The objective of this study was to develop a non-destructive approach for through-the-bottle analysis of beer composition by integrating handheld Raman spectroscopy with a pattern recognition algorithm. Sixty commercial beers representing diverse styles were analysed directly through glass bottles of three colours (clear, green, and amber) using a handheld Raman device equipped with a 1064 nm excitation laser. Five key quality parameters including alcohol content, real extract, specific gravity, colour, and bitterness (IBU), were determined using official methods from the American Society of Brewing Chemists and served as reference values for partial least squares regression (PLSR) model development. Clear and green bottles yielded strong predictive performance with R PRE = 0.92–0.97 and RPD ranging from 2.2 to 8.3 for all five quality parameters, confirming the suitability of handheld Raman devices for rapid quality screening. In contrast, amber bottles showed extensive signal attenuation, failing to capture reliable compositional signatures and leading to poor model performance (R PRE < 0.56). Overall, the results demonstrate the potential and challenges of handheld Raman spectroscopy for non-destructive, through-bottle assessment of beer quality. • Handheld 1064 nm Raman enabled through-bottle beer quality analysis • PLSR model accurately predicts alcohol, SG, extract, color, and bitterness • Clear and green bottles yield strong predictive performance for beer quality traits • Amber bottles cause severe signal attenuation, yielding poor regression performance • Demonstrates potential and limitations of handheld Raman for rapid beer QC
Bao et al. (Fri,) studied this question.