different Eliashberg spectral function a2F in epw and PHonon

Post here questions linked with issue while running the EPW code

Moderator: stiwari

Post Reply
xiaming
Posts: 5
Joined: Mon Dec 19, 2022 11:58 am
Affiliation: Student

different Eliashberg spectral function a2F in epw and PHonon

Post by xiaming »

Dear all
When I calculate the Eliashberg spectral function in EPW and PHonon(the module in QE), I found they are different. Especially, the DOS near 0 frequency jumps sharply in EPW but not in PHonon. And the lambda caculated in EPW seems too large, so strange! Maybe something wrong in the Phonon spectrum calculated in ph.x? Since the TA mode is too soft near gamma point? The pictures are showed in the attachments. Can anyone tell me why? and I should I do to get the right result.
Attachments
Phonon spectrum.png
Phonon spectrum.png (121.92 KiB) Viewed 5975 times
a2f_PHonon.png
a2f_PHonon.png (127.78 KiB) Viewed 5975 times
a2f_EPW.png
a2f_EPW.png (68.34 KiB) Viewed 5975 times
gkafle1
Posts: 31
Joined: Wed Jun 17, 2020 8:55 pm
Affiliation: Binghamton University

Re: different Eliashberg spectral function a2F in epw and PHonon

Post by gkafle1 »

Hi xiaming,

Which "electron_phonon" method are you using in the PHonon code? In my experience, electron_phonon ='interpolated' in the PHonon code gives the a2F in very good agreement with the EPW code but electron_phonon ='simple' does not. In the electron_phonon ='interpolated' method, it uses the interpolation scheme both in k- and q-grid so that your a2F agrees well with the EPW code.

Sometimes the artifact (soft mode) you see around the Gamma point may cause you to jump near zero frequency in EPW. I would suggest increasing the q-mesh along the in-plane direction to calculate phonon so that your artifact can be corrected near Gamma, and hopefully, your EPW results will also be meaningful. Also, please make sure that your phonon is converged with respect to k- and q-mesh.

Hope this helps.

Best
Gyanu
Binghamton University
Post Reply