CO₂ utilization as a soft oxidant for the synthesis of styrene from ethylbenzene over Co₃O₄ supported on magnesium aluminate spinel: role of spinel activation temperature

Abstract

Magnesium aluminate spinel (MgAl₂O₄) supported Co₃O₄ catalysts are synthesized and tested for the oxidative dehydrogenation (ODH) of ethylbenzene using CO₂ as a soft oxidant. The effect of spinel calcination temperature on the catalytic performance has been systematically investigated. With an increase in the activation temperature from 600 to 900 °C, the active presence of a single-phase MgAl₂O₄ spinel is observed. A catalyst series consisting of MgAl₂O₄ spinel with varying Co loadings (10–20 wt%) were prepared and systematically distinguished by ICP, XRD, BET, TPR, NH₃-TPD, UV–Vis DRS, FT-IR, XPS, SEM, and TEM. Among the tested cobalt catalysts, 15Co/800MA sample derived by calcination of MgAl₂O₄ support at 800 °C exhibits the most excellent catalytic performance with the maximum ethylbenzene conversion (≥ 82%). Also, high yields of styrene (≥ 81%) could be consistently achieved on the same active catalyst. Further, the catalyst exhibited almost stable activity during 20 h time-on-stream with a slow decrease in the ethylbenzene conversion from 82 to 59%. However, the selectivity of styrene (98%) stayed almost constant during the reaction. Activation of the MgAl₂O₄ spinel at 800 °C facilitates a dramatic chemical homogeneity for the alignment of Co₃O₄ nanoparticles on the surface of the active catalyst. Moreover, the isolated Co₃O₄ clusters have a strong chemical/electronic interaction with the Mg²⁺ and Al³⁺ ions on the support perform a crucial role to achieve the maximum catalytic activity.

Introduction

Styrene (ST) monomer is one of the most essential chemical compounds and huge industrial specifications on large scale for the synthesis of several useful commodities. Styrene has been largely synthesized by the oxidative dehydrogenation (ODH) of ethylbenzene (EB) using iron oxide (Fe₂O₃) with an excess of steam in the precise range of temperature from 600 to 650 °C¹. However, there are several problems associated with the ODH of EB like high energy utilization, low yield of ST together with rapid catalyst deactivation, and thermodynamic constraints owing to the endothermic nature. To overcome the above limitations, few efficient catalyst systems are developed by using different oxidizing agents such as O₂, N₂O, Ar, and CO_2, respectively². Among them, CO₂ has received more attention due to it stays in gas form during ODH and does not need latent heat for vaporization. In addition, carbon dioxide reduces the reactants partial pressure more effectively in comparison to conventional steam and other oxidants. In other words, CO₂ utilization helps in the coupling of reverse water–gas shift reaction (RWGSR) with normal dehydrogenation of EB using Fe₂O₃/Al₂O₃ catalyst³. With carbon dioxide as soft oxidant, ODH of EB was studied over the influence of cerium and its precursor on potassium promoted iron catalyst⁴, mesoporous silicalite-1 supported TiO₂-ZrO₂⁵, and V₂O₅-CeO₂/TiO₂-ZrO₂ catalyst⁶. Chang et al. claimed the CO₂ as active oxidant over supported vanadium-antimony oxide which afforded maximum and steady ST yields compared to steam-assisted process and N₂ flow conditions⁷. Meso structured nickel-based CeO₂⁸, and Al₂O₃ supported catalysts have been tested for the ODH of EB with good ST yields⁹. Significant EB conversion achieved in the presence of carbon dioxide over CeO₂–ZrO₂/SBA-15¹⁰, MnO₂–ZrO₂ and K₂O/MnO₂–TiO₂ catalysts¹¹. Significant yield of ST production has been accomplished in the ODH of EB over high-surface-area CeO₂ catalyst¹². Park et al. published a comprehensive review on the ODH activity of EB over different promoters incorporated on solid oxide catalysts at various reaction conditions¹³. Of late, cobalt nanoparticles receive potential applications in the extensive fields of biomedicine, biotechnology, material engineering, and environmental science¹⁴. In this scenario, several cobalt-based catalysts are also applied for ODH of EB such as bimetallic Co–Mo catalysts where cobalt-active species have a huge impact on getting high ST production in the presence of CO₂¹⁵. Likewise, Moronta et al. reported the use of Co–Mo mixed oxides for the dehydrogenation of EB with a good conversion¹⁶. Based on the above discussions, reduced catalysts are more efficient for the ODH of EB than supported metal oxides. Our group earlier reported the application of Co₃O₄ supported on mixed oxide materials for the oxidative-dehydrogenation of EB under mild reaction conditions¹⁷. Moreover, cobalt impregnated La/MgO¹⁸, Co, Ni/carbon nanotube¹⁹, and Co and Ni ferrites distributed on porous silica catalysts have been utilized in ST synthesis²⁰. CoFe₂O₄–MCM-41 catalyst is also active for the ODH of EB using CO₂ as a soft oxidizing agent²¹. In addition to the CO₂, N₂ also can play a significant role as co-oxidant when Co-Mo bimetallic nitrides are employed in the ODH of EB²².

During the decades, there is a growing concern in the application of magnesium aluminate (MA) spinel as active catalysts or catalyst support in the field of green catalysis, processing of petroleum, and fine-chemicals production. Moreover, the tunable acidic-basic sites, superior chemical strength, and thermal steadiness are crucial for numerous industrial applications²³. In addition, MgAl₂O₄ spinel is also used as an efficient support material for the catalytic steam reforming and dehydrogenation reactions^24,25. For industrial applications, particularly as catalyst support or catalyst, a high surface area, optimum crystallite size, and a greater number of active sites are significantly preferred²⁶. Ji et al. developed iron oxide incorporated MA spinel catalysts for ODH of EB, which is beneficial than bulk iron oxide/MgO and iron oxide/γ-Al₂O₃ catalysts²⁷. Pratap et al. reported MgAl₂O₄ supported CeO₂ catalysts with high specific surface area and improved redox properties by co-precipitation method²⁸. However, the previously mentioned catalytic systems are not attractive to some extent for desired catalytic activity. So far, ODH of EB over different calcination temperatures of MA spinel supported cobalt oxide catalysts has been unemployed.

In this perspective, for the first time, we are reporting the profound effect of calcination temperature of MgAl₂O₄ spinel supported on Co₃O₄ catalysts for the ODH of EB at atmospheric pressure using CO₂ as a soft oxidant. In notable addition, Co₃O₄ is used without any pre-reduction, as a result majorly avoids the typical usage of H₂ which makes it a cost-effective process for industrial application. The growing demand for hydrogen energy makes researchers develop an effective and advanced catalytic process to produce chemical feedstocks²⁹. Further, the synthesized samples represent an efficient alternative catalytic system for the ODH of EB than the reported metal oxides including cobalt-based catalysts^15,16. In the present study, the catalytic performance of Co₃O₄/MgAl₂O₄ catalysts has precisely correlated with the surface structural characteristics using various analytical and spectroscopic tools as follows.

Experimental

Catalyst preparation

Magnesium aluminate spinel (MgAl₂O₄) with Mg/Al molar ratio of 1:2 was prepared by the co-precipitation method at a p^H of 8.9–9.0, using 5% NH₄OH solution as a precipitating agent. The precipitate obtained was filtered, gently washed with high purity millipore water followed by oven drying at 100 °C for 12 h. The obtained solid was calcined at different temperatures like 600, 700, 800, and 900 °C for 5 h labeled as 600MA, 700MA, 800MA, and 900MA respectively. A series of cobalt catalysts are prepared by the impregnation method using MgAl₂O₄ as support and cobalt nitrate as a metal precursor. The catalysts are calcined at 600 °C for 5 h with a heat ramp of 10 °C/min, designated as x Co₃O₄/yMA where ‘x’ indicates weight percentage of cobalt content (x = 10, 15, and 20), whereas ‘y’ represents the calcination temperature of MgAl₂O₄ spinel. For the comparison studies, 15 wt% Co₃O₄/γ-Al₂O₃ (CA) and 15 wt% Co₃O₄/MgO (CM) catalysts are synthesized by the wet-impregnation method under similar conditions.

Characterization of catalyst samples

The XRD experiments for all catalyst samples are carried out in an ULTIMA-IV X-Ray Diffractometer (M/s. Rigaku Instruments, Japan) with Cu Kα monochromatic radiation operating at 40 kV voltage and a current of 30 mA. The crystalline phase of catalysts is determined by continuous scan mode with the 2θ range of 10°–80° at a sampling pitch of 0.07 and 4° min⁻¹ scan rate. Reduction patterns of calcined catalysts are investigated by TPR on a self-made reactor system¹⁹. Typically, 50 mg of the catalyst is loaded in a small quartz reactor and placed in the micro furnace. The reactor is heated linearly up to 800 °C at a ramping rate of 10 °C min⁻¹ while flowing 5%H₂/Ar gas (30 mL/min). After reaching the maximum temperature, the specimen is maintained under isothermal conditions for another 60 min. The amount of H₂ consumption is monitored by online TCD in-built GC (M/s. CIC Instruments, India)¹⁹. The amount of cobalt content present in the catalysts was accurately estimated using iCAP 6500 duo ICP-OES Analyzer (M/s. Thermo Fisher Scientific, USA). Before the ICP analysis, the specific cobalt catalyst was dissolved in a stock solution of aquaregia.

FTIR spectra of the samples were recorded on a spectrum-GX spectrometer in the scan range of 4000 to 400 cm⁻¹. UV–Vis diffused reflectance spectra (DRS) of samples are investigated on a UV Winlab spectrometer in the UV–Vis region of 200–800 nm. BET principle is employed to investigate the specific surface area of all the catalysts. Before the N₂ sorption analysis, all samples are degassed under vacuum at 350 °C for 4 h to eliminate physically absorbed gases and moisture. The surface morphology of samples is explored by high-resolution TEM (TECHNAI G2, Japan), operated at an acceleration voltage of 200 kV. A carbon-coated grid is used to disperse the powder which has been mixed with ethanol under sonication. The elementary composition is identified by EDX analysis from the emitted X-rays of the specimen during TEM analysis.

The coke content present in the used catalysts is estimated by TG analysis on TA instruments Q-500 using alumina sample holder heated from 30 to 800 °C at a ramping of 5 °C min⁻¹ under O₂ flow. The NH₃-TPD studies are performed using ASAP 2020 (M/s. Micrometrics, USA). Typically, 5%NH₃ balanced He gas is adsorbed at 100 °C for 30 min, then purged with ultra-pure He (20 mL/min) to eliminate the physisorbed water and other gases. The TPD patterns were recorded by desorption of NH₃ from 100 to 800 °C at a heat ramp of 5 °C min⁻¹³⁰. The XPS analysis is carried out on a Thermo Scientific K-Alpha Spectrometer operated with Al Kα as X-ray radiation (E_photon = 1486.7 eV) at room temperature. The B.E. values of all the elements are accurately calculated by adopting C 1s as a standard reference peak.

Results and discussion

Powder X-ray diffraction

The crystallographic structure and active phase of cobalt catalysts with different calcination temperatures of MgAl₂O₄ are typically displayed in Fig. 1. The direct XRD reflections corresponding to the lattice planes at (111), (220), (311), (222), (400), (422), (511), and (440) confirm the active presence of single-phase MgAl₂O₄ spinel (JCPDS No # 04-008-1061)³¹. As shown in Fig. 1a, the peak intensity of the MgAl₂O₄ spinel increases linearly with the gradual rise in calcination temperature, eventually, noticeable growth in crystallite size is typically observed. The 600MA sample exhibited very low-intensity broad peaks due to the partial formation of the MgAl₂O₄ spinel phase along with γ-Al₂O₃ active phase. The MgO formed by the decomposition of Mg(OH)₂, reacts instantly with Al(OH)₃ to form MgAl₂O₄ spinel since the Al(OH)₃ phase is stable up to 600 °C³². Conversely, the 700MA sample represents significant growth in the signal intensity, and the unique formation of the MgAl₂O₄ spinel phase is observed³³. It should be carefully noted that when the activation temperature is 800 °C (800MA), a highly pure MgAl₂O₄ spinel phase was typically formed. Consequently, a high calcination temperature of 900 °C promoted drastic acceleration in the crystallite size due to severe sintering of MgAl₂O₄ spinel particles and typically exhibited an extremely sharp reflection with strong intensity. In summary, calcination in the temperature between 700 to 900 °C is typically preferred for the proper formation of single-phase MgAl₂O₄ spinels.

Figure 1

XRD patterns of (a) MgAl₂O₄ spinel at various activation temperatures and (b) cobalt supported catalysts.

The XRD patterns of MgAl₂O₄ supported Co₃O₄ catalysts are as shown in Fig. 1b. All the diffractions can be assigned to Co₃O₄ (JCPDS No # 00-042-1467) with lattice planes of (111), (220), (311), (222), (400), (422), (511), and (440) respectively³⁴. No other reflections are observed, indicating the presence of an active phase with high purity. However, the diffraction lines of Co₃O₄ and MgAl₂O₄ are not well-resolved which can be attributed to the similar 2θ values of the corresponding active phases³³. Besides, 15Co/600MA, 15Co/700MA, and 15Co/800MA catalysts display XRD lines with similar intensity while the 15Co/900MA catalyst maintains extraordinary intensity. The spent 15Co/800MA catalyst displayed intact XRD peaks as observed for fresh catalysts even after several hours of activity studies (Fig. 1b). Also, the intensity of direct XRD reflections of 15Co/600MA is higher than the bare 600MA spinel due to the formation of bigger crystallites of Co₃O₄ after doping. Similarly, the micropores present in the 900MA spinel are covered by the large Co₃O₄ crystals in the 15Co/900MA catalyst, consequently, a drastic change in the crystallite size observed as illustrated in Table 1. However, the crystallite size of the 15Co/700MA catalyst stays unchanged after the doping with Co₃O₄. Remarkably, a huge contrast typically emerged between the crystallite size of 800MA spinel and 15Co/800MA catalyst. This can be typically attributed to the uniformly distributed Co₃O₄ rather than occupied micropores of the spinel support. A similar trend was also observed in 10Co/800MA and 20Co/800MA catalysts as shown in Fig. S1 (supplementary information).

Table 1 Physicochemical characteristics of MgAl₂O₄ spinel synthesized at different activation temperatures and its corresponding cobalt supported catalysts.

BET surface area and pore-size analysis

The specific BET surface area and pore-size values of as-synthesized MgAl₂O₄ spinels at variable calcination temperatures and corresponding Co₃O₄ supported catalysts are typically presented in Table 1. It is revealed that with an increase in the calcination temperature of the MA spinel from 600 to 900 °C, there is a noticeable decrease in the surface area from 152 to 52 m²/g. This can be ascribed to the generation of bigger crystallites of MgAl₂O₄ spinel at more elevated temperatures follwing the XRD results. The high surface area of the 600MA sample is majorly due to the possible formation of γ-Al₂O₃ phase along with MgAl₂O₄ spinel below the activation temperature of 700 °C³³. In addition, the pore volume and pore diameter of the samples also decreased with rising the calcination temperature owing to the gradual increase in the crystallite size of MA spinel.

However, the Co₃O₄ doped MgAl₂O₄ spinel catalysts have typically exhibited fascinating results as mentioned in Table 1. The addition of nominal cobalt content to the MA support forms a uniform layer of Co₃O₄ during the thermal treatment. This phenomenon results increase in the specific surface area and pore-size of the Co₃O₄/MgAl₂O₄ samples when compared with the bare support. From Table 1, the measured surface area of 15Co/600MA and 15Co/700MA catalysts are 146 and 121 m²/g respectively. The smooth decrease in the surface area probably due to the partial filling of Co₃O₄ species in the micropores of the MgAl₂O₄ spinel. Remarkably, the high surface area and pore-size of 15Co/800MA catalyst can be explained based on the fine dispersion of Co₃O₄ particles on the 800MA support. Likewise, 10Co/800MA and 20Co/800MA catalysts promptly follow a similar trend and typically exhibited a more specific surface area as stated in Table 1.

In contrast, the low surface area of the 900MA spinel could not provide enough external surface space for the sheer distribution of Co₃O₄ clusters. Therefore, the 15Co/900MA catalyst has typically shown an extremely low specific surface area in comparison with the other samples. As per the above discussion, 15Co/800MA represents the best catalyst for the titled reaction with a high specific surface area and large pore size.

H₂-TPR

The reduction profiles of Co/MA spinel catalysts including 15Co/γ-Al₂O₃ (CA), and 15Co/MgO (CM) are typically shown in Fig. 2a. It has been extensively investigated that the reduction pattern of Co₃O₄ could be influenced by several factors like preparation method, metal precursor, nature of the support, and activation treatment. The TPR profiles of CA, CM, Co/MA (i.e., 15Co/700MA, and 15Co/900MA) catalysts exhibit the reduction peak with high intensity at elevated temperatures compared to those present in low-temperature reduction signals. This allows a scope to think that the reduction process of Co₃O₄ phase in 15Co/γ-Al₂O₃, 15Co/MgO, 15Co/700MA, and 15Co/900MA catalysts proceed in two steps as follows:

$$ {\text{Co}}_{{3}} {\text{O}}_{{4}} + {\text{ H}}_{{2}} \to {\text{3CoO }} + {\text{ H}}_{{2}} {\text{O}} $$

(1)

$$ {\text{3CoO }} + {\text{ 3H}}_{{2}} \to {\text{3Co }} + {\text{ 3H}}_{{2}} {\text{O}} $$

(2)

Figure 2

(a) H₂-TPR patterns of (i) CA, (ii) CM, (iii) 15Co/600MA, (iv) 15Co/700MA, (v) 10Co/800MA, (vi) 15Co/800MA, (vii) 20Co/800MA, and (viii) 15Co/900MA catalysts. (b) NH₃-TPD profiles of (i) 800MA, (ii) CM, (iii) CA, (iv) 15Co/700MA, (v) 10Co/800MA, (vi) 15Co/800MA, (vii) 20Co/800MA, and (viii) 15Co/900MA catalysts.

Thereby, H₂ consumption could be higher for 15Co/900MA catalyst because of larger crystallite size than 10Co/800MA, 15Co/800MA and 20Co/800MA catalysts. The H₂-TPR of 15Co/700MA catalyst displayed two reduction signals one at a lower temperature (500–600 °C) with less intensity and another peak at a high temperature (600–900 °C) with more intensity. The low-temperature signal can be ascribed to the reduction of uniformly distributed Co₃O₄ particles. Whereas, the high-temperature peak is attributed to the complex reduction pattern of strongly interacted Co₃O₄ species. Besides, the 15Co/γ-Al₂O₃ catalyst exhibited two major reduction peaks at high temperature due to the strong interaction of Co₃O₄ species with γ-Al₂O₃ support³⁵. It is reported that transition metal (M²⁺) ions preferentially occupies tetrahedral vacancies of γ-Al₂O₃, so it is difficult to get metallic Co, compared to those occupied in octahedral vacancies of γ-Al₂O₃³⁶. In the case of Co/MgO catalyst three different reduction signals are observed with different peak intensities as described in Fig. 2a. Amongst, the first reduced signal centered at 300 °C is ascribed to the direct transition of Co₃O₄ into metallic Co species, while second reduced signal about T_max 400 °C is illustrated the transformation of strongly support interacted Co₃O₄ species into CoO and metallic Co. Then, reduction pattern at high temperature (600 °C) may probably represent the formation of some solid solution species such as MgCo₂O₄ and CoO-MgO species in Co₃O₄/MgO catalyst³⁷.

However, reduction patterns of 10Co/800MA, 15Co/800MA, and 20Co/800MA catalyst are quite different from Co/MgO, Co/γ-Al₂O₃, 15Co/700MA, and 15Co/900MA catalyst. As a result, 10Co/800MA, 15Co/800MA, and 20Co/800MA catalysts displayed facile reduction patterns owing to synergistic interface between Co₃O₄ and MgAl₂O₄ support. Therefore, 15Co/800MA catalyst is exhibited low-intense reduced signals compared to those present in Co/MgO, Co/γ-Al₂O₃, 15Co/700MA, 20Co/800MA, and 15Co/900MA catalysts. Furthermore, the surface textural properties of 15Co/800MA catalyst is better in line with smaller crystallite size, and high surface area facilitates the easy reduction of Co₃O₄ species as evidenced from H₂-TPR patterns (Fig. 2a). In the present investigation, reduction patterns of 10Co/800MA and 20Co/800MA catalysts are almost identical with the 15Co/800MA catalyst. In notable addition, highly intense TPR signals are observed for 20Co/800MA catalyst due to the reduction of bulk Co₃O₄ species. While 10Co/800MA catalyst displayed very low intensity TPR signals owing to the low cobalt oxide content. Similarly, the 15Co/600MA catalyst contain two reduction peaks as observed in 10Co/800MA catalyst, which is in agreement with the reported literature^35,38.

UV–Vis DRS

The UV–Vis spectra of MgAl₂O₄ spinel activated at various temperatures are presented in Fig. S2 (supplementary information), where the K–M functions of the bands are plotted as a function of wavelength. It is evident from the figure that absorption bands of all samples are following a similar trend with small variation in the peak intensity. The absorption maxima below 300 nm can be attributed to the charge transfer (CT) from O⁻² to Al³⁺ ions due to excitation of electrons from the valence band of O(2p) to the conduction band of Al(3d)³¹.

As shown in Fig. 3a, a split in the absorption bands below 300 nm observed for Co₃O₄ supported catalysts placed at 214 nm (λ₁) and 240 nm (λ₂) corresponding to the CT transition of O⁻² → Al³⁺ and O⁻² → Co³⁺ respectively^39,40. The broad absorption signals centered at 430 (λ₃) nm and 720 nm (λ₄) among all the cobalt catalysts indicate the presence of a large amount of Co atoms in the octahedral and tetrahedral symmetry respectively⁴¹. In general, CoAl₂O₄ spinel might be formed during high-temperature due to diffusion of Co²⁺ ions into the alumina, and their distribution possibly varies with the amount of Co loading in the lattice. However, no peaks corresponding to tetrahedral Co²⁺ symmetry (~ 620 nm) confirms the absence of the CoAl₂O₄ spinel phase among all the MgAl₂O₄ supported cobalt catalysts.

Figure 3

(a) UV–Vis spectral analysis of (i) 15Co/600MA (ii) 15Co/700MA (iii) 10Co/800MA (iv) 15Co/800MA (v) 20Co/800MA, and (vi) 15Co/900MA catalysts. (b) FTIR spectral analysis of (i) 15Co/600MA (ii) 15Co/700MA (iii) 10Co/800MA (iv) 15Co/800MA (v) 20Co/800MA, and (vi) 15Co/900MA catalysts.

FT-IR

The FT-IR spectra of MgAl₂O₄ spinel synthesized at various calcination temperatures were recorded in the frequency range from 4000 to 400 cm⁻¹ and illustrated in Fig. S2 (supplementary information). The visible FTIR bands in a range of 900 to 500 cm⁻¹ can be attributed to the metal–oxygen (M–O) stretching vibrations, where “M” denotes either Al or Mg. The signal at 700 cm⁻¹ indicates the presence of Al³⁺ ions in the octahedral sites, and another vibration band at 750 cm⁻¹ assigned to the occupation of Mg²⁺ ions in the tetrahedral sites²⁸. Therefore, can be confirmed the formation of a single-phase MgAl₂O₄ spinel with rising the activation temperature from 600 to 900 °C. Further, the signal located at 1630 cm⁻¹ is assigned to the deformation band of interlayer water molecules (H–O–H) corresponding to the MgAl₂O₄ spinel surface²⁸. The broad FTIR band at 3430 cm⁻¹ could be plausibly attributed to the O–H stretching vibration of the surface adsorbed water molecules.

The FTIR spectral analysis of Co₃O₄ incorporated on MgAl₂O₄ spinel catalysts prepared by the impregnation method is typically displayed in Fig. 3b. The Co–O stretching vibration bands typically displayed at an optimal frequency of 570 and 675 cm⁻¹ which are ascribed to the octahedral and tetrahedral symmetry of Co³⁺ ions⁴². However, no apparent difference is found between the FTIR spectra of Co/MA catalysts and MA support, because of the overlapping of Co₃O₄ signals with MgAl₂O₄ spinel.

NH₃-TPD

To explore the distribution of the surface acidic sites on MgAl₂O₄ spinel (MA) including MA supported cobalt catalysts, NH₃-TPD was performed as illustrated in Fig. 2b. The resulting basic sites analyzed from TPD patterns of samples are further classified into three categories (i.e., weak, moderate, and strong) assigning to their desorption strength. However, the bare 800MA support contains a lesser number of strong acidic sites, although the uniform distribution of weak and moderate acidic sites is investigated because of balanced Al³⁺ and Mg²⁺ ions. In the CM catalyst, weak and moderate acidic sites are massive in number, whereas strong acidic sites are relatively lesser compared to that of CA catalysts. It is majorly owing to the more significant number of Mg²⁺ species present in CM catalyst; consequently, diminish the surface acidic nature of the catalyst. Moreover, the formation of some extent solid solution species like MgCo₂O₄ also decreases the surface acidic nature of CM catalyst. Therefore, CM catalyst afforded a low conversion of EB but promoted high styrene selectivity as illustrated in Table 3.

In the case of CA, 10Co/800MA, 15Co/800MA, and 20Co/800MA catalysts strong acidic site distribution and the total acidity are relatively higher than CM, 15Co/900MA, and 800MA samples as mentioned in Table 2. The 10Co/800MA catalyst displayed a lower fraction of moderate and strong acidic sites, because of less cobalt oxide content. In contrast, 20Co/800MA catalyst with more cobalt oxide containing a more considerable fraction of strong acidic sites accordingly severe decrease of styrene monomer selectivity (Table 3). Whereas, a more significant quantity of weak and moderate acidic sites along with strong acidic sites exist in the CA catalyst is responsible for poor catalytic activity. In a similar approach, less acidity present in the 800MA and CM samples is unfavorable for high EB conversion and ST selectivity (Table 3). The 15Co/900MA catalyst despite containing less acidic sites did not yield more ST owing to the sintering of active cobalt oxide particles at high calcination temperatures. The number of weak and moderate acidic sites generated in 15Co/800MA catalyst is considerably higher in number due to the uniform alignment of Co₃O₄ species, but strong acidic sites are noticeably low compared to CA and 20Co/800MA catalyst.

Table 2 NH₃-TPD analysis of MgAl₂O₄ (MA), 15Co₃O₄/γ-Al₂O₃ (CA), 15Co₃O₄/MgO (CM), and Co₃O₄/MgAl₂O₄ (Co/MA) samples.

Table 3 ODH of EB over MgAl₂O₄ (MA), Co₃O₄/MgAl₂O₄ (Co/MA), 15Co₃O₄/γ-Al₂O₃ (CA), and 15Co₃O₄/MgO (CM) samples in CO₂ atmosphere.

Typically, based on NH₃-TPD patterns neither a low fraction of acidic sites nor strong acidic sites generated are suitable for potential catalytic activity. Therefore, balanced weak and moderate acidic nature along with adequate strong acidic sites possessed 15Co/800MA catalyst represented a significant role in getting maximum EB conversion and styrene selectivity.

XPS

The oxidation state of active cobalt species present in the calcined Co/MA catalysts is detected by the XPS technique and depicted in Fig. 4. The binding energy (B.E.) value of Co2p_3/2 at ~ 780 eV and Co2p_1/2 at ~ 795 eV with a spin-orbital splitting of 15.2 eV and the absence of intense satellite peaks demonstrate the presence of Co₃O₄ species among all the cobalt catalysts^43,44,45. The B.E. value of Al2p at 74.1 eV and Mg1s at 1303.4 eV indicates the presence of Al⁺³ and Mg⁺² ions in the form of MgAl₂O₄ spinel (Fig. S3, supplementary information)⁴⁰. Nevertheless, the 15Co/800MA catalyst exhibited a lower B.E. value than 15Co/700MA, 15Co/900MA, and 20Co/800MA catalysts owing to the synergetic interaction between Co₃O₄ and MgAl₂O₄ support⁴¹.

Figure 4

Co2p XPS spectra of (a) 15Co/600MA, (b) 15Co/700MA, (c) 15Co/800MA and (d) 15Co/900MA catalysts.

TEM

HRTEM analysis of MgAl₂O₄ spinel (MA) and 15Co/800MA catalyst are illustrated in Fig. 5. As shown in Fig. 5b, 15Co/800MA catalyst displayed homogeneously distributed nano-sized cobalt oxide particles environment on the high surface area of MA spinel. Similarly, the bare 800MA spinel (Fig. 5c) is prominently displayed a lesser number of aggregated species with related lattice fringes substantially similar to the 15Co/800MA catalyst. As shown in Fig. 5d, the lattice resolved HRTEM image of 15Co/800MA catalyst confirm the uniformly distributed nano-sized Co₃O₄ particles through crystal planes of (111), (311) and (220) with d-spacing of 0.466, 0.243 and, 0.286 nm respectively⁴⁶. However, the majority of the cobalt oxide fringes are typically covered over the surface of MA spinel. Further, SAED patterns authenticate the amorphous or poor crystalline nature of 800MA spinel and 15Co/800MA catalyst as outlined in Fig. S4 (supplementary information). Indeed, the SAED pattern of 800MA spinel showed bright circular planes compared to low-intensity circular planes in the SAED image of 15Co/800MA catalysts.

Figure 5

HRTEM images of 800MA spinel (MA) and 15Co/800MA catalysts.

The elementary composition of Co, Mg, and Al present in the fresh and spent 15Co/800MA catalysts are illustrated in the supplementary information (Figs. S5 and S6). It can be observed that no appreciable leaching of active cobalt oxide present in the used catalyst even after several hours of catalytic activity. The EDX elementary mapping of HAADF-STEM images of 15Co/800MA catalyst is as shown in Fig. S7 (supplementary information). It can be distinguished that a more considerable number of isolated and uniformly distributed Co₃O₄ species on the surface of the MgAl₂O₄ support. All these consistent findings are precisely following the HRTEM morphology results procured as illustrated in Fig. 5.

Thermogravimetric analysis

TG analysis of spent cobalt catalysts collected after 20 h of EB dehydrogenation activity is graphically displayed in the supplementary information (Fig. S8). As shown in the figure, TGA patterns of 15Co/700MA, 15Co/800MA, and 15Co/900MA catalysts exhibit a characteristic signal in the temperature range of 400–500 °C in all the catalysts. This can be attributed to the decomposition of deposited hydrocarbon moiety²⁸. The 15Co/800MA and 15Co/700MA catalysts have a low intense endothermic signal beyond 450 °C with a weight loss of 10% and 11.5% results from the catalytic oxidation of deposited carbon (coke). While 15Co/900MA catalyst has a steep decomposition peak at 470 °C with a significant weight loss of 14.5% owing to excessive decomposition of EB molecules on the cobalt oxide particles with large crystallize size. Also, elemental analysis results are in line with the TGA analysis concerning coke deposition on spent catalysts. As described in TGA patterns, less coke deposition is investigated on the surface of 15Co/800MA spent catalysts because of efficient oxidation of deposited carbon into CO₂ molecules in the course of ODH of EB. Therefore, one can expect that the 15Co/800MA catalyst would typically exhibit a steady catalytic activity compared to 15Co/700MA and 15Co/900MA catalyst.

Catalytic activity studies

The CO₂ assisted ODH of EB over cobalt catalysts takes place in two steps as shown in Eq. (3) and (4). In the first step, EB dehydrogenated into ST with the liberation of H₂. Afterward, the H₂ formed will typically react with CO₂ via reverse water–gas shift (RWGS) reaction. Therefore, the general reaction comprises a combination of EB dehydrogenation and RWGS reaction as displayed in Eq. (5).

$$ {\text{EB}} \to {\text{ST }} + {\text{ H}}_{{2}} $$

(3)

$$ {\text{CO}}_{{2}} + {\text{ H}}_{{2}} \to {\text{CO}} + {\text{H}}_{{2}} {\text{O}} $$

(4)

$$ {\text{EB }} + {\text{ CO}}_{{2}} \to {\text{ST}} + {\text{CO}} + {\text{H}}_{{2}} {\text{O}} $$

(5)

Influence of temperature on the ODH of EB activity

The influence of reaction temperature on the EB conversion (X_EB) was investigated in the presence of N₂ and CO₂ as oxidants over Co/MA catalysts as depicted in Fig. 6a,b respectively. Typically, ODH of EB is endothermic thereby mostly depends on the reaction temperature⁴⁷. Therefore, with a rise in the temperature from 450 to 650 °C, X_EB increased gradually to a maximum extent. As evident from Fig. 6, it can be observed that the remarkable activity of CO₂ assisted ODH of EB in comparison with the EB dehydrogenation under N₂ flow.

Figure 6

Influence of temperature on EB conversion over Co/MA catalysts under (a) N₂, and (b) CO₂ (conditions: T = 450–600 °C, catalyst = 0.7 g, N₂ (or) CO₂ = 30 mL/min, EB = 1.5 mL/h).

Among all the catalysts, 15Co/800MA catalyst performed outstanding activity than other samples (Fig. 6b), where the X_EB increased progressively from 8 to 82% with a change in the temperature from 450 to 600 C. Further increase in the reaction temperature to 650 °C, which showed a maximum X_EB (93%) but the selectivity of ST decreased to significant levels (not shown in Fig. 6). It can be observed that all the samples afforded very mild activity of EB at low reaction temperatures (450–550 °C), and then considerable improvement in X_EB is observed at 600 °C. After recognizing the optimum temperature, further CO₂ assisted ODH of EB reactions were carried out over all the cobalt catalysts including the MA support and the results are illustrated in Table 3.

MA spinel activated at low calcination temperature (600MA) exhibits a very low X_EB (29%) and S_ST (92%), which can be ascribed to the partial formation of single-phase MgAl₂O₄ spinel. In contrast, improved X_EB of 700MA (35%), and 800MA (42%) owing to the complete formation of single-phase MA spinel. However, the X_EB of 900MA spinel remained at 23% owing to the sintering of Mg²⁺ and Al³⁺ ions result in the formation of bigger crystallites with low specific surface area as depicted in Table 1.

After the incorporation of the optimal amount Co₃O₄ onto the 800MA spinel support, the catalytic activity increased two folds from 42 to 82%. Thus, cobalt oxide nanoparticles played an active role via synergistic interaction with the MgAl₂O₄ spinel support. Moreover, homogeneously distributed Co₃O₄ species on 800MA spinel with adequate surface acidic-basic sites has shown high EB conversion and ST selectivity. Among all, 15Co/600MA catalyst displayed the lowest X_EB and S_ST because of the lesser number of Co₃O₄ clusters on the MgAl₂O₄ spinel support and the presence of γ-Al₂O₃ phase as described in XRD (Fig. 1a). Besides, 15Co/700MA and 20Co/800MA catalysts with almost similar physicochemical characteristics (Table 1) afforded some extent equal X_EB at all reaction temperatures. The high surface area, thermal stability, and optimum surface acidic nature of the 15Co/800MA sample provide an enhanced chemical homogeneity towards the uniform distribution of Co₃O₄ nanoparticles. As a result, less coke deposition is observed on the external surface of the 15Co/800MA spinel catalyst even after several hours of catalytic activity study when compared to 15Co/700MA and 15Co/800MA catalysts, further evidenced from TGA analysis as displayed in Fig. S8 (supplementary information). It can be observed that neither low acidic strength of 10Co/800MA catalyst nor strong acidic sites of 20Co/800MA catalyst are suitable for achieving optimum catalytic activity (Table 3). Similarly, a very low specific surface area, and large crystallite size 15Co/900MA catalyst is also undesirable for the higher activity of CO₂ assisted ODH of EB.

Time-on-stream (TOS) study

To distinguish the long-term stability for the ODH of EB, TOS studies are performed over 15Co/800MA and 15Co/900MA catalysts in the presence of N₂ and CO₂ atmosphere as drawn in Fig. 7. It is evident from Fig. 7a, with an increase in the reaction time from 1 to 20 h, the X_EB decreases from 60 to 37% over the 15Co/800MA catalyst. Similarly, X_EB decreases from 43 to 23% over the 15Co/900MA catalyst with passing time. However, the S_ST stayed constant at 90% in both the samples during the reaction. Typically, the generated ST undergoes self-polymerization to yield polystyrene, which is crucial for coke deposition on the active surface of the catalyst¹². In the presence of N₂, the polystyrene will transform into graphitic carbon at elevated temperature (> 500 °C), consequently, rapid catalyst deactivation takes place (Fig. 7a).

Figure 7

TOS on 15Co/800MA and 15Co/900MA catalysts under (a) N₂ flow, (b) CO₂ flow (a) N₂ flow, (b) CO₂ flow (conditions: T = 600 °C, catalyst = 0.7 g, N₂ = 30 mL/min, EB = 1.5 mL/h).

It can be observed the X_EB decreases from 82 to 59% over the 15Co/800MA catalyst the 15Co/900MA catalyst follows a similar trend where the X_EB declined from 58 to 37%. However, the S_ST of 15Co/800MA (98%) and 15Co/900MA (95%) remains constant throughout the catalytic run. It is supposed to be due to the promotion of the RWGSR phenomenon in the presence of the CO₂ environment, which played a key role in superior catalytic activity. Besides, a greater number of uniform Co₃O₄ oxide particles on the 800MA spinel results in the high specific surface area and optimum crystallite size responsible for the enhancement in the catalytic activity. Nevertheless, the coke formation can be removed effectively by the oxidation of CO₂ as a soft oxidant. Hence, CO₂ assisted ODH of EB has a potential role to improve the catalytic activity as evident from Fig. 7b.

In the case of 10Co/800MA catalyst, a mild fraction of Co₃O₄ oxide particles leads to the marginal conversion of EB but no major changes are observed in the ST selectivity. In contrast, high Co oxide content possessed by 20Co/800MA catalyst influence the ST selectivity due to the high density of surface acidic sites as evidenced from NH₃-TPD analysis (Table 2). For convenience, we have not included the TOS study of 10Co/800MA and 20Co/900MA catalysts in the manuscript, however, corresponding X_EB and S_ST are listed in Table 3. In contrast, 15Co/900MA catalyst accomplished insignificant X_EB during the time-on-stream owing to the bigger crystallite size and subsequently displayed low surface area as listed in Table 1.

The regeneration of the used catalyst was performed on 15Co/800MA and 15Co/900MA samples, the corresponding results are as shown in. Typically, before the reusability studies, the specimens were activated in the airflow (30 mL/min) for 1 h. There is no observable change in the activity is observed and so we can assume that the 15Co₃O₄/MgAl₂O₄ catalyst possesses excellent thermal stability. The enhanced catalytic activity in the active presence of CO₂ is due to the potential elimination of H₂ formed in the dehydrogenation via reverse water–gas shift (RWGS) reaction. The amount of CO liberated during ODH of EB was accurately measured and graphically represented in Fig. S9 (supplementary information). High CO yield (≥ 9%) is obtained on 15Co/800MA catalyst compared to H₂ evolution, which specifies the great participation of H₂ in the RWGS reaction. While low yield of CO (6.4%) on 15Co/900MA catalyst indicates the low RWGS activity which is responsible for the mild X_EB (54%) as illustrated in Fig. 8. Accordingly, more fractions of carbonaceous material deposition tend to oxidize for the generation of excess CO₂ molecules on the surface of 15Co/800MA catalyst compared to that of 15Co/900MA catalyst.

Figure 8

TOS on regenerated cobalt catalysts under CO₂ flow (reaction conditions: T = 600 °C, catalyst = 0.7 g, CO₂ flow rate = 30 mL/min, EB flow rate = 1.5 mL/h).

Assessment of CO₂ assisted ODH of EB activity of 15Co/800MA with the reported catalysts

A comparison of CO₂ assisted oxidative dehydrogenation of EB activity over Co₃O₄ supported on the MgAl₂O₄ spinel system with the previously reported catalysts is summarized in Table 4. Guo et al. reported the use of a multi-walled CNT supported Co and Ni catalyst (CNT-Co-10) for the ODH of EB using CO₂ as oxidant with a maximum X_EB of 83% and S_ST of 88% at 550 °C (Table 4, entry 1). Similarly, Braga et al. synthesized a CoFeSi sample by polymeric precursor method which represents a X_EB value of 45.5% with an affordable S_ST of 98% at 550 °C (Table 4, entry 2). Moronta et al. prepared the CoMo bimetallic catalysts supported on montmorillonite Al-pillared clay and tested for the EB dehydrogenation activity. The reduced catalysts have shown more activity than unreduced samples with a X_EB of 20% and S_ST of 91% at 400 °C (Table 4, entry 3). Burri et al. explored the potential application of TiO₂-ZrO₂ binary oxide for efficient usage of CO₂ as a mild oxidant in the ODH of EB. The alkali promoted mixed oxide (K₂O/ TiO₂–ZrO₂) displayed a high selectivity of ST (99%) with a X_EB of 65.5% at 650 °C (Table 4, entry 4). Correspondingly, Burri et al. have also investigated the application of MnO₂-ZrO₂ mixed metal oxide for the ODH of EB in the active presence of CO₂. The binary metal oxide catalytic system providing a more specific surface area typically shown 73% X_EB and 98% S_ST at 650 °C (Table 4, entry 5). Madhavi et al. have examined the Co-Mo nitride catalysts (Co₃Mo₃N) for the ODH of EB using CO₂ as mild oxidant which exhibits X_EB of 62.5% and S_ST of 94.3% (Table 4, entry 6). Pratap et al. reported MgAl₂O₄ supported CeO₂ catalysts with high specific surface area and improved redox properties by co-precipitation method. The synthesized ceria catalysts exhibited a X_EB of 82% and S_ST of 98% at 600 °C (Table 4, entry 7). Likewise, Zhange et al. also investigated the high-surface-area CeO₂ for the CO₂ aided ODH of EB and achieved a noticeable improvement in the catalytic activity (Table 4, entry 8). Madduluri et al. reported the application of La₂O₃ incorporated Co₃O₄/MgO catalysts for the highly selective ODH of EB in CO₂ with a X_EB of 62% (Table 4, entry 9). It is clearly evident the 15Co/800MA achieved a remarkable enhancement in the EB dehydrogenation activity with a maximum X_EB of 82% and S_ST of 98% (Table 4, entry 10), which is much superior and better than the other catalysts stated.

Table 4 Comparison of CO₂ assisted EB dehydrogenation activity of 15Co/800MA with the reported catalysts.

Conclusions

In summary, 15Co₃O₄/MgAl₂O₄ spinel is found to be an efficient catalyst system for the ODH of EB in the presence of CO₂ as a soft oxidant. Besides, the catalyst possesses remarkable stability with a prolonged activity during 20 h TOS study. A gradual decrease in the conversion from 82 to 59% is anticipated for the mild coke formation. However, the selectivity of the styrene monomer (98%) has stayed almost the same throughout the reaction. The significant improvement can be ascribed to the calcination of MgAl₂O₄ spinel at 800 °C, which facilitates a chemical uniformity for the distribution of Co₃O₄ nanoparticles on the active surface of the catalyst. To be specific, a greater number of uniform Co₃O₄ oxide particles on the 800MA spinel results in the high specific surface area and optimum crystallite size responsible for the enhancement in the catalytic activity. Furthermore, the coke formation was suppressed effectively by the oxidation of CO₂ as a soft oxidant. It is supposed to be due to the promotion of RWGSR in the CO₂ environment, which performed a crucial role to achieve the maximum catalytic activity.