Aplicação de técnicas quimiométricas para identificação da adulteração de diesel S10-B7 por óleos vegetais brutos

• Autor(a) principal: Câmara, Anne Beatriz Figueira
• Data de Publicação: 2016
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Repositório Institucional da UFRN
• Texto Completo: https://repositorio.ufrn.br/handle/123456789/38334
• topic: Diesel S10.; Adulteração.; FT-IR.; NIR.; Quimiometria.; MCR-ALS.; Chemometrics.; Adulteration.; Química

In the national scene, diesel oil stands out as the most frequently used fuel, having their main use in road transportation of passengers and cargo. One of the biggest problems associated with his use is adulteration with low cost products, as vegetable oils and petrochemical solvents. The use of adulterated fuel brings many losses, both financial and environmental, therefore it is necessary that occurs the development of fast and efficient methods for the detection and quantification of these potential adulterants. Thus, in this work were used Spectroscopic methods associated with chemometrics techniques, such as Multivariate Curve Resolution Alternating Least Squares MCR-ALS, for the identification and quantification of vegetable oils, added as adulterants in diesel fuel commercialized. Were prepared mixtures of the fuel with the adulterant, in various volumetric concentrations, from 1 to 60%, and then were obtained the physicochemical properties specified by the National Agency of Petroleum Natural Gas and Biofuels - ANP, such as specific gravity, kinematic viscosity, flash point and atmospheric distillation, according to standards by American Society for Testing and Materials International (ASTM). Besides that, was used Fourier transform infrared, FTIR, and near infrared, NIR, for the creation of two models involving multivariate calibration, using the MCR-ALS algorithm. The results obtained for the physical-chemical properties were not efficient for the identification of adulteration with the insertion with small amounts of oil, 15% (v/v), given that under these conditions, the results obtained still remain within the specifications. Among these, distillation ASTM D86 was presented as the best parameter to identify the adulteration, since it performs a compositional analysis of the mixture, but it is an expensive technique and demand a high analysis time. Both models created with the MCR-ALS, both the FT-IR and to the NIR, were efficient in quantifying the adulterant in mixtures, presenting correlation coefficients close to 1 and low errors of calibration and prediction, which indicates a good fit. The model was also effective in identifying adulteration.