ChemicalsCopper chloride dehydrate (99.99%) was purchased from Macklin. Nano-carbon black (99.5%) was purchased from Aladdin. Potassium hydroxide (≥85.0%), sodium hydroxide (≥96.0%), and methanol (≥99.5%) were purchased from the Sinopharm Chemical Reagent Company. Ethanol (≥99.7%) was purchased from General Reagent. Cu NPs (25 nm, 99.0%), CuO (99.0%), and Cu2O (99.0%) were purchased from Sigma-Aldrich. The Nafion solution (Dupont, D-520 dispersion, 5% w/w in water, and 1-propanol) was purchased from Alfa Aesar. Freudenberg H23C9 and H14C9 gas diffusion layer (GDL) and Sigracet 39BC GDL were purchased from the Fuel Cell Store. A hydroxide exchange membrane FAB-PK-130 was purchased from Fumatech. An anion exchange membrane X37-FA was purchased from Sustainion. All chemicals, including precursors, solvents, hydrophobic agents, and ionomers, unless otherwise stated, were used without further purification.Preparation of the CuO nanopowdersThe CuO nanopowders were synthesized via a refined solvothermal method. Initially, 1.02 g of Copper (II) chloride dihydrate (6.0 mmol) and 50 mg of nano-carbon black were dissolved in 30 ml of 2 M sodium hydroxide solution and stirred for 30 min at room temperature. This mixture was then transferred to a 50 ml Teflon-lined stainless-steel autoclave and subjected to solvothermal treatment at 130 °C for 12 h. Post-treatment, the product underwent centrifugation with deionized water and ethanol, each thrice, for purification. Finally, the CuO nanopowders were dried in a vacuum oven at 80 °C for 6 h, yielding a high-purity product.Preparation of gas diffusion electrodes (GDEs)In the fabrication of gas diffusion electrodes (GDE) utilizing the nitrogen-doped CuOx (N-CuOx) pre-catalyst, a catalyst ink was prepared by homogeneously mixing 5 mg of CuO nanopowders with 1 ml of methanol and 30 μl of Nafion solution, followed by sonication to ensure uniform dispersion. This slurry was then carefully applied onto a 2 cm × 2 cm gas diffusion layer (GDL) using a drip-coating technique. The coated GDL was subsequently placed in the chamber of an atmospheric plasma-enhanced cleaning machine. During the plasma treatment process, a plasma ball was generated above the sample, maintained at a microwave power of 500–600 W. The plasma comprised ultra-high purity Argon (Ar, 99.999%) and Nitrogen (N2, 99.999%), with constant pressure and flow rates. The exposure time to the plasma was meticulously controlled to ensure the sample temperature did not exceed 150 °C. This exposure resulted in a color change of the samples, indicative of the plasma’s effect. For comparative purposes, electrodes with the CuO pre-catalyst were also prepared using a similar procedure, excluding the plasma treatment. All subsequent preparation steps were identical to those used for the N-CuOx pre-catalyst GDEs.The prepared GDEs were electrochemically activated by running them for 900 s in a CO gas environment (20 ml min−1) at −0.43 V (versus reversible hydrogen electrode, RHE) in 1 M KOH solution at 25 °C. This activation ensured that the pre-catalysts were fully electro-reduced to a stable state, suitable for CORR catalysis, as indicated by the achievement of a stable current density.The electrode potentials were rescaled to the RHE reference by the following equation:$${E}\,({{\rm{vs}}}.\,{{\rm{RHE}}})={E}\,({{\rm{vs}}}.\,{{\rm{Hg}}}/{{\rm{HgO}}})+0.098\,{{\rm{V}}}+0.0591\times {{\rm{pH}}}$$
(1)
And the potential is presented with a 70% iR correction. The uncompensated solution resistances (RΩ) were measured by extrapolating the electrochemical impedance semi-circle to the high-frequency end, which was ca. 3.5 ± 0.1 Ω for each electrode in 1 M KOH.Structural characterizationScanning electron microscopy (SEM) analysis was conducted on a Hitachi FE-SEM S-4800, operated at an accelerating voltage of 1.0 kV, providing detailed surface morphology insights. high-resolution transmission electron microscopy (HRTEM) images were captured using a JEOL JEM-2100F transmission electron microscope, functioning at 200 kV. Scanning transmission electron microscopy (STEM) investigations were performed on two sophisticated instruments: an FEI Titan Cubed 60-300, operating at a high accelerating voltage of 300 kV, and a JEOL ARM-200F equipped with a cold field emission gun and a CEOS-corrected Cs probe, at an operating voltage of 200 kV. These STEM analyses were complemented by high-resolution energy dispersive X-ray spectroscopy (EDX) utilizing a state-of-the-art super-X detector, enabling precise elemental analysis. X-ray photoelectron spectroscopy (XPS) was carried out on a PHI 5700 ESCA system, employing Al Kα X-ray radiation (1486.6 eV) for excitation. Powder X-ray diffraction (XRD) patterns were acquired using a MiniFlex600 instrument, operating in Bragg–Brentano mode. The instrument was configured with a 0.02° divergence and a scan rate of 0.1° s−1. The surface strain was quantified by HAADF-STEM images with Strain++ software70, utilizing geometric phase analysis (GPA) to map the strain71. The color spectrum in the strain maps ranges from negative (compressive) to positive (tensile) strain values.Operando X-ray absorption fine spectroscopy (XAFS)Operando Cu K-edge X-ray absorption fine structure (XAFS) measurements were executed at the 1W1B beamline of the Beijing Synchrotron Radiation Facility (BSRF), China. These measurements utilized the Quick XAFS (QXAFS) technique, spanning an energy range from 8.8 to 9.2 keV, with a detailed focus on the near-edge region using a fine step size of 0.5 eV. This approach, conducted in fluorescence mode, allowed for an efficient collection of spectra, with each QXAFS spectrum requiring ~40 s (30 s for data acquisition and 10 s for detector repositioning).The study involved the use of an LSN-Cu-loaded gas diffusion electrode (GDE), identical to the one used in electrochemical assessments. This GDE underwent a chronoamperometry process at a potential of −0.53 V (vs. RHE) within a custom-built flow-cell type reactor (Supplementary Fig. 45b). This setup was akin to that used in preference measurements, with the sole variation being the replacement of the outer surface of the gas chamber with Kapton tape. The reactor system comprised a Hg/HgO reference electrode in 1 M KOH, a Fe–S/Ni foam electrode as the anode, and a Fumatech FAB-PK-130 anion exchange membrane. The electrolyte was 1 M KOH solution, and CO (99.99% purity, Air France) was continuously supplied to the gas chamber for CORR studies. XAS data processing was conducted using Athena and Artemis software within the IFEFFIT package suite. And the potential is presented with a 70% iR correction. The uncompensated solution resistances (RΩ) were measured by extrapolating the electrochemical impedance semi-circle to the high-frequency end, which was ca. 3.5 ± 0.1 Ω for each electrode in 1 M KOH.The Fe–S/Ni foam electrodes were fabricated using a solvothermal method. The process began with ultrasonication of Fe–S/Ni foam in a mixture of acid, H2O, and ethanol. A solution containing 4.87 g of Ferric chloride hexahydrate and 7.21 g of sodium sulfide nonahydrate, dissolved in 300 ml of deionized water, was stirred for 2 h at room temperature. Subsequently, this mixture, along with the pre-treated Ni foam, was transferred to a 50 ml Teflon-lined stainless-steel autoclave for solvothermal treatment at 150 °C for 13.5 h. Following this, the sample was centrifuged three times with deionized water and then dried in a vacuum oven at 60 °C for 12 h.For comparative purposes, ex-situ Cu K-edge XAFS measurements were carried out on commercial Cu nanoparticles (NPs) and CuO powders. These reference samples were prepared by uniformly distributing the powders on a strip of 3 M tape, ensuring a standardized approach for comparative XAFS analysis.Operando surface-enhanced Raman spectroscopy (SERS)Operando SERS measurements were expertly conducted using a Horiba Scientific Xplora Raman Microscope, equipped with a water immersion objective (×100) and utilizing a 633 nm laser (numerical aperture, NA = 1.0; working distance, WD = 2.0 mm; LUMPLFLN-60X/W; Olympus Inc., Waltham, MA) and a flow-cell type reactor (Supplementary Fig. 46). Each Raman spectrum was meticulously acquired over a integration time of 5 s, with an averaging of 10 scans to ensure data accuracy and repeatability. The acquired spectra were subsequently analyzed and processed utilizing LabSpec 6.0 software, providing detailed insights into molecular interactions and structural changes.For these measurements, the same working electrode previously utilized for electrochemical performance evaluation was employed. This ensured consistency in the experimental conditions and comparability of the results. The electrochemical setup was complemented by an Ag/AgCl reference electrode in 3 M KCl and a graphite rod as the counter electrode. A 1 M KOH aqueous solution served as the electrolyte, maintaining an optimal environment for electrochemical reactions.During the SERS experiments, CO gas was continuously supplied to the gas chamber, ensuring a stable and controlled environment for the assessment of catalytic processes under realistic operational conditions. This operando approach enables the direct observation and analysis of electrocatalytic phenomena, providing valuable insights into the mechanisms and efficiency of the catalysts under study. The potentials in Raman measurements were converted to values with reference to RHE using the equation:$${E}({{\rm{vs}}}.{{\rm{RHE}}})={E}({{\rm{vs}}}.{{\rm{Ag}}}/{{\rm{AgCl}}})+0.197\,{{\rm{V}}}+0.0591\times {{\rm{pH}}}$$
(2)
where the pH value of the 1 M KOH aqueous solution in this work is 13.6 ± 0.1. And the potential is presented with a 70% iR correction. The uncompensated solution resistances (RΩ) were measured by extrapolating the electrochemical impedance semi-circle to the high-frequency end, which was ca. 3.5 ± 0.1 Ω for each electrode in 1 M KOH.Temperature programed desorption (TPD)GDEs, identical to those used in electrochemical measurements, were subjected to a chronoamperometry process at −0.53 V (vs. RHE) and subsequently dried in a vacuum oven. Following this, both the catalyst and gas diffusion layer (GDL) components of the samples were meticulously ground into a fine powder to facilitate CO desorption measurements. To isolate the specific contribution of the GDL support, a parallel CO desorption assessment was conducted on ground GDL samples under identical conditions.For the CO adsorption studies, a sophisticated TPD apparatus equipped with a thermal conductive detector (AutoChem II 2920) was employed. The catalysts underwent a degassing process at 100 °C under a continuous flow of Argon (Ar) gas, effectively removing any pre-adsorbed gases from the catalyst surface. This process lasted for 1 h, ensuring thorough preparation of the catalysts for subsequent CO adsorption. Following degassing, CO gas was introduced to the system, allowing for ample adsorption onto the catalysts. Excess CO was then purged using Argon.The TPD sequence was initiated under a steady Ar flow at a constant velocity, facilitating the transport of desorbed CO molecules to the detector. This methodology provided a detailed understanding of CO adsorption and desorption dynamics on the catalyst surfaces, crucial for elucidating their catalytic behavior and efficiency in electrochemical processes.Operando attenuated total reflection surface-enhanced infrared absorption spectroscopy (ATR-SEIRAS)ATR-SEIRAS measurements were conducted using a Bruker INVENIO infrared spectrophotometer equipped with a built-in mercury cadmium telluride (MCT) detector. Various Cu-based catalysts coated on an Au/Si substrate were used as the working electrode, with a Hg/HgO electrode and a graphite rod serving as the reference and counter electrodes, respectively (Supplementary Fig. 47a). The Au/Si substrate was prepared as follows: a hemicylindrical Si prism was purchased from IRUBIS GmbH. The Au film on the reflecting plane of the Si prism (Au/Si substrate) was prepared using a two-step wet process: First, the reflecting plane of the Si prism was mechanically polished with 1.0, 0.3, and 0.05 µm Al2O3 powder, sonicated in acetone and water, respectively, soaked in piranha solution, and thoroughly rinsed with Milli-Q water (18.2 MΩ cm). The reflecting plane was then immersed in a 40% NH4F solution for 1.5 min to terminate the Si surface with hydrogen, and subsequently immersed in a plating solution containing 0.015 M HAuCl4, 0.15 M Na2SO3, 0.05 M Na2S2O3, and 0.05 M NH4Cl at 60 °C for 3 min to deposit the Au film.ATR-SEIRAS spectra were acquired at a resolution of 4 cm−1 using unpolarized IR radiation at an incidence angle of ~70°. The electrolyte was 0.1 M KOH, saturated with CO or purged with Ar gas during the experiment. The electrode potential was held at an open circuit potential (OCP), and a background spectrum was recorded. All spectra are presented in absorbance units as −log(I/I0), where I and I0 represent the intensities of the reflected radiation of the sample and the background spectrum, respectively. The electrode potential was varied from −0.40 to −0.75 V vs. RHE, in a stepwise manner. Concurrently, infrared spectra were recorded with a time resolution of 30 s per spectrum at a spectral resolution of 4 cm−1. The potential is presented with a 70% iR correction. The uncompensated solution resistances (RΩ) were measured by extrapolating the electrochemical impedance semi-circle to the high-frequency end, which was ca. 7.5 ± 0.3 Ω for each electrode in 0.1 M KOH.Operando differential electrochemical mass spectroscopy (DEMS)The DEMS experiments were conducted in a custom-designed cell (Supplementary Fig. 47b). A cold trap, cooled with dry ice, was installed between the vacuum chamber and the electrochemical cell to trap water vapor during the experiments and protect the mass spectrometer. Platinum wire and a Hg/HgO electrode were used as the counter electrode and reference electrode, respectively. The working electrode consisted of an Au film sputtered onto a hydrophobic polytetrafluoroethylene membrane, with the catalyst ink airbrushed onto the Au film and dried at ambient temperature before the DEMS experiments. Due to the pressure difference, the online-generated products were drawn downward into the vacuum chamber and subsequently detected by the mass spectrometer. The experiments were performed in 1 M KOH, with the electrolyte saturated by CO bubbling prior to the electrochemical measurements. The catalysts were subjected to cyclic voltammetry (CV) cycles from 0 to −0.9 V (vs. RHE) with no iR correction, at a scan rate of 5 mV s−¹, while the mass signals of the products were recorded.Electrochemical measurementsIn the absence of specific conditions, the CORR performance of various catalysts was systematically evaluated at a standard temperature of 25 °C using a flow cell setup (Supplementary Fig. 45a). This configuration encompassed a gas chamber, a cathodic chamber, and an anodic chamber. The prepared working electrode was strategically positioned between the gas and cathodic chambers, ensuring that the catalyst-coated side faced the cathodic chamber, which had a geometrically active surface area of 1 cm2. For electrochemical consistency, a Fe–S/Ni foam electrode and a Hg/HgO electrode, filled with 1 M KOH, were utilized as the counter and reference electrodes, respectively.Separation of the cathode and anode chambers was achieved using an Anion exchange membrane (AEM, Supplementary Fig. 48), specifically the Fumatech FAB-PK-130 (110–140 μm). The Fumatech FAB-PK-130 membranes were activated for 12 h in 1 M KOH at room temperature before use. This configuration was carefully assembled with the combined catalyst and diffusion layer, the AEM, and the nickel anode, all held in place using PolyTetraFluoroEthylene (PTFE) spacers. This design facilitated the introduction of alkaline electrolytes into the interstitial spaces between the anode and the membrane, as well as between the membrane and the cathode. The electrolyte flow was regulated at 10 ml min−1 using a peristaltic pump.To maintain a consistent and controlled environment, CO gas (99.99% purity, Air France) was supplied to the gas chamber at a constant rate of 20 ml min−1, managed by an Alicat Scientific mass flow controller. The actual flow rate within the system was accurately determined using a bubble flowmeter located at the outlet of the cathodic chamber. This comprehensive setup ensured precise control over the experimental conditions, vital for the reliable assessment of CORR catalytic performance. Potentials are presented with 70% iR corrections. The uncompensated solution resistances (RΩ) were measured by extrapolating the electrochemical impedance semi-circle to the high-frequency end, which was ca. 3.5 ± 0.1 Ω for each electrode in 1 M KOH.Membrane electrode assembly (MEA) electrolysis experiments were conducted at a controlled temperature of 25 °C, employing a custom-built CO electrolyzer with a 5 cm2 reaction area. A specially prepared gas-diffusion electrode (GDE), measuring 2.0 cm× 2.5 cm, functioned as the cathode. To prevent short-circuiting, a PTFE insulator sheet featuring a 5 cm2 window, was strategically affixed to the cathode. Adjacent to this setup, a pre-treated sustainion membrane (X37-FA, 50 μm) was aligned with a Fe–S/Ni foam electrode of identical dimensions. The X37-FA membranes were activated for 12 h in 1 M KOH at room temperature before use. The anolyte, consisting of a 1 M KOH aqueous solution, was continuously circulated at a flow rate of 30 ml min−1 using a pump.On the cathodic side, CO gas at a flow rate of 40 ml min−1, humidified using deionized (DI) water, was introduced into the cathode chamber. Gas products were collected and analyzed via an in-line gas chromatograph equipped with a cold trap. Given the crossover of liquid products, the Faradaic efficiencies (FEs) of these products were computed based on the total amount collected at both anode and cathode.All CO reduction experiments were conducted using an Autolab PGSTAT302N electrochemical workstation, complemented with a 10 A current booster. Reactions were sustained for a minimum of 300 s before product collection to ensure complete electro-reduction of the pre-catalysts to stable CORR catalysts. Gas chromatographs (Agilent Technologies 7890B and Shanghai Ramiin GC 2060), equipped with thermal conductivity (TCD) and flame ionization (FID) detectors, quantified the gaseous products. These were sampled from both the gas chamber’s outlet and the cathode chamber for enhanced accuracy. Liquid products were examined offline via ¹H nuclear magnetic resonance (NMR) analysis (AVANCE III HD 400 MHz), with Dimethyl sulfoxide (Sigma, 99.99%) added as an internal standard. The ¹H spectra were acquired using water suppression through a pre-saturation technique. FEs of liquid products were calculated based on the total amount collected from both anode and cathode chambers, taking into account the crossover of liquids. After obtaining the n-propanol concentration of each sample from NMR quantification, FEn-propanol was calculated based on the following equation:$${{{\rm{FE}}}}_{{{\rm{n}}}-{{\rm{propanol}}}}=\frac{96,485\times 4\times {{\rm{moles}}}\; {{\rm{of}}}\,{{\rm{n}}}-{{\rm{propanol}}}}{\int i{{\rm{d}}}t}$$
(3)
where i is the stabilized total current during electrolysis measurements.Electrochemical active surface area (ECSA) calculationThe ECSAs of catalysts were calculated based on their electrical double-layer capacitor (Cdl), which were obtained from CV plots in a narrow non-Faradaic potential window from 0.14 to 0.20 V (vs. RHE) with no iR correction. The measured capacitive current densities at 0.17 V were plotted as a function of scan rate, and the slope of the linear fit was calculated as Cdl. The specific capacitance was found to be 29 μF cm−2, and the ECSA of the catalyst is calculated from the following equation:$${{\rm{ECSA}}}=\frac{{{C}}_{{{\rm{dl}}}}}{29\upmu {{\rm{F}}}\,{{{\rm{cm}}}}^{-2}}{{{\rm{cm}}}}^{2}$$
(4)
The intrinsic activity was revealed by normalizing the current to the ECSA to exclude the effect of surface area on catalytic performance. The ECSA values of the catalysts are listed in Supplementary Table 8.Cathodic energy efficiency (EE) calculationCathodic EE is calculated assuming the overpotential of anodic oxygen evolution reaction to be zero, which is calculated as follows:$${{\rm{n}}}-{{\rm{propanol}}} \, {{\rm{EE}}}_{{{\rm{half}}}-{{\rm{cell}}}}=\frac{(1.23+(-{E}_{{{\rm{n}}}-{{\rm{propanol}}}})) \times {{\rm{FE}}}_{{{\rm{n}}}-{{\rm{propanol}}}}}{1.23+\left(-{E}\right)}$$
(5)
where E is the applied potential; FEn-propanol is the measured Faradaic efficiency of n-propanol; En-propanol is the thermodynamic potential of the CO-to-n-propanol process, i.e., 0.20 V. This potential is presented with a 70% iR correction. The uncompensated solution resistances (RΩ) were measured by extrapolating the electrochemical impedance semi-circle to the high-frequency end, which was ca. 3.5 ± 0.1 Ω for each electrode in 1 M KOH.Full-cell EE calculationSimilar to cathodic energy EE, full-cell EE is calculated as follows:$${{\mathrm{n}}}-{{{\mbox{propanol}}}} \, {{\mbox{EE}}}_{{{\mbox{full}}}-{{\mbox{cell}}}}=\frac{(1.23{+}(-{E}_{{{\mbox{n}}}-{{\mbox{propanol}}}})) \times {{\mbox{FE}}}_{{{\mbox{n}}}-{{\mbox{propanol}}}}}{{E}_{{\mbox{cell}}}}$$
(6)
where Ecell is the measured cell voltage at a given current density, FEn-propanol is the measured Faradaic efficiency of n-propanol; En-propanol is the thermodynamic potential of the CO-to-propanol process, i.e., 0.20 V.DFT calculationsThe simulations were conducted utilizing the framework of density functional theory (DFT), employing the Perdew–Burke–Ernzerhof (PBE) density functional72, as implemented in the Quantum Espresso ab initio simulation package73. For wavefunction and electron density representation, a plane wave basis set was used, with kinetic energy and charge density cutoffs set at 40 and 240 Ry, respectively. The core electron regions and core–valence interactions were described through Vanderbilt ultrasoft pseudopotentials74. Three models were chosen to represent the Cu0 surface, the Cu0/Cu+ surface, and the N-doped Cu0/Cu+ surface, as shown in Supplementary Fig. 27. The Cu0 surface was constructed by a three-layer 4 × 4 Cu(111) surface. The Cu0/Cu+ surface was constructed by reducing one-quarter of a three-layer 2 × 2 Cu2O(111) surface, which is replaced by the Cu(111) surface. The N-doped Cu0/Cu+ surface was constructed by substituting an N atom for an O atom in the Cu0/Cu+ surface75. The oxygen formation energy was computed as follows:$${\Delta }_{{\mbox{f}}}{E}={E}_{{\mbox{surf}}+{\mbox{Ovac}}}-{E}_{{\mbox{surf}}}+{\frac{1}{2}{E}_{{\mbox{O}}}}_{2}$$
(7)
The adsorption energies for the OCCO* and OCCOCO* species were calculated using the following equation:$${\Delta {E}}_{{\mbox{ads}}}={E}_{{\mbox{surf}}+{\mbox{ads}}}-{E}_{{\mbox{surf}}}-n{E}_{{\mbox{CO}}}$$
(8)
where the Esurf represents the adsorption energy of the pristine surface, the Esurf+ads denote the energy of the surface with the adsorbate, the Esurf+Ovac corresponds to the surface energy with an oxygen vacancy, and the ECO is the energy of an adsorbed *CO molecule.
High-efficiency C3 electrosynthesis on a lattice-strain-stabilized nitrogen-doped Cu surface
