I had the same thought as Gerry about checking the bleed capillary on the
backflush valve. It's a delicate capillary that can crack and leak without
even looking broken. Also try tightening the nut where the capillary
connects to the valco union on the combustion reactor inside the oven. This
spot seems to be prone to leaks that are undetectable with He sniffing and
UC-Davis Stable Isotope Facility
2220 Plant & Env Sci
One Shields Ave
Davis, CA 95616
Phone: (530) 752-8100
FAX: (530) 752-4361
From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On Behalf
Of Charlotte Lehmann
Sent: Monday, January 07, 2008 10:38 AM
To: [log in to unmask]
Subject: [ISOGEOCHEM] peak size, shape and timing not as expected
We have a Delta V with GCCIII and Trace GC. Last good runs were at the
end of November with no use until 1/2/08. On 12/21/07, I turned the
source off since no one would be around during the Winter Holiday. The
PI turned on the source and geared up to run some samples, but: "On
Wednesday last week, things looked fine for the first couple of runs
of the standard, and then something happened between the third and
fourth injection -- n-alkane peaks became delayed, broader and much
smaller. (The CO2 Ref on-offs are fine -- meaning the problem lies up
stream of the ref inlet.) My initial thought was a leak somewhere.
Mass 40 was higher when back-flush is not on, suggesting a leak.
However, I have checked for leaks everywhere using the leak detector
and snooping with argon (while monitoring mass 40) and there is no
sign of a leak anywhere. Backflush valve flows are fine. I have
installed new septa, and re-oxidized the furnace...still no
I checked, at her request, to see if a mass scan gave an indication
that the oxidation or reduction furnace was spent. The mass scan, in
"straight mode" with back-flush and dilution OFF, looks the same as
with them on, and the masses all appear normal in size, shape, and
relative proportions. They look very similar to the mass calibration
scan done on 10/25/07 when the filament was replaced and the
instrument tuned through the Conflo III (on the EA side).
I even scanned without REF CO2 in, both "straight" and "not straight"
modes. Peaks are small, no leaks apparent.
Any thoughts, suggestions, etc. would be greatly appreciated.
Assistant in Instruction
Department of Geology
206A Carnegie Science Building
44 Campus Avenue
Lewiston, ME 04240