Ni Nanoparticles on Reducible Metal Oxides (Sm2O3, CeO2, ZnO) as Catalysts for CO2 Methanation

The activity of reducible metal oxide Sm2O3, CeO2, and ZnO as Ni nanoparticles support was investigated for CO2 methanation reaction. CO2 methanation was carried out between 200 °C to 450 °C with the optimum catalytic activity was observed at 450 °C. The reducibility of the catalysts has been comparatively studied using H2Temperature Reduction Temperature (TPR) method. The H2-TPR analysis also elucidated the formation of surface oxygen vacancies at temperature above 600 °C for 5Ni/Sm2O3 and 5Ni/CeO2. The Sm2O3 showed superior activity than CeO2 presumably due to the transition of the crystalline phases under reducing environment. However, the formation of NiZn alloy in 5Ni/ZnO reduced the ability of Ni to catalyze methanation reaction. A highly dispersed Ni on Sm2O3 created a large metal/support interfacial interaction to give 69% of CO2 conversion with 100% selectivity at 450 °C. The 5Ni/Sm2O3 exhibited superior catalytic performances with an apparent phase transition from cubic to a mixture of cubic and monoclinic phases over a long reaction, presumably responsible for the enhanced conversion after 10 h of reaction. Copyright © 2021 by Authors, Published by BCREC Group. This is an open access article under the CC BY-SA License (https://creativecommons.org/licenses/by-sa/4.0).


Introduction
Utilization of CO2 as carbon feedstock via the conversion to methane has been identified as a sustainable technology to produce renewable energy [1]. CO2 reduced under H2 gas at atmospheric pressure to form methane in the presence of metal catalysts. In order to overcome the thermodynamic of the reaction, investigation is carried out to synthesis catalysts with high methanation potential [2]. Modification of metal oxide with metal nanoparticles increases the hy-flexibility of the catalyst surface during CO2 methanation which mostly was associated with the presence of surface defect [12]. Depending on the degree of reducibility, reducible metal oxide, such as CeO2, decreases the CO coverage on Ru metal center, to ensure the dissociation of H2 can occur for the generation of methane [13]. However, on a non-reducible metal oxide such Al2O3, the catalytic activity is largely correlated with the basicity of the catalyst [13].
The activity of CeO2, Sm2O3, and ZnO as reducible oxides support will be investigated in improving Ni activity for CO2 methanation. CeO2 represented as among the mostly investigated metal oxide for CO2 methanation reaction [14,15]. The oxidation states of cerium are interchangeable between 4+ and 3+, with the Ce 3+ ions signified the presence of oxygen vacancies and the migration of sub-surface oxygen [16]. Rare-earth metal oxide Sm2O3 was reported as promoter for CO2 methanation that exhibited high activity, stability and coke resistance when incorporated with Ni/SBA-15 catalysts [17]. As a support, the Sm2O3 showed high activity when loaded with 39% of Ni to give ~71% of CO2 conversion to methane at 300 °C [18]. ZnO is another interesting reducible metal oxide as catalyst support, however its reducibility is associated with the ability to transfer the charge to fill the d-band of metal for the formation of metal alloy [19,20]. ZnO was reported as active support for CO2 hydrogenation to methanol when impregnated with Pd due to the formation of Pd-Zn alloy that can stabilize formate intermediate [21][22][23]. In this study, the catalytic activity of Ni supported on reducible metal oxides, CeO2, Sm2O3, and ZnO were investigated for CO2 methanation at different reaction temperatures. Reducible metal oxide has the ability to release electron under reducing environment for dissociation of C−O bond, however the reducible support can also change the structural and electronic properties of Ni. Due to the reducibility of these support, the presence of continuous hydrogen gas as reducing agent in CO2 methanation also modified the structure of the catalysts. This study elucidate the changes on Ni nanoparticles activity when exists in metallic and alloy states upon impregnation on reducible support. Comparative H2−TPR analysis is used to provide understanding on the dispersion and the reducibility of Ni, meanwhile X-ray Diffraction (XRD), N2 adsorption, and Transmission Electron Microscopy (TEM) are employed for structural and morphology analysis.

Preparation of the Catalysts
All the catalysts were prepared using wetness impregnation method. Ni(NO3)2 at 5% of weight loading was dissolved in 10 mL of water. The amount of Ni loading was kept at 5% for all catalysts. The Ni(NO3)2 solution was added into metal oxide supports, CeO2, Sm2O3, and ZnO, and stirred for 3 h at 60 °C to ensure thorough dispersion. The catalysts were dried at 200 °C for 2 h and then calcined at 500 °C for 3 h in air with a heating rate of 10 °C/min.

Characterization of the Catalysts
X-ray powder diffraction (Shimadzu XRD-700 X-ray Diffractometer) was carried out to analyze the crystalline structure of the catalysts by utilizing Cu-K radiation at  = 1.54 Å, with operating conditions of 60 kV and 50 mA. The scanning angle range, 2θ, was 10-70° with a scan rate of 2°/min. The crystalline structure size was calculated using Scherrer's equation (Equation (1)). (1) where D is the crystallite size (nm),  is the wavelength of x-ray (1.54 Å), θ is the peak position (radians), k is Scherrer constant and  is the FWHM (radians).
Surface area, average pore volume and pore diameter of the catalysts were determined by N2 adsorption at 77 K (Micromeritics ASAP 2020). The samples were degassed under carbon degassing conditions for 2 h at 300 °C. Brunauer-Emmett-Teller (BET) equation was used to calculate the catalysts' surface area and Barret-Joyner-Halenda (BJH) equation for pore volume and diameter calculation.
H2 Temperature-Programmed Reduction (ChemBET Pulsar TPR/TPD, equipped with TCD) was performed to study the catalysts' reducibility of NiO species on the catalyst surface. Catalysts undergone pre-treatment in N2 stream with flow of 75 mL/min at 100 °C for 1 h and later cooled down to 40 °C in the same gas stream. Catalysts were later heated to 900 °C with a heating rate of 10 °C/min in a 5%H2/N2 gas mixture with a flow rate of 75 ml/min.

Catalytic Testing
CO2 methanation reaction was conducted in a stainless-steel plug flow reactor at atmospheric pressure and reaction temperature cos k D    = ranging from 200 to 450 °C. 0.5 g of catalyst was placed in a tubular fixed bed reactor and reduced in situ with H2 gas (99.9%) with flow rate of 20 mL/min at 450 °C for 3 h. The catalysts were cooled down to room temperature and CO2/H2/N2 (15%/60%/25%) gases at 20 ml/min were flowed in the reactor for catalytic reaction. The gas effluent from reactor was analyzed by GC (Shimadzu GC-2014) after 1h. Reaction temperature was increased by 50 °C every hour and data were recorded. CO2 conversion (Equation (2)), CH4 (Equation (3)) and CO selectivity (Equation (4)), and CH4 yield (Equation (5)) were calculated as follows: (2) Where xCO2(%) is the CO2 conversion, SCH4(%) is the CH4 selectivity, SCO(%) is the CO selectivity, YCH4(%) is the CH4 yield, and m is the number of moles of the gases. Figure 1 showed the XRD analysis of 5Ni/CeO2, 5Ni/Sm2O3, and 5Ni/ZnO after calcination in air at 500 °C and reduction in H2 for 3 h at 400 °C. For 5Ni/CeO2, the characteristic peaks of CeO2 were observed at 2θ = 28.6°, 33 In the 5Ni/CeO2, no peak associated with NiO was observed that may be due to the formation of highly disperse Ni nanoparticles [24]. The additional peak was observed at 2θ = 23.9° on 5Ni/ZnO catalyst which indicated the formation of Ni-Zn alloy.

N2 Adsorption-Desorption Analysis
N2 analysis was carried out to investigate the textural properties of the catalysts. Figure  2 presented the N2 adsorption-desorption isotherms and pore size distributions of 5Ni/Sm2O3, 5Ni/ZnO, and 5Ni/CeO2 after calcination at 500 °C for 3 h. Based on the IUPAC isotherm classification, all the catalysts displayed Type III isotherm, which is a typical feature of a non-porous material [25]. 5Ni/Sm2O3 and 5Ni/CeO2 showed the presence of H3-hysteresis loop at high P/Po that suggested the formation porous structure [26]. The Type H3 loop at high P/Po was produced from the adsorption of multi-layer metastability and delayed capillary condensation which occurs in slit-shaped macropores or non-rigid aggregates of plate-like particles resulted from a low level of pore curvature and flexibility of the aggregate structure [26]. Table 1 summarized the textural properties of the catalysts calculated using BET equation. The 5Ni/CeO2 showed the highest surface area (15 m 2 /g) followed with 5Ni/ZnO (13 m 2 /g) and 5Ni/Sm2O3 (9 m 2 /g). The plot of pore size distribution showed the presence of pores at ~21 nm on 5Ni/TiO2 catalysts  100%   presumably due to the formation of interparticle mesopores.

H2−TPR Analysis
Temperature Programmed Reduction (TPR) was carried out on the catalysts to investigate the reducibility and the dispersion of Ni particles ( Figure 3). In general, the reducibility of Ni catalyst was determined based on the position of the reduction peak at 250-500 °C due to reduction of NiO to metallic Ni. 5Ni/ZnO showed the first reduction peak centered at 470 °C and was accompanied with a broad reduction peak between 550-700 °C. The first reduction peak was associated with the reduction of -type NiO species, generally reduced at lowtemperature region (300-500 °C) [27,28]. Meanwhile, the broad reduction peak observed between 550-700 °C was due to the segregation of ZnO to Ni-Zn alloy [29]. The presence of Ni-Zn alloy was evidenced from the XRD analysis ( Figure 1). The 5Ni/CeO2 showed the first reduction peak appeared at 350 °C and a small peak at 290 °C. The multiple H2 consumption peaks indicated the differences between interaction of NiO species with the metal oxide support [30]. Increasing the temperature to 900 °C showed a broad reduction peak started at 600 °C and was continuously increased at temperature above 900 °C. At temperature above 600 °C, the oxygen in ceria surface was removed when reacted with H2, releasing water and creating surface oxygen vacancies [30]. The vacancies were formed from the reduction of Ce 4+ to Ce 3+ under H2 flow. The 5Ni/Sm2O3 showed a broad reduction profile of NiO with the peak centered at 411 °C. Sm2O3 was reduced at temperature above 700 °C due to reduction of Sm2O3 to SmO in the absence of Ni [31]. It is suggested that the reduction peak of Sm2O3 was reduced to temperature below 700 °C due to the H2 spillover from Ni. The reduction peak of NiO was overlapped with the reduction peak of Sm2O3 thus produced a broad TPR profile on Ni/Sm2O3. The Sm2O3 was reduced at temperature below 700 °C, significantly lower than the reduction of CeO2 which was occurred at temperature above 700 °C. The results implied the tendency of Sm2O3 to generate oxygen vacancies under H2 flow was higher than CeO2.

TEM-EDX Analysis
TEM analysis was carried out to determine the morphology of 5Ni/Sm2O3 catalyst and to acquire the average diameter of Ni particles. 5Ni/Sm2O3 was calcined in air and reduced in hydrogen gas prior to the TEM analysis. Figure  4 showed the representative TEM image of 5Ni/Sm2O3 with its corresponding EDX analysis. Ni was well distributed on Sm2O3 surface evidenced from the formation of spherical shapes particles as shown in a small red circle. The particle size of Ni nanoparticles was analyzed using ImageJ and the average diameter was determined at ~2 nm. The EDX elemental analysis of 5Ni/Sm2O3 further confirmed the presence of Ni at ~5% loading. The elemental mapping also indicated the homogeneous dispersion of Ni on Sm2O3.

Catalytic Activity
CO2 methanation was carried out 200-450 °C at atmospheric pressure (Equation (6)). Prior to the reaction, the catalyst was reduced insitu at 450 °C under H2 flow for 3 h. Table 1 Figure 4. HRTEM-EDX analysis of 5Ni/Sm2O3 catalyst. summarized the physicochemical properties and the catalytic performances of all the catalysts at 450 °C. The 5Ni/CeO2 showed 67.6% of CO2 conversion with 100% selectivity towards methane. When Ni was impregnated onto Sm2O3, the conversion of CO2 was slightly increased to 69.3% at 100% selectivity to methane. If the reaction was carried out using 5Ni/ZnO, the CO2 conversion was observed at 26.6% with 99.5% selectivity to methane. Unlike CeO2 and Sm2O3, ZnO showed a relatively low catalytic conversion presumably due to the formation of Ni-Zn alloy on the catalysts that reduced the activity of Ni towards methanation reaction. In CO2 methanation reaction, it was reported that Ni must be in its zero oxidation states to catalyze hydrogen dissociation [32]. The use of ZnO as support, formed Ni-Zn alloy that changed the electronic properties of Ni. The charge transfer from Ni to Zn transformed the metallic Ni with zero oxidation state to be partially positive charged NiZn [33]. The NiZn promoted the formation of CO2 hydrogenation to C2 products due to further hydrogenation of formate intermediates [34]. The formation of Ni5Ga3 alloy was also observed on Ga2O3 support, responsible for enhanced methanol formation from CO2 [35]. There is also possibility that NiZn alloy further catalyzed the resulting methane into CO via dry reforming reaction [36]. Therefore, the efficiency of Ni alloy in 5Ni/ZnO catalyst for methanation reaction was reduced to give only 26.6% of CO2 conversion at 450 °C, with CO was also determined as the result of reverse water gas shift reaction (Equation (7)).
CO2 (g) + H2 (g) ↔ CO (g) + 2H2O (g) ΔH298K = −41.2 kJ/mol CO2 methanation was carried out further at different reaction temperatures to study the performance of Ni on reducible metal supports at low temperatures ( Figure 5, Table 2). All the catalysts were seen to be inactive at both 200 °C and 250 °C but methane formation was starting to observe at 300 °C, in exception for 5Ni/ZnO catalyst. Evidently, 5Ni/Sm2O3 showed the highest activity at low temperatures with the CO2 conversion at 300 °C was measured at 20.1%. The reactivity was then followed by 5Ni/CeO2 and 5Ni/ZnO. At the variation of reaction temperatures, both 5Ni/CeO2 and 5Ni/Sm2O3 displayed 100% CH4 selectivity. However, for 5Ni/ZnO, as reaction temperature increased over 350 °C, 5Ni/ZnO only demonstrated 73.7% selectivity towards methane, with CO was produced as a side product. The production of methane was enhanced at 450 °C, to give 87% selectivity with 26.6% conversion of CO2.

Stability Studies of Ni/Sm2O3
To further investigate the stability of 5Ni/Sm2O3, catalytic reaction was carried out for 28 h at 300 °C. Figure 6(a) showed the CO2 conversion was measured at 20% at the first 60  min into the reaction. The conversion was significantly improved after 2 h to give ~32% conversion and the conversion was stable for the next 10 h of reaction. It is interesting to note that CO2 conversion was improved to ~51% at 15 h of reaction and the conversion was maintained until 28 h. The XRD of the catalysts after 5 h and 28 h were analyzed to determine the structural changes of 5Ni/Sm2O3 ( Figure  6(b)). XRD analysis of 5Ni/Sm2O3 following 5 h reaction showed no significant differences on the structure of the catalysts. All the peaks corresponded to the Sm2O3 in the cubic crystal structure were visible, according to JCPDS 043-1029. When the XRD analysis was carried out on 5Ni/Sm2O3 after 28 h catalytic reaction, the new peak corresponded to Ni metal at 2θ = 47° was gained significant enhancement in intensity suggesting the NiO nanoparticles was sintered to form large Ni crystallites. Two new peaks corresponded to Sm2O3 with monoclinic phase were observed at 27° and 39.8° (JCPDS card no 042-1464) [37]. The transition of Sm2O3 phase from cubic to monoclinic implied the increase of catalytic activity after 10 h of reaction maybe due to the formation of more active Sm2O3 sites due to the prolonged reduction under CO2/H2/N2 mixed gas during reaction. Cubic phase was characterized by a larger distance between the ions with isometric and higher entropy, meanwhile monoclinic structure had a shorter distance between the ions with a minimal bond energy [38]. Although the transition between cubic to monoclinic phase in Sm2O3 was reported at high temperatures ~1000 °C [39], there was a possibility that impregnation with Ni and annealing under H2 gas reduced the temperature for the phase transformation. HRTEM analysis of Ni/Sm2O3 further confirmed the presence of cubic and monoclinic phases (Figure 4). The distances between planes were measured to be 0.318 nm which corresponded to the (222) planes of cubic phase Sm2O3 [31], meanwhile the monoclinic phase was identified with the shorten lattice spacing at 0.27 nm due to shorter distance between the ions with a minimal bond energy of the monoclinic phase [38]. The activity of Ni nanoparticles was significantly enhanced when Sm2O3 was used as support mainly due to the increase of Ni dispersion on Sm2O3. Despite the surface area of Sm2O3 was significantly lower than ZnO and CeO2, the number of Ni active sites on 5Ni/Sm2O3 were determined at 18.5 mmol/g, which were higher than 5Ni/CeO2 at 6.8 mmol/g and 5Ni/ZnO at 5.6 mmol/g. Apart from the dispersion, the reducibility of metal nanoparticles also played an important role during the catalytic reaction Surface oxygen vacancies improved the adsorption and dissociation of CO2. The energy required for the formation of vacancies in reducible oxides was relatively low compared to the non-reducible metal oxide. The activity of non-reducible oxide support was generally correlated with its basicity such as Al2O3 and zeolite [40]. The dissociated hydrogen on Ni, was spilled over to metal oxide support that can be used to form oxygen vacancies. The Sm2O3 and CeO2 with 4+ oxidation states reduced to 3+ under hydrogen created oxygen vacancies to facilitate CO2 adsorption. The proximity of Ni/support interfacial contact in 5Ni/Sm2O3 compared to 5Ni/CeO2, due to the formation of highly dispersed metal nanoparticles, assisted the formation of oxygen vacancies. The highly dispersed Ni nanoparticles created a larger metal/support interaction, which was identified as the active sites for CO2 methanation. Prolonged catalytic reaction further reduced the Sm2O3 via hydrogen spill over, which transformed the cubic phase into monoclinic phase. The combination of highly exposed metallic Ni and the continuous formation of oxygen vacancies on Sm2O3 created Ni-VOx-Sm sites which were essential for CO2 methanation [41].

Conclusion
The activity of 5Ni/Sm2O3 was investigated as catalysts for methanation reaction and compared to another reducible oxide supports CeO2 and ZnO. The importance of metallic Ni to catalyze methanation was further consolidated by the non-reactivity of 5Ni/ZnO catalysts due to the formation of Ni-Zn alloy. Ni was highly dispersed on Sm2O3 thus created a high interfacial contact for efficient formation of oxygen vacancies during catalytic reaction. The enhancement of CO2 conversion after 10 h of reaction implied the formation of oxygen vacancies were continuously formed during the reaction, resulting in the transition of Sm2O3 crystal phase from cubic to monoclinic.