CAREER: Electron correlation and optical spectra with a nonlocal energy-optimized (NEO) kernel

该职业奖支持旨在改进分子和材料计算特性的理论和计算方法的理论和计算研究和教育。PI的目标是开发一种计算效率高且有用的精确校正，以提出基于密度泛函理论从组成原子开始计算材料性质的计算方法。这种方法在量子力学描述许多相互作用的粒子的另一种理论公式中提出了一种近似值，这在计算上是非常昂贵的。这种近似称为随机相位近似。与许多现有的方法相比，PI的方法应该具有更高的精度和计算成本低廉。它还应该使系统的计算精度更高，其中电子的配置导致比系统的最低或基态能量更高的能量。太阳能电池的运行提供了本研究将探讨的例子。PI将通过计算材料特性，并与其他理论和计算方法以及实验的结果进行比较来验证她的方法。该奖项支持针对研究生、本科生和高中生的教育活动。这一活动的一个重点与培训高中教师的TUTeach项目有关。这个项目的学生是物理专业的学生，他们将作为PI的学习助理，了解PI的研究。这些学生也是高中的外联大使。TUTeach的学生将与PI合作，为高中生提供计算物理讲座。PI还将在当地教务长的本科生研究经验项目中指导学生。PI教育计划的另一个重要组成部分是开发一门新的基于主动学习的计算材料物理课程。该职业奖支持理论和计算研究和教育，以开发更准确的基于密度泛函理论的基态能量和材料激发态特性计算。通过这个项目，PI有两个目标：(1)开发一种对随机相位近似的基态相关能的计算效率校正。如果没有修正，随机相近似将不够精确，无法成为化学和材料科学的基准。目前存在的一些修正是准确的，但代价是高昂的计算成本。(2)使计算效率高的时变密度泛函理论与更复杂的量子多体技术在精度上具有竞争力。时变密度泛函理论是密度泛函理论对时变势的推广。时变密度泛函理论已经成为计算激发态能的一种非常流行的工具。该方法计算效率高，但与基于量子多体格林函数的方法相比，精度较低。最常用的绝热局部密度核没有激子效应。PI将进一步开发非局部能量优化核模型，该模型计算效率高，可以捕获这些效应。对于基态和激发态，在基态的频率相关线性响应函数的库仑电子-电子相互作用核中加入相同的非局域交换相关核。非局域能量优化核将应用于分子和材料。核校正将被测试并应用于许多基态性质，如原子化能、内聚能、结构相变、团簇和吸附问题。时间依赖密度泛函理论的核校正将应用于激发能，包括半导体中的激发能，特别关注太阳能电池应用的光伏材料。该奖项支持针对研究生、本科生和高中生的教育活动。这一活动的一个重点与培训高中教师的TUTeach项目有关。这个项目的学生是物理专业的学生，他们将作为PI的学习助理，了解PI的研究。这些学生也是高中的外联大使。TUTeach的学生将与PI合作，为高中生提供计算物理讲座。PI还将在当地教务长的本科生研究经验项目中指导学生。PI教育计划的另一个重要组成部分是开发一门新的基于主动学习的计算材料物理课程。

NONTECHNICAL SUMMARY This CAREER award supports theoretical and computational research and education aimed to improve theoretical and computational methods for computing properties of molecules and materials.The PI aims to develop a computationally efficient and usefully accurate correction to present computational methods of calculating the properties of materials starting from constituent atoms based on the density-functional-theory. The approach appeals to an approximation within another theoretical formulation of the quantum mechanical description of many interacting particles that would be computationally very expensive to evaluate. The approximation is known as the random phase approximation. The PI's method should have a higher accuracy and be computationally inexpensive compared to many currently existing methods. It should also enable higher accuracy calculations of systems in which the configuration of electrons leads to a higher energy than the lowest or ground state energy of the system. The operation of a solar cell provides examples which the research will explore. The PI will test her method through calculations of materials properties and comparisons with results from other theoretical and computational methods, and experiments. .This award supports an educational activity focused on graduate, undergraduate and high-school students. A thrust of this activity is linked to the TUTeach program which trains high-school teachers. The students in this program are physics majors, who will work as learning assistants for the PI and learn about the PI's research. These students are also outreach ambassadors in the high schools. TUTeach students will work with the PI to deliver computational physics lectures to high-school students. The PI will also mentor students within the local provost's research-experiences-for-undergraduates program. Another significant component of the PI's education program is to develop a new active-learning-based Computational Materials Physics course.TECHNICAL SUMMARYThis CAREER award supports theoretical and computational research and education to develop more accurate density functional theory-based calculations of ground-state energies and excited state properties of materials. The PI has two goals through this project: (1) To develop a computationally efficient correction to the ground-state correlation energy of the random phase approximation. Without a correction, the random phase approximation would not be accurate enough to become a benchmark in chemistry and materials science. Some of the currently existing corrections are accurate, but the price is a high computational cost.(2) To make the computationally efficient time-dependent density functional theory competitive in accuracy with more complex quantum many-body techniques. Time-dependent density functional theory is an extension of density functional theory to time-dependent potentials. Time-dependent density functional theory has become a very popular tool to compute excitation energies. It is computationally efficient, but its accuracy is moderate compared to quantum many-body Green's function-based methods. The most commonly used adiabatic local density kernel misses excitonic effects. The PI will further develop a nonlocal energy-optimized kernel model that is computationally efficient and can capture these effects. Both for ground and excited states the same nonlocal exchange-correlation kernel is to be added to the Coulomb electron-electron interaction kernel in the frequency-dependent linear response function of the ground state. The nonlocal energy-optimized kernel will be applied to molecules and materials. The kernel correction will be tested and applied to many ground-state properties, such as atomization energies, cohesive energies, structural phase transitions, clusters and adsorption problems. The kernel correction to time-dependent density functional theory will be applied to excitation energies including those in semiconductors with a particular focus on photovoltaic materials for solar cell applications.This award supports an educational activity focused on graduate, undergraduate and high-school students. A thrust of this activity is linked to the TUTeach program which trains high-school teachers. The students in this program are physics majors, who will work as learning assistants for the PI and learn about the PI's research. These students are also outreach ambassadors in the high schools. TUTeach students will work with the PI to deliver computational physics lectures to high-school students. The PI will also mentor students within the local provost's research-experiences-for-undergraduates program. Another significant component of the PI's education program is to develop a new active-learning-based Computational Materials Physics course.