Ciao Francesco, Thanks for the comment. As you've seen Ba has a sharp edge resonance which confuses the defaults for calculating phase shifts in FEFF. To tame it, add a little broadening and a finer grid: EXCHANGE 0 0 1.0 RGRID 0.01 and the problem seems to goes away. Also, the problem does not appear to be visible with FEFF8.4. Fortunately (I think) this is an unusual case. Cheers, J. Rehr On Mon, 3 Mar 2008, Francesco Giannici wrote:
Hi all,
when calculating paths around Ba (K-edge) in the cubic perovskite BaZrO3, the third path (Ba-Ba) is totally screwed up in the 2.6 - 3.8 ang^-1 range, with amplitude going up to 1000. You can find the input file at http://giannici.altervista.org/feff.inp
Actually, this is also the case of Gd:BaCeO3, substituting one Gd for one Ce, and calculating paths for the Gd K-edge: the Gd-Ba path shows the same issue. This applies to Feff8.2 on Debian, Fedora, and MacOSX 10.4.11. I found that playing around with FOLPs seems to help, but I�m not very confident with that, and I�d like to hear your opinion on what the cause could be. (Yeah, just throwing away all those low-k data would also be a solution).
On the other hand, Zr and In K-edge in Ba(In,Zr)O3 work just fine, as do Ce, Ba, Y and In K-edges in Ba(Y/In,Ce)O3. Also Gd K-edge in Gd2O3 works fine.
Cheers from Stuttgart, -- Francesco Giannici Dip. di Chimica Inorganica e Analitica Universit� di Palermo / ISMN-CNR giannici@pa.ismn.cnr.it
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit