[CP2K-user] [CP2K:17814] Re: magnetic systems in cp2k / AiiDA common workflows

[Leopold Talirz]

Hi Matthias,

After chatting with Sasha, he mentions that, for the body-centered cubic Fe 
structure

 a) the equation of state of the ferromagnetic state (=ground state) fit 
relatively well with the other codes for 5x5x5 k-points (see 
https://github.com/aiidateam/aiida-common-workflows/issues/117#issuecomment-799611434), 
but started to deviate substantially when increasing the number of k-points 
(!)

 b) the equation of state for the antiferromagnetic state (artificial, not 
the ground state) was always off, and total magnetization per cell was ~1/4 
Bohr magneton per atom above the result for Quantum ESPRESSO.

He mentions he tried a number of things


   1. Fix the total magnetization to 0.
   2. Set a very large k-point grid: 20 20 20
   3. Set MGRID CUTOFF to 1000
   4. Use a very large basis: TZV2P-MOLOPT-SR-GTH-q16
   5. Set EPS_DEFAULT to 1.0E-16
   6. Use different starting magnetization: +4/-4, +2/-2.

He also mentions that one thing he did not try to vary was the 
pseudopotential, and that you recently added a potential that would make 
sense to try.
Could you please point me to this potential, and perhaps the basis you 
would recommend to use with them?

I will then try to rerun the workflow and report results back (including 
the input files generated).

Cheers!
Leopold

On Wednesday, 5 October 2022 at 14:30:25 UTC+2 Aliaksandr Yakutovich wrote:

> Hey.
>
> Please find more info here: 
> https://github.com/aiidateam/aiida-common-workflows/issues/117
>
> Best,
> Sasha
>
> On Monday, 3 October 2022 at 15:21:45 UTC+2 Leopold Talirz wrote:
>
>> Hi,
>>
>> I've been looking through the AiiDA common workflows paper [1] and I 
>> noticed in table 2 of the SI [2] that both the ferromagnetic and the 
>> antiferromagnetic Fe converged to an incorrect ground state with CP2K for 
>> "as of yet unknown reasons". 
>>
>> I reached out to coauthor SashaYakutovich to ask whether this was 
>> eventually resolved, but it seems like this is not yet the case.
>>
>> I'd be curious to understand whether someone could help shine some light 
>> on what may be underlying this issue, as well as potentially some more 
>> general advice on the applicability of cp2k to magnetic systems (both 
>> periodic and clusters).
>>
>> Cheers,
>> Leopold
>>
>> P.S. Sasha may be able to provide some more plots / concrete input files 
>> as further context. 
>>
>> [1] Common workflows for computing material properties using different 
>> quantum engines | npj Computational Materials (nature.com) 
>> <https://www.nature.com/articles/s41524-021-00594-6>
>> [2] 41524_2021_594_MOESM1_ESM.pdf (springer.com) 
>> <https://static-content.springer.com/esm/art%3A10.1038%2Fs41524-021-00594-6/MediaObjects/41524_2021_594_MOESM1_ESM.pdf>
>>
>

-- 
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/7dd446a7-c841-4ebd-8e9e-c4c742c61272n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20221005/91b85328/attachment-0001.htm>