In the pursuit of high-performance energy storage systems, lithium-ion batteries have emerged as a cornerstone technology due to their high energy density and rechargeability. However, the development of advanced anode materials remains a critical challenge to enhance capacity, rate capability, and safety. Graphene, with its two-dimensional structure and excellent electrical conductivity, has shown promise as an anode material, but its pristine form suffers from limited lithium storage capacity and weak interaction with lithium ions. To address this, element doping has been explored as an effective strategy to tailor graphene’s electronic and structural properties. Among various dopants, boron is particularly interesting due to its similar atomic size to carbon and its ability to introduce holes into the graphene lattice, potentially enhancing lithium adsorption and storage. In this work, we employ first-principles calculations based on density functional theory to systematically investigate boron-doped graphene as an anode material for lithium-ion batteries. We focus on the structural stability, adsorption characteristics, diffusion behavior, and lithium storage capacity across a range of boron doping concentrations. Our goal is to provide theoretical insights into optimizing doping levels for improved battery performance, including the suppression of lithium dendrite growth, which is a key safety concern in lithium-ion batteries.
The computational methodology forms the backbone of our investigation. All calculations are performed using density functional theory as implemented in the CASTEP module. We utilize the generalized gradient approximation with the Perdew-Burke-Ernzerhof functional for exchange-correlation effects, and incorporate van der Waals corrections via the Grimme method to account for dispersion interactions. Ultrasoft pseudopotentials are adopted to describe electron-ion interactions. For structural optimizations, we set a plane-wave cutoff energy of 500 eV and a k-point mesh of 4×4×1 in the Brillouin zone. Convergence criteria include forces on atoms less than 0.03 eV/Å, energy changes below 1.0×10⁻⁵ eV/atom, and maximum displacements under 1.0×10⁻³ Å. To model the graphene sheets, we construct supercells with periodic boundary conditions and include a vacuum layer of 20 Å in the z-direction to prevent interlayer interactions. For electronic structure analysis, such as density of states calculations, we increase the cutoff energy to 600 eV and use a finer k-point grid of 6×6×1. The adsorption energy of lithium on boron-doped graphene is calculated using the formula:
$$E_{ad} = E_{LiC_mB_n} – E_{Li} – E_{C_mB_n}$$
where \(E_{LiC_mB_n}\) is the total energy of the lithium-adsorbed system, \(E_{Li}\) is the energy of an isolated lithium atom, and \(E_{C_mB_n}\) is the energy of the boron-doped graphene without lithium. A more negative \(E_{ad}\) indicates stronger adsorption. To assess diffusion properties, we compute the energy barriers for lithium migration on the graphene surface using the linear synchronous transit/quadratic synchronous transit method. The open-circuit voltage and theoretical capacity are derived from total energy calculations at different lithium concentrations, with formulas detailed later. Additionally, we perform ab initio molecular dynamics simulations at 300 K and 500 K to evaluate the thermal stability of selected structures over 4 ps with a time step of 1 fs.
We begin by examining the structural properties of boron-doped graphene. We model various doping concentrations by replacing carbon atoms with boron in graphene supercells, resulting in systems denoted as CmBn, where the boron concentration ranges from 2.08% to 25.00%. Specifically, we study C47B (2.08%), C46B2 (4.17%), C44B4 (8.33%), C28B4 (12.50%), and C24B8 (25.00%). After full relaxation, all structures maintain a planar configuration, indicating good structural stability. The bond lengths between carbon and boron atoms are found to be slightly longer than typical carbon-carbon bonds in graphene, consistent with boron’s larger atomic radius. For instance, in C47B, the C-B bond lengths are around 1.482 Å, compared to the C-C bond length of 1.420 Å in pristine graphene. This minimal distortion suggests that boron doping does not severely disrupt the graphene lattice, which is beneficial for maintaining mechanical integrity during battery cycling. The structural parameters are summarized in Table 1, highlighting the gradual changes with increasing boron content. This stability is crucial for the long-term performance of anode materials in lithium-ion batteries, as repeated lithium insertion and extraction can cause volume expansion and contraction.
| System | Boron Concentration (%) | Average C-C Bond Length (Å) | Average C-B Bond Length (Å) | Li Adsorption Energy (eV) | Distance from Li to Plane (Å) |
|---|---|---|---|---|---|
| C48 | 0.00 | 1.420 | — | -1.82 | 1.759 |
| C47B | 2.08 | 1.423 | 1.482 | -2.15 | 1.677 |
| C46B2 | 4.17 | 1.412 | 1.479 | -2.48 | 1.661 |
| C44B4 | 8.33 | 1.393 | 1.475 | -3.12 | 1.918 |
| C28B4 | 12.50 | 1.414 | 1.492 | -4.25 | 2.061 |
| C24B8 | 25.00 | 1.365 | 1.475 | -5.07 | 2.466 |
Next, we delve into the adsorption properties of lithium on boron-doped graphene. Lithium adsorption is a key step in the charge-discharge process of lithium-ion batteries, as it determines the capacity and cycling stability. We explore typical adsorption sites, including hollow sites (center of hexagons), top sites (above atoms), and bridge sites (above bonds). For boron-doped systems, we find that lithium preferentially adsorbs at hollow sites that contain boron atoms in the hexagon. This preference strengthens with the number of boron atoms in the ring, indicating enhanced lithium-graphene interaction due to boron doping. After adsorption, the structures are optimized, and we observe that as boron concentration increases, the graphene sheet undergoes noticeable deformation, with lithium atoms moving closer to or farther from the plane depending on the doping level. The adsorption energies, calculated using the formula above, show a clear trend: higher boron concentration leads to more negative adsorption energies, signifying stronger lithium binding. For example, at 25.00% boron doping, the adsorption energy is -5.07 eV, which is significantly lower than that of pristine graphene (-1.82 eV). Importantly, all adsorption energies are more negative than the cohesive energy of lithium atoms (-2.74 eV), implying that lithium clustering and dendrite formation are suppressed. This is a critical advantage for safety in lithium-ion batteries, as dendrite growth can lead to short circuits and thermal runaway. The adsorption energies and distances are included in Table 1 for comparison.
To further understand the interaction between lithium and boron-doped graphene, we analyze the electronic properties through differential charge density and electron localization function. Differential charge density, defined as \(\Delta \rho = \rho_{LiC_mB_n} – \rho_{C_mB_n} – \rho_{Li}\), reveals charge redistribution upon lithium adsorption. We observe charge accumulation around boron atoms and depletion around lithium, indicating electron transfer from lithium to the graphene sheet. This transfer is more pronounced at higher boron concentrations, correlating with the stronger adsorption energies. The Mulliken charge analysis shows that lithium atoms lose electrons, with charge loss increasing with boron content. For instance, in C24B8, the lithium atom loses approximately 0.85 electrons, compared to 0.45 electrons in pristine graphene. This charge transfer enhances the ionic character of the Li-CmBn bond, which is confirmed by electron localization function plots. The plots show regions of high electron localization between lithium and graphene, suggesting a mix of ionic and covalent interactions. These electronic changes are beneficial for lithium storage, as they facilitate stronger binding and potentially higher capacity in lithium-ion batteries.
The density of states provides insights into the electrical conductivity of the materials, which is vital for the rate performance of lithium-ion batteries. Pristine graphene exhibits a zero-bandgap semimetallic character, with the valence and conduction bands touching at the Fermi level. Boron doping introduces holes into the graphene lattice, shifting the Fermi level downward and enhancing the density of states near the Fermi level. This results in improved electrical conductivity, as boron acts as a p-type dopant. After lithium adsorption, the systems remain metallic, with the Fermi level shifting upward due to electron donation from lithium. The density of states plots for all systems show significant contributions from carbon p-orbitals and boron p-orbitals, with lithium s-orbitals contributing minimally. This indicates that the graphene framework maintains good conductivity even after lithium intercalation, which is advantageous for fast charge-discharge cycles in lithium-ion batteries. We summarize the key electronic features in Table 2, highlighting the metallicity and charge transfer effects.
| System | Band Character | Charge Transfer from Li (e) | Diffusion Barrier (eV) | Preferred Diffusion Path |
|---|---|---|---|---|
| C48 | Semimetallic | 0.45 | 0.423 | Hollow to Hollow |
| C47B | Metallic | 0.52 | 0.409 | Hollow to Hollow (near B) |
| C46B2 | Metallic | 0.61 | 0.567 | Hollow to Bridge |
| C44B4 | Metallic | 0.73 | 0.892 | Hollow to Top |
| C28B4 | Metallic | 0.79 | 1.124 | Hollow to Hollow (distorted) |
| C24B8 | Metallic | 0.85 | 1.543 | Complex path |
Diffusion behavior of lithium on the graphene surface is another critical factor for the rate capability of lithium-ion batteries. We compute the energy barriers for lithium migration along various paths using transition state search methods. The results show that for low boron concentrations (e.g., 2.08%), the diffusion barrier is slightly lower than that of pristine graphene (0.409 eV vs. 0.423 eV), indicating improved lithium mobility. This suggests that mild boron doping can enhance the rate performance of graphene anodes in lithium-ion batteries. However, as boron concentration increases, the diffusion barriers rise significantly, reaching 1.543 eV at 25.00% doping. This increase is attributed to the structural distortion and stronger lithium binding, which traps lithium atoms and hinders their movement. The diffusion paths often involve lithium moving from hollow sites to bridge or top sites, with transition states where lithium is bonded to multiple atoms. The high barriers at high doping levels may limit the rate capability, but they also contribute to stable lithium adsorption, reducing the risk of dendrite formation. Therefore, a balance must be struck between adsorption strength and diffusion ease for optimal lithium-ion battery performance. The diffusion barriers are listed in Table 2 for comparison.
Now, we turn to the lithium storage properties, specifically the open-circuit voltage and theoretical capacity. The open-circuit voltage is calculated using the formula:
$$V \approx \frac{E_{Li_{x1}C_mB_n} – E_{Li_{x2}C_mB_n} + (x_2 – x_1)E_{Li}}{x_2 – x_1}$$
where \(E_{Li_{x}C_mB_n}\) is the total energy of the system with x lithium atoms adsorbed, and \(E_{Li}\) is the energy per atom in bulk lithium metal. We evaluate voltages at different lithium concentrations for representative doping levels. The results show that for C28B4 (12.50% boron), the voltage curve is relatively flat with an average voltage of 0.35 V, which lies within the desirable range of 0.1–1.0 V for anode materials in lithium-ion batteries. This indicates good reversibility and stable operation. At lower or higher doping concentrations, the voltages deviate, with some cases dropping below 0 V at high lithium loading, suggesting over-lithiation and potential irreversibility. The voltage profiles are illustrated in Figure 1, though we avoid referencing figure numbers as per instructions. Instead, we describe trends: boron doping at 12.50% yields the most favorable voltage characteristics for lithium-ion battery applications.
The theoretical capacity is estimated using the formula:
$$C_M = \frac{x_{max} n F}{M_{GB}}$$
where \(x_{max}\) is the maximum number of lithium atoms adsorbed per formula unit, \(n\) is the valence of lithium (1), \(F\) is Faraday’s constant (26,801 mA·h/mol), and \(M_{GB}\) is the molar mass of boron-doped graphene. We find that the capacity varies with boron concentration, as lithium adsorption sites depend on boron arrangement. At 12.50% doping, the theoretical capacity near boron sites reaches 211.59 mAh/g, which is significantly higher than that of lower doping levels. For example, at 2.08% doping, the capacity is only 46.61 mAh/g. This demonstrates that optimal boron concentration can enhance lithium storage capacity in graphene-based anodes for lithium-ion batteries. However, it’s important to note that these values are specific to boron sites; overall capacity may be higher if other sites are considered, but our focus is on the impact of boron. The capacities are summarized in Table 3, along with the corresponding lithium coverage.
| System | Boron Concentration (%) | Maximum Li Adsorption per B Site | Theoretical Capacity (mAh/g) | Average Open-Circuit Voltage (V) |
|---|---|---|---|---|
| C47B | 2.08 | 1 | 46.61 | 0.45 |
| C46B2 | 4.17 | 1 | 46.69 | 0.38 |
| C44B4 | 8.33 | 2 | 93.71 | 0.31 |
| C28B4 | 12.50 | 2 | 211.59 | 0.35 |
| C24B8 | 25.00 | 2 | 142.56 | 0.28 |
To assess thermal stability, we perform ab initio molecular dynamics simulations on C24B8 with adsorbed lithium at 300 K and 500 K. The results show that the structure remains intact over 4 ps, with no bond breaking or reconstruction. The energy fluctuations are minimal, and the temperature stays close to the target values, confirming high thermodynamic stability. This is crucial for the safety and durability of anode materials in lithium-ion batteries, especially under operating conditions that may involve heat generation.
In conclusion, our first-principles study reveals that boron doping significantly improves the properties of graphene as an anode material for lithium-ion batteries. Key findings include: (1) Boron-doped graphene maintains structural stability across a range of concentrations, with planar configurations preserved. (2) Lithium adsorption strength increases with boron content, and all adsorption energies are lower than lithium’s cohesive energy, indicating effective suppression of dendrite growth—a major safety advantage for lithium-ion batteries. (3) Electronic analysis shows enhanced conductivity due to hole doping and strong lithium-graphene interactions via charge transfer. (4) Diffusion barriers are lower at low boron concentrations, suggesting improved rate performance, but increase at high concentrations, which may balance stability and mobility. (5) The optimal boron concentration for lithium storage appears to be around 12.50%, where the open-circuit voltage is stable and the theoretical capacity is maximized at 211.59 mAh/g near boron sites. These insights provide valuable theoretical guidance for designing boron-doped graphene anodes in lithium-ion batteries, emphasizing the importance of doping level control for optimizing capacity, safety, and rate capability. Future work could explore combined doping strategies or heterostructures to further enhance performance for next-generation lithium-ion batteries.
The implications of this research extend beyond basic science to practical applications in energy storage. By tailoring boron doping concentrations, manufacturers can develop graphene-based anodes that offer higher energy density, longer cycle life, and improved safety in lithium-ion batteries. This aligns with the global push towards electric vehicles and renewable energy integration, where advanced battery technologies are essential. Our computational approach serves as a cost-effective tool for screening materials before experimental synthesis, accelerating the discovery process. We hope that this work inspires further investigations into doped carbon materials for lithium-ion batteries and other energy storage systems.