Collaborative Research: DMREF: Accelerated Design, Discovery, and Deployment of Electronic Phase Transitions (ADEPT)

非技术描述：在主要依赖传统半导体的技术进步的推动下，全球连接、信息处理和信息存储出现了巨大的增长。然而，它们的基础材料平台正面临着巨大的挑战。可以使用以多种电子状态存在的材料来建立下一代电子设备。可以通过外部触发从绝缘体切换到金属的材料和设备将重振美国半导体生态系统，为低功率计算系统提供新的途径，并集成到6G及更高应用的系统中。该项目的目标是设计和发现具有这种绝缘体到金属转变(IMT)的材料，使其能够在室温下运行并显示电阻率的巨大变化。该研究团队由计算和实验材料物理、数据科学和设备工程方面的跨学科专业知识组成，旨在通过培训一支装备精良、多样化、精通数据驱动的先进材料发现的劳动力队伍，实现材料研究、开发和部署方面的文化转变。利用材料基因组计划的原则，该团队将提供紧密集成的协同设计方法，以促进具有优异性能的新IMT材料的建模和合成，并最终引导设计走向创纪录的设备性能，以加强美国在未来计算、存储和通信技术和行业中的领导地位。技术描述：加速设计、发现和部署电子相变(ADEPT)项目的目标是实施加速发现和协同设计引擎，以有效部署传统上受到稀疏先验数据和系统级限制的绝缘体-金属转变(IMT)材料。要实现这一目标，需要超越传统的、线性的材料发现方法，将其转变为循环和迭代的过程。该项目制定了新的计算方法，将计算数据与高通量材料合成和表征数据融合在一起，以克服(I)从稀疏和昂贵的数据中发现材料，(Ii)高效解码高维实验数据，以及(Iii)集成时性能与性能不匹配的关键挑战。它们的融合阻碍了未来微电子和无线通信技术的新材料平台的进步。项目目标将通过创建集成的协议来执行，这些协议将标准的顺序发现步骤(假设生成、综合、表征)转换为闭环，以克服这些挑战：(1)人工智能辅助的虚拟筛选和自适应发现，(2)加速合成和表征分析，以及(3)材料集成、器件制造和合作设计。该框架的成功将允许在器件限制内实现物质目标并提供以下结果：(1)新的单相和两相IMT材料类别(2)分布在开放访问数据库中，(3)IMT行为理论，(4)新的IMT薄膜合成方法，(5)提高产量的非接触式表征方法，(6)实现共同设计材料和器件的自适应学习方法，以及(7)对设备性能的定量了解，以利于未来的可扩展性和与工业合作伙伴的制造。该奖项反映了NSF的法定使命，并已通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Non-Technical Description:The world has seen an enormous increase in global connectivity, information processing, and information storage driven by advances in technologies that rely largely on traditional semiconductors. Their underlying material platforms, however, are facing enormous challenges. A future generation of electronic devices can be established using materials which exist in multiple electronic states. Materials and devices that can be switched from an insulator to a metal by an external trigger would revitalize the U.S. semiconductor ecosystem, providing new paths to low-power computing systems and integration into systems for 6G and beyond applications. The project goal is to design and discover materials exhibiting such insulator-to-metal transitions (IMT) that enable room-temperature operation and display large changes in electrical resistivity. The research team, which comprises interdisciplinary expertise in computational and experimental materials physics, data science, and device engineering, aims to enable a culture shift in materials research, development, and deployment through training a well-equipped and diverse workforce with proficiencies in data-driven discovery of advanced materials. Leveraging Materials Genome Initiative principles, the team will deliver a tightly integrated codesign methodology to facilitate modeling and synthesis of new IMT materials with superior properties, and ultimately guide the design towards record-setting device performance to strengthen American leadership in future computing, storage and communication technologies and industries. Technical Description:The goal of the Accelerated Design, Discovery, and Deployment of Electronic Phase Transitions (ADEPT) project is to implement an accelerated discovery and codesign engine for efficient deployment of insulator-metal transition (IMT) materials traditionally marred by sparse prior data and system-level constraints. Achieving this goal requires moving beyond conventional, linear approaches to materials discovery, transforming them into a cyclic and iterative process. The project formulates new computational approaches that fuse computational data with high-throughput materials synthesis and characterization data to overcome key challenges of (i) Materials Discovery from Sparse & Expensive Data, (ii) Efficient Decoding of High-Dimensional Experimental Data, and (iii) Property-Performance Mismatch upon Integration. Their confluence hinders the advancement of novel material platforms for future microelectronic and wireless communication technologies. The project goal will be executed by creating integrated protocols that transform the standard sequential discovery steps (hypothesis generation, synthesis, characterization) into closed-loops fashioned to overcome these challenges: (1) AI-Aided Virtual Screening and Adaptive Discovery, (2) Accelerated Synthesis and Characterization Analytics, and (3) Materials Integration, Device Fabrication, and Codesign. Success with this framework will allow for the realization of material objectives within device constraints and deliver the following outcomes: (1) new classes of single and two-phase IMT materials (2) distributed in open-access databases, (3) theories of IMT behavior, (4) novel IMT thin film synthesis methods, (5) contactless characterization methods to improve throughput, (6) adaptive learning methods to achieve codesigned materials and devices, and (7) quantitative understanding of device performance to benefit future scalability and manufacturing with industrial partners.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.