Hi Alexandre,


A question for you in line with Gerad's suggestions and comments, esp. re a
poisoned catalyst, you said in your query to the list what you were doing
(AA analytic) but not how.

What derivatisation protocol do you use to make your AAs amenable to GC? If
your derivatised AAs contain halogens (esp. fluorine), there is your
answer. Halogens such as Cl and F will seriously decrease the serviceable
life-span of your combustion catalyst.

How clean is the ethyl acetate you are using? Is it completely dry (= 100%
water free as in freshly distilled and stored over a drying agent)?

Also, you mentioned your FID trace looks OK without "too much" fronting.
What's your definition of "without too much"? Overloading the GC column not
only makes for bad chromatography but also leads to an overload of the
combustion reactor. It is perfectly possible to overload the reactor
temporarily without actually exhausting it.

There is only a finite time window available for interaction of the carrier
gas solute (= the peak) with the surface of the combustion catalyst, which
is determined by the net volume of the reactor and the actual helium flow
rate. So, in case of too much material in the system, the peak gets moved
through and out of the reactor before all material had a chance of an
intimate relationship with the oxidising agent, hence the abrupt cut-off.
But the problem doesn't stop there; this non-combusted material gets into
your reduction tube, which will become clogged eventually, and it might
even end further down-stream.

In addition, the incomplete or non-quantitative conversion of your peak is
associated with isotopic fractionation. Typically, this makes itself
noticed by uncommonly low delta13C values (e.g. -50 or worse).


All the best,

Wolfram



On Jan 20 2005, Gerard Olack wrote:

> Hi Alexandre--
>
> Have you tried looking at Ar, mass 40, while you inject air?  Your
> manual might describe this, but basically use a gastight syringe and
> inject up to 1 uL while monitoring mass 40 over time.  That should come
> through cleanly if there are no strange clogs or leaks, and it's a good
> time to blow some Ar at all the connections to check for leaks.  If
> that's ok, you can then move on to look at N2 from air and/or inject
> some CO2 or CO2 in He.  If they are OK, then you may need to recharge
> your reactor (our GCC is Cu-Pt-Ni that periodically is reoxidized by
> introducing O2 in He stream, other's are set up to bleed O2 in with
> sample peaks) or replace it if it has been poisoned or clogged.  It may
> also be time to replace the reduction column.
>
> good luck
>
> gerry
>
> Alexandre Myre wrote:
>
> > HI everyone,
> >
> > We are presently doing amino acid anaylse on a GC-C-IRMS system. Our
> > samples are recuparated in Ethyl Acetate solvent and are concentrated
> > enough to saturate the FID trace. But when the peak of interest is
> > allowed into the IRMS, the CO2 peak is very small and has a weird
> > shape. The trace of the peak is slowly and steadily increasing and
> > then abruptely ends. It looks like major fronting. I was thinking that
> > may be was overloading the reactor so I quited the splitless mode and
> > changed into split mode but even when putting less sample into the
> > column and reactor, there is not significant change in CO2 peak
> > shape... All this seems mysterious because the FID trace looks OK
> > without too much fronting....
> >
> > I,m looking for any hint !
> >
> > Thank you very much !
> >
> > Alexandre Myre, M.Sc. Earth Sciences
> >
> > Research Officer, Animal Science Department, Universite Laval.
> >
> >
>