2.1. Morphology and structure analysis
As shown in
Figure 1, Al–Cu
2O–X catalyst was prepared by a simple one-step method (Experimental section for details). In order to further characterize Al–Cu
2O–X catalysts, XRD pattern was used to study the crystal structure of Cu
2O, Al–Cu
2O, and Al–Cu
2O–2. From the XRD pattern in
Figure 2a, the peaks at 29°, 36°, 42°, 61°, 73°, and 77° correspond to the (110), (111), (200), (220), (311) ,and (222) planes of Cu
2O, respectively, which agree well with the octahedral Cu
2O (PDF#75–1535). The XRD patterns of Al–Cu
2O–X (3-6) were shown in Figure S1. During the preparation process, the catalyst synthesized by adjusting the amount of Al
3+, the concentration of NaOH, and the reaction time. Al–Cu
2O–X (3-6) were all single-phase Cu
2O nanocrystals. The morphologies of the Cu
2O, Al–Cu
2O, and Al–Cu
2O–2 were monitored by scanning electron microscopy (SEM). Cu
2O nanocrystal without Al doping showed a octahedron shape with smooth surface (Figure S2a). Due to the doping effect of Al, the Al–Cu
2O nanocrystal presented a octahedral shape with a more rough surface and formed a defect structure (
Figure 2b), which may provide abundant active sites for CO
2RR [
33]. When the concentration of Al
3+ increased from 0.02 M to 0.03 M, Al–Cu
2O–2 catalyst exhibits a cube shape (Figure S2b). However, it was reported that the resulting cube Al–Cu
2O catalyst is not conducive to the formation of C
2H
4 [
23]. The better-performing Al-Cu
2O with a homo-octahedral shape was observed by TEM (
Figure 2c), which was consistent with SEM image. High resolution transmission electron microscopy (HRTEM) image in
Figure 2d presented that the lattice stripe spacing d marked was 0.304 nm, corresponding to the (110) crystal plane of Cu
2O. The HAADF-STEM image (
Figure 2e) also exhibited an octahedral shape. The composition of Al-Cu
2O was reconfirmed by elemental mapping (
Figure 2f). The Al (red), Cu (blue), and O (green) elements uniformly distributed over the Al-Cu
2O nanocrystals.
The surface composition and valence of Cu
2O and Al–Cu
2O nanoscrystals were characterized by X-ray photoelectron spectroscopy (XPS). As shown in
Figure 3a and b, four peaks were observed in Cu 2p spectrum for both Cu
2O and Al–Cu
2O samples. For Cu
2O, the peaks at 932.78 and 952.62 eV corresponded to the binding energies of Cu 2p
3/2 and Cu 2p
1/2 of Cu
2O or Cu, respectively. The binding energies at 935.28 eV and 944.48 eV were ascribed to the peaks of Cu
2+. For Al–Cu
2O, the peaks at 932.89 and 952.73 eV corresponded to the Cu 2p
3/2 and Cu2 p
1/2 of Cu
2O or Cu, respectively. The binding energy of 935.26 and 944.46 eV belonged to the peak of Cu
2+. The above results showed that the existence of Cu
0 may be due to the partial reduction of Cu
2O in the CO
2RR process [
34]. The existence of trace CuO may be mainly due to the oxidation of a small amount of Cu
2O catalyst to CuO in the air after the synthesis of Cu
2O [
35]. When octahedral Cu
2O nanocrystals were doped with Al, the peaks of Cu 2p
3/2, Cu 2p
1/2, and Cu
2+ of Cu
2O or Cu were shifted positively. This results may be attributed the introduction of Al, which can induce charge transfer from Al atoms to Cu atoms, thus modulating the electronic structure of Al–Cu
2O. The existence of Cu
2O was also confirmed in the O 1s XPS spectra of Cu
2O and Al-Cu
2O (figure 3c-d). There were three XPS peaks in both catalysts, of which the peak at 530.5 eV corresponds to the Cu-O bond, and 532.11 and 532.77 eV correspond to Olat and C=O, respectively [
36]. In the high-resolution spectrum of Al 2p (figure 3e), the peaks at 74.55 and 77.35 eV correspond to the Al 2p
1/2 and Al 2p
3/2 of metal Al, respectively. The Al atom was 0.41% by XPS analysis, indicating that the Al–Cu
2O catalyst has been successfully prepared.
2.2. Electrocatalytic CO2RR perfomances
To further analyze the electrochemical performance of the catalyst, the linear sweep voltammetry (LSV) of Cu
2O and Al–Cu
2O–X catalysts in saturated CO
2 electrolyte and saturated N
2 electrolyte were tested. The analysis of
Figure 4a shown that the current density of Al–Cu
2O catalyst in CO
2 saturated electrolyte is higher than that in N
2, indicating that Al–Cu
2O catalyst had higher activity. The LSV curve was measured in CO2-saturated 0.1 M KHCO
3 electrolyte (figure S3a). The current density of the Al–Cu
2O catalyst in CO
2 saturated atmosphere was higher than that of Cu
2O and Al–Cu
2O–2 catalysts, indicating that the Al–Cu
2O catalyst had better electrocatalytic activity for CO
2RR. Figure S3b shows the potentiostatic electrolysis of CO
2 at various potentials. The almost constant current signal indicates that the Al-Cu
2O catalyst exhibited good electrochemical stability during the CO
2RR process. In
Figure 4b, the formation rates of three kinds of catalysts were presented for ethylene products. The Al–Cu
2O catalyst has higher current density for ethylene formation than that of Cu
2O and the Al–Cu
2O–2 catalysts in a wide potential range, and reaches the partial current density of 16.7 mA cm
-2 at -1.38 V (vs. RHE). The above results showd that the Al–Cu
2O catalyst was more conducive to the production of ethylene as the main product and has a better inhibitory effect on compete hydrogen formation.
In order to determine the CO
2RR selectivity of the Al–Cu
2O catalyst, the reduction products were qualitatively and quantitatively analyzed. In this study, the reduction products of each catalyst were determined in the wide potential range from-0.98 V to-1.38 V (vs. RHE). The analysis of
Figure 4c and S4 shows that the products of Cu
2O ,and Al–Cu
2O–X catalysts were C
2H
4, HCOO
-, CO, CH
4 and by-product H
2. When Al
3+ was not introduced in the reaction, the octahedral Cu
2O nanocrystal catalyst was synthesized, and Figure S4a shows the FE of the catalyst to the CO
2RR product. The result shown that the catalyst had a good effect on inhibiting by product H
2 at low potential, and the highest FE
C2H4 was 26.1%. If an appropriate amount of Al
3+ was introduced into octahedral Cu
2O nanocrystal (0.02 M), the selectivity of Al–Cu
2O catalyst was improved. As shown in
Figure 4b, with the increase of catalyst potential, the FE value of H
2 decreases from 35.1% to 22.1%. On the contrary, the FE value of C
2H
4 increases from 12.9% at-0.98 V (vs. RHE) to 44.9% at-1.23 V (vs. RHE). The result shown that the catalyst had good selectivity for ethylene and good inhibition effect on HER. If more Al
3+ was added to the reaction (0.03 M), the FE of the prepared Al–Cu
2O–2 catalyst for the CO
2RR product was shown in Figure S4b. The result has shown that the FE of C
2H
4 was 32.8%, indicating that the catalyst had a good selectivity for ethylene. It was worth noting that we also studied the effects of reaction time (S4cpender d) and NaOH concentration (S4ePowerf) on the selectivity of the catalyst. The result showed that the FE of the four catalysts C
2H
4 was 40.1%, 38.6%, 41.0%,and 30.8%, respectively, indicating that the optimization of reaction time and NaOH concentration can make the catalyst have a certain selectivity, but compared with the quantity conditions of introducing Al
3+. The effect of improving the selectivity of the product is weaker, mainly because the addition of different amounts of Al
3+ will form different morphology of the catalyst, resulting in different selectivity of the catalyst to the product.
Figure 4d compares the selectivity of three kinds of catalysts Cu
2O, Al–Cu
2O, and Al–Cu
2O–2 for ethylene products. The result showed that under different applied voltages, the efficiency of the Al–Cu
2O catalyst for reducing CO
2 to C
2H
4 in 0.1 M KHCO
3 electrolyte was higher than that of the other two catalysts, indicating that the amount of Al
3+ introduced into the catalyst effected the selectivity of the catalyst. This may be due to the fact that Al -doped Cu
2O will cause changes in the electronic structure and the morphology of the catalyst, thus reducing the adsorption energy of the catalyst for ethylene intermediates in the CO
2RR process and enhancing the selectivity of the reaction to the products.
The electrochemical surface area (ECSA) is also a key point for the electrocatalyst. According to the formula for calculating ECSA, it is known that this parameter is related to the C
dl and C
ds values of their catalysts, because the catalysts are coated on hydrophobic carbon paper (model 060). Therefore, the Cds of the three catalysts are the same, and only the C
dl value of the catalyst can be calculated to determine the ECSA of the catalyst. According to the Cu
2O, Al–Cu
2O, and Al–Cu
2O–2 catalysts, the cyclic voltammograms (Figure S5a-c) of 0.47 V~0.57 V (vs. RHE) at different scanning rates (20, 40, 60, 80,100, 120 mV s
-1). It can be seen from
Figure 4e that the capacitance values of Cu
2O, Al–Cu
2O, and Al–Cu
2O–2 catalysts were 0.109, 0.122, and 0.076 mF cm
-2, respectively. The largest C
dl of the Al–Cu
2O electrocatalyst suggested that the high electrochemical activity surface area of the Al–Cu
2O–2 catalyst. This high ECSA can offer a lot of catalytic active sites for improving the electrocatalytic performance of CO
2RR, thus improving the selectivity of the catalyst, which was consistent with the previous research conclusion.
The impedance of several different catalysts under open-circuit voltage was obtained (Figure S6a). Compared with the octahedral Cu
2O catalyst without Al doping, the EIS arc of the Al–Cu
2O catalyst was smaller than that of the octahedral Al–Cu
2O catalyst. The results indicate that during the reaction process, interface charges can be rapidly transferred and catalytic activity can be improved. In order to better understand the activity and kinetics of Al-Cu
2O materials on CO
2RR, the Tafel slope analysis of the local current density of the catalyst product is carried out. As shown in
Figure 4f, the Tafel slope of the Al–Cu
2O catalyst (74.3 mV dec
-1) was lower than that of Cu
2O (85.9 mV dec
-1) and the Al–Cu
2O–2 (110.4 mV dec
-1), indicating that the electron transfer rate of the catalyst is faster, which was beneficial to the rapid adsorption and desorption of the important intermediate of ethylene from the surface of Al–Cu
2O catalyst. To speed up the reaction, therefore, the introduction of Al into octahedral Cu
2O catalyst may help to reduce the activation energy of various intermediates in the CO
2RR reaction process, thus making the catalyst show better selectivity and activity for the products.
We further studied the stability of the CO
2RR material. As seen in
Figure 5a, the Al–Cu
2O catalyst was electrolyzed at-1.23V (vs. RHE) for about 25000s. The analysis chart shown that the current density of the Al–Cu
2O catalyst can be kept stable and the FE of ethylene can be kept above 40% in the first 3600s. With the change in reaction time, the current density increases gradually. By observing the stable SEM of the Al–Cu
2O catalyst, we can see that there are some small pores on the surface of the catalyst, which may provide more active sites, thus increasing the current density. However, the selectivity of the catalyst to ethylene, the main product, began to decrease obviously after two hours of testing, probably because the catalyst was dripped on hydrophobic carbon paper during the test, which may lead to the shedding of the catalyst in the long-term electrolysis process, resulting in a decrease in the FE of the catalyst. At the same time, in the current research, the stability of copper-based catalysts is poor, in later research, other strategies need to be used to improve the stability of copper-based catalysts for a long time [
37]. Through the XRD spectrum after long-term electrolysis, it was found that the composition of the electrode was consistent with that before the reaction, indicating that the Al–Cu
2O showed good electrochemical stability (
Figure 5b) in the whole CO
2RR test. It was worth noting that after the electrolysis of Al–Cu
2O catalyst for 10min, 20min, 30min and 7h (Figure S7a-d), the result showed that the morphology of the octahedron remains unchanged. With the increase of electrolysis time, some small pores appear on the surface of the catalyst. The appearance of these pores may provide more active sites, resulting in an increase in current density in the electrolysis process, but a decrease in the FE of ethylene. The result shown that the increase of these active sites was not conducive to improving the selectivity of the main products, and the above results shown that the catalyst can maintain good stability under long-term electrolysis.
2.3. DFT Computations
In order to further explore the catalytic reaction mechanism, we used Density functional theory (DFT) to calculate, simulate and compare the CO
2RR reaction path on the surface of Al-Cu
2O and Cu
2O catalysts to further understand the path from CO
2 to C
2H
4.
Figure 6 shows the spatial structure (
Figure 6a) and energy distribution of Al–Cu
2O and Cu
2O.
Figure 6b shown the energy distribution of ethylene production and by-product H
2 of Cu
2O and Al–Cu
2O catalysts. The Gibbs free energies of each intermediate along ethylene on Cu
2O and Al-Cu
2O catalysts *CHCOH, *CCH, *CCH, *CCH
2, *CHCH
2 (intermediates for ethylene production) and *H (intermediates to H
2) have been calculated. Because the Gibbs free energy of Al–Cu
2O catalyst was lower than that of Cu
2O catalyst in each reaction step, the path of ethylene production of CO
2RR was easier to occur. It can be seen that the strategy of doping Al to octahedral Cu
2O was beneficial to improve the selectivity of product C
2H
4 [
38]. At the same time, the analysis of figure 6c shows that the Al-Cu
2O catalyst doped with Al enhances the adsorption of intermediate * H and was further away from the ideal hydrogen adsorption value (0 eV). It makes the competitive reaction of HER more disadvantageous, thus inhibiting the occurrence of side effects.To further analyze the potential reason for the selective improvement of this product, the density of states (DOS) of d orbitals on Cu
2O (001) and the Al–Cu
2O (001) surfaces before CHCOH adsorption is compared (
Figure 6d,e). Since the electronic states near the Fermi level are mainly contributed by the d electrons of Cu atoms, it is indicated that the reaction is mainly caused by the interaction between Cu and C, and the d band center of undoped Al octahedron Cu
2O (001) was -2.08 7eV. The d-2.027 eV of the Al-doped Al–Cu
2O (001) surface was closer to the Fermi level (0eV), and the d-band shifts upward on the Abscissa, which makes the center of the d-band closer to the Fermi level and increases the density of electronic states, which is beneficial to the adsorption of Cu atoms through d electrons and intermediates, thus promoting the CO
2RR process and improving the selectivity of the catalyst for C
2H
4 product.