*2.2. Characterization Techniques*

The microstructures of the resulting powders were observed by scanning electron microscopy (SEM; JEOL, JSM-6060, JEOL, Tokyo, Japan) and field-emission transmission electron microscopy (FE-TEM; JEOL, JEM-2100F, JEOL, Tokyo, Japan). The crystal phases were evaluated by X-ray diffractometry (XRD; X'Pert PRO MPD, PANalytical, Almelo, The Netherlands) using Cu Kα radiation (λ = 1.5418 Å). X-ray photoelectron spectroscopy (XPS; K-Alpha, Thermo Fisher Scientific, Waltham, MA, USA) with a focused monochromatic Al Kα at 12 kV and 20 mA was used to analyze the composition of the samples. A structural characterization of carbon in the sample was performed by Raman spectra (Jobin Yvon LabRam HR800, Horiba Jobin Yvon, Paris, France, excited by a 632.8 nm He–Ne laser) at room temperature. The surface areas of the powders were measured by the Brunauer–Emmett–Teller (BET) method, using N2 as the adsorbate gas. Thermogravimetric analyses (TGA) were performed using a Pyris 1 TGA (Perkin Elmer, Waltham, MA, USA) within a temperature range of 25–650 ◦C and at a heating rate of 10 ◦C min−<sup>1</sup> under a static air atmosphere.

## *2.3. Electrochemical Measurements*

The electrochemical properties of the samples were analyzed by constructing a 2032-type coin cell. The lithium cell assembly was made in an Ar-filled glove box at room temperature where water and the oxygen concentration was kept at less than 1 ppm. The anode slurry was prepared by mixing the active material, carbon black, and sodium carboxymethyl cellulose (CMC) in a weight ratio of 7:2:1. The working electrodes were formed by coating the slurry onto copper foils and subsequently dried at 70 ◦C for 3 h. Li metal and a microporous polypropylene film were used as the counter electrode and the separator, respectively. The electrolyte was composed of 1 M LiPF6 dissolved in a mixture of fluoroethylene carbonate/dimethyl carbonate (FEC/DMC; 1:1 *v*/*v*). The discharge/charge characteristics of the samples were investigated by cycling over a potential range of 0.001–3.0 V under CC (constant-current) conditions. Cyclic voltammograms were measured at a scan rate of 0.1 mV s<sup>−</sup>1. The negative electrode measured 1.5 cm × 1.5 cm, and the mass loading of the active materials was kept at approximately 1.5 mg cm<sup>−</sup><sup>2</sup> in every electrochemical test. The electrochemical impedance spectra were obtained by performing alternating current electrochemical impedance spectroscopy (EIS; ZIVE SP1) over a frequency range of 0.01 Hz to 100 kHz.

#### **3. Results and Discussion**

Low crystalline MoO3/C composite microspheres, in which MoO3 nanocrystals were distributed homogeneously in the amorphous C matrix, were directly prepared by a one-step spray pyrolysis without any further treatment. Figure 1 shows the morphologies of the MoO3/C composite microspheres obtained after the one-step spray pyrolysis. The powders were spherical and had diameters on the order of microns because they were formed from one droplet with several tens of micrometers by drying, decomposition, and crystallization inside the hot-wall reactor, as shown in Scheme 1. Additionally, there was no aggregation between the powders because the spray pyrolysis was carried out within a very short residence time of 6 s in a hot-wall reactor maintained at 900 ◦C under a N2 atmosphere in Figure 1a,b. From a high-resolution TEM image in Figure 1c, it was hard to confirm the nanocrystal MoO3 grains formed during spray pyrolysis in a microsphere structure because the amorphous-like, very small MoO3 nanocrystals were formed during the spray pyrolysis at 900 ◦C within a short residence time of 6 s. The XRD result also showed the broad peak intensities of the β-MoO3 phase in Figure 1d. The mean crystallite size of the MoO3 powders, which was calculated from the width of the (011) peak using Scherrer's equation, was 4 nm. Grain growth of the MoO3 nanocrystals was effectively prohibited both by the short residence time of the droplet in the reactor and by being surrounded by the carbon formed by the decomposition of PVP and sucrose during the process. The elemental mapping images shown in Figure 1e exhibited a homogeneous distribution of Mo, O, and C, which implies that the ultrafine MoO3 nanocrystals were homogeneously composited with C in the microsphere structure.

**Figure 1.** The (**a**) SEM, (**b**) TEM, (**c**) high-resolution TEM images, (**d**) XRD pattern, and (**e**) elemental mapping images of MoO3/C composite microspheres.

**Scheme 1.** The formation mechanism of the low crystalline MoO3/C composite microspheres by the one-step spray pyrolysis process.

To identify the chemical composition of the MoO3/C composite microspheres, XPS analysis was carried out, as shown in Figure 2. The XPS survey spectrum of the composite microspheres confirmed the presence of Mo, O, and C, as shown in Figure 2a. In the Mo 3d spectrum of the microspheres (Figure 2b), the main peaks occurred at binding energies of 231.7/232.7 eV for Mo 3d5/2 and 234.7/235.7 eV for Mo 3d3/2; the peaks located at 232.7 and 235.7 eV are characteristic of typical values of the 3d orbital doublet Mo6+, and the minor ones centered on 231.7 and 234.7 eV corresponded to the 3d orbital doublet Mo5+, which indicated that dangling bond sites where charges could be trapped existed in MoO3 [27,28]. The C 1s XPS peak observed at 284.6 eV in Figure 2d corresponds to the binding energy of the sp<sup>2</sup> C–C bond of the carbon matrix [29–31].

The carbon matrix of the MoO3/C composite microspheres was characterized by means of Raman spectroscopy. The degree of graphitization of the carbon material can typically be evaluated according to the intensity ratio of the D and G bands of carbon at approximately 1350 and 1590 cm<sup>−</sup>1, respectively [32,33]. The peak intensity ratio between the D and G bands (ID/IG) for the MoO3/C composite microspheres was approximately 3.2, and the absence of the 2D band at approx. 2685 cm<sup>−</sup><sup>1</sup> demonstrated that the carbon formed in the composite was fairly disordered. Thus, a large amount of the amorphous carbon was formed by the decomposition of both PVP and sucrose during the spray pyrolysis. In general, amorphous carbon has more capacity as an anode for LIBs than graphitic carbon, which is mainly contributed by pores and voids in the microcavities of the structure. The Thermogravimetric (TG) curve of the MoO3/C composite microspheres in Figure 3b revealed a weight loss between 380 and 460 ◦C because of the degradation of amorphous carbon. Therefore, the content of amorphous carbon of the MoO3/C composite microspheres estimated from the TG analysis was 26 wt %.

**Figure 2.** The XPS spectra of the MoO3/C composite microspheres: (**a**) the survey XPS spectrum and high-resolution XPS spectra of (**b**) Mo 3d, (**c**) O 1S, and (**d**) C 1s.

**Figure 3.** (**a**) The Raman spectrum and (**b**) thermogravimetric analysis (TGA) curve of the MoO3/C composite microspheres.

In order to clearly prove the structural merits of MoO3/C composite microspheres as anodes for Li+ ion storage properties, bare MoO3 powders without C were also prepared from the spray solution without either PVP and sucrose by spray pyrolysis, as shown in Figure 4. The mean particle size of the resulting bare MoO3 powders, as measured from the SEM and TEM images in Figure 4a,b, was 420 nm and had no aggregation between the powders. Additionally, the resulting powders were angular, which is attributed to the crystal growth of MoO3 particles because there was no carbon surrounding the particles during spray pyrolysis to prevent the growth of MoO3 crystals during the short residence reaction time of the droplets. The high-resolution TEM image in Figure 4c shows clear lattice fringes separated by 0.23 nm, which corresponds to the (011) crystal plane of β-MoO3 (JCPDS card No. 37–1445) [34]. The XRD pattern of the bare MoO3 powders (Figure 4d) shows that they have different allotropes of MoO3 structures, with no impurities. The thermodynamically favored α-MoO3 phase was newly formed along with β-MoO3 in the bare MoO3 powders during spray pyrolysis. Bare MoO3 powders without C were further confirmed by the elemental mapping images in Figure 4e. The BET surface areas of the MoO3/C composite microspheres and of the bare MoO3 powders were 4.3 and 0.6 m<sup>2</sup> g<sup>−</sup>1, respectively, in Figure S2.

**Figure 4.** The (**a**) SEM, (**b**) TEM, (**c**) high-resolution TEM images, (**d**) XRD pattern, and (**e**) elemental-mapping images of the bare MoO3 powders.

The electrochemical properties of the MoO3/C composite microspheres are compared with those of the bare MoO3 powders in Figure 5. The cyclic voltammogram (CV) curves of the MoO3/C composite microspheres and bare MoO3 powders performed in the 0.01–3.0 V range at a scanning rate of 0.01 mV s<sup>−</sup><sup>1</sup> for the first four cycles are shown in Figure 5a. In the first cathodic scan of the MoO3/C composite microspheres, the broad peaks located at 1.16 V and 0.21 V are assigned to the interaction of Li+ ions with the amorphous carbon matrix of the MoO3/C composite and conversion reaction of Li*x*MoO3 to Mo0 and Li2O [35–37]. The peak at 0.05 V is also observed, caused by the Li+ ion's intercalation into the C matrix [38,39]. In the anodic scans of the MoO3/C composite microspheres, reversible peaks at 1.42 and 1.77 V are attributed to the monoclinic-orthorhombic-monoclinic phase

transitions in the partially lithiated Li*x*MoO2 [35–37]. In the subsequent cycles, two redox peak pairs appeared at 0.21/1.3 and 1.42/1.77 V, which corresponded to the redox reaction of MoO3 [35,40,41]. The bare MoO3 powders showed peaks at 2.03 and 1.8 V in the first cathodic scan, which correspond to the generation of Li*x*MoO3, causing an irreversible structural change from the α-MoO3 additionally formed in the bare MoO3 powders to an amorphous phase [40–42]. The subsequent peak at 0.17 V results from the conversion reaction of Li*x*MoO3 to Mo0 and Li2O [35–37,41].

**Figure 5.** The electrochemical properties of the MoO3/C composite microspheres and the bare MoO3 powders: (**a**) CV curves, (**b**) charge-discharge curves, (**c**) cycling performances, and (**d**) rate performances.

The initial discharge-charge curves of the two samples at a current density of 1.0 A g<sup>−</sup><sup>1</sup> are shown in Figure 5b. The initial discharge capacities of the MoO3/C composite microspheres and the bare MoO3 powders were 1403 mA h g<sup>−</sup><sup>1</sup> and 1478 mA h g<sup>−</sup>1, respectively, and their initial Coulombic efficiencies were 75% and 72%, respectively. Although the MoO3/C composite microspheres contained C with a high irreversible capacity loss, the initial Coulombic efficiency of the MoO3/C composite microspheres was relatively higher than that of the bare MoO3 powders. The high structural damage to the bare MoO3 powders in the first discharge and charge processes resulted in a low initial Coulombic efficiency. The discharge capacity and cycling properties of the MoO3/C composite microspheres and bare MoO3 powders at a current density of 1.0 A g<sup>−</sup><sup>1</sup> are shown in Figure 5c. Compared with bare MoO3 powders, the MoO3/C composite microspheres exhibited a satisfactorily stable cycling performance. The discharge capacity of the MoO3/C composite microspheres decreased slightly from 1066 mA h g<sup>−</sup><sup>1</sup> (533 mA h cc<sup>−</sup>1) to 808 mA h g<sup>−</sup><sup>1</sup> (404 mA h cc<sup>−</sup>1) from the 2nd cycle to the 100th cycle, whereas that of the bare MoO3 powders decreased rapidly from 1090 mA h g<sup>−</sup><sup>1</sup> (621 mA h cc<sup>−</sup>1) to 239 mA h g<sup>−</sup><sup>1</sup> (136 mA h cc<sup>−</sup>1) in the same cycle range. Additionally, the Coulombic efficiency of the MoO3/C composite microspheres increased quickly to above 99% after the second cycle. The amorphous C matrix of MoO3/C composite microspheres more effectively buffered the large volume change of the MoO3 active material during the fast charging–discharging process. On the other hand, the structural destruction of the bare MoO3 powders during repeated Li+-ion insertion and

desertion processes resulted in capacity fading continuously. Therefore, better cycling of the MoO3/C composite microspheres could be achieved because of the improved structural stability of the MoO3.

In order to evaluate the rate performances of both samples, electrochemical tests were performed at various current densities, as shown in Figure 5d. As the current densities increased from 0.5 to 1.5, 3.0, and 5.0 A g<sup>−</sup>1, the MoO3/C composite microspheres exhibited reversible discharge capacities of 999, 875, 716, and 467 mA h g<sup>−</sup>1, respectively. However, the bare MoO3 powders delivered a low reversible discharge capacity of 352 mA h g<sup>−</sup><sup>1</sup> at 5.0 A g<sup>−</sup><sup>1</sup> as shown in Figure 5d. The C matrix of the MoO3/C composite microspheres improved the electrical conductivity of the sample. Additionally, the small, amorphous MoO3 nanograins imbedded within the C matrix decreased the diffusion distance and increased the diffusion rate of the Li+ ions, thus synergistically speeding up the rate of the MoO3/C composite microspheres more than that of the bare MoO3 powders.

The superior Li+-ion storage properties of the MoO3/C composite microspheres were supported by EIS analysis, as shown in Figure 6 [43–45]. Nyquist plots of the samples before and after cycles were obtained by deconvolution with a Randle-type equivalent-circuit model (Figure 6d). The MoO3/C composite microspheres and bare MoO3 powders had similar charge-transfer resistance (Rct) values before cycling, as shown in Figure 6a. However, the cell with the MoO3/C composite microspheres obtained after 100 cycles showed a lower Rct value of 42 Ω compared to that of 134 Ω for the bare MoO3 powders, as shown in Figure 6b,c. The structural destruction of the bare MoO3 powders during the repeated Li+-ion insertion and desertion processes increased the Rct values significantly. On the other hand, the MoO3 nanograins embedded within the amorphous C were not pulverized during the repeated cycles. Moreover, the C matrix served as fast and continuous transport pathways for electrons upon cycling because of its high electrical conductivity. The high structural stabilities of the MoO3/C composite microspheres with high lithium-ion storage capacities resulted in low Rct values during cycling. The MoO3/C composite microspheres with a high structural stability during repeated lithium insertion and desertion reactions showed excellent cycling and rate performance, as shown in Figure 5.

**Figure 6.** The impedance analysis of the MoO3/C composite microspheres and the bare MoO3 powders: (**a**) before cycling, (**b**) bare MoO3 powders, (**c**) MoO3/C composite microspheres, and (**d**) the tquivalent circuit model used for AC impedance fitting: *Rct* = charge-transfer resistance, *Re* = electrolyte resistance, *Rf* = SEI layer resistance, *Q*1 = dielectric relaxation capacitance, and *Q*2 = associated double layer capacitance.

The morphologies of the MoO3/C composite microspheres and bare MoO3 powders obtained after 100 cycles are shown in Figure 7. The bare MoO3 powders were broken into several pieces after the cycles, as shown by the TEM image in Figure 7a. In contrast, the MoO3/C composite microspheres maintained their morphologies quite well even after the repeated Li+ insertion and desertion processes in Figure 7b,c. The excellent Li+-ion storage properties of the MoO3/C composite microspheres are, therefore, attributed to the improvement of the structural stability and electrical conductivity by the carbon composite.

**Figure 7.** The morphologies of (**a**) bare MoO3 powders and (**b**,**<sup>c</sup>**) MoO3/C composite microspheres obtained after 100 cycles at a constant current density of 1.0 A g<sup>−</sup>1.
