Pre-screening of 1-ethyl-3-methylimidazolium tetrachloroaluminate and influence of diethylene glycol on the ionic liquid in separation of molecular solutes: Activity coefficients at infinite dilution and COSMO-SAC modelling

One of the suitable and convenient techniques in predicting the feasibility of an extracting solvent for liquid mixtures is to compare the selectivity and capacity of the components to be separated at infinite dilution. In this work, the separation parameters (selectivity and capacity) were calculated from the activity coefficients at infinite dilution. The activity coefficients at infinite dilution were computed from the retention data acquired by a gas liquid chromatography operated at different temperatures, T = (313.15 - 373.15) K and at atmospheric pressure. The volatile organic solutes (alkanes, alkenes, alkynes, alcohols, ketones, aromatic hydrocarbons, tetrahydrofuran, thiophene and acetonitrile) were tested in 1-ethyl-3-methylimidazolium tetrachloroaluminate and an eutectic solvent (1-ethyl-3-methylimidazolium tetrachloroaluminate and diethylene glycol). The influ-ence of diethylene glycol on 1-ethyl-3-methylimidazolium tetrachloroaluminate ionic liquid in separation of molecular solutes was discussed. The results were interpreted in terms of molecular interactions occurring be-tween the investigated extracting solvents and volatile organic solvents. These were further elucidated by computing the excess thermodynamic parameters at infinite dilution including (enthalpies, Delta H1E, infinity, Gibbs free energies, Delta GE, infinity 1 , and entropy term, Tref Delta SE,1 infinity). The values of selectivity and capacity were calculated for the separation problems comprising azeotropic mixtures and the values were compared with other possible extracting solvents found in the literature. Both investigated solvents were suitable for the extraction of cyclo-hexane/ethanol, acetone/ethanol, or tetrahydrofuran/methanol. For the stated mixtures, the IL gave selectivity values approximately greater than the mixed ionic liquid with diethylene glycol. COSMO-SAC predictions were qualitatively in very good agreement with the experimental data. The model predicted the results in the correct quadrant. The signal for the effect of temperature on the activity coefficients at infinite dilution values were predicted correctly for the solutes in the ionic liquid and eutectic solvent.