HxMoO3 nanobelts with better performance as anode in lithium-ion batteries

Electrochimica Acta 213 (2016) 641–647

Contents lists available at ScienceDirect

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

HxMoO3 nanobelts with better performance as anode in lithium-ion batteries Xiaokang Ju, Peigong Ning, Xiaobing Tong, Xiaoping Lin, Xi Pan, Qiuhong Li, Xiaochuan Duan* , Taihong Wang* Pen-Tung Sah Institute of Micro-Nano Science and Technology, Xiamen University, Xiamen 361005, China

A R T I C L E I N F O

Article history: Received 6 June 2016 Received in revised form 27 July 2016 Accepted 31 July 2016 Available online 1 August 2016 Keywords: HxMoO3 anode high rate Lithium-ion batteries

A B S T R A C T

We first report the pure HxMoO3 nanobelts as anode for lithium-ion batteries by a facile hydrothermal with ammonium heptamolybdate tetrahydrate ((NH4)6Mo7O24
4H2O) and hydrochloric acid (HCl). Owing to hydrogen-doping, Mo5+ exists in the HxMoO3 nanobelt, which may release extra electrons. Therefore, the electric conductance of HxMoO3 nanobelt is enhanced greatly. Moreover, the content of hydrogen can’t be high, since the ordered structure deteriorates when amount of hydrogen increasing. The H0.28MoO3 nanobelts we designed exhibit outstanding specific capacity and rate performance. The stable capacity of 920 mAh g1 is obtained after 25 charge/discharge cycles at 100 mA g1. At high current densities such as 1, 2, 5 and 10 A g1, the H0.28MoO3 electrode delivers specific capacities of about 600, 500, 420, 300 mAh g1, respectively. Even after 450 charge discharge cycles at 1 A g1, the performance of our materials can maintain the capacity of about 550 mAh g1. Furthermore, we provide more discussion about the lithium storage mechanism of HxMoO3 nanobelts through ex situ XRD and FESEM. By comparing HxMoO3 with different X, we find that low content of hydrogen can greatly improve the performance of a-MoO3 electrodes in Li-ion batteries. ã 2016 Published by Elsevier Ltd.

1. Introduction Lithium-ion batteries (LIBs) have been widely used in various fields, such as mobile phones, laptops, and other small portable appliances [1–4]. In addition, significant improvements in rate and endurance for economic, safe electrode materials may make possible utilization in plug-in hybrid electric vehicles (PHEVs) [4– 8]. Great effort has been made to develop graphite alternatives, which deliver higher specific capacity than 372 mAh g1. A good anode material candidate should be characterized by a high specific capacity, a highly reversible Li insertion/de-insertion as well as low polarization [9–21]. Recent work has been concentrated on preparing nanostructures for improving capacity and rate index in LIBs electrode [22]. The morphology of 1D material has not only enlarged electrodeelectrolyte contact areas and relaxed strain of volume expansion, but also provided efficient 1D electron transport pathways [23]. Orthorhombic molybdenum trioxide (a-MoO3) exhibits a characteristic double-layered structure consisting of MoO6 octahedron

* Corresponding author. Tel.:+86 0592 2187196; fax: +86 0592 2197196. E-mail address: [email protected] (T. Wang). http://dx.doi.org/10.1016/j.electacta.2016.07.160 0013-4686/ã 2016 Published by Elsevier Ltd.

units with lattice constants a = 3.962 Å, b = 3.855 Å, and c = 3.699 Å [24]. Orthorhombic crystal structure is based on a series of bilayers which should be perpendicular to the [010] axis. [25]. This unique characteristic makes it be one of the most potential candidates to form large surface areas and high surface-to-volume ratio nanostructures [26]. In this paper, we employed a one-step hydrothermal approach to synthesize the 1D nanobelts MoO3 and HxMoO3. The theoretical specific capacity of MoO3 is about 1117 mAh g1, which is nearly three times that of graphite. However, its applications have been limited by its poor electronic conductivity. Some techniques have been proved to be effective, such as dimensionality reduction, synthesis of MoO3 nanobelt-graphene composites, ion-doping, coating [16,27–30], and et al., which are being reported in the past few years. The electrochemical reactivity of MoO3/C as anode material has been investigated by Xia et al. [31] and compared with pure MoO3 electrode, MoO3/C exhibits much higher specific capacity (a stable capacity of 600 mAh g1 is delivered at 100 mA g1) due to the protection of carbon layer. In another study [32], SnO2/MoO3/C nanostructure delivers a reversible discharge capacity of 500 mAh g1 after 120 cycles at a high current density of 200 mA g1, because of the 1-D nanostructure composed of nanosheets and the carbon matrix. Great progress in improving capacities of MoO3 has also been

642

X. Ju et al. / Electrochimica Acta 213 (2016) 641–647

achieved recently. B. Ahmed et al. [33] have coated MoO3 nanobelts with HFO2, and the HFO2-coated nanobelts demonstrates a high capacity of more than 600 mAh g1 at a current density of 1000 mA g1 after 50 cycles. The presence of HFO2 layer serves as protective barrier at anode/electrolyte interface. Herein, electrode design appears to be a critical issue for meeting the development of LIBs. In this work, an alternative method of preparing HxMoO3 nanobelts is proposed. As a potential candidate of the electronproton mixed conductor, the hydrogen molybdenum bronze (HxMoO3) has drawn more and more attention due to their applications in hydrogen-transfer catalysts, electrochromic displays, fuel cell, integration of nanoscale electronic devices and hydrogen storage [34–37] but rarely focusing on lithium-ion batteries. And as we all know, the transitional metallic cations are shown with mixed valences in HxMoO3 nanobelts [37]. Owing to hydrogen-doping, Mo5+ exists in the nanobelts, which may release extra electrons that are quasi-free within the MoO6 octahedral. Therefore, the electric conductance of HxMoO3 nanobelt is enhanced greatly. Indeed, there are series of compounds of layered MoO3 with different hydrogen contents. And the samples are transformed from orthorhombic to monoclinic with the increase of the hydrogen content. Intuitively, the colour of MoO3 turns blue, and gradually darken [38,39]. We find that only the low hydrogendoping sample is orthorhombic, which is similar to a-MoO3 that is a thermodynamically stable phase, while monoclinic MoO3 (b-type) belongs to metastable phase. So, it is better to obtain HxMoO3 with smaller x value. In addition, as reported by Wang et al. [40], HxMoO3 belongs to MxMoO3nH2O. The most distinguish structural feature of MxMoO3
nH2O is a wide onedimensional tunnel (3.2–3.5 Å in diameter), which should allow lithium-ions free intercalation [41]. Up to now, we have only found that song et al. [34] have reported the H0.6MoO3@C (monoclinic phase) nanobelts in LIBs. However, they can't distinguish which play a more important role between carbon coating and hydrogen doping. Moreover, the hydrogen content may be too high that can distort the structure. The inference also has been proved in the following experiment that low hydrogen-doping may be more useful in LIBs. The reversible capacity of synthetic H0.28MoO3 nanobelts is about 920 mAh g1, which is approaching the theory capacity of MoO3. When testing at a high rate, the performance of our materials can maintain the capacity of about 550 mAh g1 after 450 cycles at 1 A g1. Even at 10 A g1, the capacity can still maintain at around 300 mAh g1

morphology and crystal structure are collected by scanning electron microscopy (SEM, Hitachi S4800) and high-resolution transmission electron microscopy (HRTEM, JEM–2100). Raman spectroscopy was performed using a micro-Raman 2000 system (Renishaw, Britain) with a 10 mW laser excitation source of wavelength 633 nm. X-ray photoelectron spectrum (XPS) is carried out in ESCALAB 250 with a monochromatic Al Ka X-ray source (hn= 1486.6 eV) operating at 150 W, a multichannel plate and delay line detector under a vacuum of 109 mbar. 2.2. Electrochemical measurements The electrochemical tests are performed using a CR2016-type coin cell. Pure lithium foils are used as counter electrodes. The active materials are mixed with carbon black and carboxyl methyl cellulose at a weight ratio of 8: 1: 1, which are dispersed in distilled water and alcohol, then to form a homogeneous slurry. The mixed slurry is tape cast on Cu foil and dried at 80  C for 12 h under vacuum. The dried foil is cut into circular disks to be used as working electrode. The loading density of working electrode is about 0.9–1.2 mg cm2. The electrolyte solution is 1 M LiPF6 in ethylene carbonate–dimethyl carbonate–diethyl carbonate (1: 1: 1, in weight percent). A Celgard 2400 microporous polypropylene membrane is used as a separator. The coin-type cells are assembled in an argon filled glovebox with water and oxygen contents of less than 0.5 ppm. Galvanostatic charge/discharge cycling performance is carried out by Land CT 2001 battery tester in the voltage range of 3 V–0.01 V vs. Li+/Li at 25  C. 3. Results and discussion Fig. 1 exhibits the XRD patterns of the obtained H0.33MoO3 samples(phase III), H0.28MoO3 samples (phase II) and MoO3 samples(phase I). The three patterns in Fig. 1 are similar to each other in general, but there are still subtle differences among them. (For example, the magnified pattern of (040) diffraction peaks for H0.28MoO3 and H0.33MoO3 is shown in Fig. S1) The diffraction peaks of three samples can be well-indexed to different JCPDS cards. In addition, the cell lattices of HxMoO3 (X = 0.28 and 0.33) and MoO3 can be seen in Supporting Information of Table 1. It is easy to discover that the lattice unit is enlarged after hydrogen atoms are doped into MoO3 [42]. The phase I maintains the pristine phase without evident change and agrees well with the XRD pattern of MoO3 (JCPDS, no:76-1003, orthorhombic). When increasing HCl to

2. Materials synthesis All the chemical reagents are analytical grade and they are used without further purification. In a facile hydrothermal synthesis, 0.6 g ammonium heptamolybdate tetrahydrate ((NH4) 6Mo7O24
4H2O) is dissolved in 36 ml deionized water under continuous stirring. After 10 minutes, 2.5 ml hydrochloric acid (HCl) is added to the beaker slowly and stirred for another 5 minutes. Then the homogeneous suspension is transferred to a Teflon-lined autoclave with a capacity of 50 ml and kept inside an electric oven at 160  C for 15 h. The item is then centrifuged and washed with deionized water for several times and dried under vacuum at around 60  C. MoO3with white colour is observed. In addition, by increasing the amount of HCl to 3.5 ml and 6.5 ml respectively, we can get the synthesis of H0.28MoO3 nanobelts and H0.33MoO3 nanobelts. (Other preparation conditions are all same). 2.1. Materials characterization The powder X-ray diffraction (XRD) patterns are recorded on a Panalytical X–pert diffracto meter with Cu–Ka radiation. The

Fig. 1. The XRD patterns of the obtained H0.33MoO3 samples(phase III),H0.28MoO3 samples(phase II) and MoO3 samples(phase I).

X. Ju et al. / Electrochimica Acta 213 (2016) 641–647

643

Fig. 2. (a) and (d) are the SEM images of MoO3, (b) and (e) are the SEM images of H0.28MoO3, (c) and (f) are the SEM images of H0.33MoO3. (g) (h) and (i) are the quantitative statistics distribution of length of MoO3 and H0.28MoO3 and H0.33MoO3.

3.5 ml and 6.5 ml, respectively, the sample I is transformed into H0.28MoO3. (phase II, JCPDS, no:70-4476, orthorhombic) and H0.33MoO3 (phase III, JCPDS, no:85-1800, monoclinic). In combination with the orthorhombic a-MoO3, we consider that the structural framework of MoO3 nanobelts is maintained in H0.28MoO3 nanobelts (orthorhombic) to a great degree and hydrogen doping does not break the pristine structure. Thus, HxMoO3 inherits the morphology of MoO3. [38]. Many researchers have discovered that MoO3 can change its colour with different content of hydrogen-doping [36–39,42,43]. The colour transforms of HxMoO3 from initial white to blue can be seen in Fig. S2.

The morphology of MoO3 and HxMoO3 (X = 0.28 and 0.33) nanobelts are characterized by SEM. Fig. 2 shows SEM images of pristine MoO3 and HxMoO3 nanobelts powders. As shown in Fig. 2a–f, by comparing their SEM images, there is no obvious difference in morphology and topology between MoO3 and HxMoO3 nanobelts. Fig. 2d–f are FESEM images about MoO3 and HxMoO3. From Fig. 2, it can be seen that the nanobelts with uniform width of 100–250 nm. In Fig. 2 (g) (h) and (i), we use Image Pro Plus to give the distribution of length of the three nanobelts. It's easy to find that most of the nanobelts have the length of 4–7 mm, though with different content of hydrogen-doping. Fig. 3 shows

Fig. 3. (a) and (d) are the TEM images of MoO3, (b) and (e) are the TEM images of H0.28MoO3, (c) and (f) are the TEM images of H0.33MoO3. The inset images of Fig. 3(a) (b) and (c) are the SAEDs of MoO3, H0.28MoO3 and H0.33MoO3 nanobelts.

644

X. Ju et al. / Electrochimica Acta 213 (2016) 641–647

representative TEM images of MoO3 and HxMoO3 (X = 0.28 and 0.33) nanobelts, respectively. Similar to the SEM results (Fig. 2), the magnification TEM images of the MoO3 and the HxMoO3 nanobelts don't make much difference. The lattice spacing of 0.278 nm (Fig. 3d) corresponds to (101) plane of orthorhombic a-MoO3 structure, and meanwhile, the lattice spacing of 0.372 nm (Fig. 3e) and 0.443 nm (Fig. 3f) corresponds to (110) plane of orthorhombic H0.28MoO3 and (220) plane of monoclinic H0.33MoO3 structure, respectively. Fig. 4(a), (b) and (c) are the XPS of MoO3, H0.28MoO3 and H0.33MoO3 nanobelts, respectively. Mo (3d) level spectras are obtained and spectral analysis is carried out using Gaussian curves to fit the experimental date. In Fig. 4a, the doublets of 235.75 and 232.55 eV are attributed, respectively, to the binding energies of the 3d3/2 and 3d5/2 electrons of Mo6+ oxidation state, the energy gap between the doublets is 3.2 eV and the integral areas between the two doublets are 2:3 in ratio, which agrees well with the standard data [44]. In Fig. 4b and c, the two peaks of Mo+6 are about 232.5 and 235.6 eV for H0.28MoO3, while 232.55 and 235.6 eV for H0.33MoO3 respectively. The couples of Mo+5 (231.15 and 234.3 eV corresponding to 3d2/5 and 3d2/3 of Mo+5 for H0.28MoO3, 231.13 and 234.2 eV for H0.33MoO3) can also be detected. The peaks shift to low energy slightly, as compared to those in Fig. 4a. It is known that the extra electron screening can reduce the inner electron binding energy of core metallic cations [43],From the ratio of the areas of the Gaussian pairs, the values of 0.099 and 0.112 are obtained for the ratio of the number of Mo5+ atoms to the total Mo atoms in the samples of H0.28MoO3 and H0.33MoO3. Thus, the more protons (H+) intercalated, the more electrons injected in the meantime. Considering about the presence of different oxidation state of Mo and slight energy shift, we can draw a conclusion that hydrogen has doped into the MoO3 successfully. However, in order to know the site of doped hydrogen and the state of carbon layer, we adopt Raman spectra. As we can see, Fig. S3, shows the typical Raman

scattering bands of a-MoO3 and H0.28MoO3. The frequency at about 994.70 cm1 is attributed to the symmetric stretching vibrations of Mo = O unit, whereas the frequencies at about 818.45 cm1 and 665.18 cm1 are respectively assigned to the characteristic peaks of corner-sharing Mo2-O and edge-sharing Mo3-O [38]. The Raman spectra of two nanobelts are similar to each other, which can suggest that inserted hydrogen ions have little effect on MoO6 octahedra and chemical bond parameters. It is conceivable to infer only the presence of hydrogen ions with the interlayer space of van der Waals layers. Layer itself has little effect on the structure when the amount of hydrogen is small. The detailed electrochemical behaviors of MoO3, H0.28MoO3 and H0.33MoO3 nanobelts electrodes are investigated by cyclic voltammetry (CV). Fig. 5a and b show the CV curves of HxMoO3(X = 0.28 and 0.33), and CV curve of MoO3 is shown in Fig.S4. Electrodes are been carried out at a scan rate of 0.1 mV s1 in potential window of 0.01–3.0 V, respectively. The curves all show typical peaks of MoO3 in the oxidation and reduction reactions, which suggest the same electrochemical redox reaction and lithium-ion insertion-desertion on HxMoO3 electrode. Taking H0.33MoO3 as an example, in the first cathodic scan, peak at around 1.75 V corresponds to the intercalation of Li+ ions into the interlayer spacing between the MoO6 octahedron layers. More specifically, as reported by Brezesinski et al. [45], the cathodic peak at around 2.75 V can be attributed to an irreversible phase transition from a-MoO3 to Li0.25MoO3, and peaks at lower potentials (2.25 V) can be described to the charge-insertion sites of the intra layer, while peak at 0.27 V attributes to the reaction of lithium with formed LixMoO3 [3]. Interestingly, during the first anodic scanning process, except the relatively weaker peak at around 1.3 V, which demonstrates the desertion of Li+ from the inner structure of the MoO6 octahedron layers, the slight peak at 2.5 V of H0.33MoO3 in the cycle may be assigned to the desertion of lithium from the van der Waals gaps [23]. After the first cycle, the

Fig. 4. The XPS of (a)MoO3, (b)H0.28MoO3and(c) H0.33MoO3 nanobelts.

X. Ju et al. / Electrochimica Acta 213 (2016) 641–647

645

Fig. 5. The CV curves of (a) H0.28MoO3, (b) H0.33MoO3. (c) and (d) show the charge/discharge profiles for the 1st, 2nd and 25th cycles at a current density of 100 mA g1 between 0.01 and 3.0 V for H0.28MoO3 and H0.33MoO3.

cathodic conversion potential is slightly shifted to a lower potential and the cathodic peak current is obviously decreased. Furthermore, we adopt corresponding ex situ XRD and ex situ FESEM images of H0.28MoO3 nanobelts to show the charge and discharge process. Fig.S5 is the morphology corresponding to the different discharge voltage. When discharging at 1.95 V, 0.65 V and 0.34 V (Fig. S5(a) (b) and (c)), the morphology of nanobelts can be maintained to a large degree. While discharging at 0.01 V and charging at 3 V (in Fig.S5 (d) and (e)), we can find that there are some nanosheets on the surface of nanobelts. Fig. S5(f) is the magnifying image of Fig. S5(e). As the stability of HxMoO3 depends on the content of hydrogen, the results are different from the report of Wang et al. [40]. However, in Fig. S5(g), the corresponding XRD patterns from different stages show that the main crystalline phases are all about Li2MoO4, JCPDS:12-0763. (except for some minor phase,such as Li6Mo2O7, JCPDS: 73-2301). Li2MoO4 is a stable intermediate phase. We can reasonably infer that the high performance mainly result from the presence of the phase Li2MoO4. Fig. 5c, d and Fig. S6 show the charge/discharge profiles for the 1st, 2nd and 25th cycles at a current density of 100 mA g1 between 0.01 and 3.0 V for H0.28MoO3 H0.33MoO3 and MoO3 electrodes, respectively. The first discharge curve in all pristine HxMoO3 and MoO3 electrodes show 2 distinct voltage plateaus at 2.25 and 0.45 V. The voltage plateau at 2.25 V has already been reported [23] and corresponds to charge-insertion sites of the intra layer, which also agree well with the CV curves of Fig. 5a and b. During the lithium insertion, The electrode encounter an irreversible structural change from a phase to incomplete phase LixMoO3, while the long plateau at 0.40 V in the first cycle is related to the lithium insertion into LixMoO3 by a characteristic conversion reaction to form Mo metal and Li2O, and the formation of the SEI layer [46]. In addition, the superior conductivity of HxMoO3 and expanded crystal spacing provide more opportunities for lithium

ion insertion/desertion in spacing of MoO6 and the interior structure of the MoO6 octahedron. From Fig. 5c and d, the first cycle discharge capacities are found to be about 1546.8 and 1585.1 mAh g1 for the H0.28MoO3 and H0.33MoO3 electrodes, respectively. Almost stable capacities of 920 and 897.9 mAh g1 are obtained after 25 charge/discharge cycles for them, respectively. While it can be seen in Fig. S6 that the first discharge capacity of MoO3 is about 1324.1 mAh g1 and 820.4 mAh g1 can be delivered after 25 cycles. The rate performance of the H0.28MoO3 and MoO3 is shown in Fig. 6a. The contrastive rate of H0.33MoO3 is provided in Fig. S7. All electrodes are studied at current rates of 100 mAh g1, 200 mAh g1, 500 mAh g1, 1 A g1, 2 A g1, 5 A g1 and 10 A g1. At lower current rates of 100 mAh g1, 200 mAh g1, 500 mAh g1, they exhibit close behavior, but at higher current densities (1 A g1, 2 A g1, 5 A g1 and even higher), the effect of hydrogen-doping become notable. At 1 A g1, the H0.28MoO3 electrode shows a specific capacity of 584 mAh g1, which is about two times of pristine MoO3 nanobelts electrode (about 300 mAh g1). Even at 10 A g1, the capacity of H0.28MoO3 is about 294.3 mAh g1. The results presented in Fig. 6a clearly demonstrate the great improvements of rate performance of MoO3 nanobelts at different current densities. Fig. 6b presents the cyclic performance of MoO3 and HxMoO3 (X = 0.28 and 0.33) electrodes measured at a current density of 1 A g1. All of the three electrodes exhibit increasing capacities in the initial several cycles and capacity fading after it. The mechanism may result from the damage of SEI and a new thinner SEI has been generated [47] or the formation of a polymeric gel-like film [48,49]. The capacity delivered by H0.28MoO3 sample is about 527 mAh g1 after 450 cycles (60.67% of the first discharge capacity). In contrast, the H0.33MoO3 powder gives a lower capacity of 374.8 mAh g1 (39.73% of the first discharge capacity), while pure MoO3 drops to 310 mAh g1 after 450 cycles. Which also suggest that hydrogen doping really has a potential to improve the

646

X. Ju et al. / Electrochimica Acta 213 (2016) 641–647

Fig. 6. (a) The rate performance of the H0.28MoO3and MoO3, (b) the cyclic performance of MoO3 and HxMoO3(X = 0.28 and 0.33) electrodes measured at a current density of 1 A g1.

ability of MoO3. However, when the content of hydrogen-doping is high, the crystal lattice of oxide expands and the ordered structure deteriorates, the electrical performance may not be the best. The coulombic efficiency of H0.28MoO3 (current density of 1 A g1) is shown in Fig.S8. The initial efficiency is about 80.6%, which is really higher than other metal oxide. Furthermore, the efficiency is closing to 100% in the following cycles. Those all indicate that the H0.28MoO3 anodes possess better performance in LIBs. After cycle at 1 A g1, we carry out ex situ SEM of H0.28MoO3 electrode. The images are exhibited in the Fig.S10. We can discover the nanobelts neither pulverize nor aggregate. Instead, they formed to the nanowire clusters. According to the report by Wang et al. [40]. The original H0.28MoO3 nanobelts are gradually transformed into ultrathin nanowire clusters within the Li2O and Mo by the lithium ions due to it containing a larger ion tunnel than in the nanobelts. Moreover, the more active sites can be provided for lithium ion diffusion by the larger exposed surface area. Based on the discussion mentioned above and the ex situ XRD of the electrode material after 450th cycle (In the supporting information of Fig. S11) The Li2MoO4 and Li6MoO7 et al. (denoted as LixMoyOz) and Mo can be indexed. Though we don’t find Li2O, the electrochemical behavior should include the following process: V > 1.5: MoO3 + xLi+ + x e ! LixMoyOZ +SEI

(3)

V < 0.7: LixMoyOZ + (2z-x) Li + + (2z-x) e ! y Mo + z Li2O

(4)

The performance of our H0.28MoO3 nanobelts can be compared with other MoO3-based anodes reported in the literature previously [31–33] and the relevant data are presented in the supporting information of Table 2. As shown in Table 2, the electrode performance of the carbon coated anodes [31] present an attractive performance in respect to that of previously reported MoO3 electrodes. A stable capacity of about 600 mAh g1 is still delivered at 100 mA g1 after 100 cycles. (53.7% of the theoretical capacity). SnO2/MoO3/C [32] has also demonstrated reversible discharge capacity of 500 mAh g1 after 120 cycles at a high current density of 200 mA g1 (50.13% of the theoretical capacity), these results are a reflection of poor ionic and electronic conductivity of MoO3. It has been recently demonstrated that the electrodes based on MoO3 nanoparticles [33] can be fine tuned to achieve a high capacity more than 600 mAh g1 (53.66% of the theoretical one) at high rate of 1000 mA g1 after 50 cycle. While the performance of our materials can maintain the capacity of about 550 mAh g1 after 450 cycles at 1000 mAh g1 (676.6 mAh g1 after 50 cycles, which is 60.57% of theoretical capacity). The superior performance of this

electrode can be attributed to the improved conductivity and the expanded lattice unit arising from hydrogen-doping. In addition, There are some pioneering other publications such as carbon coated MoO2 nanobelts [50–53] and MoO2 synthesized by reduction MoO3 with ethanol vapor [54], Those all indicate that Mo-based materials have a potential to be anode of lithium ion batteries. The electrochemical impedance spectroscopy (EIS) is used to provide electrochemical insights. EIS is measured and fitted to the equivalent circuit. Fig. 7 shows the Nyquist plots of the electrodes before cycles for the three samples. Generally, the intercept at the Zre axis at high-frequency region is on behalf of the resistance of electrolyte (Rs). And semicircle, in the high-frequency region, can be put down to the charge transfer impedance (Rct). Charge transfer resistance is closely related to the conductivity of the working electrode material. The straight line in the low frequency region is related to Warburg diffusion [30]. The electrolyte resistances for the three electrodes are almost the same, and all exhibit a small Rs value about 6 V. In addition, the pristine MoO3 anode presents an Rct value about 200 V, and the HxMoO3(X = 0.28 and 0.33) nanobelts present a lower Rct of about 140 V and 107 V respectively. Therefore the electronic conductivity of HxMoO3 is higher than that of the MoO3. In addition, the conductivity of

Fig. 7. The Nyquist plots of the electrodes before cycles for MoO3 and HxMoO3(X = 0.28 and 0.33).

X. Ju et al. / Electrochimica Acta 213 (2016) 641–647

H0.33MoO3 is the best. However, we also need to take account for the stability of structure when applied to LIBs. 4. Conclusions In summary, low content of hydrogen can greatly improve the performance of a-MoO3 electrodes in Li-ion batteries, and HxMoO3 nanobelts have been synthesized by a facile hydrothermal method and first applied to LIBs as anode material, the designed H0.28MoO3 nanobeits exhibits outstanding high specific capacity and rate capability. The reversible capacity of H0.28MoO3 nanobelts is about 920 mAh g1, which is approaching the theory capacity of MoO3 nanobelts. When testing at high rate, the performance of our materials can maintain the capacity of about 550 mAh g1 after 450 cycles at 1 A g1. Even at 10 A g1, the capacity can still maintain at around 300 mAh g1. It is observed that the presence of hydrogen becomes more significant at high current densities. The excellent performance of H0.28MoO3 nanobelts can be attributed to the improved conductivity and the expanded lattice unit resulting from hydrogen-doping. The obtained results suggest that H0.28MoO3 can be required for future application with high energy density and long cycling life. Acknowledgements This work was partly supported by the National Natural Science Foundation of China (grant no. 61574118), and the Key Project of Science and Technology Plan of Fujian Province (grant No. 2015H0038). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. electacta.2016.07.160. References [1] F. Ma, A.B. Yuan, J.Q. Xu, P.F. Hu, Acs Appl Mater Inter 7 (2015) 15531–15541. [2] Z.Y. Zhang, W.Y. Li, T.W. Ng, W.P. Kang, C.S. Lee, W.J. Zhang, J Mater Chem A 3 (2015) 20527–20534. [3] Q. Xia, H.L. Zhao, Z.H. Du, Z.P. Zeng, C.H. Gao, Z.J. Zhang, X.F. Du, A. Kulka, K. Swierczek, Electrochim Acta 180 (2015) 947–956. [4] Y. Liu, B.H. Zhang, S.Y. Xiao, L.L. Liu, Z.B. Wen, Y.P. Wu, Electrochim Acta 116 (2014) 512–517. [5] S.H. Lee, Y.H. Kim, R. Deshpande, P.A. Parilla, E. Whitney, D.T. Gillaspie, K.M. Jones, A.H. Mahan, S.B. Zhang, A.C. Dillon, Adv Mater 20 (2008) 3627-+. [6] Y.D. Zhang, B.P. Lin, J.C. Wang, P. Han, T. Xu, Y. Sun, X.Q. Zhang, H. Yang, Electrochim Acta 191 (2016) 795–804. [7] W.W. Xia, Q.B. Zhang, F. Xu, L.T. Sun, Acs Appl Mater Inter 8 (2016) 9170–9177. [8] W. Zeng, G.H. Zhang, S.C. Hou, T.H. Wang, H.G. Duan, Electrochim Acta 151 (2015) 510–516. [9] N. Nitta, G. Yushin, Part Part Syst Char 31 (2014) 317–336. [10] W.X. Ji, D. Wu, R. Yang, W.P. Ding, L.M. Peng, Chinese J Inorg Chem 31 (2015) 659–665. [11] X.L. Hu, W. Zhang, X.X. Liu, Y.N. Mei, Y. Huang, Chem Soc Rev 44 (2015) 2376– 2404. [12] Y.N. Ko, S.B. Park, Y.C. Kang, Chem-Asian J 9 (2014) 1011–1015. [13] Z.X. Huang, Y. Wang, Y.G. Zhu, Y.M. Shi, J.I. Wong, H.Y. Yang, Nanoscale 6 (2014) 9839–9845. [14] D.S. Guan, J.Y. Li, X.F. Gao, C. Yuan, J Power Sources 246 (2014) 305–312.

647

[15] H.J. Zhang, T.H. Wu, K.X. Wang, X.Y. Wu, X.T. Chen, Y.M. Jiang, X. Wei, J.S. Chen, J Mater Chem A 1 (2013) 12038–12043. [16] Y. Liu, B.H. Zhang, Y.Q. Yang, Z. Chang, Z.B. Wen, Y.P. Wu, J Mater Chem A 1 (2013) 13582–13587. [17] J.P. Jegal, H.K. Kim, J.S. Kim, K.B. Kim, J Electroceram 31 (2013) 218–223. [18] T. Tao, A.M. Glushenkov, C.F. Zhang, H.Z. Zhang, D. Zhou, Z.P. Guo, H.K. Liu, Q.Y. Chen, H.P. Hu, Y. Chen, J Mater Chem 21 (2011) 9350–9355. [19] L.C. Yang, W.L. Guo, Y. Shi, Y.P. Wu, J Alloy Compd 501 (2010) 218–220. [20] E. Shembel, R. Apostolova, V. Nagirny, M. Chuiko, P. Grebenkin, P. Lytvyn, Elec Soc S (2003) 230–246. [21] J.W. Bullard, R.L. Smith, Solid State Ionics 160 (2003) 335–349. [22] L.A. Riley, S.H. Lee, L. Gedvilias, A.C. Dillon, J Power Sources 195 (2010) 588– 592. [23] L. Zhou, L.C. Yang, P. Yuan, J. Zou, Y.P. Wu, C.Z. Yu, J Phys Chem C 114 (2010) 21868–21872. [24] G.B. Wang, J.F. Ni, H.B. Wang, L.J. Gao, J Mater Chem A 1 (2013) 4112–4118. [25] D.O. Scanlon, G.W. Watson, D.J. Payne, G.R. Atkinson, R.G. Egdell, D.S.L. Law, J Phys Chem C 114 (2010) 4636–4645. [26] S.R. Cong, T. Sugahara, T.T. Wei, J.T. Jiu, Y. Hirose, S. Nagao, K. Suganuma, Cryst Growth Des 15 (2015) 4536–4542. [27] F.Y. Li, C.R. Cabrera, Z.F. Chen, J Mater Chem A 2 (2014) 19180–19188. [28] X.F. Yang, H.Y. Ding, D. Zhang, X.H. Yan, C.Y. Lu, J.L. Qin, R.X. Zhang, H. Tang, H.J. Song, Cryst Res Technol 46 (2011) 1195–1201. [29] F. Leroux, G.R. Goward, W.P. Power, L.F. Nazar, Electrochem Solid St 1 (1998) 255–258. [30] H. Zhang, X.J. Liu, R.L. Wang, R. Mi, S.M. Li, Y.H. Cui, Y.F. Deng, J. Mei, H. Liu, J Power Sources 274 (2015) 1063–1069. [31] Q. Xia, H.L. Zhao, Z.H. Du, J. Wang, T.H. Zhang, J. Wang, P.P. Lv, J Power Sources 226 (2013) 107–111. [32] L.L. Si, Z.Q. Yuan, J.W. Liang, L. Hu, Y.C. Zhu, Y.T. Qian, J Mater Chem A 2 (2014) 9784–9791. [33] B. Ahmed, M. Shahid, D.H. Nagaraju, D.H. Anjum, M.N. Hedhili, H.N. Alshareef, Acs Appl Mater Inter 7 (2015) 13154–13163. [34] Y.P. Song, H. Wang, Z.H. Li, N.Q. Ye, L.J. Wang, Y. Liu, Int J Hydrogen Energ 40 (2015) 3613–3623. [35] L. Huang, X. Gao, Q. Dong, Z. Hu, X. Xiao, T. Li, Y. Cheng, B. Yao, J. Wan, D. Ding, Z. Ling, J. Qiu, J. Zhou, J. Mater. Chem. A 3 (2015) 17217–17223. [36] J. Lu, W.S. Li, J.H. Du, J.M. Fu, J New Mat Electr Sys 8 (2005) 5–14. [37] H. Noh, D. Wang, S. Luo, T.B. Flanagan, R. Balasubramaniam, Y. Sakamoto, J Phys Chem B 108 (2004) 310–319. [38] X.K. Hu, Y.T. Qian, Z.T. Song, J.R. Huang, R. Cao, J.Q. Xiao, Chem Mater 20 (2008) 1527–1533. [39] K. Eda, A. Sukejima, N. Sotani, Journal of Solid State Chemistry 159 (2001) 51– 58. [40] Z.H. Li, H. Wang, Z.Z. Sun, J. Su, Z.Y. Wang, L.J. Wang, J Mater Chem A 4 (2016) 303–313. [41] J.D. Guo, P. Zavalij, M.S. Whittingham, Journal of Solid State Chemistry 117 (1995) 323–332. [42] H.C. Zeng, W.K. Ng, L.H. Cheong, F. Xie, R. Xu, J Phys Chem B 105 (2001) 7178– 7181. [43] X.K. Hu, D.K. Ma, L.Q. Xu, Y.C. Zhu, Y.T. Qian, Chem Lett 35 (2006) 962–963. [44] I.Z. Balta, S. Pederzoli, E. Iacob, M. Bersani, Appl Surf Sci 255 (2009) 6378–6385. [45] T. Brezesinski, J. Wang, S.H. Tolbert, B. Dunn, Nat Mater 9 (2010) 146–151. [46] M. Sasidharan, N. Gunawardhana, H. Noma, M. Yoshio, K. Nakashima, B Chem Soc Jpn 85 (2012) 642–646. [47] D.P. Cai, B. Liu, D.D. Wang, L.L. Wang, Y. Liu, B.H. Qu, X.C. Duan, Q.H. Li, T.H. Wang, J Mater Chem A 4 (2016) 183–192. [48] C.Z. Yuan, J.Y. Li, L.R. Hou, L.H. Zhang, X.G. Zhang, Part Part Syst Char 31 (2014) 657–663. [49] D.H. Li, D.J. Yang, X.Y. Zhu, D.W. Jing, Y.Z. Xia, Q. Ji, R.S. Cai, H.L. Li, Y.K. Che, J Mater Chem A 2 (2014) 18761–18766. [50] Y. Zhou, Q. Liu, D.B. Liu, H. Xie, G.X. Wu, W.F. Huang, Y.F. Tian, Q. He, A. Khalil, Y. A. Haleem, T. Xiang, W.S. Chu, C.W. Zou, L. Song, Electrochim Acta 174 (2015) 8– 14. [51] D.Y. Park, Y.K. Sun, S.T. Myung, J Power Sources 280 (2015) 1–4. [52] X.H. Sun, Y.F. Shi, X.P. Fang, H.M. Ji, X.L. Li, S. Cai, C.M. Zheng, Y.S. Hu, J Nanosci Nanotechno 14 (2014) 4278–4285. [53] L. Zhou, H.B. Wu, Z.Y. Wang, X.W. Lou, Acs Appl Mater Inter 3 (2011) 4853– 4857. [54] L.C. Yang, L.L. Liu, Y.S. Zhu, X.J. Wang, Y.P. Wu, J Mater Chem 22 (2012) 13148– 13152.