Cascaded orbital–oriented hybridization of intermetallic Pd₃Pb boosts electrocatalysis of Li-O₂ battery

Significance

Li-O₂ batteries with a high theoretical energy density are promising for powering electrical vehicles. However, their practical application is restricted by the sluggish oxygen evolution reaction (OER) kinetics that associates with the high overpotentials. For pure Pd, its 4d level is usually too close to Fermi level, which poses excessively orbital overlaps between Pd and intermediate (LiO₂), thus leading to high energetic barriers. Herein, we report a general strategy of cascaded orbital-oriented hybridization to boost the OER electrocatalytic activity of intermetallic Pd₃Pb. We demonstrate that the cascaded orbital–oriented hybridization in Pd₃Pb brings lower energy levels of antibonding states and attenuated orbital interaction with LiO₂, delivering a low OER overpotential.

Abstract

Catalysts with a refined electronic structure are highly desirable for promoting the oxygen evolution reaction (OER) kinetics and reduce the charge overpotentials for lithium–oxygen (Li-O₂) batteries. However, bridging the orbital interactions inside the catalyst with external orbital coupling between catalysts and intermediates for reinforcing OER catalytic activities remains a grand challenge. Herein, we report a cascaded orbital–oriented hybridization, namely alloying hybridization in intermetallic Pd₃Pb followed by intermolecular orbital hybridization between low-energy Pd atom and reaction intermediates, for greatly enhancing the OER electrocatalytic activity in Li-O₂ battery. The oriented orbital hybridization in two axes between Pb and Pd first lowers the d band energy level of Pd atoms in the intermetallic Pd₃Pb; during the charging process, the low-lying 4d_xz/yz and 4d_z² orbital of the Pd further hybridizes with 2π* and 5σ orbitals of lithium superoxide (LiO₂) (key reaction intermediate), eventually leading to lower energy levels of antibonding and, thus, weakened orbital interaction toward LiO₂. As a consequence, the cascaded orbital–oriented hybridization in intermetallic Pd₃Pb considerably decreases the activation energy and accelerates the OER kinetics. The Pd₃Pb-based Li-O₂ batteries exhibit a low OER overpotential of 0.45 V and superior cycle stability of 175 cycles at a fixed capacity of 1,000 mAh g⁻¹, which is among the best in the reported catalysts. The present work opens up a way for designing sophisticated Li-O₂ batteries at the orbital level.

The aprotic lithium–oxygen (Li-O₂) batteries are considered an ideal power source for electric vehicles due to their ultrahigh theoretical energy density of ~3,500 Wh kg⁻¹ (1–6). During the operation of Li-O₂ batteries, the discharge product of Li₂O₂ is generated and decomposed through oxygen reduction reaction (ORR, discharge process) and oxygen evolution reaction (OER, charge process), respectively (7–11). However, the sluggish OER kinetics, associated with the insulating nature of Li₂O₂, lead to high charging overpotentials and inferior cycling performance, which has seriously restricted the development of Li-O₂ batteries (12–19). Though Li-O₂ batteries with liquid catalysts (soluble molecules) can obviously decrease the OER overpotentials, the intrinsic decomposition (9) and shuttling effect (20) of redox mediators will degrade the cycle performances. Alternatively, various solid catalysts, including transition metal oxides (21–27), single-atom catalysts (28–30), perovskites (31–33), transition metal sulfides (34–36), and noble metals (37–41), have been explored for use in Li-O₂ batteries. Among these options, palladium (Pd) has shown particularly promising OER electrocatalytic performance, which is attributed to its tunable 4d orbital structure that allows for tailoring the coordination environment and surface structure (39, 42–45). For pure Pd, however, its 4d level is too close to the Fermi level, which poses excessive orbital overlaps between Pd and the intermediate (LiO₂), thus giving rise to high energetic barriers. Such orbital overlap can be manipulated by electronic tuning via alloying Pd with other elements. However, rational alloying necessitates an orbital-level understanding of the electronic structure–activity relationship, which remains elusive for OER in Li-O₂ batteries, with a main challenge lying in the ill-defined structural nature of prevailing Pd-based solid-solution alloys. We took the view that an intermetallic with a well-defined atomic position serves as an ideal platform for establishing the structure–activity relationship at the orbital level, which is crucial for the rational design of OER electrocatalysts in Li-O₂ batteries.

Herein, we report the cascaded orbital–oriented hybridization of intermetallic Pd₃Pb as a general strategy to greatly boost the OER electrocatalytic activity for Li-O₂ battery. While Pd₃Pb is widely utilized in various electrocatalytic applications such as ORR (46, 47), ethanol oxidation (48), NH₃ synthesis (49), hydrogen peroxide synthesis (50), and methanol oxidation reaction (51), this study presents a concept of cascaded orbital–oriented hybridization in intermetallic Pd₃Pb for enhancing the OER process in Li-O₂ batteries. This approach differs significantly from conventional tuning strategies employed in Li-O₂ batteries (30, 36–38, 52–54) and capitalizes on the unique electronic structure and bonding properties of Pd₃Pb to create a hybridized orbital cascade that enhances its catalytic activity toward OER. The density of state (DOS) analysis coupled with X-ray photoelectron spectroscopy (XPS) results indicates that the Pd 4d_xz/yz and 4d_z² orbitals directionally hybridize with Pb 6P_x/y and 6P_z, respectively, resulting in a low d band energy level of Pd atoms in intermetallic Pd₃Pb relative to that of the pure Pt metal. The 4d_xz/yz and 4d_z² orbitals of the Pd atom in Pd₃Pb with a low d state energy level will further hybridize with 2π* and 5σ orbitals of LiO₂ molecules during the galvanostatic cycles. The cascaded hybridization leads to lower energy levels of bonding/antibonding states and, thus, the attenuated orbital interaction toward LiO₂, which results in lower OER activation energy. As a consequence, the intermetallic Pd₃Pb displays a low OER overpotential (0.45 V) and excellent cycle stability (175 cycles) at a restricted capacity of 1,000 mAh g⁻¹. The present work provides universal insights to construct the effective catalysts at the orbital level by tandem orbital-oriented hybridization of noble metals.

Results

Synthesis and Structural Characterization of Intermetallic Pd₃Pb.

Pd₃Pb nanocubes were synthesized by reducing palladium(II) acetylacetonate and lead(II) acetate in oleylamine (Details in SI Appendix). The X-ray diffraction (XRD) pattern of Pd₃Pb nanocubes confirms their pure intermetallic phase without Pd or Pb impurities (SI Appendix, Fig. S1). The high-resolution transmission electron microscopy (HR-TEM in Fig. 1A) result reveals that the Pd₃Pb nanocube has a diameter of 10 to 15 nm. The atomic resolution high-angle annular dark-field scanning TEM (HAADF-STEM) (Fig. 1B) shows that intermetallic Pd₃Pb has a lattice spacing of 0.20 nm, ascribed to the (200) facets of the intermetallic Pd₃Pb. The Fourier transform (FFT) of the selected area in Fig. 1B shows different facets along the [001] zone axis (Fig. 1C), being in accordance with the XRD results (SI Appendix, Fig. S1). Both HAADF-STEM image and alternating intensity profiles further confirm the formation of ordered intermetallic Pd₃Pb nanocrystals (Fig. 1 D and E) with uniform distribution of Pd and Pb (Fig. 1F). The uniformly distributed Pd and Pb elements in the intermetallic Pd₃Pb nanocubes are further demonstrated by the TEM linescan profile and energy dispersive X-ray mappings (Fig. 1 G and H).

Fig. 1.

The characterization of intermetallic Pd₃Pb. (A and B) TEM image (A) and atomic-resolution HAADF-STEM image (B) of Pd₃Pb nanocubes. (C) The FFT pattern of Pd₃Pb nanocube from the white dash box in (B). (D) The enlarged HRTEM image of intermetallic Pd₃Pb from the white dash box in (B). (E) Integrated pixel intensity of Pd₃Pb images from green and blue dash boxes in (B). (F) Schematic illustration of the molecular structure for Pd₃Pb nanocrystals. (G) Linear elemental distribution of Pd and Pb for the Pd₃Pb nanocubes. (H) The elemental mappings of Pd and Pb in intermetallic Pd₃Pb nanocubes.

Electrocatalytic Performances of Intermetallic Pd₃Pb.

The Pd₃Pb nanocubes were then employed as the cathode catalysts to evaluate the electrochemical performance for Li-O₂ batteries, where the Li foil anode is used as a counterelectrode. The OER overpotential is defined when 50% of discharge products is decomposed. During the deep discharge and charge processes, the Pd₃Pb catalyst displays a high discharge capacity of 7,746 mAh g⁻¹ and a low charge overpotential of 0.96 V (Fig. 2A). As a comparison, the Pd cathode demonstrates a lower discharge capacity of 4,953 mAh g⁻¹ and a higher OER overpotential of 1.26 V due to the poor catalytic activities. When the cutoff capacity is restricted to 1,000 mAh g⁻¹ at 0.1 A g⁻¹, the Pd₃Pb-based Li-O₂ cell exhibits a lower OER overpotential of 0.45 V than Pd cathode–based one (1.11 V), confirming superior OER catalytic activity of the Pd₃Pb electrode (Fig. 2B).

Fig. 2.

Electrocatalytic performances of Pd and intermetallic Pd₃Pb. (A) The deep discharge and charge profiles of Pd and Pd₃Pb cathodes. (B) The discharge and charge curves of Pd and Pd₃Pb electrodes with a cutoff capacity of 1,000 mAh g⁻¹ at 0.1 A g⁻¹. (C) The CVs of Pd and Pd₃Pb cathodes at 0.05 mV s⁻¹ from 2.0 and 4.5 V. The Inset is the exchange current density of Pd and Pd₃Pb cathodes. (D and E) The cycle performances of Pd and Pd₃Pb electrodes with a fixed capacity of 1,000 mAh g⁻¹ at 0.5 A g⁻¹.

The overpotential of Pd₃Pb electrocatalyst-based Li-O₂ batteries is among the best in all the reported Li-O₂ batteries using solid electrocatalysts such as noble metals, single atom and transition metal oxide, etc (SI Appendix, Table S1). Meanwhile, it should be noted that the flat charging curve of Pd₃Pb is mainly due to the oxidation of Li₂O₂ with negligible decomposition of solvent and electrolyte, which will be discussed below. Similarly, the excellent OER catalytic performance of Pd₃Pb nanocubes for the decomposition of the discharge products is further proved by the cyclic voltammograms (CVs) at 0.05 mV s⁻¹ from 2.0 to 4.5 V since Pd₃Pb exhibits a more negative onset potential for OER than Pd (Fig. 2C). Moreover, the exchange current density of Pd₃Pb (36.81 μA cm⁻²) is higher than that of the Pd electrode (22.59 μA cm⁻²), confirming the fast OER kinetics for Pd₃Pb-based Li-O₂ batteries.

The rate ability of Pd₃Pb and Pb catalysts was investigated at a constant capacity of 1,000 mAh g⁻¹ from 0.1 to 1.0 A g⁻¹. Even at a high current density of 1 A g⁻¹, the OER overpotential of Pd₃Pb is lower than 1.1 V, better than Pd with charge overpotentials rapidly increasing even at 0.5 A g⁻¹ (SI Appendix, Figs. S2 and S3). Furthermore, with a fixed capacity of 1,000 mA h g⁻¹, there is no sharp discharge and charge voltage degradation for the Pd₃Pb cathodes even after 175 discharge–charge cycles (Fig. 2E), whereas the Pd cathode can only be maintained for 105 cycles due to the serious OER polarization (Fig. 2D). Although PdMo bimetallene shows a low charge overpotential for Li-air battery (55), the Pd₃Pb herein demonstrates excellent long-cycle stability. The satisfied cycle stability of Pd₃Pb electrode shows obvious advantages over previous reports (39, 41, 56–60).

Microstructure Analysis of Intermetallic Pd₃Pb during the ORR and OER.

To reveal the OER catalytic role of the Pd₃Pb electrode, the morphologies of discharged Pd₃Pb and Pb electrodes were tracked by scanning electron microscopy (SEM). Disc-shaped products with a diameter of approximately 5 μm were formed on the Pd electrode upon discharge (SI Appendix, Fig. S4) and were confirmed to be Li₂O₂ through XRD, Fourier-transform infrared (FTIR) spectra, and Raman results (SI Appendix, Figs. S5−S7). Meanwhile, the bowl-shaped discharged products were observed on the Pd₃Pb electrode (Fig. 3A). Although minor by-products could be detected from FTIR results due to the decomposition of solvent (SI Appendix, Fig. S8A) after the first discharge–charge process, the absence of by-products from the results of XRD, XPS, Raman spectra, and NMR could confirm the nearly reversible formation and decomposition of Li₂O₂ (Fig. 3 G and H and SI Appendix, Figs. S8B and S9). The ultraviolet–visible (UV-Vis) absorption spectra data indicate that Li-O₂ batteries featuring Pd₃Pb cathodes can achieve a high Li₂O₂ formation efficiency (68.6%) and a low residual Li₂O₂ (30.0%) even after the 20th discharge and charge process, respectively. These results point to excellent reversibility and are illustrated in SI Appendix, Fig. S10. The SEM morphologies, XPS spectra, and XRD techniques were employed to further study the long-cycle reversibility of Pd₃Pb for Li-O₂ battery under the different discharge and charge states. As shown in Fig. 3 A–F, the discharge products of Li₂O₂ can be reversibly formed on the Pd₃Pb electrode even after 10 galvanostatic cycles, demonstrating the superior reversibility. A similar phenomenon can also be proved by XPS spectra and the XRD pattern. Because the O 1s signal of by-products (Li₂CO₃) overlaps with those of the initial Pd and Pd₃Pb electrodes (SI Appendix, Figs. S11 and S12), Li 1s and C 1s spectra are used to analyze the discharge and parasitic products. Impressively, the reversible formation and oxidation of discharged products (Li₂O₂) can be achieved for the Pd₃Pb nanocubes even after 50 discharge–charge cycles (Fig. 3 G and H). Furthermore, Pd and Pb are uniformly distributed with ignorable valence change for the reversible catalysis of Li₂O₂ after 20 galvanostatic cycles, further proving the excellent stability of the Pd₃Pb catalyst (SI Appendix, Figs. S13–S16). As a comparison, typical by-products of Li₂CO₃ and residual Li₂O₂ can be observed on the Pd electrode after the first recharge procedure, as shown in SI Appendix, Figs. S17 and S18 through analysis of Li 1s and C 1s spectra.

Fig. 3.

Microstructure analysis of intermetallic Pd₃Pb during the ORR and OER. (A–F) SEM images of 1st discharged (A), 1st charged (B), 5th discharged (C), 5th charged (D), 10th discharged (E), and 10th charged (F) Pd₃Pd electrodes. The Inset on the Left Top of (A) is the enlarged SEM images of the discharged Pd₃Pd electrodes, while the Inset on the Right Top of (A) is a schematic illustration of discharged products on Pd₃Pb electrodes. (G and H) XRD patterns (G) and XPS spectra (H) of the Pd₃Pb electrode under different states.

Intramolecular p-d Hybridization between Pd and Pb in Pd₃Pb.

To deeply illuminate the redox process and explain the superior catalytic activities of Pd₃Pb, density functional theory (DFT) calculations were conducted to analyze the charge states of Pd atoms on both pure Pd and intermetallic Pd₃Pd surfaces. For the intermetallic Pd₃Pb nanocubes, there is a typical p-d orbital-oriented hybridization between Pd 4d and Pb 6p. According to the orbital symmetry, the d_z² of Pd 4d can orientedly hybridize with the p_z orbital of Pb 6p to form σ and σ* orbitals, while the d_xz/yz orbital of Pd 4d can directionally interact with the p_x/y orbital of Pb 6p to generate the π and π* orbitals (Fig. 4 A–C ). Meanwhile, the d_xy and d_x²_−y² orbitals of Pd 4d are considered to be nonbonding due to mismatched orbital orientation with Pb 6p orbitals according to previous reports (61). The orbital-oriented p-d hybridization results in the electron transfer from Pd to Pb, thereby causing a high oxidation state of Pd atoms on the Pd₃Pb surface, which can be confirmed by XPS spectra. The Pd 4f spectrum of Pd₃Pb catalysts shifts to a high binding energy by 0.2 eV compared to that of pure Pd metal (Fig. 4D), consistent with the electron transfer by Bader charge results (Fig. 4E). The oxidized Pd on the Pd₃Pb entails the low d band energy level of Pd, which can affect the orbital interaction between Pd₃Pb and the intermediates, as discussed below.

Fig. 4.

Intramolecular p-d hybridization between Pd and Pb in Pd₃Pb. (A) The typical d-p orbital hybridization distribution pattern of intermetallic Pd₃Pb. (B) DOSs of Pd d_z² and d_xz/_yz orbitals and Pb p_x, p_y, and p_z orbitals after d-p orbital hybridization. (C) The energy level of the Pd-Pb orbital hybridized density of states. (D) XPS spectra of Pd and Pd₃Pb in the Pd 4f region. (E) The Bader charge analysis for Pd₃Pb nanocubes. The negative value means higher electron density, while the positive value means lower electron density at the (111) plane.

Intermolecular Orbital Hybridization between the Pd Atom in Pd₃Pb and LiO₂.

According to previous reports, the OER overpotentials of Li-O₂ batteries are mainly determined by two factors: 1) the Li₂O₂/catalyst contact interface (35) and 2) the interaction between electrocatalysts and intermediates (28). Since the large-diameter Li₂O₂ are deposited on both Pd and Pd₃Pb during the discharge process (bowl shape for Pd₃Pb and disc shape for Pd), there is no obvious difference in the Li₂O₂/catalyst contact interface between Pd and Pd₃Pb catalysts. Hence, the catalytic activities of Li-O₂ batteries mainly depend on the interaction between LiO₂ and electrocatalysts that can tune the OER energy barrier and charge overpotential. With the orbital-oriented hybridization between Pd 4d and Pb 6p, the oxidized Pd in intermetallic Pd₃Pb (4d_z², 4d_xz, 4d_yz) can further hybridize with LiO₂ (5σ and 2π*) for improved OER catalytic activities. To be specific, the oxidized Pd with low-lying d orbitals in intermetallic Pd₃Pb due to the intramolecular d-p hybridization shows a lower d-band center of Pd (−0.58 eV) than that of the pure Pb (−0.48 eV), leading to a lower orbital overlap to form the bonding and antibonding orbitals for 4d_z2-5σ and 4d_xy/d_yz-2π* orbital hybridization (Fig. 5 A and B). Therefore, the lower 4d energy level of Pd atoms in intermetallic Pd₃Pb causes a high electron occupation in the antibonding orbitals, resulting in weaker orbital interaction toward LiO₂ (Fig. 5 C and D). It can be demonstrated that the adsorption strength between Pd₃Pb and LiO₂ (−1.69 eV) is lower than that between Pd and LiO₂ (−2.31 eV). The weak orbital interaction between Pd₃Pb and LiO₂ is also directly confirmed by the charge density difference, where the charge transfer from intermetallic Pd₃Pb to LiO₂ is less than that of the Pd cathode (Fig. 5 E and F). The interaction between superoxide and catalysts (Pd and Pd₃Pb) was investigated by the UV-Vis absorption spectra where the absorption peak around 255 nm is attributed to superoxide species. Higher absorbance of superoxide species could be observed in the presence of Pd₃Pb catalysts (SI Appendix, Fig. S19), indicating the weak adsorption strength between Pd₃Pb and LiO₂. This weak orbital interaction between LiO₂ intermediates and Pd₃Pb nanocubes is beneficial to the dominant OER catalytic activities in Li-O₂ batteries.

Fig. 5.

Intermolecular orbital hybridization between the Pd atom in Pd₃Pb and LiO₂. (A and B) The DOS for Pd 4d d_yz/d_z²/d_xz orbitals and adsorbed LiO₂ orbitals for Pd (A) and Pd₃Pd (B). (C and D) Schematic illustration of orbital interactions between LiO₂ (5σ and 2π*) and 4d orbital for Pd and Pd₃Pb catalysts. (E and F) Differential charge density on Pd-LiO₂ (E) and Pd₃Pb-LiO₂ (F) with an isosurface level of 0.002 e Å⁻³. The cyan and yellow represent electron depletion and accumulation, respectively.

Gibbs Free Energy and Activation Energy Analysis.

Analysis of the Gibbs free energy with the optimized structures of Pd₃Pb and Pd was further conducted to investigate the OER performance intuitively. The OER overpotential is determined as η = U_RC–U_eq, where the U_RC and U_eq stand for the charge and equilibrium voltage, respectively. In addition, U₀ represents the free energy profiles at a zero potential. During the charge process, the superoxide species can be confirmed as the intermediates for both Pd and Pd₃Pb cathodes by the UV-Vis spectra, which proves the two-step decomposition process of Li₂O₂ (SI Appendix, Figs. S20 and S21). For the OER process, while the energy barrier between Li₂O₂ and (Li₂O₂)₂ is greater than the gap between LiO₂ and O₂, the oxidation of LiO₂ to release O₂ remains a potential determining step due to the high ratio of the energy gap to the electron transfer number. The weak orbital interaction toward LiO₂ in the intermetallic Pd₃Pb causes a lower OER energy barrier (0.41 eV in Fig. 6A) compared to that of the Pd cathode (0.89 eV in Fig. 6B). As a consequence, the OER overpotential of Pd₃Pb (0.41 V in Fig. 6A) is lower than that of the pure Pd metal (0.89 V in Fig. 6B). The calculated energy barrier is also consistent with the activation energy results (E_a in Fig. 6C and SI Appendix, Figs. S22–S25), where the Pd₃Pb catalyst shows lower E_a values than the Pd electrode at various charge capacities. The above results highlight that the cascaded orbital–oriented hybridization from intramolecular p-d hybridization between Pd and Pb in intermetallic Pd₃Pb to intermolecular orbital hybridization between 4d_z²/4d_xz/4d_yz of Pd₃Pb and 5σ/2π* of LiO₂ can directly clarify the enhanced OER catalytic performance (Fig. 6D).

Fig. 6.

Gibbs free energy and activation energy analysis. (A and B) The Gibbs free energy profiles at zero, equilibrium, and charge potentials for Pd₃Pb (A) and Pd (B) cathodes. The Insets in (A) and (B) are the optimized molecular structures of the Pd3Pb and Pd cathodes, respectively, with adsorbates at various steps. (C) The change of activating energy for the OER process at different capacities by using Pd₃Pb and Pd. (D) Schematic illustration of cascaded orbital–oriented hybridization from intramolecular p-d hybridization between Pb and Pd in intermetallic Pd₃Pb to intermolecular orbital hybridization between Pd₃Pb (Pd 4d_xz/yz and 4d_z²) and LiO₂ (2π* and 5σ) with d_z² at the z axis as the example.

Discussion

To summarize, unconventional cascaded orbital–oriented hybridization of the intermetallic Pd₃Pb is reported to act as an advanced cathode catalyst for boosting OER kinetics in Li-O₂ batteries. We find that the intramolecular p-d orbital-oriented hybridization in two directions, including Pb 6P_z-Pd 4d_z² hybridization in z axes and Pb 6p_x/y-Pd 4d_xz/_yz hybridization in the perpendicular direction, can lead to a high oxidation state of Pd atoms with lower d orbital energy levels. In particular, we find that during the OER process, there is a further intermolecular orbital hybridization between 4d_xz/4d_yz/4d_z² orbitals of the oxidized Pd atoms and the 2π*/5σ orbitals of LiO₂, leading to decreased energy levels of antibonding orbitals and energy overlaps, thereby decreasing the activation energy to achieve a low OER overpotential. This study provides a general strategy to construct an efficient oxygen cathode by tuning the electronic configuration of noble metal–based intermetallics at the orbital level.

Materials and Methods

Synthesis of Pd Nanocrystals.

Palladium diacetylacetonate [Pd(acac)₂, 75 mg] and oleylamine (OAm, 15 mL) were mixed under a nitrogen flow before heating at 60 °C for 10 min to form the colorless Pd-OAm solution. Then, 300 mg of bromothymol blue was quickly injected into the Pd-OAm solution after being solvated in a minimum amount of OAm (~4 mL), which brings a visible color change from colorless to brown–black. Then, the temperature was raised to 90 °C with a heating rate of 3 °C/min, and the solution was cooled down to room temperature after being kept at 90 °C for 60 min. The product was collected by adding 30 mL ethanol and separated by centrifugation (8,000 rpm for 8 min) before dispersing in hexane.

Synthesis of Pd₃Pb Nanocubes.

OAm (5 mL), Pd(acac)₂ (7.6 mg), lead acetate [Pb(OAc)₂, 3.2 mg], and hexadecyl trimethyl ammonium bromide (CTAB, 36 mg) were mixed in a vial by sonicating for 2 h before transferring into a flask in an oil bath at 160 °C for another 5 h. The product was then separated with the same process as that of Pd nanocrystals before dispersing in hexane.

Preparation of Pd₃Pb/C and Pd/C.

The as-synthesized Pd₃Pb nanocrystals (40 mg) were dried and redispersed in cyclohexane (30 mL) before further mixing with Vulcan XC72R carbon (10 mg) under ultrasonication (60 min). Then, the product was collected by centrifugation and washed with mixed cyclohexane/ethanol (9/1, v/v) solvent. Afterward, the mixture was kept stirring overnight at 70 °C to remove the residual surfactants by adding acetic acid (40 mL). The loading mass of Pd₃Pb on the Pd₃Pb/C is 80 wt%. The preparation process of Pd/C is the same as that of Pd₃Pb/C.

Characterizations.

XRD patterns of Pd and Pd₃Pb were obtained on an X-ray powder diffractometer (Rigaku D/max-2400) where Cu Kα (λ = 1.5406 Å) is employed as the radiation source. The morphologies of as-prepared Pd and Pd₃Pb catalysts were investigated by HR-TEM (Tecani-G2 T20). The discharge and charge products of Pd and Pd₃Pb catalysts were characterized by SEM (HITACHI S-4800), XPS, FTIR (PerkinElmer Spectrum II FTIR Spectrometer), Raman (WITec RAMAN alpha 300R), NMR (Bruker), and UV-vis spectra (Hitachi UH4150). The valence information of the initial and discharged Pd₃Pb (or Pd) cathode was analyzed by XPS in different regions including Pd 3d, Li 1s, O 1s, and C 1s (XPS, Escalab 250XI system, Thermo Fisher Scientific, USA). The UV-Vis spectra (Hitachi UH4150) were used for Li₂O₂ quantification at 409 nm.

Battery Assembly and Its Performance.

Pd₃Pb/C (90 wt%) was mixed with a binder [poly(vinylidene fluoride), 10 wt%] in 1-methyl-2 pyrrolidone (NMP) to obtain a slurry. The slurry was heated at 60 °C for 24 h after being deposited onto the carbon paper for the preparation of Pd₃Pb/C cathodes. The Pd/C cathode was prepared with the same content as of the binder for the parallel experiments. The loading mass of Pd₃Pb (Pd) catalysts is close to 0.20 mg/cm² based on the total mass of the Pd₃Pb/C (Pd/C) cathode. The whole mass of the Pd₃Pb/C (Pd/C) cathode also determines the discharge capacity. The Pd₃Pb/C (Pd/C) cathode was utilized together with a Li anode, a separator, and electrolytes (1 M lithium trifluoromethanesulfonate in tetraethylene glycol dimethyl ether) for the assembly of coin cells (2032) under the argon atmosphere in a glove box. Those coin cells based on Pd₃Pb/C and Pd/C cathodes were investigated by a LAND cell system to obtain the full and limited discharge–charge curves. The CVs were analyzed from 2.0 to 4.5 V with a scan rate of 0.05 mV s⁻¹ by an electrochemical workstation (CHI760e). The electrochemical impedance spectroscopy was collected by the same instrument with the voltage amplitude fixed at 5 mV and frequency ranging from 10⁻² to 10⁶ Hz.

DFT Calculations.

DFT as the popular and versatile method within the framework of the Vienna Ab initio simulation package was implemented for all the quantum chemical calculations. The generalized gradient approximation under the Perdew–Burke–Ernzerhof form was used to describe the exchange-correlation potential, while the projector augmented-wave method was employed to treat the interactions between ion cores and valence electrons. The variation in energy smaller than 10⁻⁵ eV was achieved when the plane-wave cutoff energy was set to be 500 eV, and structural models were relaxed until the Hellmann–Feynman forces is smaller than −0.02 eV/Å. During the relaxation, the Brillouin zone was represented by a Γ-centered k-point grid of 3×4×1 for Pd₃Pb, 4×4×1 for Pd, and 3×3×1 for Pb. To avoid the interaction between the periodic images for two nearest neighbor unit cells, the vacuum was set to 20 Å in the z-direction. Different adsorbates on catalysts were optimized with various configurations, and the adsorption energy (ΔE) calculations of *LiO₂ were determined as

$$\mathrm{\Delta E}={\mathrm{E}}_{\mathrm{ad}}-{\mathrm{E}}_{\mathrm{slab}}-\mathrm{E}{}_{{\mathrm{LiO}}_2},$$

where E_ad is the energy of the catalysts with adsorbates (LiO₂) on the surface, E_slab is the energy of pure catalysts with slab structure, and E_LiO2 is the energy of LiO₂ molecule. The overpotential (η) of OER was calculated as η = U_C-U₀, where U_C is the lowest potential that charging is energetically downhill for all the steps, and U₀ is the equilibrium potential.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.