Qualidade físico-química de grãos de soja classificada com e sem defeitos físicos utilizando espectroscopia de infravermelho próximo

Abstract

The soy (Glycine max L.) is a nutritious grain widely used in human and animal diets, biodiesel production, among other products and by-products. Given its global importance, it is essential to apply methodologies for the efficient monitoring of the physical-chemical parameters that determine the quality of grains in order to assist in classification. Near Infrared Spectroscopy (NIR) is a fast and non-destructive method for food analysis. Thus, the study aimed to analyze the feasibility of using near-infrared spectroscopy to identify and predict the quality of soybeans with and without defects, replacing the conventional physical classification method. Defects were separated according to technical classification standards at the Agricultural Products Postharvest Laboratory (LAPOS), located at UFSM-CS. Next, the grains were subjected to grinding and the proximate composition was subsequently analyzed using NIR. Pearson's correlation analysis identified negative relationships between proteins and fibers, as well as between moisture and ash, the cluster identified similar behavior in the proximate composition of healthy grains with fermented grains. Therefore, Near Infrared Spectroscopy showed positive results to assist in the classification of soybeans, characterizing the quality of the grains according to the physicochemical composition.