[Ifeffit] Question on GDS values: amp

Bruce Ravel bravel at bnl.gov
Tue Aug 2 06:37:22 CDT 2011


Nic,
I think that you have rediscovered something many of us have found before you. If your fit is wonky -- an E0 of >50 for instance -- that usually means you have made a mistake in your starting structure. 

In this case, there was a problem in the crystal data. Sonetimes one presumes, say, a sulfide when the sample is actually oxidized. Sometimes one makes the absorption edge mistake I suggested.

The moral is that, if your fits are wonky, that is usually telling you something about your initial assumptions. To say that another way, a screwy fit is usually useful information and a hint that you need to reexamine your work carefully.

I also want to point out that when you described your problem in words, you got a bunch of interesring, but ultimately unhelpful guesses. When you provided enough information that someone else could reproduce the problem, Dominic was able to provide you the right answer.

That's a lesson for everyone reading this email!

B
-- 


Dominik Samuelis <d.samuelis at fkf.mpg.de> wrote:

Dear Nic,

I've had a short look into your Artemis file. First of all, it seems 
that something's wrong with the Gd2O3 structure data in your feff.inp.

Running feff on your feff.inp gives this error:

Feff 8.40
...
WARNING: TWO ATOMS VERY CLOSE TOGETHER. CHECK INPUT.
atoms 8 14
8 2.24848E+00 2.70250E+00 0.00000E+00
14 2.47549E+00 2.70250E+00 0.00000E+00
Run continues in case you really meant it.
WARNING: TWO ATOMS VERY CLOSE TOGETHER. CHECK INPUT.
atoms 9 15
9 -2.24848E+00 -2.70250E+00 0.00000E+00
15 -2.47549E+00 -2.70250E+00 0.00000E+00
Run continues in case you really meant it.
[...]

It seems that all your "Gd2" atoms have a duplicate in your "Gd3" atoms, 
shifted only by ~0.2 angstroms in either x,y, or z direction.

Looking further at your feff.inp, already your ATOMS input somehow looks 
broken:

(your ATOMS list)
* Gd 0.25000 0.25000 0.25000 Gd1 1.00000
* Gd 0.02100 0.00000 0.25000 Gd2 1.00000
* Gd 0.54200 0.00000 0.25000 Gd3 1.00000
* O 0.12500 0.12500 0.37500 O1 1.00000
* O 0.12500 0.37500 0.37500 O2

(my list, cif file from ICSD, Shuenn et al., J. Chin. Chem. Soc. 57 
(2010), 938)

Gd	0.25	0.25	0.25	Gd1	1.000		* (8b position)
O	0.3904	0.1513	0.379	O1	1.000		* (48e position)
Gd	0.9710	0	0.25	Gd2	1.000		* (24d position)

Using this crystal data, I was able to fit your data reasonably well 
(didn't tweak to make the fit perfect though...) - see attached file!

I've also attached the .cif file that I've used, in case you want to 
compare with yours.

BTW, enot values of >20 eV shouldn't really happen in any fit.

Best regards,
Dominik


On 02.08.2011 04:42, Nicholas.Tse at csiro.au wrote:
> Thank you all the suggestions.
>
> I believe the ATOM model used is representative of what it should be. I generated the FEFF file from the CIF file from the American Mineralogist X'stal DB. The generated FEFF has the edge L3 that should be fine. A simple summation of the scatter paths does roughly produce the pattern for the sample, except for the amplitudes.
>
> The enot was found to be 53.8995 +/- 2.7632.
>
> I have also included the .prj and .apj files for your inspection.
>
> Regards,
>
> Nic
>
>
>
>
>_____________________________________________

> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit


-- 
Dr. Dominik Samuelis
d.samuelis at fkf.mpg.de
Max-Planck-Institut für Festkörperforschung
Max Planck Institute for Solid State Research
Heisenbergstr. 1
70569 Stuttgart
Germany
Phone +49-711-689-1769
Fax +49-711-689-1722
Web http://www.fkf.mpg.de/maier/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://millenia.cars.aps.anl.gov/pipermail/ifeffit/attachments/20110802/15296f20/attachment.html>


More information about the Ifeffit mailing list