Cu₂O@CuS nanozymes as a cross-reactive sensing platform: Machine learning-enabled discrimination of sulfur compounds in milk.

Detection of sulfur compounds is important in the dairy industry for ensuring food safety. Herein, we develop a colorimetric sensor array based on peroxidase-like Cu₂O@CuS nanocubes to simultaneously discriminate multiple sulfur compounds, thereby overcoming the inherent limitations of traditional "lock-key" sensors. In this array, three concentrations of Cu₂O@CuS catalyze the oxidation of 3,3',5,5'-tetramethylbenzidine (TMB) into blue oxidized TMB in the presence of H₂O₂. The introduction of sulfur compounds inhibits the catalytic activity of the Cu₂O@CuS to varying extents, and the resulting cross-reactivity of the sensor array generates distinct colorimetric "fingerprint" patterns. When combined with machine-learning algorithms, such as linear discriminant analysis, this sensor array achieves excellent discrimination of five sulfur compounds, even at low concentrations (0.08 μM) in milk samples. This sensor array has the advantages of short analysis time (about 20 min) and easy operation, and has good application prospects in rapid on-site evaluation of food quality.