Varian 3400 GC - problems with peak heights - split flow?

[Seabiscuit]

I'm having a few problems with our Varian 3400...It was used without any problems for FAME separation a couple of weeks ago and has been on standby until last week. I ran a standard, and peak heights were very small and eluting incorrectly. Checked column nuts (loose column nuts have caused this problem before), still the same problem. Chopped off 10 cm of the head of the column (100 m column), and then elution order corrected itself but peaks still far too small.

Checked all gas flow rates, and the split flow rate seems variable....I adjust it to what it should be for the method in use (50:1 ratio, carrier gas flow 2 ml/min so split flow = 100 ml/min), and when I check it a few hours later it has changed (tends to speed up) - this shouldn't happen should it?

I've tried replacing the needle in the autosampler, cleaning out the autosampler syringe, checking the wash bottles, injecting manually, and still the peaks are far too small. I've run different standards and samples that I already have chromatograms for, but all are the same.

Do you think it is related to the split flow variability? Or should I be checking somewhere else?

Many thanks

[Peter Apps]

Do inlet maintenance - new liner and new septum. There is probably a leak through the septum. A leak seeker is a requirement in any GC lab.

Peter

[Seabiscuit]

Thanks for the reply.. I have already installed a new liner and septum, and still have the same problem..

[cjm]

Sounds like you've ruled out injection problems. Having done routine inlet maintenance, and ruling out leaks, I would check the split line for blockage before considering a possible problem control valve... Does the measured column flow fluctuate? If not it could still be an inlet leak...

[larkl]

When you checked the flows at different times and found they had changed, was the oven at the same temperature (not in the middle of a run)? You're running in constant pressure mode with these old pneumatic instruments and you'll see the flow drop and split change as the column temperature increases (if you're temp programming).

[Peter Apps]

By how much have the peaks got smaller ? 1/2, ten times ? (please give a reasonably exact number) and is this the same ratio for all the samples and standards that you have injected, and is the same ratio for all the peaks on one chromatogram ?

By how much does the split flow change ?

Have the peak shapes changed ?

When you say that you checked all the gas flows, does that include the detector gasses ?

Peter

[Seabiscuit]

Thanks for replies...

"When you checked the flows at different times and found they had changed, was the oven at the same temperature (not in the middle of a run)? You're running in constant pressure mode with these old pneumatic instruments and you'll see the flow drop and split change as the column temperature increases (if you're temp programming)."
This is something I wanted to check - I was checking the split flow at different oven temps...at standby temp (50degC) and starting temp (70degC). I (perhaps wrongly) assumed that column temp shouldn't affect split flow, but this would explain my variations.

"By how much have the peaks got smaller ? 1/2, ten times ? (please give a reasonably exact number) and is this the same ratio for all the samples and standards that you have injected, and is the same ratio for all the peaks on one chromatogram ?

I thought it was quite constant to begin with...peaks that were around 50 mV high on a chromatogram went as low as 12 mV (so a reduction of 3/4). But the last run they were much smaller than that and appeared to get even more smaller as the run continued. So no, it doesn't seem the same for all samples/standards, or all the peaks in one chromatogram.

By how much does the split flow change ?
It is set at 50:1, so with a column flow of 2 ml/min the split vents out 100 ml/min. I set it to this when in standby mode. Yesterday when a method was activated (so column temp increased to 70degC) I checked split vent again and it was about double the rate (200 ml/min!!!). I reduced it to what it was before. Then after the run had finished I checked it again and it was too slow. This could be related to the column oven temps I suppose.


Have the peak shapes changed ?
No

When you say that you checked all the gas flows, does that include the detector gasses ?
Yes - these are unchanged.
Thanks. I changed the column yesterday just to rule that out, and used new ferrules to connect etc. Ran standard again - exactly the same small peaks.

[Peter Apps]

If retention times are unchanged, and peak shapes are OK the gas flow through the column must be correct.

If the split flow accounts for a factor of four decrease in peak height it would have to increase to 400 ml/min - which is unlikely unless there is a gross problem with the pneumatics which would probably also affect retention.

A change in split ratio during a run cannot affect peak areas, since all the analytes transfer to the column within a few seconds of injection. Check that there is no pulse in split flow at injection - not likely but it needs to be eliminated from the list of suspects.

Double check that peaks actually get smaller during a run - if this is the case it is unlikely to be due to changes in gas flow at the inlet, although there is a small chance that you now have boiling point discrimination. It makes me suspect a detector problem. What is the baseline doing during a run ? Check the detector gas flow rates with the oven hot.

The 3400 had options for inlet pneumatics - do you have mechanical or electronic, front or back pressure control ?

Peter