Angeli Celestino
Dipartimento di Chimica, Universitá di Ferrara, Via Borsari 46, I-44100 Ferrara, Italy.
J Comput Chem. 2009 Jun;30(8):1319-33. doi: 10.1002/jcc.21155.
This article addresses an analysis of the physical effects required for the correct description of the ionic pi --> pi* excited states in the frame of ab initio quantum chemistry, using the ionic V state of the ethene molecule as an example. The importance of the dynamic sigma polarization (absent in methods where the sigma skeleton is treated at a mean-field level) has been recognized by many authors in the past. In this article a new physical effect is described, i.e. the spatial contraction of the pi and pi* molecular orbitals (or of the local p atomic orbitals) originated from the reduction of the ionicity due to the dynamic sigma polarization. Such an effect is a second-order effect (it appears only as a consequence of the dynamic sigma polarization) but it cannot be ignored. Many of the difficulties found in the past in the calculation of the vertical excitation energy of the ionic states are attributed to an incomplete description of this contraction, while the few successes have been obtained when it has been fortuitously introduced by ad hoc procedures or when it is described in a brute force approach. Various strategies are proposed to allow for the spatial contraction of the p atomic orbitals. If this effect is considered at the orbital optimization step, it is shown that for the V state of ethene no Rydberg/valence mixing occurs and a simple perturbation correction (to the second order in the energy) on the pi --> pi* singly excited configuration gives stable results with respect to the computational parameters and in good agreement with the experimental findings and with the best theoretical calculations. Moreover, our results confirm the indication of Müller et al. (J Chem Phys 1999, 110, 7176) that the transition to the V state of ethene conforms to the Franck-Condon principle and that it is not necessary to appeal to a nonvertical transition to interpret the experimental data. The strategy reported in this article for ethene can be in principle generalized to the pi --> pi* ionic excited states of other molecules.
本文以乙烯分子的离子V态为例，探讨了在从头算量子化学框架内正确描述离子π→π激发态所需的物理效应分析。过去许多作者已经认识到动态σ极化的重要性（在将σ骨架处理为平均场水平的方法中不存在）。本文描述了一种新的物理效应，即由于动态σ极化导致离子性降低，从而使π和π分子轨道（或局部p原子轨道）发生空间收缩。这种效应是二阶效应（仅作为动态σ极化的结果出现），但不能忽略。过去在计算离子态垂直激发能时遇到的许多困难都归因于对这种收缩的描述不完整，而少数成功案例是在通过特殊程序偶然引入或用蛮力方法描述时获得的。提出了各种策略来考虑p原子轨道的空间收缩。如果在轨道优化步骤中考虑这种效应，结果表明对于乙烯的V态，不存在里德堡/价态混合，并且对π→π单激发组态进行简单的微扰校正（到能量的二阶），相对于计算参数能给出稳定结果，并且与实验结果以及最佳理论计算结果吻合良好。此外，我们的结果证实了Müller等人（《化学物理杂志》1999年，110卷，7176页）的观点，即乙烯向V态的跃迁符合弗兰克 - 康登原理，并且无需诉诸非垂直跃迁来解释实验数据。本文针对乙烯报道的策略原则上可以推广到其他分子的π→π离子激发态。
