(540a) Thermodynamic Studies and Process Modeling for the Separation of Aromatics from Aliphatics with Ionic Liquids

The separation of aromatics from aliphatics is a routine step in oil refinery processes. Current methods for this separation include extraction, extractive distillation, and azeotropic distillation. The main problems with these methods are: 1) suitability only for aromatic concentration of the feed mixture higher than 20%, 2) low aromatic/aliphatic selectivities, and 3) low capacities. All of these problems call for the development of new technology for aromatic/aliphatic separation, with the design of an ionic liquid (IL) solvent for the extraction process being a promising research direction.

Well-known for their favorable properties including negligible vapor pressure, good thermal stability, large liquid range, and versatility for creation, ILs are highly attractive as potential substitutes for current commercial solvents [1,2,3]. Process modeling can guide the design of ILs. For example, through fundamental studies of the thermodynamic models for liquid-liquid-equilibrium (LLE) and vapor-liquid-liquid-equilibrium (VLLE), we can better understand the thermodynamic limits of the product purities one can achieve with various ILs, and we can evaluate energy and economic costs by building process models to compare the use of IL extraction solvents to traditional solvents for aromatic/aliphatic separation like sulfolane.

To study the aromatic/aliphatic extraction process, toluene-heptane mixtures are used as a model, with a current commercial solvent (sulfolane) used as the benchmark for comparison. We evaluate two methods of recovering the aromatics from the IL-based extract. The first takes advantage of the nonvolatility of ILs, by simply evaporating the aromatics out of the aromatic-IL mixture, thus recovering the IL solvent by flash evaporation. To reduce the loss of solvent, ILs with good thermal stabilities are preferred. We compare this to using a Gas-Antisolvent Process (GAP), where gaseous CO₂ is used to facilitate a liquid-liquid phase split between an aromatic-rich and an IL-rich phase [4,5]. We focus on ILs with the bis(trifluoromethylsulfonyl)imide ([Tf₂N]^−) anion paired with 1-hexyl-3-methylpyridinium ([hmpy]⁺), 1-hexyl-3-methylimidazolium ([hmim]⁺), 1-n-butylthiolanium ([bthiol]⁺), and triethyloctyl phosphonium ([P₂₂₂₈]⁺) cations.

For the thermodynamic modeling of the heptane-toluene-IL ternary LLE, COSMO-SAC and COSMO-RS are investigated for all four ILs. The ability of these models to fit the experimental ternary LLE data [6] are compared. On this basis, COSMO-SAC is chosen to model both the flash evaporation and GAP IL recovery methods.

Process models are built for the separation of toluene from heptane using an IL extraction solvent. Theoretical product purities are calculated for each of the four ILs considered, and the sensitivity of product purities to variation in the number of extractor stages, solvent flow rate, temperature and pressure of flash separation, etc. is analyzed. Capital cost and energy cost are also calculated for each IL process. Results are compared with the current commercial sulfolane process in both product purities and cost analysis. The four types of ILs are compared in terms of the desired physical properties, based on the results of the thermodynamic modeling and the process modeling.

[1] Lin, W.-C., Tsai, T.-H., Lin, T.-Y., Yang, C.-H. J. Chem. Eng. Data 2008; 53: 760-764.

[2] Cháfer, A., de la Torre, J., Font, A., Lladosa, E. J. Chem. Eng. Data 2015; 60: 2426-2433.

[3] Brennecke. J. F., Maginn. E. J. AIChE J. 2001; 47: 2384-2389.

[4] Mellein, B. R., Brennecke, J. F. J. Phys. Chem. B 2007; 111: 4837-4843.

[5] Blanchard, L. A., Brennecke, J. F.. Ind. Eng. Chem. Res. 2001; 40: 287-292.

[6] Lubben, M. J., Canales, R., Gonzalez-Miquel, M., Lyu, Y., Stadtherr, M. A., Brennecke, J. F., to be submitted.