Introduction

The future of renewable energies and electric vehicles (EVs) are tightly correlated to further progress in Li-ion batteries [1]. Along with commercialized LiCoO2 other cathode materials such as LiMn2O4 and LiFePO4 have found their ways toward practical application [2,3,4]. In contrast, graphite remains mainly used as anode material due to its low working potential, cost-effectiveness, and good cycling stability [5]. However, this anode cannot meet the high power/energy density requirements for renewable energies and EVs owing to its low theoretical capacity (372 mAh g–1) [6]. Therefore, advanced anode materials with superior electrochemical properties are urgently demanded to be designed and synthesized.

Among various possible materials, iron oxide (Fe3O4) is considered to be a promising anode material for next-generation Li-ion batteries due to its high theoretical capacity (924 mAh g–1), abundance in the Earth's crust, and environmental friendliness [7, 8]. However, bulk Fe3O4 particles suffer from rapid capacity fading owing to their low electrical conductivity and colossal volume change during lithiation/delithiation processes [9, 10]. Two main approaches have been successfully applied to overcome these issues, such as a synthesis of nanostructured Fe3O4 particles and its integration with carbon materials [11,12,13,14,15]. However, Li-ion battery application requires these composite materials to possess an appropriate morphology that ensures a high contact area between Fe3O4 nanoparticles and electrolytes; and to be prepared by simple and cost-effective methods.

Thus far, Fe3O4 dispersed/decorated carbon materials have been prepared mainly by hydrothermal/solvothermal [16, 17], co-precipitation [18, 19], and electrospinning [20, 21] methods. Although these composite materials demonstrated an improved electrochemical performance when applied as anode materials for Li-ion batteries, the preparation of these materials rely on the complicated and time-consuming procedures, and require expensive toxic reagents. Compared with these methods, spray pyrolysis (SP) can provide a simple, short processing, and continuous synthesis route for multifunctional materials with homogeneous chemical composition and diverse structures [22,23,24]. Moreover, an easy variation of a starting solution can offer numerous opportunities to synthesize various multicomponent composite materials with excellent chemical and physical flexibility. Despite many advantages of SP, there have been very few reports on a direct synthesis of C/Fe3O4 composite materials by SP [25,26,27]. For example, Choi et al. reported a macroporous Fe3O4/carbon composite synthesis by SP with polymer additives as anode materials for Li-ion batteries [26]. As a cost-effective alternative for the polymer additives, Atkinson et al. [27] used inorganic salt additives such as sodium chloride (NaCl) and sodium nitrate (NaNO3) to produce iron-impregnated porous carbon (Fe–C) spheres for catalytic application. However, to the best of our knowledge, the effect of NaNO3 additive on a phase conversion of Fe3O4 nanoparticles, morphology evolution of C/Fe3O4 microspheres, and their potential application in lithium-ion batteries have not been reported.

Recently, we reported a synthesis of porous-crystalline C/Fe3O4 microspheres by spray pyrolysis with post oxidation under steam atmosphere [28]. Here we chemically activated C/Fe3O4 microspheres with a sodium nitrate (NaNO3) additive in the precursor solution to ensure a high contact area between Fe3O4 nanoparticles and electrolyte. This approach provides a facile and cost-effective way to prepare porous C/Fe3O4 microspheres. We investigated the effect of NaNO3 additive on a phase conversion of Fe3O4 nanoparticles, morphology evolution of C/Fe3O4 microspheres, and their electrochemical performances as an anode material for Li-ion batteries in detail by various physical and electrochemical characterization techniques.

Experimental section

Synthesis of porous C/Fe3O4 microspheres

All analytically pure chemicals were purchased from FUJIFILM Wako Pure Chem. Corp., Japan, and used as-is. The precursor solutions were prepared by dissolving 0.50 M sucrose and 0.015 M iron (III) nitrate nonahydrate (FeNO3)3・9H2O) in distilled water at room temperature. Sodium nitrate (NaNO3) was then added to the solution at different concentrations ranging from 0 to 2.00 M. The solutions were magnetically stirred to dissolve the reagents completely.

A schematic diagram of the spray pyrolysis setup has been described elsewhere [24]. It consists of an ultrasonic nebulizer (1.7 MHz, Omron Co., Ltd., Model NE–U17), a laminar flow aerosol reactor (a high-quality ceramic tube with an inner diameter of 20 mm and a length of 150 cm), and an electrostatic precipitator. The nebulizer continuously atomized the starting solution at a frequency of 1.7 MHz. The generated droplets were carried into the pre-heated reactor tube (700ºC) by high-purity nitrogen gas (1.0 L min–1). The electrostatic precipitator collected the particles at 180ºC to prevent vapor condensation on the particles, while a cold trap was set to condense evolved gases. The samples prepared with NaNO3 additive were washed several times with a distilled water and ethanol mixture to remove the remaining sodium salts (Fig. S1a) and dried at 110ºC for 6 h in a vacuum oven.

Physical characterization

The crystal phase of C/Fe3O4 microspheres was identified using an X-ray (Rigaku, Ultima IV with D/teX Ultra) diffractometer with monochromated Cu Kα radiation. The morphology of the samples was observed by SEM (KEYENCE, VE-9800SP). The measurements and the calculation of the particle size distribution, geometric mean diameter dp,g, and geometric standard deviation σg have been described elsewhere [29]. The microstructure and elemental distributions of the samples were characterized by field-emission TEM (FE-TEM, JEOL Ltd., JEM2010F) at 200 kV with EDS (Genesis).

The composition of the samples was quantified by a CHNS elemental analyzer (Elementar, Vario Micro Cube), and thermogravimetry and differential thermal analysis (TG-DTA, Rigaku TG8120). The carbon structure was analyzed by Raman spectroscopy (NRS–2100, JASCO Co.). The XPS (JEOL Ltd., JPS-9010) was performed to clarify the chemical nature and valence state of the elements on the surface of samples.

The nitrogen adsorption-desorption isotherms were obtained using a Micrometrics Tristar II 3020 system to calculate the specific surface area and pore volumes of the samples. The specific surface area was calculated by the Bruneaur–Emment–Teller (BET) method. Total pore volume and micropore volume were evaluated using the Gurvich rule and t-plot method, respectively.

Electrochemical characterization

The electrochemical properties of C/Fe3O4 microspheres were evaluated in CR2032 coin-type cells with lithium metal as reference and counter electrodes. To prepare the working electrodes, 75 wt.% active materials, 15 wt.% acetylene black (AB), and 10 wt.% polyvinylidene fluoride (PVDF) were mixed in N-methyl-2-pyrrolidone (NMP) to obtain the slurry. The slurry was then coated onto a Cu foil current collector by doctor blade technique. The coated electrodes were dried at 110ºC for 5 h in a vacuum oven, roll-pressed to enhance the contact between the electrode material and the current collector, and punched into round disks of 15.9 mm in diameter. The area of the electrode material was then standardized into 1 × 1 cm2 by scratching. The mass loading of the active materials was about 1.0 ± 0.1 mg cm–2. The CR2032 coin-type cells were assembled in a high-purity-argon-filled glove box. A commercially available electrolyte of 1.0 mol L–1 lithium hexafluorophosphate (LiPF6) dissolved in a mixture of ethylene carbonate (EC) and dimethyl carbonate (DMC) was used. The separators were kept in the electrolyte for a couple of weeks before use to let the separator thoroughly soak the electrolyte. The assembled cells were galvanostatically cycled in the potential window of 0.01–3.00 V (vs. Li/Li+) at a current density of 50 mA g–1 using multichannel battery testers (Hokuto Denko, HJ101mSM8A). The rate performance tests were conducted at different current densities from 50 to 2000 mA g–1. The electrode material's current density and specific capacity were calculated based on the mass of active materials. Cyclic voltammetry (CV) analysis was performed in the potential window of 0–3.00 V (vs. Li/Li+) at a scan rate of 0.1 mV s–1. Electrochemical impedance spectroscopy (EIS) measurements were conducted in the range of 100 kHz to 0.1 Hz using a Solartron 1255B frequency response analyzer connected to a Solartron SI 1287 electrochemical interface. All electrochemical measurements were performed at room temperature.

Results and discussion

Synthesis and physical characterization of porous C/Fe3O4 microspheres

Figure 1 shows the XRD patterns of C/Fe3O4 microparticles prepared by SP with different concentrations of NaNO3 additive. The sample prepared without additive had XRD peaks corresponding to amorphous carbon and Fe3O4 without impurity [30,31,32]. The sucrose decomposition products partially reduced Fe3+ ions into Fe2+ ions during the synthesis, producing Fe3O4 instead of Fe2O3 from the single Fe3+ precursor solution [33]. The peaks of Fe3O4 and carbon remained in the XRD patterns of the samples prepared with 0.50 and 1.00 M NaNO3 additive after the aqueous removal of sodium salts. However, the sample prepared with 2.00 M NaNO3 additive had an additional peak of the impurity phase. The XRD peaks of carbon and Fe3O4 intensified in the samples prepared with NaNO3 additive due to the enhanced carbonization of sucrose and crystallization of Fe3O4, respectively.

Fig. 1
figure 1

XRD patterns of C/Fe3O4 microspheres prepared by SP with different concentrations of NaNO3 additive

Full size image

Figure 2 shows the SEM images of C/Fe3O4 microparticles prepared by SP with different concentrations of NaNO3 additive. The composite prepared without additive had a spherical morphology with a smooth surface (Fig. 2a). On the other hand, the presence of NaNO3 additive in the precursor solution significantly affected the morphology of the final microparticles. For example, the synthesis with 0.50 M NaNO3 additive produced sheet-like microparticles and spherical microparticles (Fig. 2b), but an increase in the additive concentration to 1.00 M formed mainly spherical microparticles (Fig. 2c). Such morphology evolution was attributed to a supersaturation degree of aerosol droplets, determining whether the precipitation occurs on the surface or throughout the entire droplet [34, 35]. A further increase in the concentration of NaNO3 (2.00 M) led to the collapse of microstructure (Fig. 2d). Moreover, the presence of NaNO3 in the precursor solution enlarged the geometric mean diameter of C/Fe3O4 microspheres from 0.47 to 1.11 μm (Fig. S2) owing to the increase in the amount of carbonized sucrose (Table 1). The content of carbon and Fe3O4 in the composites prepared with 1.00 M NaNO3 was further confirmed by TG/DTA analysis, as shown in Fig. S3. The prepared microspheres had ⁓20 wt.% of Fe3O4 calculated based on the weight of Fe2O3 as the final oxidation product. As this content of Fe3O4 is relatively low compared with other reported composites, the spray pyrolysis synthesis was performed with a higher concentration of iron source (0.06 M) in the precursor solution to increase the content of Fe3O4. However, the impurity peaks of FeO appeared in the XRD pattern of this sample (Fig. S4). Therefore, it is not easy to simultaneously increase the content of Fe3O4 and ensure its purity.

Fig. 2
figure 2

SEM images of C/Fe3O4 microspheres prepared by SP without (a) and with 0.50 M (b), 1.00 M (c), and 2.00 M (d) NaNO3 additive; TEM and TEM-EDX images of C/Fe3O4 microspheres prepared without (e, g) and with (f, h) 1.00 M NaNO3 additive

Full size image
Table 1 CHNS results of C/Fe3O4 microspheres prepared SP with different concentrations of NaNO3 additive
Full size table

Figure 2e and f show the TEM images of C/Fe3O4 microspheres prepared by SP without and with 1.00 M NaNO3 additive, respectively. The sample prepared without additive had ultrafine-sized Fe3O4 nanoparticles densely dispersed in the carbon matrix (Fig. 2e). The sucrose and its decomposition products sterically hindered the aggregation and growth of iron oxide nucleus [33]. On the other hand, the TEM images of the sample prepared with 1.00 M NaNO3 additive showed larger Fe3O4 nanoparticles (20–30 nm in diameter) evenly dispersed mainly on the surface of the carbon microsphere (Fig. 2f). The molten NaNO3 may increase the surface energy of Fe3O4 nanoparticles, thus leading to their diffusion and crystal growth. The corresponding elemental mapping shows the presence of carbon (C), iron (Fe), and oxygen (O) elements in both samples, further verifying the formation of carbon microspheres with evenly dispersed Fe3O4 nanoparticles (Fig. 2g, h).

Figure 3a shows the nitrogen adsorption-desorption isotherms of C/Fe3O4 microspheres prepared by SP with different concentrations of NaNO3 additive. The microspheres prepared with 0.50 and 1.00 M NaNO3 additive had an increased nitrogen uptake at a low relative pressure region (P/P0 = 0.01), which originates from the micropore filling. In addition, the former sample shows a combination of type I and II or III isotherms, suggesting the presence of micro-, meso-, and macropores. On the other hand, the sample prepared with 1.00 M NaNO3 additive had a hysteresis loop at a high relative pressure region (P/P0 = 0.5–1.0), corresponding to type IV isotherms with H3 hysteresis [36]. Such isotherms originate from the capillary condensation in the mesopores and suggest the presence of mainly mesopores along with micropores in this sample. However, a further increase in the concentration of NaNO3 (2.00 M) led to the collapse of microstructure. The BET-surface area and pore volumes of C/Fe3O4 microspheres are summarized in Table 2. The use of NaNO3 additive produced porous C/Fe3O4 microparticles with higher specific surface areas and pore volumes. Accordingly, the sample prepared with 1.00 M NaNO3 additive had the largest specific surface area of 704 m2 g-1 and total pore volume of 1.00 cm3 g-1.

Fig. 3
figure 3

N2 sorption isotherms (a), Raman spectra (b), wide-scan XPS spectra (c), and N 1s XPS spectra of C/Fe3O4 microspheres prepared by SP with different concentrations of NaNO3 additive

Full size image
Table 2 BET surface area and pore volume of C/Fe3O4 microspheres prepared by SP with different concentrations of NaNO3 additive
Full size table

Figure 3b shows the Raman spectra of C/Fe3O4 microspheres prepared with and without 1.00 M NaNO3 additive. The spectra of both samples had two typical peaks at ~1358 and ~1591 cm–1, corresponding to D and G bands of carbon, respectively. The D band was Raman active A1g mode associated with the vibration on sp3–bonded disordered carbon atoms. The G band was Raman active E2g mode associated with the vibration on all sp2–bonded ordered carbon atoms in a graphene layer [37, 38]. The peak intensity ratio of these peaks (ID/IG) reflects a graphitization degree of carbon in the microspheres. The ratio of ID/IG was lower in the sample prepared without the NaNO3 additive (ID/IG=0.78) than the one prepared with the additive (ID/IG=0.95), indicating that the carbon had a lower degree of graphitization in the sample prepared with NaNO3 additive. According to the above results, the spray pyrolysis without NaNO3 additive produced a more uniform and dense carbon. On the other hand, the synthesis with NaNO3 additive produced wide-range of pores, which might be responsible for the lower degree of graphitization in this sample [39].

Figure 3c shows the wide-scan XPS spectra of C/Fe3O4 microspheres prepared with and without 1.00 M NaNO3 additive. The spectra further confirmed the presence of C 1s, N 1s, O 1s, and Fe 2p in the prepared samples. The N 1s indicated a formation of N-doped carbon, which was consistent with CHNS results (Table 1). This spectrum was deconvoluted into three typical bonds of nitrogen: graphitic, pyrrolic, and pyridinic (Fig. 3d). These nitrogen bonds can induce the electronic structure of carbon and promote its electronic conductivity [40].

The formation mechanism of porous C/Fe3O4 microspheres prepared by SP with NaNO3 additive is graphically shown in Fig. 4. In the case of the SP synthesis without NaNO3 additive (Fig. 4a), the ultrasonic nebulizer first generates aerosol droplets that contain Fe3+ ions and partially hydrolyzed sucrose molecules. Hydrolysis of sucrose produces glucose and fructose molecules in an acidic solution (1) [33]. When the droplet enters the reactor, it starts to evaporate, creating a solid residue. The residue may contain sucrose, glucose, fructose, and Fe(OH)3 or FeOOH. The formation of iron hydroxides might be due to the hydrolysis of Fe3+ ions at higher temperatures (2), which glucose will then reduce to form Fe3O4, gluconic acid, and hydrogen. When the sucrose melts, it produces even more glucose molecules [41], which evenly adsorb on Fe(OH)3 or FeOOH, preventing their aggregation similar to conventional surfactant agents. Finally, the decomposition and carbonization of Fe(OH)3 or FeOOH/glucose can produce the C/Fe3O4 microspheres. Note that there is no carbon isolation from sucrose-only precursor solution in this synthesis. Another possible mechanism for the formation of Fe3O4 might be the reduction of Fe2O3 (a direct decomposition product of Fe(NO)3) by the decomposition products of sucrose.

Fig. 4
figure 4

Formation mechanism of porous C/Fe3O4 microspheres prepared by SP with NaNO3 additive

Full size image
$${C}_{12}{H}_{22}{O}_{11}+HOH\to {C}_{6}{H}_{12}O \left(glucose\right)+{C}_{6}{H}_{12}O (fructose)$$
(1)
$${Fe}^{3+}+HOH\to {Fe(OH)}_{3}\left(or FeOOH\right)+{H}^{+}$$
(2)

In contrast, the SP synthesis of porous C/Fe3O4 microspheres with NaNO3 additive proceeds with some additional steps: the melting and decomposition of NaNO3 (Fig. 4b-4). The molten NaNO3 may increase the surface energy of Fe3O4 nanoparticles, leading to their agglomeration and crystallization (Fig. 4b-4). Furthermore, the decomposition products of NaNO3 react with COx gases generated from the decomposition of sucrose to produce sodium compounds such as Na2CO3 (Fig. S1a). Aqueous etching can easily remove these sodium compounds to produce porous C/Fe3O4 microspheres with large surface areas and pore volumes. Moreover, NaNO3 additive catalyzes the carbonization of sucrose, leading to an increase in carbon content and particle size.

From the above results, it can be concluded that the SP with a 1.00 M NaNO3 additive can successfully produce porous C/Fe3O4 microspheres with large surface areas and pore volumes. Moreover, the systematic studies by various characterization techniques showed that NaNO3 additive catalyzes the carbonization of sucrose and enhances the crystallization of Fe3O4.

Electrochemical characterization of porous C/Fe3O4 microspheres as an anode material for Li-ion batteries

Figure 5a shows the CV curves of porous C/Fe3O4 microspheres prepared by SP with 1.00 M NaNO3 additive in the potential window of 0–3.00 V (vs. Li/Li+) at a scan rate of 0.1 mV s–1. The first cathodic scan revealed weak (at ~1.68 V) and strong (at ~0.50 V) peaks, respectively, corresponding to Li-ion insertion into Fe3O4 lattice (Fe3O4 + xLi+ + xē ↔ LixFe3O4 (0 ≤ x ≤ 2)) and its subsequent reduction to Fe0 (LixFe3O4 + (8–x)Li+ + (8–x)ē ↔ 3Fe0 + 4Li2O), as well as a formation of solid electrolyte interface (SEI) layer [42, 43]. Moreover, a sharp cathodic peak near 0 V could be attributed to a reversible lithiation of carbon (2C + Li+ + ē ↔ LiC2) [44]. On the other hand, the anodic scan revealed a broad peak ranging from ~1.1 to 1.8 V assigned to the delithiation of carbon and reversible oxidation of Fe0 to Fe2+ and Fe3+ [45]. The anodic scan also revealed a peak at ~2.3 in the CV curves of this sample. Similar peaks appeared in the CV curves of sandwich-structured carbon nanotube/CuxOy/Cu composite, which was assigned to the electrochemically active copper oxide in the interface between the Cu foil and carbon nanotube [46]. Mao et al. [47] also attributed such high potential peaks to the partially oxidized Cu formed during the drying of electrode coated Cu foil in air atmosphere. Therefore, the appearance of this peak in the CV curves of porous C/Fe3O4 microspheres might be due to these partially oxidized Cu foil or some irreversible side reactions as it disappeared in the third cycle.

Fig. 5
figure 5

CV curves (a) at a scan rate of 0.1 mV s–1, charge-discharge curves (b) at a current density of 50 mA g–1, cycling performance (c) at a current density of 50 mA g–1, C-rate performances (d), and long-term cycling performances (e) at a current density of 500 mA g–1 of C/Fe3O4 microspheres prepared by SP with and without 1.00 M NaNO3 additive

Full size image

Figure 5b shows the first three charge/discharge curves of porous C/Fe3O4 microspheres prepared by SP with 1.00 M NaNO3 additive in the potential window of 0.01–3.00 V (vs. Li/Li+) at a current density of 50 mA g–1. A sizeable irreversible capacity loss in the first cycle was attributed to the formation of SEI layer and structural modification of the electrode. Nevertheless, the specific capacity was stabilized from the second cycle with a high charge/discharge capacity of ~780 mAh g–1, confirming reversible electrochemical performance. Figure 5c shows the corresponding cycling performances of dense and porous C/Fe3O4 microspheres prepared by SP with and without 1.00 M NaNO3 additive in the potential window of 0.01–3.00 V (vs. Li/Li+) at a current density of 50 mA g–1. The porous C/Fe3O4 microspheres delivered a higher specific capacity of ~780 mAh g–1 in the initial cycles and ~520 mAh g–1 after 30 cycles than the dense microspheres. The cycle performance of porous C/Fe3O4 microspheres at a current density of 100 mA g-1 was also shown in Fig. S5. The enhanced electrochemical performances of C/Fe3O4 microspheres were attributed to the porous morphology with large surface areas and pore volumes.

Figure 5d shows the rate performance of dense and porous C/Fe3O4 microspheres prepared by SP with and without 1.00 M NaNO3 additive in the potential window of 0.01–3.00 V (vs. Li/Li+) at a series of current densities. The porous C/Fe3O4 microspheres had a better rate performance than the dense C/Fe3O4 microspheres at all current densities. Moreover, the specific capacity of porous C/Fe3O4 microspheres could recover to ~700 mAh g–1 when the current returns to a lower rate, indicating their good reversibility. The wide range of pores produced from the removal of sodium compounds might enable easy electrolyte penetration and facilitate fast Li-ion diffusion, while the N-doping can promote the electronic conductivity of the carbon. These features of porous C/Fe3O4 microspheres led to the improved rate capability of this sample.

Figure 5e shows the long-term cycling performances of dense and porous C/Fe3O4 microspheres prepared by SP with and without 1.00 M NaNO3 additive in the potential window of 0.01–3.00 V (vs. Li/Li+) at a current density of 500 mA g–1. Both samples demonstrated a stable cycling performance thanks to the carbon matrix's suppressed volume expansion of Fe3O4 nanoparticles. However, the porous C/Fe3O4 microspheres had a two-fold higher specific capacity of ~400 mAh g–1 than the dense microspheres (~200 mAh g–1) at all over cycles. The appropriate morphology of porous C/Fe3O4 microspheres ensures a high contact area between Fe3O4 nanoparticles and electrolyte, leading to a better utilization of Fe3O4 nanoparticles. Moreover, the CV measurements of this sample showed that the N-doped carbon can reversibly react with Li-ions in the specified potential window. Considering the composition of the composite material, it can be seen that the N-doped carbon has a substantial contribution to the specific capacity of the composite.

Figure 6 shows the electrochemical impedance spectra of dense and porous C/Fe3O4 microspheres prepared by SP with and without 1.00 M NaNO3 additive before cycling. The spectra of both samples had a semicircle in the high to medium frequency range and an inclined line in the low-frequency region associated with the charge transfer resistance and Warburg resistance, respectively. A lower charge transfer resistance accelerates the reaction kinetics of the electrodes. The porous C/Fe3O4 microspheres had much smaller charge transfer resistance than their dense counterparts. Accordingly, this sample demonstrated a higher specific capacity and enhanced rate performances.

Fig. 6
figure 6

Electrochemical impedance spectra of C/Fe3O4 microspheres prepared by SP with and without 1.00 M NaNO3 additive before cycling and Randles equivalent circuit (the inset)

Full size image

The morphology stability of porous C/Fe3O4 microspheres for repeated lithiation/delithiation processes was observed by SEM, as shown in Figure 7. The electrode material preserved its initial spherical morphology even after 200 cycles at a current density of 500 mA g–1. The appropriate morphology of porous C/Fe3O4 microspheres with a wide range of pores can effectively withstand substantial volume expansion of Fe3O4 nanoparticles, leading to the stable cycling stability of the sample.

Fig. 7
figure 7

SEM images of the porous C/Fe3O4 microspheres after 200 charge/discharge cycles at a current density of 500 mA g–1

Full size image

Conclusion

In conclusion, we prepared porous C/Fe3O4 microspheres by simple and cost-effective spray pyrolysis with NaNO3 additive and studied their electrochemical properties as anode material for Li-ion batteries. The systematic characterizations were performed to investigate the effects of NaNO3 on the physical and electrochemical properties of the samples. The XRD, SEM, TEM, and CHNS results showed that NaNO3 catalyzes the carbonization of sucrose and enhances the crystallization of Fe3O4. Moreover, nitrogen adsorption/desorption analysis showed that the aqueous removal of sodium compounds produces porous C/Fe3O4 microspheres with large surface areas and pore volumes. The porous C/Fe3O4 microspheres exhibited a reversible capacity of ~780 mAh g–1 in the initial cycles and ~520 mAh g–1 after 30 cycles at a current density of 50 mA g–1. Moreover, a reversible capacity of ~400 mAh g–1 was attainable after 200 cycles, even at a high current density of 500 mA g–1. The wide range of pores produced from the removal of sodium compounds might enable easy electrolyte penetration and facilitate fast Li-ion diffusion, while the N-doping can promote the electronic conductivity of the carbon. These features of porous C/Fe3O4 microspheres led to the improved electrochemical properties of this sample.