[Pw_forum] graphene conduction bands

[Nicola Marzari]

Dear Fabrizio,


1) in the band structure of graphene you not only have the \pi \pi^star
manifold, but many free electron parabolic bands overlapping (see e.g.
just googling 
http://www.intechopen.com/source/html/17504/media/image2.png ). Try 
using 15-20 bands, instead of 5

2) in the region where things go amiss, it looks like PWSCF is not
giving you always the lowest possible band, but most often around gamma
the second lowest possible band (i.e. the 6th band) and only 
occasionally the 5th band. This is somewhat related to Davidson
diagonalization finding eigenstates, but not necessarly the lowest ones
- it would be interesting to see what would happen if you'd specify 
"random" as the choice for starting wavefunctions, use a very tight 
tolerance for convergence, and maybe switch from Davidson to "cg" -
then the code would probably always get the lowest 5th band (i.e. the 
lowest parabolic free electron close to gamma, and \pi close to
K/M).

			nicola




On 04/04/2013 13:07, fabrizio gala wrote:
> Dear all,
> I am tryng to evaluate the graphene electron bands with QE, however, i
> miss (almost) completely the trend of the first conduction band.
>
> In the attachment you can find my imput file for the scf+nscf
> calculations together with the computed electron bands  (the ps file).
> Any suggestion will be greatly appreciated
> best regards
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>


-- 

----------------------------------------------------------------------
Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL