[Pw_forum] how to choose bds in PWCOND?

Gabriele Sclauzero sclauzer at sissa.it
Sat Oct 2 15:51:16 CEST 2010


Dear Manoj,

Il giorno 01/ott/2010, alle ore 23.04, Manoj Srivastava ha scritto:

> Dear All, 
> I am confused about choosing bds parameter in the transmission
> calculation of PWCOND. My question is about example on Smogunov's website
> Monatomic Ni wire with a spin reversal, which can be accessed at
> http://people.sissa.it/~smogunov/ 
> The manual says- right boundary of the scatt. reg. (left one is at 0 if
> prefixs is used and = bdl if prefixt is used) Now in this example bds is
> chosen to be at 1.0, but if I let the system relax and atom at z=1.0 moves
> away from z=1, would bds still be equal to 1.0?

I think that's not correct. If you keep the same left lead, then there must be an atom exactly at the right boundary of the Scattering Region. If you let relax the atoms in the larger cell, then you should change bds according to the position of the rightmost atom that you want to include in the S.R. for the pwcond calculation. Be careful that the scf potential in a periodic unit cell (i.e. that corresponding to the infinite, equally spaced chain) close to the right boundary of the S.R. must be, to a good approximation, equal to that in the left lead (the same applies to the right boundary). Hence you may need to include more atoms in the larger supercell and fix the atomic positions of those that will correspond to the leftmost and rightmost regions inside the S.R. (according to your choice of bdl and bds).


HTH


GS

> I am guessing so, but I am
> not sure. Would someone mind explaining this?
> 
> -Manoj Srivastava 
> Department of Physics, 
> University of Florida, 
> Gainesville, FL
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum


§ 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/20101002/ea4831bd/attachment.html>


More information about the users mailing list