Identification of Baijiu flavor types by MEMS gas sensor based on temperature programmed desorption method

Aiming to address the bottleneck issues of database complexity and signal drift in electronic nose systems, this study proposes a novel supplementary strategy for multi-component gas detection. The MEMS gas sensor developed here, based on temperature-programmed desorption (TPD), enables the extraction of multiple volatile target compounds in Baijiu (Chinese distilled spirits) through an adsorption process. Subsequently, the TPD spectrum is employed to derive characteristic signals of these target substances for identification. By incorporating this sensor into a Baijiu detection system and combining it with pattern recognition algorithms, accurate classification of aroma profiles has been successfully achieved for both known and unknown Baijiu samples across different brands. This multi-component gas detection method not only offers an effective complement to electronic nose technology but also provides valuable insights for improving the selectivity of metal oxide semiconductor (MOS) gas sensors.