Robust bifunctional oxygen electrocatalyst with a “rigid and flexible” structure for air-cathodes

Abstract

The development of highly active air-cathodes with robust stability and a low price is of crucial significance for rechargeable Zn–air batteries and remains a great challenge. Herein, for the first time, we report a “rigid and flexible” material consisting of three-dimensional (3D) porous nickel-manganese oxide (Ni₆MnO₈) coupled with 1D ultrathin Au nanowires (Au-NWs) as an efficient bifunctional oxygen electrocatalyst, adopting α-naphthol-Au(III) as a precursor of Au-NWs and pre-formed Ni₆MnO₈ as a support. Ni₆MnO₈ acts not only as a robust carbon-free support that is stable in alkaline electrochemical conditions, but also as a highly active component for the oxygen evolution reaction (OER), while flexible Au-NWs contribute to the excellent oxygen reduction reaction (ORR) activity and act as a flexible conductive electronic network. The coupling of Ni₆MnO₈ and Au-NWs plays a complementary role in the two types of oxygen electrocatalytic reactions. Accordingly, their advantages have been optimally harnessed while overcoming their deficiencies. Moreover, a Zn–air battery assembled with such a rigid and flexible air-cathode has lower charge and discharge overpotentials and a higher cyclic stability than those with a mixed Pt/C+RuO₂ catalyst.

Introduction

Oxygen-reduction and oxygen-evolution reactions (ORR and OER) are two significant reactions (O₂ + 2H₂O + 4e⁻ ↔ 4OH⁻), with oxygen reduced by electrons while discharging and the reverse process occurring during charging in rechargeable zinc (Zn)–air batteries, a promising technology to cope with future energy demands by virtue of their theoretically high energy density (1084 Wh kg⁻¹), low-price, eco-friendly nature, and safety^{1,2,3,4,5,6,7,8,9,10}. However, Zn–air batteries in practical energy devices remain less efficient in output and are unsatisfactorily stable in cycling resulting from the sluggish kinetics of oxygen reactions and the poor stability of catalysts at the air-cathode^11,12,13,14. Additionally, efficient bifunctional catalysts that are active for both ORRs and OERs is one of the great challenges faced in designing rechargeable Zn–air batteries^{15,16,17,18,19,20}. Thus far, the best-known catalysts for oxygen electrocatalysis in alkaline media are prohibitive Pt-based metals for ORRs and Ir/Ru-based metals for OERs^21,22,23, but neither catalyst has sufficient bifunctional properties. Therefore, the development of highly active and stable bifunctional catalysts in light of low-price transition metals is important for practically applying Zn–air batteries.

In contrast to all the other types of transition-metal alternatives, the oxides, encompassing perovskites (e.g., CaMnO₃, LaMnO₃)²⁴, spinels (e.g., CuCo₂O₄, ZnCo₂O₄)^25,26, pyrochlore oxides (e.g., Pb₂Ru₂O_6.5, Y₂[Ru_2−xY_x]O_7−y)^7,27, and oxysulfides (e.g., CoO_0.87S_0.13)²¹, are potential bifunctional catalysts because of their chemical versatility and favorable stability in alkaline media. The electrocatalytic performances of metal oxides are limited by their intrinsically poor electrical conductivity, which can, however, be improved via supporting them on conductive materials. A synergetic combination of oxides with carbon materials is a general strategy for designing efficient bifunctional electrocatalysts^{28,29,30,31,32,33,34,35}. Notably, such carbon-supported metal oxides are electrochemically unstable, especially under harsh OER conditions because carbon oxidation/corrosion occurs at relatively low potentials (C + 2H₂O → CO₂ + 4H⁺ + 4e⁻, E₀ = 0.207 V vs. standard hydrogen potential)^36,37. Carbon corrosion has been identified as a main issue that greatly affects the stability of catalysts^11,38. In this regard, exploring robust carbon-free bifunctional electrocatalysts while retaining satisfactory electrical conductivity and evident catalytic activity is vitally necessary^36,38,39. For instance, Chen et al. developed a new carbon-free bifunctional catalyst comprising CaMnO₃ and Pt clusters (Pt/CaMnO₃) that exhibit evidently prolonged durability and optimized activity relative to that of carbon-supported Pt and CaMnO₃ alone³⁹. A similar result was also observed for a synergetic catalyst of Co₃O₄ and Ag particles (Ag/Co₃O₄)⁴⁰. More recently, Goodenough et al. reported a robust Ni₃FeN-supported Fe₃Pt (Fe₃Pt/Ni₃FeN) that can enable Zn–air batteries to achieve long-term cycling stability with high efficiency as an air-cathode³⁶. Although these robust carbon-free-supported noble metal catalysts become markedly more stable, the morphology and structure of these catalysts are unsatisfactory because electrocatalysis is a type of structure sensitive reaction, especially for ORRs.

Herein, a new, low-cost and highly active bifunctional electrocatalyst is designed and fabricated, consisting of flexible Au nanowires cross-linked with rigid Ni₆MnO₈ microspheres (Au-NWs/Ni₆MnO₈) as an efficient air-cathode. The Ni₆MnO₈ microspheres present a 3D hierarchically porous structure, while the Au nanowires have an ultrathin 1D anisotropic structure. Such “rigid and flexible” materials were prepared using in situ deposition of Au nanowires on the oxide support in the presence of α-naphthol. Unlike the numerous Au nanowires previously reported, the Au-NWs outlined here can be rapidly synthesized with ultrathin diameters. Strongly coupled Au-NWs/Ni₆MnO₈ have a superior bifunctional electrocatalytic performance, in which the Au-NWs are active for ORRs and Ni₆MnO₈ for OERs, coinciding with the concept of “rigid and flexible coupling”. Meanwhile, Au-NWs/Ni₆MnO₈ is electrochemically stable over the potential range of an air-cathode in an alkaline medium. The synergistic interaction of the 3D porous structure and the metallic 1D nanowires provides abundant catalytic active sites and facilitates mass and electron transport. Moreover, a Zn–air battery adopting Au-NWs/Ni₆MnO₈ as the air-cathode exhibits a larger peak power density (118 mW cm⁻²), a higher specific capacity (768 mAh g_Zn⁻¹) and a longer cycle life than that of a mixed Pt/C + RuO₂ air-cathode. This paper is the first to report the synthesis of rigid Ni₆MnO₈-supported flexible Au nanowires as an efficient synergistic catalyst for air-cathode reactions.

Materials and methods

Materials

Nickel (II) nitrate hexahydrate (Ni(NO₃)₂·6H₂O), manganese (II) acetate tetrahydrate (Mn(COOH)₂·4H₂O), and hydrogen tetrachloroaurate (III) tetrahydrate (HAuCl₄·4H₂O) were purchased from Alfa Aesar. Hexamethylenetetramine (HMTA), α-naphthol (C₁₀H₈O), and ruthenium oxide (RuO₂) were supplied by Aladdin. Commercial Pt/C was purchased from Johnson Matthey Chemicals Ltd. All reagents and chemicals were used without further purification.

Synthesis of Ni₆MnO₈ samples

Typically, Ni(NO₃)₂ and Mn(COOH)₂ in a 3:1 molar ratio was dissolved into 20 mL of deionized water (metal-ion concentration 40 mM) with 560 mg of HMTA by continuously stirring. The mixture was then transferred to a Teflon-lined steel autoclave and heated in an oven at 80 °C for 12 h. Once cooled to room temperature, the obtained precipitate was collected and washed several times with deionized water and subsequently dried at 60 °C to yield the hierarchical NiMn LDHs. Then, the resulting NiMn LDHs were placed in a tube furnace and annealed in air at 600 °C for 4 h at a heating rate of 5 °C min⁻¹. After natural cooling to room temperature, the porous Ni₆MnO₈ hierarchical microspheres were collected.

Synthesis of Au-NWs/Ni₆MnO₈ and Au-NWs/C samples

Typically, 25 mg of Ni₆MnO₈ powder (or carbon black) was added into a 10 mL mixture of deionized water and ethanol (V_water:V_ethanol = 1:1) with 144 mg (1 mM) of α-naphthol in an ultrasonic bath for 10 min to form a uniform solution. Then, 2.0 mL of 0.05 M HAuCl₄ solution was added into the above mixture and further stirred for 10 min. Finally, the mixture was treated by submersion in a water bath at 60 °C for a few minutes. The resulting Au-NWs/Ni₆MnO₈ and Au-NWs/C catalysts were then centrifuged and extensively washed with deionized water.

Synthesis of pure Au-NWs and Au-particles/Ni₆MnO₈ samples

Pure Au-NWs were prepared using an ultrafast method. Specifically, 2.0 mL of 0.05 M HAuCl₄ solution was added into a mixture of 5 mL of deionized water and 5 mL of ethanol with 1 mM alpha naphthol in an ultrasonic bath for 10 min to form a uniform solution, followed by submersion in a water bath at 60 °C for a few minutes. For the synthesis of the Au-particles/Ni₆MnO₈ catalyst, Au colloids were first synthesized in a trisodium citrate solution: a 20 mL mixed solution containing 0.10 mM HAuCl₄ and 0.25 mM trisodium citrate was prepared at room temperature, and NaBH₄ was also dissolved in deionized water; the resulting solution (0.01 M) was used as a reducing agent; with the dropwise addition of 0.5 mL of the NaBH₄ solution, the red-wine solution containing Au colloids was formed. Then, an appropriate aliquot of Au colloids was mixed with Ni₆MnO₈ in a beaker while being vigorously stirred for 30 min, resulting in the deposition of Au colloids on the Ni₆MnO₈ supports. Finally, the Au-particles/Ni₆MnO₈ catalyst was obtained by filtration and washing with water to remove excess surfactant and reducing agent.

Physicochemical characterization

The phase purity and crystallinity of the products were identified using X-ray powder diffraction (XRD) on a Rigak Smartlab X-ray with Cu-Kα radiation (λ = 0.1541 nm). Scanning electron microscopy (SEM) and energy-dispersive X-ray analysis (EDX) were performed using a Hitachi S4800 SEM. Transmission electron microscopy (TEM) and scanning transmission electron microscopy (STEM) were performed using a JEOL 2010F TEM/STEM operated at 200 kV. X-ray photoelectron spectroscopy (XPS) was performed using a Thermo VG Scientific ESCALAB 250 spectrometer with an Al-Kα radiator. The binding energy (BE) was calibrated using the C 1s peak energy of 284.6 eV. The Brunauer–Emmett–Teller (BET)-specific surface area was measured at 77 K using a Micromeritics ASAP 2050 system. The electrical conductivity was measured using 4-probe conductivity measurements on an ST-2722 semiconductor resistivity of the powder tester (Suzhou Jingge Electronic Co., Ltd., China) under a pressure of 10 MPa.

Electrochemical measurements

Electrochemical measurements were carried out using a CHI 760E electrochemical analyzer (Shanghai, Chenghua Co.). A standard three-electrode system was used, including a rotating disk electrode (RDE) or rotating ring-disk electrode (RRDE) as the working electrodes (0.196 cm²), a platinum wire as the auxiliary electrode, and a saturated calomel electrode protected by a Luggin capillary with KCl solution as the reference electrode. The catalyst ink was prepared by ultrasonically dispersing the mixture of 5 mg of catalyst, 1 mL of ethanol, and 20 μL of 5 wt.% Nafion solution; 10 μL of the catalyst ink was pipetted and spread onto the electrode. The loading of the studied catalysts was 250 µg cm^–2. For the ORR test, the background current was measured under a N₂ atmosphere at an identical rotation speed and scan rate as for measurements conducted under O₂. The ORR activities of the catalysts were measured via the RRDE voltammograms in 0.1 M KOH electrolyte at the predefined rotation rate of 1600 rpm and a scan rate of 5 mV s^–1. Before testing, O₂ was passed through the electrolyte for at least 20 min to saturate the electrolyte with O₂. To remove the capacitive current of the working electrode, the background current was measured by running the above electrodes in N₂-saturated 0.1 M KOH and then was subtracted from the ORR polarization curve. The percentage of HO₂^– intermediate production (%HO₂^–) and electron transfer number (n) were determined using the following equations:

$$\% {\mathrm{HO}}_2^ - = \frac{{200I_{\mathrm{r}}}}{{NI_{\mathrm{d}} + I_{\mathrm{r}}}}\quad \quad n = \frac{{4NI_{\mathrm{d}}}}{{NI_{\mathrm{d}} + I_{\mathrm{r}}}},$$

where I_d is the disk current, I_r is the ring current, and N is the current collection efficiency of the Pt ring, which was determined to be 0.37. For the OERs, the polarization curves were also measured in 0.1 M KOH solution recorded from 1.1 to 2.0 V at a scan rate of 5 mV s^–1. Note that high-purity O₂ is bubbled through the electrolyte during testing to fix the reversible oxygen potential (or ensure the O₂/H₂O equilibrium at 1.23 V vs. RHE). To avoid the peeling of the catalyst that is caused by evolved O₂ adhesion, a rotation speed of 1600 rpm was used during the OERs.

The Zn–air battery tests were performed with a homemade Zn-air cell. The air-cathode consisted of hydroholic carbon paper with a gas diffusion layer on the air-facing side and a catalyst layer on the water-facing side. The catalyst layer was made by loading catalyst ink onto the carbon paper by drop-casting, with a loading of 10 mg cm⁻² for all the catalysts. A polished Zn plate with a thickness of 0.3 mm was used as the anode. A 0.2 M ZnCl₂ + 6 M KOH mixed solution was used as the electrolyte. The gas diffusion layer had an effective area of 0.5 cm² and allows O₂ from the ambient air to reach the catalyst sites. A Land CT2001A system was used to perform the cycling test with a 5-min rest time between each discharge and charge at a current density of 10 mA cm^‒2. Each discharge and charge period was set to be 20 min.

Results

Structure and composition of Au-NWs/Ni₆MnO₈

Figure 1a presents a schematic diagram for the fabrication process of the Au-NWs/Ni₆MnO₈ catalyst, which primarily involves three steps: (i) co-deposition of Ni and Mn precursors to form hierarchical NiMn layered double hydroxides (NiMn LDHs) with the aid of HMTA, in which HMTA was used as a hydrolysis agent to slowly alkalize the solution and homogeneously precipitate out bimetallic NiMn LDHs; (ii) pyrolysis treatment at 600 °C in air to promote the formation of the hierarchically porous Ni₆MnO₈ phase; (iii) and ultrafast in situ growth of Au nanowires on the oxide support via a novel and simple water bath method in the presence of α-naphthol. The experimental details are provided in the experimental section. The XRD pattern of NiMn LDHs (Fig. S1) shows diffraction peaks at 2θ of 9.5, 19.0, 34.7, 38.7, 44.6, and 60.0, which can be, respectively, indexed to (003), (006), (012), (015), (018), and (110) reflections of the rhombohedral LDH phase. Figure 1b shows the XRD patterns of the Au-NW, Ni₆MnO₈, and Au-NWs/Ni₆MnO₈ samples. As observed, all of diffraction peaks for the obtained oxide support are observed to be well geared into the face-centered cubic (fcc) murdochite-type Ni₆MnO₈ rigid structure (space group: Fm3m (225), JCPDS No. 83–1186, a = b = c = 8.32 Å). No other crystalline phase was detected, which indicates that the final product has a high purity. Murdochite-type Ni₆MnO₈ has a derived rock-salt structure, and one-eighth of the Ni²⁺ ions are replaced by Mn⁴⁺ ions and vacancies (Fig. 1c and Fig. S2)^41,42. This murdochite-type rigid structure was found to be chemically stable after being soaked in 0.1 M KOH solution for 2 weeks, as demonstrated by the XRD patterns (Fig. S3). There are five additional diffraction peaks of the Au-NWs/Ni₆MnO₈ hybrid appearing at 38.3^o, 44.5^o, 64.7^o, 77.7^o, and 81.9^o as compared to Ni₆MnO₈; they are consistent with the (111), (200), (220), (311), and (222) reflections of fcc Au (space group: Fm3m (225), JCPDS No. 65-8601, a = b = c = 4.072 Å). The XRD results indicate the successful reduction of the Au(III) precursor with α-naphthol as the reducing agent.

Fig. 1

a Schematic illustration of the formation of the Au-NWs/Ni₆MnO₈ catalyst; b XRD patterns of the Au-NW, Ni₆MnO₈, and Au-NWs/Ni₆MnO₈ samples, drop lines corresponding to Ni₆MnO₈ (JCPDS No. 83-1186) and Au (JCPDS No. 65-8601); c the crystal structure of the Ni₆MnO₈ phase

Micro/nanostructures of the as-synthesized NiMn LDHs, Ni₆MnO₈, and Au-NWs/Ni₆MnO₈ were characterized specifically using electron microscopy technology. Representative SEM images (Fig. 2a and Fig. S4) of the NiMn LDHs display a 3D hierarchical porous architecture comprising dozens of interconnected sheet-like subunits. After thermal pyrolysis, the Ni₆MnO₈ sample retains the structure and morphology integrity of the hierarchical NiMn LDH microspheres, while the sheet-like subunits of the NiMn LDHs were completely transformed into porous ones (Fig. 2b and Fig. S5). The porous subunits packed with particles make the surface much coarser (Fig. S5), hinting at open and accessible mesopore channels, which should provide open pathways for mass transportation^43,44. Figure 2c and d and Fig. S6 display SEM images of Au-NWs/Ni₆MnO₈. After the decorating of Au nanowires, the Au-NWs/Ni₆MnO₈ sample manifests a 3D interconnected architecture akin to the original Ni₆MnO₈ oxides. Remarkably, the randomly tortuous Au nanowires are uniformly dispersed on the surface of the oxides, filling the pores and bridging the boundary of the porous sheet-like subunits. The flexible Au nanowires can work conductive bridges to offer effective electron transport channels. Although Ni₆MnO₈ itself has a poor conductivity (2.46 × 10⁻³ S m⁻¹), the Au-NWs/Ni₆MnO₈ hybrid demonstrates a better electrical conductivity of approximately 1.25 × 10³ S m⁻¹ that can facilitate electron transport during electrocatalysis. To better investigate the structural features of Au-NWs/Ni₆MnO₈, TEM measurements were conducted, as presented in Fig. 2e and f and Fig. S7. Flexible nanowires with a high morphological yield (100%) and an ultrathin diameter of approximately 4 nm (Fig. S6d) were observed on the Ni₆MnO₈ surface. The high-resolution TEM (HRTEM) image and the corresponding fast Fourier transform (FFT) patterns show two types of crystalline domains of Au and Ni₆MnO₈ oriented along the [110] and [100] zone axes (Fig. 2g), respectively. The EDX elemental mappings (Fig. 2i and Fig. S8) reveal a uniform distribution of Ni, Mn, O, and Au in the STEM image (Fig. 2h), which demonstrates the homogeneous growth of Au nanowires on the porous Ni₆MnO₈ surface. The content of Au in the Au-NWs/Ni₆MnO₈ is approximately 22 wt.%, which was demonstrated by the EDX spectrum (Fig. S9).

Fig. 2

SEM images of a NiMn LDHs microspheres; b Ni₆MnO₈ microspheres; c, d Au-NWs/Ni₆MnO₈; e, f TEM images of Au-NWs/Ni₆MnO₈; g HRTEM image of Au-NWs/Ni₆MnO₈ and the corresponding FFT patterns from the selected region of Ni₆MnO₈ and Au-NWs; h STEM image of Au-NWs/Ni₆MnO₈; and i the corresponding EDX element mappings

A nitrogen adsorption–desorption measurement was performed to ascertain the porosity of Au-NWs/Ni₆MnO₈. The adsorption–desorption curves (Fig. 3a) manifest representative type-IV isotherms with a distinctive hysteresis loop, which indicates the existence of microporous/mesoporous structures. The BET-specific surface area was calculated to be approximately 176 m² g⁻¹. The surface state of the Au nanowires and Ni₆MnO₈ and the interaction between them were ascertained using XPS. The full XPS spectrum clearly indicates the co-existence of Au, Mn, Ni, and O elements in the Au-NWs/Ni₆MnO₈ hybrid (Fig. S10). Figure 3b shows the high-resolution Au4f spectrum of Au-NWs/Ni₆MnO₈. The Au4f peaks centered at 83.7 and 87.4 eV are assigned to the spin-orbit doublet characteristic of metallic Au. Note that the Au4f peaks for Au-NWs/Ni₆MnO₈ shifted to a lower BE by 0.48 eV in contrast to that of the non-supported Au nanowires, which was prepared via the direct α-naphthol reduction method (see Experimental for details). Given the similar morphology and surface properties of the Au particles in Au-NWs/Ni₆MnO₈ and non-supported Au-NWs (Fig. S11), the Au4f-BE shift of Au-NWs/Ni₆MnO₈ probably arose from the strong incorporation between the oxide support and Au nanowires. The XPS analysis of the Au-NWs/Ni₆MnO₈ also demonstrated that Ni (II) and Mn (IV) are the dominant species in the Ni₆MnO₈ (Fig. 3c, d). Additionally, a definite positive shift was found in the BE of the Ni2p and Mn2p peaks, for Au-NWs/Ni₆MnO₈ relative to Au-free Ni₆MnO₈. The positive shifts in the BE of the Ni2p and Mn2p peaks further demonstrate the presence of electron interactions between Au and Ni₆MnO₈ in Au-NWs/Ni₆MnO₈, consequently transferring electrons from Ni₆MnO₈ to metallic Au. Note that α-naphthol has little influence on the valence state of Ni₆MnO₈ because of its low reducing capacity. In contrast with numerous Au nanowires that have been reported previously, the Au-NWs outlined here have better structural advantages, and the α-naphthol-driven synthetic method is deemed to be more rapid and effective. Detailed comparisons are provided in Table S1. Notably, α-naphthol is critically significant in the formation of Au nanowires. In the absence of α-naphthol, the HAuCl₄ precursor cannot be reduced under the present HAuCl₄-α-naphthol-Ni₆MnO₈ conditions, while in the presence of α-naphthol and without adding Ni₆MnO₈, pure Au nanowires also can obtained rapidly. Accordingly, α-naphthol is deduced to be able to not only serve as a reducing agent but also direct the formation of ultrathin Au nanowires. To confirm the rapid formation of Au nanowires, the reaction process was continuously monitored using UV–Vis spectroscopy at the surface plasmon resonance (SPR) peak of the Au particles (approximately 550 nm). As Fig. 3e indicates, the intensity of the SPR peak progressively increases as the reaction progress until equilibrium occurs. The acquisition of the equilibrium point indicates the end of the reduction reaction. We found that it merely takes 130 s to completely reduce the HAuCl₄ precursor under the present reaction conditions. This phenomenon might stem from the strong molecular interaction between α-naphthol and HAuCl₄ that was ascertained from the UV-Vis spectra (Fig. 3f). When α-naphthol was incorporated into the HAuCl₄ solution, the characteristic absorption peak at 317 nm of the HAuCl₄ species is markedly blue-shifted. Moreover, as the nanowires were directed by α-naphthol-Au(III), due to the sufficient pre-interaction between α-naphthol and Ni₆MnO₈ support through the electrostatic binding, the Au nanowires capped by α-naphthol had the opportunity to come into intimate contact with the oxide support, resulting in the in situ deposition of the Au nanowires onto the Ni₆MnO₈.

Fig. 3

a N₂ adsorption–desorption isotherms of Au-NWs/Ni₆MnO₈ microspheres; b high-resolution Au4f, c Ni2p, and d Mn 2p XPS spectra of Au-NWs/Ni₆MnO₈ and Au-NWs; e the intensity of SPR peak changes as a function of time, the inset is the UV-Vis spectrum of a Au nanowire after a complete reaction; f UV-Vis spectra of HAuCl₄, α-naphthol, and α-naphthol–HAuCl₄ complex solutions

Bifunctional electrocatalytic activities for the OERs and ORRs

Given the unique structural advantages of the Au-NWs/Ni₆MnO₈ hybrid, that is, inclusive of numerous mesopores, high BET surface area, and prominent electrical conductivity, they should be very promising for electrocatalysis. First, the ORR activity of Au-NWs/Ni₆MnO₈ (Fig. 4a) was evaluated using a RDE. The performances of the Ni₆MnO₈ and Au-NWs samples are provided for comparison. Au-NWs/Ni₆MnO₈ shows significantly improved ORR activity in terms of its more positive half-potential (E_1/2 = 0.90 V) as compared with Au-NWs (0.88 V) and Ni₆MnO₈ (0.62 V), suggesting that oxygen was more easily reduced on the Au-NWs/Ni₆MnO₈ catalyst. Note that ultrathin Au-NWs are an active component for ORRs, and their activity is higher than those of previously reported Au-based catalysts (Table S2). To probe the effect of Au-NWs on ORR performance, the Ni₆MnO₈-supported Au nanoparticles (Au-NPs/Ni₆MnO₈) were subjected to an ORR test. The detailed characterizations of Au-NPs/Ni₆MnO₈ are presented in Fig. S12. Au-NPs/Ni₆MnO₈ exhibited inferior ORR activity in comparison to the Au-NWs/Ni₆MnO₈ catalyst (Fig. S13), further verifying that the anisotropically ultrathin 1D nanostructure accounts for the high ORR activity of Au-NWs/Ni₆MnO₈. The anisotropic 1D nanowires possess many exposed surface atoms that can work as high-efficiency active sites, thus elevating the atom utilization efficiency of the catalysts^45,46. Ni₆MnO₈ microspheres can act as a support and promoter to facilitate ORRs, due to their unique 3D hierarchically porous structure. Additionally, the Mn (IV) sites from Ni₆MnO₈ oxide for the enhanced ORR activity cannot be neglected. This is because Mn (IV) cations can effectively assist the charge transfer to adsorb oxygen on the catalyst surface, facilitating HO₂⁻ decomposition^47,48. Thus, the strong incorporation of Au-NWs with Ni₆MnO₈ (i.e., synergistic effect) makes hybridization a very promising ORR catalyst, even better than the commercial Pt/C catalyst (Fig. 4a). As Table S3 lists, the Au-NWs/Ni₆MnO₈ catalyst is also more active than the reported ORR catalysts. Additionally, Au-NWs/Ni₆MnO₈ has a smaller Tafel slope of 62 mV Dec⁻¹ (Fig. S14) than that of Au-NWs (63 mV Dec⁻¹), Ni₆MnO₈ (140 mV Dec⁻¹), and commercial Pt/C (74 mV Dec⁻¹), indicating a better kinetic process. To quantify the catalysts’ ORR efficiencies, RRDE measurements were employed. Within a wide potential range (0.4–0.8 V), the average electron transfer number (n) and HO₂⁻ yield for Au-NWs/Ni₆MnO₈ are ca. 3.85 and 7.47% (Fig. 4b and Fig. S15), respectively, which are obviously superior to those for the Ni₆MnO₈ sample (2.74 and 37.5%). This fact further indicates that Au nanowire is a favorable conductive electronic network (n: 3.78 and HO₂⁻ yield: 10.5%). These values are very similar to those for the commercial Pt/C catalyst (3.88 and 5.90%), indicating a good four-electron ORR pathway for the Au-NWs/Ni₆MnO₈ catalyst.

Fig. 4

a ORR polarization curves of catalysts in O₂-saturated 0.1 M KOH (rotation rate: 1600 rpm; sweep rate: 5 mV s^–1); b bar plots of the electron transfer number and percentage of peroxide; c OER polarization curves of the catalysts in O₂-saturated 0.1 M KOH (rotation rate: 1600 rpm; sweep rate: 5 mV s^–1); d bar plots of the E_onset and E_1/2

Apart from the remarkable ORR activity, the Au-NWs/Ni₆MnO₈ catalyst also has very good OER activity. Figure 4c shows the OER polarization curves for Au-NWs/Ni₆MnO₈ and other samples. Clearly, the Au-NWs/Ni₆MnO₈ catalyst exhibits a lower overpotential and a higher current density over the provided potential range compared to the other references (Fig. 4d). At the current density of 10 mA cm⁻², Au-NWs/Ni₆MnO₈ has a very low overpotential (E_J10) of 0.36 V, outperforming the state-of-the-art RuO₂ catalyst (0.39 V), and is even comparable to those of other non-noble metal catalysts (Table S4). Pure Au-NWs show a poor OER response, indicating that unsupported Au-NWs do not work as an active catalytic site for the OER. The excellent OER activity of Au-NWs/Ni₆MnO₈ probably principally arise from the intrinsic OER property of Ni₆MnO₈, as indicated by the lower onset-potential compared with that of Au-NWs. As shown in Fig. S16, the Tafel plots for the OER of Au-NWs/Ni₆MnO₈ exhibit a slope of 62 mV Dec⁻¹, lower than that of the Au-NW (101 mV Dec⁻¹), Ni₆MnO₈ (82 mV Dec⁻¹), and RuO₂ (76 mV Dec⁻¹) catalysts. To further explore the active site of Ni₆MnO₈ for OERs, we performed a theoretical investigation by calculating the free energies based on density functional theory. Figure 5 compares how the free energy of OER changes at the Ni sites (Fig. 5a) and the Mn sites (Fig. 5b) of Ni₆MnO₈. Generally, OERs abide by the following four-electron step process:

$${\mathrm{OH}}^ - + ^\ast \to {\mathrm{HO}} ^\ast +\,{\mathrm{e}}^ -$$

(1)

$${\mathrm{HO}} ^\ast + {\mathrm{OH}}^ - \to {\mathrm{H}}_2{\mathrm{O}} + {\mathrm{O}} ^\ast +\,{\mathrm{e}}^ -$$

(2)

$${\mathrm{O}} ^\ast + {\mathrm{OH}}^ - \to {\mathrm{HOO}} ^\ast +\,{\mathrm{e}}^ -$$

(3)

$${\mathrm{HOO}} ^\ast + {\mathrm{OH}}^ - \to {\mathrm{O}}_2 ^\ast +\,{\mathrm{e}}^ -$$

(4)

Fig. 5

Free energy diagrams for the OER obtained at zero potential (U = 0), equilibrium potential (U = 1.23 V), and at the potential for which all steps are downhill on a Ni sites and b Mn sites; c the structures of the OER intermediates on the Ni sites and Mn sites

It is seen that all the elementary reactions are positive free energy change (endothermic) at zero electrode potential. As the electrode potential increases to the equilibrium potential at 1.23 V, some elementary reactions (e.g., OH* and O* formation) become energetically downhill (exothermic). All the elementary reactions become downhill until the electrode potential increases to 2.13 V for the Ni sites and 2.48 V for the Mn sites on the Ni₆MnO₈ surface. This fact suggests the overpotential is 0.90 and 1.25 V on the Ni sites and Mn sites. It can be concluded that the overpotential on the Ni sites is lower than that on the Mn sites, indicating that the Ni sites have superior OER activity and are the preferred reaction active sites. Figure 5c shows the corresponding structures of OER intermediates on the Ni sites and Mn sites of Ni₆MnO₈.

Notably, Ni₆MnO₈ alone cannot explain such a good OER activity of the Au-NWs/Ni₆MnO₈ catalyst, which arises from its poor electrical conductivity. Previous studies demonstrated that some hydroxides, including NiFeOOH, NiOOH, and CoOOH, have better OER activities on a Au substrate than on glassy carbon; this phenomenon was attributed to the partial electron transfer from hydroxides to more electronegative Au^49,50. It is therefore reasonable to infer that Au-NWs improve the electrical conductivity of the catalytic active centers in Ni₆MnO₈ due to the difference of the work functions and the strong interfacial coupling between Au-NWs and Ni₆MnO₈. The results demonstrate the advantages of the strongly coupled Au-NWs and Ni₆MnO₈ at boosting bifunctional oxygen electrocatalysis: Au-NWs bring both catalytic active sites for the ORR and a high electrical conductivity, which facilitates electron transport during electrocatalysis, whereas Ni₆MnO₈ is catalytically active for the OER due to its intrinsic OER property and improves mass transport by the 3D hierarchically porous structure. The stability of Au-NWs/Ni₆MnO₈ as a bifunctional catalyst was investigated for the ORR and the OER using the chronoamperometric method (Fig. 6a). Au-NWs/Ni₆MnO₈ shows better electrocatalytic stabilities during both the OER and ORR than those of the commercial Pt/C and RuO₂ catalysts after 30,000 s (Fig. 6b). More importantly, the Ni₆MnO₈-supported Au-NWs clearly improved the stability relative to that of the Au-NWs/C catalyst, which apparently arises from the preferable chemical stability and structural stability of Ni₆MnO₈. A favorable chemical stability would be beneficial to the electrochemical stability of catalysts⁴³. When cycling repeatedly at a wide potential range of ORRs and OERs for 100 cycles at 0.1 M KOH (Fig. S17), there was no observable degradation of the electrochemical signal associated with Ni₆MnO₈, demonstrating that Ni₆MnO₈ is electrochemically stable. Furthermore, 1D Au-NWs within the 3D hierarchically porous architecture impede agglomeration and dissolution of catalyst particles. The remarkable electrocatalytic activity and stability of Au-NWs/Ni₆MnO₈ make it highly promising as a bifunctional oxygen catalyst, as further indicated by the overall oxygen electrode curves (Fig. 6c, d). The oxygen electrode activity (ΔE = E_J10 – E_1/2) of Au-NWs/Ni₆MnO₈ noticeably present the smallest value (0.69 V) of all the provided catalysts and is even comparable with that of other reported bifunctional electrocatalysts (Table S5), confirming that the Au-NWs/Ni₆MnO₈ catalyst is one of the most efficient bifunctional catalysts.

Fig. 6

a Current-time chronoamperometric responses for the ORR at 0.75 V and the OER at 1.60 V (percentage of current retained vs. operation time); b bar plots of the degradation percentage of the catalysts for the ORR (top panel) and the OER (bottom panel) after the stability tests; c the overall polarization curves of Au-NWs/Ni₆MnO₈, Au-NW, Ni₆MnO₈, and Pt/C+RuO₂ catalysts; d the oxygen electrode activity was evaluated by the difference in the potential between the OER current density at 10 mA cm⁻² and the ORR current density at the half-wave potential (ΔE = E_J10 – E_1/2)

Rechargeable Zn–air batteries

Using the prominent bifunctional performance of the Au-NWs/Ni₆MnO₈ hybrid, a Zn–air battery was designed and assembled to evaluate its feasibility, as exhibited in Fig. S18. The open-circuit voltage (OCV) of the Au-NWs/Ni₆MnO₈-driven Zn–air battery is approximately 1.52 V (Fig. 7a), which is in the vicinity of the theoretical voltage (1.65 V). Two cells connected in series can work as an efficient energy supplier to power a red light-emitting diode (LED, 3.0 V). For comparison, a mixed Pt/C + RuO₂ catalyst (the mass ratio of Pt/C and RuO₂ was of 1/1) was also tested as a control air-cathode. Figure 7b shows the discharge polarization curves and the corresponding power density curves. Au-NWs/Ni₆MnO₈ exhibits a decent current density of 104.9 mA cm⁻² at 1.0 V, which is superior to that of the mixed Pt/C + RuO₂ catalyst (70.2 mA cm⁻²). The maximum power density of ca. 121 mW cm⁻² exceeds that of the battery with Pt/C + RuO₂ (110 mW cm⁻²) and many other reported bifunctional catalysts (Table S6). The specific capacity of the Au-NWs/Ni₆MnO₈-based battery was up to 760 mAh g_Zn⁻¹ at a current density of 5 mA cm⁻² (Fig. 7c), corresponding to an energy density of 996 Wh kg_Zn⁻¹, which surpasses those of Pt/C + RuO₂ (692 mAh g_Zn⁻¹; 892 Wh kg_Zn⁻¹) and previously reported state-of-the-art air-cathodes (Table S7). Moreover, the Au-NWs/Ni₆MnO₈-based battery demonstrated superior cycle stability compared to that of the mixed Pt/C + RuO₂ catalyst (Fig. 7d). For instance, the 1st cycle discharge–charge voltage gap is only 0.70 V, contributing to a high energy efficiency of 65%. After 80 cycles, both the voltage gap and the energy efficiency remain at high values, 1.02 V and 50.2% (Fig. 7e), respectively. Comparatively, the Pt/C + RuO₂-based battery has a poor cycle life with a large overpotential increase after 20 cycles (Fig. 7f). After the charge/discharge cycles, the overall morphology and structure of the Au-NWs/Ni₆MnO₈ cathode material were well preserved (Fig. S19). These results verify the excellent activity and stability of the Au-NWs/Au₆MnO₈ catalyst as the air-cathode in rechargeable Zn–air batteries.

Fig. 7

a Open-circuit plots, the inset shows the photograph of a red LED powered by two home-made Zn–air batteries in series with Au-NWs/Ni₆MnO₈ as the air-cathode; b discharge polarization and power density curves; c discharge curves of Zn–air batteries at 5 mA cm⁻²; d long-term cycling performance at the charging and discharging current density of 10 mA cm⁻². The enlarged 1st and last cycle discharge and charge voltage profiles of Zn–air batteries with e the Au-NWs/Au₆MnO₈ catalyst and f the Pt/C + RuO₂ catalyst

Discussion

In summary, we have developed a simple and efficient strategy for the synthesis of a strongly coupled Au-NWs/Ni₆MnO₈ material with both rigidity and flexibility as a new, low-cost, and highly active air-cathode for rechargeable Zn–air batteries. The α-naphthol–Au(III) complex plays a critical role in regulating the morphology of Au nanocrystals and synthesizing Ni₆MnO₈-supported Au nanowires. The synergistical integration of Ni₆MnO₈ that is active for OERs and Au nanowires that are active for ORRs endows the hybrid with a superior bifunctional electrocatalytic performance. The rigid Ni₆MnO₈ also makes the electrochemical properties more stable over the potential range of an air-cathode in an alkaline media, while the flexible Au nanowires offer effective electron transport channels as conductive bridges. The strong cross-link between the hierarchically porous Ni₆MnO₈ and Au nanowires provides an additional synergistic effect to further enhance electrocatalytic activity and stability, by virtue of its unique structural advantages. Moreover, the Au-NWs/Ni₆MnO₈-driven Zn–air battery can output a high OCV of 1.52 V, a high peak power density of 118 mW cm⁻², and a large specific capacity of 768 mAh g_Zn⁻¹. This work offers an innovative strategy for developing carbon-free bifunctional electrocatalysts that have many potential applications in metal–air batteries.