Narrowing the error in electron correlation calculations by basis set re-hierarchization and use of the unified singlet and triplet electron-pair extrapolation scheme: application to a test set of 106 systems.
Narrowing the error in electron correlation calculations by basis set re-hierarchization and use of the unified singlet and triplet electron-pair extrapolation scheme: application to a test set of 106 systems.
复制标题
DOI：
10.1063/1.4903193
复制
发表时间：
2014-12
期刊：
影响因子：
--
通讯作者：
A. Varandas;A. Varandas;F. Pansini
中科院分区：
文献类型：
--
作者：
A. Varandas;A. Varandas;F. Pansini
之前建议的一种方法是通过重新分配基础分层数并使用统一的单重态和三重态对外推方案来计算完整单电子基础集极限下的相关能，该方法应用于 106 个系统的测试集，其中一些系统多达 48 个电子。该方法用于根据二阶 Møller-Plesset 微扰理论和耦合簇单双激发方法计算的原始值获得外推相关能，其中一些方法还进行了微扰三重校正。计算出的相关能量也已用于预测加法方案中的原子化能量。即使当使用 (d, t) 分层数对来执行外推时，也能与最佳可用估计获得良好的一致性。可以想象，这证明没有充分的理由在外推中排除双 zeta 能量，特别是如果基础经过校准以符合理论模型的话。
A method previously suggested to calculate the correlation energy at the complete one-electron basis set limit by reassignment of the basis hierarchical numbers and use of the unified singlet- and triplet-pair extrapolation scheme is applied to a test set of 106 systems, some with up to 48 electrons. The approach is utilized to obtain extrapolated correlation energies from raw values calculated with second-order Møller-Plesset perturbation theory and the coupled-cluster singles and doubles excitations method, some of the latter also with the perturbative triples corrections. The calculated correlation energies have also been used to predict atomization energies within an additive scheme. Good agreement is obtained with the best available estimates even when the (d, t) pair of hierarchical numbers is utilized to perform the extrapolations. This conceivably justifies that there is no strong reason to exclude double-zeta energies in extrapolations, especially if the basis is calibrated to comply with the theoretical model.