Dear Lars and low volume CO2 measurers

I'm not familiar with the set-up and plumbing of the Breath Prep on
the optima, but from the loop sizes compared to the vacutainer
sizes, it seems that you are only putting 0.5 to 10 per cent of the
sample into the source. When doing similar low ppm CO2
measurements on a Europa 20-20 ANCA, we switched from the
normal breath mode, into tracer gas mode, which directs all the
sample into the source - in standard breath mode, only one fiftieth of
the flow is directed to the source, which gave us terrible results
when we first tried to measure low CO2 samples. In tracer mode
however, this gave us precisions of about 0.1 permil (for 13C) from
10ml samples of air (350 ppm). Maybe you can adapt your system
similarly.

Now working with an Optima Prism, by directing all the sample into
the source we can get standard deviations of 0.15/0.20 for 13C and
18O from 350 ppm CO2 in 10 ml of air, and again, maximising the
sample volume going into the source (by adjusting the open split and
loop sizes) was crucial to getting good numbers.

I hope this helps you reach the often elusive quoted precisions

Jim Gillon

Dept. of Environmental Science and Energy Research
Weizmann Institute of Science
76100 Rehovot
Israel

Tel 972-8-9344227 or Fax 972-8-934 4124