Dear Roy,
thanks a lot for your detailed answer.
In the meanwhile I tried a cleaning technique with 3% HCl in an ultra sonic bath at 80°C followed
by extensive rinsing with water. After drying, the tubes were heated for 5 min in a small oven
sharply below the temperature the tubes deform. With this I had success! Whatever I removed
must be adsorbed near the surface, possibly graphite you mentioned. Although I don't think it
could be graphite because of the lack of oxygen in my tubes since pressure is far below 10-3
mbar. So I assume adsorbed CO2 on a strong inorganic adsober...

Kind regards

Jochen Schmitt

On 19 Aug 2004 at 17:53, Roy Krouse wrote:

> Dear Jochen,
> Perhaps you are aware of the following; There was wide discussion many
> decades ago regarding CO2 contamination produced during the preparation of
> SO2 for S-isotope analyses even when very pure S-containing minerals were
> combusted in quartz tubes.  This problem was attributed to the manufacturing
> process whereby molten glass was drawn over graphite rods to form a
> consistent internal diameter. i.e. the CO2 production was identified with
> oxidation of microscopic carbon on the interior of the quartz combustion
> tube.  There was also a recommended procedure whereby mild etching with HF
> was performed followed by rinsing with distilled H2O to remove a thin layer
> of glass on the interior wall of the quartz combustion tube. Practitioners
> claimed that this treatment significantly reduced the CO2 production.
> It is interesting that your blank delta 13C value was found to be near that
> of atmospheric CO2 whereas I recall that more negative values were found in
> the problem cited above. The sealing of sampling tubes is quite a different
> environment from the hotter oxygenated situation in a quartz combustion
> tube. A number of questions come to mind. Did the SO2 analysts overlook
> possible CO2 production when the prepared sample was transferred to pyrex
> tubes? Have you checked to see if your source of CO2 exists throughout the
> glass or whether it would  be reduced if the interior of the glass was
> mildly etched?
> Roy Krouse
> ----- Original Message -----
> From: "Jochen Schmitt" <[log in to unmask]>
> To: <[log in to unmask]>
> Sent: Thursday, August 19, 2004 8:01 AM
> Subject: Re: CO2 Outgassing of Glass During Heating
> > Dear Carl Johnson and others interested,
> > long ago in March there was an interesting discussion about CO2 outgassing
> > from glass tubes while sealing/heating them. I suffer from similar
> problems
> > with CO2 outgassing while sealing tubes for d13C analysis with CF-IRMS
> > (blank size ca. 20 ng CO2 per tube or as area 0.1-0.4 Vs). d13C values of
> > this blank is around -9 permil, so I rule out organic and inorganic
> > contamination on the surface but assume outgassing/desorption of
> atmospheric
> > CO2 from the glass itself.  Currently I use ordinary Pasteur pipetts tips
> > with barely 1mm diameter (the small diameter prevents me from using
> > pre-constricted tubes you mentioned) and heat them prior usage in a flame
> to
> > reduce the blank - with limited success. Since I want to change my tube
> > design I am wondering if there is a more suitable glass type. Has anyone
> > experiences with different glasses in terms of CO2 outgassing or is there
> > even a theory behind? e.g. are glasses with higher melting points better
> > because they can be heated to higher temperatures until they deform? or is
> > it a function of the glass chemistry?
> > Any helpful information is appreciated.
> >
> > Kind regards,
> >
> >
> > Jochen Schmitt
> >
> > Alfred-Wegener-Institute for Polar and Marine Research (AWI)
> > Dept. Glaciology
> > Columbusstr.
> > D-27568 Bremerhaven, Germany
> > Email: [log in to unmask]
> >
> >
> >
> > On Wed, 25 Feb 2004 11:21:16 -0500, Carl G. Johnson <[log in to unmask]>
> wrote:
> >
> > >Has anyone experienced outgassing of CO2 during the heating or sealing
> > >of quartz or pyrex tubing?
> > >
> > >We have recently been trying to seal such tubes and have noted a
> > >heating-temperature dependent CO2 background in our blanks. We can
> > >monitor this by other means (flushing the contents of the tube into a
> > >Licor CO2 meter) and see that if we even just flame the tubes above
> > >about 500c, we observe a CO2 signal. Repeated heating of the same area
> > >results in progressively lowered signals, so this is not a leak, but
> > >rather something coming out of the glass itself. Fully melting the tube
> > >to seal it generates an unacceptably large blank for our purposes. The
> > >effect can be somewhat reduced by working with tubes that are
> > >pre-constricted with a *very* thin wall in the area to be melted.
> > >
> > >Thanks for any input you may all be able to offer.
> > >
> > >Carl Johnson
> > >Research Specialist
> > >Woods Hole Oceanographic Institution
> > >Woods Hole, MA
> > >USA
> > >02543-1543

Jochen Schmitt
Alfred Wegener Institut für Polar und Meeresforschung
27568 Bremerhaven

Telefon Büro: 0471/4831-1345
email: [log in to unmask]


Moltke Strasse 22
27576 Bremerhaven
Telefon:        0471/ 9269838
                     0170/4470978 (Handy)