Broad Range FTIR Spectroscopy and Multivariate Statistics for High Energetic Materials Discrimination

The objective of this work was to measure infrared spectra of high explosive materials (HE) in wide spectral range in order to acquire information for their complete characterization and find out the regions that are the most discriminatory for each material. Four HEs were measured by means of Fourier Transform Infrared (FTIR) spectroscopy in a very broad range (from near- via mid- to far-IR). Obtained spectra were subsequently evaluated using multivariate statistical methods for dimension reduction and results grouping. Clustering was assessed in terms of compactness and stability in order to distinguish which region or regions are most suitable for the identification based on spectral signature. Based on outcomes of visualization method (silhouette plot) used to compare results of implemented chemometric methods (HCA, PAM, and PCA) done on FTIR spectra collected for four high explosive materials (PETN, C-4, RDX, and TNT) within all regions, it seems that the mid-IR region is the most informative for the distinction among analyzed HE materials based on substance spectral signatures. However, it is worth noticing that also the near-IR region can be used for good differentiation.