RIR

Helge Stanjek ( t9211ac@wsc18.lrz-muenchen.de )
Thu, 31 Jul 1997 15:25:06 +0200

Hi there!
We investigate samples, which contain iron (hydr)oxides (synthetic and
natural) with additions of other phases such as quartz, feldspars, micas,
clay minerals and all other nasty stuff occurring in soils.
We want:
An XRD measurement, which gives
a) the identity of the iron oxide phases
b) the unit cell edge lengths
c) the profile shape for mean coherence lengths
d) the quantity in the sample, especially for those samples, which contain
amorphous phases or clay minerals (kaolinite, mixed layers, ...), which can
hardly be modelled by Rietveld programmes.

We use either single peak fitting or Rietveld (preferably the latter).

Cases a, b, and c can be handled with SRM 660 (or Si from Merck), but d)
requires other materials (Al2O3, CeO2, Cr2O3, TiO2, ZnO).

The problem:
All listed RIR materials have plenty of peaks, which interfere too much
with iron oxides (Fe2O3 - Al2O3 - Cr2O3 !) or they have mass attenuation
coefficients (Co radiation), which are too high.

Solution ??
Why not using MgO, which has been measured against Si for precise unit
cells. MgO might have slight broadening, which improves fitting with
PseudoVoigt profiles. It is cubic and has at least three peaks with I >
10%. Its mass attenuation is not too far away from our sample phases.

Any objections against using this stuff? Have I overlooked something else
to think about?

Thanks in advance

Helge Stanjek