In the same situation as yours I used both cards, Polarization and Ellipticity. If you use only Ellipticity card, without Polarization, you get spherically averaged XAFS which will not take into account texture you are concerned about. If you use only Polarization card, without Ellipticity, you will assume linear polarization which is guaranteed to work correctly only if your polarization is along the surface normal to the film but may not work if it is within the plane of the film since the film is rotated. Depending on the in-plane symmetry (3-fold or higher), linear polarization will give the same result as circular (you can check that for your particular structure) but it is more safe to use circular (besides, is just one extra card, so why not to use it, just to make sure). Anatoly ________________________________________ From: ifeffit-bounces@millenia.cars.aps.anl.gov [ifeffit-bounces@millenia.cars.aps.anl.gov] on behalf of xu dongbin [binbinmirror@gmail.com] Sent: Sunday, May 13, 2012 4:06 AM To: ifeffit@millenia.cars.aps.anl.gov Subject: [Ifeffit] Question: About using polarization and ellipticity cards in FEFF Hi, I would like to ask a question which has confused me for a few days. In my current case, I'm dealing with samples with c texture in the film normal. The sample was measured with linear polarized x-ray in both directions with the surface of sample parallel/perpendicular to the electrical field of x-ray. During measurement, the sample was always rotated around the film normal. Therefore, I think the parallel direction data can be treated with random orientated grains in the film plane. In order to describe this, I need use POLARIZATION and ELLIPTICITY cards in FEFF calculation. However, I'm not sure which case below is the correct one I should include in FEFF calculation for my case: a. POLARIZATION 1 1 0 ELLIPTICITY 1 0 0 1 b. ELLIPTICITY 1 0 0 1 I think these two cases should give the same calculation results, however, when I sum the paths in these two cases respectively using the same parameters(S02, E0, delR, sigma^2). The results give substantial difference. Can anyone help me to figure out this? Thanks&Regards! _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit