Hi Roland and other interested members,

I started working on Farquhar's pyrolysis system early this year and
fiddled with the analysis of water samples. At present, I have achieved an
excellent correlation between the accepted and raw delta O-18  of solid and
water reference standards (namely, beet sucrose, IAEA-CH-6, ANU-C1, ANU-P1
and ANU-A1). The slope is 1.0025 and R-square is 0.9994. To check the yield
of CO from water, I compare both the TCD and mass spec outputs of water
standards with beet sucrose standard. CO yield from water standards are
usually 98 - 99.5 % of what's observed from beet sucrose. These results were
obtained with no modification to the method described in Farquhar (1997)
paper (at 1080 oC) other than removing the little hiccups from the previous
efforts.

The problem with our system is still memory effect which I'm looking into
as well.  I've examined your paper with Barbara Kornexl (1999) and I wonder
if the trick of eradicating memory effect lies with high temperature
(1400oC), use of glassy carbon grit and tube or a combination of all. If I
read correctly, your choice of high temperature is to permit analyses of
inorganic samples such nitrates, sulphates and phosphates. I wonder if high
temperature is also crucial in minimizing memory through efficient
pyrolysis. The answer is probably obvious and I would appreciate your
comments based on your experience in developing the technique with Barbara.
I hope my answer to your query helps.

Cheers,
Kim



**************************************************
Kim Gan

Environmental Biology
Research School of Biological Sciences
Australian National University
GPO Box 475
Canberra, ACT2602
Australia

Tel: 61-2-62492406
Fax: 61-2-62494919