Fabrication and understanding of Cu3Si-Si@carbon@graphene nanocomposites as high-performance anodes for lithium-ion batteries

Zhiming Zheng a, Hong-Hui Wu b, Huixin Chen c, Yong Cheng a, Qiaobao Zhang *a, Qingshui Xie a, Laisen Wang a, Kaili Zhang d, Ming-Sheng Wang a, Dong-Liang Peng *a and Xiao Cheng Zeng *b
aDepartment of Materials Science and Engineering, Collaborative Innovation Center of Chemistry for Energy Materials, Xiamen University, Xiamen, Fujian 361005, China. E-mail: zhangqiaobao@xmu.edu.cn; dlpeng@xmu.edu.cn
bDepartment of Chemistry, University of Nebraska-Lincoln, NE 68588 Lincoln, USA. E-mail: xzeng1@unl.edu
cXiamen Institute of Rare Earth Materials, Haixi institutes, Chinese Academy of Sciences, Xiamen 361024, China
dDepartment of Mechanical and Biomedical Engineering, City University of Hong Kong, 83 Tat Chee Avenue, Hong Kong

Received 5th September 2018 , Accepted 19th September 2018

First published on 20th September 2018

Besides silicon's low electronic conductivity, another critical issue for using silicon as the anode for lithium-ion batteries (LIBs) is the dramatic volume variation (>300%) during lithiation/delithiation processes, which can lead to rapid capacity fading and poor rate capability, thereby hampering silicon's practical applications in batteries. To mitigate these issues, herein, we report our findings on the design and understanding of a self-supported Cu3Si-Si@carbon@graphene (Cu3Si-SCG) nanocomposite anode. The nanocomposite is composed of Cu3Si-Si core and carbon shell with core/shell particles uniformly encapsulated by graphene nanosheets anchored directly on a Cu foil. In this design, the carbon shell, the highly elastic graphene nanosheet, and the formed conductive and inactive Cu3Si phase in Si serve as buffer media to suppress volume variation of Si during lithiation/delithiation processes and to facilitate the formation of a stable solid electrolyte interface (SEI) layer as well as to enable good transport kinetics. Chemomechanical simulation results quantitatively coincide with the in situ TEM observations of volume expansion and provide process details not seen in experiments. The optimized Cu3Si-SCG nanocomposite anode exhibits good rate performance and delivers reversible capacity of 483 mA h g−1 (based on the total weight of Cu3Si-SCG) after 500 cycles with capacity retention of about 80% at high current density of 4 A g−1, rendering the nanocomposite a desirable anode candidate for high-performance LIBs.

1. Introduction

Advanced rechargeable lithium-ion batteries (LIBs) with high-energy density and long cycle life are critical for the development of next-generation energy storage devices.1–12 Silicon (Si) has been regarded as one of the promising anode candidates for substituting current commercialized graphitic anodes for LIBs due to its superior advantages including highest theoretical specific capacity (∼4200 mA h g−1), low working voltage (∼0.4 V vs. Li/Li+), high natural abundance, and environmental friendliness.9,13,14 However, the main challenges for Si-based anodes include cracking and pulverization of electrodes, loss of interparticle electrical contact, and regeneration of a solid electrolyte interphase (SEI) layer on the fractured surfaces caused by large volume variation (>300%) during lithiation/delithiation processes, which often lead to rapid capacity decay and severely reduced cycle life of Si-based anodes.15–17

Various approaches have been proposed to mitigate the aforementioned critical issues. For example, designing nanostructures of Si anodes, such as nanoparticles,18 nanotubes,19 nanowires,20,21 hollow nanospheres,22 opal nanocables,23,24 and nanoporous structures,25 has been demonstrated as an effective approach because nanosized Si can relieve the strain/stress induced by the volume change, conserve the structural integrity and ensure a short pathway for Li-ion diffusion and electron transport. Song et al.19 reported a Si nanotube array electrode, which delivered stable capacity retention (80% after 50 cycles) with high initial coulombic efficiency (∼85%) at a rate of 0.05 C. Unfortunately, the poor electrical conductivity of Si (1.6 × 10−3 S m−1) seriously limits its practical utilization and hinders its rate performance. To enhance the electrical conductivity of nanostructured Si and to alleviate the formation of unstable SEI caused by direct contact between nanoscale Si and electrolyte, modifying or decorating nanoscale Si with other conductive and stable buffer materials, especially carbonaceous layers such as amorphous carbon or graphene sheets, has been proposed as one of the most effective strategies.26–29 Yushin's group29 produced a C–Si nanocomposite through a hierarchical bottom-up assembly by loading Si nanoparticles on a 3D spherical carbon-black scaffold; the as-prepared Si/C nanohybrid exhibited high specific capacity of about 1500 mA h g−1 at a rate of 1 C after 100 cycles. Although these strategies can improve the electrochemical performances of Si-based materials for LIBs to some extent, they are still insufficient to prevent capacity loss during prolonged cycling, where cracking and pulverization of the shells and electrical disconnection from the current collector may occur upon volume expansion during repeated cycles. Recently, it has been shown that the incorporation of an electrochemically inactive volume buffer phase into Si anodes to form active Si/inactive phase composite can effectively alleviate mechanical stresses arising from the drastic volume change of Si anode during repeated lithiation/delithiation processes and provide structural reinforcement for the electrodes, thereby contributing to enhanced electrochemical performance.30–34 Here, the inactive phase does not react with lithium and remains unchanged during cycling.

Inspired by the previous study, herein, we report the design and fabrication of carbon-coated Si-based nanocomposites with an inactive Cu3Si buffer phase anchored on graphene nanosheets (denoted as Cu3Si-SCG) that are directly supported on a Cu foil substrate as high-performance anodes for LIBs. This electrode configuration integrates several favorable design attributes for high performance. Specifically, the carbon coating layer acts as a mechanical and electrochemical buffer medium, which not only reduces volume expansion and ensures the formation of stable SEI, but also increases electrical conductivity and reduces the inter-particle resistance of composites.35 Highly elastic graphene sheets further suppress the volume change and particle agglomeration of Cu3Si-Si@carbon nanocomposites during the lithiation/delithiation processes and enable the construction of a 3D interconnected conductive network to improve electrical conductivity and charge transport.28,36 Moreover, the formation of the highly conductive Cu3Si buffer phase with excellent mechanical flexibility and high electronic conductivity can mitigate structural degradation and offer high conductivity.30,31 With these favorable attributes of the self-supported Cu3Si-SCG architecture, the newly optimized composite electrode exhibits outstanding electrochemical performance in terms of good long-cycle stability and superior rate capability. Moreover, when coupled with a commercial LiCoO2 cathode, the full cell (2.0–3.9 V) delivers good capacity retention of ∼70% (based on the weight of LiCoO2 in cathode) after 50 cycles at 2 C (1 C = 135 mA g−1), demonstrating its potential as an anode candidate for high-performance LIBs.

2. Experimental section

2.1 Material synthesis

2.1.1 Synthesis of Si@C@graphene nanocomposites. Si@C@graphene (SCG) composite is prepared through sol–gel coating using resorcinol-formaldehyde (RF) as precursors. In a typical process, 0.2 g Si nanoparticles and 0.46 g hexadecyltrimethylammonium bromide (CTAB) are added into 14.08 mL of deionized water. After 30 min ultrasonication and 10 min stirring, 0.7 g resorcinol, 56.4 mL absolute ethanol and 0.2 mL NH4OH are added sequentially. Then, the mixed solution is stirred at 35 °C for 30 min, followed by the addition of 0.1 mL formaldehyde. The reaction solution is further stirred for 6 h at 35 °C and polymerization is conducted via ageing overnight. The precipitates are collected by centrifugation, washed with deionized water and alcohol several times, and dried under vacuum at 80 °C for 12 h. For the synthesis of the SCG composite, 0.15 g of obtained Si@RF nanoparticles is well dispersed in 50 mL deionized water by 30 min ultrasonication. Then, 10 mL of 2 mg mL−1 graphene oxide (GO) suspension is added dropwise under vigorous stirring. Si@RF nanoparticles with distributed positive charge provided by CTAB on the surface are immediately assembled with negatively charged GO through electrostatic interaction. After further stirring for 30 min, the resulting homogeneous aqueous solution is collected by centrifugation, washed with deionized water three times, and dried under vacuum freeze-drying at −50 °C for 24 h. Finally, the freeze-dried precursor is sintered at 600 °C for 3 h under argon atmosphere in a tube furnace to yield the SCG composite.
2.1.2 Synthesis of self-supported Cu3Si-Si@carbon@graphene nanocomposite electrode. Cu3Si-Si@carbon@graphene (Cu3Si-SCG) composite electrode is prepared by mixing 60 wt% active SCG material and 40 wt% polyvinylidene fluoride (PVDF) dissolved in N-methyl-2-pyrrolidone (NMP) to form a slurry and casting the slurry on a copper foil (d = 1.2 cm) current collector; no further carbon additives are used, followed by drying at 80 °C overnight in a vacuum environment. After compaction of the SCG electrodes and heat treatment at 900 °C for 0.5, 1, 1.5, 2, or 3 h in Ar atmosphere to further enhance three-dimensional conductivity of the composite electrode by partially or completely carbonizing the PVDF binder, the electrode remains intact and well-attached to the current collector.

2.2 Material characterization

The crystal phase components of as-prepared samples are characterized by an X-ray diffractometer (XRD) using PANalytical X'pert PRO X-ray diffractometer with Cu-Kα radiation at 40 kV and 40 mA. The morphological assessments and microstructures of the as-prepared samples are performed by field-emission scanning electron microscopy (FE-SEM, Hitachi SU-70, Tokyo, Japan). TEM images and EDX mappings are captured by a high-resolution transmission electron microscope (HRTEM, FEI, Talos-F200) operating at 200 kV. Raman measurements are taken on a HORIBA LabRAM HR Evolution Raman spectrometer with 532 nm argon ion laser. TG analysis is performed using an SDT-Q600 thermal analyzer in air atmosphere at a heating rate of 10 °C min−1 from room temperature to 800 °C. In situ TEM observations of the lithiation process are conducted using the Nanofactory TEM sample holder operated at 200 kV. Cu3Si-SCG sample and a piece of Li metal are mounted to individual sides of the holder with Au and W rods, respectively. The sample holder is then transferred to TEM (FEI Talos 200s) within a few seconds, during which a thin layer of Li2O (as solid-state electrolyte) is formed due to exposure to air. Lithiation of an individual Cu3Si-SCG sample starts when negative voltage is applied to the Au end.

2.3 Electrochemical characterization

The electrochemical properties of as-obtained samples are measured using CR2025 coin-type cells. The working electrode is the as-prepared Cu3Si-SCG composite electrode mentioned above. The Cu foil is first cut into circular disks with a diameter of 12 mm, and the masses of the as-obtained Cu foil circular disk (m1) and Cu3Si-SCG nanocomposites supported on the disk (m2) are determined using a microbalance (Mettler Toledo XS3DU) with accuracy of 1 μg. The active mass of mass loading of the electrode is calculated as m2m1 and is about 0.8–1.2 mg cm−2. The SCG ordinary electrode with no heat treatment is obtained by casting a commixture of the SCG composite, Super P carbon and PVDF (weight ratio: 6[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]2) for comparison. Coin-type half cells are fitted together in an argon-filled glovebox with a lithium wafer as the counter electrode, Celgard 2300 membrane as the separator, and 1 M LiPF6 in a mixed solution of ethylene carbonate (EC), diethyl carbonate membrane (DEC) and dimethyl carbonate (DMC) (1[thin space (1/6-em)]:[thin space (1/6-em)]1[thin space (1/6-em)]:[thin space (1/6-em)]1 in volume) with the addition of 10 wt% fluoride ethylene carbonate (FEC) as the electrolyte. Prior to full cell assembly, the Cu3Si-SCG composite anode is pre-lithiated in a half-cell by discharging to 0.01 V and then charging to 1.0 V to alleviate the effect of initial low coulombic efficiency on the cycle life of the full cell. For the full cell test, the cathode is fabricated by mixing 80 wt% commercialized LiCoO2 (LCO) with 10 wt% carbon black and 10 wt% polyvinylidene fluoride dissolved in N-methyl-2-pyrrolidone (NMP) to form a slurry, which is then spread onto an Al foil current collector and dried under vacuum at 80 °C for 24 hours. The electrolytes and separator in the full cell are the same as those in the half-cells described above. The full cell is designed with an N/P ratio of 1.2[thin space (1/6-em)]:[thin space (1/6-em)]1; electrochemical analysis of the full cell is carried out in the voltage window between 2.0 V and 3.9 V. The capacity of the full cell is calculated based on the weight of active materials in the cathode.

2.4 Chemomechanical simulation

A nonlinear chemomechanical model of two-phase lithiation is adopted to gain insights into the volume expansion and stress development during lithiation.37,38 The concentration of Li, c, is governed by the standard diffusion equation ∂c(r, t)/∂t = ∇ × [D(c, r)∇c (r, t)]. The diffusivity D depends on c, i.e., D = D0[1/(1 − c) − 2Ωc], where D0 is the diffusivity coefficient and Ω controls the lithium profile near the lithiated and unlithiated phase boundary. During the lithiation process, there is a small gradient of c at the reaction front and thus, Li diffuses through the lithiated shell to the unlithiated particle center. Maximum D is capped at 104D0 to ensure numerical stability. To understand the mechanical deformation, we adopt an ideal elastic-plastic model to capture the lithiation-induced deformation.39–41 The total strain rate, [small epsi, Greek, dot above]ij, consists of three parts: [small epsi, Greek, dot above]ij = [small epsi, Greek, dot above]cij + [small epsi, Greek, dot above]eij + [small epsi, Greek, dot above]pij, where [small epsi, Greek, dot above]cij denotes the chemical strain rate caused by lithiation and [small epsi, Greek, dot above]cij = βcij, where βij is the lithiation expansion coefficient. Also, [small epsi, Greek, dot above]eij is the elastic strain rate and obeys the Hooke's law. The plastic strain rate, [small epsi, Greek, dot above]pij, obeys the classic J2-flow rule.42 Here, plastic yielding occurs when von Mises equivalent stress, σeq = (3σijσij/2)1/2, equals the yield strength, σY, where σij = σijσkkδij/3 is the deviatoric stress. The plastic strain rate is given by [small epsi, Greek, dot above]pij = [small lambda, Greek, dot above]σij, where [small lambda, Greek, dot above] is a scalar coefficient and can be determined by solving the specific boundary value problem. Such a chemomechanical model is also subject to the mechanical equilibrium equation σij,j = 0.43,44 To simplify the Cu3Si-SCG-2 model, we treat the active Si and inactive Cu3Si phases as a Cu3Si-Si composite and then choose isotropic lithiation expansion coefficients for bare Si, carbon shell, and Cu3Si-Si composite as βij = 0.32, 0 and 0.05, respectively.

3. Results and discussion

Detailed formation processes of SCG and Cu3Si-SCG-2 composites are described in Fig. 1a and b. First, a carbon (C) precursor layer and GO sheets are coated on Si nanoparticles by a sol–gel coating process using RF as precursors, followed by electrostatic interaction. After annealing, SCG composites are obtained (Fig. 1a). Subsequently, the active SCG material and PVDF are mixed with NMP to form a slurry, and the slurry is cast on copper. Finally, the copper is coated with active SCG through a controlled heat treatment under Ar atmosphere at 900 °C for 2 h, resulting in the formation of a self-supported Cu3Si-SCG-2 electrode (Fig. 1b). The morphologies and microstructures of SCG (Fig. S1) and Cu3Si-SCG-2 composites are obtained by FESEM and TEM, respectively. Fig. 1c and d exhibit low and high magnification FESEM images of the Cu3Si-SCG-2 sample, showing that homogeneous and uniform Si-based composite nanoparticles are wrapped by thin graphene sheets. The TEM images of Cu3Si-SCG-2 are shown in Fig. 1e and f; the original core–shell structure of SCG is well preserved after heat treatment. Moreover, the HRTEM image in Fig. 1g displays distinct crystal lattice fringes of both crystalline Si and Cu3Si in the composite. The interlayer spacings are 0.314 nm and 0.203 nm, indexing to the (111) plane of Si and the (012) plane of Cu3Si, respectively. The STEM and elemental mapping images in Fig. 1h–m clearly illustrate Si, C, and Cu elemental distributions, which further evidence the uniform distribution of Cu3Si in SCG nanocomposites.
image file: c8nr07207h-f1.tif
Fig. 1 (a) Schematic drawing of the fabrication of Si@C@graphene (SCG) composite and (b) schematic of the heat treatment process of self-supported Cu3Si-SCG electrode. (c, d) FESEM images of Cu3Si-SCG-2 composite. (e, f) TEM and (g) HRTEM (the inset in Fig. 3g shows the corresponding FFT image) images of Cu3Si-SCG-2 composite. (h–m) STEM image of Cu3Si-SCG-2 composite and corresponding element mappings of Si, C and Cu.

The phase purity and composition of the as-obtained composites are investigated using XRD and Raman spectroscopy. The diffraction peaks at 28.4, 47.3, 56.1, 69.1, 76.4 and 88.0° in the spectra of all samples are well-indexed to the crystalline Si phase (JCPDS 27-1402), as shown in Fig. 2a. Furthermore, the additional diffraction peaks of Cu3Si-SCG-2 at 44.5° and 44.9° can be assigned to the Cu3Si phase (JCPDS 51-0916), as shown in the inset of Fig. 2a, indicating the formation of the Cu3Si phase after heat treatment at 900 °C. The Raman spectrum from 400 to 2000 cm−1 in Fig. 2b displays two distinct peaks around 497 and 923 cm−1, which are associated with the Si phase. The peaks around 1340 and 1585 cm−1 are ascribed to D and G bands of carbon, respectively. The Si peak at 497 cm−1 of Cu3Si-SCG-2 electrodes presents lower intensity and a blue-shift to about 510 cm−1 compared to those of other samples, which may be attributed to the formation of the Cu3Si phase and phonon confinement effect or masking effect after heat treatment.45 Furthermore, ID/IG of Cu3Si-SCG-2 composite is about 1.0, which is higher than that of the SCG composite (0.826), indicating more defects in the SCG composite after annealing. The defects provide Li+ diffusion channels and react electrochemically with Si during charge/discharge.46Fig. 2c shows the EDX spectrum of Cu3Si-SCG-2 acquired from the full region of Fig. 1h and the corresponding elemental content. The ratio of Si[thin space (1/6-em)]:[thin space (1/6-em)]Cu3Si is calculated to be around 9.9[thin space (1/6-em)]:[thin space (1/6-em)]1 according to the chart inset in Fig. 2c. The carbon contents of SC and SCG are calculated to be ∼15 wt% and ∼28.5 wt%, respectively, according to thermogravimetric analysis (TGA) (Fig. S2).

image file: c8nr07207h-f2.tif
Fig. 2 (a) XRD patterns and (b) Raman spectra of pure Si and SC, SCG, and Cu3Si-SCG-2 composites. (c) EDX spectrum of Cu3Si-SCG-2 acquired from the full region of Fig. 1h and the corresponding elemental contents.

To understand the electrochemical performance of the SCG electrode after heat treatment, we carry out related electrochemical performance testing. The cyclic voltammogram (CV) curves of as-synthesized Cu3Si-SCG-2 are first measured at a scan rate of 0.1 mV s−1 for initial four cycles in the 0.01–2 V (vs. Li/Li+) voltage window. As shown in Fig. 3a, a small peak at 0.4–0.65 V appears during the first cathodic scan of Cu3Si-SCG-2 electrode, which could be due to irreversible reduction and the formation of SEI film on the carbon shell surface and it disappears in the following cycles. A clear reduction peak around 0.19 V emerges in the following cycles with increased intensity and is associated with the lithiation reaction of Si to form an amorphous LixSi (a-LixSi) phase. A sharp cathodic peak at potential <0.1 V corresponds to the formation of the Li15Si4 phase. Two distinct anodic peaks located around 0.35 and 0.51 V are detected in the first charge process and become more distinct in the following cycles; they are associated with phase transitions from Li15Si4 alloy to a-LixSi and a-LixSi to amorphous Si, respectively. Importantly, the CV curves of the final two cycles almost overlap, indicating good reversibility of the corresponding alloying/dealloying processes.47,48 Typical charge–discharge voltage profiles of the Cu3Si-SCG-2 electrode at a rate of 0.2 A g−1 in the first two cycles and 1 A g−1 in the following cycle are shown in Fig. 3b. In the first cycle, the discharge curve exhibits a long plateau at about 0.1 V, which is associated with the discharge potential of crystalline Si to form the amorphous LixSi phase.11 The voltage profiles in the following cycles exhibit the same shape between 0.2 and 0.01 V, which corresponds to the typical electrochemical behavior of the lithium alloy of amorphous Si. These results are consistent with the electrochemical mechanisms discussed above. Furthermore, the first discharge specific capacity is 1246 mA h g−1 with initial coulombic efficiency (ICE) of 71.8%; the coulombic efficiency (CE) quickly increases to 93.6% in the second cycle. The irreversible capacity loss can be ascribed to the decomposition of electrolyte and formation of SEI on the electrode surface in the first cycle and can be compensated by prelithiation through either chemical or electrochemical methods or by using stabilized Li metal powder.11Fig. 3c displays the cycling performance and corresponding CE of Cu3Si-SCG composite with different heat treatment times at 1 A g−1. All cells are first activated at 0.2 A g−1 for two cycles and then subjected to 500 consecutive cycles at 1 A g−1. SCG composite electrodes after heat treatments at 900 °C for 0.5, 1, 1.5, 2, and 3 h deliver discharge specific capacities of 289, 448, 520, 522, and 401 mA h g−1 after 500 cycles at 1 A g−1 with capacity retentions of 15.4%, 34.2%, 42.8%, 70.7%, and 97.4%, respectively. It can be clearly seen that there is remarkable improvement in cycling stability with relatively low reversible capacity with the increase in heat treatment time, resulting in increase of Cu3Si buffer phase content in the composite electrode, which may effectively suppress large structural changes and relieve mechanical stress concentration induced by Li15Si4 formation during cycling.24 For comparison, cycling performances of pure Si and SC after annealing at 900 °C for 2 h together with cycling performances of SCG without annealing and SCG electrode with heat treatment at 700 and 800 °C for 2 h are given in Fig. S3a and S3b, respectively. The Cu3Si-Si-2 electrode demonstrates the most significant capacity decay among all electrodes, delivering capacity of 317 mA h g−1 after 500 cycles at 1 A g−1 with only 25% capacity retention, whereas the Cu3Si-SC-2 electrode maintains 44.8% capacity retention (Fig. S3a), confirming the important roles of the carbon shell and highly elastic graphene nanosheets in the improvement of battery cycling. Fig. S3b shows that the capacity of the SCG electrode without annealing fades rapidly, whereas SCG electrodes heat-treated at 700 and 800 °C for 2 h without the formation of inactive Cu3Si phase exhibit 17.3% and 27.8% capacity retention after 500 cycles at 1 A g−1, respectively. Cu3Si-SCG-2 retains 522 mA h g−1 after 500 cycles at 1 A g−1 with capacity retention of 70.7%, indicating that the formation of inactive Cu3Si buffer phase can mitigate structural degradation,24 thus resulting in the improvement of cycling stability. Fig. 3d presents the rate capabilities of different Cu3Si-SCG composite electrodes at various current densities. The charge capacity of the Cu3Si-SCG-2 electrode is 512 mA h g−1 at current density of 4 A g−1 with capacity retention of 57.3% from 0.2 A g−1 to 4 A g−1. When the current rate returns from 4 to 0.2 A g−1, the charge specific capacity recovers to the original value and reaches up to 938 mA g−1; it also displays charge capacity of 737 mA g−1 after 10 repeated cycles at 1 A g−1. The charge capacities of the other three electrodes cannot reach the original values when the current rate returns from 4 to 0.2 A g−1. The remarkable rate capability of the Cu3Si-SCG-2 electrode is mainly due to increase in highly conductive Cu3Si content in the composite electrode because of the extended heat treatment time. The carbon shell and graphene sheet together with carbonization of binder form a 3D conductive network structure, effectively improving the diffusion rate of ions and electrons. Impressively, the Cu3Si-SCG-2 electrode exhibits outstanding long-cycle performance, delivering charge capacity of 483 mA h g−1 after 500 cycles with capacity retention of about 80% at high current density of 4 A g−1 (Fig. 3e); it outperforms some reported inactive phase/Si-based composite anodes (Table S1). In contrast, other electrodes except SCG-HT-900-3 h exhibit distinct capacity decay after 500 cycles at the same current density. The highest capacity retention of the Cu3Si-SCG-3 electrode, reaching almost 100% with relatively low reversible capacity, is mainly ascribed to the highest content of Cu3Si due to prolongation of annealing time. This result further demonstrates that the formation of Cu3Si alloy plays an important role in enhanced cyclic performance. To better understand the improved electrochemical performance of the Cu3Si-SCG-2 electrode compared to that of the SCG electrode, electrochemical impedance spectroscopy (EIS) measurements are obtained. Fig. S4 presents the experimental and fitted EIS curves for SCG and Cu3Si-SCG-2 electrodes before cycling together with that for the Cu3Si-SCG-2 electrode after 500 cycles at 4 A g−1. It can be seen that all samples consist of a semicircle in the high-frequency region and a sloping line in the low-frequency region, which represent interfacial electrode characteristics and impedance of ionic diffusion in electrode materials, respectively. The charge-transfer resistances (Rct) of the electrodes are extracted from the impedance data using the equivalent circuit shown in the inset in Fig. S4.Rct of Cu3Si-SCG-2 electrode before cycling is about 35.2 Ω, which is much smaller than that of the SCG electrode (64.9 Ω), suggesting better conductivity due to the formation of highly conductive Cu3Si phase after heat treatment. After 500 cycles, Rct of Cu3Si-SCG-2 composite reaches 2.3 Ω, which may be ascribed to enhanced charge-transfer kinetics, facilitating lithium-ion storage under high current density.35

image file: c8nr07207h-f3.tif
Fig. 3 Electrochemical properties of Cu3Si-SCG-2 composite in half cell: (a) CV curves of Cu3Si-SCG-2 in initial four cycles, (b) discharge–charge curves of Cu3Si-SCG-2 at 1 A g−1, (c) cycling performances of Cu3Si-SCG composite with different heat treatment times at 1 A g−1, (d) rate performances of Cu3Si-SCG composite with different heat treatment times, and (e) long-term cycling performances of Cu3Si-SCG composite with different heat treatment times at 4 A g−1.

To explain the outstanding electrochemical performance of the SCG composite after annealing for 2 h at 900 °C, in situ TEM (Fig. 4a) is employed to study the volume change and structural evolution of bare Si, SCG nanocomposite and Cu3Si-SCG-2 nanocomposite during the lithiation process. Fig. 4b presents time-resolved TEM images (captured from ESI Video S1) of the lithiation process of Cu3Si-SCG-2 composite, which do not exhibit visible volume expansion from the initial state (0 s) to the final lithiation state (120 s), indicating that the presence of Cu3Si phase after heat treatment effectively suppresses large structural changes during lithiation. As shown in Fig. 4c (captured from ESI Video S2), the diameter of the SCG composite increases from 37 nm at the initial state (0 s) to 41 nm at the final lithiation state (120 s). The slight volume change of the SCG composite also indicates that the carbon layer and graphene coating play important roles in structural expansion. Bare Si nanoparticles show distinct large volume expansion from 59 nm to 93 nm in diameter after 87 s of lithiation, as shown in Fig. 4d (captured from ESI Video S3), clearly displaying a core–shell structure with amorphous LixSi layer coated on the Si core; this indicates that lithiation is initiated from the surface of the particle and it propagates toward the center of the particles, featuring phase boundary evolution between amorphous LixSi and crystalline Si, as previously reported.49,50 The relative expansion rates of the three electrodes as a function of lithiation time, as shown in Fig. 4e, demonstrate that the coating modification and electrode heat treatment contribute to structural stability.

image file: c8nr07207h-f4.tif
Fig. 4 (a) Schematic of in-situ nano-battery configuration. In situ TEM observation of the lithiation process of (b) Cu3Si-SCG-2 composite, (c) SCG composite and (d) bare Si. (e) The relative expansion of three electrodes with respect to different lithiation times.

To better understand the structural expansion in in situ TEM observations and to evaluate the stress development during the lithiation process in detail, we carry out a continuum chemomechanical simulation by coupling the Li diffusion equation and elasto-plastic deformation equation in the finite element framework.40,41 The normalized Li concentration, c, is defined as the actual Li concentration divided by the Li concentration of the fully lithiated state (c varies from 0 to 1). The lithiation reaction front is located at the interface between pristine (blue) and lithiated (red) phases. A series of sequential snapshots of the simulated lithium distribution and von Mises stress profiles during lithiation are shown in Fig. 5. The simulation results well capture the propagation of the lithiation reaction front in the radial direction. To compare the stress evolution more intuitively, different color map ranges of von Mises stress are adopted. For the bare Si model (Fig. 5a1 and b1), with the increase in evolution time, Li+ ion diffuses from the Li-rich outer layer to the Li-poor inner layer (ESI Video S4), and the associated von Mises stress redistributes with Li+ ion evolution. Generally, this is a free-expansion process without serious stress concentration. However, the expansion volume is very large. The Li+ ion diffusion and stress development of SCG model are presented in Fig. 5b1 and b2, respectively. Since there is a carbon surface layer outside the nanoparticles, free expansion is mechanically suppressed during the lithiation process (ESI Video S5). However, the stress concentration at the interface of the Si core and carbon shell approaches 70 GPa during the lithiation process, which can crack the samples and decrease the long-cycle stability of coated Si particles. In contrast to the large volume expansion of bare Si model and serious stress concentration of SCG model, Fig. 5c1 and c2 present snapshots of simulated Li distribution and associated radial stress distribution of Cu3Si-SCG-2. It can be seen that both the volume expansion (ESI Video S6) and stress concentration are well balanced during lithiation. These simulation results suggest two dominant factors accounting for the high performance of Cu3Si-SCG-2. First, the carbon shell effectively suppresses the large radical expansion of the Si core, which reduces the risk of fragmentation during cycling lithiation/delithiation. Second, the huge tensile stress in carbon shells is largely relieved by the buffered Cu3Si component. Both results are consistent with the experimental observations that Cu3Si-SCG-2 composite displays excellent performance.

image file: c8nr07207h-f5.tif
Fig. 5 Chemomechanical modeling of lithiation evolution (a1)–(c1) and von Mises stress distribution (a2)–(c2) during lithiation process of bare Si, SCG, and Cu3Si-SCG-2. To compare the simulations more intuitively, different color maps of von Mises stress are used. The final radial expansions of bare Si, SCG, and Cu3Si-SCG-2 models are 159.8%, 139.2%, and 103.8%, respectively.

The structural stability and the morphology and structural changes of Cu3Si-SCG-2 electrode after 500 cycles at 4 A g−1 are explored by SEM and TEM, respectively. As shown in Fig. 6a and b, initial nanoparticles covered by a gel-like layer on their surfaces can be perfectly maintained after cycling and retain good interconnection of active particles, indicating that the carbonized binder contributes to enhanced electrode integrity. Moreover, TEM and HRTEM images shown in Fig. 6c–e further confirm the structural integrity of the electrode. There is no clear cracking of active nanoparticles after 500 repeated charge–discharge cycles, suggesting that Cu3Si effectively suppresses large structural changes and relieves the stress concentration in Si-Li alloying process. Furthermore, the neighboring fringe distance is 0.201 nm according to FFT patterns (inset of Fig. 6e), which corresponds to the (300) d-spacing of Cu3Si. To verify the feasibility of Cu3Si-SCG-2 electrode in practical devices, a full cell using Cu3Si-SCG-2 as the anode and commercial LiCoO2 (LCO) as the cathode is assembled. The capacity of the LCO cathode is 129 mA h g−1 at 0.2 C and 119 mA h g−1 after 200 cycles at 2 C (Fig. S5). Fig. 6f shows the typical discharge/charge profiles of the full cell measured at 2 C, which is first activated at 0.2 C for three cycles, followed by 50 consecutive cycles at 2 C (1 C = 135 mA g−1) in the voltage range of 2.0–3.9 V. The initial coulombic efficiency of the full cell is 81.5% at 0.2 C, which is higher than that of the half-cell; this can be due to effective prelithiation. As shown in Fig. 6g, the as-fabricated full cell delivers reversible capacity of 62 mA h g−1 after 50 cycles with capacity retention of 70% at 2 C, demonstrating significant potential for future practical applications.

image file: c8nr07207h-f6.tif
Fig. 6 (a, b) FESEM images of Cu3Si-SCG-2 composite after cycling. (c, d) TEM and (e) HRTEM (inset in Fig. 5e shows the corresponding FFT image) images of the Cu3Si-SCG-2 composite after cycling. Electrochemical properties of the Cu3Si-SCG-2 hybrid electrode in full cell: (f) galvanostatic discharge–charge curves and (g) cycling performance between 2.0 and 3.9 V at 0.2 C for initial three cycles and then at 2 C for the following cycles.

4. Conclusions

In summary, a high-performance LIB anode material consisting of carbon-coated Cu3Si-Si nanocomposites wrapped by graphene nanosheets anchored directly on a Cu foil is designed and successfully fabricated via a simple sol–gel method, followed by well-controlled heat treatment. The newly optimized Cu3Si-SCG anode exhibits good rate capability (512 mA h g−1 at 4 A g−1 with capacity retention of 58% from 0.2 A g−1 to 4 A g−1) and long-term cycling stability (483 mA h g−1 after 500 cycles with capacity retention of about 80% at high current density of 4 A g−1). The enhanced electrochemical performance can be due to the beneficial effects of carbon-coating, encapsulation by highly elastic graphene nanosheets, and the presence of conductive Cu3Si buffer phase in Si, which can mitigate structural degradation of Si during repeated lithiation/delithiation processes while facilitating the growth of a stable solid electrolyte interface (SEI) layer and good transport kinetics. The high structural stability of Cu3Si-Si@C nanocomposites, confirmed by in situ TEM measurements, is due to uniform encapsulation by graphene nanosheets. The chemomechanical simulation results confirm quantitatively that Cu3Si-Si@C nanocomposites can well accommodate volume expansion and stress concentration during the lithiation process, thereby enabling longer cycle structural stability. The new design will be useful for the further development of high-performance Si-based composite anode materials and for engineering other anode materials with higher performance for next-generation LIBs.

Conflicts of interest

There are no conflicts to declare.


This work is supported by the National Natural Science Foundation of China (Grants No. 21703185 and 21805278), the National Key Research Program of China (Grant No. 2016YFA0202602) and Fundamental Research Funds for the Central Universities (Xiamen University: 20720170042). The computational work is conducted on computer facility in UNL Holland Computing Center.


  1. K. Feng, M. Li, W. Liu, A. G. Kashkooli, X. Xiao, M. Cai and Z. Chen, Small, 2018, 14, 1702737 CrossRef PubMed.
  2. D. Liu, Z. J. Liu, X. Li, W. Xie, Q. Wang, Q. Liu, Y. Fu and D. He, Small, 2017, 13, 1702000 CrossRef PubMed.
  3. D. T. Ngo, H. T. T. Le, X. M. Pham, C. N. Park and C. J. Park, ACS Appl. Mater. Interfaces, 2017, 9, 32790–32800 CrossRef CAS PubMed.
  4. K. Li, X. Zhou, A. Nie, S. Sun, Y.-B. He, W. Ren, B. Li, F. Kang, J.-K. Kim and T.-Y. Zhang, Nano Lett., 2017, 17, 1282–1288 CrossRef CAS PubMed.
  5. K. Li, B. Li, J. Wu, F. Kang, J.-K. Kim and T.-Y. Zhang, ACS Appl. Mater. Interfaces, 2017, 9, 35917–35926 CrossRef CAS PubMed.
  6. X. Leng, S. Wei, Z. Jiang, J. Lian, G. Wang and Q. Jiang, Sci. Rep., 2015, 5, 16629 CrossRef CAS PubMed.
  7. X. Leng, Y. Shao, L. Wu, S. Wei, Z. Jiang, G. Wang, Q. Jiang and J. Lian, J. Mater. Chem. A, 2016, 4, 10304–10313 RSC.
  8. X. Leng, X. Ding, J. Hu, S. Wei, Z. Jiang, J. Lian, G. Wang, Q. Jiang and J. Liu, Electrochim. Acta, 2016, 190, 276–284 CrossRef CAS.
  9. F. H. Du, B. Li, W. Fu, Y. J. Xiong, K. X. Wang and J. S. Chen, Adv. Mater., 2014, 26, 6145–6150 CrossRef CAS PubMed.
  10. Y. Tang, J. Deng, W. Li, O. I. Malyi, Y. Zhang, X. Zhou, S. Pan, J. Wei, Y. Cai, Z. Chen and X. Chen, Adv. Mater., 2017, 29, 1701828 CrossRef PubMed.
  11. Y. Zhang, O. I. Malyi, Y. Tang, J. Wei, Z. Zhu, H. Xia, W. Li, J. Guo, X. Zhou, Z. Chen, C. Persson and X. Chen, Angew. Chem., 2017, 56, 14847–14852 CrossRef CAS PubMed.
  12. Y. Tang, Y. Zhang, O. I. Malyi, N. Bucher, H. Xia, S. Xi, Z. Zhu, Z. Lv, W. Li, J. Wei, M. Srinivasan, A. Borgna, M. Antonietti, Y. Du and X. Chen, Adv. Mater., 2018, e1802200 CrossRef PubMed.
  13. H. Shang, Z. Zuo, L. Yu, F. Wang, F. He and Y. Li, Adv. Mater., 2018, e1801459 CrossRef PubMed.
  14. L. Zhang, R. Rajagopalan, H. P. Guo, X. L. Hu, S. X. Dou and H. K. Liu, Adv. Funct. Mater., 2016, 26, 440–446 CrossRef CAS.
  15. C. Sämann, K. Kelesiadou, S. S. Hosseinioun, M. Wachtler, J. R. Köhler, K. P. Birke, M. B. Schubert and J. H. Werner, Adv. Energy Mater., 2018, 8, 1701705 CrossRef.
  16. J. Liang, X. Li, Z. Hou, W. Zhang, Y. Zhu and Y. Qian, ACS Nano, 2016, 10, 2295–2304 CrossRef CAS PubMed.
  17. Q. Zhang, H. Chen, L. Luo, B. Zhao, H. Luo, X. Han, J. Wang, C. Wang, Y. Yang, T. Zhu and M. Liu, Energy Environ. Sci., 2018, 11, 669–681 RSC.
  18. X. H. Liu, Li Zhong, S. Huang, S. X. Mao, T. Zhu and J. Y. Huang, ACS Nano, 2012, 6, 1522–1531 CrossRef CAS PubMed.
  19. T. Song, J. Xia, J. H. Lee, D. H. Lee, M. S. Kwon, J. M. Choi, J. Wu, S. K. Doo, H. Chang, W. I. Park, D. S. Zang, H. Kim, Y. Huang, K. C. Hwang, J. A. Rogers and U. Paik, Nano Lett., 2010, 10, 1710–1716 CrossRef CAS PubMed.
  20. X. H. Liu, L. Q. Zhang, L. Zhong, Y. Liu, H. Zheng, J. W. Wang, J. H. Cho, S. A. Dayeh, S. T. Picraux, J. P. Sullivan, S. X. Mao, Z. Z. Ye and J. Y. Huang, Nano Lett., 2011, 11, 2251–2258 CrossRef CAS PubMed.
  21. Y. Yao, K. Huo, L. Hu, N. Liu, J. J. Cha, M. T. McDowell, P. K. Chu and Y. Cui, ACS Nano, 2011, 5, 8346–8351 CrossRef CAS PubMed.
  22. Y. Yao, M. T. McDowell, I. Ryu, H. Wu, N. Liu, L. Hu, W. D. Nix and Y. Cui, Nano Lett., 2011, 11, 2949–2954 CrossRef CAS PubMed.
  23. H. Liu, H.-M. Cho, Y. S. Meng and Q. Li, ACS Appl. Mater. Interfaces, 2014, 6, 9842–9849 CrossRef CAS PubMed.
  24. H. Liu and Q. Li, ACS Appl. Mater. Interfaces, 2013, 5, 12190–12196 CrossRef CAS PubMed.
  25. M. Ge, Y. Lu, P. Ercius, J. Rong, X. Fang, M. Mecklenburg and C. Zhou, Nano Lett., 2014, 14, 261–268 CrossRef CAS PubMed.
  26. N. Lin, J. B. Zhou, L. B. Wang, Y. C. Zhu and Y. T. Qian, ACS Appl. Mater. Interfaces, 2015, 7, 409–414 CrossRef CAS PubMed.
  27. F. M. Hassan, R. Batmaz, J. Li, X. Wang, X. Xiao, A. Yu and Z. Chen, Nat. Commun., 2015, 6, 8597 CrossRef CAS PubMed.
  28. L. Wang, B. Gao, C. Peng, X. Peng, J. Fu, P. K. Chu and K. Huo, Nanoscale, 2015, 7, 13840–13847 RSC.
  29. A. Magasinski, P. Dixon, B. Hertzberg, A. Kvit, J. Ayala and G. Yushin, Nat. Mater., 2010, 9, 353–358 CrossRef CAS PubMed.
  30. S.-Ok Kim and A. Manthiram, J. Power Sources, 2016, 332, 222–229 CrossRef CAS.
  31. J. Zhou, N. Lin, Y. Han, J. Zhou, Y. Zhu, J. Du and Y. Qian, Nanoscale, 2015, 7, 15075–15079 RSC.
  32. H. Song, S. Wang, X. Song, H. Yang, G. Du, L. Yu, J. Xu, P. He, H. Zhou and K. Chen, J. Mater. Chem. A, 2018, 6, 7877–7886 RSC.
  33. L. Deng, Y. Cui, J. Chen, J. Wu, A. P. Baker, Z. Li and X. Zhang, Electrochim. Acta, 2016, 192, 303–309 CrossRef CAS.
  34. P. Zhang, L. Huang, Y. Li, X. Ren, L. Deng and Q. Yuan, Electrochim. Acta, 2016, 192, 385–391 CrossRef CAS.
  35. Z. Zheng, Y. Zao, Q. Zhang, Y. Cheng, H. Chen, K. Zhang, M.-S. Wang and D.-L. Peng, Chem. Eng. J., 2018, 347, 563–573 CrossRef CAS.
  36. Z.-S. Wu, G. Zhou, L.-C. Yin, W. Ren, F. Li and H.-M. Cheng, Nano Energy, 2012, 1, 107–131 CrossRef CAS.
  37. J. W. Wang, Y. He, F. Fan, X. H. Liu, S. Xia, Y. Liu, C. T. Harris, H. Li, J. Y. Huang and S. X. Mao, Nano Lett., 2013, 13, 709–715 CrossRef CAS PubMed.
  38. S. Zhang, K. Zhao, T. Zhu and J. Li, Prog. Mater. Sci., 2017, 89, 479–521 CrossRef CAS.
  39. L. Q. Zhang, X. H. Liu, Y. Liu, S. Huang, T. Zhu, L. Gui, S. X. Mao, Z. Z. Ye, C. M. Wang and J. P. Sullivan, ACS Nano, 2011, 5, 4800–4809 CrossRef CAS PubMed.
  40. K. Zhao, M. Pharr, J. J. Vlassak and Z. Suo, J. Appl. Phys., 2011, 109, 016110 CrossRef.
  41. S. Huang, F. Fan, J. Li, S. Zhang and T. Zhu, Acta Mater., 2013, 61, 4354–4364 CrossRef CAS.
  42. R. M. Jones, Deformation theory of plasticity, Bull Ridge Corporation, 2009 Search PubMed.
  43. J. Zhu, H. Wu, D. Wang, Y. Gao, H. Wang, Y. Hao, R. Yang, T.-Y. Zhang and Y. Wang, Int. J. Plast., 2017, 89, 110–129 CrossRef CAS.
  44. S.-G. Cao, H.-H. Wu, H. Ren, L.-Q. Chen, J. Wang, J. Li and T.-Y. Zhang, Acta Mater., 2015, 97, 404–412 CrossRef CAS.
  45. M.-S. Wang and L.-Z. Fan, J. Power Sources, 2013, 244, 570–574 CrossRef CAS.
  46. L. Yan, J. Liu, Q. Wang, M. Sun, Z. Jiang, C. Liang, F. Pan and Z. Lin, ACS Appl. Mater. Interfaces, 2017, 9, 38159–38164 CrossRef CAS PubMed.
  47. W. Luo, Y. Wang, L. Wang, W. Jiang, S. L. Chou, S. X. Dou, H. K. Liu and J. Yang, ACS Nano, 2016, 10, 10524–10532 CrossRef CAS PubMed.
  48. M. G. Jeong, H. L. Du, M. Islam, J. K. Lee, Y. K. Sun and H. G. Jung, Nano Lett., 2017, 17, 5600–5606 CrossRef CAS PubMed.
  49. M. Gu, L. R. Parent, B. L. Mehdi, R. R. Unocic, M. T. McDowell, R. L. Sacci, W. Xu, J. G. Connell, P. Xu, P. Abellan, X. Chen, Y. Zhang, D. E. Perea, J. E. Evans, L. J. Lauhon, J. G. Zhang, J. Liu, N. D. Browning, Y. Cui, I. Arslan and C. M. Wang, Nano Lett., 2013, 13, 6106–6112 CrossRef CAS PubMed.
  50. M. Gu, Y. Li, X. Li, S. Hu, X. Zhang, W. Xu, S. Thevuthasan, D. R. Baer, J.-G. Zhang and J. Liu, ACS Nano, 2012, 6, 8439–8447 CrossRef CAS PubMed.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c8nr07207h
These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2018