An enhanced oxygen electrode catalyst by incorporating CoO/SnO₂ nanoparticles in crumpled nitrogen-doped graphene in alkaline media

Abstract

To address issues concerning energy consumption and the environment, it is of great importance to design a highly efficient, durable and inexpensive oxygen electrode for energy storage and conversion devices. In this work, CoO/SnO₂ nanoparticles were successfully incorporated in nitrogen-doped graphene by a facile method. The composite was of porous structure, and exhibited efficient bifunctional activity and outstanding stability towards both oxygen reduction reaction (ORR) and oxygen evolution reaction (OER). For ORR, the catalyst favored a direct four-electron-transfer pathway, and produced kinetic current density values close to those of Pt/C. In addition, for OER, it achieved a lower overpotential at a current density of 10 mA cm⁻² and a smaller Tafel slope than RuO₂. What is more, the value of the potential difference ΔE_(OER–ORR), a measure of the overall bifunctionality of the catalyst, was 0.72 V, comparable with the values for state-of-the-art nonprecious bifunctional catalysts. The remarkable bifunctional activity of the catalyst was mainly attributed to the synergistic effects between metal oxides and nitrogen-doped graphene.

---

1. Introduction

Owing to their high efficiency, environmental friendliness and convenience, new energy resources have attracted intensive attention in order to protect the ecological environment and build a harmonious society. The oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) play decisive roles in oxygen reaction-based regenerative energy storage and conversion devices, including metal–air batteries, and water-splitting and fuel cells.^1–5 So far, Pt/C and RuO₂ have been recognized as the most efficient ORR and OER catalysts,^6,7 respectively. However, the scarcity and high cost of the noble metals hamper their large scale application. Therefore, it is desirable to develop highly efficient and low-cost nonprecious bifunctional catalysts as alternatives to noble metal catalysts.

Recently, various nonmetal and nonprecious metal materials have been reported as effective oxygen electrode catalysts, including heteroatom-doped carbon,^8–11 transition metal composites,^2,12–18 spinel materials,^19,20 perovskite materials,^4,21–23 and so on. Investigations revealed nitrogen-doped carbon afforded distinct ORR/OER performance^10,24–27 because nitrogen intrinsically created polarization in the carbon framework, configurated defects, and induced active sites.^24,28 In addition, the transition metal composites/nitrogen-doped carbon hybrids achieved attractive bifunctional performance, mainly due to electronic coupling between transition metal and carbon, and to synergistic effects.^1,29,30 Cobalt-based materials, such as porous Co₃O₄,^31,32 plate-formed Co(OH)₂,³³ porous CoS₂,³⁴ Co₂P₃ ³⁵ and so on, were promising since these materials gave a reasonable balance between storage, cost, activity and durability. Particularly, CoO nanoparticles which were of a diversity of morphologies exhibited good ORR^36–38 and superior OER performance^1,38 due to their mesoporous structure, large surface area, and the highly active surface containing Co²⁺ with oxygen vacancies.^1,36–38 When CoO nanoparticles were supported on N–carbon, the ORR and OER performance was greatly improved.^1,14,39 Additionally, tin oxide (SnO₂) itself was of inherently abundant oxygen vacancy defects and a variety of valance states of Sn,⁴⁰ so that long-term durability and kinetic stability could be obtained when tin oxide with very high surface energy was modified with metal oxides and conductive templates.^41–43 Thus, SnO₂/graphene connected to CoO could be efficient as an oxygen electrode catalyst; however, as far as we know, there are few previous reports in this respect.

In view of this, we prepared a hybrid of CoO and nitrogen-doped graphene-supported tin oxide by a facile method. In this material, the 3d orbitals of tin opened up the possibility of redistribution of electron density, resulting in the easy formation of oxygen vacancies, which play a critical role in catalyzing the OER.⁴⁴ As a consequence, the hybrid exhibited excellent ORR and OER performance – a quick four-electron-transfer pathway, a kinetic current density close to that of Pt/C, and a favorable kinetic process in the ORR. What is more, it achieved a similar overpotential (285 mV vs. saturated calomel electrode (SCE) at a current density of 10 mA cm⁻²) to RuO₂ in the OER. At the same time, the hybrid avoided the common problem of poor durability of carbon–transition-metal catalysts under corrosive conditions. The remarkable bifunctional performance mainly resulted from the porous structure with high specific surface area, which facilitated quick electron and mass transfer, and the synergistic effects between metal oxides and nitrogen-doped graphene. As far as we know, this is the first report on cobalt monoxide and tin oxide dispersed in nitrogen-doped graphene as an efficient oxygen electrode catalyst.

2. Experimental

2.1. Materials

Graphite powder was obtained from Sinopharm Chemical Reagent Co., Ltd. Commercial platinum/carbon (Pt/C) (20 wt% Pt on Vulcan XC-72) was purchased from Alfa Aesar. Both ruthenium(IV) oxide and 5% Nafion were obtained from Sigma. All other chemical reagents, including H₂SO₄ (98%), KOH, KMnO₄, Na₂SO₄, melamine, cobalt(II) acetate tetrahydrate and tin(II) chloride dehydrate, were analytically pure and were employed directly as received. Distilled water was used throughout the whole experiment.

2.2. Decorating nitrogen-doped graphene with cobalt monoxide and tin oxide

Graphene oxide (GO) was first prepared using the modified Hummers' method.⁴⁵ Then, 400 mg of GO was dispersed in 50 mL of mixed ethylene glycol and distilled water, where the volume ratio of ethylene glycol to distilled water was 9 : 1. To fabricate the final catalyst, a facile method was employed that used a combination of hydrothermal treatment and pyrolysis. In a typical process, 0.14 mmol of SnCl₂·2H₂O, 0.35 mmol of Co(OAc)₂·4H₂O, 1.07 mmol of C₃H₆N₆, 0.55 mmol of Na₂SO₄, 6.2 mL of the above GO dispersion, and 0.5 mL of ethylene glycol were dissolved in 20 mL of distilled water and vigorously stirred at 70 °C for 6 h. Herein, the epoxide in graphene promoted the hydrolysis of SnCl₂·2H₂O and accelerated the condensation of Sn(OH)₄, which further crosslinked together and yielded a homogenous mixture in a GO dispersion under strong stirring.⁴⁶ Then the mixture was transferred to a 50 mL Teflon-lined stainless autoclave, with the addition of NH₃·H₂O to adjust the pH to ∼12, and heated at 180 °C for 12 h, forming a black precipitate. After that, the wet aggregate was washed with distilled water several times until the pH was neutral, and dried in a vacuum oven at 35 °C overnight. The product was recorded as CoO/SnO₂/NG. Afterwards, the black material was placed in the center of a tube furnace, and heated to 300 °C at a rate of 5 °C min⁻¹, which was then increased to 500 °C at 10 °C min⁻¹, and calcined for 2 h in an argon atmosphere. The powder was labeled CoO/SnO₂/NG-500. For comparison, CoO/SnO₂/NG-600 and CoO/SnO₂/NG-700 were prepared. CoO/NG-500, SnO₂/NG-500, NG-500 and r-GO-500 were also obtained by a similar procedure, only with the absence of SnCl₂·2H₂O, Co(OAc)₂·4H₂O, SnCl₂·2H₂O and Co(OAc)₂·4H₂O, and SnCl₂·2H₂O and Co(OAc)₂·4H₂O and NH₃·H₂O, respectively.

2.3. Morphology and structure characterization

The morphology and structure of the as-synthesized samples were characterized by scanning electron microscopy (SEM; JSM-6701F, operating at 5 kV) and transmission electron microscopy (TEM; Tecnai G2 20 S-Twin, operated at 200 kV). SEM energy-dispersive X-ray (EDX) mapping was conducted on a focused ion beam scanning electron microscope (EDS Inca X-Max). X-ray diffraction (XRD) powder spectra and nitrogen adsorption–desorption isotherms were measured on a RIGAK X-ray diffractometer (D/MAX2550 VB/PC, Japan) with Cu K_α radiation (λ = 0.15148 nm) and an ASAP2020 volumetric adsorption analyzer at 77 K, respectively. The Raman spectra were measured using a TriVista™ 555CRS Raman spectrometer at 532 nm. X-ray photoelectron spectroscopy (XPS) was done on an ESCLAB 250 spectrometer using Al K_α as the exciting source (1486.6 eV photons), and the C 1s peak at 284.6 eV as a reference.

2.4. Electrochemical characterization

All electrochemical measurements were taken in a standard three-electrode cell system on a CHI 660E electrochemical workstation, where a catalyst-modified glassy carbon electrode was used as working electrode, a saturated calomel electrode (SCE) functioned as the reference electrode and a platinum wire acted as a counter electrode. All electrochemical testing was performed with a Pine Instrument Company AF-MSRCE modulator speed rotator. The tests were carried out in O₂/N₂-saturated 0.1 M KOH solution for ORR with a rotating disk electrode (RDE, d = 5 mm) and ring rotating disk electrode (RRDE, d = 5.61 mm), and O₂-saturated 1 M KOH electrolyte for OER with the with a rotation speed of 1600 rpm and a scan rate of 10 mV s⁻¹. The working electrode was prepared as follows. In general, 6 mg of electrocatalyst was ultrasonically dispersed in 3 mL of solution comprising 2.8 mL of ethyl alcohol and 0.2 mL of Nafion, forming a homogeneous ink, followed by pipetting a certain amount of the ink onto the surface of a glassy carbon electrode. The catalyst loading per area was kept to 0.28 mg cm⁻² and all electrochemical measurements were performed three times to avoid any incidental errors. All potentials are reported versus (vs.) the saturated calomel electrode (SCE).

3. Results and discussion

3.1. Composition and morphology characterizations of the catalyst

XRD was used to characterize the crystallinity of the samples. As shown in Fig. 1a, r-GO-500 (002), (103), (107), (511) and (200) diffraction peaks were located at 26.1°, 43.2°, 48.8°, 50.3° and 54.5°, respectively. The broad peak at 2θ ∼ 26.1° suggested the reduction of GO. After nitrogen doping (NG-500), the diffraction peaks at 26.2°, 43.4° and 55.7° fitted well to NG-500 (002), (103) and (108), respectively. Fig. S1a† depicts that the diffraction peaks at 34.1°, 51.8°, and 65.7° in the SnO₂/NG-500 sample were indexed to SnO₂ (101), (201) and (310), respectively. In addition, the Sn (111) diffraction peak centered at 32.3° was observed in SnO₂/NG-500 (Fig. S1a†), formed by the reduction of SnO₂ via contact with graphene sheets during the calcination process.⁴⁷ The presence of Sn was very important in order to derive long-term stability.⁴⁸ And for CoO/NG-500, strong CoO (111), (021), (200), (104), (220) and (204) diffraction peaks appeared at 36.7°, 42.5°, 44.3°, 51.3°, 61.6° and 75.4°, respectively. The intensity of the diffraction peaks corresponding to 2θ ∼ 26° for both SnO₂/NG-500 and CoO/NG-500 decreased (Fig. S1a†), which was probably due to the overlapping of SnO₂ or the dispersion of CoO on the NG matrix. For CoO/SnO₂/NG-500, the characteristic diffraction peak of CoO (200) was located at 44.3°, and the strong XRD peak centered at 51.5°, 62.62° and 75.9° corresponded to SnO₂ (201), (223) and (400), respectively. The diffraction peak of SnO₂ was shifted to higher angles with narrow half-peak width compared with SnO₂/NG-500, indicating the high crystallinity of the hybrid.

Fig. 1 (a) XRD profiles for 2θ ranging from 10° to 90°; (b) nitrogen absorption–desorption isotherm curve and pore distribution curve (inset); (c) Raman spectra and I_D/I_G values of CoO/SnO₂/NG, CoO/SnO₂/NG-500, CoO/NG-500, SnO₂/NG-500, NG-500 and r-GO-500; and (d) enlargement of the rectangular area outlined in (c).

Catalytic reactions often occur on the surface of catalysts, and the reaction rate is typically related to the surface area.⁴⁹ The nitrogen adsorption isotherm for CoO/SnO₂/NG-500 revealed a typical-IV type curve with an obvious hysteresis loop (Fig. 1b), and the pore size distribution curve inset in Fig. 1b indicates a mesoporous and macroporous structure. The average pore width of CoO/SnO₂/NG-500 was 45.1 nm. The nitrogen adsorption–desorption isotherm and the pore size distribution curves for CoO/NG-500, SnO₂/NG-500, NG-500 and r-GO-500 are given for comparison; see Fig. S1b–e.† The surface area values were 509.4, 359.7, 173.5, 195.5 and 282.2 cm² g⁻¹ for r-GO-500, NG-500, CoO/NG-500, SnO₂/NG-500 and CoO/SnO₂/NG-500 (see Table S1†), respectively, and the average pore sizes are shown in Fig. S1f and Table S1.† The higher BET (Brunauer–Emmett–Teller) surface area of CoO/SnO₂/NG-500 compared with CoO/NG-500 and SnO₂/NG-500 was mainly ascribed to the strong interaction between cations preventing condensation and precipitation of mixed metal oxides,⁴¹ exposing more active sites, which would be favorable for the transfer of electrolyte and oxygen molecules during oxygen catalysis.^9,50

Raman spectra (Fig. 1c) exhibited typical D bands at around 1339 cm⁻¹, attributed to the structure defects in the graphene plane, and G bands at around 1589 cm⁻¹, assigned to the sp² carbon in all of the samples. As shown in Fig. 1c, the I_D/I_G values apparently were higher for NG-500 (1.13), SnO₂/NG-500 (1.26), CoO/NG-500 (1.08), CoO/SnO₂/NG (1.12) and CoO/SnO₂/NG-500 (1.16) than for r-GO-500 (1.04), revealing an increased number defects in the graphene structure after nitrogen and metal oxide doping. Additionally, the I_D/I_G value of CoO/SnO₂/NG-500 was higher than that of CoO/SnO₂/NG, attributed to the increased vacancies in the graphene planes and defects at the edges after heat decomposition.⁵¹ In addition, CoO/SnO₂/NG-500 displayed two bands located at 640 and 482 cm⁻¹ stronger than those of CoO/NG-500 and SnO₂/NG-500 (Fig. 1d), attributed to A_1g and E_g, respectively. The A_1g and E_g were Raman-active, and the E_g band (482 cm⁻¹) indicated the vibration of oxygen, representing the oxygen defects in the Raman spectrum.^1,41 The oxygen vacancies could increase the number of active defects, which is beneficial for oxygen catalysis.^1,44

The morphology and structure of the samples were characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The SEM micrograph in Fig. 2a shows more distinct crumples in NG-500 than in r-GO-500 (Fig. S2a†), representing more defective sites in NG-500. The high number of defects could also be observed in the TEM image (Fig. 2c). For the CoO/NG-500 and SnO₂/NG-500 samples, CoO and SnO₂ nanoparticles closely covered the surface of NG (Fig. S2b and c†). Fig. 2b shows that CoO/SnO₂ nanoparticles were anchored on the NG matrix in the CoO/SnO₂/NG-500 composite. What is more, the energy-dispersive X-ray (EDX) mapping in Fig. S2d–i† shows that cobalt, tin, nitrogen, oxygen and carbon atoms were distributed uniformly. At the same time, TEM images demonstrated that CoO and SnO₂ were successfully loaded on NG, and distinct crystal lattice fringes with interplanar distances of 2.4 and 1.8 Å correspond to the (200) plane of CoO and (201) plane of SnO₂ (inset in Fig. 2d), respectively.

Fig. 2 (a and b) SEM images of NG-500 (a) and CoO/SnO₂/NG-500 (b), and (c and d) TEM images of NG-500 (c) and CoO/SnO₂/NG-500 (d). Insets in (d) show HRTEM images of CoO and SnO₂.

Information about the surface chemical composition was gained by XPS. Fig. 3a reveals the signals of Co, Sn and N in CoO/SnO₂/NG-500, demonstrating the successful doping of the three elements into the graphene matrix. The contents of Co, Sn, N, O and C in CoO/SnO₂/NG-500 were determined to be 8.0, 7.8, 3.8, 27.0 and 53.4 at% (see Table 1), respectively. After doping, the C content in r-GO-500 was greatly reduced, as presented in Table 1. The high-resolution XPS spectra shed light on the bonding configuration of each component. As exhibited in Fig. 3b, the C 1s spectrum was deconvoluted into three peaks at 284.65, 286.1 and 288.25 eV, corresponding to sp² carbon, C=N & C–O and C–N & C=O, respectively.^24,52 The high peak intensity at 284.65 eV indicated the sp² nature of graphene in the composite. The O 1s peaks centered at 530.1, 530.75, 531.6, 532.45 and 533.45 eV were attributed to Sn–O, C–O, C(O)O, Co–O and O–H, respectively (see Fig. 3c).⁵³ As seen from Fig. 3d, the N 1s peaks demonstrated the formation of pyrrolic N (398.65 eV, 2.1 at%) and pyridinic N (400.1 eV, 1.7 at%). Nitrogen doping could enhance the conductivity of the graphene sheets, and pyrridic-like defects could facilitate electrolyte penetration into the graphene, and lead to a positive shift of the onset potential for ORR.^9,50,54 Therefore, nitrogen doping was beneficial for oxygen catalysis.^9,24,28 The peak located at 780.8 eV was fitted to Co²⁺, and the shake peaks at 786.2, 796.8 and 802.9 eV (Fig. 3e) further demonstrated the existence of Co²⁺.⁵⁵ As shown in Fig. 3f, the two peaks in the high resolution Sn 3d XPS spectrum were symmetric; Sn 3d_5/2 appeared at 487.3 eV and Sn 3d_3/2 at 495.2 eV, confirming the presence of Sn⁴⁺.^46,56 The high valence state of the tin ion (Sn⁴⁺) in the tin oxide created strong stabilizing effects.^41,57 Additionally, the Sn 3d_5/2 peak in SnO₂/NG-500 was located at 486.9 eV, and the Co²⁺ peak in CoO/NG-500 was at 780.2 eV (Fig. S3a and b†). The positive shifts of Sn 3d_5/2 and Co²⁺ in CoO/SnO₂/NG-500 relative to SnO₂/NG-500 and CoO/NG-500 indicated interaction between SnO₂ and CoO. What is more, the ΔBE_O–Sn (binding energy difference between O and Sn) of CoO/SnO₂/NG-500 was significantly reduced compared with that of SnO₂/NG-500 (Table S2†), further suggesting strong electronic coupling effects between SnO and anchored CoO.⁴¹

Fig. 3 (a) XPS spectrum of CoO/SnO₂/NG-500, and (b–f) high resolution XPS spectra of C 1s (b), O 1s (c), N 1s (d), Co 2p (e) and Sn 3d (f).

Table 1 XPS element contents of CoO/SnO₂/NG-500, CoO/SnO₂/NG, CoO/NG-500, SnO₂/NG-500, NG-500 and r-GO-500

Material	C 1s (%)	O 1s (%)	N 1s (%)	Co 2p(%)	Sn 3d(%)
CoO/SnO2/NG-500	53.4	27.0	3.8	8.0	7.8
CoO/SnO2/NG	51.9	27.5	9.5	7.7	3.4
CoO/NG-500	41.6	36.8	1.4	20.2	—
SnO2/NG-500	60.5	24.3	4.1	—	11.1
NG-500	85.9	3.3	10.8	—	—
r-GO-500	90.2	9.8	—	—	—

---

3.2. Oxygen reduction reaction activity

The oxygen reduction reaction activity of the catalyst was first evaluated by cyclic voltammetry (CV) measurements. As presented in Fig. 4a, a reduction peak appeared in O₂-saturated 0.1 M KOH solution, whereas a featureless plot was observed in O₂-free electrolyte, suggesting catalytic activity for CoO/SnO₂/NG-500 towards the ORR. To check the effect of calcining temperature on the ORR performance, we also prepared CoO/SnO₂/NG by annealing at different temperatures (500, 600 and 700 °C). In Fig. S4a,† CoO/SnO₂/NG-500 shows a larger peak current density than CoO/SnO₂/NG-600 and CoO/SnO₂/NG-700 in the CV measurements, and it achieved the largest limiting current density (J_L) and the most positive onset potential (E_onset) in the linear sweep voltammetry (LSV) polarization curves (Fig. S4b†), demonstrating a pronounced ORR catalytic activity for CoO/SnO₂/NG-500. Consequently, the following tests were taken on the CoO/SnO₂/NG-500 catalyst. LSV measurements were conducted for the samples using a rotating disk electrode at 1600 rpm and 10 mV s⁻¹ in 0.1 M KOH, as shown in Fig. 4b. It can be seen that the polarization curve of r-GO-500 exhibited an onset potential of −0.122 V vs. SCE, whereas NG-500 displayed the onset potential of −0.116 V vs. SCE, which was about 6 mV positive compared with that of r-GO-500, confirming the effect of nitrogen doping on the ORR. Additionally, CoO/SnO₂/NG-500 displayed the largest J_L (5.77 mA cm⁻²), the most positive half-wave potential (E_1/2 ∼ −0.147 V vs. SCE, only 10 mV more negative than that of Pt/C; thus ΔE_1/2 ∼ 10 mV) and the largest kinetic current density (J_K: 22.11 mA cm⁻²) compared with CoO/NG-500 and SnO₂/NG-500 (Fig. 4b and c and Table 2), implying the existence of synergistic effects between CoO/SnO₂ nanocrystals and NG. The smaller Tafel slope of CoO/SnO₂/NG-500 (33 mV dec⁻¹) than of Pt/C (51 mV dec⁻¹) (Fig. 4d and Table 1) indicated its intrinsically favorable kinetic process. The value of the Tafel slope of CoO/SnO₂/NG-500 was comparable with those of recently reported state-of-the-art bifunctional catalysts, such as Co(OH)_x–NCNT (54 mV dec⁻¹),¹⁵ CoS₂(400)/N,S-GO (75 mV dec⁻¹),¹⁶ NCG–CoO (48 mV dec⁻¹)⁵⁸ and so on (see Table S3†).

Fig. 4 (a) Cyclic voltammograms of CoO/SnO₂/NG-500 in O₂/N₂-saturated 0.1 M KOH solution at a scan rate of 10 mV s⁻¹; (b and c) linear sweep voltammograms (b) and kinetic current density (c) of various catalysts; (d) Tafel plots of CoO/SnO₂/NG-500 and Pt/C; (e) rotating ring disk electrode tests; (f and g) calculated electron transfer number per oxygen molecule (f) and the HO₂⁻ yield (g) of various catalysts; (h) current–time curves of CoO/SnO₂/NG-500 and Pt/C at −0.27 V at 1600 rpm after 5 h.

Table 2 ORR analysis of the samples in O₂-saturated 0.1 M KOH at 1600 rpm

Sample	Eonset (V vs. SCE)	E1/2 (V vs. SCE)	JL (mA cm^−2)	Tafel slope (mV dec^−1)	JK (mA cm^−2)
CoO/SnO2/NG-500	−0.045	−0.147	5.77	33	22.11
CoO/NG-500	−0.118	−0.237	4.68	78	12.90
SnO2/NG-500	−0.120	−0.267	4.00	64	11.21
NG-500	−0.116	−0.254	3.56	83	7.75
r-GO-500	−0.122	−0.231	2.47	91	6.53
20 wt% Pt/C	0.020	−0.137	6.39	51	27.13

---

To gain insight into the ORR kinetics for CoO/SnO₂/NG-500, rotating disk electrode (RDE) measurements were performed at rotation speeds ranging from 200 to 2500 rpm, and ​ the corresponding LSV polarization curves obtained are shown in Fig. S4c.† Increasing limiting current densities were observed with the increasing rotation speeds due to the shortened diffusion distance at high rotating speeds.^9,17,29 In Fig. S4d,† the Koutecky–Levich (K–L) plots at various potentials exhibited good linearity and the fitting lines were almost parallel, suggesting first-order reaction kinetics and similar electron-transfer numbers (n) per oxygen molecule at different potentials.^9,17,29 The n value for CoO/SnO₂/NG-500 varied from 3.89 to 3.98 according to the K–L plots in a potential range of −0.24 to −0.72 V vs. SCE (Fig. S4e†),^9,17 implying that oxygen was directly reduced to water via a four-electron pathway with high efficiency,^9,17,24,25 similar to that for Pt/C.

To further reveal the ORR catalytic mechanism, rotating ring disk electrode measurements were taken in O₂-saturated 0.1 M KOH solution at 1600 rpm and 10 mV s⁻¹ at a polarization potential of 0.2 V vs. SCE (Fig. 4e). The LSV polarization curve of CoO/SnO₂/NG-500 resembled that of Pt/C, suggesting a similar mechanism for the two catalysts in the ORR.²⁹ CoO/SnO₂/NG-500 presented a larger disk J_L and a more positive onset potential than those of CoO/NG-500, SnO₂/NG-500, NG-500 and r-GO-500, in accordance with the LSV results (Fig. 4b). The electron transfer number per oxygen molecule and HO₂⁻ yield were calculated from the following equations according to the disk and ring polarization curves:¹⁷

where, i_D and i_R are the disk current and ring current, respectively, and N (0.37) is the efficiency of Pt ring.

The results are presented in Fig. 4f and g. r-GO-500 had a low n value (3.1) and a high HO₂⁻ yield (∼45%) (Fig. 4f and g), suggesting r-GO-500 facilitated a 2 + 4 electron-transfer route for the ORR.^24,25 In contrast, the n value for CoO/SnO₂/NG-500 was 4.0, indicating it had a four-electron-transfer mechanism in the ORR. The n values for CoO/NG-500, SnO₂/NG-500 and NG-500 are also given in Fig. 4f. The HO₂⁻ yield (<5%) of CoO/SnO₂/NG-500 was lower than that of Pt/C, suggesting the hybrid was of excellent stability under corrosive conditions.⁵⁹

In order to determine the long-term durability of CoO/SnO₂/NG-500, the current–time chronoamperometric response was measured at −0.27 V vs. SCE at 1600 rpm and 10 mV s⁻¹ in O₂-saturated 0.1 M KOH over 5 h. Fig. 4h shows that CoO/SnO₂/NG-500 retained 97.2% of the initial current after continuous measurements, whereas the current for Pt/C decreased to 67.8%, revealing CoO/SnO₂/NG-500 had better long-term stability. The better operational durability of the composite was attributed to the following. First, graphene, serving as the matrix in the catalyst, was stable enough in the corrosive alkaline medium,⁶⁰ while the active carbon used as the support in the 20% Pt/C sample, by contrast, was not stable in the corrosive conditions. Second, SnO₂ in the catalyst had abundant oxygen vacancies, which could adsorb oxygen molecules on the surface, accommodating the volume change during the electrocatalytic process.⁴⁸ Third, the synergistic coupling effects between metal oxides and the NG substrate not only hindered the aggregation of the metal oxides, but also contributed a lot to the remarkable stability due to the metal–N bonding.^42,46,60 The overall remarkable ORR performance for CoO/SnO₂/NG-500 mainly arose from the porous structure and nitrogen doping in the composite, which provided a feasible pathway to transfer mass and electrons, the role of oxygen vacancy in SnO₂, and the synergistic effects between metal oxides and NG.^28,39

3.3. Oxygen evolution reaction evaluation

To further investigate the bifunctional properties of CoO/SnO₂/NG-500 as an oxygen electrode catalyst, oxygen evolution reaction measurements were performed in 1 M KOH solution using a rotating disk electrode at 1600 rpm and 10 mV s⁻¹. CoO/NG-500, SnO₂/NG-500, NG-500, r-GO-500, Pt/C and RuO₂ electrodes were also prepared for comparison. As shown in Fig. 5a, NG-500, r-GO-500 and Pt/C displayed poor OER activities. The OER onset potential of CoO/SnO₂/NG-500 in Fig. 5a reached 0.42 V vs. SCE, which was more negative than RuO₂ (0.47 V vs. SCE), Pt/C (0.61 V vs. SCE), CoO/NG-500 (0.54 V vs. SCE) and SnO₂/NG-500 (0.65 V vs. SCE) (Fig. 5a). The overpotentials (η) at a current density of 10 mA cm⁻², a value for solar water-splitting devices,¹⁰ were compared for CoO/SnO₂/NG-500 and RuO₂. CoO/SnO₂/NG-500 generated a current density of 10 mA cm⁻² at an overpotential of 285 mV, which was lower than RuO₂ (331 mV) (Table 3), and comparable with previously reported state-of-the-art catalysts, including CoFe₂O₄/PANI–MWCNT (314 mV),¹⁹ CoSn(OH)₆ (274 mV),²¹ NCG–CoO (334 mV),⁵⁸ CNCNT (310 mV)⁶¹ and so on (see Table S4†).

Fig. 5 (a) OER linear sweep voltammogram curves of the investigated samples at 1600 rpm and 10 mV s⁻¹; (b) OER Tafel plots of the samples; (c) current–time plot of CoO/SnO₂/NG-500 (inset: cycle stability measurements by LSV polarization curves before and after 200 cycles); (d) polarization curve for CoO/SnO₂/NG-500 in the whole region of ORR and OER.

Table 3 Comparison of bifunctional catalysts

	ΔEORR–OER [V]	Reference
CoO/SnO2/NG-500	0.72	This work
Fe/C/N	0.76	62
IrO2/C	0.76	18
Co/N-CNTs-700	0.78	12
Co3O4/N-CNT–GNR	0.80	18
Co/C/N	0.80	62
CoS2(400)/N,S-GO	0.82	16
NCNT	0.83	22
NCNT	0.84	4
Co(OH)2	0.87	33
Pt/C	0.90	18
RuO2	0.98	33

---

The kinetics for oxygen evolution with the composite were examined using Tafel plots (Fig. 5b). The Tafel slope of CoO/SnO₂/NG-500 was 41 mV dec⁻¹, lower than RuO₂ (71 mV dec⁻¹), CoO/NG-500 (58 mV dec⁻¹), SnO₂/NG-500 (68 mV dec⁻¹) and NG-500 (92 mV dec⁻¹). This value was comparable with those of other Co-based OER catalysts, including CoP/rGO (66 mV dec⁻¹),² NCNT/Co_xMn_1−xO (40 mV dec⁻¹),³ Co(OH)_x–NCNT (36 mV dec⁻¹),¹⁵ Co₃O₄/N-CNT–GNR (44 mV dec⁻¹),¹⁸ CoFe₂O₄/PANI–MWCNT (30.7 mV dec⁻¹),¹⁹ CoSe (40 mV dec⁻¹),⁵⁵ and NCG–CoO (71 mV dec⁻¹),⁵⁸ suggesting this hybrid was favorable for OER kinetics. Furthermore, the stability of CoO/SnO₂/NG-500 as an OER catalyst was evaluated by chronoamperometric response; 90.3% of the original current was retained after 5 h (Fig. 5c), and 83.3% of the initial catalytic current remained after 200 continuous cycles (see the inset in Fig. 5c), implying CoO/SnO₂/NG-500 was of outstanding OER durability.

The overall oxygen electrode activity was then assessed by the potential difference of ORR and OER metrics (ΔE = E_j=10 − E_j=−3).¹² The OER potential was often measured at a current density of 10 mA cm⁻² (E_j=10), and the ORR potential was usually recorded at a current density of −3 mA cm⁻² (E_j=−3).³³ The smaller the ΔE, the more ideal is the catalyst as a reversible oxygen electrode. Remarkably, the ΔE of CoO/SnO₂/NG-500 was 0.72 V (Fig. 5d), much lower than the values for recently reported state-of-the-art bifunctional catalysts, such as Co/N-CNT (ΔE = 0.78 V),¹² Fe/C/N (ΔE = 0.76 V),⁶² and IrO₂/C (ΔE = 0.87 V).⁶² Table 3 shows a detailed comparison of different bifunctional catalysts, further highlighting the superior bicatalytic activity of the CoO/SnO₂/NG-500 catalyst. The outstanding and selective catalytic activity and long-term durability of CoO/SnO₂/NG-500 in both ORR and OER makes it a promising candidate for the oxygen electrode catalyst in energy storage and conversion devices.

Based on the relationship properties to structure and composition, the superior bifunctional performance of CoO/SnO₂/NG-500 should be attributed to the following. First, as shown by its nitrogen adsorption–desorption isotherm, the hybrid was of porous structure and high specific surface area, which facilitated quick mass and electron transfer.^48,50 Second, the presence of SnO₂ in the composite hindered the aggregation of CoO, and vice versa, supported by the higher specific surface area of CoO/SnO₂/NG-500 than those of single metal oxide/NG-500. The larger specific surface area enabled the more exposure of the active sites for the oxygen catalysis reaction. Third, the presence of Sn in the sample offered more active sites for the interfacial adsorption of alkaline electrolyte.⁴² Owing to the formation of Sn in the composite, SnO₂ nanoparticles could anchor closely to the surface of NG by means of N–Sn interaction,^42,63 resulting in the outstanding operational stability of the catalyst. At the same time, the Sn nanoparticles prevented the shrinkage of SnO₂, thus expanding the contact area between the hybrid and electrolyte, and providing extra active sites.⁴⁸ In addition, Sn played a role in the highly active OER performance of the sample due to the interconversion of Sn and SnO₂ during continuous cycling.^46,48,64 Also, both the oxygen vacancies in the composite, confirmed by Raman spectra, and 1.7 at% pyridinic N doping, shown by XPS, afforded more active sites, and resulted in enhanced ORR and OER performance.^1,24,44 Last but not least, the synergistic effects between CoO/SnO₂ and heteroatom-doped graphene not only facilitated electron transfer, consequently enhancing electrocatalytic activities, but also provided a guarantee for the long-term working stability.

4. Conclusions

In summary, a novel and highly efficient catalyst for use in oxygen electrodes was prepared by introducing CoO and SnO₂ into nitrogen-doped porous graphene nanosheets via a ‘green’ method. The fabricated hybrid achieved enhanced ORR and OER performances. The superior properties originated from the porous structure and high specific surface area of the hybrid, oxygen vacancies, heteroatom doping and the strong synergistic effects of the composites. Owing to the ease of the preparation method, inexpensive materials and highly effective oxygen catalytic activity, the catalyst is a promising candidate as oxygen electrode catalyst in energy storage and conversion devices.

Acknowledgements

This work was supported by the Nature Science Foundation of China (NSFC-No. 21273024) and Nature Science Foundation of Jinlin Province, China (No. 20160101298 JC).

Notes and references

1. R. Gao, Z. Y. Li, X. L. Zhang, J. C. Zhang, Z. B. Hu and X. F. Liu, ACS Catal., 2016, 6, 400–406.
2. L. Jiao, Y. X. Zhou and H. L. Jiang, Chem. Sci., 2016, 7, 1690–1695.
3. X. Liu, M. Park, M. G. Kim, S. Gupta, X. J. Wang, G. Wu and J. Cho, Nano Energy, 2016, 20, 315–325.
4. H. W. Park, D. U. Lee, M. G. Park, R. Ahmed, M. H. Seo, L. F. Nazar and Z. W. Chen, ChemSusChem, 2015, 8, 1058–1065.
5. L. J. Li, C. Liu, G. He, D. L. Fan and A. Manthiram, Energy Environ. Sci., 2015, 8, 3274–3282.
6. N. Jung, H. Shinc, M. Kimd, I. Jang, H. J. Kim, J. H. Jang, H. Kim and S. J. Yoo, Nano Energy, 2015, 17, 152–159.
7. V. T. T. Ho, N. G. Nguyen, C. J. Pan, J. H. Cheng, J. Rick, W. N. Su, J. F. Lee, H. S. Sheu and B. J. Hwang, Nano Energy, 2012, 1, 687–695.
8. J. T. Zhang and L. M. Dai, ACS Catal., 2015, 5, 7244–7253.
9. R. Li, Z. D. Wei and X. L. Gou, ACS Catal., 2015, 5, 4133–4142.
10. K. G. Qu, Y. Zheng, S. Dai and S. Z. Qiao, Nano Energy, 2016, 19, 373–381.
11. S. Chen, J. J. Duan, Y. Zheng, X. M. Chen, X. W. Du, M. Jaroniec, and S. Z. Qiao, Energy Storage Materials, 2015, vol. 1, pp. 17–24.
12. Y. Y. Liu, H. L. Jiang, Y. H. Zhu, X. L. Yang and C. Z. Li, J. Mater. Chem. A, 2016, 4, 1694–1701.
13. J. X. Xu, Q. M. Yu, C. X. Wu and L. H. Guan, J. Mater. Chem. A, 2015, 3, 21647–21654.
14. Y. Y. Liang, H. L. Wang, P. Diao, W. Chang, G. S. Hong, Y. G. Li, M. Gong, L. M. Xie, J. G. Zhou, J. Wang, T. Z. Regier, F. Wei and H. J. Dai, J. Am. Chem. Soc., 2012, 134, 15849–15857.
15. J. E. Kim, J. Lim, G. Y. Lee, S. H. Choi, U. N. Maiti, W. J. Lee, H. J. Lee and S. O. Kim, ACS Appl. Mater. Interfaces, 2016, 8, 1571–1577.
16. P. Ganesan, M. Prabu, J. Sanetuntikul and S. Shanmugam, ACS Catal., 2015, 5, 3625–3637.
17. L. Lin, M. Li, L. Q. Jiang, Y. F. Li, D. J. Liu, X. Q. He and L. L. Cui, J. Power Sources, 2014, 268, 269–278.
18. X. Y. Lu, H. M. Chan, C. L. Sun, C. M. Tseng and C. Zhao, J. Mater. Chem. A, 2015, 3, 13371–13376.
19. Y. Liu, J. Li, F. Li, W. Z. Li, H. H. Yang, X. Y. Zhang, Y. S. Liu and J. T. Ma, J. Mater. Chem. A, 2016, 4, 4472–4478.
20. P. X. Li, R. G. Ma, Y. Zhou, Y. F. Chen, Z. Z. Zhou, G. H. Liu, Q. Liu, G. H. Peng, Z. H. Liang and J. C. Wang, J. Mater. Chem. A, 2015, 3, 15598–15606.
21. F. Song, K. Schenk and X. L. Hu, Energy Environ. Sci., 2016, 9, 473–477.
22. K. Elumeeva, J. Masa, F. Tietz, F. K. Yang, W. Xia, M. Muhler and W. Schuhmann, ChemElectroChem, 2016, 3, 138–143.
23. H. W. Park, D. U. Lee, M. G. Park, R. H. Ahmed, M. H. Seo, L. F. Nazar and Z. W. Chen, ChemSusChem, 2015, 8, 1058–1065.
24. Y. Nie, L. Li and Z. D. Wei, Chem. Soc. Rev., 2015, 44, 2168–2201.
25. L. M. Dai, Y. H. Xue, L. T. Qu, H. J. Choi and J. B. Baek, Chem. Rev., 2015, 115, 4823–4892.
26. C. Domínguez, Fr. J. P. Alonso, M. A. Salam, S. A. A. Thabaiti, A. Y. Obaid, A. A. Alshehri, J. L. G. D. L. Fuente, J. L. G. Fierro and S. Rojas, Appl. Catal., B, 2015, 162, 420–429.
27. B. Merzougui, S. Bukola and R. Zaffou, Mater. Today, 2016, 3, 691–695.
28. J. J. Duan, S. Chen, M. Jaroniec and S. Z. Qiao, ACS Catal., 2015, 5, 5207–5234.
29. B. L. Chen, R. Li, G. P. Ma, X. L. Gou, Y. Q. Zhu and Y. D. Xia, Nanoscale, 2015, 7, 20674–20684.
30. Z. Wang, P. J. Zuo, L. Q. Fan, J. N. Han, Y. P. Xiong and G. P. Yin, J. Power Sources, 2016, 311, 68–80.
31. L. Yao, H. X. Zhong, C. W. Deng, X. F. Li and H. M. Zhang, J. Energy Chem., 2016, 25, 153–157.
32. X. M. Zhou, Z. M. Xia, Z. M. Tian, Y. Y. Ma and Y. Q. Qu, J. Mater. Chem. A, 2015, 3, 8107–8114.
33. Y. Zhan, G. J. Du, S. L. Yang, C. H. Xu, M. H. Lu, Z. L. Liu and J. Y. Lee, ACS Appl. Mater. Interfaces, 2015, 7, 12930–12936.
34. Q. M. Su, J. Xie, J. Zhang, Y. J. Zhong, G. H. Du and B. S. Xu, ACS Appl. Mater. Interfaces, 2014, 6, 3016–3022.
35. J. F. Chang, L. Liang, C. Y. Li, M. L. Wang, J. J. Ge, C. P. Liu and W. Xing, Green Chem., 2016, 18, 2287–2295.
36. J. Liu, L. H. Jiang, B. S. Zhang, J. T. Jin, D. S. Su, S. L. Wang and G. Q. Sun, ACS Catal., 2014, 4, 2998–3001.
37. X. M. Ge, A. Sumboja, D. Wuu, T. An, B. Li, F. W. T. Goh, T. S. A. Hor, Y. Zong and Z. L. Liu, ACS Catal., 2015, 5, 4643–4667.
38. B. S. Wu, H. Z. Zhang, W. Zhou, M. R. Wang, X. F. Li and H. M. Zhang, ACS Appl. Mater. Interfaces, 2015, 7, 23182–23189.
39. Y. F. Zhao, B. Sun, X. D. Huang, H. Liu, D. W. Su, K. N. Sun and G. X. Wang, J. Mater. Chem. A, 2015, 3, 5402–5408.
40. T. Toupance, H. E. Hamzaoui, B. Jousseaume, H. Riague, I. Saadeddin, G. Campet and J. Brötz, Chem. Mater., 2006, 18, 6364–6372.
41. B. Mallesham, P. Sudarsanam, G. Raju and B. M. Reddy, Green Chem., 2013, 15, 478–489.
42. H. P. Cong, S. Xin and S. H. Yu, Nano Energy., 2015, 13, 482–490.
43. Z. F. Li, Q. Liu, Y. D. Liu, F. Yang, L. Xin, Y. Zhou, H. Y. Zhang, L. Stanciu and J. Xie, ACS Appl. Mater. Interfaces, 2015, 7, 27087–27095.
44. G. Pacchioni, ChemPhysChem, 2003, 4, 1041–1047.
45. D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii, Z. Z. Sun, A. Slesarev, L. B. Alemany, W. Lu and J. M. Tour, ACS Nano, 2010, 4, 4806–4814.
46. Z. Li, J. Ding, H. L. Wang, K. Cui, T. Stephenson, D. Karpuzov and D. Mitlin, Nano Energy, 2015, 15, 369–378.
47. Q. H. Tian, Z. X. Zhang, L. Yang and S. Hiranob, J. Mater. Chem. A, 2014, 2, 12881–12887.
48. B. Luo, B. Wang, M. H. Liang, J. Ning, X. L. Li and L. J. Zhi, Adv. Mater., 2012, 24, 1405–1409.
49. H. Y. Zhu, J. C. Zhao, J. W. Liu, X. Z. Yang and Y. N. Shen, Chem. Mater., 2006, 18, 3993–4001.
50. F. H. Du, Y. S. Liu, J. Long, Q. C. Zhu, K. X. Wang, X. Wei and J. S. Chen, Chem. Commun., 2014, 50, 9961–9964.
51. R. X. Shen, Y. Z. Hong, J. J. Stankovich, Z. Y. Wang, S. Dai and X. B. Jin, J. Mater. Chem. A, 2015, 3, 17635–17643.
52. M. R. Gao, X. Cao, Q. Gao, Y. F. Xu, Y. R. Zheng, J. Jiang and S. L. Yu, ACS Nano, 2014, 8, 3970–3978.
53. R. H. Wang, C. H. Xu and J. M. Lee, Nano Energy, 2016, 19, 210–221.
54. C. H. Xu, J. Sun and L. Gao, Nanoscale, 2012, 4, 5425–5430.
55. M. Liao, G. F. Zeng, T. T. Luo, Z. Y. Jin, Y. J. Wang, X. M. Kou and D. Xiao, Electrochim. Acta, 2016, 194, 59–66.
56. Q. H. Tian, Z. X. Zhang, L. Yang and S. I. Hirano, J. Mater. Chem. A, 2014, 2, 12881–12887.
57. C. Peng, L. Gao and S. W. Yang, Chem. Commun., 2007, 4372–4374.
58. S. Mao, Z. H. Wen, T. Z. Huang, Y. Hou and J. H. Chen, Energy Environ. Sci., 2014, 7, 609–616.
59. N. Ramaswamy, U. Tylus, Q. Y. Jia, and S. Mukerjee, Science to Coordination Chemistry, 2013, vol. 135, pp. 15443–15449.
60. J. J. Duan, S. Chen, M. Jaroniec and S. Z. Qiao, ACS Catal., 2015, 5, 5207–5234.
61. Z. L. Wang, S. Xiao, Z. L. Zhu, X. Long, X. L. Zheng, X. H. Lu and S. H. Yang, ACS Appl. Mater. Interfaces, 2015, 7, 4048–4055.
62. Y. Zhao, K. Kamiya, K. Hashimoto and S. Nakanishi, J. Phys. Chem. C, 2015, 119, 2583–2588.
63. L. W. Ji, Z. K. Tan, T. Kuykendall, E. J. An, Y. B. Fu, V. Battaglia and Y. G. Zhang, Energy Environ. Sci., 2011, 4, 3611–3616.
64. Q. H. Tian, Z. X. Zhang, L. Yang and S. Hirano, J. Mater. Chem. A, 2014, 2, 12881–12887.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra07489h

---