Re: The EXCHANGE potential

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

Hi Ming

> I used the Feff8 to calculate the LDOS and absorption spectrum of the
> ZnO (L23 edge spectrum). In this case, I found when I chose the
> different potential in the exchange card, there are obviously different
> morphology in the absorption spectrum. I had tried to find some detail
> interpretations for these potential to understand what is best choice in
> each case. However, I only found the interpretation and comparison of
> HL, Dirac-Hara and Xa potential. Someone can give some information or
> reference of these potential, (0,1,2,3,5) ? 

  These exchange potentials refer to different ways of including the
photoelectron self-energy (which is a dynamically screened exchange
interaction). These are described in various references on the FEFF
code as mentioned in the doc (e.g. Appendix C of the FEFF8 doc): 

  0 HL is the default - recommended for solids
  1 DH is the Dirac Hara - sometimes better than HL,
   especially in molecules where HL is too lossy
  2 Ground State - recommended only for near edge where 
        HL or DH may be questionable
  3 DH real part + HL imaginary part - an ad hoc way of adding
       loss to DH - not at all justified
  5 Dirac-Fock for core states; HL for valence
     an approach for improving on HL which includes some non-local exchange

   None is fully satisfactory in all cases, but the FEFF project 
is working to improve the self-energy algorithms.

   Cheers,
   John Rehr