Assessment of the liquid mixture density effect on the prediction of supercritical carbon dioxide volume expansion of organic solvents by Peng-Robinson equation of state

• VLE of 13 binary systems containing CO2 and an organic solvent at high pressures were calculated by Peng-Robinson EOS.
• Classical quadratic and LCVM-UNIFAC mixing rules were compared.
• It is not possible to generalize the capacity of PR EOS to predict the liquid mixture density and volume expansion.
• There is an offset between the terms of volume expansion equation.
• The predictive capacity of the EOS to describe the liquid mixture density is hidden by this offset.

Volume expansion and liquid mixture density are key-factors to design particle precipitation processes from expanded organic solvents using supercritical fluids such as carbon dioxide. To predict volume expansion it is necessary to know the volume or the density of the liquid mixture that is in equilibrium with its vapor at a certain temperature, pressure and composition. However, equations of state that fail to accurately describe the liquid mixture density are used to predict the volume expansion and results are considered satisfactory. In this work we investigated why this occurs using vapor-liquid equilibrium data of 13 binary systems containing CO2 and an organic solvent at high pressures with wide spread application in supercritical anti-solvent particle precipitation processes. Peng-Robinson cubic equation of state was applied with quadratic (PR-QMR) and LCVM-UNIFAC mixing rules (PR-LCVM-UNIFAC), and Peneloux’s volume correction was adopted too (PR-QMR-Peneloux). The results show that there is an offset of terms during volume expansion calculation that explains why large deviations in the prediction of liquid mixture density from bubble pressure data can lead to small deviations in prediction of organic solvents volume expansion by carbon dioxide.