> >>
> >> Dear All,
> >>
> >> I am trying to carry out Rietveld refinement on a sample that has
> >> very obvious peak shape variation; probably due to subcell and
> >> supercell peaks. I believe there is a way to deal with this in GSAS,
> >> but I can't quite work out the best way.
> >>
> >> I'd be grateful for any tips
> >>
> >> Caroline
> >>
> >>
> >> Caroline Kirk
> >
> ........
> >This may not be helpful to your specified query but if you have a Bruker
> >XRD with Topas Rietveld software - Topas has an option to apply spherical
> >harmonics not only to Preferred orientation, but also anisotropic
> >peak broadening, peak displacements (e.g., due to stress/strain) and
> >some other parameters as well.
> >
> >       http://www.bruker-axs.com
> >
> ...............
> >
> >(I am not sure what other Rietveld programs allow this as well?)
> >Lachlan M. D. Cranswick
> 
> 
> The program Maud (see my web pages) also models anisotropic peak broadening
> using the Popa (see on J. Appl. Cryst. of 1998, May-June?) model. It's an
> advanced harmonic expansion method were the Laue group of the phase imposes
> some obvious constraints on the harmonics to respect symmetries.
> 
> There is another problem to know.  Is the observed strange peak shape
> simply anisotropy or there are other effects, like superstructure peak
> broadening, asymmetry and peak displacement by faulting etc.?
> In that case you need some special treatments (a.k.a Warren theories for
> twin, stacking faults and/or antiphase boundary). Maud can be used in this
> case but the actual general model only works for bcc, fcc, hcp structures.
> No general approach available up to now.
> 
> Probably you may give some more information about the material and what
> kind of strange peak shape you are observing.
> 
>               Best regards,
>                       Luca Lutterotti

As Caroline has described, the peak shape is anisotropic due to
subcell/supercell behaviour. I will explain, how I have solved
this problem with BGMN:

PARAM=b1sub=0_0^0.1 PARAM=supfactor=1_0^1 b1sup=supfactor*b1sub //
B1=ifthenelse(mod(h+k+l,2),b1sub,b1sup)

This text is part of the structure description input file.
It is for the case of cubic bodycentered subcell/cubic primitive
supercell. Of course, BGMN may handle (positiv definite) broadenings
according to a Tensor of stage 4 (spheric harmonics of 4th order),
too. See ANISO4 at www.bgmn.de. Of course, each value may declared to be
ANISO4.

J"org Bergmann, Dresden
[EMAIL PROTECTED]

Reply via email to