Metodologia para identificação dos óleos vegetais utilizados como matéria-prima na produção de biodiesel por espectroscopia no infravermelho aplicando métodos multivariados de análise.

Abstract

The main objective of this study was to develop a methodology for identification of vegetable oils used as raw material in the production of biodiesel by infrared spectroscopy applying multivariate methods of analysis. The study was conducted in three main steps: identification of vegetable oils (soybean, canola, sunflower, rice, olive oil and corn) and degummed oils (canola, sunflower, soybean, cotton, palm and corn) and identification and classification of different samples of biodiesel (canola, sunflower, soybean, cotton, palm and corn). Three samples of B100 were also evaluated obtained predominantly from soy oil. Techniques were employed in mid-infrared spectroscopy and multivariate analysis associated with the next, such as principal component analysis (PCA), hierarchical cluster analysis (HCA), principal component analysis interval (iPCA) and for independent and flexible modeling of Anolog class (SIMCA). The PCA, HCA and SIMCA were performed using the computer program Pirouette 3.11 and iPCA through the Matlab 7.11.0. To degummed oils were also used data obtained from analysis by gas chromatography using a Shimadzu GC Schimadzu - detector GCMS 2010 - QP 2010 Plus Gas chromatograph Spectrometer. The samples degummed oils and biodiesel were characterized by analysis using methods standardized by the AOCS (American Oil Chemists Society) and EN (European Norm). The mid-infrared spectra were acquired on a Perkin Elmer infrared spectrophotometer model Spectrum 400 Fourier transform from a universal attenuated reflectance accessory (UATR-FTIR). We used the interval between 4000 to 650 cm-1, with a resolution of 4 cm-1 and 32 scans. The near-infrared spectra were acquired on the same equipment from an attachment containing the integrating sphere and detector using indium-gallium-arsenic (InGaAs). We used the interval between 10000 to 4000 cm-1, with a resolution of 4 cm-1 and 32 scans. The results indicate that it is possible to develop a methodology for identifying and degummed vegetable oils used as raw material in the production of biodiesel by infrared spectroscopy applying multivariate methods of analysis. It is noteworthy that the results obtained with the degummed oil samples showed that both mid-infrared and near infrared, the palm and cotton samples differ from other samples, a fact justified by a higher amount of saturated fatty acids in their composition. Also this aspect was observed in samples of biodiesel in both spectral ranges evaluated. By comparing samples of degummed oils and biodiesel in the mid-infrared through the PCA, 10 it is clear that the source of vegetable oil has the same influence on the major components in samples corresponding biodiesel. Although it was observed that the iPCA find the best possible spectral range for separation of samples of biodiesel in the mid-infrared, and by this result, the SIMCA modeling allowed us to 100% of the batches of biodiesel from soybeans.