Aplicação do método quimiométrico PLS-DA para classificar bioquerosene de aviação de macaúba através de espectroscopia MIR

Abstract

Due to the growing global energy demand and the ongoing prevalence of non-renewable and finite sources, the search for renewable sources, primarily biomass-derived, is essential to meet this demand. One sector that has sparked significant research interest is aviation, where large volumes of kerosene are consumed as the primary fuel. Thus, research is advancing toward the development of alternative fuels to partially or fully replace kerosene. In Brazil, the National Agency of Petroleum, Natural Gas, and Biofuels (ANP) under Law No. 12,490/2011, and also in ANP Resolution No. 856 of October 22, 2021 - DOU of October 25, 2021, defines bio-kerosene as a substance derived from renewable biomass that can be used in aircraft turbine consumption and establishes a limit for the addition of up to 50% (v/v) of bio-kerosene in aviation kerosene. In this context, the project aims to develop and validate a method using MidInfrared Spectroscopy (MIR) combined with the chemometric method of Partial Least Squares Discriminant Analysis (PLS-DA) to classify samples based on the macaúba bio-kerosene content. The model showed classification accuracy of 100% in both calibration and prediction sets, with errors: RMSEC (Root Mean Square Error of Calibration) = 0.62; RMSECV (Root Mean Square Error of Cross-Validation) = 0.65, and RMSEP (Root Mean Square Error of Prediction) = 0.63, these results indicate excellent performance, making it a promising method for routine analysis to detect biokerosene content in mixtures through a reliable, robust, and easy-to-interpret approach.