1 Introduction

The growing need for heavier hydrocarbons in chemical industries has reached a new height in the twenty-first century. Over the last few decades, extensive research has been carried out to directly convert methane, the major component of natural gas, to value-added chemicals using heterogeneous catalysts [1]. An increased production of natural gas has led to a reduction in natural gas price compared to crude oil. Currently, the major use of natural gas is power generation and heating [2]. Despite extensive research for alternative uses, much of the natural gas produced as by-products of oil recovery is combusted and released [3], which is not an economic and environment friendly utilization of this major natural resource. Current industrial technologies for the conversion of natural gas are indirect processes, which are based on energy inefficient and involve complex steps involving the production of syngas (CO and H2) [4]. Direct conversion routes can eliminate the requirement of this intermediate step [5, 6] and produce valuable chemicals with high energy efficiency.

Conventional processes of producing ethane and ethylene from methane involves the presence of oxidant, i.e., O2, CO2, S. One indirect approach is known as oxidative coupling of methane (OCM) [7]. Reported in early 1980s by Bhasin and Keller [8] as well as Hinsen and Baerns [9], this mechanism involves extraction of a hydrogen atom from methane to activate it and produce methyl intermediates. This methyl radical may react with another one to produce ethane. If another hydrogen atom can be extracted from the methyl radical, it would generate CH2* intermediate, which may react with a similar radical to produce ethylene. Studies have shown that mixed metal oxides like La-based perovskites [10], MnNaW/SiO2 [11], Mo/HZSM-5 [12, 13] and noble metals [14] help oxidize methane toward ethylene and ethane. The challenge of this approach is that selectivity to C2s has been reported to be relatively low and the process involves high temperature with complex intermediates [5] which make the whole process less efficient.

A similar mechanism can also be proposed for direct activation of methane to C2 hydrocarbons. Progress has been made toward understanding how methane can be directly activated by forming methyl radicals [15], without the involvement of coreactants, which may further react to produce higher hydrocarbons. This route of methane activation to ethane and ethylene requires the presence of a solid acid catalyst [16]. The reaction has two important pathways, one converts methane to ethane and the other produces ethylene (Fig. 1).

Fig. 1
figure 1

Methane to C2 hydrocarbons possible reaction pathways

Full size image

Thermodynamically, such methane activation pathways ultimately drive the reaction spontaneously toward formation of solid carbon [1, 17], which is the most stable product at equilibrium conditions. In order to drive the selectivity toward C2s, an oxide heterogeneous catalyst is necessary [18,19,20].

One of the most important properties needed for such catalysts is the ability to form oxygen vacancies [21]. For strong oxidizing agents, oxygen vacancies are readily formed. This is the similar mechanism that enables methane during OCM to convert and selectively produce C2 hydrocarbons. A recent DFT study by Cheng et al. [22] suggested that a limited increase in concentration of oxygen vacancy on the oxide-based catalyst results in significant decrease in the energy required for the hydrogen extraction to break the methane molecule and form methyl radicals. Oxygen vacancies thus can promote the direct conversion of methane.

Here, this study focuses on using W oxide-based acid catalyst. Previously, transition metals of group VIB have shown activity for alkane activation [23, 24]. Mo/W oxide supported on ZSM-5/MCM-22 is a well-studied bifunctional catalyst for methane dehydroaromatization [2, 25]. It has been reported that the metal carbide species produced from Mo/W oxides activate methane by forming CHx species, which is dimerized into C2Hy species [13, 26]. Previously, this group demonstrated high activity and strong aromatic selectivity for methane dehydroaromatization (MDHA) using Mo oxide doped on sulfated zirconia (SZ) solid acid catalysts [15]. During these studies, it was observed that the addition of W as promoter to Mo/SZ shifted the MDHA product selectivity toward ethylene and ethane [24]. W is known to be a strong oxidizing agent [27, 28]. It reacts with rare earth elements, Fe, Cu, Al, Mn, Zn, Cr, Mo, C, H2 and can be reduced to pure tungsten metal [29].

The rationale is to dope W oxide into a stable non-reducible oxide that exhibits high resistance to the loss of oxygen and low reactivity toward hydrogen at high temperature. Although silica and alumina have been reported as strong non-reducible catalyst supports [30, 31], recent studies have revealed zirconium oxides to demonstrate higher strength in working conditions [32]. It has been reported that the sulfurized form of ZrO2 known as sulfated zirconia (SZ), actively helps the active metal sites to stabilize on the oxide support at a greater degree [29, 33]. SZ is a well-known solid acid possessing surface H + ions [29, 34], which can react with some of the excess methyl radicals and remove these in the form of benzene and heavier hydrocarbons as valuable side products [15, 35].

This present study is a follow-up of this group’s recent works with sulfated zirconia-based solid acid catalysts for methane activation. Here, we introduce a novel catalytic approach, where W oxide is supported on SZ and investigate its effect on direct activation of methane. To the best of our knowledge, no study has shown methane activation with W/SZ catalyst to selectively produce C2 hydrocarbons. Synthesis of this novel catalyst is followed by characterization using pyridine DRIFTS, BET, ammonia TPD, SEM–EDX, XPS, XRD and temperature programmed techniques. Afterward, W/SZ is tested to activate methane and its performance on C2 product selectivity will be evaluated in thermodynamically favored reaction temperatures. This is followed by subsequent temperature programmed oxidative analysis on carbon deposition.

2 Catalyst Preparation

2.1 Materials

Ammonium metatungstate hydrate, (NH4)6H2W12O40·xH2O and zirconium hydroxide, Zr(OH)4 (97%) precursors were purchased from Sigma-Aldrich Inc. H2SO4 (95–98.0%) was purchased from Malinckrodt Chemicals Inc. Ultrahigh purity grade He, H2, CH4 and 10% O2/He were ordered from Airgas Inc.

2.2 Synthesis

Sulfated zirconia was prepared by following conventional methods [24, 29]. 35 g of zirconium oxide was mixed with 500 mL of 0.5 M H2SO4 solution prepared with DI water. The mixture was stirred for 2 h, followed by vacuum filtration with excess DI water. The retentate was dried at 110 °C overnight followed by calcination at 550 °C for 4 h to prepare the final catalyst.

W/SZ was synthesized by following the preparation method of Mo/SZ [24, 32]. The catalyst was prepared using incipient wetness impregnation method with ammonium metatungstate hydrate as the precursor. W impregnation was carried out at room temperature. 5 wt% of W was added in the form of ammonium tungstate to 100 ml DI water mixture. 8 gm of prepared SZ was later added to this solution. The mixture was stirred for 2 h, followed by vacuum filtration with excess DI water. The sample was dried at 110 °C overnight and calcined at 550 °C in atmospheric condition for 4 h. The catalyst was passed through a sieve to achieve the final powdered form. Later, it was weighed and collected in glass ampule to avoid exposure to moisture and air.

3 Characterization Techniques

3.1 BET

Altamira AMI-200 characterization reactor was used for breauner emmett teller (BET) surface area measurement with N2 monolayer adsorption. To measure the catalytic surface area, a three-point BET was used in the presence of 10%, 20% and 30% N2 concentrations.

3.2 Ammonia TPD

Ammonia TPD was carried out in Altamira AMI-200 reactor system coupled with Ametek mass spectrometer. 50 mg of the prepared catalyst was loaded on a quartz U-tube reactor. Catalytic pretreatment was followed by using He as the inert gas source. Temperature was increased till 200 °C with 30 sccm of He flow and held for 30 min to eliminate any weakly physisorbed particles on catalytic surface. Sample temperature was cooled down to 50 °C under He, followed by introduction of 40 sccm of 5% NH3/He to initiate NH3 adsorption process for 90 min. 30 sccm He was flown subsequently for 40 min to get rid of any residual ammonia. Mass spec and TCD detector were later turned on, and the temperature was increased at 10 °C/min. from 50 to 700 °C. Based on the signal from mass spec and TCD, amounts of ammonia desorbed and peak positions were calculated to quantify the corresponding acid sites available on the catalyst.

3.3 Pyridine DRIFTS

Pyridine is a weak base and was used as a probe molecule for catalytic acid site characterization with diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). A Thermo Scientific Nicolet 6700 FTIR equipped with harrick praying mantis (HPM) reaction cell fitted with KBr windows was used for DRIFTS experiment.

Glovebox technique was used to load the HPM cell with catalyst powders to avoid atmospheric exposure. Helium was introduced in the cell initially to remove any residual moisture or physisorbed particles. This was followed by catalytic pretreatment by increasing the temperature to 100 °C and was held for 30–40 min to clean the surface from adsorbed impurities. To collect a background spectrum, sample temperature was reduced to 25 °C, and a spectral resolution of 4 cm−1 was recorded within the region of 2000–250 cm−1. Gaseous pyridine was introduced to saturate the surface for 180 min at 25 °C. Helium was reintroduced in the follow-up step to remove the physisorbed pyridine at the catalyst surface and the IR cell walls. Temperature was raised at 100 °C again to desorb the pyridine that was bonded to the strong acid sites of the sample. It was held at 100 °C for 10 min and cooled back to room temperature to record the pyridine adsorption spectrum. Similar spectra were recorded at 25 °C after 10 min. This procedure was repeated at 200 °C, 300 °C and 400 °C to investigate thermal stability of the acid sites.

3.4 XPS

X-ray photoelectron spectroscopy was performed at Louisiana State University shared instrument facility (SIF) to understand the oxidation sates of the active metal. Samples were characterized using Scienta Omicron ESCA 2SR XPS instrument with aluminum monochromatic X-ray source at 15 kV with pass energy of 40. XPS data were analyzed using CasaXPS licensed software.

3.5 SEM–EDS

SEM–EDS analysis of the samples was performed to determine elemental composition at the bulk level. The experiment was carried out at LSU shared instrument facilities (SIF) using FEI quanta 3D FIB/SEM coupled with Ametek EDAX accessory. Voltage of 5 kV and a resolution of 3 µm were maintained to detect elemental compositions.

3.6 XRD

PANanalytical EMPYREAN diffractometer with Cu Kα radiation was used to perform XRD analysis. Spectra were scanned between the 2 θ range of 5° to 90° with step size of 0.1°. XRD data were analyzed with licensed PANanalytical X’Pert software, and data comparison was done using MS Excel.

4 Experimental procedure

The prepared catalyst was run at a temperature range of 650–750 °C for ~ 15 h to investigate direct activation of methane. The catalysts were loaded in an Altamira AMI 200HP reactor system equipped with quartz tube reactor and reduced under H2 flow till the desired reaction temperature was reached. This was followed by carburization while flowing a gas mixture of methane and H2 (flow ratio 1:4), which was introduced into the reactor for 4 h to further reduce the catalyst. H2 and CH4 gases were stopped, and the reactor was purged with helium as inert gas. Methane was reintroduced later to carry out methane activation reaction.

Downstream reaction product gas was analyzed using Shimadzu GC2014 (FID, 2 TCDs) equipped with Restek RT-Q-bond column (30 m × 0.53 mm × 20 μm) in conjunction with Shimadzu QP2010 GC–MS system.

Conversion of methane was calculated with the following formula.

$$\% {\text{ CH}}_{4} {\text{conversion }} = \frac{{{\text{mol CH}}_{{4\,{\text{in}}}} - {\text{mol CH}}_{{{4}\,{\text{out}}}} }}{{{\text{mol CH}}_{{4\,{\text{in}}}} }} \times 100$$

Product selectivity was calculated based on the gaseous products observed from conversion of methane. Carbon deposition is excluded from this calculation as the coke formed at any instant of the reaction is not measured. Another reason for the exclusion is because the rate of coke formation differs over a period and is not discussed in this work.

$$\% {\text{ product selectivity }} = \frac{{\text{mol Product}}}{{\text{mol total products}}} \times 100$$

5 Results and discussion

5.1 Physicochemical properties

BET physisorption showed that surface area of W/SZ was less than pure SZ (Table 1). Literature reports the surface area of SZ to be 50–100 m2/g [36, 37]. When W was impregnated in SZ, a slight decrease in surface area indicates some blockage in porous regions of SZ due to loading of the active metal.

Table 1 Physico-chemical properties of W/SZ catalysts
Full size table

To determine the actual loading of W on SZ, SEM–EDS and XPS analysis were carried out. It was observed from both analysis that the actual amount of W was lower than the intended loading. This may be due to loss of W during catalytic synthesis.

5.2 Pyridine DRIFTS

Pyridine is typically used as a weak base to probe the strong acids sites of catalysts. DRIFTS analysis was carried out to investigate the stability of SZ acid sites, before and after W was loaded into the catalyst. Adsorbed pyridine sites on SZ correspond to either Brønsted, Lewis acid sites or dual acid sites, detected from the vibrational bands which can be distinguished between the types of acid sites.

Vibrations at around 1445 cm−1 and 1610 cm−1 represent Lewis acid sites, whereas vibrations at around 1545 cm−1 and 1645 cm−1 represent Brønsted acid sites [38]. Vibrations at around 1495 cm−1 represents sites that contain both Lewis and Brønsted acidity [38, 39]. All of these characteristic bands were observed in SZ catalysts, before and also after W loading (Fig. 2). This indicates that surface acidity was not compromised when active metals were impregnated into SZ. An increased intensity of Lewis acid sites on W/SZ refers to the addition of Lewis acidic W sites on SZ. The small peak at ~ 1580 cm−1 represents physisorbed pyridine [38].

Fig. 2
figure 2

DRIFTS of the prepared catalysts at 400 °C, after pyridine exposure for 3 h

Full size image

Both catalysts were tested at temperatures up to 400 °C to study the stability of these acid sites at elevated temperatures. It was observed that both catalysts showed stable acid sites, as observed from Fig. 2.

5.3 Ammonia TPD

To quantify the amount of acidity in the catalysts before and after W addition, ammonia was used as a probe molecule for temperature programmed desorption (TPD). Ammonia is a weak base molecule with pKb value of 4.5 [40].

Figure 3 shows a comparison of ammonia TPD curves for the two catalysts: base SZ and W/SZ. SZ showed a typical ammonia TPD peak at ~ 120–140 °C range, followed by a long shoulder that drops down slowly to ~ 550 °C. W/SZ showed a very similar TPD curve as well, but the intensity in the TPD curve went down after loading W onto SZ, indicating an overall loss of the total acidity.

Fig. 3
figure 3

NH3-TPD of fresh catalysts, 40 sccm 5%NH3/He flow, 50 mg

Full size image

This decrease in total acidity with the addition of W to SZ was quantified by measuring the total TPD area for both SZ and W/SZ (Table 2). This can occur due to a possible blockage of microporous SZ channels by W oxide particles, thus restricting the access of ammonia to some acid sites.

Table 2 Amount of NH3 desorbed/consumed per gram of catalyst
Full size table

5.4 XRD

SZ has distinct XRD pattern widely reported in the literature [24, 32]. Different phases of SZ have been observed based on the synthesis technique and calcination temperatures [37, 41]. In this work, since SZ was calcined at 550 °C. At this temperature tetragonal phases mostly dominate the chemical structure [29, 37]. Figure 4 shows the XRD spectra of SZ and W/SZ catalysts. Base SZ clearly showed all the characteristic peaks attributed to tetragonal phase (reference ICDD PDF # 811,544). After W was loaded onto SZ, no difference in XRD pattern is observed, suggesting that W is evenly dispersed into SZ structure in amorphous form, with little crystallinity. This also implies that base structure of SZ remained intact even after W was impregnated.

Fig. 4
figure 4

XRD patterns of SZ and W/SZ

Full size image

5.5 Activation of methane

The catalysts were tested for methane activation at different temperatures to understand their effect on activity and product selectivity. Before initiating the catalytic reaction, a blank reaction was carried out in the absence of the catalysts at reaction conditions. Only methane was observed in the product stream, detected by GC–MS. Afterward, three catalytic reactions were carried out for ~ 15 h each, at 650, 700 and 750 °C in otherwise identical reaction conditions. The products observed in all of the catalytic runs were primarily ethylene, ethane, with small amount of aromatics including BTEX (benzene, toluene, ethylbenzene and xylenes).

As expected, initial methane conversion increased with temperature, but quickly deactivated with time, reaching similar methane conversion with time at each temperature (Fig. 5). This can be attributed to catalytic deactivation due to carbon deposition. Even though methane conversion did not differ significantly with temperature after less than 100 min, product selectivity was significantly different.

Fig. 5
figure 5

Conversion of methane with 5% W/SZ, 1 g, 10 SCCM methane flow, 1 atm

Full size image

Product selectivity toward C2 hydrocarbons increased with temperature (Fig. 6). Selectivity increased initially for 650 °C and reached stability at ~ 70% after ~ 300 min of run. For 700 °C, selectivity dropped down initially and then reached a stable ~ 76% selectivity throughout the run. Highest amount of C2 selectivity was observed for 750 °C, which went up initially for a short period of time and then became stable at ~ 93% throughout the run. This increase in C2 selectivity with temperature can be attributed to more W active sites being reduced at a faster rate. When temperature increases, more W oxide are reduced [42], as a result more methyl radicals turn to C2 dimers [24], thus produce ethylene and ethane.

Fig. 6
figure 6

C2 product selectivity for methane conversion with 5% W/SZ, 1 g, 10 SCCM methane flow, 1 atm

Full size image

Benzene was a major product observed at 650–700 °C, due to the enhanced acidity provided by SZ [15] (Fig. 7). For 650 °C, benzene selectivity was initially at ~ 35%, which decreased rapidly with time. A similar trend was also observed with 700 °C, where benzene selectivity increases initially and decreased rapidly with time as well. For 750 °C, very little benzene formation was observed, product selectivity was mostly C2 hydrocarbons at this temperature. Acidity from SZ promoted methyl radicals to produce benzene instead of carbon. This transformation, however, was unstable in this case. Acidity of SZ is dependent on the available sulfate ions, which are responsible for supplying H+ ions during the reaction. It has been reported that at temperatures higher than 630 °C, sulfate ions are not stable in dynamic reaction conditions [34], and leave SZ surface in gaseous states (SOx). Thus, SZ loses the required acidity to transform methyl radicals to aromatics like benzene and is similar to ZrO2, as observed from these runs.

Fig. 7
figure 7

Benzene product selectivity for methane conversion with 5% W/SZ, 1 g, 10 SCCM methane flow, 1 atm

Full size image

5.6 Temperature programmed oxidation (TPO)

To understand the deactivation mechanism of W/SZ, TPO was carried out on each spent catalyst from the three temperature runs. Catalytic deactivation is mostly due to carbon deposition [43], which can be categorized as either amorphous, with a peak position at ~ 400 °C, or polymeric carbon generated from aromatics, with peak temperatures at 500–600 °C range, or graphitic carbon with peaks at 650–750 °C range [44].

Spent catalyst from 650 to 700 °C runs showed sharp TPO peaks at around 500–600 °C range, attributable to polymeric carbon (Fig. 8). Liu et al. [45]reported that Brønsted acid sites from SZ provide active sites for the formation of aromatics and lead to carbon formation as final product. At 650 °C, there was more aromatics formation than 700 °C, indicating that less sulfate ions are vaporized from the SZ surface at lower temperatures, thus producing more carbon precursors.

Fig. 8
figure 8

TPO data for the three spent catalysts, 15 sccm O2/He, 25 mg

Full size image

For the 750 °C spent catalyst, a small shoulder at 500–550 °C region indicates very little formation of polymeric carbon. A stronger peak at ~ 700 °C indicates the formation of graphitic carbon, which can be attributed to carbon from active metal or acid sites that are strongly attached to the catalytic surface [46]. This suggests that W oxides were greatly reduced at higher temperature to W carbides or oxycarbides, which are responsible for enhancing dehydrogenation activity [47,48,49], as observed during the 750 °C run.

Amount of carbon deposited was quantified for all three runs (Table 3). More carbon was observed from the 650 °C run with the highest aromatics formation, which is in accordance with the TPO analysis.

Table 3 Quantification for carbon deposition with TPO (after ~ 15 h of reaction)
Full size table

6 Conclusion

Activation of methane was studied with novel W/SZ catalyst to study the correlation of the effect of reaction temperature with product selectivity. Previous studies with Mo/SZ successfully demonstrated high activity for MDHA, with affinity toward C2 formation when W was present as a promoter [24]. The objective of this follow-up work was to use a similar novel approach to directly convert methane to C2 hydrocarbons by introducing WOx as the primary active site onto SZ solid acid, characterizing the catalyst and understanding the role of W toward C2 hydrocarbons selectivity.

Characterization of SZ and W/SZ was carried out using pyridine DRIFTS and ammonia TPD, which confirmed that addition of W to SZ increased Lewis acid sites in SZ but decreased the total acidity. BET surface area measurements showed high dispersion of W active sites inside SZ pores. SEM–EDS and XPS confirmed that the actual loading of W was close to the intended amount. XRD suggested that W is well dispersed throughout SZ in amorphous form.

Experimental results showed that methane activation did not vary significantly with temperature, although change in product selectivity was observed. W/SZ was found to be highly selective to ethylene and ethane (~ 73–93% selectivity) at 650–750 °C, which was the primary goal of this study. C2 selectivity increased with temperature and remained stable with time. Benzene was another major product observed at 650–700 °C range, which decreased rapidly with temperature and time due to possible loss of sulfate sites from SZ at high temperature. TPO results indicate that the primary reason of catalytic deactivation is due to polymeric(650–700 °C runs) and graphitic carbon deposition (750 °C run).