EIS: Potentiostatic or Galvanostatic Mode?
Electrochemical Impedance Spectroscopy (EIS) measurements are more often performed in potentiostatic mode than under galvanostatic mode. The reason is partly convention and partly driven by materials. For example, coatings and corrosion-resistant materials have a higher impedance, so that the application of a 10 mV sine wave results in a nA (or smaller) response. Conversely, in a low-impedance device such as a large supercapacitor or battery, the application of a 10 mV sine wave produces amperes of current-flow, possibly changing the state of the device—or worse, yet, damaging it. For devices with intermediate impedance (e.g., smaller batteries, smaller supercapacitors, and sensors), either potentiostatic or galvanostatic modes may be used.
Usually potentiostatic and galvanostatic modes are equivalent and give the same impedance diagrams. But sometimes−if the system evolves during the measurement−results from the two techniques may differ.
In corrosion applications, the polarization resistance is often determined under potentiostatic control around the open-circuit voltage (OCV). This is an appropriate approach if the corrosion potential does not change during the measurement. But if the corrosion potential drifts, a potentiostatic-mode experiment defined to run at OCV could result in a measurement being performed at an anodic or cathodic potential with respect to the true OCV.
Under galvanostatic control, drift is not a problem, for the desired zero-current condition is maintained throughout the experiment, ensuring that the measurement is performed at the true corrosion potential. In battery applications, determining the variation of internal resistance during discharge/charge is often worthwhile. Here, use of galvanostatic mode in EIS measurements may be better.
Both potentiostatic and galvanostatic modes are available in Gamry’s Framework™ software. In this application note, we compare potentiostatic and galvanostatic modes in EIS measurements on a commercial Li-ion cylindrical cell (model 18650) and a coin cell.
EIS measurements were performed in potentiostatic and galvanostatic mode on a commercial Li-ion battery (model 18650, nominal capacity 2500 mAh). After full charge, the battery was discharged at a C/10 regime to 2.7 V (see Fig. 3) and, after a rest period, the EIS measurements were performed (under potentiostatic or galvanostatic mode) around the OCV. Measurements in both modes were performed from 100 kHz to 10 mHz, with sine amplitudes of 2 mV and 10 mA for potentiostatic and galvanostatic modes, respectively. Figs. 1 and 2 show the parameter settings for these measurements.
Figure 1. Setup window for potentiostatic mode.
Figure 2. Setup window for galvanostatic mode.
The battery voltage changes with time during discharge, rest, and potentiostatic and galvanostatic EIS measurements. An example of a non-linear discharge is shown in Fig. 3.
Figure 3. Plot of voltage against time of 18650 cell during C/10 discharge.
The top graph in Fig. 4 shows the first cycle of both potentiostatic EIS and galvanostatic EIS, while the bottom shows the evolution of the potentiostatic EIS Nyquist plot during the discharge. Fig. 4 reveals that potentiostatic EIS and galvanostatic EIS show the same result in the whole range of frequencies, for both experimental conditions. The fact that these results are practically identical under potentiostatic and galvanostatic control confirms empirically that the galvanostatic EIS amplitude used was appropriate. Had the results been different, a new galvanostatic EIS amplitude could be tested and checked until a better value was found. As a rough approximation, the galvanostatic EIS amplitude can be chosen from the current modulus obtained in potentiostatic mode,|I|. This value usually is bigger than the minimum amplitude necessary to obtain the right measurements.
Figure 4. Upper: Nyquist plot of an 18650 cell under potentiostatic control (blue diamonds) and under galvanostatic control (hollow green triangles), both taken at a potential of ~3.15 V. Lower: Evolution of the EIS diagram (under potentiostatic control) during the discharge. Blue points are 4.19 V, green points are 3.15 V, and brown points are 2.87 V.
The change of internal resistance with the potential or State of Charge (SoC) is studied by EIS measurements during discharge (or charge). This resistance is determined by fitting the EIS graphs to an equivalent model circuit (constructed in Gamry’s Model Editor) in Gamry’s Echem Analyst™. Gamry’s Model Editor and Echem Analyst software provide a powerful, user-friendly tool to analyze the successive impedance measurements: Impedance Fit by the Simplex Method. In this method, the user clicks the AutoFit button to let the software iterate automatically, converging on the best fit to the data, and then superimposing the fit on the data. Fig. 6 shows the result of the fitting process (ignoring inductive effects of the cables) for the 18650 cell (Li-ion) for the potentiostatically-controlled experiment with the equivalent circuit (Fig. 5).
Figure 5. Equivalent circuit diagram from Gamry’s Model Editor, used to model the cells, with parameters attached to each element. The “null” symbol (Ø) represents a constant-phase element, and W indicates a Warburg circuit element.
Figure 6. Nyquist plots of potentiostatic EIS on an 18650 cell at various voltages (points), with fits (solid lines) provided by Gamry’s Echem Analyst software. Goodness-of-fits: χ2,2.94 V = 9.6 × 10–5; χ2,3.40 V = 9.4 × 10–5; χ2,4.19 V = 1.8 × 10–4.
EIS measurements were performed in potentiostatic and galvanostatic mode on a commercial coin cell (nominal capacity 40 mAh). After full charge, the battery was discharged under the C/10 regime (Fig. 8, upper graph), and, after a rest period, the EIS measurements were performed (under potentiostatic or galvanostatic mode) around the OCV. Measurements in both modes were performed from 100 kHz to 10 mHz, with sine amplitudes of 2 mV and 10 mA for potentiostatic and galvanostatic modes, respectively.
As with the 18650 cell, the coin cell’s voltage changes with time during discharge, rest, and potentiostatic and galvanostatic EIS measurements. Another example of a non-linear discharge is shown in Fig. 7.
Figure 7. Plot of voltage against time of coin cell during C/10 discharge.
Upper: Nyquist plot of a coin cell under potentiostatic control (blue diamonds) and under galvanostatic control (hollow red squares), both taken at a potential of 3.55 V.
Lower: Evolution of the EIS diagram (under potentiostatic control) during the discharge. Diamond-shaped points are 4.15 V, square points are 3.55 V, and triangular points are 2.90 V. Filled points are data; hollow points are the fits. Goodness-of-fits: χ2,2.90 V = 2.2 × 10–5; χ2,3.55 V = 4.2 × 10–5; χ2,4.15 V = 3.7 × 10–5.
The upper graph in Fig. 8 compares a potentiostatic EIS to a galvanostatic EIS scan. Because they are practically coincident, we can conclude that either method works for this system.
The lower graph in Fig. 8 shows the potential-dependency of potentiostatic EIS for three different open-circuit potentials. Fits (according to the model in Fig. 9) are represented by hollow points, and they match the data excellently.
Figure 9. Equivalent circuit diagram from Gamry’s Model Editor, used to model the cells, with parameters attached to each element. The “null” symbol (Ø) represents a constant-phase element. A “Bisquert” element includes a transmission line to model porous interfaces.
Gamry Instruments’ potentiostats are ideal for collecting many types of EIS data. EIS experiments in potentiostatic or galvanostatic mode are usually equivalent. For batteries we generally recommend a current amplitude of, at most, about 5% of the standard charge current as listed in the manufacturer’s specifications. One can also use (as a rough approximation) the value of the current modulus obtained from the potentiostatic EIS measurement.
As shown in this note, galvanostatic mode and potentiostatic mode are both usually valid for general EIS studies. We recommend galvanostatic control for EIS measurements on batteries with low internal resistances, and potentiostatic mode for high-impedance systems such as coatings or corrosion-resistant materials.