Oxyanions stabilized low-coordinated nickel species for efficient and durable nucleophilic electrooxidation Authors: Xiaokang Liu, Chengxiang Shi, Gong Zhang, Ruijie Gao, Lun Pan, Xiangwen Zhang, Zhen-Feng Huang, Jinlong Gong, Ji-Jun Zou Published: 2025 ABSTRACT Nucleophilic oxidation reactions (NORs) are key for high-value organics in electrosynthesis, but limited by slow dehydrogenation kinetics and competing oxygen evolution reactions (OER). While low-coordinated Ni species exhibit enhanced activity in rate-determining electrochemical deprotonation, they suffer irreversible oxidation with rapid performance degradation. Herein, we develop a class of oxyanion-stabilized low-coordinated Ni catalysts for highly efficient and durable oxidation of diverse nucleophiles including urea, methanol, and biomass-derived 5-hydroxymethylfurfural. Bicarbonate ligands form an electrostatically repulsive microenvironment, suppressing parasitic OER and Ni over-oxidation via dynamic ligand-cation coordination. These mediators form bifunctional channels, linking Ni2-δ/Ni3-δ reversible transition-based deprotonation and nucleophilic H-transfer, boosting dehydrogenation. A prototype catalyst engineered with bicarbonate ligands via electrochemical pre-reduction achieves 500 mA cm−2 at 1.42 V (versus reversible hydrogen electrode) for urea oxidation and sustains stability for 1100 hours at 100 mA cm−2. Scaling to 100 cm2 anion-exchange membrane reactor yields H2 and valorized organics, showing sustainable electrosynthesis potential. Introduction Organic electrosynthesis offers a promising approach for producing high-value chemicals and mitigating environmental pollutants1,2, which operates under mild reaction conditions, and eliminates the need for hazardous oxidizing/reducing agents3–5. Specifically, nucleophilic oxidation reactions (NORs) play a vital role in organic electrosynthesis, where nucleophiles such as urea, methanol, ethanol, glycerol, and 5-hydroxymethylfurfural undergo stepwise dehydrogenation, hydration and rearrangement6–8, but the efficiency is limited by the slow dehydrogenation kinetics and intensified competition from oxygen evolution reaction (OER)9,10. Ni-based materials are among the most promising electrocatalysts for NORs due to their low cost and tunable oxidation states of the metal centers10–12. The dehydrogenation in NORs involves two steps: electrochemical deprotonation of catalyst itself via proton-coupled electron transfer (Ni2+-OH + OH⁻ → Ni3+-O + H2O + e⁻), and chemical hydrogen transfer from the nucleophiles/intermediates to catalyst (Ni3+-O + *R-H → Ni2+-OH + *R) (Fig. 1a). The high energy barrier of electrochemical deprotonation brings an inherent limitation on Ni-based catalysts10,13. Low-coordinated metal catalysts can accelerate the electrochemical deprotonation14,15; unfortunately, the metal species are highly unstable and prone to irreversible oxidation towards coordinatively saturated states, which induces not only low activity but also intensified OER competition11,16. Moreover, rapid electrochemical deprotonation often leads to slow chemical hydrogen transfer from the nucleophiles/intermediates to the ligand oxygen center of metal species, as the rates of two steps are inversely related17–19. It is essential to stabilize low-coordinated metal species under NORs conditions while also balancing the kinetics between electrochemical deprotonation and chemical hydrogen transfer.Fig. 1Oxyanions stabilized low-coordinated Ni species for efficient and durable NORs.a Comparison of different active-site motifs for NORs. L denotes the oxyanion. b Schematic illustration of the transformation of Ni-based oxyanions to low-coordinated Ni species by electrochemical pre-reduction. c Schematic illustration of Ni(OH)x-L motif avoiding OER competition. Oxyanions stabilized low-coordinated Ni species for efficient and durable NORs. a Comparison of different active-site motifs for NORs. L denotes the oxyanion. b Schematic illustration of the transformation of Ni-based oxyanions to low-coordinated Ni species by electrochemical pre-reduction. c Schematic illustration of Ni(OH)x-L motif avoiding OER competition. Herein, we report a family of dynamically stable low-coordinated Ni-based catalysts, which are fabricated by a facile electrochemical pre-reduction from Ni-based oxyanions (e.g., bicarbonate, borate, oxalate, carbonate and phosphate) (Fig. 1b). The catalysts exhibit both high activity and stability for NORs of urea, methanol, ethanol, glycerol, and 5-hydroxymethylfurfural (HMF). Combining theoretical and experimental approaches, we demonstrate that the oxyanion ligands effectively protect low-coordinated Ni species from irreversible oxidation and suppress OER by increasing the surface charge density to electrostatically repel hydroxide ion adsorption (Fig. 1c). Additionally, the oxyanions bridge electrochemical deprotonation of catalyst and chemical hydrogen transfer from nucleophiles/intermediates, promote the reversible transition of low-coordinated Ni2-δ/Ni3-δ species and thereby accelerate the dehydrogenation kinetics. A representative Ni-based catalyst embedded with bicarbonate ligands achieved a current density of 500 mA cm−2 at only 1.42 V (versus reversible hydrogen electrode (RHE)) for the urea oxidation reaction (UOR) and maintained stable operation over 1100 h at 100 mA cm−2 (Supplementary Tables 1 and 2). Results Catalyst design and fabrication Ligand vacancies such as oxygen vacancies in metal oxides usually accelerate deprotonation14,20, however, these vacancies will strongly adsorb and be irreversibly filled by OH⁻. This transforms the low-coordinated Ni species into coordinatively saturated states, which not only reduces catalytic performance but also intensifies OER competition21,22. We employed density functional theory (DFT) calculations to probe whether oxyanion ligands, such as -HCO3, could confer improved activity and stability on low-coordinated Ni species. We constructed low-coordinated models with and without oxyanion ligands, i.e., Ni(OH)x and Ni(OH)x-HCO3 featuring with tetracoordinated Ni sites, and compared them with stoichiometric Ni(OH)2 (Supplementary Data 1). Density of states (DOS) calculations suggested that the low-coordinated Ni(OH)x significantly alters the electronic structure of metal-centered dx2-y2 and dxy orbitals (Fig. 2a and Supplementary Table 3), which leads to an enhanced contribution from the orbitals near the Fermi energy level, accompanied by overall orbital broadening (Fig. 2b, Supplementary Figs. 1, 2). Importantly, the Ni(OH)x-HCO3 further enhances the contribution of dxy orbitals at the Fermi level (circled area in Fig. 2a), which could stem from the π-d conjugation effect23,24, where the delocalized electrons from the -HCO3 interact with NiO6-x units, thereby enhancing the repulsion of dxy electrons (Supplementary Figs. 3, 4). As a result, we propose that this interaction may activate the redox activity of dxy orbitals (t2g* broader)25,26, and contribute to a decrease in the electron transfer barrier in electrochemical deprotonation that is typically the potential-determining step for NORs27,28. DFT calculations further predicted a stepwise reduction in energy barrier of electrochemical deprotonation, decreasing from 2.35 eV for Ni(OH)2 to 0.78 eV for Ni(OH)x and 0.37 eV for Ni(OH)x-HCO3 (Fig. 2c).Fig. 2Catalyst design and structural characteristics.a Density of states of catalysts with different motifs. b Schematic illustration of eg* and t2g* orbital broadening to accelerate electron transfer. c Deprotonation energy barriers. d Differential charge densities (cyan and yellow represent charge depletion and accumulation, respectively). e Adsorption energies of nucleophiles relative to OH⁻. f Fourier-transform infrared spectroscopy spectra of as-synthesized Ni HCs and PR-Ni HCs. g Normalized Ni K-edge XANES spectra of different Ni samples, and h the corresponding k3-weighted Fourier-transformed EXAFS. Catalyst design and structural characteristics. a Density of states of catalysts with different motifs. b Schematic illustration of eg* and t2g* orbital broadening to accelerate electron transfer. c Deprotonation energy barriers. d Differential charge densities (cyan and yellow represent charge depletion and accumulation, respectively). e Adsorption energies of nucleophiles relative to OH⁻. f Fourier-transform infrared spectroscopy spectra of as-synthesized Ni HCs and PR-Ni HCs. g Normalized Ni K-edge XANES spectra of different Ni samples, and h the corresponding k3-weighted Fourier-transformed EXAFS. We further investigated that the oxyanions could effectively repel the OH⁻ adsorption, which is critical for protecting the low-coordinated Ni species from irreversible oxidation and suppressing OER. Differential charge density analysis indicated that -HCO3 ligand exhibits much larger oxyanion configuration (q3O = −3.50 |e−|) than -OH (qO = −1.14 |e−|), so Ni(OH)x-HCO3 accumulates significantly more surface charge density, which could provide strong electrostatic repulsion against OH⁻ (Fig. 2d, Supplementary Figs. 5, 6). As a result, the catalyst inhibits the adsorption of OH⁻ while enhances the adsorption of organic nucleophiles (Fig. 2e). The potential-adsorption energy diagram also suggested that the -HCO3 ligands in Ni(OH)x-HCO3 effectively inhibit OH⁻ adsorption across a wider oxidative potential range (1.2–1.6 V versus RHE) than Ni(OH)x (Supplementary Fig. 7). While these DFT insights are based on computational models that may not fully capture operational conditions, they consistently indicate that Ni(OH)x-HCO3 has good potential to overcome the inherent trade-off between activity and stability in NORs. We synthesized the catalyst with Ni(OH)x-HCO3 motifs via a rapid electrochemical pre-reduction strategy using Ni bicarbonate (Ni HCs) as the precursor (Fig. 1b). During this process, the valence state of Ni was decreased, resulting in the partial removal of -HCO3 ligands from the lattice, which were subsequently partially filled by OH⁻ (or water) species from the solution. The remaining -HCO3 ligands were stabilized in the defect network of Ni HCs, named as PR-Ni HCs. Indeed, the stability of -HCO3 has been observed even under highly alkaline conditions such as 6 M KOH29–31. The pre-reduction potential and duration were optimized to synthesize PR-Ni HCs with the highest activity (Supplementary Figs. 8, 9). X-ray diffraction (XRD) pattern, Transmission electron microscopy (TEM), Raman spectroscopy and Fourier transform infrared spectroscopy (FTIR) revealed an amorphous, short-range ordered lamellar structure of Ni HCs and PR-Ni HCs (Fig. 2f and Supplementary Figs. 10–14). As expected, PR-Ni HCs retains a significant number of -HCO3 ligands after pre-reduction (Fig. 2f), which was further corroborated by C 1s and O 1s XPS spectra (Supplementary Fig. 15). Notably, the stability of the -HCO3 ligands—firmly coordinated within defect sites rather than merely physisorbed—is attributed to their origin as precursor lattice anions that undergo surface reconstruction during the pre-reduction process, ultimately becoming integral structural components. Ni K-edge X-ray absorption near-edge structure (XANES) spectra showed the absorption edge of PR-Ni HCs is shifted to lower photon energy (Fig. 2g), indicating a reduction in the valence state of Ni. Fourier-transformed EXAFS spectra of PR-Ni HCs displayed two characteristic peaks at 1.54 and 2.68 Å (Fig. 2h), attributed to Ni-O and Ni-Ni coordination, respectively. Wavelet-transformed EXAFS spectra further validate the distinct coordination environments (Supplementary Fig. 16), while quantitative structural analysis via fitting the original EXAFS data reveals that the Ni-O coordination number in PR-Ni HCs is ~4, lower than the 6 in Ni HCs (Supplementary Fig. 17 and Supplementary Table 4), confirming the presence of low-coordinated Ni species. Thus, we successfully fabricated low-coordinated Ni catalyst (PR-Ni HCs) with -HCO3 (Supplementary Fig. 18). Electrocatalytic NORs performance We evaluated the electrochemical performance of catalysts using UOR as a model reaction. As shown in Fig. 3a, PR-Ni HCs and PR-Ni(OH)2 exhibit significantly higher activities than their Ni HCs and Ni(OH)2 counterparts, implying that the formation of low-coordinated Ni species enhances catalytic performance. Among these, PR-Ni HCs stands out by delivering a current density of 500 mA cm−2 at a low potential of 1.42 V versus RHE. This performance not only surpasses all other tested samples but also remains competitive with most reported catalysts (Supplementary Fig. 19 and Supplementary Table 1). Crucially, PR-Ni HCs shows negligible degradation in cyclic tests, whereas PR-Ni(OH)2 suffers from significant performance degradation in the second cycle, accompanied by a cliff-like current drop at 1.5 V versus RHE due to competing OER (Fig. 3a). This direct comparison confirms that although low-coordinated Ni species boost activity, the presence of -HCO3 ligands is essential for maintaining stability. The long-term stability of PR-Ni HCs was further demonstrated in a continuous test, where it was operated for over 1100 h at 100 mA cm−2 without significant decay (Fig. 3b, Supplementary Figs. 20–22). The structural integrity underpinning this stability was confirmed via 13C isotopic labeling coupled with in situ differential electrochemical mass spectrometry (DEMS) analysis, which detected negligible leaching of bicarbonate ligands during the reaction (Supplementary Fig. 23). Finally, product analysis revealed that the formation of undesirable NOx− byproducts accounted for only 2.36%, demonstrating the promising potential of this system for treating urea-containing wastewater (Supplementary Fig. 24).Fig. 3Electrocatalytic performance in NORs.a LSV scan in 1 M KOH electrolyte with 0.33 M urea (catalyst-1 and catalyst-2 represent the first and second LSV profiles, respectively) at a scan rate of 5.0 mV s−1. b Long-term stability test (without iR correction). c The capability of catalysts to repel OH⁻ detected by DMSO. d In situ evaluation of O2 production using rotating ring-disk electrodes (without iR correction). Blue shaded area denotes the region where OER competition occurs significantly. e Current response under pulse voltammetry conditions. f Relationship between charge storage and potential. g Tafel slopes. h Activation energies. i Comparison of NOR activities of catalysts for different nucleophiles. j Comparison of UOR activities of Ni L, PR-Ni L and cycled PR-Ni L (after 100 h of operation) at 100 mA cm−2. Electrocatalytic performance in NORs. a LSV scan in 1 M KOH electrolyte with 0.33 M urea (catalyst-1 and catalyst-2 represent the first and second LSV profiles, respectively) at a scan rate of 5.0 mV s−1. b Long-term stability test (without iR correction). c The capability of catalysts to repel OH⁻ detected by DMSO. d In situ evaluation of O2 production using rotating ring-disk electrodes (without iR correction). Blue shaded area denotes the region where OER competition occurs significantly. e Current response under pulse voltammetry conditions. f Relationship between charge storage and potential. g Tafel slopes. h Activation energies. i Comparison of NOR activities of catalysts for different nucleophiles. j Comparison of UOR activities of Ni L, PR-Ni L and cycled PR-Ni L (after 100 h of operation) at 100 mA cm−2. Inhibiting OH⁻ adsorption is crucial for not only maintaining the stability of catalyst but also preventing OER competition4,10. We quantified the generation of •OH on the catalyst surface in electrooxidation using dimethyl sulfoxide (DMSO) as the probe molecule, where •OH reacts with DMSO to produce methanesulfonic acid (MSA)32. As shown in Fig. 3c and Supplementary Fig. 25, Ni(OH)2 and PR-Ni(OH)2 produced similar amounts of MSA (5.28 µM and 4.32 µM), indicating comparable surface hydroxylation. This suggests that PR-Ni(OH)2 is nearly fully hydroxylated and irreversibly returns to Ni(OH)2. In contrast, PR-Ni HCs produced only 0.23 µM of MSA, demonstrating that oxyanion ligands effectively suppress OH⁻ adsorption and prevent irreversible transformation into coordinatively saturated Ni species. We further evaluated whether these catalysts are accompanied by OER using ring-disk electrodes (Fig. 3d and Supplementary Fig. 26). For Ni(OH)2, the increase in ring current along with a decrease in disk current at higher potentials indicates the occurrence of OER. For PR-Ni HCs, there is no cliff-like drop in the disk current within the high potential range, and the ring current remains close to zero, indicating negligible OER. In situ DEMS further confirmed that PR-Ni HCs effectively suppresses O2 formation (Supplementary Fig. 27). To check the electrochemical deprotonation kinetics, we monitored its intrinsic oxidative charge storage capacity using pulse voltammetry in a urea-free environment33. Under different potential pulses, Ni(OH)2, PR-Ni(OH)2 and PR-Ni HCs exhibited alternating between cathodic and anodic current pulses (Fig. 3e). The oxidative charge storage capacity of different catalysts was measured by integrating the response of the cathodic current to the voltage pulse. As shown in Fig. 3f, PR-Ni HCs demonstrated a higher oxidation charge storage capacity compared with Ni(OH)2 and PR-Ni(OH)2, implying that PR-Ni HCs undergoes rapid electrochemical deprotonation. Moreover, the Ni oxidation peak (in the CV curve) of PR-Ni HCs was shifted downward compared with other catalysts, confirming that the deprotonation barrier is significantly reduced (Supplementary Fig. 28). Likewise, the Tafel slope and activation energy of PR-Ni HCs demonstrated the fastest reaction kinetics (Fig. 3g, h). PR-Ni HCs also showed high normalized activity, indicating notable intrinsic activity (Supplementary Figs. 29, 30). To demonstrate the versatility of our catalyst, we extended the electrooxidation to other organic nucleophiles such as methanol, ethanol, glycerol, and HMF. For all substrates, PR-Ni HCs exhibited improved activity and remained competitive with most reported catalysts (Fig. 3i)10,13, requiring only a potential of less than 1.5 V versus RHE to achieve a high current density of 400 mA cm−2 (Supplementary Fig. 31 and Supplementary Table 2). The Faraday efficiencies for the corresponding products—formic acid, acetic acid or FDCA—were consistently greater than 94% (Supplementary Figs. 32–36). Furthermore, to verify the generality of oxyanion stabilization strategy, we fabricated analogous catalysts embedded with other ligands, including borate, oxalate, carbonate and phosphate. These materials all show improved activity and stability for the oxidation of urea, requiring significantly lower potentials at identical current densities compared with nickel hydroxide (Fig. 3j, Supplementary Figs. 37–40). This collective evidence confirms the broad applicability of our design principle across diverse substrates and oxyanion ligands. While all tested oxyanions enhance NOR performance, the extent of this improvement varies (Supplementary Figs. 39, 40). This phenomenon may be linked with the pKa value of oxyanion ligand (Supplementary Fig. 41a). A lower pKa value generally indicates a stronger deprotonation ability of oxyanion34,35, a property crucial for the initial catalyst activation step (i.e., the formation of high-valence active species) (Supplementary Fig. 41b, c). However, an efficient NOR cycle depends not only on this initial activation but, more critically, on the subsequent chemical hydrogen transfer from the nucleophile to the catalyst. Consequently, a lower pKa does not necessarily guarantee better performance. An ideal catalyst must effectively bridge both steps: the catalyst’s electrochemical deprotonation and the chemical hydrogen transfer from the nucleophile, wherein the oxyanion ligand acts as a bifunctional bridge connecting the Ni metal center with the adsorbed substrate (or intermediate) and facilitating proton/hydrogen transfer throughout the redox cycle (Supplementary Fig. 41d). Bicarbonate, with the well-suited pKa value (6.38), strikes the optimal balance between these two competing kinetic demands, thereby exhibiting the best overall performance. A more detailed mechanistic analysis is provided in the following sections. Dynamic transition of low-coordinated Ni2-δ/Ni3-δ species We conducted in situ electrochemical characterizations with and without nucleophiles to analyze the transition of low-coordinated Ni2-δ/Ni3-δ species in electrooxidation. Initial in situ XPS analysis at open circuit potential (OCP) confirmed the presence of low-coordinated Ni2-δ species (characteristic peak at 852.74 eV) in both PR-Ni(OH)2 and PR-Ni HCs (Fig. 4a, b, Supplementary Fig. 42). Upon applying a potential of 1.5 V versus RHE, the intensity of the characteristic peak of Ni2-δ species in PR-Ni HCs decreases, accompanied by a shift of the Ni2+ peak toward higher binding energy—a phenomenon indicating the formation of high-valence Ni3-δ species. In contrast, PR-Ni(OH)2 only exhibits a reduction in its Ni2-δ signal, with negligible formation of Ni3-δ species. The modest accumulation of Ni3-δ in PR-Ni HCs is reasonable since this intermediate is rapidly consumed by urea upon its formation; nevertheless, its detectable presence indicates that PR-Ni HCs facilitate electrochemical deprotonation far more effectively than PR-Ni(OH)2. Crucially, the entire process was reversible for PR-Ni HCs. Upon cycling the potential back to the OCP, the Ni2-δ signal was fully restored (Fig. 4b). This behavior was also reflected in the O 1s and C 1s XPS spectra (Supplementary Fig. 43), which confirmed the dynamic transition of the -HCO3/-CO32- ligands. Conversely, for PR-Ni(OH)2, the number of Ni2-δ species continues decreasing during the potential variation (Fig. 4a), indicating that it undergoes irreversible oxidation and ultimately becomes coordinatively saturated Ni(OH)2.Fig. 4Dynamic transition of low-coordinated Ni2-δ/Ni3-δ species for accelerated reaction kinetics.In situ XPS spectra of Ni 2p for PR-Ni(OH)2 (a) and PR-Ni HCs (b) under open circuit, at 1.5 V versus RHE, and back to open circuit (in 1 M KOH + 0.33 M urea). Gray shaded area denotes Ni2-δ species. In situ Raman spectra of PR-Ni HCs in 1 M KOH (c) and 1 M KOH + 0.33 M urea (d), respectively. e In situ FTIR spectra of PR-Ni HCs at different potentials during UOR process. f Schematic illustration of the oxyanion-bridged hydrogen transfer and deprotonation for enhanced UOR. g Detailed free energy profiles, with each dehydrogenation including two steps: electrochemical deprotonation of catalyst (represented by △), and chemical hydrogen transfer from nucleophile to catalyst (represented by ▽). Blue and gray shaded areas denote electrochemical deprotonation and chemical hydrogen transfer processes, respectively, associated with the rate-determining step. Dynamic transition of low-coordinated Ni2-δ/Ni3-δ species for accelerated reaction kinetics. In situ XPS spectra of Ni 2p for PR-Ni(OH)2 (a) and PR-Ni HCs (b) under open circuit, at 1.5 V versus RHE, and back to open circuit (in 1 M KOH + 0.33 M urea). Gray shaded area denotes Ni2-δ species. In situ Raman spectra of PR-Ni HCs in 1 M KOH (c) and 1 M KOH + 0.33 M urea (d), respectively. e In situ FTIR spectra of PR-Ni HCs at different potentials during UOR process. f Schematic illustration of the oxyanion-bridged hydrogen transfer and deprotonation for enhanced UOR. g Detailed free energy profiles, with each dehydrogenation including two steps: electrochemical deprotonation of catalyst (represented by △), and chemical hydrogen transfer from nucleophile to catalyst (represented by ▽). Blue and gray shaded areas denote electrochemical deprotonation and chemical hydrogen transfer processes, respectively, associated with the rate-determining step. The dynamic Ni2-δ/Ni3-δ transition includes electrochemical deprotonation of catalyst to form Ni3-δ and then chemical hydrogen transfer from nucleophile to catalyst to form Ni2-δ. We conducted in situ Raman spectroscopy under different conditions to analyses this process. Ni(OH)2 exhibited the peak of Ni3+-O species within the 400–600 cm−1 range as the potential increased in a 1 M KOH solution (Supplementary Fig. 44), resulting from the electrochemical deprotonation of lattice hydroxyl groups (Ni2+-OH)36,37. However, upon the addition of urea, the Ni3+-O peak was almost not detected in the voltage range of 1.2–1.6 V versus RHE, indicating the generated Ni3+-O species were consumed by urea through chemical hydrogen transfer (Supplementary Fig. 45). Although PR-Ni(OH)2 could accelerate the deprotonation, it still required a high potential of 1.5 V versus RHE to detect the Ni3-δ-O species (Supplementary Figs. 46, 47). In contrast, for PR-Ni HCs, the signals of Ni3-δ-O species remained clearly detectable even at very low potential 1.4 V versus RHE in the presence of urea (Fig. 4c, d), indicating that the catalyst displays faster deprotonation kinetics than PR-Ni(OH)2 and Ni(OH)2. Notably, for the PR-Ni HCs, the Raman peak of Ni3-δ-O species disappeared immediately after switching the potential from 1.4 V to OCP versus RHE in the presence of urea (Fig. 4d), suggesting that the chemical hydrogen transfer from nucleophilic urea to Ni3-δ-O species occurred at a rapid rate. The same procedure was repeated by changing the potential to 1.5 V versus RHE, yielding similar results. We also performed intermittent UOR electrochemical tests to validate the above processes. Compared with other catalysts, PR-Ni HCs exhibited both a faster catalyst deprotonation under the same conditions (as described earlier) and a rapid chemical hydrogen transfer from nucleophiles/intermediates to the catalyst (Supplementary Fig. 48). Oxyanion-mediated rapid deprotonation and hydrogen transfer An effective catalyst for NORs must not only accelerate the initial electrochemical deprotonation but also balance it with the subsequent chemical hydrogen transfer from nucleophiles10,13. Through a combination of in situ experiments and DFT calculations, we elucidated the key role of oxyanion ligands in synergizing these two steps (Fig. 4e, f). First, we used in situ FTIR spectra to identify the key reaction intermediates in UOR. For PR-Ni HCs, the absorption bands at ~1109 cm−1, 1430 cm−1 and 1691 cm−1 are assigned to the stretching vibrations of N-H, C-N and C=O bonds in urea (Fig. 4e)4,38. The intensities of these bands increased with the applied potential, suggesting that higher potentials enhance the catalyst’s ability to adsorb more urea molecules. The vibration peak at 1641 cm−1 was assigned to the N-H vibration in the amide groups of *CO(NH)N intermediate39, while the vibration peaks at 1346 cm−1 and 1272 cm−1 corresponded to the C-O stretching vibrations in -HCO3 and -CO3, respectively4,40. This provides key intermediate in the reaction pathway and indicates that oxyanion ligands (-HCO3/-CO3) remain stable with the applied potential. Subsequently, we used isotope-labeled in situ DEMS measurements to monitor the mode of N2 formation during UOR (Supplementary Fig. 49). The results indicate that N2 is mainly formed by two electron-deficient nitrogen atoms in the same urea molecule, demonstrating the main intramolecular N-N coupling mechanism occurring in UOR41, and so the UOR pathway proceeds as follows: *CO(NH2)2 → *CO(NH2)NH → *CO(NH2)N → *CO(NH)N → *CONN → *CO + N2 → *COOH → *CO2 (Supplementary Fig. 50)4,6. Among these, we screened out the optimal adsorption configuration of urea as O-terminal adsorption (Supplementary Fig. 51). The rate-determining step is the dehydrogenation of *CO(NH2)N to *CO(NH)N, which involves the electrochemical deprotonation of catalyst (blue shaded area in Fig. 4g) and the chemical hydrogen transfer (gray shaded area in Fig. 4g) from *CO(NH2)N to catalyst, ultimately forming *CO(NH)N. For Ni(OH)2, PR-Ni(OH)2 and PR-Ni HCs, the electrochemical deprotonation energy barriers are 2.57, 1.59 and 0.87 eV, respectively, indicating that PR-Ni HCs with bicarbonate ligands are more prone to electrochemical deprotonation (Ni2-δ(OH)x-HCO3 + OH⁻ → Ni3-δ(OH)x-CO3 + H2O + e⁻). Since the chemical hydrogen transfer from *CO(NH2)N to catalyst is thermodynamically spontaneous, we further calculated the transition state to quantify the kinetics of this chemical step, and the energy barriers are 0.32, 0.84, and 0.46 eV for these three catalysts. Compared with Ni(OH)2, the electrochemical deprotonation in PR-Ni(OH)2 is faster, but the chemical hydrogen transfer is slower with increased energy barrier. This reflects the inverse relationship between the rates of electrochemical deprotonation and chemical hydrogen transfer at the same ligand oxygen center. Remarkably, PR-Ni HCs circumvents this kinetic trade-off. Unlike in the -OH mediated system, its accelerated electrochemical deprotonation is not penalized by a slower chemical hydrogen transfer (barrier = 0.46 eV). This capability stems from the synergistic bifunctionality of the bicarbonate/carbonate redox pair. Specifically, bicarbonate acts as an efficient proton shuttle to facilitate electrochemical deprotonation, while the subsequently generated carbonate serves as a thermodynamically favorable hydrogen acceptor for the chemical transfer step (Supplementary Fig. 41d). This dynamic interplay establishes a microenvironment that breaks the scaling relationship inherent to single-component (-OH) systems, thereby accelerating the overall reaction kinetics (Supplementary Note 1, Supplementary Figs. 40, 41). Low-energy NORs-coupled hydrogen production Coupling NORs in the anode and hydrogen generation in the cathode can enhance the economic efficiency and sustainability of electrosynthesis/electrolysis (Fig. 5a)13,42. To demonstrate the potential of PR-Ni HCs (Supplementary Fig. 52a), we assembled an electrolyzer without membrane and with anion exchange membrane (AEM), respectively, using urea oxidation coupled with hydrogen evolution reaction (HER). For the UOR//HER electrolyzer without membrane, an industrial current density of 2.0 A cm−2 was achieved at a cell voltage of just 1.79 V at 70 °C (Supplementary Fig. 52b, c and Supplementary Table 5). Similarly, it demonstrated good activity in AEM electrolyzer (Fig. 5b), requiring only 1.61 V and 1.99 V to reach the industrial current densities of 400 mA cm−2 and 1 A cm−2, respectively (Supplementary Fig. 53). We further enlarged the electrode area from 1 cm2 to a pilot scale of 100 cm2 (Supplementary Fig. 54), and achieved a current of 55 A at cell voltage of 2.17 V (Supplementary Fig. 55). Besides, the electrolyzer delivered high energy efficiencies of 91.9% at 0.4 A and 74.0% at 55 A (Fig. 5c). This system’s calculated hydrogen price per gasoline gallon equivalents (GGE) was as low as $0.86 at 0.4 A and $1.06 at 55 A, both well below the U.S. Department of Energy’s 2026 target of $2.00 per GGE and close to the 2031 target of $1.00 per GGE43,44. Moreover, the system operated stably for over 700 h at an industrial current density of 1 A cm−2 with negligible performance degradation (Fig. 5d).Fig. 5Low-energy NORs-coupled hydrogen production system.a Application prospects of NORs. b The setup for paired UOR(+)//HER(−). c Comparison of energy consumption and efficiency of AEM electrolyzers with OER//HER versus UOR//HER. d Stability of UOR//HER electrolyzer at 1 A cm−2 (area: 1 cm2). Low-energy NORs-coupled hydrogen production system. a Application prospects of NORs. b The setup for paired UOR(+)//HER(−). c Comparison of energy consumption and efficiency of AEM electrolyzers with OER//HER versus UOR//HER. d Stability of UOR//HER electrolyzer at 1 A cm−2 (area: 1 cm2). Discussion In summary, we presented oxyanion ligands can dynamically stabilize low-coordinated Ni species and achieve both high activity and long-term stability for NORs. The oxyanion ligands protect low-coordinated nickel species and suppress OER by increasing the surface charge density to electrostatically repel hydroxide ion adsorption. Moreover, these ligands promote the dynamic transition between Ni2-δ and Ni3-δ by bridging the catalyst’s electrochemical deprotonation with chemical hydrogen transfer from the nucleophile, thereby accelerating the rate-determining step in NORs. The representative low-coordinated Ni-based catalyst embedded with bicarbonate ligands exhibits high activity and stability in the electrooxidation of classical nucleophiles such as urea, methanol, ethanol, glycerol, and 5-hydroxymethylfurfural. Furthermore, the catalyst shows good potential for application as demonstrated in a 100 cm2 AEM NORs-coupled hydrogen production system. Methods Chemicals and materials Nickel chloride hexahydrate (NiCl2·6H2O, 99.99%), potassium carbonate (K2CO3, 99.99%), potassium dihydrogen phosphate (KH2PO4, 99.0%), and potassium hydroxide (KOH, 99.99%) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. Potassium bicarbonate (KHCO3, 99.99%) and potassium oxalate (K2C2O4, 99.7%) were purchased from Meryer (Shanghai) Biochemical Technology Co., Ltd. Sodium tetraborate pentahydrate (Na2B4O7·5H2O, 99.99%) was purchased from Shanghai Macklin Biochemical Technology Co., Ltd. Carbon black (VXC-72R) was purchased from Johnson Matthey Catalyst Co., Ltd. Ethanol (C2H5OH, 99.5%), urea (CO(NH2)2, 99.5%), 5-hydroxymethylfurfural (C6H6O3, 99%), and glycerol (C3H8O3, 99.5%) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd.; methanol (CH3OH, 99.9%) was purchased from Sigma-Aldrich Co., Ltd. All chemicals were used as received without further purification. Deionized water (18.2 MΩ·cm) was used in all experiments. Additionally, bipolar plates (BPPs, titanium) were purchased from Shanghai Chuxi Industrial Co., Ltd., anion exchange membranes including FAA-3-50 (thickness 50 μm) and FAA-3-PK-130 (thickness 130 μm) were purchased from FuMa Tech, Germany, and nickel foam (purity 99.9%) was purchased from Jiayisheng Electronics Co., Ltd. The waterproof and breathable membrane (PTFE material, pore size ≤20 nm, porosity ≥50%) was purchased from Linglu Instruments (Shanghai) Co., Ltd. Synthesis of Ni-based catalysts All Ni L samples (where L = bicarbonate, phosphate, borate, carbonate, oxalate, hydroxide) were synthesized via a modified chemical precipitation method. The general procedure is as follows: 100 mg of VXC-72R carbon black (used as a carrier to prevent catalyst aggregation) was dispersed in 20 mL of an aqueous solution containing 237.7 mg of NiCl2·6H2O, and the mixture was stirred for 30 min to ensure complete dispersion. Subsequently, 10 mL of a specific precipitant was added, and the mixture was further stirred for 30 min to achieve thorough mixing. The resulting mixture was centrifuged, and the solid precipitate was washed with ultrapure water until neutral pH was reached. The solid was then dried in a vacuum oven at 60 °C for 12 h, finely ground, and stored. The precipitants used for different L species are specified as follows: 0.2 M KHCO3 for nickel bicarbonate (Ni HCs); 0.1 M KOH (with the initial solution containing 272.2 mg of KH2PO4) for nickel phosphate; 0.2 M Na2B4O7·5H2O for nickel borate; 0.1 M K2CO3 for nickel carbonate; 0.2 M K2C2O4 for nickel oxalate; and 0.2 M KOH for nickel hydroxide. PR-Ni L catalysts were prepared using a universal and rapid electrochemical pre-reduction technique. The electrolytic system consisted of a working electrode (Ni L catalyst loaded on carbon paper or nickel foam), a Hg/HgO reference electrode, a carbon rod counter electrode, and 1 M KOH as the electrolyte. For PR-Ni HCs, the pre-reduction conditions were optimized by varying the potential (−0.3, −0.6, −0.9, −1.2, −1.5 V versus RHE) and duration (1, 5, 10, 20, 30, 60 min). For other PR-Ni L samples, to form low-coordinated metal species, a reduction potential of −0.9 V (versus RHE) was applied for 5 min. After electro-reduction, PR-Ni L supported on carbon paper or nickel foam was obtained. Composition and structure characterization X-ray diffraction (XRD) patterns were recorded on a Bruker D8 Focus equipped with Ni-filtered Cu Kα radiation (λ = 1.54056 Å) at a scan rate of 5° min−1. X-ray photoelectron spectroscopy (XPS) was employed to examine the valence states of various elements on the catalyst surface using Al Kα radiation (ThermoFisher Scientific K-Alpha+), and the C 1s peak of contaminated carbon (284.8 eV) was used to calibrate the binding energy. Scanning electron microscopy (SEM) images were obtained using field emission scanning electron microscope (Hitachi, Regulus 8100). Transmission electron microscopy (TEM) images were obtained using a JEM-F200 at 200 kV. An energy-dispersive X-ray spectroscopy (EDAX) analyzer attached to the TEM was used to analyze the elemental distribution. Fourier transform infrared (FTIR) spectroscopy was recorded using a Bruker INVENIO FT-IR spectrometer. The X-ray absorption spectra (XAS), including X-ray absorption near-edge structure (XANES) and extended X-ray absorption fine structure (EXAFS), of the samples at the Ni K-edge were collected at the Singapore Synchrotron Light Source (SSLS) center, where a pair of channel-cut Si (111) crystals was used in the monochromator. Electrochemical measurements All electrochemical NOR activity and stability measurements were performed on the IVIUMSTAT workstation (Ivium Technologies, Netherlands). For a better comparative study, the samples were loaded onto glassy carbon electrodes (0.196 cm2, ~0.25 mgcatalyst cm−2), carbon paper (0.5 × 0.5 cm2, ~0.5 mgcatalyst cm−2), and Ni foam (1 × 0.5 cm2, ~5 mgcatalyst cm−2) as working electrodes, respectively. For example, when focusing on the competition mechanism between UOR and OER, glassy carbon electrodes and the rotating ring-disk electrode (RRDE) were used for testing. A Hg/HgO electrode and a carbon rod were used as the reference and counter electrodes, respectively. Unless otherwise specified, all potentials were converted to the reversible hydrogen electrode (RHE, ERHE = EHg/HgO + 0.059 × pH + 0.098, which was corrected experimentally by using Pt/C as the working electrode in a H2-saturated solution), aiming to unify the benchmark for comparing electrochemical performances. Typically, 5 mg of catalyst was suspended in a mixed solution of 1 mL of isopropanol, water, and Nafion, and subjected to intense ultrasound for 3 h to form a uniform ink. The electrolyte for the UOR was a freshly prepared 1 M KOH solution containing 0.33 M urea (pH 13.8 ± 0.1). For the extension to other organic substrates, such as methanol, ethanol, glycerol, and HMF, the concentration of each was 0.1 M. The reactions were tested separately in an H-type electrochemical cell. Linear sweep voltammetry (LSV) curves were collected with the scan rate of 5 mV s−1. Unless otherwise specified, all data were reported with 90% iR compensation. Electrochemical impedance spectroscopy (EIS) measurements were performed over a frequency window spanning 100 kHz to 0.1 Hz. Double-layer capacitance (Cdl) was measured and electrochemically active surface area (ECSA, with ECSA = Cdl/Cs) was estimated via cyclic voltammetry (CV) curves in the potential range of 0.94–1.04 V versus RHE, using a standard specific capacitance (Cs) of 40 μF cm−2 and at scan rates of 20, 40, 60, 80, and 100 mV s−1. Additionally, in order to ensure sufficient accuracy at high current densities, the AEM electrolyzers with electrode areas of 1 cm2 (1 cm × 1 cm) and 100 cm2 (10 cm × 10 cm) were tested on the CORRTEST (CS350M) and LANHE (CT3001D) workstations, respectively. The detailed operating procedures for this part are described in the AEM electrolyzer testing section. Electrochemical in situ ATR-FTIR measurements The in situ ATR-FTIR spectrum was collected using a Bruker INVENIO FT-IR spectrometer equipped with an MCT-A detector cooled by liquid nitrogen. The catalyst was uniformly coated on the surface of a single-bounce silicon crystal (with Au thin film) and then assembled in a spectroelectrochemical cell with a carbon rod as the counter electrode and a Hg/HgO electrode as the reference electrode (Supplementary Fig. 56a)4. The Au thin film is used to enhance the infrared spectral signal and also serves as a protective layer to prevent the dissolution or corrosion of silicon crystal materials45. 1 M KOH and 0.33 M urea were used as the electrolyte. All spectra were obtained at a spectral resolution of 4 cm–1, and each single-beam spectrum was an average of 100 scans. Background spectra were collected under the condition of after adding the electrolyte and without applying a working voltage (i.e., in the absence of electrochemical reactions) to subtract background signals unrelated to the reaction (such as static species and free molecules in the electrolyte, etc.), thereby specifically highlighting the signals of dynamically generated intermediates during the reaction. Notably, spectral collection focused on wavenumbers beyond 1000 cm−1, as the 400–1000 cm−1 low-wavenumber range was excluded: strong intrinsic lattice vibrations of the silicon crystal, superimposed with interference from the catalyst support and electrolyte, caused severe peak overlapping (signal-to-noise ratio <1) and precluded effective spectral interpretation. Potential control was carried out using an IVIUMSTAT workstation. Electrochemical in situ XPS measurements In situ XPS measurements were performed using a custom-built electrochemical cell, and data were collected with an X-ray photoelectron spectrometer (Thermo ESCALAB 250Xi) (Supplementary Fig. 56b)46. The homemade X-ray cell consists of three parts: a reaction chamber, an analysis chamber, and a preparation chamber. Before testing, the proper electrical contact between the potentiostat and catalyst should be confirmed by measuring the resistance between the working and reference electrodes. The catalyst was first placed into a reaction chamber containing a 1 M KOH and 0.33 M urea electrolyte (which had been deaerated with argon). The reactor was then vacuumized, with the pressure reduced to <10−8 Torr, and transferred to the analysis chamber once the vacuum standard was achieved. At each set potential, the electrodes were first polarized for 5 min until stable, and XPS signals were collected at various applied potentials (ranging from OCP to 1.5 V and back to OCP, versus RHE). Electrochemical in situ Raman measurements In situ Raman spectroscopy was performed using a LabRAM HR Evolution, with a 532 nm Nd:YAG laser line as the excitation source. The in situ electrochemical cell consisted of the prepared electrocatalyst coated onto carbon paper as the working electrode, a Hg/HgO reference electrode, and a Pt wire as the counter electrode (Supplementary Fig. 56c). The electrolyte was 1 M KOH, with or without 0.33 M urea. The measured potential for the reaction ranged from 1.2 to 1.6 V (versus RHE) and was controlled using an IVIUMSTAT workstation. DEMS measurements The DEMS system (QAS 100, Linglu Instruments (Shanghai) Co., Ltd.) was employed. The in situ electrochemical cell consisted of the prepared electrocatalyst as the working electrode, Hg/HgO as the reference electrode, and Pt wire as the counter electrode (Supplementary Fig. 56d). The electrolyte consisted of 1 M KOH and 0.33 M urea. The potential was controlled using an IVIUMSTAT workstation. The gaseous products were pumped into a mass spectrometer, and signals with mass-to-charge ratios of 17, 28, 29, 30, 32, 44 and 45 were detected. During isotope labeling experiments, the natural abundance proportion of these isotopes is extremely low, such that their impact on product detection and reaction analysis is negligible (Supplementary Fig. 57). AEM electrolyzer performance measurements The AEM electrolyzer system consisted of a membrane, bipolar plates (BPPs), current collectors, gas diffusion layers (GDLs), and catalyst layers (CLs). The BPPs, made of titanium, also served as current collectors and had serpentine flow channels at their center for electrolyte transport. The anion exchange membrane was activated by separately soaking it in 0.5 M NaCl and 1 M KOH solutions for 24 h, respectively. Additionally, the anode catalyst was PR-Ni HCs, while the cathode catalyst was commercial Pt/C. The catalyst slurry comprises 50 mg of catalyst, 7.5 mL of isopropanol, 2 mL of ultrapure water, and 0.5 mL of perfluorosulfonic acid resin (Nafion), which is homogenized via intense ultrasonic treatment. These catalysts were uniformly sprayed onto nickel foam substrates (dimensions: 100 × 100 × 0.2 mm, corresponding to an area of 100 cm2), forming integrated working electrodes (CL/GDL) that served as both the CL and GDL to ensure zero-gap contact. Furthermore, the AEM electrolyzer system operated with an electrolyte flow rate of 200 mL min−1, where the anode was simultaneously supplied with KOH and urea, while the cathode was supplied only with KOH. The electrode area for the stability test was 1 cm2. The thicker and more durable FAA-3-PK-130 was used as the anion exchange membrane in the stability test, but its performance was inferior to that of FAA-3-50. The energy conversion, efficiency, and economic evaluations (hydrogen price in gasoline gallon equivalents, GGE) were calculated using the following methods47.1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{{\rm{Electrolyzer\; power}}}}={{{\rm{V}}}}\times {{{\rm{I}}}}$$\end{document}Electrolyzer power=V×I2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{{{\rm{H}}}}}_{2}\,{{{\rm{production\; rate}}}}={{{\rm{I}}}}/(1.602\times {10}^{-19}\ \mathrm C\times 2)$$\end{document}H2production rate=I/(1.602×10−19C×2)3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{{\rm{Energy\; consumption}}}}\ ({{{\rm{kW}}}}{{{\rm{h}}}}/{{{{\rm{Nm}}}}}^{3}\ {{{{\rm{H}}}}}_{2})=2\times {{{\rm{F}}}}\times {{{\rm{V}}}}/{{{{\rm{V}}}}}_{{{{\rm{H}}}}2}/3600/1000$$\end{document}Energy consumption(kWh/Nm3H2)=2×F×V/VH2/3600/10004\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{{\rm{Energy}}}}\; {{{\rm{efficiency}}}} \ (\%)={{{\rm{V}}}}_{{{\rm{neutral}}}} \times {{{\rm{T}}}}{{{\rm{h}}}} {{{\rm{eoretical}}}}\; {{{\rm{electricity}}}}\\ /{{{\rm{Energy\; consumption}}}}/1000$$\end{document}Energyefficiency(%)=Vneutral×Theoreticalelectricity/Energy consumption/10005\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{{\rm{Electricity\; cost}}}}\,({{{{\rm{H}}}}}_{2}/{{{\rm{Kg}}}}) =1\ {{{\rm{GGE}}}}\ {{{{\rm{H}}}}}_{2}/{{{{\rm{H}}}}}_{2}\ {{{\rm{production\; rate}}}} \\ \times {{{\rm{Electrolyzer\; power}}}}\times {{{\rm{Electricity\; bill}}}}$$\end{document}Electricity cost(H2/Kg)=1GGEH2/H2production rate×Electrolyzer power×Electricity bill In these formulas, I denotes the applied current in amperes (A); C is the unit of electric charge (coulomb); and F represents the Faraday constant (96,500 C). V stands for the measured voltage in volts (V); VH2 denotes the volume of hydrogen under standard conditions. Vneutral is 1.48 V, with a theoretical electricity consumption of 2390 A h/m3. The electricity price is set at 0.02 USD/kWh, and 1 GGE of hydrogen equals 0.997 kg. These formulas are used to analyze the hydrogen production performance of UOR and OER under different current conditions (e.g., 0.4 A and 55 A). Product analyses The electrolytic products were quantitatively analyzed using high-performance liquid chromatography (HPLC, LC-2050, Shimadzu, Japan) and ion chromatography (IC, Essentia IC-16, Shimadzu, Japan). The HPLC system was equipped with UV (LC-2050 UV) and RID (RID-20A) detectors, while the IC system was equipped with an anion suppressor (SHY-A-6) and a cation suppressor (CS-1000). Specifically, urea was quantified by HPLC with a mobile phase of acetonitrile-water (20:80, v/v) at a flow rate of 1.0 mL min−1. The HPLC column used was ShimNex CS C18 (4.6 × 150 mm), and detection was carried out at 195 nm. Nitrogen-containing products were analyzed by IC. For anion analysis, a Dionex IonPacTM AS20 column (4 × 250 mm) was used with a mobile phase of 35 mM KOH at a flow rate of 1.0 mL min−1. For cation analysis, a Dionex IonPacTM CS12A column (4 × 250 mm) was employed with a mobile phase of 20 mM methanesulfonic acid at a flow rate of 1.0 mL min−1. Glycerol, HMF, FDCA, formic acid, acetic acid, and other products were quantified by HPLC using an Aminex HPX-87H column (300 × 7.8 mm), with a mobile phase of 5 mM H2SO4 at a flow rate of 0.5 mL min−1. FE for the formation of the product was computed from: FE = (eF × n)/Q = (eF × n)/(I × t). In this equation, e represents the number of electrons transferred for each product, F is the Faraday constant, Q is the charge, I is the applied current, t is the reaction time, and n is the total amount of product (in moles). All measurements for each product were repeated at least three times at each applied potential to minimize any potential errors. Computational methods All DFT calculations were performed using the Vienna Ab initio Simulation Package (VASP), employing spin-polarized density functional theory with the Hubbard U correction (DFT + U), and the Perdew–Burke–Ernzerhof (PBE) exchange-correlation functional along with the projector augmented wave (PAW) pseudopotential method to describe interactions between the core and valence electrons. For the PAW pseudopotential, the following valence electron configurations were included: 3d 84s2 for Ni, 2s22p2 for C, 2s22p4 for O, and 1s1 for H. A U-J value of 5.5 eV was used for Ni to describe electron correlation in 3d metals. In addition, the Grimme’s correction method (DFT-D3) was used for describing the long-range dispersion interaction. All atomic positions were fully relaxed until the energy and force reached tolerances of 1 × 10–5 eV and 0.02 eV Å−1, respectively. The periodic slab model was optimized using the Monkhorst-Pack method with a k-point mesh of 2 × 3 × 1 for electronic structure calculations and a cutoff energy of 500 eV. All slabs were separated by at least 15 Å of vacuum space to minimize spurious interactions between atomic layers. The Gibbs free energy of the reaction was calculated using the following equation.6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\Delta {{{\rm{G}}}}}_{{{ads}}}={\Delta {{{\rm{E}}}}}_{{{{\rm{ads}}}}}+{\Delta {{{\rm{E}}}}}_{{{ZPE}}}\,-\,{{{{\rm{T}}}}\Delta {{{\rm{S}}}}}_{{{{\rm{ads}}}}}$$\end{document}ΔGads=ΔEads+ΔEZPE−TΔSads Where ΔEads is the electronic adsorption energy, ∆EZPE and T∆Sads are the change in zero-point energy of the adsorbates and the corresponding entropy difference. The zero-point energy and entropy were obtained by calculating the vibration frequency of adsorbed molecules. Supplementary information Supplementary Information Description of Additional Supplementary Files Supplementary Data 1 Transparent Peer review file Supplementary Information Description of Additional Supplementary Files Supplementary Data 1 Transparent Peer review file Source data Source data Source data