This thesis presents an efficient and robust detailed population balance framework for simulating aerosol synthesis of structured particles using a stochastic method. This is developed in the context of the industrial titania (TiO2) process to enable extensive numerical characterisation of the pigmentary product.

A reactor network model is used to provide a modular treatment of the reactor and account for key features, including multiple reactant injections, and tubular reaction and cooling zones. This approach simplifies the flow field in order to focus computational effort on resolving particle structure using a high-dimensional particle model and its modularity offers flexibility to investigate different configurations. Initial results are presented using a pre-defined temperature profile in the network, and the particulate product is characterised by its property distributions. Numerical performance is studied, highlighting the high computational cost of simulating strong phase-coupling, fast process rates, and broad particle size distributions.

A novel hybrid particle model is developed to address these challenges. The hybrid particle model employs a univariate description of small particles and switches to a detailed particle model to resolve morphology of more complicated, aggregate particles. New simulation algorithms are presented to manage interactions between particles of each type. The hybrid model is shown to improve efficiency (resolution versus computational cost) and robustness (sensitivity to numerical parameters), while generating the same solutions and convergence behaviour as earlier models.

The reactor model is extended, utilizing the superior numerical performance of the new hybrid particle model to enable inclusion of a system energy balance for more accurate study of a broad range of process conditions, and a more sophisticated particle model to resolve particle geometry. These contributions facilitate the study of particle structure and its sensitivity to reactor design and operational choices, providing insight into how operation affects characteristics of the particles and allowing direct comparison with experimental images of the pigmentary product.