Towards separator safety of lithium-ion batteries: a review

Abstract

The safety problem of lithium-ion batteries (LIBs) has restricted their further large-scale application, especially in electrical vehicles. As a key component of LIBs, separators are commonly used as an inert component to provide a migration path for the ions and prevent direct contact of the cathodes with the anodes. However, the separator damage caused by external impact puncture, local overheating, and internal and external short circuit leads to the thermal runaway and explosion accidents of LIBs. Therefore, how to enhance separator safety has been a challenge in the development of high-performance LIBs. In this review, the recent advance of high-safety separators with high mechanical strength, high thermal stability and good lithium dendritic resistance is the main focus. Various factors affecting the separator's safety are discussed, including the species of the polymer substrate, structure, synthesis and modification processes. More importantly, the designed strategies towards high safety separator are comprehensively addressed in terms of various aspects: (i) to composite ceramic materials or organic polymers to improve thermal stability, (ii) to modify the polymer and composite structure to increase mechanical strength, (iii) to design a functional separator to regulate ion transport and lithium deposition for mitigating lithium dendrites, and (iv) to intelligently control the thermal runaway of LIBs by flame retardant and thermal shutdown function. Furthermore, the component–structure–performance relationship of separators is summarized, and the impact of separator compositions and structures on the safety of LIBs is emphasized. In addition, the future challenges and perspectives of separators are provided for building high safety rechargeable lithium batteries.

---

Boli Tong

Boli Tong is currently a PhD candidate in the Institute of Advanced Electrochemical Energy at Xi’an University of Technology. He received his Master's degree from the Zhongyuan University of Technology in 2018. His research interests primarily focus on the design and synthesis of high-safety lithium-ion battery separators.

Xifei Li

Xifei Li is currently a professor at Xi’an University of Technology. He was awarded the 2018–2022 Highly Cited Researchers of Clarivate Analytics award. He is a vice-president at the International Academy of Electrochemical Energy Science and a Fellow of Royal Society of Chemistry. Dr Li's research group is currently working on safety strategies for high performance rechargeable batteries. Dr Li has authored/co-authored 358 peer-reviewed articles with more than 19 000 citations with an H index of 74.

---

1. Introduction

Mechanical damage, thermal abuse and electrical abuse have been triggering many thermal runaway accidents in lithium-ion batteries (LIBs).^1,2 These safety problems hinder the large-scale application of high-power LIBs.^3–5 The separator, which isolates the cathode and anode, plays an important role in safety accidents, preventing the stored chemical energy from converting to heat in the battery. However, separator failures caused by abuse conditions lead to battery internal short circuits (ISC).⁶ In particular, commercialized polyolefin separators are inflammable and thermally unstable. These characteristics increase the chance of the battery short circuit under local overheating conditions, which leads to a serious risk to the battery.^7,8 Hence, enhancing the separator safety for reliable LIBs is of great necessity.

The different failure mechanisms of separators in safety accidents are as follows: (i) when LIBs suffer uncontrollable thermal conditions caused by a local micro short circuit, subsequent chain-like chemical reactions are likely. Firstly, a high local temperature (∼100 °C) triggers unstable component decomposition in the solid–electrolyte interphase (SEI). When the temperature is up to 100–150 °C, the SEI decomposes and brings about drastic chemical reactions in batteries. This stage accompanies the decomposition of the electrolyte and further shrinking of the separators.^9,10 Finally, the softened, melted polymer separators generate massive heat by ISC, causing the LIBs to catch fire, emit smoke and explode. (ii) The separators break by collision and extrusion, which leads to direct contact between the anode and the cathode; this is the important mode of ISC caused by separator failure. Subsequently, chain-like uncontrollable chemical reactions caused by a large amount of internal currents occur.^11–13 (iii) Separators are pierced and cause ISC by lithium dendrites formed in a continuous cycle.^14,15 Especially, the overcharge state not only elevates the temperature inside the battery but also causes the destruction of the electrode crystal structure and the decomposition of the electrolyte; these conditions all affect the role of the separators and greatly influence battery safety.^12,16,17

According to the separator failure mechanism, modifying the separators and ensuring the safety of the LIBs under complex conditions are of great necessity.¹⁸ Firstly, the heat resistance temperature of the separator should be improved to prevent uncontrollable thermal conditions. Sufficient mechanical strength is needed to ensure the isolation of the cathode and the anode.⁴ Secondly, separators should regulate the uniform lithium deposition during the cycle to avoid further battery deterioration caused by the dendrite. In addition, separators must be chemically and electrochemically stable in LIBs.^14,19,20 They should be inert in the battery when fully discharged and charged.^21,22 In some applications, separators have to withstand the corrosive nature of the electrolyte at elevated temperatures. Recently, research to improve the battery safety from the perspective of separators focused on the simple modification of traditional polyolefin separators and the development of novel separators. Surface coating, grafting process and composite construct modified separators have been developed to improve safety. In addition, fibre separators possess good mechanical strength and safety through introducing the modified materials inside the fibre or on the fibre surface.^23,24 Despite considerable research on the modified separator, systematic reviews from the safety perspective are lacking.²⁵

This review focuses on the role of the separators in LIB safety accidents, the safety influencing factors and the failure mechanism of separators. More importantly, this paper comprehensively outlines the design strategies of high-safety separators from the aspects of improving thermal stability, mechanical strength, good lithium dendrite resistance, flame retardant and thermal shutdown functions, as shown in Fig. 1. The paper also summarizes the composition–structure–performance relationship of the separator, underlining the influence of the separator composition and structure on the safety of LIBs. This work also provides future challenges and directions for high-safety separators.

Fig. 1 Some strategies to improve separator safety.

2. Key factors of safe separator

2.1. Requirements for LIB separators

The physical structure and chemical characteristics of separators directly affect the ion transport, internal resistance, cycle performance and safety of LIBs.²⁶ The requirements of high-performance separators usually include thickness, pore size, porosity, mechanical strength, thermal stability, permeability (Gurley number), chemical/electrochemical stability, ionic conductivity, Mac-Mullin values and lithium ion migration number. In this review, we mainly focus on the key factors of safe separators, including safety requirements and failure process. So we focus on the requirement parameters of the separators’ safety, including the thickness, pore size, porosity, wettability, mechanical strength and thermal stability. Among them, the thickness, mechanical strength, pore size, porosity and thermal stability directly affect the separator's safety. Wettability affects the assembly process of LIBs and the transmission of lithium ion, which indirectly limits the battery safety. Although other parameters are important for the performance of separators, they have smaller effects on the safety of separators.

(1) Thickness. A thinner separator is conducive to lower electrochemical impedance and high lithium ion migration number.^27,28 Therefore, a thinner separator is ideal on the premise of ensuring battery safety. The progress in mechanical strength of thick separators sacrifices space for active materials and reduces the energy density.^29,30 The requirements for the separator thickness differ in various application scenarios: (a) the thickness of computer, communications and consumer electronic products and button batteries is 5–25 μm. (b) The thickness of separators used in energy storage power stations and the electric vehicles field is 25–40 μm.³¹ Due to the difference in the viscoelasticity of the polymer, the characterization methods affect the test results of the thickness. Hence, the appropriate methods must be selected to measure the thickness according to the variety of separator.^32,33

(2) Pore size. Separators with a small pore size limit the transmittance of Li⁺.^34,35 In contrast, separators with an excessively enlarged pore size can enhance the permeability of Li⁺, which is easily pierced by grown lithium dendrites and applied force, resulting in safety issues of ISC. The United States Advanced Battery Consortium requires the pore size of the lithium-ion separator to be less than 1 μm.^36–38 Pore sizes can be observed by scanning electron microscopy (SEM) or measured using a capillary flow pore size analyser or mercury porosimeters. Pore size can be obtained according to eqn (1):

d = 4γ cos θ/ΔP, (1)

where ΔP is the pressure, d is the pore diameter and flow, γ is the surface tension and θ is the contact angle.

(3) Porosity. Porosity is the ratio of the pore volume to the total volume of the separator. High porosity favours Li⁺ migration during charging and discharging, brings low impedance and high ionic conductivity but leads to serious thermal shrinkage and poor mechanical strength.^39,40 By contrast, low porosity increases the impedance and Li⁺ ion transport. The porosity of large-scale separators produced by the current manufacturers is mostly 25–45%.^41,42 Indeed, the porosity used for LIBs is about 40%.⁴³ Porosity (ε) can be obtained by using eqn (2):

E (%) = (W − W₀)/ρ_LV₀ × 100%, (2)

where W and W₀ are the separator masses before and after the electrolyte immersion, respectively, ρ_L is the liquid density and V₀ is the separator volume.

(4) Wettability. Electrolyte wettability could help achieve effective ion transmission and reduce internal resistance.^44,45 The quick absorption of electrolyte is crucial for the preparation and operation of LIBs.^46,47 Wettability can be judged by electrolyte uptake, liquid absorption height or contact angle.^48,49 Electrolyte uptake is calculated by eqn (3).

Electrolyte uptake = (m − m₀)/m₀ × 100%, (3)

where m₀ and m are the mass of the separators before and after soaking in electrolyte, respectively.

(5) Mechanical strength. Mechanical strength is determined jointly by material intrinsic properties and preparation.^50,51 The indexes of mechanical strength for separators include tensile strength and puncture strength. On the one hand, the separator must have a certain tensile strength to overcome the damage caused by the physical impact, wear and compression in the production, assembly, disassembly and actual use processes. On the other hand, during repeated charging and discharging, the volume expansion of the electrode compresses the separator, and the dendrite also punctures the separator, thus affecting battery safety. Hence, a high puncture strength was necessary to resist the destruction of the lithium dendrite. According to ASTM D882 or ASTM D638, the minimum tensile strength of separators with a thickness of 25 μm is 98.06 MPa.^52–54 Mechanical strength is highly correlated with the thickness, porosity and pore size of separators.

(6) Thermal stability. Thermal shrinkage reflects the volume change of a material owing to its inherent thermal expansion rate. The thermal shrinkages of the separators in the transverse and longitudinal directions cannot exceed 5% when the assembled LIBs undergo heat treatment at 90 °C for 1 h. Thermal shrinkage can be calculated from eqn (4):

Shrinkage = (S₀ − S)/S × 100%, (4)

where S₀ is the original separator area, and S is the separator area after heating.

2.2. Factors and failure process for separator safety

Mechanical damage, thermal abuse and electrical abuse are the main factors of separator and battery safety.¹⁸ In addition to the abuse condition, the safety failure of the separator is affected by various external factors, as shown in Fig. 2a. Such as over charge and over discharge,^55–57 external short circuit, aging and defects.^58,59 These conditions lead to separator failure and subsequent ISC.^60,61 ISC is the predominant accident cause and is difficult to control.^62–64 When a large ionic current passes through the battery, the battery is inclined to produce a large amount of heat and triggers uncontrollable chemical reactions.⁶⁵ Furthermore, accumulative heat causes decomposition of the cathode materials and results in the release of gaseous oxygen.^66,67 Oxygen and heat accumulation leads to violent oxidation of the electrolyte and the separator, which are highly exothermic reactions, and eventually causes thermal runaway, further fires and even explosions (Fig. 2b).^68–70

Fig. 2 (a) Thermal runaway of LIBs with temperature change, (b) abuse conditions related to thermal runaway, (c) three levels of internal short circuit (reproduced with permission,⁴ copyright 2020, Elsevier), and (d) temperature–time profile of thermal runaway (reproduced with permission,³¹ copyright 2022, Wiley).

ISC is an intermediate process between abuse conditions and thermal runaway, which occurs in more than 90% of abuse conditions. ISC is caused by the direct contact between the positive electrode and the negative electrode due to the failure of the battery separators, which may be caused by the following reasons: (1) separators rupture due to penetration or extrusion, (2) separators are pierced by the growing dendrites and (3) separators shrink and collapse at high temperature.

Feng et al. divided ISC into three levels, as shown in Fig. 2c. At level I, the battery discharges slowly without evident abnormal heat and with self-extinguishing characteristics. At level II, voltage drop and temperature rise are substantially accelerated. At level III, the thermal runaway process cannot be stopped, and a large amount of heat is generated due to the broken separators.

The thermal runaway process follows the three stages through three characteristic temperatures of T₁ (loss of thermal stability), T₂ (runaway trigger temperature) and T₃ (maximum temperature) according to the thermal runaway mechanism of the battery system, as shown in Fig. 2d. A separator possesses a different failure process in these three stages. T₁ stage is the beginning of the battery self-heating and ISC.

The battery begins to lose thermal stability at this temperature. Initial self-heating may be caused by the growth of lithium dendrite, separator rupture, electrode deformation and other problems. These conditions could cause the battery to change from a safety state to an abnormal state of ISC, and the battery system overheats. T₂ is the trigger temperature at which the battery starts the uncontrollable reaction to cause thermal runaway. At this stage, the dT·dt⁻¹ changes by order of magnitude. LiF and Li₂CO₃ anode SEI swell and decompose when temperatures rise to 70–120 °C. The exposed anode continuously reacts with the electrolyte. The newly formed SEI layer is in a metastable state, and its exothermic decomposition releases flammable gases. At temperatures above 130 °C, the common commercial separators shrink, and direct contact between the cathode and anode aggravates the situation and then leads to excessive heat build-up. Therefore, the exothermic heat of the cathode, adhesive and electrolyte peaks in the second stage, which breaks down and releases oxygen. T₃ is the highest temperature in the thermal runaway process, and max{dT·dt⁻¹} is the maximum heat release rate and is positively correlated with the battery energy density. The T₃ stage is the combustion and explosion process. Oxygen and heat release in the second phase causes the combustion of combustible organic electrolytes, resulting in a fire or explosion hazard. Increasing T₁ and T₂ and reducing T₃ and max{dT·dt⁻¹} can reduce the risk of thermal runaway.⁷¹

According to the above analysis, thermal runaway is affected by the state of charge, operating conditions, electrode materials, electrolyte and separators. Thermal runaway can be mitigated at different stages in the runaway process by different methods. These measures can be divided into three genres according to the different processes. The first type is preventive measures to improve the safety from the structure and material design, including the introduction of flame retardant. The second method is to prevent or reduce the damage caused by thermal runaway, which includes shutdown separator and battery exhaust. This approach is a passive defensive measure to prevent further runaway and deterioration. The third category involves measures to extinguish the fires in the thermal runaway event, including flame-retardant separators and new flame-retardant electrolytes.

3. Four strategies to improve separator safety

3.1. High thermal stability

Various modified separators have been developed to enhanced thermal stability and safety. These separators include three categories: inorganic ceramic modified separator, organic layer modified separator and novel heat-resistant substrate separators. Organic and inorganic modified separators are usually prepared by coating or grafting, such as silica (SiO₂), alumina (Al₂O₃) and polyvinylidene fluoride (PVDF), and prepared by electrospinning, casting and paper processes.⁷² Novel heat-resistant separators are prepared based on new processes and new structures, such as electro-spinning and phase separation, aramid and polyacrylonitrile (PAN), which has an inherent high heat resistance.⁷³ Amongst them, the inorganic modified separator is favoured by LIBs due to its excellent thermal stability and high safety at elevated temperature; heat-resistant separators with novel materials are the development direction in the future.

3.1.1. Surface-modified separator with ceramic layer. Coating the heat resistance ceramic particles on the polyethylene separator surface is a simple, effective method to improve heat resistance. Inorganic materials can improve thermal stability and surface performance. Al₂O₃, SiO₂ and TiO₂ are commonly used as ceramic coating materials. For example, SiO₂ nanotube layers are used to coat commercial PE separators (Fig. 3a).⁷⁴ The coating separators exhibit a ‘zero’ shrinkage rate at 120 °C for 0.5 h. The assembled battery also maintains a stable voltage at high temperatures (Fig. 3b and c), which is attributed to the synergistic effect of thermally stable SiO₂ and antideformation wrapped 1D SiO₂ tube. Furthermore, a PDA–SiO₂ composite-modified PE separator is proposed. In Fig. 3d,⁷⁵ PDA forms an overall-cover self-supporting film on the PE separator pore wall, ceramic particles and surface PDA coated on the separator surface (Fig. 3e and f), which amends the film-forming properties of the bare PE separator. The bare PE separator undergoes shrinkage of 5.0% at 110 °C, whereas the PE-SiO₂ separator shrinks by 5.0% at 150 °C. More importantly, the PE-SiO₂@PDA separator has no visual shrinkage at 230 °C. In the OCV test at 170 °C (Fig. 3g), the OCV of the LIBs with the PE separator rapidly drops sharply to 0 V (after 75 s). The LIBs with PE-SiO₂ separator shows an OCV sharp drop at 487 s. In contrast, the LIBs with the PE-SiO₂@PDA separator maintains a stable OCV up to 2 h at 170 °C. A disassembled battery proves the decrease of voltage is due to the shrinkage of the separators (Fig. 3h). Wang et al. prepared silica dispersions coated on polypropylene (PP) separators with an impregnation method.⁷⁶ Because the coated ceramic particles and adhesive PVA has excellent heat resistance, the thermal shrinkage rate of the original separators is 53% at 170 °C, whereas the shrinkage of the silica coated separators is only 27%. The silica and PVA inhibit the deformation of microporous separators. Silica nanoparticles were also combined with polyvinylidene fluoride hexafluoropropylene (PVDF-HFP) adhesive to prepare a composite separator.⁷⁷ Highly porous honeycomb structures with a layer thickness of only 700 nm are formed. Composite separators can remarkably reduce thermal shrinkage. Over a wide temperature range, the coated separators exhibit enhanced thermal shrinkage resistance, and the shrinkage of silica coated PP is only 40% at 170 °C. However, larger SiO₂ particles are the main issue limiting their further application.

Fig. 3 (a) Cross-section SEM images of silica tubes; (b) thermal shrinkage of different separators; (c) OCV curves of LIBs assembled with different separators (reproduced with permission,⁷⁴ copyright 2015, Elsevier); (d) schematic illustration of the prepared PE-SiO₂@PDA separator; (e) photographic image of the bare PE separator (left), PE-SiO₂ separator (middle) and PE-SiO₂@PDA separator (right); (f) SEM images of the cross-section of PE-SiO₂@PDA separator; (g) OCV measurement of LIBs with different separators at 170 °C (the inset is the enlarged profiles); (h) optical image of the separators after the OCV measurement (reproduced with permission,⁷⁵ copyright 2016, Royal Society of Chemistry); (i) surface SEM image for the coated layer; (j) thermal shrinkage of the separator at different temperatures (reproduced with permission,⁷⁹ copyright 2014, Elsevier).

Except SiO₂ particles, silica sol is also a modified ceramic material that exhibits thermal stability. Fu et al. prepared silica by directly hydrolysing tetraethyl orthosilicate (TEOS) and depositing it on the separators.⁷⁸ The presence of heat-resistant SiO₂ substantially reduces thermal shrinkage, improves the tensile strength, surface wettability and electrolyte absorption. Especially suitable for high-power LIBs, the increased mechanical strength and reduced thermal shrinkage are beneficial for LIB safety. The mechanism of two layers of alumina/silica composite ceramics on the thermal stability of separators was also investigated.⁷⁹ Silica possesses a low density, high surface hydroxyl content and a tendency to form chains, forming a 3D network through hydrogen bonding. Alumina is uniformly distributed within it, as shown in Fig. 3i and j. The porous network is conducive to the absorption of liquid electrolytes. The hydrophilic, heat-resistant ceramic improves the wettability and reduces the thermal shrinkage of modified separators. The original separators have a high thermal shrinkage at 150 °C and a maximum shrinkage of 60% at 170 °C. Composite separators reduce the thermal shrinkage of 30% at 170 °C.

The low thermal stability of binders used to prepare the ceramic slurry is another factor limiting ceramic modified separators. The commonly used polymer adhesives for preparing ceramic slurries include PVDF and its composite materials. However, PVDF-based polymer binder possesses poor thermal stability. Therefore, the thermal stability of coated separators must be improved by developing other adhesives with high thermal stability. An alumina-coated PP separator was developed using phenolphthalein polyether ketone (PEK-C) binder, which has good heat resistance and can withstand high temperatures of approximately 230 °C (Fig. 4a).⁸⁰ The alumina coating adheres tightly to the PP separator at high temperatures using PEK-C adhesive (Fig. 4b), which enhances the integrity of the melt. At 150 °C, it only shows 8% dimensional shrinkage (Fig. 4c). The assembled battery ensures a steady state at 150 °C (Fig. 4d). Deng et al. prepared composite separators by coating Al₂O₃ on PE separators using PVDF-HFP and carboxymethyl cellulose double adhesive (Fig. 4e).⁸¹ The prepared Al PHC/PE separator possesses ‘zero’ shrinkage) at 110 °C (Fig. 4f). Based on the variance in electrical conductivity caused by the density difference of crystalline polyolefin and their amorphous counterparts, impedance spectroscopy analysis is beneficial for analysing the thermal structural changes of the separators. In Fig. 4g, the impedance of the PE separators rapidly increases at 128 °C, and the PE membrane softens and agglomerates at this temperature. A peak occurs at 136 °C, and with the increase of temperature, the PE undergoes severe thermal shrinkage and further leads to ISC. Due to the intrinsic properties of the base PP separator, the impedance of the Al PHC/PE separator also increases at 128 °C. The impedance no longer changes from 130 °C to 160 °C. When the temperature exceeds 160 °C, the battery has a short circuit, which is attributed to the melting of PVDF (the melting point is 152 °C). The nail penetration tests are shown Fig. 4h–j. The Al-PHC/PE-2 LIBs have no expansion or rupture.

Fig. 4 (a) Molecular structure of PEK-C, (b) SEM image of the composite separator, (c) thermal shrinkage properties at elevated temperatures, and (d) OCV curve at 150 °C (reproduced with permission,⁸⁰ copyright 2015, Elsevier). (e) Cross-section image of the Al-PHC/PE-2 separator, (f) thermal shrinkage properties of the modified separator at 110 °C for 1 h, (g) impedance change curve with temperature, (h)–(j) digital images of LIBs after nail penetration tests, and (k)–(m) voltage–temperature profiles (reproduced with permission,⁸¹ copyright 2016, Elsevier).

The PE/LIBs and C-PE/LIBs expand and cause a fire due to the formation of gas (Fig. 4h). Fig. 4k–m show the relationship between voltage and battery surface temperature. The voltage and heat of the different batteries in the puncture tests vary greatly. After one minute of puncture, the voltage of PE LIBs and C-PE LIBs drops sharply to about zero, and the temperature suddenly increases to 430 °C and 360 °C. The voltage of Al-PHC/PE-2 LIBs decreases slightly and remains stable. This outcome proves ceramic Al₂O₃ is conducive to maintaining separator stability and preventing a short circuit. Colloidal PVDF-HFP could repair the wound caused by nail penetration.

Ceramic metal hydroxides with high decomposition temperature and fire resistance have been used to modify the polyolefin separators to improve their thermal stability. Mg(OH)₂ and Al(OH)₃ are usually combined with PVDF-HFP binder.⁸² The inherent thermal stability of ceramic materials results in a relatively small thermal shrinkage rate for separators at 140 °C. The thermal shrinkage of Al(OH)₃ coated separators is 6.6%, and that of Mg(OH)₂ coated separators is 16.6%, whereas the thermal shrinkage rate of the original PE separators is 27%. More importantly, the Al(OH)₃ and Mg(OH)₂ can generate water at high temperatures, and the modified separators can exhibit self-extinguished characteristics in the case of fire. Jung et al. prepared highly thermally stable non-aqueous ceramic coated membranes (NA-CCSs) based on Al₂O₃ and Mg (OH)₂ (Fig. 5a). The NA-CCS separator possesses zero thermal shrinkage at 200 °C, as shown in Fig. 5b and c.⁸³ The thermal stability of the separator was tested under the action of electrolytes and mechanical stress as a standard for determining whether the separator is thermally stable in actual batteries. The closing behaviour of an electrolyte immersion separator sandwiched between conductive metal foils was tested. The membrane was continuously heated under electrolyte immersion, as shown in Fig. 5d. When the PE membrane is heated, a sudden increase in impedance indicates the pores begin to close at 130 °C. Compared with PE, the pores of CCS begin to close at 135 °C, which is related to the cross-linking interface between the PE and the ceramic coating. After heating, CCS forms a PE ceramic composite layer and ceramic coating, which can maintain its structural stability. This outcome indicates that even in the presence of electrolytes, CCS can withstand thermal and mechanical stresses.

Fig. 5 (a) Images of PE, Al₂O₃ and Mg(OH)₂ coated separators, (b) thermal shrinkage images at 200 °C for 10 min, (c) SEM images of these separators, and (d) impedance–temperature curves (reproduced with permission,⁸³ copyright 2019, Elsevier). (e) Schematic diagram of the cross-section of the reactive plasma coating on a PE separator, (f) cross-sectional SEM images of modified separators, and (g) camera images after heat treatment at 120 °C for 1 h (reproduced with permission,⁸⁴ copyright 2018, Wiley).

Non binder coatings have also been developed to improve the thermal stability issues caused by low-melting-point polymer binders. In Fig. 5e and f, Qin et al. used a reactive atmospheric plasma coil to coil method to coat SiO_xC_yH_z mixed nanoparticle films on PE separators.⁸⁴ After 120 °C for 1 h, the thermal shrinkage rate of SiO_xC_yH_z coated PE separators is 4%. The SiO_xC_yH_z nanoparticle provides good structural support for the separators, as shown in Fig. 5g. Yuan et al. dipped phenolic compounds based on biologically stimulated surface modification.⁸⁵ A silicon layer was formed on the separators through the sol–gel process, and the inorganic organic mixed layer was coated on the PP without polymer binder. In addition, this method hardly increases the thickness of the original separator and does not sacrifice the microporous structure (Fig. 6a and b). The separators obtained by introducing a mixed layer exhibit excellent dimensional thermal stability because the thermal shrinkage at 150 °C is only 20%, whereas that of the original separator is about 80%. The electrochemical performance of LIBs was improved with modified separators (Fig. 6c). At a charging/discharging current density of 5C, the PP LIBs have no discharge capacity, and the modified separator possesses a capacity of 45.7% at 0.2C.

Fig. 6 (a) Schematic illustration of the inorganic–organic layer modified separator preparation, (b) SEM images of the SiO₂-PP separators, and (c) thermal shrinkage–temperature curves for heat treatment for 1 h (reproduced with permission,⁸⁵ copyright 2018, Wiley). (d) Fabrication of BNNT-modified separator, (e) thermal infrared images of hot point, PP and BNNT separator, (f) and (g) cycling life at 50 °C and 70 °C at 1C, (h) thermal shrinkage and combustion behaviour from 25 °C to 150 °C (reproduced with permission,⁸⁷ copyright 2019, Wiley). (i) SEM image of the PLHS separator; (j) photographs of PP, PS and PLHS separators before and after 150 °C for 1 h; (k) OCV at 150 °C. (Reproduced with permission,⁸⁸ copyright 2021, Elsevier.)

Solid fast ionic conductors are also used to boost the heat resistance of the separator.⁸⁶ Rahman used boron nitride nanotubes (BNNTs) to prevent short circuits (Fig. 6d).⁸⁷ By simply adding appropriately designed thin BNNTs without blocking the separator porous structures, the heat resistance is up to 150 °C. When a hot point is generated on the separator through a thermistor, the peak temperature at the PP separator hot spot is about 70 °C, whereas the peak temperature at the BNNT separator drops to 61.5 °C (double-sided coating), as shown in Fig. 6e, indicating the thermal diffusion of the BNNT coating is enhanced due to the excellent thermal conductivity of the nanowires. These results indicate the BNNT separator (double-sided coating) has good heat dissipation efficiency. During cycling, the high-speed capacity of BNNT membrane batteries is also remarkably improved due to the absorption of additional heat and diffusion through BNNT, as shown in Fig. 6f–h. BNNT shows an exciting new type of nanomaterial that prevents battery short circuits and verifies the practical application scenarios of high-safety separators. A mixed separator with a solid fast ionic conductor modulus dielectric aromatic amine nanofiber Li_6.75La₃Zr_1.75Ta_0.25O₁₂ (LLZTO) was prepared (Fig. 6i).⁸⁸ At 150 °C for 1 h, the PP membrane substantially shrinks, and some edges of the PS separators are curved; the PLHS separators remain stable, as the assembled battery also has a stable voltage at 150 °C for a long time, indicating PLHS has excellent thermal stability (Fig. 6j and k). Importantly, it enhances the ionic conductivity of the separators.

The ceramic coating has been widely used in various battery systems, as an effective heat-resistant modification process, which plays an obvious role in the safety separators. But its defects are still inevitable. A high density of ceramic layers reduces the energy density of the battery, although this has been noted and reduces the coating thickness as much as possible. The surface coating only improves the apparent electrolyte wettability of the separators, and the shuttling process of electrolyte ions on the based separators is still not different from the uncoated separators. Furthermore, polymers with a low melting point are the most commonly used binders in this process, which are harmful to battery safety. Therefore, modified materials with high thermal stability are still an important research direction.

3.1.2. Organic layer modified separators. Widely developed surface modification of inorganic nanoparticles can improve the thermal stability of separators and other properties, but some problems remain, for example, uneven thickness, blockage of separator pores, interface difficulties and side reactions caused by stiffness and brittleness. In addition, the binder is prone to melting or decomposition when the temperatures exceed the melting point, leading to the detachment of the coating materials. The new modification methods of atomic layer deposition (ALD) and chemical vapor deposition can hold the base separator structure and avoid the detachment of coated materials, but the cost is relatively high.⁸⁹ Organic layer surface modification separators with heat-resistant structures is a promising modification method for improving thermal stability. The organic modified layer has its unique preparation, usually including surface in situ grafting and adhesive-assisted coating technology. By grafting the different monomers, polymer films can be endowed with specific electrochemical and physicochemical properties, making them meet the technical requirements of different types of batteries for separators. This method enables the grafted material to grow on the polyolefin separators directly without increasing the thickness. Pre modification can be achieved through electron beam γ high energy irradiation such as radiation, ultraviolet (UV), and plasma. High-energy irradiation promotes the formation of cross-linked polymers in polyolefin membranes, thereby improving their structure and thermal stability. Wei et al. prepared a polymer to form a solid polymer electrolyte separator by one-step UV polymerization with the matrix consisting of lithium bis(trifluoromethane) sulfonimide (LiTFSI) and poly(ethylene glycol) methyl ether methacrylate (PEGDMA).⁹⁰ The polymer separator remains stable up to 317.1 °C with a weight loss of 5%.

The strong chemical bond between the separator and graft structure guarantees adequate mechanical support even at elevated temperatures. By grafting polymerization and condensation reactions, Zhao et al. grafted the organic/inorganic hybrid network on the separators.⁹¹ The silicon–oxygen cross-linking network was constructed and reduced the thermal contraction of the separators. At 150 °C, the original and graft-modified separators contract in the mechanical direction by 38.6% and 4.6%, respectively (Fig. 7a). At the same time, the grafted organic–inorganic hybrid cross-linking network only occupies a part of the pore of the separator, and the thickness of the grafted modified separator is equal to that of the original separator, which is favourable for electrochemical performance. Due to the strong chemical bond between the grafting material and a polyolefin matrix, the grafted separator usually has a long working life. However, commonly used high-energy irradiation has potential safety risks and increased preparation costs. He et al. coated the polyethylene terephthalamide (PPTA) on a commercial PP separator.⁹² In the absence of any additional binder, the PPTA nanofiber adheres to the PP separator through physical anchoring, conferring the composite separators with heat resistance and improved wettability. In Fig. 7b, no thermal contraction remains for 1 h at 300 °C. In addition, in the 200 °C heat treatment, the network structure is provided by heat-resistant PPTA, which can still maintain the separator mechanical structure after PP melting, realizing the shutdown effect of the composite separator and helping improve battery safety.

Fig. 7 (a) Schematic diagram and thermal shrinkage of modified separators and commercial separators (reproduced with permission,⁹¹ copyright 2015, Elsevier), (b) heat resistance of the Ps and PPTA@PPs separators at different temperatures (reproduced with permission,⁹² copyright 2015, Elsevier); (c) diagrammatic sketch of free-liquid surface electrospinning, (d) SEM image for PET/PP-22% after shutdown, (e) photography of the separator at different temperatures (reproduced with permission,⁹³ copyright 2015, MDPI); (f) SEM image of CS1; (g) thermal shrinkage of the PP separator, CS1, CS2, CS3 and CS4 at various temperatures (reproduced with permission,⁹⁴ copyright 2017, Elsevier).

Polyethylene terephthalate (PET) has excellent mechanical strength, thermodynamic properties and electrical insulation properties. Cai et al. used electrospinning technology to prepare a composite separator with PET fibre and commercial PP separators (Fig. 7c and d).⁹³ The PET fibres improve the wettability of the electrolyte, whilst the separator structure can be maintained after the thermal contraction of the PP, which helps improve the thermal safety of LIBs. At 180 °C for 30 min, the PP separator changes from white (33%) to translucent, whereas the composite separator does not change much (Fig. 7e). The PP separator acts as a base membrane to melt and bind with the PET fibres, transforming the porous ionic conductive polymer film into a non-porous insulating layer. The closing of the hole prevents ISC and improves battery safety.

Dipping processes have also been widely developed to improve the defects of the grafting process and coating methods for the surface-modified separator of organic layers. Zhang et al. prepared a silica/polyvinyl alcohol coated PP separator using the immersion coating method.⁹⁴ The prepared separator has a unique porous structure with layered pores combined with interstitial pores for optimal wettability and thermal stability. Fig. 7f and g show the thermal stability results of 0.5 h at different temperatures from 130 °C to 170 °C. At 170 °C, CS4 shrinks by only 8.3% over the original size, whereas PP shrinks by 46.9%. The severe thermal contraction of the PP separator stems from its formation of pores during stretching. With the increase of TEOS dosage, the thermal shrinkage rate of the composite material is slightly reduced because the large amount of TEOS produces more heat-resistant nano silica particles.

Except polyvinyl alcohol, thermally stable polydopamine (PDA) was coated on the PP separators.⁹⁵ The thin inorganic–organic hybrid layer was then fixed to the PDA coated separators using a tetraethoxysilane (TEOS) solution via a sol–gel process (Fig. 8a and c). The PDA midlayer has a unique adhesion behaviour, which greatly maintains the original structure of the separators. The shrinkage of the unmodified PP separator is 100%. When the COI reaches 16.4%, the longitudinal shrinkage rate is reduced to 40%. PP-PDA heat contracts by 7% at 120 °C for 4 h compared with 17% for PP. At 165 °C for 1 h, PP is completely melted, and PP-PDA shrinks by 80%, whereas the mixed layer coated PP can still maintain 55% (Fig. 8b). The further improvement in thermal shrinkage is attributed to the presence of a hybrid layer. The introduction of a PDA layer and mixed layer remarkably improves thermal stability.

Fig. 8 (a) Preparation of hybrid layer coated PP separator; (b) thermal shrinkage-COI curves at 165 °C for 1 h; (c) SEM images of PP, PP-PDA and hybrid layer coated PP separator with COI: 11.9% (reproduced with permission,⁹⁵ copyright 2014, Royal Society of Chemistry); (d) SEM images of PP without pyrogallic acid treatment PP for 24 h, (e) rate performance of different separators (reproduced with permission,⁹⁶ copyright 2014, Royal Society of Chemistry); (f) schematic illustration of the preparation of the PE/Al₂O₃/PDA separator; (g) thermal shrinkage of PE, PE/Al₂O₃, PE/PDA and PE/Al₂O₃/PDA at 140 °C for 30 min (reproduced with permission,⁹⁷ copyright 2019, Elsevier); (h) chemical structure of P84, (i) SEM image of the cross-sectional view of the 3 wt% P84-coated PE separator and digital camera images of the bare PE separators and PE separators coated with different concentrations of P84 before and after heat treatment at 140 °C for 30 min (Reproduced with permission,⁹⁸ copyright 2012, Elsevier).

To avoid the cost problem of PDA, Wang et al. reported a cheap, simple coating of hydrophilic surface-modified PP separator by replacing PDA with the low-cost coating precursor pyripionic acid (PA).⁹⁶ PA was formed spontaneously on the PP surface in a single double tris buffer aqueous solution. More importantly, PA-coated PP maintains the microporous structure (Fig. 8d). Furthermore, the thickness of the PA-PP-24 (22 ± 0.8 μm) varies only slightly from the thickness of the original PP (22 ± 0.2 μm). The porosity and Gurley values are almost unchanged for all PA-coated PP separators compared with the original PP. This finding ensures a high thermal stability whilst remaining preserved due to the separator properties. The PA-PP-24 has the best rate performance (Fig. 8e).

Another modified PE separator was prepared by Al₂O₃ and PDA coating using ALD, as shown in Fig. 8f and g. ALD technology can produce ultra-thin conformal coatings with precise thickness control, so the total thickness of the obtained PE/Al₂O₃/PDA separator does not increase compared with the original base separator.⁹⁷ Except when combining with thermal stable inorganic materials as a hybrid coating, PDA and PI can reduce the aggregation and pore blockage caused by coating materials, further improving the melt integrity and thermal separator structure stability. For example, Song et al. synthesized a polyimide (P84) by dip coating, and the polyimide was connected to toluene diisocyanate and methylene diphenyl diisocyanate.⁹⁸ The structure is shown in Fig. 8h. This polymer coating with an optimal concentration allows the PE separator to reduce thermal contraction at higher temperatures. After 0.5 h at 130 °C, the naked PE separator substantially contracts by 83.3%. The P84 coated separator with different coating ratios shows higher thermal stability. The shrinkage of the 1%, 3% and 5% coated separators are 40.0%, 10.0% and 13.3%, respectively (Fig. 8i). Shin et al. modified the PE separator with a thin hybrid coating of aluminium fluoride (AlF₃) ceramic particles and poly(3,4-ethylene dioxythiophene)-polyethylene glycol (PEDOT-co-PEG) copolymers.⁹⁹ The synergistic interaction of highly thermal stable inorganic AlF₃ particles with heat-stable PEDOT-co-PEG copolymers considerably enhances the thermal stability of the modified PP separators, and the thermal shrinkage is almost ‘zero’ after 0.5 h at 130 °C.

3.1.3. Novel separator with high thermal stability. Although surface coatings and grafting modification contribute to improving the thermal stability of the separator, the inherent low melting point of polyolefin still causes severe softening and shrinkage of the separator in a harsh hot environment, which still limits the performance of the LIBs at high temperatures. Therefore, new efforts have been committed to developing new separators based on high-heat-resistant polymers. The melting point is a decisive factor in the high-temperature resistance of polymer separators. Developing a high heat separator with an inherent high melting point is the most reliable solution to address the safety issues of LIBs. The parameters of several novel heat-resistant membrane substrates are shown in Table 1.

Table 1 Parameters of novel thermos stable separator substrate

Substrate materials	Chemical formula	Melting points (°C)	TS (MPa)	Ref.
PAN		85–317	—	100
PET		250–260	78–255	101
PI		>350	231	102
PEEK		>400	109	103
GF	SiO2, Na2O and MgO	650	—	104
Cellulose		>200	—	105
Aramid	—	400	340	106

---

Widely used high-heat-resistant polymers include PVDF copolymer (melting point 170 °C (PVDF)), PI (melting point >350 °C), PAN (melting point ∼300 °C), poly(ethylene terephthalate) (PET) (melting point >250 °C), poly(m-phenylene isophthalamide) (PMIA) (decomposition temperature ∼400 °C), PEEK (decomposition temperature ∼530 °C), poly(arylene ether ketone) (PAEK) (decomposition temperature ∼500 °C), polysulfonamide (PSA) (decomposition temperature ∼400 °C), PVA (melting point 275–280 °C), and cellulose-based polymers (melting point >250 °C). According to the different characteristics of these materials, technologies including electrospinning, reverse spinning and phase separation have been developed and studied to prepare them into LIBs separators. Thereinto, aramid possesses high-temperature resistance, ultra-high strength and excellent acid and alkali resistance. The coefficient of thermal expansion of aramid fibre is about 2.4 × 10⁻⁶ K⁻¹, and the aromatic groups in the molecular structure substantially improve the structural stability at high temperatures. Polyimide (PI) with aromatic heterocycles is very stable, with outstanding thermal stability, mechanical properties and high-temperature mechanical property retention. PI has a melting point above 350 °C and is a self-extinguished polymer that does not ignite. However, the preparation of typical PI requires specialized equipment to control the changing high-temperature conditions. If the heating time, concentration and feeding ratio are not appropriately controlled, problems such as hardness, brittleness and difficulty in processing of the PI separator arise. Improving the production capacity of the PI separator is a widely applied effective strategy. PET has a melting point of 255–260 °C, which has better development in electric vehicle fields. Polyether ether ketone (PEEK) is a linear aromatic thermoplastic engineering plastic with ketone and ether bonds in the main molecular chain, which has high thermal stability. The heat-resistant temperature is approximately 50 °C higher than PI. Glass fibre (GF) partition is made of inorganic non-metallic materials, with high heat resistance of about 680 °C. GF can be used as a high-temperature battery separator in extreme applications.

Wang et al. designed a novel PE composite separator coated by high-temperature ZrO₂@PI core–shell nano-micro spheres.¹⁰⁷ On account of the ZrO₂ shell nano encapsulation, the prepared (ZrO₂@PI)/PE separator displays negligible shrinkage at 150 °C (Fig. 9a and b). The LIBs assembled with (ZrO₂@PI)/PE separator demonstrate enhanced thermal safety performance and can work continuously for 109 min at 140 °C (Fig. 9c), which is an enormous improvement compared with the LIBs assembled using a PE separator (i.e. 21 min). Xie et al. prepared a thin polyacrylonitrile/cellulose acetate (PAN/CA) composite separator. PAN and cellulose provide a heat-resistant structure (Fig. 9d).¹⁰⁸ When the separators undergo heat treatment for 150 °C, the PE separators collapse and disappear. In contrast, as shown in Fig. 9e, the PAN/CA and PAN separators have no heat shrinkage and edge curling, suggesting the the PAN and CA polymer matrix has excellent thermal stability to maintain dimensional stability.

Fig. 9 (a) Illustrative protocol for the preparation of the ZrO₂@PI)/PE separator; (b) thermal dimensional stability test of PE (7 μm), (ZrO₂@PI)/PE (11 μm) and Al₂O₃/PE (11 μm) separator (heated at different temperature for 30 min); (c) discharge curves at 140 °C and 0.1C (reproduced with permission,¹⁰⁷ copyright 2022, Elsevier); (d) schematics of preparation of PAN/CA separators, and (e) thermal shrinkage test of PE, PAN/CA and PAN separators at 130 °C and 150 °C (reproduced with permission,¹⁰⁸ copyright 2023, Elsevier).

Novel heat-resistant separators with an intrinsic high melting point are significantly superior to the commercial polyolefin separators, which have been considered as alternatives to polyolefin separators. Various processes have been developed for preparing the novel heat-resistant separators, such as, electrospinning, phase inversion and phase separation. Electrospun novel heat-resistant separators commonly possess high porosity, which benefits electrolyte permeability and lithium ion transportation. However, this feature may also lead to self-discharge and short circuit of LIBs. Meanwhile, electrospun separators always suffer from low mechanical strength, which restricts their applications in LIBs. In addition, phase inversion and separation usually involve high costs and toxic solvents as wells as complex fabrication processes. Therefore, polymers with intrinsic high melting points and optimized new processes will be the development directions of high heat resistant separators in the future.

3.2. High mechanical strength

On the one hand, the separator must have a certain physical strength to overcome the damage caused by the physical impact, puncture, wear and compression in production, and the assembly, disassembly and actual use processes. On the other hand, during repeated charging and discharging, the volume expansion of the electrode compresses the separator, and the dendrite also punctures the separator, thus affecting battery safety. The indexes of mechanical strength include tensile strength and puncture strength. Tensile strength reflects the dimensional stability of the separator when external forces are applied during use; a low tensile strength leads to a broken separator and further ISC of LIBs. Puncture strength tests the force of the separator when the electrode mixture penetrates the separator, which is used to assess the probability of a short circuit in the battery.

3.2.1. Modified polyolefin separators to enhance mechanical strength. Although the melting point is slightly lower than the dry process prepared PP separator, the current wet process polyethylene separator occupies a major market position. The dry PP separator has minute strength in the TD direction. The wet process ensures uniform mechanical strength in the MD (longitudinal) and TD (transverse) directions.¹¹¹ Therefore, the strength enhancement study of the polyolefin separator focuses on the dry separator.

Liu et al. explored the enhanced effect of Al₂O₃/SiO₂ composite ceramics for the mechanical properties of PP separators.⁷⁶ Tensile strength and puncture force are enhanced to 109.3 MPa and 2.9144 N compared with the original PP (106 MPa and 2.3619 N). The Al₂O₃/SiO₂ ceramic layer can effectively limit the movement of the separator, so the tensile strength of the modify separator is slightly improved. The tightly packed ceramic nanoparticles layer can resist the destruction of external forces and enhance the penetration force of the composite separator. Hao et al. found PDA nanoparticles have a strong adhesion capacity and developed surface nanoparticle modification methods of biomaterials to improve the electrochemical properties and mechanical property (yield stress and failure strain) of PP and PP/PE/PP separators (Fig. 10a).¹⁰⁹ At three strain rates (0.002, 0.02 and 0.2 s⁻¹), the yield stress, failure stress and failure strain of the PP separator increase from 17.48% to 100.11%, 13.45% to 82.71% and 4.08% to 303.13% along the TD, MD and DD, respectively, whereas the yield stress, failure stress and failure strain of the PP/PE/PP separator increase by 11.77% to 296.00%, 12.50% to 248.30% and 16.53% to 32.56%, respectively. The PDA improved method strengthens the fibres and does not block the pore (Fig. 10b and c), and this improved method does not affect the electrochemical properties. Fig. 10d and e show an apparent strain rate effect of the unmodified and PDA-modified separator, with yield stress, failure stress and failure strain increasing with increasing strain rate. MD has higher strength but less toughness compared with DD and TD. Fang coated the PP separator with PDA.⁹⁵ A thin inorganic–organic hybrid layer was then fixed to a PDA-coated separator by a tetraethoxysilane (TEOS) solution via a sol–gel process. The elongation ratio was greatly reduced after the plasma treatment.

Fig. 10 (a) Schematic illustration of the preparation the PDA-modified separator, (b) SEM images of PP and PDA-modified PP separators, (c) SEM images of PP/PE/PP and PDA-modified PP/PE/PP separators, and (d) and (e) mechanical properties of unmodified and PDA-modified separators: stress–strain curves along the transverse direction (TD), machine direction (MD) and diagonal direction (DD) (reproduced with permission,¹⁰⁹ copyright 2020, MDPI). (f) Tensile curves of the PP separator, hybrid layer coated PP (COI: 11.9%) and plasma treated PP separator (reproduced with permission,⁹⁵ copyright 2014, Royal Society of Chemistry); (g) transverse tensile strength of the separator, temperature voltage curves of the cells with (h) PE, (i) PGP and (j) PGMP separators during the impact test; the inset images show the digital images of the pouch cells after the impact test (reproduced with permission,¹¹⁰ copyright 2022, Wiley).

To overcome the deficiency of a lone commercial PE separator, Chen et al. developed a PE/GF-Mg(OH)₂/PE composite separator (PGMP).¹¹⁰ Compared with the PE separator, the PGMP separator shows a high mechanical strength (250 MPa), 2.5 times the lateral tensile strength of the PE separator (100 MPa). Even for a thickness of PE up to 64 μm, their MD (160 MPa) and TD (59 MPa) tensile strength is still lower than of the PGMP separator. Nail penetration, impact, overcharge and arc safety were tested for LiNi_0.5Co_0.2Mn_0.3O₂/graphene soft-pack batteries. After the impact test, the battery voltage and temperature of the PGMP separator remain unchanged; whilst the PE separator battery is severely short-circuited after the impact, the temperature rises to 500 °C within 3 s and finally burns out. The impacted PGMP separator remained intact without any crack. The superior mechanical strength results in a strong impact resistance in the test whilst maintaining the structural integrity and effectively avoiding ISC, thus greatly improving impact safety.

3.2.2. Modified fibre-based separators to enhance mechanical strength. Because of its high porosity, large specific surface area and adjustable structural characteristics, the fibrous membrane greatly compensates for the defects of polyolefin in electrolyte affinity and thermal stability. Its sources are extensive and highly processed, but the lower mechanical strength limits its development. Thus, the demand for the mechanical strength of the separator is reflected in a variety of fibre separators. The main reason for the poor mechanical properties of the fibre separators is that the nanofibers only loosely and randomly overlap without strong interactions. The methods developed to improve the strength of the fibre membrane include thermal pressing technology, composite fibre, adding adhesive and regulating the spinning process, including the strength enhancement and modification of basement membrane features to promote the macroscopic mechanical strength.

Xu et al. designed a unique composite structure of polyimide (PI) poly(m-phenylene isophthalamide) (PMIA) and a new high-performance electrospinning separator (Fig. 11a).¹¹² PI/PMIA and H-PI/PMIA have the best thermal stability (300 °C), tensile strength (24.1 MPa PI/PMIA and 34.3 MPa H-PI/PMIA) and electrochemical capacity retention (97.9% and 99.2%), as shown in Fig. 11b. Coaxial electrospinning and thermo-pressing techniques were developed to design a nuclear@shell structure of PAN/HCNFs@PVDF/UIO-66 (Fig. 11c and d).¹¹³ The PAN/HCNF core layer can support the electrospun membrane and help maintain an intact structural stability at high temperatures. Due to the synergetic effect of hot-pressing and the PAN/HCNF core layer, the tensile strength of fibre separators reach 24.77 MPa (Fig. 11e and f). HCNF enhances the polymer strength. Li et al. increased the macroscopic mechanical strength of PI nanofiber membranes by inhibiting the slip between the fibres (Fig. 11g).⁵³ The PI nano fibre membranes treated with lithium polyacrylate (PAALi) adhesive solution shows a strength of 16.1 MPa, higher than the 5.0 MPa of the original PI membrane (Fig. 11h). The introduction of this adhesive allows the initially loose, disordered PI nanofibers to cross-connect with one another.

Fig. 11 (a) SEM images of H-PI/PMIA, (b) stress–strain curves of different separators, (reproduced with permission,¹¹² copyright 2021, Wiley), (c) cross-section SEM images of PU-PC, (d) schematic of the preparation of electrospun composite fibre separators, (e) stress–strain curves of the fibre separators, (f) TEM images of PU-PC using red circles and yellow dash line (reproduced with permission,¹¹³ copyright 2020, Elsevier), (g) schematic of the fabrication strategy for crosslinked PI nanofiber membrane, (h) stress–strain curves (reproduced with permission,⁵³ copyright 2021, Elsevier), (i) sketch of HO-PPESK-FM, (j) stress–strain curves of electrospun R-PPESK-FM, O-PPESK-FM and HO-PPESKFM (reproduced with permission,¹¹⁴ copyright 2019, Springer), and (k) SEM image of PP/Cotton (reproduced with permission,¹¹⁵ copyright 2018, Elsevier).

Gong et al. developed electrospinning combined with thermal pressure and prepared new thermo stable PPESK materials.¹¹⁴ To obtain higher mechanical properties in the vertical and horizontal directions, the oriented PPESK membrane was covered vertically with two PPESK membranes and thermos pressed. The prepared thermal pressure composite membrane has a tensile strength of 22.8 MPa in the TD and MD directions (Fig. 11i and j). The heat-oriented PPESK membrane is heat size stable at 200 °C. PP fibres were also modified by grafting and coating.¹¹⁵ The PP/cotton fibre composite nonwoven fabric was prepared by using cotton and PP fibre (Fig. 11k). The modified nonwoven PP/cotton fibre composite shows good tensile strength (1.6529 kN m⁻¹) with a stable size at 170 °C. These modification processes effectively enhance the mechanical properties of the fibrous membrane.

3.3. Inhibiting lithium dendrites

3.3.1. Dendrite–separator interactions. Lithium dendrites cause potential ISCs, reduce battery life and are the main hazards of LIBs. The morphology of the separator and the stochasticity associated with the morphology of the polymer fibres can theoretically induce ionic current localization and dendrite growth. When the separator has poor infiltration of the electrolyte and uneven pore size distribution, the uneven lithium-ion flux causes the uneven distribution on the electrode surface, which leads to the uneven nucleation in the lithium deposition and the formation of lithium dendrite. As the number of battery cycles increases, the dendrite continues to grow and eventually punctures the separator to cause the ISC, causing serious safety failure. Based on the formation mechanism of the dendrite, the reasonable design of the separator can effectively solve the dendrite problem, and the separator can prevent dendrite growth by regulating the Li⁺ distribution without substantially increasing the battery weight. Attempts to arrest dendrites in the separator focused on the selective phase transition reactions or through the addition of an impermeable ceramic nonporous layer. Although the large elastic modulus helped inhibit dendrite formation, the underlying mechanism of the separator for inhibiting dendrite growth is still not fully understood.

Theories and models that predict the dendrite nucleation and growth process in lithium-based batteries include the seminal work from Chazalviel, who proposed the concept of dendrite incubation time and critical current in the context of dendrites growing in a dilute electrolyte. Monroe and Newman demonstrated dendrite growth as a function of the applied current density and incorporated the contribution of the dendrite tip radius. Ryan et al. demonstrated the power-law growth of dendrite morphologies, and delineated the flow of current and the variation of voltage along dendrite arms.

To further reveal the underlying mechanism of the separator on dendrites, Jana et al. evaluated the effect of the separator pore on the growth of lithium dendrites using the phase field method.¹¹⁶ A critical current density exists below which the penetration of the dendrite can be completely suppressed. Fig. 12a–d show the vectorial electric field is localized at the dendrite tip and scatters away from the separator fibres. The electric field is negligible inside the dendrite and has a high value inside the polymer phase of the separator. Simulations demonstrate the electric field concentrates at the tip of the dendrite and enhances the localized electro deposition rate. Furthermore, the dendrites are advantageous for the Joule-heated condensers, potentially leading to the separator melting, SEI decomposition and the electrolyte instability. For a very small aperture, the dendrites decompose into smaller metal fragments, called ‘dead lithium’. Simulation-based analytical models show the critical current density is a function of the pore channel inclination between the separator aperture and the polymer fibres (Fig. 12e). Four states of dendrite growth are determined: (i) inhibition where dendrite growth is thermodynamically unfavourable, (ii) penetration of the first layer outside the separator, (iii) the dendrite penetrates the first layer but is trapped in the oblique channel of the separator and (iv) the short circuit and the dendrite penetrates the entire width of the separator. These four dendrite growth mechanisms serve as a guideline to explore and select the optimal geometric, chemical and characteristic pore sizes. Shimizu et al. used three different pore sizes. The 3DOM separator in Li||Cu batteries provides uniform deposition of lithium and inhibition in dendrite formation (Fig. 12g–j).³⁷ This result is due to the uniform current distribution in the pores of 3DOM PI. With a minimum pore diameter of the 3DOM PI separator (300 nm), the lithium metal anode has the highest gas crack. This outcome is also related to the uniform distribution of the current in the cell. Using the 3DOM PI separator in the Li||Cu battery, the dendrite morphology of Li metal is suppressed during the discharging and charging cycles (Fig. 12k).

Fig. 12 Dendrite growth through the separator for a current density of I = 0.11 mA cm⁻², pore radius, a = 1.05 mm and layer interspacing, h = 0.7 mm (a) t = 0 s, (b) t = 3.44 h, (c) t = 6.88 h, and (d) t = 10.32 h; (e) electrodeposition fields for lithium growth subjected to a current density of I = 0.11 mA cm⁻²; dendrite morphology at t = 10.32 h, the corresponding electric field distribution, and local electrode position and electro dissolution rates (reproduced with permission,¹¹⁶ copyright 2015, Elsevier); (f) SEM images of surface and cross-section of 3DOM PI separators with the pore sizes of 500 nm, (g)–(i) voltage profiles of Li||Cu cells using 3DOM PI separators with pore sizes of (g) 300 nm, (h) 500 nm and (i) 800 nm; (j) coulombic efficiency of Li||Cu cells using 3DOM PI separators with the different pore sizes, (k) lithium dissolution/deposition test from Li||Li cells cycled at 10.3 mA cm⁻² using 3DOMPI with the pore sizes of 300 nm and PP separators (reproduced with permission,³⁷ copyright 2019, IOP Science).

Li et al. simulated the effect of the separator thickness and surface coating on the lithium dendrite morphology.¹¹⁷ Dendrites grow relatively uniformly and have a shorter average dendrite length than the thinner separator because lower ion concentrations are more uniform within the thicker separator (Fig. 13a and b). The coating can reduce the dendrite growth rate by extending their growth path. Klein evaluated the effects of PP membranes, PP fibres and ceramic coated separator on the high-pressure NCM523 graphite batteries.¹¹⁸ PP showed good cycling performance at 4.3 V. However, substantial battery failure occurred after 50 cycles when the cut-off voltage was up to 4.5 V (Fig. 13c and d). In contrast, the use of the fibre separator results in the formation of a remarkably more uniform, more planar distribution of Li metal dendrites, which correlates with more uniform TM deposition. The results show the ceramic coating for separator and/or active materials affects the battery performance not only through its mechanical properties but also through the reactivity of the electrolyte components, which in turn can lead to the formation of new components in situ affecting the formation of CEI and SEI and change in the long-term cycle. Therefore, detailed assessment of the separator properties (e.g. aperture distribution, tortuous porosity, air permeability, wetting behaviour and electrolyte absorption) provides a better understanding of how its properties achieve uniform TM deposition (Fig. 13e and f).

Fig. 13 (a) Phase-field simulation of Li deposition under separators with different thicknesses, (b) Li⁺ concentration under a single layer separator and coated separator (reproduced with permission,¹¹⁷ copyright 2022, MDPI); (c) cycling performance of NCM523||graphite cells at 2.8–4.3 V (N/P ratio: 1.15/1.00) and 2.8–4.5 V (N/P ratio: 1.35/1.00) using the PP membrane separator, (d) cycling performance of NCM523||graphite cells at 2.8–4.5 V (N/P ratio: 1.35/1.00) using PP fibre and PP membrane separators; (e) and (f) LA-ICP-MS analysis of the graphite anodes (ϕ15 mm discs) after 100 charge/discharge cycles in NCM523||graphite full-cells operated at 2.8–4.5 V using the PP membrane separator and the PP fibre separator (reproduced with permission,¹¹⁸ copyright 2022, Wiley).

In summary, and in spite of the great progress, a thermodynamic description that includes the effect of the separator on dendritic growth, or that attempts to rationalize the porous structure of emerging or existing separator structures remains unavailable. Therefore, a systematic theory on how the complex diaphragm morphology affects the dendrite formation process is still lacking.

3.3.2. Functional layer to regulate the ion transport for inhibiting dendrites. The interaction study of separator and dendrite shows uniform ion transmission contributes to the uniform distribution of ion concentration and current density, which can reduce incomplete lithium removal/lithium removal and improve the cycle life and battery safety. The PP separator has straight pores and low pore connectivity, which facilitates ion transport but does not balance the Li⁺ concentration gradient. In contrast, the PE separator has higher pore connectivity, which is more conducive to homogenization of the ion concentration gradient. The separator with a regular pore structure can provide a uniform Li⁺ transmission channel and facilitate uniform lithium deposition. The functional separator is constructed by regulating the pore structure, improving the affinity between anode and separator, and constructing the artificial SEI, which can effectively regulate ion transmission, promote uniform lithium deposition, and then prevent the dendrite growth. Taking the separator prepared by phase conversion method and electrospinning as an example, the separator structure can be optimized by regulating the precursor and experimental parameters to improve the ion transmission rate. The ion transport characteristics of the battery can also be effectively improved by introducing polar groups (carboxyl and amino) or coating hydrophilic materials (inorganic particles and polar polymer) on the separator surface.

Plasma was used to modify the commercial PP separator, which resulted in separator surface negative functionalization (Fig. 14a and b).¹¹⁹ The Coulomb interaction of plasma-modified separator surface and electrolyte cation (anion) can adjust the lithium ion flux and form a thin and uniform SEI layer in the anode to prevent lithium dendrite growth (Fig. 14c). Zhao developed a new ion-selective separator using the phase separation method to achieve anion fixation in a liquid electrolyte and achieve a dendrite-free lithium anode (Fig. 14d).¹²⁰ The number of t_Li⁺ increased to 0.81 (Fig. 14e). At a high current density of 3 mA cm⁻², a long-term stable 1000 h lithium deposition/stripping cycle is achieved (Fig. 14f and g). High-performance batteries pursue a thinner separator, but a thinner separator increases the risk of ISC caused by lithium dendrites. Zhou et al. grew metal–organic frameworks (MOFs) in PP separator (8 μm thick) channels, which aggregate electrolytes in MOF channels.¹²² The MOF modified separator (9 μm) greatly improves the cycle stability and dendrite resistance of the battery assembled with carbonate-based electrolyte. Furthermore, Yao et al. used the –NH₂ group in MOF to immobilize anions in the electrolyte, thereby facilitating cation transport (Fig. 14h). The number of t_Li⁺ in the PP-UIO-66-NH₂ separator increased from 0.45 to 0.68.¹²¹ The superior conductivity and large t_Li⁺ can not only achieve long-term reversible lithium plating/stripping and uniform lithium deposition but also further inhibit the lithium dendrite growth (Fig. 14i).

Fig. 14 (a) and (b) SEM and AFM images of oxygen plasma modified separators at 1 min, (c) mechanism of lithium dendrite growth and SEI layer formation in a battery cell structure with unmodified separator and plasma modified separator (reproduced with permission,¹¹⁹ copyright 2022, Elsevier); (d) cell voltage evolution with time for a Li||Li battery with CS separator at current densities from 0.25 mA cm⁻² to 3 mA cm⁻², over potential profiles of galvanostatic cycling of the Li||Li cell equipped with a CS separator and Celgard separator. The inset shows the cycling of the CS separator-based cell during 995–1000 h; (e) comparisons of the Li⁺ transference number (t_Li⁺) and separated conductivity (σ_Li⁺ and σ_anion) between the CS and Celgard separators; (f) in situ observation of Li dendrites growth on the surface of Li metal with CS/Celgard separators on top sides at a galvanostatic deposition of 1 mA cm⁻², (g) mechanism of lithium electro deposition during concentration polarization. The red and blue arrows represent the transport of Li⁺ ions and anions in the electrolyte, respectively (reproduced with permission,¹²⁰ copyright 2022, Elsevier); (h) working principle of PP, PU and PUN separators. SEM images of Li anodes cycled after 400 h of Li plating/stripping cycles for Li||Li symmetric cells prepared with PP, PU and PUN separators; (i) cycling stabilities of Li||Li symmetric cells with PP, PU and PUN separators (reproduced with permission,¹²¹ copyright 2022, Elsevier).

Only a high Young's modulus makes it difficult to hinder the lithium dendrite growth, and a high ionic conductivity is also important. Wang et al. prepared a poly(vinyl alcohol) composite separator (OPVA/NHNT separator) using nanostructured halogenated nanotubes (NHNTs) (Fig. 15a).¹²³ OPVA/NHNTs have a high Young's modulus and ionic conductivity, which can effectively delay the lithium dendrite growth (Fig. 15). In promoting the uniform distribution of lithium ions, the germination of lithium whiskers can be effectively inhibited. Liu et al. prepared a composite separator with electrospinning technology, zeolite imidolate frame (ZIF-8) and PAN,¹²⁴ with enhanced ionic conductivity (1.176 mS cm⁻¹), increased lithium ion transfer number (0.306), wider electrochemical stability window (5.04 V) and superior cycle stability (600 h voltage lag, over 30 mV), which can effectively regulate ion transport and inhibit dendrite growth (Fig. 15d).

Fig. 15 (a) Schematic image of lithium dendrites for batteries with a Celgard separator and OPVA/NHNT separator, (b) Li plating/stripping processes are conducted at current densities of 1.0 mA cm⁻² and 1.0 mA h cm⁻² (reproduced with permission,¹²³ copyright 2023, Elsevier); (c) SEM image of ZIF8-PAN composite separators and (d) cycling stability of the Li||Li symmetric cells assembled with PP and PAN separators (reproduced with permission,¹²⁴ copyright 2023, MDPI.)

Zhou et al. sprayed nano diamond particles on a commercial separator, which provides good electrolyte affinity for regulating Li⁺ distribution.¹²⁵ A high Young's modulus (more than 11 GPa) provided by diamond particles is beneficial to suppressing the Li dendrite growth. A good thermal diffusion ability provides a uniform thermal environment for Li⁺ deposition and effectively avoids uneven deposition at local high-temperature points (Fig. 16a). The synergistic effect realizes a high-performance battery with no dendrite lithium metal anode (Fig. 16b). Functional porous materials have also been used to regulate the Li⁺ transmission, and Chen et al. prepared a functional separator by simple vacuum filtration of porous graphene oxide onto an electrospun polyacrylonitrile (HGO-PAN) fibrous membrane (Fig. 16c).¹²⁶ The polar group of the stacked HGO can redistribute the lithium flow on the lithium metal surface (Fig. 16d and e), regulating the formation and lithium dendrite growth, achieving 800 h long-term reversible lithium plating/stripping. Liu et al. developed a three-layer separator to extend the cycle life of batteries.¹²⁷ In Fig. 16f, a layer of silica nanoparticle is clipped by two layers of polyolefin separator. When the lithium dendrites grow and penetrate the separator, they contact the silica nanoparticles in the sandwich layer (Fig. 16g). Silica reacts with Li through a solid-state conversion reaction, thus effectively etching dangerous Li dendrites and inhibiting the dendrite growth and improving the safety and extending the battery life.

Fig. 16 (a) Fabrication of the ND modified separator, a schematic illustration of lithium dendrite growth in different separators, (b) voltage–time profiles of Li||Li symmetric cells (reproduced with permission,¹²⁵ copyright 2021, Royal Society of Chemistry), (c) SEM image of HGO-PAN, (d) schematic illustration of Li⁺ flux distribution, (e) simulation study of Li⁺ distributions (reproduced with permission,¹²⁶ copyright 2022, Elsevier), (f) schematic illustration of SiO₂ sandwiched in the polyolefin separators, and (g) mechanism for extending the battery life (reproduced with permission,¹²⁷ copyright 2016, Wiley).

3.4. Overheat protection and flame-retardant properties

3.4.1. Thermal shutdown separators. The separator with thermal shutdown function can effectively close the pores, prevent the large ionic currents at a certain temperature and stop work in time to avoid thermal runaway. In the past decade, various separators with a closure function were developed from material synthesis to structural design, but most are single acting and irreversible, and the shutdown function is irreversible. The battery is damaged after the thermal shutdown. Moreover, the difference between the closing temperature and the melting temperature is low, and playing a role under the condition of rapid thermal runaway is difficult. Therefore, the study of a separator with reversible closing or other protective functions by using new functional materials is widely used.

As mentioned previously, the polyolefin separators are thermally unstable and flammable. Because the glass transition (T_g) and melting point (T_m) of PE (T_g = −68 °C, T_m = ∼135 °C) were lower than PP (T_g = −10 °C, T_m = ∼165 °C), the self-closing separator was developed with multilayer PP and PE, and this is the simplest, most mature flame-retardant composite separator. However, the melting point difference between PP and PE is small, which narrows the shutdown temperature range of the PP/PE/PP separator. When the temperature is above 160 °C, the PP layer loses the mechanical gap, causing an ISC. To adjust the self-closing temperature of the separator, other polymers can be employed.

Simply mixing two polymers with different melting temperatures is the easiest way to prepare the thermal shutdown separator. Li et al. prepared a shutdown separator by mixing PVDF, cyanoethyl cellulose (CEC) and cellulose nanofibers (CNF), as shown in Fig. 17a.¹²⁸ The PVDF–CNF–CEC separator has shutdown functionality. The melting point of PVDF is near 170 °C, and the superior thermal stability of composite separators is up to 280 °C due to the presence of CNF and CEC. The assembled LIBs exhibit good stability (Fig. 17b). Core–shell structures with different heat resistant temperatures are also used as thermal shutdown separators. In general, core materials provide thermal capacity and mechanical properties, whereas shell materials have thermal sensitivity and low melt points. Jiang et al. developed a new poly(lactic)@poly(succinate) (PLA@PBS) core–shell separator with a thermo sensitive temperature of 110 °C (Fig. 17c–e).¹²⁹ A heat-induced shutdown separator of PAN@PBS with thermal sensitivity and stability was prepared.¹³⁰ The PAN@PBS separator has a shutdown temperature of 110 °C and a wider closing temperature range from 110 °C to 250 °C.

Fig. 17 (a) Schematic illustration of the preparation of the PVDF-CNF-CEC separator, and (b) voltage–temperature profiles of LIBs with PVDF-CNF-CEC separators. The insets are photographs of the nail penetration tests (reproduced with permission,¹²⁸ copyright 2017, American Chemical Society), (c) schematic representation of coaxial fibre separators, (d) TEM image of the PLA@PBS separator, (e) photographs of the PLA@PBS and Celgard 2325 separator before/after heat treatment at 130 °C for 30 s and 170 °C for 15 min (reproduced with permission,¹²⁹ copyright 2017, Royal Society of Chemistry), (f) molecular structure, illustration for selective swelling of SFEG and its application as the LIB separator, surface and cross-sectional morphologies of SFEG membranes (reproduced with permission,¹³¹ copyright 2021, Wiley), (g) thermal shrinkage images of PE@PI (left) and PP (right) at 110 °C, 130 °C and 150 °C (reproduced with permission,¹³² copyright 2017, IOP Science), and (h) burring test (reproduced with permission,¹³³ copyright 2017, Elsevier).

Except for simple mixtures or septa of core–shell structures, new materials and processes have been developed to create monolayers with thermal shutdown functions. Yang et al. prepared a polysulfone-block-polyethylene glycol (PSF-b-PEG) membrane by selective swelling.¹³¹ In Fig. 17f, PSF with electrolyte affinity and PEG has lithium ion affinity, and the composite separator can close the pore at 125 °C, realize thermal closure function and improve battery safety. He et al. demonstrated a new method for making PPTA@PP by coating PPTA nanofibres onto PP esters without the need for any additional adhesive.⁹² When the temperature rises above 170 °C, the shape and intensity of PPTA@PP remain unchanged whilst the melted PP blocks the ionic path, in which the ionic conductivity drops sharply and cuts the circuit.

The creation of a bilayer separator with a closing function can also be realized by using synthetic two layer polymers. A new type of hot shutdown PE@PI composite separator was prepared (Fig. 17g).¹³² The PAA meta-membranes were prepared using an electrospinning method, and the resulting PI meta membrane was applied and spin-coated LDPE with application at high temperature as a thermal closing layer. The close temperature of PE@PI separators is 95–110 °C. The PI has an excellent flame-retardant performance.

An important advantage of the two layer PET/PP separator is that the bilayer structure can avoid battery thermal runaway.⁹³ The PP separator can be heat-closed at ∼169 °C in abnormal cases, whereas the PET provides the separator structural stability to avoid the shrinkage even at 248 °C, thus allowing the battery to operate safely. Li et al. prepared a hydroxyethyl-functionalized polyether ether ketone (OHPEEK).¹³³ Due to the intrinsic stability of the PEEK base separator and the enhanced molecular interactions introduced by the functional group, the OHPEEK separator undergoes ‘zero’ shrinkage at 160 °C (Fig. 17h). Compared with PE, the OHPEEK separator undergoes smaller shrinkage in the ignition test. When the ignition source is removed, the OH-PEEK fire quickly dies out spontaneously, whereas the Celgard 2400 separators continue to burn.

3.4.2. Flame-retardant separators. Although the use of a self-stop multilayer separator can prevent the migration of lithium ions at a certain temperature to cut off the redox reaction, temporary thermal fluctuations can still lead to permanent battery failure. Furthermore, most shutdown separators are based on flammable polyolefin. Therefore, the strategy of improving intrinsic thermal stability and flame retardant is adopted. Polymers containing Cl/Br can release halogen radicals or halides to remove free chain radicals that maintain the flame. Flame-retardant polymers contain aromatic heterocycles and nitrile groups, which are more stable than any other polymer due to the strong F–C bond. Generally, the flame retardant property of polymer separators was evaluated by limiting oxygen index (LOI) and burning experiment. LOI represents the volume concentration of oxygen in oxygen and nitrogen. The higher the oxygen index, the better the flame retardant performance. A burning experiment is another simple and intuitive way to characterize the flame retardant nature of the separators, and was used to intuitively show the flame retardant nature of the separators.

Non-combustible polymers can be prepared separately as a non-combustible separator, and they can also be used to modify the commercial separator. In the combined composite separator, one polymer provides good thermal and mechanical properties, and another is highly compatible with a liquid electrolyte. Various separators have been reported recently to enhance the flame resistance. PEEK has a high melting point of 330 °C and an excellent flame resistance. However, the poor solubility limits its application in LIBs. Modification of PEEK with side chain groups can reduce its crystallization and improve its solubility. Liu et al. used the thermal induced phase separation method to prepare PEEK separators with polyether sulfone and diphenylsulfone to solve the solubility issue.¹³⁴ Due to its inherent high LOI (37%), the separator exhibits excellent flame retardant performance.

The use of a multicomponent non-combustible polymer can effectively combine the advantages of each component. Sun et al. used the electrospinning technology to prepare the PI@PBI separators with PI as the core material and PBI as the reinforced sheath material. Composite separators possess ultra-high strength and excellent fire resistance (Fig. 18a).¹³⁵ Due to the excellent bonding properties of PBI the PI fibres are tightly wrapped by PBI. The combustion test results are shown in Fig. 18b, where the flame resistance of the PI@PBI-reinforced separator is remarkably better than that of a commercial separator. The LIBs assembled with the PI/PBI separator can work stably at a high temperature of 120 °C for 1 h.

Fig. 18 (a) Schematic illustration and cross-section SEM image of the core–shell fibre, (b) combustion tests of the separators (reproduced with permission,¹³⁵ copyright 2019, Elsevier), (c) preparation of the flame-retardant separator and flame-retardant performance (reproduced with permission,¹³⁶ copyright 2019, American Chemical Society).

Although separators combined with flame-retardant materials (such as PBI, PEEK and PI) can effectively improve the flame resistance, the production processes of the non-combustible separator are still complex. The mechanical strength and electrolyte wettability of many nanofiber-based polymer separators cannot satisfy the actual requirements. Furthermore, most flame resistance polymers are much more expensive. Therefore, other flame-retardant separators must be developed. Adding a flame retardant to a polymer separator is an effective method. Flame-retardant additives can evaporate or decompose, so they can slow down the combustion reaction. The release of an inert gas reduces the oxygen concentration. Flame-retardant additives must be immobilized or fully encapsulated by a polymer separator. For example, as shown in Fig. 18c, the modified ceramic coated ammonium polyphosphate (CCS@PFR) forms a heat-proof bifunctional separator (APP-CCS@PFR). The APP particles decompose over 300 °C into a dense ammonium polyphosphate (PPA) layer, which can serve as a barrier to isolate the oxygen released from the cathode.¹³⁶ Furthermore, the NH₃ and H₂O produced by APP decomposition can dilute the oxygen concentration. The dense PPA layer in turn turns the separator and electrolyte carbon into a non-combustible dense carbon residue, thus inhibiting the transfer of heat, combustible gas and oxygen.

4. Summary and perspectives

Battery safety is increasingly important and challenging, especially in high-energy/power density LIBs. Separator safety plays a key role in battery safety, but systematic research on high safety separators is scarce. The development of heat-resistant separators can help alleviate the serious consequences caused by thermal abuse in battery systems. The heat-resistant separator generally uses heat-resistant inorganic particles or organic materials to modify the polyolefin or fibre separator, thereby providing the ideal heat-resistant performance and ensuring battery safety. For mechanical strength, the main research direction is improving the mechanical strength of fibre separators. By optimizing the structure and manufacturing process, the poor mechanical properties of the fibrous membrane can be improved. In situ welding and grafting technology can integrate the binding forces of the fibrous membrane layer, effectively improving their mechanical strength. For safety issues caused by lithium dendrites, constructing fast functional layers to realize Li⁺ diffusion and uniform ion deposition are common methods. Alternatively, optimizing the pore structure of the separators and adjusting the electrolyte compatibility can help prevent dendrites. As a passive safety protection measure, an intelligent separator with thermal shutdown and flame-retardant functions can effectively block ion transfer and cut off the large ion current, suppress the transfer of heat, combustible gas and oxygen, and prevent further serious loss of battery control.

Although the above research has made remarkable progress in the safety of separators, this issue has not yet been fully resolved. The above four strategies will continue to be developed for separator safety. The modified heat-resistant layer that maintains the original separator pore structure and the high-melting point separators with low-cost and high-strength will replace the existing ceramic coated separators. Novel processes may be developed to improve the mechanical strength of the separators, especially the fibrous separators. Functional separators will play different roles for different battery systems. Finally, for thermal-shutdown and flame-retardant separators, expanding the gap between closure temperature and thermal shrinkage temperature is the key to thermal-shutdown function. Composite flame-retardant function separators may replace the single component flame retardant separators.

In addition, the current main challenge is that the prepared separators have difficulty in simultaneously achieving all expected parameters. According to the extant challenges and the review of previously reported research results, future research directions may include the following:

(i) The mechanism of separators to improve the thermal safety of the battery must be explored: the molecular structure and macroscopic structure of separators jointly affect their safety performance, but in-depth mechanisms of action are lacking. Various advanced characterization techniques should be used to explore the positive and negative effects of substances in the separator during the thermal runaway stage. A development system must be established for separator structure–properties–applications from micro to macro levels, which provide a valuable theoretical system for the development of novel separators.

(ii) Intelligent separators must be developed to achieve thermal regulation and controlled interface reaction from the separators to the functional membranes. Examples are as follows: (a) thermal regulating separator. This separator has an endothermic effect and excellent thermal diffusion functions. When the battery releases excess heat, the thermal control separator can absorb a large amount of heat and quickly diffuse, effectively suppressing the continuous heating and local overheating of the battery. (b) Self-healing separator: lithium dendrites are suppressed, and defects caused by lithium dendrites are repaired, avoiding thermal runaway.

(iii) Low-cost, simplified processes must be developed: some high-safety separators developed have difficulties in expanding their applications, and the preparation and raw materials limit their ability to only be used for button batteries or small bag batteries. Simplifying the production and reducing costs are fundamental engineering issues that need to be solved in the preparation of novel separators.

(iv) Assessment of the safety of power LIBs: most battery safety assessment tests are only conducted under isothermal conditions or just one adverse working condition. To address future battery issues, safety assessments should be conducted using large batteries under realistic or extreme conditions, such as immersion, fire, freezing and vibration, and the problem of LIBs under abusive conditions should be deeply investigated.

Author contributions

Boli Tong: investigation, writing – original draft, writing – review & editing. Xifei Li: project administration, supervision, writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We gratefully acknowledge the National Key Research and Development Program of China (2021YFB2400202).

References

1. X. Feng, D. Ren, X. He and M. Ouyang, Mitigating Thermal Runaway of Lithium-Ion Batteries, Joule, 2020, 4, 743–770.
2. J. Kalhoff, G. G. Eshetu, D. Bresser and S. Passerini, Safer Electrolytes for Lithium-Ion Batteries: State of the Art and Perspectives, ChemSusChem, 2015, 8(13), 2154–2175.
3. M. Su, G. Huang, S. Wang, Y. Wang and H. Wang, High safety separators for rechargeable lithium batteries, Sci. China: Chem., 2021, 64, 1131–1156.
4. X. Feng, M. Ouyang, X. Liu, L. Lu, Y. Xia and X. He, Thermal runaway mechanism of lithium ion battery for electric vehicles: A review, Energy Storage Mater., 2017, 10, 246–267.
5. Z. Liu, Y. Jiang, Q. Hu, S. Guo, L. Yu, Q. Li, Q. Liu and X. Hu, Safer Lithium-Ion Batteries from the Separator Aspect: Development and Future Perspectives, Energy Environ. Mater., 2021, 4, 336–362.
6. K. L. Jungjohann, R. N. Gannon, S. Goriparti, S. J. Randolph, L. C. Merrill, D. C. Johnson, K. R. Zavadil, S. J. Harris and K. L. Harrison, Cryogenic Laser Ablation Reveals Short-Circuit Mechanism in Lithium Metal Batteries, ACS Energy Lett., 2021, 6, 2138–2144.
7. J. Gao, W. Cai, Y. Hu and C. Chen, Improving the Flame Retardancy of Polyethylenes through Palladium-Catalyzed Incorporation of Polar Comonomers, Polym. Chem., 2019, 10, 1416–1422.
8. L. Mohammed, B. Boating, M. Mwemezi, L. Hamenu, A. Madzvamuse, A. Nyarko, M. Mohammed, W. Oduro, F. B. Agyenim, Y. M. Lee and J. M. Ko, EMI-BF₄ electrolyte and Al2O3/PVDF-HFP modified PE separator for high capacitance retention and cycle stability in supercapacitors, Korean J. Chem. Eng., 2022, 39(11), 3003–3011.
9. X. Zhang, E. Sahreai and K. Wang, Li-ion Battery Separators, Mechanical Integrity and Failure Mechanisms Leading to Soft and Hard Internal Shorts, Sci. Rep., 2016, 6, 32578.
10. H. Qian, H. Ren, Y. Zhang, X. He, W. Li, J. Wang, J. Hu, H. Yang, H. M. K. Sari, Y. Chen and X. Li, Surface Doping vs. Bulk Doping of Cathode Materials for Lithium-Ion Batteries: A Review, Electrochem. Energy Rev., 2022, 5, 2.
11. J. Zhu, X. Zhang, H. Luo and E. Sahraei, Investigation of the deformation mechanisms of lithium-ion battery components using in situ micro tests, Appl. Energy, 2018, 224, 251–266.
12. X. Huang, J. Xue, M. Xiao, S. Wang, Y. Li, S. Zhang and Y. Meng, Comprehensive evaluation of safety performance and failure mechanism analysis for lithium sulfur pouch cells, Energy Storage Mater., 2020, 30, 87–97.
13. H. Wu, D. Zhuo, D. Kong and Y. Cui, Improving battery safety by early detection of internal shorting with a bifunctional separator, Nat. Commun., 2014, 5, 5193.
14. H. Lee, X. Ren, C. Niu, L. Yu, M. H. Engelhard, I. Cho, M.-H. Ryou, H. S. Jin, H.-T. Kim, J. Liu, W. Xu and J.-G. Zhang, Suppressing Lithium Dendrite Growth by Metallic Coating on a Separator, Adv. Funct. Mater., 2017, 1704391.
15. M. S. Gonzalez, Q. Yan, J. Holoubek, Z. Wu, H. Zhou, N. Patterson, V. Petrova, H. Liu and P. Liu, Draining Over Blocking: Nano-Composite Janus Separators for Mitigating Internal Shorting of Lithium Batteries, Adv. Mater., 2020, 32(12), e1906836.
16. H. Zhou, C. Fear, M. Parekh, F. Gray, J. Fleetwood, T. Adams, V. Tomar, V. G. Pol and P. P. Mukherjee, The Role of Separator Thermal Stability in Safety Characteristics of Lithium-ion Batteries, J. Electrochem. Soc., 2022, 169, 090521.
17. J. Wang, G. Cao, R. Duan, X. Li and X. Li, Advances in Single Metal Atom Catalysts Enhancing Kinetics of Sulfur Cathode, Acta Phys.-Chim. Sin., 2023, 39(5), 2212005.
18. C. Yuan, L. Wang, S. Yin and J. Xu, Generalized separator failure criteria for internal short circuit of lithium-ion battery, J. Power Sources, 2020, 467, 228360.
19. J. Li, H. Jia, H. Li, X. Zhao, G. Sun, Z. Chang, L. Li, M. Jin, Z. Wang and M. Feng, Thin buffer layer assist carbon-modifying separator for long-life lithium metal anodes, J. Energy Chem., 2021, 57, 61–68.
20. W. Yao, M. Xu, W. Qiu, J. Wang, Y. Sun, J. Xu and Q. Zhang, Ultralight PEDOT Functionalized Separators toward High-Performance Lithium Metal Anodes, ChemElectroChem, 2021, 8, 2836–2845.
21. H. Lee, M. Yanilmaz, O. Toprakci, K. Fu and X. Zhang, A review of recent developments in membrane separators for rechargeable lithium-ion batteries, Energy Environ. Sci., 2014, 7, 3857–3886.
22. F. Sun, R. Moroni, K. Dong, H. Markötter, D. Zhou, A. Hilger, L. Zielke, R. Zengerle, S. Thiele, J. Banhart and I. Manke, A Study of the Mechanisms of Internal Short Circuit in a Li/Li Cell by Synchrotron X-ray Phase Contrast Tomography, ACS Energy Lett., 2017, 2(1), 94–104.
23. Z. Li, J. Chen, F. Lan and Y. Li, Constitutive Behavior and Mechanical Failure of Internal Configuration in Prismatic Lithium-Ion Batteries under Mechanical Loading, Energies, 2021, 14, 1219.
24. B. Huang, H. Hua, L. Peng, X. Wang, X. Shen, R. Li, P. Zhang and J. Zhao, The functional separator for lithium-ion batteries based on phosphonate modified nano-scale silica ceramic particles, J. Power Sources, 2021, 498, 229908.
25. L. Zhang, X. Li, M. Yang and W. Chen, High-safety separators for lithium-ion batteries and sodium-ion batteries: advances and perspective, Energy Storage Mater., 2021, 41, 522–545.
26. E.-S. Choi and S.-Y. Lee, Particle size-dependent, tunable porous structure of a SiO₂/poly(vinylidene fluoride-hexafluoropropylene)-coated poly(ethylene terephthalate) nonwoven composite separator for a lithium-ion battery, J. Mater. Chem., 2011, 21, 14747.
27. S.-G. Lim, H.-D. Jo, C. Kim, H.-T. Kim and D.-R. Chang, Electro-Spun Poly(vinylidene fluoride) Nanofiber Web as Separator for Lithium Ion Batteries: Effect of Pore Structure and Thickness, J. Nanosci. Nanotechnol., 2016, 16(1), 956.
28. G.-L. Zhu, Y.-Y. He, Y.-L. Deng, M. Wang, X.-Y. Liu, L.-P. Wang and J. Gao, Dependence of Separator Thickness on Li-Ion Battery Energy Density, J. Electrochem. Soc., 2021, 168, 110545.
29. D. V. Horváth, R. Tian, C. Gabbett, V. Nicolosi and J. N. Coleman, Quantifying the Effect of Separator Thickness on Rate Performance in Lithium-Ion Batteries, J. Electrochem. Soc., 2022, 169, 030503.
30. J. Chen, Q. Liu, B. Wang, F. Li, H. Jiang, K. Liu, Y. Wang, M. Li, Z. Lu and D. Wang, Hierarchical Polyamide 6 (PA6) Nanofibrous Membrane with Desired Thickness as Separator for High-Performance Lithium-Ion Batteries, J. Electrochem. Soc., 2017, 164(7), A1526–A1533.
31. W. Lin, F. Wang, H. Wang, H. Li, Y. Fan, D. Chan, S. Chen, Y. Tang and Y. Zhang, Thermal-Stable Separators: Design Principles and Strategies Towards Safe Lithium-Ion Battery Operations, ChemSusChem, 2022, 15, e202201464.
32. S. Yan, X. Huang and X. Xin, Measurement of the through thickness compression of a battery separator, J. Power Sources, 2018, 382, 13–21.
33. R. Pan, Z. Wang, R. Sun, J. L. Kristina Edstrom, M. Strømme and L. Nyholm, Thickness difference induced pore structure variations in cellulosic separators for lithium-ion batteries, Cellulose, 2017, 24, 2903–2911.
34. R. Miyagawa, J. Takeyoshi, K. Kanamura, S. Taminato, D. Mori and N. Imanishi, Electrochemical Evaluation of Lithium Metal Batteries Using Separators with Different Pore Sizes, Electrochemistry, 2022, 90(10), 107005.
35. D. Lv, J. Chai, P. Wang, L. Zhu, C. Liu, S. Nie, B. Li and G. Cui, Pure cellulose lithium-ion battery separator with tunable pore size and improved working stability by cellulose nanofibrils, Carbohydr. Polym., 2021, 251, 116975.
36. Y. Li, G. Zhang, B. Chen, W. Zhao, L. Sha, D. Wang, J. Yu and S. Shi, Understanding the separator pore size inhibition effect on lithium dendrite via phase-field simulations, Chin. Chem. Lett., 2022, 33, 3287–3290.
37. Y. Shimizu and K. Kanamura, Effect of Pore Size in Three Dimensionally Ordered Macroporous Polyimide Separator on Lithium Deposition/Dissolution Behavior, J. Electrochem. Soc., 2019, 166(4), A754–A761.
38. C. Cheng, R. Yang, Y. Wang, D. Fu, J. Sheng and X. Guo, A bacterial cellulose-based separator with tunable pore size for lithium-ion batteries, Carbohydr. Polym., 2023, 304, 120489.
39. R. E. Sousa, J. Nunes-Pereira, C. M. Costa, M. M. Silva, S. Lanceros-Méndez, J. Hassoun, B. Scrosati and G. B. Appetecchi, Influence of the porosity degree of poly(vinylidene fluoride-cohexafluoropropylene) separators in the performance of Li-ion batteries, J. Power Sources, 2014, 263, 29–36.
40. Z. Fu, H. Feng, C. Sun, X. Xiang, W. Wu, J. Luo, Q. Hu, A. Feng and W. Li, Influence of solvent type on porosity structure and properties of polymer separator for the Li-ion batteries, J. Solid State Electrochem., 2013, 17, 2167–2172.
41. X. Ma, P. Kolla, R. Yang, Z. Wang, Y. Zhao, A. L. Smirnova and H. Fong, Electrospun polyacrylonitrile nanofibrous membranes with varied fiber diameters and different membrane porosities as lithium-ion battery separators, Electrochim. Acta, 2017, 236, 417–423.
42. Y. Wang, Q. M. Li and Y. Xing, Porosity variation of lithium-ion battery separators under uniaxial tension, Int. J. Mech. Sci., 2020, 174, 105496.
43. F. Toquet, L. Guy, B. Schlegel, P. Cassagnau and R. Fulchiron, Combined roles of precipitated silica and porosity on electrical properties of battery separators, Mater. Chem. Phys., 2019, 223, 479–485.
44. R. Xu, H. Huang, Z. Tian, J. Xie and C. Lei, Effects of Coated Separator Surface Morphology on Electrolyte Interfacial Wettability and Corresponding Li-Ion Battery Performance, Polymers, 2020, 12, 117.
45. M. M. U. Din and R. Murugan, Metal Coated Polypropylene Separator with Enhanced Surface Wettability for High Capacity Lithium Metal Batteries, Sci. Rep., 2019, 9, 16795.
46. Y. Wang, Separator Wettability Enhanced by Electrolyte Additive to Boost the Electrochemical Performance of Lithium Metal Batteries, Nanomicro Lett., 2021, 13, 210.
47. Y. Xie, H. Zou, H. Xiang, R. Xia, D. Liang, P. Shi, S. Dai and H. Wang, Enhancement on the wettability o flithium battery separator toward nonaqueous electrolytes, J. Membrane Sci., 2016, 503, 25–30.
48. M. Liu, P. Zhang, L. Gou, Z. Hou and B. Huang, Enhancement on the thermostability and wettability of lithium-ion batteries separator via surface chemical modification, Mater. Lett., 2017, 208, 98–101.
49. J. Tan, L. Kong, Z. Qiu and Y. Yan, Flexible, high-wettability and thermostable separator based on fluorinated polyimide for lithium-ion battery, J. Solid State Electrochem., 2018, 22, 3363–3373.
50. S. Kalnaus, Y. Wang and J. A. Turner, Mechanical behavior and failure mechanisms of Li-ion battery separators, J. Power Sources, 2017, 348, 255–263.
51. X. Zhang, E. Sahraei and K. Wang, Deformation and failure characteristics of four types of lithium-ion battery separators, J. Power Sources, 2016, 327, 693–701.
52. W. Xiao, J. S. L. Huang, Z. Yang and Q. Qiao, PVA-ZrO₂ multilayer composite separator with enhanced electrolyte property and mechanical strength for lithium-ion batteries, Ceram. Int., 2020, 46(18), 29212–29221.
53. M. Li, L. Song, R. Xu, Y. Yang, Y. Bai, S. Song, G. Liu, T. Wang, X. Huang and J. He, Enhanced the mechanical strength of polyimide (PI) nanofiber separator via PAALi binder for lithium ion battery, Compos. Commun., 2021, 25, 100607.
54. Y. Zhai, N. Wang, X. Mao, Y. Si, J. Yu, S. S. Al-Deyab, M. El-Newehy and B. Ding, Sandwich-structured PVdF/PMIA/PVdF nanofibrous separators with robust mechanical strength and thermal stability for lithium ion batteries, J. Mater. Chem. A, 2014, 2, 14511.
55. F. Xiao, B. Xing, L. Kong and Y. Xia, Impedance-based diagnosis of internal mechanical damage for largeformat lithium-ion batteries, Energy, 2021, 230, 120855.
56. Y. Liu, X. Li, H. Guo, Z. Wang, Q. Hu and W. Peng, Overcharge performance of LiMn₂O₄/graphite battery with large capacity, J. Cent. South Univ. Technol., 2009, 16, 0763–0767.
57. Q. Yuan, F. Zhao, W. Wang, Y. Zhao, Z. Liang and D. Yan, Overcharge failure investigation of lithium-ion batteries, Electrochim. Acta, 2015, 178, 682–688.
58. R. Yang, G. Yu, Z. Wu, T. Lu, T. Hu, F. Liu and H. Zhao, Aging of lithium-ion battery separators during battery cycling, J. Energy Storage, 2023, 63, 107107.
59. L. Kong, Y. Li and W. Feng, Strategies to Solve Lithium Battery Thermal Runaway: From Mechanism to Modification, Electrochem. Energy Rev., 2021, 4, 633–679.
60. N. Kasai, M. Kato, N. Ishida, N. Kitamura and Y. Idemoto, Safety Test Methods Simulating Internal Short Circuit and the Mechanism for Safety Improvement of Li-ion Batteries by Heat Resistant Separators, Electrochemistry, 2020, 88(6), 555–559.
61. J.-S. Kim, D.-C. Lee, J.-J. Lee and C.-W. Kim, Optimization of Lithium-Ion Battery Pouch Cell for Maximization of Energy Density while Preventing Internal Short Circuit Caused by Separator Failure under Crush Load, J. Electrochem. Soc., 2021, 168, 030536.
62. C.-S. Kim, J.-S. Yoo, K.-M. Jeong, K. Kim and C.-W. Yi, Investigation on internal short circuits of lithium polymer batteries with a ceramic-coated separator during nail penetration, J. Power Sources, 2015, 289, 41–49.
63. D.-C. Lee and C.-W. Kim, Detailed Layered Nonlinear Finite Element Analysis for Lithium-Ion Battery Cells to Predict Internal Short Circuits Due to Separator Fractures under Hemisphere Indentation, J. Electrochem. Soc., 2020, 167, 120511.
64. X. Zhang, Q. Sun, C. Zhen, Y. Niu, Y. Han, G. Zeng, D. Chen, C. Feng, N. Chen, W. Lv and W. He, Recent progress in flame-retardant separators for safe lithium-ion batteries, Energy Storage Mater., 2021, 37, 628–647.
65. M. Bulla, S. Kolling and E. Sahraei, An Experimental and Computational Study on the Orthotropic Failure of Separators for Lithium-Ion Batteries, Energies, 2020, 13, 4399.
66. E. Hu, S.-M. Bak, J. Liu, X. Yu, Y. Zhou, S. N. Ehrlich, X.-Q. Yang and K.-W. Nam, Oxygen-Release-Related Thermal Stability and Decomposition Pathways of Li_xNi_0.5Mn_1.5O₄ Cathode Materials, Chem. Mater., 2014, 26, 1108–1118.
67. E. Lee, W. Lee, H. Kim, M. Kim, S. Yun, J. Kim and W.-S. Yoon, Decisive Factors in the Sequential Thermal Decomposition Reactions of Ni-Based Layered Cathode Materials, ACS Energy Lett., 2023, 8, 707–714.
68. K. L. Browning, L. Baggetto, R. R. Unocic, N. J. Dudney and G. M. Veith, Gas evolution from cathode materials: A pathway to solvent decomposition concomitant to SEI formation, J. Power Sources, 2013, 239, 341–346.
69. Y. Li, W. Wang, C. Lin, C. Yang and F. Zuo, Multi-physics safety model based on structure damage for lithium-ion battery under mechanical abuse, J. Cleaner Prod., 2020, 277, 124094.
70. J. G. Lee, J. Kim, J. B. Lee, H. Park, H.-S. Kim, J. H. Ryu, D. S. Jung, E. K. Kim and S. M. Oh, Mechanical Damage of Surface Films and Failure of Nano-Sized Silicon Electrodes in Lithium Ion Batteries, J. Electrochem. Soc., 2017, 164(1), A6103–A6109.
71. S. Zhang, Z. Shen and Y. Lu, Research Progress of Thermal Runaway and Safety for Lithium Metal Batteries, Acta Phys. Chim. Sin., 2021, 37(1), 2008065.
72. L. Peng, X. Wang, J. Dai, X. Shen, B. Huang, P. Zhang and J. Zhao, A reinforced ceramic-coated separator by overall-covered modification of electron-insulated polypyrrole for safety performance of lithium-ion batteries, Mater. Chem. Front., 2021, 5, 1884–1894.
73. F. Wang, X. Ke, K. Shen, L. Zhu and C. Yuan, A Critical Review on Materials and Fabrications of Thermally Stable Separators for Lithium-Ion Batteries, Adv. Mater. Technol., 2022, 7, 2100772.
74. P. Zhang, L. Chen, C. Shi, P. Yang and J. Zhao, Development and characterization of silica tube-coated separator for lithium ion batteries, J. Power Sources, 2015, 284, 10–15.
75. J. Dai, C. Shi, C. Li, X. Shen, L. Peng, D. Wu, D. Sun, P. Zhang and J. Zhao, A rational design of separator with substantially enhanced thermal features for lithium-ion batteries by the polydopamine-ceramic composite modification of polyolefin membranes, Energy Environ. Sci., 2016, 9, 3252–3261.
76. X. Wang, Y. Hu, L. Li, H. Fang, X. Fan and S. Li, Preparation and performance of polypropylene separator modified by SiO₂/PVA layer for lithium batteries, e-Polymers, 2019, 19, 470–476.
77. H. Liu, Z. Dai, J. Xu, B. Guo and X. He, Effect of silica nanoparticles/poly(vinylidene fluoride-hexafluoropropylene) coated layers on the performance of polypropylene separator for lithium-ion batteries, J. Energy Chem., 2014, 23, 582–586.
78. D. Fu, B. Luan, S. Argue, M. N. Bureau and I. J. Davidson, Nano SiO₂ particle formation and deposition on polypropylene separators for lithium-ion batteries, J. Power Sources, 2012, 206, 325–333.
79. H. Liu, J. Xu, B. Guo and X. He, Effect of Al₂O₃/SiO₂ composite ceramic layers on performance of polypropylene separator for lithiumion batteries, Ceram. Int., 2014, 40(9), 14105–14110.
80. J. Wang, Z. Hu, X. Yin, Y. Li, H. Huo, J. Zhou and L. Li, Alumina/Phenolphthalein Polyetherketone Ceramic Composite Polypropylene Separator Film for Lithium Ion Power Batteries, Electrochim. Acta, 2015, 159, 61–65.
81. Y. Deng, X. Song, Z. Ma, X. Zhang, D. Shu and J. Nan, Al₂O₃/PVdF-HFP-CMC/PE separator prepared using aqueous slurry and post-hot-pressing method for polymer lithium-ion batteries with enhanced safety, Electrochim. Acta, 2016, 212, 416–425.
82. D. Yeon, Y. Lee, M.-H. Ryou and Y. M. Lee, New flame-retardant composite separators based on metal hydroxides for lithium-ion batteries, Electrochim. Acta, 2015, 157, 282–289.
83. B. Jung, B. Lee, Y.-C. Jeong, J. Lee, S. R. Yang, H. Kim and M. Park, Thermally stable non-aqueous ceramic-coated separators with enhanced nail penetration performance, J. Power Sources, 2019, 427, 271–282.
84. S. Qin, M. Wang, C. Wang, Y. Jin, N. Yuan, Z. Wu and J. Zhang, Binder-Free Nanoparticulate Coating of a Polyethylene Separator via a Reactive Atmospheric Pressure Plasma for Lithium-Ion Batteries with Improved Performances, Adv. Mater. Interfaces, 2018, 1800579.
85. J.-J. Yuan, Y.-Z. Song, Y. Zhang, Z.-L. Qiu, C.-C. Sun, X. Yin, L.-P. Zhu and B.-K. Zhu, Polyphenols assisted silica coating on polypropylene separators with improved wettability and heat-resistance for lithium-ion batteries, J. Appl. Polym. Sci., 2019, 136, 47277.
86. T. Chen, S. Chen, Y. Chen, M. Zhao, D. Losic and S. Zhang, MOFs Containing Solid-state Electrolytes for Lithium Metal Batteries and Beyond, Mater. Chem. Front., 2021, 5, 1771–1794.
87. M. M. Rahman, S. Mateti, Q. Cai, I. Sultana, Y. Fan, X. Wang, C. Hou and Y. Chen, High temperature and high rate lithium-ion batteries with boron nitride nanotubes coated polypropylene separators, Energy Storage Mater., 2019, 19, 352–359.
88. Y. Mao, W. Sun, Y. Qiao, X. Liu, C. Xu, L. Fang, W. Hou, Z. Wang and K. Sun, A high strength hybrid separator with fast ionic conductor for dendrite-free lithium metal batteries, Chem. Eng. J., 2021, 416, 129119.
89. S. Karimzadeh, B. Safael, C. Yuan and T. C. Jen, Emerging Atomic Layer Deposition for the Development of High-Performance Lithium Ion Batteries, Electrochem. Energy Rev., 2023, 6, 24.
90. Z. Wei, Z. Zhang, S. Chen, Z. Wang, X. Yao, Y. Deng and X. Xu, UV-cured polymer electrolyte for LiNi_0.85Co_0.05Al_0.1O₂//Li solid statebattery working at ambient temperature, Energy Storage Mater., 2019, 22, 337–345.
91. P. Zhao, J. Yang, Y. Shang, L. Wang, M. Fang, J. Wang and X. He, Surface modification of polyolefin separators for lithium ion batteries to reduce thermal shrinkage without thickness increase, J. Energy Chem., 2015, 24, 138–144.
92. L. He, T. Qiu, C. Xie and X. Tuo, A phase separation method toward PPTA–polypropylene nanocomposite separator for safe lithium ion batteries, J. Appl. Polym. Sci., 2018, 135(39), 46697.
93. H. Cai, X. Tong, K. Chen, Y. Shen, J. Wu, Y. Xiang, Z. Wang and J. Li, Electrospun Polyethylene Terephthalate Nonwoven Reinforced Polypropylene Separator: Scalable Synthesis and Its Lithium Ion Battery Performance, Polymers, 2018, 10(6), 574.
94. Z. Zhang, W. Yuan and L. Li, Enhanced wettability and thermal stability of nano-SiO₂/poly(vinylalcohol)-coated polypropylene composite separators for lithium-ion batteries, Particuology, 2018, 37, 91–98.
95. L.-F. Fang, J.-L. Shi, J.-H. Jiang, H. Li, B.-K. Zhu and L.-P. Zhu, Improving the wettability and thermal resistance of polypropylene separators with a thin inorganic– organic hybrid layer stabilized by polydopamine for lithium ion batteries, RSC Adv., 2014, 4, 22501.
96. H. Wang, L. Pan, C. Wu, D. Gao, S. Chen and L. Li, Pyrogallic acid coated polypropylene membranes as separators for lithium-ion batteries, J. Mater. Chem. A, 2015, 3, 20535–20540.
97. J. Moon, J. Y. Jeong, J. I. Kim, S. Kim and J. H. Park, An ultrathin inorganic-organic hybrid layer on commercial polymer separators for advanced lithium-ion batteries, J. Power Sources, 2019, 416, 89–94.
98. J. Song, M.-H. Ryou, B. Son, J.-N. Lee, D. J. Lee, Y. M. Lee, J. W. Choi and J.-K. Park, Co-polyimide-coated polyethylene separators for enhanced thermal stability of lithium ion batteries, Electrochim. Acta, 2012, 85, 524–530.
99. W.-K. Shin, J.-H. Yoo and D.-W. Kim, Surface-modified separators prepared with conductive polymer and aluminum fluoride for lithium-ion batteries, J. Power Sources, 2015, 279, 737–744.
100. Q.-Y. Wu, H.-Q. Liang, L. Gu, Y. Yu, Y.-Q. Huang and Z.-K. Xu, PVDF/PAN blend separators via thermally induced phase separation for lithium ion batteries, Polymer, 2016, 107(19), 54–60.
101. W. Li, X. Li, X. Xie, A. Yuan and B. Xia, Effect of drying temperature on a thin PVDF-HFP/PETcomposite nonwoven separator for lithium-ion batteries, Ionics, 2017, 23, 929–935.
102. Q. Tian, Q. L. K. Song, Y. Mei, W. Lu, J. Peng, J. Zhou and Y. Ding, Fabrication of Aligned PI/GO Nanofibers for Battery Separators, Fibers Polym., 2021, 22, 30–35.
103. X. Niu, J. Liu, G. Song, Y. Li and T. He, Evidence of high temperature stable performance of polyether ether ketone (PEEK) separator with sponge-structured in lithium-ion battery, J. Mater. Sci., 2022, 57, 7042–7055.
104. X. Zhang, S. Cheng, X. Huang and B. E. Logan, The use of nylon and glass fiber filter separators with different pore sizes in air-cathode single-chamber microbial fuel cells, Energy Environ. Sci., 2010, 3, 659–664.
105. J. Zhang, Z. Liu, Q. Kong, C. Zhang, S. Pang, L. Yue, X. Wang, J. Yao and G. Cui, Renewable and superior thermal-resistant cellulose-based composite nonwoven as lithium-ion battery separator, ACS Appl. Mater. Interfaces, 2013, 5(1), 128–134.
106. S. Hu, S. Lin, Y. Tu, J. Hu, Y. Wu, G. Liu, F. Li, F. Yu and T. Jiang, Novel aramid nanofiber-coated polypropylene separators for lithium ion batteries, J. Mater. Chem. A, 2016, 4, 3513–3526.
107. Z. Wang, X. Li, N. Dong, B. Liu, G. Tian, S. Qi and D. Wu, Novel ZrO₂@Polyimde nano-microspheres-coated polyethylene separators for high energy density and high safety Li-ion battery, Mater. Today Energy, 2022, 30, 101155.
108. X. Xie, L. Sheng, C. Arbizzani, B. Gao, X. Gao, L. Yang, Y. Bai, H. Dong, G. Liu, T. Wang, X. Huang and J. He, Multi-functional groups decorated composite nanofiber separator with excellent chemical stability in ester-based electrolyte for enhancing the lithium-ion transport, J. Power Sources, 2023, 555, 232431.
109. C. Liao, L. Han, N. Wu, X. Mu, Y. Hu, Y. Kan and L. Song, Fabrication of necklace-like fibers separator by electrospinning technique for high electrochemical performance and safe lithium metal batteries, Mater. Chem. Front., 2021, 5, 5033–5043.
110. W. Hao, D. Kong, J. Xie, Y. Chen, J. Ding, F. Wang and T. Xu, Self-Polymerized Dopamine Nanoparticles Modified Separators for Improving Electrochemical Performance and Enhancing Mechanical Strength of Lithium-Ion Batteries, Polymers, 2020, 12, 648.
111. J. Chen, T. Kang, Y. Cui, J. Zhao, J. Xue, H. Xu and J. Nan, A Nonflammable and Thermally Stable Polyethylene/Glass Fiber-Magnesium Hydroxide/Polyethylene Composite Separator with High Mechanical Strength and Electrolyte Retention to Enhance the Performance of Lithium-Ion Batteries, Energy Technol., 2022, 10, 2101040.
112. J. Xu, L. Yuan, G. Liang and A. Gu, Achieving superiorly high heat-dimensional stability, high strength, and good electrochemical performance for electrospun separators in power lithium-ion battery through building unique condensed structure based on polyimide and poly(m-phenylene isophthalamide), J. Appl. Polym. Sci., 2021, e51233.
113. Q. Fu, W. Zhang, I. P. Muhammad, X. Chen, Y. Zeng, B. Wang and S. Zhang, Coaxially electrospun PAN/HCNFs@PVDF/UiO-66 composite separator with high strength and thermal stability for lithium-ion battery, Microporous Mesoporous Mater., 2021, 311, 110724.
114. W. Gong, J. Gu, S. Ruan and C. Shen, A high-strength electrospun PPESK fibrous membrane for lithium-ion battery separator, Polym. Bull., 2019, 76, 5451–5462.
115. L. Jiang, X. Zhang, Y. Chen, L. Qiao, X. Lu and X. Tian, Modified Polypropylene/Cotton Fiber Composite Nonwoven as Lithium-Ion Battery Separator, Mater. Chem. Phys., 2018, 219, 368–375.
116. A. Jana, D. R. Ely and R. E. Garcia, Dendrite-separator interactions in lithium-based batteries, J. Power Sources, 2015, 275, 912–921.
117. Y. Li, L. Sha, P. Lv, N. Qiu, W. Zhao, B. Chen, P. Hu and G. Zhang, Influences of Separator Thickness and Surface Coating on Lithium Dendrite Growth: A Phase-Field Study, Materials, 2022, 15, 7912.
118. S. Klein, J. M. Wrogemann, S. V. Wickeren, P. Harte, P. Bärmann, B. Heidrich, J. Hesper, K. Borzutzki, S. Nowak, M. Börner, M. Winter, J. Kasnatscheew and T. Placke, Understanding the Role of Commercial Separators and Their Reactivity toward LiPF₆ on the Failure Mechanism of High-Voltage NCM523||Graphite Lithium Ion Cells, Adv. Energy Mater., 2022, 12, 2102599.
119. B. Ghanbarzadeh, A. Khatibi, A. Asadi and B. Shokri, Regulating lithium-ion flux in the solid electrolyte interphase layer to prevent lithium dendrite growth on lithium metal anode, J. Energy Storage, 2022, 47, 103668.
120. Q. Zhao, R. Zhou, C. Wang, J. Kang, Q. Zhang, J. Liu, Y. Jin, H. Wang, Z. Zheng and L. Guo, Anion Immobilization Enabled by Cation-Selective Separators for Dendrite-Free Lithium Metal Batteries, Adv. Funct. Mater., 2022, 32, 2112711.
121. Z. Chang, H. Yang, A. Pan, P. He and H. Zhou, An improved 9 micron thick separator for a 350 Wh kg⁻¹ lithium metal rechargeable pouch cell, Nat. Commun., 2022, 13, 6788.
122. H. Yao, L. Yan, J. Shen, T. Wang, P. Chen, X. Cong, S. Zhang, H. Jiang and X. Zhao, Controllably regulating ion transport in lithium metal batteries via pore effect of metal–organic framework-based separators, Appl. Surf. Sci., 2022, 589, 152885.
123. W. Wang, A. C. Y. Yuen, Y. Yuan, C. Liao, A. Li, I. I. Kabir, Y. Kan, Y. Hu and G. H. Yeoh, Nano architectured halloysite nanotubes enable advanced composite separator for safe lithium metal batteries, Chem. Eng. J., 2023, 451, 138496.
124. T. Liu, X. Hu, Y. Zhang, T. He, J. Zhou and J. Qiao, Ion Transport Regulated Lithium Metal Batteries Achieved by Electrospun ZIF/PAN Composite Separator with Suitable ElectrolyteWettability, Batteries, 2023, 9, 166.
125. Y. Zhou, K. Zhao, J. Zhang, Y. Zhu, Y. Ma, H. Zhang, D. Song, X. Shi, L. Zhang and Y. Ding, Synergistic effects of nanodiamond modified separators toward highly stable and safe lithium metal batteries, J. Mater. Chem. A, 2021, 9, 16046–16055.
126. Y. Chen, J. Li, Y. Ju, R. Cheng, Y. Zhai, J. Sheng, H. Liu and L. Li, Regulating Li-ion flux distribution via holey graphene oxide functionalized separator for dendrite-inhibited lithium metal battery, Appl. Surf. Sci., 2022, 592, 153222.
127. K. Liu, D. Zhuo, H. W. Lee, W. Liu, D. Lin, Y. Lu and Y. Cui, Extending the Life of Lithium-Based Rechargeable Batteries by Reaction of Lithium Dendrites with a Novel Silica Nanoparticle Sandwiched Separator, Adv. Mater., 2017, 29(4), 1603987.
128. L. Li, M. Yu, C. Jia, J. Liu, Y. Lv, Y. Liu, Y. Zhou, C. Liu and Z. Shao, Cellulosic Biomass-Reinforced Polyvinylidene Fluoride Separators with Enhanced Dielectric Properties and Thermal Tolerance, ACS Appl. Mater. Interfaces, 2017, 9(24), 20885–20894.
129. X. Jiang, L. Xiao, X. Ai, H. Yang and Y. Cao, A Novel Bifunctional Thermo-Sensitive Poly(Lactic Acid)@Poly(Butylene Succinate) Core-Shell Fibrous Separator prepared by Coaxial Electrospun route for Safe Lithium-Ion Battery, J. Mater. Chem. A, 2017, 5, 23238–23242.
130. Z. Wei, J. Gu, F. Zhang, Z. Pan and Y. Zhao, Core-Shell Structured Nanofibers for Lithium Ion Battery Separator with Wide Shutdown Temperature Window and Stable Electrochemical Performance, ACS Appl. Polym. Mater., 2022, 2(5), 1989–1996.
131. H. Yang, X. Shi, S. Chu, Z. Shao and Y. Wang, Design of Block-Copolymer Nanoporous Membranes for Robust and Safer Lithium-Ion Battery Separators, Adv. Sci., 2021, 8(7), 2003096.
132. C.-T. Hsieh, S.-C. Lin, C.-H. Lee, C.-F. Liu and C.-C. Hu, Designing Multifunctional Polyethylene-Polyimide Composite Separators for Rechargeable Lithium-Ion Batteries, J. Electrochem. Soc., 2019, 166(14), A3132–A3138.
133. Z. Li, T. Cao, Y. Zhang, Y. Han, S. Xu and Z. Xu, Novel lithium ion battery separator based on hydroxymethyl functionalized poly(ether ether ketone), J. Membrane Sci., 2017, 540, 422–429.
134. J. Liu, Y. Mo, S. Wang, S. Ren, D. Han, M. Xiao, L. Sun and Y. Meng, Ultra-strong and Heat-resistant Poly(ether ether ketone) Separator for Dendrite-proof and Heat-resistant Lithium-ion Batteries, ACS Appl. Energy Mater., 2019, 2(5), 3886–3895.
135. G. Sun, L. Kong, B. Liu, H. Niu, M. Zhang, G. Tian, S. Qi and D. Wu, Ultrahigh-strength, nonflammable and high-wettability separators based on novel polyimide-core@polybenzimidazole-sheath nanofibers for advanced and safe lithium-ion batteries, J. Membrane Sci., 2019, 582, 132–139.
136. Z.-B. Shao, C. Deng, Y. Tan, M.-J. Chen, L. Chen and Y.-Z. Wang, An efficient mono-component polymeric intumescent flame retardant for polypropylene: preparation and application, ACS Appl. Mater. Interfaces, 2014, 6(10), 7363–7370.

---