betaFeedback

WorksPhysics and Astronomy

Metal-Coordination Effects on the Stability and ORR/OER Activity of Layered Organometallic Single-Atom Catalysts: A Theoretical Study

Abstract

Organometallic layered materials have emerged as promising single-atom catalysts for oxygen reduction and evolution reactions, but their practical use has been limited by insufficient electrochemical stability. Here, we present a density functional theory study clarifying the relationship between catalytic activity and stability in organometallic single-atom catalysts with metal-N4_4 (MN4_4) and metal-O4_4 (MO4_4) coordination. We compare graphene-embedded MN4_4 motif and phthalocyanine-like frameworks with MO4_4-coordination frameworks, including M4_4(OHPTP)2_2 and M3_3(HHTP)2_2 (M = Mn, Fe, Co, Ni, Cu, Zn). Stability is assessed by surface Pourbaix analysis, while activity is evaluated using the computational hydrogen electrode method. MN4_4 systems show competitive overpotentials but suffer strong pH-dependent instability. In contrast, MO4_4 frameworks exhibit enhanced robustness across wide pH ranges while maintaining good catalytic performance. A proposed stability descriptor enables direct comparison across systems, identifying MO4_4 coordination structures, particularly M4_4(OHPTP)2_2 (M = Zn, Co) as optimal for balancing activity and stability in practical electrocatalysis.

Introduction

The development of sustainable energy conversion technologies critically depends on efficient electrochemical processes, particularly the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER). A central challenge in this field is the replacement of noble-metal-based catalysts, which exhibit excellent activity but face significant limitations for large-scale applications due to their high cost and limited natural abundance.

Metal cations coordinated by pyridinic nitrogen sites embedded in graphene (G-MN4_4) have been widely reported as single-atom catalysts for both ORR and OER.[1, 2, 3] However, their catalytic performance is strongly metal dependent and highly sensitive to the electrolyte environment. For instance, G-MnN4_4 exhibits favorable ORR activity in alkaline media but shows markedly reduced performance under acidic conditions.[4, 5] Similarly, G-FeN4_4 display substantial variations in both activity and stability as a function of pH, driven by changes in reaction intermediates and local electronic structure.[6, 7] G-CoN4_4 and G-NiN4_4 structures also demonstrate catalytic activity; however, their operation is typically confined to narrow pH windows.[8] Collectively, these observations indicate that, despite their intrinsic activity, G-MN4_4 structures suffer from pronounced pH-dependent limitations that significantly restrict their practical applicability across diverse electrolyte environments.[9]

On the other hand, two-dimensional metal-organic frameworks (2D MOFs) have emerged as promising materials for electrocatalysis owing to their well-defined coordination environments, high density of active sites, and pronounced structural tunability.[10, 11, 12] Within this class, phthalocyanine-like MOFs (s-MPc) have been investigated as model systems for ORR and OER.[13] These materials feature a well-defined MN4_4 coordination motif and have demonstrated favorable intrinsic catalytic activity. Nevertheless, similar to their graphene-supported counterparts, they often exhibit limited structural stability under acidic and alkaline electrolyte conditions.

Another important family of 2D MOFs is represented by the general formula M3_3(ligand)2_2. These layered frameworks predominantly adopt metal bis(dioxolene) (MO4_4) and metal bis(diamine) (MN4_4) coordination motifs, which give rise to distinct electronic structures and catalytic properties. Early experimental studies demonstrated promising electrocatalytic performance in alkaline media, exemplified by Ni3_3(HITP)2_2 (HITP = 2,3,6,7,10,11-hexaaminotriphenylene) for the oxygen reduction reaction (ORR)[14] and Co3_3(HHTP)2_2 (HHTP = 2,3,6,7,10,11-hexahydroxytriphenylene) for the oxygen evolution reaction (OER).[15] Notably, these materials exhibit activities comparable to those of state-of-the-art precious-metal catalysts. Subsequent studies have further expanded their electrocatalytic scope,[16, 17, 18] encompassing ORR,[19, 20] OER,[21, 22] hydrogen evolution reaction (HER),[23, 24, 25, 26] as well as the electroreduction of CO2_2[27, 28] and N2_2.[29, 30]

Theoretical investigations have explored the stability, electronic structure, and catalytic performance of 2D MOFs with diverse coordination environments, including MS4_4,[31] MSe4_4,[32] MN4_4,[33, 34, 35] MO4_4,[36, 37, 38, 39] and mixed motifs.[40, 41, 42] Collectively, these studies identify the metal-ligand coordination environment as a key descriptor governing electronic structure, adsorption energetics, and catalytic activity. Nevertheless, electrochemical stability and degradation under realistic operating conditions remain relatively underexplored.[43, 44]

In this work, density functional theory (DFT) calculations are employed to investigate the electronic

Computational details

Spin-polarized density functional theory (DFT) calculations were performed using the Vienna Ab initio Simulation Package (VASP 6.5).[45] Geometry optimizations and electronic structure calculations were performed using the vdW-DF2 functional, which incorporates the GGA-type PW86 exchange functional.[46] Self-interaction effects in the 3d3d states of transition metals were treated using the Hubbard UU approach. The UU values assigned to Mn, Fe, Co, Ni, Cu, and Zn were set to 3.1, 3.3, 3.3, 3.4, 3.5, and 3.5 eV, respectively, based on prior studies.[47] This treatment is critical for accurately capturing adsorbate-metal interactions, which strongly depend on the metal dd-orbital character. A vacuum region of 15 Å was introduced along the zz direction for all unit-cell structures. All simulations employed a plane-wave energy cutoff of 400 eV, with Brillouin zone sampling performed using a 2×2×12 \times2 \times1 kk-point mesh. Full geometry optimization was performed until the residual forces on all atoms were below 0.025 eV/Å, with electronic self-consistency achieved at a convergence threshold of 10−510^{-5} eV. Data processing and visualization were performed using the VASPKIT[48] and VESTA[49] software packages, respectively.

Electrochemical stability under electrolyte conditions was assessed through surface Pourbaix analysis. The standard oxidation potential of the metal sites was estimated from the surface work function. This approach enables the construction of potential-pH stability diagrams, providing a systematic framework to evaluate how electrode potential and electrolyte conditions influence surface oxidation.[35] The thermodynamic overpotentials associated with the ORR and OER at the metal centers were evaluated using the computational hydrogen electrode (CHE) approach developed by Nørskov and co-workers. Within this formalism, the adsorption free energies of the key reaction intermediates OH*, O*, and OOH* were calculated as ΔG=ΔEDFT+EZPE−TΔS\Delta G = \Delta E_{\mathrm{DFT}} + E_{\mathrm{ZPE}} - T\Delta S. where ΔEDFT\Delta E_{\mathrm{DFT}} is the adsorption

energy obtained from DFT calculations, EZPEE_{\mathrm{ZPE}} is the zero-point energy correction, and TΔST\Delta S is the vibrational entropy contribution. The latter terms were derived from vibrational frequency calculations of the adsorbed intermediates, considering only their vibrational degrees of freedom while keeping the catalyst surface fixed. Thermodynamic and vibrational parameters for the gas-phase reference molecules were taken from the NIST database. To account for solvent effects, adsorption free energies were corrected for hydrogen-bonding interactions involving the OH* and OOH* intermediates, which are stabilized through solvation. These corrections were evaluated for each structure using the continuum dielectric model implemented in VASPsol.[51]

Results and Discussions

Structural and Electronic Properties

Figure 1 show the atomic structures of the metal-N4\mathrm{N}_4 and metal-O4\mathrm{O}_4 systems investigated in this work, represented within a 2×22 \times2 surface unit cell. Figures fig:1a and fig:1b correspond to the metal-N4\mathrm{N}_4 configurations, namely s-MPc and G-MN4\mathrm{MN}_4, whereas Figures fig:1c and fig:1d depict the metal-O4\mathrm{O}_4 systems M4(OHPTP)2\mathrm{M}_4(\mathrm{OHPTP})_2 and M3(HHTP)2\mathrm{M}_3(\mathrm{HHTP})_2, respectively. The stability of the single-layer structures was evaluated by calculating the binding energy of the metal center. Our results reveal that all systems exhibit substantial metal binding energies, ranging from -2.6 to -7.6 eV, largely governed by the metal species, indicating high structural stability arising from the interplay between ionic character and covalent orbital overlap. Among them, Cu and Zn display the weakest binding, consistent with their fully occupied 3d3d orbitals and indicative of comparatively lower thermodynamic stability. Table I summarizes the calculated structural and electronic properties of the metal-N4\mathrm{N}_4 and metal-O4\mathrm{O}_4 systems.

Atomic geometry of single-layer organometallic structures

Figure 1. Atomic geometry of single-layer organometallic structures in the 2×22 \times2 surface unit-cell representation. (a) s-MPc, (b) G-MN4_4, (c) M4_4(OHTPT)2_2, and (d) M3_3(HHTP)2_2, with M=Mn,Fe,Co,Ni,Cu,ZnM = \mathrm{Mn}, \mathrm{Fe}, \mathrm{Co}, \mathrm{Ni}, \mathrm{Cu}, \mathrm{Zn}.

SystemaadM−Xd_{\mathrm{M-X}}Φ\PhimmEbE_bEgE_g
G-MnN₄19.841.964.153.0-4.270.23
G-FeN₄19.841.954.162.0-5.280.22
G-CoN₄19.841.924.171.0-5.350.23
G-NiN₄19.841.914.200.0-5.670.22
G-CuN₄19.841.964.280.0-3.600.24
G-ZnN₄19.841.974.200.0-2.630.24
s-MnPc10.721.984.433.0-6.220.0
s-FePc10.701.975.202.0-7.580.0
s-CoPc10.681.955.041.0-7.340.0
s-NiPc10.621.925.180.0-7.360.25
s-CuPc10.681.975.241.0-5.640.22
s-ZnPc10.731.995.130.0-4.290.21
Mn₄(OHTPT)₂22.121.915.0112.8-6.060.0
Fe₄(OHTPT)₂22.041.885.408.5-6.010.0
Co₄(OHTPT)₂21.981.875.414.4-5.810.0
Ni₄(OHTPT)₂21.951.875.620.0-5.690.0
Cu₄(OHTPT)₂22.031.945.430.0-2.700.0
Zn₄(OHTPT)₂22.072.005.110.0-3.750.0
Mn₃(HHTP)₂21.541.894.989.0-7.160.0
Fe₃(HHTP)₂21.731.865.186.0-7.130.0
Co₃(HHTP)₂21.491.865.173.0-7.780.0
Ni₃(HHTP)₂21.441.855.020.0-6.570.0
Cu₃(HHTP)₂21.691.954.974.7-4.010.0
Zn₃(HHTP)₂21.722.004.832.0-3.180.0

Table I. Structural, electronic and magnetic properties of the metal-N₄ and metal-O₄ organometallic structures. Lattice constant (aa, in Å), M–X bond distance (X = N, O) (dM−Xd_{\mathrm{M-X}}, in Å), work function (Φ\Phi, in eV), total magnetic moment (mm, in μB\mu_{\mathrm{B}}), metal binding energy (EbE_b, in eV), and the electronic band gap (EgE_g, in eV).

We further computed the surface work function (Φ\Phi), defined as the minimum energy required to remove an electron from the material surface to the vacuum level. This parameter is relevant for assessing the oxidation potential of the metal center, particularly in light of the unpaired electrons predominantly localized at the metal site, as discussed below. We find that Φ\Phi for G-MN4\mathrm{MN}_4 is notably lower than in the other structures, suggesting a greater propensity for electron removal in this MN4\mathrm{MN}_4 coordination environment.

Figures S1 and S2 present the calculated band structures of the metal-N4\mathrm{N}_4 systems G-MN4\mathrm{MN}_4 and s-MPc. The G-MN4\mathrm{MN}_4 structures exhibit semiconducting behavior with nearly identical band-gap energies for all metal studied, approximately 0.23 eV. The systems G-MnN4\mathrm{N}_4, G-FeN4\mathrm{N}_4, and G-CoN4\mathrm{N}_4 display magnetic moments of 3, 2, and 1 μB\mu_{\mathrm{B}}, respectively. Their magnetization is reflected in a small spin splitting of the bands. In contrast, G-NiN4\mathrm{N}_4, G-CuN4\mathrm{N}_4, and G-ZnN4\mathrm{N}_4 are nonmagnetic. For the s-MPc structures, both semiconducting and metallic behaviors are observed depending on the metal atom. The systems s-MnPc, s-FePc, and s-CoPc exhibit metallic and magnetic character, with magnetic moments of 3, 2, and 1 μB\mu_{\mathrm{B}}, respectively. In contrast, s-NiPc, s-CuPc, and s-ZnPc are nonmagnetic semiconductors with band gaps of approximately 0.22 eV. The metallic behavior is attributed to significant spin splitting near the Fermi level.

Figures S3 and S4 present the calculated band structures of the metal-O4_4 systems M4_4(OHTPT)2_2 and M3_3(HHTP)2_2. Both MOFs exhibit metallic behavior, although their magnetic properties depend on the metal atom. The systems Ni4_4(OHTPT)2_2, Cu4_4(OHTPT)2_2, Zn4_4(OHTPT)2_2, and Ni3_3(HHTP)2_2 are nonmagnetic, whereas the remaining structures display finite magnetic moments. Notably, the magnetic systems exhibit relatively large magnetizations, which can be attributed to metal-oxygen hybridization. These systems also show fractional magnetic moments and significant spin polarization on the organic ligands, indicating a delocalized magnetic character. This behavior is consistent with the spin-density maps shown in Figures S5-S8 and with previous reports in the literature.[52] Table I summarizes the structural, electronic, and magnetic properties of metal-N4_4 and metal-O4_4 systems.

Eletrochemical Stability

The oxygen reduction and evolution reactions are intrinsically sluggish due to their multistep proton-coupled electron transfer mechanisms, which impose significant kinetic barriers and lead to substantial overpotentials. Consequently, the design of catalysts that can simultaneously enhance activity and maintain long-term stability remains a major challenge. This is particularly critical under harsh operating conditions, where acidic and alkaline environments can induce catalyst degradation through dissolution, surface reconstruction, or poisoning effects. The electrochemical stability of the layered organometallic structures in aqueous environments is evaluated through the construction of surface Pourbaix diagrams, which describe the dependence of the electrode potential on pH. Stability under ORR and OER conditions is determined by comparing the relative thermodynamic stability of distinct surface states via the following redox reactions[53]:

Mz++ze−⇌M,(1)\mathrm{M}^{z+} + z e^- \rightleftharpoons\mathrm{M}, \tag*{(1)}
M+H++e−⇌H∗,(2)\mathrm{M} + \mathrm{H}^{+} + e^- \rightleftharpoons\mathrm{H}^*, \tag*{(2)}
OH∗+H++e−⇌M+H2O,(3)\mathrm{OH}^* + \mathrm{H}^{+} + e^- \rightleftharpoons\mathrm{M} + \mathrm{H_2O}, \tag*{(3)}
O∗+2H++2e−⇌M+H2O.(4)\mathrm{O}^* + 2\mathrm{H}^{+} + 2e^- \rightleftharpoons\mathrm{M} + \mathrm{H_2O}. \tag*{(4)}

Here, zz denotes the number of electrons involved in the electrochemical process. Reaction (1) represents the oxidation of the metal center in the pristine system (M). Reactions (2)–(4) describe coupled proton-electron transfer processes associated with the adsorption of H, OH, and O species on the metal site, respectively. The standard electrode potentials, referenced to the standard hydrogen electrode (SHE), were evaluated within the computational hydrogen electrode (CHE) model.[50] The explicit consideration of H, OH, and O adsorbed species in surface Pourbaix diagrams captures the fundamental proton-electron coupled surface states that govern electrochemical stability in aqueous environments. Electrochemical stability can be further assessed by analyzing the energetic relationship between the OH* adsorption potential and the metal oxidation potential. This relative alignment provides a key descriptor for determining whether surface oxidation is thermodynamically favored. Under standard conditions (298 K) and applying the Nernst equation, the oxidation potential of the metal species can be expressed as:[54]

E(Mz+)=E∘(Mz+)−0.059zlog⁡a(Mz+),(5)E(\mathrm{M}^{z+}) = E^\circ(\mathrm{M}^{z+}) - \frac{0.059}{z}\log a(\mathrm{M}^{z+}), \tag*{(5)}

where EE denotes the electrode potential of the electrochemical cell, E∘E^\circ represents the standard oxidation potential, and a(Mz+)a(\mathrm{M}^{z+}) the activity of the dissolved metal ions. In the present work, the ionic activity is fixed at a(Mz+)=1×10−8 mol dm−3a(\mathrm{M}^{z+}) = 1 \times10^{-8}\ \mathrm{mol\ dm}^{-3} and the first oxidation potential (z=1z = 1) is assumed.

The first oxidation potential of the coordinated metal centers, E∘(M1+)E^\circ(\mathrm{M}^{1+}), was derived from the calculated electronic work functions of the organometallic surfaces, under the assumption that electron removal occurs primarily from unpaired electrons localized at the coordinated metal sites.[35, 38, 55] Specifically, E∘(M1+)E^\circ(\mathrm{M}^{1+}) was obtained by exploiting the linear relationship between metal work functions and their corresponding standard redox potentials, as reported by Li et al.[56] Furthermore, within the same theoretical framework, the oxidation potential associated with the key OH* intermediate can be expressed as

E(OH∗)=E∘(OH∗)−0.059×pH,(6)E(\mathrm{OH}^*) = E^\circ(\mathrm{OH}^*) - 0.059 \times\mathrm{pH}, \tag*{(6)}

where E∘(OH∗)=ΔGOH∗/eFE^\circ(\mathrm{OH}^*) = \Delta G_{\mathrm{OH}^*}/eF, with ΔGOH∗\Delta G_{\mathrm{OH}^*} representing the adsorption free energy of the OH* intermediate on the coordinated metal site and FF the Faraday constant.

According to Pourbaix diagrams, the electrochemical stability of organometallic surface catalysts can be assessed by examining the dependence of electrode potentials on pH. In this framework, both the electrode potential of the pristine surface, E(M1+)E(\mathrm{M}^{1+}), and that of the surface with OH adsorbates on the metal sites, E(OH∗)E(\mathrm{OH}^*), are considered. Comparing these potentials as a function of pH enables the evaluation of the stability of metal sites against dissolution, which occurs when E(M1+)E(\mathrm{M}^{1+}) is lower than E(OH∗)E(\mathrm{OH}^*) across the entire pH range.[35, 54]

Figure 2 schematically represent the relative alignment between E(M1+)E(\mathrm{M}^{1+}) and E(OH∗)E(\mathrm{OH}^*) in the surface Pourbaix diagram. We define α\alpha as the pH value at which these two potentials intersect. Therefore, from equations (5) and (6) we obtain the relation

Schematic representation of the electrode potential as a function of pH in a Pourbaix diagram

Figure 2. Schematic representation of the electrode potential as a function of the pH in a Pourbaix diagram. E(M1+)E(\mathrm{M}^{1+}) is the first oxidation potential for the metal center in the pristine surface. E(OH∗)E(\mathrm{OH}^{*}) is the electrode potential of the surface with OH adsorbate on the metal sites. α\alpha represents the pH at which these two potentials intersect.

α=D0.059 V+8,(7)\alpha= \frac{D}{0.059\ \mathrm{V}} + 8, \tag*{(7)}

where D=E∘(OH∗)−E∘(M1+)D = E^\circ(\mathrm{OH}^*) - E^\circ(\mathrm{M}^{1+}). Within this framework, the pH range 0≤α≤140 \le\alpha\le14 corresponds to $0.5 \le D \le 1.31.3 V. Specifically, for D<0.5D < 0.5 V, the systems are electrochemically stable, whereas for D>1.3D > 1.3 V they are unstable. For intermediate values, 0.5≤D≤1.30.5 \le D \le1.3 V, the systems exhibit partial stability depending on the pH. Therefore, we propose DD as a descriptor of electrochemical stability based on standard oxidation potentials, enabling direct comparison of stability trends across different organometallic materials.

Table 2 summarizes the calculated DD values for the metal-N4_4 and metal-O4_4 coordination motif considered in this study. For G-MN4_4, we find that only the system with M = Mn remains stable across the entire pH range. In contrast, for M = Fe, Cu, and Zn, the structures exhibit partial stability depending on the pH, whereas for M = Co and Ni they are unstable throughout. These findings are consistent with the Pourbaix diagrams presented in Figures S9. For the s-MPc structures, none of the metal centers display stability over the full pH range. Systems with M = Mn, Fe, Co, and Zn are only partially stable depending on the pH, while those with M = Ni and Cu are unstable, as illustrated in Figure S10.

SystemE0(OH∗)E^{0}(\mathrm{OH}^{*})E0(M+)E^{0}(\mathrm{M}^{+})DD
G-MnN4_4-0.55-0.850.30
G-FeN4_4-0.09-0.830.74
G-CoN4_41.12-0.821.94
G-NiN4_41.37-0.782.15
G-CuN4_40.50-0.671.17
G-ZnN4_40.14-0.780.92
s-MnPc0.24-0.480.72
s-FePc1.160.530.63
s-CoPc1.340.321.02
s-NiPc1.860.511.35
s-CuPc2.250.581.67
s-ZnPc1.460.441.02
Mn4_4(OHPTP)2_20.710.280.43
Fe4_4(OHPTP)2_21.720.790.93
Co4_4(OHPTP)2_20.750.81-0.06
Ni4_4(OHPTP)2_21.511.080.43
Cu4_4(OHPTP)2_20.010.83-0.82
Zn4_4(OHPTP)2_20.650.410.24
Mn3_3(HHTP)2_20.340.240.10
Fe3_3(HHTP)2_21.250.510.75
Co3_3(HHTP)2_2-0.300.49-0.79
Ni3_3(HHTP)2_21.620.301.33
Cu3_3(HHTP)2_20.960.230.73
Zn3_3(HHTP)2_20.440.050.39

Table 2. Standard oxidation potential of the coordinated metal center [E0(M+)E^{0}(\mathrm{M}^{+}) vs SHE, in V] and after the OH adsorption [E0(OH∗)E^{0}(\mathrm{OH}^{*}) vs SHE, in V], together with the electrochemical stability descriptor (DD in V).

In the case of metal-O4_4 structures, for M4_4(OHPTP)2_2 we observe that for M = Mn, Ni, Cu, Co, and Zn are stable across the entire pH range. The only exception is M = Fe, which is stable only for pH >8> 8, as shown in Figures S11. Finally, for the M3_3(HHTP)2_2 structures, only the Ni-based system is unstable. The Mn and Cu systems are stable for pH >4> 4, while Co and Zn systems remain stable across the entire pH range, as shown in Figure S12. Overall, a clear trend emerges: metal-N4_4 structures tend to be less electrochemically stable than their metal-O4_4 counterparts, suggesting that oxygen coordination enhances the electrochemical stability of the metal centers, suppressing metal oxidation over a broad pH range. In contrast, nitrogen-coordinated systems tend to exhibit either partial stability only above a critical pH, or complete instability, highlighting the influence of the metal-ligand chemical environment on electrochemical robustness.

We also note that other oxidizing species, such as O*, exhibit higher oxidation potentials than OH*, making the latter the primary adsorbate competing with the metal oxidation potential of the pristine surface in contact with the solution. Additionally, reduction via H* adsorption on the metal center is also taken into account in the Pourbaix diagrams shown in Figures S9-S11.

ORR and OER catalytic activity

The initial step of the ORR process involves the adsorption and activation of O2_2 at the catalyst active site. Therefore, we first analyze the stability and electronic properties associated with O2_2 adsorption on the metal centers of the metal-N4_4 and metal-O4_4 systems. The equilibrium geometries, binding energies, and spin-coupling characteristics are summarized in tbl:S1. Our results show that the binding energies for the metal-O4_4 centers range from -0.3 to -1.6 eV, indicating a moderate interaction between O2_2 and the metal sites that is strong enough to activate the molecule while avoiding overly strong adsorption. In contrast, the metal-N4_4 systems exhibit binding energies in the narrower range of −0.1-0.1 to −0.8-0.8 eV, shifted toward weaker O2_2 adsorption relative to their oxygen-coordinated counterparts, suggesting comparatively lower catalytic activity. Furthermore, systematic changes in the magnetic ground state are observed upon O2_2 adsorption. For M = Ni, Cu, and Zn, both metal-N4_4 and metal-O4_4 systems adopt high-spin configurations, with the adsorbed O2_2 stabilized in a triplet state. In contrast, for M = Mn, Fe, and Co, the systems favor low-spin configurations, with the adsorbed O2_2 stabilized in a singlet state.

Following O2_2 adsorption and activation, the catalytic activity toward the four-electron ORR pathway was evaluated within the computational hydrogen electrode (CHE) framework.[50] In this approach, the reaction leads to the formation of hydroxide ions as O2_2 + 2 H2_2O + 4e−^- ⟶\longrightarrow 4 OH−^-.[57] Accordingly, the ORR mechanism is described by a sequence of four elementary electrochemical steps involving the successive generation of OH−^- species.

O2∗+H2O(l)+e−⇌OOH∗+OH−,(8)\mathrm{O_2^* + H_2O}(l) + e^- \rightleftharpoons\mathrm{OOH^* + OH^-}, \tag*{(8)}
OOH∗+e−⇌O∗+OH−,(9)\mathrm{OOH^*} + e^- \rightleftharpoons\mathrm{O^* + OH^-}, \tag*{(9)}
O∗+H2O(l)+e−⇌OH∗+OH−,(10)\mathrm{O^* + H_2O}(l) + e^- \rightleftharpoons\mathrm{OH^* + OH^-}, \tag*{(10)}
OH∗+e−⇌MOF+OH−.(11)\mathrm{OH^*} + e^- \rightleftharpoons\mathrm{MOF + OH^-}. \tag*{(11)}

For the OER, the reactions proceed in reverse. To assess the energetics of the ORR/OER reactions, we plotted adsorption free energy profiles for the reaction intermediates OOH∗^*, O∗^*, and OH∗^*, each bound to the metal cation of systems studies.

At zero applied potential referenced to the reversible hydrogen electrode (RHE), i.e., U=0U = 0 V, all elementary reaction steps are thermodynamically downhill. When the potential is increased to the standard equilibrium value of U=1.23U = 1.23 V, the reaction free energies are shifted downward in proportion to the number of electrons involved in each elementary step. The minimum applied potential at which every step becomes energetically favorable defines the onset potential, denoted as UonsetU_{\mathrm{onset}}. The theoretical overpotential, η\eta, which serves as a key metric for ORR/OER electrocatalytic performance, is determined from the deviation of the onset potential from the equilibrium value. Accordingly, the overpotentials for ORR and OER are given by ηORR=1.23\eta_{\mathrm{ORR}} = 1.23 V −Uonset- U_{\mathrm{onset}} and ηOER=Uonset−1.23\eta_{\mathrm{OER}} = U_{\mathrm{onset}} - 1.23 V, respectively. A detailed description of the computational procedure employed to evaluate the adsorption free energies of reaction intermediates and the associated ORR/OER overpotentials is provided in the Supplemental Material. Table III summarizes the calculated ORR and OER overpotentials for the metal-N4_4 and metal-O4_4 structures, together with the dd-band center, which is used as a descriptor of catalytic activity.

Systemεd\varepsilon_dηORR\eta_{\mathrm{ORR}}ηOER\eta_{\mathrm{OER}}
G-MnN4_4-0.801.782.51
G-FeN4_4-1.011.322.57
G-CoN4_4-2.010.411.55
G-NiN4_4-2.710.760.62
G-CuN4_4-4.550.731.50
G-ZnN4_4-7.681.091.84
s-MnPc-1.430.981.80
s-FePc-0.880.271.05
s-CoPc-1.820.380.85
s-NiPc-2.790.921.00
s-CuPc-4.500.910.75
s-ZnPc-7.530.820.88
Mn4_4(OHPTP)2_2-1.231.141.51
Fe4_4(OHPTP)2_2-1.721.650.85
Co4_4(OHPTP)2_2-2.020.480.61
Ni4_4(OHPTP)2_2-1.981.110.84
Cu4_4(OHPTP)2_2-3.151.220.91
Zn4_4(OHPTP)2_2-6.860.580.55
Mn3_3(HHTP)2_2-0.930.890.91
Fe3_3(HHTP)2_2-1.170.510.25
Co3_3(HHTP)2_2-1.281.531.04
Ni3_3(HHTP)2_2-1.560.480.39
Cu3_3(HHTP)2_2-2.641.140.82
Zn3_3(HHTP)2_2-5.740.871.37
Pt(111)-0.48-
IrO2_2(110)--0.65

Table III. Catalytic activity descriptors for the single-layer organometallic structures, including the dd-band center (εd\varepsilon_d, in eV) relative to the Fermi energy and ORR/OER overpotentials (η\eta, in V). Previously reported theoretical overpotential values for ORR on Pt(111)[58] and OER on IrO2_2(110)[59] are shown for reference.

Previously reported ORR and OER overpotentials for Pt(111) and IrO2_2(110) surfaces, obtained using a similar theoretical approach, are included as benchmarks.[58, 59] fig:S13–S20 present the corresponding ORR and OER free energy diagrams for all the structures considered. To further corroborate our results, we calculated the dd-band center for all systems, confirming that the highest ORR/OER activity is generally associated with dd-band centers closer to the Fermi energy, as shown in fig:S21–S24.

Among the metal-N4_4 systems, G-CoN4_4 and G-NiN4_4 exhibit the best catalytic performance for ORR and OER, with overpotentials of 0.41 and 0.62 V, respectively, comparable to the benchmark values. For ORR, our results are in good agreement with available experimental data.[60] In the case of s-MPc structures, the most active catalysts for ORR are s-FePc and s-CoPc, with overpotentials of 0.27 and 0.38 V, respectively, whereas none of these structures exhibit OER overpotentials lower than the benchmark. Experimental studies on Co-phthalocyanine-based MOFs report ORR activities approaching those of Pt/C benchmark catalysts,[61] and also superior activity for OER,[62] which agree well with our predictions.

For the metal-O4\mathrm{O}_{4} systems, Co4(OHPTP)2\mathrm{Co}_{4}(\mathrm{OHPTP})_{2} exhibits good catalytic activity for both ORR and OER, with overpotentials of 0.48 and 0.61 V, respectively, while Zn4(OHPTP)2\mathrm{Zn}_{4}(\mathrm{OHPTP})_{2} also shows a relatively low OER overpotential of 0.55 V. Notably, both structures display bifunctional behavior. In addition, Fe3(HHTP)2\mathrm{Fe}_{3}(\mathrm{HHTP})_{2} and Ni3(HHTP)2\mathrm{Ni}_{3}(\mathrm{HHTP})_{2} exhibit ORR activity comparable to benchmark values, while delivering significantly improved OER performance. Experimental data are available for M3(HHTP)2\mathrm{M}_{3}(\mathrm{HHTP})_{2} systems; in particular, Co3(HHTP)2\mathrm{Co}_{3}(\mathrm{HHTP})_{2} and Ni3(HHTP)2\mathrm{Ni}_{3}(\mathrm{HHTP})_{2} have been demonstrated to be efficient OER catalysts with enhanced stability,[63, 64] in agreement with our predictions.

Concerning electrochemical stability, a growing body of experimental evidence indicates that G-FeN4\mathrm{FeN}_{4} centers, although active for the oxygen reduction reaction, are intrinsically prone to degradation under operating conditions.[43, 65, 66] Recent studies have shown progressive degradation due to reactive oxygen species attack on the active site, inducing oxidative Fe demetallation and the concomitant oxidation of the surrounding carbon matrix.[67, 68, 69, 70] Similar degradation pathways have been reported for other metal centers, including G-CoN4\mathrm{CoN}_{4} and G-MnN4\mathrm{MnN}_{4},[60, 71, 72] showing metal-dependent stability trends. Collectively, these findings identify demetalation and support oxidation as the primary mechanisms limiting the durability of G-MN4\mathrm{MN}_{4} catalysts, underscoring an inherent trade-off between catalytic activity and electrochemical stability.[43]

To identify metal-N4\mathrm{N}_{4} and metal-O4\mathrm{O}_{4} structures that simultaneously satisfy catalytic activity and electrochemical stability, and to enable a direct comparison between them, we plot the stability descriptor DD against the ORR and OER overpotentials. Figure 3 and Figure 4 present the DD descriptor as a function of ORR and OER overpotentials, respectively, for the systems under study. The three colored regions of the figures, red, yellow, and green, represent electrochemically unstable, partially stable, and stable regimes, respectively. As observed, metal-N4\mathrm{N}_{4} structures are considerably less stable than their metal-O4\mathrm{O}_{4} counterparts for most metal centers, suggesting a higher susceptibility to degradation. This behavior is consistent with experimental observations reported for G-MN4\mathrm{MN}_{4} catalysts. In contrast, none of the metal-O4\mathrm{O}_{4} structures fall within the instability regime.

Electrochemical stability descriptor as a function of ORR overpotential

Figure 3. Electrochemical stability descriptor (DD) as a function of the ORR overpotential (ηORR\eta_{\mathrm{ORR}}) for the organometallic structures. The red, yellow, and green regions denote electrochemically unstable, partially stable, and stable regimes, respectively.

Electrochemical stability descriptor as a function of OER overpotential

Figure 4. Electrochemical stability descriptor (DD) as a function of the OER overpotential (ηOER\eta_{\mathrm{OER}}) for the organometallic structures. The red, yellow, and green regions denote electrochemically unstable, partially stable, and stable regimes, respectively.

Therefore, based on the combined analysis of catalytic activity and electrochemical stability across the four investigated systems, oxygen-coordinated metal frameworks emerge as the most promising candidates for practical electrocatalytic applications. These systems consistently exhibit a greater tendency toward electrochemical stability over a broad pH range, while maintaining competitive overpotential values for both the ORR and OER. In contrast,

although nitrogen-coordinated metal systems often display favorable overpotentials, indicative of strong intrinsic catalytic activity, our results suggest a higher susceptibility to degradation, in good agreement with experimental reports on G-MN4\mathrm{MN}_{4} structures. Such limitations include restricted stability depending on pH and, in some cases, complete electrochemical instability, which may ultimately hinder their practical applicability under operational conditions.

Summary and Conclusions

In summary, DFT calculations were employed to investigate the electrochemical stability and ORR/OER catalytic activity of layered organometallic materials. The studied systems comprise metal-N4\mathrm{N}_4 structures: G-MN4_4 and s-MPc, and metal-O4_4 structures: M4_4(OHPTP)2_2 and M3_3(HHTP)2_2, with M = Mn, Fe, Co, Ni, Cu, and Zn. ORR and OER activities were evaluated using the CHE model, while electrochemical stability was assessed through surface Pourbaix analysis. To quantify electrochemical stability, we introduce the descriptor DD, defined from the standard potentials, which is obtained from the electronic work function of the pristine surface and the corresponding OH-covered surface at the metal active site. This descriptor enables a direct comparison of stability across different metal coordination environments.

Our results reveal that metal-N4\mathrm{N}_4 systems often exhibit favorable overpotentials but suffer from pronounced pH-dependent stability limitations. In contrast, metal-O4_4 systems display enhanced electrochemical robustness while maintaining competitive ORR and OER performance, thereby offering a more balanced combination of activity and stability. Particularly noteworthy are the Co4_4(OHPTP)2_2 and Zn4_4(OHPTP)2_2 structures, which exhibit both high stability and bifunctional ORR/OER activity. Similarly, Fe3_3(HHTP)2_2 shows excellent ORR/OER activity, although it becomes unstable under highly acidic conditions. Finally, owing to its clear physical meaning and quantitative nature, the proposed DD descriptor serves as a transferable metric for screening the electrochemical stability of layered organometallic catalysts.

Acknowledgments

We acknowledge financial support from the National Agency for Research and Development of Chile (ANID), through projects FONDECYT 1230138 and FONDECYT POSTDOC 3240440 (P.H.S.). The computational resources were provided by Fenix HPC (UNAB) and the supercomputing infrastructure of the NLHPC (CCSS210001).

References

  1. [1]1.
  2. [2]2.
  3. [3]3.
  4. [4]4.
  5. [5]5.
  6. [6]6.
  7. [7]7.
  8. [8]8.
  9. [9]9.
  10. [10]10.
  11. [11]11.
  12. [12]12.
  13. [13]13.
  14. [14]14.
  15. [15]15.
  16. [16]16.
  17. [17]17.
  18. [18]18.
  19. [19]19.
  20. [20]20.
  21. [21]21.
  22. [22]22.
  23. [23]23.
  24. [24]24.
  25. [25]25.
  26. [26]26.
  27. [27]27.
  28. [28]28.
  29. [29]29.
  30. [30]30.
  31. [31]31.
  32. [32]32.
  33. [33]33.
  34. [34]34.
  35. [35]35.
  36. [36]36.
  37. [37]37.
  38. [38]38.
  39. [39]39.
  40. [40]40.
  41. [41]41.
  42. [42]42.
  43. [43]43.
  44. [44]44.
  45. [45]Kresse, G., & Furthmüller, J. (1996). Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B, 54, 11169.
  46. [46]Lee, K., Murray, É. D., Kong, L., Lundqvist, B. I., & Langreth, D. C. (2010). Higher-accuracy van der Waals density functional. Phys. Rev. B, 82, 081101.
  47. [47]47 T. Hamada and T. Ohno, “A new constraint DFT technique for self-consistent determination of U values,” J. Phys.: Condens. Matter. 31, 065501 (2018).DOI
  48. [48]48 V. Wang, N. Xu, J. Liu, G. Tang, and W. Geng, “VASPKIT: a User-Friendly interface facilitating High-Throughput computing and analysis using VASP code,” Comput. Phys. Commun. 267, 108033 (2021).DOI
  49. [49]Momma, K., & Izumi, F. (2008). VESTA: A three-dimensional visualization system for electronic and structural analysis. J. Appl. Crystallogr., 41, 653–658.
  50. [50][visible bibliography entry not printed on this page]
  51. [51]51 K. Mathew, R. Sundararaman, K. Letchworthweaver, T. A. Arias, and R. G. Hennig, “Implicit solvation model for density-functional study of nanocrystal surfaces and reaction pathways,” J. Chem. Phys. 140, 084106 (2014).DOI
  52. [52]Ni, X., Li, H., & Bredas, J. L. (2024). Half-metallic ferromagnetism in radical-bridged metal-organic frameworks. Chem. Mater., 36, 2380–2389.
  53. [53][visible bibliography entry not printed on this page]
  54. [54][visible bibliography entry not printed on this page]
  55. [55][visible bibliography entry not printed on this page]
  56. [56][visible bibliography entry not printed on this page]
  57. [57]Li, S., Shi, Y., Guo, J., Wang, J., Liu, D., & Zhao, S. (2024). Selective oxygen reduction reaction: Mechanism understanding, catalyst design and practical application. Chem. Sci., 15, 11188–11228.
  58. [58]Hansen, H. A., Rossmeisl, J., & Nørskov, J. K. (2008). Surface pourbaix diagrams and oxygen reduction activity of Pt, Ag and Ni(111) surfaces studied by DFT. Phys. Chem. Chem. Phys., 10, 3722–3730.
  59. [59]Xu, H., Cheng, D., Cao, D., & Zeng, X. C. (2018). A universal principle for a rational design of single-atom electrocatalysts. Nat. Catal., 1, 339–348.
  60. [60]60 N. Li, L. Li, J. Li, X. Jia, M. Arif, S. Zhou, F. Yin, G. He, and H. Chen, “Single-atom Co-N₄ catalytic sites anchored on N-doped ordered mesoporous carbon for excellent Zn-air batteries,” J. Mater. Sci. Technol. 139, 224 (2023).DOI
  61. [61]Shantharaja, Giddaerappa, & Sannaggowda, L. K. (2023). Phthalocyanine based metal-organic framework with carbon nanoparticles as hybrid catalyst for oxygen reduction reaction. Electrochim. Acta, 456, 142405.
  62. [62]62 Y. Kim, D. Kim, J. Lee, L. Y. S. Lee, and D. K. P. Ng, “Tuning the electrochemical properties of polymeric cobalt phthalocyanines for efficient water splitting,” Adv. Funct. Mater. 31, 2103290 (2021).DOI
  63. [63]63 M. Zhang, B.-H. Zheng, J. Xu, N. Pan, J. Yu, M. Chena, and H. Cao, “Fewer-layer conductive metal-organic nanosheets enable ultrahigh mass activity for the oxygen evolution reaction,” Chem. Commun. 54, 13579 (2018).DOI
  64. [64]Dong, J., Bouk hvalov, D. W., Lv, C., Humphrey, M. G., Zhang, C., & Huang, Z. (2024). Enhancing oxygen evolution reaction performance of metal-organic frameworks through cathode activation. ChemSusChem, 17, e202401176.
  65. [65]65 Y. Shao, J.-P. Dodelet, G. Wu, and P. Zelenay, “PGM-free cathode catalysts for PEM fuel cells: A mini-review on stability challenges,” Adv. Mater. 31, 1807615 (2019).DOI
  66. [66]66 D. Xia, C. Yu, Y. Zhao, H. Wei, H. Wu, Y. Kang, J. Li, L. Gan, and F. Kang, “Degradation and regeneration of Fe–Nₓ active sites for the oxygen reduction reaction: the role of surface oxidation, Fe demetallation and local carbon microporosity,” Chem. Sci. 12, 11576 (2021).DOI
  67. [67]Menga, D., Li, Y.-S., Damjanovic, A. M., Proux, O., Wagner, F. E., Fellinger, T.-P., Gasteiger, H. A., & Piana, M. (2024). On the stability of an atomically-dispersed Fe–N–C ORR catalyst: An in situ XAS study in a PEMFC. ChemElectroChem, 11, e202400228.
  68. [68]Xie, X., Bie, L., Xu, P., Xu, R., Garcia-Serres, M. T., Sougrati, X., Cullen, D. A., & al., . et . (2025). Unravelling the stability stressors of atomically dispersed Fe–N–C oxygen reduction catalysts. J. Am. Chem. Soc., 147, 48117.
  69. [69]69 W. Chen, Y. Qin, C. Heng, S. Wang, M. Ge, Y. Li, X. Wan, X. Li, J. Shui, Y. Su, , and D. Su, “Degradation of FeNC electrocatalysts for acidic and alkaline oxygen reduction,” J. Am. Chem. Soc. 147, 35730 (2025).DOI
  70. [70]Kobayashi, R., Honma, S., & Ozaki, J. (2025). Quantitative detection of hydroxyl radical-mediated degradation in Fe–N–C catalysts during the oxygen reduction reaction. Carbon Reports, 4, 270.
  71. [71]Li, J., Chen, M., Cullen, D. A., Hwang, S., Wang, M., Li, L., & al., . et . (2018). Atomically dispersed manganese catalysts for oxygen reduction in proton-exchange membrane fuel cells. Nat. Catal., 1, 935–945.
  72. [72]Li, Y., Chen, M.-Y., Lu, B.-A., Lu, B.-A., Zhang, J.-N., & Wu, H.-R. (2024). Unravelling the role of hydrogen peroxide in pH-dependent ORR performance of Mn-N-C catalysts. Appl. Catal. B-Environ., 342, 123458.

Paper details

vlnc:W_df2xagnz
Version
v1 · latest
Published
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents