Ultrathin 3D Cu/Li Composite with Enhanced Li Utilization for High Energy Density Li‐Metal Battery Anodes

Abstract

Li‐metal batteries (LMBs) are promising candidates for next‐generation energy storage devices because of their high energy densities. However, limitations of Li‐metal anodes (LMAs) such as dendrite formation hinder their practical application. This paper reports an ultrathin 3D Cu/Li composite anode (Li in 3DCu) with a thickness of <30 µm and moderate Li loading of 5 mA h cm⁻² via electrochemical etching and electrodeposition, followed by thermal infiltration of Li. The lightweight composite anode achieves a specific capacity of 514 mA h g⁻¹ while effectively reducing current density and suppressing dendritic growth, thus enabling stable performance at high current densities. Novel insights regarding the Li infiltration mechanism are obtained via an integrated analysis of forces, interfacial chemistry, and thermodynamics, offering a comprehensive understanding for Li infiltration. Electrochemical characterization indicates that the proposed composite anode (Li in 3DCu) achieves a 335% improvement in cycle life compared to that when using a conventional anode with Li on a Cu foil in a Li@Cu||LFP cell. This study establishes a robust platform for lightweight high‐performance LMAs by combining structural innovations, maximizing Li utilization, and broadens the understanding of infiltration mechanisms to develop next‐generation LMBs.

This study presents a lightweight 3D Cu/Li composite anode with a thickness of 30 µm, achieving a high specific capacity of 514 mA h g⁻¹ and exceptional cycling stability. Insights into molten Li infiltration mechanisms and advanced structural designs highlight a transformative approach for scalable, high‐performance Li‐metal batteries.

1. Introduction

Li metal is emerging as a promising next‐generation anode material because of its remarkable theoretical capacity of 3860 mA h g⁻¹, which significantly exceeds the 372 mA h g⁻¹ of graphite anodes.^{[ 1 ]} However, the growth of Li dendrites during repeated charge–discharge cycles remains challenging, primarily due to non‐uniform Li deposition and unstable solid–electrolyte interphase (SEI) formation, which lead to high local current densities and tip‐induced field enhancement that accelerate dendrite propagation.^{[ 2} , ^{3 ]} These dendrites can eventually penetrate the separator, causing short circuits and severely limiting the cycle life of Li‐metal batteries (LMBs).^{[ 4 ]} Thus far, various strategies have been proposed, including electrolyte modification,^{[ 5 ]} advanced separators,^{[ 6 ]} and anode surface modification.^{[ 7} , ⁸ , ⁹ , ¹⁰ , ^{11 ]} Among these, the structural designs of anode current collectors have shown significant potential for mitigating dendrite formation.^{[ 4} , ^{12 ]} Structured current collectors can effectively reduce the local current density at the anode, as described by Sand's equation, thus delaying the onset and slowing the progression of dendrite growth.^{[ 13} , ^{14 ]}

Although 3D structures show promise in mitigating Li dendrite growth, filling these structures with Li for preparing Li metal anodes (LMA) remains challenging, mainly because of the complex interplay between molten Li dynamics, surface wettability, and internal structure geometry, which has not been fully understood or optimized.^{[ 15 ]} Mechanical pressing, which is a low‐cost process commonly used for 2D structures such as thin foils, is unsuitable for 3D frameworks because it cannot force Li to fill internal voids in 3D structures. Alternatively, electroplating can achieve Li deposition within 3D structures. However, it requires expensive electrolytes, which makes it economically inefficient.^{[ 16 ]} To address these limitations, the infiltration of molten Li into 3D structures has gained considerable attention as a cost‐effective alternative.^{[ 17} , ^{18 ]} This method eliminates the need for expensive and hazardous electrolytes and offers a practical approach to prepare composite LMAs. The intrinsic lithiophobicity of Cu presents additional challenges for infiltrating molten Li, as it results in poor wetting behavior that hinders capillary‐driven infiltration, especially in narrow porous structures.^{[ 19} , ^{20 ]} Early studies solved this issue by employing surface modification strategies such as coating Cu with lithiophilic materials to improve wetting and enable uniform infiltration.^{[ 15} , ^{21 ]}

Despite these advances, most reported composite LMAs suffer from excessive thickness and oversized Li content. This results in limited improvement in practical energy density and inefficient use of active material. Furthermore, the absence of standardized metrics for evaluating areal capacity versus Li excess makes it difficult to assess the real contributions of structural innovation.^{[ 22} , ^{23 ]} Commonly used 3D frameworks for composite LMAs such as metal foam, carbon felt, and carbon cloth typically have thicknesses ranging from 200 µm to 1 mm. The resulting composite anodes possess corresponding thicknesses and inevitably incorporate an excessive amount of Li. For LMAs with these characteristics, it is difficult to determine if their improved performance comes from excess Li or from addressing the underlying technical challenges. These issues conflict with the core goal of LMBs, namely, achieving high energy density. Considering that the ideal areal capacity of Li‐metal content in LMA is ≈4 mA h cm⁻²,^{[ 24} , ^{25 ]} (equivalent to 20‐µm‐thick Li) composite LMA should target an optimized structure with a comparable thickness that can match such areal capacity to ensure efficient Li utilization while enhancing cycling stability.^{[ 26 ]} However, excessive thickness and oversized Li in most composite LMAs waste resources and reduce the energy density of the cell, rendering them impractical for practical applications.^{[ 22 ]} Although these LMAs demonstrated feasibility in laboratory settings, they fail to satisfy the energy density requirements for commercialization.^{[ 22 ]} This highlights the need for a thin, high‐porosity LMA that can efficiently utilize Li while maintaining structural robustness during repeated cycles, thereby ensuring both high energy density and long‐term electrochemical stability.^{[ 27 ]} This design ensures stability, delivers a high performance, and aligns with the practical demands of energy‐dense LMBs.

In this paper, we propose an ultrathin 3D Cu/Li composite anode (Li in 3DCu) that overcomes the limitations of conventional LMA and delivers novel insights into molten Li infiltration mechanisms, as shown in Figure 1a. The 3D Cu framework fabricated via electrochemical etching and electrodeposition achieves a uniform thickness of 24 µm and a low areal mass of 8.53 mg cm⁻², which enables exceptional gravimetric energy density. This study clarifies the molten Li infiltration process by analyzing the interplay of forces, interfacial chemistry, and thermodynamics, which demonstrates how capillary‐driven Li infiltration can be enhanced by lithiophilic Li₂O layers derived from low concentrations of CuO, Figure 1b. The ultrathin thickness and moderate Li infiltration of ≈4 mA h cm⁻² delivers a high specific capacity of 514 mA h g⁻¹, which significantly outperforms conventional thick frameworks with excessive Li, Figure 1c.^{[ 15} , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ , ^{46 ]} Electrochemical characterization revealed that Li in 3DCu exhibited a stable cycling performance, retaining over 80% of its initial capacity after 100 cycles and demonstrating a 335% improvement in cycle life compared to that of Li on 2DCu. In addition, the enhanced rate capability and ability to mitigate dendritic growth highlight the critical role of the lightweight 3D Cu framework in enabling stable high‐performance LMBs. This study provides a robust platform to advance LMB technology and offer new pathways for lightweight and scalable anode designs.

Figure 1.

Ultrathin Cu/Li composite anode: a) Infiltration of molten Li into ultrathin 3D Cu. b) Illustration of molten Li infiltration into ultrathin 3D Cu driven by capillary force. c) Comparison of anode thickness and Li loading for various composite LMAs fabricated by molten Li infiltration, highlighting superb Li utilization of Li in 3DCu proposed in this paper.

2. Results and Discussion

2.1. Fabrication of Ultrathin 3D Cu Framework

The fabrication process of the ultrathin 3D Cu framework is illustrated in Figure 2a–c. Figure 2a shows a commercially available Cu mesh was used as the backbone. The mesh offers high porosity and a straightforward design; however, it is thin compared to other porous sheets such as metallic foams.^{[ 47} , ^{48 ]} The commonly known method of fabricating ultra‐thin 3D copper frames through foam etching exhibits limited thickness reduction due to restricted internal etching and poor uniformity resulting from non‐homogeneous etching rates.^{[ 49 ]} In contrast, meshes fabricated with wires of constant diameter ensure uniformity during the etching process, and the reduction in wire diameter facilitates the thinning of the structure.^{[ 50 ]} Top‐view and cross‐sectional images of the Cu mesh are presented in Figure 2d,e, respectively. The wire diameter, inter‐wire spacing, and total thickness of the mesh were measured as 30, 45, and 50 µm, respectively.

Figure 2.

Fabrication of ultrathin 3D structured Cu (3DCu). Schematic showing the fabrication process where the a) Cu mesh is etched to produce b) a thin‐wired Cu mesh, and a subsequent electrodeposition for the preparation of c) 3DCu. SEM images for d and e) Cu mesh, f and g) thin‐wired Cu mesh, and h‐j) 3DCu. Figure 2d,f,h, and i correspond to the top‐view image, and Figure 2e,g,j correspond to the cross‐sectional image. k) Thickness and areal mass of the 3D structured Cu with and without etching. l) Specific capacity of Li‐infiltrated 3D structured Cu with and without etching corresponding to various Li loading values.

The structure was significantly lightweight and thinner by using galvanostatic electrochemical etching in a sulfuric acid and copper sulfate aqueous solution, as indicated in Figure 2b. The wire diameter was reduced to below 10 µm, and the overall thickness of the structure was decreased to under 20 µm, as shown in Figure 2f,g. Further, the pores of the mesh were filled with Cu nanorods by applying a voltage of 2.8 V for 4 min of electrodeposition in the same solution, as illustrated in Figure 2c,h,i. These nanorods with an average diameter of ≈1 µm offer a significant advantage because of high electrical conductivity of Cu compared to that of Li. Cu nanorods can enhance charge transport within the electrode by filling the voids in the mesh, which contributes to an improved electrochemical performance. The final thickness of the thin‐wired Cu mesh with Cu nanorods (3DCu) was measured to be ≈24 µm, with high uniformity across the structure, as confirmed in Figure 2j. This is significantly thinner than 3D Cu without electrochemical etching (Thick 3DCu) shown in Figure S1 (Supporting Information).

Structural properties with and without electrochemical etching are compared in Figure 2k and Table S1 (Supporting Information). Without etching, the Thick 3DCu exhibits a thickness of 80 µm and an areal mass of 28 mg cm⁻². In contrast, the etched 3D Cu reduces these values to 24 µm and 8.53 mg cm⁻², which makes it even lighter than a 9‐µm‐thick Cu foil. Given that Cu has a mass density of 8.96 g cm⁻³ compared to that of 0.534 g cm⁻³ for Li, reducing the mass of the Cu framework is critical for minimizing its effect on the overall gravimetric energy density of the anode. Figure 2l highlights the advantages of the lightweight 3D Cu framework, achieving a specific capacity exceeding 400 mA h g⁻¹ with a moderate Li infiltration of ≈4 mA h cm⁻². In comparison, the Thick 3DCu requires excessive Li loading to achieve the same capacity. These results indicate that optimizing the Cu framework mass is essential for achieving a practical gravimetric energy density in Cu/Li composite anodes, which is a key factor in advancing lightweight, high‐performance LMB designs.

2.2. Molten Li Infiltration into 3DCu

The infiltration of molten Li into the ultrathin and lightweight 3DCu, as indicated in Figure 3a, enables the facile preparation of a Cu/Li composite anode without wasting a high‐cost electrolyte. We provide novel insights into the mechanism of molten Li infiltration into a porous 3DCu sheet via an integrated analysis of forces, interfacial chemistry, and thermodynamics, which offers a deeper understanding of efficient Li infiltration, as shown in Figure 3b. A thin CuO layer on 3DCu enhances Li infiltration by forming lithiophilic Li₂O, generating capillary forces that drive complete Li penetration (Figure 3c). The spontaneous formation of Li₂O is thermodynamically favorable, as indicated by the negative Gibbs free energy change (∆G°<0) for the reaction between Li and CuO.^{[ 19 ]} The formation of Li₂O was confirmed by X‐ray Photoelectron Spectroscopy (XPS) (Figure S2, Supporting Information). In the Li 1s spectrum, an asymmetric peak shape slightly skewed to the higher binding energy from 54.2 eV is observed, which is attributed to the superposition of the main Li peak at 54.2 eV and the Li₂O signal appearing at 55.5 eV (Figure S2a, Supporting Information).^{[ 51} , ⁵² , ^{53 ]} In the O 1s spectrum, a peak appears at 530.5 eV corresponding to Li₂O,^{[ 51 ]} which confirms the presence of Li₂O formed during Li infiltration and the subsequent infiltration of Li facilitated by the lithiophilic Li₂O layer (Figure S2b, Supporting Information).

Figure 3.

Molten Li infiltration into 3DCu. a) Schematic for the molten Li infiltration process into 3DCu to fabricate Li in 3DCu. b) Proposed infiltration mechanism in this work. c) Optical images for 3DCu and Li in 3DCu. d) Illustration for the wetting behavior of Li for 3DCu. e) Time‐lapse optical images of 3DCu. f) XRD spectra of 3D Cu and Li in 3DCu. g) Illustration of the wetting behavior of Li for H₂‐3DCu. h) Time‐lapse optical images of 3DCu. Optical image of molten Li droplet on i) CuO and j) Cu foil. Optical images comparing the contact angles for k) CuO and l) Cu after replacing the previously dropped Li droplet with a new Li droplet. m) Time‐lapse optical images comparing the diffusion behavior of Li over time on CuO and Cu surfaces at elevated temperatures. n) Top‐view optical images at 5 min that clearly show the difference in Li diffusion on Cu or CuO.

We experimentally confirmed that CuO was critical for Li infiltration. As shown in Figure 3d,e, 3DCu rapidly underwent wetting almost immediately upon contact with molten Li. As shown in Figure 3f, the X‐ray diffraction (XRD) analysis of the Li‐infiltrated 3DCu (Li in 3DCu) revealed that the CuO peak observed in pristine 3DCu disappeared after infiltration, whereas a weak Li₂O peak emerged. This suggests that a reduction of CuO at elevated temperatures occurs as

$$CuO+2Li\to Cu+L{i}_{2}O$$ (1)

where the change in the Gibbs free energy is negative, thereby implying a thermodynamically favorable reaction.^{[ 19 ]}

The role of CuO was further confirmed by hydrogen‐treated H₂‐3DCu, where CuO was removed (Figure S3 and S4, Supporting Information). Li infiltration was completely suppressed, highlighting the necessity of CuO for enabling Li penetration (Figure 3g,h).

Molten Li can diffuse laterally on CuO surfaces, forming a lithiophilic Li₂O layer that promotes capillary‐driven infiltration. For demonstration, CuO and Cu foils were prepared by air oxidation and acid treatment,^{[ 28 ]} respectively (Figure S5, Supporting Information). As shown in Figure S6a,b (Supporting Information), the contact angle of Li on the CuO foil changed over time; however, as shown in Figure S6c,d (Supporting Information), the contact angle of Li on the Cu foil remained high without any noticeable change. This contrast indicates that a chemical reaction occurs at the interface between CuO and Li, forming a more lithiophilic surface, which corresponds with the changes in the XRD spectra shown in Figure 3f. Notably, the XRD pattern of the Li‐infiltrated 3DCu shows characteristic peaks of crystalline BCC Li, particularly the (211) and (220) planes at ≈66.0° and 73.5°, respectively. The presence of these peaks suggests that Li maintains a crystalline structure during infiltration, indicating a well‐defined interaction between Li and the CuO surface. Li spreads along the CuO surface, in contrast to its behavior on Cu, where no lateral diffusion occurs, as shown in Figure 3i,j. When the initial Li droplet is removed and replaced with a new droplet as shown in Figure 3k,l, a significantly low contact angle on CuO is observed. However, Li fails to spread on Cu and maintain a high contact angle.

The diffusion behavior of Li on the CuO surface was investigated using an additional wetting experiment, as indicated in Figure 3m. Li was placed on the CuO and Cu foil on a hot plate contracted because of the lithiophobicity of Cu. In contrast, the Li on CuO gradually spread along the surface. These changes are more clearly illustrated in the top‐view image shown in Figure 3n, which highlights how the presence of CuO enhances the lithiophilicity and determines the success of molten Li infiltration.

Among the forces influencing fluid flow inside our 3DCu, capillary force was determined to be the dominant factor, outweighing the gravitational, viscous, and inertial forces. The capillary pressure is determined using the Young–Laplace equation

$$\mathrm{\Delta }P=\frac{2\gamma \cos \theta }{r}$$ (2)

where ΔP, γ, θ, and r represent the capillary force [Pa], surface tension [n m⁻¹], contact angle, and radius of curvature [m], respectively. In the case of CuO, a positive capillary pressure was generated, which enabled Li infiltration because the contact angle was less than 90° (Figure 3k). However, the contact angle exceeded 90° on Cu without the oxide layer, generating a negative pressure that prevented Li infiltration (Figure 3l).

An analysis based on dimensionless numbers was conducted for quantifying the relative effects of various forces.^{[ 54} , ⁵⁵ , ⁵⁶ , ⁵⁷ , ⁵⁸ , ^{59 ]} (see Table S2 (Supporting Information) for the values of parameters in the details). The capillary numbers (Ca) can be determined using

$$Ca=\frac{\mu v}{\gamma }$$ (3)

where μ and v represent the viscosity of the fluid [Pa·s] and flow speed [m s⁻¹], respectively. This indicated that viscous forces are negligible compared to surface tension, which has a value on the order of 10⁻⁵. Similarly, the Bond number (Bo) is

$$Bo=\frac{\mathrm{\Delta }\rho g{L}^2}{\gamma }$$ (4)

where Δρ, g, and L represent the difference in density [kg m⁻³], gravitational acceleration [9.81 m s⁻²], and characteristic length [m], respectively, which indicates that gravitational forces are also insignificant, with a value of ≈5 × 10⁻⁵. Finally, the Reynolds number (Re) is expressed as

$$Re=\frac{\rho vL}{\mu }$$ (5)

where L represents the characteristic length, which reveals that inertial forces can be compared to are weaker than the viscous forces, with a value of 0.68. These findings demonstrate that the capillary force overwhelmingly dominates, ensuring rapid and unobstructed Li infiltration without reaching equilibrium in the middle of the infiltration.

2.3. Electrochemical Characterization of Li in 3DCu

Li in 3DCu exhibits several key advantages because of its ultrathin and lightweight 3D Cu framework; for example, extremely low thickness, high volumetric capacity, enhanced electrical conductivity from the Cu distributed throughout the Li, and improved stability under high current densities facilitated by the increased effective surface area, and reduced current density attributed to partially exposed Cu nanorods. (Figure 4a) As shown in Figure 4b and S7 (Supporting Information), slightly exposed Cu nanorods can be seen on the surface of Li in 3DCu, which enable stable charge transport across the Cu/Li composite anode. In addition, the effective surface area of the anode can be increased, which significantly reduces the current density, thereby contributing to the suppression of Li dendrite formation.^{[ 4} , ^{32 ]} Figure 4c shows that the thickness of Li in 3DCu remains below 30 µm, which is far below the Li‐infiltrated Thick 3DCu (Li in Thick 3DCu) shown in Figure S8 (Supporting Information), highlighting its potential to far enhance the volumetric energy density of LMBs. Moreover, the infiltrated Li exhibits a moderate areal capacity of ∼5 mA h cm ², as shown in Figure 4d. This value is significantly lower than that in the previously reported studies,^{[ 29} , ³³ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴⁵ , ^{46 ]} where frameworks with a thickness exceeding 100 µm exhibited an excessive Li infiltration of over 20 mA h cm⁻². This low areal capacity suggests that Li utilization can be maximized using the thin Cu/Li composite anode. Considering the high density (8.96 g cm⁻³) of Cu compared to that of Li (0.534 g cm⁻³), lightweighting the Cu framework is essential for achieving the high energy density. The large amount of Cu in Li in thick 3D Cu deteriorated the specific energy, thereby resulting in a lower specific energy than that of Li in 3DCu, as shown in Figure S9 (Supporting Information). Unlike previous studies that compensated for the excessive areal mass of Cu frameworks by overloading Li,^{[ 15} , ²⁸ , ²⁹ , ³⁴ , ^{36 ]} Li in 3DCu benefits from the ultralight areal mass of the 3DCu framework (8.53 mg cm⁻²). This structure achieves an impressive specific capacity of 514 mA h g⁻¹, which far exceeds that of graphite anodes, with only a minimal Li infiltration of 5 mA h cm⁻². Furthermore, it shows a higher energy density than that of Li‐Cu anodes formed by depositing 5 mA h cm⁻² of Li onto a 2D Cu foil, which proves its practical potential for practical applications, as confirmed in Figure S10 (Supporting Information).

Figure 4.

Structural and electrochemical characterization of Li in 3DCu. a) Schematic of Li in 3DCu with top and side views. b) Side view and c) cross‐sectional SEM image of Li in 3DCu. d) Voltage profile of Li in 3DCu for Li stripping to evaluate areal and specific capacities. EIS spectra of bare Li and Li in 3DCu e) as‐assembled and f) after 20 cycles. g) Voltage profiles of Li (5 mA h cm⁻²) on 2D Cu (9 µm foil) and Li in 3DCu at various current densities. h) Voltage profiles of Li on 2DCu and Li in 3DCu during cycling with an areal capacity of 1.0 mA h cm⁻² at 1.0 mA cm⁻².

The presence of Cu within and on the surface of the LMA significantly improves charge transport through the LMA and charge transfer at the interface. Electrochemical impedance spectroscopy (EIS) revealed a substantial reduction in the resistance of 3DCu compared with that of bare Li, as shown in Figure 4e. When Li‐electrodeposited Cu was used as control, the electrodeposition process inevitably resulted in the formation of an SEI layer, which made it difficult to isolate its effects. Therefore, a 100‐µm‐thick Li chip (Bare Li) was used as a control to eliminate any effects from the pre‐formed SEI layer. According to curve fitting with the equivalent circuit, the serial resistance, solid electrolyte interphase resistance, and charge‐transfer resistance decreased by 24, 60, and 66%, respectively, as shown in Figure S11 and Table S3 (Supporting Information). After 20 cycles at a current density of 1 mA cm⁻² and a capacity of 1 mA h cm⁻², the Li in 3DCu still outperformed the Li on 2DCu. The three types of resistances of Li in 3DCu were 25, 86, and 51% lower than those of Li in 2DCu, respectively, as shown in Figure 4f. These results suggest that the superior characteristics of Li in 3DCu are retained even after prolonged cycling.

Electrochemical impedance analysis reveals that excessive formation of lithiophilic Li₂O can deteriorate the electrochemical performance of the Cu/Li composite anode. To investigate this effect, we prepared an oxidized 3D Cu framework (O‐3DCu) by introducing an additional air oxidation step to the as‐prepared 3D Cu, thereby increasing the surface CuO content. Following molten Li infiltration, the resulting composite anode (Li in O‐3DCu) exhibited significantly higher interfacial resistance than Li in 3DCu, as shown in Figure S12 (Supporting Information). This increase is attributed to the accumulation of Li₂O, which has low electrical conductivity. These results highlight the importance of precisely controlling the oxide content: while a moderate amount of surface CuO facilitates Li wetting and infiltration, excessive CuO resulting in excessive Li₂O negatively impacts charge transport. Importantly, we demonstrate that even a small amount of CuO formed naturally during electrodeposition and ambient handling is sufficient to enable uniform molten Li infiltration while maintaining low interfacial resistance, without the need for additional oxidation processes.

The reduced current density and low resistance afforded by the 3D framework improved the reversibility of Li plating and stripping, while enabling stable cycling even at high current densities, as shown in Figure 4g. In the symmetric Li@Cu||Li@Cu cell, Li on 2DCu exhibited high overpotentials, even at relatively low current densities. At a current density of 8.0 mA cm⁻², severe voltage hysteresis occurred because of the depletion of active Li and the degradation of the electrode. In contrast, Li in 3DCu maintained a stable cycling performance with an overpotential of less than 12 mV, even at 8.0 mA cm⁻². Figure 4h shows the voltage profiles during cycling at a constant capacity and current density. For Li on 2DCu, an initial overpotential of ≈12 mV was observed, which gradually increased to over 1 V at 220 h, implying the depletion of active Li. In contrast, the Li in 3DCu exhibited a remarkably low overpotential of less than 30 mV over 690 h.

During Li deposition and stripping, the 3D Cu framework acts as a stable host for Li by uniformly distributing the current density across its structure, effectively suppressing dendritic growth during cycling via the enhanced surface area and facilitating smooth Li deposition within its interconnected pores (Figure 5a). To confirm this functionality, we observed the morphology of Li after multiple deposition and stripping cycles. We prepared the samples using bare Li as the control instead of electrodeposited Li, thereby avoiding the effect of the pre‐formed SEI layer and grain‐like Li morphology formed during the electrodeposition process. Unlike bare Li (Figure 5b), Li in 3DCu exhibited exposed Cu nanorods on the surface, as indicated in Figure 5c,d, and S13 (Supporting Information). Granular Li with an approximate diameter of 5 µm forms on bare Li, as observed in Figure 5e. In contrast, Li in 3DCu displays smooth Li deposition that filled the internal structure without granular formation, as shown in Figure 5f. Figure 5g shows a reduction in the exposure of Cu nanorods after deposition compared to that of pristine Li in 3DCu, thus indicating that the internal pores of the 3D Cu framework are filled with Li. During the stripping process, bare Li removes granular deposits, as shown in Figure 5h. However, in the case of Li in 3DCu, the previously filled Li was stripped, thus exposing a significant portion of the 3DCu, as shown in Figure 5i,j. This re‐exposure of the Cu nanorods suggests that the 3DCu framework can effectively mitigate the current density, thereby ensuring stable Li‐plating behavior in subsequent cycles.

Figure 5.

Ex situ analysis of Li morphology on the 3D Cu framework. a) Schematic of the Li deposition/stripping mechanism in 3DCu. SEM image of bare Li and Li in 3DCu b–d) as‐prepared, after e‐g) deposition and h–j) following stripping step of the 1st cycle, and k‐m) deposition step of 30th cycle. Figure 5b,e,h,k correspond to bare Li. Figure 5c,d,f,g,i,j,l, and m correspond to Li in 3DCu.

The anode surface was observed after 30 cycles to evaluate the stability of the Li plating/stripping behavior during extended cycling. As shown in Figure 5, granular and mossy Li coexisted in the case of bare Li, which indicates a grain size reduction with dendritic growth. In contrast, Li in 3DCu exhibited no evidence of structural thinning. Instead, granular Li grains were distributed within the 3D Cu framework, as shown in Figure 5l. This suggests that the 3D Cu framework successfully prevented Li dendrite formation and maintained a stable Li morphology even after prolonged cycling. From a practical battery operation perspective, we further examined the Li morphology after 100 cycles to assess its long‐term stability. As shown in Figure S14a,b (Supporting Information), Bare Li exhibited extremely thin, filament‐like Li deposits, whereas Li in 3DCu maintained a granular morphology. The sustained granular Li morphology observed in 3DCu can be attributed to the combined effects of current homogenization and mechanical confinement within the 3D structure. The interconnected Cu framework disperses local current densities and suppresses dendritic nucleation, while simultaneously providing structural guidance that promotes uniform Li deposition. This behavior contrasts with Bare Li, where the absence of confinement leads to localized current density and dendritic growth. Overextended cycling, the accumulation of dead Li and interfacial instability become more pronounced in Bare Li, whereas the preserved granular morphology in Li in 3DCu reflects improved deposition reversibility and interfacial stability. which are essential for long‐life Li‐metal batteries.

Li in 3DCu demonstrated superior electrochemical performance compared to that of Li on 2DCu, as evidenced by its enhanced cycling stability, higher rate capability, and improved reversibility (Figure 6 ). Figure 6a shows that Li in 3DCu retained 80% of its initial capacity even after 100 cycles, while Li in 2D Cu maintained 80% of its initial capacity for only 23 cycles. Further, the coulombic efficiency (CE) of Li in 3DCu remained significantly high, with an average CE of 99.80% over 100 cycles, thus highlighting its stable plating and stripping behavior. In contrast, the CE of Li in 2DCu decreased significantly with cycling, with an average CE of 97.68% over 60 cycles, indicating poor reversibility and increased degradation. Voltage profiles during plating and stripping further emphasized the advantages of Li in 3DCu. As shown in Figure 6b,c, Li in 3DCu exhibited consistently lower overpotentials than that of Li in 2DCu throughout cycling. The gaps between voltage profiles in the charging/discharging step for Li in 3DCu remained narrow and stable from the first to the 20^th cycle. In contrast, Li on 2DCu exhibited gradually widening gaps with increasing overpotential.

Figure 6.

Li@Cu||LFP full‐cell characterization. a) Areal discharge capacity with CE in Li@Cu||LFP cell in 0.5C with corresponding voltage profiles of b) Li on 2DCu and Li in 3DCu. d) Areal discharge capacity during the rate‐capability test with a C‐rate range of 0.2–2C. Corresponding voltage profiles of e) Li on 2DCu and f) Li in 3DCu. The mass loading of the LFP cathode was 13.5 mg cm⁻² and the areal discharge capacity at 0.05C was ≈2.0 mA h cm⁻².

The rate performance test with various C rates of 0.2, 0.5, 1.0, and 2.0C, as shown in Figure 6d, highlights the significant advantages of Li in 3DCu at high current densities. Li in 3DCu always shows a higher discharge capacity than that of Li on 2DCu during the C‐rate capability test because of the current density reduction characteristics. Li in 3DCu maintained over 90% of its initial capacity even after five cycles at 2C, while Li in 2DCu experienced substantial capacity fading, retaining only 70% at the same rate. This improvement was attributed to the ability of 3DCu to effectively distribute current density and suppress dendritic growth, thus enabling stable performance under high‐rate conditions using a cathode with practical areal capacity. Voltage profiles under varying current densities, as shown in Figure 6e,f, provide further evidence for the superior rate capability of Li in 3DCu. Even at high current densities, Li in 3DCu exhibited low overpotentials, ensuring efficient charge transport/transfer and stable operation. In contrast, Li on 2DCu suffers from a severe overpotential, limiting its ability to perform efficiently under high‐rate conditions. Figure S15 (Supporting Information) further confirms the usefulness of Li in 3DCu with extremely high C‐rates. To avoid ambiguity in the experimental analysis caused by the instability of carbonate electrolytes at high current densities, we fabricated an LFP cathode with a low areal capacity and conducted cycling tests at a 5C rate. Under these conditions, Li in 3DCu exhibited a specific capacity of 111.2 mA h g⁻¹, which is 7.9% higher than that of Li on 2DCu (103.1 mA h g⁻¹). While Li on 2DCu retained 80% of its initial capacity after ≈70 cycles, Li in 3DCu demonstrated superior cycling stability, maintaining an impressive 91.0% capacity retention even after 180 cycles at extremely high C‐rates, confirming its effectiveness under fast charge/discharge conditions. By comparing the coulombic efficiency, the results with low cycling capacity shown in Figure S15 (Supporting Information) can also provide a clearer assessment of irreversible Li‐ion loss associated with SEI formation. The average CE over the first 10 cycles was slightly higher for Li in 2DCu (97.37%) than for Li in 3DCu (97.14%), likely due to the smaller exposed surface area. However, over 80 cycles, the trend reversed: Li in 3DCu demonstrated a higher average CE (99.47%) than Li in 2DCu (99.36%). These results indicate that a carefully controlled integration of high‐surface‐area 3D architecture does not necessarily lead to excessive SEI formation or increased irreversible capacity. On the contrary, it improves long‐term interfacial stability and suppresses dead Li accumulation, ultimately enhancing the electrochemical reversibility of composite Li‐metal anodes. These findings collectively demonstrate that the 3D Cu framework provides critical advantages for LMA. Li in 3DCu achieved enhanced cycling stability, excellent rate performance, and high reversibility by reducing the interfacial resistance, mitigating dendritic growth, and evenly distributing the current density.

2.4. Rationalizing Thickness and Li Loading for High‐Energy‐Density Composite Anodes

The thickness and Li loading of the 3D Cu framework are critical parameters determining the practical applicability and electrochemical performance of composite Li‐metal anodes. Previous studies have frequently employed relatively thick 3D frameworks with excessive Li loading, aiming primarily at improved cycling stability. However, such designs inevitably compromise both gravimetric and volumetric energy densities, limiting their commercial feasibility. In this context, Li in 3DCu which has a thickness of ≈30 µm with moderate Li loading of 5 mA h cm⁻² addresses these limitations by significantly enhancing energy density while maintaining stable electrochemical performance. The results obtained in this study clearly demonstrate that reducing the framework thickness and appropriately controlling Li loading simultaneously improves Li utilization, reduces overall anode mass, and maintains mechanical integrity and conductivity, thus providing valuable design guidance for developing next‐generation high‐energy‐density Li‐metal composite anodes.

To systematically demonstrate the impact of thickness and Li loading, comparative analyses were conducted using two samples: Li in Thick 3DCu which has a thickness of 100 µm with Li loading of 10 mA h cm⁻², and Thick Li on 2DCu (≈20 mA h cm⁻² Li deposited on a 9 µm‐thick Cu foil). (Figure S16a–c, Supporting Information) Both samples exhibited improved cycling stability, with Li in Thick 3DCu retaining 88% capacity after 210 cycles, confirming the advantage of increased Li reservoirs. Nevertheless, these enhancements were achieved at the expense of reduced energy densities. (Figure S17 and Table S4, Supporting Information) Specifically, Li in 3DCu (197.6 W h kg⁻¹ and 439.9 W h L⁻¹) exhibited 63.0% and 45.4% improved initial gravimetric and volumetric energy density in respectively compared to Thick 3DCu (121.2 W h kg⁻¹ and 302.5 W h L⁻¹), due to its thinner and lighter framework. Compared to Li on 2D Cu (193.0 W h kg⁻¹ and 420.0 W h L⁻¹), Li in Thick 3DCu exhibited a higher energy density and maintained this advantage over prolonged cycling due to its superior electrochemical reversibility. Thick Li on 2DCu (165.2 W h kg⁻¹ and 272.1 W h L⁻¹) suffered from diminished volumetric energy density caused by excessive Li loading and low Li utilization (Figure S17 and Table S4, Supporting Information). These comparative results clearly underscore that rationally balancing thickness and Li loading, as demonstrated by Li in 3DCu, is essential for achieving practical high‐energy‐density composite anodes, moving beyond traditional strategies based solely on maximizing Li content or framework thickness, as shown in Figure 1c and Table S5 (Supporting Information).^{[ 15} , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ , ^{46 ]}

The introduction of 3D framework not only significantly reduced current densities preventing the formation of Li dendrites, but also facilitate electron transport throughout the anodes as validated by our electrochemical cycling and impedance measurements. Especially, the partial re‐exposure of conductive Cu nanorods within the 3D structure significantly mitigates local current density fluctuations and prevented the localized nucleation and propagation of Li dendrites. This structural advantage effectively stabilizes Li morphology during prolonged cycling, maintaining mechanical integrity and reducing interfacial resistance as confirmed by SEM observations and EIS analyses. Furthermore, careful control of surface chemistry is essential due to the trade‐off relationship between facilitating Li infiltration and electrical conductivity. A controlled CuO concentration ensures sufficient wettability and subsequent formation of lithiophilic Li₂O at the interface, driving uniform and complete Li infiltration via capillary action. However, an excessive CuO content must be avoided, as it may induce the formation of electrically insulating phases, negatively affecting electron transport. Hence, achieving a delicate balance in CuO content is crucial to preserving electronic conductivity while enhancing infiltration capability. Collectively, the structural, chemical, and morphological design principles presented herein offer clear and practical guidelines for researchers aiming to develop advanced, high‐performance Li‐metal composite anodes suitable for next‐generation high‐energy‐density battery applications. As the thinning of the framework can weaken the mechanical stability and reduce the long‐term cycling stability of composite anode, further research is needed to enhance the mechanical robustness of ultrathin 3D Cu frameworks while preserving their electrochemical advantages.

3. Conclusion

This study demonstrated the successful fabrication and characterization of a novel ultrathin 3D Cu/Li composite anode (Li in 3DCu) that addressed the limitations of Li‐metal anodes (LMA). The composite LMA of less than 30 µm and a moderate Li loading of 5 mA h cm⁻² with the impressive specific capacity of 514 mA h g⁻¹ was achieved by utilizing a lightweight, highly porous 3D Cu framework. These metrics were achieved without compromising the cycling stability, with over 80% of the initial capacity retained after 104 cycles with a practical 2.05 mA h cm⁻² LFP cathode. This highlighted its ability to deliver the high energy density while leveraging the advantages of the 3D framework, including effective current‐density reduction across the electrode and suppression of dendritic growth, which could ensure stable performance under high current densities. This work also provided new insights into the molten Li infiltration mechanism, moving beyond the traditional reliance on Gibbs free energy and lithiophilicity transitions. The comprehensive understanding of the molten Li infiltration process was achieved via an analysis of the interplay between capillary, viscous, gravitational, and inertial forces, along with interfacial chemical changes, clarify how these forces and the surface chemistry collectively drive infiltration into a 3D Cu framework, offering a framework for optimizing future anode designs for favorable infiltration.

4. Experimental Section

Fabrication of 3DCu

3DCu was prepared by the electrochemical etching of a Cu wire mesh followed by the electrodeposition of Cu nanorods.^{[ 4 ]} A Twill weave Cu wiremesh (Hebei Aegis Metal Materials, China) with a wire diameter of 50 µm and an open width of 30 µm was immersed in a 1 M sulfuric acid solution for 5 min, followed by rinsing with de‐ionized water. A twill‐woven wiremesh is used because it is structurally robust and resistant to wrinkles compared with a plane‐woven wiremesh, and the rate of electrochemical etching and electrodeposition can be intense on the surface protrusion by wrinkles, which easily occurs in a plane‐woven wiremesh. The Cu wiremesh was electrochemically etched in 0.2 M sulfuric acids with 0.05 M copper sulfate aqueous solution using a 25‐µm‐thick Cu foil (MTI Korea) as a counter electrode. After 21 min of etching with a current density of 40 mA cm⁻², the Cu nanorod is deposited on the mesh with a voltage of 2.8 V for 5 min after changing the electrolyte and counter electrode into new ones. Electrodeposition for less than 5 min makes the infiltration of molten Li impossible because of the insufficient filling of voids in the wire mesh. After electrodeposition, 3DCu were dried in a tube furnace (Samheung Energy, SH‐FU‐120TG) at 60 °C under vacuum and stored in an Ar‐filled glove box to avoid undesired oxidation. For the fabrication of thick 3DCu, electrochemical etching was skipped, and the nanorods were electrodeposited for 10 min.

Fabrication of Li in 3DCu

A crucible made of stainless steel 316 was heated up on a hot plate set to 350°C. Li infiltration was conducted after a contactless infrared thermometer was used to confirm that the temperature of the crucible was stabilized to 200°C. A Li chip with a diameter of 10 mm was cut in half, the surface oxide layer was removed using a nylon brush, and the chip was then placed in a crucible for melting. Subsequently, 3DCu with a diameter of 14 mm was brought into contact with molten Li to allow Li to infiltrate the 3D Cu structure. Li in 3DCu was immediately removed from the crucible upon completion of the infiltration process to prevent the corrosion of Cu due to prolonged contact with Li at high temperatures.

Other Fabrication Methods

H₂‐3DCu were fabricated by thermally reducing 3DCu in a tube furnace (Samheung Energy, SH‐FU‐120TG) under 400 °C for 1 h, with a ramping rate of 8 °C min. H₂ gas with 3.9% concentration (Ar base) was used at a flow rate of 5 SLPM under vacuum as a reductant. Two types of Cu foils were prepared to compare the affinities of the Cu and CuO foils for molten Li. The Cu foil is prepared by removing the surface oxide of a 25‐µm‐thick Cu foil (MTI Korea) in a 1 M sulfuric acid solution for 5 min, followed by rinsing with de‐ionized water. CuO films were formed on Cu foil using a previously reported method.^{[ 28 ]} The Cu foil after acid treatment were heated in a tube furnace (Samheung Energy, SH‐FU‐120TG) at 300 °C for 12 h, with a ramping rate of 8 °C min. O‐3DCu was prepared by placing 3DCu on a hot plate heated to 300°C for 5 min.

Electrochemical Cell Assembly

Various anodes were characterized using CR2032 coin cells (MTI, USA) assembled using a hydraulic crimper (MSK‐110, MTI, USA) with a crimping pressure of 7 MPa. For the Li@Cu||Li@Cu symmetric cell, 50 µl of 1 M lithium bistri(fluoromethanesulfonyl)imide in dioxolane (DOL)/dimethyl ether (1:1 vol. %) (Dongwha Electrolyte, South Korea) with 3 wt.% LiNO₃ (Sigma Aldrich) and a Celgard 2400 separator (Celgard, USA) were used. For preparing Li on 2DCu and Thick Li on 2DCu, 5 and 21 mA h cm⁻² of Li was electrodeposited on a 9‐µm‐thick Cu foil (MTI Korea) with a current density of 0.5 mA cm⁻², followed by a formation cycle in 50 µA cm⁻² with a voltage window of 0–1 V. For the Li@Cu||LFP cell, 50 µl of 1 M lithium hexafluorophosphate in ethyl carbonate/diethyl carbonate (1:1 vol. %) with 10 wt.% fluoroethylene carbonate and 1 wt.% vinylene carbonate and LFP‐coated Al foil (MTI Korea) were used as the electrolyte and cathode, respectively. The areal capacity and coating area density of LFP cathode were 13.5 mg cm⁻² and 2.0 mA h cm⁻², respectively.

Electrochemical Characterization

EIS was conducted on an electrochemical workstation (Ivium‐n‐stat, Ivium Technology, Netherlands), with a frequency range of 10⁻³–10⁵ Hz at an amplitude of 50 mV. The parameters of the components constituting the equivalent circuit were derived using the IviumSoft software. The areal capacities, cycle lives, and rate capabilities of the Li@Cu||Li@Cu and Li@Cu||LFP cells were characterized using a battery cycler (WBCS3200L; WonA Tech, South Korea). The Li@Cu||LFP cell was activated for two cycles with a C‐rate of 0.1 C, and then, it was cycled with a C‐rate of 0.5 C in a voltage window of 2.5–4.2 V.

Other Characterization Methods

Surface observations were performed using an SEM (SU5000, Hitachi, Japan) at an acceleration voltage of 10–15 kV. Elemental distribution was mapped using EDS (Octane Elite 25, EDAX). Phase analysis was performed using XRD (Smartlab, RIGAKU, Japan). To analyze the chemical bondings of various samples were analyzed by XPS (Nexsa G2, Thermo VG Scientific, USA). For contact angle analysis with molten Li droplets, the oxide layer of the molten Li droplets was removed according to a previous report.^{[ 60 ]} The oxide layer on the surface of the 3 cm × 5 cm Li foil was removed using a nylon brush. The Li foil was then melted on a Ti plate (Sigma Aldrich) at 350 °C. The oxide layer on Li was removed using a stainless‐steel spatula and spoon, and the molten Li was shaped into a spherical droplet. Then, the Li droplet was placed on Cu or CuO foil and heated on a hot plate. The masses of the fabricated samples were measured using an ultraprecision microbalance (XA 52.4Y, RADWAG, Poland), and their thicknesses were measured using a thickness meter (547‐401A, Mitutoyo, Japan) and SEM.

Conflict of Interest

The authors declare no conflict of interest.

Author Contributions

S. K. and I. Y. contributed to conceptualization. I. Yang designed the fabrication and characterization methods and conducted sample preparation, electrochemical characterization, and SEM/EDS analysis. H. Baek and D. G. Kim contributed to XRD/XPS analysis and sample preparation, respectively. S. Kim contributed to the funding and resource acquisition. I. Yang wrote and S. Kim edited the manuscript. All authors approved the final version of the manuscript.

Supporting information

Supporting Information

Yang I., Baek H., Kim D. G., Kim S., Ultrathin 3D Cu/Li Composite with Enhanced Li Utilization for High Energy Density Li‐Metal Battery Anodes. Small 2025, 21, 2501629. 10.1002/smll.202501629

Data Availability Statement

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