Utilização de métodos quimiométricos aliados a RMN na caracterização dos diferentes tipos de cafés comerciais
Abstract
Coffee is an important product of Brazilian agribusiness. Nevertheless, the commercial evaluation of its quality is based upon subjective methods and this requires the methodologies development that provides information on differences of
coffee chemical composition of different traits. The current work describes the application of multivariate analysis methods, mainly in NMR data, in order to verify chemical differences related to certain attributes that determine the coffee quality. In this work, the specie used in the brew preparation, defective coffee beans, adulteration with barley, the effect of temperature and time on the roasted coffee
composition, cup quality and production mode were studied. In the first study, it was determined the main differences between the composition of arabica and robusta coffee discriminating both species and was determined the content of each one in blends. In similar study, were discriminated blends of coffee and barley and was determined the percentage of this adulterant added to the coffee. In this case, comparing to NMR, the infrared spectroscopy also shows great efficiency for the same purpose. In the study of defective coffee beans were verified greater changes
in the chemical composition of black beans compared to the non-defective coffee besides the discrimination of all kinds of defective beans analyzed; green, sour, and black. Considering the production mode, were discriminated coffees produced
conventionally and organically. The higher acetic acid content in the organic samples was determinant to this discrimination. Considering the cup quality, efficient chemometric models were obtained to classify the samples with different qualities -
soft, hard, and rio. The lipid compounds were important to this classification and were identified by using HR-MAS technique. In the study of roast degree it was verified
greater influence of the temperature in the chemical composition changes from roasted coffees and it was possible to discriminate samples that were submitted to several conditions of time and temperature during roast process.