An?lise multivariada como ferramenta na avalia??o de res?duos de armas de fogo utilizando fluoresc?ncia de raios-X

• Ano de defesa: 2017
• Autor(a) principal: Madeira, Fernanda Bomfim
• Orientador(a): Rocha Junior, Jose Geraldo
• Banca de defesa: Rocha Junior, Jos? Geraldo, Santos, Ariana dos, Casartelli, Evelton Alves
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal Rural do Rio de Janeiro
• Programa de Pós-Graduação: Programa de P?s-Gradua??o em Qu?mica
• Departamento: Instituto de Ci?ncias Exatas
• País: Brasil
• Palavras-chave em Português: quimiometria; qu?mica forense; SIMCA; PCA; HCA
• Palavras-chave em Inglês: chemometrics; forensic chemistry; SIMCA; PCA; HCA
• Área do conhecimento CNPq: Qu?mica
• Link de acesso: https://tede.ufrrj.br/jspui/handle/jspui/4331

In this work, Wavelength Dispersion X-Ray Fluorescence Spectrometry (WDXRF) was used as an alternative technique for the analysis of gunshot residue (GSR), as it is a non-destructive, fast technique and shows chemical information related to the organic and inorganic compounds of the samples. In order for unfair convictions to occur less frequently, it is necessary to correlate the gun used in the crime with the suspect or at least to determine whether the suspect actually used any gun. To perform this correlation and to recognize the patterns in the data obtained in the analysis, chemometrics models (PCA, HCA and SIMCA) were constructed in order to discriminate the GSR according to the guns (Imbel Pistol 9 mm, Taurus Revolver .38, and rifle FN FAL ) and the cartridges used (SM Winchester, Speer copper bullet, Winchester SPL (silver) and CBC SPL + P gold), and differentiate them from residues coming from individuals who did not fire any guns (such as mechanics, painters, electricians and gas station attendants), aiming to assist the work of the police for a more reliable correlation between the suspect and the shooter. The residues were collected in swabs soaked in EDTA solution 2 g 100 mL-1 and analyzed by WDXRF. The models created with PCA, using the bands of the organic (scattering region) and inorganic elements, were able to discriminate between types of arms for the samples of 3 and 2 shots together, and to discriminate between cartridges for the samples of 3 shots. The hierarchical classification HCA together with PCA assisted in the construction of a SIMCA model, which correctly classified the innocents, by employing three training classes (1 for GSR samples and 2 for innocent samples), without classifying them in the GSR class