EAGER: MATDAT18 Type-1: Collaborative Research: Data Driven Discovery of Singlet Fission Materials

该奖项支持在MATDAT18数据马拉松活动中点燃的材料研究人员与数据科学家的持续合作。利用单线态裂变（SF）可以显著提高有机太阳能电池的效率。单线态裂变是一种量子力学过程，可以从一个光量子产生两种导电物质，例如电子。目前，很少有材料在固体状态下表现出分子间SF，它们属于受限的化学家族。SF还没有探索过可能制造的大量分子和晶体。pi将使用计算机模拟来寻找新的SF材料的许多可能性。为此，将开发一种新的方法，一种集成量子力学模拟和机器学习的前沿进展的方法。这项研究将推动材料科学和数据科学两个领域的发展。研究生和本科生将在计算材料科学和数据科学的跨学科协作环境中进行培训，并获得高需求的可转移工作技能。该奖项支持材料研究人员与MATDAT18数据马拉松活动中点燃的数据科学家的持续合作。单线态裂变（SF）是一个光产生的单线态激子转化为两个三重态激子。最近，人们对SF的兴趣激增，因为它有可能通过从一个光子中收集两个载流子来显着提高有机太阳能电池的效率。然而，目前很少有材料能够高效地表现出分子间的SF，这阻碍了固态SF基太阳能电池的实现。可能的发色团的化学合成空间是无限广阔的，很大程度上尚未被探索。为了实现SF材料的计算发现，将开发一种新的多保真度筛选方法，该方法将不同保真度水平的量子力学模拟与机器学习（ML）和数据库挖掘相结合。机器学习算法将用于分析量子力学模拟产生的数据，并引导模拟进一步的数据采集。在GW近似和Bethe-Salpeter方程中，采用多体摄动理论方法对候选发色团固态形式的激发态特性进行高成本高保真评估。用密度泛函理论对基态特征进行低成本、低保真度的评估。然后，特征选择算法将确定哪些描述符最能预测顺丰的热力学驱动力。这些描述符将用于筛选包含没有信息或只有部分电子属性信息的晶体结构的数据库。将采用优化算法来决定采样哪些数据点以及在何种保真度水平上最大化信息增益。这项研究将促进新的分子间SF发色团的发现，并将导致实验设计领域数据科学的进步。该奖项由数学和物理科学理事会的材料研究部和数学科学部共同资助。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

NONTECHNICAL SUMMARYThis award supports continued collaboration of materials researchers with data scientists kindled at the MATDAT18 Datathon event. The efficiency of organic solar cells may be enhanced significantly by harnessing singlet fission (SF), a quantum mechanical process that can lead to the generation of two conducting species, for example electrons, from one quantum of light. Presently, few materials are known to exhibit intermolecular SF in the solid state, and they belong to restricted chemical families. The vast number of possible molecules and crystals that could be made has not been explored for SF. The PIs will use computer simulations to search the many possibilities for new SF materials. To this end, a new approach will be developed, one that integrates cutting-edge advances in quantum mechanical simulations and machine learning. This research will advance both fields of materials science and data science. Graduate and undergraduate students will train in a collaborative cross-disciplinary environment at the interface of computational materials science and data science and acquire transferrable job skills in high demand. TECHNICAL SUMMARYThis award supports continued collaboration of materials researchers with data scientists kindled at the MATDAT18 Datathon event. Singlet fission (SF) is the conversion of one photogenerated singlet exciton into two triplet excitons. Recently, there has been a surge of interest in SF thanks to its potential to significantly increase the efficiency of organic solar cells by harvesting two charge carriers from one photon. However, few materials are presently known to exhibit intermolecular SF with high efficiency, hindering the realization of solid-state SF-based solar cells. The chemical compound space of possible chromophores is infinitely vast and largely unexplored. To enable computational discovery of SF materials, a new multi-fidelity screening approach will be developed, which integrates quantum mechanical simulations at different levels of fidelity with machine learning (ML) and database mining. ML algorithms will be used to analyze data generated by quantum mechanical simulations and to steer simulations for further data acquisition. High-cost high-fidelity evaluations of excited state properties of solid-state forms of candidate chromophores will be performed with many-body perturbation theory methods within the GW approximation and the Bethe-Salpeter equation. Lower-cost lower-fidelity evaluations of ground state features will be performed with density functional theory. Feature selection algorithms will then determine which descriptors are most predictive of the thermodynamic driving force for SF. These descriptors will be used to screen databases that contain crystal structures with no information or only partial information on their electronic properties. Optimization algorithms will be employed to decide which data points to sample and at what level of fidelity to maximize information gain. This research will advance the discovery of new intermolecular SF chromophores and will lead to advances in data science in the area of experimental design. The award is jointly funded through the Division of Materials Research and the Division of Mathematical Sciences in the Mathematical and Physical Sciences Directorate.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.