As mentioned, magnesium perchlorate is used in numerous continuous flow
applications without any carbon isotopic fractionation problems. I too have
found no problems using magnesium perchlorate as a water trap in two other
applications I use often. Until recently we did not have continuous flow
capability. Hence, I setup an offline means of collecting CO2. Coupled to
the EA I use an extraction line on the down stream side for capturing CO2
in ampoules for MS analyses. This procedure works very well and has been
tested using numerous stds with no carbon fractionation problems.

Taking this one step further, I have installed a sea water stripping device
on the up stream side of the EA. I use magnesium perchlorate attached to
the down stream side of the stripper and in the usual place on the EA.
Again no carbon fractionation problems. This procedure was tested using CO2
of known composition. The CO2 was passed through the magnesium perchlorate
attached to the stripping device and the EA then collected on the down
stream extraction line. The results demonstrated that the procedure and
chemicals used had no effect on the CO2 carbon. I then took the stripping
technique one step further and eliminated the EA and connected the
stripping device to the extraction line performed the same CO2 check. Again
no problems with carbon using magnesium perchlorate as a water trap.

All of these procedures I mention above were used with a UHP He carrier
flow, collecting CO2 from the He stream using LN2, evacuating the He to a
10-3 torr vacuum, then collecting the CO2 into a 6mm ampoule for MS
analyses. I have no comment on how magnesium perchlorate would perform
under vacuum. In vacuum extractions I use IPA/Dry ice to remove water.



At 08:33 PM 5/25/98 -0700, you wrote:
>I built a vacuum line to trap CO2 from atmospheric samples for isotopic
>analysis a couple years ago, and initially tried a magnesium perchlorate
>trap to remove water vapor from the samples.  The line produced poor
>results for stable carbon isotopes.  While investigating potential
>problems, we heard that magnesium perchlorate works fine near atmospheric
>pressure, but does not perform well under vacuum.
>
>We elected to replace the magnesium perchlorate trap with a dry ice/alcohol
>slush trap, and eventually got excellent results.  Unfortunately, we
>changed several other aspects of our line at the same time, so I can't
>conclusively say that this was the problem.
>
>While the isogeochem group is on this subject, I'd be very interested to
>know if others have experienced problems with magnesium perchlorate at high
>vacuum... or if these problems can be eliminated by using an anhydrous
>variety.
>
>Thanks,
>Troy
>
>PS.  Also a safety warning:  dry magnesium perchlorate can explode
>violently when shocked severely.
>
>
>Troy Baisden
>Graduate Student
>ESPM-Ecosystem Sciences
>UC Berkeley     (510) 643-6910
>
>
****************************************************************************
David A. Mucciarone
Stanford University
Department of Geological and Environmental Sciences, 320
Stanford, CA 94305-2115
650-723-0817 (office) 650-725-0979 (fax)
e-mail: [log in to unmask]
WWW: http://pangea.stanford.edu/~dam

****************************************************************************