Dear all,
the glassy carbon tube and filling are necessary to prevent contact with
O-bearing surfaces in the hot zone (for D/H, they are not crucial). Now,
because of the tube-in-tube design, some He passes along the outer
compartment and some through the core of the reactor when the He is
directly fed from the top. With a number of samples reacted the debris
(mainly Ag) remains in the hot zone. Some of the glassy carbon is
converted to amorphous graphite. Both lead to a change in the flow
resistance of the reactor core and the flow partitioning changes.
Eventually the flow through the inner core becomes too low and the peaks
broaden to an extent that furhter measurements are not feasible. This is
why we changed the He feed.
In RCM 18 (2004) 2650 we proposed to introduce the carrier from the
bottom of the reactor tube so that the full flow is forced through the
core. This makes the design much less dependent on flow changes and the
peak heights become constant (at least for water, where the injected
amount can be controlled with good precision). This also improved memory
and it allowed to apply a linearity correction. We usually observe a
dependence of the signal upon ion current intensity of about 0.3 per
mill per Volt.
We also find that, with the full flow through the reactor, our results
for organic oxygen improves dramatically to a sub-per mill level. We use
benzoic acid as an internal standard and we can certainly run more than
200 samples without deterioration of the peak shape. Memory seems to be
negligible in this case.
Regards Willi
Dr W Meier-Augenstein wrote:
> Dear All,
>
> Prompted by what Gerard wrote, I thought I ought to mention what our
> observations are (though they may not be applicable to the particualr
> problem Adam is dealing with).
>
> Running solid samples on our TC/EA system, H2 peaks start to go
> pear-shaped first (while CO peaks still look fine and the numbers
> still make sense). This is predominantly caused by build-up of very
> fine silver 'pearls' at the bottom of the reactor (and silver coating
> the glassy carbon chips) when too many capsules miss the crucible.
>
> This used to happen after 50 to 70 samples and ties in with
> observations made by Paul Brooks who as a result fitted his machine
> with a self-built 'hoover' to remove silver in-situ (while switching
> the He in reverse flow).
>
> The only samples leading to a deposition of ash are wood samples,
> which doesn't surprise much. By the way, the maximum samle size we run
> is 0.15-0.2mg for stuff such as hair, wood, waxes.
>
> Not living in a lab such as Paul's with perks like a centrally
> supplied vacuum line, I looked at the reactor tubes and crucibles more
> closely and noticed that depending which way manufacturing tolerances
> add up, there are tubes in which the crucibles sit 'slanted' instead
> of 100% up-right.
>
> So, I ordered from a 'tame' supplier crucibles with slightly wider
> dimensions figuring should they get stuck I can always file them down
> (another case of "too much not being a problem"). With these tight
> fitting crucibles, I have managed to get 150 solid samples run before
> both the H2 peak and d2H-values go south.
>
> There is another (obvious and more elegant) solution to this
> particular problem but so far I haven't found the right people to talk
> to.
>
> Changing the way the He is piped in the furnace is supposed to help too.
>
>
> Cheers,
>
> Wolfram
>
>
> On Aug 1 2005, Gerard Olack wrote:
>
>> HI Adam--
>>
>> What size samples? We were running some amino acids awhile back, ca.
>> 300 ug each run, and were only getting 40 some down to 20 runs before
>> crucible had to be changed and then the eventual cleaning of the
>> reactor. The excess carbon filling up the crucible seemed to be the
>> problem. We had pushed through some plant material previously,
>> basically CHO so no excess C, and got 49 runs (one autosampler load)
>> before changing crucible and the reactor itself wasn't getting too
>> dirty either. We won't talk about mispacking the reactor, upside
>> down crucibles or that piece of kimwipe leftover from cleaning the
>> autosampler that fell into the reactor.
>>
>> In case any other garbage makes it's way through the reactor, we have
>> a cold (liq N2) trap between the reactor and the GC, via a 4-way
>> valve so the trap can be vented when it's warmed, akin to Werner's
>> set up in Isotopes Environ. Health Studies, vol 39, #2, June 2003
>> 85-104. Don't know if it helped in this case.
>>
>> Smaller sample sizes and larger crucible (still in hot zone) are the
>> only things I've tried. I haven't really looked for any particular
>> tricks of trade in the hair and feather studies (for example, Z. D.
>> Sharp, et al., J. Archeological Science 30, 2003, 1709-1716, amongst
>> others). Nor have I tried a water or nitrate "clean up" run. Back
>> to barite and phosphate for me--for now.
>>
>> good luck
>>
>> gerry
>>
>>
>> Adam Zoltan Csank wrote:
>>
>> > Hi Wolfram,
>> >
>> > Thanks for the thoughts, It is the CO peaks primarily but both do
>> it. > And the TCEA was in perfect working order (new packing on the
>> ceramic > reaction tube, baked out GC etc...).
>> >
>> > What seems to happen is that the little graphite crucible, that >
>> catches silver preventing it from running into the ceramic tube,
>> seems > to soften and expand preventing the sample from dropping all
>> the way > into the hot spot. Because we've never had this problem
>> before but it > has now happened twice to us and both times when
>> running continuous > samples of benzoic acid...I've been doing some
>> reading online and it > seems that graphite will adsorb aromatic
>> acids, such as benzoic acid, > I'm wondering if that has something to
>> do with it. We're going to be > trying a fresh crucible but if that
>> doesn't work, well three times > conspiracy and we'll know the
>> benzoic is the culprit, just not why.
>> >
>> > Adam
>> >
>> > On Jul 30, 2005, at 10:18 AM, Dr W Meier-Augenstein wrote:
>> >
>> >> Hi Adam,
>> >>
>> >>
>> >> Do both peaks broaden (H2 and CO) or only the H2 peak?
>> >>
>> >> If its only the H2 sample peak, I'd be tempted to say something's
>> >> messing up the He flow in the reactor (usually spilled silver >>
>> starting to build up at the bottom of the reactor tube); but for that
>> >> to happen after only ten runs is pushing it a bit.
>> >>
>> >> I take it all the usual suspects have been ruled out, such as air
>> >> breaking in due to problems with the AS isolation valve or a GC >>
>> column that hasn't been baked out for a while?
>> >>
>> >> If both sample peaks broaden and eventually vanish, my money would
>> be >> on the column. Perhaps the fact it happened when you run your
>> >> benzoate samples was merely a coincidence.
>> >>
>> >>
>> >> Cheers,
>> >>
>> >> Wolfram
>> >>
>> >>
>> >> On Jul 29 2005, Adam Zoltan Csank wrote:
>> >>
>> >>> Hello all,
>> >>> I've been attempting to run some Benzoic acid on the TCEA,
>> however, >>> it seems that after a little over ten samples have been
>> run my peaks >>> begin to tail until they they disappear. The cause
>> of this problem >>> is that the graphite crucible seems to be
>> softening and degrading >>> and so is not allowing the samples to
>> reach the hot spot. I am >>> wondering if maybe the benzoic acid is
>> reacting with the graphite >>> somehow. Or if , since we have been
>> cleaning and reusing our >>> crucibles, the crucibles themselves
>> weaken over time.
>> >>> I would be curious to know if anyone else has had this problem,
>> and >>> if anyone has any ideas as to what could be the cause.
>> >>> Thanks in advance,
>> >>> Adam
>> >>> Adam Zoltan Csank
>> >>> Department of Geological Sciences,
>> >>> University of Saskatchewan,
>> >>> 114 Science Place,
>> >>> Saskatoon, SK, Canada,
>> >>> S7N 5E2
>> >>> phone (o) (306) 966-5737
>> >>> (h) (306) 665-2797
>> >>> email <[log in to unmask]>
>> >>> http://homepage.usask.ca/azc461
>> >>>
>> > Adam Zoltan Csank
>> > Department of Geological Sciences,
>> > University of Saskatchewan,
>> > 114 Science Place,
>> > Saskatoon, SK, Canada,
>> > S7N 5E2
>> > phone (o) (306) 966-5737
>> > (h) (306) 665-2797
>> > email <[log in to unmask]>
>> > http://homepage.usask.ca/azc461
>> >
>>
>
>
--
.....................................................................
Willi A. Brand, Stable Isotope Laboratory [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, October 10-12:
http://www.bgc-jena.mpg.de/service/iso_gas_lab/gasir2005/index.shtml
.....................................................................
|
