An Insight into Synergistic Metal-Oxide Interaction in CO 2 Hydrogenation to Methanol over Cu/ZnO/ZrO 2

: The metal-oxide interaction is of signiﬁcance to the construction of active sites for Cu-catalyzed CO 2 hydrogenation to methanol. This study examines the effect of ZnO and ZrO 2 composition on the Cu/ZnO/ZrO 2 catalyst structure and surface properties to further tune the catalytic activity for methanol synthesis. The ZnO/ZrO 2 ratio can impact the CuZn alloy formation from strong Cu-ZnO interactions and the surface basic sites for CO 2 adsorption at the Cu-ZrO 2 interface. The proportional correlation of the CuZn alloy content and CO 2 desorption amount with the space-time yield (STY) of methanol reveals a synergistic interaction between Cu and oxides (ZnO and ZrO 2 ) that enhances methanol synthesis. The optimized Cu/ZnO/ZrO 2 catalyst exhibits higher STY relative to the traditional Cu/ZnO/Al 2 O 3 catalyst. The obtained results presented herein can provide insight into the catalyst design for methanol synthesis from CO 2 .


Introduction
The production process of methanol, a high-value-added chemical, has historically received widespread attention.As early as 1923, BASF Company in Germany took the lead in developing high-pressure methanol synthesis technology from syngas [1].Afterward, ICI discovered a low-pressure process for the synthesis of CH 3 OH using Cu-based catalysts [2].Recently, Bai et al. [3] proposed that CO 2 from the atmosphere and H 2 from renewable energy sources can be used for CO 2 hydrogenation to methanol, which can realize chemical utilization of CO 2 and storage of renewable energy.
The existing literature endeavors to design efficient active sites for Cu-based catalysts.However, there is no consensus on the active site structure due to the catalyst complexity.A synergistic effect between Cu and oxide (e.g., ZnO, ZrO 2 , and CeO 2 ) at the interface has been proposed as the active site [7].Another view is that CuZn alloy is an active site; its formation may promote the partial reduction of ZnO to Zn q+ or Zn 0 , thereby modifying the Cu surface [9,11,27].The dissociated H atom on Cu 0 might overflow to the CuZn alloy, and then react with the adsorbed species CO 2 * by ZnO near the CuZn alloy, so as to improve the catalyst performance in the CO 2 hydrogenation to methanol [28].Compared to syngas, CO 2 hydrogenation to methanol produces water, which requires a weakly hydrophilic material to timely desorb the generated water.In previous work [29], a highly dispersed Cu−ZrO 2 interface can effectively improve the conversion rate and methanol selectivity.Meanwhile, compared to Al 2 O 3 , ZrO 2 shows weaker hydrophilicity.In this work, we propose the design of a tertiary Cu-ZnO−ZrO 2 catalyst with the construction of Cu-ZnO and Cu−ZrO 2 interactions as active centers, which potentially promotes CO 2 hydrogenation to methanol.
In this study, Cu/ZnO, Cu/ZrO 2 , and Cu/ZnO/ZrO 2 with a series of mass ratios of ZnO/ZrO 2 were constructed with the structure and surface properties studied using a range of characterizations [e.g., XRD (powder X-ray diffraction), TEM (transmission electron microscopy), XPS (X-ray photoelectron spectra), and CO 2 -TPD (CO 2 temperatureprogrammed desorption)].The optimized Cu/ZnO/ZrO 2 catalyst outperforms the conventional Cu/ZnO/Al 2 O 3 catalyst in terms of activity for CO 2 hydrogenation to methanol.The presence of CuZn alloy and Cu−ZrO 2 interaction enhances CO 2 conversion and methanol formation rates.

Characterization
A series of Cu/ZnO/ZrO 2 catalysts with different ZnO/ZrO 2 mass ratios were synthesized via oxalate coprecipitation.The catalysts were named based on element content.For example, the ZnO/ZrO 2 ratio is 2:5 in the catalyst, which is named C3Z2Z5.The element contents measured using ICP-OES (inductively coupled plasma optical emission spectroscopy) are close to the target values [Table 1, including specific surface area (S BET ), specific surface area of Cu (S Cu ), Cu dispersion (D Cu ), Cu diameters (d Cu ), and other physicochemical properties] indicative of no significant loss of Cu during the catalyst preparation.In the adsorption and desorption isotherms (Figure S1A), all the catalysts show typical Type IV hysteresis loops, indicating that the prepared catalysts are mesoporous.The pore size distribution in Figure S1B also confirms the mesoporous structure.As the ZnO/ZrO 2 ratio increases, the specific surface area shows a decreasing trend, which is not conducive to the dispersion of Cu (Table 1).This indicates that ZrO 2 plays an important role in maintaining the specific surface area of the Cu/ZnO/ZrO 2 catalyst.In the Zr-containing catalysts, the change in ZnO/ZrO 2 ratio does not have a significant impact on the pore volume.Compared with Cu/ZrO 2 and Cu/ZnO/ZrO 2 , the pore volume of Cu/ZnO is significantly reduced.The pore size shows an increasing trend with the increase of the ZnO/ZrO 2 ratio.The crystalline structures of all the catalysts after calcination and reduction were analyzed by XRD (Figure 1).The monoclinic CuO (2θ = 35.5 • and 38.7 • , PDF#48-1548) was observed in all the calcined samples in Figure 1A, while ZnO (2θ = 31.8• , 34.4 • and 38.7 • , PDF#79-0206) [30,31] can only be detected when the Zn content exceeds 20 wt%.The ZrO 2 in all samples was a-ZrO 2 , because of the low calcination temperature, and no characteristic peak of ZrO 2 was observed.The monoclinic CuO phase was reduced to a cubic metallic Cu 0 phase (2θ = 43.3• , 50.4     , PDF#65-3288) via H 2 reduction at 300 • C in Figure 1B [32,33].As the ZnO/ZrO 2 ratio increased, the Cu reflection signal intensified, indicating that metal Cu continued to aggregate.In Figure 1C, it is worth noting that the Cu characteristic diffraction peak (2θ = 43.3• ) moved to lower degrees as the ZnO content increased.This indicates that during the catalyst reduction activation process, a fraction of ZnO is reduced to Zn 0 and interacts with Cu to form a CuZn alloy [34,35].The crystalline structures of all the catalysts after calcination and reduction were analyzed by XRD (Figure 1).The monoclinic CuO (2θ = 35.5°and 38.7°, PDF#48-1548) was observed in all the calcined samples in Figure 1A, while ZnO (2θ = 31.8°,34.4° and 38.7°, PDF#79-0206) [30,31] can only be detected when the Zn content exceeds 20 wt%.The ZrO2 in all samples was a-ZrO2, because of the low calcination temperature, and no characteristic peak of ZrO2 was observed.The monoclinic CuO phase was reduced to a cubic metallic Cu 0 phase (2θ = 43.3°,50.4°, and 74.1°, PDF#4-836) and Cu2O (2θ = 37.1°, PDF#65-3288) via H2 reduction at 300 °C in Figure 1B [32,33].As the ZnO/ZrO2 ratio increased, the Cu reflection signal intensified, indicating that metal Cu continued to aggregate.In Figure 1C, it is worth noting that the Cu characteristic diffraction peak (2θ = 43.3°)moved to lower degrees as the ZnO content increased.This indicates that during the catalyst reduction activation process, a fraction of ZnO is reduced to Zn 0 and interacts with Cu to form a CuZn alloy [34,35].The TEM image (Figure S2) shows that the reduced catalyst exhibits irregular morphology, with a size ranging approximately from 400-600 nm.In Cu/ZrO2 and Cu/ZnO/ZrO2, the black area weakly increased with the ZnO content increased, due to Cu aggregation.However, when the ZrO2 content decreased from 10% (C3Z6Z1) to 0% (C3Z7Z0), the size of Cu particles increased significantly, which is consistent with the XRD analysis.The size of Cu particles in Cu/ZnO is significantly larger than that in Cu/ZrO2, which suggests that ZrO2 has a stronger dispersion effect on Cu than ZnO [36].Regular hemispherical or spherical Cu nanoparticles can be found in the HR-TEM images (Figure 2).In the C3Z1Z6 sample, Cu particles expose the (100) and (111) crystal planes corresponding to the lattice spacing = 0.182 and 0.210 nm, and ZnO particles expose the (101) crystal planes corresponding to the lattice spacing = 0.248 nm [37].As the ZnO/ZrO2 mass The TEM image (Figure S2) shows that the reduced catalyst exhibits irregular morphology, with a size ranging approximately from 400-600 nm.In Cu/ZrO 2 and Cu/ZnO/ZrO 2 , the black area weakly increased with the ZnO content increased, due to Cu aggregation.However, when the ZrO 2 content decreased from 10% (C3Z6Z1) to 0% (C3Z7Z0), the size of Cu particles increased significantly, which is consistent with the XRD analysis.The size of Cu particles in Cu/ZnO is significantly larger than that in Cu/ZrO 2 , which suggests that ZrO 2 has a stronger dispersion effect on Cu than ZnO [36].Regular hemispherical or spherical Cu nanoparticles can be found in the HR-TEM images (Figure 2).In the C3Z1Z6 sample, Cu particles expose the (100) and (111) crystal planes corresponding to the lattice spacing = 0.182 and 0.210 nm, and ZnO particles expose the (101) crystal planes corresponding to the lattice spacing = 0.248 nm [37].As the ZnO/ZrO 2 mass ratio increased, Cu (111) crystal plane lattice spacing expanded to 0.216-0.219nm.The expansion of lattice stripes indicates that during the reduction process, ZnO is reduced to Zn 0 and combines with the metallic Cu to form a CuZn alloy, which is consistent with the XRD result [35,38].No lattice fringe of ZrO 2 was found in the HR-TEM image, which is related to the amorphous morphology of ZrO 2 .EDS-mapping (Energy Dispersive Spectrometer mapping) was used to determine the distribution of Cu, Zn, and Zr in the catalyst.In Figure S3, as the ZnO/ZrO 2 ratio increased, the metallic Cu 0 particles were gradually aggregated.
expansion of lattice stripes indicates that during the reduction process, ZnO is reduced to Zn 0 and combines with the metallic Cu to form a CuZn alloy, which is consistent with the XRD result [35,38].No lattice fringe of ZrO2 was found in the HR-TEM image, which is related to the amorphous morphology of ZrO2.EDS-mapping (Energy Dispersive Spectrometer mapping) was used to determine the distribution of Cu, Zn, and Zr in the catalyst.In Figure S3, as the ZnO/ZrO2 ratio increased, the metallic Cu 0 particles were gradually aggregated.
N2O-oxidation followed by H2 titration was used to calculate the SCu, DCu, and dCu (Table 1).As the ZnO/ZrO2 ratio increases, metallic Cu 0 particles gradually aggregate, reasonably agreeing with the TEM and XRD analyses.Due to the presence of CuZn alloy in the sample, the Cu particle size calculated from the N2O titration is too large, so the results are considered principally for qualitative analysis.The H2-TPR (H2-temperature-programmed reduction) analysis (Figure 3) for the calcined catalysts generated overlapped hydrogen consumption peaks with resultant asymmetric profiles in the 100-230 °C temperature range.In the catalysts with the ZnO/ZrO2 ratio between 0:7 and 5:2, the three peaks (denoted as α, β, and γ) suggest the presence of different CuO species.Within this range, as the ZnO/ZrO2 ratio increased, α peak intensity gradually decreased, indicating a decrease in content of smaller Cu particles.At the same time, β and γ peaks gradually became dominant.The reduction temperature first decreases and then increases as the ZnO/ZrO2 ratio increases, and the C3Z2Z5 catalyst shows the lowest reduction temperature.This indicates that when the ZnO/ZrO2 ratio is 2:5, the interaction between Cu and oxide support is strong, promoting the reduction of Cu species [39].Only β and γ peaks were found in the C3Z6Z1 and C3Z7Z0 catalyst, which demonstrate the Cu species in the sample are mostly large CuO particles or exist in the form of bulk CuO.The theoretical H2 reduction consumption was calculated and compared with the actual H2 reduction consumption (Table 2).The ratio was recorded as RH2/Cu, with RH2/Cu > 1, indicating that ZnO was partially reduced to Zn 0 .N 2 O-oxidation followed by H 2 titration was used to calculate the S Cu , D Cu , and d Cu (Table 1).As the ZnO/ZrO 2 ratio increases, metallic Cu 0 particles gradually aggregate, reasonably agreeing with the TEM and XRD analyses.Due to the presence of CuZn alloy in the sample, the Cu particle size calculated from the N 2 O titration is too large, so the results are considered principally for qualitative analysis.
The H 2 -TPR (H 2 -temperature-programmed reduction) analysis (Figure 3) for the calcined catalysts generated overlapped hydrogen consumption peaks with resultant asymmetric profiles in the 100-230 • C temperature range.In the catalysts with the ZnO/ZrO 2 ratio between 0:7 and 5:2, the three peaks (denoted as α, β, and γ) suggest the presence of different CuO species.Within this range, as the ZnO/ZrO 2 ratio increased, α peak intensity gradually decreased, indicating a decrease in content of smaller Cu particles.At the same time, β and γ peaks gradually became dominant.The reduction temperature first decreases and then increases as the ZnO/ZrO 2 ratio increases, and the C3Z2Z5 catalyst shows the lowest reduction temperature.This indicates that when the ZnO/ZrO 2 ratio is 2:5, the interaction between Cu and oxide support is strong, promoting the reduction of Cu species [39].Only β and γ peaks were found in the C3Z6Z1 and C3Z7Z0 catalyst, which demonstrate the Cu species in the sample are mostly large CuO particles or exist in the form of bulk CuO.The theoretical H 2 reduction consumption was calculated and compared with the actual H 2 reduction consumption (Table 2).The ratio was recorded as R H2/Cu , with R H2/Cu > 1, indicating that ZnO was partially reduced to Zn 0 .
The chemical states of Cu, Zn, and Zr on the catalyst surface can be analyzed by XPS (Figure 4).In the Cu 2p XPS spectra of the reduced catalyst (Figure 4A), the signals at binding energy (BE) = 932.6 eV and 952.4 eV belong to the characteristic peaks of Cu 2p 3/2 and Cu 2p 1/2 .No Cu 2+ satellite peak near 943.0 eV was observed, indicating that the surface CuO species were completely reduced [40,41].The Cu LMM Auger region was used to better distinguish between Cu 0 and Cu + .However, there is a certain overlap in the characteristic peak positions of Cu and Zn, so only samples with a ZnO/ZrO 2 ratio less than 2:5 were analyzed by Cu LMM.In Figure 4C, the characteristic peaks of Cu 0 or Cu + appeared at binding energies of 569.9 eV and 573.9 eV, but the content of Cu + was relatively low.After reduction, most of the Cu on the catalyst surface existed in the form of Cu 0 .
The characteristic peaks of Zr 3d 5/2 and Zr 3d 3/2 at BE = 181.8eV and 184.2 eV were observed in the reduced catalysts (Figure 4B).At the same time, the reduction of the catalysts did not cause a significant shift in the position of the Zr 3d characteristic peak, indicating that the surface ZrO 2 was basically not reduced to metal Zr.It can be inferred that the excessive consumption of H 2 in H 2 -TPR is mainly related to the reduction of ZnO.The valence state distribution of Zn in the catalysts after reduction was analyzed by Zn LMM (Figure 4D).In C3Z2Z5, C3Z5Z2, and C3Z7Z0 catalysts, the characteristic peaks of Zn appeared at Kinetic energy = 988.0eV, 991.2 eV, and 992.4 eV, respectively, corresponding to Zn 2+ , Zn q+ , and Zn 0 .The presence of Zn 0 further clarifies that some ZnO was reduced to metal Zn during the reduction process.As the ZnO/ZrO 2 ratio increases, the content of Zn 0 shows a trend of first increasing and then decreasing.The chemical states of Cu, Zn, and Zr on the catalyst surface can be analyzed by XPS (Figure 4).In the Cu 2p XPS spectra of the reduced catalyst (Figure 4A), the signals at binding energy (BE) = 932.6 eV and 952.4 eV belong to the characteristic peaks of Cu 2p3/2 and Cu 2p1/2.No Cu 2+ satellite peak near 943.0 eV was observed, indicating that the surface CuO species were completely reduced [40,41].The Cu LMM Auger region was used to better distinguish between Cu 0 and Cu + .However, there is a certain overlap in the characteristic peak positions of Cu and Zn, so only samples with a ZnO/ZrO2 ratio less than 2:5 were analyzed by Cu LMM.In Figure 4C, the characteristic peaks of Cu 0 or Cu + appeared at binding energies of 569.9 eV and 573.9 eV, but the content of Cu + was relatively low.After reduction, most of the Cu on the catalyst surface existed in the form of Cu 0 .
The characteristic peaks of Zr 3d5/2 and Zr 3d3/2 at BE = 181.8eV and 184.2 eV were observed in the reduced catalysts (Figure 4B).At the same time, the reduction of the catalysts did not cause a significant shift in the position of the Zr 3d characteristic peak, indicating that the surface ZrO2 was basically not reduced to metal Zr.It can be inferred that the excessive consumption of H2 in H2-TPR is mainly related to the reduction of ZnO.The valence state distribution of Zn in the catalysts after reduction was analyzed by Zn LMM (Figure 4D).In C3Z2Z5, C3Z5Z2, and C3Z7Z0 catalysts, the characteristic peaks of Zn appeared at Kinetic energy = 988.0eV, 991.2 eV, and 992.4 eV, respectively, corresponding to Zn 2+ , Zn q+ , and Zn 0 .The presence of Zn 0 further clarifies that some ZnO was reduced to metal Zn during the reduction process.As the ZnO/ZrO2 ratio increases, the content of Zn 0 shows a trend of first increasing and then decreasing.
In the process of CO2 hydrogenation to methanol, the capacity of the catalyst adsorption/activation of CO2 and H2 plays an important role.In Cu/ZrO2 and Cu/ZnO/ZrO2, the  In the process of CO 2 hydrogenation to methanol, the capacity of the catalyst adsorption/activation of CO 2 and H 2 plays an important role.In Cu/ZrO 2 and Cu/ZnO/ZrO 2 , the difference in H 2 adsorption capacity (based on per gram catalyst, Table 1) is not significant, which means that the ZnO/ZrO 2 ratio has a small impact on H 2 adsorption.In the Cu/ZnO catalyst (C3Z7Z0), the adsorption capacity of H 2 is significantly lower, which is related to the larger size of Cu particles on ZnO.The larger Cu particle size reduces the adsorption sites of H 2 , causing a decrease in the adsorption capacity of H 2 .In the H 2 -TPD (H 2 chemisorption) profile (Figure 5A) for the catalysts containing Zr, a main desorption peak of 380-420 • C can be observed, which is the desorption peak of H species bound by Cu-H bonds on metal Cu [42].As the ZnO/ZrO 2 ratio increases, the peak temperature firstly shifts towards low temperature and then remains basically unchanged, indicating that the formation of CuZn alloy will reduce the desorption temperature of H 2 .In Cu/ZnO (C3Z7Z0), the desorption temperature further decreases.In the range of the ZnO/ZrO 2 ratio between 1:6 and 5:2, the high temperature peak of 500-600 • C in H 2 -TPD did not appear in the H 2 -MS spectra (Figure 5B), which may be related to the unknown species released from the catalyst decomposition under high-temperature conditions (higher than the catalyst calcination temperature).
catalyst and the change is not significant, but the desorption temperature gradually increases.The CO2 desorption amount in the high-temperature zone reaches a maximum value on C3Z2Z5 and then decreases.The desorption temperature is basically the same in the samples containing Zn but is higher than that in C3Z0Z7.The CO2 desorption amount (Table 1) depends on the ZnO and ZrO2 composition.The value firstly increases and then decreases with increasing ZnO/ZrO2 ratio and the highest amount (0.311 mmol/gcat) is recorded at the ratio of 2:5.The CO2 desorption peaks correspond to three basic sites on the catalyst surface, namely weak, moderate, and strong basic sites.At weak basic sites, CO2 adsorbs as bicarbonate and desorbs at low temperatures [6,39].The moderate basic sites are generally metal-O pairs, which adsorb CO2 and form carbonate species [43].At medium temperature, the carbonate intermediate is prone to RWGS, generating CO [44].The species generated by CO2 adsorption on strongly basic sites are not yet clear, but it can be determined that the presence of strongly basic sites is related to the Cu-ZrO2 interface [29, [45][46][47].CO2 hydrogenation to methanol is generally operated at 200-300 °C.Therefore, the moderate and strong basic sites are effective active sites for CO2 adsorption/activation and further conversion.The difference in ZnO/ZrO2 ratio affects the types and quantities of basic sites on the carrier surface and changes the desorption amount and temperature of CO2.A total of three main desorption signals were detected in the CO 2 -TPD analysis (Figure 6A), including the low-temperature (80-150 • C), medium-temperature (350-450 • C), and high-temperature (450-620 • C) ranges.The difference in ZnO/ZrO 2 ratio affects the CO 2 desorption amount and temperature within different temperature ranges.With the increase of the ZnO/ZrO 2 ratio, the desorption amount in the low-temperature range decreases to a certain extent, but the desorption temperature does not significantly increase.In the middle-temperature range, the CO 2 desorption amount is relatively low in each catalyst and the change is not significant, but the desorption temperature gradually increases.The CO 2 desorption amount in the high-temperature zone reaches a maximum value on C3Z2Z5 and then decreases.The desorption temperature is basically the same in the samples containing Zn but is higher than that in C3Z0Z7.The CO 2 desorption amount (Table 1) depends on the ZnO and ZrO 2 composition.The value firstly increases and then decreases with increasing ZnO/ZrO 2 ratio and the highest amount (0.311 mmol/g cat ) is recorded at the ratio of 2:5.
erated by CO2 adsorption on strongly basic sites are not yet clear, but it can be determined that the presence of strongly basic sites is related to the Cu-ZrO2 interface [29, [45][46][47].CO2 hydrogenation to methanol is generally operated at 200-300 °C.Therefore, the moderate and strong basic sites are effective active sites for CO2 adsorption/activation and further conversion.The difference in ZnO/ZrO2 ratio affects the types and quantities of basic sites on the carrier surface and changes the desorption amount and temperature of CO2.

Catalytic Test
Catalytic performance comparison of CuZn (C3Z7Z0), CuZr (C3Z0Z7), CuZnZr (C3Z2Z5), and commercial catalyst (CuZnAl) for the reaction is shown in Figure 7A.The CO2 conversion increases with the increasing temperature.The order of CO2 conversion follows CuZnAl > CuZnZr > CuZr > CuZn.Methanol selectivity is decreased with increasing temperature (Figure 7B).Except for CuZnAl, the methanol selectivity remains above 60% at 240 °C, indicating that the higher CO2 conversion over CuZnAl is due to the generation of more by-product CO.The methanol selectivity for the catalysts follows the order CuZr > CuZn ≈ CuZnZr > CuZnAl.STY of methanol was used to determine catalyst activity for methanol synthesis (Figure 7C).The STY of methanol rises with the increase in temperature (180-260 °C).CuZnZr shows the best catalytic performance in all four catalysts, which also ranks top in the existing literature (Table S1).In traditional CO2 hydrogenation to methanol catalysts, CuZn alloy [9,27,28] or Cu-ZnO interface [48] is generally considered as the active site.Here we can note that the CuZn sample displays the lowest The CO 2 desorption peaks correspond to three basic sites on the catalyst surface, namely weak, moderate, and strong basic sites.At weak basic sites, CO 2 adsorbs as bicarbonate and desorbs at low temperatures [6,39].The moderate basic sites are generally metal-O pairs, which adsorb CO 2 and form carbonate species [43].At medium temperature, the carbonate intermediate is prone to RWGS, generating CO [44].The species generated by CO 2 adsorption on strongly basic sites are not yet clear, but it can be determined that the presence of strongly basic sites is related to the Cu-ZrO 2 interface [29, [45][46][47].CO 2 hydrogenation to methanol is generally operated at 200-300 • C. Therefore, the moderate and strong basic sites are effective active sites for CO 2 adsorption/activation and further conversion.The difference in ZnO/ZrO 2 ratio affects the types and quantities of basic sites on the carrier surface and changes the desorption amount and temperature of CO 2 .

Catalytic Test
Catalytic performance comparison of CuZn (C3Z7Z0), CuZr (C3Z0Z7), CuZnZr (C3Z2Z5), and commercial catalyst (CuZnAl) for the reaction is shown in Figure 7A.The CO 2 conversion increases with the increasing temperature.The order of CO 2 conversion follows CuZnAl > CuZnZr > CuZr > CuZn.Methanol selectivity is decreased with increasing temperature (Figure 7B).Except for CuZnAl, the methanol selectivity remains above 60% at 240 • C, indicating that the higher CO 2 conversion over CuZnAl is due to the generation of more by-product CO.The methanol selectivity for the catalysts follows the order CuZr > CuZn ≈ CuZnZr > CuZnAl.STY of methanol was used to determine catalyst activity for methanol synthesis (Figure 7C).The STY of methanol rises with the increase in temperature (180-260 • C).CuZnZr shows the best catalytic performance in all four catalysts, which also ranks top in the existing literature (Table S1).In traditional CO 2 hydrogenation to methanol catalysts, CuZn alloy [9,27,28] or Cu-ZnO interface [48] is generally considered as the active site.Here we can note that the CuZn sample displays the lowest Cu surface area with the smallest amount for reactant uptake, which can be considered to be responsible for the poor activity for methanol formation.Meanwhile, CuZr shows the highest dispersion and largest surface area of Cu and relatively high capacity for H 2 and CO 2 adsorption but displays a lower activity than CuZnZr (C3Z2Z5).This suggests the Cu surface area is not the exclusive factor in determining the catalyst activity for CO 2 hydrogenation to methanol.Moreover, CuZr is more active than CuZn, indicating more effective active sites formed on CuZr.Based on the CO 2 -TPD results, the presence of the Cu-ZrO 2 interface can effectively increase the CO 2 adsorption capacity, thereby promoting the reaction.We consider that both ZnO and ZrO 2 play an important role in working with Cu for the construction of the active sites over CuZnZr.As indicated by the XRD and TEM analyses, CuZn alloy is formed from strong Cu-ZnO interaction over C3Z2Z5.The C3Z2Z5 catalyst exhibits a higher capacity for CO 2 adsorption relative to CuZn and CuZr, implying the involvement of the Cu-ZrO 2 interface in catalyzing the reaction.It is possible that the cooperation between Cu-oxide interaction (e.g., CuZn alloy and Cu-ZrO 2 interface) acts as the active sites, effectively improving carbon dioxide conversion rate and methanol selectivity.Since the composition of ZnO and ZrO 2 impacts the Cu-oxide interaction, the following work further investigates the optimization of ZnO and ZrO 2 contents and establishes the relationship between the catalyst structure and the activity.
active sites formed on CuZr.Based on the CO2-TPD results, the presence of the Cu-ZrO2 interface can effectively increase the CO2 adsorption capacity, thereby promoting the reaction.We consider that both ZnO and ZrO2 play an important role in working with Cu for the construction of the active sites over CuZnZr.As indicated by the XRD and TEM analyses, CuZn alloy is formed from strong Cu-ZnO interaction over C3Z2Z5.The C3Z2Z5 catalyst exhibits a higher capacity for CO2 adsorption relative to CuZn and CuZr, implying the involvement of the Cu-ZrO2 interface in catalyzing the reaction.It is possible that the cooperation between Cu-oxide interaction (e.g., CuZn alloy and Cu-ZrO2 interface) acts as the active sites, effectively improving carbon dioxide conversion rate and methanol selectivity.Since the composition of ZnO and ZrO2 impacts the Cu-oxide interaction, the following work further investigates the optimization of ZnO and ZrO2 contents and establishes the relationship between the catalyst structure and the activity.The effect of varying ZnO/ZrO 2 ratios on the catalytic performance in the reaction over Cu/ZnO/ZrO 2 at 220 • C and 240 • C reveals the highest CO 2 conversion is achieved at the ZnO/ZrO 2 ratio of 2:5 to 3:4 (Figure 7D,E), suggesting an optimized content of ZnO and ZrO 2 is required for efficient activation and conversion of CO 2 .The methanol selectivity does not vary dramatically with changes in the ZnO/ZrO 2 ratio.Due to the larger change in CO 2 conversion than in methanol selectivity, the overall trend of STY of methanol with the Zn/Zr ratio is similar to the trend of CO 2 conversion (Figure 7F).Except for 220 • C, the STY of methanol of C3Z2Z5 is the highest at all other temperatures.The trend of activity variation is similar to the results of Zn LMM and CO 2 -TPD.The catalyst activity increased before the ZnO content increased to 20%, possibly due to the increase in CuZn alloy content.However, as the ZnO content increases, the ZrO 2 content correspondingly decreases, and the Cu-ZrO 2 interface gradually decreases, which may lead to a decrease in catalyst activity.
In C3Z2Z5, the amount of CuZn alloy and Cu-ZrO 2 interface reach the highest values, demonstrating the best catalytic activity.
The C3Z2Z5 catalyst was used to investigate the effect of reduction temperature on catalytic performance (Figure 8).When the reduction temperature exceeds 260 • C, there is no significant difference in CO 2 conversion, methanol selectivity, and STY of methanol (Figure 8A-C), indicating that the catalyst had been completely reduced beyond 260 • C, which is consistent with the results of H 2 -TPR.Under the reaction conditions of 220, 240, and 260 • C, with the increase in reduction temperature, the conversion rate slightly increases and then remains unchanged, while the selectivity initially increases and then decreases (Figure 8D-F).When the reduction temperature is higher than 260 • C, the catalytic performance does not change significantly, indicating that a higher reduction temperature below 300 • C does not significantly alter the structure of the catalyst.The C3Z2Z5 catalyst was also selected to investigate the impact of gas hourly space velocity (GHSV, 6000, 12,000, 18,000, and 24,000 mL/g cat /h) on catalytic performance (Figure 9).Lower GHSV results in higher CO 2 conversion but lower methanol selectivity, indicating that low GHSV is more conducive to the production of CO through the RWGS.The STY of methanol is also increased with the increase of GHSV, reaching a maximum value at 24,000 mL/g cat /h under the conditions used here.
In the long-term stability test for 110 h (P = 3 MPa, T = 240 • C, GHSV = 18,000 mL/g cat /h), the CO 2 conversion and methanol selectivity over the C3Z2Z5 catalyst remain basically unchanged (Figure S4), and exhibit no significant deactivation.Comparing the XRD patterns of the catalyst after reduction and after 110 h reaction (Figure S4B), the intensities of the Cu peak (2θ = 43.3• and 50.4 • ) are similar, indicating that the C3Z2Z5 sample after 110 h reaction exhibits no severe sintering of Cu particles.TEM and EDS mapping of the spent catalyst (Figure S4C-F) verify that the long-term stability test does not cause aggregation of the Cu particles, indicating that the C3Z2Z5 catalyst exhibits good stability.
The adsorption capacity of catalysts on reactants will have a significant impact on the reaction performance.When the metal Cu content is fixed, except for the Cu/ZnO catalyst, there is not much difference in the H 2 adsorption amount, but the change in the CO 2 adsorption amount is more significant.The correlation analysis between the CO 2 desorption amount and catalyst performance (Figure 10A) shows that the catalytic activity is directly proportional to the CO 2 desorption amount, indicating that the ZnO/ZrO 2 ratio affects the reaction performance by adjusting the CO 2 adsorption sites.When the ZnO/ZrO 2 ratio is 2:5, the number of strong basic sites on the catalyst surface is the highest, effectively improving the CO 2 adsorption amount and enhancing the catalytic activity.
XRD, HR-TEM, and XPS analyses show that a fraction of ZnO is reduced to Zn 0 and reacts with metallic Cu to form a CuZn alloy.The content of Zn 0 in different samples can be calculated by Zn-LMM, which corresponds to the amount of CuZn alloy.By correlating the catalytic performance with the content of Zn 0 (Figure 10B), C3Z2Z5 with the highest CuZn alloy content exhibits the highest activity, while the catalysts with lower Zn 0 content also had lower catalytic performance, indicating that the presence of CuZn alloy effectively promotes the reaction.The change in the ZnO/ZrO 2 ratio not only affects the amount of CuZn alloy but also affects the number of Cu-ZrO 2 interfaces, thereby affecting the adsorption capacity of CO 2 .C3Z4Z3 and C3Z3Z4 have similar CuZn alloy content, but C3Z3Z4 exhibits better catalytic activity, which can be associated with an increase in the CO 2 adsorption capacity.Thus, we consider the oxide (ZnO and ZrO 2 ) composition mainly affects the catalytic performance in methanol synthesis by regulating the quantity of CuZn alloys and Cu-ZrO 2 interaction, and furthers the ability for adsorption/activation of reactants.The synergistic sites associated with the CuZn alloy and Cu-ZrO 2 interface in the optimized Cu/ZnO/ZrO 2 catalyst (C3Z2Z5) reach the maximum, contributing to the enhanced production of methanol.In the long-term stability test for 110 h (P = 3 MPa, T = 240 °C, GHSV = 18,000 mL/gcat/h), the CO2 conversion and methanol selectivity over the C3Z2Z5 catalyst remain basically unchanged (Figure S4), and exhibit no significant deactivation.Comparing the XRD patterns of the catalyst after reduction and after 110 h reaction (Figure S4B), the intensities of the Cu peak (2θ = 43.3°and 50.4°) are similar, indicating that the C3Z2Z5 sample after 110 h reaction exhibits no severe sintering of Cu particles.TEM and EDS mapping of the spent catalyst (Figure S4C-F) verify that the long-term stability test does not cause aggregation of the Cu particles, indicating that the C3Z2Z5 catalyst exhibits good stability.
The adsorption capacity of catalysts on reactants will have a significant impact on the reaction performance.When the metal Cu content is fixed, except for the Cu/ZnO catalyst, there is not much difference in the H2 adsorption amount, but the change in the CO2 adsorption amount is more significant.The correlation analysis between the CO2 desorption amount and catalyst performance (Figure 10A) shows that the catalytic activity is directly proportional to the CO2 desorption amount, indicating that the ZnO/ZrO2 ratio affects the reaction performance by adjusting the CO2 adsorption sites.When the ZnO/ZrO2 ratio is 2:5, the number of strong basic sites on the catalyst surface is the highest, effectively improving the CO2 adsorption amount and enhancing the catalytic activity.
XRD, HR-TEM, and XPS analyses show that a fraction of ZnO is reduced to Zn 0 and reacts with metallic Cu to form a CuZn alloy.The content of Zn 0 in different samples can be calculated by Zn-LMM, which corresponds to the amount of CuZn alloy.By correlating the catalytic performance with the content of Zn 0 (Figure 10B), C3Z2Z5 with the highest CuZn alloy content exhibits the highest activity, while the catalysts with lower Zn 0 content also had lower catalytic performance, indicating that the presence of CuZn alloy effectively promotes the reaction.The change in the ZnO/ZrO2 ratio not only affects the amount of CuZn alloy but also affects the number of Cu-ZrO2 interfaces, thereby affecting the adsorption capacity of CO2.C3Z4Z3 and C3Z3Z4 have similar CuZn alloy content, but C3Z3Z4 exhibits better catalytic activity, which can be associated with an increase in the CO2 adsorption capacity.Thus, we consider the oxide (ZnO and ZrO2) composition mainly affects the catalytic performance in methanol synthesis by regulating the quantity of CuZn alloys and Cu-ZrO2 interaction, and furthers the ability for adsorption/activation of reactants.The synergistic sites associated with the CuZn alloy and Cu-ZrO2 interface in the optimized Cu/ZnO/ZrO2 catalyst (C3Z2Z5) reach the maximum, contributing to the enhanced production of methanol.

Catalyst Preparation
A series of Cu/ZnO/ZrO2 catalysts with varying Cu contents were prepared by coprecipitation.The mixed nitrates (including copper nitrate, zinc nitrate, and zirconium nitrate) in ethanol were precipitated at 70 °C using excess oxalic acid under stirring for 1 h, then aged for 4 h at room temperature.The precipitate was separated by centrifugation, washed three times with ethanol, and dried at 80 °C for 12 h.The dried samples were calcined at 450 °C for 4 h.After calcination, the samples were subjected to compression, crushing, and sieving treatment and 40-60 mesh particles were selected for the catalysts test.The target mass fraction of Cu was fixed at 30 wt%, and a total of 8 samples were prepared by sequentially adjusting the mass ratio of ZnO/ZrO2.The calcined catalysts were named based on element content.

Catalyst Characterization
Nitrogen physisorption was tested using a Micromeritics ASAP 2460 (Micromeritics instrument Ltd., Atlanta, America) system for analysis of the specific surface area (SBET) and pore structure (pore volume and pore size) according to the standard Braeuer-Emmett-Teller (BET) method and Barrett-Joyner-Halenda desorption branch.All the samples were outgassed under vacuum at 300 °C for 4 h before the nitrogen physisorption test.

Catalyst Preparation
A series of Cu/ZnO/ZrO 2 catalysts with varying Cu contents were prepared by coprecipitation.The mixed nitrates (including copper nitrate, zinc nitrate, and zirconium nitrate) in ethanol were precipitated at 70 • C using excess oxalic acid under stirring for 1 h, then aged for 4 h at room temperature.The precipitate was separated by centrifugation, washed three times with ethanol, and dried at 80 • C for 12 h.The dried samples were calcined at 450 • C for 4 h.After calcination, the samples were subjected to compression, crushing, and sieving treatment and 40-60 mesh particles were selected for the catalysts test.The target mass fraction of Cu was fixed at 30 wt%, and a total of 8 samples were prepared by sequentially adjusting the mass ratio of ZnO/ZrO 2 .The calcined catalysts were named based on element content.

Catalyst Characterization
Nitrogen physisorption was tested using a Micromeritics ASAP 2460 (Micromeritics instrument Ltd., Atlanta, America) system for analysis of the specific surface area (S BET ) and pore structure (pore volume and pore size) according to the standard Braeuer-Emmett-Teller (BET) method and Barrett-Joyner-Halenda desorption branch.All the samples were outgassed under vacuum at 300 • C for 4 h before the nitrogen physisorption test.
For ex situ characterizations (including nitrogen physisorption, XRD, TEM, and XPS), the calcined catalyst should be reduced at 300 • C for 3 h.

Catalyst Testing
The catalytic performance was evaluated using a fixed-bed continuous-flow stainlesssteel reactor.The catalyst (100 mg, 40−60 mesh) was reduced in H 2 (10 mL/min) for 3 h before the reaction and cooled down to 180 • C, which was monitored by a thermocouple in the catalyst bed.Afterward, the hydrogen gas was turned off and the feed gases (CO 2 /H 2 = 3/1) and N 2 (internal standard) were introduced, increasing the pressure to 3 MPa.Next, the reactor temperature was raised to the test temperature (180-300 • C).The composition of the outlet gas was analyzed online using a gas chromatographic system (Shimadzu, GC-2014) equipped with an FID (flame ionization detector, for methanol) and a TCD (thermal conductivity detector, for N 2 , CO 2 , CO, and CH 4 ).The X CO 2 (conversion of CO 2 ), S CH3OH (methanol selectivity), and STY CH3OH (space-time yield of methanol) were calculated with the following equations: where F is the volumetric flow rate, M CH3OH is the molecule weight of methanol, and m cat refers to the mass of the catalyst.The deviation of the carbon balance is within 5%.

Conclusions
This study has established the effect of ZnO and ZrO 2 contents on the Cu/ZnO/ZrO 2 catalyst structure/surface characteristics and catalytic performance in the CO 2 hydrogenation to methanol.CuZn alloy was formed in the catalyst containing ZnO, and the ZnO/ZrO 2 ratio can affect the content of CuZn alloy.A correlation between the amount of CuZn alloy and the catalytic activity reveals the participation of CuZn alloy in the reaction.The ZnO/ZrO 2 ratio also affects the amount of strong basic sites on the catalyst surface.With an increase in strong basic sites related to the Cu-ZrO 2 interfaces, the amount of CO 2 desorption significantly increases.The increase in both CuZn alloy content and Cu-ZrO 2 interfaces promotes the process of CO 2 hydrogenation to methanol.The optimized catalyst C3Z2Z5 shows the highest STY of methanol (610.8 g CH3OH /kg cat /h) with no significant decrease in activity during long-term stability testing.

Figure 3 .
Figure 3. H 2 -TPR profiles for all the catalysts.

Table 1 .
Physicochemical properties of the catalysts.

/g cat ) b S Cu (m 2 /g Cu ) b H 2 Uptake (mmol/g cat ) CO 2 Desorption (mmol/g cat )
a Derived from ICP-OES.b Derived from N 2 O titration.

Table 2 .
H2 consumption during the TPR analysis.

Table 2 .
H 2 consumption during the TPR analysis.