Multivariate statistical methods in the forensic investigation of the post-blast residues measured by Fourier transform infrared spectroscopy

Abstract

Fourier transform infrared (FTIR) spectroscopy has gained significant attention among the forensic scientists because it shows high sensitivity and selectivity, and offers near-real-time detection. Application of the multivariate statistical techniques for the analysis of the spectra is necessary in order to enable feature extraction, proper evaluation and identification of obtained spectra. In this paper we show the development of a feasible procedure for the characterization of spectroscopic signatures of the explosive materials in the remnants after explosion. In our research especially designed and prepared sample catchers were used during the blasts of three various high explosives: C-4, TNT and PETN. Principal component analysis (PCA) was performed using broad spectral data range (600–4000 cm−1) for sample classification into separate classes. Most of the information contained in spectral data was compressed by PCA in few relevant principal components that explain most of the variance of spectral data. The results show that FTIR spectroscopy in combination with multivariate methods are well suited for identification and differentiation purposes even in very large data sets and could be employed by forensic laboratories for rapid screening analysis.