Harnessing Machine Learning for QSPR Modeling of Corrosion Inhibitors in HCl for Mild Steel Protection

Abstract

Background: The corrosion of Mild Steel (MS) in harsh acidic environments, such as Hydrochloric acid (HCl), is a significant industrial issue with environmental consequences. Corrosion inhibitors, particularly those containing heteroatoms and aromatic rings, are a proven method for mitigating corrosion. Traditional methods for studying corrosion inhibitors often require resource- intensive experiments. Methods: This study explores the use of Quantitative Structure-Property Relationship (QSPR) modeling, a Machine Learning (ML) technique, to predict the inhibition efficiency of organic corrosion inhibitors in HCl environments. Several ML models were employed: Linear Regression (LR), Random Forest Regression (RF), Support Vector Regression (SVR), Multilayer Perceptron Regression (MLP), and XGBoost Regression (XGB). Results: The investigation revealed that some models achieved exceptional predictive accuracy with significantly reduced errors and high precision. These models offer a promising avenue for efficient corrosion inhibitor design, reducing reliance on extensive experimentation. Conclusion: This study contributes to the advancement of corrosion science and materials engineering by introducing innovative strategies for developing effective corrosion inhibitors using machinelearning- driven QSPR models.