G0W0 convergence for metals

[DavidPolito93]

Dear all,

Looking around the various tutorials I couldn't find an example regarding the convergence of G0W0 band structure in metals.
More precisely, does the same scheme regarding convergence of the exchange and correlation self-energy applies or is it some other technical issue or input variable to take into account?

I thank you in advance,

Davide Romanin
-----------------------------------------------------
PhD student in Physics XXXIII cycle
Representative of the PhD students in Physics
Applied Science and Technology department (DiSAT)
Politecnico di Torino
Corso Duca degli Abruzzi, 24
10129 Torino ITALY
------------------------------------------------------

[Daniele Varsano]

Dear Davide,
for metals, you would probably need to include many k points.
Anyway, it is important to note that the intraband transitions are not included.
You can take them into account via a Drude model, but I'm not sure they take effect in the GW ppa approximation.

Best,
Daniele

[DavidPolito93]

Dear Daniele,

Thank you for your response!

Therefore, also the wrong Fermi level given by the band interpolation step might be related to a poor k-point convergence? (I mean wrong with respect to the PW calculation).

Moreover, when performing the G0W0 calculation I got the warning:

""[WARNING] The system is a metal but Drude term not included.""

Which I think it was the term you were talking about. So, it might play a role in this calculation?

Best,

Davide Romanin
-----------------------------------------------------
PhD student in Physics XXXIII cycle
Representative of the PhD students in Physics
Applied Science and Technology department (DiSAT)
Politecnico di Torino
Corso Duca degli Abruzzi, 24
10129 Torino ITALY
------------------------------------------------------

[Daniele Varsano]

Dear Davide,
the Fermi level is read by QE in the setup.
When it is recalculated it is possibile that it differs from QE because of the different treatment of the smearing.

Which I think it was the term you were talking about. So, it might play a role in this calculation?

Yes, it is the intraband term and it plays a role, we are working on a strategy to include it in a plasmon-pole calculation, but at the moment the only way to include it is by adding the Drude term in input when performing a full frequency calculation.
Best,
Daniele