(540c) A Molecular Design Method Based on the COSMO-SAC Model for Ionic Liquid in Extractive Distillation
AIChE Annual Meeting
2017
2017 Annual Meeting
Separations Division
Developments in Extractive Separations: Solvents
Wednesday, November 1, 2017 - 1:20pm to 1:45pm
A Molecular Design Method Based on the COSMO-SAC
Model for Ionic Liquid in Extractive Distillation
FANG Jing, Li Chunli, Wang Yijing, Zhao Rui
(School of Chemical Engineering, Hebei University of Technology, Tianjin, China)
Extractive distillation is widely used for azeotropic systems because of its superiority. At present, most organic solvents or salt are used universally as extraction solvents. Comparing with traditional organic solvents, the radio of solvent to feed using ILs could decrease 20% and the heat duty on reboilers in extractive distillation could decrease about 17%. However it is difficult to select appropriate ILs for one certain azeotropic system.
In this article, a COSMO-SAC model was used to screen for prospective solvent from a wide variety of ILs for extractive distillation. In 1983, UNIFAC model which was applied in molecular design of solvent was based on the group contribution method (GCM). Because of absence of group interaction parameters, it is not suitable for selecting ILs. In 1995, COSMO-RS which was based on the solvent molecular orbital continuum model was proposed by Klamt. This method can provide promising results but the equation for chemical potential does not converge for boundary conditions and final expression for activity coefficient fails to satisfy the thermodynamic consistency relationships. The COSMO-SAC model was introduced by Lin and Sandler, which has an obvious advantage is that the interaction parameters are too limited to perform the UNIFAC calculation.
In order to reduce the calculation workload, dividing the ILs into cations and anions. Applying materials studio to drawn the structure of ILs and optimizing its
structure and energy. Based on the COSMO-SAC model, the Ï-profile database of ILs
was established. Selectivity and solubility were used as the indexes for solvent
screening. Solvents can change the relative volatility of the components but it is difficult to calculate the activity coefficient of each component. Turning to the other aspect, the infinite dilution activity coefficient of component is easy to obtain. The larger the value, the better the separation results. Thus, selecting the solvent which has high infinite dilution activity coefficients will obtain better separation results. According to the molecular design method, three suitable extractive distillation were determined. Figure 1 shows structure of three ILs.
There are two systems-acetonitrile-water and ethanol-cyclohexane are selected to test the results of chosen ILs. The first system is a typical polar binary azeotropic system while the other is a binary azeotropic system composed of one polar component and one nonpolar component. Adding various weight fractions (15 and 30%) into two chosen systems to obtain the vapor-liquid equilibrium database. The database of acetonitrile-water system are plotted in Figures 2-4 while ethanol- cyclohexane in Figures 5-7.
From these figures, we can conclude that the azeotropic point vanished when the mass fractions of the ILs reached 30%, the accuracy of the chosen ILs are proved again. In this article, the Wilson equation was used to predict the vapor composition for molecular design. Comparing the experimental and calculated relative volatilities of two systems with three types of ILs. Relative volatility for compared of experiment data and calculated results were shown in figure 8 and figure 9. Through calculation, for acetonitrile-water-IL systems the overall average relative deviation was 4.92%, while
ethanol-cyclohexane-IL was 5.45%.