Computational modeling of ionic liquids density by multivariate chemometrics

The experimental measurements of important physico-chemical properties can be quite complex task especially in the case of the large group of chemistries such as ionic liquids. The main goal of the present study is to define the relationships between the structural characteristics of the analyzed ionic liquids described by molecular descriptors and their important physico-chemical property - density. For this purpose the multiple linear regression and artificial neural networks analysis were applied. Also, principal component analysis was performed in order to gain an overview of similarities and dissimilarities among the studied ILs. In addition, the detailed statistical validation procedures were done to select the best model which can be used for prediction of the density of ILs.