Ionic conductivity is a key characteristic of the electrolyte, which quantifies the degree to which ions are mobile and available for the ongoing electrochemical reaction, and in part determines the power output of the battery^[50]. The Li⁺ is dissolved in the polymer and moves with the movement of the polymer chain segment, and the higher the ionic conductivity is, the better the battery performance is^[51]. To achieve an ionic conductivity (10 mS·cm^-1) comparable to that of an electrolyte at room temperature, the ionic conductivity can be improved in two ways: improving the ion mobility and increasing the concentration of free ions. Therefore, a good ionic conductor must have a small resistance to the movement of ions and promote the dissociation of ion pairs. Reducing the crystallinity of the polymer increases the ionic conductivity. Ionic conductivity is usually measured by electrochemical impedance spectroscopy.

Although the total current is generated by the movement of anions and cations, the useful current to drive redox reactions is usually generated by the movement of cations, which are lithium ions in lithium batteries. Lithium ion transference number refers to the ratio of the number of mobile lithium ions to the total number of mobile ions in the electrolyte. The larger the $t_{Li}^{+}$ is, the smaller the concentration polarization is, which is beneficial to inhibit the formation of lithium dendrite. The lithium ion transference number is usually obtained from the potentiostatic polarization test and calculation of lithium-lithium symmetrical batteries.

The electrochemical window is the difference between the oxidation and reduction potentials. In order to make the battery work stably, the electrolyte can not react with the cathode and anode, that is, it is inert to both electrodes, so the oxidation potential must be higher than the intercalation potential of Li⁺ in the cathode, and the reduction potential must be lower than the intercalation potential of anode lithium metal. The electrochemical stability of the high voltage cathode in the battery system should be 4. 3 ~ 5.0 V^[52]. The electrochemical window of the electrolyte can be achieved by constructing a lithium/polymer electrolyte/stainless steel sheet cell, followed by linear sweep voltammetry at a certain scan rate over a specific potential range.

The manufacturing process of the polymer film and the assembly process of the battery require the (quasi) solid electrolyte to have a certain degree of flexibility. At the same time, because lithium dendrites are produced during the operation of the lithium battery, the polymer film must have a certain strength to inhibit the growth of lithium dendrites, prevent short circuit within the battery, and reduce potential safety hazards^[53]. According to the tensile test and rheological test, the mechanical properties including strength, modulus and elongation at break can be measured.

Reliable thermal stability (> 150 ° C) can guarantee the stable operation of the battery under high temperature conditions, which is usually tested by thermogravimetric analysis, ranging from liquid nitrogen temperature to room temperature, and then to the set temperature above room temperature^[54]. Chemical stability requires that chemical reactions do not occur during battery operation, including within the electrolyte, with the electrodes, or with the current collector and packaging materials^[55].

In addition, the electrolyte and the electrode should have good interfacial compatibility to reduce the interfacial impedance. The electrolyte membrane should have an appropriate thickness. If the thickness is too small, it is difficult to resist the penetration of lithium dendrites, which will lead to a series of safety problems. If the thickness is too thick, the impedance of the electrolyte will be too large, which is not conducive to improving the energy density and power density of the battery.

Zero-dimensional nanomaterials usually refer to very small particles (nanoparticles) and clusters. Ceramic nanoparticles can destroy the local reorganization of the polymer chain, reduce the crystallinity, expose more amorphous regions, and improve the swing ability of the polymer chain segment to promote the conduction of lithium ions. Nanoparticles have larger specific surface area, which can provide more active sites, and high Young's modulus and hardness can improve the mechanical stability of polymer electrolytes, which is conducive to the suppression of lithium dendrites. Fig. 3 is a schematic diagram of a nanoparticle promoting ion conduction by enhancing the movement of a polymer segment.

Lv et al. Used methacryloxypropyltrimethoxysilane (KH570) to modify SiO₂, and then copolymerized PEO with the modified SiO₂ under UV conditions to reduce the crystallinity of PEO, so that it could achieve 3.37×10^-4S/cm conductivity at room temperature^[57]. Similarly, Li et al. Utilized the interaction between the — NCO of toluene diisocyanate (TDI) and the — OH of SiO₂ particles to prepare TDI-SiO₂ nanoparticles, which were then grafted onto PEO to form a cross-linked network, and the TDI-SiO₂ nanoparticles provided a more continuous Li-ion conduction path while enhancing the movement of polymer segments, broadening the electrochemical stability window (5.6 V vs Li/Li⁺)^[58]. Lee et al. Synthesized nano ZrO₂ particles in situ on a polyester substrate P(CL₈₀TMC₂₀). This in situ synthesis method reduced the agglomeration of nanoparticles and made the particles disperse uniformly in the polymer matrix^[59]. The ionic conductivity of the electrolyte at 30 ° C is 1.7×10^-5S/cm, which is higher than that of the electrolyte synthesized ex situ (5.6×10^-6S/cm). Guo et al. Doped poly (ethylene glycol) (PEG) with nano-sized TiO₂ particles and triethyl phosphate (TEP) to promote the rapid conduction of Li⁺ using the electric double layer on the surface of TiO₂^[30]. When the TiO₂ particles are dissolved, the released cations are attached to the Helmholtz layer, and the Li⁺ in the lithium salt is attracted to the diffusion layer to form unstable adsorption. At the same time, the interaction between the nanoparticles and Li⁺ promotes the movement ability of PEG segments, and the migration of Li⁺ is accelerated under the action of electric field and segment movement. This electric field effect can guide the uniform deposition of lithium ions and inhibit the formation of lithium dendrites, thus enhancing the interfacial compatibility of electrolyte/electrode. The cycle performance of the lithium-lithium symmetrical battery was tested at a constant current of 0.01 mA/cm², and the battery could cycle stably for 800 H without short circuit. Although scholars have found that the introduction of nanoparticles destroys the crystal structure of some polymer skeletons and increases their amorphous regions and free volume, the mechanism of nanoparticles promoting lithium ion conduction at the micro-molecular level has not been thoroughly studied.To fill this gap, Song et al. Thoroughly investigated the effect of oxygen vacancy-rich nanofillers on the transport mechanism of lithium ions at the molecular level^[60]. Taking nanoscale Al₂O₃ and nanoscale TiO₂ as examples, they revealed that the effect of nanofillers on improving the ionic conductivity of PEO-based electrolytes containing lithium salts is due to synergistic dissociation and trapping effects. The polar filler can interact with the ion pair of lithium salt to produce a local induced dipole, which weakens the Coulomb attraction within the ion pair and dissociates the ion pair to produce abundant anions and lithium ions. The oxygen vacancy of the filler preferentially interacts with the anions to restrict the movement of the anions and increase the transference number of lithium ions. Although fillers without oxygen vacancies can also improve ionic conductivity by dissociating ion pairs, the lithium ion transference number is not high due to the inability to trap anions. The relationship between oxygen holes and ionic conductivity is also applicable to other systems, which provides a general method for the design of inorganic-polymer composite electrolytes. Table 1 summarizes the main electrochemical properties and the corresponding battery performance of the solid state electrolyte based on inert nanoparticle fillers.

Due to its high surface energy and easy agglomeration, nano-ceramic particles are separated from the polymer matrix, and the agglomerated particles will block the ionic conduction channel and reduce the ionic conductivity. In addition, the poor interfacial compatibility due to the weak interaction between inorganic ceramics and organic polymers further reduces the dispersion of ceramic particles in the polymer matrix. For this reason, researchers have proposed methods to modify the structure of ceramic particles, such as coating ceramic particles in polymers to form inorganic-organic core-shell structure, so as to improve the stability of fillers in polymer matrix.

In the early stage, Lee et al. Treated SiO₂ with vinyltrimethoxysilane (VTMS), and the surface of the modified SiO₂ contained carbon-carbon double bonds, and then carried out radical polymerization with sodium 4-toluenesulfonate to obtain a material with a core layer of SiO₂ and a shell layer of poly (sodium 4-styrenesulfonate), and then obtained the shell layer of poly (lithium 4-styrenesulfonate) by ion exchange method, and finally obtained a core-shell structure SiO₂(Li⁺) (Fig. 4)^[61]. They doped core-shell SiO₂(Li⁺) into PVDF-HFP substrate to prepare composite electrolyte. However, the particle size of this SiO₂(Li⁺) is relatively large (~ 2.0 μm), which limits its application in solid-state electrolytes. Therefore, they optimized the size of the core layer and shell layer on the basis of the above SiO₂(Li⁺)), and prepared a SiO₂ core layer with an average diameter of 200 nm and a shell layer with a thickness of 320 nm, which significantly improved the physicochemical and electrochemical properties of the electrolyte^[62,63]. Gao et al. Used a similar method to prepare filler SiO₂-PAA@Li with core-shell structure, and then blended the filler with PVDF-HFP, EMITSI, and LiTFSI to prepare gel polymer (ILGPE-SiO₂-PAA@Li)^[64]. It is found that ILGPE-SiO₂-PAA@Li has good adhesion and interfacial compatibility with anode lithium, which can provide interfacial transport channels for lithium ions and inhibit the formation of lithium dendrites. When the filler content is 15 wt%, there is almost no shrinkage deformation at more than 200 ℃, showing excellent heat resistance. The initial discharge capacity of the LiFePO₄/ILGPE-15%-SiO₂-PAA@Li/Li cell is as high as 138 mAh/G at 0. 05 C, and the capacity retention rate is 87% (120. 1 mAh/G) after 100 cycles of stable operation. Similarly, Khan et al. Found that when the outer surface of SiO₂ was coated with a layer of PMMA and the content was 10 wt%, the PVDF-HFP-based electrolyte had the highest ionic conductivity of 2.22 mS/cm, while the unmodified SiO₂ content reached the maximum ionic conductivity (1.33 mS/cm) at 4 wt%^[65].

One-dimensional nanomaterials refer to materials in which the size of one of the three dimensions is not between 0.1 and 100 nm, including nanowires, nanotubes, nanorods, nanoribbons, etc. Compared with isolated nanoparticles, one-dimensional nanostructures have larger specific surface area, larger aspect ratio and more exposed active sites, which can provide a longer distance continuous ion conduction pathway.Creating an effective percolating network, a higher surface-to-volume ratio is beneficial to reduce the crystallinity of the polymer and enhance the polymer segment movement, thereby improving the ionic conductivity. Fig. 5 is a schematic diagram of the ion conduction pathway constructed by the nanorods.

As early as 2016, Liu et al. Reported a composite electrolyte of PAN modified by Y₂O₃ doped ZrO₂ nanowires, and explored the effect of doping different concentrations of Y₂O₃ in ZrO₂ nanowires on ionic conductivity^[66]. With the in-depth understanding of one-dimensional inorganic fillers, the research of one-dimensional structure in electrolytes has become more and more extensive. Ao et al. Used a variety of characterization methods to compare CeO₂ nanoparticles and CeO₂ nanowires, and they found that the network structure formed by interwoven CeO₂ nanowires can be used as a fast conduction channel for Li⁺.PAN doped with 10 wt%CeO₂ nanowires at 60 ° C can achieve a high ionic conductivity of 1.1×10^-3S/cm, while the ionic conductivity of PAN doped with CeO₂ particles under the same conditions is only 5.9×10^-4S/cm^[67]. The plating/stripping test was carried out at a current of 0.25 mA/cm², and the results showed that the stable and low hysteresis potential was maintained within 2000 H, indicating that the CeO₂ nanowires could effectively suppress the lithium dendrite. In order to obtain more point defects and enhance the cooperative dissociation and trapping effect, Chen et al. Doped low-cost Ca in CeO₂ and prepared Ca-CeO₂ nanotubes by electrospinning technology, whose hollow structure increased the contact area with PEO matrix.Similarly, Gao et al. Reported a composite electrolyte es-PVDF-PEO-GDC formed by gadolinium-doped ceria (Gd-doped CeO₂,GDC) ceramic nanowires and PVDF-PEO composite nanofiber membrane,The doping of gadolinium increases the oxygen vacancy of CeO₂, and the ionic conductivity of es-PVDF-PEO-GDC at 30 ℃ can reach 2.3×10^-4S/cm, the lithium ion transfer number is 0. 64, the tensile strength is 10. 8 MPa, and the average coulombic efficiency is 99. 2% after 250 cycles at 0. 5 C^[68][64]. Li et al. Prepared Ti³⁺ doped TiO₂ nanowires through the bridging effect of TDI, which was applied to PEO-based electrolyte as a modifier (Fig. 6)^[69]. The oxygen vacancy in the TiO₂ nanowire can stabilize the free ion, promote the dissociation of lithium salt by Lewis acid-base interaction with the anion, and provide a continuous and long-range ionic conduction path. When the doping amount of TDI-TiO₂ is 8%, the ionic conductivity is 1×10^-4S/cm at 30 ℃, and the lithium ion transfer number is 0.36. The cross-linked network formed by the TDI-TiO₂ nanowires and the polymer substrate helps to improve the electrochemical stability and increase the electrochemical stability window (5.5 V vs Li/Li⁺). Table 2 summarizes the main electrochemical properties of the inert 1D nanowire-based solid-state electrolyte and the corresponding battery performance.

Two-dimensional materials are ultrathin materials with high anisotropy and chemical functionality, which have extremely large specific surface area and can form effective interfacial interaction with polymer matrix. The performance improvement of polymer electrolyte is highly related to the structure of the filler. Generally, the layered structure with large specific surface area, high porosity and polar materials are beneficial to the full contact or interaction between the filler and the polymer matrix and lithium salt, and improve the stability and dispersion of the filler in the polymer system. Two-dimensional materials such as MOFs, COFs, graphene and boron nitride usually have abundant functional groups and specific crystal structures. The interaction between some functional groups and lithium salts and polymers can promote the dissociation of lithium salts, inhibit the movement of anions and improve the transference number of lithium ions. The unique size selectivity of periodic porous materials also contributes to the transport of lithium ions. The larger size surface of the two-dimensional packing can form a richer active interface with the polymer matrix and construct a fast ion transport channel. Taking graphene oxide as an example, it is shown that the utilization of the functional groups of two-dimensional inert fillers and the construction of larger inert filler-polymer active interfaces can enhance ion transport, as shown in Figure 7.

Two-dimensional layered boron nitride (BN) is a honeycomb-like material composed of nitrogen atoms and boron atoms, which has a crystal structure similar to graphene. Due to the inherent physical and chemical properties of BN, it can improve the performance of solid polymer electrolytes when used as an additive. The heat resistance of boron nitride can alleviate the heating problem during the working process of the battery, the excellent mechanical stability is beneficial to inhibiting the growth of lithium dendrite and improving the safety performance of the battery, and the electronic insulation and electrochemical stability of boron nitride mean that the boron nitride can remain stable and does not participate in the battery reaction during the long-term working of the battery^[71]. When applied to quasi-solid electrolyte, the high porosity of boron nitride is beneficial to the absorption of electrolyte, and the interlayer or external channels can be used as the conduction path of lithium ions to improve the ionic conductivity. Recent studies have shown that boron nitride can adsorb polysulfides and lithium ions, which is beneficial to improve the electrochemical performance of solid-state lithium-sulfur batteries^[72]. This paper mainly focuses on the effect of BN nanosheets as an additive on the performance of the electrolyte.

Initially, Li et al. Used a template-free heat treatment process to synthesize porous g-BN nanosheets, which were used as the main body of the electrolyte to absorb ionic liquids and prepare quasi-liquid solid-state electrolytes.Its ionic conductivity at room temperature and low temperature (− 20 ° C) is 3.85×10^-3S/cm and 2.32×10^-4S/cm, respectively, which is the earliest work using boron nitride nanosheets as the electrolyte framework, and also provides an idea for the development of solid-state electrolytes for two-dimensional layered nanomaterials^[73]. Subsequently, Shim et al. First used boron nitride as an additive in solid-state electrolytes for lithium batteries, and found that it could significantly improve the performance of polymer electrolytes, especially effectively inhibit the growth of lithium dendrites^[74]. They prepared perfluoropolyether modified boron nitride nano sheets (BNNS) by ultrasound-assisted exfoliation and non-covalent functionalization, then added them into PVDF-co-HFP matrix to prepare porous composite membrane, and soaked the porous membrane in electrolyte (1 mol/L LiTFSI in EC: DEC (1:1 vol%)) to prepare composite gel electrolyte (Fig. 8). The functionalized boron nitride reduces the crystallinity of PVDF-co-HFP, and the resulting porous structure facilitates the absorption of solvent, thereby improving the ionic conductivity. Even with only 0.5 wt% doping, the prepared composite electrolyte can achieve a high Li-ion transfer number ( t_{\mathrm{L}\mathrm{i}}^{+} = 0.62), a high Young's modulus (110 GPa), and a high tensile strength (53 MPa), and the Li-Li symmetric battery can be stably cycled for more than 1940 H at a constant current of 0.1 mA/cm² and maintain a low voltage hysteresis, which indicates that boron nitride has a significant advantage in suppressing lithium dendrite. However, due to the weak chemical interaction between boron nitride and polymer matrix, the compatibility and dispersion are poor, resulting in the increase of thermal resistance at the interface, which can not give full play to the thermal properties of boron nitride. For this reason, Zhang et al. Used SiO₂ to modify the surface of hexagonal boron nitride nanosheets to prepare PEO/LiTFSI/SiO₂@BNNS composite electrolyte, which improved the electrical conductivity and thermal properties of boron nitride by introducing defects (SiO₂)^[75]. The SiO₂@BNNS composite can form a thermal channel in the polymer system to enhance the thermal stability of the electrolyte. When the temperature rises to 150 ℃, the composite membrane has almost no thermal shrinkage and the shape remains unchanged, while the Celgard 2300 membrane and PEO/LiTFSI membrane have obvious deformation. Table 3 summarizes the main electrochemical properties and the corresponding battery performance of the 2D boron nitride nanosheet-based solid-state electrolyte.

Materials with three-dimensional structure refer to composite materials composed of one or more basic units in zero, one or two dimensions, usually bulk materials. The mechanical support of the three-dimensional framework is better, which can ensure better lithium dendrite inhibition, and its connectivity and permeability are also better than those of the two-dimensional lamella, which can conduct lithium ions in more dimensions^[76]. Taking the SiO₂-PEO continuous nanonetwork constructed in Fig. 9 as an example, the three-dimensional network structure not only provides abundant ion transport channels and improves the ionic conductivity, but also effectively enhances the mechanical properties of the electrolyte membrane and improves the dimensional stability.

The main components of glass fiber are SiO₂, CaO, Al₂O₃, etc., and these oxides play an important role in improving the mechanical strength and ionic conductivity of the electrolyte. Zhang et al. filled PEO, LiTFSI and ionic liquid into a 3D glass fiber cloth (GFC) framework to prepare a composite electrolyte (Fig. 10). The 3D rigid framework gives the electrolyte good mechanical support, and its polar groups (— OH, Si — O bonds) can guide the uniform deposition of lithium.The Li/PEO @ GFC/Li symmetrical cell was cycled stably for more than 1000 H at a current of 0.2 mA/cm², and the plating/stripping voltage was stable at about 0. 21 V, while the Li/PEO/Li cell could only cycle for 400 H under the same conditions, and the voltage fluctuated greatly, which would cause internal short circuit due to the growth of lithium dendrites^[78]. The glass fiber skeleton across the PEO matrix destroyed the crystal structure of the polymer and exposed more amorphous regions, which facilitated the transport of lithium ions, and the polar groups reacting with lithium salt anions enhanced the dissociation of lithium salts, and the PEO @ GFC had an ionic conductivity of 1.1×10^-5S/cm and a lithium ion transference number of 0.30 at 30 ° C, which were superior to the electrical properties of PEO electrolyte under the same conditions (7.6×10^-6S/cm, t_{\mathrm{L}\mathrm{i}}^{+} = 0.24). Similarly, Wang et al. Replaced the ionic liquid with the plasticizer succinonitrile (SN). Although the addition of the plasticizer can greatly improve the ionic conductivity, it reduces the mechanical properties of the electrolyte. The 3D glass fiber framework can compensate for the mechanical loss to some extent^[79]. When doped with 25 wt% succinonitrile and 10 wt% glass fiber, the ionic conductivity of PEO-SN₂₅-LiTFSI₁₀-GF can reach 2.84×10^-4S/cm at room temperature, which is two orders of magnitude higher than that of pure PEO. The electrochemical stability window is 5. 5 V, which shows excellent mechanical stability and electrochemical stability. The assembled LiFePO₄/Li battery is tested at 0. 2 C, the initial discharge capacity is 150 mAh/G, and the capacity almost does not decay after 100 cycles, which is 147. 8 mAh/G, achieving a high capacity retention rate of 98. 5%.

In addition to the three-dimensional glass fiber skeleton, the researchers also explored the effect of non-lithium inorganic fillers with microsphere structure on the electrochemical performance of batteries when applied to electrolytes. Hu et al. First proposed that the polymer electrolyte (PEO-LiTFSI-0.2C₃N₄,20 wt%C₃N₄) with tightly packed polygraphitic carbon nitride (g-C₃N₄) microspheres was applied to the conversion-type FeF₃ cathode, and the Li/FeF₃ battery has a high theoretical energy density (850 Wh/kg) and no solid-liquid phase conversion of lithium polysulfide, which is considered to be a battery system with the possibility of achieving high-energy all-solid-state^[80]. The g-C₃N₄ nanosheets are self-assembled to form a highly regular microsphere structure, the microspheres are tightly stacked to form a firm porous framework structure, and the PEO molecular chain can penetrate into the pores to enable the g-C₃N₄ to be deeply crosslinked with the polymerized lithium salt, thereby enhancing the mechanical strength of the composite electrolyte.The high Young's modulus (285 MPa) and lithium dendrite inhibition function were achieved, and the electrochemical stability window was improved to 5. 12 V. Plating/stripping tests of lithium-lithium symmetric cells were carried out at currents of 0. 1 and 0.2 mA/cm², and the stable cycling time without short circuit was as long as 10 400 and 5000 H, respectively. The binding energy between g-C₃N₄ and lithium salt is larger, which is beneficial to the adsorption of lithium salt anions, increases the lithium ion transference number ( t_{\mathrm{L}\mathrm{i}}^{+} = 0.69) and reduces the potential decomposition of lithium salt anions, and enhances the stability of the passivation layer on the anode side. The microsphere filler has better interfacial compatibility and can realize high capacity retention rate and cycle reversibility when being applied to a Li/FeF₃ battery.