Development of Computer Algorithms for Density Functional Theory Simulations and Applications to Crystal Growth

Hannes Jonsson得到了理论和计算化学计划的资助，继续他在开发用于密度泛函理论模拟和半导体晶体生长应用的计算机算法方面的研究。Jonsson将并行化他自己寻找最小能量路径和过渡态激活势垒的方法。他还将研究包括非谐性在内的自由能垒，以及随系统规模线性调整的算法。他的应用将侧重于通过计算与工业相关的系统(如硅和碳化硅表面、锗-硅界面和二氧化硅覆盖层)的吸附原子扩散速率和表面团簇形成速率来解释当前的成像实验。现代密度泛函理论方法将被发展并用于研究各种具有技术重要性的表面生长机制。特别是，这些方法将得到扩展，以便能够在新出现的大规模并行计算机上研究更大、更复杂的材料模型。这些方法包括对半导体表面上许多复杂的原子尺度过程的精确能量学进行复杂的评估。

Hannes Jonsson is supported by a grant from the Theoretical and Computational Chemistry Program to continue his research in the development of computer algorithms for density functional theory simulations and applications to semiconductor crystal growth. Jonsson will parallelize his own methods for finding minimum energy paths and transition state activation barriers. He will also investigate free energy barriers which include anharmonicity, and algorithms which scale linearly with system size. His applications will focus on interpretation of current imaging experiments by calculating adatom diffusion rates and surface cluster formation rates for industrially relevant systems such as silicon and silicon carbide surfaces, germanium-silicon interfaces, and silicon dioxide overlayers. Modern density functional theoretical methods will be developed and used to study a variety of surface growth mechanisms of technological importance. In particular these methods will be extended to enable the study of larger, more complex models of materials on new and emerging massively parallel computers. The methods involve the sophisticated evaluation of accurate energetics for a number of complex atomic-scale processes on semiconductor surfaces.