Small Signal Response?

[tehzap]

Hello,

I am trying to analyze the product gases of an alkaline water electrolysis plant, being hydrogen/oxygen mixtures. In this case gas concentrations of 0-1 Vol.-% H2 in (99-100%) O2 or the same amounts of O2 in H2 occur.

I found an old HP 5890 Series II with a TCD in our stock, which should do the job. As a column I am using a new HP-Plot Molesieve with Argon carrier gas. Unfortunately the signal I get seems to be very small, so that I am not really able to measure quantitatively.

The settings are the following at the moment:

Oven Temperature 35 °C
Injector Temperature 60 °C
TCD Temperature 180 °C
Column Head Pressure 18 psi
Split Vent 23 ml/min
Purge Vent 2.6 ml/min
Column Flow 3.1 ml/min
Make-Up Flow 3.0 ml/min
Reference Flow 15 ml/min
Injection Volume 1 - 2 ml

I changed all these parameters in the last weeks and injected a testgas with following composition:

0.2 Vol.-% H2
0.2 Vol.-% O2
0.5 Vol.-% N2
Rest Argon

The chromatogram I got with a injection of 1.5 ml testgas looks like that:



In this case the S/N-Ratios for the components are the following:
H2 - 29 (0.2 Vol.-%)
O2 - 5 (0.2 Vol.-%)
N2 - 8 (0.5 Vol.-%)

The signal response seems very low to me. Is there any way to increase the signal, am I injecting too much/too less or do I have to change the method? I would appreciate any hint to improve my results somehow.

Thank you very much in advance,
tehzap

[fahim.khattak]

Dear molsieve and argon carrier are the best choices you using. you just increase the sensitivity of your desired peak in your software. hope your problem will solve.

[aidnai]

I don't see anything wrong with your parameters, and I too feel that the peaks are pretty small. I have a 5890/TCD, although with helium carrier. Based on what I see, I would think your sensitivity should be at least 100x what you are getting.

Do you have any way to verify the TCD is functioning well? e.g. you used it for different analyses recently.

If you can verify that, I would start to question what you "know". How are you injecting 1.5mL? do you have a sampling valve, and how do you know the loop volume? How was the standard made? (why is it volume % instead of mol%?) etc.

Good luck,

[tehzap]

Unfortunately I have no real option to the test whether the TCD is working well, as the GC was kept in our stock for several years now and I don't know anyone who has worked with it before. But I assume the TCD itself is causing the problem.

I injected the 1.5 ml with a gas syringe through the septum of the GC. The testgas was delivered by our gas supplier. In this case I'd assume that Mole-% equal Vol.-% as I'am only trying to analyze "ideal" gases.

[aidnai]

Have you tried a syringe full of air?
Unfortunately, the CO2 and moisture will get trapped on the molesieve, but one injection should be ok. You can also bake it out afterwards.
This would at least rule out an issue with the test gas.

How are you checking your flow rates? There can be a few pitfalls with this...