Experimental determination of the LLE data of systems consisting of {hexane + benzene + deep eutectic solvent} and prediction using the Conductor-like Screening Model for Real Solvents

Abstract Recently, deep eutectic solvents (DESs) have proven to be excellent extracting agents in the separation of aromatic components from their mixtures with aliphatic compounds. The tunable properties of the DESs allow to tailor-make optimal solvents for this application. In this work type III DESs, based on chloride quaternary ammonium salts (tetramethylammonium chloride, tetraethylammonium chloride, tetrabutylammonium and tetrahexylammonium chloride) as hydrogen bond acceptor molecules and polyols (ethylene glycol and glycerol) as hydrogen bond donor molecules were used. The liquid-liquid equilibrium (LLE) data of the system {hexane + benzene + DES} was measured at room temperature and atmospheric pressure. The solute distribution ratio and the selectivities were calculated and compared. The highest solute distribution ratios were obtained for salts with long alkyl chain length as hydrogen bond acceptor molecules and for ethylene glycol as hydrogen bond donor molecules. COSMO-RS was used for the prediction of the LLE data of the studied systems. It was found that both the LLE data and the solute distribution ratio of the ternary systems containing DESs can be qualitatively well predicted using this model. Moreover, the LLE data was quantitatively predicted within a root mean square error of 10 wt% without the need of any experimental data. This implies that COSMO-RS is an effective screening tool for the optimization of the separation process of aromatic components from {aliphatic + aromatic} mixtures using DESs as extracting agents.