[Ifeffit] Fit a mixture

[Bruce Ravel]

On 11/27/2014 04:13 AM, kaziz sameh wrote:
> Dear All,
> I look for advice or document that can help me to fit the FT EXAFS data
> of Pd2Au36(SR)24.
> The situation is as follows:
> In the sample there is 3 sites of the Pd ( in staples, in surface or in
> the core). the spectrum that I have presents  a first bond of Pd_S and
> second bond of Pd_Au, so the palladium should be in the staples or in
> the surface because in core we cannot see Pd-S, but other results of the
> same sample confirm that the palladium is in core.
> I concluded that there is a mixture of some nano (x ) with Pd in staple
> and another nano(y) with Pd in core  such y>x.
> Note that I fit the data with Pd2Au36(SR)24 structure in which I put the
> 2 palladium in the staple and I have got a good fit (attached) but I
> think that is not logical.
> If you can help me with an idea how to fit this mixture, I would be very
> grateful.

Kaziz,

I don't really understand your question.  You use a lot of jargon.  I
don't know what a "staple" is this context.  I don't know what "SR"
means in this context.  I don't know what "x" and "y" are or how they
might enter into your analysis.  And I do not know anything about the
nature of the sample you are measuring or what you did to measure it.

When you ask a question in a forum like this mailing list, it is
helpful to remember that those reading your question know a lot about
XAS, but probably don't know anything about your specific research
problem.  You need to ask a question that can be answered in that
context.

Reading between the lines, you seem to have a problem that is similar,
at least conceptually, to the question that Pushkar Shejwalker asked
last night.  That is, you seem to be anticipating one thing, but
coming to a different conclusion in your EXAFS analysis.

That happens sometimes.  In fact, that can be the exciting part of
doing research.  If our measurements always turned out exactly the way
we anticipate, then the synchrotron would be a much more boring place.

You need to first ask yourself if your expectation is merited.  What
is the evidence from other measurements supporting your expectation?
Have you been very careful to control the state of your sample
throughout its entire lifetime?  Is there any chance that the sample
could be damaged by exposure to the beam?

If you have convinced yourself that the sample is in the state you
expect, then you need to evaluate the result of your EXAFS analysis.
Is the R factor small?  Are all the fitting parameters of values that
make sense physically?  Are the error bars small?  If you are finding
the sample in a different state that you anticipate, do you think that
the results you obtained are defensible?  If so, then maybe you have
discovered something new about your sample!  If not, then you need to
critically evaluate all the steps of the experiment all the way from
sample prep until pressing the fit button in Artemis.

I encourage you to, as Matt suggested, figure out how to ask a more
targeted question.  Broad, open-ended questions are hard for the rest
of us to answer.  Well-focused questions tend to get better answers.

HTH,
B

-- 
  Bruce Ravel  ------------------------------------ bravel at bnl.gov

  National Institute of Standards and Technology
  Synchrotron Science Group at NSLS-II
  Building 535A
  Upton NY, 11973

  Homepage:    http://bruceravel.github.io/home/
  Software:    https://github.com/bruceravel
  Demeter:     http://bruceravel.github.io/demeter/