Calibration drift issue with a 5975 MSD

I have a 7980/5975 GC/MS that suffers from a significant calibration drift of over 1 amu in a few days.  It affects higher masses more but low mas (O2 at m/z 32) will also drift down to the point it registers as m/z 31. Everything looks good when I perform atune tuning initially, but the draft is affecting results within hours.  Loading different historical atune files and retuning leads to same result (drift in hours to days).

The EM was replaced last year and the instrument met specs during a recent PM, but the drift issue remains,  Last year I reseated the column, and it seemed to help for a while but I never understood why (DB-FFAP 50m x 0.32mm x 0.5 um)

The system is in otherwise good shape but makes it impossible to use MS spectral library matching!  Would also be useful if I could correct calibration in .d files after acquisition as well if anyone knows how to do that!

Thanks!

  • How many runs on the system?    7890: Service Mode , Diagnostics, Instrument Status.   What type of samples do you analyze?  What sample prep?  What concentrations?  How big are your peaks?  Please share a couple of tune reports spread out over a day or two and a chromatogram of your highest level standard if you can.

    The first simple guesses are a quad contaminated with sample residue or an electronics problem.

  • Hi Paul,

    Overall the system is lightly used for analysis of headspace samples using SPME technique (fermentation broth, fabric swatches, resins).  Overall sample load is low and liners never appear contaminated. Issue started a few years ago and is now a big issue with longer sample queues.  I will share some tune reports when I am in lab later today.

    Thanks,

    Joe

  • Everything that comes out of the column goes into the MS. A system can be contaminated very quickly - less than 50 samples, even - but is cumulative so over time and runs it continues.  Looking forward to tune reports.

  • Hi Paul - so it indeed appears to have been a column positioning (or age) issue.  I swapped the older DB-FFAP column for a new Rxi-1ms column and the calibration now appears stable (at least for past 48hrs).  I noticed this a year a go when I clipped the same FFAP column and reinstalled in the MSD - calibrations seemed more stable, but then began to slip again after several months.  Is it possible that column age or positioning can have this effect?  Still seems strange to me but now I have seen this effect twice..  Will check calibration stability over the next week to see if this is really it..

  • Mass Drift would be related to the mass filter, the quadrupole. The vacuum could cause that, too much column flow, an incorrect column configuration so the flow is too high, the wrong gas configured, or the pumping speed is not good.  How old is the rough pump?  If the turbo is spinning 100% it's working fine, but if the rough pump is not pumping out the turbo exhaust there will be issues.

    If the column is installed too far into the orthogonal ion source (SS, Inert, Extractor, HydroInert) the peaks go away and if installed way too far, the end of the column will appear burned by the electron beam.  If the column is installed too short there's every chance that no problem will be seen at all. I did a test where I pulled the column 50mm back into the transferline and only the most reactive compound in my 4 compound mix was affected.  The peaks come out of the column in the hundreds to thousands of miles per hour so a few millimeters too short is not a worry.

    DB-FFAP 50m x 0.32mm x 0.5 um ... At least it is 50 m so the inlet pressure will always be >5 psi, the necessary minimum for decent inlet/column pressure/flow control.  The Rxi-1ms column you just installed - what is that column's flow rate setpoint?

    On the historic tune reports, look at Mass Gain and Mass Offset.  Are those numbers changing appreciably over time/tune-to-tune ?   Those set the Mass Calibration.

    Amu Gain and Offset and WIdth219 adjust the tune ion peak widths.

    Does the temperature in your laboratory drift many degrees over time?  Most labs don't change temperatures all that much over the course of a day or a week. 

  • The odd thing is that I had been using the 50m x 0.32mm x 0.5 um DB-FFAP column for some time with no issues - so I wonder if degradation of the end of the column nearest the source was the root cause.  System seems stable with the new column.  Attached are historical tune readings.  I used different atune file names from 2011 to 2022.  Switched to Fil 2 om Jun 21 which made little difference until I switched column.  Thanks again for your insights!

Was this helpful?