[RMCProfile-users] Spikes in PDF when using distance window and Bragg

Fri Jan 15 09:22:10 GMT 2016

Dear Shiliang,

The mailing list software automatically removed your pictures, so can you please send them to me at helen.playford at stfc.ac.uk as I think they are an important part of your question.

Getting spikes at the edges of distance window constraints is pretty common (with and without Bragg datasets - I can't see why the Bragg data would cause this to happen) but it's concerning if you get spikes at r = 0, as this implies that your configuration starts with atoms on top of each other!

If you wouldn't mind also emailing me the files you are running with I can have a closer look.

Best wishes,
Helen

-----Original Message-----
From: rmcprofile-users [mailto:rmcprofile-users-bounces at rmcprofile.org] On Behalf Of Shiliang Zhou
Sent: 14 January 2016 19:56
To: rmcprofile-users at rmcprofile.org
Subject: [RMCProfile-users] Spikes in PDF when using distance window and Bragg

Hi all,

I have been getting spikes in my PDF model when I try to fit the PDF and Bragg datasets at the same time. The system is Fe2Mo3O12, and I have them in the order of [Fe Mo O]. The spikes usually happen at the distance window limits that I set. When I set up a distance window by including the following block in the .dat file,

DISTANCE_WINDOW ::
 > MNDIST :: 0 0 1.84 0 1.64 0
 > MXDIST :: 0 0 2.14 0 1.94 0

I got a spike at r=0 even though the starting point for the PDF to fit is at 1.61 A. (Pic #1)

When I set up the distance window by having a separate .dw file with the following cotent,

0 0 1.84 0 1.64 0
0 0 2.14 0 1.94 0

it split the Fe-O and Mo-O peaks into basically two sets of spikes at 1.84, 2.14, 1.64 and 1.94. (Pic #2)

But if I don't include the Bragg datasets, it doesn't have these issues.
(Pic #3) The Bragg fit was pretty good to start with (Pic#4). Do you have any idea what the driving force of the spikes may be?

I do find the following in the v6 EXAFS manual, which says "DISTANCE WINDOW" and "MINIMUM DISTANCES" constraints frequently cause unphysical spikes in partial PDFS at the distance limits set by these constraints....

If that's just a program bug and it coincidentally happens to me when I try to include the Bragg datasets, how exactly should I set up the restraints on the peak tails? It says to fit a fourth-order polynomial to the tail of interest. But how do you define the tail, do I just choose the data range from the onset to the highest point of the peak?

I tried including the following block in the .dat file, but the same spikes as in Pic #2 showed up.

LEFT_TAILS ::
  >START_FINISH :: 1.62 1.76 (start and end of the tail)
  >PARTIAL :: 5 (partial number 5, which is Mo-O)
  >COEFFICIENTS :: -156.617 3826.67 -25411.7 55786.8 (coefficients I obtained by fitting the PDF between 1.62 to 1.76 using Origin)
  >WEIGHT :: 0.01 (How much weight should I generally put into this?)

[image: Fe_bragg_0_2.JPG]
[image: Fe_bragg_0_dw_file.JPG]

   [image: Li_nobragg_0.JPG]
[image: bragg_fit.JPG]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20160114/81528ac9/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Fe_bragg_0_dw_file.JPG
Type: image/jpeg
Size: 38415 bytes
Desc: not available
URL: <http://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20160114/81528ac9/attachment.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Fe_bragg_0_2.JPG
Type: image/jpeg
Size: 43108 bytes
Desc: not available
URL: <http://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20160114/81528ac9/attachment-0001.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Li_nobragg_0.JPG
Type: image/jpeg
Size: 47206 bytes
Desc: not available
URL: <http://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20160114/81528ac9/attachment-0002.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bragg_fit.JPG
Type: image/jpeg
Size: 86978 bytes
Desc: not available
URL: <http://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20160114/81528ac9/attachment-0003.jpe>
_______________________________________________
rmcprofile-users mailing list
rmcprofile-users at rmcprofile.org
http://lists.rmcprofile.org/mailman/listinfo/rmcprofile-users