[Ifeffit] How to distinguish whether the coordination element is heavy or light

[Bruce Ravel]

On 07/04/2014 03:51 AM, ZHAN Fei wrote:
> In the picture you recommended,it says "The variations in functional
> form allow Z to be determined (±5or so) from analysis of the
> EXAFS".But I don't find any publication use it.

Hi Zhanfei,

The figure I referred you to was a plot of the scattering amplitude (the 
bottom panel had the phase shift) for three different scattering 
elements.  I don't know what's plotted in the figure you attached 
because there are no labels on the axes.  In any case, the scattering 
amplitude is the F(k) term in the EXAFS equation.  As such, it's 
relationship to chi(R) or chi(q) is subtle.  And, of course, in real 
data, all the different scattering paths interfere with one another.

As for the Z+/-5 rule, I don't know who first stated that.  Perhaps Teo 
and Lee...?

In any case, it is easy to test.  Measure, say, a NiO standard.  Try 
replacing O in the feff.inp by N or F and do the analysis.  Try again 
with C or Ne.  Try again with B or Na.  You will find that the fit using 
F is basically indistinguishable of the fit with O.  Ne will be a bit 
worse, but not much.  Na a bit worse, but not much.  Eventually, you 
will get far enough away from O that you can clearly see the difference 
in the fit.

Some years ago, I tried to work on a FeGa alloy.  Although 5 apart, I 
could not distinguish Fe scatterers from Ga scatterers well enough to 
say anything about how the dopant was distributed in the lattice.


> I simply try an
> example.I choose the second shell of MoS2 which is the Mo-Mo shell in
> R-space in Athena,then do the back FT to get amplitude of the second
> coordination shell,then I compare to amp described in paper
> http://pubs.acs.org/doi/abs/10.1021/ja00505a003. The peak is at the
> almost same wave number when q-space use k2
> weight(when use higher kweight,q space peak  shifts to high
> wavenumber ),see pic attached. I only get a peak without the vally
> described in paper when use kweight,but can see a vally without k
> weight,maybe stress the contribution in high k make the vally
> obscure.Can this method works in element determination correct to +-5
> Z number?
>
> And is there any method to determine whether a coordination peak in R
> space have one element contribution or  more? For example ([Ifeffit]
> path contribution to fit in low R-space position, but the fit bond
> length is much longer than that ):a cluster we expected it has both
> Ni-O Ni-S(normally Ni-S peak is in high position),and when spectrum
> has a peak between regular Ni-O and Ni-S and a shouler near Ni-O,can
> I ensure the expection of both Ni-O Ni-S？OR just maybe because of the
> multiplicity of EXAFS

You seem to be looking for a magic wand that you can wave at your data 
and have a clear answer pop out.  EXAFS analysis, sadly, isn't like 
that.  The information content of the data is quite limited, the data 
range is usually quite limited, structural and chemical disorder make 
the EXAFS signal hard to interpret.  Scott's discussion that you 
referred to in a previous email is just a tool to help disentangle all 
these problems ... it's not a solution.

EXAFS analysis does not "solve structures".  There is no mathematical 
operation that can somehow "invert" EXAFS data.  The best we can do is 
test models against real data and do a statistical analysis of the 
results.  In the end, the best we can ever say is whether a fitting 
model is *consistent* with data.  Or perhaps, whether one model is *more 
consistent* with data than another.

In your case, if I understand your description, you need to test models 
with O scatterers, with S scatterers, and with different mixtures. 
Hopefully, some of these models will be more successful than others. 
And hopefully, the successful models will be consistent with other data 
you have about your samples.

O and S are an interesting case.  If you have Feff compute Ni-O and Ni-S 
scatterers at the same distance, then plot chi(k), you sill see that 
they oscillate out of phase over much of the k-range.  This is both 
great and problematic.  It is great because it means that the contrast 
between the two is about as big as it can be.  Ni-O with only a small 
amount of S will be measurable because the presence of the S serves to 
significantly reduce the amplitude of chi in the data.  However, if you 
have similar amounts of O and S, you may be in the situation where the 
overall chi(k) has a very small amplitude due to two things mostly 
cancelling each other out.  As a result, parameters such as sigma^2 for 
the different scatterers will be highly correlated.

So, the "too-long;didn't-read" version of this is that you have to 
simply try all the different, reasonable fitting models and decide what 
works best.

HTH,
B


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

  National Institute of Standards and Technology
  Synchrotron Science Group at NSLS --- Beamlines U7A, X24A, X23A2
  Building 535A
  Upton NY, 11973

  Homepage:    http://xafs.org/BruceRavel
  Software:    https://github.com/bruceravel