NSF/DMR-BSF: Density Functionals for Predictive Excited-State Calculations of Solids (NSF-BSF Application)

非技术总结该奖项支持理论和计算研究，以及推进根据基本科学原理预测材料性质的计算方法的教育。光电子应用新材料的发现和发展受到对材料如何收集光、转换能量和传输电荷的详细了解的严重限制-所有这些现象都涉及电子激发态，其中电子的组态导致比材料的最低或基态能量更高的能量。现有的计算方法是对这种过程的预测，但它们带来了巨大的计算成本，而具有类似精度的替代方法将能够预测日益复杂的材料，并采用这种方法来发现和设计材料。这一研究项目为开发更有效的预测理论框架奠定了重要的基础，这些框架是对计算成本更高的实际材料中电子激发态的现有方法的补充。这项工作的核心是指导所有年龄段的下一代计算材料理论家，重点是积极招募和提升女性和其他代表性不足的本科生和研究生；并通过参观湾区内外的本科生、小学生和中学生以及教育工作者来推广当地的研究设施。技术总结该奖项支持理论和计算研究，以及促进预测材料性能的计算方法的教育。电子能带结构是晶体物质的基本性质。它是理解块状材料电荷输运性质的基础。此外，它也是理解材料的光学性质和使光谱测量结果合理化的先决条件。在材料和凝聚态物理中，定量确定能带结构的形式一直是多体微扰理论(MBPT)。这种形式主义已经为许多不同类别的金属、半导体和绝缘体产生了出色的电子结构预测。然而，这些预测带来了巨大的计算成本，而从密度泛函理论(DFT)中提取能带结构的能力，基于Kohn-Sham方程的解获得的单电子能量和轨道，可以减轻这一成本。这个项目涉及两国的理论和计算合作，通过将最优调谐范围分离杂化(OTRSH)密度泛函与多体微扰理论相结合，以更高的精度和效率为复杂固态材料的准粒子带隙、能带结构和光谱的第一性原理计算预测开发了一个健壮的框架。PI的工作表明，OTRSH泛函导致了一大类分子晶体以及一组IV和III-V族半导体和绝缘体的能带结构和光谱，具有前沿从头计算GW和GW-BSE方法的准确性。在这方面，私人投资机构通过同时推进两个重要战线来巩固这一成功。首先，PIS将评估OTRSH作为固体GW和GW-BSE计算的有效起点。其次，在以前工作的基础上，PI将探索无需GW计算即可确定准确的DFT-OTRSH能带结构和TDDFT-OTRSH光学性质的方法。一旦得到验证，PI将使用OTRSH起始点计算一系列复杂材料的准粒子带隙和能带结构以及线性吸收光谱，包括卤化物钙钛矿，重要的光电材料，对于这些材料，标准的GW和GW-BSE方法已被证明是不适当的或不一致的。到目前为止，OTRSH在固体的能带结构和光谱方面取得的进展令人鼓舞，因为它表明，通常可以将一个参数固定到取向平均介电常数，然后调整单个参数-距离分离参数-以产生与GW-BSE一致的能带结构和光学光谱，部分或完全降低与MBPT相关的计算成本。考虑到单参数OTRSH能带结构的质量，OTRSH能成为复杂结构和化学非均相体系的GW和GW-BSE计算的最佳起点吗？此外，人们是否可以独立于GW-BSE计算和实验来设置这一单一参数还有待观察。对于各向同性和各向异性材料，是否可以通过计算易于处理的方式从材料属性中预测它？这个项目的目的是解决这些问题，并最终开发有效的方法来理解复杂材料中现有的和预测新的激发态现象。该奖项反映了NSF的法定使命，并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。

NONTECHNICAL SUMMARYThis award supports theoretical and computational research, and education to advance computational methods for predicting the properties of materials from fundamental scientific principles. The discovery and development of new materials for optoelectronic applications is significantly limited by a detailed understanding of how materials harvest light, transduce energy, and transport charge - all phenomena involving electronic excited states where the configuration of electrons leads to a higher energy than the lowest or ground state energy of the material. Existing computational methods are predictive for such processes, but they come at significant computational cost, and alternative approaches with similar accuracy would enable predictions for increasingly complex materials and for adapting such methods for materials discovery and design. This research project lays important groundwork toward the development of more efficient predictive theoretical frameworks that are complimentary to more computationally costly existing methods for electronic excited states in real materials. Central to this effort is mentoring of next-generation computational materials theorists at all age levels, with focus on active recruitment and promotion of women and other underrepresented-minority undergraduate and graduate students; and outreach through tours of local research facilities for undergraduate, elementary, and middle-school students – as well as educators – in the Bay area and beyond.TECHNICAL SUMMARYThis award supports theoretical and computational research, and education to advance computational methods for predicting the properties of materials. The electronic band structure is a fundamental property of crystalline matter. It serves as the basis for understanding charge transport properties of bulk materials. Moreover, it is a prerequisite for understanding optical properties of materials and for rationalizing the results of spectroscopic measurements. In materials and condensed matter physics, the formalism of choice for quantitative determination of the band structure has long been many-body perturbation theory (MBPT). This formalism has yielded excellent electronic structure predictions for many different classes of metals, semiconductors, and insulators. However, these predictions come at significant computational cost, and the ability to extract band structures from density functional theory (DFT), based on the single-electron energies and orbitals obtained from the solution of the Kohn-Sham equation, could alleviate this cost. This project involves a binational theoretical and computational collaboration to develop a robust framework for the first principles computational prediction of the quasiparticle band gaps, band structures, and optical spectra of complex solid-state materials with greater accuracy and efficiency than existing methodologies, by combining optimally-tuned range separated hybrid (OTRSH) density functionals with many-body perturbation theory. The PIs’ work has shown that OTRSH functionals lead to band structures and optical spectra for a broad class of molecular crystals and a set of group IV and III–V semiconductors and insulators, with the accuracy of leading-edge ab initio GW and GW-BSE approaches. Here, the PIs build on this success by advancing two important fronts simultaneously. First, the PIs will evaluate OTRSH as an effective starting point for GW and GW-BSE calculations for solids. Second, building on prior work, the PIs will explore routes to determine accurate DFT-OTRSH band structure and TDDFT-OTRSH optical properties without GW calculations. Once validated, the PIs will use OTRSH starting points to calculate the quasiparticle gap and band structure, as well as the linear absorption spectra, of an array of complex materials, including halide perovskites, important optoelectronic materials for which standard GW and GW-BSE methods have been demonstrated to be inadequate or inconsistent.The progress made thus far with OTRSH for band structure and optical spectra of solids is encouraging, as it suggests that it is in general possible to fix one parameter to the orientationally-averaged dielectric constant and then tune a single parameter – the range-separation parameter – to yield band structures and optical spectra in agreement with GW-BSE, partially or fully mitigating computational costs associated with MBPT. Given the quality of the OTRSH band structure with a single parameter, could OTRSH be an optimal starting point for GW and GW-BSE calculations of complex structurally and chemically heterogeneous systems? Additionally, it remains to be seen whether one can set this single parameter independent of GW-BSE calculations and experiment. Can it be predicted from materials properties in a computationally tractable manner, for isotropic and anisotropic materials alike? The aim of this project is to address these questions, and ultimately develop efficient approaches for understanding existing and predicting new excited-state phenomena in complex materials.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.