Strange new peak before H2

[emorris1000]

I have an inficon microGC with a column used for detecting H2. The column is MolSieve with Argon as carrier gas. The normal elute time is around 0.38 minutes. In the last few runs I have been getting a peak at 0.36 minutes in addition to my 0.38 minute H2 Peak. At first I thought this was some strange peak splitting (as discussed here: viewtopic.php?f=2&t=18413&view=next), but the first peak area seems to have no correlation to the H2 concentration. During a run with no H2 the peak was massive. It actually may be inversely correlated...I need to look at that more.

Anyways, I can't figure out what this peak is. Does anyone have any idea what it could be?

[rb6banjo]

If you initiate a run with no injection (just press start on the GC) does the "impurity" show itself in the chromatogram?

[emorris1000]

Hmm. Not sure how to do that on this GC (using Cerity right now which isn't great for customizing individual runs), but I should be able to sort something out later today. At worst I could just run it with only N2 in injector.

Any idea what could elute faster than H2 in a (normal-phase) molsieve column? Also it doesn't look like it's inversely related to the H2 peak.

[rb6banjo]

No. That's why I'm wondering if it's an electrical artifact of some sort.

[emorris1000]

Ok, ran the system with just carrier gas, and the peak is still there (shifted a bit, but everything shifted). I ran for a good 15 samples or so, the interesting thing is that the peak started high and then leveled out pretty low. Looking back on some previous runs where this peak existed it looks like it's high when the GC is run after not being used for a while, but after a few runs it goes down.

This sounds like an artifact in some way, but I don't see how an artifact could cause a peak like this on a TCD.

[muGC]

Perhaps one of your samples had that 0.36 minute contaminant peak? It sounds like each time you "blank inject", your carrier gas is slowly cleaning out the gas sampling valve during the injection time. Any chance this peak is helium?

[GasMan]

Helium will elute before hydrogen on a Mol Sieve column. Also, does the 'strange' peak have a gaussian shape. This would give an indication if this is something coming off of the column, whereas an electrical disturbance does not normally give a gaussian response.

Gasman

[Consumer Products Guy]

Disrupting the flow with a syringe needle can cause a system peak sometimes.

[emorris1000]

It looks like it may be coming from our house nitrogen. I wonder if there is a helium contaminant in it.....Time to call our plant I guess

[emorris1000]

Ok, so it looks like we figured it out. We spiked the system with helium to confirm that it was helium and lo and behold it was. This system has one Argon column and one Helium column, and the argon column is showing helium in it. At first I thought this was due to house N2 contamination, but at the suggestion of my supervisor we shut off the helium column entirely and then ran the argon column on its own.

It looks like helium from the helium column is somehow contaminating the argon line. My guess is that this is due to a fault in the injector, sample pump, or some kind of backflow preventer in the helium column, so that when the sample pump is running it is pulling helium backwards out of the helium column. Luckily this is a microGC which uses modular column assemblies and I have a spare column so it's just a matter of ripping it out and slapping a new one in....probably.