Strategic intercalation of AB₂O₄ perovskite oxides for synergetic enhanced redox activity in sulphonated Ti₃C₂T_x MXene for energy storage applications

Abstract

Flexible supercapacitors have emerged as efficient and fast energy storage devices for new generation electronics. Two-dimensional (2D) transition metal carbides (MXene) have garnered attention as supercapacitor electrodes owing to their conductive layered sheets and the tunability of their surface functional groups. In the present work, the Ti₃C₂T_x MXene surface was sulphonated using dimethyl sulfoxide (DMSO) and intercalated with AB₂O₄ (A = Co and Ni; B = =Fe) perovskite nanoparticle (NPs). The sulphonated MXene (TMS) was processed using a sonication method in DMSO solvent to enhance the surface area and redox active sites for electrolyte (0.1 M H₂SO₄) interaction. The redox dominated enhancement in specific capacitance was observed in 3 wt% CoFe₂O₄ (CFO)-intercalated TMS (3CTMS) and 3 wt% NiFe₂O₄ (NFO)-intercalated TMS (3NTMS), as confirmed by Electrochemical Impedance Spectroscopy (EIS) and Dunn's method analysis. The specific capacitance of 3CTMS was found to be 593.81 F g⁻¹ at 5 mV s⁻¹, with an excellent cyclic stability of 81.75% after 10 000 cycles. A flexible symmetric supercapacitor fabricated with 3CTMS showed an energy density of 4.177 W h kg⁻¹ and a power density of 512.17 W kg⁻¹. The flexible supercapacitor has been utilized in real time applications by charging and discharging to power 5 Light-Emitting Diodes (LEDs) with different forward voltages.

---

1 Introduction

Flexible supercapacitors have gained much attention due to their capacity to meet the demands of wearable and portable electronic devices. One primary challenge in developing flexible supercapacitors is producing electrodes with superior cyclic stability.^1,2 Producing flexible supercapacitors with outstanding electrochemical properties is critical for meeting the increased demand for wearable and portable devices.³ Based on the charge storage mechanism, supercapacitors are classified as electrochemical double-layer capacitors (EDLCs) and pseudocapacitors. EDLCs use porous carbon electrode–electrolyte interfaces for dual-layer energy storage,⁴ whereas pseudocapacitors rely on reversible faradaic processes or ion adsorption on and within the electrode's surface. EDLCs typically exhibit lower energy density and cyclic stability compared to pseudocapacitors but they have higher power densities.⁵ The electrochemical and mechanical performance of flexible supercapacitors strongly relies on their electrode materials, specifically their composition and microstructure.^6,7 Materials with high electrical conductivity, fast redox active sites and efficient ion mobility are promising candidates for supercapacitor electrode materials. MXene, a two-dimensional transition metal carbonitride or carbide, possesses unique properties, including outstanding electrical properties. The general formula for MXene is M_n+1X_nT_x (n = 1, 2, 3, or 4), where M represents an early transition metal, X denotes carbon or nitrogen, and T_x represents functional group terminations, such as –O, –OH, and –F groups.^8,9 Due to its highly conductive layered sheets, MXene has been proven to be an excellent candidate for supercapacitor applications.^10,11 The efficient charge storage within and on these layered sheets increases the surface area for electrolyte ion interaction.¹² MXene's hydrophilic nature, along with its inherent potential to participate in redox reactions, makes it an effective material for use in electrode applications.¹³ Moreover, the tunability of surface functional groups allows for the modification of its electrochemical properties to meet specific application requirements.¹⁴ However, MXene nanosheets tend to self-restack and attach due to strong van der Waals interactions between the monolayers, limiting the accessibility of electrolyte sites and preventing complete utilization of active sites in MXene.¹⁵ To enhance the electrochemical properties of MXene, researchers have focused on preventing the self-restacking of nanosheets and modifying the electrode structure. Strategies include increasing interlayer separation, altering the surface structure, and developing 3D porous structures.¹⁶ In this regard, NP intercalation has been explored for increasing interlayer separation. For instance, gold nanoparticle (Au NP) intercalation in MXene has resulted in as specific capacitance of 278 F g⁻¹ at 5 mV s⁻¹.¹⁷ Additionally, the in situ hydrothermal intercalation of SrTiO₃ with MXene has been studied in Zn–air batteries, and La_0.5Sr_0.5CoO_3−δ combined with MXene has been explored in Li–oxygen batteries.^18,19 The electrochemical energy storage performance of Ti₃C₂T_x was enhanced by incorporating BaTiO₃ perovskite nanoparticles into the layered structure of 2D MXene in our previous work.²⁰ The challenge with functional groups lies in their stability and compatibility in the electrolyte medium. Therefore, in this work, the targeted sulphonation of the MXene surface and its electrochemical stability have been explored. The basic strategy to enhance the pseudo capacitance of a material involves introducing redox active elements, selecting suitable electrolytes for ion intercalation and deintercalation under applied potential.^21,22 Strategic functionalization of the MXene surface significantly improves redox ion interaction efficiency by increasing active sites for supercapacitor applications. Weizhai Bao et al. modified Ti₃C₂T_x MXene with sulphur during MAX phase synthesis, enhancing the performance of Na–S batteries.²³ Ruicheng Jiang et al. demonstrated efficient adsorption and desorption of hydrogen by modifying pristine MXene with sodium alanates (NaAlH₄).²⁴ The efficient removal of methylene blue by Ti₃C₂–SO₃H through the incorporation of p-aminobenzene sulfonic acid was studied by Yuan Lei et al.²⁵ Previously, multilayered MXene was delaminated by using tetrabutylammonium hydroxide (TBAOH) and dimethyl sulfoxide (DMSO) through repeated washing. In the present work, we have omitted the DMSO washing step to decorate the layered structure of MXene with sulphonate ions, thereby enhancing its redox activity. To increase the active sites and prevent the restacking of layers, perovskite nanoparticles owing to their inherent oxygen vacancies, can significantly contribute to the overall physiochemical properties.²⁶ These nanoparticles can be effectively intercalated within the layered structure of MXene to increase its active sites. Here, we have deliberately selected H₂SO₄ as the electrolyte to achieve targeted enhancement in charge storage dominated by redox mechanisms.

The present approach for preventing the restacking of layers and enhancing the electrochemical performance involves the intercalation of perovskite nanoparticles. This work addresses the restacking issue by sulphonating and drying MXene under normal oven conditions, avoiding the vacuum drying approach.²⁷ We have studied intercalation with DMSO sonication, omitting the washing step to enhance the layered structure of MXene with sulphonate ions, thereby enhancing its redox activity. DMSO has an electrochemical stability window (nearly −2.5 V to +2.0 V vs. SCE), allowing the exploration of electrochemical interaction involving redox active groups.²⁸ Perovskite is a highly promising and stable nanoparticle that can be effectively intercalated into the layered structure, increasing the active sites of MXene. The inherent oxygen vacancies in perovskites contribute significantly to the physiochemical properties of the composite. Different halide perovskites, double perovskites and oxide perovskites have been explored for energy storage applications.^29–31 The electronic and chemical properties of the perovskite crystal network can be improved by partial or complete substitution at the anion or cation sites.³² The electrochemical performance of TMS in a light electrolyte (0.1 M H₂SO₄) was enhanced by the introduction of CFO and NFO and the cationic effect on charge storage was studied. To understand the dynamics of perovskite particle intercalation within the sulphonated MXene layered structure, we explored compositions of CFO and NFO at 3, 5 and 7 weight percentage (wt%). The surface area of CFO and NFO intercalated TMS compared to pure TMS led to enhanced electrochemical performance. The enhancement in redox dominated charge storage was validated by Dunn's method calculation and Electrochemical Impedance Spectroscopy (EIS). The specific capacitance obtained for 3 wt% CFO intercalated TMS (3CTMS) and 3 wt% NFO intercalated TMS (3NTMS) was found to be 567.42 and 593.81 F g⁻¹, respectively, with stabilities of 81.75 and 80.65% after 1000 cycles. Besides, the flexible device using redox active 3CTMS was demonstrated through a charge and discharge profile, including an application where LEDs glow. The device was tested at different angles (0°, 90° and 120°), retaining 79% stability after 5000 cycles, showing great potential for use in sensors, skin-like interfaces, and self-powdered energy devices. This work presents a novel approach for sulfonating the MXene surface to enhance energy storage performance in supercapacitors. Furthermore, a simple and cost-effective strategy for stable perovskite nanoparticle intercalation is explored. The findings highlight the potential of strategically functionalizing MXene, offering promising applications in diverse fields such as sensors, water purification, energy harvesting, and storage.

2 Results and discussion

The XRD analysis of all the synthesized materials (Fig. 1a) and their respective compositions (Fig. 1b) confirmed the phases of each material. The crystal structure and Braggs shifts observed after etching can be seen in MXene from the recorded XRD patterns (Fig. 1a and b). The pristine MAX phase shows an Al peak appearing at the (104) plane with a Bragg angle of 38.85°. After etching with HF, this peak disappeared, confirming the removal of Al from the MAX layered structure and the formation of MXene sheets. The formation of the hexagonal Ti₃C₂T_x MXene (TM) layered structure with the R3̄m space group is confirmed by characteristic diffraction peaks at a 2θ angle of 8.73°, 18.18°, 27.24°, 34.25° and 60.64° corresponding to the (002), (004), (006), (101) and (110) crystal planes, respectively. The corresponding diffracted peaks matched the JCPDS card numbers for the Ti₃AlC₂ MAX phase and TM which are 96-722-1325 and 96-153-2228, respectively. Shifts in the (002) plane from 9.37° to 8.73°, the (004) plane from 19.06° to 18.17°, and the (110) plane from 60.64° to 60.14° were observed in MAX and Ti₃C₂T_x MXene due to changes in the c-lattice parameter during the exfoliation process.³³ These changes in c-lattices were due to the tensile strain. The peak broadening is ascribed to a decrease in crystallite size, providing further evidence for the formation of nanolayered structure in TM.³⁴ The multiple diffraction peaks in TMS are due to sulphur attachment on the MXene layers.

Fig. 1 XRD pattern of (a) individual materials and (b) composites, (c) FTIR spectra of composite materials, HRTEM of (d) MXene, (e) 3CTMS, and (f) 3NTMS at 5 nm magnification; SEAD patterns of (g) MXene, (h) 3CTMS nanocomposites, (i) 3NTMS nanocomposites, and (j) sulphur present in 3CTMS nanocomposites, (k) N₂ adsorption–desorption isotherms and (l) pore-volume-distribution curves of TMS and 3% CFO/NTMS nanocomposites.

The broadening of peaks at (002), (004) and (110) in sulphonated MXene (TMS) compared to pristine TM is evident. The full width at half maximum (FWHM) values calculated at {(002), (004), (110)} points for TMS and TM are {0.697, 0.6202, and 0.1774} and {0.5867, 0.5208, and 0.1618}, respectively, suggesting microstrain induced in the lattice because of sulphonation. The XRD peak reveals the formation of CFO and NFO NPs with a spinel cubic structure characterized by the space group Fd3̄m without any secondary phases, consistent with JCPDS card numbers 96-591-0064 and 96-591-0065, respectively. In the XRD pattern of all the composite samples (Fig. 1b), the peaks corresponding to TM or TMS, NFO and CFO are indicated with *, # and $ symbols, respectively. FTIR spectroscopy was conducted to confirm the surface functional groups (Fig. 1c) present in 3CTMS, 3NTMS and pristine TMS. FTIR was performed to test the extent of sulphonation at the MXene surface by scanning the test samples using infrared light in the active wavelength range of 500 to 4000 cm⁻¹. The peaks at 3436, 2927, 1629, 1025, 881 and 624 cm⁻¹ are consistent with previously reported data.³⁵ The peaks at 3436 and 1629 cm⁻¹ correspond to the hydroxyl group attached to the surface of MXene during the synthesis process.³⁵ The deformation vibration of the Ti–O bond is observed at 624 cm⁻¹, while the peak at 881 cm⁻¹ corresponds to the Ti–C bond present in MXene.³⁶ The peak at 2927 cm⁻¹ corresponds to the –CH₂ group.³⁷ It can be observed that the intensity of peak representing the –OH group decreases from TMS to 3CTMS, indicating the sulphonation of the surface by DMSO. The peak corresponding to C–O at 1025 cm⁻¹ remains nearly unaltered, suggesting no alteration in the mechanical properties of pristine TM.³⁸

Fig. 1d shows the high resolution HRTEM image of MXene at 5 nm magnification. The d-spacing is calculated to be 0.22 nm corresponding to the (110) lattice plane present in MXene. The HRTEM images of intercalated 3CTMS confirm the presence of the (110) plane of MXene with a slightly increased d spacing of 0.223 nm (Fig. 2e). Inverse Fast Fourier Transform (FFT) calculations showed the d-spacing to be 0.29 nm, corresponding to the (022) plane of CFO. Similarly, observations in different areas of 3NTMS revealed the presence of the (131) crystal plane of NFO showing 0.251 nm d-spacing and the (110) crystal plane of MXene with 0.221 nm d-spacing (Fig. 2f). This increase in d-spacing can be attributed to the intercalation of nanoparticles within the layered sheets,³⁹ confirming the successful incorporation of CFO and NFO into the sulfonated MXene structure. The SEAD pattern clearly reveals that the 3CTMS and 3NTMS particles are intercalated within the layered structure of sulphonated MXene (Fig. 1e and f). Fig. S1a† shows the unmodified pristine MXene, while successful sulphonation is evident from the sulphur attached to the surface of 3CTMS and 3NTMS (Fig. S1b and c†). The intercalation of perovskite NPs into MXene is confirmed by the presence of (353) and (131) lattice planes of NFO and CFO in the SEAD pattern (Fig. 1h and i). The EDX analysis of MXene, 3CTMS and 3NTMS confirms the successful exfoliation of MXene sheets and the intercalation of CFO and NFO in the multilayered structure of MXene (Fig. S1d–f†). The morphological analysis was performed using FESEM on Ti₃AlC₂, TM, TMS, and 3, 5, and 7 wt% CTMS and NTMS samples (Fig. S2†). The detailed intercalation of CFO/NFO NPs in TMS has been further discussed in the ESI through elemental mapping (Fig. S3†) and point EDAX (Fig. S4 and S5†).

Fig. 2 (a) CV curves of individual compositions at 5 mV s⁻¹, (b) CV curves of 3CTMS at scan rates ranging from 5 to 100 mV s⁻¹, (c) retention capability of all compositions at various scan rates, (d) GCD curve at a current density of 4 A g⁻¹ for all compositions, (e) GCD curves of 3CTMS at current densities ranging from 4 to 9 A g⁻¹, (f) retention capability of all compositions at various current densities, (g) EIS curves of individual compositions (inset-EIS of 3CTMS), (h) graphical representation of the calculated surface and diffusion-controlled capacitance (by Dunn's method) of TMS and 3C/NTMS, and (i) cyclic stability showing retention of TMS, 3CTMS and 3NTMS after 10 000 cycles.

To determine the relative oxidation states of constituent elements in the composites, X-ray Photoelectron Spectroscopy (XPS) was performed on the 3CTMS and 3NTMS composites. The sulphonation of the MXene surface is confirmed by sharp peaks of sulphur at 162.51 and 165.54 eV, corresponding to S 2p_3/2 and S 2p_1/2, respectively (Fig. 1j).⁴⁰ The extent of sulphonation at the MXene surface is corroborated by the presence of the SO₄²⁻/S₂O₃²⁻ peak. The peak position of this state at 167.36 eV aligns well with the previously reported data, confirming SO₄²⁻/S₂O₃²⁻ states.⁴¹ The areas under the S 2p_3/2, S 2p_1/2 and SO₄²⁻/S₂O₃²⁻ peaks are 77.94, 168.36 and 388.64%, respectively, indicating the dominant contribution of SO₄²⁻/S₂O₃²⁻ sulfur ions to the overall electrochemical performance. The observed % area under the peak corresponding to SO₄²⁻/S₂O₃²⁻ is larger in 3NTMS compared to 3CTMS, suggesting a higher presence of redox active sulphone ions during electrochemical interactions. The survey spectra of 3CTMS and 3NTMS, along with discussions of the individual elements, are provided in the ESI (Fig. S6†). A comparison of the XPS peak assignments and the corresponding areas of individual elements in 3CTMS and 3NTMS is presented in Table S1.†

The specific surface area and pore size distribution of the samples were characterized using Brunauer–Emmett–Teller and Barrett–Joyner–Halenda (BET–BJH) analyses. The higher nitrogen adsorption–desorption observed in 3CTMS and 3NTMS compared to TMS (Fig. 1k), indicates an increase in the surface area due to the intercalation of CFO and NFO, respectively. The specific surface areas of TMS, 3CTMS and 3NTMS are 0.721, 3.356 and 1.341 m² g⁻¹, respectively. In the lower relative pressure range from 0 to 0.8, the increased adsorbed volume for 3CTMS and 3NTMS suggests improved N₂ adsorption, indicating the effective intercalation of NFO/CFO within the multilayered, sulphonated MXene.⁴² For TMS, the desorption curve shows a flat response to the reduction in relative pressure, suggesting slower evaporation from the cavities. This can be attributed to the fact that the MXene layered interfaces cause the adsorbent trapping at the slit edges of sheets, forming cavities. However, this desorption trend disappears in 3CTMS and 3NTMS, indicating the opening of slit edges, which facilitates multilayered adsorption and desorption. The pore size distribution, illustrated in terms of pore volume to pore diameter (Fig. 1l), shows enhancement in active sites for adsorption and desorption in 3CTMS and 3NTMS in comparison to pristine TMS. This indicates the presence of more interactive surface sites for electrolyte interaction. The calculated pore volumes are 0.020, 0.032 and 0.038 cc g⁻¹ for TMS, 3CTMS and 3NTMS, respectively. The significant increase in the surface area for 3CTMS suggests the presence of more active sites for electrochemical ion interaction, leading to enhanced charge storage capacity.

The intercalation of CFO in the layered structure of TMS, as well as the sulphonate ion decoration, provides a larger surface area than TMS. The observed increase in the surface area of 3CTMS and 3NTMS confirms the successful incorporation of CFO and NFO nanoparticles into the layered structure of TMS. These nanoparticles not only contribute their own layers but also widen the layers of TMS, resulting in the enhanced surface area. The surface abundance of SO₄²⁻/S₂O₃²⁻ provides redox active sites in H₂SO₄ electrolyte, further enhancing the electrochemical performance.

The cyclic voltammetry (CV) curves of CTMS and NTMS, each with 3, 5, and 7 wt% combinations, were recorded within the active potential window of −100 to −600 mV at scan rates ranging from 5 to 100 mV s⁻¹. The CV curves of all the electrodes at 5 mV s⁻¹ (Fig. 2a), illustrate the relative voltage–current response profile, with 3CTMS showing the largest area under the curve and hence the highest specific capacitance. The CV curve corresponding to 3CTMS shows a uniform increase in the area under the curve, suggesting a consistent current to voltage response over increasing scan rates (Fig. 2b). In the present work, 3CTMS showed the highest specific capacitance and enhanced redox activity compared to TMS. The CV curves for TMS, along with 5, 7CTMS and 3, 5, 7NTMS are included in the ESI (Fig. S7a–f†), respectively. The specific capacitance for 3CTMS was calculated to be 593.81 F g⁻¹ at 5 mV s⁻¹, whereas for TMS and 3NTMS, it was 303.98 and 567.422 F g⁻¹, respectively. The equations used for calculating the specific capacitance of electrodes from the CV and Galvanic Charge Discharge (GCD) curves are provided in the experimental section. The specific capacitance observed at different scan rates shows a stable electrode and electrolyte interaction with respect to the applied voltage rate (Fig. 2c). The charge–discharge profile for all samples at 4 A g⁻¹ current density is presented (Fig. 2d), indicating longer discharge times in C/NTMS and enhanced charge storage compared to TMS. The GCD curve of 3CTMS is shown at current densities ranging from 4 to 9 A g⁻¹ (Fig. 2e). The working electrode showed incomplete discharge at 2 and 3 A g⁻¹ current densities, suggesting sluggish and insufficient charge transfer at the electrode surface. The specific capacitance values of 3CTMS, 3NTMS and TMs are calculated to be 390, 356 and 76 F g⁻¹ at 4 A g⁻¹ current density. The GCD curve of 3CTMS at current densities ranging 2 to 9 A g⁻¹ is included in the ESI (Fig. S8a†). The GCD curves of 5, 7 CTMS and 3, 5, 7 NTMS are included in the ESI (Fig. S8b–f†). The specific capacitance of all the electrodes at different scan rates and current densities is tabulated in ESI Table S2.† Stable specific capacitance performance is observed at different current densities (Fig. 2f).

Fig. 2g shows the EIS response of 3, 5, 7 CTMS and 3, 5, 7 NTMS. The inset image in Fig. 2g shows the experimental data and the fitted data of 3CTMS. The EIS curve was obtained within a frequency window of 0.01 Hz to 100 kHz with a sinusoidal signal amplitude of 5 mV. In the equivalent circuit (inset of Fig. 2g), R_s represents the equivalent circuit resistance, R_ct1 and R_ct2 stand for the charge transfer resistances, Q₁ and Q₂ are the constant phase elements (CPEs) used to understand the non-ideal behavior under ideal capacitor conditions. The CPE is given using eqn (1),

CPE = [T_CPE (jω)ⁿ]⁻¹ (1)

where ω is the angular frequency, T_CPE is the constant phase coefficient, and n is the exponent and j = (−1)^1/2. The exponent value provides information about the non-ideal charge storage mechanism, where the inductor, resistor, pseudo capacitive element and ideal capacitor are represented by the values n = −1, n = 0, n = 0.5 and n = 1, respectively.⁴³ Incomplete discharge indicates the incapability of the electrode to participate in redox activity at lower current densities due to an increase in the R_ct2 value. The EIS fitted parameters for each electrode have been provided in Table S3.† The equivalent series resistance R_s values for 3CTMS, 3NTMS and TMS are found to be 4.35, 4.59 and 4.06 Ω, indicating effective electrolyte–ion interaction. The slight increase in series resistance in 3C/NTMS compared to pristine TMS can be attributed to the additional resistance offered by CFO and NFO at the electrode–electrolyte interface. The smaller increase in R_s values in 5, 7C/NTMS is attributed to the increase in the wt% of CFO/NFO nanomaterials in the resulting compositions. The enhancement in redox dominated charge storage is confirmed by an increase in the n value corresponding to Q₁ in C/NTMS compounds compared to TMS. The n value corresponding to Q₂ of 3CTMS compared to TMS and 3NTMS suggests smaller surface dominated static charge interaction. To study the individual contributions of surface and diffusion-controlled kinetics, eqn (2) of Dunn's power method was used:

i = k₁v + k₂v^1/2 (2)

where, v is the scan rate and i is the current at a specific voltage, k₁v and k₂v^1/2 represent the electrostatic and diffusive contributions to the overall CV response.⁴⁴Fig. 2h shows the individual contributions and the enhancement in diffusion-controlled charge interaction in 3CTMS and 3NTMS. The intercalation of CFO and NFO inside the layered structure of TMS provides more active sulphonate ions that interact with the electrolyte, resulting in more redox reactions and hence diffusion dominated charge storage. This increase in the redox active sulphonated surface is in good agreement with the BET surface area analysis and the Q₁ value obtained from the EIS equivalent circuit. The calculated individual contributions of surface and diffusion-controlled behavior at 100 mV s⁻¹ and bar diagrams of TMS, 3CTMS and 3NTMS are provided in the ESI (Fig. S9a–f†). At higher scan rates, the enhancement in diffusion-controlled kinetics is observed from 17.32% in TMS to 85.15% in 3CTMS and 94.28% in 3NTMS, respectively. A smaller diffusion% value for 3CTMS compared to 3NTMS at scan rates ranging from 5 to 100 mV s⁻¹ is due to the individual surface-controlled contribution of CFO. The CV, GCD and EIS responses of CFO and NFO in 0.1 M H₂OS₄ electrolyte under the same electrochemical conditions are provided in ESI (Fig. S10†). The specific capacitance values obtained for CFO and NFO at 5 mV s⁻¹ are 10.4 and 8.2 F g⁻¹, respectively. This suggests that the enhancement observed in sulphonated MXene is solely due to the increase in sulphonated sites resulting from the intercalation of stable perovskite nanoparticles. Cyclic stability was tested for 10 000 cycles, showing excellent retention rates of 74.05%, 81.75% and 80.65% for TMS, 3CTMS and 3NTMS, respectively, suggesting the resolution of the restacking issue in 2D TMS. The morphological integrity and the retained layered structure of 3CTMS have been discussed in the ESI (Fig. S11†). A comparison of the specific capacitance and retention capability of electrodes in the present work and previously reported composites of MXene is provided in Table S5 of the ESI.† The effective interaction of H₂SO₄ electrolyte due to the enhanced availability of redox active sulphonate ions is shown in Fig. 3, with the relevant reactions represented by eqn (3) and (4):

Ti₃C₂–SO₄²⁻/S₂O₃²⁻ + H₂SO₄ + H₂O → Ti₃C₂–OH + SO₄²⁻/S₂O₃²⁻–H₂ + SO₄²⁻ (3)

Ti₃C₂–OH + SO₄²⁻ + H₂–SO₄²⁻/S₂O₃²⁻ → Ti₃C₂–SO₄²⁻/S₂O₃²⁻ + H₂SO₄ (4)

Fig. 3 Mechanism illustrating the redox active interaction of the sulphonated surface of Ti₃C₂T_x MXene in H₂SO₄ electrolyte (three electrode assembly is also depicted).

The SO₄²⁻/S₂O₃²⁻ ions present in TMS interact with the H⁺ ions in an aqueous H₂SO₄ electrolyte under cyclic applied potential, undergoing a reversible reaction.

To make the supercapacitor suitable for practical applications, a symmetric (two electrodes) and flexible supercapacitor was fabricated. The measurements of flexible devices were performed at various bending angles (0°, 90° and 120°) (Fig. 4a). The CV and GCD curves at scan rates from 5 to 100 mV s⁻¹ and current densities from 2 to 5 A g⁻¹, respectively, are shown in Fig. 4b and c. The CV curves of the device at 100 mV s⁻¹ for 0°, 90° and 120°, bending angles, demonstrate uniformity in electrode–electrolyte interaction (Fig. 4d). The EIS curve reveals significant series resistance (4.951 Ω) and charge transfer resistances (R_ct1 = 30.68 Ω and R_ct2 = 39.55 Ω) due to the gel electrolyte (Fig. 4e). The n values corresponding to Q₁ and Q₂ are 0.124 and 0.6577, indicating resistive and diffusive charge interaction at the electrode–electrolyte interface. The diffusion dominance in Q₂ confirms the redox dominated charge storage even in the polymer gel electrolyte due to the suphonation of the MXene surface. Fig. 4f shows the fabricated device, and the specific capacitance retention capabilities with respect to the scan rate and current densities showed stable electrode performance (Fig. 4g). The power and energy densities of the symmetric flexible supercapacitor were calculated using eqn (5) and (6).

where, E is the energy density, P is the power density, S_c is the specific capacitance, ΔV is the active potential window and t is the discharge time.⁴³ The energy and power densities are calculated to be 4.177 W h kg⁻¹ and 512.17 W kg⁻¹, respectively, as shown in the Ragone plot (Fig. 4h). The device showed 79% retention after 5000 cycles (Fig. 4i). The initial increase in specific capacitance is attributed to the activation of the material's surface and the gel electrolyte interface.⁴⁵ As the number of cycles increases, the active flexible carbon cloth, functioning as the current collector, becomes more effective due to enhanced wettability and ion infiltration from the electrolyte.⁴⁶ This leads to improved interaction with the electrolyte ions. However, the subsequent decline in specific capacitance is caused by the saturation of ions at the electrode surface.⁴⁷ The practical charge and discharge capability of the device was tested by illuminating LEDs with various forward voltages. After 30 s of charging, the device showed discharge times of 84, 78, 60, 45 and 41 seconds for red, yellow, green, blue and white LEDs, respectively. The forward voltage and charge–discharge time are provided in Table 1. The details of device fabrication and the assembly of the electrode, separator and sealing process are provided in Experimental section 4.7 (Fig. 5). To comprehend the band bending and variations in the density of states in MXene after sulphonation and perovskite nanoparticle intercalation, a theoretical study has been provided in the ESI.† The detailed discussion has been included in the ESI section following Fig. S12.† The band gaps for pristine Ti₃C₂ MXene, sulphonated MXene (TMS) and CoFe₂O₄ incorporated TMS were calculated to be 3.404, 3.196 and 3.667 eV, respectively. Higher-band gap materials demonstrate superior electrochemical and thermal stability, which supports reversible redox reactions over extended cycles and improves the overall charge storage capability.⁴⁸ The observed enhancement in redox-dominated charge storage in experimental findings aligns with the theoretical trend.

Fig. 4 (a)Various bending angles (0°, 90°, and 120°) of the fabricated flexible supercapacitor device, (b) CV curve at a scan rate of 5 to 100 mV s⁻¹ and (c) GCD curves at current densities ranging from 2 to 5 A g⁻¹, (d) CV curve at 100 mV s⁻¹ at various bending angles, (e) EIS spectra with the fitted equivalent circuit provided in the inset, (f) image of the fabricated flexible device, (g) retention capability at different scan rates and current densities (inset), (h) Ragone plot, and (i) cyclic stability of the device showing 79% retention of the initial value after 5000 cycles.

Table 1 Glowing time of LEDs with different forward voltages

Sl. no.	Colour and number of LEDs	Required forward voltage (V)	Charging time (s)	Discharging time (s)
1	5 red LEDs	1.8	30	84
2	5 yellow LEDs	2.1	30	78
3	5 green LEDs	2.2	30	60
4	5 blue LEDs	3.2	30	45
5	5 white LEDs	3.6	30	41

---

Fig. 5 Device fabrication process: (a) carbon cloth of dimensions 4 × 4 cm², (b) and (c) coating of the active sample using a brush. Carbon cloth (d) before and (e) after coating, (f) assembly of the electrode with a gel electrolyte and sealing it with PET tape, and (g) RC circuit assembly and lighting of LEDs with different forward voltages.

3 Conclusion

In summary, the MXene surface was sulphonated using DMSO to functionalize it with redox active sulphonate ions for fast faradaic electrolyte interaction. Further enhancement of the sulphonated surface was achieved through perovskite nanoparticle intercalation with 3, 5 and 7 wt% within the layered structure of MXene. The increase in active sites for redox reaction was confirmed by BET analysis. Specific capacitance increases due to fast redox reactions in 3CTMS and 3NTMS were confirmed by EIS and Dunn's method analysis. The practicability of sulphonated MXene (3CTMS) was tested by fabricating a flexible symmetric supercapacitor, which successfully powered 5 LEDs with various forward voltages. The energy and power densities were found to be 4.177 W h kg⁻¹ and 512.17 W kg⁻¹, respectively. This work provides a pathway for advancing scientific research on the sulphonation of 2D material sheets for various applications.

4 Experimental details

4.1 Synthesis of Ti₃C₂T_x MXene and its sulphonation

Ti₃C₂T_x MXene was synthesised using the hydrofluoric acid (HF) etching method as described in step-I (Scheme 1). HF selectively removes Al from the Ti₃AlC₂ MAX phase (99.99% purity with an average particle size 40–60 μm, NANOSHEL, India), leading to the formation of a layered MXene structure. Briefly, 30 ml of 48% concentrated HF acid (Alfa Aesar) solution was taken in a 100 ml Teflon beaker. Under steady stirring at 100 rpm, 2 g of Ti₃AlC₂ powder was added slowly to reduce heat generation at the initial reaction stage with HF. The MAX phase was treated in HF for 24 h at 400 rpm and stirred at room temperature (35 °C). After 24 h, the mixture was washed with DI water several times (5000 rpm, 15 min) until the pH of the supernatant reached 6. Then, the washed mixture was filtered using Whatman filter paper (15 cm diameter; pore size: 11 μm). The collected material was dried at 80 °C for 8 h to obtain multilayered Ti₃C₂T_x MXene. The following steps were involved the selective etching of Al from Ti₃AlC₂ through HF and the formation of Ti₃C₂T_x MXene.

Ti₃AlC₂ + 3HF → AlF₃ + 3/2H₂ + Ti₃C₂ (7)

Ti₃C₂ + 2H₂O → Ti₃C₂(OH)₂ + H₂ (8)

Ti₃C₂ + 2HF → Ti₃C₂F₂ + H₂ (9)

Scheme 1 Schematic of synthesis: the synthesis of Ti₃C₂T_x from the Ti₃AlC₂ MAX phase by the HF etching method (step-I), followed by sulphonation using DMSO (step-II). The intercalation of NiFe₂O₄ (step-III) and CaFe₂O₄ (step-IV) was carried out by sonicating in DMSO solvent and drying at 80 °C for 6 h. The complex compound in DMSO solvent is coated on nickel foam by the addition of poly-vinyl-difluoride (PVDF) and carbon black (CB).

The removal of Al and formation of AlF₃ can be seen in eqn (7). The AlF₃ formed can be easily washed away due to its hydrophilic nature.⁴⁹ The generation of hydrogen gas due to the oxidation of Ti and reduction of H is shown in eqn (8). The unreacted HF can lead to the formation of fluoride functional termination at the MXene surface as presented in eqn (9).

Sulphonation of the Ti₃C₂T_x MXene surface was carried out using DMSO solvent. 16 mg of the synthesized MXene was dispersed in 400 μl of DMSO solvent and sonicated for 2 h. This dispersion solution was then dried at 80 °C for 8 h, leading to the formation of sulphonate ions at the MXene surface. The reaction of DMSO with the hydroxyl group present on MXene leads to the formation of surface attached sulphonate ions as shown in step-II (Scheme 1). Eqn (10) shows the sulphonation of the MXene sheet surface.

Ti₃C₂–OH + 2C₂H₆SO (DMSO) → Ti₃C₂–SO₄²⁻/S₂O₃²⁻ + 4CH₃↑ (10)

This sulphonate ion termination at the MXene surface serves as a promising participant in the potential-driven redox reaction in H₂SO₄ electrolyte.

4.2 Synthesis of NiFe₂O₄

Briefly, 0.05 M nickel nitrate nonahydrate (99% purity, Sigma Aldrich) and 0.1 M ferric nitrate nonahydrate (99% purity, Sigma Aldrich) were dissolved in 50 ml of DI water in separate solutions. The two solutions were mixed for 30 minutes. A 2 M NaOH (99% purity, Sigma Aldrich) solution was prepared by dissolving 2 g if NaOH in 25 ml of DI water. The prepared 2 M NaOH solution was added dropwise to the mixed solution until the pH reached 13. This solution was vigorously stirred for 1 h and then sealed within a 100 ml Teflon lined autoclave. The hydrothermal treatment was carried out for 24 h at 120 °C. The product was then washed several times with deionized water and calcined at 600 °C for 4 h. The reactions involved are shown as follows.

Ni(NO₃)₂·9H₂O + 2NaOH → Ni(OH)₂↓ + 2NaNO₃ + 9H₂O (11)

Fe(NO₃)₃·9H₂O + 3NaOH → Fe(OH)₃↓ + 3NaNO₃ + 9H₂O (12)

Ni(OH)₂ + 2Fe(OH)₃ → NiFe₂O₄ + 4H₂O (13)

4.3 Synthesis of CoFe₂O₄

0.05 M cobalt nitrate nonahydrate (99% purity, Sigma Aldrich) and 0.1 M ferric nitrate nonahydrate (99% purity, Sigma Aldrich) were prepared by dissolving the respective compounds in 50 ml of DI water. These two solutions were then combined and stirred for 30 minutes. Separately, a 2 M NaOH (99% purity, Sigma Aldrich) solution was prepared by dissolving 2 g of NaOH in 25 ml of DI water. This 2 M NaOH solution was added dropwise to the mixed solution until the pH reached 13. The resulting mixture was vigorously stirred for 1 h before being transferred to a 100 ml Teflon lined autoclave for hydrothermal treatment at 120 °C for 24 h. After treatment, the product was washed several times with deionized water and calcined at 400 °C for 4 h. The final powder was obtained by grinding the calcined material into a fine powder. The reactions involved in this process are shown as follows.

Co(NO₃)₂·9H₂O + 2NaOH → Co(OH)₂↓ + 2NaNO₃ + 9H₂O (14)

Fe(NO₃)₃·9H₂O + 3NaOH → Fe(OH)₃↓ + 3NaNO₃ + 9H₂O (15)

Co(OH)₂ + 2Fe(OH)₃ → CoFe₂O₄ + 4H₂O (16)

4.4 Synthesis of the Ni/CoFe₂O₄ composite with sulphonated MXene

As discussed in the sulphonation of MXene sheets, the same dispersion solution of MXene was used in DMSO. To prepare the individual compositions of NiFe₂O₄ (NFO) and CoFe₂O₄ (CFO) with sulphonated MXene, 3, 5 and 7 wt% of the synthesized materials are added to the dispersion solution. The wt% of CFO/NFO with TMS was tested at a lower percentage to study the enhancement in electrochemical interaction of TMS without additional physiochemical changes in MXene caused by CFO/NFO. The mixture was sonicated for 30 minutes followed by drying at 80 °C for 8 h as shown in step-III and IV (Scheme 1). The collected powder was then taken for analysis. For electrode fabrication, the dispersion solution of CFO/NFO with TMS is coated on nickel foam with a binder followed by drying.

4.5 Material characterization techniques

X-ray diffraction (XRD) patterns were obtained on a Panalytical, X-pert powder diffractometer using Cu Kα radiation (λ = 0.15406 nm). The presence and alteration of functional groups were confirmed using Fourier transform infrared spectroscopy on an Alpha-II, Bruker, by making crack free KBr-active sample pellets. The elemental composition was analyzed by High Resolution Transmission Electron Microscopy (HRTEM, JEOL JEM-2100 LaB₆). Compositional and oxidation state analyses were conducted using X-ray photoelectron spectroscopy (XPS) (Nexsa Base, ThermoFisher Scientific spectrometer). The surface area and pore size distribution were analyzed using the Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) (Quantachrome Instruments, version 5.23) methods. The morphological and structural details were investigated by Field Emission Scanning Electron Microscopy (FESEM; JEOL JSM IT-800). The electrochemical performances of the materials were analyzed in a three-electrode set up using an Origalys 500, Origa master. The charging of the fabricated symmetric flexible supercapacitor device was carried out using a DC power supply unit (PSD7303B, Scientific, India).

4.6 Electrode preparation and experimental conditions

A 3D porous nickel foam (99.99% purity, 110 PPI Cell Size, 1.5 mm thickness, 95–98% porosity) was used as a current collector (working electrode). Prior to coating the nickel foam was cut into 1 × 3 cm² dimensions and cleaned with 3 M HCl, ethanol and DI water separately by sonicating for 30 min each, followed by drying. As mentioned above, a prepared slurry of MXene and CFO/NFO in DMSO was taken and conductive carbon black and polyvinylidene difluoride (PVDF) were added in an 8 : 1 : 1 weight ratio. This slurry was further mixed and coated on the cleaned Ni foam with 1 × 1 cm² dimensions. The coated Ni foam was then dried for 8 h at 80 °C. The coated Ni foam was then pressed at 10 MPa pressure to ensure uniform adhesion of the active material on the Ni Foam. In all the coated Ni foam samples the mass loading was adjusted to ∼3 mg.

The electrochemical performances of all the fabricated electrodes were evaluated in 0.1 M H₂SO₄ electrolyte. A light concentration of the electrolyte was chosen as it provides higher ion mobility and reduced series equivalent resistance during the voltammetry study. Using lighter electrolytes offers cost-effectiveness and safety in terms of lower flammability with very little toxicity in consumer electronics. A saturated Ag/AgCl electrode was used as the reference electrode and platinum wire was used as the counter electrode in the three-electrode setup for measurements. A saturated Ag/AgCl solution is used to avoid any additional diffusion potential. Platinum is used as a counter electrode because of its durability, resistance to corrosion and internees towards electrochemical reactions.⁵⁰

The specific capacitance of electrodes from the CV curves are calculated using eqn (17).

where S_c, ∫idV, ∇V, m and v stand for the specific capacitance, area under the CV curve (current to voltage response), potential window, active mass loading on the electrode and scan rate, respectively.

The specific capacitance of electrodes from the GCD curves are calculated using eqn (18).

where S_c, i, ∇t, ∇V, m and v stand for the specific capacitance, current, potential window, and active mass loading on the electrode, respectively.

4.7 Device fabrication

For the electrochemical two electrode device study, conducting carbon fiber cloth (0.3 ± 0.1 thickness, 300–400 m² g⁻¹ (BET) surface area) was used as the current collector at both anodic and cathodic ends. The carbon fiber cloth was cut into 4 × 4 cm² pieces. For the electrode cleaning process, carbon fiber cloth was submerged in 1 M HCl solution and sonicated for 30 min to remove any residual metal or oxide component. Following this, the carbon fiber cloth was sonicated in DI water for 30 min to remove remaining contaminants. Then this carbon fiber cloth was dried at 60 °C for 6 h prior to coating. The detailed coating and device fabrication processes are demonstrated in Fig. 5. As mentioned above, the slurry of the active sample, PVDF and CB in DMSO solvent was coated onto the fiber cloth (Fig. 5b and c). The coated electrodes were dried at 80 °C for 6 h (Fig. 5d and e). For the gel electrolyte, 1 g of polyvinyl alcohol (PVA) was added to 10 ml of DI water and heated at 90 °C with constant stirring until complete dissolution of PVA. To the above solution, 1 g of H₂SO₄ was added and stirred continuously to prepare a homogeneous mixture and the solution was allowed to cool down to room temperature. To ensure unform electrolyte contact, the coated electrodes and separator Whatman filter paper (15 cm diameter; pore size: 11 μm, dimension; 5 × 5 cm²) were soaked in the gel electrolyte for 5 min. The device was cascaded using the coated electrodes as the cathode and anode with filter paper serving as the separator, resulting in a symmetric supercapacitor assembly (Fig. 5f). To prevent leakage of electrolyte and maintain device integrity the assembly was sealed with a PET adhesive sheet. To test the practical charging and discharging performance, the device was connected to an RC circuit and a DC power supply unit. The discharging behavior was analyzed by connecting LEDs of various forward voltages to the load (Fig. 5g).

4.8 Theoretical analysis

Theoretical band bending and analysis of the band structure have been carried out utilizing quantum ESPRESSO and NanoDecal.

Data availability

Data will be made available upon reasonable request.

Author contributions

Jitesh Pani: conceptualization, investigation, writing original draft, formal analysis. Priyanka Chaudhary: validation, review & editing. Hitesh Borkar: supervision, writing – review & editing, visualization. Meng-Fang Lin: resources, review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Jitesh Pani acknowledges the Department of Science and Technology, Government of India, for providing research grants under the INSPIRE fellowship (IF200298). HB acknowledges the financial support from the Department of Science and Technology, New Delhi ((ANRF) SERB-DST, file no EEQ/2022/001055). MFL, would like to thank the National Science and Technology Council (NSTC 111-2221-E-131-019-MY3) for the financial support. JP, and HB would like to thank the Director, NIT Warangal, for his constant encouragement. JP and HB would like to thank Dr Narendar Vadthiya and Venkata Ramakrishna Kotha from the Electronics and Communication Engineering department, NIT Warangal, for providing resources and assistance to carry out the theoretical analysis.

References

1. Z. Li, W. Zhang, H. Wang and B. Yang, Electrochim. Acta, 2017, 258, 561–570.
2. M. Yousaf, H. T. H. Shi, Y. Wang, Y. Chen, Z. Ma, A. Cao, H. E. Naguib and R. P. S. Han, Adv. Energy Mater., 2016, 6, 1600490.
3. Y. Yue, N. Liu, Y. Ma, S. Wang, W. Liu, C. Luo, H. Zhang, F. Cheng, J. Rao, X. Hu, J. Su and Y. Gao, ACS Nano, 2018, 12, 4224–4232.
4. B. Xu, F. Wu, R. Chen, G. Cao, S. Chen and Y. Yang, J. Power Sources, 2010, 195, 2118–2124.
5. P. E. Lokhande, U. S. Chavan and A. Pandey, Materials and Fabrication Methods for Electrochemical Supercapacitors: Overview, Springer Singapore, 2020, vol. 3.
6. P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845–854.
7. M. F. El-Kady, Y. Shao and R. B. Kaner, Nat. Rev. Mater., 2016, 1, 16033.
8. M. Alhabeb, K. Maleski, B. Anasori, P. Lelyukh, L. Clark, S. Sin and Y. Gogotsi, Chem. Mater., 2017, 29, 7633–7644.
9. K. S. Kumar, N. Choudhary, Y. Jung and J. Thomas, ACS Energy Lett., 2018, 3, 482–495.
10. N. K. Chaudhari, H. Jin, B. Kim, D. San Baek, S. H. Joo and K. Lee, J. Mater. Chem. A, 2017, 5, 24564–24579.
11. S. Xu, Z. Li, G. Wei, Y. Wang and Y. Yang, J. Mater. Chem. A, 2022, 10, 18812–18821.
12. Y. Cai, X. Chen, Y. Xu, Y. Zhang, H. Liu, H. Zhang and J. Tang, Carbon Energy, 2024, 6, e501.
13. S. Jayakumar, P. C. Santhosh, S. Ramakrishna and A. V. Radhamani, J. Energy Storage, 2024, 97, 112741.
14. R. A. Murugesan and K. C. Nagamuthu Raja, Mater. Res. Bull., 2023, 163, 112217.
15. B. Anasori, M. R. Lukatskaya and Y. Gogotsi, Nat. Rev. Mater., 2017, 2, 16098.
16. B. Xu, S. Yue, Z. Sui, X. Zhang, S. Hou, G. Cao and Y. Yang, Energy Environ. Sci., 2011, 4, 2826–2830.
17. Z. Zheng, W. Wu, T. Yang, E. Wang, Z. Du, X. Hou, T. Liang and H. Wang, J. Adv. Ceram., 2021, 10, 1061–1071.
18. Z. Sun, M. Yuan, L. Lin, H. Yang, C. Nan, G. Sun, H. Li and X. Yang, ACS Appl. Energy Mater., 2019, 2, 4144–4150.
19. X. Hui, P. Zhang, Z. Wang, D. Zhao, Z. Li, Z. Zhang, C. Wang and L. Yin, ACS Appl. Energy Mater., 2022, 5, 6100–6109.
20. J. Pani and H. Borkar, J. Electroanal. Chem., 2024, 956, 118092.
21. Y. P. Zhu, C. Xia, Y. Lei, N. Singh, U. Schwingenschlögl and H. N. Alshareef, Nano Energy, 2019, 56, 357–364.
22. T. Zhai, S. Sun, X. Liu, C. Liang, G. Wang and H. Xia, Adv. Mater., 2018, 30, 1–9.
23. W. Bao, C. E. Shuck, W. Zhang, X. Guo, Y. Gogotsi and G. Wang, ACS Nano, 2019, 13, 11500–11509.
24. R. Jiang, X. Xiao, J. Zheng, M. Chen and L. Chen, Mater. Chem. Phys., 2020, 242, 122529.
25. Y. Lei, Y. Cui, Q. Huang, J. Dou, D. Gan, F. Deng, M. Liu, X. Li, X. Zhang and Y. Wei, Ceram. Int., 2019, 45, 17653–17661.
26. B. S. Reghunath, K. R. Sunaja Devi, S. Rajasekaran, B. Saravanakumar, J. J. William and D. Pinheiro, J. Energy Storage, 2024, 84, 110775.
27. J. Wang, J. He, D. Kan, K. Chen, M. Song and W. Huo, Crystals, 2022, 12, 1034.
28. W. Kao-ian, M. T. Nguyen, T. Yonezawa, R. Pornprasertsuk, J. Qin, S. Siwamogsatham and S. Kheawhom, Mater. Today Energy, 2021, 21, 100738.
29. V. V. Deshmukh, C. R. Ravikumar, M. R. A. Kumar, S. Ghotekar, A. N. Kumar, A. A. Jahagirdar and H. C. A. Murthy, Environ. Chem. Ecotoxicol., 2021, 3, 241–248.
30. M. Ahangari, J. Mostafaei, A. Sayyah, E. Mahmoudi, E. Asghari, A. Coruh, N. Delibas and A. Niaei, J. Energy Storage, 2023, 63, 107034.
31. M. Riaz, T. Munawar, M. S. Nadeem, F. Mukhtar, S. D. Ali, S. Manzoor, M. N. Ashiq and F. Iqbal, J. Appl. Electrochem., 2023, 53, 673–687.
32. H. shan Nan, X. ying Hu and H. wei Tian, Mater. Sci. Semicond. Process., 2019, 94, 35–50.
33. J. Zhang, K. A. S. Usman, M. A. N. Judicpa, D. Hegh, P. A. Lynch and J. M. Razal, Small Methods, 2023, 7, 1–16.
34. M. Mahmood, A. Rasheed, I. Ayman, T. Rasheed, S. Munir, S. Ajmal, P. O. Agboola, M. F. Warsi and M. Shahid, Energy and Fuels, 2021, 35, 3469–3478.
35. D. Maru, J. Pani, H. Borkar and V. S. Palaparthy, ACS Appl. Electron. Mater., 2024, 6, 330–339.
36. Y. Li, X. Zhou, J. Wang, Q. Deng, M. Li, S. Du, Y. H. Han, J. Lee and Q. Huang, RSC Adv., 2017, 7, 24698–24708.
37. R. Shafique, M. Rani, K. Batool, A. A. Alothman, M. S. Saleh Mushab, A. A. Shah, A. Kanwal, S. Ali and M. Arshad, J. Mater. Res. Technol., 2023, 24, 2668–2677.
38. Z. Han, Y. Niu, X. Shi, D. Pan, H. Liu, E. R. Elsharkawy, M. A. Amin, C. Liu and Z. Guo, Nano-Micro Lett., 2024, 1–17.
39. M. Li, Q. Fan, L. Gao, K. Liang and Q. Huang, Adv. Mater., 2024, 36, 2312918.
40. R. Wang, Y. Wang, C. Wu, T. Zhai, J. Yang, B. Sun, S. Duhm and N. Koch, Adv. Funct. Mater., 2020, 30, 1903440.
41. X. Niu, Y. Chen and H. Hu, Membranes, 2022, 12, 4–13.
42. H. Yanagihara, K. Yamashita, A. Endo and H. Daiguji, J. Phys. Chem. C, 2013, 117, 21795–21802.
43. J. Pani, D. Maru, P. Chaudhary, J. Gangwar, K. U. Kumar, B. C. Yadav and H. Borkar, Energy Technol., 2023, 2300193, 1–11.
44. R. K. Muddelwar, J. Pani, A. B. Lad, H. Borkar and V. M. Gaikwad, J. Solid State Chem., 2024, 329, 124455.
45. G. Guo, Q. Su, W. Zhou, M. Wei and Y. Wang, RSC Adv., 2023, 13, 3643–3651.
46. S. Sharma and P. Chand, Results Chem., 2023, 5, 100885.
47. N. Kumar, S. Bin Kim, S. Y. Lee and S. J. Park, Nanomaterials, 2022, 12, 3708.
48. M. R. Islam, N. Farzana, M. R. Akhond, M. Rahaman, M. J. Islam and I. M. Syed, Mater. Adv., 2024, 5, 2491–2509.
49. C. B. Cockreham, X. Zhang, H. Li, E. Hammond-Pereira, J. Sun, S. R. Saunders, Y. Wang, H. Xu and D. Wu, ACS Appl. Energy Mater., 2019, 2, 8145–8152.
50. G. Jerkiewicz, ACS Catal., 2022, 12, 2661–2670.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta05816j

---