Regenerative Artificial Solid Electrolyte Interphase via Dynamic Cross‐Linking for Stable Lithium Metal Anodes

ABSTRACT

Artificial solid electrolyte interphases (ASEIs) are widely used to stabilize lithium metal anodes, yet their protection inevitably degrades because polymer coatings crack under repeated volume fluctuation and inorganic SEI components detach during cycling. Herein, we present a dynamic crosslinked fluoropolymer (DCF) artificial interphase that simultaneously enables autonomous polymer crack healing and continuous replenishment of inorganic SEI components. The dynamic interphase is constructed from a layer of poly (2,2,2‐trifluoroethyl acrylate) cross‐linked by ureido‐pyrimidinone, whose reversible quadruple hydrogen bonds enable rapid self‐healing of the polymer network and effective closure of interfacial cracks. Additionally, the fluorine‐rich polymer matrix serves as a persistent source for LiF formation, continuously replenishing the inorganic inner layer to compensate for detachment during cycling. Benefiting from this dual regenerative mechanism of polymer self‐healing and inorganic component self‐replenishment, lithium metal anodes protected by the dynamic interphase exhibit highly stable plating and stripping for over 4000 h in symmetric cells. In full cells, a LiFePO₄||Li cell retained 90.67% of its capacity after 1000 cycles, while a high‐loading LiNi_0.6Co_0.2Mn_0.2O₂ (NCM622) pouch cell maintained 70% capacity after 700 cycles at 1C.

Dynamic crosslinked fluoropolymer (DCF) is an artificial solid‐state electrolyte interface material, which possesses dully regenerative properties. It leverages the dynamic reversibility of quadruple hydrogen bonds to achieve self‐healing of the polymer matrix, autonomously repairing physical cracks and maintaining its structural integrity. Additionally, the abundant fluorine sources in the matrix provide continuous replenishment of inorganic components such as LiF to compensate for detachment‐induced losses.

1. Introduction

Lithium metal anode is widely regarded as a key step toward next‐generation rechargeable batteries [1, 2] owing to its ultrahigh theoretical specific capacity (3860 mAh g⁻¹) and the most negative electrochemical potential among practical anode materials [3, 4]. Despite these advantages, the practical deployment of lithium metal remains impeded by severe interfacial instability during repeated plating and stripping. The high chemical reactivity of lithium metal inevitably leads to electrolyte decomposition and the formation of a solid electrolyte interphase (SEI) [5], which governs lithium deposition behavior, interfacial kinetics, and long‐term cycling stability [6]. An ideal SEI should simultaneously enable fast Li⁺ transport, block electron transfer, and maintain structural integrity under continuous volume changes. In practice, however, the native SEI formed on lithium metal is chemically heterogeneous and mechanically fragile [7]. It typically consists of a mosaic mixture of inorganic species such as LiF and Li₃N together with organic components derived from electrolyte decomposition [8, 9, 10, 11]. During cycling, this interphase experiences repeated volume fluctuation and stress concentration, leading to cracking, delamination, and continuous reconstruction [12]. As a result, fresh lithium is repeatedly exposed to the electrolyte, accelerating parasitic reactions, consuming active lithium, and generating “dead lithium”, which ultimately causes rapid capacity decay and safety concerns [13, 14, 15].

To address these challenges, extensive efforts have been devoted to constructing artificial solid electrolyte interphases (ASEIs) [16, 17, 18]. Among various strategies, polymer‐based ASEIs have attracted particular interest because of their flexibility, film‐forming capability, and good interfacial compatibility with lithium metal [16, 19, 20, 21]. Considerable progress has been made by tailoring polymer chemistry, incorporating inorganic fillers, or introducing supramolecular interactions to improve mechanical robustness and ionic conductivity. Nevertheless, most reported ASEIs still suffer from two fundamental limitations that are rarely addressed simultaneously. Polymer coatings are unable to maintain long‐term integrity on lithium metal. Repeated plating and stripping inevitably generate cracks that cannot be repaired once they form, causing permanent loss of interfacial protection [22]. Besides, inorganic SEI components such as LiF are prone to detachment and depletion during long‐term cycling, leading to the gradual loss of their protective function. Notably, polymer cracking accelerates electrolyte penetration and consumption, while the loss of inorganic components, further aggravates interfacial stress and promotes mechanical failure. Existing ASEI designs typically address only one aspect of this coupled degradation, either by enhancing polymer toughness or by enriching inorganic species at the interface. Developing an ASEI that can simultaneously repair polymer damage and restore inorganic SEI components during operation, therefore, remains a critical, but unresolved challenge.

Herein, we report a dynamically cross‐linked fluoropolymer ASEI that addresses these two degradation processes at the same time. The ASEI is constructed from poly (2,2,2‐trifluoroethyl acrylate) (TFA) dynamically cross‐linked by 2‐ureido‐4[1H]‐pyrimidinone (UPy) molecules featuring reversible quadruple hydrogen bonds. This dynamic bonding allows cracks formed during lithium plating and stripping to close spontaneously at room temperature within several minutes, thereby limiting renewed electrolyte contact with fresh lithium. Meanwhile, the fluorine‐rich TFA continuously supplies fluorine at the interface, leading to the sustained formation of LiF. As a result, inorganic SEI components that detach during cycling are gradually replenished rather than permanently lost. Benefiting from the dually regenerative processes of the intrinsic polymer and ASEI, the Li symmetric cell operates stably for more than 4000 h. In full cells, a LiFePO₄||Li cell retained 90.67% of its capacity after 1000 cycles, and a high‐loading NCM622 pouch cell with 10 mg cm⁻² cathodes maintained 70% capacity after 700 cycles at 1 C. This study provides a fundamental design strategy for constructing ASEIs with multi‐scale regenerative capabilities and offers important insights for developing highly stable lithium metal batteries.

2. Results and Discussion

2.1. Preparation and Characterization of DCF

Fluorinated polymers are employed to construct ASEIs due to their chemical and electrochemical stability and the abundant fluorine source for lithium metal [23, 24]. In this work, the TFA molecule containing abundant fluorine atoms was selected as the matrix for DCF membrane (Figure 1a–c). The abundant ─CF₃ groups facilitate the timely replenishment of a thin and uniform LiF‐rich SEI on the fresh lithium surface, providing protection for the lithium metal (Figure S1). The UPy molecules, which contain quadruple hydrogen bonds, were incorporated into the polymer as dynamic cross‐linkers. Its dynamic cross‐linking function imparts higher elastic modulus and self‐healing properties to the polymer (Figure 1b) [25, 26].

FIGURE 1.

(a) Chemical structure of DCF and DCF@Li anode. (b) DCF achieves self‐healing via the dynamic “breaking‐forming” behavior of quadruple H‐bonds in UPy. (c) DCF spontaneously replenishes the LiF inorganic layer on the lithium metal surface through the abundant fluorine elements in the TFA matrix.

The UPy monomers were synthesized by reacting 6‐methylisocytosine with 2‐isocyanatoethyl methacrylate (Figure S2). In the ¹H nuclear magnetic resonance (1H‐NMR) spectrum (Figure S3), the hydrogen atoms at 11.94 ppm (s, 1H), and 10.49 ppm (t, 1H) confirmed the successful preparation of the UPy monomers. The disappearance of the N═C═O stretching band at 2261 cm^{− 1} in the Fourier transform infrared (FTIR) spectrum indicated complete consumption of the isocyanate group (Figure S4). Meanwhile, the stretching vibration peak of the C═O shifted from 1720 to 1698 cm^{− 1}, signifying the transformation from an ester to a urea group, further evidencing the successful synthesis of the UPy monomer [27]. Components were mixed as precursor solutions in the glove box according to the proportions shown in Table S1 and then polymerized in situ on the surface of the lithium metal via ultraviolet initiation, with the polymerization reaction shown in Figure S5. Success of the polymerization reaction was confirmed by the inversion test in Figure S6. Photographs of the DCF membrane are shown in Figure S7. Energy‐dispersive x‐ray spectroscopy (EDS) mapping indicates uniform distribution of each element within the DCF, demonstrating homogeneous distribution of the constituent monomers (Figure S8). The cross‐section SEM image shows that, the thickness of the DCF layer is about 19 µm (Figure S9). Furthermore, the atomic force microscopy (AFM) image reveals that the surface of the DCF membrane is smooth (Figure S10).

2.2. Performance Characterization of DCF

The dynamic cross‐linking effect of UPy between the TFA segments was reflected in the increase of the polymer's elastic modulus. As demonstrated in Figure 2a, the incorporation of UPy molecules significantly improved the Young's modulus of the polymer. The Young's modulus of the polymer TFA (without UPy dynamic cross‐linking) was very low, even as low as approximately 0.34 MPa. When 2 mol% UPy molecules were incorporated into the polymer, the Young's modulus significantly increased to 5.93 MPa due to the dynamic cross‐linking. The resilience of DCF was further investigated through tensile‐recover tests. Five consecutive cyclic tensile tests were performed on DCF at a constant tensile strain of 200%. As shown in Figure 2b, the first cycle showed a significant hysteresis loop, but as cycling progressed, the loading and unloading curves gradually converged, and the curves for each cycle also became increasingly consistent. When unloaded after being fixed at maximum deformation for 3/5/10 min, the DCF specimen still exhibited excellent recovery capability (Figure S11). Clearly, DCF demonstrated excellent toughness and resilience, indicating its ability to effectively accommodate volume changes of the lithium anode during cycling [28, 29]. To further investigate the dynamic thermomechanical behavior of DCF, rheological tests were conducted between 25  and 200° C. As shown in Figure 2c, the storage modulus (G′) remained consistently higher than the loss modulus (G″) across the entire temperature range, indicating the dominant elastic behavior of DCF. However, as shown in Figure S12, when the mole fraction of UPy in the polymer was increased to 5 mol%, the storage modulus of the polymer rapidly decreased with rising temperature and eventually converged with the loss modulus, indicating that an excessively high mole fraction of the UPy crosslinker is detrimental to the deformation recovery performance of the polymer.

FIGURE 2.

(a) Stress‐strain curves and Young's modulus (inset) of TFA and DCF. (b) Cyclic tensile test curve of DCF in successive loading–unloading cycles with a strain of 200% and the optical images (inset) of DCF before stretching (left) and with stretching to 200% elongation (right). (c) Rheological results of DCF across a temperature range of 25 to 200° C. (d) Self‐healing process of DCF. (e) LSV curves of Li| DCF| stainless steel (SS) batteries and Li|TFA‐5% UPy| SS batteries. (f) I‐t curve and corresponding Nyquist plots (inset) before and after polarization of Li||Li symmetric cells. (g) Highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels of ethylene carbonate (EC), diethyl carbonate (DEC), fluoroethylene carbonate (FEC), LiPF₆, TFA, UPy, and DCF. (h) Optimized geometrical configurations and electrostatic potentials (ESPs) of EC, DEC, FEC, TFA, UPy, and DCF. (i) Binding energies of EC, DEC, FEC, LiPF₆, TFA, UPy, and DCF molecules with free Li⁺.

The dynamic cross‐linking, not only enhanced the elastic modulus of the polymer matrix, but also endowed it with self‐healing properties. Under optical microscopy, a scratched polymer basically healed within 1 min (Figure 2d). On a macroscopic level, a cut DCF film was reattached, and the rejoined film could lift a weight of 200 g without breaking and maintain adhesion for up to 40 s (Figures 2d and S13). As shown in Figure S14, the ultimate stress of the DCF and healed‐DCF is almost equivalent. Fast and robust self‐healing behavior enables DCF to timely repair its own cracks, preventing the lithium metal from being exposed to the electrolyte. This reduces side reactions between the lithium metal and the electrolyte, thereby helping to minimize electrolyte consumption. The electrochemical window of the DCF up to 5.2 V is measured by linear sweep voltammetry (LSV) test, indicating high electrochemical stability (Figure 2e). The Li⁺ transference number of DCF is as high as 0.67 (Figure 2f), which helps reduce interfacial polarization and promote uniform Li deposition [30].

While UPy can enhance polymer strength and provide self‐healing properties, its content is not necessarily higher the better. The thermal stability of polymers was analyzed by thermogravimetric analysis (TGA), as shown in Figure S15. The DCF film exhibited only 8.7 wt% weight loss at around 250°C, and the differential thermogravimetric (DTG) curve reveals that DCF did not exhibit significant thermal weight loss until 382°C. Excessive UPy units are detrimental to the thermal stability of the polymer. The polymer with a 5% mole fraction of UPy had already undergone 15.6% thermal weight loss at 250°C, nearly twice that of DCF. TGA tests confirm the thermal stability of DCF and highlight its superiority under high‐temperature conditions. Additionally, the differential scanning calorimetry (DSC) curves in Figure S16 indicates that the DCF film had a lower glass transition temperature (Tg) of −36.7°C, suggesting rapid chain segment dynamics and high ionic conductivity within the polymer network. However, the crosslinking density in TFA‐5%UPy is excessively high, which increases the crystallinity of the polymer. As a result, the T_g rises to −29.8° C, indicating that excessive UPy crosslinking is detrimental to lithium‐ion transport within the polymer segments.

Another important function of the ASEI is to suppress side reactions on the lithium metal surface. Bare Li and DCF@Li were assessed for stability upon air exposure (Figure S17). Bare‐Li exhibited darkening after 5 min and became completely black after 15 min, reflecting its corrosion by H₂O and O₂, while DCF@Li maintained its brightness for up to 30 min. To evaluate the affinity between DCF and the electrolyte, contact angle measurements were conducted. As shown in Figure S18, the contact angle between DCF and carbonate‐based electrolyte is as high as 63.51°, while with ether‐based electrolyte, it also reaches 62.09°, indicating that DCF is non‐wetting with these two types of commercial electrolytes. As shown in Figure S19, DCF exhibits no significant dissolution or swelling in the electrolyte, thereby maintaining its structural stability and integrity, which helps effectively isolate lithium metal from electrolyte components, thereby reducing side reactions on the lithium metal surface [31]. Air oxidation tests, contact angle tests, and the electrolyte uptake test together demonstrate that DCF can effectively reduce side reactions between lithium metal and water, oxygen, and the electrolyte, which helps protect the lithium metal anode.

2.3. Analysis of the Mechanism of DCF as an Effective ASEI

To explain the interaction between various components in DCF and lithium ions, as well as the formation mechanism of the internal inorganic layer, density functional theory (DFT) calculations were employed. To simplify the DCF structure for modeling, the TFA structural unit, the UPy structural unit, and a short chain of DCF polymer containing both UPy and TFA were selected for the calculations.

The reduction sequence of organic molecules on the Li surface is related to the energy of the lowest unoccupied molecular orbital (LUMO). As can be seen from Figure 2g, the components on the DCF film possess lower LUMO energies compared to the electrolyte components, with DCF in particular having the lowest LUMO (−0.46 eV). The above calculation results indicate that on the Li surface, the components in the DCF film are likely to decompose preferentially, facilitating the formation of an SEI rich in LiF. Because DCF decomposes preferentially over the electrolyte components, it helps minimize electrolyte consumption.

According to the electrostatic potential (ESP) calculation in Figure 2h, the carbonyl oxygen atoms in the TFA, UPy, and DCF molecules exhibit the most concentrated negative charge. Therefore, the carbonyl oxygen atoms were selected as the binding sites to calculate their binding energies with Li⁺. By comparing the binding energies of lithium ions with DCF and electrolyte components (Figure 2i), it could be found that carbonyl oxygen atoms on the DCF segments have significantly higher binding energies than those of electrolyte components such as ethylene carbonate (EC), diethyl carbonate (DEC), and fluoroethylene carbonate (FEC). This suggests that when Li⁺ ions reach the DCF | electrolyte interface, DCF can “embrace” Li⁺ ions more firmly, which is beneficial for reducing the interfacial contact resistance. Additionally, it can also be seen in Figure 2i that the binding energies of lithium ions with the carbonyl oxygen atoms in the DCF segment are lower than those with the carbonyl oxygen atoms in TFA and UPy. This indicates that Li⁺ ions can migrate more easily within the polymer segments of DCF. A moderate lithium‐ion binding energy reduces interfacial impedance as well as facilitates the rapid transport of lithium ions, enabling the excellent cycling performance of DCF@Li.

2.4. Battery Performance of DCF@Li Anode

The long‐term interfacial stability of the DCF@Li anode was assessed by the performance of symmetric cells under galvanostatic cycling. As shown in Figure 3a, under a current density of 0.5 mA cm⁻², the DCF@Li cell maintained a highly stable voltage profile for over 4000 h (more than five months). In contrast, the cell with bare Li anodes exhibited a gradually increasing polarization after 1000 h, which ultimately culminated in a sharp voltage drop caused by an internal short circuit. Elemental analysis of these cycled anodes revealed that the surface of bare lithium was composed predominantly of O and C (Figure S20a). In sharp contrast, the distinct signal of F was still obviously detectable on the DCF@Li anode even after 4000 h of cycling (Figure S20b), confirming that the DCF facilitates the formation of a stable and durable LiF‐rich layer on the Li surface. The superiority of DCF@Li was further demonstrated under higher current density. As shown in Figure 3b, the DCF@Li anode exhibited significantly superior stability to the bare Li anode under 1 mA cm⁻² and 1 mAh cm⁻². Furthermore, as shown in Figure 3c, the nucleation overpotential and stripping overpotential of bare lithium reached 32 and 36 mV, respectively, whereas the nucleation overpotential of DCF@Li was nearly 0 mV, and the stripping overpotential was only 5.9 mV. Lower nucleation and stripping overpotentials indicate that DCF facilitates uniform lithium deposition and suppresses the formation of “dead lithium,” contributing to the long‐term stable cycling of the lithium metal anode [32].

FIGURE 3.

(a) Cycling performance of symmetric cells with bare Li and DCF@Li under 0.5 mA cm⁻². (b) Cycling performance of symmetric cells with bare Li and DCF@Li under 1 mA cm⁻². (c) Magnified voltage−time curves of bare Li||bare Li and DCF@ Li ||DCF@ Li symmetric cells. In situ Nyquist plots (inset figures) and corresponding DRT analysis of the (d) bare Li||bare Li and (e) DCF@Li||DCF@Li symmetric cells after the 50th,100th, 150th, and 200th cycles under 1 mA cm⁻² and 1 mAh cm⁻². (f) Voltage profiles of DCF@ Li||DCF@ Li and bare Li||bare Li symmetric cells at various current densities with a fixed capacity of 1 mAh cm⁻². (g) Comparison of cycle time and current density of Li||Li symmetric cells reported in the literature.

The gradual decrease in polarization voltage during the initial stage of symmetric cell cycling is attributed to the formation of the SEI. To compare the SEI formation behavior of bare Li and DCF@Li, in situ EIS and distribution of relaxation times (DRT) analyses were conducted over the first 200 cycles at 1 mA cm⁻² and 1 mAh cm⁻² (Figure 3d,e). In the EIS curves, the bare Li symmetric cell exhibited a progressively enlarging semicircle, indicating a continuous increase in interfacial resistance. In contrast, the DCF@Li symmetric cell showed a consistent decrease in semicircle diameter with cycling, despite having an initially larger one. As shown in the DRT curves of bare Li (Figure 3d), apart from the high‐frequency bulk resistance R_s and the low‐frequency Warburg impedance W₀, the DRT plot exhibited two distinct peaks: one around 10⁻⁴ s and another around 10⁻² s. Among them, the peak near 10⁻⁴ s increased significantly as cycling proceeded and could therefore be identified as R_SEI [33]. In contrast, the peak around 10⁻² s showed only a slight increase and could thus be attributed to R_ct [34, 35]. In the DRT analysis of DCF@Li (Figure 3e), the peak corresponding to R_SEI progressively decreased with cycling, demonstrating that the SEI layer on DCF@Li does not undergo thickening, but rather maintains a stable configuration on the lithium metal surface, which is favorable for lithium‐ion transport. At the same time, the peak for R_ct also continuously decreased, signifying an accelerated charge transfer process on the lithium metal surface. The decrease of R_SEI and R_ct is probably related to the self‐replenishment of LiF on the DCF@Li surface. When the current density was increased to 2 mA/cm², the symmetric cells exhibited the same trend in both R_SEI and Rct. (Figure S21)

At a fixed deposition capacity of 1 mAh cm⁻², the symmetric cell with the DCF@Li anode exhibited significantly lower lithium deposition/stripping overpotentials compared to the bare Li anode. As the current density increased from 0.5 to 1, 2, 4, and 5 mA cm⁻², the overpotentials of the bare Li anode reached 81, 137, 316, 511, and 590 mV, respectively. In comparison, the DCF@ Li anode showed lower overpotentials of 31, 43, 63, 85, and 92 mV, indicating decreased Li⁺ transport resistance and less pronounced interfacial polarization. Notably, when the current density was sequentially reduced from 5 to 0.5 mA cm⁻², the DCF@Li anode maintained stable overpotential profiles without significant hysteresis (Figure 3f). The substantial reduction in overpotential, coupled with consistent responses under dynamic current conditions, highlighted the excellent rate capability of the DCF@Li anode. As shown in Figure 3g, DCF@Li exhibited ultra‐long cycling stability, surpassing various other recently reported ASEIs for Li anodes [20, 23, 36, 37, 38, 39, 40, 41, 42, 43]. The superior long‐term cycling performance of the DCF@Li anode demonstrates that DCF provides durable and stable protection for lithium metal, which stems from its self‐healing and self‐replenishing behavior.

To further illustrate the self‐healing and self‐replenishing functions of DCF, lithium symmetric batteries with TFA@Li and dynamically crosslinked ethyl acrylate (DCE)@Li as electrodes were also assembled. DCE is prepared by ethyl acrylate (which contains no fluorine) and UPy. While it possesses dynamic self‐healing ability, the absence of a fluorine‐rich inner SEI layer led to significant polarization in the lithium symmetric battery, preventing normal cycling (Figure S22a). Although TFA@Li also contained a substantial amount of fluorine, it lacks self‐healing capability, resulting in less sustained protection for lithium metal compared to DCF. As shown in Figure S22b, the lithium symmetric battery with TFA@Li coating could only cycle for approximately 400 h. Worse still, due to TFA's lower ionic conductivity compared to DCF, the TFA@Li symmetric cell exhibited a higher polarization voltage at the same current density (0.5 mA cm⁻²). These comparative experiments demonstrate the importance of the dual (self‐healing and self‐replenishing) regenerative mechanism of DCF. To visually demonstrate that DCF retains its self‐healing ability during battery cycling, the DCF@Li and TFA@Li anodes after 100 cycles were characterized by scanning electron microscope (SEM). As shown in Figure S23a, the DCF@Li anode after cycling shows complete coverage of lithium metal by the DCF film, with the polymer layer remaining intact and free of holes, detachment, or cracks. In contrast, the polymer on the TFA@Li anode (which lacks self‐healing ability) exhibits cracks and even detachment (Figure S23b). These results confirm that, the self‐healing behavior of DCF helps preserve the structural integrity of the polymer during battery cycling, thereby, benefiting the protection of the lithium metal anode.

To verify the self‐replenishing effect of the LiF component on the DCF@Li, we performed Auger electron spectroscopy (AES) tests on the cycled electrodes. As shown in Figure 4a, the peak corresponding to the F element (around 640 eV) in the AES spectrum of bare lithium metal decreased rapidly with sputtering time. In contrast, in Figure 4b, the peak around 640 eV in the AES spectrum of DCF@Li remained stable. A semi‐quantitative analysis of the AES spectra (Figure 4c) revealed that when the sputtering time reached 5 min, the fluorine content on the DCF@Li surface was 14.7%, significantly higher than that of bare lithium (9.1%), meaning that the fluorine content on DCF@Li was richer and remained stable. Further AES etching revealed that the thickness of the LiF layer formed by DCF does not exceed 25 nm. (Figure S24) Because the LiF layer is very thin, it does not adversely affect ion migration within the SEI layer.

FIGURE 4.

Depth‐profiled AES spectra data for the cycled Li anodes in (a) bare Li|| bare Li and (b) DCF@Li||DCF@Li cells. (c) Atomic percentage of F on bare Li anode and DCF@Li anode with increasing sputtering time. (d) Tafel plots and exchange current densities of bare Li||bare Li and DCF@Li||DCF@Li symmetric cells. (e) XPS spectra analyses of F 1s for bare Li anode after 50,100, and 150 cycles, conducted at 1 mA cm⁻², and 1 mAh cm⁻². (f) XPS spectra analyses of F 1s for DCF@Li anode after 50,100, and 150 cycles, conducted at 1 mA cm⁻², and 1 mAh cm⁻². In situ Raman spectroscopy displayed characteristic TFSI⁻ peaks in (g) the bare Li system and (h) the DCF@Li system, tested at 1 mA cm⁻².

To further investigate the reason for the increased proportion of fluorine on the DCF@Li surface, x‐ray photoelectron spectroscopy (XPS) tests were conducted on anodes after 50, 100, and 150 cycles. As shown in Figure 4e, after different cycles, the fluorine on the bare lithium metal surface mainly originated from C–F bonds and Li_xPF_y from the electrolyte. On the bare Li anode operating after 50 and 100 cycles, no peaks corresponding to LiF were observed in the XPS spectra. Even after 150 cycles, the intensity of the LiF peak remained significantly lower than that of the C─F peak, indicating the low content of LiF. In contrast, Figure 4f shows that after different cycles, distinct peaks were detected at around 685 eV on the DCF@Li surface, and the peak area corresponding to LiF increased significantly with the number of cycles. This suggests that the increase in fluorine on the DCF@Li surface likely stems from the replenishment of LiF, confirming the self‐replenishing effect of DCF. Additionally, Figure S25 also revealed an increase in the content of Li₃N, which may originate from the decomposition of the UPy. Benefiting from the increased inorganic content, DCF@Li demonstrated a higher exchange current density. Tafel profiles in Figure 4d identified superior kinetics of the DCF layer [44]. The higher exchange current density indicates faster ion transport at the DCF@Li surface under equilibrium conditions, which promotes uniform Li deposition and reduces interfacial impedance. This result is also consistent with the observed decrease in R_ct and R_SEI in Figure 3g. The LiF on the DCF@Li surface is believed to be ultra‐thin and polycrystalline in nature [45], and its self‐replenishment does not lead to continuous thickening, thereby not hindering ion transport within the SEI layer. The enhanced exchange current density on the DCF@Li surface suggests better rate capability, which is consistently reflected in Figure 3f.

Reducing side reactions between lithium metal and the electrolyte is another function of ASEI. To verify this capability of DCF, we performed in situ Raman spectroscopy to monitor changes in the electrolyte component (LiTFSI) on the lithium metal surface during charge—discharge processes. As shown in Figure 4g,h, under a current density of 1 mA cm⁻², the TFSI⁻ signal (738–746 cm⁻²) on the bare Li surface almost completely disappeared after about 100 s, whereas the TFSI⁻ signal on the DCF@Li surface remained stable for up to 400 s. The stable intensity of the TFSI⁻ peak near the lithium metal indicates that DCF can effectively suppress side reactions between the lithium metal and the electrolyte during charge—discharge processes, thereby enhancing the cycling stability of the lithium metal anode.

The inhibition effect of DCF against lithium dendrites was visually demonstrated through combined characterization using scanning electron microscopy (SEM), atomic force microscopy (AFM), and optical microscopy (OM). As shown in Figure 5a, the surface of bare lithium after 200 cycles exhibited loose and uneven lithium deposition, accompanied by significant cracking. The abundant microcracks could lead to continuous exposure of the underlying lithium metal to the electrolyte, resulting in persistent side reactions and lithium consumption. In contrast, as shown in Figure 5b, the DCF@Li anode exhibited a uniform and dense lithium plating morphology without obvious lithium dendrites. The cross‐section SEM images of lithium metal show that the deposition layer on the bare lithium metal surface was approximately 132.2 µm and relatively loose, whereas the deposition layer on DCF@Li was comparatively thinner (98.4 µm), denser, and more uniform. Atomic force microscopy (AFM) can provide more detailed morphological features of the electrode surface. From the local comparisons in Figure 5c,d, it was observed that the surface roughness of the cycled bare Li was much higher than that of DCF@Li. This indicates that DCF facilitated uniform lithium deposition, thereby contributing to the long‐term cycling stability of the battery.

FIGURE 5.

Top view and cross‐section SEM images of (a) bare Li and (b) DCF@Li anodes captured after 200 cycles of Li plating/stripping at 1 mA cm⁻² and 1 mAh cm⁻². AFM images of surface morphology of (c) bare Li and (d) DCF @Li anodes after 50 cycles. In situ optical microscopy of Li plating processes on (e) bare Cu and (f) DCF@Cu under 1 mA cm⁻² over various time periods.

To observe the inhibitory effect of DCF on lithium dendrites visibly, the in situ optical microscope (OM) was used for real‐time observation of the lithium plating behavior on bare Cu and DCF@Cu. The observation device used is shown in Figure S26a. As shown in Figure 5e,f, under a current density of 1 mA cm⁻², uneven lithium plating appeared on the bare Cu surface after 5 min, eventually evolving into extensive lithium dendrites. In contrast, the DCF@Cu anode maintained a flat and dendrite‐free surface throughout the 20 min lithium deposition process. The photos of the two copper foils after lithium deposition also show significant differences (Figure S26b,c). The bare copper surface exhibited a large amount of lithium dendrites along with exposed red copper foil, while the DCF@Cu surface displayed uniform lithium deposition with only very few dendrites at the edges. This indicates that DCF can effectively suppress lithium dendrite growth, reduce the risk of internal short circuits in the battery, and extend its cycle lifespan. This also demonstrates that the self‐healing behavior of DCF does not compromise the uniformity of species distribution on the lithium metal surface during cycling. The uniformity of the SEI layer on the DCF@Li surface is maintained thanks to two key factors: self‐healing that suppresses parasitic reactions between lithium and the electrolyte, and self‐replenishment that promptly patches the loss of inorganic species.

2.5. Performance of DCF@Li Anodes in Full Cells

To further evaluate the feasibility of the DCF@Li anode for practical applications, cells were assembled using LiFePO₄ (LFP) cathodes. As shown in Figure 6a, at a rate of 1 C, the LFP||bare Li cell maintained a retention of 84.59% after 550 cycles, whereas the LFP||DCF@Li cell maintained a capacity retention of 85.22% after 859 cycles and 90.67% after 1000 cycles. The charge/discharge voltage profiles in Figure S27 revealed that the polarization voltage of the DCF@Li anode remained stable during the cycling process, indicating the improved and long‐lasting interfacial charge transfer kinetics of DCF@Li. Beyond cycling stability, the high‐rate capability of LFP|| Li cells was systematically evaluated (Figure 6b). When the charge‐discharge rate was increased to 2C, 4C, and even 8C, LFP||DCF@Li demonstrated higher specific discharge capacity than LFP||bare Li. Particularly at 4C and 8C, the specific discharge capacity of LFP||bare Li decayed to nearly zero, while LFP||DCF@Li still maintained capacities of approximately 70 and 50 mAh/g, respectively. The improvement in rate performance indicates that DCF facilitates faster lithium‐ion transport on the electrode surface by replenishing the inorganic components.

FIGURE 6.

(a) Long‐term cycling performance of LFP||Li coin cell with bare Li and DCF@Li anodes. (b) Rate capability of LFP||bare Li and LFP||DCF@Li coin cells at varying discharge rates from 0.2  to 8  C. (c) Cycling performance of NCM622||bare Li and NCM622||DCF@Li pouch cells. (d) Representative charge/discharge voltage profiles of the NCM622||DCF@Li pouch cell at various cycles. (e) Cycling performance of NCM622||bare Li and NCM622||DCF@Li pouch cells at 5C/5C. (f) Representative charge/discharge voltage profiles of the NCM622||DCF@Li pouch cell at the 30th cycle. Ultrasonic transmission images of (g) NCM622||bare Li and (h) NCM622||DCF@Li during the first 10 cycles.

The cycling coulombic efficiency (CE) was used to evaluate the reversibility of lithium plating/stripping on the anode during cycling. As shown in Figure S28a,b, at a current density of 1 mA cm⁻², the DCF@Cu electrode maintained a high CE even after 70 cycles, while the bare Cu electrode showed a significant drop in CE after 40 cycles (Figure S28a,c). At a lower current density of 0.5 mA cm⁻², the DCF@Cu electrode retained a high CE of 92.31% after 180 cycles (Figure S29a,b), whereas the bare Cu electrode experienced a substantial CE decrease after 120 cycles (Figure S29a,c). By using the modified Aurbach method, the average CE has been calculated. After cycling for 30 h, the average CE of the bare Cu||Li cell was measured to be 96.67% (Figure S30a), whereas the average CE of the DCF@Cu||Li cell was maintained at 97.52% (Figure S30b). The improvement in coulombic efficiency can be attributed to the crack self‐healing capability of DCF, which reduces lithium metal loss.

Furthermore, to investigate the potential of the DCF@Li anode under challenging commercial conditions, pouch cells were assembled using NCM622 cathodes with a high mass loading of approximately 10 mg cm⁻² for cycling stability tests. As shown in Figure 6c,d, the NCM622||DCF@Li pouch cell demonstrated a capacity retention of 70% after 700 cycles at 1 C, significantly outperforming the NCM622||bare Li cell. The cycling stability and capacity retention of pouch cells at high rates are closely related to the ion transport rate at the lithium metal surface [46, 47]. Under high‐rate of 5C and a high cathode loading of 10 mg cm⁻², the NCM622||DCF@Li pouch cell also demonstrated significantly superior capacity. As shown in Figure 6e, when the charge/discharge rate was increased to 5C/5C, the specific discharge capacity of the NCM622||bare Li cell rapidly decayed. As shown in Figure 6f, by the 30th cycle, its specific discharge capacity dropped to only 47.11 mAh/g, while the NCM622||DCF@Li cell still maintained a specific discharge capacity of 92.99 mAh/g. This indicates that the protective effect of DCF facilitates ion transport at the lithium metal surface under high‐rate conditions. Furthermore, as shown in Figure S31, the NCM622||DCF@Li pouch cell remained capable of powering a bulb even after being folded and cut, demonstrating its excellent safety characteristics. Ultrasonic imaging serves as a powerful tool for assessing interfacial contact and side reactions [48, 49]. The peak‐to‐peak amplitude value (PPV) of the recorded maximum and minimum signals is converted into different colors, ranging from blue to red, to create ultrasonic images. Ultrasonic imaging was performed on NCM622 || Li pouch cells after different numbers of cycles at a 1C rate. As can be seen from Figure 6g, the NCM622 || bare Li cell initially exhibited small blue areas, which then expanded rapidly during cycling, indicating poor interfacial contact. In contrast, in Figure 6h, the NCM622 || DCF@ Li cell consistently displayed a uniform green color throughout the cycling process, demonstrating good wettability and interfacial contact. The good interfacial contact indicates that DCF can maintain its structural integrity due to its self‐healing capability and stabilize the lithium metal surface due to its self‐replenishment of the inorganic layer.

3. Conclusion

In summary, a dynamically cross‐linked fluoropolymer artificial SEI was designed to enable both polymer crack healing and continuous regeneration of inorganic species during cycling. Reversible quadruple hydrogen bonds allow the DCF polymer layer to autonomously repair physical cracks formed during cycling, while the fluorine‐rich matrix from TFA sustains the formation of LiF‐based inorganic components at the lithium surface. Consequently, the Li symmetric cell equipped with the DCF@Li anode exhibits exceptional long‐term cycling stability, operating steadily for over 4000 h. More importantly, this strategy demonstrates excellent compatibility with practical battery configurations. The assembled LFP||Li full cell delivers a remarkable capacity retention of 90.67% after 1000 cycles, underscoring its superior potential for practical application. This work highlights the importance of designing artificial SEIs that can actively recover from both mechanical and chemical degradation. Introducing regenerative functions into interfacial layers offers a practical route to sustaining lithium metal anodes for the development of long‐life lithium metal batteries.

Author Contributions

Ronghui Dou: investigation, software, methodology, conceptualization, writing – original draft, writing – review and editing, validation, formal analysis, and resources. Ruifeng Song: investigation. Kunchi Xie: investigation. Jia Lei: investigation. Jiamiao Li: software. Man Ge: investigation. Hongwen He: supervision. Yunhui Huang: supervision. Henghui Xu: supervision, funding acquisition, project administration, and writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: The authors have cited additional references within the Supporting Information [50, 51].

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.