How Membranes and Three-Electrode Configurations Affect Stability and Data Reliability in CO₂RR GDE Flow Cells

In carbon dioxide electroreduction (CO₂RR) research, it is natural to focus first on the catalyst itself: metal composition, facets, defects, coordination structures, particle size, loading, or interfacial modification. Catalysts are unquestionably important. However, in flow cells and gas diffusion electrode (GDE) systems, the measured Faradaic efficiency, selectivity, current density, and stability are not determined by the catalyst alone. When the same catalyst gives different results in different laboratories, reactors, or membrane, electrode, and flow-cell configurations, the discrepancy often reflects changes in the local reaction environment created by the entire integrated membrane, electrode, and fluidic environment rather than simple “catalyst failure” [1,2].

The previous two blog articles discussed CO₂RR platform design, fluid management, mass-transport pathways, and GDE flow-cell selection. This third article moves closer to the core of data reliability: how membranes, working electrodes, reference electrodes, and counter electrodes should be selected. Although these components may appear to be accessories, they directly influence ion-transport pathways, local pH, product crossover, ohmic losses, potential measurement accuracy, gas-liquid-solid three-phase interface stability, and long-term material compatibility.

For a GDE flow cell, the reaction no longer follows the slow pathway typical of an H-cell, where CO₂ must first dissolve in the electrolyte and then diffuse to the catalyst surface. Instead, gaseous CO₂ can reach the vicinity of the catalyst layer directly through the porous electrode, shortening the transport distance and increasing local CO₂ availability [1,3]. This advantage also makes the system more sensitive: small changes in membrane resistance can alter the cell voltage; differences in electrode pore structure can trigger flooding or salt precipitation; reference-electrode positioning can introduce substantial iR errors; and counter-electrode instability can introduce contamination or side reactions. Reliable CO₂RR data therefore require more than catalyst formulation. They also require clear reporting of the membrane, electrodes, flow rates, electrolyte, reference system, and compensation method.

1. Membrane Selection: Ion Transport, Local pH, and Product Crossover

In a CO₂RR electrolyzer, the membrane first separates the cathode and anode environments, preventing products, gases, and oxidized or reduced intermediates from directly mixing. Second, it provides an ion-conduction pathway to close the electrochemical circuit. Third, and more subtly, it helps regulate the local pH, ionic composition, and product migration near the cathode, which can be decisive for selectivity [2,4].

From the perspective of voltage distribution, the cell voltage of a practical CO₂RR electrolyzer can be expressed approximately as:

Here, E_cell is the actual electrolyzer voltage, E_thermo is the thermodynamic voltage, η_cathode and η_anode correspond to cathodic and anodic kinetic overpotentials. The term iR_Ω represents the overall ohmic voltage loss of the electrolyzer, where R_Ω denotes the overall ohmic resistance, while ηtransport reflects mass-transport limitations. The membrane contributes substantially to R_Ω through its ionic conductivity and thickness, while its ion selectivity, hydration state, and product permeability influence η_transport and long-term material balance [2,5].

In CO₂RR flow cells, the most common membranes are anion-exchange membranes (AEMs) and proton-exchange membranes (PEMs), while bipolar membranes (BPMs) are more often used for specialized reaction-environment design or mechanistic studies [2,6].

Table 1. Differences among AEM, PEM, and BPM

Note: Membrane performance depends on polymer chemistry, ion form, thickness, pretreatment, hydration state, temperature, and operating conditions. The entries above summarize general tendencies rather than universal specifications.

Beyond these qualitative differences, membrane properties can vary substantially with polymer chemistry, ion form, thickness, pretreatment, hydration state, temperature, and measurement method. AEMs are commonly selected to support anion transport and alkaline cathode environments, whereas PEMs provide strong proton conduction. BPMs combine cation- and anion-exchange layers and use interfacial water dissociation to generate H⁺ and OH⁻, enabling greater control over the pH environments on the two sides of the membrane. Because ion-exchange capacity and conductivity can vary significantly among commercial membranes and test protocols, these properties should be reported for the specific membrane used rather than treated as universal values [2,5,6].

Figure 1. Representative carbon-flow pathways in a CO₂ electrolyzer, illustrating how carbonate formation, membrane crossover, and product transport can affect overall carbon utilization. Reproduced with permission from Ref. [2].

Figure 1 illustrates why membrane selection affects more than ionic conductivity. During CO₂RR, OH⁻ generated near the cathode can react with CO₂ to form HCO₃⁻ and CO₃²⁻. In an AEM-based system, these anionic carbon species can migrate toward the anode, contributing to carbon loss from the cathode stream. Liquid products may also cross the membrane, depending on product chemistry, membrane properties, and operating conditions. PEM- and BPM-based architectures alter these transport pathways by changing the dominant charge carriers and local pH environments. The resulting trade-offs among carbon utilization, selectivity, cell voltage, and water management are therefore strongly dependent on the complete membrane–electrode configuration rather than membrane type alone [2,5,6].

In current CO₂RR research, AEMs remain the most commonly used membrane type in GDE flow cells. Nevertheless, AEMs also face challenges such as carbonate migration, decreased carbon efficiency, and liquid-product crossover [5,7]. PEMs are better suited to acidic environments or proton-involved mechanistic studies, whereas BPMs provide a route to regulate different cathode and anode pH environments, though usually with higher cell voltage and more complex water management requirements [2,6].

Recent developments in CO₂ electrolysis show that membranes are evolving from simple ion-conduction media into active regulators of the reaction environment. Membrane type determines not only ion-migration direction, but also local pH, carbonate formation and migration, water management, and product crossover. It has therefore become a central factor affecting CO₂RR performance and stability [2,4–7]. For example, AEM configurations typically create a distinct local reaction environment that can influence carbonate formation, salt precipitation, liquid-product crossover, and C–C coupling pathways. At the same time, they are more susceptible to salt precipitation, flooding, carbonate migration, and liquid-product crossover. PEM (or cation-exchange membrane) configurations provide stronger proton conduction and may alleviate salt precipitation to some extent, but they can also intensify HER or allow liquid products such as formate and methanol to permeate. BPMs, physically stacked AEM+PEM architectures, and porous composite membranes are all attempts to balance lower carbon loss and salt precipitation against the need to maintain favorable local pH [2,4–7].

First, what is the target product?

Membrane selection should be considered together with the catalyst, electrolyte, and intended reaction environment. For CO or formate production, AEM-based configurations are commonly used because they can support alkaline or locally alkaline cathode environments and reduce proton availability, which may suppress HER under suitable conditions. For deliberately acidic CO₂RR environments, PEM- or other cation-conducting configurations may be considered, provided that HER competition is adequately controlled. For C₂₊ products, particularly ethylene and ethanol on Cu-based catalysts, AEMs remain widely used, but carbonate transport, liquid-product crossover, water management, and long-term salt accumulation should be evaluated at the same time [1,2,4].

Second, is the experiment aimed at short-term activity or long-term stability?

The importance of membrane-related degradation mechanisms depends strongly on the duration and purpose of the experiment. In short-duration catalyst screening or mechanistic studies, changes in membrane resistance, hydration state, or product crossover may have a smaller effect on the overall result, but they should still be considered whenever they influence product quantification, material balance, or potential measurement. In long-term stability tests, membrane resistance, salt deposition, hydration state, electrolyte composition, and product crossover should be monitored explicitly because gradual changes in these variables can become major sources of performance drift and data uncertainty [4,5,7,8].

Third, does the membrane material match the intended operating conditions?

An ideal ion-exchange membrane should provide high ionic conductivity as well as chemical stability, mechanical robustness, and dimensional stability. For AEMs, in addition to OH⁻ conductivity, alkaline stability, ion-exchange capacity (IEC), water uptake, swelling behavior, and gas-blocking ability all affect practical CO₂RR performance. Thus, “AEM” or “PEM” is only a first-level classification; the specific brand, model, and pretreatment protocol must still be selected according to the experimental conditions.

Data comparability and experimental records

For studies that require cross-laboratory comparison or long-term data accumulation, simply stating that an “AEM” or “PEM” was used is not sufficient. Membrane model, thickness, pretreatment method, ion-form conversion procedure, soaking solution, and post-operation state can all affect membrane conductivity, cell voltage, and product selectivity. These details should therefore be fully reported in laboratory records and publications to improve reproducibility and comparability. Otherwise, even membranes of the same general type may lead to noticeably different results because of differences in pretreatment or hydration state.

2. Working Electrode Selection: More Than a Catalyst Support

In a GDE flow cell, the working electrode serves not only as an electron-conducting pathway, but also as the structural basis for forming the gas-liquid-solid three-phase interface. In a traditional H-cell, glassy carbon mainly functions as an inert conductive substrate. In a GDE system, however, porous conductive substrates such as carbon paper, carbon cloth, nickel foam, and other gas-permeable supports can directly affect CO₂ delivery, water management, local salt concentration, and current distribution [1,3,10]. By contrast, glassy carbon is a dense, nonporous substrate more commonly used for fundamental electrochemical studies and is not typically employed as a conventional GDE substrate.

A simple limiting-current expression helps illustrate why the working electrode matters:

Here, j_lim is the limiting current density, n is the number of transferred electrons, F is the Faraday constant, D is the diffusion coefficient, C is the reactant concentration, and δ is the diffusion-layer thickness. The core advantage of a GDE is that gas-phase CO₂ supply shortens the effective transport distance and increases local CO₂ availability [1,3]. Whether this advantage can be maintained depends on whether the electrode pores can deliver gas without becoming fully flooded by electrolyte.

Table 2. Common working-electrode substrates for CO₂RR

Note: Structural parameters are representative values for commonly used commercial substrates. Actual properties vary with manufacturer, substrate architecture, surface treatment, catalyst-layer integration, and cell-assembly conditions.

Different working-electrode substrates affect not only electron transport, but also gas diffusion, water management, and three-phase interface stability (Table 2). For most GDE flow-cell CO₂RR studies, carbon paper and carbon cloth remain the most common choices. Glassy carbon is better suited to fundamental electrochemical studies, while nickel foam is often used in strongly alkaline and medium-to-high-current-density systems. Importantly, the role of the working electrode has expanded from a simple catalyst support to a regulator of the local reaction environment; pore structure, wettability, and mechanical stability all influence the final experimental outcome [3,9,10].

Electrode design is shifting from “catalyst support” to “local reaction environment regulator”

In traditional electrochemical tests, the working electrode is often viewed as a substrate that carries the catalyst and conducts electrons. In GDE flow-cell CO₂RR, however, the working electrode itself has become an important engineering unit for regulating the local reaction environment. A complete GDE typically includes a gas diffusion layer (GDL), a microporous layer (MPL), and a catalyst layer (CL). Together, these structures influence CO₂ transport pathways, local ion concentration, wetting state, bubble detachment, and three-phase interface stability. Therefore, improved catalyst performance may arise not only from active sites, but also from electrode-structure optimization that enhances local CO₂ concentration, mass transport, and interfacial stability. When comparing catalyst results, researchers should report not only catalyst composition and loading, but also the GDL/MPL structure, electrode-substrate type, and electrode-preparation method; otherwise, data from different studies are difficult to compare directly [3,9,10,14].

Two common failure modes: flooding and salt precipitation

GDE instability often appears as increasing cell voltage, drifting product selectivity, enhanced HER, abnormal gas-line pressure, or local drying. One typical mechanism is flooding, in which electrolyte intrudes into pores that should remain available for gas transport, blocking the CO₂ transport pathway. Once CO₂ supply from the gas side becomes unstable, local CO₂ concentration decreases, HER becomes more competitive, and both current efficiency and stability suffer [3,10].

Another common issue is salt precipitation. CO₂RR experiments commonly use KHCO₃, KOH, or other salt-containing electrolytes. At high current densities, local pH, ion migration, and water flux change the local salt concentration. When salts deposit inside the GDE pores or at the membrane/electrode interface, both gas transport and ion conduction deteriorate, leading to increased voltage, local hot spots, interfacial delamination, and selectivity fluctuations [5,10].

Therefore, working-electrode selection should not be based only on whether the substrate is conductive or can hold catalyst ink. More practical questions include: What is the target current density? Is gas-phase CO₂ supply required? Is the electrolyte prone to salt precipitation? Will the test last 30 minutes or 100 hours? Does the setup need to work together with water-management systems, gas-liquid mixing pumps, mass-flow controllers, and pressure balancing in a GDE flow cell?

Figure 2. Typical failure mechanisms in gas diffusion electrodes (GDEs) during CO₂ electroreduction. (a) Flooding in a GDE. Reproduced with permission from Ref. [11]. (b) Salt precipitation in a flow-cell CO₂RR system. Reproduced with permission from Ref. [12].

As shown in Fig. 2, the two most common failure mechanisms during long-term GDE operation are flooding and salt precipitation. Flooding usually originates from gradual electrolyte intrusion into pores intended for gas transport, partially or completely blocking CO₂ diffusion pathways. This lowers the local CO₂ concentration and strengthens HER competition. Meanwhile, under medium-to-high-current-density conditions, OH^– generated at the cathode continuously reacts with CO₂ to form bicarbonate and carbonate species, which can further combine with cations such as K⁺ to form deposits. As salt crystals accumulate in the catalyst layer, microporous layer, and gas diffusion layer, both gas-transport and ion-transport resistances increase, eventually causing higher cell voltage, product selectivity fluctuations, and reduced long-term stability. Although flooding and salt precipitation arise through different pathways, both disrupt the stable three-phase interface inside the GDE and weaken CO₂ mass transport, making them key engineering challenges for durable medium-to-high-current-density CO₂RR [11,12].

3. Reference Electrode Selection: The Basis for Reliable Potential Measurement

One common misconception in CO₂RR is to treat the potential displayed by the workstation as the true potential at the catalyst surface. In reality, current passes through electrolyte, membrane, pores, and interfaces, producing uncompensated resistance (R_u). Reference-electrode position, salt-bridge structure, filling solution, and compatible pH range all affect the measured potential [13].

The potential drop associated with uncompensated resistance can be approximated as:

Here, ΔE_iR is the potential drop associated with uncompensated resistance, I is the total cell current, and R_u is the uncompensated resistance specifically between the working and reference electrodes. Unlike R_Ω, which denotes the overall ohmic resistance of the electrolyzer in the cell-voltage expression above, R_u refers specifically to the uncompensated resistance relevant to the working-electrode potential measurement. As operating current increases, even modest uncompensated resistance can produce a substantial potential error. Because the magnitude of the iR error depends on both the total current and R_u, and the total current scales with electrode area at a given current density, the iR-compensation method should be reported together with the potential data when comparing across different flow-cell configurations [13].

Table 3. Common reference electrodes used in CO₂RR

Note: Actual performance depends on electrolyte composition, cell architecture, and operating conditions.

Regardless of the reference electrode selected, reports should specify the reference type, filling solution, salt-bridge configuration, calibration method, potential-conversion equation, and iR-compensation approach. For cross-literature comparison, potentials should preferably be converted to reversible hydrogen electrode (RHE) or another clearly defined scale, and online or offline iR compensation should be stated. Otherwise, a claim such as “high selectivity at -0.8 V vs RHE” may mask differences in Ru and reference-electrode placement between experiments.

Figure 3. Schematic illustration of resistance gradients and reference-electrode positioning effects in a CO₂RR flow cell. Reproduced with permission from Ref. [13].

As shown in Fig. 3, under medium-to-high-current-density operation in flow cells, continuous gas evolution at the electrode surface changes the local electrolyte conductivity through bubble accumulation and flow-field distribution, creating nonuniform resistance along the channel. When the reference electrode is located farther from the working electrode, the measurement pathway includes a longer uncompensated resistance, and the measured potential may deviate from the true potential at the working-electrode surface. This error increases with current density. Therefore, in CO₂RR flow-cell experiments, the reference electrode should be placed as close as practical to the working electrode, while the Ru measurement method and iR-compensation protocol should be reported to improve data comparability [13].

4. Counter Electrode Selection: An Often Overlooked Variable

The counter electrode closes the electrochemical circuit and supports the anodic reaction, most commonly the oxygen evolution reaction (OER). Although CO₂RR studies primarily focus on the cathode, the counter electrode should not be assumed to be inert. Its catalytic activity, effective area, stability, and possible dissolution products can influence anodic polarization, overall cell voltage, and contamination of the cathode compartment. Pt-based counter electrodes are convenient for many short-duration laboratory experiments, but Pt dissolution and subsequent deposition on the cathode can become a concern in contamination-sensitive or extended tests. For longer-duration or higher-current operation, the counter electrode should therefore be selected on the basis of OER activity, chemical stability, effective area, and compatibility with the electrolyte and membrane [15,16].

Table 4. Common counter electrodes used in CO₂RR experiments

Note: Actual stability and contamination risks vary with operating conditions and cell configuration.

Counter-electrode selection can be summarized as follows: for short-term screening, prioritize stability, convenience, and low polarization; for long-term operation, prioritize material compatibility and contamination control; for medium-to-high-current-density tests, prioritize sufficient area and efficient bubble detachment; for mechanistic studies, prioritize minimizing potential interference from the anode material.

Figure 4. Possible counter-electrode contamination pathways in CO₂RR systems. Based on Refs. [15,16].

As shown in Fig. 4, counter-electrode materials may not be completely inert under certain conditions. Taking a Pt counter electrode as an example, metal species may dissolve at high anodic potentials or during long-term operation, migrate through the cell, and redeposit on the working-electrode surface, thereby changing the catalyst composition or electrochemical behavior. For CO₂RR studies aimed at evaluating intrinsic activity, even trace metal deposition can affect product selectivity, current response, and stability. Counter-electrode selection and potential contamination risks should therefore be explicitly recorded and evaluated.

It should be emphasized that Fig. 4 only illustrates potential pathways for counter-electrode contamination of the cathode and is not a quantitative result for any specific system. The actual degree of contamination depends on counter-electrode material, electrolyte composition, membrane type, current density, operation time, and reactor architecture. In long-term stability tests or studies sensitive to metal contamination, blank experiments, post-operation characterization, and elemental analysis such as ICP-MS, XPS, or SEM-EDS are recommended to determine whether the counter electrode interferes with the results and to improve reliability and reproducibility [15,16].

5. Practical Selection Matrix: From Catalyst Screening to Long-Term Stability

The following combinations can serve as starting points for CO₂RR GDE flow-cell experiments. Actual selection should still be adjusted according to target product, electrolyte pH, current density, reactor geometry, flow rate, pressure balance, and analytical method.

Table 5. Illustrative Starting Configurations for Membrane and Three-Electrode Selection in CO₂RR

Table 5 is intended as a decision-support framework rather than a universal prescription. The appropriate configuration depends on the target product, catalyst, electrolyte, current-density range, cell architecture, and duration of operation. A practical selection process should first define the desired reaction environment and operating regime, then evaluate membrane transport, GDE wettability, reference-electrode accuracy, counter-electrode compatibility, carbonate management, product crossover, and contamination risk. Membrane and electrode selection should therefore be treated as a system-level optimization problem rather than as a choice based on any single component.

6. Data Comparability Is Equally Important

In recent years, the CO₂RR community has placed increasing emphasis on standardized reporting. Reporting only Faradaic efficiency, current density, and product distribution is no longer sufficient for meaningful cross-literature comparison. To improve reproducibility, researchers should report membrane, electrode, and operating conditions as completely as possible. Many apparently contradictory catalyst-performance trends may originate from incomplete descriptions of experimental conditions rather than from the catalyst itself. High-quality CO₂RR research should therefore pursue not only higher FE or larger current density, but also experimental conditions that can be reproduced and compared [15,16].

CO₂RR results are not determined by the catalyst alone. They emerge from the combined effects of the catalyst, membrane, working electrode, reference electrode, counter electrode, fluid management, and reactor architecture. The membrane governs ion transport, local pH, product crossover, and cell voltage; the working electrode determines the three-phase interface, CO₂ supply, flooding, and salt-precipitation risks; the reference electrode determines whether the measured potential is reliable; and the counter electrode affects anodic polarization, material compatibility, and possible contamination.

Therefore, when building and optimizing a CO₂RR platform, the membrane and three-electrode system should be systematically recorded and controlled as key experimental parameters rather than treated as interchangeable auxiliary components. A high-quality CO₂RR data set should report at least the membrane model and pretreatment, working-electrode substrate and catalyst loading, reference-electrode type and calibration method, counter-electrode material, iR-compensation method, electrolyte composition, CO₂ flow rate, liquid flow rate, reactor geometry, and stability-test duration.

Once this information is systematically documented, differences among experiments become easier to interpret, and catalyst performance can be evaluated more fairly. In other words, reliable CO₂RR research is not only about achieving higher FE or larger current density; it is about establishing a reproducible, comparable, and diagnosable CO₂RR GDE platform.

As CO₂RR systems continue to move from laboratory demonstrations toward industrially relevant operation, careful control and reporting of membranes, GDE substrates, electrode configuration, flow conditions, and reference/counter-electrode setup will become increasingly important for generating reliable, reproducible, and transferable performance data.

References

[1] Burdyny, T.; Smith, W. A. CO₂ reduction on gas-diffusion electrodes and why catalytic performance must be assessed at commercially-relevant conditions. Energy & Environmental Science, 2019, 12, 1442–1453. https://doi.org/10.1039/C8EE03134G

[2] Lees, E. W.; Mowbray, B. A. W.; Parlane, F. G. L.; Berlinguette, C. P. Gas diffusion electrodes and membranes for CO₂ reduction electrolysers. Nature Reviews Materials, 2022, 7, 55–64. https://doi.org/10.1038/s41578-021-00356-2

[3] Hernandez-Aldave, S.; Andreoli, E. Fundamentals of Gas Diffusion Electrodes and Electrolysers for Carbon Dioxide Utilisation: Challenges and Opportunities. Catalysts, 2020, 10, 713. https://doi.org/10.3390/catal10060713

[4] Chang, H.-M.; Zenyuk, I. V. Membrane electrode assembly design to prevent CO₂ crossover in CO₂ reduction reaction electrolysis. Communications Chemistry, 2023, 6, 2. https://doi.org/10.1038/s42004-022-00806-0

[5] Yang, Y.; Shi, Y.; Yu, H.; Zeng, J.; Li, K.; Li, F. Mitigating carbonate formation in CO2 electrolysis. Next Energy, 2023, 1, 100030. https://doi.org/10.1016/j.nxener.2023.100030

[6] Petrov, K. V.; Koopman, C. I.; Subramanian, S.; Koper, M. T. M.; Burdyny, T.; Vermaas, D. A. Bipolar membranes for intrinsically stable and scalable CO₂ electrolysis. Nature Energy, 2024, 9, 932–938. https://doi.org/10.1038/s41560-024-01574-y

[7] Yan, J.; Wang, Y.; Xu, T. Ion leakage in bipolar membranes undermines CO₂ electrolysis. Nature Chemical Engineering, 2026, 3, 91–93. https://doi.org/10.1038/s44286-025-00325-4

[8] Liu, W.; Dunne, H.; Ballotta, B.; Massie, A. A.; Ghaani, M. R.; McKelvey, K.; Dooley, S. CO₂ Loss into Solution: An Experimental Investigation of CO₂ Electrolysis with a Membrane Electrode Assembly Cell. ACS Applied Energy Materials, 2024, 7, 7712–7723. https://doi.org/10.1021/acsaem.4c01101

[9] Higgins, D.; Hahn, C.; Xiang, C.; Jaramillo, T. F.; Weber, A. Z. Gas-Diffusion Electrodes for Carbon Dioxide Reduction: A New Paradigm. ACS Energy Letters, 2019, 4, 317–324. https://doi.org/10.1021/acsenergylett.8b02035

[10] Yang, K.; Kas, R.; Smith, W. A.; Burdyny, T. Role of the Carbon-Based Gas Diffusion Layer on Flooding in a Gas Diffusion Electrode Cell for Electrochemical CO₂ Reduction. ACS Energy Letters, 2021, 6, 33–40. https://doi.org/10.1021/acsenergylett.0c02184

[11] Wu, Y.; Rabiee, H.; Zhao, X. S.; Wang, G.; Jiang, Y. Insights into electrolyte flooding in flexible gas diffusion electrodes for CO₂ electrolysis: from mechanisms to effective mitigation strategies. Journal of Materials Chemistry A, 2024, 12, 14206–14228. https://doi.org/10.1039/D4TA01994F

[12] Chen, Y.; Wu, B.; Qian, L. Proton Pool for the Mitigation of Salt Precipitate Enhancing CO₂ Electroreduction in a Flow Cell. Catalysts, 2024, 14, 807. https://doi.org/10.3390/catal14110807

[13] Jännsch, Y.; Hämmerle, M.; Leung, J. J.; Simon, E.; Fleischer, M.; Moos, R. Gas evolution in electrochemical flow cell reactors induces resistance gradients with consequences for the positioning of the reference electrode. RSC Advances, 2021, 11, 28189–28197. https://doi.org/10.1039/D1RA05345K

[14] Rabiee, H.; Ge, L.; Zhang, X.; Hu, S.; Li, M.; Yuan, Z. Gas diffusion electrodes (GDEs) for electrochemical reduction of carbon dioxide, carbon monoxide, and dinitrogen to value-added products: a review. Energy & Environmental Science, 2021, 14, 1959–2008. https://doi.org/10.1039/D0EE03756G

[15] Seger, B.; Robert, M.; Jiao, F. Best practices for electrochemical reduction of carbon dioxide. Nature Sustainability, 2023, 6, 236–238. https://doi.org/10.1038/s41893-022-01034-z

[16] Pander, J. E., III; Ren, D.; Yeo, B. S. Practices for the collection and reporting of electrocatalytic performance and mechanistic information for the CO₂ reduction reaction. Catalysis Science & Technology, 2017, 7, 5820–5832. https://doi.org/10.1039/C7CY01785E