Modeling Solution Crystal Growth Processes: Toward Three- Dimensional, Transient Simulations on Massively Parallel Supercomputers

制造先进的光学、磁性和电子器件需要高质量的大型单晶，而这些晶体的生长是现代材料加工中最困难的挑战之一。决定器件中晶体性能的属性是由晶体结构或组成决定的。通常，这些特性在晶体生长过程中受到与分子尺度生长过程和宏观输运过程相联系的过程的影响。晶体生长建模的一个长期目标是通过与这些分子尺度和宏观尺度过程相关的长度和时间尺度层次来理解影响晶体质量的机制。本研究的直接目标是开发必要的工具来理解在溶液晶体生长过程中传输现象、晶体生长动力学和晶体学的相互作用。由于这些真正的三维、时变系统所带来的极端计算挑战，过去不可能对这些系统中的流体流动、传质和晶体生长习惯进行逼真的建模。然而，未来几代大规模并行超级计算机有望提供性能（在速度和内存方面），使这种计算可行。首席研究员将开发、测试和应用新的算法来模拟溶液晶体生长系统。这项工作将特别强调适用于大规模并行超级计算机的算法。这项工作的好处是巨大的，无论是在为这种类型的超级计算机开发适当的算法，还是在应用它们来理解解决方案增长系统的复杂性。

Large, single crystals of exacting quality are needed for the fabrication of advanced optical, magnetic, and electronic devices, and the growth of these crystals is one of the most difficult challenges in modern materials processing. The properties responsible for the performance of a crystal in a device are determined by crystalline structure or composition. Often, these characteristics are affected during the growth of the crystal by processes which link molecularscale growth processes with macroscopic transport processes. A long-term goal in crystal growth modeling is to understand the mechanisms which influence crystal quality through the hierarchy of length and time scales relevant to these molecular-scale and macroscale processes. The immediate goal of this research is to develop the tools necessary to understand the interactions of transport phenomena, crystal growth kinetics, and crystallography in solution crystal growth processes. The realistic modeling of fluid flow, and mass transfer, and crystal growth habit in these systems has not been possible in the past due to the extreme computational challenges posed by these truly three- dimensional, time-dependent systems. However, the coming generations of massively parallel supercomputers promise to deliver the performance (in terms of speed and memory) to make such computations viable. The principal investigator will develop, test, and apply new algorithms for modeling solution crystal growth systems. This work will place particular emphasis on algorithms appropriate for massively parallel supercomputers. The benefits of this work promise to be enormous, both in the development of appropriate algorithms for this type of supercomputer and in their application to understanding the complexities of solution growth systems.