DMREF: Collaborative Research: Structure Genome of Metal-Insulator Transitions

非技术描述：世界已经看到了计算能力的巨大增长，但半导体行业目前的前进道路上充满了路障。未来一代电子设备的另一种战略是基于存在于多种电子状态的材料。为了实现这一目的，需要新一代的电子材料，以及在多种电子状态之间切换的手段。这个项目的重点是基础科学，其核心是当某些材料在外部触发器的作用下通过所谓的从金属到绝缘体的转变时，其电学性质的突然变化。在过渡的一侧，该材料的行为类似于铜金属，而在另一侧，它的行为类似于绝缘木。该项目的目标是设计和发现具有这种金属到绝缘体转变的材料，使其能够在室温下运行，并显示出感兴趣的关键属性--电阻率的巨大变化。其策略是通过原子尺度的结构设计来控制属性。该方法采用了实验、理论和文献数据挖掘的紧密结合，这将使新的见解出现，并有助于设计理想的材料。该项目将提供一个研究工作流程和一套工具，以实现对发现关键材料的新概念的评估和实验验证。该项目将阐明选择高性能材料的协议，从而扩大可能影响未来技术的化合物调色板。对学生的教学和培训与该项目的发现能力交织在一起，旨在通过调查人员及其小组成员参与公共外联活动来扩大参与。为本科生和研究生课程开发模块以及让学生参与跨学科团队环境是项目计划的内在要求。该项目将产生大量关于一系列氧化物和新的计算材料方法的新数据和挖掘数据。这些将被聚集到公共门户上的开放访问数据库中。技术描述：该项目将通过在单元单元级晶体结构和宏观电子响应之间建立联系的方法，从理论、数据和综合实验的耦合受益于发现显示金属到绝缘体转变的材料的原子级遗传代码。目前，破译金属到绝缘体转变的遗传密码(通用描述符)的基本数据和结构-电子功能关系不存在于允许预测合成的格式中。S项目的意义在于，它将问题重新归结为原子结构的问题，重点关注不同类型的结构扭曲的作用，特别是呼吸模式、扬特勒扭曲，以及广泛的结构类型和化学物质中的类Peierls不稳定性。该项目将生成和收集一系列数据，以便将电子相互作用映射为原子特征，应用基于信息学的方法，实现监督和非监督学习。该项目将阐明选择高性能材料的预测规则和协议，从而扩大化合物的调色板，这些化合物可能会影响电子以外的技术。对多层次学生的教学和培训与该项目的发现能力相互交织在一起，旨在通过以下方式扩大参与：通过公共外联活动，通过为本科生和研究生课程开发单元；最后，通过让学生参与跨学科团队环境。该项目将产生大量关于一系列氧化物和新的计算材料方法的新数据。这些数据将使用为材料数据设计的新的便携文件格式汇聚到公共门户网站上的数据库中。新的数据可视化方法将允许外部用户交互、查询和分析数据，以达到本文提出的目标之外的目的。用于生成金属-绝缘体性能定量模型的数据驱动模型和信息学工作流将与上述数据和可视化工具一起托管在MIST上：金属和绝缘体结构调整平台。PIS还计划将MIST作为开源发布，并通过确保感兴趣的研究人员能够为MIST代码库做出贡献来围绕该平台建立一个用户社区。这将使该项目有更广泛的增长。这方面对高级数字基础设施办公室的软件集群特别感兴趣，该办公室为这一奖项提供了共同资金。在合成、理论和表征方面的进步将通过允许学生和学术或工业研究人员将已制定的结构-性质关系用于教育和研究目的来加强科学能力和劳动力。

Non-technical Description: The world has seen an enormous increase in computing power, but the current path forward for the semiconductor industry is beset with roadblocks. A different strategy for a future generation of electronic devices is based on materials that exist in multiple electronic states. A new generation of electronic materials are required for this purpose, as are the means for switching between multiple electronic states. The fundamental science that is the focus of this project is centered on the sudden change in the electrical properties of certain materials when they are switched through a so-called metal-to-insulator transition by an external trigger. On one side of the transition, the material behaves like copper metal, while on the other side, it behaves like insulating wood. The project goal is to design and discover materials exhibiting such metal-to-insulator transitions that enable room-temperature operation and that display large changes in the key property of interest; the electrical resistivity. The strategy is to control properties by structural design at the atomic scale. The approach employs a tightly integrated combination of experiment, theory, and data-mining of the literature, that would enable new insights to emerge and aid in the design of desirable materials. This project will deliver a research workflow with a suite of tools to enable assessment and experimental validation of new concepts for the discovery of key materials. The project will articulate protocols for selecting high-performing materials, leading to an expanded palette of compounds that could impact future technologies. The teaching and training of students and the discovery capabilities of the project are interwoven, and aimed at broadening participation through the involvement of the investigators and their group members in public outreach events. The development of modules for undergraduate and graduate courses and the involvement of students in interdisciplinary team environments are intrinsic to project plan. The project will yield a plethora of new and mined data on a range of oxides and new computational materials approaches. These will be aggregated into open-access databases on public portals.Technical description: This project will pursue discovery of the atomic-level genetic code of materials displaying metal-to-insulator transitions through approaches that establish links between unit cell level crystal structure and the macroscopic electronic response, profiting from a coupling of theory, data, and comprehensive experimentation. At the present time, the essential data and structure-electronic function relationships to decipher the genetic code (generic descriptors) of metal-to-insulator transitions do not exist in a format which permits predictive synthesis. The project?s significance is that it recasts the problem into one of atomic structure, focusing on the role of different kinds of structural distortions, notably, breathing modes, Jahn-Teller distortions, and Peierls-like instabilities across a broad range of structure types and chemistries. The project will generate and collect a range of data that will permit the mapping of electronic interactions into atomic features, applying informatics-based methods to enable supervised and unsupervised learning. The project will articulate predictive rules and protocols for selecting high-performing materials, leading to an expanded palette of compounds that could impact technologies beyond electronics. The teaching and training of students at multiple levels and the discovery capabilities of the project are interwoven and aimed at broadening participation by through public outreach events, through the development of modules for undergraduate and graduate courses; and finally, by involving students in interdisciplinary team environments. The project will yield a plethora of new data on a range of oxides and new computational materials approaches. These will be aggregated into databases on public web-portals using a new portable file format designed for materials data. New methods of data visualization will allow external users to interact, query, and analyze the data for aims beyond those proposed herein. Data-driven models and informatics workflows for generating quantitative models for metal-to-insulator performance will be hosted with the aforementioned data and visualization tools on the MIST: Metals and Insulators by Structural Tuning platform. The PIs also plan to release MIST as open source and build a user community around the platform by ensuring that interested researchers are able to contribute to the MIST codebase. This will allow a wider growth of the project. This aspect is of special interest to the software cluster in the Office of Advanced Cyberinfrastructure, which has provided co-funding for this award. Advances in synthesis, theory, and characterization will strengthen the scientific capabilities and workforce by allowing students and academic or industrial researchers to employ the formulated structure-property relationships for educational and research purposes.