Modeling of fixed-bed adsorption of Cs+ and Sr2+ onto clay–iron oxide composite using artificial neural network and constant–pattern wave approach

Abstract

A low-cost, non-toxic and effective adsorbent constituted by a montmorillonite coated by iron oxides (montmorillonite–iron oxide composite) was prepared to assess its effectiveness in the removal of Cs+ and Sr2+ from aqueous solution. Dynamic adsorption experiments were carried out at room temperature under the effect of various operating parameters such as bed depth Z (5–15 cm), initial cation concentration C0 (2–50 mg L−1) and volumetric flow rate Q (0.5–8 mL min−1). Column performance has been modeled with constant-pattern wave approach combined to the Freundlich isotherm model and artificial neural network (ANN) models. The time, initial cation concentration, bed depth and volumetric flow rate were chosen as the input variables whereas, the outlet concentration Ct was considered as the output variable. The developed network was found to be useful in predicting the breakthrough curves. Experimental data for the used system were well fitted with ANN than the combined constant–pattern wave approach.