Towards Predictive Modeling of Crystal Growth and Solid-State Synthesis

Computational models, informed by experimental characterizations, provide insights into processing and synthesis of materials and facilitates the design of processing conditions that lead to desired outcomes, such as fast synthesis time and high completion rate and high quality of solidified crystals. They offer fundamental understanding of complex, dynamic phenomena that are difficult to observe experimentally. This dissertation presents two sets of computational models and methods, first to predict heat transfer during crystal growth processes and second to understand reaction progression during solid-state synthesis.

First, a heat-transfer model for simulating temperature distribution in sample rods in optical floating zone (OFZ) crystal growth furnaces is developed and validated against experimental temperature measurements. It is then utilized to conduct parametric studies to understand the effect of experimentally controllable parameters on the sample temperature profiles, which directly determine the grown crystal quality. Second part focuses on developing computational models to simulate the reaction progression during solid-state synthesis. An electrical conductivity model, coupled with a phase-field model for the ion exchange reaction, is employed to simulate the process of conductive path formation during FeS₂ synthesis. Finally, this phase-field model, employing the effective medium approximation, is utilized to quantitatively predict the progression of a reaction to synthesize LiFeO₂. The findings of this dissertation can guide informed decisions during crystal growth and synthesis experiments to optimize experimental conditions and improve processing efficiency.

CO-CHAIRS: Katsuyo Thornton and L. Jay Guo