[Wannier] Question about the matrix elements in seedname_hr.dat

[Saeed Bahramy]

Hi Nicolas,

Thank you very much for your reply. I use a Gamma centered 10x10x10  
kmesh.

I'm quite sure that I'm looking at WFs corresponding to top and bottom  
atoms in the original supercell. If I set  
"translate_home_cell=.false.", I can clearly see some of the Wannier  
centers (those close to the lower plane of the cell) are located  
outside the supercell but they all look pretty normal with  
"translate_home_cell=.true.". This option however only works for the  
wannier centers and not for the matrix elements .....

Anyway thanks again for your help,
Best,
Saeed




On Jun 30, 2011, at 9:41 PM, nicolas poilvert wrote:

> Dear Saeed,
>
> How many k points are you using in the z direction
> of your supercell ? Are you using Gamma sampling
> for the orthogonal directions ?
> Also, the wannier code does not order the wannier
> functions at all, so make sure that the functions that
> you are looking at are indeed the ones corresponding
> to the proper locations.
> At last, if you use the "translate_home_cell = .true."
> option, the centers will be translated back to the original
> supercell that you used. Sometimes, some wannier
> functions that are very close to the plane separating
> two adjacent supercells are translated all the way back
> because of rounding.
>
> hope this helps,
>
> best,
>
> Nicolas
>
> On Thu, Jun 30, 2011 at 3:40 AM, Saeed Bahramy <bahramy at riken.jp>  
> wrote:
> Dear Wannier90 developers,
>
> I have a supercell for which I've been able to make a TB model using
> wannier functions. The TB model seems to be working  fine. However the
> transfer matrix elements listed in seedname_hr.dat show somewhat
> unexpected values. Let's suppose my supercell consists of two unit
> cells along z direction and each cell contains 4 atoms. The relative
> positioning of atoms is such that atom #1 of cell #1 (let's call it
> A1) is located near the bottom of supercell and atom #4 of cel #2
> (let' call it A2) is close to the top of the supercell. In this case,
> I expect that the transfer matrix elements between A1-A2  to be quite
> small inside the supercell   (rvec= 0 0 0), whereas they should show
> appreciable values for rvec=0 0 1. However, what I see is totally
> opposite, i.e., A1-A2 elements are large for rvec=0 0 0 and quite
> small for rvec = 0 0 0. More strangely, for other two  atoms-adjacent
> to each other and  located well inside the first cell- I strangely
> encounter the situation which I normally expected for A1 and A2. So
> this basically means that the program shifts all the atoms along z-
> direction and, hence, assumes a different ordering for atomic
> positions inside the supercell. I was wondering, if this is a bug or
> there is an option to turn off such a reordering.
>
> Thanking you in advance,
>
> -----------------------------------------------------------
> Saeed Bahramy
> Correlated Electron Research Group
> Advanced Science Institute, RIKEN
> 2-1 Hirosawa, Wako
> Saitama 351-0198 JAPAN
> TEL: +81-48-462-1111 (EXT. 6013)
> FAX: +81-48-462-4691
> Email: bahramy at riken.jp
> -----------------------------------------------------------
>
>
>
>
>
>
>
> _______________________________________________
> Wannier mailing list
> Wannier at quantum-espresso.org
> http://www.democritos.it/mailman/listinfo/wannier
>
>
>
> -- 
> POILVERT Nicolas
> PhD candidate,
> Dpt of Materials Science and Engineering
> Massachusetts Institute of Technology
> 77, Massachusetts avenue
> Cambridge, MA 02139
> USA
> work: (617) 452-4212
> nicolas.poilvert at gmail.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20110630/f949001c/attachment.html>