[Ifeffit] Normalization of XANES spectra

Scott Calvin SCalvin at slc.edu
Wed Oct 29 11:32:12 CDT 2008


Hi Jens,

I got lost in your example, somehow (I think you mixed up some letters  
and numbers), but I think I get the gist.

I would use the qualitative comparison of normalized spectrum to find  
abundances with considerable caution. Peaks can change size for a  
variety of reasons.

It's much better to use linear combination fitting to determine  
relative abundances if you have appropriate standards. If you don't  
have appropriate standards, but do have a series of samples, then pca  
can help confirm that there isn't any funny business going on.  
(Changes in peak size that are due to factors other than relative  
composition would tend to show up as a third significant component in  
pca.)

--Scott Calvin
Sarah Lawrence College

On Oct 29, 2008, at 10:02 AM, Jens Kruse wrote:

> Hi everyone,
>
> I have a general question:
> I know Normalizing is something we do so that we can compare samples
> measured
> under different experimental conditions (removes the effect of  
> different
> gains ...) but I have problems in understanding and applying
> normalization when also quantitative conclusions are needed when
> comparing different spectra.
>
> If I have two XANES spectra (a) and (b) measure under the same
> conditions but with different concentration of the absorbing atom with
> in different molecules.If both normalized spectra show a peak 1 , but
> this peak is higher in intensity in spectrum (a) than in spectrum (b),
> does this mean that compound which produced peak 1 is also absolute
> higher in concentration in sample/spectrum (a) or do I loose this
> information after normalization?
> Or can I just say: in the normalized spectrum (a) peak 1 is higher  
> than
> peak 2 suggesting that the compound leading to peak 1 is more abundant
> (just relative proportions).But I can't say: The intensity of peak 1  
> in
> the also normalized spectrum 2 is lower than in spectrum one,  
> Therefore,
> the compound leading to peak 1 is less abundant in spectrum/sample  
> 2. ????
>
>
> I hope somebody can help me to understand this issue.
>
> Thanks a lot,
>
> jens
>
> -- 
> Jens Kruse
> Institute for Land Use
> Faculty for Agricultural and Environmental Sciences
> Rostock University
> Justus-von-Liebig-Weg 6
> 18059 Rostock GERMANY
> Phone: +49(0)381-498 3190
>
> _______________________________________________
> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
>




More information about the Ifeffit mailing list