Machine learning-assisted fluorescence/smartphone dual-mode platform for lead ion detection using the novel polyimide covalent organic framework.

Unintended leakage of toxic and lead ion (Pb) ions poses high harmful to human health and environment, hence its monitoring and detection is of utmost significance. Here, we developed a covalent organic framework (COFs) that was first synthesized by condensation of monomers 4,4',4''-(1,3,5-triazine-2,4,6-triyl) triphenylamine (TTA) and perylene-3,4,9,10-tetracarboxylic dianhydride (PTCA), and then fluorescence/smartphone dual-mode point-of-care testing (POCT) platform based on COFs combined with machine learning was successfully constructed to detect Pb. The morphology, chemical structure, fluorescent properties, sensing competition and selectivity behaviors of COFs-Pb were systematically analyzed. Additionally, the synthesized COFs can also function as a fluorescence-quenching sensor for Pb detection, leveraging the synergistic effects of fluorescence resonance energy transfer (FRET), the photoinduced electron transfer (PET), and dynamic quenching effect collectively, enable the sensor to achieve a wide linear detection range from 0.1 nM to 1 μM, with an impressive detection limit of 50 pM. This limit is well below the permissible threshold of 15 μg L set by the United States Environmental Protection Agency (EPA). Notably, in contrast to liquid-phase sensors, swab-based sensors were successfully employed for Pb detection. The sensor also demonstrated reliable performance in actual water samples, yielding satisfactory recovery rates. This work introduces a novel approach to utilizing COFs with enhanced luminescent properties, offering significant potential for environmental monitoring applications.