Synthesis and regulation strategies for enhancing the electrochemical performance of sodium-ion battery anode materials

HC

Pei et al. devised a novel protocol for in-situ texture creation utilizing sulfuric acid polymerized sucrose and trithiocyanuric acid as starting materials to fabricate N and S double-doped porous HC (SNC-P) material with increased layer spacing, appropriate doping and defect concentrations, thin carbon shells, and pore classification^[41]. This material exhibited a notable reversible specific capacity of 430 and 277 mAh·g^-1 at 0.05 and 5.0 A·g^-1, along with exceptional cyclic stability for Na⁺ storage. Jin et al. synthesized N and S co-doped micro/nanoparticles as an anode material for SIB through a one-step approach involving thiourea and sodium citrate grounded on the pseudo-capacitance mechanism^[42]. Upon fine-tuning, the specific capacity of NSC2 (thiourea to sodium citrate mass ratio of 2:1) was recorded as 280 and 223 mAh·g^-1 at 0.05 and 1.0 A·g^-1 after 200 and 2,000 cycles, respectively. Xie et al. implemented a successive carbonization process on the phosphate-treated petroleum residue to expunge the irreversible configuration of heteroatoms and delineate the screening strategy for heteroatom configuration shown in Figure 2A^[43]. They retained the reversible heteroatom configuration and harmonized with free radicals, and the full battery achieved a consistently high slope capacity (250 mAh·g^-1) and ICE (80%) shown in Figure 2B. In the carbonization step following P/O doping, it was noted that the configuration of heteroatoms outweighs the amount of doping. Wu et al. synthesized N and P co-doped ultramicropores HC spheres (NPUCS) using polyvinyl alcohol (PVA) and sodium carboxymethyl cellulose as raw materials shown in Figure 2C^[44]. The increase in the ratio of active pyridine-N had been found to enhance the electrode reactivity and Na⁺ diffusion kinetics while reducing irreversible depletion. The NPUCS exhibited high specific capacity (257.7 and 157.0 mAh·g^-1 at 0.1 and 5.0 A·g^-1) shown in Figure 2D, extraordinary ICE (75%) and stable Na storage performance (over 2,000 cycles).

Figure 2. (A) Schematic illustration of the heteroatom configuration screening strategy for the synthesis of PO-SC-S as an anode material of SIBs and its corresponding (B) Rate performance of full cells coupled with an O3- or P2/O3-type cathode with C-rates from 0.1 to 20 C. Quoted with permission from Xie et al.^[43]; (C) Synthesis scheme of NPUCS and NUCS and their corresponding (D) Rate capability. Quoted with permission from Wu et al.^[44]; (E) Schematic illustration of the synthesis process of UPCs; (F) Rate capability of UPC-6. Quoted with permission from Yan et al.^[47]. PO-SC-S: The pristine SC was co-doped with P/O and subsequently carbonized on two occasions at 800 °C; SIBs: sodium-ion batteries; NPUCS: N and P co-doped ultramicropores HC spheres; NUCS: N-doped ultra-micropores carbon spheres; UPCs: UPC-6: phosphorus-doped carbon with molar ratios n (PCl₃) : n (C₆H₁₂) = 6:1; SC: soft carbon.

Mehmood et al. proposed a highly nanoporous (NP) carbon material doped with nitrogen created through the ionic carbonization of the Zn-imidazolium framework^[45]. This material demonstrated a consistent cycling capacity of 496 and 280 mAh·g^-1 at 0.03 and 5.0 A·g^-1. Tao et al. produced mesoporous HC materials with N and P co-doping that were synthesized from soybean roots (NPDC)^[46]. The material has abundant active sites, increased layer spacing, and a modified electronic structure. The pyrolysis of NPDC (NPDC-700) at 700 °C presented excellent rate capability (150 mAh·g^-1 at 5.0 A·g^-1) and exceptional cycle stability (197 mA·g^-1 at 2.0 A·g^-1 after 2,000 cycles). Yan et al. introduced an innovative synthesis technique for the fabrication of ultrahigh P-doped carbon (UPC) as an extraordinary-performance anode material for SIB^[47]. By utilizing PCl₃ and cyclohexane (C₆H₁₂) as the sources of P and C, in-situ P-doping was achieved during carbonization under N₂ flow shown in Figure 2E, resulting in a phosphorus content up to 30 wt%. Moreover, the phosphorus incorporated in the carbon structure was mainly replaced by protrusions of P bonded to three C atoms [P-(C₃)], which can remarkably enhance the interlamellar spacing and improve the sorptive ability of Na⁺. UPC-6 delivered a reversible capacity of 510.4 mAh·g^-1 at 0.1 A·g^-1 and an optimal rate capacity of 397.1 mAh·g^-1 at 10 A·g^-1 shown in Figure 2F.

Titanium-based oxides

Fan et al. used the N and S double doping method to achieve highly doped anatase-type titanium (NS-TiO₂) dioxide hollow spheres illustrated in Figure 3A^[48]. The N doping reduced the band gap width of TiO₂ while the S doping promoted the spread of Na⁺ in TiO₂. Therefore, TiO₂ co-doped with N and S exhibited significant electron conductance and accelerated Na⁺ transfer kinetics. The tested material exhibited excellent rate performance and specific capacities of 249.3 and 156.4 mAh·g^-1 at 0.1675 and 5.025 A·g^-1, respectively. Additionally, it showed remarkable cycle stability with capacity retention of 90.5% at 3.35 A·g^-1 for 2,400 cycles illustrated in Figure 3B.

Figure 3. (A) Synthesis routes for NS-TiO₂; (B) Long-term cycle performance of TiO₂ and NS-TiO₂. Quoted with permission from Fan et al.^[48]; (C) Synthesis routes for different TiO₂ spheres, and their corresponding (D) Initial charge/discharge profiles and (E) Long-term cycling performance. Quoted with permission from Xu et al.^[54]; (F) Synthetic routes for C-NTO and C-NTC microspheres; (G) Cycling stability and (H) Rate performance of C-NTO. Quoted with permission from Meng et al.^[55]; TiO₂-HS: TiO₂ hollow spheres formed by smaller nanosheets; TiO₂-YB: TiO₂ yolk-like spheres with bigger nanosheets; TiO₂-S: amorphous TiO₂ solid spheres; TTIP: tetra-n-butyl titanate; NS-TiO₂: N and S double doping method to achieve highly doped anatase-type titanium; NTO: Na₂Ti₃O₇; NTC: titanate/titanium.

Studies have shown that incorporating porous carbon materials into TiO₂ to construct TiO₂@C composite material is conducive to enhancing the electronic/ionic conductivity of the composite material and is the most common and effective method to increase the electrochemical performance of titanium-based oxide anode materials. Luo et al. developed an anode material composed of TiO₂ nanoparticles anchored on the shear carbon nanotube backbone (TiO₂/SCNT), which demonstrated distinguished rate capability and cycling stability^[49]. The TiO₂/SCNT electrode showed a specific capacity of 267 and 136 mAh·g^-1 at 1.0 and 10 C and preserved 87% of the initial capacity for 1,000 cycles. Wang et al. outlined a technique for preparing carbon fiber clothing-supported rutile TiO₂ hollow cube array composites (H-TiO₂@CFC)^[50]. This compound material served as a self-sustaining and binder-free anode for Na⁺ storage, providing a reversible specific capacity of 287.3 and 103.3 mAh·g^-1 at 0.1 and 50.0 C due to rapid ion dispersion and remarkable reactivity with Na⁺.

The kinetics of Na⁺ spread and its storage capacity within TiO₂ can be improved through the combination of heteroatom doping and composite material constructed with carbon. Lv et al. synthesized N/F co-doped TiO₂/carbon microspheres (NF-TiO₂/C) through a simple hydrothermal approach and high-temperature doping reaction^[51]. The co-doping of N and F increased the content of oxygen vacancy (VO), narrowed the band gap between TiO₂ and C, and improved the conductivity of NF-TiO₂/C. Additionally, NF-TiO₂/C showed significant Na⁺ bond energy and a low Na⁺ diffusion barrier. Hence, NF-TiO₂/C provided a marked contact density (1.51 g·cm^-3), remarkable capacity (~242 mAh·g^-1 at 1.0 C), excellent rate performance (190 and 125.9 mAh·g^-1 at 10 and 100 C, respectively), high area capacity (~4.8 mAh·cm^-2) and very lasting cycle performance (~80.2% at 10 C after 10,000 cycles). Wang et al. synthesized coadjacent micro-sheets composed of carbon nanotubes (CNTs) and S-doped TiO₂ (CNT/S-TiO₂) through ultrasonic technology followed by dead burn^[52]. The use of single-wall CNTs (SWCNT) overcame the instability of electrode structure during the preparation of thick electrodes. CNT/S-TiO₂ showed an outstanding rate performance (0.3 to 15 C). Furthermore, the material exhibited long-cycle stability of enduring 1,000 cycles under a load of 1.96 mg·cm^-2.

Introducing different components into titanium-based materials to form heterogeneous structures can significantly enhance the conductivity of TiO₂. Guan et al. devised heterogeneous structures consisting of two-dimensional (2D) nanosheets TiO₂@TiOF₂ by hydrothermal method^[53]. Among them, TiO₂@TiOF₂-30 h maintained a reversible capacity of 151.7 and 101.2 mAh·g^-1 at 0.5 and 5.0 A·g^-1 after 2,000 and 10,000 cycles. Density functional theory (DFT) simulations further showed that the hetero-structured TiO₂@TiOF₂ nanosheets had better electrical conductivity and lower Na⁺ formation energy than separated TiO₂ and TiOF₂. Xu et al. developed a program to control the formation of hetero-structured TiO₂ spheres with an adjustable interior and shell shown in Figure 3C, which enhanced the pseudo-capacitive response and Na⁺ storage dynamics and performance^[54]. The rational design of nanosheets reduced the layered structure and generated excess space, thereby increasing the packing density of TiO₂ spheres. Among the three TiO₂ spheres, TiO₂-HS had the highest surface area and the most exposed active surface and showed the most favorable electrochemical performance. The TiO₂-HS anode showed a significantly higher voltage plateau and lower polarization during the alkalization process shown in Figure 3D. At 5.0 A·g^-1 after 4,000 cycles, the TiO₂-HS electrode was still able to maintain a specific capacity of 119 mAh·g^-1 shown in Figure 3E. Meng et al. designed a sodium titanate/titanium dioxide/C (C-NTC) hetero-structural compound with VO shown in Figure 3F^[55]. C-NTC electrodes achieved the optimal cycling performance with 172 mAh·g^-1 specific capacity after 2,000 cycles at 0.1 A·g^-1 shown in Figure 3G. At 0.05 A·g^-1, the specific capacity remained constant at 181.1 mAh·g^-1, with capacity retention reaching 90.6% after 120 cycles shown in Figure 3H.

Conversion-type materials

Compositing TMSs and TMOs with carbon materials to construct composites can effectively prevent the aggregation of metal nanoparticles, improve the interconnection between the nanoparticles, and enhance the electrical conductivity of the electrode materials, which leads to improved specific capacity, rate capability, and cycling stability. The formation of bimetallic or trimetallic composites can take advantage of the synergistic effect of heterogeneous metals and achieve the enhancement of the electrochemical performance of electrode materials.

Zhao et al. prepared ultrafine TMO nanoparticles encapsulated in nitrogen-doped carbon nanofibers (uf-TMOs@N-CNFs), in which polyvinylpyrrolidone (PVP) was utilized as a highly effective dispersant for inhibiting metal agglomeration and controlling the size of TMOs^[56]. The uf-Fe₃O₄@N-CNFs offered a capacity of 234.5 mAh·g^-1 at 1.0 A·g^-1 with a contribution of 78.69% from pseudo-capacitor control. They also exhibited remarkable rate performance and durable cycle stability while retaining a capacity of 119.4 mAh·g^-1 after 5,000 cycles at 5.0 A·g^-1. Hou et al. described a TMSs-based covalent coupling strategy for anode material synthesis utilizing amide bonds to bind TMSs and CNTs shown in Figure 4A^[57]. By forming a compact ZnS-CNT heterojunction to promote electron/ion transfer and ensure the stability of the structure shown in Figure 4B. The tight coupling interaction among ZnS and CNT resulted in the formation of the composite material with outstanding pseudo-capacitance behavior and highly reversible electrochemical process and thus had excellent electrochemical performance. It shows favorable specific capacity, rate capability, and cycling stability shown in Figure 4C-E. Chen et al. prepared SnS₂/FeS₂/reduced graphene oxide (rGO) composite by introducing FeS₂in situ and growing SnS₂ simultaneously on rGO^[58]. The composition and unique structure facilitated the expeditious dissemination of Na⁺ and served to buffer the volume change. The test results indicated that the electrode had a specific capacity of 768.3 and 541.2 mAh·g^-1 at 0.1 and 1.0 A·g^-1.

Figure 4. (A) Schematic illustration of synthesis procedure of the CC-ZnS/CNT and its corresponding (B) HRTEM image, (C) Cycling performance at 0.1 A·g^-1, (D) Rate capability compared with pure ZnS and ZnS/CNT, and (E) Long-term cycling performance at 2.0 A·g^-1. Quoted with permission from Hou et al.^[57]; (F) Schematic representation depicting the synthesis process of the MoSe₂/Cr₂Se₃@N-HCSs; (G) Cycling stability of MoSe₂/Cr₂Se₃@N-HCSs in SIBs. Quoted with permission from Wang et al.^[60]. CC-ZnS/CNT: Covalently coupled ZnS/CNT composite; HRTEM: high-resolution transmission electron microscopy; SIBs: sodium-ion batteries.

Aminu Muhammad et al. successfully synthesized N-doped rGO (N-rGO) wrapped Ni₃S₂ nanocrystalline compound (Ni₃S₂-N-rGO-700 °C) through a simple phase conversion engineering process^[59]. The strong coupling effect during annealing on the structural phase transition led to the formation of Ni₃S₂ nanocrystals and N-doped decreased graphene oxide nanocrystals, which significantly influenced the Na⁺ storage properties of Ni₃S₂-N-rGO-700 °C nanocarbons. The Ni₃S₂-N-rGO-700 °C compound demonstrated a notably reversible capacity of 377 mAh·g^-1 at 2.0 A·g^-1. Furthermore, the system demonstrated a remarkably high level of coulombic efficiency, exceeding 100% over more than 400 cycles. Wang et al. developed N-doped hollow carbon nanospheres with uniformly embedded MoSe₂/Cr₂Se₃ heterojunction (MoSe₂/Cr₂Se₃@N-HCs) as SIB electrode materials shown in Figure 4F^[60]. MoSe₂ and Cr₂Se₃ heterojunctions induced lattice mismatches at the hetero-structural interface, which contributed to spin-polarized surface capacitance. MoSe₂/Cr₂Se₃@N-HCS exhibited a striking capacity of 498 mAh·g^-1 at 0.1 A·g^-1 with outstanding durability over repeated cycles (405 mAh·g^-1 capacity after 1,000 cycles and a coulombic efficiency of 99.8%) shown in Figure 4G.

Chen et al. employed a one-step hydrothermal process followed by annealing to synthesize Fe₇Ni₃S₁₁/CN^[61]. The resulting spherical solid solution Fe₇Ni₃S₁₁/CN had a hollow structure with ample active sites for Na⁺ storage. Additionally, the Fe₇Ni₃S₁₁/CN nanoparticles were constrained by external CN, which promoted structural stability. The reversible specific capacity at 0.2 A·g^-1 was 567 mAh·g^-1. When exposed at 6.0 A·g^-1, Fe₇Ni₃S₁₁/CN maintained a capacity of 320 mAh·g^-1 over 5,000 cycles.

Alloying-type materials

Nanocomposites can mitigate volume expansion/shrinking of antimony-based materials during the de/sodiation process and improve their performance as an anode for SIB. Wang et al. proposed an innovative N-doped 3D porous carbon and studied the design and fabrication of network-confined nano-FeSb alloy composites (3DFeSb@NC) with robust Fe–N–C bonds^[62]. This distinctive structure fully mitigated the volume change of Sb throughout the cycle, shortened the ion/electron proliferation distance, and enhanced the electrical conductivity. The 3D FeSb@NC anode had an excellent charge/discharge capability of 231 mAh·g^-1 at 5.0 A·g^-1 and adequate capacity retention of 85% after 750 cycles. Ma et al. synthesized an innovative Sb₂O₃@Sb nanocomposite with an ultra-fine Bi-continuous mesoporous structure shown in Figure 5A-C^[63]. It demonstrated an exceptionally high specific capacity (659 mAh·g^-1 at 0.2 A·g^-1) in Figure 5D, overwhelming rate capacity (200 mAh·g^-1 at 29.7 A·g^-1), and outstanding cyclic stability with a capacity conservation rate of 99.8% after 200 cycles at 0.2 A·g^-1 shown in Figure 5E.

Figure 5. (A) Fabrication process of the mesoporous Sb₂O₃@Sb nanocomposite and its corresponding (B) SEM image, (C) TEM image, (D) Cycling performance, (E) Rate performance. Quoted with permission from Ma et al.^[63]; (F) The one-step dealloying process for NP-SnSb alloy and its corresponding (G) SEM image; (H) Cycling performance (as benchmarked with P-Sn); (I) Rate performance at various specific currents. Quoted with permission from Ma et al.^[67]. SEM: Scanning electron microscope; TEM: transmission electron microscopy; NP-SnSb: bimetallic nanoporous SnSb; P-Sn: porous Sn.

Edison et al. incorporated Fe into SnSb by rapid solidification technology to produce a novel ternary nanocrystalline composite, which effectively reduced the stress resulting from the conversion reaction, thereby enhancing the cyclic stability^[64]. The SIB anode material provided a specific capacity of ~500 mAh·g^-1 over more than 120 cycles at 0.5 A·g^-1. Chen et al. synthesized Bi nanoparticles embedded in an N-doped carbon matrix (Bi@NC)^[65]. Even at 10 A·g^-1, Bi@NC electrodes still provide a high capacity of 338.2 mAh·g^-1 (86% retained at 0.1 A·g^-1) and a high capacity of 326.9 mAh·g^-1 after 5,000 cycles at 2.0 A·g^-1. Gao et al. successfully prepared NP Bi-Sb alloys with varying Bi/Sb ratios using Mg-based precursors^[66]. Among them, the np-Bi₂Sb₆ electrode provided a reversible capacity of 257.5 mAh·g^-1 after 2,000 cycles at 0.2 A·g^-1 and exhibited long-term cycle performance at 1.0 A·g^-1 with only 0.0072% capacity attenuation per 10,000 cycles. Ma et al. fabricated a bimetallic single-phase NP-SnSb alloy with a bicontinuous ligament channel structure shown in Figure 5F and G^[67]. The NP-SnSb alloy had a special specific capacity [Figure 5H], with a reversible capacity of 506.6 mAh·g^-1 after 100 cycles at 0.2 A·g^-1. Even at 10 A·g^-1, the reversible capacity can reach 458.2 mAh·g^-1, and when the specific current was reset to 1.0 A·g^-1, the reversible capacity fully recovered to 560.6 mAh·g^-1 and remained stable for 100 cycles [Figure 5I]. Lei et al. successfully prepared heteroatomic amorphous phosphate-rich selenide molecules encapsulated in porous CNTs (a-P₉Se@pCNTs), which showed exceptional reversible Na⁺ storage capacity with ICE up to 85.9% and specific capacity up to 2,215 mAh·g^-1 at 0.2 A·g^-1[68]. Supplemented by DFT simulations and comprehensive experimental studies, it was demonstrated that the low formation energy of Na products alleviated the issue of pulverization and prevented undesired decomposition.

Carbonaceous materials

The hydrothermal pretreatment process can facilitate the stabilization of the morphology and structure of the carbon material, thereby enhancing the performance of the battery. This is achieved by introducing a greater number of defects and nanopores, which provide additional active sites for the storage of sodium ions^[69]. Lu et al. employed a low-cost, high-carbon yield bitumen as a precursor through a pre-oxidation strategy, enabling the bitumen to undergo effective structural transformations from ordered to disordered during the subsequent carbonation process^[70]. The structural advantages of disordered carbon help to achieve excellent Na⁺ storage performance, resulting in a reversible capacity of 300.6 mAh·g^-1 at 0.03 A·g^-1. Zhao et al. synthesized an uncommon-performance HC anode with low defects and long-sequence graphite domains of 0.04 m-MnHC by the Mn²⁺ ion-assisted catalytic process, which catalyzed the growth of graphite sheets into long-sequence graphene layers^[71]. The SIB anode material had a high ICE (92.05%) and capacity (336.8 mAh·g^-1 at 0.02 A·g^-1). Wang et al. prepared carbon materials from hazelnut shells by washing with hydrochloric acid to remove impurities and alter microstructure to prepare a HC SIB anode material with a reversible capacity of 342 mAh·g^-1 at 0.02 A·g^-1 and an ICE of 91%^[72].

Meng et al. used the solvothermal method (ST) to form a precursor in liquid phenol-formaldehyde resin (PF) and ethanol (EtOH) with a volume ratio of 2:1, which was carbonized at 1,400 °C, and the synthesized sample was named HC-21-1400^[73]. It showed a high reversible capacity of about 410 mAh·g^-1 and a high ICE of 84% at 0.03 A·g^-1. The capacity retention was 96% after 40 cycles. Kamiyama et al. synthesized the high-capacity HC material HC600-1500 (F50:50) by MgO template technology [Figure 6A]^[74]. After MgO was removed by pickling, nanoscale pores were formed and these pores were the main active sites for electrochemical sodium storage [Figure 6B and C]. Increasing the pretreatment temperature led to the growth of MgO template particles, resulting in a larger size and number of open micropores. As shown in Figure 6D, HC600-1500 (F50:50) obtained by pickling and post-heating at 1,500 °C showed a high reversible capacity of 478 mAh·g^-1. It had good capacity retention and excellent ionization rate performance at 0.025 A·g^-1 [Figure 6E]. Li et al. made an HC material from waste cork (CC) as a raw material, which effectively stored sodium ions by adjusting pore types^[75]. Due to the natural porous structure of cork, CC presented a uniform layered porous morphology, which improved passivation [Figure 6F and G] (related to platform capacity), and CC-1600 provided a high capacity retention of 87% after 200 cycles at 0.03 A·g^-10.1 C, with a remaining capacity of 312 mAh·g^-1 [Figure 6H]. Li et al. proposed screening carbon with a highly adjustable nanopore tightened pore inlet, and scalable and reversible low potential platform (< 0.1 V) for high energy anode materials^[76]. The results indicated that the constricted pore inlet prevented the formation of a SEI phase within the nanopore, so the Na clusters formed a platform and the number of Na clusters increased almost linearly by generating a larger area of hydrophilic pore surface. Moreover, the control of pore diameter ensured the reversibility of sodium cluster formation and generated a screening carbon anode with a superb platform capacity of 400 mAh·g^-1 at 0.05 A·g^-1.

Figure 6. (A) Schematic illustration for the two mixing procedures for preparation of the mixtures of Mg Glu and Glc; TEM images and SAED patterns of (B) HC600-1500 (F0:100) and (C) HC600-1500 (F50:50); (D) Rate performance and (E) Cycle performance of HC600-1500 (F50:50). Quoted with permission from Kamiyama et al.^[74]; (F) SEM image and (G) HRTEM and SAED images of CC-1600; (H) Cyclic capability of CCs. Quoted with permission from Li et al.^[75]; Mg Glu: Magnesium gluconate; Glc: glucose; HC600-1500 (F0:100) and HC600-1500 (F50:50): HC by heating the Mg Glu/Glc ratio of 0:100 mol and 50:50 mol, respectively, with preheat-treatment temperature at 600 °C and post-heat-treatment temperatures at 1,500 °C; HC400-1500 (F50:50) and HC550-1500 (F50:50): HC by heating the Mg Glu/Glc ratio of 50:50 mol and preheat-treatment temperature at 400 °C and 550 °C, respectively, with post-heat-treatment temperatures at 1,500 °C; CC-Pre-800: carbonisation process at 800 °C for a period of 2 h, followed by a washing procedure with a dilute hydrochloric acid solution (30% concentration); CC-1200, CC-1400, and CC-1600: cork-derived HC pyrolyzed at 1,200, 1,400, and 1,600 °C, respectively, under argon flow for 2 h; TEM: transmission electron microscopy; SAED: selected area electron diffraction; SEM: scanning electron microscope; HRTEM: high-resolution transmission electron microscopy.

Titanium-based oxides

Anatase-type TiO₂ is characterized by an open bilateral TiO₆ octahedral crystal structure, which facilitates the insertion and removal of Na⁺. However, weak electronic conductivity and tardy ion diffusion severely limit the remarkable-rate performance of TiO₂. Various strategies have been investigated to modify TiO₂ to enhance Na storage capacity, for instance, the implementation of VO and the designing of nanoscale structures. Ma et al. prepared VOs-TiO_2-x nanosheets with folds, which increased the conductivity and storage capacity of Na by adding Ti³⁺ and abundant VO^[77]. At 20.0 A·g^-1 the capacity can reach 91.0 mAh·g^-1 with extremely long cycle stability. Han et al. proposed a two-solvent enhanced pressure strategy to prepare N-doped TiO₂/C nanocomposites by using tetrabutyl-n-titanate tetramer as titanium dioxide and reducing its size with mutual solvent ethanol and doped dimethylformamide as pressors^[78]. The material had an uncommon specific capacity of 370.4 mAh·g^-1 at 0.05 A·g^-1, a superior rate capability of 111.7 mAh·g^-1 at 17.0 A·g^-1, and an extensive cycle life of over 20,000 cycles. Hwang et al. synthesized carbon-free anatase/bronze TiO₂ microspheres [TiO₂(A/B)-MS] by ST^[79]. The highly dense structure constructed by nano-primary particles enhanced the stability of the structure shown in Figure 7A and B. TiO₂(A/B)-MS exhibited a specific capacity of 221 mAh·g^-1 at 0.025 A·g^-1 shown in Figure 7C and a superb cycle stability of more than 1,000 cycles at 1.0 and 10.0 C shown in Figure 7D. Yang et al. synthesized mesoporous floral titanium-deficient anatase TiO₂ (TDAT)^[80]. The presence of Ti vacancies enhanced the conductivity of Li and Na ions and provided additional storage sites. The TDAT electrode had a superb specific capacity of 219.9 mAh·g^-1 at 0.05 A·g^-1, excellent rate performance, and brilliant cycle performance of 15,000 cycle tests at 2.0 A·g^-1.

Figure 7. (A) Synthesis process of the TiO₂(A/B)-MS material and its corresponding (B) HRTEM image; (C) Discharge capacities at various C rates; (D) Discharge capacities for long-term cycling. Quoted with permission from Hwang et al.^[79]; (E) Schematic illustration of the synthesis process of titania-based nanoarchitectures and their corresponding SEM images and (F) Cycling performance at 0.2 C; (G) Long-term cycling performance of 2D UNs⊂TC at 50 C. Quoted with permission from Xia et al.^[82]. aTiO₂/HDA: The titania oligomers by hydrogen-bonding to condense into a compact amorphous inorganic-organic shell; aTiO₂/HDA@mcTiO₂: a hybrid shell consisting of an amorphous inorganic-organic inner shell tightly confined by a mesoporous crystalline outer shell; mcTiO₂: the HDA was partly removed, and a highly crystalline mesoporous shell; 2D UNs⊂TC: hierarchical nano configurations with the stepwise varying morphologies of 2D UNs vertically encapsulated in TiO₂/carbon hollow nanospheres; MT@C: mesoporous hollow nanospheres consisting of graphitic carbon-coated TiO₂ nanocrystals; HRTEM: high-resolution transmission electron microscopy; SEM: scanning electron microscope.

Lan et al. proposed a TiO₂ mesoporous structure with a precisely controlled mesoporous framework^[81]. The meticulously crafted mesoscopic TiO₂ model attained a maximum specific capacity of 240 mAh·g^-1. Under the high mass loading of 9.47 mg·cm^-2, the area capacity of 2.1 mAh·cm^-2 was further achieved. Xia et al. constructed 2D ultrathin nanosheets (UNs) vertically aligned inside of graphitic carbon-coated mesoporous TiO₂ hollow nanospheres (2D UNs⊂MT@C) shown in Figure 7E^[82]. It consisted of regularly alternating monomolecular layers of TiO₂ and carbon, which facilitated rapid electron and Na⁺ transportation and abundant 2D heterogeneous interface active sites. The structure demonstrated excellent mechanical stability and structural integrity, ensuring more than 500 cycles at 0.2 C and a capacity decay rate of 0.04% of the battery shown in Figure 7F. At 50 C, the long-term cycle is up to 20,000 times shown in Figure 7G. Liu et al. used strong 1D TiO₂ nanowires as a substrate to link these nanosheets together and constructed a bundle layered structure [TiO₂(B)-BH] with a capacity of 131 mAh·g^-1 at 0.5 C and a distinguished rate of 10 C^[83]. A reversible capacity of up to 500 cycles can still be maintained at 85 mAh·g^-1.

Conversion-type materials

Yang et al. synthesized nickel cobaltite@nitrogen-doped holey graphene (NCO@N-HG) and Fe₃O₄@N-HG composites by loading nickel-cobaltate and magnetite nanoparticles in a 3D nitrogenous porous graphene framework, respectively^[84]. The thin graphene sheet buffered volume changes and the unique large pore structure combined with the existence of nanoholes in the graphene sheet facilitated ion transport during the processes of charge/discharge, contributing to the observed superb specific capacity and good cycling performance of 510 mAh·g^-1 over 100 cycles at 0.1 A·g^-1. The ball milling technology^[85] was employed to adjust the lattice strain of NiS nanocomposites. The restructured anode had a superb capacity, great rate performance (495.7 and 383.2 mAh·g^-1 at 0.2 and 15.0 A·g^-1), and cycle performance (16,000 cycles at 5.0 A·g^-1). Li et al. successfully prepared micron-scale MoO₂/β-MoO_3-x Mosaic composites by in-situ micro-corrosion method^[86]. Owning to the micron-scale particle morphology, the composite with high solid density (1.68 g·cm^-3) and low surface area (8.4 m²·g^-1) exhibited a prominent specific capacity of 1,742 mAh·g^-1. Furthermore, the dual conductive design achieved by combining high electrical conductivity MoO₂ and superb ionic conductivity β-MoO_3-x resulted in long cycle stability for the micron-sized particles (87% capacity retention after 500 cycles) and impressive rate performance (440 mAh·g^-1 at 7.0 A·g^-1). Wang et al. successfully developed a coral-like structure composed of Mn₃O₄ nanotubes embedded in 2D porous graphene sheets using a double template of NaCl and SiO₂^[87]. Mn₃O₄@pGS exhibited a prominent specific capacity of 770 mAh·g^-1 at 0.1 A·g^-1 for Na⁺ storage.

Zhang et al. constructed MO@NCF (MO = Co₃O₄, Mn₃O₄, Fe₃O₄, NCF: nitrogen-enriched 3D carbon nanofiber networks) by coordinating nanofibrous carboxymethyl chitosan (CMCh) hydrogel with metal ions to create a steady metal-polysaccharide skeleton, then a multistage porous structure was formed through the pyrolysis process^[88]. The Co₃O₄@NCF exhibited an impressive Na⁺ storage capacity of 645 mAh·g^-1 at 0.1 A·g^-1 and a great rate performance of 191 mAh·g^-1 at 4.0 A·g^-1. Xu et al. synthesized R-Co₃O₄/C hybrid hollow spheres composed of Co₃O₄ nanoparticles uniformly dispersed in a carbon matrix^[89]. The R-Co₃O₄/C mixed hollow microspheres showed superior Na⁺ storage capacity (up to 712 mAh·g^-1), remarkable rate capacity (223 mAh·g^-1 at 5.0 A·g^-1), and excellent cycle life (74.5% capacity retention after 500 cycles). Zhao et al. prepared highly crystalline mesoporous Fe₃O₄ nanoparticles encapsulated in hollow nitrogenous carbon nanospheres (represented by HCM-Fe₃O₄@void@N-C)^[90]. A remarkable specific capacity of 372 mAh·g^-1 was achieved after the first 5 cycles at 0.16 A·g^-1, which can be gradually increased and maintained at 522 mAh·g^-1 over 800 cycles. In addition, brilliant rate performance of 196 mAh·g^-1 and high coulomb efficiency (≈ 100%) at 12.0 A·g^-1 were obtained.

Cao et al. successfully fabricated fine hierarchical heterostructures composed of antimonous sulfide Sb₂S₃@FeS₂ (SFS) hollow nanorods embedded in an N-doped carbon matrix shown in Figure 8A and B^[91]. The well-designed architecture facilitated the rapid diffusion of Na ions and enhanced charge transfer at heterogeneous interfaces. Simultaneously, this structure greatly accelerated electron/ion transport and inhibited volume expansion over a long cycle. Thus, the composite demonstrated outstanding rate performance of 537.9 mAh·g^-1 at 10.0 A·g^-1 and exceptional cyclic stability of 85.7% capacity retention after 1,000 cycles at 5.0 A·g^-1 shown in Figure 8C. Liu et al. investigated a general strategy for successfully preparing a variety of high entropy oxide-polyoxometalate (HEO-POM) self-assembled nanowires using POM clusters shown in Figure 8D, where the types of metal oxides and POM clusters can be flexibly adjusted^[92]. A series of experiments and molecular dynamics simulations showed that the formation of HEO-POM sub-1 nm nanowires (SNWs) followed the oriented attachment mechanism; namely, the POM clusters acted as linkers for multiple metal oxides under the action of van der Waals interactions. CoZnCuNiFeZrCeO_x-phosphomolybdic acid (PMA) SNWs can maintain a high capacity retention of about 92% even after 5,000 cycles at 10 C shown in Figure 8E.

Figure 8. (A) Schematic illustration of the fabrication process of tSFS/C composite and its corresponding (B) TEM image; (C) Cycle performance of SFS/C, SFS, and Sb₂S₃ at 5 A g^-1. Quoted with permission from Cao et al.^[91]; (D) The snapshot of a nanowire composed of CoZnCuNiFeZrCeO_x-PMA building blocks with HAADF-STEM images and AC-TEM images; (E) Cycling stability of CoZnCuNiFeZrCeO_x-PMA SNWs. Quoted with permission from Liu et al.^[92]. SFS/C: Sb₂S₃@FeS₂/N-doped graphene; TEM: transmission electron microscopy; PMA: phosphomolybdic acid; HAADF: high-angle annular dark field; STEM: scanning transmission electron microscopy; AC-TEM: aberration-corrected transmission electron microscopy; SNWs: sub-1 nm nanowires.

Alloying-type materials

Yang et al. designed an optimized Bi@Void@C nanosphere (Bi@Void@C-2) with a yolk-shell structure, which exhibited a conspicuous cycle life of 198 mAh·g^-1 after 10,000 cycles at 20.0 A·g^-1[93]. Theoretical calculations and experimental studies showed that the external layer of the thin carbon shell was of significant importance in improving electronic conductivity and preventing the aggregation of Bi nanoparticles. Fan et al. designed a distinctive 3D collector assembled from an array of 1D nanowires (1D NWAs) anchored to 3D porous copper foam for constructing core-shell Cu@Sb nanowires^[94]. The resulting 3D hierarchical anode with interconnected 3D micron-sized holes and abundant interwire spaces effectively relieved the structural stress during cycling, which ensured the structural integrity and helped to achieve a uniform ion/electron distribution over the entire surface as shown in Figure 9A. The electrode exhibited a high specific capacity of 605.3 mAh·g^-1 at 0.33 A·g^-1 [Figure 9B], and a supernormal capacity retention of 84.8% even at a prominent specific current of 3.3 A·g^-1 [Figure 9C].

Figure 9. (A) SEM images of Sb@3D Cu NWAs from (a and b) a top view, (c) a side view, and (d and e) an enlarged top view. Energy dispersive X-ray spectroscopy maps for elements (f) Cu and (g) Sb; (B) Cyclability curves; (C) Rate performance. Quoted with permission from Fan et al.^[94]; (D) Structural evolution of bulk Bi flakes. Secondary-electron images recorded at low magnification; (E) Changes in specific capacity of Na-Bi half-cells measured at various C-rates using an electrolyte containing DME; (F) Changes in the specific capacity of the Na-Bi half-cell with DME and EC/DEC solvent measured at 20.0 C as a function of the battery cycle. Quoted with permission from Kim et al.^[97]. SEM: Scanning electron microscope; Sb@3D Cu NWAs: a thin film of Sb is electrochemically deposited on 3D Cu NWAs; DME: dimethoxy-ethane; EC: ethylene carbonate; DEC: diethyl carbonate.

Carbon fiber has excellent electrical conductivity, a significant specific surface area, and superb electrochemical stability. Through electrostatic spinning and reduction roasting, an independent carbon fiber film was obtained. Owing to the superb specific surface area and porosity, the one-dimensional structure provided more active centers and greatly alleviated the volume expansion. Dang et al. developed a distinctive pod-like independent film^[95]. This film incorporated SbSn particles into an integrated carbon matrix. In addition, the following materials are included: hollow carbon spheres and N-doped carbon fibers (B-SbSn/NCFs), which not only improved the electrical conductivity and flexibility but also provided sufficient space to accommodate volume changes. Moreover, the B-SbSn/NCFs self-supporting film can serve as a direct anode without the need for polymer binders and conductors, which improves the energy density and reaction kinetics. The membrane electrode demonstrated a specific capacity of 486.9 mAh·g^-1 with a coulombic efficiency approaching 100% after 400 cycles at 0.1 A·g^-1.

Song et al. prepared dispersed in situ 3D cross-linked carbon nanofiber SnSb@CNF^[96]. The 3D cross-linked nanostructures ensured more Na storage active sites, stable structure, and increased electrical conductivity. Thus, the SnSb@CNF/CNT anode exhibited excellent cycle stability (161 mAh·g^-1 in 1,000 cycles at 1.0 A·g^-1). Kim et al. designed a massive bismuth anode that featured ultra-fast charging, long cycle stability, and substantial energy storage capacity without expensive nanomaterials or surface modifications^[97]. During the electrolyte cycle of dimethoxy-ethane (DME), the massive Bi particles gradually transformed into porous nanostructures shown in Figure 9D, thus storing Na⁺ by forming a phase with high Na diffusivity. Even at the very fast charging rates of 80 and 100 C, 94% and 93% of the capacity measured at 1 C were retained, respectively shown in Figure 9E. The Na-Bi half-cell delivered a specific capacity of 379 mAh·g^-1 (97% of 1 C measurement) at 7.7 A·g^-1 (20 C) over 3,500 cycles shown in Figure 9F. Eaves-Rathert et al. designed the microstructured Pb-Sb-Sn, demonstrating that these can be effectively controlled through simple heat treatment of the unrefined powder to achieve a max specific capacity of 522 mAh·g^-1[98].

Carbonaceous materials

HC has a substantial specific surface area and numerous defect sites and more sodium ions need to be sacrificed to form SEI films, which ultimately generates an increase in irreversible specific capacity and a decrease in coulombic efficiency. The ICE of graphite anode is more than 90% while that of HC anode is usually only 60% to 80%.

Fang et al. employed sodium diphenyl ketone (Na-DK) as a reducing agent for the bifunctional preparation of HC^[99]. The chemically synthesized sodium facilitated the formation of a robust sodium fluoride (NaF)-rich SEI layer in the electrolyte by pre-doping HC. Figure 10A analyzes the optimal structure of adsorbed and non-adsorbed Na⁺, revealing the interaction of Na⁺, dibenzoyl (DK), and HC. As illustrated in Figure 10B, the presodiated HC (pHC) anode exhibited a high reversible capacity of 251.5 mAh·g^-1 at 0.5 A·g^-1. As demonstrated in Figure 10C, it enabled a full battery of pHC//Na₃V₂(PO₄)₃ (NVP) to operate for over 700 cycles with minimal capacity degradation. Bai et al. developed a sustainable film for an SEI on an HC anode by adding a small amount of ester additive as an agent for forming the SEI film in the ether-based electrolyte is illustrated in Figure 10D^[100]. This membrane prevented the coupling of the protective SEI with the fast transmission of Na⁺, leading to improved electrochemical performance of the HC anode and effectively shielding the HC electrode from the influence of structural changes shown in Figure 10E and F. The material had a specific capacity of 211 mAh·g^-1 after 2,000 cycles at 1.0 A·g^-1. The decay rate was only 0.0023% per cycle displayed in Figure 10G.

Figure 10. (A) Optimized structures of Na⁺ adsorbed on graphene layer of 60 carbon atoms containing five/seven-membered rings; (B) Rate capability of HC and pHC; (C) Cycle performance of full cells at 0.5 A·g^-1 of pHC//NVP and HC//NVP. Quoted with permission from Fang et al.^[99]; (D) Schematic illustration of the evolution of SEI films in the DME-based and DME-0.5%VC electrolytes upon prolonged cycling on HCs; TEM images of HC anodes after 2,000 cycles in the (E) DME-based and (F) DME-0.5%VC electrolytes; (G) Long-term cycling performance. Quoted with permission from Bai et al.^[100]. DME-VC-SEI: SEI films in DME and vinylene carbonate electrolytes; HC: hard carbon; pHC: presodiated HC; NVP: Na₃V₂(PO₄)₃; SEI: solid electrolyte interphase; DME: dimethoxy-ethane; TEM: transmission electron microscopy.

Jin et al. presented an advanced electrolyte containing sodium difluorosulfonyl imide (NaFSI)-triethyl phosphate (TEP), which exhibited great stability with high voltage cathodes and was able to improve the cycling of Na⁺ batteries^[101]. The Na||NaNi_0.68Mn_0.22Co_0.10O₂ (NaNMC) battery demonstrated a capacity retention of 89% after 500 cycles with a cut-off voltage of 4.2V (vs. Na⁺/Na). Jin et al. introduced NaFSI-TEP/1,1,2,2-tetrafluoroethyl-2,2,3,3-tetrafluoropropyl ether (TTE) (molar ratio 1:1.5:2), a non-combustible electrolyte for SIBs, which stabilized the interface at the layered O₃-NaCu_1/9Ni_2/9Fe_1/3Mn_1/3O₂ (Na-CNFM) cathode and HC anode simultaneously for consistent cyclic performance^[102]. The capacity retention rate of Na||HC battery after 500 cycles was 94.8% (247.9 mAh·g^-1). The high-quality interface established on both the cathode and anode ensured HC’s exceptional cycle stability.

Titanium-based oxides

Li et al. combined 16 nm anatase TiO₂ with diglyme-based electrolyte and the reversible capacity reached 257.9 and over 100 mAh·g^-1 at 0.1 and 2.0 A·g^-1[103]. Both of them were significantly better than the ethylene carbonate (EC)/diethyl carbonate (DEC)-based electrolytes. It had been shown that the ether-based electrolyte promoted the Na⁺-induced TiO₂ structure transformation, and the charge transfer kinetics at the SEI/electrode interface played a crucial role in rate cycling performance. Meng et al. reported hollow NTO microspheres (H-NTO) with oxygen defects prepared from 2D ultra-thin nanosheets^[104]. These microspheres exhibited a unique chemical bonding NTO/C(N) interface shown in Figure 11A. When the temperature reached 60 °C, the specific capacity of the H-NTO electrode was 59 mAh·g^-1 at 20 A·g^-1 demonstrated in Figure 11B. By optimizing the electrolyte, the H-NTO electrode can be stably cycled at -40 °C for 200 days without a decrease in capacity shown in Figure 11C. The outstanding cycling stability was due to the stable SEI formed by the NTO/C(N) interface and the highly adaptable electrolyte/electrode interface.

Figure 11. (A) Synthetic route for H-NTO microspheres and its corresponding (B) Rate performance at different temperatures; (C) Cycle performance. Quoted with permission from Meng et al.^[104]; (D) Synthesis of Li⁺ pre-inserted SEI layer; (E) Rate performance; (F) Cycling performance at 1.0 A·g^-1. Quoted with permission from Cha et al.^[106]. A-TNT: Anatase TiO₂ NTs (A-TNT); Li-TiO₂ NTs: Li⁺ inserted anatase TiO₂ NTs; Li-TiO₂ NTs-De: Li⁺ inserted/de-inserted anatase TiO₂ NTs; H-NTO: hollow Na₂Ti₃O₇; SEI: solid electrolyte interphase.

Xu et al. showed that anatase TiO₂ underwent amorphous crystallization during the first alkalization process and in the subsequent cycles^[105]. The amorphous phase showed pseudo-capacitor sodium storage behavior. For the first time, it was found that the ether-based electrolyte was able to form a thin (≈ 2.5 nm) and stable SEI layer while the conventional carbonate-based electrolyte could form a thick (≈ 10 nm) and growing SEI layer. The elevated minimum molecular orbital energy of the ether solvent/ionic complexes was the underlying cause of this discrepancy. In the ether-based electrolyte, after 500 cycles, a specific capacity of 192 mAh·g^-1 can be achieved at 0.1 A·g^-1, which significantly exceeds the capacity of the carbonate electrolyte. Cha et al. obtained significantly enhanced SEI layers on TiO₂ nanotube (TiO₂ NTS) arrays, which improved the performance of sodium battery systems^[106]. Pre-insertion of Li⁺ into TiO₂NTS can make SEI meet the conditions of Na⁺ replacement shown in Figure 11D. SIBs constructed using pre-embedded lithium nanotubes had superb rate capability (132 mAh·g^-1 at 1.0 A·g^-1) shown in Figure 11E and superb Na⁺ cycle stability (capacity retention of 99.9% after 250 cycles at 1.0 A·g^-1) shown in Figure 11F.

Conversion-type and alloying-type materials

Li et al. developed a new novel type of alloying/conversion-based anode in which the metal oxide (e.g., MnO) micro dumbbell skeleton was designed to stabilize large capacity alloyed (e.g., Sn) nanoparticles^[107]. In combination with electrolyte engineering methods, the Sn-MnO@C anode exhibited outstanding Na⁺ storage performance, including an excellent specific capacity of 370 mAh·g^-1, a good rate capacity of more than 10 A·g^-1, and a favorable cycling life of more than 500 cycles. Yang et al. performed surface engineering on tin nanorods by using N-doped carbon layers (Sn@NC) and interface engineering strategy shown in Figure 12A^[108]. The uniform surface modification effectively enhanced electron and Na⁺ transport kinetics and interacted with the ether-based electrolyte to form a strong organic-inorganic SEI shown in Figure 12B. Moreover, Sn@NC anodes provided ultra-long cycle stability of more than 10,000 cycles shown in Figure 12C.

Figure 12. (A) Schematic of synthesis of Sn@NC composite and its (B) TEM image; (C) Long cycling performance of composite anodes at 5.0 A·g^-1 in the DEGDME-based electrolyte. Quoted with permission from Yang et al.^[108]; (D) Composition/nanostructure of SEIs probed by cryo-TEM and XPS. (a-f) HAADF STEM and TEM images of SEIs in (a, c, and e) NaBF₄/EC/DMC and (b, d and f) NaBF₄/diglyme with SAED as insets in (c and d). Note that the insets in (e and f) are the enlarged HRTEM and FFT images for the SEI layer region; (E) Cyclic stability of Sn microparticle electrodes; (F and G) The first three discharge/charge curves. Quoted with permission from Huang et al.^[110]. SnO₂@PDA: The SnO₂ nanorods were further encapsulated by dopamine; Sn@NC: 3D flower-like architecture consisting of nitrogen-doped carbon-coated tin nanorods; TEM: transmission electron microscopy; DEGDME: diethylene glycol dimethyl ether; SEIs: solid electrolyte interphases; XPS: X-ray photoelectron spectroscopy; HAADF: high-angle annular dark field; STEM: scanning transmission electron microscopy; TEM: transmission electron microscopy; EC: ethylene carbonate; DMC: dimethyl carbonate; SAED: selected area electron diffraction; HRTEM: high-resolution transmission electron microscopy; FFT: fast Fourier transform.

Chu et al. employed an ultra-low concentration electrolyte strategy that included a significantly high amount of sodium salts of tetra phenyl-borate ion (BPH^4-) and a DME solvent to stabilize micron tin and bismuth anodes^[109]. In the 0.1 M NaBPh₄-DME electrolyte, sodium had a faster migration rate attributed to the significant steric hindrance of BPH^4- anion, which made it difficult to approach the electrode/electrolyte interface and thus the side reactions at the electrode/electrolyte interface were reduced. With electrolyte optimization, the specific capacity of 600 mAh·g^-1 on the Sn anode can be reached at 0.5 A·g^-1. Huang et al. explored the nanostructure and nanomechanical SEI derived from carbonate-and ether-based materials on micron-sized tin particles^[110]. The structure/composition and properties of SEI derived from EC/DMC and diethylene glycol dimethyl ether were elucidated. The SEI derived from diethylene glycol dimethyl ether exhibited high elasticity and ionic conductivity, which can be attributed to the unique structure of the amorphous particles dispersed in polymer-like substances shown in Figure 12D. Compared to the NaBF₄/(EC/DMC) electrolytic liquid phase, the diethylene glycol electrolyte provided significantly improved coulomb efficiency and cycle stability shown in Figure 12E-G.

Wu et al. demonstrated the rapid production of ultrafine Bi nanoparticles embedded in a 3D carbon framework (Bi@3D-CF) by a Joule heating method with a synthesis time of just 30 s^[111]. The Bi@3D-CF exhibited excellent Na storage performance in 1,2-DME based electrolyte with a capacity of 349 mAh·g^-1 after 12,000 cycles at 5.0 A·g^-1, and 311 mAh·g^-1 at 100 A·g^-1. The outstanding performance was attributed to the formation of unique 3D porous structures during cycling and the gradient organic/inorganic SEI.