Chromatography Forum

Hello,

i know that the headspace thechnique is only for liquid and solid samples, but is it also possible to analyse gas samples with this technique? The reason is that i want to use the autosampler and the calibration that is already made for the compound.

So the question is, will there also be an equilibrium between the liquid and the gas phase if only a gas sample is injected in the vial for the autosampler, where the liquid phase is milli q water? Or is it better to inject the gas sample by hand directly in the GC?

Thanks for your help!

Why make things complicated ? Just fill the vial with the gas. Good luck with doing that repeatably, but the you will find the same problem if you also have water in the vial.

Peter

What gas makes up the bulk of your sample?

I have pretty good success with repeatability when I'm filling vials with CO2 (heavier than air). To give you an idea of my technique, check this out (dispensing from a 3 L, Restek "basketball" cylinder, source pressure about 22 psig):

http://www.flickr.com/photos/61686435@N03/11311806455/

With the gas flowing the dip tube is pulled and the cap is snapped over the top then crimped down. I make sure to dispense at least 2 psig of gas into the vial to ensure sufficient purging of the 22 mL vial. 2 psig is about 400 mL of gas at 1 atm. The gas trapped inside the headspace vial is pretty much at atmospheric pressure. Using SPME to sample the contents, I can get 5% RSD or better on the impurities I worry about in CO2.

I have used this technique on a Perkin-Elmer headspace sampler but the sensitivity is not as good as what I get with SPME. That sampler has to pressurize the vial (thus diluting the sample) to get the sample into the chromatograph. SPME doesn't dilute the sample.

I developed a method where I jury rigged a septum sampling valve huh an accurate pressure monitor connects to a tank of gas and used valved Hamilton syringes to to withdraw specific gas amounts ( used P, V, T to calc mass of gas) which I then injected into headspace vials using GC-FID. You have to be careful to let the pressure equilibrate in the syringe.

R2 of 0.999 was typical.

So yes it can be done, and done reproducibly.

Samples were gas released from suspensions that injected into headspace vials.

Filling a vial with gas certainly can be done repeatably, but the two methods presented show that the procedure and the hardware have to be adapted to the job.

A question though - what advantage does transferring the gas from one container (a Restek sampler or a cylinder) to another (a headspace vial) confer ?. Is it just so that the vials can be queued and analysed automatically ?

Peter

Filling a vial with gas certainly can be done repeatably, but the two methods presented show that the procedure and the hardware have to be adapted to the job.

A question though - what advantage does transferring the gas from one container (a Restek sampler or a cylinder) to another (a headspace vial) confer ?. Is it just so that the vials can be queued and analysed automatically ?

While that was important too it was not the only reason.

In my case i needed different volumes of gas in the headspace vial to make a standard curve... and while I was analyzing gas, the samples being tested started off as lyophilized solids, which were reconstituted into suspensions which were then injected into headspace vials and the gas evolved in the vial on heating was measured.

Does that make sense to you?

Filling a vial with gas certainly can be done repeatably, but the two methods presented show that the procedure and the hardware have to be adapted to the job.

A question though - what advantage does transferring the gas from one container (a Restek sampler or a cylinder) to another (a headspace vial) confer ?. Is it just so that the vials can be queued and analysed automatically ?

While that was important too it was not the only reason.

In my case i needed different volumes of gas in the headspace vial to make a standard curve... and while I was analyzing gas, the samples being tested started off as lyophilized solids, which were reconstituted into suspensions which were then injected into headspace vials and the gas evolved in the vial on heating was measured.

Does that make sense to you?

OK, so you were using the gas as an external calibrant. How did you correct for the partition between gas and sample when you were analysing real samples - or did you know beforehand that all the gas would migrate into the headspace ?

Peter

Filling a vial with gas certainly can be done repeatably, but the two methods presented show that the procedure and the hardware have to be adapted to the job.

A question though - what advantage does transferring the gas from one container (a Restek sampler or a cylinder) to another (a headspace vial) confer ?. Is it just so that the vials can be queued and analysed automatically ?

While that was important too it was not the only reason.

In my case i needed different volumes of gas in the headspace vial to make a standard curve... and while I was analyzing gas, the samples being tested started off as lyophilized solids, which were reconstituted into suspensions which were then injected into headspace vials and the gas evolved in the vial on heating was measured.

Does that make sense to you?

OK, so you were using the gas as an external calibrant. How did you correct for the partition between gas and sample when you were analysing real samples - or did you know beforehand that all the gas would migrate into the headspace ?

Peter

Sample was suspended in water and the gas was VERY hydrophobic. In any case an equal volume of water was added to the standard headspace vials before the gas was injected.

So you did exactly what the OP was proposing, and it worked. The difference is that your samples were suspensions in liquid, and the OPs are only gasses as far as I can tell.

Peter

Filling a vial with gas certainly can be done repeatably, but the two methods presented show that the procedure and the hardware have to be adapted to the job.

A question though - what advantage does transferring the gas from one container (a Restek sampler or a cylinder) to another (a headspace vial) confer ?. Is it just so that the vials can be queued and analysed automatically ?

Peter

For me, the GC I typically use to analyze CO2 has 2 rotary valves with fixed-volume loops on them, 2 columns (a nonpolar phase and a polar phase), and 2 FID detectors. If I need to use mass spec. to confirm the identity of an unknown, I use the vial technique and SPME to get the sample into my GCMS.

Filling a vial with gas certainly can be done repeatably, but the two methods presented show that the procedure and the hardware have to be adapted to the job.

A question though - what advantage does transferring the gas from one container (a Restek sampler or a cylinder) to another (a headspace vial) confer ?. Is it just so that the vials can be queued and analysed automatically ?

Peter

For me, the GC I typically use to analyze CO2 has 2 rotary valves with fixed-volume loops on them, 2 columns (a nonpolar phase and a polar phase), and 2 FID detectors. If I need to use mass spec. to confirm the identity of an unknown, I use the vial technique and SPME to get the sample into my GCMS.

OK, so you do not need quantitative repeatability. Presumably you need confirmations rarely enough that it is not worth putting septum ports on the Restek sample containers.

Peter

Hi SommerTag,

I am Roger Bardsley an Application Chemist for Teledyne Tekmar working with our headspace instruments. We have application notes for detecting light hydrocarbons (methan, ethane, ethene and propane) in water. While developing this method I created standards by 2 separate methods. Please feel free to contact me about soultions to your question. I think it is possbile, however I will need a little more backgorund on your samples.

Filling a vial with gas certainly can be done repeatably, but the two methods presented show that the procedure and the hardware have to be adapted to the job.

A question though - what advantage does transferring the gas from one container (a Restek sampler or a cylinder) to another (a headspace vial) confer ?. Is it just so that the vials can be queued and analysed automatically ?

Peter

For me, the GC I typically use to analyze CO2 has 2 rotary valves with fixed-volume loops on them, 2 columns (a nonpolar phase and a polar phase), and 2 FID detectors. If I need to use mass spec. to confirm the identity of an unknown, I use the vial technique and SPME to get the sample into my GCMS.

OK, so you do not need quantitative repeatability. Presumably you need confirmations rarely enough that it is not worth putting septum ports on the Restek sample containers.

Peter

The vial method is quantitative. I can fill a vial with a standard and calibrate with the SPME sampling. Vial Sampling + SPME Sampling + Injection gives me better than 5% RSD on things like acetaldehyde, ethanol, dimethylsulfide, BTEX.