Machine learning-assisted sensor array based on single trimetallic nanozyme with multienzyme-like activities for rapid and intelligent identification of multiple metal ions.

BACKGROUND

The concurrent detection and differentiation of multiple heavy metal ions is critical for water quality surveillance, but it continues to pose a significant technical challenge.

RESULTS

Herein, this study designed a machine learning-assisted sensor array based on single trimetallic nanozyme (PVP-PtCuAu NCs) with multienzyme-like activities for rapid and intelligent identification of multiple metal ions (Pb2+, Ag+, As3+, Cu2+ and Hg2+). Wherein, the trimetallic nanozyme (PVP-PtCuAu NCs) exhibited good multienzyme-mimicking activities, including peroxidase-like, oxidase-like and laccase-like catalytic properties. Notably, different metal ions could effectively regulate the three enzyme-like activities of PVP-PtCuAu NCs, generating diverse cross-reaction optical response patterns. On this basis, a three-channel colorimetric sensor array based on PVP-PtCuAu NCs nanozyme was constructed, enabling the rapid and accurate identification of five metal ions and their mixtures. Furthermore, to enhance analytical performance, machine learning algorithms were employed to integrate with the nanozyme sensor array for processing the three-channel response data, achieving highly reliable qualitative classification and quantitative prediction of five metal ions, which also demonstrated satisfactory applicability in real water samples.

SIGNIFICANCE: Overall, this study presents a facile and efficient approach for the rapid discrimination of multiple metal ions in environmental waters, offering broad application prospects for environmental monitoring and water quality assessment.