Dear all,

Reading the question by Bob and the answers by others, I like to point at
the fact that using an EA for carbonate analysis basically thermal
decomposiiton of the carbonate is applied. From early years this method was
not considered very trustful: McCrea (1950!) tested thermal decomposition
but concluded that fractionations were too large and irregular and
reproducibility was not according to desired values.
However, that was all done with 'off-line' methods.
With the introduction of laser techniques, several lasers (eg Nd:YAG, CO2
lasers) were tested for thermal decomposition of carbonates, with different
success. Both, vacuum systems and CF-He flow (about atmospheric pressure -
actually slight overpressure) systems were applied. Reproducibilities are in
order of 0.4 to 0.5 permil in the 'successful' systems. Remark with most of
these methods is that the conditions between samples and standards has to be
carefully similar - changes in laser operation (and other?) conditions may
give erroneous results.
Meanwhile also thermal decomposition in EA-CF-IRMS systems was introduced.
here similar problems as with the CF-laser system are valid in general.

In fact this relates to the general problem of the carbon standard/scale.
The carbon scale was defined by phosphoric acid digestion of the former PDB
standard. Since this is a completely different operation, one has to be
careful about values created by thermal decomposition. Almost certainly
these values will deviate from the acid digestion values... Different
physical factors are playing an important role here.
For a start, what exactly happens during the acid digestion process is not
completely understood. The method is calibrated on specific quality
(density) of acid reacting at a controlled temperature (25 degrees-C). This
makes the process sufficiently reproducible and precise. If using other
methods, such as thermal decomposition, this calibration will not be valid
in a similar way. The only thing we can do is comparing with the values from
acid digestion and correcting the thermal decomposition values, considering
they are reproducibly deviating, if we want to be on the proper C and O
isotope scale.
For oxygen the problem is even worse than for carbon, since only two third
of the carbonate oxygen is released by both methods. Oxygen isotope
fractionations at elevated temperatures in thermal decomposition (where
there is no precise control on the actual temperature!) will be clearly
different from the 25 degree-C acid digestion situation. So, different
values are expected in any case for both methods. Moreover, if applying
thermal decomposition isotopic fractionations will be fairly small at high
temperature (at least it is what one expect). The actual fractionation for
acid digestion is not known to my knowledge, but will not be similar with
the other method.
Not to mention organic 'contamination' in organically formed carbonates,
where organic-C is released (or not?) in different ways in both methods (as
also mentioned by David). Pretreatment is not a sufficient solution, since
organic carbon generally is included in the carbonate crystal structure and
hardly to be remove completely, unless decomposing the carbonate completely
in the pretreatment, which is not an option of course. To what extent the
acid will decompose (digest) organic matter, if at all, is not quite certain
either. Thermal methods will decompose them for certain.

Forefoing is also an important problem in analysing organic materials for
13C and relating to the VPDB scale based on carbonate. The methods for
analysing standard and sample material and also the quality of material is
different. If using a EA-CF-irms-system, there is a physical difference in
materials analysed there!
I do not think there is a simple solution for this, to keep the existing
scale, which certainly we must do. Even if we calibrate organic materials by
applying the same method on different materials, we not necessarily are on
the same scale...

Just some words of caution!

Since the VPDB (Vienna-PDB) is a defined scale and not based on an existing
material with that specific isotopic compostion, it might be correct to
change the name into: Virtual-PDB, just with the same acronym VPDB, and thus
not misleading as such!


Best wishes,
Pier.

****************************************************************
Delta Isotopes Consultancy

Dr. Pier A. de Groot
Pastoor Moorkensstraat 16
2400 Mol - Achterbos
Belgium
Tel. +32 (0)14 326 205
e-mail: [log in to unmask]

Visit my WEB-site about my ³Handbook of Stable Isotope Analytical
Techniques²
at:
http://users.pandora.be/handbook/index.html
last update: March 17, 2004
****************************************************************







> Folks,
> Myth or reality - Can CaCO3 be run for d13C via standard EA-IRMS methods?
> If so, what kind of accuracy and precision are readily obtained?
> I have an application where it would be advantageous to do so, but cannot
> recall if such analyses are possible, or just urban legend.
>
> thanks, Bob
>
>
>
> "Make three correct guesses consecutively and you will establish yourself
> as an expert"  (Ryan's Law)
>
> Dr. Robert F. Dias
> Assistant Professor
> Department of Chemistry and Biochemistry
> Old Dominion University
> Norfolk VA 23529
> ph (757) 683-4093
> fx (757) 683-4628
>