GCMS(CI+) Instrument Drift

[jdlh199]

Does anybody out there use GCMS in CI+ mode? I'm developing a method for fatty acid ethyl esters. We're using a deuterated internal standard which means we're using CI+ with methane to detect the M+H ion.

However, our instrument drift is extreme! We are running large batches of ~100 samples, which with a 15 minute sample run time, means our analysis batches are usually ~2 days long. Both the analytes and internal standards drift, which means the internal standard seems to successfully correct for the drift.

It's definitely not a stability issue as when the batch is finished I can re-tune the gas/instrument and rerun the same samples. We start with similar count rates as in the original run, and again see the downward drift.

Should I be concerned about this drift even though the IS seems to be doing its job and correcting....?

Here's an example of the internal standard peak area count rates drifting over a single run:

21895
12474
12405
19100
10755
10406
10166
8411
9999
9653
9944
9291
9854
7698
8163
8486
8147
7078
7821
6020
6887
6788
6660
6490
6585
7692
7792
5952
5919
4936
5483
4850
4368
4345
3948
4360
3458
3335
3224
2934
2963
2674
2757
2805
2553
2414
2390
2337
2175
1938
1664
1373
1551
1512
1449
1387
1584
1483
1607
1718

[CE Instruments]

No details re instrument ? Can we assume that re-tune/recalibrate does not involve source clean so we can eliminate that ? Is system still using an analogue flow regulator to adjust the CI reagent flow ? Used to be a major source of the problems seen on old Trio 1 systems
Has it ever worked well ?

[jdlh199]

Sorry, it's a Perkin Elmer Turbomass Gold MS with an AutoSystemXL GC.

The retune and gas flow check does happen without a source clean so it's not that.

The CI gas flow is adjusted with a little black knob, which I presume is just a simple analogue flow regulator.

I'm not sure if it's ever worked well since I've never really used the instrument in CI+ mode too much before this assay. However I did write up an OQ/PQ for CI+ mode and that seemed to work fine, although obviously that was only a few samples with a short run time. I think it's the fact that I'm dealing with such long run times with so many samples.

It's not too much of a problem since the IS does it's job and corrects, I just wondered if this was "normal" or if there was something I could do to minimise the drift. Or does it just come with CI+ mode?

[CE Instruments]

The PE TurboMass is a sister system to the MD800, Trace MS from Thermo and used the same mass flow regulator as the named Trio 1. CI reagent flow could vary and cause the problems detailed but you can also get electronic faults that give the same type of errors. Try contacting the vendor agents or if in the UK one of the many MS specialist service companies (most of whom employ ex factory people who should know the likely solution)