Efficient electrocatalytic reduction of CO₂ to CO over Ni/Y diatomic catalysts

1 Introduction

The utilization of electrocatalytic CO₂ reduction technology, powered by sustainable energy resources, stands as a promising approach for achieving CO₂ emission reduction objectives, owing to its favorable reaction parameters and the ability to modulate reaction outputs effectively, as documented in the literature [1–4]. In recent years, researchers have dedicated significant efforts to developing efficient electrocatalysts for CO₂ reduction. However, the majority of advanced electrocatalysts face challenges such as high overpotential and low selectivity, primarily due to the elevated C═O bonding energy of the CO₂ molecule and its stable molecular structure [5]. Consequently, the imperative task of designing and synthesizing catalysts with enhanced selectivity and reduced overpotential has become a pivotal focus in the industrial application of electrochemical CO₂ reduction.

Among the electrocatalysts utilized for CO₂ reduction to date, monoatomic catalysts have captured the attention of researchers due to their exceptional atom utilization efficiency and impressive catalytic activity and stability. However, the overall catalytic performance of monoatomic catalysts is constrained by the presence of a single active site. Therefore, researchers aspire to design diatomic catalysts (DACs) with the integration of additional metal atoms to enhance the catalytic efficiency of the catalysts. The catalytic performance of the catalysts can be improved by the introduction of an additional metal atom [6–9]. DACs encapsulate the full array of benefits exhibited by monoatomic catalysts, effectively mitigating the limitation of a solitary catalytic active site characteristic of single-atom catalysts. These catalysts demonstrate elevated catalytic activity and metal loading levels in comparison to monoatomic catalysts [10,11]. The incorporation of an extra metal atom in DACs is anticipated to amplify the catalytic activity surpassing that of a singular metal [12], thereby engendering an unparalleled catalytic impact. The presence of an additional metal atom in DACs enables the handling of complex reactions involving multiple intermediates, thereby expanding the repertoire of catalytic reaction modalities. DACs are extensively employed in a myriad of applications, including the hydrogen evolution reaction [13–15], the oxygen evolution reaction [16–18], the oxygen reduction reaction [19–23], the carbon dioxide reduction reaction (CO₂RR) [7,24,25], and the nitrogen reduction reaction [26–30].

The interaction between transition metal 3d orbitals and the surrounding molecular environment hinders the desorption of CO from the active site, thereby leading to a substantial decrease in their performance for CO₂RRs. Modulating the electron density localized within the 3d orbitals of transition metals is envisaged to optimize material selectivity and facilitate the attainment of high current densities. Rare-earth elements exhibit a rich abundance of electronic orbitals and possess considerable ionic radius, enabling their effective incorporation onto a variety of carriers. Furthermore, the notable contraction effect associated with rare earth elements has the capacity to induce adjustments in the electron density of neighboring atoms, thereby optimizing the electronic configuration of the carriers. This circumstance is conducive to bolstering the adsorption capacity on the catalyst surface and elevating the catalytic performance [31–33]. The incorporation of rare-earth elements in the development of transition metal and metal DACs has demonstrated the capability to yield increased current density and heightened selectivity. As far as we are aware, there exists limited documentation on the application of rare-earth DACs in the realm of CO₂ reduction.

In this study, we introduce a diatomic porous carbon Ni–Y electrocatalyst fabricated through the method of organic ligand synthesis. Subsequent characterizations provide compelling evidence for the monoatomic nature of Ni and Y in the catalyst. The electron density of the nickel element is effectively regulated by the incorporation of the rare-earth element Y, thereby optimizing its localization within the porous carbon host and enhancing the electrocatalytic performance of the composite material. The electrochemical performance assessment conclusively demonstrated that the Ni–Y bimetallic catalyst, incorporating the rare earth element Y, achieved an exceptional Faraday efficiency of 89% at −0.93 V vs reversible hydrogen electrode (RHE) potential, showcasing superior performance compared to the mono-metallic catalyst counterpart. The achieved current density peaked at 12 mA·cm⁻², with a prolonged 12 h electrolysis test demonstrating consistent current density levels and retention of over 80% Faraday efficiency, thereby indicating remarkable operational stability.

2 Results and discussion

Analysis of the complete spectral data (Figure S1a) reveals the detectable presence of the five elements C, N, O, Ni, and Y, suggesting the successful integration of Ni and Y elements into the molecular framework of 2,2-bipyridine-5,5 dicarboxylic acid. In Figure S1b, the C 1s spectrum of the synthesized complex reveals three distinctive peaks, with binding energies of 284.4, 285.5, and 287.8 eV. The peak at 284.4 eV corresponds to the presence of C–C and C═C bonds within the molecular structure of 2,2-bipyridine-5,5 dicarboxylic acid. The peak at 285.5 eV signifies the presence of C–N bonds, while the peak at 287.8 eV indicates the presence of C═O bonds in the molecule. Figure S1c illustrates the high-resolution N 1s profile of the organic complex, displaying two distinct peaks positioned at 398.9 and 399.8 eV. These peaks are assigned to the pyridine N group within the molecular framework of the Ni–N and 2,2-bipyridine-5,5 dicarboxylic acid, respectively [34]. The high-resolution O 1s spectrum of the organic complex depicted in Figure S1d reveals two distinct peaks at 530.5 and 531.3 eV, corresponding to the C–O and O═C–O bonds, respectively. Furthermore, the Ni 2p high-resolution spectrum presented in Figure S1e displays two significant peaks at 855.9 and 873.6 eV, assigned to the Ni 2p_1/2 and Ni 2p_3/2 binding energies, respectively. The characterization results suggest that the Ni component in the material is predominantly in a +2 valence state [35,36]. Moreover, the presence of two peaks at 862.2 and 880.0 eV is interpreted as arising from the satellite structures of Ni 2p orbitals. The high-resolution Y 3d spectra illustrated in Figure S1f displayed distinct peaks at 157.9 and 159.8 eV corresponding to the 3d_3/2 and 3d_5/2 orbital states, respectively. These findings indicate that the Y element is in a +3 valence state within the synthesized complexes.

For the purpose of investigating the morphology of the synthesized bimetallic complexes, we employed scanning electron microscopy (SEM) in conjunction with energy-dispersive X-ray spectroscopy (EDS) to conduct comprehensive analysis. As depicted in Figure S2, the Ni–Y bimetallic complex was observed to possess a needle-and-tube structure on a large scale. Upon further magnification, the morphology of the Ni–Y bimetallic complex was confirmed. Subsequent analysis using SEM coupled with EDS (SEM–EDS) and mapping techniques revealed the presence of five elements within the complex, namely carbon (C), nitrogen (N), oxygen (O), nickel (Ni), and yttrium (Y). Remarkably, the uniform distribution of the nickel and yttrium elements in the complex observed in the SEM–EDS and mapping data corroborated the results derived from X-ray photoelectron spectroscopy (XPS) characterization. The conformity between the findings and the XPS characterization results points to the successful integration of nickel (Ni) and yttrium (Y) into the molecular framework of 2,2′-bipyridine-5,5′-dicarboxylic acid. Detailed elemental quantification presented in Table S1 indicates a higher concentration of yttrium compared to nickel, reflecting the initial raw material proportions utilized in the experimental synthesis.

Figure 1a illustrates the X-ray diffractograms of the Ni–Y-PC-X catalysts. From the XRD patterns in Figure 1a, it is evident that both Ni–Y-PC-X-750°C and–Ni–Y-PC-X-850°C samples exhibit diffraction peaks exclusively at 2θ = 13° and 26°. The presence of diffraction peaks exclusively at 2θ = 13° and their subsequent repetition in the XRD patterns is noteworthy in both conditions. The appearance of the peak at 13° is indicative of the generation of amorphous carbon as a consequence of the pyrolysis process. In contrast, the broader diffraction peak at 2θ = 26° corresponds to the (002) crystalline plane of graphitic carbon, suggesting the initiation of a partially graphitic structure during pyrolysis. The wider diffraction peaks imply a modest level of graphitization obtained in the material. It is of significance to highlight that, in the absence of cluster peaks attributed to the Ni and Y elements beyond the aforementioned diffraction peaks, there is an indication that the Ni and Y elements could potentially exist in a monoatomic state within the substrates, as posited in prior investigations [37]. Furthermore, in the Ni–Y-PC-X-950°C specimen, discernible diffraction peaks emerged at approximately 2θ = 44.5° and 52° in the XRD pattern, indicative of the presence of Ni species. These peaks align with the crystallographic planes of Ni (111) and (200), implying the likely formation of Ni clusters as a consequence of high-temperature-induced migration processes. The prepared Ni–Y-PC-X catalysts were then characterized for their morphology and structure. Figure 1b and Figure S3a, b present the scanning electron micrographs depicting the surface characteristics of the three catalysts. Analysis of these images reveals that the bimetallic catalysts synthesized through calcination at varying temperatures demonstrate a consistent three-dimensional honeycomb surface morphology akin to that of the PC carrier (Figure S4b). This result suggests that the three catalysts have maintained the surface morphology of the carriers well, as well as that the calcination temperature does not have much effect on the surface morphology of the synthesized catalysts. For a comprehensive molecular structure analysis of Ni–Y-PC-X, Raman spectroscopy examinations were employed to investigate the sample. The results, as depicted in Figure 1e, clearly exhibit the presence of characteristic bands such as the D-band at approximately 1,350 cm⁻¹ and the G-band at around 1,580 cm⁻¹. These bands are indicative of the in-plane defects in the carbon material attributed to in-plane lattice imperfections and the in-plane vibrational modes arising from the sp² hybridization of carbon atoms, respectively. The ratio of the intensities of the D-band to the G-band (I_D/I_G) is commonly utilized to assess the extent of graphitization in carbon materials [38–40]. The calculated I_D/I_G ratios for Ni–Y-PC-750°C, Ni–Y-PC-850°C, and Ni–Y-PC-950°C were found to be 0.9, 0.94, and 0.96, respectively, suggesting a low level of graphitization resulting from the pyrolysis process, consistent with the outcomes of the XRD analysis.

Figure 1

Structural characterization of Ni–Y-PC-850°C: (a) XRD figure, (b) SEM image at 850°C pyrolysis temperature, (c) TEM at 850°C pyrolysis temperature, (d) High-resolution TEM at 850°C, (e) Raman profiles of Ni–Y-PC catalysts at different pyrolysis temperatures, (f)–(i) TEM of pyrolysis at 850°C, spherical aberration-corrected high-angle annular dark-field scanning TEM diagrams.

To investigate the presence of Ni and Y elements in the carriers more comprehensively, transmission electron microscopy (TEM) examinations were performed on the synthesized Ni–Y-PC-X materials, with the results presented in Figure S3c, d and Figure 1f–i. The synthesized Ni–Y-PC-X samples displayed a typical flocculent morphology associated with 2D materials, along with the observation of tubular structures under high-resolution TEM (Figure 1d), which suggests the presence of a bipartite material. These tubular features are likely a result of the formation of a partially graphitized structure during the pyrolysis process of the bimetallic complexes, consistent with the results obtained from Raman and XRD analyses. In addition, scanning TEM investigations did not detect nanoclusters on the surface of the carrier material. In order to bolster the comprehensive nature of the TEM analysis, we present a set of multi-angle TEM images captured at the pyrolysis temperature of 850°C, alongside spherical aberration-corrected high-angle annular dark-field scanning TEM images and the corresponding mapping spectra. As illustrated, the absence of Ni clusters is observed, while the mapping spectra (Figure S5) illustrate the existence of five elements, namely C, N, O, Ni, and Y, within the carrier material. This observation implies the effective incorporation of the metallic elements Ni and Y into the carbon framework of the carrier. For the purpose of validating the presence of Ni and Y metals in the carriers, spherical aberration-corrected SEM was employed to examine the catalysts synthesized under pyrolytic conditions at 850°C. The visualization of Ni and Y elements within the carriers is clearly demonstrated in Figure 1i, with the diatomic combinations of Ni and Y encircled in yellow within the figure.

Additionally, the specific surface area and pore size distribution of the catalysts fabricated at different pyrolysis temperatures were evaluated through the analysis of N₂ adsorption–desorption isotherms. As evidenced by the data presented in Figure 2 and Figure S6, the Ni–Y-PC-X specimens exhibited IV-type adsorption isotherms, a distinctive trait commonly associated with mesoporous materials. The pore size distribution analysis revealed that the pore diameters of Ni–Y-PC-X ranged between 15 and 20 nm, providing additional confirmation of the mesoporous nature of the synthesized Ni–Y-PC materials. As the pyrolysis temperature was elevated from 750°C to 850°C, a marked increase in the specific surface area was noted, demonstrating a significant escalation from 206 to 351 m²·g⁻¹. Nevertheless, upon further elevation of the pyrolysis temperature to 950°C, a notable decrease in the specific surface area to 269 m²·g⁻¹ was observed. This phenomenon could be ascribed to the disruption of the ordered mesoporous structure within the material. It is commonly accepted that mesoporous materials serve as advantageous carriers for obtaining catalytically active sites, and an increased specific surface area facilitates the introduction of additional active centers, thereby enhancing the catalytic activity of the materials. The specimen produced at a pyrolysis temperature of 850°C is characterized by the highest specific surface area and optimal pore size, indicating its potential to display the most favorable CO₂RR catalytic activity among the three synthesized samples.

Figure 2

N₂ adsorption and desorption curves of Ni–Y-PC-750°C, Ni–Y-PC-850°C, Ni–Y-PC-950°C and their pore size distribution: (a, c, e) adsorption–desorption curve of N₂; (b, d, f) pore size distribution of N_2.

To delve deeper into the chemical composition and elemental states of the samples, XPS tests were carried out on the synthesized Ni–Y DACs. Figure 3 illustrates the XPS test result plot for the sample subjected to pyrolysis at 850°C. Figure 3a presents the comprehensive spectrum of the sample, revealing the presence of carbon, nitrogen, oxygen, nickel, and yttrium. This observation indicates the successful integration of nickel and yttrium into the porous carbon support structure. The high-resolution C 1s spectrum displayed in Figure 3b manifests the presence of four distinct peaks, each characterized by binding energies of 284.6, 285.2, 286.2, and 289.5 eV. These peaks can be attributed to the C–C, C═C, C–OH, C–N, and C═O functional groups originating from the porous carbon carrier. In Figure 3c, the high-resolution N 1s spectrum of the sample exhibits distinct peaks corresponding to graphite N (400.9 eV), pyridine N (398.5 eV), pyrrole N (399.2 eV), and oxide N (403.8 eV). In Figure 3d, the high-resolution O 1s spectrum of the sample is depicted, exhibiting three characteristic peaks at 530.5, 531.3, and 533.7 eV. These peaks correspond to the C–O, O═C–O, and C═O bonds, respectively, and were resolved through inverse deconvolution. In the high-resolution Ni 2p spectrum depicted in Figure 3e, two prominent peaks are observed at 855.9 and 873.6 eV, corresponding to the 2p_1/2 and 2p_3/2 states, respectively. These findings suggest a Ni²⁺ oxidation state within the material. Additionally, the presence of binding energy peaks at 862.2 eV, attributed to the satellite peaks of Ni 2p, further strengthens this conclusion. Furthermore, the Y 3d high-resolution mapping, also presented in Figure 3f, showcases distinct peaks at 157.9 and 159.8 eV, corresponding to the 3d_3/2 and 3d_5/2 states of Y, signifying a +3 valence state for the Y element within the material.

Figure 3

XPS plots of Ni–Y bimetallic catalyst at 850°C pyrolysis temperature: (a) full spectrum, (b) high resolution C 1s, (c) high resolution N 1s spectra, (d) high resolution O 1s spectra, (e) Ni 2p high resolution spectra, and (f) Y 3d high resolution spectra.

Based on the unique structure of the Ni–Y DAC material, we anticipate a favorable electrocatalytic performance from this compound. Initially, we conducted linear voltammetric scanning (LVS) tests on the synthesized Ni–Y-PC-X catalysts under Ar and CO₂ atmospheres, respectively, as an initial step to ascertain the CO₂RR properties of the material. The test results are depicted in Figure 4a for reference. In the experimental range, the current density of Ni–Y-PC-X under an argon atmosphere is substantially lower than that observed under a CO₂ atmosphere. This initial finding strongly suggests the potential for excellent electrocatalytic CO₂ reduction performance of Ni–Y-PC-X, marking a significant result. The highest current density was observed under pyrolysis conditions at 850°C, suggesting that Ni–Y-PC-X is expected to demonstrate optimal CO₂RR performance at the same temperature.

Figure 4

Parameters of the electrochemical CO₂RR over the Ni–Y-PC and Ni–Y-PC-X catalysts: (a) Ni–Y-PC-X linear scanning voltammetric curve, (b) Ni–Y-PC-X CO Faraday efficiency plot, (c) Ni–Y-PC-X hydrogen Faraday efficiency plot, (d) Ni–Y-PC-X current density vs potential growth curve, (e) Ni–YPC-X electrochemical impedance spectrum, (f) Ni–Y-PC-X Tafel slope curve, (g) Ni–Y-PC-850°C long time electrolysis plots at −0.93 V vs RHE potential, (h) Ni–Y-PC-850°C CO partial current density plot under N₂ atmosphere, and (i) Faraday efficiency plots with addition of single metal CO efficiency plot.

Subsequent to the demonstration of exceptional electrocatalytic CO₂ reduction potential by Ni–Y-PC-X in the LVS test, we undertook an assessment of its CO₂RR performance. This evaluation involved the use of an H-type electrolytic cell employing a timed current–time (i–t) method in a 0.5 M KHCO₃ electrolyte at ambient temperature and pressure. The quantification of gaseous and liquid products was achieved through gas chromatography and nuclear magnetic resonance spectroscopy, respectively. Gas chromatography analysis identified hydrogen and carbon monoxide as the sole gaseous products. Remarkably, no substantial liquid products were discernible within the examined voltage range. In the graphical representation provided in Figure 4b, the CO Faraday efficiency of Ni–Y-PC-X is depicted across the range of potentials tested, with a focus on Ni–Y-PC-850°C at −0.93 V vs RHE. Notably, at this specified potential, the Faraday efficiency of CO for Ni–Y-PC-850°C registers at 89%, maintaining a level of more than 72% over an extensive potential window (−0.63 to 1.03 V vs RHE). It is noteworthy that the highest Faraday efficiencies obtained for Ni–Y-PC-750°C and Ni–Y-PC-950°C are 70% and 69% respectively, a finding which aligns with the results of the LVS test analysis, and further corroborates the outcomes of the BET characterization analysis.

In the context of Figure 4c, the curve portraying the fluctuation of Ni–Y-PC-X hydrogen Faraday efficiency as a function of potential is depicted. This visual representation underscores the effective suppression of the competing hydrogen precipitation reaction under the pyrolysis conditions at 850°C, resulting in a modest maximum Faraday efficiency of 18%. This observation is likely attributable to the conducive nature of this temperature for the carrier to selectively capture the metal elements released during the thermal disintegration of bimetallic complexes, thereby fostering the creation of a higher concentration of catalytically active sites. Figure 4d presents the current density as a function of potential for Ni–Y-PC-X. This illustration reveals that the current density of Ni–Y-PC-850°C at a potential of −1.03 V vs RHE can achieve 20 mA·cm⁻², representing an approximate twofold increase in comparison to that of Ni–Y-PC-750°C (11 mA·cm⁻²). Moreover, it is observed that with the escalation of pyrolysis temperature, the current density undergoes a notable reduction to 8.7 mA·cm⁻². This phenomenon can be ascribed to the spontaneous migration of Y elements, leading to the formation of nanoclusters at elevated temperatures due to the specific surface energy. Figure 4e depicts the electrochemical impedance spectrum of Ni–Y-PC-X. Electrochemical impedance serves as a reflection of the material’s electron transport resistance. Notably, among the Ni–Y-PC-X samples, Ni–Y-PC-850°C displays the lowest impedance value, indicating minimal electron transport resistance. This finding elucidates the observed higher current density in the measured potential range, thus underscoring the influence of electron transport resistance on the electrochemical behavior of the material. Figure 4f displays the Tafel slope curve of Ni–Y-PC-X, presenting insights into the material’s kinetic performance during the reaction process. Within the Ni–Y-PC-X series, the Tafel slope of Ni–Y-PC-850°C is recorded at 89 mV·dec⁻¹, notably lower than the Tafel slopes observed for Ni–Y-PC-750°C (110 mV·dec⁻¹) and Ni–Y-PC-950°C (135 mV·dec⁻¹). This disparity indicates that Ni–Y-PC-850°C exhibits the swiftest reaction kinetics during the electrocatalytic CO₂RR. The recorded Tafel slope value of 89 mV·dec⁻¹, approaching 60 mV·dec⁻¹, suggests that the conversion of *CO²⁻ to *COOH represents the rate-determining step of CO₂RR for this catalyst.

For the purpose of assessing the stability of Ni–Y-PC-X, an extended electrolysis of Ni–Y-PC-850°C was carried out at a potential of −0.93 V vs RHE, and the results of this evaluation are depicted in Figure 4g. During the extended 12 h electrolysis, the current density demonstrated a negligible decrease and consistently retained more than 80% of the Faraday efficiency. The observed test outcome indicates the outstanding stability of Ni–Y-PC-850°C. Subsequently, to ascertain the origin of carbon in the CO product, we replicated the aforementioned test methodology in an N₂-saturated electrolyte. Figure 4h demonstrates that the CO current density in the N₂ saturated electrolyte remains near zero within the measured potential range of −0.63 to −1.03 V vs RHE. This observation provides compelling evidence that the CO generated in the CO₂RR exclusively originates from the conversion of CO₂, rather than other conceivable carbon sources. In addition, we assessed the Faraday efficiency by introducing a single metal in the same voltage range, and the outcomes are depicted in Figure 4i. The Faraday efficiencies of CO for Ni-PC and Y-PC reached only 65% and 20%, respectively, significantly lower than that of Ni–Y-PC-850°C. This discrepancy serves as strong indication that the introduction of Y monatomic species considerably enhances the CO₂RR performance of the material.

To substantiate the impact of PC carriers on augmenting the electrocatalytic performance of the materials, we carried out the direct carbonization of the Ni–Y bimetallic complexes at 850°C and subsequently assessed their electrocatalytic performance in CO₂ reduction. The results of these tests are presented in Figure S7. Figure S7a illustrates the plot of CO Faraday efficiency for the Ni–Y bimetallic complex at the measured potential, demonstrating a maximum Faraday efficiency of only 54% at −0.83 V vs RHE potential, significantly lower than that of Ni–Y-PC. Figure S7b depicts the plot of hydrogen Faraday efficiency for the Ni–Y bimetallic complex, revealing a notable occurrence of hydrogen precipitation reaction following direct carbonization, in stark contrast to the performance of Ni–Y-PC with the inclusion of the carrier. Figure S7c displays the growth curve of the current density for the Ni–Y bimetallic complex as a function of the test potential. Notably, the current density of Ni–Y-PC, with the introduction of the carrier, increased by almost fourfold to 20 mA·cm⁻² compared to the direct carbonization of the bimetallic complex. This substantial enhancement can be attributed to the remarkable electrical conductivity of the carbon–nitrogen carrier. In Figure S7d, the electrochemical impedance spectrum of the bimetallic complex is presented, revealing that the impedance value of the bimetallic complex is nearly twice that of Ni–Y-PC with the incorporated carrier. This disparity indicates a higher resistance to charge transfer in the bimetallic complex. Overall, it is evident that the carbon and nitrogen carriers significantly contribute to the improvement of the material’s electrocatalytic CO₂ performance. This is attributed to the occurrence of metal redispersion during the pyrolysis process, wherein the carriers offer anchoring points to facilitate the redispersion of the metal elements. Additionally, the superior electrical conductivity exhibited by the carbon and nitrogen carriers, along with the well-regulated distribution of pore sizes, further underpins the heightened electrocatalytic activity of the materials.

3 Conclusion

In this study, a designed and synthesized Ni–Y DAC has been implemented on a porous carbon substrate. Various characterization techniques revealed that the Ni and rare earth Y elements were uniformly distributed as single atoms within the carrier material. Electrochemical experiments indicate that the introduction of the Y element serves to modulate the local electron density of the Ni element, thereby playing a crucial role in enhancing the CO₂RR performance of the material. This study confirms the exceptional catalytic performance of DACs, thereby broadening the potential for their utilization in electrocatalytic CO₂ applications. In addition, this study is poised to establish a novel pathway for the design and production of DACs incorporating transition metals and rare earth elements, thereby heralding a paradigm shift in catalyst innovation and advancement.