Modelling and optimization of fenton process for decolorization of azo dye (DR16) at microreactor using artificial neural network and genetic algorithm

• Published in: Heliyon 2024-07, Vol.10 (13), p.e33862, Article e33862
• Main Authors: Sadeghzadeh Ahari, Jafar, Sadeghi, Masoud, Koolivand Salooki, Mahdi, Esfandyari, Morteza, Rahimi, Masoud, Anahid, Sanaz
• Format: Article
• Language: English
• Subjects: Artificial neural network; Azo dye; Decolorizing; Direct red 16; Fenton process; Genetic algorithm
• ISSN: 2405-8440; 2405-8440
• DOI: 10.1016/j.heliyon.2024.e33862

The Fenton process is widely employed for decolorizing industrial wastewater. Therefore, it is imperative to construct a model for optimizing the operational parameters and estimating the efficiency of decolorization within this process. In this study, an artificial neural network (ANN) model was created based on experimental data provided by a previous researcher who examined the decolorization of Direct Red 16 dye (DR16) using a heterogeneous Fenton process within a microchannel reactor. This model was utilized to optimize and forecast the efficiency of decolorization in the Fenton process. The accuracy of the model was validated by comparing its outcomes with actual experimental data. To further improve the efficiency of decolorization, optimal operational parameters were ascertained utilizing the genetic algorithm method. The study revealed that as dye concentrations increased from 10 to 40 mg/l, decolorization efficiencies improved proportionately, peaking at 89.78 %. Optimal operational parameters for maximizing efficiency were identified as a feed flow rate of 1 ml/min, H2O2 concentration at 500 mg/l, Fe2+ concentration of 4 mg/l, and maintaining pH between 2.6 and 2.8. Insights derived from both experimental and model-generated data were used to analyze the impact of operational parameters on decolorization efficiency. •Development of ANN model for predicting effluent dye concentration in Fenton process for decolorizing azo dyes.•Utilization of statistical and graphical criteria to select appropriate neural network model for predicting decolorization efficiency.•Application of genetic algorithm to determine optimal operating parameters for maximizing decolorization efficiency.•Combination of ANN model with genetic algorithm to obtain optimal operating conditions for maximizing decolorization efficiency.