Optimization of Bi2O3/TS-1 preparation and photocatalytic reaction conditions for low concentration Erythromycin wastewater treatment based on artificial neural network

Abstract At present, it is a challenge to degrade antibiotics of low concentrations in wastewater environment, highly effective and eco-friendly photocatalytic processes were considered promising technologies for this degradation. In this work, Bi2O3-loaded titanium silicalite-1 molecular sieve (Bi2O3/TS-1) composites were prepared and used to degrade low-concentration erythromycin (ERM) in wastewater. The content of active components and the dose of photocatalyst are key operating parameters with major effects on photocatalytic efficiency. The optimal parameters (Bi content and photocatalyst dose) were determined through artificial neural network simulating the relationship between key operating parameters and removal efficiency (RE). The maximum RE (98.02%) was measured under optimal operating parameters (Bi % = 5.5%, catalyst dosage = 0.6 g/L). The effects of water quality parameters (such as pH and ERM concentration) were also studied under optimal experimental conditions. Artificial intelligence was used in this work to achieve optimal control of catalyst preparation and photocatalytic reaction conditions. This study provides a useful strategy for the preparation of nanocatalysts and the practical application of high-efficiency photocatalytic reactions.