[Ifeffit] N K-edge XANES, normalization parameters

[Barton, David (DG)]

Peter,

ATHENA is mainly written for EXAFS analysis, however...it is still quite useful for XANES analysis as well.  I opened your project that you sent and there were a couple of obvious challenges with this dataset.  First of all, you should read the section regarding Normalization in the ATHENA help POD if you haven't already.  

One useful feature in ATHENA is the ability to turn ON the Pre-edge and Post-edge plots.  After looking at your two Plots in the file you sent it is obvious that the choice of the Post-edge region is quite poor for the dataset "merge21to27".  The "Flattening" Normalization routine in Athena uses a Linear function for the pre-edge (Pre-edge Region) and a quadratic for the post-edge (Normalization Region).  Flattening is the process of subtracting these two functions from your experimental dataset (subtracting pre-edge function at energies below Eo and subtracting the post-edge function above Eo) and then scaling the dataset to unity over the normalization range.

Your dataset is a bit challenging because the pre-edge region has a limited range so that a when you fit a linear function to it, it crosses over your data set indicating that it is not a good representation of the sample background.  I can't think of a quick workaround for this, but you may want to consider manually subtracting a linear function with a smaller slope from your raw dataset prior to opening in ATHENA and then adding a couple of points to the begin of the dataset that could be used for the pre-edge region in ATHENA to force a zero slope.  Then you could look closer at your post-edge region (Normalization Range) to ensure the shape of the quadratic is reasonably fitting your dataset.  At this point ATHENA only allows a linear pre-edge and a quadratic post-edge, however it does looks like Matt and Bruce are working on adding additional fitting functions to IFEFFIT and ATHENA.  So if you cannot get a sufficient background normalization fitting with these functions!
  you will have to try some other tool.

Finally, ATHENA likes to work in eV so you should adjust your energy scale, especially if you want to attempt to use CLNORM background subtraction.

Hope this helps,
Dave

> -----Original Message-----
> From: Peter Leinweber [mailto:peter.leinweber at uni-rostock.de] 
> Sent: Monday, November 14, 2005 3:25 PM
> To: ATHENA Userlist
> Cc: fan.walley at usask.ca
> Subject: [Ifeffit] N K-edge XANES, normalization parameters
> 
> 
> dear all,
> i am a new XANES and ATHENA user. i didi N K-edge XANES of 
> soil samples. while feeling very good with data import, 
> deglitching, truncation, merging etc. i am insecure about 
> parameters for normalization. 
> in the attached project i did truncation of data before 
> 395900 (it is milli-eV), because these seem to be invalid, 
> and i set an E-shift of -900 sinc the beamline had -9.9 eV 
> off compared to the N K-edge of N2 gas. in this way i should 
> be right back on scale ? the i did set the pre-edge 
> parameters manually at start of the curve, and around 397000 
> since non of the previously measured N standard substances 
> had edge steps below this value. 
> however, then i am insecure how to set the other parameters 
> for normalization properly. depending on how i play with the 
> parameters sometimes i get normalized spectra which look 
> pretty reasonable, sometimes the curves disappear completely 
> from the graphics window, and sometimes i get curves of a 
> shape never senn in the literature. is there easily 
> understandable advice for the setting of the normalization 
> parameters available; or can anybody more experienced check 
> my project and give me recommendations. thanks, peter
> 
> Prof. Dr. Peter Leinweber
> Institute for Land Use
> Universitaet Rostock
> D-18051 Rostock
> 
> Present address:
> Department of Soil Science
> University of Saskatchewan
> 51 Campus Drive
> SK S7N 5A8
> Canada
> 
> phone: 001-306-966-6866
> fax: 001-306-966-6881
>