Energies and Dynamics of Defects in Copper and Aluminum ULSI Interconnects

铜和铝 ULSI 互连缺陷的能量和动力学

• 批准号: 9520371
• 负责人: Susan Atlas
• 金额: $ 23.43万
• 依托单位: University of New Mexico
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1995
• 资助国家: 美国
• 起止时间: 1995-08-15 至 2003-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9520371&HistoricalAwards=false
• 关键词: Energies; Dynamics; Defects; Copper; Aluminum

9520371 Atlas This project is part of the Computational Approaches to Real Materials (CARM95) program. The Division of Materials Research, Division of Advanced Scientific Computing, and Division of Mathematical Sciences are all supporting this research. The investigator is collaborating with researchers at Sandia National Laboratory and Los Alamos National Laboratory, and the computing will take place on the University of New Mexico parallel supercomputer located in Kihei, Hawaii. The project focuses on the microscopic properties of copper and aluminum grain boundaries and defects, and the impact of substitutional impurities and vacancies on the chemistry of these intergranular interfaces. The work is motivated by the failure of Cu and Al interconnects used in semiconductor chips. These problems arise from a phenomenon called electromigration in which the electron wind caused by high current densities results in the formation of voids that diffuse and aggregate along grain boundaries, ultimately forming cracks in the metal and leading to catastrophic failure. These calculations represent some of the largest ab initio materials calculations ever attempted. The investigator is writing the parallel molecular dynamics code using the portable, parallel C++ class library that she has developed in collaboration with the Object-Oriented Particle Simulation Project at the Los Alamos Advanced Computing Laboratory. The defect physics is coupled across quantum and mesoscopic length scales through successive ab initio refinements of the interatomic forces used in dynamic simulations. %%% This project is part of the Computational Approaches to Real Materials (CARM95) program. The Division of Materials Research, Division of Advanced Scientific Computing, and Division of Mathematical Sciences are all supporting this research. The investigator is collaborating with researchers at Sandia National Laboratory and Los Alamos National Laboratory, and the computing wil l take place on the University of New Mexico parallel supercomputer located in Kihei, Hawaii. The project is motivated by the failure of copper and aluminum interconnects used in semiconductor chips. The project simulates defects in these materials and how they evolve in the presence of high electric currents, eventually leading to failure resulting from small cracks. These calculations represent some of the largest first- principle materials calculations ever attempted. A better understanding of the atomic-level principles that apply in the simulations will help to design interconnects that have improved properties. ***

小行星9520371 该项目是真实的材料计算方法（CARM 95）计划的一部分。 材料研究部、高级科学计算部和数学科学部都支持这项研究。 调查人员正在与桑迪亚国家实验室和洛斯阿拉莫斯国家实验室的研究人员合作，计算将在位于夏威夷基平的新墨西哥州大学并行超级计算机上进行。 该项目的重点是铜和铝晶界和缺陷的微观特性，以及替代杂质和空位对这些晶间界面化学的影响。 这项工作的动机是在半导体芯片中使用的铜和铝互连的失败。 这些问题源于一种称为电迁移的现象，其中由高电流密度引起的电子风导致形成沿着沿着晶界扩散和聚集的空隙，最终在金属中形成裂纹并导致灾难性故障。 这些计算代表了有史以来最大的从头计算材料的计算。 调查员正在编写并行分子动力学代码使用的便携式，并行C++类库，她已经开发了与面向对象的粒子模拟项目在洛斯阿拉莫斯先进计算实验室合作。 缺陷物理耦合在量子和介观的长度尺度，通过连续的从头算改进的动态模拟中使用的原子间的力量。 %此项目是真实的材料计算方法（CARM 95）计划的一部分。 材料研究部、高级科学计算部和数学科学部都支持这项研究。 研究人员正在与桑迪亚国家实验室和洛斯阿拉莫斯国家实验室的研究人员合作，计算将在位于夏威夷基平的新墨西哥州大学并行超级计算机上进行。 该项目的动机是半导体芯片中使用的铜和铝互连的失败。 该项目模拟了这些材料中的缺陷以及它们在高电流存在下如何演变，最终导致小裂纹导致的故障。 这些计算代表了有史以来最大的第一原理材料计算尝试。 更好地理解在模拟中应用的原子级原理将有助于设计具有改进性能的互连。 ***