From: Steve Hillier [[email protected]]
To: [email protected]
Subject: Quantification of soil clay minerals

Dear Marcelo,

There is no simple answer to your question about using the Rietveld
method to quantify clay minerals in soils, or clay minerals by the
Rietveld method in any sample for that matter.

If your clay mineral assemblage is simple, and the clay minerals well
crystallised with no 2D diffraction effects you may be able to get by.

If on the other hand, the clay mineral assemblage is complex and the
clay minerals poorly ordered  you will face many more problems.

Here at Macaulay Institite we tend to prefer the classic RIR method
where clay minerals are concerned.  Although the RIR method has its own
drawbacks and fullpattern approaches seem much more elegant; when it
comes to clay minerals, soil clay minerals especially, it can be
considerably easier to measure a single peak that can be unequivocally
assigned to one clay mineral or another than to try and model the
contribution of a combination of poorly ordered clay minerals to a full
pattern, invariably including other minerals in addtion to clays.

At then end of the day the accuracy of which ever method you use will
also depend on how much you know about the clays in your sample.

Which ever method you decide upon I would always advise on taking the
effort to add an internal standard.  Even if you use the Rietveld method
the addittion of a standard is about the only way to have some
confidence that the quantification of an unknown is reasonable.

An age old problem you will also face with clay minerals is preferred
orientation, At the Macaulay Institute we spray dry all our samples to
be sure that we eliminate preferred orientation before we start.

http://www.mluri.sari.ac.uk/newcommercialservices/spray/index.html

This makes single/several peak RIR methods much more precise and
accurate it also means you have at least one less varaible to refine if
you use the Rietveld method.


If you want some literature on this subject the chapters by Bish and  by
Hughes et al on XRPD in the following reference are a good starting
point.

Amonette, J.E. & Zelazny, L.W. (1994) Quantitative methods in soil
mineralogy.  Soil Science Society of America Inc.  Madison, Wisconsin,
USA.

Sincerely,

Steve