Modeling of Solubility of Organic Substances in Supercritical Carbon Dioxide

Mathematical models of solubility of various groups of organic substances in supercritical carbon dioxide are proposed. The developed models are based on the method of quantitative structure–property relationship (QSPR). The key feature of the method is the construction of a dependence between the predicted property and parameters that numerically reflect the features of the molecular structure (molecular descriptors). In the study, molecular descriptors affecting solubility are selected and dependencies of the solubility of organic substances in supercritical carbon dioxide (SC–CO2) are constructed using the multiple linear regression method. The models are developed using the developed software and analytical package and an original database containing information on the solubility of substances in SC–CO2 and information on the molecular descriptors of dissolved substances. The predictive ability of the developed models is evaluated. Based on the obtained results, a conclusion about the prospects of using the developed models for predicting the solubility of substances in supercritical carbon dioxide is made.