Atomically dispersed Cu coordinated Rh metallene arrays for simultaneously electrochemical aniline synthesis and biomass upgrading

Abstract

Organic electrocatalytic conversion is an essential pathway for the green conversion of low-cost organic compounds to high-value chemicals, which urgently demands the development of efficient electrocatalysts. Here, we report a Cu single-atom dispersed Rh metallene arrays on Cu foam for cathodic nitrobenzene electroreduction reaction and anodic methanol oxidation reaction. In the coupled electrocatalytic system, the Cu_single-atom-Rh metallene arrays on Cu foam requires only the low voltages of 1.18 V to reach current densities of 100 mA cm⁻² for generating aniline and formate, with up to ~100% of nitrobenzene conversion/ aniline selectivity and over ~90% of formate Faraday efficiency, achieving synthesis of high-value chemicals. Density functional theory calculations reveal the electron effect between Cu single-atom and Rh host and catalytic reaction mechanism. The synergistic catalytic effect and H*-spillover effect can improve catalytic reaction process and reduce energy barrier for reaction process, thus enhancing electrocatalytic reaction activity and target product selectivity.

Introduction

The utilization of renewable energy is a crucial pathway for solving the increasing energy crisis and promoting the green low-carbon transformation of energy^1,2. The organic electrocatalytic conversion, driven by electricity generated from renewable energy sources at ambient temperature and pressure^3,4, is a green synthesis route for achieving controlled conversion of low-cost organic compounds into high-value chemicals and holds considerable research significance and application potential in the chemical industry and organic synthesis^5,6,7. Among the numerous organic electrocatalytic conversion reactions, the cathodic nitrobenzene (Ph-NO₂) electroreduction for aniline (Ph-NH₂) synthesis is regarded as a more low-carbon and environment-friendly green Ph-NH₂ synthesis process using H₂O and electrons as the hydrogen source and reductant, compared with high pollution and harsh conditions of traditional chemical synthesis method^8,9,10,11. Notably, the anodic reaction for the conventional cathodic Ph-NO₂ electroreduction reaction (Ph-NO₂ ERR) is a sluggish kinetic oxygen evolution reaction (OER)^12,13. In contrast, the anodic biomass electrooxidation reaction not only can achieve the upgrading of cheap biomass but also offers the advantages of low energy consumption and high electrolysis efficiency^14,15,16. Currently, the cheap and widespread methanol (CH₃OH, 350 USD/t) is considered an appropriate precursor for the high-value formate (HCOO⁻, 1300 USD/t) synthesis via anodic electrooxidation reaction^17,18, and methanol electrooxidation reaction (MOR) coupled with various electrocatalytic reactions has attracted widespread research interest worldwide^19,20. In view of this, the construction of Ph-NO₂ ERR-MOR coupled electrolytic system is a promising strategy for achieving high-value chemical synthesis from sustainable organic electrocatalytic conversion. Nevertheless, both Ph-NO₂-to-Ph-NH₂ and methanol-to-formate are multi-step reaction processes and are limited by numerous side reactions^21,22. To address these bottlenecks, it is fundamental to explore efficient bifunctional electrocatalysts for activating reaction pathways of the directed synthesis towards Ph-NH₂ and formate.

Recently, the single-atom alloys (SAAs) catalysts, consisting of exogenous isolated metal atoms dispersed on the surface of the metal host^23,24, as a promising material with the advantages of high active atom utilization of single-atom catalysts (SACs) and alloy synergistic effect^25,26, have been in the field of various electrocatalytic energy conversions^27,28,29,30. For example, Duan and co-workers designed a Ru₁Cu SAAs with isolated Ru atoms on Cu nanowires for the electrocatalytic conversion of 5-hydroxymethylfurfural (HMF) to 2,5-dihydroxymethylfuran (DHMF), where the introduction of Ru single-atom sites facilitates the dissociation of H₂O to produce H* species for the HMF hydrogenation process³¹. Thus, the electron effect and metal-support interaction of SAAs can promote effective dissociation of reactants and optimize the adsorption/desorption of key intermediates to achieve an optimum balance between reactants and intermediates, leading to high activity and selectivity^32,33. Furthermore, the electrocatalytic activity of SAAs catalysts can be further improved by the precise regulation for morphology and structure^34,35. Metallene, a group of graphene-like two-dimensional (2D) nanomaterials with a thickness less than 5 nm^36,37. Due to their flexible microstructural tunability, highly exposed metal active sites as well as simple and well‐defined structure model^38,39, metallene can serve as a desirable carrier for immobilizing single-atom, clusters, or nanoparticles, which has attracted great research interest^40,41. For example, Xiaoqiang Cui et al. report a dispersed MoO_x on Rh metallene (MoO_x-Rh metallene) for boosting alkaline HER⁴². Chu’s group prepare a single-atom Bi alloyed Pd metallene (Bi₁Pd metallene) that shows excellent NO₃⁻ electroreduction reaction activity and near 100% Faradaic efficiency of NH₃⁴³. Therefore, the design and development of SAA metallene are extremely promising for improving electrocatalytic activity and catalytic product selectivity.

In this work, we report a synthesis of Cu single-atom dispersed Rh metallene arrays on Cu foam (Cu_SA-Rh MAs/CF) by a facile and rapid one-step solvothermal approach. As a bifunctional electrocatalyst, the Cu_SA-Rh MAs/CF displays superior electrocatalytic activity for Ph-NO₂ ERR and MOR. For the constructed Ph-NO₂ ERR-MOR coupled electrocatalytic system, the low voltage of only 1.05/1.18 V achieves current densities of 50/100 mA cm⁻² for efficient conversion of Ph-NO₂-to-Ph-NH₂ and methanol-to-formate, with Ph-NO₂ conversion and Ph-NH₂ selectivity up to ~100% and HCOO⁻ FE reaching over ~90% on Cu_SA-Rh MAs/CF, which not only enables the simultaneous cathodic and anodic organic electrocatalytic conversion for the synthesis of high-value chemicals but also maximizes the energy efficiency. Moreover, density functional theory (DFT) calculations further reveal the synergistic catalysis effect and H*-spillover effect induced by the local electron change between the isolated Cu single-atom and the Rh host, which promotes the rapid conversion of the reactants to key intermediates and rapid desorption of the target products thus enhancing electrocatalytic reaction activity and targeted product selectivity.

Results

Synthesis and characterization of Cu_SA-Rh MAs/CF

The Cu_SA-Rh MAs/CF was synthesized using a straightforward one-step solvothermal method (Supplementary Fig. 1), which contains rhodium (II) acetate dimer (C₈H₁₂O₈Rh₂), N,N-dimethylformamide (DMF), potassium hydroxide (KOH), ethylene glycol (EG), diethylenetriamine (DETA) and CF as reactants. In this synthesis process, the formed DMA by the synergistic interaction of DMF and KOH can induce the 2D nanosheet structure growth by facet control as well as diamine ligand DETA can chelate with metal ions to reduce the metal reduction rate thus promoting the formation of 2D nanosheet structure⁴⁴. Supplementary Fig. 2 shows that the surface color of formed Cu_SA-Rh MAs/CF is black, which is significantly different from that of CF. Figure 1a displays uniform ultrathin nanosheets grown on the CF surface to form metallene arrays, and the nanosheets are tightly interconnected to each other to form a security wall-like structure, which contributes to providing enough active sites as well as facilitating rapid charge transfer and mass transfer during the electrocatalysis process⁴⁵. Transmission electron microscopy (TEM) images of Cu_SA-Rh MAs further display the ultrathin nanosheet-like metallene structure with lateral dimensions of around several hundred nanometers (Fig. 1b). The thickness of a single nanosheet in Cu_SA-Rh MAs was measured to be ~1.51 nm corresponding to 7~8 atomic layers (Fig. 1c and Supplementary Fig. 3), further proving the ultrathin nature for Cu_SA-Rh MAs. The ultrathin 2D nanosheet structure with certain curvature can provide highly accessible surface atoms, highly exposed active sites, and rich defect structures, which is conducive to promoting the electrocatalytic process³⁷. Figure 1d presents the aberration-corrected high-angle annular dark field scanning transmission electron microscopy (AC-HAADF-STEM) and corresponding elemental mapping images of Cu_SA-Rh MAs, revealing the homogeneous distribution of Cu atoms in Cu_SA-Rh MAs. The Rh/Cu atomic ratio was further determined to be approximately 93.6/6.4 via the TEM energy dispersive X-ray spectroscopy (TEM-EDS, Supplementary Fig. 4). The mass ratio (Rh/Cu = 95.1/4.9) and atomic ratio (Rh/Cu = 92.6/7.4) of Rh/Cu in Cu_SA-Rh MAs were further analyzed by inductively coupled plasma optical emission spectroscopy (ICP-OES) (Supplementary Fig. 5), which is close to the results obtained from TEM-EDS. Furthermore, as revealed by the electron energy loss spectrum (EELS) of Cu_SA-Rh MAs (Supplementary Fig. 6), the energy loss peak around 498.8 eV can be assigned to the Rh M electron transition (Supplementary Fig. 6a)^46,47, and the energy loss peak around 933.1 eV in Supplementary Fig. 6b can be assigned to the Cu L electron transition⁴⁸. The EELS data of Cu_SA-Rh MAs further reveal the existence of Rh and Cu elements. The AC-HAADF-STEM images of Cu_SA-Rh MAs were collected and analyzed for further investigating its structure at the atomic scale. As shown in Fig. 1e–f, the obvious vacancy defects are observed on the basal surface of Cu_SA-Rh MAs and the presence of atomic vacancies is also verified by the corresponding integrated pixel intensity profile in the red region of Fig. 1e. Moreover, the lattice spacing (d = 0.220 nm) of Cu_SA-Rh MAs can be indexed to the typical face-centered cube (fcc) Rh (111) facet and the appearance of amorphous sites in Cu_SA-Rh MAs reveals the presence of crystalline and amorphous phases (Fig. 1g). In the X-ray diffraction (XRD) pattern (Fig. 1h), the characteristic peaks of Cu_SA-Rh MAs can be assigned to a typical fcc metallic Rh phase (No. 05–0685). Notably, the characteristic peaks of Cu_SA-Rh MAs exhibit a negative shift compared with the Rh JCPDS card, originating from the curved 2D geometrical structure and the introduction of Cu atoms³⁸. Moreover, the poor crystallinity is revealed by the weak and broad characteristic peaks of Cu_SA-Rh MAs in consistent with the selected area electron diffraction (SAED, inset in Fig. 1h) result, which further indicates the presence of the crystalline and amorphous phases in Cu_SA-Rh MAs. In detail, the vacancy defects and amorphous sites as highly active low-coordination sites can break the inherent crystal arrangement to cause atomic unsaturated bonding and readjust the local electron structure for optimizing the surface electron structure of the electrocatalyst, which is beneficial for improving the electrocatalytic activity³⁷. The Cu single-atom was analyzed by AC-HAADF-STEM image and 3D topographic atom images. Figure 2a shows that some individual dark dots (Cu atoms) can be observed on the surface of Cu_SA-Rh MAs owing to the lower atomic number of Cu (29) compared with Rh (45). The low-intensity dots also indicate the dispersion situation of the isolated Cu atoms on the surface of Cu_SA-Rh MAs (Fig. 2b, c). The corresponding integrated pixel intensity profile also illustrates the isolated low-intensity Cu atoms dispersed surrounding the high-intensity Rh atoms on the crystal surface (Fig. 2d), further proving the presence of isolated Cu single-atom on the Cu_SA-Rh MAs. The strong metallic interaction between the isolated Cu single-atom and Rh metallene support can modulate the electronic structure of active sites and form the activated Rh-coordinated Cu single-atom active sites, which is beneficial for optimizing the adsorption and activation between active sites and reactants during electrocatalytic reactions^25,43.

Fig. 1: Morphological and structure characterization of Cu_SA-Rh MAs/CF.

a SEM image of Cu_SA-Rh MAs/CF. b TEM image, c 3D view for the corresponding AFM image, d AC-HAADF-STEM image, and the corresponding elemental mapping images of Cu_SA-Rh MAs. e AC-HAADF-STEM image of Cu_SA-Rh MAs. f The integrated pixel intensity profile for the selected red region in e. g AC-HAADF-STEM image of Cu_SA-Rh MAs. h XRD pattern of Cu_SA-Rh MAs and the inset in h displays the corresponding SAED pattern of Cu_SA-Rh MAs.

Fig. 2: Structure characterization of Cu_SA-Rh MAs/CF.

a AC-HAADF-STEM image of Cu_SA-Rh MAs and the inset in a displays the magnified image of the selected white region in a. b, c 3D topographic atom images of Cu_SA-Rh MAs. d The integrated pixel intensity profile for the selected region in b. e Rh 3d XPS and f Cu 2p XPS spectra of Cu_SA-Rh MAs. g Rh K-edge XANES spectra, h Fourier transformed EXAFS spectra and i EXAFS wavelet transform diagrams of Rh foil, Rh₂O₃, and Cu_SA-Rh MAs. j Cu K-edge XANES spectra, k Fourier transformed EXAFS spectra, and l EXAFS wavelet transform diagrams of Cu foil, CuO, and Cu_SA-Rh MAs.

The elemental composition, chemical valence, and coordination environment for Cu_SA-Rh MAs were investigated through the X-ray photoelectron spectroscopy (XPS), X-ray absorption near-edge structure (XANES), and extended X-ray absorption fine structure (EXAFS) tests. As shown in Fig. 2e, the Rh 3d spectrum of Cu_SA-Rh MAs reveals that four obvious characteristic peaks (307.42 eV, 308.75 eV, 312.19 eV, 313.98 eV) are ascribed to Rh⁰ 3d_5/2, Rh³⁺ 3d_5/2, Rh⁰ 3d_3/2 and Rh³⁺ 3d_3/2, respectively^46,49, indicating the presence of Rh in Cu_SA-Rh MAs primarily as the metallic state. In the Cu 2p spectrum of Cu_SA-Rh MAs (Fig. 2f), two obvious characteristic peaks are at 932.50 eV and 952.37 eV ascribed to Cu⁰⁵⁰, and other characteristic peaks at 934.76 eV, 954. 49 eV and 942.21 eV can be ascribed to Cu²⁺ and satellite peak⁵¹. As observed from Rh K-edge XANES, EXAFS, and wavelet transform (WT) spectra (Fig. 2g–i), the similar features of Cu_SA-Rh MAs with Rh foil reveal that the surface valence and coordination structure of Rh is not significantly changed by the introduction of Cu single-atom. Notably, based on the WT spectra analysis for Cu_SA-Rh MAs and Rh foil (Fig. 2i), the Rh-Rh/Cu-Rh intensity maximum of Cu_SA-Rh MAs exhibits a negative shift of ~0.33 Å⁻¹ compared with Rh-Rh intensity maximum of Rh foil, which is induced by the coordination of the Cu-Rh bond. Supplementary Figs. 7–9 shows the Rh K-edge experimental and fitting Fourier-transformed EXAFS spectra for Cu_SA-Rh MAs, Rh foil, and Rh₂O₃, indicating a good fitting result. The Cu K-edge XANES spectra indicate that the absorption edge of Cu_SA-Rh MAs is positioned between Cu foil and CuO (Fig. 2j), indicating an increase in the valence state of Rh for Cu_SA-Rh MAs owing to the electron transfer from Cu to Rh. This phenomenon reveals the electronic interaction between Cu single-atom and Rh in Cu_SA-Rh MAs. The Cu K-edge EXAFS spectra show the distinct peak of Cu_SA-Rh MAs at 2.43 Å ascribed to the Cu-Rh bond, obviously distinct with Cu-Cu (2.23 Å) and CuO (1.51 Å) bands (Fig. 2k), revealing the presence of dispersed Cu single-atom. The atomically distributed Cu atoms were also identified based on the fitting results (Supplementary Figs. 10–12 and Supplementary Table 1). As displayed in Fig. 2l, The WT spectra for Cu_SA-Rh MAs, Rh foil, and Rh₂O₃ further indicate the presence of isolated Cu single-atom, where the intensity maximum (~10.31 Å⁻¹) of Cu_SA-Rh MAs is ascribed to the Cu-Rh bond compared with Cu-Cu (~8.01 Å⁻¹) and CuO (~6.76 Å⁻¹) intensity maximums.

Electrocatalytic performance for Ph-NO₂ ERR

The cathodic electrochemical Ph-NH₂ synthesis activity of Cu_SA-Rh MAs/CF was evaluated in an H-type electrolyzer separated via a Nafion membrane under ambient conditions, consisting of a working electrode (prepared electrocatalysts), a reference electrode (Hg/HgO electrode) and a counter electrode (Pt foil). As shown in Fig. 3a, linear sweep voltammetry (LSV) curves of Cu_SA-Rh MAs/CF were recorded in 1 M KOH and 1 M KOH + 5 mM Ph-NO₂ solutions. Apparently, a rapid increase in cathodic current density is observed at ~0.4 V (vs RHE) in the presence of Ph-NO₂, suggesting the proceeding of Ph-NO₂ electroreduction on Cu_SA-Rh MAs/CF. Furthermore, at the applied potential range from −0.7 to 0.5 V (vs RHE), the significantly improved overall current density and lower onset potential in a 1 M KOH + 5 mM Ph-NO₂ solution further demonstrate the favorable Ph-NO₂ ERR activity on Cu_SA-Rh MAs/CF. To investigate the potential-dependent Ph-NO₂ conversion and Ph-NH₂ selectivity on Cu_SA-Rh MAs/CF, the chronoamperometric (i-t) measurements were performed at various applied potentials and the electrolyzed products were quantified by GC analysis. As shown in Fig. 3b, the Ph-NO₂ conversion of Cu_SA-Rh MAs/CF is as high as ~100% at various applied potentials, revealing the superior Ph-NO₂ adsorption and electroreduction ability on Cu_SA-Rh MAs/CF. For Ph-NH₂ selectivity (Fig. 3c), the Ph-NH₂ selectivity of Cu_SA-Rh MAs/CF gradually increases from the applied potentials of 0.2 to −0.2 V (vs RHE), and the Ph-NH₂ selectivity is calculated to be ~99.7% at −0.1 V (vs RHE) approaching 100%, which indicates the Ph-NO₂ can be selectively electroreduced to Ph-NH₂ via hydrogenation using H₂O as the hydrogen source on Cu_SA-Rh MAs/CF. Notably, the low Ph-NH₂ selectivity of lower cathodic potentials can be ascribed to the weak ability of H₂O dissociation to produce active H*, and the production of active H* is gradually improved on Cu_SA-Rh MAs/CF with the increase of cathodic potentials thus enhancing the Ph-NH₂ selectivity¹¹. Besides, Fig. 3d and Supplementary Fig. 13 reveal that the Ph-NO₂ can be completely converted to Ph-NH₂ as well as the high-boiling by-products are not present in the obtained electrolyte at an applied potential of −0.1 V (vs RHE), further indicating the superior activity for electrochemical Ph-NH₂ synthesis on Cu_SA-Rh MAs/CF. As revealed in Fig. 3e, f, the time-dependent conversion, and yield illustrate that the Ph-NH₂ concentration increases and the Ph-NO₂ concentration decrease over time, indicating a gradual and complete conversion of Ph-NO₂ to Ph-NH₂ on Cu_SA-Rh MAs/CF at an applied potential of −0.1 V (vs RHE) for 1 h. To further investigate the mechanism for the improved Ph-NO₂ ERR to Ph-NH₂ activity over Cu_SA-Rh MAs/CF, several comparison experiments were carried out. LSV curves for various electrocatalysts demonstrate that the Cu_SA-Rh MAs/CF possesses a stronger electroreduction activity compared with Rh metallene-CF (Rhene-CF), Rh nanoparticles-CF (Rh NPs-CF), and CF (Supplementary Fig. 14a), and the Ph-NH₂ selectivity (99.7%) of Cu_SA-Rh MAs/CF is much higher than those of Rhene-CF (64.7%), Rh NPs-CF (59.3%) and CF (28.3%) (Supplementary Fig. 14b). Moreover, with respect to the standard potential (0.89 V vs RHE) of the Ph-NO₂ ERR²¹, the overpotential (0.532 V vs RHE) of Cu_SA-Rh MAs/CF is lower than those of Rhene-CF (0.667 V vs RHE), Rh NPs-CF (0.619 V vs RHE) and CF (0.638 V vs RHE) for achieving a current density of −50 mA cm⁻² (Supplementary Fig. 15), further suggesting a superior Ph-NO₂ ERR activity on Cu_SA-Rh MAs/CF. The improved activity of Ph-NO₂ ERR to Ph-NH₂ for Cu_SA-Rh MAs/CF originates from the ultrathin metallene array structure and the synergistic effect of isolated Cu single-atom with Rh host. Moreover, the stability of electrochemical Ph-NH₂ synthesis is also a critical parameter for evaluating Ph-NO₂ ERR activity. After 5 repeated cycles of testing on Cu_SA-Rh MAs/CF, the decay was almost negligible for Ph-NO₂ conversion, Ph-NH₂ selectivity, and Ph-NH₂ intensity (Fig. 3g–i), revealing a superb stability for Ph-NO₂ ERR to Ph-NH₂ of Cu_SA-Rh MAs/CF.

Fig. 3: Electrocatalytic Ph-NO₂ ERR performance.

a IR-corrected LSV curves for Cu_SA-Rh MAs/CF (catalyst loading: 2.5 mg cm⁻²) in 1 M KOH (pH = 13.9) and 1 M KOH + 5 mM Ph-NO₂ (pH = 13.7) solutions. b Ph-NO₂ conversion and c Ph-NH₂ selectivity for Cu_SA-Rh MAs/CF at various applied potentials in 1 M KOH + 5 mM Ph-NO₂ (pH = 13.7) solutions. d GC results of products before and after Ph-NO₂ ERR for Cu_SA-Rh MAs/CF at −0.1 V (vs RHE). e Time-dependent conversion and yield for Ph-NO₂ ERR of Cu_SA-Rh MAs/CF. f GC results for Ph-NO₂ ERR of Cu_SA-Rh MAs/CF at −0.1 V (vs RHE) with various time. g Ph-NO₂ conversion, h Ph-NH₂ selectivity and i GC results for Ph-NO₂ ERR of Cu_SA-Rh MAs/CF at −0.1 V (vs RHE) measured for 5 successive cycles in 1 M KOH + 5 mM Ph-NO₂ (pH = 13.7) solutions.

Electrocatalytic performance for MOR

A single-chamber cell with a three-electrode system was utilized to investigate the anodic electrochemical methanol (CH₃OH) upgrading for high value-added formate (HCOO⁻) production on Cu_SA-Rh MAs/CF. The working electrode contained prepared electrocatalysts, while Hg/HgO electrode and carbon rod served as reference electrode and counter electrode. Figure 4a displays the LSV curves of Cu_SA-Rh MAs/CF recorded in 1 M KOH with different CH₃OH concentrations solutions. Visibly, the Cu_SA-Rh MAs/CF exhibits an optimal MOR activity when the CH₃OH concentration reaches 4 M. The oxidation potentials (E₅₀, E_100, E₁₅₀, and E₂₀₀) of only 1.40, 1.44, 1.46, and 1.47 V (vs RHE) are required for reaching the current densities of 50, 100, 150 and 200 mA cm⁻² at a 1 M KOH + 4 M CH₃OH solution (Fig. 4b). which is lower than those of other CH₃OH concentrations. Notably, the MOR activity decays significantly when the CH₃OH concentration increases to 8 M, which can originate from the poor conductivity of the mixed solution as well as excessive oxidation species adsorbed on the active site to hinder the reaction²⁰. Hence, it is considered that a 4 M CH₃OH is appropriate for this system. Figure 4c illustrates the LSV curves of Cu_SA-Rh MAs/CF recorded in MOR (with 4 M CH₃OH) and OER (without 4 M CH₃OH). Observably, the onset potential is significantly reduced after the addition of CH₃OH. The oxidation potentials (E₅₀ = 1.40, E₁₀₀ = 1.44, E₁₅₀ = 1.46 and E₂₀₀ = 1.47 V vs RHE) for MOR are lower than those for OER (E₅₀ = 1.57, E₁₀₀ = 1.60, E₁₅₀ = 1.62 and E₂₀₀ = 1.64 V vs RHE) and the Tafel slope value of MOR (126 mV dec⁻¹) is also smaller than that of OER (189 mV dec⁻¹) (Fig. 4d, e), indicating the thermodynamic favorability of MOR over OER owing to faster kinetics and smaller oxidation potential. To further investigate the products of the anodic MOR on Cu_SA-Rh MAs/CF, the i-t measurements were performed at various applied potentials and the electrolyzed products were analyzed via the nuclear magnetic resonance (NMR) spectra. As presented in Fig. 4f, the key value-added chemical HCOO⁻ is produced during the MOR process, and the HCOO⁻ concentration gradually increases with higher anodic applied potential. Figure 4g presents that the HCOO⁻ Faradic efficiencies (FEs) for Cu_SA-Rh MAs/CF reaches over 90% at various applied potentials (1.35–1.55 V vs RHE), where the HCOO⁻ FE of Cu_SA-Rh MAs/CF is ~96.5% at 1.5 V (vs RHE), which reveals the excellent selectivity for the conversion of CH₃OH to HCOO⁻ on Cu_SA-Rh MAs/CF. Moreover, the ¹³C NMR spectra of the electrolyzed products after 24 h MOR further indicate the superior HCOO⁻ selectivity for Cu_SA-Rh MAs/CF and the generated CO₃²⁻ as a by-product is almost negligible (Supplementary Fig. 16). It is worth mentioning that the Cu_SA-Rh MAs/CF possesses a superior MOR activity compared with Rhene-CF, Rh NPs-CF and CF (Supplementary Fig. 17a). Meanwhile, relative to the standard potential (0.103 V vs RHE) of MOR¹⁷, the overpotential of Cu_SA-Rh MAs/CF is 1.25 V (vs RHE) for reaching a current density of 20 mA cm⁻², which is lower than those of Rhene-CF (1.30 V vs RHE), Rh NPs-CF (1.31 V vs RHE) and CF (1.35 V vs RHE) (Supplementary Fig. 17b). Additionally, the MOR stability of Cu_SA-Rh MAs/CF was evaluated by the repeated i-t tests under a constant potential. Figure 4h shows that the Cu_SA-Rh MAs/CF consistently maintains the HCOO⁻ FEs as high as ~90% during the five repeated cycles testing, indicating excellent stability for HCOO⁻ production on Cu_SA-Rh MAs/CF. To investigate the charge transfer kinetics at the electrocatalytic interface, Electrochemical impedance spectroscopy (EIS) for various electrocatalysts was recorded in a 1 M KOH solution. As displayed in Supplementary Fig. 18, the Cu_SA-Rh MAs/CF (1.13 Ω) presents a smaller resistance (R_ct) value than Rhene-CF (1.96 Ω), Rh NPs-CF (3.34 Ω) and CF (9.68 Ω), revealing a rapid interfacial charge transfer on Cu_SA-Rh MAs/CF^52,53. Besides, the electrochemical active surface areas (ECSAs) for various electrocatalysts were evaluated by the electrochemical double-layer capacitance (C_dl) calculated based on cyclic voltammetry (CV) curves (Supplementary Fig. 19). The C_dl value for Cu_SA-Rh MAs/CF (49.9 mF cm⁻²) was calculated to be higher than those of Rhene-CF (35.9 mF cm⁻²), Rh NPs-CF (17.1 mF cm⁻²) and CF (2.1 mF cm⁻²) (Supplementary Fig. 20), which reveals the rich active sites in Cu_SA-Rh MAs/CF due to the ultrathin metallene array structure and the introduction of isolated Cu single-atom.

Fig. 4: Electrocatalytic MOR performance.

a IR-corrected LSV curves for Cu_SA-Rh MAs/CF (catalyst loading: 2.5 mg cm⁻²) and b corresponding overpotentials comparison at 50/100/150/200 mA cm⁻² in 1 M KOH containing 0.5 M (pH = 13.9), 1 M (pH = 13.8), 2 M (pH = 13.8), 4 M (pH = 13.7) and 8 M (pH = 13.7) CH₃OH solutions. c LSV curves, d corresponding overpotentials comparison at 50/100/150/200 mA cm⁻² and e Tafel plots of Cu_SA-Rh MAs/CF for MOR and OER in 1 M KOH (pH = 13.9) and 1 M KOH + 4 M CH₃OH (pH = 13.7) solutions. f ¹H NMR spectra of products after MOR at various applied potentials. g HCOO⁻ FEs of Cu_SA-Rh MAs/CF at various applied potentials in 1 M KOH + 4 M CH₃OH (pH = 13.7) solutions. h HCOO⁻ FEs of Cu_SA-Rh MAs/CF at 1.4 V (vs RHE) measured for 5 successive cycles in 1 M KOH + 4 M CH₃OH (pH = 13.7) solutions.

Electrocatalytic performance for Ph-NO₂ ERR-MOR system

Inspired by the above results of two half-reactions, a two-electrode system (Ph-NO₂ ERR-MOR) was constructed by employing the Cu_SA-Rh MAs/CF as cathode and anode. The constructed coupling system enables the simultaneous conversion of organic small molecules to targeted high-value-added chemicals and greatly optimizes energy efficiency. As shown in Fig. 5a, LSV curves of Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF were recorded in the Ph-NO₂ ERR-MOR system and overall water splitting (OWS) system. Apparently, the Ph-NO₂ ERR-MOR system is driven requiring a lower onset potential compared with OWS, indicating a faster reaction kinetics. The cell voltages (1.05, 1.18, and 1.35 V) of Ph-NO₂ ERR-MOR system are lower than OWS system (2.04, 2.40, and 2.71 V) at current densities of 50,100, and 150 mA cm⁻² (Fig. 5b), revealing the improved electrolysis efficiency and lower electrolysis energy consumption in the Ph-NO₂ ERR-MOR system. For evaluating the selectivity and activity of Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF in the Ph-NO₂ ERR-MOR system, the electrolyzed products of cathode and anode were collected and analyzed at various applied potentials. As presented in Fig. 5c, d, the cathodic Ph-NO₂ conversion and Ph-NH₂ selectivity are up to ~100% as well and the anodic HCOO⁻ FEs also reaches over ~90% under various applied potentials, which indicates the efficient and selective targeted production of Ph-NH₂ and HCOO⁻ at cathode and anode. Furthermore, almost no by-products are generated in the Ph-NO₂ ERR-MOR coupling reaction of Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF (Fig. 5e). More encouragingly, the Ph-NO₂ conversion/Ph-NH₂ selectivity/HCOO⁻ FE of Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF remain stable after 5 repeated cycles testing in the Ph-NO₂ ERR-MOR system (Fig. 5f). Furthermore, after stability testing, no significant degradation is observed for the morphology and structure of Cu_SA-Rh MAs/CF (Supplementary Fig. 21a, b), and the crystal structure of Cu_SA-Rh MAs remains stable (Supplementary Fig. 21c, d). Notably, Supplementary Fig. 22 further reveals that the elemental composition and chemical state of Cu_SA-Rh MAs show no significant change after stability testing. These conclusions indicate superior stability for the Ph-NO₂ ERR-MOR coupling system constructed by the Cu_SA-Rh MAs/CF.

Fig. 5: Electrocatalytic performance of Ph-NO₂ ERR-MOR system.

a IR-corrected LSV curves and b corresponding voltages comparison for Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF (2.5 mg cm⁻²) in the coupled Ph-NO₂ ERR-MOR system (cathode: 1 M KOH + 5 mM Ph-NO₂ solution, pH = 13.7 and anode: 1 M KOH + 4 M CH₃OH solution, pH = 13.7) and overall water splitting system (OWS, cathode and anode: 1 M KOH solutions, pH = 13.9). c Ph-NO₂ conversion and Ph-NH₂ selectivity for Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF under various potentials at the cathode. d HCOO⁻ FEs for Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF under various potentials at the anode. e GC result of products for cathodic Ph-NO₂ ERR and ¹H NMR spectra of products for anodic MOR over Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF at 1.2 V. f Ph-NO₂ conversion, Ph-NH₂ selectivity and HCOO⁻ FEs for Cu_SA-Rh MAs/CF||Cu_SA-Rh MAs/CF during the coupled reaction at 1.2 V measured for 5 successive cycles.

DFT calculations

To further reveal the electronic structure effect between the isolated Cu single-atom and Rh host as well as the mechanism of enhanced electrocatalytic activity on Cu_SA-Rh MAs/CF, DFT calculations were carried out. Supplementary Fig. 23 shows an optimized geometric structure model for Cu_SA-Rh MAs. Based on the Bader charge calculation analysis (Supplementary Fig. 24 and Supplementary Table 2), the total net charges of Rh and Cu are 1.94 e and −1.94 e, respectively, which reveals that the electron transfer is from the Cu single-atom to the Rh host. Fig. 6a, b and Supplementary Fig. 25 reveal the accumulation of negative charge around the Rh host while Cu single-atom possesses positive charge property due to the loss of charge, which further indicate the electron-rich nature of Rh localization resulting from the introduction of isolated Cu single-atom. The projected partial density of states (PDOSs) of Cu_SA-Rh MAs show the strong overlapping of the Rh-4d orbital with the Cu-3d orbital (Fig. 6c), revealing the effective site-to-site electron transfer between the isolated Cu single-atom and the Rh host, which contributes to the optimization of electronic structures thus facilitating electrocatalytic reactions^54,55. Moreover, as displayed in Fig. 6d, the d-band center of Rh-4d orbitals for Cu_SA-Rh (111) (−1.79 eV) exhibits a slight negative shift compared with Rh (111) (−1.77 eV). It is notable that the PDOS for Rh-4d of Cu_SA-Rh bulk (−1.82 eV) and Rh bulk (−1.49 eV) also reflect a similar trend (Supplementary Fig. 26). As shown in Supplementary Fig. 27 and Supplementary Table 3, The electron transfer between Cu_SA-Rh and Ph-NO₂* (0.19 e) is smaller than that between Rh and Ph-NO₂* (0.25 e) owing to the electron interaction between Cu single-atom and the Rh host, which eventually can induce the decrease of electron interaction between Cu_SA-Rh and Ph-NO₂*, thus weakening the binding of Ph-NO₂* on Cu_SA-Rh⁴³. Notably, there is electron accumulation in the antibonding orbitals for both Cu_SA-Rh and Rh, which favors the weakening of the bond energy for the N–O bond^43,56. These conclusions indicate that the electron interaction between the isolated Cu single-atom and the Rh host causes a downshift of the d-band center and a decrease in the electron interaction between the catalyst and adsorbate, thus promoting a fast conversion of the reactants to key intermediates as well as optimizing the desorption of the target product^8,43. Moreover, Supplementary Fig. 28a and Supplementary Table 4 present the optimized adsorption structure and adsorption energy (∆Eads) of Ph-NO₂ on Rh (111) and Cu_SA-Rh (111) surfaces. The calculated Ph-NO₂ ∆Eads of Cu_SA-Rh (111) is −1.08 eV, which is lower than that of Rh (111) (−1.27 eV). The weakened ∆Eads of Ph-NO₂ on Cu_SA-Rh is beneficial for the rapid conversion of Ph-NO₂ to Ph-NOOH* by coupling with the protons in H₂O, and the unstable Ph-NOOH* rapidly removes OH* to further form the stable and important intermediate Ph-NO*, thus enhancing the rapid protonation in the Ph-NO₂ ERR process⁸. Notably, Supplementary Fig. 28b and Supplementary Table 4 show that the ∆Eads of the key intermediate Ph-NO* on Cu_SA-Rh (111) (−1.34 eV) is higher than that of Rh (111) (−1.28 eV), indicating that the intermediate Ph-NO* remains stable on Cu_SA-Rh and can be steadily converted to Ph-NOH* thus reducing the desorption of Ph-NO* to avoid the generation of undesired by-product azobenzene. Moreover, the free energy profiles of optimized intermediates for Ph-NO₂ ERR pathway reveal Ph-NOH*→Ph-NHOH* and Ph-NH₂*→Ph-NH₂ as the rate-determining step (RDS) for Cu_SA-Rh and Rh, respectively (Fig. 6e, f and Supplementary Fig. 29). Obviously, the Cu_SA-Rh exhibits a lower energy barrier (0.74 eV) on the RDS compared with Rh (0.85 eV) (Fig. 6f and Supplementary Table 5), further indicating the higher ability on Cu_SA-Rh for driving the Ph-NO₂ ERR to Ph-NH₂. To further investigate the enhanced mechanism of activity during the Ph-NO₂ hydrogenation with H₂O as the hydrogen source on Cu_SA-Rh MAs, a series of DFT calculations were performed and analyzed. Figure 6g and Supplementary Fig. 30 display the calculated energy profiles of H₂O dissociation process and corresponding optimized structures of the initial, transition, and final states. The energy barrier for H₂O dissociation of Cu_SA-Rh (0.83 eV) is lower than that of Rh (1.01 eV) (Fig. 6g and Supplementary Table 6), indicating that the introduction of the Cu single-atom is beneficial for the dissociation of H₂O on Cu_SA-Rh to facilitate the formation H* for Ph-NO₂ hydrogenation. Furthermore, the H* ∆Eads of various sites (A to E) were further calculated for investigating the hydrogenation mechanism of H* with H₂O as the hydrogen source on Cu_SA-Rh MAs (Fig. 6h, i). Visually, the variation of H* ∆Eads from A to E sites on Cu_SA-Rh reveals a gradually increased H* adsorption (Fig. 6i and Supplementary Table 7), which indicates a H*-spillover process between the isolated Cu single-atom and the Rh host. This can be ascribed to local adsorption differences induced by electronic effect of Cu single-atom and Rh host. In the H*-spillover process (Supplementary Fig. 31), the formed H* from H₂O dissociation on Rh host can spontaneously and rapidly migrate on the isolated Cu single-atom thus stabilizing H* for the hydrogenation process of intermediates, which is beneficial for inhibiting competing reaction HER and avoiding the H* accumulation on the Rh host to occupy the active sites thus maximizing the utilization of H* to promote synergistic electrocatalytic effects on Cu single-atom and Rh host for Ph-NO₂ ERR. For further investigating the MOR mechanism on Cu_SA-Rh MAs, the free energy profiles of optimized intermediates for MOR pathway were analyzed by the DFT calculations (Fig. 6j, k and Supplementary Fig. 32). It can be observed that the Cu_SA-Rh possesses a lower energy barrier for RDS (CH₂O* → CHO*, 0.55 eV) compared to the RDS (CHO* → HCOOH*, 0.62 eV) of Rh (Fig. 6k and Supplementary Table 8), indicating a more favorable MOR process on Cu_SA-Rh. Additionally, the CH₃OH-to-HCOO⁻ is regarded as a continuous dehydrogenation process of CH₃OH, and the adsorption of the key intermediate CHO* is critical for the HCOO⁻ synthesis. As shown in Supplementary Fig. 33 and Supplementary Table 9, the ∆Eads of CHO* for Cu_SA-Rh (111) (−5.09 eV) is lower than that for Rh (111) (−5.33 eV). Meanwhile, Fig. 6k further shows that the energy barrier for the desorption step (HCOOH* → HCOOH) of Cu_SA-Rh (0.43 eV) is lower than that of Rh (0.54 eV). These conclusions reveal that the introduction of Cu single-atom can effectively reduce the ∆Eads of key intermediate CHO* leading to the easier desorption of the target product from the Cu_SA-Rh MAs surface, which is conducive to facilitating the HCOO⁻ formation and avoiding the unnecessary oxidation process. Moreover, the MOR mechanism on Cu_SA-Rh was speculated as illustrated in Supplementary Fig. 34. Due to the electron effect between Cu and Rh, the electron transfer from the Cu single-atom to Rh host can activate electron-rich Rh (reaction 1).

$${{{{{\rm{Cu}}}}}}+{{{{{\rm{Rh}}}}}}\to {{{{{\rm{Cu}}}}}}(+)/{{{{{\rm{Rh}}}}}}(-)$$

(1)

Fig. 6: DFT calculations analysis.

a Charge density difference on Cu_SA-Rh model. The blue and red spheres represent Rh and Cu atoms respectively, as well as the yellow and cyan indicate the charge depletion and accumulation areas. b Slice of charge density difference for Cu_SA-Rh bulk structure. c The PDOSs of Cu_SA-Rh. d The PDOSs for Rh-4d of Rh (111) and Cu_SA-Rh (111) surfaces. e Optimized structures of Ph-NO₂ ERR intermediates on Cu_SA-Rh (111). The blue, red, gray, cyan, purple, and white spheres represent Rh, Cu, C, O, N, and H atoms respectively. f Comparison of free energy profiles for Ph-NO₂ ERR pathway on Cu_SA-Rh (111) and Rh (111). g Calculated energy profiles of H₂O dissociation on Cu_SA-Rh (111) and Rh (111). h Optimized structures of adsorbed H* and i calculated H* adsorption energies variations at different sites on Cu_SA-Rh (111). The blue, red, and white spheres represent Rh, Cu, and H atoms respectively. j Optimized structures of MOR intermediates on Cu_SA-Rh (111). The blue, red, gray, cyan, and white spheres represent Rh, Cu, C, O, and H atoms respectively. k Comparison of free energy profiles for MOR pathway on Cu_SA-Rh (111) and Rh (111).

Initially, The CH₃OH tends to adsorb on the electron-rich Rh (reaction 2), and then the CHO* is produced through a series of dehydrogenation reactions on Cu_SA-Rh (reactions 3 and 4).

$${{{{{\rm{Rh}}}}}}(-)/{{{{{{\rm{CH}}}}}}}_{3}{{{{{\rm{OH}}}}}}+{{{{{{\rm{OH}}}}}}}^{-}\to {{{{{\rm{Rh}}}}}}(-)/{{{{{{\rm{CH}}}}}}}_{2}{{{{{{\rm{OH}}}}}}}^{*}+{{{{{{\rm{e}}}}}}}^{-}+{{{{{{\rm{H}}}}}}}_{2}{{{{{\rm{O}}}}}}({{{{{\rm{l}}}}}})$$

(2)

$${{{{{\rm{Rh}}}}}}(-)/{{{{{{\rm{CH}}}}}}}_{2}{{{{{{\rm{OH}}}}}}}^{*}+{{{{{{\rm{OH}}}}}}}^{-}\to {{{{{\rm{Rh}}}}}}(-)/{{{{{{\rm{CH}}}}}}}_{2}{{{{{{\rm{O}}}}}}}^{*}+{{{{{{\rm{e}}}}}}}^{-}+{{{{{{\rm{H}}}}}}}_{2}{{{{{\rm{O}}}}}}({{{{{\rm{l}}}}}})$$

(3)

$${{{{{\rm{Rh}}}}}}(-)/{{{{{{\rm{CH}}}}}}}_{2}{{{{{{\rm{O}}}}}}}^{*}+{{{{{{\rm{OH}}}}}}}^{-}\to {{{{{\rm{Rh}}}}}}(-)/{{{{{{\rm{CHO}}}}}}}^{*}+{{{{{{\rm{e}}}}}}}^{-}+{{{{{{\rm{H}}}}}}}_{2}{{{{{\rm{O}}}}}}({{{{{\rm{l}}}}}})$$

(4)

Finally, the HCOO⁻ is produced by nucleophilic attack of OH⁻ adsorbed around the positively charged Cu against the electrophilic carbocation in CHO* and rapidly desorbs owing to the weak adsorption energy of CHO* (reaction 5).

$${{{{{\rm{Rh}}}}}}(-)/{{{{{{\rm{CHO}}}}}}}^{*}+{{{{{\rm{Cu}}}}}}(+)/{{{{{{\rm{OH}}}}}}}^{-}\to {{{{{\rm{Rh}}}}}}(-)/{{{{{\rm{Cu}}}}}}(+)+{{{{{\rm{HCO}}}}}}{{{{{{\rm{O}}}}}}}^{-}+{{{{{{\rm{e}}}}}}}^{-}+{{{{{\rm{H}}}}}}2{{{{{\rm{O}}}}}}({{{{{\rm{l}}}}}})$$

(5)

Based on the above conclusion analysis, the superior Ph-NO₂ ERR and MOR activity of Cu_SA-Rh MAs/CF originates from the following points: Firstly, the stable security wall-like structure formed by the ultrathin metallene arrays provides sufficient active sites and abundant interlayer channels⁴⁵. Secondly, the inherent defect-rich structure and low-crystalline regions of Cu_SA-Rh MAs/CF can induce unsaturated coordination metallic bonds and optimize the local electron structure^37,56. Thirdly, the synergistic catalysis effect and H*-spillover effect between Cu single-atom and Rh host can optimize the catalytic reaction process, facilitate the stable and rapid conversion of reactants to intermediates as well as accelerate the desorption of target products. Fourthly, the Cu single-atom as effective adsorption sites can modulate the competition for adsorbate adsorption on Rh sites thus promoting electrocatalytic reactions.

Discussion

In summary, we have successfully synthesized the Cu_SA-Rh MAs/CF with the isolated Cu single-atom dispersed on Rh metallene that possesses excellent electrocatalytic activity towards Ph-NO₂ ERR and MOR. The Cu_SA-Rh MAs/CF exhibits the low cell voltages (1.05 and 1.18 V) to achieve the current densities of 50 and 100 mA cm⁻² in the Ph-NO₂ ERR-MOR coupled system, and the cathodic Ph-NO₂ conversion/Ph-NH₂ selectivity are up to ~100% as well as the anodic HCOO⁻ FEs also reaches over ~90%. The constructed coupled organic electrocatalytic conversion system achieves simultaneous conversion of low-value organic compounds to high-value chemicals at both cathode and anode as well as improves energy efficiency. The electron effect between Cu single-atom and Rh host and the catalytic reaction mechanism were further revealed by DFT calculations, in which the synergistic catalytic effect and H*-spillover effect are triggered by the local electronic structure change between the Cu single-atom and Rh host, which optimizes the catalytic reaction process and facilitate the rapid production of key intermediates and the rapid desorption of target products. This work provides a novel strategy for the design of SAAs catalysts applied in the sustainable green synthesis of high-value chemicals.

Methods

Materials and chemicals

Cu foam (CF) was provided from Changsha Lyrun Material Co., Ltd. Rhodium (II) acetate dimer (C₈H₁₂O₈Rh₂, Rh: 43–46%), rhodium chloride (RhCl₃, 98%), potassium hydroxide (KOH, AR, 85%), methanol (CH₃OH, AR, 99.5%), ethylene glycol (EG, AR, 99%), N,N-dimethylformamide (DMF, AR, 99.5%), diethylenetriamine (DETA, AR, 99%) and maleic acid (≥99%) were purchased from Aladdin Industrial Corporation (China). Ethyl acetate (99%), nitrobenzene (Ph-NO₂, AR, 99%) and aniline (Ph-NH₂, ≥99.5%) were ordered from Shanghai Macklin Biochemical Co., Ltd. (China). Ethanol (C₂H₅OH, 95%) and hydrochloric acid (HCl, 37%) were acquired from Sinopharm Chemical Reagent Co., Ltd. (China). High-purity Ar gas (99.99%) was obtained from Hangzhou Special Gases Co., Ltd. (China).

Synthesis of Cu_SA-Rh MAs/CF

For a typical synthesis of Cu_SA-Rh MAs/CF, the 1 × 2 cm² of CF was soaked in a 3 M HCl solution for 20 min, followed by washing several times with ethanol/water, and then dried at 50 °C in a vacuum oven for further utilization. Typically, 10 mg of C₈H₁₂O₈Rh₂ and 1 g of KOH were ultrasonically dissolved in a mixture of DMF (6 mL) and EG (4 mL). Then, 5 mL of DETA was dropped into the above solution to form a homogeneous solution, the obtained solution was transferred into a 25 mL Teflon-lined autoclave and the treated CF was placed into the autoclave. Afterward, the autoclave was heated to 200 °C and kept for 8 h. After naturally cooling to room temperature, the product was extracted from the autoclave and washed several times with water/ethanol. Finally, the product was dried in a vacuum oven at 60 °C for further characterization and electrochemical measurements.

Synthesis of Rhene-CF

For a typical synthesis of Rhene-CF, 1 g of KOH and 10 mg of C₈H₁₂O₈Rh₂ were dissolved in a 30 mL vial containing 6 mL of DMF and 4 mL of EG. Then, 5 mL of DETA was slowly added to the above solution to form a homogeneous solution. The obtained solution was transferred to a 25 mL of Teflon-lined autoclave, followed by heating to 200 °C and maintaining for 1 h. After that, the product was obtained by centrifuging and washing several times with water/ethanol and dried in a vacuum oven at 60 °C for further utilization. Finally, the Rhene-CF was prepared by covering the prepared Rhene on CF.

Synthesis of Rh NPs-CF

For a typical synthesis of Rh NPs-CF, 10 mg of RhCl₃ was dissolved in 30 mL of EG to form a homogeneous solution. The above solution was transferred to a 50 mL Teflon-lined autoclave and heated at 200 °C for 2 h. Then, the product was collected by centrifuging and washing several times with water/ethanol and dried in a vacuum oven at 60 °C for further utilization. Finally, the Rh NPs-CF was prepared by covering the prepared Rh NPs on CF.

Material characterizations

Scanning electron microscopy (SEM) images were collected on a Zeiss Gemini 500. TEM images and selected area electron diffraction (SAED) data were collected on an FEI Tecnai G2 F30 (300 kV). Aberration-corrected high angle annular dark field scanning transmission electron microscopy (AC-HAADF-STEM) images, EELS, and elemental mapping data were collected on a Thermo Scientific Themis Z. Atomic force microscopy (AFM) was conducted on a Bruker Multimode 8. XRD data were recorded with a PANalytical Empyrean powder diffractometer using Cu Kα radiation and XPS data were collected on a Thermo ESCALAB 250XI. Rh K-edge and Cu K-edge XANES and EXAFS data were acquired from a BL14W1 beamlines at the Shanghai Synchrotron Radiation Facility (Shanghai, China). The inductively coupled plasma optical emission spectroscopy (ICP-OES) data were collected from a PerkinElmer ICP 2100. Nuclear magnetic resonance (NMR) spectra were collected on a Bruker Avance NEO 600 and gas chromatograph (GC) was conducted on an Agilent 8890/7000D.

Electrochemical experiments

Electrochemical measurements were conducted on a CHI760E electrochemical workstation (Shanghai Chenhua Instrument Corporation, China) via a two/three-electrode system, where contains the reference electrode (Hg/HgO electrode), counter electrode (Pt foil and carbon rod), and working electrode. As a blinder-free electrode, the Cu_SA-Rh MAs/CF can be directly used as a working electrode, where Rhene-CF, Rh NPs-CF and CF serve as the comparison. Regarding cathodic Ph-NO₂ electroreduction reaction (Ph-NO₂ ERR), all electrochemical tests were conducted in an H-type cell with the Nafion 117 membrane separation, where the cathode chamber contains working electrodes and Hg/HgO electrode, and the anode chamber contains Pt foil. The cathode and anode electrolytes were a 1 M KOH + 5 mM Ph-NO₂ solution and a 1 M KOH solution, respectively. Before the Ph-NO₂ ERR test, Ar gas (99.99%) was continuously passed into the cathode chamber to purify the electrolyte for 20 min. More importantly, the cathode chamber needs to be continuously flooded with Ar gas during the whole Ph-NO₂ ERR tests. Regarding anodic methanol oxidation reaction (MOR), electrochemical tests were conducted in a single-chamber cell via a three-electrode system containing a working electrode, reference electrode (Hg/HgO electrode), and counter electrode (carbon rod). Regarding Ph-NO₂ ERR-MOR two-electrode system, the Cu_SA-Rh MAs/CF served as the cathode and anode electrodes, respectively. The cathode and anode electrolytes were a 1 M KOH + 5 mM Ph-NO₂ solution and a 1 M KOH + 4 M CH₃OH solution, respectively. All linear sweep voltammetry (LSV) curves were recorded at a scan rate of 5 mV s⁻¹ with 95% iR-compensation, and the current density for Ph-NO₂ ERR and MOR was acquired by normalizing to the geometric area of the CF. The applied potentials are converted into reversible hydrogen electrode (RHE) scale according to the Eq. (6):

$${{{{{\rm{E}}}}}}({{{{{\rm{vs}}}}}}.{{{{{\rm{RHE}}}}}})={{{{{{\rm{E}}}}}}}^{{{{{{\rm{\theta }}}}}}}({{{{{\rm{Hg}}}}}}/{{{{{\rm{HgO}}}}}})+{{{{{\rm{E}}}}}}({vs}.{{{{{\rm{Hg}}}}}}/{{{{{\rm{HgO}}}}}})+0.059\times {{{{{\rm{pH}}}}}}.$$

(6)

EIS was carried out in the range of 0.1 Hz to 100 kHz.

Product analysis

For Ph-NO₂ ERR, the electrolyte was collected and extracted with ethyl acetate after i-t tests. The extracted products were determined by comparing the GC retention times and mass spectra. The Ph-NO₂ conversion and Ph-NH₂ selectivity were acquired via the GC results analysis and calculated based on the following Eqs. (7) and (8):

$${{\mbox{Conversion}}}=\frac{{{{\mbox{mol}}}} \, {{\mbox{of}}} \, {{\mbox{the}}} \, {{\mbox{consumed}}} \, {{\mbox{Ph}}}{\mbox{-}}{{{\mbox{NO}}}}_{2}}{{{\mbox{mole}}} \, {{\mbox{of}}} \, {{\mbox{the}}} \, {{\mbox{added}}} \, {{\mbox{Ph}}}{\mbox{-}}{{{\mbox{NO}}}}_{2}}\times 100\%$$

(7)

$${{\mbox{Selectivity}}}=\frac{{{{\mbox{mol}}}} \, {{{\mbox{of}}}} \, {{\mbox{the}}} \, {{\mbox{as}}} {\mbox{-}} {{\mbox{formed}}} \, {{\mbox{Ph}}}{\mbox{-}}{{{{{{\rm{NH}}}}}}}_{2}}{{{\mbox{mole}}} \, {{\mbox{of}}} \, {{\mbox{the}}} \, {{\mbox{added}}} \, {{\mbox{Ph}}}{\mbox{-}}{{{\mbox{NO}}}}_{2}}\times 100\%$$

(8)

For MOR, the electrolyte after i-t tests was collected and analyzed via the NMR spectroscopy, where maleic acid served as an internal standard. The HCOO⁻ FEs were calculated based on the following Eq. (9):

$${FE}({{{\mbox{HCOO}}}^{-}})=\frac{N({{\mbox{Production}}})\times Z\times F}{Q}\times 100\%$$

(9)

Where N is the molar amount for the formed HCOO⁻, Z is the number of electrons transferred for the formed HCOO⁻ (Z = 4), F is the Faraday constant (96485 C mol⁻¹), Q is the total amount of charge consumed.

Computational methods

Details of the calculation method can be found in the Supplementary method.