Concave Fe₂O₃ nanocubes with high-index facets for ammonia production from electrocatalytic nitrate reduction

Abstract

An iron oxide catalyst with high-index (13–44) and (12–38) facets achieves a high faradaic efficiency of 96.54% and a production rate of 1.13 mmol h⁻¹ cm⁻² towards electrocatalytic nitrate reduction to ammonia at 250 mA cm⁻².

---

Ammonia (NH₃) is an essential chemical feedstock extensively utilized in the industrial production of fertilizers and other chemicals.¹ Additionally, it shows promise as a carbon-free energy carrier, potentially replacing fossil fuels. Currently, the predominant method for industrial ammonia production is the Haber–Bosch process (H₂ + N₂ → NH₃), which operates at high temperatures (300–500 °C) and pressures (15–35 MPa), consuming significant energy.^2,3 An alternative solution is the electrocatalytic nitrogen reduction reaction, which utilizes electricity to drive the conversion of N₂ to NH₃ under ambient conditions.^4,5 Unfortunately, due to the challenge of breaking the extremely stable N≡N bond (941 kJ mol⁻¹), NRR suffers from low ammonia faradaic efficiency and yield, significantly limiting its potential for large-scale industrial application.⁶ In contrast, electrocatalytic nitrate reduction reaction (NtrRR) is thermodynamically more feasible because the breaking of the N–O bond requires a much lower energy input (204 kJ mol⁻¹).^6,7 Meanwhile, soluble nitrate is widely present in industrial and agricultural effluents, causing significant environmental pollution.⁸ Therefore, converting nitrate into high value-added ammonia is crucial for both environmental protection and energy utilization.

However, the kinetics of NtrRR remain slow due to the complexity of the multi-electron coupled multi-proton transfer process. Additionally, the various reaction pathways leading to different products (e.g., NH₃, NO₂, N₂ and NO_x) reduce the selectivity for the desired ammonia products.⁹ As a result, efforts have been focused on developing NtrRR electrocatalysts with high activity and selectivity. Currently, noble metal-based catalysts (such as Ru and Pd) demonstrate superior NtrRR performance,^10–12 but their high cost and limited availability restrict their use in large-scale applications. In this context, non-noble metals such as Cu, Co, and Fe are promising alternatives,^13–15 though their performance needs further improvement.

Since electrocatalytic reactions occur on or near the catalysts’ surface, the surface atomic arrangement (i.e., the crystal facets) plays a crucial role in determining catalytic activity. Specifically, nanocatalysts with high-index facets often exhibit superior catalytic performance compared to their low-index counterparts.^16,17 Indeed, numerous electrocatalysts with well-engineered high-index facets have been developed for both the oxygen evolution reaction (OER)^18–20 and hydrogen evolution reaction (HER).^18,21 However, the exploration of NtrRR electrocatalysts with high-index facets remains rare, with most examples primarily focusing on noble metal catalysts.²²

In this work, we systematically investigated the NtrRR performance of Fe₂O₃ electrocatalysts with different crystal facets. We found that the octagonal Fe₂O₃ with high-index (13–44) and (12–38) facets (H_o-Fe₂O₃) exhibits outstanding NtrRR performance, achieving a high NH₃ faradaic efficiency of up to 96.54% and a production rate of 1.13 mmol h⁻¹ cm⁻² at 250 mA cm⁻², outperforming most previously reported non-noble metal NtrRR catalysts. Furthermore, we assembled a “batterolyzer” by coupling the anodic N₂H₄ oxidation reaction and cathodic NtrRR, following a concept previously developed by our group.²³ This batterolyzer combines the features of a traditional battery and electrolyzer, concurrently generating electricity (with a peak power density of 2.85 mW cm⁻²) and high-value ammonia.

We synthesized H_o-Fe₂O₃ catalysts with high-index (13–44) and (12–38) facets using a recipe we previously developed.²⁴ For comparison, we also synthesized Fe₂O₃ cubes with low-index (10–12) facets (L_c-Fe₂O₃)²⁵ and octahedra with low-index (10–14) facets (L_o-Fe₂O₃)²⁶ using a reported wet chemical method (see details in ESI†). Scanning electron microscopy (SEM) images show the three catalysts have well-defined morphologies (Fig. 1a–c). Transmission electron microscopy (TEM) images reveal that the average particle sizes of H_o-Fe₂O₃, L_c-Fe₂O₃, and L_o-Fe₂O₃ are about 200, 300, and 100 nm, respectively. X-ray diffraction (XRD) analysis suggests that the diffraction peaks of the synthesized Fe₂O₃ with different exposed facets align with those of hematite (JCPDS no. 33-0664). Raman spectra further confirm the successful synthesis of the three Fe₂O₃ catalysts (Fig. 1e). In addition, the Raman peaks of H_o-Fe₂O₃ shift to higher wavenumber, indicating the presence of strain and stress due to the high-index facets.²⁷ Fourier transform infrared spectroscopy (FTIR) spectra also reveal that all the characteristic peaks can be attributed to Fe₂O₃ (Fig. 1f).

Fig. 1 Physical characterizations of the Fe₂O₃ catalysts. SEM, TEM images and structure illustrations of (a) H_o-Fe₂O₃, (b) L_c-Fe₂O₃, and (c) L_o-Fe₂O₃. (d) XRD patterns, (e) Raman spectra and (f) FTIR spectra of the three catalysts.

We investigated the NtrRR performance of the three electrocatalysts. Given that NtrRR generates additional hydroxyl groups, thereby alkalizing the electrolyte, we selected 1 M KOH as the supporting electrolyte. We first assessed the NtrRR activity of H_o-Fe₂O₃ along with L_c-Fe₂O₃ and L_o-Fe₂O₃ controls, using stabilized polarization curves in 1 M KOH electrolyte with and without 0.1 M KNO₃. No significant HER is observed for H_o-Fe₂O₃, L_c-Fe₂O₃ and L_o-Fe₂O₃ (Fig. 2a). Notably, H_o-Fe₂O₃ catalysts require only −0.145 V vs. RHE to initiate NtrRR, which is much lower than that of L_c-Fe₂O₃ and L_o-Fe₂O₃ catalysts (−0.225 and −0.194 V vs. RHE, respectively, Fig. 2b). Tafel plots further reveal the fastest NtrRR kinetics on H_o-Fe₂O₃, indicated by the smallest slope (66.93 mV dec⁻¹, Fig. 2c). To exclude the impact of surface area, we performed cyclic voltammetry (CV) tests within non-faradaic regions and calculated the double layer capacitances (C_dl, Fig. S2, ESI†). The results reveal that the three catalysts have very similar C_dl values, and thus similar electrochemically active surface areas (ECSA). This indicates that the different exposure of crystal facets, rather than the surface area, primarily accounts for the varying NtrRR activities.

Fig. 2 Electrocatalytic NtrRR performance. (a) Polarization curves in 1 M KOH (with IR correction). (b) Polarization curves in 1 M KOH/0.1 M KNO₃ (with IR correction) and (c) the derived Tafel plots of the three catalysts. (d)–(f) Ammonia faradaic efficiencies of (d) H_o-Fe₂O₃, (e) L_c-Fe₂O₃ and (f) L_o-Fe₂O₃ in 1 M KOH/0.1 M KNO₃ mixed electrolyte. (g) EIS spectra of the three catalysts. (h) Chronopotentiometry curves of the H_o-Fe₂O₃ catalyst at 100 mA cm⁻² in 1 M KOH/0.1 M KNO₃ mixed electrolyte.

We then conducted a 7200 s chronopotentiometry test at 250 mA cm⁻² (Fig. S3, ESI†) and subsequently evaluated the ammonia production efficiency using ion chromatography (IC) (see ammonia and nitrate standard curves in Fig. S4, ESI†). The H_o-Fe₂O₃ catalysts exhibit superb faradaic efficiency (96.54%, Fig. S5a, ESI†) and ammonia yield (1.13 mmol h⁻¹ cm⁻², Fig. S5b, ESI†), whereas L_c-Fe₂O₃ and L_o-Fe₂O₃ show lower faradaic efficiencies (75.98% and 72.11%) and ammonia yields (0.89 and 0.84 mmol h⁻¹ cm⁻²). This evidence suggests that H_o-Fe₂O₃ exposing high-index facets not only enhances NO₃⁻ adsorption and activation but also effectively improves ammonia faradaic efficiency and yield rate. Notably, the three catalysts produced negligible nitrite byproducts during the chronopotentiometry tests. Specifically, the nitrite faradaic efficiencies of the H_o-Fe₂O₃, L_c-Fe₂O₃ and L_o-Fe₂O₃ catalysts were only 3.01%, 3.66% and 4.61%, respectively (Fig. S5a, ESI†). Overall, the NtrRR performance of H_o-Fe₂O₃ catalysts significantly outperforms the vast majority of recently reported NtrRR catalysts (Table S1, ESI†).

In addition to the chronopotentiometry tests, we also performed chronoamperometry tests (Fig. S6a–c, ESI†) to further investigate the effect of reaction potential on NtrRR faradaic efficiency and ammonia yield rate (Fig. 2d–f). The ammonia faradaic efficiency of the H_o-Fe₂O₃ catalyst remains consistently above 90% from −0.3 to −0.7 V vs. RHE (Fig. 2d). More importantly, at −0.7 V vs. RHE, the H_o-Fe₂O₃ catalyst achieves a high ammonia yield rate of 1.07 mmol h⁻¹ cm⁻². In contrast, the faradaic efficiencies of the L_o-Fe₂O₃ and L_c-Fe₂O₃ catalysts for ammonia production are below 60%, and their ammonia yield rates are less than 1 mmol h⁻¹ cm⁻² (Fig. 2e and f). These results indicate that the H_o-Fe₂O₃ catalyst maintains the highest intrinsic NtrRR activity for ammonia production among the three catalysts, suggesting that the high-index facets of H_o-Fe₂O₃ optimize the adsorption and reaction of intermediates for NtrRR. In addition, we measured the electrochemical impedance spectra (EIS), which show that both H_o-Fe₂O₃ and L_o-Fe₂O₃ exhibit lower impedance (Fig. 2g and Table S2, ESI†), indicating faster electron transfer, consistent with the Tafel results. The H_o-Fe₂O₃ catalysts could stably produce ammonia at 100 mA cm⁻² for at least 48 h (Fig. 2h). Notably, no hydrogen production was observed during this process, and the decreased potential could be restored by renewing the electrolyte.

All three catalysts exhibit decent stability over long-term tests (Fig. 2h and Fig. S7, ESI†). We further conducted a series of physical characterization studies to examine their structural stability. SEM images reveal no significant changes in bulk morphology, although the surfaces of the catalysts appear smother after NtrRR, suggesting slight surface reconstruction (Fig. S8, ESI†). However, XRD analysis (Fig. 3a and Fig. S9, ESI†) indicates that the phase composition remains unchanged for all three catalysts, explaining their outstanding stability. Additionally, the Fe 2p X-ray photoelectron spectroscopy (XPS) spectra of the catalysts after long-term stability tests are nearly identical to those of pristine samples (Fig. 3b and Fig. S10, ESI†). We further collected the LSV curve of H_o-Fe₂O₃ after stability test, which shows that the activity of H_o-Fe₂O₃ barely changes (Fig. S11, ESI†). These results again demonstrate that the exposure of Fe₂O₃ to different crystal facets is the primary factor affecting its NtrRR performance.

Fig. 3 (a) XRD pattern, and (b) Fe 2p XPS spectrum of H_o-Fe₂O₃ after long-term stability tests. (c) In situ Raman spectra of H_o-Fe₂O₃ at various potentials. (d) Adsorption energy of different crystal facets for NO₃⁻.

To further investigate the NtrRR reaction pathway on H_o-Fe₂O₃, we performed in situ Raman tests. The evolution of Raman peaks indicates the disappearance of the characteristic peaks of NO₃⁻, along with the appearance of the characteristic peaks of NO₂⁻, NH₂, and NH₃ as the applied potential gradually decreases from 0.5 to −1.0 V vs. RHE (Fig. 3c). Thus, the reaction pathway likely follows NO₃⁻ → NO₂⁻ → NH₂ → NH₃, which has been observed in NtrRR on various catalysts.^28–30In situ EIS analysis reveals that H_o-Fe₂O₃ exhibits a smaller phase angle than L_o-Fe₂O₃ and L_c-Fe₂O₃ under the same bias, indicating that more charges participated in the redox reaction (Fig. S12, ESI†). To better understand the role of crystal facets, we conducted theoretical simulations. Previous studies suggest that the adsorption of NO₃⁻ is a crucial step that directly determines the NtrRR performance.^22,31 We calculated the adsorption energies of NO₃⁻ on different crystal facets of Fe₂O₃ (Fig. 3d). The result shows that the (13–44) has the most negative value of −1.488 eV, followed by (10–14) and (10–12). Such favourable adsorption would promote the reaction kinetics, leading to enhanced NtrRR activity of H_o-Fe₂O₃.

We further assembled a batterolyzer by coupling the cathodic NtrRR and anodic hydrazine oxidation reaction (Fig. S13, ESI†).²³ Hydrazine (N₂H₄) is widely distributed in industrial wastewaters such as pharmaceuticals, dyes, and rubber, etc. We used RuO_x as the anode catalyst and H_o-Fe₂O₃ as the cathode catalyst. The batterolyzer shows a maximum discharge power density of up to 2.85 mW cm⁻² at the current density of 18 mA cm⁻² (without IR correction, Fig. 4a). This device can generate power continuously for at least 25 000 s at 10 mA cm⁻² (Fig. 4b). In short, the batterolyzer proposed in this study provide an alternative and promising green approach for practical ammonia production and sustained power output.

Fig. 4 (a) Polarization and power density curves, and (b) long-term chronopotentiometry curve under 10 mA cm⁻² output of a N₂H₄-nitrate batterolyzer.

In conclusion, we demonstrated the crystal facet-dependent NtrRR performance of non-noble metal Fe₂O₃ catalysts. Specifically, we showed that the H_o-Fe₂O₃ catalyst with high-index (13–44) and (12–38) facets exhibits optimal NtrRR activity and selectivity, achieving a high faradaic efficiency of up to 96.54% and an ammonia production rate of up to 1.13 mmol h⁻¹ cm⁻² at 250 mA cm⁻². In addition, the developed catalysts enabled the N₂H₄-nitrate batterolyzer to generate electricity at a high power density of 2.85 mW cm⁻², with sustained nitrate-to-ammonia conversion. Our work not only establishes H_o-Fe₂O₃ as a highly active non-noble metal NtrRR electrocatalyst, but also highlights the important role of high-index facets in improving electrocatalytic performance.

This work was supported by the Natural Science Foundation of Xiamen, China (Grant No. 3502Z202473021) and the Fundamental Research Funds for the Central Universities of China (Grant No.: 20720240066).

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. A. Valera-Medina, H. Xiao, M. Owen-Jones, W. I. F. David and P. J. Bowen, Prog. Energy Combust. Sci., 2018, 69, 63–102.
2. V. Kyriakou, I. Garagounis, A. Vourros, E. Vasileiou and M. Stoukides, Joule, 2020, 4, 142–158.
3. C. Smith, A. K. Hill and L. Torrente-Murciano, Energy Environ. Sci., 2020, 13, 331–344.
4. Y. W. Ren, C. Yu, X. Y. Tan, Q. B. Wei, Z. Wang, L. Ni, L. S. Wang and J. S. Qiu, Energy Environ. Sci., 2022, 15, 2776–2805.
5. G. Soloveichik, Nat. Catal., 2019, 2, 377–380.
6. J. Wang, T. Feng, J. X. Chen, V. Ramalingam, Z. X. Li, D. M. Kabtamu, J. H. He and X. S. Fang, Nano Energy, 2021, 86, 17.
7. X. Fu, Chin. J. Catal., 2023, 53, 8–12.
8. W. D. Chen, X. Y. Yang, Z. D. Chen, Z. J. Ou, J. T. Hu, Y. Xu, Y. L. Li, X. Z. Ren, S. H. Ye, J. S. Qiu, J. H. Liu and Q. L. Zhang, Adv. Funct. Mater., 2023, 33, 23.
9. Z. X. Wang, S. D. Young, B. R. Goldsmith and N. Singh, J. Catal., 2021, 395, 143–154.
10. W. Zhu, F. Yao, Q. Wu, Q. Jiang, J. Wang, Z. Wang and H. Liang, Energy Environ. Sci., 2023, 16, 2483–2493.
11. J. Lim, C.-Y. Liu, J. Park, Y.-H. Liu, T. P. Senftle, S. W. Lee and M. C. Hatzell, ACS Catal., 2021, 11, 7568–7577.
12. Y. Yao, L. Zhao, J. Dai, J. Wang, C. Fang, G. Zhan, Q. Zheng, W. Hou and L. Zhang, Angew. Chem., Int. Ed., 2022, 61, e202208215.
13. Q. Wu, W. Zhu, D. Ma, C. Liang, Z. Wang and H. Liang, Nano Res., 2024, 17, 3902–3910.
14. T. Li, C. Tang, H. Guo, H. Wu, C. Duan, H. Wang, F. Zhang, Y. Cao, G. Yang and Y. Zhou, ACS Appl. Mater. Interfaces, 2022, 14, 49765–49773.
15. Z. Hou, Y. Zhang, H. Chen, J. Wang, A. Li and P. François-Xavier Corvini, Small, 2024, 20, 2406424.
16. S. D. Sun, X. Zhang, J. Cui, Q. Yang and S. H. Liang, Nanoscale, 2019, 11, 15739–15762.
17. H. Liang, W. Chen, R. Wang, Z. Qi, J. Mi and Z. Wang, Chem. Eng. J., 2015, 274, 224–230.
18. C. S. Wang, Q. Zhang, B. Yan, B. You, J. J. Zheng, L. Feng, C. M. Zhang, S. H. Jiang, W. Chen and S. J. He, Nano-Micro Lett., 2023, 15, 41.
19. J. Wang, Y. Q. Zhang, S. J. Jiang, C. Z. Sun and S. Q. Song, Angew. Chem., Int. Ed., 2023, 62, 7.
20. M. Zhao and H. Liang, ACS Nanosci. Au, 2024, 4, 409.
21. S. Mondal, S. Sarkar, D. Bagchi, T. Das, R. Das, A. K. Singh, P. K. Prasanna, C. P. Vinod, S. Chakraborty and S. C. Peter, Adv. Mater., 2022, 34, 10.
22. Y. Zhang and H. Liang, Chem. Synth., 2024, 4, 39.
23. W. Zhu, X. Zhang, F. Yao, R. Huang, Y. Chen, C. Chen, J. Fei, Y. Chen, Z. Wang and H. Liang, Angew. Chem., Int. Ed., 2023, 62, e202300390.
24. H. F. Liang, X. D. Jiang, Z. B. Qi, W. Chen, Z. T. Wu, B. B. Xu, Z. C. Wang, J. X. Mi and Q. B. Li, Nanoscale, 2014, 6, 7199–7203.
25. X. Y. Yu, L. Yu, L. F. Shen, X. H. Song, H. Y. Chen and X. W. Lou, Adv. Funct. Mater., 2014, 24, 7440–7446.
26. P. G. He, Z. P. Ding, X. D. Zhao, J. H. Liu, S. L. Yang, P. Gao and L. Z. Fan, Inorg. Chem., 2019, 58, 12724–12732.
27. R. J. Angel, M. Murri, B. Mihailova and M. Alvaro, Z. Kristallogr. – Cryst. Mater., 2019, 234, 129–140.
28. X. Zou, J. Xie, C. Wang, G. Jiang, K. Tang and C. Chen, Chin. Chem. Lett., 2023, 34, 107908.
29. M. Karamad, T. J. Goncalves, S. Jimenez-Villegas, I. D. Gates and S. Siahrostami, Faraday Discuss., 2023, 243, 502–519.
30. Z. Shu, H. Chen, X. Liu, H. Jia, H. Yan and Y. Cai, Adv. Funct. Mater., 2023, 33, 2301493.
31. X. Zhang, X. Liu, Z.-F. Huang, L. Gan, S. Zhang, R. Jia, M. Ajmal, L. Pan, C. Shi and X. Zhang, Energy Environ. Sci., 2024, 17, 6717–6727.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cc05101g

---