H2 detection, sensitivity problems

[leine]

Hi,

we have recently purchased a 7890 GC/TCD for low level detection of H2. I am using a molsieve plot column, Ar carrier gas, TCD temp: 150 C, 40 C oven temperature, column flow rate: 2 mL/min, pulsed splitless injection. However, I can hardly detect H2 at levels under 1%. Any suggestions very welcome!!!

Thanks,

leine

[GasMan]

How is the sample being injected. What are the TCD reference and makeup flows.

Gasman

[leine]

Hi,

direct injection (10 microL volumue from headspace). And Makeup+reference are Ar.

Leine

[GasMan]

What are the flows for makeup and reference for the TCD. I think that the main problem is the amount you are injecting, it is too small. Is this being injected with a syringe, and if so, is the main body of the syringe that contains your gas sample being isolated from the atmosphere from the time you take the headspace sample to the time you inject into the GC.

Gasman

[tigerk2001]

What is the sample volume injected?
Increase sampling loop size if you use injection valve.

[larkl]

It would be unusual to inject less than 250uL for this type of testing. I'd probably use 500uL.

[tigerk2001]

I would suggest to increase the sampling loop size up to 2ml to see the difference in sensitivity.
With TCD, I already had to go up with 4,5ml sampling loop for low Helium and H2 detection in neon.
Let me know the results.
here is my contact in case
instrumentseoul@gmail.com

[Heikki]

I can detect H2 amount to as low as 2% with a mere 1µl injection (Ar carrier, TCD and a Restek ShinCarbon column). But I agree with the rest, that when injecting headspace samples, the volume can range from 10µl to a whooping 2ml.

I could even detect H2 (but only large quantities) using He as carrier with the above mentioned column.

[AICMM]

leine,

My first suggestion would be to switch to HID for this application since it would easily see the levels of hydrogen you are looking at. However, I have a strong bias in this since I make an HID.

If that is not an option, I would ask what column id you are using and what flow rates you have through the detector (make-up and reference?) The other thing I would suggest you try is a 50 ul injection with a split injection, perhaps 2 or 5:1. Does not necessarily change the amount on column but does a much better job sweeping H2 onto column and still keeps the injection volume within reason so that you don't get a pressure pulse.

Best regards.