|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
Ionic Liquid Mediated Sol Gel Sorbents for Capillary Microextraction and Challenges in Glass Microfabrication by Anne M. Shearrow A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemistry College of Arts and Sciences University of South Florida Co Major Professor: Abdul Malik Ph.D. Co Major Professor: Shekhar Bhansali Ph.D. Florencio E. Hernndez, Ph.D Milton D. Johnston, Ph.D Date of Approval: May 18, 2009 Keywords: sol gel coatings ionic liquid porogenic agent in tube SPME, preconcentration aldehydes alcohols k etones polycyclic aromatic hydrocarbons ( PAH ) sample preconcentration, gas chromatography, stationary phase, microelectromechanical systems (MEMS), microchip, deep reactive ion etching (DRIE), electrochromatography Copyright 2009 Anne M. Shearrow
DEDICATION To my husband, Frank J. B. Shearrow, my entire family, my friends, and in memory of Colleen G. Wilson.
ACKNOWLEDGMENTS I would like to thank my major professor, Dr. Abdul Malik, and my co major profe ssor, Dr. Shekhar Bhansali, for their instruction, encouragement, knowledge, and support for so many years. I thank my Ph.D. committee member Dr. Milton Johnston for his assistance and support. I especially thank my Ph.D. committee member Dr. F. Eloy Hern ndez for his training, assistance, advice, and support during my entire academic career. I also thank Mr. Bernard Batson, all of my lab mates, and colleagues for their never ending support, encouragement, assistance and humor
i TABLE OF CONTENTS LI ST OF TABLES .....................vi LIST OF FIGURES .. vii i LIST OF SCHEMES xvi iii LIST OF ABBRE VIA TIONS...xix ABSTRAC T..... ..xxi ii CHAPTER 1: SOLID PHASE MICROEXTRACTION AND IN TUBE SOLID PHASE MICROEXTRACTION .. .. ...1 1.1 Solv entless Preconcentration: SPME and In Tube SPME 1.2 Microextraction 7 1.2.1 SPME..7 1.2.2 CME.10 1.3 Theory of SPME ..... 1.4 Sorbent Coatings for SPME and In Tube SPME ........ 1.4.1 Custom Coatings ...... 5 1.4.2 Conducting Polymers ... 1.4.3 Carbon B ased ...18 1.4.4 Metal B ased ..... 20 1.4.5 SPME Coatings Based on Molecularly Imprinted P olymers (MIP)
ii 1.4.6 SPME Coatings Based on Restricted Access M aterials (RAM) ... .. 22 1.5 References for Chapter One CHAPTER 2: SOL GEL SORBENTS FOR MICROEXTRACTION .30 2.1 Sol Gel Chemistry: History, Reactions, and Sol Gel Design ....30 2.1.1 Historical Synopsis ... 2.1.2 Advantages of Sol Gel ......3 2 2.1.3 Sol G el Synthesis .....33 184.108.40.206 Catalyst and Water 4 220.127.116.11 Precursors ..37 18.104.22.168 Solvents .37 22.214.171.124 Additives ...38 2.2 Preparation of Sol Gel Sorbent Coated Microextraction Devices ..39 2.2.1 Sol G el Coated SPME F ibers ......39 2.2. 2 Sol G el coated capillaries for in tube SPME .......40 2.3 Characterizing Sol Gel Sorbent s .....41 2.4 Evaluating Sol Gel Sorbents ... 2.5 Sol Gel Sorbents: Literature Survey ..44 2.5.1 PDMS ....44 2.5.2 b Cyclodextrins .53 2.5.3 Calixarenes ....5 5 2.5.4 Crown Ethers .... 2.5.5 Poly(ethylene glycol) ....6 1
iii 2.5.6 Organic Modified Silica ... .....63 2.5. 7 Other Sol Gel Extraction M edia ......6 6 2.5.8 Metal Alkoxide Based Sol Gel Sorbents ...... 2.6 References for Chapter Two...7 4 CHAPTER 3: IONIC LIQUID MEDIATED SOL GEL SORBENTS FOR CAPILLARY MICROEXTRACTION .. .....8 1 3.1 Introduction.....8 1 3.2 Experimental...8 4 3.2.1 Equipment.. ...8 4 3.2.2 Chemicals and Materials...8 5 3.2.3 Preparation of PDMS, PolyTHF, and BMPO Sol Gel Solutions ... 6 3.2.4 Preparation of Ionic Liquid Mediated Sol Gel Microextraction C apillaries ..........8 9 3.2.5 Sol G el CME GC Analysis ... 90 3.3 Results and Discussion... 2 3.3.1 Sol Gel Immobilization of the CME Coatings .....9 2 3.3.2 Ionic Liquid Mediated Sol Gel PDMS Microextraction Capillaries .........9 2 3.3.2. 1 Morphology of Sol G el PDMS Coated Microextraction Capillaries 2 3. 3.2.2 Role of Ionic Liquids in the Sol Gel System ...9 4
iv 3. 3.2.3 CME GC Analysis using Sol Gel PDMS C oated M icroextraction C apillaries 6 3.3.3 Ionic Liquid Mediated Polar Sol Gel Microextraction Capillaries: PEG, PolyTHF, and BMPO .. .. 101 3.4 Conclusion.... 10 3.5 Acknowledgment ..1 10 3.6 References for Chapter Three...1 1 1 CHAPTER 4: IONIC LIQUID MEDIATED BIS[(3 METHYLDIMETHOXY SILYL)PROPYL] POLYPROPYLENE OXIDE SOL GEL COAT INGS FOR CAPILLARY MICROEXTRACTION ......11 6 4.1 Introduction ...11 6 4.2 Experimental .11 9 4.2.1 Equipment ..11 9 4.2.2 Chem icals and M aterials .11 9 4.2.3 Preparation of Sol Gel Solutions 20 4.2.4 Preparation of Sol Gel Coated Microextraction Capillaries ...1 22 4.2.5 Sol gel CME GC A nalysis .12 3 4.3 Results and D iscussion .12 4 4.3.1 Sol Gel Immobilization of the CME C oatings ... 4 4.3.2 Ionic Liquid Mediated Sol Gel Microextraction Capillaries ..12 6 4.3.3 Role of Ionic Liquids ..12 8 4.3.4 Extraction Profiles of Various Analytes O btained on a MOIC Mediated Sol Gel BMPO Microextraction C apillary .....1 3 1
v 4.3.5 Thermal S tability of MOIC M ediated BMPO S ol G el Co atings ..13 2 4.3.6 CME GC A nalysis of V arious C lasses of C ompounds in A queous S amples using MOIC M ediated S ol G el BMPO M icroextraction C apill ary ......13 3 4.4 Conclusion 5 4.5 References for C hapter F our .....13 6 CHAPTRER 5: TOWARDS A MICROCHIP BASED SEPARATION SYSTEM: CHALLENGES IN GLASS MICROFABRICATION ... ...13 9 5.1 Introduction ...13 9 5.2 Experimental .1 4 1 5.2.1 Equipment and M aterials ....1 4 1 5.2.2 Fabrication of M icrofluidic C hannels .....1 4 2 126.96.36.199 Pyre x W afers ......1 4 2 188.8.131.52 Fused S ilica W afers ...1 4 4 5.3 Results and D iscussion .....14 5 5.3.1 Wet E tched M icrofluidic C hannel C haracteristics 5 5.3.2 Dry E tched M icrofluidic C hannel C haracteristics ..14 7 5.3.3 Fabrication of T hrough H oles .....14 9 5.4 Final R emarks ... 50 5.5 References for C hapter F ive .....1 5 1 ABOUT THE AUTHOR End Page
vi LIST OF TABLES Table 2 1 Some advantages of the sol gel process..... Table 2 2 A few common silica based sol gel sorbents ........46 Table 3 1 Names, functions, and chemical structures of some sol soluti on ingredients used to prepare ionic liquid mediated sol gel CME coatings......................................................................................................8 7 Table 3 2 Compositions of sol gel with TTPT io nic liquid: (PDMS IL, PolyTHF IL, or BMPO IL) and without the ionic liquid (PDMS no IL, PolyTHF no IL, or BMPO no IL) used to prepare microextraction capillaries .............................................................8 8 Table 3 3 Peak area r epeatability and limit of detection data for dodecanal (200 ppb sample), heptanophenone (100 ppb sample), and pyrene (50 ppb sample) extracted from aqueous samples using three replicate measurements by CME GC using sol gel immobilized PDMS microex traction capillaries prepared with (A) and without (B) ionic liquid .......9 8 Table 4 1 Names, functions, and chemical structures of sol gel ingredients ...1 2 1
vii Table 4 2 Compositions of sol gels with ionic liquid: (BMPO TTP T, BMPO MOIC, and PDMDPS MOIC) and without ionic liquid (BMPO no IL and PDMDSP no IL) without ionic liquid used to prepare microextraction capillaries ...12 2 Table 4 3 Run to run repeatability (peak area) and detection limit data for non polar and moderately polar analytes in three replicate measurements by CME GC using sol gel BMPO MOIC coated microextraction capillaries ...13 4 Table 4 4 Run to run repeatability ( peak area) and detection limit data for polar and moderately polar analytes in three replicate measurements by CME GC using sol gel BMPO MOIC coated microextraction capillaries 5
viii LIST OF FIGURES Figure 1 1 Soxhlet extraction apparatus: (1) stirrer bar ( 2) still pot, ( 3) distillation path, ( 4) thimble, ( 5) sample, ( 6) siphon top, ( 7) siphon exit, ( 8) expansion adapter, ( 9) condensor, ( 10) cooling water in, and ( 11) cooling water out ...4 Figure 1 2 Schematic of a Solid Phase Extraction Cartridge Figure 1 3 SPME device adapted for GC Applications Designed by Supelco..8 Figure 1 4 Modes of SPME operation: (A) headspace SPME, (B) direct extraction, and (C) membrane protected SPME......9 Figure 1 5 Cross section of (A) an SPME fiber with outer sorptive coating and (B) an in tube SPME device with an inner sorptive open tubular wall coating....1 1 Figure 1 6 Properties of commercially available coatings.. 6 Figure 1 7 Popular co nducting polymers used for SPME coatings.18 Figure 1 8 SEM image of multiwalled CNT coating on f iber surface (200 x) Figure 1 9 Scanning electron micrographs of bare silica fiber (A) RAM SPME fiber coating ( B), and schematic representation of RAM particle (C).....23 Figure 2 1 Overview of the sol gel process Figure 2 2 Example of alkoxy bridging ..38
ix Figure 2 3 Example of an extraction profile ... Figure 2 4 S canning electron micrograph of a sol gel PDMS coating on a (A) fused silica and (B) glass ceramic base ..47 Figure 2 5 CME GC analysis of mixture of two alcohols and two free fatty acids on (a) sol gel CN PDMS capillary (12 cm) and (b) sol gel PDMS capillary (12 cm); extraction time, 30 min, GC analysis conditions: 5m 250 m m I.D. sol gel PDMS column; splitless injection; injector temperature: initial 30 C, final 300 C, programmed at a rate of 60 C/min; GC oven temperature programmed from 35 C to 300 C at a rate of 15 C/min; helium carrier gas; FID temperature 350 C. Peaks: (1) 1 heptanol, (2) 1 octanol, (3) o ctanoic acid, (4) nonanoic acid....52 Figure 2 6 SEM images PDMS/silica particle blended c onditioned for 30min at 300 C (magnification 150)..53 Figure 2 7 The structure of amide bridged calixarene Figure 2 8 Electron scanning micrography (600 magnification) of a sol gel Carbowax 20M ormosil fiber.62 Figure 2 9 Sol gel PEG coating chemically ancho red to the inner walls of fused silica c apillary ..6 3 Figure 2 10 SEM images of the cross sectional view of monolithic columns prepared with 130 m L (A), 150 m L (B) and 170 m L (C) of TEOS in the orig inal sols, respectively. C8 TEOS was 100 m L for every column.....6 7
x Figure 2 11 Phenyl terminated dendrimer with a triethoxysilyl root 9 Figure 3 1 Polycondensation of 3D sol gel network to fused silica capillary wall: (A ) PDMS and (B) BMPO 3 Figure 3 2 Scanning electron microscopic images of cross sections of 250 m m I.D. (A) sol gel PDMS IL (22000) and (B) sol gel PDMS no IL (20000) coated microextraction capillaries ...9 5 Figure 3 3 Comparis on of CME GC analysis of 125 ppb dodecanal, 100 ppb heptanophenone, and 50 ppb pyrene on sol gel PDMS no IL (bottom) and sol gel PDMS IL (top) microextraction capillaries. Extraction conditions: 11 cm 0.25 mm I.D. microextraction capillary; extracti on time, 45 min (gravity fed at room temperature). Other conditions: 15 m 0.25 mm I.D. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 300C; programmed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) dodecanal, (2) heptanophenone, and (3) pyrene for both chromatograms...9 7
xi Figure 3 4 Extraction profiles of heptanophenone (A) and phenant hrene (B) extracted on 11cm 0.25 mm I.D. PDMS IL and PDMS no IL sol gel coated microextraction capillaries from an aqueous sample. Extraction conditions: triplicate extraction at various time intervals; microextraction capillaries were rinsed with 1:1 v/v CH 2 Cl 2 : methanol and dried at 300C before each extraction. GC analysis conditions: 15 m 0.25 mm I.D. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 300C; programmed tempe rature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C .100 Figure 3 5 Scanning electron microscopic images of cross sections of 250 m m I.D. (A) sol gel PEG IL (12000), (B) sol gel PEG no IL (15000), ( C ) sol gel PolyTHF IL (500), ( D ) sol gel PolyTHF no IL (350) coated microextraction capillaries .10 2
xii Figure 3 6 Comparison of CME GC analysis of 100 p pb decanol sol gel PEG IL (bottom) and sol gel PEG no IL (top) microextraction capillaries. Extraction conditions: Same as Figure 3 3. Other conditions: 15 m 0.25 mm I.D. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless d esorption; injector temperature was 250C; programmed temperature GC run from 35C (1 min) to 250C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) decanol for both chromatograms 4 Figure 3 7 Comparison of CME GC anal ysis of 500 ppb decanol, 500 ppb hexanophenone, 200 ppb phenanthreme sol gel PolyTHF IL (bottom) and sol gel PolyTHF no IL (top) microextraction capillaries. Extraction conditions: Same as Figure 3 3. Other conditions: Same as Figure 3 6. Peaks: (1) de canol, (2) hexanophenone, and (3) phenanthrene for both chromatograms..10 5
xiii Figure 3 8 Comparison of CME GC analysis of 500 ppb decanol, 500 ppb hexanophenone, and 200 ppb phenanthrene on sol gel BMPO no IL (bottom) and sol ge l BMPO IL (top) microextraction capillaries. Extraction conditions: Same as Figure 3 3. Other conditions: 15 m 0.25 mm I.D. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 280C; prog rammed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) decanol, (2) hexanophenone, and (3) phenanthrene for both chromatograms ..10 8 Figure 4 1 Polycondensation of 3 D BMPO (A) and PDMDPS (B) sol gel network to fused silica 5
xiv Figure 4 2 Comparison of CME GC analysis of 500 ppb decanol, 500 ppb hexanophenone, and 200 ppb phenanthrene on (A) MOIC mediated sol gel BMPO, (B) TTPT mediated sol gel BMPO; and (C) sol gel BMPO no IL microextraction capillaries. Extraction conditions: 11 cm 0.25 mm i.d. microextraction capillary; extraction time, 45 min (gravity fed at room temperature). Other conditions: 15 m 0.25 mm i.d. Restek Cro ssbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 280C; programmed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) decanol, (2) hexanop henone, and (3) phenanthrene for all chromatograms 7
xv Figure 4 3 Comparison of CME GC analysis of 500 ppb decanol, 500 ppb hexanophenone, and 200 ppb phenanthrene on MOIC mediated sol gel PDMDPS (bottom) and sol gel PDMDPS no I L (top) microextraction capillaries. Extraction conditions: 1 1 cm 0.25 mm i.d. microextraction capillary; extraction time, 45 min (gravity fed at room temperature). Other conditions: 15 m 0.25 mm i.d. Restek Crossbond 14% cyanopropylphenyl 86% PD MS coated GC column; splitless desorption; injector temperature was 300C; programmed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) decanol, (2) hexanophenone, and (3) phenanthrene for al l chromatograms ...12 8 Figure 4 4 Scanning electron microscopic images of cross sections of 250 m m i.d. (A) MOIC mediated sol gel BMPO (370) and (B) TTPT mediated sol gel BMPO (350) coated microextraction capillaries 30
xvi Figure 4 5 Extraction profile for a mixture of decanol, hexanophenone, and phenanthrene extracted on 11cm 0.25 mm i.d. MOIC mediated sol gel BMPO microextraction capillary from an aqueous sample. Extraction conditions: triplicate e xtraction at various time intervals. GC analysis conditions: 15 m 0.25 mm i.d. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 280C; programmed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C 3 2 Figure 4 6 Effect of conditioning temperature on the performance of MOIC mediated sol gel BMPO microextraction capillary. CME GC conditions: extraction time, 45 min; 15 m 0.2 5 mm i.d. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless injection; injector: initial 40C, final (mentioned on x axis), programmed at a rate of 60C/min; GC over temperature programmed temperature from 35C (1 min) to 270 C at a rate of 20C/min; helium carrier gas: FID 350C 3 Figure 5 1 CAD designed mask: serpentine (left) and spiral (right)...14 3
xvii Figure 5 2 Process flow for wet etching of Pyrex wafers (A) Au/Cr deposition; (B) p hotoresist deposition; (C) p hotoresist p atterning; (D) d evelopment of exposed P p hotoresist; (E) a qua regia/Cr etch to remove Au/Cr layer for etching and then protection of backside with Au/Cr ; (F) HF wet etch ; and (G) stri pping of photoresist and Au/C r.... 3 Figure 5 3 Schematic of CAD designed mask with spiral channel, and alignment markers ....14 4 Figure 5 4 Process flow for fabrication of microfluidic channels: photoresist deposition (A), patternin g of photoresist (B), dry etch of fused silica (C), removal of photoresist layer (D) ...14 5 Figure 5 5 Profiler image of wet etched Pyrex wafers..14 6 Figure 5 6 S turn spiral wet etched into a Pyrex wafer 6 Figure 5 7 Pr ofile of spiral microfluidic channels in fused silica after 30 min of DRIE 8 Figure 5 8 Scanning electron microscopic image of cross section an etched wafer after 30 min of DRIE (600 ) 8 Figure 5 9 S turn spira l dry etched into Pyrex wafer 9 Figure 5 10 Through hole in fused silica wafer created via laser ablation..1 50
xviii LIST OF SCHEMES Scheme 2 1 Hydrolysis and polycondensation sol gel reactions .34 Scheme 2 2 Hydrolysis and Condensation mechanisms: (A) acid catalyzed hydrolysis, (B) acid catalyzed condensation, (C) base catalyzed hydrolysis, and (D) base catalyzed condensation
xix LIST OF SYMBOLS AND ABBREVIATIONS AAPTS N (2 aminoethyl)3 aminopr opyltrimethoxysilane ADS Alkyl diol silica AMTEOS Anilinemethyltriethoxysilane ASE Accelerated solvent extraction BET Brunauer Emmett Teller BMA Butylmethacrylate BMPO Bis[(3 methyldimethoxysilyl)propyl] polypropylene oxide BMPT N butyl 4 meth ylpyridinium tetrafluoroborate BTEX Benzene, toluene, ethylbenzene, and xylenes CAR Carboxen CD Cyclodextrin CE Capillary electrophoresis CEC Capillary electrochromatography CME Capillary microextraction CNTs Carbon nanotubes CW Carbowax D RIE Deep reactive ion etching DVB Divinylbenzene ECD Electron capture detector
xx EOF Electroosmotic flow FID Flame ionization detector FPD Flame photometric detection FT IR Fourier transform infrared spectroscopy GC Gas chromatography HMDS H examethyldisilazne HPLC High performance liquid chromatography HS Headspace ICP Inductively coupled plasma I.D. Internal diameter ILs Ionic liquids LC Liquid chromatography LLE Liquid liquid extraction LOD Limit of detection LTGC Low temp erature glassy carbon MA Methacrylate MASE Microwave assisted solvent extraction MEMS Microelectromechanical systems m EC Micro electrochromatograph m TAS Micro total analysis systems MIP Molecularly imprinted polymer MOIC 1 Metyl 3 octlyimidazolium chloride MMA Methyl methacrylate
xxi MS Mass spectrometry MTMS Methyltrimethoxysilane NMR Nuclear magnetic reson ance OCP Organochlorine pesticides ODS Octadecylsilane OH TSO Hydroxyl terminated silicone oil OPP Organophosphorous pesticides Pa Pascal PA Polyacrylate PAHs Polycyclic aromatic hydrocarbons PANI Polyaniline PCBs Polychlorinated biphenyls PEEK Polyetheretherketone PEG Poly(ethylene glycol) PDMS Polydimethylsiloxane PMHS Polymethylhydrosiloxane PolyTHF Polytetrahydrofuran PTMOS Phenyltrimethoxysilane PVC Poly(vinyl chloride) RAM Restricted access materials rpm Revolutions p er minute RSD Relative standard deviation sccm Standard cubic centimeter
xxii SEM Scanning electron microscope SFC Supercritical fluid chromatography SFE Supercritical fluid extraction SIM Selected ion monitoring SPE Solid phase extraction SPME Solid phase microextraction TEOS Tetraethoxysilane TESP N (Triethoxysilylpropyl) O polyethylene oxide urethane TFA Trifluoroacetic acid TGA Thermal gravimetric analysis TMOS Tetramethoxysilane TR Template resin TTPT Trihexyltetradecylphospho nium tetrafluoroborate Ucon Polyalkylene glycol UV Ultraviolet VTEOS Vinyltriethoxysilane W Watt
xxiii IONIC LIQUID MEDIATED SOL GEL SORBENTS FOR CAPILLARY MICROEXTRACTION AND CHALLENGES IN GLASS MICROFABRICATION Anne M. Shearrow ABSTRACT Three ionic liquids (ILs), trihexyltetradecylphosphonium tetrafluoroborate (TTPT), N butyl 4 methylpyridinium tetrafluoroborate (BMPT) and 1 methyl 3 octylimidazolium tetrafluoroborate (MOIC), were utilized to prepare sol gel sorbent coatings. Non polar polydimethylsiloxane (PDMS) and polar poly(ethylene glycol) (PEG), poly(tetrahydrofuran) (PolyTHF) and bis[(3 methyldimethoxy silyl)propyl] polypropylene oxide (BMPO) polymers were employed to develop novel ionic liquid mediated sol gel hybrid organic ino rganic sorbents. The novel sorbents were first tested as coatings for capillary microextraction off line hyphenated to gas chromatography. To gain an understanding of the role of the ionic liquids in the sol gel process the preconcentration abilities o f these novel coatings were investigated for several classes of compounds utilizing CME GC This was accomplished by comparing GC peak areas of a series of analytes extracted on the ionic liquid mediated sol gel CME coatings with that of analogous peak are as obtained on sol gel coatings prepared without the ionic liquid. The morphology of these coatings was compared using scanning electron microscopy (SEM) imaging data Overall, the ionic liquid mediated sol gel coatings had more porous morphologies than the sol gel coatings prepared without ionic liquid. The PDMS and
xxiv BMPO sol gel coatings prepared with ionic liquid in the sol solution provided enhanced extraction sensitivity reflected in higher preconcentration effects and lower detection limits than the sol gel coatings prepared without the ionic liquid. The polar IL mediated BMPO sol gel sorbent was further investigated by exploring the extraction profile and thermal stability of these coatings. A further application of ionic liquid mediated sol gel so rbents could be as stationary phases in a microchip based separation system. Towards this goal, microfluidic channels were fabricated in glass substrates using microelectromechanical engineering. Spiral and serpentine channels were etched in Pyrex and fu sed silica wafers using wet and deep reactive ion etching (DRIE) techniques. Microfabrication protocols such as the use of hard mask and etching times were investigated for both techniques. DRIE produced microfluidic channels that had an etch quality th at was superior to wet etched channels. Thus, the ultimate microchip based separation system should by fabricated using DRIE.
1 CHAPTER 1: SOLID PHASE MICROEXTRACTION AND CAPILLARY MICROEXTRACTION (CME) 1.1 Solventless preconcentration: SPME and In tu be SPME Valcrcel defines modern analytical chemistry as: A metrological science that develops, optimizes and applies material, methodological and strategic tools of widely variable nature (chemical, physical, mathematical, biochemical, biological, etc .) which materialize in measuring processes intended to derive quality (bio)chemical information of both a partial [presence concentration structure of bio( chemical) analyte species] and global nature on materials or systems of widely variable nature (che mical, bio chemical and biological) in space and time in order to solve measuring problems posed by scientific, technical and social problems . In general, analytical chemistry involves not only the analysis of diverse material systems, but also the d evelopment of methods, materials, and instrumentation that can ultimately lead to the determination of the identity and/or the quantity of solutes or analytes in a sample. In any given analytical process several steps must occur sequentially for a success ful outcome. These steps include (1) sampling, (2) sample preparation, (3) separation, (4) detection, and (5) verification with statistics . It is well known that sampling and sample preparation account for 70 80% of analysis time, and sample prepar ation is the most error prone step of the analytical process . Sample preparation is necessary in that it prepares analytes for use in delicate instrumentation,
2 isolates or removes target analytes from a complex matrix such as soil or biological fluids purifies, and concentrates the analyte(s) . Poor sample preparation can often spoil the outcomes of an experiment. Good sample preparation can ultimately enable enhanced sensitivity during detection or sensing methods. Traditional extraction based sample preparation methods include liquid liquid extraction (LLE)  Soxhlet extraction  supercritical fluid extraction (SFE) , accelerated solvent extraction (ASE) [ 8], microwave assisted solvent extraction (MASE)  and solid phase extractio n (SPE)  In these techniques target analytes are extracted from the matrix using solvents which can include an organic solvent or an inert gas as in dynamic headspace extraction. Thus, the distribution constant (K) ( the concentration of the analyte in the extracting phase compared to the concentration of analyte in the sample phase at equilibrium) can be used to determine the extent of an extraction. Most of these sample preparation methods are affected by a number of experimental conditions includ ing temperature, pH, and solvent used. In liquid liquid extraction, target analytes are preconcentrated by transferring the analyte to a solvent that it has a higher affinity for than the liquid sample matrix The extracting solvent should not be soluble in the liquid sample matrix Systems consisting of water and an organic solvent are often used. LLE experiments are usually carried out in a separatory funnel. Analytes are transferred from the aqueous sample matrix to the extracting organic solvent un til equilibrium is reached. The distribution constant of the solute between the organic and the aqueous phases dictate how well an analyte will be extracted.
3 The Soxhlet extraction was introduced in the 19 th century by Franz von Soxhlet [ 6 ]. It is usef ul for separating analytes based on solubility in a solvent. In the past, Soxhlet extractions were primarily used to analyze lipids and oils in biological matrices [ 6 ]. The apparatus for Soxhlet extractions is shown in figure 1 1. Analytes are extracted by loading a solid sample into the thimble which is placed in a partially filled inner tube. Then a flask is filled half way with extracting solvent. A condenser is placed on the top, and the solvent is refluxed. The condensed solvent drips onto the th imble and dissolves the target analyte in it When the chamber is almost full, the solvent and solute flood the siphon, which send them into the flask. This process is repeated automatically until extraction is complete. This is useful for solutes that are not volatile, that are thermally stable and for those that will not polymerize in hot solvent. Soxhlet extractions have some disadvantages including (1) the process is long, it can take hours to months to complete and (2) the solvent is often cold th us limiting the effectiveness of the extraction. To address these issues a Soxtec  extraction can be used which utilizes high pressures causing the solvent to boil. The thimble is dipped in the boiling solvent for immediate extraction. The thimble i s rinsed to remove all extracted analytes, and the extracted sample is then concentrated. However, this system is not automatic, and it is not useful for thermally labile species. Supercritical fluid extraction (SFE) utilizes high pressures and temperatu res to convert a liquid into a supercritical fluid (SCF). The supercritical fluid acts as a solvent to extract analytes from a matrix. Supercritical fluids have increased densities and thus have high solubilizing power. The SCF solvent diffuses into a s ample matrix and extracts analytes out of the matrix into the SCF. Advantages of using SFE for extractions
4 Figure 1 1 Soxhlet extraction apparatus: (1) stirrer bar, ( 2) still pot, ( 3) distillation path, ( 4) thimble, ( 5) sample, ( 6) siphon top, ( 7) siphon exit, ( 8) expansion adapter, ( 9) condensor, ( 10) cooling water in, and ( 11) cooling water out. Reprinted with permission from [ 11 ].
5 include high diffusivity, low viscosity, and low surface tension. These proper ties help the solvent penetrate into small pores. However, there are some drawbacks to using SFE including the need to maintain high temperatures and pressures which can result in high costs, and CO 2 is the only truly usable solvent, and it is nonpolar. Polar modifiers such as methanol or acetonitrile can be added to CO 2 but this requires experimental determination of the appropriate temperature and pressure to maintain the modified solvent in supercritical condition Accelerated solvent extraction i s similar to a Soxhlet extraction that occurs at high pressure and temperat ure, but the set up is similar to that for SFE. ASE utilize s small volumes of organic solvent s in rapid extractions (10 15 minutes). In ASE high temperatures and pressures are app lied to prevent boiling of the solvent; this enables hot solvent to be used for extraction. A wide variety of organic solvents can be utilized in this procedure. However, the extraction is not selective for the target analytes, and it is hard to remove t he solvent after extraction. Microwave assisted solvent extraction is used to selectively extract analytes based on the dielectric constant of a material. In this method, Microwaves affect molecules by ionic conduction and dipole radiation (realignment of dipoles in an applied electromagnetic field) . Heating of a solution occurs when resistance to ionic condition occurs Polar molecules and ionic solutions strongly absorb microwave energy due to their permanent dipole moment. Thus, molecules or s olvents can be selectively heated. This is helpful for the extraction of thermally labile analytes. For example, if a thermally labile microwave absorbing analyte is being extracted, a nonpolar solvent such as hexane with low dielectric constants (and ha ve low microwave absorbing capacity) can
6 be used. Upon heating of the sample, the analyte will be transferred to the cool hexane solution. Extraction times can range from 40 sec to 30 minutes. However, longer extraction times may degrade some samples. Solid phase extraction is a widely used sample preparation technique. In this method, a cartridge that has a solid sorbent bed is used to selectively extract target analytes from liquid samples (figure 1 2). The most common sorbent for SPE is C 18 bonde d particles also known as octadecyl bonded silica or ODS. SPE is a popular technique because ODS is also a widely used HPLC stationary phase. After analytes sorb onto the ODS particles and the sorbent is washed with a solvent inert to the analytes, the a nalytes are eluted using solvent that the analyte has a high affinity for. Figure 1 2 Schematic of a Solid Phase Extraction Cartridge Polypropylene Polyethylene fritted disk Sorbent bed Specimen reservoir Luer tip
7 1.2 Microextraction Most traditional sample preparation techniques require the u se of copious volumes of toxic organic solvents, and they require multiple steps that are error prone and may result in the loss of analytes . Ideally, a sample preparation technique should be solvent free, easy to use, cheap, efficient, selective, and easy to integrate with down stream analysis . Microextraction is a preconcentration technique that is defined by the small volume of extracting phase relative to the sample volume . Microextraction techniques virtually eliminate the use of organic solvent and enable sample extraction and preconcentration in a single step [ 13 ]. These sample preparation methods are greener in that they use little to no solvent; more portable; easier to use; faster; they allow for in situ sampling; have high sensiti vity; are non exhaustive, equilibrium techniques (little sample volume is needed); easy to automate with instrumentation; and more cost effective than other types of sample preparation . Microextraction techniques include suspended particle microextrac tion membrane/disk microextraction vessel wall based microextraction [ 14 ], and stir bar sorptive extraction (SBSE) [ 15 ]. The most common microextraction techniques include solid phase microextraction (SPME)  and in tube SPME  also known as CME . 1.2.1 SPME SPME was first developed by Belardi and Pawliszyn in 1989 [ 13 ]. It was developed to address the need [for] rapid sample preparation in [various settings] [ 14 ]. Initially, optical fibers were used [ 13 ], and later coated silica fibers wer e incorporated into a microsyringe [1 8 ]. The first commercially available SPME device was introduced by Supelco [ 14 ]. An SPME device is schematically shown in figure 1 3. The device
8 consists of a 1.5 cm piece of fused silica rod that has about 0.5 cm long sorbent coating on the end. The fiber is attached to a retractable plunger. When not is use, the fiber is encased in a stainless steel sheath that protects the fiber from mechanical damage such as scratching and fiber breakage. This steel casing also p rotects preconcentrated analytes if the device must be transported prior to analysis. Sampling via microextraction occurs by two steps. First, analytes are extracted onto the coating from the sample matrix. In SPME, this can be accomplished by direct im mersion, headspace (HS), or membrane protected SPME modes (figure 1 4 ). Second, conc entrated analytes are desorbed and delivered into analytical instrumentation. Figure 1 3 SPME device adapted for GC Applications Designed by Supelco .
9 SPME is sui table for hyphenation to various analytical techniques: GC [ 19 ], high performance liquid chromatography (HPLC) [4, 20 ], supercritical fluid chromatography (SFC) [2 1 ], capillary electrophoresis (CE) [2 2 2 3 ], mass spectrometry (MS) [2 4 ], and inductively co upled plasma mass spectrometry (ICP MS) [2 5 ]. SPME is well suited for field analysis . Desorption of analytes occurs thermally when used for GC analysis. The analytes are desorbed from the fiber in the GC injection port, and the analytes are swept on to the column by the carrier gas for separation. In the case of LC, CE, or CEC, solvent (usually the mobile phase) is used to desorb the analytes. The analyte are then transferred onto the separation column using a sampling loop or direct injection. Mic roextraction has been used to preconcentrate samples for a variety of applications including environmental, food, clinical, and forensic analys es . Figure 1 4 Modes of SPME operation: (A) headspace SPME, (B) direct extra ction, and (C) membrane protected SPME. Adapted from . Coating A B C Sample Membrane
10 1.2.2 CME The first appearance of in tube SPME was in the late 1990s [ 16, 26 27 ]. Like fiber based SPME, in tube SPM E or CME  is a non exhaustive, equilibrium process. However, unlike SPME, the sorbent coating is located inside of a fused silica tubing that has a protective outer polyimide layer (figure 1 3). In contrast to SPME analytes are directly extracted onto the coating from a sample while passing through the tubing [ 16 ]. CME was originally developed to overcome difficulties that SPME GC had with weakly volatile or thermally labile analytes [ 28 ] as CME could be easily coupled with HPLC. CME also offer s some other advantages over SPME. SPME fibers often have limited s ample capacities since only 0.5 cm is covered with the sorptive coating. Higher sample capacities can be obtained with CME because the coating is contained within a longer tube (usually 10 cm or longer depending on the analysis technique), and open tubula r wall coatings or monoliths can be used as the extracting phase. SPME devices also have issues with mechanical stability; the fiber can break, the coating can be scratched, and the needle can bend [ 29 31 ]. CME devices allow for superior mechanical stabi lity because flexible capillaries that have protective outer polyimide coatings are utilized, and the sorbent coatings are protected inside the tubing (figure 1 5 ).
11 Figure 1 5 An SPME fiber with outer sorptive coating and an in tube SPME device with an inner sorptive open tubular wall coating. 1.3 Theory of SPME The working principles of fiber SPME and in tube SPME have been detailed greatly by Pawlisyn and co workers [2, 3]. Microextraction utilizes an extracti n g phase or sorbent coating to preconcentrate or extract analytes from a sample. Extraction is complete when a distribution equilibrium between the sample matrix and the extracting phase is reached . Conditions to reach a completed extraction can be described by the following equation: ( 1 1) w here n = the extracted amount (mol), K fs = the distribution constant of the solute betw een the fiber coating and th e sample, n = K fs V f V s C 0 K fs V f + V s Sorbent Coating Fused Silica SPME In Tube SPME
12 V f = the fiber coating volume (L), V s = the sample volume (L), C 0 = the initial concentration of the analyte in the sample (M) This equation indicates that the extracted amount is directly proportional to the samp le concentration. Since microextraction is an equilibrium technique not all of the analyte needs to be extracted (i.e. it is non exhaustive). However, complete extraction can be accomplished if a small sample volume is used and the distribution constant K fs is high. According to Pawliszyn, this is a particular advantage of the SPME technique since it is difficult to work with small sample volumes using traditional sample preparation techniques . SPME is useful for field analysis because when the sample volume is large (K fs V f << V s ) n becomes independent of the sample volume and equation 1 1 takes the form: ( 1 2) T his indicates that the extracted amount (n) is directly proportional to the initial concentration of the analyte in the sample (C 0 ). SPME and in tube SPME are suitable for field analysis because no pre defined volume of sample must be collected . Samp les can be directly gathered from field matrices which help to avoid errors with transporting samples, and it also speeds up the whole analytical process. In the case of in tube SPME, extractions can occur by a dynamic or a static process. In dynamic i n tube SPME, sample is passed through a piece of sorbent coated fused silica tubing (such as a piece of GC c olumn ) that contains the extracting phase or the sorbent. The sorbent can be in the form of an open tubular wall coating a particle n = K fs V f C 0
13 packed bed or a porous monolith. The analyte front migrates through the capillary at a speed that is proportional to the linear velocity of the sample matrix which is inversely proportional to the analyte distribution constant . The minimum extraction times for sh ort capillaries is similar to the time required for the center of the analyte migration band to reach the end of the capillary . This can be described by the following equation  : ( 1 3) Where t e = the extraction time (i.e. time required to reach equilibrium), L = the length of the capillary, K fs = the distribution constant between the extracting phase and the sample matrix, V f = the volume of the extracti ng phase, V v = the volume of the tubing containing the extracting phase, u = the linear velocity of the fluid. Thus, extraction time is directly proportional to the length of the capillary, and it is inversely proportional to the linear velocity of the fluid . During static extractions the sample is not passed through the capillary but it is kept inside the capillary for a predetermined length of time. Analytes are extracted from the sample matrix onto the sorbent via a diffusion process. This type of sampling is popular in field analysis. t e = L (1 + K fs (V f /V v )) u
14 In microextraction high extraction efficiency and sensitivity is achieved when the analyte and the coating have a strong affinity for each other (a high distribution constants). Optimal extraction also depends on temperature, sample pH, matrix effects, derivatization of target analytes, and on the sorbent coating itself. 1.4 Sorbent c oatings for Fiber SPME and CME Sorbent coatings can be high viscosity liquids or solid coatings [ 32 ]. In the case, of liquid c oatings analytes are extracted via absorption of the analyte into the bulk of the liquid. However, analytes are extracted via adsorption (analytes stay on the surface of the coating) when solid coatings are used. Malik and co workers  introduced the term capillary microextraction as an alternative to in tube SPME. In that work it was expressed that capillary microextraction provides a better reflection of the techniques compared with in tube solid phase microextraction since the technique it re lates to is not necessarily limited to the use of only solids phases as the extraction media . Both liquid and solid (which includes composite coatings) coatings are commercially available. A few of the commercially available coatings include polyd imethylsiloxane (PDMS), poly(acrylate) (PA), Carboxen (CAR) PDMS/ poly divinylbenzene (PDMS/DVB), Carbowax/DVB (CW/DVB), and carbowax/template resin (CW/TR). Liquid coatings such as PDMS and PA may be non, partially, or fully cross linked . Coatings th at are non or partially cross linked often have lower thermal stability than fully cross linked coatings; thus, analytes may not be fully desorbed and carry over problems can occur. Non or partially cross linked coatings are sometimes incompatible with so lvents and may not be suitable for analysis techniques that require desorption of analytes by solvent (e.g. HPLC or CE). Composite coatings consist of a liquid phase and
15 a particulate matter that is embedded into a partially cross linked liquid phase. In these coatings, the liquid portion is named first and the particulate portion is named second . Composite coatings such as PDMS/DVB, CW/DVB, CW/TR and solid coatings may be partially or fully cross linked and, thus, have higher thermal and solvent st abilities in comparison to liquid coatings. The choice of which coating to use for an extraction depends upon a number of factors including the nature of the target analyte, sample matrix, and analysis method. The properties of the coating dictate the sel ection of a coating use for a particular analyte (figure 1 6). Coatings that have high selectivity towards the analyte should be used; this will ensure good sensitivity of the extraction. In general, non polar coatings used for the extraction of non pola r analytes, polar coatings for polar analytes In the case of polar analytes it is of particular importance to select a analyte that has a stronger affinity for the coating than water does . Selectivity of a coating is dictated by the organic or inor ganic components in a coating, polarity, and the K fs While a handful of coatings are commercially available these may not be ideal for certain situations or analytes. For example, analytes may be too polar or high temperatures may be required for compl ete desorption of analytes in GC hyphenated analysis. Thus much research effort has been focused on custom coatings (as will be explained below) that are more selective, efficient, and robust than their commercial counterparts. 1.4.1 Custom Coatings Custom coatings take aim at overcoming the limitations of traditional commercial coatings. For example, Nafion perfluorinated resin has been used as an SPME coating by Pawliszyn and co workers [ 33 ]. Nafion coatings were developed to extract polar
16 compou nds from nonpolar matrices. Other specialized coatings include liquid crystalline films [ 34 ], bonded phase silica [ 35 ], mesoporous silica [ 36 ], PVC , polyamine [ 38, 39 ], cellulose acetate on a silver wire [ 40 ], polyacrylonitrile [ 41 ], silicone glue [ 42 ], poly Figure 1 6 Properties of commercially available coatings; adapted from [ 27 ]. (phthalazine ether sulfone ketone) [ 43 ], dibenzo 18 crown 6 [ 44 ], methacrylic acid trimethylolpropanetrimethacrylate [ 45 ], ionic liqui ds [ 46 48 ] and poly(methacrylic acid ethylene glycol dimethacrylate) monolith for in tube SPME HPLC [ 49 ]. In the follow ing sections popular types of custom coatings will be described including conducting polymers, carbon based, metal based, MIP, and RAM microextraction coatings. 1.4.2 Conducting Polymers Conducting polymers are useful for molecular recognition because (a) they can include counter ions for selective interactions; (b) they have multifuctionality and properties such as hydrophobic interacti ons p p interactions, and ion exchange capacity ; Coating Properties PA 85 m m PDMS 7 m m PDMS 30 m m PDMS 100 m m PDMS/DVB 65 m m CW/DVB 65 m m CW/TR 50 m m CAR/PDMS 75 m m DVB/CAR/PDMS 50/30 m m Weak Strong Retention Low High Pola rity Bonded Non bonded Partially cross linked Fully cross linked
17 (c) they can introduce functional groups to monomers; and (d) codeposition of metals or other monomers can occur with the polymer [ 50 ]. These polymers have been utilized as SPME coatings. In most cases, th ese polymers are electrochemically deposited onto metal wires, thus, they are not prone to fiber breakage like some SPME devices that rely on fused silica. The three main types of conducting polymers used for SPME include those based on polypyrrole, polyth iophene, and polyaniline [ 50 ] (figure 1 7). The first report of the use of a conducting polymer for an SPME coating was by Pawliszyn and co workers [ 50, 51 ]. In these works polypyrrole and poly N phenylpyrrole were coated into metal wires [ 51 ] by an ele ctrochemical method or onto the inner surface of a short piece (60 cm 250 m m i.d.) of GC capillary by chemical vapor deposition [ 50 ]. In the later work, the coatings were used for in tube SPME coupled with LC electrospray ionization mass spectrometry to analyze b blockers in urine and serum [ 50 ]. The polypyrrole coating provided lower detection limits for the b blockers than commercial Omegawax 250 coatings. The coating also allowed for easy desorption of the analytes in the mobile phase flow without c arryover problems. Pawliszyn and co workers have also used polypyrrole base coatings for SPME coupled to ion chromatography [ 52 ] and for in tube SPME analysis of organoarsenic compounds [ 53 ], stimulants [ 54 ], and aromatic compounds [ 55 ]. Polypyrrole base d coatings have also been used for PAHs [ 56, 57 ], ochratoxin A [ 58 ], phenolic compounds [ 59 ], pesticides [ 60 ], and methamphetamines in human serum [ 61 ]. Polyaniline (PANI) has been used as a sorbent coating for SPME fibers [ 62 ]. As with other conducting p olymers, electropolymerization methods were used to prepare the coatings on metal wires. PANI SPME coatings have been utilized for aromatic amines
18 Polypyrrole Polythiophene Polyaniline Figure 1 7 Popular conducting polymers used for SPME coatings. [ 62 ]; aliphatic alcohols [ 63 ]; phenols [ 64, 65 ]; anatoxin a in culture media of cyanobacteria [ 66 ]; bisphenol A, 4 n nonylphenol, and 4 tert phenol [ 62 ]; phthalates [ 67 ]; PAHs [ 68 ]; organochlorine pesticides [6 9 ]; and substituted benzenes [ 70, 71 ]. In some cases, conducting polymer sorbents have been used for elect rochemically controlled SPME [ 72 ] extraction and desorption of ionic analytes [ 73, 74 ]. According to Pawliszyn et al., the electroactivity and reversible redox properties of conductive polymers allows for electrochemical switching accompanied by the movem ent of counterions in and out of the polymer film for charge balance [ 72 ]. Analytes can be extracted into coatings during oxidation and released or desorbed during reduction. 1.4.3 Carbon B ased Sorbents It is well known that carbon is a good sorbent for vo latile organic compounds. Activated carbon based materials have been used as sorbents for SPME. Mangani et al.
19 utilized graphitized carbon black as an SPME coating [ 75 ]. The fiber was used to extract organic pollutants followed by GC or GC MS analysis. D jozan et al. utilized activated powdered charcoal as an SPME coating [ 76 ]. They analyzed benzene, toluene, ethylbenzene, and xylene derivatives [BTEX]. Detection limits for each BTEX compound was 1.5 2 pg/mL. Djozan et al. also used pencil lead as an SP ME fiber [ 77 ]. They extracted PAHs from aqueous samples and analyzed them by SPME GC. As RSD% value for naphthalene was 5.3%. The fiber was stabl e to high temperature. Burk et al. utilized polycrystalline carbon (pencil lead and glassy carbon) to try a nd improve the selectivity of carbon [7 8 ]. They extracted nonionic surfactants and found that the fibers provided results similar to PDMS/DVB and CW/TR fibers. Zeng and co workers developed ceramic/carbon composite coated stainless steel SPME wires to ex tract organophosphorous pesticides in water [7 9 ]. Recently, carbon nanotubes (CNTs) have been utilized as sorbents for SPME. Wang and co workers utilized multiwalled CNT coated SPME fibers to extract and analyze polybrominated diphenyl ethers in water and milk samples [ 80 ] (figure 1 8). Lu and co workers [ 81 ] used single walled CNTs as a sorbent on a stainless steel SPME fiber. The fiber was used to analyze pesticides in water. CNTs were used as an SPME coating by incorporating them into a groove of a stainless steel rod [ 82 ]. The fiber was thermally stable up to 280C, and it was used to extract aromatic hydrocarbons. Li et al. developed platinum SPME fibers coated with single walled CNTs via electrophoretic deposition [ 83 ]. The CNT coated SPME fi bers were used in direct immersion SPME to extract phenols followed by HPLC UV. The authors found that the CNT coated fiber provided performance comparable with a PA fiber.
20 1.4.4 Metal B ased While many sorbent coatings used for microextraction utilize org anic moieties for extraction, some have utilized inorganic materials to serve as coatings. Djozan and co workers developed methods to prepare anodized CuCl 2 coating on copper wires [ 84 ], aluminum wires [ 85 ], zinc wires [ 86 ], and copper sulfide wires [ 87 ] f or use in SPME. Farajzadeh and co workers [8 8 ] utilized electrochemical means to prepare coatings of copper (I) chloride, copper (I) oxide, and copper (I) sulfide on copper wires. The SPME fibers were used to extract amines. Farajzadeh [8 9 ] and co work ers also developed alumina based coatings by using a mixture of alumina powder and PVC on a silver wire. The SPME fiber was used to extract alcohols as alumina has an affinity for hydroxyl groups. Alumina coated on fused silica has been used to extract vo latile organic compounds [ 90 ]. Fiberglass [ 91 ] and glass ceramic rods [ 92] were coated with Nb 2 O 5 Figure 1 8 SEM image of multiwalled CNT coating on fiber surface (200 x). Reprinted from [ 80 ].
21 and were used to extract alcohols and/or phen ols by HS SPME. Mehdinia et al. [ 93 ] coated platinum wires with nanostructured PbO 2 using electrochemical means. The fibers were used to extract BTEX compounds by HS SPME GC. Limits of detection were less than 0.12 m g/L. Budziak and co workers [ 94 ] electrodeposited ZrO 2 onto NiTi wires. The SPME wires were used to extract halophenols from water. The authors state that the ZrO 2 NiTi wires provided efficiency similar to commercial PDMS fibers with thicknesses of 7 m m Cao and co workers developed titania SPME fibers by oxidizing titanium wire with hydrogen peroxide [ 95 ]. A nanostructured coating with a thickness of 1.2 m m and pore 100 200 nm in size resulted. The SPME fibers were used for HS SPME GC to analyze DDT and its degradation products. Bai et al. [ 96 ] developed zirconium phosphate coated capillary tubes that were used to extract Cr(III) from natural water. 1.4.5 SPME Coatings Based on Molecularly Imprinted Polymers ( MIP ) Molecularly imprinted polymers ha ve been used as selective coa tings for microextraction. A n MIP based SPME sorbent was reported by Mullett and co workers in 2001 [ 97 ]. In this work an MIP coating that was imprinted for propranolol was developed for in tube SPME The coating was used t o extract propranolol and some b blockers from spiked serum samples. Samples were analyzed on line using in tube SPME HPLC UV. Detection limits of 0. 32 m g/mL and an RSD value of less than 5 % were obtained. An MIP baased SPME fiber was developed by Kost er and co workers [9 8 ]. The fiber was imprinted for Clenbuterol by in situ synthesis via silinization of the silica SPME fiber The fiber was use to extract brombuterol from spiked human urine samples.
22 Li and co workers [9 9 100] have developed severa l MIP SMPE fibers coupled with HPLC analysis Prometryn imprinted fibers were used to analyze tirazines in spiked soybean, corn, lettuce, and soil samples [ 99 ]. The imprinted fiber provided an extraction yield that was 10 time higher than the non imprint ed fiber. Limits of detection ranged from 0.012 to 0.090 m g/L. MIP SPME fibers were also prepared by multiple co polymerization method using tetracycline as the template [ 100 ]. The fiber had a porous, cross liked surface. The fiber was used to extract tetracycline, oxytetracycline, doxycycline, and chlortetracycline from spiked chicken feed, chicken muscle, and milk samples. Li and co workers [ 101] also used MIP SPME fibers to analyze beta blockers in urine and plasma samples 1.4.6 SPME Coatings Bas ed on Restricted Access Materials (RAM) RAMs have been used to develop SPME coatings. Pawliszyn and co workers develop RAM SPME fibers to analyze 3 H diazepam in biological fluids [ 102 ] (figure 1 9). The SPME fiber was interfaced to HPLC, and the desorbe d analytes were introduced onto the HPLC column using a 6 port injection valve. Pawliszyn and co workers [ 103 ] were also the first developed RAM coatings for in tube SPME coupled to HPLC to analyze benzodiazepines in human serum samples. The RAM coating was prepared using alkyl diol silica (ADS). This allowed for direct extraction of analytes from serum ; the proteins were excluded by the RAM and the analytes were trapped on the hydrophobic interior. No ultrafiltration or deproteinization steps were requ ired thus simplifying sample preparation. Lim its of detection ranged from 22 to 29 ng/mL for the benzodiazepines. A RAM ADS SPME fiber was used to extract diazepam and its metabolites from heparinized blood samples [ 104 ]. The fiber was coupled to LC MS.
23 The device had an RSD value of 10% over ten injections and diction limits ranged from 20 35 ng/mL. Pawliszyn and co workers also coupled RAM SPME fibers to electrospray MS [ 105 ]. The RAM was immobilized on steel and platinum wires. In some cases, try psin was immobilized on the wires and inside the sample vial to digest proteins. The RAM SPME device and the protein solution were incubated together. The peptides were then analyzed by SPME/nanospray. Up to eight peptides could be identified. Fig ure 1 9 Scanning electron micrographs of bare silica fiber (A), RAM SPME fiber coating (B), and schematic representation of RAM particle (C). Reprinted from [ 102 ].
24 Jarmalaviciene and co workers developed a biocompatible monolithic RAM beds as an SPME capillary insert [ 106]. The device was used to exclude proteins from bovine plasma while extracting the active components of caffeine. While a myriad of commercial and custom coatings are available, most traditional polymer based sorbents have the foll owing drawbacks: (1) they are possess poor solvent stablility (2) they are thermally labile, and (3) they provide limited sample capacity. These drawbacks are mainly due to the fact that the sorbents are not permanently attached to the SPME or in tube S MPE supports. On e obvious way to overcome this poor immobilization is to utilize a technique that can chemically and permanently anchor a sorbent to the microextraction support. Sol gel coating technology was developed for this very purpose [ 17, 107 ]. C hemically anchored sol gel sorbents can survive harsh conditions such as organic solvents and high temperatures. The sol gel approach offers outstanding opportunities to develop robust, sensitive, sorbent coatings for microextraction that can be utilized for a wide variety of analytes. 1.5 References for C hapter O ne  M. Valcrcel, TrAC 16 (1997) 124.  J. Pawliszyn, Solid Phase Microextraction: Theory and Practice. Wiley VCH, NY, 1997.  J. Pawliszyn, Sampling and Sample Preparation for Field and Laboratory Elsevier, NY, 2002.  Z. Zhang, M.Yang, J. Pawliszyn, Anal. Chem. 66 (1994) A844.  R. Blumberg, Liquid liquid Extraction, Academic Press, San Diego, CA, 1988.  F. Soxhlet, Die gewichtsanalytische Bestimmung des Milchfettes, Polytechnisches J. (Dingler's) 232 (1879) 46.
25  M. Caude, D. Thiebaut, Practical Supercritical Fluid Chromatography and Extraction, Harwood Academic Publishers, Australia, 1999.  H. Giergielewicz Mozajska, L. Dabrowski, J. Namiesnik, Crit. Rev. Anal. Chem. 31 (2001) 149.  C. S. Eskilsson E. Bjrklund, J. Chromatogr. A 902 (2000) 227.  N. J. K. Simpson, Solid Phase Extraction: Principles, Techniques, and Applications, Marcel Dekker, New York, 2000.  http://en.wikipedia.org/wiki/Soxhlet_extractor 27 October 2006  M. Nilsson, In Focus 11 (1 988) 20.  R. G. Belardi J. Pawliszyn, Water Pollut. Res. J. Can. 24 (1989) 179.  H. Lord, J. Pawlisyn, J. Chromatogr. A 885 (2000) 153.  E. Baltussen, P. Sandra, F. David, C. Cramers, J. Microcolumn Separations 11 (1999) 737.  R. Eisert, J. Pawliszyn, Ana l. Chem. 69 ( 1997) 3140.  S. Bigham, J. Medlar, A. Kabir, C. Shende, A. Alli, A. Malik, Anal. Chem. 74 (2002) 752.  C. Arthur, J. Pawliszyn, Anal. Chem. 62 (1990) 2145.  C. Arthur, L. Killam, K. Buchholz, J. Pawlisyzn, Anal. Chem. 64 (1992) 1960.  J. Chen, J. Pawliszyn, Anal. Chem. 67 (1995) 2530.  Y. Hirata, J. Pawliszyn, J. Microcol. Sep. 6 (1994) 443.  D. Figeys, A. Ducrte, J. Yates, R. Aebersold, Nat. Biotechnol. 14 (1996) 1579.  C. Whang, J. Pawliszyn, Anal. Commun. 35 (1998) 353.  Z. Zhang, J. Pawlisz yn, Anal. Chem. 65 (1993) 1843.  L. Moens, T. De Smaele, R. Dams, P. Van Den Broeck, P. Sandra, Anal. Chem. 69 (1997) 1604.  H. Hartmann, J. Burhenne, M. Spiteller, Fresenius Environ. Bull. 7 ( 1998) 96.
26  H. Kataoka, J. Pawliszyn, Chromatographia 50 ( 1999) 532.  W. Mullett, J. Pawliszyn, J. Sep. Sci. 26 (2003) 251.  D. Djozan, Y. Assadi, S. Haddadi, Anal. Chem. 73 (2001) 4054.  M. Huang, C. Tai, Q. Zhou, G. Jiang, J. Chromatogr. A 1048 (2004) 257.  D. Budziak, E. Martendal, E. Carasek, J. Chromatogr. A 1164 (2007) 18.  J. Pawliszyn, T. Gorecki, Applications of Solid Phase Microextraction, 1 st ed. Royal Society of Chemistry, Cornwall, U.K., 1999, pp.92 108.  T. Gorecki, P. Martos, J. Pawliszyn, J. Anal. Chem. 70 (1998) 19.  E. Out, J. Pawliszyn, Microchim. Act a 112 (1993) 41.  Y. Liu, M. Lee, K. Hageman, Y. Yang, S. Hawthorne, Anal. Chem. 69 (1997) 5001.  J. Hou, Q. Ma, X. Du, H. Deng, J. Gao, Talanta 62 (2004) 241.  A. Saraullo, P. Martos, J. Pawliszyn, Anal. Chem. 69 (1997) 1992.  T. Ding, H. Lin, C. Whang, J Chromatogr. A 1062 (2005) 49.  R. Yang, W. Xie, Forensic Sci. Int. 139 (2004) 177.  M. Farajzadeh, M. Hatami, J. Sep. Sci. 26 (2003) 802.  M. Musteata, F. Musteata, J. Pawliszyn, Anal. Chem. 79 (2007) 6903.  D. Panavaite, A. Padarauskas, V. Vickackaite, Anal. Chim. Acta 571 (2006) 45.  W. Guan, F. Xu, W. Liu, J. Zhao, Y. Guan, J. Chromatogr. A 1147 (2007) 59.  C. Jia, Y. Luo, J. Pawliszyn, J. Microcol. Sep. 10 (1998) 167.  J. B. Zeng, J. M. Chen, Y. R. Wang, W. F. Chen, X. Chen, X. R. Wang, J. Chromatogr A 1208 (2008) 34.  J. F. Liu, N. Li, G. B. Jiang, J. M. Li, J. A. Jonsson, M.J. Wen, J. Chromatogr. A 1066 (2005) 27.  Y. N. Hsieh, P. C. Huang, I. W. Sun, T. J. Whang, C. Y. Hsu, H. H. Huang, C. H. Kuei, Anal. Chim. Acta 557 (2006) 321.
27  F. Zhao, Y. J. Meng, J. L. Anderson, J. Chromatogr. A 1208 (2008) 1.  Y. Fan, Y. Feng, J. Zhang, S. Da, M. Zhang, J. Chromatogr. A 1074 (2005) 9.  J. Wu, H. Lord, J. Pawliszyn, H. Kataoka, J. Microcol. Sep. 12 (2000) 255.  J. Wu, Z. Deng and J. Pawliszyn, Polypyrrole an d poly Nphenylpyrrole films for SPME Extech., 1999, Waterloo, Canada.  J. Wu, X. Yu, H. Lord, J. Pawliszyn, Analyst 125 (2000) 391.  J. Wu, J.; Z. Mester, J. Pawliszyn, Anal. Chim. Acta 424 (2000) 211.  J. Wu.; Lord, H.; J. Pawliszyn Talanta 54 (2001) 655.  J. Wu.; J. Pawliszyn. Anal. Chem. 73 (2001) 55.  A. Mohammadi, Y. Yamini, N. Alizadeh, J. Chromatogr. A 1063 (2005) 1.  H. Bagheri, E. Babanezhad, F. Khalilian, Anal. Chim. Acta 634 (2009) 209.  J. Yu, E. Lai, Anal. Bioanal. Chem. 381 (2005) 948.  N Alizadeh, H. Zarabadipour, A. Mohammadi, Anal. Chim. Acta 605 (2007) 159.  F. Ahmadi, A. Shahsavari, M. Rahimi Nasrabadi, J. Chromatogr. A 1193 (2008) 26.  N. Alizadeh, A. Mohammadi, M. Tabrizchi, M. J. Chromatogr. A 1183 (2008) 21.  M. Huang, G. Jiang, Y. Cai, J. Sep. Sci. 28 (2005) 2218.  D. Djozan, S. Bahar, S. Chromatographia 59 (2004) 95.  H. Bagheri, A. Mir, E. Badanezhad, Anal. Chim. Acta 532 (2005) 89.  M. Mousavi, E. Noroozian, M. Jalali Heravi, A. Mollahosseini, Anal. Chim. Acta 581 (2007) 71.  A. Ghassempour, N. Najafi, A. Mehdinia, S. Davarani, M. Fallahi, M. Nakhshab, J. Chromatogr. A 1078 (2005) 120.  X. Li, M. Zhong, S. Xu, C. Sun, J. Chromatogr. A. 1135 (2006) 101.  H. Bagheri, E. Badanezhad, A. Es haghi, J. Chromatogr. A 1152 (2007) 168.
28  X Li, M. Zhong, J. M. Chen, J. Sep. Sci. 31 (2008) 2839.  Y. Wang, Y. Li, J. Feng, C. Sun, Anal. Chim. Acta 619 (2008) 202.  A. Mehdinla, M. F. Mousavi, J. Sep. Sci. 31 (2008) 3565.  J. Wu, W. Mullett, J. Pawliszyn, Anal. Chem. 74 (2002) 4855.  G. Liljegre n, J. Pettersson, K. Markides, L. Nyholm, Analyst 127 (2002) 591.  B. Yates, K. Temsamani, O. Ceylan, S. Oztemiz, T. Gbatu, R. LaRue, U. Tamer, H. Mark, Talanta 58 (2002) 739.  F. Mangani, R. Cencarini, Chromatographia 41 (1995) 678.  D. Djozan, Y. Assadi, Chromatographia 45 (1997) 183.  D. Djozan, Y. Assadi, Chromatographia 60 (2004) 313.  R. Aranda, P. Kruus, R. Burk, J. Chromatogr. A 888 (2000) 35.  J. Zeng, B. Yu, W. Chen, Zh. Lin, L. Zhang, Ze. Lin, X. Chen, X. Wang, J. Chromatogr. A 1188 (2008) 26.  J. Wang, D. Jiang, Z. Gu, X. Yan, J. Chromatogr. A 1137 (2006) 8.  J. Lu, J. F. Liu, Y. Wei, K. Jiang, S. Fan, J. Y. Liu, G. Jiang, J. Sep. Sci. 30 (2007) 2138.  E. Adomaviciute, K. Jonusaite, J. Barkauskas, V. Vickackaite, V. Chromatographia 67 (2008) 599  Q. L. Li, X. F. Wang, D. X. Yuan, J. Chromatogr. A 1216 (2009) 1305.  D. Djozan, Y. Assadi, G. Karim Nezhad, Chromatographia 56 (2002) 611.  D. Djozan, Y. Assadi, S, Haddadi, Anal. Chem. 73 (2001) 4054.  D. Djozan, L. Abdollahi, Chromatogaphia 57 (2003) 799.  D. Djozan, M. Amir Zehmi, Chromatogaphia 58 (2003) 221.  M. Farajzadeh, N. Rahmani, Talanta 65 (2005) 700.  M. Farajzadeh, N. Rahmani, Anal. Sci. 20 (2004) 1359.
29  L. Wei, Q. Ou, J. Li, Chromatographia 59 (2004) 601.  F. de Oliveira, F.; C. da Sileira S. de Campos, E. de Campos, E. Carasek, Talanta 66 (2005) 74.  B. da Silveira, A. de Oliveira, S. de Campos, E. de Campos, E. Carasek, Colloids Surfaces A: Physicochem. Eng. Aspects 259 (2005) 15.  A. Mehdinia, M. Mousavi, M. Shamsipur, J. Chromatogr. A 1134 (2006) 24.  D. Budziak, E. Martendal, E. Carasek, Anal. Chim. Acta 598 (2007) 254.  D. Cao, J. Lue, J. Liu, G. Jiang, Anal. Chim. Acta 611 (2008) 56.  H. X. Bai, Z. F. Fan, J. Enivron. Monit. 11 (2009) 326.  W. Mullett, J. Pawliszyn, Anal. Chem. 73 ( 2001) 2383.  E. Koster, C. Crescenzi, W. den Hoedt, K. Ensing, G. de Jong, Anal. Chem. 73 (2001) 3140.  H. Xiaogang, H. Yuling, L. Gongke, J. Chromatogr. A 1147 (2007) 1.  H. Xiaogang, P. Jialiang, H. Yuling, H. Yin, L. Gongke, J. Chromatogr. A 1188 (2008) 97.  X. G. Hu, I. L. Pan, Y. L. Hu, G. Li, J. Chromatogr. A 1216 (2009) 190.  W. Mullett, J. Pawliszyn, Anal. Chem. 74 (2002) 1081.  W. Mullett, K. Levsen, D. Lubda, J. Pawliszyn, J. Chromatogr. A 963 (2002) 325.  M. Walles, W. Mullett, J. Pawliszyn, J. C hromatogr. A 1025 (2004) 85.  M. Walles, Y. Gu, C. Dartiguenave, F. Musteata, K. Waldron, D. Lubda, J. Pawliszyn, J. Chromatogr. A 2005, 1067, 197.  R. Jarmalaviciene, M. Szumski, O. Kornysova, E. Klodzinska, D. Weserlund, S. Krawczyk, D. Mickevicius, B. Buszewski, A. Maruska, Electrophoresis 29 (2008) 1753.  S. Chong, D. Wang, J. Hayes, B. Wilhite, A. Malik, Anal. Chem. 69 (1997) 3889.
30 CHAPTER 2: SOL GEL SORBENTS FOR MICROEXTRACTION 2.1 Sol Gel Chemistry: History, Reactions, and Sol Gel Design 2 .1.1 Historical Synopsis A colloidal suspension of finely divided particles (sol) is formed through hydrolytic polycondensation reactions of sol gel precursors such as inorganic salts or alkoxides of certain elements. Further progression of polycondensati on reactions in the sol solution results in a liquid filled 3D network structure called: a gel. Sol gels can be developed under a variety of experimental conditions which can result in final products such as dense films, xerogels, aerogels, or ceramics (fi gure 2 1). Modern sol gel processing has a history that begins in the 1800s. The first metal alkoxide was prepared by Ebelmen in 1845 from alcohol and SiCl 4 . It wasnt until the 1930s that oxide films were prepared from metal alkoxides which was de veloped by Geffecken and the Schott glass company . During the 1930s Kistler utilized supercritical drying to produce aerogels . Homogenous powders for ceramics were developed by Roy in the 1950s . In the 1960s and 1970s Levene and Thomas  as well as Dislich  prepared multicomponent glass through control of hydrolysis and condensation. Yoldas  and Yamane et al.  developed sol gel monoliths without cracks through controlled drying. It was demonstrated that drying control chemical a dditives (DCCA) could be used to produce monoliths without cracks . In 1985, Schmidt prepared non crystalline solid inorganic organic hybrid sol gels
31 Figure 2 1 Overview of the sol gel process . [11 ]. Cortes et al. utilized a sol gel monolith for a liquid chromatography column in 1987 . Guo and Colon prepared open tubular wall coated sol gel columns for capillary electrochromatography in 1995 . The first use of sol gel stationary phases in gas chromatograp hy columns was in 1997 by Malik and co workers . Malik and co workers also developed sol gel sorbents for SPME fibers in 1997  and for in tube SPME in 2002 . Sol gel stationary phases CEC , HPLC , and GC  have been reviewed elsewh ere.
32 2.1.2 Advantages of Sol Gel Sol gel has been used in a variety of fields including, chemistry, biology, engineering, optics, physics, and materials science. Key advantages of sol gel materials allow for its use in multiple disciplines. Macke nzie initially outlined some now well known advantages of sol gel processing over traditional glass preparation methods (Table 2 1) . Sol gel materials can be prepared in different formats such as monoliths, thin films, particles, and powders. Porosi ty and surface characteristics of sol gels can be easily controlled. Sol gel technology allows for novel materials to be developed that can be applied to specific needs. This so called tunability is possible due to the wide range of sol gel precursor an d sol gel active organic ligands and polymers that are available. Since the process happens under mild conditions, sol gels can be modified or functionalized with organic components and/or thermally liable materials like enzymes . Sol gels with o rganic moieties or hybrid organic inorganic materials have been utilized in the fields of sample preparation [15, 16] and separations science [17 19]. The organic components of the sol gel network participate in chromatographic separation and sample prec oncentration by providing moieties that can interact with, and have high affinities for, target analytes. Sol gel technology is well suited for preparing such coatings and stationary phases because they can be tailored to specific analytes by utilizing pre cursors and polymers with desired properties. As mentioned, organic components can be effectively incorporated into inorganic structures under extremely mild thermal conditions (room temperature) . Sol gel coatings can be immobilized to substrates suc h as glass, fused
33 silica, or quartz via covalent bonds with a substrates surface silanol groups [23, 24]. This is particularly advantageous since covalently attached sol gel coatings and stationary phases possess superior thermal and solvent stability co mpared with traditional unbonded materials . Table 2 1 Some advantages of the sol gel process. Adapted from . Sol gel Processing Advantages Better material homogeneity and purity from raw materials Low temperatu re preparation New non crystalline solids outside the range of normal glass formation New crystalline phases from new non crystalline solids Better glass products from special properties of gel Special products 2.1.3 Sol gel Synthesis Careful design of a sol gel is required in order to obtain a coating or stationary phase with the desired characteristics and properties. Thus an understanding of reactions involved in the sol gel process is essential. The main reactions involved include hydrolysis (Schem e 2 1(I)) and polycondensation (Scheme 2 1(II)) . Reactions, overall structure, and properties of the final product are affected by experimental conditions and components in a sol gel system. Typical components that affect a sol gel system include (1) water, (2) catalysts, (3) precursors, (4) solvents, and (5) other additives (e.g. organic molecules) .
34 184.108.40.206 Catalyst and Water Catalysts are used to speed up sol gel reactions. The amount of water and catalyst can dictate if hydrolysis reactions w ill go to completion. Hydrolysis and condensation reactions occur simultaneously. Condensation reactions can occur by the loss of water or alcohol. A sol gel can have a linear or branched structure, and this generally depends on the extent of hydrolysis (i.e. if hydrolysis was full or partial) . This maybe a function of acid concentration and other experimental variables . C a t a l y s t S o l v e n t C a t a l y s t S o l v e n t I H y d r o l y s i s o f S o l g e l P r e c u r s o r s I I C o n d e n s a t i o n o f H y d r o l y z e d P r o d u c t s M ( O R ) n + H 2 O ( R O ) n 1 M O H + R O H 2 ( R O ) n 1 M O H ( R O ) n 1 M O M ( O R ) n 2 O H + R O H o r H 2 O M = S i A l Z r G e o r T i ; R = a l k y l g r o u p Scheme 2 1 Hydrolysis and polycondensation sol gel reactions. Some common catalysts include the following: (a) acids (mineral acids and organic acids), (b) bases (amines, ammonia, hydroxide ions), (c) UV light (radical initiator), and (d) fluorides (hydrofluoric acid , alkaline metal florides, ammonium fluorides, and fluoride ions [2 6]). An acid or a base can be used to catalyze hydrolysis and/or condensation reactions (Scheme 2 2 A D). Acid catalyzed hydrolysis is fast and
35 occurs by an electrophilic attack by the hydronium ion (H 3 O + ) on the Si O bond (Scheme 2 2A) . This is fo llowed by a nucleophilic attack of water on the Si atom. Reactivity decreases as the alkoxy groups are replaced by hydroxyl groups. A low degree of cross linking and micropores may result in these gels since hydrolysis is favored over condensation. Acid catalyzed condensation reactions occur by a nucleophilic substitution (Scheme 2 2B). Base catalyzed hydrolysis is slow and proceeds by a nucleophilic substitution (Scheme 2 2C). Reactivity increases as alkoxy groups are replaced by hydroxyl groups. Thes e gels are often highly cross linked and mesoporous because condensation is favored. Base catalyzed condensation occurs by attack of a nucleophilic deprotonated silanol on a neutral silicate species  (Scheme 2 2D). The ratio of silicon atoms to water can also dictate structure. For example, under acid catalyzed hydrolysis conditions low Si:H 2 O ratio leads to weakly branched materials, whereas, large Si:H 2 O ratios with base catalyzed hydrolysis will lead to highly condensed particulate materials. One of the most commonly used catalysts for sol gel sorbents or stationary phases is trifluoroacetic acid (TFA). TFA is an organic acid, and it has a pKa value of 0.3. Carboxylic acids with pKa values of less than 4 enable higher reaction rates than some stro ng inorganic acids or organic acids . Enhanced sol gel reaction speeds allow for faster GC column , SPME , or in tube SPME  coating fabrication. TFA is also an excellent solvent, and it can provide a source of water for sol gel reactions (e.g., TFA containing 5% water is typically used in sol gel reactions). Sol gel precursors include metal/metalloid salts and alkoxides [21, 29]. Alkoxides are commonly used to prepare sol gel sorbent coatings for microextraction. Alkoxides of
36 ( A ) Acid C atalyzed Hydrolysis: S i O R R O R O R O H + H 2 O S i R O O R R O O R H + O H H S i R O O R O H R O + H O R + H + ( B ) Acid Catalyzed Condensation: R S i ( O H ) 3 + H + R S i ( O H ) 2 O H 2 + R S i ( O H ) 2 + R S i ( O H ) 3 O H 2 + R S i ( O H ) 2 O S i ( O H ) 2 R + H 3 O + ( C ) Base Catalyzed Hydrolysis: H O + S i O R R O R O R O S i O R R O O R O R H O d d H O O R S i O R O R + R O ( D ) Base Catalyzed Condensation: R S i ( O H ) 3 + O H R S i ( O H ) 2 O + H 2 O R S i ( O H ) 2 O + R S i ( O H ) 3 R S i ( O H ) 2 O S i ( O H ) 2 R + O H Scheme 2 2 Hydrolysis and Condensation mechanisms: ( A ) acid catalyzed hydrolysis, ( B ) acid catalyzed condensation, ( C ) base catalyzed hydrolysis, and ( D ) base catalyzed condensation. Adapted from .
37 Ti , Zr , Al , Ge , and Si  are commonly used. Silicon alkoxides are generally used because their chemistry is well known, the Si O bond is stable, and they are widely available. However, other metal alkoxides have been used to produce sol gel materials because these precursors can help create sol gels that have advantages over their silica based counterparts. Some of these advantages include superior pH, chemical, and thermal stability . 220.127.116.11 Precursors The rate of hydrolysis can be controlled by careful selection of the precursor. Common sol gel precursors have tetra alkoxy substitution. Length of the alkyl chain of the alkoxy moiety plays a role in the rate of hydrolysis. Longer alkyl chains slow both hydrolysis and condensation reactions due to electronic and steri c influence . Also, alkoxides that are substituted with alkyl or aryl groups can slow down hydrolysis rates due to steric hindrance. Alkyl or aryl substituted precursors can also introduce flexibility into the sol gel by producing more open networks . In some cases (such as the preparation of a monolith) flexibility of a sol gel material is desirable in order to avoid cracking during the drying step. The mechanism and rate of sol gel reactions are dependent upon the coordination number and the p artial positive charge of the sol gel precursors metal atom . If the metal has a high charge density then hydrolysis will be fast such as when using transition metal alkoxides. 18.104.22.168 Solvents Solvents play a vital role in maintaining system homog eneity during initial sol gel reactions. The nature of the system ultimately dictates the choice of solvent that is utilized. In systems with intermediate polarity mixed solvent systems may be required. The choice of solvent can also play a role is affe cting the rate of sol gel reactions. For
38 example, when using a precursor such as tetramethoxysilane, one of the products of hydrolysis is methanol. Therefore, if methanol or an alcohol is used as a solvent this may promote esterification and/or favor the reactants (the reverse reaction) slowing the rate of hydrolysis . Transition metal alkoxides are known to undergo hydrolysis reactions very quickly due to increased charge density on the metal . Solvents are inert in sol gel reactions, but they can be used to control reaction kinetics. For example, a nonpolar, aprotic solvent can form an oligomer with a transition metal sol gel precursor via alkoxy bridges through a nucleophilic addition (figure 2 2) . These oligomers are more stable towar ds hydrolysis. If a polar, protic solvent is used rapid hydrolysis and precipitations might occur due to alcohol association . 2 M O R M M O R O R Figure 2 2. Example of alkoxy bridging. Reprinted with permission . 22.214.171.124 Additiv es Many additives can be included in a sol gel system for various reasons. For example, when preparing sol gel monoliths, drying control chemical additives such as N,N dimethylformamide [35, 36] or glycerol  maybe added to prevent cracking during dryi ng steps. Chelating reagents such as acetic acid or trifluoroacetic acid may be added to stabilize transition metal precursors to slow hydrolysis . In the case of sol gel sorbents or stationary phases, the incorporation of organic moieties into the s ol gel system is necessary in order for the system to interact with the analyte. Such hybrid
39 organic inorganic systems can be prepared by adding polymers, monomers, or any other organic material that can be immobilized into the 3D network of the sol gel. In the case of sol gel coatings and stationary phases, deactivating reagents are often added to the sol solution. The deactivating reagents may be needed in order to avoid tailing peaks and problems with reproducibility caused by residual silanol groups on the surface of the substrate . Polymethylhydrosiloxane or hexamethyldisilazane are common deactivating reagent. These compounds have reactive hydrogen atoms that can derivatize surface silanol groups at elevated temperatures . Since these co mpounds can be added to the sol solution no extra deactivation steps are required after preparation of the coating or stationary phase . Due to the wide range of materials available for synthesis, the compositions and applications of sol gel coatings is virtually limitless. In the coming sections, sol gel coatings that have been utilized for microextraction techniques and the preparation of such coatings will be discussed. 2.2 Preparation of Sol Gel Sorbent Coated Microextraction Devices 2.2.1 Sol gel coated SPME fibers The first sol gel coated SPME fiber was developed by Malik and co workers . Fused silica fibers or rods are often used in SPME devices. Silanol groups present on the surface of the fused silica fiber serve as binding sites for sol gel sorbents. Fibers are prepared for sol gel coating, by removing any coatings such as protective polyimide, present on the surface of an end segment of the fiber by burning in a flame or by dipping the fiber in an appropriate organic solvent (e. g. acetone) or acid for several hours. Next, silanol groups are exposed on the surface by successively rinsing the fiber with 1M
40 NaOH followed by rinsing for 30 minutes with water . A 0.1M solution of HCl is rinsed over the bare end of the fiber to n eutralize residual NaOH, and it is then rinsed with water. Following rinsing, the fiber is heated at a time and temperature dependent upon the organic components in the sol gel system while purging with inert gas in a GC injection port. Fibers are coated with sol gel materials by designing an appropriate sol solution consisting of organic component (polymer), precursor, solvent, catalyst, water, and any other additive. The fiber is then dipped vertically into the sol solution for 20 30 minutes (or any ot her designated time). If desired, the process is repeated with fresh sol solution several times in order to achieve a preferred coating thickness. The sol gel coated SPME device is then stored in a dessicator for 24 hours. Following this, it is then con ditioned under He or N 2 in a GC injection port for several hours at a temperature dictated by the thermal stability of the sol gel system. 2.2.2 Sol gel coated capillaries for in tube SPME In tube SPME (also known as capillary microextraction (CME)) often uses flexible fused silica tubing with a solvent coating on the inner surface as the extraction device. Disadvantages of traditional coating used with in tube SPME include limited sample capacity due to thin coatings in the capillaries (typically a piece of conventionally coated GC column) and poor thermal and solvent stability. To address these issues, Malik and co workers introduced sol gel CME . Briefly, the inner surface of the capillary is hyrdrothermally treated to expose surface silanol groups Next, the pretreated capillary is coated with a sol solution. Then a filling/purging device can be used to fill the capillary with sol solution . The sol solution is kept inside the capillary for an
41 optimal amount of time (about 30 min) to facilit ate the formation of a surface bonded sol gel coating. After this, the capillaries are purged for an optimal amount of time (about 60 min) with an inert gas (e.g. 20psi He) to expel any un bonded sol solution from the capillary. Capillaries are then ther mal conditioned or dried to (1) evaporate remaining solvent and (2) to cure sol gel coating. Heating conditions depend upon the temperature limits of the sol gel coated material (usually determined by the organic component). Sol gel coated microextract ion capillaries are then rinsed with organic solvents (e.g., methylene chloride and/or methanol) to clean any debris or unbonded material from the coated surface. Conditioned capillaries are then cut into smaller segments (e.g. 11 cm pieces) that are used for CME. 2.3 Characterizing Sol Gel Sorbents Sol gel sorbent coatings are characterized in order to better understand their properties and structures. The overall appearance of coating can be characterized utilizing scanning electron microscopy (SEM) [ 40]. SEM images can tell about the porosity, thickness, and stability or defects (i.e. if it is cracked) of a coating. Surfaces characteristics and morphology can also be determined using atomic force microscopy (AFM)  and x ray photoelectron spectro scopy . Fourier transform infared spectroscopy (FT IR) [43, 44] can be used to determine functional groups that are present. This is useful in verifying if additives such as polymers were incorporated into a sol gel network. FT IR is also useful in h elping to identify types of bonds present. This could be useful in mixed systems. For instance, in a system that incorporates silica and germania materials, a Si O Ge bond can be identified via IR . Types of bonds, bond formation, and kinetics of sol gel reactions can also be followed using 1 H, 13 C, or 29 Si
42 liquid or solid state NMR . Thermal stability or degradation temperatures and the ratio of inorganic to organic constituents in a sol gel material may be determined by using thermal gravimetri c analysis (TGA) . TGA follows weight change in relation to temperature change. In some cases, pore size or pore distribution must be characterized. This can be accomplished using nitrogen adsorption methods which utilizes the Brunauer Emmett Telle r (BET) equation . 2.4 Evaluating Sol Gel Sorbents Sol gel coated SPME fibers and microextraction capillaries have been used to preconcentrate a wide range of analytes. Sorbent coatings are evaluated to determine their best operating parameters. C oatings have been tested for solvent stability by rinsing with harsh organic solvents . Sol gel coatings usually remain intact after rinsing and are thus more robust than their non bonded counterparts. Coatings are also tested for their pH stability by rinsing for a period of time with acid or base. While silica based coatings are more pH stable than non bonded coatings, they still have a tendency to breakdown at low (below 2) and high pH values (above 8) . However, sol gel coatings prepared wit h transition metal alkoxides are remarkably pH stable and can survive rinsing with strong acids or bases [33, 48, 49 ] The upper temperature limits of sol gel coatings are tested in order to determine (1) their degradation point and (2) the optimum temp erature for analyte desorption. Traditional coatings sometimes have limited temperature ranges over which they are stable. For example, some traditional coated poly(ethylene glycol) materials have narrow working temperature ranges (70 270C for Carbowax 20M) [50, 51]. Sol gel coatings
43 are thermally superior to traditional coatings because they are covalently attached to substrates. SPME and CME are non exhaustive techniques. When the extraction is complete distribution equilibrium between the sample matrix and the extracting phase is reached . Extraction profile experiments are performed in order to determine when equilibrium is achieved. These experiments are conducted by extracting a sample of a constant concentration for varying interval of time. The experiments are carried out at a constant temperature. For example, a 100 ppb sample could be extracted onto a coating for 10 min, 20 min, 30 min and so on at room temperature. Initially many analyte s are sorbed leading to a s te p rise in the ex traction profile curve (linear regime). As time goes on a point is reached when the rate of analytes being sorbed and desorbed onto the coating is equal (near equilibrium). At this point the curve plateaus and the optimum extraction time is determined (f igure 2 3). Extraction times can range from a matter of minutes to hours. Ideal coatings should provide high peak areas and fast extraction times. Figure 2 3 Example of a n extraction profile. With permission [ 53].
44 For SPME fibers, it is necessar y to determine other parameters such as optimum desoprtion time and temperature to effectively release analytes from the coating for fibers used with GC. This is accomplished by testing various desoroption time intervals and temperatures. Good desorption is required in order to avoid problems with sample carryover from one run to the next. It is necessary to determine the repeatability of an extraction from one run to the next. This is accomplished by repeating extraction experiments under the same cond itions multiple (at least 3) times (i.e. run to run repeatability). The average response (usually peak area) along with the standard deviation is used to determine the relative standa rd deviation (precision) or RSD value. The reliability of the coating pr ocedure can be determined by examining the average response given for coatings prepared in different batches (i.e capillary to capillary or fiber to fiber repeatability). 2.5 Sol Gel Sorbents: Literature Survey Various types of polymers and organic fu nctionalized compounds have been used in sol gel networks. Some of the more common organic ligands used in sol gel sorbents are detailed in table 2 2. (Table 2 2 is not intended to be an exhaustive representation.) 2.5.1 PDMS The first example of a sol g el coating for SPME fibers was introduced in 1997 by Malik and co workers . A hybrid organic inorganic sol gel coating was prepared by using a solution coating containing hydroxyl terminated poly(dimethylsiloxane) (PDMS), methyltrimethoxysilane (MTMS) poly(methylhydrosiloxane) (PMHS), and trifluoroacetic acid containing 5% water (95% TFA). To form the sol gel coated fiber, a homemade syringe for SPME GC was developed. It consisted of a 15 cm piece of fused
45 silica fiber that was glued onto a 15 cm lo ng piece of PEEK tubing using silicone adhesive. The PEEK tubing served as the plunger. Exposed, pre treated fused silica fiber was dipped for 20 min in fresh sol solution three times. The coated fiber was then dipped into a solution of trimethylmethoxy silane/ methanol (4:1 v/v) to end cap the sol gel coating. Anilines, phenolic compounds, and PAHs were analyzed. The coating was characterized by enhanced thermal (320C) and solvent stability. Malik and co workers  developed fused silica capillari es with surface bonded nonpolar PDMS sol gel hybrid organic inorganic coatings were utilized and a variety of compounds were extracted with excellent reproducibility and extraction sensitivity. Specifically, the PDMS sol gel coated microextraction capilla ry was used to extract PAHs, aldehydes, and ketones, and in CME GC experiments area RSD values of less than 6% were obtained. Sol gel sorbent coatings containing PDMS have also been developed by others. Supelco developed a C 11 PDMS sol gel coated SPME f iber which was tested by Jinno et al. to analyze benzodiazepines in human urine using SPME and micro LC . Vonderheide et al. prepared a sol gel coated PDMS fiber to study seleno amino acids by direct immersion . Bagheri and co workers developed PDMS, PEG (MW 4000), and poly(ethylenepropyleneglycol) (Ucon HTF14) sol gel coated SPME fibers to study dextrorphan in human plasma . All of the coatings were prepared using the respective polymer, TMOS, and 95% TFA. SEM investigations revealed that all of the sol gel coatings were about 10 m m thick. While the PDMS and the Ucon sol gel coating
46 Table 2 2 A few common silica based sol gel sorbents Polymer/Component Name General Structure References Poly(dimethylsiloxane) H O S i O ) n C H 3 C H 3 C H 3 C H 3 O ( S i S i C H 3 C H 3 O H [15, 16, 54 58] Poly(vinyl alcohol) ( C H 2 C H ) n O H [59 62] Divinyl benzene C H C H 2 C H 2 C H [63, 64] b cyclodextrins [68 71] Calixarenes  [73 80] Crown Ethers O O O O O H [81 86] Poly(ethylene glycol) HO ( CH 2 CH 2 O ) n H [16, 56, 87, 88]
47 appeared porous, the PEG coating did not. The authors claimed that PEG coating might not be porous since PEG MW 4000 is a solid. Furthermore, t he sol gel PDMS coated fiber was found to give a better response for the analyte in question than the Ucon fiber. The sol gel PDMS coated fiber was used for headspace extraction on acetyl derivatized dextrorphan in plasma. In a more recent work , Bag heri et al used a similar PDMS sol gel coated SPME fiber to extract and analyze geosmin in water and apple juice by HS SPME GC MS analysis. Carasek and co workers used a homemade glass ceramic rod as an SPME fiber and coated it with a PDMS sol gel  (figure 2 4). A glass ceramic rod that had a thickness of 44 m m was used to overcome the issues of fused silica fiber breakage. The rod consisted of 29% Li 2 O, 1% ZrO, 5% BaO, and 65% SiO 2 A sol gel consisting of MTMS, PDMS OH, and 95% TFA was used to coat the glass ceramic rod. BTEX compounds were extracted by di rect and headspace extractions and were analyzed by GC FID. Limits of detection for the BTEX compounds were 0.2 0.7 m g/L and RSD values of 4.4 5.3 were obtained. Figure 2 4 Scanning electron micrograph of a sol gel PDMS coating on a (A) fused silica and (B) glass ceramic base. Reprinted from .
48 Sol gel sorbent coatings that consist of PDMS and other polymers have also been prepared. Augusto and company developed PDMS/poly(vinylalcohol) (PDMS/PVA) sol gel coated SPME fibers . Sol gel coated fibers were prepared using PDMS OH, MTMS, PVA, and 95% TFA. Fibers were end capped with a 20% v/v solution of methanol and trimethylmethoxysilane for 1 min. Conditioned sol gel coated fibers were characterized using TGA, FT IR, and SEM. Napthalen e, ethyl caprate, p chlorotoluene, o xylene, and PCBs were analyzed by headspace extraction followed by GC FID or GC ECD analysis. The thin, microporous PDMS/PVA sol gel coatings were found to be more thermally stable than coating prepared only with PDMS. It was also found that the PDMS/PVA sol gel had a higher affinity for heavier compounds like PCBs than PDMS alone. Ultimately, the PDMS/PVA sol gel coated fibers allowed for cleaner baselines and a wider working temperature range. The PDMS/PVA sol gel sorbent coated fibers have been used to study pesticides in plant infusions of Passiflora species [60 62]. Zuin et al. used the PDMS/PVA sol gel coated fibers to study organochlorine (OCP) and organophosphorous (OPP) pesticides in the complex infusions b y HS GC ECD . A desorption temperature of 260C for 15 minutes was used for total desorption and to avoid memory effects. da Silva et al. optimized HS SPME GC ECD extraction parameters of OCPs and OPPs from the plant infusion using a neuro genetic app roach . da Silva et al. used PDMS/PVA sol gel coated SPME fibers for direct coupling of SPME with mass spectrometry (MS) . This so called fiber introduction mass spectrometry (FIMS) was used to extract and analyze OCP and OPP in plant infusions FIMS analysis resulted in 0.6 14.9% precision and limits of detection ranging from 0.3 3.9 ng/mL.
49 Sol gel PDMS/divinylbenzene (PDMS/DVB) sorbent coatings for SPME fibers have been prepared by Zeng and co workers . In this case, the sol gel PDMS/D VB coating was prepared using VTEOS as a bridge thereby combing sol gel chemistry and crosslinking technology. Phosphates and methylphosphates were preconcentrated and analyzed. Later, Zeng and co workers developed hydroxyl terminated silicone oil (OH TS O) butyl methacrylate divinylbenzene sol gel coating for SPME fibers . The sol gel coating consisted of butylmethacrylate (BMA), DVB, OH TSO, TEOS, g methacryloxypropyl trimethoxysilane as a bridge between sol gel reactions and radical crosslinking reactions, PMHS, benzophenone, and 95% TFA. The fibers were used for HS SPME GC to analyze polar alcohols, fatty acids, and nonpolar ester. Zeng and co workers also developed sol gel coated SPME fibers consisting of a blend of OH TSO and an acrylate . Three coatings were developed and studied for their ability to extract 2 chloroethyl ethyl sulfide (CEES) (a surrogate of mustard) from red clay, sand y soil, and agricultural soil samples. Sol gel coatings contained methylacrylate, methyl methacrylate, or butyl methacrylate in addition to OH TSO, TEOS, VTEOS, PMHS, benzophenone, methylene chloride, and 95% TFA. Acrylate was present in a 3:1 ratio over OH TSO. OH TSO was used here (and in other cases) to lengthen and spread the sol gel network over the fiber and the acrylates were used as selective coating materials. VTEOS was used for radical crosslinking to bond the acrylate to the silicone oil. In this case it was determined that the BMA/OH TSO sol gel coated fiber had the best selectivity for CEES. The authors state that MA and MMA fibers would be better for more polar analytes. Zeng and co workers also used the BMA/OH TSO sol gel/cross linked coated fiber for HS SPME GC MS of medium and
50 long chain derivatized fatty acids of lung tissue from lung cancer patients and healthy controls . A BMA/OH TSO fiber was also used to analyze ephedrine derivatives from human urine by HS SPME followed by capillary electrophoresis (field amplified sample injection) . The sol gel coating consisted of BMA, OH TSO, TEOS, g methacryloxypropyltrimethoxysilane, PMHS, benzophenone, methylene chloride, and 95% TFA. After (1R, 2S) ephedrine, (1R, 2R) pseudoephedrine, and (1S, 2S) pseudoephedrine were extracted onto the fiber by HS SPME, 80 m L of solvent in a 100 m L syringe was used to back extract the analytes from the fiber. The sample was then transferred to a vial an injected into the CE for analysis. Liu et al. developed sol gel coated SPME fibers utilizing a PDMS polymer that contained 3% vinyl groups . By utilizin g this polymer, sol gel crosslinked coatings were obtained. The sol gel consisted of PDMS containing 3% vinyl groups, methylene chloride, MTMS, PMHS, and 95% TFA. The PDMS vinyl sol gel coating was used to investigate OPPs and BTX by HS SPME in water, ora nge juice, and red wine. In order to enhance selectivity for nonpolar monocyclic aromatic hydrocarbons (MAHs) and polycyclic aromatic hydrocarbons (PAHs), Hu et al. utilized anilinemethyltirethoxysilane as a sol gel precursor . Selectivity for these aromatic compounds was enhanced by p p interactions between the analytes and the phenyl rings in the sol gel network. Utilizing PDMS also helped to enhance the elasticity of the sol gel coating. The sol gel sorbent coating consisted of PDMS OH, methylen e chloride, TEOS, AMTEOS, and TFA containing 1% water. HS SPME was used for MAHs and direct SPME was used for PAHs.
51 Kulkarni et al. developed polar coatings utilizing PDMS and a cyano functionalized sol gel precursor, 3 cyanopropyltriethoxysilane . Cyano groups are considered extremely polar; however, they are thermally labile. By incorporating them into a sol gel network they become more thermally stable. The cyano PDMS coatings were utilized for CME GC, and various classes of medium polar to pola r solutes were analyzed. The sol gel coating consisted of PDMS, 3 cyanopropyltriethoxysilane, TEOS, methylene chloride, HMDS, and 95% TFA. Coatings were found to be thermally stable to 330C; beyond this temperature to cyano sol gel component began to degrade. By comparing the extraction of a sol gel cyano PDMS and a sol gel PDMS microextraction capillary, the authors were able to demonstrate that the cyano moiety was responsible for the extraction of the more polar compounds (figure 2 5). As fused sil ica SPME fibers are prone to mechanical breakage, metal wires such as NiTi (introduced by Supelco) have come onto the market. Azenha and co workers utilized a Ti alloy wire as an SPME fiber, and they coated the wire with a PDMS sol gel/silica particle bl end (figure 2 6) . The sol gel coating was prepared using PDMS OH, MTMS, and 95% TFA. After dip coating, silica particles were pressed onto the coated fiber, and the fiber was allowed to stand for 24 hours. Following this the fiber was dip coated ag ain with the PDMS based sol solution. After a 24 hour period the SPME sol gel/silica particle coated fiber was thermally conditioned to 300C in a GC injection port. Overall, the coating was 30 m m thick. The fiber was used to analyze BTEX, 2 octanone, be nzaldehyde, acetophenone, and dimethylphenol by HS SPME. By comparing a PDMS sol gel fiber and the PDMS sol gel/silica particle fiber,
52 Figure 2 5 CME GC analysis of mixture of two alcohols and two free fatty acids on (a) sol gel CN PDMS capillary ( 12 cm) and (b) sol gel PDMS capillary (12 cm); extraction time, 30 min, GC analysis conditions: 5 m 250 m m i.d.. sol gel PDMS column; splitless injection; injector temperature: initial 30 C, final 300 C, programmed at a rate of 60 C/min; GC oven temperature programmed from 35 C to 300 C at a rate of 15 C/min; helium carrier gas; FID temperature 350 C. Peaks : (1) 1 heptanol, (2) 1 octanol, (3) octanoic acid, (4) nonanoic acid. Reprinted from .
53 Figure 2 6 SEM image of PDMS/silica particle blend conditioned for 30min at 300C (magnification 150). Reprinted from . the authors obtained the same type of extraction profile. This established that the PDMS alone was responsible for the extraction. 2.5.2 b Cyclodextrins Cyclodextrins (CDs) are cyclic oligosaccharides that contain six ( a CD), seven ( b CD), or eight ( g CD) a (1,4) linked glucopyranose units . CDs can form inclusion complexes with molecules via host guest interactions . Specificall y, b CDs have been used for microextraction since they can enhance hydrogen bonding forces between the [sorptive] coating and polar aromatic compounds . b cyclodextrins have been used in sol gel coatings for in tube SPME HPLC  and for SPME fib ers [72 73]. Feng and co workers developed a sol gel coating using 3 glycidoxypropyltrimethoxysilane derived b CD, TEOS as a sol gel precursor, HCl as a catalyst, and acetonitrile as a solvent . The sol gel coating was used for in tube SPME HPLC ana lysis of non steroidal anti inflammatory drugs in urine. The authors
54 stated that the sample matrix affected the extraction due to the salts and proteins that may act as a competitive subject for analytes to form inclusion complexes with the b CDs. Li a nd co workers have developed PDMS/ b CD sol gel sorbents for membrane extraction  and SPME fibers . The sol gel sorbent consisted of a mixture of hydroxyl terminated PDMS, TEOS, 3 glycidoxypropyltrimethoxysilane, b CD, and TFA containing 1% water. PDMS/ b CD sol gel coated SPME fibers were used for HS SPME GC . Polar compounds in river, lake, and rain water were investigated. The authors claimed that increased hydrogen bonding forces to polar compounds resulted in inclusion interactions betwee n analytes with suitable dimensions to the cavity of the b CD. Zhou and Zeng utilized a b CD heptakis (2, 6 di O methyl) b cyclodextrine (DM b CD) with hydroxyl terminated silicone oil sol gel coating for SPME fibers . The sol gel coating was prepa red by mixing DM b CD, methylene chloride, silicone oil, TEOS, 3 (2 cyclooxypropoxyl)propyltrimethoxysilane), PMHS, and 95% TFA. The b CD sol gel coated fibers were used to extract ephedrine (EP) and methamphetamine (MA) from human urine. Higher extracti on efficiencies were obtained for the sol gel coated fibers compared to commercially available fibers. This was attributed to the high surface area and sample capacity afforded through the use of sol gel technology, and due to the shape selectivity, the h ydrophobic interactions, and the hydrogen bonding of the b CD cavity.
55 2.5.3 Calixarenes Calixarenes are cyclic oligomers prepared from formaldehyde and para substituted phenols by cyclic condensation under basic conditions . These molecules are useful for extractions because their structure allows for i nteractions with polar and nonpolar analytes based on hydrogen bonding, dipole dipole interactions, p p interactions, and hydrophobic interactions. The cavity shaped cyclic structure of calixarenes allows them to form inclusion complexes with some molecule s. Zeng and co workers have utilized calixarenes for sol gel sorbents for SPME fibers [77 84]. The first such sol gel sorbent was developed in 2004 . The sol gel coating was developed by combining laboratory synthesized 5, 11, 17, 23 tetra tert buty l 25, 27 diethoxy 26, 28 dihydroxycalixarene with hydroxyl terminated oil, TEOS, 3 (2 cyclooxyprpoxyl) propyltrimethoxysilane, PMHS, and 95% TFA. Ring opening polymerization of the calixarene and 3 (2 cyclooxyprpoxyl) propyltrimethoxysilane was cat alyzed by TFA. Sol gel coated SPME fibers were used in headspace format to analyze BTEX compounds, PAHs, and aromatic amines. The fibers were used in direct immersion format to analyze phthalic acid esters. Extracted compounds were analyzed by GC. The sol gel calixarene coated fibers had relatively fast extraction times (1min to 1 hour depending on the analyte), was thermally stable up to 320C, was solvent stable, and had a lifetime of 170 runs for headspace and 140 runs for direct immersion. SPME fibers based on the original calixarene sol gel coating  were used in direct immersion format followed by GC analysis to analyze chlorophenols (LOD 0.005 0.276 m g/L) in riverwater and soil samples from a lake , and were used in headspace format to analyze OCPs and their metabolites in radishes (LOD 174 ng/kg) .
56 In order to enhance sensitivity for polar compounds, a calixarene sol gel sorbent was developed that contained an amide bridge on the lower rim of the calixarene  (figure 2 7). Calixarene containing sol gel sorbents were also developed using 5, 1, 17, 23 tetra tert butyl 25, 27 dihydroxy 26, 28 diglycidyloxycalixarene. Sol gels contained the calixarene, hydroxyl terminated silicone oil, TEOS, PMHS, and 3 aminotriet hoxysilane . Other sol gel calixarene coatings  contained about 11% calixarene. Zeng and co workers state that this was because of steric hindrance of the calixarene which resulted in low reactivity with 3 (2 cyclooxyprpoxyl) propyltrimethoxys ilane at room temperature . With the diglycidyloxycalixarene sol gel composition, sol gel reactions were fast; the precursor, 3 aminotriethoxysilane, acted as a catalyst and allowed ring opening reactions to occur. This sol gel coating contained a bout 17% calixarene . Furthermore, the sensitivity to polar compounds was enhanced because epoxy groups were incorporated into the lower rim of the calixarene. SPME fibers were prepared with the sol gel sorbent, and propranolol, a b blocker, was extr acted from human urine samples by headspace and direct extraction followed by GC analysis . The diglycidyloxycalixarene sol gel sorbent was also used for HS SPME GC analysis of PAHs and aromatic amines; limits of detection in ng/L range were achie ved . The performance of the diglycidyloxycalixarene sol gel sorbent was compared to the performance of the amide bridged calixarene sol gel, silicone oil sol gel, and commercial PDMS SPME fibers for extracting chlorobenzenes followed by GC ECD . Standard addition technique was used to analyze chlorobenzenes in red clay, sandy soil, and garden soil. The diglycidyloxycalixarene sol gel sorbent had a higher
57 Figure 2 7 The structure of amide bridged calixarene. Reprinted with per mission from [ 79]. selectivity for the chlorobenzenes than the other SPME fibers. The authors attributed this to the presence of epoxy groups in the calixarenes structure. In a further expansion, the diglycidyloxycalixarene sol gel sorb ent was used for a solid phase extraction coupled to capillary electrophoresis . Headspace SPME was used to sorb propanolol from urine, and solvent was used to back extract the analyte from the sorbent. Samples were then analyzed off line with capi llary zone electrophoresis. 2.5.4 Crown Ethers Crown ethers have been utilized in chromatography because they are known to give good selectivity of polar compounds with similar boiling points due to their cavity structures and the strong electronegativi ty of heteroatoms on the crown ether ring [85, 86]. Crown ether sol gel materials have been used as coatings for SPME in several cases. Zeng and co workers utilized hydroxydibenzo 14 crown 4 (OH DB14C4) in combination with hydroxyl terminated silicone o il as an SPME coating to analyze
58 phenols . In this case, 3 (2 cyclooxypropoxyl) propyltrimethoxysilane and TFA were used for ring opening reactions of OH DB14C4. TEOS was used as a sol gel precursor. Overall, this crown ether sol gel coating had medi um polarity because of hetroatoms on the ring. Phenolic compounds from paper mill wastewater samples were analyzed, and detection limits less than 0.1 ng/mL were obtained for HS SPME GC. Zeng and co workers also developed crown ether sol gel SPME coat ings to analyze monocyclic aromatic amines . Three crown ether sol gel coating were prepared including: OH DB14C4, dihydroxy substituted saturated urushiol crown ether (DHSU14C4), and 3,5 dibutyl unsymmetry dibenzo 14 crown 4 dihydroxycrownether (DBU D14C4). Ring open reactions of the crown ethers was accomplished using 3 (2 cyclooxypropoxyl) propyltrimethoxysilane. All coatings contained silicone oil and TEOS as a sol gel precursor. Analysis was performed by HS SPME GC. Extraction efficiencies wer e found to decrease with increasing number of alkyl groups on the crown ether ring. This was due to a decrease in the coatings polarity (hence the analytes were less attracted to it) and due to an increase in steric hindrance. Amines were extracted from pharmaceutical waste water. Yun developed a sol gel open crown ether SPME coating using a w diallytriethylene glycol with silicone oil (DATEG/OH TSO) . Vinyltriethoxysilane (VTEOS) and TEOS were used as sol gel precursors. VTEOS was also used to cross link the crown ether into the sol gel network. Sol gel SPME coated fibers were prepared by dip coating followed by UV irradiation to induce cross linking reactions. The coating was used to extract phenols and BTEX compounds via headspace extractio n.
59 Wu and co workers have also prepared crown ether sol gel coatings for SPME. In one case, dihydroxy terminated benzo 15 crown 5 (DOH B15C5) was prepared and used for a sol gel coating . This crown ether was used to try to extend the linear range of phenols. The coating consisted of DOH B15C5, silicone oil, PMHS, TEOS (sol gel precursor), methylene chloride, and TFA as a catalyst. In this work, the sol solution was optimized to lengthen the time of the sol solution to coat different thicknesses of f ibers using the same sol solutions as opposed to making a fresh solution for each dip. The crown ether was found to have enhanced hydrogen bonding; therefore, high temperatures were used for desorption to mitigate the carryover problem. The coating was al so used to extract chlorobenzenes and arylamines. Overall, the DOH B15C5 sol gel coating had less steric hindrance than other crown ether coatings, and it had high thermal and solvent stability. Wu and co workers utilized allyloxy bisbenzo 16 crown 5 tr imethoxysilane with silicone oil to prepare a sol gel coated SPME fiber . The sol gel was prepared using TEOS as a sol gel precursor, PMHS as a deactivator, methylene chloride, and 95% TFA. The crown ether silane served as a second sol gel precurso r. Furthermore, 33% crown ether was contained in the sol gel whereas other coatings  contained only 3% crown ether. The SPME fiber was used in the headspace and direct immersion modes. OPPs in food matrices were investigated. Wu and co workers als o prepared radical cross linked vinyl crown ether sol gel coated SPME fibers . In this work, 4 allyldibenzo 18 crown 5 (allyl DB18C6), 3 allyldibenzo 15 crown 5 (allyl B15C5), and allyloxy ethoxymethyl 18 crown 6 (allyl PS018C6) were utilized in com bination with VTEOS and were cross linked using AIBN
60 as an initiator. Silicone oil, TEOS, PMHS, TFA with 5% water, and methylene chloride were also used in the sol gel. UV light was used to start cross linking reactions. The vinyl crown ether sol gel co ated fibers were used for headspace and direct SPME. Extractions of OPPs in food matrices were investigated using SPME and GC FPD. Extractions of the pesticides were higher for allyl B15C5 and the allyl DB18C6 fibers. The authors claim that this was bec ause the benzyl groups in the structures could have p p interactions with the analytes. Furthermore the allyl B15C5 sol gel coated fiber performed the best due to one benzyl ring on the crown ether ring thus enabling an electron distribution that deviated from symmetry. A bigger dipole moment and thus higher polarity resulted. Interestingly, Zeng and co workers developed calix open chain crown ether sol gel coating for SPME . So called calixcrowns are macropolycyclic molecules in which monocycl ic structures of calixarenes are combined with crown ethers by bridging phenolic oxygens of calixarene by a polyether chain. The combination allows for extraction of polar aromatic compounds via p p interactions, hydrophobic interactions, and the cavity shaped structure of the calixarene and because of the cavity structure and strong electronegative effect of heteroatoms on the crown ether ring. A combination of sol gel and cross linking was used to prepare coatings that contained 5, 11, 17, 23 tetra tert butyl 25, 27 di(2 allyloxyethoxyl) 26, 28 dihydroxycalixarene (C open chain crown ether), VTEOS, silicone oil, TEOS, PMHS, benzophenone, and TFA containing 5% water. The sol gel calixcrown coating was used to extract phenolic compounds, alcohols, a nd fatty acids in wine.
61 2.5.5 Poly(ethylene glycol) Poly(ethylene glycol) (PEG) is a polar polymer that has been used in a variety of areas in an effort to analyze polar compounds. However, PEG is known to have low thermal stability, and it can undergo chemical degradation in the presence of trace levels of oxygen and water . Effective immobilization of PEG can help alleviate these issues. One means of immobilizing PEG is using sol gel technology. Malik and co workers developed sol gel CME coatin g using trimethoxysilyl terminated PEG (MW 5000) . Alcohols, phenols, and amines were tested, and RSD values of less than 4% were obtained. Limit of detection in the parts per quadrillion were obtained for some analytes. Wu and co workers prepared S uperox 4 (PEG MW 4,000,000) sol gel coated fibers for SPME . The sol gel coating contained MTMS, Superox 4, acetone, and 95% TFA. The coating was found to be porous creating a high extraction capacity due to the increased surface area. The PEG sol g el coated fiber was used to analyze BTEX compounds and phenols by HS SPME GC FID. The coating was stable up to 300C and had a lifetime of over 150 uses. Augusto and co workers prepared so gel sorbent coatings for SPME fibers using Carbowax 20M (PEG MW of 14000 16000) . The sol gel coating contained MTMS, Carbowax 20M, and 95% TFA. The coating had a porous, sponge like coating, and the authors stated that it seemed to consist of an agglomerate of microspheres (figure 2 8). The PEG sol gel coated f iber was used to preconcentrate BTEX compounds. Kulkarni et al.  utilized N (triethoxysilylpropyl) O polyethylene oxide urethane (TESP) to create a sol gel coated microextraction capillary for CME GC. Low molecular weight PEGs (e.g., MW 600) are liqu ids at room temperature and their hydrophilic end groups ( OH)
62 have more pronounced effects on their properties as extraction media or chromatographic stationary phases than higher molecular weight PEGs . However, chemical immobilization of low molecu lar weight PEGs on the silica surface presents a difficult task. The sol gel coating in this case contained MTMS and TESP as sol gel precursors, methylene chloride, and 90% TFA. By utilizing TESP (which contains 4 6 low ethylene oxide repeating units) as a sol gel co precursor, a growing 3 D sol gel network was created that contained covalently bonded PEG (figure 2 9). The coating was solvent resistant and thermally stable up to 340C. The PEG sol gel coated microextraction capillaries were used to prec oncentrate aldehydes, ketones, aromatic amines, phenols, alcohols, and free fatty acids without derivatization, salting out, or pH adjustments. Figure 2 8 Electron scanning micrography (600 magnification) of a sol gel Carbowax 20M ormosil fib er. Reprinted from .
63 Figure 2 9 Sol gel PEG coating chemically anchored to the inner walls of fused silica capillary. Reprinted from . 2.5.6 Organic Modified Silica Sol gel sorbents can be prepared using sol gel precursors or co pre cursors that contain organic moieties. These moieties interact with target analytes from samples which resulted in their preconcentaration. Caruso and co workers developed the first instance of an HPLC application of sol gel coated SPME fibers [ 23 ]. In this case, sol gel coated SPME fibers were used to analyze organo As, Hg, and Sn compounds. The sol gel consisted of n octyltriethoxysilane (C 8 TEOS), MTMS, methanol, HCl, and water. Various molar ratios of precursors were tested. A ratio of 2:1 C 8 TEOS : MTMS was found to give optimum extraction efficiencies. C 8 TEOS provided hydrophobicity, and MTMS was used to avoid shrinking and cracking. Sol gel coated fibers were end capped using a 4:1 v/v solution of trimethylmethoxysilane and methanol. After dir ect immersion SPME, a static desorption technique was used in which a desorption chamber replaced the injection loop. Analytes were desorbed by soaking the fiber in the mobile phase for 5
64 min. Limits of detections of 412 m g/L for diphenylmercury, 80 m g/L for triphenylarsine, and 647 m g/L for trimethylphenyltin were obtained The sol gel coating had RSD for extraction efficiency of 21, 29, and 2 2%, respectively. The high RSD values were attributed to non uniformity of the sol gel coating and to exposed portions of activated fused silica due to difficulties in controlling the removal of the polyimide coating. Azenha and co workers [ 97 ] developed sol gel coated SPME fibers utilizing phenyltrimethoxysilane (PTMOS) and MTMS as precursors. The sol gel coating consisted of various ratios of PTMOS: MTMS, water, methanol, and a catalyst (HCl, HF, or NaOH). The sol gel coating had a thickness of 0.2 1 m m, and it had a dense, non porous microstructure. The sol gel coating was used to preconcentrate long chain and non polar aromatic compounds. Factor analysis was used to determine the significant effects and interactions in order to optimize the coating and extraction process. The water to precursor ratio was found to provide the most important contribution in order to obtain a higher response [i.e. better extraction]. Fan et al. [9 8 ] developed ordered mesoporous octadecyl modified silica coating fo r in tube SPME coupled to HPLC. Bisphenol A was preconcentrated and analyzed from tap and Donghu lake water. In this case, the inner surface of a fused silica capillary was coated with an ordered mesoporous silica film. The films were crated using a sol ution that contained TEOS, HCl, ethanol and either Pluronic P123 (EO 20 PO 70 EO 20 ) or F127 (EO 106 PO 70 EO 106 ). Then the film was modified with octadecyltrimethoxysilane by filling the capillary with a solution of the compound in toluene, sealing the ends with silicone rubber, and heating at 105C for 24 hours. Sol gel sorbent coatings prepared
65 with P123 had the most ordered pore structure, uniform distribution, and had the highest extraction ability. Zheng and Hu developed modified silica sol gel coatings fo r inductively coupled plasma mass spectrometry (ICP MS) [ 99 ] and for ICP atomic emission spectrometry (ICP AES) [ 100 ]. Capillary microextraction was combined with ICP MS to preconcentrate and analyze Cu, Zn, Ni, Hg, and Cd in rice flour, mussel flesh, hum an hair, serum, and urine [ 99 ]. A sol gel coating solution contained N (2 aminoethyl) 3 aminopropyl trimethoxysilane (AAPTS), cetyltrimethylammonium bromide as a template, a ethanol and water mixture, and TMOS. No acid or base was utilized in the sol so lution because it would cause a precipitation. A fused silica capillary was filled 3 times with the solution and was then heated in a muffle furnace at 120C for 8 hours. AAPTS was utilized since it has metal chelating ability, rapid sorption, low swelli ng, and high mass exchange. Since inorganic acids are used to desorb metals for trace analysis, and AAPTS sol gel coating was used since it can be pH resistant. In particular, this coating was useful in a pH range of 2 9, and the coating survived 8 hours of rinsing with 4M HCl. Capillary microextraction combined with ICP AES was used to preconcentrate and analyze Cu, Hg, and Pb in human urine, serum, hair, milk powder, preserved egg, and water [ 100 ]. A sol gel 3 mercaptopropyltrimethoxysilane (MPTS) m odified silica coating was used. The coating consisted of MPTS, TMOS, and an ethanol and HCl mixture. MPTS was used in the sol solution since the terminal SH had good affinity to heavy metal ions. The coating had spherical particles, was propous, and h ad a high
66 absorption capacity. However, at too high of an acid concentration (2M HCl) the coating broke down. Feng and co workers [ 101 ] developed an octyl functionalized silica monolithic column for in tube SPME in hyphenation with m HPLC to preconcentr ate and analyze PAHs. The sol gel monolith was developed using a 2 step acid/base catalyzed hydrolysis/co condensation method. First, methanol, water, HCl, TEOS and C 8 TEOS were hydrolyzed at 60C for 5 hours. After cooling to room temperature dodecyla mine was added and a capillary was filled and sealed with silicone rubber. The capillary was heated at 70C for 48 hours. The morphology of the monolith was macroporous with microglobules interconnected making large clusters of a continuous skeleton (fig ure 2 10). The authors stated that the morphology was determined by a competitive process of phase seaparation accompanied by sol gel transition. C 8 TEOS was responsible for inducing the phase separation and controlling the size of the skeleton. The mono lithic sorptive column had high permeability and low flow resistance. CME m HPLC experiments were performed using a home built system. The monolithic column was used like a sampling loop. 2.5.7 Other sol gel extraction media Sol gel sorbents have all been made using various other polymers and materials. Giardina and Olesik [ 102 ] utilized low temperature glassy carbon (LTGC) films as a sorptive coating in SPME fibers. Glassy carbon has a flat surface, and it is non porous allowing for greater shape selectivity than bonded silica phases. First, porous silica particles were coated with diethylnyl oligomer via heating at 300 1000C in a furnace. Then sol gel was used to immobilize the particles onto a stainless steel fiber in a
67 Figure 2 10 SEM images of the cross sectional view of monolithic columns prepared with 130 m L (A), 150 m L (B) and 170 m L (C) of TEOS in the original sols, respectively. C8 TEOS was 100 m L for every column. Reprinted from [ 101 ]. manner similar to making porous frits. Essentially, a sol gel consisting of formamide and Kasil #1 was coated onto t he fiber. The coated fiber was then dipped into a small amount of particles resulting in the LTGC coated fiber. The authors claimed that while the coating was likely not bonded to the steel fiber, it did have good adhesion to it. It was
68 determined th at the oligomers were the dominant extraction mechanism since bare silica particles resulted in poor extractions. The sol gel immobilized LTGC fiber was used for HS SPME GC analysis of BTEX compounds, styrene and ethylbenzene, and odor compounds 2,4,6 trichloroanisole, geosmin, and 2 methylisoborneol. Direct immersion SPME was used for cis and trans stilbene. Azenha et al. developed a silica particle coated NiTi alloy SPME wire [ 103 ]. The functionalized silica particles were held in place on the S PME wire using UV curable sol gel as a glue. Wu and co workers [ 104 ] described a hydroxyfullerene sol gel SPME coated fiber. Fullerenes have spherical shapes and a c onjugated p electron system; thus, they have good selectivity for aromatic compounds. The sol gel sorbent was created from a solution that consisted of fullerol, water, MTMS, OH TSO, and 99% TFA. The sol gel coated fiber was used to preconcentrate PCBs PAHs, and aromatic amines by HS SPME. In particular, planar conformation selectivity and molecular recognition resulted due to charge transfer between PCBs and the fullerol coating. Wu and co workers [ 105 ] also developed a sol gel polyphenylmethyl siloxane (PPMS) coated fiber. The sorbent was used for SPME with MAE to analyze OCPs in Chinese tea. PPMS was used as it had higher sensitivity and selectivity for OCPs compared to PDMS. This was because PPMS gave coating with higher surface areas and i t also contributed p p interactions which helped extract some compounds. The PPMS sol gel consisted of PPMS, silicone oil, TEOS, VTEOS, PMHS, AIBN, and methylene chloride. Analytes were preconcentrated by either HS or direct immersion SPME. MAE was used on the tea prior to SPME.
6 9 Dendrimer sol gel coatings for CME GC were developed by Malik and co workers [ 106 ]. In this case, a phenyl terminated dendrimer with a thriethoxysilyl containing root was used as a sol gel active organic ligand (figure 2 11). Besides the dendrimer, MTMS, HMDS, PMHS, TFA, and methylene chloride were used to prepare the sol gel for the microextraction capillary. The coating was used to preconcentrate PAHs, aldehydes, ketones, phenols, and alcohols. Parts per trillion detecti on limits were achieved for all the compounds. Figure 2 11 Phenyl terminated dendrimer with a triethoxysilyl root. Reprinted from [ 106 ]. Malik and co workers also developed a medium polarity sol gel sorbent coating with high sensitivity s ample preconcentration by CME GC [ 107 ]. The sol gel coating consisted of polytetrahydrofuran 250 (PolyTHF 250), MTMS, HMDS, 95% TFA, and methylene chloride. The coating was thermally stable (up to 320C) and solvent resistant. It had a thickness of 0.5 m m. The medium polarity coating was used to extract non polar and polar analytes. Several classes of compounds were analyzed including
70 PAHs, aldehydes, ketones, phenols, and alcohols. Detection limits of parts per quadrillion and parts per trillion we re achieved. Basheer et al. [ 108 ] synthesized and utilized amphiphilic and hydrophilic oligomers for sol gel coated SPME fibers. Sol gel coatings were prepared from oligomer dissolved in tetrahydrofuran, TEOS, water, and ammonium hydroxide. The coating w as used to analyze OCPs, triazine herbicides, estrogen, alkyl phenols, and Bisphenol A by direct immersion SPME followed by GC MS SIM. The oligomer coatings were found to be give better performance than commercial fibers. They were selective and sensitiv e for both polar and non polar analytes. Segro and Malik [1 09 ] developed a sol gel sorbent coating consisting only of MTMS; it did not contain any other precursor. The sorbent was created from a solution containing PMHS, MTMS, methylene chloride, and 85% TFA. The sol gel coating was used for CME HPLC. The microextraction capillary served as a sampling loop in an HPLC set up. The sol gel coating was utilized to preconcentrate PAHs, ketones, phenols, alcohols, and amines. The authors stated that the MTMS sol gel coating was able to extract such a range of analytes due to molecular level interactions between methyl group of the precursor and the aromatic rings of the compounds. It was evident that residual silanol groups in the sol gel coating might have played a role in extractions as well. Segro and Malik also developed polydimethyldiphenylsiloxane sol gel coatings for capillary microextraction [ 110 ]. The coatings were used to extract analytes from aqueous matrices and were analyzed by HPLC UV. Bagheri and co workers [ 111 ] developed amino functionalized polymers synthesized using 3 (trimethoxysilyl) proply amine as a precursor. The SPME coated
71 fibers were used to extract chlorophenols from aqueous samples followed by GC MS analysis. Bianchi and co worke rs reported sol gel SPME coatings based on (3 aminopropyl)triethoxysilane diethoxydiphenylsilane [11 2 ] and diethoxydiphenylsilane [11 3 ]. The sol gel coated fibers had excellent thermal stability (400C) and were used to extract analytes from various matri ces. Bianchi and co workers also developed quinoxaline bridged cavitand based sol gel SPME fibers [11 4 ]. The novel sol gel fibers had excellent thermal stability and were used to extract chlorobenzenez from river water. Li et al. developed molecularly im printed sol gel SPME coating [11 5 ]. The imprinted coating was developed using phenyltrimethoxysilane and TEOS as precursors. The sol gel coatings were used to selectively extract congeners of polybrominated diphenyl ethers. 2.5.8 Metal Alkoxide Based So l Gel Sorbents Recently, metal alkoxides have been utilized to develop hybrid organic inorganic sol gel sorbents for SPME and in tube SPME. Utilizing metal alkoxides in place of silicon alkoxides enables the development of sorbents that are stable to hi gh and low pH extremes [11 6 ]. Kim et al. developed PDMS titania sol gel coated microextrraction capillaries for CME HPLC . The titania sol gel sorbent was used to preconcentrate polycyclic aromatic hydrocarbons, ketones, and alkylbenzenes. Run t o ru n peak area RSD values of less than 10% were obtained for all compounds. The PDMS titania coating was also shown to be pH resistant as it survived rinsing with NaOH for 12 hours. Zeng and co workers utilized sol gel OH TSO titania as a sorbent for SPME fibers [11 7 ]. The sol gel SPME fiber was used to extract aromatic amines, phenols, and
72 PAHs followed by GC analysis. The coating was pH stable as it survived rinsing with acid and base for 12 hours. It was thermally stable up to 320C. The OH TSO titan ia sol gel coating could be used repeatably as RSD values of less than 7% were obtained. Wu et al. developed mesoporous sol gel titania coating [11 8 ]. The novel coating was used to preconcentrate V, Cr, and Cu from biological samples by CME. Preconc entrated target analytes were determined by electrothermal vaporization inductively coupled plasma mass spectrometry detection. Goo d precision was obtained as RSD values of less than 6.4% were obtained for all analytes. Recently, Farhadi et al. developed a titania graphite sol gel sorbent for SPME fibers [1 19 ]. The novel coating was used to preconcentrate and analyze BTEX compounds in indoor air and headspace of soil samples. The fiber was thermally stable up to 250C. The fiber provided LODs of 0.2 0. 7 ng/mL for gas samples and 8 20 ng/mL for soil samples. Farhadi et al. also developed titania sol gel using tetrabutyl orthotitanant as a precursor [12 0 ]. A poly(dimethyldiphenylsiloxane) zirconia sol gel sorbent coating was prepared and utilized for CME GC by Malik and co workers . The coating was used to extract PAHs, aliphatic aldehydes, and aromatic ketones. Peak area run to run RSD values were less than 7.25%. The coating was pH stable as it survived rinsing with NaOH for 24 hours. In a d ifferent approach, Xu and Lee [12 1 ] developed a zirconia hollow fiber membrane using sol gel technology. The hollow fiber had a bimodal porous substructure and throughpores. The hollow fiber membrane was used to preconcentrate nerve agent
73 degradation pro ducts by microextraction followed by LC MS analysis. The zirconia fiber was found to be selective for compounds that contained phosphonic acid. Budziak and co workers developed novel sol gel PEG zirconium oxide [12 2 12 3 ] and sol gel PDMS zirconium oxi de coatings [12 4 ] on NiTi alloy SPME wires. The coatings were robust and could be used for a wide range of analytes. Sol gel OH TSO alumina sorbents were prepared for SPME by Zeng and co workers [12 5 ]. The coating was pH stable. Alumina has ligand excha nge properties and the authors stated that alumina based coating was structurally superior for the extraction of fatty acids, phenols, alcohols, aldehydes and amines. The sol gel OH TSO alumina fiber was used to analyze alcohols and fatty acids in beer wi th precision as RSD values were below 9%. Fang et al. developed sol gel germania based sorbents for CME GC . Hybrid organic inorganic coatings were prepared by using tetramethoxygermane as a sol gel precursor and PDMS, poly(dimethyldiphenylsiloxane) or 3 aminopropyltirmethoxy silane to provide organic moieties. The capillaries were used to analyze PAHs, ketones, alcohols, acids, aldehydes, and phenols. Overall the coatings were stable to high and low pH values (1 13) and solvents. Over the last d ecade, sol gel based microextraction sorbents have been widely investigated. Hybrid organic inorganic coatings have been developed for a wide variety of analytes. Metal based precursors have been used to develop pH resistant sol gel coatings. Future dir ections include development of polar [12 6 ] sol gel sorbents. Sol gel microextraction sorbents have enabled enhanced preconcentration over traditional sorbents owing to their thermal stability, solvent resistance, and high sample capacities.
74 Overall, the sol gel approach has secured its place in sample preparation by (1) decreasing fiber and capillary preparation times and manufacturing costs by combining surface treatment, deactivation, coating, and sorbent immobilization into a single step; and (2) by fa cilitation the growth of custom designs sorbents. 2.6 References for Chapter Two  M. Ebelmen, Ann. Chim. Phys. 15 (1845) 319.  W. Geffecken, E. Berger, German Patent 736 411, May 1939.  S. Kistler, J. Phys. Chem. 36 (1932) 52.  R. Roy, J. Am. Ceram. Soc 39 (1956) 145.  L. Levene, I. Thomas, in U.S. Patent 3,640,093, 1972.  H. Dislich, Angew Chem. 10 (1971) 363.  B. Yoldas, J. Mater. Sci. 12 (1977) 1203.  M. Yamane, A, Shinji, T. Sakaino, J. Mater. Sci. 13 (1978) 865.  www.chemat.com/assets/images/Flowch at72.jpg, 2005.  L. Hench, Science of Ceramic Chemical Provessing, eds., L. Hench and D. Ulrich. Wiley, New York, 1986, pp. 52 64.  H. Schmidt, J. Non Cryst. Solids 73 (1985) 681.  H. Cortes, C. Pfeiffer, B.Richter, T. Stevens, J. High Resolut. Chromatogr. Commun. 10 (1987) 446.  Y. Guo, L. Colon, Anal. Chem. 67 (1995) 2511.  D. Wang, S. Chong, A. Malik, Anal. Chem. 69 (1997) 4566.  S. Chong, D. Wang, J. Hayes, B. Wilhite, A. Malik, Anal. Chem. 69 (1997) 3889.  S. Bigham, J. Medlar, A. Kabir, C. Shende, A. Alli, A. Malik, Anal. Chem. 74 (2002) 752.
75  W. Li, D. Fries, A. Malik, J. Chromatogr. A 1044 (2004) 23.  O. Nunez, K. Nakanishi, N. Tanaka, J. Chromatogr. A 1191 (2008) 231.  G. M. Gross, V. R. Reid, R. E. Synovec Curr. Amal. Chem. 1 (2005) 135.  J. Mackenz ie, J. Non Cryst. Solids 48 (1982) 1.  C. Brinker, G. Scherer, G. Sol Gel Science: The Physics and Chemistry of Sol Gel Processing. San Diego, Academic Press, 1990.  J. Livage, M. Henry, C. Sanchez, J. Solid State Chem. 18 (1988) 259.  P. Gbatu, K. Sutton J. Caruso, Anal. Chim. Acta 402 (1999) 67.  J. Livage, M. Henry, C. Sanchez, J. Solid State Chem. 19 (1988) 256.  J. D. Mackenzie, U. R. Ulrich, D.R. eds. Ultrastructure Processing of Advanced Ceramics, New York, Wiley, 1988.  R. J. P. Corriu, J. J. E. Moreau, P. Thepot, M. W. C. Man, Chem. Mater. 4 (1992) 1217.  Y. Attia, Sol Gel Processing a nd Applications. Plenum Press, New York, 1994.  K. Sharp, J. Sol Gel Sci. Technol. 2 (1994) 35.  J. Livage, M. Henry, C. Sanchez, Prog. Solid St. Chem. 18 (1988) 25 9.  J. Livage, C. Sanchez, M. Henry, S. Doeuff, Solid State Ionics 32 (1998) 633.  M. Yamane, J. Sol Gel Sci. Technol. 13 (1998) 821.  D. J. Suh, T. J. Park, J. H. Kim, K. L. Kim, J. Non Cryst. Solids 225 (1998) 168.  L. Fang, S. Kulkarni, K. Alhooshani, A Malik, Anal. Chem. 79 (2007) 9441  S. Dai, Y. Ju, H. Gao, J. Lin, S. Pennycook, S.; C. Barnes, Chem. Commun. (2000) 243.  T. Adachi, S. Sakka, J. Mater. Sci. 22 (1987) 4407.  T. Adachi, S. Sakka, J. Non Cryst. Solids 99 (1988) 118.  S. Wolfrum, J. Mater Sci. Lett. 6 (1987) 706.
76  C. Woolley, K. Markides, M. Lee, J. Chromatog. 367 (1986) 9.  Z. Zeng, W. Qiu, H. Xing, Z. Huang, Anal. Sci. 16 (2000) 851.  A. Fidalgo, L. Ilharco, J. Sol gel. Sci. Technol. 26 (2003) 357.  J. Jang, J. Ha, Polym. Prepr. 43 (200 2) 582.  M. Spirkova, J. Brus, D. Hlavanta, H. Kamisova, L. Matejka, A. Strachota, J. Applied Polymer Sci. 92 (2004) 937.  P. Innocenzi, J. Non Cryst. Solids 316 (2003) 309.  A. Fidalgo, L. Ilharco, J. Sol gel Sc. Technol. 26 (2003) 363.  S. A. Rodriguez, L. A. Colon, Chem. Mater. 11 (1999) 754.  R. B. H. Tahar, T. Ban, Y. Ohya, Y. Takahasi, J. App. Phy. 82 (1997) 865.  S. Brunauer, P. Emmett, E. Teller, JACS 60 (1938) 309.  T Y Kim, K. Alhooshani, A. Kabir, D. P. Fries, A. Malik, J. Chromatogr. A 1047 (200 4) 165.  K. Alhooshani, T Y, Kim, A. Kabir, Abdul Malik, J. Chromatogr. A 1062 (2005) 1.  M. Cignek, M. Dressler, J. Teply, J. Chromatogr. 588 (1991) 225.  J. de Zeeuw, J. Luong, TrAc Trends Anal. Chem. 21 (2002) 594.  J. Pawliszyn, Solid Phase Microextrac tion: Theory and Practice. Wiley VCH, NY, 1997.  G. Ouyang, J. Pawliszyn, Anal. Chim. Acta 627 (2008)184.  K. Jinno, M. Taniguchi, H. Sawada, M. Hayashida, Analusis 26 (1998) M27.  A. Vonderheide, M. Montes Bayon, J. Caruso, Analyst 127 (2002) 79.  H. Ba gheri, A. Es haghi, M. Rouini, J. Chromatogr. B 818 (2005) 147.  H. Bagheri, A. Aghakhani, A. Es haghi, Chromatographia 66 (2007) 779.
77  A. Fernandes de Oliveira, C. Berto de Silveira, S. Denofre de Campos, E. Antonio de Campos, E. Carasek, Chromatographia 61 (2005) 277.  A. Lopes, F. Augusto, J. Chromatogr. A 1056 (2004) 13.  V. Zuin, A. Lopes, J. Yariwake, F. Augusto, J. Chromatogr. A 1056 (2004) 21.  G. Antonia de Silva, F. Augusto, R. Poppi, J. Chromatogr. A 1138 (2007) 251.  R. Cesar de Silva, V. Zuin, J. Yar iwake, M. Eberlin, F. Augusto, J. Mass Spec. 42 (2007) 1358.  M. Liu, Z. Zeng, C. Wang, Y. Tan, H. Liu, Sepu 21 (2003) 460.  M. Liu, Z. Zeng, Y. Tian, Anal. Chim. Acta 540 (2005) 341.  M. Liu, Z. Zeng, H. Fang, J. Chromatogr. A 1076 (2005) 16.  D. C ha, M. Liu, Z. Zeng, D. Cheng, G. Zhan, Anal. Chim. Acta 572 (2006) 47.  H. Fang, M. Liu, Z. Zeng, Talanta 68 (2006) 979.  W. Liu, Y. Hu, J. Zhao, Y. Xu, Y. Guan, J. Chromatogr. A 1102 (2006) 37.  Y. Hu, Y. Fu, G. Li, Anal. Chim. Acta 567 (2006) 211.  S. Kulkarni, L. Fang, K. Alhooshani, A. Malik, J. Chromatogr. A 1124 (2006) 205.  M. Azenha, P. Nogueira, A. Fernando Silva, Anal. Chim. Acta 610 (2008) 205.  J. Zhou, Z. Zeng, Anal Chim Acta 556 (2006) 400.  Y. Fu, Y. Hu, Y. Zheng, G. Li, J. Sep. Sci. 29 (20 06) 2684.  Y. Feng, M. Xie, S. Da, Anal. Chim. Acta 403 (2000) 187.  Y. Hu, Y. Yang, J. Huang, G. Li, Anal. Chim. Acta 543 (2005) 17.  K. Iwamoto, K. Araki, S. Shinkai, Tetrahedron 47 (1991) 4325.  X. Li, Z. Zeng, S. Gao, H. Li, J. Chromatogr. A 1023 (2004 ) 15.  X. Li, Z. Zeng, J. Zhou, Anal. Chim. Acta 509 (2004) 27.
78  X. Li, Z. Zeng, J. Zhou, S. Gong, W. Wang, Y. Chen, J. Chromatogr. A 1041 (2004) 1.  C. Dong, Z. Zeng, X. Li, Talanta 66 (2005) 721.  X. Li, Z. Zeng, M. Hu, M. Mao, J. Sep. Sci. 28 (2005) 2489  X. Li, S. Gong, Z. Zeng, Chromatographia 62 (2005) 519.  X. Li, Z. Zeng, Y. Xu, Anal. Bioanal. Chem. 389 (2006) 1428.  X. Zhou, X. Li, Z. Zeng, J. Chromatogr. A 1104 (2006) 359.  Z. Zeng, W. Qui, Z. Huang, Anal. Chem. 73 (2001) 2429.  L. Yun, Anal. Chim Acta 486 (2003) 63.  Z. Zeng, W. Qui, M. Yang, X. Wei, Z. Huang, F. Li, J. Chromatogr. A 934 (2001) 51.  D. Wang, J. Xing, J. Peng, C. Wu, J. Chromatogr. A 1005 (2003) 1.  J. Yu, C. Wu, J. Xing, J. Chromatogr. A 1036 (2004) 101.  L. Cai, C. Shuling, M. C hen, C. Wu, Anal. Chim. Acta 559 (2006) 89.  M. Liu, Z. Zeng, Y. Lei, H. Li, J. Sep. Sci. 28 (2005) 2306.  J. Conder, N. Fruitwala, M. Shingari, J. Chromatogr. 269 (1983) 171.  Z. Wang, C. Xiao, C. Wu, H. Han, J. Chromatogr. A 893 (2000) 157.  R. Gomes da Costa Silva, F. Augusto, J. Chromatogr. A 1072 (2005) 7.  S. Kulkarni, A. M. Searrow, A. Malik, J. Chromatogr. A 1147 (2007) 50.  F. Bailey Jr., J. Koleske, Poly(ethylene oxide), Academic Press, New York, 1976, Chapter 6.  M. Azenha, C. Malheiro, A. Silva, J. Chromatogr. A 1069 (2005) 163.  Y. Fan, Y. Feng, Z. Shi, J. Wang, Anal. Chim. Acta 543 (2005) 1.  F. Zheng, B. Hu, Anal. Chim. Acta 605 (2007) 1.
79  F. Zheng, B. Hu, Talanta 73 (2007) 372.  M. Zheng, B. Lin, Y. Feng, J. Chromatogr. A 1164 (2007) 48.  M. Giardina, S. Olesik, Anal. Chem. 73 (2001) 5841.  M. Azenha, M. Ornelas, S. A. Fernando, J. Chromatogr. A 1216 (2009) 2302.  J. Yu, L. Dong, C. Wu, L. Wu, J. Xing, J. Chromatogr. A 978 (2002) 37.  L. Cai, J. Xing, L. Dong, C. Wu, J. Chromatogr. A 1015 (20 03) 11.  A. Kabir, C. Hamlet, K. Yoo, G. Newkome, A. Malik, J. Chromatogr. A 1034 (2004) 1.  A. Kabir, C. Hamlet, A. Malik, J. Chromatogr. A 1047 (2004) 1.  C. Basheer, S. Jegadesan, S. Valiyaveettil, H. Lee, J. Chromatogr. A 1087 (2005) 252.  S. Segro, A. Malik, J. Chromatogr. A 1200 (2008) 62.  S. S. Segro, A. Malik, J. Chromatogr. A 1205 (2008) 26.  H. Bagheri, E. Badanezhad, F. Khalilian, Anal. Chim. Acta 616 (2008) 49.  F. Bianchi, F. Bisceglie, F.; M. Careri, S. Di Berardino, A. Mangia, M. Musci, J. C hromatogr. A 1196 (2008) 15.  F. Bianchi, M. Careri, A. Mangia, M. Mattarozzi, M. Musci, J. Chromatogr. A 1196 (2008) 41.  F. Bianchi, M. Mattarozzi, P. Betti, F. Bisceglie, M. Careri, A. Mangia, L. Sidisky, S. Ongarato, E. Dalcanale, Anal. Chem. 80 (2008 ) 6423.  M. K. Li, N. Lei, C. Gong, Y. Yu, K. Lam, M. H. Lam, H. Yu, P. K. Lam, Anal. Chim. Acta 633 (2009) 197.  J. Nawrocki, M. Rigney, A. McCormick, P. Carr, J. Chromatogr. A 657 (1993) 229.  X J. Li, J. Gao, Z R. Zeng, Anal. Chim. Acta 590 (2007) 26.  Y. Wu, B.Hu, W. Hu, Z C. Jiang, B. Li, J. Mass spectrum. 42 (2007) 467.
80  K. Farhadi, M. Mamaghanian, R. Maleki, J. Haz. Mat. (2008) 677.  K. Farhadi, R. Tahmasebi, R. Maleki, Talanta 77 (2009) 1285.  L. Xu, H k. Lee, Anal. Chem., 79 (2007) 5241.  D. Budzi ak, E. Martendal, E. Carasek, Anal. Chim. Acta 629 (2008) 92.  D. Budziak, E. Martendal, E. Carasek, J. Chromatogr. A 1198 (2008) 54.  D. Budziak, E. Martendal, E. Carasek, J. Chromatogr. A 1187 (2008) 34.  M m. Liu, Y. Liu, Z r. Zeng, T y. Peng J. Chroma togr. A, 1108 (2006) 149.  N. Fontanals, R. M. Marce, E. Borrull, J. Chromatogr. A, 1152 (2007) 14.
81 CHAPTER 3: IONIC LIQUID MEDIATED SOL GEL SORBENTS FOR CAPILLARY MICROEXTRACTION 3.1 Introduction Pioneering research by Pawliszyn and co workers [1 ] on solid phase microextraction (SPME) about two decades ago was a significant step toward automation of sample preparation in chemical analysis. SPME techniques include traditional fiber SPME [1 3] and in tube SPME [4 7] in tu be SPME [4 7] also known as CME  These sample preparation techniques pose little risk to the environment and human health since they require no organic solvents. SPME is also suitable for hyphenation to various analytical techniques: GC [ 9 ], high per formance liquid chromatography (HPLC) [4, 10 ], supercritical fluid chromatography [1 1 ], capillary electrophoresis (CE) [1 2 1 3 ], mass spectrometry ( MS ) [1 4 ], and inductively coupled plasma mass spectrometry ( ICP MS) [ 1 5 ] It is portable and is especially suited for field analysis [1 6 ]. SPME is a non exhaustive extraction technique based on the principle of equilibrium extraction. In fiber SPME, a sorbent coated fiber is used to extract analytes from the sample, either by direct immersion or from the head space. Some disadvantages of fiber SPME include fiber breakage, mechanical damage of the coating during operation and handling of the SPME device, and limited sample capacity. These issues led to the development of in tube SPME  also called capillary microextraction [ 8 ]. In this new format, the sorbent coating is placed on the capillary inner wall. Analytes are extracted by passing the
82 sample through the coated capillary . In tube SPME has a significant advantage over traditional fiber SPME in t hat the sorbent coating is protected against mechanical damage during operation since it is secured on the inner wall o f a capillary. S hort segments of GC columns have been used to perform extraction by in tube SPME . Disadvantages of using convention ally coated GC capillaries for in tube SPME include small sample capacity due to diminutive sub micrometer thickness of GC coatings, as well as their reduced thermal and solvent stability due to a lack of chemical bonds between the coatings and the capill ary wall. To address these issues, Malik and co workers introduced sol gel capillary microextraction (CME) [ 8 ] representing in tube SPME on fused silica capillaries with surface bonded sol gel hybrid organic inorganic coatings. The use of the capillary f ormat and the covalently bonded sol gel coating helped overcome the format related shortcomings of conventional fiber SPME as well as the thermal and solvent stability issues of traditional non sol gel sorbent coatings. In recent years, ionic liquids (IL s) (organic salts that melt at or below 100C) have gained popularity in a number of fields due to their perceived advantages over traditional solvents. They are considered green solvents because they are remarkably less hazardous than their conventiona l counterparts thanks to negligible vapor pressures, low flammability, good thermal stability, tunable viscosities, low corrosion tendencies, and varying degrees of solubility with water and organic solvents . These properties have led to the use of ILs in a variety of areas including green chemistry , organic synthesis and catalysis [ 20 24 ], chemical industry , electrochemistry [ 26 29 ], amino acid and peptide chemistry , carbohydrate chemistry , and in the preparation of
83 microemulsion s [ 32 ]. Several books and extensive reviews hav e been also published on ionic liquids and their applications [33 34 ]. ILs have also found applications in a number of areas in analytical chemistry including GC [ 35, 36 ], LC [ 37 40 ], countercurrent chroma tography , CE [ 42 45 ], analytical spectroscopy , liquid liquid extractions [ 47, 48 ], solid phase extraction , micro solvent cluster extraction , SPME [ 51 53], single drop microextraction [ 54], and supercritical fluid extraction . Exten sive review s have been published in on IL applications in the areas of analytical chemistry [5 6, 57 ]. Recently, ILs have been used in the preparation of sol gel materials [5 8 65 ]. In sol gel applications, IL s have served as solvents [ 58, 60, 64 ], pore tem plates [ 59, 61 ], drying control chemical additives [ 62 ], and possibly as a catalyst [ 64 ]. In several cases, ILs had significant effects on the porous structure of sol gel materials [ 59, 62, 64 ], reduction in cracking and shrinking [ 62, 63 6 6 ] during solve nt evaporation from the sol gel pores, and sol gel reaction kinetics [ 62, 64, 67 6 8]. Ionic liquid mediated sol gels have only seldom been used in analytical separations. Yan and co workers utilized IL mediated sol gel monoliths in CEC [6 3, 65 ] for th e separation of chiral molecules. Racemic mixtures of naproxen [ 63 ] and zolmitriptan [ 65 ] were analyzed using the IL mediated sol gel monoliths. In these cases, 1 butyl 3 methylimmidazolium tetrafluoroborate IL was used to assist in a non hydrolytic sol g el process to prepare molecularly imprinted silica based monoliths. The IL might have helped mitigate the sol gel shrinking problem and acted as a template for pores [ 63 ].
84 In analytical microextraction, IL mediated sol gel hybrid organic inorganic sorbe nts look promising since they are likely to possess favorable material characteristics such as porous morphology, enhanced surface area, improved stability, and thus better extraction efficiency and superior sample preconcentration effects compared to curr ently available extracting phases However, this possibility has not yet been explored. I n this work, we investigated IL mediated polydimethylsiloxane ( PDMS ) poly(ethylene glycol) (PEG) poly(tetrahydrofuran) (PolyTHF) and Bis[(3 methyldimethoxysilyl)p ropyl] polypropylene oxide ( BMPO ) sol gel coatings for capillary microextraction. The effects of two ionic liquids, trihexyltetradecylphosphonium tetrafluoroborate (TTPT) and N butyl 4 methylpyridinium tetrafluoroborate (BMPT) were studied on the physica l characteristics and CME performance of these sol gel coatings. To the best of our knowledge, ionic liquid mediated sol gel sorbents have not been previously utilized in analytical microextraction. 3.2 Experimental 3.2.1 Equipment. A Micromax Thermo I EC OM3590 microcentrifuge (Needham Heights, MA USA ) was used for centrifugation of sol solutions. A Fisher model G 560 Vortex Genie 2 (Fisher Scientific, Pittsburgh, PA USA ) was used to mix sol solution ingredients. A Barnstead model 04741 Nanopure dei onized water system (Barnstead International, Dubuque, IA USA ) was used to obtain 15.5 M ? water. An oxy acetylene torch (Smith Equipment, Watertown, SD USA ) was used to fla me seal fused silica microextraction capillaries. Sol gel CME GC experiments were carried out on a Shimadzu (Kyoto, Japan) model GC 17 capillary gas chromatograph equipped with a flame ionization
85 detector (FID) system An in house designed liquid sample dispenser [ 8 ] was used to facilitate gravity fed flow of aqueous samples through t he sol gel microextraction c apillary. A gas pressure operated filling/purging device [ 69 ] was used to fill fused silica capillaries with sol solutions, expel the sol solutions from and to pass helium (He) through the capillaries. On line data collectio n and processing were accomplished using ChromPerfect for Windows (version 3.5) computer software (Justice Laboratory Software, Denville, NJ USA ). A Hitachi model S 800 scanning electron microscope (Hitachi, Tokyo, Japan) was used to obtain SEM images of the sol gel coated fused silica capillaries. 3.2.2 Chemicals and Materials. Fused silica capillary (250 m I.D. ) with a protective polyimide external coating was obtained from Polymicro Technologies (Phoenix, AZ USA ). Methylene chloride, methanol, Kimwipes, polypropylene microcentrifuge tubes (2.0 mL), and glass scintillation vials were purchased from Fishe r Scientific. Trifluoroacetic acid (TFA, 99%) was purchased from Acros Organics (Morris Plains, NJ USA ). Pyrene, decanol, dodecanal, phenanthrene, hexanophenone, heptanophenone, methyltrimethoxysilane (MTMS 98%), tetramethoxysilane (TMOS), tetraethoxysila ne (TEOS, 99%), formic acid (96%), PEG MW 600 and poly(methylhydrosiloxane) (PMHS) were purchased from Sigma Aldrich (St. Louis, MO USA ). BMPO was purchased from Gelest (Morrisville, PA USA ). Silanol terminated PDMS was obtained from United Chemical Te chnologies (Bristol, PA USA ). PolyTHF was a gift from BASF (Parsippany, NJ). TTPT and BMPT were purchased from Fluka (Seelze, Germany).
86 3.2.3 Preparation of PDMS, PolyTHF, and BMPO Sol Gel Solutions. Table 3 1 presents the names and chemical structure s of sol solution ingredients used to prepare s ilica based hybrid organic inorganic PDMS, PolyTHF, and BMPO sol gel coatings. Compositional details of the prepared sol solution s are listed in table 3 2 IL mediated sol solutions were prepared in microc entrifuge tubes as follows: PDMS BMPO or PolyTHF was individually weighed into a clean microcentrifuge tube in the amount shown in table 3 2 In all cases, a mixture of 250 m L of methylene chloride and 50 m L of ionic liquid TTPT was added. Further, TEOS (50 m L) and PMHS (10 m L) were added in sequence. In the case of PolyTHF and BMPO sol gels no PMHS was added. This was followed by the addition of 50 m L TFA 99%. After the addition of each chemical ingredient, the solution was vortexed for 1 min to ensur e th o rough mixing. The sol solution was further centrifuged for 4 minutes at 14000 rpm (18297g). The supernatant was decanted into a clean microcentrifuge tube. Sol gels PDMS no IL, PolyTHF no IL, and BMPO no IL were prepared in a similar manner excep t that 300 m L of CH 2 Cl 2 was used as solvent instead of a mixture of CH 2 Cl 2 (250 m L) and the IL TTPT (50 m L). An ionic liquid mediated PEG sol gel (PEG IL) and a PEG sol gel that did not contain IL (PEG no IL) were prepared in analogous manner using the ionic liqui d BMPT ( 147.8 m L ) methanol (40 m L), w ater (25 m L) MTMS (100 m L), TMOS (50 m L) and formic acid (61.8 m L).
87 Table 3 1. Names, functions, and chemical structures of some sol solution ingredients used to prepare ionic liquid mediated sol gel CME coatings Ingredient Function Chemical Structure Hydroxy terminated Poly(dimethylsiloxane) (PDMS) Sol gel active organic component H O S i O ) n C H 3 C H 3 C H 3 C H 3 O ( S i S i C H 3 C H 3 O H Poly (tetrahydrofuran) 250 (PolyTHF) Sol gel active o rganic Ligand H O [ ( C H 2 ) 4 O ] n H Bis [(3 methyldimethoxy silyl) propyl] Polypropylene Oxide (BMPO) Sol gel active organic component S i O C H 3 O C H 3 H 3 C ( C H 2 ) 3 O C H 2 C H C H 3 O S i O C H 3 O C H 3 ( C H 2 ) 3 C H 3 n Poly(ethylene glycol) MW 600 (PEG) Sol gel active o rganic l igand HO ( CH 2 CH 2 O) n H Trihexy ltetradecyl phosphonium Tetrafluoroborate (TTPT) Co s olvent H 3 C ( C H 2 ) 5 ( C H 2 ) 5 C H 3 ( C H 2 ) 5 C H 3 ( C H 2 ) 1 3 C H 3 B F 4 P + 4 Methyl N butylpyridinium tetrafluoroborate (BMPT) Co s olvent N + B F F F F Methylene chloride Co s olvent CH 2 Cl 2
88 Tetraethyl orthosilicate ( TEOS ) Sol gel precursor Table 3 1 continued S i O C H 2 C H 3 O C H 2 C H 3 C H 3 C H 2 O O C H 2 C H 3 Poly(methylhydro siloxane) ( PMHS ) Deactivating a gent H 3 C S i O ) n C H 3 C H 3 H C H 3 O ( S i S i C H 3 C H 3 C H 3 Trifluoroacetic Acid ( TFA ) 99% Catalyst CF 3 COOH Table 3 2 Compositions of sol gel with TTPT ionic liquid: (PDMS IL, PolyTHF IL, or BMPO IL) and without the ionic liquid (PDMS no IL, PolyTHF no IL, or BMPO no IL) used to prepare microextraction capillaries. Ingredient Sol gel PDMS Coating with IL no IL Sol gel PolyTHF Coating with IL no IL Sol gel BMPO Coating with IL no IL PDMS(g) 0.0505 0.0510 0 0 0 0 BMPO (g) 0 0 0 0 0.0507 0.0502 PolyTHF (g) 0 0 0. 0 25 0. 0 25 0 0 TTPT ( m L) 50 0 50 0 50 0 CH 2 Cl 2 ( m L) 250 300 250 300 250 300 TEOS ( m L) 50 50 50 50 50 50 PMHS ( m L) 10 10 0 0 0 0 TFA 99% ( m L) 50 50 50 50 50 50
89 3.2.4 Preparation of Ionic Liquid Mediated Sol Gel Microextraction Capillaries. The supernatant of the centrif uged sol solution was immediately utilized to coat the capillaries. For each sol gel composition, a hydrothermally treated [ 70 ] fused silica capillar y (50 cm 250 m m I D ) w as installed on a home built filling/purging device [ 69]. The capillary w as fill ed with the sol solution under 20 psi (1.38 10 5 Pa) of He pressure. After several drops of the coating sol solution dripped out of the capillary, its exit end was sealed with a rubber septum. The solution was allowed to reside inside the capillary for 20 min to facilitate the formation of a surface bonded sol gel coating. After this in capillary residence period, the rubber septum was removed from the capillary end and the un bonded bulk sol solution was expelled from the capillary under helium pressur e For the TTPT ionic liquid mediated sol gel coatings, the capillaries were purged under 20 psi (1.38 10 5 Pa) helium pressure for 60 min prior to thermal conditioning which was somewhat different for coatings with different organic ligands The sol ge l PDMS coated capillaries were thermally conditioned in a GC oven under He purge (1 mL/min) from 40C to 300C at 1C/min and was held at 300C for 300 min. The PolyTHF and BMPO sol gel coated capillaries were conditioned to a final temperature of 250C a nd 280C, respectively, using the same temperature program. The conditioned capillaries were rinsed with 2 mL of 1:1 v/v CH 2 Cl 2 /CH 3 OH mixture to remove any residual IL or its decomposition products. The capillaries were further dried under helium purge in a GC oven by raising the temperature from 40C to 300C (for PDMS), 250C (for PolyTHF), or 280C (for BMPO) at a rate of 10C/min and holding at the final temperature for 30 minutes.
90 For the BMPT ionic liquid mediated sol gel PEG capillaries were flame sealed with an oxy acetylene torch and then thermally conditioned in a GC oven from 40C to 110C at 5C/min holding at 110C for 100 min. Following this, the ends of the capillaries were cut open with an alumina wafer, and rinsed with 2 mL a mixture of CH 2 Cl 2 / CH 3 OH (1:1 v/v) using 5 psi (3.45 10 4 Pa) pressure in the filling/purging device to remove any remaining ionic liquid. The capillaries were further conditioned under He purge (1mL/min) in a GC oven by programming the temperature from 40C to 250 C at 1C/min, and was held at 250C for 120 min. The sol gel coated capillaries prepared without ionic liquid (TTPT or BMPT) were thermally conditioned analogous to their IL mediated counterparts for comparative purposes. The conditioned sol gel capilla ries were then cut into 11 cm long pieces that were further used for capillary microextraction. 3.2.5 Sol gel CME GC Analysis. For CME GC analysis, aqueous samples were prepared using compounds form various chemical classes (aliphatic alcohols, aliphatic aldehydes, aromatic ketones, and polycyclic aromatic hydrocarbons (PAHs)). For each analyte, a stock solution (10 mg/mL) was prepared in methanol and was stored in a surface deactivated 6 mL glass scintillation vial. Fresh aqueous samples were prepared prior to extraction by further diluting these stock solutions with DI water to ng/mL levels. CME was performed as previously described [ 8 ]. Briefly, an 11 cm long sol gel coated microextraction capillary was vertically connected to the bottom of the empt y gravity fed sample dispenser [ 8 ]. Liquid sample (15 mL) was then loaded into the dispenser from the top and allowed to flow through the capillary under gravity for 45 minutes. The capillary was then
91 disconnected from the dispenser and any remaining solu tion was removed from the capillary by touching the end of the capillary with Kimwipes tissue paper. The microextraction capillary was then installed in the GC injector such that approximately 9 cm of the sol gel capillary remained inside the GC injection port previously cooled down to 40C, and approximately 2 cm of it protruded into the GC oven. This was accomplished by providing a gas tight connection of the capillary with the lower end of the injection port with the help of a nut and a graphite ferr ule. The lower free end of the microextraction capillary, located inside the GC oven, was connected to one end of a two way press fit fused silica connector. Further, a Restek Crossbond 14% cyanopropylphenyl 86% PDMS 15 m 0.25 mm I D GC column inlet w as attached to the other end of the connector. The e xtracted analytes, residing in the sol gel coating of the microextraction capillary, were then thermally desorbed from the capillary by rapidly raising the temperature (60C/min) of the injection port fr om 40C to 300C for the sol gel PDMS, from 40C to 250C for the sol gel PEG and PolyTHF, and from 40C to 280C for the sol gel BMPO coated microextraction capillaries. Desorption of the analytes was performed in the splitless injection mode, keeping th e split closed for the entire CME GC analysis. The desorbed analytes were swept onto the GC column by the carrier gas flow and were focused at the inlet of the GC column maintained at 35C. Following this, the GC oven temperature was programmed from 35C (1 min) to 270C at a rate of 20C/min to achieve separation of the extracted analytes transferred to the GC column. The column was held at a final temperature of 250C when sol gel PEG or PolyTHF microextraction capillaries were used. Analytes were dete cted using FID at 350C.
92 3.3 Results and discussion 3.3.1. Sol Gel Immobilization of the CME Coatings. The chemical ingredients used to prepare the coating solutions for the capillaries are listed in table 3 1. The main reactions that take place in the sol solution include hydrolysis of the sol gel precursor(s) and polycondensation of sol gel active species [ 71 ]. These reactions occur simultaneously and are affected by various experimental factors such as water content, type of catalyst used, precursor identity, nature of solvent(s) and other additives (e.g. organic molecules), etc. [ 72 ]. In this case, TFA was used as the catalyst for sol gel PDMS, PolyTHF, or BMPO coatings, while formic acid was used as a catalyst for the sol gel PEG coatings. No extr a water was added to the systems that utilized TTPT as (1) it created a phase separation in PDMS IL and (2) sol gel systems that contained BMPO gelled instantly in its presence. It is reasonable to assume that small amounts of water present in the TFA and the methylene chloride have contributed to the initiation of sol gel precursor hydrolysis Water, generated from the condensation of those hydrolyzed products, could have further facilitated the hydrolysis reaction. Fragments of the sol gel networks evolv ing in the vicinity of the fused silica capillary inner walls had the opportunity to become covalently bonded to it via condensation reactions with silanol groups on the fused silica capillary inner surface ( Figure 3 1A B). 3.3.2 Ionic Liquid Mediated Sol Gel PDMS Microextraction Capillaries. 126.96.36.199 Morphology of Sol gel PDMS Coated Microextraction Capillaries Scanning electron microscopy (SEM) was used to investigate the morphology of the PDMS IL and PDMS no IL sol gel coatings (Figure 3 2 A and 3 2 B ). The IL mediated sol gel microextraction capillaries were rinsed with a 2 mL mixture of
93 (A) (B) Figure 3 1. Polycondensation of 3D sol gel n et work to fused s ilica capillary wall : (A) PDMS and (B) BMPO. methylene chloride and methanol p rior to aquiring SEM images. Rinsing cleaned the coated surface of any debris left from thermal conditioning. Since the analytical data was collected after rinsing, it is obvious that the all of the sol gel coatings survived rinsing and were solvent stab le. As the SEM images show the ionic liquid mediated sol gel
94 coating (figure 3 2 A ) appears to have a more porous texture than the PDMS no IL coat ing (figure 3 2B ). This is an indication that the IL TTPT affected the structure of the overall sol gel. 3. 3.2.2 Role of Ionic Liquids in the Sol Gel System ILs were used as co solvents and as porogens in the sol gel system. Advantages of using ILs as solvents for reactions include their ability to be recycled, high thermal stability, and the improved stabili ty of reactants in ILs [ 61 ]. Advantages of using ILs as porogens instead of organic molecules in sol gel systems include the affect that the cation and the anion portions of the IL have on pore structure and distribution [ 59 62, 7 3 7 4 ], and the ability of ILs to decompose from sol gel systems without leaving residues behind [6 2 ]. Our investigation revealed that i n the case of the PDMS sol gels, the addition of phosphonium based IL, TTPT slowed down the gelation by about 1.5 hours in comparison with the sol gel that did not contain the IL These results are in agreement with those of Karout and Pierre [6 4 ] who also observed an increase in gelation time due to the increase in relative amounts of pyridinium based and imidazolium based ILs in sol gel system s. The slower gelation in the ionic liquid mediated sol gels can be attributed to the increased viscosity of the sol solution due to the addition of the IL. The kinematic viscosity of TTPT is 1117.80 mm 2 s 1 [ 7 5 ] compared to that of methylene chloride whic h is 0.3298 mm 2 s 1 [ 7 6 ]. Further, it is reasonable to assume that the IL did not play a role in extractions since the thermal decomposition temperature of TTPT is 190C [ 7 5 ], and the ionic liquid mediated sol gel PDMS microextraction capillaries were heat ed in an inert environment to 300C. Therefore, it can be safely assumed that during thermal conditioning, the IL had decomposed and the decomposition products
95 (A) (B) Figure 3 2 Scanning electron microscopic images of cross sections of 250 m m I D ( A ) sol gel PDMS IL (22000) and ( B ) sol gel PDMS no IL (20000) coated microextraction capillaries.
96 were at least partially removed from the capillary with the purging helium flow. Any remaining products of the decomposition were further removed from th e capillary during the rinsing step. 3 3.2.3 CME GC analysis using sol gel PDMS coated microextraction capillaries The preconcentration abilities of the two types of sol gel PDMS capillaries (PDMS IL and PDMS no IL) were compared to determine the effect of the IL on the resulting sol gel sorbent Extraction of an aqueous sample containing 125 ppb dodecanal, 100 ppb heptanophenone, and 50 ppb pyrene was performed on the two types of sol gel capillaries F igure s 3 3 shows two chromatograms representing e xtraction results obtained on sol gel PDMS IL and sol gel PDMS no IL capillaries, respectively. These chromatograms show that the sol gel PDMS IL coating provide d significantly higher extraction utility than the PDMS no IL. This, in turn, translates into lower detection limits for the sol gel PDMS IL microextraction capillary (Table 3 3). Run to run and capillary to capillary repeatability data was collected for each analyte on the two types of capillaries in individual CME GC experiments (Table 3 3). In all repeatability experiments, 500 ng/mL dodecanal, 200 ng/mL heptanophenone, and 50 ng/mL pyrene aqueous samples were individually extracted using the two types of capillaries. Run to run GC peak area repeatability data was collected by extracting the s ample analytes in individual experiments on each type of capillary using three replicate measurements. For all three analytes, the sol gel PDMS IL coated capillary provided consistent run to run RSD values between 4.0% and 5.0%. On the other hand, quite sc attered RSD values (2.76% for pyrene, 6.45% for heptanophenone, and 14.1% for dodecanal) were obtained for the same analytes using the sol gel PDMS no IL coated capillary. This coating also
97 Figure 3 3 Comparison of CME GC analysi s of 125 ppb dodecan al, 100 ppb h eptanophenone, and 50 ppb pyrene on sol gel PDMS no IL ( bottom ) and sol gel PDMS IL ( top ) microextraction capillaries. Extraction conditions: 11 cm 0.25 mm I D microextraction capillary; extraction time, 45 min (gravity fed at room tempera ture). Other conditions: 15 m 0.25 mm I D Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 300C; programmed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium car rier gas: FID 350C. Peaks: (1) dodecanal, (2) heptanophenone, and (3) pyrene for both chromatograms.
98 Table 3 3. Peak area repeatability and limit of detection data for dodecanal (200 ppb sample), heptanophenone (100 ppb sample), and pyrene (50 ppb sample) extracted from aqueous samples using three replicate measurements by CME GC using sol gel immobilized PDMS microextraction capillaries prepared with (A) and without (B) ionic liquid. Run to Run Repeatability ( n = 3) Detection Limits ( S/N =3) Sol Gel A Sol Gel B Sol Gel A Sol Gel B Analyte Mean Peak Area* RSD ( % ) Mean Peak Area* RSD ( % ) (ng/L) (ng/L) Dodecanal 129.0 5.0 4.6 14.1 17.4 487.0 Heptanophenone 265.3 4.2 17.2 6.4 3.9 52.3 Pyrene 69.6 4.5 45.5 2.8 3.2 4.9 Arbitrary un it provided worse limits of detection for all three analytes. Both sol gel PDMS IL and sol gel PDMS no IL coated capillaries pr ovided significantly higher detec tion limits for dodecanal compared to pyrene and heptanophenone. It appears that the affinit y of both sol gel coatings for the aldehyde was unusually poor. Currently, this aspect of uncanny extraction behavior is being investigated. Capillary to capillary reproducibility data was obtained by extracting each sample in triplicate onto six individu ally prepared capillaries: three PDMS IL and three with PDMS no IL capillaries. This data characterized the reproducibility of the sol gel coating method The coating s prepared with ionic liquid (PDMS IL) provided excellent capillary to capillary repeata bility Peak area RSD value s of 2.26% 0.15% and 4.07% were obtained for dodecanal, heptanophenone, and pyrene respectively The sol gel coating prepared without ionic liquid (PDMS no IL) provided RSD values of 7.79% for heptanophenone and 9.72% for pyr ene. Repeated extractions were made on the
99 capillar ies without deterioration of performance over a period of 9 months. Furthermore, the coating is solvent stable and can be rinsed to ensure consistent performance. In CME GC experiments, both types of so l gel PDMS capillaries provided limits of detection (using a signal to noise ratio of 3) in the ng/L range. However, the sol gel PDMS IL microextraction capillary provided better detection limits (3.2 17.4 ng/L) than the sol gel PDMS no IL capillary (4 .9 487.0 ng/L). This is likely because the sol gel PDMS IL capillary had a more porous morphology (F igure 3 2 ), and thereby provided a higher surface area for sorption. These results clearly suggest that the ionic liquid had a positive effect on the e xtraction ability of the sol gel PDMS coating. Furthermore, PDMS sol gel coatings are known to be thermally stable beyond 350C [ 7 7 ] Since CME is a non exhaustive equilibrium extraction technique , it is important to determine the time required for the analyte to reach a sorption/desorption equilibrium between the sol gel coating and the sample. In this work, we compared the extraction profiles of the PDMS IL and PDMS no IL capillaries using heptanophenone (Figure 3 4 A ) and phenanthrene (Figure 3 4 B ) as test solutes. The extraction profiles (Figure 3 4 ) indicate that the PDMS IL coating had a higher capacity, but the equilibrium is reached at a slower pace than on sol gel PDMS no IL coating The extraction time required for the curve to reach the pla teau indicates the onset of extraction equilibrium. In the case of PDMS no IL extraction time was reached in 10 15 min for both analytes, but in the case of PDMS IL the equilibrium was attained at about 60 min for both analytes. Thus, fewer experiments w ere required for PDMS no IL since the equilibrium time was reached quickly This extraction behavior on the ionic liquid mediated coating
100 (A) (B) Figure 3 4 Extraction profiles of heptanophenone (A ) and phenanthrene ( B ) extracted on 11cm 0.25 mm I D PDMS IL and PDMS no IL sol gel coated microextraction capillaries from an aqueous sample. Extraction conditions: triplicate extraction at various time intervals; microextraction capillaries were rinsed with 1:1 v/v CH 2 Cl 2 : methanol and dried at 30 0C before each extraction. GC analysis conditions: 15 m 0.25 mm I D Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 300C; programmed temperature GC run from 35C (1 min) to 270C at a r ate of 20C/min; helium carrier gas: FID 350C.
101 can be explained by the slow diffusion of analytes in the liquid filling the porous sol gel structure. 3.3.3 Ionic Liquid Mediated Polar Sol Gel Microextraction Capillaries: PEG, PolyTHF, and BMPO. Attemp ts were made to prepare ionic liquid mediated polar sol gel coatings based on PEG and PolyTHF T he IL, BMPT was used in conjunction with the sol gel PEG coating while the IL, TTPT used to prepare the sol gel polyTHF coating. Again, t wo types of sol gel c oatings were prepared with each polymer: (a) PEG IL and PolyTHF IL and (b) PEG no IL and PolyTHF no IL. SEM was used to investigate the morphology of the sol gel coated capillaries. Cross sections of the capillaries showed that PEG IL s ol gel coating (Fi gure 3 5A ) appeared more porous than its co unterpart prepared without IL (Figure 3 5B ) The same trend was observed for the sol gel PolyTHF coatings: sol gel PolyTHF IL (Figure 3 5 C ) coating seemed more porous than sol gel PolyTHF no IL (Figure 3 5 D ) coa ting. Reproducible coating thickness could be obtained with IL mediated sol gels. For example, three individually prepared PEG IL sol gel coated capillaries had an average thickness of 5.8 microns with a standard deviation of 0.3 microns (an RSD value of 5.2%). S ol gel coating s with greater porous morphology obtained with the help of ILs can be expected to provide better performance in extraction However, extractions using the ionic liquid mediated PEG and the PolyTHF sol gel porous coatings showed that this was not the case. The PEG no IL (Figure 3 6 top) and the PolyTHF no IL (Figure 3 7 top) coated microextraction capillaries provided better performance in CME GC A peak area for 100 ppb decanol extracted on the sol gel PEG IL coating was 72329 arbi trary
102 (A) (B)
103 (C) (D) Figure 3 5. Scanning electron microscopic images of cross sections of 250 m m I D ( A ) sol gel PEG IL (12000), ( B ) sol gel PEG no IL (15000), ( C ) sol gel PolyTHF IL (500), ( D ) sol gel PolyTHF no IL (350) coated microext raction capillaries.
104 Figure 3 6. Comparison of CME GC analysis of 100 ppb decanol sol gel PEG IL ( bottom ) and sol gel PEG no IL ( top ) microextraction capillaries. Extraction conditions: Same as Figure 3 3 Other conditions: 15 m 0.25 mm I D R estek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 250C; programmed temperature GC run from 35C (1 min) to 250C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) decanol for b oth chromatograms.
105 Figure 3 7. Comparison of CME GC analysis of 500 ppb decanol, 500 ppb hexanophenone, 200 ppb phenanthreme sol gel PolyTHF IL ( bottom ) and sol gel PolyTHF no IL ( top ) microextraction capillaries. Extraction conditions: Same as F i gure 3 3. Other conditions: Same as Figure 3 6 Peaks: (1) decanol, (2) hexanophenone, and (3) phenanthrene for both chromatograms.
106 units, and the peak area provided by the sol gel PEG no IL coating was 266681 arbitrary units. As figure 3 7 shows, three analytes decanol, hexanophenone, and phenanthrene were poorly extracted on the PolyTHF IL microextract ion capillary (bottom) compared to the PolyTHF no IL capillary (top). It appears that while the sol gel PEG IL and PolyTHF IL coating s were more p orous, they might have consist ed mainly of silica with only very small amounts of polymer incorporated into the so l gel network resulting in inferior extraction performance Even though, the PEG and the Poly THF sol gel coated capillaries were prepared u tilizing different ILs and different thermal conditioning methods, they both demonstrated that a C OH terminated polymer does not create effective sol gel sorbents when mediated by an IL. An important factor in this phenomenon is the lower reactivity of t he terminal hydroxyl group s on PEG and PolyTHF compared to silanol groups on hydroxy terminated PDMS and alkoxy groups on silica based sol gel precursors [7 8 ]. Because of higher reactivity of Si OH and Si OR groups compared with C OH groups polycondensa tion reactions are likely to predominantly take place between chemical species containing the sol gel active Si OH (silanol) and/or alkoxy silane groups. Condensation reactions between a Si OH or a Si OR group and the terminal C OH (hydroxyl group) of the polymer s can be expected to occur less effectively. Apparently, condensation of C OH terminated polymers is slowed down even further or is hindered when utilizing an ionic liquid in the sol gel system. W hile the IL s help develop porous morphology in coa tings, they appear not to produce sol gel coatings that ar e effective at microextraction due to quantitative deficiency of bonded organic polymer ligands. As was the case with the PDMS based sol gel sorbents, it can be assumed that
107 the ILs played no role in extractions. As mentioned, this is because TTPT decomposes at 190C [7 5 ], and the capillaries prepared with this IL were conditioned at temperatures higher than 250C, and they were rinsed with organic solvents before use. Although the decomposition t emperature of BMPT is 295C [7 5 ], and the PEG capillaries prepared with it were conditioned at a lower temperature, this IL was removed from the capillary by rinsing with copious amounts of methylene chloride and methanol mixture. In order to determine if in fact PEG and PolyTHF were not being incorporated into the sol gel network in the presence of an IL due to their low reactivity, we investigated a sol gel system that contained (instead of PEG or PolyTHF ) bis[(3 methyldimethoxysilyl)propyl] polypropylene oxide (BMPO) a polymer with sol gel active methoxysilane termination and a flexible propylene oxide repeating unit ( Table 3 1 ). To our knowledge, BMPO has not been utilized in the preparation of microextraction coatings. It has, however, been used previ ously to synthesize hybrid inorganic organic polymer membranes [ 7 9 ]. In t he present work, we investigated two types of sol gel BMPO coatings: (a) coatings prepared with the use of ionic liquid (TTPT) (BMPO IL) and (b) coatings prepared without the us e of T TPT (BMPO no IL) (Table 3 2 ). The ionic liquid slowed the rate of gelation in the case of sol gel BMPO system just like it did in the sol gel PDMS system. BMPO IL sol solution gelled in more than 24 hours, and BMPO no IL gelled in 16 hours. U nlike the I L mediated PEG and PolyTHF sol gel coated capillaries, the IL mediated BMPO sol gel (Figure 3 8 top ) coated capillaries could extract analytes decanol, hexanophenone, and phenanthrene with significantly higher extraction efficiency than its non IL counte rpart (Figure 3 8 bottom ) Compared to the BMPO no IL
108 Figure 3 8. Comparison of CME GC analysis of 500 ppb decanol, 500 ppb hexanophenone, and 200 ppb phenanthrene on sol gel BMPO no IL ( bottom ) and sol gel BMPO IL ( top ) microextraction capillaries. Extraction conditions: Same as Figure 3 3 Other conditions: 15 m 0.25 mm I D Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 280C; programmed temperature GC run from 35C (1 min) to 27 0C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) decanol, (2) hexanophenone, and (3) phenanthrene for both chromatograms.
109 sol gel coating, the BMPO IL sol gel coating provided 3.6, 3.5, and 8.1 times more efficient extractions for d ecanol, hexanophenone, phenanthrene, respectively. Since BMPO is a sol gel active polymer that acquires Si OH termination after hydrolysis it gets effectively bonded into the sol gel network ( figure 3 1B) even in the presence of IL because of higher reac tivity terminal silanol groups Limits of detection (using a signal to noise ratio of 3) for the investigated analytes in CME GC FID analysis were determined to be in the ng/L range for the BMPO IL and BMPO no IL capillaries. The BMPO IL coating provide d a limit of detection of 53.0 ng/L for decanol, 41.0 ng/L for hexanophenone, and 3.5 ng/L for phenanthrene. The BMPO no IL capillary provided a limit of detection of 186.0 ng/L for decano l, 137.0 ng/L for hexanophenone, a nd 27.0 ng/mL for phenanthrene. Thus, the sol gel BMPO IL capillaries provided 3 5 times better detectoion limits than BMPO no IL capillaries. This work demonstrate d the possibility of using ILs in the preparation of both non polar (PDMS) and polar (BMPO) sol gel coating for CME In bo th cases, the IL mediated sol gel coatings had significantly better extraction performance than analogous coatings prepared without ionic liquids Thus, IL mediated sol gel coatings open new possibilities for effective preconcentration of analytes since bo th polar and non polar sol gel coatings can be prepared with ionic liquid mediation. However, when preparing non polar or polar hybrid organic inorganic sol gel sorbents using ILs it is vital to chose organic p olymers and sol gel precursors with similar s ol gel reactivity to ensure that the organic polymers are effectively incorporated into the sol gel network providing an efficient sol gel sorbent
110 3.4. Conclusion. Fo r the first time, IL mediated sol gel PDMS and BMPO coatings were developed for use as immobilized sorbents in capillary microextraction. Ionic liquid mediated sol gel PDMS coatings provided consistent performance in CME GC analysis ( run to run peak area RSD values of 4.2 to 5.0 %) compared with sol gel PDMS coatings prepared without ioni c liquid (2.8 to 14.1%) PDMS and BMPO IL mediated sol gel coatings also provided lower d etection limits (Table 3 3 ) compared to analogous sol gel coatings prepared without IL. S canning electron microscopy results suggest that ILs can provide a porous m orphology of sol gel extraction media when it is incorporated in the sol gel coating solution. E nhancement of porosity alone was not enough to provide effective extraction of analytes. Thus, careful choice of the polymer and precursor with comparable sol gel reactivity must be made when designing an IL mediated sol gel sorbent in order to ensure that the created sol gel coating inherently possess es the desired sorbent characteristics. IL mediated sol gel materials hold great potential for being widely us ed as sorbents and stationary phases in separation science. 3.5 A cknowledgment. This research was partially supported by the National Science Foundation under grant no. DGE 0221681. This research was performed while the first author was a Fellow under th e Department of Homeland Security (DHS) Scholarship and Fellowship Program, administered by the Oak Ridge Institute for Science and Education through an interagency agreement with the US Department of Energy under contract number DE AC05 00OR22750. All opi nions expressed in this paper are the authors.
111 3.6 References for Chapter Three.  J. Pawliszyn, S. Liu, Anal. Chem. 59 (1987) 1475.  R. Belardi, J. Pawliszyn, Water Pollut. Res. J. Can. 24 (1989) 179.  C. L. Arthur, J. Pawliszyn, Anal. Chem. 62 (1990) 2145.  R. Eisert, J. Pawliszyn, Anal. Chem. 69 ( 1997) 3140.  M. E. McComb, R. D. Oleschuk, E. Giller, H. D. Gesser, Talanta 44 (1997) 2137.  H. Hartmann, J. Burhenne, M. Spiteller, Fresenius Environ. Bull. 7 (1998) 96.  H. Kataoka, J. Pawliszyn, Chromatogr aphia 50 ( 1999) 532.  S. Bigham, J. Medlar, A. Kabir, C. Shende, A. Alli, A. Malik, Anal. Chem. 74 (2002) 752.  C. L. Arthur, L. M. Killam, K. D. Buchholz, J. Pawliszyn, J. R. Berg, Anal. Chem. 64 (1992) 1960.  J. Chen, J. B. Pawliszyn, Anal. Chem. 67 (199 5) 2530.  Y. Hirata, J. Pawliszyn, J. Microcol. Sep. 6 (1994) 443.  D. Figeys, A. Ducrte, J. R. Yates, R. Aebersold, Nat. Biotechnol. 14 (1996) 1579.  C. W. Whang, J. Pawliszyn, Anal. Commun. 35 (1998) 353.  Z. Y. Zhang, J. Pawliszyn, Anal. Chem. 65 (1993) 1843.  L. Moens, T. De Smaele, R. Dams, P. Van Den Broeck, P. Sandra, Anal. Chem. 69 (1997) 1604.  J Pawliszyn, Sampling and Sample Preparation for Field and Laboratory Elsevier New York 2002.  H. Lord, J. Pawliszyn, J. Chromatogr. A 885 (2000) 153.  S. A. Forsyth, J. M. Pringle, D. R. MacFarlane, Aust. J. Chem. 57 (2004) 113.  N. V. Plechkova, K. R. Seddon, i n : P. Tundo, A. Perosa, F. Zecchini ( E ds.) Methods a nd Reagents for Green Chemistry, Wiley, Hoboken, NJ, 2007 pp. 105 130.
112  T. Welton, Chem. Re v. 99 (1999) 2071.  R. Sheldon, Chem. Comm. 23 ( 2001) 2399.  C. M. Gordon, Appl. Catal. A 222 ( 2001) 101.  J. Dupont, R. F. de Souza, P. A. Z. Suarez, Chem. Rev. 102 (2002) 3667.  D. B. Zhao, M. Wu, Y. Kou, E. Min, Catal. Today 74 (2002) 157.  N. V. Plec hkova, K. R. Seddon, Chem. Soc. Rev. 37 (2008) 123.  D. R. Macfarlane, M. Forsyth, P. C. Howlett, J. M. Pringle, J. Sun, G. Annat, W. Neil, E. I. Izgorodina, Acc. Chem. Res. 40 (2007) 1165.  R. Hagiwara, J. S. Lee, Electrochemistry 75 (2007) 23.  D. Wei, A Ivaska, Anal. Chim. Acta 607 (2008) 126.  P. Hapiot, C. Lagrost, Chem. Rev. 108 (2008) 2238.  J. C. Plaquevent, J. Levillain, F. Guillen, C. Malhaic, A. C. Gaumont, Chem. Rev. 18 (2008) 5035.  O. A. El Seoud, A. Koschella, L. C. Fidale, S. Dorn, T. Heinz e, Biomacromolecules 8 (2007) 2629.  Z. M. Qiu, J. Texter, Curr. Opin. Colloid Interface Sci. 13 (2008) 252.  S. Chowdhury, R. S. Mohan, J. L. Scott, Tetrahedron 63 (2007) 2363  H. Weingartner, Angew. Chem. Int. Ed. 47 (2008) 654.  D. W. Armstrong, L. F. He, Y. S. Liu, Anal. Chem. 71 (1999) 3873.  J. L. Anderson i n : M. Koel ( Ed. ), Ion ic Liquids in Chemical Analysis, CRC Press, Boca Raton, FL 2009 pp. 139 165.  L. J. He, W. Z. Zhang, L. Zhao, X. Liu, S. X. Jiang, J. Chromatogr. A 1007 (2003) 39.  R. K aliszan, M. P. Marszall, M. J. Markuszewski, T. Baczek, J. Pernak, J. Chromatogr. A 1030 (2004) 263.
113  M. P. Marszall, R. Kaliszan, Crit. Rev. Anal. Chem. 37 (2007) 127.  A M. Stalcup, i n : M. Koel ( Ed .), Ionic Liquids in Chemical Analysis CRC Press, Boca Raton, FL 2009 pp. 168 183.  A. Berthod, S. Carda Broch, Anal. Bioanal. Chem. 380 (2004) 168.  G. Yanes, S. R. Gratz, M. J. Baldwin, S. E. Robinson, A. M. Stalcup, Anal. Chem. 73 (2001) 3838.  M. Vaher, M. Koel, M. Kaljurand, Electrophoresis 23 (2002) 42 6.  W. D. Qin, S. F. Y. Li, Analyst 128 (2003) 37.  M. Lopez Pastor, B. M. Simonet, B. Lendl, M. Valcarcel, Electrophoresis 29 (2008) 94.  C. D. Tran, Anal. Lett. 40 (2007) 2447.  J. G. Huddleston, H. D. Willauer, R. P. Swatloski, A. E. Visser, R. D. Rogers, Chem. Comm. 16 (1998) 1765.  M. Gharehbaghi, F. Shemirani, M. Baghdadi, Int. J. Environ. Anal. Chem. 89 (2009) 21.  G. V. Myasoedova, N. P. Molochnikova, O. B. Mokhodoeva, B. F. Myasoedov, Anal. Sci. 24 (2008) 1351.  T. Charoenraks, M. Tabat a, K. Fuji, Anal. Sci. (2008) 1239.  J. F. Liu, N. Li, G. B. Jiang, J. M. Li, J. A. Jonsson, M. J. Wen, J. Chromatogr. A 1066 (2005) 27.  Y. N. Hsieh, P.C. Huang, I. W. Sun, T. J. Whang, C. Y. Hsu, H. H. Huang, C. H. Kuei, Anal. Chim. Acta 557 (2006) 32 1  F. Zhao, Y. Meng, J. L. Anderson, J. Chromatogr. A 1208 (2008) 1.  L. Vidal, A. Chisvert, A. Canals, A. Salvador, J. Chromatogr. A 1174 (2007) 95.  S. Keskin, D. Kayrak Talay, U. Akman, O. Hortacsu, J. Supercrit. Fluids 43 (2007) 150.  S. Pan dey, Anal. Chim. Acta 556 (2006) 38
114  X. Han, D. W. Armstrong, Acc. Chem. Res. 40 (2007) 1079  S. Dai, Y. H. Ju, H. J. Gao, J. S. Lin, S. J. Pennycook, C. E. Barnes, Chem. Commun. 3 (2000) 243.  Y. Zhou, J. H. Schattka, M. Antonietti, Nano Lett. 4 (2004 ) 477.  Y. Liu, M. J. Wang, Z. Y. Li, H. T. Liu, P. He, J. H. Li, Langmuir 21 (2005) 1618.  Y. Liu, M. J. Wang, J. Li, Z. Y. Li, P. He, H. T. Liu, J. H. Li, Chem. Commun. 13 (2005) 1778.  M. A. Klingshirn, S. K. Spear, J. D. Holbrey, R. D. Rogers, J. Mater Chem. 15 ( 2005) 5174.  H. F. Wang, Y. Z. Zhu, X. P. Yan, R. Y. Gao, J. Y. Zheng, Adv. Mater. 18 (2006) 3266.  A. Karout, A. C. Pierre, J. Non Cryst. Solids 353 (2007) 2900.  H. F. Wang, Y. Z. Zhu, J. P. Lin, X. P. Yan, Electrophoresis 29 (2008) 952.  A. Safavi, N. Maleki, M. Bagheri, J. Mater. Chem. 17 (2007) 1674.  K. S. Yoo, T. G. Lee, J. Kim, Microp. Mesopr. Mater. 84 (2005) 211.  H. Choi, Y. J. Kim, R. S. Varma, D. D. Dionysiou, Chem. Mater. 18 (2006) 5377.  J. D. Hayes, A. Malik, J. Chromatogr. B 695 (1997) 3.  J. D. Hayes, Ph.D. Dissertation, University of South Florida, Tampa, FL, (2000).  C. Brinker, G. Scherer, Sol Gel Science: The Physics and Chemistry of Sol Gel Processing Academic Press San Diego 1990.  J. D. Mackenzie, D. R. Ulrich, Ultrast ructure Processing of Advanced Ceramics Wiley New York 1988.  C. J. Adams, A. E. Bradley, K. R. Seddon, Aust. J. Chem. 54 (2001) 679.  J. Zhou, Y. Ma, F. Shi, L. Liu, Y. Deng, Microporous Mesoporous Mater. 119 (2007) 97.
115  Merck, Ionic Liquids: New Mater ials for New Applications, http://ildb.merck.de/ionicliquids/en/startpage.htm 2009  California Environmental Protection Agency: Air Resources Board, http://www.arb.ca.gov/db/solvents/solvent_pages/Halogens HTML/methylene_chloride.htm 2009  D. Wang, S. L. Chong, A. Malik, Anal. Chem. 69 (1997) 4566.  H. H. Huang, G. L. Wilkes, i n : Z. L. Wang, Y. L. Zhang ( Eds .), Handbook of Nanophase and Nanostruct ured Materials, Vol.1 Synthesis, Klu wer Academic/ Plenum Publishers, New York, 2003, pp. 90.  S. W. Li, Z. Zho u, M. L. Liu, W. Li, J. Ukai, K. Hase, M. Nakanishi, Electrochim. Acta 51 (2006) 1351.
116 CHAPTER 4: IONIC LIQUID MEDIATED BIS[(3 METHYLDIMETHOXYSILYL)PROPYL] POLYPROPYLENE OXIDE BASED POLAR SOL GEL COATINGS FOR CAPILLARY MI CROEXTRACTION 4.1 Introduction Hydrophilic p olar analytes are notorio usly difficult to extract and preconcentrate from aqueous matrices. Sample preconcentration is of utmost importance in the trace analysis of these recalcitrant analytes. A variety of e xtraction based preconcentration techniques have been utilized for this purpose . With the current trend of miniaturization in analytical instrumentation, microextraction techniques are gaining popularity. Microextraction techniques include solid phase microextraction (SPME) [2,3], hollow fiber microextraction , single drop microextraction , liquid phase microextraction , extraction techniques based on suspended particle s membrane s /disk s coated v essel walls, etc. , and stir bar sorptive ex traction (SBSE) . In particular, fiber SPME and in tube SPME [9 11 ] capillary microextraction (CME)  have experienced an explosive growth over the past two decades. The main reason behind such growth lies in the fact that they provide green extr action by completely eliminating the use of organic solvents in the extraction process. Moreover, CME uses a sorbent coating located inside a small diameter tubing either in the form of a surface coating or a packed/monolithic sorbent bed. Thus, analyte s are directly extracted onto the sorbent coating/bed from a sample as they pass through the
117 tubing . CME was originally developed to overcome difficulties encountered in fiber SPME GC with weakly volatile or thermally labile analytes  as CME coul d be easily coupled with HPLC. CME also offers some other advantages over fiber SPME. SPME fibers often have limited sample capacities. Higher sample capacities can be obtained with CME because the sorbent coating/bed is contained within a longer segmen t of the tube providing higher sorbent loading Fiber SPME devices also have i ssues with mechanical stability the fiber can break, the coating can be scratched, and the needle can bend . CME devices allow for superior mechanical stability because f lexible capillaries with outer protective coatings are utilized, providing safeguard against mechanical damage to the sorbent or the tubing. Traditio nal sorbents coatings used in early in tube SPME devices consisted of GC stationary phases [9 ]: a piece of GC capillary column with a stationary phase coating on the inner walls was typically used as the extraction device. These sorbent coatings can be problematic for in tube SMPE applications since : (a) they are they are thin (typically sub micrometer thickne ss) and therefore have low sample capacity and (b) they are typically held on the capillary surface by physical forces of adhesion, and therefore, can have issues with solvent and thermal stability. One way to overcome these issues is to use sol gel mater ials as the sorbent coating [ 12,15 ] Sol gel coatings are advantageous because they can be chemically anchored to the inner wall of fused silica tubing. Thus, sol gel sorbents are stable to high temperature and harsh solvents. Furthermore, sol gel sorbe nts can be tailored to the polarity of a specific analyte by using appropriate organic and inorganic components. Polar sol gel sorbents have been developed for in tube SPME including those based on cyano , crown ether , and poly(ethylene glycol) [ 12,18
118 21] materials. While these sol gel coatings have advanced the use of polar organic polymers and achieve higher thermal and solvent stability these coatings mostly contained long chain polymers of high molecular weight s [12,18 20 ] having lower polar ity (compared to their short chain counterparts), and thus, reduce d ability to extract highly polar analytes For such capillaries, sample capacity can still be an issue . One method to enhance sam ple capacity is to create sol gel coatings with poro us morphology using porogenic agent(s) in the sol solution Traditional porogenic agents include organic solvents, polymers, surfactants, micelles, latex spheres, and inorganic salts  Recently, ionic liquids have been used as porogens and templates for sol gel materials [ 23 30]. Advantages of using ionic liquids as opposed to traditional types of porogens include their low toxicity, high thermal stability, ultra low vapor pressures, and their ability to be recycled . Moreover, ILs can have sig nificant effects on the porous structure of sol gel materials [24, 26, 31], reduction in cracking and shrinking [31 33] during solvent evaporation from the sol gel pores, and sol gel reaction kinetics [26,31,32]. In a previous work  we introduced IL mediated sol gel coatings for capillary microextraction of nonpolar and moderately polar analytes. In this work, we describe the use of a polar sol gel active polymeric precursor, BMPO ( which contains a propylene oxide repeating unit and has a molecular w eight of 500 900 mol/g ) to create a surface bonded organic inorganic hybrid sol gel coating within a fused silica capillary to extract polar analytes. W e report the use of two ionic liquids (TTPT and MOIC) as porogens in the sol gel BMPO system and illust rate the analytical capability of IL mediated sol gel BMPO coatings to serve as the extracting
119 phase in capilla ry microextraction of polar, non polar and moderately polar analytes from aqueous matrices. 4.2 Experimental 4.2.1 Equipment Nanopure deio nized water (15.5 M ? ) was acquired using a Barnstead model 04741 Nanopure system (Barnstead International, Dubuque, IA, USA). Sol solution ingredients were mixed using a Fisher model G 560 Vortex Genie 2 (Fisher Scientific, Pittsburgh, PA, USA). Sol solut ions were centrifuged with a Micromax Thermo IEC OM3590 microcentrifuge (Needham Heights, MA, USA). A gas pressure operated filling/purging device [35 ] was used to introduce a sol solution into, and to expel the solutions from fused silica capillaries. A liquid sample dispenser  was used to facilitate gravity fed flow of aqueous samples through the sol gel microextraction capillary. A Shimadzu model GC 17 capillary gas chromatograph equipped with a flame ionization detector (FID) (Shimadzu, Kyoto, J apan) was used for CME GC experiments. ChromPerfect for Windows (version 3.5) computer software (Justice Laboratory Software, Denville, NJ, USA) was used for on line data collection and processing. SEM images were obtained with a Hitachi model S 800 scan ning electron microscope (Hitachi, Tokyo, Japan). 4.2.2 Chemicals and materials The ionic liquids, TTPT and MOIC, were acquired from Fluka (Seelze, Germany). BMPO was obtained from Gelest Inc. (Morrisville, PA). Fused silica capillary (250 m I.D.) with a protective polyimide external coating was bought from Polymicro Technologies (Phoenix, AZ). Trifluoroacetic acid (TFA, 99%) and alcohols
120 (nonanol, decanol, and undecanol) were secured from Acros Organics (Morris Plains, NJ, USA). P henols ( 2,4,6 trichlorophenol, 2 tert butyl 4 methoxyphenol, pentachlorophenol), aliphatic acids (nonanoic acid, decanoic acid, and undecanoic acid), a ldehydes (d ecanal, undecanal, and dodecanal), ketones (hexanophenone, heptanophenone, decanophenone), anilines ( N butylaniline, diphenylamine, and acridine), PAHs (acenaphthene, phenanthrene, and pyrene), tetraethoxysilane (TEOS, 99%), and PDMDPS were acquired from Sigma Aldrich (St. Louis, MO, USA). Methylene chloride, methanol, polypropylene microcentrifuge tubes (2.0 mL), and glass scintillation vials (6 mL) were purchased from Fisher Scientific. 4.2.3 Preparation of sol gel coating solutions Briefly 50 mg of BMPO or PDMDPS was weighed into a clean microcentrifuge tube. A mixture consisting of 250 m L of methylen e chloride and 50 m L of an ionic liquid, TTPT or MOIC, was then added. For comparative purposes, we also prepared s ol solutions that did not contain any IL This was done in a similar manner except that 300 m L of CH 2 Cl 2 was used as the solvent instead of a mixture of methylene chloride (250 m L ) and an IL (50 m L ) Then TEOS (50 m L) followed by 50 m L TFA 99% was added to the respective sol solutions. To obtain a homogenous system, the sol solutions were vortexed for 1 min after the addition of each compone nt. Sol solutions were centrifuged for 4 minutes at 14000 rpm (18297g). The supernatant was collected for further use. Chemical structures of sol solution components are illustrated in table 4 1. Table 4 2 details the compositions of the sol solutions used in this work
121 Table 4 1. Names, functions, and chemical structures of sol gel ingredients. Ingredient Function Chemical Structure Bis [(3 methyldimethoxy silyl) propyl] Polypropylene Oxide (BMPO) Sol gel active organic component S i O C H 3 O C H 3 H 3 C ( C H 2 ) 3 O C H 2 C H C H 3 O S i O C H 3 O C H 3 ( C H 2 ) 3 C H 3 n Poly(dimethylsiloxane co diphenylsiloxane), dihydroxyterminated (PDMDPS) Sol gel active organic component ( S i O ) x ( S i O ) y C H 3 C H 3 H O H Trihexyltetradecyl phosphonium Tetrafluroborate (TTPT) Co solvent H 3 C ( C H 2 ) 5 ( C H 2 ) 5 C H 3 ( C H 2 ) 5 C H 3 ( C H 2 ) 1 3 C H 3 B F 4 P + 1 Methyl 3 octylimidazolium Chloride (MOIC) Co solvent N + N C l ( C H 2 ) 7 C H 3 C H 3 Methylene chloride Co solvent CH 2 Cl 2 Tetraethylortho silicate (TEOS) Sol gel precursor S i O C H 2 C H 3 O C H 2 C H 3 C H 3 C H 2 O O C H 2 C H 3 Trifluoroacetic Acid (TFA) Catalyst CF 3 COOH
122 Table 4 2 Compositions of sol gels with ionic liquid: (BMPO TTPT, BMPO MOIC, and PDMDPS MOIC) and without ionic liquid (BMPO no IL and PDMDSP no IL) without ionic liquid used to prepare microextraction cap illaries. Ingredient Sol gel BMPO TTPT Coating Sol gel BMPO MOIC Coating Sol gel BMPO no IL Coating Sol gel PDMDPS MOIC Coating Sol gel PDMDPS no IL Coating BMPO (mg) 50 50 50 0 0 PDMDPS (mg) 0 0 0 50 50 TTPT ( m L) 50 0 0 0 0 MOIC ( m L) 0 50 0 50 0 CH 2 Cl 2 ( m L) 250 250 300 250 300 TEOS ( m L) 50 50 50 50 50 TFA 99% ( m L) 50 50 50 50 50 4.2.4 Preparation of sol gel coated microextraction capillaries Hydrothermally treated [36 ] fused silica tubing (50 cm 0.25 mm i.d.) was utilized to prepare sol gel coated microextraction capillaries. The fused silica capillary was set up in the filling/purging device . Helium pressure (20 psi = 1.38 10 5 Pa), was used to fill the capillaries with sol solution. The ex it end of the capillary was sealed with a rubber septum after several drops of the coating sol solution trickled out of it. The solution was kept in the capillary for 20 min allowing a surface bonded sol gel coating to form. Then any un bonded sol soluti on was expelled from the capillary by purging with 20 psi (1.38 10 5 Pa) helium pressure for 60 min.
123 The BMPO sol gel capillaries (prepared with or without IL mediation) were then thermally conditioned under helium purge in a GC ove n from 40C to 280C at 1C/min and were held at the final temperature for 300 min. The sol gel PDMDPS capillaries (prepared with or without IL mediation) were thermally conditioned using a temperature programming rate of 1C/min to a final temperature of 300C for 300 min. T he conditioned capillaries were rinsed with 2 mL of 1:1 v/v methylene chloride/methanol mixture and were dried under helium purge in a GC oven by programming the temperature from 40C to 280C (for BMPO) or 300C (for PDMDPS) at 10C/min holding at a fina l temperature for 30 minutes. The finished sol gel coated capillaries were then chopped into 11 cm long pieces; they were further used for CME. 4.2.5 Sol gel CME GC analysis Stock solutions (10 mg/mL) of test analytes from various chemical classes (phen ols, acids, amines, alcohol s, aldehydes, ketones, and PAHs ) were prepared in methanol and were stored in glass scintillation vials. A queous test samples were prepared by diluting the s tock solutions to ng/mL levels with Nanopure water CME experiments wer e conducted as earlier detailed . Briefly, a n 11 cm long sol gel coated microextraction capillary was vertically connected to the bottom of the empty gravity fed sample dispenser . Liquid sample (15 mL) was allowed to flow through the sol gel micr oextraction capillary under gravity for 45 minutes. Following this, the capillary was removed from the dispenser, and the microextraction capillary was installed in the GC injector. About 9 cm of the sol gel capillary was contained inside the GC injectio n port (which was held at 40C). Only a 2 cm segment of the capillary remained in the GC oven. This was enabled by a gas tight connection of the capillary to the lower
124 end of the GC injection port. The portion of the microextraction capillary that was inside the GC oven was connected to one end of a two way press fit fused silica connector. The inlet of a Restek Crossbond 14% cyanopropylphenyl 86% PDMS GC column ( 15 m 0.25 mm i.d. ) was coupled to the other end of the connector. Analytes that were ex tracted onto the sol gel coating of the microextraction capillary were then thermally desorbed from the capillary. This was accomplished by rapidly raising the temperature (60C/min) of the injection port from 40C to 280C for the sol gel BMPO and to 300 C for the sol gel PDMDPS coated microextraction capillaries. Analytes were desorbed in the splitless injection mode, and the split was kept closed for the entire CME GC analysis. The mobile phase transferred the desorbed analytes onto the GC column, and they were focused at the inlet of the GC column maintained at 35C. The GC oven temperature was then programmed from 35C (1 min) to 270C at a rate of 20C/min to achieve chromatographic separation of the desorbed analytes that were further detected by an FID maintained at 350C. 4.3 Results and discussion 4.3.1 Sol gel immobilization of the CME coatings The chemical ingredients used to prepare the sol gel coating solutions are listed in table 1. Trifluoroacetic acid was used in sol solutions to cat alyze the sol gel reactions. Trace amounts of water in the TFA and the methylene chloride was enough to initiate the hydrolysis reaction. Sol gel reactions allowed polymeric chains of BMPO or PDMDPS to become chemically incorporated in the sol gel network as an organic component of the organic inorganic hybrid coating and also to covalently anchor the coating to the inner surface of a fused silica capillary. Portions of the sol gel BMPO (Figure 4 1A) or
125 S i C H 3 O H O ( C H 2 ) 3 O C H 2 C H C H 3 O S i O O ( C H 2 ) 3 C H 3 n S i O H S i O H S i O H + S i O H S i O S i O H C a t a l y s t S i H O O O S i C H 3 O H O ( C H 2 ) 3 O C H 2 C H C H 3 O S i O O ( C H 2 ) 3 C H 3 n S i O O Wall bonded sol gel BMPO coating (A) ( S i O ) x ( S i O ) y C H 3 C H 3 H O S i O H S i O H S i O H + S i O H S i O S i O H S i H O O O O S i O O ( S i O ) x ( S i O ) y C H 3 C H 3 H C a t a l y s t Wall bonded sol gel PDMDPS coating (B) Figure 4 1. Polycondensation of 3D BMPO (A) and PDMDPS (B) sol gel network to fused silica.
126 PDMDPS (Figure 4 1B) networks evolving near the fu sed silica capillary inner walls had the opportunity to become covalently bonded to it via condensation reactions with silanol groups on the capillary inner surface. 4.3.2 Ionic liquid mediated sol gel microextraction capillaries In this work, the effec t s of the presence of two ILs (TTPT or MOIC) in the sol gel coating solution s on the morphology and extraction behavior of the resulting hybrid organic inorganic sol gel sorbents utilized in capillary microextraction were investigated. Extraction of an a queous sample containing 500 ppb decanol, 500 ppb hexanophenone, and 200 ppb phenanthrene was performed on microextraction capillaries with different ionic liquid mediated sol gel coatings (TTPT mediated BMPO, MOIC mediated BMPO, and MOIC mediated PDMDPS). For comparison, the same extraction experiments were performed on sol gel capillaries prepared without the mediation of ILs (BMPO no IL and PDMDPS no IL). The MOIC mediated sol gel BMPO coating (Figure 4 2A) and TTPT mediated sol gel BMPO coating (Figure 4 2B) were both able to provide more efficient extractions than the BMPO no IL sol gel coating (Figure 4 2C). Likewise, the MOIC mediated sol gel PDMDPS coating (Figure 4 3 bottom) provided a superior extraction performance compared to the PDMDPS no IL c oating (Figure 4 3 top). Clearly, the ILs had an explicit effect in on the extraction capability of the prepared sol gel sorbents. Both non polar (PDMDPS) and moderately polar (BMPO) sol gel sorbent coatings can be prepared following the described proced ure. However, in the case of BMPO based sol gels, the MOIC mediated sol gel BMPO coated capillary (Figure 4 2 A) provided better extraction performance than TTPT mediated sol gel BMPO coated capillary (Figure 4 2B). The MOIC mediated sol gel BMPO coated m icroextraction capillary
127 (A) (B) (C) Figu re 4 2. Comparison of CME GC analysis of 500 ppb decanol, 500 ppb hexanophenone, and 200 ppb phenanthrene on ( A ) MOIC mediated sol gel BMPO, ( B ) TTPT mediated sol gel BMPO; and ( C ) sol gel BMPO no IL microextraction capillaries. Extraction conditions: 11 cm 0.25 mm i.d. microextraction capillary; extraction time, 45 min (gravity fed at room temperature). Other conditions: 15 m 0.25 mm i.d. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 280C; programmed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) decanol, (2) hexanophenone, and (3) phenanthrene for all chromatograms.
128 Figure 4 3. Comparison of CME GC analysis of 500 p pb decanol, 500 ppb hexanophenone, and 200 ppb phenanthrene on MOIC mediated sol gel PDMDPS (bottom) and sol gel PDMDPS no IL (top) microextraction capillaries. Extraction conditions: 11 cm 0.25 mm i.d. microextraction capillary; extraction time, 45 mi n (gravity fed at room temperature). Other conditions: 15 m 0.25 mm i.d. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 300C; programmed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C. Peaks: (1) decanol, (2) hexanophenone, and (3) phenanthrene for all chromatograms. provided enhanced GC peak areas, and the enhancement factors were 1.5 for decanol, 2.3 for hexanophenone, and 2.1 for phenanthrene. 4.3.3 Role of ionic liquids ILs are known to act as porogens in sol gel systems [ 23 30]. Thus, the IL mediated sol gel coatings provided enhanced GC peak areas because they were more
129 porous than the non IL mediated BMPO sol gel. Furthermo re, SEM investigation of the morphology of the two BMPO sol gel coatings revealed that the MOIC mediated sol gel BMPO coating (Figure 4 4A) had a more porous morphology than the TTPT mediated sol gel BMPO coating (Figure 4 4B). Enhanced GC peak areas prov ided by the MOIC mediated sol gel BMPO coating is indicative of a higher surface area of this coating compared to the TTPT mediated sol gel BMPO coating. It has been pointed out [ 23 ], that IL s with the same cation but different anions could have different effects on the porosity of mesoporous silica materials. It has also been noted that pore size of silica gel can be affected by variations of ILs . Therefore, it is logical to assume that the structural differences of the ionic liquids ( Table 4 1) li kely resulted in varying effects on porosity of the sol gel BMPO material. Ionic liquids are green solvents  and have been used by separation scientists as chromatographic stationary phases [38,39] and as extraction solvents . One important quest ion that naturally arises is what role (if any) is played by the ionic liquids in the CME extraction process using sol gel coatings prepared with the mediation of an IL (MOIC or TTPT). The answer becomes evident by looking into the decomposition temperatu res of the used ILs. The decomposition for both of these ILs takes place at 190C [41,42]. Since the CME capillaries were thermally conditioned above decomposition temperatures of these ILs (conditioning temperature s for sol gel BMPO and PDMDPS were 280 C and 300C, respectively) it is safe to assume that the ILs had
130 Figure 4 4. Scanning electron microscopic images of cross sections of 250 m m i.d. ( A ) MOIC mediated sol gel BMPO (370) and ( B ) TTPT mediated sol gel BMPO (350) coated microextraction capillaries. decomposed and the de composition products had been carried away from the capillary by the purging flow of helium Following this purging, the capillaries were rinsed with a Sol gel BMPO MOIC Coating A Sol Gel BMPO TTPT Coating B
131 mixture of 1:1 v/v CH 2 Cl 2 and CH 3 OH and dried prior to use to ensure that any debris formed on the surface of the sol gel coating during heating as well as unbonded chemicals were removed. Thus, it is logical to assert that the used ILs did not participate in the extraction process and that extraction of analytes from the sample matrix occurred by interaction with the organic inorganic hybrid sol gel coating. 4.3.4 Extraction profiles of various analytes obtaine d on MOIC mediated sol gel BMPO microextraction capillary Since the MOIC mediated sol gel BMPO coated microextraction capillary provided the best extraction performance out of all of the prepared sol gel coatings, we further investigated these sol gel coa tings. Figure 4 5 illustrates the extraction profiles of 2 t butyl 4 methoxyphenol, decanol, hexanophenone, and phenanthrene on the MOIC mediated sol gel BMPO microextraction capillary. This IL mediated sol gel coating provided a fast equilibrium time (5 10 minutes) for relatively polar compounds like 2 t butyl 4 methoxyphenol, decanol, and hexanophenone and a slower equilibrium time (60 70 minutes) for nonpolar analytes like phenanthrene. These results indicate that the phenol, alcohol, and the ketone h ad a high affinity for the MOIC mediated sol gel BMPO coating. Phenanthrene had a lower affinity for the IL mediated sol gel coating and was extracted slower making the equilibration time longer. It was somewhat unexpected considering the hydrophobicity of the PAH. From a practicality perspective, this is an important, highly desirable result considering the difficulties associated with the extraction of polar analytes from aqueous matrices .
132 Figure 4 5 Extraction profile for a mixture of decano l, hexanophenone, and phenanthrene extracted on 11cm 0.25 mm i.d. MOIC mediated sol gel BMPO microextraction capillary from an aqueous sample. Extraction conditions: triplicate extraction at various time intervals. GC analysis conditions: 15 m 0.2 5 mm i.d. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless desorption; injector temperature was 280C; programmed temperature GC run from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C. 4.3.5 Therma l stability of MOIC mediated BMPO sol gel coatings The thermal stability of the MOIC mediated sol gel BMPO microextraction coating was evaluated by conditioning the coated capillary stepwise at higher temperatures and performing extractions on the capillar y after every condit ioning step. The MOIC mediated sol gel BMPO capillary was thermally conditioned stepwise for 1 hr each at 280C, 290C, 300C, 310C, 320C, 330C, 340C, and 350C in a GC oven purging the capillary with helium (1 mL/min). The GC pe ak areas of the extracted analytes (decanol and hexanophenone) remained practically constant in this conditioning process until 330C. A slight drop in GC peak area for the analytes was observed at conditioning temperatures exceeding 330C (Figure 4 6), in dicating that the coating was stable at least up to 330C. The reduction in GC peak area can be attributed to a change in
133 the extraction performance of the BMPO polymer due to the onset of thermal degradation. BMPO is a relatively low molecular weight (50 0 900 g/mol) polyalkylene oxide material that demonstrated this remarkable thermal stability when used in the IL mediated sol gel. By comparison, conventionally prepared coatings for a polyalkylene oxide (e.g. PEG, Ucon, etc.) of similar molecular weight is unlikely to exceed 200 250C . The excellent thermal stability is due to the strong chemical bonding between the MOIC mediated sol gel BMPO coating and the inner walls of the fused silica capillary. 4.3.6 CME GC analysis of various classes of com pounds in aqueous samples using MOIC mediated sol gel BMPO microextraction capillary The sol gel coatings extraction ability was investigated using nonpolar (PAHs), moderately polar (aliphatic aldehydes and aromatic ketones) (Table 4 3), and polar (aliph atic alcohols, aromatic amines, phenols, and free fatty acids) (Table 4 4) test Figure 4 6 Effect of conditioning temperature on the performance of MOIC mediated sol gel BMPO microextraction capillary. CME GC conditions: extraction time, 45 min; 1 5 m 0.25 mm i.d. Restek Crossbond 14% cyanopropylphenyl 86% PDMS coated GC column; splitless injection; injector: initial 40C, final (mentioned on x axis), programmed at a rate of 60C/min; GC over temperature programmed temperature from 35C (1 min) to 270C at a rate of 20C/min; helium carrier gas: FID 350C.
134 solutes. For all of the studied analytes, run to run GC peak area relative standard deviation (RDS) values were determined to evaluate the repeatability of CME with the MOIC mediated sol gel BM PO coating GC peak area RSD values ranged from 0.4% to 5.7% for the nonpolar/moderately polar compounds. The RSD values ranged from 0.3% to 6.7% for the polar analytes. These reasonably small RSD values translate into excellent repeatability in CME per formance of the MOIC mediated sol gel BMPO coating for the classes of compounds investigated. This coating provided ng/L detection limits for all the analytes polar, nonpolar, and moderately polar. Furthermore, the sol gel coating was solvent resistant since it was used in all extraction experiments after it had been rinsed with organic solvents. Capillary to capillary RSD values in GC peak areas of extracted analytes (which is a measure of reproducibility of the coating procedure) was determined by obt aining GC peak area values for decanol, hexanophonone, and phenanthrene extracted on three individually prepared MOIC mediated sol gel BMPO microextraction capillaries. The Table 4 3 Run to run repeatability (peak area) and detection limit data for non polar and moderately polar analytes in three replicate measurements by CME GC using sol gel BMPO MOIC coated microextraction capillaries. Chemical Class Name of Analyte Mean Peak Area (arbitrary unit) RSD (%) Detection Limit S/N = 3 (ng/L) PAH Acena phthene 77.5 1.8 11.6 Phenanthrene 395.3 5.7 2.3 Pyrene 232.8 4.5 1.9 Ketone Hexanophenone 93.0 2.6 24.2 Heptanophenone 176.4 2.4 12.8 Decanophenone 216.2 4.5 6.2 Aldehyde Decanal 39.2 0.4 69.0 Undecanal 55.3 3.2 40.6 Dodecanal 114.7 2.3 19.6
135 Table 4 4. Run to run repeatability (peak area) and detection limit data for polar and moderately polar analytes in three replicate measurements by CME GC using sol gel BMPO MOIC coated microextraction capillaries. Chemical Class Name of Analyte Mean Peak Area (arbitrary unit) RSD (%) Detection Limit S/N = 3 (ng/L) Alcohol Nonanol 59.9 6.6 60.1 Decanol 53.8 1.7 41.8 Undecanol 182.8 6.7 12.3 Aromatic Amine N Butylaniline 32.8 5.9 109.8 Acridine 116.0 2.6 31.0 Diphenylamine 188.0 0.3 19.1 P henol 2,4,6 Trichlorophenol 23.5 2.2 153.3 2 tert Butyl 4 methoxyphenol 39.7 4.3 90.7 Pentachlorophenol 55.1 0.3 65.3 Acid Nonanoic Acid 10.9 5.0 330.5 Decanoic Acid 32.6 4.0 110.3 Undecanoic Acid 111.3 3.1 32.4 capillary to capillary GC peak ar eas obtained provided RDS values of 7.2% for decanol, 8.6% for hexanophenone, and 3.9% for phenanthrene. These RSD values are indicative of acceptable reproducibility of the used sol gel coating procedure. 4.4 Conclusion IL mediated sol gel BMPO and PDMD PS coatings were developed for use as immobilized sorbents in capillary microextraction. The ionic liquid mediated sorbents provided more efficient extractions and lower detection limits compared to analogous sol gel coatings prepared without IL. The MO IC mediated sol gel BMPO coating provided superior preconcentration performance than TTPT mediated sol gel BMPO coating. SEM investigations revealed that the use of MOIC in the sol gel system resulted in a more porous morphology responsible for a more eff icient extraction performance.
1 36 The MOIC mediated sol gel BMPO coating provided consistent extraction results in CME GC analysis (run to run peak area RSD values ranged from 0.3% to 6.7%) for nonpolar and polar analytes. The MOIC mediated sol gel BMPO coat ing was thermally stable up to 330C, and the coating had a high affinity for polar and moderately polar compounds. The main advantage of the MOIC mediated sol gel BMPO coating was that it provided fast extraction of polar analytes from aqueous media. Ov erall, the use of IL in the preparation of sol gel sorbents shows potential for producing stable, reliable, and sensitive extraction media. 4.5 References for Chapter Four  N. Fontanals, R. M. Marce, F. Borrull, J. Chromatogr. A 1152 (2007) 14.  J. Pawlisz yn, S. Liu, Anal. Chem. 59 (1987 ) 1475.  R Belardi, J. Pawliszyn, Water Pollut. Res. J. Can 24 (1989 ) 179.  Z. Zhouyao, J. Poerschmann, J. Pawliszyn, Anal. Commun. 33 (1996) 219.  M. A. Jeannot, F. F. Cantwell, Anal. Chem. 69 (1997) 235.  Y. He, H. K. L ee, Anal. Chem. 69 (1997) 4634.  H. Lord, J. Pawliszyn, J. Chromatogr. A 885 (2000 ) 153.  E. Baltussen, P. Sandra, F. David, C. Cramers, J. Microcolumn Separations 11 (1999) 737.  R. Eisert, J. Pawliszyn, Anal. Chem. 69 ( 1997 ) 3140.  M. E. McComb, R. D. Ol eschuk, E. Giller, H. D. Gesser, Talanta 44 (1997 ) 2137.  H. Hartmann, J. Burhenne, M. Spiteller, Fresenius Environ. Bull. 7 ( 1998 ) 96.  S. Bigham, J. Medlar, A. Kabir, C. Shende, A. Alli, A. Malik, Anal. Chem. 74 (2002) 752.  W. Mullett, J. Pawliszyn, J. Sep. Sci. 26 (2003 ) 2510.
137  D. Djozan, Y. Assadi, S. Haddadi, Anal. Chem. 73 (2001 ) 4054.  S. Chong, D. Wang, J. Hayes, B. Wilhite, A. Malik, Anal. Chem. 69 (1997 ) 3889.  S. Kulkarni, L. Fang, K. Alhooshani, A. Malik, J. Chromatogr. A 1124 (2006) 205.  Z. Z eng, W. Qui, Z. Huang, Anal. Chem. 73 (2001 ) 2429.  Z. Wang, C. Xiao, C. Wu, H. Han, J. Chromatogr. A 893 (2000 ) 157.  R. Gomes da Costa Silva, F. Augusto, J. Chromatogr. A 1072 (2005) 7.  H. Bagheri, A. Es haghi, M. Rouini, J. Chromatogr. B 818 (2005 ) 147  S. Kulkarni, A. M. Shearrow, A. Malik, J. Chromatogr. A 1174 (2007) 50 62.  M. Zhang, Y. Sun, J. Chromatogr. A. 922 (2001) 77.  C. J. Adams, A. E. Bradley, K. R. Seddon, Aust. J. Chem. 54 (2001) 679.  Y. Zhou, J. H. Schattka, M. Antonietti, Nano Lett. 4 (2004 ) 477.  Y. Liu, X. Sun, F. Luo, J. Chen, Anal. Chim. Acta 604 (2007 ) 107.  A. Karout, A. C. Pierre, J. Non Cryst. Solids 353 (2007 ) 2900.  H. Wang, Y. Zhu, J. Lin, X. Yan, Electrophoresis 29 (2008 ) 952.  C. He, Y. Long, J. Pan, K. Li, F. Liu, Ta lanta 74 (2008 ) 1126.  J. Zhang, Y. B. Ma, F. Shi, L. Q. Liu, Y. Q. Deng, Microp. Mesop. Mater. 119 (2009) 97.  R. K. Donato, M. V. Migliorini, M. A. Benvegnu, M. P. Stracke, M. A. Gelesky, F. A. Pavan, C. M. L. Schrekker, E. V. Benvenutti, J. Dupont, H. S. Schrekker, J. Sol Gel Sci. Technol. 49 ( 2009) 71.  M. A. Klingshirn, S. K. Spear, J. D. Holbrey, R. D. Rogers, J. Mater. Chem. 15 (2005) 5174.  J. D. Holbrey, R. D. Rogers, J. Mater. Chem. 15 (2005) 5174.  A. Safavi, N. Maleki, M. Bagheri, J. Mater. Chem 17 (2007) 1674.
138  A. M. Shearrow. G. A. Harris, L. Fang, P. K. Sekhar, L. T. Nguyen, E. B. Turner, S. Bhansali, A. Malik, J. Chromatogr. A (2009) doi:10.1016/j.chroma.2009.04.093.  J.D. Hayes, A. Malik, J. Chromatogr. B 695 (1997) 3.  J. D. Hayes, Ph.D. Dissertation, University of South Florida, Tampa, FL, 2000.  S. A. Forsyth, J. M. Pringle, D. R. MacFarlane, Aust. J. Chem. 57 (2004) 113  D. W. Armstrong, L. F. He, Y. S. Liu, Anal. Chem. 71 (1999) 3873.  J. L. Anderson, in: M. Koel (Ed .), Ionic Liquids in Chemical Analysis, CRC Press, Boca Raton, FL, 2009, pp. 139 165.  J. G. Huddleston, H. D. Willauer, R. P. Swatloski, A. E. Visser, R. D. Rogers, Chem. Comm. 16 (1998) 1765.  Merck, Ionic Liquids: New Materials for New Applications, htt p://ildb.merck.de/ionicliquids/en/startpage.htm, 2009.  B. K. M. Chan, N H. Chan, M. R. Grimmett, Aust. J. Chem. 30 (1977) 2005.  G. Castello, G. DAmato, J. Chromatogr. 90 (1974) 291.
139 CHAPTER 5: TOWARDS A MICROCHIP BASED SEPARATIO N SYSTEM: CHALLENGES IN GLASS MICROFABRICATION 5.1 Introduction Over the past several decades, a push towards miniaturization of traditional analytical instrumentation has occurred with the objective being to create devices capable operating in the fiel d or in a point of care setting. Through miniaturization, researchers seek to incorporate sample introduction, separation, detection, and data output onto a single microchip . Such devices are termed Lab on a chip and are also referred to as "Micro Total Analysis Systems" (TAS). They are a division of Micro Electro Mechanical Systems (MEMS) devices. Some advantages of Lab on a chip devices compared to traditional instrumentation include small sample consumption, decreased sample processing time increase in sample purity, a decrease in power consumption of the instrument, and a potential increase in accuracy and precision of analysis . In particular, miniaturized analytical separation instrumentation has been widely investigated. The firs t micro separation system was a gas chromatograph fabricated on a silicon (Si) wafer by Terry and co workers in 1975 [2, 3]. Other microfluidic separation systems include those based on liquid chromatography (LC) [4, 5], capillary electrophoresis (CE) [6 8], and gas chromatography [9, 10]. Extensive reviews have been published concerning these types of Lab on a chip devices [11 14]. Specifically, microchip based electrophoretic devices have been widely explored due to the inherent simplicity of fabric ation, operation, and control of these devices .
140 CE is a separation technique that is based on electrokinetic phenomena. This transport mechanis m is termed electroosmotic flow (EOF). The use of EOF simplifies the fabrication of the microfluidic dev ices because no complex pumps or valves need to be micromachined . The first demonstration of this technique in planar chip format was accomplished in the early 1990s by Harrison and Manz [4, 7]. Another technique, capillary electrochromatography (CEC ), has the ability to separate both neutral and ionic species. In CEC, ionic species are separated based on their mass to charge ratio (as is the case with CE), and neutral species are separated based on the interaction of the analytes with a stationary p hase coating (as is the case with LC). The first demonstration of micro electrochromatography ( m EC) was accomplished in 1994 by Ramsey and co workers [ 9 ]. In general, silicon, glass, or polymer substrates can be used to fabricate microfluidic devices. Silicon wafers are the most commonly used substrates for MEMS devices because Si wafers are widely available and etching techniques are well established. However in the case electrophoretic devices, Si wafers are rarely used because Si is a semiconducto r; thus it can interfere with generation of EOF. Nitride coatings have been used as insulators on Si wafers, but these films can breakdown beyond 1200V . Polymers such as PDMS  and PMMA  have been used to develop electrophoretic devices. Whi le plastic substrates allow for complex device fabrication, some polymers have poorly defined EOF, can adsorb hydrophobic analytes, and their surfaces can be difficult to modify [13, 14]. Glass substrates include Pyrex, borofloat, fused silica, soda lim e, and quartz. (From this point on we will use the term glass interchangeably with any of these types of
141 substrates.) Glass substrates are widely used to prepare electrophoretic microfluidic devices because they are insulators, biocompatible, reusable, th eir surfaces can be modified by taking advantage of silanol chem istry, and some substrates (i.e. fused silica and quartz) are transparent to the UV Visible region and thus allow for on chip optical detection of analytes. However, there are some drawbacks associated with glass microfabrication including difficult etching processes, high temperatures required for glass to glass bonding, and high wafer cost. In this work, we attempted to developed microfluidic channels in Pyrex or fused silica wafers using w et etching and deep reactive ion etching (DRIE). The aspects of the etched channels were characterized, and the challenges associated with developing glass microfluidic channels are discussed. The eventual use of the device will be as a m EC that contains ionic liquid mediated sol gel sorbents as a stationary phase. 5.2 Experimental 5.2.1 Equipment and materials A Laurel spinner (Laurel Technologies, North Wales, PA, USA) was used to coat wafer with photoresist, and a Karl Suss mask aligner (Suss Mic rotec, Garching, Germany) was used for photolithography. A Tencor Alpha 200 Automatic Step Profiler (KLA Tencor, Milpitas, CA, USA) was used to obtain profile readings of etched channels. An Alcatel Adixen AMS 100 SE DRIE (Alcatel Lucent, Paris, France) was used to etch channels into wafers. A Hitachi model S 800 scanning electron microscope (Hitachi, Tokyo, Japan) was used to obtain SEM images of etched fused silica microfluidic channels. A Despatch oven (Despatch Industries, Minneapolis, MN, USA) was u sed to bake wafers. An ArtCut M40 CO 2 laser engraver (MBKP, Albuquerque,
142 NM, USA) was used to create through holes in the wafers. AZ 4620, AZ400K, and AZ300T were purchased from Clariant (Charlotte, NC, USA). Fused silica wafers (4 i.d., 500 micron thic kness) and Pyrex 7740 wafers (2 i.d., 500 micron thickness) were purchased from Universitywafer (South Boston, MA, USA). Hydrofluoric acid (HF) and hexamethyldisilazane (HMDS) was purchased from Sigma Aldrich (St. Louis, MO, USA). 5.2.2 Fabrication of m icrofluidic channels 188.8.131.52 Pyrex wafers Microfluidic channels fabricated on a Pyrex 7740 wafer using a CAD designed mask with serpentine and spiral channels (figure 5 1). The wafer was cleaned with acetone, methanol, and water before use. Then the wafe r was annealed at 450C for 1min. Nickel (Ni) (20 nm) and gold (Au) (200 nm) metal layers were thermally evaporated onto the wafer. HMDS and S1813 positive photoresist were sequentially spun onto the wafer at 5000 rpm for 60 sec. The wafer was exposed fo r 6 sec using a Karl Suss mask aligner at an intensity of 25 mW/cm 2 and it was developed for 20 sec in MF 319. The exposed metal layers were stripped, and the backside of the wafer was protected with chromium (Cr) (15 nm) and Au (200 nm) layers. The Pyr ex wafer was etched under sonication (10 min) in a dilute hydrofluoric acid solution (15% HF). The remaining metal layers were stripped and the wafer was inspected with an Olympus microscope. The process flow is shown in figure 5 2.
143 Figure 5 1 CAD designed mask: serpentine (left) and spiral (right) Figure 5 2 Process flow for wet etching of Pyrex wafers. (A) Au/Cr deposition; (B) photoresist deposition; ( C) photoresist patterning; (D) development of exposed photoresist; (E) aqua regia/Cr etch to remove Au/Cr layer for etching, and then protection of backside with Au/Cr; (F) HF wet etch; and (G) stripping of photoresist and Au/Cr. C B A D F G E glass Au/Cr Photoresist
144 184.108.40.206 Fused silica w afers Two wafers were cleaned with acetone, methanol, water, and were blow dried prior to use. A Laurel spinner was used to (1) apply a layer of HMDS onto the wafers at 3500 rpm for 30 sec, and (2) a layer of AZ4620 photoresist onto the wafers at 500 rpm for 10 sec followed by 1500 rpm for 60 sec. The wafers were then soft baked for 20 min in an oven at 90C. The wafers were exposed to a CAD designed mask (figure 5 3) using the Karl Suss mask aligner for 55 sec at an intensity of 25 mW/cm 2 The wafers w ere developed for 3.5 min in a mixture of 250 mL of AZ400K to 150 mL of water. The wafers were hard baked at 90C for 20 min, and an edge bead removal was performed prior to etching. Wafers were loaded into the DRIE and etched for 30 min using a mixture of C 4 F 8 (17 sccm), CH 4 (13 sccm), and Ar (100 sccm). The substrate generator had a power of 550 W, and the source generator had a power of 2800 W. A b ias voltage of 185V, and a press ure of 0.85 Pa was used. Remaining photoresist was stripped using warm AZ300T or a base bath (1M NaOH in 4:1 v/v isopropanol: water). The process flow is shown in figure 5 4. Through holes were then created in the cover wafer using laser ablation. Figure 5 3 CAD designed mask with spiral cha nnel, and alignment markers.
145 Figure 5 4 Process flow for fabrication of microfluidic channels: photoresist deposition (A), patterning of photoresist (B), dry etch of fused silica (C), removal of photoresist layer (D). 5.3 Results and discussion 5.3.1 Wet etched microfluidic channel characteristics Wet etching of glass occurs isotropically, the material etches faster horizontally than it does vertically creating an undercut of the hard mask (Figure 5 2F). Often with wet et ched glass materials are undercut and have v shaped trenches form (Figure 5 2G). In order to ensure that only desired sections of the wafer are etched, a hard mask must be used to protect areas from being etched. Typically, Cr/Au hard masks are used fo r wet etching of glass since metal is impervious to HF [6, 7]. However, some problems are associated with Cr/Au masks such as pinholes and fraying of etched edges due to poor mask definition . In this case, we utilized a Ni/Au hard mask on the top si de of the wafer in order to obtain a better quality etch since Ni is know to give less pinhole than Cr . The resulting channel dimensions were 10 m m deep, 40 m m wide at the bottom, and merged at the top (Figure 5 5). The typical v shaped trench was not observed since only 10 microns was etched. Channels merged because the distance between channels on A B C D Photoresist
146 Figure 5 5 Profiler image of wet etched Pyrex wafers. the CAD mask was not wide enough to compensate for the isotropic etch rate in Pyrex. Channels also had frayed edges (indicating that the metal masks were flaking off at the edges) and some pinholes were evident (figure 5 6). Overall, the etch quality was poor, and thus these channels were deemed unfit for further use. Figure 5 6 S turn spiral wet etched into a Pyrex wafer. Glass sur face Merged Channels Individual channels 10 um
147 5.3.2 Dry etched microfluidic channel characteristics Dry etching was investigated because it has the potential to create clean features with high aspect ratios. Comm on hard masks for dry etching of glass include metal [16 18], bonded Si wafers [19, 20] amorphous Si  and polymers [21, 22]. Metal masks can be difficult to use in some machines due to back sputtering of metal leading to micro masking, and bonded wafe rs may be difficult to remove. Thus, masks such as amorphous Si or photoresist have been investigated due to their compatibility with DRIE machines and their ease of removal after the final process. While amorphous Si can be useful as a mask it must be app lied via plasma enhanced chemical vapor deposition. Creating uniform layers from batch to batch is tricky resulting in difficulties in reproducing etch processes. In this work, a thick layer of AZ 4620 photoresist (8.5 microns) was used as a hard mask for dry etching. The photoresist hard mask could withstand 30 40 minutes of dry etching. The etch rate of fused silica under these conditions was determined to be 0.5 m m/min. The etched microfluidic channels created in the fused silica wafer were about 1 5 16 microns deep (figure 5 7). Trenches were nearly rectangular in profile and did not merge (figure 5 8) as is to be expected with DRIE. In contrast to the wet etched channels (figure 5 6), the dry etched channels did not have frayed edges and they wer e not merged (figure 5 9). Some difficulties associated with this process included optimizing the lithography process to create thick masks. For example, long baking times and multiple layers of photoresist maybe required to obtain thicknesses greater tha n 10 microns. Also, etching procedures could only be accomplished for 30 min at a time when photoresist was utilized as a hard mask. If etch processes were performed for longer than 30 min,
148 then debris from the plasma chamber would fall onto the wafer. Thus, 30 min cycles of etching had to be following by 20 min of O 2 cleaning on a dummy wafer. This significantly increased the amount of time required to etch wafers compared to wet etching techniques. Figure 5 7 Profile ( m m) of spiral microfluidic channels in fused silica after 30 min of DRIE. Figure 5 8 Scanning electron microscopic image of cross section an etched wafer after 30 min of DRIE (600 ). 20 m m
149 Figure 5 9 S tur n spiral dry etched into Pyrex wafer. 5.3.3. Fabrication of through holes Through holes are often needed in microfluidic devices to create a means to get sample into and out of the system. There are a few options for creating through holes in glass wafe rs including deep wet etching, ultrasonic drilling, powder blasting, and laser ablation. Wet etching and powder blasting can be difficult to repeat. For example, holes can be too large (due to isotropic etching). Mechanical drilling has the advantage th at hole can be properly aligned, but drill heads are fragile and expensive and the amount of stress that is put on the wafer can cause it to crack. In some cases laser ablation is useful because it is fast, but it can be difficult to align and damage occu rs about 1mm around the area where the hole was created. In this work, we utilized laser ablation for its speed and simplicity. Through holes with a diameter of 400 450 microns were created in the fused silica wafer using a CO 2 laser (figure 5 10). Hole s could be easily replicated, but there was a wide area of damage around the hole.
150 Figure 5 10 Through hole in fused silica wafer created via laser ablation. 5.4 Final remarks In essence, dry etching produced etched ch annels that were superior to wet etched channels. Dry etched channels were not merged and did not have frayed edges. Such characteristics are advantageous for complex channel geometries, bonding steps, and for coating the inner wall of the channels with a sorbent material. However, dry etching processes are significantly longer than wet etching processes. Even though much remains to be done in preparing a completed microchip based separation system with in situ prepared sorbent coatings, this work has ou tlined the immense potential in utilizing DRIE to prepare glass microfluidic channels for such a device. Damaged area
151 5 .5 References for chapter five  A. Manz, H. Becker, Microsystem Technology in Chemistry and Life Sciences, Springer, New York, 1998.  S. C. Terry, Ph .D. Thesis, Stanford, Stanford, CA, 1975.  S. Terry, J. Jerman, J. Angell, IEEE Trans Electron Devices 26 (1979) 1880.  A. Manz, Y. Miyahara, J. Miura, Y. Watanabe, H. Miyagi, K. Sato, Sens. Actuators B1 ( 1990) 249.  G. Ocvirk, E. Verpoorte, A. Manz, M. G rasserbauer, H. M. Widmer, Anal. Methods Instrumen. 2 ( 1995) 74.  A. Manz, D. J. Harrison, E. M. J. Verpoorte, J. C. Fettinger, A. Paulus, H. Ludi, H. M. Widmer, J. Chromatogr. 593 ( 1992) 253.  D. J. Harrison, A. Manz, Z. H. Fan, H. Ludi, H. M. Widmer, Ana l. Chem. 64 ( 1992) 1926.  D. J. Harrison, Z. H. Fan, K. Seiler, A. Manz, H. M. Widmer, Anal. Chim. Acta 283 ( 1993) 361.  S. C. Jacobson, R. Hergenroder, L. B. Koutny, J. M. Ramsey, Anal. Chem. 66 (1994) 2369.  D. J. Harrison, P. G. Glavina, Sensors and Act uators B 10 (1993) 107.  C. S. Effenhauser, G. J. M. Bruin, A. Paulus, M. Ehrat, Anal. Chem. 69 (1997) 2035.  S. A. Soper, A. C. Henry, B. Vaidya, M. Galloway, M. Wabuyele, R. L. McCarley, Anal. Chim. Acta. 74 (2002) 87.  J. C. McDonald, D. C. Duffy, J. R. Anderson, D. T. Chiu, H. Wu, O. J. Schueller, G. M. Whitesides, Anal. Chem. 70 (1998) 27.  G. Ockvirk, M. Munroe, T. Tang, R. Oleschuk, K. Westra, D. J. Harrison, Electrophoresis 21 (2000) 107.  Y. Mourzina, A. Steffen, A. Offenhausser, Microsystem Techno logies 11 (2004) 135.  X. Li, T. Abe, M. Esashi, Sensors and Actuators A 87 (2001) 139.
152  L. Li, T. Abe, M. Esahi, J. Vac. Sci. Tech. B 21 (2003) 2545.  J. H. Park, N. E. Lee, J. Lee, J. S. Park, H. D. Park, Microele Eng 82 (2005) 119.  P. W. Leech, Vacuum 55 (1999) 191.  T. Akashi, Y. Yoshimura, J. Micromech. Microeng. 16 (2006) 1051.  K. Kolari, V. Saarela, S. Franssila, J. Micromech. Microeng. 18 (2008) 1.  Baram, M. Naftali, J. Micromech. Microeng. 16 (2006) 2287.
153 ABOUT THE AUTHOR An ne M arie Shearrow received her Bachelor of Science Degree in Chemistry from the University of Central Florida in August, 2003. In the fall of 2004, she joined the University of South Florida graduate program and began her Ph.D. studies under the guidan ce of Dr. Abdul Malik in the department of chemistry and Dr. Shekhar Bhansali in the department of electrical engineering She received a National Science Foundation IGERT fellowship a Department of Homeland Security (DHS) fellowship a University of Sou th Florida MEMS fellowship and a DHS dissertation grant during her graduate career.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200409Ka 4500
controlfield tag 001 002063189
007 cr mnu|||uuuuu
008 100316s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002988
Shearrow, Anne M.
Ionic liquid-mediated sol-gel sorbents for capillary microextraction and challenges in glass microfabrication
h [electronic resource] /
by Anne M. Shearrow.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 153 pages.
Dissertation (Ph.D.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
ABSTRACT: Three ionic liquids (ILs), trihexyltetradecylphosphonium tetrafluoroborate (TTPT), N-butyl-4-methylpyridinium tetrafluoroborate (BMPT), and 1-methyl-3- octylimidazolium tetrafluoroborate (MOIC), were utilized to prepare sol- gel sorbent coatings. Non-polar polydimethylsiloxane (PDMS) and polar poly(ethylene glycol) (PEG), poly(tetrahydrofuran) (PolyTHF) and bis[(3-methyldimethoxy-silyl)propyl] polypropylene oxide (BMPO) polymers were employed to develop novel ionic liquidmediated sol- gel hybrid organic- inorganic sorbents. The novel sorbents were first tested as coatings for capillary microextraction off-line hyphenated to gas chromatography. To gain an understanding of the role of the ionic liquids in the sol-gel process, the preconcentration abilities of these novel coatings were investigated for several classes of compounds utilizing CME-GC.This was accomplished by comparing GC peak areas of a series of analytes extracted on the ionic liquid mediated sol-gel CME coatings with that of analogous peak areas obtained on sol- gel coatings prepared without the ionic liquid. The morphology of these coatings was compared using scanning electron microscopy (SEM) imaging data. Overall, the ionic liquid-mediated sol- gel coatings had more porous morphologies than the sol-gel coatings prepared without ionic liquid. The PDMS andBMPO sol-gel coatings prepared with ionic liquid in the sol solution provided enhanced extraction sensitivity reflected in higher preconcentration effects and lower detection limits than the sol- gel coatings prepared without the ionic liquid. The polar IL-mediated BMPO sol- gel sorbent was further investigated by exploring the extraction profile and thermal stability of these coatings.A further application of ionic liquid-mediated sol-gel sorbents could be as stationary phases in a microchip-based separation system. Towards this goal, microfluidic channels were fabricated in glass substrates using microelectromechanical engineering. Spiral and serpentine channels were etched in Pyrex and fused silica wafers using wet and deep reactive ion etching (DRIE) techniques. Microfabrication protocols such as the use of hard mask and etching times were investigated for both techniques. DRIE produced microfluidic channels that had an etch quality that was superior to wet etched channels. Thus, the ultimate microchip-based separation system should by fabricated using DRIE.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Co-advisor: Abdul Malik, Ph.D.
Co-advisor: Shekhar Bhansali, Ph.D.
t USF Electronic Theses and Dissertations.