Prediction of CO2 solubility in polar aprotic solvents using the PC-SAFT equation of state

Abstract

In this work, the PC-SAFT equation of state (EoS) is used to predict the solubility of CO₂ in polar aprotic solvents (PASs). The PAS molecules are modeled as pseudo–associating molecules to mimic the polar interaction. Four scenarios are considered to study the CO₂ solubility in PASs. In scenario I, the CO₂ and PAS molecules are considered inert molecules (non-associating), in scenario II, the CO₂ molecules are modeled as inert, in scenario III the PAS molecules are considered inert, and in scenario IV all components are modeled as associating molecules. In scenario IV, the cross-association between PAS and CO₂ molecules is considered. The vapor-liquid equilibrium (VLE) of binary systems is predicted without using any additional adjustable parameters. The ARD values of scenarios I to III for all binary systems are almost the same. The average ARD values of scenarios I to III, and IV are about 22.3 %, and 11.6 %, respectively. The results show that by considering the cross-association between CO₂ and PAS the model performance improves dramatically. Therefore, considering non-auto associative molecules as associative ones is a trick that improves the model prediction capability.