[Ifeffit] Possible bug in fitting

[Matt Newville]

Hi Ditty,


On Wed, Apr 30, 2014 at 7:05 AM, Ditty Dixon <ditty.dixon at gmail.com> wrote:

>
>  Hello IFEFFIT users,
>
>
>
> I am working on LiNi0.5Mn1.5O4 cathode materials, I am pretty sure that
> some of you already worked on it. On the first look, a multiple edge
> analysis on both Ni and Mn is interesting; however it’s not that straight
> forward due to multiple phase and oxidation states. I started doing
> multiple edge fit, but I gave up briefly. Then I did individual edge
> analysis and obtained some values. Looking at both edges my Ni fit looks
> good. So I thought of taking the Ni-Mn bond distance as well as DW factor
> value to Mn edge and set this value for my Mn-Ni paths. However it seems
> like this value is further evaluated in the fit and giving a new value. Can
> someone reproduce this, please see the project file and the Log file
> attached, Is it a possible bug or am I doing something wrong?
>



To me, it looks like you've set the Mn-Ni distance and sigma2, and these
values are used and reported as you set them.    You probably meant to set
Delta_R for the Mn-Ni path, or use "RMn-Ni - reff" for the Delta_R
parameter for the Mn-Ni path.  As it is, you told it to use ~2.92 Ang as
Delta_R for the Mn-Ni path, which will get added to the starting distance.

--Matt
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://millenia.cars.aps.anl.gov/pipermail/ifeffit/attachments/20140430/c04a632c/attachment.html>