Fe-doped ZIF-8 for sunlight-driven photocatalytic degradation of methylene blue: synthesis, characterization, and AI-based process modeling

Abstract The growing environmental cost of wastewater dye effluents is reason enough to adopt effective, sustainable remediation measures. This paper reports a new, environmentally friendly synthesis of bimetallic Fe x Zn 1− x ZIF-8 structures through a fast, solvent-free, room-temperature mechanochemical synthesis without using toxic solvents or high energy expenditures. The Fe 0.075 Zn 0.925 -ZIF-8 has much higher visible-light photocatalytic activity than pristine ZIF-8, with over 95 % degradation of methylene blue (MB) under solar simulation in 40 min. Successful incorporation of Fe, prolonged absorption in visible light, decreased bandgap (2.8 eV at x = 0.075), and inhibited recombination of charges were confirmed by comprehensive characterization (XRD, SEM-EDS, XPS, UV–Vis), which explains the mechanism contributing to improved performance. More importantly, this material serves two purposes (1): as a high-efficiency, reusable photocatalyst that takes advantage of synergistic radical-mediated dye degradation using Fe/Zn sites and (2) as a model of the framework of an artificial intelligence (AI) model. The experimental degradation data were trained on machine learning algorithms (Artificial neural networks) using the radiation time and the catalyst dosage. The AI models demonstrated high predictive accuracy ( R 2 = 0.995, RMSE = 0.005) that allowed the optimization of the degradation conditions with high precision and the demonstration of the non-linearity among the parameters. This study is the first to bridge sustainable materials design and AI-based process engineering to provide a scalable, environmentally friendly solution to the dye wastewater treatment problem and an effective predictive toolkit to drive the development of a photocatalytic system.