Skip to main content

Electrochemical Impedance Spectroscopy

A practical, physics-first guide to the principles, measurement, plots, equivalent circuits and interpretation of electrochemical impedance spectroscopy.
SCIENTIFIC INSTRUMENTS · ELECTROCHEMICAL MEASUREMENTS

Electrochemical
Impedance Spectroscopy

Understand what an impedance analyser actually measures, why the response depends on frequency, and how to turn the data into defensible physical insight.

MEASUREMENT PRINCIPLEEXPERIMENTAL WORKFLOWDATA INTERPRETATION 
<span class="eg-label">THE CENTRAL IDEA</span>
<h2>EIS is a frequency-resolved measurement of how a system opposes a small AC perturbation.</h2>

Instead of measuring only one resistance value, electrochemical impedance spectroscopy (EIS) measures the amplitude and phase of the response over a range of frequencies. Different physical processes respond on different characteristic time scales. Frequency therefore becomes a way to separate overlapping contributions from the electrolyte, interfaces, charge transfer, and mass transport.

01 · What does the instrument measure?

An EIS instrument applies a small sinusoidal voltage or current to an electrochemical system and measures the resulting response. In a potentiostatic experiment, the voltage perturbation is controlled and the current response is measured. In a galvanostatic experiment, the current perturbation is controlled and the voltage response is measured.

For a sinusoidal voltage input,

[ E(t)=E_0(t) ]

the current response of a linear system is also sinusoidal at the same frequency, but it can have a different amplitude and a phase shift:

[ I(t)=I_0(t+) ]

The complex impedance is defined as

[ Z()= =Z’()+jZ’’() ]

where:

  • (Z’) is the real component: the in-phase contribution.
  • (Z’’) is the imaginary component: the out-of-phase contribution.
  • (j) is the imaginary unit, (j^2=-1).
  • (f) is angular frequency in rad s(^{-1}).

The impedance magnitude and phase are

[ |Z|= =^{-1}() ]

The instrument does not directly “see” a grain boundary, an electrode reaction, or an oxygen vacancy. It measures voltage and current, then calculates their complex ratio. Assigning a measured feature to a physical process is an interpretation that must be supported by controls, frequency dependence, temperature dependence, and a physically appropriate model.

Important distinction: resistance is a real, frequency-independent quantity for an ideal resistor. Impedance is generally complex and frequency-dependent. Reporting every EIS feature as a “resistance” without considering its capacitive or inductive behaviour loses information.

02 · Why does frequency reveal different processes?

A material or electrochemical cell contains processes with different response times. A fast process can follow a rapidly changing perturbation; a slow process cannot. At high frequency, slow processes may appear effectively inactive. At lower frequency, progressively slower processes can contribute.

Typical contributions include:

Contribution Common physical origin Typical electrical behaviour
Series / ohmic resistance Electrolyte, leads, contacts, current collectors Predominantly resistive
Bulk response Mobile charge in the crystal lattice Often represented by an R–C or R–CPE element
Grain-boundary response Grain-boundary barriers and defect segregation Often a separate, slower relaxation
Electrode response Interfacial charge transfer and chemical capacitance R–C, R–CPE, or more complex response
Diffusion / transport Mass transport or finite-length ionic transport Warburg-like or finite-length response
Inductive contribution Leads, wiring, fixture, or genuine inductive processes Positive imaginary impedance under the usual convention

The frequency ranges are not universal labels. A grain-boundary arc is not automatically “the middle-frequency arc” in every material. Its position depends on conductivity, capacitance, geometry, temperature, electrode configuration, and the instrument’s measurement window.

The time-constant connection

For an ideal parallel resistor–capacitor element,

[ =RC f_0= ]

The characteristic frequency is the frequency at which the magnitude of the capacitive and resistive reactance contributions are comparable. A smaller time constant shifts the relaxation to higher frequency; a larger time constant shifts it to lower frequency.

This is why temperature-dependent EIS is particularly useful in solid-state ionics: conductivity and relaxation times can change strongly with temperature, shifting arcs across the accessible frequency range.

03 · The equivalent-circuit model

An equivalent circuit is a compact mathematical representation of the measured response. It is not a literal photograph of the microstructure. Different circuits can sometimes fit the same finite frequency window, so circuit selection must be guided by physical plausibility and independent evidence.

Start with the simplest useful modelIdeal Randles-type response
Rₛ Rct Cdl InputOutput
Z(ω) = Rₛ + [ 1/Rct + jωCdl ]⁻¹

Here, (R_s) is the series or solution resistance, (R_{ct}) is the charge-transfer resistance, and (C_{dl}) is the double-layer capacitance. The parallel branch has impedance

[ Z_{}= ]

Therefore,

[ Z’=R_s+ ]

[ Z’’=- {1+(R_{ct}C_{dl})^2} ]

For this ideal model, the Nyquist response is a perfect semicircle:

  • The high-frequency real-axis intercept is (R_s).
  • The low-frequency real-axis intercept is (R_s+R_{ct}).
  • The semicircle diameter is (R_{ct}).
  • The maximum of (-Z’’) is (R_{ct}/2).
  • The characteristic frequency is (1/(2R_{ct}C_{dl})).

This simple model is a teaching model. Real electrodes and ceramic materials often show depressed arcs, overlapping processes, or diffusion tails. Those features require a more appropriate model rather than forcing an ideal semicircle onto the data.

Resistors, capacitors, CPEs, and diffusion elements

Element Impedance Interpretation / caution
Resistor (R) (Z=R) Ideal dissipative contribution
Capacitor (C) (Z=1/(jC)) Ideal energy storage
Constant-phase element (CPE) (Z=1/[Q(j)^n]) Non-ideal distributed response; (n=1) is an ideal capacitor
Warburg element Frequency-dependent diffusion impedance Semi-infinite diffusion gives a 45° line under ideal assumptions
Inductor (L) (Z=jL) Can arise from wiring and fixtures as well as the sample

A CPE is often used to represent a distribution of relaxation times, surface heterogeneity, roughness, or non-uniform current distribution. Its parameter (Q) is not generally equal to a capacitance unless the exponent and circuit context justify a conversion.

04 · How to read the main EIS plots

Nyquist plot

A Nyquist plot places (Z’) on the horizontal axis and (-Z’’) on the vertical axis under the common electrochemical sign convention. Each point corresponds to one excitation frequency, but frequency is not normally displayed as an axis. The direction from high to low frequency should therefore be identified from the acquisition sequence or frequency labels.

  • A high-frequency intercept is often used to estimate the ohmic contribution.
  • A semicircle can represent a relaxation process, but its physical assignment is not unique.
  • Multiple arcs may indicate processes with different time constants.
  • A low-frequency inclined tail may indicate diffusion or distributed transport.
  • A feature below the real axis can indicate an inductive response, depending on the plotting convention.

Bode plots

Bode plots show impedance magnitude and phase as functions of frequency, usually with logarithmic frequency axes. They preserve the frequency information that is hidden in a conventional Nyquist plot.

  • (|Z|) versus frequency shows how the total opposition changes.
  • Phase versus frequency shows whether the response is mainly resistive, capacitive, or inductive.
  • A phase maximum/minimum can help locate a relaxation, but overlapping processes can distort it.
  • A broad phase feature may indicate distributed time constants rather than one ideal RC process.
Practical reading rule: Do not interpret a Nyquist arc in isolation. Check the corresponding Bode magnitude and phase, confirm the frequency range covered, and examine whether the feature moves consistently with temperature or another controlled variable.

05 · How an EIS measurement is performed

  1. Prepare the sample and electrodes. Record sample dimensions, electrode area, electrode material, contact method, surface preparation, and cell configuration. For a pellet, document thickness and the effective electrode area used for geometry correction.
  2. Establish the measurement condition. Set temperature, atmosphere, humidity or gas composition where relevant. Allow the sample and furnace to reach thermal and electrical stability.
  3. Choose the perturbation mode. Use potentiostatic EIS when controlling a small voltage perturbation, or galvanostatic EIS when controlling a small current perturbation. Select a perturbation amplitude that gives adequate signal without driving the system outside its linear regime.
  4. Set the frequency window. Choose upper and lower limits based on the expected time constants, instrument capability, sample resistance, and measurement time. Use sufficient points per decade to resolve the features of interest.
  5. Check linearity and stability. Repeat selected spectra using a smaller perturbation amplitude. Compare repeated measurements and verify that the system does not drift during the low-frequency acquisition.
  6. Inspect the raw spectrum. Review Nyquist and Bode plots before fitting. Look for outliers, inductive lead effects, missing intercepts, noisy low-frequency points, and incomplete arcs.
  7. Fit a physically justified circuit. Start with the simplest model consistent with the data. Inspect residuals, parameter uncertainty, parameter correlations, and whether fitted values are physically plausible.
  8. Convert to material properties carefully. Apply geometry corrections only to the appropriate bulk or total resistance and state the convention used.

Geometry correction for a pellet

For a homogeneous specimen with thickness (L), electrode area (A), and resistance (R), the conductivity is

[ = ]

Use consistent units. If (L) is in centimetres, (A) in cm(^2), and (R) in ohms, conductivity is obtained in S cm(^{-1}). The resistance chosen must correspond to the process being reported. For example, bulk conductivity should not be calculated from a total resistance that includes electrode polarization unless that is explicitly the intended quantity.

For a pellet with several resolved contributions, a common first analysis is to estimate separate bulk and grain-boundary resistances, then calculate the relevant conductivity using the stated geometry. Whether that separation is justified depends on whether the arcs are resolved and whether the model is physically defensible.

06 · Special considerations for oxide-ion and proton-conducting ceramics

In ceramic electrolytes, EIS is frequently used to distinguish bulk, grain-boundary, and electrode contributions. These contributions may overlap, and their relative importance changes with temperature, atmosphere, hydration state, and electrode chemistry.

Bulk response: Often appears at higher frequency because the associated relaxation time is short. Its capacitance is commonly smaller than that associated with grain boundaries or electrodes, but this is a trend, not a universal identification rule.

Grain-boundary response: Can arise from space-charge layers, impurity segregation, secondary phases, porosity, or blocking barriers. A grain-boundary resistance may dominate the total measured resistance even when the bulk material is highly conductive.

Electrode response: Depends on electrode material, contact quality, gas exchange, surface kinetics, and whether the electrode blocks or conducts the mobile species. At low frequency, electrode polarization can dominate the spectrum.

Temperature and atmosphere: For proton conductors, water vapour partial pressure and thermal history can alter hydration and proton concentration. For oxide-ion conductors, oxygen partial pressure and electrode kinetics may matter. State gas composition, flow, humidity, equilibration time, and heating/cooling protocol.

A useful ceramic EIS report should specify sample dimensions, electrode area and material, atmosphere, temperature, perturbation amplitude, frequency range, points per decade, instrument configuration, equivalent circuit, fitting method, and the exact resistance used for conductivity calculations.

07 · Common errors and how to diagnose them

Symptom Possible cause What to check
High-frequency arc or intercept is missing Upper frequency limit too low; cable or fixture inductance Extend frequency if instrument permits; inspect open/short compensation and wiring
Noisy low-frequency data Drift, insufficient settling, unstable temperature or atmosphere Repeat spectrum; improve equilibration; reduce drift
Arc changes with AC amplitude Non-linear response or perturbation too large Repeat at smaller amplitudes and compare
Fitted parameters vary wildly Over-parameterized circuit, overlapping arcs, poor frequency coverage Simplify model; inspect confidence intervals and residuals
Apparent negative resistance or implausible capacitance Wrong topology, sign convention, bad data, or fitting local minimum Check raw data, model definition, units, and fit constraints
Spectrum changes between consecutive runs Sample not stationary; evolving electrode or hydration state Stabilize sample and document time/history
Inductive loop at high frequency Leads, cables, fixture, or genuine inductive behaviour Check wiring geometry and instrument compensation

Data validation before fitting

A good fit is not proof that the circuit is physically correct. Before accepting a fit:

  • Confirm that the spectrum is reproducible and the sample is stationary.
  • Inspect residuals across the entire frequency range, not only the visual overlap.
  • Check parameter confidence intervals and correlations.
  • Avoid adding extra circuit elements solely to reduce the fitting error.
  • Where available, use a Kramers–Kronig consistency test to assess whether the data are compatible with a linear, causal, stable system over the measured range. Passing such a test does not identify the correct equivalent circuit.

08 · What should be reported?

At minimum, report the measurement mode (potentiostatic or galvanostatic), perturbation amplitude, DC bias or open-circuit condition, frequency range, points per decade, temperature and atmosphere, sample geometry, electrode details, instrument and fixture configuration, equivalent circuit, fitting software/method, and uncertainty or fit-quality indicators.

When reporting conductivity, state the exact resistance used, the geometry equation, and the units. When comparing samples or temperatures, use the same measurement and fitting conventions, or explain any differences.

09 · Optional interactive model

The simulator below is a compact teaching aid for the ideal circuit (R_s+(R_{ct}C_{dl})). It calculates the complex impedance from the analytical equation and redraws the Nyquist and Bode responses as parameters change. Use it to understand how resistance changes the arc width and how capacitance shifts the characteristic frequency.

What this model can teachIt demonstrates the mathematics of one ideal RC relaxation. It does not simulate a real ceramic, predict experimental data, or replace instrument software and equivalent-circuit fitting. 
<aside class="eis-panel" id="eis-controls">
  <div class="eis-panel-head"><small>IDEAL RC MODEL</small><h2>Change one parameter at a time</h2><span class="eis-engine-status" id="eis-engine-status" role="status" aria-live="polite">Preparing the model…</span></div>
  <div class="eis-controls">
    <div class="eis-control"><div class="eis-control-top"><label for="eis-rs">Series resistance (Rₛ)</label><output id="eis-rs-out">20 Ω</output></div><input id="eis-rs" type="range" min="0" max="200" step="1" value="20"><small>Moves both real-axis intercepts.</small></div>
    <div class="eis-control"><div class="eis-control-top"><label for="eis-rct">Charge-transfer resistance (Rct)</label><output id="eis-rct-out">500 Ω</output></div><input id="eis-rct" type="range" min="10" max="2000" step="10" value="500"><small>Sets the ideal semicircle diameter.</small></div>
    <div class="eis-control"><div class="eis-control-top"><label for="eis-cdl">Capacitance (Cdl)</label><output id="eis-cdl-out">10 µF</output></div><input id="eis-cdl" type="range" min="0.1" max="100" step="0.1" value="10"><small>Shifts the characteristic frequency.</small></div>
    <div class="eis-control"><div class="eis-control-top"><label for="eis-fmin">Minimum frequency</label><output id="eis-fmin-out">0.1 Hz</output></div><input id="eis-fmin" type="range" min="-3" max="2" step="1" value="-1"><small>Logarithmic sweep lower limit.</small></div>
    <div class="eis-control"><div class="eis-control-top"><label for="eis-fmax">Maximum frequency</label><output id="eis-fmax-out">1 MHz</output></div><input id="eis-fmax" type="range" min="2" max="7" step="1" value="6"><small>Logarithmic sweep upper limit.</small></div>
    <button class="eis-reset" id="eis-reset" type="button">Reset model</button>
    <div class="eis-error" id="eis-error" role="alert"></div>
  </div>
</aside>
<div class="eis-main">
  <div class="eis-metrics" aria-live="polite">
    <div class="eis-metric"><span>High-frequency intercept</span><strong id="eis-rinf">20 Ω</strong></div>
    <div class="eis-metric"><span>Low-frequency intercept</span><strong id="eis-rzero">520 Ω</strong></div>
    <div class="eis-metric"><span>Characteristic frequency</span><strong id="eis-fchar">31.8 Hz</strong></div>
    <div class="eis-metric"><span>Peak −Z″</span><strong id="eis-zpeak">250 Ω</strong></div>
  </div>
  <div class="eis-chart-grid">
    <section class="eis-chart"><div class="eis-chart-head"><strong>Nyquist plot</strong><span>−Z″ vs Z′</span></div><svg id="eis-nyquist" viewBox="0 0 480 330" role="img" aria-label="Calculated Nyquist plot"></svg><p class="eis-chart-note">The ideal RC branch produces a semicircle.</p></section>
    <section class="eis-chart"><div class="eis-chart-head"><strong>Bode magnitude</strong><span>|Z| vs frequency</span></div><svg id="eis-bode-mag" viewBox="0 0 480 330" role="img" aria-label="Calculated Bode magnitude plot"></svg><p class="eis-chart-note">Magnitude is plotted against logarithmic frequency.</p></section>
    <section class="eis-chart"><div class="eis-chart-head"><strong>Bode phase</strong><span>Phase vs frequency</span></div><svg id="eis-bode-phase" viewBox="0 0 480 330" role="img" aria-label="Calculated Bode phase plot"></svg><p class="eis-chart-note">The phase depends on the complete series-plus-parallel circuit.</p></section>
    <section class="eis-chart"><div class="eis-chart-head"><strong>Model equations</strong><span>Frequency domain</span></div><div class="eis-reference"><div>Angular frequency: <b>ω = 2πf</b></div><div>Time constant: <b>τ = Rct Cdl</b></div><div>Characteristic frequency: <b>f₀ = 1/(2πτ)</b></div><div>Capacitive impedance: <b>ZC = 1/(jωCdl)</b></div><p class="eis-reference-note">The curves are calculated directly from the ideal circuit equation, not from measured data.</p></div></section>
  </div>
</div>
Model limitation: no CPE, diffusion, inductance, grain-boundary distribution, electrode geometry, noise, or instrument artefacts are included. Real systems may need a different circuit.

Key takeaway

EIS is not simply a tool for drawing semicircles. It is a measurement of a system’s complex frequency response. The instrument measures voltage and current; the scientist connects the measured response to physical processes through careful experimental design, data validation, and a justified model.

← Back to Electrochemical Instrumentation · Scientific Instruments library