Machine learning-assisted Eu-MOF fluorescent material for simultaneous monitoring and removal of malachite green

Many countries prohibit the use of malachite green (MG) in aquaculture, owing to its high carcinogenicity, teratogenicity, and high tendency to leave residues. Therefore, it is crucial to develop materials that can simultaneously monitor and remove MG. In this study, a stable fluorescent material (EuUQCA) was constructed as a dual-functional platform for the simultaneous monitoring and removal of MG from water. EuUQCA enabled ultrahigh-sensitivity detection of MG (with a detection limit as low as 16.8 nM) through fluorescence quenching response. The introduction of MG caused a distinct change in fluorescence color, and a paper sensor based on EuUQCA utilized the color change to enable portable and visual detection of MG. A fast and portable semiquantitative method for sensing MG in aquatic environments was established by analyzing RGB values using an integrated smartphone. Moreover, an online data analysis model was developed based on fingerprint extraction from paper sensor images and machine learning algorithms; the model was rigorously cross-validated to enhance the precision and accuracy of the visual determination of MG. Additionally, EuUQCA powder exhibited excellent adsorption capacity for MG. Kinetic studies indicated that adsorption followed a pseudo-second-order model, whereas the equilibrium data were best described by the Freundlich isotherm. This work introduces a dual-functional material that integrates simultaneous detection and adsorption of MG with machine learning-assisted portable sensors; this system can not only improve the detection performance, but also promote the intelligent development of analytic technologies, with broad application prospects in the environmental monitoring and food safety fields.