Discrimination of matches using ATR-FTIR spectroscopy and chemometrics

• Main Author: Azman, Nur Adila
• Format: Thesis
• Language: English
• Published: 2022
• Subjects: QD71-142 Analytical chemistry
• Online Access: http://eprints.usm.my/58499/1/NUR%20ADILA%20AZMAN-24%20pages.pdf; http://eprints.usm.my/58499/

Matches are one of the trace evidence that could be encountered at the crime scenes of arson cases. Forensic analysis of matches may serve as crucial corroborating evidence to link the suspect to the crime scene or the victim. Nonetheless, matches evidence has been overlooked, due to presence of other prominent evidence, often contaminated with other fire post-debris, and trace in quantity. Furthermore, there has been very limited research in the past on the analysis of matches for forensic applications. In the present study, a rapid, non-destructive, sensitive, and reliable analytical technique namely Attenuated Total Reflectance-Fourier Transform-Infrared (ATR-FTIR) spectroscopy has been utilised to differentiate 11 brands of matches samples. Principal Component Analysis- Linear Discriminant Analysis (PCA-LDA) was employed to generate objective interpretation of the results. The 95.8% accuracy of predictive PCA-LDA model was established with almost all matchstick, unburnt match head, and burnt match head correctly classified to their respective brands. The blind test was conducted with two unknown samples (pretended to be unknown to the analyst) in two different conditions, which were correctly link with their respective source. The current work demonstrated a quick, cost-effective, robust, and nondestructive method for differentiation of matches samples in forensic investigation.