What I'm going to suggest is based on a swag (scientific wild ass guess)
since I don't know what kind of bath you are using to thermostat the
How close are the glass bottoms (of the tubes) to the bottom of the
thermostat bath; 1/8" or less? Am I correct in assuming the thermometer (or
thermocouple) is positioned somewhere in the corner near tubes #16 and #24?
If so, check the temperature (manually) at postions #1, #9 and #17 near the
bottom of the bath to see if the temperature there deviates from 24C.
(a) The heating element beneath the tank may be faulty or tend to overheat
locally. (b) The water in the tank isn't circulating properly. A mini
propeller driven be a dentist's drill or similar positioned in way to
ensure it doesn't hit the equilibration tubes (or the bath wall) might
already do the trick.
As mentioned above, this is a hunch based on what you are describing.
On Aug 23 2006, Soumitri Sarkar wrote:
> I am a new graduate student at Syracuse University, working on a
> multi-tracer approach to local groundwater problems. Among these
> tracers are oxygen/hydrogen isotopes, which we measure using a water
> equilibration bath (GFL 1086; 24 samples (3 rows of 8 bottles); 24
> Deg C; 8 hrs equilibration) connected to a Finnigan MAT252 mass
> spectrometer. Despite great efforts, I keep having the following
> problems (with great consistency….):
> 1. The first sample bottle of each row (samples 1; 9; 17) yield
> consistently bad numbers (several per mil off), despite gas pressure
> readings being ok.
> 2. The spread in d18O values for samples from the first row of
> bottles (samples 1-8) is very large, with variations of up to 1.5
> permil. The variations of samples in rows 2 and 3 (samples 9-24;
> except 9 and 17 - see above) are much smaller (within 0.3 permil).
> Changes in equilibration times does not influence or improve this
> pattern (tests were done for 4, 5, 6 and 8 hours).
> 3. Another pattern we consistently see is a slight but steady increase
> in isotope values for each individual row (it goes down again at the
> first sample of the next row), resulting in a cyclic pattern.
> 4. Last but not least, once in a while the 24 sample bottles seem to
> have very different amounts of gas in them but fortunately, this does
> not happen very often!
> Has anybody experienced similar problems? Or even better: does
> anybody know how to approach/solve any of these problems?
> I truly welcome any insights and suggestions on how to improve my
> oxygen isotope measurements with the existing equipment.
> Best regards,