Facile preparation of a Ni/MgAl₂O₄ catalyst with high surface area: enhancement in activity and stability for CO methanation

Catalyst characterization

N₂ adsorption

The N₂ adsorption-desorption isotherms and the pore size distribution (PSD) curves of MgAl₂O₄ and NiO/MgAl₂O₄ are exhibited in Figure 1. For the samples, on the basis of the International Union of Pure and Applied Chemistry classification, all adsorption-desorption isotherms are qualitative of a type IV shape, which is the significant property of porous structures (Guo et al., 2004). Obviously, all the samples are mesoporous materials on the basis of the Barett-Joyner-Halenda (BJH) adsorption PSDs. The PSD curves of the supports have only one peak (Figure 1B); however, the average pore sizes of NiO-0MA, NiO-0.5MA, and NiO-1MA actually shift to higher values after addition of NiO (Figure 1D). Their PSDs are bimodal with the first PSD in the range of 3.0–6.5 nm and the second PSD in the range of 9.5–15.0 nm. In contrast, for the NiO-0.25MA, the value of average PSD is almost unchanged, which mainly locates within the range of 3.0–6.5 nm, revealing the even distribution of oxide particles (Sun et al., 2008).

Figure 1:

N₂ adsorption-desorption isotherms and pore size distribution of the xMA supports (A, B) and NiO-xMA catalysts (C, D).

Table 1 lists the physical parameters of the calcined samples, for instance, the specific surface area, total pore volume, and average pore diameter. After addition of carbon black, the specific surface area of MgAl₂O₄ increases obviously, particularly for 0.25MA, which can be as high as 235.8 m² g⁻¹. This may because of the fact that the addition of carbon black inhibits severe aggregation of particles in the high-temperature calcination step and the supports can form more pore structure after removing of carbon black (Liu et al., 2013a,b). In principle, high specific surface areas should supply a large number of reaction or interaction sites, which can enhance surface or interface-related processes such as adsorption and separation (Li et al., 2016). The specific surface area of MgAl₂O₄ (0.5MA and 1MA) decreased with addition of excess carbon black, which due to excess carbon black agglomerated into large blocks forming larger pores in the process of forming pore structures. The large pore volume for 0.25MA was up to 0.79 cm³ g⁻¹, and the average pore diameter was around 8 nm. Large pore volume has shown promise in the loading of active species. The specific surface area of catalysts becomes lower with addition of excess NiO species, perhaps because of blockage of some pores over MgAl₂O₄.

Table 1:

Physical parameters of the supports and oxidized catalysts.

Sample	SBETa (m2 g−1)	Vpb (cm3 g−1)	Dpc (nm)
0MA	147.5	0.33	8.9
0.1MA	198.7	0.73	14.7
0.25MA	235.8	0.79	13.4
0.5MA	208.7	0.66	12.1
1MA	186.7	0.63	12.8
NiO-0MA	132.8	0.33	9.9
NiO-0.1MA	159.8	0.43	10.8
NiO-0.25MA	177.3	0.46	10.4
NiO-0.5MA	158.5	0.62	15.6
NiO-1MA	142.6	0.59	16.5

1. ^aSurface area of the calcined sample derived from BET equation.

2. ^bPore volume of the calcined sample obtained from the volume of nitrogen adsorbed at the relative pressure of 0.97.

3. ^cAverage pore diameter of the calcined sample derived from BJH method using the following equation: DP=4×VPSBET.

XRD analysis

Figure 2 shows the wide-angle X-ray diffraction (XRD) patterns of the samples. As seen in Figure 2A, the XRD patterns of all the supports synthesized by the modified co-precipitation method have apparent diffraction peaks, which are located at 19.5, 31.8, 37.1, 45.2, 59.8, and 65.5°, respectively, corresponding to the characteristic diffraction peaks of MgAl₂O₄ (JCPDS, No. 73-1959). No component phases such as MgO and Al₂O₃ are observed, which indicates that the supports are formed by the single-phase spinel MgAl₂O₄. The diffraction peaks become broad and dispersed with the increase in the amount of carbon black, revealing that the supports show poor crystallinity correspondingly with addition of carbon black, which indicates that the existence of carbon black has some effect on the crystallization process of MgAl₂O₄. For the NiO/MgAl₂O₄ catalyst (Figure 2B), the diffraction peaks of MgAl₂O₄ seem to be unchanged, and the diffraction peaks at 37.1, 43.5, and 63.4° can be assigned to the (111), (200), and (220) planes of cubic-phased NiO species (JCPDS, No. 65-2901), suggesting that the addition of a small amount of nickel species does not destroy the structure of MgAl₂O₄ spinel. For these diffraction peaks, especially those located at 43.5 and 63.4°, the diffraction intensity of NiO conforms to the following order: NiO-0.25MA<NiO-0.5MA<NiO-1MA<NiO-0.1MA<NiO-0MA, demonstrating the formation of a separate nickel oxide phase and the NiO particles dispersed best on the support of NiO-0.25MA compared with other catalysts. The intensity of the NiO diffraction peaks is lower, suggesting the formation of smaller metal nickel crystallites. As indicated in Figure 2C, for the catalysts reduced at 650°C, the peaks located at 44.9, 52.0, and 76.8° correspond to (111), (200), and (220) planes of metallic Ni (JCPDS, No. 01-070-1849), respectively. The diffraction intensity of Ni at 44.9, 52.0, and 76.8° follows the order below: Ni-0.25MA<Ni-0.5MA<Ni-1MA<Ni-0.1MA<Ni-0MA, which is in keeping with the diffraction intensity of oxidized catalysts, indicating that nickel species are highly dispersed over the catalyst of Ni-0.25MA compared with other catalysts owing to its high specific surface area.

Figure 2:

XRD patterns of the catalysts: (A) xMA supports; (B) NiO-xMA catalysts; and (C) 650°C-reduced Ni-xMA catalysts.

The Ni nanoparticle sizes estimated from the XRD patterns of the reduced catalysts are listed in Table 2. For Ni-0MA, the Ni particle size is 10.6 nm, while after the addition of carbon black, the sizes of Ni particle are <7.0 nm (except for Ni-0.1MA), revealing that the incorporation of carbon black can decrease the size of Ni particles and further improve the Ni dispersion. Additionally, the Ni particle size of the catalyst of Ni-0.25MA is 5.9 nm, which is smaller than other reduced catalysts. The main reason is that the supports form a large number of pore structures with addition of moderate amount of carbon black, which promotes the dispersion of nickel particles.

Table 2:

Physical parameters of reduced catalysts.

Sample	Ni particle size (nm)		Maximumb
	By XRDa	By TEM	CO conversion (%)	CH4 selectivity (%)
Ni-0MA	10.6	15.5±5.8	72.88	68.80
Ni-0.1MA	8.4	12.4±4.6	95.38	72.87
Ni-0.25MA	5.9	8.6±2.8	99.99	85.70
Ni-0.5MA	6.1	11.5±3.9	98.57	78.42
Ni-1MA	6.8	11.9±5.2	97.73	76.48

1. ^aEstimated from the XRD diffraction peak (27#x03B8; = 44.9°) using the Debye-Scherrer equation.

2. ^bIn the temperature range of 300–480°C, 0.1 MPa, 30 000 mL g⁻¹ h⁻¹.

H₂-TPR analysis

Figure 3A shows the H₂ temperature-programmed reduction (H₂-TPR) profiles of the catalysts, and both the corresponding peak temperature and fraction of each elementary peak area are exhibited in Table 3. It is reported that the reducible NiO species could be classified into four types: α-type, β₁-type, β₂-type, and γ-type (Zhang et al., 2005; Kang et al., 2011; Hu et al., 2012; Zhao et al., 2012). The H₂ consumption peaks located in the relatively low temperature range (452–502°C) can belong to the reduction of α-type NiO species (bulk NiO or surface amorphous NiO), which have weak interaction with the support (Rynkowski et al., 1993; Diskin et al., 1998). The peaks of α-type NiO in the Ni-0MA and Ni-0.1MA catalysts are much stronger than those of other catalysts. For the α-type NiO species, the Ni particles would migrate and agglomerate due to the weak interaction between metallic nickel particles and support, leading to Ni sintering and catalyst deactivation during the processes of both catalyst reduction and methanation reaction at relatively high temperatures (Yang et al., 2010; Zou et al., 2010). The H₂ consumption peaks located in the middle temperature range (602–651°C, 749–762°C) can be attributed to the reduction of β₁-type NiO species and β₂-type NiO species, respectively, which possess stronger interaction with the support than the α-type NiO species. It is commonly known that the active NiO species of β stage have relatively high catalytic activity in CO methanation reaction (Ren et al., 2017). It can be seen that the improvement of dispersion of nickel particles can enhance the interaction between active metal nickel and support, which leads to the increase of the NiO reduction temperature (Özdemir et al., 2014). Among all the catalysts, Ni-0.25MA has the highest content NiO species of β stage (82.1%), which can facilitate CO methanation reaction. This result reveals that the catalyst of Ni-0.25MA can provide suitable interaction between active metal and support. Thus, the Ni-0.25MA catalyst can possess more suitable nickel active species, which is beneficial for CO methanation reaction. While the H₂ consumption peaks located in the relatively high temperature region (845–851°C) refer to reduction of γ-type NiO species, this part of NiO may react with the support to form spinel NiAl₂O₄ at high temperature (Sun et al., 2008; Fan et al., 2015).

Figure 3:

Profiles of the catalysts.

(A) H₂-TPR profiles of the catalysts; (B) H₂-TPD profiles of the catalysts.

Table 3:

H₂-TPR quantitative data of the catalysts.

Catalyst	Tm (°C)				Fraction of total area (%)
	α	β1	β2	γ	α	β1	β2	γ
Ni-0MA	452	602	751	849	25.7	46.3	23.6	4.4
Ni-0.1MA	474	618	749	845	17.9	52.9	24.9	4.3
Ni-0.25MA	502	645	762	851	15.6	56.9	25.2	2.3
Ni-0.5MA	490	651	753	848	18.1	52.6	23.8	5.5
Ni-1MA	468	628	749	845	19.1	56.8	21.7	2.4

SEM observation

Figure 4 shows the scanning electron microscope (SEM) microphotographs of the different materials related to the modified co-precipitation method and impregnation method. Figure 4A shows the SEM micrographs of the MA+CB_0.25 solid before the removal of carbon black. The MgAl₂O₄ and carbon black mixed together are well-proportioned, forming a compact overall structure after calcination in Ar. In Figure 4B, the abundant pore structure of the 0.25MA can be observed after removing of carbon black, which corresponds to the higher values for specific surface area and total pore volume obtained from the physical absorption of N₂. As seen in Figure 4C, the NiO-0MA particles with a structure similar to superposed platelet may be shaped by agglomeration of smaller particles. Furthermore, for NiO-0.25MA, it can be observed that the support with addition of moderate amount of carbon black can form more pore structures (Figure 4D), which facilitates the high dispersion of nickel particles on the 0.25MA support. And it is in accordance with the result of high specific surface area of this sample in Table 1.

Figure 4:

SEM images of the MA+CB0.25 (A), 0.25MA (B), NiO-0MA (C), and NiO-0.25MA (D).

Moreover, Figure 5 reveals the SEM image and the related elemental mapping of the reduced Ni-0.25MA, which confirms that Ni, Mg, and Al elements are homogeneously dispersed across the whole catalyst and the single-phase character of the samples.

Figure 5:

SEM image of the reduced Ni-0.25MA (A) and elemental mapping images of O (B), Mg (C), Al (D), and Ni (E).

TEM observation

Figure 6A exhibits the TEM microphotograph of the Ni-0MA; it can be observed that the active nickel particles are aggregated severely and most of them are loaded on the surface of the support. The addition of carbon black can change the Ni particles size obviously compared to the Ni-0MA, particularly for Ni-0.25MA, which shows higher Ni dispersion and much smaller Ni particle size (Figure 6B–E). The reason may be that the addition of proper amounts of carbon black can form more uniform pore structure, which may provide the appropriate interaction between active metal and support. This structure can promote the dispersion of nickel, resulting in more exposed active nickel species for CO methanation. According to the particle size statistical analysis, the average nickel particle sizes of Ni-0MA, Ni-0.1MA, Ni-0.25MA, Ni-0.5MA, and Ni-1MA were 15.5±5.8, 12.4±4.6, 8.6±2.8, 11.5±3.9, and 11.9±5.2 nm (Table 2), respectively. This result is consistent with that of the nitrogen adsorption and XRD characterizations, which demonstrates that the specific surface area of support has a significant impact on the dispersion of nickel nanoparticles. The reduced Ni-0.25MA catalyst has the smallest nickel nanoparticle size and hence the highest nickel dispersion and active nickel specific surface area, which will help to enhance its catalytic activity (Liu et al., 2015a,b,c).

Figure 6:

TEM and particle statistical distribution images of the reduced catalysts: (A) Ni-0MA, (B) Ni-0.1MA, (C) Ni-0.25MA, (D) Ni-0.5MA, and (E) Ni-1MA.

XPS analysis

In order to figure out the surface chemical composition of the catalysts with different amounts of carbon black addition, the X-ray photoelectron spectroscopy (XPS) characterizations for the Ni-0MA as well as Ni-0.25MA were carried out. In Figure 7A, it can be seen that the Ni⁰ peaks are at the same position (around 852.5 eV), which are attributed to the metallic nickel (Kim et al., 2007), revealing that the introduction of carbon black does not alter the binding energy of metallic nickel. The results demonstrate that the carbon black was completely removed from support after calcination in air at 600°C for 3 h. The peaks of Ni 2p_3/2 at 855.2 and 860.3 eV belong to Ni²⁺ due to the quick oxidation of Ni⁰ in air (Liu et al., 2014a,b,c). After addition of carbon black, a shift of 0.5 eV for Ni⁰ from 852.4 eV to 852.9 eV and for Ni²⁺ from 854.2 eV to 854.6 eV for the Ni-0.25MA catalyst can be observed. Furthermore, the ratio of Ni⁰/Ni²⁺ in Ni-0MA is much lower than in Ni-0.25MA, which may be due to the fact that the metallic nickel nanoparticles are embedded in the pore of the 0.25MA support; thus, the oxidation of nickel becomes more difficult.

Figure 7:

Spectra of the reduced Ni-0MA and Ni-0.25MA catalysts.

(A) The Ni 2p spectra of the reduced Ni-0MA and Ni-0.25MA catalysts; (B) the O 1s spectra of the reduced Ni-0MA and Ni-0.25MA catalysts.

The O 1s XPS spectra of the samples are shown in Figure 7B. It can be seen that the peak at 529.9 eV for the Ni-0MA and Ni-0.25MA are attributed to lattice oxygen. In addition, for Ni-0.25MA, the binding energy peak at 527.8 eV represents chemisorbed oxygen, revealing that the support formed more pore structure after the removal of carbon black, which makes the catalyst produce more oxygen vacancy compared with the catalyst without addition of carbon black.

Hydrothermal stability test

In view of H₂O being one of the byproducts in CO methanation, a high hydrothermal stability is required for the catalyst due to the fact that the higher the catalyst activity is, the higher concentration of water vapor in product gas will be. Furthermore, steam is often mixed into the feed gas to restrain both the hot spot in catalyst bed and the coking formation in industry (Li et al., 2010). Thus, the hydrothermal stability is an important character for CO methanation catalyst. The hydrothermal treatment of Ni-0MA and Ni-0.25MA was carried out at 600°C, 0.1 MPa, and weight hourly space velocity (WHSV) of 60 000 mL g⁻¹ h⁻¹ in the presence of 80 vol% H₂O/H₂ in this work. The XRD patterns and TEM image of the hydrothermally treated catalysts are exhibited in Figure 8. As shown in Figure 8A, compared with the fresh catalysts, the positions of diffraction peaks are unchanged, but the diffraction intensity of MgAl₂O₄ declined in a different extent. This may be due to the decrease in crystallinity of MgAl₂O₄ during the hydrothermal treatment process. The intensities of the Ni peaks for Ni-0MA and Ni-0.25MA increased, and the Ni nanoparticle sizes of Ni-0MA and Ni-0.25MA estimated from the XRD patterns (Figure 8A) are 19.8 and 10.2 nm, respectively, which can be ascribed to sintering of the active nickel species, thereby decreasing the uniform dispersion of nickel nanoparticles, then leading to the increase in signal of nickel diffraction peak in XRD patterns. It can be found that nickel particle size grows from 15.5±5.8 nm to 24.5±6.7 nm in Ni-0MA-H (Figure 8B), while the change in Ni-0.25MA-H is from 8.6±2.8 nm to 11.8±3.5 nm (Figure 8C). The drastic increase of Ni particle size calculated from the TEM image of Ni-0MA-H (Figure 8B) should be owing to the poor interaction between the Ni nanoparticles and the MgAl₂O₄ support. On the contrary, Ni-0.25MA-H still shows relatively high dispersion of Ni nanoparticles (Figure 8C), and the Ni particle size increased indistinctly, indicating that the formation of porous structure by introduction of moderate amount of carbon black is useful to stabilize Ni nanoparticles on MgAl₂O₄. These results suggest the better hydrothermal stability of Ni-0.25MA than Ni-0MA, which should be ascribed to the fact that addition of carbon black can improve structure stability and thermal stability of MgAl₂O₄ support.

Figure 8:

XRD patterns (A) and TEM image of the hydrothermally treated catalysts: (B) Ni-0MA-H and (C) Ni-0.25MA-H.

Catalytic performance test

As seen in Figure 9, the reaction over the Ni-xMA catalysts are carried out in the temperature range of 300–480°C at 0.1 MPa using a WHSV of 60 000 mL g⁻¹ h⁻¹. For Ni-0MA, the catalytic activity is the worst, and the maximum CO conversion and CH₄ yield at 420°C is 72.88 and 48.25%, respectively. After the carbon black addition, the CO conversion and CH₄ yield of the Ni-xMA (x=0.1, 0.25, 0.5, 1) catalysts are rapidly improved, especially for Ni-0.25MA: its CH₄ yield can reach 85.70%, and the CO conversion is close to the thermodynamics equilibrium value when the reaction temperature is higher than 360°C (Liu et al., 2014a,b,c). Compared with Ni-0MA, the addition of carbon black can considerably enhance the catalyst’s low-temperature-activities. The CO conversion over all the Ni-xMA (x=0.1, 0.25, 0.5, 1) catalysts can run up to the thermodynamics equilibrium value in the reaction temperature range of 340–380°C. In addition, the Ni-xMA (x=0.1, 0.25, 0.5, 1) catalysts show similar activity (>400°C) due to thermodynamic equilibrium limit. Furthermore, it can be seen from Figure 9A that the activities of Ni-xMA (x=0.1, 0.25, 0.5, 1) are obviously improved compared to those of Ni-0MA, revealing that the addition of carbon black also can strengthen the performance of the catalysts. However, the activities of both Ni-0.5MA and Ni-1MA are lower than that of Ni-0.25MA, demonstrating that the excess carbon black may be disadvantageous for the formation of the support structure. The activities of all the catalysts beyond 440°C decrease, because the CO methanation is a strongly exothermic reaction with thermodynamic limit (Gao et al., 2012). Among all the catalysts, Ni-0MA shows the poorest catalytic performance, corresponding to its worst dispersion of nickel nanoparticles. The addition of carbon black can similarly increase CH₄ selectivity of the catalysts as shown in Figure 9B. The CH₄ selectivity at low reaction temperatures is strongly enhanced with the addition of carbon black, especially for Ni-0.25MA. The CH₄ selectivity of Ni-0.25MA can reach 85.70% at 340°C, but with the rise of temperature, the CH₄ selectivity of Ni-0.25MA becomes lower. The detected byproduct is CO₂, which is the product of water-gas shift reaction, CO Boudouard reaction, or the reversed methane-CO₂ reforming reaction (Gardner et al., 1981; Liu et al., 2014a,b,c). The maximum CH₄ yields over Ni-0.25MA can reach 85.70% at a relatively low temperature of 360°C, while that over Ni-0MA is just 48.25% at high temperature of 420°C (Figure 9C). Clearly, the addition of moderate amount of carbon black (25%) can remarkably improve the low temperature activity performance and the CH₄ yield of the catalyst at atmospheric pressure. Combining with the aforementioned XRD and TEM results, Ni-0MA shows the poorest performance of both CO conversion and CH₄ yield, indicating that the support of Ni-0MA formed a relatively dense structure which has few pores with the absence of carbon black; thus, the active nickel particles are aggregated, and most of them are loaded on the surface of the support of Ni-0MA, which leads to the aggregation and sintering of nickel metal particles during the process of catalyst reduction at 650°C and the poorest catalytic performance. For Ni-0.25MA, the activities were dramatically improved due to incorporation of moderate amount of carbon black, which made the support produce more uniform pore structure and brought about the appropriate metal-support interaction, thereby promoting the dispersion of nickel. It was reported that Ni particles with highly dispersion could provide more active catalytic sites to facilitate methanation reactions (Liu et al., 2015a,b,c). The CO conversion is approaching 100%, and CH₄ selectivity is above 85% between 340 and 400°C for the Ni-0.25MA catalyst. The addition of a small amount of carbon black for Ni-0.1MA or excess carbon black for Ni-0.5MA and Ni-1MA caused the formation of smaller or larger pores to supports, which results in the worse dispersion of Ni particles.

Figure 9:

Catalytic properties of the catalysts at 60 000 mL g⁻¹ h⁻¹, 0.1 MPa: (A) CO conversion, (B) CH₄ selectivity, and (C) CH₄ yield.

Stability test

In order to investigate the catalytic stability of the best catalyst Ni-0.25MA, a lifetime test was performed at fixed bed reactor for 120 h at the constant temperature of 540°C, 0.1 MPa, and a high WHSV of 60 000 mL g⁻¹ h⁻¹. Generally, sintering of nickel particles will take place at relatively high reaction temperatures (>500°C), and carbon deposition can be formed at temperatures above 450°C (Bartholomew, 1982, 2001; He et al., 2016); hence, the high temperature was used in stability test. In addition, Ni-0MA was also tested as the reference sample, and the results are shown in Figure 10. ForNi-0MA, the CO conversion and CH₄ selectivity remain stable in the first 40 and 5 h; however, they decline sharply during the next 80 or 70 h, and the CO conversion and CH₄ yield is only 35% and 68% at the end of the lifetime test, respectively. Obviously, the Ni-0MA was almost completely inactivated under harsh reaction conditions in the 120-h lifetime test. In contrast, Ni-0.25MA exhibits much better catalytic activity and stability, and there is no significant decrease in activity during the whole stability test. This may be because more pores were formed after the incorporation of carbon black, which brings stronger interaction between metal and support. This result demonstrates that nickel metal particles loaded in the pore of the support showed better anti-sintering ability than those loaded in the surface of the support. Therefore, the excellent stability of the catalyst is closely associated with the anchoring of nickel nanoparticles in the pores of the catalyst, which can promote the interaction between nickel metal and support as described in H₂-TPR analysis and inhibit the rapid decrease of specific surface area owing to the property of anti-sintering in the process of stability test.

Figure 10:

Stability test of the catalysts for CO methanation: (A) CO conversion, (B) CH₄ selectivity, and (C) CH₄ yield.

Characterization of the spent catalysts

To analyze the deposited carbon and nickel sintering over the spent catalysts, a series of characterizations were carried out. The TEM images of the Ni-0MA-spent and Ni-0.25MA-spent catalysts are exhibited in Figure 11A and B. The migration and aggregation of nickel nanoparticles over the spent Ni-0MA are very serious compared with the freshly reduced Ni-0MA (Figure 6A). It can be observed from Figure 11A that the nickel nanoparticle size of Ni-0MA-spent grows from 15.5±5.8 nm (Figure 6A) to 42.6±11.9 nm, which have spherical-like shape in the freshly reduced Ni-0MA agglomerated to bulk after the stability test (see yellow oval dashed area of Figure 11A) owing to the easy migration of nickel nanoparticles over the surface of the support. In contrast, nickel nanoparticles still maintain high dispersion in the Ni-0.25MA-spent catalyst. Furthermore, as seen from the TEM image of the spent Ni-0.25MA catalyst, nickel nanoparticle size grows from 8.6±2.8 nm (Figure 6C) to 12.6±3.9 nm (Figure 11B) without any obvious agglomeration, which exhibits significant resistance to sintering compared with Ni-0MA-spent. Figure 11D shows the XRD patterns of the Ni-0MA-spent and Ni-0.25MA-spent catalysts. The nickel nanoparticle size of the Ni-0.25MA-spent is calculated to be 12.8 nm, which is slightly larger than that of the freshly reduced Ni-0.25MA (8.6 nm). In contrast, the nickel nanoparticle size of Ni-0MA-spent increases from 15.5 to 30.6 nm after the lifetime test. These results indicate that the Ni-0.25MA is thermally stable and well retained after the 120-h stability test, which brings about the high resistance toward Ni sintering during the process of CO methanation. It can be demonstrated that the confinement of the nickel nanoparticles with high dispersion in the pore and their anchoring in the MgAl₂O₄ can inhibit Ni sintering significantly. It is known that the carbon deposition may be in the forms of amorphous carbon, vermicular carbon, and graphitic carbon (Bartholomew et al., 1982), but the diffraction peak corresponding to graphitic carbon was not observed in the XRD pattern (Figure 11D), and there was no clear observation of carbon filaments on the catalysts in the TEM and SEM images of the spent catalysts (Figure 11A–C), indicating that the deposited carbon is not crystallized or the amount of graphitic carbon cannot reach the detection limit of XRD over the spent catalysts. The amount of carbon deposition over the spent catalysts is further measured by thermogravimetric (TG) analysis, and the results are presented in Figure 11E. The deposited carbon content over the spent Ni-0MA is up to 3.6 wt%. In contrast, the deposited carbon content over the spent Ni-0.25MA is just 1.3 wt%, suggesting that the latter has a stronger performance of anti-coking than the former. The weight increase between 280 and 450°C in all the spent catalysts is due to the oxidation of nickel, and the increment of weight for Ni-0MA-spent is higher than Ni-0.25MA-spent, revealing that the reduced nickel loaded on 0.25MA is more difficult to be oxidized than that loaded on 0MA due to stronger interaction between nickel and support, which is in accordance with the H₂-TPR results (Figure 3A). Besides, the weight loss above 450°C is due to the oxidation of the deposited carbon; Ni-0.25MA-spent exhibits better anti-coking property than Ni-0MA-spent. In short, the superior properties of both anti-sintering and anti-coking of the Ni-0.25MA catalyst can be attributed to the appropriate interaction between metal nickel and support as well as the porous structure of the support. Furthermore, the Ni-xMA catalysts show higher resistance to carbon deposition than those Ni-based catalysts such as bentonite, α-Al₂O₃ (Liu et al., 2015a,b,c; Lu et al., 2015), which is because Mg is alkaline earth metal and its incorporation can enhance the alkalinity of the catalyst (Zuo et al., 2013). Thus, the MgAl₂O₄ can exhibit better resistance to carbon deposition as the support.

Figure 11:

TEM images of the Ni-0MA-spent catalyst (A) and Ni-0.25MA-spent (B); SEM image of the Ni-0.25MA-spent catalyst (C); XRD analysis of the Ni-0MA-spent and Ni-0.25MA-spent (D); and TG curves of the reduced and spent catalysts in air (E).

Particle statistical distribution images ofNi-0MA-spent and Ni-0.25MA-spent are listed in (A) and (B) inset.

Synthesis of supports and catalysts

Aluminum nitrate nonahydrate (Al(NO₃)₃⋅9H₂O), magnesium nitrate hexahydrate (Mg(NO₃)₂⋅6H₂O), ammonium carbonate ((NH₄)₂CO₃), and nickel (II) nitrate hexahydrate (Ni(NO₃)₂⋅6H₂O) were purchased from Sinopharm Chemical Reagent Co. Ltd., Shanghai, China, and carbon black purchased from Alfa Aesar (product code 39724, acetylene, 50% compressed, Shanghai, China). In the process of synthesis of MgAl₂O₄ (Figure 12), 15.85 g of Al(NO₃)₃⋅9H₂O and 5.42 g of Mg(NO₃)₂⋅6H₂O were dissolved in 300 mL of deionized water at 60°C, then different mass ratio (carbon black/MgAl₂O₄=0, 0.1, 0.25, 0.5, 1) of carbon black were added and stirred 5 h to obtain a precursor slurry. Moreover, 10.00 g of (NH₄)₂CO₃ was dissolved in 50 mL of deionized water and then heated to 60°C. Subsequently, the (NH₄)₂CO₃ solution was added to the above slurry at pH value of 8.0. After being vigorously stirred for 4 h, the mixture was filtered and washed twice. The filter cake was dried at 100°C overnight and then calcined in Ar (Ar purchased from Jinan Deyang Special Gas Co., Ltd., Jinan, Shandong, China) at 800°C for 2 h at a heating rate of 5°C min⁻¹ and marked as ‘MA+CB_x’ (x=0, 0.1, 0.25, 0.5, 1; MA: MgAl₂O₄; CB:carbon black), followed by the removal of carbon black in air and denoted as ‘xMA’. As seen in Figure 13, the complex of MA and CB was detected by TG in air flow, and the CB can be removed at 600°C. Thus, in order to completely remove the carbon black, the selected calcination condition was 600°C for 3 h at a heating rate of 5°C min⁻¹.

Figure 12:

Schematic diagram illustrating the preparation of Ni/MgAl₂O₄ catalysts: (A) supports synthesis by the co-precipitation method and (B) catalysts synthesis by the impregnation method.

Figure 13:

TG curve of MgAl₂O₄ and CB in air.

The NiO/MgAl₂O₄ catalysts (20 wt% NiO) were prepared by the traditional impregnation method. About 1.56 g of Ni(NO₃)₂⋅6H₂O was dissolved in 50 mL of ethanol solvent, and then 2.00 g of MgAl₂O₄ was added. The resulting mixture was vigorously stirred at 30°C for 12 h, then evaporated at 70°C under stirring, and dried at 100°C overnight. After calcination at 500°C for 2 h in air at a heating rate of 5°C min⁻¹, the obtained samples were denoted as ‘NiO-xMA’, whose schematic diagram is shown in Figure 12B. Accordingly, the catalysts after H₂ reduction were denoted as ‘Ni-xMA’, and the corresponding spent catalysts were marked as ‘Ni-xMA-spent’.

Catalyst characterization

N₂ adsorption was measured at −196°C using a Quantachromesurface area and pore size analyzer NOVA 3200e (SSA-4300, Beijing Builder Electronic Technology Co., Ltd., Beijing, China). Prior to the measurement, the sample was degassed at 200°C for 4 h under vacuum. The specific surface area was determined according to the Brunauer-Emmett-Teller (BET) method in the relative pressure range of 0.05–0.3. The PSD was calculated based on the BJH method using the adsorption isotherm branch. XRD patterns were tested by a Rigaku Utima IV (Rigaku Corporation, Osaka, Japan) with a step of 0.02 from 5.0 to 80.0 (wide angle range) using Cu Kα radiation (λ=1.5418 Å) at 40 kV and 40 mA. The crystal size of the sample was estimated using the Debye-Scherrer equation. The morphology of the sample was observed by SEM (JSM-6700F, JEOL, Tokyo, Japan) and TEM (JEM-2010F, JEOL, Tokyo, Japan). Before TEM measurement, the H₂-reduced catalysts were cooled to room temperature in H₂ flow and then passivated in 1 vol% O₂/Ar gas mixture for 30 min to prevent bulk oxidation of the Ni nanoparticles. H₂-TPR and H₂-TPD were carried out on a Micromeritics AutoChem II 2920 (Micromeritics, Atlanta, GA, USA). Typically, prior to the TPR measurements, a 0.10 g sample was pretreated at 300°C for 1 h under He flow to remove moisture and impurities. Then the sample was cooled to room temperature, followed by heating to 1000°C at a rate of 10°C min⁻¹ in 10 vol% H₂/Ar flow (30 mL min⁻¹), and the thermal conductivity detector (TCD) signal was recorded continuously. For H₂-TPD, a 0.20 g sample was prereduced in situ by H₂/Ar flow at 700°C for 1 h and saturated with H₂ for 1 h at room temperature. After the physically adsorbed H₂ was removed by purging with Ar for 2 h, the sample was heated to 600°C at 10°C min⁻¹ in an Ar flow (30 mL min⁻¹). In order to research the properties of deposited carbon over spent catalysts, TG analysis was carried out on STA449F3 (NETZSCH-Gerätebau GmbH, Bavaria, Germany), NETZSCH-Gerätebau GmbH, Germany. About 10 mg of the sample was used and heated under air (200 mL min⁻¹) from room temperature up to 1100°C (10°C min⁻¹). The surface chemical composition was analyzed by XPS conducted on a VG ESCALAB 250 spectrometer (Thermo Electron, UK) with a nonmonochromatized Al Kα X-ray source (1486 eV).

Catalytic activity measurements

CO methanation reactions were carried out in a fixed bed reactor using quartz tube (I.D. 8 mm) at 0.1 MPa. A thermocouple was placed in the chamber near the middle position of the catalyst bed to monitor the reaction temperature. The granule catalyst sample (0.1 g) with size of 20–40 mesh blended with quartz sands (5.0 g, quartz sands were purchased from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) were loaded in the reactor and reduced at 650°C at 5°C min⁻¹ in pure H₂ for 1 h. After being cooled to the starting reaction temperature in H₂, the fixed bed reactor was passed into the mixed H₂ and CO as well as N₂ (H₂/CO/N₂=3/1/1, molar ratio). All gases were purchased from Jinan Deyang Special Gas Co., Ltd., Jinan, Shandong, China with purity >99.999% and used as received. The product gases were analyzed by a gas chromatography (Micro 3000A, Agilent Technologies) equipped with flame ionization detector and TCD. A stability test of CO methanation reaction was operated under the following conditions (540°C, 0.1 MPa, and WHSV=60 000 mL g⁻¹ h⁻¹). The CO conversion, CH₄ selectivity, and yield are defined as below:

CO conversion: XCO(%)=FCO,in−FCO,outFCO,in×100

CH4 selectivity: SCH4(%)=FCH4,outFCO,in−FCO,out×100

CH4 yield: YCH4(%)=XCO×SCH4100=FCH4,outFCO,in×100

where, X, S, and Y are the abbreviation of conversion, selectivity, and yield, respectively, and F_i,in and F_i,out are the volume flow rates of CO or CH₄ at the inlet and outlet.

Hydrothermal stability measurement

The hydrothermal treatment was carried out in a tubular furnace reactor at 0.1 MPa. The catalyst (0.3 g) was prereduced at 650°C in pure H₂ for 1 h and then subjected to 80 vol% H₂O/H₂ (300 mL min−¹) at 600°C for 12 h. The obtained samples were denoted as Ni–yMA–H (y=0 and 0.25), where ‘H’ represents hydrothermal, and ‘y’ represents the different mass ratio of carbon black and MgAl₂O₄.