DESIGN AND OPTIMIZATION OF ELECTROCHEMICAL OXIDATION INTEGRATED WITH MACHINE LEARNING IN THE ELIMINATION OF PARACETAMOL POLLUTANTS

• Main Author: Benito Tei De Mori, Benedito
• Format: Final Project
• Language: Indonesia
• Subjects: Teknik saniter dan perkotaan; teknik perlindungan lingkungan
• Online Access: https://digilib.itb.ac.id/gdl/view/87467
• Institution: Institut Teknologi Bandung

The electro-Fenton process has been a promising method in pharmaceutical wastewater treatment, especially to remove Paracetamol, a pharmaceutical compound that is often found in aquatic environments and is potentially harmful. This study evaluates the efficiency of electro-Fenton in the degradation of Paracetamol using the Platinized Mesh electrode as the anode and the Carbon Graphite Plate as the cathode and the Fe²? ion source as the catalyst. Key operational parameters, such as pH, Fe²? concentration, and reaction time, are optimized to achieve maximum degradation. The results showed that the electro-Fenton method was able to remove 70.35% of Paracetamol within 50 minutes under optimal conditions. The degradation of Paracetamol was confirmed through Total Organic Carbon (TOC) measurements and intermediate product identification by High Performance Liquid Chromatography (HPLC). To obtain an optimal design, it is necessary to analyze using machine learning modeling by including datasets from references and experiments and analyzed using Random Forest, Partial Least Square, and k-Nearest Neighbors models, and the optimal model based on RMSE, R2 , and MAE values, namely the Random Forest model with an RMSE of 7.3800247, an R2 value of 0.6722403, and an MAE of 6.2235160.