Yambo Community Forum

Dear Alejandro,

I am trying to do electron-phonon corrections on top of previously calculated g0w0 QP energies; however
when i use GfnQPdb= "E < SAVE/db.QP" , i end up with a bad termination error.

How can i perform electron-phonon calculation on top of g0w0 QP energies, in order to be able to include electron_phonon affected g0w0 QP energies in the subsequent BSE calculations ?
(Otherwise yambo_ph calculates it on top of KS energies, and i loose electron-electron effects.)

Further question, Andrea Marini talks in his seminal PRL paper from 2008 about exciton-phonon coupling (coherent contributions for hBN); However it seems that this part is missing in the Tutorial.
How can i account for exciton-phonon coupling ?

Best wishes
Martin

I also had the same questions.



Best,

Dear Haseeb,
can you elaborate a bit more what is the problem you are facing?

Daniele

martinspenke wrote: ↑Wed Dec 28, 2016 5:07 pm Dear Alejandro,

I am trying to do electron-phonon corrections on top of previously calculated g0w0 QP energies; however
when i use GfnQPdb= "E < SAVE/db.QP" , i end up with a bad termination error.

How can i perform electron-phonon calculation on top of g0w0 QP energies, in order to be able to include electron_phonon affected g0w0 QP energies in the subsequent BSE calculations ?
(Otherwise yambo_ph calculates it on top of KS energies, and i loose electron-electron effects.)

Further question, Andrea Marini talks in his seminal PRL paper from 2008 about exciton-phonon coupling (coherent contributions for hBN); However it seems that this part is missing in the Tutorial.
How can i account for exciton-phonon coupling ?

Best wishes
Martin

Dear Daniele,

I made the comment to Martin's question of this thread, which somehow remained unanswered.


Regards,

Dear Martin

I think what you want to do is not correct. Let me explain.
When you construct the electron-phonon self-energy you have at the numerator the electron-phonon
matrix elements and at the denominator the Kohn-Sham energy differences.

if you replace the KS energy differences of the denominator with the GW ones,
you should do the same with the numerator matrix elements.
Otherwise, your self-energy would not be consistent.

Moreover in many cases, but not all cases, there is a cancellation between the GW correction
in the numerator and in the denominator, see the discussion on page 3 in

https://arxiv.org/abs/0808.2285

In the end, if you want to calculate the effect of EPC on the band structure you should start from KS and then add
the relative shift and linewidths to the GW result.

If you want to include GW correction in the EPC calculation, you have to use a code like the one of Steven-Louie
(that I think it is not available)

https://journals.aps.org/prl/abstract/1 ... 122.186402

or doing it by finite differences

https://journals.aps.org/prb/abstract/1 ... .93.100301

best regards
Claudio Attaccalite