Follow up Experimental Broadening question

[David Eustace <daveus@chem.gla.ac.uk>]

Hi all,

I have a very short follow up to my recent question about using the 
EXCHANGE/CORRECTIONS cards to broaden FEFF XANES spectra in a peak fitting 
process.

We've recently been investigating the DOS and XANES of several oxide spinel 
samples, and have found that with some samples the best fit to experimental 
data is given by 1.2eV imaginary broadening, as described by Prof. Rehr in 
his response to my last email.

Experimental broadening in the data we're comparing against is ~ 0.5eV. 
This amount of imaginary broadening does not give good agreement - too much 
fine structure remains.

Searching the literature, I have found that in the published work 
"Real-space multiple-scattering calculation and interpretation of XANES" 
(Phys. Rev. B 58, 1998) - Ankudinov, Ravel, Rehr and Conradson, cubic BN 
XANES had been broadened 1.3 eV, and UF6 DOS broadened 2.0 eV.

So my 1.2eV broadening seems reasonable (?).

Can the experimental and imaginary broadening values be directly compared?

If so, does that indicate my calculation results are incorrect?

If not, is it reasonable to broaden a spectra 1.2eV with FEFF when you know 
experimental broadening can only be 0.5eV (and also - why?) ?

Thanks for any response, and any indication of what amount of broadening 
(if any) other users regularly encounter.

Cheers,

David Eustace