Rapid and Simultaneous Analysis of Qualitative Parameters of Beer by FT-NIR Spectroscopy

Abstract

Calibration models were developed for simultaneous analysis of alcohol, real extract, energy value, total saccharides and other features in beer using Fourier-transform near-infrared (FT-NIR) spectroscopy. The recorded FT-NIR spectra were correlated to the analytical data of reference methods by means of the multivariate PLS algorithm. Two groups of test beers were made from barley from two consecutive harvests. A calibration model for FT-NIR was set up using data from reference methods and was selected by means of the error of calibration (RMSEC), cross-validation (RMSECV), and the root-mean-square-error of prediction (RMSEP). It was found that calibration models for energy value, alcohol, real extract and the content of saccharides are stable and reliable. The long-term effect of calibration was evaluated by comparing external validation parameters, expressed as RMSEPex.