Calibration of GC/MS Saturn Varian to analyse Pure Methanol

[Labotek]

Dear All,

I need to analyse pure Methanol (spécification is 99.5%) by GC/MS Saturn 2100T.

The problem I have is that when I inject the methanol, I have a very bad response (little pic of methanol)

The condition I use are :

Temperature of Injector : 200°C
Injector mode : Splitless

Colum temperature : 200°C during 5min then to 300°C @10°C per min

Volume injected = 0,5 uL

Could you help me please ?

Thank you very much

[Peter Apps]

What is the lowest mass in your scan range ?

What is it that you need to know about the methanol ? its purity ?, or the identities and concentrations of impurities ?, or something else ?

Peter

[Labotek]

My lowest mass is 30.

I need to know the purity of the methanol and the identities adn concentrations of the impurities.

Thank you.

[Peter Apps]

OK, you have strong ions at m/z 31 and 32.

Are you sure that the injection, inlet and column are working as they should ? How did you test them ?

The most likely impurity in methanol is water - you will not see that on the MS. How do you plan to determine the purity of them methanol ?, are you going to calibrate its peak against methanol standards of known purity ?

Just to be sure - do you have a solvent delay on the MS to protect the filaments ?

Peter

[Labotek]

I did not test the injection, column and inlet!
How I should process ?

I test the water by Kar Fisher.

I plan to determinate the purity of the methanol injecting 0.5 uL of the pure product, and I see other peaks, I will search the indentity in Saturn Library.

I have no solvent delay on the MS, because I inject the methanol pure (there is no solvent).

[Peter Apps]

O.5 ul injected splitless will completely overload the MS; its electronics then automatically reduce the ion intensity in an attempt to avoid ion-ion interactions in the trap, and the end result is that no matter how much methanol you inject, or how pure it is, you will never see a large peak.

If you want to measure how much methanol you injected (which is what a purity determination requires) then you would need to do a split injection with a very high split ratio, and use an FID as the detector. I can save you the trouble of trying it by telling you that the variability in the peak areas will not allow you to distinguish between 99% and 100 % purity. There is also the problem that you cannot get 100.0 % pure methanol to use as a calibrant.

You could use the GC-MS to identify and quantify the impurities, but separating them from the methanol can be difficult.

Peter