Enhanced computational efficiency in the direct determination of the two-electron reduced density matrix from the anti-Hermitian contracted Schrödinger equation with application to ground and excited states of conjugated π-systems.
Enhanced computational efficiency in the direct determination of the two-electron reduced density matrix from the anti-Hermitian contracted Schrödinger equation with application to ground and excited states of conjugated π-systems.
复制标题
通过应用于共轭 π 系统的基态和激发态，提高了从反厄米收缩薛定谔方程直接确定双电子约化密度矩阵的计算效率。
DOI：
--
复制
发表时间：
2015
影响因子：
4.4
通讯作者：
D. Mazziotti
中科院分区：
文献类型：
--
作者：
A. Sand;D. Mazziotti
从反厄米收缩薛定谔方程（ACSE）的解中确定双电子约化密度矩阵（2-RDM），可以得到基态和激发态的准确能量和性质。在这里，我们开发了一种更有效的方法来解决ACSE，使用二阶信息来选择一个更优化的步骤，以解决方案。对水、氟化氢和共轭π体系基态和激发态的计算表明，改进的ACSE算法比原算法快10-20倍. ACSE可以处理单参考和多参考电子关联与初始2-RDM从一个完整的活动空间自洽场（CASSCF）计算。利用改进的算法，我们探讨了CASSCF计算中活性空间的截断与ACSE计算的能量和2-RDM精度之间的关系。ACSE的准确性，我们发现，是不太敏感的活动空间的大小比其他波函数的方法，这是有用的，当大的活动空间计算是计算上不可行的准确性。
Determination of the two-electron reduced density matrix (2-RDM) from the solution of the anti-Hermitian contracted Schrödinger equation (ACSE) yields accurate energies and properties for both ground and excited states. Here, we develop a more efficient method to solving the ACSE that uses second-order information to select a more optimal step towards the solution. Calculations on the ground and excited states of water, hydrogen fluoride, and conjugated π systems show that the improved ACSE algorithm is 10-20 times faster than the previous ACSE algorithm. The ACSE can treat both single- and multi-reference electron correlation with the initial 2-RDM from a complete-active-space self-consistent-field (CASSCF) calculation. Using the improved algorithm, we explore the relationship between truncation of the active space in the CASSCF calculation and the accuracy of the energy and 2-RDM from the ACSE calculation. The accuracy of the ACSE, we find, is less sensitive to the size of the active space than the accuracy of other wavefunction methods, which is useful when large active space calculations are computationally infeasible.