Can Yambo deal with the BSE of metallic systems?

[Guangfu Luo]

Dear friends,

Recently, I encounter an error while testing the BSE spectrum of graphene---“[ERROR]Fractional e/h occupations found. BS coupling forced.”

I searched this forum and found a former discussion on this issue (viewtopic.php?f=13&t=195). However, that discussion focused mainly on the physics of exciton and the cause of this issue was not mentioned. Though the exciton effect is inessential in bulk metals, it counts in some low-dimensional metallic systems, such as carbon nanotubes [PRL 92, 07742 (2004)].

Would you please explain the cause of such error and is it possible for Yambo to deal with excitonic effects of metallic system?

Sincerely,

[claudio]

Dear Guangfu Luo

if you have a semimetal like graphene (or nanotubes) you can choose a k-point sampling
that does not contain the metalic k-points.
For example in graphene any k-point sampling that is not divisible by 3 like 2x2x1 4x4x1,....., 100x100x1
and then calculate the normal Bethe-Salpeter using these samplings.
For example have a look to the papers:

Ab initio high-energy excitonic effects in graphite and graphene
Paolo E. Trevisanutto, Markus Holzmann, Michel Côté, and Valerio Olevano
Phys. Rev. B 81, 121405(R) (2010)
and
Excitonic Effects on the Optical Response of Graphene and Bilayer Graphene
Li Yang, Jack Deslippe, Cheol-Hwan Park, Marvin L. Cohen, and Steven G. Louie
Phys. Rev. Lett. 103, 186802 (2009)

Claudio

[Guangfu Luo]

Dear Claudio,

Thanks for your help; I got the point. (This trick is also used in the Raman spectrum calculation of graphene)

Sincerely,