[Wien] QTL

[jan kunes]

Hi Doru,

just a few suggestions. 
1) Check the unitarity test (in case.qtlout). If the case.cf matrix is 
unitary the total 4f DOS should be the same as from lapw2.
2) Use the full mode as a test. You should see that the PDOS's for 
symmetry related orbitals are identical. 
3) Check the local rotations matrices. The f orbital, on which you are 
making the projection are defined by a combination of local rotation 
matrix and case.cf file. As far as I remember the case with spin-orbit is 
a little bit tricky. The local z-axis should point along the spin 
quantization axis (define in case.inso). You can change the local rotation 
matrices in case.struct just for the qtl run to get correct results (don't 
forget to change them back if you want to continue with other 
calculations).

Good luck

Jan Kunes 

On Tue, 8 Mar 2005, Doru Torumba wrote:

> Dear all,
> 
> We are trying to calculate the 4f-DOS projected on a different basis
> (a2u,t1u,t2u) using the QTL program. To understand this program we took
> first a simple test case, bcc-Ho. After the convergence (with SO and LDA+U)
> we created the following 'case.inq' file:
> SUMA        /FULL      ( j=1/2 or 1/2,-1/2 and 1/2,1/2 extra)
> DOSYM       /NOSYM     ( symmetrization is/is not performed )
> -3. 3.                 ( energy window)
> 1.08090                ( fermi energy)
> 1                      ( number of atoms for which the DOS are calculated)
> 1 3                    ( jatom  latom)
> 
> and 'case.cf1' file with the matrix corresponding to the desired f-electron
> splitting:
> 
> *0. 0.  1. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.                  0. 0.  0.
> 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.                  a2u
>  0. 0.  0. 0.  0. 0.  0. 0.  -.61237244 0.  0. 0.  0.79056942 0.  0. 0.  0.
> 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.
>  -.79056942 0.  0. 0.  -.61237244 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0.
> 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.
> *0. 0.  0. 0.  0. 0.  1. 0.  0. 0.  0. 0.  0. 0.                  0. 0.  0.
> 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.                   t1u
>  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  1. 0.  0. 0.                  0. 0.  0.
> 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.
>  0. 0.  0. 0.  0. 0.  0. 0.  0.79056942 0.  0. 0.  0.61237244 0.  0. 0.  0.
> 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.
> *-.61237244 0.  0. 0.  0.79056942 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0.
> 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.    t2u
> *0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.                  0. 0.  1.
> 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.    a2u
>  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0.
> .  -.61237244 0.  0. 0.  0.79056942 0.
>  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  -.79056942 0.  0.
> .  -.61237244 0.  0. 0.  0. 0.  0. 0.  0. 0.
> *0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.                  0. 0.  0.
> 0.  0. 0.  1. 0.  0. 0.  0. 0.  0. 0.    t1u
>  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.                  0. 0.  0.
> 0.  0. 0.  0. 0.  0. 0.  1. 0.  0. 0.
>  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.
> 0.79056942 0.  0. 0.  0.61237244 0.
> *0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  0. 0.  -.61237244 0.  0. 0.
> 0.79056942 0.  0. 0.  0. 0.  0. 0.  0. 0.    t2u
> 
> Running the QTL program (x qtl -c -so -up/dn) we should get a 'case.qtl1'
> file with 3 columns corresponding to the a2u-t1u-t2u splitting. For spin up
> we could get the 3 columns only after we removed the last 3 'stars' (*) from
> the 'case.cf1' file (otherwise we have 6 columns and lots of ' ******* ').
> At first sight, there is no problem for spin down, we get the 3 columns in
> the qtl1-file. However, if we make the sum of the three contributions
> a2u+t1u+t2u and compare this result with the total 4f-DOS obtained in a
> classical way (x lapw2 -c -so -up/dn) we see a clear difference for spin
> down (I can also send these graphs if somebody is interested). For spin up
> the results are consistent. Therefore we think there is something wrong in
> reading the cf1-file.
> 
> Best regards,
> Doru Torumba
> 
> ____________________________________
> Doru-Cristian Torumba
> E-mail: Doru.Torumba at fys.kuleuven.ac.be
> Unit: Afdeling Kern- en Stralingsfysica
> Address: Afd. Kern- en Stralingsfysica, Celestijnenlaan 200D, B-3001 Leuven,
> Belgium
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> 

-- 
***********************************************************************
Dr. Jan Kunes                     email: kunes at yammer.ucdavis.edu
Department of Physics             Phone: (530) 752-0627
University of California Davis    Fax:   (530) 752-4717
Davis CA 95616
                       Home page:   http://www.fzu.cz/~kunes
***********************************************************************