Thermodynamic modeling for supercritical fluid process design

Abstract

The present paper describes a predictive model for ascertaining the process parameters toward establishing an economically viable supercritical fluid technology. The model utilizes the Peng-Robinson equation of state with a modified covolume-dependent (CVD) mixing rule to be able to predict the solubility of mixed solids in pure and mixed supercritical carbon dioxide from minimum information, such as the pure component solubilities. The model can predict the solubility data at various temperatures by virtue of its temperature-insensitive CVD interaction parameter. In order to make the model completely predictive, the CVD interaction parameter has been correlated in terms of the easily available pure component properties, such as solid molar volume, van der Waals volumes, and dipole moments. For mixed solvent, however, a more rigorous correlation is needed to account for stronger solute-cosolvent interactions. The validity of the model has also been tested for the solubility predictions of liquid solutes in the supercritical solvent. Slightly higher deviations of the liquid-phase compositions indicate that the mixing rule is rather well suited for the dilute supercritical solution. The utility of the model has been demonstrated for separation of palmitic acid from tripalmitin. The process parameters have been established under the equilibrium condition utilizing the well-predictedmore » crossover pressures and solubility behavior with temperature.« less