Molten salt electrolytic synthesis of silicon-copper composite nanowires with enhanced performances as lithium ion battery anode

Journal of Alloys and Compounds 751 (2018) 307e315

Contents lists available at ScienceDirect

Journal of Alloys and Compounds journal homepage: http://www.elsevier.com/locate/jalcom

Molten salt electrolytic synthesis of silicon-copper composite nanowires with enhanced performances as lithium ion battery anode Zhongren Zhou a, b, Yingjie Zhang a, c, **, Yixin Hua c, d, Peng Dong a, c, Yan Lin c, d, *, Mingli Xu a, c, Ding Wang a, c, Xue Li a, c, Lina Han a, b, Jianguo Duan a, c a

National and Local Joint Engineering Laboratory for Lithium-ion Batteries and Materials Preparation Technology, Key Laboratory of Advanced Battery Materials of Yunnan Province, Kunming University of Science and Technology, Kunming 650093, China Faculty of Material Science and Engineering, Kunming University of Science and Technology, Kunming 650093, China c Faculty of Metallurgical and Energy Engineering, Kunming University of Science and Technology, Kunming 650093, China d State Key Laboratory of Complex Nonferrous Metal Resources Clean Utilization, Kunming University of Science and Technology, Kunming, 650093, China b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 February 2018 Received in revised form 9 April 2018 Accepted 10 April 2018 Available online 11 April 2018

A facile metallurgical method named molten salt electrolysis was introduced to prepare silicon nanowires (Si NWs) and Cu9Si/Si composite nanowires that were used as the anode materials for lithium-ion batteries (LIBs). Simply utilizing SiO2 and Cu/SiO2 mixture as the raw materials, by tailoring voltage into the molten salt chamber, SiO2 was constantly deoxidized to form Si flexible nanowires with 1e3 mm in length and 100e200 nm in width. Cu ingredient further catalyzed the growth of silicon nuclei to generate straight wires with longer 10 mm and narrower 60e100 nm. Benefiting from the compositional and structural advantages, the Cu9Si/Si nanocomposite anode exhibited a better capacity retention that delivered a specific capacity of 601.3 mAh$g
1 at 200 mA g
1 after 200 cycles and 398.8 mAh$g
1 at 500 mA g
1 after 500 cycles. Notably Cu9Si/Si NWs anode with enhanced rate capability was also obtained, with a specific capacity of 747.6 mAh$g
1, 550.6 mAh$g
1 and 412.2 mAh$g
1 at the current density of 200 mA g
1, 500 mA g
1 and 1000 mA g
1, respectively. © 2018 Elsevier B.V. All rights reserved.

Keywords: Silicon-copper composite Nanowire Molten salt electrolysis Lithium-ion batteries

1. Introduction Lithium-ion batteries have become the key-enabling commodity in energy storage fields due to their high energy and long-term cycles. Presently, the anode material served in commercial LIBs is made of graphite, but the practical low capacity (less than 372 mAhg
1) limits its marketing demands [1e3]. Plenty of researches are needed to replace the traditional graphite with new anode materials that contain a relatively higher capacity, energy and power density. Nowadays, there has existed considerable alternative anode materials, including Sn with the theoretical gravimetric specific capacities of 994 mAhg
1 [4], Si with 4200 mAhg
1 [5e7], transition oxides such as Co3O4 [8e10], CuO

* Corresponding author. Present address: State Key Laboratory of Complex Nonferrous Metal Resources Clean Utilization, Kunming University of Science and Technology, Kunming, 650093, China. ** Corresponding author. Present address: Faculty of Metallurgy and Energy Engineering, Kunming University of Science and Technology, Kunming 650093, China. E-mail addresses: [email protected] (Y. Zhang), [email protected] (Y. Lin). https://doi.org/10.1016/j.jallcom.2018.04.128 0925-8388/© 2018 Elsevier B.V. All rights reserved.

[11], Fe2O3 [12], TiO2 with much higher gravimetric specific capacities [13] and spinel Li4Ti5O12 [14e17], Na2Li2Ti6O14 [18] and LiNbO3 [19]. Amongst them, silicon is deemed to be one of the most promising anode materials for LIBs because of the almost tenfold theoretical capacity higher than that of graphite, and exhibits low discharge voltage (vs. Liþ/Li) during the lithiation/delithiation process. Moreover, the rich abundance of silicon presents itself as an ideal alternative candidate for use [20]. However, constant lithium ion insertion/extraction from silicon is unavoidable accompanied by the huge volumetric change of silicon anode (>300%), leading to the formation of unstable solid electrolyte interphase (SEI), structure disintegration, particle pulverization, poor electrical contact among electrode, and finally resulting in the fast capacity fading and unsatisfactory rate performance [21]. In order to overcome the above issues, many works have been attempted based on the nanocrystallization and recombination solution. First, minimizing the particle size to nanoscale such as nanowires, nanocomposites, thin films, nanobelts and nanotubes; these types of nanosized structures provides individual geometries to help accommodate the large volumetric expansion, alleviate

308

Z. Zhou et al. / Journal of Alloys and Compounds 751 (2018) 307e315

pulverization of electrode and promote the lithium ions transmission into and out of the anode matrix [22,23]. Second, manufacturing silicon and carbonaceous agent or Si-Metal binary composite to increase the conductivity of materials [24]. Third, covering silicon nanoparticles by metal oxide layer [25]. Mostly, the Si-Metal binary composite, such as silicon-copper nanocomposite, is suggested to be an efficient way by forming an electrochemically inactive matrix, with the advantages of increasing the total conductivity of anode and mitigating the mechanical stress. For instance, using chemical vapor deposition (CVD) method, Chen et al. fabricated copper coated Si nanowire with a capacity retention of 2216 mAh$g
1 at 3.1 A g
1 over 30 cycles [26]. With the thermal reduction method, Zhou et al. synthesized Cu3Si@Si coreshell nanoparticles showed a capacity of 903.6 mAh$g
1 at 2 A g
1 over 400 cycles [27]. Jae-Young Woo et al. synthesized Cu3Si/Si nanoparticles showed the discharge capacity of 3036.4 mAh$g
1 with a coulombic efficiency of 90.49% at the first cycle [28]. By the high-energy ball milling method, Yoon et al. prepared Cu3Si doped layers with a stable capacity of 850 mAh$g
1 for 30 cycles at 100 mA g
1 [29]. However, there might still exist difficulties for preparing industrially viable copper-silicon nanoparticles when considering the cost and technique. For example, the CVD method probably seems to be complicated because this technique presents pricy reactant dependence, low productivity and the limitation of complex equipment [30]. The ball milling process could unavoidable give rise to a highly metastable state and crystal defects [31]. Therefore, developing a facile strategy for fabricating Si-based nanomaterial continues to be a hot spot. Herein, a novel synthetic route is proposed for the production of silicon and copper-silicon composite in nanoscale with molten salt electrolysis. This method is well known as the Fray-Farthing-Chen (FFC) Cambridge electro-deoxidization process [32], and has been operating at the Metalysis company (http://www.metalysis.com/). Generally, this method applies to the direct conversion of metallic oxides into metals and alloys in solid state, such as Cr [33], Ge [34], FeTi [35], Ti5Si3 [36,37]. This method provides an effective metallurgical process that the porous mineral precursor is simply ‘metallized’ in solid state and bulk quantities of meals are obtained by simply increasing the number of oxide once used, offering a potential benefit in both case of mass production and economics. Moreover, silicon nanowires have been prepared successfully with the aid of molten salt [38e40]. These investigations induce a fine way to produce silicon-based alloys in controllable nanosized architectures. Herein, we proposed a novel safe synthetic process for silicon and copper-silicon composite nanowires applied as the anode materials for LIBs through the direct electrolysis of silica and copper mixture (Cu/SiO2) in molten salt. The conversion of nascent bulk oxide precursor into silicon and silicide compound nanoparticles was investigated. During the reduction, when the injected cell voltage was much higher than the decomposition voltages thermodynamically, oxygen from SiO2 could be constantly removed via reaction (1). After Si was produced, it could spontaneously react with Cu to form Cu-Si composite at given working temperature.

SiO2 ¼ Si þ O2 ðgÞEq ð700 CÞ ¼ 1:90V

C103843) were blended by molar ratio of 1:1 to serve as the Cu/SiO2 precursor and SiO2 powders were solely served as the SiO2 precursor. The mass of the two samples was kept the same. NH4HCO3 (Aladdin, A110536) of 40% by mass was served as the pore-forming agent and added into the mixed powders. After that, the powders were pressed into pellets (diameter: 14 mm; thickness: ~4 mm), and sintered at 500  C in argon atmosphere for 3 h. The precursor was further sandwiched between two porous molybdenum foils and served as the cathode, and a graphite rod (diameter: 6 mm; length: 80 mm, Aladdin, G103921) was used as the anode. Molten salt consisting equimolar CaCl2 (Aladdin, C110766)-NaCl (Aladdin, C111549) in total 210 g was served as the electrolyte and packed in an Al2O3 crucible (inner diameter: 57 mm; external diameter: 63 mm; depth: 75 mm, Aladdin, A2590) chamber. The schematic of experimental apparatus was shown in Fig. 1. To obtain silicon and copper-silicon composite, electrolysis of SiO2 and Cu/SiO2 pellets was conducted under a constant cell voltage of 2.4 V for 10 h at 700  C. In all electrolytic occasions, the chamber was sealed in a vertical quartz tube closed at one end and continuously purged with an argon flow. After electrolytic terminals, the cathode was cooled down with the cell to room temperature under a flow of argon. The products were removed from the chamber, washed in 0.5% HCl solution, deionized water and alcohol, and dried under vacuum at 80  C for 1 h.

2.2. Material characterization Rigaku MiniFlex II X-ray diffractometer was used to collect X-ray powder diffraction pattern of the sample in the 2q range of 10e90 at a rate of 4 /min with a Cu Ka radiation source (l ¼ 0.154178 nm) operating at 40 kV, 30 mA. The samples were analyzed by X-ray photoelectron spectroscopy (XPS) for the measurement of the binding energy by the use of a PHI Quantum 2000 scanning ESCA Microprobe equipped with an Al Ka1,2 X-ray radiation source. The survey scans were used a source with power and voltage of 50 W and 15 kV, respectively. Survey scans and high-resolution scans of the Si 2p, Cu 2p, O 1s, C 1s, energy spectra were taken from samples to identify the compounds present on the surface. The spectra were calibrated to the carbon peak at 284.8 eV. The morphology and microstructure of samples were detected by scanning electron microscope equipped a FEI Nova NanoSEM 450 field emission scanning electron microscope (FESEM) at 15 kV and transmission electron microscopy (TEM, JEM2010-HT). The specific surface area and the pore structure of samples were examined by nitrogen sorption measurements at 77 K (Micromeritics ASAP2020).

(1)

2. Experimental 2.1. Electrochemical synthesis of Si and Cu-Si nanocomposite High purity SiO2 powders (particle size: ~1 mm, Aladdin, S104600) and Cu powders (particle size: ~0.5 mm, Aladdin,

Fig. 1. Schematic diagram of electrolysis apparatus.

Z. Zhou et al. / Journal of Alloys and Compounds 751 (2018) 307e315

2.3. Assembly of coin cells and electrochemical testing The electrochemical performance of samples was measured within a lithium half-cell system. The system assembled the testing electrode (surface area: 1.13 cm2), separator, lithium foil and electrolyte to form the standard CR2016 coin cell in an argon glove box. Pasting the active material (the reduced Si and Cu-Si composite) slurry, conductive carbon blacks (super P Li, Timcal Co.), styrene butadiene rubber (SBR) binder, sodium carboxymethyl cellulose (CMC, Aladdin LR) by the mass fraction of 80: 10: 5: 5 and a solvent of deionized water for 3 h and further casing on a Cu foil to constitute the working electrode. The working electrode was dried overnight at 110  C in a vacuum oven. After that, the foil was cut to disk for coin cell assembling. Lithium metal was used as the counter electrode, and the porous polypropylene (Celgard 2500) separator was loaded to separate the electrodes. The electrolyte consisted of 1 M LiPF6 in ethylene carbonate (EC), dimethyl carbonate (DMC) and ethyl methyl carbonate (EMC) with a volume ratio of 1:1:1. After that, 5 vol% fluoroethylene carbonate (FEC) was added into the electrolyte. The assembled coin cell was kept in a thermostatic chamber to soak in the electrolyte for 6 h. Galvanostatic discharge/charge cycling performance was tested in a Land BTI-40 (Wuhan, China) cell test system with the voltage range of 0.01e1.5 V. The cycling performances were also tested with current densities of 200e500 mA g
1. Notably, the first three cycles of cells were tested at a low current density of 50 mA g
1 to sufficiently activate the materials. The rate capabilities of samples were texted at current densities from 200 mA g
1 to 1000 mA g
1. The cyclic voltammetry (CV) measurement was conducted at constant scanning rate of 0.2 mV s
1 in the voltage window of 0.01e1.5 V using Metrohm Autolab PGSTAT302 N (Netherlands). Electrochemical impedance spectroscopy (EIS) measurements were tested with amplitude of 10 mV over frequency range of 100 kHZ to 0.01 Hz.

3. Result and discussion 3.1. Product characterization To clarify the electrochemical behavior of reduction, constant voltage electrolysis of SiO2 as well as the Cu/SiO2 blocks under 2.4 V in equimolar CaCl2-NaCl molten salt is carried out for 10 h. The current varied with time curves are recorded in detail, as shown in Fig. 2(a). Tendencies of the current decayed with time herein are familiar to others [41], that the current is constantly declined with

309

time, which implies that the reduction kinetics is fast initially but slows down till to the electrolytic end. Typically, as electricity is appended into the cell, the cathodic reaction occurs at the first contacting interface between the Mo wire and raw materials (SiO2 or Cu/SiO2 mixture), and hereafter proceeds inside of pellets with the passage of time, which can be well explained by the propagation of conductor/oxides/molten salt three-phase interline model. When considering the electrolysis of single SiO2, electrochemical reaction will firstly take place at the fresh Mo foils/SiO2/electrolyte three-phase interline on the surface of the pellet through the reaction of oxygen removal (SiO2þ4e
¼ Siþ2O2
). Once silicon is electrochemically formed, the nuclei is believed to be crystallized to form crystalline state. Constantly, electrolysis can be proceeded at the newly Si/SiO2/electrolyte interline inside. But the current values, in view of the smaller interfaces inside of the porous pellet, will be decreased, resulting in a sharp decline of the current till to the end. For example, the current recorded in reduction of single SiO2 is decreased from the initial 1093 mAe265.2 mA in the original 1 h of electrolytic duration. However, when introducing Cu powders into cathode, due to its terrific conductivity, e
uptake is much easier from these metals and electrolysis could be proceeded at the Cu/SiO2/electrolyte interlines, resulting in the dramatic increase of current from the beginning 2001 mAe1088 mA in 1 h. The current later encounters a slightly increase in 1e3 h and 6e6.5 h, implying the deoxidization is promoted by the addition of Cu. In such case, it is induced that once SiO2 is deoxidized at the Cu/SiO2/electrolyte interlines, the produced silicon will just ‘dissolve’ into the connected Cu matrix via the diffusing process spontaneously to form Cu-Si nanocomposite at high operating temperature. Upon constant electrolysis of individual SiO2 with/without Cu powers at 2.4 V for 10 h, the products are evidenced to be Si and Cu9Si composite according to the XRD analysis. From Fig. 2(b) can be seen that the diffraction peaks of SiO2 all disappear after electrolysis, indicating a complete reduction of SiO2. The diffraction peaks are assigned to the (111), (200) and (220) lattice planes, which reveal the formation of cubic unit cells with the space group Fd3m of Si whilst all the other peaks are well indexed on cubic unit cells with the space group  Fm3m of Cu9Si. Upon the electrolysis of the mixed oxides, SiO2 is evidenced to be deoxidized and the reduced Si will react with Cu to transform into Si and Cu9Si composite without noticeable signals of phase segregation. Fig. 3 shows XPS spectra of Si 2p and Cu 2p during the direct electrolysis of Cu/SiO2 pellet. Considering the spectra of raw precursor before reduction, there is a strong peak centered at 102.5 eV in Fig. 3(c), which can be assigned to Si4þ of the raw SiO2. After 1 h

Fig. 2. Current-time curves recorded during 10 h of electrolysis of SiO2 and Cu/SiO2 precursors and XRD patterns of products under 2.4 V electrolysis for 10 h at 700  C.

310

Z. Zhou et al. / Journal of Alloys and Compounds 751 (2018) 307e315

Fig. 3. Simulation of the high-resolution XPS core spectrum of the Si 2p and Cu 2p level in reduction: XPS spectrum of Si 2p after reduction for (a) 10 h and (b) 1 h, and (c) the raw precursor; (d) XPS spectrum of Cu 2p after reduction for 10 h and (e) before reduction.

of electrolysis, a strong Si 2p peak noticed around at 99 eV is resolved into two peaks centered at 98.8 eV and 99.5 eV, while another peak is observed centered at around 101.9 eV. The overall peaks are further fitted by the combination of respective binding energies for the different oxidation state of Si (Cu9Si: B.E. ¼ 98.8 eV, Si: B.E. ¼ 99.5 eV and SiOx: B.E. ¼ 101.9 eV). Meanwhile, the negative shift of metallic Si2p XPS spectra toward low binding energies indicates that part of the reduced Si is alloyed with Cu to form the metallic Cu-Si alloy [42,43]. After 10 h of reduction, an obvious increase in the peak intensity of Si 2p core level spectra for the metallic Si can be expected, as shown in Fig. 3(a). For the core spectrum of the Cu 2p level in Fig. 3(d)~(e), no obvious shift is found before and after reduction, expect for a hump centered at 952.7 eV in the raw precursor sample, which can be attributed to the Cu-O bond that is mainly due to the easily oxidized nano-Cu in air. Field-emission scanning electron microscopy (FESEM) and highresolution transmission electron microscopy (HRTEM) are utilized to characterize the morphology of products, as shown in Fig. 4. Upon electrolysis of individual SiO2 and Cu/SiO2 precursors, brown silicon powders as well as light red copper-silicon composite powders are produced, as shown in Fig. 4(a) and (f). The EDS analysis in Fig. 4(c) and (h) also confirm the purity of Si and Cu-Si that the corresponding EDS spectrum indicates the Cu/Si atomic ratio of 0.95. As shown in Fig. 4(a) and (b), the as-obtained Si exhibits flexible nanowires that are measured almost 1e3 mm long and 100e200 nm width. Moreover, the interconnected structures in Fig. 4(a)~(b) reveal the accumulation of silicon nuclei to form the continuous network. This is desirable for maintaining a robust mechanical support to alleviate stress development [44]. When electrolyzing Cu/SiO2 mixture at the same condition, the Cu-Si composite product shows straight nanowires that are measured to the longer 10 mm in length and narrower 60e100 nm in width. Compared to Si curving nanowires, the Cu-Si composite straight nanowires exhibit more regular shapes that the structural transition mechanism is similar to the Vapor-Liquid-Solid mechanism (VLS process) [45,46], that is the metallic copper-catalyzes the growth of Si: once silicon is produced at the Cu/SiO2/electrolyte interfaces, it will ‘dissolve’ into Cu matrix to form Cu-Si alloy (Cu9Si in present) whilst the continuous ‘dissolution’ of Si will induce the precipitation of it due to the limited dissolubility of Si in Cu matrix, and will further cause the oriented growth of Cu-Si binary composite. Consistent with XRD analysis, the HRTEM image in Fig. 4(g) shows the Si nanowires are crystalline with clear lattice fringes corresponded to a Si (111) lattice with a d-space of ~0.314 nm. Meanwhile, from Fig. 4(h), the d-spacing value of the lattice planes

is calculated ~0.307 nm, that corresponds well with that of the (111) planes of the cubic Cu9Si crystal. It is believed that these nanowires are helpful to accommodate volumetric expansion for the lithiation [47]. Moreover, with the addition of copper, the Cu9Si nanowires exhibiting a largely enhanced conductivity could be expected, which is another important facilitation for practical LIB applications. According to the N2 sorption plots shown in Fig. 5(a), the Brunauer-Emmett-Teller (BET) specific surface area of Cu-Si nanowires is calculated to be 32.1 m2g
1, slightly higher than that of Si NWs (22.1 m2g
1). The void space of sample is evaluated based on the BJH pore-size distributions from the desorption data, as shown in Fig. 5(b). The results suggest that Si NWs deliver a relative smaller interspace since there exists a pore size distribution around 1e10 nm whilst Cu-Si NWs exhibit some kind of complex mesopores with size distributions around 2e42 nm, implying a larger interspace in Cu-Si NWs. The enhanced structures of Cu-Si nanocomposite sample could contribute to providing considerable interspace for Liþ transport, which is favorable for buffering the volume change in discharge/charge process. Thus, it can be deduced that both the larger specific surface areas and enhanced void structures of Cu-Si nanowires will shorten the mass transfer process and are beneficial to the electrochemical performance of the anode. 3.2. Electrochemical performance Fig. 6(a) shows the current-voltage curves of Cu-Si nanocomposite anode from the 1st to 5th cycles at 0.2 mV s
1. An oppressive broad peak at around 0.75 V is observed in the first scan but disappears in the following scans, which can be ascribed to the formation of SEI [48]. As voltage is scanned negatively, the sharp peak appeared below 0.1 V represents the Liþ insertion into the crystallized Si. In the subsequent cycles, the peak below 0.1 V is replaced by the peak at 0.16 V, mainly ascribed to the Liþ insertion to the complete amorphization of crystalline Si [49,50]. In the anodic part from the 1st to 5th cycle, predominant peaks located at 0.33 V and 0.53 V correspond to the lithium extraction potential, which can be assigned to the delithiation reactions from Li-Si alloy and the formation of amorphous Si [51]. In addition, due to the higher reaction voltage of 2.5 V and 2.7 V (vs. Liþ/Li) for the Cu2þ/ CuO reaction, the Cu component within Cu-Si composite will not participate even contribute to the lithium-ion storage under the voltage range of 0.01e1.5 V [8,29]. Fig. 6(b)~(c) show the respective 1st, 2nd and 50th cycles' voltage varied with the capacity of Si and Cu-Si electrodes in the

Z. Zhou et al. / Journal of Alloys and Compounds 751 (2018) 307e315

311

Fig. 4. Morphologies of Si: (a)~(b) SEM images with the digital image inside, (c) EDS analysis from (b), (d) TEM and (e) HRTEM from (b); morphologies of Cu-Si nanocomposite: (f) ~(g) SEM images, (h) EDS analysis from (g), (i) TEM and (j) HRTEM from (g).

Fig. 5. (a) N2 adsorption-desorption isotherms at 77 K of Si and Cu-Si nanowires, (b) pore size distributions from desorption data.

voltage window of 0.01e1.5 V at 200 mA g
1. During the first cycle, the Si and Cu-Si electrodes exhibit sloping profiles upon lithiation and following delithiation reactions. The first discharge capacity of Si and Cu-Si is 2248.7 mAh$g
1, 1120.8 mAh$g
1 and the charge capacity is 1519.8 mAh$g
1, 990.6 mAh$g
1 with the respective initial coulombic efficiency of 67.6% and 88.4%. Though the single Si NWs anode exhibits more capacity than that of Cu-Si

nanocomposite has, when extending the cycling number from the 2nd to thereafter 50th, the Si NWs show quick capacity degradation of ~67.8% whilst the Cu-Si anode shows a much lower capacity degradation of ~10.3%. Even after 50 cycles, the capacity of Cu-Si nanocomposite anode is still above 767.2 mAh$g
1, indicating the more excellent performance than that of single Si NWs. This is reasonable because the formation of regular Cu-Si composite

312

Z. Zhou et al. / Journal of Alloys and Compounds 751 (2018) 307e315

Fig. 6. Electrochemical performance of Cu-Si anode materials between 0.01 and 1.5 V: (a) cyclic voltammetry curves scanned at 0.2 mV s
1; (b) and (c) voltage profiles at 200 mA g
1; (d) cycling performances and coulombic efficiencies at 200 mA g
1 and (e) of Cu-Si anode at 500 mA g
1; (f) rate capability. Notably, the first three cycles of cells were tested at a low current density of 50 mA g
1 to sufficiently activate the materials.

nanowires can provide enhanced ionic and electronic conductivity and mechanical buffering effects against volume expansion. Fig. 6(d) exhibits the cycling performances of the two samples at a current density of 200 mA g
1. All the two samples exhibit capacity fading. But a higher capacity retention of 55.6% after 100 cycles and of 51% after 200 cycles reveal the better cyclic property of Cu-Si nanocomposite sample and the single Si NWs electrode shows extremely poor cycling stability fading from the initial 2248.7 mAh$g
1 to 405.7 mAh$g
1 after merely 100 cycles (only 18% capacity retention). When cycling at a higher current rate 500 mA g
1, the Cu-Si nanocomposite anode could also exhibit an excellent performance of about 398.8 mAh$g
1 after 500 cycles, as shown in Fig. 6(e). The results further prove that Cu-Si

nanocomposite anode possesses outstanding cycling stability, which is ascribed to the copper component enhancing the conductivity of sample and the specific architectural nanowires adjusting structure stress upon repeated insertion and extraction of Liþ. The rate capability of the two samples were tested at various rates (200 mAh$g
1 ~ 1000 mAh$g
1) between 0.01 V and 0.15 V, as shown in Fig. 6(f). It can be seen that the Si NWs electrode exhibits the highest discharge capacity of the initial ~2261.8 mAh$g
1 at 200 mA g
1 but encounters a severe capacity drop in the first 20 cycles and delivers a limited discharge capability of ~200 mAh$g
1 at 1000 mA g
1. The worse performance can be improved by the addition of Cu. Firstly, when considering the Liþ insertion, Cu

Z. Zhou et al. / Journal of Alloys and Compounds 751 (2018) 307e315

313

Fig. 7. Nyquist plots and equivalent circuit model of (a) Si and (b) Cu-Si anodes before and after discharging at 200 mA g
1.

component with high conductivity provides electrons to fast counteract the Liþ-insertion-induced charge imbalance; secondly, upon the Liþ desertion, the metallic Cu phase will facilitate the electron collection and transport. Both of the mentioned cases are beneficial for improving the electrochemical performance. As a result, the discharge capacities of Cu-Si nanocomposite sample are increased to ~748, 550.6, 412.2 mAh$g
1 respectively at current densities of 200, 500, 1000 mA g
1, and the specific capacity can be readily restored to 739.5 mAh$g
1 when the current density is set back to 200 mA g
1. In order to further compare the electrochemical performance of Si and Cu-Si composite NWs, electrochemical impedance

spectroscopy tests were conducted as half cells were in the fully delithiated state at 200 mA g
1, as shown in Fig. 7. The equivalent circuits of electrodes inside the figure are represented to produce the fitted-model data. From Fig. 7, it can be seen that these circuits consist of serial connections involving Rel, Rsei/CPEsei, Rct/CPEdl and W that are related to the ionic resistance of the electrolyte, SEI resistance, SEI capacitance, charge transfer resistance at the electrode/electrolyte interface, double layer capacitance and Warburg diffusion impedance, respectively. By curve-fitting the Nyquist plots with this equivalent circuit model, the Nyquist plots compose of a suppressed semicircle in high frequency region, an obvious semicircle in medium frequency region and a visible uphill line in

Fig. 8. Surface and cross-sectional SEM images of Si and Cu-Si anodes: (a)~(d) before and (e)~(f) after discharging at 200 mA g
1.

314

Z. Zhou et al. / Journal of Alloys and Compounds 751 (2018) 307e315

low frequency region, where the semicircles are related to the SEI resistance (Rsei), charge transfer resistance (Rct) and the impedance of Liþ diffusion, respectively. From Fig. 7(a), it can be seen a dramatic impedance increase of the Si NWs anode just after 100 cycles at 200 mA g
1, the charge transfer resistance of which is calculated to be 100.8 U before cycling but increases to almost 165.1 U after cycling. The result implies a significant structure collapse of Si during cycling, leading to an unstable electrochemical stability. Conversely, the Cu-Si NWs anode in Fig. 7(b) exhibits a more stable state during the discharge/charge process because the electrode proves a smaller resistance increase. The charge transfer resistance is calculated 60.6 U at the initial stage and increases to 71.7 U after cycling. Comparison to the values of single Si NWs anode, the resistance transfer resistances are smaller before and after cycling at the same discharging regime. Owing to the more stable structure of Cu-Si nanowires as well as the improved conductivity of the composite, the Cu-Si nanowires anode are more suitable for LIBs storage materials. The surface and cross-sectional morphologies of Si NWs and CuSi nanocomposite anodes before and after cycling at 200 mA g
1 are observed and compared, as shown in Fig. 8. From Fig.8(a) ~ (g) can be seen that after cycling, the Si anode exhibits severe particle reunions and dramatic volumetric expansion with the increased thickness from 8.26 mm to 20.8 mm (over 252%). However, the modified Cu-Si composite shows the reduction of volume expansion particularly. After cycling, the thickness of sample is only increased from 8.6 mm to 12.6 mm (about 146.5%). Such comparative morphologies of Cu-Si anode further confirm that the volume expansion and particle reunions are restrained effectively by the unique structural Cu-Si nanocomposite. 4. Conclusion In present paper, silicon and Cu-Si nanocomposite (Cu9Si/Si) anode materials for LIBs are synthesized successfully via the molten salt electrolysis method that simply utilizes SiO2 and Cu/SiO2 as cathode and graphite as anode under constant cell voltage of 2.4 V at 700  C in equimolar CaCl2-NaCl melts. During the reduction, SiO2 is constantly deoxidized to produce silicon, leading to the formation of Si flexible nanowires with 1e3 mm in length and 100e200 nm in width. The growth of silicon is catalyzed by copper ingredient to produce straight Cu-Si composite nanowires with longer 10 mm and narrower 60e100 nm. When evaluated as the anode, the Si NWs manifest a quick capacity degradation of 67.8% from the 2nd to thereafter 50th at 200 mA g
1 and deliver a discharge capability of ~200 mAh$g
1 at 1000 mA g
1. In contrast, the Cu-Si nanowires anode exhibits the better cycling stability and rate performance, and delivers a specific capacity of 601.3 mAh$g
1 at 200 mA g
1 after 200 cycles and 398.8 mAh$g
1 at 500 mA g
1 after 500 cycles. The modified electrochemical performance is attributed to the enhanced conductivity and the straight nanowire structures of Cu-Si nanocomposite. Acknowledgement The authors acknowledge the financial support of the National Natural Science Foundation of China (Project Nos. 51764029, 51504112, 51604136, 51604132) and the Provincial Natural Foundation of Yunnan (Project No. 2017FB085). References [1] L. Hu, N. Liu, M. Eskilsson, G. Zheng, J. McDonough, L. Wågberg, Y. Cui, Siliconconductive nanopaper for Li-ion batteries, Nanomater. Energy 2 (2013) 138e145.

[2] B. Luo, T. Qiu, D. Ye, L. Wang, L. Zhi, Tin nanoparticles encapsulated in graphene backboned carbonaceous foams as high-performance anodes for lithium-ion and sodium-ion storage, Nanomater. Energy 22 (2016) 232e240. [3] H. Yin, Q.W. Li, M.L. Cao, W. Zhang, H. Zhao, C. Li, K.F. Huo, M.Q. Zhu, Nanosized-bismuth-embedded 1D carbon nanofibers as high-performance anodes for lithium-ion and sodium-ion batteries, Nano Research 10 (2017) 2156e2167. [4] T. Li, X. Li, Z. Wang, H. Guo, A short process for the efficient utilization of transition-metal chlorides in lithium-ion batteries: a case of Ni0.8Co0.1Mn0.1O1.1 and LiNi0.8Co0.1Mn0.1O2, J. Power Sources 342 (2017) 495e503. [5] U. Kasavajjula, C. Wang, A.J. Appleby, Nano-and bulk-silicon-based insertion anodes for lithium-ion secondary cells, J. Power Sources 163 (2007) 1003e1039. [6] M. Su, H. Wan, Y. Liu, W. Xiao, A. Dou, Z. Wang, H. Guo, Multi-layered carbon coated Si-based composite as anode for lithium-ion batteries, Powder Technol. 323 (2018) 294e300. [7] W.J. Zhang, A review of the electrochemical performance of alloy anodes for lithium-ion batteries, J. Power Sources 196 (2011) 13e24. [8] T. Li, X. Li, Z. Wang, H. Guo, Y. Li, A novel NiCo2O4 anode morphology for lithium-ion batteries, J. Mater. Chem. A 3 (2015) 11970e11975. [9] Y. Zhou, H. Guo, Y. Yong, Z. Wang, X. Li, R. Zhou, Introducing reduced graphene oxide to improve the electrochemical performance of silicon-based materials encapsulated by carbonized polydopamine layer for lithium ion batteries, Mater. Lett. 195 (2017) 164e167. [10] Q. Feng, S. Wen, J. Deng, W. Zhao, Combined DFT and XPS investigation of enhanced adsorption of sulfide species onto cerussite by surface modification with chloride, Appl. Surf. Sci. 425 (2017) 8e15. [11] H. Yin, X.X. Yu, Q.W. Li, M.L. Cao, W. Zhang, H. Zhao, M.Q. Zhu, Hollow porous CuO/C composite microcubes derived from metal-organic framework templates for highly reversible lithium-ion batteries, J. Alloys Compd. 706 (2017) 97e102. [12] Y. Zhu, Y. Zhu, S. Murali, M.D. Stoller, R.S. Ruoff, Nanostructured reduced graphene oxide/Fe2O3 composite as a high-performance anode material for lithium ion batteries, ACS Nano 5 (2011) 3333e3338. [13] X. Li, J. Zhao, S. Sun, L. Huang, Z. Qiu, P. Dong, Y. Zhang, The application of plasma treatment for Ti3þ modified TiO2 nanowires film electrode with enhanced lithium-storage properties, Electrochim. Acta 211 (2016) 395e403. [14] J. Leng, Z. Wang, X. Li, H. Guo, H. Li, K. Shih, G. Yan, J. Wang, Accurate construction of a hierarchical nickelecobalt oxide multishell yolkeshell structure with large and ultrafast lithium storage capability, J. Mater. Chem. A 5 (2017) 14996e15001. [15] M. Wagemaker, D.R. Simon, E.M. Kelder, J. Schoonman, C. Ringpfeil, U. Haake, D. Lützenkirchen-Hecht, R. Frahm, F.M. Mulder, A kinetic two-phase and equilibrium solid solution in spinel Li4þ xTi5O12, Adv. Mater. 18 (2006) 3169e3173. [16] J. Liu, J. Xiao, X. Zeng, P. Dong, J. Zhao, Y. Zhang, X. Li, Combustion synthesized macroporous structure MFe2O4 (M¼ Zn, Co) as anode materials with excellent electrochemical performance for lithium ion batteries, J. Alloys Compd. 699 (2017) 401e407. [17] J. Wang, G. Zhang, Z. Liu, H. Li, Y. Liu, Z. Wang, X. Li, K. Shih, L. Mai, Li3V(MoO4)3 as a novel electrode material with good lithium storage properties and improved initial coulombic efficiency, Nanomater. Energy 44 (2018) 272e278. [18] T. Li, X. Li, Z. Wang, H. Guo, Y. Li, J. Wang, A new design concept for preparing nickel-foam-supported metal oxide microspheres with superior electrochemical properties, J. Mater. Chem. A 5 (2017) 13469e13474. [19] S.Y. Yin, C.Q. Feng, S.J. Wu, H.L. Lu, B.Q. Ke, K.L. Zhang, D.H. Chen, Molten salt synthesis of sodium lithium titanium oxide anode material for lithium ion batteries, J. Alloys Compd. 642 (2015) 1e6. [20] W. Li, C. Hu, M. Zhou, K. Wang, H. Li, S. Cheng, K. Jiang, The electrochemical synthesis of LiNbO2 in molten salts and its application for lithium ion batteries with high rate capability, Electrochim. Acta 189 (2016) 231e236. [21] Y.M. Kim, J. Ahn, S.H. Yu, D.Y. Chung, K.J. Lee, J.L. Lee, Y.E. Sung, Titanium silicide coated porous silicon nanospheres as anode materials for lithium ion batteries, Electrochim. Acta 151 (2015) 256e262. [22] H. Sohn, D.H. Kim, R. Yi, D. Tang, S.E. Lee, Y.S. Jung, D. Wang, Semimicro-size agglomerate structured silicon-carbon composite as an anode material for high performance lithium-ion batteries, J. Power Sources 334 (2016) 128e136. [23] A.A. Ensafi, M.M. Abarghoui, B. Rezaei, Metal (Ni and Bi) coated porous silicon nanostructure, high-performance anode materials for lithium ion batteries with high capacity and stability, J. Alloys Compd. 712 (2017) 233e240. [24] P. Dong, D. Wang, Y. Yao, X. Li, Y. Zhang, J. Ru, T. Ren, Stabilizing interface layer of LiNi0.5Co0.2Mn0.3O2 cathode materials under high voltage using p-toluenesulfonyl isocyanate as film forming additive, J. Power Sources 344 (2017) 111e118. [25] D. Wang, X. Li, Z. Wang, H. Guo, Y. Xu, Y. Fan, J. Ru, Role of zirconium dopant on the structure and high voltage electrochemical performances of LiNi0.5Co0.2Mn0.3O2 cathode materials for lithium ion batteries, Electrochim. Acta 188 (2016) 48e56. [26] R. Teki, M.K. Datta, R. Krishnan, T.C. Parker, T.M. Lu, P.N. Kumta, N. Koratkar, Nanostructured silicon anodes for lithium ion rechargeable batteries, Small 5 (2009) 2236e2242. [27] H.X. Chen, Y. Xiao, L. Wang, Y. Yang, Silicon nanowires coated with copper layer as anode materials for lithium-ion batteries, J. Power Sources 196 (2011)

Z. Zhou et al. / Journal of Alloys and Compounds 751 (2018) 307e315 6657e6678. [28] J. Zhou, N. Lin, Y. Han, J. Zhou, Y. Zhou, J. Du, Y. Qian, Cu3Si@Si coreeshell nanoparticles synthesized using a solid-state reaction and their performance as anode materials for lithium ion batteries, Nanoscale 7 (2015) 15075e15079. [29] J.Y. Woo, A.Y. Kim, M.K. Kim, S.H. Lee, Y.K. Sun, G. Liu, J.K. Lee, Cu 3 Si-doped porous-silicon particles prepared by simplified chemical vapor deposition method as anode material for high-rate and long-cycle lithium-ion batteries, J. Alloys Compd. 701 (2017) 425e432. [30] S. Yoon, S.L. Lee, H. Kim, H.J. Sohn, Enhancement of capacity of carbon-coated Si-Cu3Si composite anode using metal-organic compound for lithium-ion batteries, J. Power Sources 161 (2006) 1319e1323. [31] M. Ben-Ishai, F. Patolsky, Shape-and dimension-controlled single-crystalline silicon and SiGe nanotubes: toward nanofluidic FET devices, J. Am. Chem. Soc. 131 (2009) 3679e3694. [32] S. Murugesan, J.T. Harris, B.A. Korgel, K.J. Stevenson, Copper-coated amorphous silicon particles as an anode material for lithium-ion batteries, Chem. Mater. 24 (2012) 1306e1315. [33] G.Z. Chen, D.J. Fray, T.W. Farthing, Direct electrochemical reduction of titanium dioxide to titanium in molten calcium chloride, Nature 407 (2000) 361e371. [34] C. Schwandt, D.J. Fray, The electrochemical reduction of chromium sesquioxide in molten calcium chloride under cathodic potential control, Z. Naturforsch. A 62 (2007) 655e670. [35] L. Rong, R. He, Z. Wang, J. Peng, X. Jin, G.Z. Chen, Investigation of electrochemical reduction of GeO2 to Ge in molten CaCl2-NaCl, Electrochim. Acta 147 (2014) 352e359. [36] Z. Zhou, Y. Hua, C. Xu, J. Li, Y. Li, L. Xiong, Z. Zhang, Electrolytic synthesis of ferrotitanium powders from ilmenite in CaCl2-NaCl melts at a lower temperature of 700 C, J. Alloys Compd. 726 (2017) 1124e1131. [37] Z.R. Zhou, Y.X. Hua, C.Y. Xu, J. Li, Y. Li, K. Gong, J.J. Ru, L. Xiong, Preparation of ferrotitanium from ilmenite by electrolysis-assisted calciothermic reduction in CaCl2-NaCl molten salt, JOM 68 (2016) 532e537. [38] X. Zou, X. Lu, C. Li, Z. Zhou, A direct electrochemical route from oxides to TieSi intermetallics, Electrochim. Acta 55 (2010) 5173e5185. [39] S. Fang, H. Wang, J. Yang, S. Lu, B. Yu, J. Wang, C. Zhao, Formation of Si nanowires by the electrochemical reduction of porous Ni/SiO2 blocks in molten CaCl2, J. Phys. Chem. Solid. 89 (2016) 1e6. [40] J. Zhao, S. Lu, L. Hu, C. Li, Nano Si preparation by constant cell voltage

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

315

electrolysis of FFC-Cambridge Process in molten CaCl2, J. Energ. Chem. 22 (2013) 819e825. W. Xiao, J. Zhou, L. Yu, D. Wang, X. Lou, Electrolytic formation of crystalline silicon/germanium alloy nanotubes and hollow particles with enhanced lithium-storage properties, Angew. Chem. Int. Ed. 55 (2016) 7427e7431. G. Rossi, T. Kendelewicz, I. Lindau, W.E. Spicer, The Si (111)/Cu interface studied with surface sensitive techniques, J. Vac. Sci. Technol., A 1 (1983) 987e990. H. Chen, Y. Xiao, L. Wang, Y. Yang, Silicon nanowires coated with copper layer as anode materials for lithium-ion batteries, J. Power Sources 196 (2011) 6657e6662. H. Zhao, H. Yin, X.X. Yu, W. Zhang, C. Li, M.Q. Zhu, In2O3 nanoparticles/carbon fiber hybrid mat as free-standing anode for lithium-ion batteries with enhanced electrochemical performance, J. Alloys Compd. 735 (2018) 319e326. H. Yin, M.L. Cao, X.X. Yu, H. Zhao, Y. Shen, C. Li, M.Q. Zhu, Self-standing Bi2O3 nanoparticles/carbon nanofiber hybrid films as a binder-free anode for flexible sodium-ion batteries, Mater. Chem. Front 1 (2017) 1615e1621. Y. Deng, D. Wang, W. Xiao, X. Jin, G.Z. Chen, Electrochemistry at conductor/ insulator/electrolyte three-phase interlines: a thin layer model, J. Mater. Chem. B 109 (2005) 14043e14051. H. Song, H.X. Wang, Z. Lin, X. Jiang, L. Yu, J. Xu, Z. Yu, X. Zhang, Y. Liu, P. He, L. Pan, Y. Shi, H. Zhou, K. Chen, Highly connected siliconecopper alloy mixture nanotubes as high-rate and durable anode materials for lithium-ion batteries, Adv. Funct. Mater. 26 (2016) 524e531. J. Wang, Z. Liu, G. Yan, H. Li, W. Peng, X. Li, L. Song, K. Shih, Improving the electrochemical performance of lithium vanadium fluorophosphate cathode material: focus on interfacial stability, J. Power Sources 329 (2016) 553e557. http://pubs.rsc.org/-/content/articlehtml/2014/ta/c4ta05098c. Q. Ye, J. Ru, J. Peng, G. Chen, D. Wang, Formation of multiporous MnO/Ndoped carbon configuration via carbon thermal reduction for superior electrochemical properties, Chem. Eng. J. 331 (2018) 570e577. Z. Yan, Q. Hu, G. Yan, H. Li, K. Shih, Z. Yang, X. Li, Z. Wang, J. Wang, Co3O4/Co nanoparticles enclosed graphitic carbon as anode material for high performance Li-ion batteries, Chem. Eng. J. 321 (2017) 495e501. H. Wu, N. Du, H. Zhang, D. Yang, Voltage-controlled synthesis of Cu-Li2O@Si coreeshell nanorod arrays as high-performance anodes for lithium-ion batteries, J. Mater. Chem. A 2 (2014) 20510e20514.