I have been trying to index the planes from an XRD pattern. But the 2theta value of the nanoparticles happens to match with various other phases of the same sample. What is the correct method to find out the planes from JCPDS software?
It is tough to do an indexing on a nanocrystalline powder, especially from lab data. What you do, however, seems a search match. not an indexing.
The peaks can be displaced wth respect to their expected positions for several reasons, some of them real (e.g. surface relaxation, non crystallograpic pahses), some of them fake (e.g. wrong analysis tools.. https://www.researchgate.net/publication/230824264_On_the_modelling_of_the_powder_pattern_from_a_nanocrystalline_material).
If you are using the ICDD software (JCPDS has changed its name into ICDD long time ago!! http://ww.icdd.com), you can simulate the effect of a small domain size, but you should always know in advance what can be inside (chemically) to limit the changes of mismatch.
Once you have the candidates, a possibile solution is usually modelling the pattern.
You can try to model the pattern using the partial information that you have (the various pahses have different intensity ratio), using appropriate microstructure models (to account for the broadening). In that case, I am sorry but (traditional) Rietveld refinement (as suggested) might not be the best tool as it works in the wrong space (see the ref I gave you above) and usually has very poor microstrcture models. You can try the WPPM we developed, that can give more accurate information (being based on physical models for the microstructure).
If you are in cases where the size is too small and multiple solutions cannot be separated, then you should use a alrge fraction of reciprocal space for the analysis and try to get more info via e.g. pair distribution function (PDF).
Ideally, combinging real space (PDF) and reciprocal space (WPPM) should solve your problem of identifying (and characterizing) your dispersed nanoparticles.
Article On the modelling of the powder pattern from a nanocrystalline material
It is tough to do an indexing on a nanocrystalline powder, especially from lab data. What you do, however, seems a search match. not an indexing.
The peaks can be displaced wth respect to their expected positions for several reasons, some of them real (e.g. surface relaxation, non crystallograpic pahses), some of them fake (e.g. wrong analysis tools.. https://www.researchgate.net/publication/230824264_On_the_modelling_of_the_powder_pattern_from_a_nanocrystalline_material).
If you are using the ICDD software (JCPDS has changed its name into ICDD long time ago!! http://ww.icdd.com), you can simulate the effect of a small domain size, but you should always know in advance what can be inside (chemically) to limit the changes of mismatch.
Once you have the candidates, a possibile solution is usually modelling the pattern.
You can try to model the pattern using the partial information that you have (the various pahses have different intensity ratio), using appropriate microstructure models (to account for the broadening). In that case, I am sorry but (traditional) Rietveld refinement (as suggested) might not be the best tool as it works in the wrong space (see the ref I gave you above) and usually has very poor microstrcture models. You can try the WPPM we developed, that can give more accurate information (being based on physical models for the microstructure).
If you are in cases where the size is too small and multiple solutions cannot be separated, then you should use a alrge fraction of reciprocal space for the analysis and try to get more info via e.g. pair distribution function (PDF).
Ideally, combinging real space (PDF) and reciprocal space (WPPM) should solve your problem of identifying (and characterizing) your dispersed nanoparticles.
Article On the modelling of the powder pattern from a nanocrystalline material
To determine the hkl reflections of a powder diagram is necessary to determine the parameters of the unit cell using the programs ITO, INDEX and DICVOL. The reliability of the unit cell and the indexed reflections is indicated in these diagrams by a series of figures of merit, usually indicated in these programs or by information published in articles.
In my opinion the question should be reformulated more clearly. In particular, the sentence “But the 2theta value of the nanoparticles happens to match with various other phases of the same sample” seems to mean that the sample is polycrystalline, i.e. it contains more than one crystalline phase. In this case it could be actually impossible to index the diffraction pattern.
In my view, finding index plane is tedious calculations from our xrd data. But you can calculate if you know its crystal system. You will get details from any solid state Chemistry/Physics book or googling too for calculation of planes. You mentioned "match with various other phases" means there is either more than one phase (already mentioned by Sir Ferraris) or impurity (Defect). I think you should repeat the experiment if you are not looking for multiphase in your NPs. I guess JCPDS/ICDD contains only single phase plane in one PDF card.
It is difficult to make indexing on a nanocrystalline powder.
The peaks can be displaced with respect to their expected positions for several reasons, some of them real (e.g. surface relaxation, non crystallograpic pahses), some of them is wrong
If you are using the ICDD software (JCPDS ) you can simulate the effect of a small domain size, but you should always know in advance what can be inside (chemically) to limit the changes of mismatch.
Once you have the possibilities, a possibile solution is usually modelling the pattern.
You can try to model the pattern using the partial information that you have (the different pahses have different intensity ratio), using a suitable microstructure models (to acI am socount for the broadening). In that case, Rietveld refinement (as suggested) might not be the best tool as it works in the wrong space and usually has very poor microstrcture models.
If you are in cases where the size is too small and multiple solutions cannot be separated, then you should use a large fraction of reciprocal space for the analysis a
Ideally, combinging real space (PDF) and reciprocal space (WPPM) should solve your problem of identifying your nanoparticles.
Dear Hatem, I am curious: can i just ask why you propose some approximate cut and paste version of my previous answer?
Yup, I am agree @Matteo Leoni. Concept can be same but words can not be same.
Another interpretation is "complete agreement". I see it as a compliment! That's good. Right?
I see a lot of expertise flying around, yet the interpretation of the conventional diffractogram is still so nebulous. Just out of curiosity, any of you Nano particle/agglomerate experts try this stuff in 2D. How come none of you "heroes" show or use the 2D XRD data or method to "really" characterize the material? What's the compulsion to wring this century old (dating back to the Braggs) conventional method further? What's the matter? Why the 'stone age" mentality? Poor SNR? Don't forget the JCPDS (old) or ICDD (new) are both compilation of conventional equatorial scans (diffractograms) that may be totally outdated and are virtually "blind" to the micro lattice strain components and "preferred orientation". Only tremendous mathematical machinations (Scherrer method, Warren-Averbach, Rachinger etc.) would yield some reasonable Nano structural parameters. Never-the-less, ambivalence in interpretation is still a challenge. "Known" calibration standards (internal/external) would ease the "pain" of interpretation dramatically in such cases.
Kakoli! The best way to leverage the tremendous resources (expertise) in the RG network is to fully describe your paradigm accurately and provide as much supplementary (SEM, TEM, Literature) information as possible. Please post your data along with details of your sample/material composition and characteristics. When you have folks like Matteo interested in your discussion, I suggest a "full court press". Expose the ignorance and eradicate! Meaning, provide as much information as possible to get the best possible & most complete answer. Please post your data and share your methodology thus far.
In principle the WPPM approach would be the best with the limitations imposed by the conventional diffractogram. With this approach the integrated intensity and any related "data normalization" would work better in order to correctly interpret your XRD pattern.
Matteo, I need "cart blanch" to plagiarize you at will. I promise to acknowledge you when doing so. Because the truth may not sound as good if words were changed. I'm sure your students take pride in using your thoughts verbatim but many of those thoughts that you espouse may have originated elsewhere as well. Your description is awesome! I pay close attention to each word.
Matteo! "It is tough to do an indexing on a nanocrystalline powder, especially from lab data". Why after a century is this still so? It ought to be easy & straight forward!
Why can't XRD be used just as GC and with similar certitude & precision? Why all the ambiguity still? XRD ought to be unequivocal!
Chief Protagonist, 2D Real Time Bragg XRD Microscopy
http://www.youtube.com/watch?v=IU0m4yI7D-k&list=PL7032E2DAF1F3941F
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction
- XRD Analysis