An Efficient Investigation and Machine Learning-Based Prediction of Decolorization of Wastewater by Using Zeolite Catalyst in Electro-Fenton Reaction

El Jery A.; Aldrdery M.; Shirode U.R.; Gavilán J.C.O.; Elkhaleefa A.; Sillanpää M.; Sammen S.S.; Tizkam H.H. · Catalysts (2023) · DOI: 10.3390/catal13071085

Resumen

The shortage of water resources has caused extensive research to be conducted in this field to develop effective, rapid, and affordable wastewater treatment methods. For the treatment of wastewater, modern oxidation techniques are desirable due to their excellent performance and simplicity of implementation. In this project, wet impregnation and the hydrothermal technique were applied to synthesize a modified catalyst. Different analysis methods were used to determine its characteristics, including XRD, BET, FT-IR, (Formula presented.), and FE-SEM. The catalyst features a spherical shape, large surface area, high crystallinity, and uniform active phase dispersion. In order to eliminate the methylene blue dye as a modeling effluent, the catalyst’s performance was examined in a heterogeneous quasi-electro-Fenton (EF) reaction. The impact of various performance characteristics, such as catalyst concentration in the reaction medium, solution pH, and current intensity between the two electrodes, was elucidated. According to the results, the best operational circumstances included a pH level of 2, a catalyst concentration of 0.15 g/L, and a current of 150 mA, resulting in the greatest elimination efficiency of 101%. The catalyst’s performance was stable during three consecutive tests. A pseudo-first-order model for the elimination reaction’s kinetics was developed, which showed acceptable agreement with the experimental results. This study’s findings help clarify how well the heterogeneous zeolite catalyst functions in the pseudo-EF reaction. The results revealed the method’s potential to be implemented in wastewater treatment. An artificial neural network model is utilized to predict the removal percentage. The hyperparameter tuning is used to find the best model, and the model achieved an MAE of 1.26% and the (Formula presented.) was 0.99.

Instituciones: Universidad Privada del Norte