Chromatography Forum

Hi everyone

I have a GC (3900) connected with a reactor which produces H2 as a major gas.
when I analyze the gases produced from this system by this GC, the chromotograph displays H2, O2 and N2 peaks, even though the atmosphere of the system is inert by purging by Ar before starting.
The system should produce H2 only and sometimes O2 but by the very small quantity.
by the way, the peak area of H2 changes with changing the experiment but peak areas of O2 and N2 are the same in all cases even if I inject a sample from a pure H2 cylinder. I inject manually.
So how I avoid the sample contamination because I want to detec any trace of O2 produced from the reactor?

Thank all of you in advance

If the nitrogen peak is much larger than the oxygen peak, then this is a sign that you have air getting into your sample. You mention that you are injecting manually. How are you doing this.

Gas man

It sounds like you are not flushing your syringe with sample several times before you inject.

You are injecting residual air from the syringe or the syringe is leaking and is not gas tight.

Improve your sampling technique and I think you will find little but a tiny trace of air in your sample. That is unless your reactor is leaking air into its system.

best wishes,

Rod

Thank you sooo much for your helping.

I have tried to do your suggestion about the flushing of syringe several times before I inject and I saw the effect significantly. The peaks of O2 and N2 are much less than before but the problem is still there because I need to take a sample every 30 min to measure the quantity of hydrogen produced, so the flushing several times may affect on the result concerning H2 quantity. I should think about a way to evacuate the air from the syringe before the sampling...

Thank you again

new future,

Eliminate the manual injection. If you are really serious about measuring small amounts of O2, then you will have to get rid of the syringe. If your reactor operates at positve pressure, burp some of the gas into the 6 port sampling loop you have or if your reactor is operating at ambient, use a small pump (like KNF) to pull a sample through your loop, allow it to equilibrate, and inject.

Best regards.