Prediction of the solid solubility of anthraquinone derivatives in supercritical CO2 from the molecular structure and melting temperature by the solution model approach

Abstract

Solubility estimations and understandings are crucial tasks for designing and developing supercritical fluid technology. This study aims to develop an approach to predict the solid solubility in supercritical CO₂ for a class of structure-related compounds, anthraquinone derivatives (AQDs), by using only the molecular structure and melting temperature. A regular solution model coupled with the Flory-Huggins equation was used to generalize the total 1362 solid solubility data of 28 AQDs in supercritical CO₂. By the developed model, the solubility behavior can be predicted successfully within order consistency with an average absolute logarithmic deviation (AALD) of 0.439. To demonstrate the extrapolation ability of the model, new solid solubility data of an AQD, 1,5-diaminoanthraquinone, in supercritical CO₂ were measured, and the experimental solubilities could be predicted to be order-consistent with AALD of 0.453. In addition, the robustness of the solution model approach was confirmed by building different tests, even in an extreme case, using only 2 % of the data (28/1362) for model generalization. In summary, only by using the molecular structure and melting temperature, the developed solution model is efficient for solubility prediction within order consistency and shows excellent extrapolation ability for screening and designing new AQD molecules for supercritical fluid applications.