Deconvoluting CO₂ Electroreduction Membrane-Electrode-Assembly Performance Via Five-Electrode Setup

For CO 2 electrolysis, a zero-gap or membrane-electrode-assembly (MEA) configuration is desirable for lowering internal resistances and enabling the use of dilute supporting electrolytes for operation at industrially relevant current densities (>100 mA cm -2 ). However, to rationally optimize individual electrolyzer components, the contributions that cell components (cathode, anode, and membrane) have on the total cell voltage and internal resistance need to be deconvoluted. Moreover, deconvolution techniques can provide operando snapshots of electrode dynamics and component potentials to help assess component degradation for accelerated stability testing protocols 1 . In this work, we present a technique for the complete deconvolution of the cell voltage and internal resistances through a five-electrode setup. Unlike similar methods 1-3 , three additional reference electrodes are introduced: two quasi-reference electrodes on each side of the MEA and one fritted reference electrode in the supporting electrolyte, fed to the anode, for validation of quasi-reference electrode potentials before and after testing. Furthermore, the technique presented in this work was optimized for ease of implementation for a standard test setup for laboratory-scale CO2 electrolysis without requiring modifications to the electrochemical cell endplates or a multi-channel potentiostat. Using this five-electrode setup, a significant membrane impedance at both high and low frequencies is identified for a CO 2 reduction MEA employing an anion exchange membrane. As shown via electrochemical impedance spectroscopy (EIS) (Fig. 1b), the membrane accounts for about 40% of the total cell impedance at high frequencies. Notably, the membrane also exhibits a low-frequency impedance attributed to concentration gradients across the membrane 4 . As an initial validation check, the individual EIS spectra of each component were summed to obtain the measured full cell EIS spectra, albeit with an introduction of measurement noise (Fig. 1b, d). To further validate the setup, the corners of the cathode and anode electrodes were cut asymmetrically to assess sensitivity towards edge effects from electrode misalignment reported for a similar technique (Fig. 1a,c) 2 . We demonstrate that the membrane impedance can be reversed by deliberately introducing this edge effect (Fig. 1b,d). In this reversed configuration, the measured membrane impedance is flipped due to the apparent reference electrode positions being swapped. Moreover, in this configuration the user inadvertently measures the cathode and anode through the membrane, thereby convoluting the signal. This is of significant concern for directly measuring the cathode mass transport dynamics measured at low frequency (<100 Hz) as the membrane has its own characteristic low-frequency impedance. We will present these results and discuss the application of this method towards catalyst layer optimization, diagnosing a working CO 2 reduction MEA employing a cation exchange membrane and comparing the performance of zero-gap and hot-pressed MEA configurations. References O. Sorsa, J. Nieminen, P. Kauranen, and T. Kallio, J. Electrochem. Soc., 166, F1326–F1336 (2019) https://iopscience.iop.org/article/10.1149/2.0461916jes. R. Zeng, R. C. T. Slade, and J. R. Varcoe, Electrochim. Acta, 56, 607–619 (2010) http://dx.doi.org/10.1016/j.electacta.2010.08.032. D. Salvatore and C. P. Berlinguette, ACS Energy Lett., 5, 215–220 (2020) https://pubs.acs.org/doi/10.1021/acsenergylett.9b02356. A. Kozmai et al., Membranes (Basel)., 11, 1–17 (2021). Acknowledgement This work was supported by the U.S. Department of Energy, Office of Energy Efficiency & Renewable Energy under award no. DE-EE0009287.0001 Figure 1