Abstract
Predicting the interactions between solvent and solute molecules is one of the most prominent methods for the evaluation of separation performance. The infinite dilution activity coefficient and s-profile, as a superior thermophysical properties, indicate the solubility of a solute and reflecting the strength of the interaction between solution’s molecules. Depending on the infinite dilution activity coefficient of linoleic acid, as a representative for rubber seed oil, Conductor like Screening Model for Real Solvent (COSMO-RS) based screening was employed to select the efficient organic solvent for rubber seed oil extraction. The s-profiles of linoleic acid and organic solvents were investigated for a better understanding of their similarity and interactions with solute. The experimental extraction of rubber seed oil was conducted in sono-reactor to validate the computational predictions. Based on the COSMO-RS prediction and experimental validation, diethyl ether and tetrahydrofuran exhibited the greatest capacity of 18.6 and 9.0 toward the solubility of linoleic acid and the highest experimental oil yields of 27.3 % and 27.0 % as well. Whereas, acetonitrile with capacity of zero revealed poor extraction efficiency of 7 (w/w)%. In conclusions general, the COSMO-RS method proves to be a helpful tool for estimating seed oil's solubility in organic solvents, allowing for the early diagnosis of the most efficient so
