Integrated In Situ Characterization of a Molten Salt Catalyst Surface: Evidence of Sodium Peroxide and Hydroxyl Radical Formation.

Sodium-based catalysts (such as Na2 WO4 ) were proposed to selectively catalyze OH radical formation from H2 O and O2 at high temperatures. This reaction may proceed on molten salt state surfaces owing to the lower melting point of the used Na salts compared to the reaction temperature. This study provides direct evidence of the molten salt state of Na2 WO4 , which can form OH radicals, using in situ techniques including X-ray diffraction (XRD), scanning transmission electron microscopy (STEM), laser induced fluorescence (LIF) spectrometry, and ambient-pressure X-ray photoelectron spectroscopy (AP-XPS). As a result, Na2 O2 species, which were hypothesized to be responsible for the formation of OH radicals, have been identified on the outer surfaces at temperatures of ≥800 °C, and these species are useful for various gas-phase hydrocarbon reactions, including the selective transformation of methane to ethane.

Gas‐phase radical chemistry involving OH radicals plays a crucial role in the oxidative coupling of methane (OCM),1 the dehydrogenation of ethane,2 atmospheric chemistry,3 and combustion reactions.4 The catalytic generation of OH radicals from O2 and H2O can occur on Pt metal and alkali earth oxides at high temperatures (>700 °C).5 Alkaline metal containing catalysts (that is, Na supported on oxide) enhanced the rate of H2O activation in the presence of O2 and enhanced both the CH4 conversion rate and C2 selectivity during OCM.1 Otsuka et al. demonstrated that Na2O2 reacts with CH4 to form methyl radicals at relatively low temperatures.6 Recently, kinetic evidence suggests that H2O is involved in the activation of CH4 via the quasi‐equilibrated formation of OH radicals when a Na‐containing catalyst is used.1 In a similar context, an in situ Raman spectroscopic study identified Ba species on MgO that formed peroxide ions and was proposed as the active sites.7 However, direct evidence of the formation of this sodium peroxide species and an OH radical product are lacking. Here, in situ studies performed at high temperatures using different in situ characterization techniques were performed to identify the authentic active species during catalysis responsible for the OCM reaction. We report evidence for the formation of Na2O2 and OH radicals during the catalysis, and these species play a significant role in high‐temperature gas‐phase chemistry.

A detailed kinetic investigation that was focused on the effects of H2O on the CH4/O2 reaction was reported in our previous studies.1 Briefly, the kinetic expression for the CH4/O2 reaction on Na‐based catalysts at low conversion levels is given in (1):

The first term corresponds to CH4 activation via surface O (O*), which is quasi‐equilibrated with gas‐phase O2 (dry condition). The second term corresponds to CH4 activation via an OH radical that is formed from O2 and H2O in quasi‐equilibrated steps (wet condition). Among the studied catalysts, the Na2WO4/SiO2 catalyst exhibited the highest contribution from the OH radical pathway relative to the surface O* pathway as well as the highest selectivity to C2 hydrocarbons from CH4.1d

In this study, the TiO2 support was chosen to replace SiO2 to immobilize Na2WO4 and avoid charging effect during characterization using TEM and AP‐XPS. The extensive heat treatment at 900 °C for 15 h in flowing air (100 cm3 min−1) was essential to ensure the stability of the catalyst during OCM. In this process, the excess Na2WO4 was evaporated. The resulting Na contents in the SiO2 and TiO2 samples, which were measured using inductively coupled plasma, were about 4 and about 0.6 wt %, respectively, and these values did not change after the OCM kinetic testing at 800 °C. The BET surface areas that were estimated by N2 sorption were 2 and 3 m2 g−1 for Na2WO4/SiO2 and Na2WO4/TiO2, respectively. In our studies, we confirmed that kinetic expression (1) applies to the Na2WO4/TiO2 catalyst, which is consistent with the data reported for the Na2WO4/SiO2 catalyst.1 The Na2WO4/TiO2 catalyst exhibits excellent OCM selectivity in the presence of an O2/H2O mixture and the catalyst most likely has the ability to form OH radicals from an O2 and H2O mixture. The details of the kinetic results are described in the Supporting Information, Figures S2. The proposed catalytic formation of OH radicals on the Na catalyst is as in (2), (3), (4):2, 4

Herein, H2O and H2O2 can be gas‐phase species or adsorbed species on the surface, even though gas‐phase H2O2 (or decomposed product) is most likely present because its boiling point is 150 °C. The Na2O2 decomposition temperature is 657 °C, which may remain kinetically formed in the presence of O2. The overall reaction for the OH radical formation (when quasi‐equilibrated in the gas phase) is according to (5):

Direct detection of OH radicals was attempted using a LIF spectrometer with the setup described in the Supporting Information, Figure S1. Owing to the detection limitations of our system, a transient condition was applied for the detection of OH radicals. The catalyst, which was placed in an alumina boat, was initially heated to 900 °C under vacuum (<1 Pa), followed by the introduction of 2.3 kPa of H2O into the system. As the 20 kPa of O2 was gradually introduced into the heated reactor with the catalyst, the change in the LIF signal was successively recorded. Figure 1 shows the resulting transient detection of OH radicals. The OH radicals were detected upon introduction of an O2 and H2O mixture at 900 and 950 °C, which is consistent with reaction (5). This result unambiguously confirms that the Na catalyst can form OH radicals from an O2 and H2O mixture. To the best of our knowledge, this result is the first evidence for the direct detection of OH radicals from an O2 and H2O mixture on a Na‐based catalyst surface.

Figure 1

Detection of OH radicals through a quartz window using LIF spectroscopy. The setup details are provided in the Supporting Information.

It is important to note that the melting point of Na2WO4 (698 °C) is lower than the typical reaction temperature for the CH4 reaction (≥800 °C). Therefore, we attempted to capture the change of crystal structure and surface morphology using in situ XRD and TEM at high temperature in air. Figure 2 a shows the XRD patterns of Na2WO4/TiO2 collected at 500–800 °C in flowing air. When the temperature was increased from 500 to 700 °C, the XRD peaks for Na2WO4 disappeared, which suggests the formation of its molten salt state. Exposure to ambient air/moisture at room temperature resulted in recovery of the original phase (that is, cubic Na2WO4; the top XRD pattern of sample cooled down to 25 °C in Figure 2 a). This in situ XRD result (see also the Supporting Information, Figure S3) is consistent with the in situ Raman spectroscopy data reported by Yu et al., which confirms the disappearance of characteristic Raman peaks for Na2WO4 (supported on CeO2) at high temperature.8 Our sample shows very small Raman peak ascribable to Na2WO4 owing to its low loading, as shown in the Supporting Information, Figure S4. The fusion temperature coincided with endothermic/exothermic behaviors for the differential scanning calorimetric measurement of the Na2WO4 salt during heating/cooling (Supporting Information, Figure S5).

Figure 2

a) XRD patterns and b)–e) HAADF‐STEM images for the Na2WO4/TiO2 catalyst at different temperatures in the presence of air. b) Low magnification measured at 500 °C; high magnification at point in the circle in (b) measured at c) 500 °C, d) 700°, e) 900 °C.

Figure 2 b–e also shows in situ the high‐angle annular dark field (HAADF)‐STEM images that were recorded at different temperatures (500–900 °C) in the presence of air. The sample pretreated at 900 °C exhibited a TiO2 surface that was uniformly covered with a Na2WO4 film, as previously proposed by Lunsford et al. using ion scattering spectroscopy.8 At 700 °C or higher, the salt‐covered surface becomes blurry as the z‐contrast is weakened in the images (Figure 2 d), which is indicative of salt melting. At 900 °C, the surface layer seems to disappear and the facet feature of the support surface (TiO2) becomes more evident, owing to the absence of lattice of the melted salt, as shown in Figure 2 e. Nevertheless, the energy dispersive spectroscopy (EDS) image confirmed the homogeneous distribution of Na and W species after the 800 °C treatment (Supporting Information, Figure S6).

The surface states of the Na2WO4/TiO2 catalyst under different gas atmospheres (O2+H2O or O2+CH4) at 800 °C were investigated using the AP‐XPS system using a flowing reaction cell built by the Tao group.9 Photoemission features of subshell electrons of surface atoms can be collected when the catalyst is heated up to 850 °C in a flowing gas. The Na 1s, O 1s, and W 4f features of the catalyst in UHV and different gas atmospheres at different temperatures are shown in Figure 3. The binding energies of these spectra were calibrated to Au 4f7/2 of Au foil, which was used to support the Na2WO4/TiO2 catalyst particles (Supporting Information, Figure S7). As shown in Figure 3, the O 1s peak of the catalyst surface in UHV at 20 °C was located at 530.6 eV, which is consistent with the reported value.10 The W 4f7/2 peak can be identified even though W 4f5/2 overlaps with Ti 2p, as shown in Figure 3 c and f. In UHV, W 4f7/2 was located at 36.0 eV, which is in good agreement with the literature.10 The peak intensity of Na 1s was located at 1072.0 eV, which corresponds to Na 1s of Na2WO4 and is consistent with the reported value.11 To explore the surface state of Na2WO4 in the OH radical generation condition at 800 °C, a mixture of O2 and H2O was introduced into the flowing reaction cell for AP‐XPS.10 The presence of free‐state O2 and H2O gases was confirmed based on the observation of the photoemission features of the O 1s peaks (pink and blue line in Figure 3 b and red line in Figure 3 e) because they appear at a relatively higher binding energy than those of adsorbed species by a few eV. The peak positions at 539.1 and 535.8 eV were assigned to O2 and H2O in the gas phase, respectively, and are consistent with the references.12 The results provide strong evidence for the existence of gas environment around the catalyst. Intriguingly, the photoemission feature of Na 1s of the catalyst in a mixture of O2 and H2O at 800 °C upshifted by 1.0 eV, in contrast to that in a mixture of O2 and H2O at 20 °C (Figure 3 d). This shift does not result from surface charging because there is no shift in W 4f7/2. This upshift suggests a definite change in the chemical environment of the Na cations after being heated from 20 to 800 °C in a mixture of O2 and H2O. The high‐binding energy peak at 1073.2 eV for Na 1s at 800 °C was assigned to peroxide species (that is, Na2O2), which is consistent with the observed binding energy of Na 1s of Na2O2 reported previously.13

Figure 3

Photoemission features of Na 1s, O 1s, and W 4f for the catalyst collected at UHV, 20 °C (black line), 66 Pa O2 and 66 Pa H2O at 800 °C (red line), 66 Pa O2 and 66 Pa CH4 at 20 °C (blue line), and 66 Pa O2 and 66 Pa CH4 at 800 °C (pink line).

To check the authentic chemical state of the Na2WO4 catalyst during OCM catalysis at 800 °C, O2 and CH4 were mixed and then introduced to the reaction cell containing the Na2WO4 catalyst for AP‐XPS. At room temperature in a mixture of O2 and CH4, the photoemission features of Na 1s, O 1s, and W 4f (blue lines in Figure 3 a–c) are the same as those in UHV at 20 °C (black lines in Figures 3 d–f), except for the observation of an O 1s peak of gaseous O2. By heating the sample to 800 °C by an infrared laser beam and maintaining the catalyst at 800 °C in a mixture of O2 and CH4 during AP‐XPS data acquisition,10 Na 1s with a binding energy at 1073.1 eV was observed during catalysis. Notably, the photoemission feature of Na 1s of the surface of the catalyst during catalysis (pink line in Figure 3 a) is very similar to that observed at 800 °C in a mixture of O2 and H2O (red line in Figure 3 d). Clearly, the AP‐XPS studies of nominal catalyst Na2WO4 supported on TiO2 during OCM at 800 °C uncovered that the authentic surface phase of the Na species during catalysis was in fact Na2O2. Therefore, this result was due to the quasi‐equilibrated generation of Na2O2 in the presence of O2 under the studied conditions.13

The formation of Na2O2 is consistent with the quantitative analysis of the photoemission features collected under (O2+H2O) and (O2+CH4) atmospheres. As shown in Figure 4 a (data are normalized to the ratio in UHV at 20 °C), the relative ratios of Na 1s to O 1s at 800 °C was higher than those at 20 °C. This result is consistent with the higher Na to O stoichiometric ratio in Na2O2 (1:1) than Na2WO4 (2:1). The peaks centered at 36.3 eV (W 4f7/2) and at 38.0 eV (W 4f5/2 and Ti 3p) were deconvoluted to obtain the relative area ratios of Na 1s and W 4f7/2, and the results are plotted in Figure 4 b (data are normalized to the ratio in UHV at 20 °C). Notably, the Na 1s/W 4f7/2 area ratios under (O2+H2O) or (O2+CH4) were relatively increased from 20 to 800 °C. These increases suggest that the molten salt film became richer in sodium cations on the outer surface and lean in W atoms on the topmost surface layers. These quantitative analyses of the surface composition of nominal Na2WO4 obtained during catalysis suggest the formation of Na2O2 layers on top of the surface, which blocks the photoelectrons of W 4f of Na2WO4 from escaping to travel to gas phase or UHV. Therefore, the Na/W ratio increased once the Na2O2 layer was formed. Thus, we can conclude that Na2O2 formed on the surface of Na2WO4 during catalysis and under the reaction conditions at 800 °C.

Figure 4

Area ratios of a) Na 1s/O 1s and b) Na 1s/W 4f7/2 of the catalyst surface at 800 °C in a mixture of 66 Pa O2 and 66 Pa H2O; and at 20 and 800 °C in a mixture of 66 Pa CH4 and 66 Pa O2, relative to the ratios obtained at 20 °C in UHV.

With regard to the OCM selectivity, the active site has been extensively discussed for Mn/Na2WO4/SiO2 (or supported on other oxides).1, 2, 9, 11, 14 In our previous study,1d we confirmed that Mn and W are not essential components for the selective OCM reaction. However, H2O was identified as a key reactant (wet condition). In contrast, the supported Na2WO4 is a poor catalyst for the activation of CH4. Mn is effective for mildly burning CH4 to generate H2O from a CH4/O2 mixture (dry condition), which is taken over by the OH radical pathway at high conversions (leading to the wet condition). WO4 anions are most likely effective at immobilizing Na cations, which otherwise would sublime easily (NaOH: m.p. 318 °C, b.p. 1388 °C, Na2WO4: m.p. 698 °C). The support material should be inert and accommodate the Na2WO4 liquid film on its surface at high temperature,8 and the topmost layer becomes Na2O2, which was observed by AP‐XPS.

In conclusion, in situ spectroscopic and microscopic techniques were employed for identifying the authentic species of the Na‐based catalyst at 800 °C under OCM relevant conditions. The catalyst was active for generation of OH radicals from O2 and H2O at high temperature, as measured by OH radical selective LIF spectroscopy. The molten salt state of Na2WO4 on oxides at high temperature was confirmed using in situ XRD measurements and STEM images performed at high temperature. The in situ studies using AP‐XPS clearly uncovered the formation of Na2O2 under reaction conditions at high temperature (ca. 800 °C), which is considered to be responsible for OH radical formation. The specific binding energy of Na 1 s and the increased Na/W atomic ratio during catalysis at 800 °C revealed a surface that is richer in newly formed Na2O2. This study provides an excellent demonstration of integration of in situ electron spectroscopic and in situ electron microscopic techniques in identifying authentic surface phase of a nominal catalyst during catalysis at high temperatures in a mixture consisting of all the reactants of a catalytic reaction.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.