A High-Capacity and Long-Cycle-Life Lithium-Ion Battery Anode

Nov 15, 2016 - Department of Materials Science and Engineering, Korea Advanced Institute of Science and ... Metal-organic frameworks derived yolk-shel...
0 downloads 0 Views 7MB Size
A High-Capacity and Long-Cycle-Life LithiumIon Battery Anode Architecture: Silver Nanoparticle-Decorated SnO2/NiO Nanotubes Chanhoon Kim,† Ji-Won Jung,† Ki Ro Yoon,† Doo-Young Youn,† Soojin Park,‡ and Il-Doo Kim*,† †

Department of Materials Science and Engineering, Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea ‡ Department of Energy Engineering, School of Energy and Chemical Engineering, Ulsan National Institute of Science and Technology (UNIST), 50 UNIST-gil, Ulsan 44919, Republic of Korea S Supporting Information *

ABSTRACT: The combination of high-capacity and long-term cyclability has always been regarded as the first priority for next generation anode materials in lithium-ion batteries (LIBs). To meet these requirements, the Ag nanoparticle decorated mesoporous SnO2/NiO nanotube (m-SNT) anodes were synthesized via an electrospinning process, followed by fast ramping rate calcination and subsequent chemical reduction in this work. The one-dimensional porous hollow structure effectively alleviates a large volume expansion during cycling as well as provides a short lithium-ion duffusion length. Furthermore, metallic nickel (Ni) nanoparticles converted from the NiO nanograins during the lithiation process reversibly decompose Li2O during delithiation process, which significantly improves the reversible capacity of the m-SNT anodes. In addition, Ag nanoparticles uniformly decorated on the m-SNT via a simple chemical reduction process significantly improve rate capability and also contribute to long-term cyclability. The m-SNT@Ag anodes exhibited excellent cycling stability without obvious capacity fading after 500 cycles with a high capacity of 826 mAh g−1 at a high current density of 1000 mA g−1. Furthermore, even at a very high current density of 5000 mA g−1, the charge-specific capacity remained as high as 721 mAh g−1, corresponding to 60% of its initial capacity at a current density of 100 mA g−1. KEYWORDS: SnO2, NiO, nanotubes, porous structure, anodes, lithium-ion batteries

S

layer; (ii) the low reversible capacity of SnO2 arising from very large initial capacity losses (∼50%) due to the formation of highly stable Li2O; and (iii) the low electrical conductivity. To solve these problems, several nanostructures (such as nanoparticles, nanofibers, nanoflakes, and nanotubes) and various porous structures have been suggested.2,4−10 However, most of the suggested nano- and porous structures exhibited limited effects on alleviating the huge volume change of SnO2, which could not contribute to the increase in the reversible capacity of SnO2. Therefore, exploration of suitable design concepts not only for mitigating the huge volume change but also for

ince the invention of transition-metal oxide anodes in 1997 by Miyasaka, various transition-metal oxide (MxOy, M = Fe, Co, Ni, Cu, etc.) anodes have been intensively investigated, owing to their desirable properties such as high theoretical capacity, safety, and much improved rate capability than those of conventional graphite anodes.1−3 Among many transition-metal oxides, tin dioxide (SnO2) has been regarded as an attractive candidate considering its high theoretical capacity of 780 mAh g−1, environmental benigity, and abundance as well as other advantages among different transition-metal oxides. However, in spite of intensive studies on the SnO2 as a next generation anode over the past decade, several apparent demerits of SnO2 have not been completely solved including: (i) the huge volume change (>300%) generally occurs during the lithiation/delithiaion process, which leads to pulverization and aggregation of SnO2 particles and formation of unstable solid-electrolyte interface (SEI) © XXXX American Chemical Society

Received: September 27, 2016 Accepted: November 15, 2016 Published: November 15, 2016 A

DOI: 10.1021/acsnano.6b06512 ACS Nano XXXX, XXX, XXX−XXX

Article

www.acsnano.org

Article

ACS Nano

Figure 1. Schematic illustration of the synthetic route of the m-SNT@Ag.

Figure 2. SEM images of (a) as-spun fibers and (b) m-SNTs. The yellow arrows indicate that the NiO/SnO2 NFs have hollow structures. (c) Magnified SEM image showing the porous hollow structure. (d) TEM image of the m-SNTs. The yellow dotted circles indicate mesopores in the m-SNTs. (e) Corresponding HR-TEM image of the red box frame in (d). (f) EDS elemental mapping of O (blue), Sn (yellow), and Ni (red) in the m-SNTs.

agents and binders should be minimized to incorporate more active materials into a limited volume of electrodes. More recently, Zhu et al. synthesized three-dimensional (3D) nanoporous SnO2−NiO@C hybrid networks using Sn−Ni cyanogels and achieved a capacity of 614.0 mAh g−1 after 100 cycles at a current density of 100 mA g−1 and a stable capacity of 421.5 mAh g−1 at a high current density of 1A g−1.1 They first developed cyanogels as precursors to prepare a few nanosized SnO2 and NiO particles, interconnected hybrid structures, as anode materials in lithium-ion batteries (LIBs). They observed that additional reversible capacity increases over 2.5 V which attributed to the catalytic processes by the Ni nanocrystals and the 3D nanoporous structure effectively alleviated the huge volume expansion of SnO2. The 3D nanoporous SnO2−NiO@C hybrid networks delivered 614.9 mAh g−1 without noticeable capacity fading after 100 cycles under a current density of 100 mA g−1, but this value is not sufficient compared with other SnO2 nanocomposite anodes. Moreover, the complicated synthetic route including formation of SnCl4−K2Ni(CN)4 cyanogel, freeze-drying, calcination, and polypyrrole coating for carbon coating is not desirable for their practical use. Furthermore, K2Ni(CN)4, a Ni precursor of 3D nanoporous SnO2−NiO, is highly toxic and relatively expensive (∼2000 $ kg−1). Hence, it is necessary to consider a

increasing the reversible capacity and electrical conductivity is in great demand. In the recent literature, several hybrids composed of different metal oxides, especially SnO2 and MxOy, have exhibited surpassing electrochemical performances such as high reversible capacity as well as long-term cyclability over single-component oxides due to their synergistic effects.1,11,12 Hua et al. observed that a few nanometer-sized NiO and SnO2 composites, which were interconnected to each other, delivered much higher reversible capacities than theoretical capacity of both SnO2 and NiO, owing to the catalytic effect arising from Ni converted from NiO during the lithiation process, which reversibly decompose to Li2O.13 They have shown that metallic Sn was electrochemically oxidized to SnO2 below 3.00 V due to the catalytic effect of nanosized metallic Ni, leading to a high reversible capacity of the nanocomposite (970 mAh g−1) compared to the theoretical capacity of SnO2 and NiO. However, the NiO/SnO2 nanocomposites delivered sufficient capacity retention only for a small number of cycles (∼20 cycles) even with the use of large amount of both conductive agent (28.6 wt %) and binder (18.9 wt %) due to the aggregated structure of nanocomposites without consideration of the huge volume change of SnO2. Additionally, from the point of view of pursuing high energy density, use of conductive B

DOI: 10.1021/acsnano.6b06512 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 3. (a) XRD patterns of the m-SNTs. (b) Nitrogen adsorption−desorption curve of m-SNTs (inset: BJH pore size distribution).

has recently demonstrated that porous SnO2 nanotubes were formed by Kirkendall effect during the rapid calcination process of composite NFs containing Sn precursor and PVP.8 Similarly, the formation of the m-SNTs can also be explained by Kirkendall effect.17 When the temperature was increased during the calcination step, PVP was decomposed, and metal precursors (SnCl2·2H2O and Ni(NO3)2·6H2O) on the outer surface of the as-spun NFs were oxidized and converted to SnO2 and NiO, respectively. Then, metal precursors located at the center of as-spun NFs were also subsequently oxidized and diffused to the outer SnO2 layer and finally formed into larger grain-sized SnO2, owing to Ostwald ripening.8 Simultaneously, vacancies diffused to the center of the as-spun NFs where metal precursors moved outward. Finally, nanotubular morphology is formed after the rapid calcination process, resulting in m-SNTs with diameters varying from 200 to 400 nm. More importantly, the wall thickness of the m-SNTs after the rapid calcination process was about 28 nm. The relative amount of SnO2 and NiO was measured via inductively coupled plasma optical emission spectrometer (ICP-OES) and a ratio of SnO2/NiO in the m-SNTs was 2.5:1 (w/w). The nanostructure of the m-SNTs was further elucidated by transmission electron microscopy (TEM) in Figure 2d. The mesopores with diameter ranging from 2 to 50 nm are observed on the SNTs (the yellow dotted circles in Figure 2d). From the high-resolution TEM image, approximately few tens-of-nanometer-sized SnO2 and NiO nanograins were continuously linked and composed of porous tubular structure (Figure 2e). The measured interplanar distances were 0.33 and 0.17 nm, which was matched well with the (110) and (211) lattice planes of SnO2, respectively.1,2 The marked interplanar distance of 0.23 and 0.20 nm also corresponded to the (101) and (012) lattice planes of NiO, respectively.18 The selected area fast Fourier transform (FFT) pattern of the m-SNTs further confirmed the polycrystalline nature of the as-synthesized sample, as shown in Figure S1.19 Energy-dispersive X-ray (EDX) mapping further demonstrates that the Sn, Ni, and O elements are well-dispersed throughout the m-SNTs (Figure 2f). This 1D hollow porous structure of as-synthesized m-SNTs is an attractive design for LIB electrodes since such a structure holds the ability to provide an efficient 1D pathway for electron transport, which facilitates electrolyte penetration and improves strain relaxation. To clearly confirm the crystal structure of SnO2 and NiO in the m-SNTs, X-ray diffraction (XRD) patterns of the m-SNTs were obtained. The XRD patterns confirm that the m-SNTs are composed of tetragonal cassiterite phase of SnO2 (space group P42/mnm) and hexagonal NiO (space group R3̅m) compo-

straightforward, scalable, less expensive, and environmentally benign method in order to overcome the aforementioned challenges of SnO2 anodes. Herein, we present a simple and facile method to prepare mSNTs via simple electrospinning method and subsequent calcination under air atmosphere. Notably, one-dimensional (1D) hollow structures have been regarded as an effective way to alleviate the huge volume change.14 1D hollow interior cavities of m-SNTs shorten the Li+ ions diffusion length as well as offer more space to alleviate the strain arising from Li+ ions insertion/deinsertion.15 Furthermore, the introduction of porous structures into 1D hollow structures can effectively accelerate Li+ ion diffusion, which further enhances the electrochemical performances.16 The m-SNT composed of SnO2 and NiO nanograins, which are continuously interconnected, not only effectively alleviates the huge volume expansion of SnO2 but also provides much higher reversible capacities than theoretical values of SnO2, owing to the catalytic effect attributed to the Ni converted from NiO during the lithiation process. Furthermore, subsequent coating of the mSNTs with highly conductive Ag nanoparticles significantly improves electrical conductivity and provides a fast electron transfer pathway, resulting in a highly improved rate capability. The m-SNT anodes exhibit a much improved reversible capacity of 1335 mAh g−1 with an initial Coulombic efficiency (ICE) of 78% and high reversible capacity of 826 mAh g−1 after 500 cycles at the high current density of 1000 mA g−1. Furthermore, the capacity of the Ag nanoparticles decorated mSNT anodes at a high current density of 5000 mA g−1 discharge/charge is 60% of its capacity at a current density of 100 mA g−1 with a high active materials portion of 80 wt % with a less portion (10 wt %) of binder and conductive agent, respectively.

RESULTS AND DISCUSSION The Ag nanoparticle decorated m-SNTs were synthesized via three-step processes, as illustrated in Figure 1. Sn/Ni/ poly(vinylpyrrolidone) (PVP) precursor nanofibers (NFs) were prepared by an electrospinning method. The m-SNTs were synthesized via a subsequent calcination step in air. Then, Ag nanoparticles were coated on the surface of the m-SNTs via a wet chemical reduction process at relatively low temperature (50 °C). During the first step, the Sn/Ni/PVP precursor NFs with a diameter of 300−500 nm were prepared by an electrospinning process, as shown in Figure 2a. Interestingly, the 1D hollow porous structure was simply developed during the calcination process without any further treatment (Figure 2b,c). Our group C

DOI: 10.1021/acsnano.6b06512 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 4. Electrochemical performances of the electrodes made of the m-2.5SNTs and the m-5SNTs. CV curves in the first three cycles of (a) the m-2.5SNTs and (b) the m-5SNTs at a scan rate of 0.1 mV S−1. (c) Galvanostatic first cycle discharge/charge voltage profiles of the m2.5SNTs and the m-5SNTs at a current density of 50 mA g−1. (d) Cycle retentions of the m-2.5SNTs and the m-5SNTs at a current density of 500 mA g−1.

nents, as shown in Figure 3a.18,20 The average size of the crystalline domains is calculated by the Scherrer formula (eq 1):2 τ = Kλ /β cos θ

however, the structure was collapsed and agglomerated. In addition, a few hundreds of nanometer-sized NiO crystals were observed among the collapsed and agglomerated structures. In order to explain the agglomeration of the fibers with the same ratio between SnO2 and NiO, we have synthesized SnO2 and NiO fibers, respectively. The Sn/PVP precursor fibers with a diameter of 600−800 nm were prepared by electrospinning process, as shown in Figure S3. Interestingly, the 1D hollow porous structure was simply developed during the calcination process without any further treatment, similarly to the m2.5SNT (Figure S3b,c). The Ni/PVP precursor fibers with a diameter of 200−400 nm, which was smaller than that of the Sn/PNP precursor fibers, were prepared by an electrospinning process, as shown in Figure S3d. Then, the Ni/PVP precursor fibers were calcined in the same manner as Sn/PVP precursor fibers. In contrast with SnO2 fibers, however, NiO fibers are composed of NiO nanoparticles ranging from a few tens of nanometers to a few hundreds of nanometers, as shown in Figure S3e,f. As observed in Figure S3, NiO shows severe agglomerated morphology after the calcination process. Therefore, with the high content of Ni in Sn/Ni/PVP precursor fibers, the structure of as-prepared sample was collapsed and agglomerated due to the agglomeration behavior of NiO. Thus, we chose two samples (the m-2.5SNTs and the m-5SNTs) which maintain mesoporous tubular structures for electrochemical tests. The XRD patterns confirm that the m-5SNTs are composed of the same crystal structures of SnO2 (tetragonal cassiterite) and NiO (hexagonal) with the m-2.5SNTs components, as shown in Figure S4. To evaluate the electrochemical characterization of the m2.5SNTs and the m-5SNTs, we fabricated CR-2032 coin cells with Li metal as the counter electrode. We used 10% fluoroethylene carbonate (FEC) as a stable SEI layer formation agent for electrodes. FEC has been widely used as a reducible additive for stabilizing the surface of the electrodes during the

(1)

where τ is crystallite size, K is a dimensionless shape factor, λ is the X-ray wavelength, β is the line broadening at half the maximum intensity, and θ is the Bragg angle.21 From the Scherrer formula using the strongest peaks (2θ = 26.579° for SnO2, 2θ = 43.286° for NiO), the crystallite size of SnO2 and NiO in the m-SNTs was about 20 and 26 nm, respectively. These results are in good agreement with the TEM analysis. The surface morphology of the m-SNTs was further investigated by N2 adsorption−desorption isotherm curves (Figure 3b). The N2 adsorption/desorption isotherms of the m-SNTs show high absorption at a high relative pressure (P/ P0) ranging from 0.75 to 1, suggesting the existence of numerous mesopores and macropores.22 Moreover, the pore size distribution curve reveals that most of the pores in the mSNTs are meso- and macrosizes, indicating that the m-SNTs are suitable for electrolyte penetration and Li+ ion diffusion (inset in Figure 3b).23 Based on Brunauer−Emmett−Teller (BET) analysis, it was verified that the m-SNTs have a specific surface area of 8.8 m2 g−1. In order to optimize the composition between SnO2 and NiO in the m-SNTs for the optimal battery performance, samples with different ratio of SnO2/NiO were prepared by controlling the amount of precursors during the synthesis, as shown in Figure S2. When the m-SNTs with high SnO2 content (SnO2/NiO = 5:1, w/w, denoted as m-5SNTs) exhibited more increased diameter compared to the samples with a SnO2/NiO ratio of 2.5:1 (denoted as m-2.5SNTs). Furthermore, mesopores of about 30 nm diameter were created on the surface of the m-2.5SNTs. When the content of NiO increased to the same content of SnO2 (SnO2/NiO = 1:1, w/w), D

DOI: 10.1021/acsnano.6b06512 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano first few cycles.24 Figure 4a,b exhibits cyclic voltammogram (CV) curves of the initial three cycles of as-synthesized mSNTs within the voltage range of 0.005−3.0 V vs Li/Li+ at a scan rate of 0.1 mV s−1. In the CV curves of the m-5SNTs, two cathodic peaks were observed in the first cathodic scan. The sharp peak near 0.77 V on the first cathodic scan is caused by the initial reduction of SnO2 to Sn and the formation of the solid−electrolyte interface (SEI) layer as well as Li2O.8 Typically, these processes are well-known to be irreversible, as shown in the reaction below:9 SnO2 + 4Li+ + 4e− → Sn + 2Li 2O

nanocrystals converted from NiO could reversibly decompose to Li2O generated from both SnO2 and NiO.13 The reduction of Li2O generated from both SnO2 and NiO occurred by metallic nickel converted from NiO.13 Although the m-5SNTs also have the NiO, we could not observe the drastic increase in reversible capacity compared to the m-2.5SNTs due to the relatively small amount of NiO in the m-5SNTs. The cycling performances and Coulombic efficiencies (CEs) of the samples with different SnO2/NiO ratio are compared in Figure 4d. Both samples were cycled under a high current density of 500 mAh g−1 discharge/charge in the voltage range from 0.01 to 3.0 V. Both samples showed the initial capacity decay in the first 40 cycles which could be originated from SEI formation.27,28 On the basis of previous works, the disproportionate formation of the SEI layer on our samples resulted in a numerous consumption of Li+ and increase of inner electrode resistance, leading to initial capacity fading during initial cycles.27−29 However, the m-2.5SNTs show highly stable cycle retention after initial 40 cycles and reach an average CE near 99%. On the contrary, the rapid capacity fade and poor reversible capacity with fluctuation of CE were observed in the m-5SNTs during 100 cycles. Such enhanced CE and outstanding cycle retention of the m-2.5SNTs may be explained as follows: First, Ni nanocrystals serve as an effective catalyst for decomposition of Li2O as well as oxidation of metallic Sn from SnO2.13 Ni nanocrystals could be sufficiently generated from the m2.5SNTs rather than the m-5SNTs. Second, according to the magnified SEM images of the samples (Figure S5), the m5SNTs exhibited a more increased diameter and thinner walls compared to the m-2.5SNTs. Consequently, the structural integrity of hollow structures of the m-5SNTs may not be enough to tolerate the internal stress induced by volume change during the discharge/charge, resulting in a rapid capacity fading. On the contrary, the m-2.5SNTs possess both thicker walls and higher aspect ratio (low diameter). Thus, the m-2.5SNTs retain good structural integrity and exhibit good cycling performance. Lastly, the different lithiation potentials of NiO and SnO2 give rise to stepwise lithiation processes of the m-2.5SNTs, which can mitigate sudden internal stress arising from volume expansion of the metal oxide in the porous hollow NFs and maintain stable cycle retention.1 The lithiation process of m-2.5SNTs occurs with many plateaus compared to SnO2 and NiO nanoparticles, as shown in the Figure S6. In addition, the m-2.5SNTs possess higher NiO contents than that of the samples with high SnO2. Therefore, gradual lithiation process occurs in the m-2.5SNTs. Meanwhile, one of the aforementioned challenges of SnO2 anodes is a poor electrical conductivity which leads to sluggish electron transportation and poor rate capability.1,30 In order to enhance electrical conductivity of SnO2 anodes, surface coating of SnO 2 with conductive materials is a simple, but straightforward route. Typically, carbon coating process has been widely used for anode materials to improve electrical conductivity.31 However, the carbon coating process is usually carried out via thermal decomposition of carbon precursors, which are explosive gases such as methane and acetylene or a carbonization of carbon source materials at a high-temperature (>700 °C) procedure at inert atmosphere.32 Furthermore, SnO2 can be easily reduced under above carbon coating condition. In this study, Ag nanoparticle coating process was carried out on the surface of the m-2.5SNT via simple reduction process in the solution containing AgNO3, ethanol, and butylamine at a low temperature of 50 °C. It is generally

(2)

A cathodic peak near 0.2 V and three anodic peaks (0.5, 1.25, and 1.8 V) account for the alloying and dealloying processes, respectively, which are known to be reversible, as shown in the reaction below:9 Sn + x Li + x e− ⇌ LixSn (0 ≤ x ≤ 4.4)

(3)

Meanwhile, no obvious peaks resulted from reactions between NiO and Li+ ions were observed in the first scan of the m-5SNTs. However, during the second and third scan, small cathodic peaks near 1.2 and 0.45 V were observed, which are attributed to reduction of NiO. The detailed reaction is as follows:25,26 NiO + 2Li+ + 2e− → Ni + Li 2O

(4)

Anodic peaks regarding the opposite reaction of eq 4 were not found, which are attributed to the less amount of NiO in the m-5SNTs. The m-2.5SNTs show different CV characteristics from the m-5SNTs. In the first cathodic process, a sharp peak was clearly observed at near 0.45 V, which corresponds to the reduction of NiO to Ni and the formation of amorphous Li2O as well as a partially irreversible SEI layer (Figure 4b). Furthermore, an anodic peak near 2.2 V was also observed in the m-2.5SNTs, corresponding to the conversion of Ni to NiO, as shown in the reaction below:26 Ni + Li 2O → NiO + 2Li+ + 2e−

(5)

During the second and third scan, especially, more increased cathodic peaks of NiO near 1.2 and 0.45 V were observed, owing to the increased amount of NiO in the m-SNTs. The results from CV characteristics of the m-2.5SNTs and the m5SNTs imply that SnO2 and NiO in the porous tubular structure separately reacted with Li+ ions. Figure 4c shows the first galvanostatic discharge/charge voltage profiles of the m2.5SNTs and the m-5SNTs in the voltage range of 0.005−3.0 V at a current density of 50 mA g−1. The m-2.5SNTs deliver 1635.5 mAh g−1, which is similar to that of the m-5SNTs (1604.9 mAh g−1). Due to the small difference of theoretical gravimetric capacities between SnO2 (780 mAh g−1) and NiO (718 mAh g−1), the m-2.5SNT and the m-5SNT show the almost same charge capacity during the first lithiation process in spite of the different SnO2 and NiO ratios in samples. On the other hands, the m-2.5SNTs exhibit higher initial discharge capacity of 1266.0 mAh g−1 than m-5SNTs samples (1144.0 mAh g−1). The corresponding ICEs were 77.4% for the m2.5SNTs and 71.3% for the m-5SNTs, respectively. The reason for higher reversible capacity of the m-2.5SNTs can be explained by further delithiation reaction observed above 2.0 V, which is attributed to the catalytic effect of metallic nickel nanocrystals converted from NiO. 13 Hua et al. have demonstrated the catalytic effect that the metallic nickel E

DOI: 10.1021/acsnano.6b06512 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 5. Morphological analysis of the m-SNT@Ag using TEM. (a) TEM image of the SNT@Ag and (b) HR-TEM image corresponding selected area in (a). (c) HAADF-STEM image and corresponding STEM elemental mappings images of Sn (green), Ni (yellow), O (sky blue), Ag (red) in the m-SNT@Ag, respectively. (h) SAED pattern of the m-SNT@Ag.

known that the solubility of AgNO3 in an ethanol is quite low.32 However, the addition of butylamine significantly enhanced the solubility of AgNO3 in ethanol as well as slowly reduced AgNO3 to Ag nanoparticles.32 The particle growth of Ag is suppressed by the following complex equilibrium:33 [Ag(R‐NH 2)2 ]+ ↔ [Ag(R‐NH 2)]+ + R‐NH 2

(6)

[Ag(R‐NH 2)]+ ↔ Ag + + R‐NH 2

(7)

referred to as m-2.5SNT@Ag. According to the elemental mapping images of the m-2.5SNT@Ag, we can clearly distinguish the uniformly distributed Ag nanoparticles with a diameter of ∼10 nm coated on the surface of the m-SNT2.5@ Ag. In the selected area electron diffraction (SAED) pattern of the m-2.5SNT@Ag (Figure 5h), however, we could not observe a clear diffraction pattern of Ag. It can be speculated that the concentration of Ag nanoparticles was too low to display a clear diffraction pattern. According to XRD patterns of the m-2.5SNT@Ag, Ag nanoparticle coating process did not affect the crystalline structures of the m-2.5SNTs (Figure S7). In accordance with the result of the SAED pattern, we could not observe the XRD patterns corresponding to Ag which were too minimal to have enough intensity to have clear peaks. The Ag content in the 2.5SNT@Ag was measured using ICP-OES, which was determined to be about ∼2 wt %. To demonstrate the effect of Ag nanoparticle coating on the electrochemical performance of m-2.5SNT@Ag, a series of electrochemical tests were carried out. The m-2.5SNT@Ag electrode showed more improved first discharge and charge capacities of 1714.9 and 1335.3 mAh g−1, respectively, than that of the m-2.5SNT electrode (the first discharge and charge capacity of 1635.5 mAh g−1 and 1266.0 mAh g−1, respectively) (Figure 6a). Furthermore, the corresponding ICE of the m2.5SNT@Ag electrode is 77.8%, which is a slightly higher value

+

The Ag is gradually released from the complex equilibrium, which leads to the formation of Ag nanoparticles with diameter of