[Wannier] Wannier90/VASP problems

[Nicola Marzari]

Hi Mario,


can you plot the WFs? It would be a safe check to see that the
centers and the WFs themselves are consistent.

When you say that you checked the DOS - is this the PDOS on
f orbitals - i.e. you have a sense of where the electrons sit,
and is it contradictory with the WFs?

Bottom line: the first step is figuring out what kind of MLWFs
you are getting, and see if everything is consistent with the PDOS.
That should be the case - otherwise it points to a computational issue.

If all is ok in the step above, the next thing to do is figuring out
what the correct chemistry should be. If HSE and PBE+U give the same
result, one should have some hope that the problem at hand is studied
correctly, but note that you are dealing with a system that is in
general difficult to describe correctly.

			nicola



On 13/02/2012 16:10, Mario Burbano wrote:
> Dear All,
>
> I am currently working on Gd-Doped ceria using vasp (PBE+U/HSE) and
> using Wannier90 as a library. So far, I've had good results with a
> similar system (Y-doped Ceria) which I've compared against those from
> CPMD. However, now that I am trying Gd-doped ceria I am getting some
> unexpected results: the wannier centres that should be associated with
> Gd are localized around Ce. I have checked the density of states from
> VASP, and everything seems ok. I have also checked the wannier90.wout
> file and the spreads seem to be fine, (between 2.5 and 3 Bohr^2, and
> there is a block of WFs with a smaller spread between 1.6 and 2 Bohr^2,
> which I'm guessing correspond to the f-electrons). This is a spin
> polarized calculation.
>
> Thanks
>
> Mario Burbano
>
>
>
> _______________________________________________
> Wannier mailing list
> Wannier at quantum-espresso.org
> http://www.democritos.it/mailman/listinfo/wannier


-- 

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