Rifle bullets comparison by wavelength dispersive X-ray fluorescence spectroscopy and chemometric analysis

• Published in: Forensic science international Vol. 325; p. 110880
• Main Authors: Saide, Victor G.P., Viegas, Gabriel M., Canuto, André V.S., Barra, Cristina M., Shimamoto, Gustavo G., Tubino, Matthieu, Rocha Junior, José G.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.08.2021; Elsevier Limited
• Subjects: Algorithms; Ammunition; Cartridges; Chemometrics; Cluster analysis; Criminal investigations; Firearms; Fluorescence; Fluorescence spectroscopy; Forensic ballistics; Forensic science; Forensic sciences; Fragments; HCA; K-NN; Manufacturing; PCA; Principal components analysis; Projectiles; SIMCA; Wavelength; WDXRF; X ray fluorescence analysis; X-ray fluorescence; X-ray scattering; K-NN; WDXRF; SIMCA; HCA; PCA
• ISSN: 0379-0738; 1872-6283
• DOI: 10.1016/j.forsciint.2021.110880

•Wavelength dispersion X-ray fluorescence was used for the analysis of rifle bullets.•The spectral data of rifle bullets were treated by chemometric methods.•Principal components analysis discriminated the five groups of rifle bullets.•Rifle projectiles were correctly classified into five groups using different models. Despite Federal Bureau of Investigation Laboratory announcing the discontinuation of bullet lead examinations, knowledge of the composition of the bullets has been used as an alternative means of identifying their origin, achieving success in some case studies. In this work, wavelength dispersion X-ray fluorescence (WDXRF) and chemometrics were used for the analysis of rifle bullets, in order to identify the spectral similarities of these samples. For this purpose, 54 lead core fragments from 7.62 mm rifle bullets from 5 different manufacturers were obtained: Companhia Brasileira de Cartuchos (CBC), Israel Military Industries (IMI), Federal Cartridge (FC), Fray Luiz Beltrán (FLB) and Zavod Vlasim (ZV). Principal components analysis (PCA) discriminated the five groups of bullets according to their manufacturers in a three-dimensional scores graph, where 3 principal components accounted for>99% of the variability between the samples. The spectral region for Sb and the scattering region together proved to be determinant for discrimination of the groups. The dendrogram presented in the hierarchical cluster analysis (HCA) showed the formation of five groups. The k-nearest neighbor algorithm (k-NN) and soft independent modeling of class analogy (SIMCA) correctly classified all samples of the test set. X-ray scattering spectrum were used for the first time in the analysis of the fragments and contributed to the grouping of samples from the same manufacturers. The results indicate that the WDXRF technique is suitable for forensic purposes in case studies, as, besides being quick and relatively simple, it has the advantage of preserving evidence.