Chromatography Forum

Hi all,

I was wondering if anyone could help and shed some light. I am attempting to analyse H2S by headsapce GC but I am unble to get reproducible peak areas.

I am using a Varian CP3800 GC with a PFPD detector and headspace autosampler (even though I have tried manual headspace as well). The column conditions I have obtained from an application note:

Column: VF5MS 15m X 0.32mm x 1um
Oven Temperature programme: 30C (1min), 5C/min to 65C
Injector: 200C
Detector: 200C

These have remained constant but I have changed the injector temperatures, detector tempeartures, injection volume, split, headspace conditions.

The sample is a capsule which encapsulates H2S in cyclodextrin.

Whatever I change and then try to reproduce I am not obtaining the same or near enough the same detector response.

I am using the wrong column? Is there something blatent which I have completely missed as I am relatively new at this.

Thanks to anyone that could shed some light - if you would like to know anything else. Just let me know.

Thanks

elw

Hi elw

Hydrogen sulphide sticks to metal, so you have to eliminate all metal from the sample flow path. This includes the headspace sampler needle, its internal plumbing, 6-port valve (if it has one) and transfer line. You cannot connect the transfer line to the carrier gas supply line to the inlet, you have to take the tip of the line in through the inlet septum.

The only practical way of doing this is to get a headspace sampler in which all the internal surfaces are silcosteel or sulfinert treated - check out the Restek web site, and even then you are going to struggle at low levels.

The only success I ever had with hydrogen sulphide was to take the sample with a gas-tight sryinge that had a piece of empty fused silica column as a needle.

Peter

Thanks Peter,

This has definitely given me something to think about especially the second option. I assume you added the empty piece of column yourself?

Thanks again

elw

Hi elw

Yes, the silica needle was a home made affair. If I remember correctly I cut the bottom 10 mm off an ordinary metal needle, threaded the silica into that and glued it in. The original needle still seals to the syringe body and the sample sees only silica, glass and the teflon plunger tip.

Good luck !

Peter