Electroreduction of CO₂ (CO₂RR) into chemical fuel is a promising method of carbon capture and utilization towards efficient utilization of renewable energy [1-4]. Despite the recent progress made in CO₂RR, there are currently some bottlenecks that need to be overcome en route to efficient energy conversion, including large activation barrier due to the stable chemical bond in CO₂, low selectivity limited by competing hydrogen evolution reaction (HER) and degradation of catalytic activity [5-11]. Although precious metal-based electro-catalysts, like Au [12, 13], Ag [14-16], Pd [17, 18] and their alloys [19-24], possess fast CO₂RR kinetics, but their low cost-efficiency impedes large-scale applications. As a result, developing the active and high-selectivity catalysts based on earth-abundant material to overcome these bottlenecks is highly desirable for CO₂RR.

Recently, transition metal nickel (Ni) has been proved to effectively convert CO₂ into CO theoretically and experimentally [25-29]. Up to now, various strategies have been proposed to design efficient Ni-based CO₂RR catalyst, including heteroatom doping engineering [30], crystal phase engineering [31, 32], strain engineering [33] and nano-scale engineering [34-36]. In particular, downsizing the transition metal-based catalysts into nano-scale has been widely applied, which can increase the number of active sites and maximum metal utilization efficiency [36-38]. Besides above strategies, the valence state engineering of transition metal has been demonstrated to effectively tune the catalytic activities. To date, extensive efforts have been made in valence state engineering in different catalytic fields, such as the valence effect of MnO_x for oxygen reduction (ORR) [39], high-valence engineering for oxygen evolution (OER) [40], and Mo-valence state adjustment for optimizing hydrogen evolution (HER) [41]. Nevertheless, the study of effect of valence state on Ni-based CO₂RR catalyst is far from satisfaction. Therefore, it is highly desired to combine the downsizing the transition metal with valence state engineering for designing a more efficient CO₂RR catalyst.

Here, we developed a simple two-step high temperature reduction strategy to construct low-valent Ni nanoparticles on N-doped carbon layer as electro-catalyst (Ni-NC). Such a Ni-NC catalyst exhibits remarkable CO₂ electroreduction performance and high selectivity for CO. It can achieve high Faradaic efficiency (FE) over 90% for CO in wide potential range from −0.6 V to −1.1 V and gives a maximum FE reaches 98% at −0.8 V, which significantly exceed the most reported Ni-based catalyst (all potential are in reference to the RHE). Detailed electrochemical tests and DFT calculation demonstrate that the low-valent Ni should be responsible for the high selectivity for CO.

The Ni-NC was synthesized via a simple two-step pyrolysis method (Fig. 1a). Simply, glucose (GC), dicyandiamide (DCDA) and NiCl₂ was used as carbon, nitrogen and nickel source, respectively, which were mixed and sintered at 550 ℃ and followed by carbonized at 900 ℃ under flowing Ar. The crystal structure was confirmed by X-ray diffraction (XRD) pattern in Fig. 1b. Two diffraction peaks corresponding to the (110) and (200) planes of metallic Ni are observed at 44.5° and 51.9° for Ni-NC (JCPDS, No. 71–4655) [42], which are not found in the NC counterpart. The Ni content was measured to be 7.2 wt% based on the inductively coupled plasma masss pectrometry (ICP-MS). The scanning electron microscopy (SEM) images of Ni-NC and NC exhibit a clear layered-structure morphology with folds and wrinkles (Fig. 1c and Fig. S1 in Supporting information).

To further reveal the nanostructure features of Ni-NC, high-resolution transmission electron microscopy (TEM) was performed. As shown in Fig. 2a and Figs. S2 and S3 (Supporting information), they display the carbon layer wrapped Ni nanoparticles. The lattice fringes of d = 0.203 nm in the catalysts of Ni-NC, which correspond to the (111) plane of the Ni (Fig. 2b), which is well in line with XRD data [43]. The TEM-energy dispersive X-ray spectra (TEM-EDS) mapping results show the existence of Ni, N and C elements over the nanosheets (Fig. 1d). The chemical states of Ni-NC were further investigated by X-ray photoelectron spectroscopy (XPS) (Fig. 2c and Figs. S4 and S5 in Supporting information). As shown in Fig. S5, four types of N species were distinguished, including pyridinic N (398.86 eV), pyrrolic N (399.7 eV), graphitic N (401.15 eV) and oxidized N (402.52 eV) species [44, 45]. Fig. 2c displays the high-resolution of Ni spectrum of Ni-NC, Ni 2p_3/2 deconvoluted in two peaks. The minor peak at 852.4 eV, which is corresponding to Ni0. The major peak at 854.6 eV is lower than Ni²⁺ (856.0 eV), suggesting the low-valent state of Ni in Ni-NC [46, 47].

The electro-catalytic activities of NC, Ni-NC were studied using a H-type electrochemical cells under CO₂ saturated 0.1 mol/L KHCO₃. Additionally, NiO_x-NC was synthesized from Ni-NC by applying positive potential in 1 mol/L KOH, with purpose of verifying the catalytic effectiveness of low-valent Ni. As presented in Fig. 3a, the metallic Ni in Ni-NC was first oxidized to form α-Ni(OH)₂ and followed by oxidized to Ni³⁺ [48]. And the α-Ni(OH)₂ irreversibly transferred to β-Ni(OH)₂ during the range from 0.6 V to 1.3 V, which cannot be reduced to metallic Ni. Linear Sweep Voltammetry (LSV) displayed in Fig. 3b shows total current density at the voltage range from 0 to −1.2 V vs. RHE. The total current density of both NC and NiO_x-NC at −1.2 V were close to 10 mA/cm², while the total current density of Ni-NC almost reached 30 mA/cm², suggesting prominent electrocatalytic activity of Ni-NC due to the existence of low-valent Ni in Ni-NC. Faradaic efficiency (FE) and partial current density of CO were displayed respectively in Figs. 3c and d. Compared with the FE of CO exceeding 92% at the voltage range from −0.6 V to −1.1 V catalyzed by Ni-NC, the maximum FE of CO over NiO_x-NC is just close to 80%. Furthermore, the maximum FE of CO over NC is as low as 53%, even falling to no more than 10% at more negative potential. Consequently, CO partial current density of Ni-NC is much higher than the other counterparts at the whole voltage range applied. This loss in performance of NiO_x-NC is likely the result of Ni oxidation [27], which diminished the active site from metallic Ni after applying positive potential. The stability of the Ni-NC catalyst was tested by continuous working at −0.8 V for more than 10 h, both the current density and Faradaic efficiency stayed stable as presented in Fig. 3e. This result evidenced the metallic Ni would provide the active sites. And XRD pattern, TEM images together with high-resolution XPS Ni 2p spectra suggest that the excellent structural and chemical stability of the Ni-NC during the CO₂RR process (Figs. S6 and S7 in Supporting information).

The oxidation of Ni during the positive potential scanning studied by XPS. The result in Fig. 4a demonstrate that the binding energy for Ni in NiO_x-NC has an obvious shakeup satellite signal at 861.1 eV, which is attributed to the Ni²⁺ [49]. This observation illustrates the surface oxidation of Ni element. The electrochemical active surface area (ECSA) of catalyst was determined to study the factor of CO₂RR activity enhancement. The ECSA of Ni-NC and NiO_x-NC were evaluated by double-layer capacitance (C_dl) depend on CV curves (Figs. S7a and b in Supporting information). The C_dl can be obtained by determining the slope of the fitted plots. The C_dl of Ni-NC is 4.3 mF/cm², which is nearly 16.5 times as high as NiO_x-NC (0.26 mF/cm²) (Fig. 4b), indicating that the oxidation process dramatically decreases the active sites. Furthermore, electrochemical impedance spectrum (EIS) was performed to evaluate the reaction kinetics. The Ni-NC exhibits smaller charge transfer resistances than that of NiO_x-NC (Fig. 4c), suggesting its fast charge transfer kinetics. These results further demonstrate the excellent electrochemical properties of Ni-NC due to the existence of low-valent Ni.

DFT calculations were performed to reveal the mechanism of the extremely high-selectivity electrocatalytic CO₂RR process on Ni-NC. According to the observed experimental results, models of (111) and (200) feces in metallic Ni were built and presented in Fig. 5a, subsequently the free energy diagram of CO₂RR process were calculated and presented in Fig. 5c. The free energy change for *COOH formation (CO₂ + (H⁺ + e⁻) → *COOH + H₂O) in Ni (111) is 0.2 eV, while in Ni (200) is −0.03 eV. After the *COOH formation, *CO formation from *COOH (*COOH + (H⁺ + e⁻) →*CO + H₂O) is exothermic on both faces. It can be seen that, the H⁺/e⁻ steps in both faces can proceed with low free energy, which indicates that the *CO species may easily form in both faces under potential condition of experiment. And the free energy change for CO desorption (*CO → CO(g) + *), which cannot be tune by the external potential, is more endothermic Ni (200) than that in Ni (111), this indicates that the Ni (200) would be more easily poisoned by *CO and the desorption of *CO in Ni (111) is more easily. We also examined the selectivity between the CO₂RR process and HER process (Figs. 5b and d). We compared the free energy change of potential-determining step (PDS) of CO₂RR and HER on both faces, the ΔG_PDS (CO₂RR) is smaller than the ΔG_PDS (HER), suggesting that the CO₂ reduction process is thermodynamically more favorable than that of HER, identical with the experimental high-selectivity of CO.

In summary, low-valent Ni-based nanoparticles were successfully constructed on carbon layer (Ni-NC) via a two-step pyrolysis method. Notably, the Ni-NC exhibited high-selectivity for CO. The Faradaic efficiency of CO was as high as 98% at overpotential of −0.8 V vs. RHE, with good stability over 10 h. In contrast, a significant decrease in CO₂ reduction activity was observed after low-valence Ni was oxidized, resulting from the decreasing active sites. Experimental and theoretical studies indicate that Ni (111) on Ni-NC may be a type of active site for CO₂RR, originating from more favorable initiation of CO₂ reduction than HER. This discovery not only pays a new path for design of high-selectivity to CO₂ reduction, but also presents new understanding of Ni-based CO₂ reduction.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors also would like to acknowledge the financial support from the Shenzhen Science and Technology Research Grant (No. JCYJ20200109140416788, China), the Chemistry and Chemical Engineering Guangdong Laboratory (No. 1922018, China) and National Key R & D Program of China (No. 2020YFB0704500).

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2021.07.016.