GC/MS Inlet Contamination

I have a 7890A/5975C with MMI. I have been seeing a huge ghost peak towards the end of my 1st run of the day. It gets smaller in the runs after but never goes away. I have performed basic inlet maintenance, cleaned it, baked it out with the split on 200ml/min. I can't get it to go away. My guess is that its coming from a part of the inlet I cannot see but not sure where. It's being identified as TXIB which is a plasticizer. I have avoided using any plastics near this instrument because of potential phthalate contamination so I cannot think of a potential source. I have made sure that the regulator has a stainless steel diaphragm, changed the copper lines, and even the split vent cartridge. Has anyone run into an issue like this before? Am I missing something? Any insight or thoughts would help. 

Thanks in advance. 

  • Laurenairchecklab - how huge in counts?  what is the spectrum of the ghost peak? Can you share a screen capture of the TIC and the spectrum of the peak?  Is it always the first run of the day?  What temperature do you leave the oven at overnight?  How long has it been since the gas traps were replaced?

    There was an old trick to see if the gas is contaminated. You leave the oven as low as it can go, typically near ambient, for a long-ish while, like an hour or more, then observe the signal as you heat it up to the method final oven temperature.  Then you cool the oven down to the same initial oven temperature and then heat it up again immediately and watch the signal. If the peak is bigger after waiting longer with the oven cold, then the peak may be getting trapped out of the gas on the front of the column. 

    Plasticizer. Do you use plastic squirt bottles for your wash solvents?  Plastic pipettes?  Plastic beakers anywhere? Plastic sample low-volume inserts?  What about the sample vial cap septa - are those silicone or?

    Some google:

    CAS   6846-50-0
    2,2,4-Trimethyl-1,3-pentanediol diisobutyrate
    2,2,4-trimethylpentane-1,3-diyl bis(2-methylpropanoate)
    Kodaflex txib

    Some NIST:

    The problem is that 43 and 71 are also part of nearly every hydrocarbon spectrum.

    I look forward to your reply.

  • I would say over well over 100 million. hard to to estimate with the integrator in Openlab CDS also its not necessarily a peak in the first run. Can you possibly explain why it looks like this at the beginning of the day? The oven sits at 40 degrees when its not running. Currently do not have any gas traps (not my choice) but have ultra high purity helium. 

    I agree with that second part considering that this instrument is for research and is not used frequently so it does sit for days sometimes weeks at a time without running. 

    We currently have a dispenser for the methylene chloride that is made from PP, PFA, and FEP. I am in the process of trying to get rid of this pump and switch to only using glass. However I have made extensive efforts to not use any DCM that comes into contact with the dispenser on this instrument. I have a primary instrument with a similar set up the main difference being that it has a S/SL inlet instead of the MMI. The primary instrument has a different contaminant that mimics the behavior of the TXIB but it appears to be a chlorinated fluorocarbon. I cannot confirm the identity of that contaminant because of the fluorine. All I can say is that the contaminants are different. The primary instrument only uses DCM that comes into contact with the pump. The contaminant also does not linger on the primary instrument (it goes away after the first run).

    The vial caps are PFTE/Silicon/PFTE. 

    The whole blob is giving the same ID. 

    The peak goes to about 50000 counts in the runs afterwards but the spectra still gives a pretty good match. 

  • Ultra High Purity Helium is good, but UHP Helium with good traps that get changed frequently enough on the line before the GC is better.

    Change your standby oven temperature to something over 100°C.  I like 111°C because that number is strange and can be seen on the 7890 display from across the room, but hotter is fine, too. Save it as a method called something like Hot_Standby.M or 111_Oven_Standby.M or ...     Does having it sit at 111°C all the time between runs change this behavior?

  • So I changed the temperature to about 110. When it sits for at least a few hours the peak goes away. Then it comes back in the next run but small just like before. I repeated this a few times just to make sure it was consistent. 

Was this helpful?