Nonlocal Density Functional Theory of Molecules and Solids

宾夕法尼亚大学的陶建民得到了化学系化学理论、模型和计算方法计划的支持，以开发改进的密度泛函理论(DFT)方法。在电子材料、器件以及化学和物理过程的计算机辅助设计中，电子性质的准确预测起着举足轻重的作用。虽然许多基于波函数的方法具有很高的精度，但它们的计算成本太高，不适用于实际应用。对于大型多原子系统，密度泛函理论是唯一具有所需精度和效率的方法，因此它是应用最广泛的电子结构方法。然而，常用的密度泛函不能达到普遍的精度。陶和同事们专注于开发具有更广泛适用性的函数。这些泛函被用来研究不同体系的性质，如(水)簇合物、发光低聚物和聚合物。这项工作既涉及方法的发展，也涉及到数值应用。这个研究项目的目标是开发和应用非局域密度泛函(DFT)，包括一种新的用于远程van der Waals相互作用的非局域关联能量泛函。这项工作的重点是解决或缓解常用密度泛函遇到的一些长期存在的困难。密度泛函方法的一个主要突出问题是，密度泛函对一类或多类体系可能是精确的，但不是对所有系统都是准确的，或者对一种或多种性质是精确的，但不是对所有性质都是精确的。该项目的一个目标是开发具有普遍精度和更广泛适用性的密度泛函，以便能够在相同的基础上以相同的密度泛函解决不同的化学和物理问题。这样的密度泛函必须是非局域的，因为库仑相互作用的性质是非局域的。本科生、研究生和博士后都参与了这个项目的各个方面。

Jianmin Tao of the University of Pennsylvania is supported by the Chemical Theory, Models and Computational Methods program in the Chemistry Division to develop improved methods for density functional theory (DFT). In the computer-aided design of electronic materials, devices, as well as chemical and physical processes, accurate prediction of electronic properties plays a decisive role. While many wave function-based methods are highly accurate, their computational costs are too expensive to be generally useful for practical applications. For large many-atom systems, DFT is the only method with both the required accuracy and efficiency and for that reason it is the most widely used electronic structure method. However, commonly-used density functionals are unable to achieve universal accuracy. Tao and coworkers focus on developing functionals with a wider range of applicability. These functionals are being used to study the properties of diverse systems such as (water) clusters and light-emitting oligomers and polymers. The work involves both method development and numerical application. The project also includes summer outreach programs aimed primarily at undergraduates.The goal of this research project is the development and application of nonlocal density functionals (DFT), including a new nonlocal correlation energy functional for long-range van der Waals interactions. The focus of this effort is on solving or mitigating some long-standing difficulties that commonly-used density functionals encounter. A major outstanding problem for DFT methodology is that a density functional may be accurate for one or more classes of systems, but not for all systems, or for one or more properties, but not for all properties. A goal of this project is to develop a density functional with universal accuracy and wider applicability so that diverse chemical and physical problems can be solved on the same footing, with the same density functional. Such a density functional must be nonlocal, because the nature of the Coulomb interaction is nonlocal. Undergraduates, graduate students and postdocs are involved in various aspects of this project.