Quantitative corrosion framework for anti-corrosive passivation design to extend calendar life in lithium metal batteries

Abstract

Reactive lithium metal renders it prone to corrosion, which severely limits calendar life and practicality in energy storage. Despite its importance, corrosion-induced degradation remains largely qualitative and lacks a clear mechanistic understanding. Here, we present a quantitatively integrated and experimentally validated framework that correlates lithium corrosion with interphase growth kinetics and interfacial morphological evolution. Guided by this model, a bi-layered anti-corrosive passivation layer composed of lithium polyacrylate embedded with lithium silver alloy-fluoride interphase is rationally designed. The outer polymer-rich layer resists swelling and blocks corrosion, while the underlying LiAg/LiF-rich interphase enhances interfacial transport kinetics. Operando X-ray microscopy reveals that calendar-aged lithium regions are particularly vulnerable to accelerated corrosion, which intensifies dendritic formation and is effectively suppressed by the passivation layer. Consequently, full-cells show a high-rate capacity of 133 mAh g⁻¹ at 10 C (6 min) and retain 74.6% capacity after 400 cycles at 0.5 C (120 min), with Coulombic efficiency above 99.9%. Under a four-hour rest protocol for calendar life evaluation, full-cells maintain 75.1% capacity after 200 cycles, and further pouch cell testing shows 85.5% capacity after 640 cycles. This study offers insights into corrosion dynamics and informs the design of passivation strategies for improving calendar life in lithium metal batteries.

Introduction

Lithium (Li) metal is widely regarded as the most promising negative electrode for rechargeable batteries^1,2,3,4, achieving a practically viable lifespan for lithium metal batteries (LMBs) remains challenging due to irreversible interfacial phenomena occurring at the Li metal surface⁵. Highly reactive Li metal is particularly susceptible to spontaneous chemical corrosion upon contact with electrolyte, leading to increased interfacial resistance and eventual electrolyte depletion^6,7. This corrosion involves oxidation of fresh Li, forming ionic species and releasing electrons that contribute to solid–electrolyte interphase (SEI) formation^8,9. While the SEI initially passivates the surface, its structural fragility, inhomogeneous solubility with electronic conductivity render it inherently unstable^10,11,12. Recent studies have shown that continuous electrolyte decomposition at the SEI interface triggers dynamic processes of swelling, dissolution, fracture, and reformation, repeatedly exposing to the electrolyte and accelerating parasitic consumption of both Li and electrolyte, thereby rapidly degrading cell performance^13,14. Consequently, SEI instability–driven Li corrosion has emerged as a critical failure mechanism for Li metal negative electrodes (LMEs), underscoring the essential role of SEI design in stabilizing LMB.

Boyle et al. further demonstrated that chemical corrosion remains an unavoidable process during calendar aging, with SEI growth proceeding uncontrollably even in electrolytes exhibiting high Coulombic efficiency (CE)¹⁵. Likewise, Zhang et al. revealed that native SEI can swell or dissolve in the electrolyte, enabling direct Li–electrolyte contact and thereby exacerbating spontaneous chemical corrosion¹⁶. Thus, constructing an artificial interfacial layer that effectively blocks electron transfer and/or tunneling from Li is essential to fully mitigate corrosion and stabilize the SEI. Various inorganic (e.g., carbon, metals, and other compounds) and organic passivation layers have been explored to improve interfacial stability and suppress dendrite growth^{17,18,19,20,21,22}. Nonetheless, reliably preventing Li corrosion remains a challenge, as even inorganic-rich SEI layers can swell within their medium, compromising their ability as corrosion barriers¹⁵.

Interestingly, these limitations have led to the development of hybrid artificial passivation layers that combine organic and inorganic materials, designed to be insoluble and swelling-resistant while promoting uniform Li deposition^23,24,25. Jin et al. reported a PVDF-MgF₂ passivation layer that mitigates lithium corrosion by stabilizing the SEI and revealed its relation to SEI dissolution^23,24,25. Nevertheless, the direct and mechanistic correlation between the structural or compositional features of such layers and their role in suppressing chemical corrosion is still insufficiently elucidated. Most prior studies have predominantly offered qualitative and phenomenological interpretations, lacking a clearly defined quantitative framework that links corrosion behavior with SEI chemistry and electrochemical performance degradation. Moreover, fundamental design principles for developing such protective architectures have yet to be comprehensively developed. Thus, a deeper and quantitatively mechanistic understanding of corrosion dynamics is urgently needed to guide the rational design of passivation strategies that not only chemically shield the Li surface with insoluble and non-swelling matrices but also ensure sufficient interfacial kinetics and mechanical stability.

In practice, the calendar life of a battery cell is as essential as its cyclic life. While artificial passivation layers focus primarily on suppressing dendrites and extending cycle life, their impact on calendar life remains underexplored. As the calendar aging in LME is largely influenced by SEI accumulation^26,27, the chemical stability of freshly deposited Li depends on the protection provided by the SEI chemistry²⁸. Since each electrodeposition exposes fresh Li surfaces, continuous SEI reformation leads to recurring degradation. This process accelerates dendrite growth and dead Li formation through a negative feedback loop between cycling and resting^5,29, ultimately compromising calendar life and limiting practical viability. Thus, revealing the dynamics of calendar-aged Li is crucial for enabling the strategic development of anti-corrosive passivation layers to improve calendar life effectively.

Here, we report a quantitatively grounded framework that directly correlates Li corrosion with SEI growth kinetics and interfacial morphological evolution. Grounded in physical principles and experimentally validated, the chemical corrosive dissipation model (CCDM) integrates interfacial resistance and surface area evolution into a unified platform, enabling predictive analysis of corrosion-induced capacity loss. Guided by CCDM insights, a bi-layered anti-corrosive passivation layer composed of lithium polyacrylate embedded with lithium silver alloy-fluoride interphase (LPLA) was rationally designed to couple corrosion protection with interfacial kinetics. Operando X-ray micro-imaging (XMI) revealed that chemically corroded regions in bare Li trigger accelerated dendritic growth leading to calendar life degradation, whereas Li-LPLA preserves a smooth and stable interface even at high areal capacities (5 mAh cm⁻² at 2 mA cm⁻²). Paired with a high-loading practical full-cell configuration (15 mg cm⁻²), the Li-LPLA full-cell demonstrates stable cycling behavior, retaining 74.6% capacity after 400 cycles at 0.5C and 76.8% after 300 cycles at 1C, with CE > 99.9%. Notably, under calendar life evaluation with 4-h rest period following each charge cycle, it retains 75.1% capacity after 200 cycles, demonstrating improvement over bare Li (7.3% after 65 cycles). Our findings clarify mechanistic aspects of corrosion behavior and provide a practical basis for designing effective passivation layers in LMBs.

Results and discussion

Characterizations of the LPLA

The as-prepared Li (bare Li) surface features inherent roughness with numerous protrusions (Supplementary Fig. 1), leading to uneven seeding of Li nucleation, inhomogeneous plating, and dendritic growth (Fig. 1a). Besides, its native SEI is unevenly distributed, making Li highly susceptible to chemical corrosion and severe side reactions, ultimately causing early cell failure⁶. To reliably protect the LME, the bi-layered anti-corrosive passivation layer of LPLA is developed to prevent Li corrosion and extend calendar life. The swelling-resistant, low-solubility lithium polyacrylate (LP) matrix is specifically designed to inhibit electrolyte-induced corrosion by isolating the electrolyte and blocking electron tunneling³⁰. Typically, polymers with low swelling and solubility tend to exhibit high interfacial resistance for Li⁺ transport³⁰. However, the kinetically stable LP matrix forms via in situ reaction between Li and polyacrylic acid (PAA)^31,32, where lithiated carboxyl groups (COOH) with Li⁺ can dramatically improve ionic conductivity, enabling rapid and uniform ion transport (Supplementary Fig. 2)^33,34. With its strong binding ability and mechanical stability, the LP polymer also offers an elastic SEI that adapts to Li deformation, thereby suppressing dendrite growth and side reactions^31,35. The embedded AgF facilitates LiF enrichment and lithiophilic LiAg alloy formation at the interface through a metal fluoride displacement reaction^23,36, which improves charge transport, reduces the nucleation overpotential for uniform Li deposition, and enables subsurface Li growth beneath the interlayer^37,38. Notably, LiF is an insulator with a wide band gap that effectively blocks electron tunneling from the LME into the SEI, contributing to suppressed corrosion with dendrite-free Li deposits^39,40,41. By a facile, one-step process and eco-friendly manufacturing encapsulation methodology (Supplementary Fig. 3), the polymer solution in dimethyl sulfoxide (DMSO) with AgF is applied to the commercial Li foil. As a low-soluble, lithiophilic fluoride, AgF was incorporated at an optimized content of 90 wt%, determined by the electrochemical performance of Li||Li symmetric cells (Supplementary Fig. 4).

Fig. 1: Materials characterization of Li-LPLA.

a Schematic illustration of the anti-corrosive Li-LPLA passivation layer strategy. b XRD pattern of the bare Li and Li-LPLA. c FTIR spectra of bare Li, LP-only and Li-LPLA. High-resolution XPS profiles for the Li-LPLA of d Ag 3d and e F 1s. f Top-view FE-SEM image of the Li-LPLA. g Cross-section SEM image of the Li-LPLA. h High-resolution TEM image showing the bi-layer structure of Li-LPLA. i Ag- and Li-containing species observed within the Li-LPLA via TEM analysis. j EELS chemical composition mapping of Ag, F and C species acquired from the Ag M_4,5-edge, F K-edge, and C K-edge.

To investigate the crystal structure and surface chemistry, X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR) were conducted on the Li-LPLA. In Fig. 1b, the artificial passivation LPLA layer displayed peaks around 2θ = 38°, 44°, and 64°, which correspond to metallic Ag (PDF no. 04-0783) in the α-phase Ag solid solution, along with peaks around 2θ = 44° associated with the β-phase LiAg alloy (PDF no. 04-0805)⁴². The FTIR spectra of the LPLA characterized the molecular structure of LP (Fig. 1c), with prominent peaks at 1580 and 1420 cm⁻¹ corresponding to asymmetric and symmetric C=O stretching groups, related to the –COOLi⁺ moiety⁴³. X-ray photoelectron spectroscopy (XPS) further analyzed the chemical states of Ag 3d, F 1s, C 1s, and Li 1s on the passivated Li metal surface (Fig. 1d, e, and Supplementary Fig. 5). The high-resolution Ag 3d XPS spectrum confirmed the presence of Ag and LiAg in the hybrid layer, which serves as an electronic/ionic conductor to enhance charge transport⁴². The F 1s spectra showed a prominent peak at 685.2 eV, corresponding to LiF⁴⁴, with significantly enriched LiF formation in the LPLA layer. A uniform and tightly adherent layer with approximately 25 μm thickness is formed on the lithium surface, exhibiting good interfacial compatibility compared with bare lithium²³ (Fig. 1f, g, and Supplementary Figs. 6 and 7). The electrolyte wettability of the LPLA layer was confirmed by its minimal contact angle (Supplementary Fig. 8), which promotes uniform Li⁺ flux and nucleation^40,45. Transmission electron microscopy (TEM) was conducted to investigate the role of the anti-corrosive passivation layer in protecting LME. A solution of AgF and PAA was directly coated onto a Cu grid, followed by lithium deposition to form the LPLA passivation layer (Supplementary Figs. 9 and 10).

High-resolution TEM revealed a bi-layered artificial SEI consisting of an outer LP-rich amorphous polymer layer and an inner LiAg/LiF-rich inorganic layer (Fig. 1h, i). Lattice-resolved analysis identified a crystalline Li domain beneath the interphase, indicating preferential Li deposition at the LiAg/Cu interface. Electron energy loss spectroscopy (EELS) confirmed the compositionally heterogeneous structure (Fig. 1j), and ToF-SIMS depth profiling validated this configuration, revealing LP-derived polymeric fragments (CHOO⁻/C⁻) preferentially enriched near the surface and Ag⁻/F⁻ species increasingly concentrated toward the Li side (Supplementary Fig. 11)^31,46. This hybrid passivation layer effectively blocks chemical corrosion while enabling rapid interfacial Li⁺ transport across the interface.

Electrochemical properties of the LPLA

To validate the LPLA passivation layer in stabilizing metallic Li, its electrochemical performance was examined in Li||Li symmetric cells. The Li-LPLA electrode exhibited a lower nucleation overpotential (ΔE_n) than bare Li (Fig. 2a), indicating that the lithiophilic LiAg solid-solution alloy reduced the energy barrier for Li nucleation²³, promoting dense and uniform Li deposition⁴⁷. At a current density of 2 mA cm⁻² with a high deposition capacity of 2 mAh cm⁻², the Li-LPLA maintained low voltage polarization throughout cycling (Fig. 2b), whereas bare Li showed erratic voltage hysteresis due to uneven deposition and unstable SEI²⁸. SEI fracture and fresh Li exposure led to rapid voltage polarization increase in bare Li after 25 cycles (100 h), eventually causing short circuits via dendrite growth. Detailed voltage profiles revealed that Li-LPLA maintained lower overpotentials and smooth voltage plateaus, reflecting even Li deposition with stable SEI (inset of Fig. 2b)²⁸. Consequently, Li-LPLA symmetric cell sustained stable cycling for over 180 h. The plating overpotential of the bare Li electrode increased drastically with cycling at a fixed capacity, while the Li-LPLA electrode maintained a consistently lower rise and continued cycling without any short circuits (Fig. 2c). To evaluate high-rate Li reversibility, the Li-LPLA symmetric cells exhibited stable cycling behavior at 5 mA cm⁻² for 1 mAh cm⁻², while bare Li rapidly failed after only a few cycles (Supplementary Fig. 12a). For high-energy-density LMBs, 40 µm-thick symmetric cells were tested (Supplementary Fig. 12b), where the Li-LPLA consistently maintained Li cycling stability compared to bare Li. These results confirm that the hybrid LPLA passivation layer effectively suppresses parasitic reactions, stabilizes LME, and enhances cycling longevity.

Fig. 2: Electrochemical performance of Li-LPLA.

a Nucleation overpotential with bare Li and Li-LPLA at 0.5 mA cm^–2 for 1 mAh cm^–2. b Voltage profiles of Li||Li symmetric cells at 2 mA cm^–2 for 2 mAh cm^–2. c Plating overpotential of the Li||Li symmetric cells during cycling. d Tafel curves of bare Li and Li-LPLA symmetric cells. Impedance spectra of bare Li and Li-LPLA symmetric cells after e 1st cycle and f 50th cycle at 1 mA cm⁻² and 1 mAh cm⁻². Steady-state current measurement and calculated effective Li⁺ transference number for g bare Li and h Li-LPLA under 10 mV polarization for 0.5 h after the 30 cycles. i Li plating/stripping voltage profile of bare Li and Li-LPLA electrodes at 0.5 mA cm⁻² and 0.5 mAh cm⁻² for 50 cycles, followed by Li stripping at 0.2 mA cm⁻².

Given its improved electrochemical performance, the charge transport kinetics of the Li-LPLA were analyzed (Fig. 2d–f). Fluoride incorporation improved the ionic conductivity of the LP matrix compared to bare LP and PAA layers (Supplementary Fig. 13)²³. Tafel plot analysis revealed a higher exchange current density (0.120 mA cm⁻²) for Li-LPLA, more than twice that of bare Li (0.053 mA cm⁻²) (Fig. 2d). This indicated enhanced reaction kinetics, enabled by lower overpotential and reduced Li nucleation barrier^48,49. It can be attributed to the uniformly dispersed LiAg sites combined with LiF-rich interphase, which enables rapid Li⁺ diffusion and promotes homogenous Li deposition at the interface. Electrochemical impedance spectroscopy (EIS) further evaluated the interfacial reaction kinetics (Fig. 2e, f)^50,51. After 50 cycles, the Nyquist plots of the Li-LPLA electrode exhibited a comparatively smaller semicircle in the high- to medium-frequency region relative to bare Li, indicating reduced overall interfacial impedance. This behavior is suggestive of a more stable artificial interphase on Li-LPLA, which may facilitate improved interfacial Li⁺ transport compared with the native SEI on bare Li^42,52. A similar tendency was consistently observed under a higher current density and areal capacity at 2 mA cm⁻² for 2 mAh cm⁻² (Supplementary Fig. 14).

The effective Li⁺ transport within the SEI was evaluated using the Bruce–Vincent–Evans method to assess SEI evolution and electrolyte stability during cycling (Fig. 2g, h, and Supplementary Fig. 15). The Li-LPLA showed a high initial effective Li⁺ transference number (0.745) due to Li⁺ coordination with carboxyl groups and the LiAg/LiF-rich interphase, and retained 0.735 after cycling, indicating stable SEI formation and uniform Li⁺ flux for effective passivation⁴⁴. In contrast, bare Li decreased from 0.394 to 0.226, accompanied by increased electrolyte resistance (R_S) from continuous decomposition and SEI thickening. Moreover, Li-LPLA exhibited lower activation energy for Li⁺ transport across the SEI (Supplementary Fig. 16), reflecting reduced migration barriers and accelerated interfacial kinetics that collectively enable uniform deposition and long-term interfacial stability⁵³. Given the limited areal capacity of ultrathin Li foil, achieving high cycling reversibility is critical. The average Coulombic efficiency (ACE) was measured for 40 µm-thick Li||Li symmetric cells using a previously developed method (Supplementary Fig. 17)¹⁷. The Li-LPLA achieved an ACE of 95.8% after 100 h at 0.5 mA cm⁻² (Fig. 2i), outperforming bare Li. Thus, the artificial LPLA passivation layer ensures uniform Li plating/stripping, suppresses side reactions, and enhances cycling longevity.

To evaluate the practical applicability of the Li-LPLA electrode, full-cells were assembled with a high-Ni positive electrode, LiNi_0.8Co_0.1Mn_0.1O₂ (NCM811). The initial charge and discharge specific capacities were measured at 209 mAh g⁻¹ and 202 mAh g⁻¹, respectively, resulting in a CE of 96.7% (Supplementary Fig. 18). This relatively high initial CE suggests that the anti-corrosive LPLA layer effectively suppresses surface side reactions with the electrolyte. The Li-LPLA||NCM811 full-cell exhibited stable performance at high-rates (Fig. 3a), delivering discharge capacities of 202, 196, 186, 178, 172, 161, and 133 mAh g⁻¹ at 0.1, 0.2, 0.5, 1, 2, 5, and 10 C, respectively. In stark contrast, the bare Li||NCM811 full-cell without the artificial passivation layer showed severely degraded rate performance and failed to operate at 5 C or higher. It can be attributed to void formation and SEI collapse, with the resulting overpotential dominated by Li⁺ transport across the SEI during high-rate stripping/plating (Fig. 3b)⁵⁴. The Li-LPLA||NCM811 full-cell exhibited stable charge–discharge profiles and notably low overpotentials at all tested C rates (Fig. 3c), further corroborated by symmetric Li-LPLA||Li-LPLA cycling up to 30 mA cm⁻² without short-circuiting (Supplementary Fig. 19). These improvements in the rate performance can be attributed to the passivating and charge-coordinating LPLA layer, which effectively enables rapid and uniform Li plating/stripping²³. However, the bare Li electrode undergoes rapid SEI thickening, which severely impedes ion transport and leads to increasingly sluggish diffusion. The proposed Li-LPLA||NCM811 full-cell exhibits electrochemical performance that is competitive with representative state-of-the-art LMBs reported in recent studies (Fig. 3d and Supplementary Table 1), especially under high current rates with high areal loading and extended cyclability.

Fig. 3: Li-LPLA||NCM811 full-cell performance.

a Rate capability of bare Li||NCM811 and Li-LPLA||NCM811 full-cells (N/P ratio = 30.9; 1C = 200 mA g⁻¹). Voltage profiles of b bare Li||NCM811 full-cells and c Li-LPLA||NCM811 full-cells under varying C rates. d Comparison of the rate performance of the Li-LPLA||NCM811 full-cell with previously reported state-of-the-art LMBs. The source of the literature data shown in this figure can be found in Supplementary Table 1. e Cyclic performance of bare Li||NCM811 and Li-LPLA||NCM811 full-cells at 0.5C (N/P ratio = 2.72; 1C = 200 mA g⁻¹). SEM images of f bare Li||NCM811 and g Li-LPLA||NCM811full-cell after 50 cycles. h Cycling performances of single-stack Li-LPLA||LiFePO₄ pouch cells (mass loading = 15 mg cm⁻²; N/P ratio = 3.23; electrolyte: 1M LiPF₆ in EC/DEC (1:1, v/v), ρ = 1.26 g mL⁻¹; E/C ratio = 24.7 g Ah⁻¹; 1C = 170 mA g⁻¹).

The Li-LPLA||NCM811 full-cell, comprising a 40 μm-thick Li metal negative electrode and a high-loading (15 mg cm⁻²) NCM811 positive electrode, exhibited stable cycling retention, maintaining 74.6% of its initial capacity and CE > 99.9% over 400 cycles at 0.5C using a conventional carbonate-based electrolyte without additives (Fig. 3e, and Supplementary Table 2). In contrast, the bare Li||NCM811 cell rapidly degraded, sustaining only 27.8% after 120 cycles with a notable CE decline. At higher C rates, the Li-LPLA||NCM811 full-cell consistently outperformed its unprotected counterpart (Supplementary Fig. 20). When cycled at 1C for 300 cycles using fluoroethylene carbonate (FEC) additives, it retained 76.8% with CE consistently >99.9% (Supplementary Fig. 21a). Even at 5C, it maintained 77.3% after 80 cycles with CE > 99% (Supplementary Fig. 21b), whereas the bare Li cell exhibited early capacity fade. The cyclic stability of the Li-LPLA negative electrode, even at high rates, is attributed to its passivating and ion-conductive interphase, which effectively stabilizes SEI and facilitates reversible Li plating/stripping. The surface morphology of both Li-LPLA||NCM811 and bare Li||NCM811 full-cells was examined after repeated cycling tests. Scanning electron microscopy (SEM) images revealed that the Li-LPLA electrode maintained a smooth surface without visible clogs, while the bare Li exhibited rough morphology with porous dendrites after 50 cycles (Fig. 3f, g). Correspondingly, Nyquist plots showed a pronounced increase in interfacial resistance for bare Li, whereas Li-LPLA displayed a reduced semicircle after 100 cycles, indicating enhanced charge transfer kinetics (Supplementary Fig. 22). Therefore, the Li-LPLA negative electrode maintains structural integrity through repeated cycling and shows high rate capability and cycling stability, indicating relevance to practical LMB. To validate practical utility, a pouch-type full-cell with a high-loading LiFePO₄ positive electrode (15 mg cm⁻², 3.0 cm × 4.0 cm) retained 85.5% capacity after 640 cycles with CE > 99.9% (Fig. 3h). These results confirm that the LPLA passivation layer effectively mitigates Li corrosion and sustains interfacial stability under practical constraints.

Understanding anti-corrosive behaviors of Li-LPLA

Lithium corrosion in LMBs is inherently complex and dynamic process, dominated by persistent interfacial reactions that promote SEI growth and progressive morphological degradation. This leads to SEI thickening and the formation of voids, cracks, and surface roughening, all of which expand the reactive area and accelerate side reactions. Over time, this self-amplifying feedback loop exacerbates lithium and capacity fade. The SEI growth rate was evaluated by tracking time-dependent interfacial resistance (R_int(t)) in symmetric cells during calendar aging (Fig. 4a, and Supplementary Fig. 23). Power-law fitting behavior (R_int(t)∝at^x)^15,55,56,57 revealed sublinear scaling (x = 0.15 for bare Li, 0.08 for Li-LPLA), indicating diffusion-limited SEI growth⁵⁶. The Li-LPLA showed the lowest R_int increase (1.7-fold over 300 h), compared to 2.6-fold for bare Li, demonstrating its effective resistance to corrosion-driven SEI formation. Surface morphological evolution, quantified by Brunauer–Emmett–Teller (BET) surface area under argon condition, showed similar sublinear power-law scaling (x = 0.17 for bare Li, 0.07 for Li-LPLA; Fig. 4b), confirming its effectiveness of Li-LPLA in suppressing corrosion-induced surface reconstruction and degradation. Importantly, ΔCE was employed as a cumulative metric of corrosion activity, evaluated via galvanostatic stripping after rest periods up to 300 h (Supplementary Figs. 24 and 25). Bare Li exhibited rapid CE decline and severe capacity loss (ΔCE: –13.3%, 0.89 mAh loss after 300 h) due to SEI degradation, resulting in dead Li accumulation and increased impedance. In contrast, Li-LPLA showed markedly reduced degradation, with just –0.8% ΔCE and 0.05 mAh loss after 8 h, maintaining CE above 94.9% and a low capacity loss to 0.33 mAh after 300 h (Fig. 4c).

Fig. 4: Anti-corrosive effect and extended calendar life of Li-LPLA.

a Time-dependent interfacial resistance in Li||Li symmetric cells. The interfacial resistance is normalized to the initial value measured at 0 h for each cell. The fitting curves represent nonlinear least squares fit to a power law (R_int ∝at^x). b Surface area evolution of bare Li, LA-only, LP-only, and Li-LPLA electrodes, as determined by BET analysis after aging. The data reveal power-law scaling behavior associated with corrosion-driven morphological degradation. c Comparison between experimentally measured and CCDM-predicted ΔCE values over increasing rest times. d Ex situ XMI images of bare Li, LA-only, LP-only, and Li-LPLA electrodes after 50-h electrolyte immersion, revealing corrosion-induced surface degradation. e Li plating/stripping voltage profiles of bare Li, LA-only, LP-only, and Li-LPLA electrodes cycled at 0.5 mA cm⁻² and 0.5 mAh cm⁻² for 50 cycles with 8-h rest after every 5th cycle, followed by Li stripping at 0.2 mA cm⁻². f Calendar life of the full-cells during cycling involving 4 h rest at the end of each charge. g Quadrant plot comparing interfacial kinetics and chemical stability for all electrode configurations. h Schematic of corrosion behavior with different passivation layers. Bare Li suffers from severe electrolyte-induced degradation. LA-only offers lithiophilicity but limited corrosion resistance, while LP-only prevents corrosion but lacks lithium reversibility. The bi-layered LPLA synergistically achieves both corrosion suppression and reversible Li cycling.

To quantitatively interpret Li corrosion behavior, the CCDM was developed to predict ΔCE based on SEI growth kinetics and surface area evolution (Supplementary Fig. 26). It assumes Li corrosion drives SEI formation via chemical dissolution (Li → Li⁺ + e^-), consuming Li and electrolyte. The SEI growth rate, denoted as $\frac{d\left({SEI}\right)}{{dt}}$, is directly linked to a kinetic descriptor of corrosion activity. Concurrently, the time-dependent surface area S(t) defines the spatial extent for parasitic reactions. Their product, $\frac{d\left({SEI}\right)}{{dt}}\cdot S\left(t\right)$, represents the instantaneous corrosive flux and is analogous to the time-dependent corrosion current density. The cumulative charge loss, expressed as ΔCE(t), is obtained by integrating this flux over time:

$$\Delta {CE}\left(t\right)=k\cdot {\int }_{0}^{t}\left(\frac{d\left({SEI}\right)}{{dt}}\cdot S\left(t\right)\right){dt}$$

(1)

where k is an empirically fitted scaling factor. The CCDM validation under four representative passivation conditions (bare Li, lithium silver alloy-fluoride (LA)-only, LP-only, Li-LPLA) showed high predictive accuracy (R² > 0.99, Supplementary Table 3), confirming that cumulative efficiency loss is governed by the interplay between SEI kinetics and reactive surface evolution. Beyond its predictive insight, the CCDM offers a mechanistic understanding of corrosion behavior and demonstrates the efficacy of anti-corrosive LPLA strategies in mitigating interfacial degradation, guiding the rational design of stable LMEs.

Interestingly, the LA-only layer also suppressed chemical corrosion but exhibited greater capacity loss and lower CE compared to both LP-only and Li-LPLA electrodes. Although the LiF-rich SEI stabilized Li, swelling compromised its anti-corrosion effectiveness¹⁶. By contrast, the outer LP-rich layer remained insoluble and structurally stable, emphasizing its critical role in anti-corrosive protection. Immersion tests with ex situ XMI analysis revealed pronounced surface roughening caused by chemical corrosion on bare Li and LA-only electrodes, while LP-only and Li-LPLA maintained smooth and intact surfaces (Fig. 4d, and Supplementary Figs. 27 and 28). All samples were freshly prepared and imaged under identical low-dose X-ray exposure (100 ms), confirming that the observed features result from chemical reactions at the Li–electrolyte interface rather than radiation-induced artifacts. Moreover, severe pitting and by-product accumulation were observed on bare Li, whereas Li-LPLA retained an undamaged morphology (Supplementary Fig. 29), demonstrating the effectiveness of the corrosion-blocking polymer layer in suppressing corrosion and preserving interfacial integrity during prolonged calendar aging.

The cyclic lifespan of LMBs is typically assessed without rest periods, failing to account for calendar aged degradation^26,27. To evaluate chemical corrosion inhibition and its impact on electrochemical performance, quantified ACE was analyzed in symmetric cells with 8-h rests after every 5th cycle (Fig. 4e). The Li-LPLA showed the highest ACE of 95.7% and minimal ΔACE (−0.1%) compared to bare Li (−2%), indicating improved corrosion suppression and improved Li reversibility. LA-only also performed well (94.7%), while LP-only showed greater cumulative loss (93.8%), suggesting limited reversibility despite corrosion resistance. To further validate calendar life, full-cells underwent cycling with 4-h rest after each charge (Supplementary Fig. 30a). While bare Li||NCM811 suffered rapid failure after 38 cycles, Li-LPLA||NCM811 retained 75.1% of its initial capacity after 200 cycles with CE > 99.8% (Fig. 4f, and Supplementary Fig. 30b). These results underscore the anti-corrosive efficacy of the LPLA layer and its potential for long-lasting LMBs. Although polymer-based passivation mitigates chemical corrosion, it does not inherently ensure Li reversibility (Fig. 4g). Achieving high CE requires not just ionic conductivity^58,59,60 but also efficient interfacial charge transport and uniform Li nucleation. LA-only layers offer limited corrosion resistance, while LP-only layers lack sufficient lithiophilicity and interfacial kinetics. The LPLA design strategically integrates both functionalities (Fig. 4h): the inner LA-rich layer forms LiF/LiAg interphases that enhance lithiophilicity and interfacial kinetics, while the outer LP-rich layer provides chemically stable and non-swelling, effectively blocking corrosion. Thus, the bi-layered LPLA architecture simultaneously suppresses chemical corrosion and promotes electrochemical reversibility, establishing a rational design strategy for extending the calendar life of LMBs.

Calendar-aged Li deposition dynamics

To elucidate the mechanism pertaining to the Li deposit dynamics on corroded Li (Fig. 5a), the morphological evolution of Li deposition after calendar aging was investigated. The inhomogeneous components of the native SEI and its dissolution nature could cause differences in the Li⁺ flux distribution, which directly impacts the surface area of exposed Li. Accordingly, the SEI can influence the interfacial kinetics of the LME, closely tied to the morphological evolution of deposited Li^28,61. SEM images of calendar-aged bare Li and Li-LPLA revealed distinct surface morphologies as Li plating capacity increased (Fig. 5b). Bare Li after aging showed extensive coverage with Li dendrites as Li deposit progressed, forming a fibrous network with an undensified structure. This sparse morphology significantly increased the exposed Li surface area, intensifying parasitic reactions with the electrolyte and contributing to the decrease in CE after calendar aging. In contrast, Li-LPLA exhibited a compact and uniform plating behavior even at a high capacity of 5 mAh cm⁻² under calendar aging. This flat and dense morphology inhibited the exposed Li surface, reducing contact with the electrolyte and thereby mitigating chemical corrosion during aging while enhancing Li reversibility. These SEM results demonstrate that Li-LPLA promotes a stable, compact, and dense morphology of Li deposits, while bare Li struggles to accommodate large volume variations, making it prone to fracture and exposing fresh Li to the electrolyte^28,62.

Fig. 5: Calendar-aged Li dynamics.

a A scheme of anti-corrosive passivation strategy of the LPLA for inhibited Li corrosion and uniform Li morphology. b Ex situ SEM images of Li plating morphology after calendar aging at 0.5, 1, and 5 mAh cm^–2 under a current density of 1 mA cm^–2 for bare Li and Li–LPLA. c Colorized operando XMI images of the calendar-aged bare Li and Li-LPLA at 2 mA cm^–2. Li growth regions are distinctly highlighted in yellow to improve the visual interpretation of Li dynamics.

Although recent studies have suggested that dendrites preferentially grow within corroded regions of Li, these claims have largely relied on ex situ methods³⁹. Such techniques, however, fall short in providing real-time insights into the specific composition of corroded Li and the mechanisms driving dendrite formation and growth. To achieve an in-depth understanding of nucleation and growth mechanisms in corroded Li, operando XMI was employed to monitor the dynamic evolution of Li deposition (Supplementary Fig. 31, and Fig. 5c). The imaged lithium was visualized through optical contrast with the surrounding electrolyte⁶³, and identified as crystalline metallic Li (Supplementary Fig. 32), indicating that its geometry can be directly observed in real-time using the operando XMI setup. After 50 h of calendar aging in electrolyte, bare Li exhibited discernible external voids and surface cavities were observed, attributed to the collapse of the flexible SEI and Li loss³⁹. To further elucidate calendar-aged lithium dynamics, Supplementary Movie 1 revealed the time-resolved evolution of Li deposition. On corroded Li, pre-existing voids and surface concavities progressively expanded further due to concurrent corrosion (marked with white dashed lines in Fig. 5c), serving as preferential sites for dendritic growth. As plating proceeds to 5 mAh cm⁻², mossy Li rapidly propagated outward from these recessed regions, eventually covering the entire electrode surface.

This behavior can be explained by the geometric effect, where Li⁺ flux becomes concentrated in the corroded regions, coupled with SEI stress-induced defects that increase local Li⁺ diffusivity^39,64. As a result, the corrosion-prone Li surface facilitates preferential Li nucleation within corroded pits during successive deposits, exacerbating dendrite formation and active Li loss⁶⁵, which in turn contributes to the degradation of calendar life. By contrast, the Li-LPLA electrode exhibited no voids or surface cavities after calendar aging. The LPLA passivation layer effectively inhibited electrolyte-induced corrosion and enabled uniform, compact lithium deposition without the dendritic formation, even at high plating capacity of 5 mAh cm⁻² at a current density of 2 mA cm⁻². These findings suggest that corrosion suppression is a decisive factor in preventing dendritic growth and establish LPLA as an emerging passivation strategy for interfacial stability in LMBs. By effectively suppressing corrosion and dendritic growth, the LPLA layer substantially contributes to the long-term stability and calendar life.

Chemical and mechanical stability analysis of Li-LPLA

The long-term structural integrity of the passivation layer is essential for maintaining its anti-corrosive behavior. High-resolution TEM revealed that the Li-LPLA electrode retained a thin, conformal SEI even after 30 cycles, indicating effective suppression of chemical corrosion and sustained interfacial protection (Fig. 6a). EELS mapping further confirmed the intact bi-layer architecture of outer LP-rich layer and inner LiAg/LiF-rich layer, with no apparent degradation or morphological collapse (Supplementary Fig. 33). It is demonstrated that the LPLA passivation layer retains its structural integrity throughout cycling, enabling it to accommodate volume changes and relieve interfacial stress.

Fig. 6: Chemical and mechanical stability of Li-LPLA.

a High-resolution TEM image of the Li-LPLA electrode after 30 cycles, showing a well-preserved bi-layer structure with a stable SEI. b XPS spectra of F 1s for the comparisons between bare Li and Li-LPLA after 50th cycle. c Contour map of the Li-LPLA showing the spatial distribution of the C 1s intensity. d Top-view and matching cross-sectional SEM images of bare Li and Li-LPLA after 50th cycle. e AFM 3D images of DMT modulus, adhesion and dissipation mapping for bare Li and Li-LPLA after 50th cycle.

To elucidate the correlation between SEI stabilization and battery performance, depth-resolved XPS analysis was conducted after 50 cycles. At the surface, the C 1s spectrum of Li-LPLA showed lower Li₂CO₃ intensity compared to bare Li (Supplementary Fig. 34a), indicating suppressed electrolyte decomposition and restrained SEI growth^44,66,67. The F 1s and Li 1s spectra revealed higher LiF-to-Li_xPF_y ratios and stronger LiF signals, respectively (Fig. 6b, and Supplementary Fig. 34b), signifying the formation of thinner and denser SEI layers¹⁸. This LiF-rich interphase in Li-LPLA efficiently passivates reactive surfaces and enhances Li⁺ transport across the SEI^19,68, as corroborated by EIS analysis showing reduced charge transfer and SEI resistances. With increasing sputtering depth, the surface-associated carbonate peaks (C–O, C=O, and ROCO₂Li) in the C 1s spectrum diminished beyond ~20 nm, while a pronounced COOLi peak emerged ⁶⁹ (Fig. 6c), indicating that organic decomposition products were confined to the near-surface region. High-resolution spectra at the etched regions further confirmed the chemical stability of the polymer matrix and preservation of the inorganic interphase (Supplementary Fig. 35), indicating its stable and well-integrated structure.

Repeated Li deposition/stripping, along with significant volume changes, compromise the cyclic stability and safety of LMBs. SEM images after 50 cycles exhibited that bare Li developed severe dendrites and cracks due to unstable and thick SEI (Fig. 6d), significantly increasing electrolyte exposure and exacerbating side reactions. In contrast, the Li-LPLA maintained a smooth and compact surface morphology with no noticeable dendrites, indicating a well-preserved and tightly connected structure. The bare Li electrode showed a significant thickness increase to 477 μm (by 59% increase), whereas Li-LPLA exhibited only 330 μm (by 10% increase), demonstrating improved structural integrity. These results are attributed to the LPLA layer, which effectively passivates the LME surface while boosting interfacial charge transport kinetics, thereby ensuring stable and reversible cycling.

Atomic force microscopy (AFM) was utilized to examine the surface morphology changes in Li-LPLA electrode before and after 50 cycles. The surface roughness of Li-LPLA remained nearly unchanged, whereas bare Li became significantly roughened (Supplementary Fig. 36). To further investigate interfacial and structural stability, local mechanical properties of the SEI were analyzed using AFM pinpoint nanomechanical mode. The Young’s modulus of SEI, calculated via the Derjaguin–Müller–Toporov (DMT) model, was higher for Li-LPLA (11.5 GPa) than for bare Li (1.6 GPa) (Fig. 6e, and Supplementary Fig. 37). This disparity suggests that the swollen SEI on bare Li softens in liquid electrolytes¹⁶, whereas the non-swollen SEI on Li-LPLA with LiF-enriched phase demonstrated a high shear modulus in its components, reflecting greater robustness and stress tolerance^66,70. Additionally, average adhesion of Li-LPLA (59.6 nN) was also higher than that of bare Li (32.5  nN), attributed to its well-preserved SEI structure⁶⁶. Electrode stability was further analyzed via energy dissipation during loading/unloading. The dissipation energy of Li-LPLA (1.4 keV) was lower than that of bare Li (1.8 keV), indicating reduced irreversible deformation, fewer microcracks, and better stress dissipation. These advanced mechanical properties effectively suppress SEI swelling, maintain structural integrity, and enhance cycling stability.

In conclusion, this study establishes a quantitatively grounded framework that directly correlates Li corrosion with SEI kinetics and morphological evolution into a unified predictive platform. By integrating a physically validated corrosion model, operando XMI visualization and calendar life evaluation, a comprehensive methodology is developed to diagnose and mitigate corrosion-driven degradation. The real-time observation further reveals corrosion suppression as a critical factor in controlling dendritic growth. Informed by mechanistic insight, a rationally designed bi-layered LPLA passivation architecture effectively couples corrosion shielding with rapid interfacial charge kinetics, which is key to enabling highly reversible and durable calendar-aged Li cycling. Calendar life evaluation confirms the corrosion-inhibiting efficacy of the LPLA layer in high-loading NCM811 full-cells, achieving 75.1% capacity retention after 200 cycles, outperforming bare Li counterparts. Our discovery reveals critical mechanistic insights into corrosion dynamics and articulates a foundational design strategy, paving a viable pathway toward the practical deployment of LMBs.

Methods

Synthesis of Li-LPLA

AgF (Aldrich, 99%), PAA (Aldrich, Mw ~450,000), and DMSO (Alfa, 99.9%) were used for the fabrication of the Li-LPLA passivation layer. DMSO was employed as the suspension solvent owing to its chemical stability, high polarity and thermal stability, which facilitated the dissolution of AgF and enabled the formation of a uniform LiF-rich and lithiophilic LiAg alloy interphase upon evaporation and subsequent in situ reaction with lithium, without inducing side reactions. AgF and PAA were each dissolved in DMSO at a concentration of 0.1 wt% and stirred for 24 h to ensure complete dissolution and homogeneity. The two solutions were then mixed in the desired ratios and stirred for an additional 24 h to form a homogeneous mixture. The LPLA passivation layer was obtained using a facile blade-casting method. The prepared solution is blade-cast onto commercial Li foil (Wellcos Corporation), followed by drying at 80 °C. All synthetic procedures were conducted in an argon-filled glove box, where the H₂O and O₂ levels were maintained below 0.1 ppm to prevent contamination.

Materials characterization

The morphology and microstructural features of the samples were examined using field-emission SEM (FE-SEM, JSM-7800F Prime, JEOL) coupled with energy-dispersive X-ray spectroscopy analysis and TEM (JEM-2200FS, JEOL) with EELS, conducted at an accelerating voltage of 200 kV. ToF-SIMS (TOF-SIMS 5, ION TOF) was performed under vacuum to characterize the composition of the LPLA layer. A Cs cluster ion beam (0.5 keV) was used for sputtering over a 250 × 250 μm² area, while a Bi ion beam (25 keV) was applied for analysis within a 50 × 50 μm² region. The chemical structure of the LPLA passivation layer was analyzed using FTIR with a Perkin Elmer FTIR Spectrum Two instrument. XPS (MultiLab 2000) measurements were performed with an Al-Kα X-ray source (E = 1486.6 eV) to determine the chemical properties and elemental composition of the surface. The surface area evolution was evaluated using the BET method from adsorption–desorption isotherms in the 0.2–0.4 P/P₀ range, conducted under argon to prevent Li reactivity (model ASPS 2010 BET/Porosimeter; Micromeritics). The crystallographic characteristics of the samples were examined using synchrotron XRD at the PLS-II 9A U-SAXS (wavelength of 0.6202 Å) and 6D C&S UNIST SAXS (wavelength of 0.65303 Å) beamline at the Pohang Accelerator Laboratory (PAL). Data were collected in transmission mode using 2D CCD detector (SX165, Rayonix and MX225-HS, Rayonix, respectively). All measured 2θ values were calibrated and converted to the Cu Kα radiation scale (1.5406 Å) for analysis. Additionally, XRD at the 3D XRS beamline (λ = 0.61992 Å) was employed to investigate Li dendrite structures observed via operando XMI, with data collected using a Mar345 image plate detector. AFM measurements were conducted to evaluate the mechanical properties of the LPLA passivation layer after 50 cycles, using an NX-10 instrument (Park Systems) equipped with an NCHR cantilever in PinPoint nanomechanical mode, within a glove box. For ex situ analyzes, cells were disassembled and transferred under an Ar-sealed environment to minimize air exposure.

Swelling characteristics of the polymer matrix were evaluated in a 1.0 M solution of LiPF₆ dissolved in a 1:1 (v/v) mixture of ethylene carbonate (EC) and diethyl carbonate (DEC). Polymer films (16 mm diameter) were immersed in the electrolyte for a specified period. After immersion, the films were removed, carefully wiped, and weighed. The swelling ratio (SR_t) at time t was calculated using Eq. (2):

$${{SR}}_{t}=\frac{{m}_{t}-{m}_{o}}{{m}_{o}}\times 100\%$$

(2)

where m₀ is the initial weight of the dried polymer, and m_t is the weight of swollen polymer after time t. The wettability of the lithium foil with the electrolyte was analyzed using a contact angle measuring instrument (SmartDrop, FemtoBiomed Inc.).

Operando XMI analysis

Operando XMI experiments were conducted at the PLS-II 6C BMI beamline of the PAL. A monochromatic 10 keV X-ray beam was used for illumination, with exposure controlled by a shutter. In this arrangement, the transverse coherence length, given by $\frac{{{\rm{\lambda }}}\cdot {{\rm{L}}}}{{{\rm{s}}}}$, where λ the X-ray wavelength, L is the source-to-object distance, and s is the source size, was calculated to be 4.5 μm. Projection images were acquired 50 mm downstream of the sample using a detector assembly consisting of a 50 μm-thick LuAG:Ce scintillation screen (CRYTUR, Turnov, Czech Republic), a 10× objective lens, and a pco.Edge 5.5 sCMOS camera (Excelitas Technologies, Pittsburgh, PA, USA). The effective imaging resolution was 2 μm, which is smaller than the transverse coherence length, thereby satisfying the conditions for phase contrast imaging. The custom-designed operando XMI cell employed symmetric Li metal electrodes (bare Li and Li-LPLA). A handmade stainless-steel rectangular bar was polished to obtain a flat substrate, onto which Li foil was pressed to ensure intimate contact and secure assembly. These Li-loaded electrodes were mounted into a custom cell holder featuring a 3 mm-diameter aperture and sealed with Kapton film window to enable X-ray transmission. A flat Li electrode with an exposed surface area of 1.7 × 0.2 mm² was obtained, and the inter-electrode distance was adjustable to about 500 μm to ensure effective X-ray penetration. To minimize X-ray scattering and enhance image clarity, separators were intentionally excluded, and excess electrolyte was added to fully fill the internal volume of the cell. The synchrotron X-ray beam was directed through the window, and time-resolved projected images were continuously acquired to monitor the real-time evolution of calendar-aged lithium dynamics. The images were flat-field corrected using a reference image captured just before time zero. However, beam drift during the acquired period introduced uneven background artifacts in the corrected images. To mitigate this, the images were further processed by normalizing them with their blurred versions, generated using non-local means filtering, to suppress background variations. The resulting images were deblurred by subtracting their Laplacian (Amira-Avizo 3D v2024.1, Thermo Fisher Scientific, Waltham, MA, USA).

Electrochemical measurements

All electrochemical performance tests were conducted using CR2032-type coin cells with stainless-steel cases, spacers, and springs assembled in an Ar-filled glove box (H₂O and O₂ < 0.1 ppm) and evaluated at 25 °C in a temperature-controlled chamber. The electrolyte used was 1.0 M LiPF₆ dissolved in a 1:1 (v/v) mixture of EC and DEC, with or without 10 wt% FEC additive (Dongwha), and Celgard 2400 polypropylene separators (Wellcos Corporation; thickness: 25 µm; porosity: 41%; one layer; diameter: 19 mm) were used. The electrolytes were stored in their original sealed Al bottles inside an Ar-filled glove box (H₂O and O₂ < 0.1 ppm). Electrolyte transfer was carried out using polypropylene (PP) pipette tips. During coin cell assembly, a total electrolyte volume of 40 μL was added in two portions to wet the separator. Lithium foil electrodes with thicknesses of 300 μm or 40 μm (12 mm diameter) were employed in Li||Li symmetric cells, which were tested at specific current densities and Li capacities. Galvanostatic charge-discharge cycling tests were carried out using a WBCS 3000 battery tester (WonATech). The full-cells were fabricated with single-side-coated NCM811 positive electrodes and both bare Li and Li-LPLA negative electrodes. The NCM811 electrodes were prepared by mixing the active material, Super P, and polyvinylidene fluoride (PVDF) in N-methyl-2-pyrrolidone (NMP, Aldrich) at a weight ratio of 8:1:1 using a Thinky mixer under ambient atmosphere. The slurry was cast onto Al foil (MTI; thickness: 15 μm) without additional surface treatment using a doctor blade, dried in a vacuum oven at 120 °C for 12 h, and subsequently compressed using a roll press. Electrodes with a diameter of 12 mm were prepared using a punching tool (Wellcos Corporation), with active material loadings of approximately 10 and 15 mg cm⁻², corresponding to an N/P ratio of 2.72 based on the theoretical capacity of 40 μm Li (16 mm diameter). All the full-cells were activated by five formation cycles at 0.05 C (1 C = 200 mA g⁻¹) within a cell voltage range of 2.7−4.2 V (vs. Li/Li⁺) under constant-current (CC) mode. For full-cell rate capability tests, discharging was performed at increasing current densities (0.1 to 10C, every five cycles), while charging was fixed at 0.1C to assess reversibility. For pouch cells, LiFePO₄ electrodes were prepared similarly by mixing active material, Super P, and PVDF in NMP (8:1:1 by weight) using a Thinky mixer. The slurry was cast onto carbon-coated Al foil, dried at 80 °C under vacuum for 12 h, followed by roll pressing. Single-stack Li-LPLA||LiFePO₄ pouch cells were assembled using a LiFePO₄ positive electrode (3.0 × 4.0 cm, active material loading of 15 mg cm⁻²), a Li metal negative electrode (3.2 × 4.2 cm), and a separator (3.3 × 4.3 cm), with an N/P ratio of 3.23. The pouch cells were cycled under an external stack pressure of ~0.1 MPa, and a degassing step was performed after electrolyte injection and pouch sealing. The assembled Li-LPLA||LiFePO₄ pouch cell delivered a total capacity of 25.5 mAh, with an electrolyte-to-capacity (E/C) ratio of 24.7 g Ah⁻¹. Full-cells were cycled under CC mode within a voltage window of 2.5–3.8 V (vs. Li/Li⁺) at 0.2C for charging and 0.5C for discharging (1C = 170 mA g⁻¹). The specific capacity (mAh g⁻¹) was calculated by normalizing the measured capacity (mAh) to the mass of active material (g) in the electrode. Tafel plots were derived from cyclic voltammetry measurements performed in Li||Li symmetric cells within a cell voltage range of −0.15 V to 0.15 V (vs. Li/Li⁺) at a scan rate of 1 mV s⁻¹. EIS was conducted in potentiostatic mode at open-circuit potential using an AC amplitude of 5 mV over a frequency range from 1 MHz to 1  mHz, and the impedance data were fitted using the ZMAN software. For each electrochemical experiment, at least three independent cells (n ≥ 3) were tested, and a representative cell closest to the average performance is shown.

The ACE was evaluated using Li||Li symmetric cells cycled at 0.5 mA cm⁻² for 0.5 mAh cm⁻² per cycle over 50 cycles, either with an 8-h rest period after every 5 cycles (to estimate calendar aging) or continuously without rest. The initial reservoir capacity (Q_R), representing the practically available lithium, was determined by fully stripping the electrode to 1.0 V (vs. Li/Li⁺) at 0.2 mA cm⁻², resulting in 6.41 mAh cm⁻² for Li-LPLA and 6.67 mAh cm⁻² for bare Li. After cycling, the remaining lithium was quantified by an exhaustive final stripping at the same current density to obtain the remaining stripping charge (Q_S). The ACE was then calculated using the following Eq. (3)¹⁷:

$${ACE}=\frac{1}{1+\left({Q}_{R}-{Q}_{S}\right)\div50}$$

(3)

The ionic conductivity (${{\rm{\sigma }}}$) of the polymer films was evaluated using symmetric stainless steel||stainless steel blocking cells assembled in CR2032-type coin cells, which effectively suppress Faradaic reactions and isolate ionic transport. Stainless-steel disk electrodes (16 mm diameter) and polymer films (19 mm diameter) were used, with 1 M LiPF₆ in EC/DEC (1:1, v/v) as the electrolyte. The ionic conductivity was calculated from the high-frequency intercept of the Nyquist plots using the following Eq. (4):

$$\sigma=\frac{t}{R\,\times A\,}$$

(4)

where t is the thickness, R is resistance, and A is the area of the polymer film. Li⁺ transfer number (t_Li+) was evaluated using the DC polarization method and calculated with the Bruce–Vincent–Evans Eq. (5):

$${t}_{{{Li}}^{+}}=\,\frac{{I}_{{ss}}(\Delta V-\,{I}_{0}{R}_{0})}{{I}_{0}(\Delta V-\,{I}_{{ss}}{R}_{{ss}})}$$

(5)

where ΔV is the polarization voltage, I_ss is the steady-state current, I₀ is the initial current, R_ss is the steady-state resistance after polarization, and R₀ is the initial resistance before polarization.