Heterogeneous seeds boosting the self-lithiophilic host with dual-phase lithium storage for a stable lithium-metal anode

Abstract

Lithium (Li)-metal anode holds great promise for high-energy-density battery applications. However, the issue of uncontrollable Li dendrite growth, which is associated with large volume expansion during cycling, remains a significant hurdle. It is well known that the uniform Li⁺ flux, rich lithiophilic nucleation sites, and low local current density are of significant importance for inducing even Li deposition. Herein, a three-dimensional (3D) composite host was constructed by decorating an ultrafine Pt-nanoparticle layer on a carbon fiber framework (CF@Pt) via sputtering. CF with a high graphitic degree was in situ transformed into a lithiophilic LiC₆ phase upon charging, endowing self-lithiophilicity with a low Li nucleation energy barrier. A reversible “dual-phase” Li storage behavior (lithiation and metallization) was spontaneously realized in this 3D host with low local current density. Highly dispersed Pt heterogeneous nano-seeds further served as the lithiophilicity and Li nucleation boosters, consequently leading to even Li⁺ flux distribution and boosting the dense and smooth Li nucleation/growth. Additionally, the as-obtained CF@Pt host shows remarkably improved electrochemical performances in half-cells, symmetrical cells and full-cells.

---

Introduction

Lithium (Li) metal batteries (LMBs) are considered a promising technology for next-generation high-energy-density storage systems owing to their ultra-high theoretical specific capacity (3860 mA h g⁻¹) and low redox potential (−3.04 V vs. a standard hydrogen electrode).^1–3 However, the issues of uncontrollable dendrite growth and vast volume change in the Li anode during cycling result in a low Coulombic efficiency (CE), substantial capacity fade, and pose safety hazard, which have grievously limited their commercialization.^4–6 To date, considerable studies have been devoted to address these drawbacks, and the overall performances of the Li anodes have indeed effectively improved using various strategies, including artificial solid electrolyte interphase (SEI) construction,^7–9 solid/liquid-state electrolyte engineering,^10,11 separator structure design/modification^12,13 and 3D Li host construction.¹⁴ Among these, fabrication of a 3D Li host with a large surface area and high porosity has received increasing attention. This unique architecture can not only reduce the local current density and thus suppress the dendrite growth based on the widely accepted Sand's time model but also efficiently confines metallic Li within its porous structure and increases the tolerance to volume variation.^14–20

In light of their intrinsic properties of lightweight, structural flexibility and high conductivity, carbon materials are regarded as the ideal option for building such 3D hosts.^21,22 However, the conventional carbon-based host materials exhibit inferior Li affinity, leading to a large nucleation overpotential and uneven Li deposition.^14,23 For that, the modification of carbon skeletons with a wide variety of lithiophilic species, such as ZnO,^24,25 Co₃O₄,²⁶ and V₂O₅,²⁷ has been verified as an effective method. Nevertheless, a universal issue is that the popular synthesis routes for introducing these lithiophilic components, like electrochemical deposition and hydrothermal reaction, are time-consuming and cumbersome to scale up.²⁸ Moreover, the uneven distribution of lithiophilic sites always incurs heterogeneous deposition processes, resulting in an uncontrollable growth of Li dendrites. It has been proved that the dispersity and uniformity of lithiophilic sites could directly modulate the ion or electron conductivity of the 3D host, further redistributing the Li⁺ flux and electric field for Li deposition/stripping processes.²⁹ In this regard, developing simple and efficient methods for introducing lithiophilic sites with uniform distribution is essential.

In this work, carbon fiber (CF) with a high degree of graphitization was adopted to construct a self-lithiophilic 3D Li host. Lithiophilic LiC₆ ^30,31 with high ionic conductivity can be in situ formed via a simple lithiation reaction between the Li ions and this highly graphitized carbon, endowing the self-lithiophilic properties for the CF host. The in situ conversion avoids the cumbersome process of additional lithiophilic modification, and ensures the uniform distribution of lithiophilic sites. Meanwhile, the lithiophilic LiC₆ can reduce the nucleation barrier and the porous 3D structure of the host can decrease the local current density, further regulating the Li plating behavior during cycling. Moreover, the enhanced Li reversibility can be achieved via the reversible “dual-phase” Li storage behavior (lithiation and metallization) in this CF host. Considering the limited mass transfer flow and the “tip effects”, the Li⁺ flux distribution is uneven at the electrode/electrode interface, and the reduced Li mostly tends to deposit unevenly on the top-surface of the conductive 3D host.³² Hence, ultra-fine Pt nanoparticles were conformally and uniformly covered on the top surface of the CF host (CF@Pt) via sputtering. Specifically, the Pt metal has been reported to possess high lithiophilicity.^33,34 These highly dispersed and heterogeneous Pt nano-seeds can serve to boost the lithiophilicity and Li nucleation, dispersing the uneven Li⁺ flux and facilitating dense and smooth Li nucleation. As expected, within a fixed areal capacity of 1 mA h cm⁻², the as-obtained CF@Pt host demonstrates a high CE of 98.6% for more than 700 cycles at 1 mA cm⁻² in the half-cell. The corresponding CF@Pt/Li electrode-based symmetrical cell also delivers a small voltage hysteresis with a long cyclic stability of 1100 h at 0.5 mA cm⁻². This impressive Li plating/stripping behavior also features the CF@Pt anode with good adaptability to achieve a competitively high mass loading cathode (10.5 mg cm⁻²). Impressively, the as-fabricated CF@Pt/Li|LFP full cell maintained a high reversible capacity of 120.5 mA h g⁻¹ even after 300 cycles, of which a high-capacity retention rate of 90.7% was achieved.

Results and discussion

The schematic illustrations of the Li plating process of the Cu foil, CF and CF@Pt electrode are shown in Fig. 1a–c, respectively. For the planar Cu foil substrate (Fig. 1a), the presence of many uneven defects (such as protuberances) accompanied with poor lithiophilicity resulted in random Li nucleation and the development of irregular Li growth. Inevitably, owing to the “tip effect”, uncontrollable Li dendrites with large volume changes still appear in the pre-existing defects during repeated cycling process.²⁹ For the CF electrode (Fig. 1b), the porous architecture enables a high specific surface area and offers room for metallic Li confinement, which can reduce the local current density and thus suppress the Li dendrite growth to some content. Moreover, the in situ formed LiC₆ during Li plating process endows the CF host with lithiophilic property, which can decrease the nucleation energy barrier of Li. However, due to the shorter ion transportation paths and the uneven Li⁺ flux on the electrode/separator interface (especially, the pristine CF skeleton displays a lithiophobic nature without any regulated ability for Li⁺ flux), some Li⁺ would be interrupted and unevenly deposited on the surface of the conductive host. After continuous cycling, Li dendrites still appeared in the CF host. As shown in Fig. 1c, via a simple sputtering method, ultra-fine and lithiophilic Pt nanoparticles were effectively introduced on the upper layer of the CF host with high dispersity. A high dispersity of Pt nanoparticles indicates a high uniformity of the lithiophilic seeds.²⁹ Therefore, these heterogeneous Pt nano-seeds acted as Li nucleation seeds, boosting the uniform Li⁺ flux distribution at the electrode/separator interface and the dense Li nucleation/growth. In conjunction with the advantages of the CF host and heterogeneous Pt nano-seeds, the deposition behavior of Li in the CF@Pt electrode is smoother and more uniform. Although there is still some Li⁺ that is inevitably trapped and deposited on the surface, due to the homogeneous Li nucleation with the well-spaced distribution of lithiophilic sites, a more compact and dense plating morphology is achieved instead of uneven dendrite growth.

Fig. 1 Schematic of the Li plating process on different substrates: (a) bare Cu foil, (b) CF host, and (c) CF@Pt host.

The fabrication process of CF@Pt is schematically illustrated in Fig. 2a. Firstly, the commercial carbon fibers were interconnected into a 3D porous structure after a typical slurry-coating method. Following the sputtering of ultra-fine Pt particles on the top surface of the 3D CF host, CF@Pt was eventually fabricated. As shown in Fig. S1,† the Pt nanoparticles introduced via sputtering could be uniformly dispersed on the selected matrix, and their size was ultra-fine, ranging from 12–20 nm. FE-SEM images in Fig. 2b and Fig. S2a† reveal that the carbon fibers with nano-sized diameters (∼80 nm) are interconnected into a robust and porous structure, which can provide plenty of void spaces for Li confinement and act as fast Li ion/electron diffusion pathways. As displayed in Fig. 2c, the thickness of the pristine CF host is about 25 μm. There is no clear apparent difference after sputtering treatment (in Fig. 2d and Fig. S2b†), and the related EDS elemental mapping and spectrum (Fig. 2e–f and S3†) show that the ultra-fine Pt nanoparticles are successfully anchored on the upper layer of the CF host. From Fig. 2f, it can be observed that the average dispersion depth of the Pt layer is about 2 μm. Fig. 2d–i also demonstrate the high dispersity and uniform distribution of ultra-fine Pt nanoparticles on the CF scaffold. The numerous lithiophilic Pt nanoparticles help achieve a uniform Li ion flux at the host/separator interface and provide abundant nucleation sites with low nucleation barriers for smooth Li plating.

Fig. 2 (a) Schematic of the preparation procedures of the CF@Pt electrode; (b, c) top-view and cross-sectional SEM images of the CF electrode, respectively. (d–f) Cross-sectional SEM image of CF@Pt, and the corresponding EDS elemental mappings of C and Pt. (g–i) Top-view SEM image of CF@Pt, and the corresponding EDS elemental mapping of C and Pt.

To determine the Li plating behavior on the Cu foil, CF and CF@Pt, the morphology evolution of these three substrates under various Li areal capacities was inspected by FE-SEM. As shown in Fig. S4† and reflected by the formation of uneven needle-like Li dendrites, random Li plating behavior is observed on the Cu foil surface at various capacities of 1, 2 and 3 mA h cm⁻². Correspondingly, under high-capacity deposit loadings (2 and 3 mA h cm⁻²), the plating thickness recorded in the cross-sectional SEM image far exceeded the theoretical value (1 mA h cm⁻² of Li, ∼5 μm), further indicating that the fatal Li dendrite formation would generate loose packing of the plated Li with large volume expansion. For the CF substrate, the local current density is reduced due to the large specific area provided by the 3D electronic transmission network. Moreover, the in situ formed LiC₆ with high ion conductivity can provide fast Li ion diffusion channels and decrease the nucleation barrier. Hence, the Li plating behavior is effectively regulated, and metallic Li is well-confined in the host, while the sparse Li deposits can be observed on the surface after 1 mA h cm⁻² of Li plating (Fig. S5a–d†). With the increase of the Li plating capacities, the cavities among the fibers are gradually filled. However, the shorter ion diffusion distance enables more Li ions to be reduced on the top-surface of the CF host. Specifically, the sparse Li deposits covered on the surface would attract more Li ions than other areas during the Li plating process, further aggravating the inhomogeneous Li ion flux distribution.³⁵ Thus, as demonstrated in Fig. S5e–f and S5i–j,† the sparse Li deposits gradually evolved into Li chunks with dendrite morphologies and unevenly covered the surface of the CF host. In addition, the corresponding cross-sectional SEM images (in Fig. S5g–h and S5k–l†) demonstrate that Li prefers to be unevenly deposited on the surface of the CF host rather than the whole electrode, giving rise to severe volume change.

It should be noted that the homogeneous Li nucleation at the initial stage is critical for the delivery of uniform Li growth/plating with no Li dendrites.³⁶ Owing to the uniform distribution of ultra-fine Pt nanoparticles, the Li ion flux at the interface can be tuned so that it is more homogeneous. Moreover, the lower Li nucleation barrier and the homogeneous Li nucleation seeds are achieved during the initial Li nucleation process. Therefore, the Li plating behavior can be further regulated in CF@Pt. As demonstrated in Fig. 3, while depositing a fixed Li capacity into CF@Pt, the nanowire structure is gradually covered with Li species with the shrinkage of the porous area (Fig. 3a–f). Although some Li is deposited on the surface, the plating morphology is smooth without any dendrite aggregations and the thickness of the electrode in the cross-sectional SEM image is almost the same as that of the pristine CF@Pt. This indicates that a more uniform deposition is realized, which prevents the formation of more bulgy Li dendrites on the host surface. As shown in Fig. 3i and j, more metallic Li is accumulated and even covers the top surface of the electrode when the plating capacity reaches 3 mA h cm⁻². However, the Li-depositing morphology is still smooth and dense, and no obvious dendrite is observed on the surface. Moreover, the overall thickness of the electrode shows minimal variation at this stage, further illustrating the ability of good Li confinement and homogeneous Li plating of the CF@Pt host (Fig. 3k and l).

Fig. 3 Top-view and cross-sectional SEM images of the CF@Pt electrode after plating: (a–d) 1 mA h cm⁻², (e–h) 2 mA h cm⁻² and (i–l) 3 mA h cm⁻² at 0.5 mA cm⁻². In situ optical microscopy visualization of Li plating on the (m) Cu foil and (n) CF@Pt at 5 mA cm⁻²; the scale bar is 50 μm.

More intuitive evidence for the homogeneous Li plating behavior in the CF@Pt host was obtained from the in situ optical observation. Apparently, along with the extended Li plating time, many Li dendrites were grown, yielding an enlarged volume expansion on the bare Cu foil (Fig. 3m). In contrast, CF@Pt always displays a smooth dendrite-free surface and exhibits barely any volume change even after discharging for 30 min, suggesting its good regulation effect in Li plating/stripping behavior (Fig. 3n). To validate the superiority of the Pt nanoparticles in guiding the Li plating, the morphology evolution of the Pt-coated Cu foil (Cu@Pt) under various Li deposition capacities was investigated. Compared with the bare Cu foil counterpart, the even and compact Li blocks are gradually linked together and tightly attached to the current collector after continuously increasing the Li deposition capacity on Cu@Pt (Fig. S6†). The much smaller Li plating thickness in the corresponding cross-sectional SEM images infers that the heterogeneous Pt nano-seeds can effectively regulate the Li nucleation/growth behavior, and thus decrease the volume expansion of the whole electrode.

To evaluate the electrochemical performance of the CF@Pt electrode, the asymmetric cells were assembled by using Li metal as the counter electrode. Fig. S7† displays the voltage–capacity curves of the as-prepared asymmetric cells during the 1st plating process at 0.5 mA cm⁻². Compared with the bare Cu foil-based asymmetric cells, there is a significant capacity contribution platform above 0 V (versus Li/Li⁺), which can be observed in both CF and CF@Pt based asymmetric cells, mainly attributed to the intercalation reaction between Li⁺ and the graphitic carbon structure. The lithiophilic LiC₆ compounds can be formed in situ at this stage, endowing the lithiophilicity property of the graphitic CF host. Once the potential drops below 0 V (versus Li/Li⁺), the metallization reaction (consisting of the Li nucleation and growth process) will immediately start. The XRD measurement results further prove that the graphitic carbon of CF can be transformed into lithiophilic LiC₆ after the lithiation process (Fig. S8†). Typically, the nucleation overpotential is stipulated as the difference between the sharp voltage dip and the later stable platform potential, which is used to elucidate the nucleation barrier. The lower the nucleation barrier, the better the lithiophilicity of the matrix, and the more favorable conditions are toward achieving uniform Li nucleation and smooth Li plating behavior.^37,38 As shown in Fig. S7,† the Cu@Pt electrode delivers a smaller nucleation overpotential of 97.2 mV than that of the bare Cu foil (142.3 mV) during the initial plating, indicating the lithiophilicity property of the Pt nanoparticles. Impressively, the CF@Pt electrode exhibits the lowest nucleation overpotential among other substrates, which can be attributed to the synergistic effects of the self-lithiophilic 3D host and the heterogeneous Pt nucleation seeds. CE, an important parameter to evaluate the reversibility and efficiency of the Li metal anode, is calculated as the ratio of the specific charge capacity to discharge capacity in each cycle. Since the value and stability of the CE heavily depend on the current density and areal capacity,³⁹ the measurement was done at different current densities and areal capacities. As exhibited in Fig. 4a, with a constant plating capacity of 1.0 mA h cm⁻² at 0.5 mA cm⁻², the bare Cu foil delivers the lowest average CE value of 97.7%, and then dramatically drops after only 220 cycles, whereas the CF can be stabilized with an average CE of 98.2% for 500 cycles. By contrast, the CF@Pt electrode shows the highest average CE of 98.7% even for 800 cycles under the same conditions. As the current density increases to 1.0 mA cm⁻² (in Fig. 4b), the CF@Pt electrode also displays the longest cycling stability of 700 cycles with the highest average CE value of 98.6%. As a contrast, the CF electrode experiences significant fluctuations in CE after 450 cycles, and the bare Cu electrode sharply drops to less than 95% after only ∼220 cycles. These results indicate that the 3D structure of the host and the lithiophilic LiC₆/Pt can synergistically regulate the Li plating behavior, and thus enable the highest Li reversibility for the CF@Pt electrode. According to the Sand's time model theory, it is widely accepted that a larger current density means a shorter initial nucleation time, and the easier formation of Li dendrites.³⁹

Fig. 4 Comparison of the coulombic efficiency of the CF@Pt, CF, and Cu foil electrodes at different current densities: (a) at 0.5 mA cm⁻², (b) at 1 mA cm⁻² and (c) at 2 mA cm⁻² with a total capacity of 1 mA h cm⁻². (d) Performance comparison among recently reported host materials. (e, f) Voltage–capacity curves of the CF@Pt, CF, and Cu foil electrodes in the 50^th and 240^th cycles at 1.0 mA cm⁻² with a constant capacity of 1 mA h cm⁻².

The higher current density produces the lower CE of the Li metal anode. Therefore, we also tested the CEs of the as-prepared electrodes at the higher current density of 2 mA cm⁻², and the corresponding results are exhibited in Fig. 4c. Significantly, the CF and CF@Pt electrodes present higher average CE values of 96.7% and 98.0% for 290 and 380 cycles, respectively. Alternatively, the bare Cu foil shows a low average CE of 96.3% for only 151 cycles and then attenuates rapidly to 0 at the high current density, owing to the uneven Li nucleation/growth behavior and the uncontrollable growth of Li dendrites. When cycling at 1 mA cm⁻² with the deposition capacity up to 2 mA h cm⁻², CF@Pt can still achieve a stable average CE of 98.6% for 500 cycles, which is relatively higher than that of the CF (97.7% for 220 cycles) and Cu foil (97.0% for 80 cycles) counterparts (in Fig. S9†). Fig. 4d and Table S1† summarize the electrochemical performances of the asymmetric cells, which are assembled with the recently reported host materials,^24,38–47 manifesting that our work holds a conspicuous position under various Li plating/stripping conditions.

The charge/discharge curves of the Cu, CF and CF@Pt electrodes after different cycles at 1 mA cm⁻² are compared in Fig. 4e and f. Apparently, the Li deposition process is solely observed on the bare Cu foil electrode, whereas a clear Li⁺ intercalation reaction occurs prior to the Li metallization reaction in both CF and CF@Pt electrodes, suggesting that the “dual-phase” Li storage behavior (lithiation and metallization) takes place in this carbon framework. As shown in Fig. 4e, the Cu foil exhibits a lower polarization overpotential (44 mV) with respect to that of the CF (96 mV) and CF@Pt (58 mV) after 50 cycles. This might be attributed to the uncontrollable formation of Li dendrites on the bare Cu foil that provides a large specific surface area, resulting in reduced polarization of the cells. However, as demonstrated in Fig. 4f, the dendrites will dissolve and transform into a large amount of insulating “dead Li” after 240 cycles, which accumulates on the surface of the Cu foil, giving rise to the greatly increased overpotential (118 mV) with fluctuating profiles. In contrast, the polarization overpotentials of CF and CF@Pt are only slightly changed even after 240 cycles, suggesting that better cycling stability is achieved for these two electrodes. Interestingly, by comparing the discharge curves in Fig. 4e and f, the Li⁺ intercalation capacity contribution in the CF electrode is decreased with an obviously shortened platform above 0 V. This can be assigned to the preferential uneven Li plating on the top surface of CF, which intensifies the reaction by-product (like thick SEI, dead Li) accumulation on the surface. The Li⁺ diffusion pathways will be blocked, thereby inhibiting the Li⁺ intercalation into a carbon structure with decreased intercalation capacity. Owing to the homogeneous Li deposition behavior, the growth of dendrites is effectively suppressed and the Li⁺ transport pathways is still unobstructed in CF@Pt. The Li⁺ can be smoothly inserted into the carbon structure, and the dual-phase Li storage behavior is well-preserved in the CF@Pt host. Therefore, even after 240 cycles, a stable discharge profile with a steady intercalation and a plating capacity contribution is observed in CF@Pt, suggesting the highest Li plating/stripping reversibility.

To assess the electrochemical kinetics of the CF@Pt/Li composite anode (where Li is pre-deposited at 0.5 mA cm⁻²), the exchange current density (j₀) was gauged by assembling the symmetrical cells. In general, the higher j₀ demonstrates the faster charge-transfer kinetics of the interface and better distribution of the Li⁺ flux, leading to the decreased electrochemical polarization and the more even Li plating behavior.^48–50 Based on the Tafel plots in Fig. 5a, the j₀ of these three as-prepared composite Li anodes is calculated and further compared in Fig. 5b. Significantly, CF@Pt/Li presents a much higher j₀ of 1.99 mA cm⁻² relative to that of CF/Li (0.89 mA cm⁻²) and Cu/Li (0.37 mA cm⁻²), suggesting that the CF@Pt skeleton has faster electrochemical kinetics. The electrochemical performances of CF@Pt/Li in symmetrical cells were assessed with a fixed capacity of 1 mA h cm⁻². As shown in Fig. 5c, CF@Pt/Li exhibits stable cycling performance at 0.5 mA cm⁻² without fluctuation and the polarization voltage stabilizes at ∼13 mV for more than 1100 h, while CF/Li and Cu/Li start fluctuating and the voltage platform changes rapidly at about 780 and 280 h, respectively. Moreover, under a higher current density of 1 mA cm⁻² (in Fig. 5d), the CF@Pt/Li also presents the optimal cyclic stability with a low average overpotential of 20 mV for 770 h, in comparison with CF/Li and Cu/Li, where the vibration of the voltage hysteresis suddenly occurs with increased polarization at 420 h and 250 h, respectively. The increased polarization is triggered by the formation of dendrites and the accumulation of the by-products (dead Li and thick SEI),^14,51 which coincides well with the Li plating morphology in Fig. S4 and S5.† As shown in Table S2,† compared to the recently reported works, the as-obtained CF@Pt/Li electrode also shows competitive performance in the symmetric cell.

Fig. 5 (a) Tafel plots and (b) the corresponding calculated exchanging current densities of the CF@Pt/Li, CF/Li and Cu/Li electrodes. Galvanostatic cycling of the CF@Pt/Li, CF/Li and Cu/Li symmetric cells at (c) 0.5 mA cm⁻² and 1 mA h cm⁻² and (d) and 1 mA cm⁻² and 1 mA h cm⁻².

To evaluate the practical application prospect of the CF@Pt/Li anode, the electrochemical performances of the full cells were investigated by coupling with the high mass loading (∼10.5 mg cm⁻²) LFP cathode. As shown in Fig. 6a, the CF@Pt/Li|LFP full cell reveals a better rate performance than the CF/Li|LFP and Cu/Li|LFP cells. Although these three full cells exhibit similar capacities at low rates (∼159 mA h g⁻¹ at 0.1C, ∼154 mA h g⁻¹ at 0.2C), much higher specific capacities at high current densities (like 133 mA h g⁻¹ at 1C, 116 mA h g⁻¹ at 2C) are recorded for the CF@Pt/Li|LFP cell with respect to the other two counterparts (1C = 150 mA g⁻¹). Furthermore, when the current density changes back to 0.2 and 0.1C, the reversible capacity of CF@Pt/Li|LFP can be fully recovered, which are comparable to the initial capacities. As a contrast, the CF/Li|LFP suddenly drops to 129 and 109 mA h g⁻¹ at the high rates of 1C and 2C, while the Cu/Li|LFP decreases to 122 (at 1C) and 100 mA h g⁻¹ (at 2C), respectively. The long-term cycling performances at 1C were further measured, and the corresponding results are showed in Fig. 6b–e. The CF@Pt/Li|LFP still maintains a reversible capacity of 120.5 mA h g⁻¹ with a high-capacity retention of 90.7% even after 300 cycles (Fig. 6b and c). Furthermore, an ultra-high average CE of 99.8% is delivered. Compared with Cu/Li, the large surface area of the porous structure enables the reduced local current density, and the Li plating behavior is regulated in CF/Li. However, the uneven Li plating still exists, which will inevitably result in dead Li accumulation and thus decrease the cyclic reversibility of the CF/Li anode. Correspondently, although a high initial capacity of 133.0 mA h g⁻¹ is achieved, CF/Li|LFP can only maintain a reversible capacity of 103.7 mA h g⁻¹ after 180 cycles with a low average CE of 99.6% (Fig. 6d). In sharp contrast, the capacity of the Cu/Li|LFP cell rapidly fades to 73.9 mA h g⁻¹ with the shortest cycling lifespan of 180 cycles, showing the lowest average CE of 99.3% and suggesting the worst Li plating/stripping reversibility (Fig. 6e).

Fig. 6 (a) Rate performances of the CF@Pt/Li|LFP, CF/Li|LFP and Cu/Li|LFP full cells ranging from 0.1C to 2C. (b) Cycling performances and (c–e) corresponding voltage profiles (at the 1^st, 20^th, 50^th, 120^th and 180^th cycle) of the CF@Pt/Li|LFP, CF/Li|LFP and Cu/Li|LFP full cells at 1C. Top-view SEM images of the CF@Pt/Li (f), CF/Li (g) and Cu/Li (h) anode in the LFP-based full cells after 120 cycles.

As shown in Fig. 6c–e, the voltage polarizations of these three as-fabricated full cells slightly increase within the repeated charge/discharge process. Interestingly, within the continuous cycling, the charge/discharge curves of the CF/Li|LFP and CF@Pt/Li|LFP cells are gradually changed from one long flat platform into one platform and one short slope (Fig. 6c and d). Considering that the 3D skeleton of CF is a self-lithiophilic host material, it can realize a dual-phase Li storage behavior via lithiation and metallization. Therefore, the emerging short slope in the charge/discharge profiles can be attributed to the capacity contribution of the lithiation/de-lithiation process, which will act as a replenishment Li source for further improving the Li cyclic reversibility. In spite of the sharp degradation in the capacity, the Cu/Li|LFP cell only presents typical Li metallization behavior with a gradually shortened charging/discharging platform, and no slope appears. Such inferior cyclic stability of the Cu/Li|LFP cell should be attributed to the worsened morphology. As shown in Fig. 6f, the cycled morphology of CF@Pt/Li is still flat and undamaged, whereas some protuberances aggregate and cover the cycled CF/Li anode (Fig. 6g), indicating that the lithiophilic Pt induces more homogenous Li plating and better Li dendrite suppression. In sharp contrast, the cycled Cu/Li anode presents a completely deteriorated interface morphology with the surface covered by rugged and loose Li species (Fig. 6h). This can be assigned to the overgrown dendrites in the Cu/Li anode, inevitably transforming into the large amount of dead Li during cycling, which results in the inferior Li reversibility with uncontrollable volume expansion.

Conclusions

In summary, 3D CF@Pt was proposed to host Li for high performance LMBs. Specially, the carbon fiber (CF) with its high degree of graphitization could be in situ transformed into the lithiophilic LiC₆ during cycling, endowing the self-lithiophilic property for the 3D porous CF frameworks. This design avoids the cumbersome process of lithiophilic modification for conventional lithiophobic carbon materials. Meanwhile, the Li nucleation barrier can be reduced and a “dual-phase” Li storage behavior of lithiation and metallization is achieved by the self-lithiophilic CF framework. On the other hand, the ultrafine, highly lithiophilic and heterogeneous Pt nanoparticles can serve as lithiophilic sites and boost Li nucleation, dispersing the Li⁺ flux and promoting the dense Li nucleation/growth. Thus, smooth Li plating/stripping behavior with dendrites-free morphology can be achieved by the synergistic effect of the 3D CF framework and the heterogeneous Pt seeds. All these benefits enable the as-obtained CF@Pt host to achieve a high average CE of 98.6% for more than 700 cycles at 1 mA cm⁻² in the half cell, and a small voltage hysteresis with a long cyclic stability of 1100 h at 0.5 mA cm⁻² in the corresponding Li-contained symmetrical cell. Even when combined with the high mass loading LFP cathode (10.5 mg cm⁻²), the as-prepared CF@Pt/Li anode shows better performance. A high reversible capacity of 120.5 mA h g⁻¹ could be maintained even after 300 cycles, with a high capacity retention rate of 90.7% achieved in a CF@Pt/Li|LFP full cell. This work may provide a new avenue to develop a promising Li metal anode for high energy density batteries.

Author contributions

Zhicui Song: methodology, data curation, investigation, software, writing-original draft, visualization. Jing Xue: software, investigation, validation, writing-review & editing. Chaohui Wei: software, methodology, validation, writing-review & editing. Donghuang Wang: data curation, funding acquisition, resources, supervision, validation. Yingchun Ding: data curation, validation. Aijun Zhou: data curation, validation, visualization. Jingze Li: conceptualization, funding acquisition, resources, project administration, supervision, writing-review & editing.

Data availability

Data will be made available on request.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work is partly supported by the National Natural Science Foundation of China (No. 22379019 and 52172184), Sichuan Science and Technology Program (No. 2024YFHZ0313), Huzhou Natural Science Foundation Project (No. 2022YZ04), and the National Science Foundation of Sichuan Province (No. 2022NSFSC0259).

References

1. T. Lyu, F. Luo, D. Wang, L. Bu, L. Tao and Z. Zheng, Carbon/Lithium Composite Anode for Advanced Lithium Metal Batteries: Design, Progress, In Situ Characterization, and Perspectives, Adv. Energy Mater., 2022, 12, 2201493.
2. J. Xing, T. Chen, Z. Wang, Z. Song, C. Wei, Q. Deng, Q. Zhao, A. Zhou and J. Li, Revisiting porous foam Cu host based Li metal anode: The roles of lithiophilicity and hierarchical structure of three-dimensional framework, J. Colloid Interface Sci., 2024, 673, 638–646.
3. Z. Wang, Z. Du, Y. Liu, C. E. Knapp, Y. Dai, J. Li, W. Zhang, R. Chen, F. Guo, W. Zong, X. Gao, J. Zhu, C. Wei and G. He, Metal–organic frameworks and their derivatives for optimizing lithium metal anodes, eScience, 2024, 4, 100189.
4. C. Wang, C. Yang, Y. Du, Z. Guo and H. Ye, Spherical Lithium Deposition Enables High Li-Utilization Rate, Low Negative/Positive Ratio, and High Energy Density in Lithium Metal Batteries, Adv. Funct. Mater., 2023, 33, 2303427.
5. S. Ye, X. Chen, R. Zhang, Y. Jiang, F. Huang, H. Huang, Y. Yao, S. Jiao, X. Chen, Q. Zhang and Y. Yu, Revisiting the Role of Physical Confinement and Chemical Regulation of 3D Hosts for Dendrite-Free Li Metal Anode, Nano–Micro Lett., 2022, 14, 187.
6. J. Xing, L. Yan, T. Chen, Z. Song, Z. Wang, Y. Liu, L. Zhou and J. Li, Highly lithiophilic and structurally stable Cu–Zn alloy skeleton for high-performance Li-rich ternary anodes, J. Colloid Interface Sci., 2023, 652, 627–635.
7. Y. Feng, B. Zhong, R. Zhang, M. Peng, Z. Hu, Z. Wu, N. Deng, W. Zhang and K. Zhang, Achieving High-Power and Dendrite-Free Lithium Metal Anodes via Interfacial Ion-Transport-Rectifying Pump, Adv. Energy Mater., 2023, 13, 2203912.
8. S. Li, J. Huang, Y. Cui, S. Liu, Z. Chen, W. Huang, C. Li, R. Liu, R. Fu and D. Wu, A robust all-organic protective layer towards ultrahigh-rate and large-capacity Li metal anodes, Nat. Nanotechnol., 2022, 17, 613–621.
9. Y. Xu, Y. Song, Z. Chen, J. Yu, J. Wang, M. He, J. Xu, J. Luo and W. Yao, Demonstration of MAX phases as triple functional artificial solid electrolyte interphase for ultralong life lithium metal anodes, J. Colloid Interface Sci., 2025, 679, 737–746.
10. L.-P. Hou, N. Yao, J. Xie, P. Shi, S.-Y. Sun, C.-B. Jin, C.-M. Chen, Q.-B. Liu, B.-Q. Li, X.-Q. Zhang and Q. Zhang, Modification of Nitrate Ion Enables Stable Solid Electrolyte Interphase in Lithium Metal Batteries, Angew. Chem., Int. Ed., 2022, 61, e202201406.
11. C. Hu, Y. Shen, M. Shen, X. Liu, H. Chen, C. Liu, T. Kang, F. Jin, L. Li, J. Li, Y. Li, N. Zhao, X. Guo, W. Lu, B. Hu and L. Chen, Superionic Conductors via Bulk Interfacial Conduction, J. Am. Chem. Soc., 2020, 142, 18035–18041.
12. Z. Chang, H. Yang, A. Pan, P. He and H. Zhou, An improved 9 micron thick separator for a 350 Wh/kg lithium metal rechargeable pouch cell, Nat. Commun., 2022, 13, 6788.
13. L. Sheng, Q. Wang, X. Liu, H. Cui, X. Wang, Y. Xu, Z. Li, L. Wang, Z. Chen, G.-L. Xu, J. Wang, Y. Tang, K. Amine, H. Xu and X. He, Suppressing electrolyte-lithium metal reactivity via Li⁺-desolvation in uniform nano-porous separator, Nat. Commun., 2022, 13, 172.
14. Z. Song, Y. Liu, Z. Wang, J. Xing, C. Wei, W. Zou, A. Zhou and J. Li, Synergistic modulation of Li nucleation/growth enabled by CNTs-wrapped lithiophilic CoP/Co₂P decorated hollow carbon polyhedron host for stable lithium metal anodes, Nano Res., 2023, 16, 4961–4969.
15. J. Yun, B.-K. Park, E.-S. Won, S. H. Choi, H. C. Kang, J. H. Kim, M.-S. Park and J.-W. Lee, Bottom-Up Lithium Growth Triggered by Interfacial Activity Gradient on Porous Framework for Lithium-Metal Anode, ACS Energy Lett., 2020, 5, 3108–3114.
16. S. Huang, H. Zhang and L.-Z. Fan, Confined Lithium Deposition Triggered by an Integrated Gradient Scaffold for a Lithium-Metal Anode, ACS Appl. Mater. Interfaces, 2022, 14, 17539–17546.
17. J. Sun, Y. Cheng, H. Zhang, X. Yan, Z. Sun, W. Ye, W. Li, M. Zhang, H. Gao, J. Han, D.-L. Peng, Y. Yang and M.-S. Wang, Enhanced Cyclability of Lithium Metal Anodes Enabled by Anti-aggregation of Lithiophilic Seeds, Nano Lett., 2022, 22, 5874–5882.
18. J. Jiang, Z. Pan, Z. Kou, P. Nie, C. Chen, Z. Li, S. Li, Q. Zhu, H. Dou, X. Zhang and J. Wang, Lithiophilic polymer interphase anchored on laser-punched 3D holey Cu matrix enables uniform lithium nucleation leading to super-stable lithium metal anodes, Energy Storage Mater., 2020, 29, 84–91.
19. L. Cao, M. Chu, Y. Li, X. Xu, Y. Qiu, Y. Dai, C. Sun, Z.-X. Huang, X.-L. Wu and H. Geng, In Situ-Constructed Multifunctional Composite Anode with Ultralong-Life Toward Advanced Lithium-Metal Batteries, Adv. Mater., 2024, 36, 2406034.
20. L. Yi, Z. Wang, X. Chen, J. Xing, H. Huang, C. Wei, Q. Zhao, A. Zhou and J. Li, An electron-insulating Li2O protection layer endowing a Li–Cu–Zn ternary alloy composite anode with high performance, Chem. Commun., 2024, 60, 5832–5835.
21. Q. Lu, Y. Jie, X. Meng, A. Omar, D. Mikhailova, R. Cao, S. Jiao, Y. Lu and Y. Xu, Carbon materials for stable Li metal anodes: Challenges, solutions, and outlook, Carbon Energy, 2021, 3, 957–975.
22. W. Zihao, C. Tao, S. Zhicui, X. Jianxiong, Z. Aijun and L. Jingze, Porous array of BaLi₄ alloy microchannels enforced carbon cloth for a stable Li composite anode, Energy Mater., 2024, 4, 400031.
23. Z. Song, J. Xue, C. Wei, Q. Zhao, A. Zhou and J. Li, Head-Mounted coating on graphite host enabling highly reversible Li Plating/Stripping in Lithium-ion/Lithium metal hybrid anode, Chem. Eng. J., 2024, 490, 151618.
24. Z. Xiao, D. Han, Y. Fu, K. Xie, W. Tian, C. Shu, K. Xi, C. Peng, Y. Wu, S. Dou and W. Tang, Eliminating concentration polarization with gradient lithiophilic sites towards high performance lithium metal anodes under low N/P ratio, Chem. Eng. J., 2024, 480, 148029.
25. T. Wei, C. Sun, X. Guo, Y. Zhou, M. Wang, X. Qiu, Q. Wang and Y. Tang, Petaloid bimetallic metal-organic frameworks derived ZnCo₂O₄/ZnO nanosheets enabled intermittent lithiophilic model for dendrite-free lithium metal anode, J. Colloid Interface Sci., 2024, 664, 596–606.
26. G. Jiang, N. Jiang, N. Zheng, X. Chen, J. Mao, G. Ding, Y. Li, F. Sun and Y. Li, MOF-derived porous Co₃O₄-NC nanoflake arrays on carbon fiber cloth as stable hosts for dendrite-free Li metal anodes, Energy Storage Mater., 2019, 23, 181–189.
27. X. Zhang, D. Jin, C. Guo, L. Ke, N. Li, X. Zhang, K. Xu, K. Rui, H. Lin, Y. Zhang, L. Wang and J. Zhu, Achieving Electronic Engineering of Vanadium Oxide-Based 3D Lithiophilic Sandwiched-Aerogel Framework for Ultrastable Lithium Metal Batteries, ACS Appl. Mater. Interfaces, 2022, 14, 33306–33314.
28. C. Guo, Y. Guo, R. Tao, X. Liao, K. Du, H. Zou, W. Zhang, J. Liang, D. Wang, X.-G. Sun and S.-Y. Lu, Uniform lithiophilic layers in 3D current collectors enable ultrastable solid electrolyte interphase for high-performance lithium metal batteries, Nano Energy, 2022, 96, 107121.
29. Y. Liu, J. Sun, X. Hu, Y. Li, H. Du, K. Wang, Z. Du, X. Gong, W. Ai and W. Huang, Lithiophilic sites dependency of lithium deposition in Li metal host anodes, Nano Energy, 2022, 94, 106883.
30. Z. Wang, Q. Deng, Z. Song, Y. Liu, J. Xing, C. Wei, Y. Wang and J. Li, Ultrathin Li-rich Li-Cu alloy anode capped with lithiophilic LiC6 headspace enabling stable cyclic performance, J. Colloid Interface Sci., 2023, 643, 205–213.
31. Z. Song, C. Wei, J. Jiang, D. Wang, X. Wang, Q. Deng, Q. Zhao, A. Zhou, H. Li and J. Li, Double-gradient host enabling bottom-up Li deposition towards hybrid lithium-ion/metal anode with long lifespan, Chem. Eng. J., 2024, 501, 157755.
32. G. Peng, Q. Zheng, G. Luo, D. Zheng, S.-P. Feng, U. Khan, A. R. Akbar, H. Luo and F. Liu, A Gradient Lithiophilic Structure for Stable Lithium Metal Anodes with Ultrahigh Rate and Ultradeep Capacity, Small, 2023, 19, 2303787.
33. P. Afzali, E. Gibertini and L. Magagnin, Improved plating/stripping in anode-free lithium metal batteries through electrodeposition of lithiophilic zinc thin films, Electrochim. Acta, 2024, 488, 144190.
34. K. Wen, L. Liu, S. Chen and S. Zhang, A bidirectional growth mechanism for a stable lithium anode by a platinum nanolayer sputtered on a polypropylene separator, RSC Adv., 2018, 8, 13034–13039.
35. X. Yan, Q. Zhang, W. Xu, Q. Xie, P. Liu, Q. Chen, H. Zheng, L. Wang, Z.-Z. Zhu and D.-L. Peng, Bottom-top channeling Li nucleation and growth by a gradient lithiophilic 3D conductive host for highly stable Li-metal anodes, J. Mater. Chem. A, 2020, 8, 1678–1686.
36. S. Cui, P. Zhai, W. Yang, Y. Wei, J. Xiao, L. Deng and Y. Gong, Large-Scale Modification of Commercial Copper Foil with Lithiophilic Metal Layer for Li Metal Battery, Small, 2020, 16, 1905620.
37. S. Huang, W. Zhang, H. Ming, G. Cao, L.-Z. Fan and H. Zhang, Chemical Energy Release Driven Lithiophilic Layer on 1 m2 Commercial Brass Mesh toward Highly Stable Lithium Metal Batteries, Nano Lett., 2019, 19, 1832–1837.
38. T. Lyu, F. Luo, Z. Wang, F. Jiang, S. Geng, Y. Zhuang, X. Lin, J. Chen, D. Wang, L. Bu, L. Tao, L. Liang and Z. Zheng, Bifunctional lithiophilic carbon fibers with hierarchical structure for high-energy lithium metal batteries, Chem. Eng. J., 2023, 466, 143357.
39. H. Liu, J. Di, P. Wang, R. Gao, H. Tian, P. Ren, Q. Yuan, W. Huang, R. Liu, Q. Liu and M. Feng, A novel design of 3D carbon host for stable lithium metal anode, Carbon Energy, 2022, 4, 654–664.
40. K. Yang, Y. He, Z. Gao, C. Li, F. Zhao, X. Xu, J. Li and X. Liu, Constructing a TiN/CNT lithiophilic scaffold for dendrite-free Li-metal anode, Carbon, 2024, 222, 118999.
41. Y. Xu, L. Wang, W. Jia, Y. Yu, R. Zhang, T. Li, X. Fu, X. Niu, J. Li and Y. Kang, Three-dimensional carbon material as stable host for dendrite-free lithium metal anodes, Electrochim. Acta, 2019, 301, 251–257.
42. G. Wang, T. Liu, X. Fu, Z. Wu, M. Liu and X. Xiong, Lithiophilic amide-functionalized carbon nanotube skeleton for dendrite-free lithium metal anodes, Chem. Eng. J., 2021, 414, 128698.
43. J. Ma, J. Yang, C. Wu, M. Huang, J. Zhu, W. Zeng, L. Li, P. Li, X. Zhao, F. Qiao, Z. Zhang, D. He and S. Mu, Stabilizing nucleation seeds in Li metal anode via ion-selective graphene oxide interfaces, Energy Storage Mater., 2023, 56, 572–581.
44. M. Gao, Q. Dong, M. Yao, X. Wang, J. Li, W. Zhang, H. Huang, H. Guo, Z. Sun, Q. Chen, X. Han and W. Hu, Dual-gradient Engineering of Urchin-like Silver@Copper Oxide Heterostructures for Highly Stable Lithium Metal Anodes, Adv. Funct. Mater., 2024, 34, 2401442.
45. X. L. Li, S. Huang, D. Yan, J. Zhang, D. Fang, Y. V. Lim, Y. Wang, T. C. Li, Y. Li, L. Guo and H. Y. Yang, Tuning Lithiophilicity and Stability of 3D Conductive Scaffold via Covalent Ag-S Bond for High-Performance Lithium Metal Anode, Energy Environ. Mater., 2023, 6, e12274.
46. D. Xie, H.-H. Li, W.-Y. Diao, R. Jiang, F.-Y. Tao, H.-Z. Sun, X.-L. Wu and J.-P. Zhang, Spatial confinement of vertical arrays of lithiophilic SnS₂ nanosheets enables conformal Li nucleation/growth towards dendrite-free Li metal anode, Energy Storage Mater., 2021, 36, 504–513.
47. H. Jiang, Y. Zhou, H. Zhu, F. Qin, Z. Han, M. Bai, J. Yang, J. Li, B. Hong and Y. Lai, Interconnected stacked hollow carbon spheres uniformly embedded with Ni2P nanoparticles as scalable host for practical Li metal anode, Chem. Eng. J., 2022, 428, 132648.
48. T. Jiang, L. Sun, Y. Zhang, X. Zhang, H. Lin, K. Rui and J. Zhu, Lithiophilic interface dynamic engineering to inhibit Li dendrite growth for intrinsically safe Li-metal batteries, Chem. Eng. J., 2023, 464, 142555.
49. D. T. Boyle, X. Kong, A. Pei, P. E. Rudnicki, F. Shi, W. Huang, Z. Bao, J. Qin and Y. Cui, Transient Voltammetry with Ultramicroelectrodes Reveals the Electron Transfer Kinetics of Lithium Metal Anodes, ACS Energy Lett., 2020, 5, 701–709.
50. M. Huang, Z. Yao, Q. Yang and C. Li, Consecutive Nucleation and Confinement Modulation towards Li Plating in Seeded Capsules for Durable Li-Metal Batteries, Angew. Chem., Int. Ed., 2021, 60, 14040–14050.
51. H. Yong Li, G. Zhen Yi, G. Jin Zhi, W. Xiao Tong, Z. Xin Xin and W. Xing Long, Research progress on the surface/interface modification of high-voltage lithium oxide cathode materials, Energy Mater., 2022, 2, 200017.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5qi00286a

---