Synthesis of Value-Added Chemicals via Oxidative Coupling of Methanes over Na2WO4-TiO2-MnO x /SiO2 Catalysts with Alkali or Alkali Earth Oxide Additives.

Na2WO4-TiO2-MnO x /SiO2 (SM) catalysts with alkali (Li, K, Rb, Cs) or alkali earth (Mg, Ca, Sr, Ba) oxide additives, which were prepared using incipient wetness impregnation, were investigated for oxidative coupling of methane (OCM) to value-added hydrocarbons (C2+). A screening test that was performed on the catalysts revealed that SM with Sr (SM-Sr) had the highest yield of C2+. X-ray photoelectron spectroscopy analyses indicated that the catalysts with a relatively low binding energy of W 4f7/2 facilitated a high CH4 conversion. A combination of crystalline MnTiO3, Mn2O3, α-cristobalite, Na2WO4, and TiO2 phases was identified as an essential component for a remarkable improvement in the activity of the catalysts in the OCM reaction. In attempts to optimize the C2+ yield, 0.25 wt % Sr onto SM-Sr achieved the highest C2+ yield at 22.9% with a 62.5% C2+ selectivity and a 36.6% CH4 conversion. A stability test of the optimal catalyst showed that after 24 h of testing, its activity decreased by 18.7%.

Introduction

Methane (CH4) is a resource that is abundantly found and serves as a major component of natural gas and biogas. As its greenhouse impact is greater than that of CO2, researchers worldwide are increasingly motivated to explore strategies for converting it into more useful chemicals. The oxidative coupling of CH4 (OCM)—the direct conversion of CH4 with air or O2 into hydrocarbons, such as ethylene, ethane, propene, propane, and so forth (C2+)—is one of the most challenging of those strategies. Its great potential hinges on the discovery of a suitable catalyst that can handle the highly exothermic conditions of the reaction, which typically result in temperatures above 600 °C, while maintaining high stability and suppressing the creation of CO and CO2.[1−4]

In the OCM mechanism, it was proposed that an atom of either adsorbed oxygen or lattice oxygen in a catalyst’s surface interacts with a hydrogen atom of the CH4 molecule to produce an OH intermediate and a methyl radical. Subsequently, the methyl radical interacts with another methyl group to generate a molecule of ethane, while two OH intermediates further couple together to generate an H2O molecule and an active oxygen atom species, which continues to abstract a hydrogen atom of a new CH4 molecule. However, the generating methyl radical is uncontrollable and other products (such as propane and propene) can also be produced. More importantly, the oxidation of CH4 and hydrocarbon products to CO2 and CO may occur, thereby resulting in a reduction in the C2+ selectivity.

In the past several years, various catalysts have been investigated for their suitability to the OCM reaction. One of the most selective catalysts reported to date is a combination of Na2WO4 and MnO on an SiO2 support (Na2WO4–MnO/SiO2).[5−14] Although a complete understanding of the nature of each catalyst component has not been well established yet,[13] each component in the catalytic material was proposed to have a special role in the mechanism, which is given as follows. Sodium was proposed to increase the basicity of the catalyst’s surface, thereby increasing the C2+ selectivity of the products.[15,16] Tungsten ions were proposed to improve the stabilization of the catalyst[14,15] and to provide the W=O and W–O–Si structures in the catalyst’s surface that are responsible for the OCM reaction.[17] Nevertheless, it was unclear how this catalytic system promotes a reaction. Manganese oxides were proposed to promote the oxygen mobility between surface-adsorbed oxygen atoms and the lattice oxygen atoms from the catalyst material. These oxygen species play an essential role in activating the CH4 molecule.[17] Moreover, the α-cristobalite phase, which occurs from a phase transformation of amorphous SiO2 during calcination in the presence of sodium, was proposed as an important phase for enhancing the product formation.[12,18−23] It is important to note that the phase transformation of amorphous SiO2 to the α-cristobalite phase in the absence of sodium generally occurs at relatively high calcination temperatures (>1500 °C), but it can easily occur at low calcination temperatures (approximately 750–900 °C) when sodium ions are present.[12,13] The performance of Na2WO4–MnO/SiO2 for OCM reaction was reported to be in the range of 5.2–20.6% C2+ yield with 36–80% C2+ selectivity and 8.2–35.4% CH4 conversion.[5−12,14,24]

Recently, various modifications of Na2WO4–MnO/SiO2 with different active components and promoters were widely investigated to improve the performance of the catalyst toward C2+ yield/selectivity. One example was the addition of metal oxides into the combination of Na2WO4 and MnO, including transition-metal oxides (Fe, Co, Ni, Ti, Ce, and La),[8,25−29] alkali metal oxides (Li, Na, K, Rb, and Cs),[30−34] or alkali earth metal oxides (Mg, Ca, Sr, and Ba).[30,32] The C2+ yields of those modified Na2WO4–MnO catalysts were in the range of 5–26% with 36–56% C2+ selectivity and 9–46% CH4 conversion. The addition of those components into Na2WO4–MnO led to the discovery of important information about the catalysts for the OCM reaction. For instance, (i) adding the catalyst with La2O3 caused an increase in the oxygen concentration of the catalyst’s surface, thereby increasing the catalytic activity;[26] (ii) adding a coactive metal (such as Pt, Ir, or Rh metal) led to the finding that the W concentration of the catalyst’s surface increased, thereby improving the catalytic activity;[8] (iii) using alkaline chlorides (e.g., NaCl or KCl) as promoters enhanced the CH4 conversion by which the alkaline chlorides generate a chloride radical so that CH4 activation was enhanced;[31] (iv) introducing an additive (such as TiO2) into the catalyst promoted low-temperature catalytic activity for the OCM reaction because of the formation of the MnTiO3 phase;[25] and (v) adding a Ce promoter to the catalyst resulted in increased activity and stability by preventing catalyst sintering.[27]

Previous results reported in the literature showed that the Na2WO4–MnO/SiO2 catalyst is highly active for OCM reaction. Thus, it is of great interest and it is challenging to modify the catalyst with promoters to improve its catalytic performance, especially with TiO2 and/or a metal oxide from alkali and alkali earth metals. Previous studies show that the use of TiO2 as a promoter in the Na2WO4–MnO/SiO2 catalyst induced a low reaction temperature[25,29] because of the formation MnTiO3, which activates the low-temperature Mn2+ ↔ Mn3+ cycle.[29] The addition of Ti and Mg into the Na2WO4–MnO/SiO2 catalyst was also reported to improve the OCM activity when Na2WO4 and MnO were on the external surface of the silica, but it was unclear how these two additives improved the activation.[35] Oxides of alkali or alkali earth metals (e.g., Li2O,[30,33,36−42] Na2O,[33,39] K2O,[39,43] Rb2O,[39] Cs2O,[39] MgO,[30,33,44−46] CaO,[30,33,45,46] SrO,[30,38,43,45−48] and/or BaO[45]) have been used as additives for various catalysts for the OCM reaction and were found to enhance the OCM activity because of their basicity property. In other words, they are unlike oxides of acidic metals, which prefer the complete oxidation of methane.[49] However, the investigation of the Na2WO4–MnO/SiO2 catalyst with TiO2 and an oxide of alkali or alkali earth metals was never reported. Therefore, in this work, we screened the alkali (Li, K, Rb, Cs) and alkali earth (Mg, Ca, Sr, Ba) oxide additives to determine whether any of these enhanced the activity of the Na2WO4–MnO–TiO2/SiO2 catalyst for the OCM reaction. Moreover, the OCM activities of the catalysts with the additives were evaluated to determine whether they have any correlation with the catalysts’ properties. After that, the catalyst and product optimizations, the effects of the operating conditions, and catalyst stability for the reaction of the most active catalyst were rigorously investigated.

Results and Discussion

Activity of Na2WO4–TiO2–MnO/SiO2 Catalysts with Alkali or Alkali Earth Oxide Additives

The Na2WO4–TiO2–MnO/SiO2 catalysts with alkali or alkali earth oxide additives were tested to determine their suitability for OCM reaction, under identical conditions. The metal-oxide additives included oxides of Li, K, Rb, Cs, Mg, Ca, Sr, and Cs. All the catalysts were synthesized by fixing a metal weight ratio of Na2WO4/TiO2/MnO/X of 5.0:5.0:0.5:0.5 (where X is the metal oxide additive and the weights of TiO2, MnO, and X are determined on the basis of their metallic forms). [Hereafter, this Na2WO4–TiO2–MnO/SiO2 catalyst will be denoted as SM.] The average activity test results with the error bars are shown in Figure and the products and byproducts are listed in Table S1. It must be noted that the data of each catalyst was collected after the catalyst was left to rest for 1 h under the feeding reactants and the setting temperature. Three batches of each catalyst were also prepared and tested in order to collect the average values for the activity test results. Moreover, each SiO2-supported additive and the SM catalysts physically mixed with each additive were tested, and the results (see Tables S2 and S3, respectively) were used for the control experiments. A comparison of the results of the control catalysts (the catalysts in Table S2) with the results of SM with and without the additives in Figure shows that the CH4 conversion and C2+ yield of all the catalysts in Table S2 were much lower than those in Figure . This indicates that the performance of each SiO2-supported additive catalyst is normally lower than that of the SM with and without additives under these conditions. Furthermore, when comparing each catalyst performance in Figure with that in Table S3, the C2+ yield of the catalyst prepared by the physical mixing method was lower than that of the catalyst in Figure , except for SM with Li in which the catalyst in Table S3 was slightly greater than that in Figure . Therefore, the catalysts prepared by the physical mixing method were not of interest in this work.

Figure 1

Activities of SM catalysts with alkali or alkali earth oxide additives. Testing conditions: catalyst amount 8 mg (gas hourly space velocity = 76,100 h–1) CH4/O2/N2 = 3:1:0, reactor temperature = 700 °C, atmospheric pressure, total feed gas flow rate = 35 mL min–1.

Considering Figure , the unmodified catalyst produced 6.7% C2+ yield with 41.5% C2+ selectivity and 16.1% CH4 conversion. The most promising catalyst was the one with Sr (SM–Sr)—yielding 12.8% C2+ with 44.4% C2+ selectivity and 28.9% CH4 conversion, followed by the catalysts with Rb and Ca, respectively. The others provided C2+ yields that were lower than that of the unmodified catalyst. The performances of the catalysts were ranked in terms of C2+ yield as: Sr > Rb > Ca > K > none > Mg > Ba > Li > Cs, providing a C2+ yield range of 3.7–12.8%. It should be noted that the activities of these catalysts were compared at only one active metal ratio. Thus, the activity results could be different at other ratios. In other words, this experiment investigated the obtainment of merely a local optimum. The characterization of the catalysts in the next section investigated how the physicochemical properties of these catalysts are related to their catalytic activities.

Characterization of SM Catalysts with Alkali or Alkali Earth Oxide Additives in Relation to Their Catalytic Activity

The crystalline phase composition of the prepared catalysts was examined using an X-ray diffraction (XRD) technique, as presented in Figure and Table S4. It was found that all the catalysts exhibited the crystalline phase of α-cristobalite, MnTiO3, Mn2O3, Na2WO4, and TiO2, except for SM–Li that did not have the MnTiO3 and Mn2O3 phases. It was confirmed that the α-cristobalite phase occurs at a low calcination temperature when sodium is present (800 °C), in good agreement with an earlier report.[50] Moreover, the characteristic peaks of SiO2 quartz were also observed in SM–Li and SM–Ba. Besides, the measurements using Fourier-transform infrared spectroscopy (FT-IR) and Raman spectroscopy (see Figure S1) confirmed that the α-cristobalite phase occurred in every catalyst. Note that the characteristics of each catalyst in terms of surface morphology and particle size were analyzed using SEM (see Figure S2), which found that all the catalysts were similar in shape but slightly different in size.

Figure 2

XRD plots of SM catalysts with different alkali or alkali earth oxide additives.

According to the catalytic activities in Figure , no clear characteristic XRD peaks of Li2O, SrO, CaO, MgO, and Cs2O were observed in the catalysts with the Li, Sr, Ca, Mg or Cs additive, potentially because they were in the amorphous phase or their particles were smaller than the limitations of the instrument. The catalysts that exhibited the XRD peaks of alkali or alkali earth oxides were SM with Ba, K, and Rb. For SM with Mg, Ca, and Sr, the crystalline phases of MgWO4, CaWO4, and SrWO4 were also detected, respectively. For the crystalline phase of Mn2O3, all the catalysts were presented, except SM–Li. Similarly, the crystalline phase of MnTiO3 was detected in all the catalysts, except SM–Li.

All of the XRD spectra shown in Figure are summarized in Table S5. The results presented in Figure , the information in Table S5, and previous studies reported in the literature,[12,21,22,29,50−54] all suggest that the catalyst must contain the crystalline phases of α-cristobalite, MnTiO3, Mn2O3, Na2WO4, and TiO2, in order to deliver a remarkable improvement in the activity in OCM reaction.

The presence of tungsten in the catalysts was proposed to be the most critical phase for the activation of CH4.[13] Therefore, the binding energies of the catalysts in the region of W 4f were determined using the X-ray photoelectron spectroscopy (XPS) technique. The XPS spectra of all the catalysts are shown in Figure S3. The observed peaks were associated with WO42– (W 4f5/2 ≈ 37.3 eV and W 4f7/2 ≈ 35.2 eV). This indicates that WO4 has a regular tetrahedral geometry in Na2WO4[22,55,56] and that the W atom is associated with the formation of W–O–Si bonds (i.e., the bridge oxygen), which has been proposed as an essential structure for the OCM reaction.[56] The presence of tetrahedral WO4 has also been found to correlate with the high CH4 conversion and C2 selectivity.[50] Besides, it can be seen that the binding energies of W 4f of each catalyst were not identical, which could have been caused by the presence of the different promoters added to the catalysts. The relationship of the CH4 conversion of each catalyst versus its binding energy of the W 4f region is plotted in Figure . It is interesting to observe that the relationships can be divided into two groups on the basis of CH4 conversion; the first group (SM–Sr, SM–Rb, and SM–Ca) had a CH4 conversion in the range of 23.6–28.9% and the second group (SM and SM with Li, Cs, Ba, Mg, and K) had a CH4 conversion in the range of 9.0–17.0%. It must be noted that the binding energy of each catalyst was compared to each other on the basis of the same weight percentage of the promoter content. The most interesting result was that the catalysts in the first group, which exhibited high CH4 conversions, had lower binding energies of W 4f7/2 than those in the second group. This may imply that the catalysts with a lower binding energy or a higher electron density of W can provide a higher CH4 conversion. This finding is similar to the report by Gu et al., in which the Na2WO4/SiO2 catalysts, enhanced by oxides of transition metals, that gave a high CH4 conversion were more likely to have a lower binding energy of W 4f7/2.[57] This suggests that the oxygen atoms that are involved in the reaction mechanism and bond to W with a low binding energy are potentially easy to detach after the hydrogen-extraction from the CH4 molecules and they leave the catalyst’s surface as H2O.[57] Deeper research is required to gain a better understanding of how the additions of Sr, Rb, and Ca result in the shifts in the W 4f binding energy.

Figure 3

Plots of CH4 conversion of each catalyst and its binding energy in the range of W 4f7/2.

Optimization of SM–Sr for the OCM Reaction

The SM–Sr catalyst was further studied for optimizing the C2+ yield in the OCM reaction by investigating the effect of varying the Sr loading on the SM catalyst, as presented in Figure . It clearly showed that the C2+ yield rapidly increased as the Sr loading increased from 0.0 to 0.25%. The maximum C2+ yield was achieved at 0.25 wt % (14.6% C2+ yield with 46.3% C2+ selectivity and 31.6% CH4 conversion). Then, the C2+ yields gradually decreased from the maximum to 7.6% with further increases in the Sr loading.

Figure 4

Effect of Sr loading onto the SM catalyst. Testing conditions: feeding gas ratio of CH4/O2/N2 = 3:1:0, total feed flow rate = 35 mL min–1, total catalyst weight = 8 mg, reactor temperature = 700 °C.

Three samples, consisting of Sr loadings onto SM at 0, 0.25, and 0.75 wt % (denoted as SM, SM–Sr0.25, and SM–Sr0.75, respectively), were chosen for characterization using XRD, SEM, and N2-physisorption to perceive how their physicochemical properties were related to their catalytic activity. These three catalysts were a reference catalyst, an optimal catalyst, and a catalyst after the optimum point, respectively. The XRD patterns of the three selected catalysts are shown in Figure S4. As observed, SM had crystalline phases of α-cristobalite, Na2WO4, Mn2O3, MnTiO3, and TiO2, which were identical to the SM presented in Figure . Similarly, SM–Sr0.25 and SM–Sr0.75 exhibited XRD patterns similar to the SM–Sr presented in Figure . However, the amount of crystalline SrWO4 became clearer as the amount of Sr loading increased. The SEM images of these three catalysts are depicted in Figure . The shapes of these catalysts were quite similar, but the overall particle size of each catalyst grew smaller as the amount of Sr loaded onto the SM increased. The BET measurements of the catalysts (Table ) confirmed that the BET surface area of SM–Sr0.75 was the greatest, followed by SM–Sr0.25 and SM, respectively. Their pore sizes were comparatively similar, in the range of 7–10 nm. Nevertheless, the adsorption isotherms plots shown in Figure S5 indicated that all the catalysts were type IV (mesoporous material), in good agreement with their pore volume, as presented in Table . Therefore, the analyses of these three samples implied that the addition of Sr onto SM can reduce the particle size of SM, thereby increasing its surface area or providing more active sites. This must be one of the reasons why adding a certain amount of Sr onto SM improved the catalytic activity. However, loading excessive amounts of Sr onto the SM resulted in reduced activity, possibly because the high content of SrWO4 on the catalyst surface caused reductions in the numbers of other important active species (α-cristobalite, Na2WO4, Mn2O3, MnTiO3, TiO2).

Figure 5

SEM images of (a) SM, (b) SM–Sr0.25, and (c) SM–Sr0.75.

Table 1

BET Surface Areas (SBET), Pore Sizes, and Pore Volumes of SM–Sr at 0.0, 0.25, and 0.75 Sr Loadings

material	SBET (m2 g–1)	pore size (nm)	pore volume (cm3 g–1)
SM	2.9	9.6	0.006
SM–Sr0.25	4.1	6.1	0.010
SM–Sr0.75	8.6	7.2	0.020

Optimization of C2+ Yield of SM–Sr0.25 in the OCM Reaction

The optimized catalyst (SM–Sr0.25) was further investigated at different catalyst amounts (8–72 mg) and feeding gas ratios (CH4/O2/N2 = 3:1:0 and 3:1:4), as shown in Figure . It was found that increasing the catalyst amounts resulted in increased C2+ production, which reached the optimal C2+ production for both feeding gas ratios. These resulted from the increased contact time and active sites of the catalyst, as expected. The highest C2+ yields of the CH4/O2/N2 feeding gas ratios at 3:1:0 and 3:1:4 were achieved at 24 mg or GHSV = 59,200 h–1 (20.0% C2+ yield with 54.2% C2+ selectivity and 36.7% CH4 conversion) and 48 mg or GHSV = 33,850 h–1 (20.3% C2+ yield with 57.6% C2+ selectivity and 35.0% CH4 conversion), respectively. At these optimal points, the oxygen feeds were completely consumed, as observed from the GC chromatograms. Normally, the experiment might be ended after all the oxygen is consumed because the expectation is that the activities of the catalyst should remain steady after reaching the optimal points. Surprisingly, however, when the amount of catalyst continued to be increased even after all the oxygen was consumed, the gradual decreases in the C2+ yields indicated on the plots were observed. This may have been because (i) the products (C2+) could partially further react in the gas phase. In other words, because they would have to diffuse through the additional catalyst bed, there is a greater chance of gas-phase reactions that create undesirable CO products or (ii) the pressure drop that slowly occurs over the reactor, or both reasons together.

Figure 6

Catalytic activity of SM–Sr0.25 for different catalyst amounts (8–72 mg = GHSV of 76,100–8,500 h–1) at two CH4/O2/N2 feeding gas ratios (3:1:0 and 3:1:4), a reactor temperature of 700 °C, and a total feed flow rate of 35 mL min–1.

It can also be noticed that, at 8–16 mg of the catalyst used under nondiluted and diluted feed conditions using N2, all the catalytic activities (CH4 conversion, C2+ selectivity, and C2+ yield) of the nondiluted feed conditions were higher than those of the diluted feed conditions. Furthermore, in the catalyst amount in the range 8–24 mg, the catalytic activities of the nondiluted feed conditions increased slowly with the increasing catalyst amounts, whereas the catalytic activities of the diluted feed conditions rapidly increased. This could have been due to the heat that was released from the reaction. Without N2, the temperature inside the catalyst bed could be higher than the operating temperature, as reported by Harold et al.[32]

The optimized catalyst amounts of SM–Sr0.25 for the conditions at CH4/O2/N2 feeding gas ratios of 3:1:0 and 3:1:4 were investigated by varying the reactor temperature from 600 to 800 °C, as presented in Figure . The activities (C2+ yield, C2+ selectivity, and CH4 conversion) continuously increased as the temperature increased from 600 to 700 °C with both CH4/O2/N2 feeding gas ratios. Then, the activities with the CH4/O2/N2 feeding gas ratio of 3:1:4 remained virtually unchanged in the reactor temperature range of 700–750 °C, followed by a decrease at reactor temperatures above 750 °C. The highest performance was observed at 750 °C, giving a C2+ yield of 15.9% with 54.7% C2+ selectivity and 29.0% CH4 conversion. The activity with the CH4/O2/N2 feeding gas ratio of 3:1:0 started to decrease immediately after the reactor temperature exceeded 700 °C. This suggested that using the inert gas with the reactant gases resulted in a reduction in the reaction temperature and the catalyst bed temperature. Consequently, the oxidation of the C2+ products was less likely to occur. In general, at relatively high temperatures (>700 °C) in the presence of O2, the C2+ species are oxidized to CO, even in the absence of a catalyst.[58] As a result, the C2+ yield and selectivity were lower at high temperatures. It must be clarified that, when the same reactor temperatures were considered, the activities of SM–Sr0.25 with nondiluted and diluted conditions using N2 were not largely different. This was the result of each condition that was carried out using its optimal catalyst amount, that is, 24 and 48 mg for the nondiluted and diluted conditions, respectively.

Figure 7

Effects of varying reactor temperature (600–800 °C) and CH4/O2/N2 feeding gas ratio (3:1:0 and 3:1:4) of SM–Sr0.25 for the OCM reaction. Reaction conditions: catalyst amount = 24 and 48 mg (GHSV of 25,400 and 12,700 h–1) for CH4/O2/N2 feeding gas ratio = 3:1:0 and 3:1:4, respectively, and total feed flow rate = 35 mL min–1.

The effect of the total feed flow rate of SM–Sr0.25 for the OCM reaction was determined to be in the range of 35–85 mL min–1 (GHSV of 12,700–30,800 h–1), as shown in Figure . The optimal feed flow rate was attained at 65 mL min–1 (GHSV = 23,500 h–1), which yielded 22.9% of C2+ with 62.5% C2+ selectivity and 36.6% CH4 conversion. Above 65 mL min–1, the performance of SM–Sr0.25 slowly decreased, which was probably because the residence time of the reactants over the catalyst surface was reduced, thereby resulting in a reduction in the chance of catalytic interaction between the reactant gases and the active sites.

Figure 8

Effect of varying the total feed flow rate of SM–Sr0.25 in the range of 25–85 mL min–1 (GHSV of 12,700–30,800 h–1). Reaction conditions: CH4/O2/N2 ratio = 3:1:4, catalyst amount of 48 mg, and reactor temperature = 750 °C.

Lastly, a stability test of SM–Sr0.25 was carried out at 750 °C for 24 h, as shown in Figure . The CH4 conversion, C2+ selectivity, and C2+ yield were minimally decreased from the second to the last hour. The C2+ yield was reduced from 22.9% in the first hour to 18.6% after 24 h (18.7% reduction), thereby indicating that the catalyst had encountered a deactivation problem. The analyses of fresh and used SM–Sr0.25 using XRD (see Figure ), SEM (see the inset image in Figure ), and N2-physisorption were then performed. The differences between these two catalysts were found as follows. (i) The XRD spectrum of the used catalyst revealed that the phase of α-tridymite had clearly appeared and the phases of Na2WO4, SrWO4, and Mn2O3 had almost disappeared. (ii) The particle size of the used catalyst (after 24 h) was much larger (>1 μm) than that of the fresh catalyst (<0.5 μm). (iii) The average surface area of the used catalyst was about 1.4 m2 g–1, as compared to 4.1 m2 g–1 of the fresh one. (iv) The pore size (12.7 nm with a pore volume of 0.005 cm3 g–1) of the used catalyst was found to be smaller than that of the fresh catalyst (6.1 nm with a pore volume of 0.010 cm3 g–1). These findings, similar to previous reports on the Na2WO4–Mn/SiO2 catalyst,[59,60] suggest that after several hours of testing the α-cristobalite phase could change to the α-tridymite and α-SiO2 phases, resulting in the decrease in the surface area of the catalyst[60] and the collapse of the pore. However, this transformation of the catalyst structure did not substantially influence the catalyst’s activity. Thus, the decrease and the catalyst sintering[31,61−63] may have only a small influence on the activity of the catalyst.[60,64] The likely cause of the gradual catalyst deactivation is the destruction of active phases from the catalyst’s surface, especially for Na2WO4 and Mn2O3 species.[52,61,65] Deeper characterizations of the fresh and used catalysts are required to understand this complex deactivation mechanism.

Figure 9

Catalytic performance of SM–Sr0.25 over 24 h of testing. Reaction conditions: CH4/O2/N2 ratio = 3:1:4, total feed flow rate = 65 mL min–1, catalyst amount = 48 mg (GHSV of 23,600 h–1), and reactor temperature = 750 °C.

Figure 10

XRD spectra and SEM images of fresh and used SM–Sr0.25.

Conclusions

The Na2WO4–TiO2–MnO/SiO2 catalyst with the alkali or alkali earth oxide additives were prepared and tested for suitability to the OCM reaction. The highest catalytic performance was that of the Sr addition at 0.25 wt % loading (SM–Sr0.25). The analyses of the XPS spectra revealed that the shift of binding energy of W 4f7/2 toward a lower energy was related to the high CH4 conversion, probably because of an enhancement of the oxygen mobility that occurred when the chemical bonding of W–O became weaker. Moreover, the crystalline phases of MnTiO3, Mn2O3, α-cristobalite, Na2WO4, and TiO2 phases described by the results of XRD, XPS, FT-IR, and Raman spectroscopy, were essential for the C2+ formation in the OCM reaction. The investigation of the operating conditions indicated that the optimal conditions that facilitated the highest C2+ yield (22.9% with 62.5% C2+ selectivity and 36.6% CH4 conversion) were a reaction temperature of 750 °C, a CH4/O2/N2 feeding gas ratio of 3:1:4, a total feed flow rate of 65 mL min–1, and 48 mg of the catalyst. Comparison of the activity of SM–Sr0.25 with other Na2WO4/SiO4-containing catalysts (see Table S6) showed that SM–Sr0.25 is a superior catalyst for the OCM reaction. The stability test of SM–Sr0.25 over 24 h and the analyses of the used catalyst showed that the activity slightly decreased over time, probably mainly because of the unstable active phases of Na2WO4 and Mn2O3 under the operating conditions over the long testing period.

Experimental Section

Catalyst Preparation

The catalysts were synthesized using incipient wetness impregnation. The precursor (chemical formula, purity, and brand) of each component is listed in Table S7. The weight percentage of each component on the SiO2 support (amorphous fume silica, surface area of 85–115 m2 g–1) was determined on the basis of the denoted component/element (see Table S8 and the description of Table S8 in detail). Our previous report found the optimized ratio of Na2WO4/TiO2/SiO2 at 5:5:90 by wt %,[66] and in some earlier papers, 0.5–3.0 wt % of Mn in Na2WO4–MnO–SiO2-containing catalysts was reported as promising.[54] Hence, for this study of the Na2WO4–TiO2–MnO/SiO2 catalysts with alkali or alkali earth oxide additives, catalysts consisting of 5 wt % Na2WO4, 5 wt % Ti, 0.5 wt % Mn, and 0.5 wt % X [X = no added (none), Li, K, Rb, Cs, Mg, Ca, Sr, or Cs] were employed. First, the SiO2 support (89.0–89.5 mg) was weighed in a ceramic crucible. Then, each component (e.g., Na2WO4, Ti, and Mn with/without a promoter) was determined from the stock solution to have the expected weight percentage. Each predetermined solution was then pipetted onto the SiO2 support. A magnetic bar was dropped into the mixture and the mixture was stirred using a hotplate-stirrer (Scilogex, MS7-H550-S) at room temperature for 2 h, followed by constant stirring at 105 °C until dry. The magnetic bar was taken out and the powder was calcined in an air furnace at 800 °C for 4 h at a heating rate of 10 °C min–1. After the sample had naturally cooled down to room temperature, the calcined powder was ground until fine powder was obtained, which was ready for testing. A similar procedure was carried out for the other catalysts presented in this work.

Catalyst Activity Test

The catalytic activities of the catalysts were evaluated in a plug flow reactor for the OCM reaction. The testing was performed at atmospheric pressure and a reactor temperature between 600 and 800 °C. Note that the temperature inside the catalyst is usually higher than the set reactor temperature because of the heat released from the reaction.[9,67] The catalysts were packed in the middle of a quartz tube (0.5 cm in inner diameter and 30 cm in length) and placed between two layers of quartz wool. The reactant gases, consisting of methane (CH4, 99.999% purity, Praxair) and oxygen (O2, 99.999% purity, Praxair), with nitrogen (N2, 99.999%, Praxair) as the inert gas at a ratio of CH4/O2/N2 = 3:1:0 or 3:1:4, were fed into the plug flow reactor at a total feed flow rate of 35–85 mL min–1. All the flow rates were controlled using mass flow controllers (Aalborg, GFC173S) and double-checked using a bubble flow meter. The effluents were analyzed using gas chromatography (Shimadzu, GC-14A) with a thermal conductivity detector for the analysis of CO, CO2, and CH4 and with a flame ionization detector for the analysis of C2+ = C2H4, C2H6, C3H6, C3H8, C4H8, and C4H10. The catalyst activity was collected 1 h after the system reached the set point. The standard deviation (R2) value of each standard curve was >0.997. The % CH4 conversion, % C2+ selectivity, % CO selectivity, and % C2+ yield were calculated using eqs –4.

Catalyst Characterization

The plots of powder XRD of the prepared catalysts were obtained using a powder X-ray diffractometer (JEOL, model: JDX-3530 and Philips X’Pert, using Cu Kα radiation at 45 kV and 40 mA, 0.02 step size, 0.5 step time).

The binding energies of the catalysts in the tungsten region (W 4f) were acquired using an X-ray photoelectron spectrometer (XPS, Kratos Axis Ultra DLD) with Al Kα for the X-ray source.

The surface morphologies of the prepared catalysts were captured using a scanning electron microscope with an energy-dispersive X-ray spectrometer (SEM/EDS, JEOL JSM7600F). The working distance was set at 4.3 mm. Each catalyst was coated with gold using a gold-sputtering technique before conducting the SEM analysis.

The BET surface areas, the pore sizes, the pore volumes, and the isotherm plots of each prepared catalyst were obtained using N2-physisorption (BET: 3Flex Physisorption Micrometrics). Each catalyst was degassed at 200 °C for 24 h prior to the measurement. The average specific surface area of each catalyst was determined to be in a range of P/P0 values between 0.05 and 0.30. The total pore volume was estimated to be at a relative pressure (P/P0) of 0.995. The pore size of each catalyst was determined using the Barrett–Joyner–Halenda method.

The presumption that the physicochemical properties of fresh and the used catalysts (used for 1 h) were similar was confirmed by SEM and XRD analyses of SM–Sr0.25 (see Figures S6 and S7, respectively).