[Q-e-developers] [Q] hybrid functional support at Quantum Espresso

[Layla Martin-Samos]

Dear Jeon, QE supports hybrid functionals. There is still no specific
information about the parameters because it is not on its final form. nqx
are the number of q vectors to use for the coulomb potential (1/(q+G)**2),
if you want to be "exact" use the same number as your kpoint grid (if you
choose to use less, you need to test the convergence). screening_parameter
is the broadening that is use for the erfc function in HSE (the short range
part of the coulomb potential), if it is not specified and you force
input_dft = "HSE" (in system namelist), then the default value for HSE06 is
used.

we are working now on improving the computational performances of hybrids.

I hope this could help

Layla

2011/7/23 B. Jeon <bjeon at seas.harvard.edu>

> Dear Developers of QE:
>
> Hi, I am a postdoctoral researcher working at Harvard, and am trying to
> use QE for hybrid functional analysis of metal oxides, like VO2 and
> SmNiO3.
>
> Seems that QE supports hybrid functionals like HSE (from the
> introduction of website) but I couldn't find any guide or explanations
> of employed parameters like nqx, screening_parameters. Are those hybrid
> functional under development yet?
>
> Also some elements like V have only uspp pseudopotential, but they seem
> not supported yet. Only Trouillers-Martin type ?
>
> Any comment regarding hybrid functionals of HSE will be greatly
> appreciated. Hope to hear from you soon.
>
> Best regards,
>
> ByoungSeon Jeon
>
> _______________________________________________
> Q-e-developers mailing list
> Q-e-developers at qe-forge.org
> http://qe-forge.org/mailman/listinfo/q-e-developers
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20110726/ca2d43ef/attachment.html>