Hierarchical CuO Nanorods via Cyclic Voltammetry Treatment: Freestanding Electrodes for Selective CO₂-to-Formate Conversion

Abstract

Electrochemical CO₂ reduction reaction (CO₂RR) represents a promising pathway for carbon neutralization. Here, we report hierarchical CuO nanorod arrays synthesized via cyclic voltammetry (CV) treatment as freestanding electrodes for selective CO₂RR. The CV activation process generates ultrathin nanosheets on CuO nanorods, creating abundant interfaces that facilitate formate production. Optimized CV-2000-CuO achieves 42% Faradaic efficiency (FE) for formate at −1.4 V vs. RHE while suppressing hydrogen evolution reaction (HER). Comprehensive characterization reveals that CV treatment promotes partial surface reduction to metallic Cu and generates high-density grain boundaries during CO₂RR operation. These structural features enhance CO₂RR activity and stability compared to pristine CuO (P-CuO). This work demonstrates a novel electrode engineering strategy combining freestanding architecture with electrochemical activation for efficient CO₂-to-formate conversion.

1. Introduction

Electrochemical CO₂ reduction reaction (CO₂RR) represents a promising pathway for converting greenhouse gases into value-added chemicals and fuels, thereby addressing both environmental sustainability and energy storage challenges [1,2,3,4,5,6,7,8]. Copper (Cu)-based catalysts are uniquely capable of producing multi-carbon (C₂₊) products such as ethylene and ethanol, yet their practical deployment is hindered by challenges in selectivity control, high overpotentials, and structural instability under operating conditions [3,9,10,11,12,13,14,15,16,17]. The dynamic reconstruction of Cu surfaces during CO₂RR further complicates the identification of active sites [10,18,19,20].

Recent advances in copper catalysts emphasize morphological and compositional engineering to enhance C₂₊ yields [21,22,23,24,25,26,27]. For instance, crystalline/amorphous dual-phase Cu structures promote *CO coverage via chloride adsorption in neutral media, achieving 81% Faradaic efficiency (FE) for C₂₊ products [28]. Similarly, carbonate-shell-regulated CuO reconstruction enriches grain boundaries, boosting ethylene selectivity to 82.8% [29], while molecular modifiers like tannic acid optimize Cu^δ+/Cu⁰ interfaces for C₂H₄ formation [30]. Among these factors, grain boundaries play a crucial role in enhancing catalytic performance. A grain boundary is the interface between adjacent grains in a polycrystalline material and can be classified into low-angle grain boundaries (<15°) and high-angle grain boundaries (>15°) according to the crystallographic orientation difference. Owing to atomic undercoordination and lattice strain, grain boundary regions possess high surface energy, which facilitates reactant adsorption and intermediate activation. The unsaturated coordination atoms and strain fields at grain boundaries can synergistically modulate the adsorption energy of reaction intermediates, thereby significantly improving the selectivity and yield of C₂⁺ products [31]. Additionally, Xia et al. systematically reviewed the latest advances, reaction mechanisms, and challenges of copper-based alloy catalysts—including conventional alloys, high-entropy alloys, and single-atom alloys in the CO₂RR. By doping with metals such as Ag, Au, and Pd, they effectively modulated the electronic structure of copper and the adsorption energies of reaction intermediates, successfully breaking the traditional linear scaling relationships and significantly enhancing catalytic selectivity and activity [32]. Despite these successes, formate production on Cu-based catalysts is often marginalized, with only sparse reports of Cu-based systems exceeding 50% FE [33,34,35,36,37,38,39,40,41,42,43,44]—typically requiring Sn or S heteroatoms [45,46] or molecular additives like cetyltrimethylammonium bromide (CTAB) [43]. For instance, Li et al. [45] reported a novel Cu₂O/CuS catalyst that could both enhance the activity and stability of formate production, which outperforms most Cu, Cu₂O or CuS catalysts. The sulfide phase proved to have an important role in tuning the selectivity of CO₂ reduction. On the other hand, given previous density functional theory calculations have shown that the key intermediate (HCOO*) for formate evolution usually binds strongly to the Cu catalyst surface and the desorption of HCOO* has been proved the rate-limiting step of CO₂RR to formate. The CTAB modification of Cu surface therefore has also been demonstrated to facilitate the desorption of HCOO* and greatly improved formate formation [43]. It should be noted that Cu-based catalysts without any heteroatoms or modifiers usually show much lower selectivity and activity of formate formation. For example, the leaf-like CuO nanosheets with only 20% FE of formate evolution [42], CuO-derived hierarchical nanostructures with a maximum formate FE of 38%, and the Cu(OH)₂-derived nanowires exhibited only 23% FE [35], etc. Therefore, there is urgent need to investigate the true activity site of formate formation on Cu-based catalysts and help to design novel structures for CO₂-to-formate. Moreover, conventional powder-based catalysts suffer from binder-induced impedance and inhomogeneous active site distribution, limiting reproducibility and mass transport. Freestanding electrodes, though promising [47,48,49,50,51], have not yet leveraged in situ cyclic voltammetry (CV) treatment for surface engineering. This gap underscores the need for integrated approaches that combine tunable reconstruction methods with freestanding architectures to unlock formate-selective pathways.

In this work, we reported CV-activated CuO (CV-CuO) nanorod arrays as a novel freestanding catalyst for efficient CO₂ conversion. The CV process generates hierarchical CuO nanosheets on pristine CuO (P-CuO) nanorods, creating abundant interfaces that favor CO₂RR, with the CV-2000-CuO sample shows the highest FE of formate formation (42%) and most suppressed hydrogen evolution reaction (HER). The X-ray diffraction (XRD) results showed that both the high valence state of Cu in the bulk phase of both P-CuO and CV-CuO are nearly reduced to metallic Cu during activated process. However, the ex situ X-ray photoelectron spectroscopy (XPS) results suggested that both of the two samples are partially oxidized on the surface, while the CV-CuO sample exhibited more metallic Cu compared to P-CuO. The transmission electron microscopy (TEM) and Scanning electron microscopy (SEM) with an electron backscatter diffraction (EBSD) model further demonstrated the CV-CuO generated more grain boundaries and interface during CO₂RR process, which might be the true active site for the improved performance of CO₂RR. This work not only demonstrates a freestanding electrode strategy but also expands the functional scope of CV engineering into CO₂RR catalysis.

2. Materials and Methods

2.1. Materials

Polycrystalline copper foil (poly-Cu, 25 µm thickness, ≥99.99%) was supplied by Alfa Aesar. potassium hydroxide (KOH, AR, ≥85%), ammonium persulfate ((NH₄)₂S₂O₈, AR, ≥98%), Sodium hydroxide (NaOH, AR, ≥98%), hydrochloric acid (HCl, 37 wt%), acetone (AR), and ethanol (AR) were procured from Adamas. Deionized water (DI water, 18.2 MΩ cm resistivity) was generated using a Milli-Q purification system (Nanjing Yuheng Instrument Equipment Co., Ltd., Nanjing, China). All chemicals were utilized as received without additional purification.

2.2. Synthesis of Hierarchical CuO Nanorod Arrays on Polycrystalline Cu Foil

The hierarchical CuO nanorod arrays were synthesized through a sequential process involving hydrothermal growth, thermal annealing, and cyclic voltammetry treatment, adapted from a reported method with a key modification in the substrate [46]. Polycrystalline copper foil (poly-Cu, thickness: ~100 µm) was employed as the substrate instead of copper foam.

Substrate Pretreatment: Rectangular pieces of poly-Cu foil (typically 2.0 × 4.0 cm²) underwent sequential ultrasonic cleaning in 1 M HCl, acetone, ethanol, and deionized (DI) water (5 min per solvent) to eliminate surface oxides and contaminants. The cleaned foils were then dried thoroughly in an oven at 60 °C.

In Situ Growth of Cu(OH)₂ Nanorod Arrays: An aqueous reaction solution was prepared by mixing equal volumes of 2.5 M NaOH (15.0 mL) and 0.125 M (NH₄)₂S₂O₈ (15.0 mL) under constant stirring until homogeneity was achieved. The pretreated poly-Cu foil was submerged in this solution at ambient temperature (~25 °C) for 20 min. During this step, the copper surface is oxidized by persulfate (S₂O₈²⁻) in the alkaline medium, leading to the formation of Cu²⁺ ions which subsequently precipitate as wafter-blue Cu(OH)₂ nanorods directly on the foil surface. After reaction, the foil was removed, rinsed copiously with DI water and ethanol, and dried at 60 °C in a vacuum oven.

Thermal Conversion to CuO Nanorod Arrays (P-CuO): The as-prepared in situ growth of Cu(OH)₂ nanorod arrays samples were converted to CuO via a controlled thermal annealing process under a nitrogen atmosphere. The transformation involved a two-stage heating profile in a tube furnace: first at 120 °C for 3 h to facilitate dehydration, followed by a ramp to 180 °C at 2 °C min⁻¹ and holding for an additional 3 h to promote crystallization. The initial ramp rate to 120 °C and cooling rates were maintained at 5 °C min⁻¹. This process yielded CuO nanorod arrays supported on poly-Cu foil, designated as P-CuO.

Cyclic Voltammetry (CV) Treatment: The P-CuO samples were subjected to electrochemical treatment using cyclic voltammetry (CV) in a three-electrode configuration with 6 M KOH electrolyte. The P-CuO served directly as the working electrode, while a platinum foil counter electrode and a Hg/HgO reference electrode were employed. CV scans were performed within a potential window of 0 V to +0.5 V (vs. Hg/HgO) at a scan rate of 50 mV s⁻¹ for specific cycle numbers (2000, 4000, or 8000 cycles). The potential range of 0 to +0.5 V (vs. Hg/HgO) was selected to induce mild redox-driven morphological reconstruction of CuO while preventing over-reduction or bulk phase collapse. This potential window enables the dynamic formation of ultrathin secondary nanosheets on the CuO nanorod surface, thereby enhancing surface area and catalytic active site exposure. The resulting samples are denoted as CV-X-CuO, where X represents the number of CV cycles applied (e.g., CV-2000-CuO, CV-4000-CuO, CV-8000-CuO). The pristine sample without CV treatment is referred to as P-CuO.

2.3. Characterization

Scanning electron microscopy (SEM, Hitachi SU8020, 10 kV acceleration voltage, Hitachi High-Tech Corporation, Tokyo, Japan) with an electron backscatter diffraction (EBSD) and transmission electron microscopy (TEM, JEOL JEM-2100F, 200 kV, Tokyo, Japan) coupled with selected-area electron diffraction (SAED) and energy-dispersive X-ray spectroscopy (EDS) mapping were employed to analyze nanorod morphology, crystallinity, and elemental distribution. X-ray diffraction (XRD) patterns were acquired on a Bruker D8 Advance diffractometer (Bruker Corporation, Karlsruhe, Germany) (Cu Kα radiation, λ = 1.5418 Å) over a 2θ range of 10°–80° to determine phase composition and crystal structure. X-ray photoelectron spectroscopy (XPS) was conducted on a Thermo Scientific K-Alpha+ spectrometer (Thermo Fisher Scientific Inc., Waltham, MA, USA) (monochromatic Al Kα source, 1486.6 eV). Binding energies were calibrated against the C 1s peak (284.8 eV). Peak deconvolution was performed using Avantage software (5.9931) with Shirley backgrounds.

2.4. Electrochemical Measurements

All CO₂ reduction reaction (CO₂RR) measurements were conducted in a custom two-compartment H-cell electrolyzer. The anodic and cathodic compartments were physically separated by a Nafion 117 proton exchange membrane, which was pre-treated by sequential boiling in H₂O₂ (3wt%), deionized water, H₂SO₄ (5wt%), and deionized water again (each step for 1 h) to ensure proton conductivity and minimize ion crossover. The catalyst materials under investigation, namely P-CuO and CV-X-CuO, were fabricated into electrodes with a defined geometric surface area of 1.0 cm × 1.0 cm (1.0 cm²). A key aspect of the electrode preparation was the direct use of these catalyst samples as the working electrodes. No polymeric binders (e.g., Nafion) or conductive additives (e.g., carbon black) were employed in the electrode fabrication. This binder-free approach eliminates potential interference from organic binders in product distribution analysis and ensures the measured activity and selectivity are solely attributable to the intrinsic properties of the P-CuO and CV-X-CuO catalysts. The electrodes were mechanically secured to ensure good electrical contact with the potentiostat lead. 0.1 M KHCO₃ aqueous solution (pH ≈ 8.3) was saturated with high-purity CO₂ (≥99.999%) for ≥30 min before all measurements to ensure saturation. Crucially, the CO₂ purge was maintained continuously throughout the entire duration of the electrochemical experiments to replenish consumed CO₂ and maintain a constant concentration near the electrode surface (The CO₂ flow rate was set to 30 sccm), preventing mass transfer limitations from dominating the observed performance.

Prior to evaluating the steady-state CO₂RR performance, each working electrode underwent an in situ electrochemical activation step. This activation was performed by applying a constant potential of −1.2 V vs. Ag/AgCl to the working electrode for 30 min. The activation process continued until the recorded current density reached a stable plateau, indicating a relatively steady surface state of the catalyst had been achieved. Electrochemical tests utilized a CHI 760E potentiostat. Linear sweep voltammetry (LSV) and chronoamperometry (CA) were conducted at room temperature. Potentials were measured against a leak-free Ag/AgCl reference electrode (saturated KCl solution) and converted to the reversible hydrogen electrode (RHE) scale using: E (vs. RHE) = E (vs. Ag/AgCl) + 0.210 V + 0.059 × pH. No ohmic drop was compensated. Gaseous products (H₂, CO, CH₄, C₂H₄) were quantified online during CA tests via gas chromatography (GC9790Plus, FULI INSTRUMENTS, Taizhou City, Zhejiang Province, China) equipped with thermal conductivity (TCD) and flame ionization (FID) detectors, using Ar carrier gas. Liquid products were analyzed by Nuclear Magnetic Resonance spectroscopy (NMR, AVANCE NEO 400 MHz, Bruker BioSpin GmbH, Fällanden, Switzerland). Faradaic efficiencies (FEs) were calculated as:

$${\mathit{FE}}_{\mathit{product}}(\%)=n_{\mathit{product}}\ast F\ast n_e/Q_{\mathit{total}}\ast 100\%$$

where n_product is moles of product, F is a constant of Faraday, n_e is electrons transferred per molecule, and Q_total is total charge passed.

3. Results and Discussion

Figure 1a schematically depicts the synthesis pathway for CV-CuO nanorod arrays. The process commences with the in situ growth of Cu(OH)₂ nanorod arrays directly onto a Cu foil substrate at ambient temperature. This initial formation is driven by the chemical treatment reaction:

Cu + 4NaOH + (NH₄)₂S₂O₈ → Cu(OH)₂ + 2Na₂SO₄ + 2NH₃ + 2H₂O

Under the alkaline conditions provided by NaOH, the Cu surface is oxidized by the persulfate anions (S₂O₈²⁻), first generating Cu²⁺ ions which rapidly precipitate as Cu(OH)₂ nanorods. Subsequently, these Cu(OH)₂ precursor arrays undergo a controlled thermal dehydration process (involving two distinct heating stages) to yield pristine CuO (P-CuO) nanorod arrays. The final step involves subjecting the P-CuO nanowires to cyclic voltammetry (CV) treatment. This electrochemical treatment, conducted in 6 M KOH electrolyte over a 0–0.5 V potential range, transforms the material into the target CV-CuO nanorod arrays.

The morphological evolution throughout the synthesis was characterized using SEM. Figure 1b,c and Figure S1 reveal that the as-grown Cu(OH)₂ nanorods possess a well-defined array structure, with individual rods averaging approximately 200 nm in diameter and ~10 μm in length. Following thermal calcination to form P-CuO (Figure 1d,e), the fundamental nanorod array architecture is largely retained, although some degree of curling is observed, likely due to dehydration-induced stress. More interestingly, a significant morphological transformation occurs upon CV treatment. As exemplified by the CV-2000-CuO sample (Figure 1f,g), the CV treatment induces the formation of a distinct hierarchical structure. The surface of the original CuO nanorods appears to partially reconstruct, generating a dense and uniform coating of ultra-thin secondary nanosheets. SEM analysis of samples subjected to different numbers of CV cycles (2000, 4000, 8000; Figure S2) indicates that while the overall hierarchical morphology is consistent, the thickness of the secondary nanosheet layer increases with the cycle number. This intricate hierarchical architecture, comprising primary nanorods decorated with secondary nanosheets, is anticipated to significantly enhance the material’s specific surface area and expose a greater density of catalytically active sites compared to the P-CuO.

We subsequently evaluated the CO₂RR performance of P-CuO and CV-CuO samples subjected to different numbers of CV cycles (2000, 4000, and 8000). Electrochemical tests were conducted in a three-electrode H-cell using 0.1 M KHCO₃ electrolyte pre-saturated with CO₂ for 30 min. All catalysts underwent electrochemical pretreatment at −1.2 V vs. Ag/AgCl for ≥30 min to achieve a stable current plateau, ensuring complete activation and maximal exposure of active sites prior to CO₂RR measurements. Gas and liquid products were quantified by gas chromatography (GC) and nuclear magnetic resonance (NMR) spectroscopy, respectively (Figure S3). As shown in Figure 2a–f, P-CuO primarily produced H₂ with minor formate (HCOO⁻), whereas CV-CuO exhibited marked suppression of H₂ production alongside concomitant enhancement of formate selectivity. Figure 2e and Figure S4 compare the potential-dependent FE trends for formate and H₂ across CV-CuO catalysts with varying cycle numbers. Notably, CV-2000-CuO achieved peak performance at −1.4 V vs. RHE, delivering the highest formate FE (42%) and lowest H₂ FE (40%). Linear sweep voltammetry (LSV) curves in Figure 2f further confirmed superior activity for all CV-CuO catalysts, evidenced by positively shifted onset potentials for CO₂ reduction. Moreover, the CV-2000-CuO catalyst also show much higher current density during stability tests for over 20 h, further demonstrated the effectiveness of CV activated method in promoting the activity of CO₂RR (Figure S5).

To elucidate the origin of performance enhancement in CV-CuO catalysts, we performed phase analysis via XRD on as-prepared P-CuO and CV-CuO samples (Figure 3a). All patterns exhibited characteristic CuO peaks at 35.5° (002) and 38.7° (111) alongside metallic Cu substrate signals at 43.3° (111), 50.4° (200), and 74.1° (220), confirming that CV activation preserved the crystalline structure of CuO nanorod arrays on Cu substrates. Post-activation XRD analysis of P-CuO and CV-2000-CuO (Figure 3b) revealed complete disappearance of CuO reflections with only metallic Cu peaks remaining, indicating bulk reduction of Cu²⁺ to Cu⁰ during electrochemical pretreatment. Complementary ex situ XPS analysis revealed a negative binding energy shift in the Cu 2p₃/₂ main peak accompanied by attenuated satellite features (Figure 3c), collectively confirming substantial depletion of surface Cu²⁺ species in both activated catalysts. Crucially, Cu LMM Auger spectra analysis uncovered distinct valence distributions (Figure 3d): significant attenuation of Cu²⁺ spectral features coupled with emerging Cu⁺ and Cu⁰ signatures confirmed substantially higher Cu⁰ content in activated CV-2000-CuO relative to P-CuO. This electronic restructuring correlates with facilitated CO₂RR kinetics, attributable to enhanced exposure of catalytic active sites within its unique hierarchical architecture.

BET surface area analysis revealed that both CV-2000-CuO and P-CuO catalysts lack mesopores and micropores. However, CV-2000-CuO exhibited a significantly higher specific surface area (0.0922 m² g⁻¹) compared to P-CuO (0.0457 m² g⁻¹) (Figure S6). Charge transfer resistance characterization further demonstrated the superior electrochemical performance of CV-2000-CuO. Prior to CO₂RR testing, CV-2000-CuO already displayed lower solution resistance (R_s) and charge transfer resistance (R_ct) than P-CuO. Post-testing, both samples exhibited decreased resistances, with a more pronounced reduction for P-CuO; notably, CV-2000-CuO achieved a further decrease (Figure 4a). This indicates that the CV treatment significantly enhanced the conductivity and interfacial reaction kinetics of CV-2000-CuO. Consistent with this, the electrochemical active surface area (ECSA), proportional to the double-layer capacitance (C_dl), was markedly higher for CV-2000-CuO than for P-CuO (Figure 4b and Figure S7), suggesting more exposed active sites and a surface more favorable for reactant adsorption and reaction. In situ Raman spectroscopy provided mechanistic insights: characteristic CuO peaks weakened and disappeared with increasing applied potential for both catalysts, indicating reduction to metallic Cu (Figure 4c,d). Crucially, CV-2000-CuO exhibited stronger carbonate signals than P-CuO, while neither sample showed obvious bicarbonate peaks. Concurrently, adsorption features of formate intermediates (COO⁻) and atop-bound CO (CO_atop) were observed (Figure 4c,d), further confirming formate as the primary product in this catalytic system.

To assess the nanoscale morphological evolution induced by CV activation, high-resolution TEM (HRTEM) analysis was performed on as-prepared P-CuO and CV-2000-CuO catalysts (Figure 5 and Figure S8). Both catalysts retained nanorod frameworks consistent with SEM observations (Figure 1). P-CuO nanorods exhibited smooth surfaces (Figure 5 and Figure S8a,b), while CV-2000-CuO displayed well-defined hierarchical nanostructures epitaxially grown on nanorod backbones (Figure 5d and Figure S8c,d). HRTEM images and corresponding selected-area electron diffraction (SAED) patterns confirmed crystalline characteristics of CuO for both catalysts, exposing dominant CuO facets including (111), (110), and (−202) (Figure 5b–f). Crucially, CV-2000-CuO exhibited broader crystallographic orientation diversity, evidenced by complex lattice arrangements in HRTEM (Figure 5e) and SAED patterns (Figure 5f).

Post-electrochemical-activation SEM characterization (Figure S9a–f) revealed distinct morphological evolution pathways under cathodic polarization: P-CuO maintained nanorod arrays with increased surface roughness (Figure S9a–c), whereas CV-2000-CuO preserved hierarchical features despite partial structural densification (Figure S9d–f). After extended CO₂RR operation, P-CuO underwent significant morphological degradation exhibiting melting-like transformation (Figure S10a–c), contrasting sharply with CV-2000-CuO’s retained hierarchical framework despite localized nanosheet coalescence (Figure S10d–f). These findings demonstrate the superior structural integrity of CV-activated architectures under electrochemical stress.

Post-electrochemical-activation microstructural characterization (Figure 6) unveiled fundamental distinctions in nanoscale architecture between CV-activated and pristine catalysts: HRTEM coupled with SAED analysis demonstrated that CV-2000-CuO developed high-density grain boundaries decorated with 10–30 nm metallic copper nanoparticles. These nanoparticles exhibited mixed crystallographic orientations, predominantly exposing (111), (002), and (022) lattice planes as evidenced by distinct diffraction spots (Figure 6d–f). This multifaceted configuration contrasts sharply with P-CuO, which formed larger crystalline domains (>50 nm) dominated by low-energy Cu (111) facets. EBSD mapping (Figure 6g–j and Figure S11) quantitatively confirmed enhanced grain boundary density and homogeneous Cu distribution in CV-2000-CuO, directly correlating with its optimized catalytic performance through increased active site accessibility.

4. Conclusions

We developed hierarchical CuO nanorod arrays via CV treatment as efficient freestanding electrodes for selective CO₂-to-formate conversion. The optimized CV-2000-CuO electrode achieves 42% formate FE at −1.4 V vs. RHE, significantly outperforming pristine CuO. Structural characterization demonstrates that CV processing generates: (a) metastable Cu species with higher surface Cu⁰ content, and (b) high-density grain boundaries during CO₂RR operation. These features collectively enhance * CO₂ intermediate stabilization and suppress HER. The preserved hierarchical structure during long-term operation further confirms the robustness of CV-engineered electrodes. This work establishes CV treatment as a powerful tool for designing interface-rich freestanding electrodes, opening new pathways for selective CO₂RR catalyst design.