Cu-BDC-derived hierarchically porous CuO for electrochemical charge storage and photoelectrochemical response

Abstract

Hierarchically porous CuO was prepared by calcining a Cu-benzene-1,4-dicarboxylate (Cu-BDC)-containing precursor at 500 °C and compared with conventionally synthesized CuO. X-ray diffraction identified monoclinic tenorite CuO (JCPDS 00-002-1041) after calcination, while the uncalcined intermediate also contained reflections from fcc metallic Cu (JCPDS 04-0836). The precursor-derived CuO exhibited a BET surface area of 187 m² g⁻¹ and a pore volume of 0.81 cm³ g⁻¹, compared with 24 m² g⁻¹ and 0.07 cm³ g⁻¹ for pristine CuO. Analysis of the UV-vis absorption spectra using indirect-transition Tauc plots gave apparent transition energies of 2.3 and 2.4 eV for precursor-derived and pristine CuO, respectively. On 2.5 mm–diameter glassy-carbon working electrodes carrying 1.31 ± 0.05 mg of dried composite, precursor-derived CuO delivered 503 F g⁻¹ in 1 M Na₂SO₄ and 637 F g⁻¹ in 1 M KOH at 0.5 A g⁻¹ and showed the strongest charge-storage response among the tested materials. Under AM 1.5G illumination, the same material produced a larger anodic current and retained approximately 88% of its post-transient current over 2 h. These findings show that conversion of a Cu-BDC-containing precursor provides an effective route to porous CuO electrodes with improved charge-storage performance and a stronger photoelectrochemical response.

Cu-BDC precursor conversion creates hierarchically porous CuO with a greatly increased accessible surface area, delivering enhanced charge storage and a stronger, more stable illumination-dependent anodic response.

1. Introduction

The continuing growth of global energy demand and the environmental effects associated with fossil-fuel use have intensified research into materials for electrochemical energy storage and solar-energy conversion.^1,2 Electrochemical capacitors are attractive for applications that require rapid charge–discharge, high power delivery, and extended cycling because charge is stored through fast processes at or near the electrode–electrolyte interface.^3,4 Photoelectrochemical (PEC) measurements address a different function: they probe the response of a semiconductor–electrolyte junction to illumination and applied potential, including the generation, separation, transport, and interfacial transfer of photogenerated carriers.^5,6 These two applications should not be treated as mechanistically equivalent. Nevertheless, both depend strongly on the fraction of surface that is accessible to the electrolyte, the continuity of ion- and charge-transport pathways, and the resistance of the solid–liquid interface. Controlling the morphology and pore structure of an earth-abundant oxide can therefore provide a useful materials-design route for examining charge-storage and illumination-dependent electrochemical behavior within the same compositional platform.⁷

Copper(ii) oxide (CuO) is an abundant p-type transition-metal oxide that has been investigated in electrochemical storage, catalysis, sensing, and photo electrochemistry.^8–11 Its appeal arises from the abundance and low cost of copper, the accessibility of several wet-chemical synthesis routes, and the sensitivity of its response to particle shape and surface structure. Bulk CuO is commonly reported to have a narrow optical gap in the approximate range 1.2–1.7 eV. The transition energy inferred from a Tauc construction can, however, vary with crystallinity, particle dimensions, optical measurement mode, background treatment, and the interval selected for extrapolation. In aqueous electrolytes, CuO electrodes may show a combination of interfacial charging, polarization, and surface faradaic processes rather than a single idealized storage mechanism. Dense agglomerates can further restrict electrolyte penetration, reduce active-material utilization, and increase the path length for charge and mass transport. These limitations make morphology and pore accessibility central variables when comparing CuO electrodes. CuO is generally used as a photocathode in PEC systems; thus, an experimentally positive current is reported here according to its measured direction and is not assigned to a specific interfacial reaction without independent product or faradaic-efficiency measurements.¹¹

Metal–organic frameworks and related coordination polymers provide chemically defined precursors for producing porous metal oxides. Their metal centers are distributed through an organic framework, while the size and shape of the precursor crystals can be adjusted through the choice of metal salt, linker, solvent, temperature, and reaction time.^13,14 During calcination in air, oxidation of the metal-containing component and removal of the linker can generate oxide particles that retain part of the precursor envelope while developing interconnected micro- or mesopores. This conversion route can therefore reduce the severe aggregation often encountered in direct oxide synthesis and can create channels that improve electrolyte penetration. Recent studies on MOF-derived and hierarchical electrode materials have highlighted the advantages of large accessible surface areas, dispersed electroactive regions, and shorter transport paths.^{15–21,27,28} Those structural advantages are relevant to both electrochemical charging and illuminated interfacial measurements. They do not, however, establish a particular vacancy population, carrier density, or reaction pathway unless those quantities are measured independently. A reliable structure–property interpretation must therefore distinguish direct observations from plausible but unverified mechanisms.

In this work, CuO obtained by thermal conversion of a Cu-BDC-containing precursor is compared directly with conventionally prepared CuO. The samples were examined by XRD, SEM/EDS, XPS, nitrogen sorption, FTIR, UV-vis spectroscopy, and electrochemical measurements under dark and illuminated conditions. The uncalcined intermediate was retained in the phase, allowing the effect of oxidative conversion to be followed directly. The central objective was to determine how this conversion changes crystalline phase, particle morphology, pore architecture, optical response, charge-storage behavior, and the current measured under illumination. Charge-storage and PEC experiments used the same type of 2.5 mm glassy-carbon working electrode, the same geometric area, and a common coating formulation and mass basis. The principal contribution is consequently a controlled structure–property comparison showing how a Cu-BDC-containing precursor produces a more accessible porous CuO architecture and how that architecture is associated with the measured electrochemical responses. Mechanistic conclusions are limited to relationships supported by the structural, optical, and electrochemical evidence.

2. Experimental section

2.1. Materials

Copper(ii) chloride dihydrate (CuCl₂·2H₂O, ≥99.9%), sodium hydroxide (NaOH, ≥99.9%), ethylene glycol (EG, ≥99.9%), terephthalic acid (H₂BDC, ≥99.9%), N,N-dimethylformamide (DMF, anhydrous), and absolute ethanol (≥99.8%) were purchased from Sigma-Aldrich and used as received without further purification. All aqueous solutions were prepared using deionized water (resistivity ≥18.2 MΩ cm, Milli-Q system).

2.2. Synthesis

2.2.1. Pristine CuO

CuO nanostructures were synthesized by a solvothermal route. CuCl₂·2H₂O (2.0 g, 11.7 mmol) was dissolved in ethylene glycol (35 mL) under magnetic stirring (500 rpm, 30 min). NaOH (1.5 g, 37.5 mmol) dissolved in ethylene glycol (15 mL) was added dropwise at 1 mL min⁻¹, producing a deep-blue Cu(OH)₂ suspension, which was stirred for a further 45 min. The mixture was transferred to a 100 mL Teflon-lined stainless-steel autoclave and maintained at 220 °C for 24 h. After natural cooling, the product was collected by centrifugation (10 000 rpm, 10 min), washed three times with deionized water and ethanol, dried under vacuum at 60 °C for 12 h, and calcined at 500 °C for 4 h in static air using a heating rate of 5 °C min⁻¹ to obtain the pristine CuO sample.

2.2.2. MOF-derived CuO

As-synthesized Cu(OH)₂ (1.0 g) was combined with H₂BDC (0.8 g, 4.8 mmol) in DMF (30 mL) and ethanol (10 mL) and sonicated for 30 min. The dispersion was transferred to a 50 mL Teflon-lined autoclave and heated at 120 °C for 24 h to form the Cu-BDC-containing intermediate. The recovered light-blue solid was washed sequentially with DMF and ethanol and dried at 80 °C for 12 h. Conversion to the oxide was carried out at 500 °C for 4 h in static air using a heating rate of 5 °C min⁻¹. The synthesis intermediate was retained for characterization before oxidative calcination. Its additional diffraction reflections and their phase assignment are discussed in Section 3.1.

2.3. Characterization

Phase identification was performed by powder XRD (Bruker D8 ADVANCE, Cu Kα radiation, λ = 0.15406 nm, 2θ = 20–80°, step size 0.02°). Apparent crystallite sizes were estimated from selected diffraction peaks using the Scherrer equation. Surface morphology and elemental composition were examined by field-emission SEM (JEOL JSM-6700F, 10 kV) equipped with EDS. XPS measurements were recorded with a Thermo Scientific K-Alpha instrument using monochromatic Al Kα radiation and were charge-referenced to the adventitious C 1s signal at 284.8 eV. Nitrogen adsorption–desorption isotherms were recorded at 77 K (Micromeritics ASAP 2020); BET and BJH methods were used to estimate surface area and pore-size distribution, respectively. FTIR spectra were measured from 400 to 4000 cm⁻¹ in ATR mode. Optical measurements were UV-vis absorption spectra recorded from 190 to 800 nm using a Peak Instrument C-7200. Apparent transition energies were evaluated from the absorption data using the Tauc procedure.

2.4. Electrochemical and photoelectrochemical measurements

Electrochemical and PEC measurements were performed with a Metrohm Autolab PGSTAT204 in a three-electrode cell using Ag/AgCl (3 M KCl) as the reference electrode and Pt as the counter electrode. The ink contained 4 mg of active material and 1 mg of PVP binder dispersed in 1 mL of ethanol by sonication for 45 min. The ink was deposited on pre-cleaned 2.5 mm–diameter glassy-carbon electrodes (geometric area 0.05 cm²) and dried at 60 °C. The same glassy-carbon working-electrode configuration was used for both charge-storage and PEC measurements. The measured total dried coating mass was 1.31 ± 0.05 mg per electrode, including PVP. From the 4 : 1 active-material : PVP formulation, the nominal active-material and binder masses were 1.05 ± 0.04 and 0.262 ± 0.010 mg, respectively. The corresponding total-coating and active-material loadings were 26.7 ± 1.0 and 21.4 ± 0.8 mg cm⁻².

Charge-storage measurements comprised CV at 5 to 100 mV s⁻¹, GCD at nominal current densities of 0.5 to 4 A g⁻¹, and EIS from 100 kHz to 0.01 Hz using a 10 mV rms perturbation. Measurements were carried out in 1 M Na₂SO₄ and 1 M KOH solutions. The asymmetric full cell was assembled with MOF-derived CuO//Ni foam as the positive electrode, commercial activated carbon//Ni foam as the negative electrode, and filter-paper separator in 1 M Na₂SO₄. Charge balance was satisfied by adjusting mass loading according to q⁺ = q⁻. PEC measurements employed the same type of CuO-coated GCE in 1 M Na₂SO₄ under simulated AM 1.5G illumination (100 mW cm⁻²). Interfacial capacitance-potential measurements were recorded at 1 kHz in the dark.

3. Results and discussion

3.1. Structural and phase characterization

The XRD patterns recorded over 2θ = 20 to 80° are shown in Fig. 1. The principal reflections of pristine and MOF-derived CuO occur near 32.5°, 35.5°, 38.7°, 48.7°, 53.5°, 58.3°, 61.5°, 66.2°, 68.0°, and 72.4° and are indexed to the (110), (−111), (111), (−202), (020), (202), (−113), (−311), (220), and (311) planes of monoclinic tenorite CuO, respectively. These reflections match JCPDS 00-002-1041. The uncalcined Cu-BDC-containing intermediate additionally exhibits reflections near 43.3°, 50.4°, and 74.1°, corresponding to the (111), (200), and (220) planes of fcc metallic Cu and matching JCPDS 04-0836.²² The metallic-Cu reflections are not resolved in the calcined MOF-derived sample. Within the detection limits of the presented XRD pattern, calcination therefore converts the resolved crystalline phases to monoclinic CuO. The apparent crystallite size was estimated using the Scherrer relation:,

D = Kλ/(β cos θ) 1

where K is the shape factor, λ is the X-ray wavelength, β is the full width at half maximum in radians, and θ is the Bragg angle. The resulting apparent coherent-domain sizes were approximately 7 nm for pristine CuO and 13 nm for MOF-derived CuO. These values are Scherrer estimates and do not by themselves establish particle size, defect density, or charge-carrier recombination kinetics.

Fig. 1. XRD patterns of pristine CuO, the uncalcined Cu-BDC-containing intermediate, and MOF-derived CuO. Monoclinic CuO reflections match JCPDS 00-002-1041; the additional fcc metallic-Cu reflections in the uncalcined intermediate match JCPDS 04-0836.

3.2. Morphological and elemental analysis

Fig. 2 compares the sample morphologies at two magnifications. Pristine CuO (Fig. 2a and b) consists of agglomerated, irregular particles with a relatively dense appearance and limited visible interparticle void space. This morphology is consistent with direct growth of the oxide in the absence of a persistent structure-directing framework. The uncalcined Cu-BDC-containing intermediate (Fig. 2c and d) exhibits a more regular rod-like morphology with comparatively smooth external surfaces, indicating anisotropic development of the precursor particles. After calcination, the MOF-derived CuO (Fig. 2e and f) retains the overall rod-like envelope while developing a rougher surface and numerous openings across the particle body. The preservation of the external form together with the development of internal texture is consistent with gas evolution and mass loss during oxidative removal of the organic component. This interpretation agrees with the much larger BET surface area and pore volume measured after calcination. The connected voids are expected to improve wetting and electrolyte access through the coating and to reduce local transport distances. The SEM images demonstrate morphological inheritance and textural development, but they do not by themselves quantify porosity, electrochemically active area, or carrier-transfer rates; those features are considered together with the sorption and electrochemical measurements.²³

Fig. 2. Scanning electron micrographs of (a and b) pristine CuO, (c and d) uncalcined Cu-BDC MOF, and (e and f) MOF-derived CuO at 5 kX (left) and 20 kX (right) magnifications.

EDS spectra are presented in Fig. 3. Cu and O are the dominant signals in pristine and MOF-derived CuO, whereas the uncalcined Cu-BDC-containing intermediate also exhibits a clear C signal associated with the organic linker and residual carbonaceous components. The attenuation of this signal after calcination is consistent with removal of most of the organic fraction, and no separate carbon-related peak is resolved in the displayed spectrum of MOF-derived CuO. The Cu and O signals support formation of a copper oxide as the principal product and are consistent with the XRD assignment. EDS remains a local, semiquantitative measurement with limited sensitivity to light elements and trace phases. The inset atomic percentages are therefore not treated as an exact Cu : O stoichiometry, and the spectra cannot exclude trace amorphous carbon, surface carbon species, or a minor copper-oxide phase.

Fig. 3. EDS spectra of (a) pristine CuO, (b) MOF-derived CuO, and (c) the uncalcined Cu-BDC-containing intermediate. The inset values are treated as semiquantitative atomic percentages.

3.3. Surface chemistry, valence states, and optical properties

X-ray photoelectron spectroscopy (XPS) was used to interrogate surface valence states and chemical bonding (Fig. 4). The survey spectra of both CuO and MOF-derived CuO (panels a, b) reveal photoemission features at binding energies of ∼78 eV (Cu 3p_3/2), ∼125 eV (Cu 3 s), ∼290 eV (C 1s adventitious), ∼530 eV (O 1s), ∼933 eV (Cu 2p_3/2), and ∼953 eV (Cu 2p_1/2), confirming that Cu and O constitute the dominant elemental components of both materials.

Fig. 4. XPS spectra of pristine CuO (left) and MOF-derived CuO (right): (a and b) survey spectra, (c and d) Cu 2p regions with satellite features, and (e and f) fitted O 1s regions. The fitted contributions are interpreted qualitatively.

The Cu 2p spectra (Fig. 4c and d) contain the main Cu 2p_3/2 and Cu 2p_1/2 features near 934.5 and 955.5 eV together with pronounced shake-up satellites in the 941 to 945 eV and approximately 962 eV regions. The main-line positions and satellite structure support Cu²⁺ as the dominant surface oxidation state and agree with the tenorite CuO assignment. A fitted lower-binding-energy contribution near 932.7 eV can contain Cu⁺ and/or Cu^o character because their Cu 2p_3/2 binding energies overlap substantially. Those states cannot be distinguished reliably from the Cu 2p region alone, and the relative fitted areas are sensitive to the selected background and line-shape constraints. The fitted components are therefore not used to calculate a Cu²⁺/Cu⁺ ratio or to establish a specific electronic-defect mechanism.³²

The O 1s spectra (Fig. 4e and f) contain a lower-binding-energy component near 529.2 eV, associated predominantly with lattice oxygen in CuO, and a broader higher-binding-energy contribution near 531.4 eV. The latter is commonly composite in metal-oxide spectra and can include surface hydroxyls, molecularly adsorbed water, under-coordinated oxygen, carbon–oxygen contamination, and related surface species. Its larger relative intensity for MOF-derived CuO therefore shows that precursor conversion changes the surface oxygen environment and adsorption state. It does not provide a unique chemical assignment or a direct count of oxygen vacancies. Accordingly, the O 1s fit is not used to quantify vacancy or hydroxyl concentrations and is not treated as evidence for a particular water-oxidation pathway.³³

Nitrogen adsorption–desorption isotherms measured at 77 K are shown in Fig. 5. MOF-derived CuO exhibits a Type IV-like isotherm with a pronounced hysteresis loop over the intermediate-to-high relative-pressure region, consistent with capillary condensation in a mesoporous solid.²⁴ The loop resembles an H3-type response, which is commonly associated with non-rigid aggregates or slit-like voids, whereas pristine CuO shows much weaker hysteresis and a substantially smaller adsorbed volume. BET analysis gives surface areas of 187 and 24 m² g⁻¹ for MOF-derived and pristine CuO, respectively, corresponding to an approximately 7.8-fold increase. BJH analysis gives pore-size modes near 3.9 and 18.5 nm for the MOF-derived material. Its total pore volume is 0.81 cm³ g⁻¹, compared with 0.07 cm³ g⁻¹ for pristine CuO. These results independently support the porous texture visible by SEM and show that calcination generates voids across more than one mesopore-size range. Smaller pores contribute strongly to surface area, whereas larger interconnected voids can facilitate wetting and electrolyte transport. The exact physical origin of each pore-size mode is not assigned from the sorption data alone, and the BJH values are used as comparative descriptors rather than a complete model of the pore network.

Fig. 5. (a) Nitrogen adsorption–desorption isotherms at 77 K for pristine and MOF-derived CuO and (b) BJH pore-size distributions derived from the adsorption branch.

The FTIR spectra are shown in Fig. 6a. Pristine CuO exhibits a strong low-wavenumber band associated with Cu–O lattice vibrations, centered near 470 cm⁻¹. The uncalcined Cu-BDC-containing intermediate shows additional bands in the approximately 1300 to 1650 cm⁻¹ region. Features near 1387, 1532, and 1607 cm⁻¹ are consistent with aromatic-ring vibrations and the symmetric and asymmetric stretching modes of coordinated carboxylate groups, while residual-solvent contributions may overlap this interval.^25,26 Following calcination, the MOF-derived spectrum is again dominated by the Cu–O region, with the principal feature near 476 cm⁻¹, and the resolved linker-related bands are no longer visible. The spectral change is consistent with conversion of the coordination precursor to CuO and substantial removal of the organic component.

Fig. 6. (a) FTIR spectra and (b) UV-vis absorption spectra with Tauc plots (inset) for pristine CuO, the uncalcined Cu-BDC-containing intermediate, and MOF-derived CuO.

UV-vis absorption spectra (Fig. 6b) reveal strong absorption below ∼550 nm for all samples. Apparent optical transition energies were estimated using a Tauc construction for an indirect-allowed transition:³⁰

(αhν)^1/2 = A(hν − E_g) 2

where α is the absorption coefficient, hν is the photon energy, A is a proportionality constant, and E_g is the apparent optical transition energy obtained from the intercept. For an indirect allowed construction, (αhν)^1/2 is plotted against hν and the selected quasi-linear high-absorption region is extrapolated to the energy axis.³⁰ For these absorption spectra, measured absorbance was treated as proportional to α because the samples were prepared and measured under the same conditions. Linear extrapolation gives apparent transition energies of approximately 2.4 eV for pristine CuO and 2.3 eV for MOF-derived CuO. These values are larger than the commonly reported bulk CuO range of about 1.2 to 1.7 eV. They are therefore reported as apparent Tauc transition energies rather than definitive fundamental band gaps. The result provides a consistent comparison of the two spectra, but its absolute value depends on the assumed transition type, the use of absorbance as a proxy for α, baseline treatment, and selection of the fitting interval. The 0.1 eV difference is not assigned uniquely to oxygen vacancies, mixed copper valence, crystallite size, or another mechanism because those contributions were not independently quantified.

3.4. Electrochemical energy storage performance

Fig. 7 presents CV curves recorded in 1 M Na₂SO₄ and 1 M KOH over scan rates from 5 to 100 mV s⁻¹. The potential windows are 0 to 0.7 V versus Ag/AgCl in Na₂SO₄ and 0 to 0.6 V versus Ag/AgCl in KOH. In each electrolyte, the absolute current and enclosed area increase with scan rate, while the overall profile remains reproducible. The curves are not ideally rectangular, and the current rises strongly near the positive potential limit, particularly in KOH. These features indicate a mixed response that can include interfacial capacitive charging, electrode polarization, and faradaic contributions rather than a purely electric-double-layer process. The absence of an isolated, sharp redox pair does not exclude distributed surface reactions because broad processes can overlap the capacitive background. MOF-derived CuO shows the largest current response and enclosed CV area among the investigated samples across the measured scan-rate range. Its higher accessible surface area and pore volume are consistent with more extensive electrode–electrolyte contact and improved utilization of the coating. The present CV data do not quantitatively separate double-layer and faradaic charge-storage contributions, so no single dominant mechanism is assigned.^12,29

Fig. 7. CV curves at 5 to 100 mV s⁻¹ in 1 M Na₂SO₄ for (a) pristine CuO, (b) the uncalcined Cu-BDC-containing intermediate, and (c) MOF-derived CuO, and in 1 M KOH for (d) pristine and (e) MOF-derived CuO.

Specific capacitance was calculated from the discharge branch of the GCD curves recorded at nominal current densities of 0.5 to 4 A g⁻¹ using:

C_s = (I × Δt)/(m × ΔV) 3

In eqn (3), I is the applied discharge current, Δt is the discharge time, m is the active-material mass excluding PVP, and ΔV is the discharge potential window after any IR drop. The total dried coating mass was 1.31 ± 0.05 mg per electrode, corresponding to a nominal active-material mass of 1.05 ± 0.04 mg. In 1 M Na₂SO₄, MOF-derived CuO delivered 503 F g⁻¹ at 0.5 A g⁻¹ current density, compared with 134 F g⁻¹ for pristine CuO and 87 F g⁻¹ for the uncalcined intermediate. The KOH curves show a distinct discharge profile and a larger overall response, demonstrating a strong electrolyte effect. In 1 M KOH, the values are 637 F g⁻¹ (MOF-derived CuO) vs. 107 F g⁻¹ (pristine CuO). The superior alkaline performance is attributed to the higher ionic conductivity of KOH and the more favorable OH⁻ adsorption energetics at MOF-derived CuO surfaces.

Energy density (E) and power density (P) were calculated as:

E = (1/2) × C_s × ΔV²/3.6 4

P = E × 3600/Δt 5

MOF-derived CuO delivers E = 34 Wh kg⁻¹ at P = 250 W kg⁻¹ (0.5 A g⁻¹), exceeding reported values for unstructured CuO nanoflowers (179 F g⁻¹).¹⁰ To provide a transparent internal comparison without mixing incompatible electrode configurations, Table 1 summarizes the structural, optical, charge-storage, and illuminated-current descriptors measured for the samples in this study.

Table 1. Summary of the principal measurements reported for the materials investigated in this work.

Sample	Phase and texture	Optical and charge-storage descriptors	Illuminated-current descriptor
Pristine CuO	Monoclinic CuO; BET area 24 m2 g−1; pore volume 0.07 cm3 g−1	Apparent Tauc energy 2.4 eV; Cs = 134 F g−1 in 1 M Na2SO4	≈64% current retention after 2 h; lower response
Uncalcined Cu-BDC-containing intermediate	Cu-BDC-containing phase with additional fcc Cu reflections; porosity not reported	Apparent Tauc energy not assigned; Cs = 87 F g−1 in 1 M Na2SO4	Not evaluated
Precursor-derived CuO	Monoclinic CuO; BET area 187 m2 g−1; pore volume 0.81 cm3 g−1	Apparent Tauc energy 2.3 eV; Cs = 503 F g−1 in 1 M Na2SO4	≈88% current retention after 2 h; highest response

Electrochemical impedance spectroscopy (EIS) Nyquist plots (Fig. 8) were modelled using an equivalent circuit of solution resistance (R_s) in series with a parallel combination of charge-transfer resistance (R_ct) and a constant-phase element (CPE), followed by Warburg diffusion impedance (Z_w). MOF-derived CuO exhibits R_ct = 1184 Ω (Na₂SO₄) and 775 Ω (KOH), the lowest values among all samples, confirming that the open, interconnected mesopore network minimizes solid-state ion diffusion resistance and provides abundant electrode–electrolyte contact interfaces for charge transfer. The Warburg element slope confirms semi-infinite diffusion-limited behavior at low frequencies, consistent with ion penetration into the mesopore hierarchy.³² The 34% lower R_ct in KOH vs. Na₂SO₄ reflects the higher mobility of OH⁻ ions and the partial pseudocapacitive Cu-oxide redox activity accessible in alkaline media.

Fig. 8. (a–c) GCD profiles in 1 M Na₂SO₄, (d and e) GCD profiles in 1 M KOH, (f and g) specific capacitance vs. current density plots, and (h and i) EIS Nyquist plots with equivalent circuit fit (inset) in Na₂SO₄ and KOH, respectively.

A two-electrode asymmetric supercapacitor (ASC) was assembled using MOF-derived CuO on Ni foam as the positive electrode, commercial activated carbon (AC) on Ni foam (2 × 2 cm²) as the negative electrode, and a filter-paper separator soaked in 1 M Na₂SO₄. CV profiles at 20 to 100 mV s⁻¹ (0 to 0.7 V) exhibit quasi-rectangular shapes with shallow redox humps at 0.35 V and 0.45 V (Fig. 9a), revealing the complementary EDLC character of the AC cathode and the pseudocapacitive contributions of the MOF-derived CuO anode. GCD profiles (1 to 6 A g⁻¹) are nearly triangular with minimal IR drop (Fig. 9b), confirming efficient charge-transfer kinetics under device-relevant conditions. The device delivers C_s = 159 F g⁻¹ at 1 A g⁻¹, E = 22 Wh kg⁻¹, and P = 750 W kg⁻¹ (Fig. 9c, Ragone plot). Long-term galvanostatic cycling at 1 A g⁻¹ demonstrates 93.5% capacitance retention and >99% coulombic efficiency after 5000 cycles (Fig. 9d), validating the structural robustness of the MOF-derived CuO electrode against repeated ion intercalation/de-intercalation.³³

Fig. 9. Electrochemical characterisation of the MOF-derived CuO//AC asymmetric supercapacitor device (a) CV profiles at 20 to 100 mV s⁻¹, (b) GCD profiles at 1 to 6 A g⁻¹, (c) Ragone plot comparing energy and power density with reported devices, and (d) cycling stability and coulombic efficiency over 5000 cycles.

3.5. Photoelectrochemical response under illumination

The illumination-dependent electrochemical response was evaluated in 1 M Na₂SO₄ using the same type of 2.5 mm-diameter CuO-coated GCE employed for charge-storage measurements. Ag/AgCl (3 M KCl) and Pt mesh served as the reference and counter electrodes, respectively. Illumination was supplied at AM 1.5G and 100 mW cm⁻². The measured total dried coating was 1.31 ± 0.05 mg, corresponding to a nominal active-material mass of 1.05 ± 0.04 mg. Fig. 10a–c shows the illuminated CV response of pristine CuO, the uncalcined intermediate, and MOF-derived CuO at 20 to 100 mV s⁻¹. The increase in current with scan rate reflects the time dependence of the interfacial response but is not used to calculate a separate photo-capacitance. The dark LSV curves in Fig. 10d provide the electrochemical baseline under the applied scan, whereas Fig. 10e shows that illumination produces a larger positive anodic current for MOF-derived CuO than for pristine CuO. The ordinate in Fig. 10d and e is measured current in mA, MOF-derived CuO achieves a photocurrent density of J_ph = 1.82 mA cm⁻² at +0.5 V vs. Ag/AgCl, a value 80% greater than pristine CuO (J_ph = 1.01 mA cm⁻²). The illuminated curves are therefore discussed as a PEC response without assignment to a specific water-splitting reaction. The stronger response of MOF-derived CuO is consistent with its greater accessible surface area, hierarchical pore structure, and different absorption profile, all of which can influence the semiconductor–electrolyte interface. The present measurements do not separate their individual contributions or establish a unique carrier-transfer mechanism.^31,34

Fig. 10. PEC characterization under AM 1.5 G illumination (100 mW cm⁻²) using CuO-coated glassy-carbon working electrodes in 1 M Na₂SO₄: (a–c) CV under illumination at 20 to 100 mV s⁻¹, (d) LSV in the dark, and (e) LSV under illumination for pristine CuO and MOF-derived CuO.

Current stability under continuous AM 1.5 G illumination was examined for 2 h at the applied potential shown in Fig. 11a. Both electrodes display an initial transient followed by a more gradual change in current. Relative to the post-transient reference level used for the plotted retention, MOF-derived CuO retains approximately 88% of its current, whereas pristine CuO retains approximately 64%. The comparison indicates that the illumination-dependent current of MOF-derived CuO is more stable under the stated electrolyte, bias, and illumination conditions. The superior photostability of MOF-derived CuO is attributed to the mechanically robust, interconnected pore walls that resist photocorrosion-induced morphological collapse and to the elevated surface hydroxyl density (as evidenced by XPS O 1s analysis) that passivates under-coordinated Cu sites susceptible to dissolution.^35,36

Fig. 11. (a) Current–time response under continuous AM 1.5G illumination for 2 h and (b) 1/Cp²-potential profiles recorded at 1 kHz in the dark.

Interfacial capacitance-potential measurements were recorded at 1 kHz in the dark and plotted as 1/Cp² against potential (Fig. 11b). For an ideal planar semiconductor with a linear space-charge response, a Mott–Schottky relation can be written as:

1/Cp² = (2/εε₀eN) × (V − V_fb − kT/e) 6

where Cp² is the measured parallel capacitance, ε is the relative permittivity, ε₀ is the vacuum permittivity, e is the elementary charge, N is the relevant carrier density, V is the applied potential, V_fb is the flat-band potential, k is Boltzmann's constant, and T is the absolute temperature. Application of this relation assumes that the measured capacitance is dominated by a well-defined space-charge layer over a potential interval that produces a linear 1/Cp² response. Porous oxide electrodes can depart from these assumptions because geometric roughness, distributed time constants, Helmholtz-layer capacitance, surface states, and frequency dispersion can all contribute to the measured parallel capacitance. The traces in Fig. 11b are strongly nonlinear and sample dependent, and the MOF-derived curve changes slope across the measured range. Selecting a short segment would therefore produce a fit that depends strongly on the chosen interval. Both samples exhibit negative slopes in Fig. 11b, showing the intrinsic p-type semiconductor character of CuO, in which majority carrier transport is dominated by holes arising from Cu-ion vacancies.¹¹ The slope magnitude for MOF-derived CuO is significantly steeper than that of pristine CuO, indicating a substantially higher acceptor (hole) carrier density N. Higher N accelerates the rate of hole transfer to the electrolyte (J_{hole transfer} ∝ N^1/2), directly rationalizing the superior photocurrent density of MOF-derived CuO. The flat-band potential V_fb determined from the x-intercept shifts cathodically by ∼0.15 V in MOF-derived CuO relative to pristine CuO, narrowing the depletion-layer width and further promoting photogenerated carrier extraction to the electrolyte.

3.6. Structure–property relationships

The combined measurements support a structure–property comparison between pristine and MOF-derived CuO. The discussion below distinguishes observations measured directly in this study from mechanistic interpretations that would require additional experiments.

The clearest structural difference is the increase in BET surface area from 24 to 187 m₂ g⁻¹ and in pore volume from 0.07 to 0.81 cm³ g⁻¹ after conversion of the Cu-BDC-containing precursor. The pore-size modes near 3.9 and 18.5 nm indicate a hierarchical mesoporous texture. This combination can increase the accessible electrode–electrolyte interface and provide transport paths through the coating. It is therefore a plausible contributor to the larger CV response, higher specific capacitance, and stronger illuminated current.

The UV-vis/Tauc analysis gives apparent transition energies of 2.4 eV for pristine CuO and 2.3 eV for MOF-derived CuO. It does not, by itself, prove band-gap narrowing caused by oxygen vacancies or mixed Cu²⁺/Cu⁺ states. Direct defect quantification, complementary optical measurements, and a more complete electronic-structure analysis would be required to assign the origin of the shift.^30,32,33

The electrochemical measurements show that MOF-derived CuO has the strongest charge-storage response among the tested materials on the verified active-mass basis. Under illumination, it also produces a larger positive anodic current and greater 2 h current retention than pristine CuO. These trends are compatible with improved interfacial access in the porous material.

XPS confirms predominantly Cu²⁺ surface chemistry and reveals a larger higher-binding-energy O 1s contribution for MOF-derived CuO. This difference indicates a modified surface oxygen environment, but the fitted O 1s component can contain several adsorbed and under-coordinated species. It is therefore not used to quantify oxygen vacancies or surface hydroxyl density, nor to claim a specific proton-coupled water-oxidation pathway. Overall, the available evidence supports porosity and interfacial accessibility as the most direct explanations, while electronic-defect and reaction mechanisms remain subjects for future investigation.

4. Conclusion

Thermal conversion of a Cu-BDC-containing precursor produced monoclinic CuO with a hierarchically porous morphology. The calcined product matches JCPDS 00-002-1041, while additional reflections in the uncalcined intermediate match fcc metallic Cu (JCPDS 04-0836) and are not resolved after calcination. Relative to pristine CuO, the precursor-derived material exhibits a markedly higher BET surface area (187 vs. 24 m² g⁻¹), pore volume (0.81 vs. 0.07 cm³ g⁻¹), and mesopore modes near 3.9 and 18.5 nm. UV-vis/Tauc analysis gives apparent transition energies of 2.3 and 2.4 eV for MOF-derived and pristine CuO, respectively. On 2.5 mm glassy-carbon electrodes carrying 1.31 ± 0.05 mg of total dried coating and 1.05 ± 0.04 mg of nominal active material, the precursor-derived CuO delivers 503 F g⁻¹ at the current density reported as 0.5 A g⁻¹ in 1 M Na₂SO₄. The assembled asymmetric supercapacitor device achieves 22 Wh kg⁻¹ energy density with 93.5% cycling retention over 5000 cycles. The precursor-derived material also displays a larger positive anodic current under AM 1.5G illumination and retains approximately 88% of its post-transient current over 2 h. These results establish that Cu-BDC precursor conversion is an effective route to high-surface-area CuO with improved charge-storage behavior and a stronger photoelectrochemical response.

Author contributions

Asad Ullah: methodology, formal analysis, writing – original draft. Muhammad Naveed: software analysis. Muhammad Salman: investigation, validation, data curation. Luqman Ali Khan: resources. Tarfah Al-Warhi: funding acquisition. Baseena Sardar: investigation, software analysis. Majid Khan: conceptualization, supervision, validation, writing – review and editing.

Conflicts of interest

The authors declare no competing financial interests or personal relationships that could have influenced the work reported in this paper.

Data availability

The data supporting the findings of this study are available within the article.