Strategic Structural Design of a Gel Polymer Electrolyte toward a High Efficiency Lithium-Ion Battery

Febri BaskoroFebri Baskoro
Institute of Chemistry, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei 11529, Taiwan
Sustainable Chemical Science and Technology, Taiwan International Graduate Program, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei 11529, Taiwan
Department of Applied Chemistry, National Chiao Tung University, Hsinchu 300, Taiwan
More by Febri Baskoro
http://orcid.org/0000-0001-8408-6726, Hui Qi Wong, and Hung-Ju Yen*
Cite this: ACS Appl. Energy Mater. 2019, 2, 6, 3937–3971
Publication Date:May 2, 2019

Abstract

Electrolytes have played critical roles in electrochemical energy storage. In Li-ion battery, liquid electrolytes have shown their excellent performances over decades, such as high ionic conductivity (∼10^–3 S cm^–1) and good contacts with electrodes. However, the use of liquid electrolytes often brought risks associated with leakage and combustion of organic electrolytes. Hence, polymer electrolytes become potential candidates to replace liquid electrolyte systems. Although solid polymer electrolytes (SPEs) offer better safety and good mechanical properties to take over liquid electrolytes, most of them only deliver low ionic conductivities (∼10^–8 S cm^–1) and poor contact with electrodes, resulting in poor cycle performance and low electrical capacity of the batteries. In addition, gel polymer electrolytes (GPEs) have received increasing research attention due to their relevant characteristics, which extend from liquid electrolytes and solid polymer electrolytes. In this review, state-of-the-art samples of gel polymer electrolytes are elucidated with respect to their structural design and electrochemical properties to determine their application potential in Li-ion batteries (LIBs). First, we present the general requirements of GPEs for LIBs applications, followed by important electrochemical properties of GPEs for LIBs including ionic conductivity, transference number, and ionic transport mechanisms. Furthermore, recent progress of common polymers, namely, polyether, polyvinyl, polynitrile, polycarbonate, and polyacrylate, as polymer host of GPEs has been carefully explained. Finally, the alternative polymers were also discussed to provide new approaches for further developments of GPEs to fulfill the demanded properties for practical applications.

1. Introduction

1.1. Overview of Lithium-Ion Battery Development

Because of the continuous increase of environmental problems related to greenhouse gases in the atmosphere and the exhaustion of nonrenewable fossil fuels, the use of alternative renewable energy sources, such as wind, solar, and geothermal, etc., has become an important research focus. However, availability issues limit the production of energy from certain renewable energy sources, such as wind and solar energy. Therefore, an efficient energy storage system is absolutely needed to provide a continuous energy supply. Among energy storage technologies, electrochemical energy storage based on batteries shows promising features that include high round-trip efficiency, flexible power, a long cycle life, low maintenance, and energy characteristics to meet different grid functions. (1) Li-ion batteries (LIBs) are a type of electrochemical energy storage based on the use of Li intercalating compounds, and they have attracted considerable research attention over the past 2 decades. LIBs were commercialized in the 1990s by Sony Corp., and these batteries have recently dominated the entire market of portable electronic devices, such as notebooks, personal telephones, and cameras. Moreover, the use of LIBs was projected for electric vehicles due to the desirable features, including their lightweight, high energy density, high open-circuit potential, minimal memory effects, fast charging, low self-discharge rate, and environmental friendliness. (2−6) As shown in Figure 1, a conventional LIB consists of a positive electrode (cathode), negative electrode (anode), nonaqueous electrolyte system, and separator to prevent physical contact between two electrodes. In principal, when the battery is being charged, the Li ion moves from the cathode to the anode through the electrolyte and carries the current, and during discharge, the Li ion will move back from the anode to the cathode. Generally, Li-intercalated materials, such as LiCoO₂ (a lithium metal oxide with layered structure), have been used as the cathode material (lithium source) and graphite has been used as the anode to hold the Li ion in its layers. Both materials for the electrodes can reversibly insert and remove the Li ion from its respective structure. (3)

For the past 2 decades, improvements have been made to fulfill the demands of energy storage systems for portable devices and electric vehicles. Inside the LIB, the chosen electrodes (mainly cathode) are the main factors for battery performance, i.e., energy density and cyclability. Meanwhile the electrolytes do determine the current (power) density, time stability, and battery safety since they directly interact with all battery components including anode, cathode, and the separator. (1) As shown in Figure 2, the development of cathode materials has been focused on high voltage materials, such as Li-rich cathode materials (Li_1–xMn_2–yM_yO₄), which are capable of reaching over 4 V of operating voltage. In addition, research on anode materials has focused on higher capacity materials, including carbon materials and 3D metal oxides, capable of reaching higher capacities (∼1000 Ah/kg). In addition, a safety issue is associated with the use of flammable organic solvents, and the leakage of electrolytes has hindered the commercialization of LIBs based on a liquid electrolyte. (1,2,7−10) Organic solvents can generate highly flammable gases, such as CO, CH₄, C₂H₄, C₂H₆, C₃H₆, and C₃H₈, etc., when LIBs are overheated due to overcharging, abuse, internal short circuits, manufacturing defects, physical damage, or other failure mechanisms. (8,11) Furthermore, the electrodes–electrolyte compatibility issue that has also arisen from the use of high voltage materials (mainly cathode) to increase the energy density of LIBs will lead to the decomposition of electrolytes because of the poor electrochemical stability at high voltage. (12,13) Since the electrode–electrolytes compatibility mainly lies on the electrochemical window of the electrolytes, (7) therefore, the compatibility between electrolytes and the high voltage cathode materials became a major direction and cannot be ignored in the electrolyte development. (4) A list of accidents related to LIBs has shown that most of the accidents were caused by overheating during operation, which leads to fire. (8,14,15) For instance, in January 2016, a Tesla Model S caught fire while fast-charging at a supercharger station in Gjerstad, Norway due to a short circuit. (15) Therefore, the replacement of liquid electrolyte in LIBs is in high demand for the safety issues.

Solid electrolytes, including inorganic solid electrolytes (such as ceramic materials) and polymer electrolytes, have attracted attention as a replacement for liquid electrolytes. Moreover, among these types of solid electrolytes, polymer electrolytes have drawn considerable attention as a replacement for liquid electrolytes due to their intrinsic properties, such as freestanding, shape versatility, security, flexibility, lightweight, and reliability. (2,11,13) In general, there are two types of polymer electrolytes: solvent-free solid polymer electrolytes (SPEs) and gel polymer electrolytes (GPEs). However, the boundary between SPEs and GPEs can become blurred due to the presence of “plasticizers” in the SPEs, which allow them to behave similarly to GPEs; moreover, polymers or oligomers may present molecular weights near the threshold (average M_w 5,000–10,000) of polymers. (12)

1.2. Solid Polymer Electrolytes

Generally, SPEs are constructed by the dissolution of lithium salts in a polymeric matrix to provide ionic conductivity. In SPEs, lithium salts are dissolved via complexation between the Li ion and the strong electron donor of the polymer backbone, in which the electron donor panel in the polymer is involved in solvation with the cation component in the dopant salt and then facilitates ion separation, which leads to an ionic hopping mechanism that generates ionic conductivity. (2,16,17) The development of SPEs began with the discovery of ionic conductivity of ions (alkali metal) inside poly(ethylene oxide) (PEO) in 1973 by Wright. (13) Since that time, the number of research works in the field of SPEs has grown tremendously, thus reflecting the progress of development and understanding of the molecular architecture, which is necessary to achieve fast ion transport in polymers. There are several advantages of SPEs over liquid electrolytes in LIBs, such as nonvolatility, flexibility, mechanical properties, improved safety, and ease of device fabrication, etc. (17−19) As shown in Scheme 1, many existing polymer backbones have been used as SPEs for LIB applications. (20) Among these polymers, PEO is the most promising candidate as a solid solvent for lithium salts because of its flexible ethylene oxide segments and ether oxygen atoms, which have a strong donor character and thus readily solvate Li⁺ cations. (2) Moreover, an alternative polymer polysiloxane (PSi) has recently been investigated, and each of its backbones provides two cross-linkable sites or functional site chains that are favorable for a comb polyelectrolytes complex. (21,22)

In SPEs, the most tremendous challenges are always related to low ionic conductivity, interfacial contact between the electrode and the electrolyte, and the poor electrochemical stability window. (12) The low ionic conductivity in SPEs is considered a trade-off for mechanical properties. Initially, these polymer–salt complex systems may exhibit mechanical properties that are similar in most ways to those of true solids due to chain entanglement of the polymer host, whereas, in a microscopic environment, a lithium ion remains liquid-like and the ion conductivity is “coupled” to the local segmental motion of the polymer at the amorphous region. (23) Consequently, the glass transition temperature (T_g) of the polymer host is the key to determining the mechanical properties and processability of polymer material for SPEs. (2,13,18) Several attempts have been made to lower the T_g to improve the ionic conductivity via copolymerization, cross-linking, blending, and plasticizer application, etc. However, lowering the T_g of the host polymer will result in reduced mechanical properties, which represents a conflict with the fundamental expectation for SPEs in LIB applications. (23) In addition, the ionic conductivities of SPEs studied to date are still in a range of 10^–8 to 10^–5 S cm^–1, which significantly hinders the commercialization of SPEs in LIBs. (10)

The second obstacle to the application of SPEs in LIBs is based on their interfacial contact between electrodes and the electrolytes. It is well-known that liquid electrolytes have higher ionic conductivity and excellent wetting ability to the electrodes, thereby extending the electrode–electrolyte interface into the porosity of the electrode. (10,23) The nature of the solid polymer prevents the SPEs from having a good interface with the electrodes as a liquid electrolyte. Therefore, the interface of SPEs with the electrodes is limited only in the area of actual contact on the surface of electrodes. This phenomenon increases the interfacial impedance of the system and directly influences the battery performance. (23) Another significant issue of SPEs for LIB applications is based on the electrochemical stability window of the polymer host. For oligoether-based SPEs, in particular, the oxidized cathode often presents an insurmountable barrier because the breakdown potential for ether linkages may be below 4.0 V. (12)

1.3. Gel Polymer Electrolytes

Compared with SPEs, GPEs have greater potential for practical applications in LIBs by combining high ionic conductivity and good mechanical properties, which are based on liquid electrolytes and SPEs, respectively. (18,23) Feuillade and Perche proposed the concept of GPEs in 1975, in which a polymer is plasticized with an aprotic solution containing an alkali metal salt in which the organic solution of the alkali metal salt remains trapped within the matrix of the polymer. (2,11,24) Until 1994, Bellcore (currently Telcordia) reported for the first time the utilization of GPEs based on coke/GPE/LiMn₂O₄ in LIBs, in which the GPEs consisted of a poly(vinylidene fluoride)-co-hexafluoropropylene (PVDF-HFP) copolymer matrix and were gelled by a solution of LiPF₆ in ethylene carbonate (EC)/dimethyl carbonate (DMC). (24) Generally, GPEs are obtained by incorporating a certain amount of liquid plasticizer and/or solvent into the polymer–salt system. (2) According to the preparation method, the GPEs are classified into two different categories: physical and chemical gels. (17) The physical gels mean that the liquid electrolytes are confined inside a polymer matrix without any bond formation between the polymer and the solvent. In addition, the chemical gels consist of a cross-linker that leads to the formation of chemical bonding between the functional group of the polymer and the cross-linker agent.

Due to the combined system of polymer–plasticizer/solvent and Li salts, the electrochemical properties of GPEs generally are determined by the liquid/plasticizer component while the safety, morphology, and mechanical properties are defined by the polymeric matrix. (16) A debate remains as to the exact nature of the polymer–plasticizer/solvent–salts interaction on the performance of GPEs, especially for Li⁺ transport properties. (19) The host polymer in GPEs is maintained in semi-solid macroviscosity due to the presence of the plasticizer/solvent; thus, the microscale of Li⁺ is similar to that of the liquid environment inside a GPE. Hence, many researchers consider that Li⁺ motion transport is completely independent from the polymeric segmental relaxation, (12) although many have pointed out that the polymer phase may contribute to ionic transport within GPEs. (19,25)

Various polymer matrices, including PEO, polyacrylonitrile (PAN), poly(vinylidene fluoride) (PVDF), poly(methyl methacrylate) (PMMA), poly(vinyl alcohol) (PVA), and poly(vinyl chloride) (PVC), etc., have been investigated for GPEs in LIBs. (2,17,19,24) Although GPEs present greater benefits over SPEs in terms of ionic conductivity, they present additional problems related to the nature of the liquid and solid electrolytes, such as dual-ion conductivity, low ionic conductivity at room temperature, and poor thermal stability. (1,10) Several attempts have been made to overcome these problems and obtain appropriate properties for LIB applications, such as inorganic filler addition, functional polymer introduction, copolymerization, combinations with different polymer matrix types, cross-linking, and single ion polymer conduction. (2,10) In this review, we first state and discuss the fundamental aspects of GPEs, such as their general requirements and significant properties, including ionic conductivity, transference number, and ionic transport mechanism. In the following sections, we systematically discuss the recent progress of GPEs for LIBs based on the structural design of their polymer matrix for improved electrochemical performance, such as compatibility with electrodes, ionic conductivity, and ionic transport mechanism. Finally, the remaining polymer alternatives and the future direction of GPEs in LIB applications are also presented.

2. Critical Properties and Ion Transport Mechanism of GPEs

2.1. General Requirements of GPEs

In LIBs, GPEs are placed between the cathode and the anode (lithium metal, graphite, and so on) and function as both an electrolyte and separator. Therefore, GPEs play critical roles in the performance of LIB. Due to their importance in the LIB system, GPEs should possess several properties.

2.1.1. Good Ionic Conductivity

A GPE should facilitate the ionic (Li⁺) transport and inhibit the self-discharge of the cathode or anode material; therefore, a GPE must exhibit good ionic conductivity and electronic insulation. (2) The ionic conductivity of GPEs is usually measured based on the electrochemical behavior and the internal resistance at various charge/discharge rates. Generally, the ionic conductivities of liquid electrolytes containing Li salts are in the range of 10^–3 to 10^–2 S cm^–1. (10,20,26) Therefore, to facilitate Li⁺ conduction through the polymer electrolyte, a GPE should have an ionic conductivity of at least one order so that the conductivity of the liquid electrolyte is greater than 10^–4 S cm^–1. (2,7,17,24,26)

2.1.2. High Cationic Transference Number (T₊)

The cationic transference number is the cation mobility measurement relative to the anion in single salts of GPEs. The low ion diffusivity and mobility of the anion in the GPEs greatly affect the LIB performance by increasing the concentration polarization phenomenon of the electrolyte during charge/discharge, thereby lowering the Li⁺-ion transference number and reducing the discharge capacity. (11,20,24,27) Therefore, the GPEs need to have a Li⁺ transference number close to unity (T₊ ≈ 1) in the electrolyte system. (2,7,17,19,20,24,26)

2.1.3. Electrochemical Stability

The electrochemical stability of GPEs determines the operational potential of LIBs during charging/discharging. GPE materials have to be inert to both electrodes (cathode/anode), which means that the oxidation potential must be higher than the embedding potential of Li⁺ in the cathode and the reduction potential must be lower than that of the lithium metal in the anode. (2) Since the LIB generally operates in various potential windows between 0 and 4 V vs Li/Li⁺ (and occasionally reaches high voltages of ∼5 V), the electrochemical stability of the GPE materials is fully required within this potential window for a truly thermodynamically stable system. (2,7,11,19,20,24,26)

2.1.4. Good Chemical and Thermal Stability

Since a GPE is sandwiched between the cathode and the anode, it should be chemically stable and not undergo undesired chemical reactions when in direct contact with the electrodes. Although GPEs could provide ionic conductivity close to that of the liquid electrolyte, due to the solvent embedded inside of GPEs, they can occasionally undergo an irreversible capacity loss associated with the formation of a passivation layer at the surface of the electrode. (7,24) The passivation reactions are usually initiated by the presence of organic solvents in the polymer electrolytes along with unstable anions of the Li salt, which result in their decomposition at electrode surfaces. (20,28,29) Furthermore, the reaction products of the anion could initiate the degradation reactions of the polymer chains. (24) Fortunately, this issue can be prevented by using an electrolyte additive (20,30) and stable Li salts, such as lithium bis(trifluoromethanesulfonyl)imide (LiTFSI). (24)

Generally, a certain amount of heat can be generated and released in the LIB when a short circuit, overcharge, or improper operation happens. Generated heat can initiate the degradation of solvent embedded inside the GPEs or cause a melting process of the polymers under conditions of poor thermal stability. GPEs represent a hybrid electrolyte system of polymer metrics and organic liquids as plasticizer; therefore, the polymer metrics play an important role in maintaining the thermal stability of the GPEs. Therefore, a thermally stable polymer is needed for the GPE system. (2,11,19,24,26) Instead of using a single polymer chain, a composite polymer could significantly enhance the thermal stability of GPEs. (31) Furthermore, a mixture of organic liquids, such as propylene carbonates (PCs) and ethylene carbonates (ECs), could increase the thermal stability as well as extend their potential window toward the lithium electrode. (32)

2.1.5. Good Mechanical Properties

GPEs play dual important roles in LIBs as a conductive medium of the two electrodes and a separator. Therefore, polymers with good mechanical properties are necessary for applying GPEs into LIBs. (2,11,17,26) However, the polymer electrolyte based on plasticized linear polymer chains often reveals poor mechanical properties. (19) This phenomenon is a trade-off between ionic conductivity and mechanical properties due to the organic liquids inside of GPEs. To increase the mechanical properties, the polymer chains with cross-linkable components could be incorporated into GPEs. (19,26) Furthermore, introducing a three-phase system consisting of a polymer network and a solution of inorganic salts could represent another approach to increasing the mechanical properties of GPEs. (19) To that end, the incorporation of inorganic fillers into GPEs is an effective method of enhancing their mechanical strength as well as their transport and electrochemical properties. (2)

2.1.6. Porosity and Electrolyte Uptake

The porosity and electrolyte uptake become important parameters in GPEs, especially for LIB applications, and both parameters directly influence the other. The porosity is defined as the ratio of the void volume to the apparent geometric volume in a polymer membrane, (11) which directly influences the electrolyte uptake ability of organic liquid electrolytes as well as the mechanical properties of a GPE. (6) With greater membrane porosity, more electrolytes can be absorbed into the GPE, although the mechanical properties will be reduced. In addition, electrolyte uptake is a measurement of the weight ratio of a dry polymer membrane that is completely swollen by absorbing the electrolyte to the weight of the dry polymer membrane. (11) As we increase the electrolyte uptake of GPEs, the ionic conductivity can be significantly enhanced. Therefore, the balance between porosity and electrolyte uptake should be carefully considered while designing GPEs for LIBs.

Therefore, the main criteria must be considered in the design and preparation of GPEs for LIB applications. Since the main function of GPEs is to shuttle ions from/to both electrodes, the ionic conductivity and transference number are the most fundamentally important properties when benchmarking a GPE material. The following section provides a further explanation and discussion of the ionic conductivity and transference number.

2.2. Ionic Conductivity

High ionic conductivity is one of the most crucial requirements of polymer electrolytes, including both SPEs and GPEs, and it represents the ability of ionic (Li⁺) conduction inside a polymeric matrix. In GPEs, the ionic conductivity is strongly related to the degree of crystallinity, porosity, and uptake ability of the polymer matrix. (11) Furthermore, the organic liquid inside the polymer matrix also represents the major contribution to higher ionic conductivity for GPEs than SPEs. In addition, the concentration of ionic species in organic liquids and polymers is generally related to the dielectric constant of the medium, which governs the shielding of charges from each other and thereby influences the ion pairing. (20) In other words, a higher dielectric constant of the medium and/or a lower lattice energy of the additive salt corresponds to a higher charge of the carrier concentration. (33,34) Therefore, the ionic conductivity, σ, is directly proportional to the concentration of charge carrier species and their mobility as follows: (20,33)

(1)where n_i is the concentration of the charge carrier species–dissolved ions or charged mobile clusters; q_i is the ionic charge; and u_i is the mobility of each charge carrier. Equation 1 indicates that the ionic conductivity (σ) can be enhanced by increasing either the charge carrier concentration (n) or the ionic species mobility in the system. (33)

Moreover, n generally relies on both the dissociation energy (U) and the dielectric constant (ε′) of the host material as given below: (35,36)

(2)where k_B is the Boltzmann constant and T is the absolute temperature. In addition, the dielectric constant (ε′) is related to the ratio of the material capacitance (C) to the empty cell capacitance (C_o) (ε′ = C/C_o) and the capacitance is also related to the amount of stored charge (C = Q/V), where Q is the total charge and V is applied voltage, which results in a direct connection between the dielectric permittivity and the charge carrier, and an increase of dielectric constant could be interpreted as a fractional increment in the charge concentration in the electrolyte. (33) According to eq 1, the ionic conductivity strongly depends on the charge carrier concentration and the ionic mobility. In addition, eq 2 shows that the charge carrier concentration could be enhanced by increasing the dielectric constant of the medium (polymer matrix and/or organic liquid in the GPE system). In conclusion, based on eqs 1 and 2, the ionic conductivity increases as the dielectric constant of the medium increases. The above explanation further supports the phenomenon that GPEs possess higher ion conductivity than SPEs. Since the organic liquids embedded in GPEs are carbonate-based organic solvents with higher dielectric constant, (12) the charge carrier concentration and ionic conductivity of GPEs can therefore be enhanced.

The ionic conductivity of GPEs is measured by using an AC impedance instrument. In this measurement, the GPE membranes are sandwiched between two stainless steel electrodes and measured in a blocking-type cell. (11) The ionic conductivity of GPEs can be calculated as follows:

(3)where σ is the ionic conductivity (S cm^–1), R_b is the bulk resistance (Ω), l is the thickness of GPE membrane (cm), and A is the effective contact area of the GPEs with the two electrodes (cm²). (11,20,37−42)

2.3. Transference Number (T₊)

The Li⁺ transference number is defined as the number of moles of Li transferred by migration inside of the GPEs per Faraday charge. (20) In GPEs, the cation (Li⁺) and its associated anion are considered mobile in the GPEs. However, because only Li⁺ is involved in the electrochemical redox process and not its anion, the migration of its anion will therefore trigger a concentration gradient at the interface between electrodes and electrolytes as the current passes through the cell. (20,43,44) A severe concentration gradient will cause significant concentration polarization in the system and ultimately result in a high internal resistance and reduced discharge capacity. (11,17,20) In general, if the transference number of Li⁺ in the system is equal to 1 (T₊ = 1), then concentration polarization would not form. (45) Therefore, it is important for GPEs to have a transference number close to 1 (T₊ ≈ 1). (2,7,17,19,20,24,26)

Initially, in a simple binary electrolyte system containing dissociated salt M⁺X^–, the cation transference number can be given as follows: (17,43,44)

(4)where T₊ is the cation transference number and I_M⁺ and I_X^– represent the currents carried by the cation and the anion, respectively. Equation 4 is based on the assumption that the Li salts will completely dissociate as cation and anion, which are Li⁺ and its associate anion, respectively. However, in most cases, the Li salts do not fully dissociate in the electrolyte system; therefore, a series of equilibria are likely to occur in which associated species are formed: (43)

In addition, when a DC current is applied to the electrolyte system, the charge will be transported by the positively and negatively charged species. Since these species may be mobile in the electrolyte system, a concentration gradient will be developed within the electrolyte due to the blocked anions toward the electrodes, and the charge will only be transported by the mobile positively charged species through the electrolyte. Furthermore, the uncharged species will be transported by the concentration gradient and thus contribute to the charge transport (see Figure 3). (44)

According to this phenomenon, Evans, Vincent, and Bruce established a method for determining the Li⁺ transference number based on potentiostatic polarization, (43,44,46) which now represents the common method of determining the Li⁺ transference number. (11,20) In this method, a polymer electrolyte was placed between a symmetric Li||Li cell as shown in Figure 4. Subsequently, a small constant potential (∼10 mV) was applied to the two electrodes, and the initial current value will decrease until a steady-state value is achieved. (20,46) If there is no redox reaction with the associated anion, the anion current will vanish in the steady-state condition and the total current will be directly related to the cation. (43) In this condition, the cationic transference number can be easily obtained by dividing the steady-state cationic current to the initial current just after turning on the polarizing voltage. Moreover, the passivation layer at the surface of the electrodes caused by anion transport usually generates an additional contact resistance. Therefore, an additional drop in voltage will occur, and it needs to be subtracted from the applied potential difference. The contact resistance caused by the passivating layer at the surface of the electrodes may vary with time. Therefore, the contact resistance has to be measured immediately before and after the potentiostatic polarization. (43) Based on this consideration, Evans et al. proposed that the transference number for the cation can be given as follows: (43)

(5)where I_ss is the steady-state current, I_o is the initial current, ΔV is the applied potential, R_ss is the steady-state resistance, and R_o is the initial resistance.

2.4. Ionic Transport Mechanism

The ionic transport mechanism in a polymer electrolyte is a very complicated phenomenon that depends on various factors, such as cation size, dielectric constant of polymer host, ion pairing, amorphous phase character, and the mobility of charge carriers (both anion and cation). However, the concept of the Li-ion transport mechanism is still unclear, and additional investigations are required to propose a reliable mechanism of ion migration inside the polymer electrolyte. Basically, polymer electrolytes are formed by dissolving the low lattice energy of Li salts in the polar polymer matrix, in which the cation (Li⁺) is responsible for ionic conductivity. (33) Consequently, the effective number of mobile ions is strongly related to the degree of dissociation of the salts inside of the polymeric matrix. Consistent with the theoretical ionic transport in high molar mass polymer electrolytes, the ion (cation) can dissociate between neighboring coordinating sites, whether it is situated on the host molecules or nearby molecules. (33,47) Therefore, the cation (Li⁺) could be assumed to form an unstable bond with polar groups in the polymer host and could be transported via segmental motion inside the polymer matrix (2,17,33) as shown in Figure 5. Nevertheless, another new concept of ion transport (Li⁺) has been recently proposed by Bruce et al., (48) who suggested that the crystalline phase of P(EO)₆-LiX (X = PF₆, AsF₆, and SbF₆) pairs within PEO chains is folded with cylindrical tunnels. In this mechanism, the Li⁺ ion is located and coordinated inside of the channel, while the associated anion is located outside of the channel, which means that the Li⁺ ion could be transported from one side to another through the channel without the help of segmental motion of the host polymer (Figure 6). (48)

In general, a polymer host inside a polymer electrolyte is both amorphous and crystalline, and ionic transport principally occurs in amorphous regions above their glass transition temperature. (2,17,33) In other words, the ion transport in the polymer electrolytes is strongly temperature dependent. According to the temperature-dependent conductivity, the ion transport mechanism can be explained by the Vogel–Tammann–Fulcher (VTF) and Arrhenius behavior. (2,11,33,49)

2.4.1. Arrhenius Behavior for Ion Transport

The famous Arrhenius model can explain the relationship between ionic conductivity and temperature. In the polymer electrolyte, the ion transport behavior based on Arrhenius behavior generally related to the ion hopping mechanism is decoupled from long-chain motion of the polymer host at amorphous and crystalline states with the temperature below its T_g. (49) The Arrhenius behavior of ionic conductivity in polymer electrolytes can usually be expressed as follows: (2,11,33,49)

(6)where σ_o is the pre-exponential factor which is related to the charge carrier, E_a is the activation energy, and k_B is the Boltzmann constant. Furthermore, the energy activation can be easily calculated from the best linear fit of the log σ vs 1/T (K^–1). (2,11,49) When the data of the temperature and conductivity are consistent with the Arrhenius relationship (a linear plot of log σ vs 1/T), then the cationic transport can be associated with the hopping mechanism, in which the ion jumps to the closest empty vacant sites to form a new coordination inside of the polymer matrix, thus promoting ionic conductivity inside of the polymer electrolyte. (33,50) Taking one of the recently studied GPEs as an example, Gong et al. (51) prepared a new type of GPE based on an environmentally friendly material (natural lignin) with typical Arrhenius behavior observed as a linear curve (Figure 7). The results indicate that the ionic transport mechanism is based on a hopping mechanism decoupled from the motion of the host material. (51)

In most cases, the polar polymer used to prepare polymer electrolytes is a semicrystalline polymer that contains both amorphous and crystalline phases. For instance, the most popular polar polyether PEO has a melting temperature (T_m) of approximately 68 °C, while T_g is −67 °C, which means that PEO is a mixture of its amorphous and crystalline phases at room temperature. The presence of the crystalline phase (high crystallinity) could prohibit segmental motion of the PEO, and this phase will be reduced when the temperature is close to T_m, in which segmental motion would be expected. (33) As shown in Figure 8, the ionic conductivity increases significantly when the temperature is close to T_m, which indicates that there are many significant contributions of the phase transition to the ionic conductivity and ionic transport mechanism inside the polymer electrolyte. Therefore, to elucidate the true mechanism of ion transport in the polymer electrolyte, the phase transition of the polymer host should be taken into account.

2.4.2. Vogel–Tammann–Fulcher Behavior for Ion Transport

The VTF theory was proposed in the 20th century to elucidate the diffusion process in glassy disordered materials (2,17,52) and to explain the effect of unusual Arrhenius behavior on ionic transport behavior related to the phase transition. The VTF definition of ionic conductivity was developed for pure polymeric systems and is usually in agreement with the free volume theory, in which the polymer dynamic is related to the temperature and amount of free space available for macromolecule motion. (20) In addition, based on this theory the ionic transport mechanism in the polymer electrolyte was caused by ion hopping motion coupled with long-range segmental motion of the polymer host and/or solvent molecules. (49) The segmental motion of the polymer host inside the polymer electrolyte could promote ionic movement by formatting and destructing the coordination of the solvated ion, thereby creating free volume and allowing the ion to diffuse inside of the polymeric matrix. (17) Furthermore, the VTF behavior is usually expressed as follows: (2,11,17,20,33,53,54)

(7)where σ is the ionic conductivity, A is the pre-exponential factor which is associated with the number of charge carriers, B is the pseudoactivation energy for the conductivity, and T_o is a reference temperature related to T_g (or an “equilibrium glass transition temperature”, with T_o ≈ T_g – (50 K)). The pseudoactivation energy could be calculated from the best fitting of the linear plot σ vs 1/T. (11) Moreover, if the plot σ vs 1/T in the polymer electrolyte is a typical nonlinear plot, then the conductivity mechanism based on the ionic hopping motion is coupled with the segmental motion of polymeric chains and/or solvent molecules. (2,33,49) The movement of the polymeric matrix is generally sufficient for promoting ionic conductivity at the T_g, and the mobility of the polymeric matrix will become very restricted to provide new favorable coordination environments for ion to move inside the polymer electrolyte at T_o. (20) As shown in Figure 8, the ionic conductivity typically increases when the temperature is above the T_g and is 2 orders of magnitude higher than that below the T_g, which indicates that the ion mobility is limited by the presence of the crystalline region within the polymer host. (55) At high temperatures, the energy would be large enough to overcome the barriers created between sites, thereby enhancing the free volume in the system so that the polymer chain will gain internal modes faster with segmental motion initiated by bond rotation. (33) Therefore, the ions will be easily transported according to the hopping mechanism. In addition, this theory also points out that the presence of a plasticizer or residual solvent could enhance the free volume and promote free space for the ion and polymer host to move, thereby increasing the ion conductivity. (56,57) This finding is consistent with GPEs having higher ionic conductivities due to more free space for ions and polymers to perform segmental motion, which is provided by the organic solvent inside the polymeric matrix. Furthermore, it can also explain why a higher concentration of salts results in decreased ionic conductivity even with an increase in the charge carrier concentration inside the polymer electrolyte, which is attributed to the induced physical cross-linking between polymer chains in the system and enhanced activation energy of the main-chain bond rotation caused by the high charge carrier, thus reducing the polymer mobility and increasing the T_g. (20)

3. Typical Polymers for GPEs

3.1. Polyether

Polyether is one of the most common polymers in the application of polymer electrolytes for LIBs. Among polyethers, PEO represents the earliest polymer electrolyte system since it was discovered in 1973 by Wright. (13) Typically, the Li ion exhibits coordination bonding with oxygen atoms on the PEO backbone, and ionic transfer could be performed by segmental motion as shown in Figure 5. Although the polymer electrolytes based on PEO present a low ionic conductivity of approximately 10^–8 to 10^–4 S cm^–1, (26,58) the main reasons for using PEO as a polymer electrolyte are related to its lower T_g (∼60 °C) and good complexing characteristics for Li ions. (20) The structural designs of polyethers for GPEs are presented through Schemes 2

–5. Generally, traditional GPEs are obtained by dissolving Li salts in gel polymers, which represents a typical strategy to incorporate new types of Li salts and ionic liquid into GPEs. Nakano et al. (59) prepared a polyether GPE based on 2a through photoinitiated radical polymerization using 2-benzoyl-2-propanol as an initiator. Furthermore, lithium bis(trifluoromethylsulfonyl)amide (LiTFSA) was used as a Li salt, while propylene carbonate (PC) was used as the plasticizer. The resultant GPE shows an ionic conductivity of 3.8 × 10^–4 S cm^–1 at nearly room temperature (293 K) and 40 mol % LiTFSA relative to the molar amount of PC. Another study from Sekhon et al. (60) used an ionic liquid 2b, PEO, and PC to prepare GPEs. The resulting GPE exhibited ionic conductivity on the order of 10^–3 S cm^–1 at room temperature as shown in Figure 9h. In this GPE, the lower viscosity, higher dielectric constant, and the plasticizing nature of the ionic liquid 2b likely resulted in increased ionic conductivity of the resulting GPE, while the role of PEO as a stiffener resulted in a small increase in ionic conductivity. (60) Another ionic liquid, 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (EMIMTFSI), has also been added in the mixed system of polymers and Li salts to enhance LIB performance. Balo et al. (61) reported that, by the addition of ionic liquid EMIMTFSI into GPE together with lithium bis(trifluoromethylsulfonyl)imide (LiTFSI), the electrochemical properties of the resulting GPE were enhanced and showed an increased ionic conductivity up to ∼2.08 × 10^–4 S cm^–1 at 30 °C, a high lithium transference number (T_Li+ = 0.39), and a high electrochemical stability window (∼4.6 V). Another study also showed that incorporating both the ionic liquid and Li salts could increase the ionic conductivity, thermal stability, and potential window. (62) Furthermore, ion–polymer interactions may occur in response to the complexation of cation in Li salts and ionic liquid with oxygen atoms. (62) Although the simple mixing of Li salts and/or ionic liquid already shows the reasonable performance of GPEs, the salt concentration gradients generated on this system lead to a poor performance, including high internal impedances, voltage losses, and undesirable reactions in the electrolyte state, such as phase transitions or salt precipitation. (10)

Another approach to improving the electrochemical performance of GPEs is reducing the internal impedance and voltage loss by forming heterogeneous GPEs. In general, heterogeneous GPEs consist of more than one phase, which can be a hybrid system of polymer-embedded liquid electrolytes or a three-phase system consisting of a liquid solvent, a gel swollen by solvents, and a polymer matrix. The preparation of heterogeneous GPEs mostly leads to two consequences: the formation of a porous polymer membrane and its activation by liquid electrolyte uptake. (2) Li et al. (63) prepared a nanofibrous GPE based on PEO and PVDF by electrospinning and activated by dipping in 1 M lithium hexafluorophosphate (LiPF₆) in EC/PC/DMC (1:1:1, v/v/v). The results showed that by mixing PVDF/PEO with the weight ratio of 1:5, the maximum ionic conductivity of 4.8 × 10^–3 S cm^–1 could be achieved with electrochemical stability up to 4.8 V vs Li/Li⁺ at room temperature. Moreover, GPEs based on blending of three polymers, PVDF-HFP (copolymerization of vinylidene fluoride (VDF) and hexafluoropropylene (HFP)), PEO, and PMMA, have been fabricated and activated by the absorption of 1 M LiPF₆ in EC/DMC (1:1, v/v). (64) The resulting GPE exhibited ionic conductivity ∼8.1 × 10^–4 S cm^–1 and a high electrochemical window up to 5.0 V vs Li/Li⁺ at room temperature. The blending of PEO-PMMA-PVDF-HFP is responsible for the formation of the amorphous structure in the polymer matrix and porous morphology with a higher pore ratio and dimension, which can accommodate much more liquid electrolyte with a more stable liquid holding capability than the pristine PVDF-HFP polymer membrane and thus could provide more Li⁺-ion transport. (64)

Furthermore, additional fillers in the composite membrane have been reported to influence the performance of polyether-based GPEs. Liao et al. (65) reported that by doping 10 wt % nano-Al₂O₃ (inorganic filler) in PEO-PVDF-HFP, the mechanical strength increased from 9.3 to 14.3 MPa, the porosity of the membrane increased from 42% to 49%, the electrolyte uptake increased from 176% to 273%, and the ionic conductivity of the corresponding GPE was improved by up to 3.8 × 10^–3 S cm^–1. The increased ionic conductivity could be associated with the ability of certain inorganic fillers to operate as Lewis acids to compete with Li⁺ ions and prevent their reaction with the polymer, thereby avoiding ion coupling, promoting the dissociation of the salt, and increasing the number of free carriers. (10) In addition, the study suggested that enhanced mechanical properties can be ascribed to the function of nano-Al₂O₃ as a temporary mechanical connection point, which helps to form the net structure and connects firmly with the PEO and PVDF-HFP polymer in the blending system. (65) Other organic fillers, such as 2c, can act as a cross-linker to plasticize porous GPE membranes and changing them into soft but tough and flexible materials based on PEO-PMMA-LiClO₄ composite GPEs, which result in a good electrochemical window (>5 V) and a low interfacial resistance. (66)

Recently, a novel structural design of heterogeneous GPEs based on polyether has been applied to improve the electrochemical performance via the cooperation with other monomers. Ryou et al. (67) fabricated a GPE 2d by initiator-free photopolymerization of a poly(ethylene glycol) dimethacrylate (acrylate monomer) and pentaerythritol tetrakis(3-mercaptopropionate) (thiol monomer) blend. The resulting GPE exhibits an ionic conductivity of 1.1 × 10^–3 S cm^–1 at 25 °C with 50 wt % thiol content and a stable electrochemical window up to 4.4 V vs Li/Li⁺ after activation by dipping in 1 M LiPF₆ in EC/diethyl carbonate (DEC)/(PC) by 30:65:5 weight ratio. Another GPE based on a PEO star-shaped polymer 2e blended with PVDF to form a microporous polymer electrolyte (MPE) was reported by Deng et al., (68) and the GPE exhibited an ionic conductivity of 3.3 × 10^–3 S cm^–1 at room temperature and its electrochemical stability window was above 5.0 V vs Li/Li⁺ after activation by dipping into 1 M LiClO₄ in EC/PC (1:1, v/v). A series of PEO star-shaped polymers have been synthesized by atom transfer radical polymerization (ATRP) using PEO-Br as the macroinitiator and divinylbenzene (DVB) as the core template to place PEO arms. The SEM images in Figure 9a–f show that the morphology of the resulting GPE has a microporous structure, where the distribution of pores on the surface becomes denser with a higher content of star polymer due to the interaction of hydrophilic segments of star-shaped PEO. In this case, star-shaped PEO could help the formation of pores and thereby promote electrolyte uptake, which results in a higher ionic conductivity of up to 3.3 × 10^–3 S cm^–1 at room temperature. (68)

Furthermore, side-chain grafting and copolymerization have also received considerable attention for their ability to enhance the performance of GPEs (Scheme 3). Pan et al. (69) prepared the single ion conductor GPE poly(ethylene-alt-maleic anhydride) (3a) via grafting of 50% 4-amino-4′-trifluoromethyl bis(benzenesulfonyl)imide as side chains. Compound 3a was further blended with PVDF-HFP to form a dense and transparent single ion conductor GPE with an ionic conductivity up to 1.04 × 10^–4 S cm^–1 at room temperature, a Li⁺ transference number of 0.92, and a stable LiFePO₄ discharge capacity of 100 mAh g^–1 at 1 C under room temperature after 1000 cycles (Figure 9g). (69) The side-chain-grafted comb-like single ion conducting polymer electrolyte was found to be largely responsible for the excellent electrochemical performance. The single ion conducting polymer represents an approach to avoiding concentration polarization, which leads to poor performance of GPEs by holding an anion to become a steady phase so that the anion free movement could be limited or totally eliminated. (10) Different strategies were applied in 3b–3d in which the polyether, especially PEO, is used as side-chain “arms” to promote coordination with Li⁺, while the backbone polymer is used to maintain the mechanical properties of the resulting GPEs. Li et al. (70) fabricated 3b through thermal cross-linking with different weight percents of polyether. After activation by 1 M LiPF₆ in EC/DMC/ethyl methyl carbonate (EMC) with a 1:1:1 by weight ratio, the corresponding GPE exhibits an ionic conductivity of 1.1 × 10^–3 S cm^–1. Boaretto et al. (71,72) studied the polysiloxane/polyether-based GPE 3c and GPE 3d, and they showed the maximum ionic conductivity of 8 × 10^–5 S cm^–1 at 30 °C, a stable electrochemical window reaching 4.5–5 V at room temperature, and good thermomechanical behavior up to 100 °C. This study found that extending the length of the mobile polyether chains could increase the ionic conductivity, although the degree of networking is therefore increased. In addition, the decreased polyether ratio in the polymer network leads to the enhancement of T_g, which can reduce the ionic conductivity. (72) An improvement has been made by synthesizing 3c through a sol–gel synthesis and subsequent radical polymerization by incorporating the inorganic filler TiO₂. The resulting GPE exhibited a storage shear modulus of 50 kPa at 100 °C and an ionic conductivity of 1.5 × 10^–4 S cm^–1 at room temperature. (71) The enhanced ionic conductivity could be attributed to the plasticizing effect of TiO₂, which decreases the T_g and allows for free space for polyether arm mobility to coordinate the Li⁺. The results also showed that the poor particle distribution could play a critical role in hindering the charge migration process. (71)

Another strategical structural design for alternating the copolymerization of aromatic polyether as the polymer backbone is presented in Schemes 4 and 5. This strategy is proposed to enhance the mechanical properties while maintaining the ability of polyether to form complexation with Li⁺ and thus promotes the charge transfer. Chen et al. (73) synthesized a single ion conductor GPE based on 4a by using lithiated poly(arylene ether) and lithiated polyamide. The resulting GPE shows that the poly(arylene ether)-based single ion conductor LiPHFE has a higher ionic conductivity than that of polyamide-based LiPACA. (73) This phenomenon is mainly due to the morphological and electrochemical differences between the two copolymers. LiPACA with a large pore size effectively reduces the contact area between GPE film and the electrode, while LiPHFE generates a uniform film due to a higher content of flexible poly(arylene ether). Furthermore, the ether segments in poly(arylene ether) may effectively facilitate ion conduction via a strong electrostatic interaction with Li ions in the polymer matrix of LiPHFE. (73) A good rate capability of LiFePO₄ full cell batteries is shown in Figure 9i, suggesting the excellent performance of single ion conductor LiPHFE. In addition, bis(4-hydroxybenzene sulfonyl)imide and bisphenol A were chosen by Chen et al. to copolymerize with bis(4-fluorine benzenesulfonyl)imide separately and form poly(arylene ether)-based copolymers as shown in 4b. (74) Furthermore, a microporous single ion conducting GPE was formed by taking advantage of the phase separation between rigid aromatic ionomers 4b and aliphatic polymer PVDF-HFP. The distinct electrochemical performance of the resulting GPE was again determined by the morphological differences between the two GPEs, LiPBIE and LiPAFE. The results showed that an interconnected micropore was formed in the polymer matrix when LiPBIE was mixed with aliphatic PVDF-HFP, and it is beneficial to solvent uptake via capillary condensation, which facilitates ion solvation and reduces the threshold temperature of segment motion of the poly(arylene ether) chains with a higher ionic conductivity at room temperature by up to 5.2 × 10^–4 S cm^–1 compared to LiPAFE (9.1 × 10^–5 S cm^–1). (74)

In addition, Voge et al. (75) studied series of poly(arylene ether) copolymers 4c and 5a with different PEO length and content and terpolymers 5b that contain a pyridine group through systemic copolymerization reactions. The pyridine group and PEO side chain on 4c are expected to promote ionic conductivity while the aromatic backbone is used to maintain the mechanical properties. Compound 4c copolymers with various PEO lengths were synthesized using the aromatic polycondensation reaction of aromatic di- and polyfluorides with diols bearing PEO side chains (M_w, 350,750) or by modifying the dihydroxy-functionalized precursor with poly(ethylene oxide) methyl ether tosylate (PEO M_w, 350,750 and 2,000). The results also showed that the T_g can be tuned by varying the PEO content, while the thermal stability can be affected by the PEO length, with a shorter PEO side chain corresponding to higher thermal stability. After doping in 1 M LiPF₆ in EC/DMC with a 1:1 volume ratio, the 5b terpolymer I with a higher PEO side chain (M_w, 750) exhibited ionic conductivity as high as 5.5 × 10^–4 S cm^–1 at room temperature, and it further increased to 2.4 × 10^–3 S cm^–1 at 80 °C. The improved lithium-ion mobility could be attributed to the longer PEO side chains (M_w, 750) and higher PEO content absorption as well as higher charge carrier concentration of PEO. However, the 5b terpolymer II faced an instability issue in liquid electrolyte because of the interaction of the polar pyridine side group and liquid electrolyte, which caused an extreme swelling phenomenon. (75)

3.2. Polyvinyl

Similar strategies, such as mixing with ionic liquid, composite formation, side-chain grafting, and copolymerization, have also been applied in polyvinyl, which is the second most popular GPE for LIBs. A summary of recent GPEs based on polyvinyl is presented in Table 1. Among the polyvinyl GPEs, PVDF is the most common for LIBs due to its favorable properties. PVDF-based GPEs are expected to be highly anodically stable due to the strongly electron-withdrawing functional group (F–C–F). Furthermore, due to the higher dielectric constant as a polymer (ε = 8.4), PVDF can produce greater ionization of Li salts, thereby providing a high concentration of charge carriers by dissolution Li⁺ inside of the polymer matrix. (26,58) The first investigation of GPEs based on PVDF was reported by Watanabe et al. in 1981, in which a homogeneous hybrid film of PVDF can be obtained with a Li salt, EC, and/or PC in the proper proportions. (76) To date, the simple preparation of GPEs by dissolving Li salts into gel polymer is still used due to its simplicity and suitability for roll-to-roll mass fabrication. Fasciani et al. (77) also prepared PVDF-based GPEs by simply dissolving PVDF in EC:DMC (1:1, v/v) under ambient air conditions to obtain a 20:80 mixed solution. The corresponding GPE exhibited good ionic conductivity of 0.5 × 10^–3 S cm^–1 at 30 °C and 1.2 × 10^–3 S cm^–1 at 60 °C. In addition, a low molecular weight organogellator, 1,3:2,4-di-O-methylbenzylidene-d-sorbitol (MDBS), has been added to prepare 6a to avoid liquid leakage and enhance liquid uptake into GPEs. (78) The resulting GPE showed that the addition of organogellator could dramatically decrease the liquid leakage and increase the ionic conductivity up to 1.12 × 10^–3 S cm^–1 at 20 °C, with a stable electrochemical window of 4.7 V vs Li/Li⁺. (78) Furthermore, an inorganic filler, such as Al₂O₃, (79) and carbon material graphene, (40) are reported to effectively improve the ionic conductivity by avoiding ion coupling, thereby promoting the dissociation of the salt and increasing the number of free carriers. Interesting results have been obtained by incorporating inorganic filler Al₂O₃/BN-nanoparticles and macroparticles, in which the ionic conductivity declined because of the pristine PVDF GPEs after activation by 1 M LiPF₆ with ionic liquid solvents mixture of 1-ethyl-3-methylimidazolium triluoromethanesufonate (EMI-TF)/EC/PC with ratio 2:1:1 (v/v/v). (79) The decreased ionic conductivity could be attributed to the reduced volume fraction of the electrolyte mixture in the PVDF matrix resulting from the loading of micro-/nanoparticles, which leads to higher internal resistance. (79) In contrast, the ionic conductivity of PVDF-based GPEs could be enhanced from 1.85 × 10^–3 S cm^–1 (pristine PVDF) to 3.61 × 10^–3 S cm^–1 with a stable electrochemical stability window up to 4.7 V for LiCoO₂/Li cells by incorporating 0.002 wt % graphene. (40) The addition of graphene to the GPE not only increases the ionic conductivity by reducing the PVDF crystallinity and enhancing the porosity so that the electrolyte uptake could be increased up to nearly three times higher but also markedly enhances the rate capability and the cycling performance due to its high conductivity. (40)

Table 1. Summary of GPEs Based on Polyvinyl Polymersa

^a

Abbreviations: CAB, cellulose acetate butyrate; EMIMFSI, 1-ethyl-3-methylimidazolium bis(fluorosulfonyl)imide; fa-SIPE, fully aromatic single-ion conducting polymer; HEC, hydroxyethyl cellulose; LiDFOB, lithium difluoro(oxalato)borate; FEC, fluoroethylene carbonate; LiPAAOB, lithium poly(acrylic acid) oxalate borate; LiPSIPA, lithium poly(bis(phenylsulfonyl imide isophthalate amide diphenylsulfoneamide)); LiPVAOB, lithium poly(vinyl alcohol) oxalate borate; MA-PVDF, maleic anhydride-grafted-poly(vinylidene fluoride); MDBS, 1,3:2,4-di-O-methylbenzylidene-D-sorbitol; MPPyrr-TFSA, 1-methyl-1-propylpyrrolidinium bis(trifluoromethylsulfonyl) azanide; PDMS, poly(dimethylsiloxane); PLS, polyurethane lithium salt; PMImTFSI, 1-propyl-3-methylimidazolium bis(trifluoromethysulfonyl)imide; (P(MVE-MA)), poly(methyl vinyl ether-alt-maleic anhydride); P(VC-VAc), poly(vinyl chloride-co-vinyl acetate); Pyr13TFSI, N-propyl-N-methylpyrrolidinium bis(trifluoromethanesulfonyl)imide; SiO₂PPTFSI, silica nanoparticle-tethered 1-methyl-1-propylpiperidinium bis(trifluoromethanesulfonyl)imide; SPS, sulfonated polystyrene

The formation of composite membranes is the second most popular strategy in the development of polyvinyl-based GPEs, especially for PVDF-based GPEs. As shown in Table 1, several polymers have been mixed with PVDF to enhance the electrochemical properties, and they include poly(dimethylsiloxane) (PDMS), (80) polyethylene (PE), (81) PEO, (82) polystyrene (PSt), (83) and cellulose. (81,84) Furthermore, Prasanth et al. (85) prepared a hybrid inorganic–organic GPE by mixing PVDF with nanoclay in different ratios. After activation by immersing the hybrid GPE into 1 M LiPF₆ in EC/DEC (1:1, w/w), the ionic conductivity enhanced to 3.08 × 10^–3 S cm^–1 at room temperature with 2 wt % nanoclay. The enhanced ionic conductivity is due to the decreased crystallinity of composite membrane and higher porosity so that the liquid electrolytes are well-swollen inside the polymer matrix with higher electrolyte uptake. (85) Another interesting design of GPEs based on a PVDF/hydroxyethyl cellulose (HEC)/PVDF sandwiched structure has been investigated by Zhang et al. (84) The GPEs exhibited ionic conductivity up to 0.88 × 10^–3 S cm^–1 and Li⁺T₊ of 0.57 at room temperature after activation by 1 M LiPF₆ solution in EC/DMC/EMC (1:1:1, w/w/w). The higher Li⁺ transference is associated with the special structural design of sandwiched GPE (Scheme 6b). Moreover, the polarity of the PVDF matrix contributes to the improvement of the Li⁺-ion transference number since H atoms within the PVDF main chain will interact with F atoms in the PF₆̅ anions to form strong H–F hydrogen bonds and hinder the movement of the anions. Moreover, there is also a strong H–F hydrogen-bonding interaction between F atoms in PF₆̅ anions and H atoms in the -OH groups of the hydroxyethyl cellulose (HEC) host, which also hinders the movement of PF₆̅ anions from passing through and avoids concentration polarization, thereby promoting a higher ionic conductivity and Li⁺ transference number. (84)

An interesting performance has been shown by mixing PVDF with 6c, which offers an ionic conductivity of 0.35 × 10^–3 S cm^–1, Li⁺T₊ of 0.58, and flame-retardant property after activation by 1 M LiPF₆ solution in EC/DMC/EMC (1:1:1, w/w/w) at room temperature (Figure 10a–d). The good Li⁺ transference is related to the property of LiPAAOB, which is capable of solvating Li⁺ and keeping the contrary anion in steady phase. The concentration polarization could therefore be minimized, which results in the increased mobility of Li⁺ and higher transference number. (86) A similar strategy of preparing hybrid microporous GPEs based on PVDF and 6d was performed by Xing et al. (87) The resulting GPE exhibited ionic conductivity of 4.49 × 10^–3 S cm^–1 at room temperature. A side-chain modification was also performed by Zhu et al. (27) for preparing the polyvinyl-based single ion conducting polymer 6e. Although the ionic conductivity of 6e was slightly low (∼6.11 × 10^–6 S cm^–1), it shows an electrochemical window that is stable up to 7 V vs Li/Li⁺ and presents great potential for higher voltage LIBs. Furthermore, the study shows that the resulting GPEs present Arrhenius behavior, indicating that the Li⁺-ion motion is dependent on the flexibility of the polymer side chain. (27)

Although the PVDF-based GPEs showed interesting results by mixing with plasticizer and organic/inorganic fillers, the incorporation of more plasticizer could reduce the mechanical properties while improving the electrochemical performance. The copolymer of PVDF and hexafluoropropylene (PVDF-HFP) could overcome the negative issues, the amorphous domains in pristine PVDF are capable of trapping large amounts of liquid electrolytes, and the crystalline regions can provide sufficient mechanical integrity for the processing of freestanding films, thereby eliminating the required cross-linking step. (26) A similar strategy has also been applied for copolymer PVDF-HFP, such as incorporating with an ionic liquid (88−94) and making a composite polymer by blending with inorganic filler (95−97) or mixing with another polymer matrix. (64,98−101) Kuo et al. (90) also fabricated hybrid GPEs by blending PVDF-HFP with oligomeric organic liquid to form 6f, which exhibited ionic conductivity of 2.0 × 10^–3 S cm^–1 at 30 °C and 6.6 × 10^–3 S cm^–1 at 80 °C and presented high thermal stability up to 150 °C. Although the electrolyte uptake was reduced to ∼50% relative to the pristine PVDF-HFP, the oligomerization in ionic liquid is responsible for enhancing the mechanical properties of the resulting GPEs and offering high conductivity with flame-retardant capabilities, which is attributed to nonflammable pyrrolidinium- and imidazolium-based ionic liquids. (90) Freestanding GPE that consists of PVDF-HFP and inorganic/ceramic filler nanosized Al₂O₃ ceramic particles was fabricated by Kim et al. via the electrospinning process. (95) After activation with 1 M LiPF₆ in EC/DMC, the resulting GPEs showed good ionic conductivity of 4.1 × 10^–3 S cm^–1 at 25 °C and 7.8 × 10^–3 S cm^–1 at 80 °C. This phenomenon is due to the effect of nanosized Al₂O₃, which increases electrolyte uptake to 20% higher than that of the pristine PVDF-HFP prepared by electrospinning. Furthermore, the Raman spectrum of the corresponding GPE shows that the incorporation of nanosized Al₂O₃ could effectively enhance the solvation of Li⁺ ions and promote Li⁺-ions transfer. (95) To enhance the electrochemical properties, Li et al. (101) used the electrospinning method to fabricate a novel GPE composite polymer 6g, which consists of PVDF-HFP and poly(methyl vinyl ether-alt-maleic anhydride) (P(MVE-MA)) and uses polyethylene (PE) as support. The ionic conductivity of the resulting GPEs shows that mixing 75 wt % P(MVE-MA) could increase the ionic conductivity up to 1.6 × 10^–3 S cm^–1 at 25 °C and provide for superior cycle stability with a negligible capacity loss after 100 cycles after activation in 1 M LiPF₆ EMC/EC/DEC (5:3:2, w/w/w) (Figure 10e). The electrochemical performance of GPEs is significantly enhanced by introducing the P(MVE-MA) copolymer into the PVDF-HFP system. The presence of P(MVE-MA) enhances the invasion of the liquid solvent, thereby creating fast ion channels that can increase the uptake ability of liquid electrolytes, which accelerates the process of intercalation/deintercalation. (101) Freestanding GPEs were also prepared by blending with single ionic conducting polymer 6h. The results show that the morphology of the GPE was significantly different between the PVDF-HFP mixed LiPSIPA and PVDF-HFP mixed LiPSIOA (Figure 10f–i). The PVDF-HFP mixed LiPSIPA showed a porous structure with interconnected channels, while PVDF-HFP mixed LiPSIOA becomes a dense film. When mixing with PVDF-HFP, the nature aromatic backbone of the LiPSIPA and aliphatic backbone of PVDF-HFP led to poor chemical compatibility, which induced phase separation during the blending process and a porous structure of the resulting membrane. The same aliphatic backbone of LiPSIOA and PVDF-HFP presented good chemical compatibility; therefore, a dense freestanding film could be achieved. (98) The electrochemical performance indicates that GPEs prepared from PVDF-HFP mixed LiPSIPA exhibited higher ionic conductivity of approximately 3.7 × 10^–4 S cm^–1 over that from PVDF-HFP mixed LiPSIOA (∼7.9 × 10^–5 S cm^–1) at room temperature after soaking in EC/PC (1:1, v/v). The higher electrochemical properties of PVDF-HFP mixed LiPSIPA are related to their porous morphology with interconnected channels, which can uptake more solvent and improve ion transfer inside the GPEs. (98) This study demonstrated that the porous structure of GPEs plays a critical role in LIB performance and suggests a facile approach to preparing porous GPE membranes for developing better LIBs with outstanding performance.

3.3. Polynitrile

Among polymers, the application of polynitriles in LIBs has attracted interest, especially as a polymer electrolyte. Polyacrylonitrile (PAN) is one of the most investigated polynitriles for GPEs. Thus, far, many scientists have suggested that PAN could offer a homogeneous and hybrid electrolyte film in which the salt and the plasticizer are molecularly dispersed. (58) Early studies reported by Watanabe et al. (106,107) showed the direct mixing of PAN with plasticizer (EC and PC) and formation of a complex with LiClO₄ to prepare PAN-based GPEs. The investigations found that PAN is inactive toward the Li⁺ transport mechanism and only acts as a matrix for structural stability. (26,58) In addition, by using Raman and infrared (IR) spectroscopy, another study demonstrated that there is a strong interaction between Li⁺ and C≡N groups on the PAN. (108) Nevertheless, the Li⁺ transport mechanism on PAN-based GPEs is still under debate and requires additional investigation to provide a complete understanding. (26) Furthermore, three strategies have been used to develop PAN-based GPEs: incorporation of Li salts/ionic liquid, (109) formation of composite polymer, (110−116) and copolymerization. (117−124) The summary of PAN-based GPEs is presented in Table 2. Patel et al. (109) prepared a novel soft GPE via free radical polymerization of an acrylonitrile monomer under room temperature with an ionic liquid electrolyte, N,N-methyl butyl pyrrolidinium-bis(trifluoromethanesulphonyl)imide-lithium bis(trifluoromethanesulphonyl)imide (LiTFSI-[Py1,4-TFSI]). The corresponding GPE was found to exhibit higher ionic conductivity up to 1.8 × 10^–3 S cm^–1 at 25 °C and good cyclability for LiFePO₄ battery performance (Figure 11a). The study found that the Li⁺ transport is associated with the presence of ionic liquid, while the crystallinity region generated from the acrylonitrile polymer matrix does not contribute to the ionic conductivity. (109)

Table 2. Summary of GPEs Based on PANa

^a

Abbreviations: OMMT, organic montmorillonite; P(AN-MA), poly(acrylonitrile-maleic anhydride); P(AN-POSS), poly(acrylonitrile-polyhedral oligomeric silses-quioxane); P(AN-VAc), poly(acrylonitrile-vinyl acetate); PE, polyethylene; PEGDME, poly(ethylene glycol)dimethyl ether; P(MMA-AN-BA), poly(methyl methacrylate-acrylonitrile-butyl acrylate); P(MMA-AN-EA), poly(methyl methacrylate-acrylonitrile-ethyl acrylate); P(MMA-BA-AN-St)/PE, poly(methyl methacrylate-butyl acrylate-acrylonitrile-styrene)/polyethylene; PYR₁₄TFSI, N-methy-N-butylpyrrolidinium bis(trifluoromethanesulfonyl)imide; SWy, organic sodium smectite clay; TEGDA-BA, triethylene glycol diacetate-2-propenoic acid butyl ester.

The formation of a PAN-based polymer composite is the most common strategy to prepare GPEs for LIBs. This strategy could be achieved by blending with another polymer matrix (110,112−114) and/or inorganic filler. (111,115,116) Another interesting strategy demonstrated by Wang et al. (113) is the mixing of PAN-based GPEs with three-dimensional network structures 7a through in situ thermal polymerization (Scheme 7). Triethylene glycol diacetate-2-propenoic acid butyl ester (TEGBA-BA) is known to enhance the ionic conductivity of GPEs due to the good dispersion of liquid electrolytes in TEGBA-BA polymer networks, which is primarily responsible for the migration of Li⁺ inside of GPEs. (125) The resulting GPEs with 5 wt % PAN exhibits ionic conductivity up to 5.9 × 10^–3 S cm^–1 at 25 °C and a high Li⁺ transference number up to 0.481. This performance is associated with the synergistic effects of cross-linked short-chain monomers (TEGBA-BA) and a linear long-chain polymer (PAN), which enhances the compatibility of electrolytes and increases the loading amount of liquid electrolyte (1 M LiPF₆ in EC/DMC/EMC (1:1:1, v/v/v)) and thus is beneficial for fast Li⁺ mobility. (113) Furthermore, Zhang et al. (114) used the coaxial-electrospinning method to prepare a GPE composite based on PAN and PEO. Taking advantage of PAN, a GPE fabricated from nanofibers with a core–shell structure was prepared (PAN, core; PEO, shell) and activated by immersing in 1 M LiPF₆ in EC/DMC (1:1, v/v). This study revealed that ionic transport occurs through the amorphous region of PEO and PAN (Figure 11b). In the PAN phase, the lithium ions are present around the C═O groups in EC and C≡N groups in PAN, and the segmental motion of EC and PAN can affect the movement of lithium ions. In the PEO phase, lithium-ion transport can occur mainly through an association–disassociation between the ions and the PEO. (114) The electrochemical performance of the resulting GPE exhibited an ionic conductivity of 4.9 × 10^–3 S cm^–1 and was slightly enhanced to 5.3 × 10^–3 S cm^–1 after it was radiated with an electron beam to form cross-linking reactions, which not only destroyed the regularity of the molecular chains but also reduced the degree of crystallinity of polymers in the nanofiber membranes, thereby accelerating the Li⁺ transfer primarily in the amorphous region. (114)

In addition, a GPE composite has recently been constructed from PAN/organic montmorillonite (MMT) through a non-solvent-induced phase separation method by He et al. (116) MMT is a well-known inorganic filler based on a layered silicate with a central alumina octahedral sheet sandwiched between two silica tetrahedral sheets, which can improve the thermal stability. (126) After activation by 1 M LiPF₆ in EC/DMC/EMC (1:1:1, v/v/v), the resulting GPE showed high ionic conductivity of 2.5 × 10^–3 S cm^–1, a stable potential window up to 4.62 V, and thermal stability up to 150 °C. This study revealed that the incorporation of organic MMT has a direct effect on the membrane morphology via the formation of a highly porous structure with high liquid uptake up to 375.5%, which is responsible for its electrochemical performance. (116) Moreover, the cycling performance of LIB based on a LiCoO₂ cathode shows a high capacity of 141 mAh g^–1 after 100 cycles and high capacity retention of 95.9% at 1 C (Figure 11c), indicating the cyclic compatibility and low interfacial resistant. (116)

The structural design of copolymerizing PAN with other monomer and/or other functional groups has been intensively investigated by many scientists. Taking advantage of poly(vinyl acetate) to solvate Li⁺ and the good mechanical strength of PAN, a copolymer poly(acrylonitrile-vinyl acetate) (P(AN-VAc)) has been successfully synthesized by Li et al. (117) The resulting copolymer was coated on PE nonwoven fabric and activated by 1 M LiPF₆ in DMC/DEC/EC (1:1:1, v/v/v). The corresponding GPE revealed an ionic conductivity up to 3.8 × 10^–3 S cm^–1 at room temperature, Li⁺ transference number as high as 0.52, and stable electrochemical windows from 5.0 to 5.6 V vs Li/Li⁺. The resultant morphology of the P(AN-VAc)/PE nonwoven material was highly porous (higher pore volume density) and beneficial for more liquid electrolyte uptake with fast Li⁺ transfer, (117) demonstrating that the morphology of the GPE membrane plays an important role in the electrochemical performance. A novel GPE based on terpolymer 7b was successfully fabricated through emulsion polymerization by Sun et al. (118) The purpose of forming a terpolymer based on polynitrile and polyacrylate is to combine the good mechanical properties and Li⁺ solvation contributed by polynitrile and polyacrylate, respectively, which are expected to exhibit good performance for LIBs. (118) The addition of ethyl acrylate is used as a soft monomer with good cohesiveness for the synthesized macromolecule. (127) Furthermore, a pore-forming agent, poly(ethylene glycol) 400 (PEG-400), was added during the phase inversion and further activated by 1 M LiPF₆ in EC/DMC (1:1, v/v). GPEs based on terpolymer P(MMA-AN-AE) show an ionic conductivity up to 3.82 × 10^–3 S cm^–1 at room temperature and are electrochemically stable up to 5.2 V vs Li/Li⁺, and these properties are suitable for high voltage cathode materials. The investigation shows that the ionic conductivity is proportional to the electrolyte uptake, which is strongly dependent on the pore structure and coupled with special polymer–solvent interactions due to the existence of the strong polar groups C≡N and C═O in the terpolymer matrix. (118)

Furthermore, inorganic fillers, such as nano-SiO₂ and nano-Al₂O₃, have been added to improve the electrochemical performance of GPEs. (119) Although they only provide a slight improvement of ionic conductivity, the combination of 5 wt % nano-SiO₂ and 5 wt % nano-Al₂O₃ inside of terpolymer P(MMA-AN-AE)-based GPEs enhanced the electrochemical stability up to 5.6 V vs Li/Li⁺. The improvement of the electrochemical stability was due to the synergic effect of nano-SiO₂ and nano-Al₂O₃ fillers. (119) The Si–O bonds or Al–O bonds as the connection points contribute to building a stronger polymer network structure by the formation of Si–O–C or Al–O–C covalent bonds, and the interaction between the polymer and liquid electrolyte is therefore strengthened. The addition of both nano-SiO₂ and nano-Al₂O₃ fillers makes the interaction between the polymer and liquid electrolyte become even stronger, which is due to the formation of a super network structure and the low consistency of the electronic charge between SiO₂ and Al₂O₃ contributed in the effective connection points (119) By using a similar strategy, a novel quaternary polymer 7c was successfully synthesized via emulsion polymerization, in which PMMA and acrylonitrile were introduced to provide a good affinity to liquid electrolyte, and butyl acrylate (BA) supplied adhesive characteristics while polystyrene offered mechanical strength. (121) However, the electrochemical performance of the resulting GPE showed a slight improvement compared with the terpolymer by presenting an ionic conductivity as high as 2.7 × 10^–3 S cm^–1 and electrochemical stability up to 5.3 V vs Li/Li⁺ at room temperature. This phenomenon could be associated with the higher internal resistance of the quaternary polymer than that of the terpolymer. For long-term operations, the quaternary polymer 7c exhibited higher internal resistance, which started at 300 Ω cm² on the first day and increased to 514 Ω cm² after 15 days of testing, (121) while the internal resistance of terpolymer 7b increases from 62 Ω cm² initially to 69 Ω cm² after 15 days. (118) These results show that the addition of polystyrene would increase the internal resistance and lead to lower ionic conductivity.

Recently, a random copolymer 7d consisting of PAN and maleic anhydride (MA) was successfully fabricated with good electrochemical properties after activation by 1 M LiPF₆ in EC/DMC/EMC (1:1:1, w/w/w), such as ionic conductivity up to 3.03 × 10^–3 S cm^–1, Li⁺ transference number of 0.57, and a high electrochemical stability window up to 5.4 V vs Li/Li⁺ at room temperature. (123) The good electrochemical performance is attributed to the additional MA, which contains a five-membered ring with strong negatively charged groups thus could disrupt the interaction between C≡N and PAN. In addition, the steric effects of MA might weaken the electrostatic interaction between the polar groups C≡N and lithium metal electrodes to some extent, thereby resulting in good compatibility with the lithium electrode. Moreover, the strong negatively charged groups of MA may also hinder the movement of negatively charged carriers, resulting in a high effective current for Li⁺ mobility. (123) The GPEs based on copolymer 7e demonstrated an outstanding electrochemical performance, such as a high ionic conductivity of 6.06 × 10^–3 S cm^–1 at ambient temperature after activation by 1 M LiPF₆ in EC/DMC (3:7, v/v), Li⁺ transference number of 0.59, and electrochemical stability up to 5.7 V vs Li/Li⁺. (124) Moreover, the better cycle capability of copolymer 7e has been performed by incorporating 8% POSS content in the GPE within the Li/GPE/LiFePO₄ half-cell at different current rates (Figure 11d). The excellent electrochemical property is due to the incorporation of polyhedral oligomeric silsesquioxane (POSS), which is an intramolecular-level organic (alkyl or vinyl groups)–inorganic (Si–O–Si bonds) hybrid nanoparticle that possesses a type of three-dimensional cage structure with a general formula of (RSiO_1.5)_n. (124,128) By incorporating POSS on the PAN backbone polymer, the organic double bonds of C═C in POSS disrupt the polymerization sequence of AN monomers and form a network-like copolymer structure, which contributes to weakening the interaction of polar groups C≡N between adjacent polymer segments, enhancing the compatibility with Li metal, expanding the amorphous region of the GPEs, improving the electrolyte uptake, and accelerating Li⁺ mobility. (124)

3.4. Polycarbonate

Because of the success of carbonate-based liquid electrolytes, polycarbonate has been selected as an alternative in polymer electrolytes to overcome the safety issues related to carbonate-based liquid electrolytes. Polycarbonates are expected to have properties similar to those of carbonate-based electrolytes and minimal/eliminated natural flammability, which is a typical problem from carbonate-based liquid electrolytes. The carbonate groups used in polymer construction are capable of providing a polar environment that is suitable for the dissociation of salts and the solvation of ions. (129) Polycarbonates can coordinate with Li⁺ via the carbonyl group oxygen, which is similar to the Li⁺-coordinating ether oxygens of PEO and other polyether electrolyte hosts. (20) Moreover, coordination by the alkoxy oxygens (sometimes referred to as “ether oxygens”) adjacent to the carbonyl group has also been implicated in carbonate solvents. (130) Because of the properties that are similar to that of PEO, more research efforts have focused on the application of polycarbonates as polymer electrolytes, especially for SPEs, and this class of alternative host materials has recently attracted the most intense research attention and progress. (20) Nevertheless, the application of polycarbonates for GPEs has recently drawn attention among the scientists since the lower ionic conductivity hinders the use of SPEs in real LIB applications. Zhou et al. (131) fabricated GPEs by mixing a poly(propylene carbonate) (PPC) host with LiClO₄ and an ionic liquid 1-butyl-3-methylimidazolium tetrafluoroborate (BMIM⁺BF₄^–). The advantages of PPC, which has a structure similar to that of carbonate-based organic solvent and a low T_g, were exploited, and a combination with a small amount of ionic liquid could turn PPC to a rubbery state, thereby facilitating the motion of Li⁺ within the polymer matrix. (131) The results show that the ionic conductivity of GPE based on PPC/LiClO₄/BMIM⁺BF₄^– with a weight ratio of 1/0.2/3 is 1.5 × 10^–3 S cm^–1 at ambient temperature. Furthermore, the study also revealed that the ionic transport inside of GPE follows the Arrhenius equation, which means that Li⁺ is transported through an ion hopping mechanism and decoupled from the long-chain motion of the polymer host at both amorphous and crystalline states below the T_g. Another strategy demonstrated by Tillmann et al. (132) is to copolymerize 8a and oligo(ethylene glycol) methyl ether methacrylate (OEGMA) to form 8b by using UV-induced free radical polymerization (Scheme 8). The electrochemical performance of the resulting GPE shows an ionic conductivity of 2.3 × 10^–3 S cm^–1 at 25 °C after activation by 1 M LiPF₆ in EC:DMC (1:1, w/w) with electrolyte uptake of nearly 550%. The results are consistent with the expectation that both monomers have interactions with the cyclic carbonate moiety in liquid electrolyte within the polymer matrix. The strong interactions could improve the gelation behavior of the GPEs compared to other ethylene glycol-based polymer systems and can therefore enhance the electrolyte uptake and Li⁺ mobility. (132)

Recently, PPCs have attracted research attention for application as a polymer electrolyte in LIBs. As GPEs for LIBs, the polar group (ester group) on the PPC backbone is beneficial for ionic conduction by trapping and storing liquid electrolytes. (133) Furthermore, because of the low T_g of PPC, the local relaxation and segmental motion of PPC chain is favorable for transporting Li⁺ inside of GPEs. However, the low mechanical properties significantly hindered the practical application of PPCs for LIBs. (133,134) Zhao et al. (134) therefore cast PPC on a cellulose membrane to improve the mechanical property. After activation by 1 M lithium oxalyldifluoroborate (LiODFB)-PC (1:1, v/v), the resulting GPE exhibited an ionic conductivity as high as 1.14 × 10^–3 S cm^–1, Li transference number up to 0.68, and stable electrochemical window up to 5.0 V vs Li/Li⁺ at 30 °C. Furthermore, the battery performance shows a much better capacity retention of up to 91.3% for LiNi_0.5Mn_1.5O₄/Li batteries compared with the common liquid electrolyte, which has a capacity of only 78% (Figure 12a). The stable electrochemical performance of GPE based on PPC-cellulose may be due to the strong polarity groups C═O in the polymer matrix, which effectively retain the liquid electrolyte and form more ionic transport channels with high dissociation of salt and Li⁺ mobility. (134) Huang et al. (133) prepared the GPE based on an electrospinning membrane made of PPC and PVDF. The corresponding GPE exhibited an ionic conductivity of 2.11 × 10^–3 S cm^–1 at 30 °C and stable electrochemical window up to 5.2 V vs Li/Li⁺. As shown in Figure 12b, the cyclic performance of Li/GPE/LiFePO₄ half-cell shows higher capacity compared to that from pure PVDF with a commercial Celgard PE separator. This study revealed that the formation of Li⁺···(δ^–) F–C (δ⁺) and PF₆^–···(δ⁺) C═O (δ^–) complexes in the polymer framework can separate the Li⁺ and PF₆^– anions, which effectively prevents the simultaneous reconnection between Li⁺ and PF₆^– anions and creates more free Li⁺, thereby improving the transport ability of lithium ions. (133)

3.5. Polyacrylate

Among polyacrylates, PMMA is the most common polymer applied as a GPE for LIBs. In early studies, PMMAs were used as plasticizer or gelatinization agent in polymer electrolyte to increase the mechanical properties. (26,58) Recently, PMMA has been widely used as a primary backbone polymer for GPEs via simple blending or copolymerization. (135−137) Similar to polyether and polycarbonates, PMMA was expected to be beneficial for Li⁺ transport and have the ability to solvate the Li⁺ or the liquid electrolyte inside of the polymer matrix by the C═O polar group in PMMA. Studies have revealed that a consistent fraction of lithium is lost upon cycling and therefore a large excess of lithium would be required to ensure an acceptable battery life. (58) Huang et al. (137) prepared a hybrid GPE 9a consisting of poly(methyl acrylate)/poly(ethylene glycol) (PMA/PEG) and formed a quasi-solid-state electrolyte with ionic conductivity of 0.57 × 10^–3 S cm^–1 after activation by LiClO₄ in DMSO (Scheme 9). Furthermore, this study also confirmed that the ionic conductivity increases as the concentration of LiClO₄ inside of GPEs increases (Figure 13a). This finding was consistent with the theory that ionic conductivity could be enhanced by increasing the charge carriers. (33)

In addition, copolymerization is the most popular strategy of structural design for polyacrylate-based GPEs. GPE 9b was prepared by the copolymerization of poly(ethylene glycol) methyl ether methacrylate (PEGM) and poly(ethylene glycol) dimethacrylate (PEGDM). (136) After introducing 60 wt % content of ionic liquid methylpropylpyrrolidinium bis(trifluoromethanesulfonyl)imide (MPPY·TFSI) as a plasticizer and mobile charge carrier, the resulting GPE copolymer exhibited ionic conductivity of 1.0 × 10^–3 S cm^–1 at 25 °C. As presented in Figure 13b, the ionic conductivity can be further enhanced by increasing the amount of ionic liquid. This study again confirmed that the enhancement of ionic conductivity is consistent with an increased charge. (33) In addition, by increasing the ionic liquid content over 60 wt %, the GPEs will lose their mechanical properties and result in a gel-like electrolyte. (136) Liao et al. (138) successfully fabricated GPE 9c by copolymerization of PMMA and PSt, and it presented a good combination of a higher affinity toward the liquid electrolytes and good mechanical properties. (138) The resultant P(BMA-St)-based GPE exhibited ionic conductivity of 1.1 × 10^–3 S cm^–1 at room temperature and a stable electrochemical window up to 4.7 V vs Li/Li⁺ after activation by immersing into 1 M LiPF₆ in EC/DMC/EMC (1:1:1, v/v/v). The ionic conductivity and electrochemical stability of P(BMA-St) could be further enhanced by incorporating with fumed silica, which resulted in an ionic conductivity up to 2.2 × 10^–3 S cm^–1 at room temperature with a stable electrochemical window up to 5.2 V vs Li/Li⁺. The improvement can be ascribed to the porous morphology generated from fumed silica, which also reveals that high affinity with liquid electrolytes thus can tightly trap the liquid electrolyte inside the membrane and results in good oxidative stability of GPEs. (138)

By exploiting polyacrylates, the copolymers 9d (139) and 9e (140) were fabricated, and they were expected to solvate the Li⁺ and exhibit strong affinity toward liquid electrolytes. After activation by immersing the resulting GPEs into 1 M LiPF₆ in EC/DMC (1:1, v/v), GPEs 9d and 9e exhibited almost similar ionic conductivities at 25 °C of 1.8 × 10^–3 and 1.2 × 10^–3 S cm^–1, respectively. Another structural design of the polyacrylate 9f-based GPE was reported by Wang et al. (135) through the formation of three comb-like methyl methacrylate copolymer matrices. In addition, the copolymer methyl methacrylate-maleic anhydride (P(MMA-MAh)) was first prepared as the main backbone for grafting poly(ethylene glycol) monomethyl ether (PEGME) with different molecular weights to form three comb-like 9f GPEs. (135) Although the three comb-like 9f GPEs exhibited relatively lower ionic conductivity at approximately 1.22 × 10^–3 S cm^–1 at 60 °C after incorporating 1 M LiClO₄ in PC, the study revealed that the ionic conductivity increased significantly by increasing the side-chain length of the copolymers. As the side-chain length of the copolymer increased, the number of pendant ethylene oxide segments will also increase in the copolymer network, which results in a decreased T_g and an increased fractional free volume with enhanced Li⁺ transport. (135) This study also confirmed that ionic transport inside of the three comb-like GPE 9f follows the VTF behavior, in which segmental motion of polymer host inside the GPE promotes the ionic movement by the formation and destruction of solvated ion coordination. (17)

4. Alternative Polymer Types for GPEs

4.1. Polyaromatic Polymer

The application of polyaromatic polymers for GPE construction has attracted considerable research attention. Polyaromatic polymers have been used to improve the mechanical and thermal properties. (141) In addition, the nature rigid structure of the polyaromatic polymers enhances ionic conductivity by increasing liquid electrolyte uptake through the formation of a more porous structure of GPEs. (98) Recently, a polyaromatic polymer with anion anchored on the backbone was applied to prepare a single ion conducting polymer, and the charges were therefore partially delocalized by an aryl group; thus, Li⁺ poorly paired with the constituent anion and facilitated transport. (98) Furthermore, ion polarization could also be minimized by keeping anions in an immobile state via the application of single ion conductor polymers. (10) Several polyaromatic polymers have been introduced as GPEs by direct blending, composite polymer formation, or copolymerization. Aromatic polysulfone (PSF) has been used as a segment of copolymer 10a by condensation polymerization (Scheme 10). (142) The resulting copolymer was then used as a GPE by introducing Li salts in succinonitrile (SN) as the plasticizer. As shown in Figure 14a, the ionic conductivity of GPEs based on copolymer 10a depends on the PEO content, indicating that PSF was only used as a framework to improve the mechanical and thermal properties of aliphatic PEO. (142) Recently, aromatic poly(arylene ether) has been used to prepare not only the polymer backbone of GPEs but also single ion conducting polymers. The poly(arylene ether)-containing pyridine group was synthesized and used as a main backbone for the GPEs 4c, 5a, and 5b. (75) To promote the ionic conductivity of the GPEs, various molecular weights of PEO were grafted on the polymer backbone of 4c, 5a, and 5b. However, the resulting GPEs exhibited a relatively low ionic conductivity of 5.5 × 10^–4 S cm^–1 at room temperature, although it increased to ∼2.4 × 10^–3 S cm^–1 by increasing the temperature to 80 °C. Furthermore, the investigation found that the mechanical properties of GPEs are still maintained even if the T_g or thermal stability are changed by tuning the PEO content or length. (75)

In addition, various single ion conducting polymers have been prepared based on aromatic poly(arylene ether) and applied as GPEs for LIBs. Poly(arylene ether) has been used as single ion conducting polymers 4a, 4b, and 6h to improve the electrochemical performance and mechanical properties. (73,74,98) In a single ion conducting polymer, the anion is kept in an immobile state (as polymer backbone) and the benzene ring (aryl group) on poly(arylene ether) could be partially delocalized Li⁺ with poor pairing, which is beneficial for Li⁺ transport inside GPEs. Because of the phase separation of the incompatible rigid structure of poly(arylene ether) with other aliphatic polymers, the single ion conducting polymer based on poly(arylene ether) could also enhance the ionic conductivity by forming a more porous structure, which results in increased electrolyte uptake. (98) Another fully aromatic polymer, poly(ether ether ketone) (PEEK), has been used to prepare a new class of single ion conducting polymer 10b for GPEs. (143) The single ion conducting polymer 10b was achieved by grafting with sulfonamide and bis(sulfonyl)-imide groups. The most interesting feature of GPEs based on the single ion conducting polymer 10b is the absence of fluorine in its repeat unit, which is highly advisible from an environmental compatibility point of view. (143) As shown in Figure 14b, the highest ionic conductivity of GPE based on the single ion conducting polymer 10b was nearly 10^–4 S cm^–1 at 20 °C and increased as the temperature increased to ∼10^–3 S cm^–1 at 65 °C after activation using 5 wt % acetonitrile as plasticizer. This study indicates that the plasticizing effect of the solvent has a considerable impact on the ionic conductivity of the resulting GPEs. In summary, polyaromatic polymers could represent a potential alternative polymer for the preparation of high performance GPEs for LIBs due to their good mechanical properties. In addition, the aryl groups can further promote the interaction with Li⁺ and minimize ionic polarization via enhanced electrochemical and mechanical properties.

4.2. Hydrogen-Bonding Polymer

Recently, a polymer capable of forming hydrogen-bonding networks has achieved popularity based on its ion conduction. The polar networks are expected to dissolve large amounts of salt and tend to show interesting ion transport behavior, such as Arrhenius-type conduction, and appreciable conductivity below T_g. (20) Furthermore, the hydrogen-bonding polymers are intended to have interactions similar to that of liquid electrolytes and can tightly hold the liquid electrolyte inside the GPEs and avoid the leakage. Several hydrogen-bonding polymers have been investigated for the application of GPEs in LIBs, such as poly(vinyl alcohol) (PVA) and polyimide (PI). A polymer blend of 25 mol % poly(acrylic acid) (PAA) and 75 mol % PVA is used to prepare GPEs by loading 70 wt % electrolyte, which contains an ionic liquid, 1-butyl-1-methylpyrrolidinium bis(trifluoromethanesulfonyl)imide (PYRTFSI), with 0.2 M LiTFSI. (144) This study focuses on blending PVA and PAA, which are self-associated, and the OH groups are involved in hydrogen-bonding interactions between the polymer chains. The results revealed that the hydrogen bonding is helpful for improving the miscibility of polymer blends and directly affects the thermal, mechanical, and optical properties. (144) In addition, the PVA-PAA GPE composites exhibited ionic conductivity of 2.0 × 10^–4 S cm^–1 at 25 °C and increased to 1.4 × 10^–3 S cm^–1 as the temperature was increased to 90 °C, which is typical of Arrhenius temperature dependence behavior (Figure 15a). Moreover, a further electrochemical study revealed that the PVA-PAA GPE composites had a Li⁺ transference number of 0.4 and stable electrochemical window up to 5.0 V vs Li/Li⁺. (145) In addition, by taking advantage of the OH groups on the PVA backbone, PVA was used and modified to prepare the single ion conducting polymer 6e. (27) The modification of PVA as a single ion conducting polymer is performed by physical cross-linking with boric acid through hydrogen-bonding and ionic interactions in the hydrated state. Although the PVA-based single ion conducting polymer possesses a relatively low ionic conductivity of 6.11 × 10^–6 S cm^–1, which is far from the practical requirement, it exhibits a stable electrochemical window up to 7 V vs Li/Li⁺, which is potentially compatible with higher voltage LIBs. (27)

In addition, recent investigations have reported the use of PIs to prepare GPEs for LIB applications. PIs are highly polar material and thus are expected to have strong coordination within polymer chains with polar solvents, such as EC and EMC, due to the electron donor and acceptor groups, which largely enhance the retention ability of electrolytes. (146) Moreover, PIs have been reported to have excellent thermal stability over 500 °C and are able to effectively avoid the short circuits caused by shrinkage at high temperature. (147,148) Wang et al. (147) used pyromellitic dianhydride (PMDA) and 4,4′-oxidianiline (ODA) to synthesize PI 11a (Scheme 11), which was then dissolved in N,N′-dimethylacetamide (DMAc) and used to prepared an electrospun fiber. After activation by immersing PI-based fiber into 1 M LiPF₆ in EC/EMC/DMC (1:1:1, v/v/v), the resulting GPE exhibited ionic conductivity of 2.0 × 10^–3 S cm^–1 at 25 °C and a stable electrochemical window up to 5.0 V vs Li/Li⁺. Furthermore, a GPE based on PI 11a exhibited higher electrolyte uptake of ∼559% and high retention ability of 0.62. Further investigation found that the high ionic conductivity could be associated with the good affinity of PI 11a, which holds liquid electrolyte via coordination through electrostatic interactions; thus, it results in higher electrolyte retention and uptake. (148) Maceiras et al. (149) reported the copolymerization of two aromatic diamines, 1,3-bis-2-cyano-3-(3-aminophenoxy)phenoxybenzene (diamine 2CN) and 1,3-bis(3-aminophenoxy)benzene (diamine 0CN) with 4,4′-oxidiphtalic anhydride (ODPA) to synthesize a new amorphous PI 11b for elucidating the effect of the cyano group (-CN) on the electrochemical performance of GPEs. (149) The resulting PI 11b was further prepared as a fiber using electrospinning process by dissolving in N,N-dimethylformamide and DMAc (1:1, v/v). After being activated by immersion into 1 M LiPF₆ in EC/DMC (1:1, v/v), the resulting GPE exhibited an ionic conductivity of 0.7 × 10^–3 S cm^–1 at room temperature and a higher degree of porosity up to ∼80%. The study revealed that the ionic conductivity values are influenced by the presence of -CN dipolar groups in PI 11b; however, the specific mechanism was not explained clearly in this study. (149)

Poly(ether imide) (PEI) 11c is expected to also have good affinity between the skeleton materials and the polar liquid electrolyte due to the existence of ether bonds and amide bonds as well as the similar structure and polarity with PI. (150) Recently, a GPE composite based on 11c and halloysite nanotubes (HNTs) as fabricated by electrospinning technology. After being activated by immersing into 1 M LiPF6 in EC/DMC (1:1, v/v), the corresponding GPE exhibited ionic conductivity of 5.30 × 10^–3 S cm^–1, a lithium transference number beyond 0.5, and a stable electrochemical window up to 5.4 V vs Li/Li⁺ with only 1 wt % HNTs. As shown in Figure 15b, the GPEs based on 1 wt % HNTs in PEI exhibited better electrochemical performance with a higher capacity in Li/GPEs/LiCoO₂ cells with different charge rates as opposed to the conventional Celgard separator and pure PEI film. The investigation revealed that the excellent electrochemical performance was due to synergic advantages of PEI and HNT filler. The abundant carboxy groups of PEI molecules formed a hydrogen-bonding interaction with the liquid electrolyte and possessed a high electrolyte uptake, which assisted in the transmission of Li⁺ ions in the electrolytes. In addition, the abundant hydroxyl groups on the external surface and the lumen of HNT were beneficial to decreasing the complex between the Li⁺ ions, thus promoting the lithium salt dissociation. (150) In summary, hydrogen-bonding polymers gave a new opportunity to further enhance the electrochemical performance of GPEs by increasing liquid electrolyte uptake and tightly holding it inside of the GPEs via hydrogen bonding.

4.3. Biodegradable Polymer

The increasing focus on clean energy storage has led to more environmentally friendly production of LIBs. However, well-known polymers for GPEs. including PEO, PAN, PVDF, and PMMA, are all derived from the petroleum industry and nonbiodegradable, which will generate new environmental waste as the LIBs are discarded. (151) Therefore, many scientists are investigating biodegradable polymers as an alternative polymer for GPEs. Cellulose has therefore drawn considerable attention due to its outstanding properties, such as biocompatibility, biodegradability, chemical stability, and good mechanical strength and thermal stability. (152) A GPE based on cellulose as a polymer host using a simple coating of PVDF on both surfaces of methyl cellulose (MC) to form a sandwich structure was reported by Xiao et al. (31) The GPE based on PVDF/MC/PVDF structure shows an ionic conductivity of 1.5 × 10^–3 S cm^–1 at 25 °C, a lithium transference number of 0.47, and a stable electrochemical window up to 4.8 V vs Li/Li⁺ after activation by immersion into 1 M LiPF₆ in EC/DEM/EMC (1:1:1, w/w/w). A stable cycling behavior of LiFePO₄ using GPE sandwiched with PVDF/MC/PVDF is presented in Figure 16a and indicates good compatibility with the cathode electrode. This study shows that the synergic effect of PVDF polarity and hydroxyl groups on the surface of MC could hinder PF₆^̅ anion mobility and promote the movement of Li⁺ ions. (31) Furthermore, carboxymethyl cellulose (CMC) has already been used to prepare porous GPE 12a by Zhu et al. (Scheme 12). (152) Although the porous CMC-based GPE resulted in relatively lower ionic conductivity of approximately 4.8 × 10^–4 S cm^–1 at 25 °C and low electrolyte uptake for only 75.9 wt %, it showed a high lithium transference number up to 0.49 and better cycle performance for CMC-2 (GPE with 2 mL of N,N-dimethylformamide (DMF), porosity agent) compared with commercial Celgard 2730 (Figure 16b). Another type of cellulose polymer, HEC, has also been explored to identify its performance as a GPE for LiBs. (153) Interestingly, the HEC-based GPE formed a dense membrane without evidence of pores, although it also showed electrolyte uptake of up to 78.3 wt %, ionic conductivity of 1.8 × 10^–4 S cm^–1 at 25 °C, and a lithium transference number of 0.48 after activation by 1 M LiPF₆ in EC/DMC/EMC (1:1:1, w/w/w), which is close to the performance of the CMC-based porous GPEs. Furthermore, the study revealed that the hydrogen-bonding interaction of carbonate-based liquid electrolytes with the polymer matrix is beneficial for retaining the liquid electrolyte inside GPE and promoting fast Li⁺ mobility as shown in 12b. (153) The interaction between the solvent and cellulose polymer host on the 12b has also been applied to explain the electrochemical performance of GPE based on bacterial cellulose (BC). (154) The BC-based GPE was prepared through an ecofriendly fast freeze-drying method that does not involve any toxic or costly solvents as shown in 12c. After activation by immersion in liquid electrolyte with 1 M LiPF₆ in EC/DMC (1:1, v/v), the BC-based GPE exhibited an ionic conductivity of 4.04 × 10^–3 S cm^–1, a lithium transference number of 0.514 at 25 °C, and a reversible capacity of 75.2 mAh g^–1, and it was maintained under a high rate of 9 C, which indicates that the BC-based GPE is suitable for fast-charging LIBs. (154)

Another biodegradable polymer, polyurethane, has also been studied for GPE applications. Liu et al. (155) used polymer 12d to fabricate novel GPE composites based on cellulose-coated polyurethane by the solution-casting method. Although the resulting GPE showed relatively low ionic conductivity of approximately 2.3 × 10^–4 S cm^–1 at 25 °C, it exhibited stable discharge capacity of the Li/LiFePO₄ cell of approximately 128.2 mAh g^–1 after 200 cycles and 91% of the capacity retention at a charge/discharge rate of 2 C at 80 °C (Figure 16c). The study suspected that the reaction between the O–H group of the cellulose and polar groups of thermoplastic polyurethane (TPU) formed a stable structure and the cellulose provided more channels of ions, thereby facilitating ion transportation and enhancing the electrochemical performance. (155) Biodegradable poly-ε-caprolactone (PCL) has also attracted attention among the scientists as a polymer electrolyte. A polymer electrolyte for zinc batteries was implemented by Sownthari and Suthanthiraraj via solution casting of PCL and zinc triflate [Zn(CF₃SO₃)₂]. (156) This study shows an ionic conductivity of 8.8 × 10^–6 S cm^–1 and electrochemical window stable up to 3.7 V at 25 °C. Although the study shows a very low electrochemical performance for zinc batteries, it opens new opportunities for the further development of PCL-based GPEs for LIB applications and provides insights for the development of environmentally friendly LIBs.

5. Conclusions and Perspectives

Because of their characteristics as liquid and solid electrolytes, GPEs have received increasing research attention because of their high ionic conductivity (above 10^–4 S cm^–1), wide electrochemical window, favorable mechanical properties, thermal stability, and compatibility with electrodes during cycling. The general requirements and crucial properties of GPEs for LIB applications have been discussed in detail. Furthermore, the transport mechanism of Li⁺ ions have also been explained in detail. With the great potential of applying GPEs for LIBs, recent progress on GPEs based on different polymer types has been carefully reviewed. In general, polyether, polyvinyl, polynitrile, polycarbonate, and polyacrylate are the most common polymer types for GPEs. However, each of these substances has pros and cons. Many studies focused on structural design of GPEs have been performed to improve the electrochemical performance and fulfill the requirements. Although the strategic design of GPEs based on composite formation and co-/terpolymer leads to improved ionic conductivity associated with the morphology and electrolyte uptake, a bulky terpolymer can lead to instability issues and result in low ionic conductivity. On the basis of structural design, the incorporation of ionic liquid and single ion conductor polymer represents a promising strategy to enhance the ionic conductivity and Li⁺ transport properties of Li batteries. The ionic liquid provides the resulting GPEs with greater electrochemical and thermal stability compared with conventional liquid electrolytes. In addition, the single ion conducting polymer represents another method of Li⁺-ion transport and provides an immobile negative charge carrier in which the concentration of polarization could be minimized during the cycling process, thereby enhancing Li⁺ transport.

Alternative polymer types, such as polyaromatic, hydrogen-bonding, and biodegradable polymers, have also been carefully reviewed for GPE applications in LIBs to further increase the electrochemical performance and environmentally friendliness. The rigid structure of polyaromatic polymers offers good mechanical properties and is expected to partially solvate the Li⁺ through aryl groups; hence, the concentration polarization can also be inhibited. Furthermore, higher electrolyte uptake could be achieved by forming GPEs with more porous structures due to the incompatibility issue observed with mixing aromatic and aliphatic polymers. Hydrogen-bonding polymer has unique characteristics for the application of GPEs to LIBs. The hydrogen-bonding interaction between the polymer chain and electrolyte allows the corresponding GPEs to tightly hold the liquid electrolyte, thereby preventing leakage and providing additional pathways for Li⁺-ion transport. This biodegradable polymer has drawn considerable interest due to increasing environmental concerns caused by the mass production of LIBs. The application of biodegradable polymers for GPEs is expected to reduce the impact of LIBs after their use and subsequent disposal in the environment. Overall, the molecular-level strategic structural design of a GPE based on its polymer type provides an opportunity to impart specific characteristics and properties that can further enhance the electrochemical performance and stability of the resulting GPE to match the properties required for LIBs.

【Article link】

https://doi.org/10.1021/acsaem.9b00295