[Ifeffit] Several Questions

Bruce Ravel bravel at anl.gov
Fri Apr 20 12:17:16 CDT 2007

On Wednesday 18 April 2007, Edward L. Kunkes wrote:
> Hello everyone,
>  I am currently using IFEFFIT to fit LIII edge data collected for  Pt, Re
>  and PtRe bi-metallic Nan particles supported on carbon, and I have several
> questions:

> 1. I get an So2 value of 0.54 from a first coordination shell fit on
> Re-foil (this So2 value also seems to work when fitting NH4ReO4-)
> Has anyone ever worked with Re before and gotten a similar value for
> So2 ?

That seems small.  What is sigma^2?

> 2. I model the path for the a single coordination shell as 6
> octahedrally arranged nearest neighbor atoms surrounding the
> absorber atom at a distance equal to that of inter -atomic spacing
> in the metal. When comparing the IFEFFIT predicted EXAFS for this
> model to one that is generated by using the first path from a
> crystallographic model, the two seem slightly different.  The
> nanoparticles are very small (amorphous to XRD), and probably
> consist of a disordered phase, so I cannot assign a crystal
> structure. What should I do in this case?

There is a sizable literature on exafs and nanoparticles, including a
number of excellent papers by folks whose names appear regularly on
this mailing list.  Reading some of those would certainly help.

Your nanoparticles, like most other exafs research problems, resemble
something for which you can find crystallography data, buit aren't
exactly like the pure, ordered material.  Because one has to start
somewhere, the pure ordered material is as good a place to start as
any.  The trick is to figure out how your system differs from the well
ordered system.  

It is highly likely that the first coordination shell in your
nanoparticles strongly resembles the bulk material, although there
might be some variability in average coordination, average interatomic
disatnce, or sigma^2.  The second coordination shell might deviate
from the bulk a little bit more and the third a little more.  Anatoly
Frenkel, Scott Calvin, and Shelly Kelly all have papers on this

> 3. I often get high correlations the coordination number and ss, or
> delEo and dR (the correlations are ~0.8). Using multiple k-weights
> doesn?t decrease the correlations. What is an acceptable value for
> these correlations?

Coordination and sigma^2 will usually be highly correlated -- they
both affect the amplitude of chi.  Similarly e0 and deltaR will
usually be highly correlated because they both effect the phase.  That
is the nature of the exafs fitting problem.  A correlation of 0.8 for
either of those pairs is quite common.

Without more information, it is hard to improve upon that situation.
One common approach to dealing with these correlations is to do
something like measure data at various temparastures.  As long as your
material does not change phase, you expect coordination and e0 to
remain (mostly) constant with temperature which sigma^2 and delatR
follow predicatable behavior.  Measuring data at multiple edges is
another excellent strategy when applicable.


 Bruce Ravel  ---------------------------------------------- bravel at anl.gov

 Molecular Environmental Science Group, Building 203, Room E-165
 MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007

 Argonne National Laboratory         phone and voice mail: (1) 630 252 5033
 Argonne IL 60439, USA                                fax: (1) 630 252 9793

 My homepage:    http://cars9.uchicago.edu/~ravel 
 EXAFS software: http://cars9.uchicago.edu/~ravel/software/

More information about the Ifeffit mailing list