Enabling the Multiscale Interfacial Stabilization for Ah-level Anode-free Lithium Metal Battery via a Prelithiation Separator strategy

doi:10.21203/rs.3.rs-5032177/v1

Download PDF

Article

Enabling the Multiscale Interfacial Stabilization for Ah-level Anode-free Lithium Metal Battery via a Prelithiation Separator strategy

https://doi.org/10.21203/rs.3.rs-5032177/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Featuring the straightforward assembly of fully lithiated cathodes with bared current collectors, anode-free lithium metal batteries (AFLMBs) nominally achieve the highest gravimetric/volumetric energy densities with minimum Li host in excess, simplified anode processing, as well as the reduced labor/cost of cell manufacturing and maintenance. Nevertheless, issues of parasitic interfacial reactions, high-voltage cathode collapse and the irreversible Li⁺ plating on the deposition substrate, collectively deplete the cation reservoir of cell models. This study thus proposes a separator strategy to enable the multiscale interfacial stabilization for Ah-level AFLMB model. Specifically, the sacrificial Li₂S@C prelithiation layer loaded on the polyolefin separator (Li₂S@C|PE), not only supplements the customized Li⁺ inventory during the formation cycle, but also establishes the lithium polysulfides containing cathode interface with the high-voltage tolerance (till 4.5V). Through the combined analysis of in-situ electrochemical impedance spectroscopy and transmission-mode operando X-ray diffraction, the enhanced Li⁺ diffusivity and reversible phase evolution of LiNi_0.8Co_0.1Mn_0.1O₂ cathode as in contact with the prelithiation separator are real-time documented. Upon the cell assembly of Li₂S@C|PE separator with the Ag modified Cu foil (Ag-Cu) and densely-packed LiNi_0.8Co_0.1Mn_0.1O₂ cathode (25.0 mg cm^− 2) under lean electrolyte condition (E/C 1.8 g Ah^− 1), the 1.22 Ah pouch-format prototype balances the robust cycling endurance, gravimetric/volumetric energy densities of 450 Wh kg^− 1/1355 Wh L^− 1, as well as extreme power output up to 830.6 W kg^− 1. This prelithiation protocol demonstrates upscaling potential and generic applicability to secure the interfacial chemistries for anode-less/-free LMB configurations.

Physical sciences/Chemistry/Electrochemistry/Batteries

Physical sciences/Materials science/Materials for energy and catalysis/Batteries

Anode-free lithium metal battery

Prelithiation strategy

Interfacial stabilization

High-voltage tolerance

Operando X-ray diffraction

High energy/power density

Owing to the unrivaled specific capacity (3860 mAh g^− 1) and lowest reduction potential (-3.04 V vs. the standard hydrogen electrode), metallic lithium is regarded as the optimal anode choice that surpasses commercial intercalation-type graphite anodes¹. Nonetheless, the reliable operation of lithium metal batteries (LMBs) is plagued by parasitic interfacial reactions across the scale, cathode degradation upon high voltage charge as well as uncontrolled metallic dendrite growth². A plethora of previous studies utilized the thick-layer Li foil of > 100 µm to counterbalance the irreversible Li consumption, which adversely compromises the Li⁺ utilization degree (typically less than 30%) and nominal energy/power densities at the cell level³. Technological challenges also pose severe safety concerns regarding Li foil production, electrode processing, as well as cell assembly procedures⁴. Aiming to maximize the practical volumetric/gravimetric energy densities, therefore, the LMBs prototyping urgently necessitates the viable solution to enhance cation utilization degree while adopting thin-layer/lightweight metallic host without excessive Li dosage.

Anode-free lithium metal batteries (AFLMBs), emerging as the ultimate lithium-constrained configurations, enable the direct Li deposition on the current collectors (CC). Additionally, the cell manufacturing eliminates the anode slurry casting procedures or demanding requirement of the dew point^{5, 6}. However, the robust cyclability of AFLMBs suffers from irreversible side reactions at multiple interfaces. In the typical model as illustrated in Scheme 1a, Li⁺ ions extracted from the cathode lattice migrate through the separator and deposit on the bared Cu foil. However, the parasitic irreversible reactions between the exposed reactive Li and organic solvents, along with the incomplete Li stripping, would collectively deplete the active Li⁺ inventory from the cathode lattice (Scheme 1a, process i). Upon the high-voltage charging, on the other hand, the cathode structure would gradually collapse, the dissolved transition metal (TM) cations from which thus cross-over along the voltage gradient and rupture the fragile solid electrolyte interphase (SEI) on the CC (Scheme 1a, process ii). The uncontrolled metallic deposits further exacerbate the irreversible Li consumption (Scheme 1a, process iii and 1c). Addressing aforementioned issues requires the design ingenuity of the deposition substrate to regulate the interfacial electrochemistry, which is also the key enabler to reconciling the trade-off between energy density and cyclability on the cell level. Various electrolyte formulations (LiFSI-1.2DME-3TTE⁷ and 0.6 M LiDFOB-0.6 M LiBF₄ in DEC-2FEC⁸), and artificial ceramic (LiPON⁹, Li_4.4Sn¹⁰, SiO_x¹¹) or polymeric (PVDF¹², PDMS¹³, PAN¹⁴) layer protection have been attempted to passivate the surface reactivity of Li deposits However, loose packing of mosaic SEI layers or physical interfacial coating tends to retard the ionic diffusivity, meanwhile the limited Li reservoir within the cell would be inevitably consumed by the lithiophilic sites or nucleophilic functional groups on the substrates¹⁵. Most of these substrate modifications, unfortunately, are still based on empirical attempts, which lack systematic analysis to elucidate the real-time structural interplay between the dynamic cathode evolution and Li deposition behavior. Therefore, it is imperative to separately regulate the interfacial electrochemistry across the multiple scale, namely the large-areal-capacity, reversible Li plating/stripping process on the deposition substrate, the high-voltage tolerance of the cathode side, and more critically, the inhibition of TM “cross-talk” phenomenon in the AFLMB configuration.

By affording the extra Li⁺ source for energy-dense cell models, the prelithiation strategies play a crucial role in alleviating the irreversible capacity fading¹⁶. Various types of Li_xMO_y (M = Fe, Ni, Co) agents, such as Li₅FeO₂¹⁷, and Li₂NiO₂¹⁸ powders were preset in the cathode processing to compensate the Li depletion during cycling. Nevertheless, these Li-rich oxides exhibit strong alkalinity and are prone to moisture contamination, resulting in severe gelation of the electrode slurry, meanwhile the decomposed oxide by-products may block electron/ion migration channels. Furthermore, Li₂O¹⁹ and Li₃N²⁰ additives could release nucleophilic O^− 2 or N^− 3 to degrade the electrolyte solvent, and generate gas species upon the de-lithiation process. All these effects adversely lead to the electrode cracking and cell swelling. On the other hand, over-lithiated intercalation oxides (for instance Li₂Ni_0.8Co_0.1Mn_0.1O₂²¹ and Li₂Ni_0.5Mn_1.5O₄²²) could afford the Li⁺ inventory without introducing gas or inactive products, however, their widespread applications are restricted by limited de-lithiation capacities. So far, the comprehensive assessment of the prelithiation agents on the Li utilization, especially at the Ah-level AFLMB model, remains unexplored, let alone the precise control of Li inventory supplementation tailored to the specific cell prototyping. More urgently, the facile prelithiation for the AFLMB construction requires a feasible protocol that is compatible with commercial roll-to-roll manufacturing, avoiding tedious electrode processing or detrimental contamination during the cell assembly.

To tackle the issues above, we herein proposed a prelithiation separator strategy to enable the simultaneous stabilization of cathode interfacial chemistry, homogenized Li plating/stripping process as well as the precise cation supplementary for pouch-format AFLMB models. Specifically, the lithium containing sacrificial agents were encapsulated within the amorphous carbon (Li₂S@C) via a facial carbothermal reduction (2C + Li₂SO₄ = 2CO₂ + Li₂S). Subsequently, the Li₂S@C layer at the optimized thickness was cast onto the polyolefin separator (Li₂S@C|PE) through the microgroove method, as illustrated in Scheme 1b. Besides the prelithiation amount up to 0.85 mAh cm^− 2, the Li₂S@C|PE also generated the lithium polysulfides (LiPS) species at the cathode/electrolyte interface, mitigating the TM dissolution upon the high-voltage charge till 4.5V. (Scheme 1b, Process I) At the anode side, the delithiated Li⁺ from the composite separator pre-alloyed with the thermally deposited Ag on the Cu foil, suppressing the Li nucleation overpotential (Scheme 1d). Conceptually, the prelithiation separator strategy operates independently of electroactive material pretreatment or electrode processing, which allows the facile cell prototyping that is compatible with streamlined battery manufacturing. The 1.22 Ah pouch-format AFLMB model was thus constructed by integrating the Li₂S@C|PE separator with densely packed NCM811 cathode (25 mg cm^− 2, double-sided) and the Ag-modified Cu foil, which exhibits the gravimetric/volumetric energy densities of 450 Wh kg^− 1/ 1355 Wh L^− 1, power output up to 830.6 W kg^− 1 as well as robust cyclability even under lean electrolyte condition (1.8 g Ah^− 1). The combined in-situ electrochemical impedance spectroscopy (EIS) analysis and transmission-mode operando X-ray diffraction validated the accelerated ion diffusivity and reversible phase evolution of the NMC811 cathode as paired with the Li₂S@C|PE separator. The prelithiation strategy present in this study affords a straightforward, generic protocol to stabilize interfacial chemistries for both the high-voltage cathode and deposition substrate, inspiring the paradigm shift towards the highly reversible, anode-free/less cell construction.

The preparation procedures of Li₂S@C prelithiation agent were schematically illustrated in Fig. 1a. Specifically, the precursors of Li₂SO₄, polyvinylpyrrolidone (PVP), and acetylene black with a weight ratio of 4:1:1 were homogeneously mixed in the ethanol solution (solid content of 6%). After the solvent evaporation, a carbothermal reduction of the dried mixed power (depicted by the equation of 2C + Li₂SO₄ = 2CO₂ + Li₂S) was implemented under argon gas atmosphere (Supplementary Fig S1). The X-ray diffraction (XRD) pattern of the as-obtained product in Fig. 1b demonstrates the isostructural cubic phase of Li₂S (PDF#23–0369). As revealed in scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images (Fig. 1c and Supplementary Fig. 2), the Li₂S@C composite exhibits the interconnected porous network of the nanocrystalline. Accordingly, the lattice fringes with 0.33 nm d-spacing correspond to the (111) plane of Li₂S, as depicted in Fig. 1d. The high-angle annular dark field (HAADF) image and elemental mapping exhibit the uniform encapsulation of sulfur species inside the thermally reduced carbon matrix (Fig. 1e). This structural feature thus enhances the electron conduction during the electrochemical-driven de-lithiation process, which in turn boosts the lithium utilization from the Li₂S@C sacrificial agent. Through the precise regulation of the roll-to-roll gravure coating, it is possible to control the thickness of the slurry casting layer on the PE substrate. Correspondingly, the delithiated capacities of the Li₂S@C|PE separators of varied thickness were compared in the Al||Li half cells, with the Li₂S@C-coated side faced to the Al foil (details in experimental section). As illustrated in Supplementary Fig. 3, the Al||Li cell paired with the Li₂S@C|PE separator (10 µm functional layer) exhibits an initial charge capacity of 1056 mAh g^− 1 at 0.1 C, which reveals 91% Li⁺ utilization of the Li₂S. With the continual increase of functional layer thickness from 5 µm to 10, 15, and 20 µm (Fig. 1f), the Li₂S@C|PP separators (designated as Li₂S@C|PE-5, Li₂S@C|PE-10, Li₂S@C|PE-15 and Li₂S@C|PE-20) exhibit the gradual increase of areal prelithiation capacities from 0.24 to 0.53, 0.74 and 0.85 mAh cm^− 2, respectively. Noting the Li₂S@C|PP-10 achieved the highest Li⁺ utilization degree of 91% and optimized Gurley value of 227 s. The dimensional specifications and intrinsic physical properties of these separators were itemized in Table S1. Apparently, although the functional layers in Li₂S@C|PE-15 and Li₂S@C|PP-20 enhanced the theoretical prelithiation capacities, the excessive thickness adversely compromised the Gurley value (245 s and 260 s) as well as the Li⁺ utilization efficiency (89.5% and 86.2%). Upon immersion in the carbonate-based electrolyte, the Li₂S@C|PE-5 and Li₂S@C|PE-10 separators demonstrated enhanced electrolyte uptake percentages of 160% and 200%, respectively (Fig. 2g) in contrast to 127% for the pristine PE separator. This increased electrolyte absorption capability would benefit the ion diffusivity and reduce the cell's direct current internal resistance (DCIR). Meanwhile, the Li₂S@C|PE-10 exhibited the highest ionic conductivity of 1.29 mS cm^− 1 among the separators. Considering the harmony balance of the prelithiation capacity, Li⁺ utilization efficiency, and ion diffusivity, the Li₂S@C|PE-10 was selected as the optimal separator, labeled as the Li₂S@C|PE for the following tests. The optical image in Fig. 1h demonstrates a roll of Li₂S@C|PE separator manufactured via the Roll-to-Roll gravure coating procedures (dimension of 60mm*100m), highlighting the upscaling potential for practical manufacturing. As shown in the cross-sectional SEM image (Fig. 1i), the thickness of the prelithiation layer is roughly estimated as ~ 10 µm on the PE substrate. Furthermore, the enhanced wettability and ionic percolation capability of the Li₂S@C|PE-10, as compared to PE, were shown in Fig. 1j. The Li₂S@C functional layer drastically reduced the contact angle (16.81°) with the carbonate electrolyte droplet on the substrate. As shown in Supplementary Fig. 4, the Li₂S@C|PE demonstrated the enhanced transference number of 0.56 in the symmetric cell, which doubled the value as compared to that of the pristine PE (t⁺=0.33) in the Li||Li model, suggesting the facile cation permeability upon the Li₂S@C functionalization.

To highlight the high-voltage tolerance and interfacial regulation effect of the prelithiation separator, the Li₂S@C|PE was paired with the commercial NCM811 (2.0 mAh cm^− 2) and LiCoO₂ cathodes (1.8 mAh cm^− 2), respectively, with reference to the Li foil. Impressively, the cycling behavior of the NCN811 that paired with the Li₂S@C|PE, maintaining 94.9% capacity retention (187.3 mAh g^− 1) after 200 cycles within the voltage range of 3.0-4.3 V at 0.5 C (Fig. 2a). In sharp contrast, the NCM811||Li model with the PE separator only preserved 80.8% capacity retention for 200 cycles. As displayed in Fig. 2b and Supplementary Fig. 5, the model with Li₂S@C|PE also exhibits rather stable voltage profiles and much higher energy efficiency. In sharp contrast, the NCM811||Li cell with PE displayed a rapid voltage fading due to the exacerbated electrolyte decomposition upon the high-voltage charge. Moreover, the LiCoO₂||Li models were assembled to evaluate even higher voltage stability within 3.0-4.5V. The cell with Li₂S@C|PE separator achieved 98.1% capacity retention (182.5 mAh g^− 1) without observable voltage polarization after 200 cycles (Fig. 2c-d). In comparison, the cell model with PE only exhibited 95.6% capacity retention with 155.1 mAh g^− 1 maintained (Fig. 2d and Supplementary Fig. 6).

X-ray photoelectron spectroscopy (XPS) measurements were conducted to elucidate the influence of the prelithiation process on the high-voltage tolerance of the cycled NCM811 cathode. As shown in the core-level F 1s spectra (Fig. 2e), the deconvoluted sub-peaks at 684.7 and 687.2 eV could be assignable to the Li_xPF_y and LiF species²³, respectively. As compared to the PE, obviously, the interfacial species of the cycled NCM as in contact with the Li₂S@C|PE separator exhibits a pronounced peak intensity of the thermal stable LiF species in the CEI layer. Moreover, the characteristic S 2p signals assignable to S-S and ROS₃Li species at the binding energies of 164.3 and 169.5 eV (Fig. 2f) were observed on the cathode as in contact with the Li₂S@C|PE^{24, 25}. The S-containing species could be generally denoted as lithium polysulfides (LiPS). As further validated with morphological characterization (Fig. 2g), the post-cycled cathode with the Li₂S@C|PE exhibits a conformal CEI layer with the sulfur elemental distribution at the interface (Supplementary Fig. 7), which could potentially insulate the electrolyte decomposition and TM dissolution from the NCM811. However, the non-uniform CEI layer, without the LiPS species, mainly consists of the organic-rich species derived from the solvent oxidation and contains less LiF species (Supplementary Fig. 8), these features barely prevented the NCM811 degradation at high voltage²⁶. Moreover. the interfacial morphology of the post-cycled Li foil (Supplementary Fig. 9–10) that disassembled from the PE paired model exhibits the mossy structure. On the other hand, the Li foil from the Li₂S@C|PE paired cell model was relatively smoother with less dendrite protrusion, suggesting the mitigation of TM “cross-talk” phenomenon. Figure 2h vividly illustrates the construction of LiPS containing layer on the NCM cathode. Upon the high voltage charge, the Li₂S sacrificial agent could afford Li⁺ source upon the delithiation process, meanwhile, the sulfur species participate to form a thin and uniform CEI. Thanks to the inherent high-plasticity and electron-insulating properties, the Li-S-O and S-S species not only suppressed the TM dissolution and electrolyte decomposition, but also promoted the facile Li⁺ diffusion with a mitigated energy barrier. Therefore, the irreversible phasic evolution of the layered oxides was significantly restrained during high-voltage charge.

The delithiation kinetics and SEI formation evolution is investigated by the galvanostatic intermittent titration technique (GITT) and in-situ electrochemical impedance spectroscopy (EIS) during the discharge/charge process^{27, 28}. As demonstrated in Fig. 3a-b, the thermodynamic potential of the electrochemical reactions was recorded with every period of 10 min charge/discharge and 20 min rest. The Li⁺ diffusion coefficient (D_Li⁺) was also calculated to investigate the electrochemical kinetics of the cathodes (Supplementary Fig. 11). At the beginning of the charging process, the Li⁺ supplementary rapidly boosted the ion diffusion kinetics. The results indicate that the Li₂S@C|PE significantly enhances Li⁺ diffusion at the NCM811 interface. More comprehensively, the in-situ EIS measurements in Fig. 3c-d reveal the in-depth mechanism of LiPS species in optimizing the stepwise charge transfer process. As reflected in typical EIS plots, three frequencies (150 kHz, 100 Hz, and 10 Hz) were selected to show the evolution of resistance with different cells. The high-frequency region (R1) is attributed to electrolyte resistance. The semi-circle frequency region (R2) could be ascribed to the interfacial resistance. The tail appearing at low frequency corresponds to a Warburg region assignable to the bulk Li⁺ diffusion within the electrode. All the impedance plots can be well-fitted by the equivalent circuit in Supplementary Fig. 12a. With the deeper SOC, the typical tail of the curve gradually vanishes, replaced by a depressed semicircle corresponding to the charge-transfer resistance (R3). During the charge and discharge period, the R1 was kept the constant, R2 and R3 decreased simultaneously upon charging and then increased synchronously upon discharging. Moreover, distribution of relaxation times (DRT) analysis was implemented to decouple the processes that governed the EIS spectra^{29, 30}. As shown in Fig. 3e-f, the DRT plots of the EIS after the discharging process shows six peaks within a timescale (τ) range of 10^− 6-10^− 1 s, denoted as τ1–τ6. To reveal the interface chemistry with Li₂S@C|PE separator, we assigned these DRT peaks to specific processes and tracked their evolution during the first cycle. Specifically, contact resistance (τ1), ion transport through the CEI layer (τ2–τ3), charge transfer (τ4–τ5), and mass transport (τ6) were assigned, respectively. As shown in Fig. 3E and S13, charge transfer resistance at the interface (τ4–τ5) dominates upon whole process, while the value of τ5 increases sharply after the complete cycle. This indicates high charge resistance between NCM811 and electrolyte, severely hindering the ion diffusion in the cathode. In comparison, the NCM811||Li with Li₂S@C|PE separator preserved relatively lower charge resistance (Fig. 3f and Supplementary Fig. 14). Noting that the difference in the τ4 peak is due to the varied Li deficiency in the structural lattice of NCM811, which leads to different Li diffusivities. Meanwhile, the τ2 and τ3 peaks could be assigned to the ionic transport at the interfacial CEI. Notably, the peak of the model with Li₂S@C|PE separator hardly increased after the initial CEI film formation, whereas the cell with PE separator showed the pronounced charge transfer resistance. As illustrated in Supplementary Fig. 12b, the interfacial resistance (R2, 15.4 Ω) of NCM811||Li paired with Li₂S@C|PE were maintained much lower as compared to that with PE (47.6 Ω), mainly owning to the interfacial stabilization effect after the introduction of Li₂S@C layer. In addition, the charge-transfer R3 values for the Li₂S@C|PE equipped cells (38.8 Ω) are smaller than the counterpart with PE (118 Ω). These results suggest the kinetics activation effect through pairing the Li₂S@C|PP separator, which favors the Li⁺ migration across the multiple interphases in the NCM811||Li model.

As a proof-of-concept to evaluate the prelithiation effect, the models were assembled by pairing the NCM811 cathode, bared Cu foil with either Li₂S@C|PE or PE separators. For the AFLMB architecture, there exists no cation reservoir at the anode part, which thus allows for the evaluation of Li⁺ compensation effect provided by the Li₂S@C agent for the cell. Meanwhile, the interfacial incompatibility between carbonate electrolyte and Li deposits on the Cu foil would aggravate the electrolyte decomposition and irreversible Li⁺ depletion on the substrate. As compared in Fig. 4a-c, the NCM811||Cu model with Li₂S@C|PE delivered an initial reversible capacity of 194.7 mAh g^− 1 and moderate capacity retention of ~ 60% for 50 cycles. In sharp contrast, the NCM811||Cu with PE cell experienced rather severe capacity fading after only 10 cycles. Accordingly, the average Li inventory retention rates (LIRR) were calculated as 99.1% and 85.0% for the NCM811||Cu models with Li₂S@C|PE or with the PE, respectively, which could be utilized as reasonable metrics to assess the “average” Li inventory loss per cycle. Therefore, the Li extracted from Li₂S@C|PE indeed offset the irreversible Li inventory and alleviated the capacity fading. Meanwhile, the transmission-mode operando XRD was employed to probe the multi-phase evolution, more importantly, the dynamic interplay between the supplement cation inventory and the cathode³¹. As revealed in the contour plot (Fig. 4d-i), the (003), (101), (104), (105), and (107) diffraction planes of the NCM811 cathode were clearly observed. The hexagonal (003) plane at 8.63° represents the c-axis lattice breathing of the layered oxide lattice. During the initial charge process, accompanied by the gradual disappearance of the Li₂S peak, the characteristic (003) peak of NCM811 begins to shift towards the lower 2θ angles. This process could be assigned to the phasic evolution from the H1 to H2 and corresponding lattice expansion. Upon the subsequent charging, the peak quickly shifts to the higher angles, indicating the subsequent transition to H3 phase³². To quantitatively assess the Li⁺ reversibility of the cathode framework, the peak shift of (101) planes was employed to calculate the variation of the a-axis. According to lattice parameters shown in Fig. 4f, the c value is similar to the curve trend of the (003) planes as the Li⁺ source extracted from the NCM811 cathode. The variations of a and c lattice parameters of the NCM811 that paired with PE were estimated as 1.9% and 2%, which indicates lithium depletion leads to the Li deficiency of the oxide lattice. In comparison, the cell model with the Li₂S@C|PE exhibits only 0.3% and 0.17% variations of a and c lattice after a complete cycle (Fig. 4i). These results demonstrate that the supplementary lithium source could alleviate the lattice distortion of the NCM. Meanwhile, the diffraction peaks that ascribed to the Li₂S gradually evolved into the sulfur precipitation upon the initial charge process (Fig. 4h), which echoed with the S-S containing species observed on the cathode (Fig. 2f).

The reversible Li plating/stripping behavior on the bared substrates hinges on the initial nucleation process, which overwhelmingly affects cycling endurance and operational safety of AFLMB models. Owning to the intrinsic lattice mismatch with metallic lithium, the copper foil exhibits lithiophobicity and thus the high nucleation overpotential³³. Therefore, a modification layer of 100 nm Ag nanoparticles (Ag NPs) was purposely deposited onto the Cu substrate, denoted as Ag-Cu. The top-view SEM images of the pristine Cu foil at various magnifications revealed a smooth and flat surface, in contrast to the grey-color, rough surface of Ag-Cu foil with interfacial particles (Fig. 5a-b). The XRD pattern confirms the presence of cubic metallic Ag with the diffraction peaks situated at 17.2° (Supplementary Fig. 15). Notably, the coating thickness did not vary significantly as compared to 6 µm Cu foil (Supplementary Fig. 16). Energy-dispersive X-ray spectroscopy (EDS) cross-sectional mapping further verified the compact, uniform coverage of Ag NPs film on the Cu foil. The Ag 3d spectra can be deconvoluted into dual subpeaks located at 367.9 eV and 373.9 eV (Ag 3d_5/2 and Ag 3d_3/2), suggesting the presence of zero-valent Ag (Fig. 5c). Supplementary Fig. 16 demonstrates the lightweight (0.1 mg cm^− 2) layer with nano-thickness. This coating layer thus has not enhanced the weight/volume of the Cu substrate too much. The EIS measurements, combined with the equivalent circuit simulation (Supplementary Fig. 17), suggest the Li||Ag-Cu model demonstrated a much lower interfacial resistance (22.5 Ω) as compared to the Li||Cu counterpart (35.3 Ω). Based on the analysis of the Li-Ag phase diagram (Supplementary Fig. 18), the formation of a series of Li-Ag alloy intermediates (LiAg, Li₉Ag₄, and Li₁₀Ag₃ phases) could reduce the nucleation barrier upon the pronounced Li solubility ^{34, 35}. Consequently, the Ag-Cu anode demonstrates a higher initial Coulombic efficiency (ICE) of 92.9% for the initial plating/stripping process, in contrast to 84.5% for the Cu substrate at 1 mAh cm^− 2 (Supplementary Fig. 19). Corresponding to the plating curve of the Ag-Cu substrate (Fig. 5d), the transmission-mode XRD characterization tracked the stepwise lithiation alloying process in operando, validating the gradual formation of LiAg with the Li⁺ solubility amount up to 0.04 mA cm^− 2, Li₉Ag₄ with 0.062 mA cm^− 2 and Li₁₀Ag₃ with 0.097 mA cm^− 2 Li contents, respectively. Based on the Aurbach CE calculation methods³⁶, the average CE of Li plating/stripping on Ag-Cu could reach the value of 99.2% from the second cycle onwards, significantly higher than the 97.9% on the Cu foil (Fig. 5e).

As scrutinized in the post-cycled SEM images (Supplementary Fig. 20–21), the Ag-Cu substrate retained less “dead Li” residue than that on the Cu foil, indicating higher Li utilization efficiency. In-situ optical microscopy was employed to visualize the Li deposition process on both substrates. Figure 5f exhibits whisker-shaped lithium deposits that appeared on the Cu anode after 10 mins of plating, which further protruded into the dendrites after 60 mins. In sharp contrast, the Ag-Cu substrate (Fig. 5g) maintains a smooth surface without obvious dendritic morphology, highlighting the beneficial Li-Ag alloy layer to favor the homogeneous nucleation. The schematic diagrams in Fig. 5h-i illustrate the lithium deposition process on both Cu and Ag-Cu substrates. During the initial charge, the lithiophobic Cu foil with a higher nucleation barrier tends to induce the localized charge transfer process at the nucleates, which easily accumulates into the sharp spike. However, the ultrathin Li-Ag intermediate alloy enhances lithium affinity and high-throughput Li⁺ influx, enabling the homogeneous Li electrodeposition. Quantitative numerical estimations of the accessible lithium source were further explored. The irreversible Li trapping in the intermediate alloy formation only consume 0.097 mAh cm^− 2 during the first cycle, which could be sufficiently counterbalanced by the prelithiation reservoir (Li₂S@C, 0.5 mAh cm^− 2) from the separator. Therefore, this cell prototyping could avoid the cathode inventory depletion during the initial formation cycle, guaranteeing the retrievable capacity of the cell.

To evaluate the performance metrics of practical relevance, various NCM811||Cu, NCM811||Ag-Cu models paired with Li₂S@C|PE separator, in the formats of the coin cells and pouch cells, were assembled under the lean electrolyte conditions (LiFSI-1.2DME-3TTE, 10 µl mAh^− 1). So far, seldom studies could balance the robust cycling endurance and the extreme energy densities for the AFLMB models at the Ah-level prototyping, in terms of the Wh kg^− 1 or Wh L^− 1 values. As shown in Fig. 6a, the NCM811||Cu cell with PE separator continuously decayed and only maintained the 40.2% capacity retention for 100 cycles, in comparison, the NCM811||Ag-Cu model with PE separator preserved 72.0% capacity retention. Noting that the NCM811||Ag-Cu cell with Li₂S@C|PE separator exhibits the superior cycling endurance (87.4% capacity retention for 100 cycles), which further validates the synergistic coupling of the Li₂S@C|PE separator and substrate modification. Moreover, the NCM811||Ag-Cu model with the Li₂S@C|PE exhibits an apparent alloying plateau, corresponding to the activation of Li-Ag alloy formation (Fig. 6b). The following discharge/charge cycles maintained cycling endurance without any overpotential fluctuation or voltage decay in the curves (Supplementary Fig. 22). To assess the upscaling potential of the prelithiation strategy, the 1.22 Ah, layered-staked AFLMB model was assembled by pairing the NCM811 cathode (double sided, 25 mg cm^− 2 for each side), and Li₂S@C|PE separator, as well as the Ag-Cu substrate (Fig. 6c). The technological specifications of the pouch cell were itemized in Fig. 6d. The NCM811||Ag-Cu model with the Li₂S@C|PE exhibits long-term cycling endurance and maintains 85% capacity retention after 100 cycles (Fig. 6e); by contrast, the NCM811||Cu model paired with PE suffered from severe capacity fading upon the cycling (only 32% capacity retention for 100 cycles). This comparison validates the effective Li⁺ supplementary from prelithiation separator strategy. Furthermore, superior rate capabilities of the NCM811||Ag-Cu cell with Li₂S@C|PE were demonstrated, with reversible capacities of 205.2, 190.2, 172.3, 139.6 and 113.7 mAh g^− 1 obtained at 0.1, 0.5, 1, 2, and 3C, respectively (Fig. 6f). Figure 6g demonstrates the Ragone plot of the as-assembled pouch cell (based on both the total mass of the whole device. The NCM811||Ag-Cu model with the Li₂S@C|PE exhibits a practical gravimetric/volumetric energy densities of 450 Wh kg^− 1/1355 Wh L^− 1 at the cell level, far surpassing the previously reported anode-less/free based battery configurations from the literature. If the calculation was based on the total mass of the electroactive materials, a superior energy density of 779 Wh kg^− 1 could be obtained. Besides, the extreme power output of 830.6 W kg^− 1 could be achieved for the cell model (Table S2). Owning to the multiscale interfacial therapy from the Li₂S@C|PE separator with the additional Li⁺ inventory, the AFLMB prototypes overcome Li⁺ consumption origins and boost the cation transfer kinetics, especially as applications move toward energy/power-dense regimes.

In this study, we developed a prelithiation separator strategy to regulate the multiscale interfacial electrochemistry of the Ah-level AFLMB model. Specifically, a carbon thermally reduced Li₂S@C functional layer was modified onto the PE substrate. Upon delithiation of the Li₂S@C composite during the initial charging, the composite separator could afford the on-demand cation source up to 0.85 mAh cm^− 2, compensating for irreversible Li depletion on the deposition substrate (Ag-Cu). The as-formed lithium polysulfide (LiPS) species at the cathode/electrolyte interface also guaranteed the high-voltage tolerance for the NCM and LCO cathodes (up to 4.5 V). The in-situ EIS measurements and operando XRD were combined to reveal the enhanced ion migration and reversible phasic evolution of NCM-811 cathode (c = 0.17%) as in contact with the prelithiation separator. Upon the layer-stacked assembly of the Li₂S@C|PE separator with the Ag-Cu substrate and densely-packed NCM811 cathode in the 1.22 Ah pouch cell, the AFLMB model balanced the gravimetric/volumetric energy densities of 450 Wh kg^− 1/1355 Wh L^− 1, power output up to 830.6 W kg^− 1 on the device level as well as 85% capacity retention for 100 cycles. This separator strategy avoids the modifications of slurry-casting procedures and streamlines with the standard roll-to-roll cell assembly, showcasing the viability for energy-dense, anode-free/less cell constructions.

Materials: Li₂SO₄ and Polyvinyl pyrrolidone (PVP) were provided by Aladdin Co., Ltd. The LiNi_0.8Co_0.1Mn_0.1O₂ cathode, LiCoO₂ cathode, acetylene black, and PVDF binder were purchased from Guangdong Canrd New Energy Technology Co., Ltd. The Li foil was obtained from Energy Lithium Co., Ltd. The carbonated electrolyte, consisting of 1.0 M lithium hexafluorophosphate (LiPF₆) in ethylene carbonate, diethyl carbonate, and dimethyl carbonate (EC/DEC/DMC, volume ratio 1:1:1) was purchased directly from Suzhou Duoduo Chemical Technology Co., Ltd. The ether-based LHCE electrolyte composed of lithium bis(fluorosulfonyl)imede (LiFSI) in mixed solvent was prepared and stored in a glove box under an Ar atmosphere. The mixed solvent was 1,2-dimethoxyethane (DME) and 1,1,2,2-tetrafluoroethyl-2,2,3,3-tetrafluoropropyl ether (TTE) (molar ratio 1.2:3).

Preparation of Li ₂ S@C composite materials. The Li₂S@C agents were synthesized according to classic carbon thermal reactions. Firstly, 1 g of PVP, 4 g of Li₂SO₄⋅H₂O and 1 g of Acetylene Black were added into 100 ml anhydrous ethanol and ultrasonicated for 30 min. Then the obtained mixture was stirred for two days at room temperature with a magnetic stirrer until ethanol was completely evaporated. The dried sample was further sintered in a tube under an argon atmosphere at 850°C for 2 h with heating to obtain Li₂S@C samples.

Prelithiation of Li ₂ S@C|PE prelithiation separator. To prepare the Li₂S@C-based separator (Li₂S@C|PE), Li₂S@C (80%), Acetylene Black (10%), and PVDF (10%) were stirred in NMP solvents. The Polyolefin separator was homogeneously coated with the obtained slurry using a scraper. Then, the separator was dried at 80 ℃ in the glove box.

Preparation of the Ag-Cu anode. For fabrication of Ag-deposited Cu foil anode (Ag-Cu), Ag was deposited onto Cu foil with a thickness of 100 nm (SHIN HYUNG E&T CORP) using a thermal evaporator (WOOSUNG HI-VAC Co.) housed within and N₂-filled glove box. The Ag granules were placed inside a Ta boat within the evaporator. All thermal evaporation depositions were conducted under a vacuum of 10^− 6 torr at an evaporation rate of ≈0.5 Å s^− 1. A thickness monitor was used to measure the thickness of the evaporated material, with the thickness adjusted to the actual thickness obtained through ellipsometry analysis using an ellipsometer (J. A. Woollam Co. Ltd).

Electrochemical measurements. Composite cathodes were fabricated by compressing active material, carbon black (super p), and polyvinylidene fluoride (PVDF) at a weight ratio of 8:1:1. Subsequently, the as-obtained cathode slurry was directly cast on a carbon-coated aluminum and vacuum drying at 80 ℃ for 12 hours. CR2016 coin cells were assembled in the Ar-filled glovebox (< 0.1 ppm of H₂O and O₂), using Li foil as the anode. The electrolyte used was carbonated electrolyte. For coin-type cells, the galvanostatic charge/discharge measurements were tested on a Neware battery Test System (CT-4008T-5V10mA-164, Shenzhen, China) at room temperature. For anode-free cells, the NCM811||Cu with PE separator model was fabricated by replacing the Li foil anode with a bare Cu current collector. NCM811||Ag-Cu with PE separator model pouch cells were assembled using NCM811 (55 mm*36 mm) as the cathodes and Ag-Cu foil (58*39 mm²) as anodes, Li₂S@C|PE as the separator, the LCHE (LiFSI-1.2DME-3TTE) as electrolyte (1.8 g Ah^− 1) and Al-plastic film as packing.

For the Aurbach CE tests on plated Li, a standard protocol was followed: 1) perform one initial formation cycle with Li deposition of 5 mAh cm^− 2 on Cu under 0.5 mA cm^− 2 current density and stripping to 2 V; 2) plating 5 mAh cm^− 2 Li on Cu or Ag-Cu anode under 0.5 mAh cm^− 2 as a Li reservoir: 3) repeatedly strip/plate Li of mAh cm^− 2 under 0.5 mAh cm^− 2 for 10 cycles; 4) strip all Li to 2 V. The average CE was calculated by dividing the total stripping capacity by the total plating capacity. Electrochemical impedance spectroscopy (EIS) was carried out on Gamry over a frequency range of 1.0*10⁶ Hz and 1.0*10^− 1 Hz.

Materials Characterization

All post-cycled electrodes were extracted from disassembled cells in an Ar-filled glovebox (< 0.1 ppm of H₂O and O₂). The morphology and structure of all materials were characterized by field-emission scanning electron microscopy (SEM) (FEI, NovaTM Nano SEM450) with energy-dispersive X-ray spectroscopy (EDX) operating at 18 kV. TEM images were acquired on an FEI Talos F200x transmission electron microscope at 200 kV. The surface chemistry on the electrode was captured with X-ray photoelectron spectroscopy (XPS, Kratos Axis Supra instrument). The crystal structure of materials and dynamic phase transition in cells upon the charging/discharging process were measured by a transmission-mode X-ray diffractometer (STADIP STOE) with a position-sensitive detector and Mo Kα₁ radiation with a wave (λ) of 0.70930 Å, operating at 50 kV and 40 mA. The contact angle was measured at an optical contact angle (SL200KB, Kino, USA), and a 2.0 ml droplet of the electrolyte was in the experiment. The visualization of lithium dendrite growth process was observed by optical microscopy (YUESCOPE YM710R).

Acknowledgements

The authors appreciate the financial support from the National Natural Science Foundation of China (52173229 and 52373229), the Natural Science Foundation of Shaanxi (2019KJXX-099 and 2023-JC-JQ-15), the Fundamental Research Funds for the Central Universities (3102019JC005 and D5000230114) and the Key Research and Development Projects of Shaanxi Province (No. 2019ZDLGY04-05). Furthermore, we would like to thank the Analytical & Testing Center of Northwestern Polytechnical University.

Li S et al (2023) Dynamic elucidation of li solubility in the alloy and metallic plating process. Adv Mater 35:2306826A Multiscale
Jie Y et al (2024) Progress and perspectives on the development of pouch-type lithium metal batteries. Angew Chem Int Ed 63:e202307802
Betz J, Bieker G, Meister P, Placke T, Winter M, Schmuch R (2018) Theoretical versus practical energy: a plea for more transparency in the energy calculation of different rechargeable battery systems. Adv Energy Mater 9:1803170
Zhao P et al (2022) Constructing self-adapting electrostatic interface on lithium metal anode for stable 400 Wh kg^– 1 pouch cells. Adv Energy Mater 12:2200568
Shao A, Tang X, Zhang M, Bai M, Ma Y, Challenges (2022) Strategies, and prospects of the anode-free lithium metal batteries. Adv Energy Sustain Res 3:2100197
Langdon J, Manthiram A (2021) Crossover effects in batteries with high-nickel cathodes and lithium-metal anodes. Adv Funct Mater 31:2010267
Ren X et al (2019) Enabling High-voltage lithium-metal batteries under practical conditions. Joule 3:1662–1676
Louli AJ et al (2020) Diagnosing and correcting anode-free cell failure via electrolyte and morphological analysis. Nat Energy 5:693–702
Sun J et al (2023) Robust Transport: An artificial solid electrolyte interphase design for anode-free lithium‐metal batteries. Adv Mater 35:2209404
Zhang Z, Luo H, Liu Z, Wang S, Zhou X, Liu Z (2022) A chemical lithiation induced Li_4.4Sn lithiophilic layer for anode-free lithium metal batteries. J Mater Chem A 10:9670–9679
Chen W, Salvatierra RV, Ren M, Chen J, Stanford MG, Tour JM (2020) Laser-induced silicon oxide for anode-free lithium metal batteries. Adv Mater 32:2002850
Luo J, Fang C-C, Wu N-L (2018) High Polarity poly(vinylidene difluoride) thin coating for dendrite-free and high-performance lithium metal anodes. Adv Energy Mater 8:1701482
Lopez J, Pei A, Oh JY, Wang G-JN, Cui Y, Bao Z (2018) Effects of polymer coatings on electrodeposited lithium metal. J Am Chem Soc 140:11735–11744
Sahalie NA et al (2020) Multifunctional Properties of Al₂O₃/polyacrylonitrile composite coating on cu to suppress dendritic growth in anode-free li-metal battery. ACS Appl Energy Mater 3:7666–7679
Wang YX et al (2023) Stable Anode-free all-solid-state lithium battery through tuned metal wetting on the copper current collector. Adv Mater 35:2206762
Zhan R, Wang X, Chen Z, Seh ZW, Wang L, Sun Y (2021) Promises and challenges of the practical implementation of prelithiation in lithium-ion batteries. Adv Energy Mater 11:2101565
Meng Q et al (2023) A functional prelithiation separator promises sustainable high-energy lithium‐ion batteries. Adv Energy Mater 13:2300507
Wu Y et al (2023) Li₂Cu_0.1Ni_0.9O₂ with copper substitution: a new cathode prelithiation additive for lithium-ion batteries. ACS Sustain Chem Eng 11:1044–1053
Qiao Y, Yang HJ, Chang Z, Deng H, Li X, Zhou HS (2021) A high-energy-density and long-life initial-anode-free lithium battery enabled by a Li₂O sacrificial agent. Nat Energy 6:653–662
Park K, Yu B-C, Goodenough JB (2016) Li₃N as a cathode additive for high-energy-density lithium-ion batteries. Adv Energy Mater 6:1502534
Lin L et al (2021) Li-Rich Li₂[Ni_0.8Co_0.1Mn_0.1]O₂ for anode-free lithium metal batteries. Angew Chem Int Ed 60:8289–8296
Lin L et al (2021) Spinel-related Li₂Ni_0.5Mn_1.5O₄ cathode for 5-V anode-free lithium metal batteries. Energy Storage Mater
Wang HS et al (2022) Liquid electrolyte: The nexus of practical lithium metal batteries. Joule 6:588–616
Zhan Y, Yu H, Ben L, Chen Y, Huang X (2017) Using Li₂S to compensate for the loss of active lithium in li-ion batteries. Electrochim Acta 255:212–219
Tian MY et al (2022) Designer cathode additive for stable interphases on high-energy anodes. J Am Chem Soc 144:15100–15110
Yuan X et al (2023) Bi-affinity electrolyte optimizing high‐voltage lithium‐rich manganese oxide battery via interface modulation strategy. Angew Chem Int Ed 62:e202304121
Shao B et al (2022) Enabling conversion-type iron fluoride cathode by halide‐based solid electrolyte. Adv Funct Mater 32:2206845
Lu Y, Zhao C-Z, Huang J-Q, Zhang Q (2022) The timescale identification decoupling complicated kinetic processes in lithium batteries. Joule 6:1172–1198
Wan TH, Saccoccio M, Chen C, Ciucci F (2015) Influence of the discretization methods on the distribution of relaxation times deconvolution: implementing radial basis functions with DRTtools. Electrochim Acta 184:483–499
Lu Y et al (2021) The carrier transition from Li atoms to Li vacancies in solid-state lithium alloy anodes. Sci Adv 7:eabi5520
Xu C et al (2021) Bulk fatigue induced by surface reconstruction in layered Ni-rich cathodes for Li-ion batteries. Nat Mater 20:84–92
Ryu H-H, Park K-J, Yoon CS, Sun Y-K (2018) Capacity fading of Ni-Rich Li[Ni_xCo_yMn_1–x–y]O₂ (0.6 ≤ x ≤ 0.95) cathodes for high-energy-density lithium-ion batteries: bulk or surface degradation? Chem Mater 30:1155–1163
Sandoval SE et al (2023) Structural and electrochemical evolution of alloy interfacial layers in anode-free solid-state batteries. Joule
Zhang M et al (2022) Boosting the temperature adaptability of lithium metal batteries via a moisture/acid-purified, ion-diffusion accelerated separator. Adv Energy Mater 12:2201390
Pyo S et al (2022) Lithiophilic wetting agent inducing interfacial fluorination for long-lifespan anode‐free lithium metal batteries. Adv Energy Mater 13:2203573
Lin L, Suo L, Hu Y-s, Li H, Huang X, Chen L (2021) Epitaxial induced plating current-collector lasting lifespan of anode-free lithium metal battery. Adv Energy Mater 11:2003709

Scheme 1 is available in the Supplementary Files section.

There is NO Competing Interest.

Scheme1.jpeg
Scheme 1. Schematic illustration of Li supplement process of the prelithiation layer. a The interface failure mechanism of AFLMBs. b The synergistic engineering of the Li₂S@C|PE prelithiation separator and Ag-Cu substrate for the AFLMB models. Li plating behavior and SEI formation on c the bare Cu and d Ag-Cu substrates, respectively.
SupplementaryInformation.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Enabling the Multiscale Interfacial Stabilization for Ah-level Anode-free Lithium Metal Battery via a Prelithiation Separator strategy

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Methods

Materials Characterization

Declarations

Acknowledgements

References

Scheme 1

Additional Declarations

Supplementary Files

Status:

Version 1