Nanomaterial-Based Sensor Arrays With Deep Learning for Screening of Illicit Drugs

Rapid and accurate screening techniques are demanded for illicit drugs that have raised international tensions. Bovine serum albumin-stabilized gold nanoclusters (BSA-Au NCs), carbon dots (C dots), thiosalicylic acid-stabilized silver nanoclusters (TA-Ag NCs), and Marquis reagent as photoluminescent sensing probes for five common illicit drugs are demonstrated in this study. Cocaine, 4-chloroethcathinone (4-CEC), and ketamine induce different degrees of photoluminescence changes of BSA-Au NCs, C dots, and TA-Ag NCs. Detection of heroin and methamphetamine (MA) is based on their formation of fluorescence polymer particles with Marquis reagent. To provide a unique pattern for each analyte, 2 × 4 sensor arrays are prepared. A deep learning-drug screening platform and system (DL-DSPS) is established and applied to differentiate the five illicit drugs, each with unique code based on its response to the probes. For example, codes of (−1, −1, 1, 0), (0, −1, 1, 0), (0, 0, 1, 0), (0, 0, 1, 1), and (0, 0, 0, 1) are for 4-CEC, cocaine, ketamine, heroin, and MA, respectively. The cost-effective and compact DL-DSPS is validated for screening of real-world illicit drug samples, with results in good agreement with that from GC-MS, showing its potential for multi-drug screening at crime scenes.