Synthesis and electrochemical performance of a coaxial [email protected] nanocomposite as a high-capacity anode material for lithium-ion batteries

Synthesis and electrochemical performance of a coaxial [email protected] nanocomposite as a high-capacity anode material for lithium-ion batteries

Accepted Manuscript Title: Synthesis and electrochemical performance of a coaxial [email protected] nanocomposite as a high-capacity anode material for lith...

586KB Sizes 0 Downloads 32 Views

Accepted Manuscript Title: Synthesis and electrochemical performance of a coaxial [email protected] nanocomposite as a high-capacity anode material for lithium-ion batteries Author: Fushan Geng Anbao Yuan Jiaqiang Xu PII: DOI: Reference:

S0013-4686(16)31943-0 http://dx.doi.org/doi:10.1016/j.electacta.2016.09.055 EA 27985

To appear in:

Electrochimica Acta

Received date: Revised date: Accepted date:

29-6-2016 9-9-2016 9-9-2016

Please cite this article as: Fushan Geng, Anbao Yuan, Jiaqiang Xu, Synthesis and electrochemical performance of a coaxial [email protected] nanocomposite as a high-capacity anode material for lithium-ion batteries, Electrochimica Acta http://dx.doi.org/10.1016/j.electacta.2016.09.055 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis and electrochemical performance of a coaxial [email protected] nanocomposite as a high-capacity anode material for lithium-ion batteries

Fushan Geng, Anbao Yuan*[email protected], Jiaqiang Xu

NEST Lab, Department of Chemistry, College of Sciences, Shanghai University, Shanghai 200444, P.R. China

*

Corresponding author: Tel.: +86 21 66138003

Graphical Abstract

Research Highlights



A [email protected] nanocomposite material is synthesized by hydrothermal reaction.



The composite has the features of coaxial structure and improved electric conduction.



The composite exhibits high performance of Li-storage as a high-capacity anode.



The synthetic method has the advantages of cost saving and environmental benignity.



A dissolutionrecrystallization mechanism is proposed for the hydrothermal reaction.

Abstract Transition metal oxides are considered promising high-capacity anode materials for lithium ion batteries (LIBs). However, their intrinsic low electric conductivity and large volumetric expansion/contraction during lithiation/delithiation can cause fast capacity degradation. Hence, modification of the electrode materials by appropriate structural design is important to their applications. Here, we present study on design, synthesis and Li-storage performance of a vapor grown carbon fiber (VGCF) enhanced [email protected] coaxial-cable nanocomposite anode material for LIBs. This material is synthesized via a novel two-step strategy involving a crucial hydrothermal reaction between ZnO and prefabricated [email protected]-MnO2 composite in pure water at 180 C for 12 h. This composite material exhibits high specific capacity, good rate performance and excellent cycling performance. The exhibited high performance should be attributed to the improved electric conductance due to the incorporation of VGCFs and the particular architecture of one-dimensional nanocomposite. Additionally, the hydrothermal reaction mechanism is specially

focused and preliminarily studied. A dissolutionrecrystallization mechanism is proposed for the hydrothermal reaction.

Keywords: [email protected] composite; Anode; Lithium ion battery; Lithium storage performance; Hydrothermal reaction

1. Introduction R & D of next-generation greener and more sustainable batteries is an exigent mission for large-scale electrical energy storage applications toward electric vehicles, solar and wind power generation, and power grid balance, etc. High-capacity Li-air (Li-O2) batteries, Li-S batteries and high-performance rechargeable lithium-ion batteries (LIBs) are considered the promising candidates [15]. However, there are some challenges facing these technologies in respect to their practical applications [6]. High energy/power density and long cycle life are generally regarded as the technically essential issuers for next-generation LIBs, especially for applications in electric vehicles. LIBs with high-capacity electrodes such as Si anode [7], can obtain higher energy and power densities. Since Tarascon and co-workers reported the nanoparticles of transition-metal oxides (MO, where M is Co, Ni, Cu or Fe) as high-capacity anode materials for LIBs in 2000 [8]. single- and mixed-transition metal oxides have attracted great attentions due to their much higher theoretical capacities (typically ca. 7001000 mAh g1) based on the so-called conversion reaction mechanism over the commercial graphite (theoretical capacity of 372 mAh g1) [912].

ZnMn2O4, a binary transition metal oxide with tetrahedral spinel structure, has been extensively investigated as a high-capacity anode material for LIBs in recent years [1316]. Zn-Mn mixed metal oxides are good choice as anode materials, owing to their high specific capacities, abundant resources, low cost, low toxicity, and relatively lower delithiation potential (ca. 1.2–1.5 V vs. Li+/Li) than many metal oxides. A specially selected metal oxide would be employed as a high-capacity anode coupled with a high-voltage cathode to construct a full LIB with higher energy and power densities like the prototype ZnMn2O4/LiMn1.5Ni0.5O4 full battery [17]. Most recently, the electrochemical performance of an alternative Zn-Mn metal oxide, namely, cubic spinel ZnMnO3, was first reported as a high-capacity anode material for LIBs by Liu et al [18]. In Liu’s work [18], the porous ZnMnO3 spherulites material was synthesized via a carbonate precursor. This material can deliver an initial discharge capacity of 1294 mAh g1 at the current density of 500 mA g1, which is higher than its theoretical capacity of 1117 mAh g1 (based on the conversion reaction and the subsequent alloying reaction between Zn and Li) due to the first-cycle side reaction of solid electrolyte interphase (SEI) formation. In the first charge process, a reversible charge capacity of ca. 825 mAh g1 is archived. After 150 cycles, the reversible charge capacity is increased to ca. 879 mAh g1, exhibiting a good cycling performance. We also conducted experimental work on preparation and electrochemical behavior of cubic ZnMnO3 as an anode material for LIBs since earlier 2014. ZnMnO3 has a higher theoretical lithiation capacity than ZnMn2O4 because of the higher valence state of Mn in ZnMnO3. Nevertheless, ZnMnO3 has a low electric conductivity like as many metal oxides. Besides, the lithiation/delithiation process based on conversion reaction can cause severe volume change and hence failure of electric contact of the electrode,

resulting in fast capacity degradation upon repeated cycling. Based on relevant literatures and our work, we think that the electrochemical Li-storage performance of ZnMnO3 electrode material would be improved by a proper design and construction of its architecture. Design and fabrication of a particularly structured ZnMnO3/carbon nanocomposite would be one of the best solutions to this problem. In fact, various carbonaceous conductive agents have been employed to fabricate metal oxide/carbon composite electrode materials, e.g., carbon nanotube (CNT) [19,20], reduced graphene oxide (RGO) [21], nitrogen-doped graphene aerogel (NDGA) [22], vapor grown carbon fiber (VGCF) [23], and ordered mesoporous carbon (OMC) [11]. However, it is not easy to successfully prepare ZnMnO3/carbon composites by conventional methods. This is because that ZnMnO3 phase is commonly formed in the final-step calcination process in air atmosphere at a higher temperature. However, under these conditions, the carbon component would be oxidized. Herein, we demonstrate a facile two-step synthetic strategy to prepare a coaxial-cable [email protected] nanocomposite. Firstly, a coaxial [email protected]-MnO2 nanocomposite was obtained by in-situ redox reaction between the surface carbons of VGCFs and potassium permanganate solution under conventional conditions. Then, the obtained [email protected]-MnO2 composite was used as a manganese precursor as well as a structural template to synthesize coaxial [email protected] composite via a simple hydrothermal reaction between ZnO and the [email protected]-MnO2 in neutral pure water. The reactants are merely two water-insoluble solid metal oxides (ZnO and -MnO2). Besides, the hydrothermal reaction medium is merely neutral pure water without addition of any a mineralizer or acid/alkali. This reaction is somewhat special and interesting. To understand the essence of the reaction, the hydrothermal reaction mechanism was also studied in this paper. The electrochemical lithium-storage

performance of the [email protected] composite was investigated as an anode material for LIBs in comparison with that of the pristine ZnMnO3 synthesized under identical hydrothermal conditions. As expected, the electrochemical performance of the [email protected] composite is much superior to that of the pristine ZnMnO3. This method for preparation of the [email protected] composite material may be scaled up for large scale production. First, the involved raw materials are merely KMnO4, ZnO and VGCFs, which are not costly (the required amount of VGCFs in the composite is very small). Second, the conventional reaction between KMnO4 and VGCFs and the subsequent hydrothermal reaction between the intermediate [email protected]-MnO2 and ZnO are simple and could be scaled up, and the required temperature for the hydrothermal reaction is relatively low compared with conventional high temperature solid reaction. Third, the whole fabrication process is environmentally friendly. 2. Experimental section 2.1. Chemicals All the chemicals used herein are of analytical grade purchased from Sinopharm Chemical Reagent Co., Ltd. 2.2. Synthesis of ZnO ZnO was prepared by a solgel method. Zn(NO3)2·6H2O and citric acid monohydrate (C6H8O7·H2O) with a molar ratio of 1:2 were dissolved in deionized water (DI water) under constant stirring at 80 °C till formation of a white sol and then a thick gel. The gel was dried at 120 oC for 12 h and then annealed at 500 °C for 2 h in air atmosphere. Thus, the ZnO powder is obtained. 2.3. Synthesis of -MnO2 δ-MnO2 was prepared according to the work of Wang et al [24]. 100 mL of KMnO4 (0.2 mol L1) and 100 mL of ascorbic acid (0.025 mol L1) solutions were

prepared first. Then the ascorbic acid solution was dribbled into the KMnO4 solution placed in an ice-water bath at a speed of 0.75 mL min1 under magnetic stirring. The final pH value of the mixed solution was approximately 2. The reaction continued for 30 min. After reaction, the precipitate was filtered, washed, and dried at 80 °C. The chemical formula of the product is K0.205MnO2·1.47H2O (determined by element analysis). This δ-MnO2 is denoted as δ-MnO2 (I) and was used as a manganese precursor for hydrothermal synthesis of ZnMnO3 comparative anode material. Besides, for the purpose of study of hydrothermal reaction mechanism, another δ-MnO2 sample (denoted as δ-MnO2 (II)) was also prepared via a hydrothermal method (described in Supporting Information). 2.4. Synthesis of ZnMnO3 ZnMnO3 samples were prepared by hydrothermal reaction. For a typical synthesis, 0.001 mol of as-prepared ZnO was mixed with 0.001 mol of as-prepared -MnO2 (I) and ground in a mortar for 15 min, and then the mixture was moved into a Teflon-lined stainless steel autoclave (40 mL) with addition of 30 mL of DI water. The autoclave was heated to 180 °C and was kept at this temperature for 12 h in an oven. After cooling to room temperature, the hydrothermal product was collected, filtered, washed, and dried at 80 °C. Thus, the ZnMnO3 material was obtained. This ZnMnO3 was investigated as a comparative anode material for lithium ion batteries. For special synthesis (for the purpose of study of hydrothermal reaction mechanism), the pH of the hydrothermal medium was adjusted to 6−14 by adding nitric acid or KOH, or the hydrothermal reaction time was controlled within 1−48 h. Specially, for the reaction with a short duration (short dwell time) such as 1 or 2 h, the autoclave was quenched (quick cooling) to room temperature by water-cooling after hydrothermal reaction.

2.5. Synthesis of [email protected]-MnO2 composite Vapor grown carbon fibers (VGCFs, Showa Denko Carbon Inc., Japan) was pretreated by refluxing in HNO3 solution (3 mol L1) at 120 °C for 3 h, followed by repeated washing and then dried at 70 °C for 12 h. Given amount of the pretreated VGCFs was added to a KMnO4 solution (with mass ratio of KMnO4/VGCFs = 8:1) under magnetic stirring. The suspension was adjusted to pH2 by glacial acetic acid, and then heated to 80 °C and kept at this temperature for in situ redox reaction between KMnO4 and VGCFs. After reacting for ca. 6 h, the purple color of the suspension faded out. Then, the product was collected, filtered, washed and dried at 80 °C. 2.6. Synthesis of [email protected] composites [email protected] composites were synthesized by hydrothermal method. For a typical synthesis, 0.001 mol of as-prepared [email protected]-MnO2 (based on -MnO2) and 0.001 mol of as-prepared ZnO were mixed and milled in a mortar for 15 min. Then, the mixture was moved into a Teflon-lined stainless steel autoclave (40 mL) with addition of 30 mL of DI water. The autoclave was then heated to 180 °C and kept at this temperature for 12 h in an oven. After cooling to room temperature, the hydrothermal product was collected, filtered, washed and dried at 80 °C. The mass content of VGCFs in [email protected] composite is of ca. 6.22 wt% (estimated by matter and mass balance, see Supporting Information). The electrochemical performance of this composite was studied as an anode material for lithium ion batteries. For special synthesis (only for the study of hydrothermal reaction mechanism), the dwell time (reaction duration) was controlled within 1−4 h. In these cases, the autoclave was quenched to room temperature by water-cooling after hydrothermal reaction.

2.7. Characterization of materials X-ray diffraction (XRD) patterns were collected at a scan rate of 6° min1 on a Rigaku D/max 2200 X-ray diffractometer (Cu Kα radiation, 40 kV/40 mA). X-ray photoelectron spectroscopy studies were conducted on an X-ray photoelectron spectroscope (XPS, ThermoFisher Scientific ESCALAB 250Xi) with monochromatic Al Kα radiation. Morphological and microstructural observations were performed using a field-emission scanning electron microscope (FE-SEM, JEOL JSM6700F) and a high-resolution transmission electron microscope (HRTEM, JEOL JEM2010 F). Thermogravimetric analysis was operated on a thermogravimetric analyzer (Netzsch TG 209 F1 Libra®) in oxygen atmosphere with a heating rate of 10 C min1 from ambient temperature to 800 C. 2.8. Electrochemical measurements Working electrode was fabricated by mixing the active material (ZnMnO3 or [email protected]), acetylene black conductive agent and sodium carboxymethyl cellulose (NaCMC) binder in the weight ratio of 75:15:10 with a few drops of DI water, and then the mixed slurry was coated onto a copper foil current collector and dried at 90 °C overnight in vacuum. CR2016 coin-type half-cell was assembled in an argon filled glovebox (MIKROUNA Super 1220/750, Mikrouna (China) Co., Ltd., with O2 and H2O concentration <1 ppm) using the fabricated working electrode, metallic lithium foil electrode, separator (Celgard 2300), and 1 M LiPF6 dissolved in a mixture of ethylene carbonate (EC), dimethyl carbonate (DMC) and methyl ethyl carbonate (MEC or EMC) (1:1:1 V/V/V) as electrolyte. The active material (ZnMnO3 or [email protected]) loading in the working electrodes is of ca. 1 mg (ca. 0.9 mg cm2). All the electrochemical measurements were conducted on the assembled CR2016 coin-type cells at ambient temperature. Cyclic voltammetry (CV) and

electrochemical impedance spectroscopy (EIS) measurements were performed on CHI Electrochemical Workstation (CHI 660C). Specifically, EIS measurements were measured at open potential with AC amplitude of 10 mV over a frequency range from 105 to 102 Hz. Galvanostatic discharge/charge tests were carried out on Land CT2001A autocycler (China). 3. Results and discussion 3.1. Structural and morphological studies The synthesis process of the pristine ZnMnO3 and the coaxial [email protected] composite is illustrated in Fig. 1. The products are synthesized via two steps. In the first step, δ-MnO2 (or [email protected]δ-MnO2) was obtained by redox reaction in solution. In the second step, the pristine ZnMnO3 (or the coaxial [email protected] composite) was prepared by hydrothermal reaction between the prefabricated δ-MnO2 (or [email protected]δ-MnO2) and ZnO, which are premixed together by grinding before hydrothermal reaction. The results of thermogravimetric analysis (TGDTG curves) for VGCF, [email protected]δ-MnO2 and [email protected] samples are shown in Fig. S1 (Supporting Information). From the TGA results, the VGCF content in the [email protected] composite can be derived to be of ca. 6.2 wt%, which is consistent with that estimated by matter and mass balance (see Supporting Information). The X-ray diffraction (XRD) patterns and the transmission electron microscope (TEM) images of the precursors (δ-MnO2 and [email protected]δ-MnO2) are presented in Fig. S2. As can be seen in XRD patterns (Fig. S2a), the crystallinity of pure δ-MnO2 is obviously poor than that of the δ-MnO2 in [email protected]δ-MnO2 composite, suggesting that the layer structure of pure δ-MnO2 is somewhat disordered. From the TEM and SEM images (Fig. S2bS2d) we can see that the pure δ-MnO2 is aggregates of nanoparticles, and while the [email protected]δ-MnO2 composite is coaxial structure with

VGCF core and δ-MnO2 shell. The shell is composed of many self-assembled δ-MnO2 nanosheets around the one-dimensional VGCF core. Before hydrothermal reaction, the [email protected]δ-MnO2 was mixed with ZnO powder by grinding. After grinding, some ZnO particles are adhered to the surface of the [email protected]δ-MnO2 (see Fig. S3). However, as shown, some fractions of the VGCF cores are uncovered by δ-MnO2 after grinding. Through hydrothermal reaction, the δ-MnO2 shell will transform into ZnMnO3 nanocrystallines. Of course, these uncovered parts of highly conductive VGCFs could provide direct electric contact between the active material particles, and hence may have a positive effect on the electrochemical performance. The phase structure of hydrothermal reaction products is determined by XRD analysis (Fig. 2a). As can be seen, all the diffraction peaks of pristine ZnMnO3 are well in accordance with the standard diffractions of cubic spinel ZnMnO3 (PDF 191461) and the diffraction pattern of ZnMnO3 reported recently [18]. As to the [email protected] composite, there are additional weak peaks appeared except for the predominant ZnMnO3 phase. These weak peaks should be assigned to tetragonal spinel ZnMn2O4 (PDF 241133), indicating that there is a small amount of ZnMn2O4 phase coexisting in the composite. This may be related to reduction of a small amount of Mn4+ to Mn3+ caused by oxidation of VGCFs under hydrothermal conditions. In addition, the weak peak occurred at ca. 26 should be ascribed to the characteristic diffraction of VGCFs (see Fig. S2a). TEM image (Fig. 2b) shows that the ZnMnO3 particles are mainly sphere-like nano/submicro particles with a particle size distribution from ca. 50 to 200 nm. SEM image (Fig. 2c) and TEM image (Fig. 2d) show the coaxial-cable structure of [email protected] composite. As shown, the composite consists of a VGCF core and a ZnMnO3 shell. The shell is composed of many interconnected nanoparticles with the

sizes of ca. 40100 nm, as can be seen clearly in the inset of Fig. 2d. High-resolution TEM (HRTEM) image (Fig. 2e) shows clearly the lattice fringes along direction of (110) plane, and the inset displays the selected area electron diffraction (SAED) pattern facing [110] zone axis, demonstrating good crystallinity of the ZnMnO3 nanocrystals in the composite. XPS survey and Mn 2p3/2 spectra of the pristine ZnMnO3 and [email protected] composite are depicted in Fig. 3. Except for the detected Zn, Mn and O elements in the survey spectra of the two samples, a strong C 1s signal is also detected in the survey spectrum of the [email protected] composite due to the existence of VGCFs. It can be seen from the fitted spectra of Mn 2p3/2 that the Mn(IV) 2p3/2 at ca. 642.2 eV can be split into two excitations which are lower and higher than 642.2 eV, respectively. This may be associated with the Mn(IV) occupation at different lattice sites in the defect spinel structure of ZnMnO3 and/or the existence of a small amount of Mn(III). Besides, the excitation at ca. 640.3 eV should be assigned to Mn(II) in ZnMnO3. By comparison, the Mn(II) content in the [email protected] composite is somewhat higher than that in the pristine ZnMnO3. This may be due to the reduction of high-valence Mn caused by oxidation of carbon (VGCFs) under hydrothermal conditions. 3.2. Electrochemical performance Electrochemical properties of the [email protected] composite as an anode material for LIBs in comparison with those of the pristine ZnMnO3 were investigated by cyclic voltammetry (CV), galvanostatic charge/discharge and electrochemical impedance spectroscopy (EIS) methods. As we know, ZnO, MnO2 and binary metal oxide ZnMn2O4 have been intensively studied as the anode materials for LIBs. Here,

the discharge and charge reaction of the binary metal oxide ZnMnO3 should be expressed as follows [18]. ZnMnO3 + 6Li+ + 6e  Zn + Mn + 3Li2O

(1)

Zn + Li+ + e  LiZn

(2)

LiZn + Mn + 2Li2O  ZnO + MnO + 5Li+ + 5e

(3)

As shown in Fig. 4a, in the first cathodic scan process, the reduction peak at 1.428 V should be ascribed to the reduction of Mn(IV) to Mn(II) [18], and the sharp peak at 0.287 V should be attributed to the reduction of Zn(II) and Mn(II) to metallic Zn and Mn, corresponding to reaction (1). The small cuspate peak approaching 0 V should be assigned to the alloying of Zn with Li, corresponding to reaction (2). In the first anodic scan process, the oxidation peaks at ca. 1.15 and 1.45 V should be assigned to oxidation of Zn and Mn to ZnO and MnO, respectively. Correspondingly, in the second cathodic scan process, the observed peaks at ca. 0.57 and 0.31 V should be appointed to reverse reduction of MnO and ZnO, respectively. Since the first anodic scan, the CV profiles achieved in the subsequent cycles are duplicated time after time according to reaction (3). The CVs of [email protected] composite (Fig. 4b) are essentially identical to those of pristine ZnMnO3 except for the features in the low-potential region. For [email protected], the cathodic peak approaching 0 V is more acute, and in the reverse scan an anodic shoulder peak is occurred at ca. 0.2 V. These are associated with lithium intercalation into and deintercalation from the interlayer of VGCFs [23]. Besides, the two cathodic peaks of MnO and ZnO reduction observed in Fig. 4a are overlapped in the CVs of [email protected] (Fig. 4b) because of enhanced MnO reduction due to the incorporation of highly conductive VGCFs. Fig. 4c and 4d show the discharge and charge curves of first three cycles of pristine ZnMnO3 and [email protected], respectively, at a current density of 100 mA

g1 (based on the mass of ZnMnO3 or [email protected]). As shown, the first discharge profiles of the two electrodes accord well with the corresponding first-cycle CVs, e.g., exhibiting a long potential plateau of ca. 0.5 V, corresponding to initial formation of Zn, Mn and Li2O according to the conversion reaction (1). The first lithiation capacities of the two electrodes are 1276 and 1357 mAh g1, respectively, which are all higher than the theoretical lithiation capacity of 1117 mAh g1 of ZnMnO3 calculated based on reaction (1) and (2) due to the formation of solid electrolyte interphase (SEI) layer [18]. The first charge (delithiation) capacity of pristine ZnMnO3 is 779 mAh g1 that is much lower than the ideal delithiation capacity of 1117 mAh g1 (provided that the reaction (1) and (2) were all reversible). Yet, it is close to the theoretical delithiation capacity of 797 mAh g1 based on reaction (3). This result suggests that the first lithiation process based on reaction (1) is not reversible, which is similar to other metal oxide anode materials such as Fe3O4 and Mn3O4 etc [25,26]. In fact, upon delithiation process, Zn is oxidized to ZnO and Mn can only be oxidized to MnO, as shown in reaction (3). By contrast, the first charge capacity of [email protected] is as high as 928 mAh g1 that is obviously higher than the theoretical delithiation capacity of 797 mAh g1 based on reaction (3). This result suggests that some Mn may be oxidized to a higher oxidation state beyond +2 valence, e.g., MnO2 [23]. Additionally, some extra capacity may originate from the capacitance contribution of the increased surface area of active material after the first lithiation process. The discharge and charge capacities of pristine ZnMnO3 electrode for the second cycle are 785 and 730 mAh g1, respectively, and those for the third cycle are 741 and 694 mAh g1, respectively. While, the discharge and charge capacities of [email protected] electrode for the second cycle are 922 and 905 mAh g1, respectively, and those for the third cycle are 926 and 910 mAh g1, respectively,

presenting much superior performance over the pristine ZnMnO3 in respect to specific capacity and capacity retention. Fig. 4e shows mainly the cycling performance of the [email protected] at 1000 mA g1 over the potential range of 0.01 to 3 V, and the rate performances at the initial stage and at the maximum-capacity stage (indicated by the two dashed rectangles and the corresponding insets). The discharge and charge capacities for the first 30 cycles at different current densities (100, 200, 500 1000 and 2000 mA g1, respectively) present the rate performance of the as-fabricated electrode, which can be seen clearly in the inset on the top left corner (Fig. 4e) or more clearly in Fig. S4a. At 100 mA g1, its reversible charge capacity is 905 mAh g1. When the current density increases to 1000 mA g1, a reversible charge capacity of 588 mAh g1 is remained. Even at 2000 mA g1, a charge capacity of 480 mAh g1 is still remained. When the current density decreases again to 100 mA g1, the specific capacity returns to the original value. After the initial-stage rate performance test, the electrode was subjected to long-term cycling test at a constant current density of 1000 mA g1. After 200 cycles at this current density (corresponding to the 230th of total cycle times), the charge capacity is increased to 957 mAh g1, the approximately maximum capacity. At that time, rate performance test was conducted once again. As shown in the inset on the lower right corner, the charge capacities at 100, 200, 500 1000, 2000, 5000 and 10000 mA g1 are 1249, 1207, 1108, 979, 826, 524 and 242 mAh g1, respectively, which are obviously higher than those obtained in the initial-stage rate performance test. After this round of rate performance test, the long-term cycling test at 1000 mA g1 is continued again. After a total of 500 cycles, a charge capacity of 809 mAh g1 is still maintained, exhibiting good cycling stability. For comparison, the cycling performance of the

pristine ZnMnO3 at 1000 mA g1 is depicted in Fig. S4b. It is evident that the [email protected] composite presents much superior electrochemical performance to the pristine ZnMnO3. To understand the phenomenon of gradual increase in capacity of the [email protected] electrode in the early stage of the long-term cycling at 1000 mA g1, a CV measurement at a scan rate of 1 mV s1 was carried out for the electrode after the total of 500 cycles, and the CV result is also presented in Fig. 4b for comparison. Comparing with the CVs of as-fabricated electrode, the two anodic peaks of the cycled electrode located in the potential range of 11.75 V are overlapped. In addition, a new shoulder peak is appeared at ca. 2.2 V and the anodic current in the range of 1.753 V is obviously increased. Correspondingly, the cathodic current is also increased in this potential range. Besides, the cathodic peak approaching 0 V is increased remarkably, suggesting enhanced alloying reaction of Zn with Li. These characteristics should account for the increase in capacity upon long-term cycling. However, these characteristics are not found for the pristine ZnMnO3 electrode (see Fig. 4a). As can be seen from the inset of Fig. 5a, for the as-fabricated electrodes, either the ohmic resistance (the high-frequency intercept on the horizontal axis, 0.00201 Ω g) or the charge transfer resistance (corresponding to the arc in the high frequency region) of the [email protected] electrode is somewhat smaller than that of the ZnMnO3 electrode (ohmic resistance, 0.00324 Ω g) due to the enhanced electron transport. The oblique lines in the low frequency region correspond to the diffusion process. We can see from Fig. 5b that the overall impedance (the impedance corresponding to the cut-off frequency, i.e., 0.01 Hz) of the [email protected] electrode is markedly decreased after 265 cycles in comparison with that of as-fabricated electrode (no

cycling) or the electrode after 40 cycles. This is mainly due to the decrease of diffusion impedance resulting from the decrease of particle size of the active material. As shown in the inset of Fig. 5b, although the ohmic resistance increases slightly with increasing cycle number (may be due to the gradual decrease of electrolyte), the charge transfer resistance is decreased somewhat. After 500 cycles, the overall impedance remains unchanged compared to that of the electrode after 265 cycles, but the Nyquist plot in the high frequency region is obviously changed; a very small arc in the high frequency region can be observed, which may be due to another electrode process. As shown in Fig. 4e, on the whole, the variation of capacity with increasing cycle number goes through three stages, namely, the stage of capacity keeping almost unchanged, the stage of capacity increasing and the stage of capacity decreasing. Fig. S5a5c show the evolution of discharge and charge profiles upon the long-term cycling at 1000 mA g1 of [email protected] electrode at the three stages, respectively. From 40th to 80th cycles (Fig. S5a), the variation of capacity is very small (almost unchanged). However, the charge potential around 2 V and the discharge potential around 1.25 V are all increase gradually with increasing cycle number. The reason is not very clear by now. From 80th to 220th cycles (Fig. S5b), the charge potential decreases (especially in the high-potential range) and the discharge potential increases gradually with increasing cycle number, suggesting gradual decrease in electrode polarization upon cycling that is in agreement with the result shown in Fig. 5, and this should account for the capacity increase at this stage. Besides, during discharging in the low-potential range, the speed of continuous drop of potential becomes slower and slower with increasing cycle number due to the gradually enhanced alloying reaction of Li with zinc that is consistent with the CV result shown in Fig. 4b. From 300th to

500th cycles (Fig. S5c), the charge potential increases with increasing cycle number, and the discharge potential above 0.6 V increases a little and while the discharge potential below 0.6 V decreases gradually with increasing cycle number. Additionally, we can see from Fig. S5a5c that the onset discharge potential increases with increasing cycle number. Besides, as shown in Fig. S5d, the most distinct difference between the two discharge/charge curves of the 2nd cycle (the initial-stage rate performance test) and 234th cycle (the maximum-capacity stage rate performance test) at 100 mA g1 lies in the higher potential region. These results suggest that except for the decrease in polarization, more and more of MnO oxidation to a higher valence state upon charging process may be another main reason for the observed gradual increase in capacity upon cycling. This can also be confirmed by the CV results shown in Fig. 4b. Based on the above results, the gradual increase in capacity of [email protected] electrode upon cycling may be explained as follows. ZnMnO3 electrode will undergo a large volumetric expansion/contraction upon discharge/charge cycling, leading to poor electric contact and hence the fast capacity degradation (see Fig. S4b). In contrast, for the [email protected] composite electrode, the highly conductive one-dimensional VGCFs incorporated in the coaxial nanostructure can maintain the conductive network upon cycling. At the early stage of cycling, the size of the active material will become smaller and will be redistributed on the VGCFs, resulting in gradual improvement of diffusion and charge transfer kinetics upon cycling. Moreover, owing to the decrease of polarization, more and more of MnO would be oxidized to a higher valence state in the charge process. In addition, the decrease of polarization could also favor the alloying of Zn with Li during discharging. As a result, a gradual increase in capacity is observed. This behavior has also been reported

for other metal oxide based composite anode materials, for example, SnO2-graphene nanocomposite [27]. The authors suggested that the graphene sheets in the SnO2/grapheme nanocomposite could prevent the aggregation of in situ formed Sn nanoparticles as in the case of bare SnO2 nanoparticles, and therefore could increase the activity of active material and the reversibility of charge/discharge reaction [27]. Fig. 6. show the SEM, TEM and HRTEM images of [email protected] electrode in charged state after 500 cycles. As shown in Fig. 6a6c, the original 1D coaxial structure of the active material is changed to the porous aggregates of nanoparticles with the majority of the VGCFs being embedded in the active material (cannot be seen). Only one VGCF can be observed either in the SEM images or in the TEM image, from which we can see clearly that the VGCF is covered by a certain amount of active material (Fig. 6b) or the VGCF connects the active material particles (Fig. 6c). The incorporated VGCFs are beneficial to electric conductance of the active material and the porous structure of the nanoscale-dispersed active material is favorable to electrolyte diffusion. As can be seen in Fig. 6d, the nanoparticles of active material are composed of amorphous nanoclusters. 3.3. Hydrothermal reaction mechanism The hydrothermal reaction mechanism was preliminarily studied. Fig. 7 shows the XRD patterns of the products of hydrothermal reaction between pure δ-MnO2 and ZnO in neutral pure water for different durations. Before hydrothermal reaction, the δ-MnO2 was mixed with ZnO by grinding. Only the characteristic peaks of hexagonal ZnO (zincite, PDF 361451) can be observed in the XRD pattern of the mixed reactants due to the disorder or poor crystallinity of pure δ-MnO2. After hydrothermal reaction for 1 h, the intensities of ZnO peaks are reduced but no ZnMnO3 is detected. After 2 h, a great amount of ZnMnO3 is formed accompanied by a small amount of

unreacted ZnO, as indicated by the peaks at ca. 36.4 and 31.8 of ZnO. The intensities of ZnMnO3 peaks increase and while those of ZnO decrease with increasing reaction time up to 12 h. In order to know more clearly the evolution of phase composition and morphology of the hydrothermal reaction system, the intermediate products of the hydrothermal system of [email protected]δ-MnO2 and ZnO obtained after reaction for short durations of 1, 2 and 4 h, respectively, are specially selected for study. The XRD patterns and TEM images of the intermediate products are shown in Fig. 8. In addition to the TEM images shown in Fig. 8b8e, the corresponding overall perspectives are also shown in Fig. S6a6d for reference. As presented in Fig. 8a, no ZnMnO3 phase can be detected after reaction for a short time of 1 h. However, after 2 h, ZnMnO3 phase is detected and the intensities of ZnO peaks fall by half. After reaction for 4 h, a large amount of ZnMnO3 is formed but it is still accompanied by a small amount of unreacted MnO2 and ZnO. The initial reactant [email protected]δ-MnO2 is of a coaxial structure with δ-MnO2 nanosheets covering the VGCF (Fig. 8b). After 1 h, its morphology remains almost unchanged except for the emergence of some small ZnMnO3 particles on the outmost of δ-MnO2 shell (Fig. 8c and Fig. S6b). Fig. S6e shows the HRTEM image of such an early formed ZnMnO3 grain on the outmost of δ-MnO2 shell (after 1 h). Judging from the lattice fringes, it should belong to (111) plane of ZnMnO3 crystal. This HRTEM image shows well the growth boundary in crystal growth process. After 2 h, more ZnMnO3 grains of ca. 80100 nm are grown on the outer margin of the [email protected]δ-MnO2 (Fig. 8d). After 4 h, the δ-MnO2 nanosheets are almost all transformed into ZnMnO3 nanograins (Fig. 8e). These observations reveal that the reaction will proceed from the outside to inside of δ-MnO2 shell layer. As we know, hydrothermal growth of ZnO crystals has been intensively studied and it was

commonly carried out in concentrated alkaline solutions [28,29]. By contrast, the hydrothermal reaction presented herein was carried out in neutral pure water. As the two initial reactants are water-insoluble solids under conventional conditions, we think that the present reaction may be carried out via a dissolutionrecrystallization process under hydrothermal conditions. The further and more detailed experimental results and discussions about the hydrothermal reaction mechanism are described in Supporting Information. 4. Conclusions In summary, a coaxial [email protected] nanocomposite material was synthesized via a two-step method involving a crucial hydrothermal reaction between ZnO and [email protected]-MnO2. This composite material can exhibit high specific capacity, excellent cycling performance and good rate performance as an anode material for LIBs. For the as-fabricated electrode cycled at different current densities of 100, 1000 and 2000 mA g1, respectively, the reversible charge capacities of 905, 588 and 480 mAh g1 are obtained. When it is repeatedly cycled at a constant current density of 1000 mA g1, its capacity is gradually increased up to 957 mAh g1 in the 200th cycle and then is gradually decreased to 809 mAh g1 in the 500th cycle. At the stage of top capacity achieved upon the long-term cycling at 1000 mA g1, its specific capacity and rate performance are obviously better than those achieved in the initial stage (as-fabricated electrode). For example, at this stage the charge capacities at 100, 1000 and 2000 mA g1 are 1249, 979 and 826 mAh g1, respectively. The increase in capacity with increasing cycle number in the early stage should be mainly attributed to the enhancement of electric conduction due to the incorporation of highly conductive VGCFs. The electrochemical Li-storage performance of [email protected] nanocomposite is much superior to that of pristine ZnMnO3, especially for the

long-term cycling performance. This composite would be a promising high-capacity anode material for lithium ion batteries. In addition, the mechanism of the particular hydrothermal reaction between ZnO and MnO2 was also preliminarily studied. A dissolutionrecrystallization mechanism is proposed for the hydrothermal reaction. This study would provide a new insight in crystal formation of binary metal oxide under hydrothermal conditions evolving a manganese dioxide precursor. Besides, this synthetic strategy would be extended to the synthesis of other composite materials for energy storage applications. Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.electacta.2016.xx.xxx. Acknowledgment The authors acknowledge the support of the Shanghai Education Commission (Peak Discipline Construction). Instrumental Analysis & Research Center of Shanghai University is gratefully acknowledged for XRD, XPS, SEM and TEM experiments.

References [1] D. Larcher, J.M. Tarascon,

Towards greener and more sustainable batteries for

electrical energy storage, Nat. Chem. 7 (2015) 19–29. [2] J.B. Goodenough, K.S. Park,

The Li-ion rechargeable battery: a perspective, J.

Am. Chem. Soc. 135 (2013) 1167–1176.

[3] T.-H. Kim, J.-S. Park, S.K. Chang, S. Choi, J.H. Ryu, H.-K. Song,

The current

move of lithium ion batteries towards the next phase, Adv. Energy Mater. 2 (2012) 860–872. [4] B.D. Adams, R. Black, C. Radtke, Z. Williams, B.L. Mehdi, N.D . Browning, L.F. Nazar, The importance of nanometric passivating films on cathodes for Li–air batteries, ACS Nano 8 (2014) 2483–2493. [5] J.H. Yan, X.B. Liu, M. Yao, X.F. Wang, T.K. Wafle, B.Y. Li,

Long-life,

high-efficiency lithium−sulfur battery from a nanoassembled cathode, Chem. Mater. 27 (2015) 5080−5087. [6] N.S. Choi, Z. Chen, S.A. Freunberger, X. Ji, Y.K. Sun, K. Amine, G. Yushin, L.F. Nazar, J. Cho, P.G. Bruce,

Challenges facing lithium batteries and

electrical double-layer capacitors, Angew. Chem. Int. Ed. 51 (2012) 9994–10024. [7] A.T. Tesfaye, R. Gonzalez, J.L. Coffer, T. Djenizian,

Porous silicon nanotube

arrays as anode material for Li-ion batteries, ACS Appl. Mater. Interfaces 7 (2015) 20495–20498. [8] P. Poizot, S. Laruelle, S. Grugeon, L. Dupont, J.M. Tarascon, Nano-sized transition-metal oxides as negative-electrode materials for lithium-ion batteries, Nature 407 (2000) 496–499. [9] Y. Kim, J.-H. Lee, S. Cho, Y. Kwon, I. In, J. Lee, N.-H. You, E. Reichmanis, H. Ko, K.-T. Lee, H.-K. Kwon, D.-H. Ko, H. Yang, B. Park,

Additive-free

hollow-structured Co3O4 nanoparticle Li-ion battery: the origins of irreversible capacity loss, ACS Nano 8 (2014) 6701–6712. [10] G. Zhou, D.-W. Wang, F. Li, L. Zhang, N. Li, Z.-S. Wu, L. Wen, G.Q. Lu, H.-M. Cheng,

Graphene-wrapped Fe3O4 anode material with improved

reversible capacity and cyclic stability for lithium ion batteries, Chem. Mater. 22 (2010) 5306–5313. [11] Z. Li, N. Liu, X. Wang, C. Wang, Y. Qi, L. Yin,

Three-dimensional

nanohybrids of Mn3O4/ordered mesoporous carbons for high performance anode materials for lithium-ion batteries, J. Mater. Chem. 22 (2012) 16640–16648. [12] C. Yuan, H.B. Wu, Y. Xie, X.W. Lou,

Mixed transition-metal oxides: design,

synthesis, and energy-related applications, Angew. Chem. Int. Ed. 53 (2014) 1488–1504. [13] L. Zhou, H.B. Wu, T. Zhu, X.W. Lou,

Facile preparation of ZnMn2O4 hollow

microspheres as high-capacity anodes for lithium-ion batteries, J. Mater. Chem. 22 (2012) 827–829. [14] J.G. Kim, S.H. Lee, Y. Kim, W.B. Kim, Fabrication of free-standing ZnMn2O4 mesoscale tubular arrays for lithium-ion anodes with highly reversible lithium storage properties, ACS Appl. Mater. Interfaces 5 (2013) 11321–11328. [15] P. Xiong, B.R. Liu, V. Teran, Y. Zhao, L.L. Peng, X. Wang, G.H. Yu, Chemically integrated two-dimensional hybrid zinc manganate/graphene nanosheets with enhanced lithium storage capability, ACS Nano 8 (2014) 8610–8616. [16] L. Yin, Z. Zhang, Z. Li, F. Hao, Q. Li, C. Wang, R. Fan, Y. Qi, Spinel ZnMn2O4 nanocrystal-anchored 3D hierarchical carbon aerogel hybrids as anode materials for lithium ion batteries, Adv. Funct. Mater. 24 (2014) 4176–4185. [17] F.M. Courtel, H. Duncan, Y. Abu-Lebdeh, I.J. Davidson,

High capacity anode

materials for Li-ion batteries based on spinel metal oxides AMn2O4 (A = Co, Ni, and Zn), J. Mater. Chem. 21 (2011) 10206–10218.

[18] X. Liu, C. Zhao, H. Zhang, Q. Shen,

Facile synthesis of porous ZnMnO3

spherulites with a high lithium storage capability, Electrochim. Acta 151 (2015) 56–62. [19] A.L.M. Reddy, M.M. Shaijumon, S.R. Gowda, P.M. Ajayan,

Coaxial

MnO2/carbon nanotube array electrodes for high-performance lithium batteries, Nano Lett. 9 (2009) 1002–1006. [20] C. Ban, Z. Wu, D.T. Gillaspie, L. Chen, Y. Yan, J.L. Blackburn, A.C. Dillon, Nanostructured Fe3O4/SWNT electrode: binder-free and high-rate Li-ion anode, Adv. Mater. 22 (2010) E145–E149. [21] D. Cai, D. Wang, H. Huang, X. Duan, B. Liu, L. Wang, Y. Liu, Q. Li, T. Wang, Rational synthesis of ZnMn2O4 porous spheres and graphene nanocomposite with enhanced performance for lithium-ion batteries, J. Mater. Chem. A 3 (2015) 11430–11436. [22] B. Wang, W.A. Abdulla, D. Wang X.S. Zhao, A three-dimensional porous LiFePO4 cathode material modified with a nitrogen-doped graphene aerogel for high-power lithium ion batteries, Energy Environ. Sci. 8 (2015) 869–875. [23] F. Ma, A.B. Yuan, J.Q. Xu,

Nanoparticulate Mn3 O4/VGCF composite

conversion-anode material with extraordinarily high capacity and excellent rate capability for lithium ion batteries, ACS Appl. Mater. Interfaces 6 (2014) 18129–18138. [24] M. Wang, P. Pang, L.K. Koopal, G. Qiu, Y. Wang, F. Liu, One-step synthesis of -MnO2 nanoparticles using ascorbic acid and their scavenging properties to Pb(II), Zn(II) and methylene blue, Mater. Chem. Phys. 148 (2014) 1149–1156.

[25] L. Ji, Z. Lin, M. Alcoutlabi, X. Zhang,

Recent developments in nanostructured

anode materials for rechargeable lithium-ion batteries, Energy Environ. Sci. 4 (2011) 2682–2699. [26] V.A. Agubra, L. Zuniga, D. Flores, J. Villareal, M. Alcoutlabi,

Composite

nanofibers as advanced materials for Li-ion, Li-O2 and Li-S batteries, Electrochim. Acta 192 (2016) 529–550. [27] P. Lian, X. Zhu, S. Liang, Z. Li, W. Yang, H. Wang,

High reversible capacity

of SnO2/graphene nanocomposite as an anode material for lithium-ion batteries, Electrochim. Acta 56 (2011) 4532–4539. [28] W.-J. Li, E.-W. Shi, W.-Z. Zhong, Z.-W. Yin,

Growth mechanism and growth

habit of oxide crystals, J. Cryst. Growth 203 (1999) 186–196. [29] S. Xu, Z.L. Wang,

One-dimensional ZnO nanostructures: solution growth and

functional properties, Nano Res. 4 (2011) 1013–1098.

Figure captions

Fig. 1. Illustration of the synthesis process of the pristine ZnMnO3 and the coaxial-cable [email protected] composite. Fig. 2. (a) XRD patterns of pristine ZnMnO3 and [email protected] synthesized by hydrothermal reaction in pure water at 180 C for 12 h. (b) TEM image of ZnMnO3. (c) SEM and (d) TEM images of [email protected] . (e) HRTEM image of a ZnMnO3 grain marked by the rectangle in the inset of 2(d).

Fig. 3. XPS spectra of (a) full scan and (b) Mn 2p3/2 of pristine ZnMnO3. XPS spectra of (c) full scan and d) Mn 2p3/2 of [email protected] composite. Fig. 4. Cyclic voltammograms of (a) pristine ZnMnO3 and (b) [email protected] composite recorded at a scan rate of 1 mV s1. Discharge/charge curves of the first three cycles of (c) pristine ZnMnO3 and (d) [email protected] composite at a current density of 100 mA g1. (e) Cycling performance of [email protected] composite. Fig. 5. (a) Electrochemical impedance spectra (EISs) of as-fabricated ZnMnO3 and [email protected] electrodes recorded in the frequency range of 105 to 0.01 Hz (inset shows the EIS data in the high frequency region). (b) EISs of [email protected] electrode measured after total cycles of 40, 265 and 500, respectively, at charged state (charged to 3 V vs. Li+/Li) over the frequency range of 105 to 0.01 Hz (inset shows the EIS data in the high frequency region). Fig. 6. (a,b) SEM images of [email protected] electrode after 500 cycles with different magnifications. (c,d) TEM and HRTEM images of [email protected] electrode after 500 cycles. Fig. 7. (a) XRD patterns of products of hydrothermal reaction in neutral pure water for different durations (reactants are ground mixture of ZnO and δ-MnO2). (b) Close-up view of 7a. Fig. 8. (a) XRD patterns of mixed reactants (mixture of ZnO and [email protected]δ-MnO2 obtained by grinding) and products of hydrothermal reaction for different durations. TEM images of (b) [email protected]δ-MnO2 before hydrothermal reaction, and the products of hydrothermal reaction for (c) 1 h, (d) 2 h and (e) 4 h, respectively.