[CP2K-user] [CP2K:18485] Re: Something wrong with TDDFT by CP2K version 9.1 using PBE0+HSE

['Frederick Stein' via cp2k]

Dear Janson,

CP2K does not provide analytical 2nd derivatives for all of its 
self-implemented functionals. This is especially true with the ADMM 
exchange correction. You have two options:
1. Add the keyword 2ND_DERIV_ANALYTICAL 
<https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/DFT/XC.html#2ND_DERIV_ANALYTICAL> 
.FALSE. within the XC section (turn off analytical 2nd derivatives). Then, 
the ADMM exchange correction resembles the HF section as closely as 
possible.
2. Set the value of the keyword EXCH_CORRECTION_FUNC 
<https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/DFT/AUXILIARY_DENSITY_MATRIX_METHOD.html#list_EXCH_CORRECTION_FUNC> 
to PBEX, PBEX_LIBXC, BECKE88X_LIBXC, OPTX_LIBXC or LDA_X_LIBXC to switch to 
different exchange correction functionals. For these corrections, we have 
2nd derivatives available but we have never really tested the quality of 
different ADMM corrections with different hybrid functionals, but 
PBEX/PBEX_LIBXC should provide decent results.

HTH
Frederick

Janson Shi schrieb am Freitag, 24. Februar 2023 um 10:23:24 UTC+1:

> Dear sir,
>
> Something is wrong when I run cp2k to calculate the TDDFT as shown in the 
> picture, and the corresponding files have been attached. Please give me 
> some instructions to solve this. Many thanks for the contributions to CP2K 
> and your help all the way.
>
> Best regards,
> Janson Shi
> [image: 1677230258487.jpg]
>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/4b11ddba-5d39-407e-b014-8b890169a52en%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20230224/f30e0871/attachment.htm>