[Pw_forum] Querry abou the kpoints in transport calculations

[Gabriele Sclauzero]

Dear Pankaj,

     if you are speaking about the k-points in the scf calculations, 
then the correct thing to do is the second.
The optimal number of k-points depends on the band structure and on the 
size of the BZ. The latter in turn depends on the size of the simulation 
cell. Since this is usually much shorter for the electrode, then the 
corresponding number of k-points along the transport direction must be 
larger.
     Regarding the k-points in the transport calculations, if you are 
simulating a one-dimensional system then you just need one k-point in 
the perpendicular directions xy (because the transmission does not 
depend on k_perp, in principle). If the system has extended leads 
instead, you need to choose a k_perp-points set that integrates the 
transmission in the 2D-BZ with sufficient accuracy. This set might be 
larger than that of the scf calculation.

HTH

GS

On 04/11/2012 09:14 AM, pankaj sahota wrote:
> Hi,
>
> Can anybody tell me that if I want to calculate the transport 
> properties of a wire then do I have to take the same kpoints for the 
> scattering region and electrode or they can be chosen differently 
> because kpoints for both of them will be generated differently due to 
> the different length along the z-direction.
>
> Thanks and regards.
> Pankaj
>
> -- 
> Pankaj Kumar
> DST Project Fellow(Ph.D.)
> Department of Physics
> The LNM IIT Jaipur
> Mob. No. +91 9736694726
>
-- 

Gabriele Sclauzero, EPFL SB ITP CSEA
PH H2 462, Station 3, CH-1015 Lausanne

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20120413/87854c8e/attachment.html>