Facile Constructing Hierarchical Fe₃O₄@C Nanocomposites as Anode for Superior Lithium-Ion Storage

Abstract

Ferroferric oxide (Fe₃O₄) is regarded to be a promising high-capacity anode material for LIBs. However, the capacity attenuates fast and the rate performance is poor due to the dramatic pulverization and sluggish charge transfer properties. To solve these problems, a simple in situ encapsulation and composite method was successfully developed to construct carbon nanotube/nanorod/nanosheet-supported Fe₃O₄ nanoparticles. Owing to the hierarchical architecture design, the novel structure Fe₃O₄@C nanocomposites effectively enhance the charge transfer, alleviate pulverization, avoid the agglomeration of Fe₃O₄ nanoparticles, and also provide superior kinetics toward lithium storage, thereby showing significantly improved reversibility and rate performance. The carbon nanotube/nanorod supported core-shell structure Fe₃O₄@C nanocomposite displays outstanding high rate capability and stable cycling performance (reversible capability of 1006, 552 and 423 mA h g⁻¹ at 0.2, 0.5 and 1 A g⁻¹ after running 100, 300 and 500 cycles, respectively).

1. Introduction

The lithium-ion battery (LIB), an effective energy storage device, is extensively applied to electric vehicles and various portable devices, e.g., mobile phones, laptops and so on [1,2,3]. However, commercial graphite as an anode material has fatal disadvantages of low specific capacity (372 mAh g⁻¹) and deficient rate performance, so the present LIB cannot support sufficient capacity for intelligent electronic devices to satisfy people’s demand for long working time [1,4]. To solve the problem, we need to explore new suitable anode materials with high specific capacity and excellent cycling stability. In the past decades, intensive efforts have been employed to explore various kinds of new anode materials, including Si, Sn, NiSb, etc.; alloy-type [5,6,7,8,9] and transition metal oxides (e.g., Fe₂O₃/Fe₃O₄, Co₃O₄, SnO₂, etc.) [10,11,12,13,14] and so on. Ferroferric oxide (Fe₃O₄) is regarded to be a promising anode material and attracts many researchers’ interest due to the merits of high reversible theoretical specific capacity (928 mAh g⁻¹), low cost and eco-friendliness [9,10,15,16,17]. Unfortunately, the capacity of the Fe₃O₄ anode decays very fast, which results from the severe pulverization of Fe₃O₄ particles because of the huge volume variation caused by lithiation and delithiation; therefore, the Fe₃O₄ anode exhibits poor cycling stability [18,19,20,21]. On the other hand, the charge transfer property of the pristine Fe₃O₄ is poor and needs to be improved in an electrode material [22,23].

To solve these disadvantages, many strategies were explored to enhance the structural stability and improve the electrical conductivity of Fe₃O₄ [23,24,25,26,27]. Concerning enhancing the structural stability, the synthesis of nanostructured Fe₃O₄, such as nanoparticles, nanorods, hollow spheres, and so on, was intensively investigated [24,25,26,27,28,29]. The other way is designing Fe₃O₄@C nanocomposites, including carbon-coated Fe₃O₄ nanocomposites [16], graphene-wrapped nano-Fe₃O₄ or 3D structured Fe₃O₄@C nanocomposites [30,31,32]. Compared with the bulk Fe₃O₄, nanostructured Fe₃O₄ is not only beneficial in alleviating the stress but also helps to reduce the Li/Li⁺ diffusion distance [33,34,35]. It was reported that the single-crystalline Fe₃O₄ and α-Fe₂O₃ nanorods could alleviate stress effectively and supply a larger electrochemical reaction area [36]. With respect to Fe₃O₄@C nanocomposites, the carbon matrix severed as a buffer for the accommodation of strain and inhibited the agglomeration and coarsening of nano-Fe₃O₄ [37]. On the other hand, attributed to its good conductivity, the introduction of carbon also enhances the charge transfer of Fe₃O₄ [38]. As a result, the Fe₃O₄@C nanocomposites always display improved cycling stability. For example the Fe₃O₄ nanoparticles encapsulated by macroporous carbon spheres exhibited a discharge capacity of 645 mAh g⁻¹ after 1000 cycles at 2.0 A g⁻¹ [28].

Recently, constructions of yolk-shell or other hollow structures became a favorable methodology to realize the special structural effects [39,40,41,42]. Metal–organic frameworks (MOFs) were considered to be the ideal precursors for the preparation of functional porous materials (metal/metal oxide and carbon composites) with desired shapes [17,41,43,44]. Using ferrocene as the precursor, Liu and co-authors synthesized a pitaya-structured Fe₃O₄@C nanocomposite and finally obtained the same structured FeS₂@C nanocomposites as a cathode material for LIBs, which displayed stable and high energy density (~1100 Wh kg⁻¹) and ultrahigh rate capability41. The yolk-shell Fe₂O₃@C composite containing mesoporous derived from MIL-53(Fe), could retain 744 mAh g⁻¹ after 500 cycles at 1 A g⁻¹ [43] Although notable successes have been achieved, these chemical synthesis methods were severely limited by their multistep, the use of organic solvents and low yield [45] In this work, we presented a facile and suitable large-scale application method to prepare Fe₃O₄@C nanocomposites with controllable morphologies by carbonizing the plasma-assisted ball-milled ferrocene (C₁₀H₁₀Fe) precursor. The leaves-like amorphous carbon-nanosheet-encapsulated ultrafine nano-Fe₃O₄ and carbon nanotubes/nanorods-supported core-shell nanostructured Fe₃O₄ composites were prepared and exhibited outstanding electrochemical performances as an anode material for LIBs. Typically the carbon nanotubes/nanorods-supported core-shell nanostructured Fe₃O₄ presented impressive cycling performance (423 mAh g⁻¹ at 1 A g⁻¹ after 500 cycles) and outstanding high-rate capability.

2. Experiments

2.1. Material Preparation and Characterization

The synthesis of Fe₃O₄@C nanocomposites is illustrated in Scheme 1. First, the purchased ferrocene (C₁₀H₁₀Fe) powders were milled for 5 h on a plasma-assisted ball miller (PA miller, homemade) to achieve partially carbonized pretreated products. Then, these pretreated products were annealed in the tube furnace at 773 and 873 K for 30 min under the atmosphere of argon and air mixtures. The obtained black powders were the final Fe₃O₄@C nanocomposites. The Fe₃O₄@C nanocomposites achieved at 773 and 873 K were named Fe₃O₄@C-773 and Fe₃O₄@C-873, respectively, in the following.

The phase components of the Fe₃O₄@C composites were examined by an X-ray diffractometer (XRD, Rigaku SmartLab 2 kW, Tokyo, Japan) using Cu-Kα radiation (λ = 1.54056 Å) at 40 kV and 40 mA. The morphology and microstructure were further investigated by a field emission scanning electron microscope (SEM, ZEISS Sigma 500) and scanning transmission electron microscopy (STEM, FEI TALOS F200S attached BRUKER Super-X EDS). The carbon contents were determined using thermogravimetric analysis (TG, Setaram Labsys) at a heating rate of 2 K/min under an airflow of 50 mL/min. The pore structure and specific surface area were evaluated by the measurement of nitrogen adsorption/desorption isotherms at 77 K on a micromeritics 3Flex instrument (USA). The Raman spectra were tested on an invia apparatus (Renishaw, Wotton-under-Edge, UK). X-ray photoelectron spectroscopy (XPS, Nexsa from ThermoFisher Inc., Waltham, MA, USA) was performed to analyze the chemical components and element valance.

2.2. Electrochemical Measurements

The slurry was obtained by grinding the Fe₃O₄@C composites, binder-polyvinylidene fluoride (PVDF) and conductive additive (Super P) with a weight ratio of 8:1:1 in a mortar with an appropriate Nmethyl-2-pyrrolidone (NMP) on a ball mill. The prepared slurry was coated on the Cu foil, which then was dried at 85 °C for 24 h under vacuum conditions. The button batteries (Li//Fe₃O₄@C) were assembled in a glove box under the protection of high-purity Ar (O₂ and H₂O < 0.1 ppm). Lithium metal was the cathode, a celgard2400 membrane was the separator and a solution of 1 M LiPF₆ dissolved in ethylene carbonate and diethyl carbonate (1:1 in volume) was the electrolyte. The galvanostatic charge/discharge measurements were performed on a LAND-CT2001A battery test system in the voltage range from 0.1 to 3.0 V (vs. Li/Li⁺). A CHI660D electrochemical workstation was used for the measurements of cyclic voltammetry (CV) profiles in a voltage range of 0.01 to 3.0 V by a scan rate of 1 mV s⁻¹, and electrochemical impedance spectroscopy (EIS) was used from 100 kHz to 0.1 Hz.

3. Results and Discussion

Figure 1a shows the XRD patterns of the Fe₃O₄@C nanocomposites. The diffraction peaks at 18.3°, 30.0°, 35.3°, 43.1°, 57.0° and 62.5° match well with the (111), (220), (311), (400), (333) and (440) crystal plane diffractions of Fe₃O₄, respectively. This result indicated that Fe₃O₄ crystalline was formed by sintering the PA-milled C₁₀H₁₀Fe at 773 and 873 K for 30 min. The diffraction intensities of Fe₃O₄ in Fe₃O₄@C-873 were comparatively higher than these of Fe₃O₄@C-773, which could result from the better crystallinity due to the higher annealing temperature for Fe₃O₄@C-873. On the other hand, the (002) peak of carbon (JCPDS no. 89-7213) was confirmed in Fe₃O₄@C-873. However, there was no observed carbon diffraction peak in Fe₃O₄@C-773, which could be attributed to its amorphous structure. Figure 1b shows the TG results for two samples. From Figure 1b, it could be seen that the major weight loss, induced by the removement of carbon due to oxidation, occurred at approximately 750 K. The weight loss starting temperature of Fe₃O₄@C-773 was obviously lower than that of Fe₃O₄@C-873. This could be because the amorphous carbon was more easily oxidated in the air condition. The weight percent of carbon was approximately 29 wt.% and 8 wt.% for Fe₃O₄@C-773 and Fe₃O₄@C-873, respectively, based on the TG results. It implied that the higher annealing temperature would result in more weight loss in carbon.

Figure 2 shows the morphology and microstructure of the Fe₃O₄@C nanocomposites. Fe₃O₄@C-773 and Fe₃O₄@C-873 showed completely different morphologies. As observed in Figure 2a,b, Fe₃O₄@C-773 showed a leaves-like nanosheet structure with the Fe₃O₄ nanoparticles (~5 nm) dispersively embedded in the carbon leaves. The insert of Figure 2b is the STEM-selected area diffraction pattern, which contained an amorphous diffraction ring of carbon together with the (111) and (220) facet families of the polycrystalline Fe₃O₄. These agreed well with the above X-ray diffraction analysis results. Figure 2c is the enlarged image of the marked part in Figure 2a, and the related mapping for C, Fe, and O are illustrated in Figure 2d–f. The mapping results further confirmed that the ultrafine Fe₃O₄ nanoparticles were dispersively incorporated into the carbon matrix. Different from Fe₃O₄@C-773, Fe₃O₄@C-873 was characterized by a nanotube/nanorod structure and coated with larger Fe₃O₄ nanoparticles (~100 nm), as seen in Figure 2g (SEM) and Figure 2h (STEM). From the insert of Figure 2h, it could be clearly seen that the Fe₃O₄ nanoparticles were wrapped by a nanolayer of carbon (~5 nm) with a typical core-shell nanostructure. The core-shell structure could be more favorable for the release of strain caused by Li/Li⁺ insertion/extraction, and the carbon matrix could serve as a buffer for the accommodation of stress resulting from the sharp volume variation in Fe₃O₄, thereby avoiding the severe pulverization of Fe₃O₄ [46]. It was regarded that this novel structure greatly benefitted the enhancement of the structural stability of nano-Fe₃O₄.

Figure 3a illustrates the Raman spectroscopies of the Fe₃O₄@C nanocomposites. Fe₃O₄@C-773 and Fe₃O₄@C-873 exhibited two bands, centering at approximately 1343 cm⁻¹ and 1589 cm⁻¹, attributed to the D-band and G-band, respectively [47]. According to the references [17,48], the D-band was the characteristic peak of the disordered amorphous structure of carbon, and the G-band was attributed to the sp2 hybridization of carbon materials. Observably, the intensity of the D-band decreased with the increased annealing temperature of the Fe₃O₄@C composites. For Fe₃O₄@C-773, the larger value of I_D/I_G (>1) implied a higher density of defects and structural disorder in the amorphous carbon [49]. On the contrary, the sp2 hybridization was enhanced accompanied by the reduction of defects and structural disorder for Fe₃O₄@C-873 (I_D/I_G < 1). This confirmed that the carbon defects and structural disorder were eliminated partially by the crystallization of the amorphous carbon. The defects and structural disorder would lower the electron transfer efficiency, although they benefited from the infiltration of the electrolyte. For further investigation of the microstructure, the isothermal N₂-adsorption/desorption at 77 K was carried out to evaluate the BET surface area and pore structure. For Fe₃O₄@C-873, the BET surface area was approximately 51.4 m²/g (Figure 3b), and there were abundant mesoporous structures with a main pore size of approximately 3 nm, together with some that were 10 nm, as illustrated in the insert of Figure 3b, which could provide active sites on the surface for the adsorption of the electrolyte [50]. However, the BET surface area of Fe₃O₄@C-773 (Figure 3c) was only 22.4 m²/g, far lower than the value of Fe₃O₄@C-873, and there was nearly no pore structure, possibly due to its leaves-like nanosheet structure. Certainly, the large surface and pore structure is more favorable to the enhancement of electrolyte entrance and infiltration and, especially, beneficial for Li/Li⁺ diffusion [51].

The surface element valence and chemical composition of the Fe₃O₄@C nanocomposites were revealed by XPS analysis. Figure 4 shows the XPS of the Fe₃O₄@C nanocomposites. Figure 4a shows the full XPS spectrum of Fe₃O₄@C-873, in which the existence of Fe, C and O species was verified by the Fe 2p, C 1s and O 1s characteristic peaks. In the fine spectra of Fe 2p (Figure 4b), the peaks centered at approximately 711 and 725 eV, attributed to the characteristic peaks of Fe 2p_3/2 and Fe 2p_1/2, which corresponded to Fe³⁺ and Fe²⁺, respectively [15,52]. This confirmed the existence of Fe₃O₄. The fine spectrum of C 1s contained four different peaks at 289.4, 287.0, 285.2 and 284.7 eV, corresponding to O-C=O, C=O, C-O, and C-C/C=C bonds, respectively [15,47,48], as seen in Figure 4c. The weak peaks derived from O-C=O, C=O and C-O bonds could be caused by the interactions of the C and O atoms located in the interface between Fe₃O₄ and carbon, which could be further confirmed by the fine spectrum of O 1s. In the high-resolution XPS spectra of O 1s (Figure 4d), the peak at 530.5 eV was attributed to the Fe-O bond in iron oxide, the other two peaks at 533.9 eV (O-C=O) and 532.0 eV (O-C/O=C) could be ascribed to the chemical bonds between carbon atoms and oxygen-containing groups on the surface of iron oxide [17,53,54]. The emergence of chemical interactions between C and Fe₃O₄ would facilitate the electron transfer between these two phases. For Fe₃O₄@C-773, the XPS analysis results are illustrated in Figure 4e–h. It was similar to Fe₃O₄@C-873, implying they had similar chemical composition and element valence states. Overall, the XPS further confirmed the preparation of Fe₃O₄@C nanocomposites and the existence of chemical interactions between C and Fe₃O₄.

Figure 5a shows the CV profiles for Fe₃O₄@C-873. Three cathodic peaks (1.63, 0.90 and 0.54 V) were observed in the lithiation process. The peak at 1.63 V could be attributed to the reaction of Li inserting into the lattice of Fe₃O₄ crystalline, which was associated with the reaction [15]:

Fe₃O₄ + xLi⁺ + xe⁻ = Li_x(Fe₃O₄) (0 ≤ x ≤ 2)

(1)

The peak located at 0.90 V could be corresponding to the reduction of Fe³⁺ and Fe²⁺ to Fe⁰ and the formation of Li₂O based on the reaction equation [55,56]:

Li_x(Fe₃O₄) + (8 − x)Li⁺ + (8 − x)e⁻ = 3Fe + 4Li₂O (0 ≤ x ≤ 2)

(2)

The peak of 0.54 V, which appeared only at the first cycle, was ascribed to the formation of the solid electrolyte interface (SEI) film that was stable in the following charge and discharge procedure [15]. For delithiation, the reduced Fe⁰ was gradually oxidized to Fe²⁺ and Fe³⁺, accompanied by the formation of Li_x(Fe₃O₄) by two anodic steps at 1.70 and 1.94 V, respectively [55,56,57,58]. After the first cycle, the CV curves and the position of the peaks almost overlapped in the second and third cycles, implying good charge and discharge reversibility for Fe₃O₄@C-873 as a LIBs anode. For Fe₃O₄@C-773 (Figure 5c), its CV curves were similar to those of Fe₃O₄@C-873. However, the cathodic peaks slightly shifted toward a higher voltage, indicating better kinetics, which was possibly related to the ultrafine Fe₃O₄ nanoparticles. Figure 5b depicts the discharge and charge curves of Fe₃O₄@C-873 at the current densities of 0.2, 0.5, 1.0, 2.0 and 5.0 A g⁻¹, respectively. For the profile of 0.2 A g⁻¹, there was one plateau at approximately 0.9 V in the discharge process, while there was one plateau containing two transitions at approximately 1.5 and 2.0 V in the charging process, which was consistent with two reaction peaks in the CV profiles; the two transitions corresponded to the oxidation of Fe to Fe²⁺ and Fe³⁺, respectively. The initial discharge capacity and charge capacity were approximately 986 and 760 mAh g⁻¹ at the current density of 0.2 A g⁻¹. The irreversible capacity loss could mainly result from the formation of SEI, the side reaction between the electrode and the electrolyte and the decomposition of the electrolyte [15]. Certainly, the reversible capacity reduced with the increase in discharge/charge current density, but it remained at a relatively high reversible capacity of 576 mAh g⁻¹ at 5.0 A g⁻¹. Unfortunately the discharge capacity and charge capacity at 0.2 A g⁻¹ were only 587 and 553 mAh g⁻¹, and it was further reduced to 291 mAh g⁻¹ when discharging at 5.0 A g⁻¹ for Fe₃O₄@C-773 (Figure 5d). That is to say, Fe₃O₄@C-873 had a higher capacity and superior rate performance compared to Fe₃O₄@C-773.

The Li⁺ storage performance of the Fe₃O₄@C electrodes was further systematically evaluated by the extended galvanostatic discharge and charge measurements, as depicted in Figure 6. Figure 6a–d show the cycling stability of Fe₃O₄@C-873 at different current densities. As observed in Figure 6a, the discharge capacity reached 1006 mAh g⁻¹ (efficiency: ~98%) from approximately 781 mAh g⁻¹ (efficiency: ~92%) after 100 cycles at 0.2 A g⁻¹. The maximum capacity was approximately 921 mAh g⁻¹ (efficiency: ~98%) at the 94th cycle and then gradually attenuated to 552 mAh g⁻¹ (efficiency: ~100%) after 300 cycles at 0.5 A g⁻¹ (Figure 6b). When the current density further increased to 1 A g⁻¹, it still had a reversible capacity of 423 mAh g⁻¹ and 100% efficiency after 500 cycles (Figure 6c). For Fe₃O₄@C-773 (Figure 6e–h), the discharge capacity increased to 662 mAh g⁻¹ (efficiency: ~98%) from 525 mA h g⁻¹ (efficiency: ~92%) at the 45th cycle and then returned to 526 mAh g⁻¹ (efficiency: ~98%) at the 100th cycle under a current density of 0.2 A g⁻¹ (Figure 6e). The maximum capacity was approximately 548 mAh g⁻¹ (efficiency: ~97%) at the 50th cycle and then reduced to 326 mAh g⁻¹ (efficiency: ~100%) after 300 cycles at 0.5 A g⁻¹ (Figure 6f). When the discharge/charge current density increased to 1 A g⁻¹, the reversible capacity attenuated to 258 mAh g⁻¹ (efficiency: ~100%) after continuous 500 cycles (Figure 6g). According to ref. [59], the capacity of pure Fe₃O₄ was only 126.8 mAh g⁻¹ after 100 cycles at 0.1 A g⁻¹. Significantly, the electrochemical properties and cycling stability of Fe₃O₄ were effectively improved by designing the unique structure Fe₃O₄@C nanocomposites. Figure 6d illustrates the rate performance of Fe₃O₄@C-873, which exhibited no capacity decay after 70 discharge and charge cycles in the current density range from 0.2 to 5 A g⁻¹. For example, the discharge capacity was 772 mAh g⁻¹ when undergoing 10 cycles at 0.2 A g⁻¹. It was approximately 580 mAh g⁻¹ after 50 cycles and the current density increased to 5 A g⁻¹; however, it could reach 978 mAh g⁻¹ at the 70th cycle when the current density reduced to 0.2 A g⁻¹ again. For Fe₃O₄@C-773 (Figure 6h), it had a similar rate performance; the corresponding capacity was approximately 531, 300 and 655 mAh g⁻¹, respectively. From the experiment results, we could see that two samples, especially Fe₃O₄@C-873, exhibited fairly good cycle stability and excellent rate performance. Certainly, it should be pointed out that Fe₃O₄@C-873 had a significantly higher capacity and better rate performance than Fe₃O₄@C-773. However, Fe₃O₄@C-773 more easily reached the maximum capacity with better kinetics (~50 cycles, less than ~94 cycles for Fe₃O₄@C-873), which could result from its large amount of defects and the superfine Fe₃O₄ nanoparticles, together with the leaves-like amorphous carbon structure, by enhancing the electrolyte absorption and shortening the Li/Li⁺ diffusion distance. For Fe₃O₄@C-873, the superior cycling stability and the outstanding high-rate performance were mainly attributed to the improved structural stability of nano-Fe₃O₄ for the robust core-shell structure supported by the nanotube/nanorod crystalline carbon.

For further understanding of the mechanism of the significantly enhanced electrochemical performances, the electrochemical impedance measurements were carried out, and the results are shown in Figure 7a. As depicted in Figure 7b, a corresponding equivalent circuit model was established to analyze the electrochemical impedance spectroscopy. In this model, R_s is the internal resistance of the analog battery (including the electrolyte impedance and the electrode impedance), R_ct represents the charge-transfer resistance, CPE₁ is related to the constant phase element, Z_w is the Warburg resistance associated with the Li/Li⁺ diffusion in the active materials [55,58,59,60,61]. From Figure 7a, it could be observed that the electrochemical impedance spectroscopies for two samples contained only one observable semicircle in the test full frequency range. It was attributed to the interfacial charge transfer (R_ct) combined with the double-layer capacitance CPE₁. However, Fe₃O₄@C-873 contained a much smaller semicircle than Fe₃O₄@C-773. The inclined lines in the low-frequency region were derived from the Li⁺ diffusion into the Fe₃O₄@C materials. The impedance parameters were achieved by fitting the experimental Nyquist plots (Figure 7a) using the established model in Figure 7b. For Fe₃O₄@C-773, the impedance parameters R_s and R_ct were achieved to be approximately 1.2 and 252.1 Ω, respectively. They were 4.0 and 103.0 Ω for Fe₃O₄@C-873. The results indicated that the Rs of Fe₃O₄@C-773 and Fe₃O₄@C-873 were very close, but the Rct of Fe₃O₄@C-773 was far higher than that of Fe₃O₄@C-873, indicating Fe₃O₄@C-873 had better charge transfer property. It could be attributed to the better structural integrity, such as fewer defects and good crystallinity for Fe₃O₄@C-873. The results suggested that the novel structure Fe₃O₄@C-873 nanocomposite had significantly improved electrochemical performances compared with Fe₃O₄.

4. Conclusions

In summary, leaves-like amorphous carbon-nanosheet-encapsulated ultrafine Fe₃O₄ nanoparticle composites and carbon nanotube/nanorod-supported core-shell structure Fe₃O₄ nanoparticle composites were fabricated via an in situ encapsulation route, which consists of PA milling the ferrocene (C₁₀H₁₀Fe) precursor and subsequent heating treatment. The carbon nanotube/nanorod-supported core-shell structure Fe₃O₄ nanocomposite has a larger specific area (51.4 m²/g), fewer defects and better crystallinity than the leaves-like amorphous carbon-nanosheet-encapsulated ultrafine Fe₃O₄ nanocomposite (22.4 m²/g), and the former also contains abundant mesoporous structures. The carbon nanotube/nanorod-supported core-shell structure Fe₃O₄ nanocomposite exhibits outstanding cycling stability and high-rate capability as anode materials for LIBs. Remarkably, it shows 1006 mAh g⁻¹ and retains 423 mAh g⁻¹ after 500 cycles at 1A g⁻¹. The improved electrochemical performances could be mainly attributed to the synergetic effects of the hierarchical architecture and the nanostructure. This work supplies a feasible route for the preparation of novel structure Fe₃O₄@C nanocomposites as high-performance electrode materials.