Machine Learning-Assisted Screening of Pyrimidine-Pyrazole Derivatives as Steel Corrosion Inhibitors in Hydrochloric Solution

2025/12/11 by Jaka Fajar Fatriansyah, Muhammad Ali Yafi Rizky, Rio Sudwitama Persadanta Kaban +3 · 1 voice
Materials Science · #Corrosion Behavior and Inhibition #Hydrogen embrittlement and corrosion behaviors in metals #Machine Learning in Materials Science

paper · doi:10.1109/isriti68345.2025.11393419

openalex publication_date 2025/12/11 · openalex created_date 2026/02/21 · openalex updated_date 2026/07/29

Abstract

Industrial corrosion poses significant economic and safety risks. This study evaluates the performance of pyrimidine-pyrazole derivatives as corrosion inhibitors for steel in hydrochloric acid. Various machine learning algorithms, including K-Nearest Neighbours (KNN), Support Vector Regression (SVR), eXtreme Gradient Boosting (XGB), Gradient Boosting (GB), Extra Trees (ET), and Artificial Neural Network (ANN), were trained on molecular descriptors from AlvaDesc to predict the inhibitor efficiency based on the electronic features. Recursive Feature Elimination (RFE) was used to optimize model performance. Model accuracy was assessed using the R2score, with XGB achieving the highest precision (96.50%). Further analysis using SHAP and Pearson Correlation identified the F02N-N descriptor as the most significant feature influencing EHOMO energy. The findings validate that machine learning, particularly the XGB model, provides an accurate and cost-effective approach for screening potential corrosion inhibitors.

Citations

• alvaDesc: A Tool to Calculate and Analyze Molecular Descriptors and Fingerprints
• A machine learning framework for screening phenyl phthalimide derivatives as corrosion inhibitors based on dataset generated by DFT and molecular dynamics simulations

Discussions

• 🔬 ML screening of pyrimidine–pyrazole corrosion inhibitors (Fatriansyah et al., 2025). Using 5,000+ alvaDesc descriptors, XGBoost (150 selected features) achieved R² = 96.5% for EHOMO prediction, cl [bsky, 1 points, 1 comments]

Related

• Predicting corrosion inhibition efficiencies of small organic molecules using data-driven techniques
• Predicting the inhibition efficiencies of magnesium dissolution modulators using sparse machine learning models
• Heterocyclic Compounds as Corrosion Inhibitors: Bibliometric analysis based on Scopus and VOS viewer
• Large language models predicting the corrosion inhibition efficiency of magnesium dissolution modulators
• Application of automated network generation for retrosynthetic planning of potential corrosion inhibitors