466265 Ionic Liquids As Entrainers for the Separation of Aliphatic/Aromatic Mixtures By Extractive Distillation Systems: Rigorous Model Based Process Optimization

Wednesday, November 16, 2016
Grand Ballroom B (Hilton San Francisco Union Square)
Ismael Diaz1, Manuel Rodriguez1 and Jose Palomar2, (1)Chemical Engineering, Technical University of Madrid, Madrid, Spain, (2)Sección de Ingeniería Química (Departamento de Química Física Aplicada), Universidad Autónoma de Madrid,, Madrid, Spain

The great energy consumptions are one of the main workhorses that the process industries are addressing in the XXI century. It is well known that the most important consumptions are due to separation processes. Besides, it is common that the more complex operation, the greater energy consumption, which is especially true for the separations of azeotropic mixtures. Azeotropes and close boiling point mixtures are quite usual in chemical industries appearing in solvent, gasoline or monomer industries, making impossible (or very expensive) the separation of individual components of the mixtures by conventional distillation. This is overcome by using, on one hand, material separation agents in operations such as adsorption or liquid-liquid extraction, or by adding an entrainer to a distillation column (azeotropic or extractive distillation). This work is focussed on the application of recently developed solvents, called ionic liquids, as entrainers for the separation of a low differential volatility mixture (n-heptane/toluene). Separation of aromatic hydrocarbons from naphtha has a great interest for the petrochemical industry. On one side, aromatic hydrocarbons represent an extra load for the naphtha crackers, negatively affecting the process economy as additional operative and capital costs. On other side, aromatic hydrocarbons have their own value as raw materials, being furtherly transformed in higher value-added products. Between all technologies studied sulfolane has resulted the most extensively used thanks to an UOP technology, which employs it as extracting solvent. In this process, sulfolane has to be recovered from both the extract and the raffinate after liquid-liquid extraction, which entails high investment and energy costs. Ionic liquids (ILs) are a novel attractive candidate for this separation since Seddon in 1997 reported the great capacity of imidazolium ionic liquids to dissolve benzene. Ionic liquids are molten salts with a melting point below (or near) ambient conditions. They are composed by an anion and an organic cation, so they can be chosen in order to tune their physical properties (tailor made solvents). The application of ionic liquids for the separation of aliphatic and aromatic components is being under research, mainly by experimental measurements of liquid-liquid or vapor-liquid equilibria (isobaric or isothermal measurements) [1]. On the other hand it is also under research the development of accurate models for the correlation and prediction of their properties. In this way, in the last decades new physically founded methods based on quantum chemistry theory have emerged and have been largely applied due to the current computational facilities. That is the case of the COSMO-RS model developed by Klamt [2] or the COSMO-SAC model of Li et al. [3]. These COSMO-based models are a priori methods that predict thermodynamic properties of fluid mixtures by solely using the molecular surface polarity distributions of the individual compounds, obtained by quantum chemical methods. Many thermodynamic properties can be calculated by COSMO-based models, such as vapour pressure, normal boiling points, excess properties and phase equilibria data. Additionally, some COSMO-based molecular descriptors can be used to estimate some properties by quantitative-structure-property-relationships (QSPR) such as viscosity, density or partition coefficients in polymers. In the last years we have successfully integrated COSMO-RS results into process simulations using Aspen Tech’s software in order to develop potential new industrial processes with mixtures of components not available in the simulators’ databases. We have applied this procedure to biotechnological processes and especially to processes involving ionic liquids (ILs), our main goal being to provide technical and economic information about the potential industrial applications of ILs. All this has the aim of enriching the design/selection of ILs with optimized properties for specific applications.

In this work, we presented the study of how ionic liquids can increase the efficiency of n-heptane/toluene mixtures separation. We have made a screening of potential ionic liquids to be used with the COSMOTherm software and selected a small group of them. Then, following the previously described methodology we have implemented such components in the Aspen Properties databanks by specifying the required properties. Rigorous process simulations of the extractive distillation column can be then developed. IL Mixture design was carried out by defining an optimization problem in Matlab and solved using derivative-free optimization methods. A COM interface was needed in order to connect Matlab with the Aspen Plus software. The objective function implemented to design the entrainer mixture took into account both separation an economic criteria.

Besides, we have correlated the experimental liquid-liquid equilibrium [4] and vapor-liquid equilibrium [5] data of some ionic liquid – solvent systems to conventional activity coefficient thermodynamic models (NRTL) in order to evaluate COSMO predictions. Binary interaction parameters were then implemented in Aspen Plus and some process simulations were carried out in order to determine the deviation between the regressive thermodynamic model (NRTL) and the predictions of the COSMO based simulations.

Sothis work is structured as follows: First a short introduction to ionic liquids and to the COSMO based model is shown, then the methodology employed to couple quantum calculations with Aspen Plus and the optimization strategy is described. Third, a detailed comparison of COSMOSAC predictions against regressive thermodynamic models (NRTL) is shown in terms of process simulation results. The next section addresses the optimization problem of the extractive distillation process selecting the best ionic liquid or mixture for the separation proposed and finally the main conclusions of the work are inferred.[1] Navarro, P., Larriba, M., García, J., Rodríguez, F. Vapor-liquid equilibria for n-heptane + (benzene, toluene, p-xylene, or ethylbenzene) + [4empy][Tf2N] (0.3) + [emim][DCA] (0.7) binary ionic liquid mixture. Fluid 540 Phase Equilibria 2016;417:421

[2] Klamt, A. and F. Eckert, COSMO-RS: a novel and efficient method for the a priori prediction of thermophysical data of liquids. Fluid Phase Equilibria, 2000. 172(1): p. 43-72.

[3] Lin, S.-T. and S.I. Sandler, A Priori Phase Equilibrium Prediction from a Segment Contribution Solvation Model. Industrial & Engineering Chemistry Research, 2002. 41(5): p. 899-913.

[4] Larriba, M., Navarro, P., Garcia, J., Rodríguez, F.. Liquid-liquid extraction of btex from reformer gasoline using binary mixtures of [4empy][tf2n] and [emim][dca] ionic liquids. Energy and Fuels 2014;28(10):6666:6676

[5] Navarro, P., Larriba, M., García, J., González, E.J., Rodríguez, F. Vapor-liquid equilibria of n-heptane+ toluene+ [emim][DCA]g system by headspace gas chromatography. Fluid Phase Equilibria 2015;387:209:216.

Extended Abstract: File Not Uploaded
See more of this Session: Poster Session: Process Development
See more of this Group/Topical: Process Development Division