Introduction

Supercapacitors are widely recognized to fall into two categories based on their energy storage principle: electric double-layer capacitors (EDLC) and pseudocapacitors. In EDLC, when voltage is applied, ions from the electrolyte are attracted to the surface of the electrode, forming a double layer of charges and resulting in purely physical energy storage1. Carbon materials are often employed for EDLCs because of their high surface area, suitable pore size, good electrical conductivity, chemical stability, and versatility2,3,4. These properties, particularly the high surface area and chemical stability, aid in effective ionic physisorption during electrochemical processes. However, their drawback lies in their low specific capacitance and energy density5. Conversely, pseudocapacitors achieve energy storage through reversible oxidation–reduction (Faradaic) reactions at/near the electrode surface. The additional contribution from these chemical redox reactions during charge/discharge leads to a higher specific capacitance6,7,8. Among the various redox-active materials, transition metal oxides have gained significant attention as electrode materials due to their multiple oxidation states, which allow for improved pseudocapacitance5,6,9. Tungsten oxide (WO3), a transition metal oxide with multiple crystal phases, exhibits favorable attributes as a pseudocapacitor10,11,12. Not only can it undergo reversible redox reactions between W5+ and W6+ ions, but the inherent voids in its crystal structure facilitate the smooth diffusion of ions from the electrolyte13. However, their limitations include relatively poor conductivity in bulk form and the tendency to aggregate during the charge/discharge process even in their nanostructure form, as is typical of transition metals14.

WO3 and carbon-based composites have been explored to address the abovementioned constraints in recent years. Combining WO3 and carbon materials creates a synergistic effect that complements each other's limitations, leading to better overall performance15. The hybrid presents a viable approach to enhance the electronic conductivity of WO3, improve the capacitance of carbon by incorporating redox reactions, lessen the aggregation of WO3 nanocrystals, and provide overall structural stability9. Most of these works involved nanotubes, nanowires, nanoplates, and nanosheets prepared in non-aqueous solutions or with expensive polymeric templates16. Nayak et al. utilized a solvothermal approach to synthesize a WO3 nanowire–graphene sheet composite17. Xiong et al.18 and Shi et al.19 prepared hierarchical ordered porous WO3–carbon using discarded biomass as a precursor, with the former using glue milling and carbonization-activation method and the latter via a solvothermal process. Di et al. also used a solvothermal technique to decorate carbon nanotubes with an array of WO3 nanosheets20.

This work synthesized carbon microspheres decorated with WO3 nanocrystals via a facile hydrothermal method using glucose as the carbon source. The simple procedure resulted in a hierarchical micro/nanostructure that could facilitate and enhance electrochemical reactions21. The crystal phase transformation of WO3 in the presence of glucose and its effect on the capacitive behavior of the WO3/C electrode was also investigated. Electrochemical tests revealed that the WO3/C nanocomposite provided more pathways for charge diffusion within its structure. These pathways appear to result from a cooperative interplay between the intricate nanocrystalline mixed phase WO3 and the porous carbon microsphere.

Experimental section

Materials

Analytical/reagent grade tungstic (VI) acid (H2WO4, Alfa Aesar), D( +)-glucose (Acros Organics), Nafion® D-521 (Alfa Aesar), ethanol (VWR Chemicals), and potassium hydroxide (Sigma-Aldrich) were used without further purification.

Synthesis of the WO3/C composite and preparation of working electrodes

The precursor solution consisted of 1 g H2WO4 (dissolved in 5 mL ethanol), 1 g glucose, and 75 mL distilled water. The solution was transferred to a Teflon-lined stainless-steel autoclave, sealed, and heated at 180 °C for 20 h. After cooling, the hydrothermal reaction product was washed with water and ethanol before drying at 60 °C. Scheme 1 shows a graphical illustration of the hydrothermal synthesis. Carbon microspheres would form due to dehydration and oligosaccharide formation. Supersaturation would then lead to nucleation and subsequent growth of WO3 nanocrystals on these spheres via heterogeneous nucleation14,22.

Scheme 1
scheme 1

Graphical illustration of the WO3/C nanocomposite synthesis via a hydrothermal treatment strategy.

Full size image

The dried as-prepared sample was calcined at 600 °C for 3 hours at a ramping rate of 3°C/min in a nitrogen environment. A control sample without glucose was also prepared for comparison. 5 mg of the calcined powder sample was ground and dispersed in 500 µL ethanol to make the working electrodes. Then, 50 µL of the binder Nafion® D-521 was added. After mixing and sonicating, 220 µL of the slurry was drop-cast on copper substrates with a working area of 1.0 cm2 and dried at 60 °C for 12 h.

Characterization

The crystal structure of the nanocomposites was studied by X-ray diffraction (XRD) using a Rigaku MiniFlex 600 diffractometer equipped with a Cu Kα radiation source and a scintillation counter detector. Powder XRD patterns were recorded from 10° to 60° (0.02° step, 2°/min speed) at 40 kV and 15 mA. Fourier transform infrared (FTIR) spectra were recorded from 400–4000 cm−1 using a Shimadzu IRTracer-100 spectrophotometer with a DLATGS detector. The Raman and X-ray photoelectron spectra were collected using a Horiba XploRA Raman confocal microscope and an ESCALAB™ XI + X-ray photoelectron spectrometer, respectively. The morphology and elemental mapping of the samples were investigated using a JEOL JSM-IT800 Schottky field emission scanning electron microscope (FESEM). For surface area and porosity analysis, N2 adsorption–desorption isotherms were measured at 77 K using Quantachrome NovaWin. The electrochemical measurements were performed using a CH Instruments workstation with a three-electrode configuration. Copper foil, Ag/AgCl, platinum wire, and the synthesized materials were used as the current collector, reference electrode, counter electrode, and working electrodes, respectively. The electrolyte used was 0.1 M potassium hydroxide. The electrochemical tests included cyclic voltammetry (CV), galvanostatic charge–discharge (GCD), and impedance spectroscopy. The specific capacitance was calculated using Eq. (1), coulombic efficiency using Eq. (2), and energy efficiency using Eq. (3),

$$C_{s} = \frac{{\mathop \smallint \nolimits_{{V_{1} }}^{{V_{2} }} Idv}}{ 2sm\Delta V}$$
(1)
$$\eta _{c} = \frac{{t_{d} }}{{t_{c} }} \times 100$$
(2)
$$\eta _{E} = \frac{{E_{int/d} }}{{E_{int/c} }} \times 100$$
(3)

where Cs is the specific capacitance (F/g), \(\mathop \smallint \limits_{{V_{1} }}^{{V_{2} }} Idv\) is the integral CV curve area (AV), s is the scan rate (V/s), m is the mass of the active material (g), ∆V is the potential window, η is the coulombic efficiency (%), td is discharging time (s), tc is charging time (s), Eint/d is the galvanostatic discharge energy, and Eint/c is the galvanostatic charge energy21,23,24.

Results and discussion

The samples’ XRD patterns showed sharp and intensive peaks, indicating a crystalline structure. In Fig. 1A, the XRD pattern of the uncalcined WO3/C composite (blue) is consistent with the orthorhombic crystal structure of WO3 (JCPDS No. 43-0679) and diffractograms reported in the literature25,26,27. The strong peaks at 16.5° and 25.6° are attributed to the (020) and (111) reflections of the orthorhombic crystal structure of tungsten oxide hydrate (WO3·H2O), respectively. These peaks were not observed in the control sample WO3 (Fig. 1A, red), which did not have glucose as a carbon precursor, suggesting that glucose aided in preserving the orthorhombic crystalline phase of WO3·H2O during the hydrothermal process. The hydroxyl group in glucose and the hydrogen in the WO3·H2O molecule formed a hydrogen bond, promoting the controlled growth of WO3·H2O crystallites28.

Figure 1
figure 1

XRD patterns of WO3 and WO3/C (A) before calcination and (B) after calcination.

Full size image

Moreover, the triplet peaks near 35.0° for the WO3/C nanocomposite (Fig. 1A, blue) could be indexed to the orthorhombic (040), (200), and (002) crystal planes. But in the uncalcined WO3 sample (Fig. 1A, red), these peaks appeared as a doublet at 34.1° and corresponded to the (202) plane, indicating a monoclinic structure29,30,31. The uncalcined nanocomposite WO3/C showed a coexistence of orthorhombic and monoclinic phases, with the former as the dominant phase, and the control WO3 exhibited a purely monoclinic phase. The sharp diffractive peaks at 23.1°, 23.7°, and 24.2° corresponded to the (002), (020), and (200) crystal planes, consistent with JCPDS No. 43–1035 for monoclinic WO3. After calcination, as shown in Fig. 1B (red), the control WO3 did not change its phase but exhibited increased crystallinity with peaks becoming sharper and more defined. For instance, the doublet peak at 34.1° diverged more clearly. The elevated temperature during calcination provided sufficient energy for adjacent tiny crystals to rearrange and coalesce into larger crystals32,33. This increase in crystallite size of WO3 from 16 Å to 23 Å after calcination appeared as narrower, more intensified XRD peaks. On the other hand, the WO3/C composite changed its phase after calcination (Fig. 1B, blue). The intense peaks at 16.5° and 25.6° characteristic of orthorhombic crystal, disappeared, indicating the removal of the hydrate water in WO3·H2O25. The XRD pattern after calcination showed a tetragonal/monoclinic phase junction. The main diffraction peaks at 23.0°, 23.9°, 28.7°, 33.5°, and 34.0° could be attributed to the (002), (110), (102), (112), and (200) planes of tetragonal WO3, aligning with COD No. 152153234,35. However, this attribution may not be absolute, and a pseudo-phase consisting of orthorhombic and tetragonal phases would also be likely, as observed from previous work13,34. The less defined peaks from 45° to 60° resembled that of monoclinic WO3 and indicated a reduced intensity due to the amorphous carbon. The energy storage performance of WO3 depends on its crystal structure which influences the intercalation of ions in an electrochemical environment5,9. Orthorhombic and tetragonal WO3 generally tend to have more cavities or open spaces within its crystal structure than monoclinic WO3. The more open structure and wider tunnels in the former allow fast, reversible intercalation of ions5,9,36.

The FTIR spectra (Fig. 2A) elucidated the different functional groups existing on the surface of the samples. For the uncalcined WO3/C, the observed peak at 3387 cm−1 corresponded to the stretching vibrations of O–H from water molecules in WO3·H2O.

Figure 2
figure 2

(A) FTIR spectra of the samples before and after calcination and (B) Raman spectra of calcined WO3 and WO3/C nanocomposite after calcination.

Full size image

This peak was not as prominent in the uncalcined WO3 sample, suggesting a significant elimination of crystalline water during the hydrothermal treatment, thereby corroborating with the XRD result. The 1616 cm−1 and 1704 cm−1 peaks confirmed carbon's presence in the nanocomposite. These spectral bands were associated with the vibrations of C=C and C=O, respectively, and supported the idea that glucose likely underwent aromatization during the hydrothermal treatment22. The peaks in the spectral range 500–1000 cm−1 were characteristic absorptions of tungsten oxide. The strong peak at about 600 cm−1 corresponded to the stretching vibrations of O–W–O. The stretching vibrations of W=O appeared as a sharp shoulder absorption peak at 802 cm−1 for the uncalcined and calcined WO3 and for the calcined WO3/C composite as well, although less sharp. Only the uncalcined WO3/C showed a major characteristic band of the terminal oxygen atom (W=O) of the WO3⋅H2O structure appearing at 937 cm−1, again showing agreement with the XRD data25,27,37.

The Raman scattering spectra of the calcined samples were also recorded. The spectrum for the control WO3 sample exhibited two intense peaks at 701 cm−1 and 791 cm−1, corresponding to the stretching vibration of tungsten atoms with neighboring oxygen atoms (O–W–O) as shown in (Fig. 2B). These peaks became less intense in the presence of carbon in the WO3/C nanocomposite. The prominent peaks at 1334 cm−1 (D band) and 1577 cm−1 (G band) could be ascribed to the absorption of sp3-hybridized carbon and sp2-hybridized carbon, respectively. The D band is linked to structural disorder and defects, while the G band indicates the graphitization of carbon. Even though the XRD peaks for graphitic carbon ((002) at 24° and (100) at 43°) were overshadowed by the highly crystalline WO3, the Raman spectrum for WO3/C confirmed its presence. The intensity ratio of the D to the G peak (ID/IG) was measured at 0.818, attributing the higher G band to the graphitic clusters in the amorphous composite3,25,38,39.

The nanocomposite morphology was observed by SEM (Fig. 3A,B) and TEM (Fig. 3C,D). Carbon spheres were derived from glucose during the hydrothermal treatment at 180 °C which is higher than the typical glycosylation temperature, resulting in aromatization and carbonization. Glucose molecules underwent dehydration and formed oligosaccharides, resembling a polymerization process. The new carbon–carbon bonds eventually formed the carbon microspheres of > 1.0 µm in diameter. It could be presumed that within the 20 h hydrothermal reaction, the solution reached a critical supersaturation, and a burst of nucleation ensued, crosslinking the previously formed oligosaccharides. This aggregation of glucose consequently acted as a spherical nucleus onto which WO3 nanocrystals grew via heterogeneous nucleation14,22. This process successfully formed a nanocomposite consisting of tungsten oxide and carbon (WO3/C), as confirmed by the elemental mapping of a single sphere by energy-dispersive x-ray (EDX) spectroscopy (Fig. 3B). The EDX mapping spectrum revealed 56.9% C, 37.2% W, and 5.9% O (Supporting Information Fig. S1). The SEM images of WO3 synthesized without using glucose are shown in Fig. S2. Additionally, the lack of peaks in the range 1000 cm−1 to 1300 cm−1 in the FTIR spectrum of the uncalcined WO3/C further supported the loading of WO3 nanocrystals onto the carbon microspheres. Peaks in this range would have indicated C–OH stretching and OH bending vibrations from residual hydroxy groups. The lack thereof suggested that the hydroxyl groups of glucose have formed hydrogen bonds with WO3·H2O, which preserved the orthorhombic structure of the latter, as previously discussed in the XRD findings.

Figure 3
figure 3

(A) Scanning electron microscopy (SEM) images of WO3/C. (B) Elemental mapping of WO3/C (red = W, teal = C; green = O) by energy-dispersive x-ray (EDX) spectroscopy. (C) TEM image WO3/C and (D) HRTEM of WO3 nanocrystal attached on carbon surface.

Full size image

Results from the elemental mapping by EDX of the WO3/C nanocomposites were further confirmed by x-ray photoelectron spectroscopy (XPS). The XPS survey spectrum of the nanocomposite showed the presence of W, O, and C elements (Fig. S3). Furthermore, a comparison of the W 4f. spectra of the WO3/C nanocomposite (Fig. 4A) with that of WO3 (Fig. 4B) shows that introducing carbon in the nanocomposite altered the chemical state of tungsten. The WO3/C nanocomposite exhibited three resolved peaks, whereas WO3 alone displayed just a pair of peaks in the deconvoluted spectra. In the WO3 spectrum, the peaks at 35.1 eV (W 4f7/2) and 37.5 eV (W 4f5/2) corresponded to the W6+ oxidation state34,40,41. These shifted to slightly higher binding energies in WO3/C (36.2 eV and 39.6 eV), indicating a change in the chemical environment of tungsten. The nanocomposite also displayed peak broadening, particularly in W 4f5/2 (FWHM of WO3 = 2.3 eV; WO3/C = 6.7 eV), further confirming alterations in the number of chemical bonds42. A third peak at 34.3 eV was also present in the nanocomposite and could be ascribed to Wx+ (where 4 < x < 6). A similar peak was also observed in another work wherein WO3-carbon nanotubes showed tetragonal WO3 in its XRD43, similar to this work. They found that the existence of Wx+ was beneficial for increasing conductivity and, thereby, electrochemical performance. It is also worth mentioning that the WO3/C nanocomposite had higher-intensity W 4f. peaks than the pristine WO3. This might be explained by the increase in the effective surface area of the nanocomposite since carbon materials often have higher surface areas than metal oxides. The more intense XPS signals could be due to the larger fraction of the surface being probed during the analysis. This higher intensity was also observed in the O 1s spectra of the samples (Fig. S4). Moreover, the WO3/C nanocomposite showed a high-intensity C 1s peak, which deconvoluted to two peaks at 284.1 eV and 288.1 eV (Fig. 4C). The former could be ascribed to C–C, C=C, and C–H bonds while the latter to C=O bonds44. The presence of these peaks agrees with the FTIR results and further proves the aromatization of carbon during synthesis.

Figure 4
figure 4

XPS spectra of W 4f of (A) WO3 and (B) WO3/C. (C) C 1s spectrum of WO3/C nanocomposite.

Full size image

The surface area and pore characteristics were examined by nitrogen adsorption–desorption analysis, as presented in Fig. 5. The isotherm of pure WO3 closely resembles a type II isotherm, suggesting that it is an aggregation with predominantly macroporous features45. The adsorption amount of N2 for the WO3/C nanocomposite significantly increased, displaying a type IV isotherm with an apparent hysteresis loop at a relative pressure range of 0.45–0.95, suggesting the presence of abundant mesopores. The presence of such mesopores was further confirmed in the pore size distribution plot, revealing a range of pore radii between 1.5 and 15.0 nm with the highest peak occurring at 2.0 nm. Mesopores (ranging from 2 to 50 nm based on the IUPAC categorization) play a crucial role in enabling the migration of ions towards smaller micropores (those less than 2 nm in size). These facilitate the smooth transportation of ionic substances and the interconnected pore structure supports the formation of an electric double-layer during the charging process 46. The surface area and pore volume, calculated by the density functional theory (DFT) method, were 4.1 m2/g and 0.02 cc/g for WO3 while 58.5 m2/g and 0.09 cc/g for WO3/C nanocomposite.

Figure 5
figure 5

Nitrogen adsorption–desorption isotherms (A) and DFT pore size distribution curves (B) of WO3 and WO3/C nanocomposite.

Full size image

The electrochemical charge storage properties of WO3 and WO3/C nanocomposite were evaluated using cyclic voltammetry (CV), galvanostatic charge–discharge (GCD) technique, and impedance spectroscopy. Figure 6A and B display the CV curves of the WO3 and WO3/C electrodes between the potential range of − 0.90 to 0.90 V at different scan rates ranging from 20 mV/s to 200 mV/s. The anodic and cathodic currents increased with higher scan rates, which is a normal occurrence in CV. The comparative CV curves (Fig. 7) at a lower scan rate of 20 mV/s show that the WO3/C electrode maintains a larger area under the CV curve compared to WO3, indicating better capacitance. Additionally, a control sample containing only carbon particles was also prepared and similarly showed lower capacitance than the nanocomposite (Fig. 7, S2B, S6, S7). Moreover, the quasi-rectangular shape of the CV curve suggests pseudocapacitance. Equation (4) represents the electrochemical charge storage mechanism of WO3 in the KOH electrolyte47:

$${\text{WO}}_{{3}} + x{\text{K}}^{ + } + x{\text{e}}^{-} \leftrightarrow {\text{K}}x{\text{WO}}_{{3}}$$
(4)
Figure 6
figure 6

Cyclic voltammetry (CV) curves of (A) WO3 and (B) WO3/C nanocomposite at different scan rates.

Full size image
Figure 7
figure 7

Comparative CV curves at 20 mV/s scan rate of WO3/C, WO3, C, and background signal from the Cu substrate.

Full size image

With background correction from the copper substrate and the carbon control taken into consideration (Fig. 7), the WO3/C electrode showed an oxidation peak at 0.56 V to 0.76 V (peak E) and a reduction peak at − 0.20 V to − 0.50 V (peak G). While for the WO3 electrode, these peaks appeared at 0.0 V to 0.20 V (peak I) and − 0.60 V to − 0.75 V (peak K). This demonstrated the existence of reversible Faradaic reactions, suggesting an ion intercalation into the crystal structure of the metal oxide37. Peaks F and J could not be considered cathodic peaks since this was also present in the carbon control (peak L).

Interestingly, the WO3/C electrode showed an extra oxidation peak at 0.0 V to − 0.05 V (peak H), suggesting an additional irreversible Faradaic reaction. The redox behavior of WO3 and WO3/C aligned with the W 4f. XPS results, revealing two oxidation states for WO3/C and only one for WO3. The two oxidation peaks of the nanocomposite could be assigned to the electroactivity of W6+ and Wx+ (4 < x < 6). It was also observed that the redox peaks became less pronounced at higher scan rates (Fig. 6), which is due to the rapid charge kinetics caused by the high electric field37.

Using Eq. (1), the specific capacitance of the WO3 and WO3/C electrodes was calculated from the CV data. As shown in Fig. S7, the specific capacitance exponentially increased with decreasing scan rate. At a high scan rate of 200 mV/s, there was only a small difference between the specific capacitance of WO3 and WO3/C electrodes (14.4 F/g and 16.2 F/g). However, as the scan rate decreased, the difference became more apparent. At 20 mV/s, the specific capacitance of WO3/C increased to 75.1 F/g while WO3 increased to only 26.8 F/g. The values are comparable to that of reported in the literature (Supporting Information Table 1). The observed increase in capacitance at lower scan rates aligns with the typical rate performance in energy storage devices. Lower scan rates allow better diffusion of electrolyte ions to reach the cavities within the electrode material's internal structure while at higher scan rates, ions may get only surface immersion47,48. Overall, the higher capacitance of the WO3/C nanocomposite affirmed that it has more pathways for charge diffusion within its structure. These pathways were most likely a synergistic effect of the complex mixed-phase (tetragonal/monoclinic) WO3 forming a hierarchical structure with the mesoporous carbon microspheres, resulting in an expanded surface area. Both the presence of mesopores, which decrease ion transport resistance, and the Faradaic reactions significantly enhance electrochemical performance throughout the charging and discharging process49.

Figure 8 shows the GCD curves of the electrodes and their cycling stability. The deviation from the typical triangular shape further supported the pseudocapacitive behavior of the electrode materials. Using Eq. (2), WO3/C showed a higher coulombic efficiency than WO3. The former exhibited 98.2% efficiency at a current density of 1 A/g while the latter showed 75.8% at the same current density. The nanocomposite material also showed higher energy efficiency (92.8% at 1 A/g) than the pure WO3 (65.1%). Additionally, the cycling stability test demonstrated that the WO3 and WO3/C electrodes retained 68.5% and 83.2% of their capacitance, respectively, after 800 GCD cycles. This enhanced capacity retention in the nanocomposite affirmed that the inclusion of carbon contributes to the material's structural stability. Notably, the particle morphology of WO3/C remained unchanged after cycling, as depicted in Figure S8. However, the decrease in the performance for both electrodes could be explained by the possible distortion of the crystal lattice of tungsten oxide which could have adversely affected the charge transport 40.

Figure 8
figure 8

Galvanostatic charge–discharge (GCD) curves for (A) WO3 and (B) WO3/C electrodes. (C) GCD cycling stability.

Full size image

The charge transfer ability and interface resistance of the WO3 and WO3/C electrodes were studied by electrochemical impedance spectroscopy (EIS). Figure 9 shows the Nyquist plot of EIS measurements performed from 1 Hz to 100 kHz. The semicircular arc in the high-frequency region is indicative of the charge-transfer resistance (Rct) attributed to Faradaic reactions at the electrode/electrolyte interface6,50. WO3/C showed a lower resistance at Rct = 11.7Ω compared to WO3 at Rct = 14.0Ω. The arc could also be attributed to bulk electrolyte resistance (R∞) while the distance from the imaginary impedance axis (− Z″), to the electrode resistance (Re). The sum of these two accounts for the total internal resistance of the electrode51. The inset in Fig. 9 shows that WO3/C also demonstrated a lower Re than WO3. The conductivity was found to be 0.55 S/m for WO3 and 0.68 S/m for WO3/C. A sloped line related to Warburg resistance or diffuse layer resistance was present at the low-frequency region. From a physical interpretation of this line, steep slopes indicate that the dominating process is electric double-layer (EDL) formation, while it is ion diffusion at low slopes51. Interestingly, the WO3/C electrode showed a low slope (b in blue), but after 24 h of stabilization in the electrolyte the slope became steeper (d in light blue). But for WO3, the slope was already steep (a in red) and did not change much after 24 h (c in light red). This observation supports more channels in the nanocomposite arising from both the inner crystal structure and the amorphous carbon network. The relatively deep insertion of ions leads to a relatively longer time for the EDL to form, and thus ion diffusion was still the dominating process at the beginning40. After 24 h, the Rct for WO3/C and WO3 were 13.6Ω and 15.7Ω, respectively. Moreover, for WO3/C, the EIS spectra after 1000 CV cycles showed an Rct of 14.9Ω and the steepest slope, suggesting effective ion physisorption at the electrode/electrolyte interface.

Figure 9
figure 9

Nyquist plot of the electrochemical impedance. Inset: Emphasis on the difference.

Full size image

Conclusion

A facile hydrothermal method successfully prepared a hierarchical nano/microstructure WO3/C nanocomposite. Glucose, used as the carbon-source precursor, also influenced the crystal phase transformation of WO3. It contained a phase junction of tetragonal/monoclinic WO3 uniformly embedded on carbon microspheres and exhibited more oxidation states. Owing to this distinctive structure, the WO3/C electrode exhibited better electrochemical performance with a specific capacitance of 75.1 F/g compared to pure WO3 with 26.8 F/g at a scan rate of 20 mV/s in 0.1 M KOH. A pseudocapacitive behavior was observed, with WO3/C showing a high coulombic efficiency at 98.2% at a current density of 1 A/g. Cyclic voltammetry and impedance spectroscopy results suggested that the nanocomposite's energy storage mechanism showed both Faradaic and non-Faradaic capacitance.