Recent advances in composite membranes modified with inorganic nanoparticles for high-performance lithium ion batteries

Abstract

Separators with excellent physical and electrochemical performances are in urgent demand for higher power-density and energy-density lithium ion batteries (LIBs). This implies that these separators must possess high porosity, thin thickness, a large electrolyte uptake, and good thermal stability and mechanical properties. Thus, it is difficult for single component separators to satisfy all of these requirements. Recently, composite membranes modified with inorganic nanoparticles have attracted great interest as LIB separators, which show better physical and electrochemical performances than pure separators. In this article, recent advances, the main types, their manufacturing and the performances of composite membranes are thoroughly reviewed, covering three important types: inorganic particle-coated composite membranes, inorganic particle-filled composite membranes and inorganic particle-filled non-woven mats. The outlooks and future direction are also described in this review. This review will hopefully stimulate more extensive and insightful studies for fabricating and designing new LIB separators with excellent physical and electrochemical performances.

---

Qiaohuan Cheng

Qiaohuan Cheng graduated from Qilu University of Technology in 2014. In 2014, she moved to the Institute of Materials Science and Engineering, Qilu University of Technology as a postgraduate student, and majored in materials science and engineering.

Wen He

Wen He is currently a full professor in the Department of Materials Science and Engineering at Qilu University of Technology. She received her PhD degree from the Department of Materials Science and Engineering from South China University of Technology in 2009. Her research interests focus on materials chemistry and technology for lithium batteries.

Introduction

The reduction of fossil fuel energy supplies and the deterioration of the environment due to global warming have profound significance in the development, research and application of renewable energy resources. Lithium ion batteries (LIBs) have attracted much attention because they possess a high energy-density and power-density, and long cycling life,^1–6 fulfilling the demand for green energy and achieving the purpose of environmental protection simultaneously. Nowadays, LIBs are used widely in our daily life, such as cellular phones, laptops and digital cameras.⁷ There is no doubt that LIBs play an important role in our lives and productivity, becoming an indispensable part of our lives.

LIBs are usually composed of an anode, a cathode, a separator and an electrolyte (Fig. 1). The separator is a crucial component, so stricter requirements are needed for its performances. Firstly, the separator provides a physical barrier to prevent electrical short circuits resulting from the physical contact between the anode and cathode. Secondly, the separator must have electrical insulation to avoid internal short-circuits and afford channels for the migration of lithium-ions between the two electrodes.^8,9 Thirdly, the separator must be chemically and electrochemically stable toward the electrolyte and electrodes since most of the electrolyte is an organic system with strong corrosive properties.¹⁰ Fourthly, the separator should be inert, that is to say, it must not participate in chemical reactions when the battery is fully charged and discharged. Finally, ideal separators for LIBs should satisfy the following properties: good mechanical properties, good wettability with liquid electrolyte, high porosity, high permeability and low internal resistance.

Fig. 1 Schematic illustration of a typical lithium ion battery. Reproduced with permission from ref. 48. Copyright 2014 the Royal Society of Chemistry.

The separator can greatly impact the electrochemical performances of LIBs, which is closely related to the capacity, internal resistance, cycling capability and safety performance of batteries.^11–14 With the development of technology and increasing human demand, the application of LIBs is more and more extensive. The performance improvement of LIBs has made great progress, but still faces great challenges. In particular, when LIBs are used in large-sized devices, the high rate capability, long-term stability and high heat-resistance should be further improved. Thus, the performance improvement of separators is also confronted with great challenges. The requirements for the performance of separators are various when LIBs are applied to different areas, including the porosity, wettability, thickness and thermal stability. Especially for LIBs with liquid electrolyte, which have an increased potential risk of fire or explosion.^15,16 So, separators must have better physical and electrochemical performances to meet the various demands of LIBs. Conventional single component separators, such as polyolefin-based separators,¹⁷ poly(vinylidene fluoride) (PVDF), poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-co-HFP),^18–21 polyacrylonitrile (PAN),²² polyethylene terephthalate (PET)²³ and other non-woven fabrics, are not meeting the numerous and stringent requirements of LIBs.

For example, PVDF and its copolymer PVDF–HFP are usually employed as host polymers to prepare separators, which have good chemical and mechanical stabilities. They are usually prepared using the conventional casting method,²⁴ Bellcore technology method,²⁵ phase inversion technique,²⁶ or electrospinning process. However, PVDF-based separators show low ionic conductivity and electrochemical stability, which can be attributed to the crystalline part of PVDF preventing the migration of lithium ions, resulting in the LIBs showing a poor electrochemical performance.²⁷ Polyolefin-based separators, which are used widely in LIBs and have been realized through industrial production, possess excellent mechanical strength, and good electrochemical and chemical stability.^28,29 However, their inherent characteristics, such as a poor affinity with electrolytes and extensive shrinkage especially at high temperatures, prevent their application in high power and capacity LIBs.^30,31 The non-woven fabrics are usually fabricated using the electrospinning method,^32–35 such as PAN,^36–38 PVDF^39,40 and PVDF–HFP.^41–43 These electrospun separators have a high porosity to ensure the electrolyte uptake is enough and the migration of lithium ions. However, the discharge capability of pure electrospun separators has been limited at a high discharge rate due to polymer degradation and leakage of organic liquid electrolyte since these electrospun separators only possess macroporous structures (pore size > 1 μm) and do not have mesoporous structures (2 nm < pore < 50 nm).⁴⁴

Hence, separators with a single component have some disadvantages when they are applied in LIBs, and they can hardly meet the requirements of high-power LIBs. Therefore, it is highly desirable to fabricate a kind of separator, which has relatively high ionic conductivity and thermostability to ensure the security of the battery when it is used in high-power devices even in harsh environments. We expect that the separators will possess excellent electrochemical performances when they are assembled into LIBs. Thus, many researchers have done a vast amount of work to overcome the drawbacks of single component separators mentioned above. Some researchers proposed that single component separators can be modified to fabricate composite membranes by introducing inorganic nanoparticles,^45–47 which can improve the physical and electrochemical performance of separators effectively. The results indicated that the composite membranes had a better physical and electrochemical performance than pristine separators. That is to say, the pure separators can be modified successfully by introducing inorganic nanoparticles. In this paper, we introduce composite membranes with inorganic nanoparticles, which can be divided into three types: inorganic particle-coated composite membranes, inorganic particle-filled composite membranes and inorganic particle-filled non-woven mats.⁴⁸ The physical and electrochemical performance of composite membranes compared with bare separators is investigated.

Inorganic particle-coated composite membranes

It has been reported that nanosized inorganic particles were coated on the surface of separators with binders to modify the separators. The inorganic particles were aluminum oxide (Al₂O₃), silicon dioxide (SiO₂) and titanium dioxide (TiO₂).^49–52 The inorganic nanoparticles can significantly improve the thermal stability of separators and reduce thermal shrinkage of the composite membrane. They can also enhance the mechanical properties and optimize the electrochemical performance of LIBs. It is easy to form highly porous structures and well-connected interstitial voids between nanosized inorganic particles,^53–55 which play an important role in improving the performance of LIBs. The ionic conductivity and wettability of the composite membranes were also improved due to the high surface area and high hydrophilicity of the inorganic particles. Furthermore, the binders bonding the inorganic particles on the surface of the separators are also important to prepare inorganic particle-coated composite membranes. The binders directly influence the performance of the composite membrane and the electrochemical performance of LIBs. The binders are various, including PVDF, PVDF–HFP, poly(methyl methacrylate) (PMMA) and phenolphthalein polyetherketone (PEK-C).⁵⁶

The dip-coating process is the most common and simple method to prepare inorganic particle-coated composite membranes. J. H. Park et al.⁵³ developed a novel composite membrane using this method and synthesized a close-packed SiO₂/PMMA binary nanoparticle-coated polyethylene (PE) separator by introducing SiO₂ and PMMA binder nanoparticles. These composite membranes can effectively improve the thermal stability and electrochemical performance of a separator due to the formation of a unique structure of the separators, which possess a highly porous structure and well-connected interstitial voids between the binary SiO₂ and PMMA nanoparticles. This special structure had good wettability with liquid electrolyte because of the high surface area and high hydrophilicity of the inorganic particles, so the discharge capacity and C-rate capability of cells were improved significantly. P. Zhang et al.⁵⁷ coated one side of a commercial PE separator with one-dimensional silica tubes (ST), which improve the thermal stability of separators due to the formation of an interpenetrating network in the separator. The brief preparation method of this composite membrane is as follows: ST, styrene butadiene rubber (SBR), carboxyl methyl cellulose (CMC), and other reagents were mixed together to form slurries, then the slurries were coated onto one side of the PE membrane using an automatic coating machine. These composite membranes coated with ST can effectively enhance the thermostability at elevated temperatures compared to the separator coated with spherical silica particles and pristine PE separator. H. S. Jeong et al.⁵⁴ coated both sides of PE separators with a ceramic coating layer to improve thermal shrinkage using a simple dip-coating process, the coating layer was composed of SiO₂ nanoparticles and PVDF–HFP binders. Fig. 2a–d show the surface morphologies and thermal shrinkage of the composite separators, indicating that they have a well-developed porous structure. This porous structure is formed of well-connected interstitial voids between SiO₂ nanoparticles and improves the anti-thermal shrinkage of the composite membranes, resulting in the composite membrane being more resistant to rising temperatures than the conventional PE separator (135 °C). What is more, they even explored the effects of nanoparticle size (530 nm and 40 nm) on the performance of the composite membrane. They found that the composite membranes with small-sized SiO₂ possessed better properties, the porosity of 40 nm SiO₂ was 68%, and it was higher than the pristine PE separator, which was only 45%.

Fig. 2 FE-SEM photographs of surface morphologies for: (a) pristine PE separator, (b) 530 nm SiO₂-coated composite separator, and (c) 40 nm SiO₂ coated composite separator, (d) thermal shrinkage of composite separators and pristine PE separator as a function of heat-treatment temperature. Reproduced with permission from ref. 54. Copyright 2011 Elsevier.

In addition to solvent evaporation directly, inorganic particle-coated composite membranes can also be obtained using a phase inversion technique. H. S. Jeong et al.⁵⁸ prepared a composite membrane, which introduced coating layers onto both sides of a PE separator to improve thermal shrinkage. The coating layers were composed of Al₂O₃ nanoparticles and PVDF–HFP polymeric binders. PVDF–HFP and Al₂O₃ nanoparticles (50/50 by weight) were mixed in acetone, and here, water was added into the coating solution as a nonsolvent to control the microporous structure using a phase inversion technique. They found that the coating layers with a higher content of nonsolvent possessed a more developed microporous structure, leading to an excellent electrochemical performance. The same year, H. S. Jeong et al.⁵⁹ published another article about Al₂O₃/PVDF–HFP/PE composite membranes, the details were the same as the above-mentioned, with the different aspect being that they explored the effect of phase inversion, especially the solvent–nonsolvent miscibility, on the microporous structure of the coating layers of the composite membrane. The coating layers were divided into three types according to the system of solvent and nonsolvent, acetone–isopropyl alcohol system, acetone–butyl alcohol system and acetone–water system. The results indicated that not only the content of nonsolvent but also the types of nonsolvent can affect the performance of the composite membrane. The performances of the composite membranes were getting better in the following order: acetone–isopropyl alcohol system, acetone–butyl alcohol system and acetone–water system. All of these composite membranes showed good thermal stability and excellent electrochemical performance due to the stable microstructure of the inorganic particle-coating layers.

Recently, M. Kim et al.⁶⁰ employed a chemical vapor deposition method to modify a PE separator, which deposited SiO₂ nanoparticles onto the PE separator. The thin SiO₂-coated layers formed on the PE separator enhanced significantly their thermal and dimensional stability. R. S. Juang et al.⁶¹ developed a microwave method to modify tri-layered polymeric membranes (PP/PE/PP) by depositing TiO₂ onto polymeric membranes. The TiO₂ particle-coated layers can effectively improve the thermal and dimensional stability due to the formation of a robust skeleton. X. Li et al.⁶² synthesized a composite membrane by coating a PP separator with ceramic layers, which were composed of SiO₂ nanoparticles and PVDF–HFP binders. The polymer matrix was not directly used as mentioned above, the PP separator needed to be pretreated. 2,2,3,3,4,4,5,5-Octafluoropentyl methacrylate (OFPMA) was grafted onto the surface of the PP separator through plasma treatment, which can improve the adhesion between the PP separator and PVDF–HFP, and then the coating solutions were coated on the PP separator. The composite membranes presented an increased electrolyte uptake of 290 wt% and an ionic conductivity of 1.76 mS cm⁻¹ at room temperature. The discharge capacity of LiFePO₄/Li half cells based on this composite membrane remained about 150 mA h g⁻¹ after 100 cycles. T. H. Cho et al.⁶³ developed SiO₂ coated-composite membranes by means of an air-laid method. They found that SiO₂ particles were homogeneously distributed over all the composite membrane, filling the large pores (Fig. 3). These composite membranes had excellent thermal stability, cycling performance and wettability compared to a polyolefin-based separator and non-woven separators. S. M. Eo et al.⁶⁴ prepared a composite membrane where a PE separator was coated with PVDF–HFP using a simple dip-coating method firstly, then the PVDF–HFP coated separators were immersed in liquid electrolyte, which contained a small amount of inorganic additive (aluminum fluoride, aluminum iodide, lithium fluoride and lithium iodide). They found that the capacity retention and cycling performance of cells with composite membranes were improved due to the formation of protective layers to cover the active material in the electrode during cycling, which reduced electrolyte decomposition, so the structural stability of the active material can be enhanced. The performance of the obtained composite membranes with various inorganic additives showed little difference.

Fig. 3 SEM photographs of (a) the silica-free non-woven separator, (b) and (c) are surface and cross section images of the silica-composite non-woven membrane, respectively. (d) Capacity retention ratios as a function of the cycle number of cells at a C rate of 0.5. Reproduced with permission from ref. 63. Copyright 2008 The Electrochemical Society.

Although most work focuses on modifying polyolefin separators, efforts have been taken toward the modification of other polymeric separators. M. Kim et al.⁶⁵ developed a novel trilayer composite membrane, ceramic coating layer/PMMA/ceramic coating layer, and here, the ceramic coating layer was composed of Al₂O₃ nanoparticles and PVDF–HFP binders. The coating solutions were prepared by dispersing Al₂O₃/PVDF–HFP (90/10 by weight) in acetone, then they were applied to the PMMA polymer matrix using a simple dip-coating method. The schematic diagram and the cross-section morphology of the composite membrane are shown in Fig. 4a and b. These figures displayed that the composite membrane had a porous structure and the inner PMMA layers were imporous resulting in high electrolyte uptake. They found that the as-prepared composite membrane had higher ionic conductivity and thermal stability than the pristine PMMA separator. Fig. 4c and d show that the cells assembled with this composite membrane have excellent electrochemical performances, but the capacity of the cells decreased as the current rate increased, which can be attributed to the electrical polarization caused by a serial resistance increase. The composite membranes experienced a lower capacity drop at high discharge current densities.

Fig. 4 (a) Schematic diagram of the organic/inorganic trilayer separator. (b) Cross-section SEM image of the inorganic particulate film/PMMA/inorganic particulate film trilayer. (c) Discharge capacities of the unit cell (anode + inorganic/PMMA/inorganic separator + cathode) at various discharge rates after charging at 0.2C, (d) comparison of rate performance of cells with a PE separator and those with an inorganic/PMMA/inorganic separator at various discharge rates (ratio of capacities at various rates normalized against that at 0.2C). Reproduced with permission from ref. 65. Copyright 2010 Elsevier.

Y. Zhai et al.⁶⁶ constructed SiO₂/polyetherimide–polyurethane (PEI–PU) composite membranes via an electrospinning technique followed by a dip-coating process. The surface morphology images of these composite membranes in Fig. 5a–d show their high porosity and interpenetrating network structure. These composite membranes possessed excellent mechanical strength (15.63 MP) and high ionic conductivity (2.33 mS cm⁻¹). The Li/LiFePO₄ cell assembled with SiO₂/PEI–PU membranes exhibited excellent cycling stability. The cell had a capacity retention of 98.7% after 50 cycles at a 0.2C rate and better rate capability than the cell based on a Celgard membrane (Fig. 5e and f). E. S. Choi et al.⁵⁵ developed SiO₂/PVDF–HFP-coated polyethylene terephthalate (PET) non-woven composite membranes. These membranes had an unusual porous structure and well-connected interstitial voids formed between close-packed SiO₂ nanoparticles. These porous structures can be tuned by controlling the SiO₂ particle size (40 nm and 530 nm nanoparticles). Fig. 6a–d show the surface morphologies of the composite membranes with different particle sizes. The novel structure formed in this composite membrane helped improve the thermal stability and liquid electrolyte wettability. Specifically, the composite membranes with 40 nm SiO₂ nanoparticles offered higher ionic conductivity due to the ion transport being easier in the channel with higher porosity and a shorter tortuous path. Fig. 6e and f show schematic diagrams of the Li⁺ transport path in NC separators (a ceramic layer-coated non-woven composite separator) with different sizes of SiO₂ particles. They indicate that Li⁺ transport can be tuned by controlling the SiO₂ particle size.

Fig. 5 FE-SEM images of as-prepared (a and b) PEI–PU and (c and d) SiO₂/PEI–PU composite membranes with (a and c) low and (b and d) high magnification. (e) Cycling stabilities of Li/LiFePO₄ cells using Celgard, PEI–PU and SiO₂/PEI–PU composite membranes at a 0.2C rate. (f) Comparison of discharge rate capabilities. Reproduced with permission from ref. 66. Copyright 2015 Elsevier.

Fig. 6 FE-SEM photographs (surface) of: (a) PE separator; (b) pristine PET non-woven separator; (c) 40 nm-SiO₂ NC separator; (d) 530 nm-SiO₂ NC separator. Schematic illustrations showing the SiO₂ powder size-dependent porous structure of the NC separators and its influence on ion transport: (e) 40 nm-SiO₂ NC separator; (f) 530 nm-SiO₂ NC separator. Reproduced with permission from ref. 55. Copyright 2011 the Royal Society of Chemistry.

Above all, the physical and electrochemical performance of the separators has been significantly improved by coating the surface with inorganic particles to form inorganic particle-coated composite membranes. The mechanical strength, thermal stability and electrolyte uptake are also enhanced remarkably compared with pristine separators, but the porosity reduced with the increasing thickness of the separator because of the introduction of the ceramic coated layers.

Inorganic particle-filled composite membranes

Inorganic particles not only can be coated on the surface of separators, but also can be incorporated directly into polymer matrices. The composite membranes formed by directly introducing inorganic nanoparticles into the polymer matrix are called inorganic particle-filled composite membranes, which is another effective way to improve the performance of separators. Various kinds of inorganic particles can be used in this type of composite membrane, including lithium aluminate (LiAlO₂), manganese oxide (MgO), Al₂O₃, TiO₂ and SiO₂. For example, Al₂O₃ has been filled into a PVDF separator to form a composite membrane, which has excellent thermal stability and cycling performance.⁶⁷ D. W. Kim⁶⁸ developed a SiO₂-filled poly(ethylene-co-methyl acrylate) (PE-co-MA) composite membrane, and its ionic conductivity can reach 5.8 × 10⁻⁴ S cm⁻¹ at room temperature. The cell assembled with carbon/composite membrane/LiCoO₂ delivered a discharge capacity of 124 mA h g⁻¹ at a rate of 0.1C and 120 mA h g⁻¹ at 0.5C. Besides, LiAlO₂, Al₂O₃ and MgO have been used as fillers in a PVDF–HFP polymer matrix and the performances of separators for LIBs were improved significantly.⁶⁹

P. Yan et al.⁷⁰ prepared a TiO₂/PVDF–HFP inorganic composite membrane using a nonsolvent evaporation method, where TiO₂ was employed as a filler to reduce the crystallinity of the composite membranes. Fig. 7a shows the preparation process of this composite membrane. The effect of nano-TiO₂ on the crystallinity of a porous polymer membrane was studied. They found that the crystallinity of the polymer membrane was reduced with the introduction of TiO₂, resulting in an increasing number of charge carriers and ionic conductivity. The LiFePO₄/Li cell assembled with this composite membrane exhibited a stable electrochemical performance and excellent rate capability (Fig. 7b–d). For example, the discharge capacity of the cell was 164, 157, 143, and 122 mA h g⁻¹ at 0.1C, 0.5C, 1C, and 3C, respectively, and the cell had only about a 6% discharge capacity loss at 1C after 50 cycles.

Fig. 7 (a) Picture of the fabrication process of a composite micro-porous polymer membrane; (b) the charge–discharge profiles of Li/PVDF–HFP/LiFePO₄ at different rates; (c) the charge–discharge profiles of Li//LiFePO₄ with the TiO₂/PVDF–HFP composite membrane at different rates; (d) the cycling performance of Li//LiFePO₄ with the TiO₂/PVDF–HFP composite membrane at 1C. Reproduced with permission from ref. 70. Copyright 2015 Springer.

The preparation process of the above-mentioned composite membranes is as follows; the inorganic nanoparticles, polymer matrix and solvent (some with nonsolvent) were mixed together to form a slurry according to a precalculated proportion of the materials. Then the slurry was cast on a clean glass plate. After evaporation of the solvent (or nonsolvent) in the slurry, the separator was separated from the glass plate and the composite membrane with a certain thickness was obtained. But the intrinsically microporous structure of these composite membranes limited the absorbing capacity of liquid electrolyte to a certain extent. Therefore, some researchers have proposed the phase inversion method to prepare composite membranes, which have a highly porous structure, to get enough liquid electrolyte. For example, K. M. Kim et al.⁷¹ developed a PVDF–HFP-based composite membrane using the phase inversion technique, which was filled with TiO₂ nanocrystalline particles. PVDF–HFP and TiO₂ in a certain ratio were dispersed into N-methyl-2-pyrrolidone (NMP) (solvent) to make the slurry. The slurry was spread on a clean glass plate, then the glass plate was dipped into a flowing water bath rapidly, and the composite membranes were peeled off from the glass plate after the onset solidification, finally the composite membranes were obtained after drying. The obtained composite membranes showed excellent electrochemical properties, higher ionic conductivity, better wettability and enhanced interfacial stability compared to cast PVDF–HFP separators. A. D. Pasquier et al.⁷² developed an SiO₂-filled PVDF–HFP composite membrane using phase inversion, with acetone and ethanol employed as the solvent and nonsolvent, respectively. SiO₂ was uniformly dispersed in the separator resulting in the formation of a stable porous structure, which reinforced the mechanical properties of the separator. The composite membrane exhibited high electrolyte uptake, which was beneficial to the discharge and rate capacity of LIBs. A. Subramania et al.⁷³ prepared a PVDF–HFP-based micro-porous composite membrane using phase inversion with ZrO₂ used as a filler. These kinds of membrane showed excellent performance, such as high ionic conductivity, good compatibility and high discharge capability.

Furthermore, the effects of different solvents and nonsolvents, and filled amounts of inorganic particles on the performances of composite membranes fabricated using the phase inversion process were also investigated. R. Miao et al.⁷⁴ researched the effect of different solvent and nonsolvent systems on the performances of a PVDF–HFP-based composite membrane in which TiO₂ nanoparticles were used as fillers. The system of solvent and nonsolvent was divided into seven types: acetone and ethanol, tetrahydrofuran (THF) and water, THF and ethanol, acetone and water, N-methyl pyrrolidone (NMP) and acetone, NMP and acetone and water, and NMP and acetone and ethanol. The results showed that the composite membranes prepared with different solvents and nonsolvents exhibited different properties, including morphology, electrolyte uptake, porosity and mechanical performance. The system of acetone and water possessed a good pore structure and pore size resulting in high electrolyte uptake. Z. Li et al.⁷⁵ studied the effects of different amounts of Al₂O₃ nanoparticles on the performances of the PVDF–HFP-based composite membranes prepared through the phase inversion process using N,N-dimethyl formamide (DMF) as the solvent and distilled water as the nonsolvent. The results showed that the crystallinity of the composite membrane decreased with the addition of Al₂O₃ and the amorphous phase content increased accordingly. This may be attributed to the cross-linking centers formed by the interaction of Lewis acid group ceramics (e.g. the –OH groups on the alumina surface) with the polar groups (e.g. the –F atoms of the polymer chains), which hindered the reorganization of the polymer chains. On the other hand, the electrochemical properties of the composite membrane filled with Al₂O₃ nanoparticles were improved significantly. For example, the lithium-ion transference number increased with the rise in Al₂O₃ content (the mass fraction of Al₂O₃ is below 10%), because the addition of Al₂O₃ nanoparticles can weaken the interaction between Li ions and the F atoms of the polymer chains.

As we all know, because there is inevitably some acidic HF in the LiPF₆-based electrolytes now widely used in LIBs, HF corrosion of cathode materials is one of the important reasons for capacity fading.^76,77 To resolve this problem, S. S. Zhang et al.⁷⁸ studied novel inorganic composite membranes, which were composed of alkali calcium carbonate (CaCO₃) powders and a small amount of polytetrafluoroethylene (PTFE) polymer binder. They selected alkali CaCO₃ as the main component for the composite membrane because it can neutralize acidic products in situ. The results showed that the composite membranes had high ionic conductivity and excellent wettability due to high porosity and good capillarity, and the LIBs with this composite membrane exhibited a stable capacity retention and good high-rate performance.

However, the single component polymer matrix still could not completely satisfy the numerous requirements of separators, such as good mechanical strength and wettability as well as thermal stability. Considering all of these factors, some researchers developed a blend polymer matrix to improve the performance of separators, and with this blending it was easy to control the properties of the separators by adjusting the content of the blended polymer matrix.^79–82 N. T. K. Sundaram et al.⁸³ prepared PVDF–HFP/polyvinyl alcohol (PVA) blend composite membranes using the phase inversion method with LiAlO₂ nanoparticles used as fillers. The results showed that the ionic conductivity was enhanced due to the increasing electrolyte uptake resulting from the increasing pore size, surface area and porosity. The composite membranes also showed an excellent ionic conductivity of 8.12 × 10⁻³ S cm⁻¹ at room temperature. H. Huang et al.⁸⁴ developed a PVDF–HFP/polystyrene (PS)-based composite membrane with SiO₂ employed as the filler. The porosity of the composite membrane increased with the increasing content of PS, and consequently, the ionic conductivity was improved.

The means of direct addition of inorganic particles in the polymer matrix can enhance the mechanical strength as well as absorb adequate liquid electrolyte, improve the affinity of separator and electrolyte, enlarge the contact surface of them, reduce internal resistance and increase the ionic conductivity. Inorganic nanoparticles can not only act as a solid plasticizer hindering the crystallization and reorganization of polymer chains but can also interact with polar groups of the polymer chains through the Lewis acid–base effect,^85–88 which plays a vital role in improving the ionic conductivity, interfacial stability and electrochemical stability of the composite membranes.

As discussed above, inorganic particles can be coated on the surface of separators as well as can be incorporated within a polymer matrix directly to fabricate composite membranes, and the physical and electrochemical performances of these two kinds of composite membrane were improved significantly. However, certain problems exist in inorganic particle-coated composite membranes: (1) it is inevitable that the thickness increased and the porosity reduced with the introduction of inorganic particle-coating layers on the separator, resulting in higher internal resistance and lower energy and power density;⁴⁸ (2) the adherence strength between the inorganic particle-coating layer and the polymer matrix is weaker, due to the difference in their physical and chemical properties; (3) the inorganic particle-coating layers on the separator can chip off easily during charge–discharge cycles because of the difference in expansion coefficients between the inorganic particle-coating layer and the polymer matrix. Nevertheless, the inorganic particle-filled composite membranes can solve these problems well. So, the results show that the comprehensive performances of the inorganic particle-filled composite membranes are superior to those of the inorganic particle-coated composite membranes.

Inorganic particle-filled non-woven mats

Non-woven-based separators have attracted much attention because they have high porosity and low cost compared with microporous polyolefin-based separators. They have the potential to replace polyolefin membranes. However, most traditional non-woven separators have a large pore size, leading to safety concerns when they are directly used in LIBs. Recently, electrospinning technology has been developed to prepare a non-woven separator, which consists of micron and submicron fibers. The electrospun separators had high porosity, fully interconnected pore structures and high surface areas, resulting in high electrolyte uptake and the facile transport of lithium ions. They have been widely used in LIBs with high rate capability.^89–91

However, one fateful shortcoming of electrospun non-woven separators is their poor mechanical stability, thus, they cannot withstand the large tension developed by the winding operation used during battery assembly. Besides, the high crystallinity of polymer separators is another major disadvantage.⁹² All of these limited their use in LIBs. Some researchers proposed that non-woven separators can be modified by introducing inorganic nanoparticle fillers to fabricate inorganic particle-filled non-woven mats, which can significantly improve the performance of separators. Inorganic nanoparticles, such as barium titanate (BaTiO₃), ZrO₂, Al₂O₃, TiO₂ and SiO₂, can be incorporated along with the polymer matrix to fabricate composite membranes.^93–99

The introduction of inorganic fillers into non-woven separators can improve the ionic conductivity of the separator by reducing the crystallinity of the polymer matrix and introducing Lewis acid/base interactions between the polar surface group on the filler surface and the ionic species in the liquid electrolyte.^100,101 The inorganic fillers can also enhance the mechanical properties of the separator, improve the cation transference number and enhance the interfacial stability between the liquid electrolyte and lithium metal electrode.^102,103

Y. Ding et al.¹⁰⁴ developed PVDF/TiO₂ composite membranes using an electrospinning process with tetrabutyl titanate as the source of TiO₂. They investigated the effect of TiO₂ on the morphology, crystallinity and electrochemical behavior of the composite membranes. The results showed that the ionic conductivity and cycling performance of the electrospun composite membranes were improved, and the ionic conductivity can reach 1.4 × 10⁻³ S cm⁻¹ at room temperature. M. Yanilmaz et al.¹⁰⁵ prepared SiO₂/PVDF hybrid composite membranes by electrospraying a SiO₂ dispersion and electrospinning PVDF solution simultaneously. The unique morphology of the SiO₂/PVDF composite membrane resulting from loading a large amount of SiO₂ nanoparticles on the fiber surface can be observed from the SEM images (Fig. 8a–d). These research results showed that SiO₂/PVDF composite membranes had better physical and electrochemical performances than both a pristine PP separator and PVDF nanofiber membranes, such as better wettability, higher ionic conductivity, lower interfacial resistance and better cycling performance. We found that the composite membranes with 24% SiO₂ exhibited better C-rate and cycling performance than others. P. Raghavan et al.¹⁰⁶ prepared PVDF–HFP-based composite membranes with different amounts of SiO₂ using the electrospinning process, and the electrospinning solution contained in situ generated SiO₂. These novel composite membranes had high porosity due to the formation of fully interconnected pores, high electrolyte uptake, high ionic conductivity, low interfacial resistance and excellent cycling performance. Especially the composite membranes with 6% in situ silica which showed a maximum ionic conductivity of 8.06 mS cm⁻¹ at 20 °C.

Fig. 8 (a–d) SEM images of SiO₂/PVDF nanoparticle/nanofiber hybrid membranes with different SiO₂ content. (e) Cycling performance of Li/LiFePO₄ cells containing SiO₂/PVDF nanoparticle/nanofiber hybrid membranes and a microporous PP membrane at 0.2C. (f) C-rate performance of Li/LiFePO₄ cells containing SiO₂/PVDF nanoparticle/nanofiber hybrid membranes and a microporous PP membrane. Reproduced with permission from ref. 105. Copyright 2014 Elsevier.

The effects of the types and content of inorganic nanoparticles were also investigated in detail. J. K. Kim et al.¹⁰⁷ investigated the effect of the content of SiO₂ nanoparticles on the performance of composite membranes prepared using the electrospinning method. The morphologies of SiO₂/PVDF–HFP composite membranes are shown in Fig. 9a–c, indicating that the composite membranes with SiO₂ have high porosity to ensure adequate electrolyte uptake. After activating with 1-butyl-3-methylimidazolium bis(trifluoromethanesulfonylimide) (BMITFSI), the composite membranes with 6% SiO₂ not only showed excellent physical and electrochemical performance, but the ionic conductivity also reached 4.3 × 10⁻³ S cm⁻¹ at 25 °C. The discharge capacity of Li/LiFePO₄ cells assembled with this composite membrane can reach 170 mA h g⁻¹ at 0.1C, which is equal to the theoretical specific capacity of LiFePO₄, and the cells showed a very stable cycling performance (Fig. 9d and e). P. Raghavan et al.¹⁰⁸ investigated the effect of different inorganic fillers on the performance of PVDF–HFP-based composite membranes prepared using the electrospinning technique. The inorganic nanoparticles in this composite membrane are BaTiO₃, Al₂O₃ and SiO₂. The resultant composite membranes showed a well-interconnected porous structure (Fig. 10a–d) and high electrolyte uptake, which also improved their mechanical stability and ionic conductivity. Especially the composite membranes with BaTiO₃, which showed the highest electrolyte uptake, ionic conductivity and electrochemical stability. For example, the Li/LiFePO₄ cell assembled with the BaTiO₃-filled composite membranes delivered a discharge capacity of 164 mA h g⁻¹ and excellent cycling performance (Fig. 10e and f). All of these can be attributed to its better interaction with the polymer matrix and compatibility with lithium metal.

Fig. 9 SEM images of electrospun P(VDF–HFP) membranes with varying SiO₂ content: (a) 0 wt%, (b) 6 wt%, and (c) 10 wt%. (d) First cycle charge and discharge capacities of Li/LiFePO₄ cells with PE based electrospun P(VDF–HFP) membranes with varying SiO₂ content (25 °C, 0.1C, 2.5–4.0 V). (e) Cycling performance of Li/LiFePO₄ cells with PE based electrospun PVDF–HFP membranes with varying SiO₂ content (25 °C, 0.1C, 2.5–4.0 V). Reproduced with permission from ref. 107. Copyright 2008 Elsevier.

Fig. 10 SEM images of electrospun P(VDF–HFP) membranes with (a) no filler, (b) SiO₂, (c) Al₂O₃, and (d) BaTiO₃. (e) Cycle performance of Li/NCPE/LiFePO₄ cells with polymer electrolyte based electrospun P(VDF–HFP) membranes containing 6% BaTiO₃ (25 °C, 0.1C, 2.5–4.0 V). (f) Initial charge–discharge properties of Li/NCPE/LiFePO₄ cells with polymer electrolyte based electrospun P(VDF–HFP) membranes containing different ceramic fillers (25 °C, 0.1C, 2.5–4.0 V). Reproduced with permission from ref. 108. Copyright 2008 Elsevier.

In addition to PVDF and its copolymer which can be employed as a host polymer, PAN^{106,109–111} can also be used as the polymer matrix to fabricate inorganic-filled non-woven composite membranes since PAN has superior mechanical stability and fast lithium ion transport. Y. Liang et al.¹¹² prepared lithium lanthanum titanate oxide (LLTO)/polyacrylonitrile (PAN)-based composite membranes using electrospinning technology. A series of tests were carried out and it was found that the composite membranes with 15 wt% LLTO provided the highest ionic conductivity of 1.95 × 10⁻³ S cm⁻¹. The obtained composite membranes had greater liquid electrolyte uptake, lower interfacial resistance and excellent electrochemical performance compared with pristine PAN separators. H. R. Jung et al.⁹⁷ developed hydrophilic fumed silica (SiO₂)/PAN composite membranes with varying SiO₂ content using electrospinning technology. Among them, the composite membrane with 12 wt% SiO₂ had the highest porosity, largest surface area and highest ionic conductivity (1.1 × 10⁻² S cm⁻¹), and the graphite/LiCoO₂ cell assembled with this composite membrane delivered a discharge capacity of 139 mA h g⁻¹ for the initial cycle, and retained a high value of 127 mA h g⁻¹ after 150 cycles, which is significantly higher than for pristine PAN separators.

Most of the above-mentioned composite membranes were mostly prepared using the electrospinning method. Another method for fabricating high-performance fiber-based separators is centrifugal spinning, which is a fast, cost-effective and safe method.^113,114 M. Yanilmaz et al.¹¹⁵ developed SiO₂/PAN composite membranes using centrifugal spinning. The preparation process comprises the following steps: firstly, PAN was dissolved in DMF, then SiO₂ powders were added to the above solution in certain ratios to form a homogeneous solution and finally, the SiO₂/PAN composite membranes were obtained using centrifugal spinning technology. Fig. 11g shows a schematic diagram of the centrifugal spinning process. Fig. 11a–e show the SEM images of the SiO₂/PAN composite membranes with varying SiO₂ content. From Fig. 11a–e, the morphologies of the composite membranes with SiO₂ nanoparticles have no obvious differences compared with bare PAN separators. But it was found that the average fiber diameters decreased with the increase in SiO₂ content because the repulsive force of SiO₂ can minimize the entanglement of polymer chains and increase porosity. Meanwhile, the obtained composite membranes showed better wettability, higher ionic conductivity, higher electrolyte uptake, lower interfacial resistance, and better rate capability and cycling performance than pristine PAN separators; the lithium/lithium iron phosphate cells based on these composite membranes exhibited good cycling performance and rate capacity (Fig. 11f).

Fig. 11 (a–e) SEM images of SiO₂/PAN membranes with varying SiO₂ content. (f) C-rate performance of Li/LiFePO₄ cells containing SiO₂/PAN membranes with varying SiO₂ content and a microporous PP membrane. (g) Schematic diagram of the centrifugal spinning process. Reproduced with permission from ref. 115. Copyright 2015 Elsevier.

In addition to the electrospun composite membranes that were applied directly to LIBs, a non-woven polymer combined with other polymer matrices can also be used as coating layers. Their unique structures can enhance the physical and electrochemical performance of separators. For example, M. Yanilmaz et al.¹¹⁶ prepared a novel composite membrane by coating a PP separator with an electrospun SiO₂/PVDF non-woven mat. The results showed that the addition of SiO₂ nanoparticles improved the physical and electrochemical performance of separators, including the wettability, ionic conductivity, electrochemical stability, interfacial resistance and cycling performance. They also found that the composite membranes with 15% SiO₂/PVDF provided the highest ionic conductivity of 2.6 × 10⁻³ S cm⁻¹. T. H. Cho et al.³⁰ synthesized a non-woven composite membrane using a hot roll press method at 150 °C, which combined a PAN nano-fiber non-woven with ceramic Al₂O₃ or SiO₂. Fig. 12a shows the hot roll press method and the SEM images of the composite membranes. From Fig. 12a, the ceramic powder is sandwiched between the PAN nano-fiber non-woven and the polyolefin non-woven layer, forming a stable structure of three layers. They investigated the physical, electrochemical and thermal properties of the composite membranes. The results indicated that the as-prepared composite membrane had higher porosity, air permeability and thermal stability than commercial polymer separators. The LiCoO₂/graphite cells assembled with this composite membrane showed better rate capabilities and cycling performance (Fig. 12b and c), which had a high capacity retention of about 88% covering 200 cycles, and no obvious thermal shrinkage was observed when it was exposed at 150 °C.

Fig. 12 (a) Schematic depiction of the combining process. (b) Discharge capacities vs. cycle numbers of the test cells. (c) Results of rate capability tests for the cells with the Celgard 2400, and the CNS no. 1 (composite membranes with SiO₂ fillers) and no. 2 (composite membranes with Al₂O₃ fillers). Reproduced with permission from ref. 30. Copyright 2010 Elsevier.

Thus, introducing inorganic nanoparticles into non-woven fabrics (PVDF and its copolymer PVDF–HFP, and PAN, polyethylene oxide (PEO),^117,118 poly(ethylene) glycol (PEG),¹¹⁹ PVDF¹²⁰ and a blend of PEO/PVDF–HFP¹²¹) can fabricate composite membranes with high performance. These composite membranes were sandwiched between an anode and a cathode to assemble cells. The cells showed excellent electrochemical performances, good C-rate performance and high cycling stability. Meanwhile, some researchers also suggest that the separators can also be modified successfully by introducing coating layers without inorganic particles. For example, S. Y. Xiao et al.¹²² prepared a composite membrane by coating electrospun PVDF membranes on the two sides of a methyl cellulose (MC)-based host polymer. Fig. 13a–d are the SEM images of this composite membrane, showing the porous structure of the as-prepared composite membranes with outer PVDF layers. This porous structure resulted in high electrolyte uptake. The MC layer was sandwiched between two PVDF layers, forming a three layer structure. The results showed that the ionic conductivity of the composite membranes can reach 1.5 mS cm⁻¹ at room temperature, which is much higher than for pure MC membranes. These composite membranes also exhibited good mechanical properties, a high lithium ion transference number and excellent thermal stability. The Li/LiFePO₄ cell assembled with the PVDF/MC/PVDF composite membranes showed a high discharge capacity and good rate capacity (Fig. 13e and f). For example, its discharge capacity was about 158 mA h g⁻¹ at 0.1C, which was also higher than that for bare MC separators, which was about 138 mA h g⁻¹. When tested at different current densities such as 0.2C, 0.5C and 1C, its discharge capacity was still higher than for MC separators, and after 40 cycles there was still no evidence of capacity fading at 0.2C.

Fig. 13 SEM micrographs of (a) the surface and (b) a cross section of MC, (c) the surface of PVDF and (d) a cross section of the sandwiched membrane PVDF/MC/PVDF. Electrochemical performance of LiFePO₄ tested using MC and the PVDF/MC/PVDF sandwiched membrane as separators, saturated with 1 mol L⁻¹ LiPF₆ electrolyte and Li metal as the counter electrode: (e) rate behavior, and (f) cycling behavior. Reproduced with permission from ref. 122. Copyright 2014 Elsevier.

G. Ding et al.¹²³ had successfully modified the cellulose non-woven membranes by coating a single ion polymer electrolyte via a dip-coating process. This membrane was composed of polymeric lithium tartaric acid borate salt (PLTB) and PVDF–HFP. This composite membrane with a unique coating layer exhibited higher conductivity, better flame retardancy and better thermal stability compared with a pristine PP separator and pure cellulose membrane. Moreover, the as-prepared composite membranes showed a high lithium ion transference number of 0.48, good rate capability and excellent cycling performance. C. Cao et al.¹²⁴ prepared hydrophilic electrospun PVDF composite membranes by coating mussel-inspired polydopamine (PDA); the schematic diagram for preparing the composite membrane is shown in Fig. 14a. They investigated the morphology and the physical and electrochemical performances of the obtained composite membranes. The results showed that the surface of the obtained composite membranes is hydrophilic, while PVDF is hydrophobic, resulting in increasing electrolyte uptake and ionic conductivity, thus improving the electrochemical performance of the cells. The charge–discharge capability, C-rate capability and cycling performance were also investigated. In the cells assembled with the PDA–PVDF composite membranes, LiMn₂O₄ was used as the cathode material and lithium metal was employed as the anode. The discharge capacity of this cell was 104.5 mA h g⁻¹, which was better than that of the cells using PVDF separators, which was about 101.1 mA h g⁻¹ at 0.5C (Fig. 14b). The cell also exhibited very good stable cycling performance, as good as the PVDF nanofiber. The results of the rate test in Fig. 14c show that the PDA–PVDF composite membranes have higher capacity retention compared with the retention of the PVDF separators at various rates.

Fig. 14 (a) Schematic illustration of the mussel-inspired polydopamine coated PVDF nanofibrous membranes. Dopamine self-polymerizes in the aqueous solution at pH 8.5 and the resultant polydopamine coated surface becomes hydrophilic. Electrochemical characterization of the Li/separator/LiMn₂O₄ cells using the PVDF nanofibrous membranes with and without polydopamine coating as separators. (b) Cycling stability and coulombic efficiency with different separators at 0.5C. (c) Discharge capacity profiles of the cells at various rates. Reproduced with permission from ref. 124. Copyright 2014 Elsevier.

Summery and outlook

Separators play an important role in LIBs. The primary function of the separator is to insulate the anode and cathode to avoid the internal short-circuit of LIBs. The ideal separator must have excellent physical and chemical properties, such as a larger porous structure with greater tortuosity to prevent the growth of dendritic lithium, thinner thickness, higher mechanical strength, higher electrolyte uptake, a reservoir to allow the transfer of lithium ions, lower internal resistance and good thermal stability. In order to satisfy the different applications of high performance LIBs, it is important to adjust the structure and properties of the separators.

Currently, the most common single separators are polyolefin-based separators, which were prepared mainly using two different manufacturing methods: a wet process and dry process. Some polyolefin separators have been industrialized, such as PP, PE and PP/PE/PP, but their physical and chemical properties still need to be further improved, such as the cycling capability, wettability and thermal stability for high performance LIBs. Other microporous separators, such as PVDF, PVDF–HFP, PAN and MC, can be prepared using a simple casting method and phase inversion technique. These separators can meet the fundamental requirements of LIBs, as they possess certain porosity, electrolyte uptake, thermal stability and electrochemical performances. However, the poor ionic conductivity, lower charge/discharge capacity and high crystallinity hinder their application in higher performance LIBs. The other most common separators, electrospun separators, have also been used in LIBs. The electrospun separators possess many advantages, such as high porosity, high electrolyte uptake and large surface area. However, one fateful shortcoming is their poor mechanical stability. As mentioned above, there is no single component separator that can meet all the requirements of the different battery applications.

In order to address these issues, some researchers proposed introducing inorganic particles (such as SiO₂, Al₂O₃, TiO₂ and ZrO₂) into a polymer matrix to fabricate composite membranes using appropriate methods. This is because the inorganic particles have high surface areas, high mechanical stability and good thermal stability. The modified separators can be divided into three main types: inorganic particle-coated composite membranes, inorganic particle-filled composite membranes and inorganic particle-filled non-woven composite membranes.

Recent studies have shown that the performances of separators are greatly improved by introducing inorganic nanoparticles with high surface areas and absorbing ability. These inorganic nanoparticles can improve the mechanical stability and reduce crystallinity, which is beneficial to the transportation of lithium ions. All of these result in improving effectively the physical and electrochemical performances of separators. So, the modified separators exhibit higher porosity, higher electrolyte uptake, stronger mechanical strength, higher ionic conductivity and better thermal stability than pristine separators. Furthermore, compared with the single component separators, these separators have better rate capability and cycling performance when they are assembled into half-cells or all-cells. Some of them even have the perfect effect of increasing the flame retardancy at higher temperatures. Thus, the excellent performance of modified separators makes it possible to apply them in high-power and high-density LIBs and to have a better market.

However, although introducing inorganic particles into a polymer matrix or on the surface of separators is an effective way to obtain high-performance LIB separators with excellent physical and electrochemical properties, the introduction of inorganic particles can also affect the performances of the separators and result in other problems. For example, inorganic coating layers increased the thickness and the internal resistance of separators, which reduced the porosity of the separators and limited their application in LIBs, even though the inorganic particle-coated composite membranes possess the above superior performances. The results also indicated that the interfacial bonding between the inorganic particles and polymer matrix is poor in the inorganic particle-filled composite membranes and inorganic particle-filled non-woven composite membranes due to the different physical and chemical properties of inorganic and organic materials. Based on these problems, this paper provides some suggestions to promote the further development of composite membranes. Firstly, the study and understanding of the mechanism of interfacial bonding between the inorganic particles and polymer matrix should be strengthened. Next, the inorganic fibre-filled composite membranes should be developed so as to reduce the thickness and the internal resistance of separators and increase the permeability of separators by forming a nanoporous network structure. Thirdly, how the surface structures and features of the inorganic particles affect the pore structure, permeability and mechanical performance of composite membranes should be investigated. Finally, the development of new simple preparation methods, that are low cost, high-yielding and safe, is important for the business applications of composite membranes. By comparing the three composite membranes modified with inorganic particles mentioned above (Table 1), it is found that the inorganic particle-filled composite membranes prepared using the phase inversion method have more outstanding properties, a simple preparation process and lower cost. These composite membranes could effectively improve the electrochemical performances when used in LIBs because their highly porous structure has good liquid electrolyte permeability.

Table 1 Summarization of composite membranes

Substrate	Inorganic particle	Binder	Preparation method	Membrane thickness	Liquid electrolyte	Ionic conductivity	Cathode, anode	Electrochemical performance	References
PE	SiO2	PMMA	Dip coating	28 μm	1 M LiPF6–EC/DEC	7.4 × 10^−4 S cm^−1	LiCoO2, Li	In 3.0–4.3 V, an initial discharge capacity of 158, 140, 122 and 70 mA h g^−1 at 0.2C, 0.5C, 1.0C and 2.0C, respectively	53
PE	SiO2	SBR, CMC	Coating by automatic machine	29 μm	1 M LiPF6–EC/DEC/DMC	0.82 mS cm^−1	LiMn2O4, Li	In 3.0–4.3 V, a discharge capacity of 102.5 mA h g^−1 after 100 cycles at 0.5C	57
PE	SiO2	PVDF–HFP	Dip coating	30 μm	1 M LiPF6–EC/DEC	0.53–0.61 S	LiCoO2, graphite	In 3.0–4.4 V, an initial discharge capacity of 153, 150 and 111 mA h g^−1 at 0.2C, 0.5C and 1.0C, respectively, a discharge capacity of 112 mA h g^−1 after 200 cycles at 0.5C	54
PE	Al2O3	PVDF–HFP	Dip coating	30 μm	1 M LiPF6–EC/DEC	0.34–0.53 mS cm^−1	LiCoO2, Li	In 3.0–4.3 V, an initial discharge capacity of 155, 149 and 90 mA h g^−1 at 0.2C, 1.0C and 2.0C, respectively	58
PE	Al2O3	PVDF–HFP	Dip coating,	30–43 μm	1 M LiPF6–EC/DEC	0.495–0.719 mS cm^−1	LiCoO2, Li	In 3.0–4.3 V, an initial discharge capacity of 154, 148, and 88 mA h g^−1 at 0.2C, 1.0C and 2.0C, respectively (composite membrane with the system of acetone–water)	59
PE	SiO2	—	Chemical vapor deposition	19 μm	1 M LiPF6–EC/DEC/EMC	7.60 × 10^−4 to 1.00 × 10^−3 S cm^−1	LiCoO2, mesocarbon microbeads	In 3.0–4.2 V, an initial discharge capacity of 128, 126, 120.5 and 60 mA h g^−1 at 0.2C, 0.5C, 1.0C, 1.5C and 3C, respectively	60
PP/PE/PP	TiO2	—	Microwave	—	1 M LiPF6–EC/PC/DMC	0.40–0.80 mS cm^−1	Li4Ti5O12, Li	In 1.0–3.0 V, an initial discharge capacity of 173 and 167 mA h g^−1 at 0.1C and 0.2C, respectively	61
PMMA	Al2O3	PVDF–HFP	Dip coating	30 μm	1 M LiPF6–EC/DEC/DMC	5.35 × 10^−4 S cm^−1	LiCoO2, mesocarbon microbeads	In 3.0–4.2 V, an initial discharge capacity of 8.0, 7.2, 6 and 4 mA h at 0.2C, 1.0C, 2.0C and 3C, respectively	65
PEI–PU	SiO2	PVDF–HFP	Dip coating	35 μm	1 M LiPF6–EC/DMC/EMC	2.3 mS cm^−1	LiFePO4, Li	In 2.5–4.0 V, an initial discharge capacity of 164.68, 163.98, 159.87 and 142.57 mA h g^−1 at 0.1C, 0.2C, 0.5C and 1.0C, respectively, and a discharge capacity of 163.25 mA h g^−1 after 50 cycles at 0.2C	66
PET	SiO2	PVDF–HFP	Dip coating	30 μm	1 M LiPF6–EC/DEC	0.91 mS cm^−1	LiCoO2, graphite	In 3.0–4.2 V, an initial discharge capacity of 135, 130, 125 and 68 mA h g^−1 at 0.2C, 0.5C, 1.0C and 2.0C, respectively, and a discharge capacity of 101 mA h g^−1 after 100 cycles at 0.1C	55
PVDF–HFP	TiO2	—	Nonsolvent evaporate	—	1 M LiPF6–EC/DEC	1.5 × 10^−3 S cm^−1	LiFePO4, Li	In 2.5–4.2 V, an initial discharge capacity of 164, 157, 143, and 122 mA h g^−1 at 0.1C, 0.5C, 1.0C, and 3.0C, respectively	70
PVDF–HFP	SiO2	—	Phase inversion	50–75 μm	1 M LiPF6–EC/DMC	0.9–3.1 mS cm^−1	LiMn2O4, C	Good discharge rate capability	72
PVDF–HFP	ZrO2	—	Phase inversion	50–80 μm	1 M LiClO4–EC/DEC	1.1 × 10^−2 S cm^−1	LiMg0.10Mn1.90O4, C	In 3.0–4.5 V, an initial discharge capacity of 135.0 mA h g^−1 at 0.1C, and a capacity fading of 2.7% at the 25th cycle	73
PVDF–HFP	Al2O3	—	Phase inversion	—	1 M LiClO4–EC/DEC	2.11 × 10^−3 S cm^−1	—	Good ion transport	75
PVDF–HFP/PVA blend	LiAlO2	—	Phase inversion	50–80 μm	1 M LiClO4–EC/DEC	8.12 × 10^−3 S cm^−1	LiCoO2, C	In 3.0–4.2 V, a discharge capacity of 148.0 and 142.5 on the first and 25th cycle at 0.25 mA cm^−2, respectively	83
PVDF–HFP/PS blend	SiO2	—	Casting	—	1 M LiPF6–EC/DEC/DMC	4 mS cm^−1	—	Increased porosity and better ionic conductivity	84
PVDF	TiO2	—	Electrospinning	30 μm	1 M LiPF6–EC/DMC	1.4 × 10^−3 S cm^−1	LiFePO4, Li	In 2.4–4.1 V, an initial discharge capacity of 148 mA h g^−1 at 0.1C	104
PVDF	SiO2	—	Electrospinning	—	1 M LiPF6–EC/EMC	1.7–2.6 mS cm^−1	LiFePO4, Li	In 2.5–4.2 V, a discharge capacity of the initial and 100th cycle of 162 mA h g^−1 and 132 mA h g^−1 at 0.2C and 1.0C, respectively	105
PVDF–HFP	SiO2	—	Electrospinning	100–120 μm	LIFSFI–BMITFSI	2.3 × 10^−3 to 4.3 × 10^−3 S cm^−1	LiFePO4, Li	In 2.5–4.0 V, an initial discharge capacity of 169 mA h g^−1 at 0.1C, and better cycling capability	107
PVDF–HFP	SiO2, Al2O3, BaCO3	—	Electrospinning	150 μm	1 M LiPF6–EC/DMC	5.9 × 10^−3 to 7.2 × 10^−3 S cm^−1	LiFePO4, Li	In 2.5–4.0 V, a discharge capacity of the initial and 30th cycle of 164 mA h g^−1 and 156 mA h g^−1 at 0.1C, respectively	108
PAN	LLTO	—	Electrospinning	—	1 M LiPF6–EC/EMC	2.0 × 10^−3 S cm^−1	LiFePO4, Li	In 2.5–4.2 V, a discharge capacity of the initial and 50th cycle of 162 mA h g^−1, and 156 mA h g^−1 at 0.2C, respectively	112
PAN	SiO2	—	Electrospinning	—	1 M LiPF6–EC/DMC	1.1 × 10^−2 S cm^−1	LiCoO2, graphite	In 2.7–4.2 V, a discharge capacity of the initial and 150th cycle of 139 mA h g^−1 and 127 mA h g^−1 at 0.1C, respectively	97
PAN	SiO2	—	Centrifugal spinning	—	1 M LiPF6–EC/EMC	3.0 × 10^−3 to 3.6 × 10^−3 S cm^−1	LiFePO4, Li	In 2.5–4.2 V, an initial discharge capacity of 163 mA h g^−1, and 85 mA h g^−1 at 0.2C and 8C, respectively, and better cycling capacity	115

---

The batteries assembled using the above-mentioned three composite membranes as separators exhibit excellent rate capability, discharge capacity and cycling performance. However, as the organic solvents in these batteries have high vapor pressures, it is easy to cause short circuits and thermal runaway in the batteries. These are all very dangerous things because they can easily lead to fires and explosions. To address this issue, many researchers have begun to pay more attention to electrolyte membranes, which are ionically conductive and can act as both separators and electrolyte within LIBs. However there still exist many problems that need to be solved to improve the mechanical strength and ionic conductivity of these separators.

With the development of industrialization, there are higher requirements on LIBs, and the performance of the separators also needs to be improved. According to the above-mentioned results, modified composite membranes are potential candidates for high-performance LIBs. Of course, it is necessary to continue improving the performance of composite membranes. In the future, new separator structures and techniques are needed to achieve advanced LIBs with excellent electrochemical performances.

Acknowledgements

The authors thank the Natural Science Foundation of China (Grant No. 51272144, 51472127 and 51172132) for the financial support. They also thank the Projects Supported by the State Key Laboratory of Microbial Technology (No. M2013-20) and the Key Laboratory of Pulp and Paper Science and Technology of Ministry of Education (No. KF2014-22).

Notes and references

1. Y. S. Zhu, F. X. Wang, L. L. Liu, S. X. Xiao, Z. Chang and Y. P. Wu, Energy Environ. Sci., 2013, 6, 618–624.
2. J. Lee, C. L. Lee, K. Park and I. D. Kim, J. Power Sources, 2014, 248, 1211–1217.
3. Y. Li, G. Xu, Y. Yao, L. Xue, M. Yanilmaz, H. Lee and X. Zhang, Solid State Ionics, 2014, 258, 67–73.
4. M. Dirican, M. Yanilmaz, K. Fu, Y. Lu, H. Kizil and X. Zhang, J. Power Sources, 2014, 264, 240–247.
5. K. Fu, Y. Lu, M. Dirican, C. Chen, M. Yanilmaz, Q. Shi, P. D. Bradford and X. Zhang, Nanoscale, 2014, 6, 7489–7495.
6. M. Li, X. Wang, Y. Wang, B. Chen, Y. Wu and R. Holze, RSC Adv., 2015, 5, 52382–52387.
7. X. Liang, Y. Yang, X. Jin, Z. Huang and F. Kang, J. Membr. Sci., 2015, 493, 1–7.
8. P. Yang, P. Zhang, C. Shi, L. Chen, J. Dai and J. Zhao, J. Membr. Sci., 2015, 474, 148–155.
9. W. Xiao, L. Zhao, Y. Gong, J. Liu and C. Yan, J. Membr. Sci., 2015, 487, 221–228.
10. C. H. Tsao and P. L. Kuo, J. Membr. Sci., 2015, 489, 36–42.
11. Y. Liang, S. Cheng, J. Zhao, C. Zhang, S. Sun, N. Zhou, Y. Qiu and X. Zhang, J. Power Sources, 2013, 240, 204–211.
12. X. Zhang, L. Ji, O. Toprakci, Y. Liang and M. Alcoutlabi, Polym. Rev., 2011, 51, 239–264.
13. Y. Liang, Z. Lin, Y. Qiu and X. Zhang, Electrochim. Acta, 2011, 56, 6474–6480.
14. H. Lee, M. Alcoutlabi, J. V. Watson and X. Zhang, J. Appl. Polym. Sci., 2013, 129, 1039–1951.
15. C. J. Orendorff, T. N. Lambert, C. A. Chavez, M. Bencomo and K. R. Fenton, Adv. Energy Mater., 2012, 3, 314–320.
16. K. Xu, Chem. Rev., 2004, 104, 4303–4418.
17. H. S. Jeong, S. C. Hong and S. Y. Lee, J. Membr. Sci., 2010, 364, 177–182.
18. J. K. Kim, L. Niedzicki, J. Scheers, C. R. Shin, D. H. Lim, W. Wieczorek, P. Johansson, J. H. Ahn, A. Matic and P. Jacobsson, J. Power Sources, 2013, 224, 93–98.
19. S. Kataria, S. K. Chaurasia, R. K. Singh and S. Chandra, J. Phys. Chem. B, 2013, 117, 897–906.
20. J. Pitawala, M. A. Navarra, B. Scrosati, P. Jacobsson and A. Matic, J. Power Sources, 2014, 245, 830–835.
21. A. Manuel Stephan, Eur. Polym. J., 2006, 42, 21–42.
22. T. H. Cho, M. Tanaka, H. Onishi, Y. Kondo, T. Nakamura, H. Yamazaki, S. Tanase and T. Sakai, J. Power Sources, 2008, 181, 155–160.
23. J. Hao, G. Lei, Z. Li, L. Wu, Q. Xiao and L. Wang, J. Membr. Sci., 2013, 428, 11–16.
24. E. Quartarone, M. Brusa, P. Mustarelli, C. Tomasi and A. Magistris, Electrochim. Acta, 1998, 44, 677–681.
25. A. S. Gozdz, C. N. Schmutz, J. M. Tarascon and P. C. Warren, US Pat., 5,456,000, 1995.
26. H. Strathmann and K. Kock, Desalination, 1997, 21, 241–255.
27. K. M. Kim, N. G. Park, K. S. Ryu and S. H. Chang, Polymer, 2002, 43, 3951–3957.
28. H. Lee, M. Alcoutlabi, J. V. Watson and X. Zhang, J. Polym. Sci., Part B: Polym. Phys., 2013, 51, 349–357.
29. M. Alcoutlabi, H. Lee, J. V. Watson and X. Zhang, J. Mater. Sci., 2013, 48, 2690–2700.
30. T. H. Cho, M. Tanaka, H. Ohnishi, Y. Kondo, M. Yoshikazu, T. Nakamura and T. Sakai, J. Power Sources, 2010, 195, 4272–4277.
31. P. Arora and Z. Zhang, Chem. Rev., 2004, 104, 4419–4462.
32. K. Gao, X. Hu, C. Dai and T. Yi, Mater. Sci. Eng., B, 2006, 131, 100–105.
33. P. Carol, P. Ramakrishnan, B. John and G. Cheruvally, J. Power Sources, 2011, 196, 10156–10162.
34. Y. Ding, P. Zhang, Z. Long, Y. Jing, F. Xu and W. Di, J. Membr. Sci., 2009, 329, 56–59.
35. A. L. Gopalan, P. Santhosh, K. M. Manesh, J. H. Nho, S. H. Kim, C. G. Hwang and K. P. Lee, J. Membr. Sci., 2008, 325, 683–690.
36. L. Lu, X. Han, J. Li, J. Hua and M. Quyang, J. Power Sources, 2013, 226, 272–288.
37. P. Raghavan, J. Manuel, X. Zhao, D. S. Kim, J. H. Ahn and C. Nah, J. Power Sources, 2011, 196, 6742–6749.
38. M. M. Atabaki and R. Kovacevic, Electron. Mater. Lett., 2013, 9, 133–153.
39. G. L. Ji, B. K. Zhu, Z. Y. Cui, C. F. Zhang and Y. Y. Xu, Polymer, 2007, 48, 6415–6425.
40. S. W. Lee, S. W. Choi, S. M. Jo, B. D. Chin, D. Y. Kim and K. Y. Lee, J. Power Sources, 2006, 163, 41–46.
41. K. S. Kim, S. Y. Park, S. Choi and H. Lee, J. Power Sources, 2006, 155, 385–390.
42. A. M. Stephan, S. G. Kumar, N. G. Renganathan and M. A. Kulandainathan, Eur. Polym. J., 2005, 41, 15–21.
43. D. Bansal, B. Meyer and M. Salomon, J. Power Sources, 2008, 178, 848–851.
44. H. R. Jung, D. H. Ju, W. J. Lee, X. Zhang and R. Kotek, Electrochim. Acta, 2009, 54, 3630–3637.
45. Y. S. Lee, Y. B. Jeong and D. W. Kim, J. Power Sources, 2010, 195, 6197–6201.
46. Y. Liu, J. Y. Lee and L. Hong, J. Power Sources, 2004, 129, 303–311.
47. W. Xiao, Y. Gong, H. Wang, L. Zhao, J. Liu and C. Yan, Ceram. Int., 2015, 41, 14223–14229.
48. H. Lee, M. Yanilmaz, O. Toprakci, K. Fu and X. Zhang, Energy Environ. Sci., 2014, 7, 3857–3886.
49. M. Y. An, H. T. Kim and D. R. Chang, J. Solid State Electrochem., 2014, 18, 1807–1814.
50. S. H. Chung, Y. Wang, L. Persi, F. Croce, S. G. Greenbaum, B. Scrosati and E. Plichta, J. Power Sources, 2001, 97/98, 644–648.
51. W. K. Shin and D. W. Kim, J. Power Sources, 2013, 226, 54–60.
52. J. A. Choi, S. H. Kim and D. W. Kim, J. Power Sources, 2010, 195, 6192–6196.
53. J. H. Park, J. H. Cho, W. Park, D. Ryoo, S. J. Yoon, J. H. Kim, Y. U. Jeong and S. Y. Lee, J. Power Sources, 2010, 195, 8306–8310.
54. H. S. Jeong and S. Y. Lee, J. Power Sources, 2011, 196, 6716–6722.
55. E. S. Choi and S. Y. Lee, J. Mater. Chem., 2011, 21, 14747–14754.
56. J. Wang, Z. Hu, X. Yin, Y. Li, H. Huo, J. Zhou and L. Li, Electrochim. Acta, 2015, 159, 61–65.
57. P. Zhang, L. Chen, C. Shi, P. Yang and J. Zhao, J. Power Sources, 2015, 284, 10–15.
58. H. S. Jeong, D. W. Kim, Y. U. Jeong and S. Y. Lee, J. Power Sources, 2010, 195, 6116–6121.
59. H. S. Jeong, S. C. Hong and S. Y. Lee, J. Membr. Sci., 2010, 364, 177–182.
60. M. Kim and J. H. Park, J. Power Sources, 2012, 212, 22–27.
61. R. S. Juang, C. T. Hsieh, P. A. Chen and Y. F. Chen, J. Power Sources, 2015, 286, 526–533.
62. X. Li, J. He, D. Wu, M. Zhang, J. Meng and P. Ni, Electrochim. Acta, 2015, 167, 396–403.
63. T. H. Cho, M. Tanaka, H. Onishi, Y. Kondo, T. Nakamura, H. Yamazaki, S. Tanase and T. Sakai, J. Electrochem. Soc., 2008, 155, A699–A703.
64. S. M. Eo, E. Cha and D. W. Kim, J. Power Sources, 2009, 189, 766–770.
65. M. Kim, G. Y. Han, K. J. Yoon and J. H. Park, J. Power Sources, 2010, 195, 8302–8305.
66. Y. Zhai, K. Xiao, J. Yu and B. Ding, Electrochim. Acta, 2015, 154, 219–226.
67. D. Takemura, S. Aihara, K. Hamano, M. Kise, T. Nishimura, H. Urushibata and H. Yoshiyasu, J. Power Sources, 2005, 146, 779–783.
68. D. W. Kim, J. Power Sources, 2000, 87, 78–83.
69. P. P. Prosini, P. Villano and M. Carewska, Electrochim. Acta, 2002, 48, 227–233.
70. P. Yan, Z. Huang, Y. Lin, X. Wu, Y. Yang, D. Wang, F. Chen, C. Zhang and D. He, Ionics, 2015, 21, 593–599.
71. K. M. Kim, N. G. Park, K. S. Ryu and S. H. Chang, Electrochim. Acta, 2006, 51, 5636–5644.
72. A. D. Pasquier, P. C. Warren, D. Culver, A. S. Gozdz, G. G. Amatucci and J. M. Tarascon, Solid State Ionics, 2000, 135, 249–257.
73. A. Subramania, N. T. Kalyana Sundaram, A. R. Sathiya Priya and G. Vijaya Kumar, J. Membr. Sci., 2007, 294, 8–15.
74. R. Miao, B. Liu, Z. Zhu, Y. Liu, J. Li, X. Wang and Q. Li, J. Power Sources, 2008, 184, 420–426.
75. Z. Li, G. Su, X. Wang and D. Gao, Solid State Ionics, 2005, 176, 1903–1908.
76. Y. Xia, Y. Zhou and M. Yoshio, J. Electrochem. Soc., 1997, 144, 2593–2600.
77. D. H. Jang, Y. J. Shin and S. M. Oh, J. Electrochem. Soc., 1996, 143, 2204–2211.
78. S. S. Zhang, K. Xu and T. R. Jow, J. Power Sources, 2005, 140, 361–364.
79. C. H. Kim, H. T. Kim, J. K. Park, S. I. Moon and M. S. Yoon, J. Polym. Sci., Part B: Polym. Phys., 1996, 34, 2709–2714.
80. Z. Y. Cui, Y. Y. Xu, L. P. Zhu, J. Y. Wang, Z. Y. Xi and B. K. Zhu, J. Membr. Sci., 2008, 325, 957–963.
81. S. Rajendran, R. Kannan and O. Mahendran, J. Power Sources, 2001, 96, 406–410.
82. N. S. Choi, Y. G. Lee, J. K. Park and J. M. Ko, Electrochim. Acta, 2001, 46, 1581–1586.
83. N. T. K. Sundaram and A. Subramania, Electrochim. Acta, 2007, 52, 4987–4993.
84. H. Huang and S. L. Wunder, J. Power Sources, 2001, 97–98, 649–653.
85. W. Wieczorek, K. Such and J. R. Steven, Electrochim. Acta, 1995, 40, 2251–2258.
86. B. Kumar, S. J. Rodrigues and S. Koka, Electrochim. Acta, 2002, 47, 4125–4131.
87. F. Croce, L. Persi, B. Scrosati, F. Serraino-Fiory, E. Plichta and M. A. Hendrickson, Electrochim. Acta, 2001, 46, 2457–2461.
88. W. Wieczorek, P. Lipka, G. Zukowska and H. Wycislik, J. Phys. Chem. B, 1998, 102, 6968–6974.
89. C. Yang, Z. Jia, Z. Guan and L. Wang, J. Power Sources, 2009, 189, 716–720.
90. L. Zhao, H. Zhang, X. Li, J. Zhao, C. Zhao and X. Yuan, J. Appl. Polym. Sci., 2009, 111, 3104–3112.
91. P. Raghavan, X. Zhao, J. K. Kim, J. Manuel, G. S. Chauhan, J. H. Ahn and C. Nah, Electrochim. Acta, 2008, 54, 228–234.
92. F. Croce, S. Sacchetti and B. Scrosati, J. Power Sources, 2006, 162, 685–689.
93. M. Yanilmaz, F. Kalaoglu and H. Karakas, Fibres Text. East. Eur., 2013, 21, 19–21.
94. M. Yanilmaz, F. Kalaoglu and H. Karakas, Res. J. Text. Apparel, 2012, 22, 212–217.
95. M. Yanilmaz, F. Kalaoglu, H. Karakas and A. S. Sarac, J. Appl. Polym. Sci., 2012, 125, 4100–4108.
96. M. Yanilmaz and A. S. Sarac, Text. Res. J., 2014, 84, 1325–1342.
97. H. R. Jung, D. H. Ju, W. J. Lee, X. Zhang and R. Kotek, Electrochim. Acta, 2009, 54, 3630–3637.
98. Y. Liang, Z. Lin, Y. Qiu and X. Zhang, Electrochim. Acta, 2011, 56, 6474–6480.
99. X. Zhang, L. Ji, O. Toprakci, Y. Liang and M. Alcoutlabi, Polym. Rev., 2011, 51, 239–264.
100. S. H. Chung, Y. Wang, L. Persi, F. Croce, S. G. Greenbaum, B. Scrosati and E. Plichta, J. Power Sources, 2001, 97–98, 644–648.
101. F. Croce, G. B. Appetecchi, L. Persi and B. Scrosati, Nature, 1998, 394, 456–458.
102. G. Jiang, S. Maeda, H. Yang, Y. Saito, S. Tanase and T. Sakai, J. Power Sources, 2005, 141, 143–148.
103. M. Wachtler, D. Ostrovskii, P. Jacobsson and B. Scrosati, Electrochim. Acta, 2004, 50, 357–361.
104. Y. Ding, P. Zhang, Z. Long, Y. Jiang, F. Xu and W. Di, Sci. Technol. Adv. Mater., 2008, 9, 015005.
105. M. Yanilmaz, Y. Lu, M. Dirican, K. Fu and X. Zhang, J. Membr. Sci., 2014, 456, 57–65.
106. P. Raghavan, J. W. Choi, J. H. Ahn, G. Cheruvally, G. S. Chauhan, H. J. Ahn and C. Nah, J. Power Sources, 2008, 184, 437–443.
107. J. K. Kim, G. Cheruvally, X. Li, J. H. Ahn, K. W. Kim and H. J. Ahn, J. Power Sources, 2008, 178, 815–820.
108. P. Raghavan, X. Zhao, J. K. Kim, J. Manuel, G. S. Chauhan, J. H. Ahn and C. Nah, Electrochim. Acta, 2008, 54, 228–234.
109. G. B. Appetecchi, P. Romagnoli and B. Scrosati, Electrochem. Commun., 2001, 3, 281–284.
110. P. P. Chu and Z. P. He, Polymer, 2001, 42, 4743–4749.
111. T. H. Cho, T. Sakai, S. Tanase, K. Kimura, Y. Kondo, T. Tarao and M. Tanaka, Electrochem. Solid-State Lett., 2007, 10, A159–A162.
112. Y. Liang, L. Ji, B. Guo, Z. Lin, Y. Yao, Y. Li, M. Alcoutlabi, Y. Qiu and X. Zhang, J. Power Sources, 2011, 196, 436–441.
113. Y. Lu, Y. Li, S. Zhang, G. Xu, K. Fu, H. Lee and X. Zhang, Eur. Polym. J., 2013, 49, 3834–3845.
114. L. A. Mary, T. Senthilram, S. Suganya, L. Nagarajan, J. Venugopal, S. Ramakrishna and V. Dev, eXPRESS Polym. Lett., 2013, 7, 238–248.
115. M. Yanilmaz, Y. Lu, Y. Li and X. Zhang, J. Power Sources, 2015, 273, 1114–1119.
116. M. Yanilmaz, C. Chen and X. Zhang, J. Polym. Sci., Part B: Polym. Phys., 2013, 51, 1719–1726.
117. F. Croce and B. Scrosati, Ann. N. Y. Acad. Sci., 2003, 984, 194–207.
118. J. Xi, X. Qiu, M. Cui, X. Tang, W. Zhu and L. Chen, J. Power Sources, 2006, 156, 581–588.
119. D. Shanmukaraj and R. Murugan, J. Power Sources, 2005, 149, 90–95.
120. M. Sethupathy, V. Sethuraman and P. Manisankar, Soft Nanosci. Lett., 2013, 3, 37–43.
121. J. Zhang, H. Han, S. Wu, S. Xu, Y. Yang, C. Zhou and X. Zhao, Solid State Ionics, 2007, 178, 1595–1605.
122. S. Y. Xiao, Y. Q. Yang, M. X. Li, F. X. Wang, Z. Chang, Y. P. Wu and X. Liu, J. Power Sources, 2014, 270, 53–58.
123. G. Ding, B. Qin, Z. Liu, J. Zhang, B. Zhang, P. Hu, C. Zhang, G. Xu, J. Yao and G. Cui, J. Electrochem. Soc., 2015, 162, A834–A838.
124. C. Cao, L. Tan, W. Liu, J. Ma and L. Li, J. Power Sources, 2014, 248, 224–229.

---