Hi Greg 

The likely culprit is a broken bleed capillary at the end of the backflush valve.  Is is usually found near the oven door.  It is often coilled to the side and hanging.  If you unroll it you will probably find it broken.  Replace it and your peaks should return to normal.  The dimentions are in the GC-C interface manual.

Regards

Gilles

 

Hi all,
I'm working with a GC-C-IRMS system that has not been in use for some
time ( a number of months), attempting to run amino acid samples for N15
determination. Our system uses an Agilent 6890 GC, a Thermo GC III
interface, and a Delta Plus XL mass spec. I'm using a Chirasil-val
column for the GC separation (50 m x 0.25mm ID). The signal amplitude
I'm getting is off the expected value by a bit better than a order of
magnitude (factor of 20, give or take). I'd initially figured my either
my preparatory chemistry was bad or there was an issue with the
reactors, so I tried the argon test outlined in the GC III manual (ie,
inject a couple of microliters of air and observe mass 40); the
amplitude on this was off by approximately the same factor. My next
guess was a dead volume, so I sequentially have tried bypassing
different fittings and observing the response. Strangely, the thing that
makes the big difference is the GC column itself. If I bypass the column
with a couple of meters of deactivated silica capillary, the argon test
comes out in the acceptable response range, so at first glance, at least
all the fittings downstream of the column appear good. Also, the problem
does not seem column-specific- I've tried other columns of the same
physical dimensions but with a different stationary phase and I seem the
same problem.

The only thing I can think of is that the issue has something to do with
operating pressure. I'm running the GC in constant flow mode, so in
order to maintain that flow, the pressure through a 50 meter, small-bore
column must be substantially greater than when the short bypass
capillary is used. I have little personal GC-C-IRMS background (I'm
learning by doing) and am running out of ideas to resolve this. Has
anyone ever run into a situation like this before, or does anyone have
sufficient expertise to hazard a guess as to the cause? Any thoughts are
greatly appreciated. Thanks very much,

Greg Ellis
Paleoceanography/Biogeochemisty Lab
University of South Florida College of Marine Science