Ultrafine Pt nanoparticle decoration with CoP as highly active electrocatalyst for alcohol oxidation

Abstract

Improving the catalytic activity and stability of Pt-based electrocatalysts for the oxidation of alcohols is essential for their commercialization. In this study, CoP was introduced as a catalytic promoter material for Pt nanoparticles, to produce bifunctional catalysts for the oxidation of methanol and ethanol. The Pt–CoP/C-30% electrode shows current densities of 1300 mA mg_Pt⁻¹ for methanol oxidation and 857 mA mg_Pt⁻¹ for ethanol oxidation, which significantly overwhelm those of the commercially available Pt/C catalyst (current densities of 291 mA mg_Pt⁻¹ for methanol oxidation and 346 mA mg_Pt⁻¹ for ethanol oxidation, respectively). The bifunctional mechanism and electronic effect caused by the addition of CoP contribute to the excellent performance of Pt–CoP/C-30%. We conclude that Pt–CoP/C-30% can serve as an effective electrocatalyst for next generation high-performance fuel cells.

---

1. Introduction

In the past few years, there has been an increased interest in direct methanol fuel cells (DMFCs) and direct ethanol fuel cells (DEFCs) as promising power sources for micro and portable applications since these fuels have advantages such as being liquid at room temperature, having high energy density, and being easy to store and distribute.^1–3 To date, most reported efficient electrocatalysts for the methanol oxidation reaction (MOR) and ethanol oxidation reaction (EOR) are Pt-based owing to their relatively high catalytic activities.^4,5 However, the relative scarcity of platinum and the instability of the electrocatalysts are the main factors that limit the commercialization of Pt-based electrocatalysts in DMFCs and DEFCs.^6–8

Recently, researchers have made great efforts to replace Pt/C with other excellent electrocatalysts for MOR and EOR, and thus to meet the challenges for the commercial application of fuel cells.^9–11 Incorporation of a second metal (Ru, Pd, Au, etc.) with Pt has been extensively addressed, and the resulting electrocatalysts exhibited high exchange current density comparable to that of commercial Pt/C electrocatalyst.^12–17 The significant role in promoting catalytic activity and stability of those dual alloy electrocatalysts in MOR and EOR has been ascribed to the so-called bifunctional mechanism and the electronic effect caused by the addition of the second metal.¹² In addition, transition metal phosphides, such as Ni₂P,^18,19 have been suggested by Xing's group as catalytic promoter materials in Pt-based catalysts, showing remarkable catalytic activity and stability during the methanol oxidation process. The current density in a positive scan for Pt–Ni₂P/C (1430 mA mg_Pt⁻¹) in MOR is significantly higher than for PtNi/CNT (682 mA mg_Pt⁻¹), PtNiP/C (362 mA mg_Pt⁻¹) and Pt–Ni–P (360 mA mg_Pt⁻¹) electrocatalysts.^20–23 The promoting effect of Ni₂P in Pt/C was similar to that of incorporating a second metal with Pt, and the so-called bifunctional mechanism model can be written as the following reactions:²⁴

MP + H₂O → MP–OH + H⁺ + e⁻

Pt–CO + MP–OH → Pt + CO₂ + MP + H⁺ + e⁻

(M represents the transition metal.) Abundant oxygen-containing species were produced by MP through the decomposition of water, which were essential to oxidize the CO-like intermediate species.^25,26 Therefore, the adsorbed CO-like species on the surface of the noble metal will be rapidly oxidized by the neighbouring oxygen-containing species on the surface of the transition metal phosphides. Furthermore, the addition of transition metal phosphides could change the electronic properties of Pt through an electronic effect and thus weaken the binding energy of Pt and CO-like species. Therefore, the introduction of transition metal phosphides into noble metal catalysts would be an ideal choice to improve the catalytic activity and stability in the oxidation of alcohols. Though many researchers have verified that the addition of transition metal phosphides into Pt has significantly enhanced the catalytic activity and durability for MOR, there is almost no literature regarding the use of such materials as bifunctional catalysts for both MOR and EOR. It is well known that the oxidation reactions of alcohols are similar. Hence, it would be interesting to investigate integrated catalysts combining the merits of transition metal phosphides with Pt as bifunctional catalysts for both MOR and EOR.

In this paper, we demonstrate the use of CoP as a promoter in the Pt nanoparticle electrocatalyst (denoted as Pt–CoP/C hereafter) for MOR and EOR. The precursor CoP and noble metal Pt were co-deposited on high specific surface area carbon black supports via solvent refluxing. The obtained Pt–CoP/C-30% showed current densities of 1300 mA mg_Pt⁻¹ for MOR and 857 mA mg_Pt⁻¹ for EOR, which would allow significant reduction of the amount of noble metal Pt because of its high catalytic activity in the application. In addition, the decoration of Pt with CoP exhibited strong resistance to CO-like species poisoning during the alcohol oxidation. These fascinating properties of the Pt–CoP/C promote them to be applied as bifunctional catalysts for new generation DMFCs and DEFCs.

2. Experiment section

2.1 Reagents and chemicals

Tetrahydrate cobalt acetate (Co(CH₃COO)₂·4H₂O), sodium hypophosphite monohydrate (≥99.0%, NaH₂PO₂·H₂O) and hexachloroplatinic acid (H₂PtCl₆·6H₂O) were purchased from Aldrich Chemical Co. (USA). Vulcan XC-72 carbon was purchased from Cabot Co. (USA). Nafion solution (5%) was purchased from Dupont Co. (USA). Nitric acid (≥95.0%), sulfuric acid (≥95.0%) and ethanol (≥99.7%) were purchased from Beijing Chemical Co. (China). All the chemicals were of analytical grade and were used as received. High purity nitrogen (≥99.99%) and carbon monoxide (≥99.99%) were supplied by Wuhan Tuteng Co. Ltd. Deionized water (18.2 MΩ) was obtained from a Millipore-Q water purification system.

2.2 Synthesis of materials

In a typical preparation, Vulcan XC-72 carbon (abbreviated as C) was firstly pretreated in 5 mol L⁻¹ nitric acid at 110 °C for 6 h. Co₃O₄/C composite was prepared according to the reported method with minor modifications.²⁷ Firstly, 70 mg of pretreated C was dispersed into a solution containing 48 mL of absolute ethanol and 1.0 mL of deionized water with ultrasonication to form a homogeneous solution. Secondly, 83 mg of Co(CH₃COO)₂·4H₂O was dissolved in the above solution, and then 1.0 mL of ammonia (25 wt%) was gradually added under vigorous stirring. The obtained suspension was refluxed at 80 °C for 14 h. Finally, the solution was transferred into a Teflon-lined stainless steel autoclave and maintained at 150 °C for 3 h. After the autoclave had cooled down slowly to room temperature, the precipitate was collected by centrifugation and washed with water and ethanol several times. The obtained precipitate was then dried at 60 °C overnight to obtain Co₃O₄/C. The as-prepared Co₃O₄/C (0.1 g) and NaH₂PO₂·H₂O (0.5 g) were mechanically mixed in a quartz boat at room temperature. The mixture was calcined at 300 °C for 2 h at a heating rate of 2 °C min⁻¹ in N₂ atmosphere, and then washed with deionized water and dried at 60 °C overnight to obtain CoP/C-30%. CoP/C-X% (X = 15, 25, 40, 45; where X represents the mass loading of CoP on C) were also prepared by the same method. Pt–CoP/C composite with 20 wt% platinum was prepared by a solvent refluxing method. In a typical synthesis, 80 mg of CoP/C-30% was ultrasonically dispersed in a solution of 45 mL of ethylene glycol and 15 mL of deionized water to form a uniform suspension. One millilitre of 53 mg mL⁻¹ H₂PtCl₆·6H₂O solution was added into the above suspension to form a uniform ink. Subsequently, the sample was refluxed at 140 °C for 4 h. After cooling to room temperature naturally, the product was filtered, washed with deionized water, and dried at 60 °C overnight to obtain Pt–CoP/C-30%. The same mass loading of Pt but with different content of CoP was prepared by the same procedure to obtain Pt–CoP/C-X% (X = 15, 25, 40, 45). Home-made Pt/C (denoted Pt/C-H) was also prepared by the same method only without the addition of CoP. The content of Pt in all as-prepared samples was 20 wt%. A commercial Pt/C (20 wt% Pt) catalyst (Johnson Matthey Company; denoted Pt/C-JM) was used as a reference.

2.3 Characterization of materials

X-ray diffraction (XRD) measurements were performed by X'pert PRO diffractometer (PANalytical B.V.) using a Cu Kα radiation source as the X-ray source for excitation, operated at 40.0 kV and 40.0 mA within the 2θ range from 15° to 85°. Transmission electron microscopy (TEM) and energy dispersive X-ray detector spectrum (EDX) characterizations were conducted on a Tecnai G2F30 microscope (FEI Company, Holland) operating at 200 kV. The prepared Pt–CoP/C samples were dispersed in ethanol and dropped onto a holey carbon support grid for observation. X-ray photoelectron spectroscopy (XPS) measurements were performed on a VG Multilab 2000 XPS instrument (Thermo Electron Co.) with Al Kα X-ray radiation as the X-ray source for excitation to study the bonding energy of Pt-based catalysts. The binding energy from XPS was referenced to the C 1s spectrum of carbon support at 284.60 eV. Cyclic voltammetry (CV) tests were carried out with a CHI 750D potentiostat (Chenhua Co., Shanghai) in a conventional three electrode electrochemical cell using a glassy carbon electrode (3 mm, Chenhua Co., Shanghai) prepared with catalyst as the working electrode, twisted platinum wire as the auxiliary electrode, and a saturated calomel electrode (SCE) as the reference electrode. All tests were performed in an ultrahigh purity argon atmosphere and the electrolyte solutions were de-aerated with ultrahigh purity argon for about 0.5 h before testing. The prepared catalyst (2 mg) was dispersed into 1 mL of solution containing 0.2 mL of water, 0.77 mL of ethanol and 0.03 mL of Nafion (5 wt% solution in a mixture of lower aliphatic alcohols and water; Aldrich). Then 7 μL of catalyst liquid was drop-dried onto a glassy carbon electrode. Before electrochemical measurements, adsorption/desorption of hydrogen on catalysts surface was evaluated in 0.5 M H₂SO₄ solution and the electrode potential was set between −0.2 V and 1.0 V at a scan rate of 50 mV s⁻¹. To measure methanol and ethanol electrochemical oxidation reaction activity, cyclic voltammetry was performed between −0.20 V and 1.0 V in a solution of 1 M CH₃OH + 0.5 M H₂SO₄ and 1 M CH₃CH₂OH + 0.5 M H₂SO₄, respectively. The electrochemical active surface area (ECSA) and the tolerance to CO poisoning were estimated by the CO stripping test, assuming that the coulombic charge required for the oxidation of the CO monolayer was 420 μC cm⁻². CO (99.99% pure) was used to purge the 0.5 M H₂SO₄ electrolyte for 30 minutes and the working electrode was held at 0.2 V vs. SCE. The system was then purged with N₂ for 30 minutes to remove non-adsorbed CO before the measurements were made. The CO stripping experiment was performed in the potential range −0.2 V to 1.0 V.

3. Results and discussion

As depicted in Scheme 1, the Pt–CoP/C electrocatalyst was prepared through a simple heating under reflux, hydrothermal and low-temperature phosphatizing process. In order to study the architecture and distribution of Pt nanoparticles in the hybrid, we performed TEM examinations on the material. Fig. 1a and b show a typical TEM image and particle size histogram of Pt–CoP/C-30%. The TEM image clearly illustrates uniformly dispersed Pt nanoparticles on the carbon supports, with a diameter of 1.98 nm in a narrow size distribution. In addition, the TEM images and particle size distribution histograms of Pt/C-JM, Pt/C-H and Pt–CoP/C-X% compounds (X = 15, 45) are shown in Fig. S1.† Obviously, the particle size of Pt nanoparticles in Pt–CoP/C-30% composites is smaller than that in Pt/C-JM, Pt/C-H and Pt–CoP/C-X% (X = 15, 45). This indirectly indicates that Pt nanoparticles have large surface area and may have more active sites in the Pt–CoP/C-30% sample. The ultrafine and uniform distribution of Pt nanoparticles on carbon black is probably attributable to the carbon black support being pretreated with HNO₃.²⁹ Many carboxyl groups formed on the surface of carbon atoms on pretreatment with HNO₃, which can provide nucleation sites for the reduction of Pt nanoparticles via strong coordinate interactions between the Pt atom and carboxyl groups. Fig. 1c shows the HR-TEM image of Pt–CoP/C-30%. A finger lattice of 0.23 nm that corresponds to the (111) crystal plane of the face-centred cubic (fcc) Pt was observed. The finger lattice of 0.190 nm corresponds to the CoP (211) lattice. In addition, TEM images of CoP/C with different magnification are shown in Fig. S2.† CoP nanoparticles with a size of about 15 nm are uniformly deposited on carbon black supports; these are much larger than the Pt nanoparticles. In addition, we find that crystals of CoP are not obvious in the HR-TEM images of CoP/C; this probably contributes to the lack of obvious crystallinity of CoP in the HR-TEM image of Pt–CoP/C-30%. Moreover, we carried out EDX analysis to verify the existence of Co and P in Pt–CoP/C composite. As seen in Fig. S3,† the composite shows the presence of platinum, cobalt, phosphorus, carbon, and oxygen elements; the content of the composite is shown in Table S1.† We ascribe the presence of the element oxygen to the superficial oxidation of CoP, and similar phenomena also occur in other transition metal phosphides owing to their instability in air.^30–32 In order to further identify the composition and crystallinity of the material, we carried out XRD characterization; the relative results of CoP/C-30% and Pt–CoP/C-30% composite are shown in Fig. 1d. The broad diffraction peak of CoP/C-30% is located at 2θ values of ca. 23.7°, 31.6°, 36.4°, 45.1°, 46.2°, 48.2°, 52.3° and 56.8°, which can be indexed to the (101), (011), (102), (210), (112), (211), (103) and (301) planes of CoP (JCPDS 29-0497), respectively. No other diffraction peaks are detected after phosphating, indicating successful conversion of the precursor Co₃O₄ into CoP during the process of phosphating. The characteristic peaks located at ca. 39.8°, 46.2°, 67.5° and 81.3° can be ascribed to (111), (200), (220) and (311) planes of face-centred cubic (fcc) Pt (JCPDS 04-0802), respectively. After depositing Pt on CoP/C composites, the observed characteristic diffraction peaks of CoP are invisible in the XRD pattern of Pt–CoP/C, which is probably caused by their being covered by strong and broad peaks of the Pt. A similar phenomenon also occurred with Pt–Ni₂P/C, in which the peaks for Ni₂P were invisible after the decoration with Pt.¹⁸ Combining the results of EDX analysis of Pt–CoP/C, TEM analysis and the XRD analysis of CoP, it is safe to conclude that CoP exists in the Pt–CoP/C composite.

Scheme 1 Schematic illustration of the synthesis of Pt–CoP/C.

Fig. 1 (a) TEM image, (b) the corresponding particle size distribution histogram, (c) HR-TEM image and (d) XRD pattern of CoP/C-30% and Pt–CoP/C-30%.

Fig. 2 shows cyclic voltammograms of Pt–CoP/C-30% and Pt/C-JM catalysts in 0.5 M H₂SO₄ solution at a scan rate of 50 mV s⁻¹. Noticeably, the Pt–CoP/C-30% has a larger hydrogen adsorption–desorption peak area than Pt/C-JM. In addition, the cyclic voltammograms of Pt–CoP/C-X% (X = 15, 30, 45), Pt/C-JM and Pt/C-H are shown in Fig. S4.† It is obvious that the Pt–CoP/C-30% has the largest hydrogen adsorption–desorption peak area, which suggests Pt–CoP/C-30% has larger electrochemical active surface area (ECSA) than other as-prepared catalysts. The electrochemical activities of Pt-based catalysts for MOR and EOR were evaluated by cyclic voltammetry in a mixture of 1 M CH₃OH + 0.5 M H₂SO₄ and 1 M CH₃CH₂OH + 0.5 M H₂SO₄, respectively. The activities of Pt–CoP/C-30% and Pt/C-JM were compared under the same test conditions. In addition, the test was also conducted on the CoP/C electrode as a control. Fig. 3a and b, respectively, compare catalysts for the MOR and EOR in terms of mass activity. Both Pt–CoP/C-30% and Pt/C-JM catalysts exhibit a well-defined methanol and ethanol oxidation peak. However, the CoP/C electrode shows negligible activity for MOR and EOR. This demonstrates that the catalytic activity for the MOR and EOR is derived from Pt nanoparticles in the composites and the CoP itself has almost no catalytic activity for the MOR and EOR. Therefore, the function of CoP in the Pt-based catalysts is probably to promote the decomposition of water and thus produce more oxygen-containing groups to accelerate the oxidation of CO-like intermediate products.¹⁸ CV curves for both Pt–CoP/C-30% and Pt/C-JM show a similar alcohol oxidation current peak in the forward scan (I_f), and an oxidation peak in the backward scan (I_b) corresponding to the removal of the residual carbonaceous species formed in the forward scan. Hence, the ratio of I_f/I_b is regarded as a useful index to evaluate the tolerance of Pt-based catalysts against CO-like intermediate species.^33,34 In the CV curves, the ratio of I_f/I_b for the Pt–CoP/C-30% electrode (0.601) is a little less than that for Pt/C-JM (0.658) but it is superior to that for Pt/C-H (0.573) in the process of MOR. In addition, the ratio of I_f/I_b for the Pt–CoP/C-30% electrode (0.954) is larger than that of the Pt/C-H electrode (0.873) and comparable to that of the Pt/C-JM electrode (0.968) for the EOR process. Most important of all, the Pt–CoP/C-30% electrode shows current densities of 1300 mA mg_Pt⁻¹ for MOR and 857 mA mg_Pt⁻¹ for EOR, which is significantly higher than the values for the commercial Pt/C-JM catalyst (current densities of 291 mA mg_Pt⁻¹ for MOR and 346 mA mg_Pt⁻¹ for EOR). In addition, the mass activity values of Pt–CoP/C-X% (X = 15, 25, 40, 45) and CoP/C-30% for MOR (Fig. S5a†) and EOR (Fig. S5b†) were all inferior to those of Pt–CoP/C-30%. The high catalytic activity for the Pt–CoP/C-30% catalyst could be attributed to several factors, as discussed below. The electronic effect between Pt and CoP was not obvious when the content of CoP was 15% in the Pt–CoP/C electrocatalyst. When the content of CoP exceeded 30%, the nanoparticles showed partial aggregation. Therefore, the optimized loading of CoP in the composite was set at 30%, with which loading the CoP can be well-situated around the Pt nanoparticles and produce an appropriate electronic effect. Moreover, the Pt–CoP/C-30% has the Pt nanoparticles with the smallest size amongst Pt–CoP/C-X% catalysts (X = 15, 30, 45), offering more active sites. Moreover, we also evaluated the specific activity for Pt–CoP/C-30% and Pt/C-JM catalysts. The peak current of the Pt–CoP/C-30% is 3.48 times and 1.91 times higher than the Pt/C-JM for MOR and EOR, respectively (Fig. 3c and d). It is significant that the mass activity and specific activity of Pt–CoP/C-30% for MOR and EOR show performance competitive with the previously reported materials. As presented in Table S2,† the mass activity of Pt–CoP/C-30% (1300 mA mg_Pt⁻¹) for MOR is obviously superior to that of PtNiP/C nanoparticles (362 mA mg_Pt⁻¹),²² and dendritic Pt–Ni–P alloy nanoparticles (360 mA mg_Pt⁻¹),²³ and even higher than that of the recently reported Pt–Co₂P/C (1236 mA mg_Pt⁻¹).²⁸ Meanwhile, the catalytic activity of Pt–CoP/C-30% for EOR is also higher than that of other Pt-based materials (Table S3†).

Fig. 2 CV of Pt–CoP/C-30% and Pt/C-JM catalysts in 0.5 M H₂SO₄.

Fig. 3 (a and b) The mass activity of Pt/C-JM, Pt–CoP/C-30% and CoP/C-30% catalysts in solutions of 0.5 M H₂SO₄ + 1 M CH₃OH and 0.5 M H₂SO₄ + 1 M CH₃CH₂OH, respectively. (c and d) The specific activity of Pt/C-JM and Pt–CoP/C-30% catalysts in solutions of 0.5 M H₂SO₄ + 1 M CH₃OH and 0.5 M H₂SO₄ + 1 M CH₃CH₂OH, respectively.

Durability is vital for the application of catalyst. Therefore, we investigated the catalytic stability of Pt–CoP/C-30% and Pt/C-JM catalysts for MOR and EOR with the chronoamperometry (CA) technique. Fig. 4a and b show the normalized constant–potential curves of methanol and ethanol electro-oxidation on the Pt–CoP/C-30% and Pt/C-JM catalysts. Obviously, the Pt–CoP/C-30% exhibits considerable long-term stability compared with Pt/C-JM in MOR and EOR. This could be caused by the better anti-poisoning property of the Pt–CoP/C-30% composite, suggesting that CoP plays a vital role in improving the stability of the catalyst.

Fig. 4 (a and b) Normalized constant–potential curves of Pt/C-JM and Pt–CoP/C-30% catalysts at a fixed potential of 0.5 V in 0.5 M H₂SO₄ + 1 M CH₃OH and 0.6 V in 0.5 M H₂SO₄ + 1 M CH₃CH₂OH, respectively.

We also performed CO stripping experiments in order to confirm the anti-poisoning property and evaluate the ECSA for the as-prepared catalysts.³⁵ Fig. S6† shows the CO stripping of Pt/C-H, Pt/C-JM and Pt–CoP/C-X% (X = 15, 30, 45). The Pt–CoP/C-30% catalyst showed the largest ECSA by integration of the CO_ad stripping peak amongst all the tested samples (Table S4†), which is consistent with analysis of the cyclic voltammogram for Pt–CoP/C-X%. The onset potential of CO_ad stripping peaks for Pt–CoP/C-30% was estimated to be 0.584 V, which is more negative than for Pt/C-JM (0.629 V). This suggests that the intermediate species including CO are prone to removal in the case of Pt–CoP/C-30% composite during the alcohol oxidation process. The more positive onset potential of CO_ad stripping peaks can be explained as follows. CoP has been verified as an excellent catalyst for water oxidation with a relatively low overpotential.^36,37 This indicates that the decomposition of water will be easier on Pt–CoP/C-30% than on Pt/C-JM owing to the function of CoP, which indirectly suggests the oxygen-containing species are formed on the surface of Pt–CoP/C-30% at a more negative potential than on Pt/C-JM. Therefore, in accordance with the bifunctional mechanism model, CO-like species can be more easily oxidized by neighbouring oxygen-containing species on Pt–CoP/C-30%, thereby enhancing the anti-poisoning effect toward the CO-like intermediates.

We further carried out XPS characterization to seek plausible causes for the profound effect of CoP on the Pt-based catalysts. Fig. 5a shows the Pt 4f XPS spectra for Pt–CoP/C-30% and Pt/C-JM composites. The Pt/C-JM catalyst exhibits two peaks at 72.32 eV and 75.67 eV for Pt 4f_7/2 and Pt 4f_5/2, respectively. The binding energy of Pt 4f decreases by 0.37 eV in the Pt–CoP/C-30% catalyst compared with the Pt/C-JM catalyst. In addition, the binding energy of Pt 4f has no obvious shift (0.08 eV) for Pt–CoP/C-15%, which may be due to the low content of CoP. When the content of CoP in the hybrid materials was 45%, the binding energy of Pt 4f was negatively shifted about 0.23 eV. Hence, the shift of binding energy for Pt–CoP/C-X% (X = 15%, 30% and 45%) compared with Pt/C-JM showed a ‘volcanic type’ trend (Fig. S7†). Meanwhile, we found that the binding energy of Co 2p shifted to a higher value compared with the standard value (Fig. 5b). The shift of binding energy can be attributed to partial electron transfer from transition metal phosphate to the noble metal.^18,19,30 The altered binding energy would change the density of the valence state of Pt, which indeed would weaken the bond energy between intermediate species and noble metal, and thus accelerate the desorption of intermediate species from the surface of the noble metal.

Fig. 5 (a and b) XPS survey spectra of Pt 4f and Co 2p, respectively.

4. Conclusions

In summary, we have demonstrated CoP can be used as a promoter for Pt nanoparticles, providing an effective electrocatalyst for the MOR and EOR. The remarkable improvement of the catalytic activity and stability of the as-prepared Pt–CoP/C-30% catalyst can be attributed to the electronic effect and the bifunctional mechanism, which accelerate the desorption of intermediate species from the surface of Pt. Owing to these intriguing features, Pt–CoP/C-30% is a promising candidate as a new catalytic material for direct methanol fuel cells and direct ethanol fuel cells.

Acknowledgements

Financial support from the Director Fund of the WNLO, 973 Program of China (2014CB643506 and 2013CB922104) and the Fundamental Research Funds for the Central Universities (HUST: 2016YXMS031) is gratefully acknowledged. The authors thank the Analytical and Testing Centre of Huazhong University of Science & Technology and Characterization (CMFC) of WNLO for performing various characterizations.

Notes and references

1. C. Koenigsmann and S. S. Wong, Energy Environ Sci., 2011, 4, 1161–1176.
2. Z. H. Wang, W. F. Xie, F. F. Zhang, J. F. Xia, S. D. Gong and Y. Z. Xia, Electrochim. Acta, 2016, 192, 400–406.
3. L. F. Dong, R. R. S. Gari, Z. Li, M. M. Craig and S. F. Hou, Carbon, 2010, 48, 781–787.
4. H. Mao, T. Huang and A. Yu, Electrochim. Acta, 2015, 174, 1–7.
5. Q. Liu, Y. R. Xu, A. J. Wang and J. J. Feng, J. Power Sources, 2016, 302, 394–401.
6. H. J. You, F. L. Zhang, Z. Liu and J. X. Fang, ACS Catal., 2014, 4, 2829–2835.
7. L. Y. Zhang, W. L. Zhang, Z. L. Zhao, Z. Liu and Z. Q. Zhou, RSC Adv., 2016, 6, 50726–50731.
8. Z. X. Yan, G. Q. He, Z. F. Jiang, W. Wei, L. N. Gao and J. M Xie, J. Power Sources, 2015, 284, 497–503.
9. A. Arun, M. Gowdhamamoorthi, K. Ponmani, S. Kiruthikab and B. Muthukumaran, RSC Adv., 2015, 5, 49643–49650.
10. J. P. Paraknowitsch and A. Thomas, Energy Environ. Sci., 2013, 6, 2839–2855.
11. F. X. Wu, D. T. Zhang, M. H. Peng, Z. H. Yu, X. Y. Wang, G. S. Guo and Y. G. Sun, Angew. Chem., Int. Ed., 2016, 55, 4952–4956.
12. M. Wang, Y. W. He, R. X. Li, Z. Z. Ma, Z. H. Zhang and X. G. Wang, Electrochim. Acta, 2015, 178, 259–269.
13. V. Comignani, J. M. Sieben, M. E Brigante and M. M. E. Duarte, J. Power Sources, 2015, 45, 28–33.
14. S. M. Kim, L. Liu, S. H. Cho, Y. H. Jang and S. Park, J. Mater. Chem. A, 2013, 1, 15252–15257.
15. R. K. Selvan and Y. S. Lee, RSC Adv., 2016, 6, 62680–62694.
16. S. Su, C. Zhang, L. H. Yuwen, X. F. Liu, L. H. Wang, C. H. Fan and L. H. Wang, Nanoscale, 2016, 8, 602–608.
17. P. Gnanaprakasam, S. E. Jeena and T. Selvaraju, J. Mater. Chem. A, 2015, 3, 18010–18018.
18. L. G. Feng, K. Li, J. F. Chang, C. P. Liu, W. Xing and X. L. Hu, Energy Environ. Sci., 2014, 7, 1628–1632.
19. L. G. Feng, K. Li, J. F. Chang, C. P. Liu and W. Xing, ChemSusChem, 2015, 8, 3340–3347.
20. A. B. A. A. Nassr, I. Sinev, M. M. Pohl, W. Grünert and M. Bron, ACS Catal., 2014, 4, 2449–2462.
21. X. Liu, G. Xu, Y. Chen, T. Lu, Y. Tang and W. Xing, Sci. Rep., 2015, 5, 7619.
22. Y. Ma, R. Wang, H. Wang, V. Linkov and S. Ji, Phys. Chem. Chem. Phys., 2014, 16, 3593–3602.
23. J. F. Zhang, K. D. Li and B. Zhang, Chem. Commun., 2015, 51, 12012–12015.
24. J. F. Chang, L. G. Feng, C. P. Liu and W. Xing, ChemSusChem, 2015, 8, 3340–3347.
25. M. Watanabe and S. Motoo, J. Electroanal. Chem. Interfacial Electrochem., 1975, 60, 267–273.
26. A. B. Anderson, E. Grantscharova and S. Seong, J. Electrochem. Soc., 1996, 143, 2075–2082.
27. L. B. Ma, X. P. Shen, H. Zhou, G. X. Zhu, Z. Y. Ji and K. M. Chen, J. Mater. Chem. A, 2015, 3, 5337–5343.
28. J. L. Zhu, G. Q. He and P. K. Shen, J. Power Sources, 2015, 275, 279–283.
29. Y. C. Xing, J. Phys. Chem. B, 2004, 108, 19255–19259.
30. E. J. Popczun, C. G. Read, C. W. Roske, N. S. Lewis and R. E. Schaak, Angew. Chem., 2014, 126, 5531–5534.
31. J. F. Chang, L. G. Feng, C. P. Liu, W. Xing and X. L. Hu, Angew. Chem., Int. Ed., 2014, 53, 122–126.
32. L. G. Feng, K. Li, J. F. Chang, C. P. Liu and W. Xing, Nano Energy, 2015, 15, 462–469.
33. Z. L. Liu, X. Y. Ling, X. D. Su, J. Y. Lee and L. M. Gan, J. Power Sources, 2005, 149, 1–7.
34. R. Mancharan and J. B. Goodenough, J. Mater. Chem., 1992, 2, 875–887.
35. Y. Ye, H. S. Kim, P. K. Babu, P. Waszczuk, A. Wieckowski and E. Oldfield, J. Am. Chem. Soc., 2002, 124, 468–473.
36. C. C. Hou, S. Cao, W. F. Fu and Y. Chen, ACS Appl. Mater. Interfaces, 2015, 7, 28412–28419.
37. Y. Bai, H. Zhang, Y. Y. Feng, L. Fang and Y. Wang, J. Mater. Chem. A, 2016, 4, 9072–9079.

---

Footnote

† Electronic supplementary information (ESI) available: Details of characterization and results of electrochemical measurements. See DOI: 10.1039/c6ra21938a

---