Are these 47mm GF/F filters? If so, I would cut them in half and run half filters. Be sure to roll and pelletize with tin disks. I use louvered glass inserts and change them out every 50-75 samples because, as Robert said, they tend to form a glob of glass and potentially destroy the reactor column. Same change out routine goes if you are running 25mm GF/Fs.  The POM samples I run are really low level N (high C/N) so I tend to be at minimums for N signal. You may also try increasing your carrier flow as this may drive off some of the unwanted miscellaneous shouldering.




Anne M. Cotter

U.S. Environmental Protection Agency

Mid Continent Ecology Division, NHEERL

6201 Congdon Boulevard

Duluth, MN  55804


[log in to unmask]


From: Stable Isotope Geochemistry [mailto:[log in to unmask]] On Behalf Of Robert Panetta
Sent: Wednesday, April 12, 2017 8:40 AM
To: [log in to unmask]
Subject: Re: [ISOGEOCHEM] d15N and d13C of the POM


Hi Sofia,

Are the samples collected on GF/F filters? If so, the metled filters are likely clogging the oxidation tube.

Take out the oxidation reactor and have a look. If there is a big blob at the top of the oxidant then you'll have to change the entire reactor. Also start using inserts (aka "ash fingers" - with a plug of quartz wool to catch this melted glass. And swap that out frequently to save the main oxidation reactor.

Good luck,




On Wed, Apr 12, 2017 at 7:08 AM, Sofia Goncalves <[log in to unmask]> wrote:

Dear Isogeochem members,

We are analysing some difficult samples of POMs for d15N and d13C on our EA 2000 IRMS Delta V Advantage. These samples are colleted and analyzed on filters and have a high C/N.

Some of these samples gives peaks with strange shape and integration. The separation is not good either.

The analytical conditions are as follows:
- carrier flow 90ml/min
- oxygen flow: 200ml/min
- reference flow: 200ml/min
- ox. reactor: 1020 Celsius
- red. reactor: 650 C
- oven: 45 C

Could anyone help me improve this type of analysis?

Thank you very much.

Best regards,