Greetings...

I have been following the recent discussions on the list concerning
correcting inorganic O and C isotope data for carbonate samples.  We
have a MAT 253 and a Kiel III device in our lab.  We recently had the
same discussion...why is the correction we are using additive rather
than multiplicative (is that a word?).

When we run NBS-19, values come out as follows: d13C = 1.82 ‰ and d18O
= -2.37 ‰.  I've always corrected my data by adding the necessary
correction factor that brings NBS-19 in line, to all of my samples as
well.  Ideally, the reference gas would be calibrated so that this step
is unnecessary, so I performed an experiment:

Our reference gas values: d13C = -13. 36 ‰ and d18O -10.95 ‰.  I pulled
up an old run of NBS-19 and changed the reference gas values by the
additive correction factor, so d13C = -13.36 + 0.13 = -13.23 ‰ and d18O
= -10.95 + 0.17 = -10.78 ‰.  Reevaluating this NBS-19 run with these
values gave d13C = 1.95 ‰ and d18O = -2.20 ‰.

I then took a few sample runs to see what would happen.  One sample:
d13C = 0.00 ‰ and d18O = 0.41‰.  Changing the reference gas values for
this sample and reevaluating gave d13C = 0.13 ‰ and d18O = 0.58 ‰,
which is additive (this was a great test sample...0.00‰ should not have
changed at all if the correction is multiplicative).  I also tested a
more depleted sample: original d13C = -5.60 ‰ and d18O = -8.01 ‰.  With
the new ref gas values this changed to d13C = -5.47 ‰ and d18O = -7.84
‰, again which is additive.

Unless there is some flaw in the way the Isodat 2.0 software is
calculating delta values, this seems to me as proof that addition to
correct works...at least in this range of sample data.  I know of one
motherland engineer who has stated that delta values are not additive
in this way.  Though the correction seems to hold within the data range
above, as we move further away from NBS-19, such as around NBS-18 (d18O
= -23 ‰), there will be an error that appears (maybe around 0.1‰ at
this point).  Would he care to comment with some formula examples to
show the exact way of correcting data when your reference gas is not
perfectly calibrated?

Any other thoughts?

Thanks...

Tim

-----------------------------------------------
Tim Prokopiuk
B. Sc. Geology/Technician
Saskatchewan Isotope Laboratory
Room 241
Department of Geological Sciences
University of Saskatchewan
114 Science Place
Saskatoon, Saskatchewan, Canada
S7N 5E2
Phone:  (306) 966-5712
Fax:            (306) 966-8593
Email:  [log in to unmask]