[Wien] Zero DOS

[Peter Blaha]

The answer was clearly given before:

You have only ONE k-point ! One can clearly see this from your first mail, where the
"band-limits" are identical.

 BAND LIMITS OF BAND  1 ARE   1.67240   1.67240
 BAND LIMITS OF BAND  2 ARE   1.67314   1.67314

Thus when you calculate a DOS at a certain energy (eg. 1.67300) it has zero DOS, only if
you would happen to hit EXACTLY an eigenvalue, you would get a "delta-function" peak.

For the DOS using TETRA you need more than 1 k-point!

There is no program which can plot a DOS with one k-point in WIEN2k. You have to
write it yourself, eg. use a "broadening method" and broaden each eigenvalue; or
a "histogramm method", i.e. count yow many eigenvalues are in a certain E-interval.

--------------------------------
 em... I found 4 kpoints in IBZ.
in scf:
Pb:
:E2_0001: E( 2)=   -0.8650   E(BOTTOM)=   -0.940   E(TOP)=   -0.790
             APW+lo
:E2_0001: E( 2)=    0.3000
             LOCAL ORBITAL
:E0_0001: E( 0)=    0.3000
             APW+lo
:E1_0001: E( 1)=    0.3000 ---? is this incorrect?
             APW+lo
 				
But this output has nothing to do with k-points !!! ??????

You find the number of k-points written in the scf-file or simply look into
case.klist, or "remember" what you specified during initialization for x kgen.

-------------------
May be I find the reason: Pb-5p semi-core is not include! so I try to use -7.6Ry to seperate the core and valence. Hope it can work.

This has nothing to do with the DOS ????
And most likely, it is NOT a good idea to include that as valence state.
---------------

So in essence:
x kgen    and specify a number which gives you a few k-points.
x lapw1 (-c)
x lapw2 -qtl (-c)
x tetra

The "quality" of the DOS will depend on your k-point number.

PPS: Often it is absolutely fine to do the scf / or min_lapw with just ONE k-point,
but for the DOS you need more.


lilonghua schrieb:
>  the case.qtl is correct. So I think the error may occure at  tetra.f!
> it doesn't work well when total bands exceed nbtot=1000! but after nbtot=2000, DOS still zero! 
> Can somebody who have read the code of tetra give me some suggestions? thanks.
> 
> ------------------				 
> lilonghua
> 2009-05-20
> 
> -------------------------------------------------------------
> 发件人：lilonghua
> 发送日期：2009-05-20 09:54:17
> 收件人：wien at zeus.theochem.tuwien.ac.at
> 抄送：
> 主题：Re:Zero DOS
> 
> It doesn't work!! the DOS still zero even include the Pb-5p. Anybody can help me?
> 
> 2009-05-20 
> 
> 
> 
> lilonghua 
> 
> _______________________________________________Wien mailing listWien at zeus.theochem.tuwien.ac.athttp://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 

                                      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/
--------------------------------------------------------------------------