Simultaneous Design of Ionic Liquids and Azeotropic Separation for Systems Containing Water

Separation of azeotropic mixtures is a very common but challenging task, covering a wide range of industrial sectors and issues. For example, most down-stream separation problems following a synthesis step of pharmaceutical and/or biochemical processes, involve the separation of azeotropes. Also, many separation tasks in the petrochemical and chemical industries involve separation of azeotropic mixtures. A common issue with the design and operation of these separation tasks is whether or not to use solvents? And, if solvents are to be used, what kind of solvent should be used and what would environmental impact would they cause? Ionic liquids show great promise for solvent-based separation, particularly for extractive distillation-based separations, due to their negligible vapor pressures and the fact that a wide range of solubilities and other properties can be obtained through structural changes. Since a large number of azeotropes encountered include water as one of the compounds, the use of ionic liquids in solvent-based separation of water in azeotropic systems has been investigated. Along with the design of the ionic liquid being used to entrain water, the extractive distillation process has also designed as an integrated ionic liquid –extractive process design. Based on the separation desired (for example, remove the water), an ionic liquid solvent and the separation process is designed or selected from a library of designed solutions, based upon the methodology that has been developed. In order to design ionic liquids with specific physical property values, a class of predictive property estimation models is required. In this work, a group contribution approach has been developed to predict the solvent-related properties of ionic liquids, highlighted by the Hildebrand solubility parameter and a special UNIFAC group contribution method for ionic liquids. The Hildebrand solubility parameter group contribution model was compared to reported literature values for the solubility parameter of several ionic liquids. The model showed good agreement with a maximum deviation of 15% and a root mean square difference of 1.56 MPa½. The predicted solubility parameters were then used to screen for ionic liquids that are soluble with water. The predicted solubility parameters ranged in value from 19.1 to 29.8 MPa½. Experimental solubility data has been compared to the parameter values to check for consistency. The ionic liquid UNIFAC model was developed for a selected set of ionic liquid cations and anions. Group volume and area parameters were calculated using a three step procedure. First, the rules of Bondi were used for any applicable molecular groups within the cation or anion. Next, existing UNIFAC main group parameters were used for remaining undefined groups. Finally, reported literature values for the group volume and area parameters were used for any groups that still had not been defined. From experimental values for the activity coefficient at infinite dilution, group interaction parameters were fitted for the newly defined ionic liquid groups. The ionic liquid UNIFAC model was used to predict vapor-liquid equilibria for several aqueous azeotropic systems. The ionic liquids were evaluated for use as an entrainer for water in binary azeotropic mixtures where the mole fraction of water at the azeotrope is less than 30%. For promising ionic liquid candidates, the extractive distillation processes were designed using a reverse simulation approach and characterized in terms of the driving force (calculated from the predicted vapor-liquid equilibria) that corresponds to the optimal design of the separation process (in terms of number of stages, feed plate location, energy used, solvent loss, environmental impact, etc.). To ensure the feasibility of ionic liquids to be used as industrial entrainers, a correlation was also made relating molecular structure to thermal decomposition temperature. For any new synthesis-design problem involving aqueous azeotropes, all it now requires is to find the azeotropic composition of water and based on it, to identify an appropriate ionic liquid. Then the driving force is calculated for the azeotrope ionic liquid and based on it, the design of the extractive distillation process is retrieved from the database. The presentation will highlight the methodology, the tools and solutions from several case studies.