English

• The oxygen reduction reaction (ORR) is an important semireaction for fuel cells and Zn–air batteries, but their slow dynamics limit their development [1,2]. Although platinum-group metal catalysts perform excellently [3-5], their high cost and poor stability have prompted the exploration of cost-effective nonprecious-metal alternatives [6]. Nitrogen-coordinated metal or transition metal sites in carbon matrices (M–N–C) are highly efficient ORR electrocatalysts with good properties and are promising alternatives to precious-metal catalysts [7,8].

  Transition metal Fe based catalysts have become the most promising catalysts in the field of electrocatalysis because of their excellent performance. Considering the instability of Fe-based catalyst, it is necessary to further adjust the electronic and geometric structure of M-N-C catalyst, which can be adjusted by the diatomic site strategy of diatomic catalyst (DAC) [9-11]. Catalysts with bimetallic atomic-level regulation are mainly Zn-based Fe, Co, Ni, and Cu species. In combination with theoretical prediction, Li et al. [12] introduced nonmagnetic Zn²⁺ into atomic sites to change the electronic structure of Fe–N–C; this resulted in the diatomic catalyst Fe/Zn–N–C, which transformed from a semiconductor to a semimetal. Consequently, spin-polarized electrons spontaneously accumulated at the Fermi level, which was favorable for trapping and binding with O₂, and the self-sacrifice of Zn–N₄ protected the Fe site during the ORR. These characteristics enabled Fe/Zn–N–C to exhibit significant ORR performance in both acidic and alkaline media. Tong et al. [13] prepared N-coordinated Cu–Zn DACs with excellent activity and stability, which were related to the regulation of Zn of the distribution of Cu orbital electrons. This electronic regulation promoted the stretching and breaking of the O–O bond at the Cu active sites, thus accelerating the rate-limiting step of the ORR. In addition, the bimetallic sacrificial pyrolysis strategy could induce the formation of additional structural defects [14,15], which is conducive to the ORR. Tian et al. [16] used tannic acid to etch a ZnFe–ZIF and conducted high-temperature pyrolysis to vaporize Zn completely and obtain a carbon-vacancy-modified Fe_H–N–C, thus efficiently utilizing the active sites and optimizing the electronic structure. The Fe_H–N–C catalyst exhibited a half-wave potential of 0.91 V and excellent durability. Experiments combined with density functional theory (DFT) calculations showed that the excellent properties of the material were attributed to the optimization of the local atomic configuration via defect engineering, the high utilization of active sites, and the optimization of the electronic structure.

  A bimetallic ZnFe–MOF precursor was prepared by introducing the transition metal Fe into a nitrogen-rich Zn–MOF precursor using a microchannel reactor. A bimetallic ZnFe–NC catalyst was then obtained through one-step pyrolysis. ORR test results show that ZnFe–NC has excellent ORR activity and stability, with a high half-wave potential of 0.902 V, which is superior to that of a Pt/C catalyst. Characterization and DFT calculation results show changes in the electron geometry and rate-limiting step of the ZnFe diatoms and suggest that the ORR activity and stability are improved by the synergistic effect of zinc and iron. The ZnFe–MOF precursor synthesized using the microchannel reactor maintains the truncated rhombohedral shape of Zn–MOF, the Fe atoms are dispersed well, and no iron or agglomeration is observed after carbonization. The green, environmentally friendly process uses water as a solvent and for washing when centrifuging; it does not use any organic solvent.

  Fig. 1a illustrates the overall preparation route of ZnFe–NC. During pyrolysis, ZIF-8 particles acted not only as a self-sacrificing macropore/mesopore template to produce a porous carbon skeleton but also as a zinc source to form a zinc site during decomposition [17]. Figs. 1b and c show SEM images of Zn–MOF and ZnFe–MOF, respectively. The two precursors show a distinct truncated rhombohedral shape, smooth surfaces, and a relatively uniform size, similar to those in a previous study [18]. The XRD patterns in Fig. 1d show that Zn–MOF and ZnFe–MOF have the same diffraction pattern, which means that the introduction of Fe did not affect the Zn–MOF crystal structure. The N₂ adsorption–desorption test (Fig. 1e) suggests that both materials have similar microporous structures. The Fe content of ZnFe–MOF is 1.2 wt%, as determined via ICP-OES. The above results indicate the successful preparation of ZnFe–MOF.

  Figure 1

  

  Figure 1.  (a) ZnFe–NC preparation process. SEM images of (b) Zn–MOF and (c) ZnFe–MOF. (d) XRD patterns and (e) Brunauer–Emmett–Teller (BET) graph of Zn–MOF and ZnFe–MOF (inset: pore volume–width graph).

  DownLoad: Full-Size Img PowerPoint

  The precursors Zn–MOF and ZnFe–MOF were carbonized under the same conditions to obtain the carbon materials Zn–NC and ZnFe–NC. Zn-NC maintained a good carbon skeleton structure (Fig. S1 in Supporting information). In contrast, the ZnFe-NC samples containing a large amount of Fe showed a certain degree of decomposition, In addition HRTEM) shows lattice streaks of amorphous carbon (Figs. S1c and f) [19], suggesting that the carbon structure in the catalyst has a low degree of graphitization and a significantly disordered amorphous structure (Fig. S1). ICP-OES analysis showed a small amount of evaporation of zinc during carbonization at 800 ℃ [20]. The ratio of the Fe content to the Zn content increases from 0.05 to 0.42.

  The XRD patterns of Zn–NC and ZnFe–NC (Fig. 2a) show two major diffraction peaks at approximately 25.6° and 42.5°, which are attributed to the (002) and (101) planes, respectively, of graphitized carbon, indicating the formation of a graphite structure in the prepared material [21]. The Zn–NC and ZnFe–NC catalysts have no correlated XRD peaks based on Fe or Zn crystalline phases, confirming the good dispersion of Zn and Fe atoms throughout the nitrogen-doped carbon material (Fig. S1g). The (002) peak of ZnFe–NC is at a slightly higher angle than that of Zn–NC, indicating better graphitization [22].

  Figure 2

  

  Figure 2.  (a) XRD patterns of Zn–NC and ZnFe–NC. (b) BET chart (inset: pore volume–width graph). (c) Raman spectra. XPS diagrams of Zn–NC and ZnFe–NC: (d) General spectra, (e) Fe 2p, (f) Zn 2p, (g) N 1s, (h) C 1s. (i) Distribution maps of nitrogen content of different species.

  DownLoad: Full-Size Img PowerPoint

  The nitrogen adsorption–desorption isotherms of both samples (Fig. 2b) are type Ⅰ isotherms. However, ZnFe-NC catalysts have richer pore structures, which may be related to the evaporation of Zn, and these rich microporous structures can accommodate a high density of active sites [23,24]. In addition, ZnFe-NC has smaller pore volume and narrower pore size distribution (1.8-70 nm) compared with Zn-NC. The rich microporous structure and narrow pore size distribution are favorable for accommodating a high density of active sites, which enables the electrons to be better transferred to the active sites and participate in the reaction by interband jumps and form more reaction intermediates, thus enhancing the activity of the reaction [25]. Raman spectroscopy was performed to analyze the defects and graphitization degree of the Zn–NC and ZnFe–NC catalysts. Fig. 2c depicts two wide peaks at 1340 and 1585 cm⁻¹, which are related to the D and G bands, respectively, of the defective nitrogen-doped carbon substrate [26]. This indicates the formation of amorphous carbon and a large number of defects at higher temperatures of the catalyst. The I_D/I_G values of Zn–NC and ZnFe–NC are 1.43 and 1.27, respectively, indicating that both have defects to a certain extent. The graphitization degree of ZnFe–NC is slightly higher, which is consistent with the XRD pattern findings. Such a large number of defects greatly benefit ORR catalysis [19,27].

  The XPS spectra of ZnFe–NC show that it has Zn, C, N, and Fe signals (Fig. 2d), confirming the coexistence of Zn and Fe in the carbon carrier. The high-resolution Fe 2p spectrum (Fig. 2e) further shows the absence of iron. The content of Fe²⁺ is very low, so its satellite peaks are merged with the satellite peak of trivalent iron. Fe³⁺ peaks at 710.88 and 723.79 eV, Fe²⁺ peaks at 707.99 and 721.49 eV, and the other two wide peaks at 716.1 and 727.89 eV are satellite peaks of Fe 2p in the spectrum [28]. The Zn 2p spectrum (Fig. 2f) peaks at 1044.7 and 1021.5 eV belong to Zn 2p_1/2 and Zn 2p_3/2, respectively, indicating the presence of abundant Zn–N_x sites. In addition, the Zn 2p spectrum of ZnFe–NC has an overall positive shift of 0.2 eV relative to Zn–NC, which may be attributed to the electronic interaction between Zn and Fe; thus, Zn has a higher oxidation state in ZnFe–NC compared with Zn–NC [29]. Zn 2p is positively shifted (i.e., it loses electrons) and is used to stabilize the Fe–N_x structure.

  The high-resolution N 1s spectrum (Fig. 2g) can be deconvoluted into five types of N species, namely, pyridinic N, metal M (M = Zn, Fe)–N, pyrrolic N, graphitic N, and oxide N, among which pyridinic N promotes the construction of metal–N_x active centers [30]. The presence of M–N peaks indicates that the metal atoms are stabilized by nitrogen, pyridinic N increases the ORR starting potential, and graphitic N is related to the limiting current density (J_L) of the ORR [31]. The M–N content of Zn–NC is 9.3%, which is lower than that of ZnFe–NC (12.2%) and indicates the existence of atomically dispersed Zn–Fe atoms. Fig. 2i summarizes the types and contents of N species, showing that pyridinic N is the main nitrogen structure. Due to the low boiling point of Zn, it partially evaporates during carbonization, and the formation of pores increases the material defect. The introduction of Fe decreases the defect degree relative to Zn–NC, which increases the graphite and nitrogen contents and improves the ORR activity. The C 1s spectrum (Fig. 2h) shows three characteristic peaks at 287.6, 285.7, and 284.7 eV, which are attributed to C–O, C–N, and C–C, respectively, confirming the existence of heteroatom N in the porous carbon matrix.

  An electrochemical test was performed for the ORR to further compare the electrocatalytic activity of Zn–NC and ZnFe–NC (Fig. S2 in Supporting information). The CV curves (Fig. 3a) show that compared with 20 wt% Pt/C, Zn–NC and ZnFe–NC show significant ORR peaks of approximately 0.76 and 0.8 V, respectively. However, the ORR peak of ZnFe–NC has a higher positive potential than that of Zn–NC, indicating that the coexistence of Fe and Zn significantly improves the ORR electrocatalytic activity. In addition, ZnFe–NC has the largest electrochemical active area and more active sites, which is conducive to electrocatalytic reactions.

  Figure 3

  

  Figure 3.  (a) CV curves. (b) LSV curves. (c) H₂O₂ yields and numbers of transferred electrons. (d) Tafel slopes derived from LSV curves. (e) J_K (measured at 0.8 and 0.85 eV) and half-wave potentials E_1/2. (f) Methanol resistance test through injection of 3 mL of methanol after 120 s. (g) Stability test. (h) Performance comparison between ZnFe–NC (this work) and previously reported catalysts.

  DownLoad: Full-Size Img PowerPoint

  The LSV curves (Fig. 3b) show that ZnFe–NC has a high half-wave potential of 0.902 V, which is better than that of the commercially available 20 wt% Pt/C (E_1/2 = 0.887 V) and suggests that ZnFe–NC has the best ORR electrocatalytic activity. These results further demonstrate the importance of maximizing the use of bimetallic ZnFe–N_X active sites and defect engineering. In addition, both the ZnFe–NC and Pt/C electrocatalysts show electron transfer numbers of approximately 4 in the alkaline medium, according the RRDE data (Fig. 3c); this reaffirms the direct four-electron transfer-mediated ORR process, which is consistent with the conclusion that ZnFe–NC and Pt/C have higher limiting current densities than Zn–NC [32]. In the potential range of 0.2–0.8 V, the H₂O₂ yield of ZnFe–NC and Pt/C remain below 2%, indicating that O₂ is effectively reduced to water through the 4e⁻ pathway, and the Tafel slope further confirms the catalytic performance enhanced by ZnFe–NC (Fig. 3d). The ZnFe–NC catalyst has a Tafel slope of (87 mA/dec), which is smaller than that of the commercial Pt/C (114 mA/dec) and indicates the better ORR kinetics of Fe–NC and the high selectivity of O₂ four-electron reduction to H₂O [29].

  According to the K–L equation calculation (Fig. 3e), the maximum kinetic current density (J_K) of ZnFe–NC (26.93 mA/cm²) is 5.4 times that of Pt/C (4.95 mA/cm²) at 0.8 V. At 0.85 V, ZnFe–NC exhibits a kinetic current density of 14.79 mA/cm², which is 4 times that of Pt/C (3.67 mA/cm²), indicating that ZnFe–NC exhibits a fast ORR kinetic process. The kinetic current densities of Zn–NC (2.4 mA/cm² at 0.8 V and 0.97 mA/cm² at 0.85 V) are significantly lower than those of the former, which is consistent with its low activity.

  The Zn–NC and ZnFe–NC catalysts show high tolerance to methanol compared with 20 wt% Pt/C, with ZnFe–NC slightly outperforming Zn–NC (Fig. 3f). After the 35,000 s stability test, ZnFe–NC retains nearly 94% of the initial current density, confirming its excellent electrochemical stability (Fig. 3g). The samples were also not visibly caked after the test (Figs. S3 and S4 in Supporting information). The long-term stability of ZnFe–NC is significantly better than that of Zn–NC. This may be because the introduction of Fe greatly improves ORR reactivity, and sacrificing the breakage of the Zn–N bond in Zn–MOF can stabilize Fe–N_x sites, which are difficult to maintain via traditional methods [14]. Compared with reported catalysts properties, the catalysts prepared by microchannel reactor have certain advantages and comparability (Fig. 3h). In addition, ZnFe-NC-mixing catalyst was prepared by direct mixing (Fig. S5 in Supporting information). The electrochemical test under the same experimental conditions found that both the activity and stability of ORR were far less than ZnFe-NC, which proves the feasibility and superiority of microchannel reaction method.

  The potential source and underlying catalytic mechanism of ZnFe–NC synergism in the ORR were further investigated via DFT calculation of Zn–NC and ZnFe–NC models under alkaline conditions [33,34]. The electronic structures of the Fe and Zn atoms in the single-atom Zn–NC catalyst and the dual-atom ZnFe–NC catalyst were studied. Their differential charge densities (Figs. 4a and b) show that ZnFe–NC shows a more obvious electron redistribution relative to Zn–NC, with a slight electron loss around the Zn atoms and more electron accumulation around the Fe atoms; thus, the coupling between Zn and Fe significantly enhances the interaction [35].

  Figure 4

  

  Figure 4.  Zn–NC and ZnFe–NC models in alkaline solution: (a, b) Differential charge densities (red color indicates electron gain, blue color indicates electron loss). (c) U = 0 V and (d) U = 1.23 V of ORR Gibbs free energy. (e, f) PDOS.

  DownLoad: Full-Size Img PowerPoint

  The intrinsic activity of the catalysts was further evaluated by comparing the free energy of the intermediate steps of the electrochemical reaction. At U = 0 V, both catalysts exhibit spontaneous heat release based on a downhill energy path (Fig. 4c) [36]. At an applied voltage of 1.23 V (ORR equilibrium potential), the dissociation step from ^*OOH to O^* on the Zn–N₄ structure has a maximum free energy difference of 0.91 eV (Fig. 4d), becoming a rate-limiting step for the Zn–NC-catalyzed ORR; this may explain the low ORR activity of Zn–NC. For ZnFe–NC, the overpotential is significantly reduced to 0.55 eV, the ^*OH desorption step is determined as the ORR rate-limiting step. The difference in the rate-limiting steps between ZnFe–NC and Zn–NC and the lowest overpotential of the catalytic ORR indicate a synergistic effect between Zn and Fe atoms that improves the catalytic activity of the ZnFe–NC catalyst.

  Figs. 4e and f show the partial density of states (PDOS) of Zn–NC and ZnFe–NC. After Fe addition, the active metal atoms have the highest density state near the Fermi level, showing high electrical conductivity and catalytic performance [37]. In addition, the degree of overlap of the s, p, and d orbitals increases significantly, which means that the degree of hybridization is better and more conducive to catalytic reactions. The PDOS diagram of the Zn d orbitals before and after Fe doping shows a significant change in electron occupancy. The peak position of the Zn d orbitals in Zn–NC is approximately −7.3 eV, that of ZnFe–NC is far away (approximately −7.7 eV), and the overall Zn d orbitals of ZnFe–NC are slightly away from the Fermi level. This means that the transition of electrons is more difficult, so the interaction between the O^* intermediates and Zn in Zn–NC is very weak, and the reduction of ^*OOH to O^* is a rate-limiting step. Because the PDOS of the Zn atoms in ZnFe–NC differs from that of the Zn atoms in Zn–NC, a direct synergistic interaction of electrons can be determined [38]. The Fe d orbitals clearly split into several peaks, two of which are far from the Fermi level and the others close to it; moreover, the reduction of ^*OH to OH− becomes a rate-limiting step. The Fe d orbitals in ZnFe–NC are very close to the Fermi level, which can reduce ^*OH adsorption and thus better promote the ORR [39]. These changes strongly suggest that electron redistribution and changes in electronic structures occur at the ZnFe–N₆ sites, which are critical for ORR performance [37].

  Further DFT calculations revealed Fe as the main active center (Fig. S6 in Supporting information). Moreover, the introduction of Fe changed the electronic structure and geometry of the active site of the single-atom catalyst. This led to significant interatomic synergies and lowered the reaction energy barriers, thus optimizing the electrocatalytic performance.

  In conclusion, a bimetallic ZnFe–MOF precursor was prepared using a microchannel reactor, and the bimetallic ZnFe–NC was obtained after carbonization. This ZnFe-NC catalyst was specifically characterized by the structural features of high porosity, iron-based active sites and high nitrogen heteroatom content, and exhibited excellent ORR electrochemical activity with a half-wave potential of 0.902 V, and maintained 94% activity after 35,000 s stability test. Various characterizations and DFT calculations showed that the introduction of Fe could synergize with Zn to change the electronic structure and geometry of the catalyst active sites, which effectively reduced the reaction energy barrier. This work provides a new method for preparing bimetallic or even polymetallic MOF materials. It will also guide the development of highly active, stable DACs through the rational modulation of metal active sites, providing insights for optimizing electronic structures to improve their electrocatalytic performance.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Pin Cui: Writing – review & editing, Writing – original draft, Supervision, Methodology, Formal analysis, Data curation. Ying Tang: Writing – review & editing, Visualization, Data curation. Jie Yu: Writing – review & editing, Writing – original draft, Formal analysis. Zhen Yang: Writing – review & editing, Formal analysis. Shouhua Yang: Writing – review & editing, Formal analysis. Boqin Li: Writing – review & editing, Data curation. Gang Wang: Visualization, Resources. Huan Pang: Writing – review & editing, Supervision, Formal analysis. Feng Yu: Supervision, Software, Resources, Project administration, Funding acquisition, Conceptualization.

  Acknowledgments

  This study was financially supported by Xinjiang Science and Technology Program (No. 2023TSYCCX0118) and Bingtuan Science and Technology Program (No. 2023AB033).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110303.

  
  
• 1. [1]

     W. Zhang, S. Yi, Y. Yu, et al., J. Mater. Chem. A 12 (2024) 87–112. doi: 10.1039/D3TA05147A

  2. [2]

     L. Guo, S. Hwang, B. Li, et al., ACS Nano 15 (2021) 6886–6899. doi: 10.1021/acsnano.0c10637

  3. [3]

     J. Long, X. Zheng, B. Wang, et al., Chin. Chem. Lett. 35 (2024) 109354. doi: 10.1016/j.cclet.2023.109354

  4. [4]

     S. Yu, Z. Levell, Z. Jiang, X. Zhao, Y. Liu, J. Am. Chem. Soc. 145 (2023) 25352–25356. doi: 10.1021/jacs.3c09193

  5. [5]

     Y.X. Xiao, H.W. Liu, X.Y. Yang, Front. Chem. Sci. Eng. 17 (2023) 1677–1697. doi: 10.1007/s11705-023-2300-5

  6. [6]

     X. Wang, J. Ding, W. Song, et al., Adv. Energy Mater. 13 (2023) 2300148. doi: 10.1002/aenm.202300148

  7. [7]

     W. Zhang, B. Yang, Z. Xiang, Chin. Chem. Lett. 36 (2025) 109630. doi: 10.1016/j.cclet.2024.109630

  8. [8]

     T. Sun, S. Mitchell, J. Li, et al., Adv. Mater. 33 (2021) 2003075. doi: 10.1002/adma.202003075

  9. [9]

     X. Yang, W. Song, T. Zhang, et al., Adv. Energy Mater. 13 (2023) 202301737.

  10. [10]

      L. Yan, P. Li, Q. Zhu, et al., Chem 9 (2023) 280–342. doi: 10.1016/j.chempr.2023.01.003

  11. [11]

      X. Bai, L. Wang, B. Nan, et al., Nano Res. 15 (2022) 6019–6025. doi: 10.1007/s12274-022-4293-7

  12. [12]

      H. Li, S. Di, P. Niu, et al., Energy Environ. Sci. 15 (2022) 1601–1610. doi: 10.1039/d1ee03194e

  13. [13]

      M. Tong, F. Sun, Y. Xie, et al., Angew. Chem. Int. Ed. 60 (2021) 14005–14012. doi: 10.1002/anie.202102053

  14. [14]

      Z. Luo, T. Zhou, Y. Guan, et al., Small 19 (2023) 2304750. doi: 10.1002/smll.202304750

  15. [15]

      L. Yin, S. Zhang, M. Sun, et al., Adv. Mater. 35 (2023) 2302485. doi: 10.1002/adma.202302485

  16. [16]

      H. Tian, A. Song, P. Zhang, et al., Adv. Mater. 35 (2023) 2210714. doi: 10.1002/adma.202210714

  17. [17]

      H. Yin, P. Yuan, B.A. Lu, et al., ACS Catal. 11 (2021) 12754–12762. doi: 10.1021/acscatal.1c02259

  18. [18]

      C. Gao, S. Mu, R. Yan, et al., Small 18 (2022) 2105409. doi: 10.1002/smll.202105409

  19. [19]

      H. Zhang, S. Hwang, M. Wang, et al., J. Am. Chem. Soc. 139 (2017) 14143–14149. doi: 10.1021/jacs.7b06514

  20. [20]

      S.Y. Lin, L.X. Xia, Y. Cao, et al., Small 18 (2022) 2107141. doi: 10.1002/smll.202107141

  21. [21]

      W.J. Jiang, L. Gu, L. Li, et al., J. Am. Chem. Soc. 138 (2016) 3570–3578. doi: 10.1021/jacs.6b00757

  22. [22]

      L. Yang, G. Xu, J. Ban, et al., J. Colloid Interf. Sci. 535 (2019) 415–424. doi: 10.1016/j.jcis.2018.10.007

  23. [23]

      Y. Yao, T. Jiang, S.Y. Lim, et al., Small 19 (2023) 2304655. doi: 10.1002/smll.202304655

  24. [24]

      E. Proietti, F. Jaouen, M. Lefèvre, et al., Nat. Commun. 2 (2011) 416. doi: 10.1038/ncomms1427

  25. [25]

      Z. Liu, B. Tang, X. Gu, H. Liu, L. Feng, Chem. Engin. J. 395 (2020) 125170. doi: 10.1016/j.cej.2020.125170

  26. [26]

      Q. Wang, Y. Yang, F. Sun, et al., Adv. Energy Mater. 11 (2021) 2100219. doi: 10.1002/aenm.202100219

  27. [27]

      Q.H. Kong, J.T. Ren, H.Y. Wang, et al., Front. Chem. Sci. Eng. 17 (2023) 1755–1764. doi: 10.1007/s11705-023-2318-8

  28. [28]

      L. Jiao, J. Li, L.L. Richard, et al., Nat. Mater. 20 (2021) 1385–1391. doi: 10.1038/s41563-021-01030-2

  29. [29]

      S. Wu, S. Jiang, S.Q. Liu, et al., Energy Environ. Sci. 16 (2023) 3576–3586. doi: 10.1039/d3ee00840a

  30. [30]

      Q. Zhou, M. Min, M. Song, et al., Small 20 (2024) 2307943. doi: 10.1002/smll.202307943

  31. [31]

      X. Liu, M. Sun, Z. Wu, et al., Appl. Catal. A: Gen. 522 (2016) 1–12. doi: 10.1016/j.apcata.2016.04.030

  32. [32]

      J. Feng, R. Cai, E. Magliocca, et al., Adv. Funct. Mater. 31 (2021) 2102974. doi: 10.1002/adfm.202102974

  33. [33]

      C. Wang, H. Yuan, F. Yu, et al., Chin. Chem. Lett. 34 (2023) 107128. doi: 10.1016/j.cclet.2022.01.021

  34. [34]

      W. Bao, K. Lu, P. Fu, et al., Chem. Engin. J. 474 (2023) 145826. doi: 10.1016/j.cej.2023.145826

  35. [35]

      S.S.A. Shah, T. Najam, J. Yang, et al., Chin. J. Catal. 43 (2022) 2193–2201. doi: 10.1016/S1872-2067(22)64089-0

  36. [36]

      G. Chen, R. Lu, C. Li, et al., Adv. Mater. 35 (2023) 2300907. doi: 10.1002/adma.202300907

  37. [37]

      S. Zaman, Y.Q. Su, C.L. Dong, et al., Angew. Chem. Int. Ed. 61 (2022) e202115835. doi: 10.1002/anie.202115835

  38. [38]

      C. Fu, X. Qi, L. Zhao, et al., Appl. Catal. B: Environ. 335 (2023) 122875. doi: 10.1016/j.apcatb.2023.122875

  39. [39]

      J. Xu, S. Lai, D. Qi, et al., Nano Res. 14 (2021) 1374–1381. doi: 10.1007/s12274-020-3186-x

• 

  Figure 1  (a) ZnFe–NC preparation process. SEM images of (b) Zn–MOF and (c) ZnFe–MOF. (d) XRD patterns and (e) Brunauer–Emmett–Teller (BET) graph of Zn–MOF and ZnFe–MOF (inset: pore volume–width graph).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) XRD patterns of Zn–NC and ZnFe–NC. (b) BET chart (inset: pore volume–width graph). (c) Raman spectra. XPS diagrams of Zn–NC and ZnFe–NC: (d) General spectra, (e) Fe 2p, (f) Zn 2p, (g) N 1s, (h) C 1s. (i) Distribution maps of nitrogen content of different species.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) CV curves. (b) LSV curves. (c) H₂O₂ yields and numbers of transferred electrons. (d) Tafel slopes derived from LSV curves. (e) J_K (measured at 0.8 and 0.85 eV) and half-wave potentials E_1/2. (f) Methanol resistance test through injection of 3 mL of methanol after 120 s. (g) Stability test. (h) Performance comparison between ZnFe–NC (this work) and previously reported catalysts.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Zn–NC and ZnFe–NC models in alkaline solution: (a, b) Differential charge densities (red color indicates electron gain, blue color indicates electron loss). (c) U = 0 V and (d) U = 1.23 V of ORR Gibbs free energy. (e, f) PDOS.

  下载: 全尺寸图片 幻灯片
•