Predictive modeling and stability analysis of tetradecanoic acid-modified zinc/zinc oxide coatings using machine learning

Himanshu Prasad Mamgain, Jitendra Kumar Pandey, Pravat Ranjan Pati, Ranjeet Brajpuriya,
Predictive modeling and stability analysis of tetradecanoic acid-modified zinc/zinc oxide coatings using machine learning,
Next Materials,
Volume 9,
2025,
101336,
ISSN 2949-8228,
https://doi.org/10.1016/j.nxmate.2025.101336.
(https://www.sciencedirect.com/science/article/pii/S2949822825008548)
Abstract: This study investigates the mechanical, thermal, and chemical stability of tetradecanoic acid-modified Zn/ZnO (Zn-MA) coatings using machine learning (ML) for predictive modeling. Abrasion tests showed that Zn-MA coatings retained superhydrophobicity (WCA > 135°) up to 600 cycles. Among ML models, XGBoost best captured nonlinear mechanical degradation (Test R² = 0.9596), polynomial regression (order 3) accurately predicted thermal stability (Test R² = 0.9870), and random forest effectively modeled chemical resistance (Test R² = 0.8666). Integrated ML predictions under combined stresses indicated that Zn-MA coatings maintain hydrophobicity under thermal and chemical challenges, with reduction in mechanical resilience. These findings demonstrate that ML-driven modeling can reliably predict coating performance.
Keywords: Stability analysis; ZnO coating; Hydrophobicity; Low energy; Machine learning