Modeling and optimization of nonylphenol removal from contaminated water media using a magnetic recoverable composite by artificial neural networks

Abstract

Herein, activated carbon impregnated iron oxide nanoparticles (Fe3O4/AC) were synthesized to determine their potentials for the adsorption of nonylphenol (NP) in aqueous solution with different experimental variables, namely the pH of the solution, contact time, adsorbent dosage and the initial NP concentration. Additionally, an artificial neural network system was used to find the relative importance of each of the aforementioned input variables on NP adsorption efficiency. Experimental findings indicated that the optimum solution pH for NP adsorption was 3.0. The equilibrium time of the adsorption process was 30 min. According to the results of isotherm and kinetic studies, among all applied models, the Liu and pseudo-first-order models showed the best fit with the experimental data. The pH of the solution, compared to other input variables, had the maximum impacts on NP adsorption efficiency. Under optimum conditions, the adsorption percentage decreased insignificantly from 99.6 to 92.6% after the fifth cycle. Also, the adsorption efficiencies of 70.7, 73.5 and 67.3% were observed for river water, tap water and wastewater effluent, respectively. Ultimately, from the findings of this study, it can be postulated that Fe3O4/AC nanoparticles can be recommended as a promising and novel adsorbent to remove NP from polluted groundwater.