[Ifeffit] Re: EXAFS question

[Bruce Ravel]

Norman,

Your question is the sort of thing that is appropriate for the Ifeffit
mailing list, so I am CCing my answer there.  You should consider
joining the mailing list.  It is a great resource and gives you access
to a large number of experts who can answer your questions.


On Friday 23 July 2004 03:38 pm, Norman White wrote:
NW> In going through Matt Neville's xas_fun_slides.pdf file that was
NW> included with the Athena software, I noticed on slide 36:

 MN > The Fe-O distance in FeO is 2.14 ° A, but the first peak is at 1.6 ° A.
 MN > This shift in the first peak is due to the phase-shift,
 MN > delta(k): sin[2kR + delta(k)] .
 MN > A shift of -0.5 ° A is typical.


NW> I asked him in an e-mail about the shift and he said (edited):

 MN > For traditional XAFS analysis, we do a relatively simple Fourier
 MN > transform.  This simple FT would give a peak in chi(r) at 'R' if
 MN > the oscillatory part was sin[2kR].  Since it's sin[2kR + delta(k)]
 MN > instead, there's a shift.
 MN > 
 MN > It turns out that, to first order, delta(k) is linear and has a
 MN > negative slope of about 1:
 MN >     delta(k) ~= -k ~= -2k (0.5),
 MN > 
 MN > so that the shift in phase is about 0.5A.

NW> I asked if this was true for all elements and he said:

 MN > It's a typical value for most bonds, but shouldn't be taken as an
 MN > accurate number (and it's easy enough to calculate).

NW>
NW> How is an accurate shift calculated?  Is this correction performed

This calculation is computed quite accurately by Feff.  Indeed,
getting that calculation (and the related calculation of the
scattering amplitude) correct is the raison d'etre of Feff.

NW> automatically in Athena?  If it is a constant, why is it not simply
NW> included in the step producing the radial distribution function from
NW> the k values? Can it be accurately calculated?  The manual had
NW> statements that seemed to imply that it could be calculated with the
NW> Atoms program but did not explain how at a level I understood.
NW>
NW> Matt neville stated in an e-mail that:
 MN > The phase shift correction can be made fairly easy with
 MN > SixPack and with Artemis as well as Ifeffit.

It's not a constant and treating it as such would be specious at best,
and it is not something that Athena intends to compute.  However,
since it is something that Feff computes, one can imagine using the
results of Feff.  How this is done is slightly different in Athena as
in Artemis.

At the level of Athena, we know the absorbing atom but it is not
actually necessary to know what the material is to use Athena
effectively.  In Artemis, on the other hand, we do have to know
something about what the materials is.  This difference between the
two explains the rationale for how phase correction works i the two
programs. 

In Athena, only a central atom phase correction is supported.  This is
enabled by clicking the button that says "phase correction" in the
forward trasform section of the mail window.  With that button
clicked, the central atom phase shift is subtracted from the data
before the transform is made.  When you plot chi(R), then, the
transform will be peaked a little bit to the right of where it was
peaked without that button clicked.

Artemis does a somewhat better job of this.  In Artemis, you can
select a path to use for pase correction.  With a path selected, the
central *and* scattering atom phase shifts will be subtracted before
transforming.  This is possible in Artemis because we have the fuill
information about the material and the full feff calculation at our
disposal.  In Athena, in contrast, we only know about the absorber and
I have never implemented (and don't really intend to do so) a way of
reading in a feff path for the purpose of phase shifted transforms. 

NW>
NW> I could not find where it explained how to do this.

Did you read any of the documentation?

For Athena:  
http://leonardo.phys.washington.edu/~ravel/software/exafs/doc/Athena/athena_misc.html
  The document is also available within Athena via the Help menu.

For Artemis:  Section 5.8.5 of
http://leonardo.phys.washington.edu/~ravel/software/exafs/doc/artemis.pdf 


NW> I am looking at As adsorption on Fe-Al oxides and other minerals
NW> I know what the central atom is As.
NW> I can also estimate the oxygen bondlengths and the As-cation distance from
NW> the "corrected" EXAFS pattern and the literature. Is it possible to
NW> simulate an EXAFs pattern?
NW> Is so, how?

Yes.

Have you read the Artemis manual?

B



-- 
 Bruce Ravel  ----------------------------------- ravel at phys.washington.edu
 Code 6134, Building 3, Room 405
 Naval Research Laboratory                          phone: (1) 202 767 2268
 Washington DC 20375, USA                             fax: (1) 202 767 4642

 NRL Synchrotron Radiation Consortium (NRL-SRC)
 Beamlines X11a, X11b, X23b
 National Synchrotron Light Source
 Brookhaven National Laboratory, Upton, NY 11973

 My homepage:    http://feff.phys.washington.edu/~ravel 
 EXAFS software: http://feff.phys.washington.edu/~ravel/software/exafs/