Decolourization of Disperse Blue 3 Dye by Electro coagulation process using Al and Fe electrodes –Application of the Artificial Neural Network Model

Abstract

Contamination in wastewater is a major issue in the present world, the Disperse blue 3 dye (DB3) removal was studied by an electro coagulation process using Al and Fe electrodes. The experiments were performed with synthetic solutions in batch mode. The effect of the operating parameters like the electrolysis time, current density, initial pH, conductivity, inter-electrode distance and initial dye concentration, has been investigated. The results show high discoloration efficiency, reaching 98 and 96% with Al and Fe electrodes respectively. At the optimum condition of the EC process were electrolysis times of 70 and 30 min, current densities of 139 and 93 mA/cm², initial pH 5, conductivity of 5.67 mS/cm and inter-electrode distance of 1.5 cm. The Artificial Neural Network (ANN) technique was used to model the experimental data of the current density. The feed-forward neural network model was optimized by using the Levenberg-Marquardt algorithms. A comparison between the predicted and experimental data gave high correlation coefficients (0.99977 and 1) with the minimum MSE value (1.55.10-7 and 1.31.10-5) respectively for Al and Fe electrodes.