ITR/AP(MPS): Collaborative Research on Large-Scale Dislocation Dynamics Simulations for Computational Design of Semiconductor Thin Film Systems

这是信息技术研究计划下的合作研究奖。合作者(DMR-0113555)是加州大学洛杉矶分校的N.Ghoniem教授。这项研究涉及大规模位错动力学模拟方法的开发和应用。随着纳米和微米结构制造和工程应用的最新进展，迫切需要能够预测这些结构的可靠性和失效倾向的方法。这笔赠款将开发基于离散位错动力学的Fortran 90/95并行计算机软件，该软件将预测亚微米半导体微电子的塑性变形和失效。开发的软件将旨在(1)设计半导体薄膜-衬底材料系统所需的机械性能以实现最佳可靠性；(2)开发新的计算机体系结构，用于并行、大规模模拟复杂拓扑结构的线缺陷的动力学，这些缺陷通过远程力场相互作用；以及(3)加强研究生和本科生的教育培训。将开展以下项目：(1)研究各向异性材料中的单个和集体位错相互作用现象，这决定了半导体器件的塑性和失效；(2)参数位错动力学与有限元和分析弹性方法的多尺度耦合；(3)在并行、可扩展的计算机集群上开发独特的软件，以模拟与长程力场相互作用的复杂的拓扑结构的线缺陷的集体行为；(4)应用所开发的软件来研究一些关键的物理机制，包括错配和穿线位错环运动；位错-位错相互作用；连接和JOG形成；位错的湮灭和增殖；位错与晶界、自由表面和双材料弹性界面的相互作用；位错与点缺陷、析出物和裂纹的相互作用；热残余应力的影响；缓冲层和超晶格的计算设计；以及(5)半导体系统的大规模模拟和优化，为新一代微电子的工程设计提供指导。这些研究将提高我们在纳米到细观尺度上对塑性流动和破坏的理解，并对高性能计算和协议提出挑战。模型将与实验数据进行比较，并将用于设计可靠的微电子设备。%这是信息技术研究计划下的合作研究奖。合作者(DMR-0113555)是加州大学洛杉矶分校的N.Ghoniem教授。这项研究涉及大规模位错动力学模拟方法的开发和应用。随着纳米和微米结构制造和工程应用的最新进展，迫切需要能够预测这些结构的可靠性和失效倾向的方法。这笔赠款将开发基于离散位错动力学的Fortran 90/95并行计算机软件，该软件将预测亚微米半导体微电子的塑性变形和失效。开发的软件将旨在(1)设计半导体薄膜-衬底材料系统所需的机械性能以实现最佳可靠性；(2)开发新的计算机体系结构，用于并行、大规模模拟复杂拓扑结构的线缺陷的动力学，这些缺陷通过远程力场相互作用；以及(3)加强研究生和本科生的教育培训。这项研究将提高我们在纳米到细观尺度上对塑性流动和破坏的理解，并代表着对高性能计算和协议的挑战。模型将与实验数据进行比较，并将被用于设计可靠的微电子设备。

This is a collaborative research award under the Information Technology Research initiative. The collaborator (DMR-0113555) is Professor N. Ghoniem of UCLA. The research involves the development and application of methods for large-scale dislocation dynamics simulations.As a result of recent progress in manufacturing and engineering utilization of nano-and micro-scale structures, there is an urgent need for approaches that are capable of predicting the reliability and propensity of these structures to failure. This grant will develop Fortran 90/95 parallel computer software, based on discrete dislocation dynamics, which will predict plastic deformation and failure of sub-micron semiconductor microelectronics. Developed software will be aimed at (1) design of desired mechanical properties of semiconductor thin film-substrate material systems for optimum reliability; (2) development of new computer architectures for parallel, large-scale simulations for the dynamics of topologically complex line defects, which interact through long-range force fields; and (3) enhanced education training of graduate and undergraduate students. The following projects will be undertaken: (1) Investigation of single and collective dislocation interaction phenomena in anisotropic materials, which determine plasticity and failure in semiconductor devices; (2) Multiscale coupling of parametric dislocation dynamics with the finite element and analytical elasticity methods; (3) Development of unique software on parallel, scalable computer clusters to simulate the collective behavior of topologically complex line defects, which interact with a long-range force field; (4) Application of the developed software to investigate a number of critical physical mechanisms, including, misfit and threading dislocation loop motion; dislocation-dislocation interactions; junction and jog formation; dislocation annihilation and multiplication; dislocation interaction with grain boundaries, free surfaces and bimaterial elastic interfaces; dislocation interactions with point defects, precipitates and cracks; influence of thermal residual stress; computational design of buffer layers and superlattices; and (5) Large-scale simulation and optimization of semiconductor systems to provide guidelines for engineering design of new generations of microelectronics.The research will improve our understanding of plastic flow and failure at the nano-to-meso scale and represents a challenge to high performance computing and protocols. Models will be compared to experimental data and will also be used to design reliable microelectronics.%%%This is a collaborative research award under the Information Technology Research initiative. The collaborator (DMR-0113555) is Professor N. Ghoniem at UCLA. The research involves the development and application of methods for large-scale dislocation dynamics simulations.As a result of recent progress in manufacturing and engineering utilization of nano-and micro-scale structures, there is an urgent need for approaches that are capable of predicting the reliability and propensity of these structures to failure. This grant will develop Fortran 90/95 parallel computer software, based on discrete dislocation dynamics, which will predict plastic deformation and failure of sub-micron semiconductor microelectronics. Developed software will be aimed at (1) design of desired mechanical properties of semiconductor thin film-substrate material systems for optimum reliability; (2) development of new computer architectures for parallel, large-scale simulations for the dynamics of topologically complex line defects, which interact through long-range force fields; and (3) enhanced education training of graduate and undergraduate students. The research will improve our understanding of plastic flow and failure at the nano-to-meso scale and represents a challenge to high performance computing and protocols. Models will be compared to experimental data and will also be used to design reliable microelectronics.***