Multilayered Variable Selection in QSPR: A Case Study of Modeling Melting Point of Bromide Ionic Liquids

Ionic liquids (ILs) are widely used in industry as green solvent alternatives because of their exceptional solvating ability and extremely low vapor pressure. For many applications of ILs, a low melting point temperature is desirable. Several simple ILs do not exhibit a clear melting point in the accessible temperature range of the DSC apparatus. Therefore, a computational approach is required to understand the relationship between the melting point of ILs and their structural characteristics. In the present study, the authors have developed predictive quantitative structure-property relationship (QSPR) models for melting point of ILs. A pool of 376 bromide ILs having quantitative melting point data were used to develop predictive models. A multilayered variable selection strategy has been adopted for development of final QSPR models. The models would provide an important guidance for the chemists to predict melting point of bromide ILs theoretically thereby saving the time and resources involved in the experimental determination.