Janus NiN₄–CuN₄ catalyst supported by a double-layered ZIF-8 structure for CO₂ electrocatalytic reduction to C₂₊ products

Abstract

Copper (Cu)-based catalysts exhibit considerable potential for the electrocatalytic CO₂ reduction reaction (CO₂RR) to multi-carbon (C₂₊) products; nevertheless, competitive products (e.g., formic acid) have also been produced by thermodynamic constraints. The attainment of selective C₂₊ products exhibiting high faradaic efficiency (FE) and current density remains a significant challenge. Here, we designed a tandem catalytic strategy and prepared a series of Janus catalysts supported by double-layered ZIF-8 with inner MN₄ and outer CuN₄ for C₂₊ products (especially n-propanol). The synthetic catalyst, NiN₄@CuN₄, achieved high faradaic efficiencies of 80.6% for C₂₊ at −1.6 V in a flow cell, with 24.4% n-propanol (CH₃CH₂CH₂OH) and 36.6% ethanol (CH₃CH₂OH). In situ Raman experiments along with mechanism analysis indicate that the synergistic interaction of inner NiN₄ and outer CuN₄ significantly facilitates CO production and CO electroreduction, thereby greatly promoting C₂₊ production.

---

1 Introduction

The transformation of CO₂ into valuable fuels and the utilization of renewable energy resources are prospective avenues for achieving carbon neutrality.^1–6 It is a well-established fact that C₂₊ products, exemplified by C₂H₄, C₃H₇OH and C₂H₅OH, exhibit superior energy density and economic value in comparison to their C₁ counterparts. The following compounds are included in this category: methane (CH₄), formic acid (HCOOH) and carbon monoxide (CO).⁷ The complex pathways of electron and proton transfer that occur in the CO₂RR give rise to the emergence of competing reactions, such as the generation of C₁ products and the hydrogen evolution reaction (HER).⁸ These competing reactions act as a thermodynamic limitation on the formative process of C₂₊ products. Additionally, the inherent slowness of the C–C coupling process represents a kinetic limitation on the C₂₊ production.^9,10 The extant literature has demonstrated the considerable potential of Cu-based catalysts for the production of C₂₊.¹¹ Significant endeavors have been undertaken to attain elevated CO₂RR selectivity and activity. Strategies such as tandem catalysis^12–14 and single atom^15–17 active sites have been extensively studied, but FE_C2+ is still below that of the C₁ products (FE_C1).

Single-atom catalysts are defined as catalysts in which high atom utilization is achieved through the dispersion of metal atoms.¹⁸ For instance, Gu et al.¹⁹ published an innovative N-triggered Zn single-atom catalyst that exhibited excellent oxygen reduction reaction (ORR) performance and low overpotential. Presently, single-atom catalysts are being researched for the CO₂RR, with most studies focusing on CO. Single-atom transition metal catalysts, including Fe,^20–22 Co,^23–25 and Ni,^26–28 primarily facilitate the electroreduction of CO₂ to C₁ products through electrocatalysis. Furthermore, C–C coupling can only be achieved effectively by Cu-based catalysts. Nevertheless, the restricted number of catalytic sites and the considerable distance between them render the conversion of C₂ products with a single Cu single-atom catalyst an arduous task.

Janus catalysts have the capacity to separate two metals, and they are frequently employed catalysts in tandem catalysis.²⁹ Nanoparticles exhibiting high activity in cascade catalytic reactions are produced by Janus catalysts with multifunctional active sites that possess diverse catalytic properties. Transition metal monoatomic materials (e.g., Fe, Co, Ni) generate high concentrations of CO during CO₂ electroreduction; this CO then promotes C₂ product formation over Cu-based catalysts. Zhang et al.³⁰ reported that a gas diffusion electrode achieved high CO utilization efficiency and high FE_C2+ through two segments catalyzed in series. Janus catalysts represent a strategy of considerable merit for efficient CO₂ tandem catalysis. Nevertheless, this material is currently the subject of less study compared to other areas, and its synthesis is extremely challenging.

Herein, in this study, we prepared a series of MN₄@CuN₄ catalysts with a two-layer ZIF-8 structure, which was capable of enriching CO and achieving high efficiency in the tandem catalysis of carbon dioxide. The synthetic catalyst, NiN₄@CuN₄, achieved 81.0% FE_C2+ at −1.6 V vs. RHE potential in 1 M KOH, with 24.4% CH₃CH₂CH₂OH and 36.6% CH₃CH₂OH. This work employs a range of transition metals (including Fe, Co, and Ni) as the CO₂ activation site, utilizes the double-layered structure to facilitate CO enrichment, and offers a high CO concentration at the outer CuN₄ site. This provides a new strategy of electrocatalyst synthesis for transformation of the CO₂RR into a C₂₊ generation reaction.

2 Experimental section

2.1 Reagents and materials

Zinc acetate dihydrate (Zn(CH₃COO)₂·2H₂O), nickel(II) chloride hexahydrate (NiCl₂·6H₂O), iron(II) sulfate heptahydrate (FeSO₄·7H₂O), 2-methylimidazole (2-MeIm), cobalt(II) chloride hexahydrate (CoCl₂·6H₂O), copper(II) chloride dihydrate (CuCl₂·2H₂O), 1,10-phenanthroline, Nafion 117 solution, and potassium hydroxide (KOH) were analytical grade and purchased from Shanghai Macklin Biochemical Technology Company. Methanol (MeOH), (methyl sulfoxide)-d₆ (DMSO-d₆), deuterium oxide (D₂O) and anhydrous ethanol (EtOH) were purchased from Shanghai Aladdin Biochemical Technology Co. Ltd. High purity CO₂ (99.999%) and N₂ (99.999%) were purchased from Zhuhai Huaxin Gas Co. Ltd. All the chemicals and reagents were used without further purification.

2.2 Synthesis of Niphen@ZIF-8

A solution of 0.357 g NiCl₂·6H₂O and 0.743 g 1,10-phenanthroline in 120 mL methanol was prepared and stirred for 0.5 h. Subsequently, 4.61 g Zn(CH₃COO)₂·2H₂O was added to the solution and stirred for 1 h. The aforementioned solution was then added dropwise to a solution of 17.2 g 2-MeIm in 10 mL of methanol, which was stirred for 5 min and allowed to precipitate for 24 h. The products were washed with methanol, centrifuged three times and then dried in a vacuum at 60 °C. The product was prepared and named Niphen@ZIF-8.

2.3 Synthesis of Niphen@ZIF-8@Cuphen

Niphen@ZIF-8 was added into 120 mL of methanol and dispersed into a 3 mg mL⁻¹ suspension. 0.341 g CuCl₂·2H₂O was added and stirred for 0.5 h. Subsequently, 0.991 g 1,10-phenanthroline was added and stirred for 1 h. The products were washed with methanol, centrifuged three times and then dried in a vacuum at 60 °C. Finally, Niphen@ZIF-8@Cuphen was obtained.

2.4 Synthesis of Niphen@ZIF-8@Cuphen@ZIF-8

Niphen@ZIF-8@Cuphen was added into 40 mL of methanol and dispersed into a 3 mg mL⁻¹ suspension. 1.76 g Zn(CH₃COO)₂·2H₂O was added and stirred for 1 h. Subsequently, 0.991 g 1,10-phenanthroline was added and stirred for 24 h. The product was washed with methanol, centrifuged three times and then dried in a vacuum at 60 °C. The final product was labelled Niphen@ZIF-8@Cuphen@ZIF-8.

2.5 Synthesis of NiN₄@CuN₄

250 mg Niphen@ZIF-8@Cuphen@ZIF-8 was heated to 900 °C under N₂ in a tube furnace at 5 °C min⁻¹ and held for 2 h. After cooling, black NiN₄@CuN₄, was collected.

2.6 Synthesis of CoN₄@CuN₄ and FeN₄@CuN₄

CoN4@CuN4 and FeN4@CuN4 were obtained by replacing 0.357 g NiCl₂·6H₂O with 0.357 g CoCl₂·6H₂O and 0.357 g FeSO₄·7H₂O, and other steps were the same as in the preparation of NiN₄@CuN₄.

2.7 Electrochemical measurements

Electrochemical tests were undertaken within an airtight flow cell, using a Chenhua 660e electrochemical workstation. The carbon paper working electrode was loaded with the catalyst at a density of 1 mg cm⁻² and measured. The counter electrode and reference electrode were a 1 cm² platinum sheet and a Hg/HgO electrode, respectively. Cathode and anode compartments in the flow cell were divided by an anion-exchange membrane (AEM). Both compartments contained 1 M aqueous KOH electrolyte, with flow rates controlled by a peristaltic pump at 10 mL min⁻¹. Flow rates of carbon dioxide (CO₂) were regulated by a digital gas flow controller, set at 20 standard cubic centimeters per minute (sccm). It is important to note that not all measured voltages were IR-compensated; however, they were all subjected to conversion to the reversible hydrogen electrode (RHE) potential. The conversion formula is E (vs. RHE) = E (vs. Hg/HgO) + 0.095 V + 0.0591 V × pH.

2.8 Quantitative analysis of gas and liquid products

The gaseous products were subjected to analysis by gas chromatography equipped with FID and TCD detectors. The FE of gaseous products was calculated as follows:w
here z represents the number of exchanged electrons for CO₂ electroreduction to a specific product, n is the molar quantity of a specific electrocatalytic product, F is the faradaic constant (96 485C mol⁻¹), the term “Q” is used to denote the charge quantity passing through the working electrode during the electrocatalytic process, x is the volume concentration of the gaseous product, the term “v” is the flow rate of CO₂, the term “P” is the standard atmospheric pressure, I is the current passing through the working electrode during electrocatalysis, R is the gas parameter (8.314 J mol⁻¹ K⁻¹), and the term “T” is the room temperature.

The liquid product was determined by ¹H NMR spectra. 500 µL of the electrolyte after the CO₂RR and 100 µL of D₂O containing 50 ppm (m m⁻¹) dimethyl sulfoxide as an internal standard were used to determine the product-specific Faraday efficiency on a Bruker 400 MHz spectrometer. The Faraday efficiency of the liquid product was calculated as follows:w

here Q_total represents the total charge consumed during the electrolysis, Q_product is the charge during the formation of the liquid product, m represents the number of exchanged electrons for the CO₂RR, n is the number of moles of a certain product, and the term “t” is electrolysis time.

3 Results and discussion

3.1 Morphological and structural investigation

The tandem catalyst, NiN₄@CuN₄, was prepared layer by layer using a bottom–up strategy. The synthesis route is shown in Fig. 1a. Subsequent to a 900 °C treatment, a double-layered carbon material containing an inner NiN₄ site and outer CuN₄ was prepared.³¹ The synthesized materials were analyzed via XRD, and the resulting pattern is presented in Fig. 2a. The standard XRD pattern of ZIF-8 is shown. It is evident that the structures of Niphen@ZIF-8, Niphen@ZIF-8@Cuphen and Niphen@ZIF-8@Cuphen@ ZIF-8 were consistent with the ZIF-8 standard card, thereby demonstrating the retention of identical structural characteristics. The examination of Ni and Cu was not conducted due to their low content.³² The X-ray diffraction (XRD) pattern of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄ (Fig. 2b) exhibited diffraction peaks consistent with the (002) and (100) planes of graphene, with peak positions of 22.6° and 43.5°, respectively.³³ The Raman spectrum illustrated in Fig. 2c also corroborated the presence of graphene peaks. The sp² graphene peaks were observed at 2700 cm⁻¹.³⁴ This ratio of disorder to graphene peak strength (I_D: I_G) for NiN₄@CuN₄ was 1.25, indicating that heat treatment caused obvious defects.³⁵

Fig. 1 (a) The synthesis route of Janus catalyst MN₄@CuN₄ supported by double-layered ZIF-8, (b) the generation process of C₂₊ products on MN₄CuN₄, and (c) schematic representation of the flow cell.

Fig. 2 (a) XRD patterns of ZIF-8, Niphen@ZIF-8, Niphen@ZIF-8@Cuphen and Niphen@ZIF-8@Cuphen@ ZIF-8; (b) XRD patterns of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄; (c) Raman spectrum of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄; SEM image of (d) Niphen@ZIF-8, (e) Niphen@ZIF-8@Cuphen, (f) Niphen@ZIF-8@Cuphen@ ZIF-8, (g) NiN₄@CuN₄, (h) FeN₄@CuN₄ and (i) CoN₄@CuN₄; TEM image of (j) FeN₄@CuN₄, (k) NiN₄@CuN₄ and (l) CoN₄@CuN₄; (m) magnified HAADF-STEM image of NiN₄@CuN₄; (n) HAADF-STEM image of NiN₄@CuN₄; and (o) EDS elemental mappings of C, Cu, N and Ni.

The prepared materials were characterized by SEM (Fig. 2d–i). The Niphen@ZIF-8 sample exhibited a standard ZIF-8 dodecahedral structure with an edge length of approximately 500 nm. The incorporation of Cuphen into Niphen@ZIF-8@Cuphen results in an uneven surface topography. After coating with ZIF-8 and heat treatment again, the NiN₄@CuN₄ surface is smooth again and the average crystal width is 550 nm. The successful coating with ZIF-8 is evidenced by the increase in material size and the structural change. The catalysts CoN₄@CuN₄ and FeN₄@CuN₄ prepared by the same process are shown in Fig. 2e and f. The morphology of the two catalysts is consistent with that of NiN₄@CuN₄.

As shown in Fig. 2j–l, NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄ have a clear porous concave dodecahedral structure in the TEM image.

The magnified high-angle annular dark-field scanning TEM (HAADF-STEM) indicates that there are carbon streaks and obvious individual bright spots in the material, with no metallic streaks present, suggesting the successful synthesis of single-atom catalysts (Fig. 2m). NiN₄@CuN₄ has a distinct concave dodecahedral structure inside and a concave dodecahedral structure outside, demonstrating the successful synthesis of the double-layered ZIF-8 skeletal structure (Fig. 2n),which is consistent with previous reports.³¹ TEM elemental mapping images (Fig. 2o) showed that Ni was uniformly distributed in the innermost layer of NiN₄@CuN₄ (inside the inner ZIF-8) and the uniform distribution of Cu was observed in the outer ZIF-8 layer, indicating the successful synthesis of NiN₄@CuN₄.

The N₂ adsorption–desorption isotherm is a key indicator of numerous physical characteristics inherent to the tandem catalyst (Fig. 3a). The values of specific surface areas of NiN₄@CuN₄, CoN₄@CuN₄, and FeN₄@CuN₄ were found to be 366.3, 203.7, and 293.5 m² g⁻¹, respectively. The elevated specific surface area of NiN₄@CuN₄ suggests numerous active sites, consistent with its remarkable catalytic activity. A clear H3 hysteresis loop was observed at medium-pressure (0.3 < P < 0.8), without an adsorption plateau at high pressure (0.8 < P < 1). The isotherms exhibited characteristics of type IV isotherms. The material displayed clear mesoporous characteristics, with a pore size distribution curve indicating a predominant pore size of 2.3 nm and 8.5 nm (Fig. 3b).

Fig. 3 (a) N₂ adsorption–desorption isotherm curves, (b) pore size distribution of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄, (c) XPS survey of NiN₄@CuN₄, (d) C 1s spectra of NiN₄@CuN₄.

X-ray photoelectron spectroscopy (XPS) analysis of the tandem catalyst provided precise insights into the elemental composition and surface valence states. Fig. S7 depicts the XPS total spectrum of NiN₄@CuN₄, which reveals the presence of discernible peaks corresponding to elements such as C, N, Ni, and Cu within NiN₄@CuN₄. Limited by the detection depth (∼5–10 nm), the photoelectron signals of Cu and Ni are significantly attenuated, especially the Ni 2p orbital cannot be deconvoluted. The contents of Cu and Ni were 1.03 wt% and 2.32 wt% (Table S1), respectively, which were measured by ICP-OES quantitative analysis (pretreatment of aqua regia digestion). Furthermore, the XPS tests of CoN₄@CuN₄ and FeN₄@CuN₄ (Fig. S1 and S2) also demonstrated that they contained Fe, Co, N and Cu. The C 1s spectrum of NiN₄@CuN₄ is displayed in Fig. 3c. Following deconvolution, the spectrum displays five distinct peaks corresponding to C=C, C–C, C–N, C–O, and C=O bonds.³⁶ As illustrated in Fig. 3d, the deconvoluted N 1s spectrum is indicative of the presence of pyridinic N (around 398.8 eV), Ni–N and Cu–N (around 399.7 eV), pyrrolic N (around 400.0 eV), graphitic N (around 402.2 eV), and oxidized N (around 404.9 eV). This observation suggests the presence of significant defects in the material and that the metallic elements are anchored by N.³⁷ This evidence confirms the presence of NiN₄ and CuN₄ species within the catalyst. A combined analysis was undertaken, incorporating scanning electron microscopy, inductively coupled plasma emission spectroscopy, and transmission electron microscopy images. This composite analysis indicated the synthesis of the target product.

3.2 CO₂ electroreduction performance

The CO₂ electroreduction properties of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄ were investigated using a three-electrode flow cell (Fig. 1c) within 1 M potassium hydroxide solution. Initially, the electrocatalysts were submitted to linear sweep voltammetry (LSV) under a N₂ or CO₂ atmosphere (Fig. S3), where all three catalysts had much higher current densities than the hydrogen evolution reaction (HER). Fig. 4a compares the current densities of the three catalysts under a CO₂ atmosphere, with NiN₄@CuN₄ having a higher current density of approximately 120 mA cm⁻² at −1.5 V vs. RHE and FeN₄@CuN₄ having a more positive onset potential.

Fig. 4 (a) LSV curves of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄ in a flow cell, (b) FE of products NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄ in the flow cell, (c) the ratios of FE of C₂₊ and C₁ products on NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄ in the flow cell, (d) FE of C₂₊ products for NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄ in the flow cell at varying positive potentials, (e) stability and FE_C2+ of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄ in the CO₂RR at −1.6 V in the flow cell, and (f) comparison of the FE_C2+ value of NiN₄@CuN₄ with other published high-performance Cu-based electrocatalysts.

Subsequently, we investigated the CO₂ catalytic performance of the catalysts by performing constant potential tests at −1.3 to −1.7 V vs. RHE. The it curves are shown in Fig. S4. The identification of the products was achieved through the utilization of gas chromatography and ¹H NMR (Fig. 4b and S5). High electron transfer products CH₃CH₂CH₂OH, C₂H₄, CH₃COOH and CH₃CH₂OH were obtained on NiN₄@CuN₄ with the highest Faraday efficiency (81.0% at −1.6 V vs. RHE) with 24.4% CH₃CH₂CH₂OH and 36.6% CH₃CH₂OH. C₂H₄, CH₃COOH and C₂H₅OH could also be obtained on CoN₄@CuN₄, which has a Faraday efficiency of 50.4% for C₂₊ with 20.8% CH₃CH₂OH and 13.7% CH₃COOH. FeN₄@CuN₄ is less active for C₂₊ products and the products are mainly C₁, which is consistent with the lower onset voltage. CH₄ and HCOOH are the main products, with 45.8% CH₄ and 30.8% HCOOH. This may be because FeN₄@CuN₄ produces less CO, which makes the reaction slower and favorable for C₁ generation. Fig. 4c displays the C₂/C₁ ratio of electrocatalytic products for the three catalysts, indicating that the tandem catalytic action of the inner and outer layers in NiN₄@CuN₄ promotes C–C bonding processes, whereas CoN₄@CuN₄ and FeN₄@CuN₄ exhibit distinct behaviors. The C₂/C₁ ratio for NiN₄@CuN₄ increased gradually with voltage, indicating that −1.3 to −1.7 V vs. RHE is more favorable for NiN₄@CuN₄ to produce C₂. The product ratios of the C₂/C₁ class of products in Fig. 4c demonstrate an enhanced C–C coupling process via a yolk–shell structure on NiN₄@CuN₄, in contrast to CoN₄@CuN₄ and FeN₄@CuN₄. This may be attributed to the smaller BET and fewer active sites of CoN₄@CuN₄. The main product of FeN₄@CuN₄ is C₁, and the value of C₂/C₁ is almost 0. The C₂₊ Faraday efficiencies of the three catalysts at different voltages showed a volcano distribution shown in Fig. 4d, in which NiN₄@CuN₄ and CoN₄@CuN₄ reached maxima, 81.0% and 50.4%, respectively, at −1.6 V.

Finally, a long-term electrocatalytic stability test was performed on each of the electrocatalysts at −1.6 V (Fig. 4e). After 20 hours, the current density did not decrease significantly, with NiN₄@CuN₄ exceeding 100 mA cm⁻², making it suitable for industrial catalytic applications.

In conclusion, the catalyst NiN₄@CuN₄ exhibited excellent CO₂RR catalytic performance, especially up to 24.4% of CH₃CH₂CH₂OH. The catalysts in this work outperformed most reported catalysts (Fig. 4f),^38–44 as shown in Table S1.

3.3 Mechanism analysis

The reaction mechanism of the synthesized electrocatalyst was obtained through electrocatalytic kinetic analysis. As shown in Fig. 5a and S6, cyclic voltammetry was executed on NiN₄@CuN₄, CoN₄@CuN₄, and FeN₄@CuN₄, and their bilayer capacitance values (C_dl) were 25.76, 4.73, and 26.73 mF cm⁻², indicating higher electrochemical active surface area (ECSA) for NiN₄@CuN₄ and FeN₄@CuN₄. Furthermore, electrochemical impedance spectroscopy (EIS) analysis demonstrated that NiN₄@CuN₄ exhibited superior conductivity as shown in Fig. 5b, thereby potentiating electron transfer at the electrolyte interface. For further exploring generation pathways of CH₃CH₂OH and CH₃CH₂CH₂OH, we plotted current density maps (Fig. 5c) and Tafel curves (Fig. 5d) to reveal their kinetic properties. CH₃CH₂OH's part current density on NiN₄@CuN₄ is −35.87 mA cm⁻² (larger than CH₃CH₂CH₂OH's 23.91 mA cm⁻²) at −1.6 V, its Tafel slope is 672.94 mV dec⁻¹ (smaller than CH₃CH₂CH₂OH's 720.57 mV dec⁻¹), and the data suggest a slower C–C coupling step for CH₃CH₂CH₂OH.⁴⁵

Fig. 5 (a) Electrochemical double-layer capacitance of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄; (b) electrochemical impedance spectra of NiN₄@CuN₄, CoN₄@CuN₄ and FeN₄@CuN₄; (c) C₂H₅OH and C₃H₇OH part current densities on NiN₄@CuN₄ and CoN₄@CuN₄; and (d) Tafel plots of C₂H₅OH and C₃H₇OH for NiN₄@CuN₄ and CoN₄@CuN₄.

To further analyze the CO₂RR mechanism of NiN₄@CuN₄, the evolution of surface intermediates over time (0–40 min) at −1.6 V was investigated employing in situ Raman spectra. With the extension of the catalytic time, the characteristic peak of NiN₄@CuN₄ disappears (Fig. 6), and the characteristic band of Cu₂O emerges at 590 cm⁻¹.⁴⁶ Subsequent to the restoration of the voltage to the open-circuit voltage, the aforementioned Cu₂O band disappears, thus indicating the consequences of high potentials on the catalyst reconfiguration. The –C–OH stretching pattern of the C₂ intermediate (*HOCCOH), observed at 1220 cm⁻¹, provides definitive evidence of C₂₊ generation and validates the easy C–C coupling of NiN₄@CuN₄.⁴⁷ The peaks observed at 1308 and 1589 cm⁻¹ are associated with surface-bound *CO₂⁻ intermediates,⁴⁸ suggesting CO₂ activation via *CO₂⁻ to *CO.

Fig. 6 (a) Carbon dioxide electroreduction at −1.6 V was studied using in situ Raman spectroscopy on NiN₄@CuN₄ (in situ Raman measurements were taken in 0.1 M KHCO₃ aqueous solution full of CO₂), (b) carbon dioxide electroreduction from OCP to −1.8 V was studied using in situ FTIR spectroscopy on NiN₄@CuN₄ (in situ FTIR measurements were taken in 1 M KHCO₃ aqueous solution full of CO₂), (c) Gibbs free energy diagrams for the hydrogenation of CO₂ to CO on NiN₄ and CuN₄ sites, (d) Gibbs free energy diagrams for the hydrogenation of *CO to *CH₃CH₂OH and *OCHCH₂CHO on CuN₄, (e) Cu Auger spectrum and (f) the Cu 2p spectrum of CuN₄ before catalysis, (g) Cu auger spectrum and (h) Cu 2p spectrum of NiN₄@CuN₄ after catalysis.

At high potentials, we identified the adsorption bands of *CO (2097–2197 cm⁻¹)⁴⁹ and found that high *CO coverage is the basis for C–C coupling. Furthermore, C–H signals were observed on the NiN₄@CuN₄ surface in the high wave number region, from 2700 to 3200 cm⁻¹.⁵⁰ However, due to the complexity of the C–H bonds, reliable conclusions could not be drawn.

To investigate the electrocatalytic mechanism of CO₂ conversion on NiN₄@CuN₄ further, we performed density functional theory (DFT) calculations (Fig. 6c and d) and in situ infrared spectroscopy measurements (Fig. 6b). First, we calculated the pathways for CO₂ hydrogenation to *CO at the NiN₄ and CuN₄ sites. As illustrated in Fig. 6c, the energy barrier for CO₂ conversion to *CO is lower at the CuN₄ site, whereas the barrier for *CO desorption to CO is lower at the NiN₄ site. This suggests that both sites are active for CO₂ activation to CO, but that the CO generated at the NiN₄ site can diffuse to the outer layer and be adsorbed by the CuN₄ site. This result correlates with the continuous depletion of the characteristic peak of gas-phase CO₂ (2346 cm⁻¹) observed in situ by infrared spectroscopy. Concurrently, two peaks appear successively at ∼2050 cm⁻¹ (first) and ∼2113 cm⁻¹, which are assigned to CO adsorbed atop the Cu and Ni sites, respectively. Subsequently, a bridged CO signal (∼1900 cm⁻¹) emerges and intensifies, creating conditions for C–C coupling.⁵¹

Secondly, we calculated the pathway at the CuN₄ site, in which *CO and *CO dimerise to form *OCCO. This is then hydrogenated to produce the vital oxygen-containing C₂ intermediate, *CH₂CHO. The appearance of a characteristic peak at ∼1276 cm⁻¹ in the infrared spectrum can be assigned to *OCCOH. Subsequently, the shift of the ∼1693 cm⁻¹ peak (C=O stretching vibration) clearly indicates the formation of oxygen-containing C₂ intermediates, such as *CH₂CHO. Simultaneously, the evolution of the C–H bending vibration (from ∼1391 to ∼1472 cm⁻¹) was also observed.⁵²

Then, DFT calculations were performed to determine the energy barriers for the hydrogenation of CH₂CHO to form CH₃CH₂OH and for C₁ + C₂ coupling with CO molecules to form OCHCH₂CHO. The results are consistent with previous literature reports and indicate that the energy barrier is *OCHCH₂CHO > *CH₃CH₂OH. This aligns with the observed product yields of 28.6% n-propanol and 36.6% ethanol. This is attributed to CO overflow from the inner NiN₄ sites forming highly covered *CO on the CuN₄ sites, thereby favoring the formation of n-propanol.⁵³ The electrocatalytic products of the physical mixture of NiN₄ and CuN₄ (NiN₄/CuN₄) (Fig. S8) are solely C₁ and C₂, which further corroborates this. In situ infrared fingerprint region data reveal that, following the appearance of the CH₂CHO C=O characteristic peak (at ∼1693 cm⁻¹), the double peak in the characteristic fingerprint region representing carbon chain elongation (at ∼1161 and 1084 cm⁻¹) appears and intensifies simultaneously, ultimately leading to a significant increase in the intensity of the methyl characteristic peak (at ∼2969 cm⁻¹). This provides complete spectral evidence of the transition from the C₂ intermediate to the methyl-containing C₃ product.

In summary, theoretical calculations and in situ spectroscopic dynamics corroborate each other, jointly indicating a tandem catalytic pathway. CO₂ is efficiently activated to CO at the inner-layer NiN₄ sites. The CO then diffuses and becomes enriched at the outer-layer CuN₄ sites, forming a bridged adsorption configuration that triggers C–C coupling. This generates an oxygen-containing C₂ intermediate, which then undergoes C₁ + C₂ coupling with a second *CO molecule. This enables carbon chain elongation and ultimately hydrogenation, yielding the target product n-propanol.

To further investigate the deactivation mechanism of NiN₄@CuN₄, we conducted XPS and Cu LMM measurements on NiN₄@CuN₄ before catalysis and after 20 h of catalysis. Cu LMM was fitted using the least squares method. As shown in Fig. 6e–h, before catalysis, Cu²⁺ was the predominant species, with an area ratio to Cu⁺ of 4.61. A distinct Cu²⁺ satellite peak (962.39 eV) was observed in the Cu 2p spectrum, consistent with the fitting results. After 20 h of catalysis, Cu²⁺ disappeared, and Cu⁺ became the predominant species. Simultaneously, no Cu²⁺ satellite peak (962.39 eV) was detected in the Cu 2p spectrum. In summary, after prolonged electrocatalytic operation, a portion of the Cu²⁺ active sites irreversibly reduced to Cu⁺ species. This valence change, accompanied by alterations in the local electronic structure and coordination environment of the sites, weakened their adsorption capacity for the key C₁ intermediate (CO).⁵⁴ Consequently, C–C coupling significantly decreased, consistent with the sustained decline in faradaic efficiency for the C₂₊ product. Therefore, the irreversible transformation of active sites from highly active Cu²⁺ into less active Cu⁺ species is the core reason for the performance decay of NiN₄@CuN₄ in electrocatalysis.

4 Conclusions

In summary, a general strategy was developed for synthesizing Janus catalysts (MN₄@CuN₄) with double-layered ZIF-8 as the backbone structure and CuN₄-wrapped MN₄. The inner layer of MN₄ facilitates the release of a high concentration of CO, while the outer layer of CuN₄ promotes carbon–carbon coupling in the CO reduction reaction, thereby achieving tandem catalysis at the nanoscale. Furthermore, the integration of physical characterization, electrocatalytic kinetic analysis and DFT calculation facilitated the observation of the dynamic stabilization signal of *CO_atop, along with the key signal of *CO_bridge, *OCCOH, during the CO₂RR process. This enabled the efficient synthesis of the C₂₊ product. Among them, the NiN₄@CuN₄ catalyst demonstrated exceptional performance in the CO₂RR to C₂₊ products in 1 M KOH, exhibiting an excellent FE of 81.0% for the C₂₊ products at a voltage of −1.6, including a noteworthy FE of 24.4% for CH₃CH₂CH₂OH. The findings of this study offer novel insights into the design of innovative catalysts and the optimization of CO₂RR performance, particularly in the synthesis of propanol.

Author contributions

Gege Zhang: methodology, conceptualization, formal analysis, data curation, and writing – original draft. Qianyun Tan: visualization, and writing – review. Xiaoyu Xu: visualization, and writing – review. Faqian Liu: supervision, funding acquisition, project administration, visualization, and writing review.

Conflicts of interest

There are no conflicts to declare.

Data availability

All data supporting this study are included in the article and its supplementary information (SI). Supplementary information: characterization, computational methods, synthetic method of NiN₄/CuN₄, X-ray photoelectron spectroscopy, electrochemical data, metal contents data and comparison with other Cu-based catalysts. See DOI: https://doi.org/10.1039/d5ta10322c.

Acknowledgements

We acknowledge the support from the National Natural Science Foundation of China (52073311) and Guangdong Basic and Applied Basic Research Foundation (2023A1515240024 and 2025A1515012908).

References

1. J. W. Ager and A. A. Lapkin, Science, 2018, 360, 707–708.
2. Y. Y. Birdja, E. Pérez-Gallent, M. C. Figueiredo, A. J. Göttle, F. Calle-Vallejo and M. T. M. Koper, Nat. Energy, 2019, 4, 732–745.
3. X. Chen, S. Jia, J. Zhai, J. Jiao, M. Dong, C. Xue, T. Deng, H. Cheng, Z. Xia, C. Chen, X. Xing, J. Zeng, H. Wu, M. He and B. Han, Nat. Commun., 2024, 15, 7691.
4. E. W. Lees, B. A. W. Mowbray, F. G. L. Parlane and C. P. Berlinguette, Nat. Rev. Mater., 2021, 7, 55–64.
5. H. Tian, J.-T. Yang, X. Wang, H. Jiao, Z.-F. Gao, K.-Y. Zhu, Q. He and Z.-L. Wang, Appl. Catal. B Environ. Energy, 2025, 375, 125411.
6. J.-Y. Chen, J.-T. Yang, Y.-S. Han, Y.-Q. Huang, N.-N. Tian, J.-H. Li and Z.-L. Wang, ACS Catal., 2025, 15, 14882–14894.
7. M. Jouny, W. Luc and F. Jiao, Ind. Eng. Chem. Res., 2018, 57, 2165–2177.
8. X. Su, Z. Jiang, J. Zhou, H. Liu, D. Zhou, H. Shang, X. Ni, Z. Peng, F. Yang, W. Chen, Z. Qi, D. Wang and Y. Wang, Nat. Commun., 2022, 13, 1322.
9. W. Liu, P. Zhai, A. Li, B. Wei, K. Si, Y. Wei, X. Wang, G. Zhu, Q. Chen, X. Gu, R. Zhang, W. Zhou and Y. Gong, Nat. Commun., 2022, 13, 1877.
10. Y. Hori, H. Wakebe, T. Tsukamoto and O. Koga, Electrochim. Acta, 1994, 39, 1833–1839.
11. S. Popović, M. Smiljanić, P. Jovanovič, J. Vavra, R. Buonsanti and N. Hodnik, Angew. Chem., Int. Ed., 2020, 59, 14736–14746.
12. L. Bian, Y. Bai, J.-Y. Chen, H.-K. Guo, S. Liu, H. Tian, N. Tian and Z.-L. Wang, ACS Nano, 2025, 19, 9304–9316.
13. W. Zheng, X. Yang, Z. Li, B. Yang, Q. Zhang, L. Lei and Y. Hou, Angew. Chem., Int. Ed., 2023, 62, e202307283.
14. Q. He and T.-T. Li, Chem.–Eur. J., 2025, 31, e202403297.
15. T. Tang, Z. Wang and J. Guan, Exploration, 2023, 3, 20230011.
16. Z. Zhang, Z. Yang, L. Liu, Y. Wang and S. Kawi, Adv. Energy Mater., 2023, 13, 2301852.
17. Q. Yang, H. Liu, Y. Lin, D. Su, Y. Tang and L. Chen, Adv. Mater., 2024, 36, 2310912.
18. X.-F. Yang, A. Wang, B. Qiao, J. Li, J. Liu and T. Zhang, Acc. Chem. Res., 2013, 46, 1740–1748.
19. P. Song, M. Luo, X. Liu, W. Xing, W. Xu, Z. Jiang and L. Gu, Adv. Funct. Mater., 2017, 27, 1700802.
20. C. Cao, S. Zhou, S. Zuo, H. Zhang, B. Chen, J. Huang, X.-T. Wu, Q. Xu and Q.-L. Zhu, Research, 2023, 6, 0079.
21. S. Chen, X. Li, C.-W. Kao, T. Luo, K. Chen, J. Fu, C. Ma, H. Li, M. Li, T.-S. Chan and M. Liu, Angew. Chem., Int. Ed., 2022, 61, e202206233.
22. L. Lin, H. Li, Y. Wang, H. Li, P. Wei, B. Nan, R. Si, G. Wang and X. Bao, Angew. Chem., Int. Ed., 2021, 60, 26582–26586.
23. K. Li, Y. Kuwahara, K. Chida, T. Yoshii, H. Nishihara and H. Yamashita, Chem. Eng. J., 2024, 488, 150952.
24. J. Pei, H. Shang, J. Mao, Z. Chen, R. Sui, X. Zhang, D. Zhou, Y. Wang, F. Zhang, W. Zhu, T. Wang, W. Chen and Z. Zhuang, Nat. Commun., 2024, 15, 416.
25. C. Wang, H. Ren, Z. Wang, Q. Guan, Y. Liu and W. Li, Appl. Catal. B Environ., 2022, 304, 120958.
26. K. Mou, Z. Chen, X. Zhang, M. Jiao, X. Zhang, X. Ge, W. Zhang and L. Liu, Small, 2019, 15, 1903668.
27. W. Ren, X. Tan, C. Jia, A. Krammer, Q. Sun, J. Qu, S. C. Smith, A. Schueler, X. Hu and C. Zhao, Angew. Chem., Int. Ed., 2022, 61, e202203335.
28. W. Sun, S. Liu, H. Sun, H. Hu, J. Li, L. Wei, Z. Tian, Q. Chen, J. Su and L. Chen, Adv. Energy Mater., 2025, 2500283.
29. F. Chang, C. M. Vis, M. Bergmeijer, S. C. Howes and P. C. A. Bruijnincx, ChemSusChem, 2021, 14, 5328–5335.
30. T. Zhang, J. C. Bui, Z. Li, A. T. Bell, A. Z. Weber and J. Wu, Nat. Catal., 2022, 5, 202–211.
31. H. Chen, K. Shen, Y. Tan and Y. Li, ACS Nano, 2019, 13, 7800–7810.
32. J. Huang, X. Xiao, S. Xiong, J. Wan and C. Guo, J. Alloys Compd., 2022, 901, 163636.
33. R. I. Jibrael and M. K. A. Mohammed, Optik, 2016, 127, 6384–6389.
34. B. Das, R. Kundu and S. Chakravarty, Mater. Chem. Phys., 2022, 290, 126597.
35. A. Kumar, S. Patil, A. Joshi, V. Bhoraskar, S. Datar and P. Alegaonkar, Appl. Surf. Sci., 2013, 271, 86–92.
36. A. Guan, Z. Chen, Y. Quan, C. Peng, Z. Wang, T.-K. Sham, C. Yang, Y. Ji, L. Qian, X. Xu and G. Zheng, ACS Energy Lett., 2020, 5, 1044–1053.
37. J. Ma, L. Huang, K. Chen, J. Wang, X. Kang and X. Cao, J. Colloid Interface Sci., 2023, 652, 1734–1742.
38. B. A, X. Jin, M. Wang, Y. Wang, W. Chen, Z. Wei, Z. Du, X. Liu, Y. Wang and L. Zhang, Chem. Eng. J., 2024, 500, 157076.
39. Z. Cai, N. Cao, F. Zhang, X. Lv, K. Wang, Y. He, Y. Shi, H. Bin Wu and P. Xie, Appl. Catal. B Environ., 2023, 325, 122310.
40. C. Guo, Y. Guo, Y. Shi, X. Lan, Y. Wang, Y. Yu and B. Zhang, Angew. Chem., Int. Ed., 2022, 61, e202205909.
41. Z. Guo, H. Zhu, Z. Yan, L. Lei, D. Wang, Z. Xi, Y. Lian, J. Yu, K. L. Fow, H. Do, J. D. Hirst, T. Wu and M. Xu, Appl. Catal. B Environ. Energy, 2025, 364, 124839.
42. X. Jiang, Y. Zhao, Y. Kong, J. Sun, S. Feng, Q. Hu, H. Yang and C. He, Chin. J. Catal., 2024, 58, 216–225.
43. J. Zhang, C. Guo, S. Fang, X. Zhao, L. Li, H. Jiang, Z. Liu, Z. Fan, W. Xu, J. Xiao and M. Zhong, Nat. Commun., 2023, 14, 1298.
44. N. Zhang and Y. Zhang, Chem. Eng. J., 2024, 499, 156694.
45. S. Hagos, Results Chem., 2023, 5, 100952.
46. M. Valvo, J. Thyr and T. Edvinsson, ChemElectroChem, 2023, 10, e202300376.
47. C. Long, X. Liu, K. Wan, Y. Jiang, P. An, C. Yang, G. Wu, W. Wang, J. Guo, L. Li, K. Pang, Q. Li, C. Cui, S. Liu, T. Tan and Z. Tang, Sci. Adv., 2023, 9, eadi6119.
48. W. Wu, K. Bhattacharyya, K. Gray and E. Weitz, J. Phys. Chem. C, 2013, 117, 20643–20655.
49. X. Kong, J. Zhao, J. Ke, C. Wang, S. Li, R. Si, B. Liu, J. Zeng and Z. Geng, Nano Lett., 2022, 22, 3801–3808.
50. J. Gao, H. Zhang, X. Guo, J. Luo, S. M. Zakeeruddin, D. Ren and M. Grätzel, J. Am. Chem. Soc., 2019, 141, 18704–18714.
51. M. Moradzaman and G. Mul, ACS Catal., 2020, 10, 8049–8057.
52. D. Zhong, Q. Fang, R. Du, Y. Jin, C. Peng, D. Cheng, T. Li, T. Zhao, S. Zhang, Y. Zheng, Q. Zhao, Y. Sun and J. Li, Angew. Chem., Int. Ed., 2025, 64, e202501773.
53. S. J. Raaijman, M. P. Schellekens, P. J. Corbett and M. T. M. Koper, Angew. Chem., Int. Ed., 2021, 60, 21732–21736.
54. J. Liu, P. Li, J. Bi, S. Jia, Y. Wang, X. Kang, X. Sun, Q. Zhu and B. Han, J. Am. Chem. Soc., 2023, 145, 23037–23047.

---