Evaluación de la calidad de cervezas artesanales usando métodos multivariados de modelamiento

Abstract

The objetive was to evaluate the quality of 12 brands of craft beer through the application of multivariate modeling methods based on 6 different physicochemical parameters. The following analyses were conducted: alcoholic strength by applying the distillation method to obtain the ethanol content, which was measured with the help of the hydrometer, determination of pH using a pH meter; determination of total acidity based on the volumetric method, performing an acid-base titration with normal 0.1 sodium hydroxide standard; it was necessary to determine the density to get the lactic acid content in each brand of beer; determination of the presence of metals (lead, arsenic and iron) by applying the spectrophotometry method of atomic absorption, which required the preparation of the sample by means of a calcination and addition of several solutions to get the desired metal; it should be emphasized that before conducting the aforementioned analyzes, a pre-treatment called degassing was conducted, the same one that in this research was standardized with a minimum time of 1 hour for each brand of craft beer. Each analysis of this research has been done in duplicate to build the matrix. When applying chemometrics, two analyses were performed: Data Grouping and Core Components. For the clustering analysis, four methods were applied: single linkage, complete linkage, average linkage and Ward’s method at two distances: Euclidean and Manhattan; conducted through the Infostat program. The analysis was based on the permissible quality índices, representing that some craft beers have; four of them do not comply with the permissible acidity, one does not comply with the iron content and one pH breach. Concluding that the craft beer with the best quality is: Mestiza. It is recommended to analyze more parameters to get a much more complete analysis.