NOVEL SOLID-PHASE MICROEXTRACTION ADSORBENT FOR THE FORENSIC DETECTION OF ACCELERANTS IN ARSON SAMPLES GEETHA SELVARAJU UNIVERSITI TEKNOLOGI MALAYSIA iii Specially dedicated to my beloved parents and brother… iv ACKNOWLEDGEMENTS This research could not have been completed successfully without the support and assistance of numerous individuals. First and foremost, Assoc. Prof. Dr. Umi Kalthom Ahmad, my inspiring and most respectable supervisor who assisted and guided me through this entire study. Followed by my co-supervisor, Assoc. Prof. Dr. Abdul Rahim Yacob. I would like to express my deepest gratitude to both of them for their cooperation, tolerances and willingness to share their valuable ideas and comments. My sincere appreciation goes to the laboratory assistants of the Department of Chemistry, UTM especially Mr. Dinda Ahmad Hairol and Mr. Hamzah Basiron for their technical assistance throughout the study. Besides that, I would like to thank all the lab assistants of the Molecular Biology and Microbiology laboratory, Faculty of Science and Materials Science Laboratory, Faculty of Mechanical Engineering for their help in handling instruments. Special thanks goes to all the academic staffs of Chemistry Department, UTM, especially Assoc. Prof. Dr. Zainab Ramli who had shared her valuable knowledge. The continuous encouragement and moral support from my beloved family members is highly appreciated as it was also important in making this research a possible success. I would also like to thank all my dear friends for their invaluable help in completing this project. My deepest gratitude goes to the Ministry of Science, Technology and Innovation, Malaysia (MOSTI) for the financial support under IRPA vote No. 74090. Last but not least, I would like to thank God for giving me the will power and strength to accomplish this project report. May my work glorify His Name. v ABSTRACT Numerous adsorbents are available commercially as coatings for solid-phase microextraction (SPME) technique. However, some analytical methodologies might demand specific properties for the extraction of selected compounds. In this study, a simple, fast, effective and environmental friendly methodology for the determination of accelerants in arson samples using headspace-SPME coupled to gas-chromatographyflame ionization detector (GC-FID) is presented. A lab-made coated fiber prepared by sol-gel method, containing 1:1 molar ratio of octyltriethoxysilane (C8-TEOS): methyltrimethoxysilane (MTMOS) was employed in this technique. The fiber was tested for the headspace extraction of n-alkane standard hydrocarbons and common accelerants. Applicability of the fiber was demonstrated through the detection of accelerants in simulated arson samples. The C8-coated fiber showed a good selectivity for accelerants. Compared with commercial PDMS/DVB fiber, the lab-made coated fiber exhibited higher extraction capability for accelerants, higher thermal stability (up to 300 °C) and longer lifetime (~ 200 times usage). Electron microscopy experiment revealed that the surface of the fiber coating was well-distributed. A porous structure was suggested for the sol-gel derived C8 coating with an approximate thickness of (3-4) µm. The underlying mechanisms of the coating process were discussed and confirmed by infra-red (IR) spectrum. HS-SPME parameters, such as extraction time, extraction temperature and desorption time were optimized. The developed headspace-SPME method using C8-coated fiber showed satisfactory reproducibility (RSD < 6%), linearity (r > 0.9869) and detection limits for accelerants (0.7-1.0) µL. The lab-made SPME adsorbent was shown to be a good alternative to commercial SPME fiber for the determination of accelerants in arson cases. vi ABSTRAK Pelbagai jenis penjerap salutan bagi teknik pengekstrakan fasa pepejal mikro (SPME) boleh didapati secara komersial. Bagaimanapun, beberapa kaedah analisis memerlukan ciri-ciri khas bagi pengekstrakan sebatian yang terpilih. Kajian ini membentangkan suatu kaedah yang mudah, pantas, efektif dan mesra alam sekitar untuk menentukan bahan penggalak kebakaran dalam sampel kebakaran yang disengajakan dengan menggunakan teknik SPME secara ruang kepala (HS) yang digabungkan dengan kromatografi gas-pengesan pengionan nyala (GC-FID). Gentian salutan buatan sendiri yang disediakan melalui kaedah “sol-gel”, dengan nisbah molar oktatrietoksisilana (C8TEOS): metiltrimetoksisilana (MTMOS) 1:1 telah digunakan dalam teknik ini. Gentian tersebut telah diuji untuk pengekstrakan hidrokarbon n-alkana piawai dan bahan penggalak kebakaran yang lazim. Pengunaan gentian tersebut telah ditunjukkan melalui pengesanan bahan penggalak kebakaran dalam sampel simulasi kebakaran yang disengajakan. Gentian salutan C8 ini menunjukkan kepilihan yang baik bagi bahan penggalak kebakaran. Berbanding dengan gentian komersil (PDMS/DVB), gentian salutan buatan sendiri menunjukkan kebolehan pengekstrakan yang lebih tinggi bagi bahan penggalak kebakaran, kestabilan terma yang lebih tinggi (sehingga 300 °C) dan jangka hayat yang lebih lama (~ 200 kali penggunaan). Eksperimen mikroskopi elektron menunjukkan bahawa permukaan gentian telah disalut dengan sempurna. Struktur berliang telah dicadangkan untuk salutan C8 terbitan sol-gel dengan ketebalan yang beranggaran (3-4) µm. Mekanisme proses penyalutan telah dibincangkan dan ditentukan melalui spektrum infra-merah (IR). Parameter HS-SPME, seperti masa pengekstrakan, suhu pengekstrakan dan masa penyahjerapan telah dioptimumkan. Kaedah HS-SPME yang dibangunkan menggunakan gentian salutan C8 menunjukkan kebolehulangan (RSD < 6%) kelinearan(r > 0.9869) dan had pengesanan (0.7-1.0) µL yang memuaskan bagi bahan penggalak kebakaran. Penjerap SPME buatan sendiri menunjukkan suatu alternatif yang baik kepada gentian SPME komersial bagi penentuan bahan penggalak kebakaran dalam kes kebakaran yang disengajakan. vii TABLE OF CONTENTS CHAPTER TITLE PAGE THESIS TITLE i DECLARATION ii DEDICATION iii ACKNOWLEDGEMENTS iv ABSTRACT v ABSTRAK vi TABLE OF CONTENTS vii LIST OF TABLES xi LIST OF FIGURES xii LIST OF SYMBOLS/ABBREVIATIONS/NOTATION/ 1 TERMINOLOGY xv LIST OF APPENDICES xvi INTRODUCTION 1.1 Background 1 1.2 Arson Crime Scene Investigation 2 1.2.1 Arson Samples and Burn Patterns 2 1.2.2 Analysis of Fire-Scene Samples 3 1.2.3 Court Presentation and Arson Evidence 4 1.3 Nature of Accelerants 5 1.4 The Evolution of Accelerant Extraction Technique 7 1.5 Solid-phase Microextraction (SPME) 9 1.5.1 Principles of SPME 11 1.5.2 Extraction Modes With Coated SPME Fibers 13 viii 1.5.3 Recovery of Accelerants by Headspace-SPME (HS-SPME) Technique 1.5.4 Identification of Volatile Accelerants by SPME-Gas Chromatography Analysis 15 1.6 Commercial SPME Fiber Coatings 17 1.7 Novel SPME Adsorbents 18 1.8 Sol-gel Technology 28 1.8.1 Sol-gel Method as a New Tool for SPME Fiber Coatings 2 14 30 1.9 Problem Statements and the Needs of Study 31 1.10 Objectives and Scope of Study 33 EXPERIMENTAL 2.1 Introduction 34 2.2 Chemicals and Materials 34 2.3 Apparatus 36 2.4 Instrumentation 37 2.5 Preparation of the Sol-gel Derived C8-coated Fiber 38 2.6 Preparation of Standard Stock Solution 39 2.7 Procedure for Extractions using Headspace SPME (HS-SPME) 39 2.8 Preparation of Spiked Fire Debris Sample 40 2.9 General Procedures for Characterization of the C8-coated Fiber 41 2.9.1 Lifetime of the Coating 41 2.9.2 Thermal Stability 42 2.9.3 Scanning Electron Microscopy (SEM) 42 2.9.4 Calculation of Fiber Coating Volume 42 2.9.5 FTIR Analysis 43 2.10 Procedure for Optimization of Sol-gel Process 43 2.11 HS-SPME Optimization and Evaluation Procedures 44 ix 2.12 3 2.11.1 Optimization of HS-SPME Operating Conditions 44 Method Validation 46 2.12.1 Determination of Limits of Detection (LOD) 46 2.12.2 Preparation of Calibration Graph 46 CHARACTERIZATION OF THE C8-COATED FIBER, OPTIMIZATION AND UNDERLYING MECHANISM OF SOL-GEL PROCESS 3.1 Introduction 47 3.2 Characterization of the C8-coated Fiber 47 3.2.1 Selectivity for Standard Hydrocarbons 47 3.2.2 Extraction Capability for Standard Hydrocarbons 48 3.2.3 Lifetime of the Coating 49 3.2.4 Thermal Stability 51 3.2.5 Scanning Electron Microscopy (SEM) Analysis 53 3.2.5.1 Surface Characteristics of the Coating 53 3.2.5.2 Estimation and Reproducibility of the Film Thickness 3.2.5.3 Determination of Extracting Phase Volume 3.2.6 3.3 Fourier Transform Infrared (FTIR) Analysis 57 57 Optimization and Possible Underlying Mechanism of Sol-gel Process 59 3.3.1 Optimization of Sol-gel Process 59 3.3.2 Possible Underlying Mechanism of the Coating Process 3.4 55 Conclusions 60 66 x 4 HEADSPACE SPME OPTIMIZATION, METHOD VALIDATION AND PRACTICAL APPLICATION OF THE C8-COATED FIBER 4.1 Introduction 68 4.2 Optimization of HS-SPME Operating Conditions 68 4.2.1 Optimization of Extraction Time 68 4.2.2 Optimization of Desorption Time 70 4.2.3 Optimization of Extraction Temperature 72 4.3 Validation of the Analytical Method 4.3.1 4.4 Accuracy of HS-SPME Method Using C8-coated Fiber 74 4.3.2 Detection Limits of Accelerants 76 4.3.3 Calibration Graph of Target Compounds 77 Analysis of Simulated Arson Samples Using C8-coated Fiber 4.5 5 74 78 4.4.1 Selectivity for Accelerants 78 4.4.2 Extraction Capability for Accelerants 83 Conclusions 85 CONCLUSIONS AND FUTURE DIRECTIONS 5.1 Conclusions 86 5.2 Future Directions 89 REFERENCES 91 APPENDICES 103 xi LIST OF TABLES TABLE NO. TITLE PAGE 1.1 Properties of common accelerants 6 2.1 Names, functions and chemical structures of the principal compounds of sol-gel coating solutions 35 Optimization of HS-SPME experimental parameters [(a) exposure time, (b) desorption time and (c) extraction temperature] 45 3.1 The extraction precision of C8-coated fiber (200 operations) 50 3.2 Reproducibility of the peak area after conditioning at a) 270 ˚C and b) 300 ˚C 52 3.3 The approximate film thickness of five different C8-fibers 56 4.1 Inter-day precision of HS-SPME method for the extraction of standard hydrocarbon compounds using C8-coated fiber under optimum conditions 75 LODs for accelerants determined using C8-coated fiber under optimum HS-SPME conditions 76 2.2 4.2 4.3 Regression line and correlation coefficient of standard hydrocarbon compounds 77 xii LIST OF FIGURES FIGURE NO. TITLE PAGE 1.1 A (a) photograph and (b) an illustration of the commercially available SPME device attached with coated fused-silica fiber. 10 1.2 Basic steps in sol-gel coating technology 29 1.3 The stages of dip coating process 31 2.1 Experimental setup for headspace SPME accelerant extraction 37 2.2 Photographs showing (a) the burning of carpet sample and (b) the burnt fire debris sample 41 3.1 SPME-GC analysis of hydrocarbon compounds using C8-coated fibers Peaks: (1) C8, (2) C10, (3) C12, (4) C14 and (5) C16 n-alkanes. 48 3.2 Comparison of extraction capability between C8-coated fiber and PDMS/DVB fiber. 49 The surface view of sol-gel C8 coated fiber at (a) 150 and (b) 1000 fold-magnification obtained by SEM 54 The cross-sectional view of sol-gel C8 coated fiber at 4000 fold-magnification obtained by SEM 55 3.5 (a) IR spectra of pure C8-TEOS 58 3.5 (b) IR spectra of sol-gel derived C8-coating 59 3.6 Extraction capability of hydrocarbon compounds using fibers prepared from sol–gel solution containing C8-TEOS:MTMOS with the ratios of 0.5:1, 1:1 and 2:1 60 3.3 3.4 xiii 4.1 Extraction equilibration profiles of n-alkane hydrocarbon compounds (C8, C10, C12, C14 and C16) at varying extraction time. Other SPME conditions : extraction temperature of 80 °C and desorption time of 80 s 70 4.2 Desorption profiles of C8-fiber for standard n-alkane hydrocarbon compounds (C8, C10, C12, C14 and C16) by HS-SPME. Extraction conditions: extraction time, 15 min; extraction temperature, 80 °C 72 Influence of extraction temperature on the signal intensity of standard n-alkane hydrocarbons (C8, C10, C12, C14 and C16). Extraction conditions: extraction time, 15 min; desorption time, 80 s 74 4.3 4.4 Calibration graph of HS-SPME for n-alkane hydrocarbon compounds 77 using C8-coated fiber 4.5 (a) GC profiles of direct injection of gasoline. Peak identities: (1) methylbenzene, (2) ethylbenzene, (3) 1, 3-dimethylbenzene, (4) 1, 2- dimethylbenzene, (5) 1-ethyl-2-methylbenzene, (6) 1, 2, 4-trimethylbenzene, (7) 1, 2, 3- trimethylbenzene 80 4.5 (b) GC profiles of gasoline spiked burnt carpet sample using C8-coated fiber. Peak identities: (1) methylbenzene, (2) ethylbenzene, (3) 1, 3dimethylbenzene, (4) 1, 2- dimethylbenzene, (5) 1-ethyl-2-methylbenzene, (6) 1, 2, 4-trimethylbenzene, (7) 1, 2, 3trimethylbenzene 80 4.5 (c) GC profiles of the blank burnt carpet sample using C8-coated fiber 4.6 (a) GC profiles of direct injection of kerosene. Peak identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, (7) 2, 6-dimethylnapthalene, 81 (8) C15 4.6 (b) GC profiles of kerosene spiked burnt carpet sample using C8-coated fibers. Peak identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, (7) 2, 6-dimethylnapthalene, (8) C15 82 4.7 GC profiles of diesel spiked burnt carpet sample using C8-coated fibers. Peak identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, 82 (7) C15, (8) C16 hydrocarbons 4.8 (a) Comparison of sol-gel derived C8-coated fiber and commercially available PDMS/DVB fiber in extracting hydrocarbon compounds from gasoline spiked burnt carpet sample. Label identities: (1) methylbenzene, (2) ethylbenzene, (3) 1, 3-dimethylbenzene,(4) 1, 2-dimethylbenzene, (5) 1-ethyl-2-methylbenzene, (6) 1, 2, 4-trimethylbenzene, (7) 1, 2, 3trimethylbenzene. 83 81 xiv 4.8 (b) Comparison of sol-gel derived C8-coated fiber and commercially available PDMS/DVB fiber in extracting hydrocarbon compounds from kerosene spiked burnt carpet sample. Label identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, (7) 2, 6-dimethylnapthalene, 84 (8) C15 4.8 (c) Comparison of sol-gel derived C8-coated fiber and commercially available PDMS/DVB fiber in extracting hydrocarbon compounds from diesel spiked burnt carpet sample. Label identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, C15, (7) C16 84 xv LIST OF SYMBOLS/ABBREVIATIONS/NOTATION/TERMINOLOGY GC - Gas Chromatography FID - Flame ionization detector GC-FID - Gas Chromatography with flame ionization detector SPME - Solid phase micro extraction HS-SPME - Headspace solid phase micro extraction PDMS/DVB - Polydimethylsiloxane/Divinylbenzene C8-TEOS - n-Octyltriethoxysilane MTMOS - Methyltrimethoxysilane TMMOS - Trimethylmethoxysilane LOD - Limit of Detection RSD - Relative standard deviation v/v - Volume per volume SEM - Scanning Electron Microscopy xvi LIST OF APPENDICES APPENDIX A B C TITLE PAGE An enlarged scanning electron micrograph of the fiber surface view at 100 fold magnification 103 Scanning electron micrograph: cross-sectional view of the fiber at 1820 fold magnification. (Measurement of the film thickness at two different points) 104 Presentations and Publications 105 1 CHAPTER 1 INTRODUCTION 1.1 Background Arson continues to be both an urgent national problem and truly a contemporary crime. Often characterized as a clandestine tool for criminals, the true arson picture is not clearly known. The Fire and Rescue Department Malaysia reported that it has responded to 103,994 fire cases between the year 1997 to 2003 (www.bomba.gov.my, 2006). Loss of life and damage of properties are always a grave concern in fires. Economically, arson impacts insurance premium rates, removes taxable property assets and degrades communities. Historically, the inner areas of large cities often are most hard-hit and the result is that much of the cost of this destructive crime is placed on those who can least afford it (Dennett, 1980). Arson can be defined as an intentional or attempted damaging or destruction, by means of fire or explosion, of the property of another without the consent of the owner, or of one’s own property or that of another with intent to defraud (Rush, 2000). The motives to commit arson are numerous, including revenge, fraud or profit, sabotage, pyromania, vandalism, concealment of another crime and terrorism. Arsons are particularly difficult crime to investigate and prosecute. Its difficulty lies in these areas (O’ Connor, 1987): 2 • Every fire scene need to be treated as a potential arson scene (from the standpoint of security, preservation and evidence) until clear proof of natural or accidental cause is discovered. • The crime itself, if successful, destroys the physical evidence at its origin. In some cases, the evidence is still there but it requires careful and methodical analysis. Arson is a crime that destroys evidence rather than creates it as it progresses and normally there is not much first hand eyewitness evidence. Therefore the findings of a scientific investigation are considered important in criminal and civil court actions. 1.2 Arson Crime Scene Investigation An important aspect of an investigation of a suspected arson case involves the chemical analysis of the debris remaining after the fire. The aim of the on-site investigation is to establish beyond reasonable doubt the cause of the fire origin from an investigation of the fire debris and remaining building structure and obtaining physical evidence pertinent to the ignition source. 1.2.1 Arson Samples and Burn Patterns The presence and distribution of a flammable liquid at a fire scene indicates a deliberately lit fire unless it can be readily explained otherwise. “Sniffers” or portable hydrocarbon detectors are used by many successful investigators (DeHaan, 1997). Burn patterns are the basis of all fire investigation. It can be used to indicate the direction of fire spread observed on structural elements or wall surfaces. The typical Vpattern associated with most fires is a result of several factors, one being that the natural 3 plume shape of a large fire is a cone. This cone’s shape depends on the size of fire with respect to a ceiling that tends to flatten and spread it horizontally. If a fire is located some distance away from the wall, a shallow U- or V-pattern may appear (O’Connor, 1987; DeHaan, 1997). Arson samples are normally collected at the suspect crime scene and submitted to the laboratory either in new aluminium cans or forensic evidence bags. In collecting samples, investigators need to look for places where the ignitable liquids would be protected from the heat because those volatile liquids are driven off by high temperature and not likely to be found in the deepest char areas. Multiple samples should be taken so that the spread of the accelerant is ascertained (Furton et al., 1995). 1.2.2 Analysis of Fire-Scene Samples The majority of fire-scene evidence submitted to the laboratory for analysis consist of various materials, such as wood, plastics, soil, carpet, cloth, metals, liquids and other organic and inorganic substances. Laboratory examinations of the fire debris includes (Almirall et al., 1996a; Harris and Wheeler, 2003; Almirall et al., 2004; Ma et al., 2004): • Macroscopical and microscopical examination of the debris - (most fire debris consists of burned or partially burnt substances, weathering has already occurred before collection. Therefore brief visual examination of fire debris will not cause any great loss of accelerants). • Recovery of the accelerants from the debris – [separation of analytes from the matrix (fire debris, soil and water) and concentrating them]. • Identification of the accelerants – (chemical analysis of the isolated substances and the identification of the mixture). 4 The difficulties with the identification of accelerants traces isolated from fire debris or environmental samples (water and soil) result mainly from the fact that the isolated traces differ qualitatively and quantitatively from the initial mixtures. These differences appear as a result of (Cavanagh et al., 2002; Sandercock and Pasquier, 2003): • Evaporation - in the isolated traces, there is a lack of volatile compounds in comparison to the high boiling ones. • Pyrolysis - as a result of high temperature effect, the big molecules can disintegrate and the disintegration products arise. • Microbiological decomposition. • In the isolated traces, the additional substances appear - for example matrix components or matrix decomposition products, extinguishing medium components. 1.2.3 Court Presentation and Arson Evidence The investigator must use laboratory analysis to support arguments in court. Opinion and theories can be readily challenged in court because of the complex and varied nature of fires, but laboratory evidence is irrefutable proof of the presence of an accelerant. These are the basic elements that need to be considered to prove the crime of arson (Icove and Dehaan, 2004): • There has been a burning of property - (this must be shown to the court to be actual destruction, not just scorching or soothing). • The burning is incendiary in origin – (proof of the existence of an effective incendiary device accomplished by showing how all reasonable natural or accidental cases have been considered and ruled out). • The burning is shown to be started with malice – (this act of burning requires a specific intent of destroying property). 5 1.3 Nature of Accelerants Accelerants are any type of material or substance added to the targeted materials to enhance the combustion of those materials and to accelerate the burning. Accelerants such as petrol, kerosene, diesel, mineral turpentine and methylated spirits are widely utilized in everyday life. Owing to their physico-chemical properties such as volatility and flammability, they are also used as ignitable liquids to commit arsons by offenders. Exotic accelerants such as industrial solvents are rarely used and if so are readily identified by chemical analysis because of their similar chemical characteristics to the common accelerants (Furton et al., 1995; Harris and Wheeler, 2003). Accelerants can be classified as either mixtures of compounds or pure compounds. The chromatographic analysis of a pure compound will feature a single peak while a mixture will give several peaks that contribute to a fingerprint which is used to identify the accelerant. The more volatile components of an accelerant evaporate at a faster rate than the heavier components so that the overall chemical profile of the accelerant will change during the fire and before sampling (Jackowski, 1997). The amount of accelerant remaining at the fire scene available for sampling is governed by factors such as the initial loading, volatility and water solubility of the accelerant, severity of the fire, porosity of the substrate material, dryness of the area after the fire and elapsed time between the fire and sampling. (Almirall et al., 1996b; Cavanagh et al., 2002; Almirall et al., 2004). Samples of a suspected accelerant are often located at the scene in a container and are supplied for analysis. Samples of this type are usually unaffected by evaporation. It is difficult to conclusively determine if a sample of an accelerant was the same as that to initiate or propagate a fire, because of the universal composition of the common accelerants (Dennet, 1980). 6 The chemical components of the common accelerants are aliphatic and aromatic hydrocarbons and oxygenated hydrocarbons such as alcohols. The oxygenated hydrocarbons are somewhat water soluble and therefore washed away during the extinguishing of the blaze, so that little trace remains (Cavanagh et al., 2002; Sandercock and Pasquier, 2002). The most common accelerants used by arsonists are generally low to medium boiling hydrocarbon liquids. Some common accelerants and their properties are shown in Table 1.1. Table 1.1: Properties of common accelerants (Lee et al., 1986). Accelerant Flash Boiling Point (°C) Point (°C) Self- Explosive Limit (%) Ignition Flammable Range Temperature Low High (%) Gasoline -50 102-230 495 1.3 6.0 1.3-6.0 Kerosene 110-165 400-572 490 1.16 6.0 1.2-6.0 Fuel oil 110-165 350-500 490 1.16 6.0 1.2-6.0 110-190 375-650 494 - - - Turpentine 95 - 488 0.8 6.0 - Petroleum 300-400 450-500 0.9 8.0 1.1-6.0 No.1 Fuel oil No.2 20-45 benzin Gasoline, the most important fuel of petroleum origin, is a mixture of volatile, low-boiling and midrange hydrocarbons. The average molecular weight of gasoline is usually taken to be close to that of n-octane, about 114. Modern automotive gasoline contains more than 100 hydrocarbons in a complex mixture. Before being sold as a motor fuel, however, a variety of additives is added by the refinery. These additives 7 often include compounds to improve the burning characteristics of the fuel and dyes to identify one product from another (Almirall et al., 1996a; Almirall et al., 1996b). Kerosene and the distillate fuels of higher boiling ranges were considered useful only for illumination until diesel motor became a common device. Although diesel fuel is similar to kerosene, it spans a wider range of less volatile components. Kerosene and heavier petroleum distillates have long been of significance in the setting of deliberate fire. Their lower volatility presents less hazard to the user than does gasoline. The liquid is more persistent so that less haste is required in its ignition and there is much less danger of explosion (Ren and Bertsch, 1999; Llyod and Edmiston, 2003). Fuel oil No.1 (Jet Fuel A) generally contains paraffinic and olefinic hydrocarbons in the C10-C16 range (this range denotes hydrocarbons having 10 to 16 carbon atoms linked together). It falls in the fraction which follows gasoline in the simple distillation of crude oil. Diesel fuel contains low sulfur content and additives to improve combustion. Diesel fuel and domestic heating oil may both be described as fuel oil No. 2. The heavier stove oil used in room heaters contains hydrocarbons from a slightly higher boiling point range, usually (375-650)°F. The product for home use is carefully controlled to minimize sulfur and water content which could corrode heating units having minimal service. Those fuels intended for industrial service have higher sulfur and ash content and cover a range of products and applications (O’Connor, 1987; DeHaan, 1997; Icove and Dehaan, 2004). 1.4 The Evolution of Accelerant Extraction Technique The forensic discipline of ignitable liquid and fire debris analysis is rapidly changing. Refinements in existing methods as well as development of new techniques are changing the routine methods of analysis. Optimization of existing accelerant extraction, analytical techniques and research into novel methods of extraction have 8 improved the recovery of ignitable liquids from debris samples, sample turnover and reduced the number of inconclusive findings (Dolan, 2003). The earliest methods of identification relied upon simple identification of the headspace odor - often referred to as a “nasal appraisal”. With improvement in analytical instruments, the extraction technique used was sampling a headspace of heated fire debris with a syringe and then injecting the sample into a GC for analysis. Heated headspace analysis is also used for sample screening because it is simple, rapid and easy technique to apply. A sample that gives a negative result when screened by heated headspace would then be subjected to a more sensitive extraction technique. The technique however, discriminates against the less volatile components in the sample which will give less data from the chromatographic analysis for interpretation (Eisert and Pawliszyn, 1997; Lord and Pawliszyn, 2000). Distillation technique are used with steam distillation being the most popular. Distillation involves heating the sample with an extraction medium and condensing the vapor to provide a sample of the accelerant in the extraction medium used. The various medium that have been used are water, ethanol and ethylene glycol. Vacuum distillation with sub ambient trapping of the volatiles can also be used. The accelerant may be further concentrated by controlled evaporation of the medium or by solvent extraction from the medium. Steam distillation can be a lengthy technique and extraction times of up to 48 hours have been reported as being necessary for some samples. The technique however, requires considerable clean up of the apparatus between samples and also considerable operator attention, therefore the sample turnover is low (Eisert and Pawliszyn, 1997; Lord and Pawliszyn, 2000). Solvent extraction is also used and involves soaking the fire debris in a suitable solvent and then filtering and evaporating the solvent to concentrate the sample. The advantages of solvent extraction is that it readily extracts the less volatile components of an accelerant and therefore does not discriminate between fire debris samples. The 9 technique however, requires the use of high purity solvents which are expensive and also matrix components such as monomers, plasticizers, glues and resins are co-extracted which may interfere with the subsequent analysis. Both distillation and solvent extraction require further concentration of the raw extracts to increase the sensitivity of the technique (Dolan, 2003). Dynamic headspace sampling is widely used for the extraction of accelerants. The debris is heated inside the headspace gas chromatograph and if any volatile hydrocarbons are present, they will vaporize and be present in the air or “headspace” directly above the debris. The transfer gas used to sweep the headspace to the adsorbent is usually nitrogen or the headspace can be drawn through the absorbent with a vacuum. Microwave ovens have also been used to heat the sample and the steam generated sweeps the headspace to the absorbent. A major problem of dynamic headspace extraction technique is the carry over of contaminants from previous samples through the gas transfer lines. It was found that by removing the outward gas transfer line and connecting the charcoal tube directly to the sample container, the contamination problem was reduced (Eisert and Pawliszyn, 1997; Lord and Pawliszyn, 2000; Dolan, 2003). 1.5 Solid-phase Microextraction (SPME) In addition to optimizing existing extraction techniques, the field of forensic science has also introduced a new extraction technique for application to fire debris analysis: solid-phase microextraction. Much interest has been shown recently in replacing conventional accelerant residues extraction technique with headspace SPME technique ever since the technique was invented in the early 1990’s by Prof. Janusz Pawliszyn. SPME is a fast, simple, time-efficient, environmental friendly and sensitive technique, which does not require the use of solvents either for sample preparation or cleanup. Analyte extraction and pre-concentration are combined in one single step 10 (Mester et al., 2001; Theodoridis et al., 2000). Figure 1.1 shows the dimensions and architecture of the commercially available SPME device attached with lab-made fiber. (a) (b) Plunger Septum piercing needle Plunger retaining screw Fiber attachment tubing Barrel 19.5 cm 1.0 cm Coated fused silica fiber Figure 1.1 : (a) A photograph and (b) an illustration of the commercially available SPME device attached with coated fused-silica fiber. In SPME, the outer surface of a solid fused-silica fiber is coated with a selective stationary phase. Thermally stable polymeric materials that allow fast solute diffusion are commonly used as stationary phases. The extraction is based on the enrichment of components on a polymer or adsorbent coated fused silica fiber by exposing the fiber either directly to the sample or to its headspace. The extraction time can be considered complete when the analyte concentration has reached a distribution equilibrium between the sample matrix and the fiber coating. When using this in experimental conditions, it means that when the equilibrium is reached, the amount of the extracted analytes is 11 constant and independent of an increase in extraction time. This is followed by thermal desorption in a gas chromatograph (Zhang et al., 1994a). The major advantages of SPME technique are in terms of speed, sensitivity, cost effectiveness and versatility (Zhang et al., 1994b). In SPME, the equilibration can be reached in only 20-30 minutes which makes it ideal for quick screening. The parts per trillion detection limits have been attained, with an ion-trap detector. The solvent purchase and disposable costs are greatly reduced in this technique. Fibers are reusable (lifetime depends on conditions of use, more than 100 uses have been achieved). SPME can be used with any GC or GC-MS with split/splitless or on-column injection. 1.5.1 Principles of SPME A traditional approach to SPME involves coated fibers. The transport of analytes from the matrix into the coating begins as soon as the coated fiber has been placed in contact with the sample. Typically SPME extraction is considered to be complete when the analyte concentration has reached distribution equilibrium between the sample matrix and the fiber coating. In practice, this means that once equilibrium is reached, the extracted amount is constant within the limits of experimental error and it is dependent of further increase of extraction time. The equilibrium can be described as (Lord and Pawliszyn, 2000; Wang et al., 2005): n= KfsVfVsCo KfsVf + Vs (1.1) where n is the number of moles extracted by the coating, Kfs is a fiber coating or sample matrix distribution constant, Vf is the volume of the fiber coating, Vs is the sample volume and Co is the initial concentration of a given analyte. Eq (1.1) indicates that, 12 after equilibrium has been reached, there is a direct proportional relationship between sample concentration and the amount of analyte extracted. Eq. (1.1), which assumes that the sample matrix can be represented as a single homogeneous phase and that no headspace is present in the system, can be modified to account for the existence of other components in the matrix by considering the volumesof the individual phases and the appropriate distribution constants. The extraction can be interrupted and the fiber analyzed prior to equilibrium. Constant convection conditions and careful timing of the extraction are necessary to obtain reproducible data (Lord and Pawliszyn, 2000). When the sample volume is very large, Eq. (1.1) can be simplified as (Wang et al., 2005) : (1.2) n = KfsVfCo which points to the usefulness of the technique for field applications. In this equation, the amount of extracted analyte is independent of the volume of the sample. In practice, there is no need to collect a defined sample prior to analysis as the fiber can be exposed directly to the ambient air, water or production stream. The amount of extracted analyte will correspond directly to its concentration in the matrix, without being dependent on the sample volume. The equilibrium equation for analysis in a vial containing headspace should be expressed as shown in Eq (1.3) (Zhang and Pawliszyn, 1995; Lord and Pawliszyn, 2000): n= KfsVfVs KfsVf + KhsVh + Vs Co (1.3) 13 Where Khs and Vh represent the headspace/sample distribution coefficient and the volume of the headspace, respectively. The fact that SPME is an equilibrium rather than an exhaustive extraction technique means that even after the extraction process has been completed, a substantial portion of the analytes usually remain in the matrix. This presents an opportunity for quantification based on internal standardization, namely the standard is loaded onto the fiber prior to extraction step, instead of spiked into the sample. 1.5.2 Extraction Modes With Coated SPME Fibers Generally, there are three types of extractions that can be performed using SPME: direct extraction, headspace configuration and membrane protection approach. In the direct extraction mode, the coated fiber is inserted or immersed into the sample and the analytes are transported directly from the sample matrix to the extracting phase. In order to facilitate rapid extraction, some level of agitation is required to transport analytes from the bulk of the solution to the vicinity of the fiber. Natural convection of air is sufficient for gaseous sample to facilitate rapid equilibration (Zhang et al.,1994b). In the headspace mode, the analytes need to be transported through the barrier of air before they can reach the coating. This modification serves primarily to protect the fiber coating from damage by high molecular mass and other non-volatile interferences present in the sample matrix such as proteins. This mode also allows the modification of the matrix, such as a change of the pH or temperature, without damaging the fiber (Zhang and Pawliszyn, 1993). 14 Another indirect SPME extraction mode is through a membrane. The main purpose of the membrane barrier is to protect the fiber against damage, similar to the use of headspace SPME when very dirty samples are analyzed. However, membrane protection is advantageous for determination of analytes having volatilities too low for the headspace approach. In addition, a membrane made from appropriate material can add a certain degree of selectivity to the extraction process. The use of thin membranes and increased extraction temperatures will result in faster extraction times (Mester et al., 2001; Theodoridis et al., 2000). 1.5.3 Recovery of Accelerants by Headspace-SPME (HS-SPME) Technique The separation and recovery of accelerant residues from fire debris are essential steps for the identification of the accelerants. HS-SPME is a simple and convenient method for the recovery of accelerant vapors from fire debris. The objectives of analyzing the debris is to determine the presence of any volatile hydrocarbon compounds. Initially, therefore, the accelerant residues (volatile compounds) need to be separated from the debris (nonvolatile matrix). The technique is employed by introducing a coated SPME fiber into the headspace above the debris which is placed in an air-tight container. The volatilized hydrocarbon compounds are extracted and concentrated in the coating. After exposure to the sample headspace, the fiber is transferred to the analytical instruments for desorption of the extracted analytes (Snow and Slack, 2002. Theoretically, any volatiles present will reach a vapor equilibrium in a sealed container. Molecules of the accelerants actually occupy the entire interior space of the container. Part of the interior is occupied by the liquid phase of the accelerant and the remainder is occupied by its vapor phase. The continuous molecular motion in liquids causes the molecules to constantly leave the liquid phase to enter the vapor phase. At the same time, molecules also continuously leave the vapor phase to reenter the liquid 15 phase. The molecules of a volatile compound at any given temperature have a specific average kinetic energy that is dependent upon the absolute temperature of the liquid. Some molecules at the liquid surface have energies greater than the average energy, therefore, they continuously leave and enter the gaseous phase (Snow and Slack, 2002b; Zhang and Pawliszyn, 1993). Heating the sample to an elevated temperature provides energy for analyte molecules to overcome energy barriers which tie them to the matrix, enhances the mass transfer process and increases the vapor pressure of the analytes. However, excessive temperatures encourage the degradation of synthetics in the debris, such as carpet. These degradation products add to the complexity of the chromatogram which is used for accelerant identification. Excessive heating also lowers the level of real accelerants in debris and this does not contribute to the sensitivity of the method (Zhang and Pawliszyn, 1995). 1.5.4 Identification of Volatile Accelerants by SPME-Gas Chromatography Analysis The analytical instrument used most frequently with SPME has been the gas chromatograph (GC). Standard GC injectors, such as split/splitless can be applied to SPME as long as a narrow insert with an inside diameter close to the outside diameter of the needle is used. The narrow inserts are required to increase the linear flow around the fiber, resulting in efficient removal of desorbed analytes. The split should be turned off during SPME injection. Under these conditions, the desorption of analytes from the fiber is very rapid (Lord and Pawliszyn, 2000). GC uses a stream of gas (nitrogen or helium) as a carrier to move a mixture of gaseous materials along a column, tube filled or coated with a separating compound. The analytes interact with the separating compound by alternatively dissolving in it and then 16 volatilizing to be swept further along the column by the carrier. The SPME fiber is introduced onto the column at one end and at the other end is a detector (Eisert and Pawliszyn, 1997). Columns for GC analysis of suspected arson residues are usually set up to offer the best separation of components having volatilities between those of pentane (C5H12) and triacontane (C30H62), since that range includes all of the commonly encountered petroleum products. Packed tubular column (3 mm in diameter) were commonly used but it has now been supplanted by a very narrow capillary columns (0.1 mm in diameter). This resulted in shorter analysis time and the usage of smaller sample sizes (DeHaan, 1997). Although the column can be selected to separate a mixture of compounds on the basis of their chemical properties, most separations in volatile accelerants analysis are made on the basis of differences in the boiling points of the various components. Almost all the components of petroleum distillates of interest are either aliphatic (straight-chain) or aromatic (ring) in structure, thus will have very similar chemical properties within those two groups. The heavier a molecule is, the less volatile it is and it will appear last on the chromatogram (Lee et al, 1986; Zeeuw and Luong, 2002). Flame ionization detector (FID) has been the standard detector for analysis of hydrocarbons but now specialized detectors can be used in arson work. Electron capture detectors, which use a minute radioactive source as a monitor, are very sensitive to particular chemical species such as organometallics and halogenated hydrocarbons. They can be used to characterize a petroleum distillate more precisely than is possible with (FID) alone. There are nitrogen-phosphorus detectors that are not sensitive to common hydrocarbons but will select those compounds in a complex mixture that contain nitrogen or phosphorus. Since the crude oil feed stocks for gasoline vary considerably in their nitrogen, phosphorus and sulfur containing components, it is very likely that such techniques may allow discrimination between gasoline from different origins (DeHaan, 1997; Zeeuw and Luong, 2002). 17 Mass spectrometer (MS) allows each molecule to break apart into small sub molecular pieces and by counting those pieces, the chemical structure of the original molecule can be established. MS is an analytical technique that needs to be fed one compound at a time while GC is a separation technique that is good at separating a mixture of compounds. Since most of the volatile residues in fire debris involve complex mixtures, coupling the two together gives the best tool. GC/MS has the capability of displaying a mass spectrum for each peak, and scan selected ions, those that are characteristic for a particular chemical species of interest (Borusiewicz et al., 2004; Lennard et al., 1995). 1.6 Commercial SPME Fiber Coatings To date, several different coating materials are available from Supelco. There are three different polydimethylsiloxane (PDMS) films of different thickness (7, 30 and 100 µm), 85 µm polyacrylate (PA), 65 µm polydimethylsiloxane/divinylbenzene (PDMS/DVB), 75 µm, Carboxen/PDMS, 65 µm Carbowax/DVB and 50 µm Carbowax template resin fibers (Dolan, 2003). For the right selection of the polymer coating, the general principles of “like dissolves like” applies. PDMS coated fibers are typically the first choice. They are very rugged and liquid coatings are used as GC stationary phases withstanding temperatures up to 300 °C. This coating is non polar showing a high affinity for non-polar compounds. Very thin fiber coatings should be used whenever the sensitivity is sufficient. The extraction time is shorter for a thinner coating and smaller distribution constants of the analytes (Lord and Pawliszyn, 2000; Dolan, 2003). The PA fiber shows a high affinity for polar compounds such as phenols and polar pesticides. PA is a solid polymer, thus the equilibration times are higher compared to the liquid PDMS fibers. The mixed phase coatings are more suitable for volatile 18 compounds. The take up of these fibers is significantly higher compared to PDMS. When changing from PDMS to carbowax using the mixed phase coatings, more polar analytes can be extracted such as ketones and alcohols (Eisert and Pawliszyn, 1997; Lord and Pawliszyn, 2000; Dolan, 2003). All the commercially available coatings can be categorized according to two types of extraction mechanism, absorption and adsorption. Both PDMS and PA extract analytes via absorption. The absorbent type fibers extract the analytes by partitioning them into a ‘liquid-like” phase in which the analytes would migrate in and out of the coating. The remaining coatings, including PDMS-DVB, Carbowax-DVB, Carbowax Template Resin and Carboxen are mixed coatings, in which the primary extracting phase is a porous solid, extracts analytes via adsorption. The adsorbent type fibers extract analytes by physically interacting with the analytes and the extraction can be accomplished by trapping the analytes in the internal pores of the stationary phase (Gorecki et al., 1999). 1.7 Novel SPME Adsorbents Recently, researches had been carried out in developing a new SPME fiber which is more effective than commercially available fibers. Works by Gbatu et al. (1999) have shown that sol–gel procedure can be used to prepare SPME fibers having desired selectivity with judicious choice of precursors. These fibers are stable in different types of organic and acidic solvents and basic solutions unlike commercial SPME fibers. Zuin et al. (2004) employed a novel SPME fiber coated with PDMS/PVA copolymer to determine organochlorine and organophosphorus. The resulting procedure was shown to be a good alternative and environmentally friendly analytical method. Hou et al. (2004) successfully used C16-MCM-41 mesoporous materials as a novel SPME 19 fiber coating. The results indicated that this fiber coating has very high extraction efficiency and good selectivity towards aromatic hydrocarbons. Yu et al. (2004) successfully used allyloxy bisbenzo 16-crown-5 trimethylsilane to prepare the sol–gel derived bisbenzo crown ether/hydroxyl terminated silicone oil (OH-TSO) SPME coating. The long lastingness of this novel fiber is longer than that of commercial fibers. Works by Lopes et al. (2004) have shown that the PDMS/PVA composite can be used as a material for SPME fiber coating. It can be synthesized by sol-gel polycondensation route. The highly microporous coating gave faster extractions and showed potentially higher analyte loadings. Several researches (Li et al., 2004a; Li et al., 2004b) have successfully developed amide bridged-C4/OH-TSO sol-gel coated novel fiber for analysis of aliphatic amines and chlorophenols. The fiber exhibits better sensitivity, high thermal stability and satisfactory fiber-to-fiber reproducibility. For the first time, sol-gel dendrimer coated capillaries were used for solventless microextraction and preconcentration in chemical analysis by Kabir et al. (2004). Due to the strong chemical bonding with capillary inner walls, sol-gel dendron coatings showed excellent thermal and solvent stability. The fiber was capable of extracting both polar and nonpolar analytes including PAHs, aldehydes, ketones, phenols and alcohols. Araanda et al. (2000) used two polycrystalline graphites (pencil lead and glassy carbon) as sorbents for solid phase microextraction of non-ionic alkyl phenol ethoxylate surfactant. Preliminary results showed that pencil leads and glassy carbon performed equally well as SPME fibers (PDMS/DVB and Carbowax-TPR) in terms of LOD, linear dynamic range and precision. Works by Yun (2003) have shown that sol–gel derived diallytriethylene glycol/hydroxyl–terminated silicone oil (DATEG/ OH-TSO) coated SPME fibers have a powerful stationary phase for extraction of polar and non-polar analytes from headspace for the sample due to the selectivity of open crown and three–dimensional network provided by sol–gel process. Owing to the strong chemical binding between the coating 20 and surface, these fibers have high thermal and chemical stability in extraction for high– boiling point compounds. Works by Zeng et al. (2001) have shown that sol–gel derived hydroxydibenzo14-crown-4 coated fiber has good affinity for several aromatic amine derivatives. Wang et al. (2003) successfully developed a new sol–gel coated dihydoxy-terminated benzo15-crown-5 (DOH-B15C5) fiber which exhibits high solvent stability, longer application lifetime, thermal stability (≥ 350 °C), improved selectivity and sensitivity towards different aromatic compounds (non-polar or polar). The porous structure increases the speed of extraction and desorption. Yu et al. (2002) successfully developed a new porous sol–gel hydroxyfullerene (fullerol) fiber which is high temperature and solvent resistance with long durability. It shows excellent extraction characteristics for both less volatile organic compounds and polar substances, especially for planar molecules. Basheer et al. (2005) successfully synthesized amphiphilic and hydrophilic oligomers and coated on fused silica capillaries using sol-gel technique. The sol-gel coated fibers were successfully used for the determination of triazine herbicides in reservoir water samples. Both types of coating were stable under high temperature (up to 280 ˚C). Compared to commercially available SPME adsorbents such as PDMS/DVB and PA, the new materials showed comparable selectivity and sensitivity towards both non-polar and polar analytes. Quantitative results obtained from these capillaries exhibited precision at LOD values comparable with commercial coating materials. The coated capillaries also exhibited longer application lifetime. Hun et al. (2005) successfully prepared a novel poly(dimethylsiloxane)/βcyclodextrin (PDMS/β-CD) coating for solid-phase microextraction (SPME) in the form of membrane instead of fiber. The fibers were capable of extracting non-polar polycyclic aromatic hydrocarbons (PAHs) and polar phenolic compounds. The membrane showed high extraction efficiency without sacrificing the analytical speed. Ugur et al. (2004) 21 described the use of SPME method with new design of functionalized stable overoxidized sulfonated polypyrrole (OSPPy) film electrode to extract metal ions (Ni (II) and Cd (II)) without derivatization from water of high silane. The extremely fast uptake and release were attained by applied potentials inducing changes in the solution pH. Fundamental results showed that (OSPPy) film offers a wide potential range for electroanalytical exploration from selective electrodes to separation and preconcentration sampling device. Chong et al. (1997) and Malik et al. (2004) successfully developed sol-gel PDMS fibers for the extraction of PAHs. The sol-gel fiber exhibited enhanced thermal stability (up to 320 ˚C) and the coating possess a porous structure and reduced thickness which provide enhance extraction and mass transfer rate in SPME. Works by Liu et al. (2005) proved that sol-gel derived co-poly(butyl methacrylate/hydroxyl terminated silicone oil) (BMA/OH-TSO) has the best affinity for 2-chroethyl ethyl sulfide (CEES) in soil. The coatings proved to be stable at high temperature (350 ˚C) and have a longer lifetime (150 times). Works by Wang et al. (2000) showed that polyethylene glycol coated fibers have a high thermal stability (300 ˚C). This work proves that the thickness of the polymeric layer can be controlled by varying the dipping time and the concentration of the sol solution. Zhou et al. (2005) successfully developed a lab-made calixarene fiber which exhibited a high affinity for the phenolic compounds and a higher extraction capability compared to commercial polyacrylate fiber. Liu et al. (2005b) developed sol-gel coated hydroxyl-terminated silicone oil-butyl methacrylate-divinbylbenzene (OH-TSO-BMADVB) and validated for the determination of volatile compounds in wine. The novel coating was found to be very effective in carrying out extraction of both polar alcohols and fatty acids and non-polar esters. Liu et al. (1997) successfully prepared porous layer–coated SPME fibers by immobilizing the C-8 and C-18 bonded silica particles (5 µm diameter, typically used 22 for reversed-phase HPLC) onto the surface of metal wires using high-temperature epoxy. The fiber was capable of extracting aromatic hydrocarbons. Porous silica particles provide a large specific surface area, therefore, the capacity of the coating is increased. It was found that the desorption time using these fibers was less than 10 s, which provides high chromatographic efficiency without cryogenic trapping. Xia and Leidy (2001) prepared C18-bonded silica-coated multifibers by applying thin film of epoxy glue over the glass fiber and pressing onto a C18-bonded silica particle bed. It was found that the absorption rate of this fiber was 10 times higher than that of a commercial PDMS fiber. It was observed that there was a 50 °C difference of desorption temperature between the PDMS polymer and the C18 bonded silica. The desorption temperature indicated that the analyte interaction with the C18-bonded silica was stronger than that with the PDMS polymer. Rodrigues et al. (2002) coated SPME fibers with modified silica particles bonded to methyl (C1), hexyl (C6), octyl (C8), and polymeric octadecyl (C18) groups. The fiber showed good selectivity for PAHs. The fibers could be used for up to 100 extraction cycles without a significant reduction of extraction performance. However, because the fibers were very brittle the lifetime of the fibers was mainly limited by breakage. When these porous layer fibers were subjected to GC at 300 °C, some extraneous peaks were observed. This is probably due to the release of some monomers in the epoxy glue used to bond the porous layer. Cai et al. (2003) have successfully prepared sol-gel dibenzo-18 crown-6 SPME fiber coating by sol-gel technology and applied for the extraction of aliphatic amines. Compared to commercial PDMS and PA fiber, the sol-gel dibenzo-18 crown-6 is more sensitive and selective for derivatized aliphatic amines. The performance of the coating is stable at high temperature (< 350 °C) and it has a long lifetime (can be used < 150 times). 23 Djozan and Bahar (2004) have investigated the efficiency of polyaniline (PANI), coated gold wire, for use as a fiber for SPME. Aniline monomers were electropolymerized on gold wires by applying a constant current to an acetate buffer containing NaClO4 as supporting electrolyte. These fibers were capable of extracting some aliphatic alcohols from gaseous samples. Mohammadi et al. (2004) used the electrochemical fiber coating (EFC) technique for the preparation of dodecylsulfate-doped polypyrrole (PPy-DS), and applied it as a new fiber for SPME. PPy-Ds film was directly electrodeposited on the surface of a platinum wire from an aqueous solution containing pyrrole and sodium dodecylsulfate, using cyclic voltammetry (CV). The fiber coating can be prepared easily in a reproducible manner, it is inexpensive and has a stable performance at high temperatures (up to 300 °C). The new fiber was evaluated for the extraction of PAHs from water samples. Bagheri et al. (2005a) prepared three different coated fibers, polydimethylsiloxane (PDMS), polyethylenepropyleneglycol monobutyl ether (Ucon) and polyethylene glycol (PEG) based on sol-gel technology. For the first time, these fibers were evaluated for the SPME of semi-volatile drugs such as, dextromethorphan (DM) and dextrorphan (DP) in plasma. High-temperature conditioning of these fibers leads to consistent improvement in peak area reproducibility. An aniline-based polymer was electrochemically prepared by Bagheri et al. (2005b). This new fiber coating was applied for solid phase microextraction of some priority phenols from water samples. The polyaniline (PANI) film was directly electrodeposited on the platinum wire surface in sulfuric acid solution using cyclic voltammetry (CV) technique. This new coating can be prepared easily in a reproducible manner and it is rather inexpensive and stable against most of organic solvents. The thickness can be precisely controlled by the number of CV cycles. 24 Sun et al. (2005) have successfully developed activated carbon fiber (ACF) for use in SPME. The fiber was used to determine benzyl chloride, benzyl dichloride and benzyl trichloride in water samples by headspace analysis. Experiments showed that ACF has a higher adsorption capacity than the commercial coated fibers except for CAR-PDMS. The compounds studied could be desorbed completely within 60 s. The fairly well-distributed surface of ACF ensures complete desorption in a short time. 3-(Trimethoxysily) propyl methacrylate (TMSPMA) was first used as precursor as well as selective stationary phase to prepare the sol-gel derived TMSPMA-hydroxylterminated silicone oil (TMSPMA-OH-TSO) SPME fibers for the analysis of aroma compounds in beer by Liu et al. (2005c). The fiber was found to be very effective in carrying out simultaneous extraction of both polar alcohols and fatty acids and non-polar esters in beer. The fiber exhibited better sensitivity to most of the investigated analytes compared to commercial PDMS, PDMS/DVB and PA fibers. High thermal stability and long lifetime are also characteristics of the new fiber. SPME fiber coated with a novel γ-Al2O3 has been prepared and used to screen gaseous samples for traces of volatile organic compounds (VOC) by Wei et al. (2004). Results showed that it has good thermal stability (up to 350 °C), a long life span (< 180 times), higher extraction capacity and good selectivity for alkanes and esters. Physical characteristics showed that the surface of the fiber was stacked with fine particles of γAl2O3 and it was porous. It was reported that the diffusion into the porous layer might be responsible for the low diffusion kinetics which prolonged the equilibration time. Djozan et al. (2004) coated gold wire with polypyrrole (PPY) by electropolymerization for use in SPME. The adsorptive property of the coating was modified by doping with tetrasulfonated nickel phtalocyanine (NiPcTS). NiPcTS as a dopant enhances the sensitivity of PPY. The fiber was found to be capable of extracting BTEX compounds from water sample. Farajzadeh and Rahmani (2005) have successfully developed a new SPME fiber by electrolysis. The anode was oxidized to 25 copper (I) to produce copper (I) chloride as a sorbent for the studied amines on the copper wire. This fiber was capable of extracting the studied amine compounds and showed excellent fiber-to-fiber repeatability. It also exhibited lower detection limits and shorter analysis time. A new kind of membrane used for SPME was prepared with amine compounds by Yang et al. (2004). The self-prepared SPME membrane was proven to be effective for analysis of forensic toxic such as dichlorvos, morphine, Phenobarbital, methylamphetamine, MDA, ephedrine and TNT. Tsuda et al. (2004) prepared a new type of fiber adsorbent attached with silica microparticles. The silica microparticles were formed by the polymerization of silica oligomers on glass fibers, which were woven into a glass filter. The surface of the silica microparticles was chemically modified by bonding C18-ligands. Djozan and Assadii (2004) have conclusively demonstrated that pencil lead fiber modified by heating with water vapour can be used for precise analysis of PAH in water by SPME followed by GC-FID. Low cost and high temperature resistance are the main advantages of this fiber. It has characteristics required for use in routine analytical procedures. Djozan and Zehni (2003) investigated a simple and inexpensive method for the preparation of new SPME fiber consisting of copper wire coated with copper sulphide. The fibers seem very effective for extraction of aliphatic alcohols and amines from aqueous samples. Oliveira et al. (2005) successfully used a glass ceramic rod as a base for the preparation of SPME fibers using sol-gel technology. The glass-ceramic rod was coated with PDMS using sol-gel reaction. The film thickness for this new base was much bigger than the fused silica base. The characteristics of this new base can be attributed to its stronger exchange capacity. The glass ceramic fiber coated with PDMS by sol-gel reaction shows good extraction characteristics for BTEX compounds. 26 Co-poly (hydroxy-terminated silicone divinylbenzene) coating was first prepared by sol-gel and cross-linking methods and applied to the solid-phase microextraction of phosphate and methylphosphonate from and water by Liu et al. (2003). The new fiber possessed high thermal (to 380 °C) and solvent stability and a long lifetime. Compared to commercial SPME fibers, the sol-gel coated OH-TSO/DVB fiber showed better selectivity and sensitivity towards phosphate and methylphosphonate. Therefore, the novel fiber may be a good choice for analyzing warfare agents with SPME. Silveira et al. (2005) developed silica-based glass fibers and modified their surfaces with Nb2O5. Fiber glass with a composition of 29 % Li2O, 1 % ZrO2, 5% BaO and 65 % SiO2 was coated with Nb2O5 using the metallo-organic decomposition (MOD) technique. The Nb2O5-coated fibers showed excellent results both in the adsorption of phenols and in the adsorption and separation of alcohols using SPME. A novel alumina-based hybrid organic-inorganic sol-gel coating was first developed for SPME from a highly reactive alkoxide precursor, aluminium sec-butoxide and a sol-gel active polymer hydroxyl-terminated polydimethylsiloxane (OH-TSO) by Liu et al. (2006a). As compared to silica based hybrid materials, the ligand exchange ability of alumina makes it structurally superior extraction sorbents for polar compounds such as, fatty acids, phenols, alcohols, aldehydes and amines. The fiber was capable of extracting volatile alcohols and fatty acids in beer. Liu et al. (2006b) demonstrated a new coating technology for the preparation of SPME coating by sol-gel method. The extraction phase of PDMS containing 3 % vinyl group was physically incorporated into sol-gel network without chemical bonding. The extraction phase itself was then crosslinked at 320 °C, forming an independent polymer network and can withstand desorption temperature of 290 °C. Application to real life sample extraction showed that the SPME fiber is suitable for the determination of both volatile and semi-volatile compounds. 27 Gierak et al. (2006) successfully prepared new carbon fibers for SPME by the direct pyrolysis of methylene chloride on the quartz core. Owing to the high partition coefficients of the studied substances obtained on carbon fibers, it was possible to do the analysis of organic substances occurring in trace amounts in different matrices. The fiber was used for the analysis of benzene, toluene and xylene (BTX) contents in petrol. Cai et al. (2006) prepared three new crown ether SPME coatings using sol-gel method. In this work, three kinds of vinyl crown ethers with different cavity and benzyl substitutions were crosslinked with vinyl triethoxysilane under radical initiator. These new coatings showed higher extraction efficiency and sensitivity for organophosphorus pesticides (OPPs) compared with commercial fibers-85 µm PA and 65 µm PDMS/DVB. The benzo-15-crown-5 (B15C5) coating with larger polarity had the best selectivity for OPPs and it was successfully used for the determination of OPPs in juice, fruit and vegetables. Panavaite et al. (2006) developed a very simple method for the preparation of a new SPME fiber consisting of stainless steel wire coated with commercial silicone based high temperature resistant glue. The construction of the fiber is very simple and can be completed in a few minutes. The developed fiber was found to be mechanically stable and exhibited a relatively high thermal stability (up to 260 °C). The proposed SPME fiber offers and attractive alternative to commercially available PDMS based fibers for the extraction of non-polar or less-polar compounds. Anilinemethyltriethoxysilane (AMTEOS) was first used as precursor as well as selective stationary phase to prepare the sol-gel derived AMTEOS/PDMS SPME fibers by Hu et al. (2006). The novel SPME fibers exhibited higher extraction efficiency, good thermal stability, high porosity and longer lifetime compared with commercial SPME coatings. Owing to the introducing of the phenyl group in AMTEOS, the fiber offers better sensitivity towards aromatic compounds, such as monocyclic aromatic hydrocarbons (MAHs) and polycyclic aromatic hydrocarbons (PAHs). 28 1.8 Sol-gel Technology Sol-gel technology provides efficient incorporation of organic components into inorganic polymeric structures in solution under extraordinarily mild thermal conditions. In general, the sol-gel process, involves the evolution of inorganic networks through the formation of a colloidal suspension (sol) and gelation of the sol to form a network in a continuous liquid phase (gel). Through this process, homogeneous inorganic oxide materials with desirable properties of hardness, chemical and thermal resistance, polarity and tailored porosity can be produced at room temperatures (Brinker and Scherer, 1990; Klein, 1988). The precursors for synthesizing these colloids consist of a metal or metalloid element surrounded by various reactive ligands. A colloid is a suspension in which the dispersed phase is so small (~ 1-1000 nm) that gravitational forces are negligible and interactions are dominated by short-range forces, such as van der waals attraction and surface charges. An alkoxy is a ligand formed by removing a proton from the hydroxyl of an alcohol, as in methoxy (•OCH3) or ethoxy (•OC2H5). Metal alkoxides are members of the family of metallorganic compounds, which have an organic ligand attached to a metal or metalloid atom. Metal alkoxides are most popular because they react readily with water. The most widely used metal alkoxides are the alkoxysilanes, such as tetramethoxysilane (TMOS) and tetraethoxysilane (TEOS). However, other alkoxides such as aluminates, titanates and borates are also commonly used in the sol-gel process, often mixed with (TEOS) (Dietz and Camara, 2005; Hofacker et al., 2002). At the functional group level, three reactions are generally used to describe the sol-gel process as shown in Figure 1.2; hydrolysis (R1), alcohol condensation (R3) and water condensation (R2). The hydrolysis reaction (R1) replaces alkoxide groups (OR) with hydroxyl groups (OH) through the addition of water. Subsequent condensation reactions (R2 and R3) involving the silanol groups (Si-OH) produce siloxane bonds (SiO-Si) plus the by-products water or alcohol. Under most conditions, condensation 29 commences before hydrolysis is complete. However, conditions such as pH, H2O/Si molar ratio (R) and catalyst can force completion of hydrolysis before condensation begins. Although hydrolysis can occur without addition of an external catalyst, it is most rapid and complete when they are employed. Mineral acids, for example HCl or ammonia are most generally used, however, other catalysts such as acetic acid, KOH, amines, KF and HF are also employed. It has been observed that the rate and extent of the hydrolysis reaction is most influenced by the strength and concentration of the acid or base catalyst (Brinker and Scherer, 1990; Brinker et al., 1992; Klein, 1988). Figure 1.2: Basic steps in sol-gel coating technology (Brinker and Scherer, 1990) Additionally, because water and alkoxides are immiscible, a mutual solvent such as an alcohol is utilized. With the presence of alcohol as the homogenizing agent, hydrolysis is facilitated due to the miscibility of the alkoxide and water. As the number of siloxane bond increases, the individual molecules are bridged and jointly aggregate in the sol, which is a colloid suspension of solid particles of a diameter of hundred nanometers in a liquid phase. When the sol particles aggregate, or cross-link into a network, a micromolecular gel is formed. Upon drying, trapped volatiles are then driven off (Brinker et al., 1992). 30 1.8.1 Sol-gel Method as a New Tool for SPME Fiber Coatings Sol-gel technology offers a simple and convenient pathway for the synthesis of advanced material systems and for applying them on as surface coatings. Many different combinations of alkoxides and organoalkoxysilanes have been prepared to produce coatings with improved mechanical properties such as abrasion resistance and high chemical stability. However, the characteristics and properties of a particular sol-gel coating are strongly related to a number of factors, such as, pH, H2O/Si molar ratio (R), nature and concentration of catalyst, ageing temperature and time and drying during the synthesis of the coating (Dietz and Camara, 2005; Arkles, 2001). Dip coating technique is commonly used in preparing sol-gel coated fibers. The substrate to be coated is immersed in sol-gel solution and withdrawn with a well defined withdrawal speed at atmospheric condition. The stages of dip coating process includes dipping of the substrate (fused silica fiber) into the coating solution, wet layer formation by withdrawing the substrate and finally gelation of the layer by solvent evaporation as shown in Figure 1.3 (Brinker and Scherer, 1990). During the withdrawal of the substrate from the solution, the inner layer moves upward with the substrate, while the outer layer is returned to the solution. The thickness of the deposited film is related to the position of the streamline dividing the upward and downward - moving layers. The faster the substrate is withdrawn, the thicker the deposited film. Evaporation is generally relied upon to solidify the coating. The most significant factor in the rate of evaporation is the rate of diffusion of the vapor away from the film surface. The evaporation rate can be controlled by the deposition ambient. Since condensation continues to occur during sol-gel coating formation, the relative condensation and evaporation rates will dictate the extent of further cross-linking that accompanies the deposition and drainage stages (Atanacio et al., 2005; Brinker et al., 1992; Klein, 1988). 31 Syringe holder Fused silica fiber Coated layer Coating solution Figure 1.3: The stages of dip coating process (Brinker and Scherer, 1990) Among many inherent advantages of sol-gel technology are its unique ability to achieve molecular level uniformity in the synthesis of organic-inorganic composites, strong adhesion of the coating to the substrate due to chemical bonding and high thermal stability which overcomes the sample carryover problem in SPME-GC analysis (Hofacker et al., 2002; Kabir et al., 2004; Li et al., 2004a; Li et al., 2004b; Lopes and Augusto, 2004; Yu et al., 2004). 1.9 Problem Statements and the Needs of Study Interpretation of forensic evidence and in particular, the ability to uniquely detect accelerant is a crucial challenge in the scientific investigation of arson. Limitations in this area are posing a significant impact on police investigations and the successful prosecution of arson cases. An improvement in accelerant extraction came with the development of SPME technique as a viable standard method in arson analysis. The qualities that enable an SPME adsorbent to be successfully used for accelerant extraction and analysis are its selectivity towards accelerant components 32 which separates and concentrates the accelerant from the headspace to yield a sample that is suitable for introduction to GC-FID. Commercial SPME adsorbent have generally good extraction properties, however some aspects of their performance need to be improved to overcome the limited use of SPME in gas chromatographic applications (Dolan, 2003). Although a number of adsorbents are commercially available for SPME but some analytical methodologies might demand specific properties for extraction of selected compounds, special coatings that have a particular volume and selectivity towards particular analytes. Commercially available fibers are rather expensive and have a short lifetime (40-100 times) (Wang et al., 2003; Hun et al., 2005; Zhou et al., 2005). These fibers have relatively low recommended operating temperatures and generally remain in the range of 200-270 °C (Chong et al., 1997; Wang et al., 2000; Huo et al., 2004; Ugur et al., 2004; Yu et al., 2004; . Zuin et al., 2004; Hun et al., 2005; Liu et al., 2005a; Zhou et al., 2005). Poor thermal stability leads to incomplete sample desorption and sample carry-over problem (Liu et al., 2005b). Conventional fibers have poor solvent stability. The equilibrium and speed of the extraction process are related to physico-chemical characteristics of the sorbent, this limits the extent of selectivity obtainable using commercial fibers (Yun, 2003). Improved mechanical and chemical resistance is in an increasing demand in coating applications such as SPME. Recently, many novel coatings have been developed using different techniques and technology for use in SPME. However, up to now, none of the novel fibers have been evaluated for the determination of accelerants in arson analysis. 33 1.10 Objectives and Scope of Study The objectives of this study are as follows: i. To investigate suitable SPME adsorbents for arson analysis. ii. To evaluate the home-made SPME fiber against commercially available fibers iii. To apply the developed fiber for detection of accelerants in fire debris. In this study, SPME fiber coated with alternative materials was prepared by using sol-gel technology. The scope of the research is limited for the extraction of common accelerants in simulated arson samples. All the extractions in this study were performed using headspace –SPME mode. 34 CHAPTER 2 EXPERIMENTAL 2.1 Introduction This chapter covers the chemicals, materials and apparatus used in this study. It also includes general experimental work in the development, characterization and evaluation of the lab-made SPME fiber for the detection of accelerants in arson samples. 2.2 Chemicals and Materials The fiber coating material, n-octyltriethoxysilane (C8-TEOS), the precursor, methyltrimethoxysilane (MTMOS) and the end capping reagent, trimethylmethoxysilane (TMMOS) were obtained from Fluka Chemika, Switzerland. Table 2.1 lists the names, functions and chemical structures of the principal ingredients of the sol-gel coating solutions. Analytical grade n-hexane was obtained from J. T. Baker, New Jersey, (U. S. A). while hydrochloric acid (HCl) and HPLC grade methanol were obtained from BDH Chemicals, Poole, England. Distilled deionized water was prepared in the laboratory, Department of Chemistry, UTM. Individual standards of n-alkanes (C8, C9, C10, C12, C13, C14, C15 and C16) were purchased from Fluka Chemika. Samples of diesel and unleaded gasoline were purchased from a petrol station in Skudai, Johor while kerosene was obtained from a grocery shop at Taman Universiti, 35 Skudai, Johor. Samples of carpet were purchased from a carpet retail shop in Taman Ungku Tun Aminah, Skudai, Johor. Table 2.1: Names, functions and chemical structures of the principal compounds of solgel coating solutions Ingredient Methyltrimethoxysilane (MTMOS) Function Sol gel precursor Chemical structure CH3 H3CO Si OCH3 OCH3 n-Octyltriethoxysilane (C8-TEOS) Coating stationary phase [CH3(CH2)6CH2]n CH3CH2O Si OCH2CH3 OCH2CH3 Trimethylmethoxysilane (TMMOS) End-capping reagent CH3 H3C Si CH3 OCH3 Hydrochloric Acid (0.1 M) Acid catalyst Methanol Solvent HCl CH3OH 36 2.3 Apparatus Two glass apparatus (400 cm3 and 125 cm3 ) for sample preparation step of headspace SPME was specially designed (Yong, 2004). SPME holder and commercially available PDMS/DVB fibers purchased from Supelco, Bellefonte, Pennsylvania, U.S.A was used for the headspace technique. A 10 µL syringe used for direct injection method was obtained from Hamilton Co, Reno, Nevada, U.S.A. Used commercial SPME fibers were employed for the coating process. Eppendorf polypropylene micro centrifuge tubes (2 mL) used to place the sol-gel solution prior to centrifuge were obtained from Hamburg, Germany. Aluminium stubs used to place fiber samples prior to electron microscopy experiments were provided by Materials Science Lab, Faculty of Mechanical Engineering, UTM. A hotplate/stirrer Model HS 0707V2 from Favorit, United Kingdom was used to heat the samples. An SGT capillary column cutter with a rotating diamond blade from Shortix, Holland was used to cut the coated fibers for SEM analyses. Other apparatus used in this study included micropipettes Model 5000DG Nichipet (20-200 µL) and (0.5-10 µL) from Japan, vials, teflon magnetic stirring heads, dessicator, aluminium foil, double edged carbon tape (for placing SPME coated fibre on aluminium stubs) and a cigarette lighter (used to remove polyimide protective layer from the fiber). An illustration of the experimental setup for headspace SPME is shown in Figure 2.1. Prior to experiments, the glass apparatus were silanized to prevent the polar analytes from being adsorbed onto the glass wall. The sample apparatus was sealed with a septum to prevent sample evaporation. Headspace was created by partially immersing the sample apparatus containing analyte into a hot water bath. Subsequently, the SPME needle was pushed through the sub seal rubber septum (1.5 cm diameter) and the fiber was exposed in the gas phase (headspace). The hot water bath was stirred in order to homogenize the heat transfer and a thermometer was used to monitor the temperature during the extraction. After the extraction, the fiber was withdrawn into the needle and 37 removed from the sample matrix. The fiber was then immediately inserted into the heated GC injector port for subsequent analyte desorption. Thermometer SPME Holder Sub seal septum Sample preparation bottle Exposed fiber Spiked carpet sample Magnetic bar Hotplate/stirrer Figure 2.1: Experimental setup for headspace SPME accelerant extraction 2.4 Instrumentation Gas chromatography analyses were conducted using a Hewlett-Packard 6890 GC (Wilmington, Delaware, U.S.A.). The HP 6890 gas chromatograph was equipped with FID and a HP ChemStation for data processing. An Ultra-1 capillary column (Agilent) of dimensions 25 m x 0.20 mm x 0.11 µm film thickness was used. Helium was used as the carrier gas at a flow rate of 1.2 mL/min. The injection port temperature was set at 250 °C and FID temperature at 310 °C. SPME injections were performed using a split mode injection (5:1). Figure 2.2 shows the oven temperature program applied for Ultra 1 column. Studies of fiber coating structure and thickness were made by means of Philips Scanning Electron Microscope Model XL 30 SEM (Philips Electronic Instruments Company, Mahwah, New Jersey) equipped with a ThermoNoran energy dispersive Xray detection system (EDX). The fourier transform infrared (FTIR) analysis was 38 performed using a FTIR-8300 spectrometer (Shimadzu Corporation, Kyoto, Japan). A Hettich Zentrifugen, Centrifuge Universal 32R was used to separate the sol solution from the precipitate (if any). 2.5 Preparation of the Sol-gel Derived C8-coated Fiber Generally, the preparation of the solid phase microextraction fiber by sol-gel coating method includes the steps of providing the fiber structure, providing a sol-gel solution comprising a sol-gel precursor, an organic material with at least one sol-gel active functional group, a sol-gel catalyst, an end-capping reagent and a solvent system. Preferably, the sol-gel precursor includes an alkoxy compound. The organic material includes a monomeric or polymeric material with at least one sol-gel active functional group. The sol-gel catalyst is taken from the group consisting of an acid, a base or a fluoride compound (Wu et al., 2001). In this study, the sol-gel solution was prepared by mixing an appropriate amount of C8-TEOS, MTMOS, methanol, 0.1M HCl and distilled de-ionized water. The sol-gel solution was stirred for 4 hours at room temperature. The mixture was then transferred into an Eppendorf microcentrifuge tube and centrifuged at 13,000 rpm for 5 min. The precipitate at the bottom of the tube, if any, was removed and the top clear sol solution was used for fiber coating. Prior to sol-gel coating, the protective polymeric layer was removed from a 1 cm segment of the used conventional fiber at its end. This was accomplished by burning off the polyimide protective layer using a cigarette lighter. The exposed fused silica surface (burnt section of the fiber) was pretreated by washing it with excess methanol and water. It was then air dried at room temperature. The pretreated portion of the fiber (bare outer surface of the fiber end) was then reacted with sol-gel solution under controlled conditions to produce a surface bonded sol-gel coating on the portion of the fiber. This was done by dipping the bare tip of fused 39 silica vertically into the sol solution. It was held inside the sol solution for ~20 min. The coating process was repeated three times, using a freshly prepared sol solution each time. This was followed by the end-capping process. The coated fiber was end-capped by dipping it into the mixture of TMMOS/methanol solution. After 1 min, the fiber was removed from the end-capping solution and placed in a dessicator at room temperature. Before use, the fiber was conditioned at 250 °C under helium for approximately 1-2 hours in the GC injection port. After removal from the injector, the fiber was cooled to room temperature. The conditioning cycle was repeated a few more times until a stable GC baseline was obtained. The fiber was then ready for SPME-GC experiments. 2.6 Preparation of Standard Stock Solution A concentration of 20 % (v/v) standard stock solution of n-alkanes was preparaed by mixing C8, C10, C12, C14, and C16 hydrocarbons in equal volume into a 25 mL volumetric flask. Another standard solution of n-alkanes was prepared by mixing C9, C13, and C15 hydrocarbons in equal volume into a 25 mL volumetric flask. This resulted in a 33.3 % (v/v) concentration stock solution. Both prepared stock solution represented the hydrocarbon compounds commonly found in gasoline, kerosene and diesel. A 1.0 µL aliquot of both the standard mixture was then introduced into the GC injection port by direct injection. 2.7 Procedure for Extractions using Headspace SPME (HS-SPME) In order to extract standard hydrocarbon compounds using the lab-made C8- coated fiber, 30 µL from the prepared standard solution of n-alkanes (C8, C10, C12, C14, and C16) was placed in the sample preparation apparatus which was immersed in a hot water bath and heated for 20 min at 100 °C (this initial experimental conditions were 40 chosen according to the previous study done in this lab by Yong, 2004). The C8-coated fiber was exposed in the headspace for an appropriate time. Finally the exposed fiber was retracted and the fiber extracts were analyzed using GC-FID. The oven temperature was initially set at 40 °C, programmed at a rate of 10 °C/min until a final temperature of 270 °C. Blank runs of the fiber were performed prior to each extraction to ensure that there were no carryover of analytes from previous extraction. The HS-SPME procedure was repeated using commercially available PDMS/DVB fiber for comparison. For the extraction of accelerants using C8-coated fiber, 30 µL of neat accelerant (petrol, kerosene or diesel) was individually placed in the sample preparation apparatus. The extraction steps mentioned above were employed. The same HS-SPME procedure was carried out again using spiked fire debris sample for the determination of accelerants in simulated arson samples. Blank analysis of the fire debris sample was carried out before the recovery experiments. 2.8 Preparation of Spiked Fire Debris Sample A sample of carpet (20 cm × 13 cm) placed on a sheet of aluminium foil was ignited with a fire starter and left to burn until about one-third remained on the aluminium foil as shown in Figure 2.2 (a). Fire was extinguished by cutting off the oxygen supply. The partially burnt carpet was then exposed to the surrounding air for 30 minutes to let it cool down [Figure 2.2 (b)]. 100 µL of neat accelerant (petrol, kerosene or diesel) was individually spiked onto the burnt carpet fire debris sample. The spiked sample was placed in the glass apparatus prior to extraction using HS-SPME. The extracted analytes were then analyzed by GC-FID. Blank samples of unspiked burnt carpet samples were prepared in a similar way. 41 (a) (b) Figure 2.2: Photographs showing (a) the burning of carpet sample and (b) the burnt fire debris sample 2.9 General Procedures for Characterization of the C8-coated Fiber 2.9.1 Lifetime of the Coating 30 µL from the prepared standard solution of n-alkanes (C9, C13, and C15) was placed in the HS-SPME sample preparation apparatus. The extraction was conducted and the analytes were desorbed in GC-FID as described in Section 2.7. In order to determine the life span of the coated fiber, the extraction and desorption cycle was repeated 200 times under the same HS-SPME and GC operating conditions. The fiber’s life span was studied by monitoring the change of its extraction ability after it had been used for 50, 100, 150 and 200 times. 42 2.9.2 Thermal Stability The effect of conditioning temperature on the stability of the C8-coated fiber was evaluated by subjecting the fiber to successively higher conditioning temperatures, 270 ° C and 300 °C respectively for 1 hour prior to extraction. The same HS-SPME extracting conditions described in Section 2.7 were applied. Triplicate GC analyses were performed for each conditioning temperature. 2.9.3 Scanning Electron Microscopy Scanning Electron Microscopy analysis was carried out to investigate the structure of the coated surface and estimate the thickness of the coating. The 1 cm segment of the coated fiber end was cut using a diamond cutter. The fiber was then placed on an aluminium stub with the help of a double edged carbon tape. The stubs were placed in a dessicator prior to SEM analysis. The SEM was carried out by the bombardment of electrons of 30 KeV on the target sample (C8-fiber). (100-4000) times magnification of the fiber was observed using the scanning electron microscope. In order to estimate the film thickness and determine the reproducibility of the coating thickness, five C8-coated fibers were prepared in a similar way as described in Section 2.5. All the fibers were subjected to SEM analysis under the same condition. The film thickness was estimated from the cross-sectional view of the fiber. 2.9.4 Calculation of Fiber Coating Volume The coating volume of both commercial PDMS/DVB fiber and lab–made C8coated fiber were determined. The extracting phase volume (Vf) can be calculated by the following equation (Perez et al., 2001): 43 Vf = π L (e2 + ae) Where L: The fiber coating length (1.0 cm) e: The film thickness of the coating polymer a: The diameter of the fused silica base (110 µm) (2.1) 2.9.5 FTIR Analysis The sol-gel derived C8-coated fibers were conditioned at 300 °C under helium protection for 2 hr and then dipped in methylene chloride for 2 hr prior to IR analysis. A section of the coating was harvested with a razor blade from the pretreated fiber, then ground and blended with potassium bromide (KBr). The KBr pellet spectra of the coatings were acquired with air as background at a resolution of 4 cm-1 over the full midIR range (4000-400 cm-1). IR spectrum of pure C8-TEOS (fiber coating material) was recorded by liquid film method over the same frequency region and at the same resolution time for comparison. 2.10 Procedure for Optimization of Sol-gel Process Optimization of the sol-gel process was affected by varying the molar ratios of the silane precursors. Optimization of the sol-gel process was done by preparing three different fibers from sol-gel solutions containing C8-TEOS to MTMOS in the ratio of 0.5:1, 1:1 and 2:1. The fiber preparation method described in Section 2.5 was employed. HS-SPME extractions of hydrocarbons (Section 2.7) was done in triplicate for each fiber. The extracted analytes were then thermally desorbed at the GC injection port. 44 2.11 HS-SPME Optimization and Evaluation Procedures 2.11.1 Optimization of HS-SPME Operating Conditions 30 µL from the prepared standard solution of n-alkanes (C8, C10, C12, C14, and C16) was placed in the sample preparation apparatus and followed by HS-SPME extraction. When the desired fiber exposure time was reached, the fiber was retracted and the analytes were thermally desorbed in GC injection port. Several variables such as fiber exposure time, desorption time and extraction temperature have to be taken into account for the optimization of the proposed HSSPME-GC method using C8-coated fiber. A set of experiments herein performed in a stepwise fashion served to optimize the experimental parameters. The variables were optimized by employing the procedures mentioned above. Triplicate analyses were carried out for each parameter. The fiber exposure time was preliminary optimized by keeping the extraction temperature (80 °C) and desorption time (80 sec) constant. The analyte desorption time at the GC injection port was then optimized by keeping the extraction temperature (80 ° C) and extraction time (15 min) constant. Finally the extraction temperature was optimized by keeping the fiber exposure time (15 min) and desorption time (80 sec) constant. The variables optimized were in accordance to Table 2.2. 45 Table 2.2: Optimization of HS-SPME experimental parameters [(a) exposure time, (b) desorption time and (c) extraction temperature] (a) (b) (c) Exposure Time (min) 5 10 15 20 25 30 35 Desorption Time (sec) 80 80 80 80 80 80 80 Extraction Temperature ( °C) 80 80 80 80 80 80 80 Desorption Time (sec) 10 20 30 40 50 60 70 80 90 Exposure Time (min) 15 15 15 15 15 15 15 15 15 Extraction Temperature ( °C) 80 80 80 80 80 80 80 80 80 Extraction Temperature ( °C) 40 50 60 70 80 90 100 Exposure Time (min) 15 15 15 15 15 15 15 Desorption Time (sec) 80 80 80 80 80 80 80 46 2.12 Method Validation 2.12.1 Determination of Limits of Detection (LOD) The detection limits of fire accelerants were investigated using the lab-made C8coated fiber. Individual accelerants (petrol, kerosene and diesel) were spiked onto a blank burnt carpet sample in decreasing amount (initially 1 µL). Similar fire debris sample preparation as described in Section 2.8 was employed. The HS-SPME extractions were carried out with the optimized conditions of the C8-fiber. Blank samples were also extracted in order to differentiate the burnt matrix profile in the presence of accelerants. 2.12.2 Preparation of Calibration Graph A series of working solutions were prepared in individual volumetric flask (5mL) by diluting the standard stock solution with n-hexane. The resultant concentrations were a series of 2.00, 4.00, 6.00, 8.00 and 10.00 % (v/v) standard mixture. The HS-SPME extractions were carried out with the optimized conditions of the C8-fiber. Triplicate analyses were performed at each concentration level. The calibration graph was constructed by comparing average peak area counts against analyte concentrations. 47 CHAPTER 3 CHARACTERIZATION OF THE C8-COATED FIBER, OPTIMIZATION AND UNDERLYING MECHANISM OF SOL-GEL PROCESS 3.1 Introduction In this chapter, the characterization of the lab-made sol-gel derived C8-coated fiber is described. The optimization of sol-gel process and the chemistry behind the solgel coating technique involved in this study were also discussed. 3.2 Characterization of the C8-coated Fiber 3.2.1 Selectivity for Standard Hydrocarbons In order to examine selectivity of the lab-made SPME fiber towards hydrocarbon compounds, a mixture of n-alkanes (C8, C10, C12, C14 and C16 ) were subjected to headspace SPME using sol-gel derived C8-coated fiber and the GC profiles were compared with that from direct injection. The GC profiles obtained from headspace SPME using C8-coated fiber was comparable with the profiles of hydrocarbon standards from direct injection. There was a significant comparison of carbon compounds in the similar retention time ranging from 4-18 min. There was no significant differences of hydrocarbon peaks and all the hydrocarbon components were well separated as shown in Figure 3.1. 48 Figure 3.1: SPME-GC analysis of hydrocarbon compounds using C8-coated fibers Peaks: (1) C8, (2) C10, (3) C12, (4) C14 and (5) C16 n-alkanes. 3.2.2 Extraction Capability for Standard Hydrocarbons The extraction capability of the sol-gel derived C8-coated fiber for hydrocarbons was determined by comparing it with the extraction capability of commercially available PDMS/DVB fiber. PDMS/DVB fiber was selected for comparison because previous work done in this lab (Yong, 2004) proved that the fiber has the highest sensitivity towards hydrocarbon compounds. Figure 3.2 shows the comparison of extraction capability between C8-coated fiber and commercially available PDMS/DVB fiber. The C8-coated fiber exhibited a slightly higher extraction capability for all the hydrocarbon compounds by contrast with conventional PDMS/DVB fiber. A higher extraction capability yielded by C8-coated fiber could be due to the existence of higher surface area for the C8-coated fibers. The result obtained is in good agreement with the previous work done by Gbatu et al. (1999). 49 Based on the previous studies, almost all the fibers prepared by sol-gel technology gave superior extraction capability when compared to commercially available fibers. Zuin et al. (2004) have stated that the extraction capacity of sol-gel PDMS/PVA fiber for pesticides was higher than that of commercially available PDMS fiber. Liu et al. (2005a) reported that sol-gel derived acrylate/silicone fiber yielded higher extraction efficiency to 2-chloroethyl ethyl sulfide in soil compared with conventional polyacrylate (PA) fiber. Works done by Cai et al. (2003) showed that the selectivity and extraction capability of the sol-gel polymethylsiloxane fiber for organochlorine pesticides was higher than that of conventional PDMS fiber. Figure 3.2: Comparison of extraction capability between C8-coated fiber and PDMS/DVB fiber. 3.2.3 Lifetime of the Coating A coating’s lifetime is important for practical application (decline of efficiency with the number of analysis). The long lastingness of the coated fiber was determined by 200 continued operations carried out with the same C8-coated fiber and oven temperature program. The C8-coated fiber have been used for hydrocarbon extractions 50 (C9, C13, and C15) and subjected to GC for more than 200 times. There was no significant differences of hydrocarbon peaks obtained in each operation. Referring to Table 3.1, all the hydrocarbon compounds gave a low relative standard deviation (RSD) value ranging from (3.8-5.4) % which shows an acceptable reproducibility. This proves that the coated surface of the fiber was not partially depleted during the continued operation. It was still stable and reusable. Such a long service life are possibly due to the strong chemical bonding between the sol-gel generated C8-coated composite coating and the silica surface (Chong et al., 1997; Zeng et al., 2001; Liu et al., 2003). Table 3.1: The extraction precision of C8-coated fiber (200 operations) Hydrocarbon compounds C9 Extraction Period 50th Average RSD Peak (%) Area (n=3) 176.5 4.9 100th Average RSD Peak (%) Area (n=3) 153.4 5.1 150th Average RSD Peak (%) Area (n=3) 238.1 5.3 200th Average RSD Peak (%) Area (n=3) 204.9 5.4 C13 340.7 3.8 433.5 4.2 488.2 3.9 404.8 4.3 C15 418.2 4.6 501.3 4.9 490.7 5.0 523.6 4.8 The result obtained is comparable with those obtained by other researches. A single sol-gel PDMS/PVA fiber was recorded to be used for more than 300 extractions during a time range of more than five months before degradation of the coating (Zuin et al., 2004). Liu et al. (2006) reported that the extraction efficiency of sol-gel aluminahydroxyl-terminated polydimethylsiloxane coating had no obvious decline after the fiber was run at least 160 sorption/desorption processes. Zeng et al. (2001) reported that the extraction ability of sol-gel hydroxydibenzo-14-crown-4 fiber showed no significant decline after it has been used 150 times. Works by Wang et al. (2000), proved that sol- 51 gel polyethylene glycol coated fiber was still reusable although it has been used for more than 150 extractions. 3.2.4 Thermal Stability Thermal stability is a crucial property of SPME coatings in practical applications. Good thermal stability will reduce the possibility of polluting the detector and expand the scope of applications of SPME. The effect of conditioning temperature on the stability of the C8-coated fiber was determined by conditioning the fiber at high temperatures (270 and 300) °C for 1 hour prior to extraction. It is apparent from Table 3.2 that the extraction capacity of the C8-coated fiber was not significantly affected by the temperature used for thermal conditioning. High temperature conditioning lead to consistent improvement in peak area repeatability for SPME-GC analysis. The RSD value of < 4.6 % can be routinely obtained for hydrocarbons on C8-coated fibers conditioned at 270 °C and 300 °C. There was a slight improvement of the peak area reproducibility achieved by using higher fiber conditioning temperature (300 °C). The C8-coated fiber can be routinely used at 300 °C without any sign of bleeding, whereas for commercial PDMS/DVB fiber, the highest temperature the coating layer can endure is less than 280 °C, and thus the range of analyte molecular weights that can be handled by SPME-GC is limited. Enhanced thermal stability of sol-gel coated fiber allows one to overcome the sample carryover problem often encountered in SPME and expand the SPME application range towards higher boiling point. Such a high thermal stability might be due to the strong chemical bonding between the sol-gel generated composite coating and the silica surface (Chong et al., 1997; Yu et al., 2002; Yun, 2003; Cai et al., 2003; Yu et al., 2004). Therefore sol-gel C8-fiber provided efficient release of extracted analytes without negative consequences from the use of high temperatures. 52 The high thermal stability of C8-fiber determined in this study is similar to those obtained by other researchers using different sol-gel fibers. Chong et al. (1997) reported that sol-gel PDMS fiber can be routinely used at 320 ˚C without any sign of bleeding. Liu et al. (2003) reported that sol-gel silicone divinyl benzene fibers can be used at 380 ˚C. Works done by Zeng et al. (2001) proved that sol-gel hydroxydibenzo-14-crown-4 fiber exhibited high thermal stability (up to 340 ˚C). Table 3.2: Reproducibility of the peak area after conditioning at a) 270 ˚C and b) 300 ˚C. a) C8-coated fiber conditioned at 270 ˚C Hydrocarbon Compounds C8 Average Peak Area (n = 3) 151.7 RSD (%) C10 248.1 4.5 C12 377.3 4.1 C14 470.3 3.9 C16 366.6 4.4 3.7 b) C8-coated fiber conditioned at 300 ˚C Hydrocarbon Compounds C8 Average Peak Area (n = 3) 174.9 RSD (%) C10 280.3 3.8 C12 417.4 2.9 C14 620.3 3.3 C16 514.7 3.6 2.9 53 3.2.5 Scanning Electron Microscopy (SEM) Analysis 3.2.5.1 Surface Characteristics of the Coating The morphology of the lab-made fiber was investigated by the scanning electron microscope. Figure 3.3 (a) and (b) represents the surface view of sol-gel C8 coated fiber at 150 and 1000 fold-magnification respectively. An enlarged scanning electron micrograph of the fiber surface viewed at 100 fold magnification is shown in Appendix A. Figure 3.4 represents the cross-sectional view of the chemically bonded polymeric layer of C8 coating at 4000 fold-magnification. As can be seen from Figure 3.3 (a), it is obvious the surface of the fiber coating was well-distributed. A homogeneous porous structure was suggested for sol-gel C8-coating [Figure 3.3 (b)]. However, the SEM technique does not possess enough resolution to show clearly both micro- and mesopores, since they fall in the (2-500) Å range. Similar unsatisfactory resolution has been repoted by Azenha et al. (2005) for sol-gel PTMOS fiber. Chong et al. (1997) and Yu et al. (2004) have reported that porous structure of the sol-gel coating should significantly increase the surface area on the fiber. Consequently, with such a porous coating structure, even an apparently thinner coating will be able to provide enhanced stationary-phase loadings and therefore, high fiber sample capacity. Wang et al. (2003) and Zeng et al. (2001) have stated the porous structure of sol-gel coating helps faster mass transfer during extraction and therefore the equilibration time is shorter. Figure 3.4 reveals that the fiber was not uniformly coated. This suggestion is agreeable to those reported by Gbatu et al. (1999). Examination of the scanning electron micrograph of the fiber prepared from 1:1 molar ratio of C8-TEOS:MTMOS showed no apparent cracks (Figure 3.3 (a) and Appendix A). This is in good agreement with the previous work done by Gbatu. Works done by Gbatu et al. (1999), showed that no apparent crack was visible for the fiber prepared from 1:1 molar ratio of C8-TEOS:MTMOS but the cracks were visible in the fiber coated with a higher molar ratio (2:1). This could be due to the presence of the 54 bulky C8-TEOS group in a very large amount. Therefore all the analyses in this study were performed using fibers prepared from a 1:1 molar ratio. (a) (b) Figure 3.3: The surface view of sol-gel C8 coated fiber at (a) 150 and (b) 1000 foldmagnification obtained by SEM. 55 Sol-gel derived C8-coating Fused silica base Figure 3.4: The cross-sectional view of sol-gel C8 coated fiber at 4000 fold magnification obtained by SEM. 3.2.5.2 Estimation and Reproducibility of the Film Thickness One of the most attractive features of sol-gel coatings for SPME is the thickness of the coatings. It is possible to obtain extremely thin films of sorbents with large extractive power. Such thinner coatings allows faster extractions and easier thermal desorption when compared to conventionally coated fibers (Zuin et al., 2004; Oliveira et al., 2005). The film thickness of five different lab-made C8 fibers were measured in order to estimate the thickness of C8 layer and determine the reproducibility of the coating thickness. As can be seen from Appendix B, the thickness of the coating varied along the length of the fiber. This result was expected and it was reported by Gbatu et al. (1999). 56 It has been reported that the varied thickness obtained might be due to the rate at which the fused silica tip is removed from sol-gel solution after coating is not controlled and the portions of activated fused silica fibers that are exposed to the sol-gel solution for coating are not uniform. Therefore the film thickness of each fiber was measured at two different points and the average coating thickness were calculated. Referring to Table 3.3, the average thickness of the C8 coating was estimated as being (~ 3-4) µm, which is considerably thinner than most conventional SPME fibers (65 µm for PDMS/DVB). The film thickness obtained in this study is comparable with previous works done by Azenha et al. (2005) which stated (~ 0.2-1) µm thickness for sol-gel derived phenyltrimethoxysilene(PTMOS) fiber, Lopes et al. (2004) reported ~ 5 µm thickness for sol-gel derived PDMS/PVA fiber and Basheer et al. (2005) obtained (~ 5-7) µm thickness for sol-gel amphilic and hydrophilic oligomers coated fibers. The reproducibility of the coating thickness obtained in this study is rather promising. Works done by Azenha et al. (2005) revealed that the film thickness reproducibility was unsatisfactory for sol-gel PTMOS fiber. Table 3.3: The approximate film thickness of five different C8-fibers Fiber Number Thickness Measured at Two Different Points (µm) Average Coating Thickness (µm) 1 3.895 4.063 ~4 2 3.498 3.921 ~4 3 3.287 4.652 ~4 4 2.576 2.824 ~3 5 2.528 3.247 ~3 57 3.2.5.3 Determination of Extracting Phase Volume The fiber coating volume need to be determined as it is one of the factor that effects the diffusion rate. Thicker coatings result in longer extraction time because diffusion is slow within the polymer extraction phase. Larger volume of extracting phase require longer equilibration times, thus the coating of choice is the thinnest one which will provide sufficient sensitivity for the determination (Mester et al., 2001). The extracting phase volume of both commercial PDMS/DVB fiber and labmade C8-coated fiber were calculated as described in equation 2.1 of Section 2.9.4. As the diameter of the fused silica base is 110 µm, the coating length is of 1.0 cm and the film thickness is of (~ 3-4) µm, the volume of extracting phase on the lab-made C8coated fiber is (~ 0.011-0.014) mm3. A 65 µm PDMS/DVB commercial fiber leads to an extracting phase volume of 0.357 mm3. It is apparent that the volume of extracting phase on the lab-made C8-coated fiber is considerably less than that of commercial PDMS/DVB fiber. Therefore the C8 coating has a larger affinity towards the analytes than regular PDMS/DVB. Similar result has been reported by Lopes et al. (2004) for sol-gel PDMS/PVA fiber. The PDMS/PVA has a coating thickness of approximately 5 µm, which leads to a volume of extracting phase of ~ 0.046 mm3. The extracting phase volume of the PDMS/PVA fiber is almost 1/3 of the coating volume of the 30 µm PDMS commercial fiber, 0.132 mm3. It has been reported that the PDMS/PVA allows faster extractions and easier thermal desorption as compared to conventional PDMS fiber. 3.2.6 Fourier Transform Infrared (FTIR) Analysis The IR spectra of sol-gel derived C8-coating and pure C8-TEOS is shown in Figure 3.5 (a) and (b) respectively. The Si-O group peak of C8-TEOS at 1390 cm-1 also 58 appeared in the sol-gel derived C8-coating. It most probably prove the successful binding of C8-TEOS (fiber coating material) to the stationary phase. The decreased intensity of the peak in sol-gel derived C8-coating indicated that it has been successfully bonded with other compounds. Figure 3.5 (a) : IR spectrum of pure C8-TEOS 59 Figure 3.5 (b) : IR spectrum of sol-gel derived C8-coating 3.3 Optimization and Possible Underlying Mechanism of Sol-gel Process 3.3.1 Optimization of Sol-gel Process In order to optimize the sol-gel process, fibers prepared from different molar ratios (0.5:1, 1:1; and 2:1) of C8-TEOS to MTMOS were used for extraction. A higher ratio of C8-TEOS:MTMOS resulted in higher extraction capability as shown in Figure 3.5. This is due to the content of C8-TEOS co–precursor, fiber coating material which serves as a hydrophobic extracting phase. The fiber prepared from 2:1 molar ratio of C8TEOS:MTMOS, displayed the highest extraction capability due to the presence of C8 groups in a very large amount. Whereas, the fibers prepared from 0.5:1 molar ratio, 60 yielded the lowest extraction capability possibly due to the low content of C8 groups. Fiber prepared from 2:1 molar ratio appeared to give higher extraction yields compared to that of the other molar ratios tested. However, fiber prepared from 2:1 molar ratio was not chosen for analysis because the scanning electron micrographs of the fiber were not good. Moreover, a higher amount of coating material (C8-TEOS) will be needed for the preparation of fiber with 2:1 molar ratio of C8-TEOS:MTMOS compared to fiber prepared with l:1 molar ratio. This may not be cost-effective. Average Peak Area 600 500 400 300 200 100 0 C8 C10 C12 C14 C16 Hydrocarbon Compounds 0.5:1 ; C8-TEOS:MTMOS 1:1 ; C8-TEOS:MTMOS 2:1 ; C8-TEOS:MTMOS Figure 3.6: Extraction capability of hydrocarbon compounds using fibers prepared from sol–gel solution containing C8-TEOS:MTMOS with the ratios of 0.5:1, 1:1 and 2:1. 3.3.2 Possible Underlying Mechanism of the Coating Process The in situ creation of the sol-gel C8 coating involved two major sets of reactions (Chong et al., 1997; Gbatu et al., 1999; Wang et al., 2000; Zeng et al., 2001; Liu et al., 61 2003; Li et al., 2004b; Lopes et al., 2004; Liu et al., 2005a): (1) hydrolysis of the precursor and (2) polycondensation of the hydrolyzed products. These reactions are catalyzed by acids or bases and lead to the formation of a polymeric network. In this study, MTMOS was used as a precursor and HCl as the catalyst. There are several steps during the sol-gel processing. The first step is the hydrolysis reaction of the precursor and co-precursor under the acid catalyst. The hydrolysis reaction of MTMOS precursor can be presented by the following equation: CH3 H3CO CH3 Si OCH3 OCH3 HCl HO + H 2O Si OH Catalyst MTMOS OH + CH3OH (3.1) In the sol gel solution, C8-TEOS is the coating stationary phase. However it also reacts as a co-precursor which contains hydrophobic alkyl chain (octyl group). The hydrolysis reaction of C8-TEOS co-precursor can be presented by the following equation: [CH3(CH2)6CH2]n CH3CH2O Si OCH2CH3 + H 2O OCH2CH3 C8-TEOS HCl Catalyst (3.2) [CH3(CH2)6CH2]n HO Si OH + CH3CH2OH OH The second step is that the hydrolyzed products undergo polycondensation reactions to produce a three-dimensional polymer network. The octyl groups of the C8- 62 TEOS co-precursor do not participate in the poly-condensation reactions, they are exposed and serve as the hydrophobic extracting phase. Polycondensation reactions involving the hydrolyzed MTMOS and C8-TEOS can be depicted by the following scheme: CH3 Si HO CH3 + OH m HO Si OH CH3 OH O Si ( O Si )m O HO OH O CH3 (3.3) [CH3(CH2)6CH2]n Si HO [CH3(CH2)6CH2]n OH p + HO OH OH Si OH [CH3(CH2 )6CH2]n HO (3.4) O Three dimensional network Si ( O Si )p O O [CH2(CH2)6CH3]n These reactions are followed by chemical binding of the coating stationary phase to the growing silica network. The chemical reaction involved can be schematically represented by the following equation. CH3 HO Si ( O Si )m O O [CH3(CH2)6CH2]n O CH3 Growing silica network + HO Si ( O O O Si )p O [CH2(CH2)6CH3]n 63 [CH3(CH2)6CH2]n CH3 O Si ( O Si )m O CH3 HO O Si ( O O (3.5) Si )p O [CH2(CH2)6CH3]n O The silanol groups on the fused-silica fiber surface can also join in the condensation reactions and provide chemical anchorage to the polymeric network in the immediate vicinity of the fiber surface (evolving sol-gel coatings). Schematically this reaction can be represented as follows: Si OH Si OH + HO Si OH CH3 O Si ( O Si )m O CH3 [CH3(CH2)6CH2]n O O Si ( O Si )p O O [CH2(CH2)6CH3]n Fused silica fiber surface Si OH CH3 O Si HO O Si ( O Si )m Si OH O CH3 [CH3(CH2)6CH2]n O O Si ( O Si )p O O [CH2(CH2)6CH3]n Sol-gel derived C8-polymeric coating (3.6) 64 Thus, a surface-bonded polymeric coating is created by dipping the exposed end of the fused silica fiber into the sol solution and allowing it to be there for a certain amount of time. The coated fibers were treated with a solution of TMMOS to reduce the silanol contents of the coating. Possessing a small molecular size, TMMOS molecules should have greater access to the porous structure of the coating. In parallel with C8-TEOS, it should be able to provide added deactivation to sol-gel coating. Deactivation of surfacebonded sol-gel C8-coating with TMMOS can be represented by the following scheme: Si OH OH Si HO O Si ( O Si )m Si OH O O O O OH Si ( O Si )p O O [CH2(CH2)6CH3]n + CH3 H3C Si CH3 TMMOS OCH3 CH3 H 3C Si CH3 CH2 Si OH O Si HO O Si ( O Si )m Si OH OH O O [CH3(CH2)6 CH2]n O O Si ( O Si )p O O [CH2(CH2)6CH3]n Deactivated, sol-gel derived C8 Coating (3.7) 65 C8-TEOS was added into this system not only to lengthen the silica network leading to the increased surface area of the fiber and impart hydrophobic character to the coating but also to help to spread the stationary phase on the fiber uniformly. Unlike the commonly used sol-gel precursors (tetraalkoxysilanes), MTMOS (a methyl derivative of the commonly used precursor tetramethoxysilane) is used. Such a choice aims at overcoming some inherent problems that are faced at the drying step of sol-gel processing. A sol-gel material, formed as a result of hydrolysis of alkoxysilane precursor and subsequent condensation of the hydrolyzed products, may undergo cracking and shrinkage during the drying step. They originate from the capillary thrust due to solvent evaporation from the gel pores (Chong et al., 1997; Gbatu et al., 1999; Wang et al., 2000; Hofacker et al., 2002). For the use in SPME, preservation of the structural integrity is necessary for the sol-gel coating to be able to provide desired material properties. Works by Mackenzie (1995) have proved that the sol-gel network originating from an alkyl derivative of a tetraalkoxysilane precursor, such as MTMOS, possess a more open structure and can more effectively release stresses during drying, minimizing the cracking tendency. Generally, sol-gel reactions in a silica-based system are rather slow and often require the use of catalysts to accelerate the process (Liu et al., 2006), therefore HCl was included in the preparation of sol sol-gel solution. It should be pointed out that the purpose of deactivation in SPME fiber technology is not identical with that in GC. In the preparation of a capillary GC column, the column surface deactivation is an important step. It is because that there are some silanol groups on the fused-silica capillary inner walls and the polar compounds are prone to undergo adsorption interactions with silanol groups (column surface interactions that are responsible for peak tailings and efficiency losses). Therefore in conventional column technology, deactivation is usually carried out as a separate step to derive the surface silanol groups. In the preparation of SPME fiber with sol-gel technology, solute interactions can be even advantageous. The silanol groups can be used as the chemical anchorage for the evolving sol-gel coatings. Such interactions will 66 provide enhanced selectivity for polar compounds in SPME. TMMOS can chemically bind with the silanol groups of the coating surface (Chong et al., 1997; Wang et al., 2000). Sol-gel coating technology is characterized by a higher degree flexibility in coating composition and selectivity. It should be noted that the surface coating obtained from sol-gel chemistry is not purely organic, it is composite in nature. The composition of this organic-inorganic coating can be controlled by varying the proportions of the sol solution ingredients. This gives the possibility of selectivity tuning in SPME fiber technology. In principle, by varying the composition of the sol solution, one should be able to prepare widely diverse surface coatings, ranging from purely inorganic to purely organic. Inorganic coatings pose great promise for SPME of volatiles (Chong et al., 1997; Yu et al., 2002; Liu et al, 2005b). 3.4 Conclusions A lab-made SPME fiber (sol-gel derived C8-coating) has been successfully developed, characterized and evaluated for the HS-SPME of standard hydrocarbon compounds. The lab-made fiber was capable of effectively extracting volatile hydrocarbon compounds. The lab-made fiber gave superior characteristics when compared to commercially available PDMS/DVB fiber. Compared with commercial PDMS/DVB fiber, the new lab-made fiber exhibited higher extraction capability for standard hydrocarbons, higher thermal stability (up to 300 °C) and longer lifetime (~ 200 times). Electron microscopy experiments revealed that the surface of the fiber coating was well-distributed but not uniformly coated and a porous structure was suggested for the sol-gel derived C8 coating with an approximate thickness of (3-4) µm. The volume of extracting phase on the lab-made C8-coated fiber is between 0.011-0.014 mm3. The reproducibility of the film thickness was found to be rather promising. The results obtained from the optimization of sol-gel process indicated that fiber prepared from a 67 higher ratio of C8-TEOS to MTMOS resulted in higher extraction capability for standard hydrocarbons. The possible underlying mechanisms of the sol-gel coating process were discussed and the successful binding of C8-TEOS to the coating stationary phase was confirmed by IR spectra. 68 CHAPTER 4 HEADSPACE SPME OPTIMIZATION, METHOD VALIDATION AND PRACTICAL APPLICATION OF THE C8-COATED FIBER 4.1 Introduction This chapter covers the optimization of headspace SPME experimental parameters (extraction time, desorption time and extraction temperature). The precision of the proposed method, linearity of the calibration graph for n-alkane hydrocarbon compounds and the detection limits of accelerants obtained by using C8-coated fiber are discussed. It also includes the discussion for the extraction of common accelerants in simulated arson samples. 4.2 Optimization of HS-SPME Operating Conditions 4.2.1 Optimization of Extraction Time SPME is based on an equilibration distribution process, the maximum amount of analyte will be extracted at the equilibration time. Therefore the extraction equilibration time (optimum fiber exposure time) is being considered as a decisive factor in improving extraction efficiency (Cai et al., 2003). The longer extraction time leads to the higher extraction efficiency and sensitivity at pre-equilibrium (Hou et al., 2004). In general, 69 equilibrium is compound dependent and can vary significantly among the different compounds (Yu et al., 2004). It also depends on the mass transfer of the analytes through the three-phase system: polymeric coating, headspace and sample matrix (Liu et al., 2005). The extraction equilibration time was initially investigated by exposing the fiber to the headspace of an apparatus containing target analytes for a variety of time, from 5 to 35 minutes, until the amounts extracted remains constant. The extraction time profiles were constructed by plotting the mass adsorbed, measured as chromatographic peak area, against extraction time as shown in Figure 4.1. The optimum fiber exposure time estimated from the profile is 25 minutes. Within the duration of 25 minutes, the analytes would have migrated in and out of the coating until the equilibration was established between the phases. Referring to Figure 4.1, the extraction efficiency reaches a plateau after 25 minutes exposure of the fiber to the headspace. Therefore it can be suggested that after 25 minutes, an equilibration has been established. According to a previous work done in this lab, Yong (2004) suggested 35 minutes equilibration for components commonly found in accelerants using commercial PDMS fiber (cooling temperature ramp approach). A slightly shorter equilibration time (25 minutes) was obtained in this study by using the lab-made C8-coated fiber as compared to those reported by Yong (2004) for commercial PDMS fiber. This might be due to the porous structure and thinner coating of the lab-made C8-coated fiber. Lopes et al. (2004) reported that coatings provided by sol-gel process are thinner and therefore saturation with the extracted analytes takes less time than with conventional fibers, that are coated with relatively thicker films of sorbents. The extraction equilibration time obtained in this study is comparable with previous works done by Gbatu et al. (1999) which stated a 20 minutes equilibration for organo-tin compounds, 25 minutes for organo-mercury compounds and 30 minutes for organo-arsenic compounds using sol-gel derived C8-coated fiber. It has been reported 70 that the equilibration time obtained using C8-fiber is much shorter than those obtained using commercial PDMS/DVB fiber (even after an hour the equilibration has not completely reached equilibration). The difference in equilibration times of the compounds in the two types of fibers can be attributed to difference in porosity. Analytes will diffuse faster through a thinner, porous coating than a thicker, non-porous coating. Chong et al. (1997) obtained 10 minutes equilibration for nonpolar analyte using sol-gel PDMS fiber. A short equilibration time of 15 minutes has been achieved for organic compounds using sol-gel PDMS/PVA fiber (Lopes et al., 2004). Average Peak Area 600 500 400 300 200 100 0 0 5 10 15 20 25 30 35 40 Time (min) C8 C10 C12 C14 C16 Figure 4.1: Extraction equilibration profiles of n-alkane hydrocarbon compounds (C8, C10, C12, C14 and C16) at varying extraction time. Other SPME conditions: extraction temperature of 80 °C and desorption time of 80 s. 4.2.2 Optimization of Desorption Time The time needed for complete desorption of analytes from the C8-coated fiber has to be carefully determined to achieve quantitative desorption and avoid carry-over 71 effects that occur among subsequent SPME analyses. The amounts of analyte desorbed from the fiber will influence detection sensitivity. Nine desorption times ranging from 10 to 90 s in 10s increments were tested. The analytes were thermally desorbed at the GC injection port set at a temperature of 250 °C, at which the volatile hydrocarbon compounds can be desorbed immediately. As can be seen from Figure 4.2 (relationship between response and desorption time), the chromatographic signal increased gradually with the desorption time and stabilized at about 50 s, which indicated that a state of equilibrium had been reached. This result revealed that the desorption process is fast and all the hydrocarbon compounds could be desorbed in 50 s. The optimized desorption time (50 s) can be considered as the minimum time for quantitative transfer of the analytes from the C8fiber to the GC column. However to ensure complete desorption and to avoid possible fiber carryover, the desorption time was set at 1 minute. The desorption equilibration time exhibited by the lab-made C8-fiber is slightly shorter than those reported by Yong (2004) using commercial PDMS fiber (cooling temperature ramp approach). A short desorption time obtained in this study may be attributed to much larger surface area of C8-coating which is favorable to rapid mass transfer from the porous and thinner coating. Wang et al. (2000), Li et al. (2004a), Silva et al. (2005), Liu et al. (2005a) and Cai et al. (2006) reported that sol-gel coatings possess a porous structure that provides faster mass transfer during extraction as well as analyte desorption processes during sample introduction, which shortens the desorption time of analytes. A short desorption time determined in this study is comparable with previous works done by Gbatu et al. (1999), a static desorption time of 5 minutes was sufficient for complete desorption of organotin and organomercury compounds from sol-gel C8coated fiber. The result is also comparable to those obtained by other researches using different sol-gel fibers. Wang et al. (2000) obtained a 20 s desorption equilibration for 72 BTEX compounds using sol-gel derived polyethylene glycol coated fiber. An optimum desorption time of 3 minutes was suggested for aliphatic amines using sol-gel calix[4]arene fiber (Li et al., 2004b). Dong et al. (2005) concluded that a 2 minutes desorption time are sufficient to allow the complete desorption of organochlorine pesticides from sol-gel calix[4]arene/hydroxyl-terminated silicone oil coated fiber. Average Peak Area 600 500 400 300 200 100 0 0 10 20 30 40 50 60 70 80 90 100 Time (Sec) C8 C10 C12 C14 C16 Figure 4.2: Desorption profiles of C8-fiber for standard n-alkane hydrocarbon compounds (C8, C10, C12, C14 and C16) by HS-SPME. Extraction conditions: extraction time, 15 min; extraction temperature, 80 °C. 4.2.3 Optimization of Extraction Temperature The effect of extraction temperature on the amounts of analytes absorbed was studied by performing HS-SPME extractions of hydrocarbons at temperatures ranging from 40 to 100 °C. The average peak area versus extraction temperature profiles for HSSPME extractions of standard hydrocarbons were determined as shown in Figure 4.3. 73 Referring to Figure 4.3, it is apparent that peak areas increase as the extraction temperature is increased. The extraction efficiency of the fiber increased with temperature until 90 °C. Increasing the extraction temperature to 90 °C might have helped transfer of analytes to the headspace and enhanced diffusion of analytes towards the fiber. However, extraction yields tended to decrease slightly after 90 °C. This might be because the distribution constant decreased with increasing temperature. In order to control the stable temperature efficiently, 90 °C was selected as the extraction temperature for analysis of volatile hydrocarbon compounds. The extraction efficiency of sol-gel fiber is mainly controlled by two factors, distribution velocity and partition coefficient. The distribution velocity, which helps analytes to move near the solid-phase coating, increases with increased temperature. Extraction is an exothermic process and the partition coefficient, which determines the ratio of analytes extracted, is inversely related to temperature (Yu et al., 2002; Li et al., 2004; Lopes et al., 2004; Basheer et al., 2005). This means that at a higher temperature, diffusion coefficients are higher and therefore a faster distribution equilibrium is achieved, but partition coefficients are lower and unavoidably result in a poorer extraction efficiency. Zeng et al. (2001) and Zhou et al. (2005) reported that the first factor (distribution velocity) is dominant in the temperature range less than the optimum extraction temperature, while the second factor (partition coefficient) becomes dominant at temperatures higher than the optimum extraction temperature. 74 Average Peak Area 600 500 400 300 200 100 0 0 5 10 15 20 25 30 35 40 Temperature (Degree Celcius) C8 C10 C12 C14 C16 Figure 4.3: Influence of extraction temperature on the signal intensity of standard nalkane hydrocarbons (C8, C10, C12, C14 and C16). Extraction conditions: extraction time, 15 min; desorption time, 80 s. 4.3 Validation of the Analytical Method The investigation of the proposed method’s precision, linearity of calibration graphs for n-alkane hydrocarbons and detection limits of accelerants was carried out by applying the optimum HS-SPME experimental conditions. 4.3.1 Accuracy of HS-SPME Method using C8-coated Fiber The accuracy of the developed method for three replicate analyses was determined by calculating the relative standard deviations (RSD) of the peak areas for all standard hydrocarbons (C8, C10, C12, C14 and C16) in burnt carpet sample spiked at the same concentration level. In order to determine the reproducibility of the analytes’s peak 75 areas, the three consecutive analyses were performed on the same day. Reproducibility of an analyte’s peak area is important in order to produce a comparable chromatographic profile for visual pattern comparison. Referring to Table 4.1, all the hydrocarbon compounds gave a low relative standard deviation (RSD) value ranging from 3.9 %-5.5 % which shows an acceptable reproducibility. Therefore the optimized HS-SPME using lab-made C8-coated fiber is expected to produce good qualitative chromatographic profiles since comparable peak area range was observed. These results indicated that quantitative analysis of hydrocarbons was possible using the external standard method and the applicability of lab-made C8-coated fiber for routine analysis. The result obtained is in good agreement with those reported by Gbatu et al. (1999). It has been reported that the C8-coated fiber exhibited good reproducibility for the analysis of organometals. Table 4.1: Inter-day precision of HS-SPME method for the extraction of standard nalkane hydrocarbon compounds using C8-coated fiber under optimum conditions. Hydrocarbon Average Peak Area RSD (%) Compounds (n=3) C8 211.3 4.8 C10 352.6 4.3 C12 487.5 5.1 C14 579.7 5.5 C16 391.4 3.9 76 4.3.2 Detection Limits of Accelerants Detection limits of accelerants (lowest detectable amount determined based on three times the baseline noise) were evaluated in order to access the performance of the proposed method. The lab-made C8-coated fiber was employed for extractions of gasoline, kerosene and diesel in spiked burnt carpet sample under optimum HS-SPME conditions. Referring to Table 4.2, for all the common accelerants, the LODs were between 0.7 to 1.0 µL. A minimum detectable amount of 0.7 µL were obtained for the C8 fiber extraction of gasoline in burnt carpet sample. A slightly higher LODs (1.0 µL) were detected for kerosene and diesel. These values are indicative of good performance of the lab-made C8-coated fiber. The detection limits of common accelerants obtained is this study using C8-fiber is comparable to those reported by Yong (2004) using commercial PDMS/DVB fiber. In the previous work done in this lab (Yong, 2004), it has been reported that LOD as low as 0.5 µL was obtained for gasoline. Detection limits of 1.0 µL were determined for the PDMS/DVB fiber extraction of kerosene and diesel in burnt carpet sample. Table 4.2: LODs for accelerants determined using C8-coated fiber under optimum HSSPME conditions Common accelerant Minimum detectable amount (µL) This work Previous work * Gasoline 0.7 0.5 Kerosene 1.0 1.0 Diesel 1.0 1.0 * Yong 2004 77 4.3.3 Calibration Graph of Target Compounds Calibration curves consisting of five different concentration of n-alkane hydrocarbons were generated as regression lines and the correlation coefficients were obtained. The analytes were extracted using lab-made C8-coated fiber under the Average Peak Area optimum HS-SPME conditions. 180 160 140 120 100 80 60 40 20 0 0 2 4 6 8 10 Analyte Concentration (% v/v) C8 C10 C12 C14 C16 Figure 4.4: Calibration graph of HS-SPME for n-alkane standard hydrocarbon compounds using C8-coated fiber Table 4.3: Regression line and correlation coefficient of standard hydrocarbon compounds. Hydrocarbon Equation Correlation Factor (R2) Compounds C8 y = 3.85x + 2.1 0.9869 C10 y = 8.31x + 2.0 0.9889 C12 y = 12.7x + 2.4 0.9958 C14 y = 16.7x + 2.8 0.9948 C16 y = 11.4x + 2.0 0.9937 12 78 As can be seen from Figure 4.4, the HS-SPME procedure with lab-made C8coated fiber yielded an excellent linear relationship between the peak area counts and analyte concentration in the range of (2.00-10.00) % v/v. Good correlations with R2 value between 0.9869-0.9958 (R2 ~ 1) were obtained as shown in Table 4.3. It is apparent that there were almost no factors in the HS-SPME extractions of n-alkane hydrocarbons using lab-made C8-coated fiber that might cause the concentration characteristics to deviate from linearity. This result is in good agreement with most of the research works done with different sol-gel fibers as it shows a positive correlation between the peak area counts and concentration of target compounds. Cai et al. (2003) reported that linear calibration curves were obtained for all the tested amines using sol-gel derived dibenzo-18-crown-6 fiber. Liu et al. (2005) reported that the HS-SPME procedure with sol-gel derived BMA/OH-TSO fiber showed excellent linearity in concentrations ranging from 0.1-10 µg/g. Yu et al. (2004) reported that the calibration curves were found to have good linearity by correlation coefficients (R2) of more than 0.99 for organophosphorus pesticides using sol-gel derived bisbenzo crown ether/(OH-TSO) fiber. 4.4 Analysis of Simulated Arson Samples Using C8-coated Fiber The feasibility of the lab-made C8-coated fiber for the extractions of common accelerants was evaluated using simulated arson samples (Section 2.8). 4.4.1 Selectivity for Accelerants In order to examine selectivity of the lab-made SPME fiber towards petroleum based accelerants, burnt carpet sample were individually spiked with known amount of gasoline, kerosene and diesel. The samples were subjected to headspace SPME using 79 sol-gel derived C8-coated fiber and the GC profiles were compared with that from direct injection. As can be seen in Figure 4.5 (b), all the hydrocarbon components in gasoline spiked burnt carpet sample were recovered by using C8-coated fiber and the chromatogram was comparable with the profile of gasoline from direct injection [Figure 4.5 (a)]. The hydrocarbon components in kerosene (Figure 4.6) and diesel (Figure 4.7) spiked burnt carpet samples were also effectively extracted using the lab-made C8 fiber and similar comparisons were obtained with direct injection of the respective accelerants. This indicated that the lab-made SPME fiber favored the extraction of hydrocarbons, thus providing a good selectivity towards hydrocarbons in petroleum based accelerants. The GC profiles of gasoline, kerosene and diesel spiked samples obtained in this study were in good agreement with those obtained by Borusiewicz et al. (2004) using Tenax as adsorbent and Yong (2004) by using commercially available fibers. In the previous work done in this lab (Yong, 2004) a complete study on identifying the hydrocarbon components in gasoline, kerosene and diesel had been carried out. The same samples were used in this study . Therefore, the chromatograms obtained in this study are comparable to those obtained by Yong (2004). 80 Figure 4.5 (a): GC profiles of direct injection of gasoline. Peak identities: (1) methylbenzene, (2) ethylbenzene, (3) 1, 3-dimethylbenzene, (4) 1, 2dimethylbenzene, (5) 1-ethyl-2-methylbenzene, (6) 1, 2, 4-trimethylbenzene, (7) 1, 2, 3- trimethylbenzene. Figure 4.5 (b): GC profiles of gasoline spiked burnt carpet sample using C8-coated fibers. Peak identities: (1) methylbenzene, (2) ethylbenzene, (3) 1, 3- 81 dimethylbenzene, (4) 1, 2- dimethylbenzene, (5) 1-ethyl-2-methylbenzene, (6) 1, 2, 4-trimethylbenzene, (7) 1, 2, 3- trimethylbenzene. Figure 4.5 (c): GC profiles of the blank burnt carpet sample using C8-coated fibers. Figure 4.6 (a): GC profiles of direct injection of kerosene.Peak identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, (7) 2, 6-dimethylnapthalene, (8) C15 82 (b) Figure 4.6 (b): GC profiles of kerosene spiked burnt carpet sample using C8-coated fibers. Peak identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, (7) 2, 6-dimethylnapthalene, (8) C15 Figure 4.7: GC profiles of diesel spiked burnt carpet sample using C8-coated fibers. Peak identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, (7) C15, (8) C16 hydrocarbons. 83 4.4.2 Extraction Capability for Accelerants The extraction capability of the sol-gel derived C8-coated fiber was determined by comparing it with the commercially available PDMS/DVB fiber. PDMS/DVB fiber was selected for comparison because previous work done in this lab, Yong (2004) proved that the fiber has the highest sensitivity towards hydrocarbon compounds and most suitable for accelerants identification. Both C8-coated fiber and PDMS/DVB fiber were capable of extracting early, middle and late eluting hydrocarbon compounds sufficiently. However, the C8-coated fiber showed a slightly higher extraction capability by contrast with conventional PDMS/DVB fiber for all the accelerants as shown in Figure 4.8. This result is comparable to those reported by Gbatu et al. (1999). A higher extraction capability could be due to the existence of higher surface area for the C8coated fibers (Gbatu et al., 1999; Cai et al., 2003; Zuin et al., 2004; Liu et al., 2005a). Average Peak area Spiked with Gasoline 30 25 20 15 10 5 0 1 2 3 C8-coated fiber 4 5 6 7 PDMS/DVB Figure 4.8 (a): Comparison of sol-gel derived C8-coated fiber and commercially available PDMS/DVB fiber in extracting hydrocarbon compounds from gasoline spiked burnt carpet sample. Label identities: (1) methylbenzene, (2) ethylbenzene, (3) 1, 3dimethylbenzene,(4) 1, 2-dimethylbenzene, (5) 1-ethyl-2-methylbenzene, (6) 1, 2, 4trimethylbenzene, (7) 1, 2, 3-trimethylbenzene. 84 A verag e P eak area Spiked with Kerosene 20 15 10 5 0 1 2 3 4 5 C8-coated fiber 6 7 8 PDMS/DVB Figure 4.8 (b): Comparison of sol-gel derived C8-coated fiber and commercially available PDMS/DVB fiber in extracting hydrocarbon compounds from kerosene spiked burnt carpet sample. Label identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, (7) 2, 6-dimethylnapthalene, (8) C15 A verage P eak A rea Spiked with diesel 8 6 4 2 0 1 2 3 4 C8-Coated fiber 5 6 7 8 PDMS/DVB Figure 4.8 (c): Comparison of sol-gel derived C8-coated fiber and commercially available PDMS/DVB fiber in extracting hydrocarbon compounds from 85 diesel spiked burnt carpet sample. Label identities: (1) C9, (2) C10, (3) C11, (4) C12, (5) C13, (6) C14, C15, (7) C16 4.5 Conclusions The headspace SPME method using lab-made C8-coated fiber has been successfully optimized. A short extraction (25 minutes) and desorption (50 s) equilibration times arised from the porous structure of the sol-gel C8-coated fiber and resulted in short analysis time. The optimum extraction temperature for standard hydrocarbons was found to be 90 °C. The developed HS-SPME method using C8-coated fiber exhibited a high degree of inter-day accuracy under the optimum experimental conditions. The limits of detection for the common accelerants studied under the optimum HS-SPME conditions were in the range of 0.7 to 1.0 µL. The linearity of the calibration graphs for all the tested n-alkane hydrocarbons was excellent (r > 0.9869). The C8 fiber was capable of effectively extracting components commonly found in accelerants and exhibited a good selectivity for accelerants. Comparison of the lab-made sol-gel derived C8-coated fiber with the commercially available PDMS/DVB fiber showed that the lab-made fiber has a higher extraction capability for fire accelerants under study. 86 CHAPTER 5 CONCLUSIONS AND FUTURE DIRECTIONS 5.1 Conclusions In this study a lab-made sol-gel derived C8-coated fiber has been developed for the headspace SPME-GC analysis of n-alkane standard hydrocarbons and common accelerants. The developed fiber has been characterized in terms of selectivity and extraction capability for standard hydrocarbons, lifetime, thermal stability and surface morphology of the coating. The reproducibility of the C8 coating has been studied in terms of coating thickness. Optimization of the sol-gel process was effected by varying the molar ratios of the silane precursors. The successful binding of the coating material (C8-TEOS) to the coating stationary phase during the formation of sol-gel C8-coating was studied through the IR analysis. The chemistry behind the sol-gel coating technique involved in this study were discussed. The HS-SPME method using lab-made C8-coated fiber was preliminary optimized using n-alkane standard hydrocarbon mixture. The experimental parameters were investigated in terms of fiber exposure time, analyte desorption time and extraction temperature. The accuracy of the proposed method, linearity of calibration curves and detection limits of accelerants were investigated. Finally, the practical applicability of the lab-made fiber was studied via the HS-SPME of common accelerants in simulated arson samples. 87 All the tested n-alkane standard hydrocarbons and hydrocarbon components in gasoline, kerosene and diesel spiked burnt carpet samples were effectively extracted using the lab-made fiber. This indicated that the lab made fiber has a good selectivity towards n-alkane hydrocarbon compounds in petroleum based accelerants. Comparison of the lab-made sol-gel derived C8-coated fiber with the commercially available PDMS/DVB fiber showed that the lab-made fiber has a higher extraction capability for standard hydrocarbons and accelerants. It has to be high-lighted that the lifetime of the sol-gel C8-coated fiber was found to be remarkably longer than that of commercially available fibers. The extraction ability of the fiber showed no significant decline even after it has been used for 200 times [the tested hydrocarbon compounds gave a low RSD value (< 5.5 %)]. This proves that the C8 coated fiber surface was not depleted after it has been used for 200 extractions. In contrast, the commercially available fibers can only be used for 40 to 100 extractions. The sol-gel C8-coated fiber exhibited higher thermal stability compared to conventionally coated PDMS/DVB fiber due to the strong chemical bonding between sol-gel generated C8 composite and the fused silica surface. The fiber can be routinely used at 300 °C without any sign of bleeding (RSD value of < 3.7 % was routinely obtained for hydrocarbons on the C8-coated conditioned at 300 ° C), whereas conventionally coated PDMS/DVB fiber begins to bleed above 270 °C [limitations listed by the manufacturer (Supelco)]. Scanning electron micrographs revealed that the developed C8-coating possesses a porous structure. It increases the surface area on the fiber, therefore, even a thin coating will be able to provide enhanced stationary phase loading and high sample capacity. The surface of the fiber coating was well-distributed but not uniformly coated. No apparent cracks were observed on the surface of the fiber coated with 1:1, C8TEOS:MTMOS which indicated a good coating material for SPME. The film thickness of the C8 coating was estimated to be between (3-4) µm, which leads to an extracting 88 phase volume of (0.011-0.014) mm3. The simplicity of the sol-gel coating procedure provided a rather promising coating thickness reproducibility. IR spectrum showed that the Si-O group peak identified in pure C8-TEOS also appeared in the sol-gel derived C8-coating but in decreased intensity. This most probably prove the successful binding of the coating material to the stationary phase. The results obtained from the optimization of sol-gel process indicated that fiber prepared from a higher ratio of C8-TEOS to MTMOS resulted in higher extraction capability for n-alkane standard hydrocarbons. The optimization of the HS-SPME method using C8 fiber revealed that the porous structure and thinner coating of the sol-gel derived C8-coated fiber curtailed the equilibration times, resulting in short analysis time. The highest amount of analytes were extracted with a 25 minutes extraction time. A minimum desorption time of 50 s was sufficient to allow complete desorption of hydrocarbon compounds from the C8 fiber. The optimum extraction temperature for analysis of volatile hydrocarbon compounds was found to be 90 °C. The proposed HS-SPME method using C8-coated fiber showed satisfactory precision, linearity and detection limits under the optimum experimental conditions. The three consecutive analyses of standard hydrocarbons (C8, C10, C12, C14 and C16) spiked in burnt carpet sample showed a good reproducibility with RSD below 6 %. The detection limits of common accelerants (gasoline, kerosene and diesel) were determined to be in the range of (0.7 -1.0) µL. HS-SPME procedure with C8-coated fiber yielded excellent linearity in concentrations ranging from (2.00-10.00) % v/v with good correlations (r > 0.9869). The sol-gel C8 composite proposed in this study as a material for SPME fiber coating can be generated in situ by a simple, straightforward sol-gel polycondensation route. The presented experimental results conclusively demonstrated the potential of solgel derived C8-coating in SPME and chromatographic separations. The resulting 89 procedure was shown to be a good alternative methodology for qualitative and quantitative analysis of accelerants in arson samples, being a simple, fast, reproducible and also an environmental friendly analytical method. 5.2 Future Directions Further studies need to be carried out for a complete assessment of the lab-made adsorbent and to improve the results obtained in this research. For future work on characterizing the lab-made fiber, thermo gravimetric analysis of the sol-gel C8 composite should be performed in order to determine the mass loss and degradation of the coating at extremely high temperatures (up to 700 °C). This study is essential as SPME fibers with extended working temperature ranges reduces the amount of bleeding from the fiber coating during thermal desorption. Nitrogen adsorption analysis need to be carried out using a bulk quantity of the C8-coating in order to determine the surface area. The values of surface area obtained can be used to analyze the porous appearance of the coating (micro- or mesopores). A study on solvent stability of the lab-made fiber can be carried out by exposing the fiber to different organic and inorganic solvents (methylene chloride, acetonitrile, acetone, hexane, xylene and distilled water) for various lengths of time prior to extraction. The pH stability of the fiber can also be estimated by comparing the extraction efficiencies of the sol-gel derived C8 coating in extracting hydrocarbons before and after rinsing with acidic and alkaline solutions of various pH values. The coating preparation reproducibility can be investigated in many aspects. In this preliminary work, the coating preparation reproducibility of the C8 fiber has been studied in terms of coating thickness. A different aspect, such as fiber-to-fiber and batchto-batch reproducibility should be investigated in future work. At least three C8 fibers prepared in the same batch and three identically prepared C8 fibers in three different 90 batches need to be subjected to HS-SPME of n-alkane hydrocarbons. The coating preparation reproducibility of the C8 fibers within batch and between batches should be determined by calculating the RSD of the peak areas for all the hydrocarbons. Although the indirect headspace-SPME extraction technique (HS-SPME) using the lab-made C8-coated fiber has been successfully employed for the extraction of nalkane hydrocarbons, a different SPME extraction mode (direct fiber immersion extraction) is suggested to be used in further study in order to investigate the applicability of the lab-made fiber for both the extraction modes. For future work, an alternative extraction approach (cooling temperature ramp) utilized by Yong (2004) can be employed for HS-SPME of accelerants. The effects of different extraction methods towards the recovery of accelerants can be studied by comparing the extraction at constant temperature (done in this study) and cooling temperature ramp. Since most volatile residues in fire debris involve complex mixtures, coupling HS-SPME with GC/MS in future work gives the best tool for arson analysis. GC/MS has the capability of displaying a mass spectrum for each peak and scan selected ions, those that are characteristic for a particular chemical species of interest. The development of the sol-gel derived C8-coated fiber is one of the latest improvement in the field of forensic science. The objectives of this work have been achieved with great success and promising results. The lab-made fiber has a high potential of improving the quality of accelerant extraction, therefore it will be beneficial for arson crime scene investigations. It is hoped that the positive attributes of the coating will widen the application range of the fiber. The proposed ideas and suggestions would hopefully pave the way for future directions in the improvement of the lab-made fiber’s characteristics, accuracy of the coating preparation method and applicability of the fiber for different SPME extraction approaches. 91 REFERENCES Almirall, J. S., Bruna, J. C. and Furton, K. G. (1996). Recovery of Accelerants in Aqueous Samples Using SPME. Science & Justice. 36: 283-287. Almirall, J. S. and Furton, K. G. (2004). Characterization of Background and Pyrolysis Products that may Interfere with the Forensic Analysis of Fire Debris. J. Anal. Appl. Pyrolysis. 71: 51-67. Almirall, J. S., Furton, K. G and Bruna, J. C. (1996). A Novel Method for the Analysis of Gasoline from Fire Debris Using Headspace Solid-phase Microextraction. J. Forensic. Sci. 71: 12-22. Aranda, R., Kruus, P. and Burk, R. C. (2000). Assessment of Polycrystalline Graphites as Sorbents for Solid Phase Microextraction of Non-ionic Surfactants. J. Chromatogr. A. 888: 35-41. Arkles, B. (May 2001) Commercial Application of Sol-gel Derived Hybrid Materials. MRS Bulletin. 402-407. Atanacio, A. J., Latella, B. A., Barbe, C. J., and Swain, M. V. (2005). Mechanical Properties and Adhesion Characteristics of Hybrid Sol-gel Thin Films. Surface & Coatings Tech. 192: 354-364. Azenha, M., Malheiro, C. and Silva, A. F. (2005). Ultrathin Phenyl-functionalized Solidphase Microextraction Fiber Coating Developed by Sol-gel Deposition. J. Chromatogr. A. 1069: 163-172. 92 Bagheri, H., Es-Hagli, A and Rouini, M. R. (2005). Sol-gel Based Solid-phase Microextraction and Gas Chromatography-Mass Spectrometry Determination of Dextromethorphan and Dextrorphan in Human Plasma. J. Chromatogr. B. 818: 147-157. Bagheri, H., Mir, A. and Babanezhad, E. (2005). An Electropolymerized Aniline-Based Fiber Coating for Solid-Phase Microextraction of phenols from water. Anal Chim Acta. 532: 89-95. Basheer, C., Valiyaveetil, S. and Lee, H. K. (2005). Sol-gel Coated Oligomers as Novel Stationary Phases for Solid-phase Microextraction. J. Chromatogr. A. 1087: 252258. Borusiewicz, R., Zadora, G. and Palus, J. Z. (2004). Application of Headspace Analysis with Passive Adsorption for Forensic Purposes in the Automated Thermal Desorption GC-MS System. Chromatographia Supplement. 60: S113-S142. Brinker, C. J. and Scherer, G. W. (1990). Sol-Gel Science. San Diego: Academic Press. Brinker, C. J. and Scherer, G. W. (1990). Sol-Gel Science: The Physics and Chemistry of Sol-Gel Processing. New York: Academic Press. Inc. Brinker, C. J., Hurd, A. J., Schunk, P. R., Frye, G. C. and Ashley, C. S.(1992). Review on Sol-gel Thin Film Formation J. Non-cryst. Solids. 147: 424-436. Cai, L. S., Zhao, Y. Y., Gong, S. L., Dong, L. and Wu, C. Y. (2003). Use of a Novel Sol-gel Dibenzo-18 Crown-6 Solid Phase Microextraction Fiber and a New Derivatizing Reagent for Determination of Aliphatic Amines in Lake Water and Human Urine. Chromatographia. 58: 615-621. 93 Cai, L., Gong, S., Chen, M. and Wu, C. (2006). Vinyl Crown Ether as a Novel Radical Crosslinked Sol-gel SPME Fiber for Determination of Organophosphorus Pesticides in Food Samples. Anal. Chim. Acta. 559: 89-96. Cavanagh, K., Pasquier, E. D. and Lennard, C. (2002). Background Interference from Car Carpets-the Evidential Value of Petrol Residues in Cases of Suspected Vehicle Arson. Forensic Sci. Int. 125: 22-36. Chong , S. L., Wang, D., Hayes, J. D., Wilhite, B. W. and Malik, A. (1997). Sol-gel Coating Technology for the Preparation of Solid-phase Microextraction Fibers of Enhanced Thermal Stability. Anal. Chem. 69: 3889-3898. DeHaan, J. D. (1997). Kirk’s Fire Investigation. 4th ed. Upper Saddle River, N. J.: Prentice Hall. Dennett, M. F. (1980). Fire Investigation: A Practical Guide for Students and Officers, Insurance Investigators, Loss Adjusters and Police Officers. Oxford: Pergamon. Dietz, C., Sanz, J. and Camara C. (2005). Review: Recent Developments in Solid-phase Microextraction Coatings and Related Techniques. J. Chromatogr. A. 1103: 183192. Djozan, D. and Assadi, Y. (2004). Modified Pencil Lead as a New Fiber for Solid-phase Microextraction. Chromatographia. 60: 313-317. Djozan, D. and Bahar, S. (2004). Solid Phase Microextraction of Aliphatic Alcohols Based on Polyaniline Coated Fibers. Chromatographia. 59: 95-99. Djozan, D. and Zehni, M. A. (2003). Copper Sulphide Wire as a Selective Fiber in Solid-phase Microextraction. Chromatographia. 58: 221-224. 94 Djozan, D., Pournaghi-Azar, M. H. and Bahar, S. (2004). Modified Polypyrrole with Tetrasulfonated Nickel Phtalocyanine as a Fiber for Solid Phase Microextraction. Application to the Extraction of BTEX Compounds from Water Samples. Chromatographia. 59: 595-599. Dolan, J. (2003). Recent Advances in The Applications of Forensic Science to Fire Debris Analysis. Anal. Bioanal. Chem. 376: 1168-1171. Dong, C., Zeng, Z. and Li, X. (2005). Determination of Organochlorine Pesticides and Their Metabolites in Radish After Headspace Solid-phase Microextraction Using Calyx[4]arene Fiber. Talanta. 66: 721-727. Eisert, R. and Pawliszyn, J. (1997). New Trends in Solid-phase Microextraction. Crit. Rev. Anal. Chem. 27: 103-135. Farajzadeh, M. A. and Rahmani, N. A. (2005). Electrolytically Produced Copper (I) Chloride on the Copper Wire as an Excellent Sorbents for Some Amines. Talanta. 65: 700-704. Fire and Rescue Department Malaysia, “Emergency Statistic”. 5th Jan 2006. www.bomba.gov.my. Furton, K. G., Almirall, J. R., Bruna, J. (1995). A Simple Inexpensive Rapid, Sensitive and Solventless Method for the Recovery of Accelerants from Fire Debris Based on SPME. J. High Res. Chromatogr. 18: 625-629. Gbatu, T. P., Sutton, K. L. and Caruso, J. A. (1999). Development of New SPME Fiber By Sol-gel Technology for SPME-HPLC Determination of Organometals. Anal. Chim. Acta. 402: 67-79. Gierak, A., Seredyeh, M. and Bartnicki, A. (2006). The Preparation, Properties and Application of Carbon Fibers for SPME. Talanta.69: 1079-1087. 95 Gorecki, T., Yu, X. and Pawliszyn, J. (1999). Theory of Analyte Extraction by Selected Porous Polymer Fibers. Analyst. 124: 643-649. Harris, A. C. and Wheeler, K. F. (2003). GC-MS of Ignitable Liquids Using Solventdesorbed SPME for Automated Analysis. J. Forensic Sci. 48: 200-207. Hofacker, S., Mechtel, M., Mager, M. and Kraus, H. (2002). Sol-gel: A New Tool for Coatings Chemistry. Progress in Organic Coatings. 45: 159-164. Hu, Y. L., Fu, Y. L. and Li, G. K. (2006). Preparation of Anilinemethyltriethoxysilane /polydimethylsiloxane Sol-gel Coatings for Solid-phase Microextraction of Aromatic Compounds. Anal. Chim. Acta. 567: 211-217. Hun, Y., Yang, Y. and Li, G. (2005). Preparation and Application of Poly(dimethylsiloxane)/B-cyclodextrin Solid-phase Microextraction Membrane. Anal. Chim. Acta. 543: 17-24. Huo, J. G., Ma, Q., Du, X. Z., Deng, H. L. and Gao, J. Z. (2004). Inorganic/ Organic Mesoporous Silica as a Novel Fiber Coating of SPME. Talanta. 62: 241-246. Icove, D. J. and DeHaan, J. D. (2004). Forensic Fire Scene Reconstruction. Upper Saddle River, N. J.: Prentice Hall. Jackowski, J. P. (1997). The Incidence of Ignitable Liquid Residues in Fire Debris as Determined by a Sensitive and Comprehensive Analytical Scheme. J. Forensic Sci. 42: 828-832. Kabir, A., Hamlet, C., Yoo, K. S., Newkome, J. R. and Malik, A. (2004). Capillary Microextraction on Sol-gel Dendrimer Coatings. J. Chromatogr. A. 1034: 1-11. 96 Klein, L. C. (1988) Sol-gel Technology for Thin Films, Fibers, Preforms, Electronics And Specialty Shapes. Noyes, William Andrew Publishing. Lee, H. C, Desio, P. J. and Gaensslen, R. E. (1986). Methods for The Recovery of Accelerants Residues from Arson Debris. In: Davies . G. Forensic Science. 2nd ed. Washington. Dc.: American Chemical Society.279-298. Lennard, C. J., Rochaix, V. T. and Margot, P. (1995). A GC-MS Database of Target Compound Chromatograms for The Identification of Arson Accelerants. Science & Justice. 35: 19-30. Li, X., Zeng, Z. and Zhou, J. (2004). High Thermal-Stable Sol-gel Coated Calix[4]arene Fiber for SPME of Chlorophenols. Anal. Chim. Acta. 509: 27-37. Li, X., Zeng, Z., Zhou, J., Gong, S., Wang, W. and Chen, Y. (2004). Novel Fiber Coated with Amide Bridged-Calix[4]arene Used for SPME of Aliphatic Amines. J. Chromatogr. A. 1014:1-9. Liu, M. M., Zeng, Z. R., Wang, C. L., Tan Y. J. and Liu H. (2003). Solid-phase Microextraction of Phosphate and Methylphosphonate Using Novel Fibers Coated with Sol-gel Derived Silicone-divinylbenzene Co-polymer. Chromatographia. 58: 597-605. Liu, M., Liu, Y., Zeng, Z. and Peng, T. (2006). Preparation and Characteristics of High pH-resistant Sol-gel Alumina-based Hybrid Organic-inorganic Coating for Solidphase Microextraction of Polar Compounds. J. Chromatogr. A. Article in Press. Liu, W., Hu, Y., Zhao, J., Xu, Y. and Guan, Y. (2006). Physically Incorporated Extraction Phase of Solid-phase Microextraction by sol-gel Technology. J. Chromatogr. A. 1102: 37-43. 97 Liu, M., Zeng, Z. and Fang, H. (2005). Preparation and Application of the Sol-gel Derived Acrylate/silicone Co-polymer Coatings for Headspace Solid-phase Microextraction of 2-chloroethyl Ethyl Sulfide in Soil. J. Chromatogr. A. 1076: 16-26. Liu, M., Zeng, Z. and Tian, Y. (2005). Elimination of Matrix Effects for Headspace Solid-phase Microextraction of Important Volatile Compounds in Red Wine using a Novel Coating. Anal. Chim. Acta. 540: 341-353. Liu, M., Zeng, Z. and Xiong, B. (2005). Preparation of Novel Solid-phase Microextraction Fibers by Sol-gel Technology for Headspace Solid-phase Microextraction–Gas Chromatographic Analysis of Aroma Compounds in Beer. J. Chromatogr. A. 1065: 287-299. Liu, Y., Shen, Y. and Lee, M. L., (1997). Porous Layer Solid Phase Microextraction Silica Bonded Phases. Anal. Chem. 69: 190-195. Lloyd, J. A. and Edmiston, P. L. (2003). Preferential Extraction of Hydrocarbons from Fire Debris Samples by Solid phase Microextraction J. Forensic Sci. 48: 130134. Lopes, A. L. and Augusto, F. (2004). Preparation and Characterization of Polydimethylsiloxane/Poly(vinylalcohol) Coated SPME Fibers Using Sol-Gel Technology. J. Chromatogr. A. 1056: 13-19. Lord, H. and Pawliszyn, J. (2000). Evolution of Solid-phase Microextraction Technology. J. Chromatogr. A.. 885: 153-193. Ma, T., Olenick, S. M., Klassen, M. S. and Roby, R. J. (2004). Burning Rate of Liquid Fuel on Carpet (Porous Media). Fire Tech. 40: 227-246. 98 Mackenzie, J. D. In .Hybrid Organic-Inorganic Composites. Marks. J. E., Lee. C. C. Y. and Bianconi, P. A. Eds, ACS Symposium Series 585. American Chemical Society. Washington DC .1995. 227-236. Malik, A. and Wang, D. (2004). Solid Phase Microextraction Fiber Structure and Method of Making. (U. S. Patent 6,759,126 B1). Mester, Z., Sturgeon, R. and Pawliszyn, J. (2001). Review: Solid Phase Microextraction as A Tool for Trace Element Speciation. Spectrochim Acta Part B. 56: 233-260. Mohammadi, M., Yamini, Y. and Alizadeh, N. (2004). Dodecylsulfate-Doped Polypyrrole Film Prepared by Electrochemical Fiber Coating Technique for Headspace Solid-pahse Microextraction of Polycyclic Aromatic Hydrocarbons. J. Chromatogr. A. 1063: 1- 8. O’ Connor, J. J. (1987). Practical Fire and Arson Investigation. U. S. A.: Elsevier Science. Oliveira, A. F., Silveira, C. B., Campos, S. D., Campos E. A. and Carasek E. (2005). The Use of a Thin-glass Ceramic Rod as a Surface for Sol-gel Coating in the Preparation of SPME Fibers. Chromatographia. 61: 277-283. Panavaite, D., Padarauskas, A. and Vickackaite, V. (2006). Silicone Glue Coated Stainless Steel Wire for Solid Phase Microextraction. Anal. Chim. Acta. 571: 4550 Perez, V. I., Cortada, M., Apraiz, C., Molto, D. and Font, G. (2001). SPME of 52 Pesticides and Polychlorinated Biphenyls: Extraction Efficiencies of the SPME Coatings Poly(dimethylsiloxane), Polyacrylate, Poly(dimethylsiloxane)Divinylbenzene, Carboxen- Poly(dimethylsiloxane) and CarbowaxDivinylbenzene. J. Separation Sci., 24: 39-48 99 Ren, Q. and Bertsch W. (1999). A Comprehensive Sample Preparation Scheme for Accelerants in Suspect Arson Cases. J. Forensic Sci. 44: 504-515. Rodrigues, R., Lacerda, C. A., and Lancas, F. M. (2002). Coating of Solid Phase Microextraction Fibers with Chemically Bonded Silica Particles : Selective Extraction of Polycyclic Aromatic Hydrocarbons from Drinking Water. J. Chromatogr. Sci. 40: 489-494. Rush, G. E. (2000). The Dictionary of Criminal Justice. 5th ed. California: McGrawHill. Sandercock, P. M. L. and Pasquier, E. D. (2003). Chemical Fingerprinting of Unevaporated Automotive Gasoline Samples. Forensic Sci. Int. 134: 1-10. Silva, R. G. C. and Augusto, F. (2004). Highly Porous Solid-phase Microextraction Fiber Coating Based on Poly(ethylene glycol)-modified Ormosils Synthesized by Sol-gel Technology. J. Chromatogr. A. 893: 157-168 Silveira, C. B., Oliveira, A. F., Campos, S. D., Campos, E. A. and Carasek, E. (2005). Glass Fibers Coated with Nb2O5 for use in SPME. Colloids and Surface A: Physicochem. Eng. Aspects. 259: 15-22. Snow, N. H. and Slack, G. C. (2002). Headspace Analysis in Modern Gas Chromatography. Trends Anal. Chem. 21: 608-617. Sun, T. H., Cao L. K., and Jia J. P. (2005). Novel Activated Carbon Fiber Solid-phase Microextraction for Determination of Benzyl Chloride and Related Compounds in Water by GC-MS. Chromatographia. 61: 173-179. 100 Theodoridis, G., Koster, E. H. M. and Jong, G. J. (2000). Review: Solid Phase Microextraction for The Analysis of Biological Samples J. Chromatogr. B. 745: 49-82. Tsuda, T., Hisanaga, M., Inagaki, T., Kitagawa, S. and Miyabe, K. (2004). Preparation of a New Type of Fiber Adsorbent Attached with Silica Microparticles. Anal. Science. 20: 1061-1064. Ugur, T., Yucel, S., Nusret, E., Yasemin, U., Kadir, P. and Attila, Y. (2004). Preparation of Sulfonated Overoxidized Polypyrrole Film Applicable as an SPME Tool for Cationic Analytes. J. Electroanal. Chem. 570: 6-12. Wang, D., Xing, J., Peng, J. and Wu, C. (2003). Novel Benzo-15-crown-5 Sol-gel Coating for Solid-phase Microextraction. J. Chromatogr. A. 1005: 1-12. Wang, Y., O’Reily, J., Chen, Y. and Pawliszyn, J. (2005). Equilibrium In-fibre Standardization Technique for Solid-phase Microextraction. J. Chromatogr. A.. 1072: 13-17. Wang, Z., Xiao, C., Wu, C. and Han, H. (2000). High-performance Polyethylene Glycol Coated Solid-phase Microextraction Fibers Using Sol-gel Technology. J. Chromatogr. A. 893: 157-168. Wei, L., Ou, Q., Li, J., and Liang, B. (2004). Preparation of a Solid-phase Microextraction Fiber Coated With γ-Al2O3 and Determination of Volatile Organic Compounds in Indoor Air. Chromatographia. 59: 601-605. Wu, C., Wang, Z. and Xing, J. (2001). Making Method of Crown Ether Solid Phase Extracting Probe. (CN 1, 141, 577 C). 101 Xia, X. R. and Leidy, R. B. (2001). Preparation and Characterization of Porous SilicaCoated Multifibers for Solid Phase Microextraction. Anal. Chem. 73: 2041-2047. Yang, R. and Xie, W. (2004). Preparation and Usage of a New Solid Phase MicroExtraction Membrane. Forensic Sci. Int. 139: 177-181. Yong, Y. T. (2004). Development of Headspace SPME-GC Technique for the Forensic Analysis of Ignitable Liquid Residues in Fire Debris. Universiti Teknologi Malaysia: M.Sc. Thesis. Yu, J., Dong, L., Wu, C., Wu, L. and Xing, J. (2002). Hydroxyfullerene as a Novel Coating for Solid-phase Microextraction Fiber with Sol-gel Technology. J. Chromatogr. A. 978: 37-48. Yu, J., Wu, C. and Xing, J. (2004). Development of New SPME Fibers By Sol-gel Technology for The Determination of Organophosphorus Pesticide Multiresidues in Food. J. Chromatogr. A. 1036: 101-111. Yun, L. (2003). High Extraction Efficiency Soild-phase Microextraction Fibers Coated with Open Crown Ether Stationary Phase using Sol-gel Technique. Anal. Chim. Acta. 486: 63-72. Zeeuw, J. and Luong, J. (2002). Developments in Stationary Phase Technology for Gas Chromatography. Trends Anal. Chem. 21: 594-606. Zeng, Z., Qiu, W., Wei, X, Yang, M., Huang, Z. and Li, F. (2001). Solid-phase Microextraction of Monocyclic Aromatic Amines using Novel Fibers Coated with Crown Ether. J. Chromatogr. A. 934: 51-57. Zhang, Z. and Pawliszyn, J. (1993). Headspace Solid Phase Microextraction. Anal. Chem.65: 1843-1852. 102 Zhang, Z. and Pawliszyn, J. (1995). Quantitative Extraction using an Internally Cooled Solid Phase Microextraction. Anal Chem. 67: 34-43. Zhang, Z., Yang, M. J. and Pawliszyn, J. (1994). New Solvent-free Sample Preparation Techniques Based on Fiber and Polymer Technologies. Environ. Sci. Technol. 28: 569A-574A. Zhang, Z., Yang, M. J. and Pawliszyn, J. (1994). Solid-phase Microextraction. Anal. Chem. 66: 844A-852A. Zhou, F., Li, X. and Zeng, Z. (2005). Determination of Phenolic Compounds in Wastewater Samples using a Novel Fiber by Solid-phase Microextraction Coupled to Gas-chromatography. Anal. Chim. Acta. 538: 63-70. Zuin, V. G., Lopes, A. L., Yariwake, J. H., and Augusto. F. (2004). Application of a Novel Sol-gel Polydimethylsiloxane-Poly(vinyl alcohol) SPME Fiber for Gas Chromatographic Determination of Pesticide Residues in Herbal Infusions. J. Chromatogr. A. 1056: 21-26. 103 Appendix A An enlarged scanning electron micrograph of the fiber surface view at 100 fold magnification 104 Appendix B Scanning electron micrograph: cross-sectional view of the fiber at 1820 fold magnification (Measurement of the film thickness at two different points). Sol-gel derived C8-coating Fused silica base 105 Appendix C Presentation and Publications Parts of this work have been presented at the following symposia and exhibition, 1. Umi K. Ahmad, Abdul Rahim Yacob and Geetha Selvaraju. “A New SPME Adsorbent for the Forensic Analysis of Accelerant Residue”. Paper presented at the Annual Fundamental Science Seminar (AFSS) 2005. Ibnu Sina Institute, UTM, Johor Bahru (4th-5th July 2005). 2. Umi K. Ahmad, Abdul Rahim Yacob and Geetha Selvaraju. “A Home-made SPME Fiber Coating for Arson Analysis”. Abstracts of the Simposium Kimia Analisis Malaysia (SKAM 18), Hyatt Hotel Regency, Johor Bahru (12th-14th September 2005). 3. Umi K. Ahmad, Geetha Selvaraju and Yong Teik Yau. “FICONSICA: A Durable Forensic Fiber for Fire Debris Analysis”, Abstracts of the 17th International, Invention, Innovation, Industrial Design and Technology Exhibition (ITEX), Kuala Lumpur Convention Centre (KLCC), (19th-21st May 2006). 4. Umi K. Ahmad and Geetha Selvaraju. “Optimization of Headspace-SPME using Lab-made C8-coated fiber for the Forensic Analysis of Arson Accelerants in Fire Debris”. Paper presented at the 1st Regional Postgraduate Conference on Engineering and Science (RPCES), Universiti Teknologi Malaysia, Skudai, Johor (26th-27th July 2006).