Abstract:Highlights • ANN, XGBoost and LightBGM are applied to predict the properties of ILs-water system. • The ANN-GC model with 8 neurons in the hidden layer is the optimal one for density. • The ANN-GC model with 10 neurons in the hidden layer is the optimal one for heat capacity. • SHAP explains the contribution of the model features to density and heat capacity. Ionic liquids (ILs), because of the advantages of low volatility, good thermal stability, high gas solubility and easy recovery, can be regarded as the green substitute for traditional solvent. However, the high viscosity and synthesis cost limits their application, the hybrid solvent which combining ILs together with others especially water can solve this problem. Compared with the pure IL systems, the study of the ILs-H 2 O binary system is rare, and the experimental data of corresponding thermodynamic properties (such as density, heat capacity, etc .) are less. Moreover, it is also difficult to obtain all the data through experiments. Therefore, this work establishes a predicted model on ILs-water binary systems based on the group contribution method (GC). Three different machine learning algorithms (ANN, XGBoost, LightBGM) are applied to fit the density and heat capacity of ILs-water binary systems. And then the three models are compared by two index of MAE and R 2 . The results show that the ANN-GC model has the best prediction effect on the density and heat capacity of ionic liquid-water mixed system. Furthermore, the Shapley Additive Explanations (SHAP) method is harnessed to scrutinize the significance of each structure and parameter within the ANN-GC model in relation to prediction outcomes. The results reveal that system components ( X IL ) within the ILs-H 2 O binary system exert the most substantial influence on density, while for the heat capacity, the substituents on the cation exhibit the greatest impact. This study not only introduces a robust prediction model for the density and heat capacity properties of IL-H 2 O binary mixtures but also provides insight into the influence of mixture features on its density and heat capacity. Graphical abstract Download : Download high-res image (363KB) Download : Download full-size image

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

Modeling Study on Heat Capacity, Viscosity, and Density of Ionic Liquid–Organic Solvent–Organic Solvent Ternary Mixtures via Machine Learning

Machine learning-assisted modeling study on the density and heat capacity of ionic liquid-organic solvent binary systems

Ionic liquid binary mixtures: Machine learning‐assisted modeling, solvent tailoring, process design, and optimization

Predictive modeling on the surface tension and viscosity of ionic liquid-organic solvent mixtures via machine learning

High-Throughput Screening and Accurate Prediction of Ionic Liquid Viscosities Using Interpretable Machine Learning

Machine learning approach for the prediction of surface tension of binary mixtures containing ionic liquids using σ‐profile descriptors

Solubility of gaseous hydrocarbons in ionic liquids using equations of state and machine learning approaches

Group Contribution Methods for Prediction of Thermophysical Properties and Phase Behavior of Ionic Liquids

Insight to the Prediction of CO2 Solubility in Ionic Liquids Based on the Interpretable Machine Learning Model

A critical methodological revisit on group-contribution based property prediction of ionic liquids with machine learning

A Critical Discussion on Developing Molecular Property Prediction Models: Density of Ionic Liquids As Example

Predictive molecular thermodynamic models for ionic liquids

Insight to the prediction of CO 2 solubility in ionic liquids based on the interpretable machine learning model

The Electrical Conductivity of Ionic Liquids: Numerical and Analytical Machine Learning Approaches

Insights into modeling refractive index of ionic liquids using chemical structure-based machine learning methods

Estimation of Molar Conductivity of Ionic Liquids by a Group Contribution Method

Prediction of Henry's law constants of CO2 in imidazole ionic liquids using machine learning methods based on empirical descriptors

Molecular fingerprint-based machine learning assisted QSAR model development for prediction of ionic liquid properties

Predicting CO$_2$ Absorption in Ionic Liquids with Molecular Descriptors and Explainable Graph Neural Networks

Hybrid data-driven and physics-based modeling for viscosity prediction of ionic liquids