NiCo alloy‐anchored self‐supporting carbon foam as a bifunctional oxygen electrode for rechargeable and flexible Zn–air batteries

The design and fabrication of flexible, porous, conductive electrodes with customizable functions become the prime challenge in the development of new‐generation wearable electronics, especially for rechargeable batteries. Here, the NiCo bialloy particulate catalyst‐loaded self‐supporting carbon foam framework (NiCo@SCF) as a flexible electrode has been fabricated through one facile adsorption‐pyrolysis method using a commercial melamine foam. Compared with the electrode with Pt/C and Ir/C benchmark catalysts, the NiCo@SCF electrode exhibited superior bifunctional electrocatalytic performance in alkaline media with a half‐wave potential of 0.906 V for oxygen reduction reaction, an overpotential of 286 mV at j = 10 mA cm−2 for oxygen evolution reaction, and stable bifunctional performance with a small degradation after 20,000 voltammetric cycles. The as‐assembled aqueous zinc–air battery (ZAB) with NiCo@SCF as a self‐supporting air cathode demonstrated a high peak power density of 178.6 mW cm−2 at a current density of 10 mA cm−2 and a stable voltage gap of 0.94 V over a 540 h charge−discharge operation. Remarkably, the as‐assembled flexible solid‐state ZAB with self‐supporting NiCo@SCF as the air cathode presented an engaging peak power density of 80.1 mW cm−2 and excellent durability of 95 h undisrupted operation, showing promise for the design of wearable ZAB. The demonstrated electrode fabrication approach exemplifies a facile, large‐scale avenue toward functional electrodes, potentially extendable to other wearable electronics for broader applications.


| INTRODUCTION
In recent years, the booming market of wearable devices and portable electronic products has dramatically promoted the demand and development of flexible energy storage devices. 1 Among them, flexible solid-state rechargeable transition metal−air batteries have received extensive attention, mainly because of their ultrahigh theoretical energy density, safety in use, and pollution-free manufacturing. 2,3 However, the current energy conversion efficiency of metal−air batteries is extremely low, mainly due to the sluggish kinetics of oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) at the air electrode. Although the use of noble metal catalysts (such as Pt, Ru, Ir, etc.) can accelerate the electrochemical reaction kinetics of ORR and OER, the scarcity and high price of these noble metal catalysts will significantly increase the cost of metal−air batteries, which is not conducive to the development and large-scale practical application of metal−air batteries. Also, the current catalyst loading protocol relies on post immobilization approach in which the inherently strong attachment of the catalyst particles to the current collector is challenging. Moreover, it is an arduous task to formulate a three-dimensional conductive flexible framework that allows effective mass transport and reactant diffusion during physical deformation without multifunction. Therefore, the development of efficient and stable low-cost bifunctional oxygen electrodes is of great significance for the development of flexible metal−air batteries and has also become one of the core challenges. 4 Transition-metal alloy catalysts are featured with alterable electronic transitions with different valence states, excellent electrical conductivity, and versatile catalytic activities. In particular, it provides surface redox centers for O 2 adsorption and activation, which can facilitate ORR/ OER electrocatalytic turnovers. [5][6][7] Alloy nanoparticles alone tend to agglomerate, which leads to the reduction of catalytically active sites. Moreover, when working for a long time in strong acid or strong alkali environment, problems such as chemical corrosion and surface state deterioration result in degraded electrocatalytic activity. The dispersibility and stability of alloy nanoparticles can be improved by compounding with carbon supports with good chemical stability, and the activity of alloy nanocatalysts for ORR/ OER electrocatalysis can be promoted. 8,9 For example, coating carbon layers on the surface of alloy nanoparticles can be used as an efficient method for high-performance ORR/OER bifunctional electrocatalysts. The carbon layer prevents the alloy nanoparticles from coming into direct contact with the electrolyte and improves the structural integrity of the alloy catalyst in extended operation. Electrons can penetrate from the alloy core to the surface of the carbon layer, increasing the charge density on the surface of the carbon layer and thereby stimulating its catalytic activity. 10 In addition, heteroatom doping of carbon layers, especially nitrogen doping, can change the surface charge distribution of carbon materials and provide more catalytically active sites. 11 Melamine foam (MF), an elastic material with a three-dimensional self-supporting structure, provides a wide range of opportunities after surface modification and functionalizations. In particular, its conductivity and porosity can be manipulated as promising catalyst support by calcining at high temperatures. [12][13][14] In this work, NiCo alloy particles supported on elastic nitrogen-doped carbon foam were prepared by pyrolysis in an argon/hydrogen atmosphere, where the pyrolyzed precursors were uniformly anchored by a simple thermochemical method: nickel and cobalt ions were used as the metal source and melamine foam as the nitrogen and carbon sources. The resultant catalyst-loaded NiCo@SCF electrode with high surface area, catalyst loadings, and electrical conductivity exhibits excellent ORR/OER bifunctional electrocatalytic activity under alkaline conditions. More importantly, this self-supporting three-dimensional elastic melamine foam can be free-clipping and mechanically machined on demand. A rechargeable zinc-air battery (ZAB) and an all-solid-state flexible ZAB were designed and assembled using NiCo@SCF as the air cathode. The excellent electrical conductivity, high surface area, well protected, highly loaded catalyst nanostructure of the self-supporting electrode endow rechargeable ZAB with excellent energy efficiency, charge−discharge performance, and cycle stability. This work lays an avenue for the design and construction of cost-effective machinable cathodes for flexible ZAB or other wearable electronic technologies.

| Electrode fabrication
Cobalt nitrate and nickel nitrate were formulated into solutions with a concentration of 10 mg/mL. Twenty-five milliliters of each of the above solutions was taken, stirred, and mixed. The MF was put into the beaker of the mixed solution and stirred at a temperature of 90°C for 12 h. After the reaction was completed, the NiCo@MF was taken out, frozen, and dried. The NiCo@MF was pyrolyzed at 400°C in a tube furnace for 2 h under an Ar/H 2 atmosphere with a heating rate of 5°C per min and the content of H 2 was controlled at 10% of the total gas. After that, blackcolor NiCo@SCF was obtained. The SCF was obtained by the same process that MF was pyrolyzed at 400°C in a tube furnace for 2 h under an Ar/H 2 atmosphere with a heating rate of 5°C per min. The synthesis of single-metal@SCF electrodes was consistent with the above process, except that only cobalt nitrate or nickel nitrate was added as a metal ion solution in the preparation of the precursor.

| Material characterization
A field emission scanning electron microscope (JSM-7100F) at a working voltage of 15 kV was used to investigate the morphology of all as-prepared samples. Fourier transform infrared spectra (FTIR) were measured on a Bruker Vector-22 FTIR spectrometer with a scan range of 400-4000 cm −1 . The X-ray photoelectron spectra (XPS) were recorded on a Kratos Axis ULTRA system. Transmission electron microscopy (TEM) was conducted on Philips F20 at 200 kV. High-angle annular dark-field scanning transmission electron microscope (HAADF-STEM) imaging was taken by a probe-corrected JEOL JEM-ARM200F with an acceleration voltage under 200 kV. XRD patterns were obtained by a Shimadzu XRD-6000 diffractometer scanning the angle range from 10°to 80°using CuKɑ radiation (λ = 1.5418 Å). Brunauer−Emmett−Teller (BET) surface areas were estimated from nitrogen adsorption/desorption isotherms in powder forms using a surface area analyzer Autosorb iQ Station 1.

| Electrochemical measurements
All electrochemical measurements were finished in a traditional three-electrode cell using an electrochemical workstation (CHI 760D). A glassy carbon (GC) rotating disc electrode (RDE, AFE5T050GC, Pine Research) with an area of 0.196 cm 2 and graphite rod were chosen as the working electrode and counter electrode for ORR and OER measurements. To evaluate the reaction selectivity, a rotating ring disc electrode (RRDE, AFE7R9GCPT, Pine Research, OD disc = 5.61 mm, OD ring = 7.92 mm, ID ring = 6.25 mm) was used as the working electrode. Ag/AgCl and Hg/HgO served as reference electrodes for evaluating ORR and OER performances, respectively. The measurements were carried out in an O 2 -saturated 0.1 M KOH electrolyte for ORR and OER. Before recording linear sweep voltammetry (LSV) for ORR and OER at a rate of 2 mV s −1 , dozens of cyclic voltammetry (CV) scans were employed to achieve stable curves. The electrocatalyst ink was made via mixing catalyst powder (4 mg), 950 μL of isopropanol, and 50 μL of where J, J d , and J k represent the measured, diffusionlimiting, and kinetic current density, respectively. The ω is the electrode's rotational rate. B can be determined from the Levich equation as follows: B = 0.2 nFC o D 2/3 v −1/6 , where n represents the number of electrons, F is the Faraday constant (F = 96485 C mol −1 ), C o is the saturated O 2 concentration (1.21 × 10 −6 mol cm −3 ), D is the diffusion coefficient of O 2 in 0.1 M KOH (1.9 × 10 −5 cm 2 s −1 ), and ν is the kinetic viscosity (0.01 cm 2 s −1 ). The constant 0.2 is adopted when the rotation speed is expressed in rpm. RRDE measurements were tested by LSV at a scan rate of 5 mV s −1 at 1600 rpm, and the ring electrode voltage was kept at 1.5 V versus RHE. The H 2 O 2 collection coefficient (N) at the ring in RRDE experiments was 0.37. Below equations were used to calculate the apparent number of electrons transferred during ORR (n), and the percentage of peroxide released during ORR (H 2 O 2 %), based on the disk current (I D ) and ring current (I R ).  The electrochemical surface area (ECSA) of catalysts was obtained by the cyclic voltammetry curves in the non-faraday current region at the scan rates from 10 to DONG ET AL. | 3 of 13 60 mV s −1 . The relation between current and scan rates is shown in the following equation: where I c refers to the nonfaraday current; v is the scan rates.

| Assembly of Zn-air battery
The NiCo@SCF was machined to the appropriate thickness and size, as the cathode. For the Pt/C-Ir/C cathode, the electrocatalyst inks were prepared by ultrasonically dispersing 4 mg of Pt/C-Ir/C (1:1, mass ratio) into 1 mL of isopropanol/ The air electrode was fabricated by dropping electrocatalyst ink onto carbon paper and dried in the oven (an average catalyst loading of 1 mg cm −2 ). Then, this catalyst-coated carbon paper was further pressed with one commercial gas-diffusion layer (GDL) as the cathode. The Zn-air battery was assembled with the above cathode, one Zn foil as the anode, and 6 M KOH solution involving 0.2 M Zn(oAc) 2 as the electrolyte. The test of the assembled Zn-air battery was carried out in a CT2001A battery assessment system (Wuhan LAND Electronics Ltd.).

| Flexible solid-state zinc metal-air battery assembly
The air electrode is processed into a size of 6 × 2.5 × 0.2 cm. The 0.1 mm thick zinc flakes were polished and cleaned as metal anodes. The electrolyte is a hydrogel electrolyte. The specific production process is as follows: One gram of polyvinyl alcohol (PVA) with a molecular weight of 19,500 was dissolved in 10 mL of hot water (95°C), continued to heat, and stirred for 2 h. Then, 1 mL was dissolved with 0.20 M Zn(CH 3 COOH) 2 , 18.0 M KOH solution was added to the above PVA solution, and heating and stirring were continued. After 40 min, the resulting solution was placed in a −20°C freezer for gelation. After 12 h, it was taken out and thawed. The gel-thaw process was repeated three times to obtain the desired PVA gel electrolyte. Then, the air electrode with Nafion 117 suspension and the metal electrode were placed on both sides of the PVA gel electrolyte to obtain a flexible solid-state ZAB.

| Zn-air battery test
All Zn-air battery tests were performed under an ambient atmosphere at room temperature. The polarized profiles were recorded on the CHI 760E electrochemistry station using LSV with a scan rate of 5 mV s −1 calibrated with 95% iR-compensation. The discharge power density was calculated using the data from the discharge polarized profiles using the following equation: where P is the discharge power density, U d is the discharge voltage, and j d is the discharge current density. Galvanostatic discharge-charge cycling tests were carried out on a Land CT2001 multichannel battery tester. In each galvanostatic cycle, the Zn-air battery was discharged for 60 min and charged for 60 min at a given current density of 10.0 mA cm −2 . The rate performance of Zn-air batteries was evaluated by galvanostatic discharge at a series of the current density of 5, 10, 20, 50, 100, and back to 5 mA cm −2 . The Zn-air batteries went for several cycles of galvanostatic discharge-charge at 20.0 mA cm −2 . Galvanostatic discharge-charge cycling tests of FZAB were carried out on a Land CT2001 multichannel battery tester. In each galvanostatic cycle, the FZAB was discharged for 30 min and charged for 30 min at a given current density of 2.0 mA cm −2 .

| Electrode characterizations
The fabrication procedures of the NiCo@SCF catalysts are depicted in Figure 1A. The Ni 2+ and Co 2+ precursors were evenly adsorbed on the melamine foam framework by a facile immersion method at 90°C followed by lyophilization drying. The NiCo@SCF was obtained by pyrolysis of the foam under an Ar/H 2 mix atmosphere at 400°C. Supporting Information: Figure S1 shows the compressibility of the NiCo@SCF electrode, and the compressed electrode can perfectly recover from deformation. In Supporting Information: Figure S2, a 200 g weight was placed on only 0.87 g of NiCo@SCF to further testify the elasticity of the NiCo@SCF. These tests prove that the NiCo@SCF has strong elasticity and could be served as the electrode for flexible devices. The SEM images of NiCo@SCF are shown in Figure 1B,C. Notably, the three-dimensional structure of NiCo@SCF can be preserved after metal loading and pyrolysis. NiCo alloys are uniformly and abundantly distributed on individual polygonal fibers and the triangular junction of the NiCo@SCF framework ( Figure 1D). A similar foam structure has been preserved when no metal is loaded onto the foam before pyrolysis (which is noted as SCF hereafter) (Supporting Information: Figure S3).
The HR-TEM image ( Figure 1E) shows that the metal alloy nanoparticles are coated in a small number of carbon layers, where a large number of defect structures and edge structures are present. The HR-TEM image of the particles witnesses the crystal planes with interplanar spacings of 0.208 nm, which correspond to the (111) plane of metallic NiCo, 15 strongly suggesting a nickel-cobalt alloy crystal structure. In addition, curved fringes with a lattice parameter of 0.417 nm are measured for the carbon layer, which can be attributed to the presence of curved graphdiyne sheets. 16,17 The element distribution in the catalyst was characterized by the corresponding HAADF-STEM EDS spectrum, and the results are shown in Figure 1F. Ni, Co, C, and N elements in the catalyst are uniformly distributed throughout the carbon foam, and the distribution of Ni/Co elements does not overlap with the distribution of C and N elements, indicating that the nickel-cobalt alloy exists in the form of a structure embedded in the NiCo@SCF. From the distribution of EDS in HAADF-STEM in Figure 1F and Supporting Information: Figure S4, it can be concluded that the atomic ratio of element content of cobalt and nickel in the alloy is 53:47. The characteristic peaks of the Raman spectrum of NiCo@SCF are shown in Supporting Information: Figure S5, in which the peak at 1388 cm −1 corresponds to the breathing vibration (D band) of the sp 2hybridized carbon atoms of the benzene ring, and the peak at 1576 cm −1 corresponds to the in-plane stretching vibration of the sp2-hybridized carbon atom of the benzene ring (G band). The intensity ratio of D peak to G peak is 0.72, indicating that the carbon has an ordered structure. The peak at 1943 cm −1 is generally believed to be related to the diyne bond and the peak at 2194 cm −1 corresponds to the stretching vibration of the graphdiyne conjugated diyne, which is a characteristic peak in the graphdiyne structure. 17,18 However, it is difficult to quantify its proportion over the carbon structure due to its defective nature. As the XRD data of NiCo@SCF shown in Figure 2A and Supporting Information: Figure S6, the XRD patterns of prominent diffraction peaks at 44.4°, 51.7°, and 76.1°correspond to the (111), (200), and (220) planes of the facecentered cubic (fcc) nickel−cobalt alloy, respectively. 19 The peak positions corresponding to the above three diffraction peaks are slightly higher than those of pure fcc phase element Co (PDF # 15-0806) but slightly lower than those of pure fcc phase element Ni (PDF # 04-0850) (Supporting Information: Figure S6), which strongly certifies the formation of Ni−Co solid solution during the pyrolysis and also a certain degree of lattice distortion. 20 The broad peak at 21.3°corresponds to the interlayer spacing of 4.17 Å, indicating a defective carbon with the presence of graphdiyne structure. 16,21 Although the existence of the NiCo@SCF framework is confirmed, its crystallinity is poor due to the conformational fluctuations of NiCo@SCF at the mesoscale. In both FT-IR spectra of Supporting Information: Figure  Co, N, C, and O on the surface (Supporting Information: Figure S8). Detailed analysis of different elements as shown in Figure 2B is the high-resolution Ni 2p spectrum. Among them, the peaks at 855.9 and 873.7 eV correspond to metallic elemental Ni. Two satellite peaks centered at 864.1 and 883.9 eV were observed in the Ni 2p XPS spectrum of NiCo@SCF. 23 Figure 2C shows a high-resolution Co 2p XPS spectrum, with peak splitting results similar to Ni 2p. The peaks at 777.8, 779.5, 782.6, and 786.2 eV correspond to metal elemental Co, Co 3+ , high-spin Co 2+ ion vibration peaks, and satellite peaks, respectively, in which the metallic Co species occupies dominates the Co 2p spectrum. 20,24 The above results echo the crystal structure characterizations, further confirming the formation of NiCo alloy structure. Figure 2E shows the high-resolution C ls XPS spectra of NiCo@SCF, which can be fitted into five peaks. The peaks at 284.7, 285.6, 286.5, 287.6, and 289.3 eV correspond to sp 2 C-C, sp C-C, C-O, C═N, and C═O, 21,25 confirming the existence of N-doped carbon material structures. The peaks at 398.8, 399.9, 400.9, and 404.5 eV of high-resolution N 1 s XPS spectra in Figure 2F correspond to pyridine nitrogen, pyrrolic nitrogen, graphitic nitrogen, and N-oxide, respectively. 26,27 It can be seen that the nitrogen in NiCo@SCF samples is mainly composed of pyridine nitrogen and graphitic nitrogen. The pyridinic nitrogen dopants in graphdiynebased carbon materials could contribute to the ORR/OER electrocatalytic activity of the electrode. 21,[28][29][30][31] Generally, the specific surface area of the catalyst determines the number of exposed catalytic active sites, and a high specific surface area is an essential factor affecting the catalytic performance. Therefore, the specific surface area and pore size distribution of the samples were analyzed. Figure 2D shows the N 2 adsorption−desorption test curves of SCF and NiCo@SCF. The N 2 adsorption-desorption isotherms of NiCo@SCF are of type II and type IV, and there is an H4-type hysteresis loop. The multipoint Brunauer−Emmett−Teller (BET) method was used to calculate the specific surface area of SCF and NiCo@SCF, which are 104.5 and 271.2 m 2 g −1 , respectively. The higher specific surface area of the NiCo@SCF electrode is beneficial to expose more electrocatalytic active sites, which can improve its catalytic activity. From the analysis of the pore size distribution curve (Supporting Information: Figure S9), it can be seen that compared with SCF, the NiCo@SCF sample has sharp pore distribution peaks at 10-13 nm, indicating that the NiCo@SCF has mesopore structure besides macropores. Hierarchical pores promote mass transportation, gas diffusion, and the contact of catalytic active sites with the electrolyte, which facilitates the transport of catalytically relevant protons and enhances the electrocatalytic activity. 32,33

| Electrochemical performance
The electrocatalytic performance of as-fabricated Ni-Co@SCF for ORR was tested in alkaline media (0.1 M KOH). As shown in the LSV curves under alkaline conditions in Figure 3A, the initial onset potential (E 0 ) and half-wave potential (E 1/2 ) of NiCo@SCF are 1.01 and 0.906 V, respectively, much better than SCF (E 0 = 0.861 V, E 1/2 = 0.61 V) and also exceed those of the electrode with commercial Pt/C catalysts (E 0 = 0.985 V, E 1/2 = 0.878 V). As shown in Figure 3B, the Tafel slopes of SCF, NiCo@SCF, and electrodes with 20 wt.% Pt/C catalysts were measured to be 375.6, 74.4, and 83.1 mVdec −1 , respectively. NiCo@SCF has the smallest Tafel slope, which is due to the highest limiting current density of NiCo@SCF, further demonstrating the fast oxygen diffusion. The much more positive E 0 and E 1/2 as well as the larger Tafel slopes of SCF in comparison to the NiCo@SCF counterpart have strongly elucidated that the NiCo alloy particles act as the dominant active sites for the ORR on the NiCo@SCF sample. The rotating ring disk electrode (RRDE) test technique was used to further explore the H 2 O 2 yield and electron transfer number (n) of different samples during the ORR process. As shown in Figure 3C, in the potential range of 0.2-0.65 V versus RHE, the H 2 O 2 yield of the sample NiCo@SCF is lower than 3.3%, indicating a dominant four-electron transfer ORR process. The calculated n is about 3.88, which is even higher than the electrode with a commercial Pt/C catalyst (3.76). The electron transfer number calculated by the K-L equation responding to LSV curves of NiCo@SCF in an O 2 -saturated solution of 0.1 M KOH at different rotation speeds is 3.87 (Supporting Information: Figures S10 and S11), which is similar to the result obtained by the above calculation method, indicating that the reaction is an efficient four-electron transfer process.
The OER catalytic performance of different samples in alkaline media is also evaluated. As the LSV curves shown in Figure 3D, at a current density of 10 mA cm −2 (E j=10 ), the OER overpotential of NiCo@SCF is only 286 mV. Under the same test conditions, the overpotential of NiCo@SCF is 44 mV lower than that of the commercial Ir/C catalyst, indicating that the NiCo@SCF sample has excellent OER electrocatalytic activity. The OER reaction kinetics of different samples were evaluated by the slope of the Tafel plots. It can be seen from Figure 3E that the sample NiCo@SCF has the smallest Tafel slope, indicating excellent OER catalytic kinetics, which originated from the abundant catalytic active sites, high specific surface area, and better electron transport synergy of the NiCo@SCF samples. The poor OER activity of the SCF electrode has been confirmed by the high overpotential (765 mV) and high Tafel slope (146 mV decade −1 ). The difference (ΔE) between the half-wave potential of ORR and the corresponding potential of OER at a current density of 10 mA cm −2 is usually used to evaluate whether a catalyst has excellent ORR/OER bifunctional catalytic activity. The smaller ΔE indicates that the catalyst has excellent ORR/OER bifunctional electrocatalytic activity in the same electrolyte. Therefore, the ORR/OER bifunctional LSV curves of NiCo@SCF in the O 2 -saturated 0.1 M KOH electrolyte were tested and the results are shown in Figure 3F. The ΔE value of NiCo@SCF is 0.61 V versus RHE. After 20,000 cycles, ΔE still maintains a small value (0.631 V). It can be seen from the LSV curves that the performance attenuation of ORR and OER is very small, the half-wave potential of ORR is reduced by 9 mV, and the E j=10 of OER is attenuated by 12 mV. Chronoamperometric responses for ORR of NiCo@SCF and Pt/C in O 2saturated 0.1 M KOH solution at 0.6 V, OER of NiCo@SCF and Ir/C in 0.1 M KOH solution at 1.55 V are shown in Supporting Information: Figure S14, respectively. After 50,000 s, the relative current density of NiCo@SCF remained at 91% and 89% of its initial values for ORR and OER, respectively, while the Pt/C counterpart maintained at 80% and Ir/C maintained at 76% of the initial values. This result can fully prove that the bifunctional electrode has excellent stability, which could probably be linked to the carbon shell on the NiCo alloy catalyst surface.
The ECSA of the NiCo@SCF sample was assessed by calculating the C dl values (Supporting Information: Figure S12). The NiCo@SCF sample exhibits a higher C dl (10.9 mF cm −2 ) than that of Pt/C-Ir/C (9.7 mF cm −2 ), indicating that the abundant heterostructures, oxygen vacancies, and unique porous structure of nanoparticles increase the ECSA and create more active sites. Since the robustness of the NiCo@SCF electrode material gradually deteriorated when the temperature was above 600°C, the NiCo@SCF electrode materials were not suitable as an electrode for zinc−air batteries (whether it was an aqueous system or a solid-state system ZAB), so comparing the ORR performance of samples pyrolyzed from 300°C to 600°C (as shown in Supporting Information: Figure S13A), the halfwave potentials of NiCo@SCF-300, NiCo@SCF-400, NiCo@SCF-500, and NiCo@SCF-600 were 0.71, 0.906, 0.86, and 0.898 V respectively. There was no obvious difference in the ORR performance of the NiCo@SCF pyrolyzed at 400°C-600°C compared with the half-wave potential; however, the NiCo@SCF electrode material pyrolyzed at 400°C showed better kinetic properties. Moreover, the mass activities of four samples are shown in Supporting Information: Figure S13B; the NiCo@SCF-400 has the highest mass activity among them, which shows that a certain degree of carbonization is beneficial to improve electrical conductivity, thereby promoting catalytic performance.
The roles of elemental and compositional information of metals in the composite catalysts were investigated. As the XRD data of Co@SCF, Ni@SCF, and NiCo@SCF shown in Supporting Information: Figure S15, compared to the three main peaks of NiCo@SCF, the peak positions of Co@SCF are slightly lower, indicating that those peaks correspond to pure fcc phase element Co (PDF # 15-0806). On the contrary, peaks of Ni@SCF are slightly higher than those of NiCo@SCF, corresponding to pure fcc phase element Ni (PDF # 04-0850), certifying the formation of Ni-Co solid solution during the pyrolysis and also a certain degree of lattice distortion of NiCo@SCF. The SEM images of Co@SCF are shown in Supporting Information: Figure S16. Notably, there are no obvious alloy particles like the sample NiCo@SCF on the carbon skeleton, and insufficient loadings may lead to poor bifunctional electrocatalytic performance. This can be seen in Figure S17 that the particle distribution of cobalt elements is extremely rare. At the same time, it can be seen from Supporting Information: Figures S18 and S19 that in the synthesized Ni@SCF, there are a large number of aggregated metal particles on the carbon skeleton. This aggregation phenomenon is not conducive to the exposure of catalytic sites and also leads to deficient catalytic performance. In Supporting Information: Figure S20, the high-resolution Co 2p spectrum of Co@SCF is shown. The peaks at 777.8, 780.9, 784.8, and 788.6 eV correspond to metal elemental Co, Co 3+ , high-spin Co 2+ ion vibration peaks, and satellite peaks, respectively, in which the metallic Co species dominates the Co 2p spectrum. Among the peaks of Ni 2p of Ni@SCF in Supporting Information: Figure S21, the peaks at 855.6 and 874.0 eV correspond to metallic elemental Ni. The contents of different metal valence states corresponding to different XPS subpeaks of three samples are shown in Supporting Information: Table S1. Compared with the sample Co@SCF, the Co 2+ of the sample NiCo@SCF was much higher, and Co 2+ was revealed to be beneficial to enhance ORR/OER activity. 34 Compared to sample Ni@SCF, the content of Ni 2+ was also significantly increased, indicating that Ni 2+ occupies a large part of the sample surface, which is conducive to OER activity. 35 And the metallic states of Co and Ni were found in the XPS spectra; the occurrence of electron redistribution after NiCo@SCF vacancy configuration, which may actively tune the energy band and electronic structure of nanomaterials, and the electron redistribution may lead to charge transfer from NiCo to N-doped carbon to form electron-rich N-doped carbon and electron-deficient NiCo species. The electrocatalytic performance of as-fabricated Co@SCF, Ni@SCF, and NiCo@SCF for ORR was tested in alkaline media (0.1 M KOH). As shown in the LSV curves under alkaline conditions in Supporting Information: Figure S22A, the half-wave potential (E 1/2 ) of Co@SCF and Ni@SCF is 0.822 and 0.84 V, respectively, much worse than NiCo@SCF (E 1/2 = 0.906 V). The OER catalytic performance of different samples in alkaline media is also evaluated. As the LSV curves shown in Supporting Information: Figure S22B, at a current density of 10 mA cm −2 (E j=10 ), the OER overpotentials of Co@SCF and Ni@SCF are 455 and 461 mV, much higher than that of the overpotential of NiCo@SCF (286 mV), indicating that the NiCo@SCF sample has the most excellent OER electrocatalytic activity.
The excellent ORR/OER bifunctional electrocatalytic activity of NiCo@SCF samples in the same electrolyte is attributed to the following points: (1) Ni and Co elements in the alloy undergo electronic transitions between different valence states, providing redox centers for the adsorption and activation of O 2 , and enhance the electrocatalytic activity of ORR; [36][37][38] (2) NiCo@SCF has a large specific surface area, which is beneficial to reduce agglomeration of nanoparticles, expose more active sites, and facilitate the electrocatalysis-related proton transport; (3) The wellpreserved conductive network of carbon promotes the rapid transport of electrons, and a large amount of nitrogen doping can increase the surface charge density, thereby enhancing the electrocatalytic activity of NiCo@SCF. [39][40][41][42] The catalytic reaction kinetics are improved, resulting in higher ORR/OER electrocatalytic activity.

| Battery performance
The above electrocatalytic performance tests certified that the NiCo@SCF material exhibited excellent bifunctional ORR/OER electrocatalytic activity. To further evaluate the performance of electrocatalysts in practical energy device tests, in this work, the NiCo@SCF was machined into a rectangular shape suitable as a self-supporting cathode for an aqueous rechargeable ZAB (Supporting Information: Figure S23). As shown in Figure 4A, the LSV curves and corresponding power density curves of aqueous ZAB cells are prepared with NiCo@SCF and Pt/C-Ir/C@carbon paper, respectively. It can be observed that the peak power density (178.6 mW cm −2 ) of NiCo@SCF-assembled ZAB is more than double of the Pt-Ir/C-assembled ZAB (87.3 mW cm −2 ). The NiCo@SCF-assembled ZAB was subjected to discharge tests at different current densities, and the results are shown in Figure 4B. After two 10 h cycles, the discharge process with the maximum current density reaching 100 mA cm −2 , when the current density returns to 5 mA cm −2 of the third cycle, compared with the first cycle, the discharge platform remains at 1.227 V, which is 98.8% of the initial value. Even when comparing the first cycle with the third cycle when the current density is under 50 mA cm −2 , the discharge platform can be maintained at 1.065 V, which is 95.7% of the initial value. The specific capacity calculated by normalizing the mass of consumed Zn by performing constant voltage discharge tests at potential platforms corresponding to current densities of 20 mA cm −2 for a long time until the zinc flakes are exhausted ( Figure 4C), which was 755 mAh g Zn −1 , indicating that the ZAB assembled with NiCo@SCF has excellent discharge stability and discharge efficiency, which originated from the efficient and stable ORR electrocatalytic performance of NiCo@SCF. The structural stability of the NiCo@SCF as an air electrode was further assessed in a long-term charge and discharge test, as shown in Figure 4D,E. After 540 h of charge-discharge (current density of 10 mA cm −2 ), the charge and discharge voltage gap of the NiCo@SCF assembled ZAB finally reached 0.94 V, and the charge-discharge plateau hardly changed. This stability exceeds that of most aqueous ZAB assembled from NiCo alloy-based catalysts (Supporting Information: Tables S2  and S3). Considering the excellent performance and flexibility feature of the NiCo@SCF, it was further assembled into flexible solid-state ZAB (FSZAB) as an emerging technology for flexible energy storage applications ( Figure 4F). As shown in Figure 4G, the resultant FSZAB with a self-supporting NiCo@SCF cathode is capable of recharge under a stable charge and discharge voltage gap of 0.805 V, which can be bent to different angles after prolonged cycling. Compared with some recently published alloy-based bifunctional electrocatalysts (Supporting Information: Tables S4 and S5), the NiCo@SCF-assembled FSZAB exhibited a relatively high power density value of 80.1 mW cm −2 and displayed a stable charge/discharge voltage gap of 0.8 V, which was 95 h more than that at the current density of 2 mA cm −2 ( Figure 4H,I). These results have verified that NiCo@SCF catalysts are promising self-supporting electrodes in rechargeable FSZAB and wearable devices, which will pave the way to more economically viable, cost-effective, and durable smart energy systems.
To further investigate the structural robustness, the spent NiCo@SCF was retrieved from the FSZAB after a 95 h charge/discharge stability test and subjected to structural characterization. The XRD data in Supporting Information: Figure S24 shows that the three main peaks that indicate the crystal face of the NiCo alloy still maintain the original spectrum, indicating that the overall structure has been effectively preserved. The flexible framework of NiCo@SCF remained intact as confirmed in the SEM image (Supporting Information: Figure S25), and the NiCo alloy particles on its framework were also observed in TEM (Supporting Information: Figure S26A,B). In particular, the grain size of NiCo alloy particles after the charge/discharge stability test was reduced probably due to the electrochemical reconstruction, as shown in the HRTEM image (Supporting Information: Figure S26C,D). In addition, the uniform distribution of NiCo-alloy particles on the SCF framework is evident in the EDS mapping results (Supporting Information: Figure S27). Supporting Information: Figure S28 shows the XPS high-resolution spectrum of the sample after the cycle test. Supporting Information: Figure S28A,B correspond to the Ni 2p and Co 2p XPS spectra, respectively, and it is evident that there are amounts of Ni and Co elements representing the valence state of the alloy in the subpeaks. As the C ls XPS spectrum shown in Supporting Information: Figure S28C, the subpeaks from 284 to 290 eV are consistent with the five subpeaks as aforementioned and shown in Figure 2E. Compared with the peak in Figure 2F, the relative content of pyrrolic-N and graphitic-N decreases, while the content of pyridinic-N increases relatively in Supporting Information: Figure S28D, which is also consistent with the literature mentioned above that the performance of ORR/OER is determined by pyridinic-N in nitrogen doping electrocatalysts, which is one of the main factors. The above results strongly indicate that NiCo@SCF is endowed with outstanding structural robustness and high cycling stability that is essential to the delivery of efficient and steady ORR/OER electrocatalytic performance. The robust, efficient NiCo@SCF herein exemplifies a useful electrode in rechargeable FSZAB and next-generation wearable electronic devices.

| CONCLUSION
In conclusion, NiCo bialloy particles have been successfully anchored onto the carbonized MF framework (NiCo@SCF) via a facile thermal conversion approach using commercially available low-cost melamine foam. The resultant selfsupporting, machinable, porous NiCo@SCF electrode showed excellent ORR/OER bifunctional activity, providing a new idea for designing other non-precious-metal-based electrodes. More importantly, the rechargeable Zn-air battery assembled with the NiCo@SCF electrode manifested an appreciable charge-discharge performance with a steady voltage gap. The unique self-supporting NiCo@SCF electrode will have many potential applications, providing more possibilities for the design and manufacturing of massproducible, cost-effective, and wearable-type Zn-air batteries. The demonstrated strategy can be readily extended to other wearable devices for wider applications.