Optimisation Studies of Naphthalene Adsorption on Bentonite Clay Impregnated on Chitosan and Surfactant using RSM–CCD, ANN–BPand ANN–PSO Techniques

Abstract

The effects of process variables on naphthalene adsorption on bentonite clay–cetyltrimethylammonium bromide (CTAB)–chitosan matrix were investigated in this study. The independent variables are initial concentration ( A ), adsorbent dosage ( B ), contact time ( C ), temperature ( D ) and pH ( E ) while the response variable is the % removal of naphthalene. The adsorption process was modelled and optimised using the response surface methodology (RSM)–central composite design (CCD), artificial neural network (ANN)–back propagation (BP) and ANN–particle swarm optimization (PSO) algorithms. The interactive terms of A and B , B and E , C and E , and D and E give a synergistic effect on the adsorption process. On account of p –value, temperature and adsorbent dosage are the most influential single terms. Among the square terms, adsorbent dosage was found to be the most controlling factor while adsorbent dosage and contact time were the most significant interactive factors for the adsorption process. The RSM–CCD, ANN–BP and ANN–PSO aptly modelled and optimised the process variables. However, the ANN-PSO algorithm excelled over the RSM–PSO and ANN–BP owing to its highest 2 R value of 0.9803, least value of the error functions considered and percentage error of validation. Hence, ANN-PSO is projected for the optimisation studies of naphthalene adsorption on the synthesised bentonite clay– cetyltrimethylammonium bromide (CTAB)–chitosan matrix. These approaches enable scalable and data-driven solutions, which minimise costly trial-and-error experiments for treating industrial wastewater contaminated with persistent pollutants of PAHs.