I am a very newbie in this Rietveld analysis. I am trying to fit my XRD pattern using (for a polycrystalline film) a CIF file. I know I am off stochiometry and I am sure my cell is deformed. Using MAUD I cant obtain a decent fit (huge discrepancies on a and c cell param and in atom site occupancy) so my first guess is I am doing something terrible wrong. My question is a Rietveld analysis can be performed for a file acquired using a Detector scan? (with a fix tube)
It always makes sense to read a bit about the technique applied. There are several books which are focused on full pattern refinement or Rietveld refinement if the crystal structure is not clear and should be determined. The applied least-square fit is only that good as the experimental data are. The better the signal to noise ration the more reliable the least square fit. The mathematic behind ALWAYS find a solution but it does not mean it is correct. Especially if there are many other parameters like experimental uncrtaities. In your case, even with no noise the result is meaningless since you do not have a highly overdetermined equation system which is the strength of this technique. I can only recomment to read first a few papers or better a book about the technique.
In your case the gracing incident is an additional challenge. I am not familiar whether it has been applied for such a geometry which has even more specific features and sources of errors but I would guess that somebody already did it. The only problem is, if people already fail with a common geometry, the probability is clearly higher for a situation which is less investigated. Try it but don't be surprised if it does not work properly. The refinement also need a lot of experience which is related to the code and also the parameters shown as criteria for an acceptable/ reliable / trustful fit. Again: If the code is well written, mathematics will result in a number. But you cannot simply believe this values since it has been calculated by a complex software which has a lot of pre-conditions which are either assumed to be obvious (like sample replacement or texture effects), or which are used as simplifications fo the calculation. Many things are also displayed as MODEL and they are in first approximation for a typical geometry acceptable but perhaps not for your material.
How homogeneous is your layer or phase? Use as many as other information from other techniques. If there is any inconsistency something (or everything) is wrong... Always doubt! Software should help you to save time but it shall not and cannot replace your competence! Finally, as long as you don't understand only a single parameter shown or used in the software, everything is meaningless.
Dear Petru,
From the three given files, it is difficult to understand exactly what you made. Please, could you decribe the experimental part (I mean the diffraction one) and explain step by step the Rietveld analysis you performed. To start Rietveld analysis with such data collected in grazing incidence is not ideal for a neophyte. Have a look on the interesting paper by Simeone et al. (J. Appl. Cryst. (2011). 44, 1205–1210) concerning Rietveld and grazing incidence.
Best regards,
Jean-Louis
Hello Petru,
Having looked at your pattern (.xy file), it is evident that there is little information of value in the pattern - just three, very weak, broad peaks. You would need to measure data with much better statistics. As it stands, there is no point in thinking about Rietveld optimisation when the measured data is so poor. In my experience, you would do much better by measuring grazing indigence SRD data.
Regards.
Brian O'Connor
Thank you all for your answers. This enlightened me a lot on the topic. My initial toughts about trying Rietveld on my data come after I saw a paper with similar data (or worste) doing that (see atached). I will try to get more precise data.
Voila
http://pubs.rsc.org/en/Content/ArticleLanding/2016/TC/C5TC03161C#!divAbstract
Hello Petru, the paper does not mention Rietveld refinement at all. For the XRD only qualitative assessment and the Scherrer equation have been used. They don't mention grazing incidence XRD either. The XRR is a separate procedure for thickness measurement. Or have I missed the point of your original question?
Petru, find the result from the attachment (open the file using ur MAUD). Strong preffered orientation due to epitaxial growth.
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction
- XRD Analysis