[Ajax]

Hi there

Out of frustration, and a determination not to call in a service engineer, I'm hoping someone here can help...

I test vacuum greases by high-temp sim-dis. Just recently the analyses have been showing a few % boiling off way below the expected range; we're not expecting anything below ~490°C, but I'm seeing a slow linear increase from about 0.1% @ 260°C to 5-10% by 490°C, then the expected broad 'peak' 

I know that these lighter fractions are not really there for 2 reasons; the manufacture conditions have not changed and I have retested an old sample which had previously passed with 0.0% below the 490 mark suddenly showing several % in this range.

My first instinct was that the new guy in the lab contaminated the hexane... but since a blank is done for every run, any contamination would be in everything made subsequently, and thus subtracted from the results.

Then did an automatic leak check on the GC which showed a possible leak.... so did a manual leak check and found none. Then ran the auto leak check again once I'd reconnected the column and it showed no leak...

Did a K-factor evaluation (don't even know what this really measures?) on the column which showed as well within the expected range...

Found out about reconditioning the column - so ramped it to it's max (445°) and left it for 24 hours!

Still got no where

Now I've not used a GC for anything other than Simdis, and I've not been doing it for that long so I don't really know if what I'm seeing in terms of signal from the FID detector is what i should be; rather than try and describe it I'll post a chromatogram with several runs overlayed, if there is anyone who thinks they may be able to help?

Please, it's doing my nut in. 
 
     

[Mr Peanut]

When you initiate a run without making an injection, do you see the unwanted profile?

In other words; will simply running the temperature program alone produce the effect?