I just wanted to add my 2 cents to this argument...
I think one big point in all the discussion on size and strain concerns
the difference between what IS in the specimen, what we see with our
probe (X-rays or neutrons, presumably) and what we reconstruct using A
model. In most cases the model do not answer the easy question: "what is
in the specimen?"
I want to stress the "A" because in any case what we get is just a
guess... subtle philosophers could speculate on this.. but I bet we're all
scientists and not mere philosophers..
Anyway, people start their analysis with simple models and try to improve
them.. easy but effective!
As far as I know (well microelectronics is reality I think!) there
is NO connection between grain shape (and therefore crystallite shape) and
symmetry (or any descriptor for it).
In this case the ellipsoid model could work, but is completely missing
reality! The good scientist, however knows the limits of validity and the
hypotheses on which the model is based (most modern scientists tend to
forget this concept...) and knows that he obtained some "effective fit".
Ok, specimens we analyse are simpler but... we are not dealing with
specimens containing a set of perfect, equal, ordered, aligned crytallites..
More likely we have a distribution of shapes, sizes and orientation
that can screw things up! In this case we can just hope that some simple
model will accommodate all the mess! And using ANY anisotropic model is in
most cases better than using none if you just care of "good fit".
As for symmetry restrictions.. well.. they are welcome if they are
consistent with the nature of diffraction (peak overlapping is always a
painful problem in line profile analysis), but they are just related to
our probe (X-rays, neutrones) and NOT with the original crystallites!!!!!
Of course music changes if we talk about strain broadening.
That's why Stephens models is good as an effective way of treating
anisotropic broadening because "it fits better", but I'd not attach any
physical meaning to the numbers you get out of it...
I have written a bit too much I bet... so better if I go back to my
size/strain modelling!
OOps I was forgetting... Armel's replies to Nicolae emails are really
great (you could be a good boxeur, Armel!), but I do not agree on one
point:
>Yes, anisotropic line broadening is rarely observed
>with cubic compounds unless in very special cases
>of faulting.
this is true if you restrict the scope to size broadening only.
Otherwise, line defects can be a source of anisotropic brodening in
cubic materials (but you need a "good amount" to appreciate the effect)...
Mat
PS. There is no unique and simple solution to the problem... there are
just scientists with their ability to attach the proper meaning to their
results!
--
wwwww
g( o 0 )g
--oOO--(_)---OOo---------------------------------------oOO-wwwww-OOo-------
Department of Materials Engineering
and Industrial Technologies
University of Trento
38050 Mesiano (TN)
Matteo Leoni, PhD ITALY
.ooo0 0ooo. Tel +39 0461 882416 e-mail: [EMAIL PROTECTED]
( ) ( ) Fax +39 0461 881977
----\ (---) /--------------------------------------------------------------
\_) (_/ | | | |
.ooo0 0ooo.
( ) ( )
\ ( ) /
\_) (_/
