1 Introduction
The excessive consumption of fossil fuels dramatically increases CO
2 concentration in the atmosphere and results in global warming and energy crisis. Developing and utilizing renewable energy sources to optimize the energy structure have received widespread attention [
1–
6]. Electrocatalytic or photocatalytic CO
2 reduction reactions (CO
2RR) can convert intermittent electricity or solar energy into available chemical energy [
7–
11], and accelerate the carbon cycle. CO
2RR to obtain C
2 products (ethylene [
12–
14] and ethanol [
15–
17], etc.) with higher energy density than C
1 products has become the focus in the energy catalysis field [
18,
19]
Copper-based (Cu-based) materials are the most widely reported CO
2 reduction catalysts to generate C
2 products [
19–
21], due to its unique surface structure [
22–
24] electronic structure [
25–
29] and intermediate adsorption ability [
30–
33]. Many researches focus on improving the Faradic efficiency (FE) of C
2 products by forming Cu-based alloys, such as Cu–Ag [
34,
35] and Cu–Zn [
36,
37] alloys. In addition, other non-copperbased alloys (Ni–Al [
38,
39] Ni-Ga [
39,
40] and Pd–Au [
41] alloy) have been reported to produce low yield of C
2 products, which indicates the possibility of non-copper-based catalysts producing C
2 products. According to the key characteristics of copper-based catalysts for C
2 products to design high-efficiency pseudo-copper catalysts is an important strategy to enhance the activity and selectivity of C
2 products. There are two criteria for evaluating pseudo-copper catalysts to generate C
2 products: (i) Copper-like surface structure and electronic structure. Cu(100) with unique surface structure sensitivity [
23,
24,
42] and electronic structure [
25–
29] can control the activity and selectivity toward C
2 products. (ii) Copper-like adsorption abilities of *CO
2, *CO and *H. These adsorption abilities are the essential descriptors for CO
2 activation [
43,
44] CO hydrogenation, C–C coupling [
31,
32], and hydrogen evolution reaction (HER) [
30,
33]. Based on above criteria, design of pseudo-copper catalysts with high-efficiency C
2 products is promising and significant.
Alloying is a feasible strategy for obtaining pseudo-copper catalysts [
45], because the alloying effect is an effective method to regulate the adsorption abilities, surface and electronic structure of metal-based catalysts [
38,
46]. The surface and electronic structure are primary consideration of component metals in pseudo-copper alloy catalysts. Nickel (Ni) and Zinc (Zn) are adjacent to Cu in the periodic table of elements. Ni has the face-centered cubic (fcc) lattice similar with Cu and its 3d states is adjustable [
29]. Zn exhibits excellent potential in adjusting electronic structure of alloy catalysts [
47]. Ni–Zn alloy can exhibit a similar electron/atom (e/a) ratio with Cu, which indicates that Ni–Zn alloy will have a stable copper-like lattice structure [
48]. In addition, the electronic structure and intermediate adsorption ability of Ni–Zn alloys can be tuned by interaction between Ni and Zn. Ni–Zn alloys with copper-like surface and electronic structure, and intermediate adsorption ability are potential to obtain C
2 products in CO
2RR.
In this work, NiZn(110) and NiZn3(001) are selected as the active facets due to their similar square geometric site structure to Cu(100). The CO2 formation and C–C coupling on NiZn(110) and NiZn3(001) are further investigated to analyze the feasibility of C2 products. The free energy profiles of C2 products formation are used to study the reaction mechanism of NiZn(110) and NiZn3(001). The calculated results indicate that the NiZn and NiZn3 alloys exhibit copper-like properties with these critical criteria. The energy barrier of C2 products formation on NiZn3(001) is even lower than Cu(100). Therefore, the NiZn and NiZn3 alloys are potential pseudo-copper catalysts for CO2 reduction to C2 products. This work proposes a feasible pseudo-copper catalyst and provides guidance to design high-efficiency catalysts for CO2RR to C2 or multi-carbon products.
2 Computational methods
The Vienna
Ab initio Simulation Package (VASP) was used to perform spin-polarized DFT [
49–
51]. The projector augmented wave (PAW) method was applied to describe the effective potential between the ion core and the electrons. The generalized gradient approximation (GGA) with Perdew–Burke–Ernzerhof (PBE) functional was adopted to calculate the exchange-correlation energy [
52]. We set the kinetic cutoff energy of the plane wave as 450 eV. The convergence criteria for force and energy were set as 0.02 eV·Å
−1 and 10
−5 eV per atom, respectively [
53]. We built the (3 × 3 × 4) surface supercells for all slab models, including NiZn(110), NiZn(101), NiZn
3(001), NiZn
3(100), Cu(100) Cu(111), Ni(100), Ni(111), and Zn(001). The corresponding structures are shown in Fig. S1. For the Brillouin zone integration, we used a Monkhorst–Pack mesh with 3 × 3 × 1 K-points. The computational details of the adsorption energy and Gibbs free energy are shown in the Supporting Information.
3 Results and discussion
3.1 Surface and electronic structures
Cu(100) is proven to be the most favorable surface orientation for C–C coupling [
22], which can be attributed to its square ensemble sites [
23,
24]. Like Cu, NiZn and NiZn
3 alloys also have a face-centered tetragonal (fct) structure. As shown in Fig. S1, the NiZn(110), NiZn(101), NiZn
3(001), NiZn
3(100), Ni(100) and Zn(001) exhibit the similar square geometric site structure to Cu(100). Furthermore, Table S1 compares the calculated lattice constants of these facets. The lattice constants of Cu(100) were calculated to be
a =
b = 11.04 Å, while the Ni(100) and Zn(001) were
a =
b = 10.53 Å and
a =
b = 11.25 Å, respectively. This structural difference makes a big difference in catalytic performance. Interestingly, the lattice constants of NiZn(110) (
a = 9.80 Å,
b = 11.59 Å) and NiZn
3(001) (
a =
b = 11.01 Å) are more closer to Cu(100) [Fig. 1(a)] due to the structure rearrangement of Zn on Ni in alloy phases. The basic site configuration of NiZn(110) and NiZn
3(001) can be same as that of Cu(100), which has a great influence on the whole catalytic C
2 formation reaction process.
As shown in Fig. 1(b), the d-band of Cu(100) mainly distributes below the Fermi level (
EF), and its top of the 3d states are located around –1.8 eV. The anti-bonding states are filled for Cu(100), which leads to Cu(100) has favorable adsorption for key intermediates [
28]. As for Ni(100), the d-band cross the
EF and part of the antibonding states is empty for Ni(100), which will lead to strong adsorption abilities of intermediates and not conducive to further activation of adsorbate [
29]. The d-band of Zn(001) mainly distributes at –8 eV, which results in too weak adsorption of intermediates. Both NiZn(110) and NiZn
3(001) exhibit copper-like electronic structure in the density of states (DOS) calculations. Moreover, the top of the 3d states of NiZn
3(001) is almost the same to Cu(100). These results suggest that the formation of the alloy with Zn efficiently bring the apparent depression of the 3d states of Ni near
EF, which makes the DOS of NiZn alloys similar to that of Cu.
The d-band position is closely related to the binding strength between intermediates and catalysts. In detail, the electronic interaction between the adsorbate and the 3d states lead to the formation of bonding and antibonding states [
29,
54]. The lower position (relative to the Fermi level) of the d-band, the weaker binding strength, due to the occupancy of anti-bonding states. Compared with Ni(100), the adjusted d-band positions of NiZn alloys indicate suitable adsorption properties of intermediates, similar to Cu(100), which is favorable to the further activation of intermediates (CO
2, CO), and avoiding CO poisoning or strong hydrogen evolution reactions (HER).
3.2 Adsorption abilities of *CO2, *CO and *H
Copper-like adsorption abilities of *CO
2, *CO and *H are regarded as the key criterion to evaluate the potential catalytic performance of pseudo-copper catalysts. The adsorption abilities of *CO
2, *CO and *H directly influence the CO formation [
43,
44] and C–C coupling [
31,
32,
55,
56], and HER [
30,
33], respectively. The adsorption energies of *CO
2, *CO and *H on different facets are shown in Table S2.
The CO2 reduction process is accompanied by the competitive HER and hydrogenation process. The relationship between the adsorption energy of *H (∆EH) and reactants and the relationship between ∆EH and crucial intermediate *CO (∆ECO) both are very important. As shown in Fig. 2(a), the upper right region indicates that both the and ∆EH are weak, not conducive to the CO2 reduction reaction. The lower left region indicates that ∆EH is much stronger than , which will lead to the active site being occupied by *H, and is not conducive to CO2 activation. In the circle area, NiZn(110) and NiZn3(001) exhibit copper-like *CO2 and *H adsorption energies. Proper and ∆EH indicate the suitable coordination of CO2 adsorption and CO2 hydrogenation process, which provides the possibility for the generation of intermediate CO. The adsorption abilities of *CO and *H are shown in Fig. 2(b). Similarly, the upper right region indicates that both the ∆ECO and ∆EH are weak, not conducive to the CO hydrogenation and C–C coupling reaction. The lower left region indicates that both ∆ECO and ∆EH are strong, which will lead to CO poisoning and severe competing HER. In the circle area, NiZn(110) and NiZn3(001) exhibit copper-like *CO and *H adsorption energies. Proper ∆ECO and ∆EH indicate the suitable coordination of CO hydrogenation and C–C coupling process, which provides the possibility for the generation of C2 products.
The Cu-like adsorption abilities of NiZn(110) and NiZn3(001) are mainly due to the fact that Zn adjusts the surface and electronic structure of Ni. By screening and evaluation, NiZn(110) and NiZn3(001) are the potential pseudo-copper catalyst to generation C2 products during CO2RR.
3.3 Reaction pathway of *CO formation
CO is the key intermediate for the formation of C
2 products, and the first intermediate for CO formation from CO
2 is *COOH [
44,
46] The reaction pathways of CO formation are analyzed on NiZn(110) and NiZn
3(001) [Fig. 3(a)]. The formation energy barriers of *CO (∆
G∗CO) on NiZn(110) and NiZn
3(001) are 0.09 eV and 0.21 eV, respectively, which both are much lower than Cu(100) (0.53 eV) (Table S3). The adsorptions of *COOH intermediate on NiZn(110) and NiZn
3(001) are much stronger than on Cu(100), which is the key factor of lower formation energy barriers of *CO on NiZn(110) and NiZn
3(001).
In addition, the competitive reaction of formate (*HCOOH) formation is also considered on NiZn(110) and NiZn3(001) [Fig. 3(b)]. The formation energy barriers of formate (∆G∗HCOOH) on NiZn(110) and NiZn3(001) are 0.62 eV and 0.53 eV (Table S4), which is higher than formation energy barriers of *CO. This result indicates that the formation of *CO on NiZn(110) and NiZn3(001) is easier than *HCOOH. As for Cu(100), the formation energy barriers of formate (∆G∗HCOOH) is 0.45 eV, which is slightly lower than ∆G∗CO of 0.53 eV. It suggests that NiZn(110) and NiZn3(001) have higher selectivity to *CO than Cu(100).
3.4 Reaction pathway of C–C coupling
C–C coupling is reported to be the crucial reaction for the C
2 formation [
57]. There are three main pathways for the C–C coupling reaction [
32] [Fig. 4(a)]: (i) *CO–CO intermediate formed via *CO coupling with CO
(g); (ii) *CHO is formed via *CO hydrogenation, and coupling with CO
(g) to form *CHO–CO intermediates; (iii) *CO hydrogenation to form *COH, then *COH hydrogenation to form *CH
2 intermediate, and the C–C bond formed via *CH
2 dimerization.
The formation energy barriers of three intermediates (*CHO, *COH and *CO–CO) during C–C coupling are calculated to study the C2 product evolution. As shown in Table S5, both the free energy barriers of *COH and *CO–CO are much higher than *CHO. Therefore, *CHO–CO is considered as the main intermediates for C–C coupling reaction on NiZn(110), NiZn3(001), and Cu(100)
In the *CHO–CO pathway, NiZn(110) and NiZn3(001) show similar reaction trends to Cu(100) [Fig. 4(b)]. Compared with Cu(100) and NiZn(110), NiZn3(001) has the lowest free energy barrier (0.60 eV) (Table S6) which can be attributed to the optimal Cu-like surface/electronic structure and intermediate adsorption abilities of NiZn3(001). For NiZn(110), although it has the lowest free energy barrier of *CO→*CHO (0.57 eV), the energy barrier of *CHO→*CHO–CO is higher due to the strong *CO adsorption abilities on NiZn(110), which is detrimental to the C–C bond formation.
3.5 Reaction mechanism of C2 products formation
To understand the reaction mechanism of C
2 products formation on NiZn(110) and NiZn
3(001), the C
2 intermediates in the reaction pathway from *CHO–CO to *CHO–CH are investigated. *CHO–CH is the crucial precursor of the C
2 product, and the *CHO–CH
2 intermediate formed by its hydrogenation is considered to be the bifurcation of the pathway for C
2H
4 and C
2H
5OH formation [
32,
58].
Three possible pathways of *CHO–CO hydrogenation on NiZn(110), NiZn3(001) and Cu(100) are considered: alternate (*CHO–CHO) (Fig. S2 and Table S7), hybrid (*CHO–COH) (Fig. S3 and Table S8) and distal (*CHOH-CO) (Fig. S4 and Table S9) pathways. The lowest free energies of *CHO–CH formation on NiZn3(001) and Cu(100) both are alternate pathway[Fig. 5(b) and Fig. S5], while that on NiZn(110) is hybrid pathways [Fig. 5(a)].
The main formation mechanism of C2H4 and C2H5OH are formed via the deoxygenation and hydrogenation of *CHO–CH2 intermediate, respectively. Figure 5(c) compares the free energies of the reaction pathway from CO2 to *CHO–CH2 on NiZn(110) (Fig. S6), NiZn3(001) (Fig. S7), and Cu(100) (Fig. S8). In detail, the step of *CO hydrogenation is the rate determination step (RDS) on NiZn3(001) and Cu(100), but the RDS on NiZn(110) is the *CHO–CO formation via *CHO and *CO coupling. Moreover, NiZn3(001) shows the lowest energy barrier (0.60 eV) for CO2RR to *CHO–CH2, compare with Cu(100) (0.70 eV) and NiZn(110) (0.82 eV). This calculation result is consistent with the screening criteria. NiZn3(001) is an ideal pseudo-copper catalyst for CO2 reduction to C2 products.
4 Conclusion
In summary, we designed the Ni–Zn alloys as the promising pseudo-copper catalysts and predicated reaction mechanism for CO2RR to C2. DFT calculations revealed that NiZn and NiZn3 had the two key criteria of pseudo-copper alloys: copper-like surface and electronic structure and intermediate adsorption abilities. NiZn3(001) alloy had lower free energy barrier (0.60 eV) than Cu(100) (0.70 eV) and NiZn(110) (0.82 eV) in the reaction from CO2 to *CHO–CH2, and the C–C bond was formed through the *CHO and CO coupling. The *CHO–CH2 intermediate was preferred formed along the alternate hydrogenation pathway. Importantly, the copper-like nature of NiZn and NiZn3 alloys were derived from the interaction of Ni and Zn, which tuned their lattice structure and 3d state toward Cu. Here, two critical criteria of the pseudo-copper catalyst are established to form C2 products from CO2, and provides guidance to design high-efficiency catalysts for CO2RR to C2 or multi-carbon products.