Introduction

Rechargeable (or secondary) air batteries, in which oxygen is used as the positive electrode active material, have high theoretical capacity and thus are expected to replace existing batteries, particularly for small electronic devices [1,2,3]. The theoretical energy density of lithium-air secondary batteries can reach 5200 Wh/kg, which is much higher than those of other secondary batteries, including lithium-ion batteries [4]. In several types of rechargeable air batteries, other metals, such as zinc [5, 6], aluminum [7], and magnesium, are used as negative electrodes [8,9,10]. However, their applications have been relatively limited because the use of metallic negative electrodes poses safety concerns and they often suffer from dendrite formation/growth and insufficient cyclability [11,12,13,14]. In addition, in those air batteries, liquid electrolytes, which have a risk of leakage, are typically used. To address these issues, redox-active organic molecules have been investigated as negative electrode materials [15,16,17,18,19,20,21]. In addition to the lack of dendrite formation, the advantages of redox-active organic molecules include low density, variable redox potentials owing to the manipulation of molecular and electronic structures, and inertness to water. For example, quinone derivatives and their polymeric compounds have been used in liquid electrolyte air batteries [15, 16]. Recently, we explored all-solid-state rechargeable air batteries (SSABs) in which redox-active organic compounds are used as the negative electrode and proton-conductive polymer membranes are used as the solid electrolyte, and we successfully demonstrated their charge/discharge performance, rate characteristics, and cyclability [22,23,24,25]. The available organic molecules include dihydroxybenzoquinone (DHBQ) and its polymer (PDBM) [22], alkyl-ether group-substituted anthraquinone (PE-AQ) [23], and naphthoquinone (NQ) [24] and its polymer (PPNQ) [25]. In the present study, we focused on anthraquinone-2-carboxylic acid (AQC; Fig. 1a) as the redox-active organic molecule because it is composed of conjugated quinone and hydrophilic -COOH groups, hence resulting in a low redox potential and thus a high cell voltage and Coulombic efficiency of the resulting SSABs. The effect of the thickness of the proton-conductive electrolyte membranes was also investigated.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

a Redox reaction and b solid-state cyclic voltammogram of anthraquinone-2-carboxylic acid (AQC)

Methods

Evaluation of the SSABs

The SSABs were operated at 40 °C and 100% RH, 80%RH or 60%RH.

Cyclic voltammetry (CV)

CV was conducted using a potentiostat–galvanostat (AUTOLAB, Metrohm Autolab B. V., The Netherlands). CV values were obtained with a potential sweep between -0.100 V and 1.00 V at a sweep rate of 20 mV s−1 with a 100 mL min−1 hydrogen supply to the oxygen electrode and nitrogen atmosphere in the AQC electrode.

Charge/discharge cycle test

Prior to the cycle test, the cell was initially discharged at a current of 4.41 mA (a current density of 1 mA cm-2) to check the discharge capacity, which was used to determine the C rate. The cycle test protocol is summarized in Table S1. Step 1: Nitrogen (100 mL min−1) was supplied to both electrodes. Step 2: The cell was connected with a potentiostat–galvanostat, and charging was performed at 15 °C until the negative electrode reached 0 V. Step 3: Oxygen was supplied to the oxygen electrode at 100 mL min−1 for 5 min. Step 4: The cell was discharged at 15 °C until the cell voltage reached 0 V. Step 5: Nitrogen was supplied to the oxygen electrode at 100 mL min−1 for 1 min. The cycle test was performed by repeating Step 2 to Step 5.

Charge/discharge test with air supplied to the oxygen electrode

The charge/discharge test protocol is summarized in Table S2. Step 1: Nitrogen (100 mL min−1) was supplied to both electrodes. Step 2: Air was supplied to the oxygen electrode at 476 mL min−1. The cell was connected with a potentiostat–galvanostat, and charging was performed at 15 °C until the negative electrode reached 0 V. Step 3: Air was supplied to the oxygen electrode at 476 mL min−1 for 5 min. Step 4: The cell was discharged at 15 °C until the cell voltage reached 0 V. Step 5: Air was supplied to the oxygen electrode at 476 mL min−1 for 1 min. The charge/discharge test was performed by repeating Step 2 to Step 5.

Rate characteristic test

The rate characteristic test protocol is summarized in Table S3: Step 1: Nitrogen was supplied to both electrodes at 100 mL min−1. Step 2: The cell was connected with a potentiostat–galvanostat, and charging was performed at 4 C until the negative electrode reached 0 V. Step 3: Oxygen was supplied to the oxygen electrode at 100 mL min−1 for 5 min. Step 4: The cell was discharged at a rate of 4, 10, 20, 40, 60, 80, or 100 C until the cell voltage reached 0 V. Step 5: Nitrogen was supplied to the oxygen electrode at 100 mL min−1 for 1 min.

Results and discussion

Two membrane–electrode assemblies were prepared with an AQC-based negative electrode, proton-conductive Nafion membranes, and a gas diffusion positive electrode in a similar manner as for our previous SSABs. A thin Nafion NRE 212 membrane (50 μm thick) and a thick Nafion 117 membrane (175 μm thick) were used, in which the cells are designated SSAB-AQC-50 and SSAB-AQC-175, respectively. The cyclic voltammogram (CV) of the negative electrode of SSAB-AQC-50, which showed a prominent redox peak centered at approximately 0.15 V vs. RHE, was based on the redox reaction of the AQC molecules with a utilization of 22%, as shown in Fig. 1b. Compared with NQ as a reference (ca. 0.44 V of the redox potential and 16% utilization) [24] in an SSAB with a similar configuration, the much lower potential and slightly higher utilization were intrinsic to AQC and advantageous for SSABs.

The charge and discharge curves of SSAB-AQC-50 and SSAB-AQC-175 at 40 °C are shown in Fig. 2, where fully humidified nitrogen was supplied to both electrodes during charging and fully humidified nitrogen and oxygen were supplied to the negative and positive electrodes, respectively, during discharging (see the Supporting Information for details). The charge curves were very similar for both cells, which indicated that the AQC-based negative electrodes functioned well in those solid cells. The charge capacity, which was calculated from the time when the negative electrode reached 0.1 V, was 65 mAh g-1 for SSAB-AQC-50 and lower than that of our previous SSAB-NQ (90 mAh g-1) [24] because the AQC molecules were not fully reduced at 0.1 V (charging was not carried out at potentials lower than 0.1 V to avoid unfavorable hydrogen evolution and proton reduction reactions). The charge capacities of SSAB-AQC-50 and SSAB-AQC-175 were comparable.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Charge (dashed) and discharge (solid) curves of SSAB-AQC-50, SSAB-AQC-175, and SSAB-AQC-175 (air) at 40 °C

During the discharge process, the open-circuit voltages (OCVs) of SSAB-AQC-50 (1.08 V) and SSAB-AQC-175 (1.16 V) were higher than that of SSAB-NQ (0.88 V), [24] which reflected the lower redox potential of AQC than that of NQ, as discussed above. The membrane thickness significantly affected the discharge capacity of the SSAB-AQC cells, with 18 mAh g-1 for SSAB-AQC-50 and 65 mAh g-1 for SSAB-AQC-175. The slightly higher OCV and much higher discharge capacity of SSAB-AQC-175 than those of SSAB-AQC-50 occurred because the thicker electrolyte membrane had less oxygen permeation, which mitigated the nonelectrochemical oxidation reaction of the AQC in the negative electrode by the oxygen that passed through from the positive electrode. This finding was also supported by the much higher Coulombic efficiency of SSAB-AQC-175 (88%) than that of SSAB-AQC-50 (30%). When air (also fully humidified) was supplied to the positive electrode in place of oxygen, SSAB-AQC-175 had an even higher discharge capacity (80 mAh g-1) and Coulombic efficiency (95%) because of the lower oxygen concentration in the air and thus lower oxygen permeation through the membrane.

The discharge rate characteristics of SSAB-AQC-175 were evaluated under two conditions: with fully humidified oxygen and partly humidified air (a relative humidity of 60% under more practical conditions) supplied to the positive electrode (Fig. 3a). The Coulombic efficiency is plotted as a function of the C rate in Fig. 3b. The Coulombic efficiency with oxygen was 88% at a C rate of 5 and nearly constant up to a C rate of 80, which indicated good rate performance. With respect to air, the efficiency at low C rates increased (to nearly 100%) but decreased slightly as the C rate increased. At a C rate of 100, the efficiency was 85% under both conditions, which suggested that the redox reaction of AQC was substantially fast in the solid-state even under low water contents (low-humidity conditions).

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

a Discharge curves at various C rates and b Coulombic efficiency as a function of the C rate of SSAB-AQC-175 with oxygen (100% RH) and air (60% RH) at 40 °C

The charge/discharge cycle stability of the cells is shown in Fig. 4, where the discharge rate and humidity were 15 C and 100% RH, respectively, for all the cases. The advantage of using a thicker electrolyte membrane was verified in the cycle test. The discharge capacity of SSAB-AQC-50 decreased rapidly as the number of cycles increased. The Coulombic efficiency of SSAB-AQC-50 was 30% in the 1st cycle and decreased as the number of charge/discharge processes decreased to 10% after 20 cycles. In contrast, the discharge capacity and Coulombic efficiency of SSAB-AQC-175 remained high during the cycles; the Coulombic efficiency was 90% in the 1st cycle and decreased to 79% in the 20th cycle. The potential changes in each electrode were carefully investigated (Fig. S1), and the deterioration as the number of cycles increased was more severe for the AQC electrode than for the oxygen electrode. Because even a thicker Nafion membrane would have a certain degree of oxygen permeability, this finding suggests that nonelectrochemical oxidation with crossed-over (permeated) oxygen could have deteriorated the redox performance of the AQC and its cyclability. Replacing oxygen with air improved the charge/discharge cyclability, and the Coulombic efficiency was 92% after 20 cycles (the average loss of efficiency for 20 cycles was only 0.15%) because of the lower oxygen partial pressure in the air and therefore lower oxygen permeation.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Charge/discharge cycle stability of SSAB-AQC-50 and SSAB-AQC-175. a Charge and discharge curves, b remaining discharge capacity and c remaining Coulombic efficiency as a function of the cycle number

Conclusions

In this study, all-solid-state rechargeable air batteries (SSABs) were fabricated using anthraquinone-2-carboxylic acid (AQC) as the negative electrode active material. AQC exhibited a redox peak at approximately 0.15 V vs. RHE in the solid state, which was lower than that of naphthoquinone (NQ), thus resulting in a higher cell voltage and better Coulombic efficiency. As a proton-conductive solid polymer electrolyte, a thicker Nafion membrane (175 μm thick) improved the discharge capacity and Coulombic efficiency because it reduced oxygen crossover from the oxygen electrode to the AQC electrode, which otherwise caused unfavorable nonelectrochemical oxidation of the AQC. Supplying air instead of pure oxygen to the positive electrode further improved the performance. By replacing oxygen with air, the SSAB achieved an even better Coulombic efficiency of 95% because the lower oxygen partial pressure reduced permeation through the membrane. The SSAB demonstrated good rate performance and maintained a Coulombic efficiency of 85% even at a high discharge rate of 100 C. The SSABs were cyclable with small losses of discharge capacity and Coulombic efficiency. The obtained results strengthened the advantages of SSABs as emerging energy devices.