[Wien] Decomposition of Density of States

[Peter Blaha]

You are absolutely right, WIEN2k uses the coordinate system given by the 
local rotation matrix.

If you read their PRL paper, you will find that they rotated the 
coordinate system in such a way, that it fits approximately to the 
distorted octahedra of O atoms around Cr. This is VERY different from 
the coordinate system used in WIEN2k, which applies STRICT symmetry (and 
not approximate one).

 From their Fig.1 one can also see that their d-xz should look very 
similar as a d-z2 orbital (plus some other admixtures), which agrees 
with your observations.

However, you can use the   qtl  program (instead of lapw2 -qtl) and 
there is an input option which allows you to rotate the coordinate 
system in any direction you like. The resulting QTLs (and PDOS) is then 
in your chosen coordinate system.

PS: If you choose your coordinate system properly, the 3 t2g and 2 eg 
orbitals should be  nicly separated.
In addition using x lapw2 -all emin emax    you can create a density 
from an energy range where one particular orbital dominates and plot the 
resulting electron density to verify that the orbital really looks like 
your expectations.


On 1/21/20 2:30 PM, Guoping Zhang wrote:
> PRL Vol. 80,
> 4305 (1998)
I have a short question on the density of states decomposition for
CrO2 with spin polarized calculation. It has rutile TiO2 structure,
but with a small distortion along the apex oxygen atom.

According to the UG, "The m-decomposed DOS (e.g. pz , py , px ) is
given with respect to the local coordinate system", but when I plot
DOS of DZ2 for Cr (atom 1) and compare it with Fig. 4 of PRL Vol. 80,
4305 (1998), Dz2 looks more like their Dxy. Then I check the rotation
matrix of Cr atom (which is the same as in the Appendix of UG) to see
whether the decomposition is for the rotated atom, but the rotation is
45 degree rotation with respect to the $c$ axis, so there is no way
Dz2 could become Dxy.

Therefore, I wonder whether WIEN2k uses a different local coordinate
from the above literature where the z axis is from the Cr at the body
center to the apex oxygen.

I have to know which axis orbitals DZ2,DX2Y2,DXY,DXZ,DYZ use.

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/TC_Blaha
--------------------------------------------------------------------------