How to read your impedance spectrum

EIS is one of the most powerful diagnostic tools available to electrochemists. It’s also one of the most misread.

Electrochemical Impedance Spectroscopy (EIS) works by sending a small electrical signal through your cell at many different frequencies, then measuring how the cell responds at each one. Because different physical processes inside the cell occur at different timescales, a single measurement can separate them – membrane resistance, electrode kinetics, and mass transport effects all show up in different parts of the spectrum.

Think of it like listening to a complex machine: a trained ear can isolate the engine from the gearbox from the exhaust, even when all three are running at once. EIS can do the same thing electrically.

What the Nyquist plot is showing you

“Very” high frequency (>~100 kHz): Parasitic signals – Often this range is dominated by contributions from the measurement setup itself rather than the electrochemical cell. Cable inductance and resistances, connector contact impedances, and stray capacitances between leads all manifest at very high frequencies while the cell behaves as a pure resistor. A good practical rule is therefore to limit the upper frequency boundary to the point where the Nyquist plot first crosses or approaches the real axis as there is little electrochemical information above this point.

High frequency (~1 kHz and above): Ohmic resistance — the high frequency intercept with the real axis (where the imaginary part is equal to 0). The combined resistance of everything that doesn’t depend on the reaction itself: membrane resistance, contact resistance between components, and electrolyte bulk resistance. If this number is rising over time, start by checking membrane condition, bolt torque, and electrolyte concentration.

Mid frequency (a few Hz to a few hundred Hz): Charge transfer resistance — how easily the electrochemical reaction gets started at the electrode surface. A large semicircle here points to slow electrode kinetics, which can come from poor catalyst activity, surface fouling, or – in alkaline electrolysis – iron content and the oxidation state of the nickel surface.

Low frequency (below ~1 Hz): Mass transport — how fast reactants reach the electrode and products leave. In a flow cell, this responds directly to flow rate, electrode pore structure, and thickness. Significant Warburg-type features here mean the reaction is running faster than fresh reactant can arrive.

Tips & common pitfalls to avoid when reading EIS data

1. 1. Understand the I-V curve before performing any EIS. You must carefully select the DC operating potential and AC perturbation amplitude depending on the interrogation scope. If EIS is performed at OCV on a system that requires significant overpotential to drive faradaic reactions, such as a PEM electrolyzer like our X-Cell, the electrode kinetics are essentially inactive, and the resulting spectrum will reflect the unactivated interfaces rather than an operating cell.
      
      
   2. Do not compare spectra taken at different current densities as if they’re equivalent – charge transfer resistance is current-dependent, so they’re not directly comparable.
      
      
   3. Look out for your parasitic signals – attributing high-frequency irregularities to chemistry when the cause is cable inductance from a poorly set-up measurement can ruin your results.
      
      
   4. Use equal axis scaling on Nyquist plots. Ensure that the real and imaginary axes are plotted on the same scale (equal aspect ratio). If the axes have different units-per-division, semicircles will appear compressed or stretched into ellipses, making it difficult to visually identify the number of time constants, resistances from arch diameters, or compare spectra across conditions. A distorted Nyquist plot is one of the most common sources of EIS misinterpretation.
      
      
   5. Do not fit equivalent circuit models without a physical interpretation – the mathematics will fit almost anything, which makes an unconstrained fit meaningless. Each element must represent a physical process.
      
      

The Zahner IM7 potentiostat covers 10 µHz to 8 MHz – the full range needed to resolve everything from membrane resistance at high frequencies down to slow diffusion processes at low frequencies, without accuracy trade-offs at either end.

Read our whitepaper on the measurement of overpotentials and liquid potentials, focusing on understanding the overall energy losses in flow battery R&D systems. The experimental setup comprises of our S-Cell, OCV cell, an 8-channel battery tester, thermally treated AVCarb felt electrodes (G475), Fumatech FS930 membranes, and flow-through electrode holders for the reference electrodes.