Germanium-doped and germanium/nitrogen-codoped carbon nanotubes with highly enhanced activity for oxygen reduction in alkaline medium

Abstract

To develop effective cathode electrocatalysts for the oxygen reduction reaction (ORR) in fuel cells, germanium (Ge)-doped and Ge/N-co-doped carbon nanotubes (CNTs) were synthesized by chemical vapor deposition in this work. Electrochemical tests demonstrated that the as-prepared Ge-doped and GeN-codoped CNTs exhibited obviously enhanced ORR activity in alkaline medium, showing that the doping of Ge into the carbon matrix could also improve the ORR activities of the CNTs and N-CNTs as in the cases of other reported heteroatoms and would be of great importance for designing more effective ORR electrocatalysts in future alkaline fuel cells.

---

1. Introduction

Developing highly active and durable non-platinum electrode electrocatalysts to replace the traditional noble metal platinum (Pt) catalysts is a great challenge in the field of fuel cells. Since the doping of nitrogen (N), phosphorus (P) or boron (B)^1–3 into the carbon matrix was verified to bring much exciting improved activity towards the oxygen reduction reaction (ORR), new doping atoms such as sulfur (S),^4,5 selenium,⁶ halogens,^7,8 silicon (Si)^9–11 and arsenic¹² were reported to show similar improvement in the ORR activity of carbon materials successively. Meanwhile, the dual or multiple doping of foreign atoms such as Si/N, N/S, B/P, N/B/P, N/S/P was proved to bring much more striking improvement^9,13–16 when compared with unitary doping. Currently, the chemical doping of foreign atoms into the framework of carbon materials has become an effective strategy to develop more effective platinum-free ORR electrocatalysts.

Germanium (Ge), as one of IVA elements, its atomic radius (122 pm) and negativity (2.01) are different from those (77 pm, 2.55) of carbon. The doping of Ge into the carbon matrix would change the electronic and physical structures of carbon materials and cause the ORR activity improvement as the case of Si in the same IVA.⁹ To develop more effective heteroatom-doped carbon electrocatalysts, in this work, we selected tetraethyl germanium (TEG) as Ge source and tried to synthesize Ge-doped and Ge/N-codoped carbon nanotubes (Ge-CNTs, Ge/N-CNTs) and estimated their ORR electrocatalytic performance in alkaline medium. Interestingly, the results demonstrated that the as-prepared Ge-CNTs and Ge/N-CNTs showed much obviously improved ORR activity in alkaline medium like other reported heteroatom-doped carbon nanomaterials. It is worth noting that, Ge-doped carbon materials have been rarely investigated, except for the Ge, P or Ge, N-codoped carbon reported just recently,^17,18 showing a greatly promising potential for designing more effective platinum-free electrocatalysts in future alkaline fuel cells.

2. Experimental section

2.1 Preparation of materials

Ge-CNTs were synthesized via the pyrolysis of the xylene solution of TEG with FeMo/Al₂O₃ as catalyst. In a typical experiment, 0.1 g of FeMo/Al₂O₃ in a quartz boat was put in the quartz tube and 150 mL min⁻¹ of Ar was introduced to remove the air. When the furnace temperature was raised to 500 °C, the H₂ was inputted to reduce the catalyst at a flow rate of 25 mL min⁻¹. The solutions with 2.0, 4.0 and 8.0 wt% of TEG were injected at the speed of 2.0 mL h⁻¹ after the temperature arrived to 800 °C, respectively. Then, the tubular furnace was cooled down to room temperature with an argon flow rate of 50 mL min⁻¹ when the solution was conveyed into the quartz tube entirely. The resultant samples were collected and denoted as Ge-CNTs1, Ge-CNTs2 and Ge-CNTs3. Meanwhile, GeN-CNTs1, GeN-CNTs2 and GeN-CNTs3 were synthesized with 2.0, 4.0 and 8.0 wt% of TEG in aniline solutions under the same conditions, respectively. For comparison, the pure CNTs and N-CNTs were also synthetized using xylene and aniline as C and N precursors.

The commercially available Pt/C (47.6 wt% on Vulcan XC-72) catalyst were purchased from BASF Fuel Cell, Inc, USA. Other chemicals were purchased and used without any further purification.

2.2 Electrode preparation and electrochemical experiments

The pretreatment procedures of glassy carbon electrodes (GCE) (5.0 mm in diameter) were as follows: prior to use, the electrodes were polished mechanically with aluminite powder on an abrasive paper to obtain a mirror-like surface, washed with ethanol and de-ionized water by sonication for 5 min and dried in a desiccator. 1.5 mg of each grinded sample was dispersed in 0.5 mL of solvent mixture of Nafion (5%), de-ionized water and acetone (V : V : V = 15 : 385 : 100) by sonication. 10.0 μL suspension was dropped onto the GCE surface. And the electrode was dried at room temperature for 2 h in a desiccator before the electrochemical measurements. Consequently, ca. 0.153 mg cm⁻² of each prepared example or commercial Pt/C catalyst was loaded onto the surface of bare GCE, respectively.

Electrochemical experiments were carried out at room temperature in a three-electrode cell connected to an electrochemical analyzer (Pine Research Instrumentation, USA). Every prepared sample/GCE and Pt/C/GCE were used as the working electrodes, an Ag/AgCl with saturated KCl as reference electrode, and a Pt electrode or a graphite carbon as counter electrode. All potentials were measured and reported vs. the potential of Ag/AgCl electrode. The cyclic voltammetry (CV) experiments were conducted in a nitrogen-protected or an oxygen-saturated 0.1 M KOH solution in the potential range from +0.2 to −1.0 V at room temperature at the scan rate of 100 mV s⁻¹. The linear sweep voltammetry (LSV) measurements were performed in the oxygen-saturated 0.1 M KOH solution in the potential range of +0.2 to −1.0 V at the scan rate of 10 mV s⁻¹. In addition, it is noted that, to remove the influence of Fe in FeMo/Al₂O₃ catalyst on the ORR activity, all prepared samples were purified by concentrated hydrochloric acid for 2 h before electrochemical tests. The Koutecky–Levich plots were obtained by I⁻¹ = I_k⁻¹ + (0.62nFCD^2/3v^−1/6ω^1/2)⁻¹, where I_k⁻¹ is the kinetic current density, ω is the rotational speed, n is the number of electron transferred, F is the Faraday constant (F = 96 485C mol⁻¹), C is the bulk concentration of O₂ (C = 1.2 × 10⁻³ mol L⁻¹ in 0.1 M KOH), D is the diffusion coefficient of O₂ (D = 1.9 × 10⁻⁵ cm² s⁻¹ in 0.1 M KOH), v is the kinetic viscosity of the electrolyte (0.01 cm² s⁻¹ in 0.1 M KOH), ω is the angular velocity of the disk (ω = 2πN, N is the linear rotation speed).

2.3 Characterizations

The morphologies of the samples and elemental compositions were characterized by scanning electron microscopy (SEM, ZEISS Merlin) and transmission electron microscopy (TEM, JEOL 2100F) and energy dispersive spectrometer (EDS, ZEISS Merlin). The surface areas of the samples were measured by the Brunauer–Emmett–Teller (BET, Autosorb IQ). X-ray photoelectron spectroscopic (XPS) measurements were performed on a Thermo Scientific ESCALAB 250XI using Al Kα radiation, and the C 1s peak at 284.8 eV was taken as internal standard.

3. Results and discussion

SEM and TEM images illustrated that the samples synthesized by the xylene solution of TEG consisted of carbon nanotubes (Fig. 1A–F). The samples prepared by the aniline solution of TEG and aniline were mainly composed of bamboo-like carbon nanotubes (Fig. 1G–N). The main elemental compositions were analyzed by EDS and inserted in their TEM images (Fig. 1). BET data indicated that the surface areas (52.1, 51.8 and 49.8 m² g⁻¹) of Ge-CNTs1, Ge-CNTs2 and Ge-CNTs3 all were a bit larger than that (47.1 m² g⁻¹) of the pure CNTs. And the surface areas (98.5, 92.1 and 88.7 m² g⁻¹) of GeN-CNTs1, GeN-CNTs2 and GeN-CNTs3 all were larger than that (87.3 m² g⁻¹) of N-CNTs as well. The reason for the increase of the BET areas of Ge- or GeN-CNTs could be ascribed to the changes of tube size, bond angle and bond length (Fig. 1A–N and 4C and D).

Fig. 1 SEM and TEM images of the Ge-CNTs1 (A and B), Ge-CNTs2 (C and D), Ge-CNTs3 (E and F), GeN-CNTs1 (G and H), GeN-CNTs2 (I and J), GeN-CNTs3 (K and L) and N-CNTs (M and N).

The successful doping of Ge and N was further confirmed by XPS measurements of Ge-CNTs2, GeN-CNTs2 and N-CNTs. In Fig. 2A, the high C 1s peak at 284.8 eV and O 1s peak at 531.9 eV were observed in the XPS spectra. Ge 3d peak at 31.4 eV was assigned to C–Ge bond.¹⁹ And N 1s peaks at ca. 398, 399.3 and 400.7 eV (Fig. 2B) were assigned to pyridinic N, pyrrolic N and graphitic N,²⁰ respectively. Graphitic N and pyrrolic N (64.3 and 3.4 at%) decreased and pyridinic N (32.3 at%) increased in GeN-CNTs2 when compared with those (67.1, 4.0 and 28.9 at%) in N-CNTs.

Fig. 2 (A) XPS surveys of Ge-CNTs2, GeN-CNTs2 and N-CNTs. The inset shows the Ge 3d spectra of Ge-CNTs2 and GeN-CNTs2. (B) N 1s spectra of GeN-CNTs2 and N-CNTs.

To investigate the electrocatalytic activity of Ge-CNTs and GeN-CNTs for the ORR, the CV measurements were performed in an aqueous solution of N₂-protected or O₂ saturated 0.1 M KOH with a flow rate of 25 mL min⁻¹ compared with the pure CNTs and N-CNTs. Fig. 3A showed that the ORR peak potentials (−0.43, −0.46 and −0.44 V) of Ge-CNTs1, Ge-CNTs2 and Ge-CNTs3 all were similar to that (−0.45 V) of the pure CNTs. However, the peak current densities (from 1.45 to 2.06 mA cm⁻²) of three Ge-CNTs were 1.86–2.64 times that (0.78 mA cm⁻²) of the pure CNTs, indicating the additional Ge doping could obviously improve the ORR activity of CNTs in alkaline medium under the conditions with the similar surface areas of three Ge-CNTs and the pure CNTs. Similarly, for three GeN-CNTs, Fig. 3B illustrated that the peak potentials (−0.43 and −0.42 V) of GeN-CNTs1 and GeN-CNTs3 were also close to that (−0.41 V) of N-CNTs except that (−0.38 V) of GeN-CNTs2. Whereas, the peak current densities (from 2.12 to 3.0 mA cm⁻²) of three GeN-CNTs all were much larger than that (1.60 mA cm⁻²) of N-CNTs, demonstrating that the Ge doping could effectively improve the ORR activity of N-CNTs in alkaline medium as well. Ge-CNTs2 and GeN-CNTs2 both displayed the highest ORR activity among the corresponding Ge-CNTs and GeN-CNTs samples, respectively, showing that the ORR activities of Ge-containing CNTs were influenced not only by Ge doping contents but also by their surface areas.

Fig. 3 Typical CVs of CNTs and three Ge-CNTs (A), N-CNTs and three GeN-CNTs (B). The LSVs of CNTs and three Ge-CNTs (C), N-CNTs and three GeN-CNTs (D) at a rotation speed of 1600 rpm.

To better understand the ORR activities of Ge-CNTs and GeN-CNTs, the LSV measurements were also carried out in an oxygen-saturated 0.1 M KOH solution with a flow rate of 25 mL min⁻¹ at a rotation speed of 1600 rpm. As shown in Fig. 3C, the ORR onset potentials (at −0.24, −0.22 and −0.23 V) of Ge-CNTs1, Ge-CNTs2 and Ge-CNTs3 all were more positive than that (−0.30 V) of the pure CNTs. And their diffusion current densities at the potential range of −0.30 to −1.0 V all were much larger than those of CNTs as well, confirming the improvement of ORR activity of CNTs after the Ge doping. For three GeN-CNTs, Fig. 3D clearly showed that the onset potentials (at −0.11, −0.12 and −0.13 V) of GeN-CNTs1, GeN-CNTs2 and GeN-CNTs3 all were much higher than that (−0.17 V) of N-CNTs. And their diffusion current densities at the potential range of −0.17 to −1.0 V all were also obviously larger than those of N-CNTs at the same potential range. Meanwhile, the diffusion current densities of GeN-CNTs2 even exceeded those of commercial Pt/C catalyst (47.6 wt%) below −0.76 V although its onset was still lower than that of Pt/C, further demonstrating the Ge doping could improve the ORR activity of N-CNTs. From XPS analyses, N-CNTs with higher proportion of graphitic N should exhibit higher ORR activity than GeN-CNTs2 with lower graphitic N proportion because graphitic N is believed to be favorable to the ORR.⁹ However, the ORR activity of GeN-CNTs2 was much more superior to that of N-CNTs. So, the considerable improvement of ORR activity of GeN-CNTs could be mainly ascribed to the doping of Ge into the framework of N-CNTs. Similar to the CV test results, Ge-CNTs2 and GeN-CNTs2 also exhibited highest ORR activity in the LSV measurements of Ge-CNTs and GeN-CNTs samples.

To gain an in-depth understanding of the electrochemical process of Ge-CNTs and GeN-CNTs, the LSV tests were further conducted in oxygen-saturated 0.1 M KOH solution with an oxygen flow rate of 25 mL min⁻¹ at different rotation speeds. The transferred electron number (n) per oxygen molecule during the ORR course of each sample was calculated on the basis of the slopes in Koutecky–Levich plots (I⁻¹ vs. ω^−1/2) at −0.8 V (Fig. 4A). As shown in Fig. 4B, the n value of each Ge-CNTs was much higher than that (2.11) of the pure CNTs. And the n values of three GeN-CNTs were also higher than that (3.0) of N-CNTs. These results meant that the Ge doing could bring the increase of the transferred electron number of the CNTs and N-CNTs. Theoretical calculations by using Gaussian 09 electronic structure program and UB3LYP/6-31G (d,p) level of theory with all atoms fully relaxed^21,22 indicated that the net charges of neighboring carbon atoms changed from neutral to negative (Fig. 4C and D), and the energy gaps between the highest occupied molecular orbits (HOMO) and the lowest unoccupied molecular orbits (LUMO) drastically decreased from 2.75 to 2.0 or 1.58 eV after the doping of Ge atom into the network of CNT and N-CNT. Meanwhile, the bond length increased from ca. 1.44 to 1.97 Å and bond angle decreased from ca. 116.6 to 90.1° with the doping of Ge into carbon matrix. These electronic and structural changes would lead to the change of adsorption style of oxygen and the increase of n value.^2,10,13

Fig. 4 (A) Koutecky–Levich plots of J⁻¹ versus ω^−1/2 at −0.8 V and (B) the n values of the pure CNTs (a), Ge-CNTs1 (b), Ge-CNTs2 (c), Ge-CNTs3 (d), N-CNTs (e), GeN-CNTs1 (f), GeN-CNTs2 (g), GeN-CNTs3 (h) and Pt/C(I). The net charge distributions of C, Ge and N atoms in Ge-CNT (C) and GeN-CNT (D).

To investigate the potential practical application of Ge-CNTs2 and Ge-CNTs2 with relatively higher ORR activities, the stability and methanol tolerance tests were performed by CVs for 5000 cycles and chronoamperometry at a constant voltage of −0.30 V for 2000 s together with those of the commercial Pt/C catalyst, respectively. Fig. 5A showed that the peak current densities of Ge-CNTs2 and GeN-CNTs2 changed little after 5000 cycles, showing long-term stabilities. In contrast, the Pt/C catalyst underwent a loss of about 20% the peak current density after the same cycles.² Meanwhile, Ge-CNTs2 and Ge-CNTs2 also exhibited remarkably excellent methanol tolerance. As can be seen from Fig. 5B, the current densities of Ge-CNTs2 and Ge-CNTs2 almost remained unchanged after the addition of 3 M methanol into the oxygen-saturated 0.1 M KOH solution. However, the current density of the Pt/C catalyst decreased drastically. These results indicated the outstanding stability and excellent methanol tolerance of Ge-CNTs2 and GeN-CNTs2, showing a great promise for the practical application in the future alkaline methanol fuel cells.

Fig. 5 (A) The CVs of Ge-CNTs2 and GeN-CNTs2 in the O₂-saturated 0.1 M KOH solution before and after 5000^th cycles. (B) I–t chronoamperometric responses of Ge-CNTs2, GeN-CNTs2 and Pt/C upon the addition of 3 M methanol into the 0.1 M KOH solution at −0.3 V. The arrow indicates the addition of methanol.

4. Conclusions

In summary, we have successfully synthesized a series of Ge-doped and GeN-codoped CNTs with FeMo/Al₂O₃ as catalyst in this work. Electrochemical tests demonstrated the Ge doping could improve the ORR activities of the CNTs and N-CNTs in alkaline medium obviously. Meanwhile, Ge-CNTs2 and GeN-CNTs2 showed relatively high ORR activity in alkaline medium, indicating that ORR activities of the CNTs and N-CNTs did not increase proportionally with the increase of Ge doping contents and showing a great potential application for the design of much more efficient cathode catalysts in future alkaline fuel cells.

Acknowledgements

We acknowledge the financial support from the National Natural Science Foundation of China (No. 21376257), the Jiangsu Provincial Natural Science Foundation of China (No. BK20131112), and National Supercomputing Center in Shenzhen for providing the computational resources and Gaussian 09 suite of programs (Revision D.01).

Notes and references

1. R. L. Liu, D. Q. Wu, X. L. Feng and K. Mullen, Angew. Chem., Int. Ed., 2010, 49, 2565–2569.
2. Z. W. Liu, F. Peng, H. J. Wang, H. Yu, W. X. Zheng and J. Yang, Angew. Chem., Int. Ed., 2011, 50, 3257–3261.
3. L. J. Yang, S. J. Jiang, Y. Zhao, S. Chen, X. Z. Wang, Q. Wu, J. Ma, Y. W. Ma and Z. Hu, Angew. Chem., 2011, 50, 7270–7273.
4. Z. Yang, Z. Yao, G. F. Li, G. Y. Fang, H. G. Nie, Z. Liu, X. M. Zhou, X. Chen and S. M. Huang, ACS Nano, 2012, 6, 205–211.
5. S. Inamdar, H. S. Choi, P. Wang, M. Y. Song and J. S. Yu, Electrochem. Commun., 2013, 30, 9–12.
6. Z. P. Jin, H. G. Nie, Z. Yang, J. Zhang, Z. Liu and X. J. Xu, Nanoscale, 2012, 4, 6455–6460.
7. G. Panomsuwan, N. Saitobcd and T. Ishizaki, J. Mater. Chem. A, 2015, 3, 9972–9981.
8. I. Y. Jeon, H. J. Choi, M. Choi, J. M. Seo, S. M. Jung, M. J. Kim, S. Zhang, L. P. Zhang, Z. H. Xia, L. M. Dai, N. J. Park and J. B. Beak, Sci. Rep., 2013, 3, 1810–1817.
9. Z. Liu, X. Fu, M. Li, F. Wang, Q. Wang, G. Kang and F. Peng, J. Mater. Chem. A, 2015, 3, 3289–3293.
10. X. Fu, Q. D. Wang, Z. W. Liu and F. Peng, Mater. Lett., 2015, 158, 32–35.
11. S. K. Ramasahayam, U. B. Nasini, V. G. Bairi, A. U. Shaikh and T. Viswanathan, RSC Adv., 2014, 4, 6306–6313.
12. Z. W. Liu, M. Li, F. Wang and Q. D. Wang, J. Power Sources, 2016, 306, 535–540.
13. Z. W. Liu, X. Fu, X. Y. Wei and F. Peng, J. Mater. Chem. A, 2015, 3, 22723–22729.
14. G. Jo, J. Sanetuntikul and S. Shanmugam, RSC Adv., 2015, 5, 53637–53643.
15. C. H. Choi, S. P. Park and S. I. Woo, ACS Nano, 2012, 6, 7084–7091.
16. Z. W. Liu, F. Wang, M. Li and Z. H. Ni, RSC Adv., 2016, 6, 37500–37505.
17. Q. Q. Li, F. Yuan, C. Y. Yan, J. J. Zhu, J. Sun, Y. J. Wang, J. Ren and X. L. She, RSC Adv., 2016, 6, 33205–33211.
18. X. L. She, Q. Q. Li, N. Ma, J. Sun, D. W. Jing, C. M. Chen, L. J. Yang and D. J. Yang, ACS Appl. Mater. Interfaces, 2016, 8, 10383–10391.
19. J. F. Moulder, W. F. Stickle, P. E. Sobel and K. D. Bomben, Handbook of X-ray photoelectron spectroscopy-a reference book of standard spectra for identification and interpretation of XPS data, Perkin-Elmer Corporation, Eden Prairie, MN, USA, 1992.
20. G. Y. Zhong, H. J. Wang, H. Yu and F. Peng, J. Power Sources, 2015, 286, 495–503.
21. M. J. Frisch, G. W. Trucks and H. B. Schlegel, et al., GAUSSIAN 09, Revision A. 01, Gaussian, Inc., Wallingford CT, 2004.
22. F. Maillard, E. R. Savinova, P. A. Simonov, V. I. Zaikovskii and U. Stimming, J. Phys. Chem. B, 2004, 108, 17893–17904.

---