Application of a computational model of natural deep eutectic solvents utilizing the COSMO-RS approach for screening of solvents with high solubility of rutin

The screening of natural deep eutectic solvents (NADES) to identify those with the ability to strongly solvate rutin was conducted using the COSMO-RS methodology. A NADES model was constructed that took into account the possible ionic and neutral forms of its constituents. The distributions of all forms were computed based on the equilibrium constants of neutralization reactions between amino and carboxylic acids. The proposed model was validated against the experimental solubilities of 15 NADES. A linear relationship between these data and the estimated activity coefficient values was found. The screening encompassed 126 different NADES. It was found that ten of them outperformed the best reference system. The most effective two-component solvent comprised proline combined with 2,3-diaminosuccinic acid, and the solubility of rutin in this solvent was found to be 130% greater than its solubility in the best reference system. The amino acids associated with the highest rutin solubilities were all cyclic, and the use of carboxylic acids with two carboxyl groups and a main chain consisting of two methylene groups with two amino substituents was observed to yield the best rutin solubilities. Because of the acidic properties of rutin, the presence of basic sites on the components of the NADES generally leads to enhanced solubility. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s00894-018-3700-1) contains supplementary material, which is available to authorized users.