Linearity problem using HPLC-ECD

[artsjeroen]

I always inject standards bracketed when using HPLC-ECD (two standards at 80% and 120% every six samples).

Now I am trying to inject the standards using a 5-point calibration curve for method validation. However, the calibration curves do not seem to be linear.
Is it possible that the flow-cell is not able to fully oxidize or reduce the analyte if higher concentrations are injected?

I was wondering how other laboratories handle the standards/calibration curve when using HPLC-ECD (bracketed/single point calibration/curve)?

[Markus Laeubli, Metrohm]

To get a useful answer, you need to specify.
What type of ECD cell, electrodes, measuring mode, analyte, concentration range etc.

[artsjeroen]

For example a UV detector there is the relation between the area response and the analyte quantity described by the Lambert Beer law. Which is linear. An Evaporative Light Scattering Detector is non linear, the data is plotted using polynomial calibrations or using log Detector response vs. log sample concentration.

How does it work for an Electrochemical Detector?
I use gold and glassy carbon working electrodes. I don't know if that would make a difference.

[Markus Laeubli, Metrohm]

Dear artsjeroen

Amperometric detection uses an electrochemical reaction and measures the current yielding from this reaction. Per se the current/concentration ration is stable and the response is linear.

Non-linearity my occur when too high concentrations reduce the diffusion of the sample to the electrode. Or when reaction products do the same.

The use of different electrode types (gold, silver, carbon) depends on the sample type and its reaction. But does not influence the linearity.