Chitosan-based magnetic adsorbent for removal of water-soluble anionic dye: Artificial neural network modeling and molecular docking insights

Herein, we report a new composite magnetic-adsorbent made of doped spinel ferrite (15%) dispersed throughout a matrix of chitosan (CS) cross-linked with glutaraldehyde (GA). The composite material was well characterized by using instrumental methods of physical-chemical analysis (SEM, EDX, FTIR and VSM). The produced adsorbent was applied for the removal of Acid Orange 7 (AO7) dye from aqueous solutions. Aspects of adsorption kinetics, isotherms and thermodynamics were detailed. According to the Dubinin-Radushkevich (D-R) isotherm, the mean free energy of adsorption ranged from 14.37 to 16.59 (kJ/mol), suggesting ion-exchange dominating phenomena. In addition, we developed an artificial neural network (ANN) model to explore extensively the effects of factors on the adsorption performance. The coupling of ANN model with a genetic algorithm provided optimal conditions of adsorption. A maximal color removal efficiency of 98.01% was observed experimentally under optimal conditions (pH 2.51, sorbent dosage 3.88 g/L, initial dye concentration 25.3 mg/L, contact time 204 min). To unveil interaction mechanism, we employed molecular docking simulations. Computational outcomes suggested the formation of hydrophobic contacts and hydrogen bonds between AO7 dye molecule and CS-GA receptor. Molecular docking results agreed with the D-R isotherm findings, highlighting that electrostatic forces were greater than Van-der-Waals interactions.