Highly mesoporous and chemically bonded Fe3O4/N-doped carbon nanocomposite with an outstanding cycling life as lithium-ion-battery anode

Owing to high theoretical capacity (926 mAh g−1), Fe3O4 has achieved much focus as a prospective anode material for lithium-ion batteries (LIBs). A one-step vapor-pressured induced approach considering the synthesis of chemically bonded Fe3O4/N-doped carbon nanocomposites (Fe3O4/NC) via Fe-O-C and Fe-N-C, together with the encapsulation of Fe3O4 nanoparticles (∼80 nm) into highly mesoporous N-doped carbon matrix via pyrolyzing the mixture of iron oxalate and dimethylformamide in a sealed vessel, does not exist at present. As LIB anode, the Fe3O4/NC presents a high capacity of 1250.2 mAh g−1 at 0.1 A g−1, an outstanding cyclability with a capacity of 600.1 mAh g−1 after 4000 cycles at 5 A g−1, and a high rate capability (244.8 mAh g−1 at 20 A g−1). Such excellent performances can be ascribed to its unique structure that Fe3O4 nanoparticles tightly encapsulated into highly mesoporous N-doped carbon matrix can increase active sites, electrical conductivity, and cyclability.


Introduction
Over the past decade improved techniques have revealed many advantages involving good cyclability, high specific energy, as well as high safe of lithium-ion batteries (LIBs) to allow them widely used in various electronics and vehicles [1,2]. Up to now graphite is still utilized as main LIB anode materials. Nevertheless, its relatively low theoretical capacity (372 mAh g -1 ) severely limits further development of higher-energy-density LIBs. At present, transition metal oxides (Fe 3 O 4 , Fe 2 O 3 , and GeO 2 , etc) have been invoked in many studies as important alternatives to graphite due to high capacity [3]. Among them, Fe 3 O 4 has focused public's attention [4]. However, challenges such as terrible volume change and bad electrical conductivity (EC), resulting in rapid capacity decay and poor ion transport rate, limit the commercial application. In order to address these problems, a series of carbon involving disorder carbon [5], carbon nanotubes [6], and graphene [7] as a buffer materials have been added into Fe 3 O 4 -based composite to suppress volume variation of Fe 3 O 4 nanoparticles, to inhibit Fe 3 O 4 particles from agglomeration on cycling, and to serve as a conductive network for boosting ion/electron  [8], nanospheres [9], nanotubes [10], nanoflakes [11], and nanobelts [12] have been designed by sol-gel route [13], electrostatic spinning [14], and solvothermal process [15], etc, to ameliorate volume expansion and EC of Fe 3 O 4 -based electrode. For instance, Fe 3 O 4 /C nanospheres exhibited a significantly enhanced discharge capacity of 930 mAh g -1 after 50 cycles at 0.1 A g -1 , in comparison with 316 mAh g -1 for bare Fe 3 O 4 nanospheres [16]. Additionally, a foam-like Fe 3 O 4 /C nanocomposite delivered a higher capacity of 1008 mAh g -1 than 540.8 mAh g -1 of bare Fe 3 O 4 nanoparticles after 400 deep cycles at 0.2 A g -1 [17]. A hierarchical Fe 3 O 4 /C hollow nanosphere fabricated via an aerosol-pyrolysis route achieved a high capacity of 383 mAh g -1 after 1000 cycles with no obvious capacity loss at 10 A g -1 [18]. The above results demonstrate that Fe 3 O 4 /C nanocomposites possess remarkable superiority as high-level LIBs anode materials. However, for practical application, a longer lifetime with thousands of cycles should be obtained, particularly at high current density. In addition, their preparation ways are extremely fussy and complicated to signal the great importance of a simple method for preparing Fe 3 O 4 /C nanocomposites.
Here, a one-step vapor-pressured induced synthesis route was proposed to prepare Fe 3 O 4 /N-doped carbon (Fe 3 O 4 /NC) nanocomposites by heating a mixture of ferric oxalate and dimethylformamide in a sealed vessel. In the nanocomposites, Fe 3 O 4 nanoparticles with an average size of 80 nm are uniformly and tightly dispersed in a highly mesoporous N-doped carbon matrix via chemical bonds of Fe-O-C and Fe-N-C. The small size of Fe 3 O 4 , the highly mesoporous N-doped carbon matrix, and the homogeneously distributed structure not only contribute to Li + diffusion and improve EC, but also inhibit aggregation and volume expansion of nanoparticles upon cycling. As LIB anode, Fe 3 O 4 /NC shows a high capacity, an outstanding cyclability, and an ascendant rate capability .

Preparation of Fe 3 O 4 /NC
Typically, ferric oxalate (0.5 g, Macklin) and dimethylformamide (0.5 g, Macklin) were placed in a vessel with a volume of 5 ml, and then sealed in a Ar-filled glove box. Subsequently, the vessel was heated at 500℃ in a tube furnace for 0.5 h at 10℃ min -1 under Ar atmosphere. After that, the tube furnace was cooled to room temperature, and the Fe 3 O 4 /NC sample was obtained.

Characterizations
X-ray diffraction (XRD) pattern was used to determine crystalline structures. Thermogravimetric analysis (TGA) was tested in air with a heating rate of 10 o C min -1 . Scanning electron microscopy (SEM, Hitachi 4800) was used to observe morphology. Transmission electron microscopy (TEM, TECNAI G2 F20) with energy dispersive spectroscopy (EDS) was used to test the microstructures. Raman spectroscopy was performed to test carbon material. X-ray photoelectron spectroscopy (XPS, ESCALAB 250) was operated to investigate the valence structure of elements and surface composite.
The EC of sample was tested by a Powder Electrical Resistivity Tester (ST-2722, Suzhou Jingge Electronic Co., Ltd., China). Specific surface area and pore size distribution were tested using an Accelerated Surface Area and Porosimetry System (Micromeritics ASAP 2010).    [15,18], and two characteristic peaks of D and G band of carbon materials at 1345.4 and 1593.7 cm -1 , respectively [19]. The I D /I G value of 0.92 proves partial graphitization of the carbon in the composite, which suggests that the carbon matrix are beneficial for the improvement of EC and Li + storage capacity [19]. As tested, the value of EC is 1.2x10 3 S/m. Such a high value can boost Li + transport to obtain superior rate performances. XPS clearly presents that the dominating elements of the   [19,20]. The appearance of Fe-N bond proves the presence of Fe-N-C bonds. The presence of Fe-O-C and Fe-N-C bonds signals that the Fe 3 O 4 nanoparticles are firmly dispersed into the carbon matrix, benefiting for inhibiting aggregation and adapting to enormous volume expansion of Fe 3 O 4 during lithiation. The pyridic N and Fe-O/N-C bonds were highly chemically active sites and can make vast defects and accommodate more Li + to the improvement of the capacity [19,20]. Besides, the Fe 3 O 4 /NC nanocomposite has a mesoporous structure with a surface area of 128.4 m 2 g -1 , a pore volume of 0.78 cm 3 g -1 , and a well-defined mesopore centered at 2.4 nm (Figure 2h), which can obtain a high contact area with electrolyte, boost ion transport, store Li + , and alleviate volume expansion of Fe 3 O 4 during cycling, thus comprehensively enhancing lithium storage performances of the Fe 3 O 4 /NC nanocomposite.  [11,13]. In the anodic scanning, two peaks at 1.64 and 1.83 V are ascribed to gradual oxidation of Fe 0 to Fe 2+ and then to Fe 3+ [15,18]. Moreover, all peaks overlap after the first scanning to prove high-level reversibility of Fe 3 O 4 /NC. The charge/discharge curves (Figure 3b) display homologous voltage plateaus with CV curves. Besides, Fe 3 O 4 /NC reveals first lithiation and delithiation capacities of 1736.4 and 1250.2 mAh g -1 , respectively, to produce a coulombic efficiency (CE) of 72.0%. Such a low CE mostly originates from the formation of SEI and Li 2 O [21,22]. The Nyquist plots at various cycling number are represented (Figure 3c), in which the value of charge transfer resistance (R ct , the diameter of depressed semicircle) originally rises after the 1st cycle owing to the construction of SEI film, and then gradually diminishes, presumably arising from activation of electrode material. The decrease of R ct is profitable for strengthening capacity and stability [21,22]. Subsequently, the cycling life at 1 A g -1 of Fe 3 O 4 /NC is assessed, which delivers a first reversible capacity of 985.2 mAh g -1 with a CE of 71.3%, and 896.1 mAh g -1 after 800 cycles with a CE of 99.82% (Figure 3d). So high capacity retention of 91.1% is gained to fully testify a good cyclability of the Fe 3 O 4 /NC. An ultralong cycling lifetime is acquired at 5 A g -1 (Figure 3e), before which the cell is activated at 0.1 A g -1 for three cycles, in which a high capacity of 600.1 mAh g -1 after 4000 cycles with a high CE of 98.9% and 89.6% capacity retention is revealed. Such a low capacity loss per cycle of 0.0026% sufficiently demonstrates a distinguished stability of electrode materials during large-current cycle. The rate capability of Fe 3 O 4 /NC is tested at 0.1, 0.2, 0.5, 1, 2, 5, 10, and 20 A g -1 (Figure 3f) primarily are contributed from its unique structure, such as smaller particle size of Fe 3 O 4 , highly mesoporous N-doped carbon matrix, and the formation of interfacial chemical bonds of Fe-O-C and Fe-N-C. To expounds ion storage mechanism and concrete capacity contribution of the Fe 3 O 4 /NC, the electrochemical kinetics and pseudocapacitive behaviors are surveyed by CV curves at various scanning rates (v). The CV (Figure 4a) curves display homologous Li + storage behavior at v of 0.1 to 2 mV s -1 to corroborate a stable pseudocapacitive behavior. In general, peak current (i) and v is directly related to the Eq. logi=blogv+loga. It can be observed that the b is the slope of logi-logv plots, where 1.0 suggests a capacitive-controlled process and 0.5 indicates a diffusion-dominated process. Based on the calculation results, b values of all the peaks range from 0.86 to 0.92 (Figure 4b), exhibiting that both diffusion-controlled and pseudocapacitive-controlled process contribute the total capacity [23,24]. To survey their particular contribution, Eq.i=k 1 v+k 2 v 0.5 is presented, where i composes of pseudocapacitive process (k 1 v) and diffusion-controlled behaviors (k 2 v 0.5 ). Obviously, the percentage of pseudocapacitive contribution raising gradually from 43.1 to 78.5% with a rise of v from 0.1 to 2 mV s -1 (Figure 4c), demonstrating high influence of pseudocapacitive ion storage on the enhancement of the capacity and rate performances. So high pseudocapacitive contribution primarily originates from extra Li + storage sites, such as interfaces, defects, and mesoporous structure [23,24].

Conclusions
To conclude, a one-step vapor-pressured induced synthesis route is developed for synthesizing Fe 3 O 4 /NC nanocomposite for the first time by pyrolysis of mixture of ferric oxalate and dimethylformamide in a sealed vessel. The structural advantages, involving small size of 80 nm, highly mesoporous N-doped carbon matrix, and Fe-N-C/Fe-O-C bonds, of the hybrid nanocomposites induce a high pseudocapacitive contribution, a rapid ion transport route, and a highly stable structure, thus bringing about excellent electrochemical performances. As LIB anode, Fe 3 O 4 /NC shows a high reversible capacity (1250.2 mAh g -1 at 0.1 A g -1 ), an outstanding cyclability (up to 4000 cycles at 5 A g -1 ), and rate capability (244.8 mAh g -1 ).