Nanosheet-Like Ho2O3 and Sr-Ho2O3 Catalysts for Oxidative Coupling of Methane

In this work, Ho2O3 nanosheets were synthesized by a hydrothermal method. A series of Sr-modified Ho2O3 nanosheets (Sr-Ho2O3-NS) with a Sr/Ho molar ratio between 0.02 and 0.06 were prepared via an impregnation method. These catalysts were characterized by several techniques such as XRD, N2 adsorption, SEM, TEM, XPS, O2-TPD (temperature-programmed desorption), and CO2-TPD, and they were studied with respect to their performances in the oxidative coupling of methane (OCM). In contrast to Ho2O3 nanoparticles, Ho2O3 nanosheets display greater CH4 conversion and C2-C3 selectivity, which could be related to the preferentially exposed (222) facet on the surface of the latter catalyst. The incorporation of small amounts of Sr into Ho2O3 nanosheets leads to a higher ratio of (O− + O2−)/O2− as well as an enhanced amount of chemisorbed oxygen species and moderate basic sites, which in turn improves the OCM performance. The optimal catalytic behavior is achievable on the 0.04Sr-Ho2O3-NS catalyst with a Sr/Ho molar ratio of 0.04, which gives a 24.0% conversion of CH4 with 56.7% selectivity to C2-C3 at 650 °C. The C2-C3 yield is well correlated with the amount of moderate basic sites present on the catalysts.


Introduction
The present energy crisis, owing to the dwindling petroleum resource and its nonrenewable feature, must be solved as soon as possible. Methane, as a major component of natural gas, coal-bed gas, and shale gas, is attracting increasing attention as a clean fossil energy and a raw material for producing chemicals. Methane conversion can proceed via nondirect and direct routes [1][2][3][4][5][6][7]. The oxidative coupling of methane (OCM) to ethylene and ethane (C 2 hydrocarbons) is an indispensable way that has great prospect in the direct conversion of methane into value-added products [5][6][7]. Ethylene is one of the most important parts in petrochemical fields. Ethylene and its derivatives are associated closely with over 70% of petroleum chemicals. Since 1982 Keller et al. [8] first reported the OCM technology, it has attracted more and more attention in catalysis, chemical industry, and oil and gas fields because of its potential economic value and application prospect.
Up to now, several types of catalysts have been tried for OCM reaction [5,7]. It is widely accepted that Li/MgO and Mn-Na 2 WO 4 /SiO 2 are the most promising catalysts for application, and they have been widely researched [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23]. In general, higher reaction temperatures (above 800 • C) are required for both kinds of catalysts to achieve the optimal OCM performance. Wang et al. [24] reduced the reaction temperature from 800 to 720 • C by doping Ti into Mn-Na 2 WO 4 /SiO 2 , and they acquired 26% CH 4 conversion with 76% C 2 -C 3 selectivity. More studies are now shifted to a low-temperature OCM process. Nanoscale rare earth oxide-based catalysts with special morphologies (nanorods and nanosheets), such as CeO 2 , La 2 O 3 , Sm 2 O 3 , and Er 2 O 3 , were found to effectively catalyze the OCM process at lower temperatures (500-650 • C) [25][26][27][28][29][30]. However, the C 2 selectivity and yield still need to improve. Ho 2 O 3 was demonstrated to have potential application for high-k dielectric material [31], photocatalysts [32,33], and energy-storage electrodes [34]. There are few reports dealing with the use of Ho 2 O 3 as a catalyst in the OCM process [35]. In this work, we synthesized Ho 2 O 3 and Sr-Ho 2 O 3 nanosheets to develop a new type of efficient catalyst system for a low-temperature OCM reaction. The catalytic performances of these catalysts were related to their characterization results.

Catalytic Performances
We first compared catalytic behaviors of Ho 2 O 3 nanosheets and nanoparticles for the OCM reaction to investigate the morphology effect of the Ho 2 O 3 catalysts. As shown in Figure 1, with the reaction temperature raised from 600 to 750 • C, the CH 4 conversion increases progressively, while the selectivity to C 2 -C 3 (C 2 H 4 , C 2 H 6 , C 3 H 6 , and C 3 H 8 ) rises more evidently. Because of this, the C 2 -C 3 yield also improves with the temperature. It is evident that the OCM performance is better over Ho 2 O 3 -NS nanosheets than Ho 2 O 3 -NP nanoparticles. For instance, the CH 4 conversion, C 2 -C 3 selectivity, and yield over Ho 2 O 3 -NS at 700 • C are 22.1%, 43.8%, and 9.7%, respectively, and those over Ho 2 O 3 -NP are 17.8%, 29.7%, and 5.3%, respectively. The shape effects of La 2 O 3 , Sm 2 O 3 , and Er 2 O 3 catalysts on the OCM reaction were also reported by Zhu et al. and in our recent work [25][26][27][28]. We then tested the catalytic performances of Sr-modified Ho 2 O 3 nanosheets (Sr-Ho 2 O 3 -NS) to investigate the impact of Sr modification on Ho 2 O 3 -NS nanosheets in the OCM reaction. As the Sr/Ho molar ratio is increased from 0 to 0.06, the CH 4 conversion, C 2 -C 3 selectivity, and the yield first rise and then diminish ( Figure 2). The 0.04Sr-Ho 2 O 3 -NS catalyst with a Sr/Ho ratio of 0.04 exhibits the best OCM performance. This catalyst yields a 24.0% CH 4 conversion and 56.7% C 2 -C 3 selectivity at 650 • C. Even at a low temperature of 600 • C, a 21.6% CH 4 conversion and 46.9% C 2 -C 3 selectivity can be achieved. Notably, the 0.04Sr-Ho 2 O 3 -NS catalyst performs better than Ho 2 O 3 -NS (20.0% CH 4 conversion and 32.8% C 2 -C 3 selectivity at 650 • C). The typical product distribution over the Ho 2 O 3 -NS and Sr-Ho 2 O 3 -NS catalysts at 650 • C is listed in Table 1. According to the literature [5,36,37], the proposed reaction mechanism of methane transformation to ethane and ethylene is shown in Scheme S1. The interaction of the adsorbed CH 4 and O 2 generates methyl radicals (CH 3 ). The coupling of CH 3 radicals generates C 2 H 6 , followed by the dehydrogenation to C 2 H 4 . Propane and propylene can be formed in the similar way, as illustrated in Scheme S1. The results shown in Figure 2 indicate that the introduction of appropriate amounts of Sr to Ho 2 O 3 -NS is beneficial for the OCM reaction. It is noteworthy that the 0.06Sr-Ho 2 O 3 -NS catalyst shows a bit lower CH 4 conversion and C 2 -C 3 selectivity than 0.04Sr-Ho 2 O 3 -NS at 750 • C and 700 • C, which could be attributed to the blockage of some active sites upon introducing excessive Sr. However, much worse OCM performance was observed for the former catalyst than the latter one at 650 • C and 600 • C, particularly at 600 • C. It was also reported that there were the optimal contents of Na and Li for Na-CaO and for the Li-promoted Bi-Mn oxide catalysts employed in the OCM reaction [38,39].  We selected the best 0.04Sr-Ho 2 O 3 -NS catalyst to investigate the lifetime for the OCM reaction, which was evaluated at 650 • C. It is clear from the results presented in Figure 3 that the 0.04Sr-Ho 2 O 3 -NS catalyst shows good stability during 60 h of the reaction, maintaining ca. 24% CH 4 conversion with 57% C 2 -C 3 selectivity. We compared the catalytic performances of our catalyst 0.04Sr-Ho 2 O 3 -NS and three reference catalysts, i.e., 0.04Sr-La 2 O 3 nanofibers, 0.04Sr-CeO 2 nanowires, and 3% Li/MgO. As shown in Figure S1, both 0.04Sr-CeO 2 and 3% Li/MgO are inactive at 600 and 650 • C. Our catalyst 0.04Sr-Ho 2 O 3 displays a higher methane conversion and C 2 -C 3 selectivity than the three reference catalysts at 600-750 • C.  Table 2 shows that in comparison with Ho 2 O 3 -NS, the Sr-modified Ho 2 O 3 nanosheets display greater lattice parameters (1.0571-1.0589 nm vs. 1.0561 nm). This observation implies that Sr is doped into the lattice of Ho 2 O 3 , considering that Sr 2+ has larger ionic radius than Ho 3+ (0.118 nm vs. 0.090 nm). The doping of Sr into the lattice of La 2 O 3 via an impregnation method, followed by calcination at high temperatures, was also displayed in former studies [29,40].   The SEM images of Ho 2 O 3 -NS and 0.04Sr-Ho 2 O 3 -NS are shown in Figure 5. Clearly, both catalysts display a nanosheet morphology. The morphology of nanoparticles with an irregular shape can be found for Ho 2 O 3 -NP ( Figure S2). The average width and thickness of Ho 2 O 3 -NS are 771 nm and 81.9 nm, respectively. Ho 2 O 3 -NP has a mean size of 17.5 nm. In addition, 0.04Sr-Ho 2 O 3 -NS has a similar size to Ho 2 O 3 -NS ( Table 2), suggesting that the introduction of a small amount of Sr to Ho 2 O 3 -NS has a little impact on the catalyst size. As illustrated in Figure 6   The Brunauer-Emmett-Teller (BET) specific surface areas of the Ho 2 O 3 -based catalysts are given in Table 2. All catalysts give low surface areas between 6.1 and 7.9 m 2 /g, which is preferred for the OCM reaction. In contrast to Ho 2 O 3 -NP, Ho 2 O 3 -NS has a lower surface area (6.1 vs. 7.9 m 2 /g). The incorporation of small amounts of Sr into Ho 2 O 3 -NS slightly increases the surface area. Figure S3 shows the XPS spectra of O1s on Ho 2 O 3 -NP, Ho 2 O 3 -NS, and Sr-Ho 2 O 3 -NS catalysts. The XPS spectra were deconvoluted into four peaks corresponding to four different oxygen species. The XPS data are listed in Table 2. Oxygen species located at~529.3,~530.7,~531.8, and~532.6 eV are O 2− (lattice oxygen), O − (peroxide ions), CO 3 2− (carbonate), and O 2 − (superoxide ions), respectively [25,[41][42][43][44]. It is generally accepted that the surface electrophilic oxygen species O − and O 2 − are beneficial for C 2 selectivity in the OCM reaction, while the lattice oxygen O 2− is responsible for the deep oxidation of CH 4 in forming CO and CO 2 [25,26,29,43,45] (Figures 1 and 2). This observation is in accord with the results reported for the OCM reaction catalyzed by the La 2 O 3 -based catalysts [26,29,40,45].

XPS and IR
Based on theoretical studies, Sayle and co-workers have disclosed that the energy required to generate oxygen vacancies over CeO 2 for different crystal planes follows the order of (110) < (310) < (111) [46]. In other words, oxygen vacancies are easier to form on the (110) plane of CeO 2 . The interaction between O 2 and oxygen vacancies generates the surface electrophilic oxygen species such as O − and O 2 − . Therefore, we think that the higher (O − + O 2 − )/O 2− ratio observed over Ho 2 O 3 -NS than Ho 2 O 3 -NP could be associated with the predominantly exposed (222) planes over the former catalyst. It was found that the OCM process was a structure-sensitive reaction [25,30,44]. Figure 7 compares the FTIR spectra of 0.04Sr-Ho 2 O 3 -NS and 0.06Sr-Ho 2 O 3 -NS catalysts after the OCM reaction at 600 • C for 1 h. Two bands that appeared at 1637 and 3445 cm −1 are assigned to the bending and stretching vibrations of the O−H groups in H 2 O [47]. The bands appearing at 858 and 1442 cm −1 correspond to the bending and asymmetric stretching vibrations of C−O in CO 3 2− [48,49], and they stemmed from the combination of catalysts with CO 2 produced during the OCM reaction. Clearly, the spent 0.06Sr-Ho 2 O 3 -NS catalyst displays a stronger intensity of CO 3 2− vibrations than the spent 0.04Sr-Ho 2 O 3 -NS, suggesting that the amount of carbonate is higher over the former catalyst than the latter one. Thus, the worse OCM performance observed for the former catalyst than the latter one at 650 • C and 600 • C ( Figure 2) is due to the blockage of active sites by carbonate.

Temperature-Programmed Desorption (TPD) of O 2 and CO 2
To further understand the activation of oxygen over the catalysts, which plays an important role in the OCM process, the TPD of O 2 was performed. Figure 8 shows that there are two desorption peaks of oxygen from the surfaces of catalysts. The low-temperature peaks located at 85-137 • C are assigned to the desorption of molecular oxygen species (i.e., loosely bounded surface oxygen), and the high-temperature peaks located at 263-426 • C are ascribed to the desorption of chemisorbed oxygen species, which could be O − , O 2 − , and O 2− [40,44,50] that stemmed from the interaction of O 2 with the Ho 2 O 3 -based catalysts. It is generally believed that the chemisorbed oxygen species benefit CH 4 activation and C 2 selectivity in the OCM process [25,44,50,51]. Table 3 shows that a greater number of chemisorbed oxygen species are achieved over Ho 2 O 3 -NS than Ho 2 O 3 -NP (15.9 vs. 12.8 µmol/g), which is responsible for the higher CH 4 conversion and the C 2 -C 3 yield observed for the former catalyst than the latter one. The incorporation of small amounts of Sr into Ho 2 O 3 -NS leads to an increase in the quantity of chemisorbed oxygen species (19.8-24.2 µmol/g), indicative of enhancing the oxygen activation. The largest quantity of chemisorbed oxygen species are achieved over 0.04Sr-Ho 2 O 3 -NS. Moreover, introducing Sr into Ho 2 O 3 -NS weakens the interaction between oxygen and the Sr-Ho 2 O 3 -NS catalysts, since the desorption peaks of chemisorbed oxygen species shift to low temperatures (from 344 • C to 263-309 • C). The doping of low-valence Sr into high-valence Ho 2 O 3 can increase the quantity of oxygen vacancies [51][52][53], which promotes the activation toward oxygen, thus leading to an increased amount of chemisorbed oxygen species. As a result, the Sr-Ho 2 O 3 -NS catalysts exhibit better OCM performances than Ho 2 O 3 -NS. The optimal CH 4 conversion and C 2 -C 3 yield are obtained on the 0.04Sr-Ho 2 O 3 -NS catalyst with a Sr/Ho molar ratio of 0.04.  In addition to oxygen activation, the basicity of the catalysts is a key factor influencing the OCM reaction [54,55]. The surface basicity of the Ho 2 O 3 -NP, Ho 2 O 3 -NS, and Sr-Ho 2 O 3 -NS catalysts was measured by CO 2 -TPD, and the results are presented in Figure 9 and Table 3. Figure 9 shows that there are two desorption peaks of CO 2 from the surfaces of the Ho 2 O 3 -NP, Ho 2 O 3 -NS, and 0.02Sr-Ho 2 O 3 -NS catalysts, while there are three CO 2 desorption peaks for the 0.04Sr-Ho 2 O 3 -NS and 0.06Sr-Ho 2 O 3 -NS catalysts. It was reported that the surface basic sites were associated closely with the O − , O 2 − and O 2− oxygen species [10,51,54,56]. Based on the peak temperature of CO 2 desorption, the peaks that are below 200 • C, between 200 and 600 • C, and higher than 600 • C correspond to basic sites with weak, moderate, and strong strength, respectively [29,40,51]. Table 3 shows that the surfaces of all catalysts are dominated by moderate basic sites. Ho 2 O 3 -NS has a greater number of weak and moderate basic sites than Ho 2 O 3 -NP. The modification of Ho 2 O 3 -NS with Sr brings about an increase in the number of weak and moderate basic sites, and the number of moderate basic sites is increased to the maximum on 0.04Sr-Ho 2 O 3 -NS. As evidenced in Figure 10, the C 2 -C 3 yield obtained at 700 • C correlates well with the number of moderate basic sites present on the Ho 2 O 3 -based catalysts. This finding is in accordance with some previous reports that the surface basic sites with moderate strength are more favorable for the C 2 product formation in the OCM process [25,26,44,55,[57][58][59][60].

Catalyst Preparation
Ho 2 O 3 nanosheets (labelled as Ho 2 O 3 -NS) were synthesized by a hydrothermal method reported by Lee and co-workers [61]. Typically, 3.79 g of HoCl 3 •6H 2 O was dissolved in 100 mL deionized water, and 1 mL of aqueous ammonia (25-28 wt%) was then added dropwise to the above solution under stirring. The obtained suspension was transferred into a Teflon-lined stainless autoclave, which was placed in an oven setting at 200 • C for 12 h. Ho 2 O 3 nanoparticles (named as Ho 2 O 3 -NP) were prepared via a conventional precipitate method; 3.0 mL of aqueous ammonia (25-28 wt%) was added dropwise to 100 mL of 0.1 M HoCl 3 solution under stirring. All the resulting precipitates were fully washed with deionized water, followed by drying at 80 • C in an oven for 12 h. Finally, the dried Ho(OH) 3 samples were calcined at 750 • C in air for 4 h in a muffle to obtain Ho 2 O 3 nanosheets and nanoparticles.
Sr-modified Ho 2 O 3 nanosheets were synthesized by an incipient wetness impregnation method. In a typical procedure, different amounts of Sr(NO 3 ) 2 were dissolved in deionized water, and then a certain amount of dried Ho(OH) 3 nanosheets were added.
After drying under an infrared lamp, the sample was dried at 80 • C in an oven for 12 h, followed by calcination at 750 • C in air for 4 h in a muffle. The resulting catalysts were labelled as xSr-Ho 2 O 3 -NS, where x represents the Sr/Ho molar ratio (x = 0.02, 0.04, and 0.06, respectively).
For comparison, 0.04Sr-La 2 O 3 nanofibers were prepared according to the literature [29]. Ce(OH) 3 nanowires were prepared according to the literature [30]. The 0.04Sr-CeO 2 nanowires were prepared in the same way as our Sr-modified Ho 2 O 3 nanosheets. The 3% Li/MgO was prepared according to the literature [62]. The calcination temperature for three reference catalysts was 750 • C. The Sr/La or Sr/Ce molar ratio was 0.04. The content of Li in the catalyst was 3 wt.%.

Characterization of the Catalyst
X-ray diffraction (XRD) patterns were recorded on a D2 PHASER X-ray diffractometer using nickel-filtered Cu Kα radiation at 30 kV and 10 mA (Brucker, Madison, WI, USA). The BET surface areas of the samples were analyzed by N 2 adsorption at −196 • C using a Micromeritics Tristar 3000 instrument (Micromeritics, Atlanta, GA, USA). X-ray photoelectron spectroscopy (XPS) measurements were performed on a Perkin-Elmer PHI 5000C spectrometer (Perkin-Elmer, Waltham, MA, USA). All binding energy values were calibrated using the C 1s peak at 284.6 eV. The surface basicity was measured by the temperature programmed desorption of CO 2 (CO 2 -TPD) using a Micromeritics AutoChem II analyzer (Micromeritics, Atlanta, GA, USA); 0.2 g of sample was preheated at 750 • C for 1 h under He (30 mL/min), then cooled down to 80 • C. CO 2 adsorption was conducted at this temperature, followed by purging with He (30 mL/min) for 2 h. The temperature was then raised from 80 to 950 • C at a ramping rate of 10 • C/min. O 2 temperature programmed desorption (O 2 -TPD) was performed on the same instrument; 0.2 g of sample was preheated at 750 • C for 1 h under He (30 mL/min), then cooled down to 50 • C. O 2 adsorption was conducted at this temperature, followed by purging with He (30 mL/min) for 2 h. The temperature was then raised from 50 to 800 • C at a ramping rate of 10 • C/min. The desorbed CO 2 and O 2 were detected with a thermal conductivity detector (TCD). Field-emission scanning electron microscopy (FESEM) images were taken using a Hitachi S-4800 instrument (Hitachi, Tokyo, Japan). Transmission electron microscopy (TEM) images were recorded on an FEI Tecnai G 2 F20 S-TWIN instrument (FEI, Hillsboro, OR, USA). Fourier transform infrared (FTIR) spectra were recorded on a Nicolet Avatar 360 spectrometer (Nicolet, Madison, WI, USA). 30 mg of the spent catalyst and 300 mg of KBr were first mixed uniformly; 40 mg of the mixture was then pressed into a self-supporting disk.

Catalytic Tests
The oxidative coupling of methane reaction was performed with a fixed-bed flow reactor at atmospheric pressure, with a quartz tube internal diameter of 6 mm. Here, 0.2 g of the catalyst (40-60 mesh) was placed in the middle of the reactor, with the downstream of the catalyst fixed with quartz wool. The catalytic performance was evaluated using a mixture of methane and oxygen (CH 4 /O 2 = 4/1 molar ratio) as feed gas, with a total flow rate of 60 mL/min, which results in a gas hourly space velocity (GHSV) of 18,000 mL/(g•h). Before the reaction, the catalyst was pretreated at 750 • C in Ar (30 mL/min) for 1 h. The reaction temperature (actually the catalyst bed temperature) was monitored by a thermocouple placed in the middle of the catalyst bed. The reaction products were analyzed by an on-line GC equipped with a TCD and a 2-m Shincarbon ST packed column (for separation of H 2 , O 2 , CO, CH 4 , and CO 2 ) and by another on-line GC equipped with a FID and a 50-m PoraPLOT Q capillary column (for the separation of CH 4 , C 2 H 4 , C 2 H 6 , C 3 H 6 , and C 3 H 8 ). Prior to the analysis using TCD, the products were passed through a cold trap at −3 • C to remove most of water generated during the reaction. The CH 4 conversion and C 2 -C 3 selectivity were calculated using the standard normalization method based on carbon atom balance. The typical GC chromatograms showing the reaction products are given in Figure S4.

Conclusions
In this work, we developed Ho 2 O 3 and Sr-Ho 2 O 3 nanosheet catalysts for low-temperature OCM reaction. The HR-TEM images revealed that Ho 2 O 3 and Sr-Ho 2 O 3 nanosheets predominantly expose (222) facets. The Ho 2 O 3 nanosheets outperformed Ho 2 O 3 nanoparticles, which could be associated with the preferentially exposed (222) facet on the surface of the former catalyst. The ratio of (O − + O 2 − )/O 2− , the amount of chemisorbed oxygen species, and the moderate basic sites were enhanced upon the addition of small amounts of Sr to Ho 2 O 3 nanosheets, as demonstrated by XPS, O 2 -TPD, and CO 2 -TPD, respectively. This, in turn, resulted in an improved catalytic performance. The optimal 0.04Sr-Ho 2 O 3 nanosheets with a Sr/Ho molar ratio of 0.04 afforded a methane conversion of 24.0% with 56.7% C 2 -C 3 selectivity at 650 • C. Moreover, this catalyst exhibited good stability in the OCM reaction for 60 h of time on stream. A good correlation between the C 2 -C 3 yield and amount of moderate basic sites on the catalysts was established.