N2O method on Shimadzu 2014

[LMFultz]

We have a Shimadzu GC-2014 which we've previously used to analyze N2O and CH4 from the soil. It is set up with one 10 port and a 4 port valve. At the moment I'm only interested in the N2O peak. When we use the previously established method (included below) we are able to see the N2O peak, however when we try to do a direct injection we're not seeing the expected peak. For direct injection we are using a combi-pal auto-sampler.

Configuration:
10-port valve - this is where previous samples were injected and fill two 1ml loops.
HayeSep Q (2mm x 6ft)
ECD (N2O analysis) and FID (methane)

Value timing:
0.01min - 10 port valve directs sample into columns for analysis.
0.5min - 10 port valve closes to instrument.
1.45min - 4-port valve opens to detector.
1.85min - 4-port valve closes to detector to vent any acetylene/contaminants

Temperature:
Detector - 300C
Column - 70C with flow of 20 mL min-1
Injector - 200C

Injection:
1.25mL

There are two columns in the N2O line (one before the 4-port valve, and one after) We've recently added the combi-pack auto sampler and have been successful in getting it set up, however when we attempt to do a direct injection the peaks are not stable and one's we're expecting (N2O standard) are not present at all. We've tried adjusting timing and temperatures, but with no luck.

Any help is greatly appreciated and if there is any additional information needed please let me know.

Any help or direction anyone can be provided.

[carlos.teixeira]

Are your system suitability test accordanly the stated method?

[LMFultz]

Yes, we have a peak at ~3.7 minutes for N2O when we inject the sample through the 10-port valve. It's when we try to direct inject in to the GC that no N2O peaks occurs.

[AICMM]

Possible you can post the plumbing diagram? Sounds to me like one of your valves (my guess would be the 10 port) is in the wrong position when you inject.

Best regards,

AICMM

[LMFultz]

This is the flow from the 10 and 4 port valves. We we were injecting into the 10 port valve we were doing so through a plastic syringe. I also included an image of the 10-port value.

We did some maintenance on the injectors yesterday and realized that when it was initially set up they hadn't put a glass insert in the injector for the ECD and the FID one was broken. We fixed the inserts (although the only one we had for the ECD is older and a little worn at the top) and replaced the septum for both.

After running it today though we are getting any peaks regardless of where the sample gets injected.

[brewerbill]

You say "Yes, we have a peak at ~3.7 minutes for N2O when we inject the sample through the 10-port valve. It's when we try to direct inject in to the GC that no N2O peaks occurs." By "direct inject" do you mean having the autosampler inject the sample? If so, it sounds to me as though the 10- and 4-port valve timings are fine but you're having a problem filling the sample loops from the Combi-PAL. How is the autosampler attached to the 10-port valve? Are you taking enough sample to flush the dead volume in the transfer line?

[LMFultz]

Actually it turns out that it was several things.
We were switching from the injection through the 10-port valve to injecting using the Combi-Pal. The current configuration has the sample coming from the loops on the 10-port to the injectors on the GC, not going from the injector to the loops (although we have discussed doing this to allow us to do both CH4 and N2O analysis at the same time).

I want to thank everyone for their help and apologize for using forum space for what turned out to be a simple problem on our end. It turns out that no one had done maintenance on this GC in....a while. The glass-liner for the ECD injector was not there and the septa were in pieces for the ECD and FID injectors. We've corrected this error and are getting the expected N2O peaks.