Tetragonal symmetric-shaped {V^IV₄O₈} cubane-encasing vanado-phosphate derivative: prototype liquid-configured device-grade supercapacitor with enhanced performance

Abstract

Vanadate complexes, owing to their versatile redox and coordination properties, are emerging as promising candidates for next-generation electrochemical supercapacitors. Our investigations have examined how vanadium oxidation states, electronic configurations, and solvent interactions dictate charge storage efficiency. Studies with devices based on capsular diphosphonate–oxo–vanadates with linear ligands, Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–PO₃}₄ ⊂ 2H₂O]·34H₂O and Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–C₆H₄–PO₃}₄ ⊂ 2DMF]·29H₂O (Capsular Generation 1), displayed good energy and power densities. However, their performance was restricted by capsule length dependency and the strength of their interactions with electrolyte molecules. These effects confined the potential window to ∼1.75 V, yielding limited energy and power densities of 5.8 mWh cm⁻³ and 350 mW cm⁻³, respectively. To overcome these challenges, we turned to polyoxometalate design strategies and developed more open diphosphonate–oxo–vanadate architectures, capable of better regulating solvent interactions. This led to the synthesis of a dumbbell-shaped capsular diphosphonate–oxo–vanadate polyanionic complex, Na₁₆(Me₃C–NH₃)₆[H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]·44H₂O (1a), derived from the bent-shaped bis(4-phenylphosphonic acid) ether ligands. Single-crystal X-ray diffraction revealed a structure consisting of homo-valent vanadium(IV) centers arranged in cubic {V₄O₈} units, interconnected by the diphosphonate ligands. The oxidation state of the vanadium atoms were independently confirmed using EPR, XPS, magnetic susceptibility, and spin density analyses. Electrochemical devices fabricated with this new complex (Capsular Generation 2) exhibited substantial improvements in their electrochemical performance. The potential window expanded to 2.6 V, with volumetric energy and power densities reaching 11.24 mWh cm⁻³ and 977.54 mW cm⁻³, respectively; nearly double the values of Generation 1 devices. Enhanced cyclic stability was also observed, supported by powder X-ray diffraction and Raman analyses. Mechanistic insights indicated contributions from both surface capacitive and diffusion-controlled processes during charge storage. Practical demonstrations highlighted the rapid charge–discharge capability: after charging for just 10 seconds, the device powered a 21-LED “VNIT” display for 64 seconds and operated a small fan for 10 seconds. These findings establish the significant promise of polyoxovanadates for supercapacitive energy storage. They also demonstrate how subtle ligand-driven structural modifications can profoundly influence electrochemical performance, offering a clear pathway towards lightweight, flexible, and fast-recharging energy storage solutions.

---

1. Introduction

Self-assembled polyoxovanadates (POVs) represent a distinctive class of anionic vanadium-oxo clusters, wherein oxo-vanadate units combine either via corner-sharing or edge-sharing to arrange into unique molecular assemblies.^1–3 Due to the redox flexibility of vanadium through multiple oxidation states (III, IV and V) and diverse coordination geometries, including tetrahedral, square pyramidal, and octahedral, POVs have been observed to form diverse architectures based on their oxidation states: viz. fully oxidized (V^V),⁴ fully reduced (V^IV),^5–9 highly reduced (V^I,II),^6,10–13 and mixed-valent (V^III/IV or V^IV/V).^6,9,14–17 This redox flexibility imparts exceptional charge-transfer efficiency and structural adaptability, rendering POVs as promising candidates for redox-active applications.¹⁸ Most interestingly, our studies into mixed-metal vanado-molybdates have shown their superior electrochemical energy storage properties with respect to isostructural chromo-molybdates and mangano-molybdates.¹⁹ As such, vanadium-based compounds, particularly polyanionic oxo-vanadates, are emerging as promising materials for advanced energy storage applications, vis-à-vis as supercapacitors.^20–22 These devices are central to modern energy infrastructure due to their high power density, rapid charge–discharge capability and excellent cycle life. Nevertheless, their relatively low energy density remains a challenge, necessitating hybrid material systems.²³ Hybrid supercapacitors typically integrate the high surface area and conductivity of carbon-based materials (e.g., multi-walled carbon nanotubes) with the redox active transition metal oxides.^24–26 In this context, POVs have emerged as high-capacity, redox-tunable components compatible with a variety of electrolytes. When embedded within conductive carbon matrices, POV-based hybrid electrodes exhibit improved performance through synergistic effects between electron-conductive and redox-active components.²⁷

Recently, incorporation of organic ligands into POV architectures have led to the formation of complexes that show a high degree of structural and electronic versatility, and therefore tunable properties.²⁸ A notable synthetic strategy involves the functionalization of POVs with rigid, bifunctional ligands, which promote the formation of structurally discrete motifs. Foundational studies by Müller, Zubieta, and Schmitt have shown that aromatic diphosphonates based on benzyl, biphenyl or naphthyl cores can induce the formation of capsular assemblies through coordination with vanadium.²⁹ The pronounced structural variety and coordination lability of the metal centers within these organic-inorganic hybrid architectures underpin their increasing relevance in contemporary research. As such, owing to their unique acid–base behavior and structural modularity, such organic ligand functionalized-POVs have demonstrated significant potential in a wide array of applications, including heterogeneous catalysis,³⁰ bioinorganic and medicinal chemistry,^31–33 photo-redox processes,³⁴ as well as electrochemical energy conversion and storage systems.^20,21,35

Schmitt and coworkers have extensively reported on diphosphonate-functionalized oxo-vanado cage-type architectures, wherein the sizes of the cages have been aptly modified by subtle changes in ligand lengths.^36–38 These isoreticular assemblies are observed to contain the mixed-valent {(V^VO)₂(V^IV₄O₈)₂} species, constructed from square pyramidal {O=VO₄} units partially reduced to a V^IV/V state, resulting in convex oligomeric frameworks; and are typically constructed using linear organic phosphate ligands based on mono-, di- or tri-phenyl architectures connected at the ortho and para-positions. Subsequently, our studies on the energy storage properties of devices prepared with these linear-shaped capsular diphosphonate-oxovanadates (Capsular Generation 1), have shown that the size of the capsule plays a very important role in their electrochemical behavior.³⁹ This is ostensibly due to changes in ionic and solvent mobility within the cages. As such, significantly distinct electrochemical properties were observed from the isoreticular linear-shaped capsular polyanionic complexes, Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–PO₃}₄⊂2H₂O]·34H₂O and Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–C₆H₄–PO₃}₄ ⊂ 2DMF]·29H₂O, constructed from phenyl- and biphenyl-based ligands, respectively. The large variations in electrochemical properties among these isoreticular compounds were associated with the differences in guest molecules present in each capsule, which defines the size of the capsular cage. The larger biphenyl-based cages accommodate the larger DMF molecule, which more effectively modulates the vanadium electron density, due to its higher electrophilicity.

Intending to enhance the electrochemical energy storage properties of close-shaped oxo-vanadates, we endeavored to obtain newer oxo-vanadate cage assemblies by varying the shape of the diphosphonate ligand. By introducing the bent-shaped ligand bis(4-phenyl phosphonic acid) ether into oxo-vanadate assemblies, the idea was to shift from changing ligand sizes to ligand shapes. With newer oxo-vanadates functionalized with such ligands, we aimed to better understand how varying the cage architecture can influence the electrochemical properties of these complexes. Thus, using a variably-shaped bifunctional ligand, we intended to obtain more open oxo-vanadate cage-architectures, which can therefore provide greater mobility for electrolytic ions and solvent molecules to pass through the cages.

Using this approach, the successful synthesis of an unprecedented dumbbell-shaped capsular polyanion, viz., [H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]²²⁻ (1) has been accomplished, which contains homovalent cube-shaped {V^IV₄O₈} units, instead of the previously observed mixed-valent {(V^VO)₂(V^IV₄O₈)₂} units in the linear-shaped capsular polyanions. This outcome highlights the crucial influence of ligand geometry and connectivity on the self-assembly of vanadium-based clusters. The encapsulated solvent molecules (DMF and water) and solvent-accessible voids further modulate the local electronic environment. This ligand-directed assembly enabled the fine-tuning of key physicochemical properties, including cavity dimensions, solubility, and thermal and chemical stability – across a wide range of pH, temperature and electrochemical environments. Polyanion 1 crystallizes as a mixed sodium-tertiarybutyl ammonium salt, Na₁₆(Me₃C–NH₃)₆[H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]·44H₂O (1a).

The {V₄O₈} cubane cluster is a well-established polynuclear motif in vanadium oxide chemistry, known for its modularity and structural versatility, wherein all vanadium atoms are shown to exist in either +4 or +5 oxidation state(s).^40–48 Earlier reports with such motifs have included both multi-dimensional coordination polymers, as well as molecular polyanions. Examples of coordination polymers with similar oxo-vanadate units include H₂V₃O₈ with the {V₃O₈} unit,⁴⁹ vanadium-oxides with {V₄O₉} units,^50,51 (NH₄)₃(4,4′-Hbpy)[(V^IVO)₆(µ₄-O)₂(µ₃-OH)₂(µ₃-SO₃)₄(µ-SO₃)]·15/2H₂O with the {V₃O₈} unit,⁴² and (C₁₀H₁₀N₂)[(VO₂)₄(PO₄)₂] with the {V₄O₈} unit,⁴⁵ wherein the oxo-vanadate clusters act as secondary building units in multi-dimensional hybrid oxide frameworks. Furthermore, the compound (NH₄)₂(Et₄N)[(V^IVO)₆(µ₄-O)₂(µ₃-OH)₂(µ₃-SO₃)₄(H₂O)₂]Cl·H₂O is a molecular polyanion containing the vanadium(IV)-sulfites cubane,^43,44 isolated as atomically defined redox units. A related natural analogue, [P₂V₄O₁₆]⁶⁻, is found in the mineral phosphovanadylite and features edge-sharing VO₆ octahedra linked via µ₃- and µ₂-oxo bridges.⁴⁰ In contrast, polyanion 1 is therefore the first assembly of multiple {V^IV₄O₈} units in a molecular anionic complex, stabilized within a rigid inorganic–organic cage.

Subsequently, devices prepared with binder-free electrodes with compound 1a (Capsular Generation 2) have been extensively studied for electrochemical energy storage evaluations, i.e., supercapacitors. The results demonstrate significant performance enhancements with respect to Capsular Generation 1 devices across multiple parameters, such as increased potential window, areal capacitance, volumetric energy and power densities, as well as cycling stability. These findings underscore the potential of Capsular Generation 2 devices as the next-generation vanadium-based materials for energy storage. The present study also elucidates the relationship between the structural features of these oxo-vanadate cages and their electrochemical properties, particularly the roles of ligand geometry and guest inclusion in modulating redox behaviour and ion transport dynamics.

2. Experimental section

2.1 Materials and methods

All chemicals used for organic ligand synthesis, metal salts, and electrolytic materials were used as purchased. Millipore water was used for the synthesis of compound 1a. Powder X-ray diffraction (P-XRD) measurements were conducted at ambient temperature using a Bruker D8ECO diffractometer equipped with a Cu-K_α radiation source (λ = 0.154 nm), over a 2θ angular range of 5–60°. Infrared spectroscopy (IR) was performed in the mid-infrared region (4000–400 cm⁻¹) in transmittance mode on a Thermofisher Scientific iS50 spectrometer at room temperature. Field-emission scanning electron microscopy (FE-SEM) was carried out on a ZEISS SUPRA 40 FE-SEM system, coupled with an energy dispersive X-ray (EDX) analyzer for elemental composition analysis. X-ray photoelectron spectroscopy (XPS) measurements were obtained with a PHI 5000 Versa Probe II, ULVAC-PHI, Inc. instrument. Raman spectra were collected on a NOST (HEDA-URSM4/5/7) Raman spectrometer with a 512 nm excitation source over the range of 100–1800 cm⁻¹. The X-band EPR spectra were recorded with JEOL JES-FA series at 9.45 GHz in solid phase. The DC magnetic measurements [M vs. T from 3–300 K under H = 0.1 T, and M vs. H at 3 K] were carried out in a VSM-magnetometer in the temperature range of 3–300 K. 15.54 mg of the ground powder sample was wrapped with 40 mg of Teflon tape for attachment to the glass holder of the sample cavity of the machine. The raw data were subjected to diamagnetic corrections for the sample and Teflon tape. Elemental analysis for C, H, and N was performed on a Thermo Scientific Flash 2000 Organic Elemental Analyzer (CHNS/O Mode). Elemental analysis for V, P and Na was performed using a quadrupole inductively coupled plasma-mass spectrometer (ICP-MS, Thermo X Series II). BET studies were performed by Quantachrome NovaTouch LX4. Imaging studies for the thickness and morphology assessment were carried out on a Nikon Eclipse Ci-POL Polarized Optical microscope.

Electrochemical analysis was performed at room temperature with a PARSTAT 4000 potentiostat/galvanostat (Princeton Applied Research, USA) within a three-electrode electrochemical cell. The working electrode consisted of the prepared sample, while a silver/silver chloride (Ag/AgCl) electrode served as the reference electrode, and a platinum (Pt) wire was used as the counter electrode. All electrochemical measurements were conducted in a ternary solvent mixture composed of ethylene carbonate, propylene carbonate, and ethyl acetate in a 1 : 1 : 1 volume ratio, containing 0.5 M NaClO₄, referred to as the 0.5 M NaClO₄ electrolyte solution.

2.2 Single-crystal X-ray diffraction

Single crystals X-ray diffraction data of 1a were collected at 180 K on a Stoe STADIVARI diffractometer (Ga-K_α = 1.34143 Å and detector: Dectris EIGER2 R 4M). Routine Lorentz and polarization corrections were applied, and absorption correction were applied using LANA.⁵² Direct methods were used to solve the structures and to locate the heavy atoms (SHELXS97), and the remaining atoms were found from successive difference maps (SHELXL-2018).⁵³ SQUEEZE option on Platon was used to refine the solvent accessible void(s).⁵⁴ The hydrogen atoms on C and N were added in the calculated positions and refined using a riding model. Crystallographic data of 1a are summarized in Table 1. CCDC 2455377 contains the crystallographic data for this paper.

Table 1 Crystal data and structure refinement parameters of the complex 1a

a R1 = Σ||Fo| − |Fc||/Σ|Fo|. .
Complex	1a
Empirical formula	V16P16C132N10O140H270Na16
Molar mass (g mol^−1)	5916.01
Crystal system	Tetragonal
Space group	P4212
a (Å)	26.1876(4)
b (Å)	26.1876(4)
c (Å)	20.1266(4)
α (°)	90
β (°)	90
γ (°)	90
Volume (Å^3)	13 802.6(5)
Z	4
Temp (K)	273(2)
D calcd (g cm^−3)	1.423
Abs coeff (µ mm^−1)	4.137
F(000)	6072
Crystal size (mm)	0.200 × 0.183 × 0.150
Theta range for data collection (°)	2.82–64.50
R(int)	0.0340
T min/Tmax	0.492/0.576
GoF	1.015
R1 [I > 2σ(I)]	0.0661
wR2^a	0.1957

---

2.3 Synthesis

2.3.1 Bis(4-phenyl phosphonic acid) ether. Bis(4-phenyl phosphonic acid) ether was synthesized using a modified version of the Michaelis–Arbuzov reaction, as outlined in the literature.⁵⁵ A mixture of bis-(4-bromophenyl)-ether (19.68 g, 60 mmol) and 150 mL of 1,3-diisopropylbenzene were heated under nitrogen atmosphere at 120 °C for 2 h with continuous stirring. 0.75 g of anhydrous NiCl₂ and 30 mL (180 mmol) of triethyl phosphite were then added. Immediately, the temperature was increased to 180 °C and left for 48 h with continuous stirring. Subsequently, fresh 0.69 g of NiCl₂ and 30 mL of triethyl phosphite were added to the solution mixture and kept at 180 °C for the next 48 h. The resulting dark solution was cooled overnight, and then subjected to vacuum distillation to remove the solvent and excess triethyl phosphite. The obtained phosphonate ester was then dissolved in 375 mL of ethanol and heated to 80 °C. Over a 3-hour period, 150 mL of concentrated hydrochloric acid was added dropwise. At the end of the first day of hydrolysis, the appearance of a white solid was observed. Hydrolysis was stopped after 5 days as no further solid formation occurred. The reaction mixture was then cooled, and the white precipitate of bis(4-phenyl phosphonic acid) ether was isolated by filtration, washed extensively with water, and dried at 60 °C. Yield 16.91 g (86.38%). ¹H (CDCl₃) 7.82 (q, ²J_H,H = 8.8, ³J_H,P = 13.1, ortho-H), 7.13 (q, ²J_H,H = 8.8, ⁴J_H,P = 3.2, meta-H). ¹³C (CDCl₃) 160.8 (d, ⁴J_C,P = 3.5, C-4, C-4′), 134.4 (d, ²J_C,P = 11.4, C-2, C-6, C-2′, C-6′), 128.5 (d, ¹J_C,P = 190.4, C-1, C-1′), 119.9 (d, ³J_C,P = 15.9, C-3, C-5, C-3′, C-5′). ³¹P{¹H} (CDCl₃) 12.46.

2.3.2 Na₁₆(Me₃C–NH₃)₆[H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]·44H₂O (1a). A mixture of V₂O₅ (1.10 g, 6.04 mmol), t-BuNH₂ (1 mL), Et₃N (1 mL), H₂O (60 mL), and DMF (20 mL) was vigorously stirred in a 100 mL beaker at 60 °C until a grey-green solution formed after two hours. Subsequently, NaN₃ (1.00 g, 15.38 mmol), H₂O₃P–C₆H₄–O–C₆H₄–PO₃H₂ (2.00 g. 6.05 mmol) and N₂H₅OH (0.12 mL, 3.86 mmol) were added to the solution, and the volume of the then dark green solution was reduced to 50 mL under stirring at 60 °C. The solution was then filtered and cooled to RT. The resulting turquoise blue crystals (which weathered slowly out of the mother liquor) were filtered off after 15 days (yield: 0.85 g, 18.19% based on V). Elemental analysis (%) calculated for 1a: V: 13.78, P: 8.38, C: 26.80, H: 4.60, N: 2.37, Na: 6.22. Found: V: 13.64, P: 8.27, C: 25.53, H: 5.35, N: 2.76, Na: 6.38.

2.4 Preparation of the MWCNT@SS substrate

A thoroughly cleaned stainless steel (SS) substrate was employed as the deposition platform for MWCNTs. To enhance the surface reactivity and facilitate subsequent crystal growth, MWCNTs were functionalized via oxidative treatment. Specifically, 250 mg of MWCNTs were refluxed in 200 mL of hydrogen peroxide at 90 °C for 48 hours.⁵⁶ This treatment introduced oxygen-containing functional groups (e.g., –COOH, –OH) onto the nanotube surfaces, thereby improving their chemical affinity for nucleation and growth of crystalline phases. Following functionalization, the MWCNTs were thoroughly washed with double-distilled water (DDW) 4–5 times to remove residual reagents and then dried at 60 °C for 12 hours. To obtain a homogeneous and stable dispersion, the dried MWCNT powder was ultrasonicated in a solution of 0.5 mL Triton X-100 and 50 mL DDW. The resulting dispersion was then applied onto the cleaned SS substrate using a dip-and-dry coating technique.⁵⁶ The resulting MWCNT-coated SS films served as conductive platforms for the subsequent growth of crystalline materials on the interface.

2.5 Electrode preparation

Electrodes for electrochemical analysis were prepared using a binder-free approach, by the direct growth of crystalline phase of 1a on MWCNT@SS substrate [1-MWCNT@SS]. The growth of compound 1a on the MWCNT-coated stainless-steel (SS) substrate was carried out using a dip-coating method. The MWCNT@SS substrate was immersed in a vial containing the reaction mixture of compound 1a. This immersion provided nucleation sites for the crystallization of the compound, resulting in the gradual formation of a blue film on the MWCNT@SS substrate. After 15 days, the MWCNT@SS substrate was removed from the solution and allowed to dry at room temperature. The resulting 1-MWCNT@SS electrode was then used for subsequent electrochemical analysis.

2.6 Liquid state symmetric supercapacitor device fabrication

The device was assembled by sandwiching an electrolyte-soaked filter paper as a separator between two identical 1-MWCNT@SS electrodes, effectively preventing direct electrical contact while enabling ionic conduction. This configuration ensures efficient charge separation and transport within the electrochemical cell. The assemblies were then placed in a 0.5 M NaClO₄ electrolyte solution for the final stage of device characterization. The electrochemical performance was evaluated using the same workstation, employing cyclic voltammetry (CV), galvanostatic charge–discharge (GCD), and electrochemical impedance spectroscopy (EIS) techniques. The operational performance of the device was evaluated by driving an LED panel. A schematic representation of the fabricated supercapacitor device, its configuration during testing, and the demonstration of the LED illumination and operating of the fan are provided in Fig. 7.

3. Results and discussion

3.1 Synthesis and structural features

Compound 1a, Na₁₆(Me₃C–NH₃)₆[H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]·44H₂O, was synthesized in a single pot reaction of vanadium pentoxide with bis(4-phenyl phosphonic acid) ether in the presence of triethyl amine, t-butyl amine and sodium azide, in a mixed water-DMF solvent mixture. Both DMF and water molecules are observed to co-crystallize in 1, and observed to be encapsulated in the polyanion pocket(s).

Compound 1a crystallizes in the highly symmetric tetragonal crystal system with the space group of P42₁2. As such, the asymmetric unit of polyanion 1 consists of only one-fourth of the complex, made of a solitary {V^IV₄O₈} unit and two diphosphonate ligands. The oxidation state of the vanadium atoms was established to be +4 using bond-valence-sum calculations [Table S3 in SI], and confirmed using spectroscopic and magnetic studies. The {V^IV₄O₈} assembly is formed from four tetravalent vanadium(IV)-atoms, four terminal O-atoms, four triply bridging O-atoms (µ₃-O) and four phosphate anions from the organic tether. In this configuration, each vanadium atom adopts octahedral coordination, with three µ₃-O atom donors, two O-atoms from two different phosphate ions and one terminal O-atom with a V=O double bond. The four (V=O) groups and four µ₃-O atoms thus form the stable {V₄O₈} cubane moiety [Scheme 1]. In the complete polyanionic structure of 1, the four {V^IV₄O₈} units are connected alternatively with each other by two diphosphonate ligands, for a total of eight diphosphonate ligands in the polyanion. Upon connecting the asymmetric units, the overall shape of the polyanion resembles a more open dumbbell-shaped capsule [Fig. 1], the significance of which is observed to directly affect the electrochemical properties (vide infra). As found with previously reported capsular complexes,³⁶ solvent molecules are also observed to be encapsulated within this polyanionic cage; which in the case of polyanion 1 is both DMF and water. This solvent and the intracapsular voids create an interaction cage for guest species. The complete polyanion formula was therefore established to be [H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]²²⁻ [Fig. S1 in SI], with the rest of the polyanion charge being balanced by sodium and t-butyl ammonium cations. Bulk-phase purity of compound 1a was confirmed using powder X-ray diffraction (P-XRD) and elemental analysis [Fig. S2 in SI].

Scheme 1 Scheme shows the assembly of polyanion 1 from building units, {V₄O₈} and H₂O₃P–C₆H₄–O–C₆H₄–PO₃H₂, respectively.

Fig. 1 Ball-and-stick representation of the polyanion [H₁₀(V^IV₄O8)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈]²²⁻. Color code: teal, vanadium(IV); purple, phosphorous; red, oxygen; and grey, carbon. Hydrogen atoms on carbon have been omitted for clarity.

3.1.1 Packing interactions. The three-dimensional packing arrangement of polyanion 1 within compound 1a is observed to be directed by strong hydrogen-bonded interactions between the polyanionic units, counterions and co-crystallized solvent molecules [Table S4 in SI]. When viewed along the c-axis, layers of polyanion units are arranged in ABAB… fashion, with each layer in perpendicular orientation with respect to the other [Fig. S3 in SI]. This creates significant voids within each layer, occupied by counterions and solvent molecules. As such, the presence of such voids within and outside of the polyanion(s) makes movement of the electrolytic ion(s) more unconstrained.

3.2 Spectroscopic characterizations

3.2.1 Infrared (IR) spectroscopy. Infrared spectroscopy was employed to identify the functional groups associated with the ligand and metal coordination present in compound 1a [Fig. S4 in SI]. Based on literature references of vanadium-oxo complexes and phosphate groups, the strong vibrational peaks at 1589, 1500 and 1404 cm⁻¹ are assigned to the C_ar–C_ar stretching vibrations; a sharp band centered at 1260 cm⁻¹ is assigned to the C_ar–O bond stretching; a set of bands in the 1200–1000 cm⁻¹ range is assigned to P–O stretching vibrations; strong peak(s) in the 1020–960 cm⁻¹ region are attributed to V(IV)=O stretching; peaks at 810 and 776 cm⁻¹ are attributed to the O–V–O vibrations; and peaks at 741 and 441 cm⁻¹ are characteristic of the µ₃-O-bridge V–O–V connectivity asymmetric vibrational band and a symmetric vibrational band, respectively.^9,57

3.2.2 UV-visible spectroscopy. Solution-phase UV-visible spectroscopy revealed characteristic absorption bands corresponding to both ligand-to-metal charge transfer (LMCT) and d–d transitions [Fig. S5 in SI]. The intense absorption bands observed in the 230–290 nm range are attributed to pπ–dπ LMCT transitions from the terminal oxo ligands (O_t) to the vanadium centers. In the visible region, two broad, relatively low-intensity bands were observed at 600 and 800 nm, which are assigned to the ²T_2g → ²E_g transitions of the d¹ (V⁴⁺) ion in a distorted octahedral geometry.⁵⁸ The reduced symmetry of the coordination environment renders these transitions partially allowed.

3.2.3 X-ray photoelectron spectroscopy (XPS). X-ray photoelectron spectroscopy (XPS) was employed to determine the surface composition and oxidation states of elements present, especially for vanadium, in 1a [Fig. S6 in SI]. The survey spectrum reveals the predominant presence of vanadium(IV) and oxygen (O), as evidenced by distinct V 2p and O 1s peaks. The high-resolution O 1s spectrum displays a prominent peak at ∼530.3 eV, attributable to lattice oxygen (O²⁻), indicative of a chemically uniform oxide framework. The V 2p core-level spectrum shows well-resolved peaks at ∼516.5 eV (V 2p_3/2) and ∼524.2 eV (V 2p_1/2), consistent with vanadium in the +4-oxidation state. No additional peaks corresponding to V⁵⁺ or V³⁺ were detected, confirming the dominance of the V⁴⁺ species. The narrow peak widths and symmetric profiles suggest a highly ordered surface environment. These spectral features collectively confirm the presence of a stoichiometrically consistent V⁴⁺ oxide phase.

3.2.4 Electron paramagnetic resonance (EPR) spectroscopy. The X-band EPR spectrum of complex 1a was recorded in the solid state at liquid nitrogen temperature [Fig. S7 in SI]. The system contains V^IV metal centres,^5–9 each with a nuclear spin quantum number I = 7/2, which typically gives rise to an eight-line hyperfine pattern due to coupling with the vanadium nucleus. However, the observed spectrum is isotropic with a g-value of 1.9674 and no resolved hyperfine structure. This absence of hyperfine splitting is attributed to strong antiferromagnetic exchange interactions between neighbouring V^IV centres, likely mediated via oxo bridges. At low temperatures, such exchange coupling leads to broadening and coalescence of hyperfine features into a single broad signal. The significant broadening is evident from the elevated Gaussian (18.26 mT) and Lorentzian (37.54 mT) line width peak-to-peak (lwpp) values, further supporting the presence of exchange-narrowed, isotropic resonance. The spin densities of [V^IV₄O₄(O₃P–Me)₂]⁴⁻ (1a′) [V^IV₄ unit] at S = 2 and 1 (considering four S = ½ spin centres)¹⁵ have been shown in Fig. 2(a–d).^5–9

Fig. 2 (Left) Mulliken α-spin density percentage (a and c) and the spin density plot (b and d) of complex 1a′ (S = 2, 1) (blue colour indicates α-spin and purple colour indicates β-spin), calculated at the B3LYP-D3(BJ)/def2-SVP level of theory (isosurface value 0.007 a.u.). (Right) χT vs. T plot of complex 1a [for (V^IV₄)₄ unit] under the application of a DC field of 0.1 T. The red triangles represent data points. Each green ball represents a V^IV atom (S = ½).

3.3 Magnetic studies

Complex 1a contains sixteen V^IV ions (3d¹). Magnetic properties and EPR spectrum of a relevant V^IV ion have been previously studied in detail.¹⁵ Sixteen V^IV ions (3d¹) in 1a are divided into four V^IV₄ units, which are connected via the organic diphosphonate linkers (Fig. 1 and Scheme 1). Spin states of S = 0, 1 and 2 for each [V^IV₄O₄(O₃P–Me)₂]⁴⁻ (1a′) unit in 1a are created via magnetic interactions between four V^IV ions (S = ½, g = 2)¹⁵via four oxide-bridges and two Ar-PO₃ bridges. The Mulliken spin densities of each V^IV₄ unit at S = 2 and 1 of the modeled 1a′ are shown in Fig. 2(a–d).

The temperature-dependent DC magnetic susceptibility measurement on the powdered sample of 1a was carried out on the temperature range of 3–300 K. The experimental value of the χT product was found to be 6.15 cm³ K mol⁻¹, which is close to the theoretically calculated spin only value [6 cm³ K mol⁻¹; 16 S = ½; g = 2] for sixteen magnetically isolated V^IV atoms (S = ½) [Fig. 2].¹⁵ The χT product slowly decreases from 6.10 cm³ K mol⁻¹ at 300 K to 5.1 cm³ K mol⁻¹ at 45 K, below which it rapidly falls to 1.12 cm³ K mol⁻¹ at 3 K. The nature of the χT vs. T plot clearly suggests that the V^IV atoms of each {V^IV₄O₈} unit are antiferromagnetically coupled [Fig. 2 inset].⁵⁹ Finally, the four individual {V^IV₄O₈} units in complex 1a are expected to interact via antiferromagnetic interaction below 10 K, since the inter-unit distance between the {V^IV₄O₈} units is large.^5–9 The M vs. H plot show that spin states of S = 0 and 1 of each {V^IV₄} unit are mostly populated at 3 K [Fig. S8 in SI].

3.4 Electrochemical energy storage studies

As mentioned earlier, the main emphasis of this work was to understand how changes in cage architectures influence the electrochemical supercapacitive behavior in close-shaped oxo-vanadate assemblies. We therefore embarked on thorough electrochemical energy storage studies with electrodes prepared with compound 1a. In order to be able to directly compare the electrochemical performance of compound 1a with the previously studied linear-shaped capsular vanadate compounds Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–PO₃}₄ ⊂ 2H₂O]·34H₂O and Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–C₆H₄–PO₃}₄ ⊂ 2DMF]·29H₂O, identical fabrication and measurement methods as reported earlier were also employed here.³⁹ Electrodes were thus fabricated using the binder-free method, and electrochemical studies were performed in the ternary solvent mixture of ethylene carbonate, propylene carbonate, and ethyl acetate in a volume ratio of 1 : 1 : 1, containing 0.5 M NaClO₄ (hitherto mentioned as 0.5 M NaClO₄ electrolyte, see Experimental section for details). A three-electrode setup was employed to perform the electrochemical studies, with 1a deposited on multi-walled carbon nanotubes-stainless steel electrodes (MWCNT@SS) as the working electrode, Ag/AgCl as the reference electrode and platinum wire as the counter electrode.

Electrochemical parameters were estimated using the following equations:

The areal capacitance (C_a) values from the cyclic voltammogram (CV) data were calculated using eqn (1), represented as^56,60

where I, ν, A, dυ and V represent the current density, scan rate, area of the electrode, area under the CV curve and potential window, respectively.

The areal capacitance, volumetric energy density (E_V) and volumetric power density (P_V) are estimated from galvanostatic charge discharge studies and expressed by the following eqn (2), (3) and (4), respectively,⁶⁰

where C_a, V, A and d represent the areal capacitance, potential window, area of the electrode and thickness of the electrode, respectively.where E_V and t_d represent the energy density and discharge time, respectively.⁵⁶

3.4.1 Electrode fabrication and characterizations. Binder-free electrode fabrication was performed using our previously reported procedure.¹⁹ In brief, functionalized MWCNT was deposited on a thoroughly cleaned stainless steel (SS) substrate to prepare the MWCNT@SS electrode. This was vertically immersed in the synthesis solution of compound 1a. The MWCNT surface provided nucleation centers for the deposition of 1a, and a uniformly 1a deposited electrode, viz., 1-MWCNT@SS was obtained after 15 days.

The phase purity of compound 1a deposited on the MWCNT@SS substrate [1-MWCNT@SS] was assessed through powder X-ray diffraction (P-XRD) analysis. The experimental diffraction patterns of the 1-MWCNT@SS electrode, compared to the combined simulated P-XRD patterns of 1a, as well as the MWCNT substrate, exhibit a peak-by-peak match [Fig. S9 in SI]. Additionally, the sharpness of the peaks observed in the experimental P-XRD patterns of 1-MWCNT@SS indicates the retention of crystallinity for the compound upon deposition onto the MWCNT@SS substrate.

In order to confirm the identical electronic nature of the deposited phase on 1-MWCNTvs. pristine 1a itself, multiple spectroscopic methods were employed, i.e., infrared (IR) and X-ray photoelectron (XPS) spectroscopy. A comparative IR spectral analysis of 1a and the 1-MWCNT@SS electrode clearly shows the anchoring of the oxo-vanadate polyanions on the MWCNT@SS substrate [Fig. S10 in SI]. As expected, no distinct peaks were observed in the IR spectrum of the MWCNT material alone. The abovementioned characteristic peaks are also observed in the spectra of 1-MWCNT@SS, with slightly red-shifted peaks, confirming the attachment of diphosphonate groups and oxovanadium clusters in compound 1a on the MWCNT@SS substrate.

X-ray photoelectron spectroscopy (XPS) was performed to compare the electronic environments of the pristine vanadium-based material and its composite coated on multi-walled carbon nanotubes (MWCNTs). Both samples exhibit well-defined O 1s and V 2p peaks, confirming the presence of lattice oxygen and vanadium in the +4-oxidation state. In the composite sample, a slight shift towards lower binding energies is observed [Fig. S11 in SI]. This shift indicates an increase in the local electron density, suggesting interfacial electronic interaction between the vanadium species and the MWCNTs. The electron redistribution is likely due to charge transfer or orbital overlap at the oxide-carbon interface. The peak positions remain sharp and symmetric, confirming the preservation of the V⁴⁺ oxidation state.⁶¹ Enhanced spectral intensity in the composite also reflects better dispersion and surface accessibility. These modifications highlight the favourable impact of MWCNTs on the electronic structure [Fig. S12 in SI]. The observed shift supports improved conductivity and charge transport.⁶² Overall, the XPS results confirm successful integration without structural disruption, making the composite promising for electrochemical applications.

3.4.1.1 Surface morphological studies. The surface morphology of the bare MWCNT@SS substrate and the 1-MWCNT@SS electrode was characterized using field emission scanning electron microscopy (FE-SEM). The FE-SEM images of the bare MWCNT@SS electrode clearly depict the characteristic multi-tube structure of the MWCNTs [Fig. 3(a)], which provides a physical interface conducive to the subsequent growth of the crystalline phase of 1a on the MWCNT@SS surface. The surface morphology of 1-MWCNT@SS demonstrates the growth of the crystalline phases of compound 1a on MWCNT@SS [Fig. 3(b) and (c)], and FE-SEM images of 1-MWCNT@SS revealed the deposition of crystalline phases of compound 1a, having dimensions in µm. Energy-dispersive X-ray (EDX) spectroscopy was performed to obtain the elemental composition within the electrode material of 1-MWCNT@SS [Fig. 3(d), the respective percentage weight composition is provided as an inset table]. The EDX spectra confirm the existence of the elements V, P, C, N, O, and Na. The uniform distribution of the elements across the electrode surface was further revealed by elemental mapping analysis of the electrode [Fig. S13 in SI], affirming the compositional homogeneity of the electrode material. In order to assess the surface roughness and thickness of the electrode material, high-magnification polarized optical microscopic studies were performed. From such studies, an average thickness of ≈210.5 µm for the 1-MWCNT@SS electrode was estimated [see Fig. 4(a and b) for the cross-sectional optical micrograph of the electrodes, Fig. S14 and Table S1 in SI provide electrode surface parameters].

Fig. 3 (a) FESEM images of bare MWCNT at magnification of 1 µm, (b) and (c) FESEM images of 1-MWCNT@SS at magnifications of 10 and 3 µm, respectively, and (d) EDX spectra of 1-MWCNT@SS, elemental composition in wt% is displayed as a table in the inset.

Fig. 4 Cross-sectional optical micrograph of 1-MWCNT@SS. (a) Side-on view showing the thickness of the deposited active layer over MWCNT@SS. (b) Top-down view illustrating the crystalline morphology and surface distribution of the active material.

In order to have a better understanding of the change in the surface architecture and therefore the electrochemical receptiveness of the electrode material upon the deposition of compound 1a, BET (Brunauer–Emmett–Teller) surface area and BJH (Barrett–Joyner–Halenda) pore size analyses were carried out on electrode materials obtained before and after such deposition. Bare MWCNT exhibits a Type II isotherm, with a high BET surface area of 85.11 m² g⁻¹ and large pore volume (1.020 cm³ g⁻¹), consistent with an open meso/macro-porous carbon network that supports rapid ion transport and EDLC-type charge storage. In contrast, material from the 1-MWCNT@SS electrode exhibits a low uptake incipient Type IV isotherm with a narrow hysteresis loop, reflecting limited mesoporosity arising from the surface growth of the oxo-vanadate compound over MWCNT@SS. Although the pore radius (1.49 nm) remains within the microporous regime, there is a significant reduction in the accessible pore volume (0.0218 cm³ g⁻¹) and surface area (1.419 m² g⁻¹). This substantial decrease in the surface area upon deposition of the compound indicates occupancy of the surface voids of the MWCNT material by the compound 1a. This leads to synergistic interactions between the electric double layer capacitive and pseudocapacitive materials, leading to an enhancement in the charge storage capability (vide infra).

3.4.2 Cyclic voltammetry. Cyclic voltammetry (CV) investigations of the 1-MWCNT@SS electrode were performed at room temperature (27 °C) in a 0.5 M NaClO₄ electrolyte. To enhance the electrochemical performance, the potential window was systematically optimized as 0–1.45 V (vs. Ag/AgCl), and electrochemical assessments were carried out at varying scan rates ranging from 5 to 100 mV s⁻¹ [Fig. 5(a)]. The pronounced redox peaks in the CV profile indicate a reversible faradaic charge storage process occurring in conjunction with electric double-layer capacitance (EDLC), confirming the hybrid nature of the charge storage mechanism. CV measurements revealed an impressive areal capacitance of 42.67 mF cm⁻² (obtained from eqn (1)) at a scan rate of 5 mV s⁻¹. Even at an elevated scan rate of 100 mV s⁻¹, the material maintained a notable capacitance of 23.72 mF cm⁻², highlighting its robust charge storage capability across varying scan rates, as illustrated in Fig. 5(b).

Fig. 5 Electrochemical performance of the prepared electrode in a 0.5 M NaClO₄ electrolyte solution. (a) CV plot of 1-MWCNT@SS with varying scan rates of 5–100 mV s⁻¹. (b) Areal capacitance variation curve of 1-MWCNT@SS with respect to varying scan rates [inset shows the 1-MWCNT@SS electrode]. (c) GCD plot with varying current densities of 0.2–1.6 mA cm⁻². (d) Areal capacitance variation as per concerning current density values, (e) iR_dropvs. specific current (A g⁻¹), and (f) coulombic efficiency at different current density.

3.4.3 Galvanostatic charge–discharge. In order to evaluate the charge dynamics and cycling stability of the electrode material, galvanostatic charge–discharge (GCD) measurements were performed for the electrode 1-MWCNT@SS. The GCD tests, performed at various current densities ranging from 0.2 to 1.6 mA cm⁻² within a fixed potential window, exhibit non-linear charge–discharge profiles. This deviation from ideal capacitive behavior suggests the presence of a reversible redox-based charge storage mechanism. As illustrated in Fig. 5(c), the GCD curves display an irregular triangular shape, with an initial voltage drop associated with the resistive kinetics at the electrode–electrolyte interface, characteristic of pseudocapacitive behavior. Fig. 5(d) depicts the variation of the areal capacitance with the increase in current density from 58.59 mF cm⁻² (obtained from eqn (2)) at the scan rate of 0.2 mA cm⁻², which outreaches to 4.61 mF cm⁻² at 1.6 mA cm⁻². The equivalent series resistance (ESR) was determined from the linear fit slope of iR_dropvs. current density plot, giving the value of 0.91 Ω [Fig. 5(e)], reflecting the efficient charge transport within the electrode system. Fig. 5(f) illustrates the coulombic efficiency as a function of the current density, yielding the value of 83.17% at 0.2 mA cm⁻² which reaches to 97.95% at 1.6 mA cm⁻², reflecting the improved charge–discharge reversibility at higher current densities.

3.4.4 Electrochemical impedance spectroscopy (EIS). Electrochemical impedance spectroscopy (EIS) was employed to investigate the interfacial electrochemical characteristics of the 1-MWCNT@SS electrode over a frequency range of 10⁵ to 0.1 Hz. EIS provides valuable insights into charge transfer processes, reaction kinetics and electrode–electrolyte interactions. The Nyquist plot [Fig. S16(a) in SI] displays the real (Z′) versus imaginary (Z″) components of impedance, and was fitted using ZSimpWin software with an equivalent circuit R(CR)(QR)W model. The intercept at high frequency on the real axis corresponds to the solution resistance (R_s), encompassing contributions from the electrolyte's ionic resistance, the electrode–electrolyte interface and the inherent resistance of the stainless steel (SS) current collector. The semicircle in the high- to mid-frequency region represents the charge-transfer resistance (R_ct), associated with faradaic redox reactions at the electrode surface.

In this system, R_s and R_ct were determined to be 23.27 and 2.22 Ω cm², respectively. A low R_ct value indicates efficient charge transfer processes at the electrode–electrolyte interface, contributing to enhanced electrochemical performance and increased capacitance. This observation is consistent with rapid and reversible redox transitions involving the vanadium(IV) oxidation state, signifying favorable electron-transfer kinetics. The observed semicircle arises from the double-layer capacitance (C_DL), formed at the electrode–electrolyte interface. The constant phase element (CPE), included to account for surface heterogeneity, follows the expression⁶³

where T is the CPE constant related to the magnitude of the pseudo-capacitance, n is the phase shift parameter (0 ≤ n ≤ 1), j is the imaginary unit, and ω is the angular frequency (rad s⁻¹).

The parameter n serves as an indicator of the deviation from ideal capacitive behavior: n = 1 corresponds to a pure capacitor, n = 0.5 indicates Warburg-type diffusion behavior, and n < 1 reflects the presence of distributed elements typically due to electrode surface heterogeneity. In this study, the CPE exponent n was found to be 0.81 (obtained from eqn (5)); indicating significant deviation from ideal capacitive behavior, and suggesting that the electrode exhibits a pseudocapacitive response rather than purely electrical double-layer capacitance. This value reflects the influence of surface inhomogeneities and redox-active sites that contribute to faradaic charge storage processes. The sub-unit n also implies the presence of a distributed time constant, which is consistent with ion diffusion through porous channels and interfacial charge transfer resistance-characteristics, typical of pseudocapacitive materials.

3.4.5 Stability tests. To evaluate the long-term electrochemical stability of the 1-MWCNT@SS electrode, a study of an extended 10 000 GCD stability cycles at 1.4 mA cm⁻² shows 66.05% of its initial capacitance retention while maintaining a high coulombic efficiency of 97.22% [Fig. 6]. This confirms the excellent charge–discharge reversibility of the electrode. Also, stability tests with 2000 continuous CV cycles at a scan rate of 200 mV s⁻¹ showed a retention of 92.8% of its initial capacitance [Fig. S16(b) in SI]. This high retention is attributed to the robust inorganic-organic hybrid framework of the electrode material, enhancing the structural integrity during repeated redox cycling. Additionally, use of an organic electrolyte minimizes dissolution or leaching of the active species.

Fig. 6 Capacity retention (left versus bottom) and coulombic efficiency (right versus bottom) at 1.4 mA cm⁻² current density. The inset shows the charge–discharge curves for different cycles.

To assess the structural integrity of the electrode material after electrochemical cycling, P-XRD analysis was performed on the 1-MWCNT@SS electrode. No significant changes in the diffraction patterns of the electrode materials were observed [Fig. S17(a) in SI], with notable variations in peak intensities only, indicating minor alterations in the structural characteristics of the material following electrochemical cycling, which are associated with some degradation of the quality of the material on the electrode surface, such as crystallinity.

To further investigate the structural evolution and degradation mechanisms of the electrode materials, Raman spectroscopic analysis was performed. The spectra exhibit characteristic peaks in the range of 344–356 cm⁻¹, attributed to V–O–V bending modes; 518 cm⁻¹, corresponding to bridging V–O–P symmetric stretching vibrations; 625–642 cm⁻¹, assigned to the V–O–P asymmetric stretch coupled with the P–O–C/V–O–C modes of aromatic diphosphonate; and 900–1000 cm⁻¹, associated with terminal V=O stretching vibrations. Additionally, peaks in the 1000–1100 cm⁻¹ range were linked to P–O stretching, while those between 1250–1650 cm⁻¹ were ascribed to aromatic C=C stretching.³⁹ A comparative analysis of the Raman spectra for 1-MWCNT@SS electrodes before and after electrochemical stability testing [Fig. S17(b) in SI] revealed some noticeable changes in peak positions and intensities, particularly in regions corresponding to V–O–V, V=O, and C=C vibrations. These spectral modifications indicate partial degradation of the vanadate species and suggest Na⁺ ion intercalation into the electrode matrix during the electrochemical cycling process.

3.5 Liquid-state configured symmetric device and performance [Capsular Generation 2 device]

Encouraged by the excellent performance of the 1-MWCNT@SS electrode, a symmetric liquid-state supercapacitor device (dimensions 3.5 × 3.5 cm²) was fabricated using two identical 1-MWCNT@SS electrodes, separated by a porous membrane soaked with 0.5 M NaClO₄ electrolyte [see Fig. 7(a) for a schematic representation of the device, as well as Fig. 7(b) and (c) for the actual working device]. All electrochemical measurements such as cyclic voltammetry (CV), galvanostatic charge–discharge (GCD), electrochemical impedance spectroscopy (EIS), and long-term cycling stability, were carried out at room temperature.

Fig. 7 (a) Device and schematic of the fabricated symmetric liquid device. Evaluation of the practical applicability of the fabricated device with the (b) light-emitting diode (LED) panel and (c) 1 V powered small fan.

CV profiles were recorded at varying scan rates ranging from 5 to 100 mV s⁻¹, within a wide potential window of 2.6 V, with distinct oxidation-reduction peaks being observed in the cyclic voltammogram. At the onset itself, this was a significant enhancement from Capsular Generation 1 devices [Fig. 8(c)]. CV measurements were conducted across scan rates ranging from 5 to 100 mV s⁻¹, revealing stable redox activity over the extended potential window of 2.6 V, indicative of the device's wide electrochemical operating range and suitability for high-voltage energy storage applications. The CV curves retained a consistent shape across scan rates, indicating good rate capability and electrochemical reversibility [Fig. 8(a)], and exhibit quasi-rectangular shape with broad redox peaks, indicating the hybrid nature of the charge storage. At a scan rate of 5 mV s⁻¹, the device delivered a high areal capacitance of 10.78 mF cm⁻² (obtained from eqn (1)) [Fig. 8(b)], which is significantly higher in comparison with Generation 1 devices [Fig. 8(d)]. To gain deeper insight into the underlying charge storage mechanism, the b-value was calculated from the relationship between the peak current and scan rate (i = aν^b), yielding a value of 0.72, consistent with that observed for the individual electrode [Fig. S18(c) in SI]. This value supports a mixed charge storage mechanism dominated by diffusion-controlled pseudocapacitive behavior. Furthermore, the total stored charge (Q_t) was deconvoluted into surface capacitive (Q_s) and diffusion-controlled (Q_d) contributions. At 100 mV s⁻¹, Q_s and Q_d contribute 40.68% and 59.32% of the total charge [Fig. S18(d) in SI], respectively, highlighting the predominance of the diffusion behavior under high-rate conditions [see Section 2.6 for details on electrode]. The analysis revealed a predominant diffusion-controlled behavior, indicative of the involvement and mobility of Na⁺ ions and the participation of faradaic redox processes (vide infra).

Fig. 8 Supercapacitor performance of the Generation 2 device. (a) Cyclic voltammogram of the Generation 2 device at varying scan rates of 5–100 mV s⁻¹. (b) Areal capacitance variation curve for the device with respect to varying scan rates. (c) Comparative CV for Generation 1 devices 1 and 2, and Generation 2 device at the scan rate of 100 mV s⁻¹. (d) Areal capacitance variation curve for Generation 1 and 2 devices with respect to varying scan rates of 5–100 mV s⁻¹.

The charge storage performance of the device was further investigated through galvanostatic charge–discharge (GCD) measurements at the current ranging from 1.5 to 3.0 mA [Fig. 9(a)], within the constant large potential window of 0 to 2.6 V. The asymmetric nature of the GCD curves reveal three distinct regions: an initial voltage-drop corresponding to the intrinsic resistance of the electrode material, a linear segment reflecting electric double-layer capacitive (EDLC) behavior, and a curved region associated with faradaic interactions, indicative of pseudocapacitive contribution. The maximum areal capacitance of 1.198 mF cm⁻² was obtained at a current of 1.5 mA and outreaches to 0.671 mF cm⁻² areal capacitance (obtained from eqn (2)) at 3.0 mA [Fig. 9(b)].

Fig. 9 Galvanostatic charge discharge (GCD) performance of the Generation 2 device, (a) GCD plot for the device with varying current values of 1.5–3.0 mA, (b) areal capacitance variation of the device as per concerning current values, (c) iR_dropvs. specific current (A g⁻¹) plot, and (d) coulombic efficiency at different current values.

The equivalent series resistance (ESR) of the system was determined to be 0.89 Ω from the slope of the iR_dropvs. specific current plot [Fig. 9(c)], indicating low internal and diffusion resistance conducive to high-power performance with minimal thermal losses. Coulombic efficiency, calculated as the ratio of discharge to charge time, reached a value of 85.14% at 3.0 mA from 58.89% at 1.5 mA, reflecting the excellent reversibility of the electrode material [Fig. 9(d)].

EIS was employed to evaluate the interfacial characteristics and electrochemical performance of the device. The Nyquist plot of the device [Fig. S18(a) in SI] was fitted with circuit R(Q(R(QR))), which accounts for both bulk and interfacial processes. The fitted EIS data further revealed key impedance parameters, including solution resistance (R_S = 2.48 Ω), charge transfer resistance (R_ct = 37.78 Ω), and leakage resistance (R_L = 5587 Ω). These values provide insights into the ionic conductivity of the electrolyte and the efficiency of charge transfer at the electrode–electrolyte interface. The EIS response is governed by two principal contributions – ionic resistance within the electrolyte and electronic resistance associated with charge transport, highlighting the complex interplay between ionic mobility and electronic conductivity in determining the overall device performance.

To assess the long-term operational stability, the device was subjected to 2000 continuous CV cycles at a scan rate of 200 mV s⁻¹ [Fig. S18(b) in SI]. After cycling, the capacitance retention was 89.75%, indicating good electrochemical durability. The slight reduction in capacity is attributed to the influence of the extended potential window, which can compromise redox reversibility over prolonged operation.

The practical applicability of the fabricated supercapacitor device was evaluated through functional performance tests. As shown in the schematic and photographic images in Fig. 7, the device was capable of powering a light-emitting diode (LED) panel comprising 21 red LEDs arranged to form the acronym “VNIT”. A brief charging duration of 10 seconds enabled the device to sustain LED illumination for approximately 64 seconds [Fig. 7(a)]. Furthermore, under identical charging conditions, the device successfully powered a small fan for about 10 seconds, as illustrated in Fig. 7(b). These demonstrations underscore the potential of the device for deployment in lightweight, flexible, and rapidly rechargeable energy storage applications.

3.5.1 Comparative electrochemical performance between Capsular Generation 1 and Generation 2 devices. Towards assessing the variations in the electrochemical performance of both generation devices, identical experimental conditions were maintained for the linear-shaped capsular compounds Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–PO₃}₄ ⊂ 2H₂O]·34H₂O and Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–C₆H₄–PO₃}₄ ⊂ 2DMF]·29H₂O,³⁹ and the dumbbell-shaped capsular compound 1a, Na₁₆(Me₃C–NH₃)₆[H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]·44H₂O.

The newly developed Capsular Generation 2 device displays markedly superior electrochemical characteristics, underscoring its potential for advanced energy storage applications [for comparison between Generation 1 and Generation 2 electrodes only, see Fig. S19 in SI]. The comparative performance metrics for all three devices are summarized in Table 2, highlighting the significant advancements offered by the Generation 2 system in terms of electrochemical functionality and design efficiency. With a significantly extended potential window of 0 to 2.6 V, the Generation 2 system operates beyond the voltage limits of Generation 1 Device 1 with Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–PO₃}₄ ⊂ 2H₂O]·34H₂O (potential window of 0 to 1.75 V) and Generation 1 Device 2 with Na₈[H₂(V^VO)₂(V^IV₄O₈)₂{O₃P–C₆H₄–C₆H₄–PO₃}₄ ⊂ 2DMF]·29H₂O (potential window of 0.13 to 1.75 V); allowing for greater energy harvesting per charge–discharge cycle. This broader electrochemical window translates into a considerable enhancement in energy density (obtained from eqn (3)), reaching 11.24 mWh cm⁻³ for Generation 2, compared to 4.92 mWh cm⁻³ and 5.77 mWh cm⁻³ for Generation 1 Devices 1 and 2, respectively. A similar trend is observed in the power density (obtained from eqn (4)), which rises sharply to 977.54 mW cm⁻³ for the Generation 2 device, far exceeding the performance of the earlier generation devices (348.46 and 353.83 mW cm⁻³, for Generation 1 Devices 1 and 2, respectively). The Generation 2 device also exhibits excellent cycling stability, retaining 89.75% of its capacitance over 2000 cycles, indicative of robust structural and electrochemical resilience. Along with a low solution resistance (R_S = 2.48 Ω) and a favorable charge transfer resistance (R_ct = 37.78 Ω), reflecting improved ionic conductivity and efficient redox kinetics, these advantages are attributed to the architectural refinement achieved through the use of a rigid bent-shaped ligand, which creates larger and more accessible internal cavities for ion transport.

Table 2 Comparative analysis details of Generation 1 and Generation 2 devices

Sl. no.	Parameters	Generation 1		Generation 2 (compound 1a)
		Device 1	Device 2
1	Potential window (V)	0 to +1.75	0.13 to +1.75	0 to +2.6
2	Energy density (mWh cm^−3)	4.92	5.77	11.24
3	Power density (mW cm^−3)	348.46	353.83	977.54
4	Stability cycle and capacity retention (%)	2000 and 88.23%	2000 and 85.01%	2000 and 89.75%
5	Charge transfer resistance (Rct, Ω)	33.89	61.7	37.78
6	Solution resistance (RS, Ω)	4.93	6.12	2.48

---

The Ragone plot presented in Fig. 10 offers a comparative evaluation of the electrochemical performance of three hybrid supercapacitor devices based on different cage architectures: 1,4-substituted linear-shaped (Generation 1 Device 1), 4,4′-substituted linear-shaped (Generation 1 Device 2), and the Generation 2 dumbbell-shaped system. The plot clearly illustrates the superior performance of the Generation 2 device, which occupies the upper-right region of the plot, indicative of its ability to simultaneously deliver high energy density and high-power density. While both linear-shaped capsular Generation 1 devices demonstrate energy densities in the range of ∼1–5 mWh cm⁻³ and power densities ranging from 300 to 2000 mW cm⁻³; the Generation 2 system significantly outperforms them, achieving energy densities close to 11 mWh cm⁻³ at comparable or even higher power outputs (holding the range from 1000 to 2000 mW cm⁻³). This favorable positioning on the Ragone plot highlights a critical trade-off achieved by the Generation 2 system: the ability to store a greater amount of energy without sacrificing the rapid delivery of that energy, which is essential for high-performance energy storage applications.

Fig. 10 Comparative Ragone plot of Generation 1 Devices 1 and 2 vs. Generation 2 device.

3.6 Understanding the underlying electrochemical reaction mechanism for energy storage

In our earlier studies we had observed that throughout the charge–discharge process, Na⁺ ions from the electrolyte diffuse toward the electrode–electrolyte interface, wherein they undergo reversible intercalation and deintercalation into the active electrode matrix.^39,64 This ion transport is accompanied by surface and near-surface faradaic redox reactions, facilitated by the presence of electrochemically active species.

The involvement of V^IV atoms in complex 1a contributes significantly to the pseudocapacitive behavior, enhancing charge storage capacity, rate capability, and cycling stability. The interplay between capacitive ion adsorption and redox-driven ion insertion mechanisms enables improved electrochemical performance. This can be realized by the continuous intercalation – deintercalation of charged species, such as sodium ions, during the electrochemical process. Thus, the electrochemical reaction happening during the charge–discharge process can be represented as follows:

[H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]²²⁻ + zNa⁺ + ze⁻ ↔ Na_z[H₁₀(V^IV₄O₈)₄{O₃P–C₆H₄–O–C₆H₄–PO₃}₈ ⊂ 4DMF·4H₂O]^(22−z)− (6)

Accumulation of sodium into the electrode 1-MWCNT@SS was confirmed with EDX studies of the electrode obtained after electrochemical stability tests [Fig. S20 in SI], which showed a significant increase in the percentage of sodium within the electrode, in comparison with pristine electrodes.

Furthermore, during cyclic voltammetry studies at elevated scan rates, the CV curves retained their characteristic shape, accompanied by a noticeable shift toward higher current densities, which indicate the electrochemical stability. With a reduction in the scan rate, we can observe enhancement of the specific capacitance, as the electrolyte ions get sufficient time to diffuse into the interlayer of electrode material. Such rate of diffusion of electrolytic ions in capsular polyanions is directly controlled by its shape and size. With a more open dumbbell shape, diffusion of ions within the cavity of polyanion 1 is more effortless, as compared with the linear-shaped capsular polyanions. This effect directly influences the electrochemical efficiency of devices prepared with 1-MWCNT@SS [Capsular Generation 2], now showing significant improvement in its performance with respect to Capsular Generation 1 devices.

In hybrid supercapacitors composed of both pseudocapacitive and faradaic materials, the charge storage is attributed to the dual nature of charge storage mechanisms, encompassing both surface-controlled and diffusion-controlled processes. As mentioned earlier, in order to investigate the charge storage mechanism and reaction kinetics of the 1-MWCNT@SS electrode, the b-value for the device was evaluated using power-law analysis, which defines the relationship between the electrode current (i) and scan rate (v) as:⁶⁵

log(i) = log(a) + b log(v) (7)

and serves as an indicator of the prevailing charge storage mechanism. A b-value of 0.5 suggests a diffusion-controlled process, typical of battery-type behavior, whereas a value of 1.0 indicates a surface-controlled process, characteristic of ideal capacitive behavior. Values between 0.5 and 1.0 denote a hybrid mechanism involving both faradaic and non-faradaic contributions.⁶⁵ For the electrode 1-MWCNT@SS, the b-value obtained from the log(i) versus log(v) plot within the potential window of 0.3 to 1.1 V [Fig. S21(a) in SI], yields an average of 0.78 [Fig. S21(b) in SI], confirming the coexistence of capacitive and diffusion-controlled processes, indicative of the hybrid nature of the electrode material.

In any pseudocapacitive electrode material, the total charge storage (Q_t) consists of two primary components: surface capacitive charge (Q_s) and diffusion-controlled charge (Q_d). These correspond to fast surface redox reactions and slower intercalation/deintercalation processes, respectively. The Trasatti method⁶⁵ was employed to quantitatively separate these contributions by plotting Q_t against the inverse square root of the scan rate (ν^−0.5) [Fig. S21(c) in SI], where the y-intercept represents Q_s. Since surface capacitive processes are independent of the scan rate, Q_s remains constant, whereas Q_d increases at lower scan rates due to enhanced ion diffusion.

In the present system, the surface capacitive contribution is attributed to multi-walled carbon nanotubes (MWCNTs), while the diffusion-controlled response originates from the oxovanadate-polyanion 1. As the scan rate increases, Q_s becomes more dominant, while Q_d decreases, indicating a shift from diffusion-controlled to surface-controlled charge storage. At a scan rate of 100 mV s⁻¹, Q_s and Q_d account for 77.3% and 22.6% of the total charge [Fig. S21(d) in SI], respectively, demonstrating the electrode's predominant capacitive behavior under high-rate conditions.

4. Conclusions

The rich and reversible redox behaviour of polyoxovandates (POVs) makes them strong candidates for pseudocapacitive energy storage. Using bent-shaped diphosphonate ligands as linkers, we have synthesized a unique dumbbell-shaped capsular polyanionic oxo-vanadate assembly composed of cubane-shaped {V^IV₄O₈} units. Magnetic studies show weak anti-ferromagnetic interactions between individual V(IV) atoms within each such {V^IV₄O₈} unit. Mulliken spin densities of each V^IV₄ unit (1a′) at S = 2 and 1 have been computed and correlated with the magnetic property and EPR spectrum of 1a. This architecture leverages both the redox activity and structural robustness of POVs, yielding a hybrid framework with promising electrochemical characteristics. Our newly synthesized {V₄O₈}-based polyanionic complex was thus evaluated for potential application in hybrid supercapacitors (SCs), which benefit from both electric double-layer capacitance (EDLC) and faradaic pseudocapacitance. Despite the intrinsic challenges posed by the high solubility and low surface area of polyanionic complexes, as well as their weak electrode adhesion, we have formulated a binder-free electrode fabrication strategy. Using solvent evaporation techniques, we achieved the growth of uniform particle films on multi-walled carbon nanotubes (MWCNTs), yielding a robust, hybrid electrode. Devices prepared with these electrodes show enhanced performance, with an improved potential window of 2.6 V and volumetric energy and power densities of 11.24 mWh cm⁻³ and 977.54 mW cm⁻³, respectively; which is a significant enhancement of previously studied capsular polyoxovanadate complexes. The cluster's structural integrity allows for operation across such broad potential windows, while its redox-active vanadium centres enable efficient charge storage through reversible faradaic processes. This dual functionality helps address limitations in energy density, while maintaining long-term cycling stability, which is essential for sustainable energy storage systems. Practical demonstration of the fabricated device has been demonstrated against an LED panel and a small fan. Our {V₄O₈} cubane-based complex thus presents a promising approach for next-generation supercapacitor design.

Author contributions

Shikha Singh: conceptualization, methodology, validation, formal analysis, investigation, data curation, writing – original draft. T. Kedara Shivasharma: methodology, validation, formal analysis, investigation, data curation, writing – original draft. Subuhan Ahamed: formal analysis, investigation, data curation, writing – original draft. Saurav Ghosh: formal analysis, investigation, data curation, DFT calculations, writing – original draft. Kartik Chandra Mondal: conceptualization, writing – review & editing, supervision, funding acquisition. Babasaheb R. Sankapal: conceptualization, writing – review & editing, supervision, funding acquisition. Abhishek Banerjee: conceptualization, writing – review & editing, supervision, funding acquisition.

Conflicts of interest

There are no conflicts to declare.

Data availability

CCDC 2455377 contains the supplementary crystallographic data for this paper.⁶⁶

The datasets supporting this article have been uploaded as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5ta07057k.

Acknowledgements

SS thanks the Director, VNIT, for providing a teaching assistantship. TKS is thankful to DST (DST/INSPIRE/03/2022/000141) for providing an INSPIRE fellowship. The authors acknowledge Prof. Dieter Fenske and Karlsruhe Nano Micro Facility (KNMF), a Helmholtz research infrastructure at the Karlsruhe Institute of Technology (KIT), for performing the single crystal X-ray diffraction measurements. The authors ackowledge Dr S. K. Nayak at VNIT Nagpur for the optical microscope imaging facility. All authors acknowledge the DST FIST projects (SR/FIST/CSI-279/2016(C) and SR/FST/PSI/2017/5(C)) for providing the characterization resources.

References

1. K. Y. Monakhov, W. Bensch and P. Kogerler, Chem. Soc. Rev., 2015, 44, 8443.
2. J. Livage, Coord. Chem. Rev., 1998, 178–180, 999.
3. Y. Hayashi, Coord. Chem. Rev., 2011, 255, 2270.
4. N. I. Gumerova and A. Rompel, Nat. Rev. Chem., 2018, 2, 0112.
5. G. G. Aleksandrov, V. S. Sergienko and E. G. Afonin, Crystallogr. Rep., 2001, 46, 46.
6. I. S. Tidmarsh, E. Scales, P. R. Brearley, J. Wolowska, L. Sorace, A. Caneschi, R. H. Laye and E. J. L. McInnes, Inorg. Chem., 2007, 46, 9743.
7. R. Ban, Y. L. Liang, P. T. Ma, D. D. Zhang, J. Y. Niu and J. P. Wang, Inorg. Chem. Commun., 2016, 71, 65.
8. S. Tomane, E. López-Maya, S. Boujday, V. Humblot, J. Marrot, N. Rabasso, J. Castells-Gil, C. Sicard, A. Dolbecq, P. Mialane and A. Vallée, Nanoscale Adv., 2019, 1, 3400.
9. D. Thakre, S. R. Ali, S. Mehta, N. Alam, M. Ibrahim, D. Sarma, A. Mondal, M. De and A. Banerjee, Cryst. Growth Des., 2021, 21, 4285.
10. M. Dorn, J. Kalmbach, P. Boden, A. Päpcke, S. Gómez, C. Förster, F. Kuczelinis, L. M. Carrella, L. A. Büldt, N. H. Bings, E. Rentschler, S. Lochbrunner, L. González, M. Gerhards, M. Seitz and K. Heinze, J. Am. Chem. Soc., 2020, 142, 7947.
11. C. Van Stappen, D. Maganas, S. DeBeer, E. Bill and F. Neese, Inorg. Chem., 2018, 57, 6421.
12. S. L. Castro, Z. Sun, C. M. Grant, J. C. Bollinger, D. N. Hendrickson and G. Christou, J. Am. Chem. Soc., 1998, 120, 2365.
13. J. P. Zobel, T. Knoll and L. González, Chem. Sci., 2021, 12, 10791.
14. H. El Moll, W. Zhu, E. Oldfield, L. M. Rodriguez-Albelo, P. Mialane, J. Marrot, N. Vila, I. M. Mbomekallé, E. Rivière, C. Duboc and A. Dolbecq, Inorg. Chem., 2012, 51, 7921.
15. P. Singh, S. Schlittenhardt, D. Thakre, S. K. Kushvaha, S. Kumar, H. S. Karnamkkott, M. Ruben, M. Ibrahim, A. Banerjee and K. C. Mondal, Cryst. Growth Des., 2022, 22, 5666.
16. K. S. Morla, D. Thakre, A. K. Mourya, P. Singh, H. S. Karnamkkott, S. Ahamed, S. Sahoo, I. Bhattacharya, S. Banerjee, D. Sarma, K. C. Mondal, S. K. Ghosh, A. V. Wankhade and A. Banerjee, Chem. Commun., 2025, 61, 8919.
17. R. H. Laye, Q. Wei, P. V. Mason, M. Shanmugam, S. J. Teat, E. K. Brechin, D. Collison and E. J. L. McInnes, J. Am. Chem. Soc., 2006, 128, 9020.
18. J. M. Gonçalves, M. Ireno da Silva, L. Angnes and K. Araki, J. Mater. Chem. A, 2020, 8, 2171.
19. K. S. Morla, D. Thakre, T. B. Deshmukh, S. Malo, S. Ahamed, R. K. Aparna, S. Sahoo, D. Sarma, K. C. Mondal, B. R. Sankapal and A. Banerjee, J. Mater. Chem. A, 2024, 12, 22013.
20. V. Augustyn, P. Simon and B. Dunn, Energy Environ. Sci., 2014, 7, 1597.
21. P. K. Sharma, A. Arora and S. K. Tripathi, J. Energy Storage, 2019, 21, 801.
22. F. Yu, T. Huang, P. Zhang, Y. Tao, F.-Z. Cui, Q. Xie, S. Yao and F. Wang, Energy Storage Mater., 2019, 22, 235.
23. M. R. Lukatskaya, B. Dunn and Y. Gogotsi, Nat. Commun., 2016, 7, 12647.
24. W. Zuo, R. Li, C. Zhou, Y. Li, J. Xia and J. Liu, Adv. Sci., 2017, 4, 1600539.
25. Y. Shao, M. F. El-Kady, J. Sun, Y. Li, Q. Zhang, M. Zhu, H. Wang, B. Dunn and R. B. Kaner, Chem. Rev., 2018, 118, 9233.
26. T. K. Shivasharma, N. Upadhyay, T. B. Deshmukh and B. R. Sankapal, ACS Omega, 2023, 8, 37685.
27. A. A. Vannathan, T. Kella, D. Shee and S. S. Mal, ACS Omega, 2021, 6, 11199.
28. A. Banerjee, B. S. Bassil, G. V. Roschenthaler and U. Kortz, Chem. Soc. Rev., 2012, 41, 7590.
29. J. Zubieta, Comments Inorg. Chem., 1994, 16, 153.
30. H. N. Miras, J. Yan, D.-L. Long and L. Cronin, Chem. Soc. Rev., 2012, 41, 7403.
31. A. Bijelic, M. Aureliano and A. Rompel, Chem. Commun., 2018, 54, 1153.
32. M. Aureliano, N. I. Gumerova, G. Sciortino, E. Garribba, A. Rompel and D. C. Crans, Coord. Chem. Rev., 2021, 447, 214143.
33. J. C. Pessoa, S. Etcheverry and D. Gambino, Coord. Chem. Rev., 2015, 301–302, 24.
34. J. M. Cameron, D. J. Wales and G. N. Newton, Dalton Trans., 2018, 47, 5120.
35. K. S. Morla, R. Sahu, S. Ahamed, S. Ghosh, K. C. Mondal, B. R. Sankapal and A. Banerjee, Dalton Trans., 2025, 54, 17322.
36. J. M. Breen and W. Schmitt, Angew. Chem., Int. Ed., 2008, 47, 6904.
37. J. M. Breen, R. Clérac, L. Zhang, S. M. Cloonan, E. Kennedy, M. Feeney, T. McCabe, D. C. Williams and W. Schmitt, Dalton Trans., 2012, 41, 2918.
38. M. B. Mahimaidoss, S. A. Krasnikov, L. Reck, C. I. Onet, J. M. Breen, N. Zhu, B. Marzec, I. V. Shvets and W. Schmitt, Chem. Commun., 2014, 50, 2265.
39. K. S. Morla, T. B. Deshmukh, R. K. Aparna, B. R. Sankapal and A. Banerjee, Energy Fuels, 2024, 38, 21482.
40. M. D. Medrano, H. T. Evans Jr, H. R. Wenk and D. Z. Piper, Am. Mineral., 1998, 83, 889.
41. X. Wang, L. Liu, G. Zhang and A. J. Jacobson, Chem. Commun., 2001, 2472,  10.1039/B108303A.
42. H. N. Miras, R. Raptis, P. Baran, N. Lalioti, A. Harrison and T. A. Kabanos, C. R. Chim., 2005, 8, 957.
43. H. N. Miras, R. G. Raptis, N. Lalioti, M. P. Sigalas, P. Baran and T. A. Kabanos, Chem.–Eur. J., 2005, 11, 2295.
44. G. I. Chilas, H. N. Miras, M. J. Manos, J. D. Woollins, A. M. Z. Slawin, M. Stylianou, A. D. Keramidas and T. A. Kabanos, Pure Appl. Chem., 2005, 77, 1529.
45. F.-N. Shi, F. A. A. Paz, J. Rocha, J. Klinowski and T. Trindade, Eur. J. Inorg. Chem., 2004, 2004, 3031.
46. G. A. Farnum and R. L. LaDuca, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2008, 64, m1602.
47. C. Dey, R. Das, P. Poddar and R. Banerjee, Cryst. Growth Des., 2012, 12, 12.
48. J. S. Maass, Z. Chen, M. Zeller and R. L. Luck, Dalton Trans., 2011, 40, 11356.
49. H. Zhou, J. Wang, L. Shen, P. Liang, X. Xu, B. Li, Z. Zhang, X. Zhu, Z. Kong and J. Guo, Adv. Mater., 2025, e14323.
50. D. Navas, Current Innovations in Chemical and Materials Sciences, 2024, vol. 5, p. 1.
51. H. Pang, Y. Dong, S. L. Ting, J. Lu, C. M. Li, D.-H. Kim and P. Chen, Nanoscale, 2013, 5, 7790.
52. J. Koziskova, F. Hahn, J. Richter and J. Kozisek, Acta Chim. Slov., 2016, 9, 136–140.
53. G. M. Sheldrick, Sadabs, 1996.
54. A. Spek, J. Appl. Crystallogr., 2003, 36, 7.
55. M. M. Gómez-Alcántara, A. Cabeza, M. Martínez-Lara, M. A. Aranda, R. Suau, N. Bhuvanesh and A. Clearfield, Inorg. Chem., 2004, 43, 5283.
56. B. Pandit and B. R. Sankapal, ACS Appl. Nano Mater., 2022, 5, 3007.
57. K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds Part B: Applications in Coordination, Organometallic, and Bioinorganic Chemistry, Wiley, 6th edn, 2009.
58. R. Borah, S. Lahkar, N. Deori and S. Brahma, RSC Adv., 2022, 12, 13740.
59. V. Chandrasekhar, A. Dey, T. Senapati and E. C. Sañudo, Dalton Trans., 2012, 41, 799.
60. S. A. Pande, B. Pandit and B. R. Sankapal, Mater. Lett., 2017, 209, 97.
61. G. Silversmit, D. Depla, H. Poelman, G. B. Marin and R. De Gryse, J. Electron Spectrosc. Relat. Phenom., 2004, 135, 167.
62. B. Pandit, D. P. Dubal, P. Gómez-Romero, B. B. Kale and B. R. Sankapal, Sci. Rep., 2017, 7, 43430.
63. S. S. Raut, J. A. Dhobale and B. R. Sankapal, Phys. E, 2017, 87, 209.
64. H. K. Rathore, M. Hariram, M. K. Ganesha, A. K. Singh, D. Das, M. Kumar, K. Awasthi and D. Sarkar, J. Colloid Interface Sci., 2022, 621, 110.
65. A. Agarwal and B. R. Sankapal, Chem. Eng. J., 2021, 422, 130131.
66. CCDC 2455377: Experimental Crystal Structure, 2026,  DOI:10.5517/ccdc.csd.cc2nf0qv..

---