Methane analysis with FID - strange peaks

[wbouma]

Hello GC experts out there.

I am a novice chromatographer tasked with analyzing greenhouse gases (CH4, N2O and CO2) on a new 7890 equipped with ECD, TCD, and FID; injections made with an autosampler. My FID signal for CH4 has recently been doing some very strange things. I am after any thoughts or explanations. I am also interested in getting some advice on troubleshooting and figuring out what might be my problem(s).

I am using packed columns ( PP-Q, 80/100) for this application. My system has got a TCD and FID in series. So my samples goes through the column into the TCD then on to the FID. The parameters for this method are as follows:

Oven temp:90
Inlet temp: 200
Inlet pressure: 20 psi

TCD temp: 200
N2: 40

FID temp: 300
H2: 30
Air: 400
make up (N2): 1

I had been using a run time of 5 mins. As I understood CH4 elutes very early and what I believe is my CH4 peak confirms that. When injecting my reference gases (Air Liquide standards, CH4 balance N2) I will see the CH4 peak and then usually nothing for the remainder of the run. Recently now, sometimes after my first sample has finished, I notice this strange instability in the baseline. It is the very gradual rise and fall that can last over 5 minutes to return to the normal baseline. On the 5 min run time this obviously carried over in to the next runs (image 1). I haven't determined it it is random or if it comes out at the same time. But it seems looks like a ghost peak or something like that. Here is a link to some chromatograms that I will be referring to http://www.flickr.com/photos/59956231@N05/

I then extended my run time to 30 mins just to see what happened. My first sample (img 2, 2ppm CH4 standard) ran out kind of normal with just this steady increase in my baseline overtime. Second sample (img 3, 10 ppm CH4 standard) shows the same peak around 10 mins. Same for the 3rd and 4th samples (img 3 & 4, these two were soil air samples). The dramatic effect of this 'peak' varies as the scale changes btw the various runs. It appears that whatever this is come out around 10 mins. However, after this I ran out 30 mins with no injection at all and still saw these strange peaks (??).

The TCD signal, also perplexing though not as bad as the FID, shows the injection/air peak and then something elutes out around 6 mins. I don't know that this peak is. As these are purchased reference standards I wouldn't have thought that it would be contaminated. Plus it shows up on both the standards and the samples. Any suggestions? My suspicion is that this is what might be causing problems on my FID as well, but I am not sure.

Any suggestions or troubleshooting would be greatly appreciated.


Thanks!

[chromatographer1]

Your peak at 6 min on the TCD and not seen on the FID is VERY likely to be water vapor.

The other peaks are something with carbon, extremely low levels,possibly solvents in the air, methanol, propane, and butane? DCM? perhaps DMS or COS. I would guess they are less than 5 carbon or oxygen or nitrogen or sulfur or hydrogen atoms in size., most likely not straight chains but branched chains, due to the broad peak shape. They may be hydrocarbons or ethers due to the lack of asymmetry.

That is all I can come up with at the moment.

Rod

[wbouma]

Thanks for your reply,

How do I get rid of water vapor? Or is this a problem?


As for the FID, any suggestions on how to get rid of the unknown peaks which are interfering with analysis of the CH4 peaks?

Thanks for your help.

[chromatographer1]

There are two things you can do.

Get a six port valve and configure to a backflush to detector column train

Or

configure it to a backflush adding a second column.

OR

Instead of adding a valve, simply extend the time of analysis and program the oven temperature to 200-230°C and hold that temperature for 5 minutes before cooling down to 90°C and beginning another injection.

The water at the levels you are seeing should not be a problem with the FID although if you try to repeat injections too quickly it may interfere with your TCD measurements.

Good luck,

Rodney George
Senior Research and Development Scientist