Mechanistic pathways of copper catalyst reconstruction in CO2 electroreduction

Mechanistic pathways of copper catalyst reconstruction in CO2 electroreduction | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Mechanistic pathways of copper catalyst reconstruction in CO2 electroreduction Qi Hao, Dexiang Chen, Guoqun Li, Xing Zhao, Yunjia Wei, Shuying Chen, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7787217/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Surface reconstruction is a common and critical process in copper catalysts during electrochemical CO2 reduction reaction (CO2RR), continuously reshaping their structures and compositions into states distinct from the pristine material. Although recognized as central to catalyst performance, the mechanistic pathways and chemical identities of the dissolved Cu species remain unclear. Employing clean and well-defined Cu nanoparticle arrays as a model platform, we directly visualize morphological evolution and simultaneously track reaction intermediates by in situ surface-enhanced Raman spectroscopy (SERS). We resolve three distinct dissolution-redeposition pathways of Cu catalyst: (i) *CO-mediated Cu+ dissolution and facet-selective redeposition at moderate potentials; (ii) field-assisted leaching of neutral Cu0 species at strongly negative potentials; and (iii) oxidative Cu+ dissolution at open-circuit potential, followed by redeposition into Cu2O cubes enabled by CO2RR-generated active sites. These findings reveal a nonlinear, potential-dependent behavior of Cu reconstruction, and provide guiding principles for directing catalyst evolution through control of adsorbates, potential windows, and bias protocols. Physical sciences/Chemistry/Catalysis/Electrocatalysis Physical sciences/Chemistry/Catalysis/Heterogeneous catalysis Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction The electrochemical CO 2 reduction reaction (CO 2 RR) provides a feasible route for converting CO 2 into valuable chemicals and fuels under mild conditions. Among available catalysts, copper (Cu) stands out for its moderate binding affinity toward key reaction intermediates, enabling the formation of C 2+ products 1 . This process is highly surface-sensitive, with reactivity strongly influenced by the oxidation state, morphology, and local microenvironment of the surface catalyst 2–4 . As a result, even minor variations in these factors can lead to substantial changes in overall product selectivity. Therefore, controlling and understanding the state of the Cu catalyst during CO 2 RR is fundamentally important for both mechanistic elucidation and catalyst optimization. Owing to its relatively low cohesive energy and high surface-atom mobility, Cu is intrinsically susceptible to surface reconstruction under electrochemical operation 5,6 . This can lead to spontaneous atomic rearrangements and compositional transformations of Cu catalyst, reshaping surface morphology and altering the distribution of active sites 7,8 . Consequently, the actual working state of Cu catalyst may differ substantially from its as-prepared or post-reaction form, which can lead to misinterpretation of catalytic mechanisms. Therefore, a detailed understanding of the processes governing surface reconstruction is essential for accurately elucidating catalytic behavior in CO 2 RR. Although surface reconstruction of Cu has been widely reported, its dynamics and underlying mechanisms remain under debate. Some studies have observed rapid structural reshaping within the first few minutes of electrolysis 9,10 , whereas others report continuous transformations throughout the reaction 11,12 . While some attribute CO 2 RR activity to the reduction of Cu oxides into metallic Cu 13–15 , others argue that oxidic species may directly contribute to catalysis 16–18 . Moreover, morphologically diverse Cu catalysts often exhibit similar reconstruction trends, for example, the appearance of secondary particles with a typically cubic shape 19,20 . However, recent evidence suggests that such features form predominantly after the removal of bias in the air, rather than under operando conditions 21,22 . Taken together, these observations imply that Cu surface reconstruction does not proceed through a single pathway, but instead arises from a complex interplay of surface adsorbates, applied bias, catalytic environment, and intrinsic thermodynamic driving forces. Building on these complexities, multiple reconstruction mechanisms have been proposed, including nanoparticle aggregation 23,24 , cathodic corrosion 25,26 , and dissolution-redeposition mechanisms 27–29 . Among these, the latter is increasingly recognized as a central pathway for surface evolution under CO 2 RR conditions. Specifically, in situ transmission electron microscopy (TEM) has revealed amorphous-crystalline transitions in Cu catalysts 30 , accompanied by the formation of Cu + species as confirmed by UV-Vis spectroscopy 31 . Complementary density functional theory (DFT) calculations and Raman analysis further suggest that adsorbed *CO intermediates play a central role in triggering and regulating these transformations. *CO can reshape the surface energies of Cu through its facet-dependent binding affinities 32 , inducing selective dissolution and redeposition of Cu species on specific facets and thereby guiding the reconstruction pathways 33 . Despite growing evidence, several critical questions remain unresolved. First, what is the chemical nature of the dissolved Cu species, and how do applied potential and adsorbed intermediates influence their dissolution pathways? Second , how can we distinguish surface changes occurring under operando CO 2 RR conditions from those emerging during the post-reaction stage, and thereby minimize interference from the latter in mechanistic studies? Third, to what extent does surface reconstruction reshape the catalyst surface and direct its evolution under CO 2 RR conditions? In this study, we employ clean and structurally well-defined Cu nanoparticle (NP) arrays as a model system to investigate surface reconstruction during CO 2 RR. This uniform, chemical-free platform enables direct visualization of morphological evolution, while its strong plasmonic response allows in situ and real-time tracking of interfacial intermediates through surface-enhanced Raman spectroscopy (SERS). By integrating this platform with complementary techniques, we reveal potential-dependent reconstruction behavior and resolve three distinct dissolution-redeposition pathways of Cu catalysts. Importantly, we identify the chemical nature of the dissolved species, revealing not only Cu⁺ but also previously unrecognized Cu 0 species, as a function of applied potential. Together, these findings help clarify the open questions outlined above, and provide new experimental evidence and perspectives on how surface reconstruction may be understood and potentially regulated under CO 2 RR conditions. Results Preparations and advantages of the Cu NP arrays Cu NP arrays with a centimeter scale (Supplementary Fig. 1) were fabricated to investigate the surface reconstruction. As illustrated in Figure 1a, patterned Cu NPs were prepared using porous anodic aluminum oxide (AAO) membrane as a mask during e-beam evaporation 34,35 . Specifically, AAO membrane with a pore diameter of ~80 nm was transferred onto a conductive substrate (Figure 1b), followed by Cu deposition through the AAO mask with e-beam evaporation (Figure 1c). After deposition, the mask was removed, exposing the resulting Cu NP arrays (Figure 1d and 1e). Scanning electron microscopy (SEM) images confirm the uniformity and hexagonal periodicity of the Cu NP arrays. The fabricated Cu NPs exhibit a face-centered cubic (fcc) lattice of metallic Cu, and were stored in vacuum prior to use. This fabrication approach offers several advantages for the investigation of surface reconstruction. First, the AAO pore diameter is intrinsically determined by the anodization voltage, ensuring precise size control and physically defined reproducibility across samples, batches, and laboratories 36 . This offers good structural uniformity and reproducibility for the Cu NPs, enabling clear identification of reconstruction-induced surface evolution. Second, the size tunability allows optimization of the localized surface plasmon resonance (LSPR), which in turn amplifies the intrinsically weak Raman signals of CO 2 RR intermediates via SERS 37 . Third, the Cu NPs prepared by physical vapor deposition are free from chemical surfactants and ligands, and remain structurally stable under reactive conditions, thereby minimizing particle aggregation and chemical interference and providing a clean model for mechanistic studies 38 . Collectively, these features provide a robust platform for tracking catalyst structural evolution during CO 2 RR while simultaneously probing intermediates by SERS for mechanistic insights. Operando and post-electrochemical morphological changes SEM was used to track the morphological evolution of the Cu NPs at different applied potentials (Figure 2a). Two distinct reconstruction stages were identified: Stage A , representing operando reconstruction under cathodic bias during CO 2 RR, and Stage B , representing post-reconstruction occurring spontaneously in CO 2 -saturated electrolyte at open-circuit potential (OCP, ~+0.4 V). The main morphological pathways are schematically summarized in Figure 2b. In Stage A : two major morphological changes were observed, both exhibiting a non-linear dependence on applied potential: (1) Progressive shrinkage of the original Cu NPs. The initially hemispherical Cu NPs gradually decreased in size. This shrinkage was minimal at -0.1 V, pronounced at moderate potentials (-0.3 V to -0.5 V), and progressively suppressed under more negative conditions. (2) Emergence of secondary particles. The secondary particles were rarely detected at both low and high potentials (-0.1 V and -0.9 V), but evident at moderate values. Two types of secondary particles were identified. Cubic-like domains were most prominent at -0.5 V and -0.3 V, while smaller satellite particles formed at -0.5 V and -0.7 V. These secondary particles were not observed in control experiments in the absence of CO 2 . The non-linear potential dependence, along with simultaneous observation of parent-particle shrinkage and secondary-particle formation in Stage A , provides direct evidence for a dissolution-redeposition mechanism. In this process, the original Cu NPs act as atom sources, releasing Cu species that redeposit into secondary particles. Notably, even at -0.9 V, where secondary particle formation was suppressed, the shrinkage persisted (see also Supplementary Fig. 2), suggesting that the dissolution can persist even when redeposition is kinetically hindered. We therefore propose two distinct dissolution pathways: (i) an electrochemically driven route governed by the interfacial environment, where surface intermediates govern the dissolution and account for the non-linear potential dependence of redeposition, and (ii) a field-assisted dissolution route that becomes increasingly dominant at high cathodic bias. We suggest that the Cu species released through these two pathways are chemically distinct, and this will be further examined in the following sections. Upon entering Stage B after bias removal, the parent Cu NPs underwent further shrinkage together with reconstruction into cubes. This process proceeded spontaneously until the parent Cu was fully consumed, suggesting a surface-energy-driven ripening process governed by the electrolyte environment 23,29 . The newly formed particles appeared brighter in SEM images and exhibited sharper edges compared with the parent Cu NPs, indicative of the formation of Cu oxides. To assess the generality of this behavior, we extended our study to evaporated and sputtered planar Cu film electrodes (Figs. S3-S7). All tested systems displayed similar morphological evolution, confirming that the cube formation in Stage B is intrinsically surface-energy driven. The oxidation state of the Cu catalyst was examined by in situ Raman spectroscopy (Figure 2c). Under applied bias, the native oxide modes disappeared, suggesting electrochemical reduction to metallic Cu. Upon bias removal, these peaks gradually reappeared, suggesting a rapid oxidation in Stage B . X-ray diffraction (XRD) results in Figure 2d support this observation: after electrolysis, only Cu(111) and weak Cu(100) reflections of fcc Cu were observed, whereas after 30 min at OCP, distinct Cu 2 O reflections emerged. Combined with the SEM (Figure 2a) and TEM data (Figs. S8 and S9), the results suggest that the cubic structures observed at OCP are predominantly composed of single crystal Cu 2 O. Interestingly, the formation of Cu 2 O cubes occurs at OCP but requires prior CO 2 RR exposure (Supplementary Fig. 10). Further analysis indicates that sufficient CO 2 RR duration, rather than the applied potential, appears to be essential to generate the conditions that facilitate cubic growth during OCP (Figs. S2-S6). We attribute this to the formation of active sites on Cu catalyst during CO 2 RR, such as defect-rich domains or intermediate-modified terraces (Supplementary Fig. 11) 21 . Notably, the formation of these Cu 2 O cubes proceeds in the electrolyte, in contrast to earlier reports where air exposure is required (Supplementary Fig. 12) 22 . Investigation of the dissolution-redeposition pathways To identify the chemical nature of the released Cu + species, we performed in situ UV-Vis spectroscopy using a custom-designed electrochemical cell (schematic in Supplementary Fig. 13). 2,9-dimethyl-1,10-phenanthroline (dmphen) was introduced as a selective chelating agent, which can form a stable and soluble complex with Cu + , yielding a characteristic absorption band at ~455 nm 31,39,40 . Spectra were recorded across a potential range from -0.1 V to -0.9 V, with -0.5 V (Figure 3a) and -0.9 V (Figure 3b) taken as representative cases (full dataset in Supplementary Fig. 14). At -0.5 V, the 455 nm band appeared rapidly during CO 2 RR, confirming the release of Cu + into the electrolyte. In the presence of dmphen, the shrinkage of parent particles persisted, yet the formation of secondary nanoparticles was remarkably suppressed (Supplementary Fig. 15), confirming that dissolved Cu⁺ attributes to secondary-particle redeposition in Stage A. As shown in Figure 3d, the 455 nm Cu + band strengthened from -0.1 V to -0.5 V, but weakened at -0.7 V, and was barely detectable at -0.9 V. This potential-dependent behavior aligns with the morphological evolution observed by SEM in Figure 2, further supporting a Cu + dissolution-redeposition pathway. At -0.9 V (Figure 3b), the 455 nm Cu + -dmphen band was not detected, and distinct sub-400 nm features emerged. These signals are attributed to quantum-confined optical transitions of Cu 0 clusters (TEM, Supplementary Fig. 16) 41,42 , suggesting that neutral Cu 0 atoms were released and subsequently aggregated into solution-phase clusters. While previous studies reported *CO-induced Cu clusters on catalyst surfaces via a Cu⁺ dissolution-reduction pathway 28,43 , our results suggest that Cu clusters also form in solution and are composed of Cu 0 atoms, revealing a distinct mechanism. Importantly, these sub-400 nm features were also observed in the absence of dmphen (Figure 3c and Supplementary Fig. 17), confirming that the release of Cu 0 species represents an intrinsic dissolution pathway. Additionally, the Cu content in electrolyte at -0.9 V was higher than at other potentials (Supplementary Fig. 18), nevertheless, redeposition into secondary particles was not observed at this potential (Figure 2a, A3). Moreover, we continued to monitor these samples during the subsequent OCP stage (Figure 3e and Supplementary Fig. 19). The 455 nm Cu + band increased continuously during OCP, suggesting sustained Cu + release into the electrolyte. According to the Cu Pourbaix diagram, metallic Cu is thermodynamically metastable at OCP and undergoes spontaneous oxidation into Cu + , initiating rapid oxidative dissolution 44–46 . The release of Cu + at OCP also exhibited a potential dependence, pronounced at moderate potentials but less significant at the extremes. This potential dependence did not correlate with the morphological trends observed by SEM (Figure 2a, B3), implying that factors beyond Cu + release govern the redeposition process. Notably, a pristine control sample without CO 2 RR pretreatment also exhibited pronounced Cu + release at OCP (red line). As shown in Figure 3f, the introduction of dmphen can suppress the redeposition into Cu 2 O cubes at OCP a s expected , suggesting a dissolution-redeposition pathway. Interestingly, for the pristine control sample, even without dmphen, no Cu 2 O cubes were observed despite clear Cu + release . Therefore, we deduce that active sites generated during CO 2 RR, like defect-rich or intermediate-modified surfaces, enable Cu 2 O cube nucleation during OCP. Together, these findings reveal distinct dissolution-redeposition pathways under different conditions. During CO 2 RR: (i) at low and moderate reductive potentials, *CO-induced Cu + dissolution dominates, giving rise to redeposition into metallic secondary nanoparticles; (ii) at more negative potentials, bias-driven dissolution of metallic Cu 0 species becomes predominant, while redeposition into secondary nanoparticles is suppressed; and at OCP: (iii) surface Cu atoms spontaneously oxidize to Cu + , which undergo oxidative dissolution and subsequent redeposition into Cu 2 O cubes through an Ostwald-like ripening process. Crucially, prior CO 2 RR treatment is essential to generate active nucleation sites that initiate the growth of Cu 2 O cubes . Intermediate-induced Cu reconstruction in CO 2 RR In situ Raman spectroscopy was performed to monitor surface intermediates at different potentials. Our fabrication strategy allows precise design of the Cu NPs to optimize their SERS efficiency, allowing ultrasensitive detection of the intrinsically weak Raman signals of the intermediates (Figs. S20 and S21). As shown in Figure 4a, a peak assigned to HCO 3 - was initially observed at 1016 cm -1 at OCP. This peak disappeared at -0.1 V, where a strong peak at 1068 cm -1 corresponding to CO 3 2- emerged. Simultaneously, bands at 2000-2100 cm -1 assigned to adsorbed *CO appeared. Upon lowering the potential from -0.1 to -0.6 V, a peak at 1386 cm -1 assigned to *COOH, and another peak at 1440 cm -1 assigned to *OCHO, were observed. At more negative potentials, the *OCHO signal weakened, while a new band at 530 cm -1 assigned to *CH 2 CHO emerged, suggesting dimerization between *CO species 37,47 (Table S1). This trend aligns with the decrease in *COOH intensity, indicating its rapid conversion to *CO and a consequent increase in *CO surface coverage. At -1.0 V, Raman acquisition was hindered by vigorous bubble formation. As shown in Figure 4b, the adsorption energies of key intermediates on Cu were calculated using DFT, among which *CO exhibits the strongest binding (see also Figs. S22 and S23, and Table S2). This strong binding localizes electrons into the Cu-*CO bonds and disrupts the lateral Cu-Cu bonding network, thereby facilitating atomic rearrangement, as evidences by Bader charge analysis (Supplementary Fig. 24). Moreover, as revealed in Figure 4b, the binding strength of *CO is facet-dependent: stronger on Cu(100) and Cu(110) but weaker on Cu(111), consistent with previous reports 48 . Consequently, at low *CO coverage, *CO preferentially adsorbs on Cu(100) and Cu(110), lowering their surface energies relative to Cu(111). As shown in Figure 4c, although Cu(111) intrinsically has the lowest surface energy, the presence of *CO shifts the equilibrium, offering a thermodynamic driving force that favors reconstruction from Cu(111) toward Cu(100) and Cu(110). On the other hand, at high *CO coverage, which corresponds to more negative potentials, *CO adsorption on Cu(100) and Cu(110) approaches saturation, while adsorption on Cu(111) increase markedly. Under these conditions, the surface energy of Cu(111) also decreases and the thermodynamic driving force for reconstruction toward Cu(100) and Cu(110) is weakened. This explains the observed suppression of redeposition at -0.9 V during Stage A (Figure 2a), where high *CO coverage leads to a loss of facet preference. The proposed mechanism of facet-dependent reconstruction under varying *CO coverage, corresponding to CO 2 RR at -0.5 V and -0.9 V, was experimentally verified through facet-specific OH - adsorption analysis in cyclic voltammetry (Figure 4d and Figs. S25 and S26). The results reveal that reconstruction from Cu(111) toward Cu(100) and Cu(110) was pronounced at -0.5 V but significantly suppressed at -0.9 V, consistent with the DFT calculations. Comparison with Ar-saturated conditions (Supplementary Fig. 26) further confirms that this facet reconstruction is driven by CO 2 RR, not directly by the applied potential. To experimentally validate this hypothesis, we introduced 18-crown-6 (18-C-6), a macrocyclic ether known to enhance the binding affinity of *CO on metal surfaces, to modulate the surface coverage of *CO 49 . The Cu-CO stretching bands at ~280 cm -1 and ~360 cm -1 were used as indicators of *CO coverage. As shown in Figure 4e and 4f, in standard CO 2 RR experiments without 18-C-6, the bands indicative of high *CO coverage appeared only at relatively negative potentials, beginning at -0.7 V. In contrast, in the presence of 18-C-6, these bands were significantly enhanced and emerged at less negative potentials, starting from -0.4 V. These results confirm that 18-C-6 effectively increased *CO coverage on Cu, particularly under unsaturated conditions at relatively low potentials. Correspondingly, upon introduction of 18-C-6, both shrinkage of the parent particle and the formation of secondary particles were markedly suppressed at -0.5 V (Figure 4g, 4h and Supplementary Fig. 27). This finding supports our view that the reconstruction from Cu(111) toward Cu(100) and Cu(110) is favored under low *CO coverage but inhibited under high *CO coverage. Discussion In summary, the use of clean and structurally well-defined Cu NP arrays enables direct visualization of catalyst morphology and simultaneous in situ SERS detection of reaction intermediates, providing a robust platform to elucidate the complex pathways of surface reconstruction during CO 2 RR. Importantly, Cu restructuring during operando conditions follows a nonlinear, potential-dependent behavior. Through the integration of SEM, UV-Vis, in situ Raman, and DFT studies, we provide direct experimental evidence that dissolution-redeposition is a central mechanism governing Cu restructuring, proceeding through three distinct pathways depending on the electrochemical environment: (i) At low and moderate reductive potentials, *CO-mediated Cu + dissolution dominates and drives redeposition into metallic secondary particles. *CO coverage acts as a molecular switch: at low *CO overage, *CO adsorption drives facet-selective reconstruction from Cu(111) toward Cu(100) and Cu(110); while at high coverage, this driving force vanishes. (ii) At more negative potentials, field-assisted dissolution of metallic Cu 0 becomes predominant, resulting in particle shrinkage without redeposition. In this case, neutral Cu 0 species are leached from the surface and aggregate into small soluble clusters. (iii) Upon bias removal, the dominant mechanism shifts to oxidative dissolution. Cu undergoes oxidative dissolution to Cu⁺, which subsequently redeposits as Cu 2 O cubes. This transformation requires CO 2 RR pretreatment, likely due to the CO 2 RR-generated active nucleation sites. These Cu 2 O cubes resemble the metallic secondary cubes observed under operando CO 2 RR but form far more rapidly and extensively, underscoring the risk of misidentifying post-reaction products as active catalytic states. Beyond mechanistic insights, our findings suggest practical strategies for rational catalyst design. By modulating *CO coverage, potential windows, and bias protocols, dissolution-redeposition pathways can be selectively promoted or suppressed. Although a comprehensive understanding of how reconstruction feeds back into catalytic activity is not yet established, our results provide indicative evidence toward this question. 1) Our results point to an intriguing possibility that, at least for fcc Cu catalysts, although reconstruction is inevitable, its direct impact on CO 2 RR efficiency may be less significant than often assumed. At strongly negative potentials, where *CO coverage is high and CO 2 RR activity is maximized, morphological and facet reconstruction appear limited within the experimental timeframe and thus exert only a minor influence on catalytic performance. 2) Beyond morphology, the nature of dissolved species may also play an important role in CO 2 RR. Dissolution-redeposition is now widely recognized as a central mechanism of Cu reconstruction. Traditionally, the dissolved species were assumed to be Cu + only, which is readily reduced under CO 2 RR conditions and thus considered irrelevant to catalytic activity. Our observation of dissolved Cu 0 species challenges this view and introduces a critical question: could neutral Cu species directly influence CO 2 RR activity and selectivity? Addressing these issues, alongside a broader understanding of operando and post-reaction reconstructions, is crucial for establishing accurate structure-performance relationships and for guiding the design of future CO 2 RR catalysts that intelligently exploit or mitigate reconstruction. Methods Materials and methods Copper (Cu, 99.999%) and Titanium (Ti, 99.999%) were purchased from Zhong Nuo Advanced Material Technology Co., Ltd (Beijing, China). Potassium bicarbonate (KHCO 3 , ≥99.5%), copper chloride dihydrate (CuCl 2 •2H 2 O, 99%, Aladdin), acetone (≥99.5%, Aladdin), 2,9-dimethyl-phenanthroline (99%, Macklin), phosphoric acid (H 3 PO 4 , ≥99%, Aladdin) and polymethyl methacrylate (PMMA, 950PMMA A4, MicroChem) were used as received without further purification. Preparation of the AAO membranes Anodic aluminum oxide (AAO) samples were produced on aluminum foils using a standard anodization protocol by commercial companies (TopMembranes Technology Ltd. and Micron Technology Co., Ltd.). The purchased AAO samples were etched in 5 wt% phosphoric acid at 30 °C for 29 min to enlarge the pore diameter, followed by spin-coating with PMMA at 1200 rpm for 60 s, and baked at 120 ℃ for 20 min. The samples were subsequently immersed in 5 wt% NaOH for 20 min and then transferred to 10 wt% CuCl 2 solution to remove the aluminum substrate. The obtained AAO membranes were treated in a mixed solution of 5 wt% phosphoric acid and 2 wt% CuCl 2 for 37 min to remove the AAO barrier layer and adjust the pore diameter. Afterwards, the PMMA protective layer was removed in acetone, and the resulting free-standing porous AAO membranes were rinsed thoroughly with distilled water before transfer onto substrates. Electrochemical Measurements Electrochemical CO 2 RR experiments were performed in an H-type cell using a conventional three-electrode configuration. A Nafion 117 membrane was used to separate the cathodic and anodic compartments and served as a proton exchange membrane. All measurements were conducted using a CHI660E electrochemical workstation (CH Instruments). A platinum wire and an Ag/AgCl (3 M KCl) electrode were employed as the counter and reference electrodes, respectively. Prior to each experiment, the electrolyte (0.1 M KHCO 3 ) was purged with high-purity CO 2 gas at a flow rate of 10 sccm for at least 15 minutes to remove dissolved oxygen and saturate the solution. Potential was converted following the relationship: E ( vs. RHE) = E ( vs. Ag/AgCl) + 0.197 + 0.0592*pH. In situ UV-Vis Measurement The UV-Vis absorption spectra were collected using a HITACHI U-3900 spectrometer equipped with a customized electrochemical liquid-cell (Supplementary Fig. 14). The measurements were conducted under CO 2 RR conditions, using a three-electrode configuration with Cu NPs array as the working electrode, Ag/AgCl as the reference electrode, and a platinum wire as the counter electrode. Spectra were collected with a scan speed of 120 nm/min, sampling interval of 0.50 nm, and slit width of 1 nm. Sample size: 1.0 cm 2 . In situ Raman Measurement In situ Raman measurements were performed in a custom-designed in situ Raman cell (Gaossunion Photoelectric Technology Co., Ltd., Tianjin, China) coupled to an electrochemical workstation. A three-electrode system was employed with a Cu NPs working electrode, Ag/AgCl reference electrode, and platinum wire counter electrode. Raman spectra were acquired using a HORIBA Jobin Yvon LabRAM HR Evolution system with a 633 nm laser (2 mW power) and a 600/mm grating. The integration time was 10 s per spectrum. Electrochemical OH- Adsorption Electrochemical adsorption of OH - was performed on the Cu NPs catalysts. The reaction follows: Cu + OH - → Cu(OH) ad + e - . Measurements were carried out in Ar-saturated 0.1 M KOH solution using a Pt counter electrode and an Hg/HgO reference electrode. All potentials were converted to RHE scale following the equation: E ( vs. RHE) = E ( vs. Hg/HgO) + 0.098 + 0.059*pH. The fractional surface coverage of Cu(111), Cu(100), and Cu(110) facets was determined by normalizing the OH - adsorption charge to the corresponding charge densities on single-crystal Cu electrodes: 2.16 μC cm -2 for Cu(111), 13.3 μC cm -2 for Cu(110), and 8.22 μC cm -2 for Cu(100), as reported in previous studies 33,50 . DFT Calculations DFT calculations were performed using the Quantum ESPRESSO package. The exchange-correlation interactions were treated within the generalized gradient approximation (GGA) using the Perdew-Burke-Ernzerhof (PBE) functional, and the projector-augmented wave (PAW) method was employed. Van der Waals interactions were included via Grimme’s DFT-D3 correction. The Kohn-Sham orbitals were expanded in a plane-wave basis set with kinetic energy and charge density cutoffs of 50 and 400 Ry, respectively. Brillouin-zone integrations were sampled with a (3 × 3 × 1) Monkhorst-Pack k-point mesh. The energy convergence threshold for electronic self-consistency was set to 10 -4 eV, and structural optimizations were performed until the residual forces on each atom were less than 10 -3 eV Å -1 . A vacuum spacing of 15 Å was applied to prevent interactions between periodic images. Other Characterizations The morphology and elemental composition were studied by a field-emission SEM (FEI Inspect F50) and a TEM (Talos F200X) equipped with an energy dispersive X-ray spectrometer (EDS) system. Crystallographic information was obtained from XRD (Rigaku smartlab (3), Cu- K α radiation with λ = 0.15406 nm) with a 2 θ range from 20° to 60° at a scan rate of 10° min -1 . ICP-MS data were obtained by an Aglient 7850 instrument. Declarations Data availability The pristine computational models are provided as Supplementary Information. All other data that support the findings of this study are available from the corresponding author upon request. Acknowledgements This work was supported by the National Natural Science Foundation of China (92477119, 12374370, and 22173018); Fundamental Research Funds for the Central Universities (2242025K30023). We acknowledge Southeast University, Key Laboratory of Quantum Materials and Devices, Ministry of Education (Southeast University), and the Center for Fundamental and Interdisciplinary Sciences of Southeast University, for the support in fabrication, measurement and simulations. Author contributions Q.H. provided the original idea. Q.H. and T.Q. jointly supervised the project. D.C. designed the experiments, performed the measurements, and drafted the manuscript. G.L. and Q.L. conducted the simulation calculations and analyzed the SERS results. X.Z. and X.F. assisted with the operando UV-Vis measurements and contributed to data analysis and interpretation. Z.S., R.W., Z.L., and R.L. carried out the electrocatalytic measurements, TEM analyses, SEM analyses, and processed the corresponding data. Y.W. provided the initial inspiration for the choice of probe molecules and offered daily guidance. S.C. and L.Y. prepared the schematics in Figures 2 and 3 and provided advice on the visual presentation throughout the manuscript. Q.H. revised the manuscript. All authors wrote, revised and approved the final paper. Competing interests The authors declare no competing interests. References Nitopi, S. et al. Progress and Perspectives of Electrochemical CO 2 Reduction on Copper in Aqueous Electrolyte. Chem. Rev. 119 , 7610–7672 (2019). Kok, J., Albertini, P. P., Leemans, J., Buonsanti, R. & Burdyny, T. Overcoming copper stability challenges in CO 2 electrolysis. Nat. Rev. Mater. 10 , 550–563 (2025). Qin, X.-R. et al. Electroreduction of diluted CO 2 to multicarbon products with high carbon utilization at 800 mA cm -2 in strongly acidic media. Nat. Commun. 16 , 4447 (2025). Tao, F. & Salmeron, M. Surface restructuring and predictive design of heterogeneous catalysts. Science 386 , 865 (2024). Kim, Y.-G., Baricuatro, J. H., Javier, A., Gregoire, J. M. & Soriaga, M. P. The Evolution of the Polycrystalline Copper Surface, First to Cu(111) and Then to Cu(100), at a Fixed CO 2 RR Potential: A Study by Operando EC-STM. Langmuir 30 , 15053–15056 (2014). Han, Z. et al. Steering surface reconstruction of copper with electrolyte additives for CO 2 electroreduction. Nat. Commun. 13 , 3158 (2022). Gao, X. et al. Intermediate-regulated dynamic restructuring at Ag-Cu biphasic interface enables selective CO 2 electroreduction to C 2+ fuels. Nat. Commun. 15 , 10331 (2024). Liu, Q. et al. Universal Formation, Dynamics, and Reactivities of *CO bridge Accompanying Spontaneous Reconstruction of Cu during Electrochemical CO 2 or CO Reduction. J. Am. Chem. Soc. 147 , 31497–31508 (2025). Li, Z., Wang, L., Wang, T., Sun, L. & Yang, W. Steering the Dynamics of Reaction Intermediates and Catalyst Surface during Electrochemical Pulsed CO 2 Reduction for Enhanced C 2+ Selectivity. J. Am. Chem. Soc. 145 , 20655–20664 (2023). Mu, S. et al. Hydroxyl radicals dominate reoxidation of oxide-derived Cu in electrochemical CO2 reduction. Nat. Commun. 13 , 3694 (2022). Möller, T. et al. Electrocatalytic CO 2 Reduction on CuO x Nanocubes: Tracking the Evolution of Chemical State, Geometric Structure, and Catalytic Selectivity using Operando Spectroscopy. Angew. Chem. Int. Ed. 59 , 17974–17983 (2020). Grosse, P. et al. Dynamic transformation of cubic copper catalysts during CO 2 electroreduction and its impact on catalytic selectivity. Nat. Commun. 12 , 6736 (2021). Lei, Q. et al. Structural evolution and strain generation of derived-Cu catalysts during CO 2 electroreduction. Nat. Commun. 13 , 4857 (2022). Zhao, Y. et al. Speciation of Cu Surfaces During the Electrochemical CO Reduction Reaction. J. Am. Chem. Soc. 142 , 9735-9743 (2020). Long, C. et al. Regulating reconstruction of oxide-derived Cu for electrochemical CO 2 reduction toward n-propanol. Sci. Adv. 9 , eadi6119 (2023). Chou, T.-C. et al. Controlling the Oxidation State of the Cu Electrode and Reaction Intermediates for Electrochemical CO 2 Reduction to Ethylene. J. Am. Chem. Soc. 142 , 2857–2867 (2020). Kim, J. et al. Spatiotemporal active phase evolution for CO 2 electrocatalysis. Joule 8 , 3373–3385 (2024). Chen, C. et al. The in situ study of surface species and structures of oxide-derived copper catalysts for electrochemical CO 2 reduction. Chem. Sci. 12 , 5938–5943 (2021). Phan, T. H. et al. Emergence of Potential-Controlled Cu-Nanocuboids and Graphene-Covered Cu-Nanocuboids under Operando CO 2 Electroreduction. Nano Lett. 21 , 2059–2065 (2021). Grosse, P. et al. Dynamic Changes in the Structure, Chemical State and Catalytic Selectivity of Cu Nanocubes during CO 2 Electroreduction: Size and Support Effects. Angew. Chem. Int. Ed. 57 , 6192–6197 (2018). Yang, Y. et al. Operando studies reveal active Cu nanograins for CO 2 electroreduction. Nature 614 , 262–269 (2023). Li, Y. et al. Electrochemically scrambled nanocrystals are catalytically active for CO 2 -to-multicarbons. Proc. Natl. Acad. Sci. 117 , 9194–9201 (2020). Lee, S. H. et al. Structural Transformation and Degradation of Cu Oxide Nanocatalysts during Electrochemical CO 2 Reduction. J. Am. Chem. Soc. 147 , 6536–6548 (2025). Huang, J. et al. Potential-induced nanoclustering of metallic catalysts during electrochemical CO 2 reduction. Nat. Commun. 9 , 3117 (2018). Liu, S. et al. Alkali cation-induced cathodic corrosion in Cu electrocatalysts. Nat. Commun. 15 , 5080 (2024). Yang, Y. et al. Elucidating Cathodic Corrosion Mechanisms with Operando Electrochemical Transmission Electron Microscopy. J. Am. Chem. Soc. 144 , 15698–15708 (2022). Kim, I. et al. Unveiling the reconstruction of copper bimetallic catalysts during CO 2 electroreduction. Nat. Catal. 8 , 697–713 (2025). Amirbeigiarab, R. et al. Atomic-scale surface restructuring of copper electrodes under CO 2 electroreduction conditions. Nat. Catal. 6 , 837–846 (2023). Vavra, J., Shen, T., Stoian, D., Tileli, V. & Buonsanti, R. Real‐time Monitoring Reveals Dissolution/Redeposition Mechanism in Copper Nanocatalysts during the Initial Stages of the CO 2 Reduction Reaction. Angew. Chem. Int. Ed. 60 , 1347–1354 (2021). Zhang, Q. et al. Atomic dynamics of electrified solid–liquid interfaces in liquid-cell TEM. Nature 630 , 643–647 (2024). Vavra, J. et al. Solution-based Cu + transient species mediate the reconstruction of copper electrocatalysts for CO 2 reduction. Nat. Catal. 7 , 89–97 (2024). Eren, B. et al. Activation of Cu(111) surface by decomposition into nanoclusters driven by CO adsorption. Science 351 , 475–478 (2016). Wang, Y. et al. Catalyst synthesis under CO 2 electroreduction favours faceting and promotes renewable fuels electrosynthesis. Nat. Catal. 3 , 98–106 (2019). Zhao, X. et al. Plasmonic trimers designed as SERS-active chemical traps for subtyping of lung tumors. Nat. Commun. 15 , 5855 (2024). Hao, Q. et al. Facile design of ultra-thin anodic aluminum oxide membranes for the fabrication of plasmonic nanoarrays. Nanotechnology 28 , 105301 (2017). Zaraska, L., Sulka, G. D., Szeremeta, J. & Jaskuła, M. Porous anodic alumina formed by anodization of aluminum alloy (AA1050) and high purity aluminum. Electrochimica Acta 55 , 4377–4386 (2010). Chen, D. et al. Unveiling the Cation Effects on Electrocatalytic CO 2 Reduction via Operando Surface‐enhanced Raman Spectroscopy. Small 21 , 2409569 (2025). Wei, Y. et al. Unveiling Plasmon-Induced Suzuki–Miyaura Reactions on Silver Nanoparticles via Raman Spectroscopy. ACS Catal. 14 , 15043–15051 (2024). Beaudelot, J. et al. Photoactive Copper Complexes: Properties and Applications. Chem. Rev. 122 , 16365–16609 (2022). Iwamura, M., Takeuchi, S. & Tahara, T. Ultrafast Excited-State Dynamics of Copper(I) Complexes. Acc. Chem. Res. 48 , 782–791 (2015). Ghosh, S. K., Rahman, D. S., Ali, A. L. & Kalita, A. Surface Plasmon Tunability and Emission Sensitivity of Ultrasmall Fluorescent Copper Nanoclusters. Plasmonics 8 , 1457–1468 (2013). Sarkar, N. & Radhakrishnan, T. P. Single Source Polychromatic Luminescent and White Emitting Copper Nanoclusters. Adv. Opt. Mater. 13 , 2402451 (2025). Yang, Y. et al. Operando probing dynamic migration of copper carbonyl during electrocatalytic CO2 reduction. Nat. Catal. 8 , 579–594 (2025). Protopopoff, E. & Marcus, P. Potential–pH diagrams for hydroxyl and hydrogen adsorbed on a copper surface. Electrochim. Acta 51 , 408–417 (2005). Huang, J. et al. Potential-induced nanoclustering of metallic catalysts during electrochemical CO 2 reduction. Nat. Commun. 9 , 3117 (2018). Yoon, A. et al. Revealing catalyst restructuring and composition during nitrate electroreduction through correlated operando microscopy and spectroscopy. Nat. Mater. 24 , 762–769 (2025). Zhao, Y. et al. Elucidating electrochemical CO 2 reduction reaction processes on Cu( hkl ) single-crystal surfaces by in situ Raman spectroscopy. Energy Environ. Sci. 15 , 3968–3977 (2022). Zhang, Y.-C. et al. Facet-switching of rate-determining step on copper in CO 2 -to-ethylene electroreduction. Proc. Natl. Acad. Sci. 121 , e2400546121 (2024). Xu, K. et al. Favoring CO Intermediate Stabilization and Protonation by Crown Ether for CO 2 Electromethanation in Acidic Media. Angew. Chem. Int. Ed. 62 , e202311968 (2023). Xie, Z. et al. Surface Facets Reconstruction in Copper‐Based Materials for Enhanced Electrochemical CO 2 Reduction. Small 20 , 2401530 (2024). Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformation.docx Supplementary Information Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7787217","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":530289723,"identity":"e637fbbf-8c5c-4f1b-bef2-dae6985221f7","order_by":0,"name":"Qi Hao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuklEQVRIiWNgGAWjYDACZh4gUQFhS5Cg5QxJWhiAWhjbSNHCd5z34IeP8+rkDA4wH7zNw2CXR1CL5GG+ZMmZ2w4bGxxgS7bmYUguJqjF4DCPgTTvtgOJGw7wmEnzMBxIbCBCi/Fv3jl19RsO8H8jWouZNG8Dc4LBAR424rRIArVYzjh22HDmYTZjyzkGyYS18J0/Y3zjQ02dPN/x5oc33lTYEdbCcADGYAa7k6B6ZC2jYBSMglEwCnABAN4LN7HD3xKxAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-5525-4417","institution":"Key Laboratory of Quantum Materials and Devices of Ministry of Education, School of Physics, Southeast University","correspondingAuthor":true,"prefix":"","firstName":"Qi","middleName":"","lastName":"Hao","suffix":""},{"id":530289724,"identity":"61ab87a4-c45c-4f6e-9dc5-1c1e57491ece","order_by":1,"name":"Dexiang Chen","email":"","orcid":"","institution":"Southeast University","correspondingAuthor":false,"prefix":"","firstName":"Dexiang","middleName":"","lastName":"Chen","suffix":""},{"id":530289725,"identity":"121aa429-7d94-4536-a080-27581b311473","order_by":2,"name":"Guoqun Li","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Guoqun","middleName":"","lastName":"Li","suffix":""},{"id":530289726,"identity":"b7a7e306-e01f-4c09-8e34-f566fc64374f","order_by":3,"name":"Xing Zhao","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Xing","middleName":"","lastName":"Zhao","suffix":""},{"id":530289727,"identity":"561d07e9-1948-4a81-adeb-57d9916bc477","order_by":4,"name":"Yunjia Wei","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Yunjia","middleName":"","lastName":"Wei","suffix":""},{"id":530289728,"identity":"b3e65f57-2a8e-49dd-b4dc-0ff14fc977cc","order_by":5,"name":"Shuying Chen","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Shuying","middleName":"","lastName":"Chen","suffix":""},{"id":530289729,"identity":"13081793-edb9-4292-87ce-acfa38cae7b4","order_by":6,"name":"Lei Yao","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Lei","middleName":"","lastName":"Yao","suffix":""},{"id":530289730,"identity":"2c1b90c6-abf9-4a0b-a297-a28834f600f8","order_by":7,"name":"Zixuan Sun","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Zixuan","middleName":"","lastName":"Sun","suffix":""},{"id":530289731,"identity":"a18c7dfe-9a1e-490f-8982-6ca2976e0561","order_by":8,"name":"Ruinan Wang","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Ruinan","middleName":"","lastName":"Wang","suffix":""},{"id":530289732,"identity":"e8d66235-5eb1-4ab9-86ff-e8c39fe4f398","order_by":9,"name":"Xingce Fan","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Xingce","middleName":"","lastName":"Fan","suffix":""},{"id":530289733,"identity":"1e406c98-917b-407f-981b-1777e7f04cc6","order_by":10,"name":"Zejun Li","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Zejun","middleName":"","lastName":"Li","suffix":""},{"id":530289734,"identity":"b90466e8-ac22-4e3e-a351-6a462d82fb65","order_by":11,"name":"Richen Lin","email":"","orcid":"https://orcid.org/0000-0001-6861-8542","institution":"Southeast University","correspondingAuthor":false,"prefix":"","firstName":"Richen","middleName":"","lastName":"Lin","suffix":""},{"id":530289735,"identity":"82f80e2c-83d2-471b-96d6-13bf9a2d2a7e","order_by":12,"name":"Qiang Li","email":"","orcid":"","institution":"Southeast University","correspondingAuthor":false,"prefix":"","firstName":"Qiang","middleName":"","lastName":"Li","suffix":""},{"id":530289736,"identity":"2b8ab351-c7f6-4bd0-a214-d20c89b53a69","order_by":13,"name":"Teng Qiu","email":"","orcid":"","institution":"School of Physics","correspondingAuthor":false,"prefix":"","firstName":"Teng","middleName":"","lastName":"Qiu","suffix":""}],"badges":[],"createdAt":"2025-10-06 01:20:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7787217/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7787217/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":93744447,"identity":"7af8f5b1-4016-44fd-8c6c-2cc8147c72dc","added_by":"auto","created_at":"2025-10-17 06:13:32","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":742821,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePreparations and characterizations of the Cu NP arrays.\u003c/strong\u003e (a) Schematics illustrating the fabrication of the Cu NP arrays. (b) Top-view SEM image of the AAO membrane before Cu deposition. (c) Side-view SEM image of the sample after Cu deposition. (d, e) SEM images of the fabricated Cu NP arrays at different magnifications.\u003c/p\u003e","description":"","filename":"image1.tiff.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7787217/v1/cb4eaf3a64b9053015cddb55.jpg"},{"id":93744658,"identity":"4d6f41cd-c4bd-4713-8fde-994f97223fde","added_by":"auto","created_at":"2025-10-17 06:21:32","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1197101,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOperando and post-reaction morphological evolution of the Cu NPs.\u003c/strong\u003e (a) Representative SEM images showing reconstruction at applied potentials from -0.1 V to -0.9 V, recorded during Stage A (operando reconstruction under applied cathodic bias) at different reaction times, and during Stage B (post-reaction reconstruction) after subsequent OCP (~+0.4 V) for varying durations. The scale bar is 100 nm. (b) Schematic illustration of the structural transformation pathways in Stage A and Stage B. (c) In situ Raman spectra collected at -0.5 V during CO\u003csub\u003e2\u003c/sub\u003eRR and the subsequent OCP period. Dashed lines mark the Cu\u003csub\u003e2\u003c/sub\u003eO vibrational modes at ~525 and ~620 cm\u003csup\u003e-1\u003c/sup\u003e. (d) XRD patterns of the pristine Cu NPs, after 30 min electrolysis at -0.5 V, and after an additional 30 min at OCP following the reaction. All experiments were performed in CO\u003csub\u003e2\u003c/sub\u003e-saturated 0.1 M KHCO\u003csub\u003e3\u003c/sub\u003e electrolyte (pH 6.8) using a standard H-type electrochemical cell. Potentials are referenced to the reversible hydrogen electrode (RHE), converted from Ag/AgCl using the Nernst equation and the bulk electrolyte pH. All samples were vacuum-sealed before CO\u003csub\u003e2\u003c/sub\u003eRR and characterized immediately after electrolysis. Morphological features were consistently reproduced in at least five independent experiments.\u003c/p\u003e","description":"","filename":"image2.tiff.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7787217/v1/0801d1e828298d5703d966a6.jpg"},{"id":93744449,"identity":"e0f6e62b-f6dd-4593-b001-2d8f51ce8fea","added_by":"auto","created_at":"2025-10-17 06:13:32","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1338398,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUV-Vis identification of released Cu species. \u003c/strong\u003e(a-c) Time-dependent UV-Vis spectra of the electrolyte containing 0.1 mM dmphen during CO\u003csub\u003e2\u003c/sub\u003eRR at -0.5 V (a), -0.9 V (b), and without dmphen at -0.9 V (c). (d) UV-Vis spectra of the electrolyte containing 0.1 mM dmphen after 30 min CO\u003csub\u003e2\u003c/sub\u003eRR at different applied potentials. (e) Time-dependent intensity of the 455 nm band obtained from OCP measurements (Supplementary Fig. 19), measured with the Cu NPs samples in (d) and a blank control sample without CO\u003csub\u003e2\u003c/sub\u003eRR pretreatment (red line). The initial intensity in (e) was offset to zero for comparison. (f) SEM images after 60 min at OCP, with and without dmphen, for pristine Cu NPs samples and samples subject to CO\u003csub\u003e2\u003c/sub\u003eRR pretreatment at -0.5 V for 60 min. The scale bar is 100 nm. (g) Schematics illustrating the distinct dissolution-redeposition pathways under different conditions. All experiments were performed in CO\u003csub\u003e2\u003c/sub\u003e-saturated 0.1 M KHCO\u003csub\u003e3\u003c/sub\u003e using Cu NPs electrodes (1.0 cm\u003csup\u003e2\u003c/sup\u003e).\u003c/p\u003e","description":"","filename":"image3.tiff.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7787217/v1/314bc5297779c1f88702c7b2.jpg"},{"id":93744450,"identity":"936385e7-3f06-4528-b834-2490369baa8e","added_by":"auto","created_at":"2025-10-17 06:13:32","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":950854,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eInvestigation of intermediate-induced Cu reconstruction. (\u003c/strong\u003ea) In situ Raman spectra from the Cu NPs during CO\u003csub\u003e2\u003c/sub\u003eRR at varying potentials. (b) Calculated adsorption energy differences among key reaction intermediates. (c) Calculated surface energies of Cu(111), Cu(110), and Cu(100) as a function of *CO coverage. (d) Relative ratios of Cu(111), Cu(110), and Cu(100) facets, quantified by analyzing deconvoluted OH\u003csup\u003e-\u003c/sup\u003e adsorption peaks from cyclic voltammetry (Figs. S25 and S26). (e, f) In situ Raman spectra from the Cu NPs with (e) and without (f) 0.1 M 18-C-6 at varying potentials during CO\u003csub\u003e2\u003c/sub\u003eRR. (g, h) SEM images of the Cu NPs after CO\u003csub\u003e2\u003c/sub\u003eRR at -0.5 V for 30 min, with (g) and without (h) 18-C-6. The scale bars are 100 nm.\u003c/p\u003e","description":"","filename":"image4.tiff.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7787217/v1/72eb61245baaf3cffd9dfc48.jpg"},{"id":98421158,"identity":"c968ce72-cfab-4df2-9626-4704d3446b37","added_by":"auto","created_at":"2025-12-17 16:24:25","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5237647,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7787217/v1/91869296-c6f5-4190-9c32-8d4f1208bf2a.pdf"},{"id":93744452,"identity":"362ddf10-2100-47ac-a383-267c7268547f","added_by":"auto","created_at":"2025-10-17 06:13:33","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":42842964,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-7787217/v1/727d291b09d3968a3fe2f847.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Mechanistic pathways of copper catalyst reconstruction in CO2 electroreduction","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction reaction (CO\u003csub\u003e2\u003c/sub\u003eRR) provides a feasible route for converting CO\u003csub\u003e2\u003c/sub\u003e into valuable chemicals and fuels under mild conditions. Among available catalysts, copper (Cu) stands out for its moderate binding affinity toward key reaction intermediates, enabling the formation of C\u003csub\u003e2+\u003c/sub\u003e products\u003csup\u003e1\u003c/sup\u003e. This process is highly surface-sensitive, with reactivity strongly influenced by the oxidation state, morphology, and local microenvironment of the surface catalyst\u003csup\u003e2\u0026ndash;4\u003c/sup\u003e. As a result, even minor variations in these factors can lead to substantial changes in overall product selectivity. Therefore, controlling and understanding the state of the Cu catalyst during CO\u003csub\u003e2\u003c/sub\u003eRR is fundamentally important for both mechanistic elucidation and catalyst optimization.\u003c/p\u003e\n\u003cp\u003eOwing to its relatively low cohesive energy and high surface-atom mobility, Cu is intrinsically susceptible to surface reconstruction under electrochemical operation\u003csup\u003e5,6\u003c/sup\u003e. This can lead to spontaneous atomic rearrangements and compositional transformations of Cu catalyst, reshaping surface morphology and altering the distribution of active sites\u003csup\u003e7,8\u003c/sup\u003e. Consequently, the actual working state of Cu catalyst may differ substantially from its as-prepared or post-reaction form, which can lead to misinterpretation of catalytic mechanisms. Therefore, a detailed understanding of the processes governing surface reconstruction is essential for accurately elucidating catalytic behavior in CO\u003csub\u003e2\u003c/sub\u003eRR.\u003c/p\u003e\n\u003cp\u003eAlthough surface reconstruction of Cu has been widely reported, its dynamics and underlying mechanisms remain under debate. Some studies have observed rapid structural reshaping within the first few minutes\u0026nbsp;of electrolysis\u003csup\u003e9,10\u003c/sup\u003e, whereas others report continuous transformations throughout the reaction\u003csup\u003e11,12\u003c/sup\u003e. While some attribute\u0026nbsp;CO\u003csub\u003e2\u003c/sub\u003eRR\u0026nbsp;activity to the reduction of Cu oxides into metallic Cu\u003csup\u003e13\u0026ndash;15\u003c/sup\u003e, others argue that oxidic species may directly contribute to catalysis\u003csup\u003e16\u0026ndash;18\u003c/sup\u003e. Moreover, morphologically diverse Cu catalysts often exhibit similar reconstruction trends, for example, the appearance of secondary particles with a typically cubic shape\u003csup\u003e19,20\u003c/sup\u003e. However,\u0026nbsp;recent evidence suggests that such features form predominantly after the removal of bias in the air, rather than under operando conditions\u003csup\u003e21,22\u003c/sup\u003e.\u0026nbsp;Taken together, these observations imply that Cu surface reconstruction does not proceed through a single pathway, but instead arises from a complex interplay of surface adsorbates,\u0026nbsp;applied bias, catalytic environment, and intrinsic thermodynamic driving forces.\u003c/p\u003e\n\u003cp\u003eBuilding on these complexities, multiple reconstruction mechanisms have been proposed, including nanoparticle aggregation\u003csup\u003e23,24\u003c/sup\u003e, cathodic corrosion\u003csup\u003e25,26\u003c/sup\u003e, and dissolution-redeposition mechanisms\u003csup\u003e27\u0026ndash;29\u003c/sup\u003e. Among these, the latter is increasingly recognized as a central pathway for surface evolution under CO\u003csub\u003e2\u003c/sub\u003eRR conditions. Specifically, in situ transmission electron microscopy (TEM) has revealed amorphous-crystalline\u0026nbsp;transitions\u0026nbsp;in Cu catalysts\u003csup\u003e30\u003c/sup\u003e, accompanied by the formation of Cu\u003csup\u003e+\u003c/sup\u003e species as confirmed by UV-Vis spectroscopy\u003csup\u003e31\u003c/sup\u003e. Complementary density functional theory (DFT) calculations and Raman analysis further suggest that adsorbed *CO intermediates play a central role in triggering and regulating these transformations. *CO can reshape the surface energies of Cu through its facet-dependent binding affinities\u003csup\u003e32\u003c/sup\u003e, inducing selective dissolution and redeposition of Cu species on specific facets and thereby guiding the reconstruction pathways\u003csup\u003e33\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDespite growing evidence, several critical questions remain unresolved. First, \u003cem\u003ewhat is the chemical\u0026nbsp;\u003c/em\u003e\u003cem\u003enature\u003c/em\u003e\u003cem\u003e\u0026nbsp;of the\u0026nbsp;\u003c/em\u003e\u003cem\u003edissolved\u003c/em\u003e\u003cem\u003e\u0026nbsp;Cu species, and how do applied potential and adsorbed intermediates influence their dissolution pathways?\u0026nbsp;\u003c/em\u003e\u003cem\u003eSecond\u003c/em\u003e,\u003cem\u003e\u0026nbsp;how can we distinguish surface changes occurring under operando CO\u003csub\u003e2\u003c/sub\u003eRR conditions from those emerging during the post-reaction stage, and thereby minimize interference from the latter in mechanistic studies? Third, to what extent does surface reconstruction reshape the catalyst surface and direct its evolution under CO\u003csub\u003e2\u003c/sub\u003eRR conditions?\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eIn this study, we employ clean and structurally well-defined Cu nanoparticle (NP) arrays as a model system to investigate surface reconstruction during CO\u003csub\u003e2\u003c/sub\u003eRR. This uniform, chemical-free platform enables direct visualization of morphological evolution, while its strong plasmonic response allows in situ and real-time tracking of interfacial intermediates through surface-enhanced Raman spectroscopy (SERS). By integrating this platform with complementary techniques, we reveal potential-dependent reconstruction behavior and resolve three distinct dissolution-redeposition pathways of Cu catalysts. Importantly, we identify the chemical nature of the dissolved species, revealing not only Cu⁺ but also previously unrecognized Cu\u003csup\u003e0\u003c/sup\u003e species, as a function of applied potential. Together, these findings help clarify the open questions outlined above, and provide new experimental evidence and perspectives on how surface reconstruction may be understood and potentially regulated under CO\u003csub\u003e2\u003c/sub\u003eRR conditions.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003ePreparations and advantages of the Cu NP arrays\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCu NP arrays with a centimeter scale (Supplementary Fig.\u0026nbsp;1) were fabricated to investigate the surface reconstruction. As illustrated in Figure 1a, patterned Cu NPs were prepared using porous anodic aluminum oxide (AAO) membrane as a mask during e-beam evaporation\u003csup\u003e34,35\u003c/sup\u003e. Specifically, AAO membrane with a pore diameter of ~80 nm was transferred onto a conductive substrate (Figure 1b), followed by Cu deposition through the AAO mask with e-beam evaporation (Figure 1c). After deposition, the mask was removed, exposing the resulting Cu NP arrays (Figure 1d and 1e). Scanning electron microscopy (SEM) images confirm the uniformity and hexagonal periodicity of the Cu NP arrays. The fabricated Cu NPs exhibit a face-centered cubic (fcc) lattice of metallic Cu, and were stored in vacuum prior to use.\u003c/p\u003e\n\u003cp\u003eThis fabrication approach offers several advantages for the investigation of surface reconstruction. First, the AAO pore diameter is intrinsically determined by the anodization voltage, ensuring precise size control and physically defined reproducibility across samples, batches, and laboratories\u003csup\u003e36\u003c/sup\u003e. This offers good structural uniformity and reproducibility for the Cu NPs, enabling clear identification of reconstruction-induced surface evolution. Second, the size tunability allows optimization of the localized surface plasmon resonance (LSPR), which in turn amplifies the intrinsically weak Raman signals of CO\u003csub\u003e2\u003c/sub\u003eRR intermediates via SERS\u003csup\u003e37\u003c/sup\u003e. Third, the Cu NPs prepared by physical vapor deposition are free from chemical surfactants and ligands, and remain structurally stable under reactive conditions, thereby minimizing particle aggregation and chemical interference and providing a clean model for mechanistic studies\u003csup\u003e38\u003c/sup\u003e. Collectively, these features provide a robust platform for tracking catalyst structural evolution during CO\u003csub\u003e2\u003c/sub\u003eRR while simultaneously probing intermediates by SERS for mechanistic insights.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOperando and post-electrochemical morphological changes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSEM was used to track the morphological evolution of the Cu NPs at different applied potentials (Figure 2a). Two distinct reconstruction stages were identified: \u003cstrong\u003eStage A\u003c/strong\u003e, representing operando reconstruction under cathodic bias during CO\u003csub\u003e2\u003c/sub\u003eRR, and \u003cstrong\u003eStage B\u003c/strong\u003e, representing post-reconstruction occurring spontaneously in CO\u003csub\u003e2\u003c/sub\u003e-saturated electrolyte at open-circuit potential (OCP, ~+0.4 V). The main morphological pathways are schematically summarized in Figure 2b.\u003c/p\u003e\n\u003cp\u003eIn \u003cstrong\u003eStage A\u003c/strong\u003e: two major morphological changes were observed, both exhibiting a non-linear dependence on applied potential: \u003cstrong\u003e(1) Progressive shrinkage of the original Cu NPs.\u0026nbsp;\u003c/strong\u003eThe initially hemispherical Cu NPs gradually decreased in size. This shrinkage was minimal at -0.1 V, pronounced at moderate potentials (-0.3 V to -0.5 V), and progressively suppressed under more negative conditions. \u003cstrong\u003e(2) Emergence of secondary particles.\u003c/strong\u003e The secondary particles were rarely detected at both low and high potentials (-0.1 V and -0.9 V), but evident at moderate values. Two types of secondary particles were identified. Cubic-like domains were most prominent at -0.5 V and -0.3 V, while smaller satellite particles formed at -0.5 V and -0.7 V. These secondary particles were not observed in control experiments in the absence of CO\u003csub\u003e2\u003c/sub\u003e.\u003c/p\u003e\n\u003cp\u003eThe non-linear potential dependence, along with simultaneous observation of parent-particle shrinkage and secondary-particle formation in \u003cstrong\u003eStage A\u003c/strong\u003e, provides direct evidence for a dissolution-redeposition mechanism.\u0026nbsp;In this process, the original Cu NPs act as atom sources, releasing Cu species that redeposit into secondary particles. Notably, even at -0.9 V, where secondary particle formation was suppressed, the shrinkage persisted (see also\u0026nbsp;Supplementary Fig.\u0026nbsp;2), suggesting that the dissolution can persist even when redeposition is kinetically hindered. We therefore propose two distinct dissolution pathways: (i) an electrochemically driven route governed by the interfacial environment, where surface intermediates govern the dissolution and account for the non-linear potential dependence of redeposition, and (ii) a field-assisted dissolution route that becomes increasingly dominant at high cathodic bias. We suggest that the Cu species released through these two pathways are chemically distinct, and this will be further examined in the following sections.\u003c/p\u003e\n\u003cp\u003eUpon entering \u003cstrong\u003eStage B\u003c/strong\u003e after bias removal, the parent Cu NPs underwent further shrinkage\u0026nbsp;together with reconstruction into cubes. This process proceeded spontaneously until the parent Cu was fully consumed, suggesting a surface-energy-driven ripening process governed by the electrolyte environment\u003csup\u003e23,29\u003c/sup\u003e. The newly formed particles appeared brighter in SEM images and exhibited sharper edges compared with the parent Cu NPs, indicative of the formation of Cu oxides. To assess the generality of this behavior, we extended our study to evaporated and sputtered planar Cu film electrodes (Figs. S3-S7). All tested systems displayed similar morphological evolution, confirming that the cube formation in Stage B is intrinsically surface-energy driven.\u003c/p\u003e\n\u003cp\u003eThe oxidation state of the Cu catalyst was examined by in situ Raman spectroscopy (Figure 2c). Under applied bias, the native oxide modes disappeared, suggesting electrochemical reduction to metallic Cu. Upon bias removal, these peaks gradually reappeared, suggesting a rapid oxidation in \u003cstrong\u003eStage B\u003c/strong\u003e. X-ray diffraction (XRD) results in Figure 2d support this observation: after electrolysis, only Cu(111) and weak Cu(100) reflections of fcc Cu were observed, whereas after 30 min at OCP, distinct Cu\u003csub\u003e2\u003c/sub\u003eO reflections emerged. Combined with the SEM (Figure 2a) and TEM data (Figs. S8 and S9), the results suggest that the cubic structures observed at OCP are predominantly composed of single crystal Cu\u003csub\u003e2\u003c/sub\u003eO.\u0026nbsp;Interestingly, the formation of Cu\u003csub\u003e2\u003c/sub\u003eO cubes occurs at OCP but requires prior CO\u003csub\u003e2\u003c/sub\u003eRR exposure (Supplementary Fig.\u0026nbsp;10). Further analysis indicates that sufficient CO\u003csub\u003e2\u003c/sub\u003eRR duration, rather than the applied potential, appears to be essential to generate the conditions that facilitate cubic growth during OCP (Figs. S2-S6). We attribute this to the formation of active sites on Cu catalyst during CO\u003csub\u003e2\u003c/sub\u003eRR, such as defect-rich domains or intermediate-modified terraces (Supplementary Fig.\u0026nbsp;11)\u003csup\u003e21\u003c/sup\u003e. Notably, the formation of these Cu\u003csub\u003e2\u003c/sub\u003eO cubes proceeds in the electrolyte, in contrast to earlier reports where air exposure is required (Supplementary Fig.\u0026nbsp;12)\u003csup\u003e22\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInvestigation of the dissolution-redeposition pathways\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify the chemical nature of the released Cu\u003csup\u003e+\u003c/sup\u003e species, we performed in situ UV-Vis spectroscopy using a custom-designed electrochemical cell (schematic in Supplementary Fig. 13). 2,9-dimethyl-1,10-phenanthroline (dmphen) was introduced as a selective chelating agent, which can form a stable and soluble complex with Cu\u003csup\u003e+\u003c/sup\u003e, yielding a characteristic absorption band at ~455 nm\u003csup\u003e31,39,40\u003c/sup\u003e. Spectra were recorded across a potential range from -0.1 V to -0.9 V, with -0.5 V (Figure 3a) and -0.9 V (Figure 3b) taken as representative cases (full dataset in Supplementary Fig. 14). At -0.5 V, the 455 nm band appeared rapidly during CO\u003csub\u003e2\u003c/sub\u003eRR, confirming the release of Cu\u003csup\u003e+\u003c/sup\u003e into the electrolyte. In the presence of dmphen, the shrinkage of parent particles persisted, yet the formation of secondary nanoparticles was remarkably suppressed (Supplementary Fig. 15), confirming that dissolved Cu⁺ attributes to secondary-particle redeposition in Stage A. As shown in Figure 3d, the 455 nm Cu\u003csup\u003e+\u003c/sup\u003e band strengthened from -0.1 V to -0.5 V, but weakened at -0.7 V, and was barely detectable at -0.9 V. This potential-dependent behavior aligns with the morphological evolution observed by SEM in Figure 2, further supporting a Cu\u003csup\u003e+\u003c/sup\u003e dissolution-redeposition pathway.\u003c/p\u003e\n\u003cp\u003eAt -0.9 V (Figure 3b), the 455 nm Cu\u003csup\u003e+\u003c/sup\u003e-dmphen band was not detected, and distinct sub-400 nm features emerged. These signals are attributed to quantum-confined optical transitions of Cu\u003csup\u003e0\u003c/sup\u003e clusters (TEM, Supplementary Fig. 16)\u003csup\u003e41,42\u003c/sup\u003e, suggesting that neutral Cu\u003csup\u003e0\u003c/sup\u003e atoms were released and subsequently aggregated into solution-phase clusters. While previous studies reported *CO-induced Cu clusters on catalyst surfaces via a Cu⁺ dissolution-reduction pathway\u003csup\u003e28,43\u003c/sup\u003e, our results suggest that Cu clusters also form in solution and are composed of Cu\u003csup\u003e0\u003c/sup\u003e atoms, revealing a distinct mechanism. Importantly, these sub-400 nm features were also observed in the absence of dmphen (Figure 3c and Supplementary Fig. 17), confirming that the release of Cu\u003csup\u003e0\u003c/sup\u003e species represents an intrinsic dissolution pathway. Additionally, the Cu content in electrolyte at -0.9 V was higher than at other potentials (Supplementary Fig. 18), nevertheless, redeposition into secondary particles was not observed at this potential (Figure 2a, A3).\u003c/p\u003e\n\u003cp\u003eMoreover, we continued to monitor these samples during the subsequent OCP stage (Figure 3e and Supplementary Fig. 19). The 455 nm Cu\u003csup\u003e+\u003c/sup\u003e band increased continuously during OCP, suggesting sustained Cu\u003csup\u003e+\u003c/sup\u003e release into the electrolyte. According to the Cu Pourbaix diagram, metallic Cu is thermodynamically metastable at OCP and undergoes spontaneous oxidation into Cu\u003csup\u003e+\u003c/sup\u003e, initiating rapid oxidative dissolution\u003csup\u003e44\u0026ndash;46\u003c/sup\u003e. The release of Cu\u003csup\u003e+\u003c/sup\u003e at OCP also exhibited a potential dependence, pronounced at moderate potentials but less significant at the extremes. This potential dependence did not correlate with the morphological trends observed by SEM (Figure 2a, B3), implying that factors beyond Cu\u003csup\u003e+\u003c/sup\u003e release govern the redeposition process. Notably, a pristine control sample without CO\u003csub\u003e2\u003c/sub\u003eRR pretreatment also exhibited pronounced Cu\u003csup\u003e+\u003c/sup\u003e release at OCP (red line). As shown in Figure 3f, the introduction of dmphen can suppress the redeposition into \u003cstrong\u003eCu\u003c/strong\u003e\u003csub\u003e2\u003c/sub\u003e\u003cstrong\u003eO\u003c/strong\u003e cubes at OCP a\u003cstrong\u003es expected\u003c/strong\u003e, suggesting a dissolution-redeposition pathway. Interestingly, for the pristine control sample, even without dmphen, no \u003cstrong\u003eCu\u003c/strong\u003e\u003csub\u003e2\u003c/sub\u003e\u003cstrong\u003eO cubes were observed despite clear Cu\u003csup\u003e+\u003c/sup\u003e release\u003c/strong\u003e. Therefore, we deduce that active sites generated during CO\u003csub\u003e2\u003c/sub\u003eRR, like defect-rich or intermediate-modified surfaces, enable Cu\u003csub\u003e2\u003c/sub\u003eO cube nucleation during OCP.\u003c/p\u003e\n\u003cp\u003eTogether, these findings reveal distinct dissolution-redeposition pathways under different conditions. During CO\u003csub\u003e2\u003c/sub\u003eRR: (i) at low and moderate reductive potentials, *CO-induced Cu\u003csup\u003e+\u003c/sup\u003e dissolution dominates, giving rise to redeposition into metallic secondary nanoparticles; (ii) at more negative potentials, bias-driven dissolution of metallic Cu\u003csup\u003e0\u003c/sup\u003e species becomes predominant, while redeposition into secondary nanoparticles is suppressed; and at OCP: (iii) surface Cu atoms spontaneously oxidize to Cu\u003csup\u003e+\u003c/sup\u003e, which undergo oxidative dissolution and subsequent redeposition into Cu\u003csub\u003e2\u003c/sub\u003eO cubes through an Ostwald-like ripening process. Crucially, prior CO\u003csub\u003e2\u003c/sub\u003eRR treatment is essential to generate active nucleation sites that initiate the growth of Cu\u003csub\u003e2\u003c/sub\u003eO cubes\u003cem\u003e.\u003c/em\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIntermediate-induced Cu reconstruction in CO\u003csub\u003e2\u003c/sub\u003eRR\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn situ Raman spectroscopy was performed to monitor surface intermediates at different potentials. Our fabrication strategy allows precise design of the Cu NPs to optimize their SERS efficiency, allowing ultrasensitive detection of the intrinsically weak Raman signals of the intermediates (Figs. S20 and S21). As shown in Figure 4a, a peak assigned to HCO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e-\u003c/sup\u003e was initially observed\u0026nbsp;at 1016 cm\u003csup\u003e-1\u003c/sup\u003e at OCP.\u0026nbsp;This peak disappeared\u0026nbsp;at\u0026nbsp;-0.1 V,\u0026nbsp;where\u0026nbsp;a\u0026nbsp;strong peak at 1068 cm\u003csup\u003e-1\u003c/sup\u003e corresponding\u0026nbsp;to\u0026nbsp;CO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e2-\u003c/sup\u003e emerged. Simultaneously, bands\u0026nbsp;at 2000-2100 cm\u003csup\u003e-1\u003c/sup\u003e assigned to adsorbed\u0026nbsp;*CO appeared.\u0026nbsp;Upon lowering the potential\u0026nbsp;from -0.1 to -0.6 V, a peak at 1386 cm\u003csup\u003e-1\u003c/sup\u003e assigned to\u0026nbsp;*COOH, and another peak at\u0026nbsp;1440 cm\u003csup\u003e-1\u003c/sup\u003e assigned to *OCHO, were observed. At more negative potentials, the *OCHO signal weakened, while a new band at 530 cm\u003csup\u003e-1\u003c/sup\u003e assigned to *CH\u003csub\u003e2\u003c/sub\u003eCHO emerged,\u0026nbsp;suggesting dimerization between\u0026nbsp;*CO species\u003csup\u003e37,47\u003c/sup\u003e (Table S1).\u0026nbsp;This trend aligns with the decrease in *COOH intensity, indicating its\u0026nbsp;rapid conversion to\u0026nbsp;*CO and a consequent increase in\u0026nbsp;*CO surface coverage. At\u0026nbsp;-1.0 V,\u0026nbsp;Raman acquisition was hindered by vigorous\u0026nbsp;bubble formation.\u003c/p\u003e\n\u003cp\u003eAs shown in Figure 4b, the adsorption energies of key intermediates on Cu were calculated using DFT, among which *CO exhibits the strongest binding (see also Figs. S22 and S23, and Table S2). This strong binding localizes electrons into the Cu-*CO bonds and disrupts the lateral Cu-Cu bonding network, thereby facilitating atomic rearrangement, as evidences by Bader charge analysis (Supplementary Fig.\u0026nbsp;24). Moreover, as revealed in Figure 4b, the binding strength of *CO is facet-dependent: stronger on Cu(100) and Cu(110) but weaker on Cu(111), consistent with previous reports\u003csup\u003e48\u003c/sup\u003e. Consequently, at low *CO coverage, *CO preferentially adsorbs on Cu(100) and Cu(110), lowering their surface energies relative to Cu(111). As shown in Figure 4c, although Cu(111) intrinsically has the lowest surface energy, the presence of *CO shifts the equilibrium, offering a thermodynamic driving force that favors reconstruction from Cu(111) toward Cu(100) and Cu(110). On the other hand, at high *CO coverage, which corresponds to more negative potentials, *CO adsorption on Cu(100) and Cu(110) approaches saturation, while adsorption on Cu(111) increase markedly. Under these conditions, the surface energy of Cu(111) also decreases and the thermodynamic driving force for reconstruction toward Cu(100) and Cu(110) is weakened. This explains the observed suppression of redeposition at -0.9 V during Stage A (Figure 2a), where high *CO coverage leads to a loss of facet preference. The proposed mechanism of facet-dependent reconstruction under varying *CO coverage, corresponding to CO\u003csub\u003e2\u003c/sub\u003eRR at -0.5 V and -0.9 V, was experimentally verified through facet-specific OH\u003csup\u003e-\u003c/sup\u003e adsorption analysis in cyclic voltammetry (Figure 4d and Figs. S25 and S26). The results reveal that reconstruction from Cu(111) toward Cu(100) and Cu(110) was pronounced at -0.5 V but significantly suppressed at -0.9 V, consistent with the DFT calculations. Comparison with Ar-saturated conditions (Supplementary Fig.\u0026nbsp;26) further confirms that this facet reconstruction is driven by CO\u003csub\u003e2\u003c/sub\u003eRR, not directly by the applied potential.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo experimentally validate this hypothesis, we introduced 18-crown-6 (18-C-6), a macrocyclic ether known to enhance the binding affinity of *CO on metal surfaces, to modulate the surface coverage of *CO\u003csup\u003e49\u003c/sup\u003e. The Cu-CO stretching bands at ~280 cm\u003csup\u003e-1\u003c/sup\u003e and ~360 cm\u003csup\u003e-1\u003c/sup\u003e were used as indicators of *CO coverage. As shown in Figure 4e and 4f, in standard CO\u003csub\u003e2\u003c/sub\u003eRR experiments without 18-C-6, the bands indicative of high *CO coverage appeared only at relatively negative potentials, beginning at -0.7 V. In contrast, in the presence of 18-C-6, these bands were significantly enhanced and emerged at less negative potentials, starting from -0.4 V. These results confirm that 18-C-6 effectively increased *CO coverage on Cu, particularly under unsaturated conditions at relatively low potentials. Correspondingly, upon introduction of 18-C-6, both shrinkage of the parent particle and the formation of secondary particles were markedly suppressed at -0.5 V (Figure 4g, 4h and Supplementary Fig. 27). This finding supports our view that the reconstruction from Cu(111) toward Cu(100) and Cu(110) is favored under low *CO coverage but inhibited under high *CO coverage.\u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn summary, the use of clean and structurally well-defined Cu NP arrays enables direct visualization of catalyst morphology and simultaneous in situ SERS detection of reaction intermediates, providing a robust platform to elucidate the complex pathways of surface reconstruction during CO\u003csub\u003e2\u003c/sub\u003eRR. Importantly, Cu restructuring during operando conditions follows a nonlinear, potential-dependent behavior. Through the integration of SEM, UV-Vis, in situ Raman, and DFT studies, we provide direct experimental evidence that dissolution-redeposition is a central mechanism governing Cu restructuring, proceeding through three distinct pathways depending on the electrochemical environment:\u003c/p\u003e\n\u003cp\u003e(i) At low and moderate reductive potentials, *CO-mediated Cu\u003csup\u003e+\u003c/sup\u003e dissolution dominates and drives redeposition into metallic secondary particles. *CO coverage acts as a molecular switch: at low *CO overage, *CO adsorption drives facet-selective reconstruction from Cu(111) toward Cu(100) and Cu(110); while at high coverage, this driving force vanishes. (ii) At more negative potentials, field-assisted dissolution of metallic Cu\u003csup\u003e0\u003c/sup\u003e becomes predominant, resulting in particle shrinkage without redeposition. In this case, neutral Cu\u003csup\u003e0\u003c/sup\u003e species are leached from the surface and aggregate into small soluble clusters. (iii) Upon bias removal, the dominant mechanism shifts to oxidative dissolution. Cu undergoes oxidative dissolution to Cu⁺, which subsequently redeposits as Cu\u003csub\u003e2\u003c/sub\u003eO cubes. This transformation requires CO\u003csub\u003e2\u003c/sub\u003eRR pretreatment, likely due to the CO\u003csub\u003e2\u003c/sub\u003eRR-generated active nucleation sites. These Cu\u003csub\u003e2\u003c/sub\u003eO cubes resemble the metallic secondary cubes observed under operando CO\u003csub\u003e2\u003c/sub\u003eRR but form far more rapidly and extensively, underscoring the risk of misidentifying post-reaction products as active catalytic states.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBeyond mechanistic insights, our findings suggest practical strategies for rational catalyst design. By modulating *CO coverage, potential windows, and bias protocols, dissolution-redeposition pathways can be selectively promoted or suppressed. Although a comprehensive understanding of how reconstruction feeds back into catalytic activity is not yet established, our results provide indicative evidence toward this question. 1) Our results point to an intriguing possibility that, at least for fcc Cu catalysts, although reconstruction is inevitable, its direct impact on CO\u003csub\u003e2\u003c/sub\u003eRR efficiency may be less significant than often assumed. At strongly negative potentials, where *CO coverage is high and CO\u003csub\u003e2\u003c/sub\u003eRR activity is maximized, morphological and facet reconstruction appear limited within the experimental timeframe and thus exert only a minor influence on catalytic performance. 2) Beyond morphology, the nature of dissolved species may also play an important role in CO\u003csub\u003e2\u003c/sub\u003eRR. Dissolution-redeposition is now widely recognized as a central mechanism of Cu reconstruction. Traditionally, the dissolved species were assumed to be Cu\u003csup\u003e+\u003c/sup\u003e only, which is readily reduced under CO\u003csub\u003e2\u003c/sub\u003eRR conditions and thus considered irrelevant to catalytic activity. Our observation of dissolved Cu\u003csup\u003e0\u003c/sup\u003e species challenges this view and introduces a critical question: could neutral Cu species directly influence CO\u003csub\u003e2\u003c/sub\u003eRR activity and selectivity? Addressing these issues, alongside a broader understanding of operando and post-reaction reconstructions, is crucial for establishing accurate structure-performance relationships and for guiding the design of future CO\u003csub\u003e2\u003c/sub\u003eRR catalysts that intelligently exploit or mitigate reconstruction.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eMaterials and methods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCopper (Cu, 99.999%) and Titanium (Ti, 99.999%) were purchased from Zhong Nuo Advanced Material Technology Co., Ltd (Beijing, China). Potassium bicarbonate (KHCO\u003csub\u003e3\u003c/sub\u003e, \u0026ge;99.5%), copper chloride dihydrate (CuCl\u003csub\u003e2\u003c/sub\u003e\u0026bull;2H\u003csub\u003e2\u003c/sub\u003eO, 99%, Aladdin), acetone (\u0026ge;99.5%, Aladdin), 2,9-dimethyl-phenanthroline (99%, Macklin), phosphoric acid (H\u003csub\u003e3\u003c/sub\u003ePO\u003csub\u003e4\u003c/sub\u003e, \u0026ge;99%, Aladdin) and polymethyl methacrylate (PMMA, 950PMMA A4, MicroChem) were used as received without further purification.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of the AAO membranes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAnodic aluminum oxide (AAO) samples were produced on aluminum foils using a standard anodization protocol by commercial companies (TopMembranes Technology Ltd. and Micron Technology Co., Ltd.). The purchased AAO samples were etched in 5 wt% phosphoric acid at 30 \u0026deg;C for 29 min to enlarge the pore diameter, followed by spin-coating with PMMA at 1200 rpm for 60 s, and baked at 120 ℃ for 20 min. The samples were subsequently immersed in 5 wt% NaOH for 20 min and then transferred to 10 wt% CuCl\u003csub\u003e2\u003c/sub\u003e solution to remove the aluminum substrate. The obtained AAO membranes were treated in a mixed solution of 5 wt% phosphoric acid and 2 wt% CuCl\u003csub\u003e2\u003c/sub\u003e for 37 min to remove the AAO barrier layer and adjust the pore diameter. Afterwards, the PMMA protective layer was removed in acetone, and the resulting free-standing porous AAO membranes were rinsed thoroughly with distilled water before transfer onto substrates.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eElectrochemical Measurements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eElectrochemical CO\u003csub\u003e2\u003c/sub\u003eRR experiments were performed in an H-type cell using a conventional three-electrode configuration. A Nafion 117 membrane was used to separate the cathodic and anodic compartments and served as a proton exchange membrane. All measurements were conducted using a CHI660E electrochemical workstation (CH Instruments). A platinum wire and an Ag/AgCl (3 M KCl) electrode were employed as the counter and reference electrodes, respectively. Prior to each experiment, the electrolyte (0.1 M KHCO\u003csub\u003e3\u003c/sub\u003e) was purged with high-purity CO\u003csub\u003e2\u003c/sub\u003e gas at a flow rate of 10 sccm for at least 15 minutes to remove dissolved oxygen and saturate the solution. Potential was converted following the relationship: E (\u003cem\u003evs.\u003c/em\u003e RHE) = E (\u003cem\u003evs.\u0026nbsp;\u003c/em\u003eAg/AgCl) + 0.197 + 0.0592*pH.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIn situ UV-Vis Measurement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe UV-Vis absorption spectra were collected using a HITACHI U-3900 spectrometer equipped with a customized electrochemical liquid-cell (Supplementary Fig.\u0026nbsp;14). The measurements were conducted under CO\u003csub\u003e2\u003c/sub\u003eRR conditions, using a three-electrode configuration with Cu NPs array as the working electrode, Ag/AgCl as the reference electrode, and a platinum wire as the counter electrode. Spectra were collected with a scan speed of 120 nm/min, sampling interval of 0.50 nm, and slit width of 1 nm. Sample size: 1.0 cm\u003csup\u003e2\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIn situ Raman Measurement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn situ Raman measurements were performed in a custom-designed in situ Raman cell (Gaossunion Photoelectric Technology Co., Ltd., Tianjin, China) coupled to an electrochemical workstation. A three-electrode system was employed with a Cu NPs working electrode, Ag/AgCl reference electrode, and platinum wire counter electrode. Raman spectra were acquired using a HORIBA Jobin Yvon LabRAM HR Evolution system with a 633 nm laser (2 mW power) and a 600/mm grating. The integration time was 10 s per spectrum.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eElectrochemical OH- Adsorption\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eElectrochemical adsorption of OH\u003csup\u003e-\u003c/sup\u003e was performed on the Cu NPs catalysts. The reaction follows: Cu + OH\u003csup\u003e-\u003c/sup\u003e \u0026rarr; Cu(OH)\u003csub\u003ead\u003c/sub\u003e + e\u003csup\u003e-\u003c/sup\u003e. Measurements were carried out in Ar-saturated 0.1 M KOH solution using a Pt counter electrode and an Hg/HgO reference electrode. All potentials were converted to RHE scale following the equation: E (\u003cem\u003evs.\u003c/em\u003e RHE) = E (\u003cem\u003evs.\u0026nbsp;\u003c/em\u003eHg/HgO) + 0.098 + 0.059*pH. The fractional surface coverage of Cu(111), Cu(100), and Cu(110) facets was determined by normalizing the OH\u003csup\u003e-\u003c/sup\u003e adsorption charge to the corresponding charge densities on single-crystal Cu electrodes: 2.16 \u0026mu;C cm\u003csup\u003e-2\u003c/sup\u003e for Cu(111), 13.3 \u0026mu;C cm\u003csup\u003e-2\u003c/sup\u003e for Cu(110), and 8.22 \u0026mu;C cm\u003csup\u003e-2\u003c/sup\u003e for Cu(100), as reported in previous studies\u003csup\u003e33,50\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDFT Calculations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDFT calculations were performed using the Quantum ESPRESSO package. The exchange-correlation interactions were treated within the generalized gradient approximation (GGA) using the Perdew-Burke-Ernzerhof (PBE) functional, and the projector-augmented wave (PAW) method was employed. Van der Waals interactions were included via Grimme\u0026rsquo;s DFT-D3 correction. The Kohn-Sham orbitals were expanded in a plane-wave basis set with kinetic energy and charge density cutoffs of 50 and 400 Ry, respectively. Brillouin-zone integrations were sampled with a (3 \u0026times; 3 \u0026times; 1) Monkhorst-Pack k-point mesh. The energy convergence threshold for electronic self-consistency was set to 10\u003csup\u003e-4\u003c/sup\u003e eV, and structural optimizations were performed until the residual forces on each atom were less than 10\u003csup\u003e-3\u003c/sup\u003e eV \u0026Aring;\u003csup\u003e-1\u003c/sup\u003e. A vacuum spacing of 15 \u0026Aring; was applied to prevent interactions between periodic images.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOther Characterizations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe morphology and elemental composition were studied by a field-emission SEM (FEI Inspect F50) and a TEM (Talos F200X) equipped with an energy dispersive X-ray spectrometer (EDS) system. Crystallographic information was obtained from XRD (Rigaku smartlab (3), Cu-\u003cem\u003eK\u003c/em\u003e\u003csub\u003e\u0026alpha;\u003c/sub\u003e radiation with \u003cem\u003e\u0026lambda;\u003c/em\u003e = 0.15406 nm) with a 2\u003cem\u003e\u0026theta;\u003c/em\u003e range from 20\u0026deg; to 60\u0026deg; at a scan rate of 10\u0026deg; min\u003csup\u003e-1\u003c/sup\u003e. ICP-MS data were obtained by an Aglient 7850 instrument.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe pristine computational models are provided as Supplementary Information. All other data that support the findings of this study are available from the corresponding author upon request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Natural Science Foundation of China (92477119, 12374370, and 22173018);\u0026nbsp;Fundamental Research Funds for the Central Universities (2242025K30023). We acknowledge Southeast University, Key Laboratory of Quantum Materials and Devices, Ministry of Education (Southeast University), and the Center for Fundamental and Interdisciplinary Sciences of Southeast University, for the support in fabrication, measurement and simulations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eQ.H. provided the original idea. Q.H. and T.Q. jointly supervised the project. D.C. designed the experiments, performed the measurements, and drafted the manuscript. G.L. and Q.L. conducted the simulation calculations and analyzed the SERS results. X.Z. and X.F. assisted with the operando UV-Vis measurements and contributed to data analysis and interpretation. Z.S., R.W., Z.L., and R.L. carried out the electrocatalytic measurements, TEM analyses, SEM analyses, and processed the corresponding data. Y.W. provided the initial inspiration for the choice of probe molecules and offered daily guidance. S.C. and L.Y. prepared the schematics in Figures 2 and 3 and provided advice on the visual presentation throughout the manuscript. Q.H. revised the manuscript. All authors wrote, revised and approved the final paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eNitopi, S. \u003cem\u003eet al.\u003c/em\u003e Progress and Perspectives of Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction on Copper in Aqueous Electrolyte. \u003cem\u003eChem. Rev.\u003c/em\u003e \u003cstrong\u003e119\u003c/strong\u003e, 7610\u0026ndash;7672 (2019).\u003c/li\u003e\n\u003cli\u003eKok, J., Albertini, P. P., Leemans, J., Buonsanti, R. \u0026amp; Burdyny, T. Overcoming copper stability challenges in CO\u003csub\u003e2\u003c/sub\u003e electrolysis. \u003cem\u003eNat. Rev. Mater.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 550\u0026ndash;563 (2025).\u003c/li\u003e\n\u003cli\u003eQin, X.-R. \u003cem\u003eet al.\u003c/em\u003e Electroreduction of diluted CO\u003csub\u003e2\u003c/sub\u003e to multicarbon products with high carbon utilization at 800 mA cm\u003csup\u003e-2\u003c/sup\u003e in strongly acidic media. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 4447 (2025).\u003c/li\u003e\n\u003cli\u003eTao, F. \u0026amp; Salmeron, M. Surface restructuring and predictive design of heterogeneous catalysts. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e386\u003c/strong\u003e, 865 (2024).\u003c/li\u003e\n\u003cli\u003eKim, Y.-G., Baricuatro, J. H., Javier, A., Gregoire, J. M. \u0026amp; Soriaga, M. P. The Evolution of the Polycrystalline Copper Surface, First to Cu(111) and Then to Cu(100), at a Fixed CO\u003csub\u003e2\u003c/sub\u003eRR Potential: A Study by \u003cem\u003eOperando\u003c/em\u003e EC-STM. \u003cem\u003eLangmuir\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 15053\u0026ndash;15056 (2014).\u003c/li\u003e\n\u003cli\u003eHan, Z. \u003cem\u003eet al.\u003c/em\u003e Steering surface reconstruction of copper with electrolyte additives for CO\u003csub\u003e2\u003c/sub\u003e electroreduction. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 3158 (2022).\u003c/li\u003e\n\u003cli\u003eGao, X. \u003cem\u003eet al.\u003c/em\u003e Intermediate-regulated dynamic restructuring at Ag-Cu biphasic interface enables selective CO\u003csub\u003e2\u003c/sub\u003e electroreduction to C\u003csub\u003e2+\u003c/sub\u003e fuels. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 10331 (2024).\u003c/li\u003e\n\u003cli\u003eLiu, Q. \u003cem\u003eet al.\u003c/em\u003e Universal Formation, Dynamics, and Reactivities of *CO\u003csub\u003ebridge\u003c/sub\u003e Accompanying Spontaneous Reconstruction of Cu during Electrochemical CO\u003csub\u003e2\u003c/sub\u003e or CO Reduction. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e147\u003c/strong\u003e, 31497\u0026ndash;31508 (2025).\u003c/li\u003e\n\u003cli\u003eLi, Z., Wang, L., Wang, T., Sun, L. \u0026amp; Yang, W. Steering the Dynamics of Reaction Intermediates and Catalyst Surface during Electrochemical Pulsed CO\u003csub\u003e2\u003c/sub\u003e Reduction for Enhanced C\u003csub\u003e2+\u003c/sub\u003e Selectivity. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e145\u003c/strong\u003e, 20655\u0026ndash;20664 (2023).\u003c/li\u003e\n\u003cli\u003eMu, S. \u003cem\u003eet al.\u003c/em\u003e Hydroxyl radicals dominate reoxidation of oxide-derived Cu in electrochemical CO2 reduction. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 3694 (2022).\u003c/li\u003e\n\u003cli\u003eM\u0026ouml;ller, T. \u003cem\u003eet al.\u003c/em\u003e Electrocatalytic CO\u003csub\u003e2\u003c/sub\u003e Reduction on CuO\u003cem\u003e\u003csub\u003ex\u003c/sub\u003e\u003c/em\u003e\u003csub\u003e \u003c/sub\u003eNanocubes: Tracking the Evolution of Chemical State, Geometric Structure, and Catalytic Selectivity using Operando Spectroscopy. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e59\u003c/strong\u003e, 17974\u0026ndash;17983 (2020).\u003c/li\u003e\n\u003cli\u003eGrosse, P. \u003cem\u003eet al.\u003c/em\u003e Dynamic transformation of cubic copper catalysts during CO\u003csub\u003e2\u003c/sub\u003e electroreduction and its impact on catalytic selectivity. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 6736 (2021).\u003c/li\u003e\n\u003cli\u003eLei, Q. \u003cem\u003eet al.\u003c/em\u003e Structural evolution and strain generation of derived-Cu catalysts during CO\u003csub\u003e2\u003c/sub\u003e electroreduction. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 4857 (2022).\u003c/li\u003e\n\u003cli\u003eZhao, Y. \u003cem\u003eet al.\u003c/em\u003e Speciation of Cu Surfaces During the Electrochemical CO Reduction Reaction. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e142\u003c/strong\u003e, 9735-9743 (2020).\u003c/li\u003e\n\u003cli\u003eLong, C. \u003cem\u003eet al.\u003c/em\u003e Regulating reconstruction of oxide-derived Cu for electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction toward n-propanol. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, eadi6119 (2023).\u003c/li\u003e\n\u003cli\u003eChou, T.-C. \u003cem\u003eet al.\u003c/em\u003e Controlling the Oxidation State of the Cu Electrode and Reaction Intermediates for Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction to Ethylene. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e142\u003c/strong\u003e, 2857\u0026ndash;2867 (2020).\u003c/li\u003e\n\u003cli\u003eKim, J. \u003cem\u003eet al.\u003c/em\u003e Spatiotemporal active phase evolution for CO\u003csub\u003e2\u003c/sub\u003e electrocatalysis. \u003cem\u003eJoule\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 3373\u0026ndash;3385 (2024).\u003c/li\u003e\n\u003cli\u003eChen, C. \u003cem\u003eet al.\u003c/em\u003e The \u003cem\u003ein situ\u003c/em\u003e study of surface species and structures of oxide-derived copper catalysts for electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction. \u003cem\u003eChem. Sci.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 5938\u0026ndash;5943 (2021).\u003c/li\u003e\n\u003cli\u003ePhan, T. H. \u003cem\u003eet al.\u003c/em\u003e Emergence of Potential-Controlled Cu-Nanocuboids and Graphene-Covered Cu-Nanocuboids under \u003cem\u003eOperando\u003c/em\u003e CO\u003csub\u003e2\u003c/sub\u003e Electroreduction. \u003cem\u003eNano Lett.\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 2059\u0026ndash;2065 (2021).\u003c/li\u003e\n\u003cli\u003eGrosse, P. \u003cem\u003eet al.\u003c/em\u003e Dynamic Changes in the Structure, Chemical State and Catalytic Selectivity of Cu Nanocubes during CO\u003csub\u003e2\u003c/sub\u003e Electroreduction: Size and Support Effects. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e57\u003c/strong\u003e, 6192\u0026ndash;6197 (2018).\u003c/li\u003e\n\u003cli\u003eYang, Y. \u003cem\u003eet al.\u003c/em\u003e Operando studies reveal active Cu nanograins for CO\u003csub\u003e2\u003c/sub\u003e electroreduction. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e614\u003c/strong\u003e, 262\u0026ndash;269 (2023).\u003c/li\u003e\n\u003cli\u003eLi, Y. \u003cem\u003eet al.\u003c/em\u003e Electrochemically scrambled nanocrystals are catalytically active for CO\u003csub\u003e2\u003c/sub\u003e-to-multicarbons. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e117\u003c/strong\u003e, 9194\u0026ndash;9201 (2020).\u003c/li\u003e\n\u003cli\u003eLee, S. H. \u003cem\u003eet al.\u003c/em\u003e Structural Transformation and Degradation of Cu Oxide Nanocatalysts during Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e147\u003c/strong\u003e, 6536\u0026ndash;6548 (2025).\u003c/li\u003e\n\u003cli\u003eHuang, J. \u003cem\u003eet al.\u003c/em\u003e Potential-induced nanoclustering of metallic catalysts during electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 3117 (2018).\u003c/li\u003e\n\u003cli\u003eLiu, S. \u003cem\u003eet al.\u003c/em\u003e Alkali cation-induced cathodic corrosion in Cu electrocatalysts. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 5080 (2024).\u003c/li\u003e\n\u003cli\u003eYang, Y. \u003cem\u003eet al.\u003c/em\u003e Elucidating Cathodic Corrosion Mechanisms with Operando Electrochemical Transmission Electron Microscopy. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e144\u003c/strong\u003e, 15698\u0026ndash;15708 (2022).\u003c/li\u003e\n\u003cli\u003eKim, I. \u003cem\u003eet al.\u003c/em\u003e Unveiling the reconstruction of copper bimetallic catalysts during CO\u003csub\u003e2\u003c/sub\u003e electroreduction. \u003cem\u003eNat. Catal.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 697\u0026ndash;713 (2025).\u003c/li\u003e\n\u003cli\u003eAmirbeigiarab, R. \u003cem\u003eet al.\u003c/em\u003e Atomic-scale surface restructuring of copper electrodes under CO\u003csub\u003e2\u003c/sub\u003e electroreduction conditions. \u003cem\u003eNat. Catal.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 837\u0026ndash;846 (2023).\u003c/li\u003e\n\u003cli\u003eVavra, J., Shen, T., Stoian, D., Tileli, V. \u0026amp; Buonsanti, R. Real‐time Monitoring Reveals Dissolution/Redeposition Mechanism in Copper Nanocatalysts during the Initial Stages of the CO\u003csub\u003e2\u003c/sub\u003e Reduction Reaction. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 1347\u0026ndash;1354 (2021).\u003c/li\u003e\n\u003cli\u003eZhang, Q. \u003cem\u003eet al.\u003c/em\u003e Atomic dynamics of electrified solid\u0026ndash;liquid interfaces in liquid-cell TEM. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e630\u003c/strong\u003e, 643\u0026ndash;647 (2024).\u003c/li\u003e\n\u003cli\u003eVavra, J. \u003cem\u003eet al.\u003c/em\u003e Solution-based Cu\u003csup\u003e+\u003c/sup\u003e transient species mediate the reconstruction of copper electrocatalysts for CO\u003csub\u003e2\u003c/sub\u003e reduction. \u003cem\u003eNat. Catal.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 89\u0026ndash;97 (2024).\u003c/li\u003e\n\u003cli\u003eEren, B. \u003cem\u003eet al.\u003c/em\u003e Activation of Cu(111) surface by decomposition into nanoclusters driven by CO adsorption. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e351\u003c/strong\u003e, 475\u0026ndash;478 (2016).\u003c/li\u003e\n\u003cli\u003eWang, Y. \u003cem\u003eet al.\u003c/em\u003e Catalyst synthesis under CO\u003csub\u003e2\u003c/sub\u003e electroreduction favours faceting and promotes renewable fuels electrosynthesis. \u003cem\u003eNat. Catal.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 98\u0026ndash;106 (2019).\u003c/li\u003e\n\u003cli\u003eZhao, X. \u003cem\u003eet al.\u003c/em\u003e Plasmonic trimers designed as SERS-active chemical traps for subtyping of lung tumors. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 5855 (2024).\u003c/li\u003e\n\u003cli\u003eHao, Q. \u003cem\u003eet al.\u003c/em\u003e Facile design of ultra-thin anodic aluminum oxide membranes for the fabrication of plasmonic nanoarrays. \u003cem\u003eNanotechnology\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 105301 (2017).\u003c/li\u003e\n\u003cli\u003eZaraska, L., Sulka, G. D., Szeremeta, J. \u0026amp; Jaskuła, M. Porous anodic alumina formed by anodization of aluminum alloy (AA1050) and high purity aluminum. \u003cem\u003eElectrochimica Acta\u003c/em\u003e \u003cstrong\u003e55\u003c/strong\u003e, 4377\u0026ndash;4386 (2010).\u003c/li\u003e\n\u003cli\u003eChen, D. \u003cem\u003eet al.\u003c/em\u003e Unveiling the Cation Effects on Electrocatalytic CO\u003csub\u003e2\u003c/sub\u003e Reduction via Operando Surface‐enhanced Raman Spectroscopy. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 2409569 (2025).\u003c/li\u003e\n\u003cli\u003eWei, Y. \u003cem\u003eet al.\u003c/em\u003e Unveiling Plasmon-Induced Suzuki\u0026ndash;Miyaura Reactions on Silver Nanoparticles via Raman Spectroscopy. \u003cem\u003eACS Catal.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 15043\u0026ndash;15051 (2024).\u003c/li\u003e\n\u003cli\u003eBeaudelot, J. \u003cem\u003eet al.\u003c/em\u003e Photoactive Copper Complexes: Properties and Applications. \u003cem\u003eChem. Rev.\u003c/em\u003e \u003cstrong\u003e122\u003c/strong\u003e, 16365\u0026ndash;16609 (2022).\u003c/li\u003e\n\u003cli\u003eIwamura, M., Takeuchi, S. \u0026amp; Tahara, T. Ultrafast Excited-State Dynamics of Copper(I) Complexes. \u003cem\u003eAcc. Chem. Res.\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 782\u0026ndash;791 (2015).\u003c/li\u003e\n\u003cli\u003eGhosh, S. K., Rahman, D. S., Ali, A. L. \u0026amp; Kalita, A. Surface Plasmon Tunability and Emission Sensitivity of Ultrasmall Fluorescent Copper Nanoclusters. \u003cem\u003ePlasmonics\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 1457\u0026ndash;1468 (2013).\u003c/li\u003e\n\u003cli\u003eSarkar, N. \u0026amp; Radhakrishnan, T. P. Single Source Polychromatic Luminescent and White Emitting Copper Nanoclusters. \u003cem\u003eAdv. Opt. Mater.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 2402451 (2025).\u003c/li\u003e\n\u003cli\u003eYang, Y. \u003cem\u003eet al.\u003c/em\u003e Operando probing dynamic migration of copper carbonyl during electrocatalytic CO2 reduction. \u003cem\u003eNat. Catal.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 579\u0026ndash;594 (2025).\u003c/li\u003e\n\u003cli\u003eProtopopoff, E. \u0026amp; Marcus, P. Potential\u0026ndash;pH diagrams for hydroxyl and hydrogen adsorbed on a copper surface. \u003cem\u003eElectrochim. Acta\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, 408\u0026ndash;417 (2005).\u003c/li\u003e\n\u003cli\u003eHuang, J. \u003cem\u003eet al.\u003c/em\u003e Potential-induced nanoclustering of metallic catalysts during electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 3117 (2018).\u003c/li\u003e\n\u003cli\u003eYoon, A. \u003cem\u003eet al.\u003c/em\u003e Revealing catalyst restructuring and composition during nitrate electroreduction through correlated operando microscopy and spectroscopy. \u003cem\u003eNat. Mater.\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 762\u0026ndash;769 (2025).\u003c/li\u003e\n\u003cli\u003eZhao, Y. \u003cem\u003eet al.\u003c/em\u003e Elucidating electrochemical CO\u003csub\u003e2\u003c/sub\u003e reduction reaction processes on Cu(\u003cem\u003ehkl\u003c/em\u003e) single-crystal surfaces by \u003cem\u003ein situ\u003c/em\u003e Raman spectroscopy. \u003cem\u003eEnergy Environ. Sci.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 3968\u0026ndash;3977 (2022).\u003c/li\u003e\n\u003cli\u003eZhang, Y.-C. \u003cem\u003eet al.\u003c/em\u003e Facet-switching of rate-determining step on copper in CO\u003csub\u003e2\u003c/sub\u003e-to-ethylene electroreduction. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e121\u003c/strong\u003e, e2400546121 (2024).\u003c/li\u003e\n\u003cli\u003eXu, K. \u003cem\u003eet al.\u003c/em\u003e Favoring CO Intermediate Stabilization and Protonation by Crown Ether for CO\u003csub\u003e2\u003c/sub\u003e Electromethanation in Acidic Media. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e62\u003c/strong\u003e, e202311968 (2023).\u003c/li\u003e\n\u003cli\u003eXie, Z. \u003cem\u003eet al.\u003c/em\u003e Surface Facets Reconstruction in Copper‐Based Materials for Enhanced Electrochemical CO\u003csub\u003e2\u003c/sub\u003e Reduction. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 2401530 (2024).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7787217/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7787217/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Surface reconstruction is a common and critical process in copper catalysts during electrochemical CO2 reduction reaction (CO2RR), continuously reshaping their structures and compositions into states distinct from the pristine material. Although recognized as central to catalyst performance, the mechanistic pathways and chemical identities of the dissolved Cu species remain unclear. Employing clean and well-defined Cu nanoparticle arrays as a model platform, we directly visualize morphological evolution and simultaneously track reaction intermediates by in situ surface-enhanced Raman spectroscopy (SERS). We resolve three distinct dissolution-redeposition pathways of Cu catalyst: (i) *CO-mediated Cu+ dissolution and facet-selective redeposition at moderate potentials; (ii) field-assisted leaching of neutral Cu0 species at strongly negative potentials; and (iii) oxidative Cu+ dissolution at open-circuit potential, followed by redeposition into Cu2O cubes enabled by CO2RR-generated active sites. These findings reveal a nonlinear, potential-dependent behavior of Cu reconstruction, and provide guiding principles for directing catalyst evolution through control of adsorbates, potential windows, and bias protocols.","manuscriptTitle":"Mechanistic pathways of copper catalyst reconstruction in CO2 electroreduction","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-17 06:13:27","doi":"10.21203/rs.3.rs-7787217/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"cd592c5f-d50d-4c57-a17f-f3bd4c4f4cb4","owner":[],"postedDate":"October 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":56372478,"name":"Physical sciences/Chemistry/Catalysis/Electrocatalysis"},{"id":56372479,"name":"Physical sciences/Chemistry/Catalysis/Heterogeneous catalysis"}],"tags":[],"updatedAt":"2025-12-10T19:21:36+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-17 06:13:27","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7787217","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7787217","identity":"rs-7787217","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}