Dear All,

In response to Steve's question about nitrogen cold fingers and Atilla's
reply, we have tried a cold finger in my lab at UEA for a period of about a
year and with a very mixed experience.

The cold finger was built to replace the standard CO2 finger on a Micromass
SIRA instrument and was packed with silica gel. The volume of the packed
finger was carefully matched with the dummy finger on the reference gas side
so that depletions remained the same for both sample and reference gas.

Sample gases were frozen onto the cold finger and desorbed carefully to
ensure 100% desorption. However we observed phenomena that we couldn't fully
account for in terms of a physical process, other than re-adsorption, and
subsequent fractionation of the gas. Briefly the salient points are:

1) For gas amounts of ca 2.5 x 10^-5 mols the cold finger worked well, with
no observeable fractionation of the sample gas.

2) For samples below this amount there was a progressive depletion in the
measured composition of the gas. At the same time we noted that following
the initial desorption, induced by heating with an electric resistance
heater wire and temperature equilibration, there was a subsequent
re-adsorption of a component of the gas back onto the gel. This was highly
repeatable and consistent between samples.

To test we used nitrogen prepared from air (removing oxygen over hot copper
etc.) and gas pipetted different sized aliquots so as to avoid freezing onto
any adsorbent or zeolite material that might compound the problems we were
trying to characterise on the mass spectrometer cold finger.

3) The magnitude of depletion increased to about -6 per mille for a 1
micromolar sample.

4) Following warming it seems that there is a subsequent re-adsorption of
nitrogen back onto the gel with a fractionation between the adsorbed and
free gas.

5) It has been possible to correct much of the very small sample data using
an empirical procedure based on the analysis of large numbers of standards
of varying size between 1 and 10 micromoles. At the same time we devised a
model involving adsorption and fractionation that accounts for the observed
depletions.

6) To test if the problems are associated with the cold finger or mass
spectrometer we also measured samples over a range of inlet pressures, with
matching and mismatching capillary crimps etc. and always observed good
results, even down to very low major beam signals.

I'm not aware of any other group observing such behaviour and would like to
hear of any experience others might have. I suspect, Steve, that the story
may be prefaced with CAVEAT EMPTOR!

In Attillas case the samples were quite large (0.2 mmol) and your
experiences match our own. We use silica gel routinely for freezing down
such samples.

Tomasz I will check your paper.

Best wishes to all.

Paul

Paul F. Dennis
Head of Stable Isotope Laboratory
School of Environmental Sciences
University of East Anglia
NORWICH NR4 7TJ
UK

Tel: 01603 593105
Fax: 01603 507719
email: [log in to unmask]