electron-phonon matrix elements

General discussion around the EPW software

Moderator: stiwari

Post Reply
AmirhTalebi
Posts: 11
Joined: Fri Jul 10, 2020 8:37 am
Affiliation: University of Luxembourg

electron-phonon matrix elements

Post by AmirhTalebi »

Hi

I have to calculate electron-phonon matrix elements for my system. I run ewp.in file, there are a set of outputs like *.mmn , *.nnkp ,etc.
However, I do not know where to find the electron-phonon matrix elements. I will get them after running epw.x or even before that during ph.x ... ?
I am just interested for q = 0. I have to say that during phonon calculation the 'dvscf' flag is activated.

Any help is appreciated
Amir

hlee
Posts: 415
Joined: Thu Aug 03, 2017 12:24 pm
Affiliation: The University of Texas at Austin

Re: electron-phonon matrix elements

Post by hlee »

Dear Amir:

You can use the input of "prtgkk" for this purpose (See https://docs.epw-code.org/doc/Inputs.html#prtgkk ).
With this command, you can print the "interpolated" electron-phonon matrix elements and since it generates huge data in epw.out, you need to use small values of "fine" grids.

Sincerely,

H. Lee

AmirhTalebi
Posts: 11
Joined: Fri Jul 10, 2020 8:37 am
Affiliation: University of Luxembourg

Re: electron-phonon matrix elements

Post by AmirhTalebi »

Dear H. Lee

Thanks for your reply. I applied the flag and it is printed now in my output. However, it is the absolute value of g (|g|).
I am wondering how I can access to all the g(k,q) and not the absolute value ?


Best regards,
Amir

hlee
Posts: 415
Joined: Thu Aug 03, 2017 12:24 pm
Affiliation: The University of Texas at Austin

Re: electron-phonon matrix elements

Post by hlee »

Dear Amir:

If you want to print the absolute value of the electron-phonon matrix elements on fine grids, you need to modify the subroutine of print_gkk (https://gitlab.com/QEF/q-e/-/blob/devel ... ng.f90#L19).

Sincerely,

H. Lee

AmirhTalebi
Posts: 11
Joined: Fri Jul 10, 2020 8:37 am
Affiliation: University of Luxembourg

Re: electron-phonon matrix elements

Post by AmirhTalebi »

Dear Dr. Lee

Thanks for your reply.

during the epw.in run I am getting this error, the source of it is not clear for me.
Is it coming from the inconsistency between nk1, nk2, nk3 and nkf1, nkf2, nk3. ? However, I've used the same nki's as in my nscf (coarse mesh) and the nkfi's are the same value as nki's (for i = 1,2,3).

Error in routine dynmat_asr (1):
inconsistent data

best regards
Amir

hlee
Posts: 415
Joined: Thu Aug 03, 2017 12:24 pm
Affiliation: The University of Texas at Austin

Re: electron-phonon matrix elements

Post by hlee »

Dear Amir:

It seems that there is some mismatch between data from prefix.dyn_qX files (X is a core index) and that from your previous runs.
See https://gitlab.com/QEF/q-e/-/blob/devel ... r.f90#L327 .

I would suggest you to study and understand carefully the basic examples in EPW/examples before the production run for your research.

Sincerely,

H. Lee

Post Reply