Investigation of Ternary CuZnM (M = Cr, Ce, Zr, Al) Catalysts in CO₂ Hydrogenation for Methanol Synthesis

Abstract

The hydrogenation of CO₂ to methanol over Cu-based catalysts is one of the attractive routes to reduce greenhouse gas emissions and generate high-value-added chemicals. Industrial CuZnAl catalysts possess some shortcomings, but various promoters can enhance the activity and durability of Cu-based catalysts for CO₂ hydrogenation. Herein, we systematically investigated the variations in the physicochemical properties of ternary CuZnM (M: Cr, Ce, Zr, Al) catalysts induced by different promoters, as well as their impact on CO₂ hydrogenation to methanol. The results demonstrate that the catalytic activity followed the order of CZCr > CZCe > CZZr > CZAl, with CZCr exhibiting the highest stability among them. Combined with XRD, SEM, TEM, H₂-TPR, TPD, XPS, Raman findings and the experimental results, the smaller Cu particle size was conducive to increasing the CH₃OH yield, and the lower Cu⁺/Cu⁰ ratio exhibited better stability in CO₂ hydrogenation to methanol reaction. This approach offers theoretical insights and practical solutions for the industrial application of ternary Cu-based catalysts.

1. Introduction

The rapid consumption of fossil fuels currently releases a large amount of the greenhouse gas carbon dioxide (CO₂) [1]. The resource utilization of carbon dioxide has always been the focus of research, particularly in the production of high-value chemicals such as methanol [2,3] and methane [4,5], which are also premium renewable energy sources. Developing a methanol production pathway from CO₂, where methanol can complement and partially act as a substitute for petroleum-based fuels, may address energy scarcity and improve resource efficiency [6]. The catalytic coupling of CO₂ and green hydrogen to produce methanol can not only reduce greenhouse gas emissions but also provide a sustainable energy path [7].

At present, the efficiency of CO₂ hydrogenation to methanol predominantly relies on the catalyst activity, and the development of a highly efficient catalyst has been the focus of researchers, including Cu-based catalysts, precious metal catalysts, other metal oxide catalysts, etc. [8]. Among them, the ternary Cu-based catalyst has been widely used because of its excellent catalytic performance. Such catalysts are usually composed of copper, zinc and other metallic elements and can efficiently promote the activation and hydrogenation of CO₂. Copper is often considered to be the main active phase, with Zn as a carrier or structural barrier and a third element as an accelerator [9]. In the past, CuZnAl catalysts have gained prominence in methanol synthesis due to their cost efficiency and exceptional catalytic performance, with a substantial body of literature dedicated to exploring doped Cu-based catalysts enhanced by various promoters [10,11,12]. However, some Cu-based catalysts can still face aggregation or re-oxidation during catalysis, leading to a decrease in active sites and consequently impacting their catalytic performance and long-term stability [13]. In order to surpass the aforementioned challenges in CO₂ hydrogenation for methanol production, an effective tactic to enhance the efficacy of Cu/Zn-based catalysts involves the incorporation of a promoter, that is, the third element. The function of the third element in the catalyst is multifaceted, including but not limited to enhancing catalyst dispersion, strengthening metal–support interactions, modulating the electronic configuration and fine-tuning the reaction pathway. Some transition metal oxides such as Fe₂O₃, Ga₂O₃, ZrO₂, CeO₂, In₂O₃, Cr₂O₃ and Al₂O₃ are often selected as promoters [14,15]. Singh et al. revealed the accelerating effect of the CeO₂ precursor in Cu/ZnO catalysts and established a relationship between the density of the strongly alkaline sites and the active H adsorption sites [16]. Zhan et al. developed a range of Cu/Zn/Zr catalysts with varying Zr/(Cu+Zn) ratios via a precipitation method and evaluated their performance in methanol synthesis from a CO₂/H₂ mixture, highlighting the reaction’s structural sensitivity [17]. Additionally, Chang et al. optimized the structure and surface properties of a Cu/ZnO/ZrO₂ catalyst by adjusting the ratio of ZnO and ZrO₂, which significantly improved the efficiency and selectivity of CO₂ hydrogenation to methanol [15]. Xiong et al. synthesized Cu/ZnCr catalysts via the incipient wetness impregnation technique and investigated the impact of the Zn/Cr ratio on methanol selectivity; the catalyst with an optimal Zn/Cr ratio exhibited a methanol space–time yield of 4.7 mmol/g_cat./h (X_CO2 = 25.1%; S_MeOH = 31.1%) at 300 °C and 2 MPa [18]. Hengne et al. synthesized copper (Cu) nanocomposite catalysts doped with gallium (Ga) and aluminum (Al) through a synchronous co-precipitation, assessing their efficacy in CO₂ hydrogenation to methanol. The incorporation of Ga was found to strengthen the catalyst’s weak base sites to a greater extent than the addition of Al [19].

Nevertheless, these advancements have yet to be comprehensively investigated to elucidate the promotional mechanisms underlying ternary Cu-based catalyst systems and to attain enhanced methanol yields alongside extended operational longevity in industrial settings. For example, the incorporation of ZrO₂ could stabilize the intermediates and reduce the energy barrier pathway by promoting the formation and conversion of formic acid intermediates [20]. Sun et al. demonstrated that Erbium (Er) doping could refine the ZnO particle size, consequently enlarging the Cu-ZnO interfacial area and fostering a robust electron metal–support interaction (EMSI) between Cu and ZnO. This enhancement was believed to expedite the activation of CO₂ adsorbates and reaction intermediates [21].

To overcome the constraints associated with Cu-based catalysts, an in-depth study was undertaken to examine the effects of Al, Ce, Zr and Cr additives on the physicochemical properties and catalytic performance in the hydrogenation of CO₂ to methanol. These four elements were the most frequently used in Cu-based catalysts, but they were not systematically studied as ternary Cu-based catalyst systems to improve catalytic efficiency more comprehensively. This study systematically assessed the efficacy of ternary Cu-based catalysts for CO₂ hydrogenation to methanol and scrutinized the impact of various dopants on the catalysts’ physicochemical features using a suite of analytical methods, including XRD, SEM, TEM, H₂-TPR, XPS, TPD and Raman spectroscopy. The findings revealed that the CuZnCr catalyst exhibited superior activity and stability in comparison with other catalysts, which was ascribed to the electronic interaction from Cu and Cr. Meanwhile, Ce, Zr and Al also played different roles in the system, such as particle size, reduction degree, adsorption capacity, etc. It might provide theoretical guidance for designing Cu-based catalysts with higher energy conversion efficiency in industrial applications.

2. Results

2.1. Crystal Structure Determined by XRD

In order to study the crystalline structure of ternary CZM catalysts and the effect of grain size of Cu species, XRD patterns of various catalysts are shown in Figure 1. All calcined catalysts exhibited characteristic CuO diffraction peaks at 2θ = 36.5° and 38.8°, as illustrated in Figure 1a, corresponding to the (002) and (111) crystal planes within the layered structure. In comparison, the CuO diffraction peaks of CZAl catalysts were sharper, indicating higher crystallinity and larger grain size. This suggested that the doping of different elements could greatly affect the grain sizes of copper species. According to Scherrer’s equation, the calculated mean CuO grain size was less than 20 nm. On the other hand, ZnO diffraction peaks located at ca. 32.2°, 38.8° and 66.8° were detected in all cases. However, there was no third element detected because of their high dispersion or amorphous property, and Raman spectroscopy (Figure S1) also failed to detect the vibration peak of the third element. Detailed descriptions of the Raman spectra can be found in the Supporting Information [22]. Their presence was demonstrated by EDS images, including Al, Ce, Zr and Cr (Figure S2). To further clarify the presence and content of each component in the catalyst sample, we characterized the sample using an X-ray fluorescence spectroscopy (XRF) technique. As shown in

Table 1. Mass percentage content (wt%) of each element in the CZM catalysts.

Catalyst	Element	Mass Percentage Content (wt%)
CZAl	Cu	60.761
	Zn	29.126
	Al	9.732
CZCe	Cu	60.169
	Zn	29.267
	Ce	9.8
CZZr	Cu	59.867
	Zn	28.475
	Zr	11.137
CZCr	Cu	61.430
	Zn	28.194
	Cr	10.078

, all target elements were effectively detected in the samples, and their composition was in excellent agreement with the theoretical design values. This indicates that the promoter elements were successfully incorporated into the catalyst system, and the metal nitrates were almost completely precipitated. The XRD spectrum of the ternary CZM catalysts in the reduced state is shown in Figure 1b. It can be seen that the distinct diffraction peaks of reduced-state Cu₂O and Cu appeared at 36.4° and 42.7°, respectively. From the diffraction peak patterns, the states and crystallinity of the copper species were different, which indicated that the CZM catalysts were in different reduced states influenced by different additives. Furthermore, no distinct CuO diffraction peaks were detected in either the CZCe or CZCr catalysts, indicating that they predominantly exist as Cu₂O and Cu, suggesting a highly reduced state.

2.2. Catalyst Surface Morphology and Particle Size

The morphological features of the ternary CZM catalysts were observed by SEM images as shown in Figure 2. The presence of amorphous flocculent structural species was detected in all samples, and was most pronounced in the CZZr catalyst. As seen in Figure 2A, a large number of cubic block particles can be observed in the CZAl catalyst, which may be related to Cu species [23,24]. There were many irregular spherical particles dominated in the CZCe catalyst, which was the common morphological structure in co-precipitation synthesis due to the existence of multiple crystal growth mechanisms, as shown in Figure 2B [25]. The presence of Zr caused the CZZr catalyst to exhibit more irregularities that the specific composition of Cu, Zn and Zr, which may lead to the development of amorphous structures [26]. Although the morphology of the CZCr catalyst in Figure 2D resembled that of the CZCe catalyst, the particles of the CZCr catalyst were notably smaller and more dispersed. The incorporation of Cr likely altered the lattice parameters of the catalyst, leading to a reduction in particle size and a more homogeneous distribution of Cu particles [27].

In order to determine the effect of each element on the particle size of the CZM catalysts, the particle size distribution statistics were carried out by TEM characterization, as shown in Figure 3. Firstly, it can be seen from the TEM images that the general morphology of each catalyst was similar to the results described by SEM images. For example, the CZAl catalyst in Figure 3a exhibits some cubic morphological structures, and the CZCr catalyst in Figure 3d shows smaller spherical particles. By counting the particle size of each catalyst, it was observed that the CZCr catalyst had a smaller particle size (11.1 nm) compared to the others. This suggested that the addition of Cr was more favorable for the formation of catalysts with smaller and more dispersed particles. The CZCe catalyst, also in granular form, had a particle size of about 14.3 nm and was exposed to the same CuO (110) crystal surface as the CZCr catalyst. However, the presence of CeO₂ actually promoted the dispersion of Cu nanoparticles to form smaller particle sizes compared to the CZAl catalyst (19.3 nm) [28]. On the other hand, m-ZrO₂ (111) was detected in the CZZr catalyst based on the lattice stripe spacing, which is one of the main reasons for the amorphous structure of the catalyst [29].

Table 2. Textural properties and particle size of CZM catalysts.

Sample	DCu 1 (%)	SCu 1 (m2/g)	dCu 1/nm	dCu 2/nm	dCu 3/nm
CZAl	6.31	42.69	16.52	17 ± 0.2	19.3 ± 0.4
CZCe	7.52	50.87	13.86	14 ± 0.4	14.3 ± 0.5
CZZr	6.08	41.15	17.13	16 ± 0.2	17.4 ± 0.3
CZCr	9.78	66.17	10.65	11 ± 0.3	11.1 ± 0.3

¹ Via N₂O chemisorption. ² Based on Scherrer equation by XRD. ³ The size of Cu particles counted by TEM.

shows the statistical particle sizes of all samples via XRD and TEM. It can be seen that the overall particle size calculated by the Scherrer formula was slightly smaller than that obtained by TEM image statistics, but the trend was consistent in that the particle size decreased as follows: CZCr < CZCe < CZZr < CZAl. Furthermore, the average particle size calculated from N₂O chemisorption data aligns well with the results derived from XRD and TEM characterization, as summarized in Table 2.

2.3. Catalyst Reduction Behavior by H₂-TPR

To study the reduction property of ternary CZM catalysts, TPR characterization tests were performed, as shown in Figure 4. All samples exhibited two reduction peaks at below 200 °C. This suggests that the process occurred predominantly with the reduction of Cu species, as the reduction of ZnO tended to occur at above 300 °C. It was reported that the reduction of CuO was easier than the reduction of Cu₂O; the apparent activation energy for the reduction of CuO was about 14.5 kcal/mol, while the value was 27.4 kcal/mol for Cu₂O reduction [30]. Specifically, all samples showed two major reduction peaks, labeled L and H. The L-peak was usually associated with the reduction process at lower temperatures, which may be related to the reduction of oxides on the catalyst surface or in smaller particles (dispersed CuO_x species). The H-peak was related to the reduction process at higher temperatures, which may be related to larger particles (bulk CuO_x species) or oxides that are difficult to reduce [31,32,33,34,35]. The CZCr catalyst showed an obvious reduction peak, indicating that the doping of Cr may promote the reduction of Cu oxide. On the contrary, the doping of Al introduced a complex reduction mechanism. In order to further understand the reducibility of the as-prepared catalyst, the fitted TPR curves were back-convolutionally resolved, as listed in

Table 3. TPR peak position temperature and concentration of ternary CZM catalysts.

Sample	Peak L		Peak H
	T/°C	Concentration/%	T/°C	Concentration/%
CZAl	154.3	51.8	180.8	48.2
CZCe	150.9	57.5	164.3	42.5
CZZr	156.8	56.6	175.1	43.4
CZCr	160.0	57.9	169.6	42.1

. It is evident that varying promoters may induce differences in reduction temperatures, consequently influencing the extent of reduction in the ternary CZM catalysts. For instance, the two reduction peaks of the CZCr catalyst were similar compared to the CZAl catalyst, which was related to the interaction between the additives and the copper species. Meanwhile, an increase in the concentration of the high-temperature peak also indicated that more Cu₂O was reduced to Cu.

2.4. Surface Electronic State of Elements on the Catalysts

In order to investigate the surface electronic properties of the ternary CZM catalysts, XPS characterization was carried out, as shown in Figure 5. The ternary catalysts before and after the reaction were first characterized for Cu 2p orbitals. It can be seen from Figure 5a that the electronic states of the surface Cu species were similar for each catalyst before the reaction, with Cu 2p_1/2 located near 954 eV and Cu 2p_2/3 located at near 934 eV. On the other hand, the satellite peaks were observed, indicating that each catalyst was in a highly oxidized state [36]. However, the reduced catalysts exhibited different surface electronic states, as shown in Figure 5b. The reduced CZAl and CZZr catalysts still retained the satellite peaks, but the 2p orbitals shifted to higher binding energies, indicating the gradual transformation of Cu²⁺ into Cu⁰ and Cu⁺. Meanwhile, the CZCe and CZCr catalysts showed strong 2p peaks and the satellite peaks disappeared, suggesting an increase in the content of Cu⁺ and Cu⁰ [37]. Considering the similar binding energies of Cu⁺ and Cu⁰, Cu LMM XAES spectroscopic characterization was chosen to further determine the relative contents of Cu⁺ and Cu⁰. Only the Cu²⁺ wave peak located at ca. 917.7 eV was detected in all cases, as presented in Figure 5c. However, the active components (Cu⁺ and Cu⁰) of the reduced ternary CZM catalysts changed significantly, as shown in Figure 5d. The contents and proportions of Cu⁺ and Cu⁰ in ternary CZM catalysts are listed in

Table 4. Deconvolution results of the Cu LMM XAES of ternary CZM catalysts after reduction.

Sample	Cu+ (%)	Cu0 (%)	Cu+/Cu0
CZAl	75.3	24.7	3.05
CZCe	66.8	33.2	2.01
CZZr	63.8	36.2	1.76
CZCr	46.8	53.2	0.88

. It can be seen that a significant difference appeared in the distribution of active Cu species among the four catalysts, with the CZAl catalyst having a higher relative Cu⁺ content, followed by the CZCe and CZZr catalysts, and with the CZCr catalyst having a relative lower Cu⁺ content, even lower than that of Cu⁰. The Cu⁺/Cu⁰ ratios were as follows: CZAl (3.05), CZCe (2.01), CZZr (1.76) and CZCr (0.88). Hence, it was favorable to investigate the electronic states of each promoter (Figure 6).

In order to investigate the reasons leading to the different distributions of active Cu species (Cu⁺ and Cu⁰) in the ternary CZM catalysts, the third elemental orbitals were characterized and analyzed by XPS, and the results are shown in Figure 6, in which the upper curves are before reduction and the lower curves are after reduction. It is obvious that no significant change occurred for the Al 2p orbitals of the CZAl catalysts and the Zr 3d orbital profiles of the CZZr catalysts before and after reduction. In the Al 2p XPS spectra of the CZAl catalyst, the observed peaks are mainly related to the chemical state of the aluminum species. The peak with a binding energy of 74.4 eV corresponds to the presence of AlOOH species, and in addition, the shoulder peak at 77.0 eV is attributed to Cu 3p_1/2, implying the presence of Cu₂O [38,39,40]. This result depicts that the aluminum species in the catalysts are mainly in the form of aluminum hydroxide (AlOOH) and may have some association with Cu⁺. In the Zr 3d XPS profile of the CZZr catalyst, the binding energy ranges of Zr 3d_3/2 and Zr 3d_5/2 appear in the range of 180~182.5 eV and 182.5~186 eV, respectively, which suggests that zirconium elements were mainly present in the valence form of Zr⁴⁺ [41]. However, the binding energy of the reduced CZZr catalyst Zr 3d gradually shifts to a lower binding energy, indicating the existence of electron transfer from Cu to Zr and an enhanced interaction between Cu and ZrO₂ [42]. In contrast, the Ce 3d orbitals of the CZCe catalysts and the Cr 2p orbitals of the CZCr catalysts changed significantly before and after reduction. Among them, a 10-peak model was fitted to the Ce 3d orbitals, which originated from different Ce oxidation states (Ce³⁺ and Ce⁴⁺) and their 4f electronic configurations [43,44]. From Figure 6b, it can see that the u’ and v’ peaks at 904.2 eV and 885.2 eV represent Ce³⁺ species, while the Ce⁴⁺ peaks are the u₀ peaks at 900.5 eV and the v peaks at 882.0 eV. The Ce³⁺ content of CZCe catalyst after reduction obviously increased, and the conversion from Ce⁴⁺ to Ce³⁺ might have accelerated the generation of oxygen vacancy [45]. On the other hand, chromium oxides mainly existed as Cr(III) and Cr(VI) valence states according to the Cr 2p XPS spectra of the CuZnCr catalyst. The 2p_1/2 and 2p_3/2 binding energies of Cr(III) appear near 586.5 eV and 576.7 eV, while the 2p_1/2 and 2p_3/2 binding energies of Cr(VI) are present near 588.7 eV and 579.4 eV, respectively. Cr(VI) usually has a high electrophilicity and can interact with molecules containing lone pair electrons such as CO₂ to promote the activation and hydrogenation of CO₂ to methanol. Meanwhile, Cr(III) may participate in the reduction process of the catalyst by reacting with hydrogen to generate Cr(II), thus promoting hydrogen activation on the catalyst surface [46]. The presence of Cr(III) on the catalyst’s surface can also affect its acid–base properties and the distribution of active sites, which in turn affect the catalytic performance. It is evident that the peak corresponding to Cr(VI) diminished, while the peak associated with Cr(III) intensified following the reduction process. Studies have indicated that the synergistic interaction between Cr(III) and Cr(VI) species modulates the properties of the catalysts, with Cr(III) stabilizing the catalyst structure and Cr(VI) imparting robust oxidative capabilities [47].

2.5. Desorption Behavior of Reactants on Catalyst Surfaces

To assess the adsorption and activation capacity of the reactants CO₂ and H₂ on ternary CZM catalysts, TPD characterization tests were conducted, and the results are shown in Figure 7. CO₂-TPD mainly detected two peaks with a low-temperature, weak alkaline adsorption peak near 200 °C and a high-temperature, strong alkaline adsorption peak at between 400 and 500 °C, respectively. The weak base sites were related to surface hydroxyl (-OH) species, while strong basic sites were attributed to low-coordination O²⁻ or oxygen defects [48]. In particular, it was obviously noted that the CZCe catalyst exhibited a strong high-temperature desorption peak, which was related to the dense oxygen vacancy. When oxygen vacancies are present, hydrogen atoms may be more inclined to bind to them, forming unstable hydrogen–oxygen complexes whose desorption requires higher energy [49]. The desorption peak intensities of the other three catalysts were comparable; however, the desorption peak temperature of the CZCr catalyst was notably lower. This may mean that the advantages of Cr-doped catalysts in CO₂ adsorption are not obvious. In contrast, for the desorption curve (right) of the other reactant, H₂, the medium–low temperature desorption peak near 300 °C is almost the only peak that can be observed, which corresponded to the hydrogen adsorption on the surfaces of Cu⁺ and Cu⁰, or the interaction between the metal and the oxide. It can be seen that the desorption peak of the CZAl catalyst was small, indicating that the adsorption and dissociation ability of the CZAl catalyst was weak. It was reported that the Al³⁺ species replaced Zn in the ZnO lattice and promoted the formation of oxygen vacancy ($V_O^{¨}$) on the oxide surface [50]. It has been proved that $V_O^{¨}$ has a positive effect on the adsorption of H₂, just like CZCe catalyst, whose H₂ desorption peak is not only lower in strength but also higher in energy barrier than other catalysts. The reason why the CZZr catalyst presented a large H₂ desorption peak area might be that Zr has good H₂ adsorption and dissociation ability [51]. Finally, the CZCr catalyst also maintained a low desorption energy barrier for H₂, which indicated that Cr may affect the adsorption energy of hydrogen by changing the electronic properties of Cu in the catalyst.

2.6. Evaluation Performance of Catalyst

Figure 8 showed the performance evaluation results of the ternary CZM catalyst. Firstly, the catalyst activity for CO₂ hydrogenation to methanol at different reaction temperatures was investigated. As depicted in Figure 8a, the CO₂ conversion rates for all samples exhibit an upward trend with increasing temperature within the range of 200–260 °C. In addition, the CO₂ conversions of all samples were above 10%, which fully reflected the high efficiency of the ternary CZM catalyst. When the full activation temperature is 240 °C, the CO₂ conversion can reach about 20%. In contrast, the selectivity of CH₃OH declined with increasing temperature due to the exothermic nature of the reaction and competition from the by-product CO, as shown in Figure 8b. At low temperatures (200 °C), the highest selectivity of CH₃OH was about 65% over the CZCr catalyst. However, whether it was CO₂ conversion or CH₃OH selectivity, the activity order under this reaction condition decreased as follows: CZCr > CZCe > CZZr > CZAl catalysts. This finding suggests that Cr played a greater role in the ternary CZM catalyst for CO₂ hydrogenation to methanol. Subsequently, the stability test of each catalyst was conducted for 100 h, and the results are shown in Figure 8c,d. It can be seen that although the activity of each catalyst in the reaction process still maintained the previous results, the stability change trend was different. After 100 h reaction, the decline rate of CO₂ conversion and CH₃OH selectivity was in the order of CZCr > CZZr > CZCe > CZAl. Additionally, the CH₃OH yields and deactivation rates of all catalysts after 1 h and 100 h are presented in Figure 8e,f. As illustrated in Figure 8e, the CH₃OH yield of the CZAl catalyst significantly declined from 130.68 mg·g_cat.⁻¹·h⁻¹ to 82.40 mg·g_cat.⁻¹·h⁻¹ after 100 h of reaction, with a deactivation rate of 36.9% (Figure 8f), indicating a relatively severe degradation. This is followed by the CZCe catalyst (21.1% deactivation), the CZZr catalyst (11.6% deactivation) and finally the CZCr catalyst deactivation rate, which was the lowest at only 3.5%. In summary, under the reaction conditions, the CO₂ conversion, CH₃OH selectivity and stability of the CZCr catalyst were superior to those of the other three catalysts. To more accurately evaluate the performance of CZM catalysts, their turnover frequencies (TOFs) were evaluated, as shown in Figure 8g. TOFs were calculated on the basis of Cu sites determined by the N₂O, with conversions below 10.0% achieved by adjusting the GHSV. The results also demonstrate a consistent trend, with the TOF values decreasing in the order CZCr > CZCe > CZZr > CZAl, among which the CZCr catalyst exhibits the highest TOF value.

3. Discussion

The performances of CZM catalysts promoted by different third elements (Zr, Ce, Cr) in CO₂ hydrogenation to methanol is better than those of the traditional CZAl catalyst. At the same time, the structure and electron promotion of the third element were observed during the evolution of the corresponding physicochemical properties of the catalyst. The correlation between the catalytic performance and the physicochemical properties of the ternary CZM catalysts was systematically investigated, as shown in Figure 9. The results showed that the CH₃OH yield was inversely proportional to the particle size of the catalyst. With the increase in Cu species’ particle size, the CH₃OH yield decreased from 194.89 mg·g_cat.⁻¹·h⁻¹ to 130.68 mg·g_cat.⁻¹·h⁻¹, as shown in Figure 9a. Thus, the addition of Cr could form smaller and dispersed particles, which increased the specific surface area of Cu species and improved the catalytic activity [52]. Compared with other catalysts, the particle size effect of the CZCr catalyst was obvious. This was also attributed to the more efficient electronic interaction between Cu and Cr when Cr was introduced as the third element, thereby enhancing the dispersion of Cu. As shown by the reduced catalyst, the electronic states of both Cu and Cr have undergone significant changes. The first was that the CZCr catalyst produces more Cu⁰ than the other catalysts. On the other hand, the catalyst stability was related to Cu species (Cu⁺/Cu⁰) in that the higher the Cu⁺/Cu⁰ ratio, the faster the deactivation rate of the CZM catalysts. This might be attributed to the REDOX reaction between Cu species and reactants during the reaction. As can be seen from Figure 9b, when the Cu⁺/Cu⁰ ratio (3.05) was high, the deactivation rate of the CZAl catalyst reached 36.9%. On the contrary, the deactivation rate of the CZCr catalyst (Cu⁺/Cu⁰ = 0.88) was only 3.5%. This finding revealed that the higher the initial reduction degree of the ternary CZM catalysts, the better the long-term stability. In addition, the apparent activation energy (E_a) of CZM (M = Al, Ce, Zr, Cr) catalysts were tested. Under the action of different third elements, the catalysts showed different intrinsic activities in the reaction. As shown in Figure 9c, compared with the traditional CZAl catalyst (E_a = 53.59 ± 6.45 kJ/mol), other catalysts showed a lower apparent activation energy, especially the CZCr catalyst (E_a = 33.00 ± 3.97 kJ/mol), which could more effectively reduce the activation energy barrier of the reaction and promote the reaction. The introduction of Cr not only optimizes the dispersity of Cu species, but also produces certain electronic interactions with Cu, resulting in a smaller particle size and Cu⁺/Cu⁰ ratio of the CZCr catalyst, which further promotes the adsorption and activation of CO₂ and H₂ (Figure 7), thus reducing the apparent activation energy of the reaction.

For the CO₂ hydrogenation to methanol, the traditional CZAl catalyst has been the most commonly used in the industry. The role of CuZnO as the reaction substrate is indisputable; however, the influence of the third element remains to be thoroughly examined, particularly elements such as Ce, Zr and Cr, which have demonstrated superior promotional effects compared to Al. This phenomenon can also be attributed to the intrinsic role of the third element as a promoter, which significantly influences the physicochemical properties of Cu-based catalysts, encompassing particle size, dispersion, surface electronic state and adsorption activation capacity, among others. More critically, it effectively lowers the activation energy of the reaction.

4. Materials and Methods

4.1. Catalyst Preparation

In this work, a range of ternary Cu-based catalysts, for which the mass ratio was 6:3:1, were prepared by conventional co-precipitation method. Firstly, copper nitrate (Cu(NO₃)₂·3H₂O), zinc nitrate (Zn(NO₃)₂·6H₂O) and auxiliary metal nitrate (among them, Cu(NO₃)₂·3H₂O (99%, AR) and Zr(NO₃)₃·5H₂O) were purchased from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China; Zn(NO₃)₂·6H₂O (99%, AR) was purchased from Xilong Scientific Co., Ltd., Shantou, China; Al(NO₃)₃·9H₂O (99%, AR) was purchased from Tianjin Kemiou Chemical Reagent Co., Ltd., Tianjin, China; and Ce(NO₃)₃·6H₂O and Cr(NO₃)₃·6H₂O (99%, AR) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China). They were dissolved in deionized water to prepare a nitrate mixed solution with a metal ion concentration of 1.0 mol/L. In addition, a certain amount of Na₂CO₃ (99.8%, AR) was purchased from Tianjin Kemiou Chemical Reagent Co., Ltd., Tianjin, China, and was dissolved in distilled water, and the concentration of 1.0 mol/L solution was also prepared. Under stirring at 70 °C, the metal salt solution and precipitant solution were simultaneously added into 200 mL of distilled water. During this period, the drop rate of sodium carbonate was adjusted to control the pH of the suspension to 7.0 ± 0.2. After the drip addition, the mixture was stirred and aged for 1 h, then filtered, washed, dried overnight at 80 °C to obtain the catalyst precursor, and then calcined at 350 °C for 4 h to achieve the catalyst parent, which was recorded as CZM (M = Al, Ce, Zr, Cr).

4.2. Catalyst Characterization

X-ray fluorescence spectroscopy (XRF) characterization was performed on a Zetium instrument produced by Malvern Panalytical B.V., Almelo, The Netherlands, to determine the elemental composition of the catalysts. The method employed in this study was the pelletizing method. During the pelletizing process, boric acid (H₃BO₃) was used as a binder to improve the formability, uniformity and stability of the samples, while also reducing matrix effects and measurement errors.

The ¹H NMR spectrum of methanol was measured using an AVANCE NEO 400 MHz instrument manufactured by Bruker, Berlin, Germany, with deuterated chloroform (CDCl₃) as the solvent. The required magnetic field strength was 400 MHz.

XRD measurements were carried out using a TD-M001 diffractometer produced by Bruker, Berlin, Germany, for copper radiation. XRD patterns were collected by performing a stepwise scan at a rate of 8° per minute, covering a 2θ range extending from 10° to 80°. The mean crystallite sizes of Cu nanoparticles were determined using the Scherrer equation as follows.

$$d_{\mathrm{Cu}}={\displaystyle \frac{k\lambda }{\beta cos\theta }}$$

The specified parameters include the full width at half maximum (β), the diffraction angle (θ), the Scherrer constant (k = 0.89) and the X-ray wavelength (λ = 0.154 nm), in that order.

The dispersion and morphology of Cu were analyzed using a Regulus-8100 scanning electron microscope from Hitachi, Tokyo, Japan, which was furnished with an energy-dispersive X-ray spectrometer. A 40 mg sample was attached to a sample tray and characterized under accelerating voltage conditions.

The micro-morphology and structure of the catalyst were characterized by transmission electron microscopy (TEM). The experimental equipment was a field-emission transmission electron microscope (TEM) of the JEM-2100F model, produced by JEOL Ltd., Tokyo, Japan, with an acceleration voltage of 200 kV. A specific quantity of catalyst powder was incorporated into an ethanolic solution. After ultrasonic treatment for 40 min, the ensuing suspension was spread onto a carbon-supported copper mesh and left to dry in the air.

H₂-TPR was performed using a FINESORB-3010 instrument produced by Beijing Jingwei Gaobo Science and Technology Co., Ltd., Beijing, China. A 40 mg portion of the sample was introduced into a U-shaped quartz reactor, which was subsequently flushed with argon (Ar) at a temperature of 300 °C for 1 h. Once the system reached ambient temperature, the gas stream was replaced with a 10 vol% H₂/Ar flow (30 mL/min). Subsequently, the temperature was incrementally increased to 600 °C at a gradient of 10 °C/min.

CO₂-TPD analysis was performed on an instrument of the Micromeritics AutoChem 2920 model, produced by Micromeritics Instrument Corporation, Norcross, GA, USA. Initially, a 60 mg catalyst was loaded into a U-shaped quartz tube and pretreated at 50 °C under a continuous argon (Ar) flow. After removing surface impurities, the catalyst was reduced by injecting 10% H₂ at 250 °C. The gas supply was subsequently switched to pure CO₂ and maintained under these conditions for 1 h. Following this period, the system was purged with argon (Ar) for 30 min to eliminate any residual gases. Ultimately, the catalyst was subjected to a temperature ramp from 50 °C to 800 °C, conducted at a rate of 10 °C/min. The H₂-TPD analysis was performed following the same methodology as the CO₂-TPD procedure, with the exception that a gas mixture containing 10 vol% H₂ in Ar was utilized.

XPS and XAES analyses were conducted on a state-of-the-art ESCALab 220i-XL VG instrument, produced by Thermo Fisher Scientific, Waltham, MA, USA. Excitations of the energy spectra were achieved with Al Kα radiation, which had a characteristic wavelength of 1486.6 eV. For the sake of enhancing the accuracy of the measurements, the binding energy (BE) readings were carefully aligned with reference to the C1s peak, set at 284.6 eV, with a precision of ±0.1 eV. The positions of photoelectron peaks are reported in terms of binding energy (BE), while the Auger peaks are referenced to the kinetic energy (KE). In order to prevent severe oxidation of copper-based catalysts, our XPS tests were performed in a low-temperature and dry environment.

Room-temperature Raman spectroscopy was conducted using a high-resolution LabRAM HR system from HYJ, Cannes, France, equipped with a CCD detector and a spectral resolution of 1 cm⁻¹. The system utilized a 532 nm laser as the excitation source, coupled with 1800 grooves per millimeter grating in a backscattering configuration. The laser power was set to 6 mW, with an exposure time of 60 s. The wavenumber range spanned from 200 to 800 cm⁻¹.

The dispersity of Cu species was determined using N₂O chemisorption on a Micromeritics AutoChem 2920 system. A 40 mg sample was heated from 50 °C to 500 °C (10 °C/min) in a flow of 10% H₂/Ar fluid (30 mL/min), and we recorded the H₂ consumption as X. After reduction, the sample was cooled down to 50 °C and re-oxidized at 50 °C with N₂O (30 mL/min) for 30 min. Subsequently, the sample was flushed with a pure Ar fluid for 30 min to remove the physically adsorbed N₂O. The catalyst was then reduced again according to the above procedure and the H₂ consumption was recorded as Y. The dispersion of Cu species and the size of exposed Cu particles were calculated using the following formulas.

CuO + H₂ → Cu + H₂O; H₂ consumption = X

(1)

N₂O + 2Cu → N₂ + Cu₂O

(2)

Cu₂O + H₂ → 2Cu + H₂O; H₂ consumption = Y

(3)

$${\mathrm{D}}_{\mathrm{Cu}}(\%)={\displaystyle \frac{2\mathrm{Y}}{\mathrm{X}}}\times 100\%$$

$${\mathrm{S}}_{\mathrm{Cu}}(\%)={\displaystyle \frac{{2\times \mathrm{Y}\times \mathrm{N}}_{\mathrm{A}}}{{\mathrm{X}\times \mathrm{M}}_{\mathrm{Cu}}\times 1{.4\times 10}^{19}}}={\displaystyle \frac{1353\mathrm{Y}}{\mathrm{X}}}{ (\mathrm{m}}^2/\mathrm{g})$$

$${\mathrm{d}}_{\mathrm{Cu}}={\displaystyle \frac{104.25}{{\mathrm{D}}_{\mathrm{Cu}}}}\times 100\%$$

where ${\mathrm{D}}_{\mathrm{Cu}}$ represents the dispersion of Cu species, ${\mathrm{S}}_{\mathrm{Cu}}$ denotes the surface area of Cu species, ${\mathrm{M}}_{\mathrm{Cu}}$ is the molar mass of Cu, ${\mathrm{N}}_{\mathrm{A}}$ stands for Avogadro’s constant (6.022 × 10²³ mol⁻¹) and 1.4 × 10¹⁹ corresponds to the number of Cu atoms per unit area (/m²).

4.3. Catalyst Evaluation

The catalytic performance was evaluated using a continuous fixed-bed reactor. A 0.5 g sample of catalyst, mesh size 20–40, was initially charged into a stainless steel reaction tube with dimensions of a 9 mm diameter and a 1450 mm length. The catalyst bed was then uniformly surrounded by quartz sand of equivalent particle size, ensuring a consistent environment above and below the bed. Before the reaction, the catalyst was reduced to 4 h under atmospheric pressure in 10% H₂/N₂ in 60 mL/min at 250 °C. Following the initial setup, a gas mixture with a composition of 72 vol% H₂, 24 vol% CO₂ and 4 vol% Ar was introduced into the reaction tube. The pressure was rapidly accumulated to 3 MPa, followed by heating the reactor between 200 and 260 °C to study the temperature’s effect. CO and CO₂ concentrations were determined using GC (Gas Chromatography, Manufacturer: PANNA, Model: A91, Manufacturer: Changzhou Panna Instrument Co., Ltd., Changzhou, China) with a TCD and two columns: a 5A molecular sieve and a GDX-104 packed column. Methanol was detected using an FID (Flame Ionization Detector) equipped with a Porapak-Q column and further characterized by ¹H NMR. The test was conducted using deuterated chloroform (CDCl₃) as the solvent, and the results are shown in Figure S3. The spectrum clearly displays the characteristic signals of methanol, including two main peaks: Peak A corresponds to the methyl hydrogen (-CH₃) with a chemical shift range of 3.3–3.5 ppm and Peak B corresponds to the hydroxyl hydrogen (-OH) with a chemical shift range of 2.0–4.0 ppm, exhibiting a broad peak shape. The integration ratio of these signals is approximately 3:1, consistent with the expected structure of methanol. Ar was included in the feedstock as a reference for analytical purposes. The CO₂ conversion rates and selectivity of the products for each catalyst were recorded. The calculation methods were as follows:

$$X_{C{O}_2}(\%)=\left(1-{\displaystyle \frac{A_{out\left(C{O}_2\right)}/A_{out\left(Ar\right)}}{A_{in\left(C{O}_2\right)}/A_{in\left(Ar\right)}}}\right)\times 100\%$$

where $A_{in\left(C{O}_2\right)}$ and $A_{in\left(Ar\right)}$ are the pre-reaction peak areas of CO₂ and Ar; $A_{out\left(C{O}_2\right)}$ and $A_{out\left(Ar\right)}$ are the post-reaction peak areas of CO₂ and Ar.

$$S\left(CO\right)(\%)={\displaystyle \frac{A_{out\left(CO\right)}/A_{out\left(Ar\right)}\times f_{CO/Ar}}{\left(A_{in\left(C{O}_2\right)}/A_{in\left(Ar\right)}-A_{out\left(C{O}_2\right)}/A_{out\left(Ar\right)}\right)\times f_{C{O}_2/Ar}}}\times 100\%$$

$$S\left(C{H}_{3}OH\right)(\%)=\left({\displaystyle \frac{A_{C{H}_{3}OH}\times f_{C{H}_{3}OH}}{\left(A_{in\left(C{O}_2\right)}/A_{in\left(Ar\right)}-A_{out\left(C{O}_2\right)}/A_{out\left(Ar\right)}\right)\times f_{C{O}_2/Ar}}}\right)\times 100\%$$

where $A_{out\left(CO\right)}$ and $A_{C{H}_{3}OH}$ are the peak areas of CO and CH₃OH, respectively. $f_{CO/Ar}$, $f_{C{O}_2/Ar}$ and $f_{C{H}_{3}OH}$, respectively, correspond to the correction factors of CO, CO₂ and CH₃OH.

The formula for calculating the turnover frequency ($TOF$) of methanol is as follows:

$$TOF \left(h^{-1}\right)={\displaystyle \frac{F_{C{O}_{2,in}\times }{X}_{C{O}_2\times }{S}_{C{H}_{3}OH}\times M_{Cu}}{W_{Cu}\times D_{Cu}}}$$

where $F_{C{O}_{2,in}}$ (mol h⁻¹) is the molar flow rate of CO₂ in pristine mixed gas; $M_{Cu}$ is the molecular weight of copper, 63.546 g mol⁻¹; and $W_{Cu}$ and $D_{Cu}$ (%) are the catalyst weight and the actual loading of copper, respectively.

5. Conclusions

In summary, the impact of various promoters (Al, Ce, Zr, Cr) on the catalytic performance was systematically investigated. The results indicate that the optimal CZCr catalyst exhibited a higher CO₂ conversion of 21.1%, CH₃OH selectivity of 56.4% and superior stability with only 3.57% deactivation over 100 h, under conditions of 3 MPa and 240 °C. The characterization results showed that the physical and chemical properties of the catalyst greatly changed under the action of different accelerators. The third element affected the dispersion and interaction degree of Cu species in different degrees, resulting in different particle sizes and Cu⁺/Cu⁰ ratios. The smaller particle size led to higher methanol yield, while the lower Cu⁺/Cu⁰ ratio resulted in a lower deactivation rate. As a whole, the introduction of high-quality third elements (such as Cr) can reduce the apparent activation energy of the ternary CZM catalysts for the hydrogenation of CO₂ to methanol. The preliminary research also provides an important way to improve the energy conversion efficiency reasonably and develop the catalyst for industrial application.