Page 1 of 1

Missing 1 volatile component in P&T

Posted: Tue Apr 03, 2007 8:02 am
by kdamme
We use an Aquatek purge and trap system with GCMS to analyse a mix of 59 volatiles.

2 Systems (Aquatek 3100) work well: 59 peaks found


1 System (Aquatek Velocity with moisture trap) gives a problem:

After injection of a standard mixture we can find 58 peaks. We miss chloroethane. Peaks before the rt we expect and peaks after the rt we expect are all visible.

We then made a standard with only chloroethane (and Internal Standard): a peak is found, but our software cannot identify it as chloroethane and the masses we find aren't the masses we expect.

Purge velocity has been varied to obtain the same settings as the other Aquateks have.


Ideas???

Kevin van Damme

Posted: Tue Apr 03, 2007 4:47 pm
by JI2002
what are the ions and their ratios of the chloroethane peak you found? What's the result if you do a library search? Do you see chloroethane peak in the mix with direct injection in the injector port (without going through P & T process)?

Posted: Thu Apr 05, 2007 6:54 am
by Peter Apps
Hi Kevin

One possibility: if you are using an ion trap MS and the analyte peak is too large, or you have a co-elution with a contaminant, the MS software cuts the ion current automatically - giving a peak that is way too small, and a distorted mass spectrum. I have seen decane invisibly cut out of a GC - ion trap chromatogram because it was co-eluting with a siloxane contaminant.

Peter

Posted: Tue Apr 10, 2007 1:09 pm
by kdamme
Thanks for your input.

JI2002: the ions we found were 49 (100%), 51 (37%), 47 (28%), 63 (19%) and 62 (14%). Our library found 1,2-ethanedithiol as most probable compound. We tried to inject manually, but the solvent we used to make the standard (methanol) makes the most volatile compounds tail so much we cannot find proper peaks.
After P&T of a solution of only chloroethane the library found chloromethanesulfonyl chloride (66%). The 5th hit with probability of only 52% is 1,2-dichloroethane.

Peter: You were very awake when you wrote your post. Much more awake than we were :oops: ... Nobody noticed we have an ion-trap system in use. The other two systems are quadropole-systems. It might be part of our problem indeed :!: In our normal standard with 59 compounds there is a co-eluting peak. But... when we made a standard with only chloroethane our software still doesn't give the correct name (and ionisation, see above).

More ideas are welcome.

Posted: Tue Apr 10, 2007 3:20 pm
by Peter Apps
Hi Kevin

To make completely sure thatyou do not have a contaminant in the solvent, or bleeding off the column, do a blank solvent injection and check with full scan TIC for a peak at the target retention time.

Peter

Posted: Mon May 07, 2007 11:42 am
by kdamme
Peter,

it turns out that the problem is indeed the ion-trap. It seems that the molecule is unstable in the ion-trap so m/z =64 and 66 will never be found. We try to work with different mass (49 and 51).

As this is a typical ion-trap "problem" our library will never find the correct name. We are even lucky that none of the other components we measure have the same problem.

Our next challenge is to find out if this instrument has a linear range that is sufficient for our application

Thanks for your replies.