I have used a molecular sieve 5A column to separate N2 and O2/Ar from air
and then measure the N2 isotope ratio.  Using a continuous flow system and
GC-IRMS this resulted in a precision of better than +-0.1 per mil for N2,
often more like +-0.02 per mil.  Using a setup like this for O2 might work.
 The Ar mixed with the O2 might cause problems.  However, Ar and O2 can be
separated on a supercooled column, or with a capillary molecular sieve.  I
can send you reference for these latter methods if you are interested.
Certainly the method for N2 was very easy to use, just like injecting
samples into a GC.

Paul Brooks.



At 09:21 AM 11/30/98 -0500, you wrote:
>At 01:29 PM 11/30/98 +0100, you wrote:
>>Dear isotope geochemists,
>>
>>we want to analyse the d18O of molecular oxygen in soil gases.
>>We use a mass spec equipped with a double inlet system.
>>The problem we are facing is to purify the oxygen for isotope analysis.
>>
>>The only procedure for separation of oxygen from other gases (including
>nitrogen and noble gases) that I am aware of, involves conversion of oxygen
>to CO2.
>>If possible, however, I would like to analyse molecular oxygen.
>>(We run molecular oxygen extracted from silicates routinely on our mass
>spec, and it works fine)
>>
>>Can anyone comment on one of the following questions or point out
>literature that could help ?!
>>
>>(1) Are there any adsorbants, molecular sieves... that could selectively
>trap nitrogen and help to purify oxygen for isotope analyses?
>>
>>(2) How clean must molecular oxygen be to get reliable results for d18O?
>>
>>(3) What are the problematic interfearing species in the mass spec that
>could influence the d18O analysed?
>>
>>Any comments are greatly appreciated!
>>
>>Thanks.
>>
>>Rainer
>>
>
>Hi Rainer,
>
>At our lab, we are measuring different species present in the atmosphere
>for oxygen isotopes and therefore need to purified it. For your special
>case, it is relatively easy to remove the N2 matrix. You can first remove
>all condensable gases with a cold trap at liquid N2. The only remaining
>gases should be N2, O2 and Ar. Cryotransfert all these gases in a molecular
>sieve (5A) cooled at liquid nitrogen temp. Then you have to slush this
>molecular sieve at -123C using  ethanol (in general ethanol freeze at -117
>but in our lab for an unknow reason some old ethanol slush can reach -125C.
>I guess it has something to do with supercooling effect like water).
>Eventually you can use pentane which freeze at -130C (Cool it in a hood).
>All O2 should be released in 50 min. Use an other molecular (13X work well)
>to trap the O2 during the extraction.  With this inexpensive method Ar came
>with O2 and it is impossible to separate these two gases. You will still
>have around 5% of N2 but no interference in the mass spec has been seen at
>this level . For more information you can read  Cliff and Thiemens, Anal
>Chem 1994, 66, 2791 and Thiemens and Meagher, Anal Chem 1984, 56, 201.
>
>The other way will be to use a GC.
>         Joel Savarino
>        UCSD                             Villa La Jolla
>        Dpt of Chem. & Biochem.          3050 G Via Alicante
>        UH 5112                          La Jolla CA, 92037
>        9500 Gilman Drive                USA
>        La Jolla CA, 92093 0356          Tel. : +1 619 625 0760
>        USA                              Fax:   +1 619 625 0760
>        Tel.  : +1 619 534 6053
>        Fax   : +1 619 534 7042
>        E-mail: [log in to unmask]
>
>