Thermal conductivity (λ) is a crucial factor in the screening and design of ionic liquids (ILs) with desired thermal properties. Given the time consumption and computational inconvenience associated with structure-based models, a model grading strategy is proposed. This strategy aims to associate λ of ILs with their three reference properties separately in a two-grade model based on the kinetic theory and previous work. Subsequently, the model parameters are derived from the gathered database, and the prediction of λ is made with an average relative deviation (ARD) less than 5.78%. Further comparisons were conducted with the other three models, illustrating that our model exhibits high accuracy and is easily applicable, requiring only one of three reference properties. Finally, this model grading strategy is expanded to the determination of the thermal conductivity of binary mixtures (λm) of ILs without necessitating the individual λ values of each component. This approach accurately predicts the λm with an ARD less than 4.20%.
Read full abstract