Challenges, strategies and prospects in interfaces between Li metal anode and polyether-based solid-state electrolytes

Adopting polyether electrolytes with enhanced elastic property

A major challenge for polyether-based electrolytes is their poor adaptability to the dynamic volume changes of lithium metal anodes, which often leads to interfacial instability during cycling. To address this, researchers have incorporated elastic networks into polyether matrices, aiming to enhance mechanical resilience and maintain stable electrode-electrolyte contact.

For instance, Cui et al.^[83] designed an elastic Li-ion conductor with dual crosslinks for a polyether-based electrolyte, consisting of poly (propylene oxide) (PPO) elastomer, SiO₂ nanoparticles, LiTFSI, and propylene carbonate (PC). The integration of dynamic amide hydrogen bonds from Jeffamine precursors within the chemically crosslinked PPO matrix imparts high mechanical resilience to the SPE (0.32 MJ·m^-3). Using this elastic electrolyte, the LiFePO₄|Li cell shows excellent cycling stability, with a capacity retention rate of 94% after 100 cycles. Similarly, Guo et al.^[84] reported a highly stretchable and elastic polyther-based electrolyte by in situ incorporating Li salts and PEO chains into an elastomeric poly(urethane-urea) matrix (PU-EO₁₂/LiTFSI/TEG41%). In this electrolyte, the reversible noncovalent crosslinking and permanent covalent crosslinking formed between poly(urethane-urea) and PEO chains impart the SPE with pronounced elastic recovery behavior and mechanical resilience. The Li|Li symmetric cell using this electrolyte demonstrated extended operational resilience with a low polarization of 80 mV over 2,200 h at 0.1 mA·cm^-2, attributed to the formation of a self-adaptive, continuous ion transport interface [Figure 5A-C]. Later, Zheng et al.^[86] synthesized an elastic polyurethane (PU) polymer electrolyte matrix via polyaddition of poly(ethylene glycol) (PEG) and isocyanate. The PU chains were crosslinked into a polymer network via hydrogen bonding (N-H···O), resulting in a breaking elongation of up to 3,000%. This elasticity allowed the electrolyte to maintain intimate electrode contact during cycling, delivering stable performance in LFP|Li cells over 700 cycles at 0.5 C.

Figure 5. (A) Schematic illustration of the elastomeric polymer network of the PU electrolyte; (B) Schematic illustration of the self-adaptive electrolyte-electrode interface; (C) Photographs showing the adhesion of the PU-EO₁₂/LiTFSI/TEG41% electrolyte to LiFePO₄ cathode material. Figure 5A-C is reprinted with permission from Ref. [84], Copyright © 2021 Royal Society of Chemistry; (D) The chemical structure of electrolyte; (E) Long-term cycling performance of Li/Li symmetric cells using PCPE-550 and PEO-SPE at 0.05 mA·cm^-2. Figure 5D and E is reprinted with permission from Ref. [85], Copyright © 2025 Wiley. PU: Polyurethane; EO: ethylene oxide; LiTFSI: lithium bis(trifluoromethanesulfonyl)imide; TEG: tetraethylene glycol dimethyl ether; PCPE: supramolecularly crosslinked polyether electrolyte; PEO: polyethylene oxide; SPE: solid polymer electrolyte; MPEG: methoxy-terminated polyethylene glycol; MBO: tri-armed boron-centered oligomer.

More recently, a novel 3D interconnected elastic polymer electrolyte (PCEE) was designed based on PEGDA incorporating a plastic crystalline succinonitrile (SN) phase^[87]. The in situ polymerized PCEE effectively accommodated volume changes during cycling, enabling a LiFePO₄|PCEE|Li full cell to maintain a discharge capacity of 93 mAh·g^-1 at 1 C over 1,000 cycles, with minimal capacity fading (only 0.005% per cycle). In a further advancement, Ren et al.^[85] constructed a supramolecularly crosslinked polyether electrolyte (PCPE) by incorporating a tri-armed boron-centered oligomer (MBO) into a PEO/LiTFSI matrix [Figure 5D]. The flexible methoxy-terminated polyethylene glycol (MPEG) arms and dynamic Lewis acid-base interactions formed a mechanically robust and elastic polymer network, which improved structural adaptability at the Li interface. As a result, the electrolyte exhibited an enhanced elastic modulus of 1.84 MPa, enabling dendrite-free cycling for over 2,600 h and stable performance in Li symmetric cells [Figure 5E].

Providing polyether electrolytes with high mechanical strength

High-modulus SPE membranes act as physical obstacles to lithium dendrite propagation. For polyether-based electrolytes, stiffness enhancement strategies include blending or copolymerizing with rigid polymer segments, designing cross-linked polyether chains, or adding inorganic fillers and mechanical scaffolds.

Blending/copolymerization with rigid polymers. Generally, PEO segments function as the ionic conductor, while the rigid block contributes mechanical strength. Li et al.^[88] blended PEO with high-strength poly(m-phenylene isophthalamide) (PMIA) in a suitable ratio, significantly enhancing the Young’s modulus of pristine PEO SPE from 0.98 to 37.07 MPa. The mechanically robust PMIA matrix effectively suppressed lithium dendrite growth, endowing the PMIA|PEO SPE-based Li|Li symmetric cell with extended cycling stability of 468 h at 0.1 mA·cm^-2. However, the simple blending of these two segments leads to poor miscibility and macro-phase separation, resulting in uneven components distribution within the membrane.

To overcome this issue, covalent linkage between polyether segments and rigid blocks in copolymers has been explored to improve compatibility and mechanical integrity^[89-91]. Zhang et al.^[92] fabricated a PEO-b-poly(trimethyl-N-{[2-(dimethylamino)ethyl methacrylate]-7-propyl}-ammonium bis(trifluoromethanesulfonyl)imide) (PEO-b-PDM-dTFSI) copolymer. Fortunately, the covalent linking of polyether and rigid segments in copolymers can significantly enhance the intrinsic compatibility between these components. The presence of the rigid PDM-dTFSI block endows the PEO-b-PDM-dTFSI/LiTFSI SPE with high mechanical strength (10⁴ Pa) even at 100 ℃, which is approximately 10,000 times greater than that of the PEO/LiTFSI electrolyte [Figure 6A and B]. Furthermore, the tendency for phase separation was nearly eliminated by adjusting the block ratio of PDM-dTFSI and PEO, as well as the content of LiTFSI salts. The Li|Li symmetric cell utilizing this copolymer SPE demonstrates long-term stability over 1,000 h at 0.05 mA·cm^-2. Building on this strategy, Duan et al.^[93] developed a novel “rigid exterior, soft interior” structured polyether electrolyte by copolymerizing DOL with fluorinated epoxy monomer 2,2’-(2,2,3,3,4,4,5,5-Octafluorohexane-1,6-diyl) bis (oxirane) (OFBO) to construct a 3D F-DOL crosslinked network [Figure 6C]. The rigid fluorinated framework enhances the mechanical robustness, resulting in a Young’s modulus of 432 MPa, significantly higher than conventional PDOL systems^[93]. The mechanically robust network markedly prolongs Li|Li symmetric-cell lifespan (exceeding 4,000 h) and enhances the cycling durability of high-voltage LiNi_0.8Co_0.1Mn_0.1O₂ (NCM811)|Li full cells.

Figure 6. (A) Schematic illustration of the fabrication of SPEs; (B) Storage moduli (G’) of the PEO₁₁₄-b-PDM_n-dTFSI/LiTFSI electrolytes with different doping ratios upon heating when n = 15. Figure 6A and B is reprinted with permission from Ref. [92], Copyright © 2022 Multidisciplinary Digital Publishing Institute; (C) Fabrication and polymerization pathways of PDOL and F-PDOL electrolytes. Figure 6C is reprinted with permission from Ref. [93], Copyright © 2025 Wiley. SPEs: Solid polymer electrolytes; PEO: polyethylene oxide; PEO-b-PDM-dTFSI: PEO-b-poly(trimethyl-N-{[2-(dimethylamino)ethyl methacrylate]-7-propyl}-ammonium bis(trifluoromethanesulfonyl) imide); LiTFSI: lithium bis(trifluoromethanesulfonyl)imide; PDOL: poly(1,3-dioxolane); F-PDOL: F-contained PDOL; DOL: 1,3-dioxolane; OFBO: 2,2’-(2,2,3,3,4,4,5,5-Octafluorohexane-1,6-diyl) bis (oxirane).

Designing the cross-linked polyether chain structures. Constructing a polyether macromolecule network via cross-linking of liner polymer chains can improve both the mechanical robustness and thermal resistance of polyether. For instance, Cui et al.^[94] developed a crosslinked poly(1,3-dioxolane) (PSiDOL) electrolyte by introducing a multifunctional siloxane-based crosslinker to construct a robust polyether network. The resulting structure significantly enhanced the tensile strength of the membrane (25.7 MPa) and promoted the development of a durable SEI layer enriched with Li_xSiO_y and LiF. Owing to its reinforced mechanical integrity, the PSiDOL electrolyte exhibited a high CCD of 0.5 mA·cm^-2 and enabled dendrite-free cycling for over 4,000 h in Li|Li symmetric cells at 0.2 mA·cm^-2. Similarly, Tang et al.^[91] developed cross-linked polymer electrolytes (CNPE) based on PEO and poly(propylene oxide) (PPO) oligomers using surface-modified SiO₂ nanoparticles as multifunctional cross-linkers. The stable cross-linked network endowed the membrane with enhanced structural stability, allowing the Li|Li symmetric cell to sustain lithium plating/stripping at a high current density of 2.4 mA·cm^-2 for up to 1,700 h.

To further strengthen the mechanical robustness and interfacial stability of polyether-based SPEs, Sun et al.^[95] proposed an innovative dual-crosslinked network design. This strategy combines a UV-induced self-polymerized network formed by trimethylolpropane triacrylate (TMPTA), which rapidly constructs a uniform and rigid crosslinked skeleton, with a secondary, randomly crosslinked network derived from PEO and triethylene glycol dimethacrylate (TEGDME). This structural feature not only enhances the homogeneity of the membrane but also significantly reinforces its mechanical strength, achieving a tensile stress of up to 1.91 MPa. Benefiting from the improved structural integrity and interfacial compatibility, LFP|Li cells employing this SPE can retain 99.6% of their initial capacity with nearly 100% coulombic efficiency over 240 cycles.

Compositing with inorganic nanofillers. The inorganic fillers play a very crucial role in affecting the mechanical property of polyether-based SPE. Based on their dimensional properties, they can be divided into four types: particles (0D), wire-like fillers (1D), two-dimensional (2D) sheet-like fillers and three-dimensional (3D) network-like fillers.

0D oxide ceramic nanoparticles, such as SiO₂^[96], Al₂O₃^[97], ZrO₂^[98], TiO₂^[99] and Li_6.4La₃Zr_1.4Ta_0.6O₁₂ (LLZTO)^[100], are commonly employed as mechanical strengthening fillers in PEO-based SPEs due to their high modulus. For example, when 6.85% fumed SiO₂ was added into the PEO-LiClO₄ SPE [Figure 7A], the tensile strength can be increased by 2.1 times (1.9 MPa)^[101]. The symmetric cell with the SiO₂-containing SPE exhibits stable plating/stripping behavior for 380 h, significantly longer than the cycling life (210 h) of the symmetric cell with pristine PEO SPE. However, these 0D fillers tend to cluster within the polymer host due to their large interfacial energy, limiting their ability to form uniform, continuous reinforcing network.

Figure 7. (A) Physical and structural diagrams of the PEO-LiClO₄ and PEO-LiClO₄-Fumed-SiO₂ SPE. Figure 7A is reprinted with permission from Ref. [101], Copyright © 2021 Royal Society of Chemistry; (B) Scanning electron microscopy (SEM) images of vertically aligned 1D LLTO nano-arrays; (C) Cross-sectional morphology of the PEO/LLTO nano-arrays with top view sample inset. Figure 7B and C is reprinted with permission from Ref. [102] Copyright © 2025 Royal Society of Chemistry; (D) Schematic diagram of the preparing process for PEO/2D MOFs CSE; (E) Stress-strain curves of SPEs; (F) Voltage profiles of PEO/2D MOF electrolyte-based Li/Li symmetric cells. Figure 7D-F is reprinted with permission from Ref. [103], Copyright © 2023 Wiley. PEO: Polyethylene oxide; SPE: solid polymer electrolyte; 1D: wire-like fillers; LLTO: Li_0.5La_0.5TiO₃; 2D: two-dimensional; MOF: metal-organic framework; CSE: composite solid electrolyte; FS: fumed silica.

1D nanostructures can provide more continuous load transfer and mechanical reinforcement compared to 0D particles. As depicted in Figure 7B and C, Li et al.^[102] synthesized vertically aligned 1D Li_0.5La_0.5TiO₃ (LLTO) nano-arrays via a sol-gel template method, which were embedded into a PEO matrix to construct a flexible composite polymer electrolyte (VLSPE). Compared to the 0D LLTO nanoparticle/PEO electrolyte with a mechanical strength of 5.4 MPa, the aligned 1D LLTO nanowire/PEO electrolyte exhibited a significantly higher strength of 7.5 MPa^[102]. This substantial enhancement in mechanical robustness effectively stabilized the interface with the lithium anode throughout prolonged cycling, enabling Li|Li symmetric cells to function reliably for over 800 h. Similarly, Wen et al.^[104] fabricated 1D HAP nanowires embedded in PEO electrolyte, where the HAP nanowires establish a 3D network. Due to the continuous inorganic phase, the PEO-HAP composite solid polymer electrolyte (SPE) achieves a Young’s modulus of 53 MPa, showing approximately a 72-fold increase in tensile strength (6.45 MPa) compared to the PEO-LiTFSI SPE. Additionally, SiO₂ nanofibers were also prepared as fillers for PEO-based SPE^[105]. The symmetric cell using the composite electrolyte demonstrates stable deep cycling (400 h) at 0.2 mA·cm^-2.

Compared to 0D and 1D nanofillers, 2D nanoflakes exhibit more effective modification effects due to their higher surface-to-volume ratio^[106]. Shim et al.^[107] incorporated 2D boron nitride nanosheets (BNs) into PEO-LiTFSI matrix. Remarkably, the tensile strength and Young’s modulus of PEO SPE increased from 0.85 and 0.72 MPa to 2.62 and 2.91 MPa, respectively, upon the addition of BN. As a result, lithium dendrite propagation is significantly restrained. In addition, the interactions between PEO/LiTFSI chains and BN nanosheets lead to the formation of multiple interfacial sites, promoting efficient ion conduction. The symmetric cell with PEO/BN SPE demonstrates stable operation for 1,200 h without short circuit at 0.05 mA·cm^-2. Furthermore, the functionalized 2D metal-organic framework (MOF) nanosheets were integrated into PEO to augment its performance [Figure 7D]^[103]. The electron-donating effect of substituents in 2D-MOFs and their 2D morphology of synthetic MOFs enhance mechanical properties and ion migration rate. The maximum strength of the composite solid electrolyte (CSE) is approximately 4 times that of PEO/LiClO₄ electrolyte [Figure 7E], suggesting the formation of a robust mechanical barrier that effectively inhibits Li dendrite penetration. As a result, the symmetric cell maintains extended dendrite-free cycling over 500 h at 0.2 mA·cm^-2 [Figure 7F].

Incorporating the scaffold supports. The 3D framework can more effectively remove clustering fillers and provide mechanical support for SPE membrane^[108,109]. Introducing solid-state electrolytes into porous frameworks can substantially improve the mechanical stability of SPEs. For example, Li et al.^[109] demonstrated a composite electrolyte by filling PEO SPE in a 3D Ga-doping LLZO (Ga-LLZO) garnet framework [Figure 8A-C]. The interconnected 3D garnet architecture, prepared via sacrificial template method, endows the composite electrolytes with a high tensile strength of 21.8 MPa. The 3D garnet composite SPE in Li|Li symmetric configuration shows a long cycling life of 360 h at 0.4 mA·cm^-2 with the reduced dendrite growth. Additionally, Mu et al.^[111] introduced a vertically aligned Li_1.3Al_0.3Ti_1.7(PO₄)₃ (LATP) scaffold into a PEGDA-based polymer matrix to construct a composite SPE. The aligned ceramic framework markedly reinforced the structural stability of the electrolyte and provided fast Li⁺ transport channels along the vertical direction, thereby reducing interfacial resistance and improving electrode-electrolyte contact. As a result, the Li|Li symmetric cell exhibited excellent cycling stability for over 4,000 h at 0.1 mA·cm^-2 and 1,600 h at 2.0 mA·cm^-2. Nevertheless, the inherent brittleness of these ceramic frameworks results in reduced mechanical durability of the electrolyte, limiting their application in bendable solid-state batteries.

Figure 8. (A) Synthesis procedure of 3D garnet/PEO composite electrolyte; (B) SEM image of the 3D garnet framework; (C) SEM image of 3D garnet composite electrolyte. Figure 8A-C is reprinted with permission from Ref. [109], Copyright © 2019 American Chemical Society; (D) Cross-section SEM images of PEPL electrolyte; (E) Voltage profiles of Li plating/stripping cycles at 0.1 mA·cm^-2. Figure 8D and E is reprinted with permission from Ref. [110], Copyright © 2019 Wiley. 3D: Three-dimensional; PEO: polyethylene oxide; SEM: scanning electron microscopy; PEPL: polyethylene-based polymer electrolyte; LLZO: lithium lanthanum zirconium oxide; PPL: polyethylene-based polymer lithium electrolyte; PL: polymer electrolyte.

Some polymer scaffolds, such as commercial polyolefin separators^[110], porous nanocellulose membrane^[112,113], and polyacrylonitrile nanofibers film^[114,115] are flexible, mechanically strong, and lightweight, and are also used as mechanical support for SPEs. For example, Wu et al.^[110] infiltrated PEO/LiTFSI polymer electrolyte into a 5 µm thick polyethylene (PE) separator to develop a high-strength polymer electrolyte (PEPL). The robust yet flexible PE separator provides structural reinforcement to the soft PEO electrolyte, leading to a more than 1000-fold increase in Young’s modulus (445 vs. 0.2 MPa). Meanwhile, the PEPL electrolyte exhibits significantly lower thickness (7.5 µm) compared to pure PEO SPE (120 µm), which shortens the Li ⁺ diffusion time and distance between cathode and anode. Due to its sufficient mechanical strength^[110], the ultrathin PEPL electrolyte supports stable lithium electrode operation in a symmetric cell for 1,500 h at 0.1 mA·cm^-2. However, the thermal stability of polyolefin separators needs further improvement to enable the mechanical stability over a wider temperature window range.

Tung et al.^[116] incorporated a heat-resistant Kevlar-derived aramid (ANF) nanofibers membrane into PEO to increase the dendrite-suppression modulus of the electrolyte. The ultimate Young’s modulus of the PEO/ANF membrane was measured as 4.95 GPa, which is higher than that of lithium (4.91 GPa). Crucially, the membranes feature pore sizes smaller than the typical dendrite growth regions, thereby hindering lithium dendrite intrusion into the SPE. Besides, the Celgard PE melted completely after being stored at 200 °C for 10 min. In contrast, the PEO/ANF membrane remained flat, intact and undeformed, implying excellent dimensional stability over a wide temperature range. The symmetric cell based on the PEO/ANF electrolyte can operate at 10.3 mA·cm^-2 under ambient conditions, while the LiCoO₂|PEO/ANF|Li cell maintains a typical discharge capacity of 130 mAh·g^-1. Similarly, our group employed the ultralight, ultrathin and mechanically strong nanocellulose scaffold for the PEO/MPEG@LLZTO composite electrolyte (PLCN)^[112]. The formed 25 µm-thick PLCN electrolyte shows high mechanical strength of 9.3 MPa and favorable heat resistance (180 ℃). More importantly, benefiting from the ultralightweight feature of nanocellulose, the CSE possesses a low areal density of 1.67 mg·cm^-2, considerably lower than the corresponding value for a liquid electrolyte-wetted 25 µm PE separator (lean electrolyte of 1.3 g·Ah^-1, equal to 3.71 mg·cm^-2) and most of thick PEO-based SPE. The Li|Li symmetric cells with the PLCN electrolyte maintain steady Li deposition/stripping for 800 h at 0.2 mA·cm^-2. In contrast, the Li|PEO/MPEG@LLZTO|Li cell experiences premature soft short-circuiting within 60 cycles. The post-mortem views [Figure 8D] reveals pronounced dendritic lithium growth on the anodes cycled with the PEO/MPEG@LLZTO electrolyte, whereas those paired with the 25 μm PLCN electrolyte maintain a relatively uniform and compact morphology [Figure 8E]. This phenomenon indicates the excellent dendrite inhibition capability of PLCN electrolyte. Nevertheless, most polymer scaffolds exhibit limited intrinsic ionic conductivity, which may disrupt the continuity of Li⁺ transport within the electrolyte.

Therefore, composite 3D frameworks composed of mechanically flexible polymer materials and ionically conductive ceramics were introduced as skeletons to overcome the above-mentioned issues of inorganic and polymer scaffolds. For example, Zhang et al.^[114] reported a composite solid polymer electrolyte (PPL) constructed by electrospinning a 3D framework of polyacrylonitrile (PAN) and LLZTO, followed by infiltration with PEO and LiTFSI. The introduction of the continuous PAN/LLZTO network substantially enhanced the mechanical performance of the electrolyte, with the tensile strength reaching 9.47 MPa, approximately 11 times that of the pristine PEO membrane. This mechanically reinforced structure effectively mitigated lithium dendrite growth, enabling the Li|Li symmetric cell to cycle stably for over 4,000 h at 0.2 mA·cm^-2. Building on this, Liu et al.^[117] engineered a nanofiber-reinforced system (PZ/PAN@UiO66), where the electrospun ionic-conductive PAN@UiO66 scaffold served as a mechanically supportive host for a ZrO₂-containing PEO matrix. With the mechanical support from the nanofiber scaffold, the tensile strength of PZ/PAN@UiO66 increased to 1.93 MPa, nearly three times higher than that of PZ electrolyte. This enhancement endowed the PZ/PAN@UiO66 with a strong ability to resist lithium dendrite propagation, preventing the penetration of lithium dendrites even after 670 h-cycling at 0.1 mA·cm^-2.

Based on the above, for 3D fillers suitable for bendable solid-state energy storage batteries, it is essential to construct a flexible and strong 3D framework with continuous, aligned, and precisely controlled structures. Therefore, establishing novel design strategies is essential for composite 3D architectures, which should combine high flexibility, robust mechanical strength, and structural resilience.

In summary, electrolytes with enhanced mechanical rigidity can restrain dendritic lithium propagation. However, it is worth noting that electrolyte with high mechanical strength may compromise interfacial contact with the electrodes, potentially enlarging the interfacial resistance and lowering the electrochemical performance. Hence, an appropriate balance between mechanical strength and flexibility is necessary to suppress lithium dendrites while maintaining low interfacial resistance.

Endowing polyether electrolytes with self-healing property

The electrolyte breakage caused by lithium dendrite growth or significant volume changes in the anode impairs Li⁺ transport at the interface, leading to battery capacity attenuation or even short circuits. To address these problems, polyether SPEs with self-repair capabilities have been developed by incorporating dynamic covalent bonds and supramolecular interactions into polymer matrices. These dynamic networks allow autonomous repair of interfacial damage and enhance their tolerance to mechanical deformation^[118-121].

For instance, Pei et al.^[122] reported a poly(ether-urethane)-based SPE (PTMG-HDI-BHDS) with intrinsic self-healing capability by introducing dynamic disulfide bonds and hydrogen bonding into the polymer network [Figure 9A]. The resulting dual-network architecture enabled spontaneous self-healing at room temperature [Figure 9B] and maintained close interfacial contact with the Li anode, delivering outstanding cycling stability over 6,000 h at 0.2 mA·cm^-2 in symmetric Li|Li cells. Similarly, Chen et al.^[124] implemented a comparable strategy by combining disulfide bonds and hydrogen bonding within a polyether framework, which significantly improved both interfacial compatibility and electrochemical reversibility, further confirming the general applicability of this approach.

Figure 9. (A) Chemical structures of self-healing SPE; (B) The interfacial self-healing process of PTMG-HDI-BHDS SPE. Figure 9A and B is reprinted with permission from Ref. [122], Copyright © 2024 Springer Nature; (C) Schematic diagram of the self-healing mechanism; (D) Schematic representation of Li dendrite evolution in a self-healing electrolyte, which is suppressed by the electrolyte. Figure 9C and D is reprinted with permission from Ref. [123], Copyright © 2022 Wiley. SPE: Solid polymer electrolyte; PTMG-HDI-BHDS: poly(ether-urethane)-based SPE.

Building upon this, Li et al.^[123] developed a self-healing polymer electrolyte by in situ copolymerizing 2-[3-(6-methyl-4-oxo-1,4-dihydropyrimidin-2-yl)ureido]ethyl methacrylate (UPyMA) and ethylene glycol methyl ether acrylate (EGMEA), constructing a dynamic network via supramolecular interactions between UPy moieties [Figure 9C]. This supramolecular structure enables the polymer matrix to autonomously repair cracks and voids within 2 h at room temperature, thereby maintaining interfacial integrity against lithium dendrite growth and electrode volume fluctuations [Figure 9D]. Moreover, the self-healing polymer electrolyte maintains stable Li plating/stripping behavior in symmetric Li|Li cells for over 650 h at 0.5 mA·cm^-2, with suppressed voltage polarization and improved interfacial contact. The results demonstrate the efficacy of supramolecular self-healing strategies in enhancing the interfacial robustness of polyether electrolytes in lithium metal batteries.

Recently, a novel self-healing PEO-based SPE was prepared through the condensation polymerization of poly(ethylene glycol) diamine with 1,3,5-triformylbenzene^[125]. Through dynamically cross-linked imine bonds, the healed electrolyte regained stretchability up to five times its initial length after 24 h of self-healing. As a result, Li|Li symmetrical cells could maintain a stable cycle for 1,200 h at 0.1 mA·cm^-2.

In summary, the incorporation of dynamic covalent and non-covalent interactions into polyether-based SPE imparts intrinsic self-healing ability, allowing autonomous repair of bulk and interfacial damage during cycling. This self-healing behavior not only improves the mechanical durability of the electrolyte, but also ensures stable interfacial contact with lithium metal, thereby suppressing dendrite growth and enhancing long-term cycling performance.

In situ creation of SEI

Polyether electrolytes are thermodynamically unstable against Li metal and tend to undergo reductive decomposition under practical operating conditions. This decomposition gives rise to SEI formation comprising Li₂CO₃ and Li₂O precipitate on Li metal surface. However, this in situ SEI is often mechanically weak and ionically insulating, promoting filamentary lithium growth and increasing interfacial resistance. Therefore, how to construct a robust and fast-ion conducting SEI by adjusting the components on the Li metal anode is strongly considered. Accordingly, representative in situ SEI engineering strategies are discussed in this section.

(1) LiF-rich SEI with high mechanical strength

Although LiF is intrinsically brittle, its high Young’s modulus (135.3 GPa) enables it to mechanically suppress the growth of lithium dendrites^[71,126]. Consequently, extensive studies have demonstrated that increasing the LiF content in the SEI leads to significantly improved anode performance. A common approach is to tailor the polyether electrolyte composition, such as the choice of lithium salt or additive, to promote the formation of LiF-rich SEI. For instance, Armand’s team found that SEI layers formed with LiFSI-added PEO (PEO-LiFSI) electrolytes contained more LiF content than those formed with LiTFSI-added PEO (PEO-LiTFSI) electrolytes^[127]. This is due to the fact that S-F bonds in FSI^- of LiFSI salts are more easily reduced than the C-F bonds in TFSI^- of LiTFSI salts. The increased LiF component is essential for achieving a mechanically and electrochemically stable SEI layer, leading to superior cycling performance in cells with PEO-LiFSI electrolytes compared to those with PEO-LiTFSI electrolytes [Figure 10A and B]. Besides, Wang et al.^[128] introduced lithium trifluoro (perfluoro-tert-butyloxyl) borate {Li[(CF₃)₃COBF₃], LiTFPFB} into PEO-based SPE. The fluoroalkoxyl groups of TFPFB^- anion in the salt also participates in the formation of SEI, thereby enriching it with LiF [Figure 10C]. As a result, the resulting Li|Li symmetric cell exhibits a long cycling life of 600 h at 0.25 mA·cm^-2, compared to only 220 h for the cell without the LiTFPFB-based interfacial layer [Figure 10D].

Figure 10. (A) XPS spectra of Li deposits formed on Cu in Li_X/DME electrolytes; (B) Depth-dependent atomic concentrations (Li, S, F, N) of the Li deposits. Figure 10A and B is reprinted with permission from Ref. [127], Copyright © 2018 American Chemical Society; (C) Multilayer design of the salt-differentiated SPE; (D) Voltage profiles of Li/Li symmetric cells with the salt-differentiated SPE at 0.25 mA·cm^-2 (60 °C). Figure 10C and D is reprinted with permission from Ref. [128], Copyright © 2019 Wiley. XPS: X-ray photoelectron spectroscopy; DME: dimethoxyethane; SPE: solid polymer electrolyte; LiFTFSI: lithium (fluorosulfonyl)(trifluoromethanesulfonyl)imide; LiFSI: lithium bis(fluorosulfonyl)imide; LiTFSI: lithium bis(trifluoromethanesulfonyl)imide; LCO: lithium cobalt oxide; PEO: polyethylene oxide; LiTFPFB: lithium trifluoro(perfluoro-tert-butoxyl)borate; DSM: designed solid-state membrane.

Similarly, inorganic additives such as lithium nitrate (LiNO₃) have also been shown to promote LiF formation in solid polymer electrolytes. Zhang et al.^[129] introduced 5 wt% LiNO₃ into a PEO/LLZTO (PLLE) solid-state electrolyte and observed a significantly enhanced interfacial stability. During cycling, the presence of LiNO₃ facilitates the decomposition of LiTFSI, thereby increasing the LiF content within the SEI and effectively mitigating undesired reactions at the PEO/Li interface. As a result, the modified PLLE electrolyte exhibits a higher CCD of 1.0 mA·cm^-2, and the corresponding LFP|Li cell retains 128.5 mAh·g^-1 (84.5% of the maximum capacity) after 200 cycles, in contrast to the rapid performance degradation observed in the LiNO₃-free system.

Fluorine-containing organic additives also play a critical role in promoting the formation of LiF-rich interface. Fluoroethylene carbonate (FEC), known for forming stable SEI in liquid systems, provides similar benefits in solid-state systems. For example, Li et al.^[130] incorporated FEC additive into PEO-Li_0.35La_0.55TiO₃ (PEO-LLTO) electrolyte, where its coordination with Li⁺ endowed more freely mobile Li⁺ ions migrate to locally damaged regions during cycling, forming a dynamic, LiF-rich, self-repairing interfacial layer. This design enabled stable lithium deposition/stripping behavior at 0.2 mA·cm^-2 for 800 h. Besides, FEC was also introduced into a DOL-based polymer electrolyte, where it synergized with perfluorodecanoic acid (PFDA) to construct a dual-layered SEI^[131]. The inner LiF-rich layer, derived from the decomposition of FEC and LiTFSI, provided mechanical robustness to resist dendritic lithium penetration, while the outer organic layer accommodated volume fluctuations. While FEC is beneficial to promote the formation of a LiF-rich SEI and stabilize the Li/electrolyte interface, its liquid nature and flammability may compromise the mechanical robustness and flame retardancy of the composite solid electrolyte, necessitating careful consideration before practical applications.

(2) Li₃N-rich SEI with high ionic conductivity

The presence of Li₃N in the SEI has been widely demonstrated to significantly enhance interfacial ion transport due to its high ionic conductivity (~1.2 × 10^-4 S·cm^-1)^[132]. To this end, various strategies have been explored to in situ construct Li₃N-rich interphases at the Li metal interface a phenomenon confirmed by the majority of studies. For example, Hu et al.^[133] developed a sandwich-structured PEO-TiN/PEO-LiYF₄/ PEO-TiN polymer electrolyte [Figure 11A]. The incorporation of TiN additives facilitates the formation of Li₃N interface components via in situ chemical reactions with the Li metal anode. This enhancement effectively promotes Li ⁺ transfer across the Li/electrolyte interface, thereby mitigating Li dendrites growth, as depicted in Figure 11B. As a result, LFP cell employing this electrolyte showed a high-capacity retention rate of 74.18% after 300 cycles. In another study, Li et al.^[135] developed an in situ Li₃N-forming interface by introducing 2,2’-azobisisobutyronitrile (AIBN) as a nitrogen-containing additive. The Li₃N interface layer effectively improves the reversibility of the Li deposition/stripping, and the corresponding Li|Li symmetric cell exhibits the superior cycling stability at 0.2 mA·cm^-2, with a cycle life at least four times that of PEO electrolytes without the AIBN additive.

Figure 11. (A) Optical image of the PEO-TiN/PEO-LiYF₄/PEO-TiN membrane; (B) Schematic illustration of the cyclic evolution between sandwich-CSE and Li metal and the suppression of Li dendrite processes in a lithium symmetric cell. Figure 11A and B is reprinted with permission from Ref. [133], Copyright © 2023 American Chemical Society; (C) Schematic of species distribution in the CSEs-1%Li₂S₆ membrane and at the lithium metal anode interface; (D and E) TOF-SIMS depth profiles and 3D reconstruction of the sputtered volume of Li₂S^-, Li₂S₂^- species. Figure 11C-E is reprinted with permission from Ref. [134], Copyright © 2021 Wiley. PEO: Polyethylene oxide; CSE: composite solid electrolyte; TOF-SIMS: time-of-flight secondary ion mass spectrometry; LiTFSI: lithium bis(trifluoromethanesulfonyl) imide; LLZTO: Li_6.4La₃Zr_1.4Ta_0.6O₁₂.

In addition to Li₃N, other inorganic species with high ionic conductivity - such as Li₃P and Li₂S - have been extensively investigated for the construction of fast ion-conducting interphases aimed at enhancing interfacial lithium-ion transport. Furthermore, the Li₃P and Li₂S components exhibit high ionic conductivity and are utilized to enhance the lithium ion diffusion rate within the SEI^[136,137]. For example, a thin Li₃P layer is constructed in situ at the Li/electrolyte interface by introducing black phosphorus (BP) additive into a crosslinked polymer electrolyte^[137]. The Li|Li symmetric cell with this Li₃P layer demonstrates low interfacial resistance and stable cycling performance at 0.3 mA·cm^-2. Fang et al.^[134] incorporated Li₂S₆ into a PEO-based polymer electrolyte, enabling the in situ generation of an ultrathin Li₂S/Li₂S₂ layer upon contact with lithium metal [Figure 11C-E]. This interphase facilitates interfacial Li⁺ transport and mitigates dendritic lithium propagation. According to R. A. Huggins, the ionic conductivity of Li₂S could reach up to 10^-5 S·cm^-1[138]. Therefore, the Li|Li symmetric cell with the Li₂S₆-containing electrolyte delivers excellent cyclability and a high CCD of 0.9 mA·cm^-2. Furthermore, lithium polysulfidophosphate (LPS) was used as an additive in a PEO-based SPE to form a highly Li ion conductive and stable Li₃PS₄/Li₂S/LiF layer^[80]. The Li|Li cell incorporating LPS into the SPE demonstrates an interfacial resistance as low as 10 Ω·cm², enabling stable operation for 3,475 h at 0.2 mA·cm^-2. Moreover, an all-solid-state LFP|Li cell maintains a high discharge capacity of 127.6 mAh·g^-1 after 1,000 cycles at 1 C.

(3) Li alloy-rich SEI with high Li ion diffusion rate

Recently, alloy phases have been constructed on Li metal surfaces to facilitate rapid Li⁺ diffusion. Therefore, some inorganic components such as SbF₃, In₂O₃, AgF, SnF₂ and CoF₂ have been introduced into polyether-based electrolytes, which react with Li metal to form Li-rich alloy phase with high lithium ion diffusion rate, thus optimizing the lithium deposition behavior by accelerating Li ⁺ transport at the eletrode-eletrolyte interface^[139-143]. For example, SbF₃ is used as PEO electrolyte additive to in-situ generate a Li-Sb alloy-rich SEI layer on the Li metal anode^[141]. The lithophilic Li-Sb component promotes Li ⁺ migration at the interface, resulting in improved CCD and an extended cycle life (980 h) for Li|Li symmetric cell at 0.2 mA·cm^-2. Similarly, a Li-In alloy layer could also be constructed in situ via the reaction between the introduced In₂O₃ electrolyte nanofillers and Li metal anode^[142]. The In₂O₃ additive has dual functions: (i) The Li-In alloy formed by In₂O₃ could also promote the diffusion of Li ions at the interface; (ii) The remained In₂O₃ could improve the ionic conductivity of PEO electrolyte. Thus, the Li|Li symmetric cell with PEO/LiTFSI/In₂O₃ SPE exhibits superior cycle stability (over 1,200 h) at 0.1 mA·cm^-2 because of the formation of stable alloy-rich SEI.

Chen et al.^[144] further demonstrated the generation of Li-In alloy-rich SEI in a gel polymer electrolyte (GPE), synthesized via in situ polymerization of DOL and dimethoxyethane (DME) initiated by indium trifluoromethanesulfonate [In(OTF)₃]. The generated Li-In alloy (e.g., Li₃In₂, Li₁₃In₃) and LiF-rich SEI effectively reduced the energy barrier for Li⁺ deposition, enabling stable symmetric cell cycling over 4,000 h. Likewise, Lu et al.^[145] developed a PDOL-based GPE polymerizing DOL initiated by gallium trifluoromethanesulfonate [Ga(OTF)₃]. The Ga^{3 +} ions not only initiated polymerization but also reacted with Li metal to form a hybrid SEI comprising LiF, Li₂O, and Li-Ga alloys. DFT calculations revealed that the Li-Ga alloy exhibited strong Li⁺ adsorption (-2.05 eV), promoting uniform Li deposition. Consequently, Li|Li symmetric cells exhibited stable cycling for 1,600 h at 0.2 mA·cm^-2, and LiFePO₄|Li cells maintained 90.2% capacity after 200 cycles at 1 C.

Besides, other metal-based additives such as bismuthine, Al(EtO)₃ and zinc bis(2-ethylhexanoate) were also introduced to in situ form Li_xBi, LiAl and LiZn alloy interface species^[146-148], which accelerate Li⁺ conduction and suppress parasitic reactions at the Li/electrolyte interface.

Based on the quantitative comparison summarized in Table 3, several trends can be discerned. LiF-containing SEI systems generally exhibit moderate to high critical current densities (typically 0.5-2 mA·cm^-2 at ~60 °C), attributed to the high mechanical modulus of LiF. SEI rich in Li₃N or Li₂S frequently show lower interfacial resistance (often 20-60 Ω·cm²), reflecting their superior intrinsic ionic conductivity. In contrast, alloy-mediated SEI layers (e.g., Li-In, Li-Sb, Li-Al) typically achieve prolonged cycling stability, often exceeding 1,000 h, by reducing Li nucleation barriers and promoting uniform Li deposition. Collectively, these results indicate that rational SEI design in polyether-based solid-state batteries requires coordinated optimization of mechanical robustness, ionic transport, and interfacial Li deposition behavior.

In addition, several application-related trends can be identified. Most stable cycling demonstrations are conducted at elevated temperatures (50-60 °C), reflecting the temperature dependence of polyether-based electrolytes, whereas only a limited number of systems exhibit stable operation near room temperature (25-30 °C). Moreover, extended lifetimes in symmetric cells are typically achieved under relatively mild testing conditions (≤ 0.2 mA·cm^-2 and ≤ 0.2-0.4 mAh·cm^-2). In contrast, reports demonstrating stable cycling at higher current densities and areal capacities (≥ 1 mA·cm^-2 and ≥ 1 mAh·cm^-2) remain scarce.

Ex situ constructing coating

Prior to battery assembly, lithium metal is often chemically pre-treated to form an ex situ interface layer, which can effectively suppress undesirable reactions between the polyether electrolyte and lithium anode. Such an interface layer can also mitigate interfacial dendrite growth by enhancing its mechanical strength, improving the uniformity of Li-ion flux, or reducing the charge transfer resistance across the interface. Up to now, inorganic, polymer, and polymer/inorganic composites serve as common options for ex situ coating.

(1) Ex situ inorganic coating with high mechanical strength

Constructing artificial SEI layer using inorganic material has been revealed as a viable route for forming stable interface, which can isolate the electrolyte from the Li anode, and physically suppress the dendrite growth.

For instance, Fan et al.^[159] fabricated a 5 nm-thick Al₂O₃ layer on PEO SPE surface via atomic layer deposition technique to mechanically inhibit the evolution of lithium dendrites. As a result, the Li|Li symmetric cell with Al₂O₃ coated PEO can cycle stably for 500 h at 0.6 mA·cm^-2, whereas the cell with bare PEO SPE cycled only for 250 h due to mechanical failure of the relatively weak electrolyte. In a similar approach, Shi et al.^[160] developed an ultrathin (~5 nm) atomic layer deposition (ALD)-deposited Al₂O₃ layer on a PEO-TiO₂ CSE, which served not only as a mechanically robust interlayer but also chemically stabilized the electrode-electrolyte interface. Consequently, the Li|Li symmetric cell with CSE/Al₂O₃ delivered an impressive ultralong cycling life over 1,200 h at 0.1 mA·cm^-2, significantly outperforming the uncoated CSE (390 h) and bare PEO electrolyte (134 h). However, a potential limitation of such metal oxide coatings lies in their intrinsically poor lithium-ion conductivity^[161], which may impede efficient Li⁺ transport across the interface and increase interfacial resistance - thereby compromising long-term electrochemical performance if not properly engineered.

(2) Ex situ inorganic coating with fast ionic transport

To overcome the limitations of metal oxides, certain inorganic compounds that can react with Li metal to form Li-ion conductive species (e.g., Li₃N) have been ex situ constructed as interface layers, thereby simultaneously enhancing Li⁺ transport kinetics and mechanical stability at the electrolyte-anode interface. For example, a Mg₃N₂ layer was employed to decorate the PEO SPE, which could in situ transform the ionic conductor Li₃N and the benign electronic conductor Mg metal after in contact with Li foil [Figure 12A]^[162]. This interface layer could not only mechanically suppress the dendrite growth, but also could alleviate the dendrite formation by homogenizing the distribution of Li⁺/electron at the interface. Thus, a stabling cycling for 1,500 h is obtained in the Li symmetrical cell with the PEO-Mg₃N₂ selectrolyte, whereas the hysteresis voltage of the cell with PEO electrolyte increases rapidly after 400 h due to the interfacial instability. Besides, the LFP|PEO-Mg₃N₂|Li cell also delivers a stable cycling with capacity retention of 86.3% after 200 cycles.

Figure 12. (A) Schematic illustration of the cycling behavior in PEO and PEO-Mg₃N₂ electrolytes. Figure 12A is reprinted with permission from Ref. [162], Copyright © 2020 Wiley; (B and C) Digital images of pristine and Zn-coated Li foils; (D and E) Cross-sectional SEM and EDS mapping of the Zn-coated Li anode; (F) The cycling stability of Zn-coated Li at 0.1 mA·cm^-2. Figure 12B-F is reprinted with permission from Ref. [163], Copyright © 2021 Wiley. PEO: Polyethylene oxide; SEM: scanning electron microscopy; EDS: energy dispersive X-ray spectroscopy.

In addition to inorganic non-metallic interface components (Li₃N, Li₂S), Li-rich alloy components can also be generated by constructing ex-situ interface layer, thereby regulating the ions transport behavior at electrolyte-anode interface. For example, Deng et al.^[163] prepared a LiZn alloy interface layer via an alloy interaction between the pre-sputtered Zn thin film and Li foil [Figure 12B-E]. Owing to the high lithium-ion diffusion coefficient of the LiZn alloy (4.7 × 10^-8 cm²·s^-1), PEO-based symmetric cells employing LiZn-protected lithium anodes exhibited stable cycling over 900 h at 0.1 mA·cm^-2 [Figure 12F]. Similarly, a dense Li-Sn alloy layer was formed by pre-sputtering a thin Sn film onto the Li metal surface, which subsequently reacted in situ with lithium^[164]. Compared to the LiZn alloy, the Li-Sn alloy offers an even higher Li⁺ diffusion coefficient (~6 × 10^-8 cm²·s^-1), resulting in an extended symmetric cell lifespan of 2,400 h at 0.1 mA·cm^-2. The corresponding LiFePO₄|Li full cell retained 85% of its initial capacity after 500 cycles.

Most recently, Liu et al.^[165] introduced a uniform and ultra-thin (~35 nm) Ag nano interlayer onto the surface of PEO/succinonitrile(SN)-based electrolytes via magnetron sputtering, which spontaneously reacted with Li metal to form a Li-Ag alloy interphase. The resulting AgLi-rich interlayer enhanced interfacial wettability and provided fast ion transport channels. Moreover, The AgLi interphase also suppressed the side reactions between SN and lithium metal, contributing to improved interface chemical stability. Benefiting from the alloy-rich and stable interfacial structure, the Ag-modified symmetric cells exhibited prolonged cycling stability over 1,700 h at 0.05 mA·cm^-2. In addition, ex situ coating of Al, Pt and ZnO layers has also been employed to generate Li-Al, Li-Pt and Li-Zn alloy interphases in situ during cycling^[166,167]. These alloy layers further accelerate interfacial Li⁺ transport and improve electrochemical performance. However, although these alloy interlayers effectively facilitate Li⁺ transport, their intrinsic electronic conductivity may compromise interface performance by inducing undesired lithium plating above the alloy layer or triggering electrolyte decomposition (e.g., PEO degradation), thereby posing challenges to long-term cycling stability.

To address the challenges posed by the electronic conductivity of pure alloy layers, Wang et al.^[168] constructed a LiF/Li₃Sb composite layer in situ by spraying the SbF₃-contained dimethoxyethane solution on fresh Li foil. Herein, the LiF component formed simultaneously with the Li₃Sb alloy bestows insulating qualities to the interface layer, preventing reduction of the Li ⁺ on the surface. Due to the improved mechanical stability and ionic diffusion rate at the interface, the SbF₃-modified Li (SbF₃@Li)|SbF₃@Li cell with PEO-Li_6.4La₃Zr_1.4Ta_0.6O₁₂ electrolyte can cycle stably for more than 300 h at 0.2 mA·cm^-2, while maintaining a stable interfacial structure between the electrolyte and SbF₃@Li. These results highlight that the suitable inorganic materials as ex situ coatings are essential for preventing the decomposition of polyether electrolytes as well as the formation and growth of dendrites.

(3) Ex situ polymer coating for improving interfacial contact

In addition to inorganic coatings, polymer materials that are chemically stable with lithium metal, capable of accommodating anode volume changes during cycling, or able to react with lithium to form lithium-ion conductors, have been explored as surface coatings to enhance the interfacial stability between the anode and solid electrolytes.

For example, Cui’s group introduced an adaptive layer (ABL) consisting of PEO and low molecular-weight polypropylene carbonate (PPC) to ensure interface stability [Figure 13A-E]^[169]. In the presence of the viscoelastic layer, the interface resistance of the Li|LiFePO₄ cell only increased by 20% after 150 cycles due to the stabilized interface contact, which is much lower than the increase (117%) in the resistance of cell without this layer [Figure 13F]. Similarly, Wang et al.^[171] proposed an asymmetric multilayered solid composite electrolyte (MSCE) design consisting of a PEO buffer layer, a PEO-LLZTO composite middle layer, and a PPC coating on the anode side. The PPC layer serves as a self-sacrificial component that reacts with lithium to form Li alkoxides and alkyl carbonates (e.g., ROCO₂Li, ROLi), thereby improving interfacial wettability and reducing resistance. Benefiting from this adaptive architecture, the MSCE enabled excellent interfacial contact and dendrite suppression, achieving a high ionic conductivity of 3.21 × 10^-4 S·cm^-1 at 30 °C. Consequently, LiFePO₄|MSCE|Li cells exhibited 94.5% capacity retention after 150 cycles at 0.1 C, while symmetric Li|MSCE|Li cells maintained stable cycling for 500 h at 0.4 mA·cm^-2.

Figure 13. (A-C) Photographs of fully dried PEO SPE, ABL, and PPC SPE; (D) Cross-sectional SEM images of Li|Li symmetric cells without ABL layer after cycling; (E) Cross-sectional SEM images of Li|Li symmetric cells with ABL layer after cycling; (F) Nyquist plots of full cell before and after cycling at 50 °C. Figure 13A-F is reprinted with permission from Ref. [169], Copyright © 2019 American Chemical Society; (G) Illustration of Tween-grafted lithium metal; Figure 13G is reprinted with permission from Ref. [170], Copyright © 2018 American Chemical Society. PEO: Polyethylene oxide; SPE: solid polymer electrolyte; ABL: adaptive layer; PPC: polypropylene carbonate; SEM: scanning electron microscopy; W/O: without; CPE: constant phase element.

Moreover, Liu et al.^[170] constructed a polymer interfacial layer by grafting tween-20 with sequential oxyethylene groups on lithium metal anode (TG-Li) to facilitate Li⁺ conduction and stabilize the interface [Figure 13G]. As a result, the cycling stability of LFP|TG-Li cell is much better than the LFP|Li cell, delivering a reversible capacity of 132.5 mAh·g^-1 after 200 cycles at 1 C. Moreover, the F-rich poly(vinylidene fluoride) (PVDF) polymer applied as a functional coating on PEO SPE can react with lithium, resulting in the in situ formation of a LiF-enriched interface layer^[172]. Besides, after constructing PVDF skin, the tensile strength of PEO electrolyte also increased nearly ten times, which is beneficial to mechanically inhabiting the Li dendrite growth. The Li|Li symmetric cell shows a stable long-term cycling performance over 1,000 h at 0.1 mA·cm^-2 and maintains stable Li/PEO interface resistance.

In addition, polymer interface layers have also been integrated into asymmetric polymer electrolyte architectures to further improve interfacial contact with lithium metal. Recent studies demonstrated that the construction of bilayer or quasi-double-layer electrolytes with tailored interfacial chemistries on the anode side can effectively stabilize the Li surface, enhance ionic conductivity, and suppress dendrite formation, thereby enabling long-term cycling stability even under high-voltage or high-rate conditions^[173,174].

(4) Ex situ polymer/inorganic coating with excellent comprehensive performance

Recently, by combining the complementary advantages of polymer and inorganic materials, composites have become a better choice for artificial layers. For instance, Wang’s group constructed a poly(vinylidene-co-hexafluoropropylene)/lithium nitrate (PVDF-HFP/LiNO₃, PFH/LN) composite layer on Li anode surface for PEO-LLZTO electrolyte^[175]. In the artificial layer, PVDF-HFP can physically suppress Li dendrites penetration and thus maintain the structural integrity of interface layer due to its high mechanical strength and moderate flexibility; the LiNO₃ additives could react with Li to form highly ionic conductive Li₃N components, thus inducing the uniform Li deposition at the interface. With the PFN/LN artificial layer, the cycle life (1,800 h) of Li|Li symmetric cell is much longer than that (96 h) of the cell with bare Li anode at 0.2 mA·cm^-2. Similarly, the PVDF-HFP/CuF₂ composite could also be employed as Li metal interface layer^[176]. The in-situ reaction of CuF₂ with Li metal anode generates LiF and Cu components, which improves the ion transport rate and t interfacial mechanical robustness. Applied in the Li|Li symmetric cell with PEO-based electrolyte, the PVDF-HFP/CuF₂ modified Li anode demonstrates extended cycle life (1,000 h) than the bare Li at 0.4 mA·cm^-2. Furtherly, the LFP full cell with the modified Li anode shows excellent cycle stability (0.032% capacity fading rate per cycle) at 1 C. Consequently, the synergistic effect of polymer and inorganic components could efficiently regulate the Li plating/stripping behavior to illustrate excellent interface stability.

Designing 3D composite anode to decrease interface mechanical deformation

Since the volume change of the Li anode during cycling is infinite due to its hostless nature, it would cause SEI damage and discontinuous electrolyte-electrode contact. Therefore, developing 3D Li composite anode is quite effective on maintaining the mechanical stability, as the 3D structure can accommodate the huge volume change of Li metal, thereby reducing the interface mechanical deformation.

Due to the hostless nature of lithium metal, the anode undergoes significant and virtually unlimited volume changes during repeated deposition/dissolution, which leads to continuous damage of the SEI and loss of contact at the electrolyte-electrode interface. To address this issue, the design of 3D lithium composite anodes has emerged as a promising strategy for enhancing mechanical stability. The porous architecture of 3D hosts can effectively buffer the drastic volume fluctuations of Li metal, thereby mitigating interfacial mechanical stress and deformation.

To stabilize the interface between lithium metal and polyether-based solid electrolytes, various 3D anode architectures have been explored. For instance, Chen et al.^[177] fabricated a 3D Li anode via infusing molten Li into Ni foam disk for dendrite-free ASSBs [Figure 14A-F]. The volume change alleviated by 3D Li extends the cycle life to 600 h at 0.5 mA·cm^-2 for the symmetric cell with poly(ethylene carbonate)-based SPE, whereas the cell with Li foil can only cycle for 315 h. Later, Ye et al.^[178] prepared a 3D-structured lithium/copper mesh composite anode for PEO-based solid-state batteries [Figure 14G-I], which endows the LFP full cell with the enhanced cycle stability with high interfacial stability.

Figure 14. (A and B) The illustration of Li deposition behaviors of Li metal cell with bare Li foil and 3D Li metal anode; (C and D) Digital images of Ni foam and Li-Ni composite anode; (E) SEM image of Ni foam; (F) SEM image of Li-Ni composite anode. Figure 14A-F is reprinted with permission from Ref. [177], Copyright © 2019 Wiley; (G) EDS mapping images of LG composite anode; (H) Photograph of copper mesh; (I) Photograph of lithium/copper mesh. Figure 14G-I is reprinted with permission from Ref. [178], Copyright © 2021 American Chemical Society. 3D: Three-dimensional; SEM: scanning electron microscopy; EDS: energy dispersive X-ray spectroscopy; PEO: polyethylene oxide.

However, in contrast to liquid electrolyte batteries, where the electrolyte readily penetrates a 3D network, incorporating such 3D anode structures in all-solid-state batteries to sustain ionic conduction remains challenging. Therefore, Zhu et al.^[179] put forward a 3D composite anode (LG), which possesses a vertical alternating array structure. This structure not only provides continuous lithium-ion transport channels into the composite anode, but also accommodates the cyclic expansion and contraction of lithium during repeated cycling. The symmetric cells based on this 3D composite anode can operate stably for 800 cycles at a large capacity of 5 mAh·cm^-2, and the corresponding solid-state cell exhibits 93.26% capacity retention rate after 830 cycles at 5 C. Besides, Liu et al.^[180] incorporated the 3D composite Li anode [metallic Li in layered reduced graphene oxide layers (Li-rGO)] with the pre-loaded flowable PEG ion-conducting interphase into the ASSB system. The 3D rGO host could compensate for the Li metal volume changes, and the penetrated PEG components could enhance the intrinsic ionic diffusion capacity of anode. As a result, the cycle life of symmetric cell with Li-rGO anode could be improved by nearly seven times compared to that of the cell a flat Li foil at 0.5 mA·cm^-2. Furthermore, Zhu et al.^[181] constructed lithium ion diffusion pathways by incorporating Li-Al alloy with strong lithium affinity and in situ-formed Li₃N within the 3D lithium/rGO composite anode (LAG), effectively suppressing dendrite growth and enabling impressive electrochemical performance (capacity retention of 86.3% after 200 cycles for the solid state LFP|LAG cell).

As a complementary approach, Park et al.^[182] explored the use of 3D porous hosts combined with viscous polymer electrolytes to stabilize lithium deposition. The viscous nature of the poly(ethylene glycol) dimethyl ether (PEGDME)-based polymer electrolyte allowed it to effectively infiltrate the 3D host structure, forming continuous ion transport pathways and accommodating lithium volume changes during cycling. This design promoted uniform Li deposition and suppressed short-circuiting even at high areal capacities (5 mAh·cm^-2). In addition, Liu et al.^[183] developed an integrated 3D lithium-boron (LiB) fiber anode combined with polymer electrolyte to construct dual ion/electron transport pathways. The electrolyte effectively infiltrated the porous LiB scaffold, forming continuous ion-conduction channels while maintaining mechanical flexibility. The 3D structure enabled more uniform lithium deposition and reduced interfacial stress, while the in-situ formed stable SEI further enhanced interfacial stability. These results highlight that 3D composite anodes can effectively buffer volume changes and improve the structural integrity of the lithium-polyether electrolyte interface.

Employing Li-alloy anode to improve interface chemical stability

Li can be mixed with Si, Ge, Mg, In, Zn, Si, Al, etc. to form Li alloys. Compared with Li metal, Li-alloy anodes exhibit slightly higher operating potentials, which can reduce parasitic reactions with SPEs and enhance interfacial chemical stability. Furthermore, many Li alloys possess higher Li diffusion coefficient, which can promote uniform Li distribution during cycling, mitigating the formation of lithium dendrites.

For example, LiAl alloys are often used as anode materials for liquid batteries because of their good air stability, high Li diffusion coefficient (6.0 × 10^-10 Ω^-1·cm^-1) and relatively high redox potential of 0.3 V versus Li/Li⁺ [Figure 15A]^[186-188]. Until 2001, Bang et al.^[189] and coworkers first investigated the cyclability of the LiAl alloy anode material in the LiAl|PEO₃₀-LiTFSI|V₂O₅ solid state cells. Due to the low lithium activity in the LiAl alloy, the parasitic reaction on the anode significantly reduced. As a result, the PEO-based ASSBs with Li-Al anodes maintained a coulombic efficiency close to 100%. Later, Pan et al.^[184] also applied Li-Al alloy as the anode material in an all-solid-state Li-S battery. The symmetric cell with Li_0.8Al alloy anode shows steady operation for over 2,500 h at 0.5 mA·cm^-2, further implying the excellent anode-electrolyte compatibility [Figure 15B]. And they concluded that the volume change (74.95%) is modest during the biphasic lithiation reaction of Al, which further promotes interfacial mechanical stability and anode structural stability. Besides, the Li₁₀GeP₂S₁₂ electrolyte exhibits excellent stability upon contact with Li_0.8Al anode because of the smaller potential difference between sulfides and alloy than that between sulfides and Li metal. However, when the Li₁₀GeP₂S₁₂ electrolyte is in direct contact with a pure lithium metal anode, Li₂S and Ge^{4 +} components are generated at the interface, resulting in the deterioration of the stability of the anode-electrolyte interface. Similarly, Li-In alloy anode could also improve the electrode-electrolyte interface chemical stability due to its high operating potential (0.6 V) versus Li/Li ⁺ . The assembled LiIn|S full cells with PEO/sulfide composite SPE exhibit high capacity retention of 93.2% after 100 cycles^[190]. Furthermore, unlike metallic Li, the high-Li content alloy anodes such as Li₁₃In₃ and Li₁₇Sn₄ also exhibit excellent chemical and electrochemical stability in contact with the sulfide-based SSE (Li₆PS₅Cl)^[191]. Both the Li₁₃In₃|Li₆PS₅Cl|Li₁₃In₃ and Li₁₇In₄|Li₆PS₅Cl|Li₁₇In₄ cells exhibit stable long-term Li cycling (1,000 h) at 1 mA·cm^-2. However, the relatively large volumetric changes [such as LiSi (320%) and LiSn (260%) alloy anodes] in alloy materials during charge/discharge would cause material degradation.

Figure 15. (A) Schematic of the practical stability window of the Li₁₀GeP₂S₁₂ electrolyte and the chemical potential of different electrodes; (B) Galvanostatic Li plating/stripping profiles of the Li-LGPS-Li cell at 0.5 mA·cm^-2. Figure 15A and B is reprinted with permission from Ref. [184], Copyright © 2022 American Association for the Advancement of Science; (C) Schematics of the two fabricated electrodes; (D and E) TOF-SIMS analysis of 7Li⁺ in the modified electrode after charging to 2.7 V (D) and 3.7 V (E). Figure 15C-E is reprinted with permission from Ref. [185], Copyright © 2023 Wiley. LGPS: Li₁₀GeP₂S₁₂; TOF-SIMS: time-of-flight secondary ion mass spectrometry; HOMO: highest occupied molecular orbital; LUMO: lowest unoccupied molecular orbital; SE: solid electrolyte.

Consequently, Yan et al.^[192] synthesized a hard-carbon-stabilized LiSi alloy anode (LiSH46), in which the hard-carbon can effectively buffer the volume change of Si during the lithiation process. The thickness of anode only increases from 44.7 to 46.4 μm after activation at 0.1 C and further to 47.1 μm after ten cycles at 1 C. Besides, the formed Li₁₅Si₄ and LiC₆ phases provide 3D Li ⁺ /e^- conductive pathways for anode, accelerating electrode kinetics. As a result, the NCM811|LiSH46 solid-state cell delivers 61.5% reversible capacity after 5,000 cycles at 1 C due to the suppressed Li growth and reduced interfacial side reaction. Besides, to further promote the uniform alloying throughout the composite anode, Kim et al.^[185] and coworkers directly integrated the LPS glass SSE or Li₃N + LiF Li conductive components into Li-Si alloy to provide more lithium ions pathways for composite alloy anode [Figure 15C]. They found that lithium alloying (or plating) tends to occur at the interface between the electrolyte and alloy anode in the absence of implanted Li ion conductive components. In contrast, the introduction of Li conductive components results in uniform lithiation or deposition throughout the anode [Figure 15D and E]. Moreover, the mechanical properties of Li alloy will also affect the integrity of anode materials. Huang et al.^[193] compared several Li-based alloy systems (Li-Zn, Li-Si, Li-Sn, and Li-Al) and evaluated their capability to mitigate dendrite growth in ASSBs. They found that the Li-Al exhibits the highest capability of inhibiting dendrite intrusion among these alloys due to its higher Young’s modulus compared to the other alloys. Therefore, the use of alloy anodes indeed improves the interfacial stability with SPEs, however, several factors such as the volume change, Li diffusion rate and mechanical integrity require careful control to more effectively mitigate dendrite growth in alloy-based anodes.

Interfacial degradation mechanisms of polyether SPE at high-voltage cathodes

While the previous sections have primarily addressed the Li metal anode/polyether-based SPE interface, the stability of polyether electrolytes at high-voltage cathodes is equally critical for practical full-cell operation. Unlike the anode interface, where dendrite growth dominates, the cathode interface is primarily challenged by the oxidative decomposition of polyether-based electrolytes under high potentials.

Polyether-based polymers, such as PEO, PEG-derivates and P(DOL), are intrinsically susceptible to oxidation. Ether oxygen units can be oxidized at ~4-4.5 V versus Li/Li⁺, forming radicals or carbocations that generate resistive products, including lithium alkoxides, carbonates, and other oxygen-containing species, which hinder Li⁺ transport^[194,195]. This oxidative degradation is further exacerbated in contact with high-voltage layered cathodes, such as NCM811, LiNi_0.6Co_0.2Mn_0.2O₂ (NCM622), and LiCoO₂, which accelerate polymer decomposition and lower the practical oxidative onset below ~4 V versus Li/Li^+[42]. Factors contributing to degradation include the oxidative activity of partially delithiated Ni- or Co-rich cathodes, conductive carbon additives, high electrode surface area, and reactive species such as HTFSI generated during electrolyte oxidation, which can further propagate polymer chain degradation. Collectively, these reactions form a resistive cathode-electrolyte interphase, causing voltage polarization and capacity decay, which underscores the dominant role of interfacial chemistry and reaction kinetics in determining electrolyte stability.

Building on this understanding, researchers have developed the following strategies to mitigate interfacial instability at high-voltage cathodes, which can be broadly categorized into cathode surface coatings and polyether electrolyte/interphase engineering.

Strategies for stabilizing the polyether SPE/high-voltage cathode interface

(1) High-voltage cathode surface coatings

Surface coatings on high-voltage cathodes (e.g., NCM811, NCM622, LiCoO₂) function as protective barriers that prevent direct contact between the polyether-based SPE and the active cathode material, thereby mitigating oxidative degradation and interfacial side reactions while preserving Li⁺ transport pathways. For example, Dai et al.^[196] applied a lithium borate (Li₂O-B₂O₃) coating that effectively isolated the cathode from PEO, suppressing parasitic reactions and facilitating ionic transport. Similarly, Chen et al.^[197] introduced a nanoscale LATP coating onto NCM811, which increased the interfacial stability of the cathode up to 4.31 V vs. Li/Li⁺ and suppressed parasitic reactions with the PVDF/PEO-LLZTO composite electrolyte. Beyond inorganic coatings, Li et al.^[198] constructed a high-voltage, high-conductivity interfacial layer of polyfluoroalkyl acrylate (PTFEMA) directly on the NCM811 cathode via in situ solvent-free electropolymerization. This layer expanded the electrochemical oxidation window of PEO SPE from 4.3 to 5.1 V, resulting in enhanced cycling stability of NCM811|PEO|Li cells, with an initial discharge capacity of 179.8 mAh·g^-1 retained at 119.7 mAh·g^-1 after 50 cycles at 0.1 C.

(2) Polyether electrolyte and interphase engineering

Parallel efforts have focused on tailoring the polyether-based electrolyte itself to enhance interfacial stability with high-voltage cathodes, mainly through three strategies: (i) Additive engineering. Chen et al.^[199] incorporated a multifunctional additive (e.g., tris(pentafluorophenyl)borane) into PEO, which reduced polymer crystallinity, enhanced ionic conductivity, increased the Li⁺ transference number via anion trapping, and optimized SEI/ cathode-electrolyte interphase (CEI) formation, thereby enabling stable operation above 4.6 V. Similarly, Kim et al.^[200] introduced a functional sulfate additive into a thermally cross-linked PEO-based gel polymer electrolyte, forming a stable interfacial layer that effectively suppressed oxidative decomposition and side reactions on Ni-rich cathodes; (ii) In situ interphase construction. Ye et al.^[201] constructed a CEI on NCM811 via pre-reaction with trace liquid electrolyte, which isolated PEO from direct contact with the cathode and suppressed its catalytic decomposition; (iii) Double-layer polymer electrolyte design. The bilayer design proposed by Arrese-Igor et al.^[202], comprising a high-voltage-tolerant poly(propylene carbonate)(PPC) on the cathode side and PEO in the separator, not only ensures suitable energy-level alignment but also promotes stable chemical interactions with both the high-voltage cathode and the lithium-metal anode. This architecture enabled LiNi_xMn_yCo_zO₂|Li cells to deliver 160 mAh·g^-1 at 1 mAh·cm^-2, retaining over 80% of the initial capacity after 80 cycles with nearly 100% coulombic efficiency.

Collectively, these strategies, including targeted cathode coatings, additive engineering, in situ interface formation, and double-layer polymer design, constitute a coherent framework for stabilizing high-voltage cathode-polyether SPE interfaces, enhancing ionic transport, and improving cycling performance in solid-state lithium metal batteries.