[Ifeffit] Low amp results

[Carlo Segre]

On Mon, 28 May 2007, Bruce Ravel wrote:

> On Monday 28 May 2007, Joji Kuniya wrote:
>> I analyzed EXAFS spectra of trace of Fe in lithium tantalate crystals. At
>> first, I created the feff.inp file of LiTaO3. Then I changed central Li
>> atom to Fe atom and added Fe atom at the Ta site. So that, the fitted
>> results is good except amp(amp results is 0.3). I think that the model
>> structure is good, but I can not understand why amp is too low. How do you
>> think about this results? And, I tried to fitting treatment using GDS
>> parameters. But, I do not know that the general method how to use the GDS
>> parameters. Please tell me the general way of GDS manipulation. My PC runs
>> under Win XP, and My Artemis runs under feff8 conditions. Thank you.
>> Joji Kuniya
>
> Hi Joji,
>
> You don't really provide enough information to give a good answer to
> your questions.  I suspect that your problem has something to do with
> the fact that the amplitude and the various sigma^2 values are highly
> correlated.  Without more information about how you are actually
> parameterizing your fits, though, it is hard to comment on that value
> of amp.
>
> I am confused that you put Fe on both the Li and Ta sites.  Is that
> actually reasonable?  Could be -- but my hunch is that Fe would prefer
> to sit on the Ta site.  Having the wrong element on a site can do bad
> things to a fit.
>

Hi Joji:

I would agree with Bruce.  In addition, I would suggest that if the Fe is 
only present in trace amounts, it is unlikley that there will be a lot of 
Fe around it.  Rather it will be surrounded by O and either Li or Ta, 
depending on which site it subsititutes on.  The first thing you need to 
figure out is where the Fe sits.  This can be done by fitting the first 
shell only to the the two different environments surrounding the Li and 
Ta and seeing which works best (all you need to see is the distance and 
the number of O atoms to really distinguish).  Once you have determined 
where things sit, you can proceed.  You may not have an amplitude problem 
then.

By the way, if you want to provide more information, you can attach your 
project file.  There are usually a number of people who will just take a 
look and give you advice.

Carlo


-- 
Carlo U. Segre -- Professor of Physics
Associate Dean for Special Projects, Graduate College
Illinois Institute of Technology
Voice: 312.567.3498            Fax: 312.567.3494
segre at iit.edu    http://www.iit.edu/~segre    segre at debian.org