Engineered S-scheme g-C₃N₄/MnO₂ heterostructures for integrated photo-rechargeable supercapacitors with enhanced energy storage performance

Abstract

Engineering a two-in-one multifunctional device that couples energy conversion and storage offers a smarter strategy to address the current global energy crisis while reducing reliance on grid electricity. Photo-rechargeable supercapacitors are perfect devices for the storage of light-induced electrochemical energy, garnering increasing attention as the next-generation energy storage technology. This study presents a novel 2D/1D g-C₃N₄/MnO₂-based photocathode architecture, reported for the first time, for the fabrication of a solid-state photo-rechargeable supercapacitor device. Here, g-C₃N₄ functions as the light-capturing component, while MnO₂ acts as the primary charge-storing element for the device. Photoluminescence (PL) results confirm that the MnO₂/g-C₃N₄ S-scheme architecture promotes efficient photoexcited charge separation and suppresses their recombination. Upon light illumination, the optimized device exhibits a ∼23% enhancement in areal capacitance, compared to its performance in the dark at 0.7 mA cm⁻². Under light exposure, the fabricated device retains double its areal capacitance after 600 cycles and achieves 100% retention after 2000 cycles under dark conditions, highlighting its outstanding cycling stability. This remarkable performance is ascribed to the presence of oxygen vacancy-mediated trap states in MnO₂, which reduce charge carrier recombination during light illumination and facilitate charge transfer kinetics. The proposed S-scheme charge transfer mechanism is further validated by the combined evidence from Scanning Kelvin Probe (SKP) and Mott–Schottky measurements. These findings emphasize the promise of the g-C₃N₄/MnO₂ S-scheme heterojunction for efficient light-assisted energy storage, making a significant advancement for an emerging class of materials. As the proof-of-concept, the device powered a red LED for 33 s in the dark and for up to 43 s under light illumination.

---

1. Introduction

Tackling the worldwide energy needs and environmental challenges due to excessive reliance on fossil fuels necessitates the engineering of sustainable technologies and green energy solutions. Among various renewable resources, solar energy is distinguished by its abundance, cleanliness, and inexhaustibility, highlighting its potential as a key green energy source.^1–3 Conventional photovoltaic devices, such as solar cells and photodetectors, can effectively convert sunlight into electric energy, but they cannot store the electric energy. This limitation often necessitates the integration of separate energy storage units, such as batteries or supercapacitors, which introduces issues like voltage mismatch, increased system bulk, extra wiring, and additional energy losses.⁴ To address these challenges, researchers have developed photo-rechargeable supercapacitors, capable of simultaneously harvesting solar energy and storing it within a single integrated system. These devices utilize a photocathode, which performs two essential functions: converting solar energy and storing electric charges effectively.⁵ The charge storage process in photo-rechargeable supercapacitors includes three key steps: (1) efficient light absorption and photoinduced charge carrier generation, (2) suppression of photoexcited charge carrier recombination, and (3) charge storage via a redox reaction or using an electric double layer.⁶

However, despite these steps, some photo-rechargeable systems suffer from sluggish mechanisms like higher recombination rates, limited ion diffusion, and energy band mismatch. These drawbacks critically impair the photo-charging efficiency and the overall device performance. To overcome these limitations, strategies such as heterojunction engineering (type-II heterojunction, S-scheme heterojunction, etc.) and conductive additives (rGO, CNTs, etc.) are widely explored to facilitate charge transfer and suppress recombination.⁷ Among various photoactive materials, graphitic carbon nitride (g-C₃N₄) is a well-known organic semiconductor with several advantages, such as facile synthesis, a narrow bandgap (∼2.7 eV), and a layered structure consisting of tri-s-triazine (heptazine) units interconnected by nitrogen atoms. However, the charge-storage ability of g-C₃N₄ is lower than those of other carbon-based electrodes. To enhance the charge storage ability and boost the photo-generated charge carrier separation, several other charge storage materials were incorporated.⁸ For example, Gupta et al. prepared a photoelectrode using a rGO–g-C₃N₄–ZnCo₂O₄ heterojunction, achieving an electrochemical enhancement of 25–30% under light illumination, attributed to the improved charge separation and conductivity.⁹ Similarly, Bai et al. synthesized a p–n heterojunction of Co₃O₄/g-C₃N₄, which exhibited a remarkable 70.6% performance enhancement under light illumination, owing to the internal electric field, which promotes redox reactions in the device.¹⁰ Among various works, Zhang et al. reported a graphene/MXene QD hybrid photo-rechargeable supercapacitor. Under UV illumination, photo-induced charge carriers generated in QDs were trapped across the graphene junction, effectively boosting charge accumulation. The capacitance of the device also remained at 105% after 5000 cycles.¹¹ Recently, Liu et al. reported an isotype heterojunction of g-C₃N₄ synthesized via a precursor-ratio-controlled polycondensation method for a photo-zinc-ion capacitor. The optimized photoelectrode achieved a specific capacity of 37.62 mAh g⁻¹, and also showed slow self-discharge with enhanced energy density.¹² The limitation of g-C₃N₄ can be addressed by incorporating it with high-charge storage materials with energy band matching. MnO₂, with its impressive theoretical specific capacitance (1370 F g⁻¹), was integrated to boost the device's charge storage properties and assist in the effective separation of charge carriers.^13,14 However, MnO₂ suffers from its limited conductivity and insufficient surface area.¹⁵ These drawbacks can be overcome by employing one-dimensional MnO₂ nanostructures, which provide rich active centers for the reaction and rapid electron mobility, thereby increasing the electrochemical performance.^16,17 Furthermore, MnO₂ combined with a hierarchical porous carbon network is widely investigated as a supercapacitor electrode.

In the current study, g-C₃N₄/MnO₂ composites were effectively synthesized through a simple hydrothermal method, utilizing different proportions of KMnO₄. The M/CN-3 : 1 electrode exhibited excellent electrochemical properties, which were further tested with photoelectrochemical studies. The M/CN-3 : 1 photocathode shows an enhanced areal capacitance of ∼23% under light illumination. The prepared solid-state device shows excellent photo-current response under light illumination, indicating that the photogenerated charge carriers enhanced the performance and exhibited a high areal energy density of 0.68 μWh cm⁻², surpassing its performance in the dark. This work provides a practical approach for developing photo-rechargeable supercapacitors with a deeper analysis of the mechanism involved, supported by SKP-derived work function measurements, marking the first report on MnO₂/g-C₃N₄-based photo-rechargeable supercapacitors.

2. Experimental details

2.1. Materials

Analytical grade reagents, such as melamine (C₃H₆N₆), potassium permanganate (KMnO₄), polyvinyl alcohol (PVA), hydrochloric acid (HCl), sodium sulphate (Na₂SO₄), potassium hydroxide (KOH), and potassium chloride (KCl), were used in the synthesis process, which were procured from Sigma-Aldrich. All chemical reagents were used without further purification. Additionally, all glassware was carefully washed with deionized water (DI) prior to use.

2.2. Synthesis of g-C₃N₄

10 g of melamine was taken in an alumina crucible and calcined at 550 °C for 4 h. After natural cooling, the obtained yellow product was finely ground using a pestle and mortar and labeled as “CN”.

2.3. Synthesis of MnO₂

The hydrothermal procedure was employed to synthesize pure α-MnO₂. In a typical procedure, 35 mM of KMnO₄ and 0.5 mL of concentrated HCl were dissolved in 65 mL of DI water. The solution was poured into an autoclave with a 100 mL capacity and heated at 150 °C in an oven for 8 h. After hydrothermal treatment, the solution was centrifuged with DI water and dried at 80 °C for 8 h, and the obtained sample was labelled as “M”.

2.4. Synthesis of MnO₂/g-C₃N₄

To synthesize MnO₂/g-C₃N₄ nanocomposites with varying compositions, 35, 105, and 175 mM of KMnO₄ were separately dissolved in 35 mL of DI water and stirred for 2 h. Additionally, 0.36 g of g-C₃N₄ was dispersed in 30 mL DI water and ultrasonicated for 30 min. The g-C₃N₄ solution was then mixed with KMnO₄ solution and stirred continuously for 1 h, during which 0.5 mL of concentrated HCl was gradually added. The solution was then poured into a 100 mL autoclave and subjected to a hydrothermal reaction at 150 °C for 8 h. The sample was then subjected to centrifugation with DI water and ethanol, followed by drying at 80 °C for 8 h. The final sample was designated as the MnO₂/g-C₃N₄ nanocomposite, with the samples labelled as M/CN-1 : 1, M/CN-3 : 1, and M/CN-5 : 1 according to the respective KMnO₄ content used during the synthesis.

2.5. Preparation of the photocathode

The photocathode was prepared by mixing M/CN-3 : 1, PVDF binder, and carbon black in an 8 : 1 : 1 ratio using NMP as the solvent. The resulting slurry was sonicated and stirred for 4 h. The prepared solution was drop-cast onto a pre-cleaned FTO substrate and subsequently dried at 80° C for 8 h.

2.6. Fabrication of the photo-rechargeable supercapacitor

To prepare a symmetric solid-state photo-rechargeable supercapacitor, the dried photocathode was assembled by sandwiching it with a separator and a PVA–KCl gel electrolyte. The constructed device was left to dry at room temperature to eliminate residual water from the gel electrolyte. The gel electrolyte was formulated by dissolving 10 wt% of PVA in 5 mL of DI water under constant stirring for 2 h at 80 °C. Subsequently, 1 M KCl was introduced into the mixture and continuously stirred to obtain a homogeneous solution (Scheme 1).

Scheme 1 Schematic representation of M/CN composites and fabrication of the photo-rechargeable supercapacitor.

3. Results and discussion

3.1. Structure and morphology analysis

Fig. 1(a) displays the XRD patterns of M, CN, and the M/CN-3 : 1 nanocomposite. For CN, the peaks observed at 12.66° and 27.03° correspond to the (100) and (002) planes, which are commonly associated with the stacking of conjugated aromatic systems and the tri-s-triazine unit, respectively (JCPDS card no: 00-087-1526).¹⁸ For the M sample, the prominent peaks at 12.6°, 17.9°, 28.6°, 36.7°, 37.3°, 41.7°, 49.6°, 55.9°, and 59.9° correspond to the (110), (200), (310), (400), (211), (301), (411), (431), and (521) planes, matching with the reference pattern of α-MnO₂ as indexed in JCPDS card no: 00-041-0141. The diffraction peaks observed in the M/CN-3 : 1 nanocomposite confirm the coexistence of both α-MnO₂ and g-C₃N₄.¹⁹ The XRD results of the M/CN-1 : 1 and M/CN-5 : 1 composites are provided in Fig. S1(a). Fig. 1(b) shows the FTIR analysis, which was employed to investigate the chemical structure and pinpoint the functional groups present in M, CN, and M/CN-3 : 1. The M sample shows peaks at 503 cm⁻¹ and 699 cm⁻¹, which are assigned to the bending and stretching vibrations of Mn–O, respectively.²⁰ Furthermore, the absorption peak located at 565 cm⁻¹ corresponds to the Mn–O–Mn vibration.²¹ In the case of the CN sample, a distinct peak observed at 810 cm⁻¹ is characteristic of the triazine ring. Numerous strong peaks were observed between 1220 cm⁻¹ and 1650 cm⁻¹, which are attributed to the stretching vibrations of C=N and C–N bonds within the aromatic heterocycle rings.²² The peaks positioned at 1536 cm⁻¹ and 1619 cm⁻¹ are attributed the C=N stretching vibration, and the broad peak observed at 3000 cm⁻¹–3400 cm⁻¹ in the FTIR spectrum of CN is ascribed to the stretching vibration of the NH bond or =NH bond.²³ The FTIR spectra of M/CN-1 : 1 and M/CN-5 : 1 are shown in Fig. S1(b). The FTIR spectra of the M/CN-3 : 1 sample show the characteristic peaks of both M and CN, confirming the successful preparation of the nanocomposite. Furthermore, zeta potential analysis was performed to measure the surface charges of the M, CN, and M/CN-3 : 1 samples. As shown in Fig. S2, all the samples exhibit negative surface charge values, where the M, CN, and M/CN-3 : 1 samples possess zeta potential values of −10.6 mV, −23.6 mV, and −26.7 mV, respectively. The M/CN-3 : 1 sample possesses a more negative zeta potential compared to the M and CN samples, which indicates local surface charge redistribution between M and CN while forming a heterojunction.²⁴ Furthermore, a higher zeta potential value indicates good stability of the material due to stronger repulsion among the particles.²⁵ The higher negative zeta potential value of the M/CN-3 : 1 sample suggests that the formed M/CN-3 : 1 heterostructure is the most stable. FE-SEM analysis was carried out to characterize the surface morphology of the as-prepared samples. As shown in Fig. 2(a), CN exhibits a two-dimensional sheet-like structure. Meanwhile, the α-MnO₂ sample exhibits a one-dimensional nanoneedle-like morphology, as depicted in Fig. 2(b). Fig. 2(c) presents the FE-SEM image of the M/CN-3 : 1 nanocomposite, where MnO₂ nanoneedles are anchored on the g-C₃N₄ sheets, resulting in the formation of a heterostructure. This intertwined morphology of MnO₂ and g-C₃N₄ promotes electron mobility during charging and discharging. Furthermore, the energy dispersive spectroscopy (EDS) mapping images of the M/CN-3 : 1 composite, as shown in Fig. 2(d), confirm the even distribution of M, O, C, and N elements across the heterostructure. The morphologies of M/CN-1 : 1 and M/CN-5 : 1 are provided in Fig. S3. Moreover, HR-TEM analysis of the M/CN-3 : 1 composite, as shown in Fig. 2(e–g), reveals a strong surface connection between MnO₂ and g-C₃N₄. This intimate contact is crucial in facilitating efficient separation and transfer of photoinduced excitons and prevents the recombination of charge carriers. The HR-TEM image of the composite displays lattice fringes corresponding to the (3 0 1) crystal facet of MnO₂, with an interplanar d-spacing of 0.30 nm, in agreement with its XRD result.²⁶ Additionally, the elemental composition from the EDS spectra (Fig. S4) confirms the existence of Mn, O, C, and N peaks with their respective weight percentages provided in the inset image.²⁷

Fig. 1 (a) XRD patterns of M, CN, and M/CN-3 : 1. (b) FTIR spectra of M, CN, and M/CN-3 : 1.

Fig. 2 (a–c) FE-SEM images of CN, M, and M/CN-3 : 1, (d) elemental mapping of M/CN-3 : 1, and (e–i) HR-TEM images of M/CN-3 : 1.

3.2. XPS analysis

X-ray photoelectron spectroscopy (XPS) was utilized to examine the chemical states and elemental composition of the M/CN-3 : 1 sample. As depicted in Fig. 3(a), the survey spectrum confirms the presence of Mn, O, C, and N as the constituent elements of M/CN-3 : 1. In the Mn 2p spectrum [Fig. 3(b)], the peaks at 642.6 eV and 654.2 eV correspond to Mn 2p_3/2 and Mn 2p_1/2, respectively. Furthermore, these two peaks are deconvoluted into four peaks via Gaussian and Lorentzian fitting, revealing the presence of Mn³⁺ at 642.43 eV and 654.3 eV, along with Mn⁴⁺ at 644.48 eV and 654.06 eV, thereby confirming the mixed-valence state of Mn.^28–31 The Mn 3s spectrum, as depicted in Fig. 3(c), reveals a spin-energy separation (ΔE) of 4.77 eV (ref. 32) obtained using the following eqn (1):

AOS = 8.956 − 1.126ΔE (1)

Fig. 3 (a) Survey spectrum of the M/CN-3 : 1 composite; (b) Mn 2p, (c) Mn 3s, (d) O 1s, (e) C 1s, and (f) N 1s XPS spectra of the M/CN-3 : 1 sample.

The resulting average oxidation state (AOS) of Mn is found to be 3.61, which further confirms the presence of Mn³⁺ and Mn⁴⁺. As shown in Table S1, the Mn³⁺/Mn⁴⁺ ratio is 2.32, indicating a higher proportion of Mn³⁺. This increased Mn³⁺ concentration correlates with oxygen vacancies, which act as electron traps and facilitate improved charge separation.³³ The deconvoluted peak of the O 1s spectrum, as shown in Fig. 3(d), exhibits three peaks at 529.87 eV responsible for lattice oxygen (Mn–O–Mn), 531.51 eV attributed to oxygen vacancies, and 533.75 eV corresponding to surface-adsorbed –OH groups (Mn–OH), respectively.^34–36 Hence, the prominent peak observed at 531.51 eV confirms the existence of oxygen vacancies in MnO₂. As displayed in Fig. 3(e), the C 1s spectrum is deconvoluted into the characteristic peaks positioned at 284.79 eV, 286.02 eV, and 288.23 eV corresponding to the C=C/C–C, C–NH_x, and N–C=N groups, respectively.^37,38 Similarly, the N 1s spectrum, as shown in Fig. 3(f), displayed three distinct peaks at 398.69 eV corresponding to the sp²-hybridized nitrogen (C=N–C), 400.43 eV attributed to (N-(C)₃), and 405.4 eV attributed to oxidised nitrogen species.^39–41

3.3. Band structure analysis and the plausible charge transfer mechanism

Work function measurements were performed using a Scanning Kelvin Probe (SKP) system, with M- and CN-coated FTO substrates positioned at a distance of 1 mm below a vibrating gold tip that served as a reference. The work function (Φ) is defined as the numerical difference between the vacuum energy level and the Fermi energy level. The average work function (Φ) of the samples was determined from the contact potential difference (CPD) measurements using eqn (2):⁴²

Φ = 5100 − CPD_Au + CPD_sample (2)

where 5100 is the work function of Au in meV, CPD_Au is the CPD between the tip and the gold reference, and CPD_sample is the CPD between the sample and the tip. As shown in Fig. 4(a), the calculated work function values of M and CN are 5.12 eV and 4.59 eV, respectively.^43,44 The higher Φ value of MnO₂ indicates that, upon contact, electrons spontaneously flow from CN to M until the Fermi level equilibrium is reached. These results from the work function measurements provide strong evidence for the formation of an S-scheme heterojunction between the M and CN samples.⁴⁵Fig. 4(b) illustrates the band bending and charge transfer between M and CN, which are further supported by the Mott–Schottky and UV–Vis spectroscopy results.

Fig. 4 (a) SKP 3D-raster scans of M and CN and (b) schematic illustration of a plausible charge transfer mechanism based on the SKP measurement.

Furthermore, UV-vis spectroscopy was used to investigate the optical properties of CN, M, and the M/CN-3 : 1 nanocomposite. As shown in Fig. 5(a), pristine CN exhibits an absorption edge at around 450 nm, and the M sample shows a broad absorption ranging from 200 nm to 800 nm. After the formation of a heterojunction, the M/CN-3 : 1 composite exhibits a higher absorbance intensity compared to M and CN, which indicates enhanced light absorption efficiency. To determine the band gap energies, the Tauc plot was obtained using the following relationship:⁴⁶

A(hν − E_g) = (αhν)^1/n (3)

where α stands for the absorption coefficient, h is Planck's constant, ν denotes the frequency of light, E_g represents the band gap, and A is a constant. The parameter n takes a value of either 1/2 or 2, depending on whether the bandgap of the material is direct or indirect. As displayed in Fig. 5(b and c), the estimated band gap values are 2.85 eV for CN and 1.69 eV for M. These results are in good agreement with the literature reports.^47–49

Fig. 5 (a) UV-Vis absorption spectra of CN, M, and M/CN-3 : 1, (b and c) Tauc plots of CN and M, (d) PL spectra of CN, M, and M/CN-3 : 1, (e and f) Mott–Schottky plots of CN and M, (g) plausible energy band position and charge carrier separation under light illumination, (h) schematic representation of the photodetector, (i) I–V plots of M/CN-3 : 1 in the dark and under 1 sun illumination, (j) schematic representation of the photo-rechargeable supercapacitor, (k) photocurrent response of the device under light illumination and dark conditions, and (l) ESR spectrum of M/CN-3 : 1 (inset: g-value).

Photoluminescence (PL) spectra highlight the recombination behaviour of the M, CN, and M/CN-3 : 1 samples. As depicted in Fig. 5(d), the PL spectrum of CN shows a broadband spectrum from 410 nm to 550 nm with two distinct peaks at 441 nm and 457 nm related to the band-to-band transition and inter-band defect states in CN.⁵⁰ Notably, CN shows the highest intensity, indicating the fastest recombination of charge carriers. Conversely, the PL intensity was notably quenched by about 50% in the M/CN-3 : 1 composite, indicating that MnO₂ incorporation effectively promotes charge separation. The lower intensity of the M sample is attributed to the photo-generated charge carriers being less compared to that of CN.⁵¹

To elucidate the band structure, Mott–Schottky measurements were performed on CN and M, as illustrated in Fig. 5(e) and (f). Both plots show a positive slope for CN and M, indicating n-type behaviour. The flat band potentials of CN and M were determined to be −1.53 V and 1.05 V, respectively, relative to Ag/AgCl using Nernst eqn (4):⁴⁸

where at 25 °C at pH = 7. The conduction band edge (E_CB) was found to be −0.9 V for CN and 1.63 V for M, relative to the RHE scale. By using the band gap values of CN and M samples from the Tauc plots, the valence band edge (E_VB) was calculated using eqn (5):

E_VB = E_CB + E_g (5)

The E_VB value was found to be 1.95 (V vs. RHE) for CN and 3.32 (V vs. RHE) for M, respectively. Finally, the E_RHE was converted into E_(vacuum) using eqn (6):⁵²

E_(vacuum) = −4.44 − E_RHE (6)

  Based on the Mott–Schottky analysis and work function measurements, the energy band alignment of M and CN samples was established. As presented in Fig. 4(b), before contact, the conduction band edge of CN lies at a higher potential than that of M. Moreover, CN possesses a lower work function compared to M, which drives the electrons from CN to M upon contact. The electron transfer occurs until the Fermi levels of both the materials reach equilibrium, inducing upward band bending in the CN sample and downward band bending in the M sample. Consequently, a built-in electric field is formed in the interfacial region. Fig. 5(g) further confirms the formation of an S-scheme heterojunction between M and CN. When exposed to light, the photogenerated holes of CN undergo recombination with the conduction band electrons of MnO₂. The photoexcited electrons in the E_CB of CN move to the external circuit and the photogenerated holes in the E_VB of MnO₂ attract the Cl⁻ ions in the electrolyte.⁵³

3.3.1. Evidence of charge separation under light conditions. Furthermore, the photoresponse and self-powered behaviour of the M/CN-3 : 1 sample were investigated using a prepared photodetector and the results are shown in Fig. 5(h). Current–voltage (I–V) measurements were conducted under both dark and light conditions, as shown in Fig. 5(i). The photodetector shows a higher photocurrent compared to the dark current, explaining the photosensitivity of the M/CN-3 : 1 sample. Also, the I–V curve doesn't intersect at 0 V, indicating the self-powered operation of the M/CN-3 : 1 photodetector.

To validate the charge carrier separation under zero voltage bias, chronoamperometry was performed on the prepared device without external voltage (V = 0). As shown in Fig. 5(k), the test was carried out under both dark and light conditions. Under light conditions, the photocurrent increased from 0.99 μA to 1.61 μA, demonstrating that the S-scheme heterojunction facilitates photoinduced charge carrier separation.

3.3.2. Electron paramagnetic resonance (EPR) analysis. To confirm the existence of oxygen vacancies in the as-prepared samples, electron paramagnetic resonance (EPR) spectroscopy was conducted. As shown in Fig. S5, MnO₂ (M sample) exhibits a resonance absorption signal at a g-value of 2.07, which closely agrees with the free electron g value of 2.0023, indicating the presence of oxygen vacancies associated with unpaired electrons. In comparison, the ESR spectrum of the M/CN-3 : 1 composite exhibits a higher intensity than that of MnO₂, as shown in Fig. 5(i). The enhanced ESR intensity indicates a higher concentration of oxygen vacancies in the composite, which facilitates photoinduced charge carrier trapping and transfer, thereby suppressing charge recombination. This contributes to improved stability and prolonged cycling durability, amplifying the overall performance of the M/CN-3 : 1 device.^54–56

3.4. Electrochemical performance of the M/CN-3 : 1 composite

3.4.1. Three-electrode system. The electrochemical behaviour of the M, M/CN-1 : 1, M/CN-3 : 1, and M/CN-5 : 1 electrodes was investigated in a conventional three-electrode setup. Fig. S6(a) displays the cyclic voltammetry (CV) curves recorded at 20 mV s⁻¹. All electrodes exhibit a pair of redox pseudocapacitive behaviour primarily arising from the reversible intercalation and deintercalation of OH⁻ ions between the KOH electrolyte and the prepared electrodes during the charge–discharge process.^57,58 Among all the samples, the M/CN-3 : 1 composite had the largest CV curve area, highlighting its better electrochemical performance relative to others. The galvanostatic charge–discharge (GCD) curve at 1 A g⁻¹, as shown in Fig. S6(b), further confirms the superior electrochemical performance of M/CN-3 : 1, as evidenced by its prolonged discharging time compared to other devices. Using eqn (S1), the specific capacitance was calculated to be 397.18 F g⁻¹ for M, 89.14 F g⁻¹ for M/CN-1 : 1, 555.12 F g⁻¹ for M/CN-3 : 1, and 523.66 F g⁻¹ for M/CN-5 : 1, as shown in Fig. S6(c). The enhanced performance of M/CN-3 : 1 is attributed to the optimized proportion between the ratio of KMnO₄ and g-C₃N₄, facilitating effective synergistic interaction and enhanced electrochemical activity.

Fig. 6(a) displays the CV curves of M/CN-3 : 1 recorded at various scan rates (5–100 mV s⁻¹). As the scan rate increases, the integral area of the CV curve expands, while the oxidation peak shifts towards more positive potential and the reduction peak shifts towards more negative potential. This behaviour signifies an increased polarisation effect and an increase in the interfacial resistance between the electrolyte and the electrode.⁵⁹ GCD curves were obtained at varying current densities ranging from 1 A g⁻¹ to 10 A g⁻¹, as displayed in Fig. 6(b). The specific capacitances of M/CN-3 : 1 were 555.12, 509.73, 480.84, 444.2, 412.58, and 373.4 F g⁻¹ at current densities of 1, 2, 3, 5, 7, and 10 A g⁻¹, respectively. The electrode retained 67.23% of its capacitance, demonstrating good rate capability. Fig. 6(c) shows the Nyquist plot of M/CN-3 : 1 measured over a frequency range between 100 kHz and 10 mHz, with the corresponding equivalent model circuit presented in the inset. Fig. S6(d) shows the Nyquist plots of the M, M/CN-1 : 1, M/CN-3 : 1, and M/CN-5 : 1 samples. The fitted values obtained from the equivalent circuit model shown in Fig. S6(e) reveal the solution resistance (R_s) and charge transfer resistance (R_ct) of the system. M/CN-3 : 1 shows a minimum R_s value of 1.03 Ω, indicating the minimum ohmic loss and increased rate capability compared to other electrodes.⁶⁰ The R_ct value of M/CN-3 : 1 (6.52 Ω) is slightly higher than that of M (5.38 Ω), which can be traded off against its high specific capacitance. In contrast, M/CN-1 : 1 and M/CN-5 : 1 show higher R_ct values, representing the less favourable electron transfer kinetics. The cycling stability of the M/CN-3 : 1 electrode was evaluated over 2000 continuous charge and discharge cycles at 7 A g⁻¹, in which the electrode maintained a cycling retention of 91.74% with a coulombic efficiency of 99.5%, as shown in Fig. 6(d). Owing to its excellent specific capacitance, superior rate performance, and low internal resistance, the M/CN-3 : 1 composite was further evaluated as a photocathode for the fabrication of a photo-rechargeable supercapacitor.

Fig. 6 (a) CV curve of M/CN-3 : 1, (b) GCD curve of M/CN-3 : 1, (c) Nyquist plot of M/CN-3 : 1 (inset: equivalent circuit), and (d) cycling stability of M/CN-3 : 1.

3.4.2. Photo-rechargeable M/CN-3 : 1 symmetric device. The electrochemical performance of the photocathode (M/CN-3 : 1) was evaluated by preparing a symmetric solid-state supercapacitor under dark and light conditions (1 sun). CV measurements were conducted at scan rates ranging from 10 mV s⁻¹ to 100 mV s⁻¹ over a potential range spanning from 0.0 V to 0.8 V. Fig. 7(a) demonstrates that exposure to light substantially boosts the charge storage performance of the device, as the photoinduced charge carriers actively contribute to charge accumulation. Importantly, as depicted in Fig. 7(b), the CV profiles remained consistent in the presence and absence of illumination, with the primary difference being an increased current density under illumination. Fig. S7 presents the CV curve recorded in the dark. Moreover, the CV curve profiles remained stable across various scan rates, further confirming the reversibility of the photocathode during charging and discharging.⁶¹ Fig. S8(a) illustrates the capacitance enhancement in the CV curve with and without light exposure at various scan rates. The capacitance enhancement was determined using eqn (7):where C_light and C_dark correspond to capacitances under light and dark conditions, respectively. Fig. S8(b) displays the highest areal capacitance of 19.8%, which was achieved at 10 mV s⁻¹. Under increased scan rate conditions, notably 100 mV s⁻¹, a notable enhancement to 15.9% was maintained, with only a 3.9% decrease, which surpasses the values reported in the previous literature. To further investigate the photocathode behaviour, GCD measurements were conducted at various current densities between 0.3 mA cm⁻² and 0.7 mA cm⁻² under both dark and illuminated conditions, as presented in Fig. 7(d) and (e). As supported by the CV results, the light illumination led to an improvement in areal capacitance.

Fig. 7 (a) CV curve of M/CN-3 : 1 under dark and light conditions, (b) CV curve of M/CN-3 : 1 at different scan rates under light illumination, (c) GCD curve of M/CN-3 : 1 under dark and light conditions, (d) GCD curve of M/CN-3 : 1 at various current densities under light illumination, (e) comparison of areal capacitance under light and dark conditions, and (f) capacitance increment under light illumination at various current densities.

For instance, upon light irradiation, the areal capacitance increased from 6.75 mF cm⁻² to 7.65 mF cm⁻² with a capacitance enhancement of 13.3% at 0.3 mA cm⁻². Moreover, as the current density increases, the enhancement in areal capacitance becomes more pronounced. At 0.7 mA cm⁻², the capacitance increased from 2.71 mF cm⁻² to 3.33 mF cm⁻² under illumination, representing a substantial enhancement of 22.9%, as depicted in Fig. 7(f). These results demonstrate the significant contribution of photogenerated charge carriers to the enhanced charge storage behaviour under light illumination.

To investigate the electrochemical kinetics and charge storage characteristics of the photo-rechargeable supercapacitor, the correlation between the current response (i) and the scan rate (ν) was analyzed using the power-law relationship:⁶²

i(ν) = aν^b (8)

log(i) = b log(ν) + log(a) (9)

where a and b are the variable parameters, with i representing the current and ν indicating the scan rate, respectively. The b-value provides a perspective on the underlying charge accumulation mechanism and the value was determined by the logarithmic transformation of eqn (8). Fig. 8(a) presents the log(i) versus log(ν) plot, where the linear fitting is used to determine the b-value, ranging from 0 to 1. A b-value of ∼1 suggests that the charge storage is mainly capacitive, governed by surface reactions, whereas a b-value of ∼0.5 corresponds to a diffusive-controlled process primarily involving ion intercalation. The obtained b-value from Fig. 8(a) is 0.55 (dark) and 0.56 (under 1 Sun). These represent the contribution mainly from diffusive processes. The slight increase in the b-value under light illumination may be attributed to photogenerated charge carriers, facilitating redox reactions and enhancing capacitive behaviour.⁶³ To further quantify the contribution process, Dunn's method was employed:⁶²

i = k₁ν + k₂ν^1/2 (10)

Fig. 8 (a) The b-value of M/CN-3 : 1 under light and dark conditions, (b and c) capacitive contribution of the CV curve at 10 mV s⁻¹ and quantitative analysis of capacitive contribution and diffusive contribution under light illumination, (d and e) capacitive contribution of the CV curve at 10 mV s⁻¹ and quantitative analysis of capacitive contribution and diffusive contribution under dark conditions, (f) Nyquist plot of M/CN-3 : 1 under dark and light conditions with fitted curves (inset: EIS equivalent circuit), (g) Z′ vs. ω^−1/2 plots under dark and light conditions, and (h) cycling stability of M/CN-3 : 1 under light and dark conditions.

At a specific potential, the total current i is the sum of the capacitive contribution (k₁ν) and the diffusive contribution (k₂ν^1/2). By dividing eqn (10) by ν^1/2,

i/ν^1/2 = k₁/ν^1/2 + k₂ (11)

Furthermore, the charge storage behaviour of the prepared device was elucidated. From eqn (11), the capacitive contribution of the device under light illumination was calculated to be 59.5% at 10 mV s⁻¹, as depicted in Fig. 8(b). Furthermore, this contribution increases from 59.5% to 82.3% as the scan rate increases from 10 mV s⁻¹ to 100 mV s⁻¹, as presented in Fig. 8(c). As illustrated in Fig. 8(d), the capacitive contribution under dark conditions reached 58.3% at a scan rate of 10 mV s⁻¹, which increased to 81.5%, as presented in Fig. 8(e). With increasing scan rates, the capacitive contribution increases and the diffusive contribution decreases, suggesting limited ion diffusion into the photocathode at higher rates.¹⁷

Further validation of light-enhanced charge storage kinetics was performed through EIS. Fig. 8(f) presents the Nyquist plots measured in the presence and absence of light, and the equivalent circuit with fitted parameters is presented in Table S2. Notably, the R_s and R_ct values measured in the dark are 54.06 Ω and 330.42 Ω, respectively, which notably exceed the values under illumination (1 Sun), where the R_s and R_ct values decrease to 34.09 Ω and 195.77 Ω, respectively. Under light conditions, the R_ct value decreased by 40.8% compared to the dark conditions, accompanied by a reduction in the Warburg resistance (W) as well. The constant phase element (Q) increased from 0.398 in the dark to 0.595 under light illumination, indicating the transition towards capacitive behaviour, which aligns with the findings from Dunn's method (Q ∼ 0.5 suggests diffusion-controlled and Q ∼ 1 suggests capacitive-controlled).⁶⁴ The diffusion coefficient of Cl⁻ ions was calculated using the Warburg impedance derived from the plot of Z′ (ohm) versus ω^−1/2, following the Warburg equation (eqn (S7)). As shown in Fig. 8(g), the Warburg coefficient (σω) decreased from 214.34 Ω s^−1/2 in the dark to 200.16 Ω s^−1/2 under light illumination, caused by the development of an internal electric field at the interface. Correspondingly, the Cl⁻ diffusion coefficient improved from 1.09 × 10⁻¹⁵ cm² s⁻¹ (dark) to 1.25 × 10⁻¹⁵ (light), confirming that the resistance decreased between the electrode and electrolyte due to the less polarization concentration.^10,65

To assess the cycling performance of the fabricated symmetric device, the stability was monitored over 2600 cycles at 0.6 mA cm⁻², as displayed in Fig. 8(h). During the initial 600 cycles under light illumination, the device showed a notable enhancement in capacitance, achieving a retention of 205.9%, reflecting a huge enhancement. However, upon turning off the illumination, the retention decreased to 112%, confirming that the enhancement is purely light-dependent. Furthermore, Ex situ FE-SEM and XPS analyses were performed to examine the morphological and chemical state changes of the M/CN-3 : 1 electrode before and after 600 cycles under illumination. As shown in Fig. S9, the overall structure of the M/CN-3 : 1 electrode remains well preserved, showing CN nanosheets and MnO₂ nanorods, confirming its structural stability. This indicates that the enhancement in the cycling stability is not due to morphological changes but rather to variations in the chemical state of the electrode. As depicted in Fig. S10, the high-resolution Mn 2p spectra reveal that the proportion of Mn³⁺ increases from 65.8% to 85.7% after 600 cycles under light illumination. This results in an enhancement in the redox activity during the cycling. A similar trend was reported by Yu et al., where an increase in the oxidation state of V⁵⁺ was observed during prolonged cycling, further supporting the chemical-state driven improvement in the charge storage behaviour.⁶⁶ After 600 cycles under light illumination, the device was further cycled in the dark, showing an excellent stability of 100.4% after 2000 cycles. Fig. 9(a) illustrates the Ragone plot, comparing the energy and power densities of the device under illumination and dark conditions against those of previously reported systems. Under light illumination, an optimal energy density of 0.68 μWh cm⁻² was delivered by the device at the corresponding power and current densities of 120.22 mW cm⁻² and 0.3 mA cm⁻² based on eqn (S4) and (S5). In contrast, under dark conditions, the corresponding energy density was 0.60 μWh cm⁻² at a power density of 120.14 mW cm⁻².

Fig. 9 (a) Comparison of the energy density and power density of M/CN-3 : 1 with other devices,^67–69 (b) plausible charge transfer process in the prepared device under light illumination, and (c and d) digital photos of powering the red LED under dark and light conditions with a series of 4 cells.

The plausible photo-rechargeable mechanism of the fabricated device, along with the band structure, is illustrated in Fig. 9(b). Upon light illumination, both CN and MnO₂ capture photon energy, resulting in the generation of electron–hole pairs. Due to the favourable alignment of their band structures, an S-scheme heterojunction is established at the g-C₃N₄/MnO₂ interface. In this configuration, the photoinduced holes present in the valence band of g-C₃N₄ recombine with the photoexcited electrons of MnO₂. Meanwhile, a portion of photogenerated electrons in MnO₂ becomes trapped in oxygen vacancies, which inhibits the fast recombination and facilitates more efficient charge separation. The conduction band electrons of CN drift towards the FTO substrate and are transferred via the external circuit. Concurrently, the Cl⁻ ions present in the electrolyte attract the valence band holes of MnO₂, resulting in the formation of an electric double layer. In the opposite electrode, electrons attract the K⁺ ions from the electrolyte. The synergistic formation of an S-scheme heterojunction coupled with defect-mediated traps enables improved photoexcited charge separation, resulting in enhanced electrochemical performance of the fabricated device. Fig. 9(c and d) display a series of four cells powering a red LED. Under light illumination, the intensity of the LED increases compared to the dark. Moreover, the LED remained powered for 43 seconds under light illumination, whereas it lasted only 33 seconds in the dark (SI Video). This proof-of-concept experiment verifies that the photo-generated carriers enhance the capacitance of the device. At a relatively higher current density of 0.3 mA cm⁻², the M/CN-3 : 1 device achieves a superior areal capacitance of 7.65 mF, outperforming previously reported photo-rechargeable devices presented in Table 1, with excellent stability under light illumination.

Table 1 Comparison of the present work with recent advancements in photo-rechargeable supercapacitors

S. no.	Electrode	Light source	Areal capacitance under light illumination	Capacitance enhancement (%)	Cycling stability	Ref.
1	Co3O4/g-C3N4	Solar simulator	82.2 mF cm^−2 at 10.6 mA cm^−2	19.2	83.3% after 5000 cycles	10
2	NM2P1	PLS-LED100 (150 mW cm^−2)	466 F cm^−3 at 10 A cm^−3	24.3	—	70
3	MXene QD/graphene	LED (365, 420, and 475 nm)	10.2 μF cm^−2 at 0.30 μA cm^−2	70.7	105% after 5000 cycles	71
4	pC3N4/MXene	LED (420 nm)	500 μF cm^−2 at 10 μA cm^−2	1960	—	67
5	rGO–g-C3N4–ZnCo2O4 (3-electrode)	Solar simulator	3472 mF g^−1 at 10 mA g^−1	30	—	9
6	M/CN-3 : 1	Solar simulator	7.65 mF cm^−2 at 0.3 mA cm^−2	22.9 at 0.7 mA cm^−2	205.9% after 600 cycles (light)	This work
					100.4% after 2000 cycles (dark)

---

4. Conclusion

In this research, a symmetric solid-state photo-rechargeable supercapacitor was successfully prepared by utilising an M/CN-3 : 1 composite coated on an FTO substrate. An areal capacitance of 7.65 mF cm⁻² was achieved for the heterostructure at 0.3 mA cm⁻² under light illumination, reflecting an enhancement compared to its performance in the dark. Furthermore, an areal energy density of 0.68 μWh cm⁻² was achieved by the device under light illumination. This enhancement is ascribed to the integration of MnO₂ and g-C₃N₄via an S-scheme heterostructure, confirmed by SKP-derived work function measurements, while PL and photocurrent studies further verify the improved photogenerated charge carrier separation. The oxygen vacancies in MnO₂ boost the stability of the device by serving as charge carrier trap states, thereby effectively prolonging their lifetime. Consequently, the device showed a retention of 206% after 600 cycles under light illumination, along with excellent stability, maintaining nearly 100% retention after 2000 cycles in the dark. Overall, this work highlights the functionality of MnO₂/g-C₃N₄ as a promising photocathode for next-generation supercapacitors.

Author contributions

P. Chinnappan Santhosh: writing – original draft, methodology, formal analysis, data curation, and investigation. Suresh Jayakumar: formal analysis and methodology. A. V. Radhamani: supervision, conceptualization, funding acquisition, and review and editing.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

Data will be made available upon reasonable request.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d5nr03958d.

Acknowledgements

A. V. R. gratefully acknowledges the financial support from the Government of India through the start-up research grant (SRG/2021/000629), which facilitated the research presented in this paper.

References

1. S. Jana, S. Karingula, A. Sajeevan, V. V. N. Phani Kumar and Y. G. Kotagiri, Nanoscale, 2025, 17, 13283–13297.
2. N. Tewari, S. B. Shivarudraiah and J. E. Halpert, Nano Lett., 2021, 21, 5578–5585.
3. S. Ghosh, S. Bera, A. Kapuria, A. Debnath, P. Das, Y. K. Su and S. K. Saha, Nanoscale, 2025, 17, 15731–15748.
4. B. Luo, D. Ye and L. Wang, Adv. Sci., 2017, 4, 1–15.
5. Z. Hosseini, M. M. Momeni, B. K. Lee and A. W. Maijenburg, Chem. Eng. J., 2025, 515, 163566.
6. P. C. Santhosh, S. Jayakumar, M. M. Mohideen and A. V. Radhamani, Mater. Res. Bull., 2024, 174, 112722.
7. S. Kumar, G. Saeed, L. Zhu, K. N. Hui, N. H. Kim and J. H. Lee, Chem. Eng. J., 2021, 403, 126352.
8. Y. Li, Z. Xia, Q. Yang, L. Wang and Y. Xing, J. Mater. Sci. Technol., 2022, 125, 128–144.
9. G. Gupta, S. Hossain, N. Khatun, K. Gaurav, S. Alwarappan and S. C. Roy, ACS Omega, 2025, 10(4), 3729–3739.
10. L. Bai, H. Huang, S. Zhang, L. Hao, Z. Zhang, H. Li, L. Sun, L. Guo, H. Huang and Y. Zhang, Adv. Sci., 2020, 7, 1–9.
11. A. Das Mahapatra, S. Kumar, P. S. Chauhan, A. Mondal, A. Sutradhar and A. Misra, ACS Appl. Energy Mater., 2024, 7, 5801–5810.
12. X. Liu, Y. Zhu, F. Alam, T. Wang, I. P. Parkin, M. Wang and B. Deka Boruah, J. Mater. Chem. A, 2024, 25801–25811.
13. T. Yue, B. Shen and P. Gao, Renewable Sustainable Energy Rev., 2022, 158, 112131.
14. T. Xiong, W. S. V. Lee and J. Xue, ACS Appl. Energy Mater., 2018, 1, 5619–5626.
15. S. Jayakumar, P. C. Santhosh, M. M. Mohideen and A. V. Radhamani, J. Alloys Compd., 2024, 976, 173170.
16. C. Poochai, C. Sriprachuabwong, J. Sodtipinta, J. Lohitkarn, P. Pasakon, V. Primpray, N. Maeboonruan, T. Lomas, A. Wisitsoraat and A. Tuantranont, J. Colloid Interface Sci., 2021, 583, 734–745.
17. Y. Li, W. Sun, B. Gao, Y. Meng, X. Wang, X. Z. Song and Z. Tan, ACS Appl. Energy Mater., 2024, 7(23), 11126–11134.
18. S. Vinoth, K. Subramani, W. J. Ong, M. Sathish and A. Pandikumar, J. Colloid Interface Sci., 2021, 584, 204–215.
19. J. Wang, X. Guo, R. Cui, H. Huang, B. Liu, Y. Li, D. Wang, D. Zhao, J. Dong, S. Li and B. Sun, ACS Appl. Nano Mater., 2020, 3, 11152–11159.
20. N. C. Maile, S. B. Mahadik, M. V. Takale and V. J. Fulari, Mater. Res. Express, 2019, 6, 045204.
21. M. Waqas, A. Shahzadi, A. Haider, A. U. Hamid, M. M. Algaradah, H. S. M. Abd-Rabboh and M. Ikram, Int. J. Biol. Macromol., 2023, 251, 126343.
22. M. Subbiah, A. Radhakrishnan, M. Tharmaraj, A. Lakshmanan, S. Venkatachalam and N. Srinivasan, J. Alloys Compd., 2024, 1008, 176689.
23. L. G. Ghanem, M. A. Hamza, M. M. Taha and N. K. Allam, J. Energy Storage, 2022, 52, 104850.
24. J. Jing, J. Yang, Z. Zhang and Y. Zhu, Adv. Energy Mater, 2021, 2101392, 1–7.
25. M. Tayyab, Y. Liu, S. Min, R. Muhammad and Q. Zhu, Chin. J. Catal., 2022, 43, 1165–1175.
26. N. Zhang, W. He, Z. Cheng, J. Lu, Y. Zhou, D. Ding and S. Rong, Chem. Eng. J., 2023, 466, 143160.
27. R. Kumar and R. Thangappan, J. Energy Storage, 2023, 70, 108149.
28. Z. X. Huang, X. L. Zhang, X. X. Zhao, H. Y. Lü, X. Y. Zhang, Y. L. Heng, H. Geng and X. L. Wu, J. Mater. Sci. Technol., 2023, 160, 9–17.
29. M. Liu, M. Lin, P. Wu, G. Owens and Z. Chen, Sep. Purif. Technol., 2023, 310, 123052.
30. B. Feng, C. Qin, Y. Shi and L. Zhang, Energy Fuels, 2022, 36, 9221–9229.
31. X. Han, W. Zhang, S. Li, C. Cheng, Q. Yu, Q. Jia, L. Zhou and G. Xiu, J. Environ. Sci., 2023, 130, 197–211.
32. A. Zhang, R. Zhao, L. Hu, R. Yang, S. Yao, S. Wang, Z. Yang and Y. M. Yan, Adv. Energy Mater., 2021, 11, 1–11.
33. H. Ma, X. Li, M. Tian, L. Li, X. Yu, X. Zhang, Z. Lu and X. Yang, J. Mater. Sci. Technol., 2024, 201, 139–148.
34. R. Dawar, S. Narang, K. Bhattacharyya, P. M. Aiswarya, D. P. Dutta and R. Mishra, Inorg. Chem., 2024, 63, 14438–14448.
35. J. Lu, H. Wang, Y. Sun, X. Wang, X. Song and R. Wang, Chem. Eng. J., 2021, 417, 127894.
36. T. Wen, J. Wang, J. Zhang and C. Long, Environ. Res., 2023, 238, 117176.
37. S. Gao, X. Wang, C. Song, S. Zhou, F. Yang and Y. Kong, Appl. Catal., B, 2021, 295, 120272.
38. T. Shan, Y. Wang, D. Luo, Z. Huang, F. Zhang, H. Wu, L. Huang, J. Li, L. Chen and H. Xiao, Appl. Catal., B, 2024, 349, 123872.
39. S. Juntrapirom, S. Anuchai, O. Thongsook, S. Pornsuwan, P. Meepowpan, P. Thavornyutikarn, S. Phanichphant, D. Tantraviwat and B. Inceesungvorn, Chem. Eng. J., 2020, 394, 124934.
40. M. Majdoub, Z. Anfar and A. Amedlous, ACS Nano, 2020, 14, 12390–12469.
41. F. Li, B. Xu, X. You, G. Gao, R. Xu, X.-L. Wang and Y.-F. Yao, Int. J. Hydrogen Energy, 2023, 48, 12299–12308.
42. M. S. Shabanur Matada, G. P. Kuppuswamy, S. Sasi, S. Velappa Jayaraman, V. Nutalapati, S. Senthil Kumar and Y. Sivalingam, ACS Appl. Mater. Interfaces, 2024, 16, 17219–17231.
43. C. Shuai, S. Zhong, Y. Shuai, W. Yang, S. Peng and C. He, J. Colloid Interface Sci., 2023, 632, 95–107.
44. Y. Che, B. Weng, K. Li, Z. He, S. Chen and S. Meng, Appl. Catal., B, 2025, 361, 124656.
45. H. Zhang, Y. Gao, S. Meng, Z. Wang, P. Wang, Z. Wang, C. Qiu, S. Chen, B. Weng and Y. Zheng, Adv. Sci, 2024, 2400099, 1–13.
46. P. Makuła, M. Pacia and W. Macyk, J. Phys. Chem. Lett., 2018, 9, 6814–6817.
47. M. Wilhelm, R. Adam, A. Bhardwaj, I. Neumann, S. H. Cho, Y. Yamada, T. Sekino, J. Tao, Z. Hong, T. Fischer and S. Mathur, Adv. Eng. Mater., 2023, 25, 1–9.
48. A. S. Golda, A. P. Varghese, N. Rabiee, B. Neppolian and S. K. Lakhera, Carbon, 2023, 215, 118441.
49. S. Lu and H. Liu, Langmuir, 2024, 40(26), 13331–13338.
50. A. S. Golda, A. P. Varghese, N. Rabiee, B. Neppolian and S. K. Lakhera, Carbon, 2023, 215, 118441.
51. M. Wang, M. Shen, L. Zhang, J. Tian, X. Jin, Y. Zhou and J. Shi, Carbon, 2017, 120, 23–31.
52. W. Guo, W. Shi, J. Cai, F. Wei, X. Lin, X. Lu, Z. Ding, Y. Hou, G. Zhang and S. Wang, Catal. Sci. Technol., 2024, 14, 2921–2928.
53. B. D. Boruah and A. Misra, ACS Appl. Energy Mater., 2019, 2, 278–286.
54. I. Chihi, M. Baazaoui, N. Hamdaoui, J. M. Greneche, M. Oumezzine and K. Farah, J. Mater. Sci.: Mater. Electron., 2021, 32, 16634–16647.
55. W. Liu, Y. Pang, Z. Shi, H. Yue, H. Dong, Z. Cao, Z. Yang, S. Yang and Y. Yin, ACS Appl. Mater. Interfaces, 2022, 14, 11575–11586.
56. B. Kanrar, N. Pathak, C. Nayak, C. L. Prajapat, R. M. Kadam, S. N. Jha, D. Bhattacharyya, G. Ravikumar and N. L. Misra, J. Mater. Chem. C, 2017, 5, 8836–8846.
57. M. Pourshahmir, S. Ghasemi and S. R. Hosseini, Int. J. Hydrogen Energy, 2023, 48, 8127–8143.
58. G. V. Dilwale, A. C. Pawar, G. Piao, Q. Wang, Z. Said, R. K. Nimat, J. Bin In, J. M. Kim and R. N. Bulakhe, Energy Fuels, 2024, 38, 13355–13364.
59. Z. Zhang, X. Ren, Y. Wang, C. Luo, G. Wang, C. Xu and H. Chen, J. Energy Storage, 2025, 116, 116091.
60. M. M. Mujawar, V. V. Patil, V. S. Kumbhar, U. M. Patil, N. R. Chodankar, A. Al Ghaferi, M. Nakayama and J.-J. Shim, J. Power Sources, 2025, 642, 236980.
61. M. Benali, J. Azadmanjiri, M. Loula, Z. Liao, R. Gusmão, A. Subramani, K. J. Sarkar, R. Boukherroub and Z. Sofer, ACS Appl. Nano Mater., 2024, 7, 14102–14114.
62. M. Najafi, M. Mohsen Momeni and B. K. Lee, Chem. Eng. J., 2024, 493, 152423.
63. D. Zheng, X. Sun, C. An, F. Pan, C. Qin, Z. Wang, Q. Deng, Y. Song and Y. Li, Chem. Eng. J., 2023, 473, 145289.
64. G. P. Scisco, M. E. Orazem, K. J. Ziegler and K. S. Jones, J. Power Sources, 2021, 516, 230700.
65. X. Gao, D. Tian, Z. Shi, N. Zhang, R. Sun, J. Liu, H. S. Tsai, X. Xiang and W. Feng, Small, 2024, 2405627, 1–8.
66. M. Yu, Y. Zeng, Y. Han, X. Cheng, W. Zhao, C. Liang, Y. Tong, H. Tang and X. Lu, Adv. Funct. Mater., 2015, 25, 3534–3540.
67. S. Kumar, A. Mondal, V. Panwar, R. Shekhawat and A. Misra, Batteries Supercaps, 2024, 7, 1–11.
68. A. Das Mahapatra, S. Kumar, A. Sutradhar, S. Sahoo and A. Misra, Electrochim. Acta, 2023, 474, 143507.
69. R. Sinha, N. Roy and T. K. Mandal, Chem. Eng. J., 2022, 431, 133915.
70. H. Wang, J. Cao, Y. Zhou, X. Wang, H. Huang, Y. Liu, M. Shao and Z. Kang, Nano Res., 2021, 1–7.
71. A. Das Mahapatra, S. Kumar, P. S. Chauhan, A. Mondal, A. Sutradhar and A. Misra, ACS Appl. Energy Mater., 2024, 7, 5801–5810.

---