Skip to main content
ACS Physical Chemistry Au logoLink to ACS Physical Chemistry Au
. 2025 Jun 25;5(5):478–489. doi: 10.1021/acsphyschemau.5c00025

Quantum Monte Carlo Approaches to Na Intercalation on Bilayer Graphene

Hyeondeok Shin †, Anouar Benali †, Christopher S Johnson ‡,*
PMCID: PMC12464765  PMID: 41019616

Abstract

We have performed Quantum Monte Carlo (QMC) simulations on Na-intercalated bilayer graphene to study the evolution of electronic and optical properties upon Na intercalation into hard carbon layers. The objective was to model the optimal configuration of Na intercalation into a hard carbon matrix containing graphene regions. Our study showed that Na intercalation can be energetically stabilized at large interlayer distances (over 6 Å) in both AA- and AB-stacked bilayer graphene. In the QMC results, we found a significant band gap opening at the equilibrium interlayer distance of Na-intercalated bilayer graphene, while corresponding density functional theory (DFT) results showed no gap. This difference between DFT and QMC results indicates that the gap opening induced by Na intercalation into a hard carbon is underestimated within the DFT framework. In addition, a zigzag configuration of Na atoms was found to be energetically stable at interlayer distances up to 10 Å, leading us to predict the existence of a local minimum of Na intercalation at large interlayer distance. These computation and modeling results can provide guidance on how to synthesize and optimize hard carbon with bilayer graphene regions that permit a zigzag intercalation configuration that will maximize and stabilize sodium hosting.

Keywords: Graphene, Li intercalation, Na intercalation, first-principles calculations, QMC, DFT


graphic file with name pg5c00025_0015.jpg


graphic file with name pg5c00025_0014.jpg

Introduction

In the past few decades, rechargeable alkali-ion battery materials have been the focus of extensive research to discover next-generation alternatives to Li-ion batteries. − Such alternative materials have been playing an increasingly important role in energy storage technologies for assisting in diversity of sources in battery applications. − The search for alternatives is motivated by many factors, including large fluctuations in the price of lithium salts, rapid growth for the energy storage field, and supply chain stability and accessibility in the North American market.

Among the various alkali-ion battery technologies being considered, the Na-ion chemistry has attracted a great deal of attention as a promising alternative to Li-ion technology. Its key advantages are low cost, effective life cycle analysis (cradle to grave), and ease of domestic sourcing. − Because early studies projected a low energy density compared to Li-ion chemistry, extensive theoretical and experimental studies have been performed to discover the optimal anode and cathode materials for Na-ion batteries. − A promising and widely studied anode candidate for a range of alkali-ion batteries (including Li-ion) is hard carbon, which is characterized by a conglomeration of regions of turbostratic porous carbon sp3 regions and graphitized sp2 regions. Another candidate is graphite, which is already widely used as the anode for Li-ion batteries; however, Na intercalation into graphite is known to be unstable. This instability arises because Na atoms have a larger atomic radius than Li atoms, and the equilibrium layer spacing of graphite (3.4 Å) is too small for intercalation of Na atoms. Hard carbon is attractive for Na-ion batteries in part because its equilibrium interlayer distance is typically larger than that of graphite.

In the search for the optimal hard carbon alternative to graphite as the anode in Na-ion batteries, various hard carbon structures have been proposed using various experimental and theoretical methods. − Density functional theory (DFT) has been applied effectively to design hard carbons for anodes and to understand the electronic and optical properties of Na-intercalated structures. However, DFT has not been able to conclusively determine the optimal hard carbon structure for Na-ion anodes or the most suitable Na intercalation configuration. This uncertainty in Na intercalation processes within hard carbon motivates the need for a theoretical approach that fully incorporates electron–electron correlation and van der Waals (vdW) interactions to provide an accurate depiction of charge transfer behavior arising from interactions between Na and the hard carbon layers.

To explicitly evaluate all electronic correlations and to reduce the number of uncontrolled approximations in simulations, we used quantum Monte Carlo (QMC) simulations to investigate electronic and optical properties of Na-intercalated hard carbon structures. QMC is a family of stochastic methods that can describe exact ground- and excited-state properties of layered materials, including carbon allotropes. − In this study, we used a bilayer graphene model as the anode and used several computational methods to explore Na intercalation nature at the large interlayer distances characteristic of hard carbon and to estimate the upper limit of Na concentration where stability is retained. We found that Na intercalation stabilizes within the graphene layer at large interlayer distances (over 6 Å) at high Na concentrations (up to C6NaC6). Additionally, we observed that DFT calculations underestimate the band gap opening due to Na intercalation: DFT suggests a closed band gap, but QMC results indicate a potential gap opening in Na-intercalated bilayer graphene. We also discovered that a zigzag configuration of Na atoms may offer an alternative local minimum for Na intercalation in the bilayer graphene. This result suggests that different Na configurations could further stabilize Na intercalation in hard carbon layers.

Methods

Within the QMC family of methods, we used the diffusion Monte Carlo method (DMC) within the fixed-node approximation as implemented in the QMCPACK code. , Slater-Jastrow–type wave functions with a single Slater determinant were used as trial wave functions in the DMC algorithm. In this study, one-, two-, and three-body Jastrow correlation coefficients were used to incorporate electron–ion, electron–electron, and electron–electron–ion correlations, respectively. Cutoffs for the one- and two-body Jastrow coefficients were set as the Wigner-Seitz radius of the given supercell, and a maximum of 5.0 Bohr was used as the cutoff for the short-range three-body term. Single-particle orbitals in the DMC trial wave functions were generated by solving the Kohn–Sham equations using DFT. All DFT calculations in this study were performed using the QUANTUM ESPRESSO code with a plane-wave basis set with a 400 Ry kinetic-energy cutoff and 12 × 12 × 1 Monkhorst–Pack k-point grids. Kohn–Sham orbitals in the Slater determinant were generated using Perdew–Burke–Ernzerhof (PBE) parametrization of the generalized gradient approximation exchange-correlation (XC) functional. The C and Na pseudopotentials were correlation-consistent effective-core potentials (ccECPs) wherein fully correlated all-electron calculationsprimarily coupled-cluster calculationswere used as references for the parametrization of the ECPs. , DMC calculations were done using 0.005Ha–1 time steps within the nonlocal T-move approximation. To reduce one-body finite-size effects from the periodic boundary conditions applied in the DMC calculations, we employed twist-averaged boundary conditions with up to a maximum of 16 twists for each graphene supercell. In addition to the one-body finite-size correction, we estimated twist-averaged DMC energies (averaged over three sizes of supercell) and extrapolated those energies to the bulk limit to further reduce two-body finite-size effects. In addition to finite-size errors, the primary source of systematic error in DMC calculations arises from the fixed-node approximation, where the nodal surface of the trial wave function is constrained to address the Fermionic sign problem in many-body electron systems. The magnitude of the fixed-node error is typically estimated to be about 5% of the correlation energy and is largely canceled out when evaluating binding energies.

Results

To investigate the stability of Na intercalation in bilayer graphene (BLG), we estimated the binding energy E b of a Na atom intercalated into either an AA- or an AB-stacked BLG structure. The Na binding energy at an interlayer distance d in BLG is calculated using the equation ENa‑BLG(d) – EBLG – ENa, where ENa‑BLG(d) is the total energy of Na-intercalated BLG at d, EBLG represents the total energy of pristine BLG at its equilibrium interlayer distance, and ENa is the energy of an isolated Na atom. The equilibrium distances for AA- and AB-stacked BLG are set at 3.5 Å and 3.4 Å, respectively, based on previous QMC studies. In this study, we varied the Na concentration by intercalating a single Na atom into bilayer structures with varying carbon compositions (C2NaC2, C4NaC4, and C6NaC6), as shown in Figure . This approach allowed us to examine the relationship between Na concentration and binding energy. Since the exact geometries of Na-intercalated BLG at specific interlayer distances d have not been experimentally determined, we optimized these geometries including atomic positions and cell parameters using the DFT-PBE XC functional, with the interlayer distance d held constant to find the optimal geometry at each specified distance.

1.

1

Top view of AA- and AB-stacked Na-intercalated bilayers of (a) C2, (b) C4, and (c) C6 graphene.

Figure shows the DMC Na binding energy in AA- and AB-stacked BLG as a function of interlayer distance d. The binding energy for the C2NaC2, C4NaC4, and C6NaC6 configurations is calculated using supercells containing 408 (120), 594 (162), and 588 (156) electrons (atoms), respectively. At high Na concentrations such as C2NaC2 and C4NaC4, the estimated DMC binding energies in PBE-optimized geometry exhibit noticeable deviations from the fitted Morse potential curve. Given the good agreement of the PBE and PBE+D3 binding energy curves with the Morse function, we infer that the geometry of Na-intercalated bilayer graphene may be more complex than the one optimized within the DFT-PBE framework. DFT binding energy curve for Na-intercalated BLG can be found in the Supporting Information. The Na intercalation remains energetically stable only for the configuration with the lowest Na concentration, C6NaC6, which consistently shows a negative Na binding energy, as seen in Figure (c). We find that the Na binding becomes energetically stable at large interlayer distances, exceeding 6 Å, for both AA- and AB-stacked configurations, with the equilibrium interlayer distance decreasing as the Na concentration decreases. Unlike pristine BLG, where AB stacking is generally preferred, AA stacking appears slightly favored in the C4NaC4 and C6NaC6 configurations. However, this preference is minor, with a maximum binding energy difference of only 0.02(2) eV within the given supercell size. To confirm the energetically preferred stacking mode of the Na-intercalated structure, we estimated the DMC Na binding energy at the thermodynamic limit under equilibrium distance conditions. To account for finite-size effects, we extrapolated DMC binding energies to the bulk limit using data from supercells of varying sizes. Further details are provided in the Supporting Information.

2.

2

DMC binding energy of AA- and AB-stacked Na intercalated bilayers of (a) C2, (b) C4, and (c) C6 graphene. Dotted line represents Morse function fit. Note that negative binding energy indicates attractive interaction.

Table summarizes the equilibrium interlayer distances and binding energies estimated from DFT-PBE, DFT-PBE with Grimme dispersion correction (PBE+D3), and DMC. For all Na configurations considered, we observe that Na intercalation stabilizes at large interlayer distances beginning at 6 Å, where pristine BLG would otherwise dissociate. This stability at large interlayer distances are due to the interplay between Na–Na and Na–C interactions. However, in the highest Na concentration, C2NaC2, the equilibrium binding energy becomes unstable at larger interlayer distances, showing a positive value at distances of 7.0–8.0 Å. This instability indicates that Na intercalation is energetically unfavorable in the C2NaC2 configuration. Since stability of Na intercalation in bilayer graphene or defective one was reported in C18NaC18 structure in previous DFT studies, , we assume that Na concentration is required to be reduced beyond C2NaC2 to stabilize Na intercalation. In the C4NaC4 and C6NaC6 configurations, where stable negative Na equilibrium binding energies are observed in both DFT and DMC results, the AA stacking mode is more favorable than the AB mode, with a binding energy difference of 4(3) meV/atom for C4NaC4. However, for C6NaC6, it is challenging to confirm a stable stacking preference with DMC because the difference in binding energy between modes is small, only 2(4) meV/cell. This difference is even smaller than the previously reported DMC binding energy difference of 6.2(12) meV/atom between pristine AA- and AB-stacked BLG. These DMC results suggest a degeneracy may exist between AA- and AB stacking modes in Na-intercalated BLG for C6NaC6 and higher Na concentrations. If so, it may be possible to use external strain or defects to tune the optical properties of Na-intercalated BLG.

1. DFT and DMC Equilibrium Layer Spacing R (Å) and Binding Energy E b (eV/atom) of Na-Intercalated Bilayer C2NaC2, C4NaC4, and C6NaC6 Graphene .

  C2NaC2
C4NaC4
C6NaC6
  AA
AB
AA
AB
AA
AB
  R E b R E b R E b R E b R E b R E b
DFT-PBE 7.9 0.56 8.2 0.58 7.7 –0.32 6.8 –0.28 6.8 –0.83 6.8 –0.81
DFT-PBE+D3 7.0 0.36 7.5 0.37 6.7 –0.43 6.5 –0.40 6.3 –0.87 6.5 –0.84
DMC 7.6(3) 0.134(2) 7.2(5) 0.138(1) 7.4(3) –0.023(2) 6.5(5) –0.019(2) 6.0(5) –0.053(2) 6.4(3) –0.051(4)
a

The unit of eV/atom represents the total energy divided by the total number of atoms in the simulation cell.

To analyze the evolution of optical properties through Na intercalation in BLG, we computed the band structure of Na-intercalated BLG using the PBE XC functional. Figures and show the band structure of Na-intercalated BLG (for AA and AB stacking, respectively) at both the equilibrium distance of pristine BLG and the distance set for the Na-intercalated system. In these band structures, multiple linear dispersions appear at the Dirac cone, similar to those seen for pure BLG; these dispersions result from electronic interlayer coupling. With Na intercalation, the Fermi level shifts above the Dirac cone in both AA and AB stacking due to strong charge transfer from the Na atom to the BLG. Interestingly, this shift remains even at the equilibrium distance of the Na-intercalated graphene, d = d eq , indicating effective charge transfer persists even at large interlayer distances above 6.0 Å.

3.

3

PBE electron band structure of AA-stacked Na-intercalated bilayers of (a) C2NaC2, (b) C4NaC4, and (c) C6NaC6 graphene at d = 3.4 Å (left) and the equilibrium layer spacing (right).

4.

4

PBE electron band structure of AB-stacked Na-intercalated bilayers of (a) C2NaC2, (b) C4NaC4, and (c) C6NaC6 graphene at d = 3.4 Å (left) and the equilibrium layer spacing (right).

Adding Na to BLG leads to gap opening at the Dirac cone at high Na concentrations like C2NaC2 and C4NaC4, with the gap becoming smaller as Na concentration decreases. However, the details of the gap opening differ between the stacking modes. For AA stacking (Figure ), as the interlayer distance increases from d = 3.5 Å to d = d eq , there is a transition from multiple linear dispersion bands to parabolic-like dispersions, owing to reduced interlayer coupling. However, a small band gap opening remains, as shown in Figure . In contrast, for AB stacking, pure and Na-intercalated BLG have different dispersion band structures. AB-stacked BLG at d = 3.4 Å (Figure , left column) displays a linear dispersion band structure, similar to that for AA stacking (see Figure (a)), whereas pure AB BLG (Figure , right column) displays parabolic dispersion bands. Like the AA stacking mode, the AB BLG also shows a band gap opening through Na intercalation. As Na concentration decreases in AB BLG, the parabolic dispersion bands begin to resemble those of Na-intercalated AA BLG with a small gap opening. These band structures confirm strong interactions between BLG and intercalated Na atoms, with Na intercalation possibly promoting gap opening.

To quantify the band gap opening in Na-intercalated BLG, we used both PBE and DMC to estimate the optical band gap for Na-intercalated AA- and AB-stacked BLG. The DMC optical band gap at the Dirac cone was computed by calculating the excitation energy difference, Eex(k bottom → kVBM) – Eex(k upper → kVBM), where Eex(k bottom → kVBM) and Eex(k upper → kVBM) represent excitation energies to the conduction band minimum from the bottom and upper band of the Dirac cone, respectively. Those DMC excitation energies are obtained by promoting an electron from the valence band to the conduction band at the k-point where Dirac cone emerges. For AA-stacked BLG, the band gap at the Dirac cone tends to decrease as the interlayer distance increases in the C2NaC2 configuration (the highest Na concentration) for both DMC and PBE results (Figure (a)). However, for C4NaC4 and C6NaC6, the DMC and PBE results exhibit qualitatively different behavior in band gap opening at large interlayer distances. The DMC band gap rapidly increases for interlayer distances larger than 6.0 Å, whereas the PBE result shows a band gap of almost zero at those distances. Additionally, the DMC band gap tends to be larger than the corresponding PBE result, indicating that the band gap opening from Na intercalation may be underestimated in the DFT-PBE framework. For AB-stacked BLG, DMC results show the band gap increases monotonically with interlayer distance, while the PBE result shows only negligible gap openings at large distances. Based on the DMC optical gap results, we predict significant band gap opening at large interlayer distances in BLG, suggesting that strong charge transfer between BLG and Na atoms can persist at large distances. The largest observed gap opening is approximately 3.0 eV, in the AB-stacked C4NaC4 configuration, though detailed discussion of this result is beyond the scope of this paper. Given the reported overestimation of optical gaps by DMC in some materials with single-reference methods, , a more thorough QMC study to quantify band gap opening is warranted, including analyzing how excited-state properties depend on the nodal surface quality in the trial wave function; we plan to discuss such an analysis in future work.

5.

5

Optical gap calculated by PBE (squares) and DMC (circles) for AA- and AB-stacked bilayers of (a) C2NaC2, (b) C4NaC4, and (c) C6NaC6 graphene as a function of layer spacing.

Since Li-intercalated graphite is a well-known stable anode material for alkali-ion batteries, it is worth comparing the binding properties of Na-intercalated graphene systems with their Li-intercalated graphite counterparts. Figure shows the DMC binding curve from a 588-electron supercell for the C6LiC6 and C6NaC6 systems. Because Na has a larger atomic radius than Li, it is not surprising that Na-intercalated BLG has a larger equilibrium interlayer distance than Li-intercalated structures. In both AA and AB stacking modes, Na shows a smaller binding energy than Li but still maintains stable binding. In Li-intercalated structures, AA stacking is energetically preferred, whereas in Na-intercalated structures, there is a non-negligible difference in equilibrium binding energy between AA- and AB stacking. This result suggests that Na-intercalated AA- and AB-stacked BLG structures may be more sensitive to external strain, allowing for potential tuning of their optical properties.

6.

6

DMC binding energy of AA- and AB-stacked bilayers of (a) C6LiC6 and (b) C6NaC6 graphene as a function of layer spacing.

To further explore the performance of Na-intercalated configurations as anodes, we estimated the intercalation voltage for Na-intercalated BLG using the following equation:

V(x1,x2)=−ENax2C4−ENax1C4−(x2−x1)Eatome(x2−x1)

where ENaxC4, E atom, and e represent the total energy of Na-intercalated bilayer graphene at a concentration x, the total energy of an Na atom derived from body-centered cubic bulk Na, and the fundamental charge, respectively. Note that x values for C2NaC2, C4NaC4, and C6NaC6 configurations are 1, 0.5, and 0.33, respectively. As shown in Table , both Li- and Na-intercalated BLG exhibit negative voltages across all configurations modeled, indicating that the battery cannot be fully charged at the given concentrations and configurations. A promising approach to increase intercalation voltage for anode use could involve reducing the Na concentration or exploring alternative Na configurations for concentrations where stable binding energies are observed (e.g., x = 0.5 and 0.33). It is noteworthy that the discrepancy between DFT and DMC results is significantly smaller for the intercalation voltage than for the binding energy. It is probably due to the cancellation of the interlayer interactions between graphene layers in the calculation of the intercalation voltage, which is based on the energy difference between Na-intercalated bilayer graphene configurations. (ENax2C4 - ENax1C4).

2. DFT and DMC Results for Intercalation Voltage V­(x1,x2) for Li- and Na-Intercalated Bilayer Graphene, Where x Is the Intercalation Concentration .

  V(1, 0.5)
V(0.5, 0.33)
V(0.33, 0.17)
  AA
AB
AA
AB
AA
AB
  Na Na Li Na Li Na Li Na Li Na
DFT-PBE –2.4 –2.5 –1.1 –1.8 –1.2 –1.7 0.0 –0.6 –0.2 –0.5
DFT- PBE+D3 –2.3 –2.4 –1.0 –1.6 –1.2 –1.6 –0.1 –0.4 –0.3 –0.4
DMC –2.6(1) –2.7(1) –1.4(1) –1.8(1) –1.7(1) –2.0(1) –0.1(1) –0.5(1) –0.4(1) –0.3(1)
a

The x values of 1, 0.5, and 0.33 correspond to the C2(Na/Li)­C2, C4(Na/Li)­C4, and C6(Na/Li)­C6 configurations, respectively.

To analyze charge transfer between BLG and the intercalated atoms, we computed the total charge density for Li- and Na-intercalated bilayer C6 graphene. We then used PBE to obtain the charge density difference ρdiff between the Li and Na systems at the equilibrium distance, calculated as ρtotal – ρBLG + ρatom, where ρtotal, ρBLG, and ρatom represent the charge densities of Li- or Na-intercalated BLG, pure BLG, and isolated Li (or Na) atoms, respectively (Figure ). The PBE charge density difference reveals similar charge transfer behavior in both AA- and AB stacking modes for both Li- and Na-intercalated graphene, confirming significant interlayer charge transfer between Na and BLG at large interlayer distances. For a more quantitative assessment of charge transfer within Li- and Na-intercalated systems, we also calculated the charge density difference using DMC (Figure ). In Li-intercalated BLG, there is substantial charge accumulation between the graphene layers and the Li atoms, indicating strong charge transfer. The Na-intercalated system shows a smaller maximum charge density accumulation (∼0.0009 Å–3) compared to Li-intercalated graphene (∼ 0.004 Å–3), likely due to larger interlayer distance imposed by Na. As shown in Figure (b), significant charge accumulation is observed between Na atoms, while only minor density accumulation occurs between Li atoms. This strong Na–Na interaction indicates repulsion occurs at an interatomic distance of around 4 Å, whereas the Li atoms remain stabilized at this distance. Based on these results, we predict that Na intercalation requires either a lower concentration or a larger anode material volume compared to the Li-intercalated system. Differences in ρdiff between the DMC and PBE results (Figure ) suggest that charge transfer between Na atoms and graphene sheets is substantially underestimated in the PBE scheme. Specifically, note the large density difference in the portion of the interlayer region occupied by Na atoms (red band in the middle of Figure (a) and (b)). That disparity indicates PBE may be underestimating the interatomic distance needed to stabilize intercalated Na atoms. This result suggests that finding an optimal Na configuration and in particular a more effective interatomic distance are crucial for improving the performance of Na-intercalated systems as anodes.

7.

7

Top and side view of PBE charge density difference for AA and AB-stacked bilayers of (a) C6LiC6 and (b) C6NaC6 graphene at the equilibrium layer spacing. Yellow and blue indicate positive and negative charge density, respectively.

8.

8

DMC density difference projected onto the (010) plane for AA- and AB-stacked bilayers of (a) C6LiC6 and (b) C6NaC6 graphene at the equilibrium layer spacing. Solid squares and circles represent C and Na­(Li) atoms, respectively.

9.

9

Difference in charge density results between DMC and PBE for (a) AA-stacked and (b) AB-stacked Na-intercalated C6 graphene bilayer. Solid squares and circles represent C and Na­(Li) atoms, respectively. Red indicates underestimation by PBE.

As a first step toward optimization, we need to understand more about how Na atoms can be positioned in the interlayer. As shown in Figure , in the C6NaC6 configuration, the equilibrium Na interatomic distance needs to be greater than in its Li counterpart to maintain stability. We know that Li atoms align along the same z axis (d z = 0), and the results reported so far assume the same alignment for Na atoms (Figure (a)). In order to achieve larger interatomic distance with minimized interatomic repulsion between Na atoms, a zigzag pattern was considered, as shown in Figure (b). Because such a zigzag geometry has not been previously measured for C6NaC6, we optimized Na atom positions for such a configuration using the PBE XC functional. Figure shows the PBE-relaxed Na interatomic distance (d Na) and its vertical component (d z ). Based on the similar trends in d Na and d z in Figure , we predict that most of the change in interatomic distance as the interlayer distance increases can be attributed to increases in d z between Na atoms. In the zigzag arrangement, both d Na and d z increase with interlayer distance, with a noticeable jump in the overall distance occurring between 8 and 9 Å. Beyond an interlayer distance of 9 Å, d Na and d z stabilize at 5.2 and 3.0 Å, suggesting these may be the maximum distances for stable Na atoms.

10.

10

Side view of Na-intercalated C6 graphene bilayer with (a) aligned and (b) zigzag configuration.

11.

11

(a) Distance between Na atoms and (b) vertical component of this distance for Na-intercalated AA- and AB-stacked C6 bilayer graphene.

Figure shows the binding energy for the zigzag and aligned configurations of Na-intercalated C6NaC6 BLG as a function of interlayer distance. Both zigzag and aligned configurations display a global minimum in binding energy at interlayer distances between 6 and 8 Å, as observed in both DFT and DMC results. At this global minimum, it is difficult to discern which configuration is energetically preferred in DMC, as the maximum difference in binding energy between the two configurations is only 0.06(4) eV. Interestingly, only the results from the PBE functional without vdW interactions (solid squares in Figure ) indicate lower binding energy in the zigzag configuration for both AA- and AB stacking modes, suggesting vdW interactions may play different roles in binding between graphene layers and Na atoms. Notably, vdW interactions remain effective at large interlayer distances beyond 9 Å in the zigzag configuration: a sudden increase in binding energy is observed at an interlayer distance of 10 Å in both DMC and PBE+D3 results. (In contrast, the PBE results without vdW interactions show a monotonic decrease in binding energy at large interlayer distances.) The DMC and PBE+D3 results suggest the potential existence of another global minimum for the zigzag configuration at distances beyond 10 Å. In the total density difference (Figure ), we observe stronger charge transfer between Na and C in the zigzag configuration than in the aligned one, which may explain the re-entrant behavior of the binding curve at large interlayer distances. This unique behavior within the modeled zigzag configuration suggests that, in addition to the aligned configuration, another stable Na configuration may exist, one that potentially offers both stable binding energy and stable intercalation voltage at large interlayer distances.

12.

12

Na binding energy in Na-intercalated (a) AA-stacked and (b) AB-stacked C6 bilayer graphene as a function of interlayer distance. Solid and open symbols represent optimized geometry with the aligned and zigzag configuration, respectively.

13.

13

DMC density difference projected onto the (010) plane of AA-stacked bilayer C6NaC6 graphene at 10 Å interlayer distance for (a) the zigzag and (b) aligned configuration.

Conclusion

We investigated Na-intercalated bilayer graphene using DMC and DFT to determine the optimal Na configuration and concentration. Because Na has a larger atomic radius and stronger interatomic interactions compared to Li, Na intercalation in bilayer graphene requires a larger interlayer distance and lower concentration to achieve stability when using the same horizontally aligned configuration for Li. However, the re-entrant behavior observed in the Na binding energy curve for the zigzag configuration suggests a zigzag arrangement may offer an additional local minimum in binding energy, distinct from the horizontally aligned configuration. Since the zigzag arrangement is expected to increase energy density compared to the aligned configuration at the same concentration, identifying the optimal zigzag configuration for Na intercalation is crucial for improving the performance of Na-ion battery anodes. We believe our findings provide valuable guidance for modeling the ideal Na intercalation mode in hard carbon and extending designs beyond the conventional aligned configuration.

Moreover, our DMC results reveal a significant gap opening induced by Na intercalation, which was not captured by the PBE approach. This observation suggests that the DFT-PBE method likely underestimates the gap-opening effect from Na intercalation. These findings underscore the need for further detailed studies on gap opening in both aligned and zigzag Na configurations at larger interlayer distances in hard carbon anodes, along with comprehensive analyses of finite-size effects and multideterminant expansions.

Supplementary Material

pg5c00025_si_001.pdf (267.3KB, pdf)

Acknowledgments

The work conducted at Argonne National Laboratory was supported by the Office of Vehicle Technologies of the U.S. Department of Energy through the Advanced Battery Materials Research (BMR) program, including the LENS Consortium under contract no DE-AC02-06CH11357. Computer time as an award was provided by the Innovative and Novel Computational Impact on Theory and Experiment (INCITE) program. This research used resources of the Argonne Leadership Computing Facility, which is a DOE Office of Science User Facility supported under contract DE-AC02-06CH11357.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsphyschemau.5c00025.

  • Finite-size analysis of DMC binding energy; DFT binding energy of AA- and AB-stacked Na-intercalated bilayer graphene (PDF)

§.

Qubit Pharmaceuticals Inc., 121 West 36th Street 237, Manhattan New York, NY 10018, United States

The authors declare no competing financial interest.

References

  1. Whittingham M. S.. Ultimate Limits to Intercalation Reactions for Lithium Batteries. Chem. Rev. 2014;114(23):11414–11443. doi: 10.1021/cr5003003. [DOI] [PubMed] [Google Scholar]
  2. Canepa P., Gautam G. S., Hannah D. C., Malik R., Liu M., Gallagher K. G., Persson K. A., Ceder G.. Odyssey of Multivalent Cathode Materials: Open Questions and Future Challenges. Chem. Rev. 2017;117(5):4287–4341. doi: 10.1021/acs.chemrev.6b00614. [DOI] [PubMed] [Google Scholar]
  3. Van der Ven A., Deng Z., Banerjee S., Ong S. P.. Rechargeable Alkali-Ion Battery Materials: Theory and Computation. Chem. Rev. 2020;120(14):6977–7019. doi: 10.1021/acs.chemrev.9b00601. [DOI] [PubMed] [Google Scholar]
  4. Dose W. M., Johnson C. S.. Cathode Pre-Lithiation/Sodiation for Next-Generation Batteries. Curr. Opin. Electrochem. 2022;31:100827. doi: 10.1016/j.coelec.2021.100827. [DOI] [Google Scholar]
  5. Eftekhari A.. Potassium Secondary Cell Based on Prussian Blue Cathode. J. Power Sources. 2004;126:221–228. doi: 10.1016/j.jpowsour.2003.08.007. [DOI] [Google Scholar]
  6. Yabuuchi N., Kubota K., Dahbi M., Komaba S.. Research Development on Sodium-Ion Batteries. Chem. Rev. 2014;114(23):11636–11682. doi: 10.1021/cr500192f. [DOI] [PubMed] [Google Scholar]
  7. Arroyo-de Dompablo M. E., Ponrouch A., Johansson P., Palacin M. R.. Achievements, Challenges, and Prospects of Calcium Batteries. Chem. Rev. 2020;120(14):6331. doi: 10.1021/acs.chemrev.9b00339. [DOI] [PubMed] [Google Scholar]
  8. Vaalma C., Buchholz D., Weil M., Passerini S.. A Cost and Resource Analysis of Sodium-Ion Batteries. Nat. Rev. Mater. 2018;3:18013. doi: 10.1038/natrevmats.2018.13. [DOI] [Google Scholar]
  9. Bauer A., Song J., Vail S., Pan W., Barker J., Lu Y.. The Scale-Up and Commercialization of Nonaqueous Na-Ion Battery Technologies. Adv. Energy Mater. 2018;8(17):1702869. doi: 10.1002/aenm.201702869. [DOI] [Google Scholar]
  10. Hu Y.-S., Komaba S., Forsyth M., Johnson C., Rojo T.. A New Emerging Technology: Na-Ion Batteries. Small Methods. 2019;3:1900184. doi: 10.1002/smtd.201900184. [DOI] [Google Scholar]
  11. Abraham K. M.. How Comparable Are Sodium-Ion Batteries to Lithium-Ion Counterparts? ACS Energy Lett. 2020;5:3544–3547. doi: 10.1021/acsenergylett.0c02181. [DOI] [Google Scholar]
  12. Yabuuchi N., Kajiyama M., Iwatate J., Nishikawa H., Hitomi S., Okuyama R., Usui R., Yamada Y., Komaba S.. P2-type Nax[Fe1/2Mn1/2]­O2 Made from Earth-Abundant Elements for Rechargeable Na Batteries. Nat. Mater. 2012;11:512–517. doi: 10.1038/nmat3309. [DOI] [PubMed] [Google Scholar]
  13. Kim H., Park I., Seo D.-W., Lee S., Kim S.-W., Kwon W. J., Park Y.-U., Kim C. S., Jeon S., Kang K.. New Iron-Based Mixed-Polyanion Cathodes for Lithium and Sodium Rechargeable Batteries: Combined First Principles Calculations and Experimental Study. J. Am. Chem. Soc. 2012;134:10369–10372. doi: 10.1021/ja3038646. [DOI] [PubMed] [Google Scholar]
  14. Kim Y., Ha K.-H., Oh S. M., Lee K. T.. High-Capacity Anode Materials for Sodium-Ion Batteries. Chem. Eur. 2014;20(38):11980–11992. doi: 10.1002/chem.201402511. [DOI] [PubMed] [Google Scholar]
  15. Hwang J.-Y., Oh S.-M., Myung S.-T., Chung K. Y., Belharouak I., Sun Y.-K.. Radially Aligned Hierarchical Columnar Structure as a Cathode Material for High Energy Density Sodium-Ion Batteries. Nat. Commun. 2015;6:6865. doi: 10.1038/ncomms7865. [DOI] [PubMed] [Google Scholar]
  16. Mu L., Xu S., Li Y., Hu Y.-S., Li H., Chen L., Huang X.. Prototype Sodium-Ion Batteries Using an Air-Stable and Co/Ni-Free O3-Layered Metal Oxide Cathode. Adv. Mater. 2015;27(43):6928–6933. doi: 10.1002/adma.201502449. [DOI] [PubMed] [Google Scholar]
  17. Song J., Wang L., Lu Y., Liu J., Guo B., Xiao P., Lee J.-J., Yang X.-Q., Henkelman G., Goodenough J. B.. Removal of Interstitial H2O in Hexacyanometallates for a Superior Cathode of a Sodium-Ion Battery. J. Am. Chem. Soc. 2015;137:2658–2664. doi: 10.1021/ja512383b. [DOI] [PubMed] [Google Scholar]
  18. Mortemard de Boisse B., Liu G., Ma J., Nishimura S.-i., Chung S.-C., Kiuchi H., Harada Y., Kikkawa J., Kobayashi Y., Okubo M., Yamada A.. Intermediate Honeycomb Ordering to Trigger Oxygen Redox Chemistry in Layered Battery Electrode. Nat. Commun. 2016;7:11397. doi: 10.1038/ncomms11397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Chen M., Hua W., Xiao J., Cortie D., Guo X., Wang E., Gu Q., Hu Z., Indris S., Wang X.-L., Chou S.-L., Dou S.-X.. Development and Investigation of a NASICON-Type High-Voltage Cathode Material for High-Power Sodium-Ion Batteries. Angew. Chemie - Int. Ed. 2020;59(6):2449–2456. doi: 10.1002/anie.201912964. [DOI] [PubMed] [Google Scholar]
  20. Park J., Ku K., Son S.-B., Gim J., Kim Y., Lee E., Johnson C.. Effect of Electrolytes on the Cathode-Electrolyte Interfacial Stability of Fe-Based Layered Cathodes for Sodium-Ion Batteries. J. Electrochem. Soc. 2022;169(3):030536. doi: 10.1149/1945-7111/ac5c41. [DOI] [Google Scholar]
  21. Cao Y., Xiao L., Sushko M. L., Wang W., Schwenzer B., Xiao J., Nie Z., Saraf L. V., Yang Z., Liu J.. Sodium Ion Insertion in Hollow Carbon Nanowires for Battery Applications. Nano Lett. 2012;12(7):3783–3787. doi: 10.1021/nl3016957. [DOI] [PubMed] [Google Scholar]
  22. Liu Y., Merinov B. V., Goddard W. A. III. Origin of Low Sodium Capacity in Graphite and Generally Weak Substrate Binding of Na and Mg Among Alkali and Alkaline Earth Metals. Proc. Natl. Acad. Sci. U.S.A. 2016;113(14):3735–3739. doi: 10.1073/pnas.1602473113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Yoon G., Kim H., Park I., Kang K.. Conditions for Reversible Na Intercalation in Graphite: Theoretical Studies on the Interplay Among Guest Ions, Solvent, and Graphite Host. Adv. Energy Mater. 2017;7(2):1601519. doi: 10.1002/aenm.201601519. [DOI] [Google Scholar]
  24. Quan B., Jin A., Yu S.-H., Kang S. M., Jeong J., Abruna H. D., Jin L., Piao Y., Sung Y.-E.. Solvothermal-Derived S-Doped Graphene as an Anode Material for Sodium-Ion Batteries. Adv. Sci. 2018;5(5):1700880. doi: 10.1002/advs.201700880. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Sun J., Sadd M., Edenborg P., Gronbeck H., Thiesen P. H., Xia Z., Quintano V., Qui R., Matic A., Palermo V.. Real-Time Imaging of Na+ Reversible Intercalation in “Janus” Graphene Stacks for Battery Applications. Sci. Adv. 2021;7(22):eabf0812. doi: 10.1126/sciadv.abf0812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Ma J., Yang C., Li Q., Li S., Xu L., Fang S., Sun X., Quhe R., Pan F., Lu J.. 3D Hierarchical Graphene-CNT Anode for Sodium-Ion Batteries: A First-Principles Assessment. Adv. Theory Simul. 2022;5(10):2200227. doi: 10.1002/adts.202200227. [DOI] [Google Scholar]
  27. Park W.-Y., Han J., Moon J., Joo S.-H., Wada T., Ichikawa Y., Ogawa K., Kim H. S., Chen M., Kato H.. Mechanically Robust Self-Organized Crack-Free Nanocellular Graphene with Outstanding Electrochemical Properties in Sodium Ion Battery. Adv. Mater. 2024;36(21):2311792. doi: 10.1002/adma.202311792. [DOI] [PubMed] [Google Scholar]
  28. Fu H., Gao B., Qiao Y., Lin P., Liu Z., Yuan S., Abdelkader A., Kamili A. R.. Nano-Crosslinked Mesoporous Graphene for Superior Na-Ion Storage. J. Energy Storage. 2024;86:111555. doi: 10.1016/j.est.2024.111555. [DOI] [Google Scholar]
  29. Pandey M., Deshmukh K., Raman A., Asok A., Appukuttan S., Suman G. R.. Prospects of MXene and Graphene for Energy Storage and Conversion. Renew. Sustain. Energy Rev. 2024;189:114030. doi: 10.1016/j.rser.2023.114030. [DOI] [Google Scholar]
  30. Peymanirad F., Majidi R., Vishkayi S. I., Soleimani H. R.. Density Functional Theory Study of Molecular Pillared Graphene for High-Performance Sodium-Ion Batteries. Appl. Surf. Sci. 2024;669:160502. doi: 10.1016/j.apsusc.2024.160502. [DOI] [Google Scholar]
  31. Foulkes W. M. C., Mitas L., Needs R. J., Rajagopal G.. Quantum Monte Carlo Simulations of Solids. Rev. Mod. Phys. 2001;73(1):33–83. doi: 10.1103/RevModPhys.73.33. [DOI] [Google Scholar]
  32. Shin H., Kang S., Koo J., Lee H., Kim J., Kwon Y.. Cohesion Energetics of Carbon Allotropes: Quantum Monte Carlo Study. J. Chem. Phys. 2014;140(11):114702. doi: 10.1063/1.4867544. [DOI] [PubMed] [Google Scholar]
  33. Ganesh P., Kim J., Park C., Yoon M., Reboredo F. A., Kent P. R. C.. Binding and Diffusion of Lithium in Graphite: Quantum Monte Carlo Benchmarks and Validation of van der Waals Density Functional Methods. J. Chem. Theory Comput. 2014;10(12):5318–5323. doi: 10.1021/ct500617z. [DOI] [PubMed] [Google Scholar]
  34. Mostaani E., Drummond N. D., Fal’ko V. I.. Quantum Monte Carlo Calculation of the Binding Energy of Bilayer Graphene. Phys. Rev. Lett. 2015;115(11):115501. doi: 10.1103/PhysRevLett.115.115501. [DOI] [PubMed] [Google Scholar]
  35. Shin H., Kim J., Lee H., Heinonen O., Benali A., Kwon Y.. Nature of Interlayer Binding and Stacking of sp-sp2 Hybridized Carbon Layers: A Quantum Monte Carlo Study. J. Chem. Theory Comput. 2017;13(11):5639–5646. doi: 10.1021/acs.jctc.7b00747. [DOI] [PubMed] [Google Scholar]
  36. Kim J., Baczewski A., Beaudet T., Benali A., Bennett C., Berrill M., Blunt N., Borda E. J. L., Casula M., Ceperley D., Chiesa S., Clark B. K., Clay R., Delaney K., Dewing M., Esler K., Hao H., Heinonen O., Kent P. R. C., Krogel J. T., Kylanpaa I., Li Y. W., Lopez M. G., Luo Y., Malone F., Martin R., Mathuriya A., McMinis J., Melton C., Mitas L., Morales M. A., Neuscamman E., Parker W., Flores S., Romero N. A., Rubenstein B., Shea J., Shin H., Shulenburger L., Tillack A., Townsend J., Tubman N., van der Goetz B., Vincent J., Yang D. C., Yang Y., Zhang S., Zhao L.. QMCPACK: An Open Source Ab Initio Quantum Monte Carlo Package for the Electronic Structure of Atoms, Molecules and Solids. J. Phys.: Condens. Matter. 2018;30(19):195901. doi: 10.1088/1361-648X/aab9c3. [DOI] [PubMed] [Google Scholar]
  37. Kent P. R. C., Annaberidiyev A., Benali A., Bennett M. C., Borda E. J. L., Doak P., Hao H., Jordan K. D., Krogel K. T., Kylänpää I., Lee J., Luo Y., Malone F. D., Melton C. A., Mitas L., Morales M. A., Neuscamman E., Reboredo F. A., Rubenstein B., Saritas K., Upadhyay S., Wang G., Zhang S., Zhao L.. QMCPACK: Advances in the Development, Efficiency, and Application of Auxiliary Field and Real-Space Variational and Diffusion Quantum Monte Carlo. J. Chem. Phys. 2020;152(17):174105. doi: 10.1063/5.0004860. [DOI] [PubMed] [Google Scholar]
  38. Giannozzi P., Baroni S., Bonini N., Calandra M., Car R., Cavazzoni C., Ceresoli D., Chiarotti G. L., Cococcioni M., Dabo I., Corso A. D., de Gironcoli S., Fabris S., Fratesi G., Gebauer R., Gerstmann U., Gougoussis C., Kokalj A., Lazzeri M., Martin-Samos L., Mazari N., Mauri F., Mazzarello R., Paolini S., Pasquarello A., Paulatto L., Sbraccia C., Scandolo S., Sclauzero G., Seitsonen A. P., Smogunov A., Umari P., Wentzcovitch R. M.. QUANTUM ESPRESSO: A Modular and Open-Source Software Project for Quantum Simulations of Materials. J. Phys.: Condens. Matter. 2009;21(39):395502. doi: 10.1088/0953-8984/21/39/395502. [DOI] [PubMed] [Google Scholar]
  39. Perdew J. P., Burke K., Ernzerhof M.. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996;77(18):3865–3868. doi: 10.1103/PhysRevLett.77.3865. Errata in Phys. Rev. Lett. 78, 1396 (1997) [DOI] [PubMed] [Google Scholar]
  40. Bennett M. C., Melton C. A., Annaberdiyev A., Wang G., Shulenburger L., Mitas L.. A New Generation of Effective Core Potentials for Correlated Calculations. J. Chem. Phys. 2017;147(22):224106. doi: 10.1063/1.4995643. [DOI] [PubMed] [Google Scholar]
  41. Bennett M. C., Wang G., Annaberdiyev A., Melton C. A., Shulenburger L., Mitas L.. A New Generation of Effective Core Potentials from Correlated Calculations: 2nd Row Elements. J. Chem. Phys. 2018;149(10):104108. doi: 10.1063/1.5038135. [DOI] [PubMed] [Google Scholar]
  42. Casula M., Moroni S., Sorella S., Filippi C.. Size-Consistent Variational Approaches to Nonlocal Pseudopotentials: Standard and Lattice Regularized Diffusion Monte Carlo Methods Revisited. J. Chem. Phys. 2010;132(15):154113. doi: 10.1063/1.3380831. [DOI] [PubMed] [Google Scholar]
  43. Lin C., Zong F. H., Ceperley D. M.. Twist-Averaged Boundary Conditions in Continuum Quantum Monte Carlo Algorithms. Phys. Rev. E. 2001;64(1):016702. doi: 10.1103/PhysRevE.64.016702. [DOI] [PubMed] [Google Scholar]
  44. Anderson J. B.. Quantum chemistry by random walk. H 2P, H+ 3  D3h  1A′1, H2  3Σ+ u, H4  1Σ+ g Be 1S. J. Chem. Phys. 1976;65(10):4121–4127. doi: 10.1063/1.432868. [DOI] [Google Scholar]
  45. Grimme S., Antony J., Ehrlich S., Krieg H.. A Consistent and Accurate Ab Initio Parametrization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu. J. Chem. Phys. 2010;132(15):154104. doi: 10.1063/1.3382344. [DOI] [PubMed] [Google Scholar]
  46. Yang S., Li S., Tang S., Dong W., Sun W., Shen D., Wang M.. Sodium adsorption and intercalation in bilayer graphene from density functional theory calculations. Theor. Chem. Acc. 2016;135:164. doi: 10.1007/s00214-016-1910-0. [DOI] [Google Scholar]
  47. Yang S., Li S., Tang S., Shen D., Dong W., Sun W.. Adsorption, intercalation and diffusion of Na on defective bilayer graphene: a computational study. Surf. Sci. 2017;658:31–37. doi: 10.1016/j.susc.2017.01.004. [DOI] [Google Scholar]
  48. Shin H., Luo Y., Ganesh P., Balachandran J., Krogel J. T., Kent P. R. C., Benali A., Heinonen O.. Electronic properties of doped and defective NiO: A Quantum Monte Carlo Study. Phys. Rev. Materials. 2017;1(7):073603. doi: 10.1103/PhysRevMaterials.1.073603. [DOI] [Google Scholar]
  49. Shin H., Gasperich K., Rojas T., Ngo A. T., Krogel J. T., Benali A.. Systematic Improvement of Quantum Monte Carlo Calculations in Transition Metal Oxides: sCI-Driven Wavefunction Optimization for Reliable Band Gap Prediction. J. Chem. Theory Comput. 2024;20(18):8175–8189. doi: 10.1021/acs.jctc.4c00335. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

pg5c00025_si_001.pdf (267.3KB, pdf)

Articles from ACS Physical Chemistry Au are provided here courtesy of American Chemical Society

RESOURCES