Response decrease over replicate injections

[Zoraku]

Column: DB-1701 30m x 250cm x 0.25u
Inlet Temp: 300ºC
FID Temp: 300ºC
Flow: Helium at a constant 35cm/s
Oven Gradient: 200ºC for 5min > 10ºC/min to 280ºC > 280ºC 2min
Injection: 2uL injection volume with a 100:1 split ratio

Last week I performed a method validation on a method with the above parameters. I probably had 50 injections in the sequence, and everything ran great!

Yesterday I ran the method again on a storage stability test of one of our formulations, using the now validated method and I am seeing a consistent decrease in response for two of my three analytes. Every injection from the same vial will produce a lower response (this is seen in both the standard vials as well as sample vials).

I have checked the inlet, replaced the septum, gold seal, inlet liner. I've watched my injections and there doesn't seem to be anything wrong with the syringe. No leaks or pressure loss anywhere.

Literally just running the same method a couple days later, with no other method or analysis being run on the GC between then and now, and now I'm seeing this phenomenon.

Right now replicate injections are coming from the same vial (as they did in the method validation procedure) but I will probably try to inject the same solution from different vials to see if that produces any different results.

Any other suggestions would be appreciated.

[HPLC chemist]

Are you sure the molecule of interest is stable or the standard was stored properly (example = refridgerated)?

[Zoraku]

Are you sure the molecule of interest is stable or the standard was stored properly (example = refridgerated)?

Yessir! Standard stored at 4ºC. Warmed up to room temperature for standard preparation.

As I had mentioned, there was no issues at all during the method validation process, and the standard was handled the exact same way...with even more injections in the sequence than the storage stability analysis.

During the method validation the drift in response was less than 2% from the start to end of the sequence...while I saw almost a 5% drift in response in the most recent analysis, with many less injections and samples between the start and end of the run.

I ran the study again with every injection from a different vial...there was still some variation but I didn't see this significant loss in response over time. Maybe I just needed to do a few injections to get the column/system equilibrated more...but I still find it odd, and think there is more of a root cause than the vial situation.

[James_Ball]

Between the initial validation and the stability study was the instrument held at the starting temperatures of the method or was it shut down or run in standby at lower temperatures?

Just wondering because if I leave my instruments at low temps over time they tend to build up higher boiling impurities that need to be baked off before they are stable again.

[Zoraku]

Between the initial validation and the stability study was the instrument held at the starting temperatures of the method or was it shut down or run in standby at lower temperatures?

Just wondering because if I leave my instruments at low temps over time they tend to build up higher boiling impurities that need to be baked off before they are stable again.

It very well could be this... I probably didn't give the instrument enough time to warm up and equilibrate at the initial temperature before running the sequence. I did in fact turn the GC off between the validation and the study. It was probably powered down and left at ambient temperature for a week.

I'll be doing another analysis of the stability study in two weeks - so I'll give the instrument much more time to sit at temperature and even run the gradient with diluent a few times to ensure everything is baked off and we'll see how my reproducibility looks.

Thanks for you input!