Hi,
I was told that it is impossible to get a perfect seal of the transfer line into a ms...and i think I do experience exactly this.
We would like to obtain the mass spectra of various oxygenates, including alcohols. Alcohols often give a very characteristic ion at m/z = 31 and therefore we would like to scan from 30 upwards. The problem is that with trace components, massive interference of residual (?)air (oxygen @32) is experienced to such an extend that library searching in many cases becomes meaningless. Yes, scanning from e.g. 35 eliminates this problem, but then the very useful 31 alcohol ion is also not available.
This interference only applies to trace components.
Any suggestions?

Make sure the seal is as good as possible. SGE has metal ferrules for this purpose, Restek has a special design connection system, likely so do others. Of course I'd subtract out the background, gets rid of 32 ions, you'll see 31, we do.

I don't think you will totally get rid of the oxygen peak in your system. The key is to get the best seal you can. I typically get a 32 ion less than 1% of the 69 peak in a tune report using the standard Agilent graphite/vespel ferrule. I routinely scan from 10 amu units up. To get reasonable spectrum on a trace component you will need to subtratct out the background. To do this double click on the peak of the compound of interest, then double click just to the right of left of the peak of interest. Then go to spectra menu and select subtract. This should greatly reduce the background peaks and give you a searchable spectrum.