Fukui Institute for Fundamental Chemistry, Kyoto University, 34-4 Takano Nishihiraki-cho, Sakyo-ku, Kyoto 606-8103, Japan.
Research Center for Computational Design of Advanced Functional Materials (CD-FMat), National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, Japan.
J Chem Phys. 2018 Feb 14;148(6):064115. doi: 10.1063/1.5012935.
The exactly analytic gradient is derived and implemented for the fragment molecular orbital (FMO) method combined with density-functional tight-binding (DFTB) using adaptive frozen orbitals. The response contributions which arise from freezing detached molecular orbitals on the border between fragments are computed by solving Z-vector equations. The accuracy of the energy, its gradient, and optimized structures is verified on a set of representative inorganic materials and polypeptides. FMO-DFTB is applied to optimize the structure of a silicon nano-wire, and the results are compared to those of density functional theory and experiment. FMO accelerates the DFTB calculation of a boron nitride nano-ring with 7872 atoms by a factor of 406. Molecular dynamics simulations using FMO-DFTB applied to a 10.7 μm chain of boron nitride nano-rings, consisting of about 1.2 × 10 atoms, reveal the rippling and twisting of nano-rings at room temperature.
针对采用自适应冻结轨道的片段分子轨道（FMO）方法与密度泛函赝势（DFTB）相结合的情况，推导出并实现了精确的解析梯度。通过求解 Z 向量方程，计算了在片段边界上冻结离域分子轨道所产生的响应贡献。在一组具有代表性的无机材料和多肽上验证了能量、梯度和优化结构的准确性。将 FMO-DFTB 应用于优化硅纳米线的结构，并将结果与密度泛函理论和实验进行比较。FMO 将含有 7872 个原子的氮化硼纳米环的 DFTB 计算加速了 406 倍。使用 FMO-DFTB 对由约 1.2×10 个原子组成的 10.7 μm 长的氮化硼纳米环链进行分子动力学模拟，揭示了纳米环在室温下的起伏和扭曲。
