Prediction of the current state of beer fermentation indicators using volatile organic compounds from fermentation gas.

Microbial fermentation is widely used in the production of food, pharmaceuticals, and bioenergy. Proper monitoring of the fermentation process is essential to ensure consistent product quality and yield. Although the indicators of fermentation progress vary among systems, they are generally evaluated by quantifying the major metabolites in the fermentation broth. However, these measurements rely on offline analyses involving time-consuming sampling and pretreatment, which hinder the real-time detection of process deviations. In this study, volatile organic compounds (VOCs) in the fermentation gas during beer fermentation were analyzed to develop a non-invasive method for the prediction of the current state of sugar, organic acid, and ethanol concentrations in the fermentation broth at each sampling point. VOCs emitted during fermentation serve as valuable indicators reflecting the metabolic state of the yeast. Beer brewing was adopted as a model system to validate VOC monitoring, because it involves sequential sugar consumption, organic acid and ethanol formation, and abundant VOC generation. We constructed multivariate regression models using the VOC profiles obtained from gas chromatography-mass spectrometry (GC-MS) analysis using Tenax TA. Parallel sampling of the fermentation gas and broth followed by orthogonal partial least squares (OPLS) regression yielded highly accurate models for predicting key fermentation indicators at corresponding time points. Furans and aldehydes abundant at early stages showed inverse correlations with higher alcohols and esters produced later, indicating distinct metabolic transitions. This study demonstrated the feasibility of applying VOC profiles for the quantitative prediction of the current fermentation progress and highlights the potential of this approach as a novel non-invasive monitoring method linking aroma chemistry and process control.