CE GC6000vega2 FID problem - peaks too small or no peaks

[rustyGCer]

Hello,

I have a problem with one of our laboratory's old CE GC's.
Without any warning chromatograms from "Carlo Erba (then Fisons/then Thermo) GC 6000 vega 2 series" went "blank".
Chromatograms from routine analysis and from blank/standard samples happen to have very small peaks or no peaks whatsoever.
I have checked column's for leakage (DB-624 and DB-XLB), septa, injector, methanol-whashed detector and still have no clue, why did those peaks disapear. There are small peaks from solvent and toluen standard, from time to time, but their area is 100x smaller than it should be.

What could trigger this behavior?

- A leakage in oxygen/air line (we had that occasionally)?

- Hardware problems with detector or electrometer? (Fissons/CE EL 980)

- Bad split/splittless ratio setup? (there was no change in settings - may only be some leak in a support line)

- Any other source of the problem? (except the usuall PEBKAC-like answer )

How to solve this? Our service vendor doesn't really want to bother with our "much too old" GC.

Thanks in advance,

Dan

PS,

System & settings are:


CE GC 6000 vega 2 series + FID detector + Electrometer Fisons El 980 + integrator CE Instruments DP 700

mode: split/splitless (40sec after injection);
solvent - CS2, standard - toluen; usual analytes - volatile compounds

inj.temp: 150'C
det. temp: 250'C
carrier: He, 30kPa
air pressure: 100kPa
H2 pressure: 50kPa

column: DB-624, ID 0.53mm, length:30m, c.flow=20mL/min (back)

oven program:
40'C - 1 min plateau
1st rise: up to 100'C with 5'C/min - 1min plateu
2nd rise: up to 230'C with 10'C/min - 1min plateau

[CE Instruments]

Peaks gone really small means leak to me. Have you leak checked the system ?
Remove FID, fit cap to detector base body. Pressurise system 150KPa then turn off carrier and fuel gases, Should stay at full pressure for at least 3 minutes.

Most likely sources of leak, around the split valve base, septa, bottom nut of injector.

Do you have another Electrometer to swap ? Unusual problem if it is a faulty EL980 but it can happen.

You have checked the integrator program to check that the atenuation AT has not been re-set too high ?

Sorry we would love to send you an engineer but the travel cost would be prohibitive

We can supply all parts for the Vega and upgrades to 8000 series if applicable.

Regards Richard CE Instruments Ltd

[AICMM]

RustyGCer,

Are you using an autosampler? If yes, try manual injection. I don't know that CE instrument but is you split valve active to make splitless? If so, could be that the split valve is not actuating and therefore all your injections are now split instead of splitless/split. CS2 is a very poor responder on FID so try an injection of a more responsive solvent like hexane to see what kind of peak you get. If you shoot 1 uL of hexane and get a good response (and nice peak shape) you can probably rule out the electrometer and data system. I agree with Richard that the electrometer is probably least likely the problem. If you flame lights well and stays lit, I would not expect the problem to be detector flows but hexane injection should help rule that out as well. If retention times are the same but response is low.... again splitter valve. If retention times are later, like Richard says, look for inlet leak.

Best regards.