Nanostructuring copper thin film electrodes for CO₂ electroreduction to C₂₊ products

Abstract

The electrochemical CO₂ reduction reaction (CO₂RR) is a promising approach for achieving carbon-neutral processes in the chemical industry. In this context, various nanostructures have been reported to enhance the C₂₊ selectivity of Cu-based catalysts. Here, we prepared Cu nanoneedles (NN) from 300 nm sputtered Cu thin films through anodization under various conditions and investigated their performance in terms of C₂₊ product selectivity. Various combinations of anodization potentials (+0.75 V_RHE, +0.85 V_RHE, and +0.95 V_RHE) and KOH electrolyte concentrations (0.1 M, 0.5 M and 1.0 M) allow the tailoring of the NN length and density that are linked to their CO₂RR product selectivity at −1.0 V_RHE. The best performance using the C₂₊ : C₁ ratio was achieved with a high NN surface density. A detailed analysis using high-angle annular dark-field scanning transmission electron microscopy and X-ray absorption fine structure spectroscopy of the best performing sample shows that the anodization of a Cu thin film produces NNs composed of a uniform 3D network of 2 nm hydroxide nanoparticles (NPs) and reconstructs into a rougher metallic Cu NP network after the CO₂RR. A high density of NNs with this inner structure may lead to an increase in the local CO concentration and thus to C₂₊ products. This systematic work demonstrates that nanostructuring the surface of copper thin film electrodes can enhance the CO₂RR selectivity to C₂₊ products while the correlation between the NN morphology and their inner structure strengthens further their applications as CO₂ electrocatalysts.

---

1. Introduction

The electrochemical CO₂ reduction reaction (CO₂RR) is a promising emerging technology to mitigate the increase in anthropogenic carbon dioxide levels responsible for global climate change. Coupled with renewable energy, the electrochemical conversion of CO₂ in aqueous media can produce valuable fuels and chemicals that are conventionally derived from petroleum processes.^1–3 Among them C₂₊ products and in particular ethylene are the most desired molecules by industry. The latter molecule features a high energy density and is an important precursor in many industrial processes.⁴

So far, the optimal binding energy for adsorbed CO (*CO) is a unique feature of Cu and is thought to enable the formation of multicarbon products, especially C₂₊.⁵ However, despite extensive research and technological advances of the state-of-the-art Cu-based CO₂RR catalysts,⁶ their insufficient selectivity and poor stability remain important challenges. Better control over the *CO intermediate may allow unlocking the selectivity of chemical pathways to C₂₊ products which could aid in the design of improved materials.⁷

Recently, nanostructuring of copper surfaces has been introduced as an efficient method to favour the formation of C₂₊ over C₁ products. In particular, Cu-derived (Cu(OH)₂ or CuO) nanoneedle (NN) structured electrodes feature a significant enhancement in C₂₊ product formation, while their faradaic efficiency (FE) for methane formation is minimized to less than 1%, allowing an efficient separation of C₁ and C₂₊ products.⁸ This remarkable selectivity improvement has been attributed to the force-field effect created by the high-curved structures of NNs that can accommodate a higher number of hydrated cations.⁹ Their presence seems crucial in promoting the hydrocarbon selectivity as demonstrated by Monteiro et al.¹⁰ However, the exact nature of the active sites on oxide-derived Cu is still debated. Some propose that dual Cu^δ+–Cu⁰ sites^11–15 participate in the enhancement of C₂₊ product selectivity. According to other studies, the reduction of copper to Cu⁰ always precedes the CO₂RR,^16–18 with a recent study highlighting the important role of the metal particle size in boosting C₂₊ production.¹⁶

Cu NNs can be formed chemically¹⁹ or more conveniently through the anodization of a metal plate, usually in an alkaline solution,^20,21 with few works using bicarbonate electrolytes.²² The anodization regime offers a versatile, fast, and economical way to nanostructure metallic surfaces, including copper. The self-organizing process allows for the growth of nanostructured mixed Cu(OH)₂ and CuO phases that increase the original metal surface area by several orders of magnitude.

The utilization of thin films produced by physical vapor deposition methods has been proven to be very appealing for industrial catalysis applications,^23–26 such as gas diffusion electrodes that are used in electrolyzers. Compared to conventional bulk foils, their film fabrication combines a better control over the film thickness with a higher flexibility regarding the type of substrate. Nonetheless, the specific morphological changes induced by their reconstruction under electrochemical treatments have not been explored in depth. The growth of NNs via the anodization of a Cu electrode is usually controlled by the applied potential or current, the electrolyte concentration and the pH used.^27,28 Typically, a high current density or potential is applied in a highly concentrated KOH electrolyte to develop the NN structures.^8,29,30 However, neither a rational justification of the applied conditions nor a systematic investigation of the link between the NN structure/morphology and its CO₂RR selectivity has been given.

In this study, the relationship between the potentiostatic anodization conditions of copper thin films and the resulting NNs’ morphology, length and density as well as their respective ethylene or C₂₊ selectivity in the CO₂RR is investigated. Pre- and post-reaction characterization of the best-performing NN electrode using a combination of electron microscopy and X-ray emission and absorption spectroscopy allows further understanding of the structural origin of its CO₂RR selectivity. The results of this work will contribute to a more rational design of selective CO₂RR electrocatalysts.

2. Experimental methods

2.1. Synthesis of copper electrodes

Cu/Pt/glass (99.99%, Neyco/Schott Benelux) electrodes were fabricated using radiofrequency (RF) sputtering following the procedure described in ref. 31. Using Kapton tape, the geometric area of the electrodes was determined to be 0.7 × 0.7 cm².

2.2. Electrochemical experiments

2.2.1. Anodization of copper thin films. The anodization experiments were conducted in a three-electrode single compartment cell. The KOH (≥85%, pellets, ACS reagent, Fluka) concentration of the electrolyte was varied in a range from 0.1 to 1.0 M. All the electrolytes were diluted to the required concentrations by mixing them with Milli-Q water (18.2 MΩ cm). For the anodization process, a Hg/HgO (1.0 M KOH) reference electrode (RE) and a Pt gauze (2.25 cm²) counter electrode (CE) were used. Applied potentials were measured with respect to the Ag/AgCl and Hg/HgO REs and converted to the reversible hydrogen electrode (RHE) scale:where E_RHE is the converted potential versus the reversible hydrogen electrode, RHE, E_(Ag/AgCl, _{KCl 3.4 M)} is the measured potential against the Ag/AgCl (KCl 3.4 M) reference electrode, and is the standard potential of Ag/AgCl at 25 °C (0.205 V_SHE). The same applies for the Hg/HgO RE with E°_(Hg/HgO, _{KOH 1 M)} at 25 °C (0.098 V_SHE). Unless otherwise specified, potentials reported hereafter are given vs. RHE (V_RHE).

The measurement protocol starts with cyclic voltammetry  (CV) in the −0.15 – +1.5 V_RHE potential window using a scan rate of 100 mV s⁻¹ to identify the Cu redox peaks. For each anodization, a fresh Cu electrode was used and was polarized at −0.1 V_RHE for 50 s to reduce the thickness of the native surface oxide layer, before the application of oxidative potentials of +0.75 V_RHE, +0.85 V_RHE and +0.95 V_RHE for 150 s. Each applied potential results in a different electrode. After anodization, the electrode was rinsed with Milli-Q water and dried under N₂.

2.2.2. Electrochemical CO₂ reduction. The electrochemical experiments were carried out in a gas-sealed H-cell controlled from an electrochemical potentiostat (CS2350 bipotentiostat, Corrtest Instruments). The two compartments were separated by a cation exchange membrane (117-Nafion). A leak-free Ag/AgCl 3.4 M KCl (LF-1.6, Innovative Instruments, Inc.) was used as the RE and a Pt gauze (2.25 cm²) as the CE. CO₂ saturated-0.1 M KHCO₃ with pH 6.8 (≥99.7%, p.a., ACS, Carl Roth) (7 ml, 3 ml headspace) and 1.0 M KOH (6 ml) electrolyte solutions were used in the cathodic and anodic compartments, respectively. Before the CO₂RR, the catholyte was bubbled with CO₂ for 40 min and stirred with a magnetic stirrer at a constant rate of ∼500 rpm. For the electrolysis, a CO₂ flow rate (4.8 grade, Air Liquide) of 5 ml min⁻¹ was used. Electrochemical impedance spectroscopy (EIS) was used to measure the uncompensated resistance R_u (where Z_i is 0) of the electrochemical system at a potential of −1.0 V_RHE in the frequency range 4 × 10⁵−1 Hz. The R_u value was used in the post-iR correction as given by E = E_RHE − iR_u.

Gaseous products were analysed with a gas chromatograph (GC, Trace 1300, Thermo Fisher Scientific) using a sampling interval of 12 min. At the end of each experiment, 3 ml of electrolyte was stored in a refrigerator to preserve the volatile compounds. An Agilent 1200 high-performance liquid chromatography system with an Agilent Hi-Plex H 7.7 × 300 mm column was used to separate the product, and an Agilent 1260 RID detector was used to detect and quantify formate in the form of formic acid. The samples were previously diluted with water or a mobile phase. Here 0.01 M H₂SO₄ was used as the mobile phase. Organic products concentrations (i.e. ethanol, propanol and acetaldehyde) were measured using headspace-GC combined with an FID and MS detector. The compounds were separated on a DB-WAXetr column 60 m × 0.25 mm × 0.25 μm. Vials were heated in a closed headspace vial and the headspace was injected into the GC. A specific temperature profile was used to separate the several alcohols. The faradaic efficiencies (FE) of the different products were calculated using:

FE_i,gas(%) = Q_exp,i/Q_total = (z_i × n_i × F)/(I × t), (3)

FE_i,liquid = (c × V × n_i × F)/(M_w × I × t), (4)

where Q_exp,i is the amount of charge consumed for product i, Q_total is the total accumulated charge, F is the Faraday constant (96 485 C mol⁻¹), z_i is the number of exchanged electrons for each specific product, and n_i is the number of moles of product i. Q_total is the product of applied current (I) and experiment duration (t), c is the concentration of liquid products, V is the volume of the electrolyte, and M_w is the molar mass of certain liquid products.

Before CO₂ electrolysis, linear sweep voltammetry (LSV) curves were recorded from 0 to −1.0 V_RHE at a scan rate of 25 mV s⁻¹. Chronoamperometry (CA) was performed at −1.0 V_RHE for 1 h. The LSV and CA curves are shown in Fig. S3†. For the total FE estimation, the first 12 min are not included since the system had not reached steady-state in the H-cell (Fig. S4a†). The reported FE values results from three independent experiments. At the end of the experiment the electrolyte pH after the CO₂RR was 7.3.

2.3. Materials characterization

2.3.1. X-ray diffraction (XRD). The crystal structure of the synthesized materials was characterized by X-ray diffraction (X'Pert PRO MRD system) using Cu Kα radiation with a 1.5405 Å wavelength at a current of 40 mA and voltage of 50 kV. The scanning rate was 2° per min and the 2θ angle ranged from 10 to 80°.

2.3.2. Scanning electron microscopy (SEM). Micrographs of the AN-(x, y) electrodes were obtained by SEM using a FEI Helios Nanolab G3 CX DualBeam FIB/SEM with an acceleration voltage of 10 kV and an electron beam current of 86 pA.

2.3.3. X-ray photoelectron spectroscopy (XPS). The chemical stoichiometry and oxidation states of Cu and O were studied with X-ray photoelectron spectroscopy (XPS). XPS measurements were performed in a home-built setup at room temperature using Mg Kα X-rays (XR4 twin anode X-ray source, non-monochromatic) and an Alpha 110 hemispherical analyser. The spectra were acquired in ultra-high vacuum (UHV) (base pressure ∼10⁻⁸ mbar) with an energy step of 0.1 eV, an energy resolution of ≈1 eV, and a beam diameter of 5 mm. The recorded spectra were corrected by using the adventitious carbon peak C 1s at 284.8 eV.

2.3.4. X-ray absorption spectroscopy (XAS). Measurements were carried out at the XAFS beamline³² of the Elettra synchrotron (Trieste, Italy) equipped with a Si (111) double crystal monochromator and water-cooled Pt-coated silicon mirrors which ensured higher order harmonic rejection. The X-ray absorption near edge spectroscopy (XANES) and extended X-ray absorption fine structure (EXAFS) of the electrodes at the Cu K-edge (8979 eV) were collected under ambient conditions in fluorescence detection mode using a silicon drift detector (SDD) and the incoming beam was measured using a nitrogen-filled ion chamber. The energy calibration was performed using the absorption spectrum of a metallic Cu foil. The XAS data were recorded up to a wavenumber of 12 Å⁻¹, with each spectrum taking approximately 30 minutes. At least two spectra were averaged to improve the signal to noise ratio. The background and normalization were carried out by fitting a straight line for the pre-edge and cubic spline, simulating the absorption of an isolated atom, for the post edge.

2.3.5. High-angle annular dark-field scanning transmission electron microscopy (HAADF STEM) and energy dispersive X-ray spectroscopy (EDS). The measurements were performed using an aberration-corrected cubed ThermoFisher Scientific Titan transmission electron microscope, operated at an accelerating voltage of 300 kV. The instrument was equipped with a Super X EDS detector. HAADF STEM was carried out ex situ after the anodization and the CO₂RR by scratching the electrode surface, sonicating the detached material in ethanol, and drop casting the suspension on ultra-thin film (UTF) carbon coated Au TEM grids. TEM lamellas were prepared using a Helios 5 Hydra UX plasma focused ion beam (PFIB) DualBeam system (Thermo Fisher Scientific Inc.)

3. Results and discussion

3.1. Anodization of sputtered copper electrodes

A series of magnetron sputtered 300 nm Cu thin films on glass substrates, Cu₃₀₀, were anodized in various KOH concentrations and under different potentials to tailor their morphology.

Potentiostatic anodization was chosen over galvanostatic to explore the anodization of copper electrodes at specific potentials and to have the complete control over the reactions happening at the surface. To determine the optimal oxidation potential, CV was performed from +1.5 to −0.15 V_RHE using a scan rate of 100 mV s⁻¹ at KOH electrolyte concentrations of c_KOH = 0.1 M, 0.5 M and 1.0 M, as shown in Fig. 1a. The cathodic peaks C and D are specific for the reduction of the oxidized species Cu²⁺ and Cu⁺ to Cu⁰, respectively. The oxidation peak A in the anodic sweep corresponds to Cu oxidation into Cu₂O, i.e. Cu⁰ → Cu⁺, while peak B is characteristic of Cu⁺ and/or Cu⁰ → Cu²⁺ oxidation.

Fig. 1 (a) Cyclic voltammograms of the sputtered Cu electrodes, Cu₃₀₀, in 0.1 M, 0.5 M and 1.0 M KOH for the potential window between −0.15 and 1.5 V_RHE, recorded at a scan rate of 100 mV s⁻¹. Inset: Cu²⁺ oxidation region, peaks B₁ and B₂. The vertical dotted lines represent the selected anodization potentials E₁, E₂ and E₃. (b) Schematic representation of the anodization process for Cu sputtered film, Cu₃₀₀. The straight arrows represent the interaction of species, i.e. Cu and Cu²⁺ with OH⁻ to give oxidation products, and the wavy arrow depicts the dissolution of Cu metal to Cu²⁺. (c) Transient j vs. t curves for the different E_an potentials in 0.1, 0.5, and 1.0 M KOH for 150 s.

Peak B splits into two subpeaks, B₁ and B₂. Notably, the former corresponds to the formation of Cu(OH)₂ and the latter to CuO oxidation products, as given by the following chemical reactions:^21,33

Cu + 2OH⁻ → Cu(OH)₂ + 2e⁻ (5)

Cu + 2OH⁻ → CuO + H₂O + 2e⁻ (6)

Three anodization potentials (E_an) were selected based on the B peak of this CV response and are indicated by the dashed lines in the inset of Fig. 1a. Considering the specific reactions, i.e.(1) and (2), taking place at potentials close to B₁ and B₂ peaks, the electrode's morphology can be altered. Therefore, the potentials selected to investigate the growth of NNs are located before peak B₁ (E₁ = 0.75 V_RHE), around B₁ (E₂ = 0.85 V_RHE) and close to B₂ (E₃ = 0.95 V_RHE) to explore the region of NN formation and to optimize their shape (inset Fig. 1a).

According to Shoesmith et al.³⁴ the anodization starts with the formation of an initial Cu₂O layer as described by reaction (7). As this porous layer develops, metal dissolution in the form of Cu²⁺ occurs given by reaction (8). Since the pores get wider with time, some dissolution of Cu₂O must also take place according to reaction (9).

2Cu + 2OH⁻ → Cu₂O + H₂O + 2e⁻ (7)

Cu + nOH⁻ → (Cu(OH)_n)²⁻ⁿ + 2e⁻ (8)

Cu₂O + H₂O + (2n − 2)OH⁻ → 2(Cu(OH)_n)²⁻ⁿ + 2e⁻ (9)

(Cu(OH)₄)²⁻ ↔ Cu(OH)₂ + 2OH⁻ (10)

Although the anodization process is not fully understood yet, it is generally claimed to be governed by the dissolution processes of soluble (Cu(OH)₄)²⁻ complexes.³⁵ The build-up of (Cu(OH)₄)²⁻ near the electrode surface can lead to the nucleation and growth of insoluble Cu(OH)₂ crystals through reaction (10). A schematic of the NN growth is shown in Fig. 1b.

The transient responses of current density j–t, depicted in Fig. 1c, can provide a mechanistic insight into the anodization process. Specifically, an initial drop in current density is observed across all KOH concentrations and anodization potentials. In the j–t curves, the timing of the drop varies, and the fastest drop is observed at 0.85 V_RHE, followed by later drops at 0.75 V_RHE and 0.95 V_RHE. The observed drop is likely due to the formation of a Cu₂O insulating layer resulting in an increase of the charge transfer resistance of the electrode–electrolyte interface. The variation in time may reflect differences in the number of NN nuclei. The ascending current density is most pronounced for 0.85 V_RHE in 0.5 and 1.0 M KOH and for 0.75 V in 1.0 M KOH. No current increase is seen for the 0.1M KOH-series and for 0.95 V_RHE at 0.5 and 1.0 M KOH. This current increase translates into the nucleation and growth of the upper layer of (Cu(OH)₄)²⁻ resulting from the native oxide layer dissolution. Once the dissolution of these species starts, channels where fresh Cu is accessible to be dissolved further are available. Gradual supersaturation of (Cu(OH)₄)²⁻ causes Cu(OH)₂ to grow into a nanoneedle-like structure as indicated by the maximum value reached by the current.¹⁸ As the Cu(OH)₂ crystals start to overlap and to cover the surface, partial passivation takes place and the current declines. Moreover, depending on the applied E_an the colour of the Cu surface changes from luminous salmon red to darker tones indicating the successful creation of anodization products on the surface (Fig. S1†).

3.2. Effect of c_KOH and E_an on the Cu NN structure

For convenience, the anodized copper electrodes will be referred to as AN-(x, y) where x is the applied potential and y the molarity of KOH in M used for the anodization. Depending on the anodization conditions, distinct Cu NN morphologies compared to that of planar Cu₃₀₀ film can be identified through SEM in Fig. 2a. Although NNs are present for the lower 0.1 M concentrations at 0.85 and 0.95 V_RHE, their growth is limited due to the insufficient supply of OH⁻. Significant higher coverages of NNs are observed on the electrode surface when the electrolyte concentration is increased to 0.5 M and especially to 1.0 M. For this reason, thorough characterization was performed for the full set of c_KOH = 0.5 and 1.0 M at 0.75 V_RHE. For AN-(0.75, 0.5), the density of the formed NNs is low and a significant fraction of the porous Cu₂O layer is exposed. With respect to the j–t transient curves, the low-density morphology can be explained by the transport of Cu²⁺ species from the dissolution sites to the crystal-growth sites. If the metal dissolution is facilitated, supersaturation of soluble Cu²⁺ complexes occurs at the electrode surface leading to fast nucleation and growth of long NNs (average length of 3 ± 1 μm).

Fig. 2 Effects of anodization potential and KOH concentration in AN-(x, y): (a) surface morphology via SEM imaging versus the planar surface of Cu₃₀₀. (b) crystallographic structure through XRD reflections. The rhombus symbol represents the characteristic cubic Cu planes (111), (200) and (220). The circle depicts the cubic Pt planes (111), (200) and (220), and the heart symbol shows the orthorhombic Cu(OH)₂ planes (020), (021), (002), (111), (041), (061), and (200), all in an ascending order of 2 theta, and (c) electronic structure from XPS Cu 2p core level spectra.

Although the NN morphology strongly relies on the applied E_an, the current density can also provide information about their growth.³⁶ Since the electrolyte concentration is relatively low (0.5 M), a low number of NN nucleation sites are formed explaining the inhomogeneous covering of the surface. The AN-(0.85, 0.5) electrode is characterized by a more homogeneous morphology with a high density of shorter NNs (average length of 0.7 ± 0.2 μm). At high overpotentials, there is a rapid increase in the number of nucleation sites leading to a fast surface coverage inhibiting the metal dissolution and resulting in the growth of smaller crystals. Electrode AN-(0.95, 0.5) has only a few NNs grown horizontally on a roughened substrate. This condition falls in the passivation region that allows for a spontaneous oxidation of the Cu₂O layer to CuO preventing the NN formation. In AN-(0.75, 1) synthesized at the highest 1.0 M KOH concentration, a very distinct morphology with long bundles of NNs of a few micrometres (∼2.7 μm) growing in different directions on the electrode is found. Recall that the j(t) transient of AN-(0.75, 1) reached a steady-state with a constant current density of 3 mA cm⁻² (Fig. 1c). This entails an equilibrium between the competing copper dissolution and the passivation reactions, slowing down the conversion of Cu to CuO.

The growth of NN structures depends on the electrolyte concentration as well as on the pH³⁷ while the use of c_KOH ≥ 0.5 M is essential. The NN growth starts with a Cu₂O layer, which, depending on the applied potential, (partially) dissolves to generate several nucleation sites. Anodization using potentials below the B₁ peak (0.75 V_RHE) results in the growth of long NNs whose surface density increases with c_KOH as illustrated by the SEM image in Fig. 2a. The anodization potential just above the B₁ peak (0.85 V_RHE) or at the onset of the B₂ peak produces densely packed shorter NNs. Once the applied anodization potential is close to the B₂ peak (0.95 V_RHE), the NN growth is inhibited, as the reaction goes under the passivation regime control. The estimated depth of anodization related to the reconstruction of Cu₃₀₀ is calculated by integrating the current of the CA curves (Table S1†). Anodization depths of ∼107, 83, 10 and 5 nm were estimated for AN-(0.75, 1), AN-(0.85, 0.5), AN-(0.95, 0.5) and AN-(0.75, 0.5), respectively (ESI S.2†).

The XRD patterns of the anodized electrodes are shown in Fig. 2b. The bare Cu₃₀₀ electrode has diffraction peaks at 43.3°, 50.4° and 74.1° which correspond to the (111), (200) and (220) planes of cubic fcc Cu (ICSD01-089-2838) shown by the rhombus symbol. The Cu thin film shows a (111)-preferred orientation. Additionally, the circle symbol is used for the peaks corresponding to the (111), (200), (220) planes of cubic Pt (ICSD01-087-0640) and observed at 2θ angles of 39.9°, 46.4°, and 67.4°, corresponding to the Pt layer supporting the sputtered copper film. The XRD peaks of the AN-(0.75, 0.5) and AN-(0.75, 1) samples located at 16.56°, 23.64°, 33.98°, 35.78°, 37.98°, 54.10°, and 63.52° correspond respectively to the (020), (021), (002), (111), (041), (061), and (200) planes of orthorhombic copper hydroxide Cu(OH)₂ (ICSD01-072-0140) as indicated by the heart symbol to indicate the presence of NNs. The lower intensity of these reflections in AN-(0.85, 0.5) suggests that a smaller amount and/or a less crystalline phase of Cu(OH)₂ is formed in this sample. In AN-(0.95, 0.5), only the Cu metal reflections corresponding to non-anodized substrate were observed suggesting that surface passivation has occurred impeding the formation of anodization products. The passivated area is either amorphous and/or of nanoscale dimension.

XPS of the Cu 2p core level was used to determine the initial surface oxidation state of the anodized Cu electrodes (Fig. 2c). The spin–orbit splitting of Cu 2p (Δ = 19.75 eV) yields Cu 2p_1/2 : Cu 2p_3/2 components with an area ratio of 1 : 2. Our analysis is based on Cu 2p_3/2 where the peak at 932.6 eV is assigned to Cu⁺/Cu⁰, and peaks located close to the 933.5/934.5 eV (Cu²⁺ region) stem from the Cu²⁺ species in the oxide and hydroxide phases, respectively. The shape of the intense Cu²⁺ satellite peaks appearing around 937–947 eV was qualitatively used to distinguish the hydroxide and oxide components.³⁹ Based on the satellite peak appearance, it can be concluded that Cu²⁺ species are present in all the AN-(x, y) electrodes, agreeing with the expected oxidation state from Pourbaix diagrams⁴⁰ and the work of Stepniowski et al.⁴¹

The copper oxidation states can be further linked to the anodization parameters. Specifically, AN-(0.75, 0.5) consists of a mixture of Cu₂O and Cu²⁺ phases comprising Cu hydroxide as detected by XRD and possibly amorphous CuO. The Cu₂O phase is ascribed to the electrode's surface since the sample consists of sparse NNs (Fig. 2a); therefore its amount will be less than that of Cu²⁺. Electrode AN-(0.85, 0.5) mainly consists of Cu²⁺, Cu(OH)₂ as indicated by XRD. Unlike in the two previous electrodes, AN-(0.95, 0.5) surface appears to be fully oxidized into CuO as pointed out by the shape of the satellite peak and the absence of a hydroxide diffraction peaks. Although the hydroxide/oxide fraction cannot be precisely quantified, it is inferred that with a decreasing E_an, the hydroxide component systematically increases as does the density of NNs since they are largely composed of Cu hydroxide. Further confirmation of the prevalence of the Cu²⁺ state in the form of CuO and Cu(OH)₂ phases is obtained from the Cu Auger LMM transition peak and the modified Auger parameter (Fig. S2 and Table S2†).

3.3. Effect of NN morphology on the C₂₊ selectivity

Fig. 3a shows the preliminary screening of the FE of the gaseous products formed under the CO₂RR at −1.0 V_RHE for the different AN-electrodes. Only gaseous products are shown since we are interested in ethylene production. Ethane, which is usually produced from porous electrodes,⁴² was also detected but accounted for less than 1% of the total (Fig. S4b†). The missing FE values are of H₂ and liquid products. Additional liquid product analysis quantified the presence of formic acid, ethanol, propanol, and acetaldehyde (vide infra). Variations in the FE follow changes in the geometrical current density, which is larger by a factor of about 2.2 for the NN-electrodes. Typical values for the NN-free electrodes are around −16 mA cm⁻² while those with NNs feature, depending on their morphology, current densities up to −33.5 mA cm⁻².

Fig. 3 Electrochemical CO₂RR performance of the AN-(x, y) electrode series: (a) CO₂RR faradaic efficiency of gaseous products, CO, CH₄, and C₂H₄ and total current densities, (b) Contour plots derived from Fig. 3a showing the effect of the anodization process parameters on the CO₂RR to CH₄ (top) and C₂H₄ (bottom), and (c) obtained total C₂₊ : C₁ ratio. The ratio for *Cu e-pol, refers to electropolished Cu, and is taken from ref. 38 (d) CO₂RR faradaic efficiency of gaseous and liquid products and respective total current density of Cu₃₀₀, AN-(0.75, 1) and AN-(250). Error bars represent the standard errors of the mean derived from three independent measurements; averages are plotted. The electrolysis experiments were held at a constant potential of −1.0 V_RHE for 1 h.

The contour plots shown in Fig. 3b reveal the variations in the FE for CH₄ and C₂H₄ under the CO₂RR with changes in the electrode morphology resulting from differences in the anodization conditions. Relating the CO₂RR performance to the electrode morphology shows that CH₄ is the dominant CO₂RR product in the absence of NNs. On the other hand, C₂H₄ is favoured on electrodes featuring NNs. For the electrodes that produced the highest amount of C₂H₄, product selectivity was further quantified by the C₂₊ : C₁ ratio (C₂₊ products are ethylene, ethanol and propanol) in Fig. 3c whose maximization can be used as a criterion to select the electrode with the best performance. This ratio that equals 1.4 in the reference electropolished Cu foil³⁸ increases by a factor of 2–3 in the different anodized Cu thin films and reaches a maximum of ∼6 for AN-(0.75, 1).

Fig. 3d depicts the CO₂RR FE of AN-(0.75, 1) versus that of the initial thin film Cu₃₀₀ and AN-(250), a new electrode with anodization time, which will be discussed later. Anodization of Cu₃₀₀ leads to a significant improvement in its C₂₊ selectivity (Fig. 3d) that reaches 36% including FE_C2H4 of 23%, FE_EtOH of 10% and FE_PrOH of 3% while the FE for C₁ products is 6.2% including FE_HCOOH of 4% and FE_CH4 of 1.6%. A minor amount of CO (< 0.5%) was detected which was not included in the C₁ products. This electrode reaches a partial geometric ethylene current density j_C2H4 of −7.6 mA cm⁻² and a total geometric current density of −33.6 mA cm⁻². The remaining current density is expected to participate in the competing H₂ evolution reaction. This result is consistent with earlier works on nanoneedle-based electrodes that generally produce more C₂₊ over C₁ products_.^8,19,29 In comparison, the original Cu₃₀₀ thin film favours the production of C₁ with a FE of 29.8% including FE_CH4 of 16.2%, FE_HCOOH of 12% and FE_CO of 1.6% while C₂ only reaches 7.2% including FE_C2H4 of 4.2% and FE_EtOH of 3%. Although relevant information regarding the relation between the NN morphology and the CO₂RR selectivity can be acquired in an H-cell, this configuration limits the FE towards C₂ products compared to gas-fed flow cells due to the low solubility of CO₂ in water, 0.034 M, resulting in mass transport limitations.

Since all electrodes have been measured in the same H-cell, differences in the selectivity are likely attributed to the morphology and the roughness (SRF) of their surface. To this end, the electrochemical surface area (ECSA) was estimated by performing CV in a suitable non-faradaic region (Fig. S5†).

The surface roughness factors (SRFs) of the as-prepared electrodes (Table S3†) increase sharply after the reaction likely due to the reconstruction process. The post-SRF rank is given in the following ascending order AN-(0.75, 0.5) < AN-(0.85, 0.5) < AN-(0.95, 0.5) < AN-(0.75, 1). However, an inverse order is revealed when the electrodes are compared according to their ECSA normalized current density, j_ECSA (Table S3†) with bare Cu₃₀₀ showing a higher j_ECSA than AN-(0.75, 1). ECSA-normalized CO₂RR activities of low and high surface area Cu catalysts generally depend on the conditions.⁴³ Higher ECSA leads to reactions that are more controlled from mass transport than from kinetics. Since rougher electrodes expose more active sites, they can deliver higher geometric currents that in turn increase the local pH.^44,45 This pH increase can impede the selectivity of the CO₂ reduction pathway toward pH-dependent products such as methane.⁴⁶

The electrode that yields the highest C₂₊ : C₁ ratio, AN-(0.75, 1), exhibits homogeneous high density and long NN features. It is also the electrode with the highest SRF. The increased length/tilting of the high curvature structure can suppress methane formation as proposed by Ma et al.⁴⁷ In our work, the formate production is reduced (corresponding FE_C1 and FE_C2+ can be found in Fig. S4c†). However, AN-(0.85, 0.5) and AN-(0.85, 1) that both feature shorter NNs compared to AN-(0.75, 1) feature similar FE_C2H4 of 18% and 19.2%, respectively, compared to 23% in the latter electrode.

As a proof of concept, we further assessed the influence of the NN length by increasing the anodization time.²⁰ With this objective, we made a new AN-(0.75, 1) electrode with prolonged anodization times from 150 to 250 s: AN-(250) where the value in brackets denotes the anodization time in s. Although this electrode shows an increased surface roughness (Fig. S5†), SEM imaging (Fig. S6b†) does not identify any apparent morphological differences from AN-(0.75, 1). No significant enhancement in FE_C2H4 was observed while the current density only slightly increased from −33.6 to −35.1 mA cm⁻². A stronger dependency of the C₂₊ : C₁ ratio arises from the electrode SRF. It increases with SRF until it reaches a plateau suggesting that mass transport may limit the reaction. Based on this, the SRF which is interconnected with the depth of NN formation and density, appears to have a larger impact on the ethylene/C₂₊ selectivity than on their length. The impact of the selected CO₂ flow rate on the mass transport conditions was assessed by comparing the CO₂RR gas selectivity of the AN-(0.75,1) sample at 5 and 10 sccm. The CO₂ flow rate was shown to only slightly affect the CO₂RR selectivity to gaseous products.

The importance of roughening Cu surface to steer its selectivity towards multicarbon species has been explored in recent AFM studies⁴⁸ as well as from the reduction of copper native oxides under the CO₂RR.⁴⁹ These morphological changes can form new under-coordinated and defective sites that may be active for C–C coupling. Another possible reason for the enhanced C₂₊ selectivity on NNs is the increased bubble hydrodynamics, which enhances mass transport of products and reactants from and towards the electrode surface, respectively.⁵⁰

3.4. Reconstruction of NNs under the CO₂RR

To gain better understanding of the correlation between the pre-reaction catalyst structure and its C₂₊ selectivity, we have investigated further the morphological evolution of the NN structure under the CO₂RR and carried out post-reaction XAS and HAADF STEM for the best-performing AN-(0.15, 1) electrode. Reconstruction of copper electrocatalysts under the CO₂RR has been demonstrated using in situ characterization techniques. For example, Lei et al. highlighted a correlation between grain size and C₂₊ selectivity on reduced oxide-derived Cu nanocrystals, peaking with 11 nm.⁵¹ In another operando STEM study, the size of nanograins formed through the agglomeration of 5–7 nm Cu nanoparticles (NPs), is evidenced to control the CO₂RR product selectivity because of their ability to form higher density metallic Cu grains.¹⁶ Those earlier results suggest that controlling the size of the Cu NPs via the anodization of the electrode may allow tuning the selectivity towards C₂ productivity.

The relative preservation of the NN shape observed by post-reaction SEM enables us to acquire meaningful structural information of our catalysts despite their air exposure between the characterization measurements. Specifically, the SEM image of post-reaction AN-(0.75, 1)-CO₂RR in Fig. 4a shows that although the aspect of the NNs is roughened they retain their original shape. Similar reconstruction of copper electrodes during the CO₂RR has been ascribed previously to the dissolution and redeposition of Cu under negative applied potentials.⁵²

Fig. 4 (a) SEM image of AN-(0.75, 1)-CO₂RR. Cu K-edge XANES (b) and EXAFS (c) spectra of AN-(0.75, 1) before and after the CO₂RR, along with the Cu₃₀₀ substrate and Cu foil, Cu₂O and CuO reference spectra. HAADF STEM image of the (d) initial NN and high-resolution image of the NN broken tip indicated by the red square. The FFT pattern inset shows the zone axis and the crystalline planes of the crystal structures in the green dashed square, confirming the orthorhombic structure (space group: Cmca) of Cu(OH)₂. (e) NN after the CO₂RR and high-resolution image of the NN edge indicated by the blue square. FFT pattern insets of the crystal structures in the orange and maroon dashed squares, confirming the cubic structure (space group: Fm3̄m) of metallic Cu and the cubic structure (space group: Pn3̄m) of Cu₂O, respectively. (f) TEM lamella of the as-prepared NNs and high-resolution image indicated by the orange square, revealing its microstructure with large nanoparticles at the core and smaller nanoparticles surrounding them. (g) TEM lamella of NNs after the CO₂RR, showing enlarged core nanoparticles and the high-resolution image of the large surface particles with smaller nanoparticles below them, indicated by the purple square.

The XANES spectra in Fig. 4b provide information on the copper local electronic structure in AN-(0.75, 1) prior and after the CO₂RR by using reference compounds as standards. The absorption edge shift increases with increasing oxidation state. As expected, the initial pre-reaction profile of AN-(0.75, 1) consists of a mixture of oxides (Cu₂O, CuO/Cu(OH)₂). Their post-reaction profile matches that of the Cu foil and the initial sputtered Cu electrode. This shows that the largely oxidized structure of AN-(0.75, 1) is reduced into metallic Cu during the CO₂RR and remains so under the ambient conditions.

The EXAFS fitting results reflecting the copper atoms local environment are summarized in Table S4.† Fourier transforms (FTs) of k²-weighted EXAFS in Fig. 4c show two peaks at 1.9 Å and 2.5 Å corresponding to Cu–O and Cu–Cu contributions, respectively in the as-prepared AN-(0.75, 1) while post CO₂RR, only a very intense Cu–Cu peak remains. Before the CO₂RR, copper atoms in AN-(0.75, 1) are surrounded by 2.4 O at 1.93 Å and 3.7 Cu at 2.54 Å corresponding to a copper oxide and copper metal shell, respectively. The metal shell accounts for the remaining part of the Cu layer that was not anodized beneath the NN. The fractions of the different copper oxides calculated using a simple combination of the coordination numbers and bond distances was ca. 30% of Cu₂O and ca. 70% of Cu²⁺ (ESI S.7†).

In post-reaction AN-(0.75, 1)-CO₂RR Cu atoms are surrounded by 0.3 O at 1.85 Å corresponding to ca. 15% of Cu₂O phase and 85% of Cu metal with 8.5 Cu at 2.53 Å (ESI S.7†). The corrected average coordination number (N_Cu,r) of Cu metal was found to be 10.0 (8.5/0.85) with a crystallite size of ca. 2.3 nm (Table S5†). From the TEM lamella image after the CO₂RR, we notice that the initial sputtered layer is no longer observed post-reaction, implying that the small crystallite size exclusively coincides with the reconstructed NNs during the CO₂RR. The whole electrode comprising the non-anodized thin film and the NNs might reconstruct upon application of very negative potential.⁵³ The original Cu₃₀₀ whose EXAFS could be fitted with 0.2 O at 1.86 Å and 10.5 (11.7) Cu at 2.54 Å corresponds to ca. 10% of Cu₂O and ca. 90% of Cu⁰ with an average crystallite size of ca. 16 nm in line with the particle size of ca. 20 nm obtained with XRD (Table S6†).

HAADF STEM was used to image the structure of the NN in AN-(0.75, 1). The results show that the as-prepared NNs consist of relatively smooth and uniform long fibres as shown in the HAADF STEM image of an individual NN in Fig. 4d. The corresponding Fast Fourier Transform (FFT) agrees well with the orthorhombic structure (space group: Cmca) of Cu(OH)₂.⁵⁴ The high-resolution HAADF STEM image of a TEM lamella prepared from the sample reveals a microstructure composed of small nanoparticles with an average size of 2.11 ± 0.10 nm (Fig. 4f). Additionally, this cross-sectional analysis shows the formation of a compact polycrystalline copper core within the NNs as shown in Fig. 4f and g and the EDS analysis for this area is shown in Fig. S9.† In as-prepared AN-(0.75, 1), this copper core may consist of Cu hydroxide crystallites of up to 12 nm as detected by XRD (Table S6†).

For AN-(0.75, 1)-CO₂RR, the HAADF STEM image in Fig. 4d shows that the NNs undergo structural and surface morphological changes, including surface roughening. FFT analysis confirms the presence of the cubic structure (space group: Fm3̄m) of Cu metal⁵⁵ and Cu₂O (space group: Pn3̄m).⁵⁶ Moreover, a change in the averaged atomic ratio of Cu : O from 1 : 1 in the as-prepared NNs to 9 : 1 after CO₂RR is observed. As EXAFS pointed out the reduction of copper to a metallic state, the latter ratio indicates that this small oxygen content is likely resulting from air exposure of the samples between the measurements, as it is detected mostly on the surface (Fig. S8 and Table S8†). High-resolution HAADF STEM of the sample TEM lamella reveals enlarged surface particles with average sizes of 5.77 ± 1.83 nm, while smaller nanoparticles of 2.01 ± 0.14 nm remain beneath the surface (Fig. 4f). The top layer consists of smaller Pt NPs deposited during the FIB process for sample protection as confirmed by the EDS maps of the TEM lamella (Fig. S9†).

During anodization, Cu₃₀₀ is transformed according to the conditions up to a specific depth into NNs in the form of a Cu(OH)₂ phase. However, the remaining copper film might participate in the reaction only if CO₂ can diffuse to this NN depth.⁵⁷ This follows the selectivity trend of the AN-(0.75, 1) electrode, where inhomogeneous NN coverage leads to a significant decrease in C₂₊ or C₂H₄ production. The Cu(OH)₂ NPs forming the NNs in AN-(0.75, 1) reassemble into slightly bigger Cu metal NPs of 5 nm in AN-(0.75, 1)-CO₂RR while their polycrystalline domains show a size of 2.3 nm. Our results are in agreement with those of Lei et al., which point to the presence of small Cu nanograins after the CO₂RR either in a heat-quench or anodized Cu surface.²⁹ However, our samples show the occurrence of an even smaller crystal size after the CO₂RR.

The NN inner structure reveals that a 3D-network consisting of 2 nm Cu(OH)₂ NPs created earlier at the anodization stage is reduced to a 5 nm metallic Cu NP network during the CO₂RR. This is supported by the dynamic transformation of catalytically inactive single Cu atoms⁵⁸ into C₂ active-metal nanoparticles. Based on previous works, it is our hypothesis that these porous Cu NP networks efficiently trap CO₂ and CO₂RR gaseous intermediates in between of NNs⁵⁹ or inside their porous structure,⁶⁰ resulting in an increase of the number of available active sites but also of the local CO concentration.⁶¹ In addition, C–C coupling is proposed to be more favourable in these NN strucures.⁴⁷ Our results suggest that a catalyst surface with a high metallic character is needed to allow the formation of C₂₊ molecules. This would also agree with the demonstrated NN tip-induced amplified electric field attracting more hydrated K⁺ cations as shown for CO electroreduction.⁶² A high density of NNs featuring similar inner structures is likely the most important parameter controlling the C₂₊ production.

Utilization of thin film electrodes in scaled-up CO₂ electrolyzers can be envisaged as the anodization process can be ideally applied to the fabrication of gas diffusion electrodes (GDEs) with larger surface areas and higher gas permeation. As zero-gap electrolyzers typically operate with concentrated KOH electrolyte, the initial anodization of the Cu thin films by applying an oxidative current density prior to the CO₂RR would conveniently produce high density NN surfaces. However, testing the long-term performance of these catalysts under industrial conditions would be required for their future application. According to Yu et al.,⁶³ the continuous agglomeration of metallic NPs that may occur during prolonged operation could lower the C₂₊ selectivity of the copper catalysts, leading to their deactivation. Improving the existing electrode configuration by adding an additional layer of a different metal to produce a tandem catalyst⁶⁴ is also a promising strategy to fully utilize the potential of NNs.

4. Conclusions

In conclusion, the influence of E_an and c_KOH anodization parameters on tailoring the density and length of Cu(OH)₂/CuO NNs grown on sputtered Cu films was investigated. Their CO₂RR performance in terms of C₂₊ : C₁ ratio measured in a H-cell surpassed that of bare Cu₃₀₀, with AN-(0.75, 1) ratio featuring a nearly 20-fold increase relative to that of the bare film at a fixed potential of −1.0 V_RHE. Our results showed that a high SRF and NN density is required to reach a high C₂H₄ (or C₂₊) selectivity. More importantly, post-reaction-microscopy and XAFS investigations highlight that the NNs consist of a very small porous 3D network with metallic Cu NPs. The high density of NNs with a similar inner porous structure trait is likely to increase their C₂₊ selectivity. Based on our comprehensive study, we present a relationship between the NN features and their inner structure on the selectivity of the CO₂RR, providing a better understanding of the nanostructuring process of copper thin film electrocatalysts.

Author contributions

D. P.: writing – original draft, conceptualization, investigation, formal analysis, and visualization; F. F.: methodology, visualization, and writing – review and editing; D. B.: investigation and writing – review and editing; I. A.: investigation and writing – review and editing; T. N.: writing – review and editing; D. P., S. B. and I. T.: writing – review and editing, and resources; D. G., E. J. and P. L.: writing – review and editing, project administration, and funding acquisition.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data of this article are shown in the figures and the ESI.†

The raw data shown in the figures are available on the KU Leuven institutional research data repository at https://doi.org/10.48804/I6WO1U.

Acknowledgements

This work was supported by the European Union under the MSCA Innovative Training Network CATCHY (grant agreement no. 955650) and by the Flanders Innovation & Entrepreneurship (VLAIO) project HBC.2021.0586 (CLUE). T.H.T. Nguyen thanks the Research Foundation Flanders (FWO) for a FWO MSCA SoE postdoctoral fellowship (12ZZI.23N). Filippo Franceschini acknowledges the support of the Research Foundation Flanders (FWO) for the PhD fellowship (1S61725N).

Access to Elettra synchrotron was provided by the European Union as part of the Horizon Europe call HORIZON-INFRA-2021-SERV-01 under grant agreement number 101058414 and co-funded by UK Research and Innovation (UKRI) under the UK government's Horizon Europe funding guarantee (grant number 10039728) and by the Swiss State Secretariat for Education, Research and Innovation (SERI) under contract number 22.00187: ‘Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the UK Science and Technology Facilities Council or the Swiss State Secretariat for Education, Research and Innovation (SERI). Neither the European Union nor the granting authorities can be held responsible for them.

Finally, we would like to thank Luca Olivi and the staff of the XAFS beamline at Elettra synchrotron for providing guidance and assistance. We would also like to thank Zviadi Zarkua for his assistance and discussions on XPS data.

References

1. I. E. L. Stephens, K. Chan, A. Bagger, S. W. Boettcher, J. Bonin, E. Boutin, A. K. Buckley, R. Buonsanti, E. R. Cave, X. Chang, S. W. Chee, A. H. M. Da Silva, P. De Luna, O. Einsle, B. Endrődi, M. Escudero-Escribano, J. V. Ferreira De Araujo, M. C. Figueiredo, C. Hahn, K. U. Hansen, S. Haussener, S. Hunegnaw, Z. Huo, Y. J. Hwang, C. Janáky, B. S. Jayathilake, F. Jiao, Z. P. Jovanov, P. Karimi, M. T. M. Koper, K. P. Kuhl, W. H. Lee, Z. Liang, X. Liu, S. Ma, M. Ma, H.-S. Oh, M. Robert, B. R. Cuenya, J. Rossmeisl, C. Roy, M. P. Ryan, E. H. Sargent, P. Sebastián-Pascual, B. Seger, L. Steier, P. Strasser, A. S. Varela, R. E. Vos, X. Wang, B. Xu, H. Yadegari and Y. Zhou, J. Phys.: Energy, 2022, 4, 042003.
2. Y. Y. Birdja, E. Pérez-Gallent, M. C. Figueiredo, A. J. Göttle, F. Calle-Vallejo and M. T. M. Koper, Nat. Energy, 2019, 4, 732–745.
3. D. Xu, K. Li, B. Jia, W. Sun, W. Zhang, X. Liu and T. Ma, Carbon Energy, 2023, 5, e230.
4. O. S. Bushuyev, P. De Luna, C. T. Dinh, L. Tao, G. Saur, J. Van De Lagemaat, S. O. Kelley and E. H. Sargent, Joule, 2018, 2, 825–832.
5. A. Bagger, W. Ju, A. S. Varela, P. Strasser and J. Rossmeisl, ChemPhysChem, 2017, 18, 3266–3273.
6. S. Nitopi, E. Bertheussen, S. B. Scott, X. Liu, A. K. Engstfeld, S. Horch, B. Seger, I. E. L. Stephens, K. Chan, C. Hahn, J. K. Nørskov, T. F. Jaramillo and I. Chorkendorff, Chem. Rev., 2019, 119, 7610–7672.
7. C. Zhan, F. Dattila, C. Rettenmaier, A. Herzog, M. Herran, T. Wagner, F. Scholten, A. Bergmann, N. López and B. Roldan Cuenya, Nat. Energy, 2024, 9, 1485–1496.
8. S. Y. Lee, H. Jung, N.-K. Kim, H.-S. Oh, B. K. Min and Y. J. Hwang, J. Am. Chem. Soc., 2018, 140, 8681–8689.
9. T. Saberi Safaei, A. Mepham, X. Zheng, Y. Pang, C.-T. Dinh, M. Liu, D. Sinton, S. O. Kelley and E. H. Sargent, Nano Lett., 2016, 16, 7224–7228.
10. M. C. O. Monteiro, F. Dattila, B. Hagedoorn, R. García-Muelas, N. López and M. T. M. Koper, Nat. Catal., 2021, 4, 654–662.
11. H. Mistry, A. S. Varela, C. S. Bonifacio, I. Zegkinoglou, I. Sinev, Y.-W. Choi, K. Kisslinger, E. A. Stach, J. C. Yang, P. Strasser and B. R. Cuenya, Nat. Commun., 2016, 7, 12123.
12. J.-J. Velasco-Vélez, T. Jones, D. Gao, E. Carbonio, R. Arrigo, C.-J. Hsu, Y.-C. Huang, C.-L. Dong, J.-M. Chen, J.-F. Lee, P. Strasser, B. Roldan Cuenya, R. Schlögl, A. Knop-Gericke and C.-H. Chuang, ACS Sustainable Chem. Eng., 2019, 7, 1485–1492.
13. M. Favaro, H. Xiao, T. Cheng, W. A. Goddard, J. Yano and E. J. Crumlin, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 6706–6711.
14. A. Eilert, F. Cavalca, F. S. Roberts, J. Osterwalder, C. Liu, M. Favaro, E. J. Crumlin, H. Ogasawara, D. Friebel, L. G. M. Pettersson and A. Nilsson, J. Phys. Chem. Lett., 2017, 8, 285–290.
15. P. De Luna, R. Quintero-Bermudez, C.-T. Dinh, M. B. Ross, O. S. Bushuyev, P. Todorović, T. Regier, S. O. Kelley, P. Yang and E. H. Sargent, Nat. Catal., 2018, 1, 103–110.
16. Y. Yang, S. Louisia, S. Yu, J. Jin, I. Roh, C. Chen, M. V. Fonseca Guzman, J. Feijóo, P.-C. Chen, H. Wang, C. J. Pollock, X. Huang, Y.-T. Shao, C. Wang, D. A. Muller, H. D. Abruña and P. Yang, Nature, 2023, 614, 262–269.
17. M. Löffler, K. J. J. Mayrhofer and I. Katsounaros, J. Phys. Chem. C, 2021, 125, 1833–1838.
18. J. Timoshenko, A. Bergmann, C. Rettenmaier, A. Herzog, R. M. Arán-Ais, H. S. Jeon, F. T. Haase, U. Hejral, P. Grosse, S. Kühl, E. M. Davis, J. Tian, O. Magnussen and B. Roldan Cuenya, Nat. Catal., 2022, 5, 259–267.
19. M. Ma, K. Djanashvili and W. A. Smith, Angew. Chem., Int. Ed., 2016, 55, 6680–6684.
20. W. J. Stępniowski, H. Yoo, J. Choi, P. Chilimoniuk, K. Karczewski and T. Czujko, Surf. Interfaces, 2019, 14, 15–18.
21. J. Ambrose, R. G. Barradas and D. W. Shoesmith, J. Electroanal. Chem. Interfacial Electrochem., 1973, 47, 47–64.
22. A. Brudzisz, D. Giziński, M. Liszewska, E. Wierzbicka, U. Tiringer, S. A. Taha, M. Zając, S. Orzechowska, B. Jankiewicz, P. Taheri and W. J. Stępniowski, Electrochim. Acta, 2023, 443, 141918.
23. R. Garg, S. Gonuguntla, S. Sk, M. S. Iqbal, A. O. Dada, U. Pal and M. Ahmadipour, Adv. Colloid Interface Sci., 2024, 330, 103203.
24. N. S. Gultom, Q. Ha, M. Z. Silitonga and D. Kuo, ChemCatChem, 2024, 16, e202301312.
25. B. Barbee, B. Muchharla, A. Adedeji, A. Karoui, K. Kumar Sadasivuni, M. S. Sha, A. M. Abdullah, G. Slaughter and B. Kumar, Sci. Rep., 2022, 12, 7507.
26. M. S. Abdel-wahab, H. K. E. Emam and W. M. A. E. Rouby, RSC Adv., 2023, 13, 10818–10829.
27. D. W. Shoesmith, T. E. Rummery, D. Owen and W. Lee, J. Electrochem. Soc., 1976, 123, 790.
28. W. J. Stępniowski, D. Paliwoda, S. T. Abrahami, M. Michalska-Domańska, K. Landskron, J. G. Buijnsters, J. M. C. Mol, H. Terryn and W. Z. Misiolek, J. Electroanal. Chem., 2020, 857, 113628.
29. Q. Lei, H. Zhu, K. Song, N. Wei, L. Liu, D. Zhang, J. Yin, X. Dong, K. Yao, N. Wang, X. Li, B. Davaasuren, J. Wang and Y. Han, J. Am. Chem. Soc., 2020, 142, 4213–4222.
30. G. Iijima, T. Inomata, H. Yamaguchi, M. Ito and H. Masuda, ACS Catal., 2019, 9, 6305–6319.
31. F. Franceschini, C. Fernandes, K. Schouteden, J. Ustarroz, J. P. Locquet and I. Taurino, Catal. Sci. Technol., 2025, 15, 3022–3035.
32. A. D. Cicco, G. Aquilanti, M. Minicucci, E. Principi, N. Novello, A. Cognigni and L. Olivi, J. Phys.: Conf. Ser., 2009, 190, 012043.
33. S. D. Giri and A. Sarkar, J. Electrochem. Soc., 2016, 163, H252–H259.
34. D. W. Shoesmith, J. Electroanal. Chem., 1983, 143, 153–165.
35. I. Platzman, R. Brener, H. Haick and R. Tannenbaum, J. Phys. Chem. C, 2008, 112, 1101–1108.
36. X. Wu, H. Bai, J. Zhang, F. Chen and G. Shi, J. Phys. Chem. B, 2005, 109, 22836–22842.
37. S. Anantharaj, H. Sugime, S. Yamaoka and S. Noda, ACS Appl. Energy Mater., 2021, 4, 899–912.
38. K. P. Kuhl, E. R. Cave, D. N. Abram and T. F. Jaramillo, Energy Environ. Sci., 2012, 5, 7050.
39. M. C. Biesinger, Surf. Interface Anal., 2017, 49, 1325–1334.
40. M. Pourbaix, Corros. Sci., 1974, 14, 25–82.
41. W. Stepniowski and W. Misiolek, Nanomaterials, 2018, 8, 379.
42. R. Kas, R. Kortlever, H. Yılmaz, M. T. M. Koper and G. Mul, ChemElectroChem, 2015, 2, 354–358.
43. E. L. Clark, J. Resasco, A. Landers, J. Lin, L.-T. Chung, A. Walton, C. Hahn, T. F. Jaramillo and A. T. Bell, ACS Catal., 2018, 8, 6560–6570.
44. Y. Pang, T. Burdyny, C.-T. Dinh, M. G. Kibria, J. Z. Fan, M. Liu, E. H. Sargent and D. Sinton, Green Chem., 2017, 19, 4023–4030.
45. M. C. O. Monteiro, A. Mirabal, L. Jacobse, K. Doblhoff-Dier, S. C. Barton and M. T. M. Koper, JACS Au, 2021, 1, 1915–1924.
46. E. L. Clark and A. T. Bell, J.Am. Chem. Soc., 2018, 140, 7012–7020.
47. M. Ma, K. Djanashvili and W. A. Smith, Angew. Chem., Int. Ed., 2016, 55, 6680–6684.
48. K.-L. C. Nguyen, J. P. Bruce, A. Yoon, J. J. Navarro, F. Scholten, F. Landwehr, C. Rettenmaier, M. Heyde and B. R. Cuenya, ACS Energy Lett., 2024, 9, 644–652.
49. S. J. Raaijman, N. Arulmozhi and M. T. M. Koper, ACS Appl. Mater. Interfaces, 2021, 13, 48730–48744.
50. T. Burdyny, P. J. Graham, Y. Pang, C.-T. Dinh, M. Liu, E. H. Sargent and D. Sinton, ACS Sustainable Chem. Eng., 2017, 5, 4031–4040.
51. Q. Lei, L. Huang, J. Yin, B. Davaasuren, Y. Yuan, X. Dong, Z.-P. Wu, X. Wang, K. X. Yao, X. Lu and Y. Han, Nat. Commun., 2022, 13, 4857.
52. J. Vavra, G. P. L. Ramona, F. Dattila, A. Kormányos, T. Priamushko, P. P. Albertini, A. Loiudice, S. Cherevko, N. Lopéz and R. Buonsanti, Nat. Catal., 2024, 7, 89–97.
53. P. Wilde, P. B. O'Mara, J. R. C. Junqueira, T. Tarnev, T. M. Benedetti, C. Andronescu, Y. T. Chen, R. D. Tilley, W. Schuhmann and J. J. Gooding, Chem. Sci., 2021, 12, 4028–4033.
54. H. R. Oswald, A. Reller, H. W. Schmalle and E. Dubler, Acta Crystallogr., Sect. C:Cryst. Struct. Commun., 1990, 46, 2279–2284 . Accessed from https://www.crystallography.net/cod/9007849.html on 26 June 2024.
55. A. D. Fortes, M. H. Lemée-Cailleau, C. J. Pickard and R. J. Needs, J. Am. Chem. Soc., 2009, 131, 13508–13515.
56. A. Kirfel and K. Eichhorn, Acta Crystallogr., Sect. A:Found. Crystallogr., 1990, 46, 241–284 . Accessed from https://www.crystallography.net/cod/1000063.html on 26 June 2024.
57. C. T. Dinh, T. Burdyny, M. G. Kibria, A. Seifitokaldani, C. M. Gabardo, F. P. García De Arquer, A. Kiani, J. P. Edwards, P. De Luna, O. S. Bushuyev, C. Zou, R. Quintero-Bermudez, Y. Pang, D. Sinton and E. H. Sargent, Science, 2018, 360, 783–787.
58. J. Zhang, T. H. My Pham, Z. Gao, M. Li, Y. Ko, L. Lombardo, W. Zhao, W. Luo and A. Züttel, ACS Catal., 2023, 13, 9326–9335.
59. A. Dutta, M. Rahaman, N. C. Luedi, M. Mohos and P. Broekmann, ACS Catal., 2016, 6, 3804–3814.
60. J. R. C. Junqueira, P. B. O'Mara, P. Wilde, S. Dieckhöfer, T. M. Benedetti, C. Andronescu, R. D. Tilley, J. J. Gooding and W. Schuhmann, ChemElectroChem, 2021, 8, 4848–4853.
61. S. Louisia, D. Kim, Y. Li, M. Gao, S. Yu, I. Roh and P. Yang, Proc. Natl. Acad. Sci. U. S. A., 2022, 119, e2201922119.
62. W. Ren, H. Zhang, M. Chang, N. Chen, W. Ma, J. Gu, M. Lin and X. Hu, Chem, 2025, 11, 102352.
63. J. Yu, Y. Zheng, B. Lv, A. Huang, J. Zhang, Z. Wang, Y. Zhang, Y. Wu, Y. Zhou, Y. Wang and W. Luo, Appl. Catal. B: Environ., 2025, 368, 125131.
64. L. Bian, Y. Bai, J.-Y. Chen, H.-K. Guo, S. Liu, H. Tian, N. Tian and Z.-L. Wang, ACS Nano, 2025, 19, 9304–9316.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5nr01514f

---