Dear all,

Having just changed and  leak tested the reference gas bellows on the inlet
to a SIRAII ms we are getting some odd results. Analysing repeated aliquots
of CO2 equilibrated with a standard water sample over a period of several
days shows the following trends:

1) delta 45 (sam/ref) is constant

2) delta 46 (sam/ref) has decreased by ca. 0.7 per mille.

The results suggest to me that there is oxygen isotope exchange between the
reference gas CO2 and a contaminant in the bellows. My first hunch would be
to suggest water vapour adsorbed to the stainless steel. However the inlet
has been baked and thoroughly leak checked. I now wonder if there is a
cleaning agent, used by the bellows manufacturer, that is difficult to
remove through routine baking that is causing the problem.

Has anyone else seen a similar problem on their instruments? I'd be
grateful for any suggestions.

Best wishes,

Paul