Hi everyone!
I am doing Rietveld refinement of synchrotron diffractograms of a nanopowder using GSAS-II.
When I refine the size of my sample, I refine it with the isotropic model (obtaining lets say 0,050 um), but then after many tests I have seen that I could use uniaxial size model choosing the crystallographic direction of the most intense peak and I obtain a much better fit of the experimental data (obtaining lets say equatorial size = 0,045 and axial size = 0,055 um), both for intensities and profile shapes.
Now I have a few questions on that model, and I struggle to find useful information on GSAS-II documentation or old GSAS manual, where there is just the mathematical formula.
What does choosing that particular [hkl] direction imply for the "real sample"? Does it mean that the sample has a shape that exposes mostly that crystallographic direction?
How do the two size values translate to the sample? I know size is an average value, and with isotropic model I could say that the average size is 50 nm, but the real sample should be better described by non-spherical shapes.
As far as I understand, the unique axis relates the orientation of your uniaxial domain size you're modelling to the orientation of your crystallographic structure defined in your CIF file, or whatever file you've used to describe your crystal system.
For example, if you have a needle-like grain that is a fully coherent scattering domain and say, is rotationally symmetrical along its long axis, this long axis may correspond to the c-axis of your material. If so, then you would put 001 in your refinement model. If it corresponds to your a-axis, you would put 100, and if it corresponds to a direction between your a and b axes, you would put 110.
It doesn't mean that the unique axis corresponds to the longest axis. You could have an oblate spheroid shaped domain size.
It's important to remember that grain size does not directly equal scattering domain. You could have a structure with very unequal lattice parameters, and so even if you have a spherical particle, the number of coherent planes for scattering you have could be much lower due to them simply having more spacing between them. Graphite is an obvious example of this, where the basal planes have a much greater spacing than the prismatic ones, and so while you can have spherical powder grains of graphite, its scattering domain is uniaxial. This is my understanding anyhow. Sorry it's 2.5 years late, hope you found your answer before this, and if you have any recent ideas on this, I'd love to hear them.
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Journalism