Endrin/Pesticide issues on Agilent 6890A

[MCAbraha]

Hi all,

I have been reading this forum the last few weeks after stumbling on it accidentally and I must say it is very informative. So, I figured it's about time I joined! I have been working on GCs for the past 4-5 years, Varians and Agilents, specializing in PCBs and Pesticides mostly various methods. I have some background in DRO, Herbicides, and ETU as well. So, on to my issue..

We recently started using an Agilent 6890A for 8081 analysis. On the Agilent, we are looking decent but we have a serious problem with Endrin Aldehyde. The curve is consistent, and passes well against 2nd source CCV on the alternate pest GC (a Varian) but on the Agilent Endrin Aldehyde recovers some 25% higher than other analytes immediately following curve.

One would suspect breakdown, but when running the breakdown mix, breakdown of Endrin is practically non existent (<1%). The recovery in the CCV doesn't seem to get better with time/injections which suggests it is not something baking off that is quenching/adding response. And this is regardless of whether we just clipped column/installed new inlet or ran 10 curves back to back to back. This GC has not yet seen samples as we are still in the initial QC stage.

My initial thought was that something in the liner/inlet is quenching the response of Endrin Aldehyde that then bakes off before the CCV runs (only 7 injections later) resulting in a "high" response. But this doesn't make sense as you would think that after that point, re-calibrating would be spot on due to whatever "gunk" present baking off and no longer affecting response. Unfortunately, this is not the case.

The only differences between the two GCs are the column types and that the Varian has silco-steel injectors while the Agilent (problem) does not. Any ideas?

Thank you for the help and suggestions.
MCA

[mckrause]

I would suspect the opposite. I would think that you were quenching active sites and by the time you get to your CCV you have passivated all of the active sites.

One way to find out would be to start a new analysis and just run replicate injections of the CCV to see if the response increases with time.

[MCAbraha]

We thought the same MCKrause but the issue here is that each subsequent calibration curve/CCV analysis sequence has the exact same results so this doesn't support the idea we are quenching the active sites but I understand the reasoning.

We tried a few curve/CCV analysis runs (5 points, instrument blank, PEM/CCV --> total 8 inj) and each curve is roughly the same. For some reason Endrin Aldehyde recovers 20+% high immediately after curve but all other 21 analytes are fine. Though it should be noted that Methoxychlor and Mirex (Mirex is a completely different source than the Mixed Pest and is added to solution) recover a bit higher than other analytes but only about 10% high immediately after curve which is compliant.

On our Varian 3800 pest GC with ZB-1/ZB-5 columns, we used to have issues with the Restek Siltek coated liners similar to what MCKrause initially thought was the problem. We would have a compliant curve and then by the time we got to the CCV, a few analytes would skyrocked 20-40% high. We then switched to SGE gooseneck split and they work fine.

My thoughts are maybe the SGE liner is to the Agilent GC as the Siltek liner was to the Varian, and possibly some active sites are being affected with the different columns in the Agilent. They are MR-1/MR-2. Any ideas?

[MCAbraha]

Also, Thank you MCKrause for replying. We aren't in too much of a bind for time as we have the Varian Pest GC up and running but we do need to get the Agilent up with different columns as a backup soon.

Also, it should be noted I re-made the pest curve completely new to troubleshoot, and also made a new Mixed Pest CCV from a 3rd source with same Endrin Aldehyde issue.

EDIT- Also, we tried the replicate CCV injections to see if response changed for this analyte and it did not. We would run the 5 point curve, and then the 2 different CCV sources 3 times each and the response for Endrin Aldehyde was always substantially greater than the other analytes. (Ex, 21 analytes <5% RSD while Endrin Aldehyde immediately 20+%).

Mixed Pest is made with all analytes but Mirex which is added seperately.