Following the definitions of the IMA nomenclature committee, mica
subcommittee, (e.g. Rieder et al., AmMin Vol. 83 11-12, 1366), see
http://www.minsocam.org/MSA/IMA/ima98%2810%29.pdf
in the system of dioctahedral micas, the borderline between celadonite
and glauconite is the layer charge of 0.85, and the "K-deficient micas"
illite and glauconite are defined by </> 50 % Fe at the octahedral
positions. Thus, celadonite and glauconite may not be distinguished by
the b-axis value (060/33l positions). To separate illite from glauconite
(and muscovite from celadonite as well) one should be sure about the
octahedral Fe content. I'm in doubt that this is possible to do
correctly from the 060/33l position only, as there is also a certain Mg
for Al substitution. Moreover, all these minerals commonly prefer the 1M
structure but tend strongly to disorder in layer stacking
(60/120°rotations, mixed-layering of cis/trans octahedral occupation).
Consequently, the diffraction patterns are very variable, and a Rietveld
refinement of complex mixtures using ideal mica structures is like gambling.
Regarding non-convergence of refining superimposing ankerite-dolomite
patterns : Could be a problem of profile modeling. If the high angle
peaks are strongly broadened/smeared by "microstrain" (i.e. chemical
gradients) of one or both "phases" and there are no physically sound
constraints of the peak shapes, always one (the dominating) phase tends
to catch all the intensity in the peak flanks by extreme "broadening".
Consequently, there are no remaining maxima what could be fitted by the
minor phase and the refinement of the lattice parameters runs into a
wrong minimum. You should simply apply the old wisdoms: add physically
meaningful constraints, for lattice parameters and profile shapes. And
try to find a meaningful model for the peak profiles.
But if there are more complicated reasons for the misfit, e.g.
continuous gradients of the chemical composition, or any disorder in
cation occupation like in "Ca excess dolomites", than it sounds like a
hard job to find (and to formulate) a reasonable model for your refinement.
Reinhard
Am 22/04/2015 um 17:28 schrieb "Łukasz Kruszewski":
Dear Friends,
I have two questions on PXRD. First, is it possible to distinguish between
illite and glauconite and celadonite in a complex mixture containing
quartz, illite, muscovite, chlorite and kaolinite (plus some
interstratified clay minerals)? I was trying with (060) reflections, and I
cannot observe anything with d(060) > 1.51, but there is a reflection
between 1.50 (that I ascribe to illite) and 1.51, thus I assumed it could
be from celadonite.
Second question. I use TOPAS (v. 4.0) for the Rietveld refinement. I have
samples with dolomite and I know that it is strongly zoned, thus the main
reflection is split (dolomite + ankerite). As an input I've used these 2
structures, but both reflections obtained during the refinement lie at the
same position, leaving the "shoulder" non-adjusted (to dolomite). I
suppose this result is wrong (?). Do you have any idea how to fit such a
"shoulder-like" reflection in such case?
Best regards!
--
TU Bergakademie Freiberg
Dr. R. Kleeberg
Mineralogisches Labor
Brennhausgasse 14
D-09596 Freiberg
Tel. ++49 (0) 3731-39-3244
Fax. ++49 (0) 3731-39-3129
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Please do NOT attach files to the whole list <alan.he...@neutronoptics.com>
Send commands to <lists...@ill.fr> eg: HELP as the subject with no body text
The Rietveld_L list archive is on http://www.mail-archive.com/rietveld_l@ill.fr/
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
