[Wien] DOS and XANES and cahge density

[Peter Blaha]

> I corrected the parameters NUME=3000 in param.inc of lapw1.
> Do I have to copile Wine2k again?

Yes.


> x lapw2 -all XX YY   (where XX YY specifies the E-range)
> 
> However, if YY is larger that certain value,
> (2.0 Ry for AmO2, 1.9 Ry for CaF2)
> the error message is made as
> 
> L2main - QTL-B Error

The "LINEARIZED" APW method uses an simple approach with du/de to describe
the E-variation of the basisfunctions.
Of course such a linearization is good ONLY in a LIMITED energy range,
usually in the valence states.

When you use eigenvalues up to 70 eV, be aware that they COULD eventually
be very bad (or "wrong") because of this.
WIEN2k dedects in a charge density calculation if a contribution from the
u-dot term (the "BLM") is too large and stops with "QTL-B Error", because
such states are usually not very accurate (or even unphysical).

Thus, first of all check for which atom and angular momentum the QTL-B error
occurs (see case.output2).
Then you may have to adjust the energy parameters, or, eventually add a
local orbital of the specific character at higher energies.

70 eV is VERY large and carefull testing of the results with various
energy parameters and adding LOs as much as possible is necessary.

-- 

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