I agree completely with your comments. The liner volume is only 120 ul for
the HP 6890 PTV, so it is very important to be aware of
pressure/temperature/volume correlations as you say. Just curious if these
problems are actually relevant to all modern GCs - do they all use the same
type of flow or pressure sensors? Thermoquest said no problem with flooding
liners with their new systems in terms of maintained flows and pressures -
not true than you think? I think it would be a relavant feature, not having
the injector performance collapse when some overflow occurs.
Lars

-----Original Message-----
From: Howard Sanford [mailto:[log in to unmask]]
Sent: fredag, november 10, 2000 18:03
To: [log in to unmask]
Subject: Re: [ISOGEOCHEM] HP 5890 flow problem


At 04:36 PM 11/10/00 +0100, you wrote:
>At least on our 6890 with electronic pressure control, using the PTV for
>cold splitless injections, injection volumes above 5 ul toluene caused
>complete loss of analyte due to decreased total flow and column flow to
zero
>ml/min for column, and 2 ml/min for total flow (=maintained fixed septum
>purge flow). We concluded that the actual flow/pressure sensors did not at
>all like solvent evaporating out of the liner into the sensor area.
>
>So, overflowing solvent and/or dirty sensors I suppose could really get you
>into trouble with the nice HP6890 at least. Would prefer a more robust
>technique I think.

The problem you describe here could effect any GC. The 5ul liquid injection
is expanding beyond the volume of your injection liner as it goes into the
gas phase. Vaporized toluene then overloads the purge lines and before it
can be swept away is condensing in your injection lines and in your column.
HP used to have a great book on split/splitless capillary injection and had
a handy table of expansion ratios for various solvents at various pressures
and temperatures. Most liners only have a volume of 200-900 ul, so
depending on your injection temperature and the sample you are injecting,
it is quite easy to flood the liner. This will cause the effects you see in
addition to causing large injection to injection variations in peak areas.
There are a few ways around this -

1. switch to a liner that has a large internal volume
2. lower your injector temp.
3. change solvents to something that doesn't expand as much when going form
liquid to gas
4. concentrate the sample and inject less


Howard Sanford
Stable Isotope Ratio Mass Spectrometry Lab
919-513-3039
North Carolina State University
Dept. of Soil Science
Box 7619
Raleigh, NC 27695