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

Mercury (Hg 2+ ), a highly toxic neurotoxin and carcinogen, endangers the environment and human health even at low exposure. Developing high-performance Hg 2+ detection is critical for food safety, especially seafood. Herein, a novel silver nanoparticles-based organic framework (COF-AgNPs) nano-enzyme colorimetric sensor combined with multimodal machine learning (ML) for dual-model rapid quantitative detection of Hg 2+ in aquatic products was developed. The synthesized COF-AgNPs can facilitate the formation of mercury-silver alloys (Hg-Ag) between Hg 2+ and AgNPs. This combination can significantly enhance the catalytic activity of the nano-enzyme, enabling the generation of the blue oxide 3, 3', 5, 5'-tetramethylbenzidine (oxTMB) from the transparent substance 3, 3', 5, 5'-tetramethylbenzidine (TMB) within 5 min. This COF-AgNPs nano-enzyme sensor addresses the drawbacks of traditional spectroscopic analysis methods, including sophisticated instrumentation, high costs, and lengthy processing times, and can achieve a detection limit of 10.3 nM while exhibiting exceptionally strong anti-interference performance. Meanwhile, to address single-modality limitations and matrix interference, ML was integrated to analyze the detection data. The results show that the support vector machine (SVM) model achieves an accuracy rate of 96.08% in detection, and the developed sensing system is of excellent performance, with a response time of 5 min, visual output, and a wide linear range (0.1–50 µM). By integrating nanozyme sensing technology with ML algorithms, this sensor not only provides an innovative solution for Hg 2+ detection but also establishes a new paradigm for intelligent sensing systems, enabling its extension to the rapid detection of other heavy metal pollutants.