(93f) Understanding the Solid-Liquid Phase Transition during the Growth of Scintillator Single Crystals Via Computational Modeling and Neutron Imaging

The Bridgman method, named after Nobel Prize winner Percy Williams Bridgman, is a broadly used and studied method to directionally solidify single crystals from the melt. The macroscopic solid-liquid interface shape in the vertical Bridgman crystal growth process is an important determinant for crystal properties and quality. However, there have been no direct methods to “see” the shape of this interface during the growth of most inorganic crystals due to the opaque, refractory materials needed for the construction of these high-temperature systems (furnace, ampoule, etc.). Thus, researchers have had to rely on computational models that can only be validated by indirect experiments.

Recently, our team has developed groundbreaking experiments that employ spallation neutrons to visualize, in situ, the compositional field that develops during the growth of a large, single-crystal mixed-halide scintillator material via the Bridgman method. These measurements provide, for the very first time, a direct observation of melt crystal growth within a system large enough to be characteristic of an industrially relevant process, thus revealing the complex interplay among phase change, solid-liquid interface shape, heat transfer, fluid flow, and dopant segregation.

In this presentation, we present initial results from our neutron imaging experiments and apply computational models to simulate the transport of heat, mass, and momentum, along with the liquid to solid phase change in this Bridgman crystal growth system. Not only do the experimental observations serve to validate the modeling results, but, more importantly, the model provides a rigorous framework in which to understand the mechanisms that are responsible for the complicated evolution of solid-liquid interface shape and dopant distribution observed in the growth experiment.

We discuss how ongoing synergies between modeling and experiment will allow for unambiguous interpretation of large scale crystal growth experiments. Ultimately, the understanding obtained by model and experiment will allow the closing of the loop between materials quality, crystal growth conditions, and process development, an undertaking that in the past has relied on empiricism and experience to typically achieve only incremental advances.

---------------

This work has been supported by DOE/NNSA DE-NA0002514 and DOE/NNSA/DNN R&D (LBNL subcontract AC0205CH11231); no official endorsement should be inferred.