The short, short version, in my estimation, is "pick two paths as a reasonable starting point and let the fitting parameters do the rest." You could even use two clones of the same path, which will end up getting different CNs, delr, and ss in the fit.
There's no law that says you have to use Feff paths that correspond to paths in your structural model (if an analogous path even exists). As long as it's close-ish to start (the program will throw a warning if it's not close enough), the fitting parameters can be used to handle small differences arising from e.g., octahedral vs tetrahedral coordination.
You can run two calculations and import the paths and scale the contributions from each site. There are a lot of examples and this has been discussed before on the mailing list if you search.
Hope that helps,
Chris
*********************************************
Dr. Chris Patridge
Assistant Professor
Department of Chemistry
SASE 315
D’Youville College
320 Porter Ave
Buffalo, NY 14201
716-829-8096
Hi
I am running an EXAFS fitting for spinel Co3O4. As we know, the Co3O4 will have two different two local coordinations. One is 6 Co-O coordinated octahedral site, and another one is 4 Co-O coordinated tetrahedral site. When we run the Feff calculation, we can only use one Co site to do the calculation once, and we can only get one site coordination by one calculation. However, the EXAFS will tell us the average local structure. How could we use these two different calculation paths to do the fitting?
Chemical Engineering
Oregon State University
541-602-1707