Hello George,

To add on to what Wenbo said, if your acid is OK and if the acid
delivery line is set up properly, which both seem to be the case on your
system, there shouldn't be any need to modify it to get good numbers.
What you can check is if the amount of acid delivered in each vial is
constant and if you can see bubbles coming out of the acid needle. If
you have just switched back to carbonate analyses, it may take some time
to fully purge the acid delivery line.
As Wenbo suggested you may have small leaks which you should detect with
usual leak checking techniques. You may also have contamination
somewhere in your system.
The external cold finger area could be a candidate for that as it is a
key part of the system and one thing you could do to test would be to
admit some CO2 on both side of the DI, like for your zero all open
tests, then freeze a single cold finger volume of that CO2 in the
external cold finger and then re-load it into the DI as if it was a
sample and compare it against the same gas left on the reference side.
You should get zero as well.
If this is OK, you can try a similar test after admitting CO2 in the
different parts of the MultiPrep (behind MI, WI or WO), freeze it and
compare it with the same gas, that should always give you zero, unless
you have some contamination somewhere, if it is the case that will tell
you where.
Water could be an issue, but the chiller should not be the cause for it.
The MultiPrep water trap is mainly used for water analyses. For
carbonate, the amount of water coming from the sample or from the acid
will be minimal, and when the sample gas is released from the external
cold finger, it is released at -70°C so that any trace of water left
over would remain in that cold finger which is baked between each
sample. If you have water, baking should get rid of it.

Hope this helps,

--
François FOUREL

UMR CNRS 5125 PEPS
Université Claude Bernard Lyon 1
2 rue Raphaël Dubois - Bât. Géode
F-69622 VILLEURBANNE Cedex - FRANCE

Tél : +33 (0) 4 72 44 62 42
FAX : +33 (0) 4 72 43 16 88
Email : [log in to unmask]



Georges L. Paradis a écrit :
> Good Morning Everybody,
>
> We're struggling to get good precision with our GV Isoprime/Dual Inlet.
> We are running 100ug carbonate samples/standards. We've achieved the
> specified +/-0.05 d18O (1 sd) on some runs, but it doesn't last. Our
> last 4 runs resulted in +/- 0.07, 0.12, 0.22, 0.09 d18O (1 sd).
>
> The mass spec seems to be working beautifully as indicated by the peak
> shape, all open analysis, and single cold finger volume anlaysis with
> ref gas on both sides of the dual inlet.
>
> We learned some techiques from our esteemed and helpful GV service
> engineers regarding the carbonate prep line. These include replacing the
> prep line and heating the prep line between runs. We saw improvements
> in performance, especially after heating the line. Unfortunately, the
> good data quality deteriorates quickly (several runs). We've
> experimented with heating the prep line continuously, heating between
> samples, and not heating at all. Our latest experiments involve
> replacing the 1/16" ss prep line with fat tubing (1/4" convoluted ss
> tubing). The fat tubing showed promising results at first, but alas,
> has deteriorated again (above data collected with fat tubing).
>
> I like the rapid pumpdown and definitive leak checking that the fat
> tubing allows. My thought was that the increased conductance and lower
> surface area to volume ratio of the fat tubing would be a great benefit.
>
> When using 1/16" tubing, it is standard 316 ss tubing that is cut with a
> dremel tool.
>
> Any ideas or suggestions for further testing would be greatly appreciated.
>
> Take care,
> Georges
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Georges L. Paradis
> ICPMS Lab Manager
> Department of Geological Sciences, Marine Science Institute
> Unviversity of California Santa Barbara
> Santa Barbara, CA 93106
> 805-893-7182
> [log in to unmask] <mailto:[log in to unmask]>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>