An Analytical Predictive Model for the Heat Capacity of Imidazolium-Type Ionic Liquids Derived Directly from Artificial Neural Network Modeling

Analytical models directly derived from artificial neural networks are for the first time used to successfully calculate the heat capacity at a constant pressure of imidazolium-type ionic liquids. Besides this important aspect of the study, the proposed model considers variables readily available for ionic liquids that have an important effect on heat capacity, according to literature information. A set of 1135 heat capacity data for 56 imidazolium-type ionic liquids have been used in the study, distributed into 41 ionic liquids for training (912 data), 11 ionic liquids for testing (172 data), and 4 ionic liquids for evaluating the predictive capabilities of the model (51 data). The results were compared with experimental data and with values reported by other available estimation methods. The analytical model keeps the good predicting capabilities of the trained network simplifying calculations since new applications do not need to run the network software again. Results show that the new simple model gives low deviations and can be used with confidence in thermodynamic and engineering calculations.