Avaliação de propriedades físicas e químicas de óleos vegetais comestíveis empregando-se análise multitabelas

Abstract

We currently have a marked increase in market demand for vegetable oils from a variety of natural sources. Vegetable oils are extracted mostly from plant seeds, consisting of glycerol esters and a mixture of fatty acids, esters, tocopherols and mineral residues. Fast and non-destructive measurements of food quality data are advancing due to the progress of spectroscopy, for research researchers to obtain important data of the chemical and physical components of food. In this sense, chemometrics collaborates to transform the analytical signals of complex samples into useful information, the chemometric method used in the studies was the multi-table analysis (ComDim), whose objective was to evaluate tables (or matrices) of data acquired for the same samples (that is, a set of data arrays with the same number of rows, but not necessarily the same number of rows). With the application of the ComDim method, informative graphs were obtained that showed the relation between the samples and their variables, besides allowing to evaluate in which of the analyzed tables the predominant information is found. As a result of the first research, the ComDim method allowed the realization of an unsupervised pattern recognition, aiding in the identification of similarities and differences between samples of oils and olive oils. For rice oils it was possible correlates UV-Vis and NIR spectral regions, also physical-chemical analysis and attribute the start and advanced stages for degradation reactions.