Spinel cobalt–manganese oxide supported on non-oxidized carbon nanotubes as a highly efficient oxygen reduction/evolution electrocatalyst

Ting Ma, Chun Li, Xiang Chen, Fangyi Cheng* and Jun Chen
Key Laboratory of Advanced Energy Materials Chemistry (Ministry of Education), College of Chemistry, Nankai University, Tianjin 300071, China. E-mail: fycheng@nankai.edu.cn; Fax: +86-22-23504486; Tel: +86-22-23506808

Received 28th June 2017 , Accepted 26th July 2017

First published on 16th August 2017


We report an in situ preparation of ultrasmall Co–Mn–O spinel nanoparticles (4.4 nm) supported on non-oxidized carbon nanotubes (denoted as CMO@CNTs) as a bifunctional catalyst for oxygen reduction and evolution reactions (ORR/OER). The preparation process involves the oxidization of divalent metal ions under alkaline conditions and the decomposition of nitrates in aqueous solution containing dispersed non-oxidized CNTs. The synthesized CMO nanoparticles strongly couple with the non-oxidized CNTs, which facilitates electron transfer and improves the catalytic activity. Other composites such as CMO@reduced graphene oxide, CMO@Vulcan X-72R, CMO@oxidized CNTs, and a physical mixture of CMO and CNTs were also prepared for comparison. Remarkably, CMO@CNTs exhibit a half wave potential of 0.91 V in 1 M KOH and higher kinetic current and better catalytic durability than Pt/C. Moreover, CMO@CNTs afford an electrocatalytic OER current density of 10 mA cm−2 at a low potential of 1.5 V and a small Tafel slope of 81.1 mV dec−1. Furthermore, CMO@CNTs display lower discharge/charge overpotential and more stable voltage plateau on cycling than that of Pt/C when employed as a cathode material in rechargeable Zn–air cells. This work indicates that CMO@CNTs are a promising, cheap and efficient bifunctional ORR/OER electrocatalyst for rechargeable metal–air batteries.


Introduction

Oxygen reduction/evolution reactions (ORR/OER) play a key role in energy conversion and storage technologies such as metal–air batteries, fuel cells, and solar or electricity-driven water splitting.1 However, the ORR is sluggish in nature and usually requires noble-metal-based catalysts, which are scarce and expensive.2 On the other side, the OER is of great importance in solar fuel synthesis.3 It is desirable to exploit highly efficient bifunctional non-precious metal catalysts for ORR/OER. Spinel oxides can be used as alternative low-cost electrocatalysts for both ORR and OER.4 Our group has developed a facile and rapid method to prepare nanocrystalline spinel CoxMn3–xO4, which can catalyze the ORR/OER and show compositional and phase-dependent performance.5 Nevertheless, the spinel oxides suffer from poor electrical conductivity and thus limited electrocatalytic activity. Combining them with a conducting additive or supporting them on a conducting substrate are efficient strategies to improve the performance of spinel oxides.6

Carbonaceous materials such as graphene and carbon nanotubes (CNTs) can be oxidized to provide surface defects and functional groups for strongly anchoring transition metal oxides.7 However, the electrical conductivity of the oxidized graphene or CNTs is inferior to that of the non-oxidized ones, which will limit the catalytic performance of the composite.8 To address this issue, mildly oxidized graphene/CNTs and nitrogenized carbon were used to prepare oxide composite catalysts.9 Non-oxidized graphene nanoflakes were also employed to immobilize Co3O4 nanofibers by noncovalent functionalization.10 Besides, strongly coupling CoMn2O4 nanodots with reduced graphene oxide (rGO) was shown to enhance the ORR/OER catalytic activity.6b The use of non-oxidized CNTs is expected to be beneficial but this has not been investigated to support cobalt–manganese-oxide spinels. Furthermore, the effect of the type of carbonaceous support on the catalytic activity of spinel oxides remains unknown.

Here we report the preparation of a nanocomposite of cubic cobalt–manganese oxide spinel (CMO) and non-oxidized carbon nanotubes (denoted as CMO@CNTs) as a bifunctional ORR/OER electrocatalyst. We also present a detailed investigation of the electrocatalytic performances of CMO supported on different carbon substrates, including Vulcan X-72R carbon (abbreviated as CMO@Vulcan), oxidized CNTs (CMO@oxCNT), and reduced graphene (CMO@rGO). The CMO nanoparticles were synthesized through a facile oxidation–precipitation and insertion–crystallization route, which allows the in situ formation of homogeneous spinel/carbon nanocomposites. A physical mixture of CMO and CNTs (CMO + CNTs) was also tested for comparison. Our results indicate that CMO@CNTs exhibit much higher ORR/OER catalytic activity than the other composites and CMO + CNTs. The combination of CMO and non-oxidized CNTs endows the nanocomposite with high electrical conductivity and a large electrochemically active surface area, thus resulting in remarkable ORR/OER activity in an alkaline electrolyte as well as lower charge/discharge overpotential and superior stability compared to Pt/C in Zn–air batteries.

Experimental section

Chemicals and materials

Cobalt nitrate (Co(NO3)2·6H2O), manganese nitrate (50 wt% aqueous solution), concentrated nitric acid (65 wt%), ethanol and ammonia water (25 wt%) were provided by Tianjin Guangfu Fine Chemical Research Institute. Nafion (5 wt%), KOH, and Teflon-coated carbon paper were purchased from Sigma-Aldrich. A carbon-supported Pt catalyst (Pt/C, 20 wt%) was supplied by Johnson Matthey. Vulcan X-72 was purchased from Carbot. Multiwall carbon nanotubes (metal impurity ≤100 ppm) were purchased from Beijing Cnano Technology.

Synthesis of oxidized carbon nanotubes and reduced graphene oxide

300 mg carbon nanotubes were soaked in 3 mL ethanol and dispersed in 20 mL deionized water in a glass beaker, where 20 mL concentrated nitric acid was added dropwise. The glass beaker was sealed with a soft piece of plastic film and the mixture was kept under magnetic stirring at 25 °C in a water bath for 12 h. Then, the mixture was slowly poured into 100 mL deionized water for diluting the nitric acid, and the suspension was filtered and washed with deionized water 10 times to form oxidized carbon nanotubes. Graphene oxide was prepared via a Staudenmaier method and reduced by a thermal exfoliation method to obtain reduced graphene oxide.11

Synthesis of cubic CMO

The synthesis of CMO was described in our previous report.5b In a typical preparation of cubic CMO spinel, 9 mL aqueous ammonia (25 wt%) was added dropwise to a mixed solution of 2 mL 0.5 M Co(NO3)2 aqueous solution and 6 mL deionized water under constant stirring in a glass beaker. The mixed solution was kept stirring for 60 min at 30 °C in a water bath. Then, 4 mL 0.5 M Mn(NO3)2 aqueous solution and 4 mL deionized water were added dropwise to the above mixed solution, which was kept stirring for another 60 min at 30 °C in a water bath. Afterwards, the glass beaker was transferred onto a heating plate, where the mixed solution was evaporated to fully decompose the nitrates by heating at 180 °C for 60 min to obtain CMO.

Synthesis of carbon supported CMO composite catalyst

0.233 g Vulcan X-72, multiwall carbon nanotubes or reduced graphene oxide was first soaked with 1–5 mL ethanol to facilitate dispersing the carbon substrates in aqueous solution. The following procedures were similar to the above-mentioned process for the preparation of CMO spinel.

Characterization

Scanning electron microscopy (SEM) images were obtained using a JEOL JSM-7500F microscope (voltage 5 kV) equipped with an energy dispersive X-ray spectroscopy (EDX) analyzer. Transmission electron microscopy (TEM) images were acquired using a Philips Tecnai G2 F20 transmission electron microscope (acceleration voltage, 200 kV). The chemical composition was determined by EDX and atomic absorption spectrometry (AAS, Hitachi 180-90 spectrophotometer). Elemental analysis of C, H and N was performed on an elementary analyzer (Vario EL cube, Elementar). Powder X-ray diffraction (XRD) patterns were measured on an X-ray diffractometer (Rigaku MiniFlex 600) with Cu Kα radiation. Raman spectra were measured using a confocal Raman microscope (DXR, Thermo Fisher Scientific) with excitation at 532 nm from argon-ion. Specific surface areas were calculated from the N2 adsorption/desorption isotherms at 77 K on a BELsorp-Mini. XPS spectra were measured using a PerkinElmer PHI 1600 ESCA system.

Electrochemical measurement

Electrochemical measurements were conducted by using a standard three-electrode electrochemical cell with a saturated calomel reference electrode (SCE), a platinum foil counter electrode and a glass carbon working electrode confined in a rotating ring-disk electrode (RRDE, Pt ring and glass carbon disk) or rotating disk electrode (RDE). The working electrode loaded with a catalyst was prepared as follows. 5 mg catalyst was ultrasonically dispersed in a solution of 1 mL ethanol and 17.5 mL Nafion (5 wt%, Sigma-Aldrich) for 30 min to obtain a well dispersed suspension. Then 4.6 μL of the suspension was loaded on the pre-dried glass carbon. The mass loading of the glass carbon was about 0.1 mg cm−2. The electrolyte (1 M or 0.1 M KOH) was kept at 25 °C using a water bath and purged with oxygen for 30 min before each measurement. An oxygen flow was maintained above the electrolyte during the tests.

Electrochemical data were recorded on a bipotentiostat (AFCBP1, PINE Instrument) with a rotating electrode system (PHYCHEM). The cyclic voltammograms (CVs) were collected at a potential scan rate of 20 mV s−1. Linear sweep voltammetry was performed at a potential scan rate of 5 mV s−1. The polarization curves in an oxygen-saturated electrolyte were corrected by subtracting the background current recorded in an Ar-saturated electrolyte and were iR-compensated. All potentials were calibrated with reference to a reversible hydrogen electrode (RHE).

The number of transferred electrons (n) and peroxide percentage (yperoxide) relative to the total product were calculated by the following equations:

 
image file: c7qi00367f-t1.tif(1)
 
image file: c7qi00367f-t2.tif(2)
where Id, Ir and N are the disk current, ring current and the current collection efficiency of the rotating ring-disk electrode, respectively.

The kinetic current density is calculated through mass–diffusion correction by using the following equation:

 
image file: c7qi00367f-t3.tif(3)
where ik, i and id are the kinetic, measured and diffusion-limited current densities, respectively.

The electrochemically active area of composite catalysts was estimated on the basis of the measurement of double layer capacitance. In the Ar-saturated electrolyte, the electrochemical response in the region from 0.24 V to 0.34 V was governed by the double layer capacitance without an obvious faradaic process, which was confirmed by the approximately rectangular shape of the CV curves and linear dependence of current vs. scanning rate.

Electrical conductivity measurements

To measure the electrical conductivity, a 200 mg sample was pressed at 20 MPa to obtain a 0.05 cm thick plate with a diameter of 1.3 cm. Two stainless steel electrodes (1.6 cm diameter, 0.2 cm thick) were placed on the two sides of the plate and held by an alligator clip. The electrical resistance test was performed using an Ametek Parstat 4000 system and the conductivity (ρ) was determined by the voltmeter–ammeter method.

Zn–air batteries assembly

Rechargeable Zn–air batteries were assembled by using a zinc plate as the anode, a catalyst supported on Teflon treated carbon paper as the air cathode, and 30 mL 6 M KOH with 0.2 M ZnO as the electrolyte. The air cathode was prepared by dropping a homogenous catalyst suspension onto the hydrophilic side of a 2 cm2 Teflon-treated carbon paper, which was dried at 100 °C for 3 h. The catalyst loading was about 2.0 mg cm−2. The charge/discharge test was performed by an Ametek Parstat 4000 system at room temperature.

Results and discussion

The preparation process of CMO@CNTs is schematically illustrated in Fig. 1. The synthesis includes three steps, following a phase-formation mechanism as in our previous report, which has revealed that cubic or tetragonal spinel can be selectively prepared by adjusting the addition order of precursors and that the cubic cobalt–manganese oxide spinel outperforms the tetragonal form in ORR electrocatalysis.5b Thus, in this study, we prepared the cubic spinel phase. Specifically, an aqueous cobalt nitrate solution was used as the precursor, which was mixed with well dispersed, ethanol-treated CNTs. Then, the addition of NH3·H2O resulted in the room-temperature oxidation precipitation of Co2+ in air (Reaction 1: Co2+ + NH3 + O2 → Co(NH3)63+). Then, manganese nitrate solution was added, forming a brown MnOOH precipitation (Reaction 2: Mn2+ + OH + O2 → MnOOH). Meanwhile, MnOOH reacts with Co(NH3)63+ to form an intermediate, Mn7O13 (Reaction 3: MnOOH + Co(NH3)63+ → Mn7O13 + Co(NH3)62+). Finally, the spinel CMO was obtained by heating at 180 °C.
image file: c7qi00367f-f1.tif
Fig. 1 Schematic synthesis of the CMO@CNTs composite.

Fig. 2a shows the XRD pattern of the obtained neat CMO and CMO@CNTs, which can be assigned to the cubic cobalt manganese spinel oxide (Joint Committee on Powder Diffraction Standards, JCPDS card no. 23-1237). The broad peaks indicate the nanocrystalline feature and the coexistence of CMO and CNTs. The Co[thin space (1/6-em)]:[thin space (1/6-em)]Mn ratio of CMO@CNTs is 1[thin space (1/6-em)]:[thin space (1/6-em)]2, as confirmed by the EDX spectrum (Fig. 2b inset). The morphology of CMO@CNTs was observed by SEM and TEM (Fig. 2b and c), showing the formation of nanoparticles with an average size of 4.4 nm on CNTs. Almost no nanoparticles were detached from the CNTs. The high-resolution TEM shows the lattice fringes of the nanoparticles, corresponding to the (111) and (311) planes (Fig. 2d). Compared with neat CMO, the particle size of CMO in CMO@CNTs is much smaller (4.4 versus ∼10 nm), which is due to the presence of the CNT substrate that suppresses the aggregation of nanoparticles.


image file: c7qi00367f-f2.tif
Fig. 2 (a) XRD patterns of CMO@CNTs and CNTs. (b) SEM image of CMO@CNTs. Inset shows the EDX spectrum. (c) Low-resolution and (d) high-resolution TEM images of CMO@CNTs.

The surface chemical state of the synthesized CMO@CNTs was investigated by X-ray photoelectron spectroscopy (XPS) analysis. The Mn 2p spectra can be fitted with three pairs of spin–orbit triplet deconvoluted peaks, indicating the oxidation states of Mn2+, Mn3+ and Mn4+ (Fig. 3a).12 Similarly, the Co 2p spectra consist of two pairs of spin–orbit doublets and two shakeup satellites, revealing the coexistence of Co2+ and Co3+ (Fig. 3b).12 These results indicate the presence of multiple valences of Mn and Co elements on the oxide surface. In addition, the peaks at ∼532 eV and ∼533.5 eV in the O 1s spectrum of CNTs belong to the oxygenic functional groups on the carbon surface, which shift towards lower binding energy in the O 1s spectrum of CMO@CNTs (Fig. 3c).13 This phenomenon indicates the strong interaction between metal cations and carbon surface functional groups.13 Such oxide/carbon coupling contributes to the good dispersion of the ultrasmall spinel particles. Furthermore, the percentage of peak area in the C 1s spectrum at 288 eV increases from 0% (CNTs) to 5.71% (CMO@CNTs), which suggests a higher content of oxygenic functional groups14 and benefits the interfacial interaction between CMO and non-oxidized CNTs through the C–O–metal bridge.15


image file: c7qi00367f-f3.tif
Fig. 3 (a) Mn 2p and (b) Co 2p XPS spectra of CMO@CNTs. (c) O 1s and (d) C 1s XPS spectra of CMO@CNTs and CNTs.

In order to investigate the effect of different carbonaceous supports on the performance of catalysts, CMO@oxCNTs, CMO@Vulcan, CMO@rGO, and CMO + CNTs were also synthesized. The XRD patterns of the above samples confirm the identical cubic cobalt manganese spinel phase (Fig. S1, ESI). The element compositions of CMO@CNTs, CMO@Vulcan, CMO@rGO and CMO@oxCNTs, determined by elemental analysis, EDX and Atomic Absorption Spectroscopy (AAS), are very similar to each other (Table S1, ESI). All the samples have similar morphology except for CMO + CNTs, which shows severe aggregation of the oxide particles and inhomogeneous mixing of the oxide with CNTs (Fig. S2 and S3, ESI). In addition, CMO@CNTs exhibit an electrical resistance of 0.26 Ω, smaller than that of the others (Fig. S4, ESI). The high electrical conductivity is attributed to the good dispersion of the oxide particles, interconnected conductive network, and the high electrical conductivity of non-oxidized CNTs. Note that the electrical resistance of CMO@oxCNTs is 19.04 Ω, suggesting a decrease of conductivity by using the oxidized CNTs. Comparing the full XPS spectra of CMO@oxCNTs and CMO@CNTs reveals an obvious increase in the oxygen content in the former composite, as summarized in Fig. S5, ESI. The effect of the increased oxygen content was also analyzed by Raman spectroscopy (Fig. S6, ESI). As the intensity of the D band was normalized, the G band decreased in oxidized CNTs and CMO@oxCNTs, indicating a higher degree of disordered carbon and implying lower conductivity.

The ORR catalytic performances of CMO@CNTs, CMO@rGO, CMO@oxCNTs, CMO@Vulcan, CMO + CNTs and commercial Pt/C were evaluated by cyclic and linear sweeping voltammetry (CV and LSV, Fig. 4a and b). CMO@CNTs show a much more positive ORR onset potential and higher cathodic currents, suggesting a higher activity. The kinetic current density (ik at 0.85 V) and half wave potential (Ehalf-wave) are compared in Fig. 4c. Among the catalysts, CMO@CNTs display a Ehalf-wave of 0.91 V and ik of 14.6 mA cm−2, showing only ∼15 mV lower Ehalf-wave as compared to Pt/C. The activity of CMO@CNTs is clearly higher than that of the others, following an order of CMO@CNTs > CMO@rGO > CMO@oxCNTs > CMO@Vulcan > CMO + CNTs. Remarkably, the ORR performance of CMO@CNTs surpasses most previously reported metal oxide catalysts, as shown in Table S2, ESI. Besides, CMO@CNTs exhibit a larger electrochemical specific area (ESA, determined by double layer capacitance measurement16) than other CMO-based nanocomposites (Fig. S7, ESI). Furthermore, the BET specific surface area of CMO@CNTs is 203.62 m2 g−1 (Fig. S8, ESI). Thus, the high ORR activity of CMO@CNTs could be ascribed to the high electrical conductivity and abundant surface active sites.


image file: c7qi00367f-f4.tif
Fig. 4 ORR electrocatalytic properties of CMO@CNTs, CMO@rGO, CMO@oxCNTs, CMO@Vulcan, CMO + CNTs and Pt/C. (a) Cyclic voltammograms of the prepared samples compared with Pt/C in oxygen-saturated 1 M KOH aqueous solution. The dashed lines were recorded in Ar. (b) LSV curves of the ORR at a potential sweeping rate of 5 mV s−1 and electrode rotating rate of 1600 rpm. (c) Comparison of ik (E = 0.85 V) and Ehalf-wave of different catalysts.

RRDE measurements were also performed to determine the ORR pathways of the CMO nanocomposites and a commercial Pt/C catalyst. Fig. 5a displays the disk and ring currents recorded at 1600 rpm in 1 M KOH. Fig. 5b shows the percentage of peroxide species with respect to the total oxygen reduction products and the electron transfer number (n). Over the potential range of 0.9–0.3 V, the calculated peroxide yield is less than 12% for all the composites except the physical mixture of CMO + CNTs. The average electron transfer number is around 3.7 for CMO@CNTs, CMO@rGO, and CMO@Vulcan, ∼3.65 for CMO@oxCNTs, and ∼3.6 for CMO + CNTs below 0.7 V (Fig. 5b). The results indicate that the ORR proceeds through apparent four-electron pathways on CMO-based electrocatalysts, similar to the case of Pt/C.


image file: c7qi00367f-f5.tif
Fig. 5 (a) Rotating ring-disk electrode (RRDE) polarization curves of the prepared samples and Pt/C at a rotating speed of 1600 rpm in oxygen saturated 1 M KOH aqueous solution. The dashed lines and solid curves correspond to ring and disk currents, respectively. (b) Calculated peroxide yield and electron transfer number.

The OER electrocatalysis of the prepared catalysts loaded on Teflon-coated carbon paper (TCP) was investigated in oxygen saturated 1 M KOH electrolyte. CMO@CNTs deliver a considerably low OER onset potential of 1.49 V and a current density of 10 mA cm−2 at 1.5 V (Fig. 6a). The Tafel slope of CMO@CNTs is 81.1 mV dec−1, which is smaller than that of other samples (Fig. 6b) and suggests higher OER kinetics. Based on the above results, CMO@CNTs exhibit high catalytic performance for both ORR and OER.


image file: c7qi00367f-f6.tif
Fig. 6 (a) Oxygen evolution currents of CMO@CNTs, CMO@rGO, CMO@oxCNTs, CMO@Vulcan, CMO + CNTs, and Pt/C in oxygen saturated 1 M KOH at a potential sweeping rate of 5 mV s−1, 1600 rpm. (b) Tafel plots of CMO@CNTs, CMO@rGO, CMO@oxCNTs, CMO@Vulcan, CMO + CNTs, and Pt/C.

The superb ORR/OER activities of CMO@CNTs motivated us to further investigate their applicability in Zn–air batteries. The prepared CMO@CNTs and the comparative 20 wt% Pt/C were loaded on TCP as the air cathode. Both TCP and the catalyst-loaded TCP electrodes show a porous structure and a film thickness of 150–200 μm (Fig. S9, ESI). The CMO@CNT electrodes exhibit an overpotential of 0.26 V for ORR and 0.27 V for OER at the catalytic current density of 10 mA cm−2 (Fig. 7a), outperforming Pt/C. The catalyst-loaded TCP electrodes were employed to assemble Zn–air batteries, as schematically shown in Fig. 7b. Fig. 7c displays the charge/discharge curves of the assembled Zn–air batteries on cycling. The lower charge plateau and higher discharge plateau of CMO@CNTs indicate that CMO@CNTs can better catalyze ORR/OER than Pt/C in the assembled rechargeable Zn–air batteries. Moreover, the overpotentials of the initial and 200th cycles of CMO@CNTs are 0.63 V and 0.71 V, lower than the values of 0.67 V and 0.96 V for Pt/C, respectively. The better rechargeability of CMO@CNTs over Pt/C is attributed to its superior bifunctional ORR/OER activity and durability. The decrease of the performance on Pt/C is likely due to the detachment of Pt nanoparticles from the carbon substrate and severe particle agglomeration upon repeated testing, as previously shown in our earlier reports.5b,6b In comparison, the strong binding of CMO on the CNT substrate favors structural integrity. These results indicate that CMO@CNTs are an effective and durable air electrode catalyst to build rechargeable Zn–air batteries.


image file: c7qi00367f-f7.tif
Fig. 7 (a) Oxygen electrode activities within the ORR and OER potential window of CMO@CNTs and Pt/C catalysts. (b) Schematic diagram of the assembled Zn–air battery. (c) Performance of rechargeable Zn–air batteries based on CMO@CNTs and Pt/C at 20 mA cm−2 (each cycle lasts 800 s).

Conclusions

In summary, we report a one-pot method to synthesize CMO@CNTs as a bifunctional electrocatalyst for ORR/OER. Investigating the effect of different carbonaceous supports on the catalytic activity of spinel oxides indicates that non-oxidized CNTs outperform oxidized CNTs, rGO, and Vulcan carbon. Remarkably, the CMO@CNTs nanocomposite features strong oxide/carbon interaction, homogeneous particle distribution, small spinel particle size and high surface areas, showing much higher electrical conductivity and electrocatalytic activity than the physical mixture of CMO + CNTs. Furthermore, CMO@CNTs exhibit higher electrocatalytic activity and stability than Pt/C as a cathode catalyst in rechargeable Zn–air batteries. The results suggest that the in situ generation of a spinel-type transition metal oxide supported on non-oxidized carbon nanostructures is an effective strategy for the design and preparation of cheap yet active ORR/OER catalysts.

Acknowledgements

This work was supported by MOST (no. 2017YFA0206700), NSFC (no. 21231005) and MOE (no. B12015 and IRT13R30).

References

  1. (a) M. K. Debe, Nature, 2012, 486, 43–51 CrossRef CAS PubMed; (b) B. C. Steele and A. Heinzel, Nature, 2001, 414, 345–352 CrossRef CAS PubMed; (c) F. Y. Cheng and J. Chen, Chem. Soc. Rev., 2012, 41, 2172–2192 RSC; (d) X. P. Han, F. Y. Cheng, C. C. Chen and J. Chen, Inorg. Chem. Front., 2016, 3, 866–871 RSC; (e) J. S. Lee, S. T. Kim, R. Cao, N. S. Choi, M. Liu, K. Lee and T. J. Cho, Adv. Energy Mater., 2011, 1, 34–50 CrossRef CAS; (f) F. H. B. Lima, J. Zhang, M. H. Shao, K. Sasaki, M. B. Vukmirovic, E. A. Ticianelli and R. R. Adzic, J. Phys. Chem. C, 2007, 111, 404–410 CrossRef CAS.
  2. (a) H. A. Gasteiger, S. S. Kocha, B. Sompalli and F. T. Wagner, Appl. Catal., B, 2005, 56, 9–35 CrossRef CAS; (b) Z. Y. Wu, X. X. Xu, B. C. Hu, H. W. Liang, Y. Lin, L. F. Chen and S. H. Yu, Angew. Chem., Int. Ed., 2015, 54, 8179–8183 CrossRef CAS PubMed; (c) K. Lei, L. Cong and X. Fu, Inorg. Chem. Front., 2016, 3, 928–933 RSC.
  3. (a) A. J. Bard and M. A. Fox, Acc. Chem. Res., 1995, 28, 141–145 CrossRef CAS; (b) M. W. Kanan and D. G. Nocera, Science, 2008, 321, 1072–1075 CrossRef CAS PubMed; (c) M. De Koninck and B. Marsan, Electrochim. Acta, 2008, 53, 7012–7021 CrossRef CAS; (d) X. Hu, X. Fu and J. Chen, Inorg. Chem. Front., 2015, 2, 1006–1010 RSC.
  4. (a) M. Hamdani, R. N. Singh and P. Chartier, Int. J. Electrochem. Sci., 2010, 5, 556–577 CAS; (b) Q. Liu, Y. Jiang, J. Xu, D. Xu, Z. Chang, Y. Yin, W. Liu and X. Zhang, Nano Res., 2014, 8, 576–583 CrossRef; (c) W. T. Hong, M. Risch, K. A. Stoerzinger, A. Grimaud, J. Suntivich and Y. Shao-Horn, Energy Environ. Sci., 2015, 8, 1404–1427 RSC; (d) S. W. Lee, C. Carlton, M. Risch, Y. Surendranath, S. Chen, S. Furutsuki, A. Yamada, D. G. Nocera and Y. Shao-Horn, J. Am. Chem. Soc., 2012, 134, 16959–16962 CrossRef CAS PubMed.
  5. (a) F. Cheng, J. Shen, B. Peng, Y. Pan, Z. Tao and J. Chen, Nat. Chem., 2011, 3, 79–84 CrossRef CAS PubMed; (b) C. Li, X. Han, F. Cheng, Y. Hu, C. Chen and J. Chen, Nat. Commun., 2015, 6, 7345 CrossRef CAS PubMed.
  6. (a) L. Wang, X. Zhao, Y. Lu, M. Xu, D. Zhang, R. S. Ruoff, K. J. Stevenson and J. B. Goodenough, J. Electrochem. Soc., 2011, 158, A1379–A1382 CrossRef CAS; (b) J. Du, C. Chen, F. Cheng and J. Chen, Inorg. Chem., 2015, 54, 5467–5474 CrossRef CAS PubMed.
  7. Y. Liang, Y. Li, H. Wang, J. Zhou, J. Wang, T. Regier and H. Dai, Nat. Mater., 2011, 10, 780–786 CrossRef CAS PubMed.
  8. (a) Y. Liang, H. Wang, P. Diao, W. Chang, G. Hong, Y. Li, M. Gong, L. Xie, J. Zhou, J. Wang, T. Z. Regier, F. Wei and H. Dai, J. Am. Chem. Soc., 2012, 134, 15849–15857 CrossRef CAS PubMed; (b) Y. Liang, H. Wang, J. Zhou, Y. Li, J. Wang, T. Regier and H. Dai, J. Am. Chem. Soc., 2012, 134, 3517–3523 CrossRef CAS PubMed.
  9. (a) A. Zhao, J. Masa, W. Xia, A. Maljusch, M. G. Willinger, G. Clavel, K. Xie, R. Schlogl, W. Schuhmann and M. Muhler, J. Am. Chem. Soc., 2014, 136, 7551–7554 CrossRef CAS PubMed; (b) T. Y. Ma, S. Dai, M. Jaroniec and S. Z. Qiao, J. Am. Chem. Soc., 2014, 136, 13925–13931 CrossRef CAS PubMed; (c) K. L. Pickrahn, S. W. Park, Y. Gorlin, H.-B.-R. Lee, T. F. Jaramillo and S. F. Bent, Adv. Energy Mater., 2012, 2, 1269–1277 CrossRef CAS; (d) Y. Gorlin and T. F. Jaramillo, J. Am. Chem. Soc., 2010, 132, 13612–13614 CrossRef CAS PubMed; (e) Y. Gorlin, C.-J. Chung, D. Nordlund, B. M. Clemens and T. F. Jaramillo, ACS Catal., 2012, 2, 2687–2694 CrossRef CAS; (f) Z. S. Wu, S. Yang, Y. Sun, K. Parvez, X. Feng and K. Mullen, J. Am. Chem. Soc., 2012, 134, 9082–9085 CrossRef CAS PubMed; (g) X. Liu, M. Park, M. G. Kim, S. Gupta, G. Wu and J. Cho, Angew. Chem., Int. Ed., 2015, 54, 9654–9658 CrossRef CAS PubMed.
  10. W. H. Ryu, T. H. Yoon, S. H. Song, S. Jeon, Y. J. Park and I. D. Kim, Nano Lett., 2013, 13, 4190–4197 CrossRef CAS PubMed.
  11. M. J. McAllister, J. L. Li, D. H. Adamson, H. C. Schniepp, A. A. Abdala, J. Liu, M. Herrera-Alonso, D. L. Milius, R. P. Car and I. A. R. Aksay, Chem. Mater., 2007, 19, 4396–4404 CrossRef CAS.
  12. L. Yu, L. Zhang, H. B. Wu, G. Q. Zhang and X. W. Lou, Energy Environ. Sci., 2013, 6, 2664–2671 CAS.
  13. (a) X. Li, H.-J. Zhang, H. Li, B. Zhao and J. Yang, J. Electrochem. Soc., 2014, 161, F925–F932 CrossRef CAS; (b) C. Zhang, M. Antonietti and T.-P. Fellinger, Adv. Funct. Mater., 2014, 24, 7655–7665 CrossRef CAS.
  14. (a) Y. Li, W. Zhou, H. Wang, L. Xie, Y. Liang, F. Wei, J. C. Idrobo, S. J. Pennycook and H. Dai, Nat. Nanotechnol., 2012, 7, 394–400 CrossRef CAS PubMed; (b) M. Xiao, J. Zhu, L. Feng, C. Liu and W. Xing, Adv. Mater., 2015, 27, 2521–2527 CrossRef CAS PubMed.
  15. G. Zhou, D. W. Wang, L. C. Yin, N. Li, F. Li and H. M. Cheng, ACS Nano, 2012, 6, 3214–3223 CrossRef CAS PubMed.
  16. S. Levine and A. L. Smith, Discuss. Faraday Soc., 1971, 52, 290–301 RSC.

Footnotes

Electronic supplementary information (ESI) available: XRD, SEM, and additional electrochemical characterization. See DOI: 10.1039/c7qi00367f
These authors contribute equally to this work.

This journal is © the Partner Organisations 2017