Reaction mechanisms, key challenges, and future prospects of zinc-ion/magnesium-based batteries: from fundamental researches to electrode/electrolyte engineering

Zn²⁺ detachment/insertion mechanism

The deintercalation of Zn²⁺ represents the most typical form of energy storage. This phenomenon is particularly prevalent in materials featuring layered structures (V₂O₅, MnO₂) or tunnel structures (α-MnO₂)^[65]. Xu et al.^[66] claim that during discharge, Zn ions embed into the tunnels of α-MnO₂ to form a spinel-structured ZnMn₂O₄. Furthermore, through the X-ray Photoelectron Spectroscopy (XPS) and X-ray diffraction (XRD) characterization [Figure 1A and B], it was confirmed that in aqueous zinc-ion batteries, the α-MnO₂ cathode exhibited a significant Zn signal at the 1.3 V insertion state and formed the ZnMn₂O₄ crystal phase. At the 1.7 V removal state, the Zn signal significantly weakened and the crystal phase reverted to pure α-MnO₂. The original state had no Zn signal or zinc-manganese oxide phase. This clearly indicates that Zn²⁺ can undergo reversible insertion and removal in the tunnel structure of α-MnO₂, accompanied by the reversible crystal phase transformation between α-MnO₂ and ZnMn₂O₄. Moreover, the reaction processes of the cathode and anode are proposed:

Figure 1. (A) Zn 2p core level spectra of cathodic crystalline α-MnO₂ electrodes in the original, extraction, and insertion states. (B) XRD patterns of cathodic crystalline α-MnO₂ electrodes in the original, extraction, and insertion states. (A and B) Reprinted with permission from^[66]. Copyright 2011, Wiley; (C) charge and discharge curves at different rates in first cycle. (D) Ex-situ XRD patterns of the MnO₂@CFP cathode at depth of discharge at 1.3 V and 1.0 V, respectively. (E) SEM and elemental mapping images of discharged MnO₂@CFP electrode. (C-E) Reprinted with permission from^[69]. Copyright 2017, American Chemical Society. XRD: X‑ray diffraction; SEM: scanning electron microscopy; CFP: carbon fiber paper.

Cathodic reaction: Zn²⁺ +2e^- + 2α - MnO₂ ↔ ZnMn₂O₄

Anode reaction: Zn ↔Zn²⁺ + 2e^-

The results demonstrate that reversible transfer of Zn²⁺ between the anode and cathode enables energy storage and release. Zinc dissolution/deposition occurs at the anode, while Zn²⁺ insertion/extraction takes place at the cathode^[67,68]. Moreover, high capacities can be effectively achieved through the reversible intercalation of Zn²⁺ into the tunnel structure of α-MnO₂. Consequently, materials exhibiting tunnel, layered, or open-framework architectures are considered ideal cathodes for ZIBs, as they provide abundant interstitial sites to accommodate zinc ions.

Co-intercalation/deintercalation mechanism of H⁺ and Zn²⁺

Sun et al.^[69] firstly reported on the sequential ion intercalation reaction mechanism in ZIBs. They developed a reversible Zn/MnO₂ battery. The cathode was fabricated via in-situ electrodeposition on carbon fiber paper using a mildly acidic ZnSO₄ + MnSO₄ mixed electrolyte. They demonstrated through a series of experiments that the discharge curve has a fast kinetic plateau at approximately 1.4 V and a slow kinetic plateau at approximately 1.3 V, which correspond to the generation of MnOOH upon H⁺ insertion and the formation of ZnMn₂O₄ upon Zn²⁺ insertion, respectively [Figure 1C]. Without Zn²⁺ electrolyte, only the H⁺ insertion plateau is retained. The non-aqueous electrolyte has almost no capacity due to the lack of H⁺, and its capacity recovers after adding water. The ex-situ XRD shows that the characteristic peak of MnOOH appears at 1.3 V when the discharge is stopped, and the characteristic peak of ZnMn₂O₄ appears at 1.0 V when the discharge continues further [Figure 1D]. Combined with the elemental mapping results, it further proves that the energy storage mechanism of this Zn/MnO₂ battery is the continuous reversible intercalation/deintercalation of H⁺ and Zn²⁺ [Figure 1E].

Zhao et al.^[70] reported a synergistic H⁺/Zn²⁺ insertion mechanism in battery systems. The novel manganese oxide nanosheets MON [chemical formula MnO₂H_0.16(H₂O)_0.27] achieve stable ion insertion/extraction through their monoclinic P21/c(14) space group and 1 × 1 tunnel structure.

Chemical conversion reaction mechanism

The aforementioned mechanisms all center on Zn²⁺ insertion, whereas Pan et al. proposed a novel concept emphasizing that the reversible chemical conversion between α-MnO₂ and H⁺ is the key process^[71]. During discharge, metallic zinc undergoes oxidation at the negative electrode, releasing electrons and dissolving into the electrolyte as Zn²⁺ ions. This process is fully reversible during charging: MnOOH is reoxidized to α-MnO₂ at the positive electrode, while Zn²⁺ ions are simultaneously reduced and electrodeposited as metallic zinc at the negative electrode^[71]. They directly observed through transmission electron microscopy (TEM) that during the discharge process, α-MnO₂ transformed from nanofibers into short nanorods and nanoparticle aggregates with characteristic interplanar spacings of MnOOH [Figure 2A]. After charging, the structure could be restored to α-MnO₂, proving that the positive electrode undergoes reversible structural evolution and transformation reactions. Combined with scanning transmission electron microscopy-energy dispersive X-ray spectroscopy (STEM-EDS) element mapping, it was found that the discharge products only contained Mn and O, and no Zn element. Zn was only distributed in the lamellar by-products, directly confirming that Zn²⁺ did not embed into the MnO₂ lattice [Figure 2B]. Additionally, XRD characterization was used to confirm the appearance of characteristic diffraction peaks of MnOOH and ZnSO₄[Zn(OH)₂]₃·xH₂O in the discharged state, without the characteristic peaks of zinc ion insertion phase. Multiple characterizations jointly confirmed that the energy storage mechanism of this aqueous Zn/MnO₂ battery is the reversible transformation reaction between MnO₂ and H⁺ (MnO₂ ↔ MnOOH), rather than the traditional understanding of the reversible insertion/deletion mechanism of Zn²⁺.

Figure 2. (A) MnO₂ electrodes discharged to 1 V. The arrow indicates the growth directions of the short nanorods; (B) STEM-HAADF image of short nanorods and STEM-EDS mappings of the elemental distributions of Mn, O and Zn in the MnO₂ electrode in the discharged state during the first cycle. (A and B) Reprinted with permission from^[71]. Copyright 2016, Springer Nature. STEM-HAADF: Scanning transmission electron microscopy-high-angle annular dark-field; STEM-EDS: scanning transmission electron microscopy-energy dispersive X-ray spectroscopy.

Furthermore, Liu et al.^[72] found that during the discharge process of ZIBs, β-MnO₂ first reacts with protons in the electrolyte to form MnOOH which further reacts with protons to form dissolved Mn²⁺. In the subsequent cycle, the reaction stabilizes as the mutual transformation between ε-MnO₂ and MnOOH or Mn²⁺, accompanied by periodic changes in electrolyte pH throughout the process.

Deposition-dissolution mechanism

In zinc-ion batteries, the deposition-dissolution mechanism refers specifically to the reversible electrochemical processes at the metallic zinc anode, involving the reduction deposition (electroplating) and oxidation dissolution (stripping) of Zn²⁺ at the electrode/electrolyte interface during charge-discharge cycles^[73]. The MnO₂ cathode achieves a high theoretical specific capacity of 616 mAh g^-1 through a redox reaction involving double electron transfer^[74]. It is worth noting that this reaction mechanism was proposed as early as the late 20th century^[75]. Recently, Wu et al.^[76] employed in situ Raman microscopy and other techniques to investigate the previously unclear relationship between Mn³⁺ Jahn-Teller distortion and Mn-O bond length changes, as well as the distortion caused by post-processing in ex-situ characterization. They achieved the first in-situ observation of manganese Jahn-Teller distortion during operation, while also directly revealing the mechanism of dissolution-redeposition core reaction mechanism of MnO₂ cathodes. Furthermore, it was demonstrated that the addition of MnSO₄ to the electrolyte effectively suppresses manganese dissolution and significantly improves the cycling stability of the battery.

In 2022, Chen et al.^[77] investigated the long-standing controversy surrounding the energy storage mechanism of aqueous zinc-manganese (Zn-Mn) batteries. They challenged the conventional view that zinc sulfate hydroxide [Zn₄SO₄·(OH)₆·xH₂O, ZSH] as merely a byproduct. They proposed a ZSH-assisted deposition-dissolution reaction model for aqueous Zn-Mn batteries in mildly acidic sulfate electrolytes. The system elucidates the ZSH-assisted deposition-dissolution reaction mechanism: In a mildly acidic sulfate electrolyte containing Mn²⁺, charging to ~1.5 V induces a two-electron transfer reaction between ZSH and Mn²⁺, forming low-crystallinity ZnxMnO(OH)₂ nanosheets^[77]. During discharge, ZnxMnO(OH)₂ nanosheets dissolve via proton diffusion reactions, while localized pH changes in the electrolyte trigger ZSH regeneration, completing a reversible cycle. This reaction constitutes the core mechanism for aqueous Zn-Mn battery energy storage. To validate this model, a MnO₂-free Zn-ZSH battery was constructed, exhibiting identical electrochemical behavior and phase evolution to the original battery^[77]. These findings directly corroborate the pivotal role of ZSH.

Cathode material for Zn-ion batteries

In the technological development of ZIBs, cathode materials serve as a pivotal determinant for regulating electrochemical energy storage, as their intrinsic properties directly dictate the device’s energy density and operational efficiency. To date, extensively investigated cathode materials for ZIBs can be broadly categorized into manganese Mn-based materials, vanadium V-based materials, Prussian blue analogues (PBAs), among others. The following sections will systematically review these major material classes alongside their respective modification strategies.

Manganese-based compounds have attracted substantial research interest owing to their abundance, exceptional environmental compatibility, high cost-effectiveness, low toxicity, and rich valence states. Among various candidates, MnO₂ stands out as a premier choice; beyond its economic and ecological merits, it delivers superior electrochemical performance, characterized by a high theoretical specific capacity and an elevated discharge voltage plateau. Furthermore, it exhibits exceptional compatibility with aqueous electrolytes and possesses an optimal redox potential window^[78]. α-MnO₂, β-MnO₂, and γ-MnO₂ all possess tunnel structures, thus allowing Zn²⁺ to reversibly insert and extract^[79-82].

α-MnO₂ possesses a one-dimensional (1D) 2 × 2 tunnel-like micro-channeled framework that facilitates Zn²⁺ deintercalation, thereby enabling rapid ion diffusion. Recently, Li et al.^[80] addressed the challenges of kinetic retardation and poor cycling stability associated with α-MnO₂ cathodes in ZIBs. They designed a magnesium-doped α-MnO₂ (MMO) structure featuring a low concentration of oxygen vacancies. This strategy effectively suppressed manganese dissolution; consequently, the MMO cathode demonstrated significantly superior performance compared to bare α-MnO₂ after 700 cycles [Figure 3A]. Furthermore, an evaluation of Mn dissolution revealed that after 100 cycles, the Mn/Zn mass ratio for the MMO-2 electrode was restricted to 0.086, whereas it reached 0.137 for bare α-MnO₂ [Figure 3B]. Tian et al.^[83] developed a PVP-Al-MnO₂ cathode by co-intercalating Al³⁺ and polyvinylpyrrolidone (PVP) into the tunnel structure of α-MnO₂ The resulting material delivered a specific capacity of 306.8 mAh g^-1 at 0.3 A g^-1, alongside an impressive capacity retention of 93.1% over 2,000 cycles [Figure 3C]. Han et al.^[84] incorporated Eu into β-MnO₂ (20EM) as the cathode material for ZIBs. Eu doping increases the lattice spacing and enhances the oxygen vacancy content. The prepared 20EM achieved a capacity of 409 mA h g^-1 after 42 cycles at 0.2 A g^-1 [Figure 3D]. After 1,000 cycles, its performance remains stable [Figure 3E].

Figure 3. (A) Cycle Performance Comparison of Bare MnO₂ and MMO-2 at 1C Rate. (B) Mn/Zn ratio in electrolyte for both bare MnO₂ and MMO electrodes at different cycles. (A and B) Reprinted with permission from^[80]. Copyright 2024, Elsevier; (C) Cycle performance of PVP-Al-MnO₂ at 1.0 A g^-1. Reprinted with permission from^[83]. Copyright 2024, Elsevier; (D) Rate performance at different current densities. (E) Cycling performance at 1 A g^-1 of β-MnO₂ and Eu-doped MnO₂ with different molar concentrations. (D and E) Reprinted with permission from^[84]. Copyright 2023, Elsevier; (F) XRD patterns of synthesized β-MnO₂ and β-MnO₂@TA powder. (G) ICP analysis results comparing β-MnO₂ and β-MnO₂@TA after 100 cycles. (H-I) Cross-section SEM images of (H) Zn metal anode after 100 cycles using β-MnO₂ as a cathode, (I) Zn metal anode after 100 cycles using β-MnO₂@TA as a cathode. (F-I) Reprinted with permission from^[85]. Copyright 2024, American Chemical Society. MMO: Mg-doped α-MnO₂ configuration; XRD: X‑ray diffraction; TA: tannic acid; SEM: scanning electron microscopy.

Key issues in ZIBs involving Mn²⁺ leaching, phase transformations, and zinc dendrite formation in β-MnO₂ cathodes. Paik et al.^[85] proposed the use of tannic acid (TA), a hydroxyl-rich polyphenolic biomolecule, as a protective coating for β-MnO₂. Powder X-ray diffraction (PXRD) analysis confirmed that the TA coating preserved the intrinsic crystal structure of β-MnO₂ [Figure 3F]. Furthermore, the TA layer effectively mitigated manganese dissolution into the electrolyte [Figure 3G] and accelerated the phase transition kinetics during charge-discharge cycling. Notably, employing β-MnO₂@TA as the cathode also led to the successful suppression of zinc dendrite formation over 100 cycles [Figure 3H-I].

The γ-MnO₂ is classified as one of the orthorhombic crystal system, exhibiting low crystallinity with numerous defects within its lattice structure. From a microstructural perspective, the tunneling structure of this material is not regular in form. Instead, it is jointly composed of 1 × 1 tunnels characteristic of manganite and 1 × 2 tunnels typical of rhodochrosite, arranged in an irregular alternating pattern^[86]. Despite this considerable structural complexity, γ-MnO₂ delivers exceptional energy storage performance. Du et al.^[87] synthesized a composite material (NCM) comprising amino-functionalized multi-walled carbon nanotubes (MWCNTs-NH₂) and γ-MnO₂ via a hydrothermal method. The introduced amino groups not only interact with H⁺/Zn²⁺ to provide additional pseudocapacitance but also promote the redeposition of Mn²⁺ through strong adsorption, thereby compensating for the loss of active material. Consequently, the NCM electrode delivered outstanding electrochemical performance. After 120 cycles, it achieved a high specific capacity of 417.9 mAh g^-1 with a capacity retention exceeding 100% [Figure 4A]. Furthermore, the NCM demonstrated remarkable long-term cycling durability, maintaining a capacity retention of 91.4% over 3,500 cycles [Figure 4B]. Additionally, Gan et al.^[88] synthesized high-valent tungsten (W⁶⁺)-doped MnO₂ nanosheets (W-MnO₂). By utilizing W-doping to modulate the MnO₂ crystal structure, they facilitated its transformation from the β to the γ phase. Simultaneously suppressing the dissolution of Mn³⁺. At a doping ratio of 1:0.01, the redox reaction of W-MnO₂ is more complete^[88]. The discharge capacity at 0.1 A g^-1 reaches 240.6 mAh g^-1. In comparison with W-MnO₂ samples with different doping ratios at an equivalent current density, the 1:0.01 sample exhibited superior cycle life and augmented discharge specific capacity [Figure 4C]. After 1,000 cycles at 1 A g^-1, the capacity retention reached 79.8% [Figure 4D].

Figure 4. (A) Cycling performance of NCM at 0.2 A g^-1. (B) Cycling performance of NCM at 3.0 A g^-1. (A and B) Reprinted with permission from^[87]. Copyright 2023, Elsevier; (C) GCD curves of MnO₂ with different W doping ratio at 0.1 A g^-1. (D) Long-cycle performance of MnO₂ and W-MnO₂. (C and D) Reprinted with permission from^[88]. Copyright 2025, Elsevier. NCM: MWCNTs-NH₂/γ-MnO₂ nanocomposite; CCM: MWCNTs-COOH/MnO₂ composites; CM: MWCNTs/MnO₂ composites; GCD: galvanostatic charge-discharge.

In 2024, Paudel et al.^[89] investigated MnO₂ cathodes for rechargeable alkaline Zn-MnO₂ batteries. Although nanostructuring and the introduction of crystal defects enhance electrochemical performance, the underlying atomic-scale mechanisms remain elusive. Previous studies have predominantly focused on bulk MnO₂, lacking a systematic comparative analysis across its mainstream crystallographic polymorphs. To address this gap, we employed first-principles density functional theory (DFT) calculations to construct bulk models of β-MnO₂ (pyrolusite), R-MnO₂ (ramsdellite), and γ-MnO₂ featuring manganese and oxygen vacancies, alongside low-energy surface models for these three polymorphs. By simulating the core processes of MnO₂ discharge in alkaline batteries, key parameters such as defect formation energy, H⁺ binding energy, and unit cell volume changes were calculated and analyzed. Results indicate that Mn and O vacancy formation energies are higher in β-MnO₂, while R-MnO₂ and γ-MnO₂ exhibit greater defect formation propensity due to the presence of 2 × 1 large tunnels^[89]. Specifically, Mn vacancies can delay H⁺ filling of the 1 × 1 tunnel in γ-MnO₂, thereby stabilizing its crystal structure. O vacancies form impurity bands near the Fermi level, enhancing material conductivity while also creating additional pathways for H⁺ insertion. This study ultimately revealed at the atomic scale the intrinsic mechanism by which crystal defects and nanostructuring enhance the electrochemical performance of MnO₂ cathodes. It provides theoretical guidance for designing cathode materials in high-performance Zn/MnO₂ alkaline batteries, and its findings can also be partially applied to weakly acidic Zn/MnO₂ batteries based on dissolution-precipitation mechanisms.

Vanadium (V)-based materials exhibit remarkable crystallographic diversity, with distinct structural features conferring unique advantages in electrochemical performance. Among these, layered structures significantly minimize diffusion resistance during ion migration via their open two-dimensional pathways, establishing the structural foundation for fast charge/discharge kinetics. Conversely, tunnel structures - owing to their highly rigid and stable frameworks - effectively mitigate volumetric strain during cycling, thereby ensuring exceptional cycling stability. From a mechanistic perspective, the multivalent nature of vanadium facilitates multi-electron redox reactions, which serves as the fundamental origin of their extraordinarily high specific capacities.

Tunnel-type vanadium oxides have emerged as promising cathode materials for ZIBs. Nevertheless, in contrast to their layered counterparts which benefit from tunable interlayer spacings, accelerating ion transport kinetics within these rigid, fixed-size tunnels remains a formidable challenge. To address this limitation, He et al.^[90] synthesized two distinct VO₂ nanostructures by tailoring the ratio of V₂O₅ to glucose precursors. They systematically investigated the influence of macroscopic architectural arrangement within the electrode on tunnel orientation and subsequent ion transport. Their findings revealed that VO₂-D nanoribbons preferentially aligned parallel to the electrode surface and stacked along the c-axis, thereby constructing highly directional ion-transport channels. In contrast, the VO₂-A counterpart exhibited a randomly oriented configuration, resulting in isotropic ion transport. The distinct advantage of this macroscopic architectural design lies in its ability to overcome the sluggish ion diffusion inherent to tunnel-type vanadium oxides, without necessitating any modifications to the intrinsic crystal structure. Furthermore, the relatively low specific surface area of the VO₂-D nanoribbons (17.50 m² g^-1) minimizes detrimental interfacial contact with the electrolyte. This restricted interfacial exposure effectively mitigates vanadium dissolution, thereby significantly enhancing the overall cycling stability^[90]. Additionally, VO₂-D exhibits excellent performance, achieving a capacity of 420.8 mAh g^-1 at 0.1 A g^-1 and maintaining 344.8 mAh g^-1 at 10 A g^-1 [Figure 5A and B]. At 10 A g^-1 and after 5,000 cycles, the capacity retention rate was 84.3% [Figure 5C]. Moreover, the flexible pouch battery prepared from this material retained a capacity of 262 mAh g^-1 after 250 cycles at 3 A g^-1, sufficient to power an electronic watch. This demonstrates the high cycling performance of VO₂-D [Figure 5D], providing an effective approach for the commercialization of ZIBs.

Figure 5. (A) Rate Performance at Current Densities Ranging from 0.1 A g^-1 to 10 A g^-1. (B) Galvanostatic charge/discharge profiles at the current densities from 0.1 A g^-1 to 10 A g^-1. (C) Cycling performance for VO₂-D and VO₂-A cathodes at 10 A g^-1. (D) Cycling performance of Zn//VO₂-D soft-packaged battery at 3 A g^-1. (A-D) Reprinted with permission from^[90]. Copyright 2024, Wiley; (E) Cycling performances and corresponding CEs of VO₂-N400 at the current density of 0.1 A g^-1. (F) Long cycling performances of VO₂-N400 at 2 and 5 A g^-1. (E and F) Reprinted with permission from^[54]. Copyright 2023, Elsevier. CEs: Coulombic efficiencies.

The inherently low electronic conductivity of vanadium-based compounds, coupled with the sluggish diffusion kinetics of Zn²⁺ within their crystal lattices, collectively constitute the primary bottlenecks impeding their transition toward practical applications. Gu et al.^[54] synthesized nitrogen-doped VO₂(B) nanoribbons (VO₂-N) via an ammonia thermal treatment strategy. The introduction of nitrogen dopants induces lattice expansion and grain refinement within the VO₂(B) matrix, generating a disordered structure that provides abundant ion storage sites and continuous transport pathways. Consequently, this approach narrows the bandgap of VO₂ from 1.3 eV to 0.7 eV, thereby boosting intrinsic electronic conductivity, while simultaneously drastically lowering the Zn²⁺ diffusion energy barrier from 2.86 eV to 0.29 eV. As a result, the optimized VO₂-N400 electrode exhibits exceptional electrochemical performance, delivering a specific capacity of 373.7 mAh g^-1 after 100 cycles without any capacity degradation [Figure 5E]. Furthermore, the assembled battery demonstrated robust long-term cycling stability, maintaining discharge capacities of 189.6 mAh g^-1 and 123.9 mAh g^-1 over 2,000 cycles at high current densities of 2 A g^-1 and 5 A g^-1, respectively, alongside a Coulombic efficiency (CE) approaching 99.8% [Figure 5F].

Vanadium pentoxide (V₂O₅) is a well-known cathode material and serves as a typical example of layered V-based oxides. Li et al.^[91] employed an interfacial coupling strategy to synthesize porous carbon. This material was then composite with V₂O₅ to construct a cathode material for aqueous Zn-V batteries. The interface coupling between layered V₂O₅ and porous carbon forms a 3D conductive network. The porous carbon enhances interfacial conductivity and shortens ion insertion-deintercalation pathways, while V₂O₅ provides Zn²⁺ transport pathways and active sites. This composite material exhibits outstanding performance, achieving a specific capacity of 244.4 mAh g^-1 and an energy density of 171 Wh kg^-1 at 50 mA g^-1 [Figure 6A]. After 7,000 cycles, the capacity retention reached 92% with a CE of nearly 98% [Figure 6B]. Wen et al.^[92] achieved dual structural and electronic control of V₂O₅ through H₂SO₄ etching combined with hydrothermal annealing, preparing V₂O₅-H₂SO₄ cathode materials with an interlaced nanowire structure. Concurrently, a protective layer comprising CuZn₂/In/InZn₃ was deposited on the zinc foil surface, thereby impeding the growth of zinc dendrites. The etching strategy significantly increases the specific surface area of V₂O₅-H₂SO₄ compared to pristine V₂O₅. The anode protective layer further reduces battery polarization and charge transfer resistance. Meanwhile, the V₂O₅-H₂SO₄ half-cell achieved a specific capacity of 515.2 mAh g^-1 at 0.2 A g^-1 with a capacity retention exceeding 99% [Figure 6C]. After 20,000 cycles at 20 A g^-1, the capacity retention reached 79% [Figure 6D].

Figure 6. (A) Ragone plots of the V₂O₅-180. (B) Stability and Coulombic efficiency of V₂O₅-180 for 7,000 cycles at 1.0 A g^-1. (A and B) Reprinted with permission from^[91]. Copyright 2024, Elsevier; (C) Rate performance at 0.2 A g^-1 with coulombic efficiency of V₂O₅ and V₂O₅-H₂SO₄. (D) Cycling performance of full-battery at 0.2 A g^-1. (C and D) Reprinted with permission from^[92]. Copyright 2025, Elsevier; (E) Cycling durability of LVO and PEO-LVO cathodes at 10 A g^-1. Reprinted with permission from^[93]. Copyright 2024, Wiley; (F) Rate performance of KVP and P-KVP electrode under different current densities ranging from 0.1 to 10 A g^-1. (G) Long-term cycling tests and coulombic efficiency of P-KVP and KVP cathodes in different electrolytes at 0.2 A g^-1. (F and G) Reprinted with permission from^[94]. Copyright 2024, Elsevier. LVO: Pure LiV₃O₈; PEO-LVO: poly(ethylene oxide)-LiV₃O₈ superlattice nanosheets; KVP: pristine K_0.5VOPO₄·1.5H₂O; P-KVP: porous K_0.5VOPO₄·1.5H₂O polyanionic cathode.

Wu et al.^[93] prepared PEO-LiV₃O₈ (PEO-LVO) superlattice nanosheet cathodes via a simplified hydrothermal method. By inserting PEO molecules into the interlayer space of LiV₃O₈ (LVO), forming an alternating “LVO monolayer-PEO chain layer” stacking structure^[93]. PEO reduces the diffusion barrier for Zn²⁺, increases active sites, and enables highly reversible H⁺/Zn²⁺ co-intercalation. This reduces alkaline byproducts on the cathode surface and accelerates interfacial kinetics. At 10 A g^-1, the PEO-LVO anode still demonstrates a reversible capacity of 163.1 mAh g^-1, maintaining an excellent performance after 3,000 cycles [Figure 6E]. Wang et al.^[94] prepared porous K_0.5VOPO₄·1.5H₂O (P-KVP) polycationic cathode. SXRD and XPS confirmed that energy storage occurs through sequential solid solution intercalation and phase transformation during Zn²⁺ storage, accompanied by anisotropic crystal plane expansion^[94]. To address the voltage drop caused by irreversible conversion to amorphous VO_x in the P-KVP cycle, a 2 M Zn(OTF)₂/triethyl phosphate (TEP)-water (4:1) mixed electrolyte was introduced. This inhibits the conversion by forming a phosphorus-rich electrolyte interface layer. P-KVP exhibits specific capacities ranging from 181.8 mAh g^-1 to 92.0 mAh g^-1 at current densities of 0.1-10 A g^-1 [Figure 6F]. After 2,000 cycles, the capacity retention rate reached 94.2% with a CE close to 100%. After optimization, stable high-voltage output exceeding 1.2 V was maintained even after 1,000 cycles [Figure 6G]. Flexible fiber-shaped ZIBs fabricated from this material exhibited stable performance under bending angles ranging from 0° to 180°.

Within the expanding landscape of cathode materials for ZIBs, PBAs have demonstrated outstanding potential alongside traditional Mn- and V-based compounds, emerging as a prominent research frontier. Hexacyanometalates, the fundamental building blocks of PBAs, feature a robust three-dimensional (3D) open-framework structure. Benefiting from the spatial accommodation provided by this architecture for Zn²⁺ migration, zinc ions can undergo highly reversible intercalation and deintercalation processes within the crystal lattice. These unique structural advantages have firmly established PBAs as a focal point in the ongoing development of advanced ZIB cathodes. The chemical formula of PBAs is A_xM[M’(CN)₆]₁-_yγ_y·wH₂O, where A represents an alkali metal ion, M and M’ are transition metal elements, and γ is the hexacyanometallate complex vacancy^[95].

Despite the immense promise of PBA materials, their practical electrochemical performance in ZIBs often exhibits substantial variability. Syed et al.^[96] synthesized copper-substituted manganese Prussian blue analog (CuMn PBA) composite nanostructures with varying copper (Cu) and manganese (Mn) concentrations. CuMn PBA-2 exhibits a monoclinic crystal structure, high crystallinity, and uniform distribution of elements such as Cu, Mn, and Fe. Its larger specific surface area and suitable loose structure enhance contact between the electrolyte and active material. Partial substitution of Cu and Mn vacancy coordination suppress Jahn-Teller distortion to stabilize the structure. CuMn PBA-2 performed exceptionally well, it maintained excellent rate capability of 89.08 mAh g^-1 at 2.4 A g^-1[96]. After 2,000 cycles at 3 A g^-1, the capacity remained at 73.15 mAh g^-1 with a coulombic efficiency close to 100% [Figure 7A]. Yang et al.^[97] proposed constructing an epitaxial core-shell Mn@FeHCF cathode via acid etching-ion exchange, achieving epitaxial growth through lattice matching. The FeHCF shell layer provides low-strain support to suppress Jahn-Teller distortion in the MnHCF core layer, while the surface amorphous layer acts as a protective barrier. This approach avoids the capacity and voltage trade-offs associated with conventional strategies. Its performance is outstanding, with a capacity retention rate of 72.4% after 400 cycles, and a discharge capacity of 117 mAh g^-1 at 2 A g^-1 with an 83.9% capacity retention after 4,800 cycles [Figure 7B]. Furthermore, the 11.86 mAh pouch cell operates stably within -30 °C to 60 °C [Figure 7C].

Figure 7. (A) Rate performance comparison of Mn PBA and CuMn PBA-2 electrodes at various current densities. Reprinted with permission from^[96]. Copyright 2024, Elsevier; (B) Cycling performance at 2 A g^-1. (C) Discharge voltage profiles and cycling performance of Zn||Mn@FeHCF cell at 0.03 A g-1 under different circumstances. (B and C) Reprinted with permission from^[97]. Copyright 2023, American Chemical Society; (D) long-term cycling stability of cells at 5C with WEH-80%, WEH-95% and WEH-100%. Reprinted with permission from^[98]. Copyright 2022, Elsevier. (E) Illustration of the anion-cation competition chemistry. Reprinted with permission from^[52]. Copyright 2024, Wiley. PBA: Prussian blue analogue.

Sun et al.^[98] addressed the practical limitations of rechargeable ZIBs due to zinc corrosion, zinc dendrite growth, and cathode dissolution. They proposed a Daniell-type zinc/Prussian blue (MnPB) battery featuring a MnPB cathode and zinc anode, employing a water-ethanol hybrid (WEH) electrolyte at standard salt concentrations to overcome these challenges. Containing 95 wt% The WEH electrolyte containing 95 wt% ethanol enables the battery to exhibit outstanding performance, achieving a capacity retention of 91.6% after 20,000 cycles at 5C [Figure 7D]. Cui et al.^[52] addressed the issues of low discharge capacity and poor rate performance in PBA cathodes caused by the traditional reliance on cation (de)intercalation mechanisms. They firstly proposed and systematically investigated high-voltage scan-triggered anion-cation competition chemistry. Using KFeMnHCF synthesized by the coprecipitation method as the cathode, the battery achieves comprehensive superior performance in a dilute KCl + ZnSO₄ electrolyte through this competitive mechanism [Figure 7E]. At 0.6 A g^-1 to 5 A g^-1, the capacity retention exceeds 96%, with cycle stability surpassing 3,000 cycles and a CE of approximately 80%. Following the phase transition, the electronic conductivity of KZnHCF increases, the bandgap decreases from 1.15 eV to 0.99 eV, and the reduction in diffusion energy barriers for Cl^- and K⁺ is key to the performance enhancement.

In aqueous ZIBs, the cathode-electrolyte interphase (CEI) plays a critical role in dictating the structural integrity of cathode materials and the long-term cycling performance of the battery. The formation and evolution of the CEI are intricately linked to the oxidative decomposition of the electrolyte and the surface reconstruction of the active material during cycling. For Mn-based cathodes, the Jahn-Teller distortion of Mn³⁺ during charge/discharge processes induces irreversible phase transitions (e.g., from layered δ-MnO₂ to spinel ZnMn₂O₄) and triggers severe Mn²⁺ dissolution. Furthermore, these structural degradation mechanisms are concurrently accompanied by the continuous oxidation of water and electrolyte anions at the cathode surface^[99,100]. The accumulation of amorphous MnO_x, sulfate precipitates, and other parasitic byproducts results in the formation of a heterogeneous and unstable CEI on the cathode surface. This defective interphase fails to effectively suppress Mn²⁺ dissolution; worse still, it obstructs the Zn²⁺ transport pathways, ultimately leading to rapid capacity degradation^[101]. Similarly, for V-based cathode materials, the continuous dissolution of active vanadium species into the aqueous electrolyte, coupled with the accumulation of electrochemically inert byproducts, induces the formation of a defective CEI. This compromised interfacial layer constitutes the primary root cause of capacity fading in V-based systems^[102].

The surface reconstruction of ZIBs cathodes acts as a “double-edged sword” with respect to electrochemical performance. On the one hand, irreversible reconstructions - such as deleterious phase transitions in MnO₂ and the amorphization of vanadium-based materials - can lead to severe structural collapse and the loss of active materials. On the other hand, controlled in situ surface reconstruction can facilitate the formation of a robust and ion-conductive CEI. For instance, a tannic acid coating on β-MnO₂ induces the generation of a uniform CEI enriched with Mn-O-C bonds; this not only suppresses Mn²⁺ dissolution but also accelerates phase transition kinetics during cycling^[85]. Furthermore, electrolyte additives, such as MnSO₄ and sodium carboxymethyl cellulose, can regulate CEI formation via specific adsorption onto the cathode surface, thereby effectively inhibiting the oxidative decomposition of water and the dissolution of transition metal ions^[103,104].

Anode material for Zn-ion batteries

The inherent instability of zinc anodes primarily stems from inhomogeneous Zn²⁺ plating/stripping behaviors and parasitic side reactions during charge-discharge cycling, which collectively manifest as three major interconnected issues. Specifically, dendrite growth originates from the non-uniform deposition of Zn²⁺ onto the anode surface. Fundamentally, this process is driven by localized fluctuations in the electric field distribution and Zn²⁺ concentration gradients at the electrode interface. According to the Chazalviel-Brissot space charge theory, once the applied current density exceeds a critical threshold (i.e., the limiting current density), the Zn²⁺ concentration at the anode-electrolyte interface rapidly depletes to zero^[105,106]. The resulting localized space-charge accumulation generates a substantial electric field, driving the preferential deposition of Zn at protruding tips and initiating dendrite growth. Furthermore, continuous dendrite propagation can ultimately pierce the separator, triggering catastrophic internal short circuits within the battery. Simultaneously, this highly porous morphology significantly amplifies the electrochemically active surface area of the anode, thereby accelerating parasitic side reactions. Additionally, the intrinsic microscopic roughness of the zinc electrode and the continuous rupture of the brittle solid electrolyte interphase (SEI) layer synergistically exacerbate this vicious cycle of dendrite proliferation.

To address zinc dendrites and side reactions in zinc-based batteries, researchers have made numerous efforts. Hu et al.^[107] constructed a highly crystalline, flower-like two-dimensional acetylene-based covalent organic framework (COF-H) as an artificial interfacial layer on zinc anodes, synergistically inhibiting dendrite growth and corrosion through both structural design and mechanism of action. It physically isolates the Zn anode from contact with the electrolyte, preventing self-corrosion induced by water/oxygen-induced self-corrosion and HER. Linear polarization testing reveals its corrosion current (20.0 mA cm^-2) is apparently lower than that of a bare zinc electrode, with the surface remaining intact after 10 days of immersion^[107]. Furthermore, the acetylene group, ketone group, and enamine group in COF-H exhibit strong affinity for Zn²⁺. The nucleation overpotential of the COF@Cu electrode is only 23 mV [Figure 8A]. Simultaneously, the ordered micropores guide uniform transport and deposition of Zn²⁺, preventing dendrite nucleation caused by localized overconcentration. Scanning electron microscope (SEM) observations reveal that the surface of the cycled COF@Zn electrode exhibits dense, planar zinc layers, whereas the bare zinc electrode is covered with vertical dendrites [Figure 8B]. The battery exhibited a cycle life exceeding 900 hours at 3 mA cm^-2, with retention rate of 73.7% after 300 cycles [Figure 8C]. Zong et al.^[108] employed diethyl phosphoramidate (DP) as a multifunctional electrolyte additive to optimize cathode interfacial chemistry, modulate the electrolyte, and achieve formation an organic-inorganic hybrid solid electrolyte interphase at the anode [Figure 8D]. Its molecules can break the hydrogen bond of H₂O and adsorb onto the zinc surface, inducing the formation of a robust crystalline-amorphous hybrid SEI layer composed of ZnS, Zn₃N₂, and Zn₃(PO₄)₂. At the cathode surface, a DP-rich dynamic CEI forms, preventing the dissolution of V-based materials caused by H₂O while enhancing intercalation stability and reversibility. Therefore, the Zn||Zn symmetric cell achieved 1,200 h at 10 mA cm^-2 and a surface capacity of 5 mAh cm^-2 [Figure 8E]. The Zn||Cu asymmetric battery maintained CE of 99.7% after 800 cycles at 1 mA cm^-2 and 0.5 mAh cm^-2 [Figure 8F]. Furthermore, the Zn||NVO full cell exhibited a capacity retention rate of 87% after 1,000 cycles at 4 A g^-1 [Figure 8G], and the assembled pouch cell maintained stable operation beyond 200 cycles.

Figure 8. (A) Nucleation Overpotential Comparison of Zn Deposition on COF@Cu and Bare Cu Electrodes. (B) After 50 cycles, surface SEM images of the bare Zn electrode and the COF-h-modified Zn anode. (C) Cycling performance of the COF@Zn||MnO₂ and bare Zn||MnO₂ batteries at 0.5 A g^-1. (A-C) Reprinted with permission from^[107]. Copyright 2022, American Chemical Society; (D) Schematic diagram of the DP additive’s action mechanism on bulk electrolyte and electrode/electrolyte interfaces. (E) Cycling performance of Zn||Zn symmetric cells. (F) CE of Zn||Cu asymmetric batteries. (G) Cycling performance at 4 A g^-1. (D-G) Reprinted with permission from^[108]. Copyright 2025, Royal Society of Chemistry. CEI: Cathode electrolyte interphase; COF: covalent organic framework; SEI: solid electrolyte interphase; SEM: scanning electron microscopy; DP: diethyl phosphoramidate.

Constructing a 3D porous structure with high specific surface area can decrease the current density and homogenize zinc ion deposition behavior, thereby suppressing dendrite formation. Recently, Hyun et al.^[109] designed and synthesized 3D-structured copper nanosheets. This material significantly enhances the wettability of zinc electrolytes, as its hydrophilic nanostructure constructs efficient transport channels that effectively promote the mass transfer process of zinc ions. Additionally, the edge surfaces of these copper nanosheets harbor uncoordinated atomic sites that exhibit exceptional zinc affinity. These zinc (002) crystal planes not only demonstrate outstanding corrosion resistance but also suppress the propagation of zinc dendrites at the root of crystal growth [Figure 9A]. The 3D Cu NS exhibits a dominant (111) crystal plane orientation. The contact angle of the copper nanosheets was significantly reduced to 10.6°, showing their pronounced hydrophilicity [Figure 9B]. Electrochemical testing revealed that the Cu NS/Zn asymmetric cell achieved a CE of 99.8% at 5 mA cm^-2 and cycled stably for 1,700 cycles [Figure 9C]. When assembled into a full cell with commercial V₂O₅, it still cycled stability for 3,500 cycles at 5 A g^-1 [Figure 9D]. Effectively resolving zinc dendrite issues while reducing zinc consumption.

Figure 9. (A) Comparative Mechanism of Zn Deposition on Cu Foil and Cu NS. (B) Comparison of contact angles on different electrodes. (C) Electrochemical performance of asymmetric batteries based on Zn||Cu Foil, Zn||CuO NS, and Zn||Cu NS. (D) Cycling ability of Zn@Cu foil||V₂O₅ and Zn@Cu NS||V₂O₅ full cells at 5 A g^-1. (A-D) Reprinted with permission from^[109]. Copyright 2025, Elsevier; (E) Long-term stripping/plating cycle curves for pure Zn, Zn-0.4Li-0.2Mn, and Zn-0.4Li-1.0Mn symmetric cells. (F) and (G) are SEM images of ZnLiMn alloy and Zn after long-term stripping/galvanizing cycles, respectively. (H) Performance Comparison Between ZnLiMn Batteries and Zn/MnO₂ Batteries. (E-H) Reprinted with permission from^[110]. Copyright 2022, Wiley. SEM: Scanning electron microscopy.

Alloying zinc with other metals alters its crystallization behavior, thereby enhancing corrosion resistance and suppressing the HER. Zhang et al.^[110] prepared a ternary Zn-Li-Mn alloy anode through minor Li, Mn, and Zn alloying. The standard electrode potentials of Li and Mn are lower than that of Zn, enabling them to preferentially react with O₂ to form Li⁺ and Mn⁴⁺. This reduces Zn oxidation and the production of zinc-based byproducts [ZnSO₄(OH)₆·H₂O], thereby preventing anode passivation. It maintained stable cycling performance for 1,000 h without dendrite formation after cycling [Figure 9E-G]. Zn-Li-Mn-based zinc-based batteries exhibit more stable performance, maintaining over 96% capacity retention after 400 cycles, with a capacity decay rate of only 0.01% per cycle [Figure 9H]. Additionally, Jiang et al.^[111] investigated a simple and scalable chemical substitution reaction strategy. The alloying process exposes more (002) crystal planes on the ZnCo anode to guide preferential zinc deposition without dendrites. The ZnCo alloy layer reduces the onset potential of the HER, significantly decreases the corrosion current, and suppresses side reactions and corrosion. Meanwhile, the ZnCo//ZnCo symmetric cell exhibits superior cycling stability exceeding 2,000 h at 4.4 mA cm^-2 and 1.1 mAh cm^-2 [Figure 10A]. Furthermore, the cycled ZnCo anode presents a smooth, uniform surface free of protrusions and Zn dendrites [Figure 10B]. Additionally, the ZnFe//ZnFe symmetric cell demonstrated a cycle life exceeding 2,500 h [Figure 10C]. After cycling, the ZnFe anode surface appeared dense and smooth, indicating its effective suppression of zinc dendrite growth [Figure 10D].

Figure 10. (A) Cycling performance of bare Zn and ZnCo symmetric cells. (B) SEM images of ZnCo and bare Zn after 50 cycles. (C) Cycling performance of ZnFe batteries. (D) SEM images of ZnFe and after 50 cycles at 1.77 mA cm^-2 for 0.44 mAh cm^-2. (A-D) Reprinted with permission from^[111]. Copyright 2024, Elsevier; (E) Galvanostatic cycling performances of symmetrical cells with Zn@NSIC and bare Zn anodes at 10 mA cm^-2. (F) Tilted SEM images of bare Zn anode and Zn@NSI Canode after cycling in symmetrical cells. (G) Cycling performances at a current density of 5 A g^-1. (H) Cycling performances of large-sized full cells at 1 A g^-1. (E-H) Reprinted with permission from^[112]. Copyright 2025, Wiley. SEM: Scanning electron microscopy.

Forming an artificial protective layer on the surface of the zinc anode physically isolates the zinc metal from the electrolyte, suppresses side reactions, and simultaneously guides the deposition of Zn²⁺. Jeon et al.^[112] designed a near-single-ion-conductive (NSIC) protective layer. Through the synergistic interaction of sulfonated polyether ether ketone (SPEEK) and amine-functionalized metal-organic framework (MOF) [MIL-101(Cr)-NH₂]. This synergistic approach achieved a high Zn²⁺ migration number of 0.91 to fundamentally suppress dendrite growth. Additionally, the symmetric cell incorporating Zn@NSIC demonstrated stable performance under sustained high current density and capacity limitation for 1,000 h [Figure 10E]. Moreover, no dendrite formation was observed, the NSICl layer is coated [Figure 10F]. At 5 A g^-1, the Zn@NSIC maintained excellent stability after 5,000 cycles, exhibiting a reversible capacity of 135 mAh g^-1 with a retention rate of 75%, whereas bare zinc failed after 2,200 cycles [Figure 10G]. A large-scale full cell equipped with an 80 cm² zinc anode coated with NSIC demonstrated outstanding stability at 1 A g^-1. After 500 cycles, the capacity retention rate has reached 96% at a capacity of approximately 300 mAh [Figure 10H].

Unlike the robust and well-defined SEI in non-aqueous lithium-ion batteries, the formation of an effective SEI on zinc anodes in aqueous electrolytes remains a subject of intense debate. This limitation is primarily governed by the intrinsically narrow electrochemical stability window of water (1.23 V) and the competitive kinetics between water decomposition and the reduction of salt anions. In traditional mild aqueous electrolytes [e.g., ZnSO₄, Zn(CF₃SO₃)₂], the thermodynamic reduction potential of H₂O is typically more positive than that of most salt anions. Consequently, the HER preferentially occurs at the anode surface, kinetically outcompeting the reductive decomposition of electrolyte components required for SEI formation^[113,114]. Consequently, the spontaneous formation of a dense and uniform SEI film on the zinc surface is kinetically hindered. Instead, the accumulation of insulating corrosion byproducts - such as Zn₄SO₄(OH)₆·xH₂O, ZnO and Zn(OH)₂ - leads to the formation of a non-uniform passivation layer. Lacking intrinsic Zn²⁺ conductivity, this layer significantly increases interfacial impedance, exacerbates the non-uniform distribution of the electric field at the electrode interface, and further accelerates dendrite proliferation as well as the irreversible formation of “dead zinc”.

At present, in aqueous zinc-ion batteries, the construction of a uniform and Zn²⁺ conducting SEI mainly relies on three types of control strategies: (1) by using “salt-in-water” type high-concentration electrolytes, solvents or functional additives to regulate the Zn²⁺ solvation structure, reducing the activity of free water, promoting the preferential reduction of anions such as OTf^-, TFSI^-, and BF₄^-, and forming an inorganic-organic hybrid SEI mainly composed of ZnF₂, ZnS, Zn₃(PO₄)₂ and organic polymers^[115,116]; (2) a gradient zinonative artificial SEI can be in-situ constructed through electrolyte additives. The DP additive mentioned in this review can prevent the contact between the Zn negative electrode and the electrolyte while ensuring the rapid transmission of Zn^2+[108]; (3) an artificial SEI protective layer can be pre-constructed on the surface of the zinc anode. Materials such as covalent organic frameworks, metal organic frameworks, and single-ion conductive polymers can be selected^[117-119]. By utilizing the regular and orderly ionic channels, the interfacial side reactions can be inhibited at the physical level, while the migration behavior of zinc ions can be regulated to guide the uniform deposition of metallic zinc.

Electrolyte engineering

The electrolyte is a crucial component in batteries, serving as the primary medium that bridges the cathode and anode while facilitating ion transport. Its composition and properties directly dictate the interfacial electrochemical behavior of the zinc anode. However, ZIBs face several critical challenges. During the charge and discharge processes, the uneven deposition of Zn²⁺ on the anode surface induces the growth of zinc dendrites. These dendrites not only threaten battery safety by potentially penetrating the separator, but also lead to the mechanical detachment of metallic zinc. This shedding generates electrochemically inactive “dead zinc”, thereby significantly compromising the CE and cycle life of the battery^[120,121]. The presence of aqueous electrolytes makes zinc anodes highly susceptible to parasitic side reactions. Specifically, the HER consumes both active materials and the electrolyte, eventually leading to severe battery swelling. Concurrently, corrosion and passivation processes generate insulating by-products that significantly increase interfacial impedance. Ultimately, these intertwined processes cause severe degradation in overall battery performance^[122]. Traditional aqueous electrolytes are inherently constrained by the narrow thermodynamic decomposition voltage of water (1.23 V). This limitation restricts the selection of high-voltage cathode materials and subsequently hinders the enhancement of the battery’s overall energy density^[123]. Furthermore, standard aqueous electrolytes suffer from severe performance degradation under extreme temperatures. At low temperatures, freezing readily occurs, causing a drastic reduction in ionic conductivity, whereas high temperatures exacerbate parasitic side reactions - both of which severely impede reliable battery operation^[124]. Consequently, developing electrolytes capable of operating over a broad temperature range is crucial for expanding their practical applications. Kang et al.^[125] designed a electrolyte by introducing the low-polarity cosolvent diethylene glycol dimethyl ether (DGM) into a zinc perchlorate [Zn(ClO₄)₂] aqueous solution. DGM break the hydrogen bond in H₂O, regulate the freezing point of the electrolyte to -105 °C^[125]. Simultaneously, it effectively suppresses dendrite growth on the zinc anode, hydrogen evolution reactions, and the formation of byproducts. In performance testing, the Zn||Zn symmetric cell demonstrated a cycle life of up to 5,200 h at 1 mA cm^-2 and 1 mAh cm^-2 under -20 °C conditions [Figure 11A]. Zn||polyaniline batteries achieved ultra-long cycling ability of 10,000 cycles at 1 A g^-1 with no capacity decay [Figure 11B]. Dong et al.^[126] constructed high-entropy electrolytes by enhancing chloride salt diversity. They utilized high mixed entropy to regulate the coordination environment, breaking the hydrogen bond of H₂O and inhibiting [ZnCl_m]_n^2n-mn cluster aggregation, forming small and diverse solvated clusters. Concurrently, the excess entropy enhancement endowed the electrolyte with low-freezing properties, maintaining liquid state at -60 °C alongside high ionic conductivity and low viscosity [Figure 11C]. This allows the Zn||Zn battery to stably cycle for over 13,320 h at 1 mA cm^-2, with an average CE of 99.97%. It demonstrates excellent full-cell performance at -60°C when paired with cathodes such as tetrachlorobenzoquinone, VOPO₄·xH₂O, and Na₃V₂(PO₄)₃.

Figure 11. (A) Voltage-time curves of cells at -20 °C. (B) Cycling of Zn||PANI cells under -20 °C at 1A g^-1. (A and B) Reprinted with permission from^[125]. Copyright 2024, Wiley; (C) Schematic diagrams of integrating entropies in electrolytes. Reprinted with permission from^[126]. Copyright 2025, American Chemical Society; (D) Schematic representation of the chemical behavior of the hydrated deep eutectic electrolyte at the bulk phase and interface. (E) Long cycle performance plots of Zn//Zn symmetric cells at -30 °C for HDES₁₀. (F) Long cycle performance plots of Zn//Zn symmetric cells at 60 °C for HDES₁₀. (D-F) Reprinted with permission from^[127]. Copyright 2024, Wiley.

A high-quality SEI film effectively suppresses the continuous interaction between the electrode and the electrolyte, thereby preventing dendrite growth and electrode corrosion. Due to the unique nature of SEI films, the prevailing consensus is that they can only be formed in non-aqueous electrolytes. Consequently, achieving the successful formation of stable SEI films in aqueous energy storage systems still presents numerous challenges. Chen et al.^[127] designed a new hydrated deep eutectic solvent (HDES)-based electrolyte. By dual confinement of free water molecules through coordination with ethylene glycol (EG) and Zn²⁺ and hydrogen bonding with water the activity of H₂O was reduced and the electrochemical window was expanded. Simultaneously, the dissociation-reduction reaction between EG and SnI₄ facilitates the formation of an organo-inorganic hybrid solid electrolyte interphase layer with a zinc-affinity gradient on the Zn surface [Figure 11D]. This electrolyte enables the zinc/zinc symmetric cell to exhibit excellent cycling performance across a wide range of temperatures [Figure 11E and F].

Introducing small amounts of functional additives into electrolytes is a cost-effective strategy for enhancing performance. Li et al.^[128] proposed the use of sodium carboxymethyl cellulose (CMC) and sodium 4-aminobenzenesulfonate (SABS) as dual anionic electrolyte additives [Figure 12A]. It was firstly adsorbed onto the zinc (101) crystal plane, reconstructs the Zn²⁺ solvation structure, reduces water molecule coordination, and the protective layer forms on the zinc surface. In light of this, the Zn||Zn symmetric cell stably cycled for 1,800 h at 1 mA cm^-2 and 1 mAh cm^-2 without short-circuiting. The Zn||CMC+SABS||KMO full cell achieved a capacity retention rate of 94.5% after 1,300 cycles at 1 A g^-1. Furthermore, to address potential leakage issues with liquid electrolytes and meet the demands of flexible/wearable electronics, research into quasi-solid gel electrolytes is gaining momentum. Shi et al.^[129] developed a freeze-resistant hydrogel electrolyte primarily composed of zinc tetrafluoroborate [Zn(BF₄)₂] and polyacrylamide (PAM) [Figure 12B]. Its core mechanism leverages the high electronegativity of fluorine atoms in BF₄^- to form O-H…F bonds with water molecules, thereby replacing the O-H…O hydrogen bonds between H₂O. This process, combined with PAM’s 3D porous network, reduces free water and disrupts ice crystal formation. The saturated Zn(BF₄)₂-PAM hydrogel exhibits a low glass transition temperature of -117 °C. At -70 °C, the capacity retention rate is 41%, and after 100 cycles, the capacity retention rate is 100%. Samanta et al.^[130] proposed a quasi-solid ZIB system featuring a “salt-in-water” (WiS)-based molecularly crowded polymer gel electrolyte coupled with a binder-free V₂O₅@MnO₂ cathode [Figure 12C]. The gel electrolyte (PBBZf) was prepared via thermal-initiated polymerization using Poly(ethylene glycol) methyl ether methacrylate (PEGMA) as the molecular crowding agent, BEMA as the crosslinking monomer, and benzoyl peroxide (BP) as the thermal initiator, dissolved in 3M Zn(OTf)₂. This zinc-iron-nickel battery exhibits a discharge capacity of approximately 422 mAh g^-1 at 0.2 A g^-1. After 5,000 cycles at 10 A g^-1, its capacity retention rate reaches 79.83%. It effectively suppresses zinc dendrite growth and maintains stable performance under harsh conditions.

Figure 12. (A) Schematic Diagram and Performance Testing of Electrolyte Additives. Reprinted with permission from^[128]. Copyright 2025, Elsevier; (B) Schematic diagram and mechanism of the hydrogel electrolytes. Reprinted with permission from^[129]. Copyright 2023, Wiley; (C) Schematic diagram of a polymer gel electrolyte. Reproduced with permission: Reprinted with permission from^[130]. Copyright 2022, American Chemical Society.

The decomposition of aqueous electrolytes is the core origin of interfacial side reactions, SEI/CEI structural defects and battery performance degradation, which mainly includes three pathways: (1) water splitting reaction. This is the most primary decomposition pathway of aqueous electrolytes, including the HER on the anode and the oxygen evolution reaction (OER) on the cathode. The HER will increase the local pH at the anode/electrolyte interface, inducing the formation of Zn-based hydroxide byproducts and anode passivation. The OER will accelerate the dissolution of transition metal ions in the cathode and the destruction of the CEI structure. The key to inhibiting water splitting is to reduce the activity of free water, which can be achieved by regulating the solvation structure of Zn²⁺ via high-concentration electrolytes, cosolvents or additives, to break the hydrogen bond network of bulk water and raise the decomposition energy barrier of water; (2) hydrolysis and decomposition of zinc salt anions. In weakly acidic aqueous electrolytes, the anions (e.g., SO₄^2-, OTf^-, TFSI^-) will undergo hydrolysis reactions to varying degrees, which will change the pH of the electrolyte and aggravate the corrosion of the Zn anode^[131,132]. Under high voltage, the anions will undergo oxidative decomposition on the cathode surface, and the decomposition products (e.g., sulfides, fluorides, organic carboxylates) are important components of the CEI. For example, the reductive decomposition of OTf^- can form ZnF₂ with high Zn²⁺ conductivity, which is the core inorganic component of high-quality SEI in aqueous ZIBs^[133]; (3) decomposition of functional additives and cosolvents: The decomposition of organic additives/cosolvents is an important way to regulate the composition of SEI/CEI. The reductive decomposition of organic molecules on the Zn anode surface will form organic polymer components in the SEI, which can enhance the flexibility of the interfacial layer and adapt to the volume change of the electrode during cycling; while the oxidative decomposition on the cathode surface can form a compact CEI to inhibit the dissolution of active materials.

Practical capacity classification and core challenges

Guided by the practical requirements of energy storage applications, we classify the areal capacity of zinc-based batteries into three tiers: (1) Laboratory-level low areal capacity (< 2 textmAh cm^-2). This level is primarily utilized for fundamental cell analysis and is characterized by a low active material loading (< 5 mg cm^-2), a massive excess of the zinc anode (negative-to-positive capacity ratio, N/P > 10), and flooded electrolyte conditions (electrolyte-to-capacity ratio, E/C is > 100 μL mAh^-1). Consequently, testing under these conditions fails to reflect the true performance of practical devices^[145]; (2) Application-oriented medium-capacity (2-4 mAh cm^-2). This type is suitable for the basic application requirements of small consumer electronics. The active material loading is 5-10 mg cm^-2, and both the N/P ratio and E/C ratio can be designed controllably; (3) Commercial-grade high-capacity (> 4 mAh cm^-2). Satisfies the requirements for large-scale stationary energy storage applications. The active material loading is > 10 mg cm^-2, the N/P ratio is controlled within 1.2-2.0, and under lean liquid conditions, the E/C ratio is < 30 μL mAh^-1.

Under practical high-capacity conditions, zinc-based all-battery systems exhibit core bottlenecks that were not evident in the laboratory half-battery tests. Under high discharge depth conditions, the failure of the zinc negative electrode becomes significantly more severe^[146]. When the discharge depth exceeds 20%, the deposition of zinc ions becomes uneven, which is highly likely to trigger dendrite growth, volume expansion and corrosion problems, thereby causing internal short circuits and a sudden drop in capacity of the battery^[147]. High loading levels are extremely prone to causing severe structural degradation. Thick electrodes will increase the impedance of ion transport and concentration polarization, accelerate the dissolution of Mn-based and V-based active substances, and lead to irreversible structural collapse and rapid capacity degradation. The imbalance in the matching of key parameters of the device, unreasonable N/P ratio, E/C ratio, and the design of thick electrode channels will lead to the mismatch of electrochemical kinetics between the positive and negative electrodes, accelerating the failure of the entire battery. The current distribution of large-sized battery cells is uneven. For pouch cells with an area exceeding 100 cm², there is a tendency for local current density and electrolyte distribution to differ, which can lead to local overcharging and overdischarging, significantly reducing the battery’s cycle life and usage safety.

The full battery configuration design strategy that is adapted to and suitable for the practical capacity of the device

To address these challenges, various engineering solutions have been developed, focusing on electrode architectural design and key parameter matching. In the realm of electrode structure engineering, 3D porous current collectors (such as copper nanosheets and carbon felt) and gradient-structured thick electrodes are the prevailing strategies for mitigating the sluggish ion transport within high-loading electrodes. Shi et al.^[148] developed a lightweight 3D-Cu architecture to achieve reversible zinc-metal cycling. A zinc-promoting graphene (Gr) coating was used to facilitate uniform and (002)-dominant Zn growth on the 3D-Cu surface, resulting in the formation of the 3DP-Cu@Gr structure.3DP-Cu@Gr||Zn exhibited stable performance with 700 cycles at 4 mA cm^-2 and 2 mAh cm^-2, with a Coulombic efficiency of > 99.6%.The zinc-loaded 3D electrode enables the symmetric battery to achieve stable cycling at 10 mA cm^-2 for 300 h, providing a specific cumulative capacity of 86.7 Ah g^-1. And the all-printed pantacle-shape full pouch cells (3.6 mAh) exhibit 91.4% capacity retention after 200 cycles at 1 C. In terms of the positive electrode design, constructing a vertically oriented porous structure and a gradient distribution of active substances can effectively shorten the ion diffusion distance within the thick electrode, enabling stable operation under high specific capacity conditions.

In terms of key parameter matching, the N/P ratio and the E/C ratio are the core factors influencing the total battery energy density and cycle life. An excessively high N/P ratio will reduce the energy density of the device volume, while an excessively low ratio will cause the zinc negative electrode to deplete rapidly and lead to premature failure of the battery^[149]. For practical zinc-based all-battery systems, the N/P ratio needs to be adjusted to the optimal range and the E/C ratio needs to be reduced to a sufficiently low level in order to meet the core requirements of commercial applications for energy density and cost^[150].

Cathode materials Mg-ion batteries

Cathode materials dictate the energy density of MIBs. Their development has primarily focused on lowering Mg²⁺ diffusion barriers, enhancing redox activity, and improving cycling stability. Currently, these cathode materials can be broadly categorized into three types: intercalation-type, conversion-type, and organic cathodes^[176].

The Chevrel phase was the first cathode family to achieve reversible Mg²⁺ insertion, laying the foundation of magnesium-ion battery research. Yoo et al. reported layered TiS₂, which was initially pillared by PY₁₄⁺ to achieve an interlayer spacing of 1.1 nm (10.9 Å); this spacing was further expanded to 18.6 Å after subsequent MgCl⁺ insertion. This structure enables rapid diffusion of MgCl⁺ (rather than bare Mg²⁺), providing a reversible capacity of 400 mAh g^-1 at 60 °C^[177]. Liang et al. designed graphene-based MoS₂ (G-MoS₂) composites with a monolayer architecture to shorten Mg²⁺ diffusion pathways, achieving a reversible capacity of 170 mAh g^-1 at 1.8 V^[178]. The performance of V₂O₅-based materials has been further improved through composite engineering. For instance, in the RFC/V₂O₅ composite cathode reported by Cheng et al., resorcinol and formaldehyde as precursors (referred as RFC) featuring a high specific surface area of 1,188 m² g^-1 and a hierarchical pore structure effectively facilitated electrolyte diffusion and electron transport^[179]. Yin et al. prepared a Mg_0.1V₂O₅·2.35H₂O cryogel, wherein crystal water fills the interlayers to form a stable hydrated environment; together with enlarged interlayer spacing (14.5 Å) and coordination-structure tuning, the electrostatic interaction between Mg²⁺ and lattice oxygen is effectively reduced, yielding an energy density of 420 mWh g^-1[180]. Subsequently, spinel and polyanionic cathodes realized advances in high voltage and stability. Zuo et al. reported Mg(Mg_0.5V_1.5)O₄ spinel as a Mg-rich cathode; extractable Mg²⁺ and fast ion kinetics enabled 250 mAh g^-1 at 100 mA g⁻¹ and 100 mAh g^-1 retained after 500 cycles at 1 A g^-1[181]. Pérez-Vicente et al. investigated olivine-type MgMn_0.5Zn_0.5SiO₄, specifically, Ca/Ni-doped systems of this olivine material could raise the average voltage to ~3.2 V, offering new insights for high-voltage cathode design^[182].

Conversion-type cathodes achieve high capacities via redox reactions between metal cations and Mg²⁺. Zhang et al. first confirmed the conversion reaction mechanism of α-MnO₂, the MnO₆ octahedra decompose into an amorphous shell containing Mn²⁺ during magnesiation, although structural collapse compromises cycling stability^[183]. Nam et al. exploited crystal water in birnessite-type MnO₂ to enhance ion diffusion; in aqueous electrolytes the material retained 62.5% of its capacity after 10,000 cycles^[184]. Sulfide conversion cathodes show even better performance: Shen et al. prepared hollow CuS nanocubes undergoing reversible Cu₂S/Cu⁰ redox, delivering 200 mAh g^-1 at 100 mA g^-1 with no decay over 50 cycles^[185]. Xiong et al. achieved 361 mAh g^-1 in Mg(ClO₄)₂/acetonitrile (AN) using CuS nanospheres; in all-phenyl complex (APC)/tetrahydrofuran (THF), capacity rose to 164 mAh g^-1 at 55 °C, demonstrating the promoting effect of temperature on conversion-reaction kinetics within the same electrolyte system^[186]. In 2022, Wang et al. reported hollow NiS₂ nanospheres; by regulating anion redox [S^2-/(S₂)^2-], the material delivered 301 mAh g^-1 at 50 mA g^-1 after 100 cycles, offering a novel example for structural design of conversion-type cathodes^[187]. Roy et al. effectively enhanced the electrochemical performance of rechargeable magnesium batteries through a dual-cation co-intercalation strategy^[188]. By employing layered TiS₂ as a model cathode material, the co-intercalation of Mg²⁺ with Li⁺ or Na⁺ was investigated in a novel fluorinated alkoxyborate-based dual-salt electrolyte. The results reveal that the Mg-Li system significantly improves the Mg²⁺ intercalation capacity and diffusion kinetics (the Mg/Ti atomic ratio increases from 0.16 to 0.24, and the migration barrier is reduced by approximately 0.2 eV). Importantly, this is achieved while maintaining the structural stability of the host material (exhibiting an O1 phase with reversible interlayer expansion and contraction). In contrast, the Mg-Na system undergoes phase transformation (O1 → P3/O3) induced by Na⁺ intercalation, leading to structural distortion and suppressed Mg²⁺ co-intercalation. Combined experimental and theoretical investigations elucidate the co-intercalation mechanism and confirm Ti as the dominant redox center, highlighting the critical roles of monovalent ion radius and structural stability in determining co-intercalation efficacy. This work provides a new pathway for designing high-performance multivalent-ion battery systems.

Organic cathodes offer abundant resources and structural tunability. Sano et al. showed that 2,5-dimethoxy-1,4-benzoquinone (DMBQ) coordinates Mg²⁺ at carbonyl sites, delivering 260 mAh g^-1 at 0.9 V and benefiting from reversible crystalline-amorphous transitions that ensure cycling stability^[189]. Pan et al. reported 1,4-poly(anthraquinone) (14PAQ); a conjugated backbone suppresses dissolution, with only minor capacity loss over 1,000 cycles and an increased discharge 1.6 V^[190]. In 2024, Zhou et al. synthesized polymeric squarate anthraquinone (PSQAQ) via polycondensation; the squarate segments reduce solubility, while the carbonyl groups provide active sites. Following activation, the material maintained a capacity of 74 mAh g^-1 over 5,000 cycles at a high current density of 5,000 mA g^-1, demonstrating its significant potential as an organic cathode^[161].

In non-aqueous magnesium-ion batteries, the formation of the CEI and the surface reconstruction of the cathode materials are the key factors limiting cycle stability and rate performance. The primary driving forces for these phenomena are the strong Coulombic interactions between the high-charge-density Mg²⁺ ions and the host lattice, coupled with the oxidative decomposition of the electrolyte at high voltages.

For intercalation-type cathodes, the repeated insertion and extraction of high-charge-density Mg²⁺ ions will induce significant lattice stress and volume changes, leading to the irreversible amorphization of the material surface and structural reconstruction^[191]. For instance, during the magnesium insertion process, layered MnO₂ undergoes irreversible phase transformation and structural collapse, accompanied by the dissolution of Mn ions. This is the core reason for the rapid capacity degradation of manganese-based oxide cathodes in non-aqueous magnesium-ion batteries. For conversion-type cathodes (such as CuS, NiS₂), the conversion reaction leads to the pulverization of the active material, forming non-uniform CEI on the surface, ultimately causing the shuttle effect of polysulfides and the loss of active materials^{[186,192-194]}.

In aqueous systems, the oxidative decomposition of the electrolyte on the positive electrode surface constitutes the primary source of CEI components. For high-voltage oxide cathodes, nucleophilic groups in the electrolyte readily react with the electrophilic high-valent transition metal ions on the cathode surface, leading to continuous electrolyte decomposition. This results in the formation of a thick, ionically insulating CEI layer on the cathode surface, which significantly increases the interfacial impedance and hinders Mg²⁺ diffusion. The design of non-precious-metal-based electrolytes incorporating high-voltage-stable magnesium salts represents the core solution to this challenge. A boron-centered magnesium salt electrolyte with an anodic stability of up to 4.3 V can form a uniform and ultra-thin CEI on the positive electrode surface, which not only suppresses the continuous decomposition of the electrolyte but also facilitates rapid Mg²⁺ transport. In addition, surface coating of the positive electrode material can prevent direct contact between the active material and the electrolyte, inhibit detrimental surface reconstruction, and regulate the formation of a high-quality CEI.

Anode materials Mg-ion batteries

The early stage of pure-magnesium anodes was characterized by the coexistence of advantages and limitations. As early as 2000, pure magnesium metal became the preferred anode material for MIBs at that time, due to its elevated volumetric capacity and dendrite-free characteristic^[195]. Aurbach et al. first paired a pure magnesium foil anode with a Chevrel-phase Mo₃S₄ cathode to construct the first complete and practical rechargeable magnesium battery prototype. At 100% depth of discharge, the prototype achieved over 2,000 cycles with less than 15% capacity decay^[151]. However, pure-magnesium anodes have two key issues: (i) they react with most electrolytes to form a passivation layer, which hinders Mg²⁺ transport and results in a CE below 80%^[196,197]; (ii) the chemical diffusion coefficient of Mg²⁺ is low in many cathode lattices, leading to pronounced polarization at high current rates^[198].

In order to surmount the inherent limitations of pure-magnesium anodes in passivation and volumetric stability, alloying emerged as a core strategy to break performance bottlenecks. Wu et al. reported spinel Li₄Ti₅O₁₂ (LTO) as a Mg²⁺ insertion anode; its low-strain feature (volume change < 1%) enabled a reversible capacity of 175 mAh g^-1 with no obvious decay over 500 cycles^[199]. Zeng et al. prepared mesoporous Li₃VO₄/C (LVO/C) hollow spheres via spray-drying. The carbon coating effectively enhances electronic conductivity, delivering an initial discharge capacity of 318 mAh g^-1 and avoiding the passivation issue associated with pure magnesium anodes^[200]. Alloying anodes further evolved to intermetallic compounds^[201]. Niu et al. developed nanoporous Bi₆Sn₄ alloys, where a biphase synergistic effect mitigates volume expansion; the material delivered 434 mAh g^-1 at 50 mA g^-1 and retained 280 mAh g^-1 after 200 cycles at 200 mA g^-1[175]. Such materials replace simple deposition of pure magnesium with alloying reactions, effectively suppressing formation of passivation layer and improving diffusion kinetics.

Subsequently, emerging composite anodes brought two-dimensional materials and artificial interfacial modification to the forefront. Majid et al. showed that a bilayer Al₂CO structure used as an anode could deliver a theoretical capacity of 1,634.73 mAh g^-1[202]. Tasawar et al. reported layered InOCl anodes with a large theoretical capacity of 4,700 mAh g^-1 and a magnesiation-induced volume expansion of only 3.6%, indicating promising structural stability^[203]. Artificial interfacial engineering further alleviates interfacial kinetic bottlenecks. Zhang et al. constructed a Zn/ZnCl₂/MgZn₂/MgCl₂ artificial interface via a ZnCl₂-Mg metal displacement reaction, which effectively resolved the compatibility issues between Mg(TFSI)₂-based electrolytes and magnesium anodes, reducing the Mg plating or stripping overpotential to ~0.2 V and accelerating Mg²⁺ transport^[204]. Wei et al. coated the Mg anode with liquid gallium to form a Ga₅Mg₂ alloy interfacial layer; this layer contains active Ga sites that lower the nucleation energy barrier and enable more stable Mg deposition and cycling^[205].

The formation of SEI on the Mg anode in non-aqueous electrolytes and the accompanying passivation issues are the most critical interface challenges in magnesium-ion batteries, which are fundamentally different from aqueous zinc-ion batteries. While metallic Li can form a Li⁺-conductive SEI in most organic electrolytes, metallic Mg reacts with traditional carbonate- or ether-based solvents, as well as conventional magnesium salts such as Mg(TFSI)₂ and Mg(ClO₄)₂, to form a dense and ion-insulating passivation layer on its surface. This passivation layer is mainly composed of MgO, Mg(OH)₂, MgCO₃, and magnesium alkoxides. Its diffusion coefficient for Mg²⁺ is extremely low, which completely hinders the reversible plating and stripping of Mg, resulting in a sharp increase in interfacial impedance and even the complete loss of electrochemical activity at the Mg anode.

To construct a high-quality SEI with satisfactory Mg²⁺ conductivity in non-aqueous electrolytes, the key lies in regulating the reduction and decomposition pathways of the electrolyte to prevent the formation of ion-insulating magnesium-based compounds. The development of noble-metal-free electrolytes represents a milestone in solving the problem of Mg anode passivation. The APC electrolyte and magnesium borate-based electrolyte, whose reduction decomposition products are Mg-Cl complexes with high Mg²⁺ electrical conductivity, can form stable and ion-permeable SEI, and will not form a passivation layer on the Mg surface^[206]. In recent years, significant progress has been made in regulating the SEI through the use of electrolyte additives. Additives such as Mg(BH₄)₂ and GeCl₄ can undergo preferential reduction on the Mg surface, forming an Mg-B- or Ge-based SEI. This process effectively removes the native passivation layer and stabilizes the Mg plating and stripping process^[207,208].

Furthermore, artificial interface engineering is an effective strategy for addressing the passivation issue of the Mg anode. Specifically, the Zn/ZnCl₂/MgZn₂/MgCl₂ artificial interface and the Ga₅Mg₂ alloy interphase layer can prevent direct contact between the Mg anode and the electrolyte. This avoids the formation of an ion-insulating layer, while simultaneously reducing the nucleation energy barrier for Mg plating and ensuring the rapid transport of Mg²⁺ ions^[204]. Unlike the core goal of SEI regulation in aqueous zinc-ion batteries (suppressing dendrite growth and HER), the core of SEI regulation in non-aqueous magnesium-ion batteries is to solve the Mg electrode passivation problem and achieve reversible deposition/solubilization of Mg.

Electrolyte engineering

Serving as the transport medium for Mg²⁺, electrolytes must simultaneously deliver high ionic conductivity, a broad electrochemical stability window, and excellent interfacial compatibility with electrodes. Broadly speaking, these electrolytes can be classified into liquid and solid-state systems.

Liquid electrolytes have evolved from nucleophilic to non-nucleophilic systems. In 1990, Gregory et al. first reported Grignard-reagent-based electrolytes that enabled reversible Mg plating/stripping, albeit with a narrow electrochemical stability window^[209]. In 2008, Mizrahi et al. developed the APC electrolyte [(PhMgCl)₂-AlCl₃/THF], which enabled highly reversible Mg plating and stripping. Featuring a stable electrochemical window and superior ionic conductivity, it emerged as an early benchmark electrolyte for magnesium batteries^[206]. Because APC contains nucleophilic species (e.g., PhMgCl) and is corrosive to non-noble metals, it corrodes current collectors and reacts with electrophilic cathodes, severely limiting its practical use in full cells^[210]. Ionic liquids and mixed-solvent electrolytes were subsequently explored to address corrosivity. Zhao-Karger et al. formulated a (HMDS)₂Mg-AlCl₃/G₄ [hexamethyldisilazide (HMDS)] electrolyte; by tuning the solvation structure of Mg²⁺ with an ionic liquid, the electrolyte showed improved compatibility with sulfur cathodes and delivered 260 mAh g^-1 over 20 cycles^[164]. Since 2020, non-nucleophilic electrolytes have become key matches for high-voltage cathodes. Zhang et al.^[211] designed a boron-centered anion-based magnesium electrolyte (BCM electrolyte) prepared in 1,2 dimethoxyethane (DME) via a one-pot formulation of tris(2H-hexafluoroisopropyl) borate (THFPB) and MgF₂. After composition optimization and electrochemical conditioning, it exhibited an ionic conductivity of 1.1 mS cm^-1 and a Coulombic efficiency of 99.8% at the 50th cycle. Furthermore, based on the 5.3 V anodic stability measured vs. Li⁺/Li in the Li system and an empirical Li-Mg conversion, the upper limit of anodic stability for Mg²⁺/Mg was estimated to be approximately 4.3 V.

Solid-state electrolytes (SSEs) have garnered significant attention since 2018 due to their intrinsic safety and leak-free characteristics. Notably, Hassan et al. developed MOF-based solid electrolytes (MOF-SEs). The structural disorder inherent in amorphous Mg-bp3dc provides additional diffusion pathways; consequently, at 30 °C, it achieved an ionic conductivity of 3.8 × 10^-5 mS cm^-1 and a Mg²⁺ transference number of 0.335^[212].

Electrolyte additives serve as a vital auxiliary strategy to optimize the electrode-electrolyte interface (EEI), thereby enhancing overall electrochemical performance. Li et al. added Mg(BH₄)₂ to Mg[B(hfip)₄]₂/DME, removing the passivation layer on the Mg anode and stabilizing the plating/stripping overpotential after cycling, which was reduced by 50 mV compared with the blank electrolyte-significantly improving the kinetics of Mg deposition/dissolution^[208]. Zhang et al. introduced GeCl₄ to Mg(TFSI)₂/DME to form an in-situ Ge-based protection layer; symmetric cells cycled for 1000 hour without obvious decay^[207]. Wang et al. used a Pyr₁₄⁺ additive to electrostatically guide lateral Mg²⁺ deposition, suppress electrolyte decomposition and interfacial side reactions, and achieve a capacity of 50 mAh g^-1 over 40 cycles with V₂O₅ cathodes^[213].

In stark contrast to aqueous systems, the decomposition of non-aqueous electrolytes is the fundamental cause of Mg anode passivation, structural defects in the CEI, and interfacial side reactions in MIBs. It mainly involves three major pathways: (1) Reduction decomposition of organic solvents occurs on the surface of the Mg negative electrode. Traditional carbonate-based solvents [such as ethylene carbonate (EC), dimethyl carbonate (DMC), and ethyl methyl carbonate (EMC)] undergo reductive decomposition on the highly reducing Mg surface, generating MgCO₃, Mg(OH)₂, and magnesium alkoxides, which are the main components of the ionically non-conductive passivation layer. The reductive stability of ether-based solvents (such as THF and DME) is higher than that of carbonate-based solvents^[214,215]. Therefore, ether-based solvents are currently the mainstream choice for magnesium-ion battery electrolytes. However, ether-based solvents can still undergo ring-opening decomposition catalyzed by trace amounts of water and impurities, eventually leading to the formation of a passivation layer on the Mg surface; (2) The decomposition of magnesium salt anions is a key factor determining the composition of SEI/CEI. For traditional magnesium salts such as Mg(TFSI)₂ and Mg(ClO₄)₂, the TFSI^- and ClO₄^- anions will be reduced and decomposed on the Mg surface, generating insulating compounds such as MgF₂ and MgO, which leads to negative electrode passivation^[216-218]. Conversely, the anions such as [B (hfip)₄]^- and AlCl₄^- in non-protic electrolytes do not form insulating magnesium-based compounds after decomposition, but instead form Mg-Cl and Mg-B complexes with high Mg²⁺ conductivity. This is the core reason why these electrolytes can achieve reversible deposition/solubilization of Mg; (3) Nucleophilic-electrophilic side reactions occur between electrolyte components and electrodes. In traditional APC electrolytes, the strong nucleophilicity of Grignard reagents induces reactions with electrophilic high-voltage oxide cathodes and current collectors, leading to electrolyte decomposition and cathode corrosion. Consequently, non-nucleophilic electrolytes have emerged as the predominant choice in MIBs research. Furthermore, trace amounts of water and oxygen within the electrolyte react with the highly active Mg metal to form passivation products such as MgO and Mg(OH)₂, which significantly accelerate anode passivation. Therefore, rigorous control of moisture and oxygen levels is a fundamental prerequisite for the stable operation of non-aqueous magnesium-ion batteries.