Nitrogen-doped graphene/tungsten oxide microspheres as an electro-catalyst support for formic acid electro-oxidation

Abstract

Tungsten trioxide (WO₃) spheres decorated with nitrogen-doped graphene (NRGO–WO₃) were synthesized by applying the spray-drying procedure and characterized for their ability to serve as an electro-catalyst support for formic acid electro-oxidation. A possible mechanism for the formation of NRGO–WO₃ was proposed based on the results of tunneling electron microscopy (TEM) and X-ray photoelectron spectroscopy (XPS). Pd nanoparticles with dimensions of 4.8 nm were loaded onto the surface of NRGO–WO₃ using a conventional sodium borohydride reduction method. The electrocatalytic performances of Pd/NRGO–WO₃ for formic acid oxidation were investigated by using cyclic voltammetry and chronoamperometry. Due to the decrease in the resistance to electron transfer resulting from the modification of N-doped graphene, which produced an excellent electrical conductor, as well as due to the hydrogen spill-over effect, which accelerated the dehydrogenation of formic acid on Pd active sites, a great enhancement of the electrochemical performances was achieved.

---

1. Introduction

With the merits of reduced toxicity, high practical power density and limited crossover through a Nafion membrane, formic acid is considered as a promising alternative fuel for liquid fuel cells.^1–3 The Pd-based catalyst has been shown to exhibit excellent activity and stability towards the formic acid electro-oxidation (FAEO) reaction by overcoming the CO poisoning effect.^4–6 In the past few decades, metal oxides used as carriers or co-catalyst have attracted considerable attention due to their bifunctional mechanism.^7–9 Among these metal oxides, tungsten trioxide (WO₃) has many attractive properties, such as its insolubility in acid solution, hydrogen spill-over effect and the bifunctional mechanism.^10–12 However, WO₃ suffers from a low specific surface area and low electrical conductivity because it is an n-type semiconductor with a reported band gap of about 2.6 to 2.8 eV.

The introduction of carbon materials and surface modifications are generally used to enhance the electrochemical performance of WO₃. Various carbon materials, such as carbon nanotubes¹³ and Vulcan XC-72R carbon,¹⁴ can decrease resistance to charge transport. However, the activity and stability of these electro-catalysts remain unsatisfactory due to the low chance of contacts forming between palladium and WO₃. In addition, our previous research demonstrated that also including some WO₂ could enhance the electrochemical performances of WO₃ for methanol electro-oxidation (MOR) because the presence of WO₂, which is a conductive oxide, decreases the resistance to charge transfer.¹⁵ Unfortunately, however, WO₂ becomes irreversibly oxidized to WO₃ during the methanol electro-oxidation. Consequently, many efforts are still being made to improve the electrochemical performance of WO₃.

Recently, three-dimensional (3D) graphene (GR) has gained much attention as a potential component material of composites with noble metals for MOR or FAEO due to its extraordinary electronic and structural properties.¹⁶ Herein, tungsten oxide microspheres modified with nitrogen-doped 3D graphene have been prepared by applying a simple spray-drying process and a subsequent thermal reduction in a nitrogen atmosphere. This work offers a facile approach for enhancing the electrochemical performance of WO₃ by combining the inclusion of carbon materials with surface modification, which can provide a highly efficient electro-catalyst support for FAEO.

2. Experimental

Preparation of NRGO/WO₃ microsphere

All chemical reagents were of analytical grade. Graphene oxide (GO) was prepared by oxidizing graphite powder (99.9% purity, Alfa Aesar, USA) using the modified Hummers' method.^17,18 Subsequently, 1000 mL aqueous suspensions of GO (0.1 mg mL⁻¹) and ammonium metatungstate (AMT, 5 mg mL⁻¹) were prepared. After being ultrasonicated for 30 min, the mixed solutions were subjected to spray drying using a SD1000 spray dryer (EYELA) under a blower flux of 0.7 m³ min⁻¹, feeding speed of 600 mL h⁻¹, and inlet temperature of 200 °C. The as-prepared powder samples above were then put into a ceramic boat under an atmosphere of nitrogen. The temperature of the furnace was raised to 673 K and held there for 1 h, and then raised to 1073 K for 3 h, with nitrogen all the while having been passed through the furnace. The samples were cooled to room temperature, and were denoted as NRGO–WO₃.

Preparation of Pd/h-WO₃ and Pd/NRGO–WO₃ electrocatalysts

WO₃ microspheres (h-WO₃) were previously synthesized by our groups.¹⁹ Subsequently, the Pd/NRGO-WO₃ catalyst was prepared by using the improved liquid phase reduction method. In a typical procedure, 50 mg of the as-prepared NRGO–WO₃ support and 11.4 mL of a 5 mM H₂PtCl₆ solution were dispersed in 50 mL of water by sonication. Then, after the pH of this resulting solution was adjusted to nearly 9 using 1 M NaOH, a volume of 10 mL of freshly prepared 0.1 M NaBH₄ was added to the solution, followed by stirring for 2 h. The resulting solution was filtered, washed, and dried at 85 °C for 10 h in a vacuum oven, yielding a 20 wt% Pd loading on the supports, which was denoted as Pd/NRGO–WO₃. For comparison, Pd supported on the h-WO₃ catalyst with the same Pd loading amount was prepared by using the same procedure described above, and was denoted as Pd/h-WO₃. Commercial 10 wt% Pd/C (JM Pd/C) was purchased from Johnson Matthy Co., Ltd. and used as contrast samples.

Characterizations

The phases present in the synthesized materials were identified using an X-ray diffractometer (XRD, Panalytical X'Pert Pro, Cu Kα1 radiation source (λ = 0.1541 nm), voltage of 40 kV, current of 40 mA). The morphologies and structures of the products were characterized using a field emission scanning electron microscope (FE-SEM, Hitachi S-4700 II), and a tunneling electron microscope (TEM, Tecnai G2 F30) with an energy-dispersive X-ray spectroscope (EDS). Thermogravimetric analysis (TGA) was performed with a TA Q600 at a heating rate of 10 °C min⁻¹ from room temperature to 1200 °C under air flow (100 mL min⁻¹). X-ray photoelectron spectroscopy (XPS) was carried out on a Kratos AXIS Ultra DLD. All XPS spectra were corrected using the C 1s line at 284.6 eV. Curve fitting and background subtraction were performed.

Electrochemical measurements

The electrochemical measurements were performed using a three-electrode cell with an Ivium electrochemical workstation at room temperature. To prepare the electrode, a mass of 2 mg of electrocatalyst was ultrasonically mixed with 200 μL of an ethanol–water solution to form a homogeneous ink, followed by administering 2 μL of the electrocatalyst ink onto the surface of a glassy carbon (GC) electrode (with a diameter of 3 mm), to which was added 5 μL of 1.0% Nafion (DuPont, USA) in ethanol to fix the electrocatalyst on the GC electrode surface. A glassy carbon disk electrode served as the substrate for the support. Prior to use, the GC electrode was polished using an aqueous alumina suspension. And then the catalyst suspension was pipetted, using a micropipette, onto the GC surface to produce a Pd loading of about 0.056 mg cm⁻². A Pt foil and saturated calomel electrode (SCE) were used as the counter and reference electrodes, respectively. The electrocatalytic activity of the catalysts for the oxidation of formic acid was studied in a 0.5 M H₂SO₄ aqueous solution containing 1.0 M HCOOH at a scan rate of 50 mV s⁻¹.

3. Results and discussion

Structure and morphology

The XRD patterns of the synthesized samples are shown in Fig. 1. The main diffraction peaks matched well with those of the orthorhombic phase WO₃ (JCPDS-20-1324) except for the weak diffraction peaks at 39.4°, 45.8° and 66.9°, which resulted from the Pd(111), (200) and (220) planes, respectively. No obvious characteristic GO peak at about 11° was observed in the diffraction pattern of the NRGO–WO₃ sample (Fig. 1b), which suggests that GO was reduced to GR during the thermal treatment.²⁰ Meanwhile the characteristic (002) peak of graphite at about 26° was also absent from this diffraction pattern (Fig. 1b), due to the overlap of the (120) peak of WO₃. Moreover, the overall intensity of the NRGO–WO₃ diffraction was weaker than that of the h-WO₃ diffraction; it was weaker because the surface of NRGO–WO₃ was encapsulated by the carbon, as shown by the TEM results described below. A thermogravimetric analysis (TGA) of the NRGO–WO₃ composite was also carried out, in air, in order to quantify its content of carbon. The TGA (Fig. 2) yielded a slight weight loss up to a temperature of 100 °C, which could be attributed to dehydration. A 2% loss of weight was observed from 200 °C to 1200 °C, and this weight loss was attributed to combustion of the graphene,²¹ and was close to the actual amount of graphene loaded.

Fig. 1 XRD patterns of (a) h-WO₃, (b) NRGO–WO₃ and (c) Pd/NRGO–WO₃.

Fig. 2 TGA curve of the NRGO–WO₃ sample in air.

The structures and morphologies of as-prepared samples were characterized by using scanning electron microscopy (SEM) and tunnel electron microscopy (TEM) analyses. The SEM images (Fig. 3) of the as-prepared samples showed well-defined spherical structures with the diameters of about 1–5 μm. Graphene displayed a three-dimensional crumpled structure (Fig. 3a and c), due to π–π stacking and van der Waals attraction, and this crumpled structure prevented its irreversible aggregation and stacking.²² The TEM results further revealed the smooth-surfaced nature of the spherical structures (Fig. 4a). The high-resolution TEM (HRTEM) image of NRGO-WO₃ showed its hierarchical core–shell structure (Fig. 4b). The C, W and N mapping images (Fig. 4e) demonstrated uniform distributions of these elements. According to the results of the TEM analysis above, it could concluded that the surface of WO₃ was decorated by nitrogen-doped graphene.

Fig. 3 SEM images of (a and c) NRGO–WO₃ and (b and d) h-WO₃.

Fig. 4 TEM images of (a and b) NRGO–WO₃ and (c and d) Pd/NRGO–WO₃, and (e) elemental mapping of NRGO–WO₃. The Pd particle size distribution of Pd/NRGO–WO₃ is shown in the inset of panel (d).

Palladium particles were loaded on the surface of NRGO–WO₃ or GR by means of the conventional sodium borohydride reduction method. As can be seen in Fig. 4c and d, palladium particles with narrow size distributions were well dispersed on the surfaces of NRGO–WO₃ and GR. The average size of the Pd/NRGO–WO₃ catalyst particles was estimated from their histogram to be approximately 5 nm. X-ray photoelectron spectroscopy (XPS) was used to investigate the composition and surface chemical state of Pd/NRGO–WO₃. Fig. 5 shows segments of the XPS spectrum of this catalyst corresponding to the C 1s, W 4f, N 1s and Pd 3d photoemissions. As shown in Fig. 5a, the C 1s peak can be deconvoluted into four peaks. The main peak located at 284.7 eV can be assigned to the C–C bond, with the other small peaks at 288.9 eV, 286.8 eV, 285.6 eV attributed to the –COOH, –C–O–C and –C–OH functional groups, respectively.²³ The Pd 3d peak was also deconvoluted into two pairs of doublets (Fig. 5d), which were assigned as metallic Pd (335.1 eV and 340.9 eV) and Pd²⁺ (337.2 eV and 342.9 eV), suggesting that the Pd was mainly in the metallic state.²⁴ In addition, the peak located at 399.5 eV (Fig. 5c) was attributed to the nitrile N species, suggesting that N was incorporated into the GR structure, consistent with the STEM (scanning transmission electron microscopy) result. No obvious characteristic peak of the low valence state of tungsten (Fig. 5c) was observed, suggesting that WO₃ was not reduced by GR during the thermal treatment.

Fig. 5 XPS spectrum of Pd/NRGO–WO₃, showing the peaks for the (a) C 1s, (b) N 1s, (c) W 4f and (d) Pd 3d elements.

Based on these TEM and XPS results, we suggest that the as-prepared NRGO–WO₃ composites formed using the following process. After the as-prepared GO was sonicated with AMT in deionized water and the mixed solution including GO and AMT was pumped into the spray-drying apparatus, the GO sheets spontaneously assembled on the surface of the droplets due to their amphiphilicity and subsequently encapsulated the AMT microspheres during the spray drying. Then, during the calcination under an atmosphere of nitrogen, ammonia gas was generated as a result of AMT decomposition, and was used as a nitrogen source to react with functional groups such as carboxyl and hydroxyl on the surface of the GO, which was also simultaneously reduced to GR by the heat treatment. Finally, the surfaces of the WO₃ microspheres were modified by the nitrogen-doped graphene to form the observed core–shell hierarchical structure.

Activity toward formic acid oxidation

Fig. 6a shows the cyclic voltammograms (CV) of three catalyst electrodes, each in a solution of 0.5 M H₂SO₄ and 1 M HCOOH at a scan rate of 50 mV s⁻¹. In each case, the current density was calculated by normalization to the mass of loaded Pd because the intercalation of protons in WO₃ to form tungsten bronzes occurred in the same potential region (−0.24 to 0.05 V) as did the hydrogen underpotential deposition (H-UPD) on Pd.^25,26 As shown in Fig. 6a, of the samples tested, the Pd/NRGO–WO₃ catalyst exhibited the best activity for formic acid oxidation, with a peak current density of 932 mA mg⁻¹. This value was specifically 2.4 and 3.5 times greater than the 380 mA mg⁻¹ and 264 mA mg⁻¹ values displayed by the JM Pd/C and Pd/h-WO₃ catalysts, respectively. It was also better than those previously reported for Pd/WO₃ catalysts,^27,28 indicating that the modification of N-doped graphene greatly enhanced the kinetics of the Pd/WO₃ catalyst for the FAEO reaction. This enhanced kinetics may have been due to the decrease of the resistance to electron transfer resulting from the modification of N-doped graphene that produced an excellent electrical conductor as well as due to the hydrogen spill-over effect accelerating the dehydrogenation of formic acid on the Pd active site.²⁶ Nitrogen and oxygen on the surface of GR provided many active sites for anchoring Pd nanoparticles, which was also beneficial for the FAEO reaction. Also, as seen in Fig. 6b, the current density decreased rapidly with increasing cycle number for the commercial JM Pd/C catalysts, in contrast to that observed for the other two electro-catalysts, indicating that the introduction of WO₃ played an important role during the FAEO reaction.

Fig. 6 (a) Cyclic voltammetric curves and (b) CV stability performances of Pd/NRGO–WO₃, Pd/h-WO₃ and JM Pd/C catalysts each in a solution of 0.5 M H₂SO₄ and 1 M HCOOH at a scan rate of 50 mV s⁻¹.

The electro-catalytic stability of the various catalysts were further investigated by performing chronoamperometry (CA) tests at a constant voltage of 0.1 V (Fig. 7) for 6000 s in a solution containing 0.5 M H₂SO₄ and 1 M HCOOH. The stability of the three catalysts for the FAEO reaction followed the order Pd/NRGO–WO₃ > JM Pd/C > Pd/h-WO₃. The significantly greater catalytic stability of Pd/NRGO–WO₃ may have been due to the hydroxyl functional groups on the surfaces of tungsten oxide and N-doped GR facilitating the oxidation of poisonous CO intermediates on the active Pd sites.^29,30 Combining these results with the CA and CV results, we conclude that, of the catalysts tested, the Pd/NRGO–WO₃ catalyst has the highest catalytic performance toward the FAEO reaction.

Fig. 7 Chronoamperometric curves of Pd/NRGO–WO₃, Pd/h-WO₃ and JM Pd/C catalysts at 0.1 V in a solution of 0.5 M H₂SO₄ and 1 M HCOOH.

4. Conclusions

In summary, we have demonstrated a facile approach for enhancing the electrochemical performance of WO₃ by combining the inclusion of carbon materials with surface modification, which can produce a highly efficient electro-catalyst support for formic acid electro-oxidation. The catalytic performance of the as-prepared Pd/NRGO–WO₃ towards formic acid oxidation was superior to the performances of commercial Pd/C and Pd/h-WO₃ catalysts. The inherent characteristics of NRGO–WO₃ composites may also make them suitable for applications in photocatalysis, photochromism and gas sensing.

Acknowledgements

The research was financially supported by the National Natural Science Foundation of China (No. 51404110 and 21506187), Education Department Project Fund of Jiangxi Province (GJJ150665 and GJJ150633) and the Jiangxi Natural Science Foundation of China (No. 20151BAB213011).

Notes and references

1. J. Zhao, Z. S. Liu, H. Q. Li, W. B. Hu, C. Z. Zhao, P. Zhao and D. L. Shi, Langmuir, 2015, 31, 2576–2583.
2. L. Y. Zhang, Z. L. Zhao and C. M. Li, Nano Energy, 2015, 11, 71–77.
3. K. Zhang, F. F. Ren, H. W. Wang, C. Q. Wang, M. S. Zhu and Y. K. Du, Chempluschem, 2015, 80, 529–535.
4. P. Kannan, T. Maiyalagan and M. Opallo, Nano Energy, 2013, 2, 677–687.
5. T. Jin, S. J. Guo, J. L. Zuo and S. H. Sun, Nanoscale, 2013, 5, 160–163.
6. H. Y. Qian, H. J. Huang and X. Wang, J. Power Sources, 2015, 275, 734–741.
7. B. Habibi and N. Delnavaz, RSC Adv., 2015, 5, 73639–73650.
8. Q. Du, J. B. Wu and H. Yang, ACS Catal., 2014, 4, 144–151.
9. H. J. Huang and X. Wang, Phys. Chem. Chem. Phys., 2013, 15, 10367–10375.
10. A. U. Rehman, S. S. Hossain, S. U. Rahman, S. Ahmed and M. M. Hossain, Appl. Catal., A, 2014, 482, 309–317.
11. L. G. Feng, L. A. Yan, Z. M. Cui, C. P. Liu and W. Xing, J. Power Sources, 2011, 196, 2469–2474.
12. K. Y. Chen, P. K. Shen and A. C. C. Tseung, J. Electrochem. Soc., 1995, 142, L54–L56.
13. C. A. Ma, Y. X. Jin, M. Q. Shi, Y. Q. Chu, Y. H. Xu, W. P. Jia, Q. H. Yuan, J. B. Chen, H. L. Pan and Q. W. Dai, Ionics, 2014, 20, 1419–1426.
14. Z. H. Zhang, Y. J. Huang, J. J. Ge, C. P. Liu, T. H. Lu and W. Xing, Electrochem. Commun., 2008, 10, 1113–1116.
15. Y. Zhou, X.-C. Hu, X.-H. Liu and H.-R. Wen, Chem. Commun., 2015, 51, 15297–15299.
16. Y. Zhou, X.-C. Hu, Q.-z. Fan and H.-R. Wen, J. Mater. Chem. A, 2016, 4, 4587–4591.
17. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
18. L. J. Cote, F. Kim and J. X. Huang, J. Am. Chem. Soc., 2009, 13, 1043–1049.
19. Y. Zhou, X. C. Hu, Y. J. Xiao and Q. Shu, Electrochim. Acta, 2013, 111, 588–592.
20. J. Z. Wang, C. Zhong, D. Wexler, N. H. Idris, Z. X. Wang, L. Q. Chen and H. K. Liu, Chem.–Eur. J., 2011, 17, 661–667.
21. X. W. Yang, J. W. Zhu, L. Qiu and D. Li, Adv Mater, 2011, 23, 2833–2877.
22. J. Y. Luo, H. D. Jang, T. Sun, L. Xiao, Z. He, A. P. Katsoulidis, M. G. Kanatzidis, J. M. Gibson and J. X. Huang, Acs Nano, 2011, 5, 8943–8949.
23. A. Halder, S. Sharma, M. S. Hegde and N. Ravishankar, J. Phys. Chem. C, 2009, 113, 1466–1473.
24. Y. Li, X. B. Fan, J. J. Qi, J. Y. Ji, S. L. Wang, G. L. Zhang and F. B. Zhang, Nano Res., 2010, 3, 429–437.
25. S. Jayaraman, T. F. Jaramillo and E. W. McFarland, Abstr. Pap. Am. Chem. Soc., 2005, 229, U1094.
26. I. A. Rutkowska, D. Marks, C. Perruchot, M. Jouini and P. J. Kulesza, Colloids Surf., A, 2013, 439, 200–206.
27. X. C. Hu, Y. Zhou, H. R. Wen and H. M. Zhong, J. Electrochem. Soc., 2014, 161, F583–F587.
28. Y. Zhou, X. C. Hu, L. Q. Li and X. R. Chen, Acta Chim. Sin., 2014, 30, 83–87.
29. P. K. Shen and A. C. C. Tseung, J. Electrochem. Soc., 1994, 141, 3082–3090.
30. S. Sharma, A. Ganguly, P. Papakonstantinou, X. P. Miao, M. X. Li, J. L. Hutchison, M. Delichatsios and S. Ukleja, J. Phys. Chem. C, 2010, 114, 19459–19466.

---