A Li metal-SiO_x hybrid anode enabling synergistic plating/alloying dual-mechanism lithium storage

Abstract

Balancing specific energy and cycle durability remains a critical challenge for practical battery systems, as conventional single-mechanism anodes struggle to optimize both simultaneously. Here, we engineer a symbiotic Li-SiO_x hybrid anode enabling spatiotemporally coordinated Li metal plating and Li-ion alloying. The preferential lithiated Li_xSi forms a Li⁺-conductive network, guiding and confining lithium nucleation beneath the SiO_x layer. This precise control over Li plating replenishes active lithium inventory, stabilizing electrochemical reactions and minimizing Li loss. When paired with a LiNi_0.8Co_0.1Mn_0.1O₂ cathode (4 milliampere-hours per square centimeter), the full cell achieves 50% higher specific energy than SiO_x-based lithium-ion batteries while retaining 80% capacity after 900 cycles at a lean negative/positive ratio of 1.5, outperforming state-of-the-art Li-ion and Li-metal battery systems. A pouch cell exhibits stable cycling over 600 cycles at 0.5C, validating its practicality. Our findings pioneered a universal hybrid anode design paradigm with coupled reaction mechanisms, effectively addressing the longstanding energy durability trade-off in next-generation batteries.

INTRODUCTION

Optimizing power cells for electric vehicles (EVs) demands a focus on maximizing total service life (1–4). For a defined service life requirement, e.g., achieving a driving benchmark of 1 million kilometers (Fig. 1A), the cycle life of the battery is inversely proportional to the range per charge, which is directly governed by the specific energy of the battery (5, 6). High-capacity anodes, such as silicon-based materials or lithium (Li) metal paired with high-nickel cathodes like LiNi_0.8Co_0.1Mn_0.1O₂ (NCM811), deliver high specific energy (7–12), potentially reducing reliance on fast-charging infrastructure (13–15). However, these systems face notable challenges in achieving sufficient cycle life required to realize extended ranges. Silicon-based anodes suffer from low initial Coulombic efficiency and substantial volume fluctuations (16–18), while Li metal anodes are plagued by dendrite formation and the accumulation of “dead” Li (7, 19, 20), both of which degrade cycling performance. Consequently, batteries using high-capacity anodes with NCM811 cathodes still struggle to meet the cycle life requirements of 1000 cycles, necessary to sustain a practical range of 1 million kilometers for EVs. To address this challenge, there is a critical need to develop anode materials that simultaneously deliver high specific energy and long cycle life.

Fig. 1. Design target and synergistic plating/alloying lithium storage of symbiotic Li/SiO_x hybrid anode.

(A) Relationship between single driving range (proportional to specific energy) and the cycle number required for power batteries to achieve 1 million kilometer total service mileage for a 350-kg battery pack, benchmarked against state-of-the-art anode-cathode configurations. Schematic representation of Li metal-silicon hybrid anode modes: (B) parasitic (blocked Li⁺ transfer pathway) and (C) symbiotic (continual Li⁺ conduction pathway) mechanisms during electrochemical cycling.

Hybrid anode systems have emerged as a promising strategy to circumvent the inherent compromises of single-component anodes (21–28), attracting notable scientific and industrial attention. Silicon-graphite hybrids exemplify this strategy, using the synergistic interplay between high theoretical capacity (3579 mAh g⁻¹ for Li_3.75Si) of silicon and high structural stability of graphite. This integration enhances specific energy while preserving cycling stability, which is a critical advancement for next-generation lithium-ion batteries (LIBs) (29–31). Parallel innovations include Li metal-graphite hybrid, which exploit a dual intercalation-plating mechanism to achieve ∼10% higher cell-specific energy than state-of-the-art LIBs (21, 25, 28, 32–34). In these systems, the porous architecture of the graphite electrode acts as a scaffold for spatially controlled Li deposition (33), while surface functionalization or electrolyte engineering mitigates dendritic growth and improves the reversibility of Li plating/stripping (21, 34, 35).

Further progress may arise from alloy-plating hybrid anodes that integrate Li metal with alloying materials such as silicon. However, these dual high-capacity systems face critical limitations due to challenges in simultaneously using the full electrochemical capacity of both components (36, 37). Unlike graphite, silicon-based electrodes are structurally constrained by limited pore volume and undergo severe volumetric expansion during lithiation (e.g., >300% for Li_3.75Si). These limitations restrict the spatial accommodation of lithium deposition, driving preferential nucleation and growth of Li metal on silicon surfaces (elucidated below). This inhomogeneous deposition fosters electrochemically isolated dead Li (38), which impedes ion transport pathways to silicon particles (39, 40), and exacerbates rapid capacity degradation, which are termed as parasitic hybrid anodes (Fig. 1B). Therefore, achieving simultaneous capacity utilization in dual high-capacity anode systems requires precise regulation of Li nucleation kinetics and growth pathways, which enables spatiotemporal coordination of lithium storage between silicon and Li metal.

Here, we present a Li-SiO_x hybrid anode that integrates a LiAg alloy–modified patterned Cu current collector with a SiO_x-coated separator, denoted as symbiotic hybrid anode. This design enables precisely controlled Li nucleation/growth via multiphysical field gradient modulation (balancing Li-ion flux distribution, electrical potential gradients, and lithiophilicity). Electrochemical evaluations and postmortem analysis validate the spatiotemporally coordinated dual Li storage mechanism. The preferential lithiated Li_xSi phase constructs a continuous Li-ion conductive pathway, guiding and spatially confining lithium nucleation/growth beneath the SiO_x layer, thus enabling full-cell operation with high capacity/voltage output while preserving SiO_x capacity utilization (Fig. 1C). Simultaneously, the orderly deposited Li metal dynamically offsets systemic Li losses upon cycling, stabilizing SiO_x electrochemistry and voltage hysteresis, thus sustaining long-term cyclability. As a result, full cell with synergetic dual-mechanism delivers 50% specific energy higher than conventional LIBs (projected by 10-Ah pouch cell with a specific energy of ∼470 Wh kg⁻¹). When paired with a high-loading NCM811 cathode (20 mg cm⁻²), the full cell demonstrates exceptional cycling stability, retaining 80% capacity retention over 900 cycles, outperforming those LMBs. We also assembled 300-mAh pouch cells coupling symbiotic Li-SiO_x hybrid anode with NCM811, which achieve stable cycling performance over 600 cycles under realistic conditions. This hybrid anode design outperforms state-of-the-art LIB or LMB battery systems in both energy density and cyclability, highlighting the viability of symbiotic Li storage mechanisms in meeting the stringent energy and cycle life requirements for next-generation EVs.

RESULTS

Design of Li metal-SiO_x symbiotic hybrid anode

The hierarchical-structured symbiotic hybrid anode is composed of two principal components: a SiO_x-contained separator and a Li metal composite with Ag-modified patterned Cu current collector, as illustrated in Fig. 2A. The SiO_x layer is fabricated by coating a polyethylene separator with a slurry comprising SiO_x powder, carbon nanotubes, and polyacrylic acid. Scanning electron microscopy (SEM) reveals a homogeneous, porous morphology (fig. S1), enabling high areal capacities up to 5 mAh cm⁻² with stable electrochemical activity (fig. S2). The porous architecture effectively accommodates the volumetric expansion of SiO_x (fig. S3 and note S1) while simultaneously endowing the SiO_x layer with robust mechanical integrity and cycling stability (fig. S4). The Li composite layer is prepared by modifying a patterned copper with a nanosilver (nano-Ag) coating via a solution-exchange reaction (fig. S5), followed by integrating Li to form Li/LiAg (fig. S6). Upon electrolyte infiltration, the SiO_x layer undergoes in situ lithiation, reacting with the Li/LiAg to form a partially lithiated Li_ySiO_x phase, accompanied by controlled consumption of the Li reservoir. Crucially, the SiO_x capacity is designed to match the areal capacity of cathode, while the Li metal reservoir dynamically offsets irreversible Li loss and actively participates in electrochemical cycling. Thus, the loading of Li metal can be precisely tailored on the basis of the Li-to-SiO_x ratio within the hybrid anodes (see note S2).

Fig. 2. Schematic illustration and the voltage profiles of the dual-mechanism symbiotic hybrid anode.

(A) Schematic of the hierarchical structure of the symbiotic Li-SiO_x hybrid anode. (B) Schematic diagram of the synergistic dual-mechanism Li storage of the hybrid anode. (C) Voltage profiles of NCM811-based full cells using Li-SiO_x hybrid anodes versus SiO_x-only anodes.

The synergistic dual plating-alloying Li storage mechanism of the Li-SiO_x hybrid anode was systematically investigated. As illustrated in Fig. 2B, during lithiation, the upper layer of prelithiated Li_ySiO_x undergoes progressive alloying to form Li_y+zSiO_x. Notably, the Li_xSi phase serves as a conductive scaffold, directing spatially confined and kinetically controlled Li metal deposition at the Cu-SiO_x interface (discussed in detail later). Li ions are preferentially extracted from the Li metal phase before dealloying from Li_y+zSiO_x, ensuring sustained high-voltage output in full-cell configurations (fig. S7). Critically, the high-potential Li_xSi phase acts as a self-compensating Li reservoir, mitigating systemic Li loss and enabling long-term cyclability by dynamically replenishing active Li. Comparative full-cell evaluations with high-nickel NCM811 cathodes demonstrate the superiority of our hybrid anode. The hybrid system achieves an average discharge voltage of 3.59 V, a marked increase over the 3.43 V observed for conventional SiO_x anodes (Fig. 2C). This voltage enhancement, coupled with the elimination of first-cycle irreversible capacity loss (attributed to the Li replenishment), translates to a 14% improvement in specific energy. In addition, the patterned Cu current collector reduces collector mass by 64% compared to conventional Cu foil, notably minimizing inactive components. Integrating these advances with the dual-mechanism synergy, the hybrid anode delivers a 50% increase in cell level–specific energy relative to state-of-the-art SiO_x-based LIBs.

Spatially confined Li deposition site in symbiotic hybrid anode

As previously emphasized, the nucleation site and deposition dynamics of Li metal in composite anodes remain pivotal challenges due to the hostless nature of Li plating. Achieving synergistic integration of alloying and plating mechanisms demands precise spatial control over Li nucleation and deposition processes. To this end, we systematically investigated Li deposition behavior in two architectures, a parasitic hybrid anode (p-Li/SiO_x) formed by a SiO_x anode (1.5 mAh cm⁻² SiO_x directly coated on separator and Cu foil) with 1.5 mAh cm⁻² Li lithiated and the symbiotic hybrid anode (s-Li/SiO_x). SEM analysis (Fig. 3, A and B) reveals obvious contrasts in Li deposition patterns. In p-Li/SiO_x, Li preferentially deposits atop the SiO_x layer (fig. S8), forming parasitic Li which evolves into electrochemically inactive dead Li after cycling. In contrast, the layered structure directs lithium deposition beneath the SiO_x layer within the s-Li/SiO_x system, as supported by imaging data from different regions of one sample and from separate independent samples (fig. S9). Furthermore, this deposition behavior is corroborated by in situ optical microscopy and x-ray computed tomography (CT) imaging (figs. S10 and S11 and notes S3 and S4). To elucidate the underlying drivers of this behavior, we analyzed the nucleation overpotential, which is the critical determinant of Li deposition sites, between the two composite anode architectures. Li deposition sites are governed by nucleation overpotential, dictated by three key resistances: electric resistance between deposition sites and the current collector (R_e), Li-ion transport resistance from separator to deposition sites (R_Li), and charge-transfer resistance at the electrode-electrolyte interface (R_ct). These resistances are influenced by local current density, Li-ion concentration, and local potential, respectively. Lower R_e, R_Li, and R_ct collectively reduce nucleation overpotential, favoring deposition at sites with minimized resistance. Figure 3 (A and B) illustrates a schematic comparison of the three resistance values at the interface between the above and beneath of the SiO_x. In p-Li/SiO_x, the impedance disparity arises from the differences in ionic and electronic concentrations on the upper and lower surfaces. In conventional architectures, the variation in conductivity is insufficient to counteract the disparities introduced by ion transport (fig. S12 and note S5), rendering the influence of R_Li more notable than that of R_e. These combined effects of R_e, R_ct, and R_Li drives Li deposition above the SiO_x layer (Fig. 3A). In s-Li/SiO_x, the patterned Cu current collector increases the initial Li⁺ concentration near the current collector side, and the rapid Li-ion transport on the Li_xSi alloy (16) ensures a uniform ionic concentration across both its upper and lower surfaces, thereby homogenizing the overall Li⁺ concentration distribution within the electrode. Under the electrolyte potential gradient, the Li⁺ flux is effectively guided toward the current collector side of the SiO_x layer, where electronic conductivity is also highest, creating a spatial overlap of ionic and electronic transport (Fig. 3C). This results in comparable R_Li values at the interface beneath and above the SiO_x layer. In this scenario, Li deposition is predominantly influenced by R_e and R_ct. Combined with a reduced R_e and R_ct at the Cu-SiO_x interface, Li preferentially deposits beneath the SiO_x rather than above it (SEM images in Fig. 3B and figs. S4 and S9). This process is further validated by in situ electrochemical impedance spectroscopy, as shown in fig. S13 and discussed in note S6. Besides, density functional theory (DFT) calculations quantify the thermodynamic and kinetic preference for Li adsorption, charge redistribution, and work function at LiAg and Li_ySiO_x interfaces (Fig. 3D and note S5). Despite the strongest adsorption, Li-Li₄SiO₄ shows poor nucleation due to highly localized electron transfer from Li to O, high Li diffusion barrier, and a large interfacial potential difference, which hinders uniform Li deposition. In contrast, Li-LiAg exhibits delocalized interfacial electrons and a low, uniform work function, promoting homogeneous nucleation. Li-Li₁₅Si₄ shows intermediate behavior, with moderate charge transfer and electrostatic heterogeneity. In addition, the main lithiation products of SiO_x are Li₁₅Si₄ and Li₄SiO₄; thus, Li nucleation on lithiated SiO_x is expected to be less favorable and associated with a higher nucleation overpotential compared to LiAg, establishing that Li nucleation is energetically favored beneath the SiO_x layer. These trends are fully consistent with the experimental observations (8 mV versus 14 mV; fig. S14). Figure 3E presents the Li deposition morphology on the current collector and the current collector side SiO_x layer (corresponding to fig. S9), together with the corresponding energy-dispersive x-ray spectroscopy (EDS) mapping. It can be observed that Li metal predominantly grows in the region where the SiO_x layer contacts the current collector. To investigate the role of the SiO_x layer in the symbiotic Li deposition process at the anode, we used the symbiotic hybrid anode lithiated with ⁶Li, followed by ⁷Li deposition. Upon disassembling the cells and extracting the electrolyte for inductively coupled plasma analysis (Fig. 3F), the results revealed a higher proportion of ⁶Li in the electrolyte from cells of symbiotic hybrid anode than natural abundance. This finding suggests that Li ions from the ⁶Li_ySiO_x effectively participates in the Li deposition process on the collector side through efficient ion percolation and conduction (refer to the schematic diagram in Fig. 3F). This highlights the essential role of fully lithiated SiO_x in enabling spatially directed Li deposition.

Fig. 3. Spatially confined Li deposition site in symbiotic hybrid anode and its impact on full-cell electrochemical performance.

Comparative analysis of Li deposition resistance (ionic resistance, electronic resistance, and charge transfer resistance) in (A) parasitic and (B) symbiotic hybrid anodes, with cross-sectional SEM images of the anode after 2 mAh cm⁻² Li plating. (C) Simulation of Li⁺ concentration distribution, electric potential distribution, and localized current density for the s-Li/SiO_x anode. (D) The adsorption energy, charge density difference, and work function of Li₄SiO₄ (010), LiAg (110), and Li₁₅Si₄ (111). (E) SEM and EDS mapping images of the current collector and the current–collector side SiO_x coating after 2 mAh cm⁻² of Li plating. (F) The ⁶Li/⁷Li ratio in the electrolyte after Li plating on ⁶Li full-lithiated symbiotic hybrid anode and natural abundance, with a schematic of the mechanism. (G) Full-cell cycling stability, (H) voltage evolutions, and (I) partitioned capacity contributions (SiO_x versus Li) quantified via dV dQ⁻¹ analysis for parasitic and symbiotic hybrid anode.

The two hybrid anodes were then paired with the NCM811 cathode to evaluate the performance comparison in full cells. As shown in Fig. 3G, although both configurations show comparable initial capacity (the areal capacities of SiO_x and Li are both 1.5 mAh cm⁻²), the full cell of parasitic hybrid anode degrades rapidly after 40 cycles, retaining only 62% capacity after 100 cycles, whereas symbiotic hybrid anode maintains stable cycling performance. To further elucidate the capacity evolution of SiO_x and Li metal within the hybrid anode, differential voltage analysis (dV dQ⁻¹) was used to distinguish the contributions of SiO_x and Li metal. The voltage profiles in Fig. 3H reveal that, because of the difference in the potential of Li metal and SiO_x, Li metal preferentially delithiated first, followed by SiO_x. During the delithiation of SiO_x, a sharp drop in voltage is observed, which corresponds to characteristic peaks in the dV dQ⁻¹ curves (fig. S15). The left side of these peaks reflects the capacity contribution from Li metal, whereas the right side represents that from SiO_x. Figure 3I illustrates the fluctuations in the capacities of Li metal and SiO_x over cycling, as determined by the dV dQ⁻¹ analysis (normalized to the full-cell capacity of the first cycle). For the symbiotic hybrid anode, the capacity of Li gradually decreases, accompanied by a slow increase in the cyclic capacity of the SiO_x to maintain a constant full-cell capacity. Subsequently, the reversible capacities of both Li and SiO_x remain stable. In contrast, for parasitic hybrid anode, the capacity of Li declines slowly, accompanied by a gradual increase in the SiO_x capacity during the initial cycles. However, after 50 cycles, the capacities of both SiO_x and Li decrease notably, consistent with the rapid capacity decay observed in the full cell. To further elaborate the underlying reason, postmortem SEM analysis of the hybrid anode after cycling was conducted. For the parasitic hybrid anode, substantial accumulation of dead Li was observed at the interface between the separator and the SiO_x layer (fig. S16). These inactive Li deposits severely impede Li⁺ transfer to the SiO_x underneath, leading to a decline in capacity of SiO_x. With SiO_x inactivated, Li continues to deposit in an irregular pattern, forming parasitic Li. This vicious cycle ultimately leads to rapid Li loss and capacity decay in the full cell (fig. S17). In contrast, for the symbiotic hybrid anode, Li deposit more uniformly at the SiO_x-Cu interface, resulting in a notably reduced impact on the cycling performance of SiO_x. Initial Li capacity decline correlates with SEI formation and irreversible SiO_x, after which symbiotic Li-SiO_x cycling stabilizes (fig. S18). These findings demonstrate that the hierarchical structure constructs a gradient in the nucleation overpotential, which enables spatially controlled Li nucleation and deposition pathways. More crucially, the lithiated Li_xSi phase acts as a fast ion-conducting scaffold, guiding Li metal to preferentially nucleate at the Cu-SiO_x interface. This well-organized Li plating/stripping, simultaneously, stabilizes the SiO_x (de)-alloy electrochemistry by constantly replenishing Li loss upon cycling. The resultant feedback loop creates synergistic effect of dual Li storage mechanisms, enabling a self-reinforcing interplay between alloying-plating mechanisms.

Precisely controlled Li growth in symbiotic hybrid anode

Beyond controlling nucleation sites, precise regulation of Li deposition dynamics is critical to achieve high electrochemical reversibility of Li plating/stripping processes. To investigate the factors affecting Li deposition kinetics, we conducted a comparative postmortem analysis of Ag-modified hybrid anodes (Li/SiO_x-Ag) and unmodified counterparts (Li/SiO_x) after Li plating. Optical image (Fig. 4A) reveals distinct spatial Li distribution in Li/SiO_x-Ag. On the SiO_x side, Li preferentially deposits at the interface between the SiO_x layer and the patterned Cu, while the Cu side exhibits uniform surface deposition. SEM further demonstrates that Li forms smooth, densely packed layers on both surfaces of Li/SiO_x-Ag, with no evidence of dendritic Li. X-ray photoelectron spectroscopy (XPS) depth profiling (Fig. 4B and fig. S19) elucidates an intriguing, directional Li growth mode. On the SiO_x side, the progressive increase in inorganic components (LiF and Li₃N) and concurrent decline in organic species (C-C, C-O, and polymer-F) with etching depth confirm a top-down Li growth pattern. Conversely, on the Cu side, the gradual increase of inorganic components diminishment of organic species indicates bottom-up Li deposition. Notably, persistent LiAg signals observed throughout the etching profiles, indicating the solid-state Li-Ag interdiffusion upon cycling which continually modulates Li deposition even under deep (de)-lithiation (41). Time-of-flight secondary ion mass spectrometry (ToF-SIMS) depth profiling of the SiO_x surface (Fig. 4C) reveals the coupled evolution of SEI formation and alloy diffusion. CN⁻ and F⁻ signals, characteristic of inorganic SEI components, exhibit surface enrichment followed by attenuation, consistent with the formation of a passivating interphase, whereas Li⁻ and Ag⁻ display nearly identical depth profiles, indicating concurrent Li deposition and Ag diffusion. Spatial mapping shows that Ag⁻ is segregated from Si⁻, while Li⁻ overlaps with both regions, confirming that Li resides in both LiAg and lithiated SiO_x (Li_ySiO_x), with SEI species uniformly distributed on both phases. Depth-resolved analysis (Fig. 4D) further indicates Ag diffusion from the current collector into the electrode, while three-dimensional reconstruction demonstrates homogeneous overlap of Li⁻, Ag⁻, evidencing uniform alloy formation and SEI confinement to Li-containing domains. Together, these results substantiate a stable, spatially coupled dual-mechanism lithium storage enabled by Ag-assisted Li deposition and active SiO_x participation. In contrast, Li/SiO_x anodes exhibit inhomogeneous Li deposition on the SiO_x surface, characterized by dendritic morphologies (fig. S20), whereas there is no Li deposition on the Cu side due to insufficient lithiophilicity.

Fig. 4. Precisely controlled Li growth in symbiotic hybrid anode and its impact on full-cell electrochemical performance.

In-depth XPS analysis of the surface of (A) Cu and (B) SiO_x side, along with corresponding optical images and SEM images after 10 cycles of the symbiotic hybrid anode. ToF-SIMS elemental depth profiles and three-dimensional spatial distribution maps of (C) SiO_x and (D) Cu side. (E) Cycling performance of the Li||Li/SiO_x and Li||Li/SiO_x-Ag half cells. (F) Comparison of cycling performance of NCM811||Li/SiO_x-Ag and NCM811||Li/SiO_x full cells. (G) Potential variations of the anode and cathode in full-cell upon cycling, analyzed using a three-electrode configuration.

We further examined the influence of precise control over lithium growth on achieving high lithium utilization and its subsequent favorable implications for the performance of the full-cell configuration. Half-cells using Li/SiO_x-Ag and Li/SiO_x anodes were paired with Li foil counter-electrodes. As shown in Fig. 4E, the Li/SiO_x-Ag anode exhibits a slower overall Li loss in the system, whereas the Li/SiO_x anode experiences rapid Li loss. Systematic Li inventory analysis in full-cell configurations further demonstrates that the Li/SiO_x-Ag system minimizes systemic Li loss during cycling. Within this architecture, cyclable Li metal participates in reversible plating/stripping, while excess Li is retained within the Li_ySiO_x matrix as a self-replenishing reservoir, dynamically compensating for irreversible losses. In contrast, the Li/SiO_x anode suffers uncontrolled Li depletion, leading to diminished Li inventory and accelerated capacity degradation (fig. S21). In addition, we investigated the lithium deposition behavior of the Li/SiO_x-Ag anode under high current densities. As shown in fig. S22, stable lithium plating/stripping is maintained even at a current density of 4 mA cm⁻². SEM analysis of the current collector side in the s-Li/SiO_x system (fig. S23A) shows that lithium continues to deposit predominantly and uniformly on the Cu-Ag substrate even at 4 mA cm⁻², forming a finer and denser granular morphology compared with low-current conditions. Local grain coarsening is observed near the SiO_x/current collector/electrolyte three-phase interface, attributable to accelerated Li⁺ consumption and localized depletion under high current density. In contrast, sustained ion supply within the patterned Cu supports continued crystal growth at open three-phase interfaces. Additional SEM images of the SiO_x-separator interface reveal limited lithium deposition in the form of sparse, large spherical particles (fig. S23B). By regulating local current distribution and promoting homogeneous lithium nucleation, the dual-mechanism architecture minimizes separator-side lithium accumulation and suppresses dendritic propagation, thereby effectively mitigating short-circuit risks under high-rate operation. To understand the effect of Li depletion rate on the full cell, we performed cycling tests on three-electrode full cells by coupling Li/SiO_x-Ag and Li/SiO_x anodes with NCM811 cathodes (the voltage window is 2.8 to 4.3 V). The Li/SiO_x-Ag system configuration exhibited exceptional stability, retaining >90% capacity after 160 cycles, while the unmodified Li/SiO_x suffered rapid capacity degradation, with capacity retention falling below 70% after 160 cycles (Fig. 4F). Anode and cathode potential profiles extracted from the full-cell configuration elucidate the mechanistic divergence between the two systems (Fig. 4G). In NCM811||Li/SiO_x-Ag, the directionally controlled Li deposition minimizes systemic Li losses, anchoring the anode lithiation endpoint at 0 V. The delithiation potential gradually increases, reflecting the progressive activation of high-potential Li reservoirs within LiSiO_x, which dynamically compensate for irreversible losses (SEI formation and SiO_x lithiation), consistent with capacity compensation observed in Fig. 4G. Under a fixed full-cell cutoff voltage (4.3 V), the cathode endpoint charging potential stabilized at 4.3 V. In NCM811||Li/SiO_x, the unregulated Li deposition induces cumulative Li losses, progressively shifting the anode lithiation endpoint to 0.2 V. In this scenario, the cathode charging cutoff potential rises to 4.5 V, triggering overcharging. This destructs the cathode structure (e.g., Ni³⁺ migration and lattice oxygen release) and accelerates electrolyte oxidation, exacerbating Li depletion and capacity fade. As shown in fig. S24, the secondary particles of the cathode in the NCM811||Li/SiO_x cell undergo pronounced fracture after cycling, whereas the cathode particles in the NCM811||Li/SiO_x-Ag cell remain structurally intact. Thus, the reversible Li storage behavior of Li metal stabilizes voltages of both anode and cathode, thereby enhancing the cycling stability of the symbiotic hybrid anode.

Practical validation of symbiotic hybrid anodes

Building on the superiority of symbiotic Li storage, we evaluated symbiotic hybrid anodes in a practical full-cell configuration pairing a high-loading NCM811 cathode (4 mAh cm⁻²) with the hybrid anode (4 mAh cm⁻² SiO_x and 2 mAh cm⁻² Li). This design optimizes specific energy while balancing Li inventory (see calculation details in fig. S25 and note S7). The full cell achieves exceptional long-term cyclability, retaining 80% capacity over 900 cycles (Fig. 5A), surpassing state-of-the-art Li metal battery benchmarks. Figure 5B presents the capacity-voltage evolution of full cell with s-Li/SiO_x, revealing distinct contributions from alloying (SiO_x) and plating (Li) mechanisms during cycling. dV dQ⁻¹ analysis (Fig. 5C) further resolves the synergistic interplay between the Li-ion and Li-metal dual storage modes. The roll of Li metal storage mechanisms can be categorized into three distinct stages (detailed in note S8). Initially, a large portion of Li metal capacity is used for irreversible consumption of active materials and the formation of the SEI, resulting in a capacity decay rate of 0.8% per cycle for Li metal. In the intermediate symbiotic cycling stage, the decay rate of Li metal drops to 0.12% per cycle as SiO_x buffers Li loss, concomitant with the release of lithium from the high-potential SiO_x reservoir and activation of SiO_x. After 150 cycles, the contribution from Li metal gradually diminishes as it is largely consumed to replenish the Li losses of the system, leading to a further reduction in the decay rate to 0.01%. These findings demonstrate the dual-mechanism synergistic Li storage that, during the initial cycling stages, the capacity of the hybrid anode is jointly contributed by Li metal and SiO_x, and a portion of Li is retained within the SiO_x. As cycling progresses, SiO_x emerges as the primary capacity contributor, with lithium from the high-potential SiO_x reservoir gradually released to compensate for ongoing Li losses, thereby achieving an exceptional cycling stability. Hence, this spatiotemporally coordinated behavior mitigates Li depletion while maximizing specific energy, constructing a paradigm for sustainable high–energy density batteries. By contrast, the cycling performances of individual SiO_x and LiAg-Cu electrodes were systematically evaluated, demonstrating that the integrated Li-SiO_x configuration delivers markedly superior cycling stability compared with either component alone (fig. S26). In addition, a parasitic Li/SiO_x hybrid anode paired with the high-loading NCM811 cathode was also tested (fig. S27), which further underscores the critical importance of the symbiotic Li/SiO_x anode design in achieving stable long-term cycling. To assess practical viability, a practical pouch cell coupling NCM811 cathode was constructed, which delivers stable cycling over 600 cycles at 0.5C, confirming the practical implementation of symbiotic Li-SiO_x hybrid anodes (figs. S28 and S29). Consequently, the symbiotic hybrid anode design delivers excellent cycling stability under harsh cycling conditions at high areal capacity and low N/P ratio.

Fig. 5. Performance of NCM811||s-Li/SiO_x full cells under a realistic test condition.

(A) Cycling performance of full cells with 4 mAh cm⁻² NCM811 coupled with symbiotic hybrid anode (the areal capacities of SiO_x and Li are 4 and 2 mAh cm⁻², respectively, with an N/P ratio of 1.5). (B) Voltage profiles and (C) corresponding capacity changes of SiO_x and Li during cycling for the NCM811||symbiotic hybrid anode full cell. (D) Comparison of total anode specific capacity versus cycle number from recent studies (29, 42–52) (circle area represents the N/P ratio; see table S2 for details).

These results are benchmarked against recent studies of full cells using high-loading cathodes with single alloy/plating mechanism silicon-based or Li-metal anodes (Fig. 5D). To enable a direct comparison of anode contributions to full-cell specific energy, we introduced a normalized anode total capacity metric, which accounts for both inactive mass and the N/P ratio (calculations in table S1). The symbiotic Li-SiO_x hybrid anode achieves a total specific capacity of 845 mAh g⁻¹ (fig. S30), which is 30% higher than that of the stand-alone SiO_x anode and comparable to 20-μm Li-Cu anodes. Consequently, a 10 Ah–level full cell with this hybrid anode achieves a projected specific energy of 470 Wh kg⁻¹ (see note S9 for details). Conventional Li-metal anodes provide high total anode specific capacity (700 to 900 mAh g⁻¹, green dashed box in Fig. 5D) but suffer from severe cycling instability, failing to meet EV endurance targets (Fig. 1A, black dashed line). A silicon-based anode system (Li-ion mechanism, orange dashed box) offers improved stability but sacrifice specific capacity (300 to 500 mAh g⁻¹), with lower-Si loading further stabilizing cyclability. In contrast, our symbiotic hybrid anode resolves these trade-offs through dual-mechanism Li storage, simultaneously delivering high capacity with prolonged cyclability. This dual-mechanism synergy directly addresses the 1-million-kilometer challenge, highlighting hybrid anodes as a transformative pathway for energy-dense, long-life batteries.

DISCUSSION

In this work, we demonstrate a symbiotic Li-SiO_x hybrid anode that achieves the synergistic coupling of alloying and plating Li storage mechanisms. By engineering the hierarchical Li-SiO_x architecture, the preferential lithiated Li_xSi phase constructs a fast Li⁺-conductive pathway, spatially confining Li nucleation and growth beneath the SiO_x layer. Simultaneously, the precisely regulated Li plating stabilizes the SiO_x alloying/dealloying electrochemistry by dynamically replenishing Li inventory to minimize Li losses upon cycling, thereby resolving the trade-offs in high-capacity anodes. The spatiotemporally coordinated dual-mechanism Li storage delivers unique systemic advantages in full-cell configurations. (i) Synergistic utilization of alloying (SiO_x) and plating (Li) mechanisms maximizes specific energy without compromising electrochemical stability. (ii) The high-potential Li_xSi phase acts as Li reservoir dynamically compensating for systemic Li loss, enabling stable long-term cycling. (iii) Precisely controlled Li deposition guarantees high Li utilization, which prevents overcharging and preserves structural integrity of cathode. As a result, full cells using this symbiotic Li-SiO_x hybrid anode paired with a high-loading NCM811 cathode (20 mg cm⁻²) demonstrate exceptional cyclic stability, maintaining 80% capacity retention over 900 cycles. Practical validation via a 300-mAh pouch cell (NCM811||s-Li/SiO_x) confirms scalability, achieving 600 cycles at 0.5C. By surpassing the energy density/durability limits of both LIBs and LMBs, this work addresses EV range anxiety while fulfilling million-kilometer durability targets. More broadly, we pioneered a universal paradigm for hybrid anodes, extensible to Sn-, Si-, and Al-based alloying materials, while using constituent fabrication processes that are all compatible with scalable manufacturing, thereby providing a practical pathway toward next-generation batteries that unify high energy density, long cycle life, and safety.

MATERIALS AND METHODS

Preparation of the symbiotic Li-SiO_x hybrid anode

The Li-SiO_x hybrid anode is constructed from a copper substrate with Li deposition (LiAg-pCu) paired with a separator uniformly coated with the SiO_x active material. The copper substrate is available in three forms: patterned copper, silver-modified copper foil, and silver-modified patterned copper. The copper foil is 8-μm thick. The patterned copper is produced via laser perforation, yielding 1-mm pores with a porosity of 67.5%. The silver-plated patterned copper is fabricated using a solution-based method: The deionized water–cleaned patterned copper is immersed in a 0.02 M silver nitrate solution for 5 min, rinsed three times with deionized water, and subsequently dried in a vacuum oven at 60°C. For the comparative experiments, the Li-containing copper substrates were prepared via electrochemical deposition. In a glove box under Ar atmosphere, cells were assembled using 300-μm-thick Li metal foils and Ag-modified patterned Cu or patterned Cu (note that a separator was placed beneath the copper substrate to prevent Li deposition on the stainless steel casing, while the current collector tabs were left in contact with the casing to facilitate electron conduction). Li metal was then deposited at a current density of 0.1 mA cm⁻² for 15 hours by Landt (CT2001A), yielding the LiAg-pCu and Li-pCu substrates, respectively.

Preparation of the SiO_x separator

SiO_x, carbon nanotubes, super P, and poly(acrylic acid) (PAA) were mixed in water in a weight ratio of 85:5:5:5. The mixture was coated onto an alumina-coated polyethylene separator, and the areal capacities are 1.5 and 4 mAh cm⁻².

Preparation of the parasitic Li-SiO_x hybrid anode

SiO_x, carbon nanotubes, super P, and PAA were mixed in water in a weight ratio of 85:5:5:5. The mixture was coated onto the separator, and the areal capacity is 1.5 mAh cm⁻². In a glove box under Ar atmosphere, cells were assembled using 300-μm-thick Li metal foils and SiO_x-Cu (SiO_x-coated separator with Cu foil). The SiO_x-Cu anode was then lithiated at a current density of 0.1 mA cm⁻² for 15 hours by Landt (CT2001A), yielding the p-Li/SiO_x hybrid anode.

Electrochemical measurements

NCM811 cathodes with two different areal capacities were prepared. The lower-capacity cathodes, with an active material loading of 7.5 mg cm⁻², were prepared by mixing NCM811, super P, and polyvinylidene fluoride in an 8:1:1 weight ratio to evaluate anode cycling stability. In contrast, high-capacity cathodes for practical system assessment were prepared by blending the same components at a 95:2.5:2.5 weight ratio, achieving an active material loading of 20 mg cm⁻². Three electrolytes were used as follows: (i) the ester electrolyte, which contained 1 M lithium hexafluorophosphate (LiPF₆) in the mixture of ethylene carbonate, dimethyl carbonate, and diethyl carbonate [1:1:1 in volume] with 5% fluoroethylene carbonate additive, (ii) the ether electrolyte, which contained 1 M lithium bis(trifluoromethanesulphonyl)imide in the dimethoxyethane (DME)], (iii) and the localized high-concentration electrolyte (LHCE) which contained lithium bis(fluorosulfonyl)imide, DME, and 1,1,2,2-tetrafluoroethyl-2,2,3,3-tetrafluoropropylether with a molar ratio of 1:1:3. All cells were assembled in an argon-filled glove box and tested using a Landt (CT2001A) battery tester.

NCM811||p-Li/SiO _x and NCM811||s-Li/SiO _x full cells

The full cells were composed of parasitic hybrid anode (Li/SiO_x-Cu) or symbiotic hybrid anode (SiO_x/LiAg-pCu) (1.5 mAh cm⁻² predeposited Li and the areal of SiO_x is 1.5 mAh cm⁻²), NCM811 (7.5 mg cm⁻²), and 50 μl of ester electrolyte. The cycling test protocol is constant current constant voltage (CCCV) at 0.5C rate (1C = 200 mA g⁻¹) with the voltage window of 2.8 to 4.3 V.

Li||Li/SiO _x and Li||Li/SiO _x -Ag half cells

The half cells were composed of SiO_x/Li-pCu or SiO_x/LiAg-pCu hybrid anode (1.5 mAh cm⁻² predeposited Li and the areal of SiO_x is 1.5 mAh cm⁻²), 300-μm Li foil, and 50 μl of ester electrolyte. The cycling test protocol is galvanostatic charge/discharge (discharge current density: 0.2 mA cm⁻², charge current density: 0.5 mA cm⁻², cycling capacity: 1 mAh cm⁻²).

Li||s-Li/SiO _x and Li||p-Li/SiO _x half cells for Li plating analysis

The half cells were composed of SiO_x/Ag-pCu or SiO_x/Cu hybrid anode (the areal of SiO_x is 3 mAh cm⁻² with 2 mAh cm⁻² prelithiation), 300-μm Li foil, and 50 μl of ether electrolyte. The Li plating deposited at a current density of 0.1 mA cm⁻² for 30 h by Landt (CT2001A).

NCM811||Li/SiO _x and NCM811||Li/SiO _x -Ag three-electrode full cells

The full cells were composed of SiO_x/Li-pCu or SiO_x/LiAg-pCu hybrid anode (1.5 mAh cm⁻² predeposited Li and the areal of SiO_x is 1.5 mAh cm⁻²), NCM811 (7.5 mg cm⁻²), 1 mm–by–20 mm copper foil doubly coated with 1 mm–by–1 mm Li (50-μm thick) as a reference electrode, and 50 μl of ester electrolyte. The cycling test protocol is CCCV at 0.5C rate (1C = 200 mA g⁻¹) with the voltage window of 2.8 to 4.3 V.

High mass load NCM811||s-Li/SiO _x full cell

The full cell was composed of symbiotic hybrid anode (the first-cycle areal capacities of SiO_x and Li are 4 and 2 mAh cm⁻², respectively), NCM811 (20 mg cm⁻²), and 20 μl of LHCE electrolyte (6 g Ah⁻¹) to evaluate the cycling stability of the full cell in practicalization criteria. The cycling test protocol is CCCV at 0.3C rate (1C = 200 mA g⁻¹) with the voltage window of 2.8 to 4.3 V.

NCM811||s-Li/SiO _x pouch cell

The symbiotic hybrid anode was constructed from SiO_x separator (4 mAh cm⁻²) and LiAg-pCu that obtained by rolling Ag-modified patterned Cu with 15-μm Li foil in the dry room (dew point temperature < −40°C). The pouch cell was fabricated in a dry room, with Ag-modified patterned Cu (three pieces, 58 mm by 45 mm), NCM811 cathodes (two pieces, 56 mm by 43 mm), and SiO_x separator (four pieces, 58 mm by 45 mm), The LHCE electrolyte (1 g Ah⁻¹, 360 mg) was injected into the pouch cell in a glove box under the Ar atmosphere. The 0.3-Ah pouch cell was cycled over a voltage range of 2.8 to 4.3 V by Landt (CT2001A) at 0.2C.

Characterizations

The SEM analyses were conducted using a Hitachi Regulus 8100 microscope operating at 10 kV, equipped with an EDS system. Cross-sectional samples were prepared using a Helios Nanolab 600i focused ion beam system. XPS measurements were performed on a Kratos Axis Supra spectrometer with an Al Kα monochromatic x-ray source. The ion composition and distribution of the anode surface and depth sputtering area were characterized by ToF-SIMS (ToF-SIMS 5 ION-ToF GmbH, Münster, Germany). ToF-SIMS was equipped with a 30-keV Bi³⁺ primary ion gun and a 2-keV Cs⁺ sputter gun, and an electron flood gun was used for charge neutralization. X-ray CT was used to investigate the SiO_x-coated separator after Li deposition. The measurements were performed using a ZEISS x-ray microscopy system. By adjusting the magnification, an effective voxel (pixel) size of 1 μm was achieved, enabling three-dimensional visualization of the internal structure and lithium deposition morphology with micrometer-scale spatial resolution. In situ OM investigations were carried out by using an optical microscope (OLYMPUS, OLS-4000) with a 20× objective lens (MPanFLN, numerical aperture value = 0.3). The specific ratios of the ⁶Li and ⁷Li were measured by inductively coupled plasma atomic emission spectrometry (Thermo Icap 6300). The after-deposition cell was washed and diluted by DME.

DFT calculations

All the calculations are performed in the framework of the DFT with the projector-augmented plane-wave method, as implemented in the Vienna ab initio simulation package. The generalized gradient approximation proposed by Perdew, Burke, and Ernzerhof is selected for the exchange-correlation potential. The long range van der Waals interaction is described by the DFT-D3 approach. The cutoff energy for plane wave is set to 500 eV. The energy criterion is set to 10⁻⁵ eV in iterative solution of the Kohn-Sham equation. A vacuum layer of 15 Å is added perpendicular to the sheet to avoid artificial interaction between periodic images. Fermi-Dirac smearing with a width of 0.1 eV was applied to aid in the convergence of electronic states. The Brillouin zone integration is performed using a 3 by 3 by 1 k-mesh. All the structures are relaxed until the residual forces on the atoms have declined to less than 0.03 eV/Å.

For the transition state calculations, the climbing image nudged elastic band (CI-NEB) method is used. Each NEB calculation consists of five images, with forces on the atoms being smaller than 0.03 eV/Å to ensure proper convergence.

The calculations of adsorption energy were conducted as follows

$$E_{\text{ads}-x}=E_1-(E_2+E_x)$$

where E₁ is the total energy of system with X adsorbed, E₂ is the energy of Li₁₅Si₄ Li₄SiO₄ or LiAg system, E_x is the energy of Li.

COMSOL Multiphysics simulations

Geometrical configuration

The structure consists of two layers. The upper layer is a standard rectangular prism with a height of 50 μm and no boundaries in the length and width directions. The lower layer is also a rectangular prism with a central hole measuring 1000 μm in diameter. The centers of these holes are spaced 1500 μm apart, and the thickness of the lower layer is 8 μm.

The geometry is a solid domain and the physical field is electrochemistry. The electron conductivity of the upper layer is 1 × 10⁴ S m⁻¹, and the lower layer is 5.9 × 10⁷ S m⁻¹.

Calculation method

A computational study was performed using COMSOL Multiphysics 6.3 to investigate the distributions of electric potential, current density, and lithium-ion concentration. The model geometry consisted of a current collector and an active material, represented as two rectangular domains with dimensions of 8 μm by 500 μm and 50 μm by 1500 μm, respectively. The simulation coupled two physics modules: the electrostatics module to resolve the electric field and the tertiary current distribution module to describe lithium-ion transport. The electrostatics module solves Gauss’s law within the insulating polymer domains: $\nabla \times ({\mathrm{\epsilon }}_0\times {\mathrm{\epsilon }}_r\times \nabla V)=0$ where ${\mathrm{\epsilon }}_0$ is the vacuum permittivity and ${\mathrm{\epsilon }}_r$ is the relative permittivity of the current collector and active material. The tertiary current distribution module solved the Nernst-Planck equation for lithium-ion transport: $\frac{\mathrm{\partial }c}{\mathrm{\partial }t}+\nabla \times (-D\nabla c-z\mathrm{\mu }\mathit{Fc}\nabla \mathrm{\phi })=0$ where c is the lithium-ion concentration, D is the diffusion coefficient, z is the charge number, μ is the mobility, and F is the Faraday constant. Boundary conditions were defined as follows. For the experimental configuration, a constant lithium-ion concentration of 1 M was applied to both the bottom and top boundaries. For the control configuration, the same concentration was applied only at the top boundary. Periodic boundary conditions were imposed on the lateral boundaries to represent an infinitely repeating electrode structure. All internal interfaces were treated as continuity boundaries for electric potential, while the remaining external boundaries were defined as electrically insulating.

Supplementary Materials

The PDF file includes:

Figs. S1 to S30

Supplementary Notes

Tables S1 and S2

Legends for supplementary files S1 and S2

Other Supplementary Material for this manuscript includes the following:

Supplementary files S1 and S2