TCD Response Factor Stability

[Foobs]

I have a Varian 3800 TCD GC with helium as carrier gas. Recently I used a 6-way valve to inject various standards of nitrogen, hydrogen, and carbon monoxide to the GC and was surprised to discover that the response factor (area / concentration) declined with increasing concentration. For the CO and N2 the decline was minor: a few percent over a near-100% range. For the hydrogen it was more like +/- 15%.

Is this something others have seen?

Thank You

[chromatographer1]

Yes, Foobs, you are seeing nothing new.

Hydrogen with helium carrier is not linear, in fact, as the concentration goes below 10% the response actually reverses and becomes negative in direction. The crossover is at about 7% and is dependent upon the design of the TCD and the size of the sample and the amount it is diluted as it passes through the column according to my present understanding of the phenomenon.

To measure hydrogen with linearity you should use argon or nitrogen. For maximum sensitivity and linearity you should use nitrogen.

With tongue in cheek, it is always good to hear of another confirmation of the laws of physics. In this tumultuous world it is comforting to note that some things don't change.

best wishes,

Rod

[Foobs]

I've read a great deal about hydrogen, but I hadn't heard anything about other components. Admittedly, those are smaller deviations, but not inconsequential if one is closing a material balance. I didn't see any reference to it in the Dietz paper...

I haven't seen a negative peak in my work with H2, though I have on occasion seen a double peak.

[chromatographer1]

If you are curious about the historical "W" hydrogen peak effect, I can share chromatograms showing this if you send me your email address.

But this fact of hydrogen response is well documented and is not necessarily pertinent to the problem at hand.

TCD response is highly dependent upon the design of your detector. Linearity over a wide range of measurement is not a hallmark of this detector in general.

You might consider using a different detector for the mass balance analysis you appear to be addressing. Other detectors are more linear and more singular in their response and might be appropriate in your project. There may be other considerations involved but I could direct you in a different direction if you are interested. I know contacts in the instrument industry who may be able to assist you.

best wishes,

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

814-359-5737 voice
814-359-5459 fax
rodney.george@sial.com

[Foobs]

If you are curious about the historical "W" hydrogen peak effect, I can share chromatograms showing this if you send me your email address.

But this fact of hydrogen response is well documented and is not necessarily pertinent to the problem at hand.

TCD response is highly dependent upon the design of your detector. Linearity over a wide range of measurement is not a hallmark of this detector in general.

You might consider using a different detector for the mass balance analysis you appear to be addressing. Other detectors are more linear and more singular in their response and might be appropriate in your project. There may be other considerations involved but I could direct you in a different direction if you are interested. I know contacts in the instrument industry who may be able to assist you.

best wishes,

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

814-359-5737 voice
814-359-5459 fax
rodney.george@sial.com

That's very good to know. The only thing I had was the Dietz paper and it said that response factors were independent of concentration (though it was only saying within 3%). We were concerned about what we saw because we didn't see anything about it in the literature. But it sounds like that is normal, so I feel better.

Thanks