Orbital-Corrected Orbital-Free Density Functional Theory

理论化学的梦想是从第一性原理高度准确地预测凝聚相现象。目前的理论方法要么不能以高精度一致地处理化学反应，要么受到系统规模或过程时间尺度的限制。传统的固体系统理论处理方法是标准的平面波周期密度泛函理论(DFT)，但受计算成本和精度水平的限制。为了实现更好的缩放和精度，我们在很大程度上依赖于DFT，并改进了现有的DFT实现。求解DFT问题的两种方法，即传统的基于轨道的Kohn-Sham(KS)方法和无轨道方法(OF)，各有优缺点。最近，我们发展了一种新的实现DFT的方法，即轨道修正的OF-DFT(OO-DFT)，它结合了OF-DFT和KS-DFT的优点并避免了缺点，允许以比传统的自洽KS-DFT方法低得多的成本研究不同化学键环境中的体系。对于立方钻石硅和面心立方银系统，OO-DFT的精度与完全自洽的KS-DFT相当，最多只需两次非自洽迭代。在本次资助期间，我们将进一步发展OO-DFT，使OO-DFT通过使用现有的线性标度KS-DFT算法实现线性标度，并成为处理不同化学键环境中数千个原子的大系统的强大工具，比其他现有的线性标度DFT方法更有效。我们将把OO-DFT应用到从头算分子动力学(AIMD)中，以便一个高效和精确的基于OO-DFT的AIMD方法将被整个理论/计算化学界广泛使用。最终，理论技术将成熟，为科学界提供一个可靠的工具，以定性和定量地提供对复杂系统中基本机制的微观理解，并获得对自然的一些方面的洞察，这些方面不容易通过实验手段来探索。

The dream of theoretical chemistry is highly accurate predictions of condensed-phase phenomena from first principles. Current theoretical methods either cannot consistently treat chemical reactions at a high level of accuracy, or are limited by system size or the time scale of the process. The conventional theoretical treatment of solid-state systems is the standard plane-wave periodic density-functional theory (DFT), but is limited by the scaling of computational cost and the level of accuracy. To achieve better scaling and accuracy, we rely heavily on DFT and improve the existing implementations of DFT. Each of the two ways of solving the DFT problem, i.e., the traditional orbital-based Kohn-Sham (KS) and the orbital-free (OF) schemes, has its own strengths and weaknesses. Recently, we have developed a new implementation of DFT, namely orbital-corrected OF-DFT (OO-DFT), which coalesces the advantages and avoids the drawbacks of OF-DFT and KS-DFT and allows systems within different chemical bonding environments to be studied at a much lower cost than the traditional self-consistent KS-DFT method. For the cubic-diamond Si and the face-centered-cubic Ag systems, OO-DFT accomplishes accuracy comparable to fully self-consistent KS-DFT with at most two non-self-consistent iterations. In this grant period, we will further develop OO-DFT, so that OO-DFT achieves linear scaling by employing currently available linear-scaling KS-DFT algorithms and becomes a powerful tool to treat large systems of thousands of atoms within different chemical bonding environments much more efficiently than other currently available linear-scaling DFT methods. We will implement OO-DFT into ab initio molecular dynamics (AIMD) so that a highly efficient and accurate OO-DFT-based AIMD method will be available for general usage by the entire theoretical/computational chemistry community. In the end, the theoretical techniques will be matured to offer the scientific community a reliable vehicle to provide qualitatively and quantitatively a microscopic understanding of basic mechanisms in complex systems, and to gain insights into aspects of nature that cannot be easily probed by experimental means.