HELP!!

[poorstudent83]

Please excuse my ignorance - complete novice here!

I'm getting 2% more of a substance than I should get. This is consistent.

My question is - how often should I be injecting a response factor solution. At the minute I'm injecting 3 response factor solutions (same vial then average the result) followed by a set of samples.

The solvent in the method is DCM - I'm concerned about evaporation of it over time - are screw cap vials really OK to use for GC?!

Any help would be appreciated!

[rb6banjo]

Yes, lots of people use screw-capped vials for GC with great success. Make sure they're as tight as you can get them. Should be fine for the short term.

You can compensate for evaporation by adding an internal standard to your solution. In that case, you measure the response of your analyte relative to the response of your analyte. If your solvent evaporates, it should affect both materials (analyte and IS) the same way and to the same degree. The internal standard should elute in your chromatogram where nothing else does and it's best to choose a concentration of IS that's somewhere close to your analyte concentration. Close is a loose term. You just don't want the IS to have a similar response to your solvent peak and you also don't want it to be so small that you have difficulty measuring it.

[Klaus I.]

[quote="poorstudent83"
The solvent in the method is DCM - I'm concerned about evaporation of it over time - are screw cap vials really OK to use for GC?!
[/quote]

You might be true with your concerns. To evaluate this, use crimped vials and inject only once per vial. Fill your response factor olution in three vials and make nor more multiple injections from one single vial. DCM as solvent is quite diffucult to handle.

[Peter Apps]

The simplest approach will be to inject a response factor solution after the samples and compare it with the results for one injected at the beginning.

That you need to take means of triplicate injections of the standard suggests that your repeatability is poor - do you also inject samples in triplicate ?

Peter