Study on the unsteady state oxidative coupling of methane: effects of oxygen species from O₂, surface lattice oxygen, and CO₂ on the C₂₊ selectivity

Abstract

This study examined the effects of oxygen species on the unsteady-state oxidative coupling of methane (OCM) using a lengthy catalyst bed of Na₂WO₄/Mn/SiO₂. The reaction conditions, including the methane-to-oxygen ratio, ratio of feed gas dilution by N₂, quantity of catalyst, and feed flow rate were adjusted for the continuous flow fixed bed reaction system. While the O₂ gas initiated methyl radical formation from methane, the surface lattice oxygen atoms improved the dehydrogenation of paraffins to olefins without significant activation of methane. The addition of CO₂ as a mild oxidizing agent was also tested and slightly improved OCM selectivity with slightly lower methane conversion were observed.

---

1. Introduction

The vast reserves of natural and shale gas have led to global interest in converting methane, their major component, to valuable chemicals. The oxidative coupling of methane (OCM) can directly produce ethylene using oxidizing agents including oxygen gas at a high reaction temperature (>600 °C), and hence it is a promising way to use the highly stable methane as feedstock. The proposed mechanism for OCM is composed of parallel and consecutive heterogeneous and gas-phase reactions.¹ The catalyst surface activated by the co-fed oxygen produces methyl radicals by hydrogen abstraction from methane. The methyl radicals can then be converted to the desired ethane in the gas phase, or further oxidized to the undesired CO and CO₂. It was suggested that an industrially feasible OCM process can be developed if the C₂₊ products (hydrocarbons containing two or more carbon atoms) achieve a yield above 25% and a selectivity of 75–80%.^2–9

Because there is a critical technological barrier for the selective formation of olefins and paraffins while suppressing the deep oxidation of methane to CO and CO₂,^6,8,10 many selective oxidation catalysts have been prepared for the OCM reaction. For unpromoted oxides of alkali, alkaline-earth, and rare earth metals, low CH₄ conversion and high C₂₊ selectivities were observed, along with lower C₂₊ yields (<12%),¹¹ although moderate OCM activity was achieved at lower reaction temperatures (<750 °C).¹² Perovskites, which are well-defined crystalline mixed oxides containing two different metals, were also used for catalyzing OCM,^13–15 exhibiting improved activity with promoter addition.^1,12 Among them, Na₂WO₄/Mn/SiO₂ is one of the most effective and stable OCM catalysts.^9,16–21 An optimized catalyst, composed of 2 wt% Mn and 5 wt% Na₂WO₄ on SiO₂, has exhibited ∼25% C₂₊ yield.⁶ The remarkable long-term stability of Na₂WO₄/Mn/SiO₂ (1000 h) was also reported for continuous-flow reactions.²¹ Although the origin of its high catalytic activity has not yet been elucidated,²² the unusual formation of α-cristobalite SiO₂ phase during calcination at 750–850 °C has been reported as a distinguishing feature of a good catalyst,²³ and the active species has been suggested to be Na⁺ or Na₂WO₄ on silica surface with distorted WO₄²⁻ tetrahedra formed during phase transition.^23–25 In addition to the roles played by Na₂WO₄, manganese oxide was reported to improve the oxygen mobility via reduction of Mn³⁺ to Mn²⁺.²⁶ Because of the proposed oxygen activation by Mn in the Na–O–Mn composites,²⁷ the selective formation of ethylene, which can occur by oxidative or non-oxidative dehydrogenation of ethane, can be controlled by tuning the surface Mn concentration.^24,28 Improved OCM activity by the addition of several additives to Na₂WO₄/Mn/SiO₂ has also been reported.^9,29

Although a number of selective catalysts have been developed, a better understanding of the reaction conditions is also required to achieve a feasible OCM process. Because of the limited supply of O₂ and the high CH₄ conversion through the reactor, in which a reactant mixture flows downward through the catalyst bed, the conditions including reactant compositions in the upper catalyst bed may be significantly different from those in the lower catalyst bed. When methane and oxidant are flown through the oxygen-rich top bed, the oxidant (mostly O₂) can be adsorbed on the catalyst surface. The adsorbed oxygen species activates methane by abstracting H to form methyl radicals, which are coupled to form ethane. In the oxygen-deficient bottom layer, methyl radicals, C₂₊ products, and carbon oxides are present and can interact with the unreacted methane and oxidant. The unreacted O₂ can activate methane by either forming methyl radical or deeply oxidize the highly active C₂₊ products to CO and CO₂, thereby decreasing the C₂₊ selectivity. The amount of initially supplied oxidants also influences the level of deep oxidation products in the bottom layer.

By using less-active oxidizing agents, including H₂O and CO₂, the selectivity of OCM products can be improved by suppressing the burning of valuable hydrocarbons to form CO and CO₂.³⁰ Water as an oxidant improved the coupling selectivity over perovskite catalysts, exhibiting increased C₂ production with increasing partial pressure of steam.³¹ CO₂ has also been suggested as an alternative oxidant to O₂.^32–37 The formation of surface lattice oxygen atoms from CO₂ has been observed on PbO–MgO and PbO–CaO catalysts, which improves the abstraction of hydrogen atoms from methane.^32,35 In Na₂WO₄/Mn/SiO₂, the presence of CO₂ caused structural modification to improve catalytic activity.³⁸ The formation of hot spots, which accelerate the deep oxidation to CO and CO₂ because of significant local exothermic reaction, was observed in the catalyst bed of a bench-scale OCM reactor,³⁹ and these hot spots can be reduced by the presence of CO₂ with high heat capacity.⁴⁰ In addition to the improved OCM activity, the presence of CO₂ stabilized the lithium-doped MgO catalyst and suppressed its deactivation.³⁶ For oxidative dehydrogenation, it was suggested that CO₂ improves selective OCM and stabilizes the catalysts by poisoning the nonselective active sites.⁴¹

The goal of this study is to understand the local OCM reaction throughout the catalyst bed under continuously changing reaction conditions, which can be attributed to the limited supply of oxidizing reagents (particularly O₂) in the fixed bed system. The temperature (750–850 °C), N₂ dilution ratio, CH₄/O₂ ratio, catalyst bed thickness, total feed flow rate, and the co-feeding rate of CO₂ were varied to study the behaviors of reactants and products under these continuously changing conditions. The effects of O₂ and CO₂ (as strong and mild oxidizing reagent, respectively) on the local OCM reaction were investigated in a long fixed bed system, which is meant to simulate possible scaled-up reactors. CO₂ as a mild oxidizing reagent can also function as a diluent in the OCM process, and it can be recycled with the unreacted methane in the industrial scale process. With continuous consumption of O₂ throughout the long catalyst bed and the formation of O₂-deficient region, the roles of lattice oxygen atoms on the OCM were also observed. Although a steady state was not achieved while the lattice oxygen atoms were being consumed for the OCM, the unsteady state reaction examined here is useful for understanding the OCM reaction using lattice oxygen atoms. The spent catalysts were characterized using X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and Raman spectroscopy to reveal changes in the catalysts during the OCM reaction.

2. Experimental

2.1. Materials

All chemicals were used without further purification. Sodium tungstate dihydrate (Na₂WO₄·2H₂O, 99+%) was purchased from Yakuri Pure Chemicals (Kyoto, Japan). Manganese nitrate hexahydrate (Mn(NO₃)₂·6H₂O, 98%) was purchased from Kanto Chemical (Tokyo, Japan). Silica gel (SiO₂, 60–325 mesh, 99.5%) was purchased from Johnson Matthey (Royston, United Kingdom). Manganese(II) tungsten oxide (MnWO₄, 99.9%) was purchased from Alfa Aesar (Haverhill, Massachusetts, USA). Manganese(III) oxide (Mn₂O₃, 99.9%) was purchased from Sigma-Aldrich (St. Louis, Missouri, USA). Methane (CH₄, 99.95%), oxygen (O₂, 99.995%), nitrogen (N₂, 99.9%), and carbon dioxide (CO₂, 99.99%) were purchased from Shinyang Sanso (Seoul, Korea). DI water (18.2 MΩ m) was prepared using an aquaMAX-Ultra 370 series water purification system (Young Lin Instruments, Anyang, Korea).

2.2. Catalyst preparation

The catalyst containing 5 wt% Na₂WO₄ and 2 wt% Mn on SiO₂ (Na₂WO₄/Mn/SiO₂) was prepared using a mixed slurry method.^39,42 Silica gel (28 g) was mixed with DI water (150 mL) and stirred at 105 °C for 1 h under reflux. Mn(NO₃)₂·6H₂O (3.146 g) dissolved in DI water (10 mL) and Na₂WO₄·2H₂O (1.684 g) dissolved in DI water (10 mL) were added drop-wise to the boiling mixture using a syringe pump over 15 min. The prepared thick paste was dried in air at 105 °C for 16 h and then calcined in air at 800 °C for 5 h. Prior to the reaction, the prepared catalyst powder was passed through 70–140 mesh.

2.3. Catalytic activity measurement

The catalytic reaction was performed using a fixed bed cylindrical quartz reactor (ID = 6 mm, length = 370 mm) (Fig. S1†). The unreacted O₂ leaving the catalyst bed can contribute to further homogeneous oxidation of the C₂₊ products to CO and CO₂ before the product mixture was inactivated by cooling,³⁸ which hinders interpretation of the reaction results. To reduce homogeneous reaction of the product mixture, a narrow exit was introduced on the reactor to encourage its fast outflow. The outlet of the reactor was narrower than the inlet (1 vs. 6 mm). The catalyst bed was located at the center of the reactor, with quartz wool placed at its upper and lower ends. ZrSiO₄ beads (Cenotec, Co. Ltd.) were positioned on the top of the catalyst bed to minimize the homogeneous reactions. A reaction mixture of CH₄, O₂, N₂, and CO₂ flowed into the catalyst bed. Notably, the furnace temperature, which was used as the reaction temperature in this study, was controlled (Fig. S2†). The measured temperature of catalyst bed was 0–40 °C higher, depending on the position in the catalyst bed and the reaction conditions. The temperature measured at the bottom of catalyst bed was closest to the furnace temperature. Because the change of temperature though the catalyst bed was similar at each reaction temperature, we used the furnace temperature as the reaction temperature for convenience. Gas hourly space velocity (GHSV) was adjusted using catalyst bed volume and reactant flow rate. Water formed during the OCM reaction was removed at a cold trap (−2 °C) connected to a chiller (VTRC-620, Jeio Tech). After an hour of isothermal operation, the product mixture was collected and analyzed using a thermal conductivity detector (TCD) and a flame ignition detector (FID) installed onto an online-gas chromatography system (Agilent 7890A). CH₄, O₂, N₂, CO, and CO₂ were analyzed using the TCD with a Carboxen 1000 column (Supelco, 60–80 mesh, 4.6 m × 1/8 in × 2.1 mm), and the hydrocarbons were analyzed using the FID with an HP-PLOT Al₂O₃ S capillary column (50 m × 320 μm × 8 μm). The catalytic performance was characterized by the following parameters:

O₂-consumption for product i, the fraction of O₂ in the feed required to convert CH₄ into each product component i, was obtained using the following relationship described in Table 1:

Table 1 Reactions during the OCM process

Reaction

CO production: CH₄ + [/]O₂ → CO + 2H₂O

CO₂ production: CH₄ + 2O₂ → CO₂ + 2H₂O

C₂H₆ production: CH₄ + ¼O₂ → ½C₂H₆ + ½H₂O

C₂H₄ production: CH₄ + ½O₂ → ½C₂H₄ + H₂O

C₃H₈ production: CH₄ + ⅓O₂ → ⅓C₃H₈ + ⅔H₂O

C₃H₆ production: CH₄ + ½O₂ → ⅓C₃H₆ + H₂O

---

2.4. Catalyst characterization

XRD results were collected to observe the catalysts' crystal structures using a Shimadzu XRD-6000 device equipped with a CuKα₁ (λ = 0.15406 nm) source. Raman spectroscopy analysis was performed using a Renishaw InVia Raman microscope. The surface electronic structures were analyzed using a PHI 5000 Versa Probe X-ray photoelectron spectroscope at Korea Basic Science Institute (Busan, Korea). The XPS spectra were calibrated using C 1s peak at 284.6 eV.

3. Results and discussion

3.1. Effects of CH₄/O₂ ratio and N₂ dilution on catalysis

Before investigating the effects of reaction conditions on the OCM results, the effects of O₂ concentration or the CH₄/O₂ ratio on the OCM activity were studied. CH₄/O₂ = 3 mol/mol was selected as an optimal reaction condition in this study because of the high C₂₊ selectivity with moderate CH₄ conversion, although the highest C₂₊ yield was not achieved using this ratio (Fig. 1 and Table S1†). Notably, in this study, C₂₊ compounds include ethane, ethylene, propane, and propylene. When CH₄/O₂ was adjusted from 1 to 6 mol/mol at 800 °C, the highest C₂₊ yield of 21.5% (44.0% CH₄ conversion and 49.0% C₂₊ selectivity) was achieved at 800 °C when CH₄/O₂ = 2 mol/mol. The C₂₊ selectivity sharply increased with increasing CH₄/O₂ ratio up to 3 mol/mol, and then steadily increased. Based on these observations, we selected CH₄/O₂ = 3 mol/mol for the good CH₄ conversion and high C₂₊ selectivity. The observed increase in C₂₊ selectivity and decrease in CH₄ conversion at a higher CH₄/O₂ ratio (i.e., reducing the limited supply of oxygen reactant) were consistent with the results of previous studies.⁴³ At higher CH₄/O₂ ratios, the amount of O₂ was insufficient for activating the large number of methane molecules, thus the CH₄ conversion decreased. At lower CH₄/O₂ ratios, the formation of methyl radicals was improved by the high partial pressure of O₂ to achieve high CH₄ conversion, but the methyl radicals were also easily oxidized to CO and CO₂ to decrease the C₂₊ yield (Table S1†). These observations suggest that a moderate CH₄/O₂ ratio is required to achieve a high C₂₊ yield by combining the CH₄ conversion and C₂₊ selectivity.

Fig. 1 CH₄ conversion (circles), C₂₊ selectivity (triangles), and C₂₊ yield (rectangles) over Na₂WO₄/Mn/SiO₂ catalyst at 800 °C and different CH₄/O₂ (mol/mol) ratios. The O₂ conversions were 100% for all cases. GHSV = 10 000 h⁻¹. CH₄ flow rate was fixed at 18 mL min⁻¹, and the flow rates of O₂ and N₂ were adjusted to achieve the total flow rate of 50 mL min⁻¹.

With the CH₄/O₂ ratio fixed at 3 mol/mol, an optimized N₂ dilution ratio of 20% (v/v) or CH₄/O₂/N₂ = 3/1/1 mol/mol/mol produced the best CH₄ conversion and C₂₊ selectivity (Fig. 2 and Table S2†). High concentrations of methane and O₂ can accelerate the reaction rates, and the deep oxidation of methane to CO and CO₂ may occur along with the formation of hot spots because of the large heat of reaction.³⁹ Diluting the reactant mixture decreases the partial pressures of methane and O₂ and thus improves the selectivity for coupled C₂₊ hydrocarbons while suppressing the formation of CO and CO₂.^39,44,45 This change in the OCM results was more significant at the higher temperatures of 800–850 °C (with CH₄ conversions > 20%) and less significant at 750 °C (with CH₄ conversions < 20%). When the feed was highly diluted at the low reaction temperature, the formed methyl radicals may not be able to undergo gas-phase coupling with other methyl radicals in the highly diluted environment, which may reduce the C₂₊ selectivity. The general trend of increasing C₂₊ selectivity with increasing CH₄ conversion, under different reaction conditions for the same catalysts, has also been observed in our previous studies.^46–48

Fig. 2 CH₄ conversion (circles), C₂₊ selectivity (triangles), and O₂ conversion (squares) using different contents of N₂ diluent at (a) 750 °C, (b) 800 °C, and (c) 850 °C. CH₄/O₂ ratio = 3 mol/mol, GHSV = 10 000 h⁻¹.

The ratios of CO/CO₂ and olefins/paraffins increased at lower dilution by N₂ or higher partial pressures of the reactants (Fig. 3, S3, and Table S2†). Because higher partial pressures of the reactants are expected to improve the formation of CO₂ and increase the olefins/paraffins ratio,⁴⁴ the improved formation of CO rather than CO₂ we observed must come from another reaction pathway, which will be discussed later in the Section 3.2. Note that the amount of O₂ required for forming CO₂ is 1.33 times larger than that for forming CO, and the production of olefins from methane requires more than twice the amount of O₂ required to produce paraffins. Because of the limited supply of O₂ in this study, the deep oxidation to CO and CO₂ consumed more O₂ and suppressed the methane activation and the dehydrogenation, and this lowered the olefins/paraffins ratio.

Fig. 3 Ratios of olefins/paraffins and CO/CO₂ for the dehydrogenation of paraffins and the oxidation of CO, depending on the content of N₂ diluent at 800 °C.

3.2. Effects of catalyst bed thickness and feed flow rate on catalysis

Using the fixed CH₄/O₂ ratio and N₂ dilution, the effects of oxygen species on the OCM activity were investigated. The oxygen species activating methane and other hydrocarbons change throughout the long catalyst bed at a fixed flow rate. Both O₂ gas and surface oxygen species contribute to OCM, while only the latter are present at the O₂-deficient bottom part of the bed. With increasing catalyst bed thickness and the same flow rates of reactants, along with increasing residence time and decreasing GHSV, the reaction results did not significantly change when the O₂ was completely converted (Fig. 4a, c, S4, Tables S3 and S4†). While incomplete conversion of O₂ was observed at lower GHSV and lower reaction temperature, the 100% O₂ conversion produced almost the same CH₄ conversion (32.1–35.9%) along with 50.9–59.8% C₂₊ selectivity and 16.8–20.5% C₂₊ yield. The negligible C₂₊ production and CH₄ conversion in the O₂-deficient region (when 100% O₂ conversion was reached) indicated that the O₂ gas was required for significant conversion of methane to the C₂₊ compounds. Compared to the smaller changes in CH₄ conversion and C₂₊ selectivity, the olefin selectivity increased with increasing catalyst bed thickness, indicating that the olefins were formed by the oxidative dehydrogenation of paraffins at the O₂-deficient region. Based on these observations, the oxidative dehydrogenation of paraffins to olefins can be initiated by the surface lattice oxygen atoms. In contrast, the production of methyl radicals that contributed to CH₄ conversion was less initiated by these surface lattice oxygen atoms because there was no significant increase of CH₄ conversion at the O₂-deficient region. Therefore, the surface lattice oxygen atoms, which were formed by diffusion of bulk lattice oxygen atoms to the catalyst surface, mainly dehydrogenated the paraffins to olefins. Meanwhile, the surface-adsorbed oxygen atoms from the bulk O₂ gas preferentially converted methane to methyl radicals, which were further coupled to produce paraffins. In addition to the formation of olefins, the production of CO₂ also slightly increased with increasing catalyst bed thickness, indicating that dehydrogenation using surface lattice oxygen species was accompanied by deep oxidation to form CO₂ as by-product.

Fig. 4 (a and b) Catalytic activities and (c and d) ratios of olefins/paraffins and CO/CO₂ for the dehydrogenation of paraffins and the oxidation of CO, depending on the catalyst bed thickness (a and c) and feed flow rate (b and d) at 800 °C.

The OCM was also studied at a fixed catalyst bed thickness (0.18 mL or 6.4 mm) and adjustable feed flow rates (15–90 mL min⁻¹) to confirm the role of each oxygen species at CH₄/O₂/N₂ = 3/1/1 mol/mol/mol (Fig. 4b, d, S5, Tables S5 and S6†). When 100% O₂ conversion was achieved, the CH₄ conversion (28.5–35.5%) and C₂₊ selectivity (56.5–64.2%) were not significantly adjusted at 800–850 °C, indicating no significant dependence of OCM activity on the feed flow rate.

While the dehydrogenation of paraffins to olefins occurred at the O₂-deficient region using surface lattice oxygen species, as depicted in Fig. 4 and Table S3,† decreasing the feed flow rate, which has the same effect as increasing the catalyst bed thickness at a the fixed flow rate, did not significantly affect the dehydrogenation activity (Table S5†). A turbulent flow in the catalyst bed may suppress formation of the O₂-deficient region, obscuring the effects of surface lattice oxygen species.

In addition to dehydrogenation, the preferred formation of CO₂ over CO was observed, which indicates deeper oxidation to CO₂ at a lower GHSV, lower feed flow rate, or longer residence time. Particularly at 850 °C, the CO/CO₂ ratio significantly decreased at lower feed flow rate or longer residence time (Table S5†). The improved formation of CO₂, which consumed a large quantity of oxygen species, also reduced the CH₄ conversion because of the lack of oxidizing reagents. The preferred formation of CO at a shorter residence time can be attributed to the gas-phase formation of CO from methyl radicals.^5,49 CO can be preferentially produced when a methyl radical reacts with O₂ instead of another methyl radical, increasing the CO/CO₂ ratio.

The reactions preferred by the adsorbed O₂ and the surface lattice oxygen species are depicted in Fig. 5. Methane molecules were activated by the surface-adsorbed O₂ to form methyl radicals, which were then coupled to ethane. Ethane molecules were oxidatively dehydrogenated by the surface lattice oxygen species to form ethylene. The activation of methane molecules to methyl radicals by the surface lattice oxygen species was negligible; however, the dehydrogenation of paraffins and deep oxidation of hydrocarbons can be preferred by the surface lattice oxygen species.

Fig. 5 Reactions by oxygen species.

3.3. Effects of the catalyst bed thickness on the catalysis with co-fed CO₂

Based on the improved formation of CO₂ at lower feed flow rates, we studied the effects of a mild oxidant such as CO₂ on the OCM selectivity (or the product distributions of paraffins, olefins, CO, and CO₂). As one of the products of OCM, CO₂ can act as a diluent or a mild oxidant replacing the diluent N₂ and the oxidant O₂.^36,38,50 As a mild oxidant, a moderate concentration of CO₂ has been reported to improve the C₂₊ selectivity and reduce the CH₄ conversion.^38,51,52 For example, MgO or Sm₂O₃-based catalysts exhibited increased C₂₊ selectivity with a low partial pressure of CO₂.^32,53 Suppressed OCM activity with CO₂ has also been reported: Li/MgO was poisoned by CO₂,⁵⁴ and Na₂WO₄/Mn/SiO₂ was deactivated by concentrated CO₂.³⁸ Nevertheless, it may be possible to modify the OCM activity by using an appropriate amount of co-fed CO₂ as mild oxidant to replace the mixture of N₂ and O₂. Thus, the OCM reaction was tested at a fixed feed flow rate, variable catalyst bed thickness, and the 20% N₂ in the feed was replaced by 16% CO₂ + 4% N₂ in order to investigate this possibility and understand the effects of the adsorbed oxygen atoms (Fig. 6, S6, Tables S7 and S8†).

Fig. 6 (a) Catalytic activities, (b) ratios of olefins/paraffins and CO/CO₂, and (c) oxygen consumption for the dehydrogenation of ethane (or propane) and the oxidation of CO, depending on the catalyst bed thickness at CH₃/O₂/N₂/CO₂ = 3/1/0.2/0.8 mol/mol/mol/mol and 800 °C.

When increasing the catalyst bed thickness (or decreasing the GHSV), CH₄ conversion and C₂₊ selectivity reached 30.8–34.5% and 51.1–61.9%, respectively at 100% O₂ conversion (Table S5†), which is similar to the results using N₂ as diluent (32.1–35.9% and 50.9–59.8%, respectively, Table S3†). Compared with the use of N₂ diluent, the co-feeding of CO₂ slightly lowered CH₄ conversion and slightly improved C₂₊ selectivity by 1.5–4.7%, achieving almost identical C₂₊ yields. The improved C₂₊ selectivity was more distinct when the O₂ was not completely consumed (O₂ conversion < 100%), particularly when GHSV >20 000 h⁻¹ at 750 °C.

In terms of the selectivity between CO and CO₂, a larger CO/CO₂ ratio was observed when CO₂ was co-fed compared to using the N₂ diluent alone (Tables S3 and S7†). The improved formation of CO with CO₂ co-feeding can be attributed to the catalytic conversion of CO₂ to CO. While this conversion can occur by either the reduction or catalytic dissociation of CO₂, a highly oxidizing environment can suppress the reduction reaction by hydrogen generated in the dehydrogenation of paraffin or reduction of water. The catalytic dissociation of CO₂ to CO and O may occur to increase the concentration of CO,^50,51 and the produced oxygen atoms may improve the C₂₊ selectivity.

The actual temperatures of the top, middle, and bottom layers of the 35.3 mm catalyst bed were measured (Fig. S2†). The temperature was observed to decrease from the top to the bottom layer, indicating that the involved reactions occurred at the entrance of the catalyst bed.

We also summarized all the reaction results obtained in this study in terms of C₂₊ yield, depending on CH₄ conversion and O₂ conversion (Fig. S7†). Regardless of reaction conditions, CH₄ conversion, O₂ conversion (for <100%), and C₂₊ yield increased proportionally with each other, indicating that the formation of C₂₊ compounds was not significantly adjusted by varying the reaction conditions, as long as O₂ was present. Thus, the effects of surface lattice oxygen species may not be significant while O₂ is present.

3.4. Identification of crystal structures

To understand the effects of catalyst structures on the OCM activity, the spent catalysts were characterized (Fig. 7). There was no significant difference between catalysts spent using N₂ as diluent alone and with CO₂ co-feeding. However, when using a longer catalyst bed or smaller GHSV, catalysts on the top and bottom of the bed exhibited different structures. The formation of MnWO₄ was distinct at the bottom (O₂-deficient region) but not clearly observed at the top (O₂-rich region). While MnWO₄ had been reported in the literature as either an active or inactive component,^{22,28,47,55,56} we observed moderate OCM activity on MnWO₄ powder (Fig. S9†). While the co-existing Na₂WO₄ and α-cristobalite SiO₂ are known as key components for good OCM catalysts,^22,23,57 it is not clear whether the formation of MnWO₄ reduced the OCM activity or not in this study.

Fig. 7 XRD results of Na₂WO₄/Mn/SiO₂ spent in the reaction with N₂ diluent from the (a) top and (b) bottom of the catalyst bed, and in the reaction with co-fed CO₂ from the (c) top and (d) bottom of the catalyst bed. The peaks were assigned to α-cristobalite SiO₂ (○), Na₂WO₄ (◆), MnWO₄ (▽), and Mn₂O₃ (★).

When heated to 800 °C in air, the Na₂WO₄ and Mn₂O₃ phases were found to disappear for the fresh catalyst at room temperature, indicating the melting of these solid materials (Fig. S8(a)†). Compared to these, MnWO₄ survived, while maintaining its solid structure at the OCM reaction temperature. α-Cristobalite SiO₂ exhibited a peak shift toward smaller 2θ or larger d-spacings, indicating thermal expansion of the crystal structure.

3.5. Study of catalyst surface by Raman spectroscopy and XPS

Raman spectra of the spent catalysts also confirmed the formation of MnWO₄ in the O₂-deficient region (bottom layer of catalyst bed) by the distinct MnWO₄ peak at 885 cm⁻¹ for both CO₂-free and CO₂-added feeds (Fig. 8). Raman bands indicating Na₂WO₄ were observed in the fresh catalyst and spent ones at all regions of the bed. Notably, the bands at or below 500 cm⁻¹ can be attributed to α-cristobalite SiO₂, and those at 500–700 cm⁻¹ can be attributed to manganese oxides.²⁴ When heated to 800 °C in air, Mn₂O₃ and α-cristobalite SiO₂ peaks became weaker than those of the fresh catalyst, indicating that these phases became weaker or less ordered at high temperatures (Fig. S8(b)†).

Fig. 8 Raman spectra of Na₂WO₄/Mn/SiO₂ catalyst: (a) in the fresh state, after the reaction with N₂ diluent at the (b) top and (c) bottom of the catalyst bed, after the reaction with co-fed CO₂ at the (d) top and (e) bottom of the catalyst bed.

In terms of surface electronic structures, the spent catalysts exhibited more intense Na 1s and Mn 2p peaks when only N₂ was used as diluent in the reaction (Table S9†). The O 1s peaks at 530.9 eV indicating the presence of Na₂WO₄ were stronger than those with co-fed CO₂ (Fig. S10†). The peaks at 532.6 eV are assigned to SiO₂. Compared to the fresh catalyst, the spent catalysts after reaction with N₂ as the only diluent or with co-fed CO₂ had much higher surface concentrations of Mn as measured by XPS (Table S9†). This growing surface concentration of Mn can be attributed to the formation of MnWO₄, even though the distinct peaks could not be identified because the Mn may have protruded to the surface by complexation with W species.⁴⁷

The spent catalyst from the O₂-deficient region (which exhibited the formation of MnWO₄) was re-used for the OCM, and the activity was compared with the fresh one (Table 2). The spent catalyst exhibited higher C₂₊ selectivity, higher C₂₊ yield, and higher olefins/paraffins ratio by 3.7–9.4%, 4.9–10.4%, and 0–7.6%, respectively. Therefore, the spent catalysts containing MnWO₄ produced more C₂₊ compounds along with a higher dehydrogenation activity. These observations confirmed that dehydrogenation occurred preferably in the O₂-deficient region, which exhibited the formation of MnWO₄.

Table 2 OCM results using spent catalyst prepared in the O₂-deficient region (bottom of catalyst bed)^a,b

Reaction temperature (°C)	X(CH4) (%)	X(O2) (%)	S(C2+) (%)	Y(C2+) (%)	Olefins/paraffins (mol mol^−1)
a GHSV = 10 000 h^−1, CH4/O2/N2 = 3/1/1 mol/mol/mol, total feed flow rate = 30 mL min^−1.b X(CH4): CH4 conversion, X(O2): O2 conversion, S(C2+): C2+ selectivity, Y(C2+): C2+ yield.
Fresh Na2WO4/Mn/SiO2
750	35.2	100	54.5	19.2	2.12
800	35.3	100	51.5	18.2	2.49
850	33.4	100	45.9	15.4	3.77
Spent Na2WO4/Mn/SiO2 prepared in the O2-deficient region
750	35.0	98.2	59.6	20.8	2.11
800	35.8	100	53.4	19.1	2.68
850	34.1	100	49.9	17.0	3.98

---

4. Conclusions

The catalytic oxidative coupling of methane to C₂₊ compounds was examined under various reaction conditions, and the following results were obtained.

(i) Oxygen species formed by the adsorption of gas-phase O₂ preferentially activated methane molecules to produce methyl radicals, while those formed by the diffusion of bulk lattice oxygen atoms preferred the dehydrogenation of paraffins to olefins.

(ii) The co-feeding of CO₂ slightly increased the C₂₊ selectivity compared to the reaction without CO₂ co-feeding; however, no significant improvement was observed exhibiting almost identical C₂₊ yield, although beneficial effects have been reported.

(iii) The formation of MnWO₄ on the catalyst was improved in the O₂-free environment at the O₂-deficient bottom of the catalyst bed, which exhibited the preferred dehydrogenation reaction. This also confirmed the diffusion of bulk lattice oxygen atoms to the catalyst surface.

In this study, we focused on the OCM reaction in the non-steady states. The roles of surface lattice oxygen atoms and the reaction-initiated changes in the catalyst bed were clearly observed. Those results will help researchers and engineers understand the scaled-up OCM processes for the conversion of methane to paraffins and olefins.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was supported by C1 Gas Refinery Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (2015M3D3A1A01064900).

Notes and references

1. S. Lim, J.-W. Choi, D. J. Suh, K. H. Song, H. C. Ham and J.-M. Ha, J. Catal., 2019, 375, 478–492.
2. B. L. Farrell, V. O. Igenegbai and S. Linic, ACS Catal., 2016, 6, 4340–4346.
3. J. C. W. Kuo, C. T. Kresge and R. E. Palermo, Catal. Today, 1989, 4, 463–470.
4. A. M. Maitra, Appl. Catal., A, 1993, 104, 11–59.
5. J. H. Lunsford, Angew. Chem., Int. Ed., 1995, 34, 970–980.
6. U. Zavyalova, M. Holena, R. Schlogl and M. Baerns, ChemCatChem, 2011, 3, 1935–1947.
7. Y. S. Su, J. Y. Ying and W. H. Green, J. Catal., 2003, 218, 321–333.
8. A. Cruellas, J. J. Bakker, M. V. Annaland, J. A. Medrano and F. Gallucci, Energy Convers. Manage., 2019, 198, 111789.
9. N. S. Hayek, G. J. Khlief, F. Horani and O. M. Gazit, J. Catal., 2019, 376, 25–31.
10. S. Arndt, G. Laugel, S. Levchenko, R. Horn, M. Baerns, M. Scheffler, R. Schlögl and R. Schomäcker, Catal. Rev., 2011, 53, 424–514.
11. J. M. DeBoy and R. F. Hicks, Ind. Eng. Chem. Res., 1988, 27, 1577–1582.
12. S. Lim, J.-W. Choi, D. Jin Suh, U. Lee, K. H. Song and J.-M. Ha, Catal. Today, 2020, 352, 127–133.
13. G. Lee, I. Kim, I. Yang, J.-M. Ha, H. B. Na and J. C. Jung, Appl. Surf. Sci., 2018, 429, 55–61.
14. D. Kwon, I. Yang, Y. Sim, J.-M. Ha and J. C. Jung, Catal. Commun., 2019, 128, 105702.
15. Y. Sim, J. Yoo, J.-M. Ha and J. C. Jung, J. Energy Chem., 2019, 35, 1–8.
16. S.-B. Li, Chin. J. Chem., 2001, 19, 16–21.
17. U. Simon, O. Gorke, A. Berthold, S. Arndt, R. Schomacker and H. Schubert, Chem. Eng. J., 2011, 168, 1352–1359.
18. S. Pak and J. H. Lunsford, Appl. Catal., A, 1998, 168, 131–137.
19. H. R. Godini, A. Gili, O. Gorke, U. Simon, K. Hou and G. Wozny, Energy Fuels, 2014, 28, 877–890.
20. V. Fleischer, R. Steuer, S. Parishan and R. Schomacker, J. Catal., 2016, 341, 91–103.
21. S. Li, J. Nat. Gas Chem., 2003, 12, 1–9.
22. S. Arndt, T. Otremba, U. Simon, M. Yildiz, H. Schubert and R. Schomäcker, Appl. Catal., A, 2012, 425–426, 53–61.
23. A. Palermo, J. P. Holgado Vazquez, A. F. Lee, M. S. Tikhov and R. M. Lambert, J. Catal., 1998, 177, 259–266.
24. S.-F. Ji, T.-C. Xiao, S.-B. Li, C.-Z. Xu, R.-L. Hou, K. S. Coleman and M. L. H. Green, Appl. Catal., A, 2002, 225, 271–284.
25. J. G. Wu and S. B. Li, J. Phys. Chem., 1995, 99, 4566–4568.
26. Y. Kou, B. Zhang, J.-z. Niu, S.-b. Li, H.-l. Wang, T. Tanaka and S. Yoshida, J. Catal., 1998, 173, 399–408.
27. D. J. Wang, M. P. Rosynek and J. H. Lunsford, J. Catal., 1995, 155, 390–402.
28. S. Ji, T. Xiao, S. Li, L. Chou, B. Zhang, C. Xu, R. Hou, A. P. E. York and M. L. H. Green, J. Catal., 2003, 220, 47–56.
29. S. Gu, H. S. Oh, J. W. Choi, D. J. Suh, J. Jae, J. Choi and J. M. Ha, Appl. Catal., A, 2018, 562, 114–119.
30. J. S. Chang, V. P. Vislovskiy, M. S. Park, D. Y. Hong, J. S. Yoo and S. E. Park, Green Chem., 2003, 5, 587–590.
31. X. H. Li, K. Tomishige and K. Fujimoto, Catal. Lett., 1996, 36, 21–24.
32. T. Nishiyama and K.-I. Aika, J. Catal., 1990, 122, 346–351.
33. C. L. Chen, Y. D. Xu, G. J. Li and X. X. Guo, Catal. Lett., 1996, 42, 149–153.
34. S. Alzahrani, Q. Song and L. L. Lobban, Ind. Eng. Chem. Res., 1994, 33, 251–258.
35. K.-i. Aika and T. Nishiyama, J. Chem. Soc., Chem. Commun., 1988, 70–71,  10.1039/C39880000070.
36. S. J. Korf, J. A. Roos, N. A. de Bruijn, J. G. van Ommen and J. R. H. Ross, J. Chem. Soc., Chem. Commun., 1987, 1433–1434,  10.1039/C39870001433.
37. J. Shi, L. Yao and C. W. Hu, J. Energy Chem., 2015, 24, 394–400.
38. J. Shi, L. Yao and C. Hu, J. Energy Chem., 2015, 24, 394–400.
39. J. Y. Lee, W. Jeon, J.-W. Choi, Y.-W. Suh, J.-M. Ha, D. J. Suh and Y.-K. Park, Fuel, 2013, 106, 851–857.
40. R. M. Freire, F. F. de Sousa, A. L. Pinheiroa, E. Longhinotti, J. Mendes, A. C. Oliveira, P. D. C. Freire, A. P. Ayala and A. C. Oliveira, Appl. Catal., A, 2009, 359, 165–179.
41. J. S. Yoo, P. S. Lin and S. D. Elfline, Appl. Catal., A, 1993, 106, 259–273.
42. J. Wang, L. Chou, B. Zhang, H. Song, J. Zhao, J. Yang and S. Li, J. Mol. Catal. A: Chem., 2006, 245, 272–277.
43. V. O. Igenegbai, R. J. Meyer and S. Linic, Appl. Catal., B, 2018, 230, 29–35.
44. T. P. Tiemersma, M. J. Tuinier, F. Gallucci, J. A. M. Kuipers and M. v. S. Annaland, Appl. Catal., A, 2012, 433–434, 96–108.
45. M. R. Lee, M. J. Park, W. Jeon, J. W. Choi, Y. W. Suh and D. J. Suh, Fuel Process. Technol., 2012, 96, 175–182.
46. W. Jeon, J. Y. Lee, M. Lee, J.-W. Choi, J.-M. Ha, D. J. Suh and I. W. Kim, Appl. Catal., A, 2013, 464–465, 68–77.
47. R. T. Yunarti, S. Gu, J.-W. Choi, J. Jae, D. J. Suh and J.-M. Ha, ACS Sustainable Chem. Eng., 2017, 5, 3667–3674.
48. S. Gu, H.-S. Oh, J.-W. Choi, D. J. Suh, J. Jae, J. Choi and J.-M. Ha, Appl. Catal., A, 2018, 562, 114–119.
49. C. Karakaya, H. Y. Zhu, C. Loebick, J. G. Weissman and R. J. Kee, Catal. Today, 2018, 312, 10–22.
50. T. Yabe and Y. Sekine, Fuel Process. Technol., 2018, 181, 187–198.
51. Y. Liu, J. Xue, X. Liu, R. Hou and S. Li, in Stud. Surf. Sci. Catal., ed. D. S. F. F. A. V. A. Parmaliana and F. Arena, Elsevier, 1998, vol. 119, pp. 593–597.
52. Y. Liu, R. L. Hou, X. X. Liu, J. Z. Xue and S. B. Li, in Natural Gas Conversion V, ed. A. Parmaliana, D. Sanfilippo, F. Frusteri, A. Vaccari and F. Arena, Elsevier Science Publ B V, Amsterdam, 1998, vol. 119, pp. 307–311.
53. S. Kuś, M. Otremba, A. Tórz and M. Taniewski, Appl. Catal., A, 2002, 230, 263–270.
54. K. P. Peil, J. G. Goodwin Jr and G. Marcelin, J. Catal., 1991, 131, 143–155.
55. S. Sadjadi, S. Jaso, H. R. Godini, S. Arndt, M. Wollgarten, R. Blume, O. Gorke, R. Schomacker, G. Wozny and U. Simon, Catal. Sci. Technol., 2015, 5, 942–952.
56. T. W. Elkins and H. E. Hagelin-Weaver, Appl. Catal., A, 2015, 497, 96–106.
57. Z. C. Jiang, C. J. Yu, X. P. Fang, S. B. Li and H. L. Wang, J. Phys. Chem., 1993, 97, 12870–12875.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0ra06065h

‡ These authors equally contributed to this work.

---