questions from Bacewicz

["John J. Rehr" <jjr@leonardo.phys.washington.edu>]

Dear Rajmund,


 RB> I have been using FEFF for quite a time now and it is a basic
 RB> tool for the XANES modelling for me. Recently I have been playing
 RB> with Mn impurity in different semiconductors and one of things I
 RB> am really interested in, is a charge state of Mn atom. From LDOS
 RB> results I got little bit strange numbers for the 3d state
 RB> occupation. My question is: how reliable is the code in such a
 RB> case? (by the way: how is the zero of energy point set in a LDOS
 RB> plot?)

   I would like to clarify and add to Bruce Ravel's comments:

   First, the zero of energy in FEFF8 in the LDOS is always the vacuum energy,
which is typically several eV above the self consistently determined
Fermi energy. The Fermi energy is printed several places, e.g.
in the header of xmu.dat and the ldosnn.dat files.

    It's not clear what you mean by "strange numbers" for the 3d occupation
- we need to see some output.  The FEFF occupation numbers are typically more or
less consistent with SCF bandstructure results for the atoms without a core hole.
For the atom with a core hole, there is generally an extra screening
electron distributed over the s,p and d states (mostly s,p). 

     If the d occupation is indeed "strange" then I would suspect
1) problems with convergence of the SCF loop - how many iterations
did FEFF use, did the run finish satisfactorily? or 2) problems in the
input file (e.g. choice of lscf). We can't address this without seeing
an input file, so we always recommend you send a feff.inp along with
the version number of the code you are running so we can try to
repeat "strange" behavior.  I would want to know whether the d occupation
numbers on the neutral atoms are also strange, or just the atom with
the core hole? If you can send us this information we will check this
out.

  Cheers,
  John Rehr