calibration gases + GC-TCD

[wvandenboer]

by means of elektrolysis we produce O2 and H2 and consume H2O. But there is always contamination from H2 in O2 and O2 in H2. And I need to know the concentrations of these contaminations.
I want to make a calibration curve for Hydrogen (0-2%) and Oxygen (0-1%) and i am using a GC with TCD and as a carriergas Ar. Now I'm wondering if I can use as calibration gases: 0-2% H2 in balance gas Ar and 0-1% O2 in balance gas Ar? Or do I have to use an other balance gas? Can you use the same gas for carrier gas and balance gas? And can I detect oxygen=O2 as I use Argon, because they have almost the same molecular weight....

[Johnny Rod]

It is always best to use the same balance gas and carrier gas. Talking about sensitivity, there will be no problem for hydrogen in argon. For oxygen you will get a low response as they have similar thermal conductivities, but maybe a larger injection volume will do it. Are you using packed or capillary column?

[wvandenboer]

yes, I am using a packed column:molecular sieve 5A
Maybe it is wiser to use He as a carrier gas and balance gas for the calibration of O2?

[chromatographer1]

Oxygen and argon are not easy to separate (I have posted chromatograms of this on the forum) when using Helium or hydrogen as the carrier gas. But if you are certain you will not have any argon present then the only problem will be the non-linearity of the hydrogen response, if fact it will change polarity entirely over a large range of concentrations.

Using Ar as a carrier when injecting O2 should give you a low response for O2 due to conductivity issues AND because there should be little or no focusing of the O2 plug since the matrix is Argon and they compete for the pores almost perfectly at the same rate. So your O2 peak is as wide as the time it takes your sample to purge the injection loop, and the height is tiny. I have not checked the conductivities, but it could be negative in height, requiring a change in polarity of your TCD.

You could use an Argon ionizing discharge dectector? ? ? AICMM, who posts regularly on the Forum, sells such a detector. He is the best one to advise you.

This might greatly increase your sensitivity.

If you want to get some better focusing of your oxygen peak you might try a Carboxen 1000 packed column. That will separate your Hydrogen and oxygen readily and may improve your oxygen peak.

best wishes,

Rod

[wvandenboer]

Thank you for the response. I think the best solution for me will be then:

To determine only the hydrogen concentrations of samples hydrogen(0-1%) in oxygen,
I need a calibration curve for Hydrogen, so I can use as calibration gases hydrogen (0-0.5-1%) in Argon and as a carrier gas Argon.
Because it is not necessary to see the oxygen peak in these samples hydrogen(0-1%) in oxygen. I only need to know the hydrogen concentrations and it is not necessary for these samples that I can see an oxygen peak.

To determine only the oxygen concentrations of samples oxygen(0-2%) in hydrogen. I need a calibration curve for Oxygen, so I can use as calibration gases Oxygen (0-0.5-1-2%) O2 in Helium and as a carrier gas Helium. Because it is not necessary to see the hydrogen peak in these samples, I only need to know the oxygen concentrations. It is not important here that I can split Helium and Hydrogen, I only want to see here my oxygen peak.

Do you think I can do it like this or do I have to approach it differently?

[chromatographer1]

Yes you can.

TO SAVE MONEY you might use hydrogen carrier instead of Helium carrier gas.

Likewise, to get better sensitivity for hydrogen you could use nitrogen carrier. This will also save you money.

best wishes,

Rod