Li_2025_J.Hazard.Mater_501_140978

Pesticides have traditionally been detected based on the inhibition of natural acetylcholinesterase enzyme activity by pesticides. This approach was costly, offered limited detection coverage for pesticides, and suffered from the inherent instability of natural enzymes. To address the instability issue in sensor arrays, artificially synthesized nanozymes with enzyme-like activities were constructed in the sensor array. The diverse enzyme-like activities of nanozymes provide an excellent foundation for developing multi-channel sensor arrays. A four-channel sensor array was constructed based on multienzyme-like activities for the identification and discrimination of nine pesticides. An accurate distinction was achieved for 5 categories of pesticides using the constructed multienzyme-like sensor array method. In addition, a concentration-independent identification model of pesticides based on machine learning was constructed to simulate the detection situation in real environments, with an accuracy of 99.20 %. The proposed sensor array has great practical application prospects to be widely used for the detection of pesticides on food surfaces and in the water environment due to its excellent anti-interference capability.