FAME by GC (Perkin Elmer AutoSystemXL help!)

[jdlh199]

I'm running FAME by GC with FID detection using a standard FAME method with a Supelco FAME37 mix. This is a 30m DB-5 column (250µm x 0.25 µm) with a straightforward 50-240°C oven ramp, a 1mL/min He flow, and a 1:50 split injection of 1µL.

I'm using a Perkin Elmer AutoSystemXL running TotalChrom 6.2, however I am getting extremely mininal analyte peak signals. I have to zoom right in to see the analyte peaks and we see the baseline drift due to the oven temp increases.

We have another GC-FID system (Agilent 5890) which seems to produce nice chromatograms.

We have not used the FID detector on the AutoSystemXL for a number of years (since the instrument is also connected to a Turbomass Gold MS which is it's main use)!

What would cause minimal analyte peaks on an FID detector? Running by MS seems to be fine suggesting its a detector end problem rather than an injector end problem.

Does the FID detector need replacing? Can I clean it or optimise it somehow?

How can I test the sensitivity of an FID detector without paying a service tech to come out and perform a PQ test?

Here's some images:

AutoSystemXL Blank: http://i30.tinypic.com/9v9uti.jpg
AutoSystemXL FAME37: http://i27.tinypic.com/2vsgqbm.jpg
AutoSystemXL FAME37: http://i29.tinypic.com/wv693p.jpg
Agilent FAME37: http://i27.tinypic.com/e9dmxh.jpg

Thanks!

[chromatographer1]

I noticed the poor peak shape of the solvent peak on the PE.

Perhaps you should reinstall the column and retrim the front end of the column.

You should also check the flow rate through the column and the splitter flow so you know exactly the actual split ratio.

I suspect the detector is fine. A little detective work should prove enlightening.

best wishes,

Rod

[jdlh199]

I suspect the solvent peak shape differences are due to the presence of a headpressure on the Agilent but not on the PE.

But I'll trim and reinstall the column anyway.

[chromatographer1]

I am curious.

Why do you have head pressure on the Agilent but not on the Perkin Elmer?

Please explain.

Thanks.

Rod

[ratwing]

I posted something similar today - does your GC have make-up gas at the detector? (It sounds like you have the same GC as me)

[dolenjc2000]

I had the same problem with 37 FAME mix, but then I realized i diluted the standard too much
But otherwisse your peaks are fine...

[jdlh199]

1. I'm not sure why the headpressure auxillary gas is not connected. It never has been. I think this is something to do with the injector type. The PE is a PSS split/splitless injector. In any case wouldn't this affect the chromatography more than the sensitivity?

2. No make-up gas at the detector end. Running 45 mL/min H and 450 mL/min Air.

3. No dilution of FAME37 mix.

4. Trimming and reinstalling the column made no difference.

Also.... If I run more concentrated samples I get peak skewing due to column overloading, but the same low sensitivity. The analyte/solvent peak ratios are about the same between the Agilent and the PE, but the PE shows this huge baseline drift. Is this really low sensitivity or enormous baseline drift? If the baseline was flat I wouldn't have a problem.

I'm going to try cleaning the FID detector and replacing the jet. I still think it's a detector issue but I'm open to any suggestions!

Ta!

[DAMI]

The detector seems to be ok.
Maybe you could change the liner and recondtionning the column because your baseline is growing too much at the begining.
The retention time are very differents between agilent and perkin. have you the same column and the same temperature programation?
If all is the same you have maybe a leak in the injector.
Try to to test it with no column and close purge septum, split flow. Put a pressure like 25 psig and if the pressure decrease rapidly you have leak.

tell us later.

regards.

[jdlh199]

1. I'm not sure why the headpressure auxillary gas is not connected. It never has been. I think this is something to do with the injector type. The PE is a PSS split/splitless injector. In any case wouldn't this affect the chromatography more than the sensitivity?

Brain-freeze

Actually it's just that the PE is set to work in flow mode. He is set to 1mL/min, whereas the Agilent is set to work in headpressure, and was set to 20psi.

Different operator's prefer different things I guess.....

Any-hoo.... I'm baking the detector at 380°C to see if cleaning it makes a difference.

I also replaced the liner and conditioned the column, and will test and try the other suggestions on Monday - it's home time!

[AICMM]

jdlh199,

You are running 1 mL/min into an FID that was probably built for 30 or 40 mL/min of column flow. Not really a good idea. The MS could handle this because it is literally pulling the analyte in rather than it being pushed like the FID. Likewise, are you using a capillary jet or a packed jet and are you sure that the column is not sticking above the jet itself? Is the Agilent running with make-up (I suspect it is?) Look at other post on building a quick make up adapter.

Couple of other things; what happens if you run splitless or decrease the split flow to 10:1? Your solvent front is pretty ugly considering you are running with a 50:1 split. Solvent front could also be related to make up and poor sweep but look how sharp the Agilent solvent front is relative to the PE.

Best regards.

[jdlh199]

Hi AICMM,

I didn't realise that a make-up gas was really required for FID analysis. The gas flows are 450mL/min for Air, 45 mL/min for H, and 1mL/min for He. The FID is a capillary jet, and yes the column is not sticking above the jet since the operator's manual tells me how far it should be inserted and I would also see the glow from the capillary touching the flame if it was in too far. It's odd that there is nothing in the PE FID users manual about using or setting a makeup gas?

If I run splitless or decrease the split to 10:1, I get larger but wide fronting peaks (column overloading?) and a much wider solvent front.

I'm not sure whether the Agilent is running a makeup gas. I'll check in the morning. But there are no settings for a makeup gas on the PE. I'll look into the other post on building a quick makeup adapter.

Does the makeup gas flow (He or N?) go directly into the FID or through the column into the FID? In other words does the makeup gas flow into the injector port or into the detector.

Thanks for the tips!

[jdlh199]

So the Agilent does use He as a makeup gas.

The PE has no makeup gas, there is no mention of a makeup gas anywhere in the FID manual.... I'm confused! Is it absolutely essential for a makeup gas when running FID? If I run the column flow at 30 mL/min instead of 1 mL/min, won't that seriously screw the chromatography?

I baked the detector at 380°C for a few hours (this process did raise the background significantly and it slowly went back to normal again - perhaps I baked it clean?). Any-hoo I'm just running some test injections today.

But I'd appreciate feedback on the makeup gas issue!

[chromatographer1]

There is a checkout test of one sort or another used by almost all GC manufacturers to determine if the installation of the new GC is satisfactory.

I would run this test and compare the expected results with yours to see if your equipment is operating correctly.

On one GC if I remember correctly it consisted of a solution of alkanes, C14-C16, dissolved in dichloromethane at 1mg/mL and was split 100:1.

best wishes,

Rod

[jdlh199]

There is a checkout test of one sort or another used by almost all GC manufacturers to determine if the installation of the new GC is satisfactory.

I would run this test and compare the expected results with yours to see if your equipment is operating correctly.

On one GC if I remember correctly it consisted of a solution of alkanes, C14-C16, dissolved in dichloromethane at 1mg/mL and was split 100:1.

best wishes,

Rod

I have no acceptance or or comparative specs/chromatograms for the PE-FID. The only way we'll get these is if we place a service call and get the service tech to come out and perform a IQ/OQ/PQ test.

My boss who holds the purse-strings would prefer us to fix this in-house, which is why I'm running the FAME test mix.

The chromatograms have improved significantly since baking the detector. I'm going to run some more tests and will come back with some better chromatograms!

[AICMM]

jdlh199,

The detector you are using was originally designed in the 50's and, frankly, very little has been done to the commercial versions since then (almost.) Yes, if you run the column at 30 mL/min you are going to mash everything together. Make-up gas is added right at the end of the column as it goes into the detector. Even though you run lots of H2 and air, they are added in other places which does not move the analyte through in quite the same way.

Look in the Supelco catalog or Restek catalog since they will offer make-up gas kits that will probably suit your needs. Pretty easy to build your own as well, I do it all the time....

Your fame mix is probably not the best test mix (potentially reactive) but if that is what you have... You definitely need make-up.

Let me know if you need more info.