English

• The Si anode has emerged as one of the most potential anode materials for lithium-ion batteries (LIBs) due to its high capacity and abundant crustal abundance [1–5]. However, the widespread application of Si anode is constrained by several formidable challenges. These include mechanical volume changes during alloying with Li, instability of the solid electrolyte interface (SEI) resulting in irreversible capacity loss [6–8], and difficulties associated with Li-ion migration and desolvation at low temperatures, all of which are particularly difficult to overcome [9]. To address the challenges of Si anodes, a widely used strategy is to combine high-capacity Si with graphite, forming silicon-carbon (Si/C) composites. This approach improves overall performance and stability [10–14]. In Si/C anodes, graphite enhances structural stability, while nanoscale Si reduces pulverization, extending battery cycle life [15–17]. Currently, experimental Si/C anodes LIBs attained remarkable results, including high coulomb efficiency [18,19] and long cycle life [20,21].

  Recently, the interface issues within the Si/C anode have been extensively investigated. Through both theoretical and experimental approaches, anisotropic electrochemical behaviors pertaining to various crystallographic orientations of Si anodes have been investigated and reported, elucidating the superior reversibility and relatively homogeneous kinetics exhibited of the Si₍₁₁₁₎ surface [22,23]. Shen et al. studied the configuration, electronic properties, and Li-intercalation behavior of Si nanoclusters deposited on monolayer graphene [24]. Wang et al. designed and reported the silicene-derived g-SiC₅ material, with first-principles calculations revealing a theoretical capacity of up to 1520 mAh/g, which facilitates high electronic and Li⁺ conductivity, enabling high-rate performance with negligible volume expansion [25]. The Ouyang’s team investigated the influence of the distance between the Si surface and monolayer graphene in the Si/C anode on the lateral migration of Li⁺. As the distance decreased from 4.715 Å to 3.844 Å, the Li⁺ migration barrier also decreased from 0.115 eV to 0.067 eV, and the Li⁺ migration coefficient at 300 K was calculated to be 2.50 × 10⁻⁵ cm²/s [26]. Meanwhile, crystalline plane-oriented etching engineering of graphite anode materials has been utilized in low-temperature studies for LIBs [27]. These research findings provide a solid theoretical foundation for the rational design of Si/C composites as novel anode materials for high-performance LIBs.

  Furthermore, numerous studies have demonstrated that due to the anisotropic crystal growth, the graphite (1010) and (0001) crystal planes exhibit varying degrees of atomic packing density and binding forces along different directions, resulting in distinct physical, electrochemical and mechanical properties [28–30]. While extensive studies have been devoted to the investigation of the Si/C anode, studies on the diverse graphite crystal planes at the Si/C interface remains inadequate. The physical intercalation mechanisms, interfacial effect, and composition of the SEI film at low temperatures for graphite with different crystal planes in contact with Si or Li have not been thoroughly explored.

  In this study, atomic-scale simulations based on density functional theory (DFT) were conducted to gain insights into the intercalation and electrochemical activity within graphite/Li/Si composites featuring different crystal planes. A systematic analysis was performed on the interactions between crystalline silicon (Si₍₁₁₁₎), amorphous silicon (a-Si), and graphite surfaces, specifically the (0001) surface (C₍₀₀₀₁₎) and the (1010) surface (C₍₁₀₁₀₎), with Li, as well as the evolution of the SEI film at low temperatures. Results show that the a-Si/C₍₁₀₁₀₎ anode interface exhibits improved low-temperature performance, electrochemical properties, and mechanical stability compared to Si₍₁₁₁₎/C₍₀₀₀₁₎ after Li intercalation. This work provides valuable insights for designing graphite and Si materials as high-capacity, high-stability electrode composites for LIBs in low-temperature applications.

  All DFT calculations and ab initio molecular dynamics (AIMD) simulations were performed using the Vienna Ab-initio Simulation Package (VASP) software. The Si₍₁₁₁₎ surface with an FCC structure was optimized using the DFT method to form the Si₍₁₁₁₎/C₍₀₀₀₁₎ surface with graphite (0001). The parameters for Si₍₁₁₁₎/C₍₀₀₀₁₎ are a = 13.03 Å, b = 7.53 Å, and c = 15.07 Å, resulting in a lattice mismatch of 4.07%, which is within the acceptable tolerance of 5%. For the C₍₁₀₁₀₎ surface, relaxation was applied to the Si atoms in the 32-atom Si₍₁₁₁₎ configuration. The atoms were randomized from their initial crystalline configuration. After relaxation, the a-Si/C₍₁₀₁₀₎ interface was constructed with graphite (1010), yielding parameters of a = 12.7825 Å, b = 6.8 Å, and c = 35 Å, with a lattice mismatch of only 1.05% (Fig. S1 in Supporting information). The electrochemical stability of both Si/C anode configurations was verified. The stability of Li intercalation at different locations within the Si/C composite material was represented by the Li binding energy (ΔE), which is expressed as [31]:

  $ \Delta E=E_{\mathrm{Si} / \mathrm{C}}+E_{\mathrm{bulk}-\mathrm{Li}}-E_{\mathrm{Li}+\mathrm{Si} / \mathrm{C}} $

  (1)

  where E_Li+Si/C represents the total energy of the structure after Li intercalation into Si₍₁₁₁₎/C₍₀₀₀₁₎ or a-Si/C₍₁₀₁₀₎. E_Si/C denotes the energy of the Si/C structure prior to Li intercalation, and E_bulk-Li is the energy of a single Li atom. For the calculation of charge differences, the POTCAR settings were configured using the Projector Augmented Wave (PAW) method [32] within the Perdew-Burke-Ernzerhof (PBE) approximation [33]. The formula for charge density difference is [34]:

  $ \Delta \rho=\rho_{\mathrm{Li}+\mathrm{Si} / \mathrm{C}}-\rho_{\mathrm{Li}}-\rho_{\mathrm{Si} / \mathrm{C}} $

  (2)

  where ρ_Li+Si/C, ρ_Li, and ρ_Si/C represent the charge densities of the Si/C anode after Li intercalation, the charge density of Li in the same framework, and the charge density of the Si/C anode, respectively. To further investigate the interface evolution process between the composite Si/C anode and electrolyte at low temperatures, the simulation box was filled with ethylene carbonate (EC), ethyl methyl carbonate (EMC), diethyl carbonate (DEC), fluoroethylene carbonate (FEC), and lithium hexafluorophosphate (LiPF₆), with each component being added in accordance with their actual proportions and densities. The optimized structure alleviated intermolecular stresses, yielding a Si/C||electrolyte interface that closely approximates reality.

  The plane-wave cutoff energy of 500 eV was employed, with an energy convergence criterion of 10^–5 eV for electronic self-consistent iterations and 0.01 eV/Å for ionic step convergence. After optimization, a 2 × 3 × 2 K-point grid centered at the Γ-point was used for the Si₍₁₁₁₎/C₍₀₀₀₁₎ structure, and a 2 × 2 × 1 K-point grid centered at the Γ-point with a Gaussian broadening of 0.05 eV was adopted for the a-Si/C₍₁₀₁₀₎ structure.

  For the AIMD studies, the canonical ensemble (NVT) was employed with temperature control via the Nosé-Hoover thermostat. The energy cutoff for plane-wave expansion was set to 450 eV. All structural optimizations adhered to convergence criteria, namely, residual forces <0.01 eV/Å and energy changes per atom <10^–5 eV. A Gaussian smearing method with a broadening parameter of 0.5 was applied. In the Monkhorst-Pack scheme, a 2 × 1 × 1 K-point grid was used for AIMD simulations to sample the Brillouin zone. The simulation time step was set to 1 fs. An electric field of 0.07 eV/Å was applied along the Z-axis to the electrolyte/electrode model to simulate electric field effects related to battery charging [35–37]. Quantitative analysis of electronic charge changes during SEI film growth at low temperatures was performed using Bader charge analysis, which determines the Bader total charge around atoms defined by zero-flux surfaces. The random migration of Li at the Si/C anode interface can be described statistically using the mean square displacement (MSD). The MSD is calculated using the following formula [38]:

  $ M S D=\left\langle\left[r_i\left(t+t_0\right)-r_i\left(t_0\right)\right]^2\right\rangle=\frac{1}{N} \sum\limits_{i=1}^N\left[r_i\left(t+t_0\right)-r_i\left(t_0\right)\right]^2 $

  (3)

  where N represents the total number of Li, the angular brackets denote the statistical average, r_i(t + t₀) and r_i(t₀) are the displacements of the Li at t + t₀ and t₀ times. The diffusion coefficient D_Li at a particular pressure is derived from the slope of the MSD according to the Einstein relation [26].

  $ D=\lim \limits_{t \rightarrow \infty} \frac{M S D}{d t} $

  (4)

  where t represents the total diffusion time and d denotes the dimensionality. To analyze the mechanical stability of Si/C, the separation work of Si/C composites with various structures was calculated, the formula for calculating the separation work is as follows [39]:

  Wsep=EslabSi+EslabC−ESi/CAi

  (5)

  where the total energy of different the lithiation degrees is denoted as E_Si/C, and A_i is the contact area between the graphite layer and the Si layer.

  During the alloying and intercalation reactions of Li in Si/C anodes, the preferred reaction sites are the interlayer positions at the Si/C interface, particularly within graphite hexagonal rings and at graphite C–C bridge sites, followed by the interior of the graphite, and finally alloying with Si, which leads to silicon lattice distortion (Fig. S2 in Supporting information). Our findings align with experimental observations, revealing different lithiation potentials for silicon and graphite in Si/C anodes, with graphite lithiation occurring at 0.1–0.2 V vs. Li/Li⁺ and Si lithiation at 0.3–0.5 V vs. Li/Li⁺. Therefore, during charging, Li intercalation preferentially occurs in the graphite region, followed by the Si region [40]. This further validates the authenticity and rationality of our Si/C interface model.

  The differential charge density analysis was used to investigate changes in charge distribution, and the results are shown in Fig. 1. The cyan regions represent charge depletion, while the yellow regions indicate charge accumulation. Figs. 1a and b clearly show charge transfer at the Si₍₁₁₁₎/C₍₀₀₀₁₎ interface, with Li intercalation into the graphite layer leading to charge accumulation and the formation of strong chemical bonds. In contrast, Figs. 1c and d demonstrate an electron shielding effect at the Si₍₁₁₁₎/C₍₀₀₀₁₎ interface, where Li intercalation into the Si lattice fails to induce interfacial charge redistribution, leading to reduced electron conduction [39].

  Figure 1

  

  Figure 1.  Differential charge density of the Li intercalation at various positions in Si₍₁₁₁₎/C₍₀₀₀₁₎ anode: (a) Top of C; (b) C–C bridge site; (c) within graphite hexagonal ring and (d) center of Si tetrahedron. Charge differences of Li intercalation at various positions in a-Si/C₍₁₀₁₀₎ anode: (e) Top of C; (f) C–C bridge site; (g) graphite hexagonal ring and (h) center of Si tetrahedron.

  DownLoad: Full-Size Img PowerPoint

  For the a-Si/C₍₁₀₁₀₎ anode (Figs. 1e-g), the shielding effect is suppressed due to abundant dangling bonds and defects at the C₍₁₀₁₀₎ interface. Notably, Li intercalation within graphite hexagonal rings establishes clear charge transfer pathways between Si and graphite layers, accompanied by robust covalent bond formation (Fig. 1g). This configuration enables graphite to contribute capacity while dangling bonds enhance electronic conductivity and stability.

  Direct Li intercalation into the Si layer (Figs. 1d and h) also triggers shielding effects. Compared to Si₍₁₁₁₎/C₍₀₀₀₁₎, the a-Si/C₍₁₀₁₀₎ anode possesses a more complex interface, exhibiting lower binding energy upon Li binding, a more stable structure, and more active electrons, which collectively enhance the electronic conductivity of Si-based materials.

  Using AIMD simulations, we investigate the separation work of Si/C interfaces under various degrees of lithiation, thereby exploring the mechanical performance of these interfaces. Additionally, we conduct a detailed study on the Li extraction and intercalation processes at the Si/C anode interface. The snapshots of the Li_0.5Si₍₁₁₁₎/C₍₀₀₀₁₎, LiSi₍₁₁₁₎/C₍₀₀₀₁₎, Li_0.5a-Si/C₍₁₀₁₀₎, and Lia-Si/C₍₁₀₁₀₎, along with corresponding differential charge analyses along the Z-axis and separation work calculations are shown in Fig. 2. Snapshots of the Si₍₁₁₁₎/C₍₀₀₀₁₎ interface in Figs. 2a and b reveal a significant electron shielding effect at the interface. In contrast, snapshots of the a-Si/C₍₁₀₁₀₎ interface in Figs. 2c and d show abundant charge accumulation regions, with electrons having more transmission pathways, indicating superior electronic conductivity of the a-Si/C₍₁₀₁₀₎ anode. Differential charge analysis along the Z-axis further confirms that electrons at the a-Si/C₍₁₀₁₀₎ interface are more active.

  Figure 2

  

  Figure 2.  The lithiation results of Si/C anodes, along with the differential charge distribution along the Z-axis, after 10 ps AIMD simulations for the following configurations: (a) Li_0.5Si₍₁₁₁₎/C₍₀₀₀₁₎; (b) LiSi₍₁₁₁₎/C₍₀₀₀₁₎; (c) Li_0.5a-Si/C₍₁₀₁₀₎ and (d) Lia-Si/C₍₁₀₁₀₎. (e) Values of separation work under various degrees of lithiation.

  DownLoad: Full-Size Img PowerPoint

  Furthermore, the separation work of two Si/C structures under different degrees of lithiation were calculated (Fig. 2e). A higher calculated value indicates greater difficulty in separating the materials and better stability. The results demonstrate that as the separation work increases, the mechanical toughness also increases. The initial separation work values for Si₍₁₁₁₎/C₍₀₀₀₁₎ and a-Si/C₍₁₀₁₀₎ are 0.003 and 0.002 J/m², respectively, suggesting that there is little difference in mechanical properties between the two before lithiation. During the lithiation process, the inserted lithium atoms become positively charged lithium ions, which attract graphite through Coulombic forces, leading to a continuous increase in separation work. After inserting 1 mol of Li into a-Si/C₍₁₀₁₀₎, the separation work value becomes 2.3 times that of Si₍₁₁₁₎/C₍₀₀₀₁₎. Therefore, from a mechanical perspective, the a-Si/C₍₁₀₁₀₎ composite structure can exhibit stronger cycling performance.

  To evaluate the low-temperature performance of Si₍₁₁₁₎/C₍₀₀₀₁₎ and a-Si/C₍₁₀₁₀₎ composite anodes, the Si/C||electrolyte interfaces were constructed. Subsequently, AIMD simulations were performed to study SEI film formation. FEC decomposition was focused on at the anode-electrolyte interface, and we compared our findings with previous work (Fig. 3) [41,42].

  Figure 3

  

  Figure 3.  Evolution of the Si/C||electrolyte interface at low temperatures (253 K and 273 K) via 3 ps AIMD Simulations. Si₍₁₁₁₎/C₍₀₀₀₁₎||electrolyte: (a) Initial structure; (b) After 3 ps at 253 K; (c) After 3 ps at 273 K. a-Si/C₍₁₀₁₀₎||electrolyte interface: (d) Initial structure; (e) After 160 fs at 253 K; (f) After 1.6 ps at 253 K; (g) After 3 ps at 253 K; (h) After 160 fs at 273 K; (i) After 325 fs at 273 K; (j) After 3 ps at 273 K. Color code: Hydrogen (white), silicon (yellow), carbon (gray), oxygen (red), phosphorus (pink), lithium (purple), and fluorine (cyan).

  DownLoad: Full-Size Img PowerPoint

  AIMD simulations of the initial Si₍₁₁₁₎/C₍₀₀₀₁₎||electrolyte structure revealed no FEC decomposition after 3 ps of relaxation at low temperatures (Figs. 3a-c).

  Figs. 3d-g show the initial structure of the a-Si/C₍₁₀₁₀₎ interface and its evolution at 253 K. At 160 fs, Li⁺ ions attack the C_1–O₁ bond in FEC, triggering electron transfer and C₃-F bond cleavage (Fig. 3e). By 1600 fs, FEC undergoes ring-opening via C_3–O₃ bond scission (Fig. 3f), ultimately decomposing into LiF and ·OCOOCH₂CH (Fig. 3g).

  Figs. 3h-j depict the evolution of the a-Si/C₍₁₀₁₀₎ interface at 273 K. Fig. 3h shows that at 160 fs, the C₃-F bond in FEC breaks. Fig. 3i displays that at 325 fs, FEC undergoes ring-opening, and the C_1–O₂ and C_1–O₃ bonds also break, forming a CO group and an ^•OC₂H₃O fragment. This decomposition pattern is consistent with previous reports by Pan [43]. The SEI evolves further under electric fields, forming CH₂OCHOLi, LiF, and CO (Fig. 3j). Therefore, FEC remains confined and does not decompose at the Si₍₁₁₁₎/C₍₀₀₀₁₎ interface, leading to poor low-temperature performance. In contrast, FEC preferentially decomposes at the a-Si/C₍₁₀₁₀₎ interface, protecting the electrolyte solvent. Rich organic components are generated in both SEI films. These organic fragments further polymerize on the SEI film surface to form soluble cross-linked structures, which can inhibit silicon expansion and stabilize the SEI film [41,42,44]. Furthermore, in simulations of SEI film growth at the a-Si/C₍₁₀₁₀₎||electrolyte interface, Si exhibits a maximum volumetric expansion of 110.3%. This value is notably lower than typical Si-based anodes in practical applications, the C₍₁₀₁₀₎ mitigates Si expansion, albeit at the expense of capacity.

  To further investigate the structural characteristics of the a-Si/C₍₁₀₁₀₎ interface, we conducted studies on the MSD of Li, radial distribution functions (RDFs) for C-F and Li-O pairs, and Bader charge analysis of C and O atoms (Fig. 4). By analyzing the MSD of Li at two types of Si/C interfaces, the diffusion coefficients of Li are calculated at various temperatures and interfaces (Fig. 4a). At low temperatures, the diffusion coefficient of Li at the Si₍₁₁₁₎/C₍₀₀₀₁₎ interface is below 10^–15 cm²/s, indicating poor Li migration due to the low temperature and poor interfacial conductivity. At 273 K, the diffusion coefficient of Li at the a-Si/C₍₁₀₁₀₎ interface is 9.19 × 10^–5 cm²/s, while at 253 K, it is 6.32 × 10^–5 cm²/s, which is close to previously reported values [26]. Therefore, the a-Si/C₍₁₀₁₀₎ anode material exhibites better low-temperature performance.

  Figure 4

  

  Figure 4.  Structural characteristics of the Si/C interface: (a) MSD of Li at low temperatures. In the Si₍₁₁₁₎/C₍₀₀₀₁₎||electrolyte interface system: (b) The radial distribution function (RDF) for C-F atoms; In the a-Si/C₍₁₀₁₀₎||electrolyte interface system: RDFs for (c) C-F and (d) Li-O atom Pairs; (e) Charge evolution of C atoms; (f) Charge evolution of O atoms.

  DownLoad: Full-Size Img PowerPoint

  Fig. 4b shows the RDF between C and F atoms at the Si₍₁₁₁₎/C₍₀₀₀₁₎ anode interface, indicating no breakage of the C-F bond. Figs. 4c and d display the RDFs for C-F and Li-O atoms at the a-Si/C₍₁₀₁₀₎ anode interface, respectively, with the breakage of the C-F bond indicating the decomposition of FEC molecules. At 273 K, the first and second g(r) peaks for Li-O pairs were observed at 1.80 Å and 1.88 Å, respectively, while at 253 K, they were located at 1.93 Å and 1.98 Å, with the first peak of initial structure at 1.85 Å. This suggests the initial formation of the SEI film at the a-Si/C₍₁₀₁₀₎ anode interface.

  To further understand the charge evolution during the formation of the SEI film at the low-temperature a-Si/C₍₁₀₁₀₎ anode, Bader analysis was used to investigate the charges surrounding C and O atoms (Figs. 4e and f).

  The FEC molecule’s most reactive site is C₃ (F-bonded), as shown in Fig. 4e. After 3 ps of AIMD simulation at 253 K, C₃ lost 0.1 e, while after 3 ps of AIMD simulation at 273 K, C₃ gains 0.13 e. This corresponds to different decomposition products, with sp³-hybridized C₃ losing electrons during the simulation at 253 K and C₃═O gaining electrons during the simulation at 273 K. Fig. 4f also indicates that different temperatures lead to differences in the composition of the SEI film. At 253 K and 1600 fs, O₁ and O₃ in FEC were observed to gain 0.12 e and 0.11 e, respectively, indicating ring-opening decomposition of FEC, which is consistent with the previous results in Fig. 3f. Fig. 4f also shows that within the first 600 fs, the charge surrounding O₃ increases from −2.07 e to −1.67 e, which is related to the formation of organic components of the SEI film from the decomposition of FEC.

  Compared to pure Si anodes, where uncontrolled SEI growth and mechanical degradation dominate at low temperatures, the a-Si/C₍₁₀₁₀₎ composite leverages its defective carbon matrix to regulate SEI composition and uniformity. The LiF-rich SEI (Figs. 3g and j) enhances ionic conductivity as evidenced in Fig. 4a. Furthermore, unlike conventional graphite anodes, which suffer from Li plating and sluggish kinetics [45–47], the a-Si/C(1010) interface promotes kinetically favorable FEC decomposition, ensuring rapid Li⁺ desolvation and reduced interfacial resistance. This defect-engineered strategy uniquely balances capacity retention and low-temperature adaptability, surpassing the limitations of both pure Si and graphite systems.

  This study investigates the Li insertion behavior and the composition of SEI films at low temperatures for Si₍₁₁₁₎/C₍₀₀₀₁₎ and a-Si/C₍₁₀₁₀₎ anode interfaces using DFT calculations and AIMD simulations. When comparing the a-Si/C₍₁₀₁₀₎ and Si₍₁₁₁₎/C₍₀₀₀₁₎, we observed significant differences. The a-Si/C₍₁₀₁₀₎ interface is abundant in structural defects, which serve as active sites to facilitate Li⁺ adsorption and electron transfer, reducing the activation energy for FEC decomposition. In contrast, the ordered Si₍₁₁₁₎/C₍₀₀₀₁₎ interface lacks such defects, resulting in weak electron localization and a lack of driving force for FEC decomposition. From a kinetic perspective, the defects at the a-Si/C₍₁₀₁₀₎ interface induce Li⁺ migration, accelerating the rupture of the solvation shield and promoting direct intercalation between Li⁺ and FEC, thereby forming a hybrid organic-inorganic bilayer SEI through kinetically favorable pathways. Conversely, the Si₍₁₁₁₎/C₍₀₀₀₁₎ interface inhibits FEC activation, exacerbating interfacial resistance. Furthermore, the defect-rich a-Si/C₍₁₀₁₀₎ interface maintains high FEC decomposition activity at low temperature, whereas the Si₍₁₁₁₎/C₍₀₀₀₁₎ interface, lacking active sites, is temperature-sensitive, resulting in SEI passivation. Therefore, these different SEI formation mechanisms originate from the interplay between a-Si/C₍₁₀₁₀₎ and Si₍₁₁₁₎/C₍₀₀₀₁₎. Defect-rich interfaces enable targeted FEC decomposition, SEI stabilization, and enhanced low-temperature performance, highlighting the critical role of atomic-level interface engineering in Si/C anode design.

  The differential charge density analysis highlights that interfacial atomic configuration critically govern charge transfer and electron shielding effects in Si/C composites. The weakened shielding at defective a-Si/C₍₁₀₁₀₎ interfaces, driven by dangling bonds, enables robust covalent bonding and enhanced electron conduction. This underscores the importance of tailoring defect-rich interfaces to mitigate electron shielding, optimize Li⁺ pathways, and stabilize composite structures. These insights provide a design principle for high-performance Si-based anodes: engineering disordered or defective carbon matrices to synergistically improve conductivity, capacity, and structural integrity.

  Furthermore, after Li intercalation into the a-Si/C₍₁₀₁₀₎ structure, the separation work significantly increases, being 2.3 times that of the Si₍₁₁₁₎/C₍₀₀₀₁₎ anode. Consequently, the a-Si/C₍₁₀₁₀₎ anode exhibits more stable cycling performance. More importantly, simulations of the Si₍₁₁₁₎/C₍₀₀₀₁₎||electrolyte interface at low temperatures reveal no formation of SEI film components. However, the a-Si/C₍₁₀₁₀₎||electrolyte interface, after undergoing AIMD simulations, produces essential SEI film components such as LiF and CH₂OCHOLi, with more pronounced lithium-ion diffusion behavior at the interface. This suggests superior low-temperature performance for the a-Si/C₍₁₀₁₀₎ anode.

  To guide future designs, we propose three optimization strategies: (1) Defect engineering to amplify interfacial charge delocalization; (2) Surface functionalization of graphite to stabilize LiF-rich SEI; (3) Hybrid interface design. Combining amorphous Si with defective carbon surfaces promotes hybridization of electronic states. Future studies should prioritize synthesis of defect-rich C₍₁₀₁₀₎ matrices, in situ characterization of SEI evolution at subzero temperatures, and machine learning-guided interface design to improve ionic/electronic transport. These efforts will bridge theoretical insights with practical Si/C anode manufacturing. In summary, Si/C composites with the graphite (1010) surface is a promising anode material that enhances the low-temperature stability, electronic conductivity, and mechanical stability of lithium-ion batteries.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  He Ma: Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Kang Liang: Methodology, Formal analysis, Data curation. Weizong Wang: Investigation, Data curation. Zhengping Ding: Writing – review & editing, Resources, Project administration, Methodology, Investigation, Funding acquisition. Lin Shi: Writing – review & editing. Yurong Ren: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition. Qian Zhao: Conceptualization, Methodology, Writing – review & editing, Software, Project administration, Investigation, Funding acquisition, Data curation. Shoukun Xu: Writing – review & editing.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. U22A20420, 52203292, 52203291), the Program for International S&T Cooperation Projects of Changzhou City (No. CZ20240026).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111125.

  
  
• 1. [1]

     N. Kim, Y. Kim, J. Sung, et al., Nat. Energy 8 (2023) 921–933. doi: 10.1038/s41560-023-01333-5

  2. [2]

     T. Mae, K. Kaneko, H. Sakurai, et al., Carbon 218 (2024) 118663. doi: 10.1016/j.carbon.2023.118663

  3. [3]

     B. Xue, H. Liu, X. Wu, J. Energy Storage 77 (2024) 109979. doi: 10.1016/j.est.2023.109979

  4. [4]

     G.G. Eshetu, H. Zhang, X. Judez, et al., Nat. Commun. 12 (2021) 5459. doi: 10.1038/s41467-021-25334-8

  5. [5]

     Q. Shi, J. Zhou, S. Ullah, et al., Energy Storage Mater. 34 (2021) 735–754. doi: 10.1016/j.ensm.2020.10.026

  6. [6]

     S. Solchenbach, X. Huang, D. Pritzl, et al., J. Electrochem. Soc. 168 (2021) 110503. doi: 10.1149/1945-7111/ac3158

  7. [7]

     C. Bommier, W. Chang, J. Li, et al., J. Electrochem. Soc. 167 (2020) 020517. doi: 10.1149/1945-7111/ab68d6

  8. [8]

     X. Li, M. Tabish, W. Zhu, et al., Small 19 (2023) 2302388. doi: 10.1002/smll.202302388

  9. [9]

     G.A. Collins, H. Geaney, K.M. Ryan, J. Mater. Chem. A 9 (2021) 14172–14213. doi: 10.1039/d1ta00998b

  10. [10]

      M. Gautam, G.K. Mishra, M. Furquan, et al., Chem. Eng. J. 472 (2023) 144916. doi: 10.1016/j.cej.2023.144916

  11. [11]

      N. Yang, J. Sun, R. Shao, et al., Cell Rep. Phys. Sci. 3 (2022) 100862. doi: 10.1016/j.xcrp.2022.100862

  12. [12]

      X. Lin, L. Yu, Y. Wen, et al., Cell Rep. Phys. Sci. 4 (2023) 101730. doi: 10.1016/j.xcrp.2023.101730

  13. [13]

      N. Xu, Y. Sun, J. Shi, et al., J. Power Sources 511 (2021) 230437. doi: 10.1016/j.jpowsour.2021.230437

  14. [14]

      W. Chen, S. Kuang, H. Wei, et al., J. Colloid Interface Sci. 610 (2022) 583–591. doi: 10.1016/j.jcis.2021.11.099

  15. [15]

      L. Shen, C. Xu, J. Gao, et al., J. Energy Chem. 77 (2023) 348–358. doi: 10.1016/j.jechem.2022.10.044

  16. [16]

      A.S. Cattaneo, S. Dupke, A. Schmitz, et al., Solid State Ion. 249-250 (2013) 41–48. doi: 10.1016/j.ssi.2013.07.013

  17. [17]

      K. Richter, T. Waldmann, M. Kasper, et al., J. Phys. Chem. C 123 (2019) 18795–18803. doi: 10.1021/acs.jpcc.9b03873

  18. [18]

      J. Zhao, B. Rui, W. Wei, et al., J. Colloid Interface Sci. 607 (2022) 1562–1570. doi: 10.1016/j.jcis.2021.09.005

  19. [19]

      Z. Shi, Q. Liu, Z. Yang, et al., J. Mater. Chem. A 10 (2022) 1380–1389. doi: 10.1039/d1ta07112b

  20. [20]

      J. Li, Z. Li, W. Huang, et al., Small 15 (2019) 1900436. doi: 10.1002/smll.201900436

  21. [21]

      B. Nie, D. Sanchez, M. Alcoutlabi, et al., Energy Adv. 2 (2023) 1381–1389. doi: 10.1039/d3ya00203a

  22. [22]

      J. Rohrer, P. Kaghazchi, Chem. Phys. Chem. 15 (2014) 3950–3954. doi: 10.1002/cphc.201402456

  23. [23]

      Y. Shi, J. Wan, J. Li, et al., Nano Energy 61 (2019) 304–310. doi: 10.1016/j.nanoen.2019.04.074

  24. [24]

      D. Shen, Y.H. Liu, S.W. Tang, et al., Acta Phys. Sin. 70 (2021) 198101. doi: 10.7498/aps.70.20210521

  25. [25]

      H. Wang, M. Wu, X. Lei, et al., Nano Energy 49 (2018) 67–76. doi: 10.1016/j.nanoen.2018.04.038

  26. [26]

      G. Wang, B. Xu, J. Shi, et al., Appl. Surf. Sci. 436 (2018) 505–510. doi: 10.1016/j.apsusc.2017.11.237

  27. [27]

      J. Xu, X. Wang, B. Hu, et al., Batteries Supercaps 6 (2023) e202200499. doi: 10.1002/batt.202200499

  28. [28]

      A.P. Fu, D.C. Feng, C.H. Deng, Chem. J. Chin. Univ. 19 (1998) 792.

  29. [29]

      C. Han, Z. Chen, B. Li, Int. J. Heat Mass Transf. 179 (2021) 121675. doi: 10.1016/j.ijheatmasstransfer.2021.121675

  30. [30]

      T. Badzian, A. Badzian, R. Roy, et al., Appl. Phys. Lett. 76 (2000) 1125–1127. doi: 10.1063/1.125958

  31. [31]

      A.K. Geim, I.V. Grigorieva, Nature 499 (2013) 419–425. doi: 10.1038/nature12385

  32. [32]

      P.E. Blöchl, C.J. Först, J. Schimpl, Bull. Mater. Sci. 26 (2003) 33–41. doi: 10.1007/BF02712785

  33. [33]

      J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865–3868. doi: 10.1103/PhysRevLett.77.3865

  34. [34]

      D. Massa, E. Kaxiras, S. Papanikolaou, Acta Mater 268 (2024) 13673.

  35. [35]

      M. Gajdoš, K. Hummer, G. Kresse, et al., Phys. Rev. B 73 (2006) 045112. doi: 10.1103/PhysRevB.73.045112

  36. [36]

      J. Neugebauer, M. Scheffler, Phys. Rev. B 46 (1992) 16067–16080.

  37. [37]

      L. Bengtsson, Phys. Rev. B 59 (1999) 12301–12304. doi: 10.1103/PhysRevB.59.12301

  38. [38]

      W. Zhang, Y. Li, Y. Qi, et al., Energy Procedia 153 (2018) 125–130. doi: 10.1016/j.egypro.2018.10.014

  39. [39]

      H. Chang, M. Park, J. Kim, et al., J. Electroanal. Chem. 931 (2023) 117179.

  40. [40]

      G. Berckmans, L. De Sutter, M. Marinaro, et al., Electrochim. Acta 306 (2019) 387–395.

  41. [41]

      K. Leung, J.L. Budzien, Phys. Chem. Chem. Phys. 12 (2010) 6583–6586. doi: 10.1039/b925853a

  42. [42]

      A.L. Michan, M. Leskes, C.P. Grey, Chem. Mater. 28 (2015) 385–398.

  43. [43]

      S.H. Pan, S. Nachimuthu, B.J. Hwang, et al., Colloid Interface Sci. 649 (2023) 804–814.

  44. [44]

      Y. Jin, N.J.H. Kneusels, P.C.M.M. Magusin, et al., J. Am. Chem. Soc. 139 (2017) 14992–15004. doi: 10.1021/jacs.7b06834

  45. [45]

      Y. Wang, J. Liu, H. Ji, et al., Adv. Mater. 37 (2025) 2412155.

  46. [46]

      Y. Huang, C. Wang, H. Lv, et al., Adv. Mater. 36 (2024) 2308675.

  47. [47]

      C. Wei, H. Fei, Y. Tian, et al., Chem. Lett. 31 (2020) 980–983

• 

  Figure 1  Differential charge density of the Li intercalation at various positions in Si₍₁₁₁₎/C₍₀₀₀₁₎ anode: (a) Top of C; (b) C–C bridge site; (c) within graphite hexagonal ring and (d) center of Si tetrahedron. Charge differences of Li intercalation at various positions in a-Si/C₍₁₀₁₀₎ anode: (e) Top of C; (f) C–C bridge site; (g) graphite hexagonal ring and (h) center of Si tetrahedron.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  The lithiation results of Si/C anodes, along with the differential charge distribution along the Z-axis, after 10 ps AIMD simulations for the following configurations: (a) Li_0.5Si₍₁₁₁₎/C₍₀₀₀₁₎; (b) LiSi₍₁₁₁₎/C₍₀₀₀₁₎; (c) Li_0.5a-Si/C₍₁₀₁₀₎ and (d) Lia-Si/C₍₁₀₁₀₎. (e) Values of separation work under various degrees of lithiation.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Evolution of the Si/C||electrolyte interface at low temperatures (253 K and 273 K) via 3 ps AIMD Simulations. Si₍₁₁₁₎/C₍₀₀₀₁₎||electrolyte: (a) Initial structure; (b) After 3 ps at 253 K; (c) After 3 ps at 273 K. a-Si/C₍₁₀₁₀₎||electrolyte interface: (d) Initial structure; (e) After 160 fs at 253 K; (f) After 1.6 ps at 253 K; (g) After 3 ps at 253 K; (h) After 160 fs at 273 K; (i) After 325 fs at 273 K; (j) After 3 ps at 273 K. Color code: Hydrogen (white), silicon (yellow), carbon (gray), oxygen (red), phosphorus (pink), lithium (purple), and fluorine (cyan).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Structural characteristics of the Si/C interface: (a) MSD of Li at low temperatures. In the Si₍₁₁₁₎/C₍₀₀₀₁₎||electrolyte interface system: (b) The radial distribution function (RDF) for C-F atoms; In the a-Si/C₍₁₀₁₀₎||electrolyte interface system: RDFs for (c) C-F and (d) Li-O atom Pairs; (e) Charge evolution of C atoms; (f) Charge evolution of O atoms.

  下载: 全尺寸图片 幻灯片
•