Engineering PtFe/LiO2 Frontier Orbital Interaction in Li–O2 Batteries

TL;DR

This paper designs a PtFe catalyst to improve oxygen evolution reaction activity in Li–O2 batteries by manipulating orbital interactions at the atomic level.

Contribution

The study introduces a new method to enhance catalytic activity by controlling electron population in Pt dz2 orbitals through Fe alloying.

Findings

dz2–dz2 orbital coupling between Fe and Pt increases electron population in the Pt dz2 orbital.

Higher Pt content in PtFe alloys reduces the electron population in the Pt dz2 orbital, lowering OER activity.

Electron population in the Pt dz2 orbital correlates with OER activity, offering a design descriptor for electrocatalysts.

Abstract

PtFe catalyst was rationally designed based on frontier molecular orbital theory to investigate orbital-level interactions for enhanced oxygen evolution reaction activity in Li–O2 batteries.The dz2–dz2 orbital coupling between Fe and Pt leads to electron donation from Fe to Pt, increasing electron population in the Pt dz2 orbital.Excess electrons from the Pt dz2 orbital occupy antibonding states with LiO2, weakening interaction strength and boosting oxygen evolution reaction kinetics.

PtFe catalyst was rationally designed based on frontier molecular orbital theory to investigate orbital-level interactions for enhanced oxygen evolution reaction activity in Li–O2 batteries.

The dz2–dz2 orbital coupling between Fe and Pt leads to electron donation from Fe to Pt, increasing electron population in the Pt dz2 orbital.

Excess electrons from the Pt dz2 orbital occupy antibonding states with LiO2, weakening interaction strength and boosting oxygen evolution reaction kinetics.

The online version contains supplementary material available at 10.1007/s40820-026-02085-z.

Elucidating the structure–activity relationship between the electronic structure of catalytic active sites and oxygen evolution reaction (OER) activity at the orbital level is critical yet challenging in lithium–oxygen (Li–O2) batteries. Herein, employing frontier molecular orbital theory, we designed a Pt-based catalyst as a model cathode to investigate the influence of frontier orbital interactions between the Pt dz2 orbital and the 5σ orbital of LiO2 on the OER activity. Specifically, compared to the pure Pt catalyst, the dz2–dz2 orbital coupling between low-electronegativity Fe and Pt in PtFe catalyst induces predominant electron transfer from Fe to the dz2 frontier orbital of Pt. As the Pt content in PtFe alloys increases progressively (from Pt58Fe42, Pt67Fe33 to Pt76Fe24), the electron population of the Pt 5dz2 orbital gradually decreases (1.92 for Pt58Fe42, 1.85 for Pt67Fe33, and 1.80 for Pt76Fe24). This leads to a gradual enhancement in the strength of interactions between the Pt dz2 orbital and the frontier orbitals of LiO2, consequently resulting in a progressive decline in the OER catalytic activity. Establishing the correlating between the electron population in the dz2 frontier orbital and OER activity provides a descriptor for designing efficient electrocatalysts in Li–O2 batteries.

The online version contains supplementary material available at 10.1007/s40820-026-02085-z.

Figures

Full text

Introduction

Lithium–oxygen (Li–O_2_) batteries have recently gained widespread attentions due to the extremely high energy density (3500 Wh kg^−1^) [1–6]. However, the sluggish oxygen evolution reactions (OER) kinetics during the charging process of Li–O_2_ batteries result in low energy efficiency and high charging overpotential [7–12]. Although conventional redox mediators, e.g., triarylmethyl cations [13], can significantly improve OER kinetics, the shuttling effect and the inherent thermodynamic instability of liquid-phase catalysts lead to an inferior lifespan [13–16]. In contrast, various solid-phase catalysts such as noble metals [17–19], high-entropy alloys [20, 21], transition metal oxides [22–24], sulfides [25–27], single atoms [28–33], MXenes [34–36], etc. are widely employed as the catalytic alternatives for the enhanced fast OER process. Due to the direct correlation between d-orbital characteristics and catalytic activity, d-band center theory has been widely accepted as a descriptor for d-orbital electronic state, enabling rational design of high-performance catalysts by orbital hybridization [37, 38], and orbital coupling [39, 40]. However, as only a statistical average of the d-electron density of states, the d-band center cannot reveal the true electronic structure of active sites at specific energy levels and thus cannot precisely resolve the differential influence of site-specific d-orbital electron occupation on the OER catalytic activity.

Herein, unlike the d-band center theory, we utilize frontier orbital theory and construct Pt-based catalyst as a model platform to investigate the influence of frontier orbital interactions between Pt dz^2^ orbital and 5σ orbital of LiO_2_ on OER activity. The selection of PtFe as the cathode catalyst for Li–O_2_ batteries is guided by three fundamental mechanistic considerations. First, the interaction between Fe and Pt induces lattice strain effects that accurately modulate the Pt d-band center and regulate electron back-donation. This leads to optimized adsorption of oxygen intermediates such as LiO_2_, thereby accelerating the OER kinetics. Second, the incorporation of Fe significantly enhances the stability of Pt by preventing its dissolution and aggregation under strongly oxidative conditions. At the same time, it facilitates high utilization of Pt through atomic-scale dispersion. Third, the catalytic performance of PtFe systems has been well demonstrated in various oxygen-related reactions, including oxygen reduction reaction in fuel cells and OER in water electrolysis. These prior studies provide a strong theoretical and experimental basis for the rational design and mechanistic application of PtFe catalysts in Li–O_2_ battery systems. In the PtFe system, the intrinsic d–d-orbital interaction between Fe and Pt enables precise modulation of the electron occupancy in the Pt d_z^2^ orbital at the active site. This, in turn, allows for fine-tuning of the binding strength between the active site and reaction intermediates (such as LiO_2), thereby effectively regulating the OER activity. Therefore, when dz^2^ frontier orbital of Pt in PtFe interacts with the 5σ orbital of LiO_2_, the excess electrons in the Pt dz^2^ orbital occupy the antibonding orbital, thereby weakening the adsorption of LiO_2_ and ultimately improving the OER catalytic activity. It is worth noting that the frontier orbital theory-based regulation strategy we employ differs from conventional hybridization engineering. Hybridization engineering is primarily based on the d-band center theory. However, the d-band center fails to capture the detailed electronic structure of active sites at specific energy levels and therefore cannot accurately elucidate the role of electron occupancy in individual d-orbitals on OER catalytic activity. In contrast, frontier orbital theory enables a more precise understanding of how specific d-orbital components contribute to OER performances. This work establishes a correlation between the electron number of dz^2^ frontier orbital and OER catalytic activity, which provides a descriptor for designing high-performance Li–O_2_ batteries.

Experimental Section

Materials

Platinum(II) acetylacetonate (Pt(acac)2, ≥ 99.99%), iron(III) acetylacetonate (acetylacetonate (Fe(acac)3, ≥ 98%), Glucose (≥ 98%), hexadecyl trimethyl ammonium bromide (≥ 99%), poly(vinylidene fluoride) (PVDF), Li anode (lithium foil), lithium trifluoromethanesulfonate (LiCF_3_SO_3_, ≥ 99.5%) were purchased from Aladdin. Oleylamine (C_18_H_37_N, 80%-90%), cyclohexane (C_6_H_12_, ≥ 99.9%), ethanol solvent (C_2_H_5_OH, ≥ 99.7%), acetic acid (CH_3_COOH, ≥ 99.5%), 1-methyl-2 pyrrolidone (NMP, C_6_H_13_NO, ≥ 99.5%), tetraethylene glycol dimethyl ether (C_10_H_22_O_5_, 99%) were obtained from Macklin.

Synthesis of PtFe (Pt58Fe42, Pt67Fe33, and Pt76Fe24) Nanowires

Dissolve 10 mg platinum(II) acetylacetonate (Pt(acac)2), 3.6 mg iron(III) acetylacetonate (Fe(acac)3), 20 mg glucose, 40 mg hexadecyltrimethylammonium bromide (CTAB), and 5 mL oleylamine in a 20-mL glass vial and cap it (without purging with inert gas). Then, sonicate the mixture for 2 h to ensure uniform dispersion. Subsequently, the vial was placed in an oil bath at 200 °C and reacted for 5 h without stirring. After the reaction is completed, cool the solution to room temperature, wash with a cyclohexane and ethanol solvent (in a 9:1 volume ratio, and centrifuge at 9000 rpm for 10 min to obtain the Pt_58_Fe_42_ catalyst. The synthesis procedures for Pt_67_Fe_33_ and Pt_76_Fe_24_ are similar to that of Pt_58_Fe_42_, except for the amount of Fe(acac)3. Specifically, 3 and 2.8 mg of Fe(acac)3 were used for the synthesis of Pt_67_Fe_33_ and Pt_76_Fe_24_, respectively.

Preparation of Pt/C and PtFe/C

The PtFe (80 mg) obtained is dispersed in 60 mL of cyclohexane, followed by uniform mixing with 20 mg of XC72R carbon and continuous sonication for approximately 60 min. Following centrifugation, the solid product is isolated and subsequently washed with a cyclohexane/ethanol mixed solution (9:1, v/v). Furthermore, the solid product is then stirred overnight in acetic acid to remove surface residues. The PtFe/C catalyst was collected by centrifugation, with a measured PtFe loading of 80 wt%. Pt/C was synthesized following an identical procedure.

Characterizations

The X-ray diffraction (XRD) patterns of Pt and PtFe were obtained using a Rigaku D/max-2400 X-ray powder diffractometer, with Cu Kα (λ = 1.5406 nm) as the X-ray source. The morphologies of initial, discharged, and charged Pt and PtFe electrode were analyzed using high-resolution transmission electron microscopy (HR-TEM, Tecani-G2 T20), scanning electron microscopy (SEM, Thermo Fisher Quattro S system), X-ray photoelectron spectroscopy (XPS, Thermo Fisher ESCALAB XI +), Fourier transform infrared spectroscopy (FTIR, PerkinElmer Spectrum II FTIR Spectrometer), Raman (WITec alpha 300 Raman System), nuclear magnetic resonance (NMR, Bruker), and UV–vis spectra (Hitachi UH4150). Before conducting ex situ characterization, the Pt- and PtFe-based batteries are transferred to a glovebox (with O_2_ and H_2_O levels below 0.1 ppm) and disassembled. The PtFe and Pt electrodes are thoroughly rinsed with dimethyl sulfoxide to eliminate electrolyte and lithium salts. To avoid contamination, the PtFe and Pt electrodes should be securely sealed with Parafilm. To ensure accurate ex situ results, the PtFe and Pt electrodes are removed from the glovebox 5 min before testing.

Battery Assembly and Performances

To prepare PtFe electrodes, PVDF (10 wt%) was dispersed in NMP at 60 °C for 20 min. Subsequently, PtFe/C (90 wt%) was incorporated into the PVDF solution, followed by continuous magnetic stirring for 8 h to ensure a uniform mixture. Subsequently, the catalyst slurry was drop cast onto carbon paper substrates (0.21 mm thick, 10 mm diameter) and vacuum-dried at 60 °C for 24 h to fabricate the Pt/C (or PtFe/C) electrodes. The PtFe/C electrode has a surface area of 0.75 cm^2^, with a loading mass of 3 to 4 mg cm^−2^ for PtFe/C or Pt. The assembly of coin cells (2032, 13 holes) was assembled in a glove box (with H_2_O and O_2_ concentration below 0.1 ppm). The coin cell includes FePt/C (Pt/C) cathode, a Li anode, separator, and electrolytes (10 μL, 1 M lithium trifluoromethanesulfonate in tetraethylene glycol dimethyl ether). The lithium anode was prepared before the assembly of the battery. Before testing, the coin cell will be placed in a specific container (with dry O_2_ content above 99.99% at 1.0 atm.) and maintained at a constant temperature of 25 °C for 24 h in a constant-temperature chamber. The assembled coin cells (including PtFe/C or Pt/C) were analyzed by the NEWARE battery testing system to obtain full and limited discharge–charge curves. The cyclic voltammetry (CVs) was analyzed by an electrochemical workstation (CHI760e) from 2.0 to 4.5 V at a scan rate of 0.3 mV s^−1^. EIS measurements were also carried out using the CHI760e electrochemical workstation, applying an amplitude voltage of 10 mV and a frequency range from 10^5^ to 0.1 Hz.

Density Functional Theory Calculations

All density functional theory (DFT) calculations were performed using the Vienna Ab initio Simulation Package (VASP). The Perdew–Burke–Ernzerhof generalized gradient approximation (PBE-GGA) was employed to describe exchange–correlation effects in all calculations. The interactions between ionic cores and valence electrons were modeled with the Projector Augmented-Wave (PAW) pseudopotential method. We set the plane-wave energy cutoff to 450 eV and relaxed all structures until the maximum Hellmann–Feynman forces fell below − 0.02 eV Å^−1^, and consecutive energy differences were less than 10^−5^ eV. Dispersion interactions between all atoms in the adsorption systems were accounted for using Grimme's DFT-D3 correction scheme. The lattice parameters used in the DFT calculations for the PtFe alloy were determined as follows. Based on the cubic unit cell of pure Pt, which aligns with the catalyst's XRD peak positions, we constructed a 100-atom supercell. In this supercell, Pt atoms were randomly replaced with Fe atoms to generate alloy models with compositions spanning Pt_58_Fe_24_, Pt_67_Fe_33_ and Pt_76_Fe_24_. The atomic positions of each model were then optimized while keeping the experimental lattice constants fixed. The d-band center (εd) was calculated as the first moment of the projected d-orbital density of states (PDOS) of the surface Pt/Fe atoms, using the formula \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varepsilon_{{\mathrm{d}}} = {{\int {E \times D(E){\mathrm{d}}E} } \mathord{\left/ {\vphantom {{\int {E*D(E){\mathrm{d}}E} } {D\left( E \right){\mathrm{d}}E}}} \right. \kern-0pt} {D\left( E \right){\mathrm{d}}E}}$$\end{document} , where D(E) is the DOS of d-band suborbital. The d-electron numbers were obtained by integrating the occupied part of the corresponding d-suborbital PDOS below Femi energy.

Results and Discussion

Synthesis and Characterization of PtFe Nanowires

The fabrication of PtFe nanowires is conducted using a typical wet chemical method. The obtained PtFe nanowires exhibit a zigzag-like morphology, with an average width of 10–20 nm (Fig. 1a, b). The XRD results demonstrate that the main peaks at 40.35°, 47.19°, 68.71°, 82.50°, and 86.71° are attributed to the (111), (200), (220), (311), and (222) planes of PtFe nanowires (Fig. 1c). It is worth noting that the peak intensity of the (111) facet of PtFe is stronger than that of the (200) and (220) facets. The results indicate that the (111) facet of Pt_58_Fe_42_ exhibits a lower surface energy (4.68 J m^−2^) compared to both the (200) facet (5.29 J m^−2^) and the (220) facet (4.74 J m^−2^). Consequently, the (111) facet of Pt_58_Fe_42_ demonstrates superior thermodynamic stability (Fig. S1, Tables S1–S3). Due to the smaller atomic radius of Fe than that of Pt, the diffraction peaks of PtFe shift to higher angles of 2θ compared to those of pure Pt. The lattice spacing of 0.136, 0.19, and 0.221 nm corresponds to the (022), (200), and (111) crystal planes of PtFe nanowires, respectively (Fig. 1d). The Fourier transform (FFT) of the selected area in Fig. 1d shows different facets along the [011] zone axis (Fig. 1e), which is in accordance with the XRD results. The energy-dispersive X-ray spectroscopy (EDS) mapping reveals a homogeneous distribution of Pt and Fe within a single PtFe nanowire, with a molar ratio of Pt to Fe at 58:42 (Fig. 1f, g). In addition, two other PtFe electrocatalysts with different Pt ratios (Pt_67_Fe_33_ and Pt_76_Fe_24_) were also successfully synthesized (Figs. S2 and S3).Fig. 1a, b TEM images of PtFe nanowires. c XRD patterns of Pt and PtFe nanowires. d HAADF-STEM image of PtFe nanowires. e FFT patterns of the dashed squares in corresponding colors in (d) for PtFe nanowires. f EDX spectra of PtFe nanowires. g EDS mapping of Pt and Fe elements based on a single PtFe nanowire

Electrocatalytic Performances of PtFe Nanowires

Pure Pt and the PtFe nanowires are further employed as cathode electrocatalysts in Li–O_2_ batteries. Under a discharge capacity limitation of 1 mAh cm^−2^ at 0.1 mA cm^−2^ with 3 mg cm^−2^ mass loading, Pt_58_Fe_42_ demonstrates the lowest charge overpotential (0.24 V) and highest energy efficiency (84%) (Fig. 2a), significantly outperforming pure Pt (0.69 V), Pt_67_Fe_33_ (0.41 V), and Pt_76_Fe_24_ (0.92 V) (Fig. S4). In the subsequent study, the best-performing Pt_58_Fe_42_ was selected for comparison with pure Pt. Since the discharge products of Li_2_O_2_ primarily form on the catalyst surface during discharge, the capacity is largely determined by the catalyst's specific surface area, which governs the deposition capacity of Li_2_O_2_. Owing to its high specific surface area of 261 m^2^ g^−1^, Pt_58_Fe_42_ consequently delivers a full‑discharge areal capacity of 11.59 mAh cm^2^ and an energy density of 1200 Wh kg^−1^ (including the total mass of the cathode, electrolyte, binder, and Li_2_O_2_; calculation details are provided in Table S4) at a current density of 0.4 mA cm^−2^ (Fig. 2b). These values exceed those of Pt/C, which exhibits a lower specific surface area of 84 m^2^ g^−1^ (Fig. S5) and achieves 981 Wh kg^−1^ and 8.72 mAh cm^−2^. Moreover, Pt_58_Fe_42_ demonstrates a charge overpotential of 0.67 V, much lower than the 1.38 V of Pt (Fig. 2b). The cyclic voltammetry (CV) spectra further demonstrate that Pt_58_Fe_42_ shows a higher exchange current density of 19.63 μA cm^−2^  than that of the pure Pt (11.96 μA cm^−2^), confirming the exceptional OER kinetics of Pt_58_Fe_42_ (Fig. 2c). Furthermore, under capacity-limited cycling (1 mAh cm^−2^), Pt could only endure cycling for 100 cycles due to severe charge polarizations (Fig. 2d). In contrast, Pt_58_Fe_42_ maintained stable charge–discharge characteristics even after 250 cycles (1000 h) at 0.5 mA cm^−2^, showing no notable performance degradation (Fig. 2d). Even at a capacity of 1000 mAh g^−1^ (3 mAh cm^−2^), the Pt_58_Fe_42_ catalyst can still maintain stable cycling for 200 cycles at a current density of 0.75 A g⁻^1^ (Fig. 2e). The combination of low charge overpotential and long cycle stability of Pt_58_Fe_42_ electrocatalysts surpasses the results of the majority of noble metal-based cathode electrocatalysts (Fig. 2f and Table S5).Fig. 2a Discharge–charge curves of Pt and Pt_58_Fe_42_ electrocatalysts with a limited capacity of 1 mAh cm^−2^. b Full discharge–charge profiles of Pt and Pt_58_Fe_42_ nanowire. c CVs curves for Pt and Pt_58_Fe_42_ electrocatalysts from 2.0 to 4.5 V. d Cycle performance of Pt and Pt_58_Fe_42_ electrocatalysts  at 1 mAh cm^−2^ under a current density of 0.5 mA cm^−2^. e Cycling stability performance of Pt_58_Fe_42_ at 1000 mAh g⁻^1^ and a current density of 0.75 A g⁻^1^. f Examining the charge overpotentials and cycle stability of Pt_58_Fe_42_ nanowires with a variety of electrocatalysts, including RuO_2_/MnO_2_ [41], Pd/CNT [42], Pt [43], HEAPtIr [44], Pd_3_Pb [45], PtIr [18], NiRu-HTP [46], PtFe_c_/NC [47], PtAu [17], and LaSr(5TM)O_3_ [48]

Reversibility Analysis of Pt58Fe42 Nanowires

To reveal the reversibility of Pt_58_Fe_42_ electrocatalysts, the morphologies of the discharged Pt_58_Fe_42_ and Pt are analyzed using SEM. When discharged at 1 mAh cm^−2^, the discharged Pt_58_Fe_42_ cathode exhibits a disk-shaped morphology with a diameter of 2–3 μm (Fig. 3a), which have been identified as Li_2_O_2_ (Figs. S6–S8). Compared to Pt_58_Fe_42_, the vast majority of discharge products on the Pt electrode are disk-shaped (4 μm in diameter), whereas a negligible portion exhibits a distinct sphere morphology (Figs. S9–S11). For Pt_58_Fe_42_, the low surface energy (0.29 eV Å^−2^) of its (111) facet facilitates the disproportionation of the intermediate LiO_2_ in solution, resulting in the formation of typical disk-shaped discharge products via the solution growth mechanism. In contrast, for the Pt catalyst, the relatively high surface energy of (111) facet (0.38 eV Å^−2^ in Fig. S12 and Table S6) promotes partial adsorption of intermediates on the surface, where they may undergo either reduction or disproportionation reactions. Consequently, a small number of spherical discharge products are observed on the Pt electrode, representing a morphology that lies between the thin films formed by the pure surface growth model and the disk structures produced via the dissolution pathway. The small disk-shaped Li_2_O_2_ indicates good interfacial contact with the Pt_58_Fe_42_ cathode, resulting in a significantly lower charge transfer resistance of 115.6 Ω compared to 811.2 Ω for larger-diameter Li_2_O_2_ on Pt_58_Fe_42_ (Fig. S13 and Table S7). The reversibility of Pt_58_Fe_42_ and Pt is further investigated under charging and discharging conditions through SEM images, XRD patterns, Raman spectra, and XPS spectra. During the initial, the 1st discharged, the 1st charged, the 10th discharged, the 10th charged, the 20th discharged, and the 20th charged processes, Li_2_O_2_ can be reversibly formed and decomposed on the Pt_58_Fe_42_ cathodes (Fig. 3a–k). Due to the overlap of the O 1s signals of Pt_58_Fe_42_ and Pt with those of O 1s in Li_2_CO_3_ in the XPS spectra (Figs. S14 and S15), Li 1s and C 1s are employed for the analysis of discharge products and byproducts (Fig. S16). In addition, negligible valence variations for Pt and Fe can be detected after the 20th charge–discharge cycle (Figs. S17 and S18), with a uniform distribution of Pt and Fe elements even after the 20th recharge cycle (Fig. 3l, m). Conversely, the high charge polarization of Pt leads to the generation of significant byproducts (Li_2_CO_3_) after the initial charging process (Figs. S19 and S20). UV–Vis absorption spectra further confirmed that the lithium–oxygen battery based on Pt_58_Fe_42_ exhibits a higher Li_2_O_2_ formation efficiency (86%) and a lower residual Li_2_O_2_ ratio (12.5%) after the first discharge–charge cycle compared to Pt (Li_2_O_2_ formation efficiency: 80%; residual Li_2_O_2_: 37.5% in Fig. S21). When the discharge capacity is increased to 1000 mAh g^−1^ (3 mAh cm^−2^), TEM images reveal a uniform distribution of Pt and Fe elements even after the 100th recharge cycle, demonstrating the cycling stability of the Pt_58_Fe_42_ catalyst (Fig. S22).Fig. 3a–h SEM images of 1st discharged (a), 1st charged (b), 5th discharged (c), 5th charged (d), 10th discharged (e), 10th charged (f), 20th discharged (g), 20th charged (h) cathode. i–k XRD patterns (i), Raman spectra (j) and XPS spectra (k) of Pt_58_Fe_42_ cathode at initial, 1st discharged and charged, 5th discharged and charged, 10th discharged and charged, and 20th discharged and charged state. l, m Elemental mapping of C, Pt, and Fe distribution for initial (l) and 10th charged (m) Pt_58_Fe_42_ cathode. n Elemental mapping of Pt, and Fe distribution for 200th charged Pt_58_Fe_42_ cathode. o EDX spectra of 200th charged Pt_58_Fe_42_ nanowires

Furthermore, with a limited capacity of 1 mAh cm^−2^, TEM characterization of the Pt_58_Fe_42_ catalyst after the 200th cycle indicates an increase in the Pt/Fe ratio from 58:42 to 75.8:24.2 (Fig. 3n, o). This change results in poor OER activity and high charge overpotentials, ultimately leading to the incomplete decomposition of Li_2_O_2_ during charging (Fig. S23) and a reduction in the battery's cycling stability.

Electron State Analysis of dz2 Frontier Orbital in Pt58Fe42 Nanowires

The Bader charge and differential charge densities are employed to analyze the OER catalytic activities of the Pt_58_Fe_42_ electrocatalysts. Because Pt is more electronegative than Fe, electron transfer occurs from Fe to Pt, resulting in a reduced valence band energy for Pt in Pt_58_Fe_42_ versus pure Pt (Fig. 4a), as confirmed by XPS measurements (Fig. S24). The lower valence of Pt in Pt_58_Fe_42_ is further supported by the differential charge density, showing that Pt exhibits a higher electron density compared to that of Pt in Pt (Fig. 4b, c). The electron localization function analysis reveals a higher degree of electron localization around Pt in PtFe compared to pure Pt, suggesting electron transfer from Fe to Pt (Fig. S25). Furthermore, Raman spectroscopy of CO adsorbed on Pt and PtFe surfaces indicates that on the PtFe alloy surface, CO preferentially a dsorbs at Pt sites with specific coordination environments. Relative to pure Pt, the d‑electron density at these Pt sites increases due to the electronic effect of Fe (reflected by a lower apparent oxidation state), which enhances π‑back-donation from the metal to CO. This weakens the C–O bond and results in a redshift of the C–O vibrational peak (Fig. 4d). Such lower valence band leads to a downward shift of the Pt 5d band center (−7.72 eV) compared to the − 7.18 eV of pure Pt from ultraviolet photoelectron spectroscopy (UPS) results (Fig. S26). Therefore, the transfer of electrons from Fe to Pt in Pt_58_Fe_42_ causes the d-band center of Fe to shift upward compared to that in pure Pt, while the d-band center of Pt shifts downward relative to pure Pt (Fig. S27). The downward shift of the Pt 5* d*-band center in Pt_58_Fe_42_ indicates a change in the electron occupancy of Pt's five d-suborbitals. For the case of pure Pt, the number of electrons in its five suborbitals dyz, dxz, dxy, dx^2^−y^2^, and dz^2^ is 1.80, 1.80, 1.76, 1.82, and 1.76, respectively (Fig. 4e). In Pt_58_Fe_42_, electron transfers from Fe to Pt involves Pt dxz /yz/xy–Fe dxz/yz/xy, Pt dx^2^−y^2^–Fe dx^2^−y^2^, and Pt dz^2^–Fe dz^2^ orbital interactions (Fig. 4f). The strength of orbital interaction between the Pt and Fe suborbitals primarily depends on the energy gap between their respective d-band centers. A smaller d-band center gap indicates stronger orbital coupling, which facilitates greater electron transfer from the Fe suborbital to the corresponding Pt suborbital. Due to the significantly lower d-band center gap between Pt dz^2^ and Fe dz^2^ (0.3 eV, Fig. 4g), compared to those of Fe dxz–Pt dxz (0.41 eV), Fe dyz–Pt dyz (0.45 eV), Fe dxy–Pt dxy (0.37 eV), and Pt dx^2^−y^2^–Fe dx^2^−y^2^ (0.40 eV), the Pt dz^2^–Fe dz^2^ coupling is the strongest in Pt–Fe interaction (Fig. 4h). Therefore, the electron transfer between Pt and Fe mainly concentrates on Pt dz^2^–Fe dz^2^ (0.16 e^−^), in contrast to those of Fe dxz–Pt dxz (0.04 e^−^), and Fe dyz–Pt dyz (0.01 e^−^), Fe d_xy–Pt dxy_ (0.07 e^−^), Pt d_x^2^−y^2^–Fe dx^2^−y^2^ (0.03 e^−^). Therefore, the d–d-orbital coupling between Pt and Fe in Pt58Fe42 leads to an increase in the electron occupancy of the Pt dz^2^ orbital to 1.92, compared to 1.76 in pure Pt. Compared to pure Pt, the additional electron in the Pt dz^2^ frontier orbital in Pt_58_Fe_42 will participate in the hybridization interaction with fully occupied 5σ interaction of LiO_2_.Fig. 4a Bader charge results of Pt_58_Fe_42_ nanowires. b c The differential charge density (b) and two-dimensional charge density (c) for Pt_58_Fe_42_ nanowires. d Raman spectra of Pt and Pt_58_Fe_42_ under a CO atmosphere. e f DOSs of Pt 5d orbital for dyz/xz, dxy, dx^2^−y^2^ and dz^2^ in Pt (e) and Pt_58_Fe_42_ (f). g The d-band center gap of Pt dxy-Fe dxy, Pt dxz-Fe dxz, Pt dyz-Fe dyz,Pt dx^2^−y^2^ -Fe dx2x2_x^2^−y^2^, and Pt dz^2^-Fe dz^2^ in Pt_58_Fe_42. h The influence of d-band center gap on the coupling strength between Pt and Fe in Pt_58_Fe_42_

Frontier Orbital Interaction between Pt58Fe42 and LiO2

During the interaction between Pt and LiO_2_, the dz^2^ and dxz/dyz orbitals of Pt interact with the 5σ and 2π orbitals of LiO_2_, respectively, leading to the formation of new σ and π bonds. Given that σ bonds are intrinsically stronger than π bonds, modulating the σ-bond interaction (involving the Pt dz^2^ orbital) is more effective in reducing the charge overpotential than tuning the π-bond interaction. Therefore, the dz^2^ orbital of Pt and the 5σ orbital of LiO_2_ are considered to be the frontier orbitals. It should be noted that the modulation of the dz^2^ electron occupancy in Pt within Pt_58_Fe_42_ primarily governs the charging overpotential and thereby the battery's energy efficiency, while contributing minimally to the discharge capacity. Due to the distinction in the electron number of the dz^2^ orbitals in Pt and Pt_58_Fe_42_ cathodes, the orbital interaction between the Pt dz^2^ frontier orbital in Pt (or Pt_58_Fe_42_) and the 5σ frontier orbital of LiO_2_ is further analyzed, which is the key factor determining the OER kinetics. The frontier orbital interaction between the dz^2^ orbital of Pt and the 5σ orbital of LiO_2_ will form a bonding orbital of dz^2^-5σ and an antibonding orbital of dz^2^-5σ** (Fig. 5a, b). The interaction strength between the Pt-based catalyst (Pt or Pt_58_Fe_42_) and LiO_2_ depends on the bond order, which is calculated as (A-B)/2, where A and B represent the electron number in the bonding and antibonding levels, respectively. As shown in Fig. 5c, d, the higher electron count in the d_z^2^ orbital of Pt in Pt_58_Fe_42 leads to more electron occupation in the antibonding level of d_z_^2^-5σ**, resulting in a lower bond order (0.02) between Pt_58_Fe_42_ and the adsorbed LiO_2_, compared to that of pure Pt (0.12). Thermodynamically, this favors the desorption and decomposition of LiO_2_, facilitating a rapid OER on the Pt_58_Fe_42_ surface. In the Pt_58_Fe_42_ catalyst, Pt serves as the primary catalytic site. The main role of Fe is to modulate the electronic state of Pt, thereby fine-tuning the adsorption of reaction intermediates and enhancing the overall catalytic performance. This is further corroborated by theoretical calculations: The Gibbs free energy profile reveals that the charging overpotential for Li_2_O_2_ decomposition on a pure Fe catalyst is as high as 2.0 V. This result explicitly excludes Fe as the active catalytic center (Fig. S28). The weaker adsorption strength between LiO_2_ and Pt_58_Fe_42_ (−5.29 eV) compared to Pt (−5.56 eV) is further evidenced by differential charge density and Bader charge analyses, where a lower electron transfer of 0.74 e^−^ from Pt_58_Fe_42_ to LiO_2_ is found, compared to the transfer of 1.03 e^−^ from Pt (Fig. 5e, f). Moreover, UV–Vis spectra suggest that, in the presence of the Pt_58_Fe_42_ catalyst, the adsorption of O_2_^−^ is higher in the solution of KO_2_ + dimethyl sulfoxide (DMSO), further confirming that Pt_58_Fe_42_ exhibits a weaker adsorption strength toward O_2_^−^ (Fig. 5g). In addition, compared to Pt, Pt_58_Fe_42_ exhibits weaker adsorption not only toward LiO_2_ (as previously noted) but also toward O_2_, Li_2_O_2_, and Li_4_O_4_, demonstrating a similar trend (Figs. 5h–j and S29). The weaker adsorption of LiO_2_ by Pt_58_Fe_42_ results in a lower charge overpotential, as evidenced by Gibbs free energy calculations. During the charging process, the charge overpotential η = Uc − U0, where Uc and U0 represent the charge overpotential and equilibrium potential. In the rate-determining step (Fig. 5l), LiO_2_ decomposes into O_2_ and Li^+^. The weak orbital interaction between LiO_2_ and Pt_58_Fe_42_ results in a lower charge energy barrier of 0.53 eV (Fig. 5l) compared to that of Pt (0.61 eV in Fig. 5k). This finding is in line with the activation energy results obtained from electrochemical impedance spectra (Figs. 5m and S30–S33). As a result, all the aforementioned results suggest that the frontier orbital interactions between the Pt dz^2^ orbital in PtFe and the 5σ orbital of LiO_2_ can effectively enhance the OER kinetics of Li–O_2_ batteries.Fig. 5a, b DOSs curves for the orbital hybridization between Pt 5d and adsorbed LiO_2_ for Pt (a) and Pt_58_Fe_42_ (b). c, d Orbital hybridization between LiO_2_ and Pt (d) or Pt_58_Fe_42_. e, f Differential charge density on Pt/LiO_2_ (e) and Pt_58_Fe_42_-LiO_2_ (f). g UV–Vis spectra for adsorbed Pt and Pt_58_Fe_42_ electrocatalysts with KO_2_/DMSO solution. k**, **l Gibbs free energy profiles of Pt and Pt_58_Fe_42_ at various potentials. h–j Optimized structures of Pt_58_Fe_42_/O_2_^^, Pt_58_Fe_42_/Li_2_O_2_^^ and Pt_58_Fe_42_/2Li_2_O_2_^*^. m Activating energy for the charge process at different capacities by using Pt and Pt_58_Fe_42_ nanowires

We further calculated DOSs and Gibbs free energy profiles for Pt_67_Fe_33_ and Pt_76_Fe_24_ to elucidate why the Pt_58_Fe_42_ catalyst exhibits superior OER performance. The results indicate that as the Pt content increases (from Pt_58_Fe_42_ to Pt_67_Fe_33_ and Pt_76_Fe_24_), the electron population in the Pt 5* d_z^2^ orbital gradually decreases, with values of 1.92 for Pt_58_Fe_42, 1.85 for Pt_67_Fe_33_, and 1.80 for Pt_76_Fe_24_ (Fig. S34). This progressive depletion in the Pt 5 d_z*^2^ orbital leads to a corresponding increase in the adsorption strength of the LiO_2 intermediate, as evidenced by the more negative adsorption energies: −5.29 eV for Pt_58_Fe_42_, −6.17 eV for Pt_67_Fe_33_, and −6.41 eV for Pt_76_Fe_24_ (Fig. S35). Moreover, the weakening interaction with LiO_2_ across the series from Pt_58_Fe_42_ to Pt_76_Fe_24_ corresponds to a decline in OER activity, which is reflected in the increasing overpotentials derived from the Gibbs free energy diagrams: 0.53 V for Pt_58_Fe_42_, 3.08 V for Pt_67_Fe_33_, and 3.41 V for Pt_76_Fe_24_ (Figs. S36 and S37).

To further establish the universality of Pt dz^2^ orbital electrons as a descriptor for OER activity, we supplemented our study with other Pt-based alloy catalysts. Using PtCo (Pt_82_Co_18_, Pt_70_Co_30_, Pt_64_Co_36_) and PtCu (Pt_78_Cu_22_, Pt_68_Cu_32_, Pt_59_Cu_41_) as typical examples, we verified the correlation between Pt dz^2^ orbital electrons and OER performance. The results consistently show that for both PtCo and PtCu systems, OER activity gradually decreases with a reduction in the number of Pt dz^2^ orbital electrons (Figs. S38–S43). This demonstrates that the identified relationship is not limited to PtFe but extends to other Pt-based alloys, confirming that Pt dz^2^ orbital electrons can serve as a universal descriptor for OER activity across Pt-based catalysts.

Conclusions

In conclusion, we have constructed a Pt-based model catalyst to investigate the relationship between the number of Pt dz^2^ frontier orbital electrons and OER activity for Li–O_2_ batteries. Subsequently, when the dz^2^ frontier orbital of Pt in the PtFe catalyst interacts with the 5σ frontier orbital of LiO_2_, the excess electrons in the Pt dz^2^ orbital occupy the antibonding orbital, thereby weakening the interaction with LiO_2_ and ultimately enhancing the OER catalytic activity. This work established a correlation between the number of electrons in the dz^2^ frontier orbital and OER activity, suggesting that the electron number in the dz^2^ frontier orbital can serve as a descriptor for OER activity.

Supplementary Information

Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 44920 kb)