rowlesmr
Hi all
I've got a few diffraction patterns that have a distinctly asymmetric peak shape. All peaks have the same asymmetry. It isn't the instrument; the sample later anneals and becomes symmetric.
I can model it using several different cell edges, and the scale factors come out as vaguely log-normal. (with all other prms identical between the strs). THe phase is cubic.
Searching for "log-normal peak shape" and other such things just leads me to symmetric peaks modelled with log-normal crystallite size distributions.
THe samples were ball-milled. I've looked up some of the WPPM type modelling, but it doesn't seem to handle asymmetric peaks.
Has anybody got any ideas? I've attached some screen shots of the peaks (17.3 and 50.4 deg 2T).
mfisch
Check the Simple_Axial_Model (circles_conv) macro and the Clay.inp example (exp_conv_const).
Some spherical harmonics approach using one of the convolutions above?
rowlesmr
Hi Martin
I've already tried all the convolutions and couldn't get a satisfactory fit. The SH approach doesn't work as the peaks aren't anisotropic, just asymmetric.
And having a convolution-based solution in this instance doesn't feel right, as it is a sample-related broadening that changes upon heating, so I'd like to try and describe it physically.
I can _sort_ of make it work by having two strs, with different cell edges and strain/size prms.
I'll keep plugging away.
Matthew
mfisch
I notice from your plot that your additional cells are actually smaller, meaning that probably something goes out during milling, depending on the compound (hydrated or similar?). Moreover, you can produce quite distinct particle/cryst size fractions in ball mills. Did you try to vary the energy of the mill? In particlular compare high E, short t vs. long t, low E.
Quite different and I didn't read it but check this: DOI 10.1016/j.matchar.2015.09.025
lgermann
Hi Matthew,
When I first saw your attached figure it reminded me a lot about a discussion with my supervisor few days ago about stacking faults. He mentioned that stacking faults my can cause asymmetric peak shapes (his drawings looked pretty much the same as your figure). Not sure if it makes sense with your structure?
cheers,
Luzia
rowlesmr
Thanks all for your suggestions.
Stacking faults did cross my mind, and it's on my reading list for next week
This is the first time we've actually noticed this broadening; the lab data doesn't show it unless you're looking for it. The funny thing is that these samples have already been annealed, but it seems that it didn't work. So we're currently looking at everything.
davewragg
This looks more like two different unit cells than stacking faults, though I'd have to take a look at the full profile to give a more reliable opinion. As Mfisch suggests, this is quite possible with a ball milled sample (though it's also possible to create stacking faults by milling!).
We've observed two different unit cells in the same crystallite in the past for coked zeolites (the coke only penetrates the outer part of the crystallites):
doi:10.1016/j.micromeso.2013.02.022
If you could post the full profile I could show it to my colleague who is the real stacking fault expert...
Cheers,
Dave