Dear Uli,

See below.  Hope this is helpful.

Tom Brenna

>         1) First of all, is there anybody else who can report on
>            experience with similar problems and a similar solution, i.e.
>            more or less a single capillary interface? Not that I want
>            Carl's contributions being depreciated in any respect; its just
>            that additional reports would give more weight!

We have used the single capillary interface for more than 5 years in
GC-combustion-IRMS and in a system with much more capillary plumbing,
involving 2 GCs connected in tandem for pyrolysis experiments.  The
performance is excellent and the reactors sometimes do not require
re-builds for periods greater than 1 year, as long as the temperature of
the surrounding furnace can be maintained continuously >400C or so.  On
occasions when power to the furnace goes out, the bare capillary in the
furnace often breaks due to thermal expansion/contraction and must be replaced.


>         2) When substituting the interface by a house made solution, how
>            long does it take to have stable system running again?

Depends on how experienced you are.  However, there are many fewer
connections to make (and to leak) so it is generally faster than the
commercial designs.  In Goodman's paper, there is a single capillary
between the rotary valve in the GC and the open split - i.e. no connections
necessary for the reactor or the water trap, which in his case was a
cryotrap.  That saves 4 connections and inner diameter changes compared to
a conventional reactor and a Nafion-trap (truth in advertising:  we use a
Nafion trap).

>            Are there any systematic deviations from the long term values
>            of the laboratory standards?

No.


>         3) Are there any recommendations where to obtain the necessary
>            parts from? I didn't find a vendor for the ceramic tubes yet.

We have reliable suppliers - I'll look it up and get back to you.


>         4) It seems to me that the tightness of the backflush valve is of
>            significant importance for the peak shape and hence for the
>            calculated d13C values. My experience is, that spraying Argon
>            is not exceptionally useful here as the problematic fitting is
>            located inside the valve. Furthermore in closed mode there is
>            still some He purge. Therefore leaks are difficult to detect.
>            The difference of m/z=40 between "Bfl on" and "Bfl off" (Thank
>            you Carl!!!) is most helpful here.
>
>            Now the questions: I don't get along using new fittings for the
>            valve. What's the usual lifetime of the whole valve? To make it
>            live longer would it be useful to put it outside the GC?
>            I'm asking because it could be a workaround for the moment.

If you put the valve outside the GC it is no longer heated and since the
analyte has not yet combusted it will condense on the walls of the
capillary leading to the valve.  In addition, Valco rotary GC valves are
made to be use at high temperature in a GC oven and must not be rotated at
room temperature or the sealing polymer (polyimide?) will be
damaged.  Lifetime for the valves in our hands vary over the years, they
have been known to perform well for many months.



>I know this is a lot of stuff, but I think there are a lot of people with
>a lot of problems as well. Many, many thanks in advance, any hint is
>appreciated!
>
>Best ...
>
>         Uli Flenker
>         Institute of Biochemistry
>         German Sports University Cologne
>         Carl-Diem-Weg 6
>
>         50933 Cologne / Germany
>
>         Phone 0049/0221/4982-493