RT shift and response drop after universal trap replacement

[Edwardchen]

I recently replaced a universal Helium trap connected to Agilent 6890N GC and 5973Network MSD. There was significant RT shift and/or undetectable signal in 5 out of 18 compounds in Biorad Lyphochek Urine Toxicology QC. Re-injection of previously acceptable QC also showed same problem. Can this problem be caused by a universal trap replacement? Alternatively, there has been tailing in some chromatographic peaks before this problem, probably because of a deteriorated column. Can a deteriorating column cause a sudden unacceptable QC?
Could any expert please advise? I appreciate your great help.

[Peter Apps]

The answers to both of your questions is yes, but a deteriorating column is more likely to give progressively worse results rather than a sudden change. If all the peaks are affected the retention time change must be related to a change in gas flow through the column or to a change in oven temperature, so you need to check the inlet pressure and the oven. I would have expected either problem to have caused a persistant not ready fault onthe GC though. Some purifiers are filled with nitrogen that takes several hours (at least overnight) to purge out, try the QC again today.

Peter

[Ron]

If you changes the trap you should have cooled down the GC and vented the mass spectrometer. If you didn't, you may have damages the column phase by the air introduced. If you did, check all fittings, the graphite vespel ferrules may have shrunk, causing a leak.

[Edwardchen]

[quote="Peter Apps"]The answers to both of your questions is yes, but a deteriorating column is more likely to give progressively worse results rather than a sudden change. If all the peaks are affected the retention time change must be related to a change in gas flow through the column or to a change in oven temperature, so you need to check the inlet pressure and the oven. I would have expected either problem to have caused a persistant not ready fault onthe GC though. Some purifiers are filled with nitrogen that takes several hours (at least overnight) to purge out, try the QC again today.

In fact, my QC run was 3 days after the trap replacement. Hence, any nitrogen gas filling the trap should have gone.
The serviceman replaced the column and I performed RT locking. Although the RT is acceptable now, the amphetamine peak is much lower than before. I see an additional phenobarbital (no TMS) peak in addition to the phenobarbital TMS peak that I used to see. In addition, in one SIM experiment in which I added 0.5ppm IS, the IS is no longer detectable now- peak not found at all. Do low amphetamine and disappeared IS reflect a drop in sensitivity of the instrument ?
Please comment. Thanks.

[Edwardchen]

If you changes the trap you should have cooled down the GC and vented the mass spectrometer. If you didn't, you may have damages the column phase by the air introduced. If you did, check all fittings, the graphite vespel ferrules may have shrunk, causing a leak.

Can a leak cause an RT shift in some of the compounds and not in others? Also, can a leak cause a drop in response? In tuning report, there is acceptably low percentage of 18 and 28 peaks. Can this exclude a leak?
I am new to the GC-MS field. Please enlighten me. Thanks.

[Peter Apps]

If replacing the column fixed part of the problem, the old column was very likely damaged. While the column was dying it probably bled a lot of high MW contaminants into the MS, and the source now needs cleaning. That shoud fix the loss of detectability.

Peter

[Edwardchen]

If replacing the column fixed part of the problem, the old column was very likely damaged. While the column was dying it probably bled a lot of high MW contaminants into the MS, and the source now needs cleaning. That shoud fix the loss of detectability.

Peter

Thanks for your comment. If the ion source is dirty, the EM voltage and repeller voltage should be high?! But they are not high in my instrument. Since I see non-TMS compounds, is there a possibility that there is moisture somewhere in the GC column path? I checked the septum (actually replaced it) and made sure the injector nut was tightened. Then run the QC. But the low response problem and non-TMS compound were still present. Do I need to wait for the column to equilibrate so that any remaining moisture inside the column is gone? In addition, is there a need for the new column to run in for better sensitivity?
Thank you very much.

[Peter Apps]

Unless you have very dirty and wet carrier gas, which will destroy the column quickly in any case, it is not likely that the underivatised compounds that you see are TMS derivatives that have subsequently hydrolysed. Far more likely they were never derivatised in the first place, due to a problem with the sample preparation. The reason that you see the non-TMS compounds with the new column is that the old one was absorbing them completely.

I would be very surprised if the sensitivity got any better as the column ran in - columns rather tend to run out.

If the ion source is dirty high EM and repeller voltages can bring the sensitivy back again, but they would need to be adjusted accordingly - the MS does not tune itself based on the performance with QC samples.

What is your inlet maintenance schedule in terms of changing septa and liners ?

Peter

[Edwardchen]

Unless you have very dirty and wet carrier gas, which will destroy the column quickly in any case, it is not likely that the underivatised compounds that you see are TMS derivatives that have subsequently hydrolysed. Far more likely they were never derivatised in the first place, due to a problem with the sample preparation. The reason that you see the non-TMS compounds with the new column is that the old one was absorbing them completely.

I would be very surprised if the sensitivity got any better as the column ran in - columns rather tend to run out.

If the ion source is dirty high EM and repeller voltages can bring the sensitivy back again, but they would need to be adjusted accordingly - the MS does not tune itself based on the performance with QC samples.

What is your inlet maintenance schedule in terms of changing septa and liners ?

Peter

I just replaced the septum, gold seal and liner. I also made sure that the injector nut is tightened.
In fact, the service engineer just made a divert connection from the He gas cylinder with a re-used copper connection pipe. Since it is re-used, can there be any micro-leakage of air from the connection into the trap?
Thanks for your comment.

[Peter Apps]

It sounds as if you need to leak check the whole system, including the inlet and transfer line connections and the septum. Use an electronic leak detector, not a liquid.

Peter

[aldehyde]

To leak check a 6890 cap the purge valve on the back and the inlet. Set the inlet pressure to 25 PSI and allow it to equilibrate, set the pressure to off (not zero) and check for a drop of more than 1 PSI over 5 minutes (the usual metric is a bit more forgiving but you shouldn't have a leak--its either going to be rock solid or drop quickly).

If you have a flow meter you can check to see you are getting the flows you expect at the end of the column. If you have an FID, TCD or something else on the instrument in addition to the MSD you can plug it in and check from there.

Replace the ferrule at the inlet and the MSD interface just to be sure.

Even if your EM voltage and repeller aren't too high.. when was the last time you cleaned your source? Do you have an ion gauge?