Abstract
[Display omitted]
•Thermodynamic knowledge is included in thermal characterizing deep eutectic solvent.•A universal model is developed for estimating molar Cp of deep eutectic solvents.•The model predicts 503 experimental data from eight references with the AARD = 0.27%•The LSSVR accuracy is much better than the existing correlations in the literature.•Temperature linearly increase the molar heat capacity of all deep eutectic solvents.
Deep eutectic solvents (DES) are a new class of green solvents. Reliable characterization of DESs is a prerequisite for their successful applications. The molar heat capacity (Cp) is likely an essential thermal property often measured through expensive and time-consuming experimentations. Hence, it is necessary to derive an accurate model for Cp calculation from readily available features. This study introduces a universal computational approach for calculating the Cp of 26 different DESs as a function of temperature, acentric factor, and critical properties. Ranking investigations over four accuracy indices approve that the least-squares support vector regression with the Gaussian kernel function (LSSVR-G) is more reliable than thirteen intelligent and two regression-based models. The LSSVR-G estimates 503 experimental data points of DES molar heat capacity with an absolute average relative deviation (AARD%) of 0.27%. The results show that the LSSVR accuracy is better than the existing empirical correlations in the literature.