High-Capacity Dual-Electrolyte Aluminum–Air Battery with Circulating Methanol Anolyte

Aluminum–air batteries (AABs) have recently received extensive attention because of their high energy density and low cost. Nevertheless, a critical issue limiting their practical application is corrosion of aluminum (Al) anode in an alkaline aqueous electrolyte, which results from hydrogen evolution reaction (HER). To effectively solve the corrosion issue, dissolution of Al anode should be carried out in a nonaqueous electrolyte. However, the main cathodic reaction, known as oxygen reduction reaction (ORR), is sluggish in such a nonaqueous electrolyte. A dual-electrolyte configuration with an anion exchange membrane separator allows AABs to implement a nonaqueous anolyte along with an aqueous catholyte. Thus, this work addresses the issue of anode corrosion in an alkaline Al–air flow battery via a dual-electrolyte system. The battery configuration consisted of an Al anode | anolyte | anion exchange membrane | catholyte | air cathode. The anolytes were methanol solutions containing 3 M potassium hydroxide (KOH) with different ratios of water. An aqueous polymer gel electrolyte was used as the catholyte. The corrosion of Al in the anolytes was duly investigated. The increase of water content in the anolyte reduced overpotential and exhibited faster anodic dissolution kinetics. This led to higher HER, along with a greater corrosion rate. The performance of the battery was also examined. At a discharge current density of 10 mA·cm−2, the battery using the anolyte without water exhibited the highest specific capacity of 2328 mAh/gAl, producing 78% utilization of Al. At a higher content of water, a higher discharge voltage was attained. However, due to greater HER, the specific capacity of the battery decreased. Besides, the circulation rate of the anolyte affected the performance of the battery. For instance, at a higher circulation rate, a higher discharge voltage was attained. Overall, the dual-electrolyte system proved to be an effective approach for suppressing anodic corrosion in an alkaline Al–air flow battery and enhancing discharge capacity.


Introduction
Aluminum-air batteries (AABs) offer an attractive combination of low cost, lightweight, and ultrahigh energy density [1,2]. AABs exhibit a high specific capacity of 2.98 Ah/g Al , which is Figure 1a demonstrates the setup employed for the measurement of HER. Hydrogen evolution tests were carried out during 60 min of immersion in methanol solutions containing 3 M KOH with different ratios of water (0%, 5%, 10%, and 20% (v/v)) at 30 • C. Before the test, Al samples were cleaned five times using methanol. Thus, an Al sample, 1.5 cm 2 in area, was put into a beaker containing 50 mL of the solution. Next, a funnel was positioned over the sample, ensuring the accumulation of all the hydrogen from the sample. Next, a burette was installed over the funnel and was initially full of the electrolyte. The hydrogen that was in the funnel passed into the burette and gradually displaced the electrolyte. The burette allowed the volume of evolved hydrogen gas to be measured as a function of time. Reaction rates were determined by the slope of the straight lines in the gasometry plots.
A half-cell test was carried out to examine corrosion of the Al anode in methanol solutions containing 3 M KOH with different ratios of water (0%, 5%, 10%, and 20% (v/v)). A three-electrode configuration was used to study electrochemical characterization. The three-electrode cell consisted of a platinum (Pt) counter electrode of 10 × 10 mm 2 , a Ag/AgCl reference electrode, and an Al plate working electrode of 10 × 10 mm 2 . The aluminum working electrode was rinsed three times with acetone before being tested at room temperature. A potentiostat/galvanostat (PAR VersaSTAT 3A, Ametek Inc., Berwyn, PA, USA) was employed to study EIS and Tafel polarization. EIS was examined at a frequency range of 200 kHz to 0.2 Hz and voltage excitation of 5 mV at open-circuit voltage (OCV). EIS was carried out to examine the corrosion of Al samples immediately after immersion in the solutions and after being immersed in the solutions for 90 min. Besides, Tafel polarization was performed at a scan rate of 5 mV·s −1 in the potential region cover at −500 mV vs. OCV to 0 V vs. Ag/AgCl. EIS using a frequency range of 1 Hz to 100 kHz with the excitation voltage of 10 mV RMS was also applied to measure ionic conductivity of the methanol solutions. The cell setup was composed of two chambers with a 10 × 10 mm 2 stainless steel (SUS316) electrode. The bulk resistance, determined from an equivalent circuit model, was then used to calculate the ionic conductivity of the solutions [25].
Further, morphology of the Al surface before and after being immersed in methanol solutions for 45 min was studied via SEM using a field emission SEM (ZEISS Sigma 500, Carl Zeiss AG, Oberkochen, Germany) at an accelerating voltage of 5 kV at a working distance between 1 and 5 mm. SEM was carried out on Al samples before and after being immersed in the solutions for 45 min.  Figure 1b shows a schematic illustration of the dual-electrolyte Al-air battery with flowing anolyte. The anode electrode was 30 × 50 × 0.19 mm 3 and weighed 0.8 g. The methanol solutions containing 3 M KOH with different ratios of water (0%, 5%, 10%, and 20% (v/v)) were used as the anolyte. The catholyte was a polymer gel electrolyte consisting of 3 M KOH dissolved in deionized water. Then, CBP940 (1.2 wt %) was added to each condition. The gel electrolyte offered satisfactory mechanical properties and reduced electrolyte loss due to evaporation. An anion-exchange membrane of 30 × 30 mm 2 was used to separate the anolyte and the catholyte. The anion-exchange membrane was activated by soaking in 5 M NaCl aqueous solution for 24 h at 30 °C. The cathode electrode consisted of three layers: a catalyst layer, a cathode current collector, and a gas diffusion layer. The cathode current collector was nickel (Ni) foam. The catalyst layer was in contact with the catholyte. The active area of the cathode reaction was 4 cm 2 . The catalyst layer was fabricated by casting a slurry mixture of MnO2 (0.3 g) as the ORR catalyst, with carbon BP2000 (0.7 g) and poly(styrene-co-butadiene) (0.1 g) dispersed in 10 mL of toluene. The coated Ni foam was pressed at 150 °C for 10 min using a manual hot press. To fabricate the gas diffusion layer, the outer surface of the air cathode was prepared by coating a mixture of carbon XC-72 and PTFE binder in 10 mL ethanol with a ratio of 80:20 wt/wt, respectively. The well-dispersed paste was applied onto the other side of  Figure 1b shows a schematic illustration of the dual-electrolyte Al-air battery with flowing anolyte. The anode electrode was 30 × 50 × 0.19 mm 3 and weighed 0.8 g. The methanol solutions containing 3 M KOH with different ratios of water (0%, 5%, 10%, and 20% (v/v)) were used as the anolyte. The catholyte was a polymer gel electrolyte consisting of 3 M KOH dissolved in deionized water. Then, CBP940 (1.2 wt %) was added to each condition. The gel electrolyte offered satisfactory mechanical properties and reduced electrolyte loss due to evaporation. An anion-exchange membrane of 30 × 30 mm 2 was used to separate the anolyte and the catholyte. The anion-exchange membrane was activated by soaking in 5 M NaCl aqueous solution for 24 h at 30 • C. The cathode electrode consisted of three layers: a catalyst layer, a cathode current collector, and a gas diffusion layer. The cathode current collector was nickel (Ni) foam. The catalyst layer was in contact with the catholyte. The active area of the cathode reaction was 4 cm 2 . The catalyst layer was fabricated by casting a slurry mixture of MnO 2 (0.3 g) as the ORR catalyst, with carbon BP2000 (0.7 g) and poly(styrene-co-butadiene) (0.1 g) dispersed in 10 mL of toluene. The coated Ni foam was pressed at 150 • C for 10 min using a manual hot press. To fabricate the gas diffusion layer, the outer surface of the air cathode was prepared by coating a mixture of carbon XC-72 and PTFE binder in 10 mL ethanol with a ratio of 80:20 wt/wt, respectively. The well-dispersed paste was applied onto the other side of the nickel foam and then pressed using a Energies 2020, 13, 2275 5 of 14 manual hot press at 350 • C for 15 min. The gas diffusion layer exhibited hydrophobicity and prevented leaking of the electrolyte whilst enabling oxygen gas to diffuse to the cell. The operation of the battery was carried out at 15 • C to minimize evaporation of the methanol solutions.

Full-Cell Test
In Figure 1c, a photographic image of the battery used in the experiment is shown. Figure 1d displays the cell design, which consisted of an aluminum anode, a chamber for the anolyte, an anion exchange membrane (acting as a separator), a chamber for the catholyte, as well as the air cathode. The distance between the cathode and anode electrodes was 18 mm. Further, both the battery discharge capacity and voltage-current polarization characteristic were analyzed by a battery testing system (CT-4008-5V20mA, Neware Technology Ltd., Shenzhen, China). Figure 2a shows the relationship between hydrogen gas evolution as a function of time. The linear variation for the volume of hydrogen gas evolved in time varied with water content in the methanol solutions. The volume of hydrogen represents how much aluminum in AABs was consumed by HER without electricity generation.

Results and Discussion
Energies 2020, 13, x FOR PEER REVIEW 5 of 15 the nickel foam and then pressed using a manual hot press at 350 °C for 15 min. The gas diffusion layer exhibited hydrophobicity and prevented leaking of the electrolyte whilst enabling oxygen gas to diffuse to the cell. The operation of the battery was carried out at 15 °C to minimize evaporation of the methanol solutions.
In Figure 1c, a photographic image of the battery used in the experiment is shown. Figure 1d displays the cell design, which consisted of an aluminum anode, a chamber for the anolyte, an anion exchange membrane (acting as a separator), a chamber for the catholyte, as well as the air cathode. The distance between the cathode and anode electrodes was 18 mm. Further, both the battery discharge capacity and voltage-current polarization characteristic were analyzed by a battery testing system (CT-4008-5V20mA, Neware Technology Ltd., Shenzhen, China). Figure 2a shows the relationship between hydrogen gas evolution as a function of time. The linear variation for the volume of hydrogen gas evolved in time varied with water content in the methanol solutions. The volume of hydrogen represents how much aluminum in AABs was consumed by HER without electricity generation. The condition of 0% water exhibited the hydrogen evolution rate of 0.017 mL cm −2 min −1 , while the condition of 5%, 10%, and 20% water showed hydrogen rates of 0.021, 0.036, and 0.074 mL cm −2 min −1 , respectively. The corrosion of Al could also take place in the case of 0% water (anhydrous) KOH methanol solution, thereby producing hydrogen gas. The corresponding reactions are shown in Equations (1)-(3) [21]:

Results and Discussion
or Al + 4OH → Al(OH) + 3e Cathode: 2CH OH + 2e → 2CH O + H ↑ Accordingly, two anodic reactions, which were related to the dissolution of Al, were involved. The cathodic reaction involved the reduction of methanol and produced hydrogen gas. However, as shown in Equation (4), when the electrolyte contained water, reduction of water could also occur, leading to hydrogen gas production. In both cases, hydrogen gas production was possible. However, the hydrogen production rate regarding the 3 M KOH anhydrous electrolyte was much slower than that of the electrolyte containing water, as shown in Equation (4): The condition of 0% water exhibited the hydrogen evolution rate of 0.017 mL·cm −2 min −1 , while the condition of 5%, 10%, and 20% water showed hydrogen rates of 0.021, 0.036, and 0.074 mL·cm −2 min −1 , respectively. The corrosion of Al could also take place in the case of 0% water (anhydrous) KOH methanol solution, thereby producing hydrogen gas. The corresponding reactions are shown in Equations (1)-(3) [21]: or Accordingly, two anodic reactions, which were related to the dissolution of Al, were involved. The cathodic reaction involved the reduction of methanol and produced hydrogen gas. However, as shown in Equation (4), when the electrolyte contained water, reduction of water could also occur, leading to hydrogen gas production. In both cases, hydrogen gas production was possible. However, the hydrogen production rate regarding the 3 M KOH anhydrous electrolyte was much slower than that of the electrolyte containing water, as shown in Equation (4): Energies 2020, 13, 2275 6 of 14 Figure 2a also indicates that HER was higher when the water content increased. The diagram further shows that the highest amount of hydrogen gas evolved was observed in the condition of 20% water, which was the highest water content studied in this work.
Tafel polarization was investigated via a three-electrode cell, namely, an Al working electrode, a Ag/AgCl reference electrode, and a platinum counter electrode. Figure 2b shows the Tafel polarization of Al in 3 M KOH methanol solutions with different ratios of water (0%, 5%, 10%, and 20% (v/v)). The OCV of the 3 M KOH anhydrous methanol electrolyte (0% water) was detected at −1.45 V. The OCV of the electrolytes containing water showed a negative shift as water content increased. The negative shift of OCV indicated a higher tendency of hydrogen evolution or corrosion of Al. The polarization resistance and other related properties were calculated using Butler-Volmer and Stern-Geary equations [26,27], as shown in Equations (5) and (6): where, I is the measured current density, I corr is the corrosion current density, E is the electrode potential, E corr is the corrosion potential, and β a and β c are the anodic and cathodic Tafel slopes, respectively. By approximating the Tafel slopes as a straight line, the polarization resistance (R p ) can be estimated. R p is the transition resistance between the electrode and the electrolyte. Table 1 displays Tafel polarization parameters consisting of corrosion potential (E corr ), corrosion current density (I corr ), anodic slope (β a ), cathodic slope (β c ) and polarization resistance (R p ).  Corrosion rate is proportionally related to polarization resistance; higher polarization resistance indicates lower corrosion current. Therefore, higher polarization resistance indicates greater corrosion inhibition property. The anodic slope shows the corrosion tendency of the anode, in other words, how a large overpotential is required to drive the reaction [28].
Results showed that polarization resistance decreased as the amount of water in the electrolyte increased. The smallest slope (0.63 V·dec −1 ) and smallest polarization resistance (36.71 Ω) provided by the electrolyte containing 20% water indicated the highest corrosion tendency of Al. The corrosion current density of the solution with 20% water was 0.006 A·cm −2 . In comparison, the highest anodic Tafel slope (1.05 V·dec −1 ) occurred in the 3 M KOH anhydrous methanol electrolyte. Thus, the corrosion current density of the 3 M KOH anhydrous methanol electrolyte was found to be 0.002 A·cm −2 , while the polarization resistance was 252.31 Ω. Though the higher polarization proved to have a decisive advantage in the corrosion inhibition of Al, it had negative effects on battery performance as higher overpotential is required for Al dissolution.
EIS technique was applied to study the characteristic of the Al dissolution process as well as the behavior of Al corrosion. Figure 3 displays Nyquist plots of Al in 3 M KOH methanol solutions containing different ratios of water (0%, 5%, 10%, and 20% (v/v)). All plots showed a similar characteristic. Similar plots were also reported for Al in alkaline aqueous solution [29]. The impedance Energies 2020, 13, 2275 7 of 14 spectra consisted of a large capacitive loop at high frequency, a small inductive loop at middle frequency, and a second small capacitive loop at low frequency. The capacitive semicircle at high frequency was ascribed to the redox reaction of Al ↔ Al + . This loop was found to be the rate-determining step in the charge transfer process [30]. The inductive loop at middle frequency resulted from an adsorption of intermediates on the Al surface. The adsorption of intermediate species, such as Al(OH) 3 and Al(OH) 4 − , were involved in the Al dissolution process. The other capacitive semicircle at low frequency was attributed to the Al + ↔ Al 3+ redox reaction [31]. It is significant that Al dissolution in the alkaline aqueous solution involves the dissolution of Al 2 O 3 film over the Al surface, leading to repeated Al 2 O 3 film dissolution and formation of a relatively thin Al 2 O 3 film [18]. Subsequently, Al 3+ ions, oxidized at the Al 2 O 3 film interface, migrate across the film to the bulk electrolyte. Simultaneously, the migration of O 2− ions, generated at the film/electrolyte interface, are responsible for continued generation of Al 2 O 3 at the film interface. Thus, the anodic dissolution of Al is controlled by the through-film dissolution of Al. In other words, such redox reactions occur according to the aluminum hydrous layer formation and subsequent precipitation of Al(OH) 3 .
Energies 2020, 13, x FOR PEER REVIEW 7 of 15 impedance spectra consisted of a large capacitive loop at high frequency, a small inductive loop at middle frequency, and a second small capacitive loop at low frequency. The capacitive semicircle at high frequency was ascribed to the redox reaction of Al ↔ Al . This loop was found to be the ratedetermining step in the charge transfer process [30].  In the case of KOH methanol electrolytes, it is possible that methyl oxide (CH3O − ) may inhibit the reaction of OH − and Al 3+ , which are the main reactants producing Al(OH) 3 and Al(OH) 4 − . Besides, CH3O − forms a more stable oxide film over the Al surface. However, the migration of Al 3+ through the film in this case was found to be slower than that of the aqueous alkaline solution. The 3 M KOH anhydrous electrolyte displayed the highest diameter of capacitive and inductive loops. However, when water content increased, it led to a decrease in the diameter size of the capacitive loop because a higher amount of OH − was produced. Moreover, an increase in water content can reduce charge In the case of KOH methanol electrolytes, it is possible that methyl oxide (CH 3 O − ) may inhibit the reaction of OH − and Al 3+ , which are the main reactants producing Al(OH) 3 and Al(OH) 4 − . Besides, CH 3 O − forms a more stable oxide film over the Al surface. However, the migration of Al 3+ through the film in this case was found to be slower than that of the aqueous alkaline solution. The 3 M KOH anhydrous electrolyte displayed the highest diameter of capacitive and inductive loops. However, when water content increased, it led to a decrease in the diameter size of the capacitive loop because a higher amount of OH − was produced. Moreover, an increase in water content can reduce charge transfer resistance. Consequently, the corrosion rate increased. In addition, the solution resistance negatively shifted. Furthermore, in Figure 3, the Nyquist plots of Al anode in 3 M KOH methanol solutions at immediate immersion of Al and after 90 min are shown. It was found that, regarding the 3 M KOH anhydrous methanol electrolyte, only the diameter of the capacitive and inductive loops increased after 90 min immersion. This was due to the slow reaction of Al-CH 3 O − . The resulting film showed a higher corrosion inhibition effect than the pristine Al surface. The film produced by the reaction of Al-CH 3 O − could effectively suppress Al corrosion. Besides, when water was introduced to the electrolyte, the width of the capacitive semicircle decreased with time due to the Al corrosion caused by the Al dissolution process. The film produced in the presence of water exhibited a lower corrosion inhibition effect than the pristine Al surface. Results thus indicated that the presence of water in the solution led to higher corrosion tendency of Al. Table 2 shows the EIS parameters for Al in 3 M KOH in methanol solutions with different ratios of water content as determined via fitting with an equivalent circuit model. The model included the effects of solution resistance (R s ), constant phase elements (CPE 1 and CPE 2 ), and charge transfer resistances (R 1 and R 2 ). Charge transfer resistances are the result of the kinetically controlled electrochemical reaction, while the constant phase elements represent the double layer capacitance, which arises from the absorption of charged species, at the electrode/electrolyte interface. Results indicated that the solution resistance of the Al dissolution process significantly decreased as water content increased. Moreover, as water content increased, there was a decrease in the double layer capacitance as well as in the charge transfer resistance at the electrode/electrolyte interface. Results thus confirmed that water helped decrease the double layer of Al in the dissolution process. In addition, the water content affected ionic conductivity of the solution. The ionic conductivity increased substantially as the water content increased. It was therefore further confirmed that water improved the Al dissolution process.   Following the examination of Al dissolution, t h e performance of the dual-electrolyte Al-air flow battery was investigated. The effects of methanol-based anolytes containing different ratios of water content were studied. Figure 5 presents the polarization characteristics of the battery. The electrochemical reactions involved in a traditional single aqueous alkaline electrolyte are shown in Equations (7)-(9) [2]:

Anode reaction:
Al → Al + 3e Cathode reaction: Overall reaction: Besides, Al also suffers a severe parasitic reaction, as shown in Equation (10): It was observed that the limiting current density of the battery (at the discharge voltage of 0.3 V) for each condition increased slightly, as follows: 15, 18, 30, and 36 mA cm −2 at 0%, 5%, 10%, and 20% water, respectively. Maximum power density was calculated via current density and voltage, providing 7.5, 8.5, 15.4, and 19.6 mW cm −2 , respectively. The battery exhibited linear polarization characteristics. The OCV of 3 M KOH anhydrous methanol electrolyte was 1.44 V. When the deionized water increased to 5%, 10%, and 20%, the OCVs were 1.45, 1.47, and 1.50 V, respectively. It was evident that by increasing the water content, OCV improved. However, it produced a Following the examination of Al dissolution, the performance of the dual-electrolyte Al-air flow battery was investigated. The effects of methanol-based anolytes containing different ratios of water content were studied. Figure 5 presents the polarization characteristics of the battery. The electrochemical reactions involved in a traditional single aqueous alkaline electrolyte are shown in Equations (7)-(9) [2]: Anode reaction : Al → Al 3+ + 3e − Cathode reaction : Overall reaction : 4Al significantly higher hydrogen evolution rate. Table 3 summarizes the maximum current density, power density, and internal resistance of the battery.   33.09 Ω at 0%, 5%, 10%, and 20% (v/v) water, respectively. Internal resistance decreased as water content increased; the results were in line with t h e EIS results. The presence of water in the electrolyte enhanced ionic conductivity. The presence of water also improved the performance of the battery because of its higher activity and ionic conductivity and thereby promoted Al dissolution. Figure 6 presents discharge profiles of the battery using 3 M KOH in methanol solution anolytes containing different ratios of water (0%, 5%, 10%, and 20% (v/v)). One hour before the test, the anion membrane was put in the cell, and the electrolyte was circulated without an Al anode. The discharge profiles of the battery discharged at a current density of 10 mA cm −2 are presented. Table 4 summarizes the specific capacity, utilization percentage, and discharge time for each condition. Besides, Al also suffers a severe parasitic reaction, as shown in Equation (10): It was observed that the limiting current density of the battery (at the discharge voltage of 0.3 V) for each condition increased slightly, as follows: 15, 18, 30, and 36 mA·cm −2 at 0%, 5%, 10%, and 20% water, respectively. Maximum power density was calculated via current density and voltage, providing 7.5, 8.5, 15.4, and 19.6 mW·cm −2 , respectively. The battery exhibited linear polarization characteristics. The OCV of 3 M KOH anhydrous methanol electrolyte was 1.44 V. When the deionized water increased to 5%, 10%, and 20%, the OCVs were 1.45, 1.47, and 1.50 V, respectively. It was evident that by increasing the water content, OCV improved. However, it produced a significantly higher hydrogen evolution rate. Table 3 summarizes the maximum current density, power density, and internal resistance of the battery. Table 3. The maximum current density, power density, and internal resistance of the dual-electrolyte AAB using an anolyte of methanol solutions containing 3 M KOH with different ratios of water (0%, 5%, 10%, and 20% (v/v)). The internal resistances observed in each case were 76.70, 65.70, 39.24, and 33.09 Ω at 0%, 5%, 10%, and 20% (v/v) water, respectively. Internal resistance decreased as water content increased; the results were in line with the EIS results. The presence of water in the electrolyte enhanced ionic conductivity. The presence of water also improved the performance of the battery because of its higher activity and ionic conductivity and thereby promoted Al dissolution. Figure 6 presents discharge profiles of the battery using 3 M KOH in methanol solution anolytes containing different ratios of water (0%, 5%, 10%, and 20% (v/v)). One hour before the test, the anion membrane was put in the cell, and the electrolyte was circulated without an Al anode. The discharge profiles of the battery discharged at a current density of 10 mA·cm −2 are presented. Table 4 summarizes the specific capacity, utilization percentage, and discharge time for each condition. Energies 2020, 13, x FOR PEER REVIEW 11 of 15  1.10 465 16% ~7 It can be seen that, in the 3 M KOH anhydrous anolyte, the battery reached the highest specific capacity of 2328 mAh/gAl, achieving utilization efficiency of 78%. Moreover, the battery could discharge at 0.87 V continuously for 40 h. Thus, this proved that the Al-air flow battery could exhibit a significantly higher coulombic efficiency using the 3 M KOH anhydrous methanol electrolyte. Accordingly, the hydrogen evolution rate was around 0.017 mL cm −2 min −1 , which was the lowest, emphasizing the fact that the parasitic reaction was significantly suppressed. However, when water content increased, specific capacity decreased successively, i.e., 1700, 1130, and 465 mAh/gAl at the condition of 5%, 10%, and 20% (v/v) water, respectively. Simultaneously, when water content increased, the average discharge voltage was enhanced, but the utilization percentage decreased. It is noted that the condition of 5% water could boost discharge voltage and provide utilization percentage of around 57% at a discharge voltage of 0.98 V for 28 h.

% Water
The battery using the anhydrous methanol anolyte proved to be the most effective in inhibiting anodic corrosion and provided the highest capacity. It was further investigated by employing different circulation rates, i.e., 10, 15, and 20 mL min −1 . Figure 6b shows the discharge profiles at different circulation rates of anolyte at a current density of 10 mA cm −2 . The highest specific capacity occurred at 10 mL min −1 , while the discharge voltage was 0.75 V. At a circulation rate of 10 mL min −1 , the average discharge voltage was found to be minimal because the circulation rate was minimal. Accordingly, this resulted in the accumulation of discharge products on the Al surface, the accumulation of which produced thick oxide films on the Al surface. Consequently, this led to greater charge transfer resistance and lower discharge voltage. In comparison, at 20 mL min −1 , specific capacity dropped significantly as a high flow rate thinned down the oxide films accumulated on the  Table 4. Summary of the specific capacity, utilization percentage, and discharge time of the batteries using 3 M KOH in methanol solution anolytes containing different ratios of water (0%, 5%, 10%, and 20% (v/v)) and a circulation rate of anolyte of 15 mL/min at discharge current density of 10 mA·cm −2 . It can be seen that, in the 3 M KOH anhydrous anolyte, the battery reached the highest specific capacity of 2328 mAh/g Al , achieving utilization efficiency of 78%. Moreover, the battery could discharge at 0.87 V continuously for 40 h. Thus, this proved that the Al-air flow battery could exhibit a significantly higher coulombic efficiency using the 3 M KOH anhydrous methanol electrolyte. Accordingly, the hydrogen evolution rate was around 0.017 mL·cm −2 min −1 , which was the lowest, emphasizing the fact that the parasitic reaction was significantly suppressed. However, when water content increased, specific capacity decreased successively, i.e., 1700, 1130, and 465 mAh/g Al at the condition of 5%, 10%, and 20% (v/v) water, respectively. Simultaneously, when water content increased, the average discharge voltage was enhanced, but the utilization percentage decreased. It is noted that the condition of 5% water could boost discharge voltage and provide utilization percentage of around 57% at a discharge voltage of 0.98 V for 28 h.

% Water
The battery using the anhydrous methanol anolyte proved to be the most effective in inhibiting anodic corrosion and provided the highest capacity. It was further investigated by employing different circulation rates, i.e., 10, 15, and 20 mL·min −1 . Figure 6b shows the discharge profiles at different circulation rates of anolyte at a current density of 10 mA·cm −2 . The highest specific capacity occurred at 10 mL·min −1 , while the discharge voltage was 0.75 V. At a circulation rate of 10 mL·min −1 , the average discharge voltage was found to be minimal because the circulation rate was minimal. Accordingly, this resulted in the accumulation of discharge products on the Al surface, the accumulation of which produced thick oxide films on the Al surface. Consequently, this led to greater charge transfer resistance and lower discharge voltage. In comparison, at 20 mL·min −1 , specific capacity dropped significantly as a high flow rate thinned down the oxide films accumulated on the Al surface and decreased charge transfer resistance. However, the higher circulation rate brought about lower utilization percentage of Al. Table 5 shows a comparison of the battery developed in this work and other AABs that have been previously reported. The dual-electrolyte AAFB showed the highest specific capacity. Results may differ depending on different parameters, such as ORR catalyst, electrolyte, and cell configuration. A previous study [17] suggested that Al passivation occurred after discharging for 4-5 h with Al in methanol-KOH solution. It was reported that the discharge time was less than 20 h, which was much lower than the system with anolyte circulation reported herein. This short discharge time was due to the passivation of the Al surface. The dual-electrolyte AAFB developed in this work exhibited lower power density. However, the power density depends on various parameters viz. ORR catalyst, catholyte, and the structure of gas diffusion layer. More advanced ORR catalysts and different catholytes can be applied to enhance the power density. In addition, further studies of other anolytes, such as ethylene glycol and glycerol solutions, implemented in dual-electrolyte AAFB are recommended. Paper-based AAB C 0.8 N/A [33] Paper-based solid electrolyte AAB Ag/C 3.8 900.8 [34]

Conclusions
In this work, corrosion of Al in 3 M KOH methanol solutions containing different ratios of water was studied and implemented as a circulating anolyte for a dual-electrolyte AAB. The results of HER revealed excellent corrosion inhibition of 3 M KOH anhydrous methanol solution. However, an increase in corrosion tendency followed the increase in water percentage. Tafel polarization and EIS confirmed that 3 M KOH anhydrous methanol solution exhibited the largest overpotential and highest polarization resistance. It was noted that overpotential could be reduced by introducing water. However, increasing deionized water caused considerably more hydrogen evolution. The dual-electrolyte system using 3 M KOH anhydrous methanol anolyte significantly suppressed parasitic reaction and provided the highest specific capacity of 2328 mAh/g Al at a current density of 10 mA·cm −2 . The battery generated Al utilization percentage of 78%. Increasing the water content led to greater discharge voltage and increased the current density but reduced the Al utilization percentage. Thus, the results affirm that the dual-electrolyte system with circulating methanol-based anolyte can significantly suppress corrosion and thereby enhance the capacity of AABs. The dual-electrolyte system is an effective approach and can be applied in other battery systems.