PCCP View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
PAPER
Cite this: Phys. Chem. Chem. Phys., 2014, 16, 20570
View Journal | View Issue
MoO2@carbon hollow microspheres with tunable interiors and improved lithium-ion battery anode properties† Xiaolin Liu, Wenxu Ji, Jiyuan Liang, Luming Peng and Wenhua Hou* MoO2 hollow microspheres with tunable inner space have been synthesized through a hydrothermal process using MoO3 microbelts instead of bulk MoO3 as the precursor. It is found that the reactant morphology has a great impact on the product morphology and the inner space can be tuned by changing the amount of NaOH aqueous solution. An interesting evolutional process from MoO3 microbelts through a rose-like intermediate to MoO2 hollow microspheres has been clearly observed, and thus the possible formation mechanism is revealed. One layer of amorphous carbon has been subsequently coated on the surface of MoO2 hollow microspheres through a simple hydrothermal approach followed by annealing in argon. As the anode material for lithium ion batteries, MoO2@C hollow microspheres manifest excellent lithium-storage properties, such as high capacity (677 mA h g 1)
Received 7th July 2014, Accepted 6th August 2014
and good cycling stability (negligible capacity fading even after 80 cycles). The significantly enhanced
DOI: 10.1039/c4cp02960g
performance of MoO2@C hollow microspheres can be attributed to its unique structures, such as nano-
www.rsc.org/pccp
scaled primary building blocks, carbon coating, hollow structure, and especially the synergy between the carbon coating and hollow structure.
Introduction Owing to their longer cycling life, better environmental benignity, higher energy and power densities, lower self-discharge rate and higher voltage compared with other secondary batteries, lithium ion batteries (LIBs) have been studied extensively and intensively.1–7 However, the current anode material, namely, graphite suffers from a low theoretical capacity (370 mA h g 1) and a large irreversible capacity.8 Therefore, it is an urgent task to search for an alternative material for graphite. Molybdenum dioxide (MoO2) has been considered as one of the promising anode materials for LIBs due to its low electrical resistivity, high chemical and thermal stability, and large theoretical capacity (838 mA h g 1).8–10 In addition, its Li+ insertion voltage is higher than that of the graphite-based anode, which also makes MoO2 attract considerable attention as a safe anode material. Unfortunately, MoO2 has two inadequacies that greatly hinder its practical application. First, a large volume variation occurs during the Li+ insertion/extraction into/from the MoO2 host, resulting in a huge loss in electrical connectivity with
Key Laboratory of Mesoscopic Chemistry of MOE, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, P. R. China. E-mail: [email protected] † Electronic supplementary information (ESI) available. See DOI: 10.1039/ c4cp02960g
20570 | Phys. Chem. Chem. Phys., 2014, 16, 20570--20577
the current collector, which is generally called electrode pulverization. Second, the slow kinetics in bulk MoO2 severely restricts the redox conversion reaction even if it is thermodynamically allowed, which will lead to a tremendous loss in its reversible capacity.11 Although these problems cannot be eliminated completely owing to the fundamental electrochemical process, several strategies have been proposed. The first one is the introduction of nanotechnology into the fabrication process. Due to the shorter lithium/electron diffusion paths, it is believed that nanostructured electrode materials better facilitate the transportation of Li+ into the electrode body center and therefore better benefit the redox conversion reaction.12 The second one is coating MoO2 with a layer of flexible and conductive substance, such as amorphous carbon or polyaniline, to counteract the pulverization effectively. At the same time, the electrical conductivity of MoO2 anode electrode can also be improved by the wrapping layer.12,13 The third one is better designing the desired structured MoO2 electrode, for example as a hollow structure or a one-dimensional structure.14–18 On the one hand, these kinds of materials can endure the volume variation effectively, thus helping to improve the lithium-storage properties. On the other hand, these materials always have a large specific surface area, which means that they can interact with Li+ ions more quickly and relieve the polarization. Overall, the best method is no doubt the combination of all the above mentioned three approaches: nanotechnology,
This journal is © the Owner Societies 2014
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
Paper
coating and designing desirable structures. Monodisperse porous LiFePO4@C microspheres,12 Fe2O3@polyaniline hollow spheres2 and SnO2@C hierarchical tube15 have been prepared, and all of them combined the aforementioned three design rationales for high-performance anode and thus exhibited excellent Li+ storage properties. However, the preparation processes involved were too cumbersome and even needed hard templates. To date, although MoO2 hollow microspheres have been reported, the synthesized MoO2 hollow microspheres have had a very broad size distribution and lacked regularity.19–21 Moreover, their lithium-storage properties have been unsatisfactory19 or remained unknown.20,21 In the present work, we have successfully fabricated MoO2@C hollow microspheres through a simple hydrothermal process. More importantly, the inner void space in the hollow structure could be tuned. It is noteworthy that the MoO2@C hollow microspheres synthesized simultaneously integrate the aforementioned three design rationales for high-performance anode materials, namely, nanostructured, carbon wrapped and unique hollow structured, and thus exhibit high specific capacity and Coulombic efficiency, excellent cycling stability and superior rate capability.
Experimental section 2.1
Materials
Bulk MoO3, (NH4)6Mo7O244H2O (AHM), EDTA–2Na (Na2H2Y), HCl and NaOH aqueous solution, glucose. 2.2
Synthesis of MoO2 agglomerated microspheres (MoO2-AMS)
MoO2 agglomerated microspheres were synthesized via the hydrothermal reaction of a mixture of 0.400 g bulk MoO3, 0.200 g EDTA–2Na and 45 mL distilled water in a 60 mL Teflonlined autoclave at 200 1C for 48 h.22 It is noteworthy that bulk MoO3 cannot be uniformly suspended in the aqueous solution. The final product was collected by centrifugation with distilled water several times, and dried at 70 1C for 24 h. The sample was denoted as MoO2-AMS. 2.3
Synthesis of MoO3 microbelts
MoO3 microbelts were synthesized by the hydrothermal reaction of a mixture solution of 0.980 g (NH4)6Mo7O244H2O (AHM), 14 mL HCl (3 M) and 31 mL distilled water in a 60 mL Teflon-lined autoclave at 180 1C for 20 h.23,24 HCl was used instead of HNO3 to avoid the residual HNO3 which would oxidize EDTA–2Na in the next procedure. The as-synthesized MoO3 microbelts were collected by centrifugation with distilled water several times, and dried at 70 1C for 24 h. 2.4
Synthesis of MoO2 hollow microspheres (MoO2-HMS)
MoO2 hollow microspheres were synthesized by using the as-prepared MoO3 microbelts and EDTA–2Na as the starting materials. In a typical synthesis, 0.400 g of MoO3 microbelts were uniformly dispersed in 25 mL distilled water with stirring for 10 min, forming the colloidal solution A. 0.200 g EDTA–2Na was dissolved in 20 mL distilled water, and then 0.15 mL NaOH
This journal is © the Owner Societies 2014
PCCP
aqueous solution (3 M), ca. 0.45 mmol, was dripped into the above solution, forming solution B. Solution B was then added into the well-stirred colloidal solution A. The mixture was allowed to stir for another 30 min at room temperature, and then the resulting mixture was transferred to and sealed in a 60 mL Teflon-lined autoclave, heated to 200 1C and maintained at this temperature for 48 h. The final product was collected by centrifugation with distilled water several times, and dried at 70 1C for 24 h. This sample was denoted as MoO2-HMS. 2.5
Synthesis of MoO2@C hollow microspheres (MoO2@C-HMS)
The as-prepared MoO2-HMS was coated with one layer of amorphous carbon through a simple hydrothermal process.12 In a typical synthesis, 0.05 g of as-prepared MoO2-HMS was dispersed into 35 mL of 0.075 M aqueous glucose solution by ultrasonication. The black suspension was transferred into a 60 mL Teflon-lined autoclave, then heated to 170 1C and maintained at this temperature for 8 h. The product was collected by centrifugation with distilled water and ethanol six times, and dried at 60 1C. Finally, the resulting black powder was carbonized at 600 1C for 4 h under an argon atmosphere to obtain the MoO2@C hollow microspheres. This sample was denoted as MoO2@C-HMS. 2.6
Characterization
The phase purity of the resulting products was identified by powder X-ray diffraction (XRD) on a Philip-X’Pert X-ray diffractometer with Cu Ka radiation (l = 1.5418 Å). The elemental analysis was carried out using a CHN-O-Rapid (Heraeus, Germany). The product morphology was observed using scanning electron microscopy (SEM JEOL JEM-6300F) and transmission electron microscopy (TEM JEOL JEM-200CX, operated at an accelerating voltage of 200 kV). Selected-area electron diffraction (SAED) and high-resolution TEM (HRTEM) spectra were examined under a JEM-2010F high-resolution transmission electron microscope. Thermogravimetric-differential thermal analysis (TG-DTA) was carried out on an SDT600 apparatus with a heating rate of 10 1C min 1 in an air atmosphere. 2.7
Electrochemical measurements
The electrochemical tests were conducted using a coin cell. An assembled coin cell consists of lithium foil as the counter electrode and a working electrode which is composed of 80% active material, 10% super P carbon black and 10% polyvinylidene fluoride (PVDF) and then made on copper foil. The mass coating of active material is ca. 2.00 mg cm 2. All cells were assembled in an argon-filled glove box (Mikarouna, Superstar 1220/750/900) by using 1 M LiPF6 solution in ethylene carbonate/ diethyl carbonate (EC : DEC = 1 : 1, v/v) as the electrolyte and a Celgard 2400 as the separator. The galvanostatic cycling test was performed on a battery test system (Land-CT2001A) at a constant current density of 100, 200, 500 and 1000 mA g 1 in the potential range from 0.005 to 3.0 V. Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (frequency range: 0.001–105 Hz) were carried out on an electrochemical workstation (CHI 660D).
Phys. Chem. Chem. Phys., 2014, 16, 20570--20577 | 20571
View Article Online
PCCP
Paper
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
Results and discussion To investigate the effect of product morphology on the Li+storage property, monoclinic MoO2 agglomerated microspheres (MoO2-AMS) were also prepared by using bulk MoO3 and EDTA–2Na as the starting materials. The powder XRD pattern of MoO2-AMS is shown in Fig. 1a, indicating that the bulk MoO3 has been entirely transformed into monoclinic MoO2 (JCPDS 86-0135). Fig. S1 (ESI†) shows the XRD pattern of MoO3 microbelts (the precursor of MoO2 hollow microspheres) and it can be indexed to pure orthorhombic MoO3 (JCPDS 05-0508). The unusually intense (020), (040) and (060) diffraction peaks demonstrate the apparent preferred orientation, which implies that the MoO3 precursor mainly grows along one dimension. The powder XRD pattern of the as-prepared MoO2 hollow microspheres (MoO2-HMS) is shown in Fig. 1a, and it can also be easily indexed to monoclinic MoO2, confirming its pure phase nature. The broadened diffraction peaks compared with those of MoO2-AMS demonstrates that MoO2-HMS is assembled from smaller building blocks. After carbon coating, the XRD pattern of MoO2@C-HMS shows no obvious change in comparison with that of MoO2-HMS, indicating that the crystal structure of MoO2 did not change and the carbon layer is amorphous. The SEM images of bulk MoO3 and MoO2-AMS are shown in Fig. S2a–c (ESI†). It is clear that MoO2-AMS is composed of severely agglomerated microspheres with a size of ca. 10 mm, and these agglomerated microspheres are assembled from stacked nanoplates with a diameter of ca. 200 nm. Fig. S2d and e (ESI†) shows the SEM and TEM images of the as-synthesized MoO3 microbelts with an aspect ratio of ca. 20, which verifies its one dimensional growth nature and is in accordance with the XRD results. When MoO3 microbelts instead of bulk MoO3 are used as the precursor while other reaction conditions remain unchanged, the product takes on the form of irregular porous and hollow quasi-microspheres (Fig. S2f, ESI†). Furthermore, when 0.15 mL NaOH aqueous solution (3 M) was dripped into the EDTA–2Na
Fig. 1 (a) XRD patterns of MoO2-AMS, MoO2-HMS and MoO2@C-HMS; (b) SEM image of MoO2-HMS; (c) SEM image of MoO2@C-HMS; (d) magnified SEM image of MoO2@C-HMS.
20572 | Phys. Chem. Chem. Phys., 2014, 16, 20570--20577
Fig. 2 (a–c) TEM images of MoO2@C-HMS with different magnifications; (d) HRTEM image recorded within the red box of (b), its inset is the corresponding SAED pattern; (e) TEM image of MoO2@C-HMS when 0.25 mL NaOH aqueous solution (3 M) was dripped into the EDTA–2Na stock aqueous solution; (f) TG-DSC curves of MoO2@C-HMS conducted under air flow.
stock aqueous solution, a considerable change in the product morphology occurred. As illustrated in Fig. 1b, the resulted product is composed of microspheres with a size in the range of 0.5–2.0 mm. After carbon coating, the product morphology shows no obvious changes (Fig. 1c). Upon closer observation, the microspheres are found to be composed of many secondary units (Fig. 1d). The morphology and structure of MoO2@C-HMS were further studied by TEM characterizations. As shown in Fig. 2a and b, the hollow structure can be clearly confirmed, and may effectively accommodate the volume change that occurs during the Li+ insertion/extraction.25–30 Additionally, the fringes of a hollow microsphere (see the magnified TEM images in Fig. S3, ESI†), clearly show these microspheres are self-assembled from nanoslabs. As shown in Fig. 2c, one layer of amorphous carbon with a thickness of ca. 6 nm was uniformly coated on the surface of MoO2 hollow microspheres. Fig. 2d shows the HRTEM image of a nanoslab (recorded within the red square of Fig. 2b), and the clear lattice fringes demonstrate that the sample is well-crystallized. The spacing of adjacent lattice fringes was determined to be 0.2408 nm, which is in good agreement with the interplanar spacing of the (200) planes of MoO2. The bright diffraction spots in the inset SAED pattern also indicate the single crystalline nature of the nanoslabs. It is worth mentioning that the inner void space of the hollow structure could be easily controlled through changing the addition amount of NaOH aqueous solution. Fig. 2e shows the TEM image of MoO2@C-HMS when 0.25 mL NaOH aqueous
This journal is © the Owner Societies 2014
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
Paper
solution (3 M) was dripped into the EDTA–2Na stock aqueous solution. By comparing Fig. 2e with Fig. 2a, one may find that the inner void space is apparently diminished. In addition, the product size is also decreased from ca. 1.3 mm to ca. 0.8 mm. Fig. 2f shows the thermogravimetric-differential scanning calorimetry (TG-DSC) curves of MoO2@C-HMS. The slight weight change that occurs between 30 and 300 1C can be attributed to the evaporation of the adsorbed water. In the next step, the weight change between 350 and 600 1C is related to the oxidation of MoO2 and combustion of coated carbon, and correspondingly a large exothermic peak emerges in the DSC curve. Finally, the endothermic peak at ca. 780 1C is caused by the sublimation of MoO3. Since the theoretical value of the mass increase from MoO2 to MoO3 is 12.5 wt%, the content of the coated carbon was accordingly estimated to be ca. 14.67 wt% in the MoO2@C-HMS nanocomposite. Subsequent CHN analysis also demonstrated that the carbon content in MoO2@C-HMS was about 14.47 wt%, which is in good agreement with the results of TG-DSC. In order to figure out the formation mechanism of MoO2 hollow microspheres, the time-dependent experiments were carried out. The XRD pattern of the intermediate after hydrothermal treatment for 12 h is shown in Fig. 3, and its phase can be indexed as a mixture of orthorhombic MoO3 (JCPDS card 05-0508) and orthorhombic [Na(H2O)5]0.25 MoO3 (JCPDS card 39-1112). The corresponding SEM images are shown in Fig. 4a–d. It mainly consists of MoO3 microbelts (see the red ellipse of Fig. 4b) and a very beautiful rose-like intermediate. Combined with the XRD results, it can be deduced that the rose-like intermediate is probably [Na(H2O)5]0.25MoO3. After a careful observation of Fig. 4c (especially within the red rectangle of Fig. 4d), it can be found that the petals of microroses are produced by the merging of MoO3 microbelts. The XRD pattern of the intermediate after hydrothermal treatment for 15 h is shown in Fig. 3, and its phase can be mainly attributed to the mixture of
Fig. 3 XRD patterns of the intermediates and product after hydrothermal treatment for 12 h, 15 h, 30 h and 48 h, respectively. The standard data for orthorhombic [Na(H2O)5]0.25MoO3 (red, JCPDS card 39-1112), orthorhombic MoO3 (black, JCPDS card 05-0508) and monoclinic MoO2 (blue, JCPDS card 86-0135) are also presented at the bottom for comparison.
This journal is © the Owner Societies 2014
PCCP
Fig. 4 SEM images of intermediates obtained after hydrothermal treatment for 12 h (a–d), 15 h (e–g) and 30 h (h). The right corner of (h) is the corresponding TEM image.
orthorhombic [Na(H2O)5]0.25MoO3 (JCPDS card 39-1112) and monoclinic MoO2 (JCPDS card 86-0135). It is noted that the orthorhombic MoO3 phase has disappeared. The corresponding SEM images are shown in Fig. 4e–g. It is mainly composed of microroses and many small spheres (see Fig. 4e). Combined with the XRD results, it can be deduced that the small spheres are monoclinic MoO2. Within the red circle of Fig. 4f, an incomplete hollow sphere can be seen among the petals. It implies that the formation of MoO2 hollow spheres is closely related to the petals. More importantly, the microrose intermediate after hydrothermal treatment for 15 h has surfaces with holes and very blurred edges (see Fig. 4f and g), while the microrose intermediate after hydrothermal treatment for 12 h has very smooth surfaces and edges. In the light of the above discussion, it can be deduced that the petals of microroses were first reduced to MoO2 nanoslabs and then, these nanoslabs self-assembled into MoO2 hollow spheres. The XRD pattern of the intermediate after hydrothermal treatment for 30 h is shown in Fig. 3, and the product phase can be mainly indexed to monoclinic MoO2. The phase of orthorhombic [Na(H2O)5]0.25MoO3 nearly disappears, and only an ultra-weak peak (2y = 7.61) can be seen. Its SEM image is exhibited in Fig. 4h, and the product is composed of many microspheres with
Phys. Chem. Chem. Phys., 2014, 16, 20570--20577 | 20573
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
PCCP
Scheme 1
Paper
An illustration of the formation process of MoO2-HMS.
a size of ca. 1 mm, and its corresponding TEM image in the right corner manifests the hollow structure of these microspheres, demonstrating that the MoO3 microbelts have been entirely transformed into MoO2 hollow microspheres. Finally, based on the experimental facts and above discussions, the formation mechanism of MoO2 hollow microspheres could be proposed as follows. First, MoO3 microbelts merged into [Na(H2O)5]0.25MoO3 microroses at high temperature and pressure in the presence of Na+ (the ‘‘templating ion’’), since MoO3 often acted as a host for intercalation chemistry.3 Second, the petals of [Na(H2O)5]0.25MoO3 microroses were reduced to MoO2 nanoslabs by H2Y2 (Na2H2Y denotes EDTA–2Na) at high temperature and pressure. Finally, MoO2 nanoslabs tended to self-assemble into a spherical shape to decrease the surface energy. The possible formation process is illustrated in Scheme 1. Nevertheless, how did these MoO2 nanoslabs self-assemble into hollow microspheres? It can be explained as follows: first, a large number of Na+ ions would be adsorbed on the surface of MoO2 nanoslabs, and the numerous nanoslabs would agglomerate into spherical shape to decrease the surface energy. It is noteworthy that the volumetric charge density, and thus the repulsion, is larger on the inside of the sphere than that on the outside. Therefore, many microspheres were cracked and only incomplete spheres were formed (see Fig. S2f, ESI†). When 0.15 mL NaOH aqueous solution was dripped into the EDTA–2Na stock aqueous solution, the pH value and OH concentration in the reaction system increased. It is noted that OH can also adsorb on the surface of the nanoslabs. The adsorbed OH lowered the repulsion between the nanoslabs through the attraction between Na+ and OH , and then completely hollow spheres were formed (see Fig. 1b). As the amount of NaOH aqueous solution was further increased, the repulsion between nanoslabs was further lowered, and then the nanoslabs were piled up more closely, as evidenced by the decreased product size. Consequently, the inner void space was diminished (see Fig. 2e). In short, EDTA–2Na has played two roles in this synthesis, i.e., as a reducing agent and a morphology modulator. The possible formation mechanism is illustrated in Scheme S1 (ESI†). Fig. 5a shows the cyclic voltammograms (CV) of MoO2@C-HMS electrode at a scanning rate of 0.2 mV s 1 between 0.005 and 3.0 V.
20574 | Phys. Chem. Chem. Phys., 2014, 16, 20570--20577
Fig. 5 (a) Cyclic voltammograms of MoO2@C-HMS at a scanning rate of 0.2 mV s 1; (b) galvanostatic discharge–charge curves of MoO2@C-HMS at a current rate of 100 mA g 1; (c) cycling performance of MoO2-AMS, MoO2-HMS and MoO2@C-HMS at a current density of 100 mA g 1; (d) rate capability of MoO2@C-HMS at different current densities.
There is an apparent peak dislocation between the first cycle and the following cycles. From the second cycle onward, the CV profiles are almost fully-overlapped, indicating the high reversibility of Li+ insertion and extraction. Two pronounced redox couples at 1.52/1.72 V and 1.25/1.48 V in the second and fifth cycles are associated with the reversible phase transitions of LixMoO2 from monoclinic to orthorhombic and then from orthorhombic back to monoclinic.31–34 Fig. 5b displays the galvanostatic charge and discharge curves for MoO2@C-HMS electrode at a current density of 100 mA g 1 in the voltage range of 0.005 to 3.0 V (vs. Li+/Li). It can be found that the initial charge and discharge capacities are 376.4 and 478.9 mA h g 1, respectively. An irreversible capacity loss of 102.5 mA h g 1 is mainly attributed to some irreversible processes such as the formation of a solid electrolyte interface (SEI), trapping of some Li+ in the MoO2 lattice and partial decomposition of the electrolyte,11,35,36 leading to a relatively low initial Coulombic efficiency of 78.6%. In the first discharge–charge cycle, there are two obvious discharge voltage
This journal is © the Owner Societies 2014
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
Paper
PCCP
plateaus at 1.54 and 1.28 V and two charge ones at 1.41 and 1.71 V, being consistent with the CV results. These observed voltage plateaus are attributed to the reversible phase transitions of LixMoO2 from monoclinic to orthorhombic and then back to monoclinic.37 It can also be observed in Fig. 5b that the reversible capacity increases gradually upon cycling until the 30th cycle. Fig. 5c shows the cycling stability of MoO2-AMS, MoO2-HMS and MoO2@C-HMS electrodes at a current density of 100 mA h g 1 in the potential window of 0.005 and 3.0 V. The MoO2-AMS electrode delivered a maximum capacity of 539.1 mA h g 1 at the 15th cycle, and the reversible capacity drastically faded to 99.6 mA h g 1 after 50 cycles. By comparison, the MoO2-HMS electrode delivered a maximum capacity of 675.1 mA h g 1 at the 15th cycle, and the reversible capacity faded to 358.7 mA h g 1 after 50 cycles. The obviously improved reversible capacity and cycling stability of MoO2-HMS could be ascribed to its nanosized building blocks and hollow structure.11,38–41 First, the nanosized building blocks provide the shorter Li+/electron diffusion path, and better facilitate the transportation of Li+ into the electrode body center, and hence the reversible capacity is improved. Second, the hollow structure could better sustain the volume variation during the lithiation and delithiation, and thus the cycling performance is enhanced.12 It is noteworthy that the MoO2@C-HMS electrode could deliver a capacity of 677.5 mA h g 1 at the 30th cycle, and a capacity of 677.4 mA h g 1 was retained even after 80 cycles with a retention close to 100%. In addition, its Coulombic efficiency has been higher than 97% since the fifth cycle (see Fig. S4, ESI†). There is no doubt that MoO2@C-HMS has a significantly improved cycling stability over MoO2-HMS. The remarkably improved cycling stability of MoO2@C-HMS can be ascribed to the synergistic effect between the porous hollow structure and carbon coating, as suggested by the Cui group.42,43 As shown in Scheme S2 (ESI†), the inner void space renders MoO2 to expand inwards instead of outwards during Li+ insertion, and thus the structure can be restored to its original state upon Li+ extraction and also the carbon layer will not be easily destroyed during lithiation and delithiation. In contrast, if MoO2 were to be a solid sphere, it had to expand outwards since there was no inner void space. As a result, the carbon wrapping would be cracked during Li+ insertion and extraction after several cycles. It is also noteworthy that the reversible capacity of MoO2@C-HMS increases drastically in the initial cycles, which is in agreement with the results of Fig. 5b. It can be attributed to the following two reasons. First of all, the process of Li+ insertion/extraction into/from the MoO2 host can be divided into the following two steps:35 xLi+ + xe + MoO2 2 LixMoO2 (0 o x o 0.98)
(1)
Li0.98MoO2 + 3.02Li+ + 3.02e 2 2Li2O + Mo
(2)
The second (latter) step is also named the conversion reaction of MoO2. Fig. S5 (ESI†) shows the discharge curves of MoO2@C-HMS at the 6th, 10th, 14th, 18th, 22th and 26th cycle, respectively. The capacity below 0.6 V in the discharge curve is
This journal is © the Owner Societies 2014
caused by the conversion reaction as suggested by many previous studies.10 It can be clearly seen that the released capacity during this stage (conversion reaction stage), increases gradually with the cycles. In other words, the conversion reaction of MoO2 becomes more complete with the cycles. This is also called the activation process of MoO2. We also note that the released maximum capacity of MoO2-HMS is at the 13th cycle, while that of MoO2@C-HMS is at the 30th cycle. The sole difference between MoO2-HMS and MoO2@C-HMS is the carbon coating. There is no doubt that the wetting of electrolyte into the electrode material will be delayed by the wrapping layer out of MoO2-HMS. It is also necessary to note that Li+ is dissolved in the electrolyte, and Li+ only moves in the electrolyte. Therefore, only after the complete wetting of electrolyte into the electrode material can Li+ freely pass through the whole electrode material and fully release its Li+ storage capacity. Similar experimental phenomena were also observed by other groups.6 The rate capability of MoO2@C-HMS electrode is presented in Fig. 5d. A steady capacity of ca. 455 mA h g 1 can be obtained even at a current density of 1 A g 1. Moreover, the capacity can reach ca. 680 mA h g 1 when the current density is decreased to 100 mA g 1 after 40 cycles. The superior rate performance of MoO2@C-HMS electrode is closely related to its porous structure and carbon wrapping.44–46 The porous structure could make Li+ penetrate the inner space and provide more active sites for Li+ intercalation. As shown in Fig. S6 (ESI†), the carbon wrapping makes MoO2@C-HMS electrode possess relatively low charge-transfer impedance. Therefore, the electrons transfer more quickly after carbon coating, being conducive to the rate capability. When compared with the previously reported results on similar systems,9,10,31,34,47 the test results of our lithium storage performance is rather attractive if one takes into account the testing conditions used in our work (e.g. the prepared MoO2@C-HMS has not been hybridized with graphene or carbon nanotubes, PVDF instead of sodium alginate was used as the binder, and a mass coating of active material was ca. 2.00 mg cm 2).
Conclusions In summary, MoO2@C hollow microspheres have been successfully synthesized through employment of MoO3 microbelts instead of bulk MoO3 as the precursor. The inner void space of the resulted hollow structure could be easily tuned by changing the addition amount of NaOH aqueous solution. The evolutional process from MoO3 microbelts through a rose-like intermediate to MoO2 hollow microspheres has been disclosed. The as-prepared MoO2@C hollow microspheres simultaneously satisfy three desirable characters for highperformance anode materials, i.e., nanoscaled primary building blocks, highly-conductive coating layer of carbon and hollow structure. Owing to these unique structure characters and the synergy among them, the prepared MoO2@C porous hollow
Phys. Chem. Chem. Phys., 2014, 16, 20570--20577 | 20575
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
PCCP
microspheres exhibited excellent Li+-battery anode properties. The nanoscaled primary building blocks contributed to its high reversible capacity; while carbon coating greatly lowered charge-transfer impedance and thus improved its rate capability. Moreover, carbon-encapsulated MoO2 hollow microspheres effectively made the anode material expand inwards instead of outwards during the Li+ insertion, rendering the material to better sustain the volume variation and thus guaranteeing its long cycling stability. The present work provides a new method for the fabrication of hollow-structured materials. In addition, MoO2 with this unique structure may also find potential application in the catalysis field.
Acknowledgements This project was supported by the National Natural Science Foundation of China (Grant No. 21073084), the Specialized Research Fund for the Doctoral Program of Higher Education (Grant No. 20130091110010), the Natural Science Foundation of Jiangsu Province (Grant No. BK2011438), the National Science Fund for Talent Training in Basic Science (No. J1103310), the National Basic Research (973) Program of China (Grant No. 2009CB623504), and the Modern Analysis Center of Nanjing University.
Notes and references 1 J. B. Goodenough and K. S. Park, J. Am. Chem. Soc., 2013, 135, 1167–1176. 2 J. M. Jeong, B. G. Choi, S. C. Lee, K. G. Lee, S. J. Chang, Y. K. Han, Y. B. Lee, H. U. Lee, S. Kwon, G. Lee, C. S. Lee and Y. S. Huh, Adv. Mater., 2013, 25, 6250–6255. 3 M. S. Whittingham, J. D. Guo, R. J. Chen, T. Chirayil and G. Janauer, Solid State Ionics, 1995, 75, 257–268. 4 X. K. Guo, Q. Fan, L. Yu and J. Y. Liang, J. Mater. Chem. A, 2013, 1, 11534–11538. 5 M. J. Armstrong, C. Dwyer, W. J. Macklin and J. D. Holmes, Nano Res., 2014, 7, 1–62. 6 H. Wu, G. Yu, L. Pan, N. Liu, M. T. McDowell, Z. Bao and Y. Cui, Nat. Commun., 2013, 4, 1943–1948. 7 H. L. Wang, L. F. Cui, Y. Yang, Y. Cui and H. J. Dai, J. Am. Chem. Soc., 2010, 132, 13978–13980. 8 A. Bhaskar, M. Deepa and T. N. Rao, ACS Appl. Mater. Interfaces, 2013, 5, 2555–2566. 9 Y. F. Shi, B. G. Guo and L. Q. Chen, Nano Lett., 2009, 9, 4215–4220. 10 B. Guo, X. Fang, B. Li, Y. Shi, C. Ouyang, Y.-S. Hu, Z. Wang, G. D. Stucky and L. Chen, Chem. Mater., 2012, 24, 457–463. 11 Z. Wang, J. S. Chen, T. Zhu, S. Madhavi and X. W. Lou, Chem. Commun., 2010, 46, 6906–6908. 12 L. Zhang, G. Zhang, H. B. Wu, L. Yu and X. W. Lou, Adv. Mater., 2013, 25, 2589–2593. 13 L. Yang, S. Wang, J. Mao, J. Deng, Q. Gao, Y. Tang and O. G. Schmidt, Adv. Mater., 2013, 25, 1180–1184.
20576 | Phys. Chem. Chem. Phys., 2014, 16, 20570--20577
Paper
14 C. Y. Nan, J. Lu, L. H. Li, L. L. Li, Q. Peng and Y. D. Li, Nano Res., 2013, 6, 469–477. 15 C. Sun, S. Rajasekhara, J. B. Goodenough and F. Zhou, J. Am. Chem. Soc., 2011, 133, 2132–2135. 16 Y. G. Li, B. Tan and Y. Y. Wu, J. Am. Chem. Soc., 2006, 128, 14258–14259. 17 B. Song, S. W. Liu and J. K. Jian, J. Power Sources, 2008, 180, 869–874. 18 Y. G. Li, B. Tan and Y. Y. Wu, Chem. Mater., 2008, 20, 567–576. 19 Y. Z. Lei, J. C. Hu, H. W. Liu and J. L. Li, Mater. Lett., 2012, 68, 82–85. 20 Z. Q. Wang, H. F. Wang, C. Yang and J. H. Wu, Mater. Lett., 2010, 64, 2170–2172. 21 H. X. Zhang, Y. F. Li, Z. S. Hong and M. D. Wei, Mater. Lett., 2012, 79, 148–151. 22 R. Q. Song, A. W. Xu, B. Deng and Y. P. Fang, J. Phys. Chem. B, 2005, 109, 22758–22766. 23 X. W. Lou and H. C. Zeng, Chem. Mater., 2002, 14, 4781–4789. 24 Y. Chen, C. Lu, L. Xu, Y. Ma, W. Hou and J.-J. Zhu, CrystEngComm, 2010, 12, 3740–3747. 25 B. K. Guo, M. F. Chi, X. G. Sun and S. Dai, J. Power Sources, 2012, 205, 495–499. 26 W. Wang, Q. Sa, J. Chen, Y. Wang, H. Jung and Y. Yin, ACS Appl. Mater. Interfaces, 2013, 5, 6478–6483. 27 A. Vu, Y. Qian and A. Stein, Adv. Energy Mater., 2012, 2, 1056–1085. 28 S. A. Klankowski, R. A. Rojeski, B. A. Cruden, J. Liu, J. Wu and J. Li, J. Mater. Chem. A, 2013, 1, 1055–1064. 29 Q. L. Wu, J. Li, R. D. Deshpande, N. Subramanian, S. E. Rankin, F. Yang and Y.-T. Cheng, J. Phys. Chem. C, 2012, 116, 18669–18677. 30 D. W. Su, S. X. Dou and G. X. Wang, Nano Res., 2014, 7, 794–803. 31 Q. Gao, L. Yang, X. Lu, J. Mao, Y. Zhang, Y. Wu and Y. Tang, J. Mater. Chem., 2010, 20, 2807–2812. 32 L. Yang, L. Liu, Y. Zhu, X. Wang and Y. Wu, J. Mater. Chem., 2012, 22, 13148–13152. 33 L. C. Yang, Q. S. Gao, Y. H. Zhang, Y. Tang and Y. P. Wu, Electrochem. Commun., 2008, 10, 118–122. 34 L. Zhou, H. B. Wu, Z. Wang and X. W. Lou, ACS Appl. Mater. Interfaces, 2011, 3, 4853–4857. 35 L. Zeng, C. Zheng, C. Deng, X. Ding and M. Wei, ACS Appl. Mater. Interfaces, 2013, 5, 2182–2187. 36 X. Zhao, M. Cao, B. Liu, Y. Tian and C. Hu, J. Mater. Chem., 2012, 22, 13334–13340. 37 X. Fang, B. Guo, Y. Shi, B. Li, C. Hua, C. Yao, Y. Zhang, Y. S. Hu, Z. Wang, G. D. Stucky and L. Chen, Nanoscale, 2012, 4, 1541–1544. 38 L. Shen, E. Uchaker, X. Zhang and G. Cao, Adv. Mater., 2012, 24, 6502–6506. 39 L. Jin, L. Xu, C. Morein, C.-H. Chen, M. Lai, S. Dharmarathna, A. Dobley and S. L. Suib, Adv. Funct. Mater., 2010, 20, 3373–3382. 40 J. R. Szczech and S. Jin, Energy Environ. Sci., 2011, 4, 56–72.
This journal is © the Owner Societies 2014
View Article Online
Paper
45 X. Chen, X. Li, F. Ding, W. Xu, J. Xiao, Y. Cao, P. Meduri, J. Liu, G. L. Graff and J. G. Zhang, Nano Lett., 2012, 12, 4124–4130. 46 S. R. Gowda, V. Pushparaj, S. Herle, G. Girishkumar, J. G. Gordon, H. Gullapalli, X. Zhan, P. M. Ajayan and A. L. Reddy, Nano Lett., 2012, 12, 6060–6065. 47 L. Yang, L. Liu, Y. Zhu, X. Wang and Y. Wu, J. Mater. Chem., 2012, 22, 13148–13152.
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
41 T. Song, H. Han and U. Paik, Nano Res., 2014, 7, 491–501. 42 H. Wu, G. Zheng, N. Liu, T. J. Carney, Y. Yang and Y. Cui, Nano Lett., 2012, 12, 904–909. 43 W. Y. Lia, Y. Yang, Z. W. Seh, N. Liu and Y. Cui, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 7148–7153. 44 L. Xue, K. Fu, Y. Li, G. Xu, Y. Lu, S. Zhang, O. Toprakci and X. Zhang, Nano Energy, 2013, 2, 361–367.
PCCP
This journal is © the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 20570--20577 | 20577
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
PAPER
Cite this: Phys. Chem. Chem. Phys., 2014, 16, 20570
View Journal | View Issue
MoO2@carbon hollow microspheres with tunable interiors and improved lithium-ion battery anode properties† Xiaolin Liu, Wenxu Ji, Jiyuan Liang, Luming Peng and Wenhua Hou* MoO2 hollow microspheres with tunable inner space have been synthesized through a hydrothermal process using MoO3 microbelts instead of bulk MoO3 as the precursor. It is found that the reactant morphology has a great impact on the product morphology and the inner space can be tuned by changing the amount of NaOH aqueous solution. An interesting evolutional process from MoO3 microbelts through a rose-like intermediate to MoO2 hollow microspheres has been clearly observed, and thus the possible formation mechanism is revealed. One layer of amorphous carbon has been subsequently coated on the surface of MoO2 hollow microspheres through a simple hydrothermal approach followed by annealing in argon. As the anode material for lithium ion batteries, MoO2@C hollow microspheres manifest excellent lithium-storage properties, such as high capacity (677 mA h g 1)
Received 7th July 2014, Accepted 6th August 2014
and good cycling stability (negligible capacity fading even after 80 cycles). The significantly enhanced
DOI: 10.1039/c4cp02960g
performance of MoO2@C hollow microspheres can be attributed to its unique structures, such as nano-
www.rsc.org/pccp
scaled primary building blocks, carbon coating, hollow structure, and especially the synergy between the carbon coating and hollow structure.
Introduction Owing to their longer cycling life, better environmental benignity, higher energy and power densities, lower self-discharge rate and higher voltage compared with other secondary batteries, lithium ion batteries (LIBs) have been studied extensively and intensively.1–7 However, the current anode material, namely, graphite suffers from a low theoretical capacity (370 mA h g 1) and a large irreversible capacity.8 Therefore, it is an urgent task to search for an alternative material for graphite. Molybdenum dioxide (MoO2) has been considered as one of the promising anode materials for LIBs due to its low electrical resistivity, high chemical and thermal stability, and large theoretical capacity (838 mA h g 1).8–10 In addition, its Li+ insertion voltage is higher than that of the graphite-based anode, which also makes MoO2 attract considerable attention as a safe anode material. Unfortunately, MoO2 has two inadequacies that greatly hinder its practical application. First, a large volume variation occurs during the Li+ insertion/extraction into/from the MoO2 host, resulting in a huge loss in electrical connectivity with
Key Laboratory of Mesoscopic Chemistry of MOE, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, P. R. China. E-mail: [email protected] † Electronic supplementary information (ESI) available. See DOI: 10.1039/ c4cp02960g
20570 | Phys. Chem. Chem. Phys., 2014, 16, 20570--20577
the current collector, which is generally called electrode pulverization. Second, the slow kinetics in bulk MoO2 severely restricts the redox conversion reaction even if it is thermodynamically allowed, which will lead to a tremendous loss in its reversible capacity.11 Although these problems cannot be eliminated completely owing to the fundamental electrochemical process, several strategies have been proposed. The first one is the introduction of nanotechnology into the fabrication process. Due to the shorter lithium/electron diffusion paths, it is believed that nanostructured electrode materials better facilitate the transportation of Li+ into the electrode body center and therefore better benefit the redox conversion reaction.12 The second one is coating MoO2 with a layer of flexible and conductive substance, such as amorphous carbon or polyaniline, to counteract the pulverization effectively. At the same time, the electrical conductivity of MoO2 anode electrode can also be improved by the wrapping layer.12,13 The third one is better designing the desired structured MoO2 electrode, for example as a hollow structure or a one-dimensional structure.14–18 On the one hand, these kinds of materials can endure the volume variation effectively, thus helping to improve the lithium-storage properties. On the other hand, these materials always have a large specific surface area, which means that they can interact with Li+ ions more quickly and relieve the polarization. Overall, the best method is no doubt the combination of all the above mentioned three approaches: nanotechnology,
This journal is © the Owner Societies 2014
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
Paper
coating and designing desirable structures. Monodisperse porous LiFePO4@C microspheres,12 Fe2O3@polyaniline hollow spheres2 and SnO2@C hierarchical tube15 have been prepared, and all of them combined the aforementioned three design rationales for high-performance anode and thus exhibited excellent Li+ storage properties. However, the preparation processes involved were too cumbersome and even needed hard templates. To date, although MoO2 hollow microspheres have been reported, the synthesized MoO2 hollow microspheres have had a very broad size distribution and lacked regularity.19–21 Moreover, their lithium-storage properties have been unsatisfactory19 or remained unknown.20,21 In the present work, we have successfully fabricated MoO2@C hollow microspheres through a simple hydrothermal process. More importantly, the inner void space in the hollow structure could be tuned. It is noteworthy that the MoO2@C hollow microspheres synthesized simultaneously integrate the aforementioned three design rationales for high-performance anode materials, namely, nanostructured, carbon wrapped and unique hollow structured, and thus exhibit high specific capacity and Coulombic efficiency, excellent cycling stability and superior rate capability.
Experimental section 2.1
Materials
Bulk MoO3, (NH4)6Mo7O244H2O (AHM), EDTA–2Na (Na2H2Y), HCl and NaOH aqueous solution, glucose. 2.2
Synthesis of MoO2 agglomerated microspheres (MoO2-AMS)
MoO2 agglomerated microspheres were synthesized via the hydrothermal reaction of a mixture of 0.400 g bulk MoO3, 0.200 g EDTA–2Na and 45 mL distilled water in a 60 mL Teflonlined autoclave at 200 1C for 48 h.22 It is noteworthy that bulk MoO3 cannot be uniformly suspended in the aqueous solution. The final product was collected by centrifugation with distilled water several times, and dried at 70 1C for 24 h. The sample was denoted as MoO2-AMS. 2.3
Synthesis of MoO3 microbelts
MoO3 microbelts were synthesized by the hydrothermal reaction of a mixture solution of 0.980 g (NH4)6Mo7O244H2O (AHM), 14 mL HCl (3 M) and 31 mL distilled water in a 60 mL Teflon-lined autoclave at 180 1C for 20 h.23,24 HCl was used instead of HNO3 to avoid the residual HNO3 which would oxidize EDTA–2Na in the next procedure. The as-synthesized MoO3 microbelts were collected by centrifugation with distilled water several times, and dried at 70 1C for 24 h. 2.4
Synthesis of MoO2 hollow microspheres (MoO2-HMS)
MoO2 hollow microspheres were synthesized by using the as-prepared MoO3 microbelts and EDTA–2Na as the starting materials. In a typical synthesis, 0.400 g of MoO3 microbelts were uniformly dispersed in 25 mL distilled water with stirring for 10 min, forming the colloidal solution A. 0.200 g EDTA–2Na was dissolved in 20 mL distilled water, and then 0.15 mL NaOH
This journal is © the Owner Societies 2014
PCCP
aqueous solution (3 M), ca. 0.45 mmol, was dripped into the above solution, forming solution B. Solution B was then added into the well-stirred colloidal solution A. The mixture was allowed to stir for another 30 min at room temperature, and then the resulting mixture was transferred to and sealed in a 60 mL Teflon-lined autoclave, heated to 200 1C and maintained at this temperature for 48 h. The final product was collected by centrifugation with distilled water several times, and dried at 70 1C for 24 h. This sample was denoted as MoO2-HMS. 2.5
Synthesis of MoO2@C hollow microspheres (MoO2@C-HMS)
The as-prepared MoO2-HMS was coated with one layer of amorphous carbon through a simple hydrothermal process.12 In a typical synthesis, 0.05 g of as-prepared MoO2-HMS was dispersed into 35 mL of 0.075 M aqueous glucose solution by ultrasonication. The black suspension was transferred into a 60 mL Teflon-lined autoclave, then heated to 170 1C and maintained at this temperature for 8 h. The product was collected by centrifugation with distilled water and ethanol six times, and dried at 60 1C. Finally, the resulting black powder was carbonized at 600 1C for 4 h under an argon atmosphere to obtain the MoO2@C hollow microspheres. This sample was denoted as MoO2@C-HMS. 2.6
Characterization
The phase purity of the resulting products was identified by powder X-ray diffraction (XRD) on a Philip-X’Pert X-ray diffractometer with Cu Ka radiation (l = 1.5418 Å). The elemental analysis was carried out using a CHN-O-Rapid (Heraeus, Germany). The product morphology was observed using scanning electron microscopy (SEM JEOL JEM-6300F) and transmission electron microscopy (TEM JEOL JEM-200CX, operated at an accelerating voltage of 200 kV). Selected-area electron diffraction (SAED) and high-resolution TEM (HRTEM) spectra were examined under a JEM-2010F high-resolution transmission electron microscope. Thermogravimetric-differential thermal analysis (TG-DTA) was carried out on an SDT600 apparatus with a heating rate of 10 1C min 1 in an air atmosphere. 2.7
Electrochemical measurements
The electrochemical tests were conducted using a coin cell. An assembled coin cell consists of lithium foil as the counter electrode and a working electrode which is composed of 80% active material, 10% super P carbon black and 10% polyvinylidene fluoride (PVDF) and then made on copper foil. The mass coating of active material is ca. 2.00 mg cm 2. All cells were assembled in an argon-filled glove box (Mikarouna, Superstar 1220/750/900) by using 1 M LiPF6 solution in ethylene carbonate/ diethyl carbonate (EC : DEC = 1 : 1, v/v) as the electrolyte and a Celgard 2400 as the separator. The galvanostatic cycling test was performed on a battery test system (Land-CT2001A) at a constant current density of 100, 200, 500 and 1000 mA g 1 in the potential range from 0.005 to 3.0 V. Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (frequency range: 0.001–105 Hz) were carried out on an electrochemical workstation (CHI 660D).
Phys. Chem. Chem. Phys., 2014, 16, 20570--20577 | 20571
View Article Online
PCCP
Paper
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
Results and discussion To investigate the effect of product morphology on the Li+storage property, monoclinic MoO2 agglomerated microspheres (MoO2-AMS) were also prepared by using bulk MoO3 and EDTA–2Na as the starting materials. The powder XRD pattern of MoO2-AMS is shown in Fig. 1a, indicating that the bulk MoO3 has been entirely transformed into monoclinic MoO2 (JCPDS 86-0135). Fig. S1 (ESI†) shows the XRD pattern of MoO3 microbelts (the precursor of MoO2 hollow microspheres) and it can be indexed to pure orthorhombic MoO3 (JCPDS 05-0508). The unusually intense (020), (040) and (060) diffraction peaks demonstrate the apparent preferred orientation, which implies that the MoO3 precursor mainly grows along one dimension. The powder XRD pattern of the as-prepared MoO2 hollow microspheres (MoO2-HMS) is shown in Fig. 1a, and it can also be easily indexed to monoclinic MoO2, confirming its pure phase nature. The broadened diffraction peaks compared with those of MoO2-AMS demonstrates that MoO2-HMS is assembled from smaller building blocks. After carbon coating, the XRD pattern of MoO2@C-HMS shows no obvious change in comparison with that of MoO2-HMS, indicating that the crystal structure of MoO2 did not change and the carbon layer is amorphous. The SEM images of bulk MoO3 and MoO2-AMS are shown in Fig. S2a–c (ESI†). It is clear that MoO2-AMS is composed of severely agglomerated microspheres with a size of ca. 10 mm, and these agglomerated microspheres are assembled from stacked nanoplates with a diameter of ca. 200 nm. Fig. S2d and e (ESI†) shows the SEM and TEM images of the as-synthesized MoO3 microbelts with an aspect ratio of ca. 20, which verifies its one dimensional growth nature and is in accordance with the XRD results. When MoO3 microbelts instead of bulk MoO3 are used as the precursor while other reaction conditions remain unchanged, the product takes on the form of irregular porous and hollow quasi-microspheres (Fig. S2f, ESI†). Furthermore, when 0.15 mL NaOH aqueous solution (3 M) was dripped into the EDTA–2Na
Fig. 1 (a) XRD patterns of MoO2-AMS, MoO2-HMS and MoO2@C-HMS; (b) SEM image of MoO2-HMS; (c) SEM image of MoO2@C-HMS; (d) magnified SEM image of MoO2@C-HMS.
20572 | Phys. Chem. Chem. Phys., 2014, 16, 20570--20577
Fig. 2 (a–c) TEM images of MoO2@C-HMS with different magnifications; (d) HRTEM image recorded within the red box of (b), its inset is the corresponding SAED pattern; (e) TEM image of MoO2@C-HMS when 0.25 mL NaOH aqueous solution (3 M) was dripped into the EDTA–2Na stock aqueous solution; (f) TG-DSC curves of MoO2@C-HMS conducted under air flow.
stock aqueous solution, a considerable change in the product morphology occurred. As illustrated in Fig. 1b, the resulted product is composed of microspheres with a size in the range of 0.5–2.0 mm. After carbon coating, the product morphology shows no obvious changes (Fig. 1c). Upon closer observation, the microspheres are found to be composed of many secondary units (Fig. 1d). The morphology and structure of MoO2@C-HMS were further studied by TEM characterizations. As shown in Fig. 2a and b, the hollow structure can be clearly confirmed, and may effectively accommodate the volume change that occurs during the Li+ insertion/extraction.25–30 Additionally, the fringes of a hollow microsphere (see the magnified TEM images in Fig. S3, ESI†), clearly show these microspheres are self-assembled from nanoslabs. As shown in Fig. 2c, one layer of amorphous carbon with a thickness of ca. 6 nm was uniformly coated on the surface of MoO2 hollow microspheres. Fig. 2d shows the HRTEM image of a nanoslab (recorded within the red square of Fig. 2b), and the clear lattice fringes demonstrate that the sample is well-crystallized. The spacing of adjacent lattice fringes was determined to be 0.2408 nm, which is in good agreement with the interplanar spacing of the (200) planes of MoO2. The bright diffraction spots in the inset SAED pattern also indicate the single crystalline nature of the nanoslabs. It is worth mentioning that the inner void space of the hollow structure could be easily controlled through changing the addition amount of NaOH aqueous solution. Fig. 2e shows the TEM image of MoO2@C-HMS when 0.25 mL NaOH aqueous
This journal is © the Owner Societies 2014
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
Paper
solution (3 M) was dripped into the EDTA–2Na stock aqueous solution. By comparing Fig. 2e with Fig. 2a, one may find that the inner void space is apparently diminished. In addition, the product size is also decreased from ca. 1.3 mm to ca. 0.8 mm. Fig. 2f shows the thermogravimetric-differential scanning calorimetry (TG-DSC) curves of MoO2@C-HMS. The slight weight change that occurs between 30 and 300 1C can be attributed to the evaporation of the adsorbed water. In the next step, the weight change between 350 and 600 1C is related to the oxidation of MoO2 and combustion of coated carbon, and correspondingly a large exothermic peak emerges in the DSC curve. Finally, the endothermic peak at ca. 780 1C is caused by the sublimation of MoO3. Since the theoretical value of the mass increase from MoO2 to MoO3 is 12.5 wt%, the content of the coated carbon was accordingly estimated to be ca. 14.67 wt% in the MoO2@C-HMS nanocomposite. Subsequent CHN analysis also demonstrated that the carbon content in MoO2@C-HMS was about 14.47 wt%, which is in good agreement with the results of TG-DSC. In order to figure out the formation mechanism of MoO2 hollow microspheres, the time-dependent experiments were carried out. The XRD pattern of the intermediate after hydrothermal treatment for 12 h is shown in Fig. 3, and its phase can be indexed as a mixture of orthorhombic MoO3 (JCPDS card 05-0508) and orthorhombic [Na(H2O)5]0.25 MoO3 (JCPDS card 39-1112). The corresponding SEM images are shown in Fig. 4a–d. It mainly consists of MoO3 microbelts (see the red ellipse of Fig. 4b) and a very beautiful rose-like intermediate. Combined with the XRD results, it can be deduced that the rose-like intermediate is probably [Na(H2O)5]0.25MoO3. After a careful observation of Fig. 4c (especially within the red rectangle of Fig. 4d), it can be found that the petals of microroses are produced by the merging of MoO3 microbelts. The XRD pattern of the intermediate after hydrothermal treatment for 15 h is shown in Fig. 3, and its phase can be mainly attributed to the mixture of
Fig. 3 XRD patterns of the intermediates and product after hydrothermal treatment for 12 h, 15 h, 30 h and 48 h, respectively. The standard data for orthorhombic [Na(H2O)5]0.25MoO3 (red, JCPDS card 39-1112), orthorhombic MoO3 (black, JCPDS card 05-0508) and monoclinic MoO2 (blue, JCPDS card 86-0135) are also presented at the bottom for comparison.
This journal is © the Owner Societies 2014
PCCP
Fig. 4 SEM images of intermediates obtained after hydrothermal treatment for 12 h (a–d), 15 h (e–g) and 30 h (h). The right corner of (h) is the corresponding TEM image.
orthorhombic [Na(H2O)5]0.25MoO3 (JCPDS card 39-1112) and monoclinic MoO2 (JCPDS card 86-0135). It is noted that the orthorhombic MoO3 phase has disappeared. The corresponding SEM images are shown in Fig. 4e–g. It is mainly composed of microroses and many small spheres (see Fig. 4e). Combined with the XRD results, it can be deduced that the small spheres are monoclinic MoO2. Within the red circle of Fig. 4f, an incomplete hollow sphere can be seen among the petals. It implies that the formation of MoO2 hollow spheres is closely related to the petals. More importantly, the microrose intermediate after hydrothermal treatment for 15 h has surfaces with holes and very blurred edges (see Fig. 4f and g), while the microrose intermediate after hydrothermal treatment for 12 h has very smooth surfaces and edges. In the light of the above discussion, it can be deduced that the petals of microroses were first reduced to MoO2 nanoslabs and then, these nanoslabs self-assembled into MoO2 hollow spheres. The XRD pattern of the intermediate after hydrothermal treatment for 30 h is shown in Fig. 3, and the product phase can be mainly indexed to monoclinic MoO2. The phase of orthorhombic [Na(H2O)5]0.25MoO3 nearly disappears, and only an ultra-weak peak (2y = 7.61) can be seen. Its SEM image is exhibited in Fig. 4h, and the product is composed of many microspheres with
Phys. Chem. Chem. Phys., 2014, 16, 20570--20577 | 20573
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
PCCP
Scheme 1
Paper
An illustration of the formation process of MoO2-HMS.
a size of ca. 1 mm, and its corresponding TEM image in the right corner manifests the hollow structure of these microspheres, demonstrating that the MoO3 microbelts have been entirely transformed into MoO2 hollow microspheres. Finally, based on the experimental facts and above discussions, the formation mechanism of MoO2 hollow microspheres could be proposed as follows. First, MoO3 microbelts merged into [Na(H2O)5]0.25MoO3 microroses at high temperature and pressure in the presence of Na+ (the ‘‘templating ion’’), since MoO3 often acted as a host for intercalation chemistry.3 Second, the petals of [Na(H2O)5]0.25MoO3 microroses were reduced to MoO2 nanoslabs by H2Y2 (Na2H2Y denotes EDTA–2Na) at high temperature and pressure. Finally, MoO2 nanoslabs tended to self-assemble into a spherical shape to decrease the surface energy. The possible formation process is illustrated in Scheme 1. Nevertheless, how did these MoO2 nanoslabs self-assemble into hollow microspheres? It can be explained as follows: first, a large number of Na+ ions would be adsorbed on the surface of MoO2 nanoslabs, and the numerous nanoslabs would agglomerate into spherical shape to decrease the surface energy. It is noteworthy that the volumetric charge density, and thus the repulsion, is larger on the inside of the sphere than that on the outside. Therefore, many microspheres were cracked and only incomplete spheres were formed (see Fig. S2f, ESI†). When 0.15 mL NaOH aqueous solution was dripped into the EDTA–2Na stock aqueous solution, the pH value and OH concentration in the reaction system increased. It is noted that OH can also adsorb on the surface of the nanoslabs. The adsorbed OH lowered the repulsion between the nanoslabs through the attraction between Na+ and OH , and then completely hollow spheres were formed (see Fig. 1b). As the amount of NaOH aqueous solution was further increased, the repulsion between nanoslabs was further lowered, and then the nanoslabs were piled up more closely, as evidenced by the decreased product size. Consequently, the inner void space was diminished (see Fig. 2e). In short, EDTA–2Na has played two roles in this synthesis, i.e., as a reducing agent and a morphology modulator. The possible formation mechanism is illustrated in Scheme S1 (ESI†). Fig. 5a shows the cyclic voltammograms (CV) of MoO2@C-HMS electrode at a scanning rate of 0.2 mV s 1 between 0.005 and 3.0 V.
20574 | Phys. Chem. Chem. Phys., 2014, 16, 20570--20577
Fig. 5 (a) Cyclic voltammograms of MoO2@C-HMS at a scanning rate of 0.2 mV s 1; (b) galvanostatic discharge–charge curves of MoO2@C-HMS at a current rate of 100 mA g 1; (c) cycling performance of MoO2-AMS, MoO2-HMS and MoO2@C-HMS at a current density of 100 mA g 1; (d) rate capability of MoO2@C-HMS at different current densities.
There is an apparent peak dislocation between the first cycle and the following cycles. From the second cycle onward, the CV profiles are almost fully-overlapped, indicating the high reversibility of Li+ insertion and extraction. Two pronounced redox couples at 1.52/1.72 V and 1.25/1.48 V in the second and fifth cycles are associated with the reversible phase transitions of LixMoO2 from monoclinic to orthorhombic and then from orthorhombic back to monoclinic.31–34 Fig. 5b displays the galvanostatic charge and discharge curves for MoO2@C-HMS electrode at a current density of 100 mA g 1 in the voltage range of 0.005 to 3.0 V (vs. Li+/Li). It can be found that the initial charge and discharge capacities are 376.4 and 478.9 mA h g 1, respectively. An irreversible capacity loss of 102.5 mA h g 1 is mainly attributed to some irreversible processes such as the formation of a solid electrolyte interface (SEI), trapping of some Li+ in the MoO2 lattice and partial decomposition of the electrolyte,11,35,36 leading to a relatively low initial Coulombic efficiency of 78.6%. In the first discharge–charge cycle, there are two obvious discharge voltage
This journal is © the Owner Societies 2014
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
Paper
PCCP
plateaus at 1.54 and 1.28 V and two charge ones at 1.41 and 1.71 V, being consistent with the CV results. These observed voltage plateaus are attributed to the reversible phase transitions of LixMoO2 from monoclinic to orthorhombic and then back to monoclinic.37 It can also be observed in Fig. 5b that the reversible capacity increases gradually upon cycling until the 30th cycle. Fig. 5c shows the cycling stability of MoO2-AMS, MoO2-HMS and MoO2@C-HMS electrodes at a current density of 100 mA h g 1 in the potential window of 0.005 and 3.0 V. The MoO2-AMS electrode delivered a maximum capacity of 539.1 mA h g 1 at the 15th cycle, and the reversible capacity drastically faded to 99.6 mA h g 1 after 50 cycles. By comparison, the MoO2-HMS electrode delivered a maximum capacity of 675.1 mA h g 1 at the 15th cycle, and the reversible capacity faded to 358.7 mA h g 1 after 50 cycles. The obviously improved reversible capacity and cycling stability of MoO2-HMS could be ascribed to its nanosized building blocks and hollow structure.11,38–41 First, the nanosized building blocks provide the shorter Li+/electron diffusion path, and better facilitate the transportation of Li+ into the electrode body center, and hence the reversible capacity is improved. Second, the hollow structure could better sustain the volume variation during the lithiation and delithiation, and thus the cycling performance is enhanced.12 It is noteworthy that the MoO2@C-HMS electrode could deliver a capacity of 677.5 mA h g 1 at the 30th cycle, and a capacity of 677.4 mA h g 1 was retained even after 80 cycles with a retention close to 100%. In addition, its Coulombic efficiency has been higher than 97% since the fifth cycle (see Fig. S4, ESI†). There is no doubt that MoO2@C-HMS has a significantly improved cycling stability over MoO2-HMS. The remarkably improved cycling stability of MoO2@C-HMS can be ascribed to the synergistic effect between the porous hollow structure and carbon coating, as suggested by the Cui group.42,43 As shown in Scheme S2 (ESI†), the inner void space renders MoO2 to expand inwards instead of outwards during Li+ insertion, and thus the structure can be restored to its original state upon Li+ extraction and also the carbon layer will not be easily destroyed during lithiation and delithiation. In contrast, if MoO2 were to be a solid sphere, it had to expand outwards since there was no inner void space. As a result, the carbon wrapping would be cracked during Li+ insertion and extraction after several cycles. It is also noteworthy that the reversible capacity of MoO2@C-HMS increases drastically in the initial cycles, which is in agreement with the results of Fig. 5b. It can be attributed to the following two reasons. First of all, the process of Li+ insertion/extraction into/from the MoO2 host can be divided into the following two steps:35 xLi+ + xe + MoO2 2 LixMoO2 (0 o x o 0.98)
(1)
Li0.98MoO2 + 3.02Li+ + 3.02e 2 2Li2O + Mo
(2)
The second (latter) step is also named the conversion reaction of MoO2. Fig. S5 (ESI†) shows the discharge curves of MoO2@C-HMS at the 6th, 10th, 14th, 18th, 22th and 26th cycle, respectively. The capacity below 0.6 V in the discharge curve is
This journal is © the Owner Societies 2014
caused by the conversion reaction as suggested by many previous studies.10 It can be clearly seen that the released capacity during this stage (conversion reaction stage), increases gradually with the cycles. In other words, the conversion reaction of MoO2 becomes more complete with the cycles. This is also called the activation process of MoO2. We also note that the released maximum capacity of MoO2-HMS is at the 13th cycle, while that of MoO2@C-HMS is at the 30th cycle. The sole difference between MoO2-HMS and MoO2@C-HMS is the carbon coating. There is no doubt that the wetting of electrolyte into the electrode material will be delayed by the wrapping layer out of MoO2-HMS. It is also necessary to note that Li+ is dissolved in the electrolyte, and Li+ only moves in the electrolyte. Therefore, only after the complete wetting of electrolyte into the electrode material can Li+ freely pass through the whole electrode material and fully release its Li+ storage capacity. Similar experimental phenomena were also observed by other groups.6 The rate capability of MoO2@C-HMS electrode is presented in Fig. 5d. A steady capacity of ca. 455 mA h g 1 can be obtained even at a current density of 1 A g 1. Moreover, the capacity can reach ca. 680 mA h g 1 when the current density is decreased to 100 mA g 1 after 40 cycles. The superior rate performance of MoO2@C-HMS electrode is closely related to its porous structure and carbon wrapping.44–46 The porous structure could make Li+ penetrate the inner space and provide more active sites for Li+ intercalation. As shown in Fig. S6 (ESI†), the carbon wrapping makes MoO2@C-HMS electrode possess relatively low charge-transfer impedance. Therefore, the electrons transfer more quickly after carbon coating, being conducive to the rate capability. When compared with the previously reported results on similar systems,9,10,31,34,47 the test results of our lithium storage performance is rather attractive if one takes into account the testing conditions used in our work (e.g. the prepared MoO2@C-HMS has not been hybridized with graphene or carbon nanotubes, PVDF instead of sodium alginate was used as the binder, and a mass coating of active material was ca. 2.00 mg cm 2).
Conclusions In summary, MoO2@C hollow microspheres have been successfully synthesized through employment of MoO3 microbelts instead of bulk MoO3 as the precursor. The inner void space of the resulted hollow structure could be easily tuned by changing the addition amount of NaOH aqueous solution. The evolutional process from MoO3 microbelts through a rose-like intermediate to MoO2 hollow microspheres has been disclosed. The as-prepared MoO2@C hollow microspheres simultaneously satisfy three desirable characters for highperformance anode materials, i.e., nanoscaled primary building blocks, highly-conductive coating layer of carbon and hollow structure. Owing to these unique structure characters and the synergy among them, the prepared MoO2@C porous hollow
Phys. Chem. Chem. Phys., 2014, 16, 20570--20577 | 20575
View Article Online
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
PCCP
microspheres exhibited excellent Li+-battery anode properties. The nanoscaled primary building blocks contributed to its high reversible capacity; while carbon coating greatly lowered charge-transfer impedance and thus improved its rate capability. Moreover, carbon-encapsulated MoO2 hollow microspheres effectively made the anode material expand inwards instead of outwards during the Li+ insertion, rendering the material to better sustain the volume variation and thus guaranteeing its long cycling stability. The present work provides a new method for the fabrication of hollow-structured materials. In addition, MoO2 with this unique structure may also find potential application in the catalysis field.
Acknowledgements This project was supported by the National Natural Science Foundation of China (Grant No. 21073084), the Specialized Research Fund for the Doctoral Program of Higher Education (Grant No. 20130091110010), the Natural Science Foundation of Jiangsu Province (Grant No. BK2011438), the National Science Fund for Talent Training in Basic Science (No. J1103310), the National Basic Research (973) Program of China (Grant No. 2009CB623504), and the Modern Analysis Center of Nanjing University.
Notes and references 1 J. B. Goodenough and K. S. Park, J. Am. Chem. Soc., 2013, 135, 1167–1176. 2 J. M. Jeong, B. G. Choi, S. C. Lee, K. G. Lee, S. J. Chang, Y. K. Han, Y. B. Lee, H. U. Lee, S. Kwon, G. Lee, C. S. Lee and Y. S. Huh, Adv. Mater., 2013, 25, 6250–6255. 3 M. S. Whittingham, J. D. Guo, R. J. Chen, T. Chirayil and G. Janauer, Solid State Ionics, 1995, 75, 257–268. 4 X. K. Guo, Q. Fan, L. Yu and J. Y. Liang, J. Mater. Chem. A, 2013, 1, 11534–11538. 5 M. J. Armstrong, C. Dwyer, W. J. Macklin and J. D. Holmes, Nano Res., 2014, 7, 1–62. 6 H. Wu, G. Yu, L. Pan, N. Liu, M. T. McDowell, Z. Bao and Y. Cui, Nat. Commun., 2013, 4, 1943–1948. 7 H. L. Wang, L. F. Cui, Y. Yang, Y. Cui and H. J. Dai, J. Am. Chem. Soc., 2010, 132, 13978–13980. 8 A. Bhaskar, M. Deepa and T. N. Rao, ACS Appl. Mater. Interfaces, 2013, 5, 2555–2566. 9 Y. F. Shi, B. G. Guo and L. Q. Chen, Nano Lett., 2009, 9, 4215–4220. 10 B. Guo, X. Fang, B. Li, Y. Shi, C. Ouyang, Y.-S. Hu, Z. Wang, G. D. Stucky and L. Chen, Chem. Mater., 2012, 24, 457–463. 11 Z. Wang, J. S. Chen, T. Zhu, S. Madhavi and X. W. Lou, Chem. Commun., 2010, 46, 6906–6908. 12 L. Zhang, G. Zhang, H. B. Wu, L. Yu and X. W. Lou, Adv. Mater., 2013, 25, 2589–2593. 13 L. Yang, S. Wang, J. Mao, J. Deng, Q. Gao, Y. Tang and O. G. Schmidt, Adv. Mater., 2013, 25, 1180–1184.
20576 | Phys. Chem. Chem. Phys., 2014, 16, 20570--20577
Paper
14 C. Y. Nan, J. Lu, L. H. Li, L. L. Li, Q. Peng and Y. D. Li, Nano Res., 2013, 6, 469–477. 15 C. Sun, S. Rajasekhara, J. B. Goodenough and F. Zhou, J. Am. Chem. Soc., 2011, 133, 2132–2135. 16 Y. G. Li, B. Tan and Y. Y. Wu, J. Am. Chem. Soc., 2006, 128, 14258–14259. 17 B. Song, S. W. Liu and J. K. Jian, J. Power Sources, 2008, 180, 869–874. 18 Y. G. Li, B. Tan and Y. Y. Wu, Chem. Mater., 2008, 20, 567–576. 19 Y. Z. Lei, J. C. Hu, H. W. Liu and J. L. Li, Mater. Lett., 2012, 68, 82–85. 20 Z. Q. Wang, H. F. Wang, C. Yang and J. H. Wu, Mater. Lett., 2010, 64, 2170–2172. 21 H. X. Zhang, Y. F. Li, Z. S. Hong and M. D. Wei, Mater. Lett., 2012, 79, 148–151. 22 R. Q. Song, A. W. Xu, B. Deng and Y. P. Fang, J. Phys. Chem. B, 2005, 109, 22758–22766. 23 X. W. Lou and H. C. Zeng, Chem. Mater., 2002, 14, 4781–4789. 24 Y. Chen, C. Lu, L. Xu, Y. Ma, W. Hou and J.-J. Zhu, CrystEngComm, 2010, 12, 3740–3747. 25 B. K. Guo, M. F. Chi, X. G. Sun and S. Dai, J. Power Sources, 2012, 205, 495–499. 26 W. Wang, Q. Sa, J. Chen, Y. Wang, H. Jung and Y. Yin, ACS Appl. Mater. Interfaces, 2013, 5, 6478–6483. 27 A. Vu, Y. Qian and A. Stein, Adv. Energy Mater., 2012, 2, 1056–1085. 28 S. A. Klankowski, R. A. Rojeski, B. A. Cruden, J. Liu, J. Wu and J. Li, J. Mater. Chem. A, 2013, 1, 1055–1064. 29 Q. L. Wu, J. Li, R. D. Deshpande, N. Subramanian, S. E. Rankin, F. Yang and Y.-T. Cheng, J. Phys. Chem. C, 2012, 116, 18669–18677. 30 D. W. Su, S. X. Dou and G. X. Wang, Nano Res., 2014, 7, 794–803. 31 Q. Gao, L. Yang, X. Lu, J. Mao, Y. Zhang, Y. Wu and Y. Tang, J. Mater. Chem., 2010, 20, 2807–2812. 32 L. Yang, L. Liu, Y. Zhu, X. Wang and Y. Wu, J. Mater. Chem., 2012, 22, 13148–13152. 33 L. C. Yang, Q. S. Gao, Y. H. Zhang, Y. Tang and Y. P. Wu, Electrochem. Commun., 2008, 10, 118–122. 34 L. Zhou, H. B. Wu, Z. Wang and X. W. Lou, ACS Appl. Mater. Interfaces, 2011, 3, 4853–4857. 35 L. Zeng, C. Zheng, C. Deng, X. Ding and M. Wei, ACS Appl. Mater. Interfaces, 2013, 5, 2182–2187. 36 X. Zhao, M. Cao, B. Liu, Y. Tian and C. Hu, J. Mater. Chem., 2012, 22, 13334–13340. 37 X. Fang, B. Guo, Y. Shi, B. Li, C. Hua, C. Yao, Y. Zhang, Y. S. Hu, Z. Wang, G. D. Stucky and L. Chen, Nanoscale, 2012, 4, 1541–1544. 38 L. Shen, E. Uchaker, X. Zhang and G. Cao, Adv. Mater., 2012, 24, 6502–6506. 39 L. Jin, L. Xu, C. Morein, C.-H. Chen, M. Lai, S. Dharmarathna, A. Dobley and S. L. Suib, Adv. Funct. Mater., 2010, 20, 3373–3382. 40 J. R. Szczech and S. Jin, Energy Environ. Sci., 2011, 4, 56–72.
This journal is © the Owner Societies 2014
View Article Online
Paper
45 X. Chen, X. Li, F. Ding, W. Xu, J. Xiao, Y. Cao, P. Meduri, J. Liu, G. L. Graff and J. G. Zhang, Nano Lett., 2012, 12, 4124–4130. 46 S. R. Gowda, V. Pushparaj, S. Herle, G. Girishkumar, J. G. Gordon, H. Gullapalli, X. Zhan, P. M. Ajayan and A. L. Reddy, Nano Lett., 2012, 12, 6060–6065. 47 L. Yang, L. Liu, Y. Zhu, X. Wang and Y. Wu, J. Mater. Chem., 2012, 22, 13148–13152.
Published on 12 August 2014. Downloaded by University of Chicago on 11/10/2014 09:51:03.
41 T. Song, H. Han and U. Paik, Nano Res., 2014, 7, 491–501. 42 H. Wu, G. Zheng, N. Liu, T. J. Carney, Y. Yang and Y. Cui, Nano Lett., 2012, 12, 904–909. 43 W. Y. Lia, Y. Yang, Z. W. Seh, N. Liu and Y. Cui, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 7148–7153. 44 L. Xue, K. Fu, Y. Li, G. Xu, Y. Lu, S. Zhang, O. Toprakci and X. Zhang, Nano Energy, 2013, 2, 361–367.
PCCP
This journal is © the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 20570--20577 | 20577
