Center for Computational Chemistry, University of Georgia, Athens, Georgia 30602, USA.
J Chem Phys. 2009 Dec 28;131(24):244116. doi: 10.1063/1.3269031.
Explicitly correlated MP2-R12 and coupled cluster R12 methods have proven to be effective in achieving the basis set limit of correlated wave function methods. However, correlated methods for high-spin open-shell states are typically based on semicanonical orbitals, leading to an unrestricted formalism, which for double excitations requires three independent sets of amplitudes. In contrast, Z-averaged perturbation theory redefines the Hamiltonian with a symmetric exchange operator, thereby allowing a spin-restricted formulation with equivalent alpha and beta subspaces. In the current work, we present a preliminary study of explicitly correlated ZAPT for second-order perturbation theory. The superior basis set convergence of R12 methods is demonstrated for a set of atomization energies, showing the R12 results to be competitive with common basis set extrapolation techniques, albeit at a fraction of the cost. Given the efficiency gains associated with the symmetric exchange operator, we suggest ZAPT as a candidate for reducing the cost of current open-shell MP2-R12 and CCSD(T)-R12 computations.
显式相关的 MP2-R12 和耦合簇 R12 方法已被证明在实现相关波函数方法的基组极限方面非常有效。然而,高自旋开壳层状态的相关方法通常基于半经典轨道,导致无限制的形式主义,对于双激发,需要三个独立的振幅集。相比之下,Z 平均微扰理论通过对称交换算子重新定义哈密顿量,从而允许具有等效的α和β子空间的自旋限制公式。在当前的工作中,我们对二阶微扰理论的显式相关 ZAPT 进行了初步研究。一系列原子化能表明 R12 方法在基组收敛性方面具有优势,表明 R12 结果与常见的基组外推技术具有竞争力,尽管成本只是其中的一部分。鉴于对称交换算子带来的效率提高,我们建议将 ZAPT 作为降低当前开壳层 MP2-R12 和 CCSD(T)-R12 计算成本的候选方法。