Enhanced CuAl₂O₄ Catalytic Activity via Alkalinization Treatment toward High CO₂ Conversion during Reverse Water Gas Shift Reaction

Abstract

CO₂ catalytic conversion to CO would likely be an important part of CO₂ mitigation and utilization. In this work, CuAl₂O₄ was developed with a spinel structure that acts as an active and stable catalyst for this reaction. Here, the fundamental characteristics of CuAl₂O₄ catalyst were studied to understand the catalytic mechanism for the Reverse Water Gas Shift reaction. Based on the catalytic mechanism, the CuAl₂O₄ catalyst was found to have exceptional catalytic activity due to the high dispersion of copper on its surface, and it could have higher catalytic activity by increasing the oxygen vacancies on the surface of the catalyst via alkalinization treatment. By combining with XPS spectra, the relationship between the Raman mode and the oxygen vacancy structure on the CuAl₂O₄ surface was proved. Through these studies, it was proved that alkalinization treatment can regulate the oxygen vacancies on the surface of the catalyst and thus enhance the catalytic activity.

1. Introduction

Excess CO₂ emissions have caused catastrophic global environmental impacts, such as climate change and ocean acidification [1]. To mitigate this threat, developing techniques for CO₂ capture and utilization has become an urgent and significant task. Catalytic conversion of CO₂ to value-added chemicals is currently being investigated extensively. Despite significant advances in direct CO₂ conversion to methanol as a green fuel, this process still results in low methanol yield due to the high stability of CO₂ [2,3]. As an alternative, the conversion of CO₂ to CO, which occurs in thermal catalysis via the Reverse Water Gas Shift (RWGS) reaction, has drawn more and more attention. Unlike thermodynamically stable CO₂, CO is much more active and it can be further converted into a variety of chemicals [4,5].

During the RWGS reaction, Cu-based and other transition/noble metallic catalysts have been widely employed due to their relatively high activity and high selectivity [6,7,8,9,10]. Previous studies have shown that increasing the dispersity of Cu can significantly improve the catalytic activity of Cu-based catalysts, because the dispersity of Cu affects the Cu-O interface, which is widely accepted as the active site of the RWGS reaction [11]. However, the traditional Cu-based catalysts are easily deactivated by the fierce aggregation of supported Cu particles at high temperatures, and their regeneration requires a complex activation process [12,13,14,15]. As a result, the poor stability of Cu-based catalysts is the primary barrier to their industrial application. How to enhance Cu dispersion and anchor the Cu particles over the substrate to prevent high-temperature sintering is a major concern for the construction of effective and long-lifetime RWGS catalysts.

Spinel oxides (AB₂O₄), in which cations are located at the tetrahedral and octahedral holes in a dense cubic packing of oxygen anions, have always been a research hotspot in the field of catalytic materials, mainly because of their high mechanical resistance, high thermal stability, low temperature sinter ability, low surface acidity, and high ability of cation diffusion. These materials exhibit high performance in various reactions such as water gas shift, CO oxidation, and steam reforming of dimethyl ether, ethanol, and methanol [16]. However, most copper-based spinel oxides (e.g., Cu-Fe, Cu-Mn, Cu-Cr spinel oxides) show a poor copper dispersion on the surface of materials. Therefore, finding an appropriate B-site metal to improve the copper dispersion on the surface of materials is a major concern for the application of copper-based spinel materials in RWGS reactions. Li et al. [17] investigated H₂ production from methanol decomposition using Cu-Al spinel catalysts, and the results indicated that CuAl₂O₄ was a reservoir of Cu, slowly releasing Cu and preventing Cu from quick sintering. Generally, the spinel with a CuAl₂O₄ composition consists of a divalent copper ion (Cu²⁺) usually occupying a tetrahedral site and a trivalent aluminum ion (Al³⁺) normally occupying the octahedral sites of a cubic oxide lattice. It is precisely because of the special structure of the spinel that the Al³⁺ can effectively prevent the sintering of copper. However, the catalytic behavior of Cu-Al spinel oxides in the RWGS reaction is still unclear.

In addition to the dispersion of copper on the material surface, the catalytic activity of catalysts is also affected by oxygen vacancies [18]. Recently, more and more studies have been conducted on regulating the oxygen vacancies on the catalyst surface to improve catalytic activity. Based on previous literature, many methods are employed to generate oxygen vacancies, such as vacuum calcining, doping, solid solution generation, hydrogen treatment, nitric acid treatment, plasma treatment, and alkaline treatment [19]. Bahmanpour et al. [18], through coprecipitation of copper and aluminum, followed by hydrogen treatment, obtained a Cu-Al spinel, and proved that it has remarkable stability and higher activity in the RWGS reaction. Subsequently, they employed alkaline treatment to create oxygen vacancies in the Co-Al catalyst and tested it in the RWGS reaction. The results indicated that the oxygen vacancies were vital for CO₂ adsorption and activation on the spinel surfaces. However, increasing the oxygen vacancies of Cu-Al spinel by alkali treatment toward its reactivity in RWGS has not been investigated.

Herein, in this study, enhanced CuAl₂O₄ catalytic activity via alkalinization treatment toward high CO₂ conversion during RWGS reaction was investigated. The objectives of this study were (1) to investigate the catalytic activity of CuAl₂O₄ via alkalinization treatment or not in the RWGS reaction, (2) to explore the catalytic mechanism of CuAl₂O₄ in the RWGS reaction, and (3) to verify that the increase in oxygen vacancies is beneficial to enhance CuAl₂O₄ catalytic activity in the RWGS reaction.

2. Results and Discussion

2.1. Catalytic Activity

The catalytic performance in the RWGS reaction of the CuAl₂O₄ catalysts was evaluated at various temperatures under 60,000 mL/g_cat/h. As shown in Figure 1a, the CO₂ conversion increased with increasing reaction temperature. The CO₂ conversion on the CuAl₂O₄ catalyst reached 37.72% when the reaction temperature increased from 300 °C to 400 °C. The significant increase in CO₂ conversion is predominantly because the CO₂ RWGS reaction is a reversible thermodynamic endothermic reaction. According to the thermodynamic equilibrium, increasing the temperature is conducive to the balance movement in the direction of generating CO to improve CO₂ conversion. Collision theory states that the amount and movement velocity of activated molecules can increase with increasing reaction temperature. This phenomenon increases the effective collision probability between active molecules and accelerates the reaction to promote CO₂ conversion [20]. Therefore, CO₂ conversion over the CuAl₂O₄ catalysts increased with increasing reaction temperature. However, the increase in catalytic performance between 350 °C and 400 °C is higher than that between 300 °C and 350 °C. The reason could be because the surface of the CuAl₂O₄ catalyst begins to decompose and precipitate highly dispersed CuO. Meanwhile, some of the CuO is reduced to highly dispersed Cu⁰ by H₂. During the reaction, the isolated surface Cu²⁺ and the highly dispersed Cu⁰ can serve as active sites, promoting the improvement of catalytic activity [21]. However, the effect of the reduction degree of copper on the catalytic performance is relatively complex. When copper is in the state of CuO, it plays the role of catalyst, and its catalytic effect is mainly based on the intermediate mechanism; when copper is reduced to Cu⁰, it also plays the role of catalyst, and its catalytic effect is mainly based on the redox mechanism [22,23]. Therefore, its specific effect needs to be further studied.

As shown in Figure 1a, the CuAl₂O₄ catalyst kept 100% CO selectivity during the reaction temperature range. These results could be attributed to the following reasons: transition metals generally display more d-orbital holes, i.e., unoccupied d orbitals in the d energy band, strong adsorption abilities for reactants, and bond forming abilities, which can promote the absorption and activation of H₂ molecules to form active hydrogen species. Consequently, deep hydrogenation of CO₂ molecules is promoted to generate CH₄, and the generated CO molecules can also facilitate the hydrogenation reaction on the active centers of the corresponding catalyst to form CH₄ molecules [24]. However, Cu has no d-orbital holes, that is, the d orbits of Cu are completely occupied, which is not conducive to the adsorption and bonding of H species. By contrast, the O of CO₂ has lone pair electrons, which can pair with the d electrons of Cu⁰ [20]. This characteristic can be beneficial to the adsorption and activation of CO₂ molecules. Therefore, the CuAl₂O₄ catalyst not only had a high CO₂ conversion rate but also exhibited 100% CO selectivity. CO₂ conversion rates of the CuAl₂O₄ catalysts were analyzed and compared to some referenced catalysts in

Table 1. Comparison of CO₂ conversion rate and CO selectivity for the as-prepared and referenced catalysts.

Catalyst	Temperature (°C)	Pressure (MPa)	Rate (×10−5 molCO2/gcat/s)	CO Selectivity (%)	Ref.
CuAl2O4	400	0.1	9.36	100	This work
Cu/Al2O3	600	0.1	4.9	100	[25]
Cu/SiO2	500	0.1	3.34	100	[13]
Cu−Fe/SiO2	500	0.1	6.7	100	[13]
Cu−K/SiO2	600	0.1	7.3	100	[26]
Cu/CeO2	400	0.1	2.38	100	[20]
Ni/CeO2	400	0.1	5.37	31.7	[20]

. The results indicated that the CuAl₂O₄ catalyst had higher conversion rates at relatively lower temperatures. The reason is due to the CuAl₂O₄ catalyst outperforming the conventional supported copper-based catalyst in copper dispersion, which is particularly conducive to carbon dioxide adsorption and activation.

The stability of the CuAl₂O₄ catalyst at different temperatures is shown in Figure 1b. It can be noted that no detectable deactivation was observed within 20 h, and the activity of the CuAl₂O₄ catalyst gradually decreased after 20 h at 400 °C. Therefore, the CuAl₂O₄ catalysts (reaction for 0, 20, and 40 h) were characterized to analyze the catalytic and inactivation mechanisms, respectively. In comparison, the catalysts were rapidly deactivated within 10 h at 500 and 600 °C. This is due to the low Taman temperature of copper oxide. For example, the Taman temperatures of CuO, Cu₂O, and Cu are 526 °C, 481 °C, and 405 °C, respectively. Therefore, the catalyst is susceptible to agglutination and sintering at a high temperature of more than 400 °C, which could lead to the reduction in its redox activity and even affect its catalytic performance.

2.2. Catalyst Characterization

In order to investigate the variations of R0, R20, and R40 catalysts (reaction for 0, 20, and 40 h, respectively), H₂-TPR was conducted in the temperature range of the catalytic activity tests. As shown in Figure 2a, two peaks were identified on the catalyst before the reaction. The first peak (180–210 °C) is assigned to the reduction of the bulk CuO crystals, produced during catalyst preparation. This is because the CuAl₂O₄ spinel is formed under the solid–solid interaction between CuO and Al₂O₃. The CuAl₂O₄ would inhibit the diffusion of CuO into the Al₂O₃ support when enough CuAl₂O₄ spinel is formed. The second peak (350–400 °C) is assigned to reduction of the CuO obtained from the decomposition of CuAl₂O₄. Therefore, it can be inferred that the catalyst begins to decompose and precipitate CuO when the reaction temperature reaches 350 °C. However, Agzamova et al. [27] reported that the CuAl₂O₄ spinel could decompose into CuO and Al₂O₃ at 800 °C. The different decomposition and reduction temperatures of CuAl₂O₄ spinel may be due to the presence of hydrogen in this study. In the case of R20 and R40, the reduction peaks appeared at (150–170 °C), (180–210 °C), and (220–240 °C), corresponding to the reduction of highly dispersed CuO, the bulk CuO, and Cu₂O [28]. It can also be noted that the second peak of R20 was higher than that of R40, and the third peak of R20 was lower than that of R40, which indicated that the main component of the catalyst changed from CuO to Cu₂O with the passage of reaction time.

In order to analyze the changes in non-reducing components in the catalyst with the reaction, XRD characterization was analyzed and the results are shown in Figure 2b. The peaks at 2θ of 31.29°, 36.87°, and 44.98° are attributed to the diffraction of the CuAl₂O₄ spinel structure (Joint Committee on Powder Diffraction File No. 33-0448). The peaks at 2θ of 35.67° and 38.71° are attributed to the diffraction of CuO. The peaks at 2θ of 29.48° and 36.26° are attributed to the diffraction of Cu₂O. This result is similar to H₂-TPR, which revealed that the main component of the catalyst before the reaction was CuAl₂O₄, and then began to decompose into CuO, and CuO content gradually decreased while Cu₂O content gradually increased as the reaction progressed. However, in addition to the patterns of the CuO and Cu₂O, diffractions of metallic Cu⁰ appeared at 43.21° and 50.34° in the R20 and R40. This indicates that the CuO produced by CuAl₂O₄ decomposition was partially reduced to Cu⁰ by H₂.

In order to better understand the valence state change of Cu (Cu⁰/Cu⁺/Cu²⁺) on the surface of the catalyst during the reaction, the surface structure was investigated using XPS (as shown in Figure 3a). Deconvolution of the acquired spectra for the Cu 2p 3/2 and Cu 2p 1/2 regions for all catalysts showed the presence of mainly a Cu²⁺ peak before reaction (933.7 eV). In the case of R20, the Cu²⁺ peak in the catalyst spectra dramatically reduced, showing partial conversion of Cu²⁺ to Cu⁰ (931.9 eV) and Cu⁺ (932.8 eV) [29]. The intensity of the Cu⁰ peak was higher than that of Cu⁺, which was the opposite in the case of R40. This indicates that the CuO produced by CuAl₂O₄ decomposition may be reduced to Cu⁰ by H₂ before being converted to Cu₂O, and the Cu⁰ content decreased as the reaction progressed during the deactivation of the catalyst.

Moreover, the structure of catalysts with different reaction times was characterized by using SEM. As shown in Figure 3b, the crystal particles of R0 were smooth, small, and numerous, which resulted in the acceptable catalytic performance of the CuAl₂O₄ catalyst. The surface of the R20 crystal particles began to precipitate out the fine particles. This is because the catalyst surface gradually decomposed and precipitated highly dispersed CuO under the condition of high temperature reduction reaction, and the CuO would be further reduced. Meanwhile, the size of the crystal particles became larger because the catalyst crystal particles agglomerated at high temperatures. In the case of R40, the precipitated particles on the catalyst surface became larger and the catalyst structure basically sintered. This is the reason why the catalytic performance gradually weakens with the passage of reaction time.

2.3. Alkalinization of the Catalyst

According to the catalytic mechanism of the catalyst, regulating the oxygen vacancy on the surface of the catalyst may improve the catalytic activity of the catalyst.

In order to confirm the effect of oxygen vacancy density on the performance of the CuAl₂O₄ catalysts in the RWGS reaction, the CuAl₂O₄ catalyst was treated with NaOH of different concentrations to create oxygen vacancies on the surface of the CuAl₂O₄ catalyst. After treatment with NaOH solution, N0, N0.25, N0.5, N0.75, and N1 (soaked with 0, 0.25, 0.5, 0.75, and 1 mol/L of NaOH, respectively) was washed thoroughly to avoid the presence of Na on the catalyst surface [30].

To confirm the formation of the oxygen vacancies, the O 1s spectra of the treated catalysts were collected using XPS. Figure 4 shows that the O 1s spectra of all catalysts are deconvoluted into two main peaks. The first peak is assigned to the lattice oxygen (O_L) on the surface, while the second peak presents the surface adsorbed oxygen (O_A), which can form due to the adsorption of O₂ on an oxygen vacancy [18]. By comparing the O_A/(O_A + O_L) ratio, the O_A peak clearly increased due to the alkalinization, which confirmed that the alkalinization increased the oxygen vacancies on the catalyst surface. In addition, the O_A/(O_A + O_L) ratio is lower for N1-deactivated (0.12) compared to N1 (0.38). This indicates that the deactivation of catalysts is partly due to the reduction in oxygen vacancies.

Previous studies have shown that oxygen vacancies may affect the Raman mode of crystals. Liu et al. [31] studied the surface oxygen vacancy position in SnO₂ nanocrystals by Raman spectroscopy and suggested that the Raman mode does not shift in frequency but increases in intensity as the number of oxygen vacancies increases. Therefore, Raman spectra were also collected to study the oxygen vacancies’ effect on the surface structure of all catalysts, and the results are shown in Figure 5a.

All catalysts show a Raman peak at 295.3 cm⁻¹, which is attributed to Cu-O of CuAl₂O₄ [32]. The peak shifted by comparing the Raman spectra of N1 and N1-deactivated. This is because most of the CuO in the catalyst was converted to Cu₂O after deactivation. It can also be noted that both N0.5 and N1 catalysts showed obvious Raman peaks at 619.8 cm⁻¹ and 780.3 cm⁻¹, the intensity of which clearly increased due to the alkalinization. These peaks were not observed in the case of N1-deactivated. Combined with the O 1s spectra of XPS, these two peaks were attributed to the lattice vibration caused by oxygen vacancies. It proved that the alkalinization increased the oxygen vacancies of the catalyst.

To demonstrate the effect of oxygen vacancy on catalytic activity, all catalysts were subsequently tested at the same conditions for the RWGS reaction. As shown in Figure 5b, the catalytic activity of catalysts clearly increased due to the alkalinization. When the concentration reached 0.5 mol/L, the increasing trend gradually flattened out, indicating that the limiting factor at this time is no longer the number of oxygen vacancies, but may be the dissociation rate of carbon dioxide. Meanwhile, the increase trend of catalytic activity decreased significantly with the catalytic reaction temperature exceeding 350 °C. Combined with the catalytic mechanism, it can be concluded that CuAl₂O₄ was gradually decomposed into CuO when the reaction temperature exceeded the decomposition temperature of the catalyst, and the oxygen vacancy structure was destroyed so that the catalytic activity of the catalysts with different concentrations of alkalinization had no significant difference. Therefore, the regulation of oxygen vacancies is more suitable for RWGS catalytic reactions at relatively low temperatures.

2.4. Possible Catalytic Mechanism

Recently, researchers have employed in situ infrared, isotope tracer experiments, and DFT theoretical calculations to investigate the adsorption species on the catalyst surface in the RWGS reaction and inferred that the catalytic mechanism is mainly divided into surface redox mechanism and intermediate mechanism [33,34,35,36]. The main difference between these is whether the active H species is involved in the generation of carbon-containing intermediates (formate, carboxylate, etc.).

For the redox mechanism, the concept of the oxidation and reduction cycle was suggested. CO₂ oxidized Cu⁰ and generated CO. H₂ reduced Cu⁺ to Cu⁰ and formed H₂O. CO can be produced from CO₂ alone over the Cu catalyst. The reactant, H₂, plays the role of activating the catalyst for the reduction of Cu²⁺ → metallic Cu⁰ and regenerating the catalyst for the reduction of Cu₂O [37]. Zhang et al. [29] studied the RWGS reaction on Cu/β-Mo₂C and suggested that the redox process dominated the reaction mechanism. The proposed reactions occurring in this study are shown below:

CO₂ + 2 Cu⁰ ↔ Cu₂O + CO

(1)

H₂ + Cu₂O ↔ 2 Cu⁰ + H₂O

(2)

In addition to variable-valence metal sites, oxygen vacancies can also act as active sites for the redox mechanism. CO₂ adsorbs on the oxygen vacancies on the oxide surface and dissociates to produce CO. The activated dissociated H species reduce the oxygen species filled on the oxygen vacancies, realizing the regeneration of oxygen vacancies and the production of H₂O, and finally completing the whole redox process [38]. This explains why the catalyst had some catalytic activity before the reaction temperature reached the temperature at which the catalyst decomposes. The proposed reactions occurring in this study are shown below:

CO₂ + O_v ↔ CO + O*

(3)

H₂ ↔ 2 H*

(4)

2 H* + O* ↔ H₂O + O_v

(5)

The other model, the intermediate mechanism, suggested that CO could be formed from the decomposition of the formate intermediate (HCOO∗) derived from the association of hydrogen with CO₂. The HCOO∗ species could be decomposed into CO and OH∗, and these OH∗ species could lead to the production of H₂O. Bahmanpour et al. [39] studied the RWGS reaction on Cu/Al₂O₃ and observed formate on the catalyst surface during the reaction by in situ DRIFTS. The proposed reactions occurring are shown below:

CO₂ + O_v ↔ CO₂*

(6)

H₂ ↔ 2 H*

(7)

CO₂* + H* ↔ HCOO*

(8)

HCOO* ↔ CO* + OH*

(9)

CO* ↔ CO + O_v

(10)

OH* + H* ↔ H₂O

(11)

Therefore, based on the above analysis, the possible catalytic mechanism of the CuAl₂O₄ catalyst was mainly divided into two stages (as shown in Figure 6). In the first stage, when the reaction temperature was lower than the decomposition temperature of the catalyst, the catalytic activity of the CuAl₂O₄ catalyst was mainly based on the intermediate mechanism of oxygen vacancies as the active site. According to the H₂-TPR profile of R0, the Cu²⁺ on the catalyst surface has not been converted to Cu⁰ at this time. The catalytic activity is mainly limited by the number of oxygen vacancies and dissociation rate of CO₂. In the second stage, when the reaction temperature reached the decomposition temperature, the CuAl₂O₄ began to decompose and precipitate highly dispersed CuO, which was gradually reduced to Cu⁰ by H₂. This is a disadvantage of common metal oxide catalysts—the metal oxide in the catalyst is easily reduced to a metallic state during the reaction process, resulting in catalyst deactivation. However, the CuAl₂O₄ catalyst turned this disadvantage into an advantage, in which CuO was reduced to Cu⁰ and used as a supported metal catalyst to further catalyze the RWGS reaction through the redox mechanism, thus continuing and improving its catalytic performance. As the active site of redox mechanism, the increase in Cu⁰ leads to the further improvement of the catalytic activity of the catalyst. However, the surface structure of the catalyst gradually changes with the passage of reaction time. The catalyst crystal particles begin to agglomerate and sinter, leading to the reduction in oxygen vacancies in the catalyst. At this time, there is relatively more highly dispersed Cu⁰ on the surface of the catalyst, which leads to the main active site of the catalytic reaction being transferred from the oxygen vacancy to the variable metal Cu⁰, so that the catalytic activity can still be maintained at a high level. Therefore, the catalyst still has a high CO₂ conversion rate after 20 h of reaction at 400 °C. However, with the passage of reaction time, the catalytic activity of the catalyst decreased gradually. There are two main reasons for catalyst deactivation. The first is that the agglomeration and sintering of the catalyst crystal particles on the surface became more and more serious, which can be seen in Figure 3b, leading to a decrease in the number of active sites of the catalyst. Secondly, with the sintering of catalyst, it is difficult for Cu⁰ to be regenerated by H₂ after being oxidized to Cu₂O, which can be seen in Figure 3a.

3. Experimental Method

3.1. Catalyst Synthesis

We employed the coprecipitation method followed by calcination for CuAl₂O₄ catalyst synthesis. As the copper and aluminum precursors, Cu(NO₃)₂∙3H₂O and Al(NO₃)₃∙9H₂O with a 2:1 M ratio of Al:Cu were dissolved in deionized water and mixed. A 1 M aqueous solution of Na₂CO₃ as a precipitant was added dropwise to the salt mixture at room temperature. The resulting suspension was stirred at 75 °C for 6 h, filtered, and washed thoroughly, after which the formed sample was dried overnight at 120 °C. Subsequently, the dried powder was calcined at 900 °C for 5 h, which is the lowest temperature for stable spinel formation. Alkalinization was employed to create further oxygen vacancies in the CuAl₂O₄ catalyst. The alkaline-treated CuAl₂O₄ catalyst was prepared by treating 0.2 g of CuAl₂O₄ with 30 mL of various concentrations NaOH aqueous solution (0, 0.25, 0.5, 0.75, 1 mol/L) at room temperature for 30 min. The resulting suspension was centrifuged and washed thoroughly and then dried at 120 °C for 4 h.

3.2. Characterizations

3.2.1. X-ray Diffraction (XRD)

XRD patterns were collected on a powder diffractometer (Empyrean, PANalytical, Almelo, Netherlands) using Cu Kα radiation (λ = 0.1541 nm). Samples were tested on 2θ = 10–90° at a scan rate of 10° min⁻¹.

3.2.2. H₂-Temperature-Programmed Reduction (H₂-TPR)

Temperature-programmed reduction (TPR) experiments were performed in a packed bed flow microreactor (AutoChem II 2920, Micromeritics, Norcross, GA, USA). Samples of 50 mg mixed with 200 mg of SiO₂ were first pretreated under He at 300 °C for 1 h. The samples were then cooled to 50 °C and heated from 50 °C to 800 °C at 10 °C/min under 10% H₂/Ar.

3.2.3. Scanning Electron Microscopy (SEM)

Samples were characterized by Field Emission Scanning Electron Microscopy (HITACHI-S-4700, Hitachi, Tokyo, Japan). Samples were observed at 10,000 and 50,000 × magnification.

3.2.4. X-ray Photoelectron Spectra (XPS)

X-ray photoelectron spectra were acquired using a scanning XPS microprobe (K-Alpha, Thermo Scientific, Waltham, MA, USA). When the sample chamber pressure was less than 2.0 × 10⁻⁷ mbar, the sample was sent to the analysis room with the spot size of 400 μm, the working voltage of 12 kV, and the filament current of 6 mA. OriginPro 2022b (64-bit 9.9.5.167) was applied for peak deconvolution.

3.2.5. Raman Spectra

Samples were characterized by a confocal Raman electron microscope (Lab RAM HR800, Horiba, Kyoto, French). Raman spectra of samples were recorded in the range of 50–3400 cm⁻¹ with a 2 cm⁻¹ spectral resolution.

3.3. Activity Tests

The catalytic performance evaluation was performed in a continuous flow fixed-bed quartz reactor at atmospheric pressure. The reactor was a quartz tube with an inner diameter (ID) of 35 mm, in which 50 mg of catalyst (40−60 mesh) diluted with inert SiO₂ (200 mg) was sandwiched between quartz wool layers. At the start of each test, the as-prepared catalysts were treated with 20 mL/min H₂ flow at 300 °C for 1 h. The main purpose of hydrogen pretreatment is threefold; the first is to dry the water vapor adsorbed on the surface of the catalyst, the second is to exhaust all the air in the gas chamber, and the third is to reduce the CuO impurities produced in the preparation process to avoid the interference of CuO impurities with the catalytic performance of CuAl₂O₄ in the first stage. After that, the catalysts were exposed to the feed gas (a stream of H₂/CO₂ at 50 mL/min with a 2:1 ratio). Stephen Dzuryk and Ebrahim Rezaei [40] studied the influence of gas ratio on the catalytic performance of CuO/ZnO/Al₂O₃ catalyst in the RWGS reaction system. This research selected a gas ratio with relatively strong catalytic performance and relatively low hydrogen concentration, which is more in line with the economic benefits of industrial production. Meanwhile, the further use of gas products (such as Fischer–Tropsch synthesis, etc.) is also considered. At each temperature, the products were analyzed after 60 min of steady-state reaction. The product was analyzed by using a gas chromatograph (GC-2014, SHIMADZU) with a TCD detector. The intake gas was first switched from the feed gas to argon after the reaction. After the temperature dropped to normal temperature, the catalyst was taken out and collected, and weighed after vacuum drying. After weighing, the catalyst was put into the sample bag with desiccant.

Performances of the catalysts were characterized in terms of CO₂ conversion. The CO₂ conversion was defined as the molar ratio of CO₂ removed in the reaction to the feed CO₂. CO₂ conversion, CO selectivity, and CO₂ conversion rate were defined as follows:

$${\mathrm{X}}_{{\mathrm{CO}}_2}\%=\frac{n_{{\mathrm{CO}}_2}^{in}-n_{{\mathrm{CO}}_2}^{out}}{n_{{\mathrm{CO}}_2}^{in}}\times 100$$

(12)

$${\mathrm{S}}_{\mathrm{CO}}\%=\frac{n_{\mathrm{CO}}^{out}}{n_{\mathrm{CO}}^{out}+n_{{\mathrm{CH}}_4}^{out}}\times 100$$

(13)

$$r=\frac{F\times {\mathrm{X}}_{{\mathrm{CO}}_2}}{W}$$

(14)

where $n_{\mathrm{CO}}^{out}$, $n_{{\mathrm{CO}}_2}^{out}$, and $n_{{\mathrm{CH}}_4}^{out}$ are the flow rates of CO, CO₂, and CH₄ in the outlet, and $n_{{\mathrm{CO}}_2}^{in}$ is the flow rate of CO₂ in the feed. F is the CO₂ flow rate (mol/s) and W is the catalyst weight (g).

4. Conclusions

In the present work, the CuAl₂O₄ catalyst of spinel structure was prepared. The catalyst showed exceptional activity at temperatures higher than 300 °C, reaching a conversion rate of 9.36 × 10⁻⁵ mol CO₂/g_cat/s and 100% of CO selectivity at 400 °C in the RWGS reaction. A complete change path of the CuAl₂O₄ catalyst composition and structure in the reaction system was learned through the characterization of the catalyst. The influence of regulating oxygen vacancies on the catalytic activity by alkalinization was studied using XPS and Raman spectroscopy. In addition, the relationship between the Raman mode and the oxygen vacancy structure on the CuAl₂O₄ surface was also proved. The possible catalytic mechanism of the RWGS reaction was analyzed. The results revealed that in the first stage of reaction, the intermediate mechanism was predominant, and in the second stage, the redox mechanism was predominant. In both stages, oxygen vacancies and active metals act as active sites. Therefore, the CuAl₂O₄ catalyst has better catalytic activity because of the high dispersion of copper on its surface, and alkalinization can regulate the oxygen vacancy structure on the catalyst surface to enhance catalytic performance.