I have recorded X-ray diffraction (XRD) patterns of the oxide samples (sintered ceramic/pellet) using Cu-Kα radiation (λ = 0.154 nm). The sample contains two different phases. I want to quantify the amount of each phase using Rietveld refinement. Following are the two way I think I can proceed:
1. Rietveld refinement using the XRD patterns recorded from the ceramic/pellet surface.
2. Or Rietveld refinement using the XRD patterns recorded from the annealed powders (powders obtained after the grounding of pellet using agate mortar-pestle).
What would be the correct/right way to proceed? Any help or suggestion would be much appreciated.
I would prefer taking the XRD from powdered sample material. So you avoid any artifacts in the pattern due to preferred orientation issues.
But why do you want to anneal your powder?
Ok, perhapes for reduction of residual stress due to grinding operation...
But please be sure to have no phase transition while annealing...
Gerhard Martens Thank you, Sir, for your suggestion. If the phase transformation takes place on grounding, Is it then justified to work with the XRD data recorded from the ceramic surface? Can I use the refined data for publication?
Yes you can. The penetration depth is some µm; so you are not only inspecting the 'surface' but also a huge amount of bulk material. However if the peak height distributions of the phases do not resemble those from a powder equivalent (e.g. own made references of phases or from literature) then you suffer from preferred orientation. But I do not know how the refinement will deal with such a situation.
Other colleagues may comment on that ...
'Refined data from 'surface' (= non-powdered) XRD for publication': why not; it is just the result from your measurement...
It should be possible to go the refinement with both measurements. Possible texturing effects might be resolved by using an adequate software like GSAS 2 or MAUD, if you are dealing with sharp textures. The latter should not be the case in standard sintering, so hopefully you don’t have to bother. In any case, keep in mind to refine the texture in the end of your process to avoid artefacts. The comparison of the refinement of your powder measurement with the sintered one might even be a powerful tool to examine possible phase shifts during your process.
The publication can of course be used for publication if you describe the measurement and treatment properly.
Dear Kundan,
Shortly speaking, you can do the Rietveld refinement directly from the surface of ceramic sample., as what I usually do. Attached is one of my samples, i.e. the Rietveld refinement on the XRD pattern of zinc-doped magnesium titanate ceramic after the pellet was sintered at 1400 C for 8 h. As seen, the fitting is very good. The most important thing before doing refinement is that all the crystalline phases on the tested sample must be identified well and the crystallographic data (powder diffraction file database, etc) occupied to create spectral model under the software you use to do quantitative analysis is accurate/high quality.
Good luck.
You will be able to do Rietveld phase composition analysis using XRD data from a solid ceramic or metal sample or crystalline organic polymers without having to grind the sample. If you were to examine the surface of the sample with appropriate electron microscopy, you would see a distribution of surface grains which diffraction-wise provide a nice XRD pattern as does a powder sample.
The same issues apply regarding systematics errors - crystallite size and strain, extinction, preferred orientation, etc.
And we would normally require the sample to have 'flatness' less than the surface 'bumps' caused by granularity. You may wish to polish the surface of the sample if surface roughness is problematic.
And the sample must be 'infinitely thick', but thinner samples may be handled with due care.
Good luck.
Brian O'Connor Thank you, Sir, for your valuable suggestions and comments.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Powders