MPhil Thesis Presentation - Modelling of Crystallization in Emulsion Droplets with Monte Carlo Simulations

Solution crystallization is a common separation and purification method. Miniaturization refers to crystallization in a confined space, which is a promising approach for improved control over crystal quality attributes such as crystal size distribution (CSD). Recently, a novel membrane-assisted emulsion crystallization process has been developed, which better enables the application of miniaturization on industrial scale. Validated process models can be important for further development of the novel process, which are still absent. The objective of this research is to develop and validate a novel process model for emulsion crystallization. The model is based on master equations which describe stochastic occurrence of primary nucleation events in emulsion droplets. The master equations describe a single Markov chain within a discretized process time interval and are solved via Monte Carlo simulations by assuming constant levels of supersaturation within a time interval. Supersaturation is updated at each time interval with material balances to account for crystal growth so that primary nucleation rate can be accurately modelled to sample a primary nucleation event in the next time interval. Kinetic parameters for primary nucleation rate expression are obtained by fitting experimental data of measured induction times from 1-ml cooling solution crystallization. A new expression has been developed based on two-step nucleation theory, which is shown to have a satisfactory quality of fit to the data. However, the kinetic model obtained cannot accurately predict behaviour in emulsion droplets. Therefore, miniaturization experimental data of average crystal length is directly fitted to the proposed model, which shows an adequate description. Yet, CSDs predicted by the model do not match experimental observations well, which may have been caused by an inaccurate description of nucleation rate or constraints for fitting. In conclusion, a novel process model for emulsion solution crystallization is presented, which can capture certain process trends after parameter estimation.