I haven't looked at the prj file, but I can say a couple of things: 1. There's no reason why metallic Fe should look like the precipitate. The binary Fe-Zr phase disgram shows several compounds, of which the most Zr-rich is Zr3Fe. This is lighter in Fe than your 40 w/o, but it's notoriously hard to quantify XRF when you're dealing with samples that aren't infinitely thick or infinitely thin. I don't know if MicroMatter makes a Zr standard, but if they do, you could use these to quantify in the thin-sample limit. 2. Overabsorption could be a problem if there's that much Fe. Similarly, the Fe-Zr mechanical mixture could be polluted by thickness/overabsorption effects if the Fe grains are more than a couple of microns across. One way around that would be to measure in electron yield, if the beamline offers it. mam On 3/1/2012 2:01 PM, adrien couet wrote:
Dear all, I am currently a PhD student working on the corrosion of zirconium alloys used in nuclear reactors. A very commonly used alloy in the industry is called Zircaloy-4 and has the following chemical formula in wt%: Zr-1.2Sn-0.2Fe-0.1Cr. The iron and chromium solubility limit in zirconium is very low. A large portion of the iron precipitates out as Zr(Cr,Fe) precipitates with approximately 40wt% of Fe, approximately 0.02wt% of Fe is left in solid solution in the zirconium matrix. Sn is completely soluble in zirconium. I am working on XANES of the Fe K-Edge absorption line in this material. To do so, I use the microbeam (0.2x0.2 microns) at the 2IDD beamline at the APS in fluorescence mode. To quantify my results, I needed proper standards of metallic Fe. I quickly realized that metallic Fe powder mixed with Zr powder did not give me the same XANES signal as the Zircaloy-4 metal (in which metallic Fe is embedded). Indeed, if you compare the "Fe powder in Zr" group to both Zircaloy-4 groups after normalization and flattening they look different. My thinking is that the Fe in precipitates and the Fe in solid solution, even if they are both metallic, could have different XANES patterns. So I decided to create a thin sample where I could probe either a precipitate or the matrix to get standards of metallic Fe in my Zircaloy-4 metal. Both signals (from Fe in a precipitate and Fe in solid solution) look similar once they are normalized and flattened (see in the attached project file). However, the data before flattening look significantly different! The post-edge of Fe in precipitates abruptly decreases whereas the post-edge of Fe in solid solution increases. I also acquired spectra in between these two extreme cases. When I scan in between the matrix and the precipitate, I see an evolution of the post edge, which looks more and more similar to the precipitate post-edge as I get closer to it. But after flattening they all look identical and the XANES spectra does not look too bad. I have some questions regarding this: - Is the flattening process changing anything such as the edge energy in a way that I should not use it when the post-edge slopes are significant as is the case here (even though my XANES spectra in solid solution and in the precipitate match after flattening)? - Have you ever seen this kind of post edge behavior before flattening? Do you think it could be due to a concentration effect since I am going from 0.02wt% of Fe in solid solution to 40wt% of Fe in the precipitate? I should mention that I have checked that no other elements present in the metal have edges that could potentially disturb the Fe post-edge. Thanks a lot for your help!!!
-- Adrien Couet Graduate Student Dept of Nuclear Engineering, Penn State University +1 814 865 9709 axc1019@psu.edu mailto:axc1019@psu.edu
_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit