Carbon Dioxide Valorization into Methane Using Samarium Oxide-Supported Monometallic and Bimetallic Catalysts

Abstract

Samarium oxide (Sm₂O₃) is a versatile surface for CO₂ and H₂ interaction and conversion. Samarium oxide-supported Ni, samarium oxide-supported Co-Ni, and samarium oxide-supported Ru-Ni catalysts were tested for CO₂ methanation and were characterized by X-ray diffraction, nitrogen physisorption, infrared spectroscopy, H₂-temperature programmed reduction, and X-ray photoelectron spectroscopy. Limited H₂ dissociation and widely available surface carbonate and formate species over 20 wt.% Ni, dispersed over Sm₂O₃, resulted in ~98% CH₄ selectivity. The low selectivity for CO could be due to the reforming reaction between CH₄ (methanation product) and CO₂. Co-impregnation of cobalt with nickel over Sm₂O₃ had high surface adsorbed oxygen and higher CO selectivity. On the other hand, co-impregnation of ruthenium and nickel over Sm₂O₃ led to more than one catalytic active site, carbonate species, lack of formate species, and 94% CH₄ selectivity. It indicated the following route of CH₄ synthesis over Ru-Ni/Sm₂O₃; carbonate → unstable formate → CO → CH₄.

1. Introduction

The global concentration of carbon dioxide (CO₂) has already reached dangerous levels, where island countries and temperature-sensitive animals are on the borderline of extinction due to global warming. It has affected the seasonal cycles and has brought natural disasters across the globe. If the scientific community did not find a way of decreasing CO₂ concentration in the nearby future, the worst days would be faced. Besides the strategy of reducing CO₂ emissions, scientific findings for the conversion of CO₂ into useful chemicals are also encouraged worldwide [1,2].

The catalytic conversion of CO₂ over metal oxide, doped metal oxide, and supported metal oxide has received the highest attention due to its catalytic potential, heterogeneity, and high-temperature sustainability. CO₂ and H₂ interaction over catalytic surface generates variety of products from hydrocarbon (olefin and paraffin) to oxygenates (CH₃OH, CH₃COOH) [3,4,5,6,7]. Formate is a surface-bonded carbon dioxide species, whereas hydrogen is easily dissociated over a metal surface such as iron, copper, silver, gold, platinum, palladium, rhodium, molybdenum, rhenium, or nickel [6,8,9,10]. Controlling catalytic surfaces becomes crucial to understanding product distribution. Formate species is not stable, except over Ni, and it is decomposed into CO under H₂ flow through reverse water–gas shift reaction. Copper oxide (CuO) showed just 13% CO₂ conversion with 10.4% CO yield and iron (III) oxide (Fe₂O₃) showed 8% CO₂ conversion with 6.3% CO yield [11], indicating that unsupported CuO and Fe₂O₃ were not effective catalysts. Thus, the concept of promoted and supported metal catalysts has come into the picture, where the catalytic efficiency is augmented in terms of the dispersion of catalytic active sites over proper support and the role of promoters in modifying the physiochemical properties of supported catalyst systems. The comprehensive reaction scheme over the supported metal catalyst is shown in Figure 1.

When a “supported active metal iron” catalyst is utilized, carbon monoxide is converted into metal carbide, switches to the chain growth mechanism, and gives C1 to higher hydrocarbon (concerned with olefin and paraffin; Route 1) [12]. For route 1, the potassium- and copper-promoted, Al₂O₃-supported iron (Fe-Cu-K-Al) catalyst has been studied extensively [8,13,14]. Over the Cu-supported catalyst, CO hydrogenation takes place in two ways: the formic acid route (Route 2) and the methanol route through formaldehyde and methoxide intermediate (Route 3) [15]. The zirconium-promoted CuO-ZnO-Al₂O₃ (1:1:1) catalyst has received wide attention for high CO₂ conversion and high CH₃OH selectivity [9]. The supported Ag and Au catalysts prefer route 3 [10]. Supported 5.0 wt.% Ru metal induces an eight electron reduction in intermediate CO into CH₄ (Route 4) [6]. The Al₂O₃-supported Ru catalyst had an atomic assembly with high dispersion of three dimensional Ru cluster. It showed about 55% CO₂ conversion with 55% CH₄ yield. Pd, Pt, Rh, Mo, and Re metals direct CO₂ and H₂ through routes 3 and 4, which lead to high selectivity to methanol and methane, respectively [10]. On the other hand, the supported Ni catalyst carries CO₂ in the CO_x-free catalytic route (CO₂ methanation route): CO₂→ carbonate → formate → –CH₂OH → –CH₃ → CH₄ (Route 5) [16]. This route is useful in generating CO_x-free CH₄. Incorporating Co in Ni led to a smaller lattice constant and higher hydrogenation activity of CO₂ [17,18]. Cobalt-based catalysts were also investigated greatly for CO₂ conversion into light hydrocarbons [19].

Al₂O₃-supported Ni catalysts showed 50% CO₂ conversion with 50% CH₄ yield [20], while 3.0 wt.% Zr-promotional addition inhibited migration of Ni into Al₂O₃ matrix or inhibited the formation of NiAl₂O₄, caused >60% CO₂ conversion and 70% CH₄ yield [21]. The activity of La₂O₃-supported Ni catalyst toward CO₂ methanation was worse because of the formation of stable lanthanum carbonate [La₂(CO₃)₃], which did not decompose into formate [20]. In addition, the SiO₂-supported Ni catalyst was not good for CO₂ methanation because of the collapse of the Si-O-Si structure by hydrolysis of water, sintering of Ni particles, and rapid coke deposition and deactivation of the catalyst [22]. Among supported catalyst systems, the Y₂O₃-supported Ni catalyst displayed the maximum CO₂ conversion (75%) with 75% CH₄ yield due to the presence of more basic sites and easily reducible NiO species. However, in terms of CH₄ selectivity, the ZrO₂-supported Ni catalyst resulted in the best (60% CO₂ conversion with 100% CH₄ selectivity) [23].

For such an important reaction, other supports should also be investigated continuously. For instance, Sm₂O₃ had various favorable surface properties for CO₂ and H₂ interaction. It had Ni-size optimizing capacity like Y₂O₃, higher Ni dispersion and a stronger metal-support interaction than Y₂O₃ [24,25]. Electrochemically reduced Sm(II) compound was found to be a powerful catalyst for CO₂ activation [26]. Sm₂O₃ was utilized in quantitative detection of CO₂ and CO [27]. Moreover, a 5 atom% Sm addition over the Ni-Zr catalyst induced a higher H₂ uptake [28]. Muroyama et al. showed that upon reduction, the Ni particle size distribution was the narrowest and their sizes were the smallest over Sm₂O₃ compared to other supports (Y₂O₃, Al₂O₃, La₂O₃, and ZrO₂) [20]. A 5.0 wt.% Ni loading over Sm₂O₃ achieved 100% CH₄ selectivity, but with ~20% CO₂ conversion at 300 °C [29]. Upon incorporation of a 5.0 wt.% Ba into the Sm₂O₃-supported Ni catalyst, the CH₄ selectivity remained 100% and CO₂ conversion was maintained at 40% for 28-h TOS at 300 °C. The CO-free, high-performance catalytic route of the 5 wt.%Ni-5 wt.%Ba/Sm₂O₃ catalyst was due to the enhanced oxygen mobility (causing more oxygen vacancy), increased CO₂ adsorption (causing the formation of more surface carbonate species), and enhanced hydrogenation of carbonate into formate species.

Inspired by its properties in favor of CO₂ and H₂ interaction, samarium oxide-supported Ni, samarium oxide-supported Co-Ni, and samarium oxide-supported Ru-Ni catalysts were prepared by the impregnation and co-impregnation method and were tested for CO₂ methanation. Catalysts were characterized by X-ray diffraction (XRD), surface area and porosity, infrared spectroscopy, H₂-temperature programmed reduction (H₂-TPR), and X-ray photoelectron spectroscopy (XPS). Fine-tuning between surface properties and characterization results was found. This thorough study would help to understand the surface behavior of Sm₂O₃-supported metal catalysts toward CO₂ methanation reaction.

2. Results

2.1. Characterization Results

Energy Dispersive X-ray (EDX) elemental analysis of 10Ni/Sm₂O₃, 20Ni/Sm₂O₃, 40Ni/Sm₂O₃, 10Ni-10Co/Sm₂O₃ and 10Ni-10Ru/Sm₂O₃ catalysts is shown in Figure S1. EDX elemental analysis showed that with increasing 10 wt.% to 40 wt.% Ni loading, surface atomic composition of Ni was increasing. The 10Ni-10Co/Sm₂O₃ catalyst showed that the surface composition of Ni over the 10Ni-10Co/Sm₂O₃ catalyst was minimum, whereas surface composition of Ni was maximum at 10Ni-10Ru/Sm₂O₃ catalyst. The X-ray diffraction results of the catalyst samples are shown in Figure 2 and diffraction peaks are matched with Joint Committee on Powder Diffraction Standards (JCPDS) reference card number. All catalysts showed low intensity diffraction peaks. The support had diffraction peaks for the Sm₂O₃ phases at 2θ = 27.68°, 28.11°, 29.04°, 29.74°, 30.64°, 31.09°, 31.86°, 32.61°, 38.59°, 40.51°, 46.73°, 50.13°, 51.44°, 53.24°, 54.70°, 55.72° (JCPDS reference card number 00-013-0244 and 00-042-1461). Upon the 10.0 wt.% Ni addition, the intensity for the diffraction peaks for the Sm₂O₃ phases decreased significantly. However, the peaks of the cubic NiO phase were not detected, indicating that up to 10 wt.% Ni loading over the Sm₂O₃, NiO was well-dispersed (Figure 2A). NiO dispersion and the reduction in the peak intensity of Sm₂O₃ in the 10Ni/Sm₂O₃ catalyst suggested the close interaction of NiO with Sm₂O₃ support. The same type of finding was observed over the Y₂O₃-supported Ni catalyst [30]. Upon increasing the NiO loading to 20.0 wt.%, the cubic NiO phase appeared at 2θ = 37.24°, 43.28°, 62.75° (JCPDS reference card number 01-073-1523) (Figure 2B). Furthermore, increasing the NiO loading to 40 wt.% led only to the observation of the diffraction peaks of Sm₂O₃ without detecting those diffraction peaks of the cubic NiO phase (Figure 2C). The diffraction patterns of 40Ni/Sm₂O₃ and 10Ni/Sm₂O₃ were comparable. However, the bimetallic catalyst of 10 wt.%Co-10 wt.%Ni, dispersed over Sm₂O₃ support, had a very similar diffraction pattern to that of 20Ni/Sm₂O₃ (Figure 2D). It indicated that in presence of CoO, the NiO crystalline phase was relatively more exposed (like 20Ni/Sm₂O₃). The bimetallic catalyst of 10 wt.%Ru-10 wt.%Ni, dispersed over Sm₂O₃ support, had no crystalline peaks for NiO, but it had diffraction peaks for the Sm₂O₃ phase and for the tetragonal RuO₂ phase (at 2θ = 35.02°; JCPDS reference card number 00-040-1290) (Figure 2E).

The N₂-adsorption isotherms, porosity distribution, BET surface area, BJH pore volume, and BJH pore diameter are depicted in Figure 3A,B and Figure S2. All samples showed type-IV isotherm with H3 hysteresis loop, which indicated non-rigid aggregates like-pore with unlimited adsorption at high p/p⁰ [31]. The Sm₂O₃-supported 10 wt.% Ni catalyst had the least surface area, and pore volume among the remaining supported catalysts. However, upon 20.0 wt.% Ni loading, the height of the hysteresis loop was amplified and the surface area, pore volume, and pore diameter were increased to four times (25.70l m²/g), six times (0.175 cc/g) and seven times (15.797 nm), respectively, implying that Ni did not only disperse over the catalyst surface, but also it was incorporated into the Sm₂O₃ framework and caused expansion of framework [29]. Upon 40.0 wt.% Ni impregnation, the height of the hysteresis loop was decreased sharply, and the surface area and pore volume were decreased sharply (to 12.7 m²/g and 0.061 cc/g, respectively) without affecting the pore diameter. It indicated the pore blocking by inclusions of NiO nanoparticles [32] upon higher loading. Co-impregnation of 10.0 wt.% Co and 10.0 wt.% Ni over Sm₂O₃ support was found to maximize the surface area and pore volume up to the highest limit (with respect to the rest catalysts). The surface area and pore volume of Co-Ni/Sm₂O₃ were enhanced more than 10 times without affecting the pore diameter (with respect to 10Ni/Sm₂O₃). The height of the hysteresis loop in the isotherm plot of the Co-Ni/Sm₂O₃ catalyst was also enhanced up to the highest extent. However, with a co-impregnation of 10.0 wt.% Ru and 10.0 wt.% Ni over Sm₂O₃ support, the surface parameters were dropped. In XRD, the tetragonal RuO₂ was detected over the Ru-Ni/Sm₂O₃ catalyst, indicating that a sharp drop of the surface parameters was due to the large deposition of tetragonal RuO₂ crystals into the pore of the catalyst. These phenomena are related mainly to the ionic radii of the employed metals, where in the case of Co ions, the ionic radius was equal to that of Ni ions, but in the case of Ru, it was bigger. These findings are coincident with our previous study on Y₂O₃ support [33]. It appeared that in the case of co-impregnation, the pore diameter of the catalyst was not modified and the pore diameter of Co-Ni/Sm₂O₃ and Ru-Ni/Sm₂O₃ and 10Ni/Sm₂O_s were within the range from 2.0 to 2.4 nm. These observations were quite informative and indicated the uptake capacity of the heteroatom in the Sm₂O₃ framework. If 10 wt.% Ni loading was taken as a standard, then further 10 wt.% loading of Ni on Sm₂O₃ caused the expansion of the entire framework. However, when 10 wt.% Co along with 10 wt.% Ni (total 10 wt.%Co-10 wt.% Ni) was loaded over Sm₂O₃, Co was incorporated in the Sm₂O₃ framework except on the pore opening of the surface. Thus, pore volume and surface area were increased, but not the pore size. During loading of 10 wt.% Ru along with 10 wt.% Ni over Sm₂O₃, RuO₂ particles were largely deposited in the pore, causing decrements in all surface parameters.

Infrared (IR) spectra of the catalyst samples are shown in Figure 3C and Figure S3. The IR band at ~1620 cm⁻¹ and 3429 cm⁻¹ could be attributed to the bending and stretching vibrations of the hydroxyl groups over the catalyst surface, respectively [34]. These two bands were present even in the bare support of Sm₂O₃ (Figure S3). Upon Ni loading over Sm₂O₃, various other prominent bands appeared. The transmittance peak was at 852 cm⁻¹ for CO₃²⁻ symmetric stretching, 1044 cm⁻¹ for the vibration of bidentate carbonate species, 1381 cm⁻¹ and 1500 cm⁻¹ for the vibration of unidentate carbonate species [35,36,37], 2850 cm⁻¹ for the stretching vibration of C-H (of coordinated formate species), and 2925 cm⁻¹ for the combination of the asymmetric stretching of –COO and the bending vibration of C–H bond (of bi-coordinated formate species) [38]. Interaction of CO₂ with samarium (through different CO₂ intermediates) was previously reported [39]. The IR spectra of the 20Ni/Sm₂O₃ catalyst could be specified by the intense spectral vibration of the unidentate carbonate species at 1500 cm⁻¹, whereas the Co-Ni/Sm₂O₃ catalyst had the most intense vibration for the bicarbonate species at 1044 cm⁻¹, and the unidentate carbonate species at 1381 cm⁻¹. It is noticeable that the Ru-Ni/Sm₂O₃ catalyst had no vibration bands for the formate species (~2850 cm⁻¹ and 2925 cm⁻¹).

The Sm₂O₃-supported 10 wt.% Ni catalyst had reduction peaks at 360 °C and 475 °C with total adsorption of 2.197 mmol/g of H₂. The earlier one was attributed to the reduction in NiO species, which interacted weakly with support, while the latter peak was attributed to the reduction in NiO species, which interacted strongly with the support [40]. For higher Ni loading (20.0 wt.%), the 20Ni/Sm₂O₃ catalyst had the highest H₂ consumption (7.28 mmol/g) among the other Ni-loaded Sm₂O₃ catalysts with a merge reduction peak in the temperature range 327–615 °C (the peak maximum ~400 °C). It indicated the presence of a high concentration of reducible NiO species, interacting with the support up to different extents. For 40.0 wt.% Ni loading, the 40Ni/Sm₂O₃ catalyst had the worst reducible profile. It had a negligible amount of reducible NiO species (0.244 mmol/g H₂ consumption). Accordingly, 40Ni/Sm₂O₃ had minimum reducibility. The reducibility profile of the bimetallic Co-Ni, dispersed over Sm₂O₃, had the maximum amount of H₂ consumption (8.827 mmol/g) among all catalysts. The Co-Ni/Sm₂O₃ catalyst had the highest intensity of reducible peak about the peak maxima at 400 °C for the reduction in NiO-interacted species, low-temperature overlapping peaks ~300 °C for the reduction in Co₃O₄ to CoO, and a high-temperature reduction peak ~500 °C for the reduction in CoO to Co [41,42]. However, the high-temperature reduction peak ~500 °C in the Co-Ni/Sm₂O₃ catalyst might be related to the combining contributions of reducible NiO and CoO species. The H₂-consumption profile of the bimetallic Ru-Ni, dispersed over Sm₂O₃, showed an additional reduction peak at a relatively lower temperature (~330 °C) for a partial reduction in dispersed RuO_x species on the support [43]. The total hydrogen consumption of the Ru-Ni/Sm₂O₃ catalyst in the H₂-temperature programmed reduction study was found 6.25 mmol/g.

There is a marked difference in the surface characteristics of both supported bimetallic catalysts. To understand the valance state and bonding character: Ni (2p), Sm (3d) and O (1s) XPS spectra of the Co-Ni/Sm₂O₃ and Ru-Ni/Sm₂O₃ catalysts were depicted in Figure 4A–D. The Ni2p spectra of both Sm₂O₃-supported bimetallic catalysts had mostly similar profiles. The absence of metallic Ni peak (about binding energy <854 eV and 869–870 eV) in Ni2p XPS spectra was confirmed in the Co-Ni/Sm₂O₃ and Ru-Ni/Sm₂O₃ catalysts [44,45]. The presence of Ni⁺² oxidation state in the form of NiO compound was confirmed by Ni2p_1/2 spectra ~871.8–873.8 eV and 879 eV, and Ni2p_3/2 spectra ~854.3–855.7 eV [46,47,48,49] (Figure 4A). Sm3d_3/2 and Sm3d_5/2 spectra at ~1110.25 eV and 1083.24 eV, respectively, indicated the presence of Sm⁺³ oxidation state (Sm₂O₃ compound) [50,51] (Figure 4C). The binding energy values of the O1s profiles indicated the overlapping contributions of oxygen from various oxide species [52]. The O1s spectra with binding energy at 531.2 eV and 530 eV were noticed in both the Co-Ni/Sm₂O₃ and Ru-Ni/Sm₂O₃ catalysts (Figure 4D). The binding energy at 531.2 eV was attributed to adsorbed oxygen [53], whereas the binding energy at 530 eV signified the oxygen in samarium oxide [51]. Both XPS peak intensities were much less in the Ru-Ni/Sm₂O₃ catalyst than to their corresponding ones in the Co-Ni/Sm₂O₃ catalyst, indicating the availability of “surface adsorbed oxygen” and oxygen in samarium oxide in the Co-Ni/Sm₂O₃ catalyst. To bring more focus on the valance state and bonding character of the second metal over the Sm₂O₃-supported Ni catalyst, the Co (2p) spectrum of the Co-Ni/Sm₂O₃ catalyst and the Ru (3d) spectrum of the Ru-Ni/Sm₂O₃ catalyst were also recorded, as depicted in Figure S4, respectively. Co2p XPS spectra did not show a peak for metallic Co at binding energy ~778 eV, but it showed a peak with a binding energy centered at 780 eV for 2p_3/2) for Co₃O₄ [42]. The Ru3d spectrum of the Ru-Ni/Sm₂O₃ catalyst showed a broad overlapping peak in the binding energy range 281.5–285 eV. It indicated the presence of Ru oxide species [54,55].

2.2. Catalytic Activity Result and Discussion

The reaction temperature suitable for CO₂ methanation was tested over the 10Ni/Sm₂O₃ catalyst and was depicted in Figure S5. It was found that the mole% of CH₄ formation from CO₂ and H₂ was negligible at 250 °C, notable at 300 °C, and prominent at 350 °C. At 350 °C temperature, the mole% of CO formation was null, which indicated that CH₄ formation was carried out without including CO in the intermediate reaction step. In infrared spectra, a peak for formate was observed in all catalysts except 10Ru-10Ni/Sm₂O₃. The literature report suggested stable formate formation over the Ni-based catalyst system, which further underwent successive hydrogenation to methane through the methoxy intermediate route [16]. Thus, 350 °C was kept standard for CO₂ methanation over our the Sm₂O₃-supported Ni catalyst system. Other research groups also found 350 °C as the suitable temperature for CO₂ methanation [56,57,58]. The catalytic activity of the xNi/Sm₂O₃ (x = 10, 20, 40) and M-Ni/Sm₂O₃ (M = Co, Ru) catalysts in the term of CO₂ conversion, CH₄ selectivity, and CO selectivity are shown in Figure 5. XRD did not detect metallic Ni or NiO phase over the Sm₂O₃-supported 10.0 wt.% Ni catalyst and H₂-TPR indicated the reduction peaks of NiO-interacted species. IR confirmed the presence of surface interacted CO₂-species such carbonate and formate over 10Ni/Sm₂O₃ surface. Overall, 10Ni/Sm₂O₃ had widely available surface-interacted CO₂-species and reducible NiO species (inducing H₂ dissociation and spillover). The CO₂ conversion remained constant at ~35% for 13 h, but constant selectivities for CO and CH₄ were attained after 100 min of reaction. Initially, the CO and CH₄ selectivities were comparative (47% and 52%), but after 100 min, the CO selectivity dropped to 5–6%, whereas the CH₄ selectivity jumped to 94–95%. Further, the constant CH₄ and CO selectivities were maintained for 13 h time-on-stream. This finding indicated that the 10Ni/Sm₂O₃ catalyst system needed some time for constant product selectivity.

For 20.0 wt.% loading of Ni over Sm₂O₃, the NiO phase appeared in the X-ray diffraction pattern. Assembly of NiO particles might sign toward less dispersion of catalytic active species and it might provide limitation in higher activity toward CO₂ methanation. The surface area of the 20Ni/Sm₂O₃ catalyst was four times that of 10Ni/Sm₂O₃ and unidentate carbonate intensity in infrared spectra was also amplified. It had formate species and the highest number of reducible NiO species among the rest of the Ni-loaded Sm₂O₃ catalysts. However, the largely extendible surface and high amount of reducible NiO species over 20Ni/Sm₂O₃ did not turn into higher CO₂ conversion, but it resulted in the highest CH₄ selectivity. The 20Ni/Sm₂O₃ catalyst showed 24–26% CO₂ conversion, ~97–98% CH₄ selectivity, and 3–2% CO selectivity over 13 h of reaction. It indicated that a large number of reducible NiO species over large catalytic surfaces did not grant the higher CO₂ conversion by H₂. Hydrogen dissociation and its spillover concisely depend on the size of Ni particles. The higher NiO crystallinity over the 20Ni/Sm₂O₃ catalyst than 10Ni/Sm₂O₃ caused the formation of a relatively larger size of metallic Ni upon reduction, which limited H₂ dissociation and thereby CO₂ conversion. However, the relatively higher density of unidentate carbonate in the infrared spectrum of the 20Ni/Sm₂O₃ catalyst indicated that limited H₂ dissociation, followed by CO₂ interaction over a largely extended surface, led to the formation of unidentate carbonate to methane via a stable formate pathway. So, 20Ni/Sm₂O₃ conveyed highest ~98% CH₄ selectivity. The 10Ni/Sm₂O₃ catalyst needed some time for constant product selectivity, but the 20Ni/Sm₂O₃ catalyst gave constant product selectivity from the beginning. The surface mechanism over the samarium oxide-supported Ni system could be outlined at this stage. CO₂ was initially adsorbed as carbonate/bicarbonate species over the support of Sm₂O₃. The carbonate/bicarbonate species were hydrogenated to formate species. The formate species over Ni was stable and did not decompose into CO and H₂. Strong binding between Ni and the formate species guided the formate species to further hydrogenation to methane through methoxy intermediate (Route 5) [16]. The presence of a very small amount of CO over 20Ni/Sm₂O₃ indicated the initiation of further reaction of CH₄ with CO₂ (dry reforming of methane) or “CH₄ with H₂O” (stream reforming of methane) over the surface as metallic Ni also catalyzes CH₄ decomposition. For 40.0 wt.% impregnation of Ni over Sm₂O₃, the catalyst crystallinity was comparable to 10Ni/Sm₂O₃, but the surface area was decreased to the half and the amount of reducible NiO species over the catalyst surface was fallen suddenly. The shortage of reducible NiO species turned into a lower number of active sites (metallic Ni) after reduction. It caused just 8% CO₂ conversion. It appeared that, at such a high Ni loading, the formate species were majorly destabilized and decomposed into CO. Furthermore, the catalyst surface, bearing fewer active sites, was not so capable to push CO toward methanation. Overall, due to the destabilization of the formate-hydrogenation pathway over 40Ni/Sm₂O₃, two third of CO₂ was converted to CO and the remaining one third to CH₄ (75% CO selectivity and 25% CH₄ selectivity) during 180 min time-on-stream. To bring more focus to the reaction mechanism, Co-impregnation of Co-Ni and Ru-Ni over Sm₂O₃ support was carried out and was tested for CO₂ methanation reaction. When 10.0 wt.% Co and 10.0 wt.% Ni were co-impregnated over Sm₂O₃, the crystallinity of the catalyst was comparable to that of the 20Ni/Sm₂O₃, whereas it had the highest surface area and pore volume among all catalysts. Including reducible nickel oxide and cobalt oxide, the 10Co-10Ni/Sm₂O₃ catalyst had the highest intensity of reducible oxide over the catalyst surface. In addition, it had an adequate amount of formate species and the highest intensity of bicarbonate/carbonate species over the catalyst surface. Altogether, the 10Co-10Ni/Sm₂O₃ catalyst showed initially high CO₂ conversion (41%), which slowly and steadily dropped to 22% at the end of 13 h. Thus, diluting the active Ni sites by Co caused inferior catalytic activity in terms of CO₂ conversion. In EDX elemental analysis also, minimum Ni distribution over the surface was found in the case of the 10Co-10Ni/Sm₂O₃ catalyst (than the remaining catalysts). The product selectivity pattern over the 10Co-10Ni/Sm₂O₃ was very similar to the 10Ni/Sm₂O₃, namely, the 10Co-10Ni/Sm₂O₃ catalyst also needed some time for constant product selectivity. After 100 min, the 10Co-10Ni/Sm₂O₃ catalyst had attained 88–89% CH₄ selectivity and 12–11% CO selectivity. The relatively higher selectivity of CO over the 10Co-10Ni/Sm₂O₃ catalyst could be due to persuading reverse water–gas shift reaction over the cobalt-based catalyst [59]. When 10.0 wt.% Ru and 10.0 wt.% Ni were co-impregnated over Sm₂O₃, the cubic NiO phase was not found, but the tetragonal RuO₂ phase appeared. It had a good amount of reducible NiO and RuO_x species. It had an easier reducibility profile than other catalysts. The 10Ru-10Ni/Sm₂O₃ had five times less surface area and less surface adsorbed oxygen than the 10Co-10Ni/Sm₂O₃. Maximum surface distribution of Ni was confirmed over the 10Ru-10Ni/Sm₂O₃ catalyst by EDX elemental analysis (than rest catalyst). Among all catalyst systems, 10Ru-10Ni/Sm₂O₃ produced peaks for the carbonate species, but no peaks for the the formate species, indicating that the unstable formate species decomposed into CO. Further CO methanation might lead to methane formation over the 10Ru-10Ni/Sm₂O₃ catalyst. The previous literature also argued the mechanism of CO₂ methanation in the same line [6]. The 10Ru-10Ni/Sm₂O₃ catalyst had a higher catalytic activity (31–33% CO₂ conversion, 93–94% CH₄ selectivity, and ~6% CO selectivity constantly from the beginning to 13 h) than 10Co-10Ni/Sm₂O₃ due to the presence of more than one active site as Ni and Ru metals over the catalyst surface.

3. Materials and Methods

3.1. Materials

All chemicals were purchased and were used as received: Sm₂O₃ (99.9% AR from Loba Chemie, Mumbai, India), Ni(NO₃)₂·6H₂O (Acros Organics-Thermo Fisher Scientific, Geel, Belgium 99.9%), Co(NO₃)₂·6H₂O (Acros Organics-Thermo Fisher Scientific, 99.9%), and RuCl₃·xH₂O (Acros Organics-Thermo Fisher Scientific, 99.9%).

3.2. Catalyst Preparation

Samarium oxide-supported Ni, samarium oxide-supported Co-Ni, and samarium oxide-supported Ru-Ni catalysts were prepared by wet impregnation and co-impregnation methods. For supported Ni catalysts, an appropriate amount of Ni(NO₃)₂·6H₂O (equivalent to 10, 20, or 40 wt.% of Ni metal) was dissolved in water. For supported M-Ni (M = Co or Ru) bimetallic catalysts, an appropriate amount of metal precursor (equivalent to 10.0 wt.% Co or Ru) and Ni (NO₃)₂·6H₂O (equivalent to 10.0 wt.% Ni) were dissolved in water. In all, 3.0 g pre-calcined Sm₂O₃ (at 500 °C for five hours) was added as support to these aqueous solutions at 80 °C under slow stirring until complete evaporation for four hours. The slurries were kept overnight at 110 °C in the oven and were then calcined at 500 °C under air for three hours. The prepared catalysts were abbreviated as: 10Ni/Sm₂O₃, 20Ni/Sm₂O₃, 40Ni/Sm₂O₃, Co-Ni/Sm₂O₃, and Ru-Ni/Sm₂O₃.

3.3. Catalysts Characterization

The surface area and porosity of catalyst samples were analyzed by NOVA 3200 apparatus (Quantachrome Corporation, Boynton Beach, FL, USA). Crystallinity and phase identification of the catalysts were studied by X-ray diffraction analysis (XRD) by using a Shimadzu XD-1 diffractometer (Kyoto, Japan). The presence of functional groups over the catalytic surface was studied by infrared spectroscopy by using a Nicolet Is-10 model, USA. The surface reducibility of the catalyst was studied under temperature programmed reduction under H₂-stream (H₂-TPR) by using a Chem BET 3000 (Quantachrome equipment, USA). The valance states of the elements and the binding energy of the bond electron were determined by X-ray photoelectron spectroscopy (XPS) by Themo Fisher Scientific, USA. The elemental mapping of catalyst samples was carried out by Energy Dispersive X-ray Spectroscopy instrument (Zeiss Smart EDX, Jena, Germany) equipped with a field-emission scanning electron microscope (FE-SEM; FEI Quanta FEG 250, Hilsboro, OR, USA). The detailed description of instrument and characterization procedure is given in the Supporting Information S6.

3.4. Catalyst Activity Test

Carbon dioxide methanation was carried out in a fixed-bed, tubular quartz silica reactor (length = 50 cm, Inner diameter = 13 mm), equipped with an electric furnace, having a programmable temperature controller. A total of 2.0 g of fresh catalyst was diluted with silicon carbide (to obtain a 5.0 cm bed height and 5 cm³ volume of catalyst bed) and was packed in the middle of the reactor. The temperature of the catalyst bed was monitored by a K-type thermocouple, placed in the center of the catalyst bed. Prior to the reaction, the catalyst sample was reduced in situ with 30 mL/min H₂ flow at atmospheric pressure for two hours at 600 °C. After the reduction, the catalysts were cooled down, and CO₂:H₂:Ar (1:4:5 volume ratio) gas feed was passed through the activated catalyst at GHSV of 6000 ccg⁻¹h⁻¹. Afterwards, the temperature was increased to 350 °C and the methanation reaction was carried out for 13 h. Gaseous reaction products were analyzed online by a quantitative gas analysis system (HIDEN ANALYTICAL QGA, England, UK). The mol % of H₂, CO₂, CH₄, and CO was detected by gas analyzer by using a matrix equation to correct the overlapped detection values from the m/z of 28 (CO), 44 (CO₂), 2 (H₂), and 16 (CH₄). The carbon dioxide conversion (${\mathrm{X}}_{{\mathrm{CO}}_2}$), methane selectivity (${\mathrm{S}}_{{\mathrm{CH}}_4}$), and carbon monoxide selectivity (${\mathrm{S}}_{\mathrm{CO}}$) were calculated using Equations (1)–(3) [21]:

$${\mathrm{X}}_{{\mathrm{CO}}_2}=\frac{{\mathrm{CO}}_{2,\mathrm{in}}-{\mathrm{CO}}_{2,\mathrm{out}}}{{\mathrm{CO}}_{2,\mathrm{in}}}\times 100$$

(1)

$${\mathrm{S}}_{{\mathrm{CH}}_4}=\frac{{\mathrm{CH}}_{4,\mathrm{out}}}{{\mathrm{CH}}_{4,\mathrm{out}}+{\mathrm{CO}}_{\mathrm{out}}}\times 100$$

(2)

$${\mathrm{S}}_{\mathrm{CO}}=\frac{{\mathrm{CO}}_{\mathrm{out}}}{{\mathrm{CH}}_{4,\mathrm{out}}+{\mathrm{CO}}_{\mathrm{out}}}\times 100$$

(3)

4. Conclusions

A Sm₂O₃-based Ni-supported catalyst was found excellent for CO₂ methanation at 350 °C reaction temperature. The 10Ni/Sm₂O₃ had the highest CO₂ conversion (35%) but needed some time (100 min) for constant high CH₄ selectivity (94–95%) over the 13 h reaction period. The 20Ni/Sm₂O₃ had larger NiO crystallite, which limited the H₂ dissociation and thereafter CO₂ conversion to 24–26%. However, the slower H₂ dissociation and interaction of CO₂ over a largely expanded surface showed the highest ~98% CH₄ selectivity from the beginning of the reaction. The formate species were stable at the Ni surface, as observed in the infrared spectra, which underwent further hydrogenation to methane. Despite the presence of the largest surface area and highest reducible metal oxide over the surface, 10Co-10Ni/Sm₂O₃ showed inferior catalytic performance because of diluting Ni sites by Co. On the other hand, diluting the Ni sites by Ru over Sm₂O₃ (10Ru-10Ni/Sm₂O₃ catalyst) caused again good catalytic activity. It had RuO₂ crystallite, very less surface parameters but easy reducibility. The absence of format species over the 10Ru-10Ni/Sm₂O₃ catalyst surface indicated an eight-electron reduction of intermediate CO into CH₄. In the mean of constant catalytic activity, the 10Ru-10Ni/Sm₂O₃ catalyst was found the best, as it showed 31–33% CO₂ conversion, and 93–94% CH₄ selectivity constantly from the beginning to the 13 h time-on-stream.