Double-shelled hollow carbon spheres confining tin as high-performance electrodes for lithium ion batteries

Double-shelled hollow carbon spheres confining tin as high-performance electrodes for lithium ion batteries

Electrochimica Acta 321 (2019) 134672 Contents lists available at ScienceDirect Electrochimica Acta journal homepage: www.elsevier.com/locate/electa...

3MB Sizes 0 Downloads 0 Views

Electrochimica Acta 321 (2019) 134672

Contents lists available at ScienceDirect

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

Double-shelled hollow carbon spheres confining tin as highperformance electrodes for lithium ion batteries Li Sun a, Tiantian Ma a, Jun Zhang a, b, Xiangxin Guo a, c, Chenglin Yan d, **, Xianghong Liu a, b, * a

College of Physics, Qingdao University, Qingdao 266071, China Key Laboratory of Advanced Energy Materials Chemistry (Ministry of Education), Nankai University, Tianjin 300071, China Shanghai Institute of Ceramics, Chinese Academy of Sciences, Shanghai 200050, China d Soochow Institute for Energy and Materials Innovations, College of Energy, Soochow University, Suzhou 215006 China b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 9 July 2019 Received in revised form 8 August 2019 Accepted 9 August 2019 Available online 10 August 2019

Alloy-type materials hold a great potential as the negative electrodes for next generation lithium-ion batteries with high energy and low cost. However, the huge volume expansion of alloy-type materials caused by lithium alloying inevitably leads to poor cycle stability. Herein, we propose a rational design of a sandwich structure of carbon/Sn/carbon hollow spheres by a template-engaged method. The structure effect of the novel carbon/Sn/carbon spheres on the lithium storage performances is elucidated by various means of characterization and electrochemical tests. A stable and high reversible capacity of 1100 mA h g1 is retained after 130 cycles at 0.1 A g1, significantly higher than that (187 mA h g1) of Sn/ carbon hollow spheres. Furthermore, a superior rate capability is obtained for carbon/Sn/carbon spheres, e.g., showing a high capacity of 430 mA h g1 at 5 A g1. The excellent electrochemical properties of carbon/Sn/carbon against Sn/carbon are ascribed to a unique nano-confinement from the double-shelled carbons with very good structure stability and contribution of pseudocapacitive storage of lithium. These results indicate that the sandwich structure of carbon/Sn/carbon is highly effective to design electrode materials with enhanced performances. © 2019 Published by Elsevier Ltd.

Keywords: Negative electrode Hollow structure Long cycling Structure design Tin

1. Introduction Lithium-ion batteries (LIBs) are considered to be one of the best conversion and storage device for various clean energy sources based on their high energy density, long cycle life, and high environmental benignity [1]. Graphite-based materials are widely used as the negative electrode materials in commercial LIBs. However, the low theoretical capacity (372 mA h g1) of graphite greatly limits the development of high energy batteries for utilization in electric vehicles, hybrid vehicles, and solar energy storage equipment [2,3]. Therefore, it is of great significance to search for electrode materials with high reversible capacity and long cycle life. Great efforts have been explored for developing high

** Corresponding author. * Corresponding author. College of Physics, Qingdao University, Qingdao 266071, China. E-mail addresses: [email protected] (C. Yan), [email protected] (X. Liu). https://doi.org/10.1016/j.electacta.2019.134672 0013-4686/© 2019 Published by Elsevier Ltd.

performance negative electrodes [4] from metal oxides [5e10], metal sulfides [11e15], silicon-based materials [16e19], Ge negative electrodes [20e22], and tin-based materials [23e31]. Among them, the alloy-type negative electrode materials such as tin have attracted much attention for their simple manufacture process, low preparation cost and high theoretical capacity (993 mA h g1) [32]. However, the huge volume expansion (300%) of tin during alloying and de-alloying with lithium ions will result in material pulverization and rapid decay of capacity, thus seriously restricting its industrial application [27,32]. To address these issues, the common method is to synthesize Sn/carbon nanocomposite with a designed structure. For example, Huang et al. reported micron-sized hierarchical fibrous bundle made up of [email protected]@C nanofibers which exhibited 580 mA h g1 at 500 mA g1 after 100 cycles [33]. Zhou et al. devised a 2D hybrid structure of [email protected] nanosheets, which can remain a capacity of 620 mA h g1 at 0.8 A g1 after 1000 cycles [34]. Gao et al. reported that the conjoined hollow spherical structure of Sn/C nanocomposite could endure mechanical stress and mitigate pulverization, thereby decreasing the

2

L. Sun et al. / Electrochimica Acta 321 (2019) 134672

contact resistance between the spherical particles and affording a capacity of 616 mA h g1 at 1 A g1 after 1000 cycles [35]. Yu et al. presented the yolk-shell [email protected] nanobox with controllable carbonshell thickness, which can maintain a capacity of 810 mA h g1 at 200 mA g1 after 500 cycles [1]. A yolk-shell structure designed by Kang et al. [29], where Sn nanoparticles are encapsulated in carbon spheres, demonstrated a capacity of 691 mA h g1 at 1 A g1 after 500 cycles. Although these works reported on the extended cycling of Sn negative electrodes, it still remains a challenge to develop an effective protocol to prevent the Sn pulverization in order to obtain a higher reversible capacity and better rate performance. In this work, we report on rational design of sandwich carbon/ Sn/carbon (C/Sn/C) hollow spheres (Fig. 1), where Sn nanoparticle are fully encapsulated between the two concentric carbon shells. The nanospace between the double carbon shells can readily accommodate the volume expansion of Sn on lithiation, which perfectly prevents the exfoliation of Sn from carbon support due to pulverization. When tested as the negative electrode for LIBs, the C/ Sn/C spheres manifest very high reversible capacity of 1100 mA h g1 after 130 cycles at 0.1 A g1 and good cycling stability (920 mA h g1 after 300 cycles at 1 A g1) compared with the Sn/C spheres. The C/Sn/C negative electrode also delivers an outstanding rate performance involving the pseudocapacitive lithium storage, which is elucidated by the kinetic analysis derived from cyclic voltammetry tests. The excellent electrochemical properties indicate the proposed sandwich structure of C/Sn/C is highly promising for application in future LIBs.

washed with deionized water and ethanol, and dried at 80  C. The final product is PDA/SiO2. Preparation of SnO2/PDA/SiO2: 0.4 g of PDA/SiO2 was dispersed into the solution (13 mL of ethanol and 26 mL of distilled water). Then, 0.4 g of Na2SnO.34H2O was added into the mixed solution and stirred for few minutes, the above solution was placed in a 50 mL of Teflon stainless vessel and heated at 180  C for 2 h. The SnO2/PDA/ SiO2 were gained after centrifugation, washing and drying. Preparation of Sandwich Carbon/SnO2/Carbon/SiO2: 0.4 g of the as-prepared SnO2/PDA/SiO2 was added into Tris-buffer (125 mL, 10 mM, PH ¼ 8.5), 0.8 g of dopamine hydrochloride were also mixed in the solution and held stirring for 12 h. Then the suspension was centrifuged and washed several times with distilled water and ethanol and dried at 80  C. Finally, the product was annealed at 500  C for 2 h in an argon atmosphere with a heating rate of 8  C/ min. Preparation of Sandwich Carbon/Sn/Carbon Hollow Spheres: With the aim of etching away the SiO2 template, the Carbon/SnO2/ Carbon/SiO2 were dispersed in 80 mL of 5% HF solution and stirring for 24 h. Then, the powder was annealed at 800  C for 2 h under H2 (10%)/Ar (90%) atmosphere with a heating rate of 4  C/min to get the sandwich Carbon/Sn/Carbon (C/Sn/C) hollow spheres. For comparison, the Sn/Carbon (Sn/C) hollow spheres were synthesized from the SnO2/PDA/SiO2 precursor using the same protocol.

2. Experimental section

The crystal structure of samples were investigated by means of X-ray diffraction (XRD, Rigaku Smartlab) with Cu Ka radiation. The microstructure of the as-prepared samples was examined using scanning electron microscopy (SEM, Zeiss sigma 300). Transmission electron microscopy (TEM) was carried out on JEM-2010. Scanning transmission electron microscopy (STEM) and element mapping were tested on FEI TALOS F200 (American). X-ray photoelectron spectroscopy (XPS) analysis was conducted on Thermo ESCALAB 250. The specific surface area was estimated by BrunauereEmmetteTeller (BET) method and the pore size distribution was investigated via the BarretteJoynereHalenda (BJH) analysis which were measured using an ASAP 2060 by high purity nitrogen as adsorbate at 77 K. Raman spectra were obtained using Renishaw with a laser wavelength of 532 nm. Thermo-gravimetric analysis (TGA) was measured on TG209 (NETZSCH) under an air atmosphere at 10  C/min.

Tetraethylorthosilicate (TEOS), Hydrofluoric acid (HF), ethanol, sodium stannate (Na2SnO.34H2O), and ammonia were purchased from Sinopharm Chemical Reagent Co. Ltd. Dopamine hydrochloride and Trismethyl aminomethane were supplied from Aladdin. All the reagents were used as received without further purification. 2.1. Material preparation Synthesis of SiO2 spheres: SiO2 spheres were prepared by Stober method [36]. In a typical synthesis, 27 mL of ammonia water, 51 mL of ethanol and 75 mL of deionized water were mixed under stirring as solution A. Then, 30 mL of TEOS and 135 mL of ethanol was mixed under stirring as solution B. Pouring B into A and then stirring for 4 h. The products was washed with deionized water and dried by freeze drying. Fabrication of Polydopamine (PDA)/SiO2: 0.4 g of silica spheres were mixed with 0.8 g of dopamine hydrochloride in Tris-buffer (125 mL, 10 mM, PH ¼ 8.5) and stirred for 12 h. The product was

2.2. Materials characterization

2.3. Electrochemical measurements Electrochemical measurements were investigated with CR2025

Fig. 1. Schematic synthesis of Sn/C and C/Sn/C hollow spheres.

L. Sun et al. / Electrochimica Acta 321 (2019) 134672

coin-type cells. The working electrode was fabricated mixing the active materials, carbon black (Super-P) and polyvinylidene fluoride (PVDF) with a weight ratio of 80:10:10. Then the mixture was coated onto a copper foil (10 mm) with a thickness of 20 mm and dried under vacuum overnight at 80  C. The loading mass of active material was about 0.9e1.2 mg/cm2. The lithium foil was used as a counter electrode and the cells were assembled in an Ar-filled glove box with LiPF6 (1 M) in a mixed solution of dimethyl carbonate (DMC), ethylene methyl carbonate (EMC) and ethylene carbonate (EC) with a volume ratio of 1:1:1. The cells were galvanostatically cycled within a potential range of 0.01e3.0 V at a constant temperature of 25  C. Cyclic Voltammetry (CV) measurements were performed within 0.01e3.0 V at a scanning rate of 0.1 mV/s and electrochemical impedance spectra (EIS) was conducted over frequency range of 0.01e100 kHz with a sigal of 5 mV. (VSP, Bio-Logic). 3. Results and discussion The synthesis of Sn/C and C/Sn/C hollow spheres which was conducted by referring to the reported procedures with modifications is illustrated in Fig. 1. The SiO2 spheres are used as the victim template, and PDA and SnO2 serve as the precursor for N-doped carbon and metallic Sn, respectively. The synthetic process mainly involves polymerization of dopamine to give PDA, hydrothermal deposition of SnO2 and thermal treatment under high temperatures to convert PDA to N-doped carbon and reduce SnO2 to metallic Sn. This Sn/carbon material combines the high capacity of Sn, as well as the high conductivity and elastic properties of carbon hollow shells [37]. Compared with Sn/C hollow spheres, the outer carbon layer of C/Sn/C hollow spheres can improve not only the conductivity but also the structure stability of the materials. Due to the confinement of the nanospace between the two carbon shells, the volume change of Sn during lithiation is adequately accommodated, thus preventing the abscission of LixSn particles from the inner carbon shells induced by pulverization. This mechanism is anticipated to afford improved cycling stability of the C/Sn/C negative electrodes with high capacity retention. The XRD patterns of both samples are shown in Fig. 2a. All diffraction peaks of C/Sn/C and Sn/C hollow spheres can be indexed to tetragonal Sn (JCPDS 04-0673 space group: I41/amd, a ¼ b ¼ 5.831 Å, c ¼ 3.182 Å). The peak at 26.61 can be ascribed the SnO2 of Sn/C hollow spheres. No diffraction peaks from carbon can be detected, indicating the carbon shells are amorphous [38]. The XPS survey of the C/Sn/C hollow spheres shows that the composite is composed of Sn, C and N. As can be seen in Fig. 2b, the Sn 3d is fitted into four peaks. The peaks located at 486.72 and 495.37 eV which are indexed to Sn0 and the peaks corresponding to Sn4þ are also detected at 487.3 and 495.83 eV [39,40]. Since no obvious diffraction from SnO2 is discovered by XRD, the XPS signal from tin oxides can be attributed to the partial oxidization of Sn surface [39,41]. Fig. 2c shows the N 1s spectrum which was fitted into three components. The peaks at 402.7, 400.9 and 398.5 eV correspond to the graphitic-N, pyrrolic-N and pyridinic-N, respectively [42]. According to the area percentage of the fitted peaks, the proportion of the corresponding nitrogen species was 4.73%, 51.34% and 43.93%, respectively. The high intensity of graphitic N marks the successful doping of nitrogen in C/Sn/C hollow spheres composites [43], at the same time, pyridine-N and pyrrole-N produce more defects and vacancies in carbon, which has been reported to conduce to improve the conductivity of carbons and produce more active sites within carbon matrix to promote the diffusion of Liþ [44,45]. In Fig. 2d, the high-resolution XPS spectrum of C 1s is deconvoluted into three peaks at 284.8, 285.6 and 287.6 eV, which can be assigned to C]C, C]N and CeN [40,43].

3

The morphology of C/Sn/C and Sn/C hollow spheres is revealed by SEM. As can be seen in Fig. S1a, the Sn/C hollow spheres exhibit a rough surface due to the attachment of melted Sn particles, while C/ Sn/C hollow spheres (Fig. S1b) have a relative smooth surface, which can be ascribed to the coating of outer carbon layers. The C/ Sn/C and Sn/C hollow spheres are further observed by TEM. As shown in Fig. 3a and b, the Sn particles are uniformly attached to the surface of the carbon shells with a mean size of 25 nm, which is consistent with the result calculated by Scherrer equation. Due to the coating of the outer carbon, the Sn particles we observed in C/ Sn/C are not as clear as that in Sn/C. It is seen in Fig. 3d that the size of the Sn particles in C/Sn/C is around 20 nm. Fig. 3c and d shows the TEM of C/Sn/C hollow spheres. In the inset of Fig. 3c, the double carbon shells with a nanospace can be clearly distinguished. In Fig. 3c and d, it is noted that the observation of Sn is highly different from that in Sn/C of Fig. 3a and b. In Fig. 3d, we can see that Sn metals are anchored on the double carbon shells due to melting of Sn in the thermal reduction process. From the STEM and elemental mapping test, we can see the uniform distribution of Sn in the Sn/C (Fig. 3eeg) and C/Sn/C (Fig. 3hej) hollow spheres. As shown in Fig. 4a and b, the N2 adsorption-desorption isotherms of both samples belong to typical type-IV with a hysteresis loop of mesoporous structure [46]. The C/Sn/C hollow spheres have a large BET specific surface area of 157.80 m2/g, which is much higher than that (44.01 m2/g) of Sn/C hollow spheres. The higher surface area of C/Sn/C is probably due to the double-shelled structure of carbon spheres. In addition, the size of Sn particle of C/Sn/C is smaller than that in Sn/C, which might also contribute to a large surface area [40,47]. Fig. S2a shows the Raman spectra of C/Sn/C hollow spheres, reveling the typical D band at 1354 cm1 and G band at 1594 cm1 for carbon materials [48]. The D band suggests that disorders and defects appear in the carbon shells, however, the G band depends on the sp2-hybridized graphitic carbon structure [49]. The intensity ratio of D to G is 0.88, which indicates that a great number of sp2bonded carbon atoms are existing in the N-doped carbon shells. The relative content of Sn and N-doped carbon in the C/Sn/C hollow spheres are obtained by TGA. As illustrated in Fig. S2b, the weight loss attributes to the removal of water before 210  C and burning of carbons after 210  C [1]. It is calculated that the Sn and carbon content in C/Sn/C hollow spheres are about 612.8% and 38.2%, respectively. Inspired by the functions endowed by the unique sandwich structure, the C/Sn/C hollow spheres have been investigated as a negative electrode in LIBs. Fig. 5a displays the CV characteristics of the C/Sn/C spheres at a scan rate of 0.1 mV/s. During the initial cathodic scan, a broad peak at 0.8e1.5 V partially caused the initial irreversible capacity, attributing to the formation of solid electrolyte interface (SEI) film and the reduction of SnO2 to Sn [50,51]. There are three peaks at about 0.37, 0.51 and 0.65 V versus Liþ/Li in the cathodic scan which correspond to the lithiation reaction (Sn þ xLiþ þ xe 4Lix Sn ð0  x  4:4Þ) to form the LixSn, while those anodic peaks at 0.53, 0.63, 0.72 and 0.81 V are associated with the dealloying of LixSn [40,52]. In the CV curves, a small tail at around 3 V is found for the charging process, which is also observed in some works from other groups [33,35,53]. According to the literatures, this plateau can be ascribed to the electrical capacity formed between the electrode and electrolyte [53]. The first and subsequent scans coincide well and there is no obvious change of the peaks position, which indicates good electrochemical stability and the same reaction occurs during the scan. The CV result tentatively confirms the high stability of the C/Sn/C nanostructures. For comparison, the CV profiles of Sn/C hollow spheres are illustrated in Fig. S3a, revealing a very large irreversible capacity between the first and the subsequent scans.

4

L. Sun et al. / Electrochimica Acta 321 (2019) 134672

Fig. 2. (a) XRD patterns of C/Sn/C and Sn/C hollow spheres. XPS spectrum of (b) Sn 3d. (c) N 1s. (d) C 1s of C/Sn/C hollow spheres.

Fig. 3. TEM images of (a, b) Sn/C, and (c, d) C/Sn/C hollow spheres; STEM and element mapping of (eeg) Sn/C and (hej) C/Sn/C hollow spheres.

Fig. 5b manifests the discharge-charge voltage curves for the 1st, 2nd, 20th, 40th, 60th, 80th and 100th cycles of C/Sn/C between 0.01 and 3.00 V at 100 mA g1. The discharge/charge specific capacity for the first cycle is 2267 and 1413 mA h g1, indicating an initial Coulombic efficiency (CE) of 62%. For Sn/C hollow spheres

(Fig. S3b), the initial discharge/charge specific capacity is 1063 and 724 mA h g1 with a CE of 68%. The large irreversible capacity in the first cycle is ascribed to the formation of SEI and some irreversible reactions [54]. The higher initial CE of Sn/C against C/Sn/C is probably due to the high surface area and high N content in the N-

L. Sun et al. / Electrochimica Acta 321 (2019) 134672

5

Fig. 4. N2 adsorption-desorption isotherms of (a) C/Sn/C and (b) Sn/C hollow spheres (the insets are the pore distribution curves).

doped carbon of C/Sn/C [40,55]. Furthermore, after the initial several cycles, the C/Sn/C displays a high and stable discharge capacity around 1000 mA h g1, while the Sn/C material shows very fast decay in capacity. The superior cycling stability of C/Sn/C over Sn/C can also be evaluated in Fig. 5c, showing the cycling performances of two materials at 0.1 A g1. They deliver a charge/discharge capacity of 1413 and 2267 mA h g1 for C/Sn/C, 724 and 1063 mA h g1 for Sn/C, corresponding to initial CE of 62% and 68%, respectively. For the Sn/ C hollow spheres electrode, it exhibits discharge capacity of 1064 mA h g1 in the first cycle and reduces to 187 mA h g1 after 130 cycles. Compared with Sn/C hollow spheres, the C/Sn/C hollow spheres deliver a high discharge capacity of 2267 and retains 1100 mA h g1 with capacity retention of 78% after 130 cycles. For the first 30 cycles, the irreversible capacity should be due to the formation of SEI film and the occurrence of some irreversible reactions such as irreversible reduction of SnO2 to Sn and stable Li2O, (SnO2 þ 4Liþ þ 4e /Sn þ 2Li2 O) and irreversible lithium storage in carbon [31,56,57]. As the cycle goes on, the charge and discharge curves are almost coincident, and the CE is close to 100%. The capacity of C/Sn/C hollow spheres keeps stable with a slight increase. In contrast, the Sn/C hollow spheres are always in a state of decay which is due to the lack of the external protective carbon layer. Furthermore, the high rate cycling performance of the C/Sn/C hollow spheres negative electrode shows in Fig. S4. The C/Sn/C hollow spheres were tested at 1 A g1 for 300 cycles after activating at 0.1 A g1 for 10 cycles. The initial capacity reaches 505 mA h g1 and gradually increases to 920 mA h g1 until 270th cycles and remains stable in the following cycles. The increasing capacity may be ascribed to the activation of the internal Sn particles and the improvement of lithium ion accessibility upon deep cycling [58,59]. In addition, a reversible formation of a polymeric gel-like layer via electrolyte decomposition might also contribute to the improved capacity [1]. The above results indicate that the outer carbon greatly improves the cycle stability and capacity retention of the C/Sn/C hollow spheres, mainly because it can protect the Sn particles fall off from the nanocomposite, and the two carbon layers formed a conductive mesh to accelerates the electron transport of the electrode material. The rate performance of the C/Sn/C and Sn/C hollow spheres at diverse current densities are illustrated in Fig. 5d. The average reversible capacities of C/Sn/C hollow spheres are measured to be 1545, 1049, 863, 725 and 618 mA h g1 which is higher than that (529, 285, 211, 165 and 131 mA h g1) of Sn/C hollow spheres at 100,

200, 500, 1000 and 2000 mA g1, respectively. When the current density returned to 100 mA g1, the capacity of C/Sn/C hollow spheres returned to 1300 mA h g1 which remained 85% of the initial capacity. Meanwhile, the capacity has increased and become more stable in the following rate cycles. To further explore the rate performance of C/Sn/C hollow spheres, we tested it at higher current densities. As revealed by the third round rate cycling, at the higher current rate of 5, 10 and 20 A g1, the corresponding reversible capacities still retain 427, 250 and 92 mA h g1 and the capacity also can recover with the current density returned to 100 mA g1. To gain further insight into the good electrochemical performance of C/Sn/C, the CV behavior was recorded with a scan rate of 0.1 mV/s after rate test. As shown in Fig. S5, the CV curves are substantially coincident, demonstrating that the battery has excellent cycling stability and good reversibility of the electrochemical reactions [60]. From the CV curves after rate test, the cathodic peak from 0.8 to 1.5 V associated with the electrolyte decomposition, which has been observed in Fig. 5a, is seen to disappear. This indicates the formation of stable SEI film in the C/ Sn/C electrode [41]. The decreased intensity of redox peaks suggests a significant contribution of pseudocapacitive storage of lithium ions. It is noted that the C/Sn/C manifests better or comparable electrochemical performance in comparison with other Sn-based materials, as summarized in Table S1. The capacity of C/Sn/C hollow spheres electrode is higher than that of most of the listed samples, while the preparation is relatively simple. The high capacity and excellent rate performance indicate that the proposed C/ Sn/C hollow spheres are highly promising to be used as electrode materials for LIBs. To probe reaction kinetics of Sn/C and C/Sn/C hollow spheres, EIS measurements were performed after cycling at 100 mA g1. The Nyquist plots of both samples are shown in Fig. 6a with the fitted equivalent circuit as an inset. The Sn/C and C/Sn/C hollow spheres exhibit a diagonal line related to the Warburg type semiinfinite diffusion of lithium ions in the low-frequency region and a semicircle owing to the diffusion of lithium ions through the SEI film in the high-frequency region [61]. The C/Sn/C hollow spheres with a smaller semicircle diameter than Sn/C hollow spheres has lower interfacial resistance. Additionally, the specific value of the impedance can be obtained after fitting the spectrum. Rf and Rc is the SEI film resistance and charge-transfer resistance, respectively. According to the fitting results, the Rf and Rc are 4.27 and 9.97 U for C/Sn/C while Sn/C hollow spheres correspond to 11.74 and 60.73 U, respectively, which demonstrates the better kinetics of C/Sn/C

6

L. Sun et al. / Electrochimica Acta 321 (2019) 134672

Fig. 5. (a) CV, (b) Charge-discharge curves of C/Sn/C hollow spheres; (c) Cycling performance of C/Sn/C and Sn/C hollow spheres at a current density of 0.1 A g1; (d) Rate performance of C/Sn/C and Sn/C hollow spheres electrode.

Fig. 6. (a) Electrochemical impedance spectra of C/Sn/C and Sn/C hollow spheres electrodes after cycling. The equivalent circuit model is shown in the inset. (b) The real part of the impedance (Zre) as a function of the reciprocal square root of the angular frequency (u1/2) plot.

L. Sun et al. / Electrochimica Acta 321 (2019) 134672

over Sn/C hollow spheres. The major reason for the improved electronic conductivity would be the second carbon layer. Based on equation (1),

W ¼ su1=2  jsu1=2

(1)

Where s is the Warburg factor, u is the angular frequency. Zre shows a linear relationship with u1/2 (Fig. 6b) and the slope of the line is s. After fitting, it is found that the values of s corresponding to C/Sn/C and Sn/C hollow spheres are 7.78 and 148.52, respectively. The Liþ diffusion coefficient (DLiþ) can be calculated with equation (2) [40].



R2 T 2 2A2 n4 F 4 C 2 s2

(2)

Where R is the gas constant, T is the Kelvin temperature at the time of the experiment, A is the contact area between the electrolyte and the electrode, n is the number of electrons transferred in the chemical reaction, F is the Faraday constant and C is the concentration of lithium ions in C/Sn/C hollow spheres electrode. The calculated results displayed that the DLiþ of the C/Sn/C hollow spheres is 2.28*1010 cm2/s while that is 6.28*1013 cm2/s of the Sn/ C hollow spheres. This is because the double shell hollow structure shortens the Li-ion diffusion distance [62]. To further demonstrate the charge-discharge storage mechanism of the C/Sn/C hollow spheres, we analyzed the pseudocapacitive behavior and calculated the proportion of pseudocapacitive contribution in the capacity. The CV curves at different scan rates from 0.1 to 2 mV/s were displayed in Fig. 7a with similar shapes. It can be determined whether the surface or the diffusion dominates the lithium storage through the relationship between the peak current (i) and the sweep rate ðnÞ according to equations (3) and (4):

7

i ¼ anb

(3)

log i ¼ b  log n þ log a

(4)

Where a is a constant. Log i shows a linear relationship with log n and b is the slope of the line. If the value of b is close to 0.5, it indicates a diffusion-dominated lithium storage, and if b approaches 1, it suggests a surface-controlled storage behavior [63]. As can be seen in Fig. 7b, the b values all exceeded 0.5 at different potentials in the cathodic scan, and when the voltage was 0.75 V, the b value was 0.939, indicating that the C/Sn/C hollow spheres was mainly controlled by surface storage. The proportion of capacitance contribution can be quantified by the following equation [63]:

iðVÞ ¼ k1 n þ k2 n1=2

(5)

The iðVÞ is composed of two parts, one part is the capacitive contribution represented as k1 n and other part is the diffusioncontrolled contribution expressed as k2 n1=2 . Fig. 7c exhibits the ratio of pseudocapacitive contribution is 31%, 40%, 46.7%, 57%, 60.2%, 63.2% and 71.4% at scan rate of 0.1, 0.2, 0.4, 0.8, 1.2, 1.6, 2.0 mV/s, respectively. We can see that as the scan rate rises, the capacitance contribution also increases and reveals a maximum value of 71.4% at 2 mV/s. Fig. 7d shows the area of capacitance contribution accounts for 71.4% of the lithium storage in the surface layer when the scan rate is 2 mV/s. Given the excellent cycling stability and high reversible capacity of C/Sn/C, it is highly interesting to figure out the underlying mechanism. Post-XRD analysis has been performed on the C/Sn/C and Sn/C materials after the cycle test. As illustrated in Fig. S6, the main characteristic peaks of Sn can still be clearly detected. Meanwhile, two strong peaks appear at 43 and 50 , which are

Fig. 7. (a) CV curves at different scan rates from 0.1 to 2.0 mV/s. (b) b values at different potentials for cathodic scans (the inset is the fitted lines between log i and log v and the slope of the line is the b value). (c) The contribution ratio of diffusion and capacitance at different scan rates. (d) Capacitance and diffusion contributions at 2.0 mV/s and the shadow region are the capacitive contribution of C/Sn/C hollow spheres electrode.

8

L. Sun et al. / Electrochimica Acta 321 (2019) 134672

attributed to Cu diffraction derived from the copper foil to which the active materials adhere. Post-TEM observation has been carried out to check the structure of C/Sn/C and Sn/C after the cycle test. As shown in Fig. S7a, the sandwich and hollow structure of C/Sn/C is well maintained. By contrast, the sphere structure of Sn/C in Fig. S7b is hard to preserve due to the shedding of the Sn particle from Sn/C electrode in the repeated discharge/charge process. These results demonstrate strong evidence for the significantly improved cycling stability and high capacity of C/Sn/C, which are enabled by the good structure stability endowed by the double carbon shells. Therefore, the superior lithium storage performances of C/Sn/C over Sn/C hollow spheres are reasonably attributed to the double carbon shells with rich N-doping functionality, which can enhance the electronic conductivity and adequately buffer the volume changes of Sn to improve the structure stability. 4. Conclusion To sum up, C/Sn/C hollow spheres with a sandwich structure, where metallic Sn is well-confined in between two N-doped carbon shells, are designed and investigated as a negative electrode material for LIBs. The C/Sn/C electrode combines the merits of the high capacity of Sn, high conductivity of carbon, large contact area for the electrolyte of the hollow structure and the pseudocapacitive contribution. Consequently, the C/Sn/C spheres exhibit a high reversible capacity (1100 mA h g1 after 130 cycles at 0.1 A g1) with very good cycling stability and excellent rate performance. The structure design proposed in this paper might be used to the manipulation of other negative electrode materials for next generation LIBs with better performances. Acknowledgment This work is financially supported by the National Natural Science Foundation of China (No. 51602167 and 21601098), Shandong Provincial Science Foundation (ZR2016EMB07 and ZR2017JL021) and Key Research and Development Program (2018GGX102033), and Qingdao Applied Fundamental Research Project (16-5-1-92-jch and 17-1-1-81-jch), and “Distinguished Taishan Scholar” project. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.electacta.2019.134672. References [1] H. Zhang, X. Huang, O. Noonan, L. Zhou, C. Yu, Adv. Funct. Mater. 27 (2017) 1606023. [2] H. Ying, W.Q. Han, Adv. Sci. 4 (2017) 1700298. [3] M. Wang, F. Zhang, C.-S. Lee, Y. Tang, Adv. Energy Mater. 7 (2017) 1700536. [4] Y. Zhong, M. Yang, X. Zhou, Z. Zhou, Mater. Horiz. 2 (2015) 553e566. [5] H. Song, J. Su, C. Wang, Adv. Energy Mater. 9 (2019) 1900426. [6] X. Liu, T. Ma, L. Sun, Y. Xu, J. Zhang, N. Pinna, ChemSusChem 11 (2018) 1321e1327. [7] M. Li, C. Ma, Q.-C. Zhu, S.-M. Xu, X. Wei, Y.-M. Wu, W.-P. Tang, K.-X. Wang, J.S. Chen, Dalton Trans. 46 (2017) 5025e5032. [8] T. Ma, X. Liu, L. Sun, Y. Xu, L. Zheng, J. Zhang, J. Energy Chem. 34 (2019) 43e51. [9] Q. Hao, J. Wang, C. Xu, J. Mater. Chem. A 2 (2014) 87e93. [10] Q. Hao, Y. Yu, D. Zhao, C. Xu, J. Mater. Chem. A 3 (2015) 15944e15950. [11] J. Xu, T. Lawson, H.B. Fan, D.W. Su, G.X. Wang, Adv. Energy Mater. 8 (2018) 1702607. [12] T. Ma, X. Liu, L. Sun, Y. Xu, L. Zheng, J. Zhang, Electrochim. Acta 293 (2019) 432e438. [13] L. Sun, X. Liu, T. Ma, L. Zheng, Y. Xu, X. Guo, J. Zhang, Solid State Ion. 329 (2019) 8e14. [14] J. Xia, L. Liu, J. Xie, H. Yan, Y. Yuan, M. Chen, C. Huang, Y. Zhang, S. Nie, X. Wang, Electrochim. Acta 269 (2018) 452e461.

[15] T. Ma, L. Sun, Q. Niu, Y. Xu, K. Zhu, X. Liu, X. Guo, J. Zhang, Electrochim. Acta 300 (2019) 131e137. [16] G.R. Zheng, Y.X. Xiang, L.F. Xu, H. Luo, B.L. Wang, Y. Liu, X. Han, W.M. Zhao, S.J. Chen, H.L. Chen, Q.B. Zhang, T. Zhu, Y. Yang, Adv. Energy Mater. 8 (2018) 1801718. [17] A. Mukanova, A. Jetybayeva, S.T. Myung, S.S. Kim, Z. Bakenov, Mater. Today Energy 9 (2018) 49e66. [18] X. Liu, J. Zhang, W. Si, L. Xi, B. Eichler, C. Yan, O.G. Schmidt, ACS Nano 9 (2015) 1198e1205. [19] C.X. Xu, Q. Hao, D.Y. Zhao, Nano Res 9 (2016) 908e916. [20] Y. Yu, C. Yan, L. Gu, X. Lang, K. Tang, L. Zhang, Y. Hou, Z. Wang, M.W. Chen, O.G. Schmidt, J. Maier, Adv. Energy Mater. 3 (2013) 281e285. [21] J. Liu, K. Song, C. Zhu, C.-C. Chen, P.A. van Aken, J. Maier, Y. Yu, ACS Nano 8 (2014) 7051e7059. [22] Q. Hao, Q. Liu, Y. Zhang, C. Xu, J. Hou, J. Colloid Interface Sci. 539 (2019) 665e671. [23] Q. Hao, J. Hou, J. Ye, H. Yang, J. Du, C. Xu, Electrochim. Acta 306 (2019) 427e436. [24] C. Wu, J. Maier, Y. Yu, Adv. Funct. Mater. 25 (2015) 3488e3496. [25] H. Wang, X. Jiang, Y. Chai, X. Yang, R. Yuan, J. Power Sources 379 (2018) 191e196. [26] Y. Cheng, J. Huang, J. Li, Z. Xu, L. Cao, H. Qi, J. Power Sources 324 (2016) 447e454. [27] Y. Lou, H. Di, C. Li, C. Liang, Y. Yu, Z. Shi, D. Zhang, X.-B. Chen, S. Feng, Electrochim. Acta 318 (2019) 542e550. [28] C. Kim, K.-Y. Lee, I. Kim, J. Park, G. Cho, K.-W. Kim, J.-H. Ahn, H.-J. Ahn, J. Power Sources 317 (2016) 153e158. [29] Y.J. Hong, Y.C. Kang, Small 11 (2015) 2157e2163. [30] X. Chang, Z. Liu, B. Sun, Z. Xie, X. Zheng, J. Zheng, X. Li, Electrochim. Acta 267 (2018) 1e7. [31] W. An, J. Fu, S. Mei, L. Xia, X. Li, H. Gu, X. Zhang, B. Gao, P.K. Chu, K. Huo, J. Mater. Chem. A 5 (2017) 14422e14429. [32] X. Zhou, L. Yu, X.-Y. Yu, X.W.D. Lou, Adv. Energy Mater. 6 (2016) 1601177. [33] X. Wang, K. Gao, X. Ye, X. Huang, B. Shi, Chem. Eng. J. 344 (2018) 625e632. [34] Y.R. Zhong, M. Yang, X.L. Zhou, J.P. Wei, Z. Zhou, Part. Part. Syst. Charact. 32 (2015) 104e111. [35] W. An, J. Fu, S. Mei, L. Xia, X. Li, H. Gu, X. Zhang, B. Gao, P.K. Chu, K. Huo, J. Mater. Chem. A 5 (2017) 14422e14429. [36] Z.T. Lin, Y.B. Wu, Y.G. Bi, J. Nanoparticle Res. 20 (2018) 304. [37] W.-M. Zhang, J.-S. Hu, Y.-G. Guo, S.-F. Zheng, L.-S. Zhong, W.-G. Song, L.J. Wan, Adv. Mater. 20 (2008) 1160e1165. [38] P. Wu, N. Du, H. Zhang, C.X. Zhai, D.R. Yang, ACS Appl. Mater. Interfaces 3 (2011) 1946e1952. [39] J. Zhu, D. Wang, L. Cao, T. Liu, J. Mater. Chem. A 2 (2014) 12918e12923. [40] X. Chang, T. Wang, Z. Liu, X. Zheng, J. Zheng, X. Li, Nano Res 10 (2017) 1950e1958. [41] Z.Q. Zhu, S.W. Wang, J. Du, Q. Jin, T.R. Zhang, F.Y. Cheng, J. Chen, Nano Lett. 14 (2014) 153e157. [42] H.W. Song, N. Li, H. Cui, C.X. Wang, Nano Energy 4 (2014) 81e87. [43] D. Su, M. Cortie, G. Wang, Adv. Energy Mater. 7 (2017) 1602014. [44] L.G. Bulusheva, A.V. Okotrub, A.G. Kurenya, H.K. Zhang, H.J. Zhang, X.H. Chen, H.H. Song, Carbon 49 (2011) 4013e4023. [45] G. Wu, C.S. Dai, D.L. Wang, D.Y. Li, N. Li, J. Mater. Chem. 20 (2010) 3059e3068. [46] H. Woo, S. Wi, J. Kim, J. Kim, S. Lee, T. Hwang, J. Kang, J. Kim, K. Park, B. Gil, S. Nam, B. Park, Carbon 129 (2018) 342e348. [47] Y.C. Liu, N. Zhang, L.F. Jiao, Z.L. Tao, J. Chen, Adv. Funct. Mater. 25 (2015) 214e220. [48] Q. Tian, Z. Zhang, L. Yang, S.-i. Hirano, J. Mater. Chem. A 2 (2014) 12881e12887. [49] C. Fu, G. Zhao, H. Zhang, S. Li, Int. J. Electrochem. Sci. 8 (2013) 6269e6280. [50] J. Hassoun, G. Derrien, S. Panero, B. Scrosati, Adv. Mater. 20 (2008) 3169e3175. [51] L. Zhang, H.B. Wu, B. Liu, X.W. Lou, Energy Environ. Sci. 7 (2014) 1013e1017. [52] R.E.A. Ardhi, G. Liu, T. Minh Xuan, C. Hudaya, J.Y. Kim, H. Yu, J.K. Lee, ACS Nano 12 (2018) 5588e5604. [53] X. Chang, Z. Liu, B. Sun, Z. Xie, X. Zheng, J. Zheng, X. Li, Electrochim. Acta 267 (2018) 1e7. [54] B.K. Cao, Z.Q. Liu, C.Y. Xu, J.T. Huang, H.T. Fang, Y. Chen, J. Power Sources 414 (2019) 233e241. [55] X. Liu, J. Zhang, S. Guo, N. Pinna, J. Mater. Chem. A 4 (2016) 1423e1431. [56] X. Wang, J.Y. Hwang, S.T. Myung, J. Hassoun, Y.K. Sun, ACS Appl. Mater. Interfaces 9 (2017) 23723e23730. [57] J. Oh, J. Lee, Y. Jeon, J.M. Kim, K.-d. Seong, T. Hwang, S. Park, Y. Piao, Chemelectrochem 5 (2018) 2098e2104. [58] C.D. Wang, Y. Li, Y.S. Chui, Q.H. Wu, X.F. Chen, W.J. Zhang, Nanoscale 5 (2013) 10599e10604. [59] X. Zhou, L.-J. Wan, Y.-G. Guo, Adv. Mater. 25 (2013) 2152e2157. [60] J.G. Wang, H.Y. Liu, H.Z. Liu, Z.H. Fu, D. Nan, Chem. Eng. J. 328 (2017) 591e598. [61] M. Mao, F. Yan, C. Cui, J. Ma, M. Zhang, T. Wang, C. Wang, Nano Lett. 17 (2017) 3830e3836. [62] J. Sun, C. Lv, F. Lv, S. Chen, D. Li, Z. Guo, W. Han, D. Yang, S. Guo, ACS Nano 11 (2017) 6186e6193. [63] T. Brezesinski, J. Wang, S.H. Tolbert, B. Dunn, Nat. Mater. 9 (2010) 146e151.