Nano-Fe₃O₄ grown on porous carbon and its effect on the oxygen reduction reaction for DMFCs with a polymer fiber membrane

Abstract

Fe₃O₄ nanoparticles grown on porous carbon are prepared as an oxygen reduction reaction (ORR) catalyst in direct methanol fuel cells (DMFCs) in which a polymer fiber membrane (PFM) is used to replace Nafion membrane. The catalyst has shown good ORR activity and high stability in alkaline methanol solution. A flexible hydrophobic cathode structure is introduced in the full cell test, the peak power density reaching 71.24 mW cm⁻² at 25 °C and 166.87 mW cm⁻² at 65 °C, respectively. This improvement is due to the combination effect of the advantages of the PFM structure and the good ORR activity and stability of the catalyst.

---

Introduction

As an alternative hydrogen power source, a direct methanol fuel cell (DMFC) offers important advantages in hydrogen reforming, refueling, and simplified system design.¹ It aims to operate at ambient temperature and pressure, which makes itself a very convenient and easy-to-use device. Cost, durability and performance, however, are three main barriers in the development of DMFC.² Hence, commercial applications of DMFC still have a long way to go.

One factor that limits the extensive utilization of DMFC is the use of acidic electrolyte membrane such as Nafion membrane, since most catalysts frequently suffer from dissolution and agglomeration during operation in an acidic circumstance.³ Alkaline polymer electrolyte membrane (APEM) also faces the problem of insufficient stability at high pH (pH > 14) and elevated temperature (>60 °C),⁴ and its performance is limited by the poor ion conductivity at an ambient temperature since the diffusion coefficient of OH⁻ ions is four times less than that of H⁺ ions at the same temperature.^5,6

Recently, a liquid permeable polymer fiber membrane (PFM) was introduced by X. D. Yang in direct borohydride fuel cell which greatly enhanced power output due to its higher ions conductivity and lower membrane resistance, and reduced the membrane cost at the same time.⁷ At present, the expensive Pt or Pt-based catalysts are still being used heavily at the electrodes for MOR (methanol oxidation reaction) and ORR respectively because of the sluggish kinetics at both electrodes.⁸ However, in view of the permeability of the PFM structure, Pt or Pt-based catalysts no longer satisfy the demand due to its non-selectivity to MOR and ORR,⁹ since the permeation of methanol severely impedes the power output of DMFC. Therefore, developing a non-noble and methanol-tolerant catalyst for PFM structured cells will provide much significant economic advantages.

To replace this costly noble Pt and design a catalyst that is suitable for PFM structure, among the various designs where non-precious catalysts were used, several rational ones have attracted our attention. For example, in recent years significant progress has been made to Fe–N–C type complexes in improving the ORR activity to the level that is comparable to Pt and Pt-based materials.^10–13 Spinel compounds like Co₃O₄,¹⁴ Mn₃O₄,¹⁵ CoMn₂O₄ (ref. 16) and NiCo₂O₄ (ref. 17) and perovskite metal oxide such as LaMO₃ (M = Fe, Co, Mn, Ni, Cr)¹⁸ exhibit the promising ORR activity due to their facilitated kinetics for ORR in alkaline solution.¹⁹ They are also highly desirable in DMFC for their low price, high stability and high tolerance to methanol poison.²⁰ Nanosized Fe₃O₄ is known as an important member of spinel type oxide that has been widely used in photothermic therapy,²¹ recording materials,²² mineral separation,²³ heat transfer applications²⁴ and so on. It is also an important property of Fe₃O₄ to act as a catalyst in ORR activity. To our knowledge, the traditional synthesis of Fe₃O₄ nanoparticles was mainly through the precipitation of nanocrystals from basic aqueous solution. The high surface energy and magnetization make Fe₃O₄ nanoparticles prone to aggregate. As for the ORR catalysts, particles agglomeration and poor electronic conductivity must be taken into account, since they greatly impede ORR activity. The early method to deal with the problem of low conductivity of oxide catalysts in fuel cell was simply to mix them with some electronic conductive materials such as carbon nanotube and acetylene black,^25,26 while nowadays, it has been an effective approach to grow or support the oxide nanoparticles on high-surface-area carbon supporters, so as to effectively overcome electronic conductivity limitations^27,28 and improve the ORR activities.

Therefore, developing a relatively monodispersed Fe₃O₄ nanoparticles on a high-surface-area carbon supporter for enhancing the stability of the catalyst and the activity of ORR which can be used for DMFC with PFM structure is the main purpose of this study.

In this work, a Fe₃O₄ nanoparticle grown on carbon matrix (Fe₃O₄-CB) was prepared. The Fe₃O₄-CB exhibited high ORR activity and high stability in alkaline methanol solution. The fuel cell with PFM structure and Fe₃O₄-CB as catalyst delivered a promising performance compared with data in other literature.²⁹ The introduction of the non-noble ORR catalyst and low-cost membrane structure can effectively maintain the cost of DMFC at an affordable level without reducing its power performance, which seems to have provided a substantial opportunity to put DMFC into practical use.

Results and discussion

Catalytic character

To understand the basic properties of this cathode catalyst, we carried out some physical characterizations. The precursor of Fe₃O₄-CB nanoparticles was prepared through hydrothermal process that confined FeCl₃ by N,N′-methylenebisacrylamide on nanoporous carbon particles. The mixed product was collected by vacuum rotary evaporation at 110 °C. XRD confirmed that a tetragonal FeOOH (ICSD card no. 031136) was formed on the carbon matrix in this process (Fig. 1a). Then, the precursor was heated to 550 °C. When having been sintered at 550 °C for 1 h in argon gas protection, the FeOOH fully transformed into a cubic Fe₃O₄ (ICSD card no. 029129). In Fig. 1b, X-ray photoelectron spectroscopy (XPS) of Fe₃O₄-CB is compared with that of carbon matrix, both showing the peaks of C_1s and O_1s. High-resolution spectra of Fe_2p for catalyst in the inset of the Fig. 1b clearly exhibits Fe_2p signals corresponding to the peaks of Fe_2p1/2 and Fe_2p3/2 for element Fe at 725 eV and 711 eV, respectively. This demonstrates the existence of both Fe(II) and Fe(III) in Fe₃O₄-CB.

Fig. 1 (a) XRD spectrum of precursors at different heating temperatures, (b) XPS spectra of Fe₃O₄-CB and carbon matrix.

The porous nature of Fe₃O₄-CB was accessed with the N₂ adsorption–desorption isotherms (Fig. 2), the loop in which indicated the mesoporous nature of this catalyst. The BET surface area of the sample was 1191 m² g⁻¹. Using BJH model, the pore size distribution was analyzed, clearly showing the existence of a large number of mesoporous at around 10 nm.

Fig. 2 The N₂ adsorption–desorption isotherm of Fe₃O₄-CB, black symbols: adsorption; red symbols: desorption. The inset graph shows the BJH pore distribution.

Transmission electron microscopic (TEM) images showed porous texture of the carbon matrix (Fig. 3a). The carbon matrix particles were in sizes ranging from 30 nm to 100 nm. Among them, many Fe₃O₄ crystals dispersed over the carbon particles, which can be seen in bright field image (Fig. 3d–f) and dark field image (Fig. 3c), and the Fe₃O₄ particle sizes was about 3–5 nm. From Fig. 3f, we could clearly observe the lattice fringe image of Fe₃O₄, and the interplanar spacing values obtained were 0.208 nm and 0.161 nm, which was in agreement with the values of [400] and [511] planes in XRD database.

Fig. 3 (a, b, d, e and f) TEM morphology of Fe₃O₄-CB. (c) Dark field image of Fe₃O₄-CB.

Considering these characterizations above, we suggest that the formation of crystal Fe₃O₄ that integrated into large surface-area and highly mesoporous carbon nano-particles may mainly be responsible for ORR activity of this material.

Catalytic activity of Fe₃O₄-CB

The electro catalytic properties of Fe₃O₄-CB for ORR were studied by cyclic voltammetry (CV) and Normal Pulse Voltammetry (NPV) with RRDE. We compared the performance of Fe₃O₄-CB and commercial Pt/C (40 wt%; JM) in O₂ and N₂ saturated solutions (0.1 M KOH), respectively. Catalyst loading was 0.85 mg cm⁻² both for Fe₃O₄-CB and Pt/C. Fe₃O₄-CB showed an oxygen mass transfer current plate at an electronic potential below −0.1 V (vs. Hg/HgO) (Fig. 4b). A reduction potential similar to that of the commercial Pt/C (40 wt%) catalyst was observed at the electrode, the half-wave potential (E_1/2) of Fe₃O₄-CB was 0.05 V (vs. Hg/HgO), which was only 60 mV lower than Pt/C (Fig. 4a). The result suggests a high ORR catalytic activity of the Fe₃O₄-CB in alkaline solution.

Fig. 4 (a) Polarization curves of Fe₃O₄-CB (magenta line), carbon matrix (black line) and Pt/C (40 wt%) (blue line) in O₂-saturated 0.1 M KOH at a sweep rate of 10 mV s⁻¹ and rotation rate of 900 rpm. The catalyst loading is 0.85 mg cm⁻². (b) CV voltammogram of Fe₃O₄-CB in O₂-saturated (magenta line) and N₂-saturated (blue line) 0.1 M KOH at a sweep rate of 10 mV s⁻¹ and rotating disk rate of 900 rpm. The catalyst loading is 0.85 mg cm⁻².

For further insight into the ORR kinetics, the electron transfer number was analyzed on the basis of the RDE measurements at various rotating speeds and the Koutecky–Levich³⁰ equation. The linearity of the Koutecky–Levich plots and near parallelism of the fitting lines suggests first-order reaction kinetics toward the concentration of dissolved oxygen and similar electron transfer numbers for ORR at different potentials.³¹ Fig. 5 a displays its steady-state diffusion plateau currents curves at a rotating speed from 300 to 1800 rpm, and the Fe₃O₄-CB samples exhibited electron transfer numbers of n = 3.93 and n = 3.84 for ORR at −0.2 V and −0.1 V (vs. Hg/HgO) respectively, calculated from the slopes of curves by theory (Fig. 5b). Data from electro chemical technology substantiate the high performance of Fe₃O₄-CB as an ORR catalyst in alkaline medium, and indicate that the Fe₃O₄-CB is one of the alternatives to Pt as a cathode catalyst in alkaline DMFC.

Fig. 5 (a) Polarization curves of Fe₃O₄-CB in O₂-saturated 0.1 M KOH at different speed of rotating disk, from 300 to 1800 rpm. The catalyst loading is 0.85 mg cm⁻². (b) Corresponding Koutecky–Levich plots of Fe₃O₄-CB at different potentials.

Performance in full cell tests

We made full cells with Nafion-211 membrane (NRE-211, PEM-structure) and the polymer fibre membrane (PFM-structure) to investigate the Fe₃O₄-CB as cathode catalyst at different membrane-structures.

For comparison, the power densities of Pt/C and Fe₃O₄-CB cathode catalysts with Nafion-211 membrane are showed in Fig. 6b. The loading of cathode catalysts was both 8 mg. The Pt/C catalyst showed better performance at both 25 °C and 65 °C, the maximum power density (P_max) being 15 mW cm⁻² and 56 mW cm⁻², respectively, which was slightly higher than the P_max of Fe₃O₄-CB (12.5 mW cm⁻² at 25 °C and 40 mW cm⁻² at 65 °C). The result is in agreement with RRDE tests (Fig. 4a).

Fig. 6 (a) Polarization and power density plots of 8 mg and 40 mg Fe₃O₄-CB cathode loading with PFM structure at different temperature. Anode loading was 5 mg Pt/Ru. (b) Polarization and power density plots of 8 mg Pt/C and 8 mg Fe₃O₄ cathode loading with Nafion-211 membrane at different temperature. Anode loading was 5 mg Pt/Ru.

However, in PFM-structured full cell, and under the same condition of loading of Fe₃O₄-CB (which is 8 mg), the P_max could reach 100 mW cm⁻² at 65 °C (38.6 mW cm⁻² at 25 °C) (Fig. 6a) which was much higher than Pt/C catalyst in PEM structure. We increased the cathode catalysts loading up to 40 mg. From Fig. 6a, the slope of the polarization curve rises with the catalysts loading, and the P_max reaches 71.24 mW cm⁻² at 25 °C and 166.87 mW cm⁻² at 65 °C.

This improvement could be firstly correlated to catalytic activity and methanol-tolerance of Fe₃O₄-CB cathode catalyst. At the same condition among the two catalysts, Pt/C and Fe₃O₄-CB. Pt/C has catalytic nature for both O₂ and methanol. The crossover of fuel through the membrane could severely affect the power output of the cells when using Pt/C as cathode catalysts. So it is not suitable as a cathode catalyst for DMFC with PFM. As for Fe₃O₄-CB, it does not react with methanol, so the crossover does not affect the ORR in cathodic performances. To test the methanol-tolerance of the catalyst, an RRDE test was carried out in an O₂ saturated solution with 5 M methanol and 0.1 M KOH. After 1000 cycles, apparent changes could hardly be observed (Fig. 7d), which fully revealed the high stability and methanol tolerance of the Fe₃O₄-CB catalyst in the alkaline situation during ORR process.

Fig. 7 (a) Polarization plots of PFM-DMFC with different concentration of KOH, from 1 M to 5 M. (b) 100 mA cm⁻² constant current discharge of DMFC with PFM structure. (c) Electrochemical impedance spectrum of the cells with different structure, PFM (yellow dot) and PEM (blue dot). (d) CV voltammogram of Fe₃O₄-CB in O₂-saturated 2 M MeOH 0.1 M KOH solution, the scanning rate is 10 mV s⁻¹.

In the fuel cell durability test, 4 M KOH and 5 M methanol solution were used in PFM-structure. The potential showed a small decrease of about 8% after the fuel cell had been running for 100 h at 25 °C at a constant current discharge of 100 mA cm⁻² (Fig. 7b). This result also firmly exhibited the high stability of both catalysts and PFM membrane structure.

Secondly, different membrane structure could be attributed as another important factor. In PFM structure, the power density of full cell greatly depends on the electrolyte or fuel concentration. Due to the permeability of PFM, solution was allowed to penetrate, and the solution in membrane therefore contained many OH⁻ ions. During this process, the PFM layer acted as an anion transporter. On one hand, the more OH⁻ the solution contained, the more capable the membrane can be to transport OH⁻. Fig. 7a shows the changes of the polarization curves of the PFM structure with 1 M methanol and 1–5 M KOH at ambient temperature. It can be observed that the output current significantly increases with the rise of concentration of KOH from 1 M to 5 M, with the maximum current increasing from 75 mA to 200 mA correspondingly. The rate of the increase accelerates first and then slows down, and finally becomes unnoticeable until the concentration reaches 5 M. The highest peak power density of the DMFC was achieved in 5 M methanol and 4 M KOH solution after several tests. In the test of electrochemical impedance spectroscopy, PFM structure demonstrated a smaller Ohm resistance than PEM structure, which was represented by the intercept on the real-axis (Fig. 7c). The value was 0.32 Ω cm⁻² and 0.79 Ω cm⁻² for PFM and PEM, respectively. This property of PFM enables it to provide higher current output, and in turn higher power output.

PEM and PFM structure

To clearly explain the other advantages of PFM structure over the PEM structure, a schematic diagram was prepared in Fig. 8. In general, in PEM structure the cathode and anode need to be squeezed tightly together because PEM is a solid polymer electrolyte membrane, and the contact resistance will lead to a voltage drop across the interface.³³ The drawback was directly reflected by the higher membrane resistance of Nafion-211 membrane in the impedance test and lower current output of the PEM structure in full cell test compared with PFM structure. When catalyst is in relatively small amount, and in one layer as shown in Fig. 8a(1), the polymer interface can wrap the catalyst and make a good contact. If the amount of catalyst increases, forming several layers as shown in Fig. 8a(2), it is difficult for polymer electrolyte to stretch into the micro tunnels or holes of second or third layer of catalyst. Thus, the catalysts on second or third layer lose contact with the PEM and the ion transport channel is cut off, and the diffusion and transportation of ions from the anode to the cathode is limited by the catalysts distributed in PEM. In this way, a large contact resistance develops.³⁴

Fig. 8 (a) Schematic diagram of PEM structure, (1) small cathode mass loading, (2) large cathode mass loading (b) schematic diagram of PFM structure, (1) small cathode mass loading, (2) large cathode mass loading.

The PFM structure systematically avoids the limitation of solid electrolyte mass transfer by using liquid-permeable membrane and liquid electrolyte as ion conductor. Unlike Nafion-211 and other solid anion-exchange membrane, polymer fiber membrane is a permeable membrane, which means the electrolyte of membrane depends on the solution used. Hot pressing is not necessary in the preparation of three-in-one membrane electrode, because the liquid solution in the permeable membrane can provide an effective ion conductive method for the ions exchange. The most important problem in this system is to avoid the cathode from solution drowning, so a hydrophobic cathode layer was introduced in this structure. The hydrophobic layer which is up to 300 μm thick (Fig. 8b black area) possesses two important functions: one is to form a large mass loading matrix for cathode catalysts; the other is to block the solution from flooding cathode catalysts and provide adequate room for the oxygen to diffuse in.³² From Fig. 8b, ions in the solution could freely transfer from anode to cathode without being blocked by the cathode hydrophobic layer. This kind of mechanism effectively facilitated the accessibility of reactant and kinetics for ORR in alkaline circumstance, and consequently enhanced the overall power density.

Experimental

Catalysts and materials

Nafion-211 membrane was purchased from Shanghai Hephas Energy Corporation. Nippon Kodoshi Corporation (Japan) supplied polymer fiber membrane. PtRu/C and Pt/C (HISPEC 12000, atomic ratio 1 : 1) were purchased from Johnson Matthey (UK). Multiwall carbon nanotubes were obtained from Chengdu Organic Chemicals Co. Ltd. EC-600J Ketjen-black carbon matrix was purchased from Lion Company (Japan). FeCl₃·6H₂O and N,N′-methylenebisacrylamide were purchased from Sino-pharm Group Co. Ltd.

Synthesis of the nano-Fe₃O₄ nanoparticles grown on porous carbon

A solution of 100 mL water, 1 g Ketjen-600 carbon black, 1 g N,N′-methylenebisacrylamide and 6 g FeCl₃, were mixed in a flask and heated to 80 °C under magnetic stirring and kept at this temperature for 3 h. Then, the solvent was removed through rotary evaporator at 110 °C and further dried in oven for 12 h at 80 °C. The resulting powder was subjected to the heat treatment in a quartz tube at 550 °C for 1 h under argon atmosphere protection. The pyrolyzed sample was then put in a deionized water solution followed by centrifugation and washing for at least 4 times, then dried in oven for 12 h at 80 °C for further use.

Electrochemical catalyst measurements

Electrochemical characterization was conducted in 0.1 M KOH. Pt wire and Hg/HgO were used as counter and reference electrodes, respectively. Cyclic voltammograms (CVs) were obtained by scanning between −0.6 V and 0.2 V vs. Hg/HgO at a scan rate of 10 mV s⁻¹ in O₂ and N₂. ORR polarization curves were obtained by Normal Pulse Voltammetry (NPV) scanning from −0.4 V to 0.1 V vs. Hg/HgO at a scan rate of 20 mV s⁻¹ in O₂ with RDE at different rpms (300–1800 rpm). The electron transfer number (n) was calculated from the slopes of the Koutecky–Levich³⁰ plots according to:

where I, I_K and I_L are measured current, kinetic current and diffusion-limiting current respectively, ω is the angular velocity, F is the Faraday constant (96 485 C mol⁻¹), A is the electrode surface area, C_O2 is the concentration of dissolved O₂ (1.2 × 10⁻⁶ mol cm⁻³), D_O2 is the diffusion coefficient of O₂ (1.9 × 10⁻⁵ cm² s⁻¹) and υ is kinetic viscosity of the electrolyte (0.01 cm² s⁻¹)³. The constant 0.62 is applied when the rotation speed is expressed in radius per second (rad s⁻¹)². The electrochemical impedance spectroscopy of the cell was measured at open circuit voltage with an oscillating amplitude of 5 mV over a frequency range of 1 Hz to 10⁵ Hz.

Preparation of membrane electrode structure

Both anode and cathode catalysts were loaded on nickel foam. A gas diffusion layer was employed in the cathode. To prepare gas diffusion layer, 50 wt% acetylene black and 50 wt% polytetrafluoroethylene (PTFE) (30 wt% PTFE solution) were mixed into slurry. Then press the slurry into a 0.3–0.5 mm membrane and heat at 340 °C for 1 hour. The cathode electrode was fabricated by pressing the gas diffusion layer and the nickel foam under 2 MPa pressure. The catalyst ink was prepared by mixing 80 μL 5 wt% Nafion solution, 450 μL deionized water and 470 μL ethanol with 20 mg catalysts. The loading of anodic catalyst (PtRu/C) was 5 mg cm⁻². The loading of cathode catalyst was ranging from 8–40 mg cm⁻².

Conclusions

Nano-Fe₃O₄ grown on porous carbon for ORR of DMFCs with PFM structure has been prepared. Its high ORR activity and methanol-tolerance enable it to be effectively applied in such DMFC. The highly dispersed Fe₃O₄ nanoparticles on carbon matrix have shown good ORR activity that is comparable to commercial Pt/C. A permeable polymer fiber membrane with a hydrophobic cathode structure has greatly enhanced fuel cell performance. The peak power densities (71.24 mW cm⁻² at 25 °C and 166.87 mW cm⁻² at 65 °C) were achieved, and the durability test of fuel cell with PFM structure also demonstrated high stability of the catalyst in 4 M KOH and 5 M methanol solution. Due to the low cost and abundance of carbon matrix, iron and fiber materials, the DMFC with Fe₃O₄-CB catalysts and PFM possesses a high price/performance ratio and can bring the total cost down to an acceptable level, which will definitely be a promising step towards the commercial application of DMFCs.

Acknowledgements

This work was supported by the China Postdoctoral Science Foundation (no. 2012M521760) and the Fundamental Research Funds for the Central Universities (no. xjj2014052).

Notes and references

1. V. Neburchilov, J. Martin, H. Wang and J. Zhang, J. Power Sources, 2007, 169, 221–238.
2. M. Kathi Epping, P. K. John and W. M. Kevin, in Fuel Cell Chemistry and Operation, American Chemical Society, 2010, vol. 1040, ch. 1, pp. 1–13.
3. B. Wu, G. E. Gray and T. F. Fuller, Electrochem. Solid-State Lett., 2007, 10, B101–B104.
4. W. Yan-Jie, Q. Jinli, B. Ryan and Z. Jiujun, Chem. Soc. Rev., 2013, 42, 5768–5787.
5. A. V. Anantaraman and C. L. Gardner, J. Electroanal. Chem., 1996, 414, 115–120.
6. Y. Sone, P. Ekdunge and D. Simonsson, J. Electrochem. Soc., 1996, 143, 1254–1259.
7. X. Yang, Y. Liu, S. Li, X. Wei, L. Wang and Y. Chen, Sci. Rep., 2012, 2, 753–756.
8. X. Yang, Y. Liu, F. Yuan, W. Li, S. Li and X. Wei, J. Power Sources, 2013, 234, 272–276.
9. M. B. Vukmirovic, J. Zhang, K. Sasaki, A. U. Nilekar, F. Uribe, M. Mavrikakis and R. R. Adzic, Electrochim. Acta, 2007, 52, 2257–2263.
10. F. Vigier, C. Coutanceau, F. Hahn, E. M. Belgsir and C. Lamy, J. Electroanal. Chem., 2004, 563, 81–89.
11. E. Proietti, F. Jaouen, M. Lefèvre, N. Larouche, J. Tian, J. Herranz and J. P. Dodelet, Nat. Commun., 2011, 2, 73–86.
12. G. Wu and P. Zelenay, Science, 2011, 332, 443–447.
13. Z. S. Wu, Y. Sun, Y. Z. Tan, S. Yang, X. Feng and K. Mullen, J. Am. Chem. Soc., 2012, 134, 19532–19535.
14. F. Jaouen, J. Herranz, M. Lefèvre, J. P. Dodelet, U. I. Kramm, I. Herrmann, P. Bogdanoff, J. Maruyama, T. Nagaoka and A. Garsuch, ACS Appl. Mater. Interfaces, 2009, 1, 1623–1639.
15. 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.
16. J. Feng, Y. Liang, H. Wang, Y. Li, B. Zhang, J. Zhou, J. Wang, T. Regier and H. Dai, Nano Res., 2012, 5, 718–725.
17. M. Prabu, P. Ramakrishnan, H. Nara, T. Momma, T. Osaka and S. Shanmugam, ACS Appl. Mater. Interfaces, 2014, 6, 16545–16555.
18. U. L. Dong, B. J. Kim and Z. Chen, J. Mater. Chem. A, 2013, 1, 4754–4762.
19. M. A. Rahman, X. Wang and C. Wen, J. Electrochem. Soc., 2013, 160, A1759–A1771.
20. G. Nam, J. Park, S. T. Kim, D.-b. Shin, N. Park, Y. Kim, J.-S. Lee and J. Cho, Nano Lett., 2014, 14, 1870–1876.
21. J. K. Dombrovskis, C. Prestel and A. E. C. Palmqvist, APL Mater., 2014, 2, 2097.
22. T. Qiwei, H. Junqing, Z. Yihan, Z. Rujia, C. Zhigang, Y. Shiping, L. Runwei, S. Qianqian, H. Yu and L. Xiaogang, J. Am. Chem. Soc., 2013, 135, 8571–8577.
23. M. Langlet, M. Labeau, B. Bochu and J. C. Joubert, IEEE Trans. Magn., 1986, 22, 151–156.
24. E. Tahmasebi and Y. Yamini, Microchim. Acta, 2014, 181, 543–551.
25. L. S. Sundar, M. K. Singh and A. C. M. Sousa, Int. Commun. Heat Mass Transfer, 2013, 44, 7–14.
26. Y. Fang, X. Yang, L. Wang and Y. Liu, J. Power Sources, 2014, 267, 33–38.
27. L. Wang, Y. Liu, X. Yang, Y. Fang, Y. Chen and B. Wang, J. Mater. Chem. A, 2013, 1, 1834–1839.
28. D. Sebastián, I. Suelves, R. Moliner and M. J. Lázaro, Carbon, 2010, 48, 4421–4431.
29. G. Dong, K. Zhao and L. Zhang, Chem. Commun., 2012, 48, 6178–6180.
30. A. J. Bard and L. R. Faulkner, Electrochemical methods: fundamentals and applications, Wiley, New York, 1980.
31. K. J. J. Mayrhofer, D. Strmcnik, B. B. Blizanac, V. Stamenkovic, M. Arenz and N. M. Markovic, Electrochim. Acta, 2008, 53, 3181–3188.
32. G. Nam, J. Park, T. K. Sun, D. B. Shin, N. Park, Y. Kim, J. S. Lee and J. Cho, Nano Lett., 2014, 14, 1870–1876.
33. V. Mishra, F. Yang and R. Pitchumani, J. Fuel Cell Sci. Technol., 2004, 1, 2–9.
34. H. A. Gasteiger, J. E. Panels and S. G. Yan, J. Power Sources, 2004, 127, 162–171.

---