Machine learning models for the density and heat capacity of ionic liquid–water binary mixtures

• Developed ANN-GC model outperforms XGBoost and LightGBM for predicting density and heat capacity of IL-water binary mixtures. • SHAP analysis identifies mole fraction of IL as the most influential factor for density, while cation substituents dominate heat capacity predictions. • The proposed models enable accurate estimation of thermophysical properties, reducing need for extensive experimental measurements. • The study provides a robust framework for designing IL-water systems with desired properties for industrial applications.