Aplicação de espectroscopia no infravermelho médio e análise discriminante por mínimos quadrados parciais para classificação de biodieseis e misturas biodiesel /diesel

Abstract

Faced with the economic stimulus generated by agribusiness around the production of biofuels and environmental benefits, biodiesel production has been gaining market making it necessary quality control throughout its chain of production and distribution. In Brazil, such supervision is done by the National Agency of Petroleum, Natural Gas and Biofuels (ANP) establishing parameters based on international standards, eg, American Society for Testing and Materials (ASTM). It is found, in literature, great amount of studies aimed quality control in the production of biodiesel and blends with diesel. However, most of these techniques require large quantities of reagents, used are time consuming and ue expensive analytical instruments. Within this context, Spectroscopic techniques, especially in the infrared are presented as a rapid, direct, non-destructive, clean and efficient alternative, and can be applied in situ quality control of biodiesel in all stages of the production chain and supervision. The potential of these techniques become almost limitless when combined with the use of chemometric methods to extract relevant information from complex data making them simpler to be applied in routine laboratories. Given the situation described, the present work ims at the use of spectroscopy in the mid infrared, combined the technique of discriminant analysis classification by partial least squares (PLS-DA) for the development of analytical method for classifying samples according biodiesels with their energy relative to pure biodiesels (B100) and their blends with diesel content of 5% biodiesel (B5). , Classification models for compliance with the contents of B5 were also developed. The usage of Spectroscopy in the Middle Infrared allied to the method of discriminant analysis by Partial Least Squares (PLS-DA) was able to sort through nearly 100% efficiency the samples of interest of the models developed in this work. So your application becomes feasible for the monitoring and supervision of the quality of these types of fuels.