Comparable analysis of flavor compounds and quality assessment of fermented bean curd using HS-SPME-GC/MS and colorimetric sensor array

This study aims to develop an effective tool by integrating traditional analysis methods with advanced sensor array technology and multiple algorithms, to achieve rapid characterization, qualitative identification, and flavor quality evaluation of fermented bean products. Firstly, the flavor compounds and flavor profiles of fermented bean curd from different manufacturers were identified using headspace solid-phase microextraction gas chromatography-mass spectrometry (HS-SPME-GC/MS) and quantitative descriptive analysis (QDA). A total of 63 volatile compounds were identified, of which 13 volatile compounds were identified as the key differential compounds in selected samples. Eight descriptors were selected to represent the aroma characteristics of fermented bean curd. Based on the key differential compounds, a low-cost colorimetric sensor array (CSA) was designed and constructed. Furthermore, the feasibility of CSA combined with multiple algorithms was investigated for brand discrimination and flavor quality assessment. Linear discriminant analysis (LDA) exhibited a better performance in distinguishing different flavor quality samples, and the recognition accuracy of its prediction set was 97.22%. The study suggests that CSA combined with multiple algorithms can be an effective tool for flavor characterization and qualitative discrimination of fermented bean curd, and can provide a new multi-dimensional analysis method for food analysis.