求助关于TDDFT和UDFT选择计算三重激发态的文献

Flower

大家好，我是计算磷光的，所以需要计算最低三重激发态。我看到文献（Inorganic chemistry, 2014, 53(20): 11015-11019）上面为了得到精确地三重激发态，先使用TDDFT计算了前8个态，然后再使用UDFT算了上一步优化好的8个态，根据能量找到最低的三重激发态。我计算三重激发态是模仿这一篇。我自己也不是很懂。审稿人给我提了意见：the triplet excited states were optimizedby TD-B3LYP and UB3LYP, and the later was chosen as the final structure. Whatwas the basis for selecting? 我自己不太会解决。导师让我找关于UDFT和TDDFT对于优化三重激发态的优点或是不足（我理论知识特别弱，对于我来说有点大海捞针，回稿周期有点短）。所以希望各位老师能帮一下我，谢谢大家了。

liyuanhe211

如果分子的基态是单重态，仅对T1来说，UDFT比TDDFT计算快、且考虑了轨道的弛豫，但对其他的态没有道理：用UDFT算激发态（△SCF）只能算“某多重度下的基态”，如UDFT的SCF稳定之后算出来的肯定是“三重态中的基态” T1，怎么用UDFT算“前八个态”？

小范范1989

师妹，我对磷光不太懂哈。那你选择这泛函的依据是什么？不是对比波长或者什么？
如果是对比波长，可以说UDFT计算的波长和实验符合的好，所以我们后续的计算都采用的UDFT.
我的理解哈，不一定对。

sobereva

没彻底搞懂文献前绝对不要盲目效仿文献，否则到时候自己的文章被审稿人问及都不好给予解释

UDFT只能算T1，除非几个三重态由于不可约表示不同能用UDFT算不同的态，这种情况对大体系几乎碰不上

同时用TDDFT和UDFT来优化是很不合适的，明显后者更快原理更严格，用前者是完全多余的。

Flower

谢谢各位老师的回复。之前我其实也是不理解为什么要先TDDFT再UDFT，一直困惑着我，这次审稿人的问题也让我不知所措。我参考的文献里面是这样解释的，希望老师们能帮我看一下，看看他说的合理吗？我其实不是很理解。辛苦老师们了。To get insight into the emissive processes in complex 1 we explored its lowest triplet excited state potential energy surfaces (PES). Toward this aim, we determined the minimum geometries of the first eight triplet excited states at the TD-B3LYP level of theory. These TD-DFT optimized geometries served as an initial guess for the final UDFT (B3LYP/6-31G*) optimizations. For these calculations UDFT is preferred over TD-DFT, due to the problems of the latter to provide a balanced description of excited states of different character (especially if charge transfer states are involved).31 With this protocol we were able to obtain the optimized geometries of four different triplet excited states (see below), whereas the optimizations for the other higher-lying triplet states failed due to variational collapse of the DFT calculation.

冰释之川

对于用UDFT和TDDFT两种方法进行的T1结构优化，“ due to the problems of the latter to provide a balanced description of excited states of different character ”这么回答似乎有点夸大TDDFT优化结果的不可靠性了。其实这两种方法都可以优化T1结构，只不过UDFT更加好一点，不能说用TDDFT优化T1就存在“problem”，只不过TDDFT没考虑轨道弛豫会带来一些偏差（一般用这两种方法优化出的T1极小点偏差不会很大），而且用TDDFT优化会比较time consuming。

针对T1结构优化，基本上都采用UDFT方法，你可以这么说：“We eventually adpoted UDFT method to optimize T1 structures since it includes orbital relaxation effect during optimization and meanwhile it is computationally efficient. However,  this effect is not considered in the TDDFT scheme, which might bring forth a little deviation. Note that the TDDFT scheme is still acceptable in most cases especially when the UDFT method is inaccessible such as the optimization of high-lying triplet states”

Flower

谢谢冰释之川老师。您英语真好，随手就来。以后向您好好学习。如果您有相关文献，也很希望您能给我推荐哦。没有的话，我再好好找找。辛苦了。

yezhonghua

小范范1989 发表于 2017-5-18 19:30
师妹，我对磷光不太懂哈。那你选择这泛函的依据是什么？不是对比波长或者什么？
如果是对比波长，可以说UD ...

貌似UDFT和TD-DFT相同的泛函和基组下，UDFT对比TD-DFT会更加接近实验光谱么？同时两者的大小也不是绝对的吧?谢谢