English

• Ammonia (NH₃) is of great significance and widely used in agriculture, industry, and sustainable energy conversion [1, 2]. In industry, Haber–Bosch method is currently used for large-scale ammonia production at high temperature and high pressure over Fe-based or Ru-based catalysts, which requires high energy input and emits a large amount of greenhouse gases [3, 4]. Therefore, it is urgently-needed to develop green and sustainable ways for NH₃ production under mild conditions [5-8]. Among the NH₃ synthesis methods, electrochemical nitrogen reduction reaction (NRR) using water as the hydrogen source offers a promising way to replace the high-energy-consuming and environment-polluting Haber–Bosch method [9].

  However, it is difficult to cleave and dissociate the N≡N bond due to its high total bond energy (941 kJ/mol) [10]. For this reason, the performance of the NRR electrocatalyst has very big upgrade space due to a low yield rate of NH₃ and faradaic efficiency (FE) [11-13]. Multiple materials including pure metals [14, 15], alloys [16], metal compounds [17, 18] and nonmetals [19, 20] have been applied for experimental studies of NH₃ synthesis in both theory and experiment. Single-atom catalysts (SACs), isolated metal atoms anchored to supports, have been applied in many fields such as oxygen reduction reaction (ORR) [21], CO₂ reduction reaction (CO₂RR) [22] and oxidation of formaldehyde [23, 24] due to the 100% atomic utilization, high activity and selectivity, and durable stability [25-27]. In a series of single metal atom catalysts, many metal-nitrogen-carbon (M-N-C, M = Fe, Co, Ni, Mn, Mo, Y, Sc, etc.) have successfully developed for electrocatalysis [28-31]. In particular, inspired by the natural metalloenzyme called cytochrome c oxidase (CcO) with the adjacent Cu and Fe sites [32-34], several dual metal single atom catalysts with adjacent M-N-C dual active centers have been synthesized recently and exhibited excellent stability and catalytic performance [35-40]. For example, a highly dispersed Fe–Cu dual-atom nanozyme has been successfully constructed to mimic Cytochrome c oxidase for catalyzing ORR [37]. The dual-metal catalysts with neighboring Fe-N₄-C and Co-N₄-C active centers are also reported as efficient ORR catalysts [36, 39]. In addition, a dual metal single atom catalyst consisted of Cu-N₄ and Zn-N₄ on the N-doped carbon support was prepared and showed high ORR activity [35].

  As far as we know, there is no report of dual-metal catalysts for the NRR by now, and the in-depth analysis and understanding of structure-property relationship for the neighboring M-N-C catalysts are also insufficient. Inspired by the successful synthesis of adjacent M-N-C catalysts and their potential catalytic activity, we designed a series of non-precious metal-based neighboring M-N-C catalysts (denoted as MN₄/M'N₄-C), and employed the density functional theory (DFT) method to explore the NRR activity. Firstly, the stability of catalysts is evaluated, and NRR selectivity was investigated by considering *N₂/*H adsorption. According to the adsorption configuration of N₂, we then explore possible reaction pathways in the NRR process and screen promising catalysts for the NRR. Finally, the electronic structure analysis is performed to further understand the interaction on the adjacent dual metal single atom to achieve good NRR catalytic performance.

  Seven non-precious metal atoms (Cr, Mn, Fe, Co, Ni, Cu and Zn) were selected to anchor on N-doped graphene in pairs, constructing 28 MN₄/M'N₄-C catalysts in the present study as shown in Fig. 1a, motivated by that the corresponding M-N-C catalysts with N₄-coordinated structure have been synthesized [41, 42] and this may contribute to the development of adjacent M-N-C dual single atom catalysts. In the optimized geometries, each metal atom bonds with four surrounding N atoms on the graphene substrate with bond length of 1.85~1.97 Å (Table S1 in Supporting information). In particular, the Cu-N and Zn-N bond lengths in CuN₄/ZnN₄-C were calculated to be 1.91 and 1.94 Å (Table S1), which are in agreement with the experimental results from the extended X-ray absorption fine structure (EXAFS) measurement for Cu/Zn-NC with the Cu-N and Zn-N bond lengths of 2.01 Å [35]. The dual metal single atom centers of M-N-C and M'-N-C were located in an adjacent position with the distance between two metal atoms ranging from 4.96 Å to 5.10 Å (Table S1), consistent with the corresponding experimental results that the distance between the two metal atoms is around 5.0 Å for the synthesized MN₄/M'N₄-C catalysts such as CuN₄/ZnN₄-C, FeN₄/CuN₄-C and FeN₄/CoN₄-C [35-37]. These show that our computational models for the as-designed catalysts are reliable and reasonable.

  Figure 1

  

  Figure 1.  (a) The structures of adjacent dual metal active sites on MN₄/M'N₄-C and (b) the calculated E_bind values.

  DownLoad: Full-Size Img PowerPoint

  Based on the optimized geometries of the as-designed catalysts, we then evaluate their stabilities by calculating the binding energies (E_bind) using equation: E_bind= E(MN₄/M'N₄-C) – E(NC)– E(M) – E(M'), where E(MN₄/M'N₄-C) and E(NC) represent the electronic energies of catalyst and N-doped graphene substrate, respectively; E(M) and E(M') are the electronic energies of M and M' atoms, respectively. According to this definition, a lower E_bind indicates a higher thermodynamics stability of MN₄/M'N₄-C. As shown in Fig. 1b, one can see that the E_bind values were all negative ranging from –15.72 eV to –4.48 eV, indicating neighboring dual metal single atom can be stably anchored in the N-doped graphene. In addition, they have the similar stabilities as the experimentally synthesized FeN₄/CoN₄-C, FeN₄/CuN₄-C and CuN₄/ZnN₄-C which have the binding energies calculated in this study with the values of –15.33, –12.54 and –7.45 eV, respectively. It may imply that the as-designed catalysts could also be synthesized in future.

  We next investigated the adsorption of N₂ molecule on MN₄/M'N₄-C surfaces. As shown in Fig. 2, one can see that N₂ molecule can adsorb on the surface in perpendicular end-on or parallel side-on configurations. However, two metal atoms locate so far with the distance of around 5 Å so that it is not suitable for the formation of the bridge conformation in which two N atoms in N₂ molecule adsorb on different metal atoms, respectively. Based on the calculated ΔG values of N₂ adsorption listed in Table S2 (Supporting information), we found that the end-on configurations are more energetically favorable than the side-on configurations for both M and M' sites. Therefore, the 11 MN₄/M'N₄-C with negative ΔG values via N₂ end-on adsorption configuration are summarized in Table 1. Although the adsorption ability of N₂ on the 11 MN₄/M'N₄-C was relatively weak (–0.21 eV to –0.02 eV), it was still an exothermic reaction which could proceed spontaneously. In contrast, the remaining catalysts will be excluded from following study because the adsorption and activation of N₂ on these catalysts hardly take place at room temperature.

  Figure 2

  

  Figure 2.  The N₂ adsorbed on M atom by (a) end-on and (b) side-on configurations. The N₂ adsorbed on M' atom by (c) end-on and (d) side-on configurations.

  DownLoad: Full-Size Img PowerPoint

  Table 1

  

  Table 1.  The adsorption configurations of N₂ molecule and the adsorption energies of N₂ molecule, H and N₂H on 11 MN₄/M'N₄-C (The '*' donates the atom adsorbed on the surface) in eV.

  DownLoad: CSV

  As well known, hydrogen evolution reaction (HER) is the key competition reaction toward NRR, thus, the competitive adsorption between H and N₂ on the catalysts was studied by comparing the changes of Gibbs free energies (ΔG(*H) and ΔG(*N₂)). The results listed in Table 1 showed that the ΔG(*N₂) are all negative, whereas the ΔG(*H) values are all positive. It means that N₂ is preferentially adsorbed onto the screened 11 MN₄/M'N₄-C rather than H, preventing the accumulation of H-adatoms and exhibiting good selectivity for electrochemical NRR.

  The first hydrogenation step (*N₂ + H⁺ + e⁻→*N₂H) is usually considered as the potential determining step (PDS) for electrochemical NRR, and high-activity catalysts should have ΔG(*N₂H) values of less than the ΔG_max of the well-established Ru(0001) stepped surface (0.98 eV) [12], which was set as the criterion for metal-based catalysts due to the highest NRR theoretical activity among bulk metal surfaces [43, 44]. As shown in Table 1, 5 CrN₄/M'N₄-C (M' = Cr, Mn, Fe, Cu and Zn) presented the ΔG(*N₂H) values of about 0.60 eV, which suggests that they meet the requirements above and will be further studied and discussed.

  Since N₂ molecule prefers the end-on adsorption configuration on the 5 CrN₄/M'N₄-C, the typical NRR reaction pathways including distal and alternating pathways were considered to reveal the catalytic mechanism (Fig. 3a). In the distal pathway, the first three proton-electron (H⁺ + e⁻) pairs are preferentially added on the distal N atom, leading to the generation of the first NH₃ molecule, and then another three proton-electron (H⁺ + e⁻) pairs attack the remaining N atom to release the second NH₃. In the alternating pathway, six proton-electron (H⁺ + e⁻) pairs alternately hydrogenate two N atoms. The produced NH₃ can be easily protonated to NH₄⁺ and released into solution under the electrochemical conditions [45, 46], so the further protonation of *NH₃ into NH₄⁺ was not considered.

  Figure 3

  

  Figure 3.  (a) Schematic representation of distal and alternating pathways for NRR in end-on configuration. Gibbs free energy diagrams of the NRR at zero potential (blake lines) and an applied potential (red lines) via the (b) distal and (c) alternating pathway on CrN₄/MnN₄-C.

  DownLoad: Full-Size Img PowerPoint

  By taking the CrN₄/MnN₄-C as an example, as shown in Figs. 3b and c, it can be seen that the first two steps of the alternating and distal pathways, namely N₂ adsorption and *N₂H formation, were the same. The ΔG values of the first two steps were –0.05 and 0.62 eV, respectively. After adsorbed N₂ is hydrogenated, the generated *N₂H intermediate can proceed via distal or alternating pathways. In the distal pathway (Fig. 3b), the next hydrogenation to form *N-NH₂ is slightly endothermic with a free energy change of 0.14 eV. One can see that *N-NH₂ can be hydrogenated to release the first NH₃ to form *N with free energy change of −1.02 eV. The free energy changes for the continuous hydrogenations of *N to form *NH, *NH₂ and *NH₃ were calculated to be −0.05, −0.56 and −0.20 eV, respectively. Regarding the alternating pathways on CrN₄/MnN₄-C (Fig. 3c), the second hydrogenation on the other nitrogen to form *NH-NH will proceed after overcoming a positive free energy change of 0.35 eV. In the subsequent steps, the intermediates *NH-NH₂ and *NH₂-NH₂ are formed with the free energy change values of −0.19 and 0.06 eV, respectively. A negative free energy change of −1.71 eV is obtained with the formation of the first NH₃. Finally, the second NH₃ molecule can be formed with a downhill step of −0.20 eV. Therefore, the formation of *N₂H species is the PDS due to the maximum ΔG values (0.62 eV) among all the elementary steps for CrN₄/MnN₄-C. When U = −0.62 V is applied, all elementary reactions become downhill, and the whole electrochemical NRR processes turn to be spontaneous.

  According to the detailed ΔG values of CrN₄/CrN₄-C, CrN₄/FeN₄-C, CrN₄/CuN₄-C and CrN₄/ZnN₄-C in Table S3 (Supporting information), we can find that the free energy change of PDS for these catalysts are 0.63, 0.63, 0.64 and 0.63 eV, respectively. Hereby, the computed U_L for 5 CrN₄/M'N₄-C is from −0.64 V to −0.62 V, which are much lower than that for recently reported ruthenium SAC Ru₁-N₃ (−0.73 V) with the yield rate of 120.9 µg_NH mg_cat.⁻¹ h⁻¹ [47]. The U_L results also indicate the better NRR activity of 5 CrN₄/M'N₄-C by comparing with the single atom catalysts with the same metal atom such as Cr anchored on defective graphene (−0.98 V) [44], single Mn-N₄ sites anchored on porous carbon (−1.04 V) [48], Fe-N₄/graphene (−1.35 V) [49], Cu on N-doped carbon (−1.85 V) [50]. Besides, considering that NRR is conducted in aqueous electrolytes, the solvation effect has been also studied on CrN₄/M'N₄-C (M' = Cr, Mn, Fe, Cu and Zn) along distal pathway. As shown in Fig. S1 (Supporting information), the solvation effect could reduce the limiting potential in the range of −0.49 eV to −0.37 V. Here, CrN₄/M'N₄-C (M' = Cr, Mn, Fe, Cu and Zn) is expected as promising candidates for electrochemical NRR.

  To clarify the origin of the catalytic activity of the 5 CrN₄/M'N₄-Cs (M' = Cr, Mn, Fe, Cu and Zn), we first analyzed corresponding charge density difference (Fig. 4a) and partial density of states (Fig. 4b) of the adsorbed N₂ molecule on CrN₄/MnN₄-C as an example. The charge density difference results confirmed that both positive and negative charges accumulate around the adsorbed N₂ molecule and Cr atom of CrN₄/MnN₄-C, which is beneficial to promote the "acceptance–donation" process [5, 51] that metal atom donates electrons to the antibonding orbitals of N₂ and accepts lone-pair electrons of N₂. In addition, it is noted that a small part of positive charge accumulated on Mn atom, which suggests that Mn atom probably plays a role in electron transfer to Cr atom due to the so-called modulation effect [38] and promotes the N₂ activation on Cr active site. The modulation effect of CrN₄/MnN₄-C, where Cr atom is the only active center, while Mn atom is considered as the modulator, is completely different from the previously reported synergetic effect on double-atom catalysts (DACs) such as (Fe, Co)/CNT [52] and Ni/Fe-N-C [53], in which binary metal is demonstrated as one active center to catalyze reactions. Additionally, the computed partial density of states (PDOS) of N₂ adsorption on CrN₄/MnN₄-C showed that the hybridization mainly occurred between Cr d-orbital and N₂ p-orbital, and there was a small part of orbital overlap between Mn d-orbital and N₂ p-orbital. These results clarified the origin of the activation of N₂ molecule on CrN₄/MnN₄-C. To deeply understand the catalytic mechanism, the charge transfer variations on the basis of Bader charge differences of all reaction intermediates N_xH_y adsorbed on CrN₄/MnN₄-C via the distal pathway are summarized in Fig. 4c. Obviously, the charge was transferred from CrN₄ to the adsorbed N_xH_y species for most of the elementary reaction steps. Even though the N_xH_y intermediates were mainly located at the CrN₄ active site, the neighbored MnN₄ acted as a reservoir to store or release electrons when needed. Consequently, the modulation effect between the dual metal single atoms is effective in electron transfer, leading to the activation of N₂ and then promotes the subsequent hydrogenation step of NRR.

  Figure 4

  

  Figure 4.  (a) Charge density difference maps of N₂ adsorption on CrN₄/MnN₄-C, where the isosurface value is set to be 0.005 e/ų. The positive and negative charges are shown in cyan and yellow, respectively. (b) Partial density of states (PDOS) of N₂ adsorption on CrN₄/MnN₄-C. (c) The Bader charge variation of the CrN₄/MnN₄-C along distal pathway.

  DownLoad: Full-Size Img PowerPoint

  In conclusion, a series of energetically stable adjacent dual metal single atom supported on N-doped graphene (MN₄/M'N₄-C) are designed for electrochemical NRR by means of DFT calculations. Based on the results of N₂ adsorption free energy, and Gibbs free energy change of NRR, it was predicted that 5 CrN₄/M'N₄-C (M' = Cr, Mn, Fe, Cu and Zn) exhibited promising NRR activity with the limiting potential of −0.64 V to −0.62 V. The modulation effect is observed in the adjacent dual metal single atom catalysts based on the analysis of charge distribution and PDOS. This work not only provides new insight for developing neighboring dual single atom catalysts at atomic level, but also highlights the importance of the modulation effect between the multi-active centers.

  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.

  Acknowledgments

  This work was supported by the open project of State Key Laboratory of Advanced Welding and Joining, Harbin Institute of Technology (No. AWJ-19M07) and the National Natural Science Foundation of China (No. U2067216).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2021.08.102.

  
  
• 1. [1]

     S.L. Foster, S.I.P. Bakovic, R.D. Duda, et al., Nat. Catal.1 (2018) 490-500. doi: 10.1038/s41929-018-0092-7

  2. [2]

     M. Kitano, Y. Inoue, Y. Yamazaki, et al., Nat. Chem. 4 (2012) 934-940. doi: 10.1038/nchem.1476

  3. [3]

     A. Banerjee, B.D. Yuhas, E.A. Margulies, et al., J. Am. Chem. Soc. 137 (2015) 2030-2034. doi: 10.1021/ja512491v

  4. [4]

     C.J.M. van der Ham, M.T.M. Koper, D.G.H. Hetterscheid, Chem. Soc. Rev. 43 (2014) 5183-5191. doi: 10.1039/C4CS00085D

  5. [5]

     C. Ling, X. Niu, Q. Li, A. Du, J. Wang, J. Am. Chem. Soc. 140 (2018) 14161-14168. doi: 10.1021/jacs.8b07472

  6. [6]

     Y. Zhao, Y. Zhao, G.I.N. Waterhouse, et al., Adv. Mater. 29 (2017) 1703828. doi: 10.1002/adma.201703828

  7. [7]

     N. Lazouski, M. Chung, K. Williams, M.L. Gala, K. Manthiram, Nat. Catal. 3 (2020) 463-469. doi: 10.1038/s41929-020-0455-8

  8. [8]

     J.S. Anderson, J. Rittle, J.C. Peters, Nature501 (2013) 84-87. doi: 10.1038/nature12435

  9. [9]

     X. Guo, H. Du, F. Qu, J. Li, J. Mater. Chem. A7 (2019) 3531-3543. doi: 10.1039/c8ta11201k

  10. [10]

      W. Song, J. Wang, L. Fu, et al., Chin. Chem. Lett. 32 (2021) 3137–3142. doi: 10.1016/j.cclet.2021.02.043

  11. [11]

      S. Licht, B. Cui, B. Wang, et al., Science345 (2014) 637. doi: 10.1126/science.1254234

  12. [12]

      E. Skúlason, T. Bligaard, S. Gudmundsdóttir, et al., Phys. Chem. Chem. Phys. 14 (2012) 1235-1245. doi: 10.1039/C1CP22271F

  13. [13]

      J.H. Montoya, C. Tsai, A. Vojvodic, J.K. Nørskov, ChemSusChem8 (2015) 2180-2186. doi: 10.1002/cssc.201500322

  14. [14]

      H. Wang, H. Yu, Z. Wang, et al., Small15 (2019) 1804769. doi: 10.1002/smll.201804769

  15. [15]

      G. Rostamikia, S. Maheshwari, M.J. Janik, Catal. Sci. Technol. 9 (2019) 174-181. doi: 10.1039/c8cy01845f

  16. [16]

      R. Manjunatha, A. Schechter, Electrochem. Commun. 90 (2018) 96-100. doi: 10.1016/j.elecom.2018.04.008

  17. [17]

      T. Wu, X. Zhu, Z. Xing, et al., Angew. Chem. Int. Ed. 58 (2019) 18449-18453. doi: 10.1002/anie.201911153

  18. [18]

      J. Sun, W. Kong, Z. Jin, et al., Chin. Chem. Lett. 31 (2020) 953-960. doi: 10.1016/j.cclet.2020.01.035

  19. [19]

      W. Qiu, X.Y. Xie, J. Qiu, et al., Nat. Commun. 9 (2018) 3485. doi: 10.1038/s41467-018-05758-5

  20. [20]

      R. Wang, C. He, W. Chen, C. Zhao, J. Huo, Chin. Chem. Lett. 32 (2021) 3821–3824. doi: 10.1016/j.cclet.2021.05.024

  21. [21]

      J. Li, S. Chen, N. Yang, et al., Angew. Chem. Int. Ed. 58 (2019) 7035-7039. doi: 10.1002/anie.201902109

  22. [22]

      H. Yang, Y. Wu, G. Li, et al., J. Am. Chem. Soc. 141 (2019) 12717-12723. doi: 10.1021/jacs.9b04907

  23. [23]

      J. Zhou, G. Liu, Q. Jiang, et al., Chin. J. Catal. 41 (2020) 1633-1644. doi: 10.1016/S1872-2067(20)63571-9

  24. [24]

      G. Liu, J. Zhou, W. Zhao, Z. Ao, T. An, Chin. Chem. Lett. 31 (2020) 1966-1969. doi: 10.1016/j.cclet.2019.12.023

  25. [25]

      A. Wang, J. Li, T. Zhang, Nat. Rev. Chem. 2 (2018) 65-81. doi: 10.1038/s41570-018-0010-1

  26. [26]

      L. Zhang, Y. Ren, W. Liu, A. Wang, T. Zhang, Natl. Sci. Rev. 5 (2018) 653-672. doi: 10.1093/nsr/nwy077

  27. [27]

      L. Fu, R. Wang, C. Zhao, et al., Chem. Eng. J. 414 (2021) 128857. doi: 10.1016/j.cej.2021.128857

  28. [28]

      J. Liu, X. Kong, L. Zheng, et al., ACS Nano14 (2020) 1093-1101. doi: 10.1021/acsnano.9b08835

  29. [29]

      X. Wang, X. Peng, W. Chen, et al., Nat. Commun. 11 (2020) 653. doi: 10.1007/978-981-15-2914-6_62

  30. [30]

      M. Zhang, Y.G. Wang, W. Chen, et al., J. Am. Chem. Soc. 139 (2017) 10976-10979. doi: 10.1021/jacs.7b05372

  31. [31]

      L. Chen, C. He, R. Wang, et al., Chin. Chem. Lett. 32 (2021) 53-56. doi: 10.1016/j.cclet.2020.11.013

  32. [32]

      M.R.A. Blomberg, P.E.M. Siegbahn, G.T. Babcock, M. Wikström, J. Am. Chem. Soc. 122 (2000) 12848-12858. doi: 10.1021/ja002745a

  33. [33]

      T. Tsukihara, H. Aoyama, E. Yamashita, et al., Science272 (1996) 1136. doi: 10.1126/science.272.5265.1136

  34. [34]

      M. Wikström, K. Krab, V. Sharma, Chem. Rev. 118 (2018) 2469-2490. doi: 10.1021/acs.chemrev.7b00664

  35. [35]

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

  36. [36]

      H. Li, Y. Wen, M. Jiang, et al., Adv. Funct. Mater. 31 (2021) 2011289. doi: 10.1002/adfm.202011289

  37. [37]

      C. Du, Y. Gao, H. Chen, et al., J. Mater. Chem. A8 (2020) 16994-17001. doi: 10.1039/d0ta06485h

  38. [38]

      D. Wang, A. Han, X. Wang, et al., Angew. Chem. Int. Ed. 60 (2021)19262-19271. doi: 10.1002/anie.202105186

  39. [39]

      D. Wang, H. Xu, P. Yang, et al., J. Mater. Chem. A9 (2021) 13678-13687. doi: 10.1039/d1ta03176g

  40. [40]

      Y. Luo, J. Zhang, J. Chen, et al., J. Catal. 397 (2021) 223-232. doi: 10.1016/j.jcat.2021.03.030

  41. [41]

      D. Wang, X. Pan, P. Yang, et al., ChemSusChem14 (2021) 33-55. doi: 10.1002/cssc.202002137

  42. [42]

      J. Zhang, H. Yang, B. Liu, Adv. Eng. Mater. 11 (2021) 2002473. doi: 10.1002/aenm.202002473

  43. [43]

      X. Guo, J. Gu, S. Lin, et al., J. Am. Chem. Soc. 142 (2020) 5709-5721. doi: 10.1021/jacs.9b13349

  44. [44]

      C. Choi, S. Back, N.Y. Kim, et al., ACS Catal. 8 (2018) 7517-7525. doi: 10.1021/acscatal.8b00905

  45. [45]

      A. Clayborne, H.J. Chun, R.B. Rankin, J. Greeley, Angew. Chem. Int. Ed. 54 (2015) 8255-8258. doi: 10.1002/anie.201502104

  46. [46]

      H.J. Chun, V. Apaja, A. Clayborne, K. Honkala, J. Greeley, ACS Catal. 7 (2017) 3869-3882. doi: 10.1021/acscatal.7b00547

  47. [47]

      Z. Geng, Y. Liu, X. Kong, et al., Adv. Mater. 30 (2018) 1803498. doi: 10.1002/adma.201803498

  48. [48]

      L. Han, M. Hou, P. Ou, et al., ACS Catal. 11 (2021) 509-516. doi: 10.1021/acscatal.0c04102

  49. [49]

      D. Jiao, Y. Liu, Q. Cai, J. Zhao, J. Mater. Chem. A9 (2021) 1240-1251. doi: 10.1039/d0ta09496j

  50. [50]

      L. Han, Z. Ren, P. Ou, et al., Angew. Chem. Int. Ed. 60 (2021) 345-350. doi: 10.1002/anie.202010159

  51. [51]

      M.A. Légaré, G. Bélanger-Chabot, R.D. Dewhurst, et al., Science359 (2018) 896. doi: 10.1126/science.aaq1684

  52. [52]

      J. Wang, W. Liu, G. Luo, et al., Energy Environ. Sci. 11 (2018) 3375-3379. doi: 10.1039/c8ee02656d

  53. [53]

      W. Ren, X. Tan, W. Yang, et al., Angew. Chem. Int. Ed. 58 (2019) 6972-6976. doi: 10.1002/anie.201901575

• 

  Figure 1  (a) The structures of adjacent dual metal active sites on MN₄/M'N₄-C and (b) the calculated E_bind values.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  The N₂ adsorbed on M atom by (a) end-on and (b) side-on configurations. The N₂ adsorbed on M' atom by (c) end-on and (d) side-on configurations.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Schematic representation of distal and alternating pathways for NRR in end-on configuration. Gibbs free energy diagrams of the NRR at zero potential (blake lines) and an applied potential (red lines) via the (b) distal and (c) alternating pathway on CrN₄/MnN₄-C.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Charge density difference maps of N₂ adsorption on CrN₄/MnN₄-C, where the isosurface value is set to be 0.005 e/ų. The positive and negative charges are shown in cyan and yellow, respectively. (b) Partial density of states (PDOS) of N₂ adsorption on CrN₄/MnN₄-C. (c) The Bader charge variation of the CrN₄/MnN₄-C along distal pathway.

  下载: 全尺寸图片 幻灯片

  Table 1.  The adsorption configurations of N₂ molecule and the adsorption energies of N₂ molecule, H and N₂H on 11 MN₄/M'N₄-C (The '*' donates the atom adsorbed on the surface) in eV.

  下载: 导出CSV
•