Hi Torsten,

For oxygen isotope analysis of ul-scale water sample, you could also see
the paper by W. Yang, H.R. Krouse and R.J. Spencer (1996), Chemical Geology
(Isotope Geoscience Section), v. 130, p.139-145.


At 08:12 AM 10/29/97 -0500, you wrote:
>>Dear Listmembers,
>>
>>I am planning on an oxygen extraction line for small water samples
>>(1-10mg). The liberated oxygen shall be analysed on an isotope ratio
>>mass spectrometer for 18-O, 17-O and 16-O. So far, I’ve only found the
>>papers by O’Neil & Epstein (1966) and Clayton & Mayeda (1983) related
>>to milligram quantities of water and 17-O, respectively. Now, I wonder
>>if the isogeochem pool can help me with the following questions:
>>
>>1.) Many laser ablation lines use fluorine gas instead of BrF5 for
>>oxygen extraction on minerals. How suitable is fluorine gas for water
>>samples? What is the minimum temperature required in order to avoid
>>the production of any oxy-halogen compounds?
>>
>>2.) After the recent lively discussion on Kel-F, I wonder if I can use
>>it for reaction vessels. I wouldn’t like to use Ni or stainless steel
>>since metal fluoride is quite hygroscopic. Does Kel-F stand
>>temperatures of up to 200°C for 10-20 minutes?
>>
>>3.) Who supplies Kel-F in Europe?
>>
>>4.) In case of fluorine gas: Does anybody know if N2 and F2 combine to
>>NF3 at 100-200°C? In other words, do I need to purify the fluorine gas
>>first?
>>
>>5.) Who can recommend me a source for molecular sieves suitable oxygen
>>gas purification?
>>
>>Any comments, further references, ideas, etc. are very much
>>appreciated!
>>
>>
>>Thorsten
>>
>>
>>Thorsten Agemar
>>Institut fuer Umweltphysik
>>Universitaet Heidelberg
>>INF 366
>>D-69120 Heidelberg
>>Germany
>>
>>+49 (0) 6221 54 6334
>>+49 (0) 6221 54 6405 (FAX)
>>[log in to unmask]
>
>Thorsten:       We still do water fluorination on few-milligram samples
>(for another reference, see Matsuhisa et al. GCA 42,173,1978). We use BrF5
>(why would you want to use fluorine?) in a nickel reaction vessel. Heating
>to about 200degrees C is necessary to decompose metastable products, such
>as ozone and peroxides formed in the explosive reaction.
>        If NF3 were produced, it can be recognized in the mass spectrometer
>by its fragment at mass 52 (NF2+). We purify oxygen before mass analysis
>using 13X molecular sieve (30/60 mesh, Anasorb), to remove any NF3 or CF4.
>                                                Bob Clayton
>