E ﬀ ect of Support and Chelating Ligand on the Synthesis of Ni Catalysts with High Activity and Stability for CO 2 Methanation

: Carbon dioxide methanation was carried out over Ni-based catalysts on di ﬀ erent supports and chelating ligands in microreactors. To investigate the inﬂuence of chelating ligands and supports, the Ni catalysts were prepared using di ﬀ erent support such as CeO 2 , Al 2 O 3 , SiO 2 , and SBA-15 by a citric acid (CA)-assisted impregnation method. The properties of the developed catalysts were studied by X-ray di ﬀ raction (XRD), Transmission electron microscope (TEM), and X-ray photoelectron spectroscopy (XPS) measurement, and the results show that the addition of CA in the impregnation solution improved the dispersion, reﬁnes the particle size, and enhanced the interaction of nickel species. The catalytic performance of the developed Ni catalysts were evaluated by CO 2 methanation in microreactors in the temperature range of 275 ◦ C–375 ◦ C under 12.5 bar pressure. All the catalysts exhibit high CO 2 conversion and extremely high selectivity to methane. However, the catalysts prepared via CA-assisted method exhibited excellent activity and stability, compared with Ni catalysts prepared by a conventional impregnation method, which could be attributed to highly dispersed nickel particles with strong metal–support interaction. The activity of CO 2 methanation followed the order of Ni / CeO 2 -CA > Ni / SBA-15-CA > Ni / Al 2 O 3 -CA > Ni / SiO 2 -CA > Ni / CeO 2 . The Ni / CeO 2 catalysts have also been prepared using di ﬀ erent chelating ligands such as ethylene glycol (EG), sucrose (S), oxalic acid (OA) and ethylene diamine tetra acidic acid (EDTA). Among the tested catalysts prepared with di ﬀ erent support and chelating ligands, the Ni / CeO 2 catalyst prepared via CA-assisted method gave superior catalytic performance and it could attain 98.6% of CO 2 conversion and 99.7% methane selectivity at 325 ◦ C. The partial reduction of the CeO 2 support generates more surface oxygen vacancies and results in a high CO 2 conversion and methane selectivity compared with other catalysts. The addition of CA as promoter favored the synergistic e ﬀ ect of Ni and support, which led to high dispersion, controls the size, and stabilizes the Ni nanoparticles. Furthermore, the Ni / CeO 2 -CA catalyst yields high CO 2 conversion in a time-on-stream study due to the ability of preventing the carbon deposition and sintering of Ni particles under the applied reaction conditions. However, the Ni / Al 2 O 3 -CA and Ni / SBA-15-CA catalysts showed stable performance for 100 h of time on stream.


Introduction
Since the CO 2 emissions into the atmosphere cause greenhouse effects and hence global warming, its fixation has received much attention in recent years and has become a world challenge and Catalysts 2020, 10 priority [1]. Hence, intensive efforts have been devoted to mitigating CO 2 emissions through various strategies, such as CO 2 capture and storage [2,3] and chemical recycling [4][5][6][7][8][9][10][11]. Indeed, many industries consider CO 2 as a valuable resource rather than a waste [11][12][13]. However, it is necessary to find new paths enabling fuel production without releasing additional CO 2 in the atmosphere. In recent years, increased research efforts have been dedicated to the chemical conversion of CO 2 into useful chemicals such as methane, methanol, ethanol, and dimethyl ether through catalytic hydrogenation [14]. However, CO 2 methanation is considered to be one of the most promising ways of transforming CO 2 into methane, which is an important chemical feedstock as well as a fuel for power generation by gas turbines. However, the reaction temperature needs to be controlled since the exothermic character of methanation reaction makes the temperature control of methanation reactors difficult, especially under conditions of dynamic operation [15]. Plate heat-exchanger reactors offer benefits here of improved temperature control [16]. However, the amount of catalyst which can be incorporated into such a reactor is limited. Therefore, the development of catalysts of high activity is the key factor for the CO 2 methanation reaction at low temperature [17,18]. Though various catalysts have been developed for the catalytic hydrogenation of CO 2 , especially metal catalysts containing Ru [19,20], Rh [21][22][23], Pd [24], Pt [25], and Ni [26][27][28][29][30] have been proved to have good catalytic activity in CO 2 methanation reaction. Among them, nickel-based catalysts have been widely investigated due to their high catalytic activity, high selectivity for methane, and relatively low cost [14,26,[31][32][33][34]. Many researchers have reported that Ni catalysts were highly active for CO 2 methanation, including Ni/CeO 2 -ZrO 2 [35][36][37], Ni/SiO 2 [31], and Ni/Al 2 O 3 [15]. In general, most of the Ni-based catalysts were prepared by the conventional impregnation method, which has the advantages of a simple synthesis process, low cost, and an environmentally friendly procedure. However, large-size particles are usually formed through the traditional impregnation method, which leads to lower catalytic performance in the methanation reaction. Similarly, a strong metal-support interaction is also an important concern for Ni-based catalysts because the thermal sintering of metallic Ni active sites can easily take place under CO 2 methanation reaction conditions. In view of these challenges, preparing highly dispersed Ni catalyst on the supports with strong metal-support interaction can be considered to be the potential solution for effectively promoting the activity of CO 2 methanation, especially at low temperature [15]. Recent research shows that the addition of chelating ligands, for example citric acid (CA), enhances the promotion of the catalysts and resulted in the preparation of highly active Ni catalysts with small crystallite size and good dispersion [38,39]. Suárez-Toriello et al. [40] reported that the addition of CA to the impregnation solutions during the catalyst synthesis has led to a remarkable increased formation of the metal-support phase and as a result to increased catalytic activity. The Ni/SBA-15 catalyst prepared by the chelating ligand-assisted impregnation method yields the small size Ni particles and outstanding catalytic performance for CH 4 reforming with CO 2 [41]. In our previous work, we introduced sucrose as chelating ligand, which contributes majorly to the distribution of Ni species by preventing their sintering on SBA-15 silica support [42]. Therefore, size controlled and highly dispersed Ni crystallites with strong metal-support interaction need to be prepared for high CO 2 activation at low temperature.
It is also important to select the nature of support materials to improve the specific surface area of the active metal, reducibility of the catalyst, and dispersion of the active site, thereby affecting the catalytic performance of the CO 2 methanation reaction [43]. Since the morphology of the support has a significant influence in the catalytic properties, effective support with well-dispersed Ni particles must be prepared to meet these requirements. Alumina is known to be the most commonly employed support material due to its high thermal stability and strong resistance to attrition [44]. However, the resulting catalysts did not show superior catalytic stability especially at low reaction temperature mainly due to the sintering of Ni particles and deposition of carbon [45]. Indeed, nickel catalysts have been prepared on various supports such as Al 2 O 3 [15,17,44], SiO 2 [46][47][48][49], MCM-41 [26], SBA-15 [27], ZrO 2 [28] and CeO 2 [29,32,[50][51][52] and tested in methanation reaction. Among them, CeO 2 has recently emerged as the best support for metal-based catalysts and it has the capability of promoting the dispersion of Ni species and change metal properties through strong metal-support interactions. Le et al. [53] also studied the CO 2 methanation over Ni-based catalysts on some selected supports such as Al 2 O 3 , SiO 2 , TiO 2 , CeO 2 , and ZrO 2 . Owing to its high oxygen storage capacity and thermal stability CeO 2 could create the additional driving force for the CO 2 conversion at low temperature [54]. Tada et al. [29] investigated the CO 2 methanation over Ni catalysts loaded on different supports and found the following activity ranking: Ni/CeO 2 > Ni/α-Al 2 O 3 > Ni/TiO 2 > Ni/MgO. The formation of oxygen vacancies on the CeO 2 support by the redox circulation existing in the Ce 3+ /Ce 4+ ion pairs can activate the CO 2 molecules effectively to improve the CO 2 methanation [55]. Although some reports are available on Ni-based catalysts for the methanation reaction, those catalyst had been prepared by conventional impregnation methods and poor stability of the catalysts was observed, which impaired their performance of the Ni-based catalyst at low temperatures due to the formation of mobile nickel sub-carbonyl, which has to be removed quickly to maintain the catalytic activity. Moreover, these investigations had been carried out in packed bed reactors, which have the drawback of poor heat and mass transfer. In this work, we prepared various metal oxide-supported Ni catalyst by the impregnation method modified with chelating ligands. The catalytic performance of these catalysts was investigated in CO 2 methanation in microreactors and the results are presented.

Catalysts Characterization
X-ray diffraction was used to determine the crystalline structure of the developed catalysts and the results are shown in Figure 1. All the calcined samples exhibited the diffraction peaks at 2θ ≈ 37.2 • , 43.1 • , 62.8 • , 75.3 • and 79.4 • , which are attributed to (111), (200), (220), (311) and (222) facets of cubic crystalline structure of NiO, respectively [1,56]. However, the intensity of NiO peaks was too weak for the Ni/CeO 2 -CA sample, while that can be highly visible in the other samples, indicating a high degree of metal dispersion over the support surface. The difference of the peak intensity can be attributed to the different size and morphology of NiO present on the different supports. It is reported that the relatively small particle size of highly dispersed NiO on the support may also lead to the weak intensity of diffraction peaks [57]. The average crystallite sizes of the NiO were estimated from the main peak of (200) plane using the Scherrer equation. The average crystallite sizes of NiO were determined for Ni/CeO 2 , Ni/CeO 2 -CA, Ni/Al 2 O 3 -CA, Ni/SiO 2 -CA and Ni/SBA-15-CA and their results presented in Table 1 [58]. Additional peaks appeared in Ni/Al 2 O 3 -CA samples at 2θ ≈ 29.2 • , 45.8 • and 68.8 • corresponding to (220), (400) and (440) diffractions of crystalline alumina [40]. Ni/SiO 2 -CA and Ni/SBA-15-CA samples show no peaks corresponding to silica support due to amorphous nature of silica.  Nitrogen adsorption-desorption isotherms of Ni/CeO2-CA, Ni/Al2O3-CA, Ni/SiO2-CA and Ni/SBA-15-CA are shown in Figure 2. The adsorption branches of isotherms for Ni/CeO2-CA and Ni/SiO2-CA resemble that of type V with H1 hysteresis loops and are associated with the existence of uniform pores with a narrow size distribution. The Ni/Al2O3-CA catalyst shows a type IV isotherm with H1 type hysteresis loops, suggesting there are mesopores in the sample. The Ni/SBA-15-CA catalyst also exhibited IV-type isotherms with a well-expressed H1-type hysteresis loop, which is corresponding to a well-developed mesoporous structure. However, this catalyst shows two hysteresis loops due to the presence of both cylindrical and ink-bottle shape pores. It is assumed that the pores were significantly blocked by nickel species during the impregnation procedure. The surface area obtained for Ni/CeO2-CA, Ni/Al2O3-CA, Ni/SiO2-CA and Ni/SBA-15-CA catalysts from the N2 adsorption-desorption measurements and the results are presented in Table 1.   Figure 2. The adsorption branches of isotherms for Ni/CeO 2 -CA and Ni/SiO 2 -CA resemble that of type V with H1 hysteresis loops and are associated with the existence of uniform pores with a narrow size distribution. The Ni/Al 2 O 3 -CA catalyst shows a type IV isotherm with H1 type hysteresis loops, suggesting there are mesopores in the sample. The Ni/SBA-15-CA catalyst also exhibited IV-type isotherms with a well-expressed H1-type hysteresis loop, which is corresponding to a well-developed mesoporous structure. However, this catalyst shows two hysteresis loops due to the presence of both cylindrical and ink-bottle shape pores. It is assumed that the pores were significantly blocked by nickel species during the impregnation procedure. The surface area obtained for Ni/CeO 2 -CA, Ni/Al 2 O 3 -CA, Ni/SiO 2 -CA and Ni/SBA-15-CA catalysts from the N 2 adsorption-desorption measurements and the results are presented in Table 1. Nitrogen adsorption-desorption isotherms of Ni/CeO2-CA, Ni/Al2O3-CA, Ni/SiO2-CA and Ni/SBA-15-CA are shown in Figure 2. The adsorption branches of isotherms for Ni/CeO2-CA and Ni/SiO2-CA resemble that of type V with H1 hysteresis loops and are associated with the existence of uniform pores with a narrow size distribution. The Ni/Al2O3-CA catalyst shows a type IV isotherm with H1 type hysteresis loops, suggesting there are mesopores in the sample. The Ni/SBA-15-CA catalyst also exhibited IV-type isotherms with a well-expressed H1-type hysteresis loop, which is corresponding to a well-developed mesoporous structure. However, this catalyst shows two hysteresis loops due to the presence of both cylindrical and ink-bottle shape pores. It is assumed that the pores were significantly blocked by nickel species during the impregnation procedure. The surface area obtained for Ni/CeO2-CA, Ni/Al2O3-CA, Ni/SiO2-CA and Ni/SBA-15-CA catalysts from the N2 adsorption-desorption measurements and the results are presented in Table 1.  X-ray photoelectron spectroscopy (XPS) was used to investigate the nature and surface exposure of all the elements present in the developed catalysts. The oxidation states of Ni present on the support can be determined from the binding energy (BE) of Ni 2p 3/2 XPS spectrum. As can be seen in Figure 3A, all the samples exhibited the Ni 2p 3/2 BE between 855.2 and 855.7 eV along with the broad peak around 860.5 eV which are attributed to NiO species [38]. In case of the Ni/CeO 2 -CA, Ni/Al 2 O 3 -CA, Ni/SiO 2 -CA and Ni/SBA-15-CA catalysts, the main peak at 855.7 eV was slightly shifted to higher BE indicating that the nickel species might have strong interaction with the support. Figure 3B shows the XPS result of O 1s core electron levels for all the catalysts. The Ni/CeO 2 catalyst exhibited the main peak at 529.1 eV with two shoulders at 530 eV and 531.1 eV implying the presence of different oxide species. The main peak appeared at 529.1 eV and the shoulder peak at 530 eV can be attributed to the lattice oxygen and weakly bonded oxygen species present in CeO 2 [40]. The peak appearing at higher BE (531.1 eV) belongs to the surface absorbed oxygen. For the Ni/CeO 2 -CA sample similar observations indicate the combination of lattice oxygen and weakly bonded oxygen species and/or chemisorbed oxygen along with the shoulder peak (532.7 eV) corresponding to the adsorptive oxygen species present in this catalyst. The main peak appearing in the range between 530.5 eV and 532.2 eV for Ni/Al 2 O 3 -CA, Ni/SiO 2 -CA and Ni/SBA-15-CA samples could be attributed to adsorptive oxygen species, whereas the small peak at higher BE might be due the other oxygen species [39]. The presence of different oxygen species on all these supports due to their structure and chemical properties. The availability of high oxygen surface mobility by redox Ce 4+ /Ce 3+ on CeO 2 support could create more oxygen vacancies, which can enhance the catalytic reactions. X-ray photoelectron spectroscopy (XPS) was used to investigate the nature and surface exposure of all the elements present in the developed catalysts. The oxidation states of Ni present on the support can be determined from the binding energy (BE) of Ni 2p3/2 XPS spectrum. As can be seen in Figure 3A, all the samples exhibited the Ni 2p3/2 BE between 855.2 and 855.7 eV along with the broad peak around 860.5 eV which are attributed to NiO species [38]. In case of the Ni/CeO2-CA, Ni/Al2O3-CA, Ni/SiO2-CA and Ni/SBA-15-CA catalysts, the main peak at 855.7 eV was slightly shifted to higher BE indicating that the nickel species might have strong interaction with the support. Figure 3B shows the XPS result of O 1s core electron levels for all the catalysts. The Ni/CeO2 catalyst exhibited the main peak at 529.1 eV with two shoulders at 530 eV and 531.1 eV implying the presence of different oxide species. The main peak appeared at 529.1 eV and the shoulder peak at 530 eV can be attributed to the lattice oxygen and weakly bonded oxygen species present in CeO2 [40]. The peak appearing at higher BE (531.1 eV) belongs to the surface absorbed oxygen. For the Ni/CeO2-CA sample similar observations indicate the combination of lattice oxygen and weakly bonded oxygen species and/or chemisorbed oxygen along with the shoulder peak (532.7 eV) corresponding to the adsorptive oxygen species present in this catalyst. The main peak appearing in the range between 530.5 eV and 532.2 eV for Ni/Al2O3-CA, Ni/SiO2-CA and Ni/SBA-15-CA samples could be attributed to adsorptive oxygen species, whereas the small peak at higher BE might be due the other oxygen species [39]. The presence of different oxygen species on all these supports due to their structure and chemical properties. The availability of high oxygen surface mobility by redox Ce 4+ /Ce 3+ on CeO2 support could create more oxygen vacancies, which can enhance the catalytic reactions.
A B  Table 1 which was consistent with the particle sizes observed by X-ray diffraction (XRD). In the case of Ni/SBA-15-CA, the nickel particles arranged regularly in the nanochannels of the mesoporous SBA-15 support with much smaller particle size, which has the advantages of enhance the resistance of sintering due to arising of confinement effect from the wall of mesoporous SBA-15 [22]. TEM image of Ni/CeO 2 catalyst prepared without CA shows that NiO particles of larger size (9 nm) are aggregated around the CeO 2 support ( Figure S1). TEM-EELS images also confirm that the NiO particles are highly dispersed on the catalysts prepared by addition of CA ( Figure 5) than on the Ni/CeO 2 catalyst prepared without CA ( Figure 5A). The addition of CA as chelating ligand in the impregnation solution could form the chelated metal complexes and increase the viscosity of the solution, which prevents the particle aggregation and stabilizes the NiO species due to the presence of the steric hindrance effect [41]. This phenomenon proved that the presence of chelating ligand has increased the dispersion and controlled the size of the NiO particles. The TEM images of Ni/CeO 2 catalysts prepared with different chelating ligands such as sucrose, ethylene glycol, oxalic acid, and EDTA are presented in Figure S2. The Ni/CeO 2 -S and Ni/CeO 2 -OA catalysts show the uniform dispersion of NiO with average particle size of 8 nm, whereas the Ni/CeO 2 -OA and Ni/CeO 2 -EDTA catalysts possess the slight accretion of NiO with an average particle size of 11 nm.
results are shown in Figure 4. The TEM images of Ni/CeO2-CA, Ni/Al2O3-CA, Ni/SiO2-CA and Ni/SBA-15-CA catalyst show that the small size NiO particles are well dispersed on the support. The average particles sizes of NiO nanoparticles are presented in Table 1 which was consistent with the particle sizes observed by X-ray diffraction (XRD). In the case of Ni/SBA-15-CA, the nickel particles arranged regularly in the nanochannels of the mesoporous SBA-15 support with much smaller particle size, which has the advantages of enhance the resistance of sintering due to arising of confinement effect from the wall of mesoporous SBA-15 [22]. TEM image of Ni/CeO2 catalyst prepared without CA shows that NiO particles of larger size (9 nm) are aggregated around the CeO2 support ( Figure S1). TEM-EELS images also confirm that the NiO particles are highly dispersed on the catalysts prepared by addition of CA ( Figure 5) than on the Ni/CeO2 catalyst prepared without CA ( Figure 5A). The addition of CA as chelating ligand in the impregnation solution could form the chelated metal complexes and increase the viscosity of the solution, which prevents the particle aggregation and stabilizes the NiO species due to the presence of the steric hindrance effect [41]. This phenomenon proved that the presence of chelating ligand has increased the dispersion and controlled the size of the NiO particles. The TEM images of Ni/CeO2 catalysts prepared with different chelating ligands such as sucrose, ethylene glycol, oxalic acid, and EDTA are presented in Figure S2. The Ni/CeO2-S and Ni/CeO2-OA catalysts show the uniform dispersion of NiO with average particle size of 8 nm, whereas the Ni/CeO2-OA and Ni/CeO2-EDTA catalysts possess the slight accretion of NiO with an average particle size of 11 nm.

Catalytic Performance Test
The effect of chelating ligand and different supports on the performance of the developed Nibased catalysts was investigated for CO2 methanation. The catalytic activity was tested under different reaction temperature on the catalytic conversion of CO2 over Ni/CeO2, Ni/CeO2-CA, Ni/Al2O3-CA, Ni/SiO2-CA and Ni/SBA-15-CA catalyst between 275 °C and 375 °C. All the catalysts were reduced with H2 gas at 450 °C prior to the CO2 methanation tests. Figure 6 displays the percentage of CO2 conversion against the reaction temperature for all the catalysts. It can be seen that the CO2 conversion increases with the reaction temperature gradually then reached a maximum value, except for the Ni/CeO2-CA catalyst. It is worth mentioning that owing to thermodynamic limitations, no significant increase of the conversion could be observed above 375 °C with a maximal CO2 conversion of 96.3% for all the catalysts. Methane was determined to be a major product for all the catalysts. In addition to methane, small quantities of carbon monoxide were also detected especially at low temperatures. Among all the tested catalysts, the Ni/CeO2-CA catalyst exhibited the best CO2 conversion of maximum 98.6% already at 325 °C when compared with the other catalysts. A further raise in temperature led to a slight decrease in CO2 conversion due to the exothermicity of methanation reaction. Since the high temperature has a negative effect on CO2 methanation reaction, 325 °C could be the optimum temperature to achieve maximum CO2 conversion and high selectivity towards CH4 for this catalyst under the conditions of this test. The Ni/CeO2-CA catalyst yields highest CO2 conversion of all samples at a low reaction temperature of 325 °C [34]. Among the catalysts investigated, the citric acid addition in Ni/CeO2 catalyst preparation showed substantial improvement of the catalytic performance compared to the Ni/CeO2 catalyst prepared without addition of CA. Apparently, the Ni/CeO2-CA catalyst exhibited much higher CO2 conversion than Ni/CeO2. As confirmed by the TEM results (Figure 4), the Ni dispersion could be the critical factor to control the catalytic activity in these Ni catalysts. It implied that the addition of chelating ligand enhanced the dispersion of Ni species and control the particle size, resulting in more exposed active species, which was responsible for the superior catalytic activity [41]. The carboxylic groups of CA could deprotonate easily and generate the citrate ligands, which are capable of coordinating with nickel ions and prevents the aggregation of large-size Ni nanoparticles during the calcination process. Suárez-Toriello et al. [40] reported that the addition of CA drastically changed the chemical equilibria in the solution mixture and the deprotonation of the carboxylic acids groups increased the amount of citrate ligands, which is capable of forming Ni-citrate complexes. The catalytic performance of all

Catalytic Performance Test
The effect of chelating ligand and different supports on the performance of the developed Ni-based catalysts was investigated for CO 2 methanation. The catalytic activity was tested under different reaction temperature on the catalytic conversion of CO 2 over Ni/CeO 2 , Ni/CeO 2 -CA, Ni/Al 2 O 3 -CA, Ni/SiO 2 -CA and Ni/SBA-15-CA catalyst between 275 • C and 375 • C. All the catalysts were reduced with H 2 gas at 450 • C prior to the CO 2 methanation tests. Figure 6 displays the percentage of CO 2 conversion against the reaction temperature for all the catalysts. It can be seen that the CO 2 conversion increases with the reaction temperature gradually then reached a maximum value, except for the Ni/CeO 2 -CA catalyst. It is worth mentioning that owing to thermodynamic limitations, no significant increase of the conversion could be observed above 375 • C with a maximal CO 2 conversion of 96.3% for all the catalysts. Methane was determined to be a major product for all the catalysts. In addition to methane, small quantities of carbon monoxide were also detected especially at low temperatures. Among all the tested catalysts, the Ni/CeO 2 -CA catalyst exhibited the best CO 2 conversion of maximum 98.6% already at 325 • C when compared with the other catalysts. A further raise in temperature led to a slight decrease in CO 2 conversion due to the exothermicity of methanation reaction. Since the high temperature has a negative effect on CO 2 methanation reaction, 325 • C could be the optimum temperature to achieve maximum CO 2 conversion and high selectivity towards CH 4 for this catalyst under the conditions of this test. The Ni/CeO 2 -CA catalyst yields highest CO 2 conversion of all samples at a low reaction temperature of 325 • C [34]. Among the catalysts investigated, the citric acid addition in Ni/CeO 2 catalyst preparation showed substantial improvement of the catalytic performance compared to the Ni/CeO 2 catalyst prepared without addition of CA. Apparently, the Ni/CeO 2 -CA catalyst exhibited much higher CO 2 conversion than Ni/CeO 2 . As confirmed by the TEM results (Figure 4), the Ni dispersion could be the critical factor to control the catalytic activity in these Ni catalysts. It implied that the addition of chelating ligand enhanced the dispersion of Ni species and control the particle size, resulting in more exposed active species, which was responsible for the superior catalytic activity [41]. The carboxylic groups of CA could deprotonate easily and generate the citrate ligands, which are capable of coordinating with nickel ions and prevents the aggregation of large-size Ni nanoparticles during the calcination process. Suárez-Toriello et al. [40] reported that the addition of CA drastically changed the chemical equilibria in the solution mixture and the deprotonation of the carboxylic acids groups increased the amount of citrate ligands, which is capable of forming Ni-citrate complexes. The catalytic performance of all the catalysts studied at different reaction temperatures is in the following order: Ni/CeO 2 -CA > Ni/SBA-15-CA > Ni/Al 2 O 3 -CA > Ni/SiO 2 -CA > Ni/CeO 2 . As shown in Figure 6, ceria-supported Ni catalyst prepared by the CA modified method showed higher CO 2 conversion than Al 2 O 3 , SiO 2 , and SBA-15-supported Ni catalysts. The higher catalytic activity of Ni/CeO 2 -CA is associated with the medium basicity of CeO 2 support, compared to other supports used in this work, allowing stronger CO 2 adsorption. The adsorbed CO 2 on CeO 2 support can be reduced easily due to the sufficient amount of oxygen vacancies available on the CeO 2 support. Pan et al. [36] reported that Ce 0.5 Zr 0.5 O 2 -supported Ni catalyst is highly active when compared with Ni/γ-Al 2 O 3 . Similarly, Tada et al. [29] investigated the catalytic performance of different supported Ni catalysts on CO 2 methanation and reported that the partial reduction of the CeO 2 surface, which was covered by CO 2 -derived species are responsible to yield higher CO 2 conversion over Ni/CeO 2 compared with Ni/γ-Al 2 O 3 , Ni/TiO 2 , and Ni/MgO. When compared with other supports the unique superiority of CeO 2 is that the existence of Ce 3+ /Ce 4+ ion pairs on its surface can enhance the formation of the oxygen vacancies, which can promote the adsorption and activation of CO 2 [32,55]. The SBA-15-supported Ni catalyst with highest surface area of 380 m 2 /g showed however lower activity than Ni/CeO 2 -CA with a lower surface area of 44 m 2 /g, indicating that the surface area of the developed catalyst has no significant effect on the catalytic activity. As seen in Figure 7, all the developed catalysts show high selectivity towards methane in the temperature range of 275−375 • C. Compared with Ni/CeO 2 , Ni/CeO 2 -CA exhibits higher selectivity towards methane. The maximum selectivity of methane was found to be 99.7% for Ni/CeO 2 -CA. the catalysts studied at different reaction temperatures is in the following order: Ni/CeO2-CA > Ni/SBA-15-CA > Ni/Al2O3-CA > Ni/SiO2-CA > Ni/CeO2. As shown in Figure 6, ceria-supported Ni catalyst prepared by the CA modified method showed higher CO2 conversion than Al2O3, SiO2, and SBA-15-supported Ni catalysts. The higher catalytic activity of Ni/CeO2-CA is associated with the medium basicity of CeO2 support, compared to other supports used in this work, allowing stronger CO2 adsorption. The adsorbed CO2 on CeO2 support can be reduced easily due to the sufficient amount of oxygen vacancies available on the CeO2 support. Pan et al. [36] reported that Ce0.5Zr0.5O2supported Ni catalyst is highly active when compared with Ni/-Al2O3. Similarly, Tada et al. [29] investigated the catalytic performance of different supported Ni catalysts on CO2 methanation and reported that the partial reduction of the CeO2 surface, which was covered by CO2-derived species are responsible to yield higher CO2 conversion over Ni/CeO2 compared with Ni/-Al2O3, Ni/TiO2, and Ni/MgO. When compared with other supports the unique superiority of CeO2 is that the existence of Ce 3+ /Ce 4+ ion pairs on its surface can enhance the formation of the oxygen vacancies, which can promote the adsorption and activation of CO2 [32,55]. The SBA-15-supported Ni catalyst with highest surface area of 380 m 2 /g showed however lower activity than Ni/CeO2-CA with a lower surface area of 44 m 2 /g, indicating that the surface area of the developed catalyst has no significant effect on the catalytic activity. As seen in Figure 7, all the developed catalysts show high selectivity towards methane in the temperature range of 275375 °C. Compared with Ni/CeO2, Ni/CeO2-CA exhibits higher selectivity towards methane. The maximum selectivity of methane was found to be 99.7% for Ni/CeO2-CA.  To investigate the influence of chelating ligands on catalytic performance, the methanation reaction was performed over Ni/CeO2 catalysts prepared by using various chelating ligands such as sucrose (S), EG, OA, and EDTA and the obtained results are shown in Figure 8. The CO2 conversion for Ni/CeO2-S and Ni/CeO2-EG increased with increasing reaction temperature up to 325 °C and remained then stable close to complete conversion with further increasing reaction temperature. The selectivity towards methane is very close to 100% for these catalysts at all reaction temperatures ( Figure 8). Notably, the conversion of CO2 and selectivity of methane was initially much lower and increased then with increasing reaction temperature for the Ni/CeO2-OA and Ni/CeO2-EDTA catalysts. Among the catalysts tested, Ni/CeO2-S and Ni/CeO2-EG exhibited much higher CO2 conversions around 97.5% and 94.8%, respectively, at 325 °C. However, the CO2 conversion obtained for these catalysts was slightly lower than for the Ni/CeO2-CA ( Figure 6). Ni/CeO2-OA and Ni/CeO2-EDTA catalysts showed poorer catalytic activity under the same conditions, indicating that the nature of chelating ligands had affected the dispersion of Ni species on the support which was confirmed by TEM measurement ( Figure S2). To investigate the influence of chelating ligands on catalytic performance, the methanation reaction was performed over Ni/CeO 2 catalysts prepared by using various chelating ligands such as sucrose (S), EG, OA, and EDTA and the obtained results are shown in Figure 8. The CO 2 conversion for Ni/CeO 2 -S and Ni/CeO 2 -EG increased with increasing reaction temperature up to 325 • C and remained then stable close to complete conversion with further increasing reaction temperature. The selectivity towards methane is very close to 100% for these catalysts at all reaction temperatures ( Figure 8). Notably, the conversion of CO 2 and selectivity of methane was initially much lower and increased then with increasing reaction temperature for the Ni/CeO 2 -OA and Ni/CeO 2 -EDTA catalysts. Among the catalysts tested, Ni/CeO 2 -S and Ni/CeO 2 -EG exhibited much higher CO 2 conversions around 97.5% and 94.8%, respectively, at 325 • C. However, the CO 2 conversion obtained for these catalysts was slightly lower than for the Ni/CeO 2 -CA ( Figure 6). Ni/CeO 2 -OA and Ni/CeO 2 -EDTA catalysts showed poorer catalytic activity under the same conditions, indicating that the nature of chelating ligands had affected the dispersion of Ni species on the support which was confirmed by TEM measurement ( Figure S2). To investigate the influence of chelating ligands on catalytic performance, the methanation reaction was performed over Ni/CeO2 catalysts prepared by using various chelating ligands such as sucrose (S), EG, OA, and EDTA and the obtained results are shown in Figure 8. The CO2 conversion for Ni/CeO2-S and Ni/CeO2-EG increased with increasing reaction temperature up to 325 °C and remained then stable close to complete conversion with further increasing reaction temperature. The selectivity towards methane is very close to 100% for these catalysts at all reaction temperatures ( Figure 8). Notably, the conversion of CO2 and selectivity of methane was initially much lower and increased then with increasing reaction temperature for the Ni/CeO2-OA and Ni/CeO2-EDTA catalysts. Among the catalysts tested, Ni/CeO2-S and Ni/CeO2-EG exhibited much higher CO2 conversions around 97.5% and 94.8%, respectively, at 325 °C. However, the CO2 conversion obtained for these catalysts was slightly lower than for the Ni/CeO2-CA ( Figure 6). Ni/CeO2-OA and Ni/CeO2-EDTA catalysts showed poorer catalytic activity under the same conditions, indicating that the nature of chelating ligands had affected the dispersion of Ni species on the support which was confirmed by TEM measurement ( Figure S2). Owing to the highly exothermic nature of CO2 methanation, the development of Ni catalyst must be stable enough against the thermal sintering to avoid deactivation of the catalyst. The catalytic stability of the Ni/CeO2-CA, Ni/SBA-15-CA, Ni/Al2O3-CA and Ni/SiO2-CA catalysts was investigated for duration of 100 h at 325 °C. As shown in Figure 9, the Ni/CeO2-CA catalyst exhibited highest CO2 conversion. However, the conversion decreased slowly with time and ended up with a value of 85% CO2 conversion after 100 h on stream. It is well known that the methanation catalysts are also subject to carbon deposition on the active sites of the Ni nanoparticles that leads the deactivation of the catalysts. To investigate the carbon deposition on these spent catalysts, TEM and Thermogravimetric analysis (TGA) were performed, and their results are presented in Figure 10. From TEM images, a thin layer of carbon was identified on the surface of these catalysts. However, the TGA result shows no weight loss curve of carbon. Therefore, carbon deposition seems to be very less and that could not be the main reason for the catalytic deactivation. The TEM-EELS image ( Figure 5F) of spent Ni/CeO2-CA catalyst confirms that the Ni particles were not much agglomerated under this reaction conditions. This result indicates that the Ni/CeO2-CA catalyst presented comparatively better activity under this reaction conditions. As seen in Figure 8, Ni/Al2O3-CA and Ni/SBA-15-CA catalysts show considerably stable performance than Ni/CeO2-CA catalyst for 100 h of time on stream. However, the overall CO2 conversion was found to be maximum 70%, which is lower when compare with Ni/CeO2-CA catalyst. The conversion over the Ni/SiO2-CA decreased drastically with time on stream. According to TEM and TGA results, the Ni/Al2O3-CA and Ni/SBA-15-CA catalysts are resistant against high coke formation and shows stable performance which is attributed to the strong metal−support interaction and lower sintering under methanation reaction conditions, thereby leading to long-term catalytic stability. Owing to the highly exothermic nature of CO 2 methanation, the development of Ni catalyst must be stable enough against the thermal sintering to avoid deactivation of the catalyst. The catalytic stability of the Ni/CeO 2 -CA, Ni/SBA-15-CA, Ni/Al 2 O 3 -CA and Ni/SiO 2 -CA catalysts was investigated for duration of 100 h at 325 • C. As shown in Figure 9, the Ni/CeO 2 -CA catalyst exhibited highest CO 2 conversion. However, the conversion decreased slowly with time and ended up with a value of 85% CO 2 conversion after 100 h on stream. It is well known that the methanation catalysts are also subject to carbon deposition on the active sites of the Ni nanoparticles that leads the deactivation of the catalysts. To investigate the carbon deposition on these spent catalysts, TEM and Thermogravimetric analysis (TGA) were performed, and their results are presented in Figure 10. From TEM images, a thin layer of carbon was identified on the surface of these catalysts. However, the TGA result shows no weight loss curve of carbon. Therefore, carbon deposition seems to be very less and that could not be the main reason for the catalytic deactivation. The TEM-EELS image ( Figure 5F) of spent Ni/CeO 2 -CA catalyst confirms that the Ni particles were not much agglomerated under this reaction conditions. This result indicates that the Ni/CeO 2 -CA catalyst presented comparatively better activity under this reaction conditions. As seen in Figure 8, Ni/Al 2 O 3 -CA and Ni/SBA-15-CA catalysts show considerably stable performance than Ni/CeO 2 -CA catalyst for 100 h of time on stream. However, the overall CO 2 conversion was found to be maximum 70%, which is lower when compare with Ni/CeO 2 -CA catalyst. The conversion over the Ni/SiO 2 -CA decreased drastically with time on stream. According to TEM and TGA results, the Ni/Al 2 O 3 -CA and Ni/SBA-15-CA catalysts are resistant against high coke formation and shows stable performance which is attributed to the strong metal−support interaction and lower sintering under methanation reaction conditions, thereby leading to long-term catalytic stability. Catalysts 2020, 10, x FOR PEER REVIEW 11 of 16

Catalysts Preparation
The Ni catalysts supported on different supports such as CeO2, Al2O3, SiO2, and SBA-15 were prepared by a impregnation method modified with different chelating ligands such as CA, ethylene glycol (EG), sucrose (S), oxalic acid (OA) and EDTA. Ni(NO3)2·6H2O was used as the metal oxide precursor and the commercial CeO2 (99.9%, Sigma-Aldrich, Steinheim am Albuch, Germany), Al2O3 (99.99%, Sigma-Aldrich, Steinheim am Albuch, Germany) and SiO2 (99.8%, Sigma-Aldrich, Steinheim am Albuch, Germany) were used as supports. The SBA-15 support was prepared by a method describing in the literature [42]. To prepare 20 wt. % Ni catalysts, 2.92 g of Ni(NO3)2·6H2O and 2.89 g of citric acid (Ni:CA = 1:1.5 ratio) were dissolved in 20 ml of water. Then, 3 g of the support were added and then the mixtures were stirred at a constant speed for 5 h. The mixtures were dried at 80 °C for 6 h and 120 °C for 12 h. The brownish power was carbonized at 500 °C for 4 h in N2 followed by calcination at the same temperature for 3 h in the presence of air. All other Ni catalysts on different supports were prepared using CA by fixing the molar ratio of Ni to CA to 1:1.5. Similarly, Ni/CeO2 catalysts with different chelating ligands such as EG, S, OA, and EDTA were also prepared by following the same procedure with fixed molar ratio of Ni to chelating ligands to 1:1.5.

Catalysts Preparation
The Ni catalysts supported on different supports such as CeO2, Al2O3, SiO2, and SBA-15 were prepared by a impregnation method modified with different chelating ligands such as CA, ethylene glycol (EG), sucrose (S), oxalic acid (OA) and EDTA. Ni(NO3)2·6H2O was used as the metal oxide precursor and the commercial CeO2 (99.9%, Sigma-Aldrich, Steinheim am Albuch, Germany), Al2O3 (99.99%, Sigma-Aldrich, Steinheim am Albuch, Germany) and SiO2 (99.8%, Sigma-Aldrich, Steinheim am Albuch, Germany) were used as supports. The SBA-15 support was prepared by a method describing in the literature [42]. To prepare 20 wt. % Ni catalysts, 2.92 g of Ni(NO3)2·6H2O and 2.89 g of citric acid (Ni:CA = 1:1.5 ratio) were dissolved in 20 ml of water. Then, 3 g of the support were added and then the mixtures were stirred at a constant speed for 5 h. The mixtures were dried at 80 °C for 6 h and 120 °C for 12 h. The brownish power was carbonized at 500 °C for 4 h in N2 followed by calcination at the same temperature for 3 h in the presence of air. All other Ni catalysts on different supports were prepared using CA by fixing the molar ratio of Ni to CA to 1:1.5. Similarly, Ni/CeO2 catalysts with different chelating ligands such as EG, S, OA, and EDTA were also prepared by following the same procedure with fixed molar ratio of Ni to chelating ligands to 1:1.5.

Catalysts Preparation
The Ni catalysts supported on different supports such as CeO 2 , Al 2 O 3 , SiO 2 , and SBA-15 were prepared by a impregnation method modified with different chelating ligands such as CA, ethylene glycol (EG), sucrose (S), oxalic acid (OA) and EDTA. Ni(NO 3 ) 2 ·6H 2 O was used as the metal oxide precursor and the commercial CeO 2 (99.9%, Sigma-Aldrich, Steinheim am Albuch, Germany), Al 2 O 3 (99.99%, Sigma-Aldrich, Steinheim am Albuch, Germany) and SiO 2 (99.8%, Sigma-Aldrich, Steinheim am Albuch, Germany) were used as supports. The SBA-15 support was prepared by a method describing in the literature [42]. To prepare 20 wt. % Ni catalysts, 2.92 g of Ni(NO 3 ) 2 ·6H 2 O and 2.89 g of citric acid (Ni:CA = 1:1.5 ratio) were dissolved in 20 ml of water. Then, 3 g of the support were added and then the mixtures were stirred at a constant speed for 5 h. The mixtures were dried at 80 • C for 6 h and 120 • C for 12 h. The brownish power was carbonized at 500 • C for 4 h in N 2 followed by calcination at the same temperature for 3 h in the presence of air. All other Ni catalysts on different supports were prepared using CA by fixing the molar ratio of Ni to CA to 1:1.5. Similarly, Ni/CeO 2 catalysts with different chelating ligands such as EG, S, OA, and EDTA were also prepared by following the same procedure with fixed molar ratio of Ni to chelating ligands to 1:1.5.
To make the catalysts suspension to fill in the microreactors, 1.25 g of Polyvinyl alcohol (PVA) was dissolved in water under constant stirring at 65 • C for 3 h. Then, 2.5 g catalyst powder and 0.19 g of acetic acid was added under the same conditions and continued the stirring for another 3 h. After stirring the mixture for three days at room temperature a homogenous suspension was obtained, which was then filled in the microchannels of the microreactors and dried at room temperature and calcined at 450 • C for 6 h in air [59].

Catalysts Characterization
XRD patterns were recorded on a Stoe Stadi P diffractometer using Cu Kα 1 radiation from a sealed tube x-ray source operating at 40 kV and 30mA. N 2 adsorption-desorption isotherms of the developed catalysts were obtained from Sorptomatic 1900 analyzer. To remove the moistures the catalysts were degassed at 150 • C for a few hours before the measurement. Specific surface areas were obtained by the Brunauer-Emmett-Teller (BET) method in the relative pressure range of P/P 0 = 0.0-1.0. The surface nature of the developed catalysts was investigated by XPS using a multi-chamber Ultra high vacuum (UHV) system (PREVAC, Poland). The system equipped with a monochromated Al source (XM 650 X-ray monochromator) source and operated at 360 W. The constant pass energy of 200 eV was fixed to collect the survey scans. High-resolution scans of the separate regions were measured at 50 eV. The background pressure for ultra-high vacuum chamber was fixed as 5 × 10 −8 mbar and the calibration was done at 284.7 eV by setting the position of the C1s carbon line. CasaXPS (ver. 2.3.16 PR 1.6) software was used to process the recorded spectra. A LIBRA 120 microscope was used to collect the TEM images by suspending a small amount of catalysts in ethanol/water and placed onto the copper grids. Thermogravimetric analysis was carried out in a Mettler Toledo, TGA/DSC 1100.

Catalytic Activity Measurements
The CO 2 methanation tests were carried out in a microchannel reactor in the temperature range between 275 • C and 375 • C. The reaction mixture of H 2 :CO 2 :CH 4 (molar ratio of 4:1:1) was introduced into the test microreactors through mass flow controllers. The flow rate of the reaction mixture was fixed to 45 L/h g catalyst and experiments were performed at a pressure of 12.5 bar. The sandwich type of three microchannel platelets of microreactors were coated with the developed Ni catalysts and subsequently sealed by laser welding along with inlet and outlet capillaries [60]. Before starting the experiment, the catalysts were reduced by passing H 2 gas at 400 • C for 2 h, then the microreactor was by-passed by reaction mixture until a stable feed composition were reached. The obtained product composition was analyzed by on−line Gas chromatography (GC, Agilent Technologies 7890A GC System).

Conclusions
In this work, Ni catalysts supported on different supports were prepared by chelating ligand-assisted impregnation method and their catalytic activity was investigated by CO 2 methanation at different temperatures. A CeO 2 -supported Ni catalyst was determined to be the most active for CO 2 methanation compared with Ni catalysts supported on different support such as Al 2 O 3 , SiO 2 , and SBA-15. The formation of chelated metal complexes by strong interaction of CA chelating ligand with nickel ions inhibited the accumulation of the nickel particles and contributed to the well-dispersed Ni particles. The Ni/CeO 2 -CA catalyst exhibited the highest CO 2 conversion among all the catalysts with 99.7% of CH 4 selectivity. The higher catalytic performance was closely related to the enhanced metal-support interactions between Ni and Ce species, the size of Ni nanoparticles, and the synergistic effect of the nickel and ceria supports. Furthermore, the medium basicity of CeO 2 is beneficial for strong adsorption of CO 2 and resulted in high catalytic activity. In the stability test, the Ni/CeO 2 -CA catalyst yields high CO 2 conversion even though there is slight deactivation during 100 h of time on stream. The CO 2 conversion was above 85% at the end of 100 h of time on stream. From the observed results, it can be concluded that ceria-supported Ni catalyst prepared via the citric acid-assisted method can be applied to keep the metal active at lower reaction temperatures in any other thermodynamically controlled exothermic reactions.