[Wannier] charge of wannier function

[Hu Shuanglin]

Dear all,

I have something to update.
>From the plot.F90 file, we can see that the wannier function is gotten
by unitary
transfromation from the wave function read from UNK files. The UNK files are
written by VASP.
If I sum the charge density calculated from this wave function, it
will neither give
electron number of 1, nor equal to each other.

      UNK Function Num:    1       Number of Charge =  0.108717E+01
      UNK Function Num:    2       Number of Charge =  0.109259E+01
      UNK Function Num:    3       Number of Charge =  0.817972E+00
      UNK Function Num:    4       Number of Charge =  0.776551E+00
      UNK Function Num:    5       Number of Charge =  0.748194E+00
      UNK Function Num:    6       Number of Charge =  0.748194E+00
      UNK Function Num:    7       Number of Charge =  0.685133E+00
      UNK Function Num:    8       Number of Charge =  0.685133E+00

It maybe says that the wave function from UNK files are already not
normalized. Is it
correct?
If I normalized each wave function, and do the unitary transformation
again, the
electron number of each wannier function is still not the same, but
the total number
of electrons is now equal to 8.

      Wannier Function Num:    1       Number of Charge =  0.910610E+00
      Wannier Function Num:    2       Number of Charge =  0.910610E+00
      Wannier Function Num:    3       Number of Charge =  0.103225E+01
      Wannier Function Num:    4       Number of Charge =  0.103214E+01
      Wannier Function Num:    5       Number of Charge =  0.103209E+01
      Wannier Function Num:    6       Number of Charge =  0.102505E+01
      Wannier Function Num:    7       Number of Charge =  0.103220E+01
      Wannier Function Num:    8       Number of Charge =  0.102505E+01

So is there anything wrong with my understanding and method? Can I use the
renormalized wannier function to construct total charge density?

Thank you!

Best Regards,
Shuanglin Hu

On Fri, Jun 29, 2012 at 12:19 PM, Hu Shuanglin <hushuanglin at gmail.com> wrote:
> Dear all,
>
> I am new to wannier90, and try to use wannier90 and vasp together.
> I thought since wannier function is the unitary transfromation of
> Bloch functions,
> 1, each wannier centers here should correspond to one charge center,
> and each could
> be occupied by 2 electrons (without spin-polarization).
> 2, from the occupied wannier functions, I can do summation to get the
> overall charge
> density that equal to vasp CHGCAR data.
> Is this understanding correct?
>
> Then in order to test this, I added some lines like following in the
> plot.F90 file:
>
> wann_rho(nxx,nyy,nzz,loop_w)=
> wann_func(nxx,nyy,nzz,loop_w)*conjg(wann_func(nxx,nyy,nzz,loop_w)
> wann_chg(nxx,nyy,nzz)= wann_chg(nxx,nyy,nzz)+wann_rho(nxx,nyy,nzz,loop_w)
> sum_rho(loop_w)=sum_rho(loop_w)+wann_rho(nxx,nyy,nzz,loop_w)
>
> After the summation, the "electron number of each wannier function" is
> printed out
> sum_rho(loop_w)=sum_rho(loop_w)/(ngs*ngx*ngs*ngy*ngs*ngz)
>
> I am studying a insulator LiOH, in which there are 16 valence
> electrons in one unit cell,
> and vasp gave a default num_wann=8 in the .win file. The loop_w is
> looped over all the 8
> occupied band in this model, and only Gamma point is used in k-space.
> However, I get a number lower than 1 for one wannier band, and they do
> not equal to each
> other between wannier bands.
>
>       Wannier Function Num:    1       Number of Charge =  0.843700E+00
>       Wannier Function Num:    2       Number of Charge =  0.843700E+00
>       Wannier Function Num:    3       Number of Charge =  0.833333E+00
>       Wannier Function Num:    4       Number of Charge =  0.832989E+00
>       Wannier Function Num:    5       Number of Charge =  0.832841E+00
>       Wannier Function Num:    6       Number of Charge =  0.810597E+00
>       Wannier Function Num:    7       Number of Charge =  0.833188E+00
>       Wannier Function Num:    8       Number of Charge =  0.810594E+00
>
> In this case, the total number of electrons will certainly not to be 8*2=16.
>
> Is there anything wring in my understanding or inplementation? Any
> comments could be helpful.
> Thank you!
>
> Best Regards,
> Shuanglin Hu
> --
> Department of Chemistry - the Angstrom Lab
> Uppsala University
> Box 538, SE 751-21
> Uppsala, Sweden