Take care in heating the Nafion though: besides melting, exposure to
even moderate temperatures over time can lead to degradation of the
membrane: in which case it bleeds mass 46!
At 08:24 PM 07/07/2010, you wrote:
>I recommend a solid heating of everything, and with this much water,
>with the needle valve closed. I will define gently as 100*C. I keep
>an arsenal of heat guns, space heaters, and modified gutter heat
>tape. After at least a day of this, open the needle valve, set your
>lowest amplified cup to mass 18, and heat the capillary from the
>interface to the IRMS. I bet your nafion is fine...interface nafions
>are designed for small amounts of water. Once you have backgrounds
>to a tolerable level, try to find where it is accumulating and keep
>that part hot all the time. If you know its leak free and you know
>your helium is dry, HEAT HEAT HEAT.
>Earth and Space Sciences
>University of Washington
>Seattle, WA 98195
>[log in to unmask]
>On Wed, 7 Jul 2010, Rachel Schwartz-Narbonne wrote:
>>We are getting high water background levels on our GC/C/IRMS and have had
>>difficulty reducing them. The mass 18 signal on the IRMS was at saturation
>>(14000 mV), and we have had limited success lowering it to about 7500 mV by
>>allowing the system to flush over a number of days. This is this case in
>>backflush-mode with the open split in (take the split out and it only drops
>>to 3500mV). We see little difference when we turn off the backflush so it
>>does not seem to be coming from the GC-combustion side. We have replaced the
>>nafion trap in the GC/CIII interface and this only reduced the background
>>slightly but not significantly (to the present 7500mV we see presently). We
>>have been running liquid samples (amino acids in ethyl acetate). However,
>>we have not injected anything in more than two weeks, and have seen high
>>water levels throughout this time.
>>The background levels for argon, nitrogen and carbon dioxide are normal for
>>a leak-free system in backflush and non-backflush-mode, suggesting that we
>>do not have a leak. The He gas is dry, since the water problem disappears
>>when the IRMS is switched over to our other peripherals.
>>This is the first time we have done liquid samples of this nature on this
>>system and are wondering whether this is a ?normal? state of affairs for such
>>samples or has anyone else had this problem before? Do you have any
>>suggestions where our water is coming from, or how to get rid of it?