help! heartbeat peaks from an old GC

[nan-tze]

Hi,

My supervisor and I are trying to fix up an old GC system (Perkin-Elmer 8310, FID detection) with autosampler, which is older than I am, so that i can use it to measure CO2 concentrations (working on soil respiration)

We got hold of a PeakSimple data system to replace the old data printer, and we've set that all up, but when we took it for a test drive it gave these odd shaped peaks- after it peaks it drops back below the baseline and then recovers, like a heartbeat. The area of the inverse peak is about the same as the peak above the baseline.

Someone suggested we re-align the valco valve, and after we did that we managed to get the right kind of peak, by injecting test samples manually into the sample loop and manually controlling the valve on the computer (which hadn't worked before)

But then, as soon as we reconnected the autosampler, the old negative peaks came back.


Neither me nor my supervisor have worked with a GC before, and we're exhausting the patience of people trying to help us. If anyone has any ideas about what might cause peaks with valleys after them, we'd be very very grateful....

[HbJ]

I presume your GC is equipped with a methanizer as you detect CO2 with an FID.

Regarding the peak shape I'd look at the syringe of the autosampler: Is it new _and_ gas-tight (Teflon tips indicate a gas tight syringe)?

And please tell something about your instrument conditions (carrier gas, sample size, column etc.). Thanks!

[Peter Apps]

Despite what their sales agents tell you, GCs are not machines that you can just switch on and start analysing samples - as you are now finding out. Life is even more difficult whan you have to resuscitate an old piece of hardware. You really need a chromatographer to help you with this - they can probably do in a couple of days what you will struggle for weeks with. In the end you might even find that it works out more effective to get the samples analysed in a chromatography lab - though I am sure that there are all sorts of budget constraints and other reasons that look as if they make that option unattractive.

An FID cannot detect carbon dioxide - you have to have a thermal conductivity detector, or as already mentioned, a methanizer.

Good luck

Peter

[nan-tze]

Thanks both!

in answer to questions, yes we have a thermal conductivity detector, not a methanizer, the carrier gas is He and the syringe on the autosampler certainly doesn't have a teflon tip but as it's the same syringe as i used to manually inject and get a good peak, I don't think that's where the leak is if there is one. Sample size- we've tried two different sample loops, one 200ul and one 20ul.

Since I last wrote, we got a positive peak (although a bit lagging) using the 20ul with the errant autosampler, by increasing the inject time, suggesting maybe it's our timing that's wrong. but then again when we tried to manually inject standards to check if it was the right peak, we went back to getting the peak-dip shape, and the 'peaks' for two different concs were the same size/shape. I'm taking that to mean there was something wrong with my manual injection, not the machine, but that may be optimistic.

Someone suggested negative peaks should mean the gas flow is being disrupted. Could this be because our evacuation sequence/timing is wrong, as well as the injection? I'll test the times again, but what else could disrupt the flow?

And yes, of course we need a real chromatographer, we're just trying to ration the patience of the over-stretched technicians that there are around.
Out-sourcing analysis is not really an option, for practical reasons as well as budget. Also I was kind of hoping I might learn how to use the thing over the next 3 years, so that when I'm Dr. Unemployed-Overqualified I can help out the next new student. But if we don't fix it maybe it'll never happen.

[Peter Apps]

If we are going to help we need details of your set up......

You are analysing CO2 from soil, so I presume that you are injecting gas (air) samples. You have a syringe that you use for both manual and autosampler operations - to inject gas you need a syringe with a teflon tip on the plunger to make it gas tight (I presume that you are using a special GC sampling syringe). You talk about sample loops, which suggests that you are using a gas sampling valve, which is exactly what you need, and so I presume that you are using the syringe to fill the loop. What volume do you inject from the syringe into the loop ? Does the autosampler run the syringe or the valve or both ? Where are you using vacuum ? Are you heating the sampling valve ?, what temperature ?

What column are you using - length, diameter and stationary phase ?. What is the carrier gas flow rate and the column temperature ?

That will do for a start

Peter

[AICMM]

nan-tze,

Colunn information would be very useful. I would guess, and nothing more at the moment, that you might be dealing with water issues but I cannot really say without more information. Do you have filters on your helium and have you leak checked your plumbing?

Best regards.

[nan-tze]

Thanks everyone for your offers of help,

We fingered the culprit- a clogged up old drying precolumn. Taken it off. Finally working!

N