Downwards drifting baseline

[sianq]

Hi,

We're running ananlysis of Dioxane for In-process material: packed column (Chromosorb 101, 60-80 mesh, 3.2mm x 1m), oven/injector/FID detector 160C), liquid autosampler, DMF as solvent, Helium carrier gas.

We recently observed a downwards drifting baseline for 3 out of 32 injections midway through analysis. I know the most likely cause is oven temperature drift but how can I be sure that this is the cause? We can't monitor the oven temp through the run.
How much of a temperature drift would need to occur for the baseline to drift downwards?

Are there any other possible causes?

Thanks in advance,
Sian

[chromatographer1]

Can you post a chromatogram? A picture is worth a thousand words.

More information would be helpful. How much drift? Did it drift back up after going down, or did it stay down?

Your GC doesn't have a temperature readout? You can read the oven temperature or it is not convenient for you to do so?

Baseline drifts can be caused by leaks and by a drop in the amount of bleed from a column. You could also be losing your FID? detector due to contamination, lowering its sensitivity and signal.

best wishes,

Rod

[sianq]

Hi Rod,

I will try to get a chromatogram for you of good vs bad.

We had a total of 27 'good injections' with expected baselines, three then drifted down and stayed drifting down, obscuring our dioxane peak. The final 2 injections were fine again!

The software we're using to run the instrument doesn't monitor temperature in the oven, the analysis was being run overnight so nobody was here to monitor or check (it isn't a practice here anyway).

Thanks,
Sian

[GOM]

Hi Sian,

One possible explanation is that it is not your oven temperature, but the result of injecting (and therefore zeroing) on a carryover peak. Is this analysis 1,4-dioxane in SLES?

Regards,

Ralph

[chromatographer1]

Your instrument probably autozeroes the detector output at the initiation of a run.

If something was 'coming out' at that time the baseline would then go below the integrator zero and you would not see the true baseline and peaks could be missed. I suspect that is all that is 'wrong'.

Just rerun the samples. Hey, life happens. Want are you going to do?

best wishes,

Rod

[Peter Apps]

Hi Sian

This might be connected to gas supply to the FID. Do you by any chance have automatic switch over on your gas supplies ?

Peter

[sianq]

Good =


Drift =


We have a zero set in the method prior to every injection. The run time is one hour long to clean off the effects of the DMF solvent.

I have just noticed a carryover peak that was being hidden at the very end of the injection.....possible cause. Investigation to begin.

Thanks everyone for your help! [/img]

[BradVM]

Sian, your pictures are not showing up.