Abstract:
Crystallization is a widely used separation and purification technique. In the crystallization process, the change of growth conditions, such as temperature and supersaturation, may lead to changes in the size and morphology of the crystallized product, which not only affects the crystal quality, but also affects the subsequent process operations. In this work, cellular automata were used to simulate the batch crystallization process of the solution at the meso-scale. The method was built in a two-dimensional space, in which each cell was represented by a square, and neighbors adopt Moore's law, that was, eight cells around the central cell were regarded as neighbors. The evolution rule was based on the crystal classical diffusion theory, and the state of each cell was described by two parameters, i.e., concentration and crystal state (crystal or solution). The change of concentration was determined by Fick's law. For simplicity, the concentration of cell was the average of its neighbors. Whether the solution was converted into crystals was measured by the Monte Carlo method. By simulation, the size distribution and morphology of the crystal at the mesoscale were obtained. At the same time, the crystallization process was visualized, which gave an intuitive understanding of the morphological feature of each crystalline state during the crystallization. The validity of our method was compared to the high-resolution finite volume method. This work provides a theoretical reference for the identification, modeling and analysis of crystallization, facilitating the industrial practice of specific crystals, and providing new insights into fine-controlling crystal morphology.