Doctor of Philosophy (Applied Science) University of Canberra

Transcription

1 A thesis submitted for Doctor of Philosophy (Applied Science) University of Canberra Brenda Woods Bachelor of Technology (Forensic and Analytical Chemistry) Bachelor of Science (Honours) An Examination Procedure for Forensic Soil Analysis in Trace Evidence Laboratories December 2014

2 Abstract Currently, in Australia, the majority of forensic soil examinations are not conducted within forensic facilities but by a limited number of experienced soil scientists. The research presented in this thesis developed an examination procedure for forensic soil samples in a trace evidence laboratory utilising instrumentation commonly available in such a laboratory, including microspectrophotometry (MSP), attenuated total reflectance (ATR) Fourier transform infrared spectroscopy (FTIR) and elemental analysis using X-ray fluorescence spectroscopy (XRF), scanning electron microscopy/energy dispersive X-ray spectroscopy (SEM/EDX) or laser induced breakdown spectroscopy (LIBS). Further, a suggested soil examination protocol that integrates the skill set of crime scene examiners, trace evidence scientists and specialist geologists was established. In this study, 29 soil specimens were analysed, with 12 specimens coming from six geologically similar sites in the Canberra area and the remaining 17 specimens previously collected from other sites around Australia by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) and stored as part of the CSIRO National Soil Archive. The soil specimens used in this study were selected to assess the examination procedure against both soil specimens from similar geological environments (the Canberra area soils) and soil specimens from vastly different geological environments (the CSIRO soils). The 29 Australian soil samples analysed could be classified into 21 different groups based on colour analysis using MSP and L*a*b* chromaticity coordinates. Using ATR-FTIR, the organic and inorganic components of the soil sample can be readily analysed and compared, resulting in the generation of an IR spectrum of the whole soil sample. Based on this spectrum, the ATR-FTIR can be used to screen and discriminate soil samples with as little as micrograms of material. Across the 29 soil samples analysed in this way, a very high discrimination of 99.7% differentiation was achieved. iii

3 A high level of discrimination of the soil sample set analysed was achieved for all the elemental techniques (92.4% discrimination for LIBS, 98.5% discrimination for XRF and 99.5% discrimination for SEM/EDX). This study has demonstrated that these elemental profiling techniques can be readily applied to detect differences in the elemental compositions of soil specimens. Individually, each of the instrumental techniques demonstrated high discrimination of the soil set analysed. However, when combined, full discrimination of the soil set analysed was achieved with MSP followed by IR and then one of the elemental profiling methods. Utilising the proposed instrumental examination procedure from this study (i.e. dry sieving followed by MSP, ATR-FTIR and elemental analysis), the following is a suggested soil examination protocol that integrates the skill set of crime scene examiners, trace evidence scientists and specialist geologists to enable best practice for the triaging and examination of forensic soil samples. This protocol consists of four distinct stages, the first three of which could take place in typical forensic trace evidence laboratories: Stage 1 Known soil sample collection and preservation by crime scene examiners. Stage 2 Questioned soil sample collection and preservation by trace evidence chemists during the examination of items submitted for forensic analysis. Microscopic examination of the soil sample for exogenous material followed by preliminary screening of dry sieved questioned and known soil samples for differences using instrumentation currently available in trace evidence laboratories, including MSP, IR and a form of elemental analysis, either LIBS, XRF or SEM-EDX. Stage 3 Limited identification of minerals present in soil specimens coupled with the instrumental examination procedure from Stage 2. Note that, in order to complete Stage 3, typical trace evidence analysts would require some level of specialist training with respect to basic mineralogy. iv

4 Stage 4 Geological and/or palynological examinations by a subject matter expert for casework that is complex, where provenancing is required, when the questioned sample is particularly small, or when additional instrumental expertise is required. The proposed instrumental examination procedure (dry sieving the soil sample, followed by MSP, ATR-FTIR and an elemental analysis) allows for high discrimination and differentiation of soil samples, without destructively altering the soil, thereby retaining an unadulterated soil sample for the soil scientist so they can complete their analyses. The proposed instrumental examination procedure will develop stage 2 skills and promote the value of soil examination with investigators. Validation of the proposed instrumental examination procedure is underway for implementation of the procedure in the AFP Chemical Criminalistics trace evidence laboratory. The proposed soil protocol would ensure integration of the stages undertaken by the trace evidence examiner in a forensic science laboratory and the higher-level analyses that could be performed by geological and/or palynological experts. The re-integration of soil examinations into forensic laboratories using the proposed instrumental examination procedure as part of the larger soil examination protocol will allow for the objective differentiation of soil samples in a timely and efficient manner. This would facilitate soil examinations to assist in intelligence-led policing, rather than the more typical reactive forensic analyses. v

5 Acknowledgements I would like to thank Forensics, Australian Federal Police and the Chemical Criminalistics Team for allowing me the time and support required to enable me to take this scientific journey. For their assistance with keeping my scientific journey in perspective, I truly need to thank my friends and family. I would like to thank previous PhD recipients for their infinite words of wisdom, especially Dr. Vincent Otiento-Alego, Dr. Naomi Speers, Dr. Sonia Taflaga and Dr. Jane Hemmings. Some noteworthy wisdom included: Don t be disappointed by the false peaks. Acknowledge them for the challenge they were and congratulate yourself on your achievement. Then get back to work. Never make a life changing decision while doing your PhD. Yep. That s normal Finally and most importantly, to my wonderful supervisors, James Robertson, Chris Lennard and Paul Kirkbride. Without their constant enthusiasm, support, assistance and nurturing the three journal articles, the book chapter and this thesis never would have been written. Thank you for holding my hand during this journey. I do not think I will ever be able to thank you enough for your assistance in my growth as a scientist. Thank you. vii

6 Publications and Presentations Arising from this Thesis Peer Reviewed Journal Articles Brenda Woods, Chris Lennard, K. Paul Kirkbride and James Robertson, Soil examination for a forensic trace evidence laboratory - Part 1: Spectroscopic techniques. Forensic Science International, : p DOI: /j.forsciint Brenda Woods, K. Paul Kirkbride, Chris Lennard, and James Robertson, Soil examination for a forensic trace evidence laboratory - Part 2: Elemental analysis. Forensic Science International, : p DOI: /j.forsciint Brenda Woods, Chris Lennard, K. Paul Kirkbride and James Robertson, Soil Examination for a Forensic Trace Evidence Laboratory - Part 3: From Research to Casework. Forensic Science International, Submitted for publication in Forensic Science International. Book Chapter Brenda Woods, Chris Lennard, K. Paul Kirkbride and James Robertson, Reinstating Soil Examination as a Trace Evidence Sub-discipline. Criminal and Environmental Soil Forensics, CRC Press, Submitted for Publication Conference Presentations Brenda Woods, Criminalistics Gets Dirty Preliminary Investigation of Microspectrophotometry as a Screening Test for Soils. 6 th European Academy of Forensic Science Conference, The Hague The Netherlands, August 2012 Brenda Woods, Criminalistics Gets Dirty Preliminary Investigation of Microspectrophotometry as a Screening Test for Soils. Australian and New Zealand Forensic Science Society 21 st International Symposium on the Forensic Sciences, Hobart - Australia, September 2012 ix

7 Brenda Woods, Ground breaking dirty research An examination procedure for forensic soil samples. Australian and New Zealand Forensic Science Society 22 nd International Symposium on the Forensic Sciences, Adelaide - Australia, September 2014 Brenda Woods, An examination procedure for forensic soil analysis in trace evidence laboratories. The Geological Society - Forensic Geoscience Group Conference, London - England, December 2014 x

8 Table of Contents Certificate of Authorship of Thesis... i Abstract... iii Acknowledgements... vii Publications and Presentations Arising from this Thesis... ix Table of Contents... xi List of Figures... xvii List of Tables... xxiii Chapter 1 Introduction Summary Introduction Soil Forming Factors... 2 Parent Rock Material... 2 Climate... 3 Topography... 3 Biota... 3 Time Australian Soil Classification Development of Soil as Forensic Evidence Current Methods for Australian Forensic Soil Analysis Macroscopic Examination Colour Texture Particle Sizing Mineral Identification Microscopic Examination Optical Crystallography xi

9 Powder X-ray Diffraction X-ray Florescence Spectroscopy Scanning Electron Microscopy/Energy Dispersive X-ray Spectroscopy Organic Examination Fourier Transform Infrared Spectroscopy Non-instrumental Techniques Forensic Soil Examination in Trace Evidence Laboratories The Role of a Trace Evidence Scientist References Chapter 2 Materials and Methods Soil Samples Canberra Area Soil Samples CSIRO National Soil Archive Soil Samples Analytical Instrumentation Statistical Analysis Principal Component Analysis Linear Discriminate Analysis References Chapter 3 Microspectrophotometry Examinations Summary Introduction Colour in Soils The Munsell Colour System Microspectrophotometry xii

10 CIE L*a*b* Colour Space Materials and Methods Potassium Bromide and Polyethylene Discs Unadulterated Soil Discs Unadulterated Soil in Holder Results Potassium Bromide and Polyethylene Discs Unadulterated Soil Discs Unadulterated Soil in Holder - Milled Unadulterated Soil in Holder - <63 µm Fraction Unadulterated Soil in Holder - <38 µm Fraction Discussion Conclusions References Chapter 4 Infrared Examinations Summary Introduction Infrared Spectroscopy Current Infrared Soil Analysis using KBr and DRIFTS Infrared Examination in a Trace Evidence Laboratory Attenuated Total Reflectance Theory Materials and Methods Results Discussion Conclusions xiii

11 4.7 References Chapter 5 Elemental Analysis Summary Introduction Inductively Coupled Plasma X-ray Fluorescence Spectroscopy Additional Elemental Analysis Techniques Scanning Electron Microscopy/Energy Dispersive X-ray Spectroscopy Laser Induced Breakdown Spectroscopy Elemental Analysis in a Trace Evidence Laboratory Materials and Methods Laser Induced Breakdown Spectroscopy Method Micro-X-ray Fluorescence Spectroscopy Method Scanning Electron Microscopy/Energy Dispersive X-ray Spectroscopy Method Results Laser Induced Breakdown Spectroscopy Results Micro-X-ray Fluorescence Spectroscopy Results Scanning Electron Microscopy/Energy Dispersive X-ray Spectroscopy Results Discussion Conclusions References Chapter 6 Scene Variability Study Summary Scene Variability Study Scenario xiv

12 6.3 Materials and Methods Results MSP Results Scene to Scene Discrimination ATR-FTIR Results Scene to Scene Discrimination XRF Results Scene to Scene Discrimination MSP Results Sample to Scene Discrimination ATR-FTIR Results Sample to Scene Discrimination XRF Results Sample to Scene Discrimination Discussion Conclusions References Chapter 7 From Research to Casework Summary Quality Requirements for Australian Forensic Laboratories Soil Method Validation Reliability Study Repeatability Study Robustness Studies Representative Soil Sample Aggregate Examinations Sample Fraction Wet Sieving Appropriate Use of the Proposed Soil Examination Procedure Scene Sampling Investigative Policing Further Research and Development Directions xv

13 7.4 References Chapter 8 Conclusions Introduction Instrumental Examination Procedure Aggregate Soil Processing Microspectrophotometry Examinations Infrared Examinations Elemental Analysis Examinations Scanning Electron Microscopy/Energy Dispersive X-ray Spectroscopy Micro-X-ray fluorescence Spectroscopy Laser Induced Breakdown Spectroscopy Summary Quality Testing of the Proposed Soil Examination Procedure Scene Variability Blind Trial Repeatability and Robustness Studies Proposed Soil Examination Protocol References Appendix 1 Microspectrophotometry Spectra Appendix 2 Infrared Spectra Appendix 3 Elemental Analysis Spectra Appendix 4 Journal Article Appendix 5 Journal Article Appendix 6 Journal Article 3 Manuscript Appendix 7 Book Chapter Manuscript xvi

14 List of Figures Figure 1.1. Image depicting common Australian soil horizons Figure 1.2. Map of Australia using the Australian Soil Classification system Figure 2.1. Sample collection grid for the Canberra area soil samples Figure 2.2. Google Earth image of the Canberra area with the Canberra area soil collection sites marked Figure 2.3. Photomicrographs of the <38 µm sieve fraction for the Canberra area soil samples Figure 2.4. Google Earth image of Australia with the Australia wide soil collection sites marked Figure 2.5. Photomicrographs of the <38 µm sieve fraction for the CSIRO soil samples Figure 3.1. Soil colours representative of Australian soils Figure 3.2. An image of a Munsell colour chart page Figure 3.2. A representation of a MSP operating in the reflectance mode Figure 3.3. An image of the MSP used in this study Figure 3.3. A representation of the CIE L*a*b* colour space Figure 3.4. A picture of a DRIFTS sample holder empty and containing a soil sample Figure 3.5. MSP spectra of unadulterated soil discs Figure 3.6. MSP spectra of milled soil samples in DRIFTS holder Figure 3.7. Matrix plot of the CIE L*a*b* chromaticity coordinates for milled Canberra area soil samples Figure 3.8. Matrix plot of the CIE L*a*b* chromaticity coordinates for the Canberra area <63 µm sieve fraction soil samples Figure 3.9. Matrix plot of the CIE L*a*b* chromaticity coordinates for the Canberra area <38µm soil fraction Figure MSP spectra for soil samples from CSIRO soil sample set Figure Matrix plot of the CIE L*a*b* chromaticity coordinates for the <38 µm CSIRO soil fractions Figure 4.1. A representation of an ATR-FTIR instrument configuration Figure 4.2. An image of the ATR-FTIR instrument used in this study xvii

15 Figure 4.3. IR spectral variation observed due to the presence of kaolin minerals Figure 4.4. IR spectral variation observed due to the presence of quartz Figure 4.5. Score plots of the first two principal components for the original spectral data for the 12 Canberra area soil samples Figure 4.6. Score plots of the first two principal components for the 1 st derivative spectral data for the 12 Canberra area soil samples Figure 4.7. A section of the score plot of the first two principal components for the original IR spectral data for the CSIRO soil samples showing the grouping into 10 groups Figure 4.8. A section of the score plot of the first two principal components for the 1 st derivative IR spectral data for the CSIRO soil samples showing the grouping into 15 groups Figure 5.1. An image of the SEM/EDX instrument used in this study Figure 5.2. An image of the µxrf instrument used in this study Figure 5.3. An image of the LIBS instrument used in this study Figure 5.4. LIBS elemental ratio data for the Canberra area soil specimens Figure 5.5. LIBS elemental ratio data for the Australia-wide soil specimens Figure 5.6. Micro-XRF data for the Canberra area soil specimens Figure 5.7. Micro-XRF data for the Australia-wide soil specimens Figure 5.8. SEM/EDX data for the Canberra area soil specimens Figure 5.9. SEM/EDX data for the Australia-wide soil specimens Figure 6.1. Image of the known soil sample locations from Site Figure 6.2. Image of the known soil sample locations from Site Figure 6.3. Image of the soil sample location from Site Figure 6.4. Image of Shoes A prior to trace collection Figure 6.5. Image of Shoes B prior to trace collection Figure 6.6. Image of Shoes C prior to trace collection Figure 6.7. Image of Shovel X prior to trace collection Figure 6.8. Image of Shovel Y prior to trace collection Figure 6.9. Image of Shovel Z prior to trace collection Figure Matrix plot of the L*a*b* chromaticity coordinates of the MSP data for the known soil samples from Sites 1 and xviii

16 Figure Score plot for the first two principal components of the ATR-FTIR spectra for the known soil samples from Sites 1 and Figure Score plot for the first two principal components of the 1 st derivative ATR-FTIR spectra for the known soil samples from Sites 1 and Figure Score plot for the first two principal components of the XRF elemental data for the known soil samples from Sites 1 and Figure Matrix plot of the L*a*b* chromaticity coordinates of the MSP data for the soil recovered from the shoes and the shovels as well as the known soil samples from Sites 1 and Figure Matrix plot of the L*a*b* chromaticity coordinates of the MSP data for the soil recovered from Shoes A and C, and Shovels X and Y as well as the known soil samples from sites 1 and Figure Score plot for the first two principal components of the 1 st derivative ATR-FTIR spectra for the soil recovered from Shoes A and C, and Shovels X and Y as well as the known soil samples from Sites 1 and Figure Score plot for the first two principal components of the elemental data for the soil recovered from the shoes and the shovels as well as the known soil samples from Sites 1 and Figure 7.1. Matrix plot of CIE L*a*b* chromaticity coordinates for the MSP data from the blind trial samples and the soil database specimens from which the blind trial samples could not be differentiated Figure 7.2. PCA score plot of the first two principal components of the ATR-FTIR spectral data for the blind trial samples and the soil database specimens from which the blind trial samples could not be differentiated Figure 7.3. LIBS data for the blind trial samples and the soil database specimens from which the blind trial samples could not be differentiated Figure 7.4. XRF data for the blind trial samples and the soil database specimens from which the blind trial samples could not be differentiated Figure 7.5. SEM-EDX data for the blind trial samples and the soil database specimens from which the blind trial samples could not be differentiated Figure 7.6. Matrix plot of the MSP results for the repeatability study xix

17 Figure 7.7. PCA score plot of the first two components of the original ATR-FTIR spectral data for the repeatability study Figure 7.8. PCA plot of the first two components of the 1 st derivative ATR-FTIR spectral data for the repeatability study Figure 7.9. Graph of the elemental XRF data for the repeatability study Figure Graph of the elemental LIBS ratios for the reapeatability study Figure Graph of the elemental SEM/EDX data for the repeatability study Figure Matrix plot of the MSP results for the robustness study Figure PCA plot of the first two components of the original ATR-FTIR spectral data for the robustness study Figure PCA plot of the first two components of the 1 st derivative ATR-FTIR spectral data for the robustness study Figure Matrix plot of the MSP results for the dry sieved Canberra area soil samples. 133 Figure Matrix plot of the MSP results for the wet sieved Canberra area soil samples Figure PCA plot of the first two components of the 1 st derivative ATR-FTIR spectral data for the dry sieved Canberra area soil samples Figure PCA plot of the first two components of the 1 st derivative ATR-FTIR spectral data for the wet sieved Canberra area soil samples Figure A1.1. MSP spectra using KBr method in reflectance mode Figure A1.2. MSP spectra using polyethylene disc method in transmission mode Figure A1.3. MSP spectra using milled method in reflectance mode Figure A1.4. MSP spectra using <38 µm method in reflectance mode Canberra area soils Figure A1.5. MSP spectra using <38 µm method in reflectance mode Australia wide soils Figure A2.1. IR spectra using DRIFTS mode Figure A2.2. IR spectra using KBr disc mode Figure A2.3. IR spectra using milled ATR mode Figure A2.4. IR spectra using <38 µm ATR mode Canberra area soils Figure A2.5. IR spectra using <38 µm ATR mode Australia wide soils xx

18 Figure A3.1. Site 1 surface XRF spectrum Figure A3.2. Site 4 surface XRF spectrum Figure A3.3. Site 6 surface XRF spectrum Figure A3.4. Site 7 XRF spectrum Figure A3.5. Site 11 XRF spectrum Figure A3.6. Site 17 XRF spectrum Figure A3.7. Site 21 XRF spectrum Figure A3.8. LIBS spectra Canberra area soils Figure A3.9. Site 1 surface LIBS spectrum Figure A3.10. Site 4 surface LIBS spectrum Figure A3.10. Site 6 surface LIBS spectrum Figure A3.11. LIBS spectra Australia wide soils Figure A3.12. Site 7 LIBS spectrum Figure A3.13. Site 11 LIBS spectrum Figure A3.14. Site 17 LIBS spectrum Figure A3.15. Site 21 LIBS spectrum Figure A3.16. Site 1 surface SEM/EDX spectrum and associated SEM image Figure A3.17. Site 4 surface SEM/EDX spectrum and associated SEM image Figure A3.18. Site 6 surface SEM/EDX spectrum and associated SEM image Figure A3.19. Site 7 SEM/EDX spectrum and associated SEM image Figure A3.20. Site 11 SEM/EDX spectrum and associated SEM image Figure A3.21. Site 17 SEM/EDX spectrum and associated SEM image Figure A3.22. Site 21 SEM/EDX spectrum and associated SEM image xxi

19 List of Tables Table 1.1. Australian Soil Classification System Table 1.2. Five class divisions of the Atterberg Scale Table 2.1. Canberra area soil collection site details Table 2.2. CSIRO soil sample details Table 3.1. Classification of the milled Canberra area soil samples into 11 groups based on MSP Table 3.2. Classification of the <38 µm Canberra area soil fractions using MSP Table 4.1. Classification of the 12 Canberra area soil samples into 3 groups based on IR spectra Table 4.2. Classification of the 12 Canberra area soil samples into groups based on the first two principal components for both the original and 1 st derivative IR spectral data Table 4.3. Classification of the CSIRO soil samples into groups based on the first two principal components for both the original and 1 st derivative IR spectral data Table 5.1. Peak ratios used for LIBS data analysis Table 5.2. Summary of the discrimination of LIBS, XRF and SEM/EDX for the soil specimens analysed in this study Table 6.1. Scene variability sample collection site details Table 6.2. The amount of soil trace collected from the shoes and shovels Table 6.3. Summary of the scene variability study Table 7.1. Summary of LDA classification of the CIE L*a*b* chromaticity coordinates of the MSP data for the blind trial Table 7.2 Summary of LDA classification of the original IR spectral data for the blind trial Table 7.3. Means and standard deviations for the CIE L*a*b* chromaticity coordinates from the MSP repeatability study Table 7.4. Means and standard deviations for the elemental XRF data from the repeatability study Table 7.5. Means and standard deviations for the elemental LIBS ratios from the repeatabilitystudy xxiii

20 Table 7.6. Means and standard deviations for the elemental SEM/EDX data from the repeatability study Table 7.7. The means and standard deviations for the CIE L*a*b* chromaticity coordinates from the MSP robustness study Table 7.8. Summary of LDA classification of the CIE L*a*b* chromaticity coordinates for the MSP robustness study Table 7.9. Summary of LDA classification of the original IR spectral data for the robustness study Table Summary of LDA classification of the 1 st derivative IR spectral data for the robustness study Table 8.1. Discrimination values for each analytical method and for the combined procedure when applied to the soil sample set xxiv