What's the difference in XRD pattern between the belite polymorphs, or there is any 2-theta window that turns it clear to identify it?
Whilst the paper suggested by Kundan Meshram is very interesting (note that is a manuscript and not the final paper - that was published in Powder Diffraction 2003, DOI: http://dx.doi.org/10.1154/1.1523079). It does not deal with the question as posed by William Fernandes, which was concerning belite, not alite.
Where the two phases co-exist it can be a challenge to identify belite polymorphs from single peaks. A better way is to use full pattern refinement which can deal with the problem of overlaps. The most intense peaks of belite are almost completely overlapped by alite.
However... here are some pointers if you wish to do it by peak identification (the old way - before we had computers!!)
- beta-C2S has a small clear peak just above 31 2-th, (the ~29.3 is overlapped, as are those around 32-33 2-th, though there is a very weak peak just below 32 that may be visible; the weak peak at ~34.4 is mostly overlapped, but may be visible).
- alpha-C2S does not have the clear peak at ~31 2-th but has a distinct peak just below 32, which should be distinct form the alite peak. It also has a distinct peak just above 33, which again should be visible - these two together should allow identification.
- alpha'-C2S again has the clear peak at just above 31 2-th, though the main peak at 23.6 will be overlapped. There is a distinct peak at ~33.4 which, together with that at 31, should be diagnostic.
- gamma-C2S main peak at 29.6 is overlapped. It has a clear peak just below 31 2-th. The main peaks around 32.5 - 33 are overlapped, however there is a very weak peak at 34.6 that is clear, and together with that just below 31 should be diagnostic.
Dear sir,
pl check this paper, i hope it will useful for you.
https://hal-univ-artois.archives-ouvertes.fr/hal-00278373/document
Whilst the paper suggested by Kundan Meshram is very interesting (note that is a manuscript and not the final paper - that was published in Powder Diffraction 2003, DOI: http://dx.doi.org/10.1154/1.1523079). It does not deal with the question as posed by William Fernandes, which was concerning belite, not alite.
Where the two phases co-exist it can be a challenge to identify belite polymorphs from single peaks. A better way is to use full pattern refinement which can deal with the problem of overlaps. The most intense peaks of belite are almost completely overlapped by alite.
However... here are some pointers if you wish to do it by peak identification (the old way - before we had computers!!)
- beta-C2S has a small clear peak just above 31 2-th, (the ~29.3 is overlapped, as are those around 32-33 2-th, though there is a very weak peak just below 32 that may be visible; the weak peak at ~34.4 is mostly overlapped, but may be visible).
- alpha-C2S does not have the clear peak at ~31 2-th but has a distinct peak just below 32, which should be distinct form the alite peak. It also has a distinct peak just above 33, which again should be visible - these two together should allow identification.
- alpha'-C2S again has the clear peak at just above 31 2-th, though the main peak at 23.6 will be overlapped. There is a distinct peak at ~33.4 which, together with that at 31, should be diagnostic.
- gamma-C2S main peak at 29.6 is overlapped. It has a clear peak just below 31 2-th. The main peaks around 32.5 - 33 are overlapped, however there is a very weak peak at 34.6 that is clear, and together with that just below 31 should be diagnostic.
Dear Kundan, thank you for your contribution.
Dear Ian, your tips are very useful for a fast (not so much!) visual identification. Thank you. I would appreciate to know as well how to choose appropriate refinement strategy and structural model.
William, for refinements I use TOPAS, I don't know what software you have but I guess it will differ. I would load all polymorph structures to start with, fix cell parameters to start with and just have a look at the visual fit of the various polymorphs to your data. You will most likely see some that do not fit at all, switch those off. Go back and refine cell parameters - be wary of any values that move to any predetermined limits, if there are no limits on cell parameters then you can get silly results. After you have decided on what phases are actually present then start looking at microstructure (size and strain), then preferred orientation. Check all parameters after each round to look for values that make no sense. If you can calculate errors, do this and watch out for any that go large. Your aim would be to try and get GoF values (Rwp/Rexp) down to near 2.
Thank you, Ian. I use TOPAS and your tips are very useful. I have some difficulty to acchieve GOF and Rwp values down to 7, it happens often when the sample has too many phases (OPC as example) and different polymorphs are present. Is it normal? Maybe I am doing something wrong. What are the equipment requirements and sample preparation to aim such good results as you told?
We always micronise samples, and side load the powder to try to avoid the preferred orientation; do not top load samples. Use extended counting times to get good peak shapes and pull out the minor phases from the background - what detector are you using? Can you post up a raw file so I can see what you are dealing with? Rwp below 8 is ok.
A lynxeye detector and front load was used. I don't have a side load available, so I will try with back fill to avoid preferred orientation. I will do a new round with this changes and then I'll post up the raw file here. Thank you.
We also have LynxEye, 3 degree opening 176 active channels, LynxIris 6.5mm, primary Soller 4 deg, secondary Soller 2.4 deg, using 0.02 step. Parallel beam Goebel mirror. 1 second per step - equivalent to 176 s total time for quants. Sample rotation 15 rpm.
To do side loading we sacrificed some of the standard PMMA holders and ground out a small part. Place a glass slide against the holder, hold it in place with big clip, have the aperture made by the ground out part uppermost, fill the powder through this and tap the whole assembly to get it to settle. A bit fiddly at first but its better than top or back loading. Micronise wet in a solvent, some use hexane for OPC.
Another thought occurred to me - have you tried selective dissolutions to concentrate the calcium silicates? Do this to get a better idea of your polymorph.
Ian, I am trying to create a side load holder by following your hint. Thanks! Here goes a RAW file of a commercial cement made with a front load sample holder. I am trying to identify and quantify appropriate phases and polymorphs. I've found the following phases (with ICSD code) with an Rwp=6,97 and GOF=2,12 :
Phase 1 : "C3S M1 Noirfontaine (2006)" 37.602 %
Phase 2 : "C2S A 81097" 6.445 %
Phase 3 : "C3A O 100220" 6.574 %
Phase 4 : "C4AF O 9197" 4.771 %
Phase 5 : "Mayenite C 241243" 1.564 %
Phase 6 : "Lime C 75785" 0.367 %
Phase 7 : "Portlandite R 15471" 0.841 %
Phase 8 : "Periclase C 104844" 3.820 %
Phase 9 : "Calcite R 79673" 1.841 %
Phase 10 : "Gypsum M 151692" 1.531 %
Phase 11 : "Hemihydrate 380286" 3.487 %
Phase 12 : "Syngenite M 157072" 2.150 %
Phase 13 : "Anhydrite I O 16382" 0.857 %
Phase 14 : "Dolomite T 31335" 1.226 %
Phase 15 : "Quartz R 41414" 0.582 %
Phase 16 : "C3S-M3 64759" 26.342 %
Can you give me your opinion?
Rietveld refinement is the way to deal with this kind of overlaps
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction