Catalytic Performances of Co/TiO₂ Catalysts in the Oxidative Dehydrogenation of Ethane to Ethylene: Effect of CoTiO₃ and Co₂TiO₄ Phase Formation

Abstract

Co/TiO₂ catalysts with different cobalt loadings (3.8, 7.5 and 15 wt%) were prepared by impregnation method of Co(NO₃)₂ 6H₂O over titania. Samples containing Co(NO₃)₂·6H₂O and TiO₂ in stoichiometric proportions in order to obtain CoTiO₃ and Co₂ TiO₄ phases were also synthesized. The effect of the calcination treatment at two different temperatures, 550 and 1150 °C, was investigated. Characterizations by several techniques, such as XRD, UV–vis–NIR, DRS, Raman and XPS, were carried out. XRD showed the coexistence of three phases: CoTiO₃; Co₂TiO₄ and Co₃O₄ after calcination at 550 °C, while calcination at high temperature (1150 °C) led to single-phase systems (CoTiO₃ or Co₂TiO₄). Diffuse reflection and XPS spectroscopy showed that divalent cobalt occupies octahedral sites in the ilmenite phase, and both tetrahedral and octahedral sites in the spinel phase. The catalytic performances of the prepared catalysts were evaluated in the oxidative dehydrogenation reaction (ODH) of ethane to ethylene, as a function of the Co content for Co/TiO₂ catalysts and as a function of the calcination temperatures for the CoTiO₃ and Co₂TiO₄ phases. Co(7.5)/TiO₂ was the most active, although the conversion of ethane decreased in the first 150 min of the reaction, reaching values comparable to those of Co₂TiO₄ and CoTiO₃; however, Co(7.5)/TiO₂ was confirmed as having the best selectivity to ethylene in comparison with the bulk phases, CoTiO₃ and Co₂TiO₄. The influence of the reaction mixture composition, specifically the presence of water, at different percentages, was investigated. There is a decrease in the overall ethane conversion and an increase in the ethylene selectivity when the percentage of water increases. This behavior can likely be attributed to an increase in the surface concentration of hydroxyl species (OH), resulting in heightened surface acidity.

1. Introduction

The chemical industry heavily relies on unsaturated hydrocarbons as a raw material for various industrial processes. However, the current industrial capacity for light alkenes such as ethene, propene and butene is expected to be insufficient due to the increasing demands of the petrochemical industry [1,2]. Traditional methods for producing light alkenes, such as catalytic or steam cracking of naphtha and fluid catalytic cracking (FCC) of petroleum crude oil, have negative impacts on the environment [3,4,5]. With the increasing concern of climate change, there is a growing need for new, environmentally friendly methods of producing these chemicals [4,6]. It should be noted that in recent years, several commercial and academic proposals have been put forth as alternatives, promising to produce high-value chemicals from sustainable raw materials [7]. The revolution of shale gas and the prevalence of liquefied natural gas (LNG) technologies have increased the production of light alkanes, thus enhancing the interest in their dehydrogenation as a potential way to apply liquefied natural gas. Alkane dehydrogenation can be mainly achieved by nonoxidative or oxidative dehydrogenation [8,9,10]. It should be noted that thermodynamic dehydrogenation at high temperatures can result in the formation of undesirable byproducts. While the oxidation process can improve propylene yield, it can, however, also increase undesirable deep oxidation. Various catalytic formulations have been tested in oxidative dehydrogenation of alkanes, often consisting of complex mixtures of oxides. Thus, catalysts often contain oxides of vanadium, molybdenum and niobium. For example, the Mo–V–Nb–O system has been shown to have a 100% selectivity at 10% conversion for ODH from ethane to ethylene [11]; however, the best results have been obtained using boron oxides on different supports. In alumina catalysts containing 30% B2O3 in weight, the yield of ethylene was 14.6% at 38% conversion [12]. Additionally, a system composed of oxides of Co, Zr, P, Fe and K was found to have a selectivity of 74% at 53% conversion [12]. Oxidative dehydrogenation of ethane (ODH-E) is regarded as a highly promising approach for the conversion of ethane into valuable products [13]. When sustainable and easily accessible sources of ethane are employed in this process, the resulting ethylene production has the potential to satisfy both industrial production standards and environmental requirements. It is important to note that such processes should achieve an ethylene yield of greater than 60% in a single pass, with minimal formation of carbon oxides as a byproduct. Others transition metals have also demonstrated promising outcomes as well. For instance, research on various chromium-based catalysts supported on lamellar phosphates has revealed that CrPO₄ exhibits ethylene selectivity ranging from 60% to 50% and a conversion range from 30 to 20%. This result highlights the role that phosphorus can play as an additive in certain catalytic formulations [14]. ElIdrissi et al. [15,16] have reported that the selectivity for ethylene using a catalyst system composed of V/TiO₂, modified with phosphorus at a 1.64 of P/V ratio, was close to 50% at a conversion rate of 33%. They attribute the high selectivity to the acid–base properties and the V^4+/V⁵⁺ ratio and suggest that the presence of phosphorus adjusts the surface acidity, leading to an increase in the amount of isolated V⁴⁺ ions. These vanadium-based catalysts are still the subject of ongoing research, and several recent articles have been published on their capacity for oxidative dehydrogenation [17].

Cobalt-based supported catalysts have been proven to be effective in various processes such as hydro treatment and Fischer–Tropsch reactions, among others [18]. They have also been used in the coupling of methane to form longer-chain hydrocarbons [19]. The incorporation of cobalt in perovskite-like materials as LnCoO₃ leads to catalysts with an effective performance in total oxidation and pollution control reactions [20]. The cobalt’s ability to act as an oxidant during the oxidation process is due to the presence of oxygen on the surface of the catalyst, when oxygen is present in the process [21]. In ODH of ethane, cobalt has been used either as a supported catalyst to increase stability and selectivity, incorporated as a promoter to improve catalytic performance or explored in new materials [22,23,24,25,26,27,28,29].

Despite the promising conversion and selectivity rates reported in the literature, the industrial application of oxidative dehydrogenation (ODH) is still limited. The current literature reports an ethylene yield of approximately 20%, whereas industrial applications necessitate a productivity exceeding 1 kg C₂H₄·kg cat⁻¹ h⁻¹ [30]. Previous research on cobalt supported on titania has shown promising results for the oxidative dehydrogenation of ethane [12,31]. Furthermore, XRD analysis of the catalyst at the stationary state revealed the presence of the defined phases such as ilmenite CoTiO₃ and spinel Co₂TiO₄. The best ODH performance of ethane was obtained at 550 °C, with a catalyst containing 7.5% wt of Co. Notably, the reaction starts with a conversion of 33% and a selectivity around 75%, and then it decreases to reach a stationary state of 22% conversion and 60% selectivity after 150 min of running [12,31]. In view of climate change and the need to update industrial processes, research has focused on developing catalytic materials applied specially in the recovering and valorization of petrol-derived compounds. The purpose of this work is to investigate Co/TiO₂ systems at different cobalt loadings for providing information on the formation of the defined phases, CoTiO₃ and Co₂TiO₄, and their catalytic properties.

With this aim, we synthesized and characterized the ilmenite CoTiO₃ and spinel Co₂TiO₄ phases using different spectroscopic techniques. Then, we investigated their behavior in the oxidative dehydrogenation (ODH) reaction of ethane in order to correlate their activity and selectivity with the environment and the oxidation state of the active sites.

2. Materials and Methods

2.1. Chemical Products

The TiO₂ anatase used as support in this work was supplied by Rhone-Poulenc and is characterized by a specific surface area of 89 m² g⁻¹.

Other reagents, including Co(NO₃)₂.6H₂O, (Analytical Reagent, Sinopharm Chemical Reagent Co., Ltd., Ningbo, China.), were used without any further purification.

2.2. Preparation of Catalysts

Co/TiO₂ catalysts with cobalt loadings equal to 3.8, 7.5, and 15.0 wt% were prepared according to a previously described method [12,31]. The TiO₂ support was put in contact with a Co(NO₃)₂.6H₂O solution at suitable concentration in order to obtain the desired cobalt content. The suspension was kept under stirring at 80° C until complete evaporation of the water. The recovered solid was dried at 120° C, then calcined at 550° C or at 1150 °C for 4 h under air. The CoTiO₃ and Co₂TiO₄ phases were synthesized as previously described by using TiO₂ and Co(NO₃)₂·6H₂O solution in stoichiometric proportions in order to form the desired phases composition. Calcination at 550 °C and or 1150 °C was performed as reported above.

2.3. Characterization Techniques

Elemental analysis of the catalysts was carried out using an X-ray fluorescence wave-dispersive spectrometer (XRF-1800, Shimadzu, Japan). The source was an X-ray tube with a Rh anode, a voltage of 40 kV, a current of 95 mA and a diaphragm of 10 mm. The real chemical composition corresponded well to the nominal one within ±5%.

Specific surface areas of the samples were measured at −196 °C with a nitrogen sorption technique using the ASAP 2020 equipment (Micromeritics, Norcross, Catalysts 2023, 13, 60,615 of 18 GA, USA). Before the measurements, the powder samples were degassed at 250 °C for 2 h. The specific surface area was calculated via the Brunauer–Emmett–Teller (BET) method in the standard pressure range 0.05–0.3 P/P₀.

The X-ray diffraction patterns were used to identify the phases present in the samples and to determine their crystalline structure. They were collected using a Siemens D500 high-resolution spectrometer with Cu Kα radiation (Kα = 1.540598 A) at room temperature. The data were recorded with a step of 0.05° (2θ) and a counting time of 10 s per step. The assignment of the crystalline phases was based on the JCPDS powder diffraction file cards.

The morphology, composition and elemental mapping distribution of the two selected samples, CoTiO₃ and Co₂TiO₄, calcined at 1150 °C, were analyzed by scanning electron microscopy (SEM) and energy-dispersive X-ray (EDX) analysis on QUATTRO S-FEG-Thermo Fisher equipment.

Diffuse reflection UV–visible–NIR spectra were recorded in the wavelength range of 190 to 2500 nm using a Varian Cary 5E double monochromator spectrometer, which was equipped with an integrating sphere. To establish a reference, polytetrafluoroethylene was utilized.

Laser-Raman (LR) spectra were obtained using a Dilor XY spectrometer in backscatter geometry with a 514.5 nm excitation line from a Spectra Physics Model 165 coherent argon laser, with power kept below 20 mW. The spectrometer was equipped with a monochromator for selecting specific spectral ranges and a Jobin-Yvon 1024 × 56 CCD matrix thermoelectrically cooled as a multichannel detector.

XPS spectra were recorded on a Vacuum Generators MK I spectrometer with an un-monochromated Al Kα (1486.6 eV, 200 W) under about 10⁻⁹ Torr. The spectra were digitized, summed, smoothed and reconstructed using Gauss–Lorentzian components. The measurements were carried out on powdered samples dispersed on an indium plaque using the C1s peak at 285 eV as a reference.

2.4. Catalytic Tests

The oxidative dehydrogenation of ethane (ODH) was carried out in a continuous-flow microreactor made of quartz, operating at atmospheric pressure, in a temperature range of 450–550 °C. The catalyst, which was sieved to a particle size between 125 and 180 μm, was placed in the reactor between two plugs of quartz wool and was purged with pure nitrogen before the reaction. The reaction mixture consisted of 6 vol.% ethane, 3 vol.% O₂ and 91 vol.% N₂. The effluent gases were analyzed using two online chromatographs, one (FID) equipped with a Porapak Q column for hydrocarbons and the other (TCD) with a silica gel column for oxygenated products. At the experimental conditions chosen, the reaction light-off occurred at around 300 °C and produced only ethylene and CO_x. The effect of water presence at different volume percentages ranging between 1.2 and 3.1 vol.%, in reaction mixture at reaction temperature of 550 °C, was also investigated.

3. Results and Discussion

3.1. Characterization

The activity study of the defined CoTiO₃ and Co₂TiO₄ phases was imposed after their formation in the stationary state during a study of the x%/TiO₂ catalysts in ODH of ethane, hence the importance of associating the latter to this study. The X-ray diffraction patterns of the prepared x% Co/TiO₂ samples are displayed in Figure 1. An examination of the obtained results reveals that the sample diffractogram with a low concentration of cobalt 3.8 wt % (Figure 1, curve b) is characterized by peaks of the anatase titania (JCPDS file #84-1286). For the 7.5 and 15.0 wt% cobalt samples, in addition to anatase, a few peaks were observed (Figure 1c,d), denoting Co₃O₄ formation (JCPDS file no. 78-1970). The intensity of the system exhibits a direct correlation with the concentration of cobalt, whereby higher cobalt concentrations lead to increased intensity levels. Additionally, when the Co content increases from 7.5% to 15%, the average particle size ranges from 17 to 19 nm, which is in agreement with the findings reported by Ho et al. [32]. The study revealed the presence of Co₃O₄ crystallites in samples that contain as little as 1.5% of Co within the anatase and rutile mixture, which also has a surface area of 50 m² g⁻¹. The findings of this work are different from those reported previously and by other researchers, which could be a result of the greater surface area of anatase present in our samples. This leads to improved dispersion of Co₃O₄ and potentially the formation of some amorphous species that are not identifiable by X-ray diffraction.

The XRD patterns of CoTiO₃ and Co₂TiO₄ samples calcined at different temperatures are shown in Figure 2A,B, respectively. For both samples, it can be observed that the calcination at 550 °C leads mainly to a mixture of phases: CoTiO₃, Co₂TiO₄ and Co₃O₄.Co₂TiO₄ is in the minority compared with the other phases. Increasing temperature to 750 °C for the CoTiO₃ sample decreases the intensity of the Co₂TiO₄ lines. Similarly, a significant decrease in the Co₃O₄ lines can be observed. Calcination at higher temperature (1150 °C) leads to single-phase systems: the ilmenite CoTiO₃ and spinel Co₂TiO₄ phases.

On the other hand, the study of the ilmenite CoTiO₃ phase by XRD showed that the Ti⁴⁺ and Co²⁺ ions occupy slightly distorted octahedral sites. The ilmenite structure is found in oxides with the formula ABO₃, when the A and B ions have similar sizes and a total charge of +6. The name ilmenite comes from the mineral Fe^IITi^IVO3, which has a similar structure to corundum. It consists of a compact hexagonal arrangement of oxygen ions and cations occupying two thirds of the octahedral sites. Divalent and tetravalent ions form alternating layers. On the other hand, Co₂TiO₄ has an inverse spinel structure, which can be represented by the formula Co²⁺(Co²⁺Ti⁴⁺)O₄, where half of the Co²⁺ ions occupy the tetrahedral sites; the other half of the Co²⁺ ions and Ti⁴⁺ ions occupy the octahedral sites.

In

Table 1. Specific surface areas (m²/g) of Co(x)/TiO₂ and defined phases, CoTiO₃ and Co₂TiO₄.

Samples	Co(3.8)/TiO2	Co(7.5)/TiO2	Co(15)/TiO2	CoTiO3	Co2TiO4
SSA values (m2/g)	65	60	56	a 46	a 45
	-	-	-	b 9.4	b 10

^a: calcined at 550 °C ^b: calcined at 1150 °C.

, the specific surface areas (SSAs) (m²/g) are listed for Co(x%)/TiO₂ samples calcined at 550 °C, as well as for the CoTiO₃ and Co₂TiO₄ phases treated at two different calcination temperatures, 550 and 1150 °C. The SSA values of the CoTiO₃ and Co₂TiO₄ phases are lower than those for Co(x)/TiO₂ samples, and the values decrease by increasing the calcination temperature of CoTiO₃ and Co₂TiO₄. This finding suggests that sintering of the exposed surface and lowering of the active sites is available for catalysis.

In Figure 3, the adsorption/desorption profiles of selected samples are shown. The isotherms are type IV according to the IUPAC classification. According to the decreased values of specific surface area, after calcination at high temperature, a very low amount of N₂ was adsorbed by the CoTiO₃ calcined at 1150 °C, as can be seen from looking at the curves (d) in Figure 3.

SEM and EDX analyses (see Figure 4) were performed to obtain more insight into the morphology and chemical composition of the CoTiO₃ and Co₂TiO₄ samples calcined at 1150 °C. CoTiO₃ has an almost regular morphology, although some grains appear fused and with not well-defined contours (Figure 4A). The Co₂TiO₄ sample shows some distinctive differences: the grains are more defined, compact and uniform than those of CoTiO₃ (Figure 4B). Moreover, the shape is cubic or octahedral, according to the spinel structure of Co₂TiO₄. In both cases, semiquantitative EDX analysis shows a chemical composition very close to the stoichiometric formulae of the solids (Figure 4C,D).

Diffuse reflection UV–visible–NIR spectroscopy offers valuable insights into the oxidation state of transition metals, their coordination and the symmetry of their surrounding environment. Co/TiO₂ absorption spectra in the UV–visible–NIR are displayed in Figure 5. It should be noted that distinguishing between different species is challenging for two distinct considerations: (i) In the 1100–1500 nm wavelength range, the v1 transition of octahedral ions partially overlaps with the v₂ transition of tetrahedral species. In addition, the high coefficient of absorption of the tetrahedral species makes it difficult to detect the first transition. (ii) In TiO₂, a Co²⁺ → Ti⁴⁺ charge transfer can take place. Indeed, an absorption of Co²⁺-modified MgTi₂O₅ in the 400–500 nm wavelength region was assigned to this transition mode [12,33]. It is important to consider Co²⁺ to Ti⁴⁺ charge transfers, since cations of the active phase and support have variable valence. The UV spectrum of TiO₂ shows strong absorption related to the interband transition, and maxima around 320 and 220 nm are attributed to Co²⁺ to Ti⁴⁺ charge transfers. The 3.8 wt% Co sample has slight absorption in the visible range (Figure 5); however, for the 7.5 and 15 wt% Co samples, new bands appear and intensify with increasing Co load. The near-infrared (NIR) region has a broad band in the 1200 and 1600 nm interval, as well as the band of 2″(OH) harmonics of residual hydroxyl groups, which is located close to 1380 nm. The visible range has another band around 700 nm, and a shoulder with a maximum at 420 nm in the range of 400 and 600 nm observed only in the loaded 15 wt% Co sample (Figure 5, curve d). These results indicate the presence of different cobalt species [12].

The UV–visible–NIR spectra of the CoTiO₃ and Co₂TiO₄ samples calcined at different temperatures (550, 750 and 1150 °C) are shown in Figure 6A,B. In general, the calcination temperature drastically changes the position, shape and relative intensities of the bands.

In the case of the CoTiO₃ sample calcined at 550° C, two broad bands, A and B, appear at 596 and 760 nm, respectively. The calcination temperature has no effect on the position of these bands; however, a variation in their relative intensities is observed after calcination. In the NIR part, a notable modification of the maxima of the bands was observed. Indeed, the sample calcined at 550 °C is characterized by the appearance of bands at 1250, 1350 and 1530 nm and a shoulder at 1750 nm (Figure 6A; spectrum a), which is already attributed to the transition ⁴A₂(W) → ⁴ T₁ (F) of divalent cobalt in the tetrahedral site. The treatment at 1150 °C leads to the appearance of an intense and broad band centered at 1530 nm (Figure 6A, spectrum c). Also, note that in the ilmenite CoTiO₃ structure_, the Co²⁺ ions are in octahedral symmetry. In the olivine Co₂SiO₄ structure taken as reference, Co²⁺ occupies the octahedral interstices. The UV–visible spectrum of this compound (data not shown) has a similar appearance to that of the ilmenite CoTiO₃ phase (Figure 6, spectrum c). The bands at 600, 760 and 1530 nm are assigned to the three (⁴ T _1g → ⁴ T _1g (P), ⁴T _1g →^{_ 4} A _2g (F) and ⁴ T _1g → ⁴ T _2g (F)) divalent cobalts in the octahedral sites.

The fundamental ⁴F level of the Co²⁺ ion in the tetrahedral field breaks down into the lowest ones,⁴A₂ ⁴ T₁ and ⁴ T₂. The transition ⁴A₂ (F) → ⁴T₂ (F) is forbidden to the electric dipole approximation and allowed for the magnetic dipole approximation. It is indeed very weak and often appears in the infrared region. The two transitions ⁴A₂ (F) → ⁴T₁ (F) and ⁴A₂ (F) → ⁴T₁ (P) give very high intense absorption bands, often in three components because of the spin–orbit coupling.

The spectrum of the Co₂TiO₄ sample calcined at 550 °C is similar to that of CoTiO₃ calcined at the same temperature of 550 °C (Figure 6B, curve a). After treatment at 750 °C (Figure 6B, curve b), its appearance changes radically in the visible part. Indeed, the appearance of several absorptions is noticed: 588,638 and 688 nm and a shoulder around 750 nm.

At higher calcination temperature, a broad absorption spreads between 1300 nm up to 1820 nm, which has already been observed in the case of soda–lime glasses with divalent cobalt and in tetrahedral symmetry [12]. The two shoulders around 535 and 720 nm are, respectively, attributed to the ⁴T_1g → ⁴T_1g (F) and ⁴T_1g → ⁴T_1g (F) of Co²⁺ in octahedral sites. The band at 720 nm has already been observed in the case of CoNaY zeolites [34].

Laser-Raman spectroscopy is a widely used tool in the field of supported and unsupported catalysis. The obtained data for the three Co/TiO₂ catalysts are depicted in Figure 7A. The bands observed at 397, 514 and 640 cm⁻¹ are characteristic of the TiO₂ anatase structure, (curve a). When 3.8% wt of cobalt was added, the spectrum was not impacted (spectrum b). In contrast, higher cobalt level spectra show bands close to 692 cm⁻¹ and two shoulders at 487 and 528 cm⁻¹, suggesting the formation of Co₃O₄ [12,35].

The Raman spectra of the two CoTiO₃ and Co₂TiO₄ phases calcined at 1150 °C are shown in Figure 7B. As far as the spinel is concerned, no Raman vibration mode was observed in the frequency range between 200 and 1000 cm⁻¹ (Figure 7Ba). On the other hand, the Raman spectrum of the ilmenite CoTiO₃ phase (Figure 7B, spectrum b) shows several bands, respectively, at 688, 450, 379, 332, 263 and 233 cm⁻¹ and two shoulders at 602 and 204 cm⁻¹. The mode of vibration at 688 cm⁻¹ is the most intense. Several isostructural compounds, such as (Ni; Fe; Mg)TiO_3, exhibit the same Raman line around 700 cm⁻¹. It corresponds to the mode of vibration A_1g of symmetrical elongation of octahedron MO₆ [36].

The XPS spectra of the CoTiO₃ and Co₂TiO₄ phases are illustrated in Figure 8A,B, respectively. The spectral lines of the Co2p of these compounds, such as the Co2p_3/2 binding energies and the spin–orbit coupling ΔE (Co2p_1/2-Co2p_3/2), as well as the S/M ratios (satellite intensity/main peak intensity ratio), are reported in

Table 2. XPS characterization data for the Co(x)/TiO₂ samples calcined at 550 °C and crystalline phases, CoTiO₃ and Co₂TiO₄ calcined at 1150 °C.

Samples	Co2p3/2		ΔE b
	B.E. S/M a Ratio
Co(3.8)/TiO2	781.0	0.42	15.4
Co(7.5)/TiO2	780.2	0.27	15.2
Co(15)/TiO2	779.9	0.19	15.1
Co2TiO4	781	0.40	15.8
CoTiO3	781.2	0.47	15.9
Co3O4	779.9	0.18	15.1
CoO	780.4	0.4	16.0

^a S/M: (satellite intensity/main peak intensity); ^b ΔE: spin–orbit coupling (Co2p_1/2-Co2p_3/2).

.

The Co2p_3/2 binding energies (CoTiO₃: 781.2 eV; Co₂TiO₄: 781 eV), as well as the ratio (S/M), are in good agreement with the presence of paramagnetic ions (Co²⁺). The energy difference ΔE (Co2p_1/2-Co2p_3/2) is 16 eV in the case of CoO [12]. It is close to the CoTiO₃ phase ΔE value, which contains only Co²⁺ in octahedral sites. For the spinel phase, a ΔE value of 15.8 eV is observed. It should be noted that this difference in energy of the 2p_1/2 and 2p_3/2 electrons usually allows the distinction of the tetrahedral sites from the octahedral sites; however, in the case of the mixed spinel, this discrimination is practically impossible.

3.2. Catalytic Results and Discussion

In order to find a Co/TiO₂ compound adequate to compare the studied CoTiO₃ and Co₂TiO₄ phases, the experiments were conducted at temperature of 550 °C. Firstly, the conversion with pure TiO₂ as support was less than 5%. At steady state, for some Co/TiO₂ catalysts, the ethane conversion increased with reaction temperature in the 450–550 °C range (Figure 9). The conversion rate increased as the cobalt amount increased, almost reaching a plateau from 7.5 to 15 wt% of Co. The catalyst loaded with 7.5 wt% of Co shows a conversion at 550 °C of around 22.2% with an ethylene yield close to 13.3% (60% selectivity).

This result allows us to select the sample containing 7.5 wt% Co as a reference to compare the catalytic performances of the CoTiO₃ and Co₂TiO₄ phases. Figure 9 shows the variation of ethane conversion as a function of time on Co(7.5)/TiO₂ at 550 °C compared with CoTiO₃ and Co₂TiO₄ catalytic activities. For Co (7.5)/TiO₂, the conversion decreased from 33% to around 22.2% when it reached steady state after 3 h of reaction. During this time, the specific surface was significantly reduced, and the color of the catalyst changed from light gray to green, probably due to the change in the coordination and/or the oxidation state of the cobalt cations. Figure 9 also displays the conversion of ethane with time over the defined phases CoTiO₃ and Co₂TiO₄ calcined at 550 and 1150 °C, respectively. The evolution of the ethane conversion shows that these solids are stable over time regardless of the temperature of their calcination. At 550 °C, the ethane conversion is on the same order of magnitude as that measured with the Co (7.5)/TiO₂ sample. Therefore, it can be surmised that the octahedral Co²⁺ ions present are responsible for the activity [12,31], while Co³⁺ ions play a role in the ethylene selectivity, in agreement with the behavior observed for phosphorus-modified samples [12].

In Figure 10, the conversion curves of ethane vs time, at the temperature of 550 °C, are plotted for CoTiO₃ and Co₂TiO₄ calcined at 550 °C and at 1150 °C, shown in Figure 10A. For comparison, the behavior of Co(7.5)/TiO₂ calcined at 550 °C is displayed, and the selectivity to ethylene is also reported, in Figure 10B. Co(7.5)/TiO₂ appeared as the most active at the beginning of the reaction; however, the conversion of ethane decreased in the first 150 min of the reaction, from the initial value of ~33% to around 22%, reaching values comparable to those of Co₂TiO₄ and CoTiO₃, which achieved conversions ranging between 19.7–21.6% as a function of the calcination temperature (see

Table 3. Catalytic results of CoTiO₃ and Co₂TiO₄ at steady state and activation energy.

	Samples Calcined at 550 °C
Reaction Temperature	CoTiO3				Co2TiO4
	αg	αene	Sene	SCOx	αg	αene	Sene	SCOx
450 °C	12.9	4.1	32	72	13.8	3.8	28	72
500 °C	16.8	5.2	31	69	16.8	4.5	27	69
550 °C	19.9	6	30	70	21.4	4.6	21.5	78.5
Activation Energy, Ea (kJ·mol−1)	21.6				21.6
	Samples Calcined at 1150 °C
Reaction Temperature	CoTiO3				Co2TiO4
	αg	αene	Sene	SCOx	αg	αene	Sene	SCOx
450 °C	13.1	2.5	19	71	13.8	2.6	19	81
500 °C	17	3	18	72	17.7	3.3	19	81
550 °C	21.6	3.8	16	84	19.7	3.8	18	82
Activation Energy, Ea (kJ·mol−1)	24.5				24.4

α_g: overall conversion of ethane; α_ene: ethylene yield; S_ene: selectivity towards ethylene.

). In Figure 10B, the selectivity to ethylene versus time is reported for the Co(7.5)/TiO₂ sample. It resulted in a selectivity that was as high as 75% at the beginning of the reaction, then decreased to around 60% after 150 min of reaction, following a similar trend as the ethane conversion. Anyway, such selectivity values are significantly higher than those registered for the bulk samples, CoTiO₃ and Co₂TiO₄ (see Table 3).

By comparing the catalytic data obtained for the CoTiO₃ and Co₂TiO₄ samples calcined at 550 °C and 1150 °C and summarized in Table 3, the following observations can be made:

• The evolution of overall conversion (Figure 10) shows that these materials are stable over time regardless of their calcination temperature. At 550 °C, the overall conversion of ethane is on the same order of magnitude as that measured with the Co(7.5)/TiO₂ sample (α_g = 22.2%).
• For the calcination temperature of 550 °C, selectivity towards ethylene (Table 3) was slightly higher in the case of CoTiO₃ in comparison with Co₂TiO_4, while the superior behavior of the Co(7.5)/TiO₂ sample was confirmed.
• No significant modification of catalytic activity was observed after calcination at 1150 °C, but the production of CO_x became higher, while selectivity to ethylene dropped down.

The apparent activation energy, Ea, of C₂H₆ oxydehydrogenation on the defined phase catalysts was calculated using the Arrhenius equation (Equation (1)). The relationship between the reaction rate r and the reaction rate constant k was obtained using the theory of reaction series. The expression for the catalytic reaction rate r is identical to that used by Wang et al. [37].

$$k=A. e^{-\frac{Ea}{RT}}$$

(1)

$$r=k.C_1^{\alpha }{C}_1^{\beta }\dots C_n^{\gamma }$$

(2)

$$r=\frac{F\ast \alpha }{22.414\ast 60\ast 1000\ast M}$$

(3)

k is the reaction rate constant, A is the pre-exponential factor, Ea is the apparent activation energy and R is the molar gas constant. In Equation (2), α, β and γ are reaction orders; C₁, C₂ and C_n are reactant concentrations. In Equation (3), α is the conversion of reactants, F is the gas flow rate and M is the mass of the catalyst.

By plotting ln(r) versus 1/T, the apparent activation energy Ea was calculated and the values are listed in Table 3. The values are quite low, slightly increasing for the samples calcined at 1150 °C.

The selectivity to ethylene versus the ethane conversion registered at the reaction temperature of 550 °C on the CoTiO₃ and Co₂TiO₄ samples calcined at different temperatures (550 °C and 1150 °C) is reported in Figure 11. The common trend is that the selectivity to ethylene decreased by increasing ethane conversion. The samples calcined at 550 °C showed similar curves, with almost overlapping values. Samples calcined at 1150 °C exhibited higher selectivity, the CoTiO₃ phase being the most efficient.

Impact of Water Vapor in ODH reaction

In general, water increases the surface acidity and, in this sense, facilitates desorption.

Table 4. Catalytic data from the effect study of H₂O presence in reaction mixture at reaction temperature of 550 °C.

	Samples Calcined at 550 °C
Experimental Conditions	CoTiO3				Co2TiO4
	αg	αene	Sene	SCOx	αg	αene	Sene	SCOx
Dry Mixture	19.9	6	30	70	21.4	4.6	21.5	78.5
1.2% H2O	15.5	6.2	40	60	17.1	4.7	27	73
1.7% H2O	14.7	6.5	44	56	13.9	4.9	35	65
3.1% H2O	13	6.1	47	53	11.9	4.6	39	61
	Samples Calcined at 1150 °C
Experimental Conditions	CoTiO3				Co2TiO4
	αg	αene	Sene	SCOx	αg	αene	Sene	SCOx
Dry Mixture	21.6	3.8	16	84	19.7	3.8	18	82
1.2% H2O	16.7	3.7	22	78	17.6	3.7	21	79
1.7% H2O	15.3	4	26	74	15.6	3.9	25	75
3.1% H2O	13.6	3.8	28	72	12.8	3.6	28	72

summarizes the results obtained when steam was added to the reaction mixture at different percentages. It was observed that both phases, CoTiO₃ and Co₂TiO₄, show the same catalytic behavior in the presence of water, independent of their calcination temperature.

It can be pointed out that there is a decrease in overall ethane conversion and an increase in ethylene selectivity when the percentage of water increases (Table 4). On the other hand, it is also noticeable that the yield of ethylene increases slightly with a reaction mixture containing 1.7% of H₂O.

The comparison of selectivity obtained for the CoTiO₃ and Co₂TiO₄ catalysts can be appreciated from the plots shown in Figure 12. It is possible to classify the samples into two categories: The first one, less selective in ethylene, includes the samples tested with the dry mixture. The second category, which is more selective, is obtained in the presence of steam.

From these obtained results, it can be considered that the limiting step is the desorption of the ethylene molecule from the catalyst surface. The presence of water vapor in the reaction mixture promotes this desorption through competition for site occupation. However, at high percentages of water vapor, a significant portion of the active sites are occupied, leading to a significant decrease in overall conversion. Therefore, the increase in C₂H₄ selectivity is likely due to an increase in the OH* species at the catalyst surface [38] relative to the O* species, which favors total oxidation.

4. Conclusions

In this work, Co/TiO₂ catalysts with varying cobalt loadings (3.8, 7.5, and 15 wt%) were successfully prepared using the impregnation method of Co(NO₃)₂.6H₂O over titania. Additionally, the CoTiO₃ and Co₂TiO₄ phases were synthesized by maintaining stoichiometric proportions of the precursors. The effect of calcination treatments at different temperatures (550 °C and 1150 °C) was investigated. Characterization using various techniques, such as XRD, UV–vis–NIR, DRS, Raman and XPS, provided valuable insights. XRD analysis revealed the coexistence of three phases (CoTiO₃, Co2TiO4 and Co3O4) after calcination at 550 °C and depending on cobalt loading, while high-temperature calcination (1150 °C) resulted in single-phase systems (CoTiO₃ or Co₂TiO₄).

Diffuse reflection and XPS spectroscopy detected the presence of divalent cobalt ions that occupied octahedral sites in the ilmenite phase, as well as both tetrahedral and octahedral sites in the spinel phase. The catalytic performances of the prepared catalysts were evaluated for the oxidative dehydrogenation reaction (ODH) of ethane to ethylene. The Co(7.5)/TiO₂ catalyst exhibited higher activity and superior selectivity to ethylene compared with the bulk phases CoTiO₃ and Co₂TiO₄.

The influence of the reaction mixture composition, particularly the presence of water at different percentages, was also investigated. The overall ethane conversion decreased, and ethylene selectivity increased as the percentage of water in the mixture increased. This behavior is likely attributed to an increase in the surface concentration of hydroxyl species (OH), leading to heightened surface acidity.