In the present work removal of arsenic from aqueous solution using zinc oxide nanoparticle embedded on activated silica as adsorbent was studied. The initial As(III) concentration was varied from 0.01N to 0.1 N with varying amount of Zinc oxide nanoparticle embedded onto activated silica 1 – 8 g in laboratory batch adsorption experiment. The maximum adsorption efficiency was found at As(III) initial concentration of 0.06 N, adsorption dose of 8 g/L and pH of the solution of 5.0. The equilibrium contact time was found at 100 min. A three layer feed forward artificial neural network (ANN) with back propagation training algorithm was developed to model the adsorption process of As(III) in aqueous solution using Zinc oxide nanoparticle embedded onto activated silica as adsorbent. The neural network architecture consisted of tangent sigmoid transfer function at hidden layer with 10 hidden neurons, linear transfer function at output layer and Levenberg-Marquardt (LM) back propagation training algorithm. The neural network model predicted values are found in close agreement with the batch experiment result with correlation coefficient (R) of 0.999 and mean squared error (MSE) 4.39.
Keywords: Artificial Neural Network; Adsorption; Arsenic; Back Propagation Algorithm; Levenberg-Marquardt Algorithm