DMREF: Data Driven Discovery of Conjugated Polyelectrolytes for Neuromorphic Computing

DMREF：用于神经形态计算的共轭聚电解质的数据驱动发现

• 批准号: 1922042
• 负责人: Gang Lu
• 金额: $ 174.96万
• 依托单位: The University Corporation, Northridge
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-10-01 至 2024-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1922042&HistoricalAwards=false
• 关键词: DMREF; Data; Driven; Discovery; Conjugated

DMREF: Data Driven Discovery of Conjugated Polyelectrolytes for Neuromorphic ComputingNon-Technical Description: As a potentially disruptive technology, neuromorphic computing breaks away from the current performance-limiting conventional computer architectures (i.e. von Neumann paradigm) by developing biologically inspired computational devices with artificial intelligence capabilities. Organic electronic materials have recently emerged as attractive alternatives to inorganic counterparts in neuromorphic computing owing to their low-energy switching, excellent tunability, low fabrication costs, and biocompatibility. In this project, we will establish a collaborative, multidisciplinary and data-centric research program to accelerate the discovery of novel conjugated polyelectrolytes (CPEs) with chemical structures tailored for the demands of neuromorphic computing. The project will bear direct impact on applications ranging from neuromorphic computing, to energy generation (photovoltaic and thermoelectric materials), sensing, robotics, and pathogen mitigation. The project will also provide cutting-edge educational and training opportunities to students and postdoctoral fellows who will gain valuable experience in data science, materials informatics, and data driven material research. The PIs are fully committed to broadening participation and enhancing diversity in materials research and education by strengthening opportunities for underrepresented groups in STEM fields. Exposing the underrepresented groups to data centric research and education is a key component of the project. A set of courses will be developed at various levels in both participating institutions to prepare the students and postdoctoral fellows for the proposed research. The PIs will reach out to local high schools and community colleges through targeted recruitment, workshops, and summer camps for high school teachers.Technical Description: The proposed research will significantly accelerate the discovery of CPE materials specifically designed for neuromorphic computing. The research effort integrates high-throughput computation, machine learning, multiscale modeling, chemical synthesis, and materials and device characterization executed in a "closed loop" manner. The team will construct the first comprehensive database dedicated to CPEs, which includes a collection of structural, elastic, vibrational, electronic, dielectric, and energetic properties for over ten thousand CPEs. Based on the CPE database, the team will explore the correlations between the materials properties and formulate a set of molecular design rules. Materials characterization will be performed to validate the design rules and once validated, they will provide guidance for predictions of promising CPEs. Based on the predictions, the team will synthesize the most promising CPEs and examine their performance in the neuromorphic devices. The project has three deliverables: (1) The first comprehensive CPE database. (2) A fundamental understanding on how the backbone structure and adjacent electrostatic forces control CPE properties and a set of design rules for accelerated materials discovery. (3) A set of highly optimized and promising CPE structures for neuromorphic and optoelectronic applications. Successful completion of the project not only impacts neuromorphic computing, but also research areas, such as photovoltaics, light-emitting diodes, thermoelectrics, sensors and robotics. The potential to affect transformative breakthroughs in the rational design of neuromorphic and organic electronic materials makes a compelling case for the project.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.

DMREF：用于神经拟态计算的共轭聚电解质的数据驱动发现非技术描述：作为一种潜在的颠覆性技术，神经拟态计算通过开发具有人工智能功能的受生物启发的计算设备，摆脱了当前性能限制的传统计算机架构（即冯·诺依曼范式） 。有机电子材料最近因其低能量开关、优异的可调谐性、低制造成本和生物相容性而成为神经形态计算中无机材料的有吸引力的替代品。在这个项目中，我们将建立一个协作、多学科和以数据为中心的研究计划，以加速新型共轭聚电解质（CPE）的发现，其化学结构适合神经形态计算的需求。该项目将对神经形态计算、能源产生（光伏和热电材料）、传感、机器人和病原体缓解等应用产生直接影响。该项目还将为学生和博士后提供前沿的教育和培训机会，他们将在数据科学、材料信息学和数据驱动材料研究方面获得宝贵的经验。 PI 完全致力于通过增加 STEM 领域代表性不足群体的机会来扩大材料研究和教育的参与并增强多样性。让代表性不足的群体接受以数据为中心的研究和教育是该项目的关键组成部分。两个参与机构将在各个层面开发一套课程，为学生和博士后研究员进行拟议的研究做好准备。 PI 将通过针对高中教师的有针对性的招聘、研讨会和夏令营来接触当地的高中和社区大学。 技术描述：拟议的研究将显着加速专门为神经形态计算设计的 CPE 材料的发现。该研究工作集成了高通量计算、机器学习、多尺度建模、化学合成以及以“闭环”方式执行的材料和器件表征。该团队将构建第一个专门针对 CPE 的综合数据库，其中包括一万多个 CPE 的结构、弹性、振动、电子、介电和能量特性的集合。基于CPE数据库，团队将探索材料特性之间的相关性，并制定一套分子设计规则。将进行材料表征以验证设计规则，一旦验证，它们将为预测有前景的 CPE 提供指导。根据预测，该团队将合成最有前途的 CPE，并检查它们在神经形态设备中的性能。该项目有三个可交付成果： (1) 第一个综合 CPE 数据库。 (2) 对主干结构和相邻静电力如何控制 CPE 性能的基本理解以及一套加速材料发现的设计规则。 (3) 一组高度优化且有前景的 CPE 结构，用于神经形态和光电应用。该项目的成功完成不仅影响神经形态计算，还影响光伏、发光二极管、热电、传感器和机器人等研究领域。在神经形态和有机电子材料的合理设计中影响变革性突破的潜力为该项目提供了令人信服的理由。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Data driven discovery of conjugated polyelectrolytes for optoelectronic and photocatalytic applications

• DOI: 10.1038/s41524-021-00541-5
• 发表时间: 2021-05
• 期刊: npj Computational Materials
• 影响因子: 9.7
• 作者: Yangyang Wan;F. Ramírez;Xu Zhang;Thuc‐Quyen Nguyen;G. Bazan;G. Lu