If I understand you correctly, the IRMS chromatogram of injected n-hexane looks the same as your last FAME run.
By the way, what did the FID chromatograms look like of both the last FAME and the hexane injection?
Or asking outright, do the FID chromatograms continue to look OK, i.e. as expected and as before the problem started?
If the answer is yes to the above your problem is most likely interface related.
You see, problems at the injector (leaking septum, leaking split valve) or a leaking backflush valve (connected to the X-piece) will result in reduced peak size but usually not in a complete disappearance of all (major) peaks. Incidentally, to avoid fractionation in the injector it's a good idea to inject splitless and open the split subsequently (e.g. after 15s).
So, either a teensy-weensy piece of fused silica (FS) is blocking the X-piece or the X-piece is leaking He like it's goign out of fashion. If the X-piece is a full metal Valco job leaks at the screw connectors that develop after a few runs ending at 300+ C are not uncommon. That would be option 1.
Option 2, the FS connecting the X-piece to the reactor tube has become blocked reactor-side, either by "condensed" partially combusted sample material or the tip has melted and collapsed. Positioning of this piece of capillary is crucial.
Option 3, the oxidation reactor is exhausted and wants replacing.
Considering you have run a few samples after the problem manifested itself, my suggestion would be rebuild the interface and heat out the interface capillaries anyhow since there is a change of uncombusted sample material break-through into the interface.
Step 1: Replace both ox and red reactor tubes making sure that up-stream connecting capillaries come to end within the ceramic tubes but don't touch the metal wires. Also replace the capillary between X-piece and ox reactor.
Step 2: Check the X-piece for any debris before re-assembly; inspect the ends of all capillaries to make sure they are not crushed; neither should they have extraneous FS prodruding from the cut.
Step 3: With the IRMS disconnected from the interface (moving capillary open split), the reactors at operational temperature (with the ox reactor oxidised) and the back-flush closed (i.e. GC connected to the interface), work from the ox reactor towards the IRMS heating the capillaries with a heatgun (carefully).
Step 4: Connect the IRMS to the interface and run a magnet scan from mass 12 to mass 50.
Step 5: If step 4 gives your system a clean bill of health, load your CO2 settings, set the GC injector to split mode and inject some gas (collected in a test tube from a gas tap for a bunsenburner in your lab) using you current GC standby parameters and see if within the dead volume time equivalent of your system you detect a CO2 peak.
All going well you should then be back in business.
From: Stable Isotope Geochemistry [[log in to unmask]] On Behalf Of Melissa Bautista [[log in to unmask]]
Sent: 20 January 2008 00:11
To: [log in to unmask]
Subject: [ISOGEOCHEM] GC-c-IRMS
Thank you for replying. I have answered your questions as best I can below. Our GC is a Thermo Trace Ultra, The combustion interface is the Finnigan III, and the IRMS is the Delta V Plus.
1. What is the temperature setting for the injector? 260C
2. What's the average concentration per FAME in your FAME mix (in nmol/microL)? Unsure. Samples were extracted from both seagrass and mangrove sediments, but I only received them ready for methylation. I was told that the total lipid weight was between 1-2mg.
3. Have you looked what the chromatogram looks like for a dummy run (no injection or neat solvent only)? I have run hexane only with similar results (can send you the Isodat run file if that would help).
4. Same as point 3 but for single compound sample, say 10:0 or 12:0. I have not done this, but will do so now.
5. Is your system fitted with a FID and a cross-piece prior to the oxidation recator? Yes, FID and cross-piece.