Comparison of headspace solid-phase microextraction high capacity fiber coatings based on dual mass spectrometric and broadband vacuum ultraviolet absorption detection for untargeted analysis of beer volatiles using gas chromatography

Despite the same basic ingredients used in brewing, there is a significant variation in beer styles. With the rapid increase in craft brewing, beer styles have become even more numerous and complex in the recent past. A GC-MS/VUV (post-column split for dual detection) instrument with headspace high capacity SPME was used to investigate 21 different beers which represent three beer styles - India pale ales, blondes, and hefeweizens. Since results from untargeted studies can be affected by the sorbent material used, the extraction performances of three high capacity SPME fibers, i.e., polydimethylsiloxane, polydimethylsiloxane/carbon wide range, and polydimethylsiloxane/carbon wide range/divinylbenzene, were evaluated. Good reproducibility (<10% RSD) was obtained for each high capacity fiber using both detectors. The tandem MS/VUV detection coupled with GC separation proved to be particularly valuable for compound identification, especially for isomers and compounds with similar structures. The evaluation of VUV detection for untargeted analysis led to similar performances as MS detection. Both the VUV and the MS were able to effectively differentiate between beer styles using principal component analysis. In addition, the use of 3 different statistical approaches, one-way ANOVA (p-value < 0.05), partial least square discriminant analysis, and random forest, universally identified 12 of the components most influential in distinguishing the three beer styles (e.g., beta-myrcene, linalool, isopentyl acetate, 2,4-di-tert- butylphenol). This is the first reported evaluation of VUV detection and the first comparison of simultaneous VUV and MS detection for untargeted classification of complex mixtures using GC. (C) 2020 Elsevier B.V. All rights reserved.

Automated summary

Despite using the same basic ingredients, beer styles exhibit significant variation, especially with the rise of craft brewing. A study using GC-MS/VUV to analyze 21 different beers found VUV detection to have similar performance as MS detection for untargeted analysis, and both VUV and MS detection effectively differentiated beer styles through principal component analysis. This research marks the first evaluation of VUV detection and comparison of simultaneous VUV and MS detection for untargeted classification of complex mixtures using GC.