Decolorization of Azo Dye Solution by Ozone Based Advanced Oxidation Processes: Optimization Using Response Surface Methodology and Neural Network

The decolorization of acid black azo dye by three ozone-based advanced oxidation processes, viz. O3, O3/UV, O3/Fe (II) were studied by developing prediction model using Response Surface Methodology (RSM) and Artificial Neural Network (ANN) techniques. The decolorization is determined through the absorbance wavelength of dye using UV-Vis spectrophotometer. The central composite design experiment was used to develop a mathematical correlation between four independent parameters such as ozone dose, initial pH, initial dye concentration, reaction time and decolorization efficiency as process response. The decolorization efficiency of 95.5% obtained at optimized conditions (ozone concentration = 70 mg L−1, initial dye concentration of 200 mg L−1, initial pH 6 and reaction time of 20 min) was comparable to 94.15% predicted by RSM. The predicted results obtained from both RSM and ANN were found to be in good agreement with experimental values (R2 = 0.982 for O3; 0.981 for O3/UV; 0.976 for O3/Fe (II) process for RSM modeling and R2 = 0.994 for O3; 0.992 for O3/UV; 0.991 for O3/Fe (II) for ANN modeling). Thus, both RSM and ANN would effectively predict the performance of O3, O3/UV, O3/Fe (II) processes while ANN predictions found to be better than RSM in statistical comparison. All the three AOPs studied found to be competent for the decolorization of acid black azo dye. UV promotes decolorization by ozone but it adds to initial cost. Fe (II) can enhance ozonation efficiency significantly and found to be more effective compared with O3 and O3/UV processes.