10 Replies Latest reply on Apr 18, 2018 6:41 PM by koos

    GC/MS Loss of signal over time

    koos

      Good morning everyone,

       

      I am measuring some 'dirty' samples (flavour products) over a couple of days to measure impurities.

      Every day I am creating a new calibration curve and injecting a couple of samples.

      I am noticing that every day the signal of my calibration standard is lower than the day before.

      My question is what could be causing this loss of signal, and how can i fix it?

       

      Some information:

      7890B GC with a 5977B single quad MSD

      Max temperature: 240 (20C below column max)

       

      Result for highest standard, in areas:
      Day 1     100.000

      Day 2     85.000

      Day 3     70.000

       

      Thanks in advance,

      Koos

        • Re: GC/MS Loss of signal over time
          gchaplain

          1) Have you searched for an air leak through either a sniffer or through an air water check on the MS?

           

          If there is a leak somewhere then this would account for your loss of signal.

           

          We have experienced this previously with a leak on the inlet column ferrule, septum, and reducing nut.

           

          2) If you system is leak free, it could be your split vent trap filling, hindering pressure control in the inlet. It may need changing.

           

          3) When was your system last tuned? You maybe experience loss of signal becase of loss of detector sensitivity, especially if the areas are low to begin with (<10'000 counts)

           

          4) Failing this, it could be something as simple as you have a small blockage in your syringe

           

          I have a list of possible areas it could be:

           

          GC Inlet Ferrule
          Column
          Transferline ferrule
          Syringe
          HED (High Energy Dynode) horn
          HED (High Energy Dynode) power supply
          Septum
          Standards need replacing
          Source
          Split Vent Trap
          EPC
          Empty wash vials
          GC Inlet nut
          Mass Spec chamber dirty
          Gold Seal
          Liner
          Transferline nut
          Mass Spec O-ring
          Split Vent Line
          • Re: GC/MS Loss of signal over time
            james_jenkins

            Hi koos,

            Tags have been updated for better visibility in the community.

            Regards

            James

            • Re: GC/MS Loss of signal over time
              stevenwlloyd

              How dirty are your samples? We analyzed smokeless tobacco flavors in glycerin (for vaping) and the source would become completely contaminated (no tune possible) after one injection.You may need to clean your source. Check your tune results to be sure.

                • Re: GC/MS Loss of signal over time
                  koos

                  Thanks for the help gchaplain, james and steven,

                   

                   

                   

                  1) Have you searched for an air leak through either a sniffer or through an air water check on the MS?

                   

                  Performed self-evaluation on the MS. It appears to be water free and no problems where detected

                   

                  it could be ypour split vent trap filling, hindering pressure control in the inlet.

                  What does this mean exactly? I checked the inlet liner and while it had a couple of "grains' of dirt in it, the liner-cotton still appeared white and clean. Slightly coloured where the needle entered the cotton. how dirty does an inlet need to be before replacing? Or is this a simple manner of trial and error testing?

                   

                   

                  3) When was your system last tuned? You maybe experience loss of signal becase of loss of detector sensitivity, especially if the areas are low to begin with (<10'000 counts)

                  This fixed the problem. last tune was on 27mar18

                  Performed a run, tuned the MS, performed automatic tune self-evaluation (all OK and passed)

                  Area run pre-tune: 80.000

                  Area run post-tune: 150.000

                  I still dont understand why this fixed the problem. if the signal was lower due to detector sensitivity after +-2 weeks, would that mean I would have to re-preform the tune every week(?!) Wouldnt that just mean i am polluting my MS with every measurement?

                  On a small side note, is an area of 150.000 'relatively' normal for a 100 ppm sample in SIM? Or should i provide more information in order to answer this?

                   

                  How dirty are your samples? We analyzed smokeless tobacco flavors in glycerin (for vaping) and the source would become completely contaminated (no tune possible) after one injection.You may need to clean your source. Check your tune results to be sure.

                  We're on the same team here! Im analyzing raw flavours on a glycering base in a 10 times dilution in MeOH, used for vaping. I have analyzed +-50 samples so far with alot of blanks in between (old habit).

                  Is there a way to 'bake out' the MS for cleaning? Or are you talking about removing the vacuum, opening the MS, cleaning it with solvents and turning it back on? (Which I have never done before) What do I need to check on the tune results to be sure?

                   

                  THanks for the assistance everyone. Although the direct issue has been fixed, I would be really happy to discover the root cause of the loss of signal, and further assistance is much appreciated!

                   

                  Koos

                    • Re: GC/MS Loss of signal over time
                      gchaplain

                      "I still dont understand why this fixed the problem. if the signal was lower due to detector sensitivity after +-2 weeks, would that mean I would have to re-preform the tune every week(?!) Wouldnt that just mean i am polluting my MS with every measurement?

                      On a small side note, is an area of 150.000 'relatively' normal for a 100 ppm sample in SIM? Or should i provide more information in order to answer this?"

                       

                      Your answer is that over time detector drifts. In an ion source there are many parts that drift or misalign over time due to temperature and vibration fluctuations, so the instrument needs to be tuned, like tuning a guitar to the right note. The auto tune goes through an algorithm to optimise the ion peak shapes, intensity and mass location. Your detector could have drifted in one or more of these points. We tune our instruments weekly, followed by a calibration weekly. Every day we perform a tune evaluation to ensure the detector is doing what it is supposed to be doing. If it isn't we tune it and calibrate it.

                       

                      What abundance on the tune are you seeing for the 69 ion peak and what are you getting on the EMV parameter( should be on the report around 400'000 abundance, and between 1000-2000 volts)

                        • Re: GC/MS Loss of signal over time
                          koos

                          thanks for the explanation gchaplain.

                           

                          The 69 ion peak abundance during tune is 334.000 with an EM Volts of 965.8

                            • Re: GC/MS Loss of signal over time
                              gchaplain

                              Not bad. I would tune it two more times just to be sure.

                               

                              Also just a quick thought, I assume you are using an internal standard? If you are then at least fluctuations in your area response shouldn't matter t oo much for your results.

                                • Re: GC/MS Loss of signal over time
                                  koos

                                  Fortunately yes, so I am covered from that point of view

                                   

                                  The GC over here is a combo MS-FID machine which can be build to either MS of FID mode. It was in FID mode the last couple of days. Just rebuild the GC to MS mode, but I didnt perfom a full-on cleaning. My measurement series is nearly done, and i prefer to wait and not change anything major until it is complete.

                                   

                                  Performed a couple more tunes after the reset:

                                  All tunes are within 1 hour of each other

                                  Tune 1, tune 2, tune 3

                                  • Peak shape smooth & symmetric x3
                                  • Etune ion abundance m/z 69 508k, 512k, 513k
                                  • Etune ion abundance m/z 219 456k, 515k, 517k
                                  • Relative ion abundance m/z 502 7.5%, 7.4%, 7.4%
                                  • EM voltage 988, 1067, 1081
                                  • #background peak 167, 220, 203
                                  • Isotopic ratio m/z 502 10.4%, 11.0%, 9.7%
                        • Re: GC/MS Loss of signal over time
                          stevenwlloyd

                          Koos:

                           

                          150,000 counts seems very low in my experience.

                          About source cleaning: The manuals should give instructions on how to clean the source. We have a 5975 which hasn't needed cleaning and 2 5973s that I have cleaned many times over the last 20 years. I have no experience with the 5977. Briefly, you have to break vacuum, remove the source, disassemble it, clean dirty parts with green sand paper and/or aluminum oxide, sonicate in solvents, reassemble, reattach and pump down. I assume that all 5977 documentation is on line and there are probably source cleaning videos available if you search for them. The whole process usually takes 3-4 hours. I just watched an Agilent webinar (Simple Strategies for GC/MS Success ) that explains how to examine your autotune report to determine if it needs cleaning. If you can find the first autotune report printed at installation, you can get a good idea what it should look like. I think you can view the webinar through these links:

                          The View Recording button will allow you to sign-in and view the recording.
                          https://agilenteseminar.webex.com/agilenteseminar/lsr.php?RCID=110b45c8ad3ae6d80560abcbd24efd0d

                            If you wish to download a PDF copy of the presentation, please visit:
                          www.agilent.com/cs/library/eseminars/public/AFO_Simple_Strategies_for_GCMS_04112018.pdf

                           

                          This was sent after I attended the free webinar, but I think it is OK to share it. I recommend Agilent's webinars on basic GC, LC and MS procedures.

                           

                          I tried to bake out the glycerin contamination when it happened on our 5973. No luck. You are probably "polluting" the source. The autotune sets the electron multiplier voltage to achieve maximum sensitivity and as the source gets dirtier, that voltage will probably increase. If you keep your old reports, you can see it go up. Weekly autotunes are now considered a little too often but I started out on a 5970 that was tuned every day.