GW counts semiconductors as metals

[WangFuyou]

Dear yambo developers，
When I calculate GW for the first time, QPkrange is set to 1 1, which then my GW result is semiconductor. However, I changed the QPkrange to 1 30, that is, after including all k points, I did the optical calculation and found that the output file told me that it was metal.
And I am sure that there is no problem with my electronic structure, I would like to ask what is the situation and how I can solve it.
Because my GW output file is too large, I can't upload it. But the output file from optics shows that the material was calculated as metal.

Yours,
WangFuyou

[Daniele Varsano]

Dear WangFuyou,
it is not straightforward to spot what the problem is.

I suggest you to inspect your QP band structure and see if you have some weird correction that makes the system as metallic.

In general anyway, only states across the gap are of interest for optics, and you do not need to calculate corrections for deep state. Also because for deep states, the plasmon-pole approximation is usually not appropriate.

I suggest you to check the convergence of your QP structure considering fewer states, across the gap e.g.

Code: Select all

%QPkrange                        # [GW] QP generalized Kpoint/Band indices
1|30|40|50|
%

Best,
Daniele