agilent GCMS signal drops out

[sepscientologist]

I'm having trouble with reproducibility of my Agilent GCMS. Out of 12 injections, 9 were around 2.2 million in area and 3 were around 1.3 million in area
for the latest eluting peak. Zooming in on the baseline shows that the 3 injections that were much smaller had a sudden drop in baseline. Does anyone
have an idea of what is going on? I'll attempt to post images of the chromatogram and the zoom of the baseline. I know... I know... conditions, columns
compounds, concentrations etc. I'll post later if anyone is curious. Basically it is isothermal of three alchohols on an FFAP column at 80C

[Don_Hilton]

It looks like a drop in baseline on the yellow trace in the tail of the third peak as well?

What is solvent in use, the column, GC conditions and mass range in use?
And, what happens if you make a run by pushing the start button without an actual injection?

Does sequence of injections make any difference? (like does this only happen after the first or second sample inection in a sequence?)

[AICMM]

sepscientologist,

I would not be sure this is an MS problem, it could be an inlet problem.... What action brings you back to your starting point? Can you toss in another compound (possibly not a FAME) that elutes out there somewhere that you think is rock solid stable and see if it also tanks?

Best regards,

AICMM

[sepscientologist]

The solvent is methanol which elutes before my solvent delay. The C4, C5 and C6 alcohols are at 1%.

I ran an FID method just before this and the C.V. was less than 1% for 24 injections. The CV for the first
two alcohols are around 4%. The CV for the late alcohol is 22% because the baseline drops out three
times. The three times it dropped out was on the 6th, 8th and 11th injection of twelve trials.

Seems like it must be the MS since the FID data is so solid.

[Don_Hilton]

It looks like I failed to ask about the mass range collected. This almost looks like ion supression - particualry peak shape changing in the affected peak (more apparent surpression towart the front of the peak than later in the peak) - when one compares the blue trace to the others. It is as if something ouside the mass range detected is eluting at a some what irreproducable retention time and causing surpression through this last peak. However, I assume that it is too late to be water coming off the column, as two alcohols have already come off... Beyond that, I am coming up short a guessing. It may be worth it to inject methanol to see the shape of the trace of the blank run.

The mass range used for acquisition would give a clue as to what we might detect or what might not be detected, but would result in ion supression.

[Peter Apps]

As an extra symptom - the blue peak is way more tialed than any of the others. First guess on this is an inlet or injection problem. I presume at 1% that you are running split, but the last peak is front tailed which perhaps suggests split-splitless, and maybe the splitless time is a bit too short. With methanol as solvent the Agilent autoinjector defaults are far from optimal.

Peter

[tmammen]

sepscientologist,

Did you take care of the issue? I am also facing similer issue.

[sepscientologist]

As is so often the case... I didn't take care of it but I haven't seen it lately either.
Drop me a line if you like mfoster@micromidas.com

[ktr]

Hi everybody, did anyone solve the problem. I also experience similar one on Agilent 5975.
Sometimes from one analysis to another is a case of sensitivity drop (5-20%) for one or all analytes (not an injection cause-checked).
How is your presision in common GCMS analysis in a series of injections ?

Best Regards!
Chris