Optimization of Ionomer Content in Anode Catalyst Layer for Improving Performance of Anion Exchange Membrane Water Electrolyzer

Anion exchange membrane (AEM) water electrolyzers, which are considered next-generation hydrogen production energy devices, generate hydrogen using a nonprecious metal as the electrocatalyst. However, most current studies tend to focus on the development of highly active electrocatalysts based on nonprecious metals


Introduction
Anion exchange membrane (AEM) water electrolyzers are the culmination of fundamental studies on various components and processes, such as electrocatalysts, ionomers, anion exchange membranes, electrode fabrication processes, membrane electrode assembly (MEA) fabrication processes, and cell operation protocols [1][2][3].In particular, because AEM electrolyzers can be operated under alkaline conditions, they can use nonprecious metals as catalysts [4].Hence, they are considered next-generation hydrogen production energy devices.In addition, although AEM water electrolyzers have a shorter history than proton exchange membrane (PEM) water electrolyzers, AEM electrolyzers are becoming more commercially viable owing to the efforts of many researchers.The main challenges impeding the commercialization of AEM electrolyzers are the catalytic activity of the electrocatalyst and the stability of the AEM.Although AEM electrolyzers are still in their early stages of research, significant progress has been made in the field of electrocatalysis.For instance, recently developed nonprecious metal electrocatalysts, such as those based on transition metal phosphides [5][6][7], transition metal hydroxides [8][9][10][11][12], transition metal sulfides [13][14][15], perovskite oxide [16][17][18], and transition metal oxides [19][20][21], exhibit activities greater than those of precious metal electrocatalysts in half-cells.However, most studies on these materials have tended to focus on halfcells, and there have been few reports on the use of these materials in AEM electrolyzers.In addition, with respect to the results obtained in the case of half-cells, although the developed nonprecious metal electrocatalysts outperform precious metal electrocatalysts, the performance of AEM electrolyzers remains lower than that of PEM electrolyzers [22].In order to realize catalytic activity in the AEM electrolyzer, research on manufacturing high-performance electrodes is necessary.
The catalyst layer of an electrode is composed of an electrocatalyst and an ionomer.The electrocatalyst has a significant effect on the kinetics of AEM electrolyzers [19,23].Therefore, the use of highly active electrocatalyst is required to high-performance AEM electrolyzers [8,9,24,25].Ionomers are polymeric binders used to form the catalyst layer on a substrate.Similar to the electrocatalyst used, the ionomer significantly affects the ohmic losses in AEM electrolyzers.When the ionomer is used to form a catalyst layer on a substrate, the electrical conductivity of the catalyst layer is reduced because the ionomer is an electrical insulator.Therefore, the electrical conductivity of the AEM electrolyzer is also reduced, resulting in performance degradation.However, the ionomer improves the ionic conductivity of the catalyst layer because of its high ionic conductivity, meaning that the ionic conductivity of the AEM electrolyzer is increased, resulting in improvements in its performance [26][27][28][29].Therefore, to realize high-performance AEM electrolyzers, the amount of ionomer used must be optimized so that the ionic conductivity can be maximized while minimizing the decrease in the electrical conductivity.In addition to affecting the ohmic losses, the ionomer also has a determining effect on the mass transport losses.AEM electrolyzers operate at current densities above 0.5 A/cm 2 , and under these conditions, oxygen and hydrogen are produced vigorously.When the generated gases are efficiently removed from the catalyst layer, the unnecessary losses due to mass transport can be minimized.To efficiently remove the gases generated on the surface of the catalyst layer as well as within it, it is necessary to ensure that the gases (H 2 and O 2 ) and liquid (electrolyte) have a clear passage [30,31].In particular, the ionomer has a significant effect on the pore distribution within the catalyst layer, and the use of an excessive amount of the ionomer can cause the pores to be blocked, thus significantly restricting mass transport [32][33][34][35][36]. Therefore, it is necessary to study the effects of the ionomer content because the performance of AEM electrolyzers can be improved by minimizing the ohmic and mass transport losses by optimizing the ionomer content of the catalyst layer.
In this study, we investigated the effect of the ionomer contents in the catalyst layer in order to realize a highperformance AEM electrolyzer.In particular, we attempted to determine whether the performance of AEM electrolyzers can be improved dramatically by optimizing the ionomer content in the catalyst layer of the oxygen evolution reaction (OER) electrode, which is considered the bottleneck of the water electrolysis reaction.In addition, by analyzing the electrochemical behavior of the fabricated AEM electrolyzer, it was confirmed that the ionomer content has a significant effect on the ohmic and mass transport losses in the AEM electrolyzer.We also confirmed that the optimization of the ionomer content not only improves the performance of AEM electrolyzers but also increases their stability.

Experimental
2.1.Synthesis of CuCo-Oxide.CuCo-oxide was used as the electrocatalyst for the OER.In a previous study, we synthesized CuCo-oxide by the coprecipitation method.Copper sulfate pentahydrate (CuSO 4 •5H 2 O, ACS reagent ≥ 98 0%, Sigma-Aldrich, USA) (12.5 mmol) and cobalt nitrate hexahydrate (Co(NO 3 ) 2 •6H 2 O, ACS reagent ≥ 98%, Sigma-Aldrich, USA) (25 mmol) were used as the precursors in the coprecipitation method [20].The precursors were dissolved in deionized (DI) Millipore water (18.2MΩ•cm) under vigorous stirring.The pH of the solution was controlled to 9.5 using NH 4 OH (28%, Samchun, Korea).The mixture was aged for 4 h and washed with DI water.After being washed, the aged powder was freeze-dried at −90 °C for 72 h.The dried powder was then annealed at 300 °C for 4 h in air; this powder is denoted as CuCo-oxide.Finally, using high-power ball milling (8000D and 8000M Mixer Mill, SPEX), the synthesized CuCooxide powder was ground to a homogeneous state.

Characterization of Physical
Properties.The morphologies and compositions of the fabricated anodes were analyzed using a field emission scanning electron microscopy (FE-SEM) system (JSM-7001F, JEOL, Japan) coupled with an energy-dispersive X-ray spectrometry (EDS) attachment.Prior to the SEM analysis, the samples were coated with gold for 60 s.High-resolution transmission electron microscopy (HR-TEM) images were obtained using a TALOS F200X (Thermo Fisher Scientific, USA) system.X-ray diffraction (XRD) patterns were recorded using an X-ray diffractometer (D/MAX 2500, Rigaku, Japan).

Electrochemical Analysis of Half-Cell.
To confirm the OER activity of CuCo-oxide, linear sweep voltammetry (LSV) was performed using a rotating disk electrode (RDE) at 2500 rpm.The ink solution for the half-cell was prepared using the catalyst powder (20 mg), 100 μL of a 5 wt% Nafion suspension (Alfa Aesar, USA), and 900 μL of ethanol.Before being drop-cast on the RDE, the ink solution was sonicated for 10 min.Subsequently, 5 μL of the catalyst ink was dropped on the RDE and dried in an oven at 80 °C for 1 min.
2.5.Preparation of Cathode.Pt/C (40 wt%, HISPEC 4000, Johnson Matthey) was used as the electrocatalyst for the hydrogen evolution reaction.To fabricate the cathode, an ink solution was prepared using Pt/C, DI water, IPA, and a Nafion.The weight ratio of Pt/C to Nafion was 60 : 40.The cathode catalyst layer was prepared on a piece of carbon cloth with microporous layers (Toray) by ultrasonic spraying.The amount of Pt/C used was 1 mg Pt /cm 2 .
2.6.Fabrication of AEM Electrolyzer.The MEA was fabricated using the cathode (Pt/C-coated carbon cloth), AEM (X37-50 Grade T, dioxide materials), and anode (CuCooxide-coated pressed nickel foam).The AEM electrolyzer was fabricated using the anode, the cathode, the AEM, a porous transport layer (nickel foam), and a stainless steel current collector.The active area of the AEM electrolyzer was 4.9 cm 2 .The AEM electrolyzer was supplied with 1 M KOH at 50 mL/min and was operated at 50 °C.The AEM electrolyzer was evaluated using a potentiostat (BP2C, ZIVE LAB), and its electrochemical performance was analyzed by chronopotentiometry and LSV measurements.Electrochemical impedance spectroscopy (EIS) was performed over the frequency range of 1 kHz to 100 kHz with an amplitude of 50 mV.Finally, the durability of the AEM electrolyzer was evaluated at 0.5 A/cm 2 for 100 h.

Results and Discussion
Figure 1 shows the schematic of the AEM electrolyzer, MEA, and catalyst particles (CuCo-oxide) coated with different amounts of ionomer.All electrodes were prepared by CCS, a method of coating a catalyst layer on a porous transport layer (PTL).MEA consists of cathode, AEM, and anode.The amount of ionomer used in the anode catalyst layer was controlled to be approximately 2, 7, or 12 wt% of the catalyst.The results of the physical and electrochemical analyses of CuCo-oxide are presented in Figures S1 and S2.The XRD pattern of the synthesized CuCo-oxide sample showed peaks corresponding to Cu 0.76 Co 2.24 O 4 (JCPDS: 01-076-1886).In addition, its selected area electron diffraction pattern could be indexed to the Cu 0.76 Co 2.24 O 4 phase.Moreover, the presence of a ring-like pattern indicated that CuCo-oxide was polycrystalline.The synthesized CuCo-oxide sample showed higher OER activity than that of IrO 2 in the half-cell.A photograph of the nickel foam and anode is shown in Figure S3.SEM images of the pressed nickel foam are shown in Figure S4.
The results of EDS mapping confirmed that the catalyst layer was formed successfully and that the catalyst and ionomer were uniformly distributed in it, as shown in Figure S5.The surface morphology of the catalyst layer was also observed by SEM; the results are shown in Figures 2(a)-2(c).The anode with 2 wt% ionomer exhibited a homogeneous distribution of CuCo-oxide.Moreover, even when the amount of ionomer was increased from 2 to 7 wt%, the surface morphology of the catalyst layer was observed to be similar.However, in the anode with 12 wt% ionomer, CuCooxide exhibited significant aggregation.In addition, the pores of the catalyst layer can be partially blocked if the ionomer is present in excess [33,34].Interestingly, the high- 3 International Journal of Energy Research magnification SEM images of the anodes with 7 and 12 wt% ionomers showed that these anodes were different from the anode with 2 wt% ionomer in that the surfaces of the former contained sputtered gold nanoparticles.As stated in the previous experimental section, all the anodes were sputtered with gold before the SEM analysis.The presence of the gold nanoparticles can be attributed to the formation of a thin ionomer film on the catalyst particles.When the gold particles were sputtered on the ionomer thin film, the gold particles underwent heterogeneous nucleation, which is observed by SEM [37].However, the anode with 2 wt% ionomer did not contain any gold nanoparticles and only showed CuCo-oxide.This can be explained as follows.In the case of the anode with 2 wt% ionomer, the ionomer was not uniformly distributed in the catalyst layer as the ionomer content was low.Thus, the ionomer network could not be completely formed.In the case of the anode with 7 wt% ionomer, on the other hand, the amount of ionomer was sufficient to form a complete ionomer network within the catalyst layer.However, in the case of the anode with 12 wt% ionomer, the surfaces of the CuCo-oxide particles in the catalyst layer were fully covered owing to the high ionomer content, and even the pores were blocked.Schematics of the anodes with the different ionomer contents are shown in Figure 2(d).
The contact resistances of the anodes with the different ionomer contents were measured at 100 kg/cm 2 , as shown in Figure 3(a).The contact resistance of the anode increased with the ionomer content, indicating that the ionomer acted as an electrical insulator in the catalyst layer.Therefore, when an anode with a large amount of ionomer is used in an AEM electrolyzer, the resistance of the AEM electrolyzer will increase.However, when the fabricated anodes were used in AEM electrolyzers, the AEM electrolyzer resistance showed a different trend from that seen in the case of the anode contact resistance, as shown in Figure 3(b).The high-frequency resistance (HFR), which is the intercept of the real impedance axis in the Nyquist plot, is the sum of the electronic resistance and the ionic resistance of the AEM electrolyzer [27].The electronic resistance is caused by the bulk and interfacial resistances, such as the resistances at the anode-porous transport layer (PTL) and PTL-current collector interfaces as well as those related to the components of the external electrical circuit.The ionic resistance includes the ionic resistance of the membrane and the ionic resistance of the catalyst layer.In this study, all the conditions other than the ionomer content in the catalyst layer were kept constant.Therefore, the change in the HFR was related to the difference in the ionomer content in the anode catalyst layer.The ionomer has the ability to conduct OH - ions and affects the ionic conductivity within the catalyst layer [26,38].The anode with 2 wt% ionomer exhibited a relatively high electrical conductivity (Figure 3(a)).However, the ionomer network was not well formed in this anode, and the ionic conductivity was not sufficiently high, resulting in a relatively high HFR value.Although the electrical conductivity 4 International Journal of Energy Research of the anode with 7 wt% ionomer was lower than that of the anode with 2 wt% ionomer, the ionomer network was well formed in the former, and the ionic conductivity inside its catalyst layer was higher.Therefore, its HFR value was the lowest.In the case of the anode with 12 wt% ionomer, the ionomer network was well formed.However, the electrical conductivity of the catalyst layer itself was extremely low owing to the excessively high ionomer content, resulting in a high HFR.
The performances of AEM electrolyzers with the anodes with the different ionomer contents were evaluated by LSV.The results are shown in Figure 3(c).The AEM electrolyzer with the anode with 7 wt% ionomer showed the best performance.The cell voltage (V cell ), to achieve a current density of 1.0 A/cm 2 , was 1.736, 1.724, and 1.765 V cell in the cases of the anodes with 2, 7, and 12 wt% ionomers, respectively; this trend was similar to that seen in the case of the HFR values of the AEM electrolyzers.In addition, the current densities of the AEM electrolyzers at 1.8 V cell are shown in Figure 3(d).To evaluate the effect of the resistance of the AEM electrolyzer, the ohmic loss (ƞ ohmic ) was corrected using the value of the HFR, which was obtained from EIS, as shown in Figures S6 and S7.Because the ohmic loss was corrected, the LSV results for all the iR-corrected AEM  International Journal of Energy Research electrolyzers were expected to be similar.However, interestingly, as the cell voltage was increased, a difference was observed in the increase in the current density (Figure S7).At low current densities (<0.5 A/cm 2 ), the cell voltages of all the AEM electrolyzers were similar.In general, the mass transport loss would be negligible at low current densities, indicating that the performance of the AEM electrolyzers is dominated by the kinetic overvoltage.Therefore, these results are reasonable.However, differences were observed in the performances of the AEM electrolyzers at high current densities.In this region, the performance is affected by the kinetics (ƞ kinetic ) and mass transport (ƞ mass ) because the ohmic loss is corrected.Therefore, these differences are probably attributable to the mass transport losses.It is likely that the pores in the catalyst layer limit mass transport because the pores act as passages through which the generated oxygen gas can escape.Consequently, the anodes with 2 and 7 wt% ionomers would be expected to have small mass transport losses, while the anode with 12 wt% ionomer would have a large mass transport loss.However, although the pores of the catalyst layer with 2 wt% ionomer were relatively not filled compared to the pores of the catalyst layer with 7 wt% ionomer, in the high current density region, the mass transport loss of this anode was higher than that for the anode with 7 wt% ionomer.On the other hand, the AEM electrolyzer with the anode with 7 wt% ionomer showed the smallest mass transport loss.To analyze the effect of the ionomer content on the electrochemical behavior of the AEM electrolyzer, the overvoltage of the AEM electrolyzer was deconvoluted based on ƞ kinetic and ƞ mass , as shown in Figures 3(e) and 3(f).Over the entire current density range, the ƞ kinetic values of all the AEM electrolyzers were similar (Figure 3(e)).This result was consistent with that shown in Figure S7.In addition, this indicated that the difference in the high current density region originated from the mass transport loss.At the same time, it can be seen that the amount of ionomer used also affected the mass transport loss.Over the entire current density range, the mass transport loss of the AEM electrolyzer with the anode with 7 wt% ionomer was the lowest (Figure 3(f)).For current densities lower than 1.25 A/cm 2 , the AEM electrolyzer with the anode with 12 wt% ionomer showed the highest mass transport loss.Interestingly, for current densities higher than 1.25 A/cm 2 , the AEM electrolyzer with the anode with 2 wt% ionomer showed the highest mass transport loss.Thus, the mass transport losses of the AEM electrolyzers with the anodes with 2 and 12 wt% reversed at a threshold current density of 1.25 A/cm 2 .This result was not in keeping with the 6 International Journal of Energy Research expectations in that it was assumed that the presence of an abundant number of pores not filled by the ionomer would reduce the mass transport loss.The main reason for this is the permeability of the catalyst layer.When the AEM electrolyzers were assembled, the area of the catalyst layer in direct contact with the PTL was compressed, and the permeability of the catalyst layer in this region was reduced.This limited the flow of water to the area of the  International Journal of Energy Research catalyst layer in contact with the PTL, resulting in a mass transport-limited region [39].In particular, for the anode with 2 wt% ionomer, the effect of this transport-limited region was more pronounced in the high current density region.If the ionomer network is not fully formed, the bonding force between the catalyst particles will be weak, and the degree of contact between the catalyst particles will be low or nonexistent as oxygen gas is generated violently.Therefore, only the catalyst particles located near the substrate participated in the OER.As a result, a significant mass transport loss occurred because of the high ratio of the area of the transport-limited region to that of the active catalyst layer, as shown in Figure 4(a).In the case of the anodes with 7 and 12 wt% ionomers, the ionomer network was sufficiently formed, and the bonding force between the catalyst particles was strong.Therefore, the catalyst particles were well connected to each other even at the high current densities, at which oxygen gas was generated violently.As a result, the entire catalyst layer remained active, and the effect of the transport-limited region was relatively small, as shown in Figure 4(b).To confirm this, the HFR values of the AEM electrolyzers were determined by in situ EIS measurements.Figure 4(c) shows the HFR values at different current densities.It can be seen that the HFR value decreased as the current density was increased.This decrease in the HFR value can be attributed to the generation of heat [40,41].Therefore, these results are reasonable.However, as the current density was increased, a difference was observed in extent of change in the HFR.The difference in the HFR values at 1.75 and 0.25 A/cm 2 is shown in Figure 4(d).The changes in the HFR values of the AEM electrolyzers using the anodes with 7 and 12 wt% ionomers were similar at approximately -1.86 and -1.72 mΩ, respectively.On the other hand, the change in the HFR of the AEM electrolyzer using the anode with 2 wt% ionomer was approximately -0.71 mΩ.These small changes in the HFR can be attributed to the fact that the oxygen gas generated within the catalyst layer reduces or prevents contact between the CuCo-oxide particles, resulting in a decrease in the electrical conductivity.Thus, the reduction in the resistance originating from heat production within the anode catalyst layer was offset.
Durability tests were performed at 0.5 A/cm 2 for 100 h, as shown in Figure 5(a).The initial cell voltages of the AEM electrolyzers showed a trend similar to that seen in the case of the LSV results (Figure 3(c)).The AEM electrolyzer using the anode with 7 wt% ionomer showed the lowest cell voltage (1.626 V cell ) while the electrolyzer with the anode with 12 wt% ionomer showed the highest cell voltage (1.652 V cell ).In addition, the initial cell efficiencies in the cases of the anodes with 2, 7, and 12 wt% ionomers were 76.49, 77.14, and 75.92%, respectively.During the durability tests, the cell voltage increased continuously.The AEM electrolyzers with the anodes with 7 and 12 wt% ionomers showed similar degradation rates (approximately 0.48 mV/h).However, the AEM electrolyzer with the anode with 2 wt% ionomer showed rapid degradation (approximately 0.82 mV/h within 100 h).The degradation rates of the AEM electrolyzers are shown in Figure 5(b).In addition, we also observed the changes in the cell efficiency.The efficiency deterioration rate of the AEM electrolyzer with the anode with 2 wt% was the highest.The efficiency deterioration rates of all the devices are shown in Figure 5(d).The differences in the degradation rates are also related to the differences in the ionomer contents of the catalyst layer because all the other conditions corresponding to the anodes were identical.The CuCo-oxide particles in the anodes with 7 and 12 wt% ionomers experienced a strong binding force owing to the presence of a well-formed ionomer network.However, those in the anode with 2 wt% ionomer experienced a weak binding force owing to the presence of a poor ionomer network.Therefore, some of the CuCo-oxide in the anode catalyst layer with 2 wt% ionomer was physically lost during the vigorous generation of oxygen gas.In addition, this result is consistent with the observed decrease in the electrical conductivity of the catalyst layer (Figure 4(a)) because of the loss of contact between the CuCo-oxide particles owing to the vigorously generated oxygen gas.The amounts of H 2 gas produced from the AEM electrolyzers were measured during the durability tests, and their Faradaic efficiencies were calculated, as shown in Figures 5(e) and 5(f).The calculated Faradaic efficiencies of all the AEM electrolyzers were approximately 99.2%.

Conclusions
In this study, the effects of the ionomer content within the anode catalyst layer of AEM electrolyzers were investigated through half-cell tests.The ionomer content of the anode catalyst layer affected the electrical conductivity and ionic conductivity of the AEM electrolyzers.The optimal ionomer content of the anode catalyst layer was determined to be 7 wt%.As the ionomer content in the anode catalyst layer was increased from 2 to 12 wt%, the ohmic resistance of the anode increased, resulting in an increase in the resistance of the AEM electrolyzer.However, the anode with an insufficient amount of the ionomer (2 wt%) exhibited relatively poor ion conductivity because the ionomer network was not fully formed.This, in turn, resulted in an increase of the resistance of the AEM electrolyzer.In addition, we found that the mass transport loss is also affected by the ionomer content of the anode catalyst layer.The AEM electrolyzer using the anode with 2 wt% ionomer showed high mass transport loss in the high current density region owing to an increase in the area of the transport-limited region.The AEM electrolyzer with the anode with 12 wt% ionomer showed high mass transport loss over the entire current density range because the pores of the anode catalyst layer were blocked owing to the high ionomer content.On the other hand, the AEM electrolyzer with the anode with 7 wt% ionomer showed the lowest mass transport loss over the entire current density range, as it contained a large number of pores not filled by excessive ionomer and a well-formed ionomer network.The durability of the AEM electrolyzers was also affected by the ionomer content in the anode catalyst layer.The AEM electrolyzers with the anodes with a sufficient amount of the ionomer (7 and 12 wt%) showed high stability during the durability tests.This too was owing to the presence of a well-formed ionomer network.However, the AEM electrolyzer with the anode with an insufficient

Figure 5 :
Figure 5: (a) Durability tests of AEM electrolyzer at 0.5 A/cm 2 for 100 h.(b) Degradation rates of AEM electrolyzer during the durability tests.(c) Cell efficiency of AEM electrolyzer at 0.5 A/cm 2 for 100 h.(d) Cell efficiency decrease rates during the durability tests.(e) Hydrogen gas production rates at 0.5 A/cm 2 during the durability tests.(f) Faradaic efficiency of AEM electrolyzer.