Development of high-performance zero-gap carbon dioxide electrolysis cells using a hydrophilic porous membrane

doi:10.21203/rs.3.rs-3250835/v1

Download PDF

Article

Development of high-performance zero-gap carbon dioxide electrolysis cells using a hydrophilic porous membrane

https://doi.org/10.21203/rs.3.rs-3250835/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

A hydrophilic porous membrane is adopted as the diaphragm in zero-gap CO₂ electrolysis cells in order to improve their efficiency and durability. The hydrophilic porous membrane is sparse and nonpolar, which are features opposite to those seen in the anion-exchange membranes commonly used as the diaphragm in CO₂ electrolysis cells. These characteristics result in high ion permeability, which contributes to decreased power consumption. A cell with an active area of 16 cm² achieves high Faradaic efficiency of 90% at a current density of 1000 mA/cm² and a lower cell voltage compared with anion-exchange membranes. Optimization of the operating temperature achieves maximal energy efficiency of around 50% at 200 mA/cm² under neutral electrolyte conditions. Excellent chemical stability of the porous membrane is demonstrated through experiments in a simulated practical environment and under long-term operation with application of a variable power source and introduction of impurity gases. The cells using the membrane was operated for an unprecedented 1000 h in a 400 mA/cm² durability test. Furthermore, a CO₂ electrolysis cell scaled up to a 10 × 100 cm² cell stack is demonstrated as a step toward the development of an industrial-scale CO₂ electrolysis cell. These results suggest that the hydrophilic porous membrane is a promising diaphragm option for the practical application of CO₂ electrolyzers.

Physical sciences/Materials science/Materials for energy and catalysis/Electrochemistry

Physical sciences/Energy science and technology

Reducing carbon dioxide (CO₂) emissions is common issue worldwide for realizing sustainable development. CO₂ capture and storage (CCS) technology has potential for large-scale implementation.¹ Aside from the option of storing captured CO₂ deep in the ground, CO₂ capture and utilization (CCU) is important as well. A CCU system with CO₂ electrolysis cells at its center, which we call “Power to Chemicals”, is one of the most promising options.² In such a system, CO₂ captured from the atmosphere or industrial facilities that emit large amounts of CO₂ (e.g., factories and thermal power plants) is converted into value-added compounds by using renewable energy. The products of the CO₂ reduction reaction (CO2RR) varies depending on the catalyst and applied electrochemical potential.^3–5 One of the most promising CO2RR products is carbon monoxide (CO) due to its chemical value and the high energy efficiency of the reaction.⁵ CO can be converted to hydrocarbons or alcohols by further thermo- or electrocatalytic reactions.^6,7 We have set four axes for the development of CO₂ electrolysis cells: high current density, scale-up (both large electrode area and high cell stacking), high efficiency, and long-term durability. All elements contribute to reducing CO generation costs and realizing the practical application of CO₂ electrolysis cells. The energy efficiency of the CO2RR for CO synthesis (EE_CO) is described as follows: EE_CO = FE_CO × V° / V_cell, where FE_CO is Faradaic efficiency for the CO evolution reaction (COER), V° is the ideal reaction voltage (1.33 V at 25°C for COER), and V_cell is the applied cell voltage. A basic setup for electrochemical study is a liquid-type cell with a cathode solution chamber and an anode solution chamber separated by a diaphragm. However, the CO2RR rate is inherently hampered by the low solubility of CO₂ in aqueous solution (33 mM) and high resistance due to long cathode-anode distance (several millimeters), which leads to large power consumption. Flow-type and zero-gap cells are more suitable structures to study for practical use. The flow-type structure contains both catholyte and anolyte separated by a diaphragm.^8–11 The zero-gap structure omits the catholyte from the flow-type structure and is similar to a polymer electrolyte membrane fuel cell.^12–15 A membrane-electrodes assembly (MEA) comprising a cathode, diaphragm, and anode in the zero-gap cell achieves low V_cell due to the short distance between the cathode and anode equal to the thickness of the diaphragm (several hundred micrometers). Recently, we successfully demonstrated a zero-gap CO₂ electrolysis cell using an anion-exchange membrane (AEM), achieving high FE_CO of 93% at a high current density of 700 mA/cm² under near-neutral conditions.¹⁶ Both high FE_CO and current density could be achieved by the highly porous catalyst layer, which forms a three-phase interface leading to favorable supply of CO₂ and water into the catalytic reactors. Because the hydrogen-evolution reaction (HER) competes with the CO2RR, the concentration of protons in the solution (i.e., the pH) is an important factor governing the reaction. The cathodic and anodic half-reactions of the CO2RR should proceed under alkaline conditions to achieve high FE_CO, whereas the HER is favored under lower pH conditions (higher proton concentration). Recently, remarkably advances have been made in the development of low-temperature CO₂ electrolysis cells by using AEMs as the diaphragm with an alkaline electrolyte.^12,14,17 These membranes are designed for low-temperature CO₂ electrolysis cells and can be purchased from several manufacturers. These specially designed membranes are thin and have moderate water uptake to achieve two key features of a diaphragm simultaneously: ion conductivity and gas/liquid barrier function.¹⁷ The high ion conductivity rapidly transfers reactive ions for the CO2RR and the high gas/liquid barrier function prevents loss of product gases and excess water migration (i.e., flooding of the cell). In addition, some durability issues have been noted with the AEMs, particularly degradation of ion-exchange capacity due to attack of hydroxyl ions.^18,19 Furthermore, recent research indicates that the cation present may play an important role in the CO2RR, although AEMs enable selective anion exchange.^20,21

To achieve higher EE_CO and long-term stability, we have explored a novel candidate for the diaphragm instead of AEMs, finally focusing on hydrophilic porous membranes (HPMs). HPMs have chemical and physical properties different from those of AEMs, enabling ion transfers with different types and amounts of ions (Fig. 1). AEMs are dense membranes containing cation groups, so almost no water permeates them and anions are selectively transferred by a hopping mechanisms including the well-known Grotthuss mechanism.¹⁷ By contrast, HPMs are low-density membranes with many pores and no ion-exchange groups and ion transport can occur only via vehicular transport with water permeation. Not only the anions required for the reaction, but also the cations that assist the reaction, can move freely and rapidly. The fast, nonpolar ion conduction by electrolyte filling pores is not decreased by degradation of ion-exchange capacity as often occurs with AEMs.^18,19 Although the porousness conflicts with gas/liquid barrier function, product gas crossover and flooding can be prevented by applying pressure and selecting the membrane material, thickness, and pore diameter. Moreover, the HPMs can tolerate long-term operation and large-scale CO₂ electrolysis cells because the rigid molecular structures of HPMs have highly thermal resistivity and mechanical strength compared with those of AEMs. In this paper, we report fundamental studies using a cell with an active area of 16 cm² (below, the 16-cm² cell) and compare use of an HPM and ion-exchange membranes under high current density and variable temperatures. Experiments using a variable power source and introducing impurity gases related to practical use in an industrial environment were also conducted. Unprecedented operation for 1000 h in a durability test at 400 mA/cm² was achieved with little degradation of efficiency in the 16-cm² cell. Furthermore, we demonstrated a CO₂ electrolysis cell stack contain ten 100-cm² cells. The cell stack has attained 500 h of operation with a high EE_CO comparable to that of the 16-cm² cell.

Comparison of CO₂ electrolysis cells using HPM and ion-exchange membranes: high-current-density measurements

Many commercial porous membranes ranging from organic polymers to ceramics with varying hydrophilicity, thickness, and pore diameter were tested as the diaphragm of CO₂ electrolysis cells. These included hydrophilized polytetrafluoroethylene, cellulose, and nylon polymers that are used for membrane filtration and even ceramics that are often used as a separator in electrochemical reactions.²² Although moderate hydrophilicity was an essential property for transporting reactive ions through the electrolyte, chemically hydrophilized membranes deteriorated during operation because of desorption of the applied treatment material. Completely hydrophilic membranes, on the other hand, caused excessive electrolyte migration, specifically flooding. The amount of ion/water permeation depends not only on the hydrophilicity, but also on the thickness and pore diameter of the membrane. Most porous membranes used for filtration and electrochemical reactions tend to have high water permeability, and membranes which is thinner and have smaller pore are desirable. As a result of many tests for HPMs, we concluded that a polyethersulfone membrane were suitable as a diaphragm in CO₂ electrolysis cells. This chemical backbone showed moderate hydrophilicity, heat resistivity, and chemical resistivity. Notably, since there was no water uptake in the HPMs, assembly of the cell was easy and it was resistant to dryness.

The ion conductivity of a diaphragm influences V_cell and EE_CO of a CO₂ electrolysis cell. We investigated a diffusion coefficient from Fick’s laws of diffusion by determining the amount of potassium ions transferred per hour. The HPM showed a diffusion coefficient 7-fold that of an AEM for potassium ions. Although the dense cationic groups of an AEM interrupt cations, the nonpolar HPM allowed both cations and anions to cross. Chronopotentiometric measurements with current density up to 1000 mA/cm² were conducted for 16-cm² cells using various membranes. The present cell structure is shown in Fig. 2(a). Reactant gas (> 99.995% pure CO₂ gas) and 0.1 M potassium hydrogen carbonate (KHCO₃) electrolyte were continuous flowed in the cathode and anode flow units, respectively. To prevent flooding of CO₂ electrolysis cells using the HPM, the cathode pressure should be more positive than the anode pressure. However, an excessively large positive pressure in the cathode was not favorable for CO₂ electrolysis cells because it could damage the membrane, cause crossover of product gases, and decrease of water and ion transfer. The applied pressures were optimized to 200 kPaG and 100 kPaG for the cathode and anode inlets, respectively, considering the pressure drop across the HPM and the water repellency of the cathode. For comparison, measurements were also performed using an AEM (PiperION® membrane) and a cation-exchange membrane (CEM; Nafion® 211 membrane).¹² The measurement temperature was 50°C and no pressure was applied when using the AEM and CEM.

We previously reported high FE_CO > 90% when using a 4-cm² cell with an AEM.¹⁶ Fig. 2(b) shows that FE_CO > 90% was maintained up to 1000 mA/cm² in the present 16-cm² cell when using the HPM, if pressurized. FE_CO reached a maximum of around 600 mA/cm² and slightly decreased after that. When the CO₂ electrolysis cell using HPM was operated without applying pressure, a precipitous decrease in FE_CO rapidly occurred. This decrease was caused by accelerated flooding due to the pressure becoming more positive on the anode side compared with the cathode side as a result of oxygen evolution, especially at high current density. The CO₂ electrolysis cell using the AEM showed almost the same FE_CO behavior as the CO₂ electrolysis cell using the HPM. In the case of the CEM, however, the HER became dominant with increasing current density. The CO₂ electrolysis cell using the HPM showed clearly lower V_cell compared with cells using the AEM and CEM over the whole current density range in Fig. 2(c). The low V_cell critically contributed to high efficiency as shown in Fig. 2(d). The low V_cell can be attributed to the higher ion diffusion coefficient, which decreases cathode overpotential rather than ohmic losses, as discussed in the next section. There was large difference when using the HPM with and without pressure application, not only for FE_CO but also for V_cell. Increased anode pressure due to promotion of the oxygen evolution reaction at high current density caused flooding in the cathode, leading to both a decrease in FE_CO and an increase in V_cell. Given that both the cell with the HPM and the cell with the AEM showed almost the same behavior with FE_CO maintained above 90%, EE_CO strongly depended on V_cell. The CO₂ electrolysis cell using the HPM with pressurization achieved the highest EE_CO over the entire current density range and EE_CO decreased with increasing current density. EE_CO of 48.8% at 200 mA/cm² and 33.5% at 1000 mA/cm² were very high values among CO₂ electrolyzers using a near-neutral electrolyte. Comparison of EE_CO at 200 mA/cm² is shown in Fig. 2(e) with error bars. Reproducibility was verified through three repeated measurements, which showed error not exceeding 3%. EE_CO when using the HPM became 4% larger than that when using the AEM due to the lower value of V_cell.

Comparison of CO₂ electrolysis cells using an HPM and AEM: temperature dependence

We conducted evaluations using a 16-cm² cell at various cell temperatures under constant current density of 200 mA/cm² to reveal the temperature at which the maximum EE_CO was attained for the HPM and AEM. For the cell using the HPM, measurements were only performed with pressurization. Figure 3(a) shows the temperature dependence of Faradaic efficiency for COER, HER, and their total. FE_CO decreased with increasing temperature whereas FE_H2 increased. A slight decline of FE_Total was observed at high temperatures, suggesting that product gases exited from the anode side. The decrease in FE_Total was larger for the cell using the AEM compared with the HPM, indicating that this result could be attributed to the differing thermal stability of the diaphragms. The thermal stability of the HPM would be favorable for large-scale CO₂ electrolyzers. Prevention for crossover of product gases by selecting appropriate materials (pore size) and experimental conditions attained FE_Total of around 95% with the HPM, a value comparable to that with the AEM. V_cell can be decomposed into several voltage components; ideal reaction potentials (E^°_C for cathode and E^°_A for anode), overpotentials associated to transportation of reactive ions and electrodes kinetics (η_C for cathode and η_A for anode), and IR losses as follows E = − E^°_C + E^°_A − η_C + η_A + IR.^23,24 The temperature dependence of V_cell in Fig. 3(b) is displayed with an absolute value of pseudo-cathode overpotential (|η_C|) and IR losses in order to reveal parameters affected by different diaphragms. |η_C| and IR losses were evaluated by using a reference electrode and an AC resistance meter. V_cell of the CO₂ electrolysis cell using the HPM was constantly about 0.2–0.3 V lower than that of the AEM. This difference arose from |η_C| rather than η_A and IR. This suggested that improved ion permeability reduced |η_C| as a result of the effect on the CO2RR. Although the decrease in the AC resistance component was small because of the zero-gap cell structure, low resistance of ~ 30 mΩ was basically achieved, and effect is expected to increase with increasing applied current due to scale-up. As a result of the balance between decreased FE_CO and V_cell, EE_CO showed a gradual shape with a peak around 50°C as shown in Fig. 3(c). EE_CO of the CO₂ electrolysis cell using the HPM was about 5% higher than that of the AEM over the entire temperature range owing to the low V_cell. Maximum EE_CO was obtained at 50°C for both diaphragms: 48.8% for the CO₂ electrolysis cell using the HPM and 44.8% for the one using the AEM.

Performance of HPM for simulated practical environments

In contrast to the experiments using lab-scale stabilized DC power sources and pure CO₂ inlet gas as presented above, operation in practical environments with renewable energy power sources and recovery of CO₂ from power plants would have a negative impact on operation of the CO₂ electrolysis cells. Renewable energy from solar, wind, and other sources includes both long-period and instantaneous fluctuations.²⁴ CO₂ electrolysis cells acting as a regulator for renewable energy must be responsive to the instantaneous fluctuations. Recovered CO₂ is contaminated with N₂, O₂, H₂O, CO, SO_x, NO_x, and small amounts of volatile organic compounds.²⁵ In particular, O₂, SO_x, and NO_x are known to inhibit the CO2RR by being reduced themselves and degrading MEAs.^25–27 To investigate the impact of these environmental factors on CO₂ electrolysis cells, we assumed periodic variable power sources and inlet CO₂ gases including small amount of SO₂ and NO₂ to simulate a practical environment. The effects were evaluated during long-term operation. First, the responsiveness to current fluctuation of the CO₂ electrolysis cell using the HPM and AEM was evaluated by a following procedure based a start-up protocol: Constant current density initially held at 100 mA/cm² and then increased to 200 mA/cm² over 10 s. The applied 200 mA/cm² was held until V_cell stabilized as shown in Fig. 4(a). Immediately after increasing the current density, V_cell reached a maximum of 2.85 V rapidly when using the CO₂ electrolysis cell with the HPM and a maximum of 3.05 V when using the cell with the AEM. V_cell of the cell with the HPM stabilized at 2.62 V within 300 s but the cell using the AEM looked needed about 1200 s to stabilize at 2.90 V. We defined the relaxation time as the time to reach the inverse of Napier’s constant (1/e) for the voltage difference between the top and bottom. The relaxation time was 35 and 106 s for the cells using the HPM and AEM, respectively, indicating that rapid startup could be possible when using the HPM. This good response of the CO₂ electrolysis cell with the HPM was attributed to its high ion permeability. Further research into the good responsiveness of the cell using the HPM was confirmed by a current-controlled triangle wave applied to 16-cm² cells to clarify the impact of instantaneous power source fluctuations. The triangle wave was varied from 0 mA/cm² to 200 mA/cm² with a 40-s period over 1000 h (90000 cycles) for the CO₂ electrolysis cell using the HPM (Fig. 4(b-d)). Figure 4(b) shows the time evolution of the partial current density for the COER (J_CO), which was averaged over 10 h because gas chromatography analysis was not responsive to fluctuation of the power source. Since the average current density of the applied triangle wave was 100 mA/cm², J_CO in Fig. 4(b) should be 100 mA/cm² for ideal CO generation. Indeed, J_CO of the CO₂ electrolysis cell using the HPM was nearly 100 mA/cm² and the cell achieved 1000 h of operation under the triangle wave source with no degradation in J_CO. Figure 4(c) shows the time evolution of V_cell over 1000 h, and Fig. 4(d) shows an enlarged portion of Fig. 4(c). V_cell varied according to the period of the triangle wave. The fluctuation of V_cell between 1.8 V and 2.8 V for the CO₂ electrolysis cell using the HPM did not change even after 1000 h. The fluctuation of V_cell can be clearly seen in the enlarged portion of V_cell after 100 h in Fig. 4(d). Delays in the response between the applied current density and V_cell were not seen on the present timescale (recorded at 1-s intervals). Thus, we confirmed that the CO₂ electrolysis cell with the HPM can operate stably with fluctuating power supply for a long period of time and can therefore be a regulating force for renewable energy.

Next, a CO₂ electrolysis cell was operated when the CO₂ inlet gas contained SO₂ and NO₂ in order to understand the impact of contamination. CO₂ gas mixed with 15 ppm SO₂ and 30 ppm NO₂ to simulate post-combustion gases captured by amine scrubbing was supplied as the inlet gas after passing through a humidifying water tank.²⁸ Results of a durability test for 450 h at constant current density of 200 mA/cm² are shown in Fig. 5. FE_CO of 98% and FE_H2 of 2% were attained and then maintained during 450 h of operation of the CO₂ electrolysis cell using the HPM. The results were comparable to the test with pure CO₂ (see Supporting Information). Also, FE_Total (i.e., the sum of FE_CO and FE_H2) was almost 100%, implying that no other electrochemical reduction reactions involving SO₂ and NO₂ occurred. The cell using the AEM also showed FE_CO of 98% and FE_H2 of 2% initially, but FE_CO gradually decreased and FE_H2 increased. The electrolyte was replaced at 204 h (blue triangle) and both the AEM and electrolyte were replaced at 300 h (green diamond) in order to identify the cause of the degradation. Although FE_CO was not affected by replacing only the electrolyte, FE_CO was improved from 91–98% by replacing both the AEM and electrolyte. Although this fully restored FE_CO to its initial level, the decline recurred. Since pure CO₂ operation for more than 450 h had been conformed without degradation, the degradation of FE_CO was attributed to the impurity gases affecting the AEM. For the AEM, FE_CO decreased, but there was no drop in FE_Total. Many reports about the impact of impurity gases on the CO2RR have shown a decrease in CO₂ reduction products due to the reduction of impurity gases in side reactions.^25–27 Here, the concentration of SO₂ and NO₂ in the simulated post-combustion gases captured by amine scrubbing was relatively low. Moreover, it has been reported that the decrease in Faradaic efficiency varies depending on the type of catalyst. It is possible that, instead of the catalysts for which the effects of impurities were previously examined (Ag, Cu, Sn, etc.), a gold catalyst may be more favorable for promoting the COER over the reduction of impurity gases.^25–27 The time evolution of V_cell for the CO₂ electrolysis cell using the HPM is shown in Fig. 5. The initial V_cell of 2.60 V increased to 2.82 V over 450 h. The observed logarithmic-like behavior, namely, a large initial increase followed by gradual increase, was uniformly obtained in CO₂ electrolysis cells, the same as in the durability test with pure CO₂. Given that the results were comparable with both pure and contaminated CO₂ gas, the contamination with SO₂ and NO₂ had no effect on the HPM during the 450 h of operation. In contrast to the logarithmic-like behavior observed for the HPM, the initial V_cell was 2.82 V in the CO₂ electrolysis cells using the AEM and it rapidly increased, with V_cell reaching 3.45 V during the first 200 h. V_cell did not fully recover despite replacement of the electrolyte at 204 h, it further increased from 3.24 V to 4.53 V during next 100 h. Drastic recovery of V_cell to 3.03 V resulted from replacement of both the AEM and electrolyte at 300 h. Thus, SO₂ and NO₂ affected the AEM during the durability test with impurity gases, resulting in the degradation of Faradaic efficiency and V_cell. These recovered almost to their initial values after replacing the AEM. One possible cause of the poor performance due to AEM degradation was that the potassium ion concentration in the electrolyte significantly decreased (Fig. 5(d)). We measured dependence of FE_CO and V_cell on electrolyte concentration, and found that V_cell increased with dilution of the electrolyte (Figure S4). Although significant performance degradation due to lower potassium ion concentration in the electrolyte was observed in the cells using the AEM, no such performance degradation was observed in the cells using the HPM. The rate of decrease in potassium ion concentration (slope of the plot in Fig. 5(d)) was more pronounced for longer operating times and was somewhat improved by replacing the AEM. To reveal the possible mechanisms by which the introduction of SO₂ and NO₂ caused a decrease in potassium ion concentration, ion-exchange capacity measurement and compositional analysis of the AEM were performed before and after the test with impurity gases. The results showed a decrease in the ion-exchange capacity and thickness of the AEM after the test with impurity gases (see Supporting Information). These changes were expected to increase permeation of water, accompanied with potassium ions, across the AEM from the anode to the cathode. SO₂ and NO₂ reacted aggressively with the AEM and degraded the diaphragm, whereas the more chemically stable HPM was not affected by the impurity gases.

Durability tests with high current density and a large-scale CO₂ electrolyzer using the HPM

CO production rate is an important factor in CO₂ electrolysis cells. A CO₂ electrolysis cell with a high CO production rate could process a large amount of CO₂ in a short time and with a small footprint, which would be economically advantageous. The CO production rate is determined by the product of the total electrode area and current density. We performed a durability test with high current density, and a large-scale CO₂ electrolysis cell was used to consider both increasing the electrode area and stacking using the HPM. Representative fault modes of the CO₂ electrolysis cells are known to be an increase in the HER due to catalyst failure, flooding, and blockage of the CO₂ flow path due to salt precipitation.²⁹ Because these fault modes are accelerated at higher current densities, it is generally difficult to achieve both high current density and long-term durability. In particular, there have been few reports of long-term testing at high current densities of 400 mA/cm² or higher. The results of a durability test on a 16-cm² cell using the HPM under application of 400 mA/cm² is shown in Fig. 6(a). Generally, it is necessary to humidify the inlet CO₂ gas to eliminate the precipitation of salts during long-term use of AEMs. In this test, however, the HPM was successfully operated over 1000 h without humidification. The high ion permeability of the HPM quickly transported carbonate ions to the anodes and prevented the generation of salt. FE_CO and FE_H2 were maintained at around 90% and 2% even after 1000 h of operation. V_cell and EE_CO are displayed in Fig. 6(b). The initial V_cell of 2.70 V rapidly increased to 2.83 V during first 100 h, and then increased slowly to 2.95 V up to 1000 h. The logarithmic-like behavior of the increase in V_cell was similar to the previous measurement shown in Fig. 5(c) and Figure S4. EE_CO was 43% initially and decreased to 41% with increasing V_cell.

To increase the amount of CO production per hour, the CO₂ electrolysis cell must be scaled up in addition to increasing the current density. Both increasing the size of the active area and stacking are established strategies for pilot-scale equipment.^13,15,30 We have developed a large-scale CO₂ electrolysis cell by these two strategies. A comparison between a small cell (16-cm² cell) and a scaled-up cell (10 × 100 cm² cell stack) is shown in Fig. 6(c). The cell stack contains 10 cathode/anode pairs, each with an active area of 100 cm². Thus, the total electrode area was increased from 16 cm² to 1000 cm². Each cathode and anode were supplied with a CO₂ gas feed and electrolyte in parallel and power was supplied in series. The cell stack was equipped with coolant channels between anode and cathode flow units to avoid excessive temperature increases. The detailed stacking structure is explained in the Supporting Information. Figure 6(d) shows V_cell and temperature for each cell under operation at 200 mA/cm² operation. A small amount of cooling water was introduced into the cell stack to keep the temperature around 50°C, which is where the 16-cm² cells performed with the highest efficiency as shown in Fig. 3(c). The temperature varied from 50°C at the ends of the cell stack to 61°C at the center of the cell stack. V_cell showed an inverse trend with respect to temperature, increasing from 2.55 V at the center to 2.61 V at the ends, comparable to V_cell of the 16-cm² cell. Such small variations in V_cell and temperature are desirable for long-term operation so that the degradation rate of each cell is uniform. Following a preliminary check, a durability test was performed at 200 mA/cm² for 500 h as shown in Fig. 6(e) and (f). In contrast to the tests of the 16-cm² cells, degradation of FE_CO due to salt precipitation was observed without CO₂ humidification. In the large-scale test, a slight increase in FE_H2 due to salt precipitation was observed when the system was operated at a dew point of around 50°C. Increasing the dew point to around 60°C eliminated the HER at 130 h. FE_CO attained a value of 98%, with no decrease observed during 500 h of operation. The overall cell voltage (V_stack) increased from 26.4 V initially to 31.0 V at 500 h. In addition to the increase in V_stack over time, there was a jump at 280 h due to increased cooling of the cell as a result of equipment failure. EE_CO was 48% initially and decreased to 40% with increasing V_stack. The initial EE_CO of around 48% for the 10 ×100 cm² cell stack was comparable to that for the 16-cm² cell (Fig. 3(c)) despite the scale-up. The chemical and mechanical durability of the HPM resulted in excellent scalability. Degradation of FE_CO did not occur even in the cell stack, meaning that the increase in V_stack was directly related to the increase in EE_CO. The large increase in V_stack particular to this cell stack was caused by migration of humidifying water from the cathode to the anode. A decrease in potassium ion concentration was confirmed to be due to electrolyte dilution by the migrated water, not degradation of the diaphragm (see Supporting Information). Although humidification was essential to prevent salt precipitation, excess humidifying water accelerated deterioration. Water management to deal with such issues will be important for high-efficiency operation over long periods of time.

Graphs comparing the present results with previous reports of long-term and industrial-scale current density operation are shown in Fig. 6(g) and (h).^{8, 10–15,29} Figures were plots of the operating time and energy efficiency versus the total current, which was the product of the electrode active area, number of stacked cells, and current density, and equals the CO generation rate. Whole previous reports used AEMs as the diaphragm of the CO₂ electrolysis cell. For high-current-density operation at over 400 mA/cm² (filled symbol), the present 16-cm² cell achieved continuous operation for 1000 h and high efficiency, whereas previous results achieved operation for only 100 h. For the present cell stack (diamond), we demonstrated CO₂ electrolysis cells with the fastest CO generation rate, the longest durability, and the highest energy efficiency when using the HPM. Although higher EE_CO has been reported for liquid-flow small cells using high-alkalinity electrolyte, EE_CO of around 50% at 200 mA/cm² in a large cell stack using neutral electrolyte was a notable result of this study in terms of practical use.⁹

Adopting the HPM as a diaphragm improved the durability and efficiency of CO₂ electrolysis cells. Pressure regulation prevented crossover of the produced gas and flooding, allowing operation up to 1000 mA/cm². The HPM had higher ion permeability as well as higher chemical and mechanical stability compared with the AEMs commonly used as a diaphragm, owing to the pores and nonpolar features of the HPM. The high ion permeability lowered V_cell and prevented salt precipitation. Almost no degradation was observed in tests of the HPM that simulated the practical environmental conditions, such as variable power sources and inlet gases containing SO₂ and NO₂ impurities. Unprecedented operation for 1000 h at 400 mA/cm² using the 16-cm² cell and for 500 h at 200 mA/cm² using the 10 × 100 cm² cell stack were attained. EE_CO of the cell stack was ~ 48% at 200 mA/cm², which was comparable to that of the 16-cm² cell indicating good scalability, and this was in the optimal range for practical cells. Electrolyte dilution was found to be the main cause of efficiency loss in stacked operation. In addition to the use of chemically/physically stable diaphragms, water management was essential to prevent salt precipitation and excessive migration of humidifying water in order to achieve stable operation over a long period of time. Advantages of the HPM include not only efficiency and durability but also ease of handling and low cost, and thus the HPM is a promising option for the future development of electrolyzers.

Electrodes and diaphragm

The cathode was fabricated by spray coating of catalyst ink in the same manner as previously reported.¹⁶ The ink was prepared by mixing of a commercial Au loaded carbon powder, 5wt% Nafion^® solution, water and 2-propyl alcohol with ultrasonic for 20 min. The Au particles act as the catalyst of CO2RR have approximately 2 nm diameter in average.¹⁶ The homogeneous solution ink was supplied to an air atomizer spray nozzle with 0.6 mL/min and coated on a gas diffusion electrode (AvCarb MB30) on 120°C heater over a geometric area of 20 × 20 cm². the coated electrode was washed with water and 3% hydrogen peroxide aqueous solution to eliminate organic residues, dried at room temperature, and divided into an actual cell size. The anode was a commercial Iridium Oxide coated Ti mesh. The HPM was adopted a commercial polyethersulfone membrane. The AEM was a PTFE supported PiperION^® membrane which was immersed in 1 M potassium hydroxide (KOH) solution over 1 h prior to use. The CEM was a Nafion^® 211 membrane which was immersed in pure water over 1 hours prior to use.

Ion permeability measurement

Ion permeability for the HPM and AEM were determined based on Fick's laws of diffusion. An H-type cell in which pure water and 1 M KOH solution were separated by a membrane was maintained at a constant temperature, and the potassium ion concentration in the pure water was measured at time intervals. According to Fick’s laws of diffusion, flux J was defined by J = -D × (dC/dx), where D is a diffusion coefficient and (dC/dx) is a concentration change per distance. Above equation could be vary to D/d = J/-ΔC, where d was a thickness of membrane and -ΔC was a concentration change over time. D/d was a permeability coefficient which can be comparable between membranes.

Electrolyzer structure

Present CO₂ electrolysis cell had a zero-gap structure as shown in Fig. 1(a). An MEA was composed of the cathode, the diaphragm and the anode with optimal thickness gaskets. The MEA was sandwiched between a flow unit, a current collector, a temperature control unit, an insulating plate and an outer plate from both sides. Reactant (99.995% pure) CO₂ gas and 0.1 M KHCO₃ electrolyte induced into a cathode and an anode flow unit, respectively. A Power supply and AC resistivity meter were connected to current collectors for electrochemical operation. The temperature control unit was composed of a heater and coolant channel in a copper block to optimize cell temperature. The outer plate was tightened to a defined torque of 8 Nm for the cell using the HPM and 4 Nm for the cell using the AEM and CEM. A 10 × 100 cm² cell stack had same concept. The cell-stacking structure displayed and explained in Supporting Information.

General method for CO₂ reduction reaction of CO₂ electrolysis cell.

Experimental setup for CO2RR of present CO₂ electrolysis cell was displayed in Supporting Information. Electrolysis parameters such as a flow rate, pressure, temperature, applied current and so on were varied according to the cell size and purpose of experiments.

Current density dependence measurement upon 1000 mA/cm²

The current density dependence measurements were conducted with the 16-cm² cell. CO₂ and electrolyte flow rate were 320 sccm and 10 mL/min. For the measurement using the HPM, 200 and 100 kPaG were applied for the cathode and the anode inlet, respectively. After the cell temperature reached 50°C, constant current density was applied. The current density was varied with 12.5, 50, 100 to 1000 each 100 mA/cm² with 10 min interval. Generated gas was quantified by gas chromatography each 6 min after applied current change. Reproducibility was checked by repeated measurements three times.

Temperature dependence measurement

The temperature dependence measurements were conducted with the 16-cm² cell. CO₂ and electrolyte flow rate were 320 sccm and 10 mL/min. For the measurement using the HPM, 200 and 100 kPaG were applied for the cathode and the anode inlet, respectively. After the cell temperature set to 30°C, constant current density of 200 mA/cm² was applied. The temperature was varied with 30, 50, 70 and 90°C with 30 min interval. Generated gas was quantified by gas chromatography each 6, 16, 26 min after temperature change. After the 90°C measurement, waited until the cell temperature fell below 30°C, and then performed the measurement at 30°C again to check the reproducibility.

Variable power source measurement

The variable power source measurements were conducted with the 16-cm² cell. CO₂ and electrolyte flow rate were 40 sccm and 10 mL/min. CO₂ humidification with dew point of 40°C was conducted for the measurement using the AEM. For the HPM measurement, 200 and 100 kPaG were applied for the cathode and the anode inlet, respectively. The cell temperature set to 60°C for the HPM and 40°C for the AEM. For the experiment about responsiveness of CO₂ electrolysis cells as shown in Fig. 3(a), constant current density 100 mA/cm² were held in advance. After stabilization of the cell voltage was checked, the current density increased to 200 mA/cm² within 10 s and the 200 mA/cm² was maintained. For the variable current density measurement, the programmed variable current was supplied by a potentiostat. Applied triangle waves varied 0 mA/cm² to 200 mA/cm² with 40 s period as shown in Fig. 3(d). Generated gas was quantified by gas chromatography every 10 min and the cell voltage was recorded every 1 s. Because the gas quantification took 4 min and did not follow the current fluctuation, the partial current density of CO was obtained as an average value (Fig. 3(b)).

Impurity gases measurement

The impurity gases measurements were conducted with the 16-cm² cell. SO₂ and NO₂ gases mixed with CO₂ were prepared in a gas mixer to 15 ppm and 30 ppm, respectively, and fed into the CO₂ electrolysis cell. Because SO₂ and NO₂ could corrode metals, pipes through which cathode gas flowed were changed to resin or titanium for the experiment. CO₂ and electrolyte flow rate were 80 sccm and 10 mL/min. CO₂ humidification with dew point of 40°C was conducted. For the measurement using the HPM, 100 kPaG was applied for the cathode inlet and no pressure was applied for the anode inlet. The cell temperature set to 50°C for the HPM and 40°C for the AEM. Constant current density 200 mA/cm² was applied and the generated gas was quantified by gas chromatography every 10 min. The electrolyte was replaced at 204 h and both AEM and electrolyte were replaced at 300 h for the measurement using the AEM. The potassium ion concentration was measured by a potassium ion meter (HORIBA, K-11).

Durability test at 400 mA/cm²

The durability test with a high current density of 400 mA/cm² was conducted with the 16-cm² cell. CO₂ and electrolyte flow rate were 320 sccm and 10 mL/min. 200 and 100 kPaG were applied for the cathode and the anode inlet, respectively. After the cell temperature set to 60°C, constant current density 400 mA/cm² was applied. Generated gas was quantified by the gas chromatography every 30 min.

Durability test for the large-scale cell

the durability test using the 10 × 100 cm² cell stack was conducted. CO₂ and electrolyte flow rate were 5000 sccm and 100 mL/min to each manifold inlet. CO₂ humidification with dew point of 50°C was conducted, and the dew point increased to around 60°C at 130 h. 50 kPaG was applied for the cathode manifold inlet and no pressure was applied for the anode manifold inlet. After constant current density 200 mA/cm² was applied, temperature was increased by heat generation itself. The cell temperature set to 50°C by flowing a small amount of cooling water through the coolant channel. The cell voltages and temperatures for each cell were recorded and generated gas was quantified by gas chromatography every 30 min.

Acknowledgements

Part of this work was commissioned by Ministry of the Environment, Government of Japan.

Bui, M. et al. Carbon capture and storage (CCS): the way forward. Energy Environ. Sci. 11, 1062–1176 (2018).
Masel, R. I. et al. An industrial perspective on catalysts for low-temperature CO2 electrolysis. Nat. Nanotechnol. 16, 118–128 (2021).
García de Arquer, F. P. et al. CO2 electrolysis to multicarbon products at activities greater than 1 A cm−2. Science 367, 661–666 (2020).
Tamura, J. et al. Electrochemical reduction of CO2 to ethylene glycol on imidazolium ion-terminated self-assembly monolayer-modified Au electrodes in an aqueous solution. Phys. Chem. Chem. Phys. 17, 26072–26078 (2015).
Nitopi, S. et al. Progress and Perspectives of Electrochemical CO2 Reduction on Copper in Aqueous Electrolyte. Chem. Rev. 119, 7610–7672 (2019).
Bussche, K. M. V. & Froment, G. F. A Steady-State Kinetic Model for Methanol Synthesis and the Water Gas Shift Reaction on a Commercial Cu/ZnO/Al2O3Catalyst. J. Catal. 161, 1–10 (1996).
Romero Cuellar, N. S. et al. Advantages of CO over CO2 as reactant for electrochemical reduction to ethylene, ethanol and n-propanol on gas diffusion electrodes at high current densities. Electrochim. Acta 307, 164–175 (2019).
Krause, R. et al. Industrial Application Aspects of the Electrochemical Reduction of CO2 to CO in Aqueous Electrolyte. Chemie Ing. Tech. 92, 53–61 (2020).
Edwards, J. P. et al. Efficient electrocatalytic conversion of carbon dioxide in a low-resistance pressurized alkaline electrolyzer. Appl. Energy 261, 114305 (2020).
Wen, G. et al. Continuous CO2 electrolysis using a CO2 exsolution-induced flow cell. Nat. Energy 7, 978–988 (2022).
Yang, H. et al. Carbon dioxide electroreduction on single-atom nickel decorated carbon membranes with industry compatible current densities. Nat. Commun. 11, 593 (2020).
Endrődi, B. et al. High carbonate ion conductance of robust PiperION membrane allows industrial current density and conversion in zero-gap carbon dioxide electrolyzer cell. Energy Environ. Sci. 13, 4098-4105 (2020).
Endrödi, B. et al. Multilayer Electrolyzer Stack Converts Carbon Dioxide to Gas Products at High Pressure with High Efficiency. ACS Energy Lett. 4, 1770–1777 (2019).
Liu, Z. et al. CO2 Electrolysis to CO and O2 at High Selectivity, Stability and Efficiency Using Sustainion Membranes. J. Electrochem. Soc. 165, J3371–J3377 (2018).
Endrődi, B. et al. Operando cathode activation with alkali metal cations for high current density operation of water-fed zero-gap carbon dioxide electrolysers. Nat. Energy 6, 439–448 (2021).
Kofuji, Y. et al. Efficient Electrochemical Conversion of CO2 to CO Using a Cathode with Porous Catalyst Layer under Mild pH Conditions. Chem. Lett. 50, 482–484 (2021).
Salvatore, D. A. et al. Designing anion exchange membranes for CO2 electrolysers. Nat. Energy 6, 339–348 (2021).
Li, D. et al. Durability of anion exchange membrane water electrolyzers. Energy Environ. Sci. 14, 3393–3419 (2021).
Mustain, W. E. et al. Durability challenges of anion exchange membrane fuel cells. Energy Environ. Sci. 13, 2805–2838 (2020).
Monteiro, M. C. O. et al. The Role of Cation Acidity on the Competition between Hydrogen Evolution and CO2 Reduction on Gold Electrodes. J. Am. Chem. Soc. 144, 1589–1602 (2022).
Resasco, J. et al. Promoter Effects of Alkali Metal Cations on the Electrochemical Reduction of Carbon Dioxide. J. Am. Chem. Soc. 139, 11277–11287 (2017).
VERMEIREN, P. Evaluation of the ZirfonS separator for use in alkaline water electrolysis and Ni-H2 batteries. Int. J. Hydrogen Energy 23, 321–324 (1998).
Hansen, K. U., Cherniack, L. H. & Jiao, F. Voltage Loss Diagnosis in CO2 Electrolyzers Using Five-Electrode Technique. ACS Energy Lett. 7, 4504–4511 (2022).
Weng, L. C., Bell, A. T. & Weber, A. Z. Towards membrane-electrode assembly systems for CO2 reduction: A modeling study. Energy Environ. Sci. 12, 1950–1968 (2019).
Harmon, N. J. & Wang, H. Electrochemical CO2 Reduction in the Presence of Impurities: Influences and Mitigation Strategies. Angew. Chemie Int. Ed. 61, e202213782 (2022).
Luc, W. et al. SO 2 -Induced Selectivity Change in CO2 Electroreduction. J. Am. Chem. Soc. 141, 9902–9909 (2019).
Jiao, F. et al. The impact of nitrogen oxides on electrochemical carbon dioxide reduction. Nat. Commun. 11, 5856 (2020).
Porter, R. et al. The range and level of impurities in CO2 streams from different carbon capture sources. Int. J. Greenh. Gas Control 36, 161–174 (2015).
Wakerley, D. et al. Gas diffusion electrodes, reactor designs and key metrics of low-temperature CO2 electrolysers. Nat. Energy 7, 130–143 (2022).
Edwards, J. P. et al. Pilot-Scale CO2 Electrolysis Enables a Semi-empirical Electrolyzer Model. ACS Energy Lett. 8, 2576–2584 (2023).

There is NO Competing Interest.

NatEnergykiyotaSupportingInformation.docx
Development of high-performance zero-gap carbon dioxide electrolysis cells using a hydrophilic porous membrane

Download PDF

Version 1

posted

You are reading this latest preprint version

Development of high-performance zero-gap carbon dioxide electrolysis cells using a hydrophilic porous membrane

Status:

Version 1

Abstract

Figures

Introduction

Conclusion

Method

Declarations

Acknowledgements

References

Additional Declarations

Supplementary Files

Status:

Version 1