Bimetallic (Co/Ni, Ce) MOF decorated V₂CT_x MXene/CNT for high energy flexible zinc-ion capacitor

Abstract

This study reports the design and fabrication of a high-performance flexible all-solid-state zinc-ion capacitor based on a novel ternary composite, comprising V₂CT_x MXene (T_x = F, O, Cl, and OH), functionalized carbon nanotubes, and cobalt/nickel–cerium bimetallic metal–organic frameworks. The optimal binary composite ratio was determined through structural, morphological, and electrochemical analyses. Integration with bimetallic (Co and Ce or Ni and Ce) metal–organic framework yielded a ternary composite, which exhibited outstanding electrochemical performance in ZnSO₄/KCl electrolyte, achieving a specific capacitance of 1163.2 F g⁻¹ at 2 A g⁻¹. The synergistic combination of high electrical conductivity from functionalized carbon nanotubes, multiple redox-active centers from Co/Ni, Ce, and V species, and the ion intercalation capability of MXene contributed to superior energy storage performance. The assembled flexible all-solid-state zinc-ion capacitor device delivered a remarkable specific capacitance of 667.7 F g⁻¹, an energy density of 133.5 Wh kg⁻¹, and a power density of 1799.5 W kg⁻¹, with 92% capacitance retention over 10 000 cycles and excellent mechanical stability under repeated bending. It establishes the development of promising electrode materials for next-generation flexible and wearable energy storage devices. This study first investigates such a unique combination of V₂CT_x/functionalized carbon nanotube supported bimetallic metal organic framework for a flexible zinc-ion capacitor.

---

1. Introduction

Global energy demands necessitate the development of efficient and sustainable electrochemical storage systems.^1–6 Lithium-ion batteries (LIBs) dominate due to their high energy density, long cycle life, and stability, yet safety concerns and limited lithium resources hinder large-scale use.⁷ Zinc-ion capacitors (ZICs) have emerged as promising alternatives, combining the high energy density of batteries with the high power density and durability of supercapacitors. Aqueous Zn²⁺ electrolytes offer intrinsic safety, high ionic conductivity, and low cost, but achieving high energy density requires electrodes that couple rapid surface reactions with bulk ion storage while maintaining structural and electrical integrity during extended cycling.^8–10 Metallic zinc anodes, with a theoretical capacity of 820 mAh g⁻¹ and cost competitiveness, are widely used, but suffer from low utilization rates, dendrite growth, and parasitic reactions, which reduce energy density, efficiency, and stability, and risk short-circuiting.^11,12 Despite advances in anode design and electrolyte engineering, developing dendrite-free zinc anodes with high reversibility remains a challenge.¹³

Two-dimensional (2D) MXenes are attractive for electrochemical energy storage due to their layered structure, rich surface redox chemistry, metallic conductivity, and hydrophilicity.^14–16 Vanadium-based MXenes, especially V₂CT_x, offer high mass-specific capacitance from vanadium's low atomic mass, larger surface area, and shorter ion diffusion paths than Ti₃C₂T_x, and multiple oxidation states (+2, +3, +4), enabling strong pseudocapacitive activity.^17,18 Low-valence vanadium compounds also exhibit fast redox kinetics, allowing V₂CT_x cathodes in ZICs to undergo spontaneous oxygen redox for self-charging.¹⁹ However, the natural self-restacking of 2D V₂CT_x nanosheets hinders ion transport and leads to sluggish electrochemical kinetics.^20,21 Interlayer space engineering, such as incorporating conductive carbon nanotube (CNT) spacers, effectively improves ion transport, structural stability, and EDLC contributions.²² For instance, Wu et al. introduced 1D CNTs into 2D V₂CT_x, which shortened ion diffusion pathways and enhanced conductivity, resulting in a high specific capacitance of 246.88 mF cm⁻² at 0.5 mA cm⁻² for a zinc-ion micro-supercapacitor.²³ Further, the introduction of some hydrophilic functionalities into the sidewalls of the carbon nanotubes can improve their interfacial bonding as well as dispersibility.²⁴ Moreover, functionalization can enhance their surface area and act as a more effective spacer between the layers of V₂CT_x sheets. Saikia et al. studied that graphene/–COOH functioned CNT delivered superior supercapacitor performance compared to graphene/CNT composite.²⁵

Metal–organic frameworks (MOFs), consisting of metal clusters and organic ligands, have emerged as promising candidates for energy storage due to their tunable porosity, structural diversity, and well-organized active sites.²⁶ The charge storage performance of MOFs can be further optimized by introducing multi-metal centers, which regulate local electronic structures and enhance electrochemical activity.²⁷ Transition-metal-based bimetallic MOFs, particularly those incorporating Co, Ni, and Mn, are widely investigated because their multiple redox-active centers and strong electronic interactions significantly improve electron density near the Fermi level.²⁸ In contrast, Ce-MOFs stand out for their remarkable chemical stability, attributed to the high coordination number of Ce ions, diverse ligand chemistry, and strong bonding with O- and N-donor ligands.^29,30 Despite these advantages, the synergistic combination of Co or Ni with Ce in MOFs has not yet been systematically explored for supercapacitor applications. To overcome the intrinsic limitations of pristine MOFs, integrating them with conductive matrices such as MXenes or CNTs has proven highly effective. Such hybrids form hierarchical architectures with abundant electroactive sites and efficient ion/electron transport pathways. For instance, Liu et al. reported that Ti₃C₂T_x combined with MOFs including Cu-BTC, Fe, Co-PBA, ZIF-8, and ZIF-67 exhibited enhanced stability and superior electrochemical properties.³¹ Similarly, Zhang et al. demonstrated that CNT incorporation significantly improved the capacitive performance of pristine MOFs in metal-ion capacitors.³²

In this work, Co–Ce-MOF and Ni–Ce-MOF were rationally grown on an optimized V₂CT_x/functionalized CNT (M/CNT) matrix, and their electrochemical properties were systematically investigated as positive electrodes for all-solid-state flexible zinc-ion capacitors. The composite combines: (i) the conductivity and Zn²⁺ intercalation capability of V₂CT_x, (ii) the conductive network and EDLC behavior of CNT, and (iii) the redox activity of Co and Ce centers in the MOF and vanadium from MXene. The heterojunctions generated within the bimetallic MOF and binary matrix facilitate electron transport, enhancing the electrochemical performance. The optimized electrode delivers exceptional electrochemical performance in ZnSO₄/KCl electrolyte and, when assembled into a flexible all-solid-state zinc-ion capacitor, retains high capacitance and energy density after extensive cycling and mechanical deformation. This study introduces, for the first time, a bimetallic MOF integrated with an M/CNT binary system for energy storage, presenting a viable pathway toward the rational design of advanced electrodes for flexible zinc-ion capacitors.

2. Results and discussion

2.1. Optimization of binary composites

Fig. 1 illustrates the schematic representation of the overall synthesis process for the samples. The morphology of the synthesized materials was examined using field-emission scanning electron microscopy (FESEM). As shown in Fig. 2(a and b), exfoliated 2D V₂CT_x sheets are clearly visible, exhibiting distinct layer separation due to the removal of the aluminum layer from the parent MAX phase V₂AlC. This confirms successful etching of Al and the formation of delaminated MXene layers, which provide a larger number of accessible surface-active sites, beneficial for both composite formation and electrochemical performance. The nanotubular structure of f-CNT is observed in Fig. S1(a). Functionalization of CNT helps to restrict agglomeration of nanotubes as well as beneficial for composite formation and interaction with MXene. The FESEM images of the M/CNT binary composites Fig. 2(c–h) reveal homogeneous mixing of V₂CT_x nanosheets and f-CNTs without noticeable phase separation, indicating strong interfacial interaction between the two components. For M/CNT-5 : 1 Fig. 2(c and d), a thicker coating of f-CNTs is observed on the MXene nanosheets, consistent with the higher f-CNT content used in synthesis. In contrast, M/CNT-15 : 1 Fig. 2(g and h) displays only partial coverage of the MXene surface, reflecting the lowest f-CNT content among the composites. The M/CNT-10 : 1 sample Fig. 2(e and f) exhibits a uniform f-CNT coating over the MXene layers, while the high-magnification image Fig. 2(f) shows f-CNT intercalation between MXene sheets, suggesting both a balanced composition and strong structural integration within the binary composite.

Fig. 1 Schematic illustration of the overall synthesis process of the samples.

Fig. 2 FESEM images of (a and b) V₂CT_x, (c and d) M/CNT-5 : 1, (e and f) M/CNT-10 : 1, (g and h) M/CNT-15 : 1; (i) XRD comparison of V₂AlC, V₂CT_x, M/CNT-5 : 1, M/CNT-10 : 1, and M/CNT-15 : 1, (j) zoomed version of the same XRD plot at the low 2θ region, (k) cyclic voltammetry (CV) at 50 mV s⁻¹, and (l) galvanostatic charging discharging (GCD) comparison at 2 A g⁻¹ of V₂AlC, V₂CT_x, f-CNT, M/CNT-5 : 1, M/CNT-10 : 1, and M/CNT-15 : 1.

The crystal structure comparison of V₂AlC, V₂CT_x, and three binary composites M/CNT-5 : 1, M/CNT-10 : 1, and M/CNT-15 : 1 was studied using XRD analysis, as shown in Fig. 2(i). The XRD pattern of V₂AlC shows the characteristic peaks at 2θ values of 13.3°, 35.7°, 41.2°, 55.6°, 63.9°, 75.1°, and 79.0° corresponding to (002), (100), (103), (106), (110), (109), (116) planes, respectively matching with the V₂AlC MAX phase (JCPDS No. 29-0101).^33,34 After etching, the characteristic (103) plane of aluminium got diminished for V₂CT_x and the (002) plane got broadened and left shifted from 13.3° to 8.4° indicating successful MXene synthesis. The increment of d-spacing of 3.86 Å in V₂CT_x compared to V₂AlC is attributed to the layer separation due to aluminium elimination. After combination with f-CNT, the (002) plane of MXene got wider and slightly left shifted because of the intercalation of f-CNT inside V₂CT_x layers, resulting in an increment in layer spacing compared to pristine V₂CT_x. This increment in layer spacing can be associated with abundant surface-active sites, which are beneficial for further composite formation as well as enhanced electrochemical activity. The zoomed XRD patterns at low 2θ angle shown in Fig. 2(j) depict the characteristic (002) plane arising around 7.89°, which is left shifted by 0.68 Å compared to V₂CT_x in M/CNT composites. Also, a broad peak around 2θ = 25.4° arises in three M/CNT binary composites, which is attributed to the (002) plane of graphitic sp² carbon in f-CNT.^23,35

Three-electrode electrochemical measurements were performed in a mixed electrolyte of 1 M ZnSO₄ and 0.5 M KCl within a potential window of 0.0–1.0 V to determine the optimal binary composite. Fig. 2(k and l) compares the electrochemical performance of f-CNT, V₂CT_x, and M/CNT composites with ratios of 5 : 1, 10 : 1, and 15 : 1 using cyclic voltammetry (CV) at 50 mV s⁻¹ and galvanostatic charge–discharge (GCD) at 2 A g⁻¹. For reference, the bare carbon cloth current collector (without active material) exhibited a negligible CV area Fig. 2(k), confirming its minimal contribution to the overall electrochemical activity. All electrodes were prepared with an active material loading of 0.3 mg. Among all tested electrodes, M/CNT-10 : 1 displayed the largest CV area and the longest discharge time in the GCD profiles, indicative of superior capacitive performance. CV curves of the M/CNT composites showed distinct and reversible redox peaks, most prominently for M/CNT-10 : 1 in the voltage range of 0.3–0.6 V, which correspond to reversible redox transitions of vanadium species.³⁶ The incorporation of f-CNT into V₂CT_x notably enhanced electrical conductivity, leading to improved charge storage compared with pristine MXene. At a current density of 2 A g⁻¹, the M/CNT-10 : 1 electrode achieved the highest specific capacitance of 920 F g⁻¹, outperforming f-CNT (308.8 F g⁻¹), V₂CT_x (449.9 F g⁻¹), M/CNT-5 : 1 (678 F g⁻¹), and M/CNT-15 : 1 (838 F g⁻¹). This superior performance is attributed to the optimal balance between the electrical conductivity of f-CNT and ion-intercalation capability of V₂CT_x, as supported by the FESEM observations. Based on these results, M/CNT-10 : 1 was selected as the binary composite for the subsequent development of ternary composites incorporating bimetallic MOFs.

2.1.1 Optimization of ternary composites. The ternary composites were first characterized by XRD analysis shown in Fig. 3(a). All the characteristic XRD peaks of CCMOF and NCMOF are in good agreement with single metal containing MOFs such as Co-MOF, Ce-MOF and Ni-MOF, Ce-MOF. Typically, the crystalline peaks of CCMOF at 2θ = 17.3°, 24.6°, and 34.3° are associated with Co-MOF and the peaks at 2θ = 9.97°, 13.2°, 18.88°, and 20.6° are associated with Ce-MOF.^37,38 Further, the crystalline peaks of NCMOF at 2θ = 17.3°, 22.2°, 24.66°, and 34.3° are associated with Ni-MOF and peaks at 2θ = 10.22°, 13.37°, 18.88°, and 20.7° are associated with Ce-MOF.^38,39 The ternary composites M/CNT-CCMOF and M/CNT-NCMOF consist only of the prominent XRD peaks of CCMOF, and NCMOF, respectively, with little shifting, indicating successful growth of MOF over M/CNT-10 : 1 and interaction among the individual components. The (002) plane of V₂CT_x become weak due to the presence of intense peaks of MOF in the ternary composites. The broad peak at 25.4° confirms the presence of f-CNT in the ternary composites. Therefore, XRD study confirms the successful synthesis of ternary composites without any change in the crystal structure of the individual components.

Fig. 3 (a) XRD pattern comparison of CCMOF, NCMOF, M/CNT-CCMOF, and M/CNT-NCMOF; FESEM images of (b) CCMOF, (c) NCMOF, (d and e) M/CNT-CCMOF, (f) M/CNT-NCMOF; (g) FESEM mapping area, and (h–l) elemental mapping of V, C, Co, Ce, and O, respectively of M/CNT-CCMOF.

The surface morphology of CCMOF, NCMOF, M/CNT-CCMOF, and M/CNT-NCMOF was investigated using FESEM, as shown in Fig. 3(b–f). Both CCMOF Fig. 3(b) and NCMOF Fig. 3(c) exhibit nanorod-like structures, with some nanorods assembling into flower-like aggregates, clearly visible in the high-magnification images Fig. S1(b and c). The average diameter of the nanorods is approximately 300 nm. The growth of CCMOF and NCMOF nanorods on the M/CNT-10 : 1 matrix is clearly observed in the FESEM images of M/CNT-CCMOF Fig. 3(d) and M/CNT-NCMOF Fig. 3(f), confirming the successful synthesis of the ternary composites. The high-magnification image of M/CNT-CCMOF Fig. 3(e) shows CCMOF nanorods (diameter ∼200–300 nm) intertwined with f-CNTs and anchored onto V₂CT_x nanosheets, indicating strong interfacial contact between all components. This well-integrated architecture is expected to facilitate efficient electron transport and ion diffusion during electrochemical operation. Elemental mapping further validates the composition and uniformity of the synthesized materials. CCMOF Fig. S1(d–h) shows homogeneous distribution of Co, Ce, C, and O, while NCMOF Fig. S1(i–m) exhibits uniform dispersion of Ni, Ce, C, and O. For the ternary composites, M/CNT-CCMOF Fig. 3(g–l) displays even distribution of Co and Ce over the V₂CT_x/f-CNT matrix (V, C, and O), confirming uniform MOF growth. Similarly, the elemental mapping of M/CNT-NCMOF Fig. S2(a–g) reveals a homogeneous distribution of Ni and Ce over the V₂CT_x/f-CNT network (V, C, F, and O), verifying the successful formation of M/CNT-NCMOF.

The bulk topography of M/CNT-CCMOF was examined using TEM, as shown in Fig. 4(a–c). The images reveal a uniform distribution of f-CNTs across the V₂CT_x nanosheets, with rod-like CCMOF structures grown over the M/CNT-10 : 1 matrix. The nanorods have an average diameter of ∼200 nm, consistent with the FESEM observations. In the high-magnification TEM image Fig. 4(b), hollow f-CNT nanotubes with diameters of a few nanometers are clearly visible. HRTEM analysis Fig. 4(c) reveals distinct heterojunctions between CCMOF nanorods, V₂CT_x nanosheets, and f-CNTs, indicating an interfacial contact that facilitates efficient electron transport and thereby enhances electrochemical performance. The lattice fringe spacing of 0.35 nm, measured from the HRTEM image, corresponds to the (002) plane of f-CNT. TEM-based elemental mapping Fig. 4(e–j) confirms the homogeneous distribution of all constituent elements, further evidencing the uniform integration and strong interaction among the individual components in the M/CNT-CCMOF composite.

Fig. 4 (a and b) TEM images, (c) HRTEM images with lattice fringes, (d) ADF-STEM image, (e–j) TEM elemental mapping of Co, Ce, V, C, O, and F, respectively of M/CNT-CCMOF.

XPS analysis was performed to evaluate the valence state and bonding nature of different components in M/CNT-CCMOF. Fig. 5(a) shows a full surface survey scan of M/CNT-CCMOF, which confirms the presence of C, O, V, Co, and Ce with their atomic percentages. The absence of Al in the survey scan indicates the successful etching of V₂CT_x from V₂AlC. The high-resolution XPS spectra of Ce 3d shown in Fig. 5(b) exhibit two Ce³⁺ doublets due to the spin–orbit coupling of Ce³⁺ 3d_5/2 (882.7 and 886.4 eV) and Ce³⁺ 3d_3/2 (902.0 and 905.4 eV). The spin–orbit splitting was around 19 eV and the intensity ratio of I 3d_5/2/I 3d_3/2 was fixed to 1.05.^37,40Fig. 5(c) depicts the high-resolution Co 2p spectrum with distinct characteristics of Co²⁺ and Co³⁺ in spin–orbit doublets, along with two shake-up satellite peaks (labeled as Sat.).⁴¹ The first spin doublet at binding energies of 781.6 and 796.9 eV, attributed to 2p_3/2 and 2p_1/2 degenerate states of Co³⁺ and the second spin doublet at binding energies of 785.3 and 796.9 eV, attributed to 2p_3/2 and 2p_1/2 degenerate states of Co²⁺. Moreover, the binding energy difference between 2p_3/2 and 2p_1/2 degenerate states of Co³⁺ is ∼15.3 eV, which validates the Co 2p XPS profile. The three valence states of vanadium (V²⁺, V³⁺, and V⁴⁺) in spin–orbit doublets are also depicted by the deconvoluted V 2p spectrum shown in Fig. 5(d). The first spin orbit doublet at 516.9 and 523.0 eV is attributed to 2p_3/2 and 2p_1/2 degenerate states of V²⁺, the second spin doublet at 517.5 and 524.5 eV is attributed to 2p_3/2 and 2p_1/2 degenerate states of V³⁺, and the third spin orbit doublet at 519.2 and 526.4 eV is attributed to 2p_3/2 and 2p_1/2 degenerate states of V⁴⁺.⁴² The deconvoluted C 1s spectrum shown in Fig. 5(e) depicts peaks with binding energies of 284.6, 286.6, and 288.4 eV, corresponding to C–C, C–O, and C=O bonds, respectively.³⁷ In Fig. 5(f) the O 1s peaks located at 530.5 eV correspond to C=O, C–OH, V–O bonds, 531.7 eV correspond to C–O, C–O–C, COOH, V–C–(OH) bonds, and 533.3 eV represent acid anhydride bonds, respectively.⁴² Therefore, XPS reveals the presence of multiple valence states of Co, Ce, and V as well as the successful formation of M/CNT/CCMOF.

Fig. 5 XPS analysis of M/CNT-CCMOF (a) survey scan, with atomic percentages of elements present, deconvoluted XPS spectra of (b) Ce 3d, (c) Co 2p, (d) V 2p, (e) C 1s, and (f) O 1s.

In EDLC-type charge storage, ions are primarily adsorbed at the electrode surface; therefore, a high surface area with uniform pore size comparable to the electrolyte ion dimensions is essential for efficient charge accumulation. A multipoint BET analysis was conducted to determine the surface area and pore size distribution of M/CNT-10 : 1, M/CNT-CCMOF, and M/CNT-NCMOF, as shown in Fig. S3(a and b), with the corresponding data summarized in Table S1. Among the samples, M/CNT-CCMOF exhibited the highest surface area (84.6 m² g⁻¹), surpassing M/CNT-10 : 1 (50.6 m² g⁻¹) and M/CNT-NCMOF (80.6 m² g⁻¹). This enhancement is attributed to the highly porous architecture introduced by MOF incorporation into the M/CNT-10 : 1 framework. M/CNT-CCMOF also showed the largest total pore volume (0.09 cm³ g⁻¹). The pore size distribution confirms a predominantly mesoporous structure, with pore sizes centered between 0–5 nm Fig. S3(b). Overall, the surface area and pore structure analyses demonstrate that M/CNT-CCMOF provides a more favorable electrolyte–electrode interaction area, supporting enhanced charge storage performance.

The electrochemical performance of the synthesized samples was initially evaluated using a three-electrode configuration in 1 M ZnSO₄/0.5 M KCl within a potential window of 0–1.0 V. In this electrolyte system, Zn²⁺ ions act as the primary charge carriers, while KCl likely enhances ionic conductivity without directly participating in charge storage. This phenomenon was established by comparing the CV and GCD performance of M/CNT-CCMOF in only 1 M ZnSO₄ and 1 M ZnSO₄/0.5 M KCl mixed electrolytes shown in Fig. S6(c and d). The electrode (M/CNT-CCMOF) shows quite poor performance in only ZnSO₄ electrolyte in terms of CV area and discharge time compared to ZnSO₄/KCl mixed electrolytes. Thus, the enhancement in electrochemical performance in case of ZnSO₄/KCl electrolyte was due to the presence of KCl with ZnSO₄ which justified the incorporation of KCl in the electrolyte system. Fig. 6(a) compares the CV profiles of CCMOF, NCMOF, M/CNT-10 : 1, M/CNT-CCMOF, and M/CNT-NCMOF at a scan rate of 50 mV s⁻¹. The CV curves exhibit characteristics of both pseudocapacitance and electrochemical double-layer capacitance (EDLC), indicating that charge storage occurs via a combination of surface redox reactions and electrostatic adsorption involving Zn²⁺ ions.¹ Among the synthesized samples, the largest CV area and highest current response are exhibited by M/CNT-CCMOF because of the synergistic mixing of V₂CT_x, f-CNT, and CCMOF. The incorporation of CCMOF and NCMOF in the binary M/CNT-10 : 1 composite enhances the electrochemical performance, as indicated by the CV area and current response. The charge storage behavior of the M/CNT/CCMOF composite arises from a synergistic combination of surface redox reactions, ion intercalation, and double-layer capacitance.^23,42 The Co and Ce centers within the MOF undergo reversible redox transitions (Co²⁺/Co³⁺ and Ce³⁺/Ce⁴⁺), while the vanadium in V₂CT_x contributes additional redox activity. Concurrently, Zn²⁺ ions intercalate into the MXene interlayers, enhancing pseudocapacitive charge storage. The CNT network supports electrical conductivity and provides electric double-layer capacitance through the adsorption of ions. This hybrid mechanism results in enhanced capacitive behavior with fast kinetics and high energy storage efficiency. The following reactions clearly show different charge storage reactions present for M/CNT/CCMOF in ZnSO₄/KCl electrolyte.

Fig. 6 (a) CV, and (b) GCD comparison of CCMOF, NCMOF, M/CNT-10 : 1, M/CNT-CCMOF, and M/CNT-NCMOF, (c) CV, and (d) GCD comparison of M/CNT-CCMOF at three different electrolytes, (e and f) comparison of Nyquist plot of the synthesized electrodes, inset of (f) depicts fitting circuit, (g) bar plot of percentage of capacitance contribution of the synthesizes samples, (h) contribution of EDLC and pseudocapacitance towards the total capacitance of M/CNT-CCMOF, and (i) CV scan of M/CNT-CCMOF and V₂CT_x at 50 mV s⁻¹ used as positive and negative electrode, respectively in three-electrode setup.

Redox reaction:

Co²⁺ ⇌ Co³⁺ + e⁻ (1)

Ce³⁺ ⇌ Ce⁴⁺ + e⁻ (2)

V²⁺ ⇌ V³⁺ + e⁻ or V³⁺ ⇌ V⁴⁺ + e⁻ (3)

Intercalation/deintercalation:

V₂CT_x + Zn²⁺ + 2e⁻ ⇌ Zn·V₂CT_x (4)

EDLC:

CNT surface⁻ + Zn²⁺ ⇌ CNT-Zn²⁺ (5)

CNT surface⁺ + SO₄²⁻ ⇌ CNT-SO₄²⁻ (6)

Fig. 6(b) presents the galvanostatic charge–discharge (GCD) profiles of the synthesized electrodes in ZnSO₄/KCl electrolyte at a current density of 2 A g⁻¹. Among all tested samples, M/CNT-CCMOF exhibited the highest specific capacitance of 1163.2 F g⁻¹ (calculated using eqn S1), surpassing CCMOF (609.9 F g⁻¹), NCMOF (501.4 F g⁻¹), M/CNT-10 : 1 (896.0 F g⁻¹), and M/CNT-NCMOF (1021.8 F g⁻¹). The electrochemical analyses were carried out thrice to assure reproducibility, and the reported specific capacitances have the error range within ±20 F g⁻¹. This superior performance highlights the synergistic contribution of high electrical conductivity from f-CNTs, abundant redox-active sites from Co and Ce species in CCMOF, and the ion-intercalation capability of V₂CT_x. The role of f-CNTs in improving the conductivity and structural integrity of the electrode was evaluated by comparing the electrochemical performance of MXene–MOF composites with and without f-CNTs. As shown in Fig. S6(e and f), the CV and GCD profiles of M-CCMOF, M-NCMOF, M/CNT-CCMOF, and M/CNT-NCMOF demonstrate that the incorporation of f-CNTs significantly increases the CV area, current response, and discharge time, confirming their positive contribution to electrode performance. Electrochemical impedance spectroscopy (EIS) was conducted to evaluate the resistance behavior at the electrode–electrolyte interface. Nyquist plots for the different electrodes Fig. 6(e and f) were fitted using the equivalent circuit model shown in the inset of Fig. 6(f), with the extracted parameters summarized in Table S2. The equivalent series resistance (R_e), representing the sum of the electrode–electrolyte interfacial resistance and the intrinsic ohmic resistance of the electrolyte, was lowest for M/CNT-CCMOF. This minimal R_e value indicates reduced interfacial resistance and more efficient charge transfer, consistent with its superior electrochemical performance in both CV and GCD analyses.

The electrochemical performance of the optimum electrode was further studied in two mostly used electrolytes for supercapacitors, including 1 M H₂SO₄, and 3 M KOH. Fig. 6(c and d) shows the comparison of electrochemical efficiency of M/CNT/CCMOF in three different electrolytes, and it is clearly depicted that the highest potential range, CV area, current response, and discharge time were obtained for ZnSO₄/KCl electrolyte. It is interesting to note that the pseudocapacitive redox peaks are only prominent for ZnSO₄/KCl while EDLC-like charge storage is observed for two other electrolytes. This result signifies the participation of Zn²⁺ ions in the charge storage mechanism for the ZnSO₄/KCl electrolyte, indicating the synthesized composite is a potential zinc ion capacitor (ZIC) electrode. The specific capacitance obtained for KOH electrolyte was 466.7 F g⁻¹ and for H₂SO₄ it was 781 F g⁻¹ at 2 A g⁻¹ which are much lower compared to ZnSO₄/KCl electrolyte.

The cycling stability study of M/CNT-CCMOF shown in Fig. S6(b) demonstrates an excellent retention of 98% of its initial specific capacitance after 10 000 continuous charge–discharge cycles indicating the superior intrinsic stability of the optimum electrode. Fig. S4 presents the cyclic voltammetry (CV) curves and galvanostatic charge–discharge (GCD) profiles of M/CNT-CCMOF in three different electrolytes. In all cases, the CV curves Fig. S4(a, c and e) display increasing area and current response with increasing scan rate, a typical feature of supercapacitor electrodes. Corresponding GCD curves Fig. S4(b, d and f) follow a similar trend, with higher current densities producing shorter discharge times. Notably, M/CNT-CCMOF exhibits the largest CV area and longest discharge time in ZnSO₄/KCl electrolyte, indicating superior electrode–electrolyte interaction and enhanced charge-storage capability in this medium. To further elucidate the charge storage mechanism, the capacitive behavior was analyzed using the Trasatti method (Fig. S5), which separates contributions from electric double-layer capacitance (EDLC, surface-controlled) and pseudocapacitance (diffusion-controlled). At high scan rates, the capacitance predominantly originates from the rapid adsorption/desorption of electrolyte ions on the outer surface (EDLC), as ion penetration into the bulk is limited. Conversely, at low scan rates, electrolyte ions have sufficient time to access both surface and bulk active sites, enabling diffusion-controlled redox processes characteristic of pseudocapacitance. Based on this analysis, the optimized M/CNT-CCMOF electrode demonstrates a balanced energy-storage mechanism, with 62.3% pseudocapacitive and 37.7% EDLC contributions Fig. 6(g). The CV-area-based contribution plot Fig. 6(h) further supports this finding. The EDLC component arises primarily from the conductive carbon network (f-CNT) and V₂CT_x layers, while the pseudocapacitance originates from reversible redox transitions of V, Co, and Ce species. This synergistic mechanism ensures both high capacitance and rapid charge–discharge kinetics, critical for high-performance zinc-ion capacitors.

Based on the three-electrode studies, the M/CNT-CCMOF nanocomposite was identified as the optimal electrode material, prompting further evaluation of its practical applicability by assembling an all-solid-state zinc-ion capacitor (ASSZIC) device operating within a voltage range of 0.0–1.2 V. The electroactive material slurry was prepared using the same procedure as in the three-electrode tests and coated onto carbon cloth (2 cm × 2 cm). The coated electrodes were dried at 60 °C for 12 h under vacuum before use. Digital photographs of the bare carbon cloth, electroactive-material-coated electrodes, and the assembled ASSZIC device are shown in Fig. S9(a–c). In the device configuration, M/CNT-CCMOF served as the positive electrode, while V₂CT_x was employed as the negative electrode. Fig. 6(i) compares the CV responses of V₂CT_x in the negative voltage range (0.0 to −1.0 V) and M/CNT-CCMOF in the positive range (0.0–1.0 V) at 50 mV s⁻¹ in ZnSO₄/KCl electrolyte under a three-electrode setup. The suitability of V₂CT_x as a negative electrode was further validated through comparative CV measurements Fig. S6(a) at 50 mV s⁻¹ against commercial carbon, which is typically used as a standard negative electrode in supercapacitors. V₂CT_x exhibited a significantly larger CV area within the negative potential window, confirming its superior charge-storage capability. Moreover, an effective negative electrode for ZICs must support Zn²⁺ deintercalation/intercalation during charge–discharge processes. In this context, the layered structure of V₂CT_x MXene provides more favorable ion-accessible pathways than carbon, making it a more efficient and suitable negative electrode material. For ASSZIC fabrication, the active mass loading for the positive electrode was 0.6 mg. A PVA/ZnSO₄–KCl polymer gel served as both electrolyte and separator. The gel was applied to both active electrodes and dried to complete device assembly. Fig. 7(a) presents the individual components of the fully assembled ASSZIC device.

Fig. 7 (a) Schematic illustration of different components of the assembled ASSZIC, (b) photographs of the assembled device at different bending angles, (c) CV profile, (d) GCD profile, and (e) Nyquist plot of the assembled ASSZIC under different bending states, (f) representation of probable electron transfer mechanism in the ASSZIC during charging.

Here, PVA serves as a polymer matrix forming a three-dimensional network that physically entraps the electrolyte solution, provides mechanical stability, flexibility, and prevents electrolyte leakage. Also, the abundant –OH groups present in PVA can form hydrogen bonds with water and coordinate with Zn²⁺ ions by facilitating ion transport. On the other hand, Zn²⁺ ions act as the main charge carriers for pseudocapacitive and intercalation processes at the electrodes. While KCl acts as an inert supporting electrolyte as K⁺ and Cl⁻ ions enhance ionic conductivity by reducing charge-transfer resistance. Water in the gel maintains Zn²⁺ and K⁺ mobility via proton hopping and diffusion.

The working voltage window of the assembled device was first optimized via CV at 100 mV s⁻¹ Fig. S7(a), and GCD at 5 A g⁻¹ Fig. S7(b) with increasing the potential range and fixed it up to 0 to 1.2 V. The flexibility and mechanical robustness of the assembled ASSZIC device were evaluated under various bending conditions. Fig. 7(c–e) presents the CV curves at 100 mV s⁻¹, GCD profiles at 5 A g⁻¹, and Nyquist plots, respectively, for the device bent at angles of 0°, 60°, 90°, 120°, and 180°, as illustrated in Fig. 7(b). Across all bending states, the CV curves exhibited negligible variation in current response or enclosed area, and the GCD curves maintained identical discharge times. Furthermore, the Nyquist plots overlapped almost completely, indicating that bending did not significantly affect the charge-transfer or ohmic resistance of the device. Long-term mechanical stability was further assessed through repeated bending cycles. The CV profiles and corresponding capacitances (calculated using eqn S8) after different numbers of bending cycles are shown in Fig. 8(a and b). The capacitance change was minimal, with only a 1% increase observed after 250 bending cycles, which is negligible and may be attributed to measurement variations. Fig. 8(c and d) displays the corresponding GCD curves and specific capacitances (calculated using eqn S2), showing only an 8% capacitance loss after 250 bending cycles, which are likely due to minor aging effects or environmental influences. The Nyquist plots recorded after each bending cycle Fig. 8(e) and the extracted equivalent series resistance (R_e) values Fig. 8(f) reveal only a 0.4% increase in resistance after 250 cycles, confirming minimal structural or interfacial degradation. Collectively, these results demonstrate that the ASSZIC device possesses excellent flexibility, mechanical durability, and electrochemical stability, making it well-suited for integration into flexible and wearable energy storage systems.

Fig. 8 (a) CV profiles at 100 mV s⁻¹, (b) capacitance corresponding to each CV profile, (c) GCD curves at 5 A g⁻¹, (d) specific capacitance corresponding to each GCD curve, (e) Nyquist plots, and (f) equivalent series resistance (ESR) corresponding to each Nyquist plot after continuous bending test cycles.

To analyse ion diffusion kinetics, Nyquist plot for the device Fig. S8(b) were fitted using the equivalent circuit model with the extracted parameters shown in the inset of Fig. S8(b). The equivalent series resistance (R_e), representing the sum of the electrode–electrolyte interfacial resistance and the intrinsic ohmic resistance of the electrolyte, was low for the device which indicates low interfacial resistance and efficient charge transfer, consistent with its superior electrochemical performance in both CV and GCD analyses. The capacitive and diffusion-controlled processes of the flexible ASSZIC device was analysed using the Trasatti method to distinguish between EDLC (surface-controlled), and pseudocapacitance (diffusion-controlled) contributions. Based on the Trasatti analysis Fig. S8(c and d), the ASSZIC device exhibited 4.4% EDLC and 95.6% pseudocapacitive contribution, suggesting that charge storage is largely governed by the pseudocapacitance. This dominance of pseudocapacitance behavior is due to the presence of different metal centers and Zn²⁺ ion intercalation/deintercalation. Fig. S8(e) further illustrates the respective EDLC and pseudocapacitive contributions in terms of the CV area of the device at 10 mV s⁻¹.

Fig. S7(c) shows the CV profiles of the ASSZIC device at different scan rates, which depict pseudocapacitive as well as EDLC-type charge storage in the system. A similar result was obtained from the GCD analysis at varying current densities, as shown in Fig. S7(d). Under applied potential, electrons flow through the external circuit, while Zn²⁺ and K⁺ ions migrate through the gel electrolyte to maintain charge balance. The charge storage combines EDLC (surface adsorption of K⁺ and Zn²⁺) and pseudocapacitance (faradaic redox involving V, Co, and Ce, plus Zn²⁺ intercalation into MXene). During charging, metal centers oxidize in the positive electrode, releasing electrons to the external circuit, and the negative electrode accepts electrons, enabling Zn²⁺ intercalation into MXene, while K⁺ can balance charge at the positive electrode. The specific capacitance of the device was calculated via eqn S2. The energy density and power density of the device were calculated via eqn S3 and S4, respectively. The calculated specific capacitance, energy density, and power density of the ASSZIC device are 667.7 F g⁻¹, 133.5 Wh kg⁻¹, and 1799.5 W kg⁻¹, respectively, at 1.5 A g⁻¹. Interestingly, the specific capacitance, energy density, and power density of this assembled device are much superior to those of previously reported MXene-based zinc-ion capacitor (ZIC) devices (Table 1). Furthermore, the calculated volumetric energy density of the ASSZIC device was 0.87 mWh cm⁻³ at 1.5 A g⁻¹, which appears to be good for real-time application. Hence, the fabricated device is one of the best for practical applications in the current scenario.

Table 1 Comparison of the electrochemical performance of the fabricated ASSZIC device with a few previously reported ZIC devices^a

ZIC device (+//–)	Potential window (V)	Electrolyte	Sp. C (F g^−1)	ED (Wh kg^−1)	PD (W kg^−1)	Current density	% Of Sp. C retention	No. of cycles	Ref.
a CAC-carbon cloth; COC-cotton cloth; NFC-nanofibrillated cellulose; RGO-reduced graphene oxide.
MnO2-CNTs//Ti3C2Tx	1.9	2 M ZnSO4/1 M MnSO4	115.1	98.6	77.5	1 mV s^−1	83.6	15 000	43
δ-MnO2@CAC//Ti3C2Tx@COC	1.9	Gelatin-borax/ZnSO4/MnSO4	126	90	239	1 mV s^−1	80.7	16 000	44
Ti3C2Tx/NFC//Zn	1.3	2 M ZnSO4	303.1	—	—	1 mA cm^−1	92.84	10 000	45
RGO-V2O5//RGO-Ti3C2Tx	1.6	2 M ZnSO4	175	107.2	321.6	0.5 mV s^−1	81	10 000	46
Ti3C2Tx aerogel//Zn	1.7	1 M ZnSO4	233.5	129.83	1000.8	1 A g^−1	98.3	5000	47
Ti3C2Tx/NFC//Zn	1.2	3 M ZnSO4	250.6	—	—	0.5 A g^−1	83.2	20 000	48
Nb2CTx//Zn	1.0	2 M ZnSO4	27	2.4	40	30 mA g^−1	85	3000	49
M/CNT-CCMOF//V 2 CT x	1.2	ZnSO4/KCl	667.7	133.5	1799.5	1.5 A g^−1	92	10 000	This work

---

Fig. 7(f) presents a schematic illustration of the probable electron transfer mechanism during the charging process. The flexible ASSZIC device demonstrated outstanding cycling stability, retaining 92% of its initial specific capacitance after 10 000 continuous charge–discharge cycles Fig. 9(a). The inset of Fig. 9(a) shows CV profiles recorded at 100 mV s⁻¹ before and after the cycling test. The nearly identical CV areas, with only minor shape changes, confirm the excellent structural and electrochemical stability of the electrode material under prolonged operation. Nyquist plots obtained before and after cycling Fig. 9(b) reveal only a 0.1 Ω increase in equivalent circuit resistance, indicating minimal degradation of interfacial and intrinsic conductivity. Notably, the increased slope in the low-frequency region after cycling suggests improved electrolyte penetration into the porous structure during repeated charge–discharge processes. The XRD analysis after the cyclic stability study shown in Fig. S8(a) depicts that there is no change in peak positions indicating stable crystal structure of the sample after cyclic study. Post-cycling FESEM analysis Fig. S9(d and e) of the M/CNT-CCMOF electrode shows that the rod-like CCMOF structures remain uniformly anchored on the M/CNT-10 : 1 matrix, with no significant morphological deterioration. This structural integrity aligns with the high capacitance retention and minimal resistance increase, confirming the robustness of the electrode design for long-term operation.

Fig. 9 (a) Cyclic life analysis of the ASSZIC, inset shows the CV profile before and after the cyclic life study at 100 mV s⁻¹, (b) Nyquist plots of the device before and after the cyclic life study, (c) Ragone plot; photograph of (d–f) flexibility study of ASSZIC for practical application, and (g) lighting up a green LED.

Furthermore, Fig. 9(c) shows a Ragone plot, which shows that the energy density and power density of the assembled ASSZIC device are much greater than that of previously reported MXene-based ZIC devices (+//–) including MnO₂-CNTs//Ti₃C₂T_x (98.6 Wh kg⁻¹ and 77.5 W kg⁻¹),⁴³δ-MnO₂@carbon cloth//Ti₃C₂T_x@cotton cloth (90 Wh kg⁻¹ and 239 W kg⁻¹),⁴⁴ reduced graphene oxide (RGO)-V₂O₅//RGO-Ti₃C₂T_x (107.2 Wh kg⁻¹ and 321.6 W kg⁻¹),⁴⁶ Ti₃C₂T_x aerogel//Zn (129.83 Wh kg⁻¹ and 1000.8 W kg⁻¹),⁴⁷ and Nb₂CT_x//Zn (2.4 Wh kg⁻¹ and 40 W kg⁻¹).⁴⁹ Interestingly, M/CNT-CCMOF exhibited the highest specific capacitance, and energy density in two electrode study among the above-mentioned electrodes indicating a benchmark ZIC performance. This superior performance highlights the synergistic combination of high electrical conductivity, abundant redox-active sites, and efficient ion transport pathways in the M/CNT-CCMOF configuration, making it a competitive candidate for next-generation high-energy, high-power flexible energy storage devices. The real-time applicability of the optimized electrode was further demonstrated through LED illumination tests. Fig. S9(f) shows the charging configuration for three flexible ASSZIC devices connected in series. As depicted in Fig. 9(g), the charged devices were capable of powering a green LED (operating voltage 3.2 V). Notably, bending the devices into different configurations did not affect the LED's brightness Fig. 9(d–f), confirming that mechanical deformation had no adverse effect on their operational performance.

These results highlight the excellent flexibility, mechanical robustness, and stability of the assembled ASSZIC devices in practical scenarios, underscoring their potential for integration into portable and wearable electronic systems. Overall, the M/CNT-CCMOF electrode demonstrated outstanding performance for flexible ZIC applications, delivering high specific capacitance, energy density, and power density. This superior performance is attributed to the synergistic effect of the highly conductive M/CNT-10 : 1 matrix and the ordered growth of CCMOF nanorods, which collectively facilitate rapid electron transfer and efficient ion transport within the electrode structure. Comprehensively, the excellent electrochemical efficiency of the assembled flexible ASSZIC with high specific capacitance, energy density, and power density can be attributed to the following factors:

(i) Presence of conducting f-CNT matrix helps in facile electron transfer.

(ii) Transition metals like Co, Ce, and V incorporate pseudocapacitance to the system by acting as redox active centers (Co²⁺/Co³⁺, Ce³⁺/Ce⁴⁺, V²⁺/V³⁺ or V³⁺/V⁴⁺).

(iii) Intercalation/deintercalation of Zn²⁺ inside the V₂CT_x layers (negative electrode) induces a faradaic and intercalation reaction of the electrolyte and KCl incorporates additional ionic conductivity of the electrolyte.

(iv) The ordered heterostructure and conducting matrix of M/CNT-10 : 1 facilitate electrode–electrolyte interaction.

(v) The synergistic effect of individual components and the presence of different interfaces helps to enhance electrochemical performance by proving facile electron transfer pathways.

3. Conclusion

In summary, a ternary composite electrode material, M/CNT-CCMOF, was successfully synthesized via the in situ growth of Co–Ce bimetallic MOFs on an optimized V₂CT_x/f-CNT (10 : 1) binary matrix. Structural and morphological studies confirmed the uniform integration of components, while electrochemical characterization revealed a synergistic charge storage mechanism combining EDLC and pseudocapacitance from multiple redox-active species. The optimized electrode exhibited superior electrochemical performance in ZnSO₄/KCl electrolyte, significantly outperforming its binary and single-component counterparts. The flexible ASSZIC device fabricated using M/CNT-CCMOF as the positive electrode and V₂CT_x as the negative electrode demonstrated high energy and power densities, outstanding cycling stability, and robust mechanical flexibility. The exceptional performance can be attributed to the conductive f-CNT network, the ordered heterostructure facilitating rapid electron/ion transport, and the coexistence of multiple redox reactions. This work not only provides a benchmark design for high-performance zinc-ion capacitors but also opens new avenues for the development of flexible, durable, and high-energy-density energy storage devices suitable for portable and wearable electronics.

4. Experimental section

4.1. Materials

All the chemicals used in the study were analytical grade and used without any further purification. MAX phase (V₂AlC) and multi-walled carbon nanotubes (>95 wt%, CNT) were purchased from MSE Supplies, USA. Nickel(II) nitrate hexahydrate (99%) was acquired from ACROS ORGANICS, USA. Nitric acid (HNO₃), hydrochloric acid (HCl), and sulphuric acid (H₂SO₄) were obtained from Fisher Chemical, USA, and N,N-dimethylformamide (DMF) was purchased from Fisher Bioreagents, USA. Polyvinyl alcohol (PVA, 98–99%, high molecular weight), hydrofluoric acid (HF, (48-51)%), ethanol (96%, v/v), and potassium hydroxide (KOH, 99.98%), potassium chloride (KCl, 99%), triethylamine (Et₃N, 99%), Zinc(II) sulphate, and cobalt(II) nitrate hexahydrate (97.7%) were obtained from Thermo Fisher Scientific, USA. N-Methyl-2-pyrrolidone (99.5%, NMP), 1,3,5-benzenetricarboxylic acid (H₃BTC, 98%), cerium(III) nitrate hexahydrate (99.5%), and carbon black were procured from Alfa Aesar, and polyvinylidene fluoride (PVDF) were purchased from Sigma-Aldrich, USA. All the solutions were prepared using deionized (DI) water.

4.2. Characterizations

The crystal structure of the prepared samples was characterized via X-ray diffraction (XRD) analysis using a Rigaku MiniFlex 600 with Cu Kα at a wavelength of 1.5406 Å. Morphology and elemental mapping were examined using field emission scanning electron microscopy (FESEM) and energy dispersive X-ray spectroscopy (EDS)-Mapping, respectively, on a JEOL JSM-IT800 Schottky instrument. The bulk topography of the sample was studied via high-resolution transmission electron microscopy (HRTEM) on a Talos F200X instrument at 200 kV with an EDS facility. Kratos Axis Supra X-ray photoelectron spectroscopy (XPS) was used to characterize the elemental composition and bonding state of the sample. Electrochemical analysis was conducted on a BioLogic VSP-50 electrochemical workstation.

4.3. Synthesis

4.3.1 Synthesis of MXene (V₂CT_x). MXene (V₂CT_x) was synthesized from its MAX phase precursor (V₂AlC) via selective aluminium etching, following a previously reported method.⁵⁰ In a typical procedure, 1 g V₂AlC was gradually added to 20 mL of an etchant solution containing 12 mL HF and 8 mL 12 M HCl. The mixture was stirred constantly at 200 rpm for 8 days at 45 °C. After completion of the etching process, the product was repeatedly washed with deionized water and ethanol through centrifugation until the pH of the supernatant reached approximately 6. The resulting solid was then vacuum dried at 60 °C for 12 hours to obtain the final V2CTx powder, which is dark brown in color compared to the grey MAX phase.

4.3.2 Functionalization of CNT. Multi-walled carbon nanotube (500 mg) was added slowly to a mixture of 75 mL concentrated H₂SO₄ and 25 mL concentrated HNO₃. The solution mixture was heated at 100 °C for 6 h under constant stirring. After cooling down to room temperature, the solution was neutralized and centrifuged with DI water to separate the product. The obtained product was dried at 60 °C in vacuum and named as functionalized CNT (f-CNT).

4.3.3 Synthesis of V₂CT_x/f-CNT (M/CNT) binary composites. For the synthesis of M/CNT binary composites, V₂CT_x and functionalized carbon nanotube (f-CNT) were dispersed in a 1 : 1 (v/v) mixture of deionized water and ethanol, maintaining a specific weight ratio. The dispersion (total volume: 50 mL) was subjected to continuous agitation with argon (Ar) purging in an orbital shaker at 140 rpm for 6 hours to ensure homogeneous mixing. Following this, the composite material was collected by centrifugation, washed, and subsequently vacuum-dried at 60 °C for 12 hours to obtain the final M/CNT composite. The V₂CT_x and f-CNT ratios (w/w) were varied by 5 : 1, 10 : 1, and 15 : 1, and the products were designated as M/CNT-5 : 1, M/CNT-10 : 1, and M/CNT-15 : 1, respectively. The ratio of these three binary composites was further optimized via three-electrode electrochemical analysis.

4.3.4 Synthesis of V₂CT_x/f-CNT/mixed metal MOF ternary composites. The ternary composites of M/CNT combined with mixed metal (Co, Ce and Ni, Ce) MOFs were prepared at room temperature under simple stirring conditions using a common organic ligand H₃BTC as a linker to bind metal ions to form the framework structure. The multiple carboxylic acid groups present in H₃BTC help to bind more than one metal center, resulting in three dimensional structures. First, a homogeneous dispersion of M/CNT-10 : 1 was prepared in 40 mL DMF. Then, metal salt precursors (cobalt(II) nitrate hexahydrate and cerium(III) nitrate hexahydrate or nickel(II) nitrate hexahydrate and cerium(III) nitrate hexahydrate) in 1 : 1 metal ratio were added into the M/CNT-10 : 1 dispersion keeping 1 : 1 w/w ratio of metal and M/CNT-10 : 1. Then, H₃BTC (1 mmol) was added to the solution mixture under constant stirring. After that, 2 mL of Et₃N was added to maintain the pH which initiates the product formation. The resulting solution mixture was then aged for 2 days, and then the product was washed with water and ethanol via centrifugation. The obtained samples were then dried in vacuum at 60 °C for 12 h. The samples were named as M/CNT/CCMOF, and M/CNT/NCMOF for Co, Ce metal precursors and Ni, Ce metal precursors used, respectively. For comparison of characteristics and performance, mixed metal MOFs were also prepared without the M/CNT-10 : 1 matrix and named CCMOF (containing Co and Ce) and NCMOF (containing Ni and Ce).

4.3.5 Synthesis of gel electrolyte (PVA/ZnSO₄–KCl). The PVA/ZnSO₄–KCl gel electrolyte was prepared following the typical method of gel electrolyte preparation.⁵¹ Typically, 20 mL 10 wt% PVA solution was first prepared at 90 °C. Then, ZnSO₄ (0.85 g) and KCl (0.15 g) were added to the PVA solution and stirred for 30 min to dissolve. The obtained gel was used as both a separator and an electrolyte for the fabrication of flexible devices.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data will be made available upon request.

Supplementary information (SI): (1) electrochemical measurements for supercapacitors, (2) Trasatti method analysis, (3) FESEM, and elemental mapping, (4) CV and GCD profiles, (7) digital pictures of the electrodes and assembled ASSZIC, (8) charging set up of ASSZIC devices connected in series. See DOI: https://doi.org/10.1039/d5ta08281a.

Acknowledgements

This work was supported by DOE BES award number DE-SC0024611. The authors thank Mr Rabin Dahal for FESEM imaging and Prof. Debasish Kuila and Dr Sujoy Bepari for BET analysis.

References

1. P. Chai, Y. Li, Q. Guan, J. Li, L. Li, L. Bao, J. Peng and X. Li, Chem. Eng. J., 2024, 490, 151594.
2. B. R. KC and B. P. Bastakoti, Small Struct., 2025, 6, 2500237.
3. S. De, B. R. Kc, G. Pathiraja and B. P. Bastakoti, Nanoscale, 2025, 17, 16308–16321.
4. M. D. Ashie, G. Pathiraja, S. Mantripragada and B. P. Bastakoti, Sustainable Energy Fuels, 2025, 9, 2707.
5. S. M. M. Alonzo, M. D. Ashie, G. Pathiraja and B. P. Bastakoti, Chem. Eng. J., 2025, 505, 159208.
6. R. Dahal, R. Srivastava and B. Prasad Bastakoti, ChemistryOpen, 2025, 14, e202400173.
7. L. Zhu, H. Xie, W. Zheng and K. Zhang, Electrochim. Acta, 2022, 435, 141384.
8. Y. Liu and L. Wu, Nano Energy, 2023, 109, 108290.
9. H. Tang, J. Yao and Y. Zhu, Adv. Energy Mater., 2021, 11, 2003994.
10. Y. Liu, L. Song, C. Li, C. Song and X. Wu, Batteries, 2024, 10, 396.
11. Y.-Y. Zuo, K.-L. Wang, P.-C. Pei, M.-H. Wei, X.-T. Liu, Y. Xiao and P.-F. Zhang, Mater. Today Energy, 2021, 20, 100692.
12. W. Zheng, H. Xie, L. Zhu, H. Zhou and K. Zhang, J. Energy Storage, 2024, 76, 109808.
13. C. Wang, X. Zeng, P. J. Cullen and Z. Pei, J. Mater. Chem. A, 2021, 9, 19054–19082.
14. P. Liu, W. Liu and K. Liu, Carbon Energy, 2022, 4, 60–76.
15. H. Wang, W. Ye, Y. Yang, Y. Zhong and Y. Hu, Nano Energy, 2021, 85, 105942.
16. Y. Cheng, Y. Xie, S. Yan, Z. Liu, Y. Ma, Y. Yue, J. Wang, Y. Gao and L. Li, Sci. Bull., 2022, 67, 2216–2224.
17. X. Zhu, Z. Cao, W. Wang, H. Li, J. Dong, S. Gao, D. Xu, L. Li, J. Shen and M. Ye, ACS Nano, 2021, 15, 2971–2983.
18. S. Zhao, X. Luo, Y. Cheng, Z. Shi, T. Huang, S. Yang, H. Zheng, Y. Bi, J. Zhang and Q. Shi, Chem. Eng. J., 2023, 454, 140360.
19. M. Chen, Q. Liu, Z. Hu, Y. Zhang, G. Xing, Y. Tang and S. L. Chou, Adv. Energy Mater., 2020, 10, 2002244.
20. H. Zhang, Z. Wei, J. Wu, F. Cheng, Y. Ma, W. Liu, Y. Cheng, Y. Lin, N. Liu and Y. Gao, Energy Storage Mater., 2022, 50, 444–453.
21. Y. Cheng, Y. Xie, Y. Ma, M. Wang, Y. Zhang, Z. Liu, S. Yan, N. Ma, M. Liu and Y. Yue, Nano Energy, 2023, 107, 108131.
22. K. Yang, M. Luo, D. Zhang, C. Liu, Z. Li, L. Wang, W. Chen and X. Zhou, Chem. Eng. J., 2022, 427, 132002.
23. D. Wu, S. Zhao, T. Huang, W. He, X. Zhou, G. Wang, M. Guo, X. Luo, M. Cao and Y. Yue, Chem. Eng. J., 2024, 490, 151589.
24. N. Rani, M. Goyal and N. Jaggi, J. Mater. Sci.: Mater. Electron., 2024, 35, 592.
25. P. Saikia, K. Dutta, A. K. Guha, S. Dolui, P. Barman and L. J. Borthakur, Mater. Chem. Phys., 2021, 258, 123786.
26. F. Zhan, H. Wang, Q. He, W. Xu, J. Chen, X. Ren, H. Wang, S. Liu, M. Han and Y. Yamauchi, Chem. Sci., 2022, 13, 11981–12015.
27. W. Mu, A. Wei, Z. Hou, S. Zhang, X. Lei, Q. Wang and S. Luo, ACS Sustainable Chem. Eng., 2025, 13, 9251–9268.
28. A. Kazemi, E. Torabi, M. Tamtaji, F. Manteghi, S. Rohani and W. A. Goddard III, J. Power Sources, 2025, 652, 237616.
29. S. Fatemi and M. R. Ganjali, J. Energy Storage, 2022, 55, 105545.
30. J. Jacobsen, A. Ienco, R. D'Amato, F. Costantino and N. Stock, Dalton Trans., 2020, 49, 16551–16586.
31. C. Liu, Y. Bai, W. Li, F. Yang, G. Zhang and H. Pang, Angew. Chem., Int. Ed., 2022, 61, e202116282.
32. Y. Zhang, Q. Xie, R. Shao, J. Ding, J. Liu, W. Xu and Y. Wang, J. Energy Storage, 2024, 86, 111156.
33. F. F. Ridha and A. Al-Janabi, Opt. Quantum Electron., 2023, 55, 897.
34. M. Tahir, Energy Fuels, 2024, 38, 10183–10202.
35. D. Zhang, L. Shi, J. Fang, X. Li and K. Dai, Mater. Lett., 2005, 59, 4044–4047.
36. S. A. Zahra, B. Anasori, M. Z. Iqbal, F. Ravaux, M. Al Tarawneh and S. Rizwan, APL Mater., 2022, 10, 060901.
37. L. Ai, Y. Luo, W. Huang, Y. Tian and J. Jiang, Int. J. Hydrogen Energy, 2022, 47, 12893–12902.
38. S. Ma, Y. Shi, X. Xia, Q. Song and J. Yang, Inorg. Chem. Commun., 2023, 152, 110664.
39. S. A. Rasal, P. P. Dhavan, B. L. Jadhav and N. G. Shimpi, Appl. Organomet. Chem., 2020, 34, e5317.
40. D. Aggarwal, H. Kaur, T. Garg, V. Kumar, K. Tikoo, B. Singh, A. Kaushik and S. Singhal, Process Saf. Environ. Prot., 2024, 187, 1295–1310.
41. N. B. Velhal, N. C. Maile, C. Paeng, H. Lee, T. Kim, J. Kim and C. Yim, J. Energy Storage, 2024, 90, 111764.
42. T. Huang, B. Gao, S. Zhao, H. Zhang, X. Li, X. Luo, M. Cao, C. Zhang, S. Luo and Y. Yue, Nano Energy, 2023, 111, 108383.
43. S. Wang, Q. Wang, W. Zeng, M. Wang, L. Ruan and Y. Ma, Nano-Micro Lett., 2019, 11, 70.
44. J. Shi, S. Wang, Q. Wang, X. Chen, X. Du, M. Wang, Y. Zhao, C. Dong, L. Ruan and W. Zeng, J. Power Sources, 2020, 446, 227345.
45. J. Chen, H. Chen, M. Chen, W. Zhou, Q. Tian and C.-P. Wong, Chem. Eng. J., 2022, 428, 131380.
46. Y. Wang, J. Cao, J. Guo, J. Zhang, G. Liu, D. Wang, W. Si, J. Song, X. Meng and G. Wen, J. Alloys Compd., 2022, 915, 165418.
47. A. Mateen, M. Z. Ansari, I. Hussain, S. M. Eldin, M. D. Albaqami, A. A. A. Bahajjaj, M. S. Javed and K.-Q. Peng, Compos. Commun., 2023, 38, 101493.
48. Q. Wang, M. Chen, Q. Tian, X. Han and J. Chen, Chem. Eng. J., 2023, 467, 143481.
49. J. Azadmanjiri, J. Regner, J. Sturala and Z. Sofer, ACS Mater. Lett., 2024, 6, 1338–1346.
50. K. Matthews, T. Zhang, C. E. Shuck, A. VahidMohammadi and Y. Gogotsi, Chem. Mater., 2021, 34, 499–509.
51. X. Zhou, S. Huang, L. Gao, Z. Zhang, Q. Wang, Z. Hu, X. Lin, Y. Li, Z. Lin and Y. Zhang, Angew. Chem., Int. Ed., 2024, 63, e202410434.

---