HI Dave--
Well, Willi does better than I on the TC/EA end of things. And there
have been earlier threads that can be helpful, H-dev: 7/14/04 (Paul
Brooks--proper drift corrections) and 8/18/05 (Chromium), TCEA:
6/19/03.... There are others too, along with hints to speed things
along (e.g. don't do a peak center every run--but remember that can
through off the timing since the peak centering resets the clock).
Here I go...
For the H-device, we inject 0.5 uL samples, get about 60+ injections per
ref gas load--or 20 samples in duplicate, and each run goes for ~15
min. After correction (memory effect), duplicate values usually within
1 per mil of each other . There is a little memory effect--we are doing
multiple needle rinses after each injection, including doing a acetone
rinse an injection into evacuated waste vial, as well as rinsing needle
with sample (got this off list too, but can't find thread right
now)--and some drifting over time. We run in "manual" mode, that is we
set the mass 2 voltage to 3500mV (400 mV window) for both sample and
standard, we have to "pre-compress" the sample bellows prior to the step
in acquistion.isl script that leads to automatic bellows adjustment
(Isodat has issues with adjusting the sample bellows). Once the ref gas
bellows drops bellow 30 percent max volume, the "drift" picks up. If
you use a plunger-in-needle syringe, like we are, don't insert it too
deeply when injecting into the H-device--it's too hot and can damage the
needle. Our setup is not fully optimized for the H-dev, so it can do a
bit better.
For the TC/EA, I took out the graphite funnel that's normally on top and
stuck in piece of glassy C tubing, topped with a silver wool plug (think
GC insert), and it ran ok. I'm still running in quadruplicate (larger
memory corrections, and I still a little leary), ~10 min run times, 300
to 400 nL injections, no sample dilution, inject as deep as possible to
keep water from condensing in the injection port (thanks Jon), slow flow
rates a little and have GC at 85C. We are usually getting std dev < 2
per mil on the HD, and <0.2 for the 18O, and could do ~40 samples (~180
runs total, with standards) per set. There was some drifting,
especially near the end in 18O, mainly because of CO broadening and some
tailing. Occassionally, there was the bad run--like the water didn't
inject cleanly and slowly evaporated into the system. Again, not fully
optimized. I'd expect better results with the glassy C tube designed
for water injections--especially if it's used with the bottom He-inlet
connector (IVA is still working on their's). There have been a couple
postings earlier-one reporting problems with salt solutions, and another
saying they had no problems running serum samples--if I recall correctly.
Sorry, never tried equilibration for DH on gasbench--only CO2 with water
(usually std dev ~0.05 per mil)
Other things to note, our TC/EA furnace was replaced recently--it was
warping, and we replaced it while we could still swap out reactors. It
lasted ~3+ years. (rebuilt, from Thermo, <$4K)
take care
gerry
David A. Mucciarone wrote:
> Can anyone jump in and compare the Finnigan H/Device with the gas
> bench/TCEA or GV equivalent?
>
> Thanks,
>
> Dave
>
> At 11:20 PM 2/16/2006, you wrote:
>
>> Hi,
>> I just would like to comment to Anindya's remark that "/certainly
>> TCEA cant be an alternative to more precise gas bench/". The daily
>> routine precision in our TC/EA is 0.5 per mill for D/H from water
>> samples. We run more than 200 injections per day with a net of 30
>> samples, each of them injected 3 times. The other half of the
>> injections is standards (the sequences are optimized for performance,
>> not throughput; given the precision we could afford to inject each
>> sample only once, increasing the net number of samples processed for
>> D/H to 90 on a single mass spec). For 18O, precision is in general
>> better than 0.1 per mill with fewer injections due to the longer
>> elution time of the CO. With a single glassy carbon tube we have made
>> more than 3000 injections.
>> My conclusion: This is my method of choice for the times to come. It
>> is so much more convenient compared to all other techniques (I think
>> I have tried them all) with zero compromise in precision.
>> Willi
>>
>> Anindya Sarkar wrote:
>>
>>> Hi Penny
>>> With a relatively new machine and gas bench we get typically 1.5 %o
>>> SD that is the kind of SD one gets in dual inlet as well. But we
>>> have been able to bring it down to even 1 %o level also. I hope u
>>> are using a fixed volume of water (we use 500ul and our experiments
>>> show that if you go below 200ul, values systematically change). we
>>> also bake the column after say about every 100 samples. But ur
>>> problem it looks more with platinum catalyst. The sticks supplied by
>>> thermo are not like hokko beads but a thin coating. we have reused
>>> each for at least 10 analyses and yet got very good results. but we
>>> also noticed that from some sticks coatings are peeling off. why not
>>> try with absolutely new sticks and run several of ur lab standards
>>> and see the reproducibility? as i understand TCEA will have more
>>> errors than gas bench (~2%o) but we got ~3%o on limited analyses;
>>> possibility of water evaporating in up part of furnace b4 it reacts.
>>> certainly TCEA cant be an alternative to more precise gas bench.
>>> ---------------------------------------------------------------------
>>> Anindya Sarkar
>>> Associate Professor
>>> National Facility of Stable Isotope Geochemistry
>>> Department of Geology and Geophysics
>>> Indian Institute of Technology
>>> Kharagpur 721302
>>> West Bengal, INDIA
>>> Tel.: 0091-3222-283392 (O) 283393 (R) 220184 (R)
>>> Cell: 09434043377
>>> Fax.: 0091-3222-282268
>>> http://anindya-sarkar.tripod.com
>>> ------------------------------------------------------------------
>>> ----- Original Message ----- From: "Penny Higgins"
>>> <[log in to unmask]> <mailto:[log in to unmask]>
>>> To: <[log in to unmask]> <mailto:[log in to unmask]>
>>> Sent: Friday, February 17, 2006 3:57 AM
>>> Subject: [ISOGEOCHEM] Gas Bench hydrogen measurements
>>>
>>>
>>>> Hello Isotopers,
>>>>
>>>> We've been attempting to use our Gas Bench to do hydrogen analysis
>>>> of waters using platinum catalysts and headspace equilibration at
>>>> about 32degreesC. When we started, we got acceptable numbers with
>>>> good within analysis (10 peaks) standard deviations. Values plotted
>>>> sensibly along (or very near) the meteoric water line. As time went
>>>> on, our standard deviations have gotten much worse (as much as 7
>>>> permil) and our data very seldom plot reasonably close to the
>>>> meteoric water line. Often, hydrogen data plot well above the
>>>> meteoric water line. We have gotten great results from oxygen
>>>> analyses of the same samples.
>>>>
>>>> We've just spent a great deal of time and effort to ensure that our
>>>> sample gas is totally dry before it enters the mass spectrometer. I
>>>> also let the catalysts sit in our oven (60degreesC) for two days
>>>> before placing them into the vials with our water samples. We got
>>>> better standard deviations for our analyses this time (between 2
>>>> and 3), but still have unreasonable scatter around the meteoric
>>>> water line.
>>>>
>>>> Questions:
>>>>
>>>> 1) What is a generally acceptable standard deviation for d2H from
>>>> water?
>>>>
>>>> 2) Could the problem be with our catalysts? I thought that, by
>>>> definition, catalysts were reusable .. but maybe not. We've only
>>>> done about 5 analyses with each catalyst.
>>>>
>>>> 3) Are we going to have this same problem when we start running
>>>> waters using our TC/EA instead? or are we going to be happy that we
>>>> made the change?
>>>>
>>>> If anyone out there in cyberspace has dealt with and corrected this
>>>> problem, we would love to hear from you!
>>>>
>>>> Thanks,
>>>>
>>>> ~Penny
>>>>
>>>> *******************************************************************
>>>> Dr. Pennilyn Higgins
>>>> Research Associate
>>>>
>>>> "SIREAL"
>>>> Stable Isotope Ratios in the Environment Analytical Laboratory
>>>>
>>>> Department of Earth and Environmental Sciences
>>>> University of Rochester
>>>> 227 Hutchison Hall
>>>> Rochester, NY 14627
>>>>
>>>> [log in to unmask]
>>>> <mailto:[log in to unmask]>
>>>>
>>>> Office: 209b Hutchison Hall Lab: 209 Hutchison Hall
>>>> Voice : (585) 275-0601 Outer lab: (585) 273-1405
>>>> FAX : (585) 244-5689 Inner lab: (585) 273-1397
>>>>
>>>> http://www.earth.rochester.edu/SIREAL/index.html
>>>> *******************************************************************
>>>>
>>>>
>>>>
>>>> --
>>>> No virus found in this incoming message.
>>>> Checked by AVG Anti-Virus.
>>>> Version: 7.0.344 / Virus Database: 267.15.6/257 - Release Date:
>>>> 2/10/2006
>>>>
>>>
>>
>>
>>--
>>.....................................................................
>>Willi A. Brand, Stable Isotope Laboratory [log in to unmask] <mailto:[log in to unmask]>
>>Max-Planck-Institute for Biogeochemistry (Beutenberg Campus)
>>Hans-Knoell-Str. 10, 07745 Jena, Germany Tel: +49-3641-576400
>>P.O.Box 100164, 07701 Jena, Germany Fax: +49-3641-577400
>>http://www.bgc-jena.mpg.de/
>>http://www.bgc-jena.mpg.de/service/iso_gas_lab/
>>.....................................................................
>>GASIR 2005 in Jena:
>>http://www.bgc-jena.mpg.de/service/iso_gas_lab/gasir2005/index.shtml
>>.....................................................................
>>
> ***********************************************************
> David A. Mucciarone
> Stanford University
> Dept. of Geological & Environmental Sciences
> 450 Serra Mall
> Building 320, Room 118
> Stanford, CA 94305-2115
> 650-723-0817 (O) 650-725-2199 (F)
> http://pangea.stanford.edu/isotope/dam/sil_home.html
> ***********************************************************
>
|
