[CP2K-user] [CP2K:21530] Re: XAS Calculation and NTO Analysis

[Augustin Bussy]

Hello,

The only thing that XAS_TDP can print is the linear-response orbitals 
(equation (11) in 
https://pubs.rsc.org/en/content/articlehtml/2021/cp/d0cp06164f) in cube 
file format. I am not familiar with Multwfn, and what kind of file it 
expects. From the error message you get, it's probably not a cube file (no 
electronic excitation info in there). Therefore, I don't think you can use 
CP2K and XAS_TDP in the way you were imagining.

That being said, your input file is suboptimal in multiple ways:
1) You are localizing all occupied orbital, and this takes the majority of 
the run time. This is unnecessary since you are interested in the Fe 1s 
state, which is very easy to identify (lowest energy orbital) and is 
already localized. You should use N_SEARCH 1, and you can remove the 
LOCALIZE keyword
2) Do not set EPS_PGF_ORB to 1.E-32. This is way too small, and it will 
massively impact performance, especially in a PBC setting. The default 
values of SQRT(EPS_DEFAULT) is good enough. See this keyword instead: 
https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/DFT/QS.html#CP2K_INPUT.FORCE_EVAL.DFT.QS.MIN_PAIR_LIST_RADIUS
3) XAS_TDP is a perturbative method that assumes a well converged SCF. It 
is dangerous to use a EPS_SCF of 1.0E-4, which is very loose. I would 
recommend 1.0E-6 at least.

Best,
Augustin

On Tuesday, 10 June 2025 at 14:59:19 UTC+2 Tian Lu wrote:

> As far as I know, the &XAS_TDP module is unable to print configurational 
> coefficient information to output file like &TDDFPT. You aim cannot be 
> achieved without modifying the CP2K source code.
>
> On Tuesday, June 10, 2025 at 2:55:43 PM UTC+8 qing zhou wrote:
>
>> Hello everyone,
>>
>> I am currently performing XAS calculations for a periodic system and 
>> would like to conduct Natural Transition Orbital (NTO) analysis. My XAS 
>> calculations are performed using the &XAS_TDP section, employing the PBE 
>> functional with a 45% Hartree-Fock (HF) component, and the def2-TZVP basis 
>> set for the excited Fe atom. After completing the calculation, I 
>> successfully generated the molden file, cube files, spectrum file, and out 
>> file. However, when I attempted to perform NTO analysis using Multiwfn, the 
>> program reported an error after loading the out file, indicating that 
>> electronic excitation information was missing.
>>
>> I would like to ask for your advice: how should I configure the settings 
>> to ensure that the out file contains the necessary electronic excitation 
>> information? I have attached my inp and out files for reference.
>>
>> Thank you all for your responses!
>>
>

-- 
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 visit https://groups.google.com/d/msgid/cp2k/2f0ce572-2f34-4ff5-af8e-e8c62114c51bn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20250610/1b93cd1b/attachment.htm>