Prediction of solubility of biomolecules in supercritical solvents

Abstract

The supercritical .uid extraction (SFE) is considered an appropriate alternative for separation of biomolecules from food and pharmaceutical products. A major di2culty in utilizing the SFE for biomolecules has been the di2culty in measurement and prediction of their solubilities in supercritical solvents at various pressures and temperatures for process optimization. Lack of data for intermolecular energy parameters and=or critical properties, acentric factors, and molar refractions limits us to the use of the simple equations of state for prediction of their solubilities in supercritical solvents. In this report, six di8erent cubic equations of state are used to predict the solubility of cholesterol and beta-carotene, as two representative biomolecules, in supercritical .uids. They are the van der Waals, Redlich–Kwong, Mohsen-Nia–Moddaress–Mansoori (MMM), Peng–Robinson (PR) and Patel–Teja and modi=ed PR equations. It is shown that the two-parameter MMM equation is more accurate than =ve of the other equations and comparable to the modi=ed PR equation in predicting the solubility of cholesterol and -carotene in supercritical .uids.