In addition to what our colleagues have suggested I would use a slow
fill speed and also see if increasing the working temperature of the
reactor and/or GC column makes a difference. A fast fill speed for
water samples/rinsing solutions can introduce tiny amounts of air in
the sample that will be injected into the reactor, therefore
interrupting the continuity of water sample delivery, and also
negatively affect the syringe plunger in the long run. I am using a
1.2 microliter syringe and a fill speed of 250 nl/sec. Cheers. Mihai
Lefticariu (Southern Illinois University Carbondale)
On Mon, Mar 25, 2013 at 5:20 AM, Robert van Geldern
<[log in to unmask]> wrote:
> Hi isotope folks,
> we run an Thermo TC/EA with a PAL autosampler for water injection.
> We recently modified it with a bottom feed adapter that reverses the helium flow which should reduce the memory effect of the system .
> A stainless steel insert is used on top of the glassy carbon reactor and the top fan is switched off. The top He capillary is shortened (2cm) and plugged with a 16" swagelock union with blind cap.
> Before the reversal of the helium flow we used a 0.5uL syringe that was recommended by Thermo but this delicate model tend to fail after a couple of injections (plunger blocks even with DI water). So we changed to a 10uL gastight SGE (#002987) syringe as recommend for the bottom feed design in  and set it to 700nL injection volume and about 80% dilution in the ConFlo IV.
> However, we always observe a "double" peak for hydrogen (or a peak with a shoulder).
> We tried all parameters (injection speed, post injection delay, hot needle injection, injection volume etc.) with not much success. All will influence the shape somehow but after a couple if injections we always back to the "shoulder peaks" for hydrogen.
> We tried again a 0.5uL syringe with 400nL injection volume and things looked better, but as most of the times the syringe failed after a couple of injections.
> Finally we used a standard 10uL non-gastight syringe (SGE #002980) from our Picarro and it seems that we got rid of the double peak.
> So it seems that our problem is related more to the syringe type and not so much to the other parameters as we kept them identical.
> You may see pictured here:
> My questions:
> *Has anyone used and tested different types of syringes and can recommend one?
> *Any other tips and tricks around the bottom feed design that tend to improve results in your system?
>  Gehre M, Geilmann H, Richter J, Werner RA, Brand WA. Continuous flow 2H/1H and 18O/ 16O analysis of water samples with dual inlet precision. Rapid Commun Mass Spec 2004;18:2650-60.
> Dr. Robert van Geldern
> GeoZentrum Nordbayern / Applied Geosciences
> University of Erlangen-Nuremberg
> Schlossgarten 5
> 91054 Erlangen, Germany
> [log in to unmask]
> fon: +49-9131-85-22514
> fax: +49-9131-85-29294
> room: O 2.112 (office)