SI2-SSE: Genetic Algorithm Software Package for Prediction of Novel Two-Dimensional Materials and Surface Reconstructions

SI2-SSE：用于预测新型二维材料和表面重建的遗传算法软件包

• 批准号: 1440547
• 负责人: Richard Hennig
• 金额: $ 34.47万
• 依托单位: Cornell University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2015
• 资助国家: 美国
• 起止时间: 2015-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1440547&HistoricalAwards=false
• 关键词: SI2; SSE; Genetic; Algorithm; Software

The ability to control structure and composition at the nanoscale has introduced exciting scientific and technological opportunities. Advances in the creation of nanomaterials such as single-layer materials and nanocrystals have led to improved understanding of basic structure-property relationships that, in turn, have enabled impressive progress in a broad range of nanotechnologies with applications for energy storage, catalysis and electronic devices. Yet, significant knowledge gaps persist in what single-layer materials could be synthesized and in our understanding of the nature of the surfaces of nanocrystals, particularly in the complex environment of solvents and ligands. The discovery of potentially stable novel single-layer materials and the prediction of nanocrystal surface structures are arguably among the most critical aspects of nanoscale materials. This research will provide the computational tools for the detailed prediction of the structure of two-dimensional materials and nanostructure surfaces in complex environments. This will impact the development and the design of novel nanomaterials with properties optimized for applications ranging from catalyst for chemical reactions, to energy conversion materials, to low-power and high-speed electronic devices. Progress in the field requires better computational methods for structure prediction. This project will (i) transform the Genetic Algorithm for Structure Prediction (GASP) software package developed by the PI into a sustainable scientific tool, (ii) extend its functionality to 2D materials and materials interfaces, and (iii) increase its performance by coupling to surrogate energy models that are optimized on the fly. These complementary goals will be achieved through expansion of the developer and user base, transition to portable software interfaces and data structures, and the addition of modular algorithms for functionality and performance enhancements. To enhance the functionality, the GASP algorithms will be extended to two two-dimensional materials and materials surfaces with adsorbates and ligands. To enhance the performance of the genetic algorithm, the optimization approach will be coupled to surrogate energy models such as machine-learning techniques and empirical energy models that are optimized on the fly. The publication of user tutorials, and documentation on the data structures and software interfaces will enhance the GASP codes overall utility, increase the user and developer base, and enable further extension to other data-mining and structure prediction approaches. The students involved in this project will receive extensive training and experience in algorithm development, scientific computation, and structure/property determination of complex nanomaterials. As part of the education and outreach component of the project, the PI will develop a course module on Materials Structure Predictions and widely distribute it. A weeklong workshop for students and postdocs in the third year of the project on Materials Discovery and Design will broaden the research?s impact beyond the creation of new software and the discovery of novel single-layer materials and nanocrystal surface and ligand configurations.

在纳米尺度上控制结构和成分的能力带来了令人兴奋的科学和技术机遇。纳米材料（如单层材料和纳米晶体）的创造取得了进步，提高了人们对基本结构-性质关系的理解，这反过来又使纳米技术在能源存储、催化和电子设备的广泛应用方面取得了令人印象深刻的进展。然而，在什么单层材料可以合成，以及我们对纳米晶体表面性质的理解方面，特别是在溶剂和配体的复杂环境中，仍然存在重大的知识空白。潜在稳定的新型单层材料的发现和纳米晶体表面结构的预测可以说是纳米材料最关键的方面之一。该研究将为复杂环境下二维材料和纳米结构表面结构的详细预测提供计算工具。这将影响新型纳米材料的开发和设计，这些材料的性能优化应用范围从化学反应的催化剂到能量转换材料，再到低功率和高速电子设备。该领域的发展需要更好的结构预测计算方法。该项目将(i)将PI开发的结构预测遗传算法（GASP）软件包转变为可持续的科学工具，（ii）将其功能扩展到2D材料和材料界面，以及（iii）通过与动态优化的替代能量模型耦合来提高其性能。这些互补的目标将通过扩展开发人员和用户基础，过渡到可移植软件接口和数据结构，以及增加功能和性能增强的模块化算法来实现。为了增强功能，GASP算法将扩展到两个二维材料和具有吸附剂和配体的材料表面。为了提高遗传算法的性能，优化方法将与替代能量模型（如机器学习技术和动态优化的经验能量模型）相结合。关于数据结构和软件接口的用户教程和文档的出版将增强GASP代码的整体效用，增加用户和开发人员基础，并能够进一步扩展到其他数据挖掘和结构预测方法。参与该项目的学生将在算法开发、科学计算和复杂纳米材料的结构/性质确定方面接受广泛的培训和经验。作为项目教育和推广的一部分，PI将开发一个关于材料结构预测的课程模块并广泛分发。在材料发现与设计项目的第三年，为学生和博士后举办为期一周的研讨会，将扩大研究范围。它的影响超出了新软件的创建和新的单层材料、纳米晶体表面和配体结构的发现。

项目成果

Implicit self-consistent electrolyte model in plane-wave density-functional theory

• DOI: 10.1063/1.5132354
• 发表时间: 2019-12-21
• 期刊: JOURNAL OF CHEMICAL PHYSICS
• 影响因子: 4.4
• 作者: Mathew, Kiran;Kolluru, V. S. Chaitanya;Hennig, Richard G.
• 通讯作者: Hennig, Richard G.

Split-vacancy defect complexes of oxygen in hcp and fcc cobalt

hcp 和 FCC 钴中氧的分裂空位缺陷配合物

• DOI: 10.1103/physrevmaterials.4.103608
• 发表时间: 2020
• 期刊: Physical Review Materials
• 影响因子: 3.4
• 作者: Honrao, Shreyas J.;Rizzardi, Quentin;Maaß, Robert;Trinkle, Dallas R.;Hennig, Richard G.
• 通讯作者: Hennig, Richard G.