**3. Results and Discussion**

The morphology and architecture of the HCS and the HCS with metal oxides were presented (Scheme 1) and characterized using TEM analysis. Based on the TEM images (Figure 1), it was found that the cores of solid silica spheres were ~250 nm in diameter. The mesoporous shell has a thickness of ~100 nm. Therefore, the obtained HCS have a size diameter of ~450 nm. The iron oxide nanoparticles were evenly distributed throughout the hollow carbon spheres, and their size was ~15 nm. Microscopic analysis revealed that MnO2 was deposited onto HCS in a flat form of thin rods with an irregular surface (Figure S1). Their diameter was ~20 nm while the size of the iron oxide was ~20 nm (Figure 2). To confirm the elemental composition of the sample, EDS mapping was performed. Figure 3 reveals that the Mn and Fe was distributed homogeneously of the carbon shell.

**Scheme 1.** Scheme presentation of FexOy/MnO2/HCS synthesis.

**Figure 1.** Transmission Electron Microscopy (**A**) images of SiO2, (**B**) SiO2@mSiO2, (**C**) hollow carbon spheres, (**D**) FexOy/MnO2/HCS and (**E**) the lab-scale lithium cell prototype.

**Figure 2.** High resolution TEM images of (**A**) MnO2 and (**B**) FexOy.

**Figure 3.** Scanning Transmission Electron Microscope image and high-angle annular dark-field scanning transmission electron microscopy and energy-dispersive X-ray spectroscopy (HAADF-STEM-EDS) mapping images of FexOy/MnO2/HCS.

In the next step, XRD patterns acquired from HCS, FexOy/HCS, MnO2/HCS and the corresponding hybrid material of FexOy/MnO2/HCS were depicted (Figure 4). The black line which corresponds to carbon spheres has significant and broad peaks at 2θ = 24.9◦ and 42◦ in response to graphitic carbon planes (002) and (100). FexOy/MnO2/HCS exhibits further diffraction peaks which reflect the peaks appearing on the patterns of individual components. The peaks at 2θ = 37.9◦ and 57.8◦ correspond to the (211) and (600) MnO2 planes, respectively [21]. Based on the obtained pattern, it was found that the sample contains a mixture of iron oxides: Fe2O3 and Fe3O4. Diffraction peaks characteristic for Fe2O3 were identified at 2θ = 24.4◦, 35.7◦, 44.2◦, 49.6◦ and 62.9◦. They are related to the (211), (110), (024) and (214) planes [22]. Two peaks assigned to Fe3O4 were found, at 2θ = 36.6◦ (311) and 2θ = 57.8◦ (511) [23].

**Figure 4.** X-ray diffraction patterns of HCS, FexOy/HCS, MnO2/HCS and the hybrid material of FexOy/MnO2/HCS.

First, the BET method was used to investigate the specific surface area of the HCS and the corresponding hybrid composite. The N2 adsorption/desorption isotherms are shown in Figure 5A. For HCS, the BET specific surface area is 571 m2/g. After functionalization with metal oxides, the surface area decreased to 177 m2/g. A reduction in the specific surface area is related to the fact that nanoparticles of iron oxide are both on the surface and in the pores of nanospheres. Additionally, the presence of MnO2 on the surface of the nanospheres may cause blockage of the pores [24]. From the adsorption branch, the related mesopore size distribution determined using the BJH approach gives average pore sizes at ~5.53 nm with a predominance of pores of a size at 2.2 nm in the case of the HCS, and 4.97 average pore size and most pores with size 2.57 nm for FexOy/MnO2/HCS (Figure 5B). This result suggests the mesoporous nature of the material (Table 1).

**Figure 5.** (**A**) N2 adsorption/desorption isotherms and (**B**) pore size distribution of HCS and FexOy/MnO2/HCS.


**Table 1.** Specific surface area, pore volume and pore sizes of hollow carbon spheres (HCS) and FexOy/MnO2/HCS.

The purity of HCS and the quantitative analysis of the nanocomposite was verified by thermogravimetric analysis as shown in Figure 6A. At 540 ◦C, HCS began to decompose in air. The weight loss accelerated as the temperature rose, until all the carbon spheres were depleted at approximately 735 ◦C. HCS has an ash percentage of 0 wt.% after combustion at 900 ◦C, indicating that it is of high purity. In comparison to the pristine HCS, the stability of FexOy/MnO2/HCS was weaker. During heating, the nanocomposite burned at 230 ◦C and ended at 430 ◦C. About 50 wt.% of the sample decomposed; thus, it can be concluded that the metal oxides are half the mass of the sample. Raman spectroscopy is commonly used to characterize all sp2 carbons. The Raman spectra of HCS and the corresponding nanocomposite shows two prominent peaks at 1320 and 1600 cm−<sup>1</sup> (Figure 6B). The former, named D-band, reveals the defect of the C atomic lattice, while the latter peak, called Gband, represents the stretching vibration of C atom sp2 hybrid plane. The relative intensity of D to G provides an indicator for determining the in-plane crystallite size or the amount of disorder in the sample [25,26]. Upon deposition of the metal oxides' nanoparticles, the relation between the D-band and G-band intensities increases; thus, additional defects formed in the HCS structures.

**Figure 6.** (**A**) Thermogravimetric analysis profiles, and (**B**) Raman spectra of HCS and FexOy/MnO2/HCS.

The prepared materials (FexOy/MnO2/HCS) were further evaluated as anode material for Li-ion batteries. The CV curves of the FexOy/MnO2/HCS electrode recorded at a scan rate of 0.5 mV s−<sup>1</sup> are shown in Figure 7A. During the first discharge cycle of the FexOy/MnO2/HCS electrode, a strong reduction peak in cathodic scan was observed at ~0.5 V, which is in agreement with the reduction of Mn2+ and Fe3+ to their metallic states due to the formation of Li2O, as illustrated in Equation (1),

$$\text{FeMnO}\_3 + \text{6Li}^+ + \text{6e}^- \rightarrow \text{Fe} + \text{Mn} + \text{3Li}\_2\text{O} \tag{1}$$

accompanied by electrolyte decomposition into a solid electrolyte interphase (SEI) layer. During the first anodic [27] charge process of the FexOy/MnO2/HCS electrode, only two

anodic peaks (1.19 and 2.08 V) can be attributed to the oxidation of metallic Mn and Fe, which are illustrated in Equations (2) and (3):

$$\text{Fe} + \text{xLi}\_2\text{O} - 2\text{xe}^- \rightarrow 2\text{FeO}\_2 + 2\text{xLi}^+ \tag{2}$$

$$2\text{ Mn} + \text{xLi}\_2\text{O} - 2\text{xe}^- \rightarrow 2\text{MnO}\_\text{x} + 2\text{xLi}^+ \tag{3}$$

These two peaks shift to 0.75 and 0.3 V in the following reduction step, indicating improved kinetics. The large peaks at 1.2 and 1.7 V in the charge process are attributable to the two-step oxidation of Mn(0) and Fe(0) to MnOx and FeOx, respectively [28]. The two pairs of reduction and oxidation peaks that correspond to the FeOx/Fe and MnOx/Mn conversions appear to be well overlapped, indicating that the two-step electrochemical reactions are highly reversible.

**Figure 7.** (**A**) cyclic voltammetry curves performed over a potential window from 0.05 to 3 V at a scan rate from 0.5 mVs–1, (**B**) galvanostatic charge/discharge profiles at a current density of 50 mAg−<sup>1</sup> in the voltage range of 0.05–3.0 V, (**C**) voltage–capacity curves, (**D**) gravimetric specific capacities vs. cycle number and (**E**) Coulombic efficiency of FexOy/MnO2/HCS.

Figure 7B shows the first, the second and the fifth galvanostatic discharge/charge curves of the FexOy/MnO2/HCS between 0.05 and 3.0 V (versus Li/Li+). The orange line is attributed to the first cycle charge and discharge capacity of FexOy/MnO2/HCS 625 mAhg−<sup>1</sup> and 1100 mAhg<sup>−</sup>1, respectively. It can be assigned to irreversible effects such as the formation of the SEI layer. After cycling, thin SEI form on the FexOy/MnO2/HCS electrode, and additional mesopores form in the hollow structure, resulting in the establishment of linked spaces that are conducive to fast Li<sup>+</sup> ion and electron transport. The stable SEI layer and hollow space on electrodes can help to stabilize lithiation/delithiation and reduce mechanical deterioration caused by discharge volume expansion. In the next step, the charge/discharge profile, with different current densities, was measured. The charge/discharge curves of the FexOy/MnO2/HCS composite at various rates are shown in Figure 7C. On both discharge and charge profiles, multiple plateaus can be seen, which are in good agreement with the CV curves. A sequential decay in reversible capacities as the rate increases can be observed. The electrode delivered reversible capacities of 1100, 610, 320, 126, 75 and 42 mAhg−<sup>1</sup> at current densities from 50 to 1000 mAg−1. As shown in Figure 7D, the new anode material exhibits good Li+-ion storage capacity

and cyclic stability at each current density from 50 to 1000 mAg–1. Notably, the fused FexOy/MnO2/HCS presented much higher capacities at each stage compared to pristine HCS. Charge-discharge profiles obtained by different applied current densities were used to further study the rate behavior of the FexOy/MnO2/HCS electrode. As the rate increased, there was a sequential decrease in reversible capabilities. Figure 7D shows that the electrode delivered reversible capacities of 611, 323, 135, 83 and 46 mAhg−<sup>1</sup> at current densities from 50 mAg−<sup>1</sup> to 1000 mAg<sup>−</sup>1. When the current density was reduced to 50 mAg<sup>−</sup>1, the capacity immediately returned to 675 mAhg−1. The above results indicate that the hybrid material of the FexOy/MnO2/HCS electrode has an excellent rate capability. Additionally, the FexOy/MnO2/HCS electrode displays high CE, which in many cycles exceeds 100% (Figure 7E). This can be associated with the irreversible side reaction during the charge process or with an irregular amount of transported Li+ ions during charge–discharge processes. In the first case, the side reaction can suggest more capacity is generated than the amount of Li+ ions released from the active material. Capacity does not reference the actual storage ability of the electrode; it is estimated based on coulomb counting, e.g., to integrate current vs. time until the cut-off potential is reached. Therefore, if there are any side reactions that consume current without affecting the voltage (e.g., the charge is not actually intercalating to a site in the electrode), then this current is integrated, adding to the capacity. If this occurs in the discharge step, then the Coulombic efficiency (CE) result can be >100%. In the second case, if the Li+ ion is less intercalated due to structural interference during the discharge process and the maximum amount of Li-ion is released during the charge process, the CE may exceed 100%. Both scenarios can lead to a gradual degeneration of the structure of the electrode active material and, thus, to a reduced stability (Figure S2).

Hollow carbon spheres decorated by iron and manganese oxides with large surface area and a conductive network enable high accessibility of the active material. FexOy/ MnO2/HCS composite initially reaches the full theoretical capacity, but degradation effects lead to poor cycle stability. A comparison with the electrochemical performance of other reported HCS composites with FexOy, or MnO2, shows that this is a promising approach to optimize the cycling stability of the battery. Graphene-wrapped Fe3O4 synthesized by Zhao et al. [29] shows a better charge/discharge capacity but not satisfying stability. Zhu et al. [30] obtained porous olive-like carbon decorated by Fe3O4, which presented lower dis- and charge capacities. Wu et al. [31] presented a novel foam-like Fe3O4/C composite made with gelatin as the carbon source and ferric nitrate as the iron source using a sol-gel process. As a result, the Fe3O4/C composite electrode demonstrates good rate performance with a reversible capacity of 660 and 580 mAhg−<sup>1</sup> at 3 and 5 ◦C, respectively, whereas all composite manganese oxide/carbon presented lower capacity and stability compared to our data [32,33]. Therefore, a combination of two oxides (iron and manganese) significantly improved the capacity of obtained electrodes. The state-of-the-art process provides information about the synergistic effect of such a combination [34]. The synergistic effect of combining such components manifests in improving the reversibility of the electrochemical reaction, buffering large distortions and stresses during discharge-charging processes and preventing aggregation of the active material. This results in high reversible capacity, excellent cycling performance and excellent rate capabilities. The unique MnO2 nanorods morphology has been reported as anode material for lithium-ion batteries [35–37]. This rod-like morphology is reported to enhance electrochemical properties and was proven in our study. These features, along with the high performance of iron oxides, recommend this hybrid structure as promising for boosting the performance of energy storage devices.
