Investigation of photocatalytic activity of synthesized zinc stannate for tetracycline antibiotic degradation: modelling and optimization through RSM, ANN and genetic algorithm

Abstract

The present study aims at investigating the ability of the synthesized zinc stannate, especially its photocatalytic activity, for degradation of tetracycline (TC) antibiotic under UV light irradiation. The process is assessed by four independent variables including pH, reaction time, initial TC concentration and photocatalyst dosage where TC removal is considered the response parameter. First, for optimization of the process, the response surface methodology (RSM) is implemented. Then, a second-order RSM model is developed based on experimental results. Subsequently, an artificial neural network (ANN) and a genetic algorithm are, respectively, applied for simulation and optimization of TC removal from aqueous solutions. Afterwards, an ANN model is trained by applying three different algorithms (scaled conjugate gradient, gradient descent and Levenberg-Marquardt algorithms), and the best algorithm is taken into account to develop a predictive model. Next, the optimal number of hidden layers is determined. Finally, to optimize effective input parameters and percentage of TC removal from aqueous solutions, the ANN model is used along with the genetic algorithm for the process optimization.