FID/Dead Volume Peak Inversion

[jcoutts]

Hi all,

We recently experienced some FID issues on our Thermo Trace GC and replaced a bunch of parts, and are now up and running. However, the peak associated with the dead volume is now much higher than our baseline (baseline = ~20,000 counts, and dead volume peak = ~45,000) rather than being lower as is expected - running air samples and it has always previously been lower than the baseline. Our zero-air, nitrogen, and hydrogen are all from generators that are functioning properly.

Any ideas on the cause/solution?

Thanks!

[Peter Apps]

An FID cannot detect air - it has 300 or so ml/min of air going into it all the time, so injecting air is not the way to measure dead time. The simplest explanantion is that the syringe has something detectable in it that has been carried over from previous injections, or that you are seeeing a pulse of septum bleed.

Peter

[chromatographer1]

The peak you are seeing may be from the pressure/flow upset from the injection causing a change in the composition of the flame (especially if your air injection is of a large volume).

Or it might be from something in the vapor sample you are injecting (more likely).

The larger size of the upset may be caused by the detector being more sensitive from the repairs you performed (very likely).

best wishes,

Rod

[Johnny Rod]

If you inject a large volume of oxygen you can see a blip from this, I can't remember the reason but it does show up as a tiny peak by FID. What are you injecting? Levels etc.?

[jcoutts]

Peter - by air, I meant gas samples as opposed to liquid samples where the solvent front would come into play.

Johnny - We are using a 6-way valve accessory by Valco for on-column injections of an ethanol (~50 ppm gas). The air that is used for our ethanol stream is supplied through a zero-air generator, and also runs through a CO2 and water scrubber. The valve does have an associated pressure with it when it flushes to take a sample, but we have been running in this configuration for about 2 years, and this is the first time to see the chromatogram like this.

Obviously there are sensitivity changes since we have replaced most of the FID with new parts, we have also found that it has become more sensitive to Acetaldehyde (one of the intermediates we look for from the reaction we run with the degradation of ethanol) but less sensitive to ethanol, which I also find curious...I would think you would either gain sensitivity or lose sensitivity overall.

Testing samples from our stream on our other GC hasn't shown any contamination in our stream, though the change in sensitivity may be picking some some other trace molecule in the system, I suppose.