Page 1 of 1

"Checkered" Pattern in Signals

Posted: Tue Sep 15, 2020 7:58 am
by dashert123
Hello all, I am recently having problems with my MS signals in our Headspace GC-MS setup. My peaks are well seperated, but the peaks themselves have either small sharp "shark-tooth" peaks on top, while some others have "checkered" patterns, like the wall of a medieval castle (my best descricption of them).

I have consulted several Trouble-shooting manuals but have not been able to find the exact description of my problem. The problem is with the MS part, as the FID-detector (we use a splitpoint) have nice, round signals.

We have tried the following:
- Purging the Column
- Cleaning the Ion-source of the MS
- "Purging" the quadropole (heating it to the limit for a couple of days).

Hope the question is understandable, and any input is welcomed.

Edit: Picture of the spectrum Image

Cheers, Carl

Re: "Checkered" Pattern in Signals

Posted: Tue Sep 15, 2020 9:27 am
by sbashkyrtsev
Could you attach pictures of chromatograms and spectra to your question?

If signal seems overly discrete - maybe the problem is the sampling rate of the detector too low? Meaning it has too few points in the chromatogram.

UPDATE: Since I can't edit dashert123's post, linking the image he tried to attach:
Image

Re: "Checkered" Pattern in Signals

Posted: Tue Sep 15, 2020 1:21 pm
by dashert123
sbashkyrtsev wrote:
Could you attach pictures of chromatograms and spectra to your question?


Thank you for the suggestion, I believe it is linked now.

Re: "Checkered" Pattern in Signals

Posted: Tue Sep 15, 2020 3:22 pm
by dashert123
sbashkyrtsev wrote:

If signal seems overly discrete - maybe the problem is the sampling rate of the detector too low? Meaning it has too few points in the chromatogram.



I tried turning the scanning rate up by lowering the mass range for detection - but it seems that the pattern only because more concentrated with even more "walls".

Forgot to say in my original post: The peaks shown when tuning the MS looks good, at most of the other peaks in the spectra does not show the same pattern, so I am deeply puzzled.

Re: "Checkered" Pattern in Signals

Posted: Tue Sep 15, 2020 8:13 pm
by LALman
Do the masses under the peaks have the same shape or are they flat-topped? Try running the same sample at a large dilution and inspect what comes out at these elution times.

Re: "Checkered" Pattern in Signals

Posted: Wed Sep 16, 2020 8:43 am
by dashert123
I am uncertain what you mean. The MC-yields the correct masses under the peaks (in this case 44.01 for acetonitrile-d3). The peaks from the FID are nice and round. I tried diluting up to 4x times through the split/ratio (going from 10:1 to 40:1) - the peaks looks much nicer and does not show the same pattern, but are still "spiked"

Image

Re: "Checkered" Pattern in Signals

Posted: Wed Sep 16, 2020 6:50 pm
by James_Ball
What is the mass range you are scanning during acquisition?

If scanning with no cal gas present in manual tune do you see mass 44 with any abundance? If so it could be a small air leak contributing an unsteady amount of CO2(mass 44) to your chromatogram causing the peak to be jagged.

Re: "Checkered" Pattern in Signals

Posted: Wed Sep 16, 2020 10:18 pm
by benhutcherson
To me, it's REALLY worth taking a close look at the complete mass spectrum at each of those "spikes" on the peak.

They look to me like what I've seen with co-eluting compounds. It's entirely possible that you don't see them on the FID because the co-eluted compounds have the same response factory in that detector, but have different ionization efficiencies or whatever that makes them show up distinctly in the MS.

This is one of the beauties of mass specs-you can see if things like this are happening.

Re: "Checkered" Pattern in Signals

Posted: Thu Sep 17, 2020 10:06 am
by dashert123
benhutcherson wrote:
To me, it's REALLY worth taking a close look at the complete mass spectrum at each of those "spikes" on the peak.


I believe I found the solution - indeed the mass spectrum at the top of the peak and the "bottom" of the peaks were different.

The spectrum at the top of the peak shows two peaks at 42 and 44 (the d3-acetonitrile at 44 and most likely d1-acetonitrile at 42).

The spectrum at the bottom shows only 44.

Our scan range were 42-150 au, meaning not all of the 42-contributions were detected all the time. I can imagine small electronic differences could cause this. Changing the scanning range to 41.50 to 150 made the chromatogram smooth again.

Thank you all for your suggestions and help, at was most appreciated!

Cheers, Carl

Re: "Checkered" Pattern in Signals

Posted: Thu Sep 17, 2020 8:38 pm
by James_Ball
dashert123 wrote:
benhutcherson wrote:
To me, it's REALLY worth taking a close look at the complete mass spectrum at each of those "spikes" on the peak.


I believe I found the solution - indeed the mass spectrum at the top of the peak and the "bottom" of the peaks were different.

The spectrum at the top of the peak shows two peaks at 42 and 44 (the d3-acetonitrile at 44 and most likely d1-acetonitrile at 42).

The spectrum at the bottom shows only 44.

Our scan range were 42-150 au, meaning not all of the 42-contributions were detected all the time. I can imagine small electronic differences could cause this. Changing the scanning range to 41.50 to 150 made the chromatogram smooth again.

Thank you all for your suggestions and help, at was most appreciated!

Cheers, Carl


You will get that noise if your scan range is very close to the target you are looking for. I would probably set the scan range even a little lower just to make sure nothing interferes. The mass axis can vary a couple tenths of an amu as it scans so that was what was probably causing the problem.