This is exactly what I did in Joe's lab.  I was pretty sure his combustion
column temp was up around 1000.

-----Original Message-----
From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On Behalf
Of Terri Rust
Sent: Monday, December 04, 2006 12:01 PM
To: [log in to unmask]
Subject: Re: [ISOGEOCHEM] Problems with GF/F filters

Hi Maria-
     We analyze thousands of gf/f  filters 
through our Carlo Erba EA/DeltaS mass spec 
system, and though they are not the easiest 
samples to deal with, we have developed a 
procedure that works well for us and gives us 
good results.  Our oxidation reactor is set at 
1020 degrees C and we use quartz inserts from 
Costech which we pack  ~1/2 inch of quartz wool 
into the bottom end.  For whole 24 or 25 mm size 
filters,  we analyze about 30 filters plus 
associated stds and then change out the insert 
quite soon after the sequence ends.  For  smaller 
subsamples of filters, we can analyze up to about 
50 filter sub-samples before changing the 
insert.  It is important to remove the insert 
fairly soon after analysis to prevent the molten 
glass from seeping through the quartz wool plug 
at the bottom of the insert and reacting with the 
inner wall of the quartz reactor tube.
   Good luck with your analysis.

Aloha- Terri

Terri Rust
Isotope Biogeochemistry Lab
University of Hawaii






At 11:24 AM 12/4/2006, you wrote:
>Hi Maria,
>
>I have not done this type of analysis in a 
>while, but at the time I considered banning GF/F 
>filters from the EA-IRMS in my lab.  The glass 
>caused too much trouble.  Whenever possible, I 
>encouraged lab users to use quartz fiber 
>filters.  They are more expensive, though.
>If you have to use GFF, make sure to reduce the 
>amount of glass fiber as much as possible (cut 
>away the excess filter, peel off the bottom half 
>of the filter, or scrape the sample off, if 
>possible).  I would still use the inserts, 
>because they will help keep the combustion tube 
>from cracking due to the glass.  Also increase 
>the run time per sample, because random late 
>sample peaks will come off as tiny gas pockets 
>are expelled from molten glass in the combustion 
>tube.  You don't want those tiny peaks interfering with your next sample.
>
>I never tried the 840degC temp suggested by 
>another list member.  Maybe that does work.  But 
>the lower combustion temp would worry me.  Worth a try, though.
>
>Good luck,
>Julie
>
>Marja Tiirola wrote:
>
>>Hello!
>>We have filtered lake water with Whatman GF/F filters, but the stable
>>isotope analysis results have not been promising. When the filters (1/3 of
>>the 4.5 cm diameter filter) are burned in the mass spectrometer, it seems
>>that the reactor is blocked even after a few filter pieces, since the
>>nitrogen values start to oscillate. Carbon values, however, seem to be OK.
>>We have bought quartz inserts, since we thought that the glass fiber could
>>remain there, but in our case it seems that inserts do not help us. And
>>when we change the insert after 10-15 filter pieces and cool it, we see
>>that the material which is remained in the insert is stone hard and
>>extremely difficult to remove, and furthermore, the insert has became
>>brittle and white and is easily broken.
>>
>>The furnace of our Thermo Delta Plus is set in a temperature of 900
>>degrees Celsius. GF/F filters are said to resist 500 degrees. Those who
>>measure GF/F filters, do you know the temperature of your furnace? Can it
>>be that we have too high a temperature for GF/F, which makes it amorphous
>>and blocks everything? Or do you have any other ideas, which can make the
>>analysis of GF/F so difficult? I would be glad to hear the opinion of more
>>experineced SIRMS users.
>>
>>
>>Marja Tiirola
>>Researcher, Ph.D.
>>University of Jyväskylä, Finland
>>
>
>--
>Julia Cox
>Dept. of Geology
>University of Georgia
>Athens, GA  30602
>706-542-1005