CO and CO₂ Methanation over CeO₂-Supported Cobalt Catalysts

Abstract

CO₂ methanation is a promising reaction for utilizing CO₂ using hydrogen generated by renewable energy. In this study, CO and CO₂ methanation were examined over ceria-supported cobalt catalysts with low cobalt contents. The catalysts were prepared using a wet impregnation and co-precipitation method and pretreated at different temperatures. These preparation variables affected the catalytic performance as well as the physicochemical properties. These properties were characterized using various techniques including N₂ physisorption, X-ray diffraction, H₂ chemisorption, temperature-programmed reduction with H₂, and temperature-programmed desorption after CO₂ chemisorption. Among the prepared catalysts, the ceria-supported cobalt catalyst that was prepared using a wet impregnation method calcined in air at 500 °C, and reduced in H₂ at 500 °C, showed the best catalytic performance. It is closely related to the large catalytically active surface area, large surface area, and large number of basic sites. The in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) study revealed the presence of carbonate, bicarbonate, formate, and CO on metallic cobalt.

1. Introduction

CO₂ capture, utilization, and storage have become increasingly important in terms of mitigating global climate change caused by escalation in CO₂ atmospheric concentrations since the first industrial revolution. CO₂ utilization can be categorized into physical and chemical methods. The latter is preferred over the former as CO₂ can be converted into other chemicals through chemical transformations [1]. Among the various chemical utilization methods of CO₂, CO₂ hydrogenation has attracted significant attention as it can produce various platform chemicals, including methane, methanol, and formic acid. It can also utilize hydrogen generated from water with a water electrolysis system powered by renewable energy [2,3,4].

In particular, CO₂ methanation is considered an energy storage alternative to address variability in renewable energy as the produced methane can be transported through a gas grid. Furthermore, CO₂ methanation (Reaction 1) is thermodynamically more feasible than other CO₂ hydrogenation reactions (e.g., reverse water-gas-shift reaction (Reaction 2) and methanol synthesis (Reaction 3)) [5,6,7].

Reaction 1:

$${\mathrm{CO}}_2\left(\mathrm{g}\right)+4{\mathrm{H}}_2\left(\mathrm{g}\right) \to {\mathrm{CH}}_4\left(\mathrm{g}\right)+2{\mathrm{H}}_2\mathrm{O}\left(\mathrm{g}\right) \Delta {\mathrm{G}}_{298 \mathrm{K}}^{\mathrm{o}}=-113.2 \mathrm{kJ}/\mathrm{mol}, \Delta {\mathrm{H}}_{298 \mathrm{K}}^{\mathrm{o}}=-165 \mathrm{kJ}/\mathrm{mol}$$

Reaction 2:

$${\mathrm{CO}}_2\left(\mathrm{g}\right)+{\mathrm{H}}_2\left(\mathrm{g}\right) \to \mathrm{CO}\left(\mathrm{g}\right)+{\mathrm{H}}_2\mathrm{O}\left(\mathrm{g}\right) \Delta {\mathrm{G}}_{298 \mathrm{K}}^{\mathrm{o}}=28.6 \mathrm{kJ}/\mathrm{mol}, \Delta {\mathrm{H}}_{298 \mathrm{K}}^{\mathrm{o}}=41.2 \mathrm{kJ}/\mathrm{mol}$$

Reaction 3:

$${\mathrm{CO}}_2\left(\mathrm{g}\right)+3{\mathrm{H}}_2\left(\mathrm{g}\right) \to {\mathrm{CH}}_3\mathrm{OH}\left(\mathrm{g}\right)+{\mathrm{H}}_2\mathrm{O}\left(\mathrm{g}\right) \Delta {\mathrm{G}}_{298 \mathrm{K}}^{\mathrm{o}}=3.5 \mathrm{kJ}/\mathrm{mol}, \Delta {\mathrm{H}}_{298 \mathrm{K}}^{\mathrm{o}}=-49.5 \mathrm{kJ}/\mathrm{mol}$$

Since CO₂ methanation is an exothermic and thermodynamically limited reaction, active catalysts at low temperatures should be developed. To date, some metal catalysts, including Ru [8,9,10,11], Ni [12,13,14,15,16,17,18,19], Co [20,21,22,23,24,25,26], and Fe [23,27,28,29] have been reported to be active in this reaction. However, Ni-based catalysts have been more thoroughly investigated due to their high activity and comparatively low cost [3,4]. In addition to the active metal, the nature of the support also affects the catalytic properties, such as the morphology of the active metal, metal dispersion, and reducibility of active metal/metal oxides [1,2,7]. Several metal oxide supports have been reported for CO₂ methanation reactions, such as Al₂O₃ [30,31,32], CeO₂ [33,34,35], SiO₂ [36,37,38], ZrO₂ [39,40,41,42], TiO₂ [43,44,45], La₂O₃ [45,46,47,48], Y₂O₃ [31,49], and Sm₂O₃ [50,51]. Among them, CeO₂ is a well-known support due to its redox properties and high surface oxygen vacancies. In particular, two oxidation states of cerium (e.g., Ce³⁺ and Ce⁴⁺) are interconvertible between CeO₂ and CeO_2−x under oxidized and reduced environments to release oxygen vacancy [52,53,54,55]. This is favorable for the dissociation of CO₂ into CO and O species on the catalyst surface in the dissociative CO₂ methanation mechanisms [52,53,54]. The other representative CO₂ methanation mechanism is the associative one, in which the chemisorbed CO₂ (CO_{2 ads}) and chemisorbed H₂ (H_{2 ads}) species are found on the support surface and the active sites, respectively. In this CO₂ associative reaction mechanism, several intermediates such as carbonate, bicarbonate, and formate are observed [1].

Additionally, catalytic performance is influenced by the catalyst preparation methods, such as precipitation [56], co-precipitation [57], wet impregnation [21,58], and co-wet impregnation [59]. The wet-impregnation method is a well-known traditional method that is simple and economical. However, this generally provides a weaker metal-support interaction than the co-precipitation method [60]. Pretreatment conditions are also key factors in controlling the structural properties of the catalysts [61,62].

In this study, cobalt and ceria were chosen as the active metal and support, respectively. Although cobalt is more expensive than nickel, cobalt is more stable under low-temperature reaction conditions than nickel as using nickel can result in the formation of unstable nickel carbonyl complexes in the presence of CO, especially at low temperatures. The cobalt content was fixed to low enough so that the catalyst price could be comparable with those of typical nickel-based catalysts with high nickel content (higher than 10 wt.%). Two preparation methods, wet impregnation and co-precipitation, were conducted in this study. The effect of pretreatment temperature on the catalytic performance was also investigated.

2. Results and Discussion

2.1. Physicochemical Properties of the Prepared Catalysts

The textural properties of the prepared catalysts were determined using N₂ physisorption. Figure S1 shows the N₂ adsorption and desorption isotherms of each catalyst. The impregnated samples (Co/CeO₂) show a Type IV physisorption isotherm with an H2(a) hysteresis loop. The co-precipitated catalysts (Co_0.1Ce_0.9O_x) exhibited a Type III physisorption isotherm with an H3 hysteresis loop. As listed in

Table 1. Physicochemical properties of ceria-supported cobalt catalysts ^a.

Catalysts	BET Surface Area b (m2/g)	Pore Volume b (cm3/g)	Average Pore Diameter b (nm)	Cobalt Dispersion c (%)	CASA c (m2/gcat)	CO2 Uptake d (μmol/gcat)
Co0.1Ce0.9Ox(300)	90	0.23	10.2	4.65	0.74	124
Co0.1Ce0.9Ox(500)	72	0.28	18.9	5.98	0.95	146
Co0.1Ce0.9Ox(700)	34	0.22	26.3	3.27	0.52	59
Co/CeO2(300)	130	0.12	3.9	3.87	0.94	196
Co/CeO2(500)	128	0.14	4.3	4.65	1.13	186
Co/CeO2(700)	60	0.13	8.6	3.72	0.90	122

^a All catalysts were calcined in air at different temperatures and reduced in H₂ at 500 °C. ^b The data were determined with N₂ physisorption. ^c The data were calculated based on H₂ chemisorption. ^d The data were quantified based on pulsed CO₂ chemisorption.

, the Brunauer-Emmett-Teller (BET) surface area of the catalyst decreased, however the average pore diameter of the catalyst increased with increasing calcination temperature, irrespective of the preparation method.

The cobalt contents of the impregnated and co-precipitated samples were determined to be 5.3 and 3.3 wt.%, respectively. The catalytically active surface area (CASA) of each catalyst was determined by H₂ chemisorption. Interestingly, the CASA and cobalt dispersions show a volcano plot as a function of the calcination temperature, irrespective of the preparation method. Co_0.1Ce_0.9O_x(500) and Co/CeO₂(500) have the largest CASA among the catalysts prepared using the co-precipitation and wet impregnation methods, respectively. The number of surface basic sites was measured using pulsed CO₂ chemisorption. Of the co-precipitated catalysts, the largest number of surface basic sites was obtained for Co_0.1Ce_0.9O_x(500). The number of surface basic sites decreased with increasing calcination temperature for the Co/CeO₂ catalysts prepared using the wet impregnation method.

The bulk crystalline structures of the ceria-supported cobalt catalysts before and after reduction were probed using X-ray diffraction (XRD). Figure S2 reveals that the XRD peaks due to CeO₂ (JCPDS 34-0394) are dominant for the samples before reduction. The crystallite size of CeO₂ for the co-precipitated sample increased from 6.5 to 17.7 nm with increasing calcination temperature from 300 to 700 °C, respectively. Similarly, the crystallite size of CeO₂ for the impregnated samples also increased from 6.7 to 12.1 nm with increasing calcination temperature from 300 to 700 °C, respectively. As a result of the low cobalt content, the XRD peaks due to Co₃O₄ (JCPDS 42-1467) were found only for samples calcined at 700 °C, irrespective of the preparation method.

Temperature-programmed reduction with H₂ (H₂-TPR) was conducted to determine the reducibility of metal oxides in each catalyst. The H₂-TPR peaks can be divided into two regions: low-temperature and high-temperature. The low-temperature H₂-TPR peaks range from 100 to 600 °C. These H₂-TPR peaks are related to the reduction in cobalt oxides with different sizes and interactions with ceria. Generally, smaller metal oxides with weak interactions with a support can be reduced more easily at lower temperatures. Conversely, metal oxides with a strong interaction with a support can be reduced at high temperatures, even though the size of metal oxides is small. In the case of Co₃O₄, a two-step reduction process from Co₃O₄ to cobalt through CoO is well known [63]. The high-temperature H₂-TPR peak above 600 °C was ascribed to the reduction in surface ceria. As shown in Figure 1, the Co/CeO₂ catalysts prepared using the wet impregnation method have H₂-TPR peaks at lower temperatures than the Co_0.1Ce_0.9O_x catalysts prepared using a co-precipitation method. This indicates that the co-precipitation method can provide a stronger interaction between cobalt oxides and ceria than the wet impregnation method. It is noteworthy that the H₂-TPR peak position moves toward a high temperature with increasing calcination temperature, irrespective of the preparation method. This reveals that a stronger interaction between cobalt oxides and ceria can be achieved at higher calcination temperatures.

Temperature-programmed desorption after CO₂ chemisorption (CO₂-TPD) was performed to probe the surface basicity of each catalyst. As shown in Figure 2, there are two desorption peaks. The first occurs at temperatures less than 200 °C, and the second occurs at temperatures ranging from 200–600 °C. The low-temperature CO₂-TPD peaks originate from weakly adsorbed CO₂ on the catalyst surface, while the high-temperature CO₂-TPD peaks indicate the strong adsorption of CO₂ onto the catalyst surface. The CO₂-TPD peak observed at higher temperatures implies stronger chemisorption of CO₂ onto the catalyst surface. The Co_0.1Ce_0.9O_x(300) catalyst presents no low-temperature CO₂-TPD peak, although it has two CO₂-TPD peaks at high temperatures. However, all other catalysts exhibited low-temperature and high-temperature CO₂-TPD peaks. These two CO₂-TPD peaks were quantified, and the amounts of desorbed CO₂ for each catalyst are listed in Table S1. The amount of strongly chemisorbed CO₂ decreased with increasing calcination temperature for the co-precipitated catalyst (Co_0.1Ce_0.9O_x). The smallest amount of strongly chemisorbed CO₂ was obtained for the Co_0.1Ce_0.9O_x(700) catalyst. Similarly, the amount of strongly chemisorbed CO₂ decreased with increasing calcination temperature for the impregnated catalysts (Co/CeO₂). The smallest amount of strongly chemisorbed CO₂ was obtained for the Co/CeO₂(700) catalyst.

2.2. CO Methanation

The catalytic activity for CO methanation was evaluated using ceria-supported Co catalysts. As shown in Figure 3, the Co/CeO₂(500) catalyst exhibited the highest CO conversion at low temperatures of the tested catalysts, whereas Co_0.1Ce_0.9O_x(700) showed the lowest CO conversion even at high temperatures. As shown in Figures S3 and S4, CH₄ is the main product, and a small amount of C₂H₆ is also detected as a byproduct. Separately, the specific reaction rate of CO methanation was also obtained under a kinetically controlled regime. The specific reaction rate decreased in the following order: Co/CeO₂(500) > Co/CeO₂(300)~Co_0.1Ce_0.9O_x(300) > Co_0.1Ce_0.9O_x(500)~Co/CeO₂(700) > Co_0.1Ce_0.9O_x(700). This ranking order in the specific reaction rate is closely related to that of CASA (Table 1) for Co/CeO₂ catalysts. This implies that the catalytic activity for CO methanation is directly related to the catalytically active surface area.

The apparent activation energy (E_a) for CO methanation, calculated from the Arrhenius plot in Figure S5 over Co/CeO₂(300), Co/CeO₂(500), Co/CeO₂(700), Co_0.1Ce_0.9O_x(300), Co_0.1Ce_0.9O_x(500), and Co_0.1Ce_0.9O_x(700) were 128, 107, 129, 128, 146, and 162 kJ/mol, respectively. It is worth mentioning that the lowest apparent activation energy is obtained for the catalyst with the largest CASA among Co/CeO₂ catalysts. The fact that the co-precipitated Co_0.1Ce_0.9O_x catalysts have higher apparent activation energies than the impregnated Co/CeO₂ catalysts seems to be related to the H₂-TPR results that the co-precipitation method can provide a stronger interaction between cobalt oxides and ceria than the wet impregnation method. Castillo et al. [64] reported that the CO consumption turnover rate was significantly lower for smaller cobalt nanoparticles during CO methanation over Co/SiO₂ catalysts with different cobalt particle sizes ranging from 4 to 33 nm. The apparent activation energy was lower for smaller cobalt nanoparticles. The apparent activation energies for CO methanation over 10 wt.% Ni/CeO₂ and 58 wt.% Ni-CeO₂ were reported to be 115 and 133 kJ/mol, respectively [65].

2.3. CO₂ Methanation

The catalytic activity for CO₂ methanation was examined using ceria-supported cobalt catalysts. As shown in Figure 4, the tested catalysts are classified into three groups based on their catalytic activity for CO₂ methanation. The most active catalyst groups include Co/CeO₂(300), Co/CeO₂(500), and Co/CeO₂(700). The Co_0.1Ce_0.9O_x(300) catalyst was inferior to the most active catalyst group in this reaction. However, it was better than the least active catalyst groups (Co_0.1Ce_0.9O_x(500) and Co_0.1Ce_0.9O_x(700)). Unlike CO methanation, the catalytic activity for CO₂ methanation over ceria-supported cobalt catalysts tested in this study was not directly correlated with that of CASA (Table 1). In addition to CASA, the amount of moderate basic sites has been reported to be an additional factor affecting the catalytic activity for CO₂ methanation [66]. It is worth mentioning that the least active catalyst (Co_0.1Ce_0.9O_x(700)) had the lowest CO₂ uptake among the prepared catalysts. Methane was produced as the main product, and CO was also detected as a byproduct due to the reversed water-gas-shift side reaction (Figures S6 and S7). The apparent E_a for CO₂ methanation calculated from the Arrhenius plot in Figure S8 over Co/CeO₂(300), Co/CeO₂(500), Co/CeO₂(700), Co_0.1Ce_0.9O_x(300), Co_0.1Ce_0.9O_x(500), and Co_0.1Ce_0.9O_x(700) were 81, 77, 86, 81, 84, and 86 kJ/mol, respectively. Interestingly, all catalysts have similar apparent activation energies. The reported activation energies for CO₂ methanation over nickel-based catalysts range between 55 and 106 kJ/mol [47,67]. The apparent activation energies for CO₂ methanation over 58 wt.% Ni-CeO₂ were reported to be 95 kJ/mol [65]. Xu et al. [41] reported a higher activation energy for CO₂ methanation over Ni/ZrO₂ (99.7 kJ/mol) than over Ni/Y_0.1Zr_0.9O_x (84.2 kJ/mol). Furthermore, the activation energy of the Co/Ce_0.8Zr_0.2O catalyst was reported to be less than that of Ni/Ce_0.8Zr_0.2O catalysts [68].

2.4. In Situ DRIFTS Study

To probe the surface species during CO₂ adsorption and methanation, in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) study was conducted over the most active catalysts selected from co-precipitated and impregnated catalysts (e.g., Co_0.1Ce_0.9O_x(300) and Co/CeO₂(500)). The molecularly absorbed CO₂ was observed at ν = 2400-2200 cm⁻¹ [69]. CO₂ adsorption was initially performed at 40 °C (Figure 5). The typical IR peaks observed in this study are listed in

Table 2. FT-IR peaks during CO₂ adsorption and hydrogenation over Co_0.1Ce_0.9O_x(300) and Co/CeO₂(500).

Vibrational Mode Assignments	IR Peak Position (cm−1)							References
	Co0.1Ce0.9Ox(300)			Ce/CeO2(500)			Literature
	Adsorption		Hydrogenation	Adsorption		Hydrogenation
	40 °C	300 °C		40 °C	300 °C
Absorbed CO* species on metallic cobalt		2184 2142 2045	2115 2075		2119	2115	2181, 2168, 2118, 2084, 2048	[70,71]
Absorbed CO2							2200–2400	[69]
Bicarbonate (HCO3−)								[33,72,73,74]
ν(OH) νasy(OCO) νs(OCO) ν(COH)	1602 1408 1217	1217	1216	1654 1434		3636, 3667 1216	3600–3627, 3620 1655, 1652, 1650 1440, 1424, 1435 1225, 1220, 1228
Carbonate (CO3−)							1550 and 1380	[33,69,74]
Monodentate νasy(OCO) νs(OCO) ν(CO)	1450 1385		1394	1475 1380	1457 1354	1452 1349, 1396 1075	1446–1590 1380–1395 1040
Bidentate νasy(OCO) νs(OCO)	1293	1288	1575 1286	1545 1265 1331	1577 1279	1284	1535–1670 1243–1355
Formate ν(CH) ν(OCO)		2850	2713, 2849 2944		2848 1523	2838, 2719 2940	2756–2866, 2905 1540, 1510, 1450	[69,72,73]
Methane			3014			1305, 3014	1305 and 3015	[73]

. Bicarbonate and carbonate species were detected during CO₂ adsorption. For the bicarbonate intermediate species, asymmetric and symmetric stretching of OCO (ν_asy(OCO) and ν_s(OCO)) were observed at ν = 1602 and 1408 cm⁻¹ over Co_0.1Ce_0.9O_x(300). The same bands were observed at ν = 1654 and 1434 cm⁻¹ over Co/CeO₂(500). In addition, COH stretching for bicarbonate species (HCO₃⁻) clearly appeared at ν = 1217 cm⁻¹ over Co_0.1Ce_0.9O_x(300), while it was not observed for Co/CeO₂(500). For the carbonate species, monodentate and bidentate were confirmed over Co/CeO₂(500) at ν_asy(OCO) = 1475 and 1577 cm⁻¹ and ν_s(OCO) = 1380, 1331, and 1265 cm⁻¹, while the symmetric OCO peak centered at ν = 1293 cm⁻¹ was detected for Co_0.1Ce_0.9O_x(300). During CO₂ adsorption at 300 °C (Figures S9 and S10), the bicarbonate and carbonate species also occurred at the same position as at 40 °C. However, the noticeable difference during CO₂ adsorption at low and high temperatures is the weak and broad peaks of adsorbed CO* species on the metallic cobalt at ν = 2184, 2142, 2045, and 2119 cm⁻¹ over Co_0.1Ce_0.9O_x(300) and Co/CeO₂(500). This implies that the C=O double bond in the CO₂ molecule can be broken down to generate CO* intermediates at high temperatures.

The in situ DRIFTS study was also conducted during CO₂ methanation with increasing reaction temperature. Figure 6 and Figure 7 show the IR spectra of Co_0.1Ce_0.9O_x(300) and Co/CeO₂(500), respectively. The observed IR bands are summarized in Table 2. The sharp IR peak at ν = 1216 cm⁻¹, which can be ascribed to COH stretching at temperatures below 300 °C, is responsible for the bicarbonate species observed over both Co_0.1Ce_0.9O_x(300) and Co/CeO₂(500). Furthermore, carbonate species were found in the monodentate and bidentate forms, whereas the OCO and CO vibrations were obtained in the range of ν = 1575 and ν = 1075 cm⁻¹. In particular, the specific formate species, which are common intermediates during CO₂ hydrogenation, were also found in the stretching of CH and OCO at the high-frequency adsorption band. The vibrations at ν = 2944–2713 cm⁻¹ are ascribed to CH stretching. The weak vibration of adsorbed CO* species on metallic cobalt was also detected at ν = 2115 and 2075 cm⁻¹ over Co_0.1Ce_0.9O_x(300), while Coⁿ⁺-CO bridged species were only observed at ν = 2115 cm⁻¹ over Co/CeO₂(500). In addition, methane adsorbed on the catalyst surface was detected at ν = 1305 and 3014 cm⁻¹. Based on the IR study, it can be proposed that the CO₂ associative pathway is the main route as a result of the predominant appearance of bicarbonate, carbonate, and formate intermediates. However, the CO₂ dissociative route cannot be excluded to produce methane over ceria-supported cobalt catalysts.

3. Experimental Sections

3.1. Materials

All chemicals, including cobalt (II) nitrate hexahydrate (Junsei Chemical Co. Ltd., Tokyo, Japan), cerium (III) nitrate hexahydrate (Junsei Chemical Co. Ltd., Tokyo, Japan), and ceria (Rodia, S_BET = 250 m²/g), were reagent grade and used as received.

3.2. Preparation of the Catalysts

Co_0.1Ce_0.9O_x catalysts were prepared using the co-precipitation method. The calculated amount of each precursor was dissolved in distilled (DI) water to achieve 0.1 M total metal precursor solution. The pH was then increased by adding 0.5 M NH₄OH aqueous solution dropwise to reach a final pH of nine with vigorous stirring. The slurry was aged at room temperature for 12 h, filtered, and washed several times with DI water. The recovered solid cake was dried at 110 °C overnight under vacuum conditions. The dried sample was further calcined in air at different temperatures and finally reduced in H₂ at 500 °C.

For comparison, the Co/CeO₂ catalysts were prepared using a wet impregnation method. Cobalt precursor (1.30 g) was dissolved in 50.0 mL DI water in a round flask. Ceria (5.00 g) was added to the cobalt precursor solution. This slurry was mixed for 6 h at 60 °C in a rotary evaporator (BUCHI Labortechnik AG, Flawil, Switzerland), and then the excess water was evaporated while maintaining low pressure. The recovered powder was dried at 110 °C overnight under vacuum conditions. The dried sample was further calcined in air at different temperatures and finally reduced in H₂ at 500 °C.

The calcination temperature was included in the sample name to differentiate each catalyst calcined at different temperatures. For example, Co_0.1Ce_0.9O_x(300) and Co/CeO₂(300) are catalysts prepared by co-precipitation and wet impregnation, respectively. These catalysts were calcined in air at 300 °C.

3.3. Catalyst Characterizations

N₂ physisorption was performed at −196 °C on a Micromeritics ASAP 2020 (Micromeritics Ltd.), surface area and pore size analyzer after degassing at 200 °C under vacuum for 6 h. The specific surface area and pore size distribution were determined using the BET and Barret-Joyner-Halenda desorption methods, respectively.

XRD patterns were measured using a Rigaku D/Max instrument with a Cu K_α source. The crystallite size of ceria was determined using the Scherrer Formula (1):

$$\mathrm{d}=\frac{\mathrm{K}\mathsf{\lambda }}{\mathsf{\beta }\cos \left(\mathsf{\theta }\right)}$$

(1)

where $\mathrm{d}$ is the average particle size (nm), $\mathrm{K}$ is the dimensionless shape factor (0.9), λ is the wavelength of the X-ray radiation (0.15406 nm), β is the full width at half maximum (FWHM) of a peak in rad, and θ is the Bragg angle.

The temperature-programmed reduction with H₂ (H₂-TPR) was conducted using a Micromeritics 2920 Autochem instrument. A sample of 100 mg was loaded in a quartz tube, and the temperature was increased from room temperature (RT) to 900 °C at a heating rate of 10 °C/min under a flow of 10 mol% H₂/Ar.

The H₂-chemisorption was performed using a Micromeritics ASAP 2020 instrument (Micromeritics Ltd.), to determine the metal dispersion and CASA. A sample of 200 mg was placed in a quartz tube, reduced in H₂ at 500 °C for 1 h, cooled down to room temperature, and typical H₂-chemisorption was conducted.

Pulsed CO₂ chemisorption and temperature-programmed desorption after CO₂ chemisorption (CO₂-TPD) were conducted to measure the amount of basic sites and adsorption strength of CO₂ in each sample, respectively. They were performed using a Micromeritics 2920 Autochem instrument. A sample of 100 mg was loaded into a quartz tube, reduced under 10 mol% H₂/Ar flow at 500 °C for 1 h, cooled to room temperature, and received pulsed injection of CO₂. After saturation, CO₂ desorption was conducted using a He steam flow rate of 30 mL/min. The mass ion signals recorded at m/e = 44, m/e = 18, and m/e = 16 were monitored to detect desorbed CO₂, H₂O, and CH₄, respectively.

The metal content in each sample was determined by inductively coupled plasma-optical emission spectroscopy (ICP-OES) using a Thermo Scientific iCAP 6500 instrument.

The evolution of surface species during CO₂ adsorption and CO₂ hydrogenation was probed with in situ DRIFTS on a NICOLET 6700 (Thermal Scientific, Waltham, MA, USA) spectrometer with a ZnSe window at a resolution of 3.857 cm⁻¹. The sample was initially reduced in situ in the DRIFTS cell at 500 °C for 1 h under a H₂ flow rate of 20 mL/min and then cooled to the desired temperature. The cell was purged and stabilized with He flow for 20 min. The background was then collected under a He flow. The CO₂ adsorptions at 40 and 300 °C were separately performed under a flow of 10% CO₂/He, and the spectra were recorded simultaneously. In addition, CO₂ hydrogenation was examined at different temperatures under the same operating conditions mentioned above, and the spectra were collected after 15 min.

3.4. Catalytic Performance

The catalyst (100 mg) was placed in a fixed-bed reactor (internal diameter = 3 mm and length = 345 mm) and reduced under H₂ flow at a flow rate of 30 mL/min for 1 h. The reactor was then cooled to the desired temperature, and the catalyst was placed in contact with a feed composed of 1 mol% CO (or CO₂), 49 mol% He, and 50 mol% H₂ at a flow rate of 100 mL/min. A mass flow rate controller (MFC) (Brooks Instrument) was used to control the flow rate of each gas.

The kinetic experiment was conducted separately using 80 mg of the catalyst and 200 mg of α-alumina powder as a diluent. The CO (or CO₂) conversion was controlled to be less than 20%, and the E_a was calculated using the Arrhenius Equation (2):

$$\mathrm{k}=\mathrm{Aexp}\left(-\frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{RT}}\right)$$

(2)

where k is the reaction rate constant, A is the frequency factor, E_a is the apparent activation energy, R is the gas constant, and T is the temperature.

The effluent gas was separated using a packed column filled with carbonate for thermal conductive detector (TCD) and a capillary Poralot Q column for the flame ionization detector (FID) in a gas chromatograph (YL Instrument 6100GC). The CO (or CO₂) conversion (${\mathrm{X}}_{\mathrm{CO}}$ or ${\mathrm{X}}_{{\mathrm{CO}}_2}$, %) and the product yields, including CH₄ and CO (${\mathrm{Y}}_{{\mathrm{CH}}_4}$ and ${\mathrm{Y}}_{\mathrm{CO}}$, %) were determined according to the carbon balance as follows:

$${\mathrm{X}}_{\mathrm{CO}}=\frac{{\mathrm{C}}_{\mathrm{CO} \mathrm{input}}-{\mathrm{C}}_{\mathrm{CO} \mathrm{output}}}{{\mathrm{C}}_{\mathrm{CO} \mathrm{input}}}\times 100\%$$

(3)

$${\mathrm{X}}_{{\mathrm{CO}}_2}=\frac{{\mathrm{C}}_{{\mathrm{CO}}_2 \mathrm{input}}-{\mathrm{C}}_{{\mathrm{CO}}_2 \mathrm{output}}}{{\mathrm{C}}_{{\mathrm{CO}}_2 \mathrm{input}}}\times 100\%$$

(4)

$${\mathrm{Y}}_{{\mathrm{CH}}_4}=\frac{{\mathrm{C}}_{{\mathrm{CH}}_4 \mathrm{output}}}{{\mathrm{C}}_{\mathrm{CO} \mathrm{input}}+{ \mathrm{C}}_{{\mathrm{CO}}_2 \mathrm{input}}}\times 100\%$$

(5)

$${\mathrm{Y}}_{\mathrm{CO}}=\frac{{\mathrm{C}}_{\mathrm{CO} \mathrm{output}}}{{\mathrm{C}}_{{\mathrm{CO}}_2 \mathrm{input}}}\times 100\%$$

(6)

The input and output concentrations of each gas (i) are denoted as ${\mathrm{C}}_{\mathrm{i} \mathrm{input}}$ and ${\mathrm{C}}_{\mathrm{i} \mathrm{output}}$, respectively.

4. Conclusions

The catalytic activity for CO and CO₂ methanation was dependent on the preparation method for ceria-supported cobalt catalysts with low cobalt contents. Stronger interactions between cobalt oxides and ceria were observed for the co-precipitated catalysts than for wet-impregnated interactions. A moderate calcination temperature is plausible for achieving a high CASA. The total number of basic sites decreased significantly with an increasing calcination temperature of 500 to 700 °C for the co-precipitated and impregnated catalysts. CO methanation activity appears to be closely related to CASA, and the CASA and the number of strong basic sites affect the catalytic activity for CO₂ methanation. According to the in situ DRIFTS results, CO₂ methanation proceeds mainly through the CO₂ associative pathway, in which carbonate, bicarbonate, and formate intermediates are detected. Additionally, the CO₂ dissociative route worked simultaneously as we also observed the adsorbed CO species on the metallic cobalt surface sites.