密度泛函理论
轨道自由密度泛函理论
物理
离散化
电子结构
周期边界条件
有限元法
能量最小化
空位缺陷
边值问题
空格(标点符号)
原子轨道
局部密度近似
统计物理学
电子
量子力学
数学分析
凝聚态物理
热力学
数学
计算机科学
操作系统
作者
Sambit Das,Mrinal Iyer,Vikram Gavini
标识
DOI:10.1103/physrevb.92.014104
摘要
We propose a local real-space formulation for orbital-free density functional theory (DFT) with density dependent kinetic energy functionals and a unified variational framework for computing the configurational forces associated with geometry optimization of both internal atomic positions as well as the cell geometry. The proposed real-space formulation, which involves a reformulation of the extended interactions in electrostatic and kinetic energy functionals as local variational problems in auxiliary potential fields, also readily extends to all-electron orbital-free DFT calculations that are employed in warm dense matter calculations. We use the local real-space formulation in conjunction with higher-order finite-element discretization to demonstrate the accuracy of orbital-free DFT and the proposed formalism for the Al-Mg materials system, where we obtain good agreement with Kohn-Sham DFT calculations on a wide range of properties and benchmark calculations. Finally, we investigate the cell-size effects in the electronic structure of point defects, in particular, a monovacancy in Al. We unambiguously demonstrate that the cell-size effects observed from vacancy formation energies computed using periodic boundary conditions underestimate the extent of the electronic structure perturbations created by the defect. On the contrary, the bulk Dirichlet boundary conditions, accessible only through the proposed real-space formulation, which correspond to an isolated defect embedded in the bulk, show cell-size effects in the defect formation energy that are commensurate with the perturbations in the electronic structure. Our studies suggest that even for a simple defect like a vacancy in Al, we require cell sizes of $\ensuremath{\sim}{10}^{3}$ atoms for convergence in the electronic structure.
科研通智能强力驱动
Strongly Powered by AbleSci AI