[Wannier] center of MLWFs

[Jonathan Yates]

On Sun, 8 Jun 2008, nafise rezaei wrote:

> I have a question about center of MLWFs.
> Do center of MLWFs  change for valence states , when MLWFs be
> calculated in these two cases: once be calculated only for valence
> states and  once be calculated  for valence + conduction states?

Nafise,

  In general, yes they will change.

Just to address one possible confusion: When you create MLWF to describe 
valence and conduction states it isn't possible to talk about MLWF for the 
valence states - each MLWF has partial occupancy. You can look at the 
contribution of a given WF to a particular band at a particular k-point. 
(eg see bands_plot_project)

The example from the original paper (Maximally localized Wannier functions 
for entangled energy bands I. Souza, N. Marzari and D. Vanderbilt, Phys. 
Rev. B 65, 035109 (2002)) is that of silicon. For the valence states the 
MLWF are centred on the bond centres. For MLWF to span the valence and low 
lying conduction states we can choose sp3 like MLWF - which are not bond 
centred.

  Hope this helps.

  Yours
   Jonathan

-- 
Dr Jonathan Yates         |    Theory of Condensed Matter Group
Corpus Christi College    |    Cavendish Laboratory
Cambridge, CB2 1RH, UK    |    Cambridge, CB3 OHE, UK
email jry20 at cam.ac.uk     |    Tel +44 (0)1223 337461