Chemicals and materials

All reactants are analytical grade without purification as them received. Ammonium bifluoride (NH₄HF₂), copper nitrate trihydrate (Cu(NO₃)₂ ·3H₂O), ammonium chloride (NH₄Cl), ammonium bromide (NH₄Br), ammonium iodide (NH₄I), tert-butylalcohol (t-BuOH), Tetramethylammonium hydroxide solution (TMA⁺OH^–) were purchased from Beijing InnoChem Science&Technology Co., Ltd. Ethanol (C₂H₆OH), potassium hydroxide (KOH), sodium hydroxide (NaOH), N,N-dimethylformamide (DMF), 25% ammonia solution (NH₃.H₂O) were purchased from Sinopharm Chemical Company. Nafion® D-521 dispersion was purchased from Alfa Aesar. The GDL (SGL, 39BB), nickel foam (1.0-mm thickness), anion exchange membrane (fumasep® FAA-3-50) were purchased from Fuel Cell Store. The ultrapure water involved in experiments was prepared with a resistivity of 18.2 MΩ.

Preparation of Cu(OH)F

Cu(OH)F was prepared according to previous report²⁹. Typically, 114.1 mg (2 mmol) NH₄HF₂ was dispersed in 50 mL DMF with vigorous stirring at room temperature for at least 40 min, then the solution correspondingly changed to blue after 483.2 mg (2.0 mmol) Cu(NO₃)₂·3H₂O was introduced and stir for 30 min. The homogeneous liquid was poured into 100 mL hydrothermal reactor and sealed, then heated at 160 °C for 4 h. The green powder was collected by employing centrifuge process with ethanol and water washing in turn three times after it cooled down. After dried at 60 °C in vacuum, Cu(OH)F is obtained.

Preparation of Cu(OH)X (X = Cl, Br, I)

Cu₂(OH)₃Cl, Cu₂(OH)₃Br, and CuI were fabricated also as previously reports with some modifications²⁹. Generally, 300 mg NH₄Cl, NH₄Br, and NH₄I were respectively added in three-neck flask with a reflux condenser followed by placed 50 mg Cu(OH)F. Then 50 mL C₂H₅OH/H₂O (v/v, 49:1) was filled to form suspension liquid. After stir 1 h, the mixture was heated to 80 °C for 36 h. Through similar procedure of purification mentioned above, the concentrated precursor finally obtained by dried in vacuum at 60 °C.

Preparation of Cu(OH)₂ precursor

Cu(OH)₂ precursor could be prepared through a typical method³⁶. In total, 1.3 g Cu(NO₃)₂ ·3H₂O was dissolved in 100 mL water, then 30 mL 0.15 M NH₃.H₂O was also introduced in above Cu²⁺ solution. Subsequently, Cu(OH)₂ precipitate was presented after 10 mL 1 M NaOH solution added in it with stirring for 30 min. Through filtration and vacuum drying, blue solid product was harvested.

Preparation of Cu–F gas diffusion electrode

The working electrode was prepared through in situ electro-derivation of the relevant precursor on GDL surrounding 1 M KOH. Here, take the process of fabricating Cu–F GDE as an example. 5 mg Cu(OH)F precursor was dispersed in 0.75 mL C₂H₅OH/H₂O (v/v, 1:1) with 25 μl nifion binder, then uniform suspension ink was formed by ultrasonic dispersion method within 1 h. Afterward, all ink spray onto 0.5 × 2.0 cm GDL on the top of a heating plat at 65 °C to evaporate solvent with mas loading controlled at 1.0–1.2 mg. For 5 × 5 cm GDE, Cu(OH)F mass loading was about 25.0–30.1 mg. GDL with loading Cu(OH)F precursor as working electrode was in situ reduced in a flow cell and was immersed in Ar (30 mL min⁻¹) and 1 M KOH at 1.6 V (vs. Ag/AgCl) for 300 s. The obtained catalyst was labeled as Cu–F.

Preparation of Cu-X (X = Cl, Br, I) and Cu NP gas diffusion electrode

Cu-X GDEs were prepared as the same procedures as those of Cu–F. The GDL was loaded Cu₂(OH)₃Cl, Cu₂(OH)₃Br, CuI, and Cu(OH)₂ precursor, respectively. Then the catalysts were marked as Cu–Cl, Cu–Br, Cu–I, Cu NP after in situ electroreduction.

Electrochemical measurements

CHI 1130c and 1120c were employed and coupled with a typical flow cell consisted of gas chamber, cathodic chamber, and anodic chamber in all experiments. Generally, GDL with catalyst as a working electrode, nickel foam as a counter electrode, both chambers were separated by an anion exchange membrane, and Ag/AgCl electrode as a reference electrode constituted electrocatalytic system. Ar and 70% C₂H₂ gas rate were set as 30 mL min⁻¹, otherwise mentioned in this work and electrolyte flow rate was set at 1 mL min⁻¹. The single-cell ESAE experiments were conducted at different potential utilizing i–t curve, then the obtained gas product was directly provided access to gas chromatography to quantitatively analyze component and liquid product was analyzed via ¹H NMR spectroscopy by mixing 500 μl sample with 200 μl D₂O and 0.1 μL DMSO. All potentials of LSV curvy were converted to the RHE scale according to Eq. (1) without solution resistance compensation.

For a three-electrode flow cell tandem system, mainly consisted of fore-cell (1 cm²) and post-cell (25 cm²), 70% C₂H₂ feed gas flew into the fore-cell then immediately transformed into C₂H₄ and C₂H₂ mixed gas at 500 mA cm⁻² and 6 mL min⁻¹. Afterward, outlet gas of fore-cell was served as feed gas of post-cell which the connecting line between them was kept to the minimum to shorter the dead volume. Residual C₂H₂ could be completely converted to C₂H₄ at 20 mA cm⁻² and 6 mL min⁻¹ in post-cell. Individual pumps provided 1 M KOH electrolyte to tandem system as fore-cell at 1 mL min⁻¹ and post-cell at 5 mL min⁻¹. The whole process carbon balance reached 97–99%.

For a two-electrode flow cell system, assembled as above with larger geometric area, chronopotentiometry was introduced to purify ethylene through 25 cm² Cu–F GDE at 20 mL min⁻¹ and 40 mA under ethylene-rich feed gas (1% C₂H₂, 20% C₂H₄, Ar compensation). Electrolyte flow rate set at 5 mL min⁻¹. The R_u resistances at working conditions in electrode system are listed in the Supplementary Table S4.

All the electrochemical performances are presented without the IR compensation, except the stability test in the 1 cm² flow cell (Fig. 2e). As the flow cell (1 cm²) stability test, the curvy was compensated at 200 mA with the solution resistance was about 9.4 Ω.

Performance assessment

Gas products (C₂H₄, C₂H₂, C₄, H₂) were analyzed through gas chromatography (Panna, A6) coupled with a FID detector and a TCD detector. The Plot Al₂O₃ column separated ethylene, acetylene and C₄, while Porapak Q and Molecular Sieve 5 A columns separated H₂. An external standard method was applied to estimate the concentration of component of gas products. The partial current density of products was calculated as Eq. (2)³⁷

Where, v: the flow rate of feed gas, mL s⁻¹,

δ: gas product concentration calculated by calibration curve,

n: the number of electrons transferred of species,

F: Faraday’s constant, 96,485 C mol⁻¹,

V_m: 24 L mol⁻¹.

Faraday efficiency was calculated as Eq. (3)³⁷

Where, J_total: the total current density,

Q_total: the charge number.

The formation rate of every component was based on Eq. (4)²⁹

Where, t: the electrolysis time (h) corresponding to Q_total,

S: the geometric area of the working electrode (cm²).

The calculation of C₂H₂ conversion and C₂H₄ selectivity was premise on carbon balance and according to Eqs. (5)²² and (6)¹⁹

Where, c_feed: concentration of C₂H₂ feed gas,

\(c{\prime}\), \({c}_{{C}_{2}{H}_{6}}\), \({c}_{{c}_{4}}\): the concentration of C₂H₂, C₂H₆, C₄ in gas products.

H₂ generated in ESAE process was estimated as Eq. (7)¹⁹

Where, v_out: the H₂ volume in gas products,

v_feed: the volume of feed gas.

Theoretical conversion current was premise on entirely C₂H₂ conversion based on Eq. (8)¹⁹

Where, P: the atmospheric pressure, 101.3 × 10³ Pa,

R: the molar gas constant, 8.314 J (mol K)⁻¹,

T: the temperature, 293.15 K,

v_x: the velocity of acetylene in mixed gas. As simulated feed gas (1% C₂H₂, 20% C₂H₄) at 20 mL min⁻¹, the v_x could be calculated as 3.3 × 10⁻⁶ S⁻¹ and theoretical conversion current was evaluated as 26.7 mA. When the velocity of feed gas came to 10, 30, 40, and 50 mL min⁻¹, v_x was calculated as 13.4, 40.1, 53.4, and 66.8 mA, respectively.

Characterizations

The X-ray diffraction (XRD) patterns were performed on Bruker D8 advance diffractometer with Cu Kα radiation. X-ray photoelectron spectra (XPS) were performed on a ESCALAB Xi⁺ photoelectron spectrometer with monochromatic Al Kα X-rays to verify the valence state of Cu and halogen. Casa XPS was introduced to analyze spectra calibrated by the C 1 s spectrum (284.8 eV). Scanning electron microscopy (SEM) was harvested by Apreo S instrument to reveal morphology of catalysts. Transmission electron microscopy (TEM), high-resolution TEM (HRTEM) images, the corresponding energy-dispersive X-ray spectroscopy (EDS), and selected area electron diffraction (SAED) were measured on a FEI Talos F200X. Aberration-corrected HAADF-STEM images were performed on a JEOL JEMARM200F TEM/STEM system. The X-ray absorption fine structure spectra were collected at the Beijing Synchrotron Radiation Facility (BSRF) in China. The acquired EXAFS data were extracted and processed according to the standard procedures using the ATHENA module implemented in the IFEFFIT software packages, and copper foil and copper oxide were used as references to identify elaborate valence and coordination environment of Cu in catalysts. Operando Raman spectra were measured in a three-electrode observable window electrochemical cell with a counter electrode of Pt wire and Ag/AgCl under controlled potentials in 1 M KOH electrolyte, and a controlled active area of 0.384 cm² by an insulation layer on carbon paper sprayed with 1 mg Cu(OH)F was used as the working electrode. Raman spectra were collected using a Raman spectrometer (Horiba labRAM HR Evolution) by a 532 nm laser focusing on the organic materials distributed on sample surface after precursor in situ derivation under Ar then switched to C₂H₂.

Computational details

In DFT calculations, a 2 × 2 supercell of the Cu (111) slab model was constructed. Subsequently, a fluorine (F) atom was introduced onto the hollow site of the Cu (111) surface, resulting in the formation of Cu (111)-F. To prevent interactions between images, a vacuum layer with a thickness of 15 Å along the z-direction was implemented. Structural optimization calculations were practiced via the Vienna Ab-initio Simulation Package (VASP) with the projector augmented wave (PAW) method. The Perdew–Burke–Ernzerhof (PBE) functional, in conjunction with the DFT-D3 correction, was employed to handle the exchange function. The plane-wave basis cut-off energy was set at 450 eV. For geometry and lattice size optimization, Brillouin zone integration utilized a Gamma k-point mesh of 3 × 3 × 1. Self-consistent calculations adhered to a convergence energy threshold of 10^-5eV. Equilibrium geometries and lattice constants were optimized, with a maximum stress on each atom kept within 0.02 eV Å^-1.

In the computation of Gibbs free energy, the hydrogen adsorption model was constructed using the computational hydrogen electrode (CHE) model. The C₂H₂ hydrogenation steps to C₂H₄ were delineated as follows:

* + C₂H₂ → *C₂H₂

*C₂H₂ + H⁺ + e⁻ → *C₂H₃

*C₂H₃ + H⁺ + e⁻ → *C₂H₄

*C₂H₄ → * + C₂H₄

The hydrogen combination proceeded through these steps:

* + H⁺ + e⁻ → *H

*H + H⁺ + e⁻ → * + H₂

While the water dissociation process involved the following steps:

* + H₂O → *H₂O

*H₂O → *H–OH

*H–OH → *H + OH⁻

The Gibbs free energy of the H atom was computed based on H₂ → H⁺ + e⁻, where G(H⁺) = 1/2 G(H₂). Entropies of free molecules H₂ and H₂O were referenced to the NIST database, while those of free molecules C₂H₂ and C₂H₄ were obtained from the vaspkit interface. The Gibbs free energy of intermediates was calculated as G = E + E_zpe – TS, where E, E_zpe, and S represent the energy, zero-point energy, and entropy of surface adsorbing intermediates, respectively. In addition, the Kelvin temperature T was set at 298.15 K, with both E_zpe and TS acquired through the vaspkit interface. The pH value of 14 was set to simulate the reaction conditions.