. 1997, 69, 2613-2616.

6.43 Ho, M.; Pemberton, J. E. Anal. Chem. 1998, 70, 4915-4920.

6.44 Carr, J. W,; Harris, J, M. Anal. Chem. 1987, 59, 2546-2550.


6.46 Albert, K.; Brindle, R ; Martin, P.; Wilson, ID. J. Chromatogr., A 1994, 665, 253-258.

628

6.47 MacPhail, R. A.; Synder, R. G.; Strauss, H. L. J. Chem. Phys. 1982, 77, 1118-1137.

6.48 Snyder, R. G ; Schachtschneider, J. H. Spectrochim. Acta 1963, 19, 117-168.

6.49 Levin, W. W.; Hill, 1. R. J. Chem. Phys. 1979, 70, 842-851.

6.50 Snyder, R G ; Hsu, S. L., Krimm, S. Spectrochim. Acta 1978, 34A, 395-406.

6.51 Snyder, R. G.; Scherer, J. R J. Chem, Phys. 1979, 71, 3221-3228.

6.52 Synder, R. G.; Strauss, H. L.; Elliger, C. A. J. Phys. Chem, 1982, 86, 5145-5150.

6.53 Lin-Vien, J, G.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G.; In The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules, Academic Press: New York, 1991

6.54 Wallach, D. F. H.; Varma, S. P.; Fookson, J. Biochim. Biophys. Acta 1979, 559, 153-208.

6.55 Kamlet, M. J.; Abboud, J. M.; Abraham, M. H.; Taft, R. W. J. Org, Chem. 1983, 48, 2877-2887,

6.56 Kamlet, M. J.; Carr, P. W.; Taft, R. W.; Abraham, M. H. J. Am. Chem. Soc. 1981, 103, 6062-6066.

6.57 Kamlet, M. J., Doherty, R.; Taft, R. W.; Abraham, M. H. J. Am. Chem. Soc. 1983, 105, 6741-6743.

6.58 Taft, R. W.; Abboud, J. M.; Kamlet, M. J.; Abraham, M. H. J. Sol. Chem. 1985, 14, 153-175.

6.59 Sander, L, C , Glinka, C J.; Wise, S. A. Anal. Chem. 1990, 62, 1099-1101.

6.60 McGuffin, V. L.; Chen, S. J. Chromatogr., A. 1997, 762, 35-46.

6.61 Gilpin, R. K.; Gangoda, M. E.; Krishen, A. E. J, Chromatogr. Sci. 1982, 20, 345-348.

629

Chapter 7

7.1 Kamlet, M.J.; Doherty, R.; Taft, R.W.; Abraham, M.H. J. Am. Chem. Soc. 1983, 105, 6741-6743.

7.2 Kamlet, M.J.; Abboud, J.M.; Abraham, M.H ; Taft, R.W. J. Org. Chem. 1983, 48, 2877-2887.

7.3 Kamlet, M.J ; Carr, P.W.; Taft, R.W.; Abraham, M.H. J. Am. Chem. Soc, 1981, 103, 6062-6066.

7.4 Taft, R.W.; Abboud, J.M.; Kamlet, M.J.; Abraham, M.H. J. Sol, Chem. 1985, 14, 153-175.

7.5 Sander, L, C.; Callis, J. B.; Field, L. R. Anal. Chem. 1983, 55, 1068-1075.

7.6 Sander, L.C.; Wise, S.A Anal. Chem. 1995, 67, 3284-3292.

7.7 Pursch, M.; Sander, L.C.; Albert, K. Anal. Chem. 1996, 68, 4107-4113,

7.8 MacPhail, R,A,; Synder, R.G ; Strauss, H.L. J. Chem. Phys. 1982, 77, 1118-1137.

7.9 Lin-Vien, J.G.; Cohhup, N.B.; Fateley, W.G.; Grasselli, J.G.; In The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules, Academic Press; New York, 1991.

7.10 Wallach, D.F H,; Varma, S.P.; Fookson, J. Biochim, Biophys. Acta 1979, 559, 153-208,

7.11 Gaber, B P.; Peticolas, W.L. Biochim. Biophys. Acta 1977, 465, 260-274.

7.12 Harrand, M. J, Chem. Phys, 1983, 79, 5639-5651.

7.13 Yellin, N.; Levin, I.W. Biochim. Biophys. Acta 1977, 489, 177-190.

7.14 Snyder, R.G.; Schachtschneider, J.H. Spectrochim. Acta 1963, 19, 117-168,

7.15 Levin, W W,; Hill, I.R. J Chem. Phys. 1979. 70, 842-851.

7.16 Snyder, R.G ; Hsu, S.L., Krimm, S. Spectrochim. Acta 1978, 34A, 395-406.

7.17 Snyder, R.G ; Scherer, J R. J. Chem. Phys. 1979, 71, 3221-3228.

630

7.18 Synder, R.G.; Strauss, H.L., Elliger, C.A. J. Phys. Chem. 1982, 86, 5145-5150.

7.19 Pursch, M.; Vanderhart, D. L.; Sander, L. C.; Gu, X.; Nguyen, T.; Wise, S. A.; Gajewski, D. A. J. Am. Chem. Soc. 2000,122, 6997-701 ].

Chapter 8

8.1 Khong, T.M.; Simpson, C.F. Chromatographia 1987, 24, 385-394.

8.2 Smith, R.M.; Burr, C.M. J. Chromatogr. 1989, 475, 57-74.

8.3 Tanaka, N.; Kimata, K.; Ho soya, K.; Miyanishi, H.; Takeo, A. J. Chromatogr., A 1993, 656, 265-287.

8.4 Sander, L.C.; Wise, S.A. Anal. Chem. 1995, 67, 3284-3292.


8.6 Kamlet, M.J.; Abboud, J.M.; Abraham, M.H.; Taft, R.W J. Org. Chem. 1983, 48, 2877-2887.

8.7 Hansch, C.; Leo, A. Substituent Constants for Correlation Analysis in Chemistry and Biology, Wiley: New York, 1979.

8.8 Sangster, J. J. Phys. Chem. Ref. Data 1989, 18, 1111-1229.

8.9 Abraham, M.H.; Whiting, G.S.; Fuchs, R.; Chambers, E.J. J. Chem. Soc. Perkin Trans 2 1990, 2, 291-300.

8.10 Nobeli, I., Yeoh, S.L.; Price, S.L., Taylor, R. Chem. Phys. Lett. 1997, 280, 196-202.

8.11 Kitaigorodski, A.I. Organic Chemical Crystallography; Consultants Bureau; New York, 1961.

8.12 Molecular volumes estimated from values in Gavezzotti, A. J. Am. Chem. Soc. 1983, 105, 5220-5225.

8.13 Nahum, A.; Horvath, C. J. Chromatogr. 1981, 203, 53-63.

8.14 Gill, R.; Alexander, S.P.; Moffat, A.C. J. Chromatogr. 1982, 247, 39-45.

631

8.15 Sekulic, S.; Haddad, P.R.; J. Chromatogr. 1988, 459, 65-77.

8.16 McCalley, D.V. J. Chromatogr. 1993, 636, 213-220,

Chapter 9

9.1 Hofmeister, F. Arch. Exp. Pathol. Pharmakol. 1888, 24, 247-260.

9.2 Cacace, M. G.; Landau, E. M.; Ramsden, J. J. Quart. Rev. Biophys. 1997, 30, 241-277.

9.3 von Hippel, P. H.; Schleich, T. Acc. Chem. Res. 1969, 2, 257-265.

9.4 Baldwin, R, L. Biophysical Journal 1996, 71, 2056-2063.

9.5 Kalra, A.; Tugcu, N.; Cramer, S. M.; Garde, S. J. Phys. Chem. B 2001, 105, 6380-6386.

9.6 Smith, P. E. J Phys. Chem. B 1999, 103, 525-534,

9.7 Dussaud, A,; Vignes-Adler, M. Langmuir, 1997, 13, 581-589.

9.8 McDevit, W, F,; Long, F, A. J, Am, Chem. Soc. 1952, 74, 1773-1777,

9.9 Takahashi, M,; Tezuka, T, J. Chromatogr,, B 1997, 688, 197-203.

9.10 Nagata, Y ; Yamamoto, K.; Shimojo, T. J. Chromatogr. 1992, 575, 147-152.

9.11 Meynial, I.; Paquet, V.; Combes, D. Anal. Chem, 1995, 67, 1627-1631.

9.12 Lim, C. K.; Peters, T. J. J. Chromatogr. 1989, 461, 259-266.

9.13 Allinquant, B.; Musenger, C.; Schuller, E. J. Chromatogr. 1985, 326, 281-291.

9.14 Lurie, I. S.; Conver, T. S,; Ford, V, L, Anal. Chem. 1998, 70, 4563-4569.

9.15 Wolff, M.; Derstein, D ; Goeber, B. Pharmazie 1983, 38, 891-892.

9.16 Lurie, I. S.; Demchuk, S. M. J. Liq. Chromatogr. 1981, 4, 337-355,

632

9.17 Blanco, D.; Quintanilla, M. E.; Mangas, J. J.; Guitierrez, M. D. J. Liq. Chromatogr. Relat. Technol. 1996, 19, 2615-1621.

9.18 Schneider, A.; Gerbi, V.; Redoglia, M. Am. J. Enol. Vitic. 1987, 38, 151-155.

9.19 Jandera, P.; Churacek, J.; Bartosova, J. Chromatographia 1980, 13, 485-492.

9.20 Van de Venne, J. L. M.; Hendrikx, J. L. H. M.; Deelder, R. S. J. Chromatogr. 1978, 167, 1-16.

9.21 Distler, W. J Chromatogr. 1978, 152, 250-252.

9.22 Court, W A. J. Chromatogr. 1977, 130, 287-291.

9.23 Sander, L. C ; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292.


9.25 Nagaoka, S.; Hirota, N; Matsushita, T.; Nishimoto, K. Chem. Phys. Lett. 1982, 92, 498-502.

9.26 Shamsul Huq, A. K. M.; Lodhi, S. A. K. J. Phys. Chem. 1966, 70, 1354-1358.

9.27 dementi, E.; Mehl, J. von Niessen, W. J. Chem. Phys. 1971, 54, 508-520.

9.28 Sangster, J. J. Phys. Chem. Ref. Data, 1989, 18, 1111-1229.

Chapter 10

10.1 McCalley, D V. J. Chromatogr. 1993, 636, 213-220.

10.2 Nawrocki, J. J. Chromatogr., A 1997, 779, 29-72.

10.3 Miyabe, K.; Guiochon, G. J. Chromatogr., A 1999, 830, 263-274.

10.4 Gotmar, G.; Fornstedt, T.; Guiochon, G. J. Chromatogr., A 1999, 831, 17-35.

10.5 Hennion, M. C.; Picard, C.; Caude, M. J. Chromatogr. 1978, 166, 21-35.

10.6 Berendsen, G. E.; De Galan, E. J. Chromagotr. 1980, 196, 21-37.

633

10.7 TanakakN.; Sakagami, K.; Araki, M. J. Chromatogr. 1980, 199, 327-337.

10.8 Issaq, H. J. J. Liq. Chromatogr. 1981, 4, 1917-1931.

10.9 Krstulovic, A. M.; Colin, H.; Tchapla, A.; Guichon, G. Chromatographia 1983, 17, 228-230.


10.11 Hennion, M. C ; Picard, C.; Caude, M. J. Chromatogr. 1978, 166, 21-35.



10.14 Sander, L C.; Wise, S. A. Anal. Chem. 1999, 71, 4821-4830.

10.15 Nawrocki, J. Moir, D. L.; Szczepaniak, W. Chromatographia 1989, 28, 143-147.

10.16 Kimata, K.; Tanaka, N.; Araki, T. J. Chromatogr. 1992, 594, 87-96.

10.17 Engelhardt, H.; Lobert, T. Anal. Chem. 1999, 71, 1885-1892.

10.18 Horvath, Cs.; Melander, W ; Molnar, I. J. Chromatogr. 1976, 125, 129-156.

10.19 Colin, H.; Diez-Mesa, J. C.; Guiochon, G.; Czajkowska, T.; Miedziak, I. J. Chromatogr. 1978, 167, 41-???

10.20 Scott, R. P. W.; Simpson, C. F. J. Chromatogr. 1980, 197, 11-20.

10.21 Morel, D.; Serpinet, J. J. Chromatogr. 1981, 214-202-228.

10.22 Hennion, M. C.; Picard, C.; Combellas, C.; Caude, M.; Rosset, R. J. Chromatogr. 1981,210,211-228.

10.23 Gilpin, R. K.; Gangoda, M. E.; Krishen, A. E. J. Chromatogr. Sci. 1982, 20, 345-348.


10.25 Tchapla, A.; Colin, H.; Guiochon, G. Anal. Chem. 1984, 56, 621-625

634

10.26 Morel, D.; Serpinet, J.; Letoffe, J. M.; Claudy, P. Chromatographia 1986, 22, 103-108.

10.27 Dill, K. A. J. Phys. Chem. 1987, 91, 1980-1988.

10.28 Tchapla, A.; Heron, S.; Collin, H.; Guiochon, G, Anal. Chem. 1988, 60, 1443-1448.

10.29 Dorsey, J. G.; Dill, K. A. Chem. Rev. 1989, 89, 331-346.

10.30 Sentell, K. B.; Dorsey, J. G. Anal. Chem. 1989, 61, 930-934.

10.31 Cole, L. A; Dorsey, J. G.; Dill, K. A. Anal. Chem, 1992, 64, 1324-1327.

10.32 Lochmuller, C. H.; Reese, C.; Aschman, A. J.; Breiner, S. J. J. Chromatogr. A 1993, 656, 3-18.

10.33 Schunk, T. C.; Burke, M. F. J. Chromatogr. A 1993, 656, 289-316.

10.34 Scott, R. P. W. J. Chromatogr. A 1993, 656, 51-68.

10.35 Tchapla, A.; Heron, S. J. Chromatogr. A 1994, 684, 175-188.

10.36 Lochmuller, C. H.; Colbom, A. S.; Hunnicutt, M. L.; Harris, J. M. J. Am. Chem. Soc. 1984, 106, 4077-4082.

10.37 Rutan, S. C.; Harris, J. M. J. Chromatogr. 1993. 656, 197-215,

10.38 Ludes, M. D.; Wirth, M. J. Anal. Chem. 2002, 74, 386-393.

10.39 Wirth, M. J. Swinton, D. J.; Ludes, M. D. J. Phys. Chem. B 2003, 107, 6258-6268.

10.40 Pizzolatto, R. L.; Yang, Y. J.; Wolf, L. K.; Messmer, M. C. Analytica Chim. Acta 1999. 397, 81-92,

10.41 Li, X., Messmer, M. C. J, Chromatogr. A 2003, 984, 19-28.

10.42 Henry, M. C. Wolf, L. K.; Messmer, M. C. J. Phys. Chem. B 2003, 107, 2765-2770.

10.43 Sindorf, D. W.; Maceil, G. E. J. Am. Chem. Soc. 1983, 105, 1848-1851.

635

10.44 Bayer, E.; Paulus, A.; Peters, B.; Laupp, G.; Reiners. J.; Albert, K, J. Chromatogr. 1986, 364, 25-37.

10.45 Sentell, K. B.; Bliesner, D M.; Shearer, S. T. J. Chem. Modif. Surf. 1994, 139, 190-202.

10.46 Albert, K. J. Sep. Sci. 2003, 26, 215-224.

10.47 Sander, L. C.; Callis, J. G; Field, L. R. Anal. Chem. 1983, 55, 1068-1075.

10.48 Ohtake, T.; Mino, N.; Ogawa, K. Langmuir 1992, 8, 2081-2083.

10.49 Tripp, C. P.; Hair, M. L. Langmuir 1992, 8, 1120-1126.

10.50 Jinno, K.; Wu, J.; Ichikawa, M.; Takata, I. Chromatographia 1993, 37, 627-634.


10.52 Ranatunga, R. P. J. and Carr, P. W. Anal. Chem. 2000, 72, 5679-5692.

10.53 Abraham, M. H.; Whiting, G S., Fuchs, R.; Chambers, E. J. J. Chem. Soc. Perkin Trans. 2 1990, 291-300.

10.54 Abraham, M. H.; Whiting, G. S.; Carr, P. W.; Quyang, H. J. Chem, Soc. Perkin Trans. 2 1998, 1385-1390.

10.55 Abraham, M. H. J. Chem. Soc. Faraday Trans. 1 1984, 80, 153.

10.56 Abraham, M. H. J. Am. Chem. Soc. 1984, 104, 2085.

10.57 Functional group volumes estimated from values given in Gavezzotti, A. J. Am. Chem. Soc. 1983, 105, 5220-5225.

10.58 TSuzuki, S.; Houjou, H.; Nagawa, Y.; Hiratani, K. J. Phys. Chem. A 2000, 104, 1332-1336.

10.59 Konschin, J.; Tylli, H., Grundfetl-Forsius, C. J. Mol. Struct. 1981, 77, 51-57.

10.60 Schaefer, T.; Penner, G. H. J. Mol. Struct. 1987, 152, 179-183.

636

10.61 Kamlet, M. J.; Doherty, R., Taft, R. W.; Abraham, M. H. J. Am. Chem. Soc. 1983, 105, 6741-6743.

10.62 Taft, R. W.; Abboud, J. M.; Kamlet, M. J.; Abraham, M. H. J. Sol. Chem. 1985, 14, 153-175.

10.63 Klatte, S. J.; Beck, T. L. J. Phys. Chem. 1996, 100, 5931-5934.

10.64 Slusher, J. T.; Mountain, R. D J. Phys. Chem. B 1999, 103, 1354-1362.

10.65 Lochmuller, C. H.; Wilder, D. R. J. Chromatogr. Sci. 1979, 17, 574-579.

10.66 Gilpin, R. K.; Squires, J. A. J. Chromatogr. Sci. 1981, 1, 561-586,

10.67 Melander, W. R.; Chen, B.; Horvath, C. J. Chromatogr, 1979, 185, 99-109.

10.68 Crushka, E.; Colin, H.; Guiochon, G. J. Chromatogr, 1982, 248, 325-339.

10.69 Miyabe, K.; Guiochon, G. Anal. Chem. 2002, 74, 5982-5992.

Chapter 11

11.1 Jungbauer, A.; Kaltenbrunner, O. Biotechnol. Bioeng. 1996, 52, 223-226.

11.2 Gerberding, S.J.; Byers, C.H. J. Chromatogr. A 1998, 808, 141-151.

11.3 Dubinina, N.I.; Kurenbin, O.I.; Tennikova, T.B. J, Chromatogr, A 1996, 753, 217-225.

11.4 Kurganov, A.A.; Davankov, V.A.; linger, K.K. J. Chromatogr. 1991, 548, 207-214.

11.5 Carter, A J.; Muller, R.E. J. Chromatogr. 1990, 527, 31-39.

11.6 Hansen, L.D.; Richter, B E.; Rollins, D.K., Lamb, ID.; Eatough, D.J. Anal. Chem. 1979, 51, 633-637.

11.7 Dabrowski, A. Adsorption and its Applications in Industry and Environmental Protection, Vol. 2. Elsevier Science; New York, 1999.

11.8 Cassidy, R.M.; Elchuk, S. J. Liq. Chromatogr. 1981, 4, 379-398.

637

11.9 Barkley, D.J ; Blanchette, M.; Cassidy, R.M.; Elchuk, S. Anal. Chem. 1986, 58, 2222-2226.

11.10 Boland, D.M.; Burke, M.F.; Mitchell, T.; Madley, P. J. Anal. Tox. 2001, 25, 602-606.

11.11 Weiss, J. Ion Chromatography, Second Edition VCH Publishers, Inc.: New York, 1995.

11.12 Small, H. Ion Chromatography, Plenum Press; New York, 1989.

11.13 Nightingale, E.R, J. Phys. Chem. 1959, 63, 1381-1387.

11.14 Gordon, J.E. J. Phys. Chem. 1962, 66, 1150-1158.

11.15 Hofmeister, F. Arch. Exp. Pathol. Pharmakol. 1888, 24, 247-260.

11.16 Cacace, M. G.; Landau, E. M.; Ramsden, J. J. Quart. Rev. Biophys. 1997, 30, 241-277.

11.17 von Hippel, P. H.; Schleich, T. Acc. Chem. Res. 1969, 2, 257-265.

11.18 Baldwin, R. L. Biophysical Journal 1996, 71, 2056-2063.

11.19 Kalra, A.; Tugcu, N.; Cramer, S. M.; Garde, S. J. Phys. Chem. B 2001, 105, 6380-6386.

11.20 McDevit, W. F.; Long, F. A. J. Am. Chem. Soc. 1952, 74, 1773-1777.

11.21 Selemenev, V.F.; Zagorodni, A. A. React. Funct. Polym. 1999, 39, 53-62.

11.22 Hume, A.S.; Holland, W.C.; Fry, F. Spectrochim. Acta 1968, 24A, 786-788.

11.23 Berg, R.W. Spectrochim. Acta 1978, 34A, 655-659.

11.24 Ratcliffe, C.I.; Ripmeester, J.A. Can. J. Chem. 1986, 64, 1348-1354.

11.25 Blears, D.J.; Danyiuk, S.S.; Bock, E. J. Phys. Chem. 1968, 72, 2269-2270.

11.26 Dufourcq, J.; Lemanceau, B. J. Chim. Phys. et Phys. Chim. Biol. 1970, 67, 9-18.

11.27 Slusher, J.T.; Cummings, P.T. J. Phys. Chem. B 1997, 101, 3818-3826.

638

11.28 Mo, H., Wang, A.; Wilkinson, P.S.; Pochapsky, T.C, J. Am. Chem. Soc. 1997, 119, 11666-11673.

11.29 Luzhkov, V.B.; Osterberg, F.; Acharya, P.; Chattopadhyaya, J.; Aqvist, J Phys. Chem. Chem. Phys. 2002, 4, 4640-4647.

11.30 Garcia-Tarres, L.; Guardia, E. J. Phys. Chem. B 1998, 102, 7448-7454.

11.31 Pal, M.; Agarwal, A.; Khandelwal, D P.; Bist, H.D. J. Mol. Stract. 1984, 112, 309-316.

11.32 Degen, I.A.; Newman, G.A. Spectrochim. Acta 1993, 49A, 859-887.

11.33 Edwards, H. G. M. Spectrochim. Acta 1989, 45A, 715-719.

11.34 Edwards, H G. M.; Lewis, I. R. J. Mol. Struct. 1993, 301, 37-45.

11.35 Lowry, S. R.; Mauritz, K A. J. Am. Chem. Soc. 1980, 102, 4665-4667.

11.36 Whittington, D.; Millar, J. R. J. Appl. Chem. 1968, 18, 122-128.

11.37 Buijs, K. J. Inorg. Nucl. Chem. 1962, 24, 229-238.

11.38 Buijs, K.; Strasheim, A. Spectrochim. Acta 1961, 17, 388-392.

11.39 Ratcliffe, C I.; Dunell, B. A. J. Chem. Soc., Faraday Trans. 2, 1981, 77, 2169-2180.

11.40 Crumrine, D. S.; Gillece-Castro, B. J. Org. Chem. 1985, 50, 4408-4410,

11.41 O'Connel, E. M.; Peiffer, D. G., Root, T. W.; Cooper, S. L. Macromolecules 1996, 29, 2124-2130.

11.42 Ennari, J.; Neelov, I; Sundholm, F. Comp. Theor. Polym. Sci. 2000, 10, 403-410.

11.43 Ennari, J.; Pietila, L.; Virkkunen, V.; Sundholm, F. Polymer 2002, 43, 5427-5438.

11.44 Covington, A. K.; Thompson, R. J. Solution Chem. 1974, 3, 603-617.

11.45 Edwards, H. G. M.; Smith, D. N. J. Mol. Struct. 1990, 238, 27-41.

639

11.46 Mattsson, B.; Brodin, A.; Torell, L. M.; Rinne, H ; Hamara, J.; Sundholm, F., Jacobsson, P. Solid State Ionics, 1997, 97, 309-314.

11.47 Edwards, H. G. M.; Brown, D. R., Dale, J. A.; Plant, S. Vibrational Spectroscopy, 2000, 24, 213-224.

11.48 Edwards, H. G M.; Brown, D. R.; Dale, J A.; Plant, S. J. Mol, Struct. 2001, 595, 111-125.

11.49 Sollinger, S.; Dianantoglou, M. J. Raman Spectosc. 1997, 28, 811-817.

11.50 Ho, M.; Cai, M.; Pemberton, J. E. Anal. Chem. 1997, 69, 2613-2616.

11.51 Ho, M.; Pemberton, J. E. Anal. Chem. 1998, 70, 4915-4920.

11.52 Petersen, S. Ann. N. Y. Acad. Sci. 1953, 57, 144-158.

11.53 Piccoli, R. F. PhD Dissertation, University of Arizona, 1998.

11.54 Gonzalez, R. D. PhD Dissertation, University of Arizona, 2001.

11.55 Boland, D. M. PhD Dissertation, University of Arizona, 2001.

11.56 Marcus, Y. J. Chem. Soc., Faraday Trans. 1 1987, 83, 339-349.

11.57 Bard, A. J.; Faulkner, L. R. Electrochemical Methods, John Wiley and Sons; New York, 1980.

11.58 Unger, K. K. Porous Silica; Its Properties and Use as Support in Column Liquid Chromatography, Elsevier; New York, 1979.

11.59 de Beer, W. H. J. Spectrochim. Acta 1981, 37A, 1099-1107.

11.60 Nakagawa, I; Walter, J. L. J. Chem. Phys. 1969, 51, 1389-1397.

11.61 Vollmar, P. M. J. Chem. Phys. 1963, 39, 2236-2248.

11.62 Turner, D. J. J. Chem. Soc. Faraday Trans. 2 1972, 68, 643-648.

11.63 Goypiron, A.; de Villepin, J.; Novak, A. J. Raman Spec. 1980, 9, 297-303.

11.64 Preston, C. M.; Adams, W. A. J. Phys. Chem. 1979, 83, 814-821.

640

11.65 Miller, A. G.; Macklin, J. W. J. Phys. Chem. 1985, 89, 1193-1201.

11.66 Pracht, G.; Lange, N.; Lutz, H. D Thermochim. Acta 1997, 293, 13-24.

11.67 Byoung-Koo, C.; Lockwood, D. J. Solid State Commun. 1989, 72, 133-137.

11.68 Niaura, G,; Gaigalas, A. K.; Vilker, V. L. J. Phys. Chem. B 1997, 101, 9250-9262.

11.69 Mercadie, J. Compt. Rend. 1945, 221, 581-583.

11.70 Samoilov, O. Y.; Hu, K.; Nosova, T. A. Zhumal. Strukturnio KJiimii, 1960, 1, 131-134.

11.71 Conway, B. E. Ionic Hydration in Chemistry and Biophysics, Elsevier: New York, 1981,

11.72 Kielland, J. J. Am. Chem. Soc. 1937, 59, 1675-1680.

11.73 Bartucci, R.; Belsito, S,; Sportelli, L, Chem. Phys. Lipids, 1996, 79, 171-180.

Chapter 12

12.1 Zhao, J.; Carr. P. W. Anal. Chem. 2000, 72, 302-309.

12.2 Nikolov, Z. L.; Reilly, P. J. J. Chromatogr. 1985, 325, 287-293.

12.3 Waksmundzka-Hajnos, M.; Petruczynik, A.; Hawryl, A. J. Chromatogr., A 2001, 919, 39-50,

12.4 Gauvin, A.; Pinguet, F.; Poujol, S.; Astre, C.; Bressolle, F. J. Chromatogr., B 2000, 748, 389-399.

12.5 Kim, I. W,; Lee, H, S.; Lee, Y. K.; Jang, M, D ; Park, J. H J. Chromatogr,, A 915, 2001, 35-42.

12.6 Gelsema, W. J.; De Ligny, C. L, J, Chromatogr. 1990, 498, 237-240.

12.7 Durig, J. R.; Drew, B. R.; Koomer, A., Bell, S. Phys. Chem. Chem. Phys. 2001, 3, 766-775.

641

12.8 Ho, M.; Cai, M.; Pemberton, J.E. Anal. Chem. 1997, 69,2613-2616.

12.9 Ho, M.; Pemberton, IE. Anal. Chem. 1998, 70, 4915-4920.

12.10 Swanson, B.; Shriver, D. F ; Ibers, J. A. Inorg. Chem. 1969, 8, 2182-2189.

12.11 Swanson, B.; Shriver, D. F. Inorg. Chem. 1970, 9, 1406-1416.

12.12 Shriver, D. F.; Swanson, B. Inorg. Chem. 1971, 10, 1354-1365.

12.13 Cha, J. -N., Cheong, B. S.; Cho, H. -G. J. Phys. Chem. 2001, 105, 1789-1796.

12.14 Fawcett, W. R.; Liu, G.; Kessler, T. E. J. Phys. Chem. A 1993, 97, 9293-9298.

12.15 Kamlet, M. J.; Abboud, J. M.; Abraham, M. H.; Taft, R W. J. Org. Chem. 1983, 48, 2877-2887.

12.16 SchafF, J. E.; Roberts, J. T. Langmuir 1999, 15, 7232-7237.

12.17 Lin-Vien, D.; Cothup, N. B.; Fately, W. G.; Grasselli, J. G. The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules Academic Press: New York^ 1991, pp 105-113.

12.18 Tanabe, K. Chem. Phys. 1979, 38, 125-129.

12.19 Griffiths, J. E. J. Chem. Phys. 1973, 59, 751-758.

12.20 Ishida, H.; Koeing, J. L. Appl. Spectrosc. 1978, 32, 469-4479.

12.21 Durig, J. R.; Hellams, K. L.; Mulligan, J. H. Spectrochim. Acta, 1972, 28A, 1039-1057.

12.22 Palafox, M. A.; Boggs, J. E. J. Mol. Struct., 1993, 284, 23-35.

12.23 Szafranski, C. A,, Tanner, W.; Laibinis, P. E.; Garrell, R. L. Langmuir 1998, 14, 3570-3579.

Chapter 13

13.1 Abel, E. W.; Pollard, F. H.; Uden, P. C ; Nickless, G. J. Chromatogr. 1966, 22, 23-28.

642

13.2 Roumeliotis, P.; Linger, K. K. J. Chromatogr. 1978, 149, 211-224.

13.3 Unger, K. K.; Gimpel, M. G. J. Chromatogr. 1979, 180, 93-102.

13.4 Kirkland, J. J. Anal. Chem. 1971, 43, 36A-48A.

13.5 Kirkland, J. J. J. Chromatogr. Sci. 1972, 10, 593-599.

13.6 Kirkland, J. J. Chromatographia 1975, 8, 661-668.


13.8 Claudy, P.; Letofte, J. M ; Gaget, C ; Morel, D.; Serpinet, J. J. Chromatogr. 1985, 329, 331-349.

13.9 Jaroniec, M.; Lin, S.; Gilpin, R. K. Chromatographia 1991, 32, 13-18.

13.10 Dong, M. W.; DiCesare, J. L. J. Chromatogr. Sci. 1982, 20, 517-522.

13.11 Fetzer, J. C ; Biggs, W. R.; Jinno, K. Chromatographia 1986, 21, 439-442,

13.12 Mattus, A.; Ohmacht, R. Chromatographia 1990, 30, 318-322.

13.13 Lesellier, E.; Tchapla, A.; Krstulovic, A. M. J. Chromatogr. 1993, 645, 29-39.

13.14 Sander, L. C.; Sharpless, K. E.; Craft, N. E.; Wise, S. A. Anal. Chem. 1994, 66, 1667-1674.

13.15 Bell, C. M.; Sander, L. C.; Fetzer, J. C.; Wise, S. A, J. Chromatogr., A 1996, 735, 37-45.

13.16 Sander, L. C.; Sharpless, K. E.; Pursch, M. J. Chromatogr., A 2000, 880, 189-202.

13.17 Hennion, Claude, Picard, M. C.; J.

13.18 Sander, L. C.; Wise, S. A. Anal. Chem, 1984, 56, 504-510.



643

13.21 Sander, L. C.; Wise, S. A. HRC CC, J. High Resolut. Chromatogr. Chromatogr. Commun. 1988, 11, 383-387,

13.22 Sadek, P. C.; Koester, C. J.; Bowers, L. D. J. Chromatogr. Sci, 1987,25, 489-493.

13.23 Nawrocki, J.; Moir, D L.; Szczepaniak, W. Chromatographia 1989, 28, 143-147.

13.24 Kimata, K.; Tanaka, N.; Araki, T. J. Chromatogr. 1992, 594, 87-96.

13.25 Cecchi, T.; Pucciarelli, F.; Passamonit, P.; Ferraro, S. J. Liq. Chromatogr. & Rel. Technol. 1999, 22, 429-440.

13.26 Meyer, R. F.; Hart wick, R. A. Anal. Chem. 1984, 56, 2211-2214.

13.27 McNair, H.; Bowermaster, J. J. High Resolut. Chromatogr. Chromatogr. Commun. 1987, 10, 27-31.

13.28 Parikh, A. N.; Schivley, M. A.; Koo, E., Seshadri, K.; Aurentz, D., Mueller, K.; Allara, D L. J. Am. Chem. Soc. 1997, 119, 3135-3143.

13.29 Zhao, Z.; Zhang, L.; Lin, Y. J. Colloid Interface Sci., 1994, 166, 23-28.

13.30 Doering, J,; Lagaly, G.; Beneke, K; Dekany, I. Colloids and Surf., A, 1993, 71, 219-231.

13.31 Fuji, M., Takei, T.; Watanabe, T.; Chikazawa, M. Colloids and Surf., A, 1999, 154, 13-24.


13.33 Sander, L. C.; Wise, S. A. certificate of analysis, SRM 869a, Standard Reference Materials Program, NIST, Gaithersburg, MD, 1990.

13.34 Sander, L. C.; Wise, S. A. HRC CC, J. High Resolut. Chromatogr. Chromatogr. Commun. 1988, 11, 383-387.

13.35 Guo, Y.; Colon, L. A. Anal. Chem. 1995, 67, 2511-2516.


13.37 Wang, D.; Chong, S. L.; Makik, A. Anal. Chem. 1997, 69, 4566-4576.

644

13.38 Hayes, J. D.; Malik, A. Anal. Chem. 2000, 72, 4090-4099.

13.39 Tang, Q.; Lee, M. L. J. High Resol. Chromatogr. 2000, 23, 73-80.

13.40 Ratnayake, C. K.; Oh, C. S.; Henry, M. P. J. High Resol. Chromatogr. 2000, 23, 81-88.

Chapter 14

14.1 Ranatunga, R P. J. and Carr, P. W. Anal. Chem. 2000, 72, 5679-5692.

14.2 Gilpin, R. K.; Squires, J. A. J. Chromatogr. Sci. 1981, 1, 561-586.

14.3 Melander, W. R ; Chen, B.; Horvath, C. J. Chromatogr, 1979, 185, 99-109.

14.4 Crushka, E.; Colin, H.; Guiochon, G. J. Chromatogr, 1982, 248, 325-339.

14.5 Miyabe, K.; Guiochon, G Anal. Chem. 2002, 74, 5982-5992.

14.6 Sander, L C ; Wise, S. A. Anal. Chem. 1984, 56, 504-510.


14.8 Gordon, J.E. J. Phys. Chem. 1962, 66, 1150-1158.

14.9 Lowry, S. R.; Mauritz, K. A. J. Am. Chem. Soc. 1980, 102, 4665-4667.

14.10 Guo, Y.; Colon, L, A. Anal. Chem. 1995, 67, 2511-2516.


14.12 Hayes, J. D ; Malik, A. Anal. Chem. 2000, 72, 4090-4099.

14.13 Tang, Q , Lee, M. L, J. High Resol. Chromatogr. 2000, 23, 73-80.

14.14 Ratnayake, C. K.; Oh, C. S.; Henry, M. P. J. High Resol. Chromatogr. 2000, 23, 81-88.

Documents

repository.arizona.edu · 2020. 4. 2. · THE UNIVERSITY OF ARIZONA ® GRADUATE COLLEGE As members of the Final Examination Ccnnnittee, we certify that we have read the dissertation