Surface Gradient Doping Enables High‐Capacity and Long‐Life Manganese‐Based Prussian Blue Cathodes for Sodium‐Ion Batteries

ABSTRACT

Manganese‐based Prussian blue analogues (MnPBAs) are attractive cathodes for sodium‐ion batteries (SIBs) due to their high operating voltage, low cost, and sustainability. However, their practical deployment is limited by rapid capacity decay caused by Jahn–Teller distortion and phase‐transition‐induced lattice strain, which trigger Mn dissolution and structural collapse during cycling. Herein, we introduce a surface Cu concentration‐gradient‐doped MnPBA (Cu@MnPBA) synthesized via a two‐step coprecipitation strategy to stabilize the MnPBA framework while preserving high capacity. A ∼200 nm Cu‐gradient surface layer effectively suppresses Jahn–Teller distortion, reduces lattice strain, and inhibits Mn dissolution, while an outer ∼10 nm Cu‐rich amorphous layer enhances Na⁺ surface diffusion and suppresses interfacial side reactions. Unlike conventional coating strategies, the gradient‐doping architecture eliminates interfacial mismatch, ensuring robust structural integrity during prolonged cycling. As a result, Cu@MnPBA delivers a high specific capacity of 143.9 mAh g⁻¹ and retains 78% of its capacity after 500 cycles at 100 mA g⁻¹. When paired with a hard carbon anode in a pouch cell configuration, the system maintains 75% capacity after 3000 cycles, demonstrating exceptional durability under practical operating conditions. This work establishes gradient surface doping as an effective design principle for realizing high‐capacity, long‐life Prussian blue cathodes for SIBs.

Modification mechanism of MnPBA via Cu gradient doping: formation of an amorphous layer for fast Na⁺ transport, suppression of the Jahn–Teller effect, and enhanced capacity.

1. Introduction

Over the past few decades, lithium‐ion batteries (LIBs) have dominated the market for secondary energy storage; however, the limited availability of lithium resources poses a significant challenge to their large‐scale deployment [1]. In contrast, sodium‐ion batteries (SIBs) have emerged as promising alternatives owing to their low cost, abundant resources, favorable low‐temperature performance, enhanced safety, and environmental friendliness [2, 3, 4]. The cathode material plays a pivotal role in determining the electrochemical performance of SIBs [5]. To date, three major classes of cathode materials have been extensively investigated: polyanionic compounds [6], transition metal layered oxides [7], and Prussian blue analogs (PBAs) [8]. Among them, PBAs are particularly attractive because of their open framework structure, high theoretical capacity, and environmentally friendly synthesis routes, rendering them well‐suited as cathodes for SIBs [9].

The general chemical formula for PBAs is represented as A_{2− x} M[Fe(CN)₆]_{1− y} ·zH₂O, where A denotes alkali metal ions, and M represents transition metals [8]. Among various PBAs, Mn‐based PBA (MnPBA) is particularly attractive because it features two redox couples (Mn²⁺/Mn³⁺ and Fe²⁺/Fe³⁺), enabling a high theoretical specific capacity (∼170 mAh g⁻¹) and a high operating voltage (∼3.45 V) [10]. However, Na⁺ extraction converts Mn²⁺ to Jahn–Teller–active Mn³⁺ (t_2g ³e_g ¹), causing significant lattice expansion [9]. Repeated cycling then subjects the lattice to continuous strain, which accumulates mechanical stress within the particles and ultimately leads to cracking or fragmentation (Figure 1a) [11]. Particle fracture accelerates Mn dissolution. The dissolved Mn species migrate to the anode, inducing irreversible capacity loss, and simultaneously catalyze electrolyte decomposition at the cathode surface [12]. This process consumes active Na and electrolyte, releases harmful gases such as (CN)₂, and further degrades electrochemical performance [13]. Moreover, the decomposition products form an organic‐rich cathode electrolyte interphase (CEI) with low ionic conductivity and poor mechanical stability, hindering Na⁺ transport and increasing polarization [14]. As a result, rapid capacity fading and shortened cycle life are observed. Therefore, suppressing the Jahn–Teller distortion and the associated structural and interfacial degradation is essential for the practical application of MnPBA.

FIGURE 1.

Schematic illustrations highlighting the challenges associated with a) pristine MnPBA, b) elemental doping, and c) heterogeneous coating, as well as the advantages of d) Cu surface gradient doping.

To address these challenges, conventional strategies have mainly focused on elemental doping [15, 16, 17, 18, 19, 20, 21, 22]. Theoretically, the coordination strength between N and transition metals in PBAs follows the order: Mn–N < Co–N < Fe–N < Ni–N < Zn–N < Cu–N [23]. Owing to its relatively weak coordination, the Mn–N bond is susceptible to dissociation, which accelerates Mn dissolution and subsequent structural degradation. However, such dopant elements are typically electrochemically inert, and while a doping amount of at least 5% is generally required to ensure satisfactory capacity retention after 500 cycles, it reduces the density of redox‐active sites and leads to a pronounced loss (>20 mAh g⁻¹) of reversible capacity (Figure 1b) [17, 18, 19, 24, 25, 26, 27]. Alternatively, thin surface coating with various materials (e.g., NiHCF [28], CoB [29], polypyrrole (ppy) [30], and sodium alginate (SA) [12]) has shown promise in suppressing interfacial side reactions and mitigating Mn dissolution without sacrificing reversible capacity. However, lattice mismatch between the coating layer and the bulk PBA generates interfacial stress, which can cause cracking of the coating and a gradual loss of protective function during prolonged cycling (Figure 1c). Moreover, most coating materials exhibit insufficient ionic conductivity to support efficient Na⁺ transport and lack long‐term resistance to electrolyte corrosion [31, 32]. As a result, achieving PBAs that simultaneously deliver high capacity and long‐term cycling stability in SIBs remains highly challenging, and to date, has yet to be effectively demonstrated.

Herein, we adopt a surface concentration‐gradient doping design, enabling the enhancement of structural stability in MnPBA without compromising capacity. As illustrated in Figure 1d, a Cu gradient doping with a thickness of ∼200 nm is engineered on MnPBA via a two‐step co‐precipitation strategy. This architecture integrates three cooperative functions: (i) a Cu‐rich amorphous outer layer (∼10 nm) that accelerates Na⁺ surface transport and suppresses interfacial side reactions; (ii) a subsurface Cu gradient that continuously relieves lattice strain and mitigates phase‐transition‐induced volume variation; and (iii) a gradual compositional transition that eliminates interfacial mismatch and preserves mechanical integrity during cycling. Crucially, because electrochemically active Mn is retained in the bulk, this gradient design avoids the intrinsic capacity loss associated with uniform doping while achieving long‐term stability. Benefiting from the above advantages, our surface Cu gradient‐doping strategy significantly improves the battery performance compared with previously reported gradient‐doping strategies [33]. As a result, Cu@MnPBA exhibits markedly enhanced Na⁺ transport kinetics (70.4 mAh g⁻¹ at 20 C vs. 47.9 mAh g⁻¹ for MnPBA) and suppressed Jahn‐Teller distortion, leading to reduced Mn dissolution and stabilized framework evolution. The cathode delivers 78% capacity retention after 500 cycles at 100 mA g⁻¹, and a practical hard‐carbon (HC)||Cu@MnPBA pouch cell retains 75% capacity after 3000 cycles. These results demonstrate that surface gradient doping provides a general, scalable design paradigm for simultaneously achieving high capacity, fast kinetics, and long‐term durability in low‐cost Prussian blue cathodes for SIBs.

2. Results and Discussion

A two‐step co‐precipitation strategy was employed to synthesize MnPBA with a Cu surface concentration gradient (Cu@MnPBA, Figure S1a, Supporting Information). Specifically, a Cu²⁺‐containing solution was gradually pumped into a Mn²⁺ solution at a rate of 0.25 mL min⁻¹, generating a mixed solution with a progressively increasing Cu²⁺/Mn²⁺ ratio. This mixture was subsequently and continuously injected into a Na₄Fe(CN)₆ solution at 1 mL min⁻¹. At the initial stage, the high Mn²⁺ concentration favored preferential precipitation of MnPBA, forming the particle core. As the Cu²⁺ concentration increased in the later stage, Cu was gradually incorporated into the outer region of the particles, leading to the formation of concentration‐gradient structures with a Mn‐rich core and a Cu‐rich surface. For comparison, pristine MnPBA was synthesized via a conventional one‐step co‐precipitation process. (Figure S1b, Supporting Information).

Rietveld refinement of the X‐ray diffraction (XRD) patterns confirms that the synthesized Cu@MnPBA crystallizes in a rhombohedral phase (space group R –3) with lattice parameters a = b = 6.581 Å, c = 18.86 Å, and α = β = 90°, γ = 120°, which are comparable to those of pristine MnPBA (Figures 2a and S2, Supporting Information). Detailed lattice parameters and refinement results are summarized in Tables S1, S2, and S3 (Supporting Information). To analyze the coordination characteristics in Cu@MnPBA, wavelet transform analysis was performed on the K‐space data [34, 35] (WT‐EXAFS) (Figure S3, Supporting Information). The Cu, Mn, and Fe elements exhibit typical Prussian blue‐type coordination features [36]. Both Cu@MnPBA and MnPBA exhibit a uniform particle size distribution in the range of 3∼5 µm (Figure S4, Supporting Information). Moreover, the characteristic cubic morphology is well preserved after introducing the Cu concentration gradient doping, indicating that the surface modification does not alter the overall particle morphology.

FIGURE 2.

a) Rietveld refinement of XRD pattern for Cu@MnPBA. b) Cu 2p and c) Mn 2p XPS spectra for Cu@MnPBA. d) TOF‐SIMS depth profiles and 3D views of the sputtered volume of Cu@MnPBA. e) STEM image of Cu@MnPBA and f) the corresponding Cu EELS spectra. g) HAADF‐TEM image of Cu@MnPBA. h) The lattices obtained from HAADF image and the corresponding FFT diffraction patterns.

Transmission electron microscopy (TEM) combined with energy‐dispersive X‐ray spectroscopy (EDS) mapping reveals that Fe and Mn are uniformly distributed throughout the Cu@MnPBA particles, whereas Cu is predominantly enriched at the particle surface, confirming the formation of a Cu‐coated MnPBA architecture (Figure S5, Supporting Information). Consistently, EDS line‐scan analysis shows that the Cu concentration gradually decreases from the surface toward the interior over a depth of approximately 200 nm, while Mn remains essentially constant (Figure S6, Supporting Information). X‐ray photoelectron spectroscopy (XPS) was employed to further probe the elemental composition and chemical states (Figure S7, Supporting Information). In the Cu 2p spectrum prior to Ar⁺ etching, peaks at 932.9 and 952.9 eV are assigned to Cu⁺ 2p_3/2 and 2p_1/2, while peaks at 935.6 and 955.7 eV correspond to Cu²⁺ 2p_3/2 and 2p_1/2, respectively. The presence of Cu⁺ is attributed to the use of excess chelating agents, such as sodium citrate, during synthesis. During crystallization, citrate anions adsorb on Cu²⁺ species, and the hydroxyl (–OH) and carboxyl (–COOH) groups donate electrons [37], partially reducing surface Cu²⁺ to Cu⁺. The Mn 2p spectrum exhibits characteristic peaks at 641.9 and 653.8 eV, corresponding to Mn²⁺ 2p_3/2 and 2p_1/2, while the Fe 2p peaks at 708.6 and 721.8 eV indicate the presence of Fe²⁺. To further verify the Cu concentration gradient, XPS depth profiling was performed on the powder sample (Figure 2b,c). The Cu⁺ signal disappears after the initial etching. As the etching time extends to 300 s, the Cu²⁺ signal intensity progressively decreases, whereas the Mn²⁺ intensity gradually increases, confirming the surface‐enriched Cu distribution. In addition, time‐of‐flight secondary ion mass spectrometry (TOF‐SIMS) depth profiling provides complementary evidence. As shown in Figure 2d, the intensity of Cu(CN)₃ ^– fragments continuously decreases with increasing sputtering time, further confirming the Cu concentration gradient localized at the particle surface. These results are consistent with those of scanning transmission electron microscopy (STEM) image and the corresponding electron energy loss spectroscopy (EELS) profiles (Figure 2e,f).

To further investigate the surface structure within the ∼200 nm Cu gradient‐doping region of Cu@MnPBA, high‐angle annular dark‐field scanning transmission electron microscopy (HAADF‐STEM) was conducted, as shown in Figure 2g. A surface amorphous layer with a thickness of ∼10 nm was clearly observed in Area 1 in Figure 2g,h, which is consistent with the high‐resolution TEM (HRTEM) results (Figure S8a,b, Supporting Information). The formation of this amorphous surface layer is expected to effectively suppress electrolyte erosion and facilitate Na⁺ transport at the interface. [38, 39]. With decreasing Cu concentration toward the interior, Cu@MnPBA gradually transitions to a well‐defined rhombohedral crystalline phase, accompanied by a reduction in the (210) lattice spacing from 0.391 to 0.383 nm (Figure 2g,h). In contrast, pristine MnPBA retains a fully crystalline surface structure (Figure S8c,d, Supporting Information) with the (210) lattice spacing of 0.385 nm.

To monitor the structural evolution of the Cu@MnPBA electrode during Na⁺ extraction and insertion, in situ X‐ray diffraction (XRD) was conducted on Na||Cu@MnPBA cells during the first charge‐discharge cycle at a current density of 10 mA g⁻¹ (Figure 3a). Initially, Cu@MnPBA exhibits a characteristic rhombohedral structure. Upon Na⁺ extraction, the intensity of the rhombohedral diffraction peaks gradually decreases, accompanied by the emergence of tetragonal‐phase reflections. At the fully charged state (4 V), the rhombohedral peaks nearly vanish, indicating a complete transformation to the tetragonal phase. During the subsequent discharge, the Cu@MnPBA electrode reversibly converts back to the rhombohedral structure, demonstrating good structural reversibility. The full phase evolution during the charge‐discharge process is summarized in Figure 3b. For comparison, pristine MnPBA undergoes a similar reversible phase transition (Figure S9a, Supporting Information) [10, 29]. However, analysis of the in situ XRD data using FullProf refinement reveals that the overall volume change of Cu@MnPBA is approximately 10% smaller than that of MnPBA (Figure S9b, Supporting Information). It is worth noting that for Cu@MnPBA (a = b = 10.045 Å, c = 10.058 Å), the elongation of the c‐axis caused by the Jahn–Teller effect is significantly reduced compared to that of MnPBA (a = b = 10.214 Å, c = 10.485 Å), indicating that Cu@MnPBA exhibits a weakened Jahn–Teller effect (Figure S9c,d, Supporting Information). The above phenomena indicate that the Cu gradient modification effectively mitigates lattice strain during Na⁺ cycling, thereby enhancing structural stability. In‐situ Raman spectroscopy (Figure S10, Supporting Information) shows that Cu@MnPBA undergoes fully reversible Fe²⁺/Fe³⁺─C≡N─Mn²⁺/Mn³⁺ transformations during charging and discharging, confirming its structural stability and long‐term cycling performance.

FIGURE 3.

a) In‐situ XRD patterns of Cu@MnPBA during the first charge/discharge process at 10 mA g⁻¹; the corresponding voltage‐time profile is shown as an inset (right). b) The schematic description of the phase transformation evolutions of Cu@MnPBA. c) Theoretical calculations of Mn–N bond length variations and bond‐length ratios under different Cu doping concentrations. d) Calculated DOS for different Cu doping concentrations. e) –ICOHP values of the Mn–N bond for different Cu doping concentrations. f) Modulated von Mises stress distributions in Cu@MnPBA and MnPBA from COMSOL simulation.

Density functional theory (DFT) calculations were conducted to elucidate the underlying stabilization mechanism induced by the Cu gradient‐doped layer. For computational simplicity, model systems with Cu doping levels of 0%, 25%, 50% and 75% in the fully desodiated state were constructed (Figure S11, Supporting Information). The influence of Cu content on the local structure of MnN₆ octahedra in MnPBA was systematically examined [40, 41, 42]. In particular, the Mn–N bond lengths along the three principal axes (denoted as L₁, L₂, and L₃) were analyzed. As shown in Figure 3c, pristine MnPBA exhibits a strong Jahn–Teller distortion, with Mn–N bond lengths of L₁ = L₂ = 1.99 Å, and a significantly elongated axial bond L₃ = 2.23 Å. This distortion originates from the anisotropic orbital occupation of Mn³⁺ in an octahedral crystal field [15, 28, 43]. With increasing Cu doping, the Jahn–Teller distortion is progressively suppressed. In the fully desodiated state, the axial Mn–N bond length (L₃) decreases, and the distortion ratios (L₃/L₁ and L₃/L₂) approach unity, indicating that the MnN₆ octahedra evolve toward a more regular geometry (Figure S12, Supporting Information). Electronic structure analysis further supports this stabilization. Density of states (DOS) calculations reveal reduced band gaps in Cu‐doped systems (Figure 3d), implying enhanced electronic conductivity, while Crystal Orbital Hamilton Population (COHP) and the corresponding integrated COHP (–ICOHP) analyses show strengthened Mn–N bonding upon Cu doping (Figure 3e) [44]. Consistently, simulations of Mn dissolution demonstrate that pristine MnPBA undergoes severe structural distortion after Mn removal, whereas Cu@MnPBA maintains an intact framework (Figure S13, Supporting Information).

To assess structural stability at larger length scales, finite element simulations were conducted [33, 45]. At 50% state of charge (SOC), Cu@MnPBA exhibits a more uniform Na⁺ concentration distribution than MnPBA, indicating faster and more homogeneous Na⁺ transport (Figure S14, Supporting Information). Analysis of the von Mises stress (σ_v) shows that the maximum stress in Cu@MnPBA is only 58 MPa, markedly lower than the 257 MPa observed in MnPBA (Figures 3f and S15, Supporting Information). Consequently, damage accumulation is significantly reduced, with a maximum damage value of 0.8% for Cu@MnPBA compared with 1.65% for MnPBA (Figure S15). These results collectively demonstrate that gradient Cu doping effectively enhances both structural and mechanical robustness during electrochemical cycling.

To elucidate the enhanced Na⁺ transport induced by the surface Cu concentration gradient, cyclic voltammetry (CV) measurements were first performed at a scan rate of 0.1 mV s⁻¹. As shown in Figure 4a, both Cu@MnPBA and MnPBA exhibit well‐defined oxidation and reduction peaks [10, 46]. Notably, Cu@MnPBA delivers higher peak currents than MnPBA, indicating improved electronic conductivity and faster electrochemical reaction kinetics. The reaction kinetics were further analyzed using CV at varying scan rates (Figure S16, Supporting Information). As illustrated in Figure 4b, the peak current shows a linear dependence on the square root of the scan rate. According to the Randles‐Sevcik equation, the fitted slopes for both oxidation and reduction processes are larger for Cu@MnPBA, confirming more rapid Na⁺ insertion/extraction kinetics. This enhancement is supported by the narrowed band gap shown in DOS calculations (Figure 3d), which promotes electron transport. To directly quantify Na⁺ diffusion, galvanostatic intermittent titration technique (GITT) measurements were carried out between 2.0 and 4.0 V at 10 mA g⁻¹ (Figure S17, Supporting Information). As shown in Figure 4c, the Na⁺ diffusion coefficients (D _Na+) of Cu@MnPBA (10⁻¹⁰∼10⁻¹¹ cm² s⁻¹) are an order of magnitude higher than those of MnPBA (10⁻¹¹∼10⁻¹² cm² s⁻¹), further demonstrating the accelerated ion transport enabled by the Cu gradient design.

FIGURE 4.

a) CV profiles at a scan rate of 0.1 mV s⁻¹. b) Linear fitting Randles‐Sevcik curves by plotting the maximal current as a function of the square root of the sweep rate. c) D _Na+ values obtained by GITT profiles at 10 mA g⁻¹. DRT results of d) MnPBA and e) Cu@MnPBA during the first charge/discharge process at 10 mA g⁻¹.

Distribution of relaxation times (DRT) analysis was employed to probe the evolution of Na⁺ transport and interfacial kinetics during electrochemical cycling [47, 48]. In situ galvanostatic electrochemical impedance spectroscopy (GEIS) was conducted during the first charge‐discharge cycle (Figure S18, Supporting Information), and the corresponding impedance spectra were converted into DRT profiles (Figure S19, Supporting Information). Distinct relaxation times (τ) correspond to different electrochemical processes. Owing to the n‐type semiconducting nature of Prussian blue, lower impedance is observed during charging than discharging, resulting in asymmetric impedance behavior (Figure 4d,e) [49]. Based on relaxation times, the impedance response can be divided into four distinct regions: bulk Na⁺ diffusion resistance (R _diffusion, τ = 10∼10² s), charge‐transfer resistance (R _ct, τ = 0.5∼10 s), cathode–electrolyte interphase resistance (R _CEI, τ = 10⁻³∼0.5 s), and ohmic resistance (R ₀, τ = 10⁻⁵–10⁻³ s) [50]. No significant differences are observed in R ₀ or R _CEI between Cu@MnPBA and MnPBA. In contrast, pronounced differences emerge in R _diffusion, which is strongly influenced by the state of charge (SOC) [51]. R _diffusion increases near 0% and 100% SOCs due to limited Na⁺ vacancies or ions, while intermediate SOCs enable faster diffusion (Figure S20, Supporting Information). Notably, Cu@MnPBA exhibits significantly lower maximum R _diffusion values at extreme SOCs (480.3 Ω at ∼0% SOC and 536.0 Ω at ∼100% SOC) than MnPBA (731.7 Ω and 663.0 Ω, respectively), indicating accelerated bulk Na⁺ transport. A similar SOC dependence is observed for R _ct, where Cu@MnPBA shows substantially reduced R _ct at both low and high SOC (322.0 Ω and 486.6 Ω, respectively) compared with MnPBA (642.8 Ω and 525.8 Ω), and remains lower even at intermediate SOCs. These results collectively demonstrate that the surface Cu gradient doping significantly enhances both bulk Na⁺ diffusion and interfacial charge‐transfer kinetics.

To evaluate the electrochemical performance of Cu@MnPBA, a series of samples with surface Cu concentration gradient doping was synthesized using different Cu/Mn feed ratios (0.05, 0.2, and 0.35). X‐ray absorption near‐edge structure (XANES) analysis indicates that the Cu²⁺/Cu⁺ redox couple is electrochemically active during cycling (Figure S21, Supporting Information). However, its capacity contribution is substantially smaller than that of the Mn²⁺/Mn³⁺ redox couple, resulting in a gradual decrease in specific capacity with increasing Cu/Mn feed ratios. In contrast, higher Cu substitution strengthens metal–N coordination, thereby enhancing the cycling stability of MnPBA. Among the investigated compositions, Cu@MnPBA with a Cu/Mn feed ratio of 0.2 achieves an optimal balance between specific capacity and long‐term cycling stability (Figure S22, Supporting Information). Unless otherwise specified, Cu@MnPBA hereafter refers to this optimized sample.

Figure 5a presents the galvanostatic charge‐discharge curves measured at 0.1 C (1 C 100 mA g⁻¹). Cu@MnPBA delivers an initial discharge capacity of 143.9 mAh g⁻¹ with an initial Coulombic efficiency (ICE) of 94.8%. For comparison, pristine MnPBA exhibits an initial discharge capacity of 146.6 mAh g⁻¹ and an ICE of 96.5%. The poor redox reversibility of the Cu element causes a slight decrease in the ICE, which is acceptable. Additionally, the slightly lower capacity of Cu@MnPBA indicates that the surface Cu gradient doping only marginally sacrifices specific capacity. Importantly, Cu@MnPBA exhibits a longer cycling life, with a capacity retention rate of 78% after 500 cycles at a current density of 1C, significantly higher than the 20% retention rate observed for MnPBA (Figures 5b and S23, Supporting Information). The differential capacity (dQ/dV) profiles collected over extended cycling further reveal the structural stability of the cathodes. Cu@MnPBA shows a much slower increase in polarization, with significantly smaller reductions in peak intensity and integral area compared with MnPBA (Figures 5c and S24, Supporting Information). Consistently, the discharge median voltage decay of Cu@MnPBA is markedly suppressed, showing only a 1.4% decay after 500 cycles, compared with a 6.5% decay for MnPBA (Figure 5d).

FIGURE 5.

a) Galvanostatic charge/discharge curves at 0.1C (10 mA g⁻¹). b) Long‐term cycling performance and the corresponding Coulombic efficiency of Cu@MnPBA and MnPBA electrodes during 500 cycles. c) dQ/dV curves of Cu@MnPBA at 1C (100 mA g⁻¹). d) Median voltage during long cycling at 1C (100 mA g⁻¹). e) Rate capabilities of Cu@MnPBA and MnPBA. f) Galvanostatic discharge curves of Cu@MnPBA at different specific currents. g) Cycling performance of Cu@MnPBA||HC pouch cell at 1C (100 mA g⁻¹). h) Comparison of the long‐term cycling performance with previously reported PBA||HC full cells.

Rate performance was evaluated over a wide range of current densities, as shown in Figure 5e,f. Cu@MnPBA exhibits superior reversible capacities of 145.6, 134.4, 129.5, 123.8, 119.0, 111.6, 102.2, 96.3, 84.2, and 70.5 mAh g⁻¹ at current densities from 0.1 C to 20 C, respectively. In contrast, MnPBA exhibits capacities of 146.6, 132.2, 125.3, 114.4, 105.4, 93.7, 80.8, 72.7, 59.2, 47.9 mAh g⁻¹ at the same current densities. These results demonstrate excellent rate capability and highlight the clear advantage of Cu@MnPBA for high‐rate SIB applications compared with MnPBA (Figure S25, Supporting Information). In addition, using the same Cu/Mn ratio as that of the optimized Cu@MnPBA, a uniformly Cu‐doped MnPBA (MnCuPBA) sample was synthesized via a one‐step co‐precipitation method identical to that used for MnPBA. The SEM results (Figure S26, Supporting Information) show that its morphology is similar to that of Cu@MnPBA and MnPBA. The EDS mapping results from TEM further reveal that Cu is uniformly distributed throughout the particles (Figure S27, Supporting Information). XRD analysis (Figure S28, Supporting Information) confirms that MnCuPBA crystallizes in a cubic phase, which is characterized by high symmetry, exhibits excellent structural stability during the intercalation and deintercalation of Na⁺. However, its Na⁺ content is lower (Table S1, Supporting Information), resulting in a lower theoretical specific capacity compared to the rhombohedral phase [52]. XPS analysis (Figure S29, Supporting Information) of the MnCuPBA powder shows that Mn, Fe, and Cu were all in the +2 state, consistent with the cubic phase reported previously [16]. Electrochemical measurements show that MnCuPBA delivers an initial discharge capacity of 120 mAh g⁻¹ at 0.1 C, with a Coulombic efficiency of 89%. Upon prolonged cycling at 1 C, the capacity retention reaches 68% after 500cycles (Figure S30, Supporting Information). Although the median‐voltage decay rate of the MnCuPBA sample is slower than that of the MnPBA sample, it is still higher than that of the Cu@MnPBA sample (Figure S31, Supporting Information). These results indicate that while uniform Cu doping effectively enhances cycling stability, this improvement is accompanied by a marked reduction in specific capacity, underscoring the advantage of the surface Cu gradient‐doping strategy adopted in Cu@MnPBA.

To evaluate the practical applicability of Cu@MnPBA, the material was assembled into full pouch cells, using hard carbon (HC) as the anode and Cu@MnPBA as the cathode. The pouch cells were configured with a cathode area of 4.3 × 5.6 cm², an anode area of 4.5 × 5.8 cm², an N/P ratio of 1.2, and a designed capacity of approximately 400 mAh. After an initial pre‐cycling activation process, long‐term cycling tests were conducted within a voltage window of 2.0∼3.6 V at a current density of 1 C (100 mA g⁻¹). Figure S32 (Supporting Information) shows the representative charge‐discharge profiles of the pouch cells. Notably, the cells maintain an average Coulombic efficiency of 99.9% and a capacity retention of 75% after 3000 cycles (Figure 5g), with the corresponding detailed parameters of the pouch cell provided in Table S4 (Supporting Information). These results demonstrate exceptional cycling stability and highlight the superior durability of Cu@MnPBA‐based pouch cells compared with previously reported Prussian blue cathode systems (Figure 5h and Table S5, Supporting Information) [18, 37, 49, 53, 54, 55, 56, 57, 58, 59, 60].

We further investigated the evolution of the particle surface chemistry and bulk structural integrity of Cu@MnPBA and MnPBA after prolonged cycling. TOF‐SIMS, including 3D surface reconstruction, 2D overlay mapping, and depth profiling, was performed on both electrodes after 300 cycles to elucidate the effect of the Cu gradient doping layer on the formation of the cathode‐electrolyte interphase (CEI) (Figures 6a and S33, S34, Supporting Information). The detected anionic fragments, including C₂HO⁻, CHO₂ ⁻, and NaF₂ ⁻, originate from the decomposition of organic solvents and NaPF₆ at the cathode surface. In addition, dissolution of transition‐metal ions during cycling, followed by reactions with acidic species (e.g., HF) in the electrolyte, leads to the formation of metal fluorides (MnF₃ ⁻, FeF₃ ⁻) at the interface [39]. Notably, the intensities of these species are substantially lower for Cu@MnPBA compared to MnPBA, indicating suppressed parasitic interfacial reactions. This behavior can be attributed to the mitigation of the Jahn−Teller distortion in Cu@MnPBA, which reduces transition‐metal dissolution. Moreover, the amorphous Cu‐rich surface layer effectively inhibits side reactions with HF, resulting in a more stable and robust CEI.

FIGURE 6.

a) TOF‐SIMS analysis of cathodes after 300 cycles: the 3D views and corresponding 2D mappings of the NaF⁻ and MnF₃ ⁻ intensity for MnPBA and Cu@MnPBA electrodes. b) Depth profiling XPS analysis of F elements ingredient proportion, etching time ranging from 0 to 180 s. c) Cryo‐HRTEM image of CEI on the surface of MnPBA and Cu@MnPBA after 300 cycles. d) The Raman spectra of Cu@PBA and MnPBA cathodes for d) pristine and e) after 300 cycles. The (CN)₂ evolution spectra obtained from DEMS for f) MnPBA and g) Cu@MnPBA.

The CEI composition was further analyzed by XPS (Figures 6b and S35, Supporting Information). Normalized F 1s spectra reveal that the inorganic NaF component accounts for ∼55% of the CEI in Cu@MnPBA at 0 s etching time, significantly higher than the ∼30% observed for MnPBA. A NaF‐rich inorganic CEI is known to provide superior interfacial stability [61] compared with organic C–F containing species [62]. With increasing etching depth, although the NaF fraction in MnPBA gradually increases, it remains consistently lower than that in Cu@MnPBA. The enhanced structural robustness of Cu@MnPBA suppresses Mndissolution during cycling, thereby reducing the catalytic decomposition of organic electrolytes at the interface [63]. Cryogenic HRTEM (cryo‐HRTEM) provides direct evidence supporting these observations. As shown in Figures 6c and S36 (Supporting Information), Cu@MnPBA exhibits a markedly thinner and more uniform CEI layer than MnPBA after 300 cycles, underscoring its superior interfacial stability during extended cycling. Consistent with finite‐element simulation results, the MnPBA particles display extensive microcracks that propagate through the entire crystal, whereas Cu@MnPBA shows significantly fewer intragranular cracks, with the overall particle morphology remaining largely intact (Figure S37, Supporting Information).

Raman spectroscopy was employed to probe structural degradation by monitoring the –C≡N– stretching vibrations. For pristine electrodes, both Cu@MnPBA and MnPBA exhibit characteristic peaks at 2066 and 2110 cm⁻¹ (Figure 6d). However, the peak intensities of Cu@MnPBA remain substantially higher than those of MnPBA after 300 cycles, indicating superior framework integrity and higher structural reversibility [64] (Figure 6e). Furthermore, gas evolution during cycling was monitored in situ using differential electrochemical mass spectrometry (DEMS). As shown in Figure 6f,g, both electrodes exhibit a signal at m/z = 52, which reaches its maximum intensity upon charging to 4 V and corresponds to the oxidation of cyanide ligands in the PBA framework [13] to dimeric cyanogen (CN)₂ [13]. Importantly, Cu@MnPBA generates a lower amount of (CN)₂ compared with MnPBA, reflecting its enhanced structural stability and more resilient surface chemistry.

3. Conclusion

In summary, we establish a surface Cu gradient doping design principle to overcome the intrinsic trade‐off between specific capacity and cycling stability in MnPBAs. By confining Cu doping to a ∼200 nm surface region while preserving an Mn‐rich bulk, this strategy spatially decouples structural stabilization from electrochemical activity. The resulting Cu@MnPBA integrates a Cu‐rich amorphous outer layer (∼10 nm) that enhances interfacial Na⁺ transport and suppresses parasitic reactions, together with a subsurface Cu concentration gradient that alleviates Jahn‐Teller‐induced lattice strain and mitigates phase‐transition‐driven volume variation. Electrochemically, Cu@MnPBA delivers a high initial discharge capacity of 143.9 mAh g⁻¹ with an initial Coulombic efficiency of 94.8%, while maintaining superior rate capability (70.4 mAh g⁻¹ at 20 C vs. 47.9 mAh g⁻¹ for MnPBA). The cathode exhibits markedly improved durability, retaining 78% capacity after 500 cycles at 1 C, accompanied by a substantially reduced median discharge voltage decay (1.4% vs. 6.5% for MnPBA). Kinetic analyses further reveal accelerated Na⁺ transport, with diffusion coefficients on the order of 10⁻¹⁰∼10⁻¹¹ cm² s⁻¹, nearly an order of magnitude higher than those of MnPBA. Overall, this work identifies surface gradient doping as a scalable and general electrochemical design paradigm for PBA cathodes, enabling the simultaneous optimization of capacity, kinetics, and long‐term stability for SIBs.

Author Contributions

Hong Kang: conceptualization, writing – original draft. Lifeng Wang: methodology, software. Ling Li: data curation. Yixuan Wang: investigation. Yuhang Lou: validation. Junpeng Sun: formal analysis. Jialong Shen: data curation. Nan Hu: formal analysis. Xianhong Rui: visualization. Xiaojun Wu: software. Hai Yang: visualization, writing – original draft. Hanyu Huo: writing – review and editing, resources. Yan Yu: writing – review and editing, funding acquisition, project administration.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: anie72767‐sup‐0001‐SuppMat.docx

Data Availability Statement

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