Hi Shiju,
    It is a nice coincidence that I am going to collect some Mo K edge EXAFS data at beamline X18B in NSLS starting this Thursday :-))  Your approach sounds fine to me in general except that you can set the sigma2 values of your unknown samples to the values obtained from your standard only if your standard is a very good representation of your unknown sample. Else, it might be hard to get the correct coordination numbers in absolute sense. By fixing the sigma2 values to the one obtained from standards, you can however reliably obtain the relative change in the coordination number of an unknown samples series. So this could be one among several other reasons for not getting a value close to 6, when you *know* it should be 6. What happens if you do not fix the sigma2 of your unknown sample? Do you get a reasonable fit with slightly higher sigma2 but correct CN? Are the CN and sigma2 values highly correlated? If you are measuring a series of samples, you could try doing a simultaneous fit (of the series of samples each at three different k weight) - by floating the sigma2 of each path for one sample and fixing it to that value for the rest of the samples. This approach does not always work, but is worth a try.

 Wrong choice of Eo, and/or improper energy calibration could lead to similar problems too. Moreover, if the sample is highly concentrated, and the data was collected in fluorescence mode at 45 degree incidence, self absorption effect could very well be the likely cause.

 Without looking at the data, I cannot comment anything more than that. But if you are not able to resolve these issues in next few days, I can take a quick look at your data as I would be playing with Mo K edge EXAFS anyways. 

Hope That Helps,
Bhoopesh


Research Associate
Geosciences Department
Princeton University, NJ 08544


 

On 4/15/07, Shiju N. R. <shiju_nr@yahoo.co.uk> wrote:
Hi,
I am Shiju and I work as a research associate at the Chemical Engineering department of University of Cincinnati. We are working on the conversion of light alkanes using mixed oxide catalysts. We are trying to understand the structure of a Mo based mixed oxide catalyst that is highly active for this reaction using XAFS and have some data of this catalyst, collected at NSLS.
Since I am new to XAFS, I am trying to learn it myself.  I am using Athena and Artemis. I am trying to use MoO3 as a model compound for understanding the Mo local structure in my catalyst. By following Scott Calvin's ZnO example (which is highly helpful for beginners like me), I was able to get a reasonable fit for MoO3 structure. For the next step (fitting my Mo catalyst), I would like to seek help regarding the procedure I should follow. I would like to describe my understanding to approach the fitting of my Mo catalyst and please let me know if I am right or wrong. I am going to take same MoO3 structure to form the theory and import the same paths which were used to fit model MoO3 structure. Then I will set the values of S02 and sigma2 to that obtained for MoO3 and allow E0 and delr to vary during the fitting. Please let me know if this is right approach.
 
I tried once this way, also introducing the concept suggested by Bruce Ravel to find the coordination in one of the earlier posts. I used the expression S02 = ni*amp, where i represents each path, keeping the degeneracy of each path as 1. I was able to get a fit, however, ni are fractions and their sum is much less than 6, while I am expecting an octahedral coordination of Mo in my catalyst. Is it due to any mistake in my approach? what should I do to correct it?
 
Please excuse if these questions seem silly... I am just a beginner in XAFS!
 
Shiju


Dr. Shiju. N. R.

Research associate

Chemical and Material Engineering

University of Cincinnati

Ohio

USA

Yahoo! Mail is the world's favourite email. Don't settle for less, sign up for your free account today.


_______________________________________________
Ifeffit mailing list
Ifeffit@millenia.cars.aps.anl.gov
http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit