Hanshuo Wu,
Yuxin Wu,
Xi Chen,
Yongjin Ma,
Mingquan Xu,
Weifeng Wei,
Jun Pan* and
Xiang Xiong*
State Key Laboratory of Powder Metallurgy, Central South University, Changsha 410083, P. R. China. E-mail: jun.pan@csu.edu.cn; xiongx@csu.edu.cn
First published on 22nd February 2016
A heterostructure of MoSe2 nanosheets decorated on C/TiO2 nanobelts was successfully fabricated through a facile hydrothermal process with the assistance of glucose. Ultrathin MoSe2 nanosheets fully covered on TiO2 nanobelts were coupled with carbon components as strong skeletons, which possess high contact surface area, efficient electronic transport frameworks and sufficient void spaces. As expected, the MoSe2@C/TiO2 heterostructures exhibit remarkable cycling stability (discharge capacity of 987.4 mA h g−1 after 100 cycles at a current density of 500 mA g−1) and rate capability (retained capacity of 860 mA h g−1 even at a current rate of 3000 mA g−1) as anode materials in lithium-ion batteries. The excellent electrochemical performance can be attributed to the structural advantages of MoSe2@C/TiO2 heterostructures, enabling great potential as anode materials in lithium-ion batteries.
As one of the representative TMDs materials, MoSe2 reveal relatively high theoretical specific capacity and have been appreciated as good candidate anode materials for energy storage devices.2–4 However, MoSe2 nanostructure anodes still expose some inherent problems impeding its practical applications. Accompanying the process of repeated discharge and charge, large volume changes will cause severe structure collapse. Meanwhile particle aggregation can form unstable solid electrolyte interface (SEI) film, which results in fast cycle capacity fading, low coulombic efficiency and limited rate capacity.5,6 In order to avoid and resolve these obstacles, an effective strategy is to develop synthesis of MoSe2 hybrid nanostructures with various substrate matrices. Most of the previous researches focused on MoSe2/carbon nanocomposites, including carbon spheres,7 graphene,8 carbon cloth9 and organic carbon.10 In these nanocomposites, the improved electrochemical performance demonstrates the synergistic effect of MoSe2 and carbon matrices. However, the excessive interface between carbon matrices and electrolytes leads to considerable side reactions and forms a thick SEI film on the carbon matrices.11 In this respect, optimized framework materials for MoSe2 based heterostructures with effective synergy are highly desirable and imperative.
Alternatively, TiO2 has been has been widely used as a backbone material for hybrid systems owing to its structural integrity during the charge/discharge process (low volume change of less than 4%) and chemical stability.12–14 In addition, TiO2 anode exhibits a relatively high lithium insertion/extraction voltage at about 1.7 V, which can efficiently avoid the formation of SEI layers and thus improve safety compared with carbon matrices.
In addition, recently some hybrid systems based on 1D TiO2 nanostructures have been adopted to overcome the demerits of the individual materials, thus improving the anode performance in LIBs.15–18
Based on the strategies mentioned above, we rationally designed and successfully synthesized new MoSe2 nanosheets@C/TiO2 nanobelts hybrid nanostructures by a facile glucose-assisted hydrothermal process, and used them as electrode materials for lithium-ion batteries. Here, the stable skeletons of TiO2 nanobelts showed good structural stability as expected and further buffered severely volume change of MoSe2 nanosheets by providing sufficient void spaces. The low cycle life of MoSe2 nanosheets and the low specific capacity of TiO2 nanobelts could be simultaneously improved by the synergistic effect between MoSe2 and TiO2. Secondly, what is particularly worth mentioning is that the addition of glucose skillfully achieved killing two birds with one stone: glucose could not only play a critical role during synthesis process to promote uniform cover of few-layered MoSe2 nanosheets on TiO2 skeletons, but also couple with TiO2 nanobelts to provide fast and shortened pathways for electron transportation. Moreover, comparing with agglomerated MoSe2 nanoflowers, MoSe2 nanosheets could uniformly stretch with the robust mechanical support of the stable skeletons, resulting high specific surface area with electrolyte. The advisable combination of these two components takes advantages of each other and reveals favorable synergistic effect. As a result, such composites synergistically enhance the effective utilization of MoSe2, buffer the volume variation and achieve good cycling stability.
![]() | ||
Fig. 1 XRD patterns of MoSe2@C/TiO2 heterostructures, pure MoSe2 nanosheets and bare TiO2 nanobelts. |
In order to study the morphologies and microstructures of all the samples, field emission scanning electron microscopy (FSEM) and field emission transmission electron microscopy (FTEM) were performed. Fig. 2a presents that monodisperse TiO2 nanobelts with diameters of 100–300 nm are obtained. The low magnification TEM image shown in the inset of Fig. 2a reveals rough surface of pure TiO2 nanobelts. As can be seen in Fig. 2b, the morphology of MoSe2 can be described as a flower-like sphere with an average diameter of 100 nm. Fig. 2c and d show the SEM and TEM images of the as-synthesized MoSe2@C/TiO2 hybrid nanostructures after hydrothermal and annealed process. It reveals TiO2 nanobelts are uniformly covered by MoSe2 nanosheets with the assistant of glucose. The enlarged image is shown in Fig. 2e. It can be observed that the MoSe2 nanosheets are flexible, and curly, indicative of their ultrathin 2D nature. Furthermore, the rings in SAED (Fig. 2f) arise from the stacking of MoSe2 nanosheets with different crystallographic orientations, consistent with XRD results. As illustrated in Fig. 2g, the scattering diffraction spots in inner circle can be ascribed to interplanar spacing of anatase TiO2 (101), which is also regarded as the most thermodynamically stable crystal facet of anatase TiO2.19 The concentric rings can be assigned as diffraction from (100), (103) and (110) planes of hexagonal phase MoSe2.
From HRTEM image of composites in Fig. 3a, it can be seen that the interplanar distance of 0.65 nm is apparently larger than the d-spacing of (002) planes of hexagonal phase MoSe2. It is expected that MoSe2 nanosheets with an enlarged interlayer distance can accommodate more Li+. Meanwhile, the interlayered spacing of approximate 0.23 nm and 0.19 nm are observed, which can be perfectly indexed as (103) and (105) plane of the hexagonal phase MoSe2, respectively. The top view high-resolution TEM image of heterostructures in Fig. 3b indicates that most of the shells in the hybrid nanostructures consist of 3–7 layers of MoSe2 nanosheets. Furthermore, to investigate the effect of glucose on the morphology of heterostructures, one control sample was synthesized according to the same conditions except for the addition of glucose. Fig. S1† shows that TiO2 nanobelts are covered by MoSe2 thick flakes, simultaneously a large amount of them tend to aggregate into dispersive nanoparticles. Based on these results, it can be inferred that glucose is beneficial to suppress clustering of MoSe2 nanosheets, maintain ultrathin structure superiority and act as a binder to promote uniform cover of few-layered MoSe2 nanosheets on TiO2 nanobelts.11,20
The detailed local elemental mapping (Fig. 4) demonstrates that Mo and Se elements are homogeneously dispersed in the entire sample region. Remarkably, Ti elements orderly assemble to belts acting as a backbone, and the majority of O elements distribute at center belts in consistent with Ti. Meanwhile, it is deserved to be mentioned that C elements exhibit uniformly distribution in accordance with Mo and Se. In combination with synthesis procedure, it can be inferred that this is likely to come from residual glucose carbon groups after hydrothermal process.11 The energy-dispersive spectrometry (EDS) analysis of the prepared MoSe2@C/TiO2 is presented in Fig. S2.† It can be seen that the composite mainly contains Mo, Se, Ti, O and C. The atomic ratios of Se:
Mo and O
:
Ti is 1.92 and 1.99, respectively, which is very close to the theoretical value of MoSe2 and TiO2. Furthermore, the molar ratio of Ti
:
Mo is about 2.18.
In addition, TGA measurement was employed under air flow to further determine the loading amount of single component in the composite. As shown in Fig. 5, the TiO2 nanobelts show a slight weight change about 2.49%, while both pure MoSe2 and MoSe2@C/TiO2 have notable weight loss. The weight increase in the range of 300–400 °C can be attributed to the formation of SeO2 during the oxidation of MoSe2. As temperature increase, significantly weight loss is associated with the gasification of SeO2, the further oxidation of MoSe2 and redundant carbon groups, following the final products MoO3. So, approximate amount of MoSe2 can be calculated from TGA analysis by assuming that the whole weight loss is due to the oxidation of TiO2 and MoSe2 to MoO3. X represents the mass fraction of MoSe2 content in the heterostructures in the following equation.
X × 28.68% + (1 − X) × 2.49% = 20.53% | (1) |
X = 68.9% |
We also carried out X-ray photoelectron spectroscopy (XPS) to investigate the chemical states of Mo, Se, Ti, O and C in the MoSe2@C/TiO2 heterostructures and element composition on the surface (Fig. 6). Fig. 6a shows the C 1s spectrum, which can be fitted into three peaks: the carbon in C–C and C–OH appeared at 284.8 and 286.7 eV respectively, and the peak at 282.5 eV is assigned to carbon in C–Mo.10 In the high resolution Mo 3d spectrum of the composites (Fig. 6b), four separate peaks are presented indicating the Mo atoms are in two different chemical states. Two peaks at 228.8 and 232.0 eV can be ascribed to the expected Mo (+4) 3d5/2 and 3d3/2 respectively,21 while the other two peaks at 233.0 and 236.0 eV can be attributed to Mo (+6) 3d5/2 and 3d3/2 respectively, which are likely due to the existence of small amount of MoO42− that is not completely reduced.19 Meanwhile, in Fig. 6c the binding energies of Se 3d5/2 at 54.3 eV and Se 3d3/2 at 55.0 eV indicated the (−2) oxidation chemical state of Se. The 2p peak of Ti element exhibits two peaks in Fig. 6d at 459.0 and 464.7 eV, corresponding to the Ti 2p3/2 and Ti 2p1/2 respectively, confirming the existence of Ti (+4) state. The XPS results further prove the co-existence of MoSe2 and TiO2 in the heterostructures.
![]() | ||
Fig. 6 XPS spectra comparison of MoSe2@C/TiO2 heterostructures, pure MoSe2 nanosheets and bare TiO2 nanobelts: (a) C 1s peaks, (b) Mo 3d peaks, (c) Se 3d peaks and (d) Ti 2p peaks. |
By comparison, in pure MoSe2 nanosheets, Mo (+4) 3d5/2, 3d3/2, Se (−2) 3d5/2 and Se 3d3/2 peaks are located at 229.04, 232.24, 54.64 and 55.34 eV, respectively. After TiO2 nanobelts were decorated on the MoSe2 nanosheets, the binding energy of Mo 3d and Se 3d of MoSe2@C/TiO2 shifted about 0.24 and 0.34 eV to the lower energy direction. Moreover in pure TiO2 nanobelts, Ti (+4) 2p3/2 and 2p1/2 are located at 458.2 and 463.9 eV, respectively, while after loaded MoSe2 nanosheets, the binding energy shifted about 0.80 eV to higher energy direction. Therefore, the collective shifts of these peaks can be interpreted as mutual effect of varied electron environment, which can further confirmed the successful synthesis of heterostructures. In addition, the detailed compositional analysis reveals that surface Mo and Se atomic ratio is 5.91% and 11.61%, respectively, corresponding to the formula of MoSe2. Moreover, the molar ratio of Ti:
Mo is 2.26, which is also close to results of EDS.
The annealed MoSe2@C/TiO2 heterostructure was assembled into coin cells and evaluated for lithium-ion electrochemical performance. The MoSe2/TiO2 nanocomposites, pure TiO2 nanobelts and MoSe2 nanosheets were also employed to the same tests for comparison in assistant to understand discharge–charge process in detailed. Fig. 7a shows the first three cyclic voltammetry (CV) profiles of MoSe2@C/TiO2 heterostructures in the voltage range 0.01–3 V vs. Li+/Li at scan rate of 0.2 mV s−1. As can be seen, the CV behavior is generally in agreement with MoSe2 and TiO2 nanostructures reported previously.11,20,22 During the initial discharge process, three main cathodic peaks appear at around 1.45 V, 0.93 V and 0.37 V. The first cathodic peak at 1.45 V corresponds to the lithiation of TiO2 nanobelts according to reaction (2). The pronounced peak at 0.93 V is due to the intercalation of Li+ into the layered MoSe2 lattice accompanied with the formation of LixMoSe2 (reaction (3)). The other cathodic located at 0.37 V is associated to decomposition reaction from LixMoSe2 to Mo metal and Li2Se (reaction (4)).
TiO2 + xLi+ + xe− ↔ LixTiO2 | (2) |
MoSe2 + xLi+ + xe− → LixMoSe2 | (3) |
LixMoSe2 + Li+ + e− → Mo + Li2Se | (4) |
In the reversible charge process, the first two anodic peaks observed at 1.56 V and 1.65 V can be assigned to pseudo-capacitive lithium-ion storage behavior of TiO2(B), which are in pairs of S-peaks with subsequent discharges peaks at 1.48 V and 1.53 V.23–25 The sharp anodic peak at 2.13 V indicates the lithium extraction process of Li2Se to Se. Moreover in the successive cathodic sweeps, the broad peak at around 1.9 V can be interpreted as the lithiation process of Se to form Li2Se, which is similar with the well-known peak of formation Li2S in lithium-sulfur battery systems.26 This apparent change between the first and subsequent discharges is consistent with the previous reports, illustrating that MoSe2 experiences an irreversible phase transition in the initial discharge. Even more, a small pair of successive cathodic peak at 1.25 V and anodic peak at 1.4 V is attributed from transformation of Mo and MoSe2 which is in accordance with pure MoSe2 nanosheets. Notably, the areas of CV curves of MoSe2@C/TiO2 heterostructures in Fig. 7a are obviously larger than pure MoSe2 nanosheets and bare TiO2 nanobelts (Fig. S3†), further indicating hybrid composites exhibit higher specific capacity.
Fig. 7b shows the representative galvanostatic discharge–charge (GDC) voltage profiles of the MoSe2@C/TiO2 heterostructures for the first, second and fiftieth cycles at a constant current density of 500 mA g−1. The plateaus in the voltage profiles are in accordance with the distinct peaks in the CV curves. For comparison, the pure MoSe2 nanosheets and bare TiO2 nanobelts are also tested (Fig. S4†). As can be seen, two apparent voltage plateaus at 1.4 V and 0.9 V, and small plateau at 0.5 V can be observed in the initial discharge, and the subsequent GDC cycles generally move up with plateaus at 2.0 V and 1.5 V. The MoSe2@C/TiO2 heterostructures exhibit the initial discharge and charge specific capacities of 1153.3 mA h g−1 and 861.8 mA h g−1 respectively, giving a coulombic efficiency (CE) of 74.4%, which can be mainly ascribed to the gel-like polymeric layer formation on the MoSe2@C/TiO2 and side reactions of electrolyte reduction in the initial cycle.20,27 The discharge and charge specific capacities in the second cycle are 866.8 mA h g−1 and 831.7 mA h g−1, respectively, with a higher coulombic efficiency of 96%. This value can be retained as 98% even after 50 charge–discharge cycles. It should be noted that the total mass of heterostructures was used in calculating specific capacities. Comparatively, the pure MoSe2 nanosheets exhibited initial discharge and charge specific capacities of 991.3 mA h g−1 and 726.1 mA h g−1, respectively, and the corresponding initial CE is 73.2%. The initial discharge and charge specific capacities of TiO2 are 282 mA h g−1 and 140.8 mA h g−1, respectively, with corresponding initial CE of 49.5%. It can be seen that the MoSe2@C/TiO2 heterostructures possess superior Li+ insertion and extraction capabilities to the single component, further demonstrating the favorable synergistic effect of the MoSe2@C/TiO2 heterostructures.
The MoSe2@C/TiO2 heterostructures also exhibit excellent cycling stability and rate performance. Fig. 7c displays the comparative rate capability of MoSe2@C/TiO2 heterostructures with MoSe2/TiO2 nanocomposites, pure MoSe2 nanosheets and bare TiO2 nanobelts. MoSe2@C/TiO2 heterostructures exhibit a reversible discharge specific capacity of around 920 mA h g−1 in the first ten cycles at current density of 500 mA g−1, and retain a capacity of around 860 mA h g−1 even at the maximum current rate of 3000 mA g−1. After cycling under high current densities, the current density decrease to 500 mA g−1 and the MoSe2@C/TiO2 electrodes reveal an increased discharge specific capacity of 1070 mA h g−1. While the MoSe2/TiO2 nanocomposites electrodes deliver a combined rate performance of pure MoSe2 and bare TiO2. The behavior of capacity change is very similar to that of bare TiO2 during the early cycle, which is 309 mA h g−1 of 500 mA g−1 to 263 mA h g−1 of 3000 mA g−1. After the current density return to 500 mA g−1, an increased capacity is obtained, and this is the significant characteristic of MoSe2. The results confirmed that both the ability of Li+ store and exposed contact area of MoSe2 thick flakes and nanoparticles are utterly inadequate to few-layered MoSe2 nanosheets.
Apart from the excellent rate capability and high reversible capacity, MoSe2@C/TiO2 heterostructures also exhibit good cycling behavior. Fig. 7d displays the cycling performance of MoSe2@C/TiO2 heterostructures, MoSe2/TiO2 nanocomposites, pure MoSe2 nanosheets and bare TiO2 nanobelts at a current density of 500 mA g−1 after 100 cycles. The MoSe2@C/TiO2 heterostructures and MoSe2 nanosheets electrodes deliver increased capacity in accordance with expectation. In the case of MoSe2@C/TiO2 heterostructures, the discharge specific capacity increased from 866 mA h g−1 to 1050 mA h g−1 with coulombic efficiency over 97%. After transitory float of capacity, the discharge capacity of the MoSe2@C/TiO2 heterostructures electrodes remains at 987.4 mA h g−1 in 100th cycle with coulombic efficiency of 98.9%. While pure MoSe2 nanosheets exhibit severe increase of specific capacities up to 1007 mA h g−1 in first sixty cycles. However there is a sharp decline of nearly 50% specific capacities during following cycling process, and this represents destructive structural collapse. In contrast, the specific capacity of bare TiO2 nanobelts retains at about 130 mA h g−1 over 100 cycles, which shows the remarkable cycling stability. Note that MoSe2/TiO2 nanocomposites exhibit synthetic results of individual components as expected. Due to the compact structure, structural collapse of thick flakes and particles is limited, following relatively stability instead of sharp decline. Therefore, the heterostructures capacities significantly exceed those of either individual component even over 100 cycles, thus the superior performance of MoSe2@C/TiO2 heterostructures is confirmed. In addition, the SEM images (Fig. S5†) of MoSe2@C/TiO2 heterostructures electrodes after cycling for 100 cycles show that original 1D morphological characteristic still retained, demonstrating that robust skeletons could accommodate stress from cycling.
The increased specific capacity could be explained in detail as following reasons: (i) the layered structures of few-layered MoSe2 nanosheets provide more contact of electroactive components and electrolyte, which may exist activation of Li+ during cycling process; (ii) as described in CV analysis, similar with MoS2 electrodes,28 Li–Se charge–discharge process makes the major contribution to the reversible capacity during later cycling. These gradual float of specific capacity is owing to the presence of these changes in electrochemical process. Thus, with more clustered few-layered nanosheets, it is accessible that MoSe2 electrodes exhibit further increase capacity and larger extent structural collapse than unfold few-layered nanosheets in heterostructures.
In order to better understand the kinetics of electrode reaction between electrodes and electrolyte, EIS measurements were carried out for all electrodes. Fig. 8 shows the Nyquist plots of MoSe2@C/TiO2 heterostructures and pure MoSe2 nanosheets, in which the inset represents the equivalent circuit model fitting well with experimental results. Re represents the internal resistance of the test battery, Rf and Q1 are related to the resistance of the SEI film corresponding to the high frequency semicircle. The medium frequency semicircle could be attributed to the charge transfer resistance Rct and double-layer capacitance Q2 of the electrode–electrolyte interface, and a sloping line at low frequency region is associated with lithium ion diffusion processes. Obviously, compared to pure MoSe2 nanosheet (Rct = 554.8 Ω), the MoSe2@C/TiO2 heterostructures (Rct = 44.8 Ω) show a smaller semicircle diameter, indicating that MoSe2@C/TiO2 remarkably enhanced the rapid electron transport owing to the close contact between few-layered MoSe2 nanosheets and conductive skeletons with the addition of glucose.
It is deserved to be mentioned that the average specific capacity of our MoSe2@C/TiO2 heterostructures is superior to previously reported MoSe2 based products (Table S1†). The impressive electrochemical performance of MoSe2@C/TiO2 can be assigned to the rationally design of heterostructures. Firstly, the addition of glucose plays a critical role in synthesize process to maintain few-layered MoSe2 nanosheets uniformly covering on TiO2 skeletons. Ultrathin feature of the MoSe2 nanosheets can store more Li+ and facilitate fast Li+ diffusion; on the other hand, unfolded ultrathin MoSe2 nanosheets expose more contact surface, which avoid the aggregation of MoSe2 nanosheets and accelerate penetration of electrolyte. Moreover, TiO2 nanobelts skeletons could accommodate stress from cycling, maintain structural integrity and buffer severely volume change of MoSe2 nanosheets. Further coupled with carbon compounds, robust skeleton act as beneficial electronic co-conductor and provide fast pathways for electron transportation.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra01579d |
This journal is © The Royal Society of Chemistry 2016 |