Nano ceria supported nitrogen doped graphene as a highly stable and methanol tolerant electrocatalyst for oxygen reduction

Abstract

Ceria (CeO₂) nanoparticles with ellipsoid shape are coupled on a nitrogen doped reduced graphene oxide sheet through a single step solvothermal procedure. This non-precious-metal based nanocomposite material displayed enhanced electrochemical oxygen reduction activity with a 4e⁻ process similar to commercial Pt/C but with much higher stability and methanol tolerant properties.

---

Direct Methanol Fuel Cells (DMFCs) are electrochemical systems that convert chemical energy from methanol and oxygen to electrical energy without combustion. DMFCs are expected to be a next generation green energy technology. However, the poor kinetics of the cathodic oxygen reduction reaction (ORR) significantly obstructs its commercialization.¹ Generally for DMFCs and other electrochemical technologies, ORR catalysis by a 4-electron pathway (O₂ + 2H₂O + 4e⁻ → 4OH⁻) is highly preferred over the 2-electron peroxide pathway (O₂ + H₂O + 2e⁻ → HO₂⁻ + OH⁻).² Conventionally, platinum based materials are considered to be the most efficient electro-catalyst for ORR.³ On the other hand, their limited availability, high cost, long term electrochemical instability and poor tolerance to fuel molecules (methanol) are building a technological hurdle and henceforth, creating a golden opportunity for the development of non-noble metal based electrocatalysts.⁴ Recently, transition metal oxide (TMO) based catalysts (Co, Mn, Cu, Fe oxides) on sp² hybridized framework of graphene/nitrogen doped graphene, have shown dramatically high ORR activity with nearly 4e⁻ reduction pathway.⁵ It is well known that TMOs have electro-catalytically active facets but they suffer from low electrical conductivity, which seriously affects the electron transfer process during ORR.⁶ Interestingly, 2D graphene layer having large surface area, high electrical conductivity and excellent chemical stability appeared to be an impeccable choice as support material to enhance the catalytic performance of TMOs.⁶ Recently H. Dai group revealed that in case of N-graphene, the nitrogen atoms not only acts as nucleation site for TMOs but also acts as complementary site for ORR.^5a Again, better coupling in TMOs/N-graphene composite prevents aggregation of TMOs and enhance overall stability of the composite.

In another view, rare-earth metal oxides (RMOs) not only possess unique electronic properties but also possess catalytically active sites. CeO₂ is one of the most important rare earth material due to its excellent catalytic activity and was studied under several nano-technological branches like heterogeneous catalysis,^7a,b biosensors,^7c,d semiconductor devices,^7e battery materials,^7f,g etc. Very recently Lee et al. and Rao et al. successfully synthesized CeO₂ on graphene and carbon nanotube (CNT) and studied its electrocatalytic activities for hydrazine oxidation and acetaldehyde detection respectively.^8a,b On the contrary, rare-earth radioactive metal composites having long half-life period (uranium and thorium oxide doped graphene) also displayed an excellent ORR activity.^9a,b In another ground breaking work, ORR activity of lanthanide based transition metal perovskite was meticulously studied by Sho-Horn et al.^9c Although extensive experimental and theoretical studies on ORR activity of (TMOs)/N-graphene composite have been done, no one has ever tried to evaluate the ORR activity of RMOs/N-graphene composite.

Here, we present a single step solvothermal method for synthesis of CeO₂ on N doped reduced graphene oxide (CeO₂/NrGO) and studied its electrocatalytic ORR activity in alkaline medium. For comparison CeO₂, rGO, CeO₂/rGO, NrGO, was also prepared via a similar route (detailed in ESI†). In a typical reaction, graphene oxide (GO) is prepared by a modified Hummer's method which is further reduced, N-doped in presence of cerium precursor via solvothermal method. To prepare CeO₂/NrGO nanocomposite, a mixture of 250 mg cericammonium nitrate, 50 mg graphene oxide, 2 mL of 6.4 M NH₄OH and 20 mL 1,4-butanediol as solvent were solvothermally treated by microwave at 180 °C for 20 minutes, where NH₄OH is used as nitrogen source (Scheme 1).

Scheme 1 Synthesis of CeO₂/NrGO nanocomposite and ORR.

The morphology of CeO₂/NrGO composite was characterized by transmission electron microscopy (TEM). Fig. 1a shows the typical TEM image of CeO₂/NrGO. It is observed that large numbers of CeO₂ nanoparticles having size 10–15 nm were uniformly distributed on NrGO surface, which indicates the presence of graphene sheet obstructs agglomeration of CeO₂ nanoparticles. The crystallinity of CeO₂ nanoparticles was further examined by high resolution TEM (HRTEM). From HRTEM image it is clear that the nanocrystals are ellipsoid in shape with a well-defined d-spacing of 0.27 nm, which corresponds to (200) plane of CeO₂ (Fig. 1b). The selected area electron diffraction (SAED) patterns of the nanocomposite displayed bright rings, confirming the growth of polycrystalline CeO₂ on NrGO sheet (inset of Fig. 1b). The diameters of the typical CeO₂ nanoparticles are of approximately 10 nm. The crystallographic structure of CeO₂/N-rGO composite was further investigated by powder X-ray diffraction (XRD) technique (Fig. 1c). Three strong XRD peaks observed can be correlated to (111), (200) and (220) peaks of face centred cubic fluoride structure of CeO2 (JCPDS card no. 34-0394).

Fig. 1 (a) TEM (b) HRTEM image of CeO₂/N-rGO composite. The inset of (b) corresponds to the SAED pattern of the composite. (c) XRD pattern of CeO₂/N-rGO. (d) Raman spectra of GO, rGO, CeO₂/rGO and CeO₂/N-rGO.

Raman spectroscopy is a non-destructive and useful tool for characterization of graphene based materials. It can be seen that two remarkable Raman peaks at about 1300 cm⁻¹ and 1590 cm⁻¹ is observed for all graphene based samples, which can be assigned to D and G band of graphene (Fig. 1d). The D band is associated with A_1g mode of structural defects in sp² network and G band is associated with E_2g Raman mode of graphitic carbons in graphene. Generally, Tunistra–Koening relation (I_D/I_G ratio) is used to determine the structural defects in graphene sheets.¹⁰ The observed I_D/I_G ratio (∼1.3) is almost same for all the materials, indicating that growth of CeO₂ and N doping in rGO didn't affect the crystallinity of sp² network.

To further investigate the structural modification process from GO to CeO₂/N-rGO, FTIR spectra of GO, rGO, NrGO, CeO₂/rGO and CeO₂/N-rGO were taken (Fig. 2a). The FTIR spectra of GO exhibits characteristics transmittance band at 1738 cm⁻¹, which is attributed to C=O stretching of COOH. The bands at 1222 and 1057 cm⁻¹ are attributed to C–O stretching of C–O–C and C–OH respectively and band at 3434 cm⁻¹ is for O–H stretching vibration. Again, the bands at 1624 and 1397 cm⁻¹ are attributed to bending vibration of water molecules and OH groups. A significant decrease in the intensity of peaks for all the oxygen functionalized groups in FTIR spectrum of rGO, NrGO, CeO₂/rGO and CeO₂/N-rGO confirms the successful removal of most of the oxygen species from the graphene backbone. A new band at around 460 cm⁻¹ is observed for both CeO₂/rGO and CeO₂/N-rGO can be assigned to Ce–O stretching vibrational mode. Furthermore, X-ray photoelectron spectroscopy (XPS) measurement was performed to understand the composition and chemical nature of CeO₂/N-rGO composite. The peaks at about 286, 399, 531, 885 and 900 eV corresponds to C 1s, N 1s, O 1s, Ce 3d_5/2 and Ce 3d_3/2 respectively (Fig. 2b). The N 1s spectrum at 399 eV is deconvoluted to investigate the nature of nitrogen doping, which indicates the presence of 56.4% pyridinic N (399.1 eV) and 43.6% pyrrolic N (400.2 eV) species (Fig. 2c). The deconvolution of C 1s spectrum showed that there were 46.2% C=C (284.5 eV), 34.6% C=N & C–O (286.2 eV) and 19.2% C–N & C=O groups (288.6 eV), indicating the carbon atom is bonded to N and O (Fig. 2d). The O 1s spectrum is deconvoluted into three peaks at 529.2 (CeO₂), 531.0 (COOH) and 533.2 eV (OH) suggesting formation of CeO₂ on NrGO (Fig. 2e). Moreover, the chemical oxidation states of Ce in CeO₂/NrGO are investigated by deconvoluting Ce 3d peaks. The peaks at 881.6 and 884.7 eV can be attributed to 3d_5/2 of Ce⁴⁺ and Ce³⁺ core electrons respectively, whereas the peaks at 897.5 and 900.1 eV corresponds to 3d_3/2 of Ce³⁺ and Ce⁴⁺ respectively (Fig. 2f). It is well known that CeO₂ has a cubic fluorite crystal structure, in which Ce⁴⁺ cation is surrounded by eight O²⁻ anions in a cube, with each O²⁻ coordinated to four Ce⁴⁺ cations. Therefore the change in oxidation state from Ce⁴⁺ to Ce³⁺ causes oxygen vacancies in the crystal lattice, which can play a vital role in ORR catalysis. Furthermore, the concentration of N and Ce in the CeO₂/NrGO composite were calculated by taking the ratio of the integrated peak area for N and Ce to the total area under XPS spectrum. In CeO₂/NrGO, the concentration N and Ce were estimated to be 6.4 and 33.3 at% respectively.

Fig. 2 (a) FTIR spectrum of GO, rGO, NrGO, CeO₂/rGO and CeO₂/N-rGO. (b) XPS survey scan of CeO₂/N-rGO. Deconvoluted XPS spectrum of (c) N 1s, (d) C 1s, (e) O 1s and (f) Ce 3d.

To investigate ORR performances of CeO₂/NrGO, linear sweep voltammetry (LSV) measurements on a rotating disc electrode (RDE) for all the materials along with commercial 20 wt% Pt/C were done at 1000 rpm in O₂ saturated 0.1 M KOH solution (scan rate 10 mV s⁻¹) (Fig. 3a). It is observed that CeO₂ and rGO have similar ORR onset potential (E_onset = −0.23 V vs. Ag/AgCl). But a positive shift of 30 mV for CeO₂/rGO (E_onset = −0.2 V) indicates a synergetic interaction between the individual counterparts, which facilitate ORR. The onset potential of NrGO is measured to be −0.15 V vs. Ag/AgCl. Again, a further positive shift of 50 mV in E_onset for CeO₂/NrGO (−0.15 V) in comparison to CeO₂/rGO suggests strong interaction between CeO₂ and NrGO through nitrogen species in NrGO which facilitate O₂ adsorption and subsequent reduction. Although a similar onset potential has been observed for both CeO₂/NrGO and NrGO, in case of CeO₂/NrGO the increased current density in all the potentials and enhanced E_1/2 (40 mV) as compared to NrGO confirms the synergic effect. Furthermore, the diffusion limited current density of −4.0 mA cm⁻² at −0.9 V for CeO₂/NrGO, which is very close to commercial Pt/C (−4.2 mA cm⁻²) and is also considerably larger than that of CeO₂/rGO (−3.6 mA cm⁻²), NrGO (−2.6 mA cm⁻²), rGO (−1.9 mA cm⁻²) and CeO₂ (−1.8 mA cm⁻²) at −0.9 V. From the above observations we conclude that free CeO₂ nanoparticles shows very poor ORR catalytic activity due to low electrical conductivity. But when it is coupled with rGO or NrGO having high electrical conductivity facilitate the electron transfer process to the active sites and results in the higher ORR activity.

Fig. 3 (a) Linear sweep voltammograms of CeO₂, rGO, NrGO, CeO₂/rGO, CeO₂/NrGO and commercial Pt/C at a rotation rate of 1000 rpm. (b) RDE curves of CeO₂/NrGO in various rotation rates. (c) Koutecky–Levich plot of NrGO–CeO₂ at different potentials. (d) Kinetic limiting current (J_K) value for CeO₂, rGO, NrGO, CeO₂/rGO and CeO₂/NrGO at −0.6 V with corresponding electron transfer number. (e) (Current–time) chronoamperometric responses for ORR on CeO₂/NrGO and commercial Pt/C at −0.35 V at a rotational rate of 1000 rpm. (f) (i–t) chronoamperometric responses of CeO₂/NrGO and Pt/C under methanol addition at 600 s during ORR (potential −0.35 V). The RDE measurements were done at a scan rate of 10 mV s⁻¹ in O₂-saturated 0.1 M KOH solution.

To obtain insight into the electron transfer kinetics during ORR process, we carried out LSV measurements on CeO₂/NrGO at different rotation speeds (Fig. 3b). It is seen that the measured current density gradually increased with increase in rotation rate, which can be explained as; with increasing rotational speed the rate of diffusion of oxygen molecules towards the electrode increases and results in gradual increasing current values. These current values at a constant potential for different rotational rates can be correlated through classical Koutecky–Levich (K–L) equation.¹¹

1/J = 1/J_L + 1/J_K = 1/Bω^1/2 + 1/J_K

where,

B = 0.62nFC₀(D₀)^2/3ν^1/6

where J is the measured current density, J_L and J_K are the diffusion limiting and kinetic current densities respectively, ω is the angular velocity, n is transferred electron number per O₂ molecule, F is Faraday constant (F = 96 485 C mol⁻¹), C₀ the bulk concentration of O₂ (1.2 × 10⁻³ mol L⁻¹), D₀ is the O₂ diffusion coefficient (1.9 × 10⁻⁵ cm² s⁻¹), ν is the kinematic viscosity of the electrolyte (0.01 m² s⁻¹).¹¹

Based on K–L equation, J⁻¹ vs. ω^−1/2 at various electrode potentials were plotted for CeO₂/NrGO (K–L plot, Fig. 3c). The slopes of their best linear fit lines were used to calculate number of electron transferred per O₂ molecule. Good linearity and near parallelism of the plots show that ORR on CeO₂/NrGO follows first order kinetics with respect to the concentration of dissolved O₂.¹¹ It is observed that in case of CeO₂ and/or rGO as catalyst, the electron transfer number (n) is always ∼2 and for NrGO this value varies between 2 and 3 suggesting partial peroxide pathway (Fig. S1†). But interestingly, dramatic shifts in ‘n’ values were observed for CeO₂/rGO and CeO₂/NrGO composite indicates a single step 4-electron O₂ reduction pathway. Furthermore kinetic limiting current density (J_k) values were also calculated from the intercepts of K–L plots (Fig. 3d). CeO₂/NrGO shows highest J_k value of 9.5 mA cm⁻² at −0.6 V, as compared to those of CeO₂ (1.8 mA cm⁻²), rGO (3.5 mA cm⁻²), CeO₂/rGO (4.6 mA cm⁻²) and NrGO (6.2 mA cm⁻²). The facile 4-electron transfer process and highest J_k value of CeO₂/NrGO composite clearly demonstrates its electro catalytic ORR efficiency. These results indicate that effective coupling between CeO₂ and NrGO in CeO₂/NrGO not only improve the ORR catalytic activity but also significantly modify the catalytic pathways. Therefore, in terms of onset potentials, half wave potentials, current density values and electron transfer numbers, it can be generalised that the ORR activity increases as follows CeO₂ ≤ rGO < NrGO ≤ CeO₂/rGO < CeO₂/NrGO. Interestingly, this trend is very similar to that of ORR trends for TMOs/NrGO composites.

In addition, the electro-catalytic stability of CeO₂/NrGO and 20 wt% Pt/C towards ORR were compared through current–time (i–t) chronoamperometric technique at −0.35 V vs. Ag/AgCl. As shown in Fig. 3e after 12 500 s of test, relative current value for Pt/C decreased by 40%, while a decrease of only 19% was observed for CeO₂/NrGO. This result suggests the superior stability of CeO₂/NrGO to that of commercial Pt/C catalyst, which could be ascribed to strong interaction of CeO₂ to NrGO and highly stable active sites in alkaline medium. But during practical test some methanol crosses over the electrolyte membrane and contaminate the cathode compartment which abruptly disturbs the ORR process. So methanol tolerant capability of ORR catalyst is another important factor needs consideration prior to practical application. Subsequently, 10 wt% addition of 3 M methanol at 600 s, resulted in the decrease in relative current by 29% for Pt/C, whereas only 4% decrease in relative current was observed for CeO₂/NrGO at similar conditions (Fig. 3f). All the chronoamperometric data's for methanol tolerant test are recorded after 500 s of ORR to avoid initial rapid fall in current density. These results clearly demonstrate the better ORR stability and methanol tolerance capability of CeO₂/NrGO over commercial Pt/C in alkaline medium.

Conclusions

In summary, CeO₂/NrGO nanocomposite electrocatalyst synthesised via single step microwave solvothermal method displayed very good oxygen reduction activity with nearly four electron transfer pathway in alkaline medium. Furthermore, the CeO₂/NrGO exhibit superior electrochemical stability and methanol tolerance capability to that of commercial Pt/C. Hence, from the above ground of analysis we anticipate that, doping or chemical modification of CeO₂ may further enhance the ORR activity. Therefore, this work opens up a new area of non-precious metal oxide electrocatalyst.

Acknowledgements

This work is funded and supported by BRNS, India (Grant no. 2013/37C/53/BRNS/2152). Miss S. Mishra acknowledges BRNS for fellowship. Mr S. Soren and B. D. Mohapatra acknowledge UGC-RGNF for fellowship.

Notes and references

1. (a) A. S. Arico, S. Srinivasan and V. Antonucci, Fuel Cells, 2001, 1(2), 133–161; (b) X. Zhao, M. Yin, L. Ma, L. Liang, C. Liu, J. Liao, T. Luc and W. Xing, Energy Environ. Sci., 2011, 4, 2736–2753.
2. X. Huang, Z. Zhao, L. Cao, Y. Chen, E. Zhu, Z. Lin, M. Li, A. Yan, A. Zettl, Y. M. Wang, X. Duan, T. Mueller and Y. Huang, Science, 2015, 348(6240), 1230–1234.
3. S. K. Bikkarolla, P. Cumpson, P. Joseph and P. Papakonstantinou, Faraday Discuss., 2014, 173, 415–428.
4. (a) Y. Bing, H. Liu, L. Zhang, D. Ghosh and J. Zhang, Chem. Soc. Rev., 2010, 39, 2184–2202; (b) F. A. Bruijn, V. A. T. Dam and G. J. M. Janssen, Fuel Cells, 2008, 8(1), 3–22.
5. (a) Y. Liang, Y. Li, H. Wang, J. Zhou, J. Wang, T. Reiger and H. Dai, Nat. Mater., 2011, 10, 780–786; (b) J. Duan, Y. Zheng, S. Chen, Y. Tang, M. Jaroniecd and S. Qiao, Chem. Commun., 2013, 49, 7705–7707; (c) Y. Su, H. Jiang, Y. Zhu, X. Yang, J. Shen, W. Zou, J. Chen and C. Li, J. Mater. Chem. A, 2014, 2, 7281–7287; (d) Z. S. Wu, S. Yang, Y. Sun, K. Parvez, X. Feng and K. Mullen, J. Am. Chem. Soc., 2012, 134, 9082–9085; (e) S. Bag, K. Roy, C. S. Gopinath and C. R. Raj, ACS Appl. Mater. Interfaces, 2014, 6, 2692–2699.
6. D. U. Lee, B. J. Kim and Z. Chen, J. Mater. Chem. A, 2013, 1, 4754–4762.
7. (a) K. Fukui, Y. Namai and Y. Iwasawa, Appl. Surf. Sci., 2002, 188, 252–256; (b) R. K. Pandey and R. V. Jayaram, Catal. Sci. Technol., 2012, 2, 1113–1125; (c) M. Das, C. Dhand, G. Sumana, A. K. Srivastava, N. Vijayan, R. Nagrajan and B. D. Molhotra, Appl. Phys. Lett., 2011, 99, 143702–143704; (d) D. Patil, N. Q. Dung, H. Jung, S. Y. Ahn, D. M. Jang and D. Kim, Biosens. Bioelectron., 2012, 31, 176–181; (e) D. Joung, V. Sing, S. Park, A. Schulte, S. Seal and S. I. Khondaker, J. Phys. Chem. C, 2011, 115, 24494–24500; (f) R. S. Kalubarme, Y. H. Kim and C. J. Park, Nanotechnology, 2013, 24, 365401; (g) X. Lin, L. Zhou, T. Huang and A. Yu, Int. J. Electrochem. Sci., 2012, 7, 9550–9559.
8. (a) M. Srivastava, A. K. Das, P. Khanra, M. E. Uddin, N. H. Kim and J. H. Lee, J. Mater. Chem. A, 2013, 1, 9792–9801; (b) R. M. Shereema, S. R. Nambiar, S. S. Shankar and T. P. Rao, Anal. Methods, 2015, 7, 4912–4918.
9. (a) Z. Sofer, O. Jankovsky, P. Simek, K. Klímova, A. Mackova and M. Pumera, ACS Nano, 2014, 8(7), 7106–7114; (b) D. Gao, Z. Zhang, L. Ding, J. Yang and Y. Li, Nano Res., 2015, 8, 546–553; (c) J. Suntivich, H. A. Gasteiger, N. Yabuuchi, H. Nakanishi, J. B. Goodenough and Y. S. Horn, Nat. Chem., 2013, 3, 546–550.
10. F. Tunistra and J. L. Koenig, J. Chem. Phys., 1970, 53(3), 1126–1130.
11. (a) B. Zheng, J. Wang, F. B. Wang and X. H. Xia, J. Mater. Chem. A, 2014, 2, 9079–9084; (b) R. Huo, W. J. Jiang, S. Xu, F. Zhanga and J. S. Hu, Nanoscale, 2014, 6, 203–206.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental and physical characterization procedures, additional LSV curves. See DOI: 10.1039/c6ra13218a

---