Size-effect induced controllable Cu⁰-Cu⁺ sites for ampere-level nitrate electroreduction coupled with biomass upgrading

Ammonia (NH₃), as a versatile industrial feedstock, is widely used for the production of nitrogen-based fertilizers, chemicals, pharmaceuticals, and polymers. Industrial NH₃ production relies on the Haber-Bosch process, under high temperature and high pressure (i.e. 350–450 °C and 150–200 bar)¹. The carbon-intensive and highly energy-consuming Haber-Bosch process consumes about 1% of the world’s annual energy output and generates 1.3% of global carbon dioxide emissions^2,3. It is urgent and necessary to develop environmentally sustainable and affordable routes for NH₃ synthesis. Recently, the nitrate (NO₃⁻) electrochemical reduction reaction (NO₃⁻RR) has attracted increasing attention to produce NH₃ under mild conditions⁴. However, the performance of NO₃⁻RR is hindered by the sluggish kinetics, which involves an eight-electron transfer, and the complex reaction pathways with multiple by-products, such as NO₂⁻, N₂, N₂H₄, and H₂⁵. Moreover, the competitive hydrogen evolution reaction (HER) also hampers the Faradaic efficiency⁶. Thus, it is crucial to explore and develop a highly efficient electrocatalyst to compress the HER process and enable the desired conversion of NO₃⁻ into high-valued NH₃.

Recently, oxide-derived (OD) Cu catalyst has been widely studied due to excellent NO₃⁻RR activity towards valuable NH₃⁷. Mechanistic studies have demonstrated that OD Cu catalyst would be electrochemically reduced to unsaturated Cu species under negative potential, and the synergistic Cu⁰–Cu⁺ sites is regarded as the active species towards NH₃ synthesis during the NO₃⁻RR process^8,9. Therefore, numerous researches has been conducted to synthesize OD Cu catalyst with Cu⁰–Cu⁺ sites, and revealed the synergistic mechanism in the NO₃⁻RR process^9,10,11,12. The studies also showed that the interfacial motifs would determine the adsorption mode of nitro-intermediates and protons to significantly influence nitrogen selectivity^10,11. However, the reaction mechanism and individual roles of interfacial Cu⁰–Cu⁺ sites are still unclear. One of the big obstacles is that it is challenging to effectively regulate the interfacial Cu⁰–Cu⁺ structure, since the structural motif of OD Cu is unstable under negative potential. Some studies have be devoted to control interfacial structure of Cu⁰−Cu⁺ sites by the strong metal-support interaction¹³, while the electronic interaction between support and active Cu sites would bring to inevitable interference to identify the individual roles of interfacial Cu⁰–Cu⁺ sites. Therefore, with the goal of enhancing the NO₃^-RR performance of Cu-based electrocatalysts, it is of great necessary to understand the interface behavior and reaction mechanism by controllable Cu⁺–Cu⁰ sites.

Herein, we developed a feasible and efficient strategy to construct tunable Cu⁺–Cu⁰ interfacial motifs by modulating the size effects of Cu₂O nanocube catalysts. We revealed the design principle of Cu⁺–Cu⁰ pairs by correlating the macroscopic particle size with the microscopic localized coordinated structure properties by in situ electrochemical Raman and X-ray absorption near edge structure (XANES) characterization. Based on the controllable construction of the Cu⁺–Cu⁰ interfacial structure, we elucidated the individual and synergic roles of Cu⁺ and Cu⁰ sites during the NO₃⁻RR process by combining electrochemical measurement and DFT calculations. Then, we developed an efficient paired-electrolysis system to simultaneously achieve the efficient production of NH₃ and 2,5-furandicarboxylic acid (FDCA) by coupling the cathodic NO₃⁻RR process and anodic HMF electrochemical oxidation reaction (HMFOR) process at an industrially relevant current density, while maintaining high Faradaic efficiencies (FE_NH3 75.6%, FE_FDCA 71.2%), yield rates of NH₃ (5.20 mmol h⁻¹ cm⁻²) and FDCA (0.47 mmol h⁻¹ cm⁻²), and long-term operational stability (20 h) in a 100 cm² anion exchange membrane (AEM) electrolyzers, indicating promising practical applications. The techno-economic analysis demonstrates the potential of this system.

Structural characterization

The Cu₂O nanocubes with different particle sizes are synthesized through the liquid phase reduction method, by adjusting the localized pH environment of the synthetic solution¹⁴, three kinds of Cu₂O catalysts with different particle sizes were obtained, denoted as small Cu₂O nanocube (S-Cu₂O), medium Cu₂O nanocube (M-Cu₂O) and large Cu₂O nanocube (L-Cu₂O), respectively (Supplementary Fig. 1). The Cu⁺–Cu⁰ sites is controllably constructed by the in situ electrochemical reduction of Cu₂O nanocube precursors. The morphology and particle size of the Cu₂O nanocube precursors were characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), both Cu₂O sample exhibit cube-like nanocrystals (Fig. 1a–c), and the average size are concentrated in ~166 nm (S-Cu₂O), ~235 nm (M-Cu₂O) and ~760 nm (L-Cu₂O), respectively (Fig. 1d). Energy dispersive spectroscopic (EDS) elemental mapping further revealed that Cu and O elements are evenly distributed in the Cu₂O nanocubes and the ratio of Cu to O elements is close to 2:1. (Supplementary Fig. 2-4). The crystal structure of the Cu₂O nanocube was confirmed by high-resolution transmission electron microscopy (HRTEM) and X-ray diffraction (XRD). The lattice distance of 0.242 nm was found in the HR-TEM images (Supplementary Figs. 2–4), indicating that the dominant lattice plane is not influenced by the particle size, and (111) lattice plane is dominated on both Cu₂O samples. The XRD patterns confirmed the Cu₂O cubic phase (JCPDS: 78-2076) on both samples without the impurity phase (Fig. 1e).

a–c SEM images of S-Cu₂O, M-Cu₂O, and L-Cu₂O. Inset is the corresponding TEM images. d Statistical size distribution of nanoparticles of S-Cu₂O, M-Cu₂O and L-Cu₂O. e XRD patterns of S-Cu₂O, M-Cu₂O, and L-Cu₂O. f EPR patterns of S-Cu₂O, M-Cu₂O and L-Cu₂O. Source data for d–f are provided as a Source Data file.

Full size image

X-ray photoelectron spectroscopy (XPS) was used to investigate the chemical states and electronic structure of Cu and O on the various Cu₂O nanocubes (Supplementary Fig. 5). In Cu 2p XPS spectrum, there are two peaks in both Cu₂O samples at 952.42 and 932.6 eV, which are attributed to Cu⁺/Cu⁰ species (Supplementary Fig. 6)¹⁵. Further, the Auger electron spectroscopy (AES) was employed to distinguish Cu⁺ and Cu⁰ (Supplementary Fig. 7), it was found that Cu⁺ (~569.6 eV) characteristic Auger peaks dominated on the Cu₂O surface, the signal of Cu⁰ was not detected on the AES spectrum⁹. The content of surface oxygen species was identified by the O 1s XPS spectra (Supplementary Fig. 8), it is interesting that the higher content of chemisorbed oxygen groups (~531.0 eV), which belonged to defect-oxide groups, was found on the M-Cu₂O and S-Cu₂O sample¹⁶. The concentration of defect sites was further evaluated by the electron paramagnetic resonance (EPR), a representative EPR signal of g = 2.003 was found in both Cu₂O samples (Fig. 1f), revealing the electron trapped by oxygen vacancies⁹. The sequence of signal intensity is S-Cu₂O > M-Cu₂O > L-Cu₂O, indicating the higher content of oxygen vacancies on the small-size Cu₂O nanocubes, which is consistent with the XPS results. According to the principle of surface chemistry, the nanosized material not only shows a higher specific surface area, but also exposes higher content of unsaturated coordination sites, such as grain boundaries and edge sites, on the surface, and the concentration of unsaturated coordination sites is negatively related to the crystal size of materials. Therefore, it is rational speculate that the negative correlation of the concentration of defect sites and the particles size of Cu₂O nanocubes.

Then, the Cu⁰–Cu⁺ sites was constructed by a constant potentiostatic reduction on Cu₂O precursor in 1 M KOH and 0.1 M NO₃^- solution (pH =13.8) under a negative potential of −0.6 V vs RHE, ensuring to acquire the stable Cu-Cu⁺ interfacial sites. The morphology structure and composition was observed in the SEM and TEM images, it was found that the Cu₂O catalysts retained the morphology of the nanocube, while the surface roughness increased after the electrochemical reduction process (Supplementary Figs. 9–12). Then, the structural evolution is investigated by the XRD patterns and HRTEM images, the results show that the crystal structure of Cu₂O and Cu could be simultaneously observed on both Cu₂O samples, indicating that the Cu₂O precursor is partially transferred to metallic Cu species for the formation of Cu/Cu₂O interface structure during the electrochemical reduction process (Supplementary Figs. 10–13), the as-obtained Cu₂O catalysts with Cu/Cu₂O interface structure is denoted as S-Cu/Cu₂O, M-Cu/Cu₂O, and L-Cu/Cu₂O, respectively.

Further, the in situ Raman spectroscopy was conducted to identify the dynamic phase transformation during the electrochemical reduction process. As shown in Fig. 2a and Supplementary Fig. 14, the characteristic peaks at 147, 218, 415, and 639 cm⁻¹ were observed at the open circuit potential, which is attributed to the T_1u, E_u, A_u, and T_2g structural mode of Cu₂O precursors¹⁷. When the bias potential is tuned to −0.6 V_RHE, the signal of Cu₂O is rapidly dropped in the first two minutes and gradually tends to be stable, suggesting the structural evolution of Cu₂O to Cu/Cu₂O interface structure. The relative intensity of the Cu₂O signal could be used for evaluating the relative content of Cu₂O and identifying the Cu/Cu₂O interfacial structure after the electrochemical reduction process. Interestingly, the relative intensity follows the order of S-Cu/Cu₂O < M-Cu/Cu₂O < L-Cu/Cu₂O (Fig. 2b), indicating that the small particle size Cu₂O (S-Cu₂O), which owns a high concentration of defect sites, is more tended to transform to metallic Cu under electroreduction process.

a In situ Raman spectra collected during the electrochemical reduction process on various Cu₂O samples. b The relationship between the intensity of characterized Raman signals and the time of electrochemical reduction. c AES spectra of Cu LMM over various Cu₂O nanocube catalysts after the electrochemical reduction process. d Normalized Cu K-edge XANES spectra. Inset is the magnification of Cu K-edge spectra. e Fourier transforms of Cu K-edge EXAFS spectra with optimal fitting results for various Cu₂O nanocube catalysts after electrochemical reduction process. f Structural coherence changes in EXAFS coordination number of Cu–Cu bonds and Cu–O–Cu bonds. g Schematic diagram of structural transformation of Cu₂O with different sizes. Source data for a–f are provided as a Source Data file.

Full size image

Though in situ Raman spectroscopy is an efficient strategy to identify the bending mode for Cu–O bonds, it is difficult to accurately identify metallic Cu species during the electrochemical reduction process. Thus, the electronic structure and valence states of the Cu/Cu₂O structure is carefully evaluated by the AES spectrum. By quantifying the content of Cu⁺ and Cu⁰ in the Cu LMM spectra (Fig. 2c), the highest proportion of Cu⁰ was found in S-Cu/Cu₂O is 78.9%, followed by M-Cu/Cu₂O (53.1%) and L-Cu/Cu₂O (38.4%), which is consistent with the in situ Raman spectra. Further, the XANES and extended X-ray absorption fine structure (EXAFS) were used for investigating the coordination environment and local electronic structure of Cu₂O catalysts after electroreduction (Fig. 2d, e). The secondary adsorption band (8983 eV), corresponding to a 1s → 4p transition concurrent with ligand to metal charge transfer character, is used to evaluate the coordination structure of Cu⁰–Cu⁺ sites in XANES spectra¹⁸. After the electrochemical reduction process, the lower intensity of the secondary adsorption band was found on both Cu₂O samples, suggesting the appearance of unsaturated coordination structures. It was found that the S-Cu/Cu₂O shows the lowest adsorption intensity than M-Cu/Cu₂O and L-Cu/Cu₂O indicates a high degree of local structural disorder with a low oxidation state. Further, the fine coordination environment and geometric interface structure on the Cu⁰–Cu⁺ sites is observed in the EXAFS spectra. As shown in Fig. 2e, the characteristic coordination shell of the Cu–O bond (~1.5 Å) and Cu–O–Cu bond (~2.8 Å) were observed in Cu₂O precursors¹⁹. Interestingly, after reduction electrolysis, an extra coordination shell was found at ~2.5 Ų⁰, which contributed to the Cu⁰–Cu bond on the metallic Cu species, indicating the appearance of Cu⁰–Cu⁺ interfacial structure during the electroreduction process. To accurately identify the fine coordination structure, the EXAFS spectra were fitted according to the optimal structure parameters, the corresponding fitting curves and structural information are displayed in Supplementary Figs. 15 and 16 and Supplementary Table 1. According to the EPR results, the S-Cu/Cu₂O has the higher concentration of unsaturated coordination structure, confirmed by the coordination number in Fig. 2f. The unsaturated defect Cu–O species on the Cu₂O surface, which own higher surface energy, tended to be reduced to metallic Cu species during the electrochemical reduction process. Thus, it was found that the coordination number (CN) of the Cu–O–Cu bond is dramatically reduced to metallic Cu species on S-Cu/Cu₂O with the appearance of the higher content of the Cu–Cu bond (Fig. 2f), indicating that the formation of metallic-Cu-dominated Cu/Cu₂O interface structure after the electroreduction process. On the contrary, owing to low defect concentration on large-sized Cu₂O, we found that the Cu₂O is slightly reduced to metallic Cu species on L-Cu/Cu₂O sample to form Cu⁺-dominated Cu/Cu₂O interface structure. Therefore, by rational tuning the particle size of Cu₂O precursors, the Cu/Cu₂O interface structure with a flexible ratio of Cu⁰ and Cu⁺ could be effectively constructed.

Electrocatalytic performance for ammonia synthesis

Above, the electrocatalyst with Cu/Cu₂O interface structure has been constructed by modulating the size effect of Cu₂O catalysts. Then, the NO₃⁻RR performance of as-synthesized Cu/Cu₂O catalysts was studied in a standard three-electrode H-type electrolytic cell. The electrochemical behavior was identified by the linear sweep voltammetry (LSV) curve. In the 1 M KOH solution, L-Cu/Cu₂O exhibits higher hydrogen evolution reaction (HER) activity than M-Cu/Cu₂O and S-Cu/Cu₂O (Supplementary Fig. 17), which is attributed to the higher content of Cu⁺ species on the L-Cu/Cu₂O. The Cu⁺ species has a lower energy barrier for the water-dissociation process than the Cu⁰ species, resulting in rapid HER dynamics²¹. The NO₃^-RR performance is evaluated in the 1 M KOH with 0.1 M KNO₃ solution. As shown in Fig. 3a, the M-Cu/Cu₂O exhibits a higher current density than S-Cu/Cu₂O and L-Cu/Cu₂O, which is inconsistent with HER activity, indicating that the NO₃^-RR activity is not only influenced by the formation of adsorbed *H from water dissociation, but also strongly related to the adsorption of NO₃^- and NO₃⁻RR intermediates. The intrinsic NO₃⁻RR performance was evaluated by normalized by the electrochemically active surface area (ECSA), which investigated by non-faradaic double layer capacitance (C_dl) in Supplementary Fig. 18. It was found that M-Cu/Cu₂O still shows the best intrinsic activity, indicating that the M-Cu/Cu₂O has the optimal Cu⁰–Cu⁺ interface structure for catalyzing NO₃⁻RR process (Supplementary Fig. 19). To investigate the NO₃⁻RR dynamic process, the Tafel slope and electrochemical impedance spectroscopy (EIS) were conducted. The results of Tafel slopes shown that the M-Cu/Cu₂O has a faster electron transfer rate and rapid kinetics for NO₃⁻RR (Supplementary Fig. 20). The electrochemical impedance was performed and fitted by optimal equivalent circuit, and the resistance of each component was shown in Supplementary Table 2. M-Cu/Cu₂O showed the lowest charge transfer resistance (R_ct), indicating that the appropriate Cu⁰–Cu⁺ interface structure would accelerate the charge transfer process and promote the NO₃⁻RR kinetics (Supplementary Fig. 21).

a LSV curves of various Cu/Cu₂O catalysts in the 1 M KOH with 0.1 M KNO₃ solution with the scan rate of 5 mV s⁻¹. The potential value is non-iR corrected. Faradaic efficiency (b) and yield rate (c) of NH₃ over various Cu/Cu₂O catalysts at different applied potentials. Data are presented as mean ± s.d. (n = 3) with the error bars representing the s.d. d Comparisons of NH₃ yield rates and Faraday efficiency between the M-Cu/Cu₂O and typically reported NH₃-synthesis electrocatalysts. e Faradaic efficiency of NH₃ production and NO₃^- reduction rates in the 1 M KOH with various concentrations of KNO₃ solution over M-Cu/Cu₂O catalysts at −0.2 V vs. RHE. f The Faradaic efficiency and yield rate of NH₃ over M-Cu/Cu₂O catalysts during the stability measurement. Source data for a–f are provided as a Source Data file.

Full size image

The performance for the transformation of NO₃⁻ to NH₃ is evaluated by the electrochemical chronoamperometry measurement over S-Cu/Cu₂O, M-Cu/Cu₂O, and L-Cu/Cu₂O. The substrates, intermediates, and products were quantified by ultraviolet-visible (UV–vis) spectrophotometry, and the calibration curves are displayed in Supplementary Figs. 22–24. It was found that M-Cu/Cu₂O exhibits a maximum NH₃ Faradaic efficiency (FEs) of 95% at −0.2 V_RHE in Fig. 3b, which is much higher than that of S-Cu/Cu₂O (80%) and L-Cu/Cu₂O (86%). Moreover, an increasing yield rate of NH₃ and NO₃⁻ reduction rate was observed with the increase of applied negative potentials (Fig. 3c and Supplementary Fig. 25), while reached 0.29 mmol h⁻¹ mg^-1 and 0.25 mmol h⁻¹ mg^-1 towards NO₃⁻ reduction and NH₃ production, respectively, at the potential of −0.2V_RHE on M-Cu/Cu₂O, which is better than recently reported NO₃RR electrocatalysts (Fig. 3d and Supplementary Table 3). To enhance the reliability of size strategy, the M2-Cu/Cu₂O is synthesized with size (~402 nm) between M-Cu/Cu₂O and L-Cu/Cu₂O (Supplementary Fig. 26). Then, the electrocatalytic NO₃⁻RR performance towards NH₃ production is evaluated in the 1 M KOH with 0.1 M KNO₃ solution. As shown in Supplementary Fig. 27, the NH₃ Faradaic efficiency (FEs) of M2-Cu/Cu₂O (87.6%) is slightly lower than M-Cu/Cu₂O (95%) and higher then L-Cu/Cu₂O (86%), indicating that size-effect strategy is convincing.

To confirm the adaptability of NH₃ production capability, the NO₃RR performance was studied at different NO₃⁻ concentrations over M-Cu/Cu₂O catalysts (Fig. 3e). It was found that the FE of NH₃ would also keep ~80% in a low concentration of NO₃^- solution (5 mM), indicating that M-Cu/Cu₂O also has a potential application prospect in the electroreduction of wastewater to NH₃. Furthermore, it was found that the NH₃ FE and yield rate remained stable after 10 electrolytic cycles (Fig. 3f and Supplementary Fig. 28). Notably, we observed that the crystal structure and phase composition of M-Cu/Cu₂O catalysts after electrolysis is similar to the fresh M-Cu/Cu₂O catalysts, indicating excellent stability of M-Cu/Cu₂O (Supplementary Fig. 29).

To verify the source of NH₃, controlled measurements were conducted by electrolysis in 1 M KOH solution with the absence of NO₃⁻. It was found that the NH₃ produced is negligible, indicating that the acquirement of NH₃ originated from NO₃⁻RR (Supplementary Figs. 30 and 31). Moreover, isotope labeling experiments were carried out with K¹⁴NO₃ and K¹⁵NO₃ as nitrates (Supplementary Fig. 32), and the products were determined by ¹H NMR spectroscopy. The results show that the typical two peaks of ¹⁴NH₄⁺ and the typical three peaks of ¹⁵NH₄⁺ in ¹H NMR, which confirms that the synthesized NH₃ comes from NO₃⁻ in solution rather than environmental contamination.

Mechanistic studies

Firstly, the possible reaction pathway and multiple intermediates were preliminarily evaluated via in situ Attenuated Total Refraction Fourier Transform Infrared Spectrometer (ATR-FTIR). As displayed in Fig. 4a and Supplementary Fig. 33, under the applied potential of −0.2 V_RHE, the peak at 1240 cm^–1 and 1640 cm^–1 were assigned to the N−O antisymmetric stretching of NO₂^– and the adsorption of *NO intermediates, and the doublets of 1423 cm^–1 and 1457 cm^–1 is attributed to NH₄⁺ species^22,23, indicating that the pathway of *NO₃⁻ → *NO₂⁻ → *NO → *N → *NH → *NH₂ → *NH₃ is dominated on Cu/Cu₂O interfacial catalysts. Moreover, in situ differential electrochemical mass spectrometry (DEMS) is used to confirm the key intermediates during the NO₃RR process. As shown in Supplementary Fig. 34, the six possible NO₃RR intermediates could be observed on the DEMS signals at m/z 46 (*NO₂), 30 (*NO), 14 (*N), 15 (*NH), 16 (*NH₂), and 17 (*NH₃), respectively²⁴.

a In situ FTIR spectra of M-Cu/Cu₂O catalyst during the nitrate reduction at various electrolytic times. b The evaluation of nitrogen selectivity and conversion rate over various Cu/Cu₂O catalysts at −0.2 V vs. RHE. c, d ESR spectra of DMPO adducts over various Cu/Cu₂O nanocube catalysts in the absence and presence of nitrate, the signal is collected after 15 min electrolysis at -0.2 V vs. RHE. e The transformation of NO₃⁻ over various Cu/Cu₂O nanocube catalysts with the presence of 10 mM TBA solution. f LSV curves of M-Cu/Cu₂O catalyst in the absence and presence of SCN⁻ solution. The potential value is non-iR corrected. g Gibbs free energy diagrams of the conversion of nitrate to ammonia over various Cu/Cu₂O catalysts; h the kinetic energy barriers of potential limiting steps over the various Cu/Cu₂O catalysts with CI-NEB method. Source data for a–h are provided as a Source Data file.

Full size image

To identify the mechanism of NO₃⁻RR on various Cu⁰–Cu⁺ interfacial catalysts, the product distribution of NO₃⁻RR is evaluated at the applied potential of −0.2 V_RHE, it was found that the M-Cu/Cu₂O shows the highest NH₃ selectivity (93.6%) and NO₃⁻ conversion (86.2%) than S-Cu/Cu₂O and L-Cu/Cu₂O in Fig. 4b. Besides, the production of NO₂⁻, derived from two-electron NO₃⁻ reduction process, is founded in the solution, which is attributed to the weakly reduction ability of NO₂− to NH₃. Further, by revealing potential-dependent production selectivity, it was found that S-Cu/Cu₂O tended to generate NO₂⁻ at low potential, resulting in higher NO₂⁻ FEs than M-Cu/Cu₂O and L-Cu/Cu₂O at the same potential (Supplementary Fig. 35), suggesting that the NO₂⁻ electrochemical reduction step is sluggish on the metallic-Cu-dominated Cu/Cu₂O interface structure, which leads to the abundant NO₂⁻ accumulation during the NO₃⁻RR process. On the contrary, the rapid NO₂⁻ electrochemical hydrogenation is observed on the Cu⁺-dominated Cu/Cu₂O interfacial catalysts, indicating that the Cu₂O is the active species for the NO₂⁻ → NH₃ process.

It is known that two possible mechanisms of NO₃⁻ reduction reaction: direct electron transfer process and atomic H*-mediated indirect reduction process. For the atomic *H-mediated indirect reduction process, the water molecule would be firstly dissociated to *H and *OH on the surface of electrode, then a sequential electrochemical NO₃^_ hydrogenation process is conducted by utilizing the adsorbed *H species. Different from the *H-mediated pathway, the NO₃^- adsorbed on catalysts would be electrochemically reduced by directly accepting electrons from catalysts without the involvement of water dissociation via the indirect reduction process. The electron spin resonance (ESR) is used to confirm the pathway during the NO₃^->RR process. Without adding NO₃⁻ in catholyte, the ESR spectra display typical 9 signals with intensity ratios of 1:1:2:1:2:1:2:1:1 (Fig. 4c), which confirms the formation of DMPO-H in the solution. However, when NO₃⁻ was added to the solution, the typical 9 signals were decreased (Fig. 4d), indicating that the adsorbed H* is consumed by the NO₃⁻RR intermediates via atomic *H-mediated indirect reduction pathways over various Cu/Cu₂O catalysts. The sequence of DMPO-H signal intensity is S-Cu/Cu₂O < M-Cu/Cu₂O < L-Cu/Cu₂O, indicating Cu₂O species is tended to produce sufficient atomic *H adsorbed to participate in the hydrogenation of NO₃RR intermediates. The quenching experiments are conducted to identify the possible pathway on the various Cu-Cu₂O interfacial catalysts, and tertiary butyl alcohol (TBA) serves as a scavenger to capture atomic *H²⁵. As shown in Fig. 4e, a decrease of NO₃⁻ kinetics on various Cu/Cu₂O catalysts was observed when 10 mM TBA was added into the solution, indicating that the atomic *H-mediated indirect reduction pathway is dominated. Compared to S-Cu/Cu₂O, M-Cu/Cu₂O and L-Cu/Cu₂O catalysts showed a rapid NO₃⁻ RR kinetics suggesting that Cu₂O species would provide sufficient atomic hydrogen for NO₃RR process, which is consistent with ESR results.

Besides, it is expected to identify the roles of Cu sites on the Cu⁰–Cu⁺ interfacial catalysts, a poisoning experiment was implemented using KSCN species¹⁹. Since the SCN⁻ is a strongly Lewis alkalinity, which would be selectively adsorbed on the metallic Cu sites and competed with NO₃⁻, when SCN^- ions were added to the solution, the current density of NO₃RR decreased sharply (Fig. 4f), suggesting that the metallic Cu sites, in Cu⁰–Cu⁺ interface, play a vital role during the NO₃⁻RR process.

Further, the DFT calculations were performed to elucidate the mechanism over the Cu⁰–Cu⁺ interface structure. Owing to strong electron correlation, the Hubbard-U parameter is used to prevent the delocalization of Cu d-orbitals. To accurately simulate the Cu⁰–Cu⁺ interface structure, the model of S-Cu/Cu₂O, M-Cu/Cu₂O, and L-Cu/Cu₂O catalysts is constructed and optimized based on the atomic ratio of Cu⁰/Cu⁺ species according to the AES and EXAFS results (detailed in the Supplementary information and Supplementary Data 1). Based on the ATR-FTIR results, the pathway of *NO₃⁻ → *NO₂⁻ → *NO → *N → *NH → *NH₂ → *NH₃ is proposed to investigate the reaction mechanism, and the optimal adsorption configuration of the reaction intermediates on various Cu/Cu₂O catalysts is displayed in Supplementary Figs. 36–40.

The adsorption behavior of NO₃⁻ reactant is investigated on the active Cu⁰–Cu⁺ sites, it was found that NO₃⁻ tends to adsorb on the metallic Cu⁰ sites rather than Cu⁺ sites or Cu⁰–Cu⁺ interface sites over various Cu/Cu₂O catalysts (Supplementary Fig. 36). Owing to the oxyphility of metallic Cu species, its tends to produce strong binding interactions with oxygenated species, resulting to the strong adsorption of NO₃⁻ reactant (Supplementary Fig. 37)^11,26. It was found that the NO₃^‒ adsorption energy of L-Cu/Cu₂O (−0.35 eV) is lower than M-Cu/Cu₂O (−0.70 eV) and S-Cu/Cu₂O (−0.79 eV), suggesting that introduction of Cu₂O benefits to weaken the NO₃⁻ adsorption on the surface of Cu sites. The open circuit potential (OCP) measurement is also used to identify the adsorption of NO₃⁻ on the various Cu/Cu₂O electrocatalysts^27,28. After the injection of KNO₃, the S-Cu/Cu₂O (0.14 V) showed a significant decrease in OCP than the M-Cu/Cu₂O (0.11 V) and L-Cu/Cu₂O (0.06 V), indicating that Cu⁺ helps to weaken NO₃⁻ adsorption (Supplementary Fig. 41), which is consistent with the results of DFT calculation.

Then, according to the proposed reaction pathway, the free energy of NO₃RR was investigated to identify the mechanism of synergistic Cu⁰–Cu⁺ motifs on various Cu/Cu₂O electrocatalysts. As shown in Fig. 4g, it was found that the hydrogenation step of *NH₂ → *NH₃ is regarded as the potential limiting step (PLS) on the M-Cu/Cu₂O and S-Cu/Cu₂O, while the hydrogenation step of *NH → *NH₂ is considered as PLS on the L-Cu/Cu₂O. It is interesting that the free energy change of PLS for M-Cu/Cu₂O (0.11 eV) is slightly lower than S-Cu/Cu₂O (0.22 eV) and L-Cu/Cu₂O (0.20 eV), indicating that the NO₃RR process is thermodynamically favorable on the M-Cu/Cu₂O electrocatalysts. Moreover, owing to the high energy barrier (0.96 eV), we have excluded possible pathway of *NO coupling for the formation of nitrogen gas (N₂) on the M-Cu/Cu₂O materials (Supplementary Fig. 42).

It was noticed that the water-dissociation reaction (H₂O → *H + *OH) process is crucial to provide atomic hydrogen species for the hydrogenation of NO₃RR intermediates^29,30, while the H₂O molecules may be competitive adsorbed on the surface of electrode. Therefore, we considered the adsorption and activation behavior of H₂O molecules, it was found that the dissociation of the H₂O molecule needs to overcome a high kinetic energy barrier of 1.21 eV on the M-Cu/Cu₂O (Supplementary Fig. 43). Such a high energy barrier clearly reflects the sluggish water-dissociation kinetics and poor HER activity (Supplementary Fig. 17). Besides, the OH^- species may also be competitive adsorbed on the Cu sites, we found that the direct reduction of adsorbed *OH species also requires endothermic 0.24 eV, which is higher than NO₃RR (0.11 eV), demonstrating that M-Cu/Cu₂O exhibits the high selectivity for NO₃RR.

Further, we also explore the kinetic energy barrier for the PLS of the different Cu/Cu₂O materials. As shown in Fig. 4h, the kinetic energy barrier of the PLS is 0.51 eV on the M-Cu/Cu₂O, which is lower than S-Cu/Cu₂O (0.64 eV), and L-Cu/Cu₂O (1.29 eV), indicating that the M-Cu/Cu₂O, with the optimal interface structure, is favorable in kinetics for NO₃ electrochemical reduction process.

Paired-electrolysis NO₃ ⁻RR//HMFOR system

Traditional electrocatalytic nitrate reduction systems involve a cathodic nitrate reduction reaction and anodic oxygen production reaction (OER). Though the efficient transformation of NO₃⁻ to NH₃ has been achieved by modulating the size effect of the Cu₂O electrocatalyst, the sluggish anodic OER process would dramatically enhance the energy consumption for NH₃ production. Thus, it is expected to explore promising anodic reactions for the replacement of the OER process to reduce energy consumption. The electrooxidation of 5-hydroxymethylfurfural (HMF) to 2,5-furandicarboxylic acid (FDCA) has been attracted to replace the traditional OER process to reduce cell voltage and achieve the production of high-value chemicals. Since the Ni-based catalyst has excellent performance of HMF electrochemical oxidation reaction (HMFOR), the Ni₃Se₄, as a representative Ni-based material^31,32, was selected as anode electrocatalyst to assemble a paired electrolysis NO₃^-RR//HMFOR system (Fig. 5a).

a, b Scheme and digital photo of NO₃⁻RR//HMFOR paired-electrolysis system. c The LSV curves of overall water splitting systems and NO₃^-RR//HMFOR Paired-electrolysis system. Potential-dependent Faradaic efficiency and yield rate of NH₃ (d) and FDCA (e) in AEM electrolyzer. f Stability measurement of NO₃⁻RR//HMFOR Paired-electrolysis system at the current density of 2 A cm⁻². g Techno-economic analysis of NO₃⁻RR//HMFOR paired-electrolysis system. Source data for c–g are provided as a Source Data file.

Full size image

To further verify the NO₃⁻RR coupled HMFOR strategy, a simple scale-up test was carried out in an AEM electrolytic cell with a working area of 100 cm², and the M-Cu/Cu₂O and Ni₃Se₄ served as cathode and anode catalyst, respectively. The crystal structure and morphology is characterized by XRD pattern, SEM, and TEM images in Supplementary Figs. 44 and 45, and the HMFOR performance is evaluated in Supplementary Fig. 46. Digital photos and structural diagrams of AEM are shown in Fig. 5b. When coupling anodic HMFOR process and cathodic NO₃RR process, the cell voltage is significantly reduced (Fig.5c), indicating that the high energy conversion efficiency of NO₃RR//HMFOR system. The product measurement was performed at the different applied potentials as shown in Fig. 5d, e. The highly NH₄⁺ yield rates (5.20 mmol h⁻¹ cm⁻²) and FE (75.6 %) with excellent FDCA yield rates (0.47 mmol h⁻¹ cm⁻²) and FE (71.2 %) were achieved at the potential of 2.1 V. Moreover, the NH₄⁺ and FDCA could be continuously and stably produced at the current density of 2 A cm^-2, indicating the excellent stability of this electrolysis system (Fig. 5f). Further, we assembled a product extraction system to extract the cathode product (NH₄⁺) and the anode product FDCA from the electrolyte. The purified product was confirmed by the XRD pattern in Supplementary Fig. 47. Based on the above results, it can be determined that the NO₃RR//HMFOR system would significantly improve energy efficiency, ensuring the economy and practicability, TEA results show that the economic benefits generated by the system are greater than the cost input (Fig. 5g).

In summary, we developed a promising strategy to achieve the controllable construction of Cu⁰/Cu⁺ interfacial structure by modulating the size effect of Cu₂O nanocube electrocatalysts. Combining the in situ electrochemical Raman and X-ray absorption near edge structure (XANES) characterization, the design principle of Cu⁰–Cu⁺ pairs has been revealed by correlating the macroscopic particle size with the microscopic localized coordinated structure properties. Based on the controllable construction of the Cu⁰–Cu⁺ interfacial structure, we elucidated the synergic roles of Cu⁺ and Cu⁰ sites during the NO₃⁻RR process by combining electrochemical measurement and DFT calculations. The results reveal that Cu⁰ is the main active site in the Cu⁰–Cu⁺ motifs, Cu⁺ helps to weaken NO₃⁻ adsorption, provides atomic *H species to accelerate potential limiting step of the hydrogenation of NH₂ to NH₃. Based on the design of Cu⁺–Cu⁰ interfacial electrocatalyst, we developed an efficient NO₃⁻RR//HMFOR paired electrolysis system to simultaneously achieve the efficient production of NH₃ and FDCA at an industrially relevant current density, while maintaining high Faradaic efficiencies (75.6% for NH₃, 71.2% for FDCA), yield rates (5.20 mmol h⁻¹ cm⁻² for NH₃, 0.47 mmol h⁻¹ cm⁻² for FDCA), and long-term operational stability (20 h) in AEM electrolyzers with an area of 100 cm². This work provides a strategy to design energy-effective and eco-friendly electrocatalysts for large-scale industrial electrolytic synthesis of high-value-added products.

Chemical reagents

KNO₃ (Kermel, GR, 99%), KNO₂ (Kermel, GR, 99%) NH₄Cl (Aladdin, AR, 99%), 5-hydroxymethylfurfural (98%) were purchased from Energy Chemical. 5-hydroxymethyl-2-furancarboxylic acid (98%), 5-formyl-furancarboxylic acid (99%), 2,5-diformyfuran (98%), and 2,5-furandicarboxylic acid (98%) were obtained from Aladdin. Cu(CH₃COO)₂·H₂O (Aladdin, AR, 99%), L-Ascorbic acid (Macklin, AR, 99%), and NaOH (Macklin, AR, 99%) were used without further purification. Deionized water (UPR series super pure water purification system, resistivity >18.2 MΩ cm) was used to prepare all solutions.

Materials preparation

Synthesis of Cu₂O nanocubes

In a typical synthesis, a mixture of 10.0 mL of 0.009 M ascorbic acid solution and 16.0 mL of 0.113 M NaOH solution was added into 16.0 mL of 0.005 M copper acetate solution. The mixed solution was stirred vigorously for 30 min, enabling the solution became yellow and turbid, indicating the formation of Cu₂O nanocubes. Finally, the precipitates were collected and washed through centrifugation. All the experiments were carried out at room temperature (298 K).

Synthesis of Cu/Cu₂O nanocubes

The Cu/Cu₂O nanocubes were synthesized from Cu₂O nanocubes by a constant potentiostatic reduction process at the potential of −0.6 V vs RHE for 300 s in 1 M KOH and 0.1 M NO₃⁻ solution. This process was implemented in a three-electrode system, where carbon paper, loaded on the Cu₂O nanocubes, graphite rod, and Hg/HgO electrode was used as the working electrode, counter electrode, and reference electrode. After the electrochemical reduction process, the obtained Cu/Cu₂O nanocubes were washed with absolute ethanol and blown dry by nitrogen.

Materials characterization

XRD patterns were collected via an X-ray diffractometer (Shimadzu XRD-7000) using monochromatic Cu K_α radiation (λ = 1.5406 Å) to investigate the phase structure transformation. Scanning electron microscopy (SEM; JSM-7610F) and transmission electron microscopy (TEM; JEOL F200 equipped with energy dispersive spectrometer (EDS)) were employed to observe the microstructure and elemental distribution of the catalysts. The chemical composition and surface valence of the catalysts were analyzed through X-ray photoelectron spectroscopy with an Al-K_α X-ray source (E = 1486.6 eV, XPS, Thermo Scientific Escalab), and the binding energy was corrected with a C 1s spectral of 284.8 eV. Raman spectroscopy was collected on a laser Raman spectrometer (Horiba LabSpec6, 532 nm laser). The ultraviolet-visible (UV-Vis) absorbance spectra were measured on a spectrophotometer (Beijing Purkinje General T6 new century).

Electrochemical measurement

Typically, 1 mg catalyst was dispersed in a mixed solvent containing 500 μL of ethanol, deionized water (450 μL), and 50 μL of 5% Nafion solution. After ultrasonic treatment for 20 min, the catalyst slurry was dropped uniformly onto a piece of carbon paper (1 cm², 1 mg cm⁻²). After drying at room temperature (298 K), the catalyst-covered carbon paper was utilized as the working electrode in NO₃⁻RR measurements. The NO₃⁻RR measurements were performed with a three-electrode system in an H-type electrochemical cell separated by Nafion 117 membrane using a CHI electrochemical workstation (CHI 760E) at room temperature (298 K). The Nafion 117 membrane was pre-treated by immersing it in water, 5% H₂O₂ solution, and 0.5 M H₂SO₄ at 80 °C for 1 h. The Cu/Cu₂O was used as the working electrode, a carbon rod as the counter electrode, and Hg/HgO as the reference electrode. The NO₃⁻RR measurement was conducted in 1.0 M KOH with 0.1 M KNO₃ solution (pH was 13.8 ± 0.2, which determined by pH meter). All potentials were recorded against the reversible hydrogen electrode (RHE, E_RHE = 0.0591 × pH + 0.098 V + E_Hg/HgO) without special explanation. The linear sweep voltammetry (LSV) test was performed at a rate of 5 mV s⁻¹, and the potential was not iR corrected. Constant potential electrolysis was conducted at each potential for 3 h. Electrochemical surface area (ECSA) was calculated from scan rate dependence of CV with scan rates of 10–90 mV s⁻¹. The electrochemical impedance spectroscopy (EIS) measurements were carried out with frequencies from 0.01 Hz to 100 kHz at an amplitude of 10 mV. For flow electrochemical measurement, the customized two-electrode flow cell (as shown in Fig. 5b) is used for the NO₃^-RR//HMFOR paired-electrolysis system, the cathode and anode were separated by a anion exchange membrane (Fumasep FAA-3-20), which is directly used for electrochemical measurement without pre-treatment, and the voltages was not iR corrected. The Ni₃Se₄, dropped on the carbon felt (10 × 10 cm², 2 mg cm^-2), and M-Cu₂O, dropped on the carbon felt (10 × 10 cm², 2 mg cm^-2), were used as anode and cathode, respectively. Catalyst inks (Ni₃Se₄ or M-Cu₂O, 200 mg) were prepared by dispersing catalyst powder in a mixture of isopropanol (10 mL), pure water (9 mL), and Nafion (1 mL). Subsequently, the catalyst inks were dropped on the carbon felt and dried at room temperature (298 K). The anodic electrolyte of 0.1 M HMF in 2.0 M KOH was prepared before pumping into the anode chamber with a flow rate of 20.0 mL min⁻¹, and the cathodic electrolyte of 0.1 M KNO₃ in 2.0 M KOH was prepared before pumping into the cathode chamber with a flow rate of 20.0 mL min⁻¹. The electrolyte pumping out from the anode and cathode were collected, and the product was analyzed using HPLC and UV–Vis, respectively.

Product quantification

The UV–Vis spectrophotometer was used to quantified the concentration of nitrate, nitrite, and ammonia by same colorimetric methods. In nitrate quantification, the electrolyte is neutralized and diluted to suitable concentrations. Then, 2 mL of 0.8 w% sulfamic acid solution, 2 mL of Nessler’s reagent followed by 1 M HCl were added to the above solution. After standing at room temperature (298 K) for 20 min, record the absorbance at a wavelength of 420 nm with a UV–Vis spectrophotometer. In nitrite quantification, a mixture of p-aminobenzenesulfonamide (4 g), N-(1-Naphthyl) ethylenediamine dihydrochloride (0.2 g), ultrapure water (50 mL), and phosphoric acid (10 mL, 1.70 g mL⁻¹) was used as a color reagent. A certain amount of electrolyte was extracted from the cathodic chamber and diluted to suitable concentrations. Then, 0.1 mL of color reagent was added to the above solution to further react for 20 min. The absorbance was recorded at a wavelength of 540 nm. In ammonia quantification, Nessler’s reagent was used as the color reagent to determine the quantification of ammonia. Firstly, a certain amount of electrolyte was extracted from the cathodic chamber and diluted to suitable concentrations. Then, 0.1 mL of potassium sodium tartrate solution (ρ = 500 g L⁻¹) and 0.1 mL Nessler’s reagent were subsequently added into the above solution to further react for 20 min. The absorbance was recorded at a wavelength of 420 nm.

Calculation of the yield, selectivity, and Faradaic efficiency

The conversion of NO₃⁻ was calculated using:

The selectivity of NH₃ and NO₂⁻ calculated using:

The yield of NH₃ was calculated using:

The Faradaic efficiency was calculated using:

Where \({{{{\rm{C}}}}}_{{{{\rm{NH}}}}_{3}}\) and \({{{\rm{C}}}}_{{{{{\rm{NO}}}}}_{{2}}^{-}}\) are the concentration of \({{{{\rm{NH}}}}}_{3}\) and \({{{{\rm{NO}}}}}_{{2}^{-}}\), \(\Delta {{{\rm{C}}}}_{{{{{\rm{NO}}}}}_{{3}}^{-}}\) is the concentration difference of \({{{{\rm{NO}}}}}_{{3}^{-}}\) before and after electrolysis, C₀ is the initial concentration of \({{{{\rm{NO}}}}}_{{3}^{-}}\), V is the electrolyte volume, t is the electrolysis time, m is the mass of catalyst, F is the Faradaic constant (96,485 C mol⁻¹), Q is the total charge passing the electrode.

Theoretical simulation

All density functional theory (DFT) calculations were performed using the Vienna ab initio simulation package (VASP)³³. The ion-electron interaction was described with the projector-augmented plane-wave (PAW) method³⁴. The strongly correlated electronic states of Cu-3d orbitals were corrected via a DFT + U method. The effective Hubbard-U parameter (U_eff) for the Cu atom was set to be 5.0 eV³⁵. The kinetic energy cutoff for the plane-wave expansion was set to 400 eV, and the Brillouin zone was sampled with 3 × 3 × 1 for geometry optimization. All the structural models were fully relaxed to the ground state with the convergence of energy and forces setting to 5 × 10⁻⁵ eV and 0.05 eV Å⁻¹, respectively. DFT-D3 (D stands for dispersion) with the empirical correction in Grimme’s scheme was employed to describe van der Waals (vdW) interactions³⁶. Solvation effect was considered with the Poisson-Boltzmann implicit solvation model, and the dielectric constant of water was set to 80³⁷.

The Cu₂O and Cu crystallographic information files were downloaded from the Materials Project crystal material library. For the Cu₂O catalyst, the typical (111) slab with a (3 × 3 × 3) supercell was constructed; the top two layers and NO₃RR intermediates were fully relaxed, while the bottom layer was fixed. The bulk Cu was established as nanoclusters to represent the metal Cu species formed during the NO₃RR progress. Especially, considering the different structure the Cu⁰–Cu⁺ sites, we combine the Cu₂O (111) planes and Cu clusters to build the three types of Cu/Cu₂O catalysts. Using the Cu cluster of 15, 27, and 43 atoms, to represent L-Cu/Cu₂O, M-Cu/Cu₂O, and S-Cu/Cu₂O according to according to the Auger electron spectroscopy results. The length of the vacuum layer in the Z direction was set to 15 Å to avoid the interactions between periodic images. The Gibbs free energy of each species was calculated by the computational hydrogen electrode (CHE) model³⁸. Here, the chemical reaction considered can be summarized with the reaction equations below.

Where * represents the active site. Then, the reaction free energy can be obtained with the equation below:

Where ΔE is the total energy difference before and after intermediate adsorbed, ΔE_ZPE and ΔS are the differences of ZPE zero-point energy and entropy, respectively, which can be obtained from the vibrational frequency calculations. T is room temperature (298 K). Besides, the climbing-image nudged elastic band (CI-NEB) method was used to simulate the energy barrier of the NO₃RR process³⁹.