Re: [sdpd] Interesting (?) Problem

[Armel Le Bail <alb...@cristal.org>]

>It seems to me that this question could be answered by exafs at either
>the Nb or Te edge, since the two models would predict different partial
>g(r)'s.

Interesting old problem. A lot of literature about it.

Do not expect too much from EXAFS about the pyrochlore structure.
There is at least one good paper on that question (good because I am
the main author ;-) : J. Solid State Chem. 61 (1986) 188-196. The problem
with this paper is that it is in French (but translators are not for dogs).

In the above paper, the compounds studied by EXAFS were AM(ii)M(iii)F6
pyrochlores, with A = Rb, Cs; M(ii) = Zn, Ni, Mn and M(iii) = V, Cr, Fe, Ga.
All seen as cubic Fd3m by diffraction methods. The mean diffraction
structure says that distances M(ii)-F should be equal to distances M(iii)-F
since M(ii) and M(iii) occupy statistically the same site, and F atoms
occupy only one site. EXAFS at least confirms that M(ii)-F and M(iii)-F
have their usual values (so that they can be very different, for instance,
Zn-F = 2.01 A and Ga-F = 1.89 A in CsZnGaF6 by EXAFS). However,
concerning the M-M first distances, EXAFS said that there was few
dispersion (M(ii)-M(iii), M(ii)-M(ii) and M(iii)-M(iii) distances are very 
close
to each other)... Then nothing could be concluded about a possible local
M(ii)-M(iii) order by EXAFS. With a near-constant M-M distance, locally
different M-F distances lead to variable M-F-M angles and can explain
the large thermal parameters (pseudo, because representing the local
disorder of F positions)) values found for the F atoms in the X-ray structure
determinations (Fd3m). This also explains the almost normal thermal
parameter for the M atoms.

Now, by diffraction methods, disorder is disorder, and if you do not see any
peak splitting that really allows you to quit the cubic Fd3m space group,
then any of your attempts to deal with the data in an orthorhombic (and
why not triclinic ?) system is pure waste of time.

EXAFS just said that the Fd3m space group does not apply at the
local scale because of significant differences between M(ii)-F and
M(iii)-F distances. But the Fd3m space group applies at medium and
large scales says diffraction, as a mean...

Best regards,

Armel





 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/