Daily increasing global warming and environmental pollution have attracted worldwide attention to exploring eco-friendly renewable energy to replace traditional fossil fuels [1]. With the rise of the new energy industry and the continuous demand for energy transformation, rechargeable batteries as a key energy storage technology will usher in great development opportunities. For many years, lithium-ion batteries (LIBs) have had an overwhelming role in the electrochemical energy storage field due to their high energy density and long cycle life ^[2][3][2,3]. However, the restricted lithium salt in the earth’s crust and unevenly distributed resources impede the wide application of LIBs ^[4][5][4,5]. Therefore, researchers have turned their attention to exploring alternative energy storage technologies to partly replace LIBs.

Na, K, and Li are part of the same main class and have semblable equilibrium potential and physicochemical characteristics. Furthermore, the luxuriant reserves of Na/K in the earth and lower cost make SIBs and PIBs an ideal substitute for LIBs ^[6][7][6,7]. However, due to the larger ionic radii of Na⁺ (1.02 Å) and K⁺ (1.38 Å) than that of Li⁺ 0.76 Å), the electrode materials undergo severe volumetric alteration during the ion insertion process, thereby resulting in serious pulverization [8]. Thus, it is difficult to use existing mature LIB anodes directly in SIBs/PIBs [9]. Graphite, a traditional anode material, has been commercialized in LIBs. However, Na⁺ cannot form the corresponding intercalation compound, and K⁺ can only form KC₈ at a capacity of 280 mA h g⁻¹, seriously limiting its application in SIBs and PIBs. Oxides, sulfide, and selenides with a conversion-reaction mechanism could provide a high capacity of over 400 mA h g⁻¹. However, these transition-metal compounds have a high voltage terrace of around 1.5V, which will dramatically decrease the output voltage and energy density in full cells. Therefore, the design of the materials’ structure is more demanding when exploring anode materials for SIBs/PIBs. Bismuth (Bi) has a suitable redox potential (Bi³⁺/Bi = 0.308 V vs. SHE), insignificant thermoconductivity, low fusion point, and a high theoretical specific capacity (384 mAh g⁻¹) and volume energy density (3800 mAh cm⁻³) ^[10][11][10,11]. Bi metal belongs to the hexagonal structure, and the large layer spacing along the c-axis (d (003) = 0.395 nm) is beneficial to the diffusion of Na⁺/K⁺ ions, which has attracted considerable attention to SIBs/PIBs ^[12][13][12,13].

2. Bismuth Metal

2.1. Bulk Bismuth

Decreasing the bulk to the nanometer size shortens the Na⁺/Li⁺ diffusion distance and suppresses the structural fragmentation of the electrode ^[14][19]. Bulk Bi finally forms a three-dimensional (3D) porous network structure after repeated cycles in ether electrolyte. Kim et al. ^[15][20] use commercial bulk bismuth without any pretreatment as an anode for SIBs. During the cycle based on glycol dimethyl ether (DME) electrolytes, bulk Bi eventually forms a nano-porous structure. DFT results reveal that Bi stores Na⁺ ions by forming an intermediate (NaBi and Na₃Bi) phase with high Na⁺ diffusivity. When applied as an anode for SIBs, it has 379 mAh g⁻¹ after 3500 cycles at 7.7 A g⁻¹. DFT calculations proved that the porous structure formed by the stacking of Bi NPs not only improved the structural stability and provided a sodium-ion transport path but also achieved fast reaction kinetics and long-term cycle performance.

2.2. Nanostructured Bismuth

Designing efficient nanostructures enhances the reversible kinetics of alloying/dealloying reactions, shorten diffusion distances, and alleviate huge volume expansion. Generally, the recent structural design of bismuth metal anodes in scale can be divided into zero-dimensional (0D), one-dimensional (1D), two-dimensional (2D), and three-dimensional (3D) structures. One of the efficient strategies to improve the performance of Bi anodes is to construct hollow nanostructures. The interstitial space can accommodate the destructive volume contraction/expansion during the repeated insertion/extraction of Na⁺/K⁺, enhancing cycling stability. Pu et al. ^[16][23] synthesized hollow Bi nanotubes (Bi NTs) via a simple iodide-assisted galvanic replacement method. The TEM and SEM images reveal that the precursor Cu has a nanotube structure. SEM and TEM images manifest that the galvanic replacement sample preserved the nanotube structure well, and the Bi and Cu elements distribute on the shell. Bi NTs achieved a superior rate (319 mAh g⁻¹ at an extremely high density of 150 A g⁻¹) and high capacity retention (90% after 15,000 cycles at 20 A g⁻¹). It is effective for hollow nanotube structures to alleviate structural strain and provide ultra-fast Na⁺ transport. The Bi NT anode exhibits extraordinary rate capability and excellent cycle stability. The constructed Bi NTs/DME-based electrolytes/Na₃(VOPO₄)₂F (NVOPF) sodium-ion full battery has a high capacity of 181 mAh g⁻¹ after 550 cycles at 10 A g⁻¹ (based on anode mass calculation).

3. Bi/C Composite

3.1. Zero-Dimensional Bi/C Composite

A zero-dimensional Bi/C composite means that the size is from 1 nm to 100 nm. The carbon material is a typical soft material to restrain the volumetric strain and control the aggregation of Bi nanoparticles. Recently, yolk–shell ^[17][26], hollow, and core–shell structures have been designed as anodes for SIBs/PIBs.

3.1.1. Hollow Structure

For example, Zhu et al. ^[18][27] prepared pomegranate-like Bi@C nanospheres (PBCNSs) by a solvothermal method combined with carbothermal reduction treatment. This composite material completely wrapped ultrasmall Bi nanoparticles (~7 nm) into carbon spheres with a size of ~50nm, which achieved fast kinetics, effectively relieved volume strain, and prevented accumulation and crushing of Bi particles. PBCNS anodes achieve excellent sodium storage and rate performance (400.3 mA h g⁻¹ at 0.2 A g⁻¹ and 372.8 mAh g⁻¹ at 25 A g⁻¹) and long-term cycling stability (340 mA h g⁻¹ after 16 000 cycles at 20 A g⁻¹). Coupled with the Na₃V₂ (PO₄)₃/C(NVP@C) cathode, the PBCNSs could maintain a high specific capacity of 370 mA h g⁻¹ at 1 A g⁻¹.

3.1.2. Yolk–Shell/Core–Shell Structure

The yolk–shell/core–shell structure (the core is Bi and the shell is conductive carbon) has extra void space to better accommodate volume changes ^{[19][20][21][22]}[33,34,35,36]. Yao et al. ^[23][37] synthesized BiOCl nanosheet precursors via a solvothermal method; subsequently, dopamine coating and annealing prepared the Bi@C. The TEM results confirm that BiOCl nanospheres decompose into lesser nanoparticles, resulting in a looser nanosphere structure, and the gaps formed between the conductive carbon layers can relieve volume changes and expose more active sites. Therefore, Bi@C displays better K⁺ storage performance. Yang et al. ^[24][38] encapsulated Bi nanoparticles in N-doped carbon shells by dopamine coating and annealing and prepared core–shell carbon nanospheres (denoted as Bi@N-C). A series of characterization shows that a N-doped carbon shell can improve electrochemical activity, nanoscale Bi particles can lessen the diffusion distance of electrons/ions, and the unique core–shell structure makes Bi@N-C have an excellent rate performance in SIBs and PIBs and a longer cycle life. The close connection of active materials and conductive materials in the yolk–shell/core–shell Bi/C composites may be unable to fully release the strain stress caused by Na⁺/K⁺ ion insertion. Partial comminution of active materials may still occur at the electrodes. These structures can be further optimized by modifying the structures with some void space ^[25][26][27][39,40,41]. Gao et al. ^[28][42] constructed a multi-core shell heterostructure (Bi@Void@TiO₂⊂CNF) using Bi, hollow TiO₂, and 1D carbon nanofibers (CNFs). The cavity between the TiO₂ shell and the Bi core can relieve the volume expansion of Bi and inhibit the agglomeration of Bi particles, and the one-dimensional nanofiber structure can shorten the diffusion distance. 

3.2. One-Dimensional Bi/C Composite

A one-dimensional Bi/C composite means that the structure has a high length-to-diameter aspect ratio including tubes ^[29][44], rods ^[30][31][45,46], wires ^[32][47], and other morphologies ^[33][34][48,49] which have been proved to efficiently moderate the material volume expansion effect, improve the structural stability, and enhance the electrochemical performance. After combining 1D carbon nanostructures with Bi metal, the carbon can resist the accumulation of Bi nanoparticles, and the 1D structure can dramatically shorten the ion diffusion distance, thus significantly improving the electrochemical performance. For example, Hu et al. ^[35][50] prepared a Bi/C nanotubes (Bi/CNTs) composite via a one-step electro-deoxidation method and applied it as an anode for SIBs. The incorporation of heteroatoms (N, S, etc.) improves the electrochemical activity and electronic conductivity and further improves the storage performance of Na⁺/K⁺. ^[36][37][38][51,52,53]. The direct pyrolysis of a metal–organic framework (MOF) can yield uniformly dispersed metal nanoparticles on carbon materials ^[39][57]. Liang et al. ^[40][58] synthesized the Bi@C composite (denoted as Bi@C⊂CFs) assembled from carbon nanoribbons by direct pyrolysis of 1D Bi-based metal–organic framework (Bi-MOF) nanorods.

3.3. Two-Dimensional Bi/C Composite

Two-dimensional (2D) structural materials have many advantages, such as a large specific surface area, increasing the contact area between active material and electrolytes, shortening the diffusion path of Na ions, and so on ^[41][42][43][60,61,62]. Shen et al. ^[44][63] proposed an ultrasound-assisted electrochemical stripping method for the preparation of an ultra-thin multilayer Bi nanosheet (FBN). Xiang et al. ^[45][64] prepared a 2D carbon-coated Bi nanosheet (Bi@C) using (BiO)₂CO₃ nanosheets as templates by in situ conversion reaction. The carbon layer effectively prevented the accumulation of Bi and moderated the volumetric strain during the charge/discharge process. Designing Bi nanoparticles combined with carbon nanosheets to form a unique 2D structure could provide abundant active sites for the diffusion of Na⁺/K⁺ ions. Wang et al. ^[46][66] anchored 0D Bi nanoparticles on 2D N-doped carbon nanosheets (NCSs) from the ML-Bi@NCS composites. The NCSs with Bi nanospheres are stacked layer by layer to form multi-layer structures (denoted as ML-Bi@NCSs). This layered structure has enough voids to provide abundant active sites and ion migration paths. When used as the anode of SIBs, ML-Bi@NCSs can provide an outstanding capacity of 378 mA h g⁻¹ at 200 m A g⁻¹. The SEM results after cycles verified that the anode forms porous frameworks under repeated discharging/charging.

3.4. Three-Dimensional Bi/C Composite

A three-dimensional porous network structure can offer a short ion spread distance and abundant transmission channels for Na⁺ ions, can augment the contact with the electrolytes, and can relieve volumetric strain, thus improving the anode’s structural stability and the ion migration ability of materials ^{[47][48][49][50][51]}[70,71,72,73,74]. Qiu et al. ^[52][75] prepared uniformly distributed Bi nanospheres (Bi-NS) of 20~30 nm and embedded them in a spongy 3D porous conductive framework (Bi-NS@C). This structure can form a conductive network to prevent particles from gathering. Cheng et al. ^[53][76] encapsulated Bi nanoparticles in porous conductive graphene frameworks (Bi@3DGFs), which can achieve an extremely long cycle life and rate performance of SIBs. Designing 3D interconnected porous nanostructures can connect active Bi particles, buffer volume expansion, prevent Bi anode collapsing, and generate a steady SEI. Nitrogen-doped carbon could supply added active sites and promote the transport of Na⁺/K⁺ ions. Wang et al. ^[54][77] prepared N-doped carbon/porous Bi nanocomposites (Bi/N-Cs) by simple solution combustion synthesis (SCS). The porous structure can promote the diffusion of Na and electrolytes through the electrode. Sun et al. ^[55][78] directly annealed the 2DBi-MOF precursor and successfully encapsulated Bi nanoparticles in N-doped carbon nanocages (Bi@N-CNCs) with considerable gaps. The K⁺ storage mechanism of Bi@N-CNCs was studied by in situ TEM, SAED, and XRD characterization. Glycol-dimethyl-ether-based (DME) electrolytes and porous nano-networks benefit from achieving rapid kinetics during cycling.

4. Bismuth Alloys

To solve the material crushing problem caused by the increase in the volume of the anodes’ alloy type (Bi, Sb, Si, Pb), another effective method is to make binary ^[56][57][81,82] or ternary alloys ^[58][83], the other metals acting synergistically as a buffer substrate, which could provide abundant active sites, relieve the volumetric strain, and accelerate the diffusion kinetics. Nevertheless, a simple alloying strategy cannot meet the requirements of electrode material properties, and further modifications are needed to enhance the performance of Bi alloys, such as nanostructure design ^[59][84], a coupled 3D conducting carbon matrix, etc. ^[60][61][85,86]. Sb and Bi are the same main group and have similar structures. Alloying Bi and Sb as an anode for PIBs and SIBs can enhance material stability. Among binary alloys, the BiSb alloy is one of the most promising candidates ^[62][87]. Xiong et al. ^[63][88] obtained a Bi-Sb alloy composite nanosheet in which Bi-Sb alloy nanoparticles were evenly dispersed in a porous carbon matrix (BiSb@C). This composite possesses a layered nanosheet structure. Due to the introduction of carbon in composites, BiSb@C effectively inhibits the volume variation during cycles. As an anode for PIBs, this BiSb@C delivered excellent electrochemical performance and provided a highly reversible capacity of 320 mA h g⁻¹ after 600 cycles at 500 mA g⁻¹. By matching the K₄Fe(CN)₆ Prussian blue cathode with the composite bismuth–antimony alloy anode, the full PIB can still deliver a high discharge capacity of 396 mA h g⁻¹ after 70 cycles at 200 mA g⁻¹.