Classification of beer by thermogravimetric and chemometric techniques

Abstract

Thermogravimetric (TG) technique combined with partial least squares discriminant analysis (PLS-DA) was applied to distinguish five pilsner beers from different brands (Antarctica, Bohemia, Brahma, Budweiser and Imperio). Herein, we develop a TG methodology that use a small volume (40 µL), and an analysis time of 25 min. The major gases evolved during the thermal decomposition were water, ethanol and carbon dioxide. Energy-dispersive X-ray spectrometer of the ashes detected the major constituents: oxygen, sodium, magnesium, silicon, phosphorus, sulfur, chlorine, potassium and calcium. TG and PLS-DA technique together were able to classify the five brands with a classification rate for the model of 97% with a confidence interval of 92–99%, achieving high sensitivity and specificity between the calibration, cross-validation and predicted results.