Purge&Trap GCMS internal standard increase, new observations

[danicrd]

Just a question..or two What's the meaning of "cleaning OI" ? Which parts are you cleaning or replacing? We have almost all brands of purge e trap and also OI in the lab...
Thank you

[01310040231]

Daniele:

We rinse the sample path with methanol; replace the water management, trap (using #10), sample filter and adapter to sample filter on OI 4660.

[mdevay]

This was a recurring issue in our 524.2 analysis. We calibrated 74 compounds (including IS and Surrogates) from .5ppb to 100ppb (we're lazy and hate dilutions for xylenes). The problem was far more pronounced on our more sensitive 5975c's and 5977s than our 5973s.

Cleaning the source and performing inlet maintenance improved performance for a short while, but within about 6 weeks, the higher standards would start to creep up to the point that the 10ppb and 20ppb CCV standards we ran would fail surrogate recovery. This was coupled with a gradual decrease in our IS/SS recoveries, and a corresponding increase in recovery (though a nearly stable response) in most of our lighter analytes, along with a stable recovery/lowered response in our ring based analytes, particularly the aromatics. Calibrations would go from extremely linear to grossly quadratic for our heavier analytes, and responses for these became heavily dependent on the total concentration of heavier compounds in the sample. Therefore, the recoveries and responses for anything aromatic in our samples could be wildly different from what they should be.

We ended up figuring that there was preferential adsorption happening throughout the system for our aromatics. Lowering the split ratio and increasing the amount of internal standards mitigated this problem significantly. This, however, leads to more water entering the system and affecting repeatability of our more polar compounds (eg ketones, thf). We ended up biting the bullet and dropping our calibration curve high point from 100 to 20, which improved linearity and slowed the problem down somewhat, but didn't entirely solve it. Using inert silcosteel unions for our transfer line also helped a bit, but the problem seemed to be rather systemic, and the only way to get rid of it entirely was doing a complete maintenance cycle every 6 weeks or so (source cleaning, liner, gold seal, column trim). Another interesting thing that was happening was that our gold seals would develop a large amount of black deposits.

Last I heard, they were considering experimenting with changing the inlet and transfer line temperatures, but I have since left state employment for better opportunities and couldn't tell you if it worked. If the inlet temp was too high and causing gradual degradation of the more fragile compounds after purging, sooty deposits could definitely cause adsorption at the inlet. Our sources never got so dirty that a semis-volatile analyst would even think twice about it, but the same kind of issue may have been happening on the back end of the column and the opening to the source itself, as the MS interface temperature is very high as well.

[Bigbear]

I know I sound like a broken record, but I feel the problem is water. I fought with it for close to 18 months after transitioning from a 5970 to 5973. I hammered on my Tekmar Velocity, bought a Stratum and the problem persisted, we even lost certification for 524.
We got the "water management kit" from Agilent which contains parts, method parameter changes, and a new BFB tune.
I feel the significant changes are:
6mm draw out plate
0.18 X 1um column
1mm split less liner
150:1 split

All I can say is that it works, I clean source every 6 months as I need to vent to do a rough pump oil change anyway.