Exploring the ternary interactions in Cu–ZnO–ZrO2 catalysts for efficient CO2 hydrogenation to methanol

The synergistic interaction among different components in complex catalysts is one of the crucial factors in determining catalytic performance. Here we report the interactions among the three components in controlling the catalytic performance of Cu–ZnO–ZrO2 (CZZ) catalyst for CO2 hydrogenation to methanol. The in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) measurements under the activity test pressure (3 MPa) reveal that the CO2 hydrogenation to methanol on the CZZ catalysts follows the formate pathway. Density functional theory (DFT) calculations agree with the in situ DRIFTS measurements, showing that the ZnO–ZrO2 interfaces are the active sites for CO2 adsorption and conversion, while the presence of metallic Cu is also necessary to facilitate H2 dissociation and to provide hydrogen resource. The combined experiment and DFT results reveal that tuning the interaction between ZnO and ZrO2 can be considered as another important factor for designing high performance catalysts for methanol generation from CO2.


 Response to Reviewers
We thank the reviewers for their thoughtful comments, which are valuable in improving the quality of our manuscript. As described below, we have made a detailed explanation and performed additional experiments to address all the comments.

 Reviewer #1
The article deals with a very "hot" topic -valorization of CO 2 into methanol. The catalysts of this reaction are being constantly improved, the mechanism is under investigation, many works are connected to the possible implementation and its industrial potential.
Even if the most active catalysts are known, the mechanism and the action of catalyst's parts is under questions. The authors claim the mechanistic study of the ternary effect between catalyst's components (Cu, Zn and Zr oxides). The authors insist in the title on the synergy between the 3 oxides, they also insists on the synergy in CO 2 adsorption, they say that the synergy effect is the key of CO 2 conversion and MeOH selectivity. To explain the nature of those synergetic actions this only postulate is used: "The synergy among Cu, ZnO and ZrO 2 can be ascribed to that the ZnO-ZrO 2 interface is responsible for the adsorption and activation of CO 2 , and the Cu related species contribute to the dissociative adsorption of hydrogen..." In my opinion this postulate says about two different actions, no real prove of "synergy" or "ternary effect" that the authors insist so much in the title and in the body of tha article.
Despite the general quality of the results on quite high level I am afraid that the article in this shape is imprecise and not suitable for publication.

[Response]
The conclusion on the ternary synergistic action among Cu, ZnO and ZrO 2 are based on the combined results of catalytic activity, in-situ DRIFTS experiments and DFT calculations. As shown in Figure R1a, no methanol is detected over the ZnO-ZrO 2 catalyst for CO 2 hydrogenation, but both Cu-ZnO and Cu-ZrO 2 catalysts show good activity for this reaction. This indicates that the Cu/ZnO and Cu/ZrO 2 interactions are crucial for the methanol synthesis. On the other hand, the Cu-ZnO-ZrO 2 ternary catalyst exhibits much higher methanol yield than either Cu-ZnO or Cu-ZrO 2 even though it shows lower surface area of Cu (S Cu ) than the Cu/ZnO catalyst (see Figure R1b), suggesting that the ZnO-ZrO 2 interaction should also play an important role in the Cu-ZnO-ZrO 2 catalyst for CO 2 hydrogenation. The    (Figure R2d), the transformation of carbonate species to formate species is also detected, but they are furhter converted to methoxy speceis (2930,2821,1147,1046 cm -1 ) in the prsence of H 2 (which is not dected on ZnO-ZrO 2 ). These phenomena indicate that the ZnO-ZrO 2 may bind the formate intermediates and the presence of Cu could promote the further hydrogenation of formate speceis to methoxy speceis, which generally accepted as the last intermediate for methanol generation from CO or CO 2 hydrogenation [refs: Behrens et al. Science 2012, 336, 893;Graciani et al. Science 2014, 345, 546;Kattel et al. Science 2017, 357, 1296Kuld et al. Science 2016, 352, 969.]. Cu-ZnO-ZrO 2 systems in the designed conditions. In-situ DRIFT spectra over different catalysts at 493 K after switching feed gas from CO 2 (after introducing CO 2 into the reaction camber for 10 min) to H 2 with a flow rate of 40 mL/min under atmosphere pressure.
The DFT calculations also support the FRIFTS findings. As shown in Figure 5A of the manuscript, the reaction intermediates prefer the ZrO 2 -ZnO interface rather than either oxide alone. For instance, the CO 2 adsorption adopts a conformation with C bound with Zn and one of O anchored on Zr. Such configuration provides a stronger binding energy (-2.32 eV) than that on ZnO-Cu(111) (-0.13 eV), ZrO 2 -Cu(111) (-1.18 eV), ZrO 2 cluster on ZnO(110) (-1.95 eV) and ZnO(110) (-1.94 eV), indicating that the ZrO 2 /ZnO interface facilitates the activation and transformation of CO 2 , a key step for CO 2 activation.
Although the Cu component was not specifically considered in the DFT calculations, the H 2 dissociative at the ZnO-ZrO 2 interface is an endothermic process (ΔE = 0.47 eV), which is less favorable than that at the Cu-oxide interface (ΔE = formation of *H at the Cu-oxide interface under reaction conditions, which facilitates the subsequent hydrogenation processes by providing the surface *H species.  Figure R3 shows a schematic diagram for the differences in catalytic activity (the detail data of the catalytic activity can be found in Tables S1, S2 and S5) and reaction pathways among Cu-ZnO, Cu-ZrO 2 and Cu-ZnO-ZrO 2 catalysts for CO 2 hydrogenation to methanol. The reaction pathway for Cu-ZnO and Cu-ZrO 2 catalysts are based on the findings in related references. For the Cu-ZnO catalyst, the active sites of CO 2 hydrogenation to methanol are related to the Cu-ZnO species (Cu-ZnO interface or Cu-Zn alloy), while the hydrogenated ZnO is the active sites for CO 2 hydrogenation to CO [Tisseraud et al. J. Catal. 2015, 330, 533]. The presence of abundant isolated ZnO results in relatively low methanol selectivity. For the Cu-ZrO 2 catalyst, the oxygen vacancies of ZrO 2 could improve the Cu-ZrO 2 interaction and the CO 2 adsorption ability play a very important role for the conversion of CO 2 to methanol [Fisher et al. J. Catal., 1997, 172, 222;Pokrovski et al. Langmuir, 2001, 17, 4297;[Rhodes et al. J. Catal., 2005, 233, 198]. However, the oxygen vacancy concentration in ZrO 2 , especially for tetragonal ZrO 2 (t-ZrO 2 ), is relatively low, which should be responsible for the low CO 2 conversion. In the case of the Cu-ZnO-ZrO 2 catalyst, the strong ZnO-ZrO 2 interaction creates more oxygen vacancies (as revealed by the XPS characterization shown in Figure S17 in the Supplementary Information) that can enhance the CO 2 adsorption, contributing to the relatively high CO 2 conversion. In addition, the presence of ZrO 2 on ZnO reduces the surface proportion of exposed ZnO, which eliminates the active sites for the reduction of CO 2 to CO, improving the methanol selectivity.
Overall, the Cu-ZnO-ZrO 2 catalyst shows much higher activity for CO 2 hydrogenation to methanol than the binary catalyst systems, and the interplay among ZnO, ZrO 2 and Cu is essential to enable the high conversion of CO 2 and high selectivity toward methanol. The ZnO-ZrO 2 interaction contributes to the adsorption of CO 2 and binds the formate intermediate, and the interaction of Cu with the ZnO-ZrO 2 support provides the hydrogen source for the further step hydrogenations of intermediate species to methanol.

[Action]
Even though our results clearly show that the ternary Cu-ZnO-ZrO 2 catalyst demonstrates enhanced rate for methanol production over the corresponding binary catalysts, we understand the caution raised by the Reviewer. Based on the Reviewer's suggestion we have decided to put less emphasis on the term "ternary synergy". Accordingly, we have changed "ternary synergistic action" to "ternary interactions" in the title of the manuscript; we have also changed "ternary synergy" to "ternary interactions" in the Abstract and Introduction. Such changes would allow the readers to reach unbiased conclusions regarding the origin of the unique catalytic properties of the Cu-ZnO-ZrO 2 catalyst.
We have also rewritten the final discussion to emphasize on the strong interplay among Cu, ZnO and ZrO 2 in pages 21 and 22: "For the Cu/ZnO system, the Cu-ZnO interface or the Cu-Zn surface alloy is considered as the active sites for CO 2 hydrogenation to methanol 3,[5][6][7]24,28,29 . In the case of Cu/ZrO 2 , the Cu-ZrO 2 interface plays a very important role for methanol formation 10,26,27,33 . For both the binary catalysts, the catalytic activity is determined by the Cu-ZnO or Cu-ZrO 2 interaction that is closely related to the physicochemical features (e.g., Cu particle size and surface area of Cu) of Cu spices. As shown in the comparison of the catalytic activity of Cu/ZnO, Cu-ZrO 2 and Cu-ZnO-ZrO 2 in Figure   S13, the Cu-ZnO-ZrO 2 ternary catalyst exhibits much higher methanol yield than either Cu-ZnO or Cu-ZrO 2 even though it shows a lower surface area of Cu (S Cu ) than the Cu/ZnO catalyst, suggesting that the ZnO-ZrO 2 interaction should also play an important role in the Cu-ZnO-ZrO 2 catalyst for CO 2 hydrogenation. Combining the results of XPS ( Figure S17) and CO 2 -TPD (Figure 4d), it can be concluded that the ZnO-ZrO 2 interaction promotes the formation of oxygen vacancies, which should be the active sites for CO 2 adsorption. The in-situ DRIFTS (Figures 3 and S12) experiments reveal that the ZnO-ZrO 2 interface is crucial for the transformation of carbonate to formate during CO 2 hydrogenation. However, no surface methoxy, which is a crucial intermediate species for methanol synthesis, is detected on the ZnO-ZrO 2 catalyst (see Figures 4a-c), while it is abundant on the Cu-ZnO-ZrO 2 catalysts (see Figures S5 and S9). These results indicate that the presence of Cu is necessary for the formate hydrogenation to methoxy in methanol synthesis from CO 2 +H 2 . It is reasonable to propose that, in the Cu-ZnO-ZrO 2 system, the ZnO-ZrO 2 interaction contributes to the adsorption of CO 2 and binds the formate intermediate, and the interaction of Cu with the ZnO-ZrO 2 support is responsible for the dissociative adsorption of hydrogen and the subsequent hydrogenation of carbonaceous intermediate species (e.g., formate and methoxy) to methanol." Figure   S27 shows an illustration to emphasize on the role of Cu, ZnO and ZrO 2 in the ternary interaction and a full discussion is also provided.
In addition, Figures R1and R3 were added into the Supplementary Information as Figure S13 and S27, respectively. The related discussions were also added into the SI as follows: "As shown in Figure S13a, no methanol is detected over the ZnO-ZrO 2 catalyst for CO 2 hydrogenation, but both Cu-ZnO and Cu-ZrO 2 catalysts show good activity for this reaction. This indicates that the Cu/ZnO and Cu/ZrO 2 interactions are crucial for the methanol synthesis. On the other hand, the Cu-ZnO-ZrO 2 ternary catalyst exhibits much higher methanol yield than either Cu-ZnO or Cu-ZrO 2 even though it shows a lower surface area of Cu (S Cu ) than the Cu/ZnO catalyst (see Figure S13b), suggesting that the ZnO-ZrO 2 interaction should also play an important role in the Cu-ZnO-ZrO 2 catalyst for CO 2 hydrogenation." "For the Cu-ZnO catalyst, the active sites of CO 2 hydrogenation to methanol are related to the Cu-ZnO species (Cu-ZnO interface or Cu-Zn alloy), while the hydrogenated ZnO is the active sites for CO 2 hydrogenation to CO 17 . The presence of abundant isolated ZnO results in relatively low methanol selectivity. For the Cu-ZrO 2 catalyst, the oxygen vacancies of ZrO 2 play a very important role for the conversion of CO 2 to methanol 18-20 . However, the oxygen vacancy concentration in ZrO 2 , especially for tetragonal ZrO 2 (t-ZrO 2 ), is relatively low, which should be responsible for the low CO 2 conversion. In the case of the Cu-ZnO-ZrO 2 catalyst, the strong ZnO-ZrO 2 interaction creates more oxygen vacancies (as revealed by the XPS characterization shown in Figure S17) that can enhance the CO 2 adsorption, contributing to the relatively high CO 2 conversion. In addition, the presence of ZrO 2 on ZnO reduces the surface proportion of exposed ZnO, which can eliminate the active sites for the reduction of CO 2 to CO, improving the methanol selectivity"  Reviewer #2 The manuscript described the hydrogenation of CO 2 over 3DOM Cu-Zn-ZrO 2 catalyst and found that the catalyst exhibited the best catalytic performance, 18% conversion of CO 2 and the 80% selectivity to methanol. The reaction mechanism was investigated by using in-situ techniques at the reaction conditions. I read the revised manuscript and the corresponding letter and found that the author has revised the manuscript carefully according to the reviewers' commends. The result is interesting and I would like to recommend it to be published in this journal after minor revision. The related phenomenon can be found in Figures 3 and S5, which is performed by switching the CO 2 feed gas (after introducing CO 2 into the reaction camber for 10 min) to H 2 at 493 K and 0.1 MPa. Under this condition, the formed carbonate species were converted into the formate species without the presence of CO. However, the 80% selectivity of methanol is obtained in the activity testing of catalysts that is performed in the flow of CO 2 /H 2 mixture at 3.0 MPa. Under this condition, the apparent gaseous CO (2175 and 2115 cm -1 ) and methanol are observed in the in-situ DRIFTS experiment as shown in Figure 2b, which is consistent with the catalytic activity testing. This indicates that the reaction pressure is crucial for the kinetics of CO 2 hydrogenation. The reaction products (CO and methanol) of CO 2 hydrogenation are hardly detected by IR at low reaction pressure (0.1 MPa).

[Action]
We added the following sentence on page 10 to emphasize the effect of pressure on methanol production: "It is also noted that no CO intermediate is detected during the DRIFTS experiment under atmospheric pressure (see Figure S5), which is inconsistent with the formation of CO in the CO 2 hydrogenation at 3.0 MPa (see Figure 2). This reveals that the reaction pressure also afftects the production of CO from CO 2 ." Comment 2: In Fig3, the carbonate species decrease quickly over Cu-ZnO, Cu-ZrO 2 , no formate species was observed. Is the carbonate species converted into CO which was desorbed quickly?