N-Doped carbon supported Co₃O₄ nanoparticles as an advanced electrocatalyst for the oxygen reduction reaction in Al–air batteries

Abstract

Low-cost and high-performance catalysts are highly desirable for the oxygen reduction reaction (ORR) in metal–air batteries. Herein, a Co₃O₄/N-doped Ketjenblack (Co₃O₄/N-KB) composite is proposed as a high performance catalyst for Al–air batteries. The synergistic effect between Co₃O₄ and N-KB enables the Co₃O₄/N-KB composite to have a much higher cathodic current, a much more positive half-wave potential and more electron transfer in comparison with Co₃O₄ or N-KB alone. The Co₃O₄/N-KB composite favors a direct four electron pathway in the ORR process, and its ORR current density even outperforms the commercial Pt/C (20 wt%). Al–air batteries using the as-prepared catalyst in the air electrode were constructed, which displayed a high discharge voltage plateau of ∼1.52 V, comparable to that of the commercial Pt/C. The developed Co₃O₄/N-KB composite here could meet the requirements of large-scale application of metal–air batteries due to the much cheaper cobalt source and more economically commercialized carbon source compared to the commercial Pt/C.

---

Introduction

To meet the demands of modern society, energy conversion and storage systems with low-cost, high performance and environmental benignity are greatly needed.^1–3 The performance of these devices depends on the properties of their electrode materials.⁴ The current bottleneck of fuel cells and metal–air batteries lies in the sluggish ORR on the cathode side.^5,6 Despite tremendous effort, developing catalytic materials for the ORR with high activity remains a great challenge.^5,7 Although the precious metal Pt has been considered as the most promising catalyst, the scarce resources and high cost limit its long-term practical applications.^8,9 Accordingly, it is very important to develop non-precious metal catalysts with high electrocatalytic activities for ORR.^10,11

Surface chemistry and structure of the support materials can greatly influence the activity of the resultant catalysts.^12,13 Recently, catalyst supports such as Ketjenblack (KB), BP2000, and others carbon materials have been widely used in ORR because of their good electrical conductivity, surface properties, high surface area and stability.¹⁴ Note that the catalytic activities of these kinds of carbon supports are far insufficient for ORR. Nitrogen-doped carbon materials could affect the electron transportation in electrodes and increase active sites for ORR, leading to improved catalytic activity.¹⁵ These properties have been attributed to electronic or structural changes caused by the nitrogen incorporation into carbon framework.¹⁶ Although the compositions and structures of the active sites remain unclear, it has been well known that the doped nitrogen atoms (such as pyridine-like, pyrrole-like, graphite-like, and quaternary nitrogen atoms, etc.) play a decisive role for ORR.¹⁷ Nam et al.¹⁸ reported metal-free Ketjenblack incorporated nitrogen-doped carbon sheets for efficient catalysts of ORR in alkaline solution. They considered that the N dopant sites could provide positive catalytic effect for ORR. The nitrogen-doped gelatin with Ketjenblack carbon composites (GK) could adsorb more oxygen molecules owing to more exposed edge sites of GK sheets, where the oxygen adsorption was more favorable than that in basal plane. Lee et al.¹⁹ synthesized the carbonized melamine foam (Fe/Fe₃C–melamine/N-KB) with higher positive onset potential than pure Ketjenblack. It is reasonable to infer that C–N group showed higher performance for ORR.

Recently, low cost transition metal oxides (such as manganese oxides, iron oxides, cobalt oxides, and nickel oxides, etc.) with good catalytic abilities towards ORR have been considered as effective components for composite electrocatalysts.²⁰ Cobalt oxides have been investigated as non-precious ORR catalysts and applied in ORR in terms of their low cost, low electrical resistance and environmental friendship. The active sites of Co³⁺ ions in Co₃O₄ play a determinant role in the performance for ORR.^21,22 Li et al.²³ reported a series of M-doped polypyrrole-modified BP2000 via the hydrothermal method, the synergy among Co, N, and C effectively enhances the performance of ORR in alkaline medium, which depends on the nature of the transition element and the corresponding anion. Cobalt oxide has been well known as a promising electrocatalyst for metal–air batteries such as Li–air and Al–air batteries.^9,24,25 Note that the natural abundance of aluminium in the earth's crust is much higher and is non-toxic and environmental friendly.²⁶ In addition, aluminum has highly negative standard electrode potential [E⁰ = −1.7 V vs. standard hydrogen electrode (SHE)].²⁷ Therefore, Al–air battery has been considered as a promising energy storage and conversion device because of its high practical energy density (600–800 W h kg⁻¹).^28–30 Phinergy and Alcoa developed an electric car using this kind of battery, which could run more than 1600 km by only adding water several times.³¹ The success of this technology has drawn intensive attention.^32,33 In short, aluminum–air batteries turn out to be the more efficient energy systems for electric vehicles. One of the critical challenges is to create a stable and highly efficient air electrode under certain discharge current density.

In view of its low resistance, high mesoporous area and porosity, surface clean, KB has been employed as a promising catalyst support in alkaline fuel cell.^25,34 In our recent work, we reported a MnO_x–CeO₂/KB hybrid, which exhibited a comparable activity and better stability towards ORR in comparison with the commercial Pt/C.³⁵ Here, we reported a composite of Co₃O₄ nanoparticles grown on N-KB as a high-performance catalyst for ORR. Although KB alone has little ORR activity, N-KB exhibits surprisingly higher performance in ORR in alkaline solutions. Synergetic effect between Co₃O₄ and N-KB enables the as-prepared composite superior catalytic activity towards ORR. The high performance Co₃O₄/N-KB with low cost shows a good potential for large-scale commercial applications in metal air batteries.

Experimental section

Synthesis of Co₃O₄/N-KB

2.0 g of KB (Ketjenblack carbon, EC-300J) was treated in HNO₃ at 80 °C for 8 h to remove other impurities and introduce functional groups on the carbon surface. After that, the mixture was centrifuged at 8000 rpm for 5 min to remove the high concentration of HNO₃ at least repeated 10 times until the solution became neutral, and then dried in air at 110 °C. For nitrogen doped KB (N-KB), 1 g of acid-treated KB and 3.0 g of urea were added into agate mortar and firstly ground for 10–15 min then heated at 700 °C for 2 h under an argon atmosphere at a rate of 5 °C min⁻¹. Co₃O₄/N-KB composite was prepared by a simple hydrothermal method. Briefly, a certain quantity of Co(NO₃)₂·6H₂O was first dissolved into 80 mL distilled water, then 0.3 g of N-KB was added into the Co(NO₃)₂ solution. This suspension was stirred for at least 10 min and then 0.0668 g of NH₄HCO₃ was added. Finally, the mixed suspension was rapidly transferred into a 100 mL Teflon-lined autoclave and then kept at 140 °C for 12 h. The as-prepared Co₃O₄/N-KB was filtered and dried in air at 110 °C. After grinding in agate mortar, the samples were heated at 300 °C for 2 h in a muffle furnace with a slow heating rate of 5 °C min⁻¹, finally Co₃O₄/N-KB was obtained. For comparison, Co₃O₄ was synthesized by the similar method as Co₃O₄/N-KB composite but without adding N-KB. Co₃O₄/KB was prepared through the similar procedure as Co₃O₄/N-KB using KB instead of N-KB. Control experiments with different contents of Co₃O₄ (5 wt%, 10 wt%, 15 wt%, 20 wt%) in Co₃O₄/N-KB composite were also performed to optimize the catalytic activity of the composite.

Material characterizations

X-ray diffraction (XRD) patterns of as-prepared samples were recorded by X-ray diffractometer (Rigaku D/Max 2500) utilizing a CuKα source with a step of 0.02°. X-ray photoelectron spectroscopys (XPS, K-Alpha1063) were applied to measure the chemical compositions of the samples. Scanning electron microscope (SEM) images of as-prepared Co₃O₄/N-KB were conducted using a Nova NanoSEM 230 SEM. Transmission electron microscope (TEM), high resolution TEM (HRTEM) images, scanning TEM (STEM) and elemental mapping of as-prepared Co₃O₄/N-KB were obtained using a FEI Tecnai G2 F20 S-TWIX TEM. The Raman spectra were performed using a Jasco Laser Raman Spectrophotometer NRS-3000 Series, with an excitation laser wavelength of 532 nm, at a power density of 2.9 mW cm⁻². Differential scanning calorimetry and thermogravimetric analysis (DSC/TGA) was performed on a STA 449C with a heating rate of 10 °C min⁻¹ from 30 to 800 °C in air.

Electrode preparation and electrochemical test

For the rotating disk electrode (RDE) measurements, 4 mg of as-prepared catalyst was dissolved in 1900 μL ethanol and sonicated 20 min, then 100 μL of Nafion (5 wt%) was added into the above solution and sonicated 20 min again to form a homogeneous ink. Then 10 μL of the catalyst ink was loaded onto on a glassy carbon rotating disk electrode of 5 mm in diameter. Cyclic voltammetry (CV) and linear sweep voltammetry (LSV) were conducted using saturated calomel electrode (SCE) as the reference electrode, a platinum wire as the counter electrode and the sample modified glassy carbon electrode as the working electrode. Electrolyte was 0.1 M KOH solution, oxygen (99.999%) was accessed for 20 min to make the electrolyte saturated with oxygen. All potentials initially in this study measured by the SCE were calibrated with respect to reversible hydrogen electrode (RHE). LSV were measured at a scan rate of 10 mV s⁻¹ with varying rotating rates of 400, 625, 900, 1225, and 1600 rpm. CV were measured at a scan rate of 50 mV s⁻¹. The obtained current data were normalized to the area of the catalyst used (i.e. given in mA cm⁻²) for direct comparison between samples. Koutecky–Levich plots (j⁻¹ vs. ω^−1/2) were analyzed at various electrode potentials. The slopes of their linear fit lines were used to calculate the electron transfer number (n) on the basis of the Koutecky–Levich equation. The number of electrons transferred per oxygen molecule was calculated by the Koutecky–Levich equation given below:^36–38

B = 0.62nFC₀D₀^2/3ν^−1/6

For the Tafel plot, the kinetic current was calculated as follows:

where j the measured current density of the experiment, j_L is the diffusion-limiting current density, j_K is the kinetic current density, ω is the electrode rotating speed in rpm, F is the Faraday constant (F = 96 485 C mol⁻¹), C₀ the bulk concentration of O₂ (1.2 × 10⁻³ mol L⁻¹), D₀ the diffusion coefficient of O₂ in 0.1 M KOH (1.9 × 10⁻⁵ cm² S⁻¹), and ν is the kinematic viscosity of the electrolyte (0.01 cm² S⁻¹). B is the slope of K–L plots, n represents the number of electrons transferred per oxygen molecule.

Rotating ring-disk electrode (RRDE) technique was used to further estimate the electron transfer number (n). The peroxide percentage and the transferred electron number (n) were calculated based on the following equations:¹⁶

where I_d is disk current, I_r is ring current, and N is current collection efficiency of the Pt ring (0.37).

Al–air battery fabrication and testing

Al–air full batteries in home-made model were built to investigate the practical electrochemical activities of the as-prepared catalysts. It consists of a polished aluminum electrode and an air electrode as the anode and cathode, respectively. 6 mol L⁻¹ KOH adding 0.01 mol L⁻¹ Na₂SnO₃, 0.0005 mol L⁻¹ In(OH)₃, 0.0075 mol L⁻¹ ZnO as corrosion inhibitors was used as the electrolyte. The air electrodes with three layered structure, consisting of catalytic layer, current collector (nickel foam), and gas diffusion layer was fabricated using the hot pressure method. The catalytic layers were fabricated as follows: catalysts, Ketjenblack carbon and acetylene black, polytetrafluoroethylene (PTFE) were mixed well in a weight ratio of 3 : 3 : 1 : 3, then agitated for 30 min to form a homogeneous slurry. When the above slurry turned into paste, it was rolled to the layer until the thickness was ∼0.2 mm.³⁹ The thickness of this air electrode was pressed to a range of 0.4–0.6 mm (3 cm × 5 cm) at 15 MPa. Finally, the air electrode was dried in vacuum at 60 °C for 12 h. The home-built electrochemical cells was used for Al–air battery measurements on a Neware battery testing system (Shenzhen, China). The details have been given in our previous work.⁴⁰

Results and discussion

Fig. 1a shows typical XRD patterns of the as-prepared N-KB and Co₃O₄/N-KB composite. A broad peak (2θ = 24°) can be observed in the XRD pattern of N-KB, which is related to the characteristic peak of graphitic carbon. The tiny broad peak (2θ = 45°) is indexed to an amorphous structure;^19,41 whereas, for what concerns Co₃O₄, the XRD pattern indicates a more ordered structure compared to N-KB and all other sharp crystalline peaks can be well indexed to the crystalline spinel structure of Co₃O₄ [JCPDS: 43-1003]. However, the intensities of the peaks related to Co₃O₄ species are weak owing to low Co₃O₄ content. According to the Scherrer's equation, the crystal size of Co₃O₄ particle on N-KB is determined to be 15.6 nm. The lattice parameters are calculated to be a = b = c = 8.08 Å. XRD analyses indicate that Co₃O₄ and N-KB co-exist in the as-prepared Co₃O₄/N-KB composite.

Fig. 1 (a) XRD patterns of N-KB and Co₃O₄/N-KB composite, XPS spectra of (b) survey spectrum, (c) Co 2p, (d) O 1s, (e) N 1s and (f) C 1s for Co₃O₄/N-KB composite.

XPS measurement was performed to obtain more detailed elemental composition and the chemical element valence of Co₃O₄/N-KB composite. The full survey of Co₃O₄/N-KB composite in Fig. 1b indicates the existence of Co 2p, O 1s, N 1s and C 1s. By using a Gaussian fitting method, the Co 2p emission spectrum (Fig. 1c) was well fitted considering two spin–orbit doublets characteristic of Co²⁺ and Co³⁺. The fitting peaks at binding energies of 795.1 eV and 780.1 eV are ascribed to Co²⁺, while the peaks at 796.8 eV and 781.7 eV are due to Co³⁺.^42–44 The high resolution spectrum for O 1s shows three oxygen species (Fig. 1d), marked as O_I, O_II and O_III. According to previous reports, the fitting peak of O_I at 533.2 eV can be ascribed to a multiplicity of physically and chemically bonded water on and within the surface,⁴⁵ that of O_II at 531.5 eV corresponds to a high number of defect sites with a low oxygen coordination in the material with small particle size.⁴⁶ And that of O_III at 530.2 eV is a typical metal–oxygen bond.^45,47,48 The N 1s emission spectrum can be divided into three main peaks (Fig. 1e), 398.5, 400.3 and 401.3 eV, corresponding to pyridinic nitrogen, pyrrolic nitrogen and graphitic nitrogen.^49–51 It should be noted that the edge plane nitrogen groups (pyrrolic or pyridinic) in Co₃O₄/N-KB composite may provide additional active sites for catalytic process due to the lone pair electrons.⁵² The high resolution C 1s (Fig. 1f) peak is centered at 284.8 eV, related to the sp² graphitic carbon.⁵³ Apparently, the electron couples of Co³⁺/Co²⁺ in spinel Co₃O₄/N-KB, which may provide good electrocatalytic activity towards ORR.

The morphology and microstructure of the as-synthesized Co₃O₄/N-KB were characterized by SEM, TEM and HRTEM. Fig. 2a shows the typical SEM images of Co₃O₄/N-KB composite. As observed, Co₃O₄ nanoparticles are well dispersed on the surface of N-KB. In order to deeply investigate the morphology and microstructure of the Co₃O₄/N-KB, low and high-magnification TEM images of the composites are presented in Fig. 2b–d. As seen from Fig. 2b, homogeneous distribution of nanosized Co₃O₄ with intimate contact on the N-KB is further confirmed. Note that there are some deep dark areas, which should be resulted from Co₃O₄ supported on N-KB, consistent well with the reference.⁵ HRTEM images of the Co₃O₄/N-KB composite show clearer lattice fringes of Co₃O₄ in Fig. 2c and d. The lattice spacing of 0.244 nm, 0.202 nm, and 0.143 nm correspond to Co₃O₄ (311), (400), and (440) planes, respectively (Fig. 2c). The lattice spacing of 0.286 nm in Fig. 2d corresponds to (220) planes of Co₃O₄ by lattice analysis. The Co₃O₄ particles promise their intimate contact on the surface of N-KB, leading to better synergistic effect. To determine the distribution of Co₃O₄ in N-KB composite, elemental identification of C, O, N and Co elements is performed by STEM-EDS. As seen from the EDS mapping, uniform elemental distribution of C, O, N and Co elements was achieved in the Co₃O₄/N-KB composite (Fig. 3).

Fig. 2 (a) SEM images and (b) low magnification TEM images of Co₃O₄/N-KB composite, (c and d) show the magnified HRTEM images for clearer lattice fringes in corresponding square region marked in (b).

Fig. 3 (a) STEM elemental mapping images of the Co₃O₄/N-KB composite taken from the left selected region marked in (a), (b–e) the corresponding elemental mapping images of (b) C–K, (c) O–K, (d) N–K, and (e) Co–K.

To further identify the structural changes of the carbon framework of KB, N-KB, and Co₃O₄/N-KB composite, Raman spectroscopy results are given in Fig. 4a. Clearly, the D and G bands of KB show a shift to 1325.8 and 1592.7 cm⁻¹, respectively. N-KB shows the similar D and G bands at 1321.6 and 1591.4 cm⁻¹, respectively. A general feature of the N-KB is that the D band and the G band shift to lower wavenumber than those of the KB.⁵⁴ Co₃O₄/N-KB composite has two prominent peaks at 1329.2 and 1591.7 cm⁻¹ which correspond to the well-documented D band and G band, respectively. After hydrothermal treatment, the D and G bands shift to higher wavenumber in the Co₃O₄/N-KB composite. From the Raman spectra, we know that the intensity ratio of D band to G band (i.e. I_D/I_G) for Co₃O₄/N-KB composite is 1.389, which is relatively higher in comparison with that of N-KB (I_D/I_G = 1.373) and KB (I_D/I_G = 1.347). The increasing of I_D/I_G ratio for N-KB is attributed to the structural defects and edge plane exposure caused by nitrogen atom incorporation into the carbon layers present on the surface of N-doped carbon materials.^54–56 The I_D/I_G ratio is larger for Co₃O₄/N-KB composite than N-KB, indicates a decrease in the average size of the sp² domains formed during a hydrothermal process.⁵⁷ This can be allowed that the structure of the carbon layers has rebuilded in Co₃O₄/N-KB composite, which implies that the size of the in-plane-sp² domains and the defects increased during Co₃O₄/N-KB composite.^58,59 The increase of I_D/I_G in Co₃O₄/N-KB composite can also further increase defect sites of such systems.⁶⁰ As a matter of fact, the sites exhibit high electrocatalytic activities for ORR.

Fig. 4 (a) Raman spectra of undoped KB, N-KB, and Co₃O₄/N-KB composite, (b) DSC/TG curve of the as-prepared Co₃O₄/N-KB composite.

The mass ratio of N-KB in the composite was investigated by DSC/TG curve (Fig. 4b). A large weight loss of 88.6% is observed in the temperature range between 261 and 631 °C in TG curve, corresponding to the burning of N-doped Ketjenblack.⁶¹ There is a big exothermic peak at about 420 °C in DSC curve which should be related to the reaction between N-KB and O₂.⁶² When the temperature is over 631 °C, there is no weight loss, indicating that the content of N-KB in the composite is about 88.6%.

To obtain further insight into the ORR performance of each catalyst, the as-prepared catalysts were loaded onto glassy carbon electrodes for the linear sweeping voltammograms (LSVs) at 1600 rpm in O₂-saturated 0.1 M KOH in Fig. 5a. All samples were scanned cathodically at a rate of 10 mV s⁻¹. It clearly indicates that the pure Co₃O₄ and KB exhibit very poor onset potential. Compared to pure Co₃O₄, Co₃O₄/KB shows higher positive onset potential and much higher limiting current density. The results suggest that cobalt oxides supported on carbon materials play a significant role in increasing the performance of ORR. N-KB also shows better activity towards ORR than pure KB. It has been well proved that N doping in KB could facilitate chemisorption of oxygen, leading to a relatively high catalytic activity towards ORR.⁶³ Two plateaus (from 0.8 to 0.5 V and 0.5 to 0 V) are observed for Co₃O₄, KB and N-KB, indicating that Co₃O₄, KB and N-KB underwent a two-electron pathway (see Fig. 6a–c) and proceeded via a hydroperoxide anion (HO₂⁻) intermediate.⁶⁴ The assessed H₂O₂ yield is found to be significantly high for Co₃O₄ (nearly 100%), KB (nearly 50%) and N-KB (nearly 30%) in the potential region of 0.3 to 0.85 V (vs. RHE). The derived electron transfer numbers from the RRDE estimation are found to be 1.9, 2.4 and 2.7 for Co₃O₄, KB and N-KB, respectively (see Fig. 5d). Interestingly, Co₃O₄/N-KB composite possesses a more positive half-wave potential (E_1/2, 0.79 V), hence resulting in better ORR activity, which is higher than those of Co₃O₄ (0.28 V), KB (0.56 V), N-KB (0.64 V), Co₃O₄/KB (0.72 V) and only 40 mV negative shift compared with the commercial Pt/C catalyst. Moreover, the limiting current density of Co₃O₄/N-KB is up to about −5.6 mA cm⁻², which is much higher than the other samples, and even outperforms the commercial Pt/C (−5.2 mA cm⁻²). In addition, the catalytic activity of Co₃O₄/N-KB composite depends crucially on the amount of Co₃O₄. Too many Co₃O₄ may cause the free N-KB, which will weaken the ORR performance. A control experiment was fulfilled to optimize the catalytic activity of the Co₃O₄/N-KB composite by methodically changing the content of Co₃O₄. The results in Fig. 5b demonstrate that the Co₃O₄/N-KB composite with 5 wt% Co₃O₄ displays the best ORR activity. It should be noted that N-KB and Co₃O₄ exhibit inferior ORR catalytic activity, while Co₃O₄/N-KB presents significantly improved electroactivity to ORR. After the incorporation of N-KB, the ORR kinetics of Co₃O₄ could be significantly enhanced. The strongly synergistic effect of Co₃O₄ and N-KB in the composite was the key towards improving electrochemical performance. It should be noted that the integration of Co₃O₄ and N-KB was feasible to improve the catalytic activity of cobalt oxides based cathode in fuel cells and metal–air batteries.⁶⁵ The synergistic coupling between Co₃O₄ and N-doped carbon materials establish more active sites to obtain high ORR activity owing to interaction between carbon structure and other molecules.⁶⁶ Meanwhile, it has revealed that the synergistic performance enhancement results from an improved carbon-catalyst binding and increased electrical conductivity, which leads to a large number of the active sites of the Co₃O₄/N-KB composite.^23,67 We may conclude that the Co₃O₄ grown on N-KB prevented Co₃O₄ from restacking to lead to uniform nucleation during hydrothermal process, which made Co₃O₄ better dispersion on N-KB.⁶⁸ Such a synergistic coupling led to not only enhanced current density but also plenty of active sites for ORR.

Fig. 5 (a) ORR polarization curves for different catalysts at 1600 rpm, (b) the linear polarization curves of air electrodes with different contents of Co₃O₄ in Co₃O₄/N-KB composite, (c) CVs of Co₃O₄, KB, N-KB, Co₃O₄/KB, Co₃O₄/N-KB and Pt/C in Ar and O₂ saturated 0.1 M KOH solution, (d) the electron transfer number (dotted line) and percentage of peroxide (solid line) of different catalysts at different potentials.

Fig. 6 ORR polarization curves of (a) Co₃O₄, (b) KB, (c) N-KB, (d) Co₃O₄/KB, (e) Co₃O₄/N-KB and (f) Pt/C at different rotating speeds (400–1600 rpm) in O₂-saturated 0.1 M KOH solution at a scan of 10 mV s⁻¹. Insets: K–L plots at different potentials.

CV measurements were carried out in an O₂-saturated or Ar-saturated 0.1 M KOH solution with a scan rate of 10 mV s⁻¹. The tested voltage ranges from 1.2 to 0 V (vs. RHE). As shown in Fig. 5c, CV curves of all samples recorded in Ar-saturated electrolyte show no obvious peaks, however, when the electrolyte was saturated with O₂, an obvious reduction peak clearly appears, confirming the electrocatalytic activity for ORR. The CV curve for Co₃O₄/N-KB electrode in O₂ shows a reduction peak potential of ∼0.78 V (vs. RHE), which is much more positive than those for Co₃O₄ (0.49 V), KB (∼0.74 V), N-KB (∼0.75 V), Co₃O₄/KB (∼0.76 V), and only negative shift of about 20 mV compared to the commercial Pt/C. The positive shifting of the reduction peak of Co₃O₄/N-KB in comparison with Co₃O₄ and N-KB indicates the existence of synergistic effect between Co₃O₄ and N-KB. This was in agreement with the LSVs and CVs results. It was reported that Co₃O₄ was an efficient synergistic component for ORR catalysts owing to possessing the function of the intermediate disproportionation reaction in oxygen reduction.⁶⁵ Liu et al.⁶⁹ brought insights into the synergistic coupling effect between Co₃O₄ and MWCNT, and demonstrated that the enhanced stability of Co₃O₄/MWCNT composite material could be affected by Co₃O₄ in the composite assisting the prevention of carbon corrosion. Furthermore, the good synergistic interaction between N-doped carbon and Co NPs in the Co/N–C-800 composite has been demonstrated by Su et al.,⁷⁰ it can be inferred that the smaller charge transfer resistance, higher specific surface area, and better synergistic interaction could be the causes for the better catalytic activity of Co/N–C-800 than other Co/N–C samples. Olson et al.^5,71 have put forward a dual-site mechanism for cobalt–polypyrrole/C ORR catalyst, where oxygen is firstly reduced to HO₂⁻ at Co–N–C sites and further reduced to OH⁻ at Co_xO_y/Co sites in alkaline media. Herein, our current research system may be similar to the above mechanism, suggesting synergistic coupling between Co₃O₄ and N-KB is critical to ORR activity of the composite.

To obtain further information about ORR kinetics, the Koutecky–Levich plots (j⁻¹ vs. ω^−1/2) of each catalyst are obtained from LSVs at various potentials. All plots show good linearity (R² > 0.99) at various rotation speeds (Fig. 6e). The n value for Co₃O₄/N-KB was calculated to be ∼4 at a large potential range, which suggests a four-electron pathway for ORR. The electron transfer numbers (n) of Co₃O₄/KB and the commercial Pt/C were also calculated to be ∼4 (Fig. 6d and f), indicating the existence of a four-electron pathway toward the formation of OH⁻ ions.⁸ To further prove H₂O₂ yield of Co₃O₄/KB, Co₃O₄/N-KB and Pt/C, we obtained reliable data from the RRDE. The amount is found to be in a low level of 10% in the potential region of 0.3 to 0.85 V (vs. RHE). The derived electron transfer numbers (Fig. 5d) are close to 4, indicating the direct 4e⁻ transfer reaction in ORR. It showed that cobalt oxides supported on carbon materials led to a clear enhancement of the electron-transfer kinetics of ORR, similar to the behavior of the commercial Pt/C. Tafel plots of Co₃O₄/KB and Co₃O₄/N-KB derived from LSVs data are shown in Fig. 7a. The Tafel slope of Co₃O₄/N-KB is 74.7 mV per decade, which is superior to the 83.1 mV per decade of the Co₃O₄/KB, indicating a good ORR kinetic process for Co₃O₄/N-KB. In addition, the stability of Co₃O₄/N-KB for the ORR was examined via the chronoamperometric method in O₂-saturated at 1600 rpm. The potential was selected as 0.7 V. As seen in Fig. 7b, the Co₃O₄/N-KB catalyst is more stable than Co₃O₄/KB catalyst and the commercial Pt/C catalyst at the same testing condition. Apparently, the Co₃O₄/N-KB here exhibits high onset potential, high positive half-wave potential, high electron transfer number (∼4) and can be considered as a suitable catalyst for high power metal–air batteries.

Fig. 7 (a) Tafel plots of kinetic current for Co₃O₄/KB and Co₃O₄/N-KB, (b) current–time (it) chronoamperometric response of Co₃O₄/KB, Co₃O₄/N-KB and Pt/C at 0.7 V in O₂-saturated 0.1 M KOH.

Al–air batteries with Co₃O₄/KB, Co₃O₄/N-KB and commercial Pt/C catalyst as the cathodes were fabricated for further comparison of their practical activity in 6 M KOH solution. We recorded the galvanostatic discharge curves (Fig. 8) at a constant current discharge of 20 mA cm⁻² for 15 h. As can be seen, all cells show the increased discharge voltage in the beginning due to the activation process of Al anode, on which the passive film was dissolved gradually. Before ∼2.5 h, the working voltage plateau of the cells with Co₃O₄/KB as an electrocatalyst is close to that of Co₃O₄/N-KB, which is higher than that of Pt/C. After ∼2.5 h, the cell with the as-prepared Co₃O₄/KB catalyst always shows a lower discharge voltage than those of Co₃O₄/N-KB and commercial Pt/C catalyst. However, in the initial 5 h, it is interesting to note that the cell with the as-prepared Co₃O₄/N-KB catalyst even exhibits a little higher discharge voltage than that with Pt/C catalyst. As follows, their average working voltages (1.52 V) are almost the same. Generally, it could be concluded that the practical electrocatalytic activity of the Co₃O₄/N-KB composite here in Al–air battery is comparable to that of the commercial Pt/C, enabling it a very promising catalyst with high efficiency and low cost for Al–air batteries.

Fig. 8 Galvanostatic discharge curves of the Al–air batteries with Co₃O₄/KB, Co₃O₄/N-KB and Pt/C as the cathode catalysts at a discharge current density of 20 mA cm⁻².

Conclusions

In summary, Co₃O₄/N-KB composite was successfully prepared by a facile strategy and used as a high performance catalyst for ORR. In comparison with the alone Co₃O₄ and N-KB, the as-prepared Co₃O₄/N-KB showed much higher current density and more positive half-wave potential. Note that Co₃O₄/N-KB composite underwent a four-electron reaction mechanism, whose catalytic activity approached to the commercial Pt/C. The improved performance of Co₃O₄/N-KB should be attributed to the increased active sites resulted from synergetic chemical coupling effect between Co₃O₄ and N-KB. Al–air full battery using Co₃O₄/N-KB as catalyst was also evaluated, which exhibited comparable discharge voltage plateau to the commercial Pt/C counterpart at the same current density. Our work provides a simple but efficient approach to produce low-cost Co₃O₄/N-KB composite as efficient ORR electrocatalyst for Al–air battery.

Acknowledgements

This work was financially supported by the National Nature Science Foundation of China (No. 21571189, No. 21271187, No. 51304077 and No. 21301193), the Opening Project of Guangxi Key Laboratory of Information Materials (No. 131006-K), and the Open-End Fund for Valuable and Precision Instruments of Central South University (CSUZC201622).

Notes and references

1. Y. G. Guo, J. S. Hu and L. J. Wan, Adv. Mater., 2008, 20, 2878–2887.
2. G. H. Yu, L. B. Hu, M. Vosgueritchian, H. L. Wang, X. Xie, J. R. McDonough, X. Cui, Y. Cui and Z. N. Bao, Nano Lett., 2011, 11, 2905–2911.
3. J. Yan, Q. Wang, T. Wei and Z. J. Fan, Adv. Energy Mater., 2014, 4, 1400500–14005005.
4. A. S. Aricò, P. Bruce, B. Scrosati, J. M. Tarascon and W. V. Schalkwijk, Nat. Mater., 2005, 4, 366–377.
5. Y. Y. Liang, Y. G. Li, H. L. Wang, J. G. Zhou, J. Wang, T. Regier and H. J. Dai, Nat. Mater., 2011, 10, 780–786.
6. J. W. Xiao, C. Chen, J. B. Xi, Y. Y. Xu, F. Xiao, S. Wang and S. H. Yang, Nanoscale, 2015, 7, 7056–7064.
7. Y. Y. Liang, H. L. Wang, P. Diao, W. Chang, G. S. Hong, Y. G. Li, M. Gong, L. G. Xie, J. G. Zhou, J. Wang, T. Z. Regier, F. Wei and H. J. Dai, J. Am. Chem. Soc., 2012, 134, 15849–15857.
8. Z. S. Wu, S. B. Yang, Y. Sun, K. Parvez, X. L. Feng and K. Müllen, J. Am. Chem. Soc., 2012, 134, 9082–9085.
9. R. G. Cao, J. S. Lee, M. L. Liu and J. P. Cho, Adv. Energy Mater., 2012, 2, 816–829.
10. Z. W. Chen, D. Higgins, A. P. Yu, L. Zhang and J. J. Zhang, Energy Environ. Sci., 2011, 4, 3167–3192.
11. Y. L. Yu, Y. P. Hu, X. W. Liu, W. Q. Deng and X. Wang, Electrochim. Acta, 2009, 54, 3092–3097.
12. Y. Y. Shao, J. H. Sui, G. P. Yin and Y. Z. Gao, Appl. Catal., B, 2008, 79, 89–99.
13. Y. Y. Shao, J. Liu, Y. Wang and Y. H. Lin, J. Mater. Chem., 2009, 19, 46–59.
14. P. Kichambare, S. Rodrigues and J. Kumar, ACS Appl. Mater. Interfaces, 2012, 4, 49–52.
15. S. B. Yang, X. L. Feng, X. C. Wang and K. Mullen, Angew. Chem., Int. Ed., 2011, 50, 5339–5343.
16. R. A. Sidik, A. B. Anderson, N. P. Subramanian, S. P. Kumaraguru and B. N. Popov, J. Phys. Chem. B, 2006, 110, 1787–1793.
17. D. Geng, Y. Chen, Y. Chen, Y. Li, R. Li, X. Sun, S. Ye and S. Knights, Energy Environ. Sci., 2011, 4, 760–764.
18. G. Nam, J. Park, S. T. Kim, D. B. Shin, N. Park, Y. Kim, J. S. Lee and J. Cho, Nano Lett., 2014, 14, 1870–1876.
19. J. S. Lee, G. S. Park, S. T. Kim, M. Liu and J. Cho, Angew. Chem., 2013, 125, 1060–1064.
20. H. Zhang and W. Yang, Chem. Commun., 2007, 7, 4215–4217.
21. Y. Wang, X. Lu, Y. Liu and Y. Deng, Electrochem. Commun., 2013, 31, 108–111.
22. K. Song, E. B. Cho and Y. M. Kang, ACS Catal., 2015, 5, 5116–5122.
23. Z. P. Li, Z. X. Liu, K. N. Zhu, Z. Li and B. H. Liu, J. Power Sources, 2012, 219, 163–171.
24. W. H. Ryu, T. H. Yoon, S. H. Song, S. k. Jeon, Y. J. Park and I. D. Kim, Nano Lett., 2013, 13, 4190–4197.
25. G. Wang, G. Sun, Q. Wang, S. Wang, J. Guo, Y. Gao and Q. Xin, J. Power Sources, 2008, 180, 176–180.
26. D. Gelman, B. Shvartsev and Y. Ein-Eli, J. Mater. Chem. A, 2014, 2, 20237–20242.
27. R. Mori, RSC Adv., 2014, 4, 30346–30351.
28. Q. Li and N. J. Bjerrum, J. Power Sources, 2002, 110, 1–10.
29. L. Fan, H. Lu and J. Leng, Electrochim. Acta, 2015, 165, 22–28.
30. M. Wang, Y. Lai, J. Fang, J. Li, F. Qin, K. Zhang and H. Lu, Int. J. Hydrogen Energy, 2015, 40, 16230–16237.
31. L. Fan and H. Lu, J. Power Sources, 2015, 284, 409–415.
32. R. Revel, T. Audichon and S. Gonzalez, J. Power Sources, 2014, 272, 415–421.
33. D. Gelman, B. Shvartsev and Y. Ein-Eli, J. Mater. Chem. A, 2014, 2, 20237–20242.
34. J. S. Lee, G. S. Park, H. L. Lee, S. T. Kim, R. Cao, M. Liu and J. Cho, Nano Lett., 2011, 11, 5362–5366.
35. J. J. Chen, N. Zhou, H. Y. Wang, Z. G. Peng, H. Y. Li, Y. G. Tang and K. Liu, Chem. Commun., 2015, 51, 10123–10126.
36. Z. Lin, G. H. Waller, Y. Liu, M. Liu and C. P. Wong, Nano Energy, 2013, 2, 241–248.
37. Y. Ma, R. Wang, H. Wang, J. Key and S. Ji, J. Power Sources, 2015, 280, 526–532.
38. R. Liu, D. Wu, X. Feng and K. Mullen, Angew. Chem., Int. Ed., 2010, 49, 2565–2569.
39. Y. Tang, H. Qiao, H. Wang and P. Tao, J. Mater. Chem. A, 2013, 1, 12512–12518.
40. H. Zhang, H. Qiao, H. Wang, N. Zhou, J. Chen, Y. Tang and C. Huang, Nanoscale, 2014, 6, 10235–10242.
41. X. Yang, Y. Xu, H. Zhang, Y. A. Huang, Q. Jiang and C. Zhao, Electrochim. Acta, 2013, 114, 259–264.
42. J. Zhu and Q. Gao, Microporous Mesoporous Mater., 2009, 124, 144–152.
43. J. Li, S. Xiong, Y. Liu, Z. Ju and Y. Qian, Nano Energy, 2013, 2, 1249–1260.
44. L. E. Gómez, A. V. Boix and E. E. Miró, Catal. Today, 2013, 216, 246–253.
45. X. F. Lu, D. J. Wu, R. Z. Li, Q. Li, S. H. Ye, Y. X. Tong and G. R. Li, J. Mater. Chem. A, 2014, 2, 4706–4713.
46. R. Ding, L. Qi, M. Jia and H. Wang, J. Appl. Electrochem., 2013, 43, 903–910.
47. J. F. Marco, J. R. Gancedo, M. Gracia, J. L. Gautier, E. Ríos and F. J. Berry, J. Solid State Chem., 2000, 153, 74–81.
48. T. Choudhury, S. O. Saied, J. L. Sullivan and A. M. Abbot, J. Phys. D: Appl. Phys., 1989, 22, 1185–1195.
49. H. Wang, T. Maiyalagan and X. Wang, ACS Catal., 2012, 2, 781–794.
50. N. P. Subramanian, X. Li, V. Nallathambi, S. P. Kumaraguru, H. C. Mercado, G. Wu, J. W. Lee and B. N. Popov, J. Power Sources, 2009, 188, 38–44.
51. Y. Mao, H. Duan, B. Xu, L. Zhang, Y. Hu, C. Zhao, Z. Wang, L. Chen and Y. Yang, Energy Environ. Sci., 2012, 5, 7950–7955.
52. Z. Chen, D. Higgins and Z. Chen, Carbon, 2010, 48, 3057–3065.
53. Z. Zhang, L. Su, M. Yang, M. Hu, J. Bao, J. Wei and Z. Zhou, Chem. Commun., 2014, 50, 776–778.
54. A. Zahoor, M. Christy, Y. J. Hwang, Y. R. Lim, P. Kim and K. S. Nahm, Appl. Catal., B, 2014, 147, 633–641.
55. L. S. Panchakarla, A. Govindaraj and C. N. R. Rao, ACS Nano, 2007, 1, 494–500.
56. Z. H. Sheng, L. Shao, J. J. Chen, W. J. Bao, F. B. Wang and X. H. Xia, ACS Nano, 2011, 5, 4350–4358.
57. M. S. Arif Sher Shah, A. R. Park, K. Zhang, J. H. Park and P. J. Yoo, ACS Appl. Mater. Interfaces, 2012, 4, 3893–3901.
58. H. W. Wang, Z. A. Hu, Y. Q. Chang, Y. L. Chen, H. Y. Wu, Z. Y. Zhang and Y. Y. Yang, J. Mater. Chem., 2011, 21, 10504–10511.
59. D. W. Chang, E. K. Lee, E. Y. Park, H. Yu, H. J. Choi, L. Y. Jeon, G. J. Sohn, D. Shin, N. Park, J. H. Oh, L. Dai and J. B. Baek, J. Am. Chem. Soc., 2013, 135, 8981–8988.
60. D. Sen, R. Thapa and K. K. Chattopadhyay, ChemPhysChem, 2014, 15, 2542–2549.
61. Y. Su, A. Pan, Y. Wang, J. Huang, Z. Nie, X. An and S. Liang, J. Power Sources, 2015, 295, 254–258.
62. J. W. Min, A. K. Kalathil, C. J. Yim and W. B. Im, Mater. Charact., 2014, 92, 118–126.
63. S. Wang, E. Iyyamperumal, A. Roy, Y. Xue, D. Yu and L. Dai, Angew. Chem., Int. Ed., 2011, 50, 11756–11760.
64. I. Hwang, E. Ahn and Y. Tak, Int. J. Electrochem. Sci., 2014, 9, 5454–5466.
65. J. Ahmed, Y. Yuan, L. Zhou and S. Kim, J. Power Sources, 2012, 208, 170–175.
66. M. Prabu, P. Ramakrishnan and S. Shanmugam, Electrochem. Commun., 2014, 41, 59–63.
67. G. Vijayaraghavan and K. J. Stevenson, Langmuir, 2007, 23, 5279–5282.
68. S. Zhang, H. Zhang, Q. Liu and S. Chen, J. Mater. Chem. A, 2013, 1, 3302–3308.
69. Y. Liu, D. Higgins, J. Wu, M. Fowler and Z. Chen, Electrochem. Commun., 2013, 34, 125–129.
70. Y. Su, Y. Zhu, H. Jiang, J. Shen, X. Yang, W. Zou, J. Chen and C. Li, Nanoscale, 2014, 6, 15080–15089.
71. T. S. Olson, S. Pylypenko, P. Atanassov, K. Asazawa, K. Yamada and H. Tanaka, J. Phys. Chem. C, 2010, 114, 5049–5059.

---