Hi  paul:

Have you tried to shift eo in athena by 10 ev? If you have and the fit still gives a large deltaE then the theory and the data do not agree.  DeltaE changes the spectra most strongly at low k, so another thing to try is to increase kmin in the ft. 

You can also use the theory to help remove the background in athena.  If you have tried all these and it doesn't work.  I'd try a different theory.  Either a different se oxide or a different atom type.

Of course my book chapter goes over how to do these thing. I'll send you a copy if you ask.  I've also posted example athena projects on xafs.org.

Good luck.



-- Sent from my Palm Prē


On Jun 8, 2010 5:55 PM, Paul Bingham <p.a.bingham@sheffield.ac.uk> wrote:

Thanks Bruce,

I agree that the Se-O bond lengths are short but in SeO2, and Na2SeO3 as Se4+
"standards" there are Se-O bond lengths average around 1.7A. For example when I
run feff on the SeO2 cif file it gives 1 Se-O bond at 1.623A and 2 at 1.793A,
averaging at roughly 1.7A.

I'm confident that the XANES I have for this sample show that the Se is pretty
much all present as Se4+ and I think the first shell is very likely to consist
of oxygens.

I also tried another shell from the same feff as included in the data I sent,
using O1_1 instead of O2_1 and the fit including Delr still points to an
average Se-O bond length of about 1.7A with an Enot of around 10.

I will try fitting with feffs based on different cif files with some different
Se-O bond lengths, but if you have any other suggestions I'd be grateful.

Thanks very much for your advice - it's appreciated.

Regards

Paul.
--
Dr. Paul Bingham
Postdoctoral Research Associate
Immobilisation Science Laboratory
Dept. of Engineering Materials
University of Sheffield
Mappin Street
Sheffield
S1 3JD
UK

Email: p.a.bingham@sheffield.ac.uk
Direct Line: (0114) 2225473

_______________________________________________
Ifeffit mailing list
Ifeffit@millenia.cars.aps.anl.gov
http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit