A Study on the Crystallization Mechanism of Distillative Freezing

Low-pressure crystallization at three-phase equilibrium (distillative freezing, DF) technology is first introduced by Cheng and Cheng (1980) to separate the mixture of the volatile compounds with close boiling temperatures at a reduced pressure. Basically, the DF process is operated at triple point condition, in which the liquid mixture is simultaneously vaporized and solidified due to the three-phase equilibrium. It results in the formation of pure crystals, and liquid phase and vapor phase of mixtures. The process is continued until the liquid phase is completely eliminated and only the pure solid crystals remain in the feed. The low-pressure vapor formed in the process is condensed and removed. Therefore, it is a distillative crystallization technology, which combines distillation and crystallization to result in pure crystal products. Shiau et al. (2005, 2006, 2008) described the basic principle of DF and successfully applied DF in the separation of the xylene mixtures. In the conventional crystallization, filtration or centrifugation is needed to separate the solid crystals from the mother liquor. Then the crystalline mass is purified by a partial melting of the crystals to wash out adhering impurities on the crystal surfaces. However, the DF process is continued until the liquid phase is completely eliminated and only the pure crystals remain in the feed. The low-pressure vapor formed in the process is condensed and removed. Subsequently filtration or centrifugation is not needed since no mother liquor is present with the pure crystals. In addition, crystal washing is not required since only the pure solid crystals remain in the feed and no impurities are adhered on the crystal surfaces at the end of the operation. As the heat of vaporization is supplied by the heat of crystallization in DF, it is an energy-conserving separation method. In addition, no chemicals are added in DF. It is also a clean separation technology. This two-year proposal is to investigate the feasibility of low-pressure crystallization at three-phase equilibrium in separating the close-boiling compounds for industrial application. In the first year, the basic principle will be examined in detail. The previous simulation results are obtained based on the assumption that each stage is operated at the three-phase equilibrium. However, as the experiments take a finite time and then run under kinetic conditions, the equilibrium might not always be attained during the whole experimental run. In fact, thermodynamic considerations are insufficient to describe the DF process. In the actual DF operation, kinetic considerations should also be considered. For example, although the pure crystal should be formed based on the thermodynamic equilibrium, impurities can be incorporated into the final crystals under actual kinetic conditions. Application of this method for various close-boiling compounds will be studied. The mechanism of the impurity incorporated into crystals will be investigated. Besides, this technique will be extended for separation and purification from a multi-component system. In the second year, the model will be developed for a continuous low-pressure crystallization system. As the design of industrial-scale equipment is a challenging problem since an efficient process is required to provide the liquid mixture with a large exposure surface area available for simultaneous vaporization and crystallization during a series of three-phase equilibrium at reduced temperature and pressure. A continuous low-pressure crystallization system will be developed for industrial application. In the previous experiment, a single temperature probe is positioned in the center of the liquid feed. As only a small amount of liquid feed is used, it is difficult to measure the temperature gradients in the resulting mixture of crystal and liquid. In the future scale-up experiments, multiple temperature probes will be used to measure the temperature gradients. The effects of temperature gradients on the DF performance will be examined.

Project ID:PB9808-2377
External Project ID:NSC98-2221-E182-030