[Wien] Projected density of states

[Peter Blaha]

Let me start with basic comments:

For "cubic" cases, there is an eg/t2g switch automatically invoked. Thus the
"real" eg/t2g splitting is obtained "automatically".

Sometimes symmetry is lowered (small distortion,...), but still it might be
usefull to alanyle the states in eg/t2g, ALTHOUGH this is no longer a
valid (from group theory point of view) irreducible representation and thus
the splitting will have some "arbitrariness".

Anyway, for the latter case you could try    QTL.

> I would like to calculate the DOS on a different basis (e_g/t_2g)
> respect to the standard lms basis used in the scf calculation. As I
> understood from the 'comment' file in the 'SRC_qtl' subdirectory, the
> main steps to follow are 3:
>
> 1)create the ' case.inq'  file
>
> 2)copy the file  $WIENROOT/SRC_templates/case.cf_d_eg_t2g onto the file
> 'case.cf-1' , where -1 is the atom index.

I think you should just use   case.cf1  (without the -)

> 2)modify the local rotation matrix in the 'case.struct' file  according
> to the new basis (e_g/t_2g)
>
> About the third point: does it mean that I should open the structure
> file after the scf calculation and modify by hand the local rotation
> matrices written there?

Maybe. What you try to do here is "delicate" and not straight forward, although
it could be possible. Most likely you want to put the local coordinate system
such that it coincides with the coordinate system of an undistorted
octahedron.

                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------