Pesticide decontamination using UV/ferrous-activated persulfate with the aid neuro-fuzzy modeling: A case study of Malathion

In the current study, the Malathion decontamination by the aid of the UV/ferrous-activated persulfate (PS) was investigated and the effects of pH, persulfate (PS) concentration, ferrous concentration, Malathion concentration, and different inorganic ions were evaluated. Also, the Adaptive Neuro-Fuzzy Inference System (ANFIS) was applied to model Malathion degradation data. The maximum degradation efficiency was associated with pH = 3, PS concentration of 1.2 mM, the ferrous concentration of 0.6 mM, Malathion concentration of 20 mg/L for 60 min. The degradation efficiency was decreased in the presence of Cl⁻ (23%), NO₃⁻ (13.5%), HCO₃⁻ (35.4%) and H₂PO₄⁻ (48.7%) ions. Results revealed that persulfate radical (52%) plays a more important role in Malathion degradation while compared with hydroxyl radical (15%). The low root mean square error (RMSE = 6.451), mean absolute error (MAE = 3.8306), absolute-average-deviation (AAD = 0.1005) and high coefficient of determination (R² = 0.972) correlated with the proposed ANFIS models confirmed the model accuracy. Besides, the process optimization was conducted by using ANFIS to predict the best operating circumstances, which resulted in the maximum Malathion degradation (95.54%).