Injection repeatability

[themont]

Hi to all im new to the forum,
Ive got a problem with an Agilent GC. If the instrument is not used for a couple of hours and I run a 2 stage sample I get a result, but if I repeat the injection out of the same vial a further 6 times injection one is different to the other six. When the GC is calibrated in the morning it is therefore prone to failure if its not been used much during the night. Ive tried new septa, liner, gold seal, syringe and split line. If you do a hexane blank that doesn't cure it either. The instrument set up is as per the EP. Once its had one sample run on it it works fine but you cant justify ignoring a result.

Any help and advice would be much appreciated.

[kubowicz.tomasz]

Hello

If you inject std/sample from the same vial that you used couple of hours before perhaps solvent (is it hexane?) evaporates and concentration is slighltly different.
Just prepare 2 vials with the same standard/sample and inject one in the morning and one in the afternoon.
You mentioned that GC is not in use for couple of hours...Does it mean that you switch everything off? (inlet/detector temp)

Regards

Tomasz Kubowicz

[GasMan]

You need to tell us a little bit more information.

Is the first run results higher or lower than the next 6 runs.

What is in your sample.

Gasman

[Bigbear]

I was told by Agilent, when using 300ul vial inserts, not to make more than 2 injections from the vial.

[MSCHemist]

You could have an active site and the first injection causes material to adsorb to it covering it up. This is especially common with silylation as the silylation reagents deactivate active sites and derivatizes junk and cause it to be eluted.

Another possibility is puncturing the vial allows it to equilibrate with the atmosphere and stabilizes it.

[tina.dang]

a

[Peter Apps]

If you are doing one analysis, then there should be no dead time in between injections. You should not be comparing data of morning injections to that of night injections when the GC instrument is not running any injections in between. If your intention is to compare the data, the instrument must be continually running injections, especially if you are running analysis with a temperature program.

Welcome to the forum - do you have hard evidence for these assertions ?

Peter

[kubowicz.tomasz]

Hello

If you are doing one analysis, then there should be no dead time in between injections. You should not be comparing data of morning injections to that of night injections when the GC instrument is not running any injections in between. If your intention is to compare the data, the instrument must be continually running injections, especially if you are running analysis with a temperature program.

I'm sorry but I can't agree...
It depends on method. For some methods you need to perform re-calibration after couple of samples and for other methods calibration is stable for months. Control samples can be run to check calibration. I wouldn't say that GC must be running all the time.
For example: what would you advice if control lab runs air monitoring method 6am and 6pm. Can they compare results?

Regards

Tomasz Kubowicz