Trapping Lithium into Hollow Silica Microspheres with a Carbon

Dec 22, 2017 - Chemical Energy Release Driven Lithiophilic Layer on 1 m Commercial Brass Mesh toward Highly Stable Lithium Metal Batteries. Shaobo ...
2 downloads 0 Views 4MB Size
Download PDF
Letter Cite This: Nano Lett. XXXX, XXX, XXX−XXX

pubs.acs.org/NanoLett

Trapping Lithium into Hollow Silica Microspheres with a Carbon Nanotube Core for Dendrite-Free Lithium Metal Anodes Tong-Tong Zuo,†,‡ Ya-Xia Yin,†,‡ Shu-Hua Wang,† Peng-Fei Wang,†,‡ Xinan Yang,§ Jian Liu,†,‡ Chun-Peng Yang,† and Yu-Guo Guo*,†,‡ †

CAS Key Laboratory of Molecular Nanostructure and Nanotechnology, CAS Research/Education Center for Excellence in Molecular Sciences, Institute of Chemistry, Chinese Academy of Sciences (CAS), Beijing 100190, P. R. China ‡ University of Chinese Academy of Sciences, Beijing 100049, P. R. China § Beijing National Laboratory of Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences (CAS), Beijing 100190, P. R. China S Supporting Information *

ABSTRACT: Li metal anodes, which have attracted much attention for their high specific capacity and low redox potential, face a great challenge in realizing their practical application. The fatal issue of dendrite formation gives rise to internal short circuit and safety hazards and needs to be addressed. Here we propose a rational strategy of trapping Li within microcages to confine the deposition morphology and suppress dendrite growth. Microcages with a carbon nanotube core and porous silica sheath were prepared and proved to be effective for controlling the electrodeposition behavior. In addition, the insulative coating layer prevents concentrated electron flow and decreases the possibility of “hot spots” formation. Because of the Li trapper and uniform electron distribution, the electrode with delicate structure exhibits a dendrite-free morphology after plating 2 mA h cm−2 of Li. As the dendrite growth is suppressed, the as-obtained electrode maintains a high plating/stripping efficiency of 99% over 200 cycles. This work delivers new insights into the design of rational Li metal anodes and hastens the practical application of Li metal batteries. KEYWORDS: Li trapper, microcages, heterogeneous structure, Li metal anodes

T

Li metal anodes.38 Although the heterogeneous structure plays a significant role in regulating the deposition behavior, the elaborate regulation mechanism of Li metal is restricted by the deposition conditions (i.e., deposition capacity and current density). Therefore, the heterogeneous structure needs to be improved to guide uniform deposition in the case of excess deposition capacity. Herein we propose a heterogeneous structure with a conductive core and a porous insulative sheath for dendritefree Li metal anodes. The heterogeneous structure shows two advantages: (i) the conductive core provides nucleation sites for initial Li deposition, and the porous insulative sheath guarantees enough Li ion flux and constrains Li deposition within the microcages; (ii) the uniform insulative coating layer prevents locally high electric field and guides uniform deposition on the electrode. Following the rational idea, a composite microcage with a carbon nanotube core and a porous silica sheath was designed to subsequently accommodate Li metal. As a consequence, Li metal can be

he ever-increasing demands for sustainable energy facilitate the exploration for high-energy storage systems.1 Among all possible candidates, lithium metal batteries (LMBs) have received much attention because of their high energy density.2,3 In LMBs, the high specific capacity (3860 mA h g−1) and low electrochemical potential (−3.04 V vs standard hydrogen electrode) of Li metal hold promise for its use as an ideal anode material. However, dendrite formation induced by inhomogeneous deposition causes poor cycling efficiency and internal short circuiting, which further lead to battery failure and safety hazards. Since the dendrite issue severely impedes practical applications of LMBs, attempts must be made to address this fatal problem of Li metal anodes.4−8 Recently, quite a few approaches, such as electrolyte additives,9−17 stable interfacial layers,18−23 and modified electrodes,24−34 have been proposed to resolve the critical issues of Li metal anodes. As one of the most efficient strategies, regulating Li deposition with an elegant structure was proven to be effective.35−37 To control Li deposition over morphology and position, a nanocapsule structure was designed to realize the selective deposition of metallic Li through heterogeneous seeded growth, thus suppressing dendrite formation and improving the cycling performance of © XXXX American Chemical Society

Received: September 26, 2017 Revised: December 14, 2017 Published: December 22, 2017 A

DOI: 10.1021/acs.nanolett.7b04136 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 1. Preparation and characterization of H-SiO2/CNTs. (a) Schematic illustration of the procedure to prepare H-SiO2/CNTs. (b) SEM and (c) TEM images of PS/CNTs. (d) SEM and (e) TEM images of H-SiO2/CNTs. (f) Bright-field scanning TEM (STEM) image and the corresponding (g) C, (h) Si, (i) O, and (j) overlapped element mapping images of H-SiO2/CNTs.

accommodated in the microcage, where the heterogeneous structure serves as a Li trapper. Furthermore, the even electron transmission stemming from the insulative coating layer is beneficial for a relatively uniform Li distribution without direct deposition on the surface of the electrode. As a result of the Li trapper and insulative coating layer, the dendrite growth is effectively suppressed, and the Li metal anode maintains a high plating/stripping efficiency of 99% after 200 cycles. The microcage preparation process is schematically presented in Figure 1a. The polystyrene microsphere/carbon nanotubes (PS/CNTs) composite was first synthesized by polymerizing styrene on carbon nanotubes. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images (Figure 1b,c) show that PS microspheres are adhered on the CNTs. The UV−vis spectra (Figure S1) demonstrate that the absorbance of PS/CNTs in the visible range is remarkably decreased by the superficial PS microspheres on the CNTs. Subsequently, the PS/CNTs were homogeneously coated with a porous silica layer through a solution-based method and then calcined under an argon-filled atmosphere to prepare hollow silica microspheres/carbon nanotubes (H-SiO2/CNTs). The morphology and structure of H-SiO2/CNTs were confirmed by SEM and TEM images (Figure 1d,e). The hollow SiO2 microsphere with a diameter of ∼150 nm is anchored on CNT and inherits its morphology from the PS microsphere. As shown in the energy-dispersive Xray spectroscopy mapping images (Figure 1g−i), Si and O are uniformly distributed across the CNT, suggesting that the SiO2 sheath is homogeneously coated on the CNT. This is reconfirmed by the overlapped image (Figure 1j). Figure 2a shows a high-resolution TEM image of H-SiO2/CNTs, in which lattice strings with a fringe spacing of ∼0.34 nm are observed, corresponding to the (002) crystalline planes of CNTs. In addition, a silica layer with a thickness of ∼5 nm is observed on the CNT. These results agree well with the X-ray diffraction pattern (Figure 2b), where an obvious decrease in the intensity of the (002) diffraction peak is seen after silica coating. Raman spectroscopy (Figure S2) was also used to demonstrate the distribution of the silica layer. The intensities of the D and G bands in H-SiO2/CNTs are decreased by the

Figure 2. Structure characterizations of H-SiO2/CNTs. (a) Highresolution and (c) low-resolution TEM images of the H-SiO2/CNTs. (b) XRD patterns of CNTs, SiO2, and H-SiO2/CNTs powders. The impurity peaks in the CNTs pattern corresponding to the Ni catalyst are marked with black asterisks. (d) Pore size distribution of CNTs and H-SiO2/CNTs.

silica sheath, suggesting that the silica layer is distributed uniformly on CNTs. Nitrogen adsorption/desorption isotherms were measured to determine the pore structure. As shown in Figure 2d, the CNTs material with a low specific surface area of 57 m2 g−1 exhibits no obvious porous structure. In contrast, the asobtained H-SiO2/CNTs composite has a higher Brunauer− Emmett−Teller specific surface area of 1018 m2 g−1 and an average pore size of 2−3 nm, suggesting that the pores mainly originate from the porous sheath. This analysis is consistent with the experimental result. The mesoporous structure of the silica sheath, which can be observed in the TEM image (Figure 2c), facilitates Li+ ion diffusion through the insulative silica layer. To investigate the deposition behavior of Li with the heterogeneous structure, the H-SiO2/CNTs electrode was applied to plate Li metal (Figure 3a). We examined the B

DOI: 10.1021/acs.nanolett.7b04136 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

clearly recorded in the supporting video, which is consistent with the behavior of Li metal in previous literature;38 (iii) a broad peak around 55 eV was captured in the electron energy loss spectroscopy (EELS) spectrum (Figure 3e), corresponding to the K-edge of metallic Li.39,40 These results illustrate that the heterogeneous structure can trap Li metal in the microcages and restrict the deposition morphology. The electrode morphology after plating 2 mA h cm−2 of Li was observed to investigate the deposition behavior with different electrodes. As shown in Figure S4, the bare Cu anode is covered with numerous Li dendrites, which accelerates internal short circuiting and battery failure. The H-SiO2− CNTs composite electrode (the mass ratio of H-SiO2 and CNTs is 9:1 on the basis of the H-SiO2:CNTs mass ratio in the thermogravimetric analysis in Figure S5) was prepared and then applied to test the deposition morphology for comparison. For the H-SiO2−CNTs composite electrode as the anode, a number of dendrites exist underneath the electrode materials (Figure 4b), which can be clearly observed in the cross-sectional-view SEM image (Figure S6b). The phenomenon may be associated with the inhomogeneous distribution of electrons (Figure 4c). However, the H-SiO2/ CNTs electrode as the anode presents an even surface without dendrite formation (Figure 4d). As the electrode material is removed, the plated Li metal exhibits a smooth morphology instead of dendrites (Figure 4e), which coincides with the cross-sectional-view SEM image (Figure S6d). To explain the uniform deposition behavior, the electronic conductivities of different materials were measured with a four-contact method. Compared with pure CNTs (54 mS cm−1), H-SiO2/CNTs powders display a reduced electron conductivity of 2.1 × 10−3 mS cm−1. This result demonstrates that the insulative coating layer with lower conductivity prevents the formation of local “hot spots” and direct deposition on the electrode surface (Figure S7), which is beneficial for uniform deposition of Li metal (Figure 4f). Moreover, the H-SiO2/CNTs anode was observed by SEM after 10 plating/stripping cycles. As shown in Figure S8, the surface morphology of the anode remains smooth with no discernible dendrites. To investigate the maximal capacity of the H-SiO2/CNTs electrode, we further increased the deposition capacity to 6 mA h cm−2 (Figure S9). Li metal with dendrite-free morphology was plated on the surface of the electrode, suggesting that the maximal

Figure 3. Illustration of the Li trapper in the electrochemical deposition process. (a) Schematic presentation of Li deposition into the microcage. (b) TEM image of the microcage after Li deposition. (c) TEM image of the Li-plated microcage after in situ electron beam irradiation. (d) Bright-field STEM image and (e) the corresponding EELS spectra of the Li-plated microcage. The spectra were collected at the positions marked with red and blue circles in (d).

structure change of H-SiO2/CNTs after plating and confirmed the presence of Li metal in the microcages on the basis of the following evidence: (i) a translucent substance exists in the hollow silica microspheres after Li plating (Figure 3b), whereas the material without CNT remains unchanged after Li plating (Figure S3); (ii) the translucent substance melts and reshapes after the electron beam is focused on it (Figure 3c), and the volatile property of the substance during electron irradiation is

Figure 4. Surface morphology of Li metal anodes. (a, b) SEM images of (a) the electrode surface and (b) the current collector surface of the HSiO2−CNTs electrode after plating 2 mA h cm−2 of Li. (d, e) SEM images of (d) the electrode surface and (e) the current collector surface of the H-SiO2/CNTs electrode after plating 2 mA h cm−2 of Li. The insets in (a) and (d) show the optical images of the corresponding electrodes after plating of Li. (c, f) Schematic diagrams of Li deposition behavior on (c) H-SiO2−CNTs and (f) H-SiO2/CNTs electrodes. C

DOI: 10.1021/acs.nanolett.7b04136 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters deposition capacity is about 6 mA h cm−2. As a consequence, the heterogeneous structure mitigates dendrite formation and promises dendrite-free Li metal anodes. The solid−electrolyte interphase (SEI) layer shows a vital influence on the electrochemical performance. X-ray photoelectron spectroscopy (XPS) was peformed to analyze the chemical compositions of the SEIs on different electrodes (Figure S10). According to the previous literature,24,26 the features of ROLi, ROCOOLi, LiNxOy, and LiF correspond to the formed SEI constituents. The favorable SEI layer shows a synergistic effect on improving the Coulombic efficiency and cycling stability of the H-SiO2/CNTs electrode. Galvanostatic charge/discharge tests were also performed to evaluate the cycling stability and Li plating/stripping efficiency under a constant current density of 0.2 mA cm−2 and an areal capacity of 2 mA h cm−2. As shown in Figure 5a, the voltage

electrode displays a stable Coulombic efficiency of ∼94% after 40 cycles, where the hollow silica coating layer presents a beneficial effect on stabilizing the SEI layer and suppressing dendrite formation.43 Subsequently, the Coulombic efficiency of the H-SiO2−CNTs electrode decreases because of gradually deteriorating dendrite growth. However, the H-SiO2/CNTs electrode maintains a high plating/stripping efficiency of ∼99% after 200 cycles, which is attributed to the high reversibility of the H-SiO2/CNTs electrode due to the favorable SEI layer and reduced Li loss. When the measurement was conducted at a higher current density of 0.5 mA cm−2 (Figure 5d), the bare Cu and H-SiO2−CNTs electrodes showed rapid decays in Coulombic efficiency, indicating the ceaseless Li loss. However, the H-SiO2/CNTs electrode still maintained a high Coulombic efficiency of approximately 99% after 200 cycles, which demonstrates the stable framework and high reversibility of the H-SiO2/CNTs electrode. When the current density was increased to 1 mA cm−2 (Figure S13), the H-SiO2/ CNTs electrode exhibited higher and more stable Coulombic efficiency than the bare Cu and H-SiO2−CNTs electrodes. The improved stability stemming from effective suppression of dendrite growth promises Li metal batteries with long lifespans, as further confirmed by the high capacity retention (94.8% after 50 cycles) of the Li|LiFePO4 full cell (Figure S14). In summary, we have proposed a rational design of microcages with heterogeneous structure for dendrite-free Li metal anodes. With the conductive core providing electrons and the insulating porous sheath transporting Li+ flux, this distinctive structure induces homogeneous deposition instead of dendrite formation. Thereupon, we successfully prepared the hollow mesoporous silica microspheres anchored on carbon nanotubes to serve as a Li trapper. Li metal within the microcages and its morphology change were observed with TEM, suggesting that Li metal could be trapped into the hollow silica microspheres. The characteristic peak of Li metal in the EELS spectrum reconfirms the result. In addition, the reduced electronic conductivity of H-SiO2/CNTs is beneficial for uniform Li deposition, which prevents the “direct deposition” on the surface of electrodes when plating excess Li. The fatal issue of dendrite growth is mitigated by the delicate heterogeneous structure and rational electrode design. With the effective suppression of dendrite formation, the HSiO2/CNTs electrode displays a high plating/stripping efficiency of 99% over 200 cycles. The exceptional cycling stability and high Li utilization further demonstrate the effectiveness of the H-SiO2/CNTs electrode. Hence, the effective entrapment of Li metal guarantees favorable deposition behavior and excellent electrochemical performance. We believe that the concept of trapping Li within the microcages can realize dendrite-free Li metal anodes and hasten the practical application of rechargeable Li metal batteries.

Figure 5. Electrochemical performance of Li metal anodes. (a) Average voltage hysteresis of different electrodes at a current density of 0.2 mA cm−2. (b) Galvanostatic cycling profiles of the H-SiO2/ CNTs electrode at a current density of 0.2 mA cm−2. (c, d) Coulombic efficiencies of different electrodes at current densities of (c) 0.2 and (d) 0.5 mA cm−2 at a total capacity of 2 mA h cm−2.

hysteresis curves of bare Cu and H-SiO2−CNTs electrodes show a rapid increase after 25 and 60 cycles, respectively, which may result from the impedance change aroused from uncontrollable SEI formation. For the H-SiO2/CNTs electrode, the initial nucleation overpotential is lower than that of bare Cu (Figure S11), which is consistent with the nucleation behavior on a carbon substrate.38 Electrochemical impedance spectroscopy (EIS) analyses show that the charge transfer impedance decrease after five cycles (Figure S12), which agrees well with the change in voltage hysteresis. The H-SiO2/ CNTs electrode maintains a low voltage hysteresis (