[Ifeffit] FEFF6L input keyword question

[Bruce Ravel]

Hi Zin,

No one bit on this, so I'll respond.

If you are using feff6l, then I presume you are interested in
simulating or analyzing exafs rather than doing xanes calculation.  If
you are more interested in XANES, then my comments in this email won't
make a lot of sense or, indeed, be particularly useful.  And, if you
are interested in the XANES, you should be using feff8 or feff9 rather
than feff6.

I have to question your decision to simulate the effect of sigma^2 at
the level of feff.  In my experience, you can do a much better job by
using ifeffit and/or artemis to model the effect of sigma^2.  The
exact same correlated debye function is available in ifeffit/artemis,
so at the very least you can do the same thing as what you seem to
want to do.  But ifeffit/artemis is way better because it gives you
MUCH better control over how sigma^2 gets applied to the paths.

Strictly speaking, the correlated Debye model is only applicable to a
monoatomic material.  That is not to say it is not a useful
approximation in other materials, just that it is not strictly
correct.  In practice, you'll find that different kinds of scatterers
need to be treated differently in a real exafs problem.
Ifeffit/artemis gives you the tools to do so. Feff6l, by itself, does
not.

As for question #3, a trivial internet search will turn up the Feff6
document:  http://lmgtfy.com/?q=feff6+document

In the document you find, you see that there is an ION keyword that
claims to do what you want.  Use this with extreme caution.  In my
experience, it has surprising and often unphysical results on the
calculation.  This isn't so surprising.  In the exafs region, the
photoelectron has enough kinetic energy that it is not so sensitive to
the details of the potential surface.  EXAFS really is an atomic
structure tool -- while the XANES certainly tells you a lot about the
local electronic structure, the exafs signal is mostly sensitive to
where the atoms are and what kind they are.

So, in short, each of your questions makes me worry that you are
heading down the primrose path.

B

On Wednesday, March 02, 2011 05:31:35 pm YOON Zin Seok wrote:
> Hi, I am a beginner of FEFF6L. Usually depending on the manual provided by
> web site, I am managing to calculate spectrum, and there are too many
> difficulties...I would like to be advised from you, senior members.
> 
> I have three questions.
> (1) I would like to control the temperature. In the manual, I see
> 
> DEBYE 190 315 (Calculate Debye-Waller factors for Cu at 190 K??).
> 
> Can I change the temperature simply by modifying the number "190"? Can I
> find a reference paper for this?
> 
> (2) Where can I find Debye-Waller factors in the output files?
> 
> (3) How can I change the ionic state? I would like to compare the output
> file, chi dat., between SF6 and SF6+ molecules.
> 
> I hope I could have answers for my questions...

-- 

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

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

 My homepage:    http://xafs.org/BruceRavel
 EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/