周期性自举嵌入
Periodic Bootstrap Embedding
摘要:
自举嵌入(BE)是近年来发展起来的一种电子结构方法,它在处理分子中的电子关联方面取得了巨大的成功。在这里,我们扩展BE治疗表面和固体的波函数表示在周期性的边界条件下使用倒易空间总和(即$k$点采样)。这种方法的主要好处是,所得到的片段哈密顿携带没有明确的依赖于倒易空间和,允许一个应用传统的非周期性电子结构代码的片段,即使整个系统需要仔细考虑的周期性边界条件。使用耦合簇单双(CCSD)作为一个例子的方法来解决的片段哈密顿,我们提出了最小基组CCSD-在HF的结果在一维导电聚合物。我们发现,周期性的BE-CCSD通常可以恢复$\sim$ 99.9%的电子相关能。我们进一步证明,即使对于有机太阳能电池感兴趣的复杂供体-受体聚合物,周期性BE-CCSD也是可行的-尽管单体足够大,即使是$Γ-$点周期性CCSD计算也是禁止的。我们的结论是,BE是一个很有前途的新工具,应用分子电子结构工具的固体和界面。
英文摘要:
Bootstrap embedding (BE) is a recently developed electronic structure method that has shown great success at treating electron correlation in molecules. Here, we extend BE to treat surfaces and solids where the wave function is represented in periodic boundary conditions using reciprocal space sums (i.e. $k$-point sampling). The major benefit of this approach is that the resulting fragment Hamiltonians carry no explicit dependence on the reciprocal space sums, allowing one to apply traditional non-periodic electronic structure codes to the fragments even though the entire system requires careful consideration of periodic boundary conditions. Using coupled cluster singles and doubles (CCSD) as an example method to solve the fragment Hamiltonians, we present minimal basis set CCSD-in-HF results on 1D conducting polymers. We show that periodic BE-CCSD can typically recover $\sim$99.9% of the electron correlation energy. We further demonstrate that periodic BE-CCSD is feasible even for complex donor-acceptor polymers of interest to organic solar cells - despite the fact that the monomers are sufficiently large that even a $Γ-$point periodic CCSD calculation is prohibitive. We conclude that BE is a promising new tool for applying molecular electronic structure tools to solids and interfaces.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。