Have you checked your peak shapes for gases coming out of the Nu Carb,
and whether your peak-centering routine is doing its job properly?
Running on the shoulder(s) of peaks can certainly lead to errors on the
order of tens of percent. One would expect precision to be degraded in
that case as well.
On 9/18/2018 10:31 PM, Florian Dux wrote:
> Hello all,
> I'm a post-doc in Melbourne, Australia, We run a Nu Instruments - Perspective mass spectrometer in dual-inlet mode in combination with the Nu Carb prep. device for both high precision conventional and clumped isotope measurements.
> Im hoping if anyone can offer some guidance and or advice in relation to a problem we are having with the Nu Carb prep device, in regards to conventional measurements.
> In essence the problem was twofold, our standards stated to look very bad about a month ago (whilst analysing forams).
> At the time I didn’t think much of it and put it down to a bad vial seal. However, after cleaning and replacing that and running a whole bunch of standards the machine stated producing very strange results. In the order of 600 per mill in 13C and 18O. Since then I have tried running more standards to see if it was just a “fluke”.
> However, it is still occurring however to a lesser extent now it starts art about 100 to 200 per mill on C and O and drops to about 40 and 50 per mill after 30 vials of standard. However, once a new batch starts the values tart high again and become lower as the run continues.
> I have tried the following to try to narrow the problem down:
> I have run zero’s (i.e. ref gas in both sample and bellows side at the same pressure. This gives 0.001 and -0.002 on delta 45 and 46 respectively (with SD at 0.004 and 0.007).
> And the 13C and 18O results come out at the present reference gas value. Which I know as I use a calibrated working gas.
> I have also run time scans over 1-2 hours and the depletion etc is very stable.
> This leads me to believe that the Dual-inlet side of things and the mass spectrometer are working fine and the problem isn’t there.
> I have then also introduced CO2 (water equilibrated CO2 with water) from an offline gas extraction rig into the sample bellows both without and with using the prep system (i.e. gas going through the water trap and the two cold fingers). Both of these methods produced values in the order that I would expect ~ 5 per mill C and 3 per mill O.
> The timing of the transfers might be slightly off since I’m doing it manually but, I would have thought that if there was a problem here it would have also done the same thing as before.
> This leads me to think that the problem is somewhere in the oven box or in the tubing just behind the sample valve block.
> I haven’t taken anything apart as yet apart from the normal valve block cleaning process and septum seal change. I guess it could be in the cold fingers that there is some “gunk” in there which is causing problems. Having said that I baked the cold fingers and water trap over the weekend for 24 hours so I guess its unlikely that it would be the cause.
> I’ve also considered yield and have plotted sample weigh, initial beam size and transducer pressure all against each other and have found no anomalies when comparing them to previous runs that were producing perfectly good results.
> I have also considered vial contamination but have now ruled that out, acid contamination also crossed my mind but is unlikely as I’ve used that vial for a while and havnt had any issues prior.
> At present I’m reaching the end of options of what I can easily do without taking too much apart.
> Have you ever come across such a stage behaviour or can you think of a contamination source/problem that would exhibit such symptoms.
> I would very much appreciate your input and thoughts.
> Thank you very much in advance, Florian.