Dear Listmembers,

        Once again I hope to enlist your help.  I have built a conventional carbonate line and
all seems to be working well now except one item--the reaction vessels--which seem to
contain high residual gas pressures [about 40 microns upon expansion into the manifold].
The pressures are high enough that the gas transfer rates are slow.  It takes 7-8 minutes
to freeze down 100 micromoles of CO2 and clearly that is too long.  This is after rough
pumping overnight on the acid/sample and then pumping for at least a couple of hours with
the diffusion pump and a cold trap before reaction.  I have been letting samples react
overnight.  Once the CO2 is trapped and the residual gas is pumped away, the CO2 will
freeze again very rapidly if I re-expand it.

        My two legged reaction vessel design was copied from a colleage.  The upper half
contains the teflon/o-ring stopcock and the lower half contains the two legs for reagent
and sample.  They are joined with a ACE o-ring connection and a clamp rather than a
ground glass fitting.

        I ran blank vessels to determine whether the residual gas was related to the reagent or
the reaction.  I rough pumped on the vessels overnight and then with the diffusion pump
and a cold trap, then I closed the stopcocks.  I left the vessels on the manifold but let
them sit overnight [assuming the problem is not in the stopcocks].  I also torched the
bottom of the vessels before closing the stopcocks.  Overnight a residual pressure of
about 40 microns resulted.  Hence, the source of the residual gas appears to be the
reaction vessel itself.

        It seems to me that if the problem were leakage, that it would be unlikely that all of
the vessels would produce such similar residual pressures or leak the same amount.  Any
ideas as to the cause or a possible solution would be appreciated.  I've struggled with
this for quite some time now.



*********************************
Steve Nelson
Dept. of Geology
673 WIDB
Brigham Young University
Provo, UT  80602

801-378-8688 voice
801-378-8143 FAX

"Intel Inside" is a warning label
**********************************