[Wien] Fwd:

[tran at theochem.tuwien.ac.at]

Hi,

If you look at the basic formula of the density matrix, you can see that
there is no reason to expect zero non-diagonal elements. For instance,
see Eq. (9) of
http://journals.aps.org/prb/abstract/10.1103/PhysRevB.62.16392

F. Tran

On Mon, 26 Jan 2015, huimei liu wrote:

> 
> Dears all:
>  Recently I have calculated density matrix for bcc-iron using WIEN2k. In
> case.dmatup for atom 1 and l=2, there should be a 5 by 5 matrix,
> 
> with m labelling the column and row. As I know, there should not be
> nondiagonal element for when we integrate the density matrix within the
> atomic sphere of atom 1 as the spherical harmonics are othorgonal
> sets,i.e.,integral corresponding to different m's should be zero. However, I
> get nonzero matrix elements in the position(-2,2) and (2,-2) from the
> case.dmatup. Maybe something is wrong within my understanding of the Bloch
> eigenfunction?
> 
>  The eigenfunction can be partitioned to two parts : one is in the
> intersitial region one is in the atoms spheres. I just consider that in the
> atomic spheres, which can be a linear combination of spherical harmonics. So
> is there something wrong?
> 
>  Best regards!
> 
>  Feng Tang
> 
>  From Nanjing University, China
> 
> 
> 
>