Research Article www.acsami.org
Cite This: ACS Appl. Mater. Interfaces 2018, 10, 20387−20395
Highly Reversible Li Plating Confined in Three-Dimensional Interconnected Microchannels toward High-Rate and Stable Metallic Lithium Anodes Wei Deng,†,‡ Wenhua Zhu,† Xufeng Zhou,*,† Xiaoqiang Peng,† and Zhaoping Liu*,† †
Downloaded via UNIV OF TOLEDO on June 21, 2018 at 14:29:06 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Key Laboratory of Graphene Technologies and Applications of Zhejiang Province and Advanced Li-Ion Battery Engineering Laboratory, Ningbo Institute of Materials Technology & Engineering, Chinese Academy of Sciences, Zhejiang 315201, P.R. China ‡ College of Materials Science and Optoelectronic Technology, University of Chinese Academy of Sciences (UCAS), Beijing 100049, P.R. China S Supporting Information *
ABSTRACT: Practical application of metallic Li anode in Liion batteries has been restricted because of dendrite growth of Li which induces poor stability and safety issues. Despite various hosts for Li having been developed to address these issues, it is still a challenge to achieve highly reversible and stable stripping/plating behavior of Li, especially at high-rate conditions. Herein, we propose a simple method of incorporating Li in commercial carbon fiber cloth (CFC) to realize high-rate and stable metallic Li anodes by confining stripping/plating of Li in microchannels of ZnO-decorated CFC (CFC/ZnO) and dissipating high current densities through conductive carbon fiber networks. The symmetrical cell using this novel anode can run stably for over 1800 h (900 cycles) under 1 mA cm−2 and even 320 h (800 cycles) at 5 mA cm−2, which has rarely been achieved previously through structural evolution of Li-metal anode. When it is paired with commercial activated carbon, the as-made Li-ion capacitor coin cell can deliver high-rate capability (up to 30 A g−1) and long-term cycling stability for over 5000 cycles at 10 A g−1, and a large pouch cell can operate as ultrafast charge (∼1 min) battery with high-energy density of ∼50 W h kg−1. KEYWORDS: lithium-metal anode, carbon fiber cloth, interconnected channels, high rate, long-term stability
■
INTRODUCTION Tremendous requirements of high-energy density rechargeable Li-ion batteries draw worldwide research interests in advanced anode materials.1−4 For many perspectives, the most attractive anode for a rechargeable battery is metallic lithium (Li) owing to its lightweight, highest theoretical specific capacity, and lowest negative electrochemical potential.5−8 However, its high chemical reactivity and characteristic dendrite growth phenomenon have paused practical application.9,10 Especially, high local current density will significantly accelerate the formation and growth velocity of Li dendrites, which causes steep drop of cycle life and severe safety hazards, thus restricting utilization of Li-metal anode at high-rate conditions.11,12 Consequently, worldwide researchers are focusing on how to solve safety issues and prolong life span of Li-metal anode at high current densities for realizing ultrafast charging/discharging and high power density without sacrificing energy density. Previous research studies mainly concentrated on the design of tough SEI films by using functional electrolytes or forming artificial layers on the surface of Li metal to slow down the growth of Li dendrites.6,13−16 Despite inspiring breakthroughs having been made for past decades, nearly all efforts have been applied to bulk Li-metal anodes with two-dimensional planar © 2018 American Chemical Society
morphology, which still inevitably induced battery failure because only surface protection is not effective in accommodating volume change of bulk Li and dissipating high current densities, and poor mass transfer of bulk Li is also not solved. In fact, according to Chazalviel’s model, high current density applied on electrodes results in a near zero ion concentration at the negative electrode, which easily triggers the formation of Li dendrites.17 Given the factor of ion concentration gradient, creating more contact interfaces between Li and electrolytes is extremely important to alleviate ion concentration issues by producing more uniform ion flux. Thus, recently a new strategy has been developed to realize this, which employs porous hosts to accommodate volume expansion of Li metal. Hosts such as graphene-based materials,18,19 porous Cu,20 nickel foam,21,22 and nanofiber matrix23,24 also can dissipate high current densities to inhibit nucleation of Li dendrites, and facilitate diffusion of electrolytes. However, hosts simply possessing porous morphology and conductivity is insufficient to comply Received: February 12, 2018 Accepted: May 29, 2018 Published: May 29, 2018 20387
DOI: 10.1021/acsami.8b02619 ACS Appl. Mater. Interfaces 2018, 10, 20387−20395
Research Article
ACS Applied Materials & Interfaces
energy densities of 65 and 50 W h kg−1 at the ultrafast charge/ discharge time of 3 and 1 min, respectively.
with requirements at high-rate conditions. More elaborate structural design of the hosts is needed. For instance, metal (Cu, Ni, Ni−Fe, etc.) foams or meshes recently have been used to host Li metal as anode materials,20,21,25 but metallic Li inside has relatively large size of hundreds of micrometers because of the large pore sizes of such metal hosts. Thus, the restriction effect of the host for guiding the plating behavior of Li metal is limited, and the irreversible deposition of Li metal inside such metal hosts inevitably generates formation of “dead Li” on the surface of the electrode. Thus, it is better that the hosts can not only distribute Li in small scales but also restrict Li plating in a local area. On the other side, abundant pathways for electrolyte infiltration are also required to create more electrolyte/Li interfaces, which can prolong Sand’s time, and is beneficial for stable and homogeneous Li plating.17 Despite that some reported scaffolds could achieve fine distribution of Li metal, they usually lack a highly interconnected porous microstructure for continuous electrolyte infiltration. In most cases, the infiltration of electrolytes inside the electrode is suppressed to a large extent during charge/discharge processes, as the Li metal fully blocks the original pore structure of the hosts. Thus, it is important to choose appropriate hosts to realize fine distribution of Li for spatial restriction of Li stripping/plating and maximum exposure of electrolyte/Li interfaces. By comparing various possible host materials for Li metal, we found that commercial carbon fiber cloth (CFC), which has been widely used in energy conversion or storage devices, is a candidate with the most potential because of its unique structure.23,26,27 The characteristic three-dimensional (3D) interconnected channels with spatial confinement of ∼5 μm resulting from two-step weave manner in CFC, plus its high conductivity, making CFC suitable for hosting Li metal according to aforementioned requirements for appropriate hosts. More specifically, dozens of carbon fibers twinning together will help to minimize the size of Li and restrict the plating area of Li deeply inside CFC, and can also provide a steady stream of electrons for highly reversible and high-rate Li stripping/plating behavior. Meanwhile, the abundant channels among carbon fibers can provide 3D interconnected pathways for electrolyte infiltration. In addition, the macrovoids among vertically weaved carbon fiber bundles are able to facilitate electrolyte infiltration into an inner network to create more electrolyte/Li interfaces. On the basis of the advantages of CFC, we prepared highperformance CFC/ZnO-hosted Li-metal anode (CFC/Li) by simply infusing molten Li (@300 °C, inert atmosphere) into CFC/ZnO without any further structure modification. The combination of CFC/ZnO and Li metal effectively lowers voltage hysteresis of Li symmetrical cells and enhances the rate (0.5−30.0 A g−1) capability of lithium-ion capacitors (LICs) cells (paired with commercial activated carbon, AC). Since Li plating behavior was guided and restricted by 3D interconnected channels, the Li/SEI interphase was firmly confined without extrusion from the CFC/ZnO host, and ample supply of electrons by carbon fibers together gave rise to highly reversible stripping/plating of Li, rather than growth of “dead Li”. As a result, Li symmetrical cells could cycle for over 1800 h at 1.0 mA cm−2, and were extremely stable at a high current density of 5.0 mA cm−2 for 320 h (800 cycles) in unmodified ether-based electrolyte. Moreover, LIC pouch cells assembled by large-area CFC/Li anodes and AC cathodes could deliver
■
RESULTS AND DISCUSSIONS CFC (Figure S1a, Supporting Information) was modified with ZnO by dipping CFC in zinc acetate solution and annealing in an inert atmosphere to improve its lithiophilicity.28,29 Figure 1a
Figure 1. (a, b) Schematic illustration of the structure of CFC/ZnO (upper right in panel a is the top view of the CFC and lower right is the magnified illustration showing ZnO nanoparticles on individual carbon fibers), and the preparation and morphology of CFC/Li. (c−e) SEM images with different magnifications showing the morphology of CFC decorated with ZnO nanoparticles. (f−h) SEM images with different magnifications of CFC/Li electrode.
illustrates the typical inner structure of CFC/ZnO; the top view shows the macroporous structure of CFC/ZnO and the magnified image demonstrates the nano ZnO particles decorated on CFC skeleton. It consists of vertically weaved carbon fiber bundles, and the corresponding SEM image is shown in Figure 1c. Each bundle is composed of numerous carbon fibers twinning together with abundant channels as shown in Figure 1d. Surface modification of carbon fiber with ZnO is confirmed in Figure 1e. The XRD pattern of CFC/ZnO is shown in Figure S2. The main diffraction peaks confirm the transformation of zinc acetate into ZnO after annealing. Meanwhile, the high-magnification SEM images (Figure S3) demonstrate that the ZnO nanoparticles with uniform sizes of 100−200 nm are distributed evenly on each carbon fiber to enhance the lithiophilicity of CFC. The digital photographs in Figure S4 show that pure CFC without ZnO decoration has no ability for the infusion of molten Li metal. In contrast, the CFC/ZnO scaffold can enable fast incorporation of Li metal since ZnO NPs can react with molten Li to form LiZn, which enhances the lithiophilicity of CFC. After Li (∼12.0 mg for a CFC disc with diameter of 11 mm) was incorporated with CFC/ZnO (Figure S1b), channels in each carbon fiber bundle were filled with liquid metallic Li which then solidified to be Li metal immobilized by carbon fibers, as illustrated in Figure 1b,f. It should be noted that CFC/ZnO has not been fully filled with Li. Macrovoids between weaved bundles still persevered after incorporation of Li (Figure 1g), which facilitate infiltration of electrolytes into CFC/ZnO and lay the foundation for generating more electrolyte infiltration pathways during initial stripping of Li. High-magnification SEM image (Figure 1h) confirms that Li metal covers on single carbon fibers and 20388
DOI: 10.1021/acsami.8b02619 ACS Appl. Mater. Interfaces 2018, 10, 20387−20395
Research Article
ACS Applied Materials & Interfaces embeds in microchannels between carbon fibers. Compared to conventional porous hosts for Li metal, such as metal foams, microchannels with much smaller width in CFC/ZnO not only confine the size and uniformity of Li metal but also serve as restriction area to guide plating behavior of Li. To demonstrate the functions of CFC/ZnO scaffold, nonporous CFC/Li electrode (∼21.0 mg of Li in a CFC disc with diameter of 11 mm) in which all the pores are fully filled with Li by elongating infusion time has been fabricated as the control sample (named n-CFC/Li). Macrovoids between bundles have been thoroughly filled by Li metal in n-CFC/Li, and the cross-sectional SEM image shows that extra Li-metal layer has covered the entire surface of CFC/ZnO (Figure S5). Stripping/plating behaviors of CFC/Li were characterized to demonstrate the functions of highly interconnected channels in CFC/ZnO. As illustrated in Figure 2a, the macrovoids in CFC/
and exposure of original CFC/ZnO substrate. Microscale channels after stripping are clearly observed in Figure 2d, whose morphology is similar to original CFC/ZnO. More importantly, the discharge process did not remove all Li metal, but still left a thin layer of Li covering on individual carbon fibers (Figure 2e), which is helpful to reduce interfacial energy for plating of Li to eliminate nucleation overpotential.30 During subsequent Li plating process, lithium ions nucleated on the surface of unstripped Li metal layer. The same crystal structure and the reduction of current density by CFC synergistically lowered the nucleation overpotential and promised the uniform nucleation of Li. The ion concentration gradient was effectively alleviated by uniform distributed and interconnected microchannels in CFC/ZnO,31,32 resulting in highly reversible stripping/plating behavior, which can be confirmed by SEM observation of the structure of CFC/Li at a fully charged state after 20 cycles at 2.0 mA cm−2. The crosssectional SEM image (Figure 2f) of the electrode material demonstrates that the CFC/ZnO scaffold was refilled with Li metal, but there was no extra Li layer on the outer surface of CFC/Li electrode, though it showed metallic luster (inset in Figure 2f). SEM images of higher magnification (Figure 2g,h) more clearly displays that the microchannels between carbon fibers were fully filled with Li, and no dendrite-like lithium metal was visualized. The similar morphology of the CFC/Li electrode after 20 cycles compared to that of the fresh one suggests the prominent effect of CFC/ZnO in guiding highly reversible stripping/plating behavior of Li due to infinity between Li-wrapped carbon fibers after the initial stripping process and Li metal, and spatial confinement provided by microchannels between carbon fibers. Symmetrical cells were assembled using two identical CFC/ Li electrodes and ether-based electrolyte for evaluation of longterm cycling stability at high current densities of CFC/ZnOhosted Li. Various current densities were employed to compare the interfacial resistance and nucleation overpotential of bulk Li and CFC/Li cells (Figure 3a). The dashed lines represent the charge curves of bulk Li foil from 0.5 to 10.0 mA cm−2. The nucleation overpotential occurs at a relatively low current density of 2.0 mA cm−2 with voltage hysteresis of 60 mV. The voltage profiles show apparent fluctuation and inclining trends when the current density is above 5.0 mA cm−2, which is also accompanied by more obvious nucleation overpotential. The voltage hysteresis (sum of charge and discharge potential) reaches 150 mV with nucleation overpotential of 15 mV at a current density of 10.0 mA cm−2. The voltage hysteresis of CFC/Li (solid lines at discharge area) is much smaller compared to that of Li foil at all current densities. Specifically, its voltage hysteresis at low current densities below 2.0 mA cm−2 (11 mV at 0.5 mA cm−2, 18 mV at 1.0 mA cm−2, and 29 mV at 2.0 mA cm−2) are less than half of those for Li foil at the same current density. Even at a high current density of 15.0 mA cm−2, its voltage hysteresis of 120 mV is still 20% lower than that of the Li foil at 10.0 mA cm−2. Moreover, the nucleation overpotential is almost invisible for CFC/Li cells. Cells were also cycled at various current densities (1, 2, 5, and 10 mA cm−2) at a fixed areal capacity of 1.0 mA h cm−2 for monitoring their stability, and the corresponding voltage hysteresis of galvanostatic charge/discharge were measured and presented in Figure 3b,c,e,g. At a current density of 1.0 mA cm−2, voltage hysteresis of CFC/Li cell (Figure 3b) stabilizes at 20 mV (−10 to 10 mV) for over 1800 h (900 cycles). This ultralong lifetime has rarely been achieved by structural modification of Li-metal
Figure 2. (a, b) Illustration of exposure of highly interconnected microchannels in CFC/Li during stripping of Li. The thin layers on the carbon fibers with purple color represent remaining metallic Li layer after stripping. (c) Cross-sectional SEM image of CFC/Li, (d) top-view SEM image of CFC/Li, and (e) high-magnification SEM image of single carbon fiber after first stripping of Li for 6 h at 2 mA cm−2. Inset in (c) shows the digital photograph of CFC/Li electrode after stripping. (f) Cross-sectional SEM image of CFC/Li, (g) topview SEM image of CFC/Li, and (h) high-magnification SEM image of single carbon fiber after 20th plating of Li at 2 mA cm−2. Inset in (f) shows the digital photograph of CFC/Li electrode after plating.
Li enable initial infusion of electrolyte to reach the surface of carbon fiber bundles embedded with Li at the beginning of the stripping process. Then the continuous exposure of fresh microvoids and surface in the bundles, which are generated by continuous stripping of Li embedded in the microchannels, facilitates deeper infusion of electrolyte into the interior of carbon fiber bundles as illustrated in Figure 2b. This selfpromoting process continuously pushes forward the reaction interface between Li and electrolyte deep inside, resulting in smooth and controllable stripping of Li. To confirm this, CFC/ Li electrode was discharged at 2.0 mA cm−2 for 6 h at the first cycle to dissolve Li metal. SEM image in Figure 2c is an indication that top Li around CFC/ZnO was dissolved preferentially and the surface of carbon fiber bundles was exposed to interconnected channels after discharge. The inset digital photograph indicates that the color of CFC/Li also changed to black after Li stripping, suggesting the removal of Li 20389
DOI: 10.1021/acsami.8b02619 ACS Appl. Mater. Interfaces 2018, 10, 20387−20395
Research Article
ACS Applied Materials & Interfaces
Figure 3. (a) Voltage profiles of Li foil (dashed line, charge profiles) and CFC/Li anode (solid line, discharge profiles) at varying areal current densities. Galvanostatic curves of symmetrical cells at areal current densities of 1 mA cm−2 (b), 2 mA cm−2 (c), 5 mA cm−2 (e), and 10 mA cm−2 (g) using Li foil and CFC/Li electrodes at the areal capacity of 1 mA h cm−2. The voltage profiles of Li foil and CFC/Li cells from different cycles during long-term testing at the current density of 2 mA cm−2 (d), 5 mA cm−2 (f), and 10 mA cm−2 (h).
overpotential of nucleation and stable voltage for the CFC/Li cell even after long-term cycling. However, fluctuation of voltage and obvious overpotential are observed from bulk Li foil cells. The n-CFC/Li cell displays lower voltage hysteresis than that of bulk Li, but can also be stable for only 200 h with abrupt voltage fluctuation afterward (Figure S8). For further proof of the effect of CFC/ZnO in stabilization of Li stripping/ plating, areal capacity of 5.0 mA h cm−2 at 2.0 mA cm−2 current density has also been applied. The voltage profile (Figure S9) shows low voltage hysteresis of ∼38 mV and remains stable for over 300 h at such high areal capacity. More aggressive areal current densities (5.0 and 10.0 mA cm−2) were then used to further confirm high-rate capability and stability of CFC/Li anode. As is shown in Figure 3e, CFC/ Li cells cycled with extremely high stability for over 800 cycles (320 h) at 5 mA cm−2, which has rarely been achieved for Li symmetrical cells by using unmodified electrolyte. Figure 3f shows that the voltage of CFC/Li is stable for over 500 cycles at 5.0 mA cm−2 without polarization and fluctuation. As expected, step-by-step polarization, huge nucleation overpotential, and significantly expanded voltage hysteresis are observed in bulk Li foil and n-CFC/Li (Figure S10). At the same current density of 5.0 mA cm−2, a much higher areal capacity (3.0 mA h cm−2) was employed to further test the stability of CFC/Li (Figure S11), which could still work
anode. Similar cycle life has only been realized by modification of gel electrolyte with boron nitride nanoflakes.33 In contrast, cells of bulk Li have a much higher voltage hysteresis of 50 mV, and can stabilize only for 300 h even with the addition of LiNO3 in the electrolyte to enhance the formation of stable SEI.34−36 The rapid increase of voltage hysteresis after 150 cycles indicates the aggregation of “dead Li” on the surface of bulk Li, which causes high resistances and failure of cells.37 The control sample n-CFC/Li displays slightly better stability for over 420 h than that of Li foil, but its voltage hysteresis still increases during sequent cycles (Figure S6). The voltage profiles of CFC/Li and bulk Li foil at 1.0 mA cm−2 are compared in Figure S7. The nucleation of Li metal on CFC/Li electrode shows zero overpotential, and no evident increasing of voltage hysteresis is detected even after 500 cycles (1000 h); however, bulk Li foil still displays obvious nucleation overpotential and dramatic increase of voltage along with cycling, and its voltage−capacity curve also slightly fluctuates at such a low current density. When an elevated areal current density of 2.0 mA cm−2 was applied to symmetrical cells (Figure 3c), the voltage of bulk Li foil keeps increasing since the beginning of the cycling without a stable period, and the voltage hysteresis soon reaches 100 mV only after 200 h. In contrast, CFC/Li anode exhibits stable voltage for over 750 h. Figure 3d demonstrates zero 20390
DOI: 10.1021/acsami.8b02619 ACS Appl. Mater. Interfaces 2018, 10, 20387−20395
Research Article
ACS Applied Materials & Interfaces
densities from 1.0 to 10.0 mA cm−2, which manifests the repeatability of the excellent stability of CFC/Li at such high current densities. To further manifest the function of CFC on guiding stripping/plating behavior of Li metal, the Coulombic efficiency (CE) of pure CFC current collector at the current densities of 1.0, 2.0, and 5.0 mA cm−2 under the areal capacity of 3.0 mA h cm−2 was measured and shown in Figure S15. The initial CE of all cells is lower than 90%, and even approaching 80% at 5.0 mA cm−2. The relatively low initial CE is mainly ascribed to the large specific surface area of CFC (BET specific surface area of 1093.8 m2 g−,1 Figure S16), which induces the formation of relatively large amounts of SEI on the surface of CFC after initial charge/discharge processes. However, the CE keeps increasing in the subsequent cycles. It soon approaches 100% only after a few cycles and remains steady afterward. More specifically, the CE of the cell tested at 1.0 mA cm−2 quickly rises to 99.0% after several cycles, and lasts for over 140 cycles. The average CE at 2.0 mA cm−2 is also as high as ∼98.0% (over 100 cycles). Even at a relatively high current density of 5.0 mA cm−2, the CE also quickly climbs up from ∼80% in the first cycle to over 95% within 5 cycles, and remains stable for over 70 cycles. In great contrast, the CE of bare Cu foil fluctuates severely during the initial 15 cycles, and remains at ∼70% in the following cycles, indicating the poor reversibility of Li stripping/plating on planar Cu foil. The relatively high and stable CE of cells using CFC current collector suggests that the impact of exposed surface area of CFC on the reversibility of the cells is almost negligible. It can be understood that most micropores on CFC is filled by metallic Li after initial Li plating; thus, the surface area of CFC is enormously reduced and its effect on stripping/plating of Li is minimized. For fully revealing functions of CFC/ZnO in improving long-term stripping/plating behavior of Li, morphology of fully charged CFC/Li and bulk Li electrode after 150 h of cycling at 5.0 mA cm−2 was characterized. The digital photographs in Figure 5a,b show the disassembled CFC/Li cell and bulk Li cell. The separator in CFC/Li cell is still wetted by electrolyte, while the electrolyte is almost completely consumed in bulk Li cell. Bulk “dead Li” stripped from the Li foil is also observed on the separator because of the continuous formation and aggregation of dead Li during cycling of Li foil. SEM images of CFC/Li electrode after cycling are shown in Figure 5c−e. CFC/Li (Figure 5c) shows no extra Li layer covering on the surfaces even after 150 h of cycling, compared to the obvious “dead Li” layer on bulk Li (Figure 5f) and n-CFC/Li (marked by red dashed lines in Figure S17c,d). Such thick “dead Li” layers with huge resistance are the main cause for the polarization and voltage increase of the Li symmetrical cells. Instead of forming “dead Li”, the stripping/plating reaction of Li metal in the CFC/Li electrode is strictly confined in the microchannels in CFC/ZnO as demonstrated in Figure 5d,e. Low-magnification top-view SEM image (Figure S18) shows that most Li/SEI interphase is embedded in the channels or voids of CFC/ZnO with visualized CFC/ZnO network on top. No Li dendrites protruding on the surface are observed, suggesting that the deposition of Li is confined inside CFC. The top-view image (Figure 5d) of CFC/Li electrode demonstrates that intertwining carbon fibers are evidently seen, accompanied by Li/SEI interphase firmly embedded inside. The cell using CFC/Li anode can maintain stable voltage for over hundreds even thousands of hours. Though the Li metal inside microchannels of CFC/ZnO is pulverized after cycling (Figure 5e), the
properly for over 100 h. The electrochemical impedance spectroscopy (EIS) was employed to further indicate the excellence of CFC/Li on guiding dissolution/deposition behaviors of metallic Li (Figure S12). Fresh symmetrical cells of CFC/Li anode and Li foil were both observed with high resistance level due to the inert layer on the surface of the electrode. After cycling at 5.0 mA cm−2 for 50 cycles, the oxide layer was broken and the resistance of both cells reduced (19.8 Ω for the CFC/Li cell and 30.3 Ω for the Li foil cell). Along with the cycling, the resistance of the CFC/Li cell slowly increased to 36.7 Ω after 300 cycles, whereas that of the Li foil cell dramatically increased to 58.6 Ω. The much steadier resistance of the cell using CFC/Li anode implies that the CFC/ZnO scaffold is beneficial for enhancing the reversibility of Li stripping/plating. As an ultrahigh areal current density, 10.0 mA cm−2 was almost destructive to Li metal anodes reported in the literature.24,38 Few research studies have realized stable voltage longer than 20 h (100 cycles) for Li symmetrical cells based on unmodified ether-based electrolyte at such a high current density. Thanks to the 3D interconnected microchannels in CFC/ZnO that confine stripping/plating of Li in restricted areas and conductive CFC network that dissipates high current density, in this work, the CFC/Li symmetrical cell was stable at voltage profiles for over 60 h with only slightly increased polarization (Figure 3g). In contrast, the fluctuation of voltage (Figure 3h) of Li foil emerged after less than 5 h because of severe Li dendrite formation and it shows large voltage hysteresis of over 600 mV. The voltage hysteresis of the nCFC/Li cell also rapidly increases from 160 to 400 mV after less than 50 h (Figure S13). Figure 4 compares the life at
Figure 4. Comparison of this work to those in the literature about cycling time of Li symmetrical cells at different working areal current densities.
different areal current densities of Li symmetrical cells from previous literature and this work with fixed areal capacity of 1 mA h cm −2 (specific data was compared in Table S1).18,20,21,23,24,38−42 Most of previous research studies in the literature showed their stability in low areal current density condition (
