"Gyroscope"-type portable device based on a bimetallic Fe/Zn-ZIF8 MOF nanozyme@agarose hydrogel with deep learning for monitoring Hg pollution.

Mercury (Hg) pollution in Hg mine waste residues under precipitation leaching causes long-term pollution to the surrounding environment, posing a serious threat to environmental and public health. To achieve rapid monitoring of Hg pollution in mining areas, a system (Fe/Zn-ZIF8@agarose hydrogel) was developed in this study using an agarose hydrogel as a bimetallic Fe/Zn-ZIF8 metal-organic framework (MOF) nanozyme carrier. This system leverages the specific recognition of Hg2+ by the thiol groups in the antioxidant glutathione (GSH) to regulate the peroxidase-like catalytic activity of Fe/Zn-ZIF8 in the H2O2 and 3,3',5,5'-tetramethylbenzidine (TMB) colorimetric reaction, enabling an "on-off-on" signal response. To further integrate deep learning-assisted image analysis algorithms, the sensing unit was embedded into a 3D-printed "gyroscopic" device to develop a portable platform for the visual detection of Hg2+ in water environments. This portable device exhibited excellent reliability for Hg2+ detection, with a spiked recovery rate of 95.8∼112.8% and a relative standard deviation (RSD) < 5.72%. It was successfully applied for onsite Hg2+ detection in mine leachate, surrounding river water, and paddy soil water. Among these samples, the Fe/Zn-ZIF8@agarose hydrogel was used to detect an Hg2+ concentration of 5.3 μM in mine leachate, whereas Fe/Zn-ZIF8 was used to detect Hg2+ concentrations of 0.36 μM and 0.12 μM in certain river water and paddy soil samples from mining areas, respectively. Through multimodule collaboration, this system significantly reduces the subjective errors associated with traditional colorimetric methods and provides a rapid, accurate, and user-friendly portable analytical platform for the onsite emergency monitoring of mercury pollution in mining areas.