Sodium-affinitive coatings improve the sodiophilicity of substrates by employing materials that can form alloys with sodium as the base coating, thereby reducing the nucleation barrier while increasing the number of nucleation sites. Researchers have conducted extensive studies on the sodiophilic properties of Au, Sn, Sb, Cu, Cr, Bi and Mo. Results indicate that among these elements, Au, Sn, Bi and Sb can form alloys with Na, whereas Cu, Cr and Mo cannot^[60]. However, due to their high cost, elements such as Au are difficult to apply in industrial production. Yu et al.^[61] constructed a Bi-modified layer on copper foil through a simple liquid-phase in-situ reduction reaction (see Figure 2c). After activation treatment, the Bi nanoparticles transformed into Na-rich alloys with sodiophilic properties (Cu@Bi). As shown in Figures 2a and 2b, under conditions of 1 mA·cm^-2 and 1 mAh·cm^-2, Cu@Bi achieved a long lifespan of 700 h with an average CE as high as 99.5%. During the initial stage of sodium deposition, Cu@Bi exhibited a nucleation overpotential of only 3.5 mV, significantly lower than that of copper foil. Cohn et al.^[28] constructed a nano-carbon nucleation layer using conductive carbon black and sodium carboxymethyl cellulose on aluminum foil. Compared with aluminum foil, the nucleation overpotential was reduced from 19 mV to 12 mV, indicating a lower nucleation barrier. Thanks to the presence of highly reactive sp³ carbon sites and oxygen-containing functional groups, the deposited morphology of sodium ions exhibited a uniform block structure. Over 1000 deposition-stripping cycles on this current collector revealed low voltage hysteresis (average 14 mV) and a high average CE of 99.8%. In addition to single coatings, composite coatings have also attracted significant attention. Our research group^[64] proposed a co-sputtering strategy to construct a Cu₂Sb composite coating. The Cu₂Sb coating creates a sodiophilic and stable current collector–sodium interface by suppressing alloying transformation and alleviating volume changes in the coating. Compared with unstable Sb coatings, Cu₂Sb alloy coatings with low antimony content retain the sodiophilic Cu₂Sb (2 0 0) crystal plane and form low-expansion and irreversible sodium alloys during cycling, maintaining structural stability and sodiophilicity of the coating, inducing the formation of a thin and dense fluorinated SEI layer, greatly enhancing the uniformity of sodium deposition (see Figure 2d, f) and improving ion transport at the interface. As shown in Figure 2e, after sodium stripping, no residual sodium signal was observed on the Cu₂Sb@Cu substrate, whereas a strong Na (1 1 0) residual signal was detected on the Cu foil, indicating that Cu₂Sb@Cu effectively prevents the formation of dead sodium, unlike the Cu foil which produces significant amounts of dead or low-activity sodium. As shown in Figures 2h and 2i, Cu₂Sn@Cu||Na₃V₂(PO₄)₃ (NVP) AFSBs retained a discharge capacity of 75.4 mAh·g^-1 after 600 cycles with a capacity retention rate of 74.2%, significantly higher than untreated copper foil (capacity retention rate less than 30% after 200 cycles). Mitlin et al.^[66] prepared a novel ternary sodium-antimony-tellurium alloy composite, Na₂(Sb_2/6Te_3/6Vac_1/6), known as NST-Na, dispersed within electrochemically active metallic sodium via repeated cold rolling and folding (see Figure 2i). This new alloy compound possesses a thermally durable face-centered cubic structure and abundant vacancies. Its surface consists of Na atoms (rather than clusters), thus enhancing planar wetting ability and thermodynamic stability. Additionally, its structural stability during cycling effectively prevents self-degradation. An anode-free NST||Na full battery delivered a stable reversible specific capacity after 100 cycles at a current density of 1 C (see Figure 2j). By balancing the alloying capability and stability of the substrate material, or by controlling the content of sodiophilic substrate materials^[79-80], the volume expansion effect of the substrate during sodium deposition can be suppressed, effectively ensuring the long-term stability of the substrate^[64].

In the field of composite carbides, Ma et al.^[42] anchored Sn₄P₃ nanocrystals on a reduced graphene oxide (rGO) framework to regulate Na deposition (Figure 3a). Based on galvanostatic electrochemical tests, the Na/Sn₄P₃@rGO symmetric battery exhibited a long lifespan of up to 500 hours with an extremely low overpotential (<15 mV), at a current density of 1 mA·cm^-2 and a deposition capacity of 2 mAh·cm^-2, along with minimal nucleation overpotential (4 mV), demonstrating that Sn₄P₃ with uniform sodiophilic properties can significantly reduce the nucleation barrier and deposition overpotential of Na. Under mechanical bending conditions, stable cycling for 150 cycles with a capacity retention rate of 95.6% was achieved in a full battery with 1.1 mAh (Figure 3b), while also exhibiting excellent rate performance (Figure 3c). Passerini et al.^[65] utilized MOF-derived Cu-C composites, as shown in Figure 3d, to construct nucleation buffer layers on conventional current collectors. The abundant nucleation sites provided by the composite materials guided uniform Na deposition, and the carbon framework provided void volume to suppress volume changes during sodium plating and stripping processes. AFSBs using Al-Cu@C as the anode and carbon-coated NVP as the cathode exhibited a high CE value of up to 99.5% and long cycle life (Figure 3e).

It is worth noting that the sodium-philic coating strategies often require complex material synthesis and coating procedures. In addition, excessive use of coating materials or polymer binders may reduce the energy density of AFSBs. Modifying traditional current collectors themselves by using other metals or alloys has the advantage of adapting to existing manufacturing facilities, showing promising prospects for large-scale applications. Cheng et al.^[27] found that molten sodium on zinc foil exhibited a minimum contact angle of ~32°, in stark contrast to copper's sodium-repelling characteristic with a contact angle reaching ~115° (Fig. 4a, b). Studies revealed that nearly completely reversible Na deposition/stripping could be achieved on the zinc surface, where Zn current collectors formed a NaZn₁₃ alloy interface with initially deposited metallic Na, guiding subsequent Na deposition/stripping behavior. During 500 cycles at 3.0 mAh·cm^-2, the efficiency remained almost consistently above 99.9%, and dead sodium formation on the Zn surface was negligible. At 0.5 C, about 90% capacity retention was maintained after 100 cycles (Fig. 4c, d). Matsumoto et al.^[63] prepared aluminum current collectors through sequential annealing and fluorination processes (Fig. 4e), optimizing their crystal orientation and surface properties, thereby achieving a highly reversible sodium deposition/stripping process. Using density functional theory calculations, the binding energies of Na atoms adsorbed on Al (1 1 1), Al (1 0 0), oxidized Al (1 0 0), and fluorinated Al (1 0 0) planes were determined as -1.18, -1.37, -1.54, and -2.02 eV, respectively (Fig. 4f). Compared to other substrates, fluorine-treated Al (1 0 0) showed lower binding energy, indicating that annealing and fluorination treatments on the Al (1 0 0) plane played a crucial role in enhancing substrate sodium-philic properties. As shown in Fig. 4g, a full battery composed of this aluminum foil and an NVP cathode achieved high average Coulombic efficiency (~98%) over 50 cycles. Constructing sodium-philic coatings is a powerful measure to address uneven deposition at the current collector-sodium interface; however, considering that most sodium-philic coating designs typically provide only one-dimensional frameworks for sodium metal, structural instability of the substrate adversely affects the long-term cycling performance of batteries as cycling reactions proceed.

This strategy was initially inspired by porous carbon materials, such as graphene and carbon nanotubes. Their high porosity significantly increases the substrate surface area, reduces the local current density, and provides abundant pore structures that serve as additional nucleation sites, effectively accommodating the volume expansion of sodium metal during deposition. This design not only alleviates mechanical stress caused by volume changes but also promotes uniform sodium deposition, suppressing dendrite formation^[66-67,84]. To date, various methods have been developed to construct porous skeleton structures. Jiang et al.^[44] fabricated a three-dimensional porous copper current collector using a simple chemical dealloying method (see Figure 5b). Compared with traditional two-dimensional copper foil, this structure significantly suppresses dendrite formation (see Figure 5a). After 400 cycles at a current density of 1 mA·cm^-2, it exhibits a high cycling efficiency of 99.4% and voltage hysteresis below 20 mV (see Figures 5c and 5d). The superior performance is attributed to the larger specific surface area and excellent electrolyte wettability of the three-dimensional porous copper current collector, which facilitates reducing the actual current density and inhibiting dendrite growth. Liu et al.^[67] constructed Cu₃P nanowires on copper foil via an in situ growth approach as substrates for sodium deposition (see Figure 5e). Cryo-electron microscopy results demonstrate that the Cu₃P component successfully transforms into nano-Cu and Na₃P, subsequently forming sodium nuclei. First-principles calculations reveal that Na₃P possesses low sodium adsorption energy and diffusion barriers, and the Cu₃P@Cu structure demonstrates favorable sodium deposition morphology. An anode-free NVP||Cu₃P@Cu full battery delivers a high specific capacity of 76.1 mAh·g^-1 after 75 cycles at 0.5 C, with a cycling efficiency reaching 99.5% (see Figures 5f and 5g). A pouch cell shows an initial discharge capacity of 29 mAh at 0.5 C, decreasing to approximately 15 mAh after 30 cycles and remaining stable even after 170 cycles (see Figure 5h).

In addition to structural optimization of copper foil, porous structures can also be constructed on its surface to increase sodium nucleation sites. Wang et al.^[69] fabricated a porous fluorinated covalent triazine framework (FCTF) on the current collector surface, which provided excellent sodiophilicity and reduced Na⁺ consumption during cycling. This could be attributed to the highly ordered layered channels and periodic structure that lower the nucleation overpotential and enable crack-free Na deposition, avoiding excessive consumption of active sodium (see Figure 6a). To evaluate the practical feasibility of FCTF, an assembled pouch cell exhibited 300 stable cycles with nearly no capacity decay under low N/P ratio (1.5), high current density (2 C), and high cathode loading, achieving an energy density of 325 Wh·kg^-1 and outstanding rate performance (see Figures 6b, c). Compared with metallic current collectors, lightweight carbon-based current collectors also represent a significant advantage. Cao et al.^[68] reported a lightweight three-dimensional carbon current collector obtained through a fungus-assisted biosynthesis method. As shown in Figure 6d, the authors first inoculated basswood with fungi, followed by selective lignin skeleton etching via hydroxyl free radical oxidation reactions spontaneously conducted by the fungi. This fungus-treated basswood (FBW) could be converted into a self-supporting carbon electrode with short-range order after heat treatment (FBWC). The current collector features vertically aligned symmetric channels and high porosity, which reduce local current density and enable uniform sodium deposition. The Na||FBWC asymmetric battery demonstrated stable plating/stripping for over 4500 h with an average CE reaching 99.5%. Although the authors did not present full battery performance, this work provides a new solution for constructing porous skeletal structures for novel AFSBs.

The design principles for a stable artificial SEI include: (1) effectively blocking the reaction between the electrolyte and metallic sodium; (2) possessing good ionic conductivity to facilitate ion transport; (3) having a high elastic modulus to accommodate volume changes during cycling and suppress dendrite growth^[75-76]. To date, various inorganic or organic-based artificial SEIs have been reported in the field of sodium metal batteries, including graphene^[48], poly(vinylidene fluoride) (PVdF)^[36], metal-organic framework materials^[73], carbon nanotube paper^[45], metal oxide layers, and some alloy-based artificial layers^[34,74], which can provide directional guidance for AFSBs. Zhang et al.^[39] constructed an organometallic artificial layer composed of PVdF and Bi, achieving smooth and dense sodium deposition (Figure 7a). The authors also demonstrated that Bi in the artificial SEI layer can form Na₃Bi with sodium during the first deposition process. In addition to directing the sodium ion flux, Na₃Bi can help enhance the mechanical properties of the film. Moreover, the flexible PVdF can suppress volume expansion effects during deposition-stripping cycles and promote the formation of a NaF-rich SEI. At 1 mA·cm^-2, the cycle life of the PB@Cu current collector was extended to ~2500 h, showing a high average CE of 99.92%, which contrasts sharply with the fluctuating CE obtained using a copper foil current collector (Figure 7b). PB@Cu AFSBs assembled with an NVP cathode achieved a stable capacity of >90 mAh·g^-1 over 150 cycles, significantly higher than the 37.2 mAh·g^-1 obtained for NVP||Cu at the 150^th cycle (Figure 7c). The electrochemical performance of the PB@Cu current collector originates from the alloyed Na₃Bi phase with high ionic conductivity and mechanical strength, as well as the formation of a NaF-rich SEI and rapid sodium ion migration, thereby enabling dendrite-free morphology of sodium deposition at the sodium-electrolyte interface. Furthermore, based on density functional theory (DFT) calculations and ab initio molecular dynamics (AIMD) simulations, Yu et al.^[71] predicted that Na₂Te exhibits a low Na⁺ diffusion barrier and a high diffusion coefficient; thus, they designed a Na₂Te protective layer with high ionic conductivity, low electronic conductivity, and high mechanical stability. Experimental results showed that Na₂Te used as a protective layer can increase the exchange current density in the SEI, reduce the activation energy for sodium migration, and successfully inhibit the growth of sodium dendrites. The Na@Na₂Te anode could sustainably cycle for 700 h (1 mA·cm^-2, 1 mAh·cm^-2) in a low-cost carbonate electrolyte, providing beneficial insights into the design of artificial SEI layers for AFSBs. Organic artificial SEI interfaces exhibit high chemical stability and low sodium ion diffusion barriers; their application in the field of AFSBs not only acts as a robust SEI promoting dendrite-free sodium deposition but also effectively isolates the sodium metal from the liquid electrolyte, reducing irreversible sodium loss during cycling. Zheng et al.^[70] adopted a method of modifying the copper foil surface with formate. As shown in Figure 7d, during the first deposition process, the coordinated formate ligands on the Cu surface could spontaneously convert into HCOONa, and the formation of an HCOONa interface on the SF-Cu foil was confirmed using nuclear magnetic resonance (NMR) spectroscopy and surface-enhanced Raman scattering (SERS) spectroscopy. This interface possesses high chemical stability and a low sodium ion diffusion barrier, promoting uniform sodium deposition. Meanwhile, it functions as a robust SEI separating deposited sodium from the liquid electrolyte, reducing irreversible sodium loss (Figure 7e), thereby enhancing the rate capability and cycling stability of the battery. AFSBs composed of SF-Cu||NVP maintained a high capacity retention rate of 88.2% after 400 cycles, with a capacity decay of only 0.03% per cycle and an average CE remaining at 99.97% (Figure 7f).

The in-situ optimization design of the SEI layer mainly refers to adjusting the electrolyte formulation or adding functional additives, utilizing the spontaneous reaction between sodium metal and the modified electrolyte, to guide the formation of a robust SEI layer and achieve uniform deposition of sodium.

Huang et al.^[41] added a small amount of SnCl₂ additive into a conventional carbonate-based electrolyte, achieving in situ formation of Na-Sn alloy phases and a dense NaCl-rich SEI layer, thereby enabling long-term cycling of sodium metal anodes in conventional carbonate electrolytes. Upon contact with sodium metal, due to its high reductive nature, Sn²⁺ is chemically reduced to Sn, which subsequently alloys with bulk Na to form an alloy interlayer (Figure 8b). Meanwhile, the anion Cl^- facilitates the formation of a thin and dense NaCl-rich SEI instead of dendritic or mossy SEI layers (Figure 8a), allowing rapid interfacial ion transfer. As shown in Figure 8c, after 500 h of cycling, the electrode dendrite deposition morphology and voltage hysteresis are significantly reduced in Na||Na symmetric batteries. In the field of AFSBs, Zhang et al.^[71] developed a novel electrolyte composed of 1 mol·L^-1 NaPF₆ and a solvent mixture of diglyme (G2) and tetraglyme (G4) at a volume ratio of 9:1, named NG24₁₀. By optimizing the volumetric ratio of two common linear ether solvents, the cation solvation structure formed by the binary electrolyte promotes flat and dendrite-free planar growth of Na metal on the anode current collector. The SEI formed by unmodified electrolytes typically contains a high proportion of Na₂O/Na₂O₂, often resulting in a porous structure and low mechanical strength under the Kirkendall effect. As shown in Figure 8d, converting Na_xO into Na₂CO₃ can suppress the Kirkendall effect and facilitate the formation of a robust and stable SEI, thus extending battery cycle life. Simultaneously, at 1 mol/L NaPF₆, the electrolyte promotes "co-symbiosis" of NaF-rich SEI during Na metal electrochemical deposition on Al-C current collectors, effectively passivating deposited Na metal and enabling dendrite-free planar epitaxial metal growth. Al-C||NG24₁₀||P2-NCO (carbon-coated aluminum foil as current collector, 10 wt% Na₂C₂O₄ (NCO) incorporated into P2 as cathode) AFSBs retained up to 97% capacity retention after 80 cycles with high CE (Figure 8f). As illustrated in Figure 8g, a 40 mAh pouch cell exhibited highly stable charge-discharge curves, achieving an energy density of up to 180 Wh·kg^-1. Li et al.^[73] demonstrated that by adjusting the electrolyte formulation using an electrolyte composed of 0.9 M NaPF₆ and 0.1 mol/L NaBF₄ (BPG), decomposition products of flaky NaBF₄ displayed a two-dimensional distribution within the SEI, effectively suppressing the formation of dead sodium or sodium dendrites during sodium nucleation growth, and repairing cracks generated during sodium deposition and stripping processes. As shown in Figure 8h, Ah-level cylindrical AFSBs employing BPG electrolyte achieved a cycle life of 260 cycles, delivering high energy densities of 180 Wh·kg^-1 (0.91 Ah) and 205 Wh·kg^-1 (1.03 Ah) within a voltage range of 2-4 V, comparable to the energy density of lithium iron phosphate-based lithium-ion batteries. Experiments proved that dual-sodium/multi-sodium salt electrolytes contribute to forming more stable SEI layers, effectively guiding uniform sodium deposition and stripping at the sodium-electrolyte interface, providing a key strategy for constructing long-life AFSBs.