repository.arizona.edu · 2020. 4. 2. · the university of arizona ® graduate college as members...
TRANSCRIPT
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. N^al 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
TABLE OF CONTENTS - Continued
CHAPTER 4; SOLVENT EFFECTS ON THE CONFORMATIONAL ORDER OF COMMERCIAL SOLID PHASE EXTRACTION ALKYLSILANE STATIONARY PHASES 160
Introduction 160
Goals of This Work 162
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
Goals of This Work 179
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
TABLE OF CONTENTS - Continued
Conclusions and Future Directions 209
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
Goals of This Work 218
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
Conclusions and Future Directions 249
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
TABLE OF CONTENTS - Continued
Goals of This Work 279
Methodology 280
Raman Spectroscopy of High-Density Alkylsilane Stationary Phases in Solvent 281
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
Goals of This Work 320
Methodology 321
Raman Spectroscopy of High-Density Alkylsilane Stationary Phases in Solvent 321
Effect of Aromatic Compounds on the Order of These High-Density Stationary Phases 325
Effects of Temperature 340
Conclusions 347
11
TABLE OF CONTENTS - Continued
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
Goals of This Work 362
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
Conclusions and Future Directions 381
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
TABLE OF CONTENTS - Continued
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
Goals of This Work 467
Methodology 468
13
TABLE OF CONTENTS - Continued
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
Conclusions and Future Directions 529
CHAPTER 12: PRELIMARY STUDIES ON THE CHARACTERIZATION OF OTHER COMMERCIAL STATIONARY PHASES FOR REVERSED-PHASE LIQUID CHROMATOGRAPHY 532
Introduction 532
Goals of This Work 533
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
TABLE OF CONTENTS - Continued
Interactions Between Solvent and Isolute Phenyl 549
Conclusions and Future Directions 552
CHAPTER 13; ALKYLSILANE-BASED STATIONARY PHASES VIA CONTROLLED SOLUTION MODIFIER TECHNIQUES: SYNTHETIC APPROACH, CHARACTERIZATION, AND CHROMATOGRAPHIC PERFORMANCE 554
Introduction 554
Goals of This Work 556
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
Conclusions and Future Directions 595
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
TABLE OF CONTENTS - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF FIGURES - Continued
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
LIST OF TABLES - Continued
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
LIST OF TABLES - Continued
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,' ^infrared (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 analj^es
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. ^^ot 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^ /^O
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
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 analj^es, 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 layers^and 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 ^iflJ
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.
Conclusions and Future Directions
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
Octadecylsilane-modified silica-based stationary phases were prepared and
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).
Conclusions and Future Directions
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
Structure-Function Relationships in Octadecylsilane Stationary Phases by
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
Octadecylsilane-modified silica-based stationary phases were prepared and
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.
Samples were prepared by placing between 25 and 100 mg of stationary phase
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
acquisition. Raman spectra were collected using 100 mW of 532 nm 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 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 fi^om
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
Acetonitrile 0.75 0.31 0.19
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 ! ^ ^Acetonitrile
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
This appendix contains Raman spectra in the v(C-H) for the five high-density
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
Structure-Function Relationships in Octadecylsilane Stationary Phases by
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
Octadecylsilane-modified silica-based stationary phases were prepared and
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.
Samples were prepared by placing between 25 and 100 mg of stationary phase
material into a 5-mm dia NMR tube; 200 piL 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, 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
This appendix contains Raman spectra in the v(C-H) for the five high-density
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.
Samples were prepared by placing between 25 and 100 mg of stationary phase
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.
Conclusions and Future Directions
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 ^iL,
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
75:25 Methanol/ Water
50:50 Methanol/ Water
Aniline 100% Methanol
75:25 Methanol/ Water
50:50 Methanol/ Water
-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
% Water in Methanol [v/v]
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 O^An,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
% Water in Methanol [v/v]
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
% Water in Methanol [v/v]
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
% Water in Methanol [v/v]
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)
^Assignments 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" ) ^Assignment
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. ^Assignments 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, ^Assignments 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. ^Assignments 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.
Conclusions and Future Directions
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.
Samples were prepared by placing between 25 and 100 mg of stationary phase
material into a 5-mm dia NMR tube; 200 pL 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.
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) + ^a(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
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
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
^Acetonitrile
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.
Conclusions and Future Directions
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 ^iL 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.
Sample CJ04-1 was prepared by adding 440 mg of Krom-200 silica to 20 mL
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.
Sample CJ05-1 was prepared by adding 440 mg of Krom-200 silica to 20 mL
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 ^iL. 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^) ^ACCHJ)
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 ^On I ^OH
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 ^ ^O^ ' ""OH ^ HO'^ ^
OH OH OH OH OH OH
Si ./ I \
Water
OH
HO,P?O.P? J Si Si
OH
OH
OH
OH HOs^iyOH OH
OH ^
OH Si Si
OH OH OH Si
HO^i^O^OH OH HO'A^OH 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.
Conclusions and Future Directions
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
Conclusions and Future Directions
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.
1.17 Sander, L. C.; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292.
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.
1.72 Doyle, C.A.; Vickers, T.J.; Mann, C.K.; Dorsey, J.G. J. Chromatogr. A 2000, 877, 25-39.
1.73 Doyle, C.A.; Vickers, T.J.; Mann, C.K.; Dorsey, J.G. J. Chromatogr. A 2000, 877, 41-59.
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.
3.24 Sander, L. C.; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292.
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.
4.4 Morel, D.; Serpinet, J. J. Chromatogr. 1983, 248, 231-240.
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.
4.14 Wong, A. L.; Hunnicutt, M. L.; Harris, J. M. Anal. Chem. 1991, 63, 1076-1081.
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.
5.21 Albert, K.; Bayer, E. J. Chromatogr. 1991, 544, 345-370.
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.
5.38 Doyle, C.A.; Vickers, T.J.; Mann, C.K.; Dorsey, J.G. J. Chromatogr. A 2000, 877, 25-39.
5.39 Doyle, C.A.; Vickers, T.J.; Mann, C.K.; Dorsey, J.G. J. Chromatogr. A 2000, 877, 41-59.
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,
5.44 Pursch, M.; Sander, L.C.; Albert, K. Anal. Chem. 1996, 68, 4107-4113.
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.
5.63 Morel, D.; Serpinet, J. J. Chromatogr. 1980, 200, 95-104.
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
6.1 Sander, L. C.; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292.
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.
6.16 Morel, D.; Serpinet, J. J. Chromatogr. 1981, 214, 202-208.
6.17 Hennion, M. C.; Picard, C.; Combellas, C.; Caude, M.; Rosset, R. J. Chromatogr. 1981,210,211-228.
6.18 Morel, D.; Serpinet, J. J. Chromatogr. 1982, 248, 231-240.
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.
6.25 Wong, A. L.; Hunnicutt, M. L.; Harris, J. M. Anal. Chem. 1991, 63, 1076-1081.
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.
6.45 Albert, K.; Bayer, E. J. Chromatogr. 1991, 544, 345-370.
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.
8.5 Pursch, M.; Sander, L.C.; Albert, K. Anal. Chem. 1996, 68, 4107-4113.
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.
9.24 Pursch, M.; Sander, L. C.; Albert, K. Anal. Chem. 1996, 68, 4107-4113.
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.
10.10 Sander, L. C.; Wise, S. A. Anal. Chem. 1987, 59, 2309-2313.
10.11 Hennion, M. C ; Picard, C.; Caude, M. J. Chromatogr. 1978, 166, 21-35.
10.12 Sander, L. C.; Wise, S. A. Anal. Chem. 1984, 56, 504-510.
10.13 Sander, L. C.; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292.
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.
10.24 Morel, D.; Serpinet, J. J. Chromatogr. 1982, 248, 231-240.
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.
10.51 Pursch, M.; Sander, L. C.; Albert, K. Anal. Chem. 1996, 68, 4107-4113.
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.
13.7 Morel, D.; Serpinet, J. J. Chromatogr. 1982, 248, 231-240.
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.
13.19 Sander, L. C.; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292.
13.20 Sander, L. C.; Wise, S. A. LC-GC 1990, 8, 378-390.
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.
13.32 Sander, L. C.; Wise, S. A. LC-GC 1990, 8, 378-390.
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.
13.36 Guo, Y.; Colon, L. A. J. Microcolumn Sep. 1995, 7, 485-491.
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.
14.7 Sander, L. C.; Wise, S. A. Anal. Chem. 1995, 67, 3284-3292.
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.
14.11 Guo, Y.; Colon, L. A. J. Microcolumn Sep. 1995, 7, 485-491.
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.