Chromatography Forum

I have just had an ECD detector installed on my GC, replacing an FID. I made this upgrade to increase the sensitivity for a specific application (perfluorooctylbromide) that previously employed the FID.

The only modification to my method involved configuring the instrument set-up parameters to reflect the change in detector. However, I now find that the standard curves that I generate for calibration are quadratic rather than linear. Is this appropriate?

My instrument is a Thermoelectron TraceGC Ultra. One Thermo person says no way, it should be linear. Another says that the quadratic behaviour that I observe is correct. I must admit that I get wonderful fits to my data, and the numbers I generate for my actual samples agree well with theory. However, I'd like to have confidence that this is indeed the way that the detector is supposed to behave.

Any input/advice would be appreciated!

Thanks

ECD should give a linear response. FPD is quadratic in nature. Something is not right if you do not get a linear response.
Where are you in the world ?
Regards Richard
CE Instruments Ltd

Richard;

I'm in St. Louis, MO, US of A!

The more I look into this, the more the consensus seems to be that I should have linear behaviour. However, everyone seems to have a story of "someone" who once had a compound that demonstrated the same phenomenon, so it could therefore be "compound specific", yadayadayada.

I'm considering the possibility that it's linear over only a narrow range. My calibrants span the range of 0.1 - 75 ug/ml, so a little less than three orders of magnitude. However, the non-linearity is apparent even at lower concentrations; e.g., 0.1 - 2.5 ug/ml still gives a much better quadratic fit than linear.

I should have the parts in-house in a week or two to actually stack my detectors, so that the FID will be in-stream after the ECD and I can run both simultaneously. Then I can repeat this and compare the response of both detectors. That should prove interesting!

I wonder if the amp board is looking for FPD instead of ECD input - time for a jumper/dip switch check?

I wonder if the amp board is looking for FPD instead of ECD input - time for a jumper/dip switch check?

Not possible on a Thermo Trace GC

First I would run an ECD test mix and see if you get a linear response. once stacked you should get direct corolation between FID and ECD assuing the FID sees the compound at the injected levels.

You do not mention conditions Thermo ECD should be set up as column flow (Helium ? approx 1-2ml/min ?) 30ml/min Nitrogen make up
Base Temperature = 250 °C
ECD temperature = 300 °C
Reference Current = 1 nA
Pulse Amplitude = 50 V
Pulse Width = 1μs

Test compounds recomended are Lindane and Aldrin @ 0.030 ug/ml
These details should be in your SOP manual that came with the GC.
I would advise discussing this further with your supplier, the ECD is usually linear to about 4 orders of magnitude for most compounds. If installed by the agent they should have run a test standard etc. to check the operation.
We are based near Wigan in the UK so a service visit from us would prove expensive

I agree with jblackledge: ECD detector is nonlinear. The following is copied from Agilent 6890 GC manual:

"The ECD response factor versus concentration curve is nonlinear and the shape of the curve varies with the compound. A multipoint calibration is advised. When calibrating from the ChemStation, we suggest that you specify a second-order curve fit. When using an Agilent integrator, specisy a nonlinear fit"

I remember seeing a similar comment from HP GC manual. Based on my experience with ECD, it's a very sensitive detector, other than that, it sucks.

Why does it not surprise me
The Carlo Erba now Thermo ECD quotes 4 orders of linearity although for the best results I would still advise multipoint calibration not just a single point To investigate ECD design and why Carlo Erba claimed a superior ECD you can investigate the following reference
Ref. A. Zlatkis and C.F. Poole; Electron Capture: Theory and practice in
chromatography, Journal of Chromatography library Vol.20, 1981 I do not personally have a copy but only an internal sales memo mocking the launch of the micro ECD in 1998