It is a very good idea to put a scrubber inline between the Gilson and the
gas box on the Europa.
We built such scrubbers for AMS from polycarbonate pipe: 1/2" OD, 3/8" ID
and cut about 8" lengths.
We used SS swagelok fittings: 1/2" plugs and nuts, and the reducers that
have 1/16" threads and 1/4" tube (NOT reducing unions). We used BRASS
ferrules to avoid damage to the polycarbonate (PTFE is too soft).
Our machine shop drilled through the plug and brazed in the reducer for a
gas tight joint. The 1/16" takes the PEEK from the Gilson.
Now you have a choice of sorbents for this trap: if you really only want N2
then use ascarite, it traps CO2 and other urea-generated gases as well as
water. You might also use something that traps the vapours of LiOBr.
Finally, this is a related heads up for the folks at SerCon: the pairs of
solenoids on the Europa ABCA have gross deadspace and therefore carryover.
They could be used to drive a Valco 6 port valve. The sample loops would
be customised for headspace, breath or other samples. We have used these
valves on everything from tanks gas to high vacuum with good results.
University of Toronto
On Fri, 16 Sep 2005 10:20:07 -0500, Juan C. Marini <[log in to unmask]>
>Dear list members,
>I have a question for those with experience on sampling the head space of
>liquid samples. I am reacting LiOBr with a urea solution to generate N2
>gas. The sample is analyzed with an Europa 20/20 IRMS attached to a Gilson
>autosampler. I am concerned with moisture (and LiOBr) creeping into the
>machine. The sample goes through a few valves and plumbing before reaching
>the water scrubber and I am afraid that with time it is going to cause some
>The option I see, is installing another water trap immediately after the
>autosampler. The trap could be a liquid N trap or another magnesium
>perchlorate scrubber. Any suggestions?