[Ifeffit] Peak at short distance in FT - YbVO4 nanoparticles

[Bruce Ravel]

On 05/29/2015 11:54 AM, Raj kumar wrote:
> Recently, i have performed EXAFS for YbVO4 bulk (prepared through
> sintering process) and nanoparticles (at room temperature).
> Nanoparticles were synthesized by co-precipitation technique and their
> EXAFS was measured by passing the prepared solution using peristaltic
> pump. After all subtraction (background and solvent), nanoparticles XAS
> data were recovered and processed for the fitting. After Fourier
> transform, it was observed that both nanoparticles posses an additional
> short peak (or shoulder) at about 1.2 A from bulk. Since the peak
> appears much closer to 1 A, i suspect this may due to artifact or bad
> background subtraction. Hence, i would like to know the way to eliminate
> this peak at short distance.

Hi Raj,

A few comments:

1. Your bulk sample (the first one in the list) may have a bit of Hf
    in it it.  It's hard to tell, but the mu(E) seems to be trending
    upward right at the end of the data, which is the approximate edge
    energy of Hf L3.  I got a slightly better background removal by
    changing the end of the spline range to 600.  But this a minor
    detail.  The real story seems to be that ....

2. You data are pretty noisy.  If you plot chi(R) out to 10 Angstroms
    (which is, presumably, well beyond where you should expect to see
    good signal in your data), you get a sense of how the level of
    noise manifests itself in chi(R).  To my eye, the peak/shoulder at
    about 1.2 that you are asking about is of about the same size as
    the level of noise in your data.

You are asking how to get rid of a feature in the data that is
aesthetically unpleasing.  I don't think that's the right question.

Your data are what they are.  Your data are noisy and measured over a
rather short energy range.  That doesn't give you a lot of choices
about how aesthetically appealing you can make the Fourier transform
be.

You seem to be operating Athena correctly and there aren't any magic
buttons that will make your short, noisy data longer or less noisy.  I
don't see anything obviously wrong in how you are processing your data
and -- short of using PhotoShop ;) -- I don't see any defensible way of
making the feature at 1.2 go away.



Am I correct in thinking that these data are measured in the
dispersive geometry?  If so, my suggestion is to reconsider how you
make your samples.  This kind of experiment is VERY sensitive to the
homogeneity of the sample.  You have to take extraordinary care to make
your samples as homogeneous as possible.  I suspect most of the
glitchy little points in your data are indicative of sample
inhomogeneity.

You also should worry about purity of your source material.  Granted,
it canbe very difficult to obtain rare earth metal samples that are
elementally pure.  But if your data range is short because of other
edges, then you need samples that are more pure.

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/