Sequential co-reduction of nitrate and carbon dioxide enables selective urea electrosynthesis

Despite the recent achievements in urea electrosynthesis from co-reduction of nitrogen wastes (such as NO3−) and CO2, the product selectivity remains fairly mediocre due to the competing nature of the two parallel reduction reactions. Here we report a catalyst design that affords high selectivity to urea by sequentially reducing NO3− and CO2 at a dynamic catalytic centre, which not only alleviates the competition issue but also facilitates C−N coupling. We exemplify this strategy on a nitrogen-doped carbon catalyst, where a spontaneous switch between NO3− and CO2 reduction paths is enabled by reversible hydrogenation on the nitrogen functional groups. A high urea yield rate of 596.1 µg mg−1 h−1 with a promising Faradaic efficiency of 62% is obtained. These findings, rationalized by in situ spectroscopic techniques and theoretical calculations, are rooted in the proton-involved dynamic catalyst evolution that mitigates overwhelming reduction of reactants and thereby minimizes the formation of side products.

The color reagents were prepared as follows: A. Acid-ferric solution: 50 mL of concentrated H3PO4,150 mL of concentrated H2SO4 and 300 mL of distilled water were mixed, and then 50 mg of FeCl3 was dissolved in the above solution.B. Diacetyl monoxime thiosemicarbazide (DAMO-TSC) solution: 2.5 g of DAMO and 50 mg of TSC were dissolved in distilled water and diluted to 500 mL.A series of standard urea solutions with concentrations of 0 ppm, 0.2 ppm, 0.5 ppm, 1.0 ppm, 2.0 ppm and 5.0 ppm in 0.1 M KHCO3 and 0.1 M KNO3 electrolyte were prepared.The following procedure was conducted: 1 mL of standard urea solution, 2 mL of acidferric solution and 1mL of DAMO-TSC solution were mixed and heated under 100 °C for 15 min.The absorbance measurement was performed at λ = 525 nm with an ultraviolet-visible spectrophotometer.The calibration curve (y = 0.260, x = 0.051, coefficient of determination, R 2 = 0.999) showed a good linear relationship of absorbance values with the urea concentration in three independent calibrations.The inset shows the pink solutions with different urea concentrations.The Griess reagent was prepared as follows: 1.0 g of sulfonamide, 0.1 g of N-(1-naphthyl) ethyldiamine dihydrochloride, and 2.94 L of H3PO4 were mixed within 50 mL deionized water.Subsequently, a series of NO2 − standard solution (0 ppm, 0.2 ppm, 0.5 ppm, 1.0 ppm, 2.0 ppm, 5.0 ppm) were prepared.
For the colorimetric assay, 1.0 mL of NO2 − solution, 1 mL of Griess reagent and 2 mL of H2O were mixed and reacted for 10 min.The absorbance was measured by ultraviolet-visible spectroscopy at 540 nm.The color reagent was prepared by dissolving para-(dimethylamine) benzaldehyde (5.99 g) with a mixture of 30 mL HCl (12 mol L -1 ) and ethanol (300 mL). 2 mL of standard hydrazine solution in different concentrations was mixed with 2 mL of color regent.After 20 min, the absorbance of the colored solutions was measured at 455 nm by UV-vis.Then, 2 mL of the electrolyte after electrolysis was mixed with 2 mL of color regent.

bSupplementary Figure 10 .Supplementary Figure 11 .Supplementary Figure 12 .Supplementary Figure 13 .SupplementaryFig. 4 16 . 19 .
Supplementary Fig.2SEM images of NG and CuSAs/NG.EPR spectra of NC and Cu1/NC.As the nitrogen atom can redistribute the extra electron to adjacent carbon via the delocalized π-conjugated network of carbon layer, an EPR signal at g = 2.003 observed in nitrogen-doped carbon can be assigned to the unpaired electron on the carbon atom.This result is well in line with Raman spectrum and STEM observation. 1 Raman spectrum of NC.Two first-order Raman peaks centered at 1320 cm -1 and 1586 cm -1 can be assigned to D and G bands, respectively.The G band originates from the ordered sp 2 bonded carbon, representing the formation of graphitic carbon.The D band arises from the disordered sp 3 carbon in the defects.The ratio of integrated intensity of D and G bands (ID/IG) is ~1.13 for NC, indicating abundant structural defects and disordered carbon. 2 N2 adsorption and desorption isotherm for NC.The BET surface area of NC is ~879.59m 2 g −1 .Inset: pore size distribution for NC.N2 adsorption and desorption isotherm for Cu1/NC.The BET surface area of Cu1/NC is ~798.46m 2 g −1 .Inset: pore size distribution for Cu1/NC.Full XPS spectrum (a) NG and (b) CuSAs/NG (a) High-resolution Cu 2p XPS spectrum of Cu1/NC.(b) Cu LMM spectrum of Cu1/NC.Supplementary Figure 17.Free energy diagram of hydrogen binding on NC at 0 V versus RHE, and the corresponding atomic structures.Color code: N, blue; C, black; H, yellow.Quantification of urea by the diacetyl monoxime method 3 in 0.1 M KHCO3 and 0.1 M KNO3 electrolyte.(a) UV-vis absorption spectra.(b) Standard curve of urea concentration.

Supplementary Figure 20 .Supplementary Figure 21 .
Quantification of NH3 by the indophenol blue method 4 in 0.1 M KHCO3 and 0.1 M KNO3 electrolyte.(a) UV-vis absorption spectra.(b) Standard curve of NH3 concentration.The color regents were prepared as follows: A. Oxidation solution: 0.75 M NaOH and sodium hypochlorite (available chlorine, 4.00-4.99%);B. Coloring solution: 0.4 M sodium salicylate and 0.32 M NaOH; C. Catalyst solution: 0.1 g Na2[Fe(NO)(CN)5]•2H2O diluted in 10 mL deionized water.A series of standard NH4Cl solutions with concentrations of 0.0, 0.2, 0.5, 1.0, 2.0, 4.0, and 10.0 μg mL −1 in 0.1 M KHCO3 and 0.1 M KNO3 electrolyte were prepared.The following procedure was conducted: 4 mL standard solutions were separately mixed with 50 µL of oxidation solution, 500 µL of coloring solution, and 50 µL of catalyst solution.The absorbance measurement was performed at λ = 655 nm with an ultraviolet-visible spectrophotometer.The calibration curve (y = 0.260, x = 0.051, coefficient of determination, R 2 = 0.999) showed a good linear relationship of absorbance values with the NH3 concentration in three independent calibrations.The calibration curve was used to calculate the NH3 concentration.The inset shows the green solutions with different NH3 concentrations.Quantification of NO2 − by Griess test 5 in 0.1 M KHCO3 and 0.1 M KNO3 electrolyte.(a) UV-vis absorption spectra.(b) Standard curve of NO2 − concentration.

Supplementary Figure 23 .Supplementary Figure 24 .Supplementary Figure 35 . 7 Supplementary Figure 38 .
UV-vis absorption spectra of standard solutions containing 2 ppm urea with various NO2 − concentrations in DAMO-TSC measurements.The results indicate that the influence of NO2 − on urea determination is trivial when the NO2 − concentration is relatively low.Performance of urea synthesis on NC and Cu1/NC.Chronoamperometric curves of (a) NC and (b) Cu1/NC at different potentials.The insets show the color of the solutions after co-reduction.Individual CO2RR experiments on NC and Cu1/NC.(a and b) Chronoamperometric curves of (a) NC and (b) Cu1/NC at different potentials.The inset in (a) shows that the color of the solution determined by the diacetyl monoxime method after individual CO2RR is unchanged, suggesting no urea formation during the CO2RR process over NC. (c and d) Faradaic efficiencies and total current densities on (c) NC and (d) Cu1/NC at different potentials.LSV curves of individual NtrRR and individual CO2RR on NC.The onset potentials for NtrRR and CO2RR are estimated to be −0.32 and −0.50 V, respectively.Supplementary Figure 39.ATR-SEIRAS spectra for Cu1/NC under different applied potentials during co-reduction of CO2 and NO3 − .