Sr₂TiO₄ Prepared Using Mechanochemical Activation: Influence of the Initial Compounds’ Nature on Formation, Structural and Catalytic Properties in Oxidative Coupling of Methane

Abstract

Methane oxidative coupling (OCM) is considered a potential direct route to produce C₂ hydrocarbons. Layered perovskite-like Sr₂TiO₄ is a promising OCM catalyst. Mechanochemical activation (MA) is known to be an environmentally friendly method for perovskite synthesis. Sr₂TiO₄ were synthesized using MA of the mixtures containing SrCO₃ or SrO and TiO₂ or TiO(OH)₂ and annealing at 900 and 1100 °C. XRD and FT-IRS showed that MA leads to the starting component disordering and formation of SrTiO₃ only for SrO being pronounced when using TiO(OH)₂. After annealing at 900 °C, Sr₂TiO₄ was mainly produced from the mixtures of SrCO₃ or SrO and TiO(OH)₂. The single-phase Sr₂TiO₄ was only obtained from MA products containing SrCO₃ after calcination at 1100 °C. The surface enrichment with Sr was observed by XPS for all samples annealed at 1100 °C depending on the MA product composition. The OCM activity of the samples correlated with the surface Sr concentration and the ratio of the surface oxygen amount in SrO and perovskite (O_o/O_p). The maximal CH₄ conversion and C₂ yield (25.6 and 15.5% at 900 °C, respectively), and the high long-term stability were observed for the sample obtained from (SrCO₃ + TiO₂), showing the specific surface morphology and optimal values of the surface Sr concentration and O_o/O_p ratio.

1. Introduction

With large fossil resources of natural and shale gas, gas hydrates promote the development of processes for methane conversion into valuable chemicals. Among them, the direct catalytic conversion of methane to C₂ hydrocarbons by oxidative coupling (OCM) is considered a potential route for processing methane into useful products [1,2,3]. Researchers generally agree that a simplified heterogeneous–homogeneous mechanism of OCM occurs via the activation of methane to methyl radicals on the catalyst surface, the subsequent homogeneous coupling of two methyl radicals to ethane in the gas phase and oxidative dehydrogenation of ethane to ethylene. In parallel, nonselective (homogeneous and/or heterogeneous) oxidation can proceed to give COx. The reaction complexity hinders the development of efficient OCM catalysts [1,2,3]. On the whole, OCM activity of the catalysts is ensured by oxygen activation creating the active sites for generation of methyl radicals and high basicity for fast desorption of methyl radicals that could be tuned by catalyst composition [1,2,3,4]. Although many studies devoted to OCM have been performed and various catalytic materials have been investigated for this reaction, commercial implementation of OCM has not yet been reached, since the C₂ yield is relatively low [1,5,6,7]. Thus, design of the stable OCM catalysts with a high methane conversion rate and C₂ selectivity remains a relevant problem.

A number of previous studies revealed that some perovskite-type oxides containing alkaline, alkaline-earth and rare-earth elements exhibit certain activity and C₂ selectivity for the OCM reaction [1,8,9,10,11,12,13,14,15,16,17,18,19]. Perovskites have high thermal and chemical stability, while concentration of surface defects, active oxygen species and basic sites can be tailored by adjusting their chemical composition. Along with chemical composition, the surface and bulk properties could be affected by the perovskite crystalline structure. All these properties make perovskites promising catalysts for OCM. Among them, much attention has been paid to titanate and stannate perovskite catalysts based on regular ABO₃ [9,10,11,12,13,14,15,16] and layered Ruddlesden–Popper A_n+1B_nO_3n+1 [17,18,19] structures (A = Ca, Ba, Sr, B = Ti, Sn). The perovskite-like Ruddlesden–Popper-type oxides are comprised of alternating layers of ABO₃ perovskite and AO rock salt and exibit high oxygen mobility due to the peculiarities of its fine structure [10,20,21]. It was found in an earlier report by Yang et al. [17] that the layered Sr₂TiO₄ and Sr₂SnO₄ perovskites demonstrated better performance in OCM at 800 °C than the corresponding SrTiO₃ and SnTiO₃. Recently, it was shown that the fine crystal structure of different strontium stannates influences their OCM performance, which follows in the order of Sr₂SnO₄ > Sr₃Sn₂O₇ > SrSnO₃ [10]. The results of multiple methods reveal that the coordination environment of Sr cations is altered depending on perovskite type, making the formation of surface oxygen vacancies easier for Sr₂SnO₄ than for Sr₃Sn₂O₇ and SrSnO₃. This explains why Sr₂SnO₄ has the largest quantities of active surface oxygen and basic sites, as well as the best OCM performance.

The high OCM activity and selectivity of doped Sr₂TiO₄ (C₂ yield up to 25% and C₂ selectivity around 66% at 850–900 ℃) were found in [18] and related to the segregation of SrO on the catalyst surface under the reaction conditions. SrO increases the basicity of the surface and lowers the amount of weakly adsorbed molecular oxygen species that can shift the reaction to the deep oxidation. The effect of Sr surface enrichment resulting from the surface reconstruction of the model SrTiO₃ on OCM activity was studied in [14]. CH₄ conversion, C₂ selectivity and the ratio of C₂H₄/C₂H₆ was found to increase at the Sr enrichment of the surface up to Sr/(Sr + Ti) of 0.66 and then levels off. Thus, the results evidence that the optimal surface concentration of Sr could promote the high OCM activity of the catalysts based on strontium titanates.

It is known that the presence of certain active oxygen species and basic sites determining the high catalytic performance of perovskite-type oxides in methane oxidative reactions are strongly influenced by preparation method and synthesis parameters, such as the nature of the raw chemicals, pH and temperature, etc. [1,2,3]. Along with the optimal surface properties, the effective OCM catalysts could have a low specific surface area (SSA) to prevent unselective homogeneous reactions [1,2,3]. Different methods can be used for synthesis of titanates, including: solid state reaction [11,12,13,15,17], spray and glycine-nitrate combustion [10,11], hydrothermal synthesis [15], coprecipitation [15,19], the polymer precursor method [16,22], the sol-gel method [19,23] and mechanochemical activation [18,19,24,25,26]. Fu et al. [11] investigated Ca, Sr, Ba titanates prepared by solid-state and spray combustion methods for the OCM reaction. They found that the samples prepared by the solid-state method, being of lower crystallinity, showed the better OCM performance due to the high surface concentration of alkaline earth metal cations, which is beneficial for C₂ formation. The study of SrTiO₃ samples synthesized using a solid-state reaction, molten salt and sol-precipitation hydrothermal treatment also shows that the surface atomic structure of perovskite determined by the morphology of crystalline nanoparticles depends on the synthesis method used [15]. The dependence of the texture and morphology of Sr₂TiO₄ prepared via the ultrasonic modified sol-gel method on the alkaline agent adjusting the pH was shown in [21]. It was shown in [19] that the phase composition of the OCM catalysts based on Sr₂TiO₄, their texture features (specific surface area, pore size and volume) and the distribution of active surface oxygen species depends on the synthesis method: co-precipitation, sol precipitation, citrate or mechanochemical activation (MA) methods with further calcinations at 1100 °C. The catalysts prepared by sol precipitation and MA were the most effective for OCM, giving a C₂ yield of about 12% at 800–900 °C. Doped Sr₂TiO₄ possessing high OCM activity and selectivity were also prepared using MA of the mixtures containing TiO₂, oxides and carbonates in [18].

Usually, the sintered complex oxides with a low SSA are prepared by a solid-state method which requires prolonged heating at high temperatures with the intermediate homogenization of the powder precursors yielding phases with nonuniform particle size distribution and variations in the stoichiometry [11,12,13,15,17]. So called “wet” methods, such as coprecipitation, the sol-gel method, spray combustion, etc., involve several steps which take many hours to achieve the formation of the target mixed oxide phase and require the use of various chemicals or special complex equipment. Mechanochemical activation (MA) of starting compounds through high energy milling is the alternative route to avoid these problems and it is an environmentally friendly method due to the absence of any wastes [25,26,27,28,29,30]. During the milling processes, the homogeneity of the mixture increases, the particle size decreases and the new contacts arise. A high-energy input into the reaction zone leads to localized heating and high pressure, resulting in disordering of the crystal structure and the generation of various types of defects in solids that increase their reactivity [29].

In the course of milling, the raw compounds can be transformed into precursors or directly into the target products, depending on their nature and synthesis conditions. The MA effect on the formation of strontium titanates was studied in the early report of Berbenni et al. [25]. The physical mixtures containing SrCO₃ and TiO₂ (rutile) (at a ratio of SrCO₃:TiO₂ = 1:1, 1:2, 3:2) were dry milled for 110–240 h, depending on their composition. SrTiO₃ and Sr₂TiO₄ were not formed during milling and were obtained only after the annealing of the activated mixtures at 800–850 °C for 12 h. MA was applied to obtain Ruddlesden–Popper titanates Sr₂[Sr_n-1Ti_nO_3n+1] (n = 1–4) in [26]. Stoichiometric mixtures of SrO and TiO₂ (anatase) were activated in a planetary mill for 35–300 h. During MA, SrTiO₃ was formed after 35–150 h for all mixtures, except the two SrO:TiO₂ compositions. The kinetic study revealed that traces of Sr₂TiO₄ are observed at 70 h; increasing milling time up to 125 h results in the formation of the very low crystallinity Sr₂TiO₄ phase that is transformed into SrTiO₃ during further milling. After annealing of the corresponding MA products at 800–1200 °C, the single crystalline phases were only obtained for SrTiO₃, Sr₂TiO₄ and Sr₃Ti₂O₇. However, the study showed that MA of the initial mixtures leads to the substantial temperature decrease in the synthesis of all members of the layered Sr₂[Sr_n-1Ti_nO_3n+1] series. It is noted in the paper that this is a result of the grains fracture and defects generated during grinding, which leads to a higher internal energy and reduces the thermal barrier for any subsequent reaction [26].

Thus, the analysis of the previous data shows that perovskite-type layered Sr₂TiO₄ are perspective catalysts for the OCM reaction due to peculiarities of its structure that provide for the large quantities of active surface oxygen and basic sites formed as a result of SrO segregation on the surface. Synthesis of Sr₂TiO₄ with mechnochemical activation is a prospective environmentally friendly method. However, systematic studies on the influence of the raw chemicals’ nature on Sr₂TiO₄ synthesis using MA are absent in the literature. In this work, the interaction of raw compounds during MA in the high-energy planetary ball mill, the impact of MA product peculiarities on phase and surface composition, morphology and microstructure of Sr₂TiO₄ layered perovskite obtained after annealing and its catalytic activity in the OCM reaction are studied.

2. Results and Discussion

2.1. Study of MA Products by XRD and FT-IRS

The influence of the starting compounds nature on their interaction and phase composition of the samples after MA has been studied by XRD and FT-IR spectroscopy.

XRD data for the samples after mechanical activation are presented in Figure 1. The patterns of both MA-1 (SrCO₃ + TiO₂) and MA-2 (SrCO₃ + TiO(OH)₂) show the reflections of the starting SrCO₃ [PDF 05–418], in addition to the reflections of TiO₂ (rutile) [PDF 21-276] for MA-1. TiO(OH)₂ is an amorphous compound. The reflections of starting SrO [PDF 06-0520] and TiO₂ are presented in the pattern of MA-3 (SrO + TiO₂). For MA-4 (SrO + TiO(OH)₂), along with the reflections of SrO, a number of the wide peaks in the regions of 2θ~25–29° and 36–40° are observed that could evidence the presence of a minor admixture of strontium carbonate and hydroxide. In addition, the reflections of SrTiO₃ [PDF 35-0734] appear for MA-3 and MA-4. In the case of MA-4, their intensity is slightly higher, which could be due to some stronger interaction of SrO with TiO(OH)₂ compared to TiO₂.

The FT-IR spectroscopy data give additional information about the influence of the starting compounds’ nature on their interaction during MA and the phase composition of MA products. The FT-IR spectra of the initial TiO₂, TiO(OH)₂, SrCO₃, SrO, and products of MA are presented in Figure 2, Figure 3 and Figure 4. The wide absorption band (a.b.) at ~3470 cm⁻¹ and a.b. at 1640 cm⁻¹ in the spectrum of TiO₂ (Figure 2) could be accordingly assigned to the stretching and bending vibrations of physically absorbed H₂O or its hydroxyl groups [31,32]. The bands in the range of 1250–1050 cm⁻¹ are related to the bending vibrations of Ti-OH, while the a.b. in the range of 950–350 cm⁻¹ are the characteristic peaks of skeletal stretching and bending vibrations of Ti-O bonds in TiO₂ [32]. The spectrum of TiO(OH)₂ shows practically the same absorption bands as TiO₂ but the peaks related to H₂O and OH groups are noticeably higher in intensity [33,34]. Furthermore, two low-intensity a.b. at 1120 and 1050 cm⁻¹ related to SO₄²⁻ [35] are observed (Figure 5), which stems from the synthesis method of TiO(OH)₂ using TiOSO₄ hydrolysis.

The spectrum of the initial SrCO₃ shows absorption peaks at 1770, 1434, 1071, 855, 704 and 702 cm⁻¹ (Figure 2 and Figure 3). The absorption band at 1770 cm⁻¹ is assigned to the bond stretching vibration of C=O in CO₃²⁻ [36]. The intensive a.b. at 1434 cm⁻¹ and the weak a.b. at 1071 cm⁻¹ are related to asymmetric and symmetric stretching vibration of C-O band in carbonate anion, while the a.b. at 855, 704 and 699 cm⁻¹ can be attributed to the bending out-of-plane and in-plane vibrations [36,37,38]. After activation in the spectra of the MA-1 (SrCO₃ + TiO₂) and MA-2 (SrCO₃+ TiO(OH)₂) samples (Figure 2), the bands of SrCO₃ and the wide a.b. at 850–350 cm⁻¹ assigned to the characteristic vibrations of Ti-O are observed. For the MA-2 sample, the wide a.b. of the low intensity at 3400 cm⁻¹ and 1200–1100 cm⁻¹ attributed to vibrations of H₂O and OH groups in TiO(OH)₂ (Figure 2) are presented as well. Thus, in accordance with the XRD data, FT-IR spectra confirm the presence of the initial compounds in the MA products. However, the bands of SrCO₃, TiO₂ and TiO(OH) become wider and they are shifted compared with the starting compounds, which evidences the disordering of their crystal structure during MA. All alterations are more pronounced in the spectrum of MA-2, especially for the absorption bands corresponding to the vibrations of the H₂O and OH groups in TiO(OH)₂ at 3470 cm⁻¹, 1640 cm⁻¹ and 1250–1050 cm⁻¹ (Figure 2).

In the spectrum of the initial SrO a.b. at 3616 and 3472 cm⁻¹ related to the stretching vibrations of OH groups, a.b. at 785 cm⁻¹ and 725 cm⁻¹ attributed to the bending vibrations of the Sr-OH bond and the bands at 1445 cm⁻¹, 850 cm⁻¹ and 705 cm⁻¹ assigned to SrCO₃ are observed (Figure 2 and Figure 4). The high frequency of the bands corresponding to the OH group’s stretching vibrations and their small width evidence the presence of strontium hydroxide [39,40]. Thus, the low intensity of the corresponding absorption bands implies that a small impurity of strontium carbonate and hydroxide presents in SrO, although the starting SrO was obtained shortly before the synthesis. This is explained by its high ability to hydration and carboxylation when exposed to atmospheric CO₂ and H₂O at room temperature and standard pressure [26]. The band at 600 cm⁻¹ and the bands at 550–350 cm⁻¹ correspond to the skeletal stretching and bending vibrations of Sr-O [39,40]. The spectra of the activated MA-3 (SrO + TiO₂) and MA-4 (SrO + TiO(OH)₂) samples are mainly similar (Figure 2 and Figure 4). They show the bands at 3616 and 3472 cm⁻¹ corresponding to the vibrations of OH groups, the bands at 1460 cm⁻¹ and 900–705 cm⁻¹ assigned to SrCO₃ and a pronounced a.b. at ~560–586 cm⁻¹, along with the bands in the range of 500–350 cm⁻¹ that characterize the stretching and bending vibrations of Me (Sr, Ti)-O bonds. Furthermore, the bands shift and the substantial decrease in the intensity of a.b. characterizing the starting TiO₂ and TiO(OH)₂ (Figure 2) are observed, which could be a result of their disordering and interaction with SrO. Thus, the pronounced band at ~560–586 cm⁻¹ assigned to the vibrations of TiO₆ octahedron [41,42] evidences the formation of SrTiO₃ in accordance with XRD data (Figure 1). The bands of SrCO₃ at 900–705 cm⁻¹ are presented in the spectrum of MA-3, while they are absent in the case of MA-4 (Figure 4). This indicates a less effective interaction of SrO with TiO₂ compared to TiO(OH)₂ during MA, as the XRD data show (Figure 1).

Therefore, the XRD and FT-IRS data for MA products demonstrate that, during mechanical treatment, SrCO₃ possesses a lower reactivity compared with SrO, and only the disordering of the starting components is observed in the mixtures containing SrCO₃ (MA-1 and MA-2). Low SrCO₃ reactivity was also found in [25] when strontium titanates were not formed during prolonged milling of SrCO₃ + TiO₂ (rutile) mixture using a high-energy planetary mill at 400 rpm rotation. In the case of SrO-containing mixtures (MA-3 and MA-4), along with disordering of the starting components, formation of SrTiO₃ is observed in MA products. The formation of traces and the very low crystallinity Sr₂TiO₄ after milling of the 2SrO + TiO₂ mixture in the planetary mill at 200 rpm rotation speed for 70 and 125 h, correspondingly, were demonstrated by Hungrıa et al. [26]. In so doing, they did not observe the formation of intermediate SrTiO₃, in contrast with our results (Figure 1 and Figure 4). Such a difference could be because a very-high power planetary ball mill at a 800 rpm rotation rate and acceleration of 40 g was used in our work. The high-energy input in the reaction zone results in the more effective dispersing and mixing of the components [43], as well as their disordering, which increases the number of contact sites and components reactivity, thus facilitating the appearance of SrTiO₃ in 10 min of milling.

XRD and FT-IRS data for MA products have also shown that the changes of the components structure and their interaction under milling are more pronounced in the mixtures containing TiO(OH)₂ (MA-2 and MA-4) compared to the ones with TiO₂. This difference could be attributed to such factors as the amorphous phase of TiO(OH)₂ and a large quantity of OH groups contained in it. It is well known that amorphization of the solid chemicals that can often occur in the course of mechanical treatment increases their reactivity [27,28,29,30]. In addition, the interaction during MA is more effective in the mixtures containing hydrated compounds due to the high reactivity of the OH group and the liberation of water increasing the efficiency of energy consumption in comparison with dry milling [29,44,45]. FT-IR spectra for MA-2 and MA-4 (Figure 2) showed a substantial decrease in the intensity of OH–Ti absorbance bands, which could be due to liberation of water from TiO(OH)₂ facilitating an acid–base reaction at the interface between acidic TiO(OH)₂ and basic SrCO₃ or, especially, SrO, which easily formed hydroxide.

2.2. Calcined Samples

2.2.1. Structural Properties

The data on qualitative and quantitative phase composition of MA products annealed at 900 and 1100 °C obtained by the analysis of XRD patterns are presented in

Table 1. Some characteristics of the samples after annealing of MA products at 900–1200 °C.

Sample/ T °C	Initial Compounds	Phase Composition/ Content,%	Lattice Parameters *, Å			Crystallite size, nm	Specific Surface Area, m2/g
			a	b	c
MA-1 900 1100	SrCO3 + TiO2	Sr2TiO4 (27.1) SrTiO3 (26.4) Sr4Ti3O10 (11) SrCO3 (35.5) Sr2TiO4 (100)	3.8786	3.8786	12.593	120	3 1.2
MA-2 900 °C 1100 °C	SrCO3 + TiO(OH)2	Sr2TiO4 (92.5) Sr4Ti3O10 SrTiO3 SrCO3 Sr2TiO4 (100)	3.8850 3.8756	3.8850 3.8756	12.580 12.561	120 120	1.7 1
MA-3 900 °C 1100 °C	SrO + TiO2	Sr2TiO4 (40) SrTiO3 (57) Sr4Ti3O10 Sr2TiO4 (40) SrTiO3 (56) Sr4Ti3O10	3.8735 3.9049 3.8856	3.8735 3.8856	12.653 12.647	50	1.9 1.4
MA-4 900 °C 1100 °C	SrO + TiO(OH)2	Sr2TiO4 (70) SrTiO3 (14) Sr4Ti3O10 (16) Sr2TiO4 (68) SrTiO3 (14) Sr4Ti3O10 (18)	3.8861 3.8862	3.8861 3.8862	12.549 12.564	50 120	1.6 1.3

* Sr₂TiO₄ [39-1471]: a = b = 3.8861, c = 12.5924.

and Figure 6 and Figure 7. The XRD patterns of MA-1 (SrCO₃ + TiO₂) and MA-2 (SrCO₃ + TiO(OH)₂) calcined at 900 °C (Figure 6) show that they comprise different strontium titanates and the initial SrCO₃, but the quantity of each phase in the samples varies significantly. Thus, MA-1 contains comparable amounts of Sr₂TiO₄ [39-1471], SrTiO₃ [35-0734] and SrCO₃, while for MA-2, the main phase is Sr₂TiO₄ (Table 1 and Figure 6). After annealing at 1100 °C, MA-1 and MA-2 are the single-phase Sr₂TiO₄ (Table 1 and Figure 7). There are no reflections of other phases (SrO or SrSO₄) in their XRD pattern but their presence as highly dispersed or surface compounds cannot be excluded. The phase composition of the MA-3 and MA-4 annealed at 900 °C differs considerably. The comparable quantity of SrTiO₃ and Sr₂TiO₄ is observed in the case of MA-3 (SrO + TiO₂), while MA-4 (SrO + TiO(OH)₂) comprises mainly Sr₂TiO₄. Their phase composition varies little after annealing at 1100 °C (Table 1 and Figure 7).

As an example, the results of Rietveld refinement for MA-2 (1100 °C) and MA-3 (1100 °C) are presented in Figure 7 and Figure 8. The experimental pattern of MA-2 (1100 °C) and the theoretical one obtained using the known structural data for Sr₂TiO₄ [39-1471] (Figure 7) are qualitatively fit (the reliability factor Rwp = 20.12%), confirming the presence of layered Sr₂TiO₄ of tetragonal structure. The comparison of the experimental pattern for the multiphase MA-3 sample (1100 °C) and the theoretical pattern being superposition of the calculated curves for Sr₂TiO₄, SrTiO₃ and Sr₄Ti₃O₁₀ is illustrated in Figure 9.

The lattice parameters and the average size of Sr₂TiO₄ crystallites calculated for the samples contained mainly layered perovskite (Table 1). On the whole, the values of the Sr₂TiO₄ lattice parameters are similar and are closer to the ones known in the literature [21,25]. The narrow peaks in the XRD patterns of MA-1, MA-2 and MA-4 calcined at 1100 °C (Figure 7 and Table 1) evidence the presence of the large, well-crystallized particles with a mean size of about 120 nm. In the case of MA-3 (1100 °C), the broader diffraction peaks of perovskite stem from the smaller crystallites of a size of ~50 nm, which could be due to its multiphase composition.

Therefore, Sr₂TiO₄, along with other titanates, is formed in all samples after annealing at 900 °C because the interaction of the components in the activated mixtures mainly occurs at the temperatures up to 850–900 °C, was shown for mixtures containing SrO or SrCO₃ and TiO₂ [25,26]. However, the phase composition of the samples differs considerably in line with the results for mechanical activation of the corresponding mixtures. Thus, a more pronounced interaction of SrCO₃ or SrO with TiO(OH)₂ compared to TiO₂ during milling (MA-2, MA-4) leads to the formation mostly of Sr₂TiO₄ after annealing at 900 °C (Table 1 and Figure 6). The MA-1 sample obtained from the mixture of SrCO₃ and TiO₂ contains the initial SrCO₃, while only titanates are observed in the MA-3 (SrO + TiO₂) sample. Such a difference is due to the lower activity of SrCO₃ compared with SrO during activation. Indeed, SrTiO₃, being the intermediate in the formation of layered titanates [25,26], is already formed during activation of the mixtures with SrO (Figure 1 and Figure 4).

2.2.2. Surface Composition

The surface composition of the samples calcined at 1100 °C was studied by XPS. The survey spectra of all samples and spectra of Ti 2p_3/2, Sr 3d, O 1s are presented in Figure 10 and Figure 11. For all samples, a symmetric Ti 2p_3/2 peak at 457.4–458.1 eV corresponding to the Ti⁴⁺ in perovskite is observed [46]. All spectra of Sr 3d show two doublets of peaks corresponding to Sr²⁺ 3d_5/2 and 3d_3/2 in oxide and perovskite (Figure 10). The peaks related to Sr²⁺ 3d_5/2 are located at 132.2–132.5 eV and 132.7–133.0 eV for the Sr²⁺ in the oxide and perovskite phase, respectively [47,48,49]. The position of both peaks Sr²⁺ 3d_5/2 (at 132.4 and 132.9 eV) is the same for the single-phase MA-1 and MA-2 annealed at 1100 °C. Some variation in the position of the peaks in the spectra of other samples could be due to their multiphase composition (Table 1), which determines the surface structure features and, as a result, a different charging effect in the position of XPS peaks. O 1s spectra show three peaks (Figure 10) corresponding to the oxygen species contained in perovskite (Op) at 529.1–529.4 eV, in oxide or carbonate (O_o) at 531.4–531.6 eV and in hydroxyl groups at 533.8–534 eV [46,47,50]. The low-intensity peak of S 2p is also observed for the MA-2 and MA-4 samples prepared using TiO(OH)₂ (Figure 10 and Figure 11). The S 2p line is barely seen in the spectra of MA-2 and MA-4, which creates difficulties for quantitative analysis by XPS (Figure 10, S 2p line ). The S/O atomic ratio is ~0.02, so oxygen from sulfate ions is about 8% of the total O 1s line intensity, which is close to the XPS error. The binding energy of oxygen from strontium sulfate is 531.8 eV [51], which overlaps with the 531.4 eV peak (Figure 11). Experimental spectra of O 1s regions do not allow the addition of extra components in deconvolution that are in agreement with sulfur quantity. At the same time, experimental O 1s spectra of MA-2 and MA-4 do not contradict the presence of strontium sulfate.

The surface composition and concentration of elements calculated using the XPS spectra of the samples are presented in

Table 2. Atomic ratio of elements in the surface layers and binding energy of oxygen in the different states for the samples calcined at 900–1100 °C obtained from XPS data.

Sample	Sr/Ti	Op/Ti	Oo/Ti	Oh	Oo/Op	EOp	EOo	EOOH
MA-1(1100)	2.5	3.2	4.6	0.086	1.4	529.8	531.1	533.5
MA-2 1,2(1100)	2.1	3.1	3.3	0.1	1.1	529.2	531.4	533.8
MA-3(1100)	2.9	3.6	6.1	0.2	1.9	528.7	531.0	533.4
MA-4 1,3(1100)	1.9	3.2	3.2	0.1	1.0	529.1	531.4	533.8

¹—traces of sulfur; ^2,3—Sr/Ti~1.9 and 1.83 (Sr bulk content); O_o, O_p, O_h,—oxygen in Sr oxide, perovskite and hydroxyl groups, correspondingly.

. XPS analysis shows the presence of sulfur traces in MA-2 (SrCO₃ + TiO(OH)₂) and MA-4 (SrO + TiO(OH)₂) annealed at 1100 °C, which is due to its segregation from the initial TiO(OH)₂. The Sr/Ti ratio varies depending on the genesis of the samples. The enrichment of the surface layers with Sr compared with the bulk is observed for all samples, being more marked in the case of MA-1 (SrCO₃ + TiO₂) and MA-3 (SrO + TiO₂); the Sr/Ti ratio is equal to 2.5 and 2.9, correspondingly, at the stoichiometric value Sr/Ti = 2. The difference in the Sr/Ti value for MA-1 and MA-3 could result from the phase composition of the samples. Thus, MA-1 consists of the single-phase Sr₂TiO₄, while MA-3 is the multiphase system.

(Table 1) that can lead to an easier enrichment of the surface with Sr. For MA-2 and MA-4, the smaller value of the Sr/Ti ratio (2.1 and 1.9, respectively) results from some lower Sr content in the samples, as shown by the chemical analysis (Table 2). The values of O_p/Ti equal to 3.1–3.6 show that all samples comprise SrTiO₃ as the layers in Sr₂TiO₄ or as the single phase. The high values of the O_o/O_p ratio (1.4 and 1.9) for MA-1 and MA-3 correlate with the Sr enrichment of the surface, evidencing the segregation of SrO. In the case of MA-2 and MA-4, the strontium sulfate formation could be a reason for the smaller O_o/O_p values of 1–1.1 (Table 2).

2.2.3. Textural and Morphological Properties

The specific surface area (SSA) of the calcined samples obtained using the BET method is in the range of 1–3 m²/g (Table 1). The SSA of the samples annealed at 900 °C is higher for multiphase MA-1 and MA-3 samples, at which the highest SSA was shown by MA-1 (3 m²/g), including a large amount of SrCO₃. By increasing the annealing temperature from 900 to 1100 °C, the value of SSA decreases for all samples. The most marked decline of SSA is observed for MA-1 and MA-2 prepared from the MA products containing SrCO₃, which could be a result of its decomposition.

The morphology and the elemental mapping of the samples calcined at 1100 °C were characterized by FE-SEM with EDX. The SEM images show that the form and the size of the particles in the samples depend on their genesis (Figure 12). The typical micrographs of the single-phase MA-1 sample synthesized from the SrCO₃ + TiO₂ mixture demonstrate the presence of large, well-crystallized particles with a size of ~500 nm and plates with a thickness of ~50 nm (Figure 12a,b). The micrographs of all other samples show well-crystallized, three-dimensional particles with a size in the range of 200–1000 nm (Figure 12b–h). For the MA-1 sample, the ordered rows of the light spots of ~10 nm in size are clearly visible on the surface of the plates (Figure 12b), which could be related to the SrO nanospecies in accordance with previous HRTEM data for Sr₂Ti_0.9Mg_0.1O₄ [18]. In the case of MA-2 and MA-4 prepared from the mixtures containing TiO(OH)₂, slightly cambered, clear formations are observed. They could be attributed to strontium sulfate, which can be formed on the surface due to decomposition of the sulfate impurity from TiO(OH)₂, as XPS data show (Table 2). The elemental maps of Sr and Ti indicate their homogeneous distribution in MA-1 and MA-2 calcined at 1100 °C (Figure 13).

2.2.4. Catalytic Activity

The activity characteristics of the samples annealed at 1100 °C in methane oxidative coupling at 800–900 °C are presented in Figure 14. For all samples, methane and oxygen conversion increases, with the temperature rise being maximal at 900 °C, but their values are significantly higher in the case of MA-1 and MA-3, reaching up to 24–25.6% and 82–92%, correspondingly. The activity of the MA-2 and MA-4 catalysts is considerably lower (CH₄ and O₂ conversion of 16–17% and 54–63% at 900 °C, respectively). For the yield of sum C₂ (ethane + ethylene) Y_C2, the same trend is observed. Thus, its value of ~15–15.5% is observed for MA-1 and MA-3, while Y_C2 is only ~9% in the case of MA-2 and MA-4. For all the catalysts, excluding MA-2, the selectivity of C₂ hydrocarbons (S_C2) varies with the temperature within 50–63%, being maximal at 850 °C (Figure 14). In the case of MA-2, S_C2 increases with the temperature rise, reaching ~55% at 900 °C. The selectivity of C₂H₄ grows with the temperature for all catalysts but it is highest In the case of MA-1 and MA-3, reaching up 40 and 37%, respectively. Compared to MA-1 and MA-3, the C₂H₄ selectivity for MA-2 and MA-4 is noticeably lower at 800–850 °C and is close at 900 °C. The ratio of CO/CO₂ selectivities for MA-1 and MA-3 changes with the temperature rise conversely to C₂H₄ selectivity, while it tends to the increase in line with S_C2H4 for MA-2 and MA-4. Thus, on the whole, the MA-1 and MA-3 samples obtained from MA products of SrCO₃ + TiO₂ and SrO + TiO₂ mixtures, correspondingly, are substantially active compared with MA-2 and MA-4 prepared from the mixtures containing TiO(OH)₂. This difference could be attributed to the presence of sulfate traces in the surface layers of two latter catalysts (Table 2), blocking methane activation centers and thus reducing their activity. On the whole, the performance of the most active catalysts, MA-1 and MA-3, is comparable with the literature results (

Table 3. Some recent studies on the OCM performance of different titanates with a perovskite structure.

Catalyst	Reaction Conditions	Methane Conversion, %	C2 selectivity(S)/ Yield(Y),%	Ref.
CaTiO3 SrTiO3 BaTiO3	700 °C, CH4/O2 = 3, GHSV = 10,000 h−1	13 24 19	2/ 12/ 36/	[16]
SrTiO3 SrTi0.8Sn0.2 O3 SrTi0.8Nd0.2 O3	800 °C, CH4/O2 = 3, GHSV = 10,000 h−1	32.5 30.8 30.5	48.9/ 52.1/ 54.4/	[52]
SrZrO3	775 °C, CH4/O2 = 3, GHSV = 10,000 h−1	30.7	45.4/	[8]
Sr2TiO4	850 °C, CH4/O2 = 4, GHSV = 75,000 h−1	19.8	59.6/11.8	[19]
SrTiO3 Sr2 TiO4 Sr2 Ti0.9 Mg0.1O4	850 °C, CH4/O2 = 4, GHSV = 75,000 h−1	-	65.9/12.8 68.5/17.3 71.2/18.2	[18]
Sr2TiO4 (MA-1) Sr2TiO4 (MA-3)	850 °C, CH4/O2 = 4, GHSV = 75,000 h−1	24.1	62.7/15.1	This work

). Note that the analysis and comparison of the catalytic testing data is problematic due to different reaction conditions used. Thus, the most data were obtained at a GHSV of 10,000–18,000 h⁻¹, while our experiments were conducted at a higher GHSV of 75,000 h⁻¹.

OCM is mainly considered to be a heterogeneous–homogeneous process including activation of methane to methyl radicals on the catalyst surface; their subsequent coupling to ethane in the gas phase is then converted into ethylene or CO_x, depending on the peculiarities of the catalyst [1,2,3,4]. It is generally implied that formation of methyl radicals by abstraction of hydrogen and oxidative dehydrogenation of ethane to ethylene occur over the active sites if the oxygen species is available on the oxide catalyst surface. Meanwhile, the reagent conversion and product selectivity are affected by the type and concentration of the oxygen species present on the surface. XPS results for all catalysts studied revealed two main types of surface oxygen, corresponding to oxygen in SrO and perovskite SrTiO₃ (Table 2). The MA-1 and MA-3 samples demonstrating the most effective performance in OCM show the high ratio of these oxygen forms (O_o/O_p = 1.44 and 1.87) and surface Sr enrichment (Sr/Ti = 2.47 and 2.86, correspondingly). In so doing, the MA-1 sample exhibits some higher oxygen conversion, C₂ selectivity and yield (Figure 14). The presence of Sr cations in the surface layers of the Sr-Ti perovskites are considered to enhance methane conversion and promote C₂ selectivity, especially C₂H₄ [11,14,18]. Bai et al. investigated the effect of the different surface compositions caused by the surface reconstruction of SrTiO₃ on its OCM activity [14]. They revealed that in CH₄ conversion, S_C2 and S_C2H4 depend on the Sr surface concentration, and they reach maximal values at a certain Sr concentration. It may be assumed that the higher activity of MA-1 compared with MA-3 could be due to an optimal surface concentration of SrO (Sr/Ti = 2.47) and an O_o/O_p ratio equal to 1.44. This difference can result from the different structural and morphological peculiarities of the samples. MA-1 is the single-phase catalyst, while MA-3 is the multiphase one comprising a comparable quantity of SrTiO₃ and Sr₂TiO₄. Furthermore, the presence of the plate-like particles with ordered SrO nanospecies on the surface (Figure 12a,b) could contribute to the optimal activity of the surface oxygen and thus to the effective performance of the MA-1 catalyst in OCM. The activity of the multiphase MA-3 catalyst can be attributed to both the surface oxygen of SrO and, possibly, to more weakly bound active oxygen species arising out of the interface boundaries. The latter could be involved in CH₄ partial oxidation that leads to some higher CO selectivity and lower C₂ selectivity for MA-3 compared with the MA-4 catalyst (Figure 14).

The low activity of MA-2 and MA-4 synthesized from mixtures containing TiO(OH)₂ is correlated with the low values of the surface Sr concentration (Sr/Ti ~ 2) and O_o/O_p ~ 1 ratio (Table 2). This could be caused by blocking the active oxygen sites with strontium sulfate, which is observed on the surface of the catalysts, as SEM and XPS data shows (Figure 12d,k and Table 2). The methane conversion and sum C₂ yield for both MA-2 and MA-4 are close, while oxygen conversion and C₂ selectivity clearly differs, especially at temperatures below 900 °C: the lower oxygen conversion and the higher C₂ selectivity are observed for MA-4 compared with MA-2 (Figure 14). Such a difference at the close O_o/O_p ratio could be due to the lower surface Sr concentration in the case of MA-4, along with blocking the oxygen active sites to some larger extent compared to MA-2. Thus, the lower CO/CO₂ ratio for MA-2 could suggest the presence of a larger amount of surface oxygen species being active in the deep oxidation of CH₄ to CO₂.

The most active catalysts (MA-1 and MA-3) were tested for 10 h in OCM at 850 °C. The time dependence of the methane conversion and C₂ yield (Figure 15) shows the high long-term stability of the catalyst performance under the highly concentrated reaction mixture and short contact time.

3. Experimental

3.1. Sr₂TiO₄ Preparation

SrCO₃, SrO, TiO₂ (rutile) and TiO(OH)₂ prepared by hydrolysis of TiOSO₄ were used to prepare the Sr₂TiO₄ samples. SrO was obtained by calcination of Sr(NO₃)₂ at 900 °C just before the synthesis. To provide a target stoichiometry of samples, corresponding amounts of starting compounds were taken on the basis of their thermal analysis.

The stoichiometric mixtures of the starting chemicals were mixed and then activated in a high-power planetary APF-5 ball mill with two steel drums (25 cm³ volume) at a 800 rpm rotation rate and an acceleration of 40 g. Mechanochemical activation (MA) of powders was conducted for 10 min under the following conditions: air atmosphere, zirconium balls of 5 mm diameter and a powder-to-ball mass ratio of 1:10. Before each synthesis, a preliminary treatment of drums and balls with the corresponding mixture was performed to cover the surface of the drums and balls by a layer of the initial mixture to minimize contamination of the samples with Fe and Zr due to their rubbing during MA.

Activated mixtures were pressed in tablets and annealed at 900 and 1100 °C.

The composition of the starting mixtures, some characteristics of the samples and their abbreviations are presented in Table 1.

3.2. Catalysts Characterization

Thermal analysis of the starting compounds was carried out with a Q-1500D thermoanalyzer. The phase composition of the samples after MA and annealing were examined by using powder X-ray diffraction (XRD) and IR spectroscopy. XRD patterns were recorded in an X’TRA (Thermo ARL) diffractometer with Cu Ka (λ = 1.5418 Å) radiation in the 2θ angle range 10−70° with a step of 0.05° and an exposure time of 5 s at each step, and graphically processed with the Fityk program. Qualitative phase analysis was carried out by using PDF-2–ICDD files and the ICSD/retrieve database. The quantitative phase analysis was performed through the full pattern simulation method using PCW 2.4 software (http://powdercell-forwindows.software.informer.com/2.4, access date 16 February 2022). The callculated XRD patterns were obtained using the crystal structure database ICSD. To identify the phase structure of the samples, the lattice parameters of the prepared catalysts were obtained by the XRD Rietveld refinement method. FT-IR spectra were recorded in the range of 200–4000 cm⁻¹ using a BOMEM MB-102 FT-IR spectrometer.

The specific surface area (S_BET, m²/g) was determined by a routine BET procedure using the Ar thermal desorption data. The morphology and elemental mapping of the samples annealed at 1100 °C were studied using a Field Emission Scanning Electron Microscope (FE-SEM) Hitachi Regulus SU8230 equipped with an X-ray microanalysis.

The XPS experiments were performed with a SPECS (Germany) spectrometer equipped with a hemispherical PHOIBOS-150-MCD-9 analyzer. Non-monochromatic MgKα radiation (hν = 1253.6 eV) at 200W was used as the primary excitation. The spectrometer was calibrated using the Au4f_7/2 (84.0 eV), Ag3d_5/2 (368.3 eV) and Cu2p_3/2 (932.7 eV) peaks from metallic gold and copper foils [20]. The binding energies of detected peaks were calibrated by the position of the C1s peak (BE = 284.5 eV). The binding energy values and the areas of XPS peaks were determined after Shirley background subtraction and analysis of the line shapes. The ratios of the surface atomic concentrations of elements were calculated from the integral photoelectron peak intensities corrected by the corresponding relative atomic sensitivity factors based on the Scofield’s photo-ionization cross sections and transmission function of the analyzer.

3.3. Activity Tests

The catalytic activity in the OCM reaction was studied in a fixed-bed quartz tube reactor (5 mm inner diameter) at 750–900 °C and ambient pressure. The catalyst (0.2 mL) of 25–50 mesh was diluted with the quartz (1:3) to prevent a temperature gradient in the catalyst bed. The temperature of the catalyst was measured by a chromel–alumel thermocouple placed in the quartz well which was located at the middle of the catalyst bed. Methane was mixed with air so that the reaction mixture was CH₄:O₂:N₂ = 46:11.5:42.5% vol., CH₄:O₂ = 4. The total feed was 15 L/h corresponding to a gas hourly space velocity (GHSV) of 75,000 h⁻¹. Reactant and product concentrations were analyzed by on-line gas chromatography with Porapack Q (i.d. = 3 mm, l = 3 m) and CaX (i.d. = 3 mm, l = 2 m) columns using a thermal conductivity detector. The reaction products were ethane, ethylene, water, CO, hydrogen and CO₂. Water was removed from the probe with an SiO₂ trap. A blank run with the inert SiO₂ particles (0.25–0.5 mesh) showed no conversion in the reaction conditions.

The methane conversion (X_CH4), C₂ selectivity (S_C2) and C₂ yield (Y_C2) in this study were calculated using the standard normalization method [18], defined as the following equations:

$$X_i=\left(1-\frac{C_i}{C_i^{0 }}\Delta V\right), i={\mathrm{CH}}_4,{\mathrm{O}}_2,$$

$$S_j=\frac{2C_j}{2{\mathrm{C}}_2{\mathrm{H}}_6+2C_{{\mathrm{C}}_2{\mathrm{H}}_4}+2C_{\mathrm{C}\mathrm{O}}+2C_{\mathrm{C}\mathrm{O}2}}, j={\mathrm{C}}_2{\mathrm{H}}_6,{\mathrm{C}}_2{\mathrm{H}}_4,$$

$$Y_{C_j}=X_{\mathrm{C}{\mathrm{H}}_4}\times S_{C_j}\times 100\%,$$

$$S_k=\frac{2C_k}{2{\mathrm{C}}_2{\mathrm{H}}_6+2C_{{\mathrm{C}}_2{\mathrm{H}}_4}+2C_{\mathrm{C}\mathrm{O}}+2C_{\mathrm{C}{\mathrm{O}}_2}}, k=\mathrm{C}\mathrm{O},\mathrm{C}{\mathrm{O}}_2,$$

$$C-balance=\frac{2C_{{\mathrm{C}}_2{\mathrm{H}}_6}+2C_{{\mathrm{C}}_2{\mathrm{H}}_4}+2C_{\mathrm{C}\mathrm{O}}+2C_{\mathrm{C}{\mathrm{O}}_2}}{C_{\mathrm{C}{\mathrm{H}}_4}^0}, \Delta V=\frac{C_{N_2}^0}{C_{N_2}},$$

where ΔV is the molar flow rate change, $C_{N_2}^0, C_{N_2} \mathrm{is}$ the inlet and outlet concentration of N₂, C_i is a component mole fraction. The average carbon balance was not less than 98% in all the tests.

4. Conclusions

The samples of Sr₂TiO₄ were synthesized using mechanochemical activation (MA) of the four mixtures: (SrCO₃ + TiO₂), (SrCO₃ + TiO(OH)₂), (SrO + TiO₂) and (SrO + TiO(OH)₂) in a high-energy planetary ball mill, with subsequent annealing at 900 and 1100 °C. XRD and FT-IRS data show that MA for 10 min led to the disordering of the starting component structure in all mixtures and the formation of SrTiO₃ only in the case of SrO. These effects were more pronounced in the mixtures with amorphous TiO(OH)₂ containing a lot of OH groups.

The XRD Rietveld refinement method showed that, after annealing of the MA products at 900 °C, Sr₂TiO₄ was contained in all samples but its amount was considerably different in line with the peculiarities of MA products. Sr₂TiO₄ was mainly formed in the case of SrCO₃ or SrO and TiO(OH)₂, in contrast to the mixtures with TiO₂. The single-phase Sr₂TiO₄ is obtained after annealing at 1100 °C only from MA products containing SrCO₃.

FE-SEM data demonstrated that the single-phase sample synthesized from the SrCO₃+TiO₂ comprised of plates with a thickness of ~50 nm and large, well-crystallized particles with a size of ~500 nm, while all other samples contained only three-dimensional particles of 200–1000 nm. In addition, the presence of the SrO nanospecies of ~10 nm on the plate surface was assumed, while strontium sulfate was probably formed on the surface of the samples prepared using TiO(OH)₂ comprised of sulfate impurities.

The surface enrichment with Sr was observed by XPS data in all samples, being more marked in the case of the ones obtained from (SrCO₃ + TiO₂) and (SrO + TiO₂): the Sr/Ti ratio was equal to 2.47 and 2.86, correspondingly, at the stoichiometric value Sr/Ti = 2. The testing of the samples in OCM revealed that the CH₄ conversion and C₂ yield correlated with the surface Sr concentration and the ratio of oxygen amount in SrO and perovskite (O_o/O_p). The low activity of the catalysts obtained using TiO(OH)₂ could be related to the sulfate formation on the surface after annealing at 1100 °C due to blocking the active sites. The maximal CH₄ conversion and C₂ yield (25.6 and 15.5% at 900 °C, respectively) were observed for the sample obtained from (SrCO₃ + TiO₂), showing the specific surface morphology and optimal values of the surface Sr concentration (Sr/Ti = 2, 47) and a ratio of O_o/O_p = 1.44. The most active catalysts (MA-1 and MA-3) showed a high long-term stability performance under the highly concentrated reaction mixture and short contact time.