Chromatography Forum

Happy New Year!

I have noticed for some time that we have been getting some evidence of analyte peaks in blank samples. I have done all the usual troubleshooting things and have come to the conclusion that an errant setting during some large volume injection optimisation a while back has most likely led to backflash causing our carryover. Since then I have of course changed all relevant consumables several times over and have cleaned all parts of the inlet that I can access suggesting that the contamination is most likely sitting in one of the lines.

During a recent visit by an engineer I discussed this and possible solutions and his viewpoint was that although it is possible to change out or clean the split vent line, because the system (Agilent 8890) has a PAL3 unit sitting on top, it's too difficult and time consuming to do and suggested 'injecting 10ul of hexane' with the 'detector [MSD] switched off'. At the time, I nodded sagely and we moved on to other issues that were more pressing, but realise in hindsight that I have no idea what he was actually suggesting I do...

Any suggestions as to what his intention was?

Kind Regards

TD2

Pretty sure he's suggesting that you run a very large solvent blank while making sure the MSD is not physically on and acquiring data.
In fact, it might be a good idea to run multiple solvent blanks because the contamination could possibly be carry over still on the column itself and not in the lines. Multiple solvent blanks can have the effect of "rinsing" the injector and the column and diluting away contaminations.
If you have the time, it wouldn't be a bad idea to replace the column with either a known good used one or a brand new column and run a known clean blank.
If the carryover disappears then the contamination is on the old column and not in the lines.
Your contamination problem could also be coming from contaminations in the equipment/solvent/standards that are used during sample preparation and that is a much more difficult and time consuming problem to deal with.

TomC