The problem was a very small leak between the valve of H2-reference
gas tank (valve 1) and the second valve connecting it to the vacuum
line (valve 2). So the H2-ref gas was diluted by air in the space
between the two valves. The longer the gas paused between the two
valves the lower was my signal (because the H2 was more diluted). One
can detect the leak by pumping the whole line up to the valve on top
of the reference bottle. Close valve 2 and wait (time depends on leak
size). Over the time the space will fill with some air. Open valve 2
and check the pressure gauge of the bellow which will show pressure
in case of a leak.
I suspect that the H2-reference gas in the tank was probably
contaminated by air in the moment I opened the valve of the gas tank.
I usually evacuate the space between the two valves before opening
the ref gas valve but to make sure not to get any fractionation I let
the gas expand for some time. So it may be necessary to refill the
gas tank with fresh reference gas.
Robert van Geldern
Leibniz Institute for Applied Geosciences (GGA)
S3: Geochronology and Isotopehydrology
Stilleweg 2, 30655 Hannover, Germany
phone: +49-(0)511-643-2313 (office), -2539 (lab)
mailto:[log in to unmask]