Development of a novel method for the removal of diazinon pesticide from aqueous solution and modeling by artificial neural networks (ANN)

•Sodalite zeolite crystals were successfully synthesized from perlite with NaOH solution.•It was found that sodalite zeolite can be considerable potential for the removal of diazinon from aqueous solutions.•The synthesized zeolite was modified with Cu2O nanoparticles and also the ANN model was applied to predict diazinon removal efficiency using ATZ and MZ.•Diazinon pesticide which contains sulfur group interacts with the cupper site very efficiently. Removal efficiency of the diazinon pesticide was highly increased with the modified zeolite.•The kinetic data indicated that the adsorption process was controlled by Morris Weber equation. Thermodynamic studies demonstrate negative ΔG and ΔH and positive ΔS. In this study, perlite was used as a low-cost source of Si and Al to synthesize the sodalite zeolite via hydrothermal synthesis method. Cu2O nanoparticles (30–60nm) were coated on a bed of sodalite zeolite. Toward this aim, a series of batch adsorption experiments was carried out and the sorption of diazinon from aqueous solutions on acid treated zeolite (ATZ) with dilute H2SO4 solutions and modified zeolite by Cu2O nanoparticles (MZ) were also determined. Results showed that Cu2O nanoparticles have a significant effect on the diazinon removal processing from aqueous solution. Maximum adsorption rates were 98.2% (with 0.2, 20, 6 of adsorbent dose, contact time and pH, respectively, for MZ) and 63.4% (with 0.3, 80, 6 of adsorbent dose, contact time and pH, respectively, for ATZ). Three equations, i.e., Morris–Weber, Lagergren (pseudo first order) and pseudo second order have been applied to study the kinetics of removal process. The diazinon sorption process was well described by the pseudo second order (type 2) kinetic model. The Langmuir, Freundlich, Temkin and Dubnin–Randkovich (D–R) models were tested on sorption data to estimate the sorption capacity, intensity and energy. Langmuir (type 1) isotherm provided the best fit to the equilibrium data with maximum adsorption capacity of 61.73 and 15.10mg/g for MZ and ATZ, respectively. The thermodynamic parameters ΔH, ΔS and ΔG were evaluated. Thermodynamic parameters showed that the sorption of diazinon onto zeolite was feasible, spontaneous and exothermic under studied conditions. Artificial neural network (ANN) model was also applied for modeling of diazinon removal from aqueous solution by ATZ and MZ. There was a good agreement between the experimental and predicted values with seven neurons in hidden la