An untargeted gas chromatography-mass spectrometry method followed by multivariate data analysis to evaluate the volatile generated by different brewing methods

Beer is one of the most produced alcoholic beverages worldwide [1]. Different technologies have been applied for its production, so the study of the volatile profile can be a powerful tool to evaluate their effect onto the final product. Gas chromatography (GC) coupled with mass spectrometry (MS) is the technique of choice for the evaluation of the volatile profile of food. According to the nature of the samples, a variety of extraction techniques has been proposed, among which static or dynamic headspace, purge&trap and solid-phase microextraction [2]. In this study, an untargeted GC-MS method followed by multivariate data analysis was developed to evaluate the volatile profile generated by different brewing methods using both traditional technologies and auto-fermentation approaches. Different beers i.e. Bohemian Pilsner (BP), American IPA (AI), Vienna Lager (VL) and Cream Ale (CA) were analyzed. Solid phase microextraction (SPME) using a 50/30 μm DVB/CAR/PDMS fibre by operating in the headspace mode was applied for the extraction and preconcentration of the volatile compounds. Finally, the identification of the volatile molecules was carried out by comparing: i) the spectra obtained experimentally with those stored in the NIST library [3], ii) the calculated Kovats indices with those reported in literature or stored in proprietary databases [4], iii) by the injection of pure standards. For each type of beer and treatment, four independent replicate samples were analysed for a reliable evaluation of repeatability. A total of 59 volatile compounds were identified, being the AI characterised by the richest volatile profile (57 out of 59 compounds identified in the sample). Multivariate data analysis, namely principal component analysis (PCA) was carried out to evaluate the presence of clusters among the samples. A 74% of the variance was explained by the first two PCs. As shown in the score plot of Figure 1a, PC1 allowed to differentiate the AI beer, whereas PC2 mainly explained the difference between CA and BP. An additional grouping of the samples was observed for each typology of beer according to the technological treatment applied for their production.