Rapid determination of the attenuation limit of beer using middle-infrared (MIR) spectroscopy and a multivariate model.

A new approach for the determination of the attenuation limit of beer samples using the specific fingerprint region of middle-infrared (MIR) spectroscopy in combination with multiple regression by partial least-squares (PLS) was developed using an attenuated total reflectance (ATR) module. A specific spectral region between 1200 and 800 cm(-1) was identified as highly informative for the quantification of the limit of attenuation. The absorptions in this region are induced by vibrational bands of ethanol (1080, 1040, and 880 cm(-1)) and dissolved extract, in majority maltotriose (1160-1140 and 1040-980 cm(-1)). The multivariate calibration results in a root mean squared error of calibration (RMSEC) of 0.40% and a validation procedure with independent samples results in a root mean squared error of validation (RMSEV) of 0.50%. A repeatability test, concerning the precision of the developed MIR method as well as the reference method, was analyzed using Student's t test. The test has shown no significant difference between the two random samples.