Octamethylcyclotetrasiloxane

[i.horsting]

I think that your reproducibility should be raised with removing the
glass wool piece.

You don't require glass wool(most of the time, pretty active, because of breaking) with this amount of injected sample.
The space is big enough for your solvent to expand.

There should be no siloxane from the septum, because of the septum-purge.

Remove the Split 1:10. This will reduce your analytes with a 10 fold.
better to do this splitless. and after the sample transfer open the split to vent off some of the remaining stuff in the injector. (this will reduce your background a bit).

[Peter Apps]

"There should be no siloxane from the septum, because of the septum-purge"

When working at trace levels I see siloxanes frequently unless I take the most stringent precautions to reduce them, which includes eliminating silicone cap seals, and increasing septum purge flows from the recommended 2-4 ml/min to 20 ml/min. I have a suspicion that the outside of the syringe needle picks up siloxanes from the septum, which then get desorbed in the hot inlet to give nice sharp contaminant peaks, so I made a gizmo that opens a path for the needle to go through the septum and then lets the septum seal again. It seems to help.

Peter

[Bruce Hamilton]

As noted by i.horsting, remove any insert packing.

I would review some of the earlier literature, and I would probably consider using a hydrocarbon solvent if others have used it ( HCs often greatly reduce unwanted junk - perhaps making splitless more of an option ).

I would seriously review your sample preparation, perhaps concentrating any extraction/extract so you're not as such trace levels.

I agree with Peter, septum purge will not remove trace material stuck to the needle, or fragments that fall into the injector. You need to eliminate the silicone in the septa - as noted above, I'd just use a different material. Your septa requirements are not to onerous.

I've always assumed that the reason I saw one nice baseline after conditioning and new injector insert and the column was because the withdrawing hot needle fried some silicone onto the surface that subsequently volatilised on the next injection.

I once carefully wiped a tapered needle outside with warm 60:40 CHCl3/MeOH after each injection, and the spurious peaks didn't appear for quite a few runs. I assumed the needle finish wasn't smooth enough to stop a fine film appearing.

I used to be a great fan of the SGE triple layered ( TCS ) septa, because that problem wasn't such an issue - compared to the instrument vendor-supplied single layer "high temperature" septa.

More recently, some of the high temperature, preconditioned and washed septa had lower backgrounds, but I'd never trust them for trace silicone analysis.

Murphy's Law tells me that a fragment of septa would somehow end up in the injector.

Once again, search out the literature, the type of analysis you are conducting will have been undertaken before.

Bruce Hamilton

[gmattz]

How did it turn out? Were you able to validate a method with D4? I am trying to validate a method now with D4, D5, and D6 and am running into a lot of problems, however I am limited to a fairly difficult drug product. Did you have any luck switching to no septum? A low bleed septum did nothing for me. As mentioned before I have tried everyother column to no resolve, selectivity seems to be best with 5MS. Carryover, carryover.................. As for literature, I have found not one method that has truly been validated or remotely passes ICH guidlines. It does me no good to find methods for detection since cyclosiloxane are the easiest species to detect but seemingly the hardest to quantitate. Recovery and precision seem to be the most difficult. Let me know if you found anything that works.