Machine learning assisted sensor array based on ligand-regulated activity of nanozymes for the precise identification...

The development of reliable strategies for discrimination of polyphenols is crucial and challenging in the food and pharmacal fields. Herein, an effective regulation strategy for enzyme-mimic activities of nanozymes was proposed and applied into the establishment of sensor array to identify different polyphenols. The peroxidase (POD)-like and laccase (LAC)-like activities of three Cu-based nanozymes exhibited a directed order associated with the length of ligand branched chains. Based on the diverse and characteristic catalytic properties of the nanozymes, a colorimetric sensor array comprised of six sensing units (two activities × three nanozymes) for the simultaneous recognition of ten polyphenols was established. Not only the pure polyphenols at 5∼100 μM respectively displayed unique patterns with means of statistics algorithms, but the samples under complex and interfering conditions could be accurately discriminated by the sensor array. By integrating machine learning (ML) methodology, the proposed sensor array could achieve identification with almost 100% accuracy for polyphenols regardless of the component and concentration. It offered an effective approach for both nanozyme regulation and polyphenol detection, bridging and expanding the application of nanozyme materials and food testing field.