Separate convergence for two spin modes

Concerns issues with computing quasiparticle corrections to the DFT eigenvalues - i.e., the self-energy within the GW approximation (-g n), or considering the Hartree-Fock exchange only (-x)

Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, andrea marini, Daniele Varsano

Post Reply
8813204602
Posts: 58
Joined: Tue Jun 15, 2021 5:53 pm

Separate convergence for two spin modes

Post by 8813204602 » Sat Oct 09, 2021 9:12 pm

Hello,

In calculating the band structure of a magnetic material, should the convergence tests of the parameters relative to the HF gap or GW gap for both spin states be checked separately?

Will using double precision instead of single precision in compilation have an effect on performing convergence tests, in other words will using double precision make convergence easier?

Best regards,
Mitra
Mitra Helmi,Ph.D student
Ferdowsi university of Mashhad
Mashhad,Iran

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: Separate convergence for two spin modes

Post by Daniele Varsano » Sun Oct 10, 2021 10:31 am

Dear Mitra,
in general there is no need to perform separate convergence calculation anyway you double check the convergence calculating the gap for each spin channel as this does not take more much computational time.
will using double precision make convergence easier?
Single precision is sufficient for convergence tests, the precision one wants generally reach in convergence test is far smaller than the the difference between a SP/DP results.

Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

8813204602
Posts: 58
Joined: Tue Jun 15, 2021 5:53 pm

Re: Separate convergence for two spin modes

Post by 8813204602 » Tue Jan 18, 2022 8:53 am

Dear Daniele,

I performed convergence tests for both up and down spin channels. The converged values for some parameters, such as Gbndarnge and BndsRnXP, were obtained different for up and down spins. For example, as shown in figure, the converged values of GbndRnge for the up and down spins are 110 and 40,respectively.Should I consider the higher value as the converged value for GbndRnge, ie GbndRnge=110 or the value of the down spin can be considered because the down spin gap is less than the up spin?
The DFT gap values for these two spin channels are:
Direct Gaps [ev]: 1.207025 1.795936
[spin up]: 4.889193 7.215230
[spin dn]: 1.315511 2.047675

Best,
Mitra
You do not have the required permissions to view the files attached to this post.
Mitra Helmi,Ph.D student
Ferdowsi university of Mashhad
Mashhad,Iran

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: Separate convergence for two spin modes

Post by Daniele Varsano » Tue Jan 18, 2022 9:26 am

Dear Mitra,

this is reasonable as the two spin channels can have a very different localization. About the convergence parameter, you want to adopt this really depends on what you are interested in, if you are interested in the spin-down gap only you can use the lower parameters, if you are interested in both gaps or e.g. you want to perform BSE optics the larger one needs to be used.
Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

Post Reply