Optimization Strategies Toward Functional Sodium‐Ion Batteries

Exploration of alternative energy storage systems has been more than necessary in view of the supply risks haunting lithium‐ion batteries. Among various alternative electrochemical energy storage devices, sodium‐ion battery outstands with advantages of cost‐effectiveness and comparable energy density with lithium‐ion batteries. Thanks to the similar electrochemical mechanism, the research and development of lithium‐ion batteries have forged a solid foundation for sodium‐ion battery explorations. Advancements in sodium‐ion batteries have been witnessed in terms of superior electrochemical performance and broader application scenarios. Here, the strategies adopted to optimize the battery components (cathode, anode, electrolyte, separator, binder, current collector, etc.) and the cost, safety, and commercialization issues in sodium‐ion batteries are summarized and discussed. Based on these optimization strategies, assembly of functional (flexible, stretchable, self‐healable, and self‐chargeable) and integrated sodium‐ion batteries (−actuators, −sensors, electrochromic, etc.) have been realized. Despite these achievements, challenges including energy density, scalability, trade‐off between energy density and functionality, cost, etc. are to be addressed for sodium‐ion battery commercialization. This review aims at providing an overview of the up‐to‐date achievements in sodium‐ion batteries and serves to inspire more efforts in designing upgraded sodium‐ion batteries.

abundance of raw materials, high rate, and good electrochemical reversibility in a wide temperature range, SIBs have great promises for grid applications, [21][22][23] and the renaissance of SIBs has been witnesses in the 21st century after the invention of Na-S battery by Ford in 1960s. [27][28][29][30] Given the similar electrochemistry between LIB and SIB, the strategies and experiences accumulated in the development of LIBs can be transferred toward assembly of superior SIBs. There have been many excellent reviews discussing the advancements of SIBs in electrode materials, characterization techniques, functionality, structure, etc. [2,13,18,[21][22][23]27,[30][31][32][33][34][35][36] Here, as presented in Figure 1b, this review mainly focuses on the available strategies for achieving enhanced electrochemical performance in cathode and anode (defects, nanostructuring, composites, etc.), electrolyte (additives, pseudo-solid, water-insalt, etc.), separator, current collector, etc., the cost, safety, and commercialization issues of SIBs, and the challenges and developments of advanced functional (flexible, stretchable, self-healable, and selfchargeable), and integrated SIBs (SIBs actuator, SIBs sensor, and electrochromic EC SIBs). This review aims at providing an overview of the up-to-date development of SIBs and pointing out the future directions for SIB commercialization.

Cathode
Cathode material is one of the key components of a SIB as it governs more than 60% of SIB performance. [37] Most cathode materials work on the intercalation/extraction mechanism and experience smaller volume variations than the anodes during (de)sodiation. However, the large size of Na + still causes unavoidable structure changes in cathode during extended cycling. As the cathode materials primarily govern the energy density of SIBs, constructing high-energy SIBs mostly relies on cathode materials with numerous active sites, high reaction potential, unhindered ionic pathways, and stable structures. [35] The promising cathode material families for SIBs include layered oxides, polyanionic compounds, and Prussian blue analogs (PBAs). The sodium-based layered transition metal oxide (LTMO) Na x MO 2 (M corresponds to transition metal) has a characteristic structure of alternately stacked edgesharing octahedral layers, MO 6 and NaO 6 . Based on whether the Na + ion is in an octahedral or a prismatic coordination environment, these layered materials may be categorized as O3 or P2 using Delmas' nomenclature. These materials undergo complex phase transitions during cycling with significant contraction/expansion of the crystal structure resulting in structural collapse and a fast capacity decline. [38,39] Polyanionic materials are mainly made up of two different entities linked together by a corner-or edge-sharing scheme: octahedral metal MO x (M = Fe, Mn, Co, etc.) and tetrahedral (XO 4 ) or trigonal (XO 3 ) anionic groups (X = P, Si, S, etc.). [27,38,39] Stable frameworks endow polyanionic compounds with extended cycle life. However, because of the polyanion group's large molecular weight, these insertion compounds have poor conductivity and compromised gravimetric capacity. [40] PBAs are composed of a metal-organic framework (MOF) with the general formula Na x M[M 0 (CN) 6 ] y ÁnH 2 O (where M, M 0 corresponds to Fe, Mn, Ni, etc.). [41] PB and PBAs have shown potential sodium storage capabilities because of their strong electrochemical reversibility and Jingwei Chen joined the School of Materials Science and Engineering, Ocean University of China, as an Associate Professor in 2021. He obtained his Ph.D. from Nanyang Technological University (NTU) in 2019 under the supervision of Prof. Lee Pooi See, and did his postdoctoral research at NTU and the Singapore-HUJ Alliance for Research and Enterprise, Campus for Research Excellence and Technological Enterprise. His research focuses on the design, fabrication, and mechanism understanding of novel electrode materials for electrochemical energy storage and energy savings devices, including metal-ion batteries and electrochromic devices. grid open architecture with large interstitial spaces. [42] Although these materials are considered to have potential prospects, their inherent issues associated with vacancies and water coordination result in low capacity and rate performance, preventing their practical application. [38,39] 2.1. 1

. Defect Engineering
In order to improve the electrochemical performance of cathode, defect engineering has emerged as a potent approach for modifying the electrical and crystal structures. Vacancy engineering and heteroatom doping-dominated defect engineering can tune materials' inherent characteristics in unique ways. [43] Feng et al. recently reported a stable magnesium-doped cobalt-free layered P2-transition metal oxide cathode, Na 0. 67 Ni 0.18 Mg 0.15 Mn 0.67 O 2 , with an initial discharge capacity of 123 mAh g −1 and retention of 92% after 100 cycles at 0.1 C. Mg doping reduced the polarization by decreasing the volume expansion coefficient, hence rendering improved cyclic stability. The phase transition between P2 and O2 is inhibited by replacing Ni 2+ with electrochemically inert Mg 2+ , which enables more Na + to remain in the triangular prism during charging, allowing high rate performance (67 mAh g −1 at 20 C). [44] Apart from alleviating phase transformation in LTMO-type cathode materials, heteroatom doping could also enhance electronic transport for polyanion-type cathodes. For example, Jiang et al. prepared an Mg/Ti co-doped Na 3 V 2 (PO 4 ) 3 nanoparticles coated with carbon nanotube that showed a high preliminary capacity of 71.9 mAh g −1 at a high rate of 100 C, 76% of the initial capacity was retained after 1000 cycles. A favorable doping effect is generated by replacing V 3+ with Mg 2+ /Ti 4+ . The introduction of Mg 2+ generates an additional acceptor level and increases the electronic conductivity due to the presence of cavities. Due to the n-type substitution, Ti 4+ can bring in excess electrons when it replaces V 3+ in order to satisfy the charge compensation. The difference in ionic radius of V 3+ and Mg 2+ /Ti 4+ provides an enlarged interstitial site for facilitated Na + mobility. [45] Cation and anion co-doping provide synergistic effects in improving electrodes' structural stability and electrochemical performance. Recently, Wen et al. developed a Ti/F co-doped-layered P2-type Na 0.7 MnO 2.05 cathode with capacity of 227 mAh g −1 at 20 mA g −1 (~0.09 C). The optimized electrode, NMO-0.1TF (1 at.% Ti, 8 at.% F), showed a capacity of 133 mAh g −1 at 1 A g −1 with a retention of 96.2% after 200 cycles. [46] In order to achieve enhanced electrochemical performance, Ti 4+ /F − codoping suppresses the order of Na + /vacancy and the phase transition from P2-O2. A strong M-F bond can be formed with the electronegative F − , which can impede the gliding of the TMO 2 layers. Additionally, Ti 4+ is homogenously incorporated in Mn 4+ sites due to similar ionic radius and same valence, which disorders Na + /vacancy arrangement and suppresses the P2-O2 phase transformation. [47] The improved Na diffusion and air stability were achieved by introducing sodium vacancies in LTMO cathodes. An environmentally stable and kinetically superior O3-type layered oxide cathode material was prepared by introducing Na vacancies in NaLi 0.12 Ni 0.25 Fe 0.15 Mn 0.48 O 2 (NaFNM) (Figure 2a). [48] The Na-deficient Na x FNM cathode showed no charge capacity decay after 3 days of air exposure, while NaFNM suffered from a capacity decrease of 9.6 mAh g −1 . The deficiency of Na results in increased valence states of transition metal ions, which enhances the resistance against oxidation and suppresses unfavorable spontaneous reactions with atmosphere. Na vacancies in the lattice further boost kinetics by creating additional Na diffusion sites and facilitating the transition of Na + between ODH (hard oxygen dumbbell hop) and TSH (facile tetrahedral site hop) (Figure 2b). The Na 0.93 FNM cathode displayed a discharge capacity of 99.2 mAh g −1 at 1600 mA g −1 and a high capacity retention of 82.8% after 200 cycles as shown in Figure 2c. [48] This strategy is more applicable for layered oxide and polyanionic compound-type cathode materials as it mitigates the complex phase transition, and increases the electronic conductivity, respectively. However, excess doping could initiate capacity fading due to extra dopant ions occupying the minor sodium sites. [49]

Nanostructuring
Nanoengineering or nanostructuring improves electrochemical characteristics, such as high specific capacity, superior rate capability, and stable cycle life of the electrode materials by increasing the specific surface area and Na + storage sites. These could be achieved through designing different nanostructural features, including nanoparticles [0D], nanofibers/rods [1D], nanosheets [2D], and incorporation of composites, for example, carbon/graphene coating. [42] Sengupta et al. synthesized hexagonal nanocrystals of layered P2-type metal oxide, Na 0. 66 Ni 0.33 Mn 0.5-Ti 0.16 O 2 (NMTNO nano ), through microwave-assisted solid-state technique (Figure 3a,b). The nanocrystals provided lower diffusion-Daniel H.C. Chua received his B.Sc from National University of Singapore (NUS) and his Ph.D. from University of Cambridge. He was a Post Doctoral Research Fellow and has worked in the semiconductor and hard disk media industry as an Applications Engineer. Currently, he is an Associate Professor in Department of Materials Science and Engineering in NUS. His research focuses on developing alternative growth techniques to design and fabricate various types of low dimensional materials for electronics and clean energy applications.
induced stress at a high C rate due to a porous secondary structure which allows better electrolyte infiltration with shortened Na-ion diffusion pathways. The NMTNO nano showed a reversible discharge capacity of 152.65 mAh g −1 at 0.1 C, improved rate performance, and 94.8% capacity retention after 100 cycles and 90.87% over 1000 cycles at 0.5 C as shown in Figure 3c-e. [50] Similarly, multilayer-oriented stacked nanoflakes of P2-type layered oxide, Na 0. 66 Ni 0.16 Mn 0.66-Cu 0.11 Mg 0.125 O 2 , manifests superior cycling stability (81.4% capacity retention at 5 C after 500 cycles). [51] Various synthesis methods are adopted to realize structural modification of PBA-based cathode materials. Well-dispersed nanocubes of monoclinic nickel hexacyanoferrate (NiHCF-NCs) were synthesized using diethylenetriaminepentaacetic acid disodium (Na 2 DTPA)-assisted coprecipitation, [52] during which the rate of coordination of Ni 2+ and [Fe(CN) 6 ] 4− is controlled by DTPA 2− chelation. Taking advantage of the slow crystallization process for the formation of nanocubes, NiHCF-NCs displayed high sodium contents and less vacancies in the [Fe(CN) 6 ] atoms, leading to excellent cyclic stability (capacity retention of 97.1% after 2000 cycles) and rate performance (64.5 mAh g −1 at 4000 mA g −1 ). [52] Using a facile solid-state approach, a mesoporous electrode with sponge-like structure engaged in a dual-carbon matrix, Na 4 Fe 3 (PO 4 ) 2 (P 2 O 7 ) (NFPP)@C@rGO, was prepared. The amorphous carbon layer (~4 nm) suppresses the NFPP nanoparticles from growing and accommodates volume expansion, and graphene nanosheets provide conductive pathways for charge transfer. As a result of nanostructure regulation and the dual-nanocarbon coating, high rate performance (80.7 mAh g −1 at 20 C) and ultra-stable cyclic stability of 30 000 cycles at 20 C with 86.7% capacity retention can be achieved. [53] Nanostructuring of cathode materials is an optimal technique for all three types of cathode because it provides shorter electronic/ionic diffusion pathways, complete infiltration of electrolytes, good conductivity, and structural integrity, leading to superior cyclic stability and rate performance. Despite enhanced electrolyte infiltration, these structures can lead to electrolyte decomposition due to side reactions. Additionally, nanostructuring often produces certain defects, which may compromise their sodium storage capacity. Thus, developing nanostructured cathode materials with low defect density, limited side reaction, and high crystallinity remains difficult. [42]

Composition Tuning
Through chemical composition control, the unique advantages of different elements or phases are collectively harvested to optimize the electrochemical performance. Li et al. prepared a tri-phase LTMO cathode, P3/P2/O3-Na 0.674 Ni 0.319 Mn 0.590 O 2 (NNMO), by modifying the calcination temperature and gas atmosphere. The staggered design and gradient Naextraction/insertion voltages of different phases improve the electrochemical performance. The transition metal layer's slip is restricted by the neighboring phase structure, inhibiting the phase change (P2 to O2) that causes capacity fading. The tri-phase P3/ P2/O3-NNMO showed capacity of 80 mAh g −1 over 200 cycles and superior rate performance (100 mAh g −1 at 750 mA g −1 ) compared to biphase NNMO material. [54] On the other hand, a Co/Ni-free biphasiclayered P2/O3 transition metal oxide cathode, Na 0.8 Li 0.2 Fe 0.2 Mn 0.6 O 2 , was prepared through the sol-gel method followed by calcination. [55] The electrode showed capacity retention of 82% after 100 cycles, with an initial capacity of 174 mAh g −1 (at 0.1 C), mostly contributed by the Fe 3+ /Fe 4+ redox pair in addition to the oxygen and Mn 3+/4+ redox. [55] A "water-in-salt" nanoreactor strategy has been adopted to synthesize highly crystalline Mn-based PBA (MNHCF-S-170), [56] with reduced defects and water content. The unique monoclinic MNHCF-S-170 provided high sodium-ion content during the charge and discharge process, and fast Na-ion transportation, leading to better kinetics and specific capacity. The MNHCF-S-170 displayed a high specific capacity of 164 mAh g −1 at 10 mA g −1 and remarkable reversible discharge capacity at −10°C (127 mAh g −1 at 20 mA g −1 ) and 50°C (141 mAh g −1 at 20 mA g −1 ). [56] Meanwhile, electrostatic spray deposition method was used to allow the coexistence of amorphous and Figure 2. a) Schematic representation of the variation in parameters following Na vacancy introduction. b) The effect of various dual performance techniques. The schematic diagram of optimizing dual performance demonstrates the effect of adding Na vacancies technique to boost storage/rate performance. c) Cyclic stability of Na 0.93 LNFM and NaLNFM at 1600 mA g −1 . Reproduced with permission. [48] Copyright 2022, John Wiley & Sons. nanocrystalline Na 3 V 2 (PO 4 ) 3 (NVP) particles that create a disordered structure. [57] The disordered structure weakened the link between oxygen and sodium ions, activating the V 5+ /V 4+ redox couple and accomplishing a three-electron reaction with a specific capacity of 179.6 mAh g −1 at 0.2 C. [57] Composition tuning focuses on providing a more comprehensive Na + diffusion pathway by increasing disorderliness in the structure through varying the composition of different phases. This strategy was primarily developed for layered-oxide cathodes to achieve synergistic effects of P2, P3, and O3 phases, however, PBA and polyanion cathode materials could also benefit from this composition tuning of phases like monoclinic and amorphous/crystalline, respectively. Despite the fact that this modification technique has certain advantages, the preparation of different compositions and the mechanism behind the interactions between distinct phases require more investigation.
The above-outlined strategies of cathode optimization comprising defects engineering, nanostructuring, and composition tuning aimed at achieving higher effective electrochemical active surfaces for high energy densities and shortened ionic diffusion path lengths for high rate capabilities. In addition to the design of effective cathode materials, other factors like the evolution of complementary anode and compatible electrolytes, additives, selection of appropriate binders, separators, and optimum operating conditions are also to be considered.

Anode
Sodium-ion battery anodes exhibit lower voltage and unstable performance than cathodes, slowing SIB development and commercialization. Na metal is unsuitable as an anode due to dendrite growth, severe reactivity with electrolytes, and a low melting point of 97.7°C. [58] In order to identify a suitable anode for SIBs, various materials, including carbonaceous materials, transition metal oxides/sulfides, alloying compounds, and organic compounds, are explored. [35] SIB anodes should have a high capacity, cycling stability, rate capability, initial coulombic efficiency (ICE), and a low-charge/-discharge plateau to avoid sodium dendrites. [58,59] These materials could be classified based on reaction mechanisms, including insertion, conversion, and alloying reactions. Carbonaceous materials like graphite, hard carbon, graphene, and soft carbon mostly experience an insertion-based mechanism. Among them, hard carbon shows better electrochemical capacity, low charge/ Figure 3. Nanostructuring of cathode. a) A schematic of the microwave-irradiated rapid synthesis of NMTNO nano for Na-ion battery cathodes. b) SEM image of primary crystal particle of NMTO nano . c) Galvanostatic charge-discharge curves of NMTO nano at 0.1 C, potential range 2.5-4.3 V. d) Rate performance and e) cycling stability of NMTO nano and NMTO micro . d) Reproduced with permission. [50] Copyright 2023, John Wiley & Sons.
Energy Environ. Mater. 2023, 6, e12633 5 of 23 discharge potential plateau, and an economical manufacturing process. However, hard carbon anodes show low reversibility of the sodiation/ desodiation process and electrolyte decomposition in the initial cycle leading to low ICE. The stability and stoichiometry of Na-intercalated compounds limit the capacities of carbon-based materials, which seldom exceed 200-400 mAh g −1 . [58] These materials generally have less volume change. Alloying reactions can generate Na-M intermetallic binary compounds (M ≈ Sn, Sb, Bi, Si, Ge, and P). [59] Unlike an insertion process, a single atom can alloy with numerous Na atoms, resulting in capacities between 300 and 2000 mAh g −1 . [58] These materials have significant volume variations that reduce cyclability. Conversion reactions utilize transition metal oxides/sulfides as anode materials. During the sodium insertion procedure, the conversion reaction is typically accompanied by intercalation or alloying. These materials have large specific capacities of 200-1800 mAh g −1 because each electropositive atom can transfer several electrons. These materials have substantial irreversible capacity losses, resulting in poor ICE and swelling during the cycles. [58]

Defect Engineering
Heteroatom doping is a viable technique for enhancing the electrochemical performance of SIB anodes. Doping improves the intrinsic electronic conductivity and enhances the accessibility of active sites with Na ions. [43] Also, it is well known that the ICE is affected by defects in hard carbon and its capacity is affected by the arrangement of turbostratic nanodomains in hard carbon. Huang et al. developed a nitrogendoped hard carbon anode (PTA-Lys-800) with a high proportion of active pyrrolic nitrogen by the Mannich reaction employing tannic acid (TA) and amino acid as precursors. The high percentage of N-doping introduces a hierarchical pore structure to provide shorter Na + diffusion leading to rapid transport and enhanced electrical conductivity for prolonged cycle life. The PTA-Lys-800 displayed a capacity of 131.1 mAh g −1 at a high rate of 4 A g −1 after 5000 cycles. [60] Similarly, the rich phosphorous-doped carbon balls (PCBs) were developed by in situ doping via solution plasma process. The hierarchical pore structure of PCB and high P-doping content (4 at%) provided abundant active sites and shorter diffusion lengths for sodium ions. The solvated Na + in ether-based electrolyte renders a thin and robust SEI layer. The advantage of structural design combined with ether-based electrolyte provided high ICE of 75% and a reversible capacity of 83 mAh g −1 at 100 A g −1 after 40 000 cycles. [61] Lee et al. prepared N-doped crumpled graphene (NCG) anode via an aerosol spray-drying process (Figure 4a, c) with controlled composition of four different N precursors, including graphitic graphene, pyridinic, pyrrolic, and amino-functionalized graphene ( Figure 4b). The NCG prepared with ethylenediamine and 1-(3-Dimethylaminopropyl)-3-ethyl carbodiimide (EDC), NCG-EE, incorporated more defects into graphene and showed higher content of N doping. The synergistic impact of high N doping with large numbers of defects promoted surface-controlled charge storage in NCG-EE, showing specific capacities of 328 mAh g −1 at 0.05 A g −1 at room temperature and 107 mAh g −1 at 0.01 A g −1 at −40°C as shown in Figure 4d. [62] S/N co-doped graphite nanosheets (SNGNS) were synthesized through a three-roll-milling (TRM) exfoliation method followed by heating with thiourea. After 15-fold TRM exfoliation, the micro-nanostructure and increased interlayer spacing of graphite increased the number of accessible active sites for Na + insertion. The SNGNS 15 -600 showed an initial discharge capacity of 173 mAh g −1 at 1 A g −1 and high cyclic stability of 6000 cycles with~94% retention at 10 A g −1 . The co-doping provides the combined benefits of the pyrrole and pyridine N-atoms as great electron donors, leading to improved electrical conductivity, and the electrochemically active, covalently bonded S-atom contributes to pseudocapacitance. [63] Metal heteroatom doping of metal oxides/sulfides/selenides-based anodes can confer increased electronic conductivity by generating numerous defects, leading to the exposure of active sites. For example, Hu et al. reported a facile hydrothermal process with a calcination technique to fabricate Sndoped MoSe 2 nanosheets on graphene in which Sn altered the electronic structure and expanded interlamellar spacing to enhance the capacity and cyclic stability. The 9.6% Sn-MoSe 2 @GN exhibits superior electrochemical performance for SIBs. [64] After 1600 cycles, the specific capacity remained at 268.5 mAh g −1 at 1 A g −1 . The elemental doping alters the charge/ion state, bandgap, and crystal structure stability of active materials by modifying their intrinsic crystal structure, which increases the electronic conductivity and interlayer distance and provides additional pseudocapacitance. Despite certain established mechanisms, each heteroatom's unique role requires additional investigation. To have a deeper understanding of N and S doping, it is necessary to figure out the concentration and arrangement of N and S.

Nanostructuring
Nanostructuring of conversion-or alloy-type anode materials is a promising method for improving cycling/rate performance by minimizing absolute volume change, structural stress, and ion diffusion distance. [65] Li and co-workers demonstrated a facile chemical reduction process for the synthesis of ultrasmall SnSb nanoparticles (<2 nm) confined in MOF (ZIF-8)-derived N-doped carbon nanocages (SnSb@3D-NPC). The SnSb nanoparticles provide a high specific surface area to enhance the pseudocapacitive contribution and reduce the Na + diffusion path. The MOF-based nanocage prevents the self-agglomeration of nanoparticles and provides elastic voids to accommodate the large volume change, leading to superior cyclic stability. These synergies allow SnSb/3D-NPC to thrive as a promising anode material, as evidenced by its high reversible capacity of 693.6 mAh g −1 after 100 cycles at 0.2 A g −1 , excellent rate performance of 359.1 mAh g −1 at 20 A g −1 , and capacity retention of 266.6 mAh g −1 at 5 A g −1 after 15 000 cycles. [66] Similarly, Bi 2 O 3 and carbon nanofibers (CNF) formed a 1D nanocomposite via a facile preoxidation-assisted technique, in which Bi 2 O 3 nanodots were confined in the CNF structure, reducing the ions/electrons diffusion pathways and alleviating the effect of volume expansion. [67] To create hierarchical V 2 C/Fe 7 S 8 @C nanocomposites, Xiong et al. reported an in situ hydrothermal approach in which the Fe 7 S 8 nanoparticles are evenly attached to the V 2 C-MXene surface. V 2 C-Mxene is a highly ion-conductive substrate, building a rapid Na + migration network, enhancing charge transport, and providing polysulfide adsorption to improve reversibility. Many nucleation sites on the V 2 C surface promote Fe 7 S 8 particle nanometerization, which exposes additional active sites and makes Na + more accessible. The V 2 C/Fe 7 S 8 @C showed a high reversible capacity of 524.9 mAh g −1 after 200 cycles at 0.5 A g −1 . [68] Zhu et al. fabricated a 3D crumpled paper ball-like structure of carbon-pillared MoS 2 nanosheets and carbon nanotubes (CNTs) through a hydrothermal-annealing process. Highly elastic and resistant to aggregation, the crumpled MoS 2 nanosheet balls effectively release stress during cycling. Na + diffusion is aided by increased interlayer spacing due to pillared carbon. A reversible capacity of 295 mAh g −1 is maintained after 1500 cycles at 2.5 A g −1 and a superior rate capability of 299 mAh g −1 at 10 A g −1 . [65] Hollow structures are effective in promoting stability and rate capability. Hollow NaTi 2 (PO 4 ) 3 nanocubes anode prepared by hydrothermal approach could perform well in deep eutectic electrolyte when assembled with compatible cathode of Na 0.44 MnO 2 . [69] The assembled battery delivered an ultralong cycle life of up to 3500 cycles (with 90% capacity retention), an excellent energy density of 50.0 Wh kg −1 , and superior rate capability (maximum power density of 1500 W kg −1 ). [69] Hollow nanospheres of Cu 2 MoS 4 have also shown improved rate and cycling performance benefitted from fast and reversible Na + -ion storage. These bimetallic sulfide nanospheres can function via an intercalation process in an ether electrolyte or via a conversion mechanism in a carbonate electrolyte. [70] The characteristics such as shorter electronic/ionic diffusion pathway, higher electric conductivity, and improved strain accommodation due to mitigated volume change stem mostly from the unique nanoarchitecture, desirable composition, and the existence of porous/hollow structures. Aside from nanostructured anode materials' low volumetric energy density due to lower electrode tap density, other concerns of unwanted side reactions due to larger exposed surface area, and high cost due to complicated manufacturing methods must be addressed.

Heterostructure Engineering
The rational design of heterostructures can effectively regulate the movement of electrons, holes, and other excitons, resulting in the emergence of novel interfacial features. Heterojunction materials are primarily composed of semiconductors. The built-in electric field (due to the difference in bandgap and work function) is responsible for the acceleration of ion migration leading to better kinetics, and the chemically less reactive reaction intermediates formed at the junctions alleviate the volume change of active material. [71] Li et al. synthesized a heterostructured multilayer SnS-SnS 2 @GO nanosheets by a one-step solvothermal method to wrap SnS and SnS 2 nanosheets in GO. The SnS-SnS 2 @GO anode displays a discharge capacity of 601.1 mAh g −1 at 0.1 A g −1 and a capacity of 450.6 mAh g −1 after 100 cycles. The electronic/ionic conductivity of the SnS-SnS 2 @GO anode is increased due to the unique internal electric field in heterostructured SnS-SnS 2 p-n junctions and the GO modification. [72] Wang et al. (0.11 eV) than that in the MoS 2 @MoS 2 interface (1.28 eV). MoS 2 nanosheets with perforated channels provide short-ion migration paths and expose more active sites, improving electrolyte wettability and faster Na-ion storage. As a result, the MXene-MoS 2 anode demonstrated high rate capability and high cycling stability over 1000 cycles (capacity retention of 252.0 mAh g −1 at 1.0 A g −1 ). [73] Recently, Zhang et al. prepared a VS 2 /VO x heterostructure using the hydrolysis-oxidationcoupled reaction to construct a VO x layer on the VS 2 surface. This novel heterostructure's built-in electric field improved electrical conductivity and reaction kinetics and provided adequate active sites. Furthermore, the volumetric strain of the electrode can be buffered by the two-phase material, especially at high current densities. The use of VS 2 /VO x heterostructure as an anode led to significant improvement in electrochemical performance, a high specific capacity of 878.2 mAh g −1 at 0.2 A g −1 , superior rate capability of 654.8 mAh g −1 at 10 A g −1 , and excellent cyclic stability (721.6 mAh g −1 at 2 A g −1 after 1000 cycles) are achieved. [74] Despite the fast reaction kinetics, heterointerjunction suffers from separation and elimination due to continuous phase transition and substantial morphology changes. It is essential to address the longterm stability of heterointerfaces and the fine-tuning of bonding interactions between constituents.
Anode optimization methods including defect engineering, nanostructuring, and heterostructure engineering seek to minimize volume change, reduce charge diffusion distance, and enhance electronic conductivity, to provide excellent cyclic stability and faster kinetics. At the cell level, the electrochemical performance of SIBs is also affected by the choice of electrolytes, even though this is often overlooked in research. For example, the composition and properties of electrolytes have a major effect on the electrochemical reduction of electrolytes on hard carbon anodes, which will be discussed in detail in the next section.

Electrolyte
Electrolytes, being fundamental components of SIBs, should have broad electrochemical windows, good thermal stability, superior ionic conductivity, and render stable solid electrolyte interface (SEI). These characteristics are governed by the solvation effects and physicochemical qualities of the electrolyte components (solvents, salts, additives, and concentration). The inherent physicochemical property, such as ionic conductivity and electrochemical stability, directly influences the SIBs performance. On the other hand, the solvation effect controls the composition, thickness, and structure of SEI, and the electrolyte participation in the battery reaction indirectly influences the battery performance. Therefore, advanced electrolytes based on a wide range of materials and compositions are formulated to address the functional requirements of SIBs. [75] These electrolytes can be classified as aqueous, non-aqueous, or solid-state (polymer and inorganic electrolytes). Aqueous electrolytes have better ionic conductivities and are easier to handle during cell manufacturing under ambient settings. Furthermore, they are ecologically friendly albeit its inability to form a solid electrolyte interface (SEI) layer that restricts the electrochemical stability window (ESW) of the aqueous electrolytes. [76] The non-aqueous system is more promising for widespread utilization because of its high ionic conductivity, broad ESW, and good electrode wettability. However, the high-temperature instability and flammability of organic electrolytes may pose significant obstacles to their widespread use. [75] Polymer solid-state electrolytes based on polyethylene oxide (PEO), polyacrylonitrile (PAN), polymethylmethacrylate (PMMA), and polyvinylidene fluoride (PVDF) are flexible and have better interfacial contact, but their low ionic conductivity limits their uses. [77] Inorganic solid electrolytes typically consist of crystalline ceramics and amorphous glasses and are commonly metal based, for example, β″-Al 2 O 3 , Na 3 PS 4 , and NASICON (Na 1+x Zr 2 Si x P 3−-xO12 ). [78] Consequently, they are usually quite brittle and costly. The primary benefits of inorganic electrolytes are safety related, particularly their superior mechanical strength and great selectivity for Na + conduction. However, the high interfacial resistance of inorganic electrolytes restricts their electrochemical activity. [79] 2.3.1. Modification/Addition of Salt, Solvent, or Additives The electrochemical characteristics of electrolytes can be modified by addition of different salts, solvents, and additives to enhance ionic conductivity, ESW, and reduce flammability. Aqueous-based electrolyte prepared by Dong et al. benefited from the incorporation of conducting salt made of sodium (2,3,5,6-tetrafluorophenoxy) diethane sulfonate (Na-TFP) through a two-step derivatization method of perfluorobiphenol. The 0.5 M Na-TFP showed higher ionic conductivity (>60 mS cm −1 at 20°C and 70 mS cm −1 at 30°C) compared to 0.5 M NaClO 4 and 0.5 M Na 2 SO 4 , leading to superior rate performance of 38 mAh g −1 in a full cell having Na 2 VTi(PO 4 ) 3 cathode and activated carbon (AC) anode. This system provided better cyclic stability of 400 cycles at 2 C with a capacity retention of 90%. [80] To address the flammability issue of non-aqueous electrolytes, Colbin et al. reported a fluorine-free and flame-retardant non-aqueous electrolyte using 0.38 M sodium bis(oxalato)borate (NaBOB) in triethyl phosphate (TEP) and also provided satisfactory passivation of hard carbon at low salt concentration and without the use of any additives. This electrolyte exhibited ionic conductivity of 5 mS cm −1 at room temperature and a viscosity of 2.674 mPaÁs. The full cell, having hard carbon as anode and Prussian white as cathode, showed an average discharge capacity of~107 mAh g −1 after 1005 cycles at 0.05 C. [81] Similarly, 0.5 M NaBOB in trimethyl phosphate (TMP) showed effective passivation on hard carbon compared to NaPF 6 in TMP. [82] The introduction of additives can adjust the SIBs performance by altering SEI layer, modifying the solvation structure of Na ions, and providing additional functionalities (such as flame retardancy, overcharge protection, and enhanced conductivity). Recently, Lin et al. introduced an anion receptor additive, tris (2,2,2-trifluoroethyl) borate (TTFEB), in a 1,3-dioxolane (DOL)-based electrolyte, which improved the cyclic stability and rate capability of hard carbon electrodes. In this system, TTFEB acts as an anchor for the PF À 6 anion, preventing the onset of DOL polymerization as the Boron atom has a significant coordination effect with PF À 6 and forms a complex. In addition, a strong and homogenous SEI layer composed of B-O species and inorganicrich species are produced on the hard carbon electrode, preventing electrolyte decomposition, facilitating Na + ions diffusion, and enhancing electrochemical performance. The TTFEB addition significantly improved the cycling stability of hard carbon, with 89.7% capacity retention after 250 cycles, and showed a capacity of 258.6 mAh g −1 at 0.25 A g −1 . [83] Solvent or co-solvent addition is an approach to improve interface stability and reduce parasitic reactions. (trimethylsilyl)phosphite (TMSPi), and succinonitrile (SN). The 20% MA electrolyte showed better ionic conductivity of 13.65 mS cm −1 than EC:PC:DMC due to decreased drag force encountered by ions, previously dissociated by EC-PC, during migration. The decrease in charge transfer resistance with the addition of 20% MA was supported by the increased ionic conductivity and the altered solvation energy during insertion/extraction as MA participates in the primary or secondary solvation shell of sodium. Furthermore, the 20% MA (with additives) electrolyte guaranteed outstanding power capabilities, allowing rapid charging of 18650 Na-ion batteries to 84% state of charge (SOC) iñ 10 min. [84] In solid-state electrolytes, incorporating interstitial sodium ions or vacancies by aliovalent ion doping serves to boost the concentration of freely moving sodium ions and cause them to migrate together consistently. For example, Sc 3+ doping has been introduced into Na 3 Zr 2-Si 2 PO 12 (NASICON) electrolyte for a solid-state sodium battery. [85] A monoclinic-to-rhombohedral phase transition occurred because of larger ionic radii of Sc 3+ (0.745Å) than the Zr 4+ (0.72Å) in octahedral coordination, and more Na + ions are incorporated into the lattice for charge compensation. The rhombohedral structure exhibits excellent symmetry and great Na-ion conductivity because of its extensive diffusion channels for sodium conduction. At optimized doping concentration, Na 3.4 Zr 1.6 Sc 0.4 Si 2 PO 12 showed increased Na-O bond length and rhombohedral triangular bottlenecks areas, leading to high ionic conductivity at room temperature (1.77 × 10 −3 S cm −1 ) and reduced activation energy (0.26 eV). [85] To improve the interfacial interaction and kinetic mass transfer, coating of different metal layers (such as Pb, Sn, Cu, and Cr) onto the surface of NASICON-structured Na 3 Hf 2 Si 2 PO 12 (NHSP) electrolyte has been carried out to promote interfacial reaction. [86] It was found that lead-coated electrolyte enabled homogeneous charge distribution, dendrite prevention, and rapid charge transfer. The Pb-coated electrolyte rendered a discharge capacity of 103.7 mAh g −1 at 3 C compared to 10.5 mAh g −1 for bare NHSP. [86]

Pseudosolid Electrolyte
The merits of organic electrolytes, such as high ionic conductivity and outstanding interfacial properties, are combined with those of polymer electrolytes, such as adaptability, mechanical robustness, and thermal stability, to create gel polymer electrolytes. Compared to conventional solvents, ionic liquids comprised only of ions that exhibit favorable features, including reduced flammability and excellent thermal and chemical stability. [77] Ionogel electrolyte composed of NaBF 4 , 1-ethyl-3methylimidazolium tetrafluoroborate (EMIMBF4), and poly(vinylidene difluoride-co-hexafluoropropylene) (PVDF-HFP) shows remarkable ionic conductivity of 8.1 mS cm −1 at room temperature for micro-SIBs. [87] The NaBF 4 -IE displayed larger t Na þ (sodium-ion transference number) = 0.49 due to smaller sodium clusters facilitating high mobility. After 3000 cycles at 20 C, sodium titanate (NTO)|NaBF4-IE|NVP displayed a high volumetric capacity of 10.1 mAh cm −3 . [87] To achieve thin and conformal electrolytes that penetrate the porous electrodes with low contact resistance, a gel electrolyte has been formulated by encapsulating NaBF 4 and tetraglyme solvent in a mesoporous silica/ polymer matrix via the sol-gel method, in which silane precursor (TEOS) is combined with PVDF-HFP. [88] This gel electrolyte possesses excellent electrochemical properties, such as ionic conductivity of 0.9 mS cm −1 , ESW of 4 V, and an acceptable t Na þ (0.3). With this pseudosolid electrolyte, the as-assembled Na 3 (VO) 2 (PO 4 ) 2 F-based Na metal battery reaches energy density of 400 Wh kg −1 at 0.5 C. [88]

Modification of Salt Concentration
In traditional electrolytes, it is widely believed that the Na + solvation sheath has at least two layers, with the primary layer being the most firmly connected of the two. As concentration increases, the ion interaction is strengthened, and the solvent molecules are insufficient to fully solvate the cations. In addition, the solvation shells of different ions will be pressed together and made to share the solvent, allowing anions to be preferentially recruited into the Na ion's solvation sheath. Hence, anions decompose instead of the solvent, forming anion-derived inorganic-rich SEI layers that protect the SEI and electrodes and improve cycle performance. [89] This serves as the basis for water-in-salt electrolyte, for example, having 17 mol kg −1 NaClO 4 and 2 mol kg −1 NaOTF in the electrolyte could render an ESW of 2.8 V. [90] An SEI layer composed of sodium fluoride (NaF), sodium oxide (Na 2 O), and sodium hydroxide (NaOH) was found on the anode, which effectively limits further water decomposition. This electrolyte showed high ionic conductivity (95.25 mS cm −1 ) and lower viscosity (12.8 MPaÁs). [90] Following that, an all-climate aqueous SIBs with 17 mol kg −1 NaClO 4 water-in-salt electrolyte was demonstrated with exceptional cycle stability (12 000 cycles at 10 A g −1 ) of the full cell using Ni-based Prussian blue (NiHCF) cathode and 5,7,12,14-pentacenetetrone (PT) anode. The "water-in-salt" electrolyte efficiently inhibits the solubility of the organic anode, and the full cell can operate in a wide temperature range due to the low freezing point (−31.3°C) and excellent thermal stability (<200°C) of water-in-salt electrolyte. [91] Similarly, the concept of highly concentrated electrolytes for organic electrolytes was widely studied for the optimization of solvation structure and SIB performance. A highly concentrated electrolyte (3 M NaPF 6 ) was found to produce thin and more robust SEI on graphite cathode. This was ascertained through concentration optimization of sodium hexafluorophosphate (NaPF 6 ) in EMC/PC (1:1) and 10 vol.% fluoroethylene carbonate (FEC). [92] This thin layer allows rapid diffusion kinetics and therefore unaffected ion migration under various current densities. With 3 M NaPF 6 electrolyte, carbon fiber mat anode, and graphite cathode, the as-assembled cell showed a power density of 6494 W kg −1 at 269 Wh kg −1 at a high current density of 1.6 A g −1 . [92] Yan Jin et al. reported an advanced highly concentrated electrolyte using sodium bis(fluorosulfonyl)imide (NaFSI) salt in TEP for highvoltage SIBs. The full cell based on hard carbon anode and NaNi 0.68 Mn 0.22 Co 0.10 O 2 cathode showed a high capacity retention of 83.5% after 200 cycles (at 0.2 C) because of the stability of CEI (cathode electrolyte interphase) and SEI layers. The limited solubility of transition metals for the cathode cycled in NaFSI-TEP was attributed to the highly concentrated electrolyte with few free solvent molecules, which is only marginally capable of dissolving other species. Thus, inorganic-rich CEI can inhibit the dissolution of transition metals and also enhances the mechanical strength of SEI layer (rich in N-and S-based inorganic species), promoting the electrochemical performance of hard carbon. [93] The electrochemical performance of SIBs can be optimized by tuning the electrolyte's physicochemical properties (ionic conductivity and ESW) and solvation effect by adding various kinds of salts, solvents, and additives, altering the salt concentration and forming a gel. Apart from the above properties, optimizing mechanical properties of Energy Environ. Mater. 2023, 6, e12633 9 of 23 electrolyte is also essential for solid-state SIBs, as the mechanical strength and ionic conductivity are often inversely related.

Separator and Interface Membrane
The separator plays an important role as a barrier that separates the electrodes from short circuits while allowing Na-ions transport. There are many ways to improve the performance of SIBs through separator modification, where physical, material, and various chemical approaches have been reported.
In order to overcome the volume expansion yet maintain its mechanical stability, electrospinning technique can be used to fabricate expandable polymer-coated separator membranes. For example, Celgard membranes can be coated with PVDF nanofibers, reaching a high ionic conductivity of 1.25 × 10 3 S cm −1 and rendering enhanced discharge capacities of SIBs as compared to non-coated membranes. [94] The interface structure helps to lower interfacial impedance and thus excellent compatibility with sodium electrodes. The electrospun nanofiber coating also provided a 3D interconnected porous structure, which could capture a large amount of liquid electrolyte, and the attraction of the PVDF-co-CTFE polymer toward the liquid electrolyte also led to complete wetting of the coated separator. [94] Similarly, the highly versatile nanofiber-reinforced membranes were also reported in other nanostructures, such as the formation of mono-comb poly(siloxane-gethylene oxide) fiber membrane. This fiber membrane exhibits a high room-temperature ionic conductivity of 1.06 mS cm −1 with an ESW of 4.8 V. [95] A rechargeable Na/NVP solid-state full cell based on this membrane exhibits a high initial discharge capacity of 114.3 mAh g −1 at 0.2 C and capacity retention of 86.5% after 1000 cycles at 1 C.
Material modifications have been the key approach in optimizing compatibility of separators with different electrolytes. For example, a chemically stable polyacrylonitrile (PAN)/silica-aerogel separator (M-PSA) can be prepared using electrospinning and is further modified by hydrolyzing the nitrile group of PAN. [96] As such, it is made compatible with electrolytes of EC/PC, EC/DMC, and diglyme. The contact angles of this separator are, respectively, 9°, 12°, and 16°for EC/PC, EC/DMC, and diglyme, indicating superior wettability. The Na/NVP half cells assembled with M-PSA separator show high ion diffusion rate (1.98 mS cm −1 ), high capacity (85.15 mAh g −1 at 5 C), and long cycling life (capacity retention of 96.9% after 1000 cycles at 5 C).
To improve the porosity and increase electrolyte uptake, metal oxide reinforced framework structures such as TiO 2 dispersed into PVdF-HFP using phase inversion technique have been shown to reduce the crystallinity of the polymer. [97] The pore density increased with TiO 2 concentration and became maximum with 0.50 wt% TiO 2 . Using sodium hexafluorophosphate (NaPF 6 ) in EC/PC, the ionic conductivity of the membranes has been found to be~1.3 mS cm −1 . Cyclic voltammetry studies revealed a Na + transport number~0.31 with ESW of~3.5 V. [93] Other metal oxide reinforced structures such as ZrO 2 -reinforced nanofibrous membrane for SIB were also reported, through fabricating ZrO 2 -reinforced layer on the surface of the modified cellulose acetate membrane (MCA). [98] This membrane exhibited high chemical stability and excellent wettability (contact angle decreased from 26.8°to 7.4°), no dimensional change after being kept at 250°C for 30 min, and high tensile strength up to 1.15 MPa. This ZrO 2 @MCA-0.5 h (immersed in ZrO(NO 3 ) 2 solution for 0.5 h) exhibits a high ionic conductivity of 2.23 mS cm −1 . In addition, the as-assembled hard carbon/Na cell displayed a high discharge capacity of 224.8 mAh g −1 at 0.5 C. On the other hand, SiO 2 -reinforced polymers are matched to PVDF to form hybrid composite separator membrane, which is highly promising due to the high dielectric constant (8.4) of PVDF. The SiO 2 -PVDF membrane exhibits higher ionic conductivity of 4.7 × 10 −2 S cm −1 at room temperature. An initial discharge capacity of 178 mAh g −1 at 0.1 C can be achieved, with 98% columbic efficiency and 81% capacity retention after 50 cycles. [99] The more common chemical approaches in improving separator membranes are using salt incorporation into polymers through chemical dissolution, resulting in high ionic conductivity. PMMA-based sodium-ion conducting polymer gel electrolyte membranes containing sodium triflate salt dissolved in tetraethylene glycol dimethylether (TEGDME) has been reported to have room-temperature ionic conductivity of 3.6 × 10 −3 S cm −1 . [100] These membranes are also highly flexible while having an ESW of~4.4 V and t Na þ close to 0.37, and remain stable in the gel phase up to 150°C. The as-assembled Na-S battery cells displayed a stable open circuit potential of 2.24 V and delivers first discharge capacity of~677 mAh g −1 . GO (0.9 wt.%) nanosheets incorporated PEO/PVA (50/50 wt.%) composite membranes complexed with NaIO 4 salt at different concentrations were prepared for sodium ions transport. [101] Mechanical properties such as tensile strength have been shown to increase 1.8 times with 1.4 times increment in Young's modulus and room-temperature ionic conductivity of 1.36 × 10 −6 S cm −1 . This was attributed to strong interfacial interactions between functional groups associated with PEO, PVA polymer, and various oxygenated function groups over GO sheets.

Development of Binders
Through the development of suitable binders, the performances of SIB can be improved tremendously due to the ability of binders to accommodate volume expansion during sodiation. Recent advances in hybrid binder mixtures aim to enhance reaction kinetics. For example, benefitting from hybrid binder of poly(acrylic acid) PAA:PVDF, SbTe bimetallic anode was found to deliver superior cycling stability and rate capabilities. [102] In another report, PAA is found to be an effective adhesive binder in a FeF 2 -reduced graphene oxide nanocomposite. The electrode exhibits superior electrochemical performance including high capacity of 175 mAh g −1 at 0.2 A g −1 , high rate capability of 78 mAh g −1 at 10 A g −1 , and good cycling stability. [103] Green and watersoluble materials such as lignosulfonate can be used as binders and were found to affect the surface layer and SEI thickness upon cycling for Prussian white cathode cells. This cost-effective binder also leads to improved capacity retention and cycling stability. [104] The binding properties such as adhesion and conductivity can be harnessed from water-soluble natural biopolymers including guar gum, sodium alginate, and xathan gum. The rich -OH and COO-functional groups in the binder lead to better adhesion and conductivity, thereby enhancing electrochemical performance of P2-type Mn-based cathode for SIB, compared to PVDF binder. The dissolution of transition metal followed by structural degradation and capacity fading during cycling can be alleviated. [105] Binder chemistry based on carboxymethyl cellulose (CMC) and PVDF was investigated for TiO 2 electrode in SIB. It was found that cleavage of C-F and formation of NaF over the electrode caused the poor performance of PVDF binder, which is attributed to the electrochemical defluorination of PVDF. On the contrary, adoption of CMC Energy Environ. Mater. 2023, 6, e12633 10 of 23 binder rendered higher coulombic efficiency (CE). [106] Reduced graphene oxide (rGO) is an environmentally friendly and multifunctional conductive binder. The rGO not only acts as a binder and flexible backbone in the film which aids in the development of free-standing electrodes (which will be discussed in Section 3.1) but also as an active material for Na-ion storage and a conductive additive. [107]

Binder-Free Approach
Increasing efforts are directed to avoid the use of binders to minimize electrochemically inactive and electrical insulating components in the electrodes. SIBs based on binder-free electrodes have been reported with comparatively good properties. Many reports adopted carbon nanofibers or carbon-based nanostructured materials that provide conductive networks to enhance charge transfer efficiency. For example, mesoporous metal phosphide nanoarray on carbon felt (CF) was prepared as binder-free anodes for SIBs. A full SIB with the CoP 4 /CF anode and Na 3 V 2 (PO 4 ) 2 F 3 cathode delivers an average operating voltage of ≈3.0 V, a reversible capacity of 553 mAh g −1 , and very high energy density of ≈280 Wh kg −1 for SIB. The flexible battery attained a cycling life of up to 1000 times (>90% retention rate) and high rate capability of 535 mAh g −1 at a current density of 4 A g −1 . [108] Similarly, a binderfree, flexible, and self-supporting P-doped carbon cloth (FPCC) achieved a sodium storage capacity up to 242.4 mAh g −1 , and relatively good ICE of~72%, excellent rate capabilities (123.1 mAh g −1 at 1 A g −1 ), and long-term cycle life (~88% capacity retention after even 600 cycles at 0.2 A g −1 ). [109] Furthermore, SIB anodes based on TiO 2 nanorods grown on carbon fiber cloth (CFC) delivered a high discharge capacity of 148.7 mAh g −1 after 2000 cycles at 1 A g −1 . [110] Twodimensional SnS 2 nanosheet array formed on carbon cloth (CC) could well maintain the structural integrity, providing favorable electronic/ ionic transport kinetics. [111] Another approach is to embed active material within a carbon array, for example, bismuth nanodots are confined in carbon arrays derived from MOF. The obtained SIB anode could well accommodate the volume changes during sodiation/desodiation. Additional capacity with fast Na + diffusion kinetics is observed, thanks to the nitrogen-doped carbon arrays that have abundant active sites at the boundaries with bismuth nanodots. [112] On the other hand, a binder-free electrochemical co-deposited 3D Bi nanosheets and conductive CNT networks (Bi NS/ CNTs) were developed. The Bi-nanosheet/CNTs manifest strong interfacial adhesion and accommodated volume change upon alloying/dealloying, delivering capacity of 383.72 mAh g −1 at 0.1 A g −1 after 200 cycles, and 311.68 mAh g −1 at 1 A g −1 after 1000 cycles. A full SIB assembled with Bi NS/CNTs anode and Na 3 (VOPO 4 ) 2 F@rGO cathode shows a high operating voltage (above 4 V) and a high energy density of 221.99 Wh kg −1 , which outperforms many reported Bi-based analogs. [113] Recently, binder-free nanostructure composites Ge-Co-P with brutto-composition CoGe2P0.1 or CoGe2@GeP were prepared via electrolysis of aqueous electrolyte. The composites realized specific capacity of 425 mAh g −1 , reaching CE of 0.74 and 0.97 in the first and tenth cycles, respectively. [114] Other than carbon nanofibers, hybrid carbon foams with metal oxide as self-supporting structure are used as a binder-free electrode. Fe 2 O 3 @3DGF annealed at 400°C exhibited superior sodium storage properties with a high reversible capacity of 398 mAh g −1 after 200 cycles at 500 mA g −1 . [115] Furthermore, it has been shown that hard carbon fibers decorated with different electrochemical active materials can be used as the supporting framework and 3D conductive network. A porous NVP-coated hard carbon fiber cathode and a pre-sodiated graphene/SiC/hard carbon fiber anode formed a flexible SIB that delivered high output voltage (3.34 V), high energy density (234.1 Wh kg −1 at a high-current rate of 0.5 A g −1 ), ultralong cyclability (over 2905 cycles at 0.5 A g −1 and 1000 cycles at 5 A g −1 ), and high CE (approaching 100%). [116] The binder-free approach benefits from minimized mass of the inactive binder, however, it is challenging for binder-free electrodes to reach high loading mass and large volumetric changes, and dissolution of active materials might also occur during ions insertion and extraction.

Current Collector
Al and Cu current collectors are widely used in SIB due to their high conductivity. Surface properties of the metallic current collector are important to maintain high capacity in SIB. For example, electrochemically grown Cu sheet was formed by electroplating in a CuSO 4 -based aqueous solution with PAA, allowing a noticeable increased reversible capacity by 210 mAh g −1 from the first to the second cycle. The roughened Cu substrate also suppressed the detachment of the active material, thus maintaining a high capacity of 685 mAh g −1 with good capacity retention of more than 90%. [117] Aside from metals, alternative carbon-based and metal-carbon mixture current collectors have also been reported due to other desirable properties, for example, carbon-based current collector endows high mechanical flexibility, lower density, high conductivity, large surface area, and good chemical stability. One such non-metallic current collector is the Ti 3 C 2 T x current collector used for NVP electrode, which could achieve enhanced rate capability (81.9 mAh g −1 at 500 mA g −1 ) and cycling performance (81.5% over 200 cycles at 1 C) than Al foil. [118] Modified Al foil with graphene nanosheets formed a robust connection, reducing the electrode-current collector interfacial resistance by 20-fold compared with that of pure Al foil. At a high rate of 5 C, the capacity retention of NVP cathode with G-Al current collector is 74%, which is much higher than that with Al foil (22%). [119] The porous 3D graphene aerogel allowed embedding of Prussian blue (PB) nanocubes. The graphene framework not only offers mechanical support but also plays the role of a binder-free current collector. The composite electrode attains a relative high capacity (>70 mAh g −1 ) with weight percentage of PB <72%. [120] 2.5. Cost, Safety, and Commercialization

Cost
According to USGS data (https://www.usgs.gov/), as of 2021, the reserves of core raw materials of sodium batteries are higher, and Na is less difficult to exploit in contrast with Li. The content of lithium in the earth's crust is only 0.0065%, which is much lower than that of sodium resources (2.75%). The cost for LIB and SIB with same amount of energy stored was compared, using sodium carbonate and lithium carbonate as raw materials. As shown in Figure 5, the cheaper sodium carbonate allows a cost saving of $2.13 in the electrolyte and $36.82 in the cathode, with a materials cost saving of 3.8% and a total saving of 1.3% for the complete battery ($2981). Replacing Cu current collector with cheaper and lighter Al foils can bring in a more significant material cost saving of 8.7% and a total saving of 3.0% for the complete battery. [18,121] Energy Environ. Mater. 2023, 6, e12633 In addition to battery cost reduction ( Figure 5), recycling raw materials and the extension of service life can also reduce costs. The recyclable battery could be developed to further reduce costs. [122] Using Na 3 V 2 (PO 4 ) 3 as an electrode material and aluminum foil as the shared current collector, a high-power and durable SIB with solid-component recycling efficiency of more than 98.0% can be assembled. Electrolyte additives can stabilize the SEI layer as well as extend the battery life, thus further contributing to battery cost reduction. [123]

Safety
As promising candidates for large-scale grid applications, SIBs will suffer from large temperature fluctuation during long-term operation and thus should possess good safety features. Compared to LIBs, SIBs possess better safety features with lower probability of dendrite growth, higher internal resistance, and higher thermal runaway temperature. In addition, SIBs sustain a wider operating temperature range (−40°C to 80°C, with a capacity retention rate of nearly 90% at −20°C). [124] However, similar to LIBs, repeated cycling will also possibly induce dendrite growth in SIBs, leading to short circuits, and finally, thermal runaway of the cells. Overcharging/discharging may result in gas production in the sealed cells, leading to swelling and finally explosion of the cells. The interplay among dendrite growth, short-circuiting, gas evolution, and thermal runaway is detrimental to the health of SIBs and should be carefully monitored using intelligent diagnostic tools. [125,126] Design of durable and safe SIBs should be implemented from electrodes and cells to packs. [21] Although sodium metal is softer than Li, and Na dendrite is less likely to pierce through separator, continuous cycling can still render growth of Na dendrites and short circuiting in SIBs. [127][128][129] Short circuiting can also be induced under circumstances of mechanical failure and separator meltdown. [89] Separator/electrolyte/SEI engineering is the commonly adopted methods to hinder dendrite growth and short circuiting. [129][130][131][132][133] Viscous electrolyte, thicker separator, and stable SEIs can mitigate the dendrite growth and thus avoid short circuits.
Gas evolution in SIBs can be caused by several factors. For example, at elevated temperatures, charged oxide cathode will release oxygen gas. [134,135] Overcharging/discharging will also cause oxidative/reductive electrolyte decomposition with gas production, breaking the CEI/ SEIs, and progressively consuming sodium ions in electrolyte, leading to failure of the device. [89,126,[136][137][138][139][140][141] Electrode materials, for example, PBAs with high content of coordinated water, will cause side reactions with electrolyte and gas evolution. [142] Sodiumcompensating (presodiation) cathode additives, such as NaN 3 , Na 2 CO 3 , and NaNO 2 , can increase the initial coulombic efficiency and cathode capacity, however, might cause gas production upon charging. [143,144] Electrolyte engineering, including composition tuning and incorporation of additives, can ensure improved interfacial stability and safety of SIBs. Non-flammable aqueous electrolytes and solid electrolytes have higher safety. [145] The use of stable sodium salts, highly concentrated electrolytes, and non-flammable or flame-retardant solvents can improve the safety of SIBs. Adoption of solid-state electrolytes, in a way, can largely suppress dendrite growth and thus avoid short circuiting and gas evolution, posing a promising option for constructing safe SIBs. [146]

Commercialization
The main factors currently restricting the commercialization of SIBs are energy density and power density, and the cycle life of SIBs is also lower than that of ternary-based LIBs. The main reason is the larger ionic radii of sodium ions than lithium ions. CATL has now developed a SIB with an energy density of 160 Wh kg −1 , the highest level in the world at present, and the company plans to hit 200 Wh kg −1 (close to the level of conventional lithium iron phosphate batteries) in the future. [124] Faradion has achieved a sodium-ion pouch battery with a charge/discharge rate of 32 AÁh and an energy density of 160 Wh kg −1 by the end of 2020. [121] SIBs are advantageous than LIBs in terms of cost, sustainability, application scope, and safety, and have a broad development prospect. The R&D experience in LIBs will contribute to the rapid development and commercialization of SIBs.

Functional Sodium-Ion Battery
Sodium-ion batteries with comparable electrochemical performance to LIBs and the advantage of cost-effectiveness are deemed promising energy storage systems for grid applications. Nonetheless, integration of multifunctionalities will definitely broaden the application scenarios of SIBs, for example, in wearable, biocompatible, deformable, and sustainable electrochemical and electronic devices. Functional SIBs with deformability, self-healable ability, self-rechargeability, and multifunctionalities have shown intriguing advantages over practical coin cells, cylindrical cells, and pouch cells, yet at the cost of additional intricacy in electrode design and battery assembly.

Flexible Sodium-Ion Battery
Traditional electrode preparation often relies on slurry-casting methods, involving addition of insulating binder and electrochemically inactive Figure 5. Diagram of battery raw material costs. Reproduced with permission. [18] Copyright 2018, Nature Publishing Group.
Energy Environ. Mater. 2023, 6, e12633 12 of 23 carbon additives, resulting in rigid, heavy, and bulky electrodes with compromised gravimetric/volumetric capacity. [147] Such heavy and rigid SIB electrodes cannot meet the requirements of deformability in the era of smart electronics, thus, development of deformable (flexible and stretchable) SIB electrodes and devices is garnering increasing research interest. [147] The assembly of flexible SIBs can be achieved through fabrication of flexible battery components (electrode, [147] electrolyte/separator, [148,149] current collector, [147,150] etc.) and adoption of alternative battery configurations (sandwich type, [151] planar, [87] and fiber shaped [152] ).
Compared to conventional slurry-coated electrodes, self-supporting flexible electrodes needless of binder and carbon additives can reach potentially superior capacity and facilitated kinetics. [147] As discussed in Sections 2.4.3 and 2.4.4, active materials grafted on carbon matrix (carbon felt, carbon fiber, carbon cloth, etc.) or other conductive hosts (e.g., Mxene) can often deliver improved electrochemical performance due to the absence of insulating and inert binders and incorporation of conductive and stable matrix. Interestingly, aside from elevated electrochemical performance, mechanical deformability can also be incorporated in these composite electrodes, allowing preparation of flexible electrodes as will be discussed in this section. Flexible electrodes can be obtained by coupling 1D, 2D, or 3D matrix (e.g., carbon nanofibers, CNTs, graphene, Mxene, and graphene-CNTs) [147,150] with various active materials through different synthesis procedures (filtration, hydrothermal, thermal annealing, electrospinning, etc.), [150] to ensure mechanical flexibility while maintaining reasonable capacity. As shown in Figure 6a-c, 0D red phosphorous nanodots are grown on 2D reduced graphene oxide (rGO) through thermal vapor deposition, rendering flexible P@RGO electrodes that withstand bending radius as small as 2 mm, while attaining a high specific capacity of 914 mAh g −1 after 300 cycles at 1593.9 mA g −1 . [153] 1D Fe 1−x S@porous carbon nanowires are embedded in 2D rGO paper through hydrothermal and sulfurization synthesis, forming self-standing flexible Fe 1−x S@porous carbon nanowires/rGO (Fe 1−x S@PCNWs/rGO) paper electrodes that sustain various mechanical deformations of bending, twisting, folding, and rolling, as shown in Figure 6d-f. The Fe 1−x S@PCNWs/rGO SIB anode shows enhanced gravimetric and volumetric capacity of 565 mAh g −1 and 424 mAh cm −3 (@0.2 A g −1 ) than that of Fe 1−x S@C and pure Fe 1−x S electrodes. [154] 0D nitrogen-doped graphene quantum dots-decorated 2D WS 2 nanosheets anchored on a porous 3D carbon foam were obtained via solvothermal and electrophoresis processes, forming bendable and flexible NGQDs-WS 2 /3DCF electrodes with outstanding sodium-ion storage performance, as displayed in Figure 6g-i. The NGQDs-WS 2 /3DCF electrode manifests specific capacity of 460.9 mAh g −1 at 50 mA g −1 and realizes capacity retention of 97.1% after 1000 cycles. [155] Aside from flexible electrodes, development of flexible electrolyte/separators is also being pursued. A flexible hybrid solid-state electrolyte (HSE) was obtained by immersing Na 3 Zr 2 Si 2 PO 12 (NZSP)/PVDF-HFP-blended membrane in ether electrolytes, harvesting high ionic conductivity and small interfacial resistance while sustaining bending, twisting, and rolling without cracking. The flexible HSE enables assembly of pouch-type flexible hard carbon/HSE/NaFePO 4 SIB full cell that works well under bent conditions. [148] Similarly, a flexible HSE was constructed by transplanting interpenetrating poly(ether-acrylate) network into NZSP/PVDF-HFP, forming flexible composite electrolytes that allow fabrication of NVP and Na 2/3 Ni 1/3 Mn 1/3 Ti 1/3 O 2 cathode-based sodium metal batteries. [156] Ionic liquid gel membranes of Na + -based (C 3 mpyr)(FSI)/P (VdF-co-HFP) ionogel were synthesized as both electrolyte and separator, for the construction of a flexible NVP@C-based sodium metal battery. [149] In addition to constructing flexible components, the assembly of flexible SIB full cells with different configurations has also been attempted. Conventional SIB configuration is normally the "face-toface" sandwich type that allows fabrication of flexible SIB pouch cells. [154] As shown in Figure 7a, Peng et al. reported the assembly of belt-shaped flexible SIBs based on Na 0.44 MnO 2 cathode and NaTi 2 (-PO 4 ) 3 @C anode in aqueous Na 2 SO 4 electrolyte. [151] The flexible SIB delivers outstanding volumetric energy density of 23.8 mWh cm −3 that outperforms many reported flexible LIBs and supercapacitors. After 100 times repeated bending at different bending angles (0°, 45°, 90°, 135°, and 180°), the flexible SIB maintains relatively stable capacity output (from~45 to~35 mAh g −1 ). Similar sandwich configurations have been adopted for assembly of flexible SIB full cells based on the paired NVP/rGO cathode-Fe 1−x S@PCNWs/ rGO, [154] Na 3 (VO) 2 (PO 4 ) 2 F/graphene foam-VO 2 /graphene foam, [158] Na 0.7 CoO 2 /CC-HC/CC, [159] etc.; these flexible SIB pouch cells sustain stable electrochemical performance after repeated folding/unfolding. Another typical configuration for assembly of flexible SIBs is planar type, which is suitable for miniaturization and microelectronics applications. [87] Through interdigitated alignment of electrodes, the ion exchange between cathode and anode can be facilitated. As shown in Figure 7b, Wu, et al. reported the assembly of interdigitated planar flexible SIBs based on NVP cathode and NTP anode using micropatterned exfoliated graphene current collector and NaBF 4 -based ionogel electrolyte. The as-assembled flexible SIB exhibits high volumetric energy density of 30.7 mAh cm −3 as well as excellent mechanical deformability, maintaining unchanged capacity under different bending angles. [87] Deng et al. also reported an interdigitated symmetric SIB based on Na 2 VTi(PO 4 ) 3 electrode and biocompatible simulated body fluid (SBF) electrolyte, which can be bent after encapsulation by silk fibroin (SF) hydrogel. [160] As for wearable electronics applications, 1D fiber-shaped configuration is preferred, [155] including parallel, coaxial, and twisted configurations (as shown in Figure 7c) based on the relative position of cathode and anode. [161] These different configurations have their respective advantages in terms of fabrication process, wearability, mechanical stability, scalability, etc. [152] The flexible NGQDs-WS 2 /3DCF anode has been paired with Na 0.44 MnO 2 /nickel foam cathode to assemble coaxial flexible SIB full cells, lighting up a LED for 9 h even when bent to 120°. [155] As schemed in Figure 7c, Ni-coated cotton textile was employed to graft Prussian blue/GO, which was subsequently paired with sodium metal to fabricate co-axial flexible sodium metal batteries. This coaxial sodium metal battery sustains unchanged capacity under different bending angles and repeated bending cycles, and can be curved into a necklace device. [157] Similarly, NiS 2 nanoparticles embedded in porous carbon fiber were paired with Na to assemble coaxial flexible sodium metal battery, which only loses~11% of the capacity when bent from 0°(~570 mAh g −1 ) to 150°(~510 mAh g −1 ). [162] Another type of twisted fiber-shaped flexible SIBs was also showcased with CNT/NMO cathode and CNT/MoO 3 /PPy (polypyrrole) anode, which was further injected into the subcutis of a mouse, manifesting specific capacity of 43.05 mAh g −1 (at 1000 mA g −1 ) and with capacity retention of 72.3% after 100 cycles of in vivo charging/ discharging. [163] Albeit the developments of flexible SIB electrodes and devices achieved as discussed above, there are still several issues to be considered. There are no standards to evaluate mechanical deformability, for Energy Environ. Mater. 2023, 6, e12633 13 of 23 example, the bending radius or bending angle for flexible electrodes/ devices with different sizes/shapes. Safety issues brought by the strong acid/alkaline or toxic organic electrolyte used in flexible SIBs should be taken note of, especially for wearable applications. [151] The electrochemical performance of flexible electrodes/devices under deformation is also to be further optimized. The weight content of electrochemically active materials might be compromised in order to achieve flexibility, for example, the weight content of WS 2 (57.7%) in NGQDs-WS 2 / 3DCF is only slightly above 50%. [155] In addition, the overall performance of flexible SIB full-cell devices is still to be optimized, considering mechanical deformability, energy density, wearability, scalability, safety, etc. Figure 6. Preparation of flexible electrodes. a) Schematics, b) microstructure, and c) photograph of flexible red phosphorus nanodots/rGO electrodes, Reproduced with permission. [153] Copyright 2017, American Chemical Society. d) Schematics, e) microstructure, and f) photograph of flexible Fe 1−x S@porous carbon nanowires/rGO electrode. Reproduced with permission. [154] Copyright 2019, John Wiley & Sons. g) Schematics, h) microstructure, and i) photograph of flexible NGQDs-WS 2 /3DCF (nitrogen-doped graphene quantum dots-decorated WS 2 nanosheets anchored on a porous 3D carbon foam) electrodes. Reproduced with permission. [155] Copyright 2018, Royal Society of Chemistry.

Stretchable Sodium-Ion Batteries
Stretchable energy storage devices have received enormous attention due to their potential for wearable applications, in spite of the critical challenges including energy density, integration, delamination, scalability, packaging, etc. [164,165] Nonetheless, exploration of stretchable energy storage devices has been attempted and many achievements have been made. There are a few excellent reviews summarizing the achievements made in stretchable energy storage devices from different aspects, for example, materials, [166][167][168] structures, [169,170] and strategies. [164,165,171] The strategies to fabricate stretchable energy storage devices have been categorized into component level (buckling, 2D/3D porosity, rigid islands, and intrinsic stretchability) and device level (wavy, folding, fiber-like, and intrinsically stretchable structures). [164,165] These strategies are widely adopted for fabricating stretchable LIBs and supercapacitors. Stretchable SIBs and sodium-ion capacitors have also been assembled adopting these strategies. As shown in Figure 8a, Yu et al. prepared an rGO-modified 3D porous poly (dimethylsiloxane) (PDMS) sponge as the stretchable and conductive current collector to load hard carbon anode and VOPO 4 nanosheets cathode, sandwiching the stretchable electrolyte/separator of P(VDF-HFP) membrane soaking NaClO 4 /PC-FEC, followed by PDMS encapsulation. The as-assembled stretchable SIBs maintain capacity of 92 mAh g −1 when stretched with 50% strain and retain 89% of the initial capacity after 100 times repeated strain (50%)-release cycles, demonstrating excellent stretchability. [172] Stretchable sodium-ion capacitors are discussed here due to the limited reports of stretchable SIBs. The intrinsically stretchable acrylic rubber matrix (ARM) mixed with carbon black and Ag nanowires are prepared as the stretchable and conductive current collector to load CNTs supported Na 2 Ti 3 O 7 anode and AC cathode, sandwiching the gel electrolyte of ARM-bacterial cellulose membrane soaking NaPF 6 /EC-DMC-EMC electrolyte. The as-assembled SICs show almost unchanged capacity output when stretched with 100% strain and retain 89% of the initial capacity after 500 strain (100%)-release cycles. [176] Electrospun 3D porous polyurethane (PU) mat was pre-strained to 30% to spray coat Ag nanowires, further coating of MoSe 2 /MXene and AC/ MXene results in stretchable anode and cathode, respectively. Stretchable SICs are then obtained by sandwiching quasi-solid electrolyte of 3D PU soaking NaPF 6 -DME, which shows a relatively limited capacity retention of 58% at 30% strain. [177] Strategies including 3D porous structure, pre-straining, and intrinsic stretchability have been adopted for assembly of stretchable SIB/SICs, other methods, for example, rigidisland structure, are to be further explored in stretchable SIBs.

Self-Healable Sodium-Ion Battery
Sodium-ion battery electrodes often suffer from volume changes upon sodiation. For alloy-based electrodes, the theoretical volume change can be up to 300-400%, leading to severe electrode pulverization and significant capacity degradation. [6] Through nano/microstructure design, [173] combination of conversion and alloying reaction, [178,179] and adoption of proper binder, [174] the mechanical damages caused by volume change can be "self-healed," thus ensuring prolonged cycling stability in SIB electrodes. 3D continuous bulk porous bismuth (3DPBi) was obtained through liquid phase reduction. As shown in Figure 8b, the 3D interconnected bi-nanoligaments ensure electronic conduction and "self-heal" the volume changes during sodiation/desodiation, rendering outstanding rate performance and cycling stability of 3DPBi anode. [173] A combination of conversion and alloy reactions in electrodes can also induce improved cycling stability. The notably higher cycling stability of Sn 4 P 3 electrode than SnO and Sn can be attributed to the combined reversible conversion and alloying reaction, during which the conversion-produced intermediate Na 3 P can heal the cracks formed by Sn alloying/dealloying. [178,179] Liu et al. quantified the ultralow formation energy of −0.19 eV to restore the original layered structure of GeP, thus allowing self-healable SIB anode that manifests outstanding cycling stability. [180] The deformable and self-healing liquid metal can intrinsically buffer the volume changes of metal electrodes.  [151] Copyright 2017, Cell Press. Elsevier. b) Planar flexible SIBs assembled with interdigitated NVP cathode and NTP anode on exfoliated graphene substrates. Reproduced with permission. [89] Copyright 2020, Royal Society of Chemistry. c) Fiber-shaped flexible SIBs. ① Parallel NMO-CNT//Na 2 SO 4 //NTP-CNT fibershaped SIBs. Reproduced with permission. [151] Copyright 2017, Cell Press. Elsevier. ② Co-axial flexible sodium metal battery with PB@GO@NCT cathode. Reproduced with permission. [157] Copyright 2017, Fusible multielement alloy (Bi-Sn-In) can lower the melting temperature to 100°C through disordered arrangement of atoms and redistributed charge density, [181] while the in situ formed NaK alloy through galvanic replacement reaction maintains liquid state at room temperature. [182] Pasini et al. reported a self-healing binder composed of ureidopyrimidinone(UPy)-telechelic-poly(ethylene oxide) blends. Through dynamic quadruple hydrogen bonding, UPy-PEO helps to repair the mechanical damages of black phosphorous anode during sodiation, enabling impressively improved capacity retention compared to traditional PAA-CMC binder. [183] Similarly, as shown in Figure 8c, Na-alginate (SA)-based binder also enables prolonged cycling stability in SnS anode than PVdF and CMC, due to the abundant "self-healing" carboxylic groups and polar sites that withstand the volume changes in SnS, without formation of cracks at electrode surfaces. [174] During sodiation of oxygen-rich sodium rhodizonate dibasic (SRD), hydroxylrich SA binder can fill the cracks formed by volume expansion and binds the pulverized SRD together by hydrogen bonding. SRD electrode shows obviously higher cycling stability with hydroxyl-rich binders than with hydroxyl-free PVDF and polytetrafluoroethylene (PTFE) binders. [184] Construction of self-healable SIB electrodes has been achieved as discussed above, yet a self-healable SIB full cell is seldom reported, probably due to the difficulties in ensuring electronic/ionic conduction within electrode/electrolyte while avoiding short circuit.

Self-Chargeable Sodium-Ion Battery
Sodium-ion batteries as promising energy storage devices, when applied as power sources for other wearable and flexible electronics, still require frequent charging. As such, selfchargeable SIBs that harvest energy from ambient environments (mechanical, thermal, solar energy, etc.) are more sustainable with higher energy efficiency. [7,185,186] Self-chargeable SIBs have been demonstrated with incorporation of piezoelectric nanogenerator (PENG), [175,187] triboelectric nanogenerator (TENG), [188] solar energy harvesting, [185,189] etc. As shown in Figure 8d, BaTiO 3 nanoparticles are mixed with P(VDF-HFP), forming piezoelectric BaTiO 3 -P (VDF-HFP)-NaClO 4 gel-electrolyte. With NVP@C cathode and hard carbon anode, a flexible piezoelectric SIB is thus assembled (Figure 8e), which can be charged through static compression, repeated bending, and palm patting. The piezoelectric SIB can be charged from 0.09 to 0.31 V after 100 h static compression at 5 N as displayed in Figure 8f. [175] Similarly, piezoelectric potassium sodium niobite (KNN) particles are mixed with styreneethylene-butylene-styrene (SEBS) to form an elastic separator, enabling assembly of flexible piezo-SIBs with NVP@C cathode and Ni 2 P@Ndoped carbon network anode. The piezo-SIBs are chargeable via repeated bending and palm patting. The SIBs can be charged to~0.15, 0.22, and 0.32 V with 3600 s of static compression at 1.6, 3.2. and 6.4 N, respectively. [187] Sun et al. reported a Na 0. 67 Ni 0.23 Mg 0.1 Mn 0.67 O 2 cathode and solid polymer electrolyte-based Na metal battery, which can be charged by rotary TENG. After being charged by TENG rotating at 800 rpm, energy conversion efficiency of 62.3% was achieved when the SIBs are discharged at 5 mA with capacity of~60 mAh g −1 . [188] Solar energy harvesting was also realized by incorporating TiO 2 photoelectrode into sodium polysulfide/iodide battery (SPIB), with S 2À =S 2À 4 anolyte and I − /I 3− catholyte. The charging potential of the SPIB can be greatly reduced to 0.08 V due to the narrower potential difference between TiO 2 and S 2À =S 2À 4 (0.07 V) than that between S 2À =S 2À 4 and I − /I 3− (~1.0 V), saving~90% of input electric energy. After 4 h of photo-assisted charging, the SPIB can deliver capacity of 110 mAh g −1 . [185] Except for the above self-chargeable SIBs that harvest mechanical and solar energy, thermoelectric devices that transform thermal energy into electricity through Seebeck effect might also be of interest to self-chargeable SIBs.

Sodium-Ion Battery Sensors
Battery materials, for example, electrodes with electronic conductivity, abundant defects, porous structure, and unique physical/chemical properties, are also favorable for sensing applications. [190] There are several excellent reviews summarizing the application of battery materials for both energy storage and sensing, including 2D Reproduced with permission. [173] Copyright 2021, John Wiley & Sons. c) Enhanced cycling stability of SnS electrode by self-healing binder. Reproduced with permission. [174] Copyright 2020, John Wiley & Sons. d) Structure, e) photograph, and f) the charge/discharge profile of flexible piezo-chargeable SIBs. Reproduced with permission. [175] Copyright 2020, Royal Society of Chemistry.
Energy Environ. Mater. 2023, 6, e12633 16 of 23 materials, [190,191] 2D black phosphorous, [192,193] MoS 2 , [194] carbon materials, [195] ionic liquids, [196] etc. SIB electrodes/electrolytes are also applicable for sensing. For example, sodium-incorporated alumina can be used both for SIB solid-state electrolyte and gas sensing, [197] Na 2 Ti 7 O 15 nanowires for SIB anode and gas sensing, [198] NiFe 2 O 4 nanoparticles for SIB anode and biomolecule sensing (uric acid), [199] and CoS 2 /rGO for SIB anode and glucose sensing. [200] Electrospun porous PVDF/PAN membrane can function both as a piezoelectric pressure sensor for human motion detection and a separator for SIBs with CE of 98.9%. [201] Instead of deploying SIB components for sensor applications, on the other hand, optic fiber sensors and pressure sensors have been integrated into SIBs to monitor the internal thermal and pressure behaviors. Microcantilever electrodes were used to monitor the interfacial stress of a thin-film MoS 2 electrode during sodiation, revealing staged behaviors (compressive stress of 2.1 N m −1 from 1.0 to 0.85 V, 9.8 N m −1 from 0.85 to 0.4 V, and 43 N m −1 when discharged to <0.1 V). [202] As shown in Figure 9a, a fiber Bragg grating (FBG) sensor was installed in the center voids of an 18650 SIB cell of NVPF//hard carbon electrodes.
The reflected peak wavelength shift in FBG is dependent on the local thermal, strain, and pressure environment. As a result, the thermal events and pressure variation within the SIB are decoded by employing an FBG sensor written in a microstructured optical fiber (MOF-FBG), enabling the tracking of SEI formation and cell lifetime. [203] Fiber optic sensors have also been integrated into SIBs to study the effect of electrolytes on SEI/CEI formation, [207][208][209] chemical evolution of electrolytes, [209,210] sodium plating behaviors, [211] electrode structure variation, [209] etc., showing great promises of this optic fiber probes for in situ health monitoring in SIBs.

Sodium-Ion Battery Actuator
Actuators capable of deformation upon energy input (e.g., thermal, electrical, and chemical) have found applications in drug delivery, soft robotics, automobiles, artificial muscles, etc. [204,212] Assembly of highperformance actuator relies on a trade-off between the deformability and stiffness of electrode materials. [213] In metal-ion batteries, ion insertion into electrodes often causes geometry variation and generates strain in electrode materials. [212,213] Electrochemical actuators based on Li + insertion have been demonstrated in alloy-based electrode materials, including Ge/PI/Ge thin films [214] and Si nanowires. [215] The volume changes of Si in LiFePO 4 /Si full cell are also associated with voltage-dependent Li insertion in the Si anode, producing mechanical stress up to 10 MPa. [213] Similarly, sodium-ion insertion also enables construction of electrochemical actuators. Zhang et al. synthesized a hybrid MnO 2 with abundant cation vacancies and high density of Na + -accessible lattice tunnels, forming a hybrid MnO 2 /Ni electrode. As shown in Figure 9b, in three-electrode testing in aqueous Na 2 SO 4 , the actuator electrode (fabricated at pH 1.3) generated 5.3% intrinsic strain in 6 s with a rapid strain response of 0.88% s −1 and a large actuation stress of 244 MPa, while delivering capacitance of 272 mF cm −2 at 10 mV s −1 . [204] The axial expansion/ shrinkage of carbon fiber during sodiation/desodiation shows a voltage-strain coupling, reaching maximum axial strain of 0.16% while reaching capacity of 120 mAh g −1 . [216] Adoption of other electrode materials (e.g., alloybased P and Sn) for sodiation-induced electrochemical actuation should be further explored.

Electrochromic Sodium-Ion Battery
Similar to metal-ion batteries, the coloration/ bleaching of electrochromic devices is mostly based on ion insertion/extraction, rendering dual-functional electrochromic energy storage devices. [217,218] Although most of the insertionbased electrochromic devices employ H + /Li +based electrolytes, [205,217,219] Na + -based Figure 9. a) Illustration of an optical fiber integrated into a jelly-roll-type battery. Reproduced with permission. [203] Copyright 2020, Nature Publishing Group. b) The structure of a MnO 2 /Ni electrochemical actuator electrode and the optical images under 0 V actuation. Reproduced with permission. [204] Copyright 2022, American Chemical Society. c) The color changing of a Ni-CHNDIbased EC display in Na + electrolyte at different voltages. Reproduced with permission. [205] Copyright 2018, American Chemical Society. d) Sodiation-induced electrochromism in carbon nanofoam paper (scan rate of 0.5 mV s −1 ). Reproduced with permission. [206] Copyright 2022, IOP Publishing Limited.
Energy Environ. Mater. 2023, 6, e12633 electrolytes have been adopted for electrochromic WOx, [220,221] V 2 O 5 , [222] Prussian white, [223] carbon nanofoam, [206] Ni-MOF, [205] etc. Albeit the fact that the ionic radii of Na + are larger than Li + , V 2 O 5 shows faster reduction response in NaClO 4 /PC than in LiClO 4 /PC, due to the higher ionic conductivity of NaClO 4 /PC (2.2 × 10 −8 S cm −1 ). [224] WO 3 manifests poor stability in aqueous Na 2 SO 4 (0.1 M) electrolytes due to progressive dissolution, which can be effectively mitigated by a NASICON (Na 1+x Zr 2 Si x P 3−x O 12 ) capping layer. [221] An atomical layer-deposited protective Al 2 O 3 coating was also introduced to enhance the stability of monoclinic WO 2.72 nanorods in PC electrolytes, and higher coloration efficiency is achieved in Na + instead of Li + electrolytes, due to the selective occupation of optically active hexagonal tunnel sites through a capacitive-dominant process. [220] The coloration efficiency and capacity of hexagonal Cs-doped WO 3 in NaClO 4 /PC are found to be shape dependent. Cs-WO 3 nanorods deliver lower capacity (19.6 mC/10 17 nm 3 ) and higher coloration efficiency (212 cm 2 C −1 ) than Cs-WO 3 nanoplatelets (25 mC/10 17 nm 3 , 156 cm 2 C −1 ) due to less amount of accessible hexagonal sites caused by Cs + blocking. [225] Two types of 1D Ni-MOF were constructed with different pores sizes, namely Ni-CHNDI with hexagonal channel (diameter of~33Å, pore volume of 2077 m 2 g −1 ) and Ni-BINDI with square-shaped channels (diameter of~10Å, pore volume of 556 m 2 g −1 ). For both NI-CHNDI and NI-BINDI, higher electrochemical performance (coloration efficiency and switching time) is obtained in NaClO 4 /PC, than in TBA-ClO 4 /PC, LiClO 4 /PC, and Al(ClO 4 ) 3 /PC electrolyte. Ni-CHNDI EC films manifest higher coloration efficiency and faster switching time than Ni-BINDI in NaClO 4 /PC due to the larger channel size. As shown in Figure 9c, a Na + -based EC display was thus constructed, showing multicolor according to different voltages applied. [205] Aside from traditional insertion-based electrochromic materials, carbon nanofoam paper (CNFP) also shows electrochromic phenomenon along with sodiation. As displayed in Figure 9d, the CNFP electrode demonstrates reversible black-blue-red/gold color change when discharged, which is correlated with electronic state changes in CNFP caused by deep sodiation (<0.5 V vs Na/Na + ). [206] Based on the above discussion, it is found that although the ionic radii of Na + are larger than Li + , Na + still enables favorable electrochromic performance in V 2 O 5 , [222] WO 3 nanoplates, [225] Ni-CHNDI, [205] etc. However, most of the above-discussed Na + -based electrochromic materials employ three-electrode testing, lacking discussion of SIB electrochromic dual-functional performance at device level.

Multifunctional SIBs
The above discussions of classified functional SIBs have exemplified the great promises of functional SIBs that are flexible, stretchable, selfhealable, self-chargeable, and capable of sensing, actuating, and color changing (electrochromic). The functionalities incorporated into SIBs definitely empower the applications of SIBs to wearables, smart electronics, display, robotics, etc. other than grid energy storage. Higher level of integration of multifunctional SIBs is also being explored to satisfy the ever-increasing requirements of smart electronics and bionics. For example, flexible and self-chargeable SIBs have been demonstrated via incorporation of piezoelectric separators. [175,187] Self-chargeable and electrochromic SPIBs are assembled with incorporation of TiO 2 photoelectrode into color-changing S 2À =S 2À 4 anolyte and I − /I 3− catholyte, yet the electrochromic phenomenon in this SPIB is not well explored. [185] Wang et al. pointed out the safety issue in conventional acid/alkaline/flammable toxic organic electrolytes and attempted cell culture medium (Dulbecco's modified Eagle medium [DMEM]) as the electrolyte for assembly of flexible SIBs for possible biocompatible applications. Similar capacity outputs are harvested in DMEM and normal saline electrolyte for the flexible SIBs (sandwich-type and fibershaped [parallel]). [151] Deng, et al. adopted SBF as the electrolyte for the flexible micro-SIBs with symmetric interdigitated NVTP electrodes. Such a micro-SIB was subsequently implanted into the dorsal subcutaneous region of the SD rats. After 1 month in vivo charging/discharging (once/day), the implanted micro-SIB shows stable capacity output, certifying excellent biocompatibility. [160] Peng et al. reported a twisted fiber-shaped SIB based on CNT/NMO cathode and CNT/ MoO 3 /PPy anode, which is further injected into the subcutis of a mouse. The biocompatible, flexible SIB delivers specific capacity of 43.05 mAh g −1 (at 1000 mA g −1 ) in vivo, powering an implanted sensor for respiration monitoring. [163]

Future Perspectives
As one of the most promising energy storage devices that implement LIBs, SIBs are pivotal among various developing energy storage systems. With years of research and development, SIB system with energy output capacity of 30 kW/100 kWh (HiNa) has been installed for grid applications. [226] Companies including Faradion, AGM batteries, Natron Energy, CATL, etc. are also working toward SIB commercialization. [227] The available strategies for optimization of battery components (cathode, anode, electrolyte, separator, binder, current collector, etc.) are classified, and functional (flexible, stretchable, self-healable, and selfchargeable) and integrated sodium-ion batteries (−actuator, −sensors, electrochromic, etc.) have been exemplified. Given these up-to-date achievements, there are still several challenges and issues for future development of SIBs.

Energy Density
Theoretically, the energy density of SIBs is not as competitive as LIBs, thus SIBs are more suitable for applications where energy density is not of foremost importance, for example, stationary grid applications. Nonetheless, higher gravimetric/volumetric energy density of SIBs can open the market of portable electronics and electric vehicles with a lighter and smaller package. The development of Si-incorporated anode has elevated the energy density of LIBs, and thus alloyed-based anode, for example, P, might be of interest for developing high-energy-density SIBs. [18] It should also be noted that electrochemical performance optimization of battery components and functionality integration might affect the battery energy density output. For example, flexible electrodes necessarily contain higher content of low-capacity carbon materials, composite electrodes with high surface area and porosity might reduce the tap density, and both will compromise the gravimetric/volumetric energy density. Thus, energy density should always be one of the key metrics for SIB assembly.

Mechanism Understanding
It is well understood that the charge/discharge of SIBs is based on Na + shuttling between cathode and anode. However, the accompanied physical/chemical/structural changes in the electrodes, electrolyte, SEIs, Energy Environ. Mater. 2023, 6, e12633

of 23
and separators are quite complex and require further understanding. In situ and operando characterization techniques, with advantages of accuracy and resolution over ex situ methods, are powerful tools in unveiling electrochemical mechanisms in SIBs. [6,7] Various in situ spectroscopic and microscopic characterization techniques, including Xray-based techniques (X-ray diffraction, X-ray absorption spectroscopy, and X-ray photoelectron spectroscopy), Raman/Infrared spectra, transmission electron microscopy, electrochemical quartz crystal microbalance, etc., can be adopted to detect the evolution of crystal structure, valence state, coordination environment, geometry, and weight evolution of electrodes/electrolytes. However, the suitability of respective characterization techniques for specific samples should be considered. For example, in situ XRD is more suitable for crystalline rather than amorphous materials, NMR is limited for elements with non-zero spin nuclei, etc. [6,7,228] It is also suggested to combine theoretical calculations with in situ characterizations to provide a more comprehensive understanding of electrochemical mechanisms in SIBs both theoretically and experimentally.

Functionality
Aside from the above showcased functional and integrated SIBs. There are several other intriguing functionalities to be further explored in SIBs. As mentioned above, thermoelectric-integrated self-chargeable SIBs harvesting thermal energy can be evaluated. Pollution-free biodegradable SIBs, and implantable and biocompatible SIBs are also interesting for bionics. Yet it should be noted that the assembly of functional and integrated SIBs normally required extra processes than conventional SIBs, which will definitely affect the electrochemical performance, cost, and scalability. There should be a trade-off between functionality and electrochemical performance when designing such novel SIBs.

Scalability
Development of novel and advanced SIBs with fascinating functionalities and improved electrochemical performance is occasionally reported in literature. Lab-based research is aimed at performance optimization and often neglects scalability. The materials preparation process, device assembly, packaging procedures, etc. can be tedious and complicated. However, for practical applications, preparation of SIBs has to be repeatable and scalable. Methods for electrode preparation, including hydrothermal reactions, electrochemical deposition, thermal annealing, etc., are adoptable for scalable productions. Nonetheless, manual sealing and packaging for device assembly should be replaced by more automated manner. Technology transfer from laboratory to industry is also the key step to embodying research outputs into commercial products.

Cost-Effectiveness
The major advantage of SIBs over LIBs is the cost-effectiveness brought by abundance and uniform distribution of Na sources. The costly Li resources and transition metal-containing compounds, for example, Ni and Co, are the major contributors to the high cost of LIBs. Adoption of Ni/Co-containing electrodes might sacrifice the cost-effectiveness of SIBs and thus should be avoided. Currently, cost-effective SIBs are appreciated for stationary applications. It is predicted that with development of P-incorporated hard carbon anode and future cathode, the cost of SIBs can be lower than that of LIBs with the same capacity, and will fill in the markets of household appliances and electric vehicles. [18] On the other hand, assembly of functional SIBs requires additional materials/components to be incorporated. For example, some of the flexible SIBs incorporated excessive graphene, CNTs, or Mxene flakes instead of normal carbon blacks to ensure flexibility of electrodes/devices, resulting in inevitably increased cost of SIBs. Cost analysis for SIBs aiming at different applications might be needed to provide better guidance for SIB development.
Challenges still pertain before proliferation of commercial SIBs. With scalable production, cost-effective, high-energy density, and functional SIBs will solve the supply risks of LIBs, and finally fit into the energy storage market to supplement LIBs.