Development of Distillation Sensors for Spirit Beverages Production Monitoring Based on Impedance Spectroscopy Measurement and Partial Least-Squares Regression

During spirit beverages production, the distillate is divided into three parts: the head, the heart, and the tail. Acetaldehyde and ethanol are two key markers which allow the correct separation of distillate. Being toxic, the elimination of the head part, which contains a high concentration of acetaldehyde, is crucial to guarantee the consumer's health and security. Plus, the tail should be separated from the heart based on ethanol concentration. Nowadays, online or in-line sensors for acetaldehyde monitoring during distillation do not exist, and the online sensors for alcohol monitoring, based on density measurement, remain expensive for producers. In this work, we demonstrate the development of distillation monitoring sensors based on electrical impedance spectroscopy (EIS) measurements, combined with PLS-R (partial least-squares regression) modeling. Four types of sensors are proposed and tested with wine-based distillates. Using PLS-R, the best correlations were found for one electrode, named SpotsSym. With an R2 up to 89.9% for acetaldehyde concentration prediction and an R2 up to 86.8% for ethanol, the obtained results indicate the promising potential of the proposed approach. To our knowledge, this is the first report of sensors capable of simultaneously measuring ethanol and acetaldehyde concentrations. Furthermore, these sensors offer the advantages of being low cost and nondestructive. Based on these results, the development of an in-line distillation monitoring system is possible in the near future, providing a promising tool for spirit beverages producers.

Automated summary

During spirit beverages production, distillate is divided into head, heart, and tail parts based on the concentration of acetaldehyde and ethanol. The elimination of the head part containing high acetaldehyde concentration is important for consumer health. Online sensors for acetaldehyde and ethanol monitoring are not available, but in this study, distillation monitoring sensors based on electrical impedance spectroscopy and PLS-R modeling were developed and tested with wine-based distillates. The results showed promising potential with high correlations and the advantages of low cost and nondestructiveness.