Accurate and Efficient Atomic-Scale Simulation of Structural Evolution in Materials: Metal Thin-Film Growth

9617122 Fichthorn这项理论研究将通过开发一系列新的基于智能蒙特卡罗的动态方法，提高当前长期、原子尺度、材料动态模拟的能力。这些方法适用于演化受稀有事件动力学控制的系统。这些方法的固有优点是它们的准确性，可以与分子动力学模拟相媲美，并且计算效率高，可以模拟几分钟到几小时的长时间尺度。智能蒙特卡罗的新功能和扩展功能可以应用于许多不同种类材料的设计和处理，催化和分离，其中原子尺度动力学决定宏观结构和功能。这些方法将用于探索金属薄膜外延中动力学和形貌之间的关系。具体来说，我们将研究团簇扩散及其在三种模型系统的亚单层金属薄膜外延中的作用：Rh/Rh(001), Rh/Rh（111）和Pt/Rh（111）。最近的研究表明，在这些系统中，复杂的多原子机制可能介导团簇扩散，并导致大型团簇的意外高迁移率。这些发现对薄膜形态学以及簇扩散和岛屿生长理论的发展产生了影响，因为目前的理论没有考虑多原子扩散机制和大簇迁移率。这项研究包括发展新的计算技术和应用这些技术来研究金属原子在金属表面的沉积和运动。这些模拟和相关理论将帮助我们了解材料薄膜是如何生长和形成特定结构的。这项工作的结果将推动计算方法的发展，并为理解材料的生长提供重要的设计工具。＊＊＊

9617122 Fichthorn This theoretical research will advance the current capabilities for long-time, atomic-scale, dynamical simulations of materials by developing a family of new dynamical methods, based on Smart Monte Carlo. These methods are applicable to systems whose evolution is governed by rare-event dynamics. The inherent advantages of these methods are their accuracy, which is comparable to molecular dynamics simulations, and computational efficiency, which allows for simulation of long time scales on the order of minutes to hours. The new and extended capabilities of Smart Monte Carlo could have applications in the design and processing of many different kinds of materials, in catalysis, and in separations, where atomic- scale kinetics dictate macroscopic structure and function. These methods will be used to probe the relationship between kinetics and morphology in metal thin film epitaxy. Specifically, we will study cluster diffusion and its role in submonolayer metal thin film epitaxy for three model systems: Rh/Rh(001), Rh/Rh(111), and Pt/Rh(111). Recent studies indicate that complicated, many-atom mechanisms may mediate cluster diffusion in these systems and lead to unexpectedly high mobilities for large clusters. These findings have ramifications for thin film morphology, as well as for the development of theories for cluster diffusion and island growth, since current theories do not account for multiple-atom diffusion mechnaisms and large cluster mobilities. %%% This research is comprised of both the development of new computational techniques and the application of these techniques to the study of the deposition and movement of metal atoms on metal surfaces. These simulations, and associated theory, will help us understand how material films grow and form particular structures. The results of this work will advance computational methods and provide an important design tool for understanding the growth of materials. ***