Heßelmann Andreas
Lehrstuhl für Theoretische Chemie, Universität Erlangen-Nürnberg , Egerlandstr. 3, D-91058 Erlangen, Germany.
J Chem Theory Comput. 2015 Apr 14;11(4):1607-20. doi: 10.1021/acs.jctc.5b00024. Epub 2015 Mar 27.
Molecular excitation energies have been calculated with time-dependent density-functional theory (TDDFT) using random-phase approximation Hessians augmented with exact exchange contributions in various orders. It has been observed that this approach yields fairly accurate local valence excitations if combined with accurate asymptotically corrected exchange-correlation potentials used in the ground-state Kohn-Sham calculations. The inclusion of long-range particle-particle with hole-hole interactions in the kernel leads to errors of 0.14 eV only for the lowest excitations of a selection of three alkene, three carbonyl, and five azabenzene molecules, thus surpassing the accuracy of a number of common TDDFT and even some wave function correlation methods. In the case of long-range charge-transfer excitations, the method typically underestimates accurate reference excitation energies by 8% on average, which is better than with standard hybrid-GGA functionals but worse compared to range-separated functional approximations.
利用含时密度泛函理论(TDDFT),采用随机相位近似海森矩阵并结合不同阶次的精确交换贡献,计算了分子激发能。据观察,如果将这种方法与基态Kohn-Sham计算中使用的精确渐近修正交换相关势相结合,就能得到相当准确的局域价激发。在内核中包含长程粒子-粒子与空穴-空穴相互作用,对于一系列三个烯烃、三个羰基和五个氮杂苯分子的最低激发,仅导致0.14 eV的误差,从而超过了许多常见TDDFT甚至一些波函数相关方法的精度。对于长程电荷转移激发,该方法通常平均低估准确的参考激发能8%,这比标准杂化广义梯度近似泛函要好,但与范围分离泛函近似相比要差。