detect hydrogen in 5890 using TCD

[hank76]

Folks,

I am trying to detect hydrogen ~1% in argon and air using HP 5890. I am using Ar as carrier gas. The hydrogen signal is strong however the tailing problem makes hydrogen peak overlap with those of O2. Can you tell me how to solve the tailing problem? What is the operation conditions that you used for this instrument, such as TCD and Injector temp, carrer and refernece gas flow rate, and headspace pressure?. The column is packed molecular sieve.

Thanks.

Hank

[mahona]

Hank,

Is it possible to reduce the size of the sample you are injecting on the column? If you are using a gas sampling valve, reduce the sample loop size or using an inlet dilute the sample some. You can try running the column cooler to separate the peaks, but it may make things tail even more. I would aim for 35 to 40 ml/min flow through the column at your operating temp. I hope this helps.

[hank76]

Thank you for your suggestion.

Actually, I am using a low oven temperature, which is 35 C. The low column temperature does help peak separation. But the peak tailing is still there and peaks overlap when they are big. The carrier flow has been tried from 5-20 sccm. They don;t solve problem.

Could it be the leaking or heating part problem on TCD?

Thanks.

Hank

[mahona]

Hi Hank,

The first thing I would try is: Reduce the sample size if you can. You may need to give up a little sensitivity up to get the good separation.

Next (if you haven't already done it), change the sensitivity of your detector to low. This may help a little.

I assume the column is new and properly conditioned??

Is this the first time you've tried this analysis on this instrument with your column setup? Maybe you need a different kind of column or different length. I can detect hydrogen in argon on a 24' MS 13X packed column with carrier of 35ml/min (same as 35 sccm). I use a gas sampling valve with a 5mL loop at 20psi to get my sample onto the column. My detector temp is 250C.

On a side note: Agilent's recommendation for setting the TCD reference flow is 1.5X the carrier flow.

I would not think that its a problem with the detector malfunctioning since the separation of the components takes place in the column...but ensuring the correct flows thru the detector is a good idea... What is your baseline signal on your TCD??

Best of luck...

[AICMM]

Hank,

From the fact that you describe an Inlet temperature I would guess that you are trying to do manual injections. If the volume of the injection approaches 1 mL, then my guess is that you are putting too much pressure in the inlet and it is smearing out the peaks. If your sensitivity is good, reduce the injection volume. If the sensitivity is only sufficient or you need to go lower, buy a sampling valve to do the injection. I had a customer doing hydrogen in ground water this way and we tried both inlet and valve injection and the valve injection provided much sharper peaks.

Best regards.