Data-Driven Modeling and Multi-Objective Optimization of the Uv/H₂o₂ Process for Diazinon and Malathion Removal From Drinking Water Publisher



Beikmohammadi M ; Moteallemi A ; Ravanipour M ; Nazmara S ; Hasanzadeh A
Source: Journal of Environmental Chemical Engineering Published:2026

Abstract

The optimization of UV/H₂O₂ processes for pesticide removal remains challenging because treatment performance is governed by complex interactions among operational variables rather than their independent effects. In this study, a predictive framework integrating Response Surface Methodology (RSM) with desirability based multi-objective optimization was developed to optimize the removal of Diazinon and Malathion from drinking water. A total of 288 experiments were conducted to investigate the effects of solution pH, initial pesticide concentration, H₂O₂ dosage, UV irradiance, and reaction time on process performance. Third-order polynomial models accurately described the experimental data and demonstrated strong predictive capability for both pesticides (Diazinon: R² = 0.959, Adj-R² = 0.953, Pred-R² = 0.948; Malathion: R² = 0.948, Adj-R² = 0.941, Pred-R² = 0.933). Response surface analysis revealed that solution pH, initial pesticide concentration, and UV irradiance were the dominant variables controlling pesticide degradation, while interactions among operational parameters exerted a greater influence than their individual effects. Multi objective optimization identified compound specific optimum operating conditions. Maximum Diazinon removal (98.34%) was achieved at pH 4 using 25 mg L−1 H₂O₂, low UV irradiance, and a reaction time of 28.63 min, whereas complete Malathion removal (100%) was obtained at pH 10 under the same H₂O₂ dosage and UV intensity with a reaction time of 9.51 min. The optimization results further demonstrated that increasing oxidant dosage or UV irradiance beyond the optimum operating range did not improve treatment performance but increased chemical and energy demand. Overall, the proposed predictive framework provides a practical decision support tool for designing energy-efficient UV/H₂O₂ systems and can facilitate the optimization of advanced oxidation processes for pesticide removal from drinking water. Copyright © 2026. Published by Elsevier Ltd.