Open Access  |  Research Article
J Mater Inf 2023;3:24. 10.20517/jmi.2023.32 © The Author(s) 2023.

Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe3@NxC20-x (x = 0~4): a DFT exploration

Views: 179 |  Downloads: 28 |  Cited:  0

School of Physical Science and Technology, Inner Mongolia University, Hohhot 010021, Inner Mongolia, China.

*Correspondence to: Prof. Fengyu Li, School of Physical Science and Technology, Inner Mongolia University, No. 235 Daxue West Road, Saihan District, Hohhot 010021, Inner Mongolia, China. E-mail:

This article belongs to the Special Issue Theory-Assisted Developing Electrocatalysis
© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License (, which permits unrestricted use, sharing, adaptation, distribution and reproduction in any medium or format, for any purpose, even commercially, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.


The Haber-Bosch (H-B) process, which is widely used in industry to synthesize ammonia, leads to serious energy and environment-related issues. The electrochemical nitrogen reduction reaction (eNRR) is the most promising candidate to replace H-B processes because it is more energy-efficient and environmentally friendly. Atomic-level catalysts, such as single-atom and double-atom catalysts (SACs and DACs), are of great interest due to their high atomic utilization and activity. The synergy between the metal atoms and two-dimensional (2D) support not only modulates the local electronic structure of the catalyst but also controls the catalytic performance. In this article, we explored the eNRR performance of 2D Fe3@NxC20-x (x = 0~4), whose structure was based on the experimentally synthesized Ag3@C20 sheet, by means of density functional theory calculations. Through calculations, we found that the 2D Fe3@N4C16 with Fe2 site coordinated with four N is a promising eNRR catalyst: the limiting potential is as low as -0.45 V, and the competing hydrogen evolution reaction can be effectively suppressed. Our work not only confirms that the coordination environment of the metal site is crucial for the electrocatalytic activity but also provides a new guideline for designing low-cost eNRR catalysts with high efficiency.


Electrochemical nitrogen reduction reaction, two dimensions, single-atom and double-atom catalysts, active site, coordination, density functional theory


Ammonia is a raw material vital to the global economy because of its wide range to be converted into fertilizers, chemicals, future fuel substitutes, and hydrogen storage[1]. The direct conversion of nitrogen in the air into ammonia is of paramount importance to both human beings and the planet’s ecosystem[2]. Yet, the Haber-Bosch (H-B) process, the main traditional process for producing ammonia in industry, requires extremely demanding conditions, such as high temperatures and pressures[3-5]. Furthermore, the hydrogen in the H-B process is derived from natural gas or methane[6], which means that the process will release large amounts of CO2[7]. Electrocatalytic nitrogen reduction reactions (eNRR) have been attracting increasing interest because of the mild reaction conditions compared to those of the H-B process; the reaction can be carried out under ambient conditions with the assistance of renewable electricity[8,9]. In addition, the source of hydrogen from water and the energy resource required for the eNRR process can be renewable, thus eliminating the negative impact of fossil fuels as the source of hydrogen and energy[1]. However, eNRR is a complex reaction involving six electron transfer processes, which is kinetically and thermodynamically more difficult compared to the hydrogen evolution reaction (HER), thus making the protons and electrons on the electrocatalyst susceptible to the production of H2 rather than NH3[10]. Therefore, it is desirable to develop catalysts for eNRR with high conversion and effective HER inhibition[11].

Many candidates could be used as eNRR electrocatalysts, for example Chen et al. used plasma-etched Ti2O3 (OV-Ti2O3) with oxygen vacancies to obtain maximum NH3 yield up to 37.24 μg·h-1·mgcat.-1 in 0.1 M HCl[12], and they proposed a mixture of titanium dioxide and effusus-derived carbon microtubes with three-dimensional cross-linked hollow tubular structures, which gave NH3 yield as high as 20.03 μg·h-1·mgcat.-1 in 0.1 M Na2SO4[13]. The single-atom catalysts (SACs)[14], which offer a new way to maximize the efficiency of metal atom use, have emerged as promising candidates[15]. The localized structures of the metal atoms allow them to exhibit excellent electronic properties and catalytic performance in a variety of chemical reactions[16]. However, the surface free energy of metals increases significantly with decreasing particle size, and metal atoms tend to aggregate[17]. Additionally, the stability of SACs in practical reactions depends on the surface configuration, the substrate, the reactants, and the limited temperature/pressure, which are not well defined and vary from one system to another[18]. The appropriate choice of substrates is a major (and in some cases a critical) factor in determining the catalytic performance of the catalyst, which should provide a large specific area to accommodate the metal sites and be able to modulate the local geometry and electronic structure of the metal atoms in the reaction, thus determining the overall catalytic performance[19]. Two-dimensional (2D) carbon[11,20], N-doped carbon[21-24], etc.[25-28], used as substrates for metal atoms as atomic catalysts, exhibit excellent efficiency toward eNRR. As an extension of SACs, double-atom catalysts (DACs) became another research hotspot due to their higher metal loading than SACs[22]. In terms of the overall reaction, DACs have multiple active sites (top and bridge) since their active centers are composed of two adjacent metal atoms, which can significantly improve the Faraday efficiency[29]. The bridge-site adsorption behavior of metal-metal atoms was found to exhibit a strong synergistic effect of modulating the electronic structure and promoting adsorption[30]. In addition, differences in surface states in different homonuclear or heteronuclear DACs can have an important impact on the selectivity and activity of the reactions[31]. Thus, both SACs and DACs may be excellent electrocatalysts if the right substrate is chosen.

Previous studies revealed that excellent electrocatalytic performance for the nitrogen reduction reaction (NRR) can be obtained by loading iron onto different substrates, whether as monoatoms[32-36], dimers[22,37-39], or other sizes[40-44]. In addition, given that the iron-molybdenum cofactor (FeMoco) is thought to be the site of dinitrogen reduction[45], we focused on Mo/Fe-containing catalysts. The chosen substrate was prompted by the recent experimental preparation of the 2D material Ag3C20[46], a planar structure with a rectangular lattice containing three Ag atoms and 20 C atoms. The three metal atoms in the lattice can be considered as a single atom (M1) and a bimetallic atom (M2) site. In this work, we replaced the precious metal Ag atoms in Ag3@C20 with Mo/Fe atoms and obtained the Mo3/Fe3@C20 monolayer. Through theoretical studies of the eNRR process, we found that Mo3@C20 is not suitable for eNRR due to its strong N2 adsorption; in contrast, Fe3@C20 is a good eNRR catalyst (the limiting potential UL = -0.59 V). With the demonstration that Fe3@C20 can be used as an electrocatalyst for eNRR, we substituted the C bonded with Fe by N to tune the coordination environment of the metal sites in order to further improve the catalytic efficiency. The limiting potential was further changed to -0.45 V when the N doping of the Fe2 site was the maximum; i.e., all the four C bonded with Fe2 sites were replaced by N with the stoichiometry of Fe3@N4C16.


All density functional theory (DFT) calculations were carried out using the Vienna ab initio Simulation Package (VASP)[47]. The Perdew, Burke, and Ernzerhof (PBE)[48] parameterization of the generalized gradient approximation (GGA) was used to describe the exchange-correlation function. The DFT-D2 dispersion correction scheme was applied for van der Waals (vdW) interactions[49]. A cut-off energy of 600 eV was used (convergence tests are shown in Supplementary Figure 1). The convergence parameters for force and energy were 0.02 eV/Å and 10-5 eV, respectively, and the convergence parameters were set the same as those in previous NRR calculations[50]. The Brillouin zone was sampled using a Monkhorst-Pack k-point grid[51] of 3 × 5 × 1. The Bader charge analysis[52] was employed to evaluate the charge transfer.

The adsorption energy of N2 was calculated by Eads = Etot - E* - $$ E_{\mathrm{N}_{2}} $$, where Etot, E*, and $$ E_{\mathrm{N}_{2}} $$ are the total energy of N2 adsorbed on the catalyst, the clean catalyst, and free N2, respectively. The free energy G was obtained based on the computational hydrogen electrode (CHE) model proposed by Nørskov et al.[53], with the formula:

$$ \Delta G=\Delta E+\Delta E_{Z P E}-T \Delta S+\Delta G_{U}+\Delta G_{p H} $$

In the above equation, E is the DFT energy, EZPE and S are the zero-point energy and entropy[54], respectively, EZPE and TS can be obtained by VASPKIT post-processing program[55], ΔGU = -eU, ΔGpH = -kBTln10 × pH, kB is the Boltzmann constant, T is 298.15 K[56], and the pH is zero in this study. The corrected values of the gas molecules and intermediates were given in Supplementary Tables 1 and 2, respectively. The limiting potential (UL) of the entire reduction process was determined by the potential limiting step and was computed by UL = -ΔGmax/e[57].

We used the following equation for the formation energy of N doping:

$$ E_{f}=E_{\mathrm{Fe}_{3} @ \mathrm{~N}_{X} \mathrm{C}_{20-X}}-E_{\mathrm{Fe}_{3} @ \mathrm{C}_{20}}-x E_{\mathrm{N}}+x E_{\mathrm{C}} $$

where $$ E_{\mathrm{Fe}_{3} @ \mathrm{~N}_{X} \mathrm{C}_{20-X}} $$ and $$ E_{\mathrm{Fe}_{3} @ \mathrm{C}_{20}} $$ are the energies of N-doped and pure Fe3@C20, respectively, x is the number of N atoms, and EN and EC represent the energies of a free N/C atom. All the details of the calculations and the results of the energy difference calculations are presented in Supplementary Table 3.


The structure and the eNRR performance of the M3@C20 (M = Mo and Fe) monolayers

After structural optimization, the structures are represented by the Fe3@C20 shown in Figure 1A with rectangular lattice constants of a = 11.14 Å and b = 7.95 Å, and a = 11.59 Å and b = 8.45 Å for Mo3@C20, respectively. Due to the different loading positions of the metal atoms, they can be characterized as single-atom (M1) and double-atom (M2) sites (the Mo-Mo/Fe-Fe bond lengths at the diatomic sites are 2.85/2.22 Å, which also demonstrates the formation of metallic bonds in the M2 sites). The results of the electron localization function (ELF: ELF values of 1.0 and 0.5, representing fully localized and free electron gases respectively, while ELF values close to 0 represent the low electron density region, Supplementary Figure 2) show that electron transfer from the Mo/Fe atoms to the C atoms, which is in line with the Bader charge analysis [Mo1 (Mo2): 1.16 (1.46) |e| and Fe1 (Fe2): 0.96 (1.34) |e|]. The covalent bond between C and C atoms (1.32~1.47 Å) can also be identified. The total density of states (TDOS) [Figure 1B and Supplementary Figure 3] reveals that there are electronic states crossing the Fermi energy level, indicating the metallicity of both Fe3@C20 and Mo3@C20, which is beneficial to electrocatalysis.

Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe<sub>3</sub>@N<sub><i>x</i></sub>C<sub>20-<i>x</i></sub> (<i>x</i> = 0~4): a DFT exploration

Figure 1. (A) Top and side views of the optimized 2D Fe3@C20 monolayer with the unit cell marked by black solid lines; (B) the partial density of states of the d orbitals of Fe atoms (red), and the DOS of the Fe3@C20 monolayer (black) with the Fermi level indicated by the red dashed line.

For the electrochemical synthesis of ammonia, the adsorption of N2 on the catalyst surface is the first step in the process. We considered two possible N2 adsorption configurations, i.e., the side-on and end-on structures, on both M1 and M2 sites, respectively. The results are displayed in Figure 2 and Supplementary Figure 4. After N2 adsorption onto Mo3@C20, the distances between N2 and Mo after structural optimization ranged from 1.95 to 2.25 Å. It is noteworthy that when N2 adsorbs to the Mo2 site in the side-on conformation, it will spontaneously shift to the end-on conformation; thus, we only considered the end-on adsorption structure. The N−N bond lengths of N2 at the Mo1 (Mo2) site in the side-on and end-on conformations are 1.21 and 1.14 (1.12) Å, respectively. The adsorption energies (Eads) of N2 at the Mo1 (Mo2) sites are all larger than 5.00 eV; such strong adsorption is not conducive to subsequent reactions or nitride releasing, and therefore Mo3@C20 is not suitable for the eNRR. In contrast, the distances between N of N2 and Fe range from 1.88 to 3.13 Å. The N−N bond lengths of the adsorbed N2 adopting side-on and end-on configurations are 1.11 (1.18) and 1.13 (1.14) Å at the Fe1 (Fe2) site, respectively, and the Eads values of N2 are -0.10 (-0.57) eV and -0.41 (-0.52) eV at the Fe1 (Fe2) site, respectively. In comparison with the 1.11 Å of N−N length of an isolated N2 molecule, the N−N length is elongated, and the adsorbed N2 is activated (also indicated by the considerable adsorption strength), except for the case of N2 adsorbing on the Fe1 site with side-on structures. The Bader charge analysis showed that the number of electrons transferred from Fe1(Fe2) to the adsorbed N2 with side-on and end-on configurations are 0.00 (0.49) and 0.16 (0.31) |e|, respectively, agreeing well with the adsorption strengths, which also suggests that the activation of N2 at the Fe2 site is more efficient than that at the Fe1 site.

Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe<sub>3</sub>@N<sub><i>x</i></sub>C<sub>20-<i>x</i></sub> (<i>x</i> = 0~4): a DFT exploration

Figure 2. Different adsorption structures of N2 on Fe3@C20: (A) end-on and (B) side-on configurations at Fe1 sites; and (C) end-on and (D) side-on configurations at Fe2 sites, with the unit cells marked by black lines.

Before conducting the eNRR pathway test on the Fe2 site, we evaluated the stability of Fe3@C20 by carrying out first-principles molecular dynamics (FPMD) simulations at 300 K for 5 ps (with a time step of 1 fs). The result showed that both the Fe1 and Fe2 moieties in the 2D structure were well preserved during the FPMD simulations [Supplementary Figure 5], which demonstrates that Fe3@C20 has good thermodynamic stability.

The two eNRR pathways on the Fe2 site of Fe3@C20 to produce ammonia started from side-on (end-on) adsorption are shown in Figure 3A and B, which are the enzymatic (alternative) mechanism [* → *N2 → *NNH → *NHNH → *NH2NH → *NH2NH → *NH2NH2 → *NH2 → *NH3 → NH3(g)] and the consecutive (distal) mechanism [*→ *N2 → *NNH → *NNH2 → *N → *NH2 → *NH3 → NH3(g)], respectively. The main reasons for considering the above pathways can be assigned to the following two aspects: (1) N2 has different adsorption configurations; (2) the proton-electron (H+ + e-) pairs attack N atoms in different manners [whether they attack one N atom consecutively first (consecutive/distal route) or attack the two N atoms alternatively (alternating/enzymatic path)].

Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe<sub>3</sub>@N<sub><i>x</i></sub>C<sub>20-<i>x</i></sub> (<i>x</i> = 0~4): a DFT exploration

Figure 3. Different eNRR mechanisms on Fe3@C20 (A and B) and the free energy diagrams (C and D). Data denote the ΔG of each elementary step. eNRR: Electrochemical nitrogen reduction reaction.

Following the above mechanisms, the structural illustrations of each intermediate and free energy diagram are given in Figure 3. The results show that the change in free energy is preferred along the distal pathway. In detail, starting from the N2 adsorbing with side-on configurations, the free energy change (ΔG) follows the sequence of -0.09, 0.64, 0.00, -0.12, -0.04, -1.82, 0.07, and 0.53 (-0.09, 0.64, 0.50, -1.05, -0.57, -0.87, 0.07, and 0.53) eV for each elementary step along the enzymatic (consecutive) route; while starting from the end-on adsorption of N2, ΔG goes through -0.09, 0.59, -0.06, -0.20, 0.14, -1.83, 0.08, and 0.53 (-0.09, 0.59, -0.04, -0.47, -0.57, -0.87, 0.08, and 0.53) eV for each elementary step, along the alternative (distal) path. The potential determining step (PDS) is the first hydrogenation step of N2 (*NN → *NNH) in the four eNRR pathways with the largest ΔG of 0.64 or 0.59 eV, which is comparable to that over Fe3@gra (0.60 eV)[58] and smaller than that on Fe-C2N (1.06 eV)[22]. The free energy change of releasing NH3 [*NH3 → NH3(g), ΔG = 0.53 eV] is less than the ΔG of the first hydrogenating (ΔG = 0.64 or 0.59 eV), which is beneficial in facilitating the recovery of the catalyst.

The HER, requiring proton-electron pairs to proceed with the reaction, is the primary competitor against the eNRR. Therefore, we calculated the adsorption energy of *H and found Eads of *H (-0.30 eV) > Eads of *N2 (-0.57/-0.52 eV) on the Fe2 site. In addition, to exclude the possibility that the Fe1 site may also proceed HER, lowering the selectivity towards eNRR, we also calculated Eads of *H on the Fe1 site (0.16 eV). All results suggest that HER would be effectively suppressed over the Fe3@C20.

The geometry and the eNRR performance of the Fe3@NxC20-x (x = 1~4)

Previous studies showed that the catalytic performance of catalysts is associated with the coordination environment of the active site. Thus, 2D Fe3@NxC20-x (x ≤ 4) materials were constructed by replacing Fe2-bonded C with N, and all the optimized structures remain well with the initial plane [Supplementary Figure 6]. The formation energy of Fe3@NxC20-x decreases from -0.20 eV at x = 1 to -0.65 eV at x = 4 [Supplementary Table 3]. The lattice constants for these structures were given in Supplementary Table 4. Note that the N doping at x = 2 yields three structures, i.e., Fe3@N2C18-I, Fe3@N2C18-II and Fe3@N2C18-III [Supplementary Figure 6B-D], with the lowest Ef (-0.50 eV) of asymmetric substitution Fe3@N2C18-II compared to the other two structures (Ef = -0.22 eV for I and Ef = -0.35 eV for III). The lattice constant a/b decreases from 11.14/7.95 Å at x = 0 to 10.87/7.74 Å at x = 4, which can be understood by the smaller radius of N than that of C.

The adsorption energies of N2 with side-on configurations on Fe2 sites of most N-substituted monolayers are also stronger than those with end-on structures, and the adsorption strength decreases as the content of N increases, as demonstrated by the data in Supplementary Table 4: the Eads values of the side/end-on adsorption configurations on Fe2 sites of 2D Fe3@NxC20-x (x ≤ 4) monolayers are -0.62/-0.55 eV for Fe3@N1C19, -0.59/-0.55 eV for Fe3@N2C18-I, -0.50/-0.48 eV for Fe3@N2C18-II, -0.59/-0.52 eV for Fe3@N2C18-III, -0.54/-0.52 eV for Fe3@N3C17, and -0.20/-0.32 eV for Fe3@N4C16, respectively. The N2 adsorption structures are shown in Figure 4A and B, and Supplementary Figure 7. In the proceeding section, the PDSs for the considered routes on Fe3@C20 are the first hydrogenation step of N2 (*N2 → *NNH). Then, we evaluate the ΔG values of the first hydrogenation step of N2 on the Fe2 site of 2D Fe3@NxC20-x (x ≤ 4) [Supplementary Table 5]. The free energy changes on these N-substituted materials range from 0.45 to 1.23 eV. The ΔG values of the Fe3@N1C19 (0.74 eV), Fe3@N2C18-I (0.61 eV), Fe3@N2C18-II (0.73 eV), Fe3@N2C18-II (0.65 V), and Fe3@N3C17 (0.67 V) are much smaller than those of the FeCN2 (0.94 eV) and FeCN3 (0.92 eV)[59], suggesting the more pronounced effect of the DAC environment than the SAC environment on the NRR performance. Unfortunately, all of them have free energy change values higher than 0.59 eV of Fe3@N20 [Figure 4C], and therefore, these subsequent pathways will not be considered. In contrast, the free energy change for the first hydrogenation of N2 of side-on adsorption on Fe3@N4C16 (0.45 eV) is smaller than that of Fe3@C20. It is worth noting that after the first hydrogenation of end-on adsorbed N2, the *NNH configuration is automatically transformed to the side-on adsorbed *NNH configuration after structural optimization. Therefore, we only concentrated on the enzymatic and consecutive eNRR paths over Fe3@N4C16.

Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe<sub>3</sub>@N<sub><i>x</i></sub>C<sub>20-<i>x</i></sub> (<i>x</i> = 0~4): a DFT exploration

Figure 4. Top and side views of N2 adsorbed at the Fe2 site of Fe3@N4C16 with the side-on configuration (A); the end-on configuration (B) with the unit cells marked by black lines; ΔG for the *NN → *NNH step of the energy preferred path for the Fe3@NxC20-x(x = 0~4) materials and the black dashed line indicates the free energy change value for Fe3@C20 (C) and the eNRR free energy diagram over the Fe2 site of Fe3@N4C16 (D). eNRR: Electrochemical nitrogen reduction reaction.

On the Fe3@N4C16, the free energy change for each elementary step along the enzymatic mechanism was 0.21, 0.45, -0.07, 0.01, 0.09, -2.12, 0.19, and 0.40 eV, and along the consecutive mechanism, it was 0.21, 0.45, 1.08, -1.43, -0.36, -1.38, 0.19 and 0.40 eV, respectively [Figure 4D]. Notably, although the ΔG of nitrogen adsorption is a positive value (0.21 eV), the adsorption energy before correction is negative (-0.20 eV), and the elongation of the N−N bond (1.17 Å) further suggests that the nitrogen was activated. However, in terms of the overall reaction, this small positive value is smaller than the free energy change in the first step of hydrogenation, i.e., *NN → *NNH (0.45 eV). As for the eNRR on the Fe2 site of Fe3@N4C16, the step *NN → *NNH becomes the PDS in the enzymatic mechanism, possessing a free energy change value of 0.45 eV, which is smaller than the values on Fe2Mo@C6N6 (0.53 eV)[60] and Fe/Mo-N-C (0.63 eV)[61]. Furthermore, the free energy change of releasing the second NH3(g) (0.40 eV) is also smaller than ΔG of the PDS (0.45 eV), indicating the feasibility of recovering the clean catalyst. In addition, we used the Poisson-Boltzmann implicit solvation model to simulate an aqueous solvent environment (with a dielectric constant of ε = 80)[62] to further explore the solvation effect on the eNRR performance of Fe3@N4C16, and the results are shown in Supplementary Figure 8. The value of ΔGmax in the consecutive/enzymatic pathway on Fe3@N4C16 with the solvent effect is reduced from 1.08/0.45 eV to 1.04/0.34 eV, which is consistent with the path selectivity of our calculations without solvent effect and demonstrates the selectivity of Fe3@N4C16 to be an excellent NRR catalyst.

In addition, since the 2D Fe3@NxC20-x (x ≤ 4) includes both Fe1 and Fe2 sites, to make sure that the Fe2 site is the active site for eNRR to proceed, we also calculated the adsorption energy of N2 at the Fe1 site for comparison [Supplementary Table 4]. The adsorption energies of N2 for the side-on/end-on adsorption configuration at the Fe1 site are -0.11/-0.43 eV for Fe3@N1C19, -0.06/-0.39 eV for Fe3@N2C18-I, -0.08/-0.47 eV for Fe3@N2C18-II, and -0.17/-0.43 eV for Fe3@N2C18-III. For x = 1, 2, it indicates that N2 prefers to adsorb at Fe2 sites, while N2 may prefer to adsorb on the Fe1 site of Fe3@N3C17 (-0.08/-0.58 eV) and Fe3@N4C16 (0.15/-0.44 eV). Therefore, we further tested for further hydrogenation of N2 (*NN → *NNH) on the Fe1 site for x = 3, 4. However, the free energy changes for the first step of hydrogenation of N2 end-on adsorbed on the Fe1 site are 0.95 (Fe3@N3C17) and 1.24 eV (Fe3@N4C16), respectively. Such large free energy change is also unfavorable for eNRR, and for the reaction on Fe3@N4C16, even though nitrogen tends to adopt end-on adsorption on the Fe1 site, it is difficult to proceed the next step of the hydrogenation after this active site is occupied by nitrogen, and thus, the preferred N2 adsorption on Fe1 site will not affect the reaction on the Fe2 site. Besides, to guarantee the selectivity towards eNRR instead of HER, the adsorption of *H at the Fe1 site for x = 1~4. The adsorption strength of N2 at Fe2 sites prevails that of *H at Fe2 (Fe1) site: -0.62/-0.55 vs. -0.28 (0.24) eV for Fe3@N1C19, -0.59/-0.55 vs. -0.48 (0.25) eV for Fe3@N2C18-I, -0.50/-0.48 vs. -0.30 (0.65) eV for Fe3@N2C18-II, -0.59/-0.52 vs. -0.24 (0.09) eV for Fe3@N2C18-III, -0.59/-0.52 vs. -0.39 (0.23 eV) for Fe3@N3C17, and -0.20/-0.32 vs. 0.02 (0.45) eV for Fe3@N4C16.

All the above calculations show that the Fe2 site remains the active center with N-doping varying from 1 to 4, and N-doping will not change the selectivity toward eNRR to HER. Most importantly, with the lowest ΔGmax and the preferent adsorption of N2 against *H, Fe3@N4C16 distinguishes itself as the most promising candidate for eNRR.

The origin of different eNRR performance of Fe3@NxC20-x (x = 0~4)

As an eNRR catalyst, the free energy change results show that the limiting potential of Fe3@N4C16 (0.45 eV) is smaller than that of Fe3@C20 (0.59 eV) and is even better than the reported UL values FeN4/G (FeN4 moiety embedded in graphene, 1.30 eV)[63] and FeN4/FeMoPPc (FeN4 moiety embedded with polyphthalocyanine, 0.66 eV)[64]. To figure out the origin of the significantly improved eNRR activity of Fe3@N4C16, we analyzed the geometric and electronic structures of Fe2 sites in Fe3@NxC20-x(x = 0~4). As illustrated in Supplementary Table 6, the FeFe bond length of the Fe2 site shows slight variation for x = 0~3 (dFe−Fe = 2.22~2.24 Å) and an obvious drop to 2.16 Å for x = 4, and the Fe2-d orbitals exhibit a better hybridization with the adsorbed N2-p orbitals [Supplementary Figure 9], which further leads to more efficient activation of N−N as demonstrated by the lower ΔG of the first hydrogenating step. Besides, both the d-band center and the p-band center of Fe2-bonded C/N in Fe3@N4C16 have moderate values among Fe3@NxC20-x(x = 0~4), showing an inverse volcano relationship [Figure 5 and Supplementary Table 6].

Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe<sub>3</sub>@N<sub><i>x</i></sub>C<sub>20-<i>x</i></sub> (<i>x</i> = 0~4): a DFT exploration

Figure 5. The variation of ΔG for the first hydrogenating of side-on N2 against the d-band center (A) and p-band center of Fe2-bonded C/N (B) for the Fe3@NxC20-x(x = 0~4).


In summary, we theoretically investigated the eNRR performance of Mo3@C20 and Fe3@C20 monolayers based on the experimentally synthesized 2D Ag3@C20 structure. Our calculations revealed that Fe3@C20 exhibits good activity and selectivity for eNRR with a predicted limiting potential of -0.59 V. By further investigation of the synergy between the active Fe2 site and its coordination, we found that the N-containing environment would tune the electronic properties and, thus, enhance the catalytic efficiency, where 2D Fe3@N4C16 is demonstrated as the superior electrocatalyst with a limiting potential of -0.45 V. Moreover, the Fe2 sites in our designed Fe3@NxC20-x materials have much smaller ΔGmax than those of Fe-based SACs such as FeN4/G (1.30 eV)[63], FeCN2 (0.94 eV) and FeCN3 (0.92 eV)[59], and the best candidate Fe3@N4C16 also has smaller ΔGmax than that of the Fe-based triple-atom catalyst Fe3@C2N (0.57 eV)[65]. These findings provide new ideas for the development of new eNRR catalysts following experimental fabrication to theoretical prediction to experimental guidelines and highlight the important role of the metal-coordination environment.


Authors’ contributions

Performed the computations and drafted the manuscript: Han B

Provided guidance and revised the manuscript: Li F

Availability of data and materials

Supplementary materials are available from the Journal of Materials Informatics or from the authors.

Financial support and sponsorship

This work was supported by the National Natural Science Foundation of China (11964024), the “Grassland Talents” project of the Inner Mongolia autonomous region (12000-12102613), and the young science and technology talents cultivation project of Inner Mongolia University (21200-5223708). We thank the computational support from Beijing PARATERA.

Conflicts of interest

All authors declared that there are no conflicts of interest.

Ethical approval and consent to participate

Not applicable.

Consent for publication

Not applicable.


© The Author(s) 2023.

Supplementary Materials


1. Foster SL, Bakovic SIP, Duda RD, et al. Catalysts for nitrogen reduction to ammonia. Nat Catal 2018;1:490-500.

2. Wang S, Ichihara F, Pang H, Chen J, Ye J. Nitrogen fixation reaction derived from nanostructured catalytic materials. Adv Funct Mater 2018;28:1803309.

3. Liu H. Ammonia synthesis catalyst 100 years: practice, enlightenment and challenge. Chinese J Catal 2014;35:1619-40.

4. Guo X, Huang S. Tuning nitrogen reduction reaction activity via controllable Fe magnetic moment: a computational study of single Fe atom supported on defective graphene. Electrochim Acta 2018;284:392-9.

5. Chen ZW, Lang XY, Jiang Q. Discovery of cobweb-like MoC6 and its application for nitrogen fixation. J Mater Chem A 2018;6:9623-8.

6. Ritter SK. Iron’s star rising. Chem Eng News 2008;86:53-7.

7. Bai J, Huang H, Li FM, et al. Glycerol oxidation assisted electrocatalytic nitrogen reduction: ammonia and glyceraldehyde co-production on bimetallic RhCu ultrathin nanoflake nanoaggregates. J Mater Chem A 2019;7:21149-56.

8. Wang F, Mao J. Effect of N-doping on graphene: NRR activity and N-source. Diam Relat Mater 2021;118:108494.

9. Choi C, Back S, Kim NY, Lim J, Kim YH, Jung Y. Suppression of hydrogen evolution reaction in electrochemical N2 reduction using single-atom catalysts: a computational guideline. ACS Catal 2018;8:7517-25.

10. Liu Q, Xu T, Luo Y, et al. Recent advances in strategies for highly selective electrocatalytic N2 reduction toward ambient NH3 synthesis. Curr Opin Electrochem 2021;29:100766.

11. Li M, Cui Y, Zhang X, Luo Y, Dai Y, Huang Y. Screening a suitable Mo form supported on graphdiyne for effectively electrocatalytic N2 reduction reaction: from atomic catalyst to cluster catalyst. J Phys Chem Lett 2020;11:8128-37.

12. Chen H, Xu Z, Sun S, et al. Plasma-etched Ti2O3 with oxygen vacancies for enhanced NH3 electrosynthesis and Zn-N2 batteries. Inorg Chem Front 2022;9:4608-13.

13. Chen H, Liang J, Dong K, et al. Ambient electrochemical N2-to-NH3 conversion catalyzed by TiO2 decorated juncus effusus-derived carbon microtubes. Inorg Chem Front 2022;9:1514-9.

14. Qiao B, Wang A, Yang X, et al. Single-atom catalysis of CO oxidation using Pt1/FeOx. Nat Chem 2011;3:634-41.

15. Zhang W, Fu Q, Luo Q, Sheng L, Yang J. Understanding single-atom catalysis in view of theory. JACS Au 2021;1:2130-45.

16. Wang Z, Yu Z, Zhao J. Computational screening of a single transition metal atom supported on the C2N monolayer for electrochemical ammonia synthesis. Phys Chem Chem Phys 2018;20:12835-44.

17. Chen Z, Zhao J, Cabrera CR, Chen Z. Computational screening of efficient single-atom catalysts based on graphitic carbon nitride (g-C3N4) for nitrogen electroreduction. Small Methods 2019;3:1800368.

18. Liu JC, Tang Y, Wang YG, Zhang T, Li J. Theoretical understanding of the stability of single-atom catalysts. Natl Sci Rev 2018;5:638-41.

19. Pan Y, Zhang C, Liu Z, Chen C, Li Y. Structural regulation with atomic-level precision: from single-atomic site to diatomic and atomic interface catalysis. Matter 2020;2:78-110.

20. Zou H, Arachchige LJ, Rong W, et al. Low-valence metal single atoms on graphdiyne promotes electrochemical nitrogen reduction via M-to-N2 π-backdonation. Adv Funct Mater 2022;32:2200333.

21. Long J, Fu X, Xiao J. The rational design of single-atom catalysts for electrochemical ammonia synthesis via a descriptor-based approach. J Mater Chem A 2020;8:17078-88.

22. Chen ZW, Yan JM, Jiang Q. Single or double: which is the altar of atomic catalysts for nitrogen reduction reaction? Small Methods 2019;3:1800291.

23. Zhang X, Chen A, Zhang Z, Zhou Z. Double-atom catalysts: transition metal dimer-anchored C2N monolayers as N2 fixation electrocatalysts. J Mater Chem A 2018;6:18599-604.

24. Wang YH, Dong JH, Tan Z, Wang XF, Song XZ. The journey of iron-based electrocatalytic materials for nitrogen reduction reaction: from current status to future prospects. J Mater Chem A 2023;11:11048-77.

25. Sathishkumar N, Chen HT. Regulating the coordination environment of single-atom catalysts anchored on thiophene linked porphyrin for an efficient nitrogen reduction reaction. ACS Appl Mater Interfaces 2023;15:15545-60.

26. Hou P, Huang Y, Ma F, et al. Screening of single-atom catalysts of transition metal supported on MoSe2 for high-efficiency nitrogen reduction reaction. Mol Catal 2023;537:112967.

27. Xu T, Liang J, Li S, et al. Recent advances in nonprecious metal oxide electrocatalysts and photocatalysts for N2 reduction reaction under ambient condition. Small Sci 2021;1:2000069.

28. Cheng H, Ding LX, Chen GF, Zhang L, Xue J, Wang H. Molybdenum carbide nanodots enable efficient electrocatalytic nitrogen fixation under ambient conditions. Adv Mater 2018;30:1803694.

29. Li R, Wang D. Superiority of dual-atom catalysts in electrocatalysis: one step further than single-atom catalysts. Adv Energy Mater 2022;12:2103564.

30. Yang W, Jia Z, Chen L, et al. Effects of intermetal distance on the electrochemistry-induced surface coverage of M-N-C dual-atom catalysts. Chem Commun 2023;59:10761-4.

31. Yang W, Jia Z, Zhou B, et al. Why is C–C coupling in CO2 reduction still difficult on dual-atom electrocatalysts? ACS Catal 2023;13:9695-705.

32. Chen Z, Liu C, Sun L, Wang T. Progress of experimental and computational catalyst design for electrochemical nitrogen fixation. ACS Catal 2022;12:8936-75.

33. Liu Y, Fan X, Bian W, et al. High-loading Fe1 sites on vanadium disulfides: a scalable and non-defect-stabilized single atom catalyst for electrochemical nitrogen reduction. J Mater Chem A 2022;10:21142-8.

34. Zhao X, Zhang X, Xue Z, Chen W, Zhou Z, Mu T. Fe nanodot-decorated MoS2 nanosheets on carbon cloth: an efficient and flexible electrode for ambient ammonia synthesis. J Mater Chem A 2019;7:27417-22.

35. Banerjee A. Computational screening of χ3 borophene based single-atom catalysts for N2 reduction. Catal Today 2023;418:114079.

36. Sahoo SK, Heske J, Antonietti M, Qin Q, Oschatz M, Kühne TD. Electrochemical N2 reduction to ammonia using single Au/Fe atoms supported on nitrogen-doped porous carbon. ACS Appl Energy Mater 2020;3:10061-9.

37. Zhang H, Cui C, Luo Z. MoS2-supported Fe2 clusters catalyzing nitrogen reduction reaction to produce ammonia. J Phys Chem C 2020;124:6260-6.

38. Yao X, Zhang Z, Chen L, Chen ZW, Zhu YF, Singh CV. Work function-tailored nitrogenase-like Fe double-atom catalysts on transition metal dichalcogenides for nitrogen fixation. ACS Sustainable Chem Eng 2023;11:4990-7.

39. Zhang Z, Huang X, Xu H. Anchoring an Fe dimer on nitrogen-doped graphene toward highly efficient electrocatalytic ammonia synthesis. ACS Appl Mater Interfaces 2021;13:43632-40.

40. Liu JC, Ma XL, Li Y, Wang YG, Xiao H, Li J. Heterogeneous Fe3 single-cluster catalyst for ammonia synthesis via an associative mechanism. Nat Commun 2018;9:1610.

41. Chen S, Gao Y, Wang W, Prezhdo OV, Xu L. Prediction of three-metal cluster catalysts on two-dimensional W2N3 support with integrated descriptors for electrocatalytic nitrogen reduction. ACS Nano 2023;17:1522-32.

42. Xie K, Wang F, Wei F, Zhao J, Lin S. Revealing the origin of nitrogen electroreduction activity of molybdenum disulfide supported iron atoms. J Phys Chem C 2022;126:5180-8.

43. Chen ZW, Chen LX, Jiang M, et al. A triple atom catalyst with ultrahigh loading potential for nitrogen electrochemical reduction. J Mater Chem A 2020;8:15086-93.

44. Dai T, Lang X, Wang Z, Wen Z, Jiang Q. Rational design of an Fe cluster catalyst for robust nitrogen activation. J Mater Chem A 2021;9:21219-27.

45. Lancaster KM, Roemelt M, Ettenhuber P, et al. X-ray emission spectroscopy evidences a central carbon in the nitrogenase iron-molybdenum cofactor. Science 2011;334:974-7.

46. Zhong Q, Niu K, Chen L, et al. Substrate-modulated synthesis of metal-organic hybrids by tunable multiple aryl-metal bonds. J Am Chem Soc 2022;144:8214-22.

47. Kresse G, Furthmüller J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys Rev B Condens Matter 1996;54:11169-86.

48. Perdew JP, Burke K, Ernzerhof M. Generalized gradient approximation made simple. Phys Rev Lett 1996;77:3865-8.

49. Bučko T, Hafner J, Lebègue S, Ángyán JG. Improved description of the structure of molecular and layered crystals: ab initio DFT calculations with van der Waals corrections. J Phys Chem A 2010;114:11814-24.

50. Wang M, Huang Y, Ma F, et al. Theoretical insights into the mechanism of nitrogen-to-ammonia electroreduction on TM/g-C9N10. Mol Catal 2023;547:113391.

51. Monkhorst HJ, Pack JD. Special points for Brillouin-zone integrations. Phys Rev B 1976;13:5188.

52. Henkelman G, Arnaldsson A, Jónsson H. A fast and robust algorithm for Bader decomposition of charge density. Comput Mater Sci 2006;36:354-60.

53. Nørskov JK, Rossmeisl J, Logadottir A, et al. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J Phys Chem B 2004;108:17886-92.

54. Tian Y, Wang Y, Yan L, Zhao J, Su Z. Electrochemical reduction of carbon dioxide on the two-dimensional M3(Hexaiminotriphenylene)2 sheet: a computational study. Appl Surf Sci 2019;467-8:98-103.

55. Wang V, Xu N, Liu JC, Tang G, Geng WT. VASPKIT: a user-friendly interface facilitating high-throughput computing and analysis using VASP code. Comput Phys Commun 2021;267:108033.

56. Chen H, Handoko AD, Xiao J, et al. Catalytic effect on CO2 electroreduction by hydroxyl-terminated two-dimensional MXenes. ACS Appl Mater Interfaces 2019;11:36571-9.

57. Skúlason E, Bligaard T, Gudmundsdóttir S, et al. A theoretical evaluation of possible transition metal electro-catalysts for N2 reduction. Phys Chem Chem Phys 2012;14:1235-45.

58. Liu X, Li C, Ma P, Zhang W, Jia M, Song W. A first-principles study of transition metal clusters supported on graphene as electrocatalysts for N2 to NH3 reaction. Mater Today Commun 2023;35:106353.

59. Guo Z, Liu C, Sun C, Xu J, Li H, Wang T. Tuning the coordination environment of single-atom iron catalysts towards effective nitrogen reduction. ChemCatChem 2023;15:e202300669.

60. Wang S, Zhao T, Yan L. Tailoring of three-atom metal cluster catalysts for ammonia synthesis. Catalysts 2023;13:869.

61. Gao S, Liu X, Wang Z, et al. Spin regulation for efficient electrocatalytic N2 reduction over diatomic Fe-Mo catalyst. J Colloid Interface Sci 2023;630:215-23.

62. Mathew K, Sundararaman R, Letchworth-Weaver K, Arias TA, Hennig RG. Implicit solvation model for density-functional study of nanocrystal surfaces and reaction pathways. J Chem Phys 2014;140:084106.

63. Yuan D, Wu D, Zhang J, et al. Effect of oxygen coordination on the electrocatalytic nitrogen fixation of a vanadium single-atom catalyst embedded in graphene. New J Chem 2022;46:22936-43.

64. Wang Y, Cheng W, Yuan P, et al. Boosting nitrogen reduction to ammonia on FeN4 sites by atomic spin regulation. Adv Sci 2021;8:2102915.

65. Han B, Meng H, Li F, Zhao J. Fe3 cluster anchored on the C2N monolayer for efficient electrochemical nitrogen fixation. Catalysts 2020;10:974.

Cite This Article

OAE Style

Han B, Li F. Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe3@NxC20-x (x = 0~4): a DFT exploration. J Mater Inf 2023;3:24.

AMA Style

Han B, Li F. Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe3@NxC20-x (x = 0~4): a DFT exploration. Journal of Materials Informatics. 2023; 3(4): 24.

Chicago/Turabian Style

Han, Bing, Fengyu Li. 2023. "Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe3@NxC20-x (x = 0~4): a DFT exploration" Journal of Materials Informatics. 3, no.4: 24.

ACS Style

Han, B.; Li F. Regulating the electrocatalytic performance for nitrogen reduction reaction by tuning the N contents in Fe3@NxC20-x (x = 0~4): a DFT exploration. J. Mater. Inf. 2023, 3, 24.



Open Access Research Article
Investigation of dual atom doped single-layer MoS2 for electrochemical reduction of carbon dioxide by first-principle calculations and machine-learning
Available online: 20 Nov 2023
Open Access Research Article
Physics infused machine learning force fields for 2D materials monolayers
Available online: 2 Nov 2023
Open Access Review
Recent advances in the interface structure prediction for heteromaterial systems
Available online: 17 Oct 2023
Open Access Research Article
Understanding oxidation resistance of Pt-based alloys through computations of Ellingham diagrams with experimental verifications
Available online: 11 Oct 2023
Open Access Review
Characterizing the wetting behavior of 2D materials: a review
Available online: 13 Sep 2023
Open Access Research Article
Machine learning for prediction of CO2/N2/H2O selective adsorption and separation in metal-zeolites
Available online: 5 Sep 2023
Open Access Review
Recent advances and applications of machine learning in electrocatalysis
Available online: 30 Aug 2023
Open Access Research Article
Effects of solutes on thermodynamic properties of (TMZrU)C (TM = Ta, Y) medium-entropy carbides: a first-principles study
Available online: 21 Aug 2023
Open Access Review
Virtual sample generation in machine learning assisted materials design and discovery
Available online: 12 Jul 2023
Open Access Research Article
Assessment of phase equilibria and thermodynamic properties in the Fe-RE (RE = rare earth metals) binary systems
Available online: 11 Jul 2023


Comments must be written in English. Spam, offensive content, impersonation, and private information will not be permitted. If any comment is reported and identified as inappropriate content by OAE staff, the comment will be removed without notice. If you have any queries or need any help, please contact us at

Cite This Article 1 clicks
Commentary 0 comments
Like This Article 3 likes
Share This Article
Scan the QR code for reading!
See Updates
Hot Topics
machine learning |
Journal of Materials Informatics
ISSN 2770-372X (Online)
Follow Us


All published articles are preserved here permanently:


All published articles are preserved here permanently: