一篇关于DFT的tutorial review Int. J. Quantum Chem. 2020; e26332.

yflchx

P. Morgante, R. Peverati, The devil in the details: A tutorial review on some undervalued aspects of density functional theory calculations. Int. J. Quantum Chem. 2020; e26332.

https://doi.org/10.1002/qua.26332

一篇关于DFT的tutorial review。

ymygca

qua.26332.pdf (2.37 MB, 下载次数 Times of downloads: 54)

2020-6-29 03:01 上传 Uploaded

点击下载
Click to download

sobereva

看了一遍，感觉此文中有价值的信息非常有限（也就是图1可以以后用在别的地方），在SI里对计算耗时的对比非常有误导性，居然很多例子里B3LYP耗时比M06-2X还高。整篇文章基本只涉及热化学数据计算，而激发态、响应属性等等一概没提，相对来说这个标题起得太大了。看完后的感觉是，内行人看了有不少槽想吐，而初学者看了，容易被SI里的信息带偏，还是没法正确认识该怎么计算、怎么合理选择计算级别。

yflchx

sobereva 发表于 2020-6-29 03:30
看了一遍，感觉此文中有价值的信息非常有限（也就是图1可以以后用在别的地方），在SI里对计算耗时的对比非 ...

怎么计算，怎么合理选择计算级别：
作者的说法：
“• Always follow recommendations from experts in the field when choosing any combination of functional and basis sets.
• Avoid picking a functional/basis set only because they are popular.
• The best way to pick a functional is to compare the chemical problem under consideration with similar problems collected in databases of high-level computational data. Databases can also give an indication on the expected accuracy of the results (i.e., estimate the error bars associated with a calculation).
• Basis sets are responsible for two types of errors—BSIE and BSSE. Both errors can be alleviated by choosing a suitably large basis set and, in some specific situations, by using appropriate techniques (CP corrections).
• Do not use B3LYP/6-31G* blindly. It cannot be universally trusted.”

yflchx

补一篇参照：

Lars Goerigk and Nisha Mehta, A Trip to the Density Functional Theory Zoo: Warnings and Recommendations for the User.

Australian Journal of Chemistry 72(8), 563-573.  https://doi.org/10.1071/CH19023

naonao5205

Lars Goerigk and Nisha Mehta, A Trip to the Density Functional Theory Zoo: Warnings and Recommendations for the User.

sobereva

游客，本帖隐藏的内容需要积分高于 25 才可浏览，您当前积分为 0

yflchx

sobereva 发表于 2020-6-29 21:07
**** 本内容被作者隐藏 ****

特定的人有特定的倾向，批判着看。

两厢对比就能感觉到倾向性，这也是我发这个参照文献的原因。

对任何形式的泛函推荐已经不太关注，更关心所谓高精度泛函的缺陷和反例。