>I have a problem refining neutron data on MnSe due to stacking faults. Can
>any one tell me how to account for stacking faults and which refinement
>program to use. I have refine the chemical as well as magnetic cell.
To my knowledge, there is no Rietveld software able to face
automatically any stacking fault and/or antiphase domain case.
Predigested "simple" cases for fcc, bcc and hcp metals do not explain all.
The program that would absorb any data and say : "Hey you have
reflections with h + k =2n + 1 that are especially broadened
if l is different from 0, but don't worry, the program will handle
it and give you clear explanations about the phenomenon" (that
case exists, with a fit by the Rietveld method, see J. Phys. C, Solid
State Phys. 19, 1986, 4623-4633).
The old good method is to first identify the broadening law(s), and
then try to adapt more or less well a Rietveld program to it. The best
would be to not only fit with a phenomenological model, but also
to get probabilities of faulting, mean sizes, also taking account of
instrumental broadening, etc.
Literature is rather scarce but growing. I could find 45 papers with
"Rietveld" and "stacking" and "fault*" or "antiphas*".
There is a challenging and unexplained school case example in Powbase
(cubic HNbO3), you may get the neutron data at :
http://sdpd.univ-lemans.fr/powbase/index.html
See a part of the pattern with unbroadened and broadened reflections
at : http://sdpd.univ-lemans.fr/powbase/31.gif
Eveybody can fit that pattern more or less well by using several
groups of reflections with different broadening. But who can fit it
with a stacking fault or antiphase domain model with a Rietveld program ???
Something that would really explain what happens physically inside
the sample ??? Not me, I have only an idea about the phenomenon...
Armel Le Bail
http://www.cristal.org/
