[Ifeffit] doubts

Carlo Segre segre at iit.edu
Wed Nov 9 09:55:16 CST 2005

Hello Marcos:

On Wed, 9 Nov 2005, Marcos Eleoterio wrote:

> 1) After a normalization of the data, i put the data in "Athena" turn on 
> the options: "pre-edge line", "post-edge line" and then "normalized" (in 
> fact i usually turn the option "flatten" on, in dispite of the fact that 
> i don´t know what that means, what that means?), i save using the option 
> "Save chi(k)" in "file".

You can also save an Athena project fie and just that to load data into 

> 2)Then i open the program "Artemis" and insert the data saved in the 
> step "1)" in the column of the middle i click with the right button of 
> the mouse and select the option "theory", "atoms" and insert the spatial 
> group, unit cell parameter, and atomic location (here its important to 
> say that the compound that i'm studying is LaCaMnRuO6 the measure of 
> x-ray absorption was performed in the K-edge of the Mn, we are 
> interested in Mn-O ligands so in "Artemis" i only put the data of the Mn 
> and O, right?)

If you are using the known structural parameters for the material, I 
suggest that you put in all atoms.  You can restrict your fitting to the 
Mn-O first shell if you want but having the entire structure in is 
helpful.  Alternatively, if you are only interested in a quick firt shell 
fit, there are options for that too in the Theory tab.

> 3)I click in the option "run atoms" for each atoms that i inserted the 
> information, choosing 10 paths in each step, i click the in the option 
> "run feff" for each atom that i inserted the information (again), after 
> that, i click in "Fit" then appears in the screem the advise: "there is 
> something wrong with your math expressions" (or something like that) i 
> ignore that advise and go on, but the big problem is that the fit 
> appears very poor!

Since you have excluded the non Mn and O atoms, you cannot really fit the 
data in this simple way.  WIth the method you explain, only your first 
shell paths have any chance of being correct.  You will need to restrict 
your fit to those paths only (the first one or two).  Also, you will need 
to restrict the fitting range to that of the fitrt shell paths.  You 
cannot include the entire range in R-space with this model.

In general, even if you include all atoms, it is useful to start only by 
including the first shell and looking at how the data can be fit in that 
range.  I would not immediately try to fit multiple shells, you need to 
take it step by step.

> I have to say that the information that i use to insert the atomic 
> position, unit cell parameters, was obtained from a Riteveld refinament, 
> that i performed with the data from a cooper x-ray tube, do i need a 
> data from a beam line, at synchroton?

No, not at all, you just need to put all the atoms in.


Carlo U. Segre -- Professor of Physics
Associate Dean for Special Projects, Graduate College
Illinois Institute of Technology
Voice: 312.567.3498            Fax: 312.567.3494
Carlo.Segre at iit.edu    http://www.iit.edu/~segre

More information about the Ifeffit mailing list