Hoda Keshtkar - Department of Chemistry ,Shahrood University of Technology , Shahrood , Iran

In this study, densities of ionic liquids have been estimated using a combined method that includes a support vector machine and a simple group contribution method (SVM-GCM). A total data set of 6004 experimental density data points under a wide range of temperature and pressure for 390 ionic liquids (ILs), is employed to train and verify the model. A total of 4357 data points of density at several temperatures and pressures, corresponding to 300 ionic liquids, have been used to train the support vector machine. To discriminate between the different substances, the structure of the molecule, were given as input variables. Then, the ρ–T–P values of 90 other ionic liquids (1647 data points) were predicted and results compared to experimental data from the literature. The results show that the chosen support vector machine and the group contribution method represent an excellent alternative for the estimation of densities of ionic liquids with acceptable accuracy (AARD = 0.41; R2 = 0.9976; RMSE = 7.53), for a wide range of temperatures and pressures (239.754 K to 414 K and 0.1 MPa to 298,9 MPa).

Support vector machine, Ionic liquids, Density, Group contribution method