View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/D4TA04795H (Paper) J. Mater. Chem. A, 2024, 12, 23612-23621

2D conductive metal–organic frameworks for NO electrochemical reduction: a first-principles study†

Xing Chen‡ a, Xiangyu Zhu‡ a, Zhiyuan Xia a, Shiting Qian a, Yanan Zhou *b, Qiquan Luo *a and Jinlong Yang *c
aInstitutes of Physical Science and Information Technology, Anhui University, Hefei 230601, China. E-mail: qluo@ustc.edu.cn
bSchool of Material Science and Chemical Engineering, Institute of Mass Spectrometry, Ningbo University, Fenghua Road 818, Ningbo 315211, China. E-mail: zhouyanan@nbu.edu.cn
cKey Laboratory of Precision and Intelligent Chemistry, Department of Chemical Physics, Hefei National Research Center for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, Anhui 230026, China. E-mail: jlyang@ustc.edu.cn

Received 11th July 2024 , Accepted 5th August 2024

First published on 6th August 2024

Abstract

Designing flexible single-atom catalysts with tunable single-atom centers and coordination environments is crucial for highly active and selective electrochemical catalysis. Using density functional theory calculations, a range of 32 two-dimensional conductive metal–organic frameworks (cMOFs: TMX4–HTPs, TM = Sc–Ni, X = N, O, P, S) were investigated as efficient catalysts for electrocatalytic nitric oxide reduction reaction (eNORR) towards ammonia. By screening for stability, selectivity, and activity, eight TMX4–HTPs are identified as potential high-performance catalysts. Among them, MnO4–HTP delivers the lowest overpotential of only 0.21 V. Using this system as an example, the solvent effect, protonation ability of the potential-limiting step, and constant potential model were additionally considered and simulated. The computed results further verify the predicted potential of these catalysts as eNORR catalysts. Furthermore, descriptors are obtained to evaluate the competitive ability of key adsorbates using the sure independence screening and sparsifying operator method. All parameters are related to the intrinsic properties of TM (such as electronegativity, valence electron number, first ionization energy, and relative atomic mass) without extensive calculations. This work paves the way for highly active and selective cMOF-based electrocatalysts for eNORR.

1. Introduction

Nitric oxide (NO) is a common pollutant emitted by fossil fuel power plants and automobile engines, with its excessive emissions posing potential threats to human health and the ecological environment.1–4 Conventional methods for reducing NO emissions, such as selective catalytic reduction, require continuous ammonia or urea supply, leading to high operating costs and commercially non-valuable nitrogen byproducts.5–8 A recently proposed method involves synthesizing ammonia through electrocatalytic NO reduction (eNORR), which shows promise in alleviating the high energy consumption and significant CO2 emissions associated with the traditional Haber–Bosch process.9–13 However, research on eNORR catalysts has mainly focused on metal catalysts,14,15 such as Pt,16–18 Au,19 and Ru.20 These catalysts usually suffer from slow reaction kinetics and competition with the hydrogen evolution reaction (HER), hindering the large-scale industrial application of eNORR.21–23

Single-atom catalysts (SACs),24–26 with their unique physical and chemical properties, can effectively address the issues of selectivity and activity in various electrocatalytic reactions, including, but not limited to CO2 reduction,27,28 N2 reduction,4,29 oxygen reduction,30,31 HER,25 and eNORR.4,29 Among various types of SACs, metal–nitrogen–carbon (M–N–C) have been extensively studied and mass-produced due to their excellent electrical conductivity and large specific surface area.32,33 However, synthesizing M–N–C typically involves pyrolyzing precursors containing carbon, nitrogen, and metal at high temperatures (700–1100 °C).34 This process can make catalysts structurally non-uniform, complicating the precision adjustment of catalysts and understanding the catalytic mechanism.35 Fortunately, recent studies show that new two-dimensional conductive metal–organic frameworks (cMOFs) offer a promising solution.36–38 They have uniformly distributed binding sites that “lock” single atoms, exhibit high structural tunability, and are synthesized typically by hydrothermal reaction.30,39–41 This method helps prevent metal agglomeration and facilitates the rational design of the single atom's coordination environment. Therefore, cMOFs attract many researchers to explore their catalytic properties in various reactions.30,42–46 For example, Dincă et al. synthesized Ni-based 2, 3, 6, 7, 10, 11-hexaiminotriphenylene (NiN4–HTP), which showed high catalytic performance for oxygen reduction, comparable to the best non-Pt ORR catalysts (onset potential up to 0.82 V).45 Song et al. developed Co-doped CuO4–HTPs, which demonstrated outstanding eNORR performance (NH3 production rate of 299.9 μmol h−1 cm−2 and a faradaic efficiency of 96.4%).47 Many 3d metal-based cMOFs have been successfully synthesized.47–58 Although some of them show good catalytic performance in eNORR,47 the reaction mechanism remains unclear. Additionally, two intriguing questions have been raised regarding the system: (1) can we obtain or design better cMOFs-based catalysts? (2) Can we quickly predict their catalytic performance without extensive calculations? Owing to cMOFs' excellent structural controllability, systematic and atomic-level design and mechanism studies of the catalysts are expected to solve these questions and are crucial for developing highly active and selective SACs for eNORR.

In this study, 32 TMX4–HTP monolayers are constructed by changing the transition metal (TM) atoms (TM = Sc–Ni) and the coordination atoms X (X = N, O, P, S) around TM. 8 candidates are identified with high stability, selectivity, and activity using a four-step screening criterion. The NO activation mechanism and the potential-limited steps (PLS) behavior are elucidated through electronic structure analysis. Comparison of the PLS of 8 candidates reveals that MnO4–HTP is the best-performing eNORR catalyst with a relatively low potential of −0.21 V via the N-terminal mechanism. The excellent catalytic activity of MnO4–HTP is verified by simulating the electrochemical environment, considering the solvent effect, protonation ability, and constant potential model. Additionally, descriptors found without additional calculations could serve as an effective and fast screening strategy for eNORR catalysts.

2. Computational methods

All spin-polarized density functional theory calculations are performed by using the Vienna Ab initio Simulation Package (VASP).59 The ion–electron interaction is treated with the projector augmented-wave (PAW) method.60 The generalized gradient approximation (GGA)61 with the Perdew–Burke–Ernzerhof (PBE)62 functional is used to describe the electronic exchange–correlation interactions and a plane-wave cutoff energy of 450 eV is adopted.63 The van der Waals (vdW) interactions are accounted for through Grimme's DFT-D3 correction method to amend dispersion effects.62 The geometric optimization convergence threshold of the largest atoms is set to be 0.01 eV Å−1 and self-constant accuracy is set to be 10−5 eV, respectively. The gamma-centered mesh of 3 × 3 × 1 samples the first Brillouin zone. The vacuum space of 12 Å is used to avoid the interlayer interaction in the z-direction. Additionally, the effects of 15, 20, and 30 Å vacuum spaces on system energy were tested (see Table S1† for detailed data). It is found that a 12 Å vacuum space is sufficient for this system. The climbing-image nudged elastic band (CI-NEB) method is employed to locate the transition states (TS).64,65 To gain insights into the local orbital interactions, the crystal orbital Hamilton population (COHP) analysis is performed using the LOBSTER package.66Ab initio molecular dynamics (AIMD) simulations are carried out via the Nose–Hoover thermostat using the canonical ensemble (NVT) at 350 K, with a time step of 0.5 fs, and a total of 10 ps standard AIMD simulations are performed. The implicit solvation model subsequently computed is considered to be implemented in VASPsol, using the dielectric constant of water as 78.4.67,68

The free energy change (ΔG) of electrochemical reactions in eNORR is computed by:

ΔG = ΔEDFT + ΔEZPE − TΔS + eU
where ΔEDFT represents the change in DFT total energy for the species involved, T is system temperature and set to be 298.15 K, and ΔEZPE and ΔS refer to the zero-point energy change and entropy change, respectively.69 The effect of electrode potential (U) is treated as eU (e represents the elementary charge) and the total energy of (H+ + e−) is equal to that of 1/2H2 under the standard electrode condition based on the computational hydrogen electrode (CHE) method.70 In the CPM,71 potential-dependent energies are calculated by simulating system charge from −2e− to 2e− in a step of 0.5e− within the implicit solvent framework (detailed definitions in the ESI†).

The formation energy (Ef) and dissolution potential (Udiss) are defined as:72

Ef = (Etotal − Esubstrate − 3μTM)/3

image file: d4ta04795h-t1.tif

E total and Esubstrate are the total energies of the TMX4–HTP system and X4–HTP substrate, respectively. μTM is the total energy of a metal atom in its most stable bulk structure. The TMX4–HTP unit cell contains 3 TM atoms (complete unit cell structure diagram see Fig. S1†). image file: d4ta04795h-t2.tif is the standard dissolution potential of bulk metal, and n is the number of electrons involved during the dissolution process. Since μTM is referenced for its bulk metal, systems with negative values of Ef are evaluated to be thermodynamically stable against the clustering of TM atoms. Systems with positive values of Udissvs. standard hydrogen electrode (SHE) are considered to be electrochemically stable. Udiss > 0 V vs. SHE is deemed electrochemically stable.

The limiting potential (UL) for the eNORR process is calculated using:

UL = −ΔGmax/e
where ΔGmax is the maximum free energy change among all the elementary steps, and this step is defined as the PLS.

The current density (jk(U)) can be written as:73

image file: d4ta04795h-t3.tif
where NA is the density of activity sites on the catalyst, and k0 is the prefactor determined to be 200 s per site.74

3. Results and discussion

3.1 Screening catalysts

As shown in Fig. 1a (left panel), the 2D cMOF of TMX4–HTP is composed of in-plane TMX4 (TM = Sc–Ni; X: O, N, S, P) and conjugated organic parts. The details of the optimized structures and the corresponding structural parameters are given in Table S2.† In light of the eNORR process involving multi-intermediates' transformation, a universal four-step screening criterion for identifying promising candidates of such TMX4–HTPs for eNORR is proposed, as shown in the right panel of Fig. 1a.
image file: d4ta04795h-f1.tif
Fig. 1 (a) Schematic diagram of the TMX4–HTP structures for 32 candidates and the four-step screening strategy for efficient eNORR SACs. Formation energy, electronegativity (b), and dissolution potential (c) of TM atoms in TMX4–HTP systems. Hydrogen atoms are shown in white, X atoms surrounding the TM in red, carbon atoms in gray, and TM atoms in purple. 30 candidates were selected based on the thermodynamic (formation energy) and electrochemical (dissolution potential) stability.
Stability. To begin, the thermodynamic and electrochemical stability of TMX4–HTPs was assessed theoretically, involving the computation of Ef and Udiss, with their values listed in Table S3.† As shown in Fig. 1b, all catalysts can exhibit relatively high thermodynamic stability, with Ef ranging from −9.09 to −2.62 eV. The overall trend of Ef in TMX4–HTPs is related to the electronegativity (χ) of TM atoms, with Ef increasing as χ increases. Fig. 1c illustrates that most TMX4–HTPs are highly stable in electrochemical environments, with Udiss values ranging from 0.19 to 2.45 V. Only ScP4–HTP and TiP4–HTP, with potentials of −0.85 V and −0.04 V respectively, are prone to dissolution. As a result, subsequent calculations focus exclusively on the remaining 30 2D TMX4–HTP monolayers.
NO adsorption. Stable chemisorption of NO on catalysts is the initial step and prerequisite for efficient eNORR, crucial for subsequent protonation and reduction steps. Three typical adsorption modes, namely N-end, O-end, and side-on patterns, are investigated to determine the most stable adsorption configuration of NO on TMX4–HTPs, as illustrated in Fig. S2† and their adsorption free energy (ΔG*NO) are presented in Fig. 2a–d and Table S4.† All TMX4–HTPs prefer to capture NO in the N-end pattern, with ΔG*NO ranging from −0.09 to −3.59 eV, and an average of −1.51 eV, except for NiN4–HTP (ΔG*NO = 0.07 eV). Therefore, subsequent calculations only target the 29 2D TMX4–HTP monolayers that could effectively capture NO.
image file: d4ta04795h-f2.tif
Fig. 2 Free energies of *NO, *H, *H2O, *NH3 on (a) TMN4–HTPs, (b) TMO4–HTPs, (c) TMP4–HTPs, and (d) TMS4–HTPs. (e) Free energy diagram for the electrochemical NO-to-NH3 conversion on MnO4–HTP. (f) Energy profile from ab initio molecular dynamics simulation on MnO4–HTP after 10 ps at 350 K. Through the selectivity of TMX4–HTPs, 26 candidates were screened.
Selectivity. Under typical eNORR conditions, the competitive adsorption abilities of species in the reaction environment to the catalyst are crucial for overall catalytic performance. For example, the competitive adsorption ability between NO and protons is a criterion for assessing Faraday efficiency toward the target product NH3 or the byproduct H2. As shown in Fig. 2a–d and Table S5,† almost all candidate catalysts have a much stronger adsorption ability for NO over protons, indicating that HER is unfavorable. The other crucial factor is the adsorption strength between NO and the products (NH3 and H2O), which might influence the reaction rate of the entire eNORR. Therefore, systems with a stronger adsorption ability for the reactant NO compared to the products are selected as potential catalysts. Using the adsorption ability criteria of NO compared protons, NH3, and H2O in these systems, a total of 26 2D TMX4–HTPs are screened out.
Activity. To identify potential active catalysts, the UL should be smaller than that of benchmark platinum (Pt, |−0.5| V).75 In addition, using the same calculation settings as in this work, the Gibbs free energy diagram of eNORR on Pt(100) (see ESI and Fig. S3† for details) was calculated. Consistent with previous reports,1,23,75 the first protonation step is the PLS, with a free reaction energy difference of 0.55 eV (0.52 eV in ref. 75). In addition, the first protonation step is reported to be the PLS for most investigated eNORR catalysts.1,4,23,29,76 Therefore, ΔG1 was primarily computed as a screening criterion (ΔG1 < 0.5 eV), as shown in Fig. S4 and Table S6.† As a result, 12 TMX4–HTPs are screened out as potential efficient catalyst candidates. Subsequently, possible reaction routes (Fig. S5†) for the conversion of NO to NH3 were computed, and the corresponding energetic data are summarized in Table S7.† Consequently, 8 models (FeN4–HTP, Mn(Fe, Co)O4–HTP, Ti(V, Fe, Co)S4–HTP) with UL less than |−0.5| V are selected as eNORR catalysts, and their free energy profiles are shown in Fig. S6.†
Thermodynamic stability. To further confirm the structural stability of these 8 candidates during the catalytic process, AIMD simulations are conducted as depicted in Fig. 2f and S7,† respectively. The results show that the total energy remains close to equilibrium, indicating that these structures are effectively preserved.

3.2 Exploring the origin of activity

To understand the activity origin of the 8 catalysts, both the NO activation mechanism and the PLS behavior were analyzed through electronic structure analysis. Since all the catalysts prefer the N-end adsorption pattern, we hypothesize that this system follows the “acceptance-donation” mechanism, as shown in Fig. 3a. This mechanism consists of both electron donation and backdonation: the donor mechanism occurs through the transfer of the lone pair of electrons in the σ orbital (a σ orbital on the N atom, σ2p in molecular orbital, Fig. S8†) of NO to the d orbital of TMX4–HTPs, forming d–σ hybrid orbitals; the reverse conjugated donor mechanism involves the transfer of electrons from the d orbitals of TM atoms to the π* orbitals of NO, forming d–π* hybrid orbitals. As a result, when NO is adsorbed on TM, the bond strength between TM–N is often higher than that of TM–O.
image file: d4ta04795h-f3.tif
Fig. 3 (a) Schematic interaction mechanism between TM and NO. (b) PDOS of Mn-3d orbitals in MnO4–HTP, PDOS of NO-2p orbitals in free NO molecule, and PDOS of Mn-3d and NO-2p orbitals during NO adsorption on MnO4–HTP, respectively. The Fermi level is set to 0 eV. (c) COHP of Mn–N in MnO4–HTP, COHP of a free NO molecule, and COHP of NO on MnO4–HTP, respectively. Clarifying the origin of NO activation through PDOS and COHP.

The interactions of d–π* and d–σ can be validated through the partial density of states (PDOS) and COHP analysis (Fig. 3b and c). Taking the MnO4–HTP system as an example, the π* orbitals of NO molecule interact with 3d orbitals of Mn atom upon NO adsorption. An obvious shift in the NO energy levels is observed, matching well with the energy levels of TM-d orbitals of the candidate materials, indicating a strong d–σ interaction. Similar cases can also be found in the remaining 7 catalysts. Interestingly, this mechanism can also be extended to all the NO chemisorbed systems. Fig. S9† shows the d orbital of TM before NO adsorption, while Fig. S10† shows the p orbital of NO and the d orbital of TM after NO adsorption. The integrated COHP (ICOHP) values, integrated up to the Fermi level, are −9.04 eV for free NO and −6.64 eV for adsorbed NO (Fig. 3c), validating activation of NO. Additionally, Bader charge analysis quantifies the charge transfer ability from the catalyst to the adsorbed NO, showing that NO gains electrons ranging from 0.04 to 0.51e− (Table S8†). Furthermore, there is a linear relationship (coefficient of determination: R2 ≥ 0.90) between the N–O bond length and the amount of transferred charge, as depicted in Fig. S11.†

After confirming the activation mechanism of NO molecules, the PLS behavior was investigated as it directly reflects the eNORR activity. As shown in Fig. S5,† NO* can either protonate into *HNO or alternatively to *NOH for the PLS. The computed result indicates *HNO is more favorable energetically for all the 8 TMX4–HTPs (Fig. S12 and Table S9†). As shown in Fig. S13,† still using MnO4–HTP, as an example, it is obvious that near the Fermi level, the hybridization of the s orbital of H with the p orbital of N (*HNO) is stronger than the hybridization of the s orbital of H with the p orbital of O (*NOH). This makes the adsorption of H on N more stable, explaining why *HNO forms more easily. The other 7 catalysts show the same trends (Fig. S13†). In addition, the electronic structure of MnO4–HTP using the hybrid functional HSE06 was calculated, which generally provides band structures in good agreement with experimental results. As shown in Fig. S14,† both HSE06 and PBE show similar electronic structure trends and conductive properties for the catalyst. Based on this result, the PBE functional is reliable for studying the electronic properties of this system.

3.3 Simulation and description of electrochemical conditions

Considering that the reaction occurs under electrochemical conditions, factors such as the solvent effect, protonation ability, and constant potential may impact the reaction. Additionally, simulated values like overpotential and polarization current density can be directly compared with experimental values. Moreover, using intrinsic property values to predict or screen potential catalysts without extensive calculations could aid experimental synthesis. Given the optimal UL of MnO4–HTP, it is chosen as an example of the 8 catalysts to explore electrochemical condition-dependent catalytic performance.
Solvent effect. Considering that eNORR primarily takes place in aqueous solutions, the impact of the solvent on the eNORR activity of TMX4–HTPs was examined. The eNORR free energy diagram of MnO4–HTP was reevaluated utilizing the implicit solvation model integrated in VASPsol. As depicted in Fig. 4a, the potential energy surface maintains a consistent trend and the reaction energy of PLS is the same both with and without the implicit solvation model, indicating its negligible impact. Subsequently, considering the transition metals have a strong correlation effect, corrected the overpotential for the MnO4–HTP system by carrying out the PBE + SOL + U approach. The U value of Mn used to be 3.18 eV, according to the previous publication.77 The findings of the PBE + SOL + U are presented in Fig. S15.† It has been determined that the overpotentials of eNORR for the MnO4–HTP should be revised to be 0.23 V (PBE + SOL: 0.21 V). These findings suggest that the PBE functional and PBE + SOL + U level calculations produce similar results for the eNORR overpotential, consistent with previous calculations.
image file: d4ta04795h-f4.tif
Fig. 4 (a) Free energy diagrams of electrochemical NO-to-NH3 conversion on MnO4–HTP using PBE + SOL. (b) Calculated activation energy of the PLS on MnO4–HTP. (c) Reaction intermediates of MnO4–HTP as a function of the applied electrode potential. (d) Free energy curves of MnO4–HTP under varying electrode potentials at pH = 7. (e) pH- and potential-dependent contour plot of the reaction energy for the first protonation step. (f) Comparison of DFT-calculated ΔG*NO values with SISSO-predicted values. The ΔG*NO is depicted as: image file: d4ta04795h-t4.tif Detailed data in the ESI.† The excellent performance of MnO4–HTP was further confirmed by electrochemical condition simulation. Additionally, the descriptors of competitive adsorption on TMX4–HTP were explored using the SISSO method.
Protonation ability. Additionally, to accommodate the transition in the protonation process of MnO4–HTP, the corresponding explicit model was employed. This model comprises 5 water molecules and 1 hydrated proton (H13O6+), forming a ring-like hydrogen bond network. This model's effectiveness has been extensively validated in various other investigations of 2D-material-supported metallic SAC systems and our previous work.12,78–82 The reaction energy was computed to be 0.22 eV, which is slightly than that of the without considering the solvent effect. For MnO4–HTP, the calculated PLS activation energy amounts to 0.29 eV, as illustrated in Fig. 4b. Therefore, it is safe to conclude that the protonation step would be accessible under mild reaction conditions.
Polarization current density simulation. To facilitate a more intuitive comparison of the eNORR performance between MnO4–HTP and Pt(100), the polarization current density using the Tafel equation was computed. The NA on Pt(100) is estimated to be 1.44 × 1014 sites per cm2 and on MnO4–HTP is estimated to be 7.21 × 1013 sites per cm2.75 As shown in Fig. S16,† the polarization current output on MnO4–HTP is much stronger than that of Pt(100), showing excellent eNORR performance. Therefore, the theoretically predicted MnO4–HTP has higher catalytic efficiency than Pt(100) and thus has great potential as an alternative catalyst to Pt for eNORR.
Constant potential model. The simple metal/vacuum model used in the constant charge method cannot fully represent the charged electrochemical interface because it ignores surface charge effects, introducing uncertainties in electrocatalysis, particularly for two-dimensional materials. To address these uncertainties, the eNORR activity of MnO4–HTP was re-evaluated using the constant-potential method (CPM) with the double reference method, considering the surface charge. The computed total energies of MnO4–HTP and its surface with adsorbed eNORR intermediates (*NO, *HNO, *HNOH, *NH, *NH2, and *NH3) as a function of the applied electrode potential (standard hydrogen electrode, SHE) are shown in Fig. 4c. Consistent with previous studies, the charge capacity depends on the potential of each intermediate.60 Fig. S17† demonstrates that as the potential becomes more negative, the number of surface charges increases. The impact of applied potential in the five-step protonation process on ΔG was assessed at pH = 1, pH = 7, and pH = 13, illustrated in Fig. 4d and S18.† At pH = 1 and pH = 7, MnO4–HTP shows UL values of −0.19 V and −0.27 V, respectively, with the PLS being the first protonation step, consistent with previous CHE results. In contrast, under strongly alkaline conditions (pH = 13), the PLS transitioned from the first protonation step to the second step.

To gain a deeper understanding of the varying pH values and electrode potential effects on eNORR, an isoline diagram was constructed. Fig. 4e and S19† show the first protonation free energy change (ΔG1) and the second protonation free energy change (ΔG2) of MnO4–HTP plotted as functions of pH and potential. The results indicate that as pH increases, PLS transitions from the first protonation step to the second. Additionally, both ΔG1 and ΔG2 increase with higher pH or lower applied potential. This occurs because the energy required for the first protonation step to approximate that of the subsequent protonation step, and the alkaline conditions make eNORR protonation more challenging.

Descriptors of competitive adsorption. Relevant descriptors are explored to reveal the relationship between structural properties and adsorption competition. Employing the sure independence screening and sparsifying operator (SISSO) method,83 input features include some structural and electronic properties of TMX4–HTP, specific methods and detailed data are shown in ESI and Table S10.† The best descriptors for ΔG*NO, ΔG*H, ΔG*H2O, and ΔG*NH3 are identified from a pool of 5.84 × 109 one-dimensional descriptors, as outlined in ESI.† The predicted ΔG*NO (SISSO) values in the training set, shown in Fig. 4f, align with the DFT calculation results, showing a high R2 value of 0.94 and a low root mean square error (RMSE) of 0.19 eV. The predicted ΔG*NO values on TMX4–HTPs mostly fall within the range of −0.09 to −3.59 eV. Moreover, the RMSEs of ΔG*H2O (SISSO), ΔG*H (SISSO), and ΔG*NH3 (SISSO) are 0.23, 0.26, and 0.29 eV, respectively (Fig. S20†). More importantly, the descriptors may have the potential to be extended to other systems (Fig. S21†). For example, the adsorption competition can be predicted well using these descriptors in the cMOFs with different TM coordination environments (NiOxN4–x–HTP, 0 < x < 4, and NiO2S2–HTP), cMOFs with 4d transition metal (PdO4–HTP), and M–N–C (FeN4) (the structural data of FeN4 is shown in Fig. S22†). These descriptors enable a rapid assessment of the competition between eNORR reactants and products, allowing the reaction's feasibility to be evaluated without extensive computation. This approach facilitates the quick identification of high-quality eNORR catalysts.
Discussion with experimental results. Because CuO4–HTP, CoO4–HTP, and CuO4–HTP doped with Co (with Cu0.5Co0.5O4–HTP being the best) have been investigated as effective eNORR catalysts,47 it is important to compare and discuss our findings with the experimental results. Therefore, additional calculations were performed for the CuO4–HTP and Cu0.5Co0.5O4–HTP models, and the computed energy profiles are summarized in Fig. S23.† The computed UL values are −0.34 and −0.38 V for Cu0.5Co0.5O4–HTP and CuO4–HTP, respectively. This computed activity trend agrees well with experimental findings, showing that Cu0.5Co0.5O4–HTP has the best activity, followed by CuO4–HTP and CoO4–HTP. In all models, the first step protonation usually acts as the PLS via the *HNO moiety. Especially if we use Cu0.5Co0.5O4–HTP as the benchmark, three proposed catalysts: MnO4–HTP (UL = −0.21 V), FeN4–HTP (UL = −0.30 V), FeO4–HTP (UL = −0.32 V), are expected to have better or comparable catalytic performance. Although doping with heteronuclear TMs could affect the catalytic activity of neighboring TMs, this methodology is beyond the scope of the current work and will be discussed in ongoing research. More importantly, among the three catalysts, FeN4–HTP and FeO4–HTP have been successfully synthesized experimentally. The MnO4–HTP (its 3D configuration was successfully prepared experimentally53), with its more negative Ef and thermodynamic stability (AIMD), is likely to be synthesized experimentally in the near future.

4. Conclusions

This study uses first-principles calculations to investigate the electrocatalytic activity of structure-tunable TMX4–HTPs SACs to reduce nitric oxide to ammonia. Using criteria i.e., catalyst stability, NO capturing ability, product selectivity, and catalytic activity, 8 potential eNORR catalysts (FeN4–HTP, Mn(Fe, Co)O4–HTP, Ti(V, Fe, Co)S4–HTP) are identified with high selectivity toward ammonia and overpotential less than 0.5 V. AIMD simulations were used to evaluate the stability of these candidates further. Electronic structure analysis reveals the high activity of the chemisorbed NO, which follows the “acceptance-donation” mechanism, facilitating the N-end pattern *NO protonation into *HNO. Additionally, using the most active catalyst MnO4–HTP (UL = – 0.21 V) as an example, factors such as solvent effects, protonation ability, and constant-potential conditions were considered. The simulated results further confirm its excellent performance. Additionally, the computed polarization current density of MnO4–HTP is also significantly better than that of the benchmark Pt(100). Importantly, descriptors obtained without extensive calculations are found for identifying key adsorbates' adsorption ability, enabling a preliminary assessment of potential catalysts. Interestingly, when comparing these 8 potential eNORR catalysts with the experimentally reported Cu0.5Co0.5O4–HTP, 3 catalysts still perform better than Cu0.5Co0.5O4–HTP, and the experimentally reported systems also follow the first step protonation acts as the PLS via the *HNO moiety. This work investigates structurally tunable cMOFs for enhancing eNORR in ammonia synthesis but also provides a four-step screening method with descriptors to quickly identify candidates and their adsorption capacities, paving the way for highly active and selective electrocatalysts for eNORR.

Data availability

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Author contributions

Xing Chen: conceptualization, investigation, analysis, methodology, writing – original draft. Xiangyu Zhu: methodology, writing. Zhiyuan Xia: methodology. Shiting Qian: methodology. Yanan Zhou: structure analysis. Qiquan Luo: conceptualization, writing – review & editing, supervision, funding acquisition, project administration, resources. Jinlong Yang: project administration, writing – review & editing. Xing Chen and Xiangyu Zhu contributed equally to this work.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

We are grateful to Dr Chang Xu for valuable discussions. This work was supported by the National Key R&D Program of China (2019YFA0210004), and National Natural Science Foundation of China (Grants No. 22373001, 22288201, and 22102167). The authors would like to acknowledge the Supercomputer Center of the Anhui University of China, the University of Science and Technology of China, and the Shanghai Supercomputer Center of China for Scientific Computing for supercomputer access.

References

  1. Y. Zang, Q. Wu, S. Wang, B. Huang, Y. Dai and Y. Ma, J. Phys. Chem. Lett., 2022, 13, 527–535 CrossRef CAS PubMed .
  2. S. S. Markandaraj, D. Dhanabal and S. Shanmugam, J. Mater. Chem. A, 2023, 11, 23479–23488 RSC .
  3. Y. Xiao, C. Shen and W. B. Zhang, Mater. Today Chem., 2022, 25, 100958 CrossRef CAS .
  4. H. Niu, Z. Zhang, X. Wang, X. Wan, C. Kuai and Y. Guo, Small, 2021, 17, e2102396 CrossRef PubMed .
  5. J. Long, S. Chen, Y. Zhang, C. Guo, X. Fu, D. Deng and J. Xiao, Angew Chem. Int. Ed. Engl., 2020, 59, 9711–9718 CrossRef CAS PubMed .
  6. P. Guo, D. Luan, H. Li, L. Li, S. Yang and J. Xiao, J. Am. Chem. Soc., 2024, 146, 13974–13982 CrossRef CAS PubMed .
  7. Y. Shan, J. Du, Y. Zhang, W. Shan, X. Shi, Y. Yu, R. Zhang, X. Meng, F. S. Xiao and H. He, Natl. Sci. Rev., 2021, 8, nwab010 CrossRef CAS PubMed .
  8. J. Shao, H. Jing, P. Wei, X. Fu, L. Pang, Y. Song, K. Ye, M. Li, L. Jiang, J. Ma, R. Li, R. Si, Z. Peng, G. Wang and J. Xiao, Nat. Energy, 2023, 8, 1273–1283 CrossRef CAS .
  9. Y. Lin, Y. Feng, H. Zhou, Y. Han, H. Sun, L. Shi, L. Meng, M. Zhou, Y. Liu and X. Zhang, Appl. Surf. Sci., 2022, 593, 153338 CrossRef CAS .
  10. Y. Lin, L. Shi, Y. Chen, X. Yao, L. Meng, Y. Han, X. Zhao, M. He, Y. Liu and X. Zhang, Chemistry, 2023, 29, e202202925 CrossRef CAS PubMed .
  11. C. He, R. Sun, L. Fu, J. Huo, C. Zhao, X. Li, Y. Song and S. Wang, Chin. Chem. Lett., 2022, 33, 527–532 CrossRef .
  12. L. Zhou, X. Chen, M. He, X. Zhu, Y. Zhou, Z.-H. Lu and Q. Luo, FlatChem, 2023, 39, 100507 CrossRef CAS .
  13. Q. Wu, W. Wei, X. Lv, Y. Wang, B. Huang and Y. Dai, J. Phys. Chem. C, 2019, 123, 31043–31049 CrossRef CAS .
  14. J.-X. Liu, D. Richards, N. Singh and B. R. Goldsmith, ACS Catal., 2019, 9, 7052–7064 CrossRef CAS .
  15. W. Qu, X. Wang, Y. Sun, L. Dang, X. Wang and K. Chu, New J. Chem., 2023, 47, 8960–8964 RSC .
  16. C. A. Farberow, J. A. Dumesic and M. Mavrikakis, ACS Catal., 2014, 4, 3307–3319 CrossRef CAS .
  17. I. Katsounaros, M. C. Figueiredo, X. Chen, F. Calle-Vallejo and M. T. M. Koper, ACS Catal., 2017, 7, 4660–4667 CrossRef CAS .
  18. S. Zhou, B. Varughese, B. Eichhorn, G. Jackson and K. McIlwrath, Angew Chem. Int. Ed. Engl., 2005, 44, 4539–4543 CrossRef CAS PubMed .
  19. J. Choi, H.-L. Du, C. K. Nguyen, B. H. R. Suryanto, A. N. Simonov and D. R. MacFarlane, ACS Energy Lett., 2020, 5, 2095–2097 CrossRef CAS .
  20. Y. Li, C. Cheng, S. Han, Y. Huang, X. Du, B. Zhang and Y. Yu, ACS Energy Lett., 2022, 7, 1187–1194 CrossRef CAS .
  21. Y. Zhao, Q.-K. Li, C.-L. Chi, S.-S. Gao, S.-L. Tang and X.-B. Chen, J. Mater. Chem. A, 2024, 12, 8226–8235 RSC .
  22. X. Sun, J. Zheng, Z. Yao, S. Deng, Z. Pan, S. Wang and J. Wang, J. Phys. Chem. C, 2022, 126, 17598–17607 CrossRef CAS .
  23. Y. Xiao and C. Shen, Small, 2021, 17, e2100776 CrossRef PubMed .
  24. W. Zhang, Q. Fu, Q. Luo, L. Sheng and J. Yang, JACS Au, 2021, 1, 2130–2145 CrossRef CAS PubMed .
  25. M. Ren, X. Zhu, Q. Luo, X. Li and J. Yang, Nanoscale, 2023, 15, 10661–10666 RSC .
  26. B. Qiao, A. Wang, X. Yang, L. F. Allard, Z. Jiang, Y. Cui, J. Liu, J. Li and T. Zhang, Nat. Chem., 2011, 3, 634–641 CrossRef CAS PubMed .
  27. M. Wang, Y. Yao, Y. Tian, Y. Yuan, L. Wang, F. Yang, J. Ren, X. Hu, F. Wu, S. Zhang, J. Wu and J. Lu, Adv. Mater., 2023, 35, e2210658 CrossRef PubMed .
  28. S. Shu, T. Song, C. Wang, H. Dai and L. Duan, Angew Chem. Int. Ed. Engl., 2024, e202405650,  DOI:10.1002/anie.202405650 .
  29. Y. Zang, Q. Wu, S. Wang, B. Huang, Y. Dai and Y. Ma, Mater. Horiz., 2023, 10, 2160–2168 RSC .
  30. Y. Zhou, L. Sheng, Q. Luo, W. Zhang and J. Yang, J. Phys. Chem. Lett., 2021, 12, 11652–11658 CrossRef CAS PubMed .
  31. X. Zhu, M. He, X. Chen, Y. Zhou, C. Xu, X. Li, Q. Luo and J. Yang, J. Phys. Chem. Lett., 2024, 15, 2815–2824 CrossRef CAS PubMed .
  32. P. Li, Y. Jiao, Y. Ruan, H. Fei, Y. Men, C. Guo, Y. Wu and S. Chen, Nat. Commun., 2023, 14, 6936 CrossRef CAS PubMed .
  33. L.-J. Yuan, X.-L. Sui, H. Pan and Z.-B. Wang, Renewables, 2023, 1, 514–540 CrossRef .
  34. X. Qu, Y. Han, Y. Chen, J. Lin, G. Li, J. Yang, Y. Jiang and S. Sun, Appl. Catal., B, 2021, 295, 120311 CrossRef CAS .
  35. E. Sargeant, P. Rodriguez and F. Calle-Vallejo, ACS Catal., 2024, 8814–8822,  DOI:10.1021/acscatal.4c00727 .
  36. L. Wang, A. Sarkar, G. L. Grocke, D. W. Laorenza, B. Cheng, A. Ritchhart, A. S. Filatov, S. N. Patel, L. Gagliardi and J. S. Anderson, J. Am. Chem. Soc., 2023, 145, 8486–8497 CrossRef CAS PubMed .
  37. C. Zhang, H. Yang, D. Zhong, Y. Xu, Y. Wang, Q. Yuan, Z. Liang, B. Wang, W. Zhang, H. Zheng, T. Cheng and R. Cao, J. Mater. Chem. A, 2020, 8, 9536–9544 RSC .
  38. J. Wang, X. Zhou, Y. Wang and Y. Li, J. Phys. Chem. C, 2023, 127, 14733–14740 CrossRef CAS .
  39. H. Huang, Y. Zhao, Y. Bai, F. Li, Y. Zhang and Y. Chen, Adv. Sci., 2020, 7, 2000012 CrossRef CAS PubMed .
  40. M. Wang, R. Dong and X. Feng, Chem. Soc. Rev., 2021, 50, 2764–2793 RSC .
  41. Y. Gao, J. Wang, Y. Yang, J. Wang, C. Zhang, X. Wang and J. Yao, Nano-Micro Lett., 2023, 15, 158 CrossRef CAS PubMed .
  42. R. Dong, M. Pfeffermann, H. Liang, Z. Zheng, X. Zhu, J. Zhang and X. Feng, Angew Chem. Int. Ed. Engl., 2015, 54, 12058–12063 CrossRef CAS PubMed .
  43. X. Huang, S. Zhang, L. Liu, L. Yu, G. Chen, W. Xu and D. Zhu, Angew Chem. Int. Ed. Engl., 2018, 57, 146–150 CrossRef CAS PubMed .
  44. A. J. Clough, J. W. Yoo, M. H. Mecklenburg and S. C. Marinescu, J. Am. Chem. Soc., 2015, 137, 118–121 CrossRef CAS PubMed .
  45. E. M. Miner, T. Fukushima, D. Sheberla, L. Sun, Y. Surendranath and M. Dincă, Nat. Commun., 2016, 7, 10942 CrossRef CAS PubMed .
  46. H. Zhong, M. Ghorbani-Asl, K. H. Ly, J. Zhang, J. Ge, M. Wang, Z. Liao, D. Makarov, E. Zschech, E. Brunner, I. M. Weidinger, J. Zhang, A. V. Krasheninnikov, S. Kaskel, R. Dong and X. Feng, Nat. Commun., 2020, 11, 1409 CrossRef CAS PubMed .
  47. P. Liu, J. Yan, H. Huang and W. Song, Chem. Eng. J., 2023, 466, 143134 CrossRef CAS .
  48. F. B. Ajdari, E. Kowsari, M. N. Shahrak, A. Ehsani, Z. Kiaei, H. Torkzaban, M. Ershadi, S. K. Eshkalak, V. Haddadi-Asl, A. Chinnappan and S. Ramakrishna, Coord. Chem. Rev., 2020, 422, 213441 CrossRef .
  49. A. J. Clough, J. M. Skelton, C. A. Downes, A. A. de la Rosa, J. W. Yoo, A. Walsh, B. C. Melot and S. C. Marinescu, J. Am. Chem. Soc., 2017, 139, 10863–10867 CrossRef CAS PubMed .
  50. A. J. Clough, J. W. Yoo, M. H. Mecklenburg and S. C. Marinescu, J. Am. Chem. Soc., 2014, 137, 118–121 CrossRef PubMed .
  51. R. Dong, P. Han, H. Arora, M. Ballabio, M. Karakus, Z. Zhang, C. Shekhar, P. Adler, P. S. Petkov, A. Erbe, S. C. B. Mannsfeld, C. Felser, T. Heine, M. Bonn, X. Feng and E. Cánovas, Nat. Mater., 2018, 17, 1027–1032 CrossRef CAS PubMed .
  52. R. Dong, M. Pfeffermann, H. Liang, Z. Zheng, X. Zhu, J. Zhang and X. Feng, Angew. Chem., Int. Ed., 2015, 54, 12058–12063 CrossRef CAS PubMed .
  53. S. Feng, M. Zhang, Y. Ma, X. Ding, T. Yan, Y. Wu, W. Huang, L. Zhang, H. Ye, Y. Ji, Y. Li and Y. Li, Adv. Mater., 2023, 35, 2307736 CrossRef CAS PubMed .
  54. Y. Gao, J. Wang, Y. Yang, J. Wang, C. Zhang, X. Wang and J. Yao, Nano-Micro Lett., 2023, 15, 158 CrossRef CAS PubMed .
  55. M. Hmadeh, Z. Lu, Z. Liu, F. Gándara, H. Furukawa, S. Wan, V. Augustyn, R. Chang, L. Liao, F. Zhou, E. Perre, V. Ozolins, K. Suenaga, X. Duan, B. Dunn, Y. Yamamto, O. Terasaki and O. M. Yaghi, Chem. Mater., 2012, 24, 3511–3513 CrossRef CAS .
  56. N. Hoshino and T. Akutagawa, Chem.–Eur. J., 2018, 24, 19323–19331 CrossRef CAS PubMed .
  57. R. Iqbal, M. Q. Sultan, S. Hussain, M. Hamza, A. Tariq, M. B. Akbar, Y. Ma and L. Zhi, Adv. Mater. Technol., 2021, 6, 2000941 CrossRef CAS .
  58. S. Wang, F. Huang, Z. Zhang, W. Cai, Y. Jie, S. Wang, P. Yan, S. Jiao and R. Cao, J. Energy Chem., 2021, 63, 336–343 CrossRef CAS .
  59. G. Kresse and J. Hafner, Phys. Rev. B: Condens. Matter Mater. Phys., 1993, 48, 13115–13118 CrossRef CAS PubMed .
  60. P. E. Blochl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 17953–17979 CrossRef PubMed .
  61. J. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77(18), 3865–3868 CrossRef CAS PubMed .
  62. S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799 CrossRef CAS PubMed .
  63. X. Chen, J. Sun, J. Guan, J. Ji, M. Zhou, L. Meng, M. Chen, W. Zhou, Y. Liu and X. Zhang, Int. J. Hydrogen Energy, 2022, 47, 23191–23200 CrossRef CAS .
  64. G. Henkelman and H. Jónsson, J. Chem. Phys., 2000, 113, 9978–9985 CrossRef CAS .
  65. G. Henkelman, B. P. Uberuaga and H. Jónsson, J. Chem. Phys., 2000, 113, 9901–9904 CrossRef CAS .
  66. R. Dronskowski and P. E. Bloechl, J. Phys. Chem., 1993, 97, 8617–8624 CrossRef CAS .
  67. K. Mathew, R. Sundararaman, K. Letchworth-Weaver, T. A. Arias and R. G. Hennig, J. Chem. Phys., 2014, 140, 9519–9825 CrossRef PubMed .
  68. M. Fishman, H. L. Zhuang, K. Mathew, W. Dirschka and R. G. Hennig, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 87, 245402 CrossRef .
  69. X. Chen, Y. Liu, W. Gao, G. Yang, A. H. Qureshi, M. Chen, X. Yao, M. Zhou, X. Zhang and Y. Liu, J. Phys. Chem. C, 2021, 125, 15292–15300 CrossRef CAS .
  70. B. Hammer and J. K. Norskov, in Advances in Catalysis, Vol 45: Impact of Surface Science on Catalysis, ed. B. C. Gates and H. Knozinger, 2000, vol. 45, pp. 71–129 Search PubMed .
  71. Z. Duan and G. Henkelman, Langmuir, 2018, 34, 15268–15275 CrossRef CAS PubMed .
  72. X. Guo, J. Gu, S. Lin, S. Zhang, Z. Chen and S. Huang, J. Am. Chem. Soc., 2020, 142, 5709–5721 CrossRef CAS PubMed .
  73. X. Huang, J. Wang, J. Gao, Z. Zhang, L. Y. Gan and H. Xu, ACS Appl. Mater. Interfaces, 2021, 13, 17075–17084 CrossRef CAS PubMed .
  74. J. K. Norskov, T. Bligaard and A. Logadottir, J. Electrochem. Soc., 2005, 152, J23–J26 CrossRef CAS .
  75. H.-J. Chun, V. Apaja, A. Clayborne, K. Honkala and J. Greeley, ACS Catal., 2017, 7, 3869–3882 CrossRef CAS .
  76. Y. Zang, Q. Wu, S. Wang, B. Huang, Y. Dai and Y. Ma, Phys. Rev. Appl., 2023, 19, 024003 CrossRef CAS .
  77. D. Li, A. Zhang, Z. Feng and W. Wang, ACS Appl. Mater. Interfaces, 2024, 16, 5779–5791 CrossRef CAS PubMed .
  78. M. He, X. Chen, Y. Zhou, C. Xu, X. Li, Q. Luo and J. Yang, J. Phys. Chem. Lett., 2023, 14, 7100–7107 CrossRef CAS PubMed .
  79. L. Li, J. M. P. Martirez and E. A. Carter, ACS Catal., 2020, 10, 12841–12857 CrossRef CAS .
  80. Y. Wang, Y. J. Tang and K. Zhou, J. Am. Chem. Soc., 2019, 141, 14115–14119 CrossRef CAS PubMed .
  81. M. Zafari, M. Umer, A. S. Nissimagoudar, R. Anand, M. Ha, S. Umer, G. Lee and K. S. Kim, J. Phys. Chem. Lett., 2022, 13, 4530–4537 CrossRef CAS PubMed .
  82. X. Li, Q. Zhou, S. Wang, Y. Li, Y. Liu, Q. Gao and Q. Wu, J. Phys. Chem. C, 2021, 125, 11963–11974 CrossRef CAS .
  83. R. Ouyang, S. Curtarolo, E. Ahmetcik, M. Scheffler and L. M. Ghiringhelli, Phys. Rev. Mater., 2018, 2, 083802 CrossRef CAS .

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta04795h
‡ These authors contribute equally to the work.
This journal is © The Royal Society of Chemistry 2024