Zohal Safaei Mahmoudabadi

University of Tehran, Tehran, Iran

Abstract Title:

Artificial Intelligence-Driven Optimization and Prediction of Alcohol Vinasse Treatment Using Porous α-Fe₂O₃ Nanoparticles Coupled with Advanced Oxidation and Coagulation–Flocculation

Research Interests:

Alcohol vinasse is a highly complex and recalcitrant wastewater characterized by high organic loading and chemical oxygen demand (COD), making its effective treatment a significant environmental challenge. In this study, an integrated treatment strategy combining advanced oxidation and coagulation–flocculation was developed for efficient COD removal from alcohol vinasse using highly porous α-Fe₂O₃ nanoparticles and polyacrylamide as a coagulant. Highly porous α-Fe₂O₃ nanoparticles were synthesized through a chemical precipitation method and comprehensively characterized using FT-IR, Raman spectroscopy, X-ray diffraction (XRD), scanning electron microscopy (SEM), and N₂ adsorption–desorption analysis. The influence of α-Fe₂O₃ nanoparticle dosage on COD removal was initially investigated, and a dosage of 3000 ppm provided the highest removal efficiency. Subsequently, an artificial intelligence-assisted modeling and optimization framework was developed to predict COD removal and determine the optimal operating conditions. The effects of key process variables, including pH, reaction time, oxidant dosage, coagulant dosage, and temperature, were evaluated using experimental data and response surface methodology. Under the optimized conditions of pH 7.36, reaction time of 90 min, oxidant dosage of 17.89 wt.%, coagulant dosage of 1.6 wt.%, and a temperature of 70 °C, the integrated process achieved a maximum COD removal efficiency of 98.64%. The AI-based prediction demonstrated the potential of data-driven modeling to accurately describe the nonlinear relationships between process variables and treatment performance and to facilitate process optimization. The superior catalytic performance of porous α-Fe₂O₃ nanoparticles was attributed to their highly porous structure and enhanced surface reactivity, which promoted the generation of reactive species during the advanced oxidation process. Overall, the proposed AI-assisted integrated process provides an efficient and promising approach for the treatment of high-strength alcohol vinasse and offers a predictive framework for improving process performance and reducing experimental requirements.