Impressão digital e identificação de adulterações em amostras de interesse forense aplicando paper spray mass spectrometry: perfume e uísque

• Ano de defesa: 2017
• Autor(a) principal: Janaina Aparecida Reis Teodoro
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: autenticidade de perfume; espectrometria de massas com ionização por paper spray; autenticidade de uísque; PLS-DA; Identificação de falsificações; fingerprints; Química analítica; Uísque Adulteração e inspeção; Perfumes Adulteração e inspeção; Espectrometria de massa; Quimica legal; Quimiometria
• Link de acesso: http://hdl.handle.net/1843/SFSA-ALCQSL

The authenticity of food, cosmetics and other matrices has been a worldwide concern because of the economic losses and potential health risks of individuals who consume these products manufactured under bad conditions and with improper inputs. In Brazil, this situation is observed in cases of drinks, perfumes and other samples of forensic interest forgeries that occur frequently and in different ways.These frauds are hardly perceived by consumers because of the similarity and quality of counterfeits. Methods for detecting such frauds are not readily available to regulatory organs and chemical analysis is an alternative that guarantees unambiguous proof of the authenticity of a particular product. In this work, mass spectrometry with paper spray ionization (PS-MS) is suggested as a viable option for obtaining chemical profiles (fingerprints) for fast, simple and low-cost for counterfeitdetection in perfumes and whiskeys. The counterfeited samples used were real and obtained from police agencies through seizure. The obtained fingerprints presented different profiles for the two classes of samples (original and counterfeited), being possible the visual determination of authenticity, but for a clear and assertive view ofall the results simultaneously, the spectrometric data were submitted to chemometric tools for the differentiation of their classes using partial least squares discriminant analysis (PLS-DA). Merit figures (FOM) to evaluate the performance of this methodology were calculated. The optimized models have resulted in the correct classification of all samples and perfect FOM values for perfumes and values veryclose to the ideal for whiskeys, having only one sample misclassified. The method presents potential to be applied in routines of inspection laboratories and criminal investigations to detect falsifications in perfumes and whiskey and could be expanded for application in other types of fraud.