Synergistic interface between Co₃O₄ and MgAl₂O₄ in CO₂ assisted continuous vapour phase oxidative dehydrogenation of ethylbenzene to styrene monomer

2.1 Catalyst preparation

Co₃O₄/MgAl₂O₄ catalysts with different Co atomic concentrations were shown in Table1, and prepared by co-precipitation method at a pH in a range of 8.9–9.0, using 5% NH₄OH solution as a precipitating agent. The precipitated mass was filtered with several washings with deionized water followed by oven drying at 100 °C for 12 h and calcination at 650 °C for 5 h. These catalysts were designated as ‘x Co/MgAl₂O₄’ where ‘x’ represents the Co atomic concentration (x = 0.5, 0.75, 1.0 and 1.25) with a total atomic concentration of Co + Mg + Al fixed at 3.0. For the sake of comparison, Co₃O₄/γ-Al₂O₃ and Co₃O₄/MgO catalysts with Co concentration of 1 with a total atomic concentration of 3 were also prepared by co-precipitation method. A bare support, MgAl₂O₄ without Co content was also prepared under similar experimental conditions.

2.2 Catalyst characterization

XRD patterns of Co₃O₄/MgAl₂O₄ catalysts were recorded on a Ultima IV, X-ray powder diffractometer, (M/s. Rigaku Corporation, Japan) with a scanning step of 0.02° using Ni filtered Cu Kα radiation (λ = 1.5406 Å) at a scan speed of 4° min⁻¹ and a scan range of 10–80° at 40 kV and 20 mA. Surface area of catalysts was carried on an Autosorb-SI Instrument (M/s. Quantachrome, USA). Prior to N₂ adsorption-desorption experiment, the catalyst was degassed under vacuum at 300 °C for 3 h to remove the physisorbed moisture. Morphological structure of Co₃O₄ particles were systematically investigated by transmission electron microscopy (TEM) (Model TECHNAI G2 USA) operated at 200 kV. A copper grid coated with carbon and formava film was used to disperse the powder that has been mixed with ethanol and irradiated in an ultrasonic bath. FT-IR spectra of the catalysts were recorded on spectrum GX spectrometer (M/s. PerkinElmer, USA) in a scan range of 4000–400 cm⁻¹. The UV–Vis diffused reflectance spectra was recorded on Perkin Elmer UV Win Lab spectrometer with an integrating sphere reflectance accessory in a UV–Vis region of 200–800 nm. Temperature programmed reduction (H₂-TPR) profiles of cobalt catalysts were recorded on a homemade reactor interfaced with a TCD equipped gas chromatograph. Catalyst (∼30 mg) particles were placed between two quartz wool plugs at the centre of a quartz reactor and 5% H₂/Ar flow (rate of 30 cm³ min⁻¹) was maintained while increasing the temperature up to 800 °C at a ramp of 10 °C min⁻¹. The H₂ consumption was monitored with a standard GC software. TGA investigation of the spent catalysts were performed in air on a M/s. Q500 thermo gravimetric analyzer (M/s. Q500) using an alumina pan from RT to 800 °C at a heating rate of 10 °C/min. TGA patterns were obtained for 1.0CMA and 1.25CMA spent catalysts under CO₂ and N₂ atmosphere to estimate the carbonaceous deposite. CO₂ pulse chemisorption carried on an Autosorb iQ (Quantachrome USA) unit. Catalyst (100 mg), taken in a quartz reactor, was first reduced in H₂ gas at a flow rate of 60 cm³ min⁻¹ with heating rate of 10 °C/min up to 500 °C for 90 min. Then catalyst was flushed with He for 1 h followed by holding at 600 °C. Pure CO₂ (99%) gas was then introduced in pulses and the adsorption uptake analyzed using an inbuilt thermal conductivity detector. Temperature programmed desorption (TPD) of NH₃ (5% NH₃ in helium) studies were performed using an Autosorb IQ (M/s. Quantachrome USA). NH₃ was adsorbed on catalyst at 100 °C by maintaining (5% NH₃ in helium) flow for 30 min followed by purging at the same temperature with helium gas (20 cm³ min⁻¹) for the removal of the physisorbed NH₃. Desorption of NH₃ was conducted by maintaining helium gas flow and simultaneously increasing the temperature from 100 to 700 °C at a heating rate of 5 °C/min. Temperature programmed desorption (TPD) of CO₂ (5% CO₂ in helium) studies were performed using an Autosorb IQ (M/s. Quantachrome USA). CO₂ was adsorbed on catalyst at 100 °C by maintaining (5% CO₂ in Helium) flow for 30 min followed by purging at the same temperature with helium gas (20 cm³ min⁻¹) for the removal of the physisorbed CO₂ molecules. Desorption of CO₂ was conducted by maintaining helium gas flow and simultaneously increasing the temperature from 100 °C to 700 °C at a heating rate of 5 °C/min.

3.1 Xrd

Fig. 2, shows the XRD patterns of Co catalysts, MgO, γ-Al₂O₃ and MgAl₂O₄ support. XRD pattern of MgAl₂O₄ indicate the signals at 2θ values of 36.8, 44.7° and 65.1 corresponding to the crystalline phase of MgAl₂O₄ (ICDD No.86-2258). The clear absence of MgO reflections (No. 77-2364) at 2θ values of 42.9° (200 plane) and 63.3° (220 plane) was due to mild calcination temperature of MgAl₂O₄ catalyst (650 °C), in addition MgO particles were in poor crystalline state. XRD analysis of Co catalysts displayed reflections due to proper interaction between Co₃O₄ and MgAl₂O₄ which were resulted in considerable insertion of Co particles into MgAl₂O₄ spinel support. In addition to these, reflection at a 2θ value of 58.8° was visible due to the presence of MgCo₂O₄ crystallines (ICDD No. 02-1073). It was noteworthy to mention here that the peaks of MgAl₂O₄, Co₃O₄ and MgCo₂O₄ phases were very close to each other. Possible arrangement of MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species due to the fractional replacement of Mg by Co ions adopting the similar spinel structure of MgAl₂O₄ that cannot be ruled out (Profeti et al., 2009). It was reported the formation of crystalline MgAl₂O₄ phase owing to an elevated calcination temperature of Mg Al precursors (Nassar et al., 2014). The noticeble absence of MgO crystalline phase in Co incorporated MgAl₂O₄ catalyst (CMA) further strengthen the fact of mostly replacement of Mg ions by Co ions to form MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ phase and little bit of fine dispersed MgO particles. Further, no XRD peaks pertaining to CoAl₂O₄ species were found in all these catalysts, obviously confirms the complete formation of isolated solid solution (MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄) particles. In addition, no releated peaks to Co₃O₄ was observed, thus because of no clear distinguish between 2θ values of Co oxide and solid solution species. Crystalline size of MgAl₂O₄ or MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ phase in these catalysts were determined by debye-scherrer equation with high intense XRD reflection at 2θ value at 36.8 and values were shown in Table 1. Crystalline size of 18.3 for MgAl₂O₄ has been enhanced to 21.2 nm upon deposition of 1.0Co ions (1.0CMA). Beyond this 1.0Co loading (1.0CMA) the crystalline size remarkably increases to 35.7 nm in 1.25CMA catalyst. Thus owing to existence of bulk MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ and cobalt oxide species supported by substantial increment in peak intensities (see XRD Fig. 1). The crystalline texture of 1.0CMA catalyst (21.2 nm) was significantly smaller than its counterpart, 1.25CMA (35.7 nm). Hence, fine dispersion of solid solution particles were investigated in 1.0CMA compared to that of 1.25CMA catalyst. XRD patterns of MgO, γ-Al₂O₃, CA and CM catalysts were also included in Fig. 2, wherein γ-Al₂O₃ and MgO catalysts were displayed their characteristic peaks. CA catalyst shows the corresponding Co₃O₄ and γ-Al₂O₃ peaks with mild intensity indicate the strong interaction between Co₃O₄ and γ-Al₂O₃ species, besides accumulation of more number of cobalt oxide particles in the tetrahedral sites on γ-Al₂O₃ surface (Shibiao et al., 2007). It was clearly endorsed by high temperature reduction signals in the TPR patterns of CA catalyst. While, Co₃O₄/MgO pattern indicate the existence of low intense Co₃O₄ and MgO peaks, which results in uniform dispersion of cobalt oxide particles and formation of MgCo₂O₄ solid solution species (Fattah et al., 2014).

Close

Fig. 1

The preparation diagram of Co/MgAl₂O₄ (CMA) catalysts.

Export to PPT

Close

Fig. 2

XRD (a) CM, (b) CA, (c) γ-Al₂O₃, (d) MgO, (e) MgAl₂O₄, (f) 0.5CMA, (g) 0.75CMA, (h) 1.0CMA and (i) 1.25CMA catalysts.

Export to PPT

3.2 Surface area

Surface area of 0.5CMA, 0.75CMA, 1.0CMA, 1.25CMA, CM, CA and MgAl₂O₄ catalysts were shown in Table 1. High surface area of MgAl₂O₄ support (150 m²/g) was beneficial for well dispersed Co particles and formation of MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species. The surface area of MgAl₂O₄ (calcined at 650 °C for 5 h) was found to be 150 m²/g, more or less similar to reported ones (Nassar et al., 2014). The gradual decrease in the surface area of MgAl₂O₄ after deposition of Co might be due to blockage or formation of active MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ particles via slight increment in crystalline pattern. In addition, mutual interaction between Co₃O₄, and MgAl₂O₄ particles leads to smooth increment in crystalline texture up on increasing Co loading upto 1.0Co atomic concentration on MgAl₂O₄ support.

No drastic changes in surface area of Co doped catalysts (average ∼80 m²/g, Table 1) against Co loading that significantly indicate the presence of either well uniform formation of MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ or small fraction of Co oxide particles. Though, surface area of 1.25CMA catalyst (61 m²/g) was noticebly low compared to that of 1.0CMA catalyst (81 m²/g) due to more fraction of larger size Co₃O₄ ions along with adverse solid solution particles. Considerable growth in larger crystalline size (35.7 nm), of the 1.25CMA catalyst was evidenced by high temperature reduction texture (H₂-TPR analysis), wherein bulky MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ and Co₃O₄, and species from TEM image.

3.3 Reductive behavior of cobalt catalysts

Temperature programmed reduction (H₂-TPR) profiles of CA, CM, 0.5CMA, 0.75CMA, 1.0CMA and 1.25CMA catalysts were shown in Fig. 3. TPR patterns shows the presence of two reduction maxima, one in a temperature range of 300–500 °C and other in a range of 600–800 °C. Bulk Co₃O₄ ions gets reduced at around 380 °C (Meng et al., 1997). Since pure Co₃O₄ ions gets reduced at around 380 °C, one can presume that the lower temperature signal results in reduction of Co₃O₄ particles. Some reports claim that the mild temperature reduction owing to conversion of more number of Co₃O₄ species into metallic Co via CoO (Jongsomjit et al., 2001; Schanke et al., 1995). The high intense high temperature signals might be due to the reduction of Co³⁺ species, along with strongly interacted cobalt oxide species with the support. In the present study, TPR patterns of supported Co catalysts displayed a facile three reduction maxima, except 1.25CMA catalyst (two reduced peaks shown at temperature maxima i.e, 300–650 °C and 700–1100 °C) because of strong interaction between Co₃O₄ and MgAl₂O₄ with larger size MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species. However, there was a disagreement for this proposal. It was proposed that the low temperature signals are due to the reduction of Co₃O₄ to CoO with only small portion of CoO being reduced to metallic Co (Das et al., 2003; Jacobs et al., 2004). The broad signal was due to reduction of CoO to Co⁰. If the step wise reduction of Co³⁺ to Co²⁺ to Co⁰ occurs, the ratio between the intensities of low temperature and high temperature signals should be 1:3 (Lin and Chen, 2004). In the present TPR patterns of 1.0CMA catalyst the ratio of intensities of low and high temperature signals were not 1:3 (Profeti et al., 2009). It was because of partial reduction of Co₃O₄ species and solid solution particles. Other possibility for high temperature reduction (∼<1000 °C) might be ascribed to the reduction of MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ phase (Profeti et al., 2009; Wang and Ruckenstein, 2001). In the TPR section, high temperature signal was ascribed to the reduction of solid solution or Co²⁺ reduction to Co⁰ species. The reduction texture in 1.25CMA was absolutely contrary to 1.0CMA catalyst because of drastic accumulation of MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species as well as Co₃O₄ particles. However, three reduction signals were observed in 0.5CMA, 0.75CMA and 1.0CMA catalysts; the first reduction signal at low temperature region was due to the reduction of Co₃O₄ to CoO species. The next reduction signal describe the reduction of CoO to Co, while broad high temperature signal results in reduction of bulk Co⁺³ species and Co⁺² species along with MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ particles in the 0.5CMA, 0.75CMA and 1.0CMA catalysts. Herein, the reduction trend in 0.5CMA and 0.75CMA more less similar to 1.0CMA catalyst, except higher temperature (700–1100 °C) reduction signal. Which was significantly due to stronger diffusion of cobalt particles into MgAl₂O₄ lattice of 0.5CMA and 0.75CMA catalyst comparision to the 1.0CMA. On the other hand, 1.25CMA reduction manner in which, low temperature signal might be due to conversion of Co₃O₄ to CoO species. Whereas, high temperature broad reduction signal owing to reduction of CoO to Co and MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species with bulk cobalt oxide particles. Thus owing to either strong interface between Co₃O₄ and MgAl₂O₄ particles or partial reduction of Co ions. In addition, surface area of MgAl₂O₄ support did not provide enough space for uniform dispersion of 1.25 cobalt loading. Thereby, 1.25CMA catalyst clearly facilitate layers of cobalt oxide particles on the catalyst surface results in two broad reduced peaks with high H₂ consumption. Therefore H₂ consumption seems be to higher on 1.25CMA comparision to the reduction atmosphere of 1.0CMA catalyst (see Fig. 3, TPR pattern). TPR patterns of CM catalyst displayed two reduction maxima, first one indicates the reduction of surface Co₃O₄ to CoO to Co species (<400 °C), while later peak ascribed to the reduction of MgCo₂O₄ species (∼800 °C) (Fattah et al., 2014). In addition, CM catalyst also showed two reduction peaks, first one reveals the reduction of Co₃O₄ to CoO species, whereas, second one deals with reduction of CoO to Co species along with strongly interacted cobalt species with alumina support.

Close

Fig. 3

H₂-TPR (a) 0.5CMA, (b) 0.75CMA, (c) 1.0CMA, (d) 1.25CMA, (e) 1.0CM and (f) 1.0CA catalysts.

Export to PPT

3.4 UV–Vis DRS patterns

The absorption signals in series of 1.0CMA, 1.25CMA and MgAl₂O₄ catalysts were complied in Fig. 4. Broad absorption signals at 430 nm and 720 nm in all Co catalysts are ascribed to the presence of Co atoms in octahedral and tetrahedral environment (Profeti et al., 2009; Liotta et al., 2003). Besides, signal at 250 nm reflect the charge transfer from O⁻² to Co⁺³ in Co embedded MgAl₂O₄ catalysts (Pochamoni et al., 2015; Profeti et al., 2009). Absorption reflection in MgAl₂O₄ beyond 200 nm was described to charge transfer band from O⁻² to Al³⁺, due to excitation of electrons from valence band of O (2p) to conduction band of Al (3d) similar to reported studies (Profeti et al., 2009; Nassar et al., 2014). No absorption reflection for the inactive CoAl₂O₄ phase at 620 nm indicates the complete formation of CMA active phase. Further, CMA spinel formation is experimentally evidenced from XRD and TPR characterizations (Profeti et al., 2009). 1.25CMA reflected a similar UV absorption environment as 1.0CMA catalyst.

Close

Fig. 4

UV–Vis DRS patterns of (a) MA, (b) 0.5CMA, (c) 0.75CMA, (d) 1.0CMA, (e) 1.25CMA catalysts.

Export to PPT

3.5 FT-Ir

Fig. 5, shows the FT-IR spectra of CMA and MgAl₂O₄ catalysts and recorded between the range of 400 and 4000 cm⁻¹. IR absorption reflections at 3430, 1630, 1384, 670 and 569 cm⁻¹ are visible in all these catalysts. The broad band at 3430 cm⁻¹ could be attributed to stronger absorption stretching band of OH groups. Signal at 1630 cm⁻¹ was assigned to deformation band of interlayer (H—O—H) water molecules O—H coordination to surface of CMA and MgAl₂O₄ species.

Close

Fig. 5

FT-IR patterns of (a) 0.5CMA, (b) 0.75CMA, (c) 1.0CMA, (d) 1.25CMA and (e) MA catalysts.

Export to PPT

The absorption band at 1384 cm⁻¹ might be attributed to NO₃⁻ stretching vibrations, it was due to an inadequate decomposition in nitrate precursor. Absorption bands in a range of 900–500 cm⁻¹ was mostly ascribed to metal-oxygen (M—O) stretching vibrations, where M-denotes either Al or Mg. The signal at 700 cm⁻¹ describe the Al³⁺ in the form of octahedral sites, more over absorption signal at 750 cm⁻¹ was explained to Al³⁺ in tetrahedral sites (Nassar et al., 2014; Tripathy and Bhattacharya, 2013), explained the formation of MgAl₂O₄ single active phase at different calcination temperatures from 500 to 900 °C. Complete formation of MgAl₂O₄ single active phase has been pragmatic at 900 °C. While the mild calcination temperature results in disorder MgAl₂O₄ spinel form (<500 °C). In general the absorption bands at 569 and 670 cm⁻¹ was clearly ascribed to Co₃O₄ that Co-O stretching vibrations in octahedral and tetrahedral environment of Co ions (Pochamoni et al., 2015, Taghavimoghaddam et al., 2012)

3.6 TEM analysis

TEM images of fresh 1.0CMA and 1.25CMA catalysts with selective area electron diffraction patterns (SAED) were shown in Fig. 6a–d. In Fig. 6a, the fine dispersion of medium size dense MgCo₂O₄ and Mg_xCo_(1−x)Al₂O₄ species could be seen along with few isolated Co₃O₄ particles which cannot be ruled out. Whereas in Fig. 6b, the dark black spots are observed indicating the presence bulk Mg_xCo_(1−x)Al₂O₄ and accumulated Co₃O₄ species it was further evidence from XRD, CO₂ pulse chemisorption, and H₂-TPR techniques. The SAED patterns in 6c and 6d images clearly revealed that the planes exhibited for 1.0CMA and 1.25CMA catalyst are in close correlation with existed peaks in XRD texture. Whereas, Mg and Al species exist as Mg_xCo_(1−x)Al₂O₄ species after the significant cobalt content doping.

Close

Fig. 6

TEM image of fresh (a) 1.0CMA, (b) 1.25CMA catalyst, (c) fresh SAED image of 1.0CMA, (d) 1.25CMA catalysts.

Export to PPT

3.7 TGA (Thermogravimetric analysis)

As shown in Fig. 7, the significant weight loss of carbonaceous (30.2%) observed on spent 1.0CMA catalyst under N₂ gas flow, owing to the combustion of deposited hydrogenated carbon. While changing the carrier gas from N₂ to CO₂ in the EB oxidative dehydrogenation significant decrease in carbonaceous deposit from 30.2 to 20.5% was observed. 1.25CMA obtained high coke deposite (26.4%) was responsible for mild thermal stablity and surface area due to bulk solid solution species and cobalt oxide ions. Herein, CO₂ molecules could effectively hinder the development of significant carbonaceous material over the surface of 1.0CMA rather than 1.25CMA catalyst (some extent coke gets oxidized into CO₂ molecules during the course of reaction, those involved in an improving catalytic efficiency via RWGSR). Hence, it was apparent that the RWGSR phenomena observed in existed catalysts (Madhavi et al., 2014; Guo et al., 2011; Gonzáleza and Moronta, 2004; Braga et al., 2011; Raveendran Shiju et al., 2011; Holtz et al., 2008; Liu et al., 2008; Sakurai et al., 2000; David Raju et al., 2007; David Raju et al., 2006; Madduluri et al., 2020; Gurram et al., 2016; Jong-San, et al., 2003). Considerable CO yield obtained from remarkable RWGSR involvement in 1.0CMA compared to that of 1.25CMA catalyst. After 20 h of ODH-EB catalytic test thus 1.0CMA catalyst afforded 20.5% carbon weight loss (see Fig. 7), while 1.25CMA stayed at 26.4% indicating that prolonged reaction does not result in severe coking in the presence of CO₂ rather than N₂ gas flow.

Close

Fig. 7

TGA analysis of 1CMA and 1.25CMA catalysts.

Export to PPT

3.8 CO₂ pulse chemisorption

The collective amounts of CO₂ adsorbed according to repetitive injections of CO₂ pulse at 600 °C on 0.75CMA, 1.0CMA and 1.25CMA catalysts were plotted in Fig. 8. It was confirmed from the pulse experiment the dissociative adsorption of CO₂ was done on reduced CMA catalysts. Judging from the pulse results, CO₂ adsorption on the surface of 1.0CMA was noteworthy beneficial than 1.25CMA (>41 µmoles) and 0.75CMA (>49 µmoles/g) catalysts. Hence, 1.0CMA adsorbed higher ∼70 µmoles/g of CO₂ at 600 °C, while 1.25CMA exhibited only >41 µmoles/g. This results in facile re-oxidizability of more number of cobalt ions present in the 1.0CMA compared to that of 1.25CMA and 0.75CMA catalysts. But, CO₂ adsorption on 1.25CMA perceptibly poorer than 0.75CMA catalyst, because of drastic increment in crystalline size as well as noticeable decline in surface area (61.0 m²/g). As per efficient CO₂ utilization, 1.0CMA catalyst was prompt for remarkable Co particles alignment on surface of MgAl₂O₄, which leads to desired EB conversion and ST selectivity. Beyond 1.0Co atomic ratio that was 1.25CMA, drastic changes in the catalytic texture with accumulation of more number disordered MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species along with cobalt oxide ions. Moreover, we have clearly investigated the surface catalytic properties of 1.25CMA catalyst by applying H₂-TPR, surface arae, TPD, TEM and XRD techniques.

Close

Fig. 8

CO₂ pulse chemisorption patterns of 1.0CMA, 0.75CMA and 1.25CMA catalysts.

Export to PPT

3.9 NH₃-TPD analysis

The acidic properties have been determined by the well-known NH₃-TPD technique for the CMA, MA, CA and CM catalysts. As shown in Fig. 9, and Table 2, the NH₃-TPD patterns indicate the wide distribution of acidic sites from mild to strong acidic sites in CMA and CA catalysts. The strong acidic sites in MgAl₂O₄ catalyst are much less in number, but homogeneous distribution from weak to moderate acidic sites due to the presence of higher Mg⁺² ions. In CM catalyst weak to moderate acidic sites were more in number, whereas the strong acidic sites were comparatively lower to the CA catalyst. The basic nature of MgO in CM catalyst noticebly decrease in acidic nature of cobalt oxide particles. In the case of CA and CMA catalysts the acid site distribution and their number were quite different from CM and MgAl₂O₄ (MA) catalysts.

Close

Fig. 9

NH₃-TPD patterns of (a) MA, (b) CM, (c) CA and (d) CMA catalysts.

Export to PPT

The number of weak to moderate acidic sites in CMA catalyst was very high, but the strong acidic sites are lower comparable to that in CA catalyst. The higher number of acidic sites on acidic sites on CA catalyst has adverse effect on ST selectivity; it was clearly results in optimal acidic nature of alumina species. In the case of CMA catalyst only weak to marginal acidic sites were present in adequate fraction, which was originated from synergistic effect in the catalytic atmosphere. The solid solution properties seem to be major responsible in yielding superior catalytic activity in terms of better EB and ST selectivity. More number of strong acidic sites and few weak to moderate acidic sites present in the CA catalyst, which stayed at poor activity in terms of EB conversion and ST selectivity. While, MA and CM catalysts were consisted lesser number of weak to mild acidic sites compared to that of CMA, these are not exhibiting good activity, thus because of balanced acidic-basic properties needed for sustained catalytic efficiency. Remarkable, CMA catalyst possesses desired fraction of weak to moderate acidic sites along with strong acidic sites facilitated enhanced activity both in terms of EB conversion and ST selectivity.

5.1 Effect of Co loading on the activity and product distribution

All catalysts were tested in CO₂ flow at 600 °C and obtained results were tabulated in Table 4. MgAl₂O₄ exhibited lesser oxidative dehydrogenation activity with 35.3% EB conversion and ∼80% ST selectivity that was not prompt in the view of better catalytic affordance. Major by-products on MgAl₂O₄ spinel were toluene and benzene, facilitated by adverse hydrogenolysis of EB. It was due to significant absence of surface acidic properties and synergestic interface which results of active cobalt oxide content on the catalytic surface to get beneficial EB conversion and ST selectivity. In this perspective adequate active MgCo₂O₄ and Mg_xCo_(1−x)Al₂O₄ particles were required to enhance EB conversion along with ST selectivity via superior chemical interface and mechanical stability. Herein, 1.0CA catalyst in which γ-Al₂O₃ as a substantial acidic support appears to maintain mild (92.8%) ST selectivity with mariginal 53.3% EB conversion (see NH₃-TPD, Fig. 9). MgO as a basic support in CM catalyst favor towards inferior 42.2% EB conversion accompanied by 99.0% ST selectivity (from CO₂-TPD, Fig. 10). An efficient interaction between Co₃O₄ and γ-Al₂O₃ is much higher towards uniform Co particles dispersion rather than MgO support, because MgO obviously possess mild surface area, shown in Table 1. Though, basic nature of MgO and MgCoO₄ species in the Co₃O₄/MgO catalyst might clearly helpful in beneficial ST selectivity compared to the Co₃O₄/Al₂O₃. As reported by (Mohan et al., 2012), MgO easily transformed into Mg(OH)₂ in the presence of H₂O as by product during the course of reaction. But in the current study, majorly avoid the formation of MgO particles via survival of MgAl₂O₄ species in all CMA catalysts, consequently better catalytic activity comparision to the Co₃O₄/MgO catalyst. To overcome adverse catalytic properties on CA and CM catalysts, we made cobalt consisted MgAl₂O₄ catalysts which were afforded outstanding EB conversion and ST selectivity. Therefore, addition of 0.5Co content to MgAl₂O₄, increase the EB conversion steadily up to 61.5% with an improved ST selectivity compared to that of bare MA catalyst. While, 0.75CMA achieved 70.4% conversion with 96.1% ST selectivity owing to mild surface Co particles and adequate synergistic properties in the catalyst environment. Outstanding improvement in catalytic action of Co loading reaches to maximum 81.2% EB conversion on 1.0CMA catalyst, because of more number of active Co particles distribution evidence from CO₂ pulse chemisorption and facile reducible cobalt particles in the form of solid solution species (from TPR pattern, see Fig. 3). Therefore 1.0CMA catalyst in which formation of fine synergistic MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ particles plays a major role in better catalytic efficiency for prolonged TOS measurements. In contrast, agglomerated disorder MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species and Co₃O₄ particles were observed in 1.25CMA catalyst (see TEM image 6). Hence, 1.25CMA catalyst achieved a mild 67.0% EB conversion, which was appearent evidence from low surface area and complex reduction texture. In addition larger crystalline size (crystalline size 35.7 nm) facilitated negative catalytic features in the 1.25CMA catalyst, interms of mild CO₂ utilization during the course of reaction (investigated by CO₂ pulse chemisorption). As a result, TOS on 1.25CMA did not afford prolonged catalytic stability, thereby EB conversion decline to >50% at 20 h with 95.0% styrene selectivity. Whereas, uniform Co moieties distribution leads to facile reduced MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ texture (from TPR pattern, surface area and TEM analysis), which beneficial responsible for EB conversion and ST selectivity over 1.0CMA catalyst. Among the all CMA catalysts, 1.0CMA stayed at 81.2% EB conversion with 98.0% ST selectivity compared to other Co catalysts (0.5CMA, 0.75CMA, 1.25CMA, CA and CM catalysts). The large crtstalline pattern influence of MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ along with Co species in 1.25CMA catalyst makes it high compared to the 1.0CMA catalyst. As per the aforementioned Co catalysts comparison study, 1.0CMA was remarkably better catalyst in yielding superior ST monomer along with EB conversion.

5.2 Influence of reaction temperature in catalytic activity

The EB oxidative dehydrogenation on 1.0CMA and 1.25CMA catalysts were performed at various temperatures such as, 450–650 °C and results were depicted in Fig. 11a and b. Increase in EB conversion with increasing temperature was due to endothermic nature of oxidative dehydrogenation reactions and showed that temperature had prominent requirement in these reactions. ST selectivity at different reaction temperatures of 450 °C, and 550 °C was 100% and an irrelevant decrease in ST selectivity observed until 600 °C, since high temperature favors cracking of EB C—C bond on surface of the catalyst. The high temperature (>600 °C) oxidative dehydrogenation results in a smooth EB cracking into carbonaceous deposits (CO_X). But, ST selectivity better on 1.0CMA at 600 °C (better thermal and mechanical stablity) compared to the 1.25CMA catalyst, it was due to much better synergistic effect of cobalt, magnesium and alumina species as solid solution texture (MgCo₂O₄ and Mg_xCo_(1−x)Al₂O₄). Therefore, 1.0CMA afforded a remarkable catalytic properties in terms of an optimised acidic-basic texture (from NH₃ and CO₂-TPD analysis) and formation of active MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species along with Co particles (from CO₂ pulse chemisorption, see Fig. 8). On the other hand, during online analysis beneficial CO (>11% at 1 h) yield gained on 1.0CMA rather than 1.25CMA catalyst (>7.3% at 1 h), thus because of mutual interaction between 1.0CMA catalyst active species and CO₂ molecules. Herein, systematically compared the catalytic efficiency of 1.0CMA with existed cobalt based (Do-Young et al., 2005; Moronta et al., 2006) and other solid oxide catalysts, see Table 5.

Close

Fig. 11

A and B, Temperature effect in CO₂ and N₂ flow on (a) 1.0CMA and (b) 1.25CMA. Catalysts. Reaction conditions: catalyst weight: 1 g, reaction temperature: 600 °C, CO₂ and N₂ flow: 25 cm³ min⁻¹, EB feed flow: 1 cm³ h^−1.

Export to PPT

5.3 Time on stream study in CO₂ and N₂ flow

1.0 CMA and 1.25CMA catalysts time on stream study (TOS) were tested continuously in both N₂ and CO₂ oxidants and results were displayed in Figs. 12–15. Even though, in N₂ flow the EB conversion was maintained at 60% (at 1 h TOS) but ST selectivity come down to ≤90% after initial high selectivity. The low activity was because of considerable carbonaceous deposits accumulated on the catalyst surface through over hydrogenolysis of EB during the course of reaction. Despite, N₂ as carrier gas 1.0CMA catalyst attained initial ∼64% EB conversion and 90% ST selectivity until 20 h TOS, it was due to thermal steady ness of the catalyst (in TOS Fig. 10). On the other hand, better EB conversion and ST selectivity in CO₂ flow was a great responsible in combination of acidic-basic properties (from NH₃ and CO₂-TPD analysis) along with synergistic effect in 1.0CMA catalyst that results in high CO liberated via reverse water gas shift reaction (RWGSR). Superior CO₂ molecules conversion during online analysis, clearly endorsed in terms of (from CO₂ pulse chemisorption, Fig. 8), better CO yields depicted in Fig. 12, (Madhavi et al., 2014; Madduluri et al., 2020; Jong-San, et al., 2003). CO₂ consumption at 600 °C was significantly better on 1.0CMA, but it was contrast in 1.25CMA catalyst under similar reaction conditions. Hence, 1.0CMA afforded excellent 81.2% EB conversion in the presence of CO₂ flow due to active homogeneous MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ species as well as cobalt oxide ions. More number of CO₂ molecules available on 1.0CMA catalyst surface that could be convertd into CO species (evidence from lesser coke deposited on spent 1.0CMA catalyst) with involvement of more CO₂ molecules in EB dehydrogenation via RWGSR (CO₂ + H₂ = CO + H₂O). Thus results in more fraction of carbonaceous deposits could be oxidized in to CO₂ molecules on 1.0CMA surface rather than 1.25CMA catalyst, it was better line with reported study (David Raju et al., 2006; Li et al., 2015; Madduluri et al., 2020). As a result, >11% CO yield was high compared to H₂ evolution during the catalytic run, wherein this scenario conclude that the great involvement of H₂ in RWGSR over 1.0CMA catalyst. While, 1.25CMA achieved mild >60% EB conversion with 95% ST selectivity at 1 h owing to lower number of active CO₂ molecules utilization as a result 7.9% CO yield evidence by online gas analysis. More over, considerable increment in crystallite size (crystalline size-35.7 nm) of 1.25CMA catalyst was key responsible for an improper MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ clusters alignment, which was negative effect on catalytic efficiency. The adverse influence in catalytic efficiency and lower surface area properties (61 m²/g) of 1.25CMA catalyst was throughly investigated. As a result, dissimilarity in the catalytic efficiency was found between 1.0CMA and 1.25CMA catalysts in terms of 1.0CMA attained 65% EB conversion at 20 h, while 1.25CMA obtained >40% EB conversion at 20 h, thus results in higher coke deposits on the catalyst surface. RWGSR was also endothermic in which the overall energy consumption for yielding 1.0 ton of styrene was one order magnitude lower than the commercial steam assisted process (Madhavi et al.,2014; David Raju et al., 2007; David Raju et al., 2006). Considerable decline in carbonaceous development in the presence of CO₂ flow (carbon depositwas 20.5% after 20 h TOS on 1.0CMA) significantly lower than N₂ atmosphere (coke deposite was >30.2% after 20 h TOS) (see Scheme 1).

Close

Fig. 12

Analysis of CO yield on 1.0CMA and 1.25CMA catalysts. Reaction conditions: catalyst weight: 1 g, reaction temperature: 600 °C, CO₂ flow: 25 cm³ min⁻¹, EB feed flow: 1 cm³ h^−1.

Export to PPT

Close

Fig. 13

TOS on 1.0 CMA catalyst in N₂ gas flow. Reaction conditions: catalyst weight: 1 g, reaction temperature: 600 °C, N₂ flow: 25 cm³ min⁻¹, EB feed flow: 1 cm³ h⁻¹.

Export to PPT

Close

Fig. 14

TOS of fresh1.0 CMA and 1.25CMA catalysts in CO₂ flow. Reaction conditions: catalyst weight: 1 g, reaction temperature: 600 °C, CO₂ flow: 25 cm³ min⁻¹, EB feed flow: 1 cm³ h^−1.

Export to PPT

Close

Fig. 15

TOS on regenerated 1.0 CMA and 1.25CMA catalysts in CO₂ gas flow. Reaction conditions: catalyst weight: 1 g, reaction temperature: 600 °C, CO₂ flow: 25 cm³ min⁻¹, EB feed flow: 1 cm³ h^−1.

Export to PPT

Close

Scheme 1

Ethylbenzene oxidative dehydrogenation on Co₃O₄/MgAl₂O₄ under CO₂ gas flow.

Export to PPT

Superior ST selectivity to 98% under CO₂ atmosphere rather than N₂ gas flow afforded <90% ST selectivity. As a result, significant reverse water gas shift reaction (RWGSR) observed in the presence of CO₂ flow rather than N₂ gas flow. Therefore, EB cracking under N₂ atmosphere causes much high carbonaceous material deposits on surface catalyst (evidence from TGA, see Fig. 7), while mild carbon deposit and much better activity results were observed in the presence of CO₂ atmosphere. Hence, CO₂ was vital soft oxidant in terms of superior 81.2% EB conversion along with 98% ST selectivity until 20 h TOS over the 1.0CMA catalyst. In further TOS study after 20 h, the considerable decrease in EB conversion was noticed since spent catalyst consists of coke plugging MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ as well as Co₃O₄ ions. To regain its catalytic activity, the spent 1.0CMA catalyst has been subjected to re-oxidation in air, for overnight at 450 °C. Then TOS was performed in CO₂ flow at 600 °C, represented the similar catalytic activity was obtained as fresh 1.0CMA catalyst because of it possesses unusual chemical steadiness with mild coke deposit. In this context, MgAl₂O₄ contributed remarkable solid solution properties in 1.0CMA catalyst with respect to vital resistance towards severe catalytic deactivation with accessible surface area, an adequate porosity, mechanical and thermal properties (Nassar et al., 2014). The reusability of 1.0CMA catalyst and effortless isolation with good styrene yields makes the present study sustainable and an advantageous compared to existed Co and several metal oxide catalysts as shown in Table 3.

5.4 Conclusions

In summary Co content of 1.0 atomic concentration afforded superior 81.2% EB conversion with beneficial 98% ST selectivity. The prompt synergistic influence between Co₃O₄ and MgAl₂O₄ species leads to active solid solution phase (MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄) formation. That was absolute influence in dramatic 81.2% EB conversion rather than individual 1.0CM and 1.0CA catalysts. The survival of active MgCo₂O₄ or Mg_xCo_(1−x)Al₂O₄ clusters identified by XRD (significant insertion of Co₃O₄ particles into lattice of MgAl₂O₄ spinel), UV–Vis DRS and H₂-TPR analysis. On the other hand, disorder cobalt oxide particles alignment on MgAl₂O₄ causes noteworthy increase in crystallite size consequently irregular reduction texture in 1.25CMA catalyst. Spent catalyst (1.0CMA) has been proved to be on par with the fresh catalyst in maintaining unbelievable stability. The competent 1.0CMA catalyst with high surface area and better CO₂ utilization compared to that of 0.5CMA, 0.75CMA and 1.25CMA, results in superior CO formation. Therefore, 1.0CMA concluded as selective catalyst in remarkable catalytic measurements with respect to regeneration after 20 h TOS compared to that of 1.25CMA catalyst. Combined surface area, XRD, H₂-TPR and CO₂ pulse chemisorption results confirm that the 1.0CMA catalyst was better comparision to 1.0CM, 1.0CA, 0.5CMA, 0.75CMA and 1.25CMA catalysts. Hence, 1.0CMA facilitated an easy reducible solid solution species and Co₃O₄ particles with an improved thermal stability for prolonged time. Mild acidic-basic sites in 1.0CMA comparsion CM and CA catalysts via NH₃-CO₂ TPD analysis was major role in acquiring sustainable catalytic activity.