Hollow Silicon Nanospheres Encapsulated with a Thin Carbon Shell: An Electrochemical Study

Electrochimica Acta 215 (2016) 126–141

Contents lists available at ScienceDirect

Electrochimica Acta journal homepage: www.elsevier.com/locate/electacta

Hollow Silicon Nanospheres Encapsulated with a Thin Carbon Shell: An Electrochemical Study Maziar Ashuria,b , Qianran Hea,b , Yuzi Liuc , Kan Zhanga,b , Satyanarayana Emania,b , Monica S. Sawickia,b , Jack S. Shamiea,b , Leon L. Shawa,b,* a b c

Department of Mechanical, Materials and Aerospace Engineering, Illinois Institute of Technology, Chicago, IL, USA Wanger Institute for Sustainable Energy Research (WISER), Illinois Institute of Technology, Chicago, IL, USA Center for Nanoscale Materials, Argonne National Laboratory, IL 60439, USA

A R T I C L E I N F O

Article history: Received 30 May 2016 Received in revised form 20 July 2016 Accepted 12 August 2016 Available online 20 August 2016 Keywords: Lithium-ion battery Anode Silicon Hollow Si nanospheres Sol-gel processing

A B S T R A C T

In this study we have investigated the electrochemical properties of hollow silicon nanospheres encapsulated with a thin carbon shell, HSi@C, as a potential candidate for lithium-ion battery anodes. Hollow Si nanospheres are formed using a templating method which is followed by carbon coating via carbonization of a pyrrole precursor to form HSi@C. The synthesis conditions and the resulting structure of HSi@C have been studied in detail to obtain the target design of hollow Si nanospheres encapsulated with a carbon shell. The HSi@C obtained exhibits much better electrochemical cycle stability than both micro- and nano-size silicon anodes and deliver a stable specific capacity of 700 mA h g 1 after 100 cycles at a current density of 2 A g 1 and 800 mA h g 1 after 120 cycles at a current density of 1 A g 1. The superior performance of HSi@C is attributed to the synergistic combination of the nanostructured material, the enhanced conductivity, and the presence of the central void space for Si expansion with little or no change in the volume of the entire HSi@C particle. This study is the first detailed investigation of the synthesis conditions to attain the desired structure of a hollow Si core with a conductive carbon shell. This study also offers guidelines to further enhance the specific capacity of HSi@C anodes in the future. ã 2016 Elsevier Ltd. All rights reserved.

1. Introduction Silicon is one of the most promising anode candidates for nextgeneration Li-ion batteries. This is due to its low voltage profile and high theoretical capacity (3590 mA h g 1 for Li15Si4 phase at room temperature), which is about ten times that of carbonaceous materials including graphite, pyrolytic carbon and meso-phase pitch (about 372 mA h g 1) [1]. In addition, silicon is the second ample element in the earth’s crust. Therefore, mass production of silicon with low cost is not an issue. However, practical application of silicon anodes is currently hindered by multiple challenges including the enormous volume change (300%) during lithiation/ delithiation processes, low intrinsic electrical conductivity, and instability of the solid electrolyte interphase (SEI) [2,3]. The large volume change can result in particle pulverization, loss of electrical contact with the conductive additive or current collector, and even

* Corresponding author at: Department of Mechanical, Materials and Aerospace Engineering, Illinois Institute of Technology, Chicago, IL, USA. E-mail address: [email protected] (L.L. Shaw). http://dx.doi.org/10.1016/j.electacta.2016.08.059 0013-4686/ã 2016 Elsevier Ltd. All rights reserved.

peeling off from the current collector. The repeated volume expansion and shrinkage also lead to fracture and re-formation of the SEI layer around the particles, resulting in continuous consumption of the electrolyte, increased impedance, and capacity fading [2–8]. Significant efforts have been devoted to addressing the issues mentioned above. The strategies investigated include Si material design through nanostructures [9–13], porous structures [14–17], or nanocomposites [18–20], Si electrode design with combined nano- and micro-particles [21] or with 3D micro-channels [22], nanoporous SiOx microparticles embedded in PANI-Ag layers [23], mesoporous SiOx nanorods [24], addition of electrolyte additives [25], and use of novel binders [26,27]. Thanks to the worldwide efforts, rapid advancements have been made in the properties and performance of Si anodes [28]. Examples of such breakthroughs include exceptional specific and volumetric capacities at 1160 mA h g 1 and 1270 mA h cm 3, respectively, after 1000 cycles at the C/2 rate accomplished through a pomegranate-inspired nanoscale design [13], high specific and volumetric capacities at 1600 mA h g 1 and 1088 mA h cm 3, respectively, after 150 cycles at 400 mA g 1 achieved via a

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

micro-sized porous Si/C material [16], and a specific capacity of 1450 and 1230 mA h g 1 after 100 cycles at 100 and 500 mA g 1, respectively, obtained via a low cost, double-shelled Si@SiOx@C core-shell structure [29]. Among various methods for Si material design, Si particle-based structures have attracted significant attention. Examples of Si particle-based structures include solid nanoparticles [3,29], solid core-shell structures [30–33] or core double-shell structures [34], and yolk-shell structures [10,11,35–38]. Lately, Chen, et al. [39] have investigated another type of structure which is composed of a hollow Si core encapsulated by a conductive shell (termed as hollow core-shell structures hereafter). Their study shows that the hollow Si core encapsulated with an Ag shell can deliver outstanding performance (i.e., 2,900 mA h g 1 after 100 cycles at the current density of 0.5 A g 1) because of the synergistic combination of the nanostructured material, the formation of adequate void inside for volume expansion, and the enhanced conductivity. In contrast, the hollow Si core encapsulated with a carbon shell exhibits less optimized performance than the hollow Si core encapsulated with an Ag shell, which is most likely due to the higher electrical conductivity of Ag than that of carbon [39]. In spite of the less optimized performance, the hollow Si core encapsulated with a carbon shell deserves additional investigation because carbon is more stable than Ag at the anodic condition. As reported in Ref. [39], the Ag shell starts to degrade after 100 cycles, indicating that the long-term cycle stability of the hollow Si core encapsulated with an Ag shell is questionable. In addition, a carbon shell can be deposited through a wide range of techniques such as chemical vapor deposition (CVD), carbonization of polymers, and wet chemistry methods with different precursors (sucrose, glucose, polydopamine, pyrrole, etc.) [3,30,31,37,40–42], making the hollow Si core encapsulated with a carbon shell attractive from the viewpoints of both low cost synthesis and long-term cycle stability. In this study, we have investigated synthesis and properties of the hollow Si core encapsulated with a carbon shell (HSi@C) by forming hollow Si spheres first, followed by carbon coating via carbonization of a pyrrole precursor. The hollow Si spheres are formed using a templating method, while the carbon coating formation is a facile carbonization treatment that does not require expensive equipment such as chemical vapor deposition (CVD) reactors. HSi@C anodes exhibit much better electrochemical cycle stability than both micro- and nano-size silicon anodes and deliver a stable specific capacity of 700 mA h g 1 after 100 cycles at a current density of 2 A g 1 and 800 mA h g 1 after 120 cycles at a current density of 1 A g 1. These properties are better than those obtained from many other Si material designs [36,41,43–48]. With the aid of a range of advanced analytical methods, the superior performance of HSi@C has been attributed to the unique engineered design of HSi@C nanospheres, i.e., a conductive carbon

127

shell for enhanced conductivity with the central void space for Si expansion. The structural defects in HSi@C have also been identified to provide guidelines for further improvement of HSi@C in the near future. 2. Experimental Details 2.1. Synthesis of carbon-coated hollow silicon spheres The synthesis steps of this study are shown schematically in Fig. 1. The synthesis started with polystyrene (PS) nanoparticles as templating agent onto which SiO2 was deposited via a sol-gel reaction (Fig. 1a and b) to form SiO2-coated PS nanoparticles (PS@SiO2). The PS core was then removed to generate hollow SiO2 spheres (HSiO2) which were converted to hollow Si spheres (HSi) via magnesiothermic and hydrogen reductions (Fig. 1c and d). Finally, HSi was encapsulated with a thin carbon shell to form hollow silicon nanospheres encapsulated with a thin carbon layer (HSi@C, Fig. 1e). The synthesis details are described below. Commercial positively charged polystyrene (PS) nanoparticles suspension (with the diameter of PS nanoparticles at 200 nm) was purchased from Sciventions Inc. and used as templating agent (Fig. 1a). The suspension was dispersed in a mixture of deionized water (16 mL) and ethanol (10 mL). Then, cetyltrimethylammonium bromide (CTAB, 10 mg) was added to the mixture and let the mixture stir for 10 min. This was followed by addition of concentrated ammonia aqueous solution (5 wt. %, 0.0428 mL) and then the addition of 0.48 mL of tetraethyl orthosilicate (TEOS) as silica precursor. The reacting sol-gel solution was stirred at room temperature for 3 h to form PS@SiO2 nanoparticles (Fig. 1b). The suspension was then centrifuged and washed with ethanol three times. The precipitates were collected and dried at room temperature for two days. The collected PS@SiO2 nanospheres were calcined inside a tubular furnace at 600  C for 6 h to burn out the PS core (Fig. 1c). The hollow silica spheres obtained were subsequently mixed with magnesium hydride (MgH2) powder and placed in a tantalum boat crucible inside an argon-filled glovebox. MgH2 powder was used because its particle size was finer than Mg powder and thus resulted in better mixing with PS@SiO2 nanospheres. The loaded crucible was then transferred to a homemade Swagelok1 pipe reactor which was put into a tubular furnace and subject to heating up to 700  C for 5 h. This step resulted in reduction of hollow SiO2 spheres to hollow Si spheres by H2 and Mg (Fig. 1d). Next, the reduced HSi nanospheres were dispersed in 1 M hydrochloric acid (HCl) and sonicated for 3 h or 6 h to dissolve residuals of Mg and also MgO compound formed during reduction. The nanospheres were washed with pure ethanol and deionized water 3 times. Eventually, HSi nanospheres were mixed with pyrrole (in 8:1 to 11:1 molar ratio) in a quartz crucible and then loaded into a stainless steel autoclave inside an argon-filled glove

Fig. 1. Schematic of synthesis steps: (a) PS nanoparticles, (b) PS@SiO2, (c) HSiO2, (d) HSi, and (e) HSi@C. See the text for details.

128

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

box. The loaded autoclave was heated to 600  C for 5 h to form HSi@C (Fig. 1e). The powder was collected after this step and utilized as the active material for electrochemical analysis. 2.2. Electrochemical measurements HSi@C spheres were employed as the active material and mixed with carbon black (CB, TIMCAL) and polyacrylic acid (PAA). The ratio of HSi@C to CB and PAA was 60:20:20 weight percent. The mixture was grinded in the agate mortar and pestle and then Nmethyl-2-pyrrolidone (NMP) was added to form a uniform slurry. The slurry was applied to a copper foil. The painted foil was vacuum-dried in an oven at 120  C for 12 h. After drying, the foil was punched to make the working electrode. CR2032 coin cells were fabricated using the working electrode and lithium metal chips as the counter and reference electrode. The electrolyte was 1 M LiPF6 salt in ethylene carbonate (EC)-diethylene carbonate (DEC) with 1:1 volume ratio and 10 vol. % fluoroethylene carbonate (FEC) as the additive. Celgard1 2325 was used as the separator. Batteries were discharged/charged with the Neware battery test system between 0.005 and 1.5 V vs. Li/Li+ at different current densities. Specific capacities and current densities were calculated based on the weight of the active material (HSi@C). The loading of the active material per electrode was typically 0.75 mg. Cyclic voltammetry (CV) test was performed with Parstat 4000 (Princeton Applied Research) from 0.005 to 1.5 V vs. Li/Li+. The same machine was used to collect electrochemical impedance spectroscopy (EIS) data from cells before and after cycling in the frequency range of 100,000 to 1 Hz. To fit the EIS data to an equivalent electronic circuit, ZSimpWin software package was employed. For the comparison purpose and to obtain a better understanding of the properties of HSi@C, three additional sets of batteries were fabricated. One was half cells with the working electrode made of micro-sized silicon powder ( 325 mesh, 99% trace metals) from Sigma-Aldrich1. The second one was half cells with the electrode made of HSi nanospheres without the carbon coating. The last half cells used the working electrode made of silicon nanoparticles (>99%, 70–130 nm) from NanoAmor1. Li metal chips were used as the counter and reference electrode in all cases, and the entire process of coin cell fabrication (such as the ratios of the active material to binder and carbon black, type of the separator and electrolyte, etc.) was identical to that of HSi@C half cells. 2.3. Material Characterization JEOL JSM-5900LV scanning electron microscope (SEM) and JEOL JSM-6701F field emission scanning electron microscope (FESEM) were used to investigate the change in particle size and morphology. Samples were first coated with gold using a sputter coater. JEOL JEM-3010 transmission electron microscope (TEM) was employed to determine particle size and morphology. In addition, energy-filtered TEM (EFTEM) at the Center for Nanoscale Materials (CNM) in Argonne National Laboratory (ANL) was utilized to obtain elemental mappings and determine the thickness of the carbon coating layer. X-ray diffraction (XRD) was carried out with Bruker D2 Phaser in the 2u range of 20 to 120 with Cu Ka radiation (1.54056 Å). To determine the carbon content of the sample, thermogravimetric analysis (TGA) was carried out under air with Mettler-Toledo TGA-SDTA851e. Brunauer, Emmett and Teller (BET) measurement was performed to determine the specific surface area (SSA) of samples at different processing stages with a two-channel Nova Quantachrome 2200e surface area & pore size analyzer. Raman spectra was collected using Renishaw inVia confocal Raman microscope equipped with a CCD detector. The excitation wavelength was 514 nm with the grating of 1800 lines/mm. The data was collected and processed

with Wire 3.4 software. Fourier transform infrared (FTIR) analysis was conducted with Thermo/Nicolet Nexus 470 FT-IR ESP spectrometer using potassium bromide (KBr) pellet method. The powder was ground in mortar and pestle with KBr at a ratio of 1:100. Then the mixture was pressed in to a die to form a pellet. The measurement range was between 4000 and 400 cm 1. 3. Results and Discussion The size and shape of the particles at different processing stages are tracked with both SEM and TEM (Figs. 2 and 3, respectively). As can be seen in these figures, PS nanoparticles with an average diameter of 200 nm are coated with sol-gel silica uniformly (Figs. 2b & b). Based on the size difference between PS and PS@SiO2 particles (Fig. 3a vs. b), it can be deduced that the thickness of the SiO2 shell is about 18 nm at the as-synthesized condition. After burnout of the PS core at 600  C the size of HSiO2 particles shrink to 200 nm, but the thickness of the SiO2 shell increases slightly to 25 nm (Fig. 3c). This is ascribed to removal of chemically bonded water in the sol-gel SiO2, which results in reduced particle sizes and thicker shell at the same time. Note that burnout of the PS core does not lead to cracked HSiO2 particles (Fig. 3c) and all HSiO2 particles remain to have similar sizes (Fig. 2c). This is due to the multi-step programmed heating schedule used in which PS@SiO2 nanospheres are heated to 400, 500 and 600  C successively and held at each temperature for 1 h. Without this multi-step holding schedule, significant cracking could occur to HSiO2 particles because of the rapid evolution of gases derived from burning of the PS core. The prevention of cracking is also helped by slow heating and cooling rates which are kept at 5  C min 1 in the PS burning treatment. After the magnesiothermic reduction with MgH2 and washing with HCl, HSiO2 nanoparticles change to HSi nanoparticles (to be confirmed later by other analytical methods) (Fig. 3d). The carbon coating formed at the final step is about 25 nm thick, determined using the carbon element mapping (Fig. 3f). However, it is noted that not all HSi nanoparticles are coated with carbon uniformly. As shown in Fig. 4, some HSi nanoparticles are agglomerated together and do not have a uniform carbon coating. Furthermore, some free carbon particles are found in the sample (see the upper right corner of Fig. 4a), indicating that the carbon coating using pyrrole as the precursor is not uniform throughout the entire sample. In spite of the non-uniformity, some HSi nanoparticles are coated very well with carbon (e.g., Fig. 3d and f). It should be pointed out that about 25% HSi@C nanospheres contain cracks, as shown in Fig. 2(e). We attribute this phenomenon to two major factors acting at the same time. The first is that the shell of HSi nanospheres are porous and the second is the nature of the reaction pathway in forming the carbon coating using the pyrrole precursor. Our TGA study of pure pyrrole (see Fig. S1 in Supporting Material) indicates that the reaction pathway in forming the carbon coating using the pyrrole precursor is vaporization of liquid pyrrole at 129  C to form gaseous pyrrole which then carbonizes in an argon atmosphere at 550  C or higher, depending on the final heating temperature used. Because of the porous nature of HSi nanospheres, liquid pyrrole not only stays outside HSi nanospheres, but also infiltrates into the nanospheres. During the carbon coating process heating above 130  C converts liquid pyrrole to gaseous pyrrole, leading to rapid volume expansion and thus causing cracks of some HSi nanospheres. The FTIR spectra of various nanospheres at different processing stages are illustrated in Fig. 5. Silica peaks are present after the solgel coating, confirming the formation of a SiO2 shell on the surface of PS nanospheres. However, it is noted that SiO2 peaks are present in all samples. This is not a surprise because Si spheres after the magnesiothermic reduction are nano-sized and can get oxidized easily to have a thin layer of oxide around the spheres. The peak at

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

1628 cm 1 is related to C-O bending, while the peak around 3400 cm 1 is due to the water absorbed by the powder (H-O-H stretching) [49–51]. No new peaks are observed after carbon coating. Although SiO2 peaks are present in all samples, most of the SiO2 peaks become weaker after the magnesiothermic reduction and carbon coating. This is particularly obvious with the Si-O-Si stretching vibration at 1076 cm 1 [30]. Fig. 6 presents XRD patterns of HSi nanospheres before and after 6 h wash with 1 M HCl. It can be seen that the magnesiothermic reduction has resulted in the formation of magnesium

129

oxide (MgO, ICDD PDF card 04-008-8202) along with the formation of crystalline silicon in the powder (Fig. 6a), indicating that the reaction of SiO2 + 2 Mg = 2 MgO + Si has taken place during the magnesiothermic reduction. However, no magnesium silicide (Mg2Si) is observed. After 6 h washing with HCl, all the XRD peaks in the final product match silicon (ICDD PDF card 04-014-8844) very well, indicating that the MgO byproduct has been removed from the sample. However, 3-h washing with HCl is not sufficient, leading to some residual MgO as indicated by the presence of the most intensive peak of MgO in the XRD pattern (Fig. S2).

Fig. 2. SEM micrographs: (a) PS nanoparticles, (b) PS@SiO2, (c) HSiO2, and (d & e) HSi@C at different magnifications. Note the uniform spherical shape of the final product, HSi@C, but some have cracks as shown in (e).

130

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

Fig. 3. TEM images of: (a) PS nanoparticles, (b) PS@SiO2, (c) HSiO2, (d) HSi, (e) HSi@C, and (f) carbon element mapping of the rectangular region in (e). The bright ring with the arrow in (f) delineates the thin carbon coating (20 nm), while the dark region is the hollow Si nanosphere.

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

131

Fig. 4. EFTEM images of (a) HSi@C and (b) the corresponding carbon element mapping of the entire region in (a). The bright spots and their intensities in (b) indicate the carbon distribution and its concentration. See the text for discussion.

Combining with SEM and TEM images (Figs. 2 and 3), we can conclude that the magnesiothermic reduction followed by HCl washing for 6 h has led to the formation of hollow Si nanospheres with no residual MgO. The TGA curve of the carbon-coated sample, HSi@C with 6 h washing, is shown in Fig. 7. Temperature was scanned from room temperature to 1000  C with the air flowing during the experiment. The weight loss of the sample at the temperature range from 400 to 600  C reveals that it contains 16.0 wt. % carbon. With this TGA data we have concluded that HSi@C nanospheres contain 84 wt. % Si. Fig. 8 depicts the Raman spectrum of 6h-washed HSi@C nanoparticles. The peaks at 1360 and 1595 cm 1 are designated as D and G bands, respectively. It is well known that the G band arises

from the in-plane vibration of sp2 graphitized carbon atoms, whereas the D band arises because of the disorder induced in sp2bonded carbon [43]. Furthermore, the peak intensity ratio between the D- and G-bands usually provides a useful index for comparing the degree of crystallinity of carbon materials, that is, the smaller the ratio of ID/IG, the higher the degree of ordering in the carbon material [52]. In the present case, the ID/IG value is 0.70, indicating that the carbon coating is amorphous or crystalline with very high defect concentration [52]. This conclusion is supported by the fact that the D-band width is larger than 100 cm 1, suggesting that the carbon is in amorphous state [53]. Note that the intensity of silicon is relatively low considering its 84 wt. % in the sample. However, this is not a surprise because Raman is a technique relying on reflection of the laser beam. Since the carbon is on the surface,

Fig. 5. FTIR spectra of HSi nanospheres after each synthesis step.

132

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

Fig. 6. XRD patterns of HSi: (a) before and (b) after 6-h wash with HCl.

most of the beam is reflected by carbon. As a result, the D band and G band signals from the carbon are strong, while the Si signal is not strong even though the Si weight fraction is much higher than the C in the HSi@C sample. The specific surface areas of samples at different stages of processing determined from the BET method are summarized in Table 1. PS nanospheres have a SSA of 24.82 m2 g 1 which is close to the estimation of the specific surface area of PS nanospheres with a diameter of 200 nm divided by their weight. After the sol-gel processing to form PS@SiO2, the SSA increases slightly to 26.61 m2 g 1. However, after burning the PS core the SSA increases dramatically to 847.28 m2 g 1. This drastic increase is due to two mechanisms. One is the removal of the PS core but with little change in the particle size. The contribution of this factor can be estimated by taking the diameter of HSiO2 nanospheres to be 200 nm (Fig. 3c) and dividing their surface area by the weight of the porous SiO2 shell only. This estimation leads to a specific surface area of 65 m2 g 1 which is about 150% higher than the SSA of PS@SiO2, but still dramatically lower than the measured SSA of

847 m2 g 1. The second mechanism responsible for the dramatic increase in SSA is the removal of chemically bonded water within the SiO2 gel. It is well known that sol-gel processing is a powerful approach to synthesize porous materials with enormous specific surface area [54,55]. It has been shown that sol-gel processed SiO2 gels can reach SSA as high as 1,000 m2 g 1 after calcination at 450  C [56]. Thus, the dramatically increased SSA observed in this study is consistent with the literature. It is interesting to note that the SSA drops from 847.28 to 46.33 m2 g 1 after the magnesiothermic reduction at 700  C for 5 h (46.33 m2 g 1). This new SSA suggests that most of the micro- and meso-pores (pore size < 50 nm) created in HSiO2 have been eliminated because of the sintering effect at 700  C for 5 h. However, the new SSA value (46.33 m2 g 1) indicates that HSi nanospheres still contain some micro- or meso-pores. Otherwise, dense HSi nanospheres with an average diameter of 225 nm and a shell thickness of 19 nm (Fig. 3d) can only have SSA at 27.5 m2 g 1 (if we calculate the SSA by taking the Si shell being fully dense with a theoretical density of 2.328 g cm 3). The porous nature of HSi

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

133

Fig. 7. TGA curve of 6h-washed HSi@C nanospheres under air atmosphere.

Fig. 8. Raman spectrum of HSi@C nanospheres.

Table 1 Specific surface area of nanospheres after each step of synthesis. Material

Specific Surface Area (m2 g

PS PS@SiO2 HSiO2 HSi HSi@C

24.816 26.608 847.283 46.330 189.581

1

)

nanospheres is consistent with the fact that the conversion of SiO2 to Si is associated with about 50% volume change. It is also in good agreement with the presence of cracked HSi@C nanospheres after the carbon coating using the pyrrole precursor, as discussed before (Fig. 2e). Finally, it is equally interesting to note that after coating

HSi nanospheres with pyrrole, SSA increases again to 189.58 m2 g 1. This indicates that the carbon coating is also porous because the particle size has not changed much from HSi to HSi@C. The porosity in the carbon coating can facilitate lithium diffusion during electrochemical cycling. The charge-discharge performance of half cells made of 6hwashed HSi@C have been evaluated at 0.5 A g 1 and shown in Fig. 9(a). The first discharge and charge capacities are 2,230 and 1,610 mA h g 1, respectively. The first cycle Coulombic efficiency (CE) is only about 70%. This value is due to the formation of solid electrolyte interface (SEI) layer around the particles. However, the efficiency in the subsequent cycles is improved gradually and eventually becomes higher than 99% (to be discussed more later). The voltage profiles during charge and discharge are in good accordance with other researches [1,3,9], i.e., a sloping voltage

134

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

Fig. 9. (a) Voltage profile of the 6h-washed HSi@C half cell for cycles 1, 2, 5, 10, 15 and 20, and (b) its cyclic voltammogram (CV) for the first five cycles between 0.005 and 1.5 V (the scan rate was set to 0.05 mV s 1).

profile from 0.25 V to 0.005 V during discharge and a voltage plateau at 0.45 V during charge. The half cells made of 3h-washed HSi@C have also been evaluated and shown poor charge and discharge capacities (Fig. S3) than 6h-washed HSi@C counterparts. This phenomenon can be attributed to the presence of residual MgO in the 3h-washed sample. To investigate the electrochemical reactions upon charge/ discharge processes, we have performed cyclic voltammetry (Fig. 9b). Since the insertion and de-insertion of lithium ions into silicon is a very slow process, the scan rate is set to 0.05 mV s 1. In the CV experiments the first operation is the cathodic scan starting from the open circuit voltage (2.5 V vs. Li/Li+) and ending at 0.005 V. The subsequent scans are between 0.005 and 1.5 V. The peaks at 0.45 V and 1.1 V in the first cathodic scan are due to the SEI layer formation around the HSi@C nanospheres, while the peak between 0.05 and 0.2 V are related to alloying of Li and Si (LixSi) [18,36]. The 0.45 V and 1.1 V peaks disappear in the subsequent scans, indicating that most of SEI layer formation takes place in the

first scan. The broad peak in the first cathodic scan at 1.6 V may be related to the electrochemical reactions of Li ions with trace amounts of SiO2 or MgO in the sample [36,57]. Two anodic peaks centered at 0.35 and 0.56 V are associated with the delithiation of Li-Si alloy [48]. Both of these peaks become stronger with increasing the number of scans, consistent with other studies [18,36,58,59]. When comparing the CV curves of HSi@C (Fig. 9b) with those of HSi (Fig. 10a), it is interesting to note that the two anodic peaks of HSi half cells are less obvious than those of HSi@C counterparts and the cathodic peak between 0.05 and 0.2 V disappears from HSi half cells. We attribute this phenomenon to the fact that HSi spheres without carbon coating have lower electronic conductivity and thus slower electrochemical reaction rates than HSi@C spheres. This hypothesis is supported by the CV curves of Si nanoparticle half cells (Fig. 10b) which display very similar CV curves as HSi@C half cells. Si nanoparticles (100 nm) can mix with carbon black nanoparticles (50 nm) better than HSi spheres (200 nm) and

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

Fig. 10. CV curves of (a) HSi half cells and (b) Si nanoparticle half cells for the first five cycles between 0.005 and 1.5 V (the scan rate was set to 0.1 mV s

thus large portion of Si nanoparticles in the electrode can participate in electrochemical reactions at the same time. As a result, the anodic and cathodic peak intensities are stronger for HSi@C and Si nanoparticle half cells than HSi half cells even though they have the same amount of carbon black in the electrodes. Fig. 11 displays the cycle stability of 6h-washed HSi@C cells at different current densities and the associated coulombic efficiency as a function of cycles. Here the specific capacity is normalized by the HSi@C weight. Note that it takes about 20 cycles for the Coulombic efficiency to reach >98%, indicating the presence of some irreversible processes in the first 20 cycles. Accompanied with the change in the Coulombic efficiency, the discharge capacity in the first 20 cycles also displays significant decrease (changing from 2,230 mA h g 1 in the first cycle to 1,300 mA h g 1 in the 20th cycle). However, after 20 cycles the charge/discharge capacities become much more stable, showing smaller and gradual decreases with cycles and suggesting a great potential for long cycle life. The capacity decay in the first 20 cycles is attributed to two major factors. The large decay from the first cycle to the second

135

1

).

one is due to the formation of the SEI layer in all HSi@C nanospheres, while the gradual decay in the subsequent cycles is due to the gradual formation and fracture of the SEI layer in cracked HSi@C nanospheres. Fig. 12 shows schematically the different behaviors of HSi@C nanospheres with and without cracks during cycles. For HSi@C nanospheres (Fig. 12a) the hollow Si will expand inward in the lithiation process because of the mechanical constraint of the outer carbon shell. In the delithiation process the Li-Si alloy will convert back to Si and recover to the original HSi@C structure. The SEI layer formed in the lithiation process will remain intact during delithiation process and subsequent lithiation process. As a result, HSi@C nanospheres have a stable SEI layer with a long cycle life. In contrast, in the case of HSi@C nanospheres with cracks (Fig. 12b) the SEI layer formed inside the crack(s) during the lithiation process will fracture in the delithiation process. This in turn exposes new Si to the liquid electrolyte and causes SEI layer formation again in the next lithiation process. Repetition of such SEI layer formation and fracture can result in continuous consumption of the electrolyte, increased impedance,

136

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

Fig. 11. Specific capacity and Coulombic efficiency of HSi@C half-cells vs. cycle number at different current densities.

Fig. 12. Schematic of the structure change of HSi@C nanospheres upon charge and discharge processes: (a) HSi@C without cracks and (b) HSi@C with cracks. Every component is color coded. Note that the SEI layer fractures inside the cracked HSi@C nanospheres during delithiation. See the text for further discussion.

and capacity fading. About 25% HSi@C nanospheres are cracked (Fig. 2) under the synthesis condition in this study, and these cracked HSi@C nanospheres will lead to capacity decay over cycles until they no longer contribute to the capacity. Non-uniform carbon coatings could also contribute to the gradual decay of the capacity in cycles beyond the first cycle. It should be mentioned that Fig. 12 only highlights the SEI layer formation on the external surface of HSi@C and the internal surface of HSi@C with cracks. However, there is a possibility of forming a thin SEI layer in the internal surface of the hollow Si because both

the hollow Si and C shell are porous which could allow the liquid electrolyte to penetrate into the central void space before cycling. However, the effect of this thin internal SEI layer, if present, is negligible comparing to that of the SEI layer formed in HSi@C spheres with cracks. In the latter case the SEI layer can continue to fracture and form over cycles. In contrast, the internal SEI layer in HSi@C spheres without cracks is unlikely to grow after the first cycle because lithiation and delithiation of the Si are likely to remove the meso- and micro-pores in the Si or the SEI films formed inside the meso- and micro-pores of the carbon shell could prevent

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

the electrolyte from penetrating into the Si in the subsequent cycles. In spite of these possibilities, the exact situation remains to be investigated in the future. Fig. 11 also reveals the rate capability of HSi@C nanospheres. When the rate is increased from 0.5 A g 1 to 1 A g 1 at the 20th cycle, the capacity decreases from 1,220 mA h g 1 to 1,110 mA h g 1. The capacity decreases again from 980 mA h g 1 to 710 mA h g 1 when the rate increases from 1 A g 1 to 2 A g 1 at the 40th cycle. Similarly, the capacity decreases again at the 60th cycle when the rate increases. After 80 cycles the rate is switched back to 2 A g 1 and the capacity is recovered to 700 mA h g 1 which is similar to the capacity at the 60th cycle with the rate at 2 A g 1. When the rate is switched further back to 1 A g 1 at the 100th cycle, the capacity increases again. All of these illustrate the possibility of fast kinetic reactions and robust stability of HSi@C electrodes. Because of the unique structure and properties of HSi@C nanospheres without cracks the performance of HSi@C cells is better than many recent studies [36,41,43–48,58,59]. For example,

137

Su et al. [43] have synthesized a Si/graphite@graphene composite with the structure of a Si/graphite mixture wrapped by graphene sheets. The composite exhibits an initial charge capacity of 803 mA h g 1 at 0.1 A g 1 and retain 500 mA h g 1 after 50 cycles, which is significantly lower than the specific capacity of HSi@C (700 mA h g 1 at 2 A g 1 after 60 cycles). Another case for comparison is the Si/Cu core-shell structure [45]. This material contains a solid Si core with a Cu shell and exhibits only 400 mA h g 1 capacity at 0.1 A g 1 after 20 cycles [45]. In another case for comparison a yolk-shell Si@SiO2 with void space between the Si core and SiO2 shell has been synthesized [36]. This yolk-shell structure only delivers 687 mA h g 1 at 0.05 A g 1 after 30 cycles and its specific capacity decreases to less than 400 mA h g 1 after 25 cycles if the current density is increased to 0.8 A g 1 [36]. In short, the specific capacity, cycle stability and rate capability of HSi@C is better than many Si material designs reported in Refs. [36,41,43–48,58,59]. The mechanism responsible for this superior performance of HSi@C has been depicted in Fig. 12a.

Fig. 13. Electrochemical impedance spectra of (a) 6h-washed HSi half cells and (b) 6h-washed HSi@C half cells before cycles.

138

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

Fig. 14. Specific capacity and Coulombic efficiency of 6h-washed HSi half cells vs. cycle number at different current densities.

To provide direct assessment of the advantages of HSi@C, we have also assembled three additional types of half cells made of Si nanoparticles, Si micro-crystalline particles, or HSi nanospheres. Fig. 13 compares the electrochemical impedance spectra of HSi and HSi@C half cells. It can be seen that both types of half cells exhibit a semicircle in the high frequency range, which is assigned to the charge transfer resistance (Rct) [60]. However, the Rct of HSi is about 6,000 V, whereas the corresponding value for HSi@C is only about 360 V, indicating significantly lower charge transfer resistance of HSi@C due to its improved electronic conductivity enabled by the carbon coating. The ohmic resistance of the HSi@C half cell is also much smaller than that of the HSi half cells, as displayed by the intercepts of their respective semicircles with the real axis at the high frequency end. Fig. 14 shows the specific capacity and Coulombic efficiency of HSi half cells as a function of cycle numbers at different current

densities. Although HSi half cells deliver a high first discharge capacity (3,130 mA h g 1), their capacities fade very fast with only 650 mA h g 1 at the 20th cycle. When the rate is increased to 1 A g 1 at the 20th cycle, the capacity delivered is only 250 mA h g 1. After 40 cycles, the capacity is approaching zero. These results are dramatically worse than the properties offered by HSi@C half cells (Fig. 11), unambiguously revealing the critical advantages offered by the carbon coating. Fig. 15 compares the discharge capacities of 6h-washed HSi@C, 3h-washed HSi@C, Si nanoparticle and Si micro-crystalline particle half cells. Although both micro- and nano-size Si particles have very high first discharge capacities (3,600 mA h g 1) similar to HSi nanospheres, their capacities fade very fast. The specific capacity of the micro-crystalline Si cell drops to below 100 mA h g 1 only after 20 cycles, while the Si nanoparticle cell has a better performance than the micro-crystalline Si cell, but its capacity also

Fig. 15. Discharge capacity vs. cycle number of 6h-washed HSi@C cells in comparison with 3h-washed HSi@C, silicon nanoparticle and micron-size silicon particle cells.

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

139

Fig. 16. Impedance spectra of 3h-washed HSi@C cells before and after 5 cycles. The insert is the equivalent circuit used to fit the impedance data of HSi@C cells.

drops below 500 mA h g 1 after 40 cycles. The rapid capacity fading of the micro-crystalline Si cell is likely due to Si particle cracking, whereas the Si nanoparticle cell has a better performance because Si nanoparticles (70–130 nm) have lower propensity for cracking. However, the Si nanoparticle cell still has an inferior performance to the HSi@C cells because nanoparticles, although may not crack, still undergo volume expansion and shrinkage during lithiation and delithiation, which can result in repeated formation and fracture of the SEI layer and thus increased impedance, continued consumption of the electrolyte and capacity fading [28]. In sharp contrast, the void space inside HSi@C nanospheres can allow Si to expand during lithiation with little or no change in the volume of the entire particle. This minimized volume expansion and shrinkage prevents particle fracture and leads to a stable SEI layer on the surface of the particle (Fig. 12a), while keeping all or most of particles in contact with conductive pathways, thereby better performance than many other Si material designs. The 3h-washed HSi@C cell has poorer performance than the 6h-washed HSi@C cell because the former contains residual MgO and the latter does not. To further establish the relationship between cycling performance and electrode kinetics of HSi@C nanospheres, the electrochemical impedance spectra of 3h-washed HSi@C half cells before and after 5 cycles are measured and compared (Fig. 16). The impedance data are fitted into an equivalent circuit shown as the insert in Fig. 16. This model is chosen based on the fundamental studies conducted by Wu, et al. [60]. In this model Rint represents the ohmic resistance of all components of the cell including the separator, current collectors, electrode materials, ionic resistance of the electrolyte, and resistance at interfaces among them. Rct is the charge transfer resistance, Cdl is the double layer pseudocapacity, and Warburg element is the solid state diffusion. The fitted data are summarized in Table 2. Note that both Rint and Rct have increased after 5 cycles. This is attributed to the formation of Table 2 Equivalent circuit fitted data. Fitted data Cell status

Rint (V)

Rct (V)

Cdl (F)

Before cycling After 5 cycles

87.47 157.4

306.4 346.0

4.20  10-7 2.41 10-7

the SEI layer around the particles and may also be caused by SEI layer formation inside cracked HSi@C nanospheres as schematically shown in Fig. 12b. Finally, it should be pointed out that in spite of the better performance of HSi@C than many other Si material designs including Si nanoparticles, solid core-shell structure [44,45], some yolk-shell structures [36,41], Si/graphene composite [46] and Si/ graphite@graphene composite [43], HSi@C in the present study is still inferior to some recent outstanding studies such as those mentioned earlier in the introduction [13,14,16,29,61,62]. On one hand, these comparisons demonstrate the great potential of the HSi@C design in enhancing the specific capacity of Si anodes with long cycle life. On the other hand, these comparisons also highlight the critical need to optimize the synthesis conditions in the future to further improve the electrochemical properties of HSi@C. Elimination of cracked HSi@C nanospheres, improving the coating uniformity, and possibly increasing the thickness of the carbon coating are all worthy of further investigation in the near future. 4. Conclusions We have investigated synthesis, structure and electrochemical properties of HSi@C by forming hollow Si spheres using a templating method first, followed by carbon coating via carbonization of a pyrrole precursor. HSi@C anodes exhibit much better electrochemical cycle stability than HSi anodes as well as micro- and nano-size silicon anodes and deliver a stable specific capacity of 700 mA h g 1 after 100 cycles at a current density of 2 A g 1 and 800 mA h g 1 after 120 cycles at a current density of 1 A g 1. These properties are also better than those obtained from many other Si material designs such as Si coreshell structure, some yolk-shell structures, Si/graphene composite, and Si/graphite@graphene composite. The superior performance of HSi@C is due to its unique engineered design. Its outer carbon shell offers conductivity and mechanical constraint simultaneously, while the central void space allows the hollow Si core to expand inwards during lithiation with little or no change in the volume of the entire HSi@C particle. This minimized volume expansion prevents particle fracture and leads to a stable SEI layer on the surface of the particle, while keeping all or most of the particles in contact with conductive

140

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141

pathways, thereby providing better performance than many other Si material designs. This study also proves that by having a thin carbon shell around a hollow silicon core, one can obtain high capacity with good cycle stability. However, further optimization in synthesis is still needed to completely eliminate cracking of HSi@C nanospheres and improve carbon coating uniformity, thereby enhancing their electrochemical performance. Acknowledgements MA and LS are grateful to the Rowe Family Endowment Fund, and QH acknowledges Tang Fellowship. Use of the Center for Nanoscale Materials, an Office of Science user facility, was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC0206CH11357. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. electacta.2016.08.059. References [1] M. Obrovac, L. Krause, Reversible cycling of crystalline silicon powder, Journal of The Electrochemical Society 154 (2007) A103–A108. [2] C.K. Chan, H. Peng, G. Liu, K. McIlwrath, X.F. Zhang, R.A. Huggins, Y. Cui, Highperformance lithium battery anodes using silicon nanowires, Nature nanotechnology 3 (2008) 31–35. [3] L. Pan, H. Wang, D. Gao, S. Chen, L. Tan, L. Li, Facile synthesis of yolk–shell structured Si–C nanocomposites as anodes for lithium-ion batteries, Chemical Communications 50 (2014) 5878–5880. [4] S.D. Beattie, D. Larcher, M. Morcrette, B. Simon, J.-M. Tarascon, Si electrodes for Li-ion batteries—a new way to look at an old problem, Journal of The Electrochemical Society 155 (2008) A158–A163. [5] M. Holzapfel, H. Buqa, L.J. Hardwick, M. Hahn, A. Würsig, W. Scheifele, P. Novák, R. Kötz, C. Veit, F.-M. Petrat, Nano silicon for lithium-ion batteries, Electrochimica acta 52 (2006) 973–978. [6] B. Liang, Y. Liu, Y. Xu, Silicon-based Materials as High Capacity Anodes for Next Generation Lithium Ion Batteries, Journal of Power Sources (2014) 469–490. [7] H.K. Liu, Z. Guo, J. Wang, K. Konstantinov, Si-based anode materials for lithium rechargeable batteries, Journal of Materials Chemistry 20 (2010) 10055–10057. [8] W.-J. Zhang, A review of the electrochemical performance of alloy anodes for lithium-ion batteries, Journal of Power Sources 196 (2011) 13–24. [9] H. Kim, M. Seo, M.-H. Park, J. Cho, A Critical Size of Silicon Nano-Anodes for Lithium Rechargeable Batteries, Angewandte Chemie International Edition 49 (2010) 2146–2149. [10] N. Liu, H. Wu, M.T. McDowell, Y. Yao, C. Wang, Y. Cui, A yolk-shell design for stabilized and scalable Li-ion battery alloy anodes, Nano letters 12 (2012) 3315–3321. [11] L.Y. Yang, H.Z. Li, J. Liu, Z.Q. Sun, S.S. Tang, M. Lei, Dual yolk-shell structure of carbon and silica-coated silicon for high-performance lithium-ion batteries, Sci. Rep. 5 (2015). [12] J. Graetz, C. Ahn, R. Yazami, B. Fultz, Highly reversible lithium storage in nanostructured silicon, Electrochemical and Solid-State Letters 6 (2003) A194–A197. [13] N. Liu, Z. Lu, J. Zhao, M.T. McDowell, H.-W. Lee, W. Zhao, Y. Cui, A pomegranateinspired nanoscale design for large-volume-change lithium battery anodes, Nat. Nano 9 (2014) 187–192. [14] R. Yi, F. Dai, M.L. Gordin, S. Chen, D. Wang, Micro-sized Si-C Composite with Interconnected Nanoscale Building Blocks as High-Performance Anodes for Practical Application in Lithium-Ion Batteries, Advanced Energy Materials 3 (2013) 295–300. [15] N. Liu, K. Huo, M.T. McDowell, J. Zhao, Y. Cui, Rice husks as a sustainable source of nanostructured silicon for high performance Li-ion battery anodes, Sci. Rep. 3 (2013). [16] J. Song, S. Chen, M. Zhou, T. Xu, D. Lv, M.L. Gordin, T. Long, M. Melnyk, D. Wang, Micro-sized silicon-carbon composites composed of carbon-coated sub-10 nm Si primary particles as high-performance anode materials for lithium-ion batteries, Journal of Materials Chemistry A 2 (2014) 1257–1262. [17] G. Hwang, H. Park, T. Bok, S. Choi, S. Lee, I. Hwang, N.-S. Choi, K. Seo, S. Park, A high-performance nanoporous Si/Al2O3 foam lithium-ion battery anode fabricated by selective chemical etching of the Al–Si alloy and subsequent thermal oxidation, Chemical Communications 51 (2015) 4429–4432.

[18] M. Li, J. Gu, X. Feng, H. He, C. Zeng, Amorphous-silicon@silicon oxide/ chromium/carbon as an anode for lithium-ion batteries with excellent cyclic stability, Electrochimica Acta 164 (2015) 163–170. [19] D.-H. Lee, H.-W. Shim, D.-W. Kim, Facile synthesis of heterogeneous Ni-Si@C nanocomposites as high-performance anodes for Li-ion batteries, Electrochimica Acta 146 (2014) 60–67. [20] H.K. Han, C. Loka, Y.M. Yang, J.H. Kim, S.W. Moon, J.S. Cho, K.-S. Lee, High capacity retention Si/silicide nanocomposite anode materials fabricated by high-energy mechanical milling for lithium-ion rechargeable batteries, Journal of Power Sources 281 (2015) 293–300. [21] M. Wu, J.E.C. Sabisch, X. Song, A.M. Minor, V.S. Battaglia, G. Liu, In Situ Formed Si Nanoparticle Network with Micron-Sized Si Particles for Lithium-Ion Battery Anodes, Nano Letters 13 (2013) 5397–5402. [22] J.S. Kim, W. Pfleging, R. Kohler, H.J. Seifert, T.Y. Kim, D. Byun, H.-G.W. Jung, J.K. Choi Lee, Three-dimensional silicon/carbon core–shell electrode as an anode material for lithium-ion batteries, Journal of Power Sources 279 (2015) 13–20. [23] P. Zhang, L. Wang, J. Xie, L. Su, C.a. Ma, Micro/nano-complex-structure SiOx– PANI–Ag composites with homogeneously-embedded Si nanocrystals and nanopores as high-performance anodes for lithium ion batteries, Journal of Materials Chemistry A 2 (2014) 3776–3782. [24] L. Su, J. Xie, Y. Xu, L. Wang, Y. Wang, M. Ren, Effect of pore lengths on the reduction degree and lithium storage performance of Mesoporous SiOx nanomaterials, Journal of Alloys and Compounds 663 (2016) 524–530. [25] N.-S. Choi, K.H. Yew, K.Y. Lee, M. Sung, H. Kim, S.-S. Kim, Effect of fluoroethylene carbonate additive on interfacial properties of silicon thin-film electrode, Journal of Power Sources 161 (2006) 1254–1259. [26] A. Magasinski, B. Zdyrko, I. Kovalenko, B. Hertzberg, R. Burtovyy, C.F. Huebner, T.F. Fuller, I. Luzinov, G. Yushin, Toward Efficient Binders for Li-Ion Battery SiBased Anodes: polyacrylic Acid, ACS Applied Materials & Interfaces 2 (2010) 3004–3010. [27] I. Kovalenko, B. Zdyrko, A. Magasinski, B. Hertzberg, Z. Milicev, R. Burtovyy, I. Luzinov, G. Yushin, A Major Constituent of Brown Algae for Use in HighCapacity Li-Ion Batteries, Science 334 (2011) 75–79. [28] M. Ashuri, Q. He, L.L. Shaw, Silicon as a potential anode material for Li-ion batteries: where size, geometry and structure matter, Nanoscale 8 (2016) 74– 103. [29] D. Wang, M. Gao, H. Pan, J. Wang, Y. Liu, High performance amorphousSi@SiOx/C composite anode materials for Li-ion batteries derived from ballmilling and in situ carbonization, Journal of Power Sources 256 (2014) 190– 199. [30] Y. Hwa, W.-S. Kim, S.-H. Hong, H.-J. Sohn, High capacity and rate capability of core–shell structured nano-Si/C anode for Li-ion batteries, Electrochimica Acta 71 (2012) 201–205. [31] T. Zhang, L. Fu, J. Gao, L. Yang, Y. Wu, H. Wu, Core-shell Si/C nanocomposite as anode material for lithium ion batteries, Pure and applied chemistry 78 (2006) 1889–1896. [32] T. Zhang, J. Gao, H.P. Zhang, L.C. Yang, Y.P. Wu, H.Q. Wu, Preparation and electrochemical properties of core-shell Si/SiO nanocomposite as anode material for lithium ion batteries, Electrochemistry Communications 9 (2007) 886–890. [33] L. Su, Y. Jing, Z. Zhou, Li ion battery materials with core-shell nanostructures, Nanoscale 3 (2011) 3967–3983. [34] L. Su, Z. Zhou, M. Ren, Core double-shell Si@SiO2@C nanocomposites as anode materials for Li-ion batteries, Chemical Communications 46 (2010) 2590– 2592. [35] H. Tao, L.-Z. Fan, W.-L. Song, M. Wu, X. He, X. Qu, Hollow core-shell structured Si/C nanocomposites as high-performance anode materials for lithium-ion batteries, Nanoscale 6 (2014) 3138–3142. [36] Z. Sun, X. Song, P. Zhang, L. Gao, Controlled synthesis of yolk-mesoporous shell Si@SiO2 nanohybrid designed for high performance Li ion battery, RSC Advances 4 (2014) 20814–20820. [37] Z. Sun, S. Tao, X. Song, P. Zhang, L. Gao, A Silicon/Double-Shelled Carbon YolkLike Nanostructure as High-Performance Anode Materials for Lithium-Ion Battery, Journal of The Electrochemical Society 162 (2015) A1530–A1536. [38] L. Su, J. Xie, Y. Xu, L. Wang, Y. Wang, M. Ren, Preparation and lithium storage performance of yolk-shell Si@void@C nanocomposites, Physical Chemistry Chemical Physics 17 (2015) 17562–17565. [39] D. Chen, X. Mei, G. Ji, M. Lu, J. Xie, J. Lu, J.Y. Lee, Reversible Lithium-Ion Storage in Silver-Treated Nanoscale Hollow Porous Silicon Particles, Angewandte Chemie International Edition 51 (2012) 2409–2413. [40] Y. Yao, M.T. McDowell, I. Ryu, H. Wu, N. Liu, L. Hu, W.D. Nix, Y. Cui, Interconnected silicon hollow nanospheres for lithium-ion battery anodes with long cycle life, Nano letters 11 (2011) 2949–2954. [41] X.-y Zhou, J.-j Tang, J. Yang, J. Xie, L.-l Ma, Silicon@carbon hollow core?shell heterostructures novel anode materials for lithium ion batteries, Electrochimica Acta 87 (2013) 663–668. [42] L. Chen, Y. Liu, M. Ashuri, C. Liu, L.L. Shaw, Li2S encapsulated by nitrogen-doped carbon for lithium sulfur batteries, Journal of Materials Chemistry A 2 (2014) 18026–18032. [43] M. Su, Z. Wang, H. Guo, X. Li, S. Huang, W. Xiao, L. Gan, Enhancement of the Cyclability of a Si/Graphite@ Graphene composite as anode for Lithium-ion batteries, Electrochimica Acta 116 (2014) 230–236. [44] S. Murugesan, J.T. Harris, B.A. Korgel, K.J. Stevenson, Copper-Coated Amorphous Silicon Particles as an Anode Material for Lithium-Ion Batteries, Chemistry of Materials 24 (2012) 1306–1315.

M. Ashuri et al. / Electrochimica Acta 215 (2016) 126–141 [45] T. Cetinkaya, M. Uysal, M.O. Guler, H. Akbulut, A. Alp, Improvement cycleability of core–shell silicon/copper composite electrodes for Li-ion batteries by using electroless deposition of copper on silicon powders, Powder Technology 253 (2014) 63–69. [46] D. Chen, R. Yi, S. Chen, T. Xu, M.L. Gordin, D. Wang, Facile synthesis of graphene–silicon nanocomposites with an advanced binder for highperformance lithium-ion battery anodes, Solid State Ionics 254 (2014) 65– 71. [47] X. Li, P. Meduri, X. Chen, W. Qi, M.H. Engelhard, W. Xu, F. Ding, J. Xiao, W. Wang, C. Wang, J.-G. Zhang, J. Liu, Hollow core-shell structured porous Si-C nanocomposites for Li-ion battery anodes, Journal of Materials Chemistry 22 (2012) 11014–11017. [48] H.-C. Tao, X.-L. Yang, L.-L. Zhang, S.-B. Ni, Double-walled core-shell structured Si@SiO2@C nanocomposite as anode for lithium-ion batteries, Ionics 20 (2014) 1547–1552. [49] K. Zheng, A. Solodovnyk, W. Li, O.-M. Goudouri, C. Stähli, S.N. Nazhat, A.R. Boccaccini, Aging Time and Temperature Effects on the Structure and Bioactivity of Gel-Derived 45S5 Glass-Ceramics, Journal of the American Ceramic Society 98 (2015) 30–38. [50] B. Shokri, M.A. Firouzjah, S. Hosseini, FTIR analysis of silicon dioxide thin film deposited by metal organic-based PECVD, Proceedings of 19th international symposium on plasma chemistry society, Bochum, Germany, 2009. [51] B.J. Saikia, G. Parthasarathy, Fourier transform infrared spectroscopic characterization of kaolinite from Assam and Meghalaya Northeastern India, Journal of Modern Physics 1 (2010) 206. [52] J. Liang, S. Chen, M. Xie, Y. Wang, X. Guo, X. Guo, W. Ding, Expeditious fabrication of flower-like hierarchical mesoporous carbon superstructures as supercapacitor electrode materials, Journal of Materials Chemistry A 2 (2014) 16884–16891.

141

[53] K. Crosby, L.L. Shaw, Dehydriding and re-hydriding properties of high-energy ball milled LiBH4 + MgH2 mixtures, International Journal of Hydrogen Energy 35 (2010) 7519–7529. [54] R. Takahashi, S. Sato, T. Sodesawa, M. Kawakita, K. Ogura, High Surface-Area Silica with Controlled Pore Size Prepared from Nanocomposite of Silica and Citric Acid, The Journal of Physical Chemistry B 104 (2000) 12184–12191. [55] C. Shen, L.L. Shaw, L.C. Zhang, E.L. Thomas, Porous gadolinia-doped ceria with adjustable pore sizes using PI-b-PEO copolymer as the structure-directing agent, J. Sol-Gel Sci. Technol. 63 (2012) 72–84. [56] H. Izutsu, F. Mizukami, T. Sashida, K. Maeda, Y. Kiyozumi, Y. Akiyama, Effect of malic acid on structure of silicon alkoxide derived silica, Journal of NonCrystalline Solids 212 (1997) 40–48. [57] N. Yan, F. Wang, H. Zhong, Y. Li, Y. Wang, L. Hu, Q. Chen, Hollow Porous SiO2 Nanocubes Towards High-performance Anodes for Lithium-ion Batteries, Sci. Rep. 3 (2013). [58] T. Cetinkaya, M. Uysal, M.O. Guler, H. Akbulut, Developing lithium ion battery silicon/cobalt core-shell electrodes for enhanced electrochemical properties, International Journal of Hydrogen Energy 39 (2014) 21405–21413. [59] T. Cetinkaya, M. Uysal, H. Akbulut, Electrochemical performance of electroless nickel plated silicon electrodes for Li-ion batteries, Applied Surface Science 334 (2015) 94–101. [60] J.J. Wu, W.R. Bennett, Fundamental investigation of Si anode in Li-Ion cells, Energytech, IEEE, 2012 (2012) 1–5. [61] H. Yao, G. Zheng, W. Li, M.T. McDowell, Z. Seh, N. Liu, Z. Lu, Y. Cui, Crab Shells as Sustainable Templates from Nature for Nanostructured Battery Electrodes, Nano Letters 13 (2013) 3385–3390. [62] H. Wu, G. Chan, J.W. Choi, I. Ryu, Y. Yao, M.T. McDowell, S.W. Lee, A. Jackson, Y. Yang, L. Hu, Y. Cui, Stable cycling of double-walled silicon nanotube battery anodes through solid-electrolyte interphase control, Nat. Nano 7 (2012) 310–315.