COF-AgNPs nanozyme combined with multimodal machine learning for dual-mode detection of Hg2+ in aquatic products.

Abstract

A covalent organic framework-supported silver nanoparticle (COF-AgNPs) nanozyme-based colorimetric sensing platform integrating complementary UV-vis spectral and smartphone-derived RGB signals with multimodal machine learning (ML) was developed for dual-mode rapid Hg2+ determination in aquatic products. Combined spectroscopic and microscopic evidence supports Hg-Ag amalgam on the AgNPs surface, indicating that COF-AgNPs facilitate the formation of Hg-Ag interfacial species upon exposure to Hg2+. This interaction significantly enhances the catalytic activity of the nanozyme, enabling the oxidation of colorless 3,3',5,5'-tetramethylbenzidine (TMB) to blue oxidized TMB (oxTMB) within 5 min. The resulting concentration-dependent colorimetric response enabled sensitive Hg2+ determination with a limit of detection of 10.3 nM and good selectivity against the tested interfering ions. The UV-vis method exhibited two linear ranges of 0.05-0.5 and 5-50 µM, while the smartphone RGB-based method showed a linear range of 0.5-15 µM. To utilize the complementary analytical information contained in these optical responses, ML methods were employed to model and analyze the data. Among the evaluated classification models, the Random Forest (RF) model achieved an accuracy of 99.47% for Hg2+ concentration discrimination. Overall, the integration of COF-AgNPs nanozyme-based colorimetric sensing with multimodal ML analysis provides a preliminary strategy for rapid Hg2+ determination in aquatic products and may support the future development of portable data-driven sensing strategies for environmental contaminant determination​.

Abstract as supplied by the indexing source. © the authors / publisher.