I am trying to determine the ionic resistance from the electrolyte between working and reference electrode. The electrolyte is flowing between these two. I cross-check potentiostatic EIS (PEIS) and current-interrupt (CI) method. While starting conditions are equal between the collection of such measurements, I keep noticing that whenever there is a semicircle for positive imaginary impedance (at the high frequency limit), the CI method fails to give a conclusive result (blue data in the figure). When this semicircle is absent, both methods seem to work appropriately (red data). The red data for example gave me 3.29 Ω by CI method, and the zero imaginary impedance axis cut-off gave a value of ~3.1 Ω. I am not sure whether experimental noise could be a problem, but I found it remarkable that failure to use the CI method always corresponds with this particular feature in the impedance spectrum.
How can this semicircle be explained in qualitative terms, and what implication can it have for the current-interrupt method?
Some experimental details follow here:
- The spacing between the electrodes is approx. 8 mm in 1.0 M KOH.
- I run PEIS at open circuit potential (OCP) from 0.1 MHz to 1 Hz with 50 mV perturbation.
- The CI method steps from +0.1 V versus OCP to zero current condition, with sampling rate at 0.5 MHz (2 μs).
- Both PEIS and CI are conducted after stabilization of the OCP to within 1 mV drift/min.