Towards the prediction of CPA pure parameters for imidazolium, ammonium, and pyridinium based-ionic liquids using QSPR study: A comparative study

The ‘Quantitative Structure-Property Relationship’ (QSPR) method for the prediction of CPA pure parameters has been applied on three datasets. New predictive QSPR models including new molecular descriptors have been developed and have been compared with former models. According to the obtained results of statistical parameters (R² > 0.90 and Q²_LOO-CV > 0.90), the predictive capabilities of the QSPR models were better for both of training and test sets than former models. It was shown that CPA parameters for some new ILs can be predicted using QSPR models. The capability of the QSPR models were assessed by calculating density and vapor pressure of ILs. It was shown that the predicted parameters by QSPR models could predict density for none-studied ILs with AARD of 5.0 % and have qualitatively low vapor pressures. A general QSPR model with cationic (ECCEN) and anionic descriptors (MW) for the prediction of ‘b’ parameter of huge number of ILs has been developed.