Heterostructures for advanced rechargeable batteries

BASIC PROPERTIES OF HETEROSTRUCTURES

Fundamental natures of heterostructures

Heterostructure materials integrate high-conductivity and high-stability components at the atomic scale, endowing the interfaces with multiple roles (such as electron conduction, ion transport, and stress buffering). Heterointerfaces often exhibit complex interfacial coupling effects due to inherent differences in Fermi levels, band structures, and lattice constants, with the most notable feature being the spontaneous formation of electric fields at the interface. Figure 1A illustrates the formation principle of a typical p-n junction. The p-type side contains more holes, and the n-type side contains more electrons. Electrons and holes migrate at the interface until the Fermi levels reach equilibrium. Figure 1B further shows the formation process of a heterointerface between two materials with different work functions. Upon contact, electrons spontaneously flow from the side with the lower work function to the side with the higher work function, leading to band bending and simultaneous generation of an electric field. This field not only provides the driving force and highly efficient charge-transfer pathways for ion/electron transport, but also endows the interface with rich cooperative catalytic active sites. The introduction of heterostructures into rechargeable batteries is critical for breaking through the kinetic limitations and enhancing structural stability of electrode materials.

Figure 1. (A) Schematic illustration of p-n junction formation. (A) is reprinted with permission from Ref.^[15], Copyright © 2022 Wiley; (B) Schematic illustration of the mechanism of heterointerface; (C) Number of articles dedicated to heterostructures in batteries. Database: Web of Science (https://webofscience.clarivate.cn).

Constructing an appropriate heterointerface is beneficial for accelerating charge transfer and buffering the volume expansion of active materials. In addition, optimizing the pathways and energy barriers of multiphase reactions (e.g., oxygen redox and sulfur conversion) enhances the rate performance, cycling stability, and energy efficiency of batteries^[14]. The constituent materials that form a heterostructure often differ in their physical and chemical properties, such as work functions, lattice parameters, and ion-diffusion barriers. The judicious matching of these distinct parameters enables precise control over charge transfer and structural stability of the electrode material. In recent years, research on heterostructures for batteries has flourished [Figure 1C].

Built-in electric fields effect

Built-in electric fields originate from interfacial charge redistribution driven by differences in work function. When the direction of the built-in electric fields aligns with that of ion diffusion, the ion diffusion barrier can be significantly reduced. Meanwhile, a space charge layer forms at the interface, which alters the local ion concentration distribution and enhances the ion diffusion coefficient^[6-8]. The ion concentration in the space-charge layer follows the Poisson-Boltzmann equation. The ratio of interfacial conductivity to bulk conductivity can reach 10²-10³. For example, in the Mo₂C/MoO₃ heterostructure, the ratio can be 240 at 300 K and remains above 50 at 230 K, demonstrating that the built-in electric fields significantly promote ion transport even at low temperatures^[16]. In the MXene/ZnTe heterostructure, the built-in electric fields raise the adsorption energy of K⁺ and reduce the migration barrier^[17]. In conversion-type materials and alloy anodes, the built-in electric fields accelerate electron migration from the current collector to the active materials and promote ion desolvation, thereby significantly reducing ohmic and electrochemical polarization^[18]. In addition, the space charge layer often modifies the local concentration and migration energy barrier of ions. This reduces the diffusion barrier of Li⁺ at the interface by an order of magnitude compared to the bulk phase^[19]. The effective range of the built-in electric fields is limited by the Debye length (λ_D). When the characteristic size (d) of the heterostructures approaches or is smaller than λ_D, band bending extends throughout the entire particle, leading to bulk polarization. Indeed, a properly designed d/λ_D ratio can optimize the built-in electric field’s effectiveness in promoting ion migration.

Lattice strain and defect concentration

When two materials with different lattice constants form an atomic-scale heterointerface, lattice distortion occurs to maintain atomic continuity. This generates lattice strain that directly alters bond lengths and bond angles, thereby influencing the band structure and d-band center of the material. These undercoordinated atomic configurations often act as highly active sites for electrochemical reactions, which exhibit a similar electrocatalytic effect for multi-electron reactions and effectively lower the reaction energy barrier. The Ni-MoO₂/NiMoO_4-x optimizes the electronic structure by enriching nickel defect sites and utilizing the strain effect induced by the defects. This significantly reduces the adsorption energy barrier of hydrogen/oxygen intermediates in metal-air batteries^[20]. The interfacial lattice mismatch δ is approximately 2.1%, corresponding to a strain energy of about 0.34 J m^-2. This strain energy can absorb approximately 16% of the volume-expansion work, thereby improving capacity retention from 42% for pristine MoO₃ to 89% after 300 cycles. The linear correlation between strain and capacity (R² = 0.93) indicates that strain exerts a dominant structure-buffering effect. The heterointerface also serves as a thermodynamically preferred site for defect formation, which realizes modulation of the electrochemical performance^[21]. Defect-rich regions exhibit stronger chemical adsorption for LiPSs and effectively suppress their shuttle effect. Consequently, heterostructures not only provide electron/ion transport channels but also optimize the kinetic pathways of electrode reactions through defect engineering.

Interface buffering effect

To accommodate the volume changes during the charging/discharging of high-capacity materials, an interfacial buffering mechanism has been established. The combination of the “hard” components (such as carbon frameworks and inert oxides) and the “active” components in the heterostructures forms a 3D confinement effect and serves as an elastic buffer layer, which can prevent the agglomeration and pulverization of active materials and extend the cycle life^[22]. When the active materials undergo volume expansion, the interface absorbs part of the strain energy through lattice slip or interface reconfiguration and naturally inhibits crack initiation and propagation^[23]. Furthermore, the rigid components serve as the structural framework to maintain the mechanical integrity of the electrode as a whole. In summary, the interfacial buffering effect is crucial for cycling stability. Experimental results have proven that the capacity retention of heterostructure electrodes is substantially higher than that of single active materials.

Regulation of reaction kinetics

The built-in electric fields significantly reduce the activation energy of charge transfer and promote the directional transport of electrons from the low-conductivity active materials to high-conductivity components. For instance, the band gap of the composite electrode is significantly reduced compared to that of the single electrode in ZnCo₂O₄/SnO₂^[32], which greatly improves electron transport efficiency. Furthermore, the electric fields accelerate electron transport and increase ionic migration channels in MoB/Si₃N₄^[33]. For MXene-based heterostructures, Ti₃C₂T_x has a high electronic conductivity similar to metals (up to ~ 8,000 S cm^-1), and it can be combined with active materials such as transition metal sulfides or metal-organic frameworks to construct a continuous 3D conductive network^[34], which effectively overcomes the insufficient conductivity.

Interfacial built-in fields

The heterointerface, which serves as an ion transport channel, converts conventional solid-phase diffusion into rapid interface-dominated migration. This is especially significant for potassium-ion batteries (PIBs) due to the large ionic radius of K⁺ (1.38 Å)^[35]. The heterointerface in NiTe₂/MoS₂ was also reported to reduce the diffusion energy barrier for K^+[36]. Moreover, Zhang et al. reported that the enhanced built-in electric fields at the heterointerface not only promote electron transmission but also accelerate the directional migration of Li^+[37], which demonstrated the lowest Li⁺ diffusion impedance during cycling.

Reaction pathway optimization

The effectiveness of heterostructures in electrochemical reaction kinetics stems primarily from the regulation of reaction pathways. Heterointerfaces can induce uniform nucleation and the growth of discharge products for electrodes in conversion reactions, which prevents the formation of large insulating crystal domains and improves the reaction reversibility. In LSBs, the MoO₂ and N-deficient carbon nitride (MoO₂/DCN) catalyst can adsorb LiPSs with an adsorption energy of −4.85 eV, which is much higher than that of a single DCN (1.15 eV), thus effectively suppressing the shuttle effect^[38]. In-situ ultraviolet visible (UV-vis) spectroscopy tests indicate that the heterostructures generate more S₃^- free radicals during discharge, with faster conversion rates of Li₂S₈ and Li₂S₆, and more thorough solid-liquid conversion, which can consequently improve sulfur utilization and cycling stability^[39].

Structural buffering effect

Numerous conversion-reaction-based electrode materials undergo volume changes even exceeding 300%, and the mechanical stress will lead to the pulverization and failure of the active materials^[40]. Through integrating two or more materials with distinct mechanical properties and crystal structures at the nanoscale to form heterostructures, stress-dissipation and buffering zones are created at the interfaces to effectively mitigate volume strain. For example, in a CoS₂/Co heterostructure, the ordered checkerboard-like architecture provides sufficient space to accommodate volume changes, delivering a capacity retention of 900 mAh g^-1 with high S loading of 7.13 mg cm^-2 and a low E/S ratio of 4.5 mL g^-1[41]. In the Si₃N₄/MoB@GO sandwich heterostructure, the introduction of the graphene intermediate layer not only improves conductivity, but also buffers volume changes and provides excellent structural stability^[42]. Moreover, chemical bonding and electrostatic forces at the heterointerface can further enhance the stability of the structure and inhibit interface degradation during long-term cycling.

Synergistic interfacial effects

The built-in electric fields in heterostructures can not only accelerate the interfacial migration of electrons and ions but also promote synergistic interfacial effects. Particularly, 2D/2D vertical stacking or 3D conductive framework-loaded heterostructures are beneficial for preventing the agglomeration and rearrangement of nanomaterials, which can generate a stronger synergistic effect through interfacial coupling^[43]. Firstly, the high conductivity components are combined with the high capacity components, and it helps to achieve a balance between rate and capacity. Secondly, interface defects and dangling bonds provide additional active sites, which can enhance the pseudocapacitive contribution^[44,45]. Thirdly, the combined action of multiple components promotes multi-electron reactions and contributes to high capacity. Finally, the design of heterostructures can improve battery safety and reduce potential hazards.

Metal-ion batteries

Metal-ion batteries operate on the “rocking-chair” principle, wherein metal ions reversibly intercalate and deintercalate between electrodes during charging/discharging, thereby converting electrical and chemical energy via redox reactions^[46]. Solid-state diffusion in the electrodes and structural stability constitute the rate-limiting steps that dictate the energy output and cycle life of the battery^[47]. Li-rich layered oxides (LLOs) are widely employed as cathodes for LIBs owing to their high specific capacity. However, they suffer from severe structural degradation caused by the out-of-plane migration of transition metals (TMs) from the TM layers to the alkali metal layers, and this process is further exacerbated by the formation of vacancy clusters and O-O dimers upon deep delithiation. To address this, Wang et al. designed an O2/O3 biphasic heterostructure consisting of a non-Li-rich O2 phase and a Li-rich O3 phase^[48]. The heterointerface features a face-shared configuration and provides a strong electrostatic repulsion that can suppress the out-of-plane migration of TM ions from the Li-rich O3 domains. Simultaneously, the absence of Li vacancies in the TM layers of the non-Li-rich O2 phase further reduces the source of vacancies. This synergistic effect minimizes the formation of vacancy clusters and prevents the generation of detrimental O-O dimers, thereby stabilizing the anionic redox chemistry and preserving the integrity of lattice oxygen. Furthermore, LiZr₂(PO₄)₃ (LZP) was reported as a coating for LLOs to form an LLO/LZP heterostructure^[49] [Figure 3A], which modulates the antibonding and O 2p non-bonding bands in LLOs [Figure 3B]. This modulation enhances Li⁺ migration and electron transfer while suppressing irreversible phase transitions and side reactions. Owing to the enhanced lattice oxygen stability provided by the LZP layer, the generation of O₂ and CO₂ is reduced, and voltage decay is also alleviated.

Figure 3. (A) High-resolution transmission electron microscope (HRTEM) image of LLO/LZP with FFT patterns; (B) Schematic illustration of energy band modulation through surface chemistry modification of LLO/LZP. (A and B) are reprinted with permission from Ref.^[49], Copyright © 2025 Elsevier; (C) HRTEM image of the Si/MoSe₂@C heterostructure; (D) Optimized atomic structure in side view with overlaid Li⁺ diffusion trajectories; (E) K_eff of Si and Si/MoSe₂@C-based matrix as a function of porosity (ε). (C-E) are reprinted from Ref.^[50], under CC BY 4.0 license; (F) HRTEM image with insets showing magnified views of Zig-Zag and P2; (G) Schematic illustration of the P2/Zig-Zag heterostructure with atomic interlocking; (H) Lattice parameter evolution derived from XRD of the P2/Zig-Zag heterostructure. (F-H) are reprinted with permission from Ref.^[25], Copyright © 2025 Wiley; (I) Schematic illustration of the benefits of the layered-to-rocksalt atomic reconfiguration to the NFM/RS; (J) NFM and NFM/RS cathodes changed in lattice parameters a, c and V; (K) The thickness of CEI for NFM/RS after 200 cycles at 1 C. (I-K) are reprinted with permission from Ref.^[51], Copyright © 2025 Elsevier. UHB: Upper hubbard band; LHB: lower hubbard band; TM-O: transition metal-oxygen; RS: rocksalt; NFM: NaNi_1/3Fe_1/3Mn_1/3O₂; CEI: cathode electrolyte interface; NFM/RS: NaNi_1/3Fe_1/3Mn_1/3O₂; LLO: Li-rich layered oxide; LZP: LiZr₂(PO₄)₃; XRD: X-ray diffraction.

Silicon anodes possess a high theoretical capacity of 4,200 mAh g^-1 but suffer from a volume change of ~ 300% during cycling^[52]. Zhu et al. developed a highly stable Si/MoSe₂@C heterostructure that reconciles the high capacity with superior stability^[50] [Figure 3C]. Lattice-matched MoSe₂ bridges the porous silicon and carbon coating through covalent bonds to form stable Si-Se-Mo linkages that optimize Li⁺ transport pathways and stabilize the structure. In this heterostructure, MoSe₂ serves as an electron-conductive and ion-permeable interlayer [Figure 3D] to reduce the maximum Li⁺ migration energy barrier from 0.81 eV (Si@C) to 0.61 eV (Si/MoSe₂@C). Moreover, thermal conductivity measurements demonstrate that the heterointerface in Si/MoSe₂@C reduces interfacial thermal resistance, which results in a 27% increase in thermal conductivity [Figure 3E].

SIBs are considered a promising alternative to LIBs due to their low cost and superior low-temperature performance^[46]. However, most electrode materials exhibit sluggish kinetics and poor cycling stability. Layered transition metal oxides are particularly attractive cathode materials for SIBs due to their high energy density and elemental abundance^[53]. Chen et al. designed a P2/Zig-Zag biphasic heterostructure cathode featuring an atomic interlocking architecture^[25] [Figure 3F]. This structure ensures continuous Na⁺ migration pathways, while the “interlocking effect” suppresses interlayer gliding [Figure 3G], and the P2 phase reduces the Na⁺ migration energy barrier. In-situ X-ray diffraction (XRD) results reveal that the volume expansion of the P2 phase within the biphasic heterostructure is effectively counterbalanced during charging, resulting in a smaller volume change as compared to the pure P2 phase [Figure 3H]. Owing to the low-resistance Na⁺ migration pathways enabled by the heterostructures, the Na⁺ diffusion coefficient increases to 10^-9 cm² s^-1 upon the P2/Zig-Zag to P2/P3 phase transition. For O3-type cathodes, the electrochemical performance is severely compromised when the pristine layered structure transforms into a cation-disordered rock-salt phase. To solve this issue, an O3-type NaNi_1/3Fe_1/3Mn_1/3O₂ (NFM/RS) layered-rock-salt heterostructure was developed based on an atomic reconstruction strategy^[51] [Figure 3I], which exhibited superior hydrophobicity and imparted moisture resistance in air while simultaneously achieving typical quasi-zero-strain behavior with an overall volume change of only 0.72% [Figure 3J]. As a result, the NFM/RS cathode showed a thin and uniform cathode electrolyte interface (CEI) and maintained enhanced rate performance [Figure 3K].

Among anode materials for SIBs, transition metal sulfides (TMSs) offer high energy density due to their multi-electron transfer characteristics. However, the aggregation and deactivation of active particles remain major obstacles^[57]. To address this, Zhang et al. designed an isotropic Co₉S₈/MoS₂ heterostructure to alleviate internal stress^[54]. The preferential crystallographic orientation provides rapid Na⁺ diffusion pathways, while the precisely synthesized isotropic heterostructures enable uniform Na⁺ deposition [Figure 4A]. In this isotropic structure, Na⁺ rapidly diffuses and distributes uniformly to the bottom, while the stress distribution remains dispersed, which avoids stress accumulation [Figure 4B]. Additionally, the abundant heterointerface effectively enhances electronic conductivity, synergistically improving reaction kinetics. In electrodes, vacancies not only increase reactive sites but also serve as bridges facilitating electron transfer between different semiconductors. The combination of heterostructures with vacancy engineering offers synergistic enhancements in both conductivity and kinetics. A ZnS-MoS₂ heterostructure with tunable S vacancy content can be prepared by temperature control to achieve superior sodium storage performance^[55]. During heat treatment, the strong electronegativity of Mo favors bonding with S and generates the ZnS-containing S vacancies. Notably, the optimal vacancy signal was observed at 800 °C [Figure 4C]. Density functional theory (DFT) elucidates that the charge redistribution at the ZnS-MoS₂ interface effectively closes the bandgap, which yields the lowest Na⁺ diffusion energy barrier of 0.15 eV [Figure 4D and E].

Figure 4. (A) Na⁺ concentration distributions and (B) Na⁺ concentration comparisons of the isotropic and anisotropic structure at the specific time in Co₉S₈/MoS₂ heterostructure. (A and B) are reprinted with permission from Ref.^[54], Copyright © 2025 Wiley; (C) Electron paramagnetic resonance (EPR) spectra of ZnS/MoS₂-600, ZnS/MoS₂-700, and ZnS/MoS₂-800; (D) Na⁺ migration paths and (E) diffusion energy barriers in ZnS/MoS₂. (C-E) are reprinted with permission from Ref.^[55], Copyright © 2025 American Chemical Society; (F) Schematic diagram of work functions of Co/FeSe and Co/Fe₃Se₄ in Co/Fe₄Se₅ heterostructure; (G) SKPM image of Co/Fe₄Se₅@C heterostructure and (H) the concerned surface potential plots. (F-H) are reprinted with permission from Ref.^[56], Copyright © 2024 Wiley. SKPM: Scanning kelvin probe force microscopy.

The potential of potassium (K⁺/K: -2.93 V) is very close to that of lithium, which implies that the theoretical operating voltage of PIBs can be comparable to LIBs (approximately 3.5-4.0 V) ^[58]. According to the energy density formula E = Q × V, the voltage advantage endows PIBs with the potential to surpass SIBs in terms of both power density and the upper limit of energy density^[59]. Current research on PIBs focuses on addressing the severe volume expansion and sluggish diffusion kinetics arising from the large ionic radius of K^+[60]. In anode materials, constructing a heterointerface has proven to be an effective strategy to tackle these two core issues. Li et al. designed an N/CoTe₂ featuring self-catalytic N-Co bonds and a low-tortuosity 3D tunnel architecture for efficient potassium storage^[61]. This heterostructure provides short-range and efficient electron/ion transport pathways. The in-situ formed N/Co@C self-catalytic centers significantly enhance the adsorption capability toward K_xTe_γ and accelerate their conversion kinetics.

The quantitative modulation of built-in electric fields within heterostructures represents a significant breakthrough in recent research. Song et al. achieved experimental quantification and tuning of the field intensity in FeSe/Fe₃Se₄^[57] [Figure 4F]. By combining scanning Kelvin probe microscopy [Figure 4G], Zeta potential measurements, and transient photocurrent density analysis, they directly quantified the increase in intensity from 17.24 mV to 62.84 mV upon uniform Co doping [Figure 4H]. The enhanced electric fields reduce the K⁺ diffusion barrier from 0.53 eV to 0.34 eV and enable the electrode to maintain a high reversible capacity of 145.8 mAh g^-1 at 10 A g^-1, with 95.1% capacity retention after 3000 cycles. Notably, this strategy demonstrates universality across heterostructures doped with other atoms (Cu/Ni), which provides a general approach for modulating interfacial electric field strength.

Lithium-sulfur batteries

LSBs are promising candidates for next-generation energy storage systems due to their high theoretical energy density (2,600 Wh kg^-1) as well as the resource abundance and environmental friendliness of sulfur^[62,63]. However, the practical application of LSBs is hindered by three factors: the insulating nature of sulfur and its discharge products (Li₂S₂/Li₂S)^[64], the shuttle effect and the sluggish reaction kinetics of LiPSs^[65]. Heterostructure materials typically exhibit a synergistic “adsorption-catalysis” effect^[66], which offers unique advantages in regulating LiPSs and accelerating redox kinetics. Although incorporating catalysts into the sulfur cathode can enhance the reaction kinetics, their adsorption-desorption kinetics toward LiPSs often remain unsatisfactory. To address this issue, Song et al. constructed a RuP₂-RuP heterostructure through a temperature-controlled strategy^[67] [Figure 5A], which exhibits excellent electronic conductivity, desirable adsorption capability for LiPSs, and high catalytic activity. DFT calculations reveal that the RuP₂-RuP exhibits the strongest adsorption affinity toward Li₂S₆, and the S-S bond length is elongated from 2.10-2.12 Å to 2.15 Å after adsorption, which facilitates the subsequent conversion of LiPSs [Figure 5B]. Moreover, the RuP₂-RuP heterostructure achieves a high Li₂S deposition capacity of 267 mAh g^-1, which further confirms its superior catalytic activity for the liquid-solid conversion. This work clearly demonstrates the critical role of interfacial coupling effects in heterostructures for simultaneously enhancing polysulfide anchoring and catalytic conversion. GeS₂/NiS₂ heterostructure can be designed as a catalytic sulfur host^[68] [Figure 5C]. Metallic NiS₂ forms an ohmic contact with semiconducting GeS₂, which can create a highly conductive heterointerface [Figure 5D]. During reactions, NiS₂ acts as an emitter, rapidly injecting a large number of electrons into LiPSs through the GeS₂ base, thereby achieving a maximum reaction current amplification factor (β_R) of 105. Consequently, it can enhance the electron transfer kinetics [Figure 5E] and lead to a swift conversion of S₈ into LiPSs and uniform deposition of Li₂S on the surface of the S@GeS₂/NiS₂@rGO electrode.

Figure 5. (A) HRTEM image of RuP₂/RuP heterostructure; (B) The simulated charge density difference isosurface of Li₂S₆ adsorbed on the surface of different catalysts. (A and B) are reprinted with permission from Ref.^[67], Copyright © 2024 Elsevier; (C) HRTEM image of GeS₂/NiS₂@rGO; (D) Electron-triode-like GeS₂/NiS₂ heterostructure achieves 105.87 times current amplification; (E) Schematic diagram of the lithium-sulfur pouch battery, and the continuous reaction of transferred electrons results in an uneven electron distribution. (C-E) are reprinted with permission from Ref.^[68], Copyright © 2025 Royal Society of Chemistry; (F) Scenario of d-p orbital hybridization and (G) current-time profile recorded under a constant potential of 2.05 V for charging in CoO/Mo₂C. (F and G) are reprinted from Ref.^[71], under CC BY 4.0 license; (H) Corresponding planar time-resolved Raman spectra of pure PP and WB/WC-modified separator. (H) is reprinted from Ref.^[74], under CC BY 4.0 license. rGO: Reduced graphene oxide; CNF: carbon nanofiber; HRTEM: high-resolutiontransmission electron microscope; LS: low spin; HS: high spin.

Current interpretations of the role of sulfur hosts predominantly focus on the relationship between electronic energy levels and catalytic properties, while the contribution of electron spin is often overlooked. In fact, the spin-state configuration of transition metals is a key factor influencing the electronic structure of catalysts, which determines their orbital occupancy, activity, and selectivity^[69]. Inspired by this, a hollow NiS₂/NiSe₂ heterostructure with a high-spin configuration was developed by Huang et al^[70]. At the NiS₂/NiSe₂ heterointerface, the octahedral crystal field splitting energy of Ni³⁺ is modified, thus triggering electron transitions from the t_2g to e_g orbitals and inducing a low-spin to high-spin transition. In the high-spin state, the increased e_g electron occupancy upshifts the d band center, which strengthens orbital hybridization with the S 3p states of LiPSs. This heterostructure achieves significantly improved electrical conductivity and enhanced adsorption toward Li₂S₆. The nickel serves as the active site to supply abundant electrons to LiPSs to facilitate their reduction, which enables rapid reaction kinetics. The role of electron spin states is similarly evident in CoO/Mo₂C@CNFs (carbon nanofibers)^[71] [Figure 5F]. In contrast to the high-spin state of nickel, the incorporation of molybdenum reduces the spin state of cobalt due to differences in crystal structure^[72], and the resulting low-spin cobalt effectively mitigates the LiPSs shuttle effect through strong chemisorption of Co-S bonds. As a result, the CoO/Mo₂C@CNFs achieves a higher Li₂S precipitation capacity (165 mAh g^-1) than that of CoO@CNFs (60 mAh g^-1) and Mo₂C@CNFs (78 mAh g^-1) [Figure 5G].

For LSBs, the separator not only physically isolates the cathode and anode but also plays a critical role in regulating the LiPSs and Li⁺ ion deposition behaviors^[73-75]. A common strategy is the functional modification of the separator with a coating layer that imparts adsorption and catalytic capabilities. Heterostructured coating designs are particularly advantageous as they offer a catalytically active heterointerface while ensuring efficient Li⁺ transport. Wu et al. designed a WB/WC heterostructure using an MBene material^[74], where B forms strong coordination bonds (B-S) with the S atoms in LiPSs, which increases the density of catalytically active centers. In-situ Raman spectroscopy provides intuitive evidence that the WB/WC coating effectively blocks the LiPSs shuttle and promotes their reduction to Li₂S [Figure 5H]. Similarly, Ma et al. reported a MIL-88A/CdS heterostructure coating featuring nanoscale catalytic sites anchored on a micron-scale substrate^[75] [Figure 6A], where the significant size disparity enhances the catalytic activity of CdS. MIL-88A/CdS shows a substantially lower diffusion energy barrier of Li₂S₆ (0.92 eV) than that on MIL-88A (1.80 eV) [Figure 6B], which indicates rapid migration to CdS catalytic sites. Time-of-flight secondary ion mass spectrometry (TOF-SIMS) reveals that LiPSs do not accumulate extensively on the surface but exhibit transient fluctuations followed by rapid stabilization [Figure 6C]. This observation reflects a strong interaction between the LiPSs and the heterostructure, which can significantly enhance sulfur utilization.

Figure 6. (A) Optimized configuration of Li₂S decomposition on MIL-88A/CdS; (B) Diffusion energy barrier of Li₂S₆ in MIL-88A/CdS; (C) TOF-SIMS 3D reconstructed images for LiS^- and LiS₂^- secondary ions of CdS and MIL-88A/CdS. (A-C) are reprinted from Ref.^[75], under CC BY 4.0 license; (D) Energy-band diagrams of MoS_2-x/MoO₂/CoP ternary heterostructure before and after contact formation of the built-in electric fields; (E) Mo 3d spectra of MoS_2-x/MoO₂/CoP before and after adsorbing Li₂S₆. (D and E) are reprinted with permission from Ref.^[78], Copyright © 2025 Elsevier; (F) TEM image of Bi₂Te₃/Sb₂Te₃. Figure F is reprinted with permission from Ref.^[79], Copyright © 2025 Wiley; (G) Energy diagrams of Bi₂Te₃/Sb₂Te₃; (H) Discharge/charge curves with current density 500 mA g^-1 of Ru/RuO₂ cathodes. (G and H) are reprinted with permission from Ref.^[80], Copyright © 2025 Wiley. TOF-SIMS: Time-of-flight secondary ion mass spectrometry; TEM: transmission electron microscope.

Beyond separator coating modifications, heterostructures employed as interlayers also effectively mitigate the shuttle effect and enhance the performance of LSBs. Xu et al. developed a ternary heterostructure of MoS_2-x/MoO₂/CoP with built-in electric fields and abundant S vacancies^[78]. Each of the three components contributes distinct functionalities: MoS_2-x and CoP exhibit strong adsorption and catalytic capabilities, respectively, while MoO₂ possesses an intermediate work function that serves as a bridge between the electronic pathways [Figure 6D]. Adsorption experiments reveal that after interaction with Li₂S₆, the Mo 3d peaks of MoS_2-ₓ/MoO₂/CoP shift toward lower binding energies, which highlights the strong interaction with Li₂S₆ [Figure 6E]. Moreover, MoS_2-x/MoO₂/CoP exhibits higher binding energies with LiPSs than other materials throughout six lithiation steps, implying outstanding adsorption performance. When employed as an interlayer in high-loading LSBs (7.13 mg cm^-2), the MoS_2-ₓ/MoO₂/CoP delivers a stable capacity of 6.6 mAh cm^-2 after 100 cycles.

Metal-air batteries

Metal-air batteries achieve energy conversion through the oxygen reduction reaction (ORR) at the cathode and the metal oxidation reaction at the anode, which has two fundamental advantages^[81]. Firstly, they theoretically deliver exceptionally high energy densities (e.g., Li-O₂ > 3,500 Wh kg^-1). Secondly, the O₂ is sourced directly from the environment without the need for harsh reaction conditions, which can fundamentally reduce material and construction costs. On the cathode, both the ORR and the OER suffer from intrinsically sluggish kinetics, which necessitates efficient catalysts to lower the overpotentials. One critical advance is the design of coherent heterostructures with minimal lattice mismatch, which enables strong interfacial interactions and enhanced structural stability. Feng et al. developed a 2D Bi₂Te₃/Sb₂Te₃ heterostructure with exposed (001) facets as a catalyst for Li-O₂ batteries^[79] [Figure 6F]. Owing to the low lattice mismatch (< 3%) between Bi₂Te₃ and Sb₂Te₃, the coherent interface ensures seamless atomic arrangement and minimizes stress accumulation. The built-in electric fields modulate the asymmetric charge distribution of Te atoms and enhance the adsorption of intermediate LiO₂, which facilitates charge transfer between the adsorbed species and the catalyst [Figure 6G]. Consequently, the Bi₂Te₃/Sb₂Te₃ cathode delivers remarkable cycling stability in both pure oxygen and ambient air.

Compared to coherent heterostructures, Mott-Schottky heterostructures offer useful platform for accelerating ORR/OER kinetics by leveraging the spontaneous electron transfer at metal-semiconductor interfaces. Sun et al. constructed a Ru/RuO₂ heterostructure by partial oxidation^[80]. The Ru/RuO₂ heterointerface modulates the d-band center to an optimal position, which balances the adsorption energies of oxygen-containing intermediates. The moderate binding strength for LiO₂ and Li₂O₂ enables efficient conversion along a two-electron pathway with minimal overpotential [Figure 6H]. Beyond metal oxide systems, multi-component Mott-Schottky heterostructures have also demonstrated remarkable catalytic capabilities. Xia et al. designed a self-supported NiCo₂O₄/MnO₂ heterostructure with vertically aligned nanosheet arrays on Ti paper^[82] [Figure 7A]. The Mott-Schottky heterointerface weakens the adsorption of the LiO₂ intermediate to an optimal level, which can promote the formation of chip-like Li₂O₂ intimately embedded within the nanosheet arrays. Besides, the unique morphology provides abundant contact sites for catalysts and achieves efficient decomposition during charging [Figure 7B]. These merits enable the NiCo₂O₄/MnO₂-based electrode to achieve 800 cycles and a remarkably low overpotential of 0.73 V that outperforms the NiCo₂O₄.

Figure 7. (A) HRTEM image and (B) schematic reaction mechanism of NiCo₂O₄/MnO₂. (A and B) are reprinted from Ref.^[82], under CC BY 4.0 license; (C) ORR free energy diagrams for FeN₄ and Fe₅/FeN₄ sites at 0 and 1.23 V vs. RHE. (C) is reprinted with permission from Ref.^[83], Copyright © 2025 Wiley; (D) HRTEM image of CeO₂/ZnCoS; (E) Charge/discharge polarization and corresponding power density curves of ZABs employing CeO₂/ZnCoS and Pt/C + RuO₂ as air cathodes. (D and E) are reprinted with permission from Ref.^[84], Copyright © 2024 Elsevier; (F) Energy density curve of Co/Co₂P at 20 mA cm^-2; (G) Reaction mechanism of Zn-Co/air hybrid batteries. (F and G) are reprinted with permission from Ref.^[85], Copyright © 2025 Wiley; (H) TEM image of p-NiSe₂/Ni/Ni₃N@NCNT@CC. (H) is reprinted with permission from Ref.^[86], Copyright © 2025 Wiley; (I) Gibbs free energy diagrams of ORR for Fe₂N@NC and Fe₂N/CrNx@NC at U = 0 V. (I) is reprinted with permission from Ref.^[87], Copyright © 2026 Wiley. PNCF: P/N-codoped carbon nanofiber; NC: N-doped carbon; HRTEM: high-resolution transmission electron microscope; TEM: transmission electron microscope; RHE: ORR: oxygen reduction reaction.

Zinc-air batteries (ZABs) possess high theoretical energy density and inherent safety. However, the reactions involve multi-step electron transfer processes and high overpotentials^[88]. An essential approach to achieve exceptional bifunctional activity is the precise engineering of atomic-scale heterostructures that combine dual-atom sites with adjacent nanoclusters. Lu et al. developed a porous carbon fiber membrane decorated with atomically dispersed CoN₄/FeN₄ dual sites and neighboring Co₂Fe₂/Fe₅ nanoclusters^[83]. The Fe₅ cluster significantly weakens the Fe-OH coupling at the FeN₄ site, which lowers the energy barrier for the rate-determining step of ORR from 0.59 eV to 0.46 eV, while the Co₂Fe₂ cluster enhances Co-OH interaction to promote OER kinetics [Figure 7C]. This synergistic effect endows the catalyst with outstanding ORR (0.87 V) and OER (1.58 V) performance. Zhang et al. constructed a Janus CeO₂/ZnCoS heterostructure^[84] [Figure 7D], in which the intimate heterointerface induces strong electronic coupling, thus generating abundant O and S vacancies that serve as additional active sites. The CeO₂/ZnCoS with S vacancies exhibits up-shifted d-band centers and optimized adsorption energies for intermediates, resulting in a reduced overpotential for the rate-determining step of ORR (0.32 V) as compared to single-phase ZnCoS (0.82 V). When integrated into ZABs, it delivers a peak power density of 168.7 mW cm^-2 [Figure 7E], and exhibits enhanced cycling stability over 865 cycles.

Another way to enhance air cathode performance is the design of self-reconstructed heterostructures. A Co/Co2P heterostructure can be embedded in P/N-codoped carbon nanofibers (Co/Co2P@PNCF) as a cathode for Zn-Co/air hybrid batteries ^[85]. During the first cycle, the Co/Co₂P undergoes deep self-reconstruction and generates high-valent Co species (Co³⁺/Co⁴⁺) that serve as highly active redox centers. The reversible transformation between Co³⁺/⁴⁺O_x(OH)_y and K_xCo²⁺/³⁺O_y during cycling enhances OER activity and provides additional charging-discharging plateaus, which promotes the Zn-Co/air hybrid batteries to deliver high power density and cycling stability [Figure 7F and G]. A key kinetic bottleneck in ORR that has been largely overlooked is the acceleration of the proton-coupled electron transfer. Zhang et al. developed an Fe₂N/CrNₓ heterostructure catalyst supported on N-doped carbon (Fe₂N/CrN_x@NC) for aluminum-air batteries (AABs)^[87]. In this heterostructure, Fe₂N serves as the primary ORR active sites, while CrN_x clusters enhance H₂O dissociation kinetics, which acts as a continuous proton supply to accelerate the protonation of intermediates^[89]. Meanwhile, the Fe₂N/CrN_x heterointerface induces strong d-d orbital hybridization, which upshifts the d-band center of Fe atoms, thereby optimizing the adsorption energy of ORR intermediates. The synergistic effect reduces the overpotential from 0.68 V (Fe₂N@NC) to 0.55 V (Fe₂N/CrN_x@NC). Using a plasma-assisted nitridation strategy, Xu et al. designed a Se vacancy-rich ternary heterostructure of NiSe₂/Ni@Ni₃N catalyst^[86] [Figure 7H]. The plasma treatment can introduce abundant Se vacancies and carbon defects. This unique ternary architecture combines the metallic conductivity of Ni, the catalytic activity of NiSe₂, and the structural stability of Ni₃N, while Se vacancies modulate the electronic structure of the active Ni sites^[87]. This enables the AABs to exhibit a small bifunctional potential gap (ΔE = 0.74 V) and deliver a peak power density of 106.8 mW cm^-2 with stable cycling over 1,000 cycles [Figure 7I].

Aqueous batteries

Aqueous batteries have attracted considerable attention owing to their intrinsic safety, environmental friendliness, and cost-effectiveness^[90,91]. However, their development has been persistently hindered by low energy density and insufficient stability. The narrow thermodynamic stability window of water molecules (~ 1.23 V) makes the hydrogen evolution reaction (HER) a critical bottleneck limiting Coulombic efficiency and cycle life. Concurrently, metal anodes are prone to corrosion and dendrite growth in an aqueous environment. In addition, the aqueous environment readily induces dissolution and structural collapse of active materials, and most cathode materials exhibit insufficient intrinsic electronic conductivity, resulting in poor rate capability. These interconnected issues collectively constitute the core obstacles hindering the practical application of aqueous batteries.

Heterostructure strategies offer a promising way to overcome these barriers. For instance, MnO₂ is widely utilized in aqueous zinc-ion batteries (ZIBs) due to its high theoretical capacity of 308 mAh g^-1 and a suitable voltage window (1.3-1.4 V vs. Zn²⁺/Zn)^[92]. Nevertheless, Mn³⁺ is prone to both disproportionation reactions (2Mn³⁺ → Mn²⁺ + Mn⁴⁺) and Jahn-Teller distortion, which exacerbate manganese dissolution and structural instability. Therefore, constructing heterostructures to enhance the electrochemical activity of Mn-based materials is an effective strategy for improving stability. Based on this, Zhao et al. proposed a dynamic transformation strategy that effectively reduces the formation of the irreversible byproduct ZnMn₂O₄, which suppresses Mn dissolution and enhances electronic conductivity^[93]. The prepared Bi_12.53Mn_0.47O_19.85/R-MnO₂ (BiO/MnO₂) heterostructure features BiO acting as a reservoir for Bi³⁺ [Figure 8A]. During reaction, BiO can directly supply Bi³⁺ to R-MnO₂, leading to the in-situ formation of Bi₂Mn₄O₁₀ (BMO) [Figure 8B], which facilitates a dynamic transformation from BiO/MnO₂ to BMO/MnO₂. During this process, the concentration of dissolved Bi³⁺ in the electrolyte tends to stabilize, and the dynamic transformation gradually reaches equilibrium, which creates a more abundant heterointerface, avoids spontaneous dissolution of Mn, and enhances reaction kinetics. Consequently, the BiO/MnO₂ based aqueous ZIB exhibits a high rate capability (204.5 mA h g^-1 at 4 A g^-1) and superior cycling stability (no capacity decay after 2,000 cycles at 2 A g^-1).

Figure 8. (A) HRTEM image of BiO/MnO₂ with SAED pattern; (B) Schematic of Zn||BiO/MnO₂ battery. (A and B) are reprinted with permission from Ref.^[93], Copyright © 2024 Royal Society of Chemistry; (C) HRTEM of TiS₂/TiO₂ heterostructure; (D) Energy barriers for the desolvation process of the [Zn(H₂O)₅]²⁺(CF₃SO₃)^- solvated structure on different crystal structures. (C and D) are reprinted with permission from Ref.^[95], Copyright © 2025 Wiley; (E) Schematic illustrations of Zn deposition at different interfaces of bare Zn and ZHC/Cu@Zn. (E) is reprinted with permission from Ref.^[94], Copyright © 2024 Wiley; (F) Schematic illustration of interfacial electronic interaction in PNC/TiN; (G) UV-vis absorbance spectra of PNC/TiN-I₂ before and after being immersed in electrolyte for 4 months. (F and G) are reprinted with permission from Ref.^[96], Copyright © 2026 Wiley; (H) Characterization of Sn/Al; (I) CV curves of Al||Al_xMnO₂ and Sn/Al||Al_xMnO₂ cells at a scan rate of 0.1 mV s^-1. (H and I) are reprinted with permission from Ref.^[97], Copyright © 2024 Elsevier; (J) Illustration of the improvement in conductivity and NH₄⁺ conduction in crystalline-amorphous VO heterostructure.(J) is reprinted with permission from Ref.^[98], Copyright © 2025 American Chemical Society. ZHS: Zinc hydroxysulfate; ZHC: zinc hydroxide chloride; HRTEM: High-resolution transmission electron microscope; SAED: selected area electron diffraction; CV: cyclic voltammetry; PNC: porous N-doped carbon.

During the zinc deposition process, the hydrated zinc ion [Zn(H₂O)₆]²⁺ first overcomes a substantial desolvation energy barrier to strip off its coordinated water molecules, which allows the bare Zn²⁺ to gain electrons and plate onto the anode surface^[94]. Sluggish desolvation kinetics cause the accumulation of Zn²⁺ at the interface, which can induce uneven Zn deposition. Moreover, the intercalation of large-radius hydrated Zn²⁺ causes irreversible structural evolution of the electrode and results in diminished cycling stability. Consequently, the desolvation step constitutes a critical bottleneck for interfacial reaction kinetics. To address this, a TiS₂/TiO₂ heterostructure was designed to simultaneously promote interfacial desolvation and ion/electron transfer^[95] [Figure 8C]. The built-in electric fields in TiS₂/TiO₂ provide a separating force for charged Zn²⁺ and electrically neutral water molecules, which effectively reduces the desolvation barrier [Figure 8D] and promotes the transformation of [Zn(H₂O)₆]²⁺ into Zn²⁺ during the desolvation process. The resultant mitigation of lattice distortion contributes to an extended cycle life. Furthermore, the synergistic effect of the built-in electric fields and vacancies accelerates Zn²⁺ diffusion within the bulk structure. This contributes to a high reversible capacity of 160.9 mAh g^-1 even at a high current density of 5 A g^-1. Moreover, Wang et al. subsequently designed a zinc hydroxide chloride (ZHC) and Cu heterostructure, which effectively guides efficient Zn deposition [Figure 8E]^[94]. ZHC with hydrophilic/zinophobic properties can facilitate the desolvation of hydrated Zn²⁺ prior to Zn nucleation, while the adjacent Cu serves as a zincophilic site for subsequent nucleation and deposition. This accelerates Zn²⁺ deposition kinetics and effectively suppresses the hydrogen evolution side reactions.

Heterostructure materials capable of suppressing the shuttle effect are also applicable to aqueous zinc-iodine batteries. A biomass-derived porous N-doped carbon (PNC) coated with TiN (PNC/TiN) has recently been reported as an advanced iodine host^[96]. During the reaction, the upshift of the Ti d-band center enables orbital coupling between the Ti 3d and I 5p states, which enhances the affinity for iodine while suppressing polyiodide shuttling [Figure 8F]. UV-vis spectroscopy reveals that no iodine absorption peaks are detected in the PNC/TiN-I₂ solution even after four months of rest, which corroborates the exceptional chemical immobilization of iodine species within the host structure [Figure 8G]. This design ultimately achieves an average Coulombic efficiency of 99.93% over 66,000 cycles.

Aqueous aluminum-metal batteries offer high theoretical energy density because Al³⁺ transfers three electrons during redox reactions^[99]. Nevertheless, they face critical challenges, including poor reversibility of Al deposition, the formation of a passive layer, and the competing HER. A proven strategy involves introducing foreign metals to interact with Al. The Sn/Al heterostructure using Sn with a higher work function (4.42 eV) than Al (4.28 eV) can guide the underpotential deposition (UPD) of Al^3+[97]. The UPD process, in which metal deposition occurs on a substrate at a potential less negative than its thermodynamic deposition potential, realizes efficient and controlled Al deposition. This Sn/Al heterostructure effectively promotes Al stripping/plating and reduces internal resistance. Remarkably, the surface oxide layer on Sn/Al is prone to fracture and fragmentation during rolling, resulting in a uniform distribution of the oxide throughout the composite structure without completely covering the composite surface [Figure 8H]. These results confirm that the Sn framework not only helps maintain structural stability but also generates SnO₂ during cycling, which introduces higher electronic conductivity and enhances ion diffusion. As a result, the Sn/Al||Sn/Al symmetric cells exhibit stable cycling for over 900 hours with a low overpotential < 0.5 V. In full cells, the Sn/Al (1.40 V) anode shows a higher reduction peak than that observed for bare Al (0.99 V), and the energy density is improved [Figure 8I].

Aqueous ammonium-ion batteries leverage the adsorption characteristics of non-metallic charge carriers to achieve superior volumetric energy density and rate capability^[100]. V-based materials have been widely adopted due to their excellent structural integrity and ability to accommodate the large ionic radius of NH₄⁺. Yang et al. developed a crystalline-amorphous VO₂ heterostructure (SR-VO) via controlled NaBH₄ etching of VO₂ nanobelts^[98]. The heterointerface formed between metallic amorphous VO_x and semiconducting crystalline VO₂ substantially improves electronic conductivity and reduces the NH₄⁺ adsorption energy to -2.20 eV, thereby accelerating charge transfer and ion diffusion [Figure 8J]. Critically, the amorphous layer acts as a dynamic lattice-self-adaptation buffer that can relieve intercalation-induced mechanical stress, in stark contrast to the severe lattice expansion observed in pristine VO₂. The synergistic effect of the built-in electric fields and the mechanical buffering of the amorphous layer endows SR-VO with exceptional cycling stability, which delivers a capacity of 83.4 mAh g^-1 after 10,000 cycles at 10 A g^-1 with a decay rate of only 0.004% per cycle. This strategy has also been demonstrated to be universally applicable to other V-based materials (e.g., NH₄V₆O₁₀, V₆O₁₃, V₂O₅).