I hope that you get a lot of answers from the very experienced labs but here
are my two cents. Our lab is less than a year old and I spent the last
several months getting the water analyses going for H and O using the
gasbench II with a MAT 253. I have no GV experience so I can't offer you any
contrasts between the systems. I have found our instrument and software to
be pretty easy to use. If you look through the isogeochem archives you will
find some long discussions about issues with Isodat hanging up. We had such
issues too and followed the advice of Gilles St. Jean to reinstall windows
with two partitions on the hard drive, and reinstall Isodat. The other thing
was to dump McAfee antivirus and go with Norton/Symantec instead. Since then
we have not experienced system hangups with isodat (knock on wood).
To answer some of your questions:
The equilibration heating block has 96 wells. You initially set the system
up to flush all of the vials with a mixture of CO2 in helium. This removes
the air from the headspace and provides CO2 for the equilibration. Once
flushed, the vials need to sit for about 18-20 hours (at 25C) before
equilibrium is reached. So you have to figure that it will take 24 hours for
the flush/fill stage. Then... the next day you can tune the mass spec and
start the analyses. This takes an additional 12 minutes per sample which
very roughly amounts to another 24 hours. So... if analyzing only oxygen
isotopes, your throughput is 96 samples per two days. There is plenty in the
literature that states that you can dramatically reduce the equilibration
time using a sonicator. Too bad they haven't integrated one into the heating
block. Maybe the next version?
We have done mostly 18-O measurements. As far as analytical precision goes,
that depends on what is meant by this. We perform 8 analyses each of three
different standards spread throughout the run and then perform a drift
correction on them. This gives us standard deviations of about 0.1 to 0.2
permil on 8 analyses spread out over a 24 hour period. So... technically you
can say that the standard error (standard deviation divided by the square
root of the number of measurements) of the measurements is less than 0.1
permil for multiple runs of the same sample. That being said, if you are
performing single analyses of samples in a sequence (as opposed to
triplicates) then it seems to me that the correct way to report the
precision is to use the average of the standard deviation (not standard
error) observed for the standards.
For H, I only have the data from the calibration of our working standards on
the gasbench (we are running our first large batches of samples for 18-O at
the moment). From our working standard calibrations I can tell you that we
can easily get standard deviations on 8 or 9 runs of less than 2 permil
after applying the SMOW/SLAP correction (this tends to increase differences
between measurements by a factor of 3 to 3.5). I like the gasbench so far,
but I'd like to be able to analyze smaller sample volumes in the future. For
18-O analysis you need 0.5mL of water. For 2H you need about 0.3mL.
I hope that some of this helps.
Anthony L. Michaud
W. M. Keck Foundation Laboratory for Environmental Biogeochemistry
Department of Geological Sciences
Arizona State University
Tempe, AZ 85287-1404
Arizona State University
Tempe, AZ 85287-1404
From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On Behalf
Of Valerie Claymore
Sent: Tuesday, September 06, 2005 7:58 PM
To: [log in to unmask]
Subject: [ISOGEOCHEM] water isotope analysis
We are in the process of deciding which company to purchase a new water
isotope mass spectrometer from, GVI or Finnigan.
We are looking at one mass spectrometer to run both isotopes.
The two systems that we are looking at are GVI's Isoprime coupled with the
AquaPrep and PyrOH or the Finningan Delta+XP coupled with the GasBench II
I am looking to contact all labs that run both water isotopes on a single
mass spectrometer, and hopefully close to one of these two configurations.
I would love to speak with you, and if possible be able to come and visit
your lab, to be able to get a better feel for each of these systems be for
we make the big purchase.
The things that we would like to know, are the normal questions.
What is the precision and acuracy?
What is the throughput, both in samples per day, and time per
Do you switch systems regularly, and how long does it take?
What type of downtime have you had?
How is the ease of use, both with software and hardware?
Please let me know if you have any experience with this type of
configuration, and if I may come and see it.
I am planning for a trip in late September, early October.
Thank you for your time,
Stable Isotope Lab
30 Gracefield Road
[log in to unmask]