Bottom-up design of a stable CO-tolerant platinum electrocatalyst with enhanced fuel cell performance in direct methanol fuel cells

Abstract

Sluggish methanol oxidation reaction (MOR) and CO poisoning of platinum electrocatalysts are critical problems in direct methanol fuel cells (DMFCs). Here, we design a stable CO tolerant platinum electrocatalyst via a bottom-up method, in which the platinum nanoparticles are deposited on carbon black after coating with polybenzimidazole (PBI) and poly(vinyl pyrrolidone) (PVP). By comparison with the PVP post-coated electrocatalyst (CB/PBI/Pt/PVP), the PVP pre-coated electrocatalyst (CB/PBI/PVP/Pt) exhibits comparable durability and CO tolerance due to the similar amount of PVP in the electrocatalyst, suggesting the PVP pre-coating method shows negligible effect on CO tolerance and durability, while the Pt utilization efficiency, methanol oxidation activity and power density of CB/PBI/PVP/Pt are 1.6 times higher than those of CB/PBI/Pt/PVP. Thus, the PVP pre-coated electrocatalyst has better activity due to the non-coated Pt nanoparticles. Meanwhile, CB/PBI/PVP/Pt exhibits highly stable CO tolerance during the durability test, while the CO tolerance of the commercial CB/PtRu seriously deteriorates during the durability test due to the dissolution of Ru nanoparticles. To the best of our knowledge, the maximum power density of CB/PBI/PVP/Pt (104 mW cm⁻²) is one of the highest values in recent publications.

---

Introduction

Due to the energy crisis, sustainable energy sources attract much attention due to their high energy conversion efficiency and environmental friendliness. Direct methanol fuel cells (DMFCs) receive considerable attention since methanol has a higher energy density and easier transportation as well as storage compared to hydrogen.^1,2 However, there are still two main problems that should be addressed before widespread commercialization of DMFCs, namely the sluggish methanol oxidation reaction (MOR) and CO poisoning of platinum (Pt) nanoparticles caused by the incomplete oxidation of the methanol. The generated CO species cover the Pt nanoparticles to degrade the fuel cell’s performance as well as its lifetime.³

Aside from the sluggish MOR activity, alloying Pt with ruthenium (Ru) was found to be an efficient way to eliminate the CO poisoning problem in DMFCs,^4,5 while, as well known, Ru is dissolvable in the acidic medium and accordingly the CO anti-poisoning of PtRu electrocatalysts is unable to be maintained during the long-term operation of the DMFCs due to the highly acidic environment.⁶ Changing Ru to other transition metals, such as Sn,^7,8 Co,⁹ Au,¹⁰ Mo,¹¹ Fe¹², etc. has been systematically studied, while the alloyed electrocatalysts still suffered from low durability and sluggish MOR activity since transition metals exhibit low MOR activity.¹³ Considering the sluggish MOR and CO tolerance simultaneously, design and fabrication of a CO tolerant Pt electrocatalyst is seriously considered because Pt exhibits the highest MOR activity among all the metals. Doping carbon materials with phosphorus¹⁴ or boron¹⁵ was carried out and it was found that the phosphorus/boron weakened the binding energy between the Pt nanoparticles and CO species. Also, we have reported a promising method to eliminate the CO poisoning problem, in which the Pt nanoparticles were coated with poly(vinylphosphonic acid) (PVPA)^16–18 or poly(vinylpyrrolidone) (PVP).¹⁹ Due to the presence of the water soluble polymer in the electrocatalyst, the formation of Pt(CO)_ads species was accelerated resulting in high CO tolerance; meanwhile, the sluggish MOR is unable to be solved in this method because the polymer partially covers the Pt active sites and the degradation in MOR activity is unavoidable.

In this work, we try to synthesize a new electrocatalyst via a bottom-up design (Fig. 1) in order to maintain high CO tolerance of the electrocatalyst and simultaneously address the sluggish MOR by polymer coating. The carbon black (CB) is firstly coated with polybenzimidazole (PBI) and PVP before the Pt deposition. The PVP layer can be stabilized by the PBI layer due to hydrogen bonding and the Pt nanoparticles will be anchored by PVP via Pt–N bonding.²⁰ Due to the presence of the PVP layer, the CO tolerance would be enhanced because of the acceleration of water adsorption, which is important for enhancement in CO tolerance of the Pt electrocatalyst.^21,22 Meanwhile, the Pt nanoparticles are more active compared to the PVP-coated Pt electrocatalyst. The comparison between the PVP pre-coated and post-coated electrocatalysts is systematically studied.

Fig. 1 Schematic illumination of post- (upper) and pre-coating (lower) of PVP polymer on the synthesized electrocatalyst.

Experimental

Materials

H₂SO₄, methanol, isopropanol, N,N-dimethylformamide (DMF), hydrogen hexachloroplatinatehexahydrate (H₂PtCl₆·6H₂O), ethylene glycol (EG) and poly(vinylpyrrolidone) (PVP, K30) were purchased from Sinopharm Chemical Reagent Co., Ltd. Commercial CB/Pt (40 wt%) and CB/PtRu (Pt: 40 wt%; Ru: 20 wt%) were purchased from Alfa Aesar. Nafion 117 membrane and Nafion solution (5 wt%) were provided by Sigma-Aldrich. The synthesis of PBI was as reported by Xiao et al.²³ Aqueous solutions were prepared using Milli-Q water and all chemicals were used as received without further purification.

Synthesis of CB/PBI/Pt/PVP

The CB/PBI/Pt was synthesized according to our previous reports.^16,17,24 Briefly, 20 mg of CB was coated with PBI (10 mg) using sonication for 1 h in DMF. The composite was then collected by filtration, and then dried under vacuum at 80 °C. 20 mg of CB/PBI was dispersed in EG aq. (v/v = 3 : 2). The Pt loading was carried out by reduction of H₂PtCl₆ (28 mg) at 140 °C for 6 h under N₂ atmosphere. The obtained product was filtered, washed, and then dried overnight under vacuum at 80 °C to completely remove the solvent. 20 mg of CB/PBI/Pt was dispersed in 20 mL of water by sonication for 10 min, to which 20 mg of the PVP was added, then ultrasonicated for 1 h followed by filtration using 0.2 μm PTFE filter paper to collect the product, which was washed several times with Milli-Q water to remove free PVP, then dried overnight at 80 °C under vacuum to remove any residual solvent.

Synthesis of CB/PBI/PVP/Pt

20 mg of CB was coated with PBI (10 mg) using sonication for 1 h in DMF. The composite was then collected by filtration, and then dried under vacuum at 80 °C. 20 mg of CB/PBI was dispersed in water and then 20 mg of PVP was added. After sonication for 30 min, the resultant solution was filtered, washed and dried. The procedure of Pt deposition was the same as described above.

Gas diffusion electrode (GDE) preparation

GDE was prepared as follows. Two different electrocatalysts were used to prepare GDEs for the anode sides. The electrocatalyst was dispersed in 20 mL of isopropanol aqueous solution (v/v = 4 : 1) by sonication for 30 min, and then filtered using the gas diffusion layer (GDL). The Pt loading amount on the GDL was controlled at 2 mg cm⁻². The obtained GDE was dried overnight under vacuum at 60 °C to remove isopropanol. The cathode GDE was prepared using commercial CB/Pt with 2 mg_Pt cm⁻².

Membrane electrode assembly (MEA) preparation

MEA was prepared by hot pressing the prepared GDEs (anode: 2 mg_Pt cm⁻², CB/PtRu or the synthesized electrocatalyst; cathode: 2 mg_Pt cm⁻², commercial CB/Pt) and Nafion 117 membrane. The geometrical area of the MEA was 5 cm².

Material characterization

X-ray photoelectron spectroscopy (XPS) was carried out using a ThermoScientific K-Alpha instrument. The pressure in the XPS analysis chamber was kept at 10⁻⁹ Pa or lower during the measurements. During the calibration process, the binding energy (BE) of the C1s peak was fixed at 284.5 eV as standard. Elemental analysis was performed based on the peak area of N1s (395–407 eV), Ru (458–478 eV) and Pt4f (68–83 eV) with a pass energy of 40 eV. Thermogravimetric analysis (TGA) was conducted using a TGA analyzer (NETZSCH5) at a heating rate of 10 °C min⁻¹ and 100 mL min⁻¹ oxygen-flow. The TEM micrographs were measured using a JEM-2010 (JEOL, acceleration voltage of 120 kV) electron microscope.

Electrochemical measurements

The electrochemical measurements were performed using a glassy carbon electrode (GCE) with a conventional three-electrode system in a vessel at 25 °C. A GCE with a geometric surface area of 0.1257 cm² was used as the working electrode. A Pt wire and Ag/AgCl were used as the counter and reference electrodes, respectively. The potential of the electrode was controlled using a CHI604e potentiostat. The electrocatalyst ink was prepared as follows. The electrocatalyst (1.0 mg) was ultrasonically dispersed in 80% isopropanol aq. (v/v = 4 : 1, 2.0 mL) to form a homogeneous suspension, which was casted on the GCE and the loading amount of Pt was controlled at 14 μg cm⁻², then air-dried. Cyclic voltammetry (CV) measurements were carried out at a scan rate of 50 mV s⁻¹ in N₂-saturated 0.5 M H₂SO₄ electrolyte after activation of the electrocatalyst and electrochemical surface area (ECSA) values were determined from CV curves. All the potentials were referenced to the reference hydrogen electrode (RHE).

Methanol oxidation reaction (MOR)

The MOR was evaluated in N₂-saturated 0.5 M H₂SO₄ and 1 M methanol at 25 °C with a scan rate of 50 mV s⁻¹.

Pt stability test

The Pt stability was tested using the protocol of the Fuel Cell Commercialization Conference of Japan (FCCJ)^25,26 (measured in N₂-saturated 0.5 M H₂SO₄ at 25 °C), in which the potential was kept at 0.6 V vs. RHE for 3 s, then applied up to 1 V vs. RHE for another 3 s. The procedure was cycled, and the CV measurement was carried out after every 600 cycles (see ESI, Fig. S1†).

Fuel cell test

The FC performance of the fabricated MEAs was evaluated at 60 °C using a computer-controlled fuel cell testing system (Model 850e, Scribner Associate, Inc.). The polarization and power density curves were obtained at atmospheric pressure by flowing 1 M methanol (flow rate = 3 mL min⁻¹) and 100% humidified air (flow rate = 100 mL min⁻¹) to the anode and cathode, respectively.

Results and discussion

After the synthesis of the two electrocatalysts, the XPS measurement was carried out to determine the typical elements in the electrocatalysts. Commercial CB/PtRu (Pt: 40 wt%, Ru: 20 wt%, Alfar Aesar) was used as the control sample in this study. The C1s peak was calibrated to 284.5 eV as standard as shown in Fig. S2.† As shown in Fig. 2a, clear peaks were observed at 400 eV for CB/PBI/Pt/PVP and CB/PBI/PVP/Pt coming from the PBI and PVP. Fig. 2b shows two clear peaks at 71.4 eV and 75.0 eV attributed to Pt4f_7/2 and Pt4f_5/2, respectively, suggesting that the dominant valences of the Pt species in the two prepared electrocatalysts were zero.^27,28 The Ru3p peak of the commercial CB/PtRu was detected at 463 eV as shown in Fig. 2c.²⁹

Fig. 2 XPS spectra of narrow scan in N1s (a), Pt4f (b) and Ru3p (c) regions of CB/PtRu (black line), CB/PBI/Pt/PVP (blue line) and CB/PBI/PVP/Pt (red line).

In order to determine the PVP amount in the two electrocatalysts, TGA measurements were carried out and the results are shown in Fig. 3. The PVP amount in CB/PBI/Pt/PVP was evaluated from the decrease in weight residue between CB/PBI/Pt (40.5 wt%) and CB/PBI/Pt/PVP (38.6 wt%). The PVP amount was 5.4 wt% based on a constant weight ratio between CB/PBI and Pt, which is consistent with the result from the weight loss of CB/PBI/Pt/PVP before 300 °C (blue line). The PVP amount in CB/PBI/PVP/Pt was evaluated from the weight loss of CB/PBI/PVP/Pt before 300 °C (red line) since the decomposition temperature of CB and PBI was higher than 300 °C. The PVP amount in CB/PBI/PVP/Pt was determined to be 5.0 wt%. The XPS and TGA results suggested that PVP existed in both of the two electrocatalysts due to the hydrogen bonding between PBI and PVP.³⁰

Fig. 3 TGA curves of CB/PBI/Pt (green line), CB/PBI/Pt/PVP (blue line) and CB/PBI/PVP/Pt (red line) measured from room temperature to 900 °C under stable oxygen atmosphere.

It is of importance to measure the morphologies of the two prepared electrocatalysts before the electrochemical measurements. The commercial CB/PtRu was used as the control sample and the PtRu nanoparticles were homogeneously deposited on CB with a diameter of 3.2 ± 0.2 nm as shown in Fig. 4a (for histogram, see ESI, Fig. S3a†). As shown in Fig. 4b and c, the Pt nanoparticles were also uniformly deposited on the carbon materials due to the Pt–N bonding between Pt and PBI or PVP reported by us.^20,24 The Pt sizes were 3.0 ± 0.2 nm and 3.2 ± 0.1 nm for CB/PBI/Pt/PVP and CB/PBI/PVP/Pt, respectively (for histogram, see ESI, Fig. S3b and c†), indicating that the three electrocatalysts showed similar particle size. It should be noted that the PVP layer on CB/PBI/Pt was approximately 1 nm.¹⁹ Also, the PVP layer was detected by HR-TEM as shown in Fig. S4.†

Fig. 4 TEM images of CB/PtRu (a), CB/PBI/Pt/PVP (b) and CB/PBI/PVP/Pt (c) before the durability test.

Electrochemical surface area (ECSA) is an essential parameter to evaluate the activity of the electrocatalyst, which is calculated from the equation: ECSA = Q_H/(210 × Pt loading amount), where Q_H is the charge of the hydrogen electro-adsorption from 0.05 to 0.3 V vs. RHE.^31,32 From Fig. 5, the initial ECSAs of the commercial CB/PtRu, CB/PBI/Pt/PVP and CB/PBI/PVP/Pt were calculated to be 71.0 m² g_Pt⁻¹, 50.3 m² g_Pt⁻¹ and 81.7 m² g_Pt⁻¹, respectively. By comparison with CB/PBI/Pt/PVP, CB/PBI/PVP/Pt showed 1.6 times higher ECSA due to the coverage of Pt active sites by PVP in CB/PBI/Pt/PVP. The Pt utilization efficiency of CB/PBI/PVP/Pt reached 94%, which is calculated from the equation: Pt utilization efficiency = 6/(ρd × ECSA), where ρ and d are the Pt density and diameter from TEM, respectively.³³ However, the Pt utilization efficiencies of commercial CB/PtRu and CB/PBI/Pt/PVP were only 81% and 55%. The lower Pt utilization efficiency of CB/PBI/Pt/PVP is attributed to the coverage of Pt active sites by the PVP layer. The higher Pt utilization efficiency of CB/PBI/PVP/Pt suggested that the pre-coating of PVP polymer was vital for the enhancement in the Pt activity. Subsequently, the durability test was carried out based on the protocol from Fuel Cell Commercialization of Japan (FCCJ, ESI, Fig. S1†), in which the potential was dynamically scanned from 0.6 V to 1.0 V vs. RHE in order to measure the stability of the Pt nanoparticles. As shown in Fig. 5a, the hydrogen adsorption/desorption peaks of the commercial CB/PtRu showed a sharp decrease, which was due to the dissolution of Ru nanoparticles in the acidic medium and easy aggregation of Pt nanoparticles. After the durability test, the electrocatalyst was collected from the electrode and measured using XPS. As shown in Fig. S5,† the Ru3p peak disappeared in CB/PtRu after the durability test indicating that Ru is not stable in the acidic medium. After 4200 potential cycles, the ECSA of the commercial CB/PtRu lost almost 50%. For the newly synthesized CB/PBI/Pt/PVP and CB/PBI/PVP/Pt, the ECSAs were almost stable during the potential cycling and lost only 17% and 20%, respectively, as shown in Fig. 5d. CB/PBI/Pt/PVP showed the highest stability because the Pt nanoparticles were structurally sandwiched by the PBI and PVP polymers. Interestingly, CB/PBI/PVP/Pt showed almost similar loss in ECSA compared to CB/PBI/Pt/PVP, suggesting that the pre-coating or post-coating of PVP polymer showed almost negligible effect on the Pt stability. After the durability test, the nanoparticle size had grown to 6.1 ± 2.7 nm for the commercial CB/PtRu as shown in Fig. 6a, while the Pt sizes increased to 3.8 ± 0.3 nm and 4.0 ± 0.3 nm for CB/PBI/Pt/PVP and CB/PBI/PVP/Pt as shown in Fig. 6b and c, respectively (for histograms, see ESI, Fig. S6†). After the durability test, the Pt utilization efficiency of CB/PBI/PVP/Pt was still 93%, suggesting that the CB/PBI/PVP/Pt showed stability and the highest Pt utilization efficiency during the durability test.

Fig. 5 Durability results of the commercial CB/PtRu (a), CB/PBI/Pt/PVP (b) and CB/PBI/PVP/Pt (c). (d) Plots of the normalized ECSAs of CB/PtRu (black line), CB/PBI/Pt/PVP (blue line) and CB/PBI/PVP/Pt (red line) as a function of the number of potential cycles from 0.6 V to 1.0 V vs. RHE.

Fig. 6 TEM images of the commercial CB/PtRu (a), CB/PBI/Pt/PVP (b) and CB/PBI/PVP/Pt (c) after the durability test.

The CO tolerance of the electrocatalyst was evaluated before and after the durability test. As shown in Fig. 7a, the CO oxidation peak of the commercial CB/PtRu was located at 679 mV vs. RHE before the durability test, which was much lower compared to those of CB/PBI/Pt/PVP (780 mV) and CB/PBI/PVP/Pt (775 mV) due to the presence of Ru facilitating the removal of CO species on the Pt nanoparticles by Ru(OH)_ads species. The comparable location of the CO oxidation peak was attributed to the similar PVP amounts in the electrocatalysts, indicating that the pre-coating of the PVP layer showed a negligible effect on the CO tolerance of the electrocatalyst. Meanwhile, the CO oxidation peaks of these two electrocatalysts were negatively shifted compared to that of CB/PBI/Pt as shown in Fig. S7† due to the presence of PVP facilitating the water adsorption to form Pt(OH)_ads species because the surfaces of the electrocatalysts are hydrophilic after the introduction of PVP to the electrocatalysts. And Pt(OH)_ads species are also important for the enhancement in the CO tolerance of the electrocatalysts since Pt(OH)_ads species react with Pt(CO)_ads species to recover the CO-poisoned Pt (Pt(OH)_ads + Pt(CO)_ads → 2Pt + CO₂ + H⁺ + e⁻). After the durability test, due to the absence of Ru, the CO oxidation peak of the commercial CB/PtRu largely shifted to 1113 mV vs. RHE; in sharp contrast, the CO oxidation peaks of CB/PBI/Pt/PVP and CB/PBI/PVP/Pt were almost stable after the durability test due to the presence of PVP, since PVP plays a prominent role in the CO tolerance of the electrocatalysts.¹⁹ Although the commercial CB/PtRu showed better CO tolerance before the durability test, the CO tolerance had seriously deteriorated after the durability test, which shortens the lifetime of DMFCs and degrades the fuel cell performance. Stable CO tolerance is essential for long-term operation of DMFCs. It should be noted that the ECSAs of the commercial CB/PtRu, CB/PBI/Pt/PVP and CB/PBI/PVP/Pt calculated from CO stripping curves (ECSA = Q_H/(420 × Pt loading amount), Q_H is the charge of the CO oxidation peak) were 68.5, 47.3 and 79.1 m² g_Pt⁻¹, respectively, which were comparable to the ECSA values evaluated from the cyclic voltammetry curves.

Fig. 7 CO stripping voltammograms of the commercial CB/PtRu (a), CB/PBI/Pt/PVP (b) and CB/PBI/PVP/Pt (c) before (solid line) and after (dotted line) the durability test.

The methanol oxidation reaction (MOR) was carried out to evaluate the activity of the electrocatalyst. A shown in Fig. 8, the current density of CB/PBI/PVP/Pt was 1.35 A mg_Pt⁻¹, which was higher compared to those of CB/PBI/Pt/PVP (0.94 A mg_Pt⁻¹) and commercial CB/PtRu (0.8 A mg_Pt⁻¹). Relative to CB/PBI/Pt/PVP, CB/PBI/PVP/Pt showed better MOR activity because the Pt nanoparticles were not covered by any polymer and the methanol can be easily adsorbed and oxidized on the Pt nanoparticles. The PVP coating in CB/PBI/Pt/PVP partially blocked the Pt active sites and showed unavoidable loss in activity. Subsequently, the membrane electrode assemblies (MEAs) were fabricated from the three different electrocatalysts and the fuel cell performance was measured at 60 °C. The power density of MEA_{CB/PBI/PVP/Pt/PVP} reached 104 mW cm⁻², which was much higher compared to MEA_{CB/PBI/Pt/PVP} (73 mW cm⁻²) and MEA_CB/PtRu (63 mW cm⁻²) (Fig. 9). Thus, the pre-coating of PVP polymer (CB/PBI/PVP/Pt) enhanced the fuel cell performance and maintained comparable CO tolerance to the post-coating of PVP polymer (CB/PBI/Pt/PVP). To the best of our knowledge, the power density of MEA_{CB/PBI/PVP/Pt} was one of the highest values in recent publications as listed in Table 1.

Fig. 8 Methanol oxidation reaction (MOR) curves of the commercial CB/PtRu (black line), CB/PBI/Pt/PVP (blue line) and CB/PBI/PVP/Pt (red line) before the durability test.

Table 1 Comparison of the power densities reported in recent literature

Electrocatalyst	Temp. (°C)	Pt loading (mg cm^−2)	Power density (mW cm^−2)	Ref.
CB/PBI/PVP/Pt	60	2	104	This work
C/PtGe	70	1	28	34
C/PtRu	60	0.3	15.3	35
Graphene–TiO2/Pt	80	2.5	12	36
MC/Pt	60	3	67	37
Ni2P/Pt	60	1.2	65	6
NiO/Pt	60	1	46	38
Ti–C/Pt	80	1	50	39
Pt–NF/C	75	4	64	40

---

Fig. 9 I–V and power density curves of MEAs fabricated from CB/PtRu (black line), CB/PBI/Pt/PVP (blue line) and CB/PBI/PVP/Pt (red line) as the anodic electrocatalyst measured at 60 °C with 1 M methanol and 100% humidified air to anode and cathode, respectively.

Conclusions

In summary, we synthesized a stable CO tolerant electrocatalyst via bottom-up design. The PVP pre-coated electrocatalyst (CB/PBI/PVP/Pt) showed comparable CO tolerance and durability to the PVP post-coated electrocatalyst (CB/PBI/Pt/PVP), while CB/PBI/PVP/Pt exhibited higher methanol oxidation reaction activity and fuel cell performance due to the non-coated Pt nanoparticles. Meanwhile, the Pt utilization efficiency as well as the CO tolerance of CB/PBI/PVP/Pt was stable during the durability test. The newly fabricated electrocatalyst is utilizable in real DMFC applications.

Acknowledgements

Z. Yang gratefully acknowledges support of a start-up grant from China University of Geosciences Wuhan (CUG).

Notes and references

1. X. Zhao, M. Yin, L. Ma, L. Liang, C. Liu, J. Liao, T. Lu and W. Xing, Energy Environ. Sci., 2011, 4, 2736–2753.
2. C. Koenigsmann and S. S. Wong, Energy Environ. Sci., 2011, 4, 1161–1176.
3. Y. Paik, S. S. Kim and O. H. Han, Angew. Chem., Int. Ed., 2008, 47, 94–96.
4. L. Cao, F. Scheiba, C. Roth, F. Schweiger, C. Cremers, U. Stimming, H. Fuess, L. Chen, W. Zhu and X. Qiu, Angew. Chem., Int. Ed., 2006, 45, 5315–5319.
5. D. Sebastián, I. Suelves, E. Pastor, R. Moliner and M. J. Lázaro, Appl. Catal., B, 2013, 132–133, 13–21.
6. J. Chang, L. Feng, C. Liu, W. Xing and X. Hu, Energy Environ. Sci., 2014, 7, 1628–1632.
7. B. N. Grgur, G. Zhuang, N. M. Markovic and P. N. Ross, J. Phys. Chem. B, 1997, 101, 3910–3913.
8. D.-H. Lim, D.-H. Choi, W.-D. Lee and H.-I. Lee, Appl. Catal., B, 2009, 89, 484–493.
9. M. Wakisaka, S. Mitsui, Y. Hirose, K. Kawashima, H. Uchida and M. Watanabe, J. Phys. Chem. B, 2006, 110, 23489–23496.
10. S. Zhou, G. S. Jackson and B. Eichhorn, Adv. Funct. Mater., 2007, 17, 3099–3104.
11. E. I. Santiago, G. A. Camara and E. A. Ticianelli, Electrochim. Acta, 2003, 48, 3527–3534.
12. Z. Liu, G. S. Jackson and B. W. Eichhorn, Energy Environ. Sci., 2011, 4, 1900–1903.
13. H. Huang and X. Wang, J. Mater. Chem. A, 2014, 2, 6266–6291.
14. X. Xue, J. Ge, C. Liu, W. Xing and T. Lu, Electrochem. Commun., 2006, 8, 1280–1286.
15. H. Lv, T. Peng, P. Wu, M. Pan and S. Mu, J. Mater. Chem., 2012, 22, 9155–9160.
16. Z. Yang, M. R. Berber and N. Nakashima, J. Mater. Chem. A, 2014, 2, 18875–18880.
17. Z. Yang, I. H. Hafez, M. R. Berber and N. Nakashima, ChemCatChem, 2015, 7, 808–813.
18. Z. Yang, C. Kim, S. Hirata, T. Fujigaya and N. Nakashima, ACS Appl. Mater. Interfaces, 2015, 7, 15885–15891.
19. Z. Yang and N. Nakashima, ChemCatChem, 2016, 8, 600–606.
20. Z. Yang and N. Nakashima, J. Mater. Chem. A, 2015, 3, 23316–23322.
21. D. He, K. Cheng, H. Li, T. Peng, F. Xu, S. Mu and M. Pan, Langmuir, 2012, 28, 3979–3986.
22. S. Sharma, A. Ganguly, P. Papakonstantinou, X. Miao, M. Li, J. L. Hutchison, M. Delichatsios and S. Ukleja, J. Phys. Chem. C, 2010, 114, 19459–19466.
23. L. Xiao, H. Zhang, T. Jana, E. Scanlon, R. Chen, E. W. Choe, L. S. Ramanathan, S. Yu and B. C. Benicewicz, Fuel Cells, 2005, 5, 287–295.
24. Z. Yang, T. Fujigaya and N. Nakashima, J. Mater. Chem. A, 2015, 3, 14318–14324.
25. A. Ohma, K. Shinohara, A. Iiyama, T. Yoshida and A. Daimaru, ECS Trans., 2011, 41, 775–784.
26. J. Speder, A. Zana, I. Spanos, J. J. K. Kirkensgaard, K. Mortensen, M. Hanzlik and M. Arenz, J. Power Sources, 2014, 261, 14–22.
27. B. Fang, N. K. Chaudhari, M.-S. Kim, J. H. Kim and J.-S. Yu, J. Am. Chem. Soc., 2009, 131, 15330–15338.
28. H. Huang, H. Chen, D. Sun and X. Wang, J. Power Sources, 2012, 204, 46–52.
29. C. Zhou, F. Peng, H. Wang, H. Yu, C. Peng and J. Yang, Electrochem. Commun., 2010, 12, 1210–1213.
30. H. Pu, Q. Liu, L. Qiao and Z. Yang, Polym. Eng. Sci., 2005, 45, 1395–1400.
31. D. Wang, H. L. Xin, R. Hovden, H. Wang, Y. Yu, D. A. Muller, F. J. Disalvo and H. D. Abruña, Nat. Mater., 2013, 12, 81–87.
32. S. Chen, Z. Wei, X. Qi, L. Dong, Y. G. Guo, L. Wan, Z. Shao and L. Li, J. Am. Chem. Soc., 2012, 134, 13252–13255.
33. I. H. Hafez, M. R. Berber, T. Fujigaya and N. Nakashima, Sci. Rep., 2014, 4, 6295.
34. N. S. Veizaga, V. A. Paganin, T. A. Rocha, O. A. Scelza, S. R. de Miguel and E. R. Gonzalez, Int. J. Hydrogen Energy, 2014, 39, 8728–8737.
35. Y. Shimizu, Y. Suda, H. Takikawa, H. Ue, K. Shimizu and Y. Umeda, Electrochemistry, 2015, 83, 381–385.
36. L. Zhao, Z.-B. Wang, J. Liu, J.-J. Zhang, X.-L. Sui, L.-M. Zhang and D.-M. Gu, J. Power Sources, 2015, 279, 210–217.
37. M. M. Bruno, M. A. Petruccelli, F. A. Viva and H. R. Corti, Int. J. Hydrogen Energy, 2013, 38, 4116–4123.
38. T.-Y. Chen, P.-C. Huang, Y.-F. Liao, Y.-T. Liu, T.-K. Yeh and T.-L. Lin, RSC Adv., 2016, 6, 72607–72615.
39. A. Schlange, A. R. dos Santos, B. Hasse, B. J. M. Etzold, U. Kunz and T. Turek, J. Power Sources, 2012, 199, 22–28.
40. K. Lin, Y. Lu, S. Du, X. Li and H. Dong, Int. J. Hydrogen Energy, 2016, 41, 7622–7630.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra24025a

---