Hi MJacob,

One note that might help--it is not strictly necessary to fit the nearest-neighbors in order to fit the iron-iron path. In a system like this, I would think you might get lucky and have a big contribution from near-neighbors, then a bunch of smeared out stuff (as Matthew Marcus suggested) that effectively cancels, and then a little iron-iron peak because the iron is a much better scatterer than all the organics. If that's what the Fourier transform looks like, then you can fit from say 4 to 5 angstroms and just use an iron path. It sounds weird, but I've done things like that once or twice with decent results. There are certainly pitfalls; for instance I might be a little nervous about some focused MS path that happens to line up with the Fe-Fe distorting the results, but it's worth a shot.

--Scott Calvin
Sarah Lawrence College

Could you perhaps post an image of the Fourier transform of your data to the list? That would help give a sense of what might work and what wouldn't work.

On Mar 11, 2009, at 5:49 PM, Frenkel, Anatoly wrote:

I would like to be proven wrong but my feeling is, without even knowing anything about this particular system, that one cannot observe Fe-Fe pair at that distance in such complex case.
However, to approach it more rigorously, the first step would be to examine the relative "importance" factors of different paths in the output of FEFF. If Fe-Fe path is well isolated in distance and importance from others (in reality, it probably is not the case) it can be done, otherwise - not.

Anatoly


From: ifeffit-bounces@millenia.cars.aps.anl.gov <ifeffit-bounces@millenia.cars.aps.anl.gov>
To: ifeffit@millenia.cars.aps.anl.gov <ifeffit@millenia.cars.aps.anl.gov>
Sent: Wed Mar 11 09:40:31 2009
Subject: [Ifeffit] (no subject)

Hi all,

I measured Fe(II) edge EXAFS data for my protein that has two iron sites (Fe-Fe distance ~5A). My question is how to fit the model with the data? Can I discard the contribution of multiple scattering effects to the fitting? One most important information that I want to obtain is on the Fe-Fe distance. Is it enough to conclude from the raw data FT that the Fe-Fe  distance in my protein sample has been changed in comparison to the Fe-Fe distance in the model (created with Feff using the crystal structure coordinates)? I try fitting using arthemis but got bad and weird results. is there any manual or step-by step guide that I can learn how to fit model of two metal sites to my data?

Anther question that I have how can I get to the past discussions in the forum- is there any database for questions and answers that I can learn on the analysis of EXAFS data?

Any help will be appreciated,

MJacob

 ars.aps.anl.gov
http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit