A new semi-empirical model for correlation of solubility of ligands in supercritical carbon dioxide

Model correlation of solubilities of task-specific ligands in supercritical carbon dioxide medium (SC-CO 2 ) plays an important role in development of green separation methods for recovery of metal ions. In the present study, a new semi-empirical model was proposed to predict the solubility of ligands in SC-CO 2 medium and is a modified version of existing Chrastil model. The correlation ability of new model was compared with existing density-based models (Chrastil, Adachi–Lu, del Valle–Aguilera, Sparks, Sung–Shim, Mendez-Santiago–Teja, and Bartle). New model evaluation studies were carried out using reported solubility data of 31 metal ion-specific ligands. The prediction ability of new model was compared using different objective functions, coefficient of determination ( R 2 ), absolute average relative deviation (AARD), and Akaike information criteria. The proposed model is accurate for correlating the solubility of ligands with a global AARD of 8.47%. The model was found to be most suitable for predicting the solubility of organophosphorus ligands with a mean AARD of 4.67%. Therefore, the proposed model can be employed for successful prediction of solubility of ligands in SC-CO 2 medium. Graphical Abstract