Ultrasmall Fe2O3 nanoparticles/MoS2 nanosheets composite as high-performance anode material for lithium ion batteries

Coupling ultrasmall Fe2O3 particles (~4.0 nm) with the MoS2 nanosheets is achieved by a facile method for high-performance anode material for Li-ion battery. MoS2 nanosheets in the composite can serve as scaffolds, efficiently buffering the large volume change of Fe2O3 during charge/discharge process, whereas the ultrasmall Fe2O3 nanoparticles mainly provide the specific capacity. Due to bigger surface area and larger pore volume as well as strong coupling between Fe2O3 particles and MoS2 nanosheets, the composite exhibits superior electrochemical properties to MoS2, Fe2O3 and the physical mixture Fe2O3+MoS2. Typically, after 140 cycles the reversible capacity of the composite does not decay, but increases from 829 mA h g−1 to 864 mA h g−1 at a high current density of 2 A g−1. Thus, the present facile strategy could open a way for development of cost-efficient anode material with high-performance for large-scale energy conversion and storage systems.

achieved by the large surface area provided by the ultrasmall particles 39 . More importantly, due to the extremely short distance for lithium ions transportation within ultrasmall particles, the rate capability of lithium insertion/ removal can be significantly enhanced 40 . However, the nanostructured Fe 2 O 3 exhibited a poor cycling stability due to structural damage during charging/discharging process 37 .
Herein, we report a facile method to grow ultrasmall Fe 2 O 3 nanoparticles on 2H-MoS 2 nanosheets, where MoS 2 nanosheets in the composite can serve as scaffolds, efficiently buffering the large volume changes of Fe 2 O 3 during charging/discharging process, whereas the ultrasmall Fe 2 O 3 nanoparticles mainly provide the specific capacity of the anode as well as the enhanced electrical conductivity. Furthermore, strong coupling between Fe 2 O 3 and MoS 2 nanosheets, elucidated by X-ray photoelectron spectrum measurements, facilitates a rapid charge transfer. In addition, MoS 2 nanosheets in the composite can also contribute to the total capacity of the anode. As a consequence, the composite prepared here exhibited superior electrochemical performance for anode material for Li-ion battery.

Results
SEM images ( Fig. 1a and b) show that the as-prepared MoS 2 exhibits sheet-like morphology with a thickness and a lateral length of about 10 and 400 nm, respectively, similar to that reported previously 41 . After Fe 2 O 3 coating, the MoS 2 exhibits a similar morphology and lateral length to that of the pristine MoS 2 nanosheets, but the surface becomes drastically rough (Fig. 1c), in sharp contrast to the smooth surface of the pristine MoS 2 nanosheets (Fig. 1b). From the high-magnification SEM image of the composite (Fig. 1d)), it can also be found that many ultrasmall particles are anchored on both sides of basal planes of MoS 2 nanosheets. Figure 2a shows a TEM image taken from the basal plane of MoS 2 nanosheets in the composite. It can be found that Fe 2 O 3 nanoparticles are uniformly and densely deposited on the surface of the MoS 2 nanosheets. The size distribution plot ( Figure S1) indicates that the average size of Fe 2 O 3 nanoparticles is about 4.0 nm. Most lattice fringes of the Fe 2 O 3 nanoparticles in the high-resolution TEM (HRTEM) image (Fig. 2b) are not resolved well, revealing the weak degree of  Figure S2). Cross-section TEM image (Fig. 2c) reveals that the utrasmall Fe 2 O 3 nanoparticles mainly disperse on the basal planes of the MoS 2 nanosheets, in which the lattice fringes corresponding to (002) plane can be clearly observed. HRTEM image (Fig. 2d) reveals that the interlayer distance of the (002) plane of the MoS 2 nanosheets is about 0.707 nm, larger than the value (0.615 nm) of bulk MoS 2 .
The crystal structures of the samples were examined using X-ray diffraction (XRD) measurement. Figure 3a shows the XRD pattern of the pristine MoS 2 nanosheets, in which the peaks located at 2θ = 32.2° corresponds to the (100) and (101) planes, 2θ = 38.3° corresponds to the (103) plane and the peaks located at 2θ = 57.3° corresponds to the (110) and (008) planes of 2H-MoS 2 (JCPDS No. . Compared to 2H-MoS 2 bulk, these peaks slightly shift toward low-angle region, revealing the slightly enlarged lattice distances along the basal planes of 2H-MoS 2 . Similar to the previous report 41 , two additional peaks located at 9.2° and 18.5°, marked by "#", are also observed at low-angle region. The corresponding d-spacings calculated according to the Bragg equation are 0.96 and 0.48 nm, respectively. The diploid relation between the d-spacings reveals that the MoS 2 nanosheets possess a new lamellar structure with a larger interlayer spacing of 0.96 nm than that of 0.615 nm in bulk 2H-MoS 2 34,41,42 . The enlarged interlayer spacing may be related to the synthesis conditions 28,34,41,42 . As previously reported 41 , when the temperature was lower than 180 °C, the MoS 2 nanosheets with enlarged interlayer spacing (0.95 nm) could be obtained in alkaline media; while the temperature was increased to 220 °C, the interlayer distance of the nanosheets kept the same value as that in bulk MoS 2 41 . On the other hand, the enlarged interlayer spacing could be achieved in the media containing urea at 220 °C; however, when the pH in the media was decreased by replacing urea with ammonium fluoride, even at the same temperature the phenomenon did not occur 42 . Therefore, the alkalinity and the synthetic temperature seriously affect the interlayer distance of the MoS 2 nanosheets. Under the experimental conditions such as strong alkaline media and low temperature, oxygen species may incorporate more easily with MoS 2 , leading to the different lamellar structure with a enlarged interlayer spacing than that of 0.615 in bulk 2H-MoS 2 34,41,42 . However, the new lamellar structure is thermodynamically unstable. After annealing the MoS 2 nanosheets at 500 °C for 3 h under an Ar flow, XRD analysis was carried out. As shown in Figure S3, the diffraction peaks at 9.2° and 18.5° disappear, while all the resolved peaks can be assigned to thermodynamically stable 2H-MoS 2 (JCPDS No. 37-1492). After Fe 2 O 3 coating, the peak corresponding to the (002) plane is suppressed significantly, further suggesting that uniform and dense nanoparticles are deposited on the both sides of the basal plane of the MoS 2 nanosheets (Fig. 3b). In addition, the peaks from the Fe 2 O 3 and MoS 2 can also be identified in the XRD pattern. The diffraction peaks at 2θ = 33.4°can be assigned to (100) and (101) planes, 2θ = 39.5°can be assigned to (103) plane and 2θ = 58.5° can be assigned to (110) Fig. 3a is different from that in Fig. 3b. This is because the pristine MoS 2 nanosheets with the enlarged interlayer spacing are thermodynamically unstable phases, while the MoS 2 nanosheets in Fe 2 O 3 /MoS 2 composite, which were annealing at 500 °C for 3 h under an Ar flow, are thermodynamically stable. They had different lamellar structures, leading to significantly difference in XRD results ( Fig. 3 and Figure S4). Compared to the XRD pattern of the annealed MoS 2 nanosheets with that of the Fe 2 O 3 /MoS 2 composite, the peak position of (110) plane is almost identical ( Figure S5). The above results demonstrate that crystalline Fe 2 O 3 nanoparticles are successfully anchored on the surface of MoS 2 nanosheets. Figure 3c) shows the XRD pattern of the physical mixture Fe 2 O 3 + MoS 2 , in which the diffraction peaks from both Fe 2 O 3 and MoS 2 can be observed. Notably, the diffraction peak corresponding to expanded (002) plane of MoS 2 nanosheets is still visible, significantly different from that of the Fe 2 O 3 /MoS 2 composite. X-ray photoelectron spectroscopy (XPS) analysis was carried out to determine surface chemical compositions and valence states of the Fe 2 O 3 /MoS 2 composite and the MoS 2 nanosheets. Figure 4a shows the high-resolution XPS spectra of Mo 3d core level for the two samples. Two peaks at 231.9 eV and 228.7 eV are observed in the Mo 3d spectrum of the MoS 2 nanosheets, corresponding to Mo 4+ species. After coating Fe 2 O 3 , the two peaks shift to high binding energy side by an approximately 0.3 eV. The shift of the binding energy indicates that electron transfer from MoS 2 to Fe 2 O 3 occurs. It can be concluded that the strong coupling between MoS 2 and Fe 2 O 3 is presented in the composite. However, such shift in the physical mixture MoS 2 + Fe 2 O 3 does not occur ( Figure S6), revealing the advantage of our method for preparation of the Fe 2 O 3 /MoS 2 composite. In addition, two additional peaks at 235.6 eV and 232.5 eV are assigned to Mo 6+ species [43][44][45] , suggesting the surface oxidization of MoS 2 due to the electron transfer. In the XPS spectrum of S 2p core level for the pure MoS 2 nanosheets, the main doublet located at binding energies of 161.6 and 162.9 eV correspond to the S 2p 3/2 and S 2p 1/2 , respectively ( Fig. 4b) 41 . There are no obvious shift of the binding energy of the two peaks for the Fe 2 O 3 /MoS 2 composite, implying that the Fe 2 O 3 coating has little effect on the valence states of the S species. In the Fe 2p core level spectrum for the Fe 2 O 3 / MoS 2 composite, the peaks at 711.3 eV, 719.2 eV and 724.8 eV represent the binding energies of Fe 2p 3/2 , shake-up satellite Fe 2p 3/2 , and Fe 2p 1/2 of Fe 3+ species, respectively ( Fig. 4c) 46,47 . These values are consistent with the data of Fe 2 O 3 reported by the previous literatures 48 41 , respectively (Fig. 4d). Besides, the peak at 531.5 eV is associated to the hydroxyl oxygen. Cyclic voltammogram (CV) tests for coin cells of the pristine MoS 2 nanosheets were recorded at ambient temperature in the voltage range of 0.01-3 V at a scan rate of 1 mV s −1 for the initial five cycles, as shown in Fig. 5a. The peak of 0.87 V in the first cathodic scanning is ascribed to the intercalation of lithium ion on different defect sites in MoS 2 to form Li x MoS 2 . In the following cathodic scanning, two new reduction peaks at approximately 1.65 V and 1.15 V are observed, which are due to the conversion of S to Li 2 S and the association of Li with Mo respectively 52,53 . During the anodic scans, two peaks at 1.92 and 2.42 V are clearly observed and maintain for the subsequent sweeps, which are related to the conversion reaction of Mo and Li 2 S to MoS 2 phase 54,55 . As for the pure Fe 2 O 3 in Fig. 5b, the reduction peaks at 0.36 V is observed in the first cycle, and its position shifts to 0.57 V at the following scanning, which is attributed to the reduction of Fe(III) to Fe(0). In the anodic scans, the oxidation peak at 1.95 V is the oxidation of Fe to Fe 2 O 3 . As for the Fe 2 O 3 /MoS 2 composite, three reduction peaks at 0.36 V, 0.87 V and 1.20 V are observed in the first cycle (Fig. 5c). The reduction peaks locate at 0.36 V and 0.87 V during the first anodic scan can match anodic scan peaks of pure Fe 2 O 3 and pure MoS 2 , respectively. The peak located at 1.20 V shows a same start shoulder at ~1.05 V with pure MoS 2 , which suggest the same lithiation process of  To confirm the superiority of the Fe 2 O 3 /MoS 2 composite as an anode material over the pristine MoS 2 nanosheets and the physical mixture Fe 2 O 3 + MoS 2 in the lithium storage performance, we compared their cycling behaviors at different current densities (Fig. 6). Clearly, MoS 2 nanosheets deliver an initial capacity of 854 mAhg −1 at a current density of 100 mA g −1 (Fig. 6a), higher than its theoretical value due to its ultra-thin nanosheets for lithium storage. However, obvious capacity decay is witnessed when cycled at a high current density (Fig. 6a). For example, the capacity decreases from 638 mA h g −1 to 449 mA h g −1 at 1 A g −1 only after 40 cycles. This phenomenon is probably due to the exfoliation of the ultra-thin MoS 2 nanosheets during discharging/charging process. Similarly, the pure Fe 2 O 3 delivers an initial discharge capacity of 1219 mAh g −1 at a current densities of 100 mA g −1 , and the capacity decreases from 533 mAh g −1 to 287 mAh g −1 at 1 A g −1 only after 55 cycles (Fig. 6b). As the MoS 2 nanosheets are mixed with Fe 2 O 3 mechanically, slightly better cycling stability at the current densities of 100 mA g −1 and 1 A g −1 than the pristine MoS 2 nanosheets can be achieved, as shown in Fig. 6c. However, the physical mixture still shows poor cycling stability at a high current density. For example, the discharge capacity of the physical mixture decreases from 698 mA h g −1 to 545 mA h g −1 after 80 cycles at a high current density of 1 A g −1 (Fig. 6c). In contrast, the Fe 2 O 3 /MoS 2 composite exhibits drastically enhanced rate capability (Fig. 6d,e and Figure S8). Furthermore, the Fe 2 O 3 /MoS 2 composite shows an excellent cycling durability at different current densities. For example, the Fe 2 O 3 /MoS 2 composite delivers an initial discharge capacity of 1366 mAh g −1 at current densities of 100 mA g −1 . The capacity does not decay after 150 cycles, but gradually increases to 1350 mA h g −1 with a high Coulombic efficiency of > 98.7% (Fig. 6d). Surprisingly, even at high current densities of 1 and 2 A g −1 , the composite also exhibits excellent cycling stability (Fig. 6e). The capacity of the Fe 2 O 3 /MoS 2 composite increases from 908 mA h g −1 to 1011 mA h g −1 at 1A g −1 , and from 829 mA h g −1 to 864 mA h g −1 at 2 A g −1 after 140 cycles (Fig. 6e). The cycling performance is inferior to that of MoS 2 /graphene composite with the capacity of 907 mA h g −1 after 400 cycles 29 , however, comparable or superior to most other Fe 2 O 3 and MoS 2 or their composites, which is summarized in Table S1 8,13,21,37,[61][62][63][64][65][66][67][68][69][70][71] . For example, the capacity of Fe 2 O 3 nanoparticles was only about 300 mAh g −1 at ca. 100 m A g −1 after 100 cycles 37 ; the capacity of CNTs-MoS 2 was 737 mAh g −1 at 100 m A g −1 after 30 cycles 69 . Furthermore, when the current density is increased to 5 A g −1 , the composite shows a relatively bad cycling durability, but still delivers a capacity of 481 mAh g −1 after 140 cycles. To reveal the charge/discharge stability of anode, the SEM and elemental mapping analyses of Fe 2 O 3 /MoS 2 composite after 100 cycles were carried out. After the cycling, the utrasmall Fe 2 O 3 nanoparticles are still resolved in the basal planes of the MoS 2 nanosheets, as shown in Figure S9. Elemental mapping images ( Figure S10) reveal that Fe and Mo elements are uniformly distributed in the composite. These results above demonstrate that the Fe 2 O 3 /MoS 2 composite exhibit significantly enhanced capacity, rated capability and cycling stability compared to the pristine MoS 2 nanosheets, the pure Fe 2 O 3 and the physical mixture Fe 2 O 3 + MoS 2 . Notably, the electrochemical performance of Fe 2 O 3 /MoS 2 composite was measured at room temperature. It is well known that the electrochemical performance of the anode material for LIBs is suppressed significantly at the ambient temperature lower than 0 °C. However, as previously reported, the specific capacity of MoS 2 /G electrode at − 20 °C still remained ca. 700 mAh g −1 at 100 mA g −129 . This result indicates that MoS 2 -based anode material may be used at low-temperature environment. The electrochemical performance of our Fe 2 O 3 /MoS 2 composite at such low-temperature environment is studied under way.

Discussion
The excellent electrochemical properties of the Fe 2 O 3 /MoS 2 composite, as evidenced by a remarkably increased reversible capacity, improved rate capability, and robust long-term stability even at a high current density, indicates that the Fe 2 O 3 /MoS 2 composite is favorable for superior anode materials for Li-ion battery. The following factors can be attributed to the improved electrochemical properties of the Fe 2 O 3 /MoS 2 composite. First, unlike some designed composites 21,61,63,68,69,71 , both Fe 2 O 3 and MoS 2 in our composite can contribute to the total capacity of the anode, elucidated by CV measurements (Fig. 5c). The high reversible capacity of the Fe 2 O 3 /MoS 2 composite may also be related to its unique heteronanostructure character. As we know, during the first discharge process, Fe 2 O 3 reacts with Li + , and then Fe and Li 2 O will gradually produce (Equation 1) 72 ; however, only partial Li 2 O can reversibly converse to Li + during the subsequent charging process, leading to a high irreversible capacity of Fe 2 O 3 -based anodes. On the other hand, during the first discharging process of MoS 2 , amorphous Mo metal clusters will form (Equation 2) and disperse on the surface of MoS 2 . The Mo metal clusters have highly electrochemical activity 73,74 . Considering the unique heteronanostructure of the Fe 2 O 3 /MoS 2 composite, the Mo metal clusters on the MoS 2 surface can efficiently contact with Li 2 O and make the irreversible Li 2 O converse to Li + , as shown in Fig. 7a. As a result, the Fe 2 O 3 /MoS 2 composite shows a low irreversible capacity and a high Coulombic efficiency of 88.4% at the first cycle. As for the physical mixture Fe 2 O 3 + MoS 2 , since the efficient contact between Mo and Li 2 O is more difficultly available, the conversion of Li 2 O to Li + will be greatly suppressed (Fig. 7b). Consequently, the physical mixture Fe 2 O 3 + MoS 2 shows a high irreversible capacity and a low Coulombic efficiency at the first cycle ( Fig. 5d and h).
Second, the average Fe 2 O 3 size is approximately 4.0 nm, which can shorten the Li ion transfer length and then facilitate the improvement of the rate capability. Additionally, nitrogen adsorption-desorption isotherms show that Brunauer-Emmett-Teller (BET) surface areas and cumulative volume of pores were 12.5 m 2 g −1 and 0.06 cm 3 g −1 for the pristine nanosheets, around two times lower than those of the composite (23.1 m 2 g −1 and 0.12 cm 3 g −1 ), as shown in Figure S11 and Figure S12. The bigger BET surface area and larger pore volume not only allow for fast Li-ion diffusion, but also buffer the volume changes accompanying the Li charging and discharging processes 28 . Third, the strong coupled interfaces boost a rapid interfacial charge transfer, leading to excellent rate capability of the Fe 2 O 3 /MoS 2 composite, as evidenced by electrochemical impedance measurements. Nyquist plots (Fig. 8) shows that the Fe 2 O 3 /MoS 2 composite has a charge transfer resistance (R ct ) of 39.9 Ω, greatly smaller than that of the physical mixture Fe 2 O 3 + MoS 2 (238.9 Ω), the pure Fe 2 O 3 (136.0 Ω) and the In summary, a facile and cost-effective strategy was developed to anchor ultrasmall Fe 2 O 3 nanoparticles on the surface of MoS 2 nanosheets. Due to the synergistic effect of two excellent anode materials and the unique structural feature, the Fe 2 O 3 /MoS 2 composite exhibits excellent electrochemical properties, including a remarkably increased reversible capacity, improved rate capability, and long-term stability even at a high current density. After 140 cycles the reversible capacity of the composite does not decay, but increases from 829 mA h g −1 to 864 mA h g −1 at a high current density of 2 A g −1 , outperforming other MoS 2 -and iron oxide-based anode materials previously reported. Thus, the facile strategy may open a way for development of cost-efficient anode material with high-performance for large-scale energy conversion and storage systems.

Methods
Synthesis of samples. MoS 2 nanosheets were first synthesized by a solution-based method 40    nanosheets (30 mg) was dispersed in ethyl alcohol (75 mL) and then iron acetylacetone (0.5 mmol), distilled water (1.8 mL) and ammonia (2 mL) were added. After sonication for 15 min at room temperature the mixture was heated at 80 °C for 10 h in a water bath. The precipitates were separated by centrifugation, washed with distilled water and ethanol, and dried at 40 °C for 24 h under vacuum. The growth of ultra-small Fe 2 O 3 nanoparticles on the MoS 2 nanosheets was achieved after the dried powder was thermally treated at 500 °C for 3 h under an Ar flow. For convenience, the obtained sample denoted as Fe 2 O 3 /MoS 2 composite. As shown in Figure S13, the Fe and Mo atom ratio was 2:1. The Fe 2 O 3 was prepared by heating graphene-hollow iron oxide at 500 °C for 1.5 h under air atmosphere 72  Step 0.02). X-ray photoelectron spectra (XPS) were carried out by using a spectrometer with Mg Kα radiation (PHI 5700 ESCA System). The binding energy was calibrated with the C 1 s position of contaminant carbon in the vacuum chamber of the XPS instrument (284.6 eV). The pore diameter distribution and surface area were tested by nitrogen adsorption/desorption analysis (TRISTAR II3020).
Electrochemical measurements. The electrochemical tests were performed at ambient temperature using two-electrode coin cells (CR 2016) with lithium foils serving as the counter electrode. The active material was mixed with a conductive acetylene black, and a commercial polymer binder (LA133) at a weight ratio of 70:15:15. The mixture was painted onto a Cu foil and dried in air, then cut into 14 mm diameters of round piece. Finally the electrode pieces were dried in vacuum at 60 °C for 12 h to adequately evaporate the residual moisture. The average thickness and mass loading of the electrode were ~3.5 μm ( Figure S14) and 1.5 ± 0.2 mg, respectively. The electrolyte was made of LiPF 6 (1 M) in a mixture of ethylene carbonate (EC) and dimethyl carbonate (DMC) with the volume ratio 1:1. The 2016 coin-type cells were assembled in an Ar filled glove box, and pure Li foils were used as the counter electrodes. The charge-discharge cycles were carried out on a battery measurement system (LAND-BT2013A) at various current densities of 100-5000 mA g −1 in the cutoff voltage range of 3 to 0 V versus Li/Li + at room temperature (~20 °C). Cyclic voltammetry measurements were carried out on a CHI660D electrochemical workstation over the potential range of 3.0 to 0.01 V at a scan rate of 1 mV s −1 .