Predicting the solubility of a solid in supercritical carbon dioxide (scCO2) can be useful for many applications. However, an accurate prediction is extremely difficult since current methods require knowledge about the properties of the solid in question or lengthy computational calculations and often yield significant deviations. Here, a new approach is proposed that allows the solubility of any solid in supercritical carbon dioxide to be rapidly predicted based solely on the structural formula of the solute. This approach is based on the prediction of Chrastil’s parameters using group contribution methods (GCMs) followed by their use in Chrastil’s equation. For this purpose, group contribution methods to estimate Chrastil’s parameters were developed using more than 6000 experimental data points from 210 solids. These parameters were used to predict the solubility data of 30 solids (around 1300 experimental points) in pure scCO2 with Chrastil’s equation, obtaining a relatively accurate fit for 70 % of the experimental points. The average absolute relative deviation was 51.22 % (less than 8.31 · 10−4 in absolute deviation), while the logarithmic deviation was 10.83 %. The deviations obtained were substantially lower than other estimations proposed in the literature. This methodology is easily extended, faster, and more accurate than other methods and does not require the use of computational methods or the estimation of solid properties. Moreover, new ways to predict the solubility of solids in supercritical carbon dioxide can be developed in the future based on this methodology.
Read full abstract