Application and structure of carbon nanotube and graphene-based flexible electrode materials and assembly modes of flexible lithium-ion batteries toward different functions

Abstract

In recent years, the rapid development of portable/wearable electronics has created an urgent need for the development of flexible energy storage devices. Flexible lithium-ion batteries (FLIBs) have emerged as the most attractive and versatile flexible electronic storage devices available. Carbon nanotubes (CNTs) are hollow-structured tubular nanomaterials with high electrical conductivity, large specific surface area, and excellent mechanical properties. Graphene (G) is to some extent comparable to CNTs, because both have unlimited value in flexible electrodes. Herein, a systematic summary of the application of CNT and G in FLIBs electrodes is presented, including different functional applications and services at different temperatures. Furthermore, the effects of electrode structures, including powder, wire-shaped, and film-shaped structures, on electrochemical properties is highlighted. The assembly structures of the FLIBs consisting of CNT and G-based flexible electrodes to realize different functions, including bendability, stretchability, foldability, self-healing, and self-detecting, are systematically reviewed. The current challenges and development prospects of flexible CNT and G-based flexible electrodes and corresponding FLIBs are discussed.

Graphical abstract

Keywords

flexible lithium-ion batteries (FLIBs) / carbon nanotubes (CNTs) / graphene (G) / electrode structure / function

1 Introduction

With the gradual miniaturization, lightness, and flexibility of portable electronic products, flexible electronic products have emerged [1,2] and are widely used in communications [3], healthcare [4,5], wearable consumer electronics [6–8], electronic skin, bendable displays [9], flexible smart watches, etc. [10–12]. High-performance flexible lithium-ion batteries (FLIBs) play an indispensable role as a power source in flexible electronic devices. FLIBs are the core components of flexible electronic products, with a high energy density, high power density and stable cycling performance [13–19] and an outstanding flexibility such as bendability, stretchability, and foldability [20–30]. FLIBs have great advantages over conventional batteries in terms of resistance to mechanical deformation, allowing electronic products to perform well under various deformation conditions [31–34]. To overcome the problem of low efficiency, lots of researchers have been investigating how to achieve high energy densities by increasing the loading of active material and how to reduce the contact resistance by maintaining good contact between nano-active materials and other components. Carbon-based materials have excellent conductivities and can be used simultaneously as current collector, active material, and support matrices in flexible electrodes of FLIBs. The inert binder between the current collector and the electrode material in conventional FLIBs is eliminated. Therefore, the poor interfacial bonding between the collector and the electrode materials and side reactions can be effectively avoided, and the active material is better combined with the carbon-based collector to achieve high-rate performance and excellent cycle stability.

The current collector is a metal, such as aluminum (Al), which has a poor adhesion to the electrode material, resulting in the loss of capacity and low energy density when used in FLIBs. A large number of carbon-based electrodes are widely used in flexible energy storage devices, and carbon-based materials can significantly increase the proportion of active materials compared to conventional metal collector fluid, which in turn improves the overall performance of the battery. Carbon exists in the form of isomerism due to its valence state [35]. Carbon nanotubes (CNTs) are an advanced flexible energy storage material [36–38], with a high conductivity, an excellent electrochemical performance, excellent toughness, and a chemically adjustable surface [39,40], and are easy to be processed into flexible films/papers [41–48]. Widely used in medicine, electronic communication and other fields, especially in environmental protection and energy storage, CNTs have a promising prospect [35,49–52]. As a lithium embedded electrode material, CNTs have a tubular porous network structure, which can be internally embedded and externally deposited with other materials, and are the only choice for preparing FLIBs electrode materials [53,54]. Graphene (G) can be compared with CNTs to some extent as an excellent conductive substrate [32,53,55–59]. Its high stability, mechanical properties, and biocompatibility are compatible with FLIBs [60]. Therefore, CNT and G are often used as flexible substrates to eliminate the need for metal collector [30,58,61]. The capacity, rate performance, and cycle stability of FLIBs depend directly on the electrode materials. Related studies on the application of carbon materials to cathode electrode materials have been conducted by various teams. For the development of electrode materials, the improvement of the anode electrode is particularly urgent. Carbon materials usually play different roles in electrode materials as conductive agent [62–70], supporting skeleton [71–77] and active material [78–86], preparing electrodes with different structures [87–104], improving the mechanical stability and electrochemical performance of electrodes. The assembled FLIBs have different functions while achieving a high performance [97,99,105–118].

CNT and G have a wide range of promising applications in FLIBs [15,119]. For FLIBs, previous researchers have reviewed the preparation methods of flexible electrodes (vacuum filtration, chemical vapor deposition (CVD), solvothermal method, in-situ regeneration and drip) [15,21,120,121], selection of flexible electrode (CNT, G, carbon fiber, carbon cloth (CC), polymer and novel transition metal carbon/nitride two-dimensional nanolayered materials (MXene)) and electrolyte materials (composite solid polymer, flexible gel electrolyte) [15,30], surface modification [122,123], methods to mitigate lattice defects [124], FLIBs structural design (film type, fiber type, wavy type, island joint, and bamboo sheet) [21,30], and assembly methods (lamination, vacuum filtration, coating, winding, and printing) [15,21]. However, the functional application of new carbon materials in the electrodes of FLIBs lacks systematic summaries, and the insightful understanding of the relations of the structures of the FLIBs and their special functions is still lacked. In this paper, the applications of CNT and G in FLIBs electrodes are systematically summarized, including different functional applications and services at different temperatures. In addition, the effects of electrode structures, including powder, wire-shaped, and film-shaped structures, on electrochemical properties are highlighted. Moreover, the relations between the assembly structures of FLIBs composed of CNTs and G-based flexible electrodes, and some special functions including bendability, stretchability, foldability, self-healing, and self-detection are systematically reviewed.

2 CNT and G in FLIBs electrodes

CNT and G favor fast charging and discharging and reduced polarization. As a lithium-embedded material, CNT has a short length, a small depth, and a short stroke when embedded or unembedded. The polarization degree of the electrodes during charge and discharge at large current is reduced, which improves the fast charge and discharge capability of the battery. The application of CNT and G to FLIBs electrodes is illustrated in Fig.1.

In cathode and anode materials, CNT and G serve as conductors, supporting structures, and active materials. They usually have all of the three functions, but one dominates. However, there are few applications in cathode materials, mainly because the cathode materials tend to be lithium composites and polymers, and the types of lithium materials are limited. However, FLIBs suffer severe capacity losses under extreme conditions, i.e., low and high temperature batteries. The FLIBs anode and its flexibility are shown in Fig.2. The electrode is often used under low and high temperature conditions to show excellent electrochemical performance.

2.1 Functional applications

2.1.1 Conductive agent

To increase the conductivity of electrodes, carbon materials are repeatedly introduced to increase the contact between the active substances, reducing the contact resistance of the electrodes and reducing the polarization. Sun et al. [67] prepared paper composed of multi-walled carbon nanotubes (MWCNTs) and cellulose by utilizing the paper making method. Comparing the as-produced MWCNT paper (AMP) and the carbonized MWCNT paper (CMP), it was found out that carbonization improves the crystallinity and flexibility of MWCNT, with the decrease of macropores and the increase of micropores. The CMP capacity is increased. The impedance of CMP is smaller than that of AMP. The CMP electrode has an excellent cycling performance, an anti-current impact ability, and a reversible capacity of up to 500 mAh/g. This is due to the conversion of cellulose after carbonization treatment into amorphous carbon. Cheng et al. [64] prepared flexible anodes of Fe₃O₄/CNTs using cellulose oriented hetero-assembly, and cross-linked Fe₃O₄, nanoparticles (NPs) and CNTs to form conductive networks. At a strain of 7.1%, the average tensile strength (T_s) is 5.2 MPa, and the elastic modulus (E) is 73 MPa. From the 10th time, the capacity continues to increase to 898 mAh/g. Si/CNT electrodes were prepared with an initial discharge capacity of 1380 mAh/g. The carboxyl group from cellulose and CNTs form strong interfacial contact through hydrogen bonding, CNTs conductive frame, and heterogeneous assembly, ensuring the integrity of the electrode after repeated electrochemical reaction. The hydrogen bond network shows self-healing properties, which can repair the mechanical damage of the electrode during the cycle. Zeng et al. [70] prepared the GeTe/G/CNTs anode by using the ball milling method. After 100 cycles, a large number of uneven holes appeared on the surface of the GeTe/G electrode. When the electrode was fixed on CNTs, the conductivity and flexibility of the electrode were improved. After 100 cycles at 0.1 A/g, the area specific capacity remained at 0.99 mAh/cm², mainly because G alleviated the stress and hindered the accumulation of GeTe. CNTs act as electron transport lines. During the grout process, the −OH group connects with the −OH or −COOH groups of hydrophilic MWCNTs to form strong hydrogen bonds. Cellulose is pyrolyzed to form carbon fibers, which form conductive networks with MWCNTs through dehydration condensation during heat treatment.

CVD is commonly used to deposit active materials on flexible substrates. Yi et al. [69] prepared CNTs crosslinked flexible Si microspheres using CVD to improve electrode wettability. Compared to pure Si (32.6°), the cross-linked Si/CNT has a minimum contact angle of 0° and an improvement of 2.9 GPa in Young’s modulus. The wettability of the electrolyte is improved, making it easier to enter Si/CNTs. Even at 0.2 A/g, the capacity of 100 cycles is about 1.7 times that of Si. This is due to the three-dimensional (3D) cross-linked structure of the CNTs formation and increased wettability.

The mutual support and synergy between single-walled carbon nanotubes (SWCNTs), vertically aligned carbon nanotubes (VACNTs), CNTs, carbon nanofibers (CNFs), graphene foam (GF), and reduced graphene oxide (rGO) act as both conductive agents and ion transport pathways. Ren et al. [66] reported the preparation of SWCNT-coated graphene foam (SWCNT-GF) by dip coating, which formed a binary network structure and could be curled (Fig.2(a)). The SWCNT-GF has a specific surface area of 467 m²/g, an initial discharge capacity of 2492 mAh/g, and a capacity retention rate of up to 90% after 1000 cycles at 1 A/g. This is due to the formation of GF, SWCNT, and binary network structures, which provides a high conductivity, expands the contact area between electrodes and electrolyte, and efficiently utilizes the active material. Abdollahi et al. [62] grew VACNT on the rGO paper by plasma CVD to prepare a flexible film with an all-carbon structure. The VACNT/α-rGO electrode has about 11 times the specific surface area of GO compared to GO. The discharge capacity reached 459 mAh/g, and the Coulomb efficiency (CE) is 100% after 100 cycles. This stems from the fact that low-defect sp²-carbon-hybrid VACNT with “tip effect” growth acts as a conductive pathway in addition to surface enhancement, and the porosity of rGO facilitates ion transport. Huang et al. [65] prepared self-sustaining N-doped CNF/CNT films with a high conductivity by electrospinning. The specific surface area of the film is 357.4 m²/g, about 2.5 times that of CNF, and the initial capacity is more than 3 times that of G. The reversible capacity of 100 cycles at 0.05 A/g exceeds 1099.5 mAh/g, and the conductivity is 6 times that of CNF, respectively. After the battery is made, it is bent at different angles (0°, 45°, 90°, 135°, and 180°) and the impedance of 60 cycles is slightly reduced from 0°, which can provide a capacity of 463.6 mAh/g. Based on CNF/CNT and LiNi_0.5Co_0.2Mn_0.3O₂, FLIB can light a group of 12 “C” type leads under flat, 90°, 180°, and 360° bends. This is due to the increase of defects after mixing N, the enhancement of electrochemical reactions, and the cross-linking of CNT and CNF. As can be seen, the double carbon electrode performs better.

Since then, researchers have frequently used di-carbon to modify NPs with a large volume dilation. The surface area of the layered CoS₂/CNT/G electrode (CCG) prepared by Xu et al. [68] based on hydrothermal reaction is about 271 times that of CoS₂, and the initial discharge capacity is as high as 993 mAh/g at 100 mA/g. After 100 cycles, the discharge capacity of CCG is about 5 times that of CoS₂, even at 2000 mA/g. In addition, the resistance of CoS₂ is about 1.5 times that of CCG. This is due to the fact that the abundant oxygen-containing groups of the G provide sufficient nucleation sites in the hydrothermal reaction process, effectively protecting CoS₂ from aggregation and high degree of graphitization with CNTs. Cao et al. [63] prepared MoS₂/CNF/CNT flexible films based on solution papermaking and the in-situ carbonization method (Fig.2(b)). After the film is folded four times, it can still maintain a good mechanical stability. After carbonization, the fracture is relatively flat and the lamellar structure is obviously weakened. During the carbonization process, the microstructure of the heterogeneous matrix is rearranged and the ordered stacked structure is transformed into a disordered porous structure with a specific surface area 13 times larger than the original one and a pore size of 3.83 nm. The initial specific discharge capacity of the film is up to 930 mAh/g, and its porous and fibrous structure is maintained even after 100 cycles at 400 mA/g. This stems from the fact that CNF is used as a bio-based binder and skeleton, while CNT is a conductive filler that avoids MoS₂ aggregation. The introduction of CNT and G further enhances the conductivity. In addition to increasing electrical conductivity, it can sometimes serve as a flexible skeleton to support the active substance.

When CNT and G are used as conductive agents, the conductivity of the electrode is increased, the transmission of Li⁺ is accelerated, and the impedance is reduced.

2.1.2 Flexible skeleton

When CNT and G are installed in a matrix of materials with a high specific capacity, the specific surface area and pore volume of carbon skeleton can be increased to improve the mass load of active materials and the energy density of batteries. The hollow structure of CNT can be filled inside or deposited outside, which can solve the problem of decreasing CE due to material volume change.

The film electrodes are prepared by filtration and heat treatment using the solution method, and can be expanded and contracted with a certain fracture strength in addition to bending. Zhao et al. [77] (Fig.2(c)) used filtration and heat treatment to prepare stretchable CoSnO₃/G/CNT composite paper with a free ternary layered stacked structure of 32 μm in thickness. The film can be bent and the specific surface area is about 2.5 times that of pure CoSnO₃, and the reversible capacity of CoSnO₃ nanotubes/G/CNT and CoSnO₃ nano-boxes/G/CNT at 0.1 A/g are 925.1 and 1098.7 mAh/g, respectively, and the reversible capacity is 676.7 mAh/g when increased to 2 A/g. This results from the cross-linking of G and CNT to form a 3D network structure, and the amorphous CoSnO₃ is uniformly and tightly distributed between the surface of the 3D G/CNTs and the scaffold. LiFePO₄ (LFP) and LiMn₂O₄ (LMO) are often added to combine with CNT or G to improve electrode performance. The modification of double carbon greatly limits the shedding and volume expansion of the active substance. The LFP/CNT/rGO film prepared by Zhang et al. [124] through vacuum filtration exhibits a mesoporous structure and can be bent (Fig.2(i)), the specific surface area is 11 times that of LFP, the capacity of 100 cycles at 0.2 C is almost not reduced, and the capacity remains 98 mAh/g at 10 C. The conductive network constructed by CNT and rGO not only enhances the electrical conductivity but also acts as a mechanical support and current collector. There are van der Waals forces between LFP and carbon material. However, the ion diffusion rate is poor, which limits the research and development of LFP, and still needs further research.

At the same time, Bao et al. [125] prepared LiMnTiO₄/MWCNT anode by vacuum filtration, and studied the electrochemical performance of the electrode with a CNT content of 30% (CNT30) and 50% (CNT50). The film diameter is about 40 mm and can be bent into an arc shape without obvious fracture. The T_s are 1.34 and 2.04 MPa, respectively. The weight of CNT30 is only half that of the conventional electrode, and the capacity of CNT50 is 1.26 times that of the original electrode after 50 cycles at 0.5 C. Fang et al. [126] embedded LiNi_0.5Mn_1.5O₄ into MWCNTs by vacuum filtration, which provides a capacity of more than 80% at 20 C, and the polarization resistance is less than 25% that of conventional metal-based electrodes. Compared with conventional LiCoO₂ (LCO) and LFP, the voltage is reduced by only 0.026 V at a bending radius of 1.6 cm. The electrical conductivity and power density of the conventional electrode are more than 2 times. Gu et al. [113] grew LMO crystal in-situ on the CNTs film as the cathode electrode of stretchable LIBs. The discharge capacity of the battery composed of electrodes reaches 97 mAh/g, and the performance is stable after 150 cycles, with a capacity loss of 0.04% per cycle. Even after 1000 stretch-release cycles, the resistance does not increase significantly (less than 6%). This electrode is highly reversible compared to the stretchable LIBs, which are interconnected with snakes, and other “wave” electrodes, which are not chemically bonded, due to the chemical bonds between the flexible substrate of CNT and the LMO.

Polymer electrode has become a research hotspot of FLIBs anode materials due to its high energy density, adjustable structure, and good flexibility. Because of its rich green resources, it is considered as a kind of material to replace traditional inorganic materials. At present, conducting polymer, carbonyl polymer, self-healing polymer and imine polymer are widely studied. Yang et al. [127] prepared the cathode electrodes of perylene dimer (PDI) and SWCNTs by acid-assisted vacuum filtration. After acid treatment, the active substances were evenly distributed in SWCNTs, the π–π bond between PDI and SWCNTs formed a highly interconnected network, so that the electrode could be charged and discharged for more than 2000 times at 500 mAh/g. The capacity decay rate was almost zero each time. In addition, the capacity was retained 90.6% after 300 cycles at 0° to 180° bends. Polyimide (PI) is an insoluble, thermally stable, and electrochemically active conjugated carbonyl material. Wu et al. [128] prepared PI/SWCNT cathode (Fig.2(j)) by in-situ polymerization. The initial capacity of PI/SWCNT was 206 mAh/g, and the capacity remained 85% after 200 cycles. The electrical conductivity was 10 times that of pure PI, and the film did not change significantly even after 100 bending. In addition, the polyimide derivative pyromellitic dianhydride tris(2-aminoethy)amine (PMTA)/SWCNT@SWCNT cathode was prepared from PMTA by using the rolling method [129]. When the content of SWCNT was 10.9%, the SWCNT film was tightly bound to the active material layer, and the nucleation sites increased. When the reaction temperature reached 200 °C, the capacity of the electrode was 163 mAh/g, and 80% of the capacity could be retained even bending 1000 times. In addition, the T_s of the electrode after being pressed by the SWCNT film was increased by 8 times, which was due to the porous structure constructed by the SWCNT, ensuring the elasticity and electron transport of the polymer. It can be seen that the flexibility of the cathode electrode of the polymer has been greatly improved, but its capacity is relatively low, and further research is needed.

In addition, Xiang et al. [74] grew 1T-MoSe₂ in-situ on SWCNT thin films using the solvothermal method. The C−O−Mo bond between the two formed a strong electrochemical coupling, so that the electrode was cycled 100 times at 300 mA/g with a capacity of 971 mAh/g, and the structure was not broken. The elongation of the film could reach 22.3%, and the fracture strength was about 34 MPa. This is due to the network structure constructed by oxygen-containing functional groups and bonding on SWCNT, which provides fixation and nucleation sites for MoSe₂ growth. The increase of MoSe₂ layer spacing is conducive to the deintercalation of Li⁺ and provides abundant active sites. Wei et al. [117] prepared LCO/CNT film, and when bending to 1 mm and cycling 1000 times, LCO/CNT did not crack and delamination. This is due to the rough porous surface of the CNT film, which facilitates rolling embedding of the active substance (Fig.2(h)). With the increase of load after bending 1000 times, the cracks between the active materials become more obvious. The strain Al of the CNT film was 10 times, and the weight energy density of CNT film was about 1.6 times that of metal foil. This is because the introduction of CNT not only improves the conductivity of the electrode, but also acts as a flexible skeleton to prevent electrode deformation.

Moreover, the prepared paper-like anode not only increases the flexibility of the electrode, but also restores its original state. Fan et al. [71] prepared nickel sulfide paper/CNT (NS@CNT) by electrodeposition. The NS@CNT paper was subjected to 0°, 180°, and curled state, and it was found that it could be restored to the original state even after being curled into a ball. In addition, the capacity was about twice that of Ni₃S₂ at 60 mA/g, and the CE was as high as 98% after 100 cycles at 300 mA/g. This is due to the reversible conversion reaction of NS and the capacitive effect of the CNT skeleton. However, the tensile property of the paper is low and inferior to the G@CNT@MoS₂ composite foam synthesized by using the hydrothermal method by Ren et al. [72], which can be easily bent without cracking, with a slight increase in resistance at 180° of bending, but can be fully recovered after straightening, with negligible change in resistance after 5000 cycles of bending-straightening. The T_s of the electrode was 6.6 Pa, which was 2.2 times that of pure GF, and provided a specific capacity of 606 mAh/g even after 200 cycles at 0.2 A/g. This arises from the superposition of the π–π bonds between the GF endoskeletons and the network constructed by the van der Waals forces between the CNT coatings.

Based on carbon coating, Yang et al. [75] transformed mesoporous SiO₂ coated on CNTs network into a mesoporous Si layer covered with carbon to obtain sandwich C/meso-Si/CNT skeleton composed of core-shell structure of composite sponge electrode (Fig.2(d)) which could be compressed to large deformation without collapse and could be restored to the original shape and compressed up to 60% of strain. When compressed to 40% cycled 1000 times without stress degradation and structural damage, the specific discharge capacity of the sponge at 0.5 A/g was about 6 times that of pure CNTs, which corresponded to a volumetric capacity of 19.5 mAh/mL for the electrode, and the capacity remained to be 1800 mAh/g for 500 cycles, and remained to be 1700 mAh/g even at 4 A/g. This stems from the CNTs backbone and the flexible C coating to mitigate the Si volume change, which protects the structural integrity. Yildiz et al. [76] deposited polymethyl methacrylate (PMMA)-Si nanofibers on CNTs sheets by electrospinning and CVD, and deposited a 20 nm thick pyrolytic carbon layer on the surface of CNT-Si hybrid plate to form CNT-Si-C fabric. Si was dispersed throughout the CNTs fabric in a lamellar structure that could be repeatedly bent or folded without any damage (Fig.2(e)). The CE of CNT-Si-C fabric was 73.9% at 100 mA/g in the first cycle, and the discharge capacity of 100 cycles was 7.5 times that of CNT. Although CNT-Si-C has an excellent performance, there were some wrinkles on the surface after repeated bending and folding, both of which resulted from CNTs as mechanical support, high porosity structure, and flexible carbon coating. The carbon coating can be seen to adequately relieve the stress from volume change, resulting in a superior flexibility and electrochemical properties of the electrode. The addition of polymers provides a high capacity while acting as a flexible protective coating similar to the dual carbon modification. Wang et al. [73] prepared a one-dimensional coaxial polyaniline (PANI)@SnO₂@MWCNT anode. The capacity retention of 100 cycles at 0.2 A/g was about twice that of SnO₂/MWCNT, even though the CE of 150 cycles and 350 cycles at 1 A/g was over 99%, much higher than that of graphite, and there was no obvious structural fracture, with a conductivity of about 20000 times that of SnO₂. This is due to the fact that the MWCNT skeleton and the conductive PANI act as flexible protective coatings, the conductive network formed by the π bond between them, and the hydrogen bond between SnO₂ and PANI, which jointly promotes the tight coating to improve the performance of the composite.

CNT and G provide a flexible substrate for the active material as a skeleton, making the electrode exhibit superior mechanical properties and reducing the crushing and falling off of the active materials.

2.1.3 Active material

As an active material, carbon has a large specific surface area and fine pores, which is another big leap forward in terms of LIBs, with enough space to provide a large number of active sites when performing electrochemical reactions.

Wu et al. [130] constructed crossed graphitized CNT (G-CNT) anode electrodes by template synthesis. When compressed into thin paper, the G-CNT structure exhibited relatively good mechanical properties. Meanwhile, the electrode structure showed a hydrophobic surface with a static contact angle of 146.7°, on which water droplets can stay for a certain time. The electrode had a specific capacity of up to 624 mAh/g at 100 mA/g and a CE of up to 99.4% after 400 cycles at 500 mA/g. This stems from the layered structure/increased interlayer spacing to effectively insert Li⁺. Wang et al. [131] demonstrated a super-arranged CNT (SACNT) film collector. The surface density of the film was as low as 0.04 mg/cm², with a thickness of less than 1 μm, which could be folded. Compared with the conventional graphite-Cu electrode, the contact angle (19°) at the graphite-CNT interface was much smaller than that of graphite-Cu, the shear strength was 3 times that of graphite-Cu, Young’s modulus and T_s were increased by about 10 times, and the weight energy density was increased by more than 180%. This is due to the porosity and interfacial wettability of CNT, making the graphite layer and CNT in closer contact. This shows that CNT can not only be used as a conductive substrate, but also as a supporting skeleton. The carbon-based materials in electrode materials often play two or more roles, including conductive agents, flexible skeleton, and active materials. Lee et al. [79] found through direct prelithiation (DP) of the CNTs film that DP caused changes in the crystal structure of CNTs, which made lithium directly contact with the CNTs film. The DP-CNT electrode was cycled 350 times at 0.5 C, and its capacity increased to 1520 mAh/g, which was almost 3 times of the initial value, and the impedance was half of the original CNT. In addition, made of the sac cell, the fold could work normally. The vertically oriented CNT were more ordered than the chaotic ones. Wang et al. [82] prepared the VACNT-Si/CC electrode, although the initial discharge- and charge-area capacity of the electrode were 4.25 and 3.3 mAh/cm², respectively, which were relatively low for practical applications. However, after DP, the electrode could provide a reversible area capacity of up to 3.33 mAh/cm², which was more than twice that of CC, and the capacity retention rate was as high as 94.4% at 1 mA/cm² after 200 cycles. When the current density increased to 25 times of 0.2 mA/cm², the area capacity could still retain 55.3%. The LED lit up even when the battery was folded. This stems from the physical advantages of DP processing the layered 3D composite array formed by CNT and CC.

The coating-based protection prevents the electrode structure from cracking. Yu et al. [84] prepared polydopamine-normal dodecanethiol coated MWCNT film (PCMF) coated with polydopamine n-dodecyl-mercaptan (NDM) with a thickness of 13 μm. The film showed a low mass density and could bend or roll without cracking and wrinkling. The CE was still up to 99.5% even after 500 cycles, but the specific capacity was low. Thus, the PCMF electrode prepared by mixing hollow structural silicon (HSS) into PCMF showed little change in capacity even when bent at 30°, 90°, or 180°, and even when cycled 200 times with a specific capacity up to 750 mAh/g. Fu et al. [78] successfully prepared the oriented CNT-based Si composite flexible electrode. The Si coating was uniformly deposited on the CNT to form a porous structure, showing a good flexibility and the diameter of the oriented CNT reached 300 nm, 6‒30 times that of CNT. Even after 30 cycles, the CE reached 97%. At the same time, the repeatable deformation of CNT-Si-C was studied, and the hypothesis that the wavy deformation could reduce the strain and stress caused by the large volume change of crystalline silicon was proposed, which was helpful to maintain the structural stability of the electrode. Liang et al. [81] used silk as natural carbon source to prepare sandwich-like C@Si@G electrode by using the chemical solution method and carbonization method. Si was loaded on the fiber surface to form a flexible network and the electrode could be bent (Fig.2(f)). Compared with C@Si, the reversible discharge capacity of 120 cycles at 1 A/g was 7 times that of Si. After 300 cycles at 0.2 A/g, the capacity reached up to 1070 mAh/g, showing a high-rate performance and conductivity, and its resistance was significantly less than Si. This is due to the heteroatomic and functional groups in the outer layer of G and silk that can increase the active site.

Zhang et al. [85] prepared the SnS_2/GO-CNT (SGC) paper electrode through the combination of vacuum filtration and thermal reduction. The SGC paper was not damaged after bending back and forth, and its surface area was 10 times that of SnS₂ paper. Compared with SG electrode, The optimized SGC paper has a capacity of up to 1118.2 mAh/g at 0.1 A/g and a rate performance of up to 634.6 mAh/g at 2 A/g, with a lower contact resistance and charge transfer resistance (R_CT). This is due to the increased porosity and specific surface area of the sandwich structure and its addition of CNT and GO. At the same time, Li et al. [80] constructed MnO NPs@MWCNTs/rGO sandwich-shaped anode electrode through electrostatic interaction and vacuum filtration. The electrode could be bent. Even after 500 cycles at 2 A/g, it showed a “U” shape, but the capacity stabilized at 530 mAh/g after 500 cycles, despite of the conventional capacity attenuation. This results from the fact that the acidified MWCNTs act as carbon coatings, adapting to the huge volume expansion. Additionally, rGO is the reason for the formation of independent paper, which improves the integrity of the entire electrode and transverse conductivity.

In summary, it can be seen that CNT and G act as conductive agents, scaffolds, and active substances in flexible electrodes. Although one action is often dominant in the electrode material, it also plays a dual role. The non-self-supporting effect of the electrode material is fully improved, and the electrode exhibits superior mechanical and electrochemical properties (Tab.1).

2.2 Applications at special temperatures

2.2.1 Application at low temperatures

Capacity loss of batteries is severe at low temperatures. Therefore, researchers have been studying how to maintain or even increase the capacity of batteries at low temperatures while maintaining excellent mechanical properties. Liu et al. [116] prepared horizontally oriented carbon nanotube macrofilm (HOCNM) by utilizing the surface/interface engineering method, which had a good wettability, a load capacity of 10 mg/cm² which exceeded 700 mAh/g, and a weight energy density of 1.5 times that of metal-based LIB (MLIB). Even after bending and folding, the voltage did not change significantly. When the FLIB was worn on the wrist, the light glow did not change significantly. After 300 h, the operating voltage dropped from 2.623 to 2.52 V, and the capacity retention rate was about 86% that of the MLIB. At a low temperature (−35 °C), the capacity retention rate of the battery was 100%. When the temperature dropped to −40 °C, the capacity retention rate of the battery was more than 87%, which was lower than that of −35 °C. However, the battery could work stably, and could recover to the original performance after returning to room temperature. This stems from the high conductivity and porous array of HOCNM, and the wettability of electrolytes.

Electronic devices badly need to be powered for investigations in the South and North Poles. Therefore, the use of batteries at ultra-low temperatures has become the future direction for FLIBs development.

2.2.2 Application at high temperatures

Since the capacity and flexibility of batteries are maintained at low temperatures, the researchers set out to see if their performance would be compromised at higher temperatures. Yuan et al. [132] prepared 3D NCF@CNT to carry TiO₂ to increase the service life of the battery. TiO₂ was in-situ anchored in the NCF@CNT skeleton with macro and mesoporous structure (Fig.2(h)). The skeleton was rough and slightly folded, with a specific surface area of about 1.5 times that of TiO₂, and electrode flexibility increased. The impedance was about a quarter of the original. Even after 2500 cycles at 10 C, the capacity attenuation of each cycle was only 0.0037%, especially when the thermal stability of more than 100 cycles was achieved at 55 °C, the capacity retention rate was as high as 94%. This results from a dense, electrically conductive network of CNT wraps and NCF interconnects. The specific discharge capacity of HOCNM cells prepared by Liu et al. [116] at a high temperature (70 °C) was similar to that at room temperature. In addition, the electrodes of LCO/CNT and LTO/CNT studied by Wei et al. [117] in gel polymer electrolyte (GPE), polyvinylidene fluoride-tri-fluoroethylene-chlorofluoroethylene (PTC polymer), Li_6.4La₃Zr_1.4Ta_0.6O₁₂ (LLZTO), LiTFSi, and liquid electrolyte gel electrolyte and diaphragm could eliminate the safety threat caused by internal short circuit when the temperature was below 160 °C and improved the performance of the battery. This is mainly due to CNT and the electrolyte stability, which extends the temperature range of the battery. The high temperature use of the FLIB is the way forward, but it needs to be further explored if it can be used at ultra-high temperature.

The studies found that high temperature has an impact on the performance of electrodes and batteries. However, the offset of the effect of temperature has become the main problem for researchers to solve.

3 Correlation between electrode structure and performance

Flexible electrode materials are commonly used in dielectric elasticity or FLIBs. These materials must retain excellent conductivity while being thin, capable of significant deformation, and exhibiting high stretchability. These electrodes can provide highs specific capacity even after multiple cycles. Ideally, electrode materials are highly conductive, fully compliant, and the substrate thickness can be made thinner. Depending on the different structural types of electrode materials, they can be classified as powder, thin film, and wire-shaped electrodes.

3.1 Powder electrode

3.1.1 Conventional powder electrode

Based on the stretchable LIBs structure, researchers designed serpentine interconnects, origami and spring fiber structures. Although the electrodes can withstand strains of about 300% along different axial directions without failure, the tensile durability of the structure is poor at high strains, for which the powder is deposited on the substrate. Yu et al. [87] prepared super-stretched CNT composite electrodes CNT/Ni_1/3Co_1/3Mn_1/3 (NCM) and CNT/LTO by coating the CNTs film and active material powder on a biaxial pre-stretched polydimethylsiloxane (PDMS) substrate. The cleaning of the CNTs surface and the van der Waals force enhance the interface adhesion between the PDMS surface and the CNTs film, and the structure of the curved film remains stable even after 10000 tensile cycles at 150% strain. When the 45° axial strain is 0‒150%, the normalized resistance of the CNTs/LTO electrode increases by only 2.1% after 2000 drawing cycles. The CNT/LTO and CNT/NCM fold structures (Fig.3(a)) extended along the strain axis to a continuous CNTs conductive network enables the electrode to have a specific capacity of up to 138 mAh/g at different coaxial 150% strain 5 C. In situ synthesis, Cai et al. [86] prepared flexible self-supported Si-CNT/G paper electrodes by mixing acid-etched micro-sized Al−Si alloy powder with CNT/G. The electrode had a sandwich structure (Fig.3(b)) that effectively buffered the volume change of Si. The 3D carbon structure constructed by CNT/G could be embedded with Si particles to increase the conductivity of the electrode. The Si-CNT/G paper is not only flexible, but also has a specific capacity of up to 1100 mAh/g at 200 mA/g, twice that of Si/G. The capacity of the electrode remains almost constant even after 100 cycles.

3.1.2 Core-shell structure powder electrode

The electrochemical performance of the powder electrodes has been improved. However, the random accumulation and uneven dispersion lead to the aggregation of the active material, which reduces the surface area involved in the reaction. This result increases the thickness of the material and reduces the overall efficiency. In contrast, the layered core-shell structure design can effectively address these challenges [133,134].

In recent years, Si has become a research hotspot. Wang et al. [91] prepared coral-like silicon (CL-Si @C/rGO) by high temperature magnesium thermal reduction of SiO₂ and wrapped with C/rGO coating. Si was coated with a uniform carbon layer (Fig.3(c)) and there were pores on the surface. Compared with Si, CL-Si @C/rGO could absorb free water, which would lead to water loss during initial heating and decrease its weight. The specific surface area was 228.94 m²/g, and the resistance at the interface was minimal. An eversible capacity of 739.1 mAh /g could be provided even at 2 A/g. At 1 A/g, and even 100 cycles, the capacity was about 2.6 times that of CL-Si, and the capacity loss of 3‒100 cycles was only 0.5%, which resulted from the synergy between coralline and derived carbon and the accelerated charge transfer of the outer pyrolytic carbon and rGO. However, the defect of this process is the excessive melting of the initial nanostructure in the process of magnesium thermal reduction. Since Si and Sn are in the same main family, Si and Sn base oxides are often combined according to the principle of similarity and compatibility. Abnavi et al. [88] synthesized SnO₂@α-Si nanowires on the CNTs paper. The initial charge−discharge capacity of the electrode with a thickness of 30 µm was about twice that of SnO₂/CNT. The 25 cycles demonstrated a mass load of up to 5.9 mg/cm², an area capacity of 5.2 mAh/cm², and a volume capacity of 1750 mAh/cm³. Alaf et al. [89] prepared Sn/SnO₂/MWCNT electrode by thermal evaporation and plasma oxidation. The initial discharge capacity of the electrode was up to 1241 mAh/g, and the capacity was up to 390 mAh/g even after 100 cycles. But the Sn/SnO₂ electrode was pulverized (Fig.3(d)), which could be solved by adding MWCNTs. Core-shell structure and carbon coating were used as a flexible substrate to reduce the quality and thickness of the electrode. Although Si can eliminate cracks, further exploration is needed due to electrochemical agglomeration, irreversible defect capture and side reactions.

Moreover, the volume expansion of the active material is limited by the carbon coating. The capacity provided by the G/MnO@C/G electrode prepared by Zhang et al. [92] using the citric acid assisted and thermal reduction method at 0.1 A/g was 1.5 times that of the theoretical value, and the capacity hardly declined even after 4000 cycles. The electrode structure is protected by carbon shell to withstand high internal stresses and is used for intelligent reconstruction of nanostructures. Li et al. [90] deposited a 10 nm thick poly-pyrrole (PPY) coating on MnO₂/rGO/CNT and synthesized a composite material with a capacity of up to 1748.1 mAh/g after 200 cycles at 100 mA/g, which was about 4.6 times that of MnO₂/rGO/CNT. It could cycle 1500 times even at 3000 mA/g. This is because CNT and rGO are both reaction materials and elastic substrates. The PPY coating is used to construct the shell, maintain the stability of the structure and interface, improve the surface chemical state of MnO₂, which is conducive to the formation of solid electrolyte interface (SEI) layer and alleviate the volume change, and is better than the Mn-based electrode reported in the past.

3.2 Wire-shaped electrodes

Film electrodes exhibit favorable electrochemical properties and flexibility, but their tensile properties are limited. Wired-shaped electrodes, on the other hand, possess superior tensile properties compared to film. Wired-shaped electrodes are available in fiber and yarn types, where the fibers can be readily spun into yarns, which are subsequently woven into textile structures capable of solidifying to withstand substantial deformations [110,135,136].

The flexible yarn-like electrodes prepared as shown in Fig.4 can not only be torsional but also have a certain T_s and can even be woven into a fabric. The Si-CNT yarn made by Sun et al. [137] could be woven into fabrics. The process was highly flexible and could be wound around copper rods without breaking (Fig.4(a)), while the torsion of the Si-coated CNT film was limited, and the α-Si was stripped from the CNTs surface, showing a low T_s, but much higher than that of traditional textile yarns. The yarn has an initial charge and discharge capacity of up to 2440 and 2200 mAh/g, and CE of 90%, five times that of graphite. This stems from the flexibility, porosity, and Si coating of CNT, which limits the sliding between CNT, increasing the stored energy of the battery by more than eight times. Ren et al. [115] encapsulated LTO and LMO NPs in two aligned MWCNT yarns. The yarn was light in weight, the linear density was 2 and 10 mg/m, respectively, and the structure was not significantly damaged even when the electrodes were changed into different shapes (Fig.4(b)). After 200 cycles of the yarn at 0.05 mA, the capacity retention exceeded 80%. The structural stability of the MWCNT/LTO composite yarn was enhanced by coating it with GO, while the capacity retention of the uncoated yarn was only 60%. This is caused by the network formed by the MWCNT and the close contact between the LTO NPs and the MWCNT. The above results indicate that the addition of the MWCNT powder to the LMO suspension is a more efficient way to prepare such composite yarns.

In addition, researchers conducted research on CNTF. Zhang et al. [138] spun CNTs (d = 12 μm) fibers (CNTF) by CVD, and formed 10 CNTF into spring-like fibers (Fig.4(c)). The fibers (0.1 μg) had a linear density of 0.033 mg/cm, and the CNTs were highly arranged in the coil along the spiral direction, with a diameter of 65 μm and a pitch of 30 μm, which made the fibers elastic and could be easily stretched and released to 100%. When the fibers were stretched, the coils were gradually elongated, the CNTs remained highly aligned, and the fibers returned to their original coiled structure after release. The initial resistance of the fiber was 0.19 kΩ/cm, when the tensile strain was up to 100%, the resistance was up to 0.25 kΩ/cm, and the resistance returned to 0.20 kΩ/cm after release. When the elongation of the fiber reached 305%, the ultimate tensile stress was 82.7 MPa. When the tensile strain was less than 100%, E was 9.8 MPa. This is due to the spiral arrangement of CNTs. Zhang et al. [139] used the solvothermal method to directly grow nano-flower-LITi₂(PO₄)₃ on CNTF as the anode electrode. The electrodes could be bent extensively without significant mechanical damage and loss of the active material. At the same time, the electrode provided a reversible volume capacity of 85.2 mAh/cm³ at 0.4 A/cm³, and the CE approached 100% when the current density exceeded 1.6 A/cm³. After 2000 cycles at 8.0 A/cm³, the electrode still had a capacity retention rate of 74.7% and a CE of 100%. This is due to the fact that the nano-flower shape is conducive to electrolyte penetration and improves the transport of Li⁺. It can be seen above that the CNT base electrode is made into a fiber or yarn shape to facilitate the stretching and bending of the electrode.

3.3 Film-shaped electrodes

Powder-coated electrodes prepared on the metal substrates, and the use of electrolytes, deformations can cause the active substance to detach, electrolytes to leak, affect the battery performance, and cause safety issues. To avoid the loss of active substances and the destruction of inter-particle connections due to deformation, rigid electrodes are often thin films, as shown in Fig.5. To achieve both self-support and high flexibility, assembly into a battery or attachment to a flexible matrix is a relatively efficient means. While realizing flexibility, the battery can be miniaturized to adapt to the application of more flexible micro-devices [95,140–142].

Because of the advantages of the structure and capacity provided by the pure CNT film, the electrode presents a superior flexibility and electrochemical performance. Chew et al. [143] prepared CNT film by vacuum filtration, and compared SWCNT, double wall (DWCNT), and MWCNT film. The film was flexible and the capacity of the SWCNT and DWCNT was 3.2 and 2.8 times of that of MWCNT, respectively. But the capacity attenuation of the SWCNT and DWCNT was much higher than that of MWCNT, which allowed a capacity of up to 10 C of 175 mAh/g. The film is a network structure formed by the interaction of curved CNT. MWCNT had a large diameter and the thickness of the film was up to 61 μm, which solved the influence of polymer adhesive. Yoon et al. [100] used CVD and spinning to synthesize CNT film, which could be wound on a round rod with a diameter of 7 mm after high temperature heat treatment. Although the specific surface area was reduced, the tension exceeded 13%. When the current density increased from 0.5 C to 10 C and then to 1 C, the electrode capacity basically recovered to about 2 times of the original CNT, which was due to the original of the CNT beam and high crystal integrity. Although the performance of the electrode is improved, it is not as good as the self-healing function of the electrode. Kim et al. [144] prepared rGO (spray-rGO) film by using the supersonic dynamic spray method. The tension applied to the electrode increased with displacement, and adhesion prevented separation. Compared with slurry rGO, the adhesion energy of spray rGO was increased by two times due to its large contact area. CE was nearly 100% with the increase of the number of cycles at 0.1 C. This is due to the high-energy impact generated during the injection process, and the rGO is transformed into a hexagonal structure under the in-plane tensile action, resulting in a self-healing effect. Both CNT and G-based films show superior electrochemical and mechanical properties, but their capacity is low and requires the addition of high-capacity particles.

Si can provide a high capacity for the electrode, but the volume changes greatly. Therefore, researchers add CNT or G to the Si based electrode to relieve the stress caused by the volume change. Zhang et al. [102] prepared Si/rGO films respectively. rGO covered the surface of Si to form a layered structure (Fig.5(a)), which played a conductive and mechanical protective role, so that the film could be curled back and forth on the glass rod without any fracture (Fig.5(a)), and its capacity was about twice that of graphite at 200 mA/g. Even if the specific capacity of more than 150 cycles was more than 650 mAh/g, the capacity of the installed battery decreased slightly with the increase of the number of cycles in the flat state, but the increase of the number of cycles in the bending state almost did not decrease. Li et al. [145] prepared GF/Si film with a foam thickness of 130 nm by CVD, which could be bent (Fig.5(b)) with a reversible capacity of 1.4 mAh/cm² at 0.22 mA/cm², a weight capacity of up to 620 mAh/g, 7 times that of GF. Even if the current density increased to 0.45 mA/cm², the area capacity was up to 0.95 mAh/cm², and the corresponding weight capacity was up to 620 mAh/g. The surface folds were caused by the rough surface of the nickel foam template and the stress-induced dislocation during GF preparation, while the hollow structures and the dense layers coated on the GF surface served as mechanical supports and ion transport channels. Xie et al. [99] prepared Si/CNT films through blade coatings. The monodisperse CNTs network gave the film a T_s of up to 3.75 MPa and a fracture strain of 43.9%. The area capacity of the electrode was up to 10.6 mAh/cm² at 0.06 mA/cm², which fully met the actual requirements. Additionally, the reversible capacity reached 5.64 mAh/cm² after 200 cycles at 0.3 mA/cm². Although the electrode prepared by this technique has obtained an excellent performance, it is difficult to guarantee the uniformity of coating.

Compared to Si, SiO_x combines the excellent capacity properties of Si, but introduces oxygen-containing groups, making the capacity of the electrode slightly lower than that of the Si-based electrode, but with a relatively reduced volume effect. Guo et al. [146] prepared SiO_x/CNT film with core sheath structure (Fig.5(c)). Microscopically, some particles were observed to be dispersed between parallel strings. The structure of the film did not break when it was subjected to extreme deformation (repeated folding, rolling, and twisting). Reversible capacities were up to 1240 and 441 mAh/g at 100 mA/g and 2 A/g, respectively. Even after 200 and 500 cycles, the material did not undergo any deformation, and the SiO_x on the CNT became loose after 1000 cycles. However, Kang et al. [147] prepared 3D SiO_x-based film using PVDF and CNT as supports. The film was flexible and had an excellent crystal structure. There were pores in the film, and the impedance was minimum when the content of CNT was 30%. When the SiO_x load was increased to 1.25 mg/cm² and 100 cycles were performed at 500 mA/g, the capacity retention was 820 mAh/g, which was more than twice that of commercial graphite. The capacity was up to 1245 mAh/g at 1200 mA/g. This high capacity and flexibility are mainly derived from CNT substrate, which forms a 3D frame with buffering performance with PVDF to prevent SiO_x from contacting the electrolyte directly.

In addition, Fu et al. [148] prepared Si/CNT paper by electrodeposition at room temperature, which could be bent or twisted multiple times without any damage and had a volume capacity of up to 1400 mAh/g at 200 mA/g. This is due to the porous network structure constructed by CNTs, which can illuminate LEDs even when bent. Compared with other methods of CVD or etching, electrodeposition is easy to operate and less expensive. On the basis of Si, other particles were added to change the film structure to improve the performance of the electrode. Guo et al. [149] developed a camellia structure Ni_xCo_y-Silicate@CNT (NC/SC) hybrid film through CVD, which could withstand various severe deformation (curling, twisting, and folding) without any change in structure (Fig.5(d)). The T_s of the film was 2.9 times of that of CNT and could withstand a pressure of 4.08 MPa. The reversible capacity of the film is as high as 1047 mAh/g at 0.1 A/g, and the capacity retention rate of 140 cycles is as high as 78.13%. The battery can still light the LED when it is bent at 0°, 90° and 180° bending or even “fold in half.” Compared with other silicate electrode materials, the specific area capacity of NC/SC films was much higher. This is due to the high load density and camellia structure of bridging CNTs, which achieves mechanical stability but questions the higher tensile requirements. However, Zhang et al. [103] prepared CNT/(Fe@Si@SiO₂) film with a core plasma skin nanostructure, which could be wound around a glass rod, withstand a 500 g object, had a maximum tensile rate of 18%, and an ultimate strength (UTS) of 35 MPa. The mechanical properties were high enough for the manufacture of batteries. The reversible capacity of CNT/Fe@Si@SiO₂ at 5 A/g was 2.6 times that of CNT/Si, and the capacity retention rate of 500 cycles at 1 A/g was 4.6 times that of CNT/Si. The electrode structure could still be maintained after 1000 and 5000 times of bending, due to the core-pulp-skin nanostructure and the mechanical support of CNT.

The addition of other particles, such as Zn, Co, Fe oxides, CNT, and G, was used in different ways to prepare films. The G/porous ZnCo₂O₄-60% (G-PZCO, 60% representing mass fraction) film prepared by Cao et al. [150] after annealing could be repeatedly bent without cracking. The capacity loss was only 2.7% at 1 A/g for 1000 cycles, even at 8 A/g. A capacity of 322 mAh/g was also available, and the G-ZCO capacity was 20% lower than the G-PZCO film electrode. This is due to the random connection of ZnCo₂O₄ nanocrystals on the G plane to form sheet morphology and a large number of nanopores. Han et al. [151] proposed the cellulosic acetate-CNT (CA-CNT) film, where CNTs served as the substrate to prevent the aggregation of LCO/LTO particles, making the electrolyte more easily impregnated into the network, and enabling the battery to maintain an excellent electrochemical performance and a high flexibility without internal short circuit or severe electrode breakdown even after extreme deformation or repeated bending. Additionally, an area capacity of up to 5.4 mAh/cm² was obtained by stacking six self-supporting film electrodes. Ren et al. [93] used solvent evaporation to produce the TiO₂/G/PVDF film with a stacked structure, which could be rolled up and bent, and the capacity reached 165 mAh/g after 40 cycles at 60 mA/g. The RCT increased slightly with the increase of the cycle. This stems from PVDF as the packaging material and G as the flexible substrate. It is not difficult to see that the electrode capacity remains low at about 200 mAh/g. Therefore, it needs further exploration to realize high flexibility on the basis of maintaining a high stable capacity.

The conductive substrate was assembled into a thin film material similar to paper [47]. SnO₂ has a specific capacity of up to 782 mAh/g, but it also faces nanoparticle expansion. Therefore, researchers added G or CNT to limit its expansion. Shang et al. [94] prepared the flexible graphene/SnO₂ paper (FGSP) with the help of freeze-drying. The layered FGSP curved to 180°, 360°, and 720°, which not only increase the spacing of the G-sheets, but also mitigate the aggregation of SnO₂ NPs. The FGSP had a capacity of 406 mAh/g at 2 A/g. This stems from the lamellar structure of G. Noerochim et al. [152] prepared the SWCNT/SnO₂ paper by using the polyol method. The electrode could still work normally even if it was bent to 180°. SnO₂ NPs were mainly distributed on the surface of the SWCNT beam with a grain size less than 5 nm. The discharge capacity of more than 100 cycles was 2.4 times that of SWCNTs, which was due to the fact that SWCNTs acted as a flexible mechanical support and high conductivity matrix, forming a 3D nanopore network. Wang et al. [98] synthesized a polycrystalline SnO₂-G nanocomposite (SGN, Fig.5(f)) paper by using the hydrothermal method, which could be bent. The discharge capacity of the SGN paper was 1195 mAh/g at 0.1 C for the 10th cycle, and 91.5% of the original capacity after 80 cycles. Even at 1 C, the power capacity was as high as 540 mAh/g because of the synergy between GP and SnO₂, where SnO₂ prevented the agglomeration of G. The flexibility and structure of G facilitates the anchoring of SnO₂, providing a buffer space and conductive channels for SnO₂ to effectively prevent aggregation, cracking, and fragmentation.

In addition, other NPs such as Bi₂Se₃, pMn₃O₄, and MoS₂ can also provide capacity for the battery. Chen et al. [153] prepared the Bi₂Se₃/G (BSG) composite paper using vacuum filtration and thermal reduction. The BSG presented a layered structure and was flexible (Fig.5(h)), with a capacity of 203 mAh/g for 100 cycles at 50 mA/g. This is due to the synergistic effect of G acting as a conductive material and buffer layer while preventing the volume expansion of Bi₂Se₃. Park et al. [154] used vacuum filtration and heat treatment to prepare a lamination pMn₃O₄ nanorod (NR)/rGO paper, which could be bent. It had a specific capacity of about 2.9 times that of rGO after 100 cycles at 100 mA/g, and a specific capacity of 618 and 196 mAh/g at 100 and 2000 mA/g, far superior to the rGO. This is due to the hydrogen bond network between the carboxyl, hydroxyl, epoxy, carbonyl, and other oxygen-containing functional groups on the surface of the conductive rGO and water, so as to present a 3D stacked porous structure and rGO as a mechanical support. Liu et al. [155] prepared an amorphous MoS₂ nanosheet/N-doped carbon microtubule/rGO composite paper (NCMTs@A-MoS₂/rGO) by vacuum filtration, which was foldable and had a high surface hydrophobicity, with a conductivity of 700 times that of A-MoS₂, and an aperture of 3 times that of [e-mail] capacity was maintained at 544 mAh/g for 1000 cycles at 1 A/g, and the ensured diffusion coefficient was higher than that of NCMTs@A-MoS₂. This is due to the hollow space of the NCMT and the fact that the tubular structure of the open channel can be used as the skeleton connection A-MoS₂ and its rGO introduction.

At the same time, SWCNT and polymer form excellent electrodes. Wang et al. [96] prepared a microporous single-walled carbon nanotube/polycellulose papers (SWCNT/PPs) (Fig.5(e)). The SWCNT coated on cellulose fibers fully absorbed electrolytes and acted as an electrolyte reservoir. The electrodes were also flexible and had a capacity of 102.6 mAh/g at 10 C. The SWCNT/PP electrodes had a higher capacity and a lower electrochemical interfacial resistance than metal current collectors. This results from the conductive interconnection network formed by the interaction of larger van der Waals forces and hydrogen bonds between SWCNT and polymer. Yu et al. [101] studied Young’s modulus of Ga discharge at different cutoff potentials based on the self-healing ability of Ga to repair cracks, and understood the mechanical property attenuation of Ga in the process of lithium. A 3D flexible independent CNF/CNT paper @EGaIn NPs had a network structure and was flexible (Fig.5(g)). The reversible capacity remained at 420 mAh/g after 100 cycles at 800 mA/g. This is due to the mesh structure effectively buffering the volume expansion of EGa-In and the CNT layer preventing the EGa-In from falling off the conductive substrate. It can be seen from the above that the thin-film of the electrode and the improvement of its internal structure can prevent the electrode from falling off the substrate after realizing self-support, and achieve a high flexibility on the basis of maintaining high stable capacity (Tab.2).

In summary, powder is not a self-supporting material and needs to be coated with fluorine on the substrate, which easily falls off the substrate after large deformation and has a poor stability. In theory, wire and foil electrodes are superior to the non-self-supporting foil support, the protection is better than wire, and the linear electrodes have good mechanical properties compared to foil and can be bent, folded and braided.

4 Assembly structure and function of CNT and G-based FLIBs

Conventional LIBs are mainly composed of hard batteries. However, the batterie will be slightly deformed during use and may be folded when carrying heavy loads. Therefore, the assembled battery is required to be bent and folded. The battery assembled in Fig.6 can be bent, stretched, and folded so that it can still power electrical appliances after multiple bends. When batteries are damaged, they need to be repaired in time, which consumes huge manpower and material resources as well as financial resources. Based on the power supply, the battery needs to realize health self-detection and can monitor the health status at any time. This indicates that batteries are expected to be micro, soft, light, and multi-functional.

4.1 Bendability

With the continuous development of technology, conventional LIB has been continuously improved to accommodate bendability and high security properties. Zhong et al. [112] prepared an MnO@C-rGO anode by layer assembly of dopamine (PDA) coated MnO₂ nanowires (MnO₂@PDA NWS) and GO NS, and formed sandwich FLIB with the cathode electrode of LFP/Al film through diaphragm (Fig.7(a)). The discharge curves of the battery under different deformation states almost overlap, with a capacity as high as 2.6 mAh/cm² at 0.5 mA/cm². Even after bending 100 times, the battery still showed a retention of 90% and easily lit up the LED in the bent state and without dimming after 30 min. Li et al. [106] stacked LFP/GF cathode and LTO/GF anode electrodes with a thickness of 100 μm on both sides of the film, and sealed the FLIB assembled with a 250 μm thick PDMS. The thickness of the battery is less than 800 μm, and the capacity is only 4% lost at 100 cycles at 10 C. No change in the battery structure was observed even when the battery was repeatedly bent to a radius of less than 5 mm. The volumetric energy density of batteries at low discharge rates is not very high and becomes better at high discharge rates compared to batteries without GF. The high capacity and flexibility of the battery is derived from the GF flexible substrate and the constructed conducting network, and the small thickness of the GF allows the battery to operate at a high power and be fully charged in a very short time. The battery assembled by Shi et al. [108] used G-LTO and G-LFP electrodes, LiPF₆ (1 mol/L) in ethylene carbonate and diethyl carbonate as the electrolyte, and sealed polypropylene diaphragm with PDMS into a flat FLIB, which had a capacity up to 112.1 mAh/g even at 10 C. The battery powers a red light-emitting diode, which could be easily bent without structural or performance failures due to reversible deformation of the integrated electrode. With the same amount of active material added, the energy density of G-LFP was as high as 325 Wh/kg, which was 1.76 times that of Al-LFP. It can be seen that the battery assembled from the G base electrode performs well. Wang et al. [98] used SGN composite paper and LCO on Al sheet as electrodes to form a planar FLIB battery pack (Fig.7(b)). The SGN paper was light in weight, the surface density was as low as 0.6 mg/cm², and the FLIB unbent and bent 30° capacity change little in the first 10 cycles at 5 C, which could supply normal power to the LED. The battery can obtain superior performance mainly by the synergistic effect between the flexible substrate and the various components.

Additionally, Hu et al. [105] integrated all the components into a paper to make a flat FLIB with a thickness of 300 μm, where the paper was both a diaphragm and a mechanical support. After bending the battery to 6 mm, its morphology and conductivity remained unchanged, with an energy density of up to 108 MWh/g. An SWCNT/PP-based FLIB was developed by Wang et al. [96], which immersed PP in the CNTs ink for 10 min. Flexible SWCNT/PP could be customized in various shapes. A complete battery using SWCNT/PPs based on LTO and LFP electrodes showed a discharge capacity of 153.3 mAh/g at 0.1 C and 102.6 mAh/g at 10 C. The van der Waals force and hydrogen bond the SWCNTs tightly to the PP, forming a conducting network through large pores. The 50 μm CNTs microfiber paper concentrator prepared by Aliahmad et al. [104] with a mass load of 10.1 μg/cm², loaded CNTs microfiber paper coated with LTO and LCO into a flat FLIB. Even after being bent (300°) 20 times, the crack was shallow and closed, and the maximum charging capacity reached 126 mAh/g at 0.2 C. The superior performance of the battery stems from the porous structure of CNTs, but at the same time, the utilization rate of CNTs is significantly reduced.

In view of the superior properties of the fiber, the battery in combination with the fiber inherits not only the flexibility, breathability, and ductility of the fiber, but also the electronic capabilities. Yarn linear electrochemical energy storage devices are candidates for wearable and scalable storage device. Based on the composite of yarn and carbon-based materials, and due to its stable electrochemical performance under mechanical deformation, and its advantages of soft, light and small, researchers explored electrodes by twisting two fibers, but its poor performance including low energy density hindered its development. Therefore, yarn-shaped LIBs (YLIBs) with higher energy density are considered to be its replacement. In the case of stretching and deformation, the safety issue is even more prominent. Song et al. [109] formed a coaxial fiber-type LIB by wrapping a cotton core yarn with a CNTs film and a nano-sheath diaphragm. The CNTs film was a collector, and the cotton yarn was used as an electrolyte reservoir and a fiber-like skeleton, absorbing up to 9 times its own weight of electrolyte and expanding to 20 percent of its diameter after absorption, fully improving the interface contact between the battery components. In addition, the nanoweb was chosen as a diaphragm to avoid buckling by adjusting the in-plane deformation, resulting in a volumetric energy density of up to 144.82 mW/cm³. It worked well even under severe mechanical deformations such as bending and knotting, which relied on cotton yarn, and when the fiber optic battery was knotted, the amount of electrolyte and the interfacial force between the battery components increased, but the distance between the electrodes and the impedance decreased. The conductivity of the electrode was improved. Electrical conductivity is a prerequisite for optical fibers to be used as electronic fibers. Therefore, it is necessary to make conductive fibers or to give conventional fibers electrical conductivity.

Although lithium has advantages in terms of packaging and energy density, it uses standard liquid carbonic acid electrolyte, which can produce gas, battery expansion, and fire during battery aging and direct or non-penetrating damage to the battery. Therefore, gel electrolytes have been used in recent years, reducing the need for rigid or bulky packaging [107,156]. Zhang et al. [111] prepared continuous composite fibers by impregnation and online modification of direct spun CNTs fibers with LFP and LTP. The fibers were immersed in gel electrolyte mixed with sodium carboxymethyl cellulose (CMC) and Li₂SO₄, and assembled into a fiber water-based FLIB in parallel. The specific capacity could be as high as 23.7 mAh/g even at 1 A/g. With a high energy density of 30.12 Wh/kg, an excellent flexibility and cycling performance, even when bending to 90°, the initial capacity of 55% could still be maintained at 0.5 A/g for 100 cycles, and the structure of the electrode did not change significantly after cycling. This is due to the fact that there are holes and excellent flexibility on the surface of CNTs fiber, and the addition of CNTs fiber forms a conductive network, which contributes to the load of active substances. The synergistic effect with gel electrolyte improves the ratio performance of the battery, so that they can withstand bending, twisting and braiding, showing a huge application potential of composite fiber in flexible energy storage devices.

The PVDF-based electrolyte has a good flexibility and workability, and has a great potential to encapsulate high energy density flexible batteries, which can weaken the interaction between ion pairs and facilitate the dissociation of lithium. Weng et al. [110] took cotton fiber as the base material, covered the base material with a layer of shrink tube as the protective layer, wound CNT-LMO and CNT-Si/CNT composite yarn on the cotton fiber successively, and separated the gel electrolyte formed by mixing LiClO₄ solution with polyvinylidene fluoride-co-hexafluoro-propylene (PVDFHFP) solution. Electrodes and gel electrolytes were assembled into coaxial fiber LIBs. The CNT winding on the outside allowed the composite yarn to have a higher T_s than the cotton fiber, reaching a specific capacity of 106.5 mAh/g for the first cycle of discharge, which was maintained at 87% after 100 cycles. The battery achieved a linear energy density of up to 0.75 MWh/cm² and a weight energy density of 27.7 MWh/g. Although the linear capacity density was less than that of a cable-type LIB with copper wire as the skeleton, the linear capacity density was increased by reducing the bar spacing or increasing the diameter of the composite yarn. The LED could be easily lit and woven into an energy storage textile with a surface energy density of 4.5 MWh/cm² (Fig.7(c)). Although electronic fibers can be directly woven into fabrics, giving full play to the advantages of electronic fibers, the mechanical properties of many electronic fibers cannot meet the requirements of automatic integration process, easy to damage in the weaving process, the preparation method is still limited, and the processing efficiency is low, which limits the use of electronic fibers.

At present, the FLIBs prepared by most researchers can achieve bendability, but there are still technical challenges in folding, stretching, self-healing and other aspects, which need further research.

4.2 Stretchability

The development of flexible/stretchable power supply has become one of the key technologies to achieve a fully flexible integrated system [157], but layered planar devices make LIBs too bulky to meet the requirements of a fully flexible system. Therefore, researchers conducted research on device structure. Based on solid polymer electrolytes composed of polyethylene oxide (PEO) and GO, Kammoun et al. [158] reported a stretchable spiral film (< 1 mm) LIBs (Fig.7(d)), which could have a large plane deformation of 1300% and withstand 9000 tensile cycles. At about 650% out-of-plane deformation, its energy density was as high as 4.862 mAh/cm³. Under different tensile configurations, the average capacity of the battery was more than 0.1 mAh/cm². This helical structure of the battery enhances the safety. Gu et al. [113] grew LMO nanocrystals in-situ in 3D CNTs film, and the assembled battery had a stable performance after 150 cycles with a capacity loss of only 0.04% per cycle, and there was no significant increase in resistance after 1000 stretch-release cycles. Compared with the stretchable LIBs with snake interconnect and the “wave” electrode without chemical bonding, Al-based battery had a mechanical stability and high reversibility.

The introduction of gel electrolytes has greatly increased the safety, mechanical and electrochemical properties of the battery. Ren et al. [115] used MWCNT/LTO and MWCNT/LMO composite yarns as electrodes, coated them with gel electrolyte mixed with ethylene oxide, CH₂Cl₁₂, butyronitrile and disulfonimide to make stretchable YLIBs, and sealed the YLIBs in a heat-shrink tube, making the YLIBs withstand thousands of repeated deformations. When the bending cycle was 1000 times, the capacity of the YLIBs remained at 97%, and could be woven into various flexible structures. When it was made into a spring structure, it could be easily deformed without causing physical damage. After 200 times of 100% stretching, the resistance fluctuation of the two composite yarns was within 1%, and the capacity remained at 84%. The energy density was 27 Wh/kg and the power density was 880 W/kg which was an order of magnitude greater than that of lithium thin film batteries. This is because the MWCNTs forms a network structure, which reduces the mass of the composite yarn. In addition, the spring structure gives the battery a greater extension, making it safe for mass production. Zhang et al. [138] spun CNTF (d = 12 μm) by CVD and mixed it with LTO and LMO to form CNT/LTO and CNT/LMO hybrid fibers. Ten CNT/LTO and CNT/LMO fibers were wound into spring-like hybrid fibers as electrodes and impregnated with gel electrolyte formed by PEO, butyronitrile, lithium trifluoromethane sulfonimide (LiTFSi), methylene chloride and acetone to prepare fibrous LIBs. The reversible capacity of the battery was 2.2 mAh/m at 0.1mA/cm, and the capacity remained at 92.1% after 100 cycles. At the same time, the battery was stretchable, and the capacity was maintained at 85% under 100% strain, with a capacity change of less than 1% after repeated stretching of 300 times under a strain of 50%. Compared with stretchy fiber LIBs on an elastic polymer substrate, the spring-like fiber electrode reduced the volume and weight of the battery by approximately 400% and 300%, respectively, and increased the linear specific capacity by 600%. In addition, such fibrous electrodes were also suitable for supercapacitors and had excellent electrochemical properties against bending and tensile deformation. This is due to the fact that the CNTs are arranged as conductive scaffolds and current collectors, ensuring efficient charge transfer. Jung et al. [114] prepared two CNT yarns (CNTYs) by using the direct spinning method. One CNTY was coated with LFP, and the other was heat treated to obtain Fe₂O₃ doped CNTY (A-CNTY) as electrodes. Then 10% of PVDF (mass fraction is 10%) hexafluoro propylene (PVDF-HEP) was dissolved in tetrahydrofuran (THF) and mixed with LiPF₆ in the ethylene carbonate(EC) or diethyl carbonate (DEC) solution to obtain polymer gel electrolyte, and coated it on the electrode to make yarn FLIBs. At 800 mA/g, the A-CNTY had a capacity of up to 634.2 mAh/g and remained to be 423.4 mAh/g after 100 cycles, with a capacity loss of about 4 mAh/g per cycle. The A-CNTY can be wound, twisted, rolled, and even knotted without fault. It has a high specific strength of 132 MPa/(g·cm^–3) and a specific stiffness of 28.8 GPa/(g·cm^–3). The conductivity was as high as 1.59 × 10⁵ S/m, and the assembled battery could be stretched to 30 cm without using any substrate to withstand deformation. Even though the LED could be powered normally under bending and binding, the capacity loss was negligible after 1000 cycles of bending. This is due to the close contact between Fe₂O₃ and CNT to promote charge transfer. In addition, CNTY provides a pathway for electron conduction as a mechanical support.

Although the battery can be stretched repeatedly, this results in resource consumption. The rapidly growing battery market requires high energy density and waste management solutions to deal with battery waste worldwide. However, the long-term stretching cycle of the LIB reduces the capacity and mechanical performance of the battery, which requires manual maintenance but wastes a large amount of human and material resources. Therefore, the function of the battery poses higher requirements.

4.3 Foldability

After meeting the requirement of bendability and stretchability, occasionally a large folding of the battery is needed. However, folding causes the structure of the battery to deform. Therefore, it is urgent to develop a foldable FLIB on the basis of ensuring safety. The HOCNM prepared by Liu et al. [116] using HOCNM as the substrate, coated with LCO and LTO, remained intact after drying. The mass load of LCO and LTO in the HOCNM was greater than 10 mg/cm². The active material on the HOCNM coating had a flat form and a good durability. In addition, LCO and LTO were infiltrated into the HOCNM, increasing the superior wettability of the HOCNM. A foldable battery was assembled with LiPF₆ in a mixture of EC or DEC as electrolyte and a polyethylene diaphragm on Cu and Al foils based on the slurry technology. The LED continuously emitted light during the folding and release of the battery. The battery powered a smart bracelet worn on the wrist, and the glow of the lamp did not change. Replacing the battery placed in the core of the wristband with a uniform shape battery could improve the battery life of the wristband. This is due to the high electrical conductivity, porous array, and electrolyte wettability of the HOCNM prepared by surface/interface engineering. Hu et al. [159] used CNT macro film (CMF) as the collector fluid, LCO and Li₄Ti₅O₁₂ as the cathode and anode electrodes, polypropylene diaphragm and LiPF₆ (1 mol/L) mixed in EC or DEC as the electrolyte to prepare a collapsible LIB. The energy density of the battery at 1 C was 170 Wh/kg, which was twice the energy density of the traditional lithium. The battery could power the light-emitting diode, and could folded five times, even though there was no significant change in the luminescence of the released rear lamp. The initial discharge capacity of the battery at 0.2 C was 164.3 mAh/g. Whether the battery was flat or folded, the CE was close to 100%. This stems from the fact that the activity permeates the pores of the CMF to form a strong interface that resists layering even when the cell is folded repeatedly. Mu et al. [118] report a surface/interface modification strategy to obtain an electrolytic CMF flexible collector, CMF/LCO and coated CMF(CCMF)/Gra as a cathode and anode electrode, polypropylene diaphragm, mixing LiPF₆ (1 mol/L) with EC or DEC to prepare electrolytes, assembled into foldable LIBs. The battery had a high open circuit voltage of 4.04 V in the flat or folded state, and maintained this high voltage through 500-fold/release cycles, with a weight/volume energy density of 252 Wh/kg and 332.05 Wh/L, respectively. This is due to the interface electrolytic CCMF collector, which prevents ineffective insertion of Li⁺ in CNTs. This work can significantly improve the voltage and energy density of FLIBs and has important implications for the future development of FLIBs.

Liquid electrolytes are prone to leakage and safety risks. Consequently, the vast majority of researchers prefer gel solid electrolytes. Wei et al. [117] designed a planar collapsible FLIB using LCO and LTO as electrodes, CNTs film as flexible collector, and porous gel polymer electrolyte (P-GPE) and diaphragm composed of PTC polymer, LLZTO, LiTFSi, and liquid electrolyte. The battery was bent 90 times and 10000 times in situ at 1 C with a bending radius of 1 cm at 3 and 200 mm/s, respectively, but the charge and discharge curve hardly unchanged. The battery could provide a capacity of 138.3 mAh/g, even after 150000 bends, with only a 7% of the specific capacity lost. Unlike traditional coin, cylinder, or square batteries, the battery was still able to power LED after folding and punching several times (Fig.7(e)). This may result from the loss of current in the process of folding and punching, and the fact that the porous structure of CNTs film and the PTC-based GPEs prepared by the combination of LLZTO and LiTFSi effectively prevent the propagation of lithium dendrites.

From the above, it can be seen that the mechanical properties of the foldable LIB are relatively superior, but special attention needs to be paid to the short circuit and open circuit of the battery during the folding process to improve the safety factor of the battery.

4.4 Self-healing ability

Although the performance of the above batteries is superior, for the sustainable consideration of resources, researchers extend the service life of energy storage devices by improving their ability to withstand the external stress [156,158,160]. However, it is unrealistic to realize economic sustainability through artificial repair. Therefore, it should be realized from self-healing after suffering external damage. The self-healing water-type LIB manufactured by self-healing polymers [161–164] that rebuild fractures based on reversible chemical bonds or specific interactions, has become the key to the next generation of wearable electronics due to their suitable electrodes and special battery structures to achieve efficient self-healing. Zhao et al. [165] designed an array of CNTs on a self-healing polymer substrate, a sandwich structure containing LMO and LTP as electrodes, and an aqueous solution of lithium sulfate (Li₂SO₄)/sodium CMC as a gel electrolyte and diaphragm to create a planar all-solid-state flexible water-based LIB. Its energy density was as high as 32.04 Wh/kg, which was five times higher than the highest reported value. The battery was cut into two separate parts that, over a few seconds, could heal back to the normal function (Fig.8(a)). Prior to the cut, the specific capacity was 28.2 mAh/g at 0.5 A/g, which was reduced to 17.2 mAh/g after the fifth healing. After 100 charge−discharge cycles, the capacity retention rates for the prepared FLIB and the fifth healing were 82.8% and 69.3%, respectively. Even if the specific capacity was maintained at 90.3% after 200 cycles of bending at 60°, this self-healing water-based FLIB could quickly and efficiently solve the fracture problem in use, because the polymer substrate was rich in multiple hydrogen bonds, which could be rebuilt after the hydrogen bond was broken at room temperature. The oriented CNT sheets were reconnected by the van der Waals force to restore the conductivity. The non-toxic gel electrolyte avoided the safety hazards caused by electrolyte leakage, and such self-healing water-based FLIB was expected to be used in future wearable devices. Furthermore, polymer-made battery packaging layers and chemical bonds will greatly increase battery reliability, flexibility and safety. Rao et al. [166] used spring-like LCO@rGO (ASLG) fiber as the cathode and SnO₂ quantum dot-@rGO (ASG) fiber as the anode. PVDF-HEP was immersed in LiClO₄ and EC or DEC mixture as gel electrolyte, and self-healing carboxylated polyurethane (PU) as encapsulation layer to design a self-healing all-fiber quasi-solid-state LIB. The battery structure greatly improved the permeability of the electrolyte to the active material, giving the battery a capacity of 82.6 mAh/g after bending and knotting 50 cycles at 0.1 A/g. The capacity of the battery after the fifth repair at 0.1A/g was 50.1 mAh/g. Additionally, even if the battery was bent, twisted, and knotted after charging, the LED brightness did not change significantly. Even if the battery was completely cut, the LED lit up again after the two parts came into contact for some time. The superior performance of the electrode is due to the stacking and combination of rGO, which makes the electrode have plasticity and flexibility. This self-healing property is due to the presence of the hydrogen bonds in the supramolecular network at the fracture surface, and the stress generated by PU recovery pulls the disconnected rGO fiber back to the interconnected state. Therefore, self-healing quasi-solid-state LIBs can quickly solve the fracture problem in practical applications, and its successful preparation may provide a method for designing the next generation of self-healing all-fiber quasi-solid state power supplies for wearable and portable devices.

The successful preparation of the self-healing FLIBs avoids the consumption of human, material, and financial resources caused by battery damage. Self-healing FLIBs are a promising and challenging research direction.

4.5 Self-detecting ability

In order to solve the energy and environmental dilemma to increase the smart effect of batteries, Kuznetsov et al. [167] improved the energy density, specific capacity, and resource recovery of the battery on the basis of stretchability to achieve sustainability, mainly by in situ mixed growth of SWCNT and atomization of Li-Ni_0.5Mn_0.3Co_0.2O₂ (LiNMC) and graphite particles to prepare electrodes, which could withstand a bending radius of 2 mm, a torsion angle 180°, and a tensile strain of 12%. The battery could withstand mechanical interference, providing an energy density of up to 40% and a specific capacity of 70%. Two anode electrodes could share a basic block created by a cathode electrode, when the battery powered the device. The mechanical perturbation was perceived as a sharp change in potential difference when the battery was subjected to a bend/twist cycle. The battery created in the shape of a flexible wristband (folded battery) powered the smartwatch to detect heart rate, and could be integrated into the functional fabric to power LED (Fig.8(b)). This is mainly due to the piezoresistive effect of the SWCNT conductive stent to achieve in-situ/operational self-detection of battery health by detecting structural changes in the electrode. The solution-free battery manufacturing technology and architecture eliminate the components that are harmful to the environment, mainly through the combination of CNT and LiNMC and graphite in the electrode through the van der Waals force, which can be recycled through mechanical separation. This not only reduces the consumption of natural resources, but also promotes the environmentally friendly circular economy of batteries.

The twistable and self-detecting FLIB provides a new idea for the development of the future energy storage technology. It can not only achieve large deformation capability, but also realize the intelligentization of the battery, improve battery safety and reliability, and avoid sudden accidents. However, how to intelligentize FLIB is still a technical bottleneck in this field.

The above studies show that FLIBs perform different mechanical, self-healing and self-detecting functions, which rely on the hydrogen bonds between materials and the reconstruction of hydrogen bonds after major breaks to achieve self-healing and good contact between parts. LIBs must be flexible, miniaturized, lightweight, and multi-functional to power wearable electronics. The related performance of batteries with different structures and functions is summarized in Tab.3.

5 Conclusions and outlook

This paper reviewed the role of CNT and G in the fabrication of FLIB electrodes, particularly as conductors, scaffolds, and active materials to enhance the electrode electrochemical performance and flexibility of the electrode. In addition, it discussed the effects of CNTs and G on battery performance at low and high temperatures. Moreover, it highlighted the effects of electrode materials with different structures on electrochemical and mechanical properties. Furthermore, it emphasized the effects of electrode materials with different structures on electrochemical and mechanical properties. Electrode-based nanostructures can be divided into powder, wire-shaped, and thin film structures, which has advantages and disadvantages, among which wire-shaped and film structures often have excellent mechanical properties. Based on the above discussion, CNTs and G are clearly proven to be promising materials for the construction of FLIBs.

Wearable electronic devices are required not only to have bendability, but also to have extremely high performances in folding, twisting, stretching, and compressing. Foldability is an extremely severe deformation and is still a bottleneck limiting the development of FLIBs. Moreover, FLIBs can lead to short circuits under severe deformation conditions. Therefore, the combination of safety and super-flexibility should be an important research direction of FLIBs in the future. The all-solid flame-retardant electrolyte has better safety than the liquid electrolyte and therefore can be an important research direction in the future. Linear FLIBs are easier to be woven into wearable electronics than FLIBs of other shapes. Wearable electronics pay attention to beauty and comfort. Therefore, how to achieve excellent electrochemical performance, super flexibility, beauty and comfort of linear FLIBs through structural design also poses a technical challenge. The safety of lithium-ion batteries is a concern. Therefore, the development of self-detecting and self-healing intelligent FLIBs will be an important research direction in the future. At present, the research of FLIBs is still in the laboratory stage, mainly focusing on the structural design of electrodes and batteries, but the cost of making batteries is relatively high. Cost reduction can be achieved by manufacturing materials with small size, light weight, long cycle life, good mechanical stability and resistance to environmental erosion. Additionally, the development of short cycle and simple process technology is also an effective way to achieve cost reduction. Moreover, harsh environments such as ultra-high temperature, ultra-low temperature, corrosion and oxidation should be considered in the application of CNT and G-based FLIBs. Therefore, in a word, the future FLIBs and their electrode materials need to achieve miniaturization, multi-functional integration and good environmental applicability. The outlook for key research directions in the future can be described by a CNT and G-based FLIB tree, as shown in Fig.9. The red fruit represents the relatively mature research direction, while the cyan fruit represents the relatively immature research direction, that is, the direction that needs to be focused on in the future. The branches below the cyan fruit represent pathways or factors corresponding to this research direction.

References

[1] Zhu Y H, Yang X Y, Liu T. . Flexible 1D batteries: Recent progress and prospects. Advanced Materials, 2020, 32(5): 1901961

[2] Li Y, Liu Y, Sun J. . Progress of electrode designs for flexible energy storage devices. Materials Reports, 2020, 34(1A): 1177–1186 (in Chinese)

[3] Nathan A, Ahnood A, Cole M T. . Flexible electronics: The next ubiquitous platform. Proceedings of the IEEE, 2012, 100(Special Centennial Issue): 1486–1517

[4] Lochner C M, Khan Y, Pierre A. . All-organic optoelectronic sensor for pulse oximetry. Nature Communications, 2014, 5(1): 5745

[5] Misra V, Bozkurt A, Calhoun B. . Flexible technologies for self-powered wearable health and environmental sensing. Proceedings of the IEEE, 2015, 103(4): 665–681

[6] Jia W, Wang X, Imani S. . Wearable textile biofuel cells for powering electronics. Journal of Materials Chemistry. A, Materials for Energy and Sustainability, 2014, 2(43): 18184–18189

[7] Song Z, Wang X, Lv C. . Kirigami-based stretchable lithium-ion batteries. Scientific Reports, 2015, 5(1): 10988

[8] Stoppa M, Chiolerio A. Wearable electronics and smart textiles: A critical review. Sensors, 2014, 14(7): 11957–11992

[9] Wang X, Lu X, Liu B. . Flexible energy-storage devices: Design consideration and recent progress. Advanced Materials, 2014, 26(28): 4763–4782

[10] Wu Z, Wang Y, Liu X. . Carbon-nanomaterial-based flexible batteries for wearable electronics. Advanced Materials, 2019, 31(9): 1800716

[11] Miao J, Fan T. Flexible and stretchable transparent conductive graphene-based electrodes for emerging wearable electronics. Carbon, 2023, 202: 495–527

[12] Wang F, Zhao S, Jiang Q. . Advanced functional carbon nanotube fibers from preparation to application. Cell Reports. Physical Science, 2022, 3(8): 100989

[13] Choi K H, Ahn D B, Lee S Y. Current status and challenges in printed batteries: Toward form factor-free, monolithic integrated power sources. ACS Energy Letters, 2018, 3(1): 220–236

[14] Deng Q, Fu Y, Zhu C. . Niobium-based oxides toward advanced electrochemical energy storage: Recent advances and challenges. Small, 2019, 15(32): 1804884

[15] Tao T, Lu S, Chen Y. A review of advanced flexible lithium-ion batteries. Advanced Materials Technologies, 2018, 3(9): 1700375

[16] Tarascon J M, Armand M. Issues and challenges facing rechargeable lithium batteries. Nature, 2001, 414(6861): 359–367

[17] ChenYLiuSBiZ, . Reviewing electrochemical stability of ionic liquids-/deep eutectic solvents-based electrolytes in lithium-ion, lithium-metal and post-lithium-ion batteries for green and safe energy. Green Energy & Environment. 2023, early access, doi:10.1016/j.gee.2023.05.002

[18] Jetybayeva A, Aaron D S, Belharouak I. . Critical review on recently developed lithium and non-lithium anode-based solid-state lithium-ion batteries. Journal of Power Sources, 2023, 566: 232914

[19] Wang Y, Yang Q, Zhao Y. . Recent advances in electrode fabrication for flexible energy-storage devices. Advanced Materials Technologies, 2019, 4(7): 1900083

[20] Fan X, Liu B, Ding J. . Flexible and wearable power sources for next-generation wearable electronics. Batteries & Supercaps, 2020, 3(12): 1262–1274

[21] Fang Z, Wang J, Wu H. . Progress and challenges of flexible lithium-ion batteries. Journal of Power Sources, 2020, 454: 227932

[22] Fu K K, Cheng J, Li T. . Flexible batteries: From mechanics to devices. ACS Energy Letters, 2016, 1(5): 1065–1079

[23] Gong X, Yang Q, Zhi C. . Stretchable energy storage devices: From materials and structural design to device assembly. Advanced Energy Materials, 2021, 11(15): 2003308

[24] Jeong I, Han D Y, Hwang J. . Foldable batteries: From materials to devices. Nanoscale Advances, 2022, 4(6): 1494–1516

[25] Kim S D, Sarkar A, Ahn J H. Graphene-based nanomaterials for flexible and stretchable batteries. Small, 2021, 17(48): 2006262

[26] Liu W, Song M S, Kong B. . Flexible and stretchable energy storage: Recent advances and future perspectives. Advanced Materials, 2017, 29(1): 1603436

[27] Song W J, Yoo S, Song G. . Recent progress in stretchable batteries for wearable electronics. Batteries & Supercaps, 2019, 2(3): 181–199

[28] Yan C, Lee P S. Stretchable energy storage and conversion devices. Small, 2014, 10(17): 3443–3460

[29] Zhai Q, Xiang F, Cheng F. . Recent advances in flexible/stretchable batteries and integrated devices. Energy Storage Materials, 2020, 33: 116–138

[30] Zhou G, Li F, Cheng H M. Progress in flexible lithium batteries and future prospects. Energy & Environmental Science, 2014, 7(4): 1307–1338

[31] Lee S M, Kim J H, Ahn J H. Graphene as a flexible electronic material: Mechanical limitations by defect formation and efforts to overcome. Materials Today, 2015, 18(6): 336–344

[32] Zhang Y, Jiao Y, Liao M. . Carbon nanomaterials for flexible lithium-ion batteries. Carbon, 2017, 124: 79–88

[33] Islam J, Chowdhury F I, Raza W. . Toward polymer composites based and architectural engineering induced flexible electrodes for lithium-ion batteries. Renewable & Sustainable Energy Reviews, 2021, 148: 111302

[34] Marriam I, Tebyetekerwa M, Xu Z. . Techniques enabling inorganic materials into wearable fiber/yarn and flexible lithium-ion batteries. Energy Storage Materials, 2021, 43: 62–84

[35] Jayaraman T, Murthy A P, Elakkiya V. . Recent development on carbon based heterostructures for their applications in energy and environment: A review. Journal of Industrial and Engineering Chemistry, 2018, 64: 16–59

[36] Lee H, Yoo J K, Park J H. . A stretchable polymer-carbon nanotube composite electrode for flexible lithium-ion batteries: Porosity engineering by controlled phase separation. Advanced Energy Materials, 2012, 2(8): 976–982

[37] Wang J, Ma C, Tang J. . Facile fabrication of Fe₂O₃-decorated carbon matrixes with a multidimensional structure as anodes for lithium-ion batteries. Energy & Fuels, 2021, 35(1): 816–826

[38] Kumar S, Nehra M, Kedia D. . Carbon nanotubes: A potential material for energy conversion and storage. Progress in Energy and Combustion Science, 2018, 64: 219–253

[39] De Volder M F L, Tawfick S H, Baughman R H. . Carbon nanotubes: Present and future commercial applications. Science, 2013, 339(6119): 535–539

[40] Zhang Q, Huang J Q, Qian W Z. . The road for nanomaterials industry: A review of carbon nanotube production, post-treatment, and bulk applications for composites and energy storage. Small, 2013, 9(8): 1237–1265

[41] Devi R, Tapadia K, Kant T. . A low-cost paper-based flexible energy storage device using a conducting polymer nanocomposite. New Journal of Chemistry, 2020, 44(31): 13446–13457

[42] Endo M, Muramatsu H, Hayashi T. . ‘Buckypaper’ from coaxial nanotubes. Nature, 2005, 433(7025): 476

[43] Nguyen T H, Fraiwan A, Choi S. Paper-based batteries: A review. Biosensors & Bioelectronics, 2014, 54: 640–649

[44] Seo Y, Hwang B. Mulberry-paper-based composites for flexible electronics and energy storage devices. Cellulose, 2019, 26(16): 8867–8875

[45] Shen L L, Zhang G R, Etzold B J M. Paper-based microfluidics for electrochemical applications. ChemElectroChem, 2020, 7(1): 10–30

[46] Thakur A, Devi P. Paper-based flexible devices for energy harvesting, conversion and storage applications: A review. Nano Energy, 2022, 94: 106927

[47] Yao B, Zhang J, Kou T. . Paper-based electrodes for flexible energy storage devices. Advanced Science, 2017, 4(7): 1700107

[48] ZhouWZhangYCuiC, . Research progress on paper-based flexible electrode material. New Chemical Materials, 2020, 48(6): 6−8 (in Chinese)

[49] Chen S, Qiu L, Cheng H M. Carbon-based fibers for advanced electrochemical energy storage devices. Chemical Reviews, 2020, 120(5): 2811–2878

[50] Liu Y, Shen X, Wang X. . Fiber-supported alumina separator for achieving high rate of high-temperature lithium-ion batteries. Journal of Power Sources, 2020, 477: 228680

[51] Nagpure A S, Gogoi P, Lucas N. . Novel Ru nanoparticle catalysts for the catalytic transfer hydrogenation of biomass-derived furanic compounds. Sustainable Energy & Fuels, 2020, 4(7): 3654–3667

[52] Seman R N A R, Azam M A, Mohamad A A. Systematic gap analysis of carbon nanotube-based lithium-ion batteries and electrochemical capacitors. Renewable & Sustainable Energy Reviews, 2017, 75: 644–659

[53] Feng T, Chen W, Li W. . Microstructure properties, modification treatments and application prospects of graphene. Journal of Functional Materials, 2020, 51(4): 4028–-4039

[54] Zhu S, Sheng J, Chen Y. . Carbon nanotubes for flexible batteries: Recent progress and future perspective. National Science Review, 2021, 8(5): nwaa261

[55] Chen X, Ma Y. Wearable lithium-ion batteries based on carbon nanotubes and graphene. Advanced Materials Technologies, 2018, 3(10): 1800041

[56] Kumar R, Joanni E, Savu R. . Fabrication and electrochemical evaluation of micro-supercapacitors prepared by direct laser writing on free-standing graphite oxide paper. Energy, 2019, 179: 676–684

[57] Sharma V, Kagdada H L, Jha P K. Four-fold enhancement in the thermoelectric power factor of germanium selenide monolayer by adsorption of graphene quantum dot. Energy, 2020, 196: 117104

[58] Wen L, Li F, Cheng H M. Carbon nanotubes and graphene for flexible electrochemical energy storage: From materials to devices. Advanced Materials, 2016, 28(22): 4306–4337

[59] Zhao X, E J, Wu G. . A review of studies using graphenes in energy conversion, energy storage and heat transfer development. Energy Conversion and Management, 2019, 184: 581–599

[60] Li L, Zhang D, Deng J. . Carbon-based materials for fast charging lithium-ion batteries. Carbon, 2021, 183: 721–734

[61] He Y, Chen W, Gao C. . An overview of carbon materials for flexible electrochemical capacitors. Nanoscale, 2013, 5(19): 8799–8820

[62] Abdollahi A, Abnavi A, Ghasemi S. . Flexible free-standing vertically aligned carbon nanotube on activated reduced graphene oxide paper as a high performance lithium ion battery anode and supercapacitor. Electrochimica Acta, 2019, 320: 134598

[63] Cao S, Shi L, Miao M. . Solution-processed flexible paper-electrode for lithium-ion batteries based on MoS₂ nanosheets exfoliated with cellulose nanofibrils. Electrochimica Acta, 2019, 298: 22–30

[64] Cheng Y, Chen G, Wu H. . Use of regenerated cellulose to direct hetero-assembly of nanoparticles with carbon nanotubes for producing flexible battery anodes. Journal of Materials Chemistry. A, Materials for Energy and Sustainability, 2017, 5(27): 13944–13949

[65] Huang L, Guan Q, Cheng J. . Free-standing N-doped carbon nanofibers/carbon nanotubes hybrid film for flexible, robust half and full lithium-ion batteries. Chemical Engineering Journal, 2018, 334: 682–690

[66] Ren J, Ren R P, Lv Y K. A new anode for lithium-ion batteries based on single-walled carbon nanotubes and graphene: Improved performance through a binary network design. Chemistry, An Asian Journal, 2018, 13(9): 1223–1227

[67] Sun X, Liu Z, Li N. . Carbon nanotube paper as anode for flexible lithium-ion battery. Nano, 2016, 11(11): 1650120

[68] Xu C, Jing Y, He J. . Self-assembled interwoven CoS₂/CNTs/graphene architecture as anode for high-performance lithium-ion batteries. Journal of Alloys and Compounds, 2017, 708: 1178–1183

[69] Yi Z, Lin N, Zhao Y. . A flexible micro/nanostructured Si microsphere cross-linked by highly-elastic carbon nanotubes toward enhanced lithium-ion battery anodes. Energy Storage Materials, 2019, 17: 93–100

[70] Zeng T, Feng D, Peng Q. . Nano-GeTe embedded in a three-dimensional carbon sponge for flexible Li-ion and Na-ion battery anodes. ACS Applied Materials & Interfaces, 2021, 13(13): 15178–15189

[71] Fan P, Liu H, Liao L. . Flexible and high capacity lithium-ion battery anode based on a carbon nanotube/electrodeposited nickel sulfide paper-like composite. RSC Advances, 2017, 7(78): 49739–49744

[72] Ren J, Ren R P, Lv Y K. A flexible 3D graphene@CNT@MoS₂ hybrid foam anode for high-performance lithium-ion battery. Chemical Engineering Journal, 2018, 353: 419–424

[73] Wang M S, Wang Z Q, Chen Z. . One dimensional and coaxial polyaniline@tin dioxide@multi-wall carbon nanotube as advanced conductive additive free anode for lithium-ion battery. Chemical Engineering Journal, 2018, 334: 162–171

[74] Xiang T, Tao S, Xu W. . Stable 1T-MoSe₂ and carbon nanotube hybridized flexible film: Binder-free and high-performance Li-ion anode. ACS Nano, 2017, 11(6): 6483–6491

[75] Yang Y, Yang X, Chen S. . Rational design of hierarchical carbon/mesoporous silicon composite sponges as high-performance flexible energy storage electrodes. ACS Applied Materials & Interfaces, 2017, 9(27): 22819–22825

[76] Yildiz O, Dirican M, Fang X. . Hybrid carbon nanotube fabrics with sacrificial nanofibers for flexible high performance lithium-ion battery anodes. Journal of the Electrochemical Society, 2019, 166(4): A473–A479

[77] Zhao X, Wang G, Zhou Y. . Flexible free-standing ternary CoSnO₃/graphene/carbon nanotubes composite papers as anodes for enhanced performance of lithium-ion batteries. Energy, 2017, 118: 172–180

[78] Fu K, Yildiz O, Bhanushali H. . Aligned carbon nanotube-silicon sheets: A novel nano-architecture for flexible lithium-ion battery electrodes. Advanced Materials, 2013, 25(36): 5109–5114

[79] Lee S, Song H, Hwang J Y. . Directly-prelithiated carbon nanotube film for high-performance flexible lithium-ion battery electrodes. Fibers and Polymers, 2017, 18(12): 2334–2341

[80] Li Y, Wang P, Bao Y. . A flexible nanostructured paper of MnO NPs@MWCNTs/r-GO multilayer sandwich composite for high-performance lithium-ion batteries. Ceramics International, 2017, 43(10): 7588–7593

[81] Liang M, Wang W, Jiang Y. . Fabrication of C@Si@G for flexible lithium-ion batteries. Journal of Alloys and Compounds, 2021, 878: 160357

[82] Wang X, Sun L, Susantyoko R A. . A hierarchical 3D carbon nanostructure for high areal capacity and flexible lithium-ion batteries. Carbon, 2016, 98: 504–509

[83] Wu K, Xu G, Pan D. . Red phosphorus confined in MOF-derived N-doped carbon-based composite polyhedrons on carbon nanotubes for high-areal-capacity lithium storage. Chemical Engineering Journal, 2020, 385: 123456

[84] Yu X, Jiang Y, Yang X. . Dodecanethiol coated multi-walled carbon nanotube films as flexible current collector for lithium-ion batteries. Materials Letters, 2021, 291: 129508

[85] Zhang L, Huang Y, Zhang Y. . Three-dimensional nanoporous graphene-carbon nanotube hybrid frameworks for confinement of SnS₂ nanosheets: Flexible and binder-free papers with highly reversible lithium storage. ACS Applied Materials & Interfaces, 2015, 7(50): 27823–27830

[86] Cai H, Han K, Jiang H. . Self-standing silicon-carbon nanotube/graphene by a scalable in situ approach from low-cost Al-Si alloy powder for lithium-ion batteries. Journal of Physics and Chemistry of Solids, 2017, 109: 9–17

[87] Yu Y, Luo Y, Wu H. . Ultrastretchable carbon nanotube composite electrodes for flexible lithium-ion batteries. Nanoscale, 2018, 10(42): 19972–19978

[88] Abnavi A, Sadati Faramarzi M, Abdollahi A. . SnO₂@a-Si core-shell nanowires on free-standing CNT paper as a thin and flexible Li-ion battery anode with high areal capacity. Nanotechnology, 2017, 28(25): 255404

[89] Alaf M, Tocoglu U, Kayis F. . Sn/SnO₂/MWCNT composite anode and electrochemical impedance spectroscopy studies for Li-ion batteries. Fullerenes, Nanotubes, and Carbon Nanostructures, 2016, 24(10): 630–634

[90] Li Y, Ye D, Liu W. . A three-dimensional core-shell nanostructured composite of polypyrrole wrapped MnO₂/reduced graphene oxide/carbon nanotube for high performance lithium-ion batteries. Journal of Colloid and Interface Science, 2017, 493: 241–248

[91] Wang T, Ji X, Wu F. . Facile fabrication of a three-dimensional coral-like silicon nanostructure coated with a C/rGO double layer by using the magnesiothermic reduction of silica nanotubes for high-performance lithium-ion battery anodes. Journal of Alloys and Compounds, 2021, 863: 158569

[92] Zhang J, Zhang W, He T. . Smart reconstruction of dual-carbon decorated MnO for anode with high-capacity and ultralong-life lithium storage properties. Carbon, 2017, 115: 95–104

[93] Ren H M, Ding Y H, Chang F H. . Flexible free-standing TiO₂/graphene/PVdF films as anode materials for lithium-ion batteries. Applied Surface Science, 2012, 263: 54–57

[94] Shang Y, Liu X, Zhang J. . Freeze-drying-assisted fabrication of flexible graphene/SnO₂ for high-rate lithium-ion batteries. Ionics, 2021, 27(5): 1967–1976

[95] Wang H, He M, Zhang Y. Carbon nanotube films: Preparation and application in flexible electronics. Acta Physico-Chimica Sinica, 2019, 35(11): 1207–1223

[96] Wang J, Li L, Wong C L. . Flexible single-walled carbon nanotube/polycellulose papers for lithium-ion batteries. Nanotechnology, 2012, 23(49): 495401

[97] Wang J, Wang G, Wang H. Flexible free-standing Fe₂O₃/graphene/carbon nanotubes hybrid films as anode materials for high performance lithium-ion batteries. Electrochimica Acta, 2015, 182: 192–201

[98] Wang Q, Xing L, Xue X. SnO₂-graphene nanocomposite paper as both the anode and current collector of lithium-ion battery with high performance and flexibility. Materials Letters, 2017, 209: 155–158

[99] Xie C, Xu N, Shi P. . Flexible and robust silicon/carbon nanotube anodes exhibiting high areal capacities. Journal of Colloid and Interface Science, 2022, 625: 871–878

[100] Yoon S, Lee S, Kim S. . Carbon nanotube film anodes for flexible lithium-ion batteries. Journal of Power Sources, 2015, 279: 495–501

[101] Yu J, Xia J, Guan X. . Self-healing liquid metal confined in carbon nanofibers/carbon nanotubes paper as a free-standing anode for flexible lithium-ion batteries. Electrochimica Acta, 2022, 425: 140721

[102] Zhang H, Jing S, Hu Y. . A flexible freestanding Si/rGO hybrid film anode for stable Li-ion batteries. Journal of Power Sources, 2016, 307: 214–219

[103] Zhang M, Li L, Jian X. . Free-standing and flexible CNT/(Fe@Si@SiO₂) composite anodes with kernel-pulp-skin nanostructure for high-performance lithium-ion batteries. Journal of Alloys and Compounds, 2021, 878: 160396

[104] Aliahmad N, Agarwal M, Shrestha S. . Paper-based lithium-ion batteries using carbon nanotube-coated wood microfibers. IEEE Transactions on Nanotechnology, 2013, 12(3): 408–412

[105] Hu L, Wu H, La Mantia F. . Thin, flexible secondary Li-ion paper batteries. ACS Nano, 2010, 4(10): 5843–5848

[106] Li N, Chen Z, Ren W. . Flexible graphene-based lithium-ion batteries with ultrafast charge and discharge rates. Proceedings of the National Academy of Sciences of the United States of America, 2012, 109(43): 17360–17365

[107] Mezzomo L, Ferrara C, Brugnetti G. . Exploiting self-healing in lithium batteries: Strategies for next-generation energy storage devices. Advanced Energy Materials, 2020, 10(46): 2002815

[108] Shi Y, Wen L, Zhou G. . Graphene-based integrated electrodes for flexible lithium-ion batteries. 2D Materials, 2015, 2(2): 024004

[109] Song H, Jeon S Y, Jeong Y. Fabrication of a coaxial high performance fiber lithium-ion battery supported by a cotton yarn electrolyte reservoir. Carbon, 2019, 147: 441–450

[110] Weng W, Sun Q, Zhang Y. . Winding aligned carbon nanotube composite yarns into coaxial fiber full batteries with high performances. Nano Letters, 2014, 14(6): 3432–3438

[111] Zhang T, Han S, Guo W. . Continuous carbon nanotube composite fibers for flexible aqueous lithium-ion batteries. Sustainable Materials and Technologies, 2019, 20: e00096

[112] Zhong G, Yu J, Zhuang P. . Ultralong MnO@C nanowires with internal voids anchored between graphene as a robust high performance anode for flexible Li-ion battery. Electrochimica Acta, 2019, 296: 276–282

[113] Gu T, Cao Z, Wei B. All-manganese-based binder-free stretchable lithium-ion batteries. Advanced Energy Materials, 2017, 7(18): 1700369

[114] Jung Y, Jeong Y C, Kim J H. . One step preparation and excellent performance of CNT yarn based flexible micro lithium-ion batteries. Energy Storage Materials, 2016, 5: 1–7

[115] Ren J, Zhang Y, Bai W. . Elastic and wearable wire-shaped lithium-ion battery with high electrochemical performance. Angewandte Chemie International Edition, 2014, 53(30): 7864–7869

[116] Liu T, Zhang M, Wang Y L. . Engineering the surface/interface of horizontally oriented carbon nanotube macrofilm for foldable lithium-ion battery withstanding variable weather. Advanced Energy Materials, 2018, 8(30): 1802349

[117] Wei D, Shen W, Xu T. . Ultra-flexible and foldable gel polymer lithium-ion batteries enabling scalable production. Materials Today. Energy, 2022, 23: 100889

[118] Mu K W, Liu K X, Wang Z Y. . An electrolyte-phobic carbon nanotube current collector for high-voltage foldable lithium-ion batteries. Journal of Materials Chemistry. A, Materials for Energy and Sustainability, 2020, 8(37): 19444–19453

[119] Jiang X, Chen Y, Meng X. . The impact of electrode with carbon materials on safety performance of lithium-ion batteries: A review. Carbon, 2022, 191: 448–470

[120] Liu Y, Zhang R, Wang J. . Current and future lithium-ion battery manufacturing. iScience, 2021, 24(4): 102332

[121] Guo L, Zhao N, Li J. . Surface double phase network modified lithium rich layered oxides with improved rate capability for Li-ion batteries. ACS Applied Materials & Interfaces, 2015, 7(1): 391–399

[122] Zhang C X, Mei S L, Chen X H. . Electrochemical construction of functional polymers and their application advances in lithium batteries. Journal of Materials Chemistry. C, Materials for Optical and Electronic Devices, 2021, 9(48): 17182–17200

[123] Liu Z, Qin L, Cao X. . Ion migration and defect effect of electrode materials in multivalent-ion batteries. Progress in Materials Science, 2022, 125: 100911

[124] Zhang X, Li Y, Lin Y. . A flexible LiFePO₄/carbon nanotube/reduced graphene oxide film as self-supporting cathode electrode for lithium-ion battery. Ionics, 2020, 26(3): 1537–1546

[125] Bao Y, Zhang X, Zhang X. . Free-standing and flexible LiMnTiO₄/carbon nanotube cathodes for high performance lithium-ion batteries. Journal of Power Sources, 2016, 321: 120–125

[126] Fang X, Shen C, Ge M. . High-power lithium-ion batteries based on flexible and light-weight cathode of LiNi_0.5Mn_1.5O₄/carbon nanotube film. Nano Energy, 2015, 12: 43–51

[127] Yang P, Xi X, Huang T. . An acid-assisted vacuum filtration approach towards flexible PDI/SWCNT cathodes for highly stable organic lithium-ion batteries. Electrochimica Acta, 2020, 338: 135771

[128] Wu H, Shevlin S A, Meng Q. . Flexible and binder-free organic cathode for high-performance lithium-ion batteries. Advanced Materials, 2014, 26(20): 3338

[129] Wu H, Meng Q, Yang Q. . Large-area polyimide/SWCNT nanocable cathode for flexible lithium-ion batteries. Advanced Materials, 2015, 27(41): 6504

[130] Wu S, Wu H, Zou M. . Short-range ordered graphitized-carbon nanotubes with large cavity as high-performance lithium-ion battery anodes. Carbon, 2020, 158: 642–650

[131] Wang K, Luo S, Wu Y. . Super-aligned carbon nanotube films as current collectors for lightweight and flexible lithium-ion batteries. Advanced Functional Materials, 2013, 23(7): 846–853

[132] Yuan W, Wang B, Wu H. . A flexible 3D nitrogen-doped carbon foam@CNTs hybrid hosting TiO₂ nanoparticles as free-standing electrode for ultra-long cycling lithium-ion batteries. Journal of Power Sources, 2018, 379: 10–19

[133] Feng H, Tang L, Zeng G. . Core-shell nanomaterials: Applications in energy storage and conversion. Advances in Colloid and Interface Science, 2019, 267: 26–46

[134] Jiang L, Yuan X, Liang J. . Nanostructured core-shell electrode materials for electrochemical capacitors. Journal of Power Sources, 2016, 331: 408–425

[135] Kwon Y H, Woo S W, Jung H R. . Cable-type flexible lithium-ion battery based on hollow multi-helix electrodes. Advanced Materials, 2012, 24(38): 5192–5197

[136] Ren J, Li L, Chen C. . Twisting carbon nanotube fibers for both wire-shaped micro-supercapacitor and micro-battery. Advanced Materials, 2013, 25(8): 1155–1159

[137] Sun C F, Zhu H, Baker E B III. . Weavable high-capacity electrodes. Nano Energy, 2013, 2(5): 987–994

[138] Zhang Y, Bai W, Cheng X. . Flexible and stretchable lithium-ion batteries and supercapacitors based on electrically conducting carbon nanotube fiber springs. Angewandte Chemie International Edition, 2014, 53(52): 14564–14568

[139] Zhang X, Xu Z, Kong S. . Nanoflowers-like LiTi₂(PO₄)₃ on carbon nanotube fibers as novel binder-free anodes for high-performance fiber-shaped aqueous rechargeable lithium-ion batteries. Journal of Energy Storage, 2023, 64: 107249

[140] Ahmad Y, Colin M, Gervillie-Mouravieff C. . Carbon in lithium-ion and post-lithium-ion batteries: Recent features. Synthetic Metals, 2021, 280: 116864

[141] Cheng X, Pan J, Zhao Y. . Gel polymer electrolytes for electrochemical energy storage. Advanced Energy Materials, 2018, 8(7): 1702184

[142] Guo X, Chen R, Wu F. Use of thin film materials in flexible lithium-ion batteries. Journal of the Chinese Ceramic Society, 2019, 47(10): 1386–1395

[143] Chew S Y, Ng S H, Wang J. . Flexible free-standing carbon nanotube films for model lithium-ion batteries. Carbon, 2009, 47(13): 2976–2983

[144] Kim S D, Lee J G, Kim T G. . Additive-free electrode fabrication with reduced graphene oxide using supersonic kinetic spray for flexible lithium-ion batteries. Carbon, 2018, 139: 195–204

[145] Li F, Yue H, Yang Z. . Flexible free-standing graphene foam supported silicon films as high capacity anodes for lithium ion batteries. Materials Letters, 2014, 128: 132–135

[146] Guo W, Yan X, Hou F. . Flexible and free-standing SiO_x/CNT composite films for high capacity and durable lithium-ion batteries. Carbon, 2019, 152: 888–897

[147] Kang T, Ma Z, Zuo X. . Preparation of flexible self-supporting 3D SiO_x-based membrane anodes with stabilized electrochemical performances for lithium-ion batteries. Energy Technology, 2019, 7(3): 1800635

[148] Fu J, Liu H, Liao L. . Ultrathin Si/CNTs paper-like composite for flexible Li-ion battery anode with high volumetric capacity. Frontiers in Chemistry, 2018, 6: 624

[149] Guo W, Si W, Zhang T. . Ultrathin Ni_xCo_y-silicate nanosheets natively anchored on CNTs films for flexible lithium-ion batteries. Journal of Energy Chemistry, 2021, 54: 746–753

[150] Cao H, Zhou X, Deng W. . Layer structured graphene/porous ZnCo₂O₄ composite film for high performance flexible lithium-ion batteries. Chemical Engineering Journal, 2018, 343: 654–661

[151] Han J H, Shin K H, Lee Y J. Scalable binder-free freestanding electrodes based on a cellulose acetate-assisted carbon nanotube fibrous network for practical flexible Li-ion batteries. ACS Applied Materials & Interfaces, 2021, 13(5): 6375–6384

[152] Noerochim L, Wang J Z, Chou S L. . Free-standing single-walled carbon nanotube/SnO₂ anode paper for flexible lithium-ion batteries. Carbon, 2012, 50(3): 1289–1297

[153] Chen X, Tang H, Huang Z. . Flexible bismuth selenide/graphene composite paper for lithium-ion batteries. Ceramics International, 2017, 43(1): 1437–1442

[154] Park S K, Seong C Y, Yoo S. . Porous Mn₃O₄ nanorod/reduced graphene oxide hybrid paper as a flexible and binder-free anode material for lithium-ion battery. Energy, 2016, 99: 266–273

[155] Liu X, Zhang X, Ma S. . Flexible amorphous MoS₂ nanoflakes/N-doped carbon microtubes/reduced graphite oxide composite paper as binder free anode for full cell lithium-ion batteries. Electrochimica Acta, 2020, 333: 135568

[156] Ji Z, Wang H, Chen Z. . A both microscopically and macroscopically intrinsic self-healing long lifespan yarn battery. Energy Storage Materials, 2020, 28: 334–341

[157] Zhao Y, Guo J. Development of flexible Li-ion batteries for flexible electronics. InfoMat, 2020, 2(5): 866–878

[158] Kammoun M, Berg S, Ardebili H. Stretchable spiral thin-film battery capable of out-of-plane deformation. Journal of Power Sources, 2016, 332: 406–412

[159] Hu J W, Wu Z P, Zhong S W. . Folding insensitive, high energy density lithium-ion battery featuring carbon nanotube current collectors. Carbon, 2015, 87: 292–298

[160] Tong X, Tian Z, Sun J. . Self-healing flexible/stretchable energy storage devices. Materials Today, 2021, 44: 78–104

[161] Ezeigwe E R, Dong L, Manjunatha R. . A review of self-healing electrode and electrolyte materials and their mitigating degradation of lithium batteries. Nano Energy, 2021, 84: 105907

[162] Nam J, Jang W, Rajeev K K. . Ion-conductive self-healing polymer network based on reversible imine bonding for Si electrodes. Journal of Power Sources, 2021, 499: 229968

[163] Nam J, Kim E, Rajeev K K. . A conductive self-healing polymeric binder using hydrogen bonding for Si anodes in lithium-ion batteries. Scientific Reports, 2020, 10(1): 14966

[164] Wang Y, Xu H, Chen X. . Novel constructive self-healing binder for silicon anodes with high mass loading in lithium-ion batteries. Energy Storage Materials, 2021, 38: 121–129

[165] Zhao Y, Zhang Y, Sun H. . A self-healing aqueous lithium-ion battery. Angewandte Chemie International Edition, 2016, 55(46): 14384–14386

[166] Rao J, Liu N, Zhang Z. . All-fiber-based quasi-solid-state lithium-ion battery towards wearable electronic devices with outstanding flexibility and self-healing ability. Nano Energy, 2018, 51: 425–433

[167] Kuznetsov O A, Mohanty S, Pigos E. . High energy density flexible and ecofriendly lithium-ion smart battery. Energy Storage Materials, 2023, 54: 266–275

RIGHTS & PERMISSIONS

Higher Education Press