Efficient Oxidative Dehydrogenation of Ethylbenzene over K/CeO₂ with Exceptional Styrene Yield

Highlights

What are the main findings?

• A styrene yield of 91.4% was found for 10% k/CeO₂ at 500 °C and CO₂-O₂ mixed atmosphere. The excellent catalytic performance of 10% k/CeO₂ is attributed to the alkali metal oxide modified cerium oxide and carbon dioxide induced oxygen vacancies to promote the dehydrogenation of ethylbenzene.

What is the implication of the main finding?

• The proposed ODH strategy by using oxygen vacancies enriched catalysts offers an important insight into the efficient dehydrogenation of ethylbenzene at mild conditions.

Abstract

Oxidative dehydrogenation (ODH) is an alternative for styrene (ST) production compared to the direct dehydrogenation process. However, ODH with O₂ or CO₂ suffers from either over-oxidation or endothermic property/low ethylbenzene conversion. Herein, we proposed an ODH process with a CO₂-O₂ mixture atmosphere for the efficient conversion of ethylbenzene (EB) into styrene. A thermoneutral ODH is possible by the rationalizing of CO₂/O₂ molar ratios from 0.65 to 0.66 in the temperature range of 300 to 650 °C. K modification is favorable for ethylbenzene dehydrogenation, and 10%K/CeO₂ achieved the highest ethylbenzene dehydrogenation activity due to the enhanced oxygen mobility and CO₂ adsorbability. The catalyst achieved 90.8% ethylbenzene conversion and 97.5% styrene selectivity under optimized conditions of CO₂-4O₂ oxidation atmosphere, a temperature of 500 °C, and a space velocity of 5.0 h⁻¹. It exhibited excellent catalytic and structural stability during a 50 h long-term test. CO₂ induces oxygen vacancies in ceria and promotes oxygen exchange between gaseous oxygen and ceria. The ethylbenzene dehydrogenation in CO₂-O₂ follows a Mars-van Krevelen (MvK) reaction mechanism via Ce³⁺/Ce⁴⁺ redox pairs. The proposed ODH strategy by using oxygen vacancies enriched catalysts offers an important insight into the efficient dehydrogenation of ethylbenzene at mild conditions.

1. Introduction

Styrene was an important organic raw material for the production of plastics and synthetic rubber [1]. By the end of 2021, the global total styrene production capacity was close to 39.8 million tons/year, and that of China was 14.789 million tons/year. China has become the largest styrene production country in the world. At present, the catalytic dehydrogenation process of ethylbenzene accounts for about 90% of the global styrene production. However, it still suffers from challenges including the high feed ratio of steam to ethylbenzene, the limited conversion rate by thermodynamic equilibrium, complicated product separation, etc., which leads to high energy consumption [1]. Therefore, the development of a new process to improve the production efficiency of styrene and solve the problems existing in the catalytic dehydrogenation process of styrene is urgently required.

In recent years, researchers have focused on the ethylbenzene ODH with O₂ [2,3,4,5], N₂O [6,7,8] and CO₂ [9,10,11,12,13,14,15,16,17,18,19] as oxidants. The advantage of ethylbenzene ODH is that the hydrogen produced in the reaction process can be completely oxidized into water by the oxidant, which avoids the limitation of the equilibrium conversion rate in traditional dehydrogenation and makes full use of the heat generated in hydrogen combustion, thus reducing the process energy consumption [1]. Ethylbenzene ODH with CO₂ as an oxidant could convert CO₂ into CO while dehydrogenating, which could inhibit the deep oxidation of hydrocarbons and provide a new way for the preparation of styrene and the resource utilization of CO₂. At present, the active catalysts for the ODH reaction of CO₂ ethylbenzene were mainly doped/mixed oxides of Fe [20,21], V [22,23], Co [24], Ce [10,16,25,26], Mn [27], and Ti [28]. CO₂ ethylbenzene ODH follows the Mars-van Krevelen redox mechanism [23], but CO₂ activation as a mild oxidant and/or oxygen transfer agent faced obvious challenges in heterogeneous catalysis due to its high kinetic inertia and thermodynamic stability [29,30]. The consumed surface lattice oxygen could not be effectively replenished, which led to the loss of active sites and the deactivation of catalysts [31,32]. Therefore, it was crucial to ensure the reversibility of lattice oxygen on the catalyst surface effectively for the stability of the catalyst [9].

CeO₂ reveals excellent oxygen storage/release ability and high activity in ethylbenzene ODH because of its rapid Ce⁴⁺/Ce³⁺ redox cycle [25,26]. It was reported that CeO₂ exhibited significant activity in ethylbenzene ODH with O₂ [33]. However, due to the safety problems and excessive oxidation of O₂, the selectivity of styrene remained low. The ODH reaction of ethylbenzene using CO₂ as an oxidant was effective in improving the yield of styrene and inhibiting the deposition of coke. In a CO₂ atmosphere, the consumed lattice oxygen could not be effectively replenished. Therefore, O₂ was combined with CO₂ to replenish the consumed lattice oxygen of CeO₂, thus promoting the ethylbenzene conversion in the ODH reaction. Currently, the oxidative dehydrogenation of ethylbenzene using CO₂-O₂ remains unknown. Besides, it was also reported that LaSrFe promoted by alkali metal Li/Na/K could significantly improve the product selectivity of ODH by the modification of lattice oxygen [34]. CeO₂ could improve the catalytic activity of the Fe₂O₃-K₂O-based catalyst, accelerating the formation of styrene [35]. Thus, it is expected that the modification of CeO₂ using alkali metals (Li/Na/K) might promote the ethylbenzene conversion in ODH in a mixed atmosphere of CO₂-O₂.

In this paper, composite oxide catalysts Li/CeO₂, Na/CeO₂, and K/CeO₂ synthesized by co-precipitation and over-impregnation methods were investigated in the ethylbenzene ODH process to explore the effect of alkali metals’ modification on CeO₂. The candidate selected for the catalyst to explore the effect of the oxidants and the reaction temperature was 10%K/CeO₂. In combination with the surface composition and microstructure of the catalyst, the reaction pathway of ethylbenzene over the 10%K/CeO₂ catalyst was clarified further. The 10%K/CeO₂ catalyst exhibited excellent ODH activity and stability in a CO₂-O₂ mixed atmosphere at 500 °C. The reaction of ethylbenzene with O₂ was the main reaction, and the reaction of ethylbenzene with CO₂ was the auxiliary reaction with the following equations.

C₈H₁₀ + 0.5 O₂ = C₈H₈ + H₂O

(1)

C₈H₁₀ + CO₂ = C₈H₈ + CO + H₂O

(2)

The proposed ceria-based catalyst and novel ODH process might offer a new insight into the improvement of ethylbenzene conversion with high styrene selectivity.

2. Results and Discussion

2.1. Theoretical Analysis of Thermoneutral Oxidative Dehydrogenation

Currently, most CO₂/EB in CO₂ ethylbenzene ODH is >10:1 [9,18,22,25,36], and the high CO₂ usage led to the complicated subsequent product separation and low ethylbenzene conversion. It has been reported that O₂ ethylbenzene ODH was an exothermic reaction with high ethylbenzene conversion, but it suffers from over-oxidation and low selectivity [2]. CO₂ ethylbenzene ODH is a heat-absorbing reaction which can avoid deep oxidation and achieve high styrene selectivity, but low ethylbenzene conversion [25]. If the CO₂-O₂ mixed atmosphere is used as the oxidant for EB ODH, it will achieve the goals of reducing the amount of CO₂, simplify the product separation, improve the conversion of ethylbenzene, and reduce the energy consumption of the reaction. If this solution is feasible, it will be of great significance for the study of mixed oxidants for EB-ODH.

The enthalpy changes of the reaction of ethylbenzene with CO₂ or O₂ alone was simulated by integrating the heat absorption and exotherm of the reaction. As shown in Figure 1, the specific heat uptake/exhaustion values of the enthalpy change of the reaction of ethylbenzene with CO₂ or O₂ alone, and the enthalpy change of the reaction of both was within one order of magnitude, and the simulation that resulted provided theoretical data support for the reaction of the CO₂-O₂ mixed atmosphere with ethylbenzene to reach the thermoneutral condition.

In order to reduce the energy consumption of CO₂ ethylbenzene ODH, the enthalpy change of the reaction between CO₂/O₂ and ethylbenzene ODH was simulated in HSC by using a fixed molar amount of CO₂. In combination with O₂-ODH and CO₂-ODH, an enthalpy change ΔH = 0 (dynamically 0) will be obtained to achieve an ideal thermoneutral condition. The ratios of each reactant obtained are shown in

Table 1. Ratio of molar amounts of each reactant required for the enthalpy change ΔH = 0 for the reaction with EB oxidative dehydrogenation under the mixed atmospheric conditions of CO₂ and O₂.

Temperature (°C)	XC8H10 (g) + CO2 (g) + YO2 (g) = XC8H8 (g) + CO (g) + XH2O (g)
	EB (mol)	CO2 (mol)	O2 (mol)	ΔH (kJ) (Dynamic to 0)
300	2.3226	1	0.6613	0
350	2.3245	1	0.6622	0
400	2.3256	1	0.6628	0
450	2.3259	1	0.6629	0
500	2.3254	1	0.6627	0
550	2.3239	1	0.6619	0
600	2.3214	1	0.6607	0
650	2.3176	1	0.6588	0

. The results showed that it was feasible to achieve the theoretical thermoneutral conditions for the reaction of the CO₂-O₂ mixed atmosphere with ethylbenzene ODH. It shows that O₂-to-CO₂ molar ratios that ranged from 0.65 to 0.66 are able to achieve a thermoneutral condition. The reaction equation under mixed atmosphere at 500 °C is:

2.3254 C₈H₁₀ (g) + CO₂ (g) + 0.6627 O₂ (g) = 2.3254 C₈H₈ (g) + CO (g) + 2.3254 H₂O (g)

(3)

2.2. Synthesis of K/CeO₂ Catalyst

2.2.1. Effect of Alkali Metals Modification on CeO₂

The reaction performance of alkali metal Li/Na/K loaded CeO₂ catalysts were evaluated by EB-TPR experiments. Figure 2a shows the XRD patterns of CeO₂ with different alkali metals loaded (X = Li/Na/K), with the surface loading of CeO₂ by Li/Na/K. There was no obvious shift for the characteristic peaks of CeO₂ (PDF 78-0694), which indicated that the loading of alkali metals did not change the crystalline phase of CeO₂. The peaks corresponding to Li/Na/K oxides were also not observed in the XRD patterns, suggesting that the oxides of alkali metals were evenly distributed on the CeO₂ surface. The intensity of the characteristic peaks for the Li-loaded CeO₂ was larger than that of all other samples, which meant the grain size of CeO₂ became larger after Li was loaded.

The reduction ability of CeO₂ with different alkali metal modifications was tested for H₂-TPR, as shown in Figure 2b. At 638 °C, the reduction peak for CeO₂ was due to the reduction of lattice oxygen [37,38]. The reduction peak area and the intensity of Li/Na/K loaded CeO₂ was much stronger than CeO₂. It illustrated that the amount of reducible oxygen was increased after alkali metal modification, which corresponded to the significant enhancement in the amount of lattice oxygen [13]. This also demonstrated that the loading of alkali metals could greatly enhance the oxygen storage capacity of the catalyst.

Figure 2c showed the styrene generation in EB-TPR (EB-temperature program reduction) experiment at 300–600 °C. As shown in the results, styrene started to be generated at 350 °C on the pure CeO₂, and the generation amount of styrene was increased rapidly after the temperature reached 500 °C. For 10%K/CeO₂, it started to generate styrene from 400 °C. Styrene’s generation performance increased sharply after the temperature reached 450 °C, while the styrene generation rate slowed down after 550 °C. In contrast, there was almost no styrene generation on the Li/Na modified CeO₂. This showed that the oxygen storage capacity was not a decisive factor to enhance the catalytic activity. Figure 2d shows that 10%K/CeO₂ exhibited a strong rising trend of the H₂O mass spectrum after 450 °C. Combined with the rising trend of the CO₂ mass spectrum (Figure S1), this suggests that the lattice oxygen release rate of 10%K/CeO₂ was accelerated by the increase in reaction temperature, which led to the partial oxidation. Therefore, it was necessary to select a suitable reaction temperature to reduce the excessive oxidation of EB on the catalyst. These results reveal that the doping of the alkali metal Li/Na/K to the CeO₂ surface could greatly improve its oxygen storage capacity.

From Figure 3a, the amount of styrene generation was gradually increased at 500 °C with the reaction proceeds. Figure 3b shows the product distribution of the oxidative dehydrogenation of ethylbenzene. The generation amounts of by-products on 10%Li/CeO₂ were the highest, with only 54.5% styrene selectivity and 17.3% ethylbenzene conversion. The styrene selectivity of 10%Na/CeO₂ could reach 86.9%, while ethylbenzene conversion was only 29.0%. The best catalytic performance was from 10%K/CeO₂ with 92.4% styrene selectivity and 71.9% ethylbenzene conversion. The catalytic performance of ethylbenzene oxidative dehydrogenation was related to the type of alkali metal. CeO₂, with the doping of Li and Na, suffered from a serious decrease in catalytic performance and an increase in excessive oxidation. Therefore, alkali metal K was chosen for the surface modification of CeO₂ to further improve the catalytic performance.

2.2.2. Effect of K Loading Amounts on CeO₂

Figure 4a indicates the XRD pattern of the K/CeO₂ catalyst with different loading K content. As Figure 4a shows, the catalysts of CeO₂ and n%K/CeO₂ (n = 5, 10, 15, 20) appeared at the CeO₂ characteristic diffraction peaks (PDF 78-0694), which indicates that all of the catalysts are mainly of CeO₂ phase with a fluorite structure. The diffraction peak of the phase associated with potassium was not detected either because potassium is uniformly distributed on the surface of CeO₂, or the loading did not reach the detection limit of XRD.

The H₂ temperature programmed reduction (H₂-TPR) test was performed to detect the reducibility of the CeO₂ and n%K/CeO₂. The H₂-TPR curves are shown in Figure 4b. The curve of CeO₂ was one reduction peak at 638 °C, which was ascribed to the reduction of the CeO₂ surface lattice oxygen [13,37,38]. When a small amount of K was loaded (5%K/CeO₂), the reduction peak at 638 °C shifted to a lower temperature (583 °C), and the intensity of the peak increased slightly. Compared to the pure CeO₂, the reduction temperature of the surface lattice oxygen rose by about 30 °C for the 10%K/CeO₂ catalyst, but the peak intensity was significantly increased. The results suggest that the K modification of CeO₂ promoted the release of the CeO₂ surface lattice oxygen and the following hydrogen consumption. The temperature of the reduction peak at 638 °C increases continuously with the K loading. However, the amount of hydrogen consumption did not obviously add as well as temperature did. The explanation may be that the overloading of potassium led to the enhanced interaction between K₂O and CeO₂, which inhibited the release of the surface lattice oxygen. The H₂-TPR results show that the catalytic performance of K/CeO₂ in the oxidative dehydrogenation of ethylbenzene can be promoted only when the appropriate amount of K was loaded.

The basicity of the catalyst surface plays a key role in the activation of ethylbenzene C-H with CO₂ during ethylbenzene ODH. The CO₂ temperature programmed desorption (CO₂-TPD) was conducted to test the surface basicity of the n%K/CeO₂ catalysts, as shown in Figure 5. For the pure CeO₂, the CO₂ desorption peak appeared at temperatures between 70–200 °C, which was assigned to the weak base site. After loading the alkaline metal potassium, the CO₂ desorption peak was detected between temperatures of 600–700 °C. As expected, the K loading increased the alkaline sites on the K/CeO₂ catalyst surface [39]. The 10%K/CeO₂ has the strongest desorption peak, which indicates that the 10%K/CeO₂ catalyst reveals the strong adsorption and desorption capacity of CO₂, which might be the beneficial ethylbenzene conversion.

TEM was used to observe the effect of K loading on the morphology of CeO₂. As shown in Figure 6a–c, both K modified and unmodified CeO₂ reveal an irregular polygonal structure, which was consistent with the XRD test results. Lattice fringes with d values of 0.19 nm and 0.27 nm were observed on the 10%K/CeO₂ and 20%K/CeO₂ catalysts, corresponding to the (311) and (200) crystal planes of K₂O, respectively. This indicated that K was uniformly distributed on the surface of CeO₂ in the form of K₂O, while the heavy load of K made the interaction between K₂O and CeO₂ too strong, which was not conducive to the release of surface lattice oxygen. This further conformed the detection results of H₂-TPR. Therefore, an appropriate load of K might be conducive to the ethylbenzene ODH.

In order to study the effect of K loading on the reactivity of K/CeO₂ catalysts, the ethylbenzene dehydrogenation tests were conducted at 500 °C. Figure 7a shows the mass spectrum trends of styrene production over the K/CeO₂ catalyst with different K loading amounts. The production rate of styrene at 10%K/CeO₂ was significantly higher than that of other catalysts after 5 min of reaction, and reached a peak at about 25 min. As seen in the product distribution diagram of each catalyst in Figure 7b, 10%K/CeO₂ shows the highest dehydrogenation activity among all samples, achieving a high styrene yield of 71.9%. Combined with the above characterization, it might be speculated that surface K₂O is conductive to ethylbenzene activation and enhanced dehydrogenation activity. In the industrial ethylbenzene dehydrogenation process, surface K species in the commercial K-Fe-O catalysts have been found to promote the formation of KFeO₂ in the active phase from iron oxide, while K⁺ ions promoted the polarization of Fe-O bonds and enhanced the alkalinity of the active center [40], after which they enhanced the selectivity of styrene.

2.3. Ethylbenzene ODH Performance

2.3.1. Optimization of Reaction Conditions in ODH

The effect of oxidants on the ODH performance over 10%K/CeO₂ was investigated, as shown in Figure 8. In the CO₂ atmosphere (orange icon), the high ethylbenzene conversion at 80% and styrene selectivity (>95%) were obtained at CO₂ excess coefficients of 3CO₂ and 4CO₂. In the O₂ (red icon) atmosphere, high reactivity was obtained at O₂ excess coefficients of 4O₂. However, the styrene selectivity remained relatively low compared with the CO₂ atmosphere. This is because that, as a strong oxidant, higher oxygen partial pressure is conducive to ethylbenzene conversion, but it will also lead to excessive oxidation and reduce styrene selectivity [41,42]. According to Hongbo et al. the lattice oxygen released within the catalyst controls the reaction process, the C₃H₈ conversion decreases with decreasing oxygen concentration, and the selectivity of C₃H₆ increases due to decreasing peroxidation [43]. In Figure 8, CO₂-4O₂, 3CO₂-4O₂ and 4CO₂-4O₂ all have stable and efficient ethylbenzene conversion due to the high oxygen partial pressure provided by 4O₂. Conversely, as a soft oxidant, CO₂ will inhibit excessive oxidation and thus improve styrene selectivity, but with relatively low ethylbenzene conversion.

Considering the advantages of O₂ and CO₂ oxidants in ODH, the dehydrogenation performance of the CO₂-O₂ mixture (Blue icon) was studied as oxidants in ODH over the 10%K/CeO₂ catalyst. Among them, CO₂-4O₂ shows the highest dehydrogenation activity with 90.8% ethylbenzene conversion and 97.5% styrene selectivity due to the synergistic effect of CO₂ and O₂. On the one hand, CO₂ dilutes the O₂ concentration, thus suppressing the excessive oxidation and improving styrene selectivity. On the other hand, O₂ in CO₂ enhances ethylbenzene conversion in the CO₂ atmosphere.

Combined with the mass spectrum curves of CO, H₂O and O₂ (Figure S2), EB-ODH with O₂ (Equation (1)) in a CO₂-4O₂ atmosphere is the main reaction which dominates ethylbenzene dehydrogenation to generate styrene, while EB-ODH with CO₂ (Equation (2)) in a CO₂-4O₂ atmosphere is the secondary reaction which reduces the concentration of O₂ and suppresses excessive oxidation. Figure 8b compares the EB-ODH performance of 10%K/CeO₂ under the CO₂-4O₂ atmosphere at different temperatures and space velocities. It is obvious that 10%K/CeO₂ shows the highest dehydrogenation activity in the CO₂-4O₂ atmosphere, following with 90.8% ethylbenzene conversion and 97.5% styrene selectivity under 500 °C and 5.0 h⁻¹ space velocities.

2.3.2. Reaction Stability

The long-term stability of the 10%K/CeO₂ catalyst was tested under the optimal experimental conditions of CO₂-4O₂, a reaction temperature of 500 °C, and a space velocity of 5.0 h⁻¹ for 50 h, as shown in Figure 9. The average ethylbenzene conversion and styrene selectivity remain at approximately 96.1% and 94.2%, respectively. In a 50 h long-term stability test, the styrene yield reaches as high as 91.4%. The 10%K/CeO₂ catalyst exhibited excellent activity and stability in the 50 h long-term test, which was clearly superior to the ODH performance reported in the available literature [10,12,15,25,26,44]. Furthermore, a slight decrease in the conversion of ethylbenzene of about 2% was observed after the reaction reached 43 h.

The XRD and H₂-TPR were conducted to investigate the evolution of the structure and reducibility of the catalyst before and after the 50 h stability test. Figure 10a shows the XRD plots of the samples before and after the 50 h stability test for 10%K/CeO₂. In Figure 10a, the characteristic peaks of CeO₂ (PDF 78-0694) were observed for both the fresh and spent catalyst. The intensity of the main diffraction peaks of the spent catalyst were increased after the stability test, suggesting the crystallization of ceria. Furthermore, a weak diffraction peak at 2θ = 31.8° corresponding to K₂CO₃ was observed for the spent catalyst. The formation of carbonate on the surface of ceria might lead to a slight decrease of ethylbenzene conversion in a long-term stability test. Figure 10 shows that one weak reduction peak was observed at 563 °C for the spent catalyst, corresponding to the reduction of Ce⁴⁺ on the surface. Compared to fresh catalyst, the reduction peak was obviously decreased and shifted to higher temperature due to the decreased reducibility. This will be ascribed to the formation of carbonate on the surface of ceria, which hindered oxygen mobility and donation ability.

Figure 11 shows the TEM and EDS images of the spent catalyst after the 50 h stability test. As shown in Figure 11a, both the fresh and spent catalyst are composed of hexagonal crystals with a size of approximately 40 nm. CeO₂ mainly exposes the (111) crystalline surface and K₂O mainly exposes the (200) crystalline surface. After the long-term stability test, the boundary of ceria grain was corroded and the grain grew bigger compared to fresh ones due to the slight sintering of material in ODH (Figure S3). The EDS mapping suggests that all elements (Ce, O and K) are well dispersed and no obvious agglomeration was observed. The Ce distribution is well-matched with the morphology of ceria grain in the catalyst, while K tends to disperse in a large region. This reveals that K₂O or K₂CO₃ are dispersed on the surface of ceria grain. Those results confirmed the high structural stability during the ODH process.

2.4. Surface Reaction Mechanism

The surface reaction of ethylbenzene over the 10%K/CeO₂ catalyst was explored by a temperature programmed surface reaction equipped with a Fourier transform infrared spectrometer. In Figure 12a, several band features of ethylbenzene were observed after the adsorption of ethylbenzene/Ar flow at 100 °C. The bands at 3040 and 3020 cm⁻¹ correspond to the C-H extension in the aromatic ring. The bands at 2980 and 2940 cm⁻¹ correspond to methyl and methylene in ethyl. The band at 1600 cm⁻¹ was ascribed to the skeletal C-C stretching pattern of the aromatic ring. All of those bands suggest that ethylbenzene was absorbed on the surface of the catalyst. In Figure 12b, the very weak band corresponding to the vinyl overtone was observed at 1760 cm⁻¹, which reveals that ethylbenzene was dehydrogenated into styrene [25] at temperatures as low as 200 °C. The bands of surface CO₂ and carbonate were observed, which might be ascribed to the full oxidation of hydrocarbons or adsorption from ambient CO₂. According to the band feature at 200 °C, ethylbenzene was absorbed on the surface the catalyst, and then reacted with surface absorbed gaseous oxygen to produce styrene and/or complete oxidation products (H₂O and CO₂). The reaction behaviors of ethylbenzene on the surface of the catalysts are similar under different oxidation atmospheres.

The appearance of reduced cerium bands (corresponding to Ce³⁺ and oxygen vacancies) was observed at different oxidizing atmospheres starting from 400°C. In addition, the peak of the reduced cerium is stronger than the others (O₂ and Ar) in both the CO₂ and CO₂-O₂ atmospheres, indicating that more oxygen vacancies were generated in these two atmospheres. CeO₂ with higher oxygen kinetics promotes the oxidation of carbonate, and there is an inhibitory effect on the formation of carbonate which promotes the reaction; the dehydrogenation of alkanes occurs on the CeO₂ surface and subsequently generates oxygen vacancies on the CeO₂ surface. The role of CeO₂ lattice oxygen in the partial oxidation of the reaction is crucial for oxidative dehydrogenation [45]. O₂, the main force of CeO₂ lattice oxygen supplementation, formed less oxygen vacancies on the CeO₂ surface than CO₂ and the mixed CO₂-O₂ atmosphere, indicating that CO₂ has an induced effect on the formation of oxygen vacancies on the CeO₂ surface. With the increasing reaction temperature, the band of reduced ceria at 500 °C became stronger with a similar intensity of the catalysts under four atmospheres due to the enhanced oxygen mobility at elevated temperatures. The induced oxygen vacancies in the catalyst will accelerate the consumption of lattice oxygen and the replenishment of oxygen vacancies using O₂ or CO₂, thus improving the dehydrogenation activity, even though CO₂ will combine with K₂O to produce some carbonates on the surface of ceria. The catalyst retained its high activity in the CO₂-O₂ atmosphere. In this CO₂-O₂ mixture, it can be predicted that oxygen vacancies are the driving force of oxygen exchange during ODH, while O₂ is the main oxygen source for lattice oxygen replenishing. CO₂ plays a very important role in inducing oxygen vacancies and enhancing oxygen exchange during ethylbenzene ODH. It can also be speculated that the ethylbenzene ODH with CO₂-O₂ follows a Mars-van Krevelen reaction mechanism by using Ce³⁺/Ce⁴⁺ redox pairs.

3. Materials and Methods

3.1. Materials

Cerium nitrate hexahydrate (Ce(NO₃)₃·6H₂O, AR), Sodium hydroxide (NaOH, AR), Lithium nitrate (LiNO₃, AR), Sodium nitrate (NaNO₃, AR), Potassium nitrate (KNO₃, AR), Cetyl trimethyl ammonium bromide (CTAB, AR), Ethylbenzene (C₆H₅CH₂CH₃, AR), styrene (C₆H₅CH₂, AR), Benzene (C₆H₆, AR), and Toluene (C₆H₅CH₃, AR) were purchased from Aladdin Industrial Inc. O₂/Ar (9.99 vol% O₂), CO₂/Ar (5.02 vol% CO₂), high purity Ar (≥99.99 vol% Ar), high purity He (≥99.99 vol% He), and liquid nitrogen were purchased from the Kunming Mercer Gas Co., Ltd. (Kunming, China)

3.2. Catalyst Synthesis

About 4.34 g Ce(NO₃)₃·6H₂O and 2.18 g CTAB were dissolved in 200 mL water, and then 300 mL, 0.16 mol·L⁻¹ NaOH solution (1.92 g) was added to the above mixture and stirred at 70 °C for 24 h. After stirring, the water bath was aged at 90 °C for 3 h, and then the CTAB was removed by hot filtration to obtain a light yellow solid. The solid was dried at 100 °C for 6 h and then calcined at 600 °C for 4 h in a muffle furnace to obtain CeO₂ samples.

The prepared CeO₂, the different mass percentage nitrate, and 15 mL of deionized water were placed in a 50 mL beaker, placed on a magnetic stirrer at room temperature until the water was completely evaporated, and dried in a 100 °C oven for 6 h. The dried samples were calcined in a muffle furnace at 600 °C for 4 h, ground, pressed, granulated (60–80 mesh) and named n%X/CeO₂ (X means nitrate, X = Li, Na, K; n% means different mass percentages = 5, 10, 15, 20).

3.3. Catalyst Characterization

The phase structure and lattice size of the catalyst were characterized by X-ray diffraction (XRD; D8 Advance, Bruker, Karlsruhe, Germany), where λ was 0.15406 nm, and the tube current and voltage were 40 kV and 30 mA, respectively. The samples were scanned at a scanning rate of 5°/min, and the scanning area was selected from between 10° to 90°.

The Chembet Pulsar automated chemisorption instrument (Pulsar, Quantachrom Instruments, Boynton Beach, FL, USA) was used to study the reducibility of the catalyst by temperature programmed reduction (H₂-TPR). The 50 mg sample was loaded into a quartz reactor, pretreated at 300 °C for 30 min, and then cooled to 50 °C under a continuous flow of Ar (50 mL·min⁻¹). After pretreatment, the temperature was changed from 50 °C to 850 °C at a rate of 10 °C/min in a mixed flow of 10 vol% hydrogen (50 mL·min⁻¹) in argon. The H₂ consumption was recorded by an online thermal conductivity detector (TCD, Quantachrom Instruments, Boynton Beach, FL, USA).

The basicity of the catalyst was studied by CO₂-temperature programmed desorption (CO₂-TPD). The 30 mg sample was placed in a quartz reactor and pretreated at 300 °C for 30 min. The catalyst was then cooled to 100 °C under a continuous flow of Ar (50 mL·min⁻¹). After pretreatment, 10 vol% CO₂ was kept in the mixed flow of Ar (50 mL·min⁻¹) for 30 min. Subsequently, the excess gas was purged with Ar, and the temperature increased from 100 °C to 900 °C at a rate of 10 °C/min. The desorption amount of CO₂ was recorded by TCD.

The morphology and crystallinity of fresh and recycled catalysts were detected by transmission electron microscopy (JEM 2100F, JEOL (Beijing), Beijing, China), and the lattice spacing was detected by high-resolution transmission electron microscopy (JEM 2100F). The element content on the surface of the used catalyst was detected by an energy dispersive spectrometer (EDS, XFlash 5030T, JEOL (Beijing), Beijing, China).

In situ diffuse reflectance infrared Fourier transform spectroscopy (Nicolet iS50 FTIR, Termo Scientific, Waltham, MA, USA) was used to test the changes of functional groups during the reaction. The samples were first loaded into an in situ cell, and the background was collected in an Ar atmosphere at 25 °C. During the temperature-programmed in situ infrared test, the sample was raised to 100 °C in an Ar flow, and ethylbenzene was adsorbed on the sample surface by EB/Ar, EB/(CO₂ + Ar), EB/(O₂ + Ar), and EB/(CO₂ + O₂ + Ar) flow through a room temperature ethylbenzene foaming device for 30 min. The flow of ethylbenzene vapor was then stopped, the gas flow was switched to Ar, and the in situ pool was heated from 100 °C to 500 °C at a rate of 10 °C/min, and the data was recorded every minute. When the isothermal in situ infrared test was carried out at 500 °C, the in situ cell was kept at 500 °C under Ar. Next, the ethylbenzene vapor stream flowed into the in situ cell through the ethylbenzene bubbler at room temperature under Ar, CO₂ + Ar, O₂ + Ar, and CO₂ + O₂ + Ar atmospheres, and the data was recorded every minute.

3.4. Catalytic Performance Testing under Oxygen-Free Conditions

This part of the experiment tested the performance of the catalyst under oxygen-free conditions. The purpose was to test the effect of lattice oxygen on the properties of alkali metal loaded cerium oxide samples. The n%X/CeO₂ was used as the catalyst to carry out the catalytic activity experiment. The experimental device mainly consists of an electric furnace, a multi-directional gas mass flowmeter, an ethylbenzene steam generator, a fixed bed reactor with a temperature controller, a waste gas processor, and an online gas analysis mass spectrometer for gas product analysis. A gas chromatography-mass spectrometry coupler is used to quantitatively study the gas composition of waste gas.

In the isothermal redox test, 500 mg of sample and 100 mg of quartz sand (60–80 mesh) were placed in the reactor and purged with Ar (≥99.99 vol%, 50 mL·min⁻¹) at 200 °C to remove all possible adsorbed impurities. At the beginning of the measurement, the temperature increased from 200 °C to 500 °C at a rate of 10 °C/min. When the temperature reached 500 °C, the gas was switched from Ar to EB/Ar (25 mL·min⁻¹), and the unconsumed EB, styrene (ST), toluene (TOL), benzene (B) and carbon oxide (COX) were recorded and analyzed in real time by MS, and quantified by GC-MS. At the end of the reaction, the reactor was cooled in the Ar environment. EB conversion and styrene selectivity were calculated by the following formula:

$$\mathrm{EB} \mathrm{Conversion} (\%)=\frac{N_{EB,in}-N_{EB,out}}{N_{EB,in}}\times 100\%$$

(4)

$$\mathrm{ST} \mathrm{Selectivity} (\%)= \frac{N_{ST,out}}{ N_{B,out}+N_{TOL,out}+N_{COX,out}+N_{ST,out}}\times 100\%$$

(5)

where:

N_EB,in: The amount of EB introduced, mol;

N_EB,out: The amount of EB emitted, mol;

N_ST,out: The amount of ST emitted, mol;

N_B,out: The amount of byproduct B produced by the reaction, mol;

N_TOL,out: The amount of byproduct TOL produced by the reaction, mol;

N_COX,out: The amount of byproduct COX produced by the reaction, mol.

3.5. Performance Testing of Oxidative Dehydrogenation of Ethylbenzene

A total of 1000 mg of the sample was placed in the reactor and purged with Ar (gas flow rate 50 mL/min) at 200 °C for 30 min to remove all possible adsorbed impurities. At the beginning of the measurement, the temperature was increased from 200 °C to 500 °C at a heating rate of 10 °C/min. When the temperature reached 500 °C, the gas was switched from Ar to EB/(Ar-CO₂-O₂) (gas flow rate 25 mL/min) for a reaction time of 3 h/5 h/50 h. The unconsumed EB and the ST, TOL, B and COX products were recorded and analyzed by MS every half hour and quantified by GC-MS. At the end of the reaction, the reactor was cooled in Ar. The EB conversion was calculated by Equation (1), and ST selectivity was calculated by the following equation.

$$\mathrm{ST} \mathrm{Selectivity} (\%)=\frac{N_{ST,out}}{ N_{B,out}+N_{TOL,out}+N_{COX}+N_{ST,out}}\times 100\%$$

(6)

N_COX is the instantaneous molar amount of carbon oxide. Since the CO₂-O₂ mixed atmosphere was passed into the reactor and the flow rate of CO₂ and O₂ was fixed, there was a base molar amount of CO₂, and when CO₂ was consumed below the base molar amount as the reaction proceeded, the conversion rate of CO₂ was calculated using the following equation.

$${\mathrm{CO}}_2 \mathrm{Conversion} (\%)=\frac{N_{CO2,in}-N_{CO2,out}}{N_{CO2,in}}\times 100\%$$

(7)

When N_CO2,in > N_CO2,out, N_COX = 0 (CO₂ was consumed and CO was the reaction product), the CO₂ conversion rate was calculated by Equation (4); When N_CO2,in < N_CO2,out, the reaction process produces CO₂, which was considered to be the by-product CO_X, N_COX = N_CO2,in − N_CO2,out (CO was the reaction product).

4. Conclusions

We proposed an efficient ODH of ethylbenzene into styrene using a K/CeO₂ catalyst under a CO₂-O₂ mixture atmosphere at 500 °C. The thermodynamic analysis illustrates that a thermoneutral ODH is possible by the rationalizing of CO₂/O₂ molar ratios from 0.65 to 0.66 in the temperature range of 300 to 650 °C. The modification of alkali metals (Li, Na and K) oxides could greatly enhance the oxygen storage capacity ceria, and the K modification is favorable for ethylbenzene dehydrogenation. 10%K/CeO₂ reveals the highest oxygen mobility and strong CO₂ adsorbability, achieving the highest ethylbenzene dehydrogenation activity. The catalyst achieved 90.8% ethylbenzene conversion and 97.5% styrene selectivity under the optimized conditions of the CO₂-4O₂ oxidation atmosphere, a temperature of 500 °C, and a space velocity of 5.0 h⁻¹. It maintained high catalytic and structural stability during the 50 h long-term test. The slight crystallization of ceria and some surface carbonates were observed for the spent catalyst. In situ DRIFTS experiments suggested that CO₂ induces oxygen vacancies in ceria and promotes oxygen exchange between gaseous oxygen and ceria. The ethylbenzene ODH with CO₂-O₂ follows a Mars-van Krevelen reaction mechanism by using Ce³⁺/Ce⁴⁺ redox pairs. The proposed ODH strategy of using oxygen vacancies enriched catalysts offers an important pathway for the efficient dehydrogenation of ethylbenzene.