Is iron nitride or carbide highly active for oxygen reduction reaction in acidic medium?

Abstract

Currently, iron nitride (Fe₂N) and iron carbide (Fe₃C), which are regarded as the new non-precious metal electrocatalysts for the oxygen reduction reaction (ORR) with high activity and stability in acidic medium, are attracting increasing attention. Herein, by systematic comparison of the ORR activities for Fe₂N- and Fe₃C-based catalysts designed with or without a solid nitrogen source, we found that only the former is highly active for ORR while the latter is quite poorly active despite their similar crystalline phases. This result indicates that the Fe–N related species are responsible for the high ORR activities of the Fe-based catalysts, similar to the case of Fe/N/C catalysts. Density functional theory calculations demonstrate that the Fe–N₄/C moiety has a far superior ORR activity to that of Fe₂N and Fe₃C. The experimental and theoretical results mutually support that the high activities of the Fe-based catalysts originate from Fe–N_x/C moieties (x ≥ 4) rather than Fe₂N or Fe₃C phases, which is significant for exploring advanced Fe-based electrocatalysts.

---

Proton exchange membrane fuel cells (PEMFCs) are prospective power sources for mobile and stationary applications due to their large power density, high efficiency and low carbon emissions.¹ To date, the commercialization of PEMFCs is still hindered by the high cost and limited operational stability of Pt-based catalysts. Now, the main challenge is the design of cheap and stable electrocatalysts for the oxygen reduction reaction (ORR).^1,2

Great effort has been devoted to exploring non-precious metal electrocatalysts working in acidic medium, especially the three types, namely metal/nitrogen/carbon (M/N/C),³ nitrides⁴ and chalcogenides.⁵ Among them, Fe/N/C catalysts exhibit the closest activity to the Pt-based catalysts and have long-term stability, thus becoming the most promising candidate for replacing noble-metal catalysts.^3,6 Many experimental results suggest that Fe–N_x/C moieties (x ≥ 4) are the ORR active sites,⁷ although their structures have not been completely identified to date.⁸

Recently, it was reported that iron carbide (Fe₃C) was a new type of ORR electrocatalyst with a high onset potential of 0.92 V (vs. RHE) and good stability in acidic medium,⁹ which has attracted increasing attention.¹⁰ In our study on this topic, we noticed that nitrogen-containing sources such as cyanamide and pyrrole were introduced in the preparation of Fe₃C catalysts in all the related literature, although Fe–N_x/C moieties (x ≥ 4) were not detected.^9,10 It is known that high-temperature pyrolysis of the coexisting iron salt, the nitrogen-containing source, and the carbon support can produce Fe/N/C catalysts with abundant Fe–N_x/C moieties.^6,11 And it is reported that the Fe(III) species of high oxidation state in Fe–N_x/C moieties is active for ORR at high potential (0.7 V vs. RHE).^11b This situation raises a critical question, i.e., does the ORR activity of the so-constructed Fe₃C catalyst originate from the Fe₃C phase or the Fe–N_x/C moiety? Very recently, iron nitride (Fe₂N) was also reported to be ORR active in acidic medium.¹² We noted that a carbon source was used in the preparation of the Fe₂N catalyst. Hence, the same critical question also exists, i.e., does the ORR activity stem from the Fe₂N phase or the Fe–N_x/C moiety? The clarification of such questions is scientifically interesting and significant, especially for exploring the Fe-based non-precious metal ORR electrocatalysts. Herein we intentionally designed Fe₂N- and Fe₃C-based catalysts with or without a nitrogen source during preparation, and four typical Fe-based catalysts were obtained. By detailed characterization and systematic comparison, it is demonstrated that the high ORR activities of this kind of new catalysts originate from the trace of Fe–N_x/C moieties rather than the Fe₂N or Fe₃C phases, which is supported by our theoretical simulations.

Carbon nanocages (CNCs) and nitrogen-doped CNC (NCNC, N content: ∼11.4 at%), with comparable specific surface area (∼1500 m² g⁻¹) and similar structure, were used as pure and N-doped carbon supports, respectively, which were prepared by an in situ MgO template method as described in our recent papers (ESI-1†).¹³

For the Fe₂N-based catalysts, Fe₂N nanoparticles were loaded on NCNCs and CNCs by impregnation and nitridation, similar to our recent synthesis of alloyed Co–Mo nitride ORR catalyst, designated as Fe₂N/NCNC and Fe₂N/CNC, respectively. The same Fe loading of 10 wt% was used (experimental section).^4aFig. 1 displays the transmission electron microscopy (TEM) and X-ray diffraction (XRD) characterization of the catalysts. Both catalysts possess the pure Fe₂N phase with particle sizes of 20–100 nm (Fig. 1a, b and e). The Fe₂N nanoparticles can be removed by leaching with 0.5 mol L⁻¹ H₂SO₄, denoted as R-Fe₂N/NCNC and R-Fe₂N/CNC (Fig. 1c, d and f, ESI-2†).

Fig. 1 TEM images of (a) Fe₂N/NCNC, (b) Fe₂N/CNC, (c) R-Fe₂N/NCNC, and (d) R-Fe₂N/CNC. (e and f) XRD patterns of the indicated samples. Insets in (a) and (b) are the corresponding high resolution TEM images.

Fig. 2 shows the cyclic voltammograms (CVs) and rotating disk electrode (RDE) voltammetry of the two catalysts as-prepared and after the removal of Fe₂N nanoparticles. In comparison with the featureless cases in the N₂-saturated electrolyte, obvious ORR signals appear on the CVs measured in the O₂-saturated H₂SO₄ solution, indicating their apparent ORR activities (Fig. 2a). The ORR peak potential of Fe₂N/NCNC is 0.704 V (vs. RHE), much higher than that of Fe₂N/CNC (0.464 V). RDE curves also reflect the distinct ORR activities with a much higher onset potential (E_on) for Fe₂N/NCNC (0.879 V) than for Fe₂N/CNC (0.729 V) and a dominant 4e⁻ process for both catalysts (Fig. 2b, ESI-3†). These results present the much superior ORR activity of Fe₂N/NCNC to that of Fe₂N/CNC, indicating the different active sites due to the different supports, i.e., N-containing NCNC versus N-free CNC. This difference is confirmed by the ORR activities of the two catalysts after removing Fe₂N nanoparticles (Fig. 2b). A much deteriorated ORR activity is observed for the R-Fe₂N/CNC catalyst with a large decrease in E_on and current density at 0.2 V, i.e., ∼166 mV and ∼1.21 mA cm⁻², respectively (ESI-3†). This activity is very close to that of the NCNC or NH₃-treated CNC (marked as CNC-NH₃, experimental section), indicating the complete removal of Fe₂N nanoparticles by acid etching, in agreement with the TEM and XRD analysis (Fig. 1). In contrast, only a slight decrease in E_on (∼40 mV) and current density at 0.2 V (∼0.28 mA cm⁻²) is observed for the R-Fe₂N/NCNC, implying the existence of other active species different from Fe₂N for Fe₂N/NCNC. Energy dispersive X-ray spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS) analyses shown in Fig. 3 do reveal that, after removing the Fe₂N species, the Fe species is absent in Fe₂N/CNC while still present in Fe₂N/NCNC. These results suggest that the ORR-active Fe-N_x/C moieties exist in Fe₂N/NCNC, which are formed by heat-treating the iron salt and N-containing NCNC support during preparation of the catalyst.^6e Actually, even a very trace amount of Fe–N_x/C could lead to a remarkable ORR activity.¹⁴ In addition, we found that Fe–N_x/C moieties are indeed difficult to form by the same process with the pure CNC support (ESI-4†). Hence it is deduced that the high ORR activity of Fe₂N/NCNC originates from the Fe–N_x/C moieties rather than the Fe₂N particles, and actually the Fe₂N phase presents poor ORR activity as demonstrated by Fe₂N/CNC and is unstable in acidic medium (Fig. 1 and 2).

Fig. 2 ORR performance of the Fe₂N/NCNC and Fe₂N/CNC catalysts as-prepared and after removing Fe₂N nanoparticles. (a) CV curves in N₂- and O₂-saturated 0.5 mol L⁻¹ H₂SO₄ solution. (b) RDE curves. In (b), the curves for CNC, NCNC and CNC-NH₃ are presented for reference.

Fig. 3 (a and b) EDS and (c and d) XPS spectra of R-Fe₂N/NCNC and R-Fe₂N/CNC, respectively. Note: in (c), the 2p_3/2 peak of Fe species is located at ∼711 eV, corresponding to the ORR-active Fe(III).^6e

For the Fe₃C-based catalysts, the highly ORR-active species should also be reconsidered similar to the case for the Fe₂N-based catalysts. Two Fe₃C catalysts were prepared by heating a mixture of the iron source (FeCl₂·4H₂O), carbon source (polymethylmethacrylate, PMMA), and CNC support with or without a solid nitrogen source (cyanamide, CN₂H₂) at 800 °C in Ar, denoted as Fe₃C–N/CNC and Fe₃C/CNC respectively. Two control catalysts, N/CNC and PMMA/CNC, were also prepared by heating a mixture of CNC and PMMA with or without CN₂H₂, respectively, without an iron source (experimental section).

Fig. 4 shows the ORR performance of the catalysts. Fe₃C–N/CNC (with a N source in the preparation) presents the highest ORR activity with E_on of 0.920 V, much better than Fe₃C/CNC (without a N source in the preparation) with E_on of 0.481 V (Fig. 4a). This catalyst presents a dominant 4e⁻ process with an electron transfer number (n) of 3.90 ± 0.05 and a corresponding H₂O₂ yield below 8% in the range of 0.08–0.80 V, obviously superior to Fe₃C/CNC (Fig. 4b). The ORR performance of Fe₃C–N/CNC is located at the top level among the reported Fe₃C-based catalysts,^9,10 and at the normal level among the Fe/N/C catalysts.^11,15 XPS analysis reveals that N species exist in Fe₃C–N/CNC while not in Fe₃C/CNC. Hence, the high ORR activity of the Fe₃C–N/CNC catalyst comes from the Fe-based species with N participation, i.e., the Fe–N_x/C moieties. Even if the Fe₃C nanoparticles were removed by acid leaching, the resultant catalyst still shows a high E_on of 0.876 V, also much better than the Fe₃C/CNC catalyst without N participation (ESI-5†). The comparison results here indicate that the excellent ORR activity of the Fe₃C–N/CNC catalyst with a N source in preparation originates from the Fe–N_x/C moieties, and the Fe₃C itself has poor ORR activity.

Fig. 4 ORR performance of the Fe₃C–N/CNC and Fe₃C/CNC catalysts. (a) RDE curves. 1, Fe₃C–N/CNC; 2, Fe₃C/CNC. (b) Electron transfer number (n) and H₂O₂ yield at different potentials. In (a), the RDE curves for CNC, PMMA/CNC and N/CNC are presented for reference.

To get a deep insight into the different activities of Fe–N_x/C, Fe₂N and Fe₃C species, the density functional theory (DFT) method was employed to calculate their free energy diagrams of ORR.¹⁶ Fe–N_x/C was modeled by an Fe–N₄ structure in a carbon matrix, denoted as Fe–N₄/C.¹⁷ Fe₂N (010) and Fe₃C (001) slabs were chosen for Fe₂N and Fe₃C (ESI-6†). The free energy diagrams of the ORR process, including the successive steps of O₂(g) protonation to OOH*, O–O bond scission of OOH* to O*, O* protonation to OH*, and OH* removal to form H₂O, were calculated for all models, as plotted in Fig. 5. The dashed line is the theoretically ideal free energy diagram of ORR, with an equal free energy drop of −1.23 eV for each step.¹⁸ As can be seen, the deviations of the free energy diagrams from the ideal one become more and more serious in the order Fe–N₄/C < Fe₂N < Fe₃C. Fe–N₄/C possesses the closest profile to the ideal one, indicating the best ORR activity among the three models, while Fe₂N holds the second and Fe₃C the last. For Fe₃C, such a deviation is so huge that the step of OH* to H₂O shows a prohibitively high free energy increase of ∼0.785 eV, indicating inert ORR activity. We noted that the step of O* protonation to OH* has the maximum free energy change (ΔG₃) either for Fe–N₄/C or for Fe₂N, which is likely the rate-limiting step around the onset potential (ESI-7†).¹⁶ Accordingly, the thermodynamic ORR overpotentials for Fe–N₄/C and Fe₂N are calculated to be 0.65 and 1.01 V, respectively.^16,19 The much lower overpotential of Fe–N₄/C suggests the obvious superior activity of Fe–N₄/C to Fe₂N. Thus it is deduced that the relative order of ORR activities should be Fe–N₄/C ≫ Fe₂N ≫ Fe₃C. Our experimental results fully support this result (ESI-8†).

Fig. 5 Free energy diagrams of the ORR process for Fe–N₄/C, Fe₂N and Fe₃C. Insets show the corresponding theoretical models. C, grey; N, blue; Fe, purple.

The preceding experimental and theoretical results demonstrate that the high ORR activities of the Fe₂N- and Fe₃C-based catalysts, which have been intensively studied recently, come from the Fe–N_x/C moieties actually. The Fe₃C and Fe₂N phases are not the good ORR active sites and are unstable or dissolvable in acidic medium.

In summary, we have designed four typical Fe₂N- and Fe₃C-based catalysts with or without a solid nitrogen source during the preparation. By structural characterization and systematic comparison of their electrochemical performance, we demonstrate that the high ORR activities of these new catalysts actually originate from the Fe–N_x/C moieties rather than the Fe₂N or Fe₃C phases themselves. DFT calculations suggest the relative order of ORR activity, i.e., Fe–N₄/C ≫ Fe₂N ≫ Fe₃C, which is in high agreement with our experimental results. The understanding of the ORR active sites of this kind of eye-catching catalyst is significant for exploring advanced Fe-based electrocatalysts.

Acknowledgements

This work was jointly supported by the National Natural Science Foundation of China (No. 51232003, 21373108, 51571110, 21473089, and 21573107), the National Basic Research Program of China (No. 2013CB932902), and Suzhou Science and Technology Plan Projects (No. ZXG2013025).

Notes and references

1. R. Borup, J. Meyers, B. Pivovar, Y. S. Kim, R. Mukundan, N. Garland, D. Myers, M. Wilson, F. Garzon, D. Wood, P. Zelenay, K. More, K. Stroh, T. Zawodzinski, J. Boncella, J. E. McGrath, M. Inaba, K. Miyatake, M. Hori, K. Ota, Z. Ogumi, S. Miyata, A. Nishikata, Z. Siroma, Y. Uchimoto, K. Yasuda, K. Kimijima and N. Iwashita, Chem. Rev., 2007, 107, 3904–3951.
2. (a) H. A. Gasteiger and N. M. Markovic, Science, 2009, 324, 48–49; (b) M. K. Debe, Nature, 2012, 486, 43–51.
3. (a) F. Jaouen, E. Proietti, M. Lefèvre, R. Chenitz, J. P. Dodelet, G. Wu, H. T. Chung, C. M. Johnston and P. Zelenay, Energy Environ. Sci., 2011, 4, 114–130; (b) G. Wu, K. L. More, C. M. Johnston and P. Zelenay, Science, 2011, 332, 443–447; (c) M. Lefèvre, E. Proietti, F. Jaouen and J. P. Dodelet, Science, 2009, 324, 71–74.
4. (a) T. Sun, Q. Wu, R. C. Che, Y. F. Bu, Y. F. Jiang, Y. Li, L. J. Yang, X. Z. Wang and Z. Hu, ACS Catal., 2015, 5, 1857–1862; (b) J. Qi, L. H. Jiang, Q. Jiang, S. L. Wang and G. Q. Sun, J. Phys. Chem. C, 2010, 114, 18159–18166.
5. (a) M. R. Gao, J. Jiang and S. H. Yu, Small, 2012, 8, 13–27; (b) H. L. Wang, Y. Y. Liang, Y. G. Li and H. J. Dai, Angew. Chem., Int. Ed., 2011, 50, 10969–10972.
6. (a) J. Masa, W. Xia, M. Muhler and W. Schuhmann, Angew. Chem., Int. Ed., 2015, 54, 10102–10120; (b) G. Wu and P. Zelenay, Acc. Chem. Res., 2013, 46, 1878–1889; (c) Z. W. Chen, D. Higgins, A. P. Yu, L. Zhang and J. J. Zhang, Energy Environ. Sci., 2011, 4, 3167–3192; (d) A. Morozan, B. Jousselme and S. Palacin, Energy Environ. Sci., 2011, 4, 1238–1254; (e) T. Sun, Q. Wu, O. Zhuo, Y. F. Jiang, Y. F. Bu, L. J. Yang, X. Z. Wang and Z. Hu, Nanoscale, 2016, 8, 8480–8485.
7. (a) U. I. Kramm, I. Herrmann-Geppert, J. Behrends, K. Lips, S. Fiechter and P. Bogdanoff, J. Am. Chem. Soc., 2016, 138, 635–640; (b) U. I. Kramm, M. Lefèvre, N. Larouche, D. Schmeisser and J. P. Dodelet, J. Am. Chem. Soc., 2014, 136, 978–985; (c) N. Ramaswamy, U. Tylus, Q. Jia and S. Mukerjee, J. Am. Chem. Soc., 2013, 135, 15443–15449; (d) U. Tylus, Q. Y. Jia, K. Strickland, N. Ramaswamy, A. Serov, P. Atanassov and S. Mukerjee, J. Phys. Chem. C, 2014, 118, 8999–9008; (e) H. R. Byon, J. Suntivich, E. J. Crumlin and Y. S. Horn, Phys. Chem. Chem. Phys., 2011, 13, 21437–21445.
8. (a) Z. Andrea, V. Goellner, V. Armel, M. T. Sougrati, T. Mineva, L. Stievano, E. Fonda and F. Jaouen, Nat. Mater., 2015, 14, 937–942; (b) Y. S. Zhu, B. S. Zhang, X. Liu, D. W. Wang and D. S. Su, Angew. Chem., Int. Ed., 2014, 53, 10673–10677; (c) Q. Y. Jia, N. Ramaswamy, H. Hafiz, U. Tylus, K. Strickland, G. Wu, B. Barbiellini, A. Bansil, E. F. Holby, P. Zelenay and S. Mukerjee, ACS Nano, 2015, 9, 12496–12505.
9. (a) Y. Hu, J. O. Jensen, W. Zhang, L. N. Cleemann, W. Xing, N. J. Bjerrum and Q. F. Li, Angew. Chem., Int. Ed., 2014, 53, 3675–3679; (b) M. L. Xiao, J. B. Zhu, L. G. Feng, C. P. Liu and W. Xing, Adv. Mater., 2015, 27, 2521–2527.
10. (a) W. X. Yang, X. J. Liu, X. Y. Yue, J. B. Jia and S. J. Guo, J. Am. Chem. Soc., 2015, 137, 1436–1439; (b) J. Wei, Y. Liang, Y. X. Hu, B. Kong, G. P. Simon, J. Zhang, S. P. Jiang and H. T. Wang, Angew. Chem., Int. Ed., 2016, 55, 1355–1359; (c) Z. Y. Wu, X. X. Xu, B. C. Hu, H. W. Liang, Y. Lin, L. F. Chen and S. H. Yu, Angew. Chem., Int. Ed., 2015, 54, 8179–8183; (d) G. Y. Ren, X. Y. Lu, Y. N. Li, Y. Zhu, L. M. Dai and L. Jiang, ACS Appl. Mater. Interfaces, 2016, 8, 4118–4125; (e) W. X. Yang, X. Y. Yue, X. J. Liu, L. L. Chen, J. B. Jia and S. J. Guo, Nanoscale, 2016, 8, 959–964; (f) S. Marzorati, J. M. Vasconcelos, J. Ding, M. Longhi and P. E. Colavita, J. Mater. Chem. A, 2015, 3, 18920–18927; (g) G. Y. Zhong, H. J. Wang, H. Yu and F. Peng, J. Power Sources, 2015, 286, 495–503; (h) X. B. Wang, P. Zhang, W. Wang, X. Lei and H. Yang, Chem. – Eur. J., 2016, 22, 4863–4869.
11. (a) C. Z. Zhu, H. Li, S. F. Fu, D. Du and Y. H. Lin, Chem. Soc. Rev., 2016, 45, 517–531; (b) Q. Wang, Z. Y. Zhou, Y. J. Lai, Y. You, J. G. Liu, X. L. Wu, E. Terefe, C. Chen, L. Song, M. Rauf, N. Tian and S. G. Sun, J. Am. Chem. Soc., 2014, 136, 10882–10885; (c) S. Yasuda, A. Furuya, Y. Uchibori, J. Kim and K. Murakoshi, Adv. Funct. Mater., 2016, 26, 738–744; (d) A. G. Kong, X. F. Zhu, Z. Han, Y. Y. Yu, Y. B. Zhang, B. Dong and Y. K. Shan, ACS Catal., 2014, 4, 1793–1800; (e) H. R. Byon, J. Suntivich and Y. S. Horn, Chem. Mater., 2011, 23, 3421–3428; (f) A. Muthukrishnan, Y. Nabae, H. Hayakawa, T. Okajima and T. Ohsaka, Catal. Sci. Technol., 2015, 5, 475–483; (g) K. Artyushkova, A. Serov, S. Rojas-Carbonell and P. Atanassov, J. Phys. Chem. C, 2015, 119, 25917–25928; (h) J. Tian, L. Birry, F. Jaouen and J. P. Dodelet, Electrochim. Acta, 2011, 56, 3276–3285.
12. L. Liu, X. F. Yang, N. Ma, H. T. Liu, Y. Z. Xia, C. M. Chen, D. J. Yang and X. D. Yao, Small, 2016, 12, 1295–1301.
13. (a) S. Chen, J. Y. Bi, Y. Zhao, L. J. Yang, C. Zhang, Y. W. Ma, Q. Wu, X. Z. Wang and Z. Hu, Adv. Mater., 2012, 24, 5593–5597; (b) K. Xie, X. T. Qin, X. Z. Wang, Y. N. Wang, H. S. Tao, Q. Wu, L. J. Yang and Z. Hu, Adv. Mater., 2012, 24, 347–352; (c) J. Zhao, H. W. Lai, Z. Y. Lyu, Y. F. Jiang, K. Xie, X. Z. Wang, Q. Wu, L. J. Yang, Z. Jin, Y. W. Ma, J. Liu and Z. Hu, Adv. Mater., 2015, 27, 3541–3545; (d) Z. Y. Lyu, D. Xu, L. J. Yang, R. C. Che, R. Feng, J. Zhao, Y. Li, Q. Wu, X. Z. Wang and Z. Hu, Nano Energy, 2015, 12, 657–665.
14. Y. G. Li, W. Zhou, H. L. Wang, L. M. Xie, Y. Y. Liang, F. Wei, J. C. Idrobo, S. J. Pennycook and H. J. Dai, Nat. Nanotechnol., 2012, 7, 394–400.
15. (a) M. Zhou, C. Z. Yang and K. Y. Chan, Adv. Energy Mater., 2014, 4, 1400840; (b) A. Serov, K. Artyushkova and P. Atanassov, Adv. Energy Mater., 2014, 4, 1301735; (c) D. Zhao, L. J. Shui, C. Chen, X. Q. Chen, B. M. Reprogle, D. P. Wang and D. J. Liu, Chem. Sci., 2012, 3, 3200–3205.
16. (a) J. K. Nørskov, J. Rossmeisl, A. Logadottir, L. Lindqvist, J. R. Kitchin, T. Bligaard and H. Jonsson, J. Phys. Chem. B, 2004, 108, 17886–17892; (b) Y. F. Jiang, L. J. Yang, T. Sun, J. Zhao, Z. Y. Lyu, O. Zhuo, X. Z. Wang, Q. Wu, J. Ma and Z. Hu, ACS Catal., 2015, 5, 6707–6712.
17. (a) S. Kattel and G. Wang, J. Phys. Chem. Lett., 2014, 5, 452–456; (b) S. Kattel, P. Atanassov and B. Kiefer, Phys. Chem. Chem. Phys., 2014, 16, 13800–13806.
18. F. Studt, Catal. Lett., 2013, 143, 58–60.
19. (a) J. D. Baran, H. Grönbeck and A. Hellman, J. Am. Chem. Soc., 2014, 136, 1320–1326; (b) Y. Wang and H. P. Cheng, J. Phys. Chem. C, 2013, 117, 2106–2112.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental section, additional experimental characterization, and theoretical calculations. See DOI: 10.1039/c6cy01921h

‡ T. Sun and Y. F. Jiang contributed equally to this work.

---