We recently discussed problems related to lattice parameter determination with EBSD and the missing accuracy in comparison to other techniques, especially XRD. I pointed out that many phases, e.g. minerals, have a wide quite range of lattice parameters so that the question arises, how accurate lattice parameters have to be in order to find the corresponding phase (or an isostructural) in a database. Which tolerance you are using if you try to find a phase using powder XRD? There are certainly people who are using SAED in TEM to identify phases. How big are the acceptable deviations there? What are the criterias then for a reliable phase identification?
Is there a ISO norm, how phase identification has to perform?
Dear Prof. Gert Nolze
Have a nice time and good day
Yes, the only one which gives an accurate lattice parameter is XRD technique. Therefore, when you determine it by any other way You will find differences among the obtained values and those obtained by XRD.
But about what are the tolerance at the boundary between the different phases of the materials, I don't know
Good luck
Dear Ahmed, I think you have missed the point of Gert's very interesting question. If you read carefully you will see he says that XRD has greater accuracy than, for example, EBSD. This is not in question.
The point is about tolerances of lattice parameters when performing searches to identify minerals. Minerals display many 'solid solution series' and there is often a continuum of compositions from end members which mineralogists place arbitrary names on according to boundaries defined by the chemical composition. We see these names enshrined in the databases. In powder diffraction it becomes difficult sometimes to determine exactly where one phase ends and another begins due to subtly varying lattice parameters as result of substitutions in the chemistry.
I suppose the answer could be found by creating simulated structures of idealised compositions. But again what should the tolerance at the boundary be? A long time ago I wrote some software for identification of opaque minerals using reflectance data and I came upon a very similar problem. That code is long gone so I can't remember what I did exactly. I need to do some searching and have a think about it - good question!
Dear Gert Nolze
Dear Ian J Slipper
Yes really .. I have read the question again and found that Prof. Gert raises a good question. The question really needs to be answered from Researchers who are specialized in crystallography. I hope to find a definite answer to this question and what are the tolerance at the boundary between the different phases.
Thanks
Dear Giovanni Ferraris, yes, you are right. In SEM no problem (EDS), but for powder diffraction not automatically available. I guess the best technique is wet chemistry (in case of powder)? It might be not easy for a mixture of phases already grinned. Anyway, the higher the accuracy of lattice parameter the higher the probability that you can identify it. The only excuse are minor phases lower that about 2%...if they are essential for the material. For technical materials they are often important since they weaken the entire material. This I understood as primary task for techniques used in TEM or SEM. But there commonly the accuracy is not that high (except of Kossel diffraction) but this is no simple technique as well (for several reasons). Micro-X-ray diffraction is a promising alternative...but also then one has to search in databases.
Well, let's assume two cases... a) you know roughly the chemistry, and b) you don't have any idea. Which deviations you would enter? Which is your typical strategy?
Dear Gert Nolze, in recent years the identification programs are feeded with complete XRPD patterns either as a list of interplanar spacings or (better) as a profile of the pattern. I have not a real experience in that, but I am quite sure that if you perform a search with only cell parameters, without any other type of information, the risk of mis-identification is high, particularly if the allowed tolerance is higher tna 2%.
International Centre for Diffraction Data is one of the methods I recommend to identify crystalline phases.
Isn't the ICDD a company celling the PDF databases? How this can be a method? Or do they have some kind of recommendations, how to identify reliably phases?
http://www.icdd.com/products/2015SalesCatalog.pdf
Gert, I have been using BGMN-Profex recently and I noticed a setting that may feed into your question. In the BGMN Preferences section is a "Limits" page and it has Default Upper and Lower Refinement limits set at 1% for unit cell parameters. This value is adjustable but it could be useful for you.
Hi Ian, I guess this limit describes the maximum deviation of the lattice parameters of the candidate file during refinement? As far as I know BGMN or Autoquant, they don't use a free-available or commercial structure database but their "own" confirmed phases. You can add your own if you discovered some reliable phase description. BGMN also does not search in databases, doesn't it?
Thanks!
Yes, that's what I meant. A phase search one has to do with the acquisition software. I also guess that programs like Fullprof and MAUD do not offer the acquisition part as well. They are all only for post-processing which excludes any phase search. But their focus I also different...
There are several good systems to detect a phase into a solid material. One of them is produced by ICDD.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science