Horizontal Lithium Electrodeposition on Atomically Polarized Monolayer Hexagonal Boron Nitride

: Both uncontrolled Li dendrite growth and corrosion are major obstacles to the practical application of Li-metal batteries. Despite numerous attempts to address these challenges, effective solutions for dendrite-free reversible Li electrodeposition have remained elusive. Here, we demonstrate the horizontal Li electro-deposition on top of atomically polarized monolayer hexagonal boron nitride (hBN). Theoretical investigations revealed that the hexagonal lattice configuration and polarity of the monolayer hBN, devoid of dangling bonds, reduced the energy barrier for the surface diffusion of Li, thus facilitating reversible in-plane Li growth. Moreover, the single-atom-thick hBN deposited on a Cu current collector (monolayer hBN/Cu) facilitated the formation of an inorganic-rich, homogeneous solid electrolyte interphase layer, which enabled the uniform Li + flux and suppressed Li corrosion. Consequently, Li-metal and anode-free full cells containing the monolayer hBN/Cu exhibited improved rate performance and cycle life. This study suggests that the monolayer hBN is a promising class of underlying seed layers to enable dendrite-and corrosion-free, horizontal Li electrodeposition for sustainable Li-metal anodes in next-generation batteries.


INTRODUCTION
Li-metal batteries have attracted considerable attention as a promising energy-dense alternative outperforming commercial Li-ion batteries owing to the use of Li-metal anodes with the potential of simultaneously providing the highest theoretical capacity (3860 mAh g −1 ) and lowest redox potential (−3.04 V vs standard hydrogen electrode). 1,2However, their practical application has been plagued by the inferior electrochemical reversibility, susceptibility to corrosion, and vigorous reactivity of Li-metal anodes.Furthermore, the uncontrolled Li dendrite growth often results in a short circuit, which poses safety risks and causes immediate cell failure. 3,4−12 Therefore, most previous works on Li-metal anodes have focused on the introduction of artificial SEI layers 13−15 and the regulation of electrolyte solvation sheaths 16,17 and the surface chemistry of substrates 18−20 to facilitate the formation of uniform, inorganic-rich SEI layers.
−24 However, achieving dendrite-free Li deposition remains elusive due to the difficulty in precisely controlling the underlying substrates responsible for Li adsorption energy, Li diffusion barrier, and interfacial energy.Moreover, there are few studies reporting the simultaneous manipulation of SEI layers and underlying substrates in terms of thermodynamic engineering.
Here, we report atomically polarized monolayer hexagonal boron nitride (hBN) as a class of scalable underlying seed layers for achieving horizontal Li deposition.The single-atomthick dense hBN was deposited on a Cu foil (serving as a current collector) to allow a tunneling electron flow, 25,26 enabling electrochemical reactions on top of the hBN surface.In addition, the ultraflat surface of hBN, free of dangling bonds and with negligible defects, 27−29 favors uniform Li deposition.Moreover, the monolayer hBN, due to its highly ordered hexagonal lattice configuration and polarity, reduced the energy barrier for Li surface diffusion while allowing the balanced Li adsorption energy and the negative interfacial energy.Consequently, the horizontal and compact growth of Li was promoted.This in-plane Li deposition on top of the monolayer hBN is quite different from a previous study that reported Li deposition beneath an electrically insulating multilayered hBN with defects acting as a Li protective layer. 30Additionally, monolayer hBN facilitated the formation of an inorganic-rich, homogeneous SEI layer when exposed to electrolytes.Consequently, uniform Li + flux was achieved while suppressing chemical and galvanic Li corrosion.The monolayer hBN/Cu anode and a LiFePO 4 (LFP) cathode (used as a model cathode, areal capacity = 3.7 mAh cm −2 ) were used to fabricate Li-metal and anode-free full cells, and the cells exhibited high rate performance and stable cycle retention.These results demonstrated that the implementation of the underlying seed layer based on the monolayer hBN to overcome the thermodynamic challenges of reversible Li plating/stripping is a promising approach to achieve the enhanced reversibility of Li utilization for energy-dense Limetal batteries.

Horizontal Li Deposition on the Monolayer hBN/Cu.
The high activation energy for Li surface diffusion on a Cu foil leads to sluggish planar migration, 31,32 resulting in irregular and dendritic Li deposition (Figure 1a).In contrast, the monolayer hBN/Cu allowed for the reduction of the activation energy for Li surface diffusion, resulting in uniform and compact in-plane Li deposition (Figure 1b).The monolayer hBN was grown on a Cu foil by flowing borazine (at a flow rate of 0.12 standard cubic centimeters per minute (sccm)) for 2 min into a growth furnace held at 1025 °C (see the Experimental Section and Figure S1).To determine the phonon features and sample uniformity, the Raman mapping of the monolayer hBN on a SiO 2 (300 nm)/Si substrate was conducted over a mapping area of 4 × 4 μm (Figure 1c).An E 2g peak was observed at 1367 cm −1 , verifying the successful growth of hBN.The average intensity of the peak (E 2g mode at 1367 cm −1 ) for a total of 1734 spectra measured at six marked positions was 295.6 ± 21.2, indicating the large-area uniformity of the hBN thickness (monolayer over the entire hBN film).The monolayer hBN on the Cu foil was transferred to a SiO 2 / Si substrate and investigated using atomic force microscopy (AFM).Figure 1d shows that the thickness of hBN was 0.6 nm, corresponding to the single-atom thickness value.Additionally, the difference in the electrical resistivity between the monolayer hBN/Cu and the pristine Cu was negligible (Figure 1e).
To compare the Li deposition behavior on the pristine Cu, monolayer graphene/Cu (as a control sample without polarity), and monolayer hBN/Cu, density functional theory (DFT) calculations were performed.It is known that Li deposition can be described in terms of Li adsorption energy (E ad ), Li diffusion activation energy, and interfacial energy with substrates. 32,33The Cu (111) surface was used as a model substrate in all calculations, as it exhibited the most stable surface energy (Figure S2).The Li diffusion activation energy on the monolayer hBN/Cu was calculated to be 0.894 kcal mol −1 , which was significantly lower than those on the pristine Cu (1.860 kcal mol −1 ) and monolayer graphene/Cu (7.559 kcal mol −1 ) (Figures 1f and S3).Subsequently, the adsorption energies of the Li atom on the substrates were examined.Previous works revealed that a balanced Li adsorption energy, marginally exceeding the Li−Li interaction energy, is preferred to facilitate both Li nucleation and diffusion. 32,34The monolayer hBN/Cu exhibited a slightly stronger Li adsorption energy (−0.89 eV) than the Li−Li interaction energy (−0.88 eV), but much weaker than those on the pristine Cu surface (−2.46 eV) and monolayer graphene/Cu surface (−1.76 eV) (see the Experimental Section).The calculated diffusion activation energies and adsorption energies on the substrates demonstrate the uniform deposition of Li metal on monolayer hBN/Cu.
To further confirm the Li deposition behavior on the substrates, the interfacial energies of Li adsorbed on the pristine Cu (111), monolayer graphene/Cu (111), and monolayer hBN/Cu (111) were calculated (Figures 1g and  S4).The interfacial energy of Li represents the thermodynamic stability of Li layers on substrates. 35,36The monolayer hBN/ Cu exhibited a substantially low interfacial energy (−0.56 eV nm −2 ) compared to the pristine Cu (13.98 eV nm −2 ) and monolayer graphene/Cu (1.30 eV nm −2 ).It is noteworthy that a negative interfacial energy implies the thermodynamic stability of the newly formed layers.This result indicates that the formation of Li layers on monolayer hBN/Cu is thermodynamically more favorable compared to the pristine Cu and monolayer graphene/Cu.These findings confirm that the low Li diffusion activation energy and the balanced adsorption energy of the monolayer hBN/Cu, along with its negative interfacial energy, facilitate the deposition of a uniform 2D Li layer on the monolayer hBN/Cu.Inorganic-Rich, Homogeneous SEI Layer Formed on the Monolayer hBN/Cu.The SEI layer plays a critical role in protecting the Li-metal anode by suppressing Li dendrite growth and Li corrosion. 11,37,38To elucidate the dependence of substrates on the topological and compositional attributes of the SEI layers, asymmetric galvanic cells (substrate||Li metal) 38 with an externally connected ammeter were fabricated (Figure S5).After the samples were stored for 250 h at a substrate potential of 0 V, we analyzed the morphology of the SEI layers formed on the substrates (Figure 2a−c).The SEI layers on the pristine Cu and monolayer graphene/Cu exhibited a nonuniform thickness with randomly distributed organic-rich components (Figure 2a,b).Such SEI layers often induce an uneven and localized Li + flux toward the anode, resulting in Li dendrite growth and Li corrosion. 10,38In contrast, a uniform and smooth SEI layer (thickness ∼40 nm) with homogeneously distributed LiF crystals (lattice spacing ∼0.21 nm) and Li 2 CO 3 crystals (lattice spacing ∼0.31 nm) was formed over the monolayer hBN/Cu (Figures 2c and S6).This phenomenon can be attributed to the atomic-scale polarity of the monolayer hBN owing to the locally charged B and N within the lattice. 39This indicates the strong interaction between hBN and Li salt anions (herein, bis-(trifluoromethanesulfonyl)imide (TFSI − )). 40The strong anion adsorption on the hBN surface can lead to a reduction in the bond dissociation energy of anions, which is similar to those observed in previous works, 41,42 resulting in the formation of anion-derived, inorganic-rich SEI layers.The homogeneous distribution of inorganic-rich compounds in the SEI layers is known to promote uniform Li + flux and improve electrochemical stability and structural integrity. 10,43he chemical composition of SEI layers on the different substrates was analyzed by using X-ray photoelectron spectroscopy (XPS) analysis (Figure 2d−f).A high intensity of the peak (283.8 eV) corresponding to C−C bonds was observed in the XPS profiles of the pristine Cu, and this peak represents organic species mainly derived from electrolyte solvents.Moreover, the high intensities of C−F species (688.0 eV) were observed in the SEI formed on the pristine Cu and monolayer graphene/Cu, revealing the difficult cleavage of C− F bonds in TFSI − . 44In contrast, trace amounts of C−C and C−F species were observed in the presence of monolayer hBN/Cu, but it was abundant in Li−F species (684.2 eV), indicating a viable role of the monolayer hBN in promoting the cleavage of C−F bonds to generate the LiF-rich SEI.The fractions of inorganic components (including Li−F and Li− CO 3 ) in the SEI layer formed on the pristine Cu, monolayer graphene/Cu, and monolayer hBN/Cu were 21.9, 26.9, and 51.1%, respectively, suggesting the promotion of the formation of an inorganic-rich SEI layer by the monolayer hBN/Cu (Figure 2g).
To confirm the structural homogeneity of the SEI layer formed on the monolayer hBN/Cu, we measured the slope (k) in the plots of capacity loss (Q loss ) as a function of the current density using asymmetric cells, where a low k value reflects the uniformity of Li + flux through the SEI layer. 10The monolayer hBN/Cu exhibited a significantly smaller k value (0.41) compared to those of the pristine Cu (1.63) and monolayer graphene/Cu (1.60), which can be attributed to the presence of the inorganic-rich, homogeneous SEI layer (Figures 2h and  S7).
Spontaneous corrosion is another challenge facing Li metals, owing to their thermodynamic instability in organic solvents.To quantify the rate of Li corrosion, the time-dependent corrosion currents of the galvanic cells were measured (Figure 2i).The galvanic currents in the cells containing the pristine Cu and monolayer graphene/Cu did not drop to zero even after 250 h of storage, indicating the continuous galvanic corrosion of Li.In contrast, the galvanic current of the cell with monolayer hBN/Cu reached zero, indicating the inhibition of galvanic corrosion.
Li Deposition and Corrosion on the Monolayer hBN/ Cu.To verify the effect of monolayer hBN/Cu on the electrochemical performance, the Li deposition behavior on the different substrates was investigated by using asymmetric cells (substrate||Li metal) at a current density of 0.01 mA cm −2 (Figure 3a).The Li nucleation overpotential of the monolayer hBN/Cu (10 mV) was lower than those of the pristine Cu foil (22 mV) and the monolayer graphene/Cu (15 mV).Particularly, there was no significant change in the Li nucleation overpotential of monolayer hBN/Cu over the entire current densities ranging from 0.01 to 1 mA cm −2 , demonstrating the role of monolayer hBN/Cu in regulating stable Li nucleation (Figure 3b).In contrast, there was a substantial increase in the nucleation overpotential of the cells with the pristine Cu and monolayer graphene/Cu with increasing current density.These results demonstrated the combined effects of the reduced Li diffusion barrier and the inorganic-rich, homogeneous SEI layers on the monolayer hBN/Cu.Additionally, the versatility of the monolayer hBN for the reduction of the Li nucleation overpotential was observed using various metallic current collectors, including stainless-steel (SUS) and Ni foils (Figure 3c).This result highlights the role of the monolayer hBN in stabilizing Li nucleation under different current densities and current collectors.
The top and cross-sectional scanning electron microscopy (SEM) images of the Li metals deposited on different substrates were obtained to investigate their morphology (Figure 3d−f).During the deposition process, each substrate was plated with 5 mAh cm −2 of Li, corresponding to a theoretical thickness of 25 μm.The pristine Cu and monolayer graphene/Cu exhibited a porous and dendritic Li morphology with thicknesses of 52 and 47 μm, respectively.In contrast, a 25 μm-thick compact and dendrite-free Li metal was observed over the monolayer hBN/Cu, mainly due to the inorganic-rich, homogeneous SEI layers together with the facilitated Li diffusion.In addition, the change in the XPS profiles before and after Li deposition on the monolayer hBN/Cu was investigated (Figure S8).The intensities of the characteristic peaks corresponding to the B−N bonds and Cu decreased, while the peak assigned to Li was intensified after Li deposition.Moreover, the cross-sectional TEM analysis exhibited that hBN/Cu was almost completely covered by uniform and dense Li deposition (Figure S9).
To investigate the electrochemical reversibility of Li deposition/stripping, the change in the Coulombic efficiency (CE) of asymmetric cells was monitored at a current density of 1 mA cm −2 and an areal capacity of 1 mAh cm −2 at various interval rest times between Li deposition and striping (0.1, 1, and 12 h; Figure 3g).The interval rest time reflects the calendar aging, which should be critically considered for practical battery operating conditions. 11,38There was a sharp drop in the CE of the cells fabricated with the pristine Cu and monolayer graphene/Cu after only 50 cycles at an interval of 0.1 h.Furthermore, as the interval rest time increased, the CE value decreased rapidly due to sufficient time for Li corrosion reactions.In contrast, monolayer hBN/Cu maintained high CE values over 200 cycles regardless of the interval rest time, demonstrating the reversible Li deposition/stripping reactions and corrosion resistance.Moreover, the monolayer hBN/Cu maintained high CE values even when the interval rest time was increased to 200 h (Figure 3h).Additionally, the CE values of the monolayer hBN/Cu remained almost unchanged, even when the interval rest time was increased up to 200 h.This result was confirmed by analyzing the electrochemical impedance spectroscopy (EIS) results as a function of the interval rest time (Figure 3i).The interfacial resistance of the pristine Cu and monolayer graphene/Cu increased significantly with an increase in the interval rest time, whereas the monolayer hBN/Cu effectively suppressed the increase in the interfacial resistance.These results indicate the ability of the inorganic-rich, homogeneous SEI layer on monolayer hBN/Cu to suppress Li corrosion, thus enabling the practical use of Limetal anodes.Even at a higher current density of 5 mA cm −2 and an areal capacity of 5 mAh cm −2 , the monolayer hBN/Cu exhibited stable CE values compared to the pristine Cu and monolayer graphene/Cu (Figure S10), further verifying the electrochemical superiority of the monolayer hBN/Cu.
Electrochemical Performance of the Li-Metal and Anode-Free Full Cells Assembled with the Monolayer hBN/Cu.To assess the practicability of the monolayer hBN/ Cu, the monolayer hBN/Cu was paired with an LFP cathode (areal capacity = 3.7 mAh cm −2 ), in which the monolayer hBN/Cu anode was precharged with a limited amount of Li (3.7 mAh cm −2 (Figure S11), N/P ratio = 1).The Li-metal full cell with the monolayer hBN/Cu exhibited a stable cycle performance (85% cycle retention after 300 cycles; Figures 4a  and S12).In contrast, the control Li-metal full cell with a precharged Cu anode (N/P ratio = 1) exhibited rapid capacity degradation (7% cycle retention after 160 cycles).Furthermore, the Li-metal full cell with the monolayer hBN/Cu exhibited higher discharge capacities over a wide range of current densities compared to the control Li-metal full cell, indicating a superior rate capability (Figure S13).
Next, we prepared an anode-free full cell composed of the monolayer hBN/Cu (without precharging) and the LFP cathode, in which the Li sources solely lie in the cathode.The anode-free full cell with the monolayer hBN/Cu exhibited stable cycling retention (84% after 50 cycles), whereas the control anode-free cell with the pristine Cu showed a rapid capacity decay and only retained 5% of its capacity after 50 cycles (Figures 4b and S14).
The surface morphology of the cycled monolayer hBN/Cu and pristine Cu was investigated to further understand the difference in their cycling performance (Figure S15).The cycled monolayer hBN/Cu retained its original morphology, whereas numerous randomly formed Li dendrites and inactive Li were observed on cycled pristine Cu.In addition, we analyzed the local electrochemical impedance spectra (LEIS) 45,46 of the monolayer hBN/Cu and pristine Cu after the cycling tests (Figure 4c,d).The cycled pristine Cu exhibited high and irregular charge transfer resistance (R ct ) values, whereas the cycled monolayer hBN/Cu showed low and homogeneous R ct values over a wide electrode area, indicating uniform Li deposition with minimal generation of inactive Li enabled by monolayer hBN/Cu.
−54 The anode-free full cell with monolayer hBN/Cu was observed to exhibit improved areal capacity, CE, cyclability, and coating thickness, highlighting its viability as a promising underlying seed layer for anode-free full cells.

CONCLUSIONS
In summary, we have demonstrated a class of scalable underlying seed layers based on atomically polarized monolayer hBN to enable reversible planar Li electrodeposition while suppressing Li corrosion.The single-atomthick dense hBN deposited on the Cu foil allowed a tunneling electron flow with a negligible increase in the Cu foil thickness.The DFT calculations revealed that the monolayer hBN reduced the activation energy for Li surface diffusion and the Li−substrate interfacial energy, facilitating the planar and compact deposition of Li.The inorganic-rich, homogeneous SEI layer formed over the monolayer hBN played an important role in promoting uniform Li + flux and suppressing Li corrosion.Consequently, the Li-metal (N/P ratio of 1) and anode-free full cells, composed of monolayer hBN/Cu and the LFP cathode, exhibited improved rate performance and cycling retention.The underlying seed-layer strategy, driven by the monolayer hBN, is promising as a platform technology that enables dendrite-and corrosion-free, horizontal electrodeposition and can be further extended to emerging energydense metal battery systems.

EXPERIMENTAL SECTION
Preparation of the Monolayer hBN/Cu.hBN was grown by using a low-pressure chemical vapor deposition method.First, a Cu foil was placed in the middle of a 2-in quartz tube system, and borazine (Gelest, Inc.) was placed in a bubbler.The Cu foil was preannealed in a furnace heated to 1025 °C for 20 min under H 2 gas flow (12 sccm) to remove impurities.The growth of hBN on the Cu foil was initiated by flowing borazine (2 sccm).During the growth, the pressure was maintained at 0.1 Torr.After the growth was completed, the furnace was rapidly cooled to room temperature under a H 2 gas atmosphere.A commercial monolayer graphene/Cu (ALPHA Graphene) was used as a control sample for comparison.
Transfer of the Monolayer hBN to Various Substrates.The monolayer hBN was transferred to various substrates by using the electrochemical bubbling method.First, the monolayer hBN grown on the Cu foil was coated with a poly(methyl methacrylate) (PMMA) layer.The PMMA-coated monolayer hBN/Cu was used as the cathode, and bare Pt foil was used as the anode.The setup was immersed in an aqueous solution of 1 M NaOH for an electrochemical delamination process.During the bubbling transfer, a constant current of 1 A was applied, corresponding to an electrolytic voltage range of 5−12 V.The process lasted for 1−5 min.After the transfer, the PMMA-coated monolayer hBN layer was peeled off from the Cu foil and rinsed with deionized water to remove any remaining NaOH solution.Subsequently, the layer was transferred to various substrates (e.g., SUS, Ni, and SiO 2 (300 nm)/Si) and then treated with acetone to remove the PMMA layer.
Characterizations.The surface and cross-sectional morphologies of monolayer hBN/Cu were characterized by using field-emission SEM (S-4800, Hitachi), and the thickness of hBN was characterized by using AFM (Dimension Icon, Bruker).The Raman spectrum was obtained using a micro-Raman spectrometer (alpha300, WITec GmbH) with a laser excitation wavelength of 532 nm and power of approximately 2 mW.The electrical resistivities of the substrates were measured by using a four-point probe technique (CMT-SR1000N, Advanced Instrument Technology) at multiple points to obtain reproducible data.The morphology of the SEI layer and the deposition of Li metal on the monolayer hBN/Cu were investigated using TEM equipment (JEM-2100F, JEOL).The TEM sample was fabricated using a focused ion beam system (Helios NanoLab 450, FEI) and was rapidly transferred to the TEM holder to minimize any air-contact-induced changes.The composition of the SEI layer and the change in the XPS profile before and after the deposition of Li metal on current collectors were confirmed by using XPS (ESCALAB 250 Xi, Thermo Scientific).The local resistance of the cycled monolayer hBN/Cu (or pristine Cu) was characterized by using a scanning electrochemical workstation (M470, BioLogic) connected to a Pt dual microelectrode (LEIS scanning probe, BioLogic) and a potentiostat (SP-300, BioLogic).
Electrochemical Measurements.Li deposition was performed using an asymmetric cell (substrates (pristine Cu, monolayer graphene/Cu, or monolayer hBN/Cu)||Li metal (200 μm, Honjo Chemical Co.)) with a liquid electrolyte (1 M LiTFSI in 1,3dioxolane (DOL)/1,2-dimethoxyethane (DME) = 1/1 (v/v) with 2 wt % lithium nitrate (LiNO 3 ) additive).The CE of the current collectors was measured using the asymmetric cell based on the equation: CE (%) = (discharge capacity/charge capacity) × 100.The cell configuration of the galvanic cell was the same as that of the asymmetric cell, but current collectors and Li metal were connected externally via an ammeter.The EIS data of the cells were recorded using a potentiostat/galvanostat (VSP classic, BioLogic) in the frequency range of 10 −2 −10 6 Hz at an applied voltage of 10 mV.Limetal full cells and anode-free full cells were fabricated by coupling the monolayer hBN/Cu (or pristine Cu) with the LFP cathode (areal mass loading = 24 mg cm −2 , provided by LG Energy Solution) containing a polyethylene separator (thickness = 16 μm, areal mass loading = 1.2 mg cm −2 ) and an electrolyte loading of 60 μL.Prior to the fabrication of the Li-metal full cell, the monolayer hBN/Cu (or pristine Cu) was precharged at a capacity of 3.7 mAh cm −2 (areal mass loading of deposited Li = 0.96 mg cm −2 ) to obtain an N/P ratio of 1.The electrochemical performance of the cells was investigated using 2032-type coin cells (cathode, separator, and anode diameters = 10, 18, and 12 mm, respectively) and measured at ambient temperature under various charge/discharge current densities and a voltage range of 2.5−3.7 V using a cycle tester (PNE Solution).
DFT Calculations.−57 For the Dmol 3 calculations, the exchange-correlation potential of the electrons was described by using the generalized gradient approximation (GGA) with the Perdew−Burke−Ernzerhof (PBE) functional.Spin-unrestricted calculations were performed using the DNP 4.4 basis set with all-electron relativistic core treatment.The Tkatchenko−Scheffler method was employed to include dispersion effects. 58The convergence criterion for the self-consistent field (SCF) calculations was set to 1 × 10 −5 Ha, and the convergence criteria for geometry optimization were set to 1 × 10 −5 Ha, 0.002 Ha Å −1 , and 0.005 Å for energy, force, and displacement, respectively.Various monolayer stacking configurations were initially considered.The most stable configurations of monolayer graphene/Cu and monolayer hBN/Cu were used in all calculations.To calculate the Li diffusion activation energy, we considered the diffusion paths between the possible Li adsorption sites for each substrate (Figure S16), and linear synchronous transit and quadratic synchronous transit methods, as implemented in Dmol 3 , were employed to determine the Li diffusion activation energy. 59,60For the CASTEP calculations, the GGA with the PBE functional 59,61 was used to treat unreactive core electrons, and the energy cutoff was set to 400 eV.The "two-point steepest descent" algorithm 62 was used for geometry optimization, and the convergence thresholds for geometry optimization and SCF density convergence were 1 × 10 −5 and 1 × 10 −6 eV/atom, respectively.The convergence precision of geometry optimization for maximum force, displacement, and maximum stress was set to 0.03 eV Å −1 , 0.001 Å, and 0.05 GPa, respectively.For bulk structure optimization, a Monkhorst−Pack 63 k-point mesh was used and set to 8 × 8 × 8 for the bulk unit cell.
Surface Energy Calculations.The surface energy was calculated by using a slab model with a sufficiently large vacuum thickness (20  Å) to avoid interactions between periodic images.The surface energy (γ) can be expressed as where E slab is the total energy the relaxed slab model, E Bulk is the bulk total energy, N s is the number of atoms in the slab unit, N b is the number of atoms in the bulk unit, and S is the surface area.The coefficient 2 is used here because of the two equivalent surfaces in the slab model.Interfacial Energy Calculations.The interfacial energy calculations are based on the works of Liu et al. and Wang et al. 35,36 Briefly, the interfacial energy was evaluated using the following scheme.First, the full relaxation (atomic coordinates and cell-vector relaxations) of the external stress-free states was applied to the interface systems to obtain the total energy of the interface system (i.e., Li (111) on the current collector surface (monolayer hBN/Cu or monolayer graphene/Cu) E CC/Li ).Subsequently, both the current collector surface (CC) and Li (111) bulk structures with identical shapes and numbers of atoms, as used in the full relaxation step, were subjected to normal (z) direction relaxation (i.e., the in-plane (x and y) lattice parameters were kept fixed).Lastly, the interfacial energy was calculated using

=
where E CC/Li(xyz) is the fully relaxed total energy of the interfacial structure; E CC(z) and E Li(z) are the energies of the pure CC and Li (111) bulk structures (with the same number of atoms as in the interface) after constrained relaxation along the interface normal direction (z direction) and fixed x and y lattice parameters, respectively.A coefficient of 2 was used because of the two equivalent interfaces in the periodic system.
Fabrication of the monolayer hBN/Cu (SEM, TEM, and Raman); theoretical calculations (surface energy, adsorption energy, and interfacial energy); characterization (XPS and TEM); electrochemical analysis; and supporting Table S1 (comparison of the electrochemical performance) (PDF)

Figure 1 .
Figure 1.Preparation and characterization of monolayer hBN/Cu.Schematic representation of the Li deposition behavior on (a) pristine Cu and (b) monolayer hBN/Cu.The inset image of (a) illustrates the formation process of Li dendrites, where planar Li diffusion is restricted on the Cu surface owing to a high diffusion activation energy, and subsequent Li atoms are preferentially deposited on the adsorbed Li clusters.The inset image of (b) depicts the formation process of a dense Li film, as Li atoms can move freely on the hBN surface owing to the low diffusion activation energy.(c) Photograph of the monolayer hBN (3 × 6 cm) transferred to a SiO 2 (300 nm)/Si substrate and the Raman mapping images of the E 2g peak intensity for the monolayer hBN at six marked positions are measured and the mapping area at each position is 4 × 4 μm.(d) AFM image of an edge of hBN transferred to a SiO 2 (300 nm)/Si substrate.(e) Electrical resistivity of the pristine Cu and monolayer hBN/Cu.(f) Calculated Li diffusion activation and adsorption energies of the pristine Cu, monolayer graphene/Cu, and monolayer hBN/Cu.(g) Average interfacial energies (σ) of the pristine Cu, monolayer graphene/Cu, and monolayer hBN/Cu.

Figure 2 .
Figure 2. Inorganic-rich, homogeneous SEI layer formed on monolayer hBN/Cu.Cross-sectional transmission electron microscopy (TEM) images of (a) pristine Cu, (b) monolayer graphene/Cu, and (c) monolayer hBN/Cu.XPS profiles of (d) pristine Cu, (e) monolayer graphene/Cu, and (f) monolayer hBN/Cu.(g) SEI chemical composition obtained from the XPS F 1s and C 1s spectra on each current collector.The inorganic components originated from Li−F and Li−CO 3 , and the remaining components were organic components.The XPS profiles and TEM images were analyzed using galvanic cells after storage for 250 h at a substrate potential of 0 V (vs Li + /Li).(h) Fitted plots of Q loss vs current density.(i) Current density vs time curves showing the galvanic corrosion rate of Li on the different substrates.

Figure 3 .
Figure 3. Li deposition and corrosion on monolayer hBN/Cu.(a) Voltage profiles of Li deposition on pristine Cu, monolayer hBN/Cu, and monolayer graphene/Cu at a current density of 0.01 mA cm −2 .(b) Li nucleation overpotentials of pristine Cu, monolayer graphene/Cu, and monolayer hBN/Cu at different current densities (0.01, 0.1, 0.5, and 1 mA cm −2 ).(c) Li nucleation overpotentials of different current collectors with and without monolayer hBN at a current density of 0.01 mA cm −2 .SEM images of (d) pristine Cu, (e) monolayer graphene/ Cu, and (f) monolayer hBN/Cu after Li deposition at a current density of 1 mA cm −2 and an areal capacity of 5 mAh cm −2 .The insets of (d), (e), and (f) show the top-view morphologies.(g) Coulombic efficiency of the monolayer hBN/Cu, monolayer graphene/Cu, and pristine Cu at a current density of 1 mA cm −2 and an areal capacity of 1 mAh cm −2 at various interval rest times of 0.1, 1, and 12 h between Li deposition and stripping.(h) Initial Coulombic efficiency of the monolayer hBN/Cu, monolayer graphene/Cu, and pristine Cu as a function of the interval rest time (0.1, 1, 12, 50, 100, and 200 h) at a current density of 1 mA cm −2 and an areal capacity of 1 mAh cm −2 .(i) Change in the electrochemical impedance spectroscopy (EIS) profiles of the asymmetric cells (Li deposited substrate||Li metal) as a function of the interval rest time at a current density of 1 mA cm −2 and an areal capacity of 1 mAh cm −2 .

Figure 4 .
Figure 4. Electrochemical superiority of the Li-metal and anode-free full cells with the monolayer hBN/Cu.(a) Cycling performance of the Li-metal full cells (pristine Cu vs monolayer hBN/Cu) at charge/discharge current densities of 0.4/4 mA cm −2 , in which the full cells were composed of LFP cathodes (areal capacity = 3.7 mAh cm −2 ) and precharged monolayer hBN/Cu or pristine Cu anodes (areal capacity = 3.7 mAh cm −2 ), resulting in an N/P ratio of 1.(b) Cycling performance of the anode-free full cells (pristine Cu vs monolayer hBN/Cu) at charge/discharge current densities of 0.4/0.8mA cm −2 .(c) Local electrochemical impedance spectra (LEIS) area scan mapping images showing the local R ct of the pristine Cu and monolayer hBN/Cu after 50 cycles in the anode-free full cells.(d) Mean and standard deviation of the local R ct of the pristine Cu and monolayer hBN/Cu obtained from (c).