Tratamento metrológico e quimiométrico de dados de ressonância magnética nuclear para amostras de biodiesel em diesel

• Ano de defesa: 2012
• Autor(a) principal: Gambarra Neto, Francisco Fernandes
• Orientador(a): Ferreira, Antonio Gilberto
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Química - PPGQ
• País: BR
• Palavras-chave em Português: Ressonância magnética nuclear; Medição; Quimiometria; Química analítica
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::QUIMICA
• Link de acesso: https://repositorio.ufscar.br/handle/ufscar/6228

The development of instrumentation in Nuclear Magnetic Resonance (NMR) increased the sensitivity of the technique, allowing samples in micro and nanogram scale can be examined. These new technologies enabled the NMR technique was seen as an analytical tool for the quantification of substances in addition to their use in identifying as traditionally did. Quantitative Nuclear Magnetic Resonance (qNMR) is a very important area that is not widely used in analytical chemistry, with the need for a validation study to verify its ability to quantify by a metrological point of view. The application and validation of methods minimize the potential problems associated with the analytical method and also lends credence to the method studied, providing comprehensive information about many important aspects that reflect the behavior and reliability of the method. The validation studies of methods RMNq were performed in accordance with ISO / IEC 17025 and ISO GUM. The validation of the method is followed by measuring the uncertainty of measurement qNMR and its importance has been widespread in the analytical chemists area and also in the chemical society in general. The rules for establishing chemical activity as a source of patterning the selectivity studies, investigating the effects of interferents likely in the samples, and for the NMR technique, it is superior to many other analytical instruments. Therefore, the objective of this study was to provide quantitative and alternatives tools, taking into account the metrological quantification of substances by 1H NMR, which is something absolutely new for the Brazilian market. The measure of the selectivity was performed using quantification chemometrics in the spectra of 1H and 13C NMR. Metrology of the main interfering in the 1H NMR technique was studied to quantify the levels of biodiesel in commercial samples and standards of diesel in different proportions in the mixture of biodiesel BX. This study led to the separation of two sets of samples which, by classification using the chemometrics algorithm Successive projection algorithm - Linear Discriminant Analysis (SPA-LDA), are widely separated in diesel fuel containing sulfur in 1800 and 500 ppm . The selectivity was calculated by the quantification chemometric by the algorithm SPA - Multiple linear regression (SPA-MLR) of sulfur content, the density at 20°C, the distillation temperature to 50% and 80% of diesel oil, the cetane number and flash point compared to the results obtained by laboratory (Laboratório de Combustíveis do CCDM/DEMa Universidade Federal de São Carlos) accredited by the ANP (Agência Nacional de Petróleo e Biocobumstíveis).