(346g) Simulation of Morphological Population Balance for Crystallization Processes Using Cellular Automata

In the crystallization process, the morphologies and sizes of final products are affected by different growth environments, by which different properties can be obtained, such as stability and solubility, biocompatibility and pharmacodynamics in pharmaceutical industry. Through the simulation of crystallization process, better understanding on process dynamic can be extracted for the quality control of crystal products. In traditional population balance (PB) models, a crystal is characterized by only one-dimensional feature size, but no morphological information is available. Morphological population balance (MPB) models have been adopted to incorporate crystal morphology information into PB simulations so that the morphology of a crystal at any moment can be reconstructed. However, the general numerical solution of MPB model can only provide the distribution of crystal morphology and size in the macroscopic, which cannot show the specific morphology of a certain crystal particle and the difference among crystal morphologies in the microscopic. Cellular automaton (CA) method was first proposed in biosystem research to represent a system through simple cell change and interaction among cells. It does not require the solution of complex differential equations, and is suitable for any chemical and physical system, especially strongly coupled systems. By using it many complex phenomena under different conditions can be visualized by computer programming. In this work, a CA method is introduced for the simulation of crystallization processes with the consideration of growth and nucleation in MPB equations, and a general simulation platform based on CA-MPB was established. Potassium dihydrogen-phosphate (KDP) crystallization was used as a case study to demonstrate the proposed approach. The performance of the proposed scheme is verified by benchmark examples in the literature. By proposed method, the lower computation time can be achieved with similar accuracy. In addition, a visual display can be obtained at the same time, i.e. the morphology, quantity, diameter and distribution of particle during crystallization process are available at each moment, which provide a theoretical reference for identification, modelling and analysis of crystallization, and eventually can be used in the industrial practice of specific crystals.