Machine learning-assisted sensor array based on dual enzyme-like activities of SMX-Cu for accurate discrimination of...

The toxicity of multiple chromium species varies greatly, and they may transform into each other by oxidation-reduction reactions in the water bodies. Therefore, achieving on-site discrimination of multiple chromium species is of great significance in the environmental field. In our previous study, we achieved the recognition of multiple chromium species based on the significant inhibitory effect of sulfonamide ligand nanozymes with laccase-like activity. However, this method has an inherent disadvantage, that CrO₄^2- and Cr₂O₇^2- will interfere with each other at certain concentration ranges, resulting in misclassification. Based on this, we selected SMX-Cu with the highest inhibition rate and combined its oxidoreductase-like activity for the first time, greatly improving the ability to discriminate multiple chromium species (Cr³⁺, CrO₄^2-, Cr₂O₇^2-), precise discrimination of multiple chromium species mixed in different ratios was achieved. Constructed a dual enzyme-like activities sensor array for concentration-independent discriminating multiple chromium species combining K-Nearest Neighbors (KNN) algorithm, accuracy has been significantly improved compared with our previous method. Combined with paper-based sensor array, on-site discrimination of real water samples was achieved, offering promising applications in environmental monitoring fields.