Microwave-assisted Hydrothermal Synthesis of Magnetite Nanoparticles with Potential Use as Anode in Lithium Ion Batteries

*e-mail: mauricio.bomio@ct.ufrn.br


Introduction
Magnetite Fe 3 O 4 is one of the most fascinating natural ore for technological and scientific reasons 1 .It is a magnetic iron oxide that presents an inverse spinel structure with face-centered cubic (FCC) unit cell, formed by oxide anions and iron cations occupying the tetrahedral and octahedral sites inside the structure 2 .The presence of iron ions in both divalent and trivalent oxidation states, provides several different applications of magnetite materials.The magnetic properties of magnetite nanoparticles have been intensively studied due to its relevance to magnetic recording and biomedical applications [3][4][5] .
The large demand for energy storage devices has encouraged studies on rechargeable Li-ions batteries (LIBs).They have been widely used as high power source for several portable electronic devices and electric vehicles [6][7][8] .The research efforts have push up the interest in developing new electrode materials with high capacity and cycling stability for a new generation of lithium ion batteries.Numerous binary M x O y compounds have been studied as low voltage working electrodes versus lithium 8,9 with the challenge of overcoming the limited capacity provided by graphite.Among these potential binary compounds, Fe 3 O 4 deserves great attention for its high theoretical capacity (926 mAhg -1 ), low cost, being environmental-friendly and naturally abundant worldwide [10][11][12] .
Several reports have proposed that conversion reactions in lithium ions batteries are affected by composition, particle size and morphology, which in turn are directly related to the preparation method of active materials 13,14 .The development and implementation of nanostructured materials has led to great improvements in Li-ion battery performance, since size reduction leads to improvements in the Li-ion battery intercalation capability by increasing the specific surface area for interfacial Faradaic reactions and the Li + flux across the electrode-electrolyte interface; these effects ultimately enhance the mass and charge diffusion paths and modify the thermodynamics (compared to bulk), which facilitates phase transitions 15 .Hydrothermal techniques are suitable for the synthesis of ultrafine materials with homogeneous composition and adequate morphology for use in multiple technological purposes.Microwave hydrothermal method (MH) combines the advantages of both hydrothermal and microwave-irradiation techniques such as very short reaction time, production of small particles with a narrow size distribution and high purity 16 .This method have been extensively used for the preparation of pure phase of several material with different applications such as, copper oxide (CuO) 17,18 , zinc oxide (ZnO) [19][20][21] hafnium oxide (HfO 2 ) 22 and cerium oxide (CeO 2 ) 23,24 .
In this paper, we report a simple strategy to synthesize magnetite nanoparticles (Fe 3 O 4 ) by microwave hydrothermal method in a short time of processing without further treatment.The material obtained was tested as anode active material for lithium ions batteries and a high reversible capacity of 768.5 mAhg -1 at 1C over 50 cycles was demonstrated, suggesting its prospective use as anode material for high power lithium ion batteries.

Results and Discussion
Figure 1 shows the XRD patterns of the magnetite samples processed under microwave-assisted hydrothermal conditions for 8 and 15 min.Cubic Fe 3 O 4 (JCPDS No. 19-629) was present in all samples.No secondary phase peaks were found in samples.
Figure 2 presents HRTEM and TEM images of Fe 3 O 4 nanoparticles processed under microwave-assisted hydrothermal conditions for 15 min.The TEM image in Figures 2a-b indicates that the samples are mainly composed of nanoparticles with average size of 25 nm.The analysis of the expanded HRTEM image inset in Figure 2c shows that the distance between neighboring planes is about 0.25 nm, related to the (311) crystallographic plane of cubic Fe 3 O 4 .
In the initial MH processing stages, the microwave radiation is able to promote localized superheating in the aqueous solution as well as accelerate the solid particles to high velocities, leading to an increase of the interparticle collisions and inducing effective fusion at the point of collision.Moreover, this kind of electromagnetic energy can induce a uniform heating inside the primary particles formed after nucleation stage.In principle, all these effects caused by the microwave radiation favor the formation of aggregated particles with irregular shapes 25 .
The lithium storage mechanism of iron oxides here explored is based on redox conversion reaction.Iron oxides are completely reduced into mettalic Fe nanocrystals dispersed into the Li 2 O matrix upon lithiation and then reversibly restored to their initial oxidation states during delithiation 26 .Figure 3 exhibits the voltage profiles of the first cycle of magnetite samples processed under microwaveassisted hydrothermal conditions for 8 and 15 min.During the first discharge of Fe 3 O 4 nanoparticles, the cell voltage decreases steeply to ca. 1.5 V, being then followed by a small slope.When the voltage continues to decrease to ca. 1.0 V, a short plateau appears, which delivers capacity of ca. 100 mAhg -1 .Then, the voltage continues to decrease to ca. 0.8 V until the appearance of an obvious large voltage plateau for ca.800 mAhg -1 .Finally, a gradual voltage decrease is observed up to the cut-off value.Particularly, the short plateau that appears at 1.0 V is due to lithium intercalation into the Fe 3 O 4 framework before the conversion reaction proceed.The large discharge plateau around 0.8 V appears in both Fe 3 O 4 samples, which is in accordance with the reduction of Fe 3 O 4 by lithium to the metallic state 27 .The gradual potential decay from ca. 0.8 V to the cut-off voltage has been ascribed to the reversible decomposition of the electrolyte and the partial formation of SEI layer on the surface of the Fe+Li 2 O matrix 27 .
Besides the high capacity, rate capability is also important for the high performance of LIBs. Figure 4 shows the voltage profiles of Fe 3 O 4 nanoparticles cycled at current rates of 1 and 2 C. The capacity of Fe 3 O 4 nanoparticles on cycling showed two types of variation at current rate of 1C:  Microwave-assisted Hydrothermal Synthesis of Magnetite Nanoparticles with Potential Use as Anode in Lithium Ion Batteries for samples processed for 8 min, the reversible capacity presented a slight increase up to the 9 th cycle followed by a continuous decay, whereas for samples processed for 15 min the reversible capacity remained constant up to the 10 th cycle followed by a continuous decay up to the last cycle.
As the current rate increases to 2C the charge/discharge performance for sample processed for 8 min showed no significant change, while for those processed for 15 min a decrease in performance was observed.
In general, samples processed for 8 min were less influenced by the increased kinetic rate.Undoubtedly, their low particle size is likely to be directly related to this behavior.While for samples processed for 15 min, the charge/discharge performance decreases with increasing kinetic rate, probably due to the kinetic hindering, caused by the growth of bulk particles during hydrothermal processing.
Comparing the performance of Fe 3 O 4 nanoparticles with other reported ferrite-based anodes, shown in Table 1, it can be seen that the values for the first discharge capacity and the capacity retention after 50 cycles are higher for our samples than for most of other Fe 3 O 4 based anodes.These results indicate that Fe 3 O 4 nanoparticles presented in this study can be considered as a good choice of anodes for lithium ions batteries.
The internal cell resistance usually affects the electrochemical performance of electrode materials upon cycling.These parameters can be evaluated from the impedance spectra of electrodes.Figure 5 shows the Nyquist plots for cells assembled with the studied samples and after one, five and ten discharges.The profiles reveal two semicircles at high and intermediate frequencies.These features are commonly ascribed to the lithium migration through the surface layer and charge transfer reactions.The depression of semicircles reveals a frequency dispersion that must be described by a constant-phase element (CPE).The fitting of spectra according to the equivalent circuit in Figure 5b allowed calculating the resistance values associated to the processes mentioned above (Table 2).The nearly linear-type frequency variation observed at low frequencies is fitted by introducing a Warburg component (W) describing the system response to the Li + diffusion through the iron oxide.
The results shown in Table 2 reveal that similar RSL values were recorded for both samples after the first discharge.On cycling, these values slightly increased.The most remarkable differences were observed for R CT values.Thus, the sample treated for 15 minutes showed significantly lower resistance values.These results could be explained by a lower content of surface defects which are removed during thermal treatment.Therefore, the differences in cell performance on cycling observed between samples cannot be strictly correlated to cell resistance.Most probably, other factors such as the crystallinity of pristine electrodes or the stability of the polymeric layer are essential to describe this behavior.

Conclusion
The microwave-assisted hydrothermal method has demonstrated to be an efficient method for the synthesis of magnetite nanoparticles.XRD measurements revealed the presence of cubic Fe 3 O 4 for the synthesized samples.Impurity phases were not observed.TEM images showed that these samples consist of nanoparticles with average size of 25 nm.When tested as the anode in lithium batteries, Fe 3 O 4 nanoparticles exhibited a high reversible capacity, excellent cycling performance, and good capability rate.According to impedance spectroscopy, the differences in cell performance on cycling observed between samples can be due to the crystallinity of pristine electrodes or the stability of the polymeric layer.The results reported here are of interest with respect to possible applications.

Experimental Section
The chemical materials used here were all commercially available, and all of the chemical reagents were of analytical grade and used in this experiment without any further purification.
The typical procedure employed in order to obtain Fe 3 O 4 nanoparticles was precipitation of the material and further microwave hydrothermal processing.In a becker, FeSO 4 .7 H 2 O was dissolved in deionized water with finally concentration 0.04 mol.L -1 and stirring for 15 minutes.Then, the previous solution was neutralized by the addition of mineralizing agent ammonium hydroxide with concentration of 1 M, then, the black precipitated was obtained.These resulting materials were sealed in a 100 ml Teflon lined autoclave and treated by microwave hydrothermal procedure for 8 and 15 minutes at 100 o C. The black products were washed several times with deionized water and separated by centrifugation until total removal of the mineralizing agent and dried in conventional furnace at 60 °C for some hours.
The Fe 3 O 4 powders were structurally characterized by X-ray diffraction (XRD) patterns using a DMax/2500PC diffractometer (Rigaku, Japan) with Cu Kα radiation (λ = 1.5406Å) in the 2θ range from 20 o to 80 o with scanning rate of 0.2 o /s.The morphologies were investigated through a FEG-SEM of Carl Zeiss, model Supra 35-VP (Germany), operated at 6 kV and a transmission electron microscope (TEM), model CM200 (Philips, USA), operated at 200 kV.In the preparation of TEM samples, the obtained powders were first dispersed in acetone using ultrasonic bath for 20 minutes.Then, the suspensions were deposited on copper grids via fast immersion.
Electrochemical tests were performed in two and three electrodes Swagelok-type cells were assembled inside a MBraun LabMaster 130 glove box under argon atmosphere with less than 2 ppm of H 2 O and O 2 .Lithium metal disk of 9 mm in diameter, was used as auxiliary and reference.The electrolyte solution (1 M LiPF 6 in 1:1/EC:DEC) was supported by two porous glass-paper (GF/A-Whatman) disks also acting as separator.Electrodes containing the studied samples were fabricated from a mixture containing 75% active material, 10% carbon black, 10% graphite, and 5% polyvinylidene fluoride (PVDF).A paste was prepared by soaking this mixture in -methyl-2-pyrrolidone (NMP) which is eventually spread onto a 9-mm diameter copper foil (Good fellow 99.999%).The electrode was dried at 120 °C and 7 mbar for 4 h and pressed under 1 ton load to ensure uniform surface.Galvanostatic battery testing was carried out in two Li/1M LiPF 6 (EC:DEC)/Fe 3 O 4 electrodes.Current control and cell potential measurements during charge and discharge were carried out with a multichannel galvanostat.This equipment controls current conditions during charging

Figure 2 .
Figure 2. TEM (a-b) and HRTEM (c) images of Fe 3 O 4 nanoparticles processed under microwave-assisted hydrothermal conditions for 15 min.The inset shows crystallographic planes (c).

Figure 4 .
Figure 4. Discharge/charge cycling performance at different current densities of Fe 3 O 4 nanoparticles.

Figure 5 .
Figure 5. (a) Nyquist plots for lithium cells assembled with samples treated for 8 and 15 minutes and measured after 1 st , 5 th and 10 th discharge.(b) Equivalent circuits used for the fitting of the spectra.

Table 2 .
Surface layer and charge transfer resistance values calculated from impedance spectra of lithium cells assembled with samples treated for 8 and 15 minutes.Cells were measured after one, five and ten discharges.

Table 1 .
Comparison of published literature data with that obtained in this work.