Dear Hector,

Your question raises a number of competing issues with the result that no
single method is likely to provide a unique solution to your problem. The
main difficulty lies in the reaction rates of the component carbonates
(combined with the relative differences in the isotopic composition of the
end-members). My own feeling is that cross-contamination is inevitable and
the best you can do is minimise the effects for your own individual
circumstances. This means that some mixture will perform better than others
when it comes to trying to measure the isotopic composition of the
individual components.

Firstly, we all know that calcite and dolomite react at very different
rates, but it is common for different calcites and dolomites to react at
different rates depending on their physical and chemical characteristics.
Although the relative difference in reaction rates of the same phase may be
smaller, these differences may be critical when attempting to optimise
reactions conditions for differential reaction methods. Most experiments
that attempt to validate the differential reaction method will have used
highly crystalline, near-stoichiometric materials because these materials
are pure and readily available. However, these materials often have the
greatest differences in reaction rates. This is particularly the case for
dolomite - compare the reaction rates of highly-crystalline, stoichiometric
hydrothermal or metamorphic dolomites with disorderd calcian dolomites. In
addition to these compositional factors reaction rates are strongly
dependent on particle size (together with temperature), the degree of
separation achieved by differential reaction will depend on particle size
statistics. A uniform particle size would provide the best circumstances,
but clearly this is impractical.

My suggested solution requires a lot more work than the differential
reaction method, but might resolve your problem. Much depends on the
quantity of sample you have available, but if you have the order of 100mg
or more you could try the following method. The procedure is based on: (1)
preparing a sub-sample that contains only dolomite by chemically removing
the calcite; and (2) measuring the isotopic composition of calcite only by
differential reaction.

First hand crush the sample and pass it through a 45 micron sieve (this
ensures that you have no anomalous coarse particles), then determine the
quantity of dolomite and calcite in your sample by XRD analysis. Next split
the sample into appropriate amounts. Dissolve out the calcite from one
split using the method (with necessary modifications) described by Babcock
et al (American Mineralogist, 52, 1563-1567) using Na-EDTA. If performed
correctly this should yield a pure dolomite separate that can be analysed
following conventional procedures.

To measure the isotopic composition of calcite prepare the sample in the
conventional way, but weigh out at least double the amount of calcite
required for your routine isotopic analysis (obviously you need to take
into account the proportion of calcite in your dolomite-calcite mixture)
because you will obtain only about 50% of your normal gas yield. I react
the samples using a technique in which the reaction vessel is equilibrated
at 25oC with the vessel connected directly (using a flexi-section) to a
conventional gas extraction line (the constant temperature bath needs to be
next to the extraction line so that the vessel remains in the bath at all
times). I then react the sample for 30 minutes and after this time collect
the gas released by the reaction. Tests show that providing your mixture
contains >30% calcite the relative isotopic contribution from the dolomite
will be effectively zero unless there are big differences (>10o/oo) between
the isotopic composition of the component carbonates.

The method described above worked very well for mixtures of dolomite and
calcite in metamorphosed limestones, and I believe it would be fine for
other systems with appropriate minor modifications. I think you have to
accept that you will always get some cross contamination and the magnitude
of the effect will depend on the isotopic composition of the end members,
the relative difference in reactivity, and the relative abundance of the
component phases. People may think that this method is unnecessarily
complicated and obviously more time consuming, but I feel that it can be
more readily adapted to individual circumstances than differential reaction
methods that attempt to measure the isotopic composition of mixtures from a
single analysis. Obviously much depends on the number of samples you have
to run and the accuracy of the data that you're prepared to accept.

Finally, an alternative method would be a simple mass balance procedure in
which the isotopic composition of pure dolomite was determined as described
above followed by the measurement of a total CO2 yield from a
dolomite-calcite mixture. This would effectively do away with the
differential reaction procedure altogether. Providing you measured the
proportion of dolomite to calcite in the sample the isotopic composition of
the calcite could be calculated (taking into account the different acid
fractionation factors). Here the accuracy of your XRD method becomes
important, although in the right hands it is possible to measure calcite
and dolomite concentrations in simple mixtures with a high degree of
accuracy and precision.

I'd be interested in any comments anyone may like to suggest.

Regards,

Steve Crowley

Dept of Earth Sciences
University of Liverpool
UK