The Effect of Preparation Method of Ni-Supported SiO 2 Catalysts for Carbon Dioxide Reforming of Methane

: Reforming methane to produce syngas is a subject that generates considerable interest. The process requires catalysts that possess high-performance active sites to activate stable C–H bonds. Herein, we report a facile synthetic strategy to prepare Ni-based catalysts by complexation– impregnation (Ni-G/SiO 2 -C) and precipitation–impregnation (Ni-G/SiO 2 -P) methods using glycine as a complexing agent. The particle size of Ni in both types of catalysts is decreased by adding glycine in the preparation process. Nevertheless, the preparation methods and amount of glycine play a signiﬁcant role in the particle size and distribution of Ni over the Ni-based catalysts. The smaller particle size and narrower distribution of Ni were obtained in the Ni-G/SiO 2 -P catalyst. The catalysts were comparatively tested for carbon-dioxide reforming of methane (CDR). Ni-G/SiO 2 -P showed better CDR performance than Ni-G/SiO 2 -C and Ni/SiO 2 and increased stability because of the smaller particle size and narrower distribution of Ni. Moreover, a high-performance Ni-based catalyst was prepared by optimizing the amount of glycine added. An unobservable deactivation was obtained over Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P for CDR during TOS = 20 h. Thus, a new promising method is described for the preparation of Ni-based catalysts for CDR.


Introduction
With the rapid development of technology and social economy in recent years, the greenhouse effect caused by the burning of fossil fuels has become progressively serious [1,2].CO 2 is one of the main greenhouse gases.In this connection, the reduction and effective use of CO 2 has attracted significant attention.Carbon dioxide reforming of methane (CDR) technology has also been substantially examined [2][3][4].Except for the simultaneous use of the two major greenhouse gases, CO 2 and CH 4 , the H 2 /CO ratio of the produced syngas is close to 1, which can be used to synthesize carbonyls, organic oxygenates, and liquid fuel by Fischer-Tropsch synthesis [2,[4][5][6].More importantly, CDR technology can directly convert CH 4 and CO 2 of flue gas into synthesis gas without pre-separation of CO 2 and CH 4 [3,[7][8][9] Therefore, accelerating the industrialization process of CDR reaction is very important for CO 2 emission reduction and efficient utilization.
The catalysts applied in CDR are mainly noble metal (Rh, Ru, Pt, and Pd) and transition metal (Ni, Co, and Fe) catalysts [10,11].Although noble metal catalysts have excellent catalytic activity and stability, their high cost limits their industrial application.Ni-based catalysts are the most promising catalysts for CDR because of their high catalytic activity and low cost [12][13][14][15].However, the sintering at high temperatures and the serious coking of the Ni-based catalysts lead to rapid deactivation, limiting its industrial application development [16].Therefore, reasonably designed Ni-based catalysts with high antisintering and anti-coking capacities are very urgent.
A comprehensive analysis of related literature reports indicates that the particle size of Ni is an essential factor affecting the formation of carbon deposits.Small Ni particles can inhibit the formation of coking deposits and affect CH 4 decomposition [10,11,14,15].Moreover, the sintering of the active component Ni at high-temperature reduction and reaction conditions facilitates the formation of carbon deposits, which can cover the active site of Ni, resulting in the deactivation of catalysts.Therefore, the key to improving the catalytic performance of the Ni-based catalysts for the CDR is to decrease the Ni particle size and prevent the sintering of Ni at higher temperatures [10,14,15].
Among the above methods, the co-impregnation method of nickel nitrate and a chelating agent has been proven to be an effective method to reduce the Ni particle size.Zhao et al. successfully prepared Ni/ZrO 2 catalysts with smaller Ni particle sizes by using L-arginine as a complexing agent, which showed excellent anti-sintering and anti-coking performance, resulting in higher catalytic activity and stability for CDR [34].Moreover, many organic ligands, such as ammonia [35,36], citric acid [37][38][39], glucose [40], and glycine [41][42][43], etc., have been used as complexing agents to prepare Ni-based catalysts.The basis of these methods is to replace the H 2 O ligand in the metal complex with an organic ligand, resulting in the complex of metal ions with the chelating agent.The structure and coordination characteristics of the complexes complicate during impregnation and are influenced by the ratio of the chelating agent and metal ion and the pH of the solution [42,44].
Glycine is the smallest amino acid, which is inexpensive and readily available, making it a promising chelating agent for synthesizing highly dispersed metal-supported catalysts [36,38,42].Glycine molecule contains both amino and carboxyl coordination groups.Metal nitrates or chloride salts are generally not chosen as the metal precursor to synthesize the complex because they would release H + in the aqueous solution, generating protonated forms of the amino acid in significant quantity.In this work, nickel hydroxide was used as the precursor to avoid the pH change when added to the aqueous solution [42].Although the solid nickel hydroxide is insoluble in the impregnating solution, it can be dissolved by complexing with the amino acid to form a complex.Thus, the complex synthesized by glycine and nickel hydroxide was used as the impregnating solution to prepare the SiO 2 -supported Ni-based catalyst.For comparison, supported Ni catalysts were also prepared by the traditional impregnation method and the co-impregnation of nickel nitrate and glycine.The effects of the preparation method, coordination characteristics, and coordination number on the Ni particle size and dispersion were investigated.The CDR performance of the prepared Ni-based catalysts was evaluated under 1.0 atm pressure, 750 • C temperature, CO 2 /CH 4 ratio of 1.0, and GHSV of 120,000 mL•g −1 •h −1 .Regardless of the preparation method, the smaller Ni particle size and higher dispersion of Ni were obtained after complex impregnation with glycine, compared with the traditional impregnation method.However, the preparation method significantly influenced Ni particle size and particle distribution.The catalyst prepared by glycine and nickel hydroxide exhibited the smallest Ni particle size and narrower size distribution, resulting in higher activity and stability in the CDR reaction.Thus, a new preparation method of Ni-based catalysts for CDR is presented, which has more industrial application prospects.

Textural Properties of Ni-Based Catalysts
Table 1 lists the textural properties of all catalysts, including SiO 2 , such as specific surface area (S BET ), pore volume (V p ), and pore diameter (d p ), tested by the N 2 adsorptiondesorption isotherms.SiO 2 exhibits the highest BET surface area (231.3 m 2 •g −1 ).After the loading of Ni, the BET surface area, pore volume, and pore diameter of all catalysts significantly decreased due to the deposition of NiO on the pore walls of the support Q-50.Careful comparison of the above properties of the prepared catalysts shows a negligible difference in the pore volume and pore diameter, i.e., 0.89-0.93cm 3 •g −1 of pore volumes and 20.6-21.0 nm of pore diameter.In contrast, an observable difference of S BET was obtained for all Ni-supported catalysts.Compared with Ni/SiO 2 (176.9 m 2 •g −1 ), larger S BET was obtained for Ni-G/SiO 2 -C (188.3 m 2 •g −1 ) and Ni-G/SiO 2 -P (193.6-201.8m 2 •g −1 ).Some of the new pores are formed by the decomposition of the frame of Ni-glycine complexes of Ni-G/SiO 2 -P and Ni-G/SiO 2 -C.In addition, the smaller NiO particle size in Ni-G/SiO 2 -P and Ni-G/SiO 2 -C contributes to the larger surface area, which is consistent with the results of XRD (Figure 1).For Ni-G/SiO 2 -P catalysts, with an increasing molar ratio of glycine to nickel, the S BET slightly increased, which shows that the amount of glycine used for complexation plays an unimportant role in the S BET of Ni-G/SiO 2 -P.Since metallic Ni is the active species for the CDR, the Ni particle size was also characterized by X-ray diffraction after reducing the catalysts at 700 o C for 2.5 h, and the results are shown in Figure 2. All the reduced catalysts exhibited the diffraction peaks of amorphous silica, suggesting that the SiO2 support did not change by the reduction conditions.In addition, the characteristic diffraction peaks of metallic Ni with different intensities were obtained at 2θ values of 44°, 52°, and 76° for Ni/SiO2, Ni-G-2/SiO2-C, and Ni-G-1/SiO2-P, indicating that the NiO reduced to Ni at the reduction conditions [39].However, no diffraction peaks of Ni species were observed for Ni-G-2/SiO2-P and Ni-G-3/SiO2-P, suggesting that particle sizes were too small to be detected by XRD.These results demonstrate that the Ni particle size was significantly influenced by the preparation method and

Structural Properties of Ni-Based Catalysts
The structural properties of all catalysts were characterized by X-ray diffraction and TEM, and the respective XRD patterns and TEM images are shown in Figures 1-3.All catalysts exhibited a broad diffraction peak at the 2θ of about 23 • (Figure 1), the characteristic diffraction peak of amorphous silica [21], implying that the utilized preparation methods of supported Ni catalysts do not affect the structure of the support.Moreover, clear diffraction peaks at 2θ values of 37 • , 43 • , 62 • , and 74.5 • were obtained for Ni/SiO 2 and Ni-G-2/SiO 2 -C (Figure 1a,b), attributed to the (111), ( 200), (220), and (311) cubic NiO planes, respectively [34,39].The diffraction peaks of Ni-G-2/SiO 2 -C were significantly weaker than Ni/SiO 2 , indicating that relatively smaller-sized NiO nanoparticles were formed on the Ni-G-2/SiO 2 -C catalyst because the Ni 2+ complexed with glycine during the impregnation process.Catalyst Ni-G-1/SiO 2 -P showed even weaker peaks of cubic NiO diffraction, at 2θ values of 37 • , 43 • , and 63 • , whereas no observable peaks of Ni species were obtained for Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P.These observations indicate that Ni species were dispersed well in Ni-G/SiO 2 -P than Ni-G/SiO 2 -C and Ni/SiO 2 .The better dispersion occurs due to the increased viscosity of the impregnating solution caused by the addition of glycine that suppresses the outward flow of the solution, leading to a more uniform distribution of Ni species inside the pores of the support [36,41,42].Moreover, the formation of the Ni complex prevents the aggregation of Ni particles during the hightemperature calcination process, affording smaller NiO particles [41], and therefore, no obvious diffraction peaks were observed in XRD.Since metallic Ni is the active species for the CDR, the Ni particle size was also characterized by X-ray diffraction after reducing the catalysts at 700 o C for 2.5 h, and the results are shown in Figure 2. All the reduced catalysts exhibited the diffraction peaks of amorphous silica, suggesting that the SiO2 support did not change by the reduction conditions.In addition, the characteristic diffraction peaks of metallic Ni with different intensities were obtained at 2θ values of 44°, 52°, and 76° for Ni/SiO2, Ni-G-2/SiO2-C, and Ni-G-1/SiO2-P, indicating that the NiO reduced to Ni at the reduction conditions [39].However, no diffraction peaks of Ni species were observed for Ni-G-2/SiO2-P and Ni-G-3/SiO2-P, suggesting that particle sizes were too small to be detected by XRD.These results demonstrate that the Ni particle size was significantly influenced by the preparation method and the amount of glycine used.In order to verify the particle size and distribution of metallic Ni in the catalysts, the reduced Ni/SiO2, Ni-G-2/SiO2-C, and Ni-G-2/SiO2-P were subjected to TEM characterization.The obtained TEM images and particle size distribution of Ni are shown in Figure 3. Relatively larger Ni particle size, about 19.63 nm average particle size, and broad distri- Since metallic Ni is the active species for the CDR, the Ni particle size was also characterized by X-ray diffraction after reducing the catalysts at 700 • C for 2.5 h, and the results are shown in Figure 2. All the reduced catalysts exhibited the diffraction peaks of amorphous silica, suggesting that the SiO 2 support did not change by the reduction conditions.In addition, the characteristic diffraction peaks of metallic Ni with different intensities were obtained at 2θ values of 44 • , 52 • , and 76 • for Ni/SiO 2 , Ni-G-2/SiO 2 -C, and Ni-G-1/SiO 2 -P, indicating that the NiO reduced to Ni at the reduction conditions [39].However, no diffraction peaks of Ni species were observed for Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P, suggesting that particle sizes were too small to be detected by XRD.These results demonstrate that the Ni particle size was significantly influenced by the preparation method and the amount of glycine used.
In order to verify the particle size and distribution of metallic Ni in the catalysts, the reduced Ni/SiO 2 , Ni-G-2/SiO 2 -C, and Ni-G-2/SiO 2 -P were subjected to TEM characterization.The obtained TEM images and particle size distribution of Ni are shown in Figure 3. Relatively larger Ni particle size, about 19.63 nm average particle size, and broad distribution of Ni particles of about 0-70 nm were obtained for Ni/SiO 2 , while smaller Ni particle sizes and narrower distribution were observed for Ni-G-2/SiO 2 -C and Ni-G-2/SiO 2 -P.The Ni particle distribution and the average particle sizes were about 3-12 nm and 6.95 nm for Ni-G-2/SiO 2 -C, and 1-6 nm and 2.96 nm for Ni-G-2/SiO2-C, respectively, suggesting that the uniform particle size of Ni was obtained by adding glycine.These results are consistent with the XRD results shown in Figure 2.
bution of Ni particles of about 0-70 nm were obtained for Ni/SiO2, while smaller Ni particle sizes and narrower distribution were observed for Ni-G-2/SiO2-C and Ni-G-2/SiO2-P.The Ni particle distribution and the average particle sizes were about 3-12 nm and 6.95 nm for Ni-G-2/SiO2-C, and 1-6 nm and 2.96 nm for Ni-G-2/SiO2-C, respectively, suggesting that the uniform particle size of Ni was obtained by adding glycine.These results are consistent with the XRD results shown in Figure 2.

Reduction Ability of Ni-Based Catalysts
The reduction ability of Ni-based catalysts is characterized by H2-TPR, and the results are illustrated in

Reduction Ability of Ni-Based Catalysts
The reduction ability of Ni-based catalysts is characterized by H 2 -TPR, and the results are illustrated in Figure 4.Only Ni/SiO 2 shows a significant reduction peak at about 370 • C, while Ni-G-2/SiO 2 -C and all Ni-G/SiO 2 -P display two broad reduction peaks at about 370 • C and 530 • C. Generally, the reduction ability of the larger NiO particles should be lower than that of the smaller NiO particles.The reduction occurs below 400 • C for larger NiO particles and at higher temperatures for smaller NiO particle sizes and NiO with strong interaction and support [38,39].Thus, reduction peaks at lower temperatures are attributable to NiO species with bigger particle sizes, whereas the reduction peaks at higher temperatures could be attributed to the smaller NiO particle sizes or NiO species having strong interaction with supports.Furthermore, the ratio of higher temperature to lower temperature reduction peaks in Ni-G-2/SiO 2 -C is lower than all Ni-G/SiO 2 -P catalysts, suggesting larger NiO particle sizes for Ni-G-2/SiO 2 -C.These results are in agreement with the previous XRD and TEM results.
strong interaction and support [38,39].Thus, reduction peaks at lower temperatures are attributable to NiO species with bigger particle sizes, whereas the reduction peaks at higher temperatures could be attributed to the smaller NiO particle sizes or NiO species having strong interaction with supports.Furthermore, the ratio of higher temperature to lower temperature reduction peaks in Ni-G-2/SiO2-C is lower than all Ni-G/SiO2-P catalysts, suggesting larger NiO particle sizes for Ni-G-2/SiO2-C.These results are in agreement with the previous XRD and TEM results.

Catalytic Activity and Stability Tests
The CH4 and CO2 conversions, H2/CO, and H2 yield for all catalysts toward CDR at 750 o C using CH4 to CO2 molar ratio of 1 are presented in Figure 5.The results demonstrate that the CH4 and CO2 conversions significantly improved with the preparation method and the addition of glycine.The Ni/SiO2 catalyst exhibited initial CH4 and CO2 conversion of 62.7% and 75.1%, respectively.As expected, Ni-G/SiO2-C and Ni-G/SiO2-P catalysts showed higher conversions of CH4 and CO2, which were about 83% and 90%, respectively.The higher activity of these catalysts can be attributed to the synergistic effect of high surface area and smaller Ni particle size that substantially improved the accessibility of active sites.
Moreover, the CH4 and CO2 conversions over all the catalysts varied upon extending the reaction time, as shown in Figure 5A,B.The conversions of CH4 and CO2 over Ni/SiO2 catalyst linearly decreased after 20 h to 24% and 28% from the initial 62.5% and 75.1%, respectively.Interestingly, the stability of Ni/SiO2 prepared by the addition of glycine significantly improved.The stability of Ni-G-2/SiO2-C decreased during the 20 h CDR reaction, accompanied by the reduction in CH4 and CO2 conversions from 82.5% and 87.5% to 65.4% and 70.3%, respectively.However, the stability of Ni-G-1/SiO2-P decreased slightly, and the CH4 and CO2 conversions dropped marginally from 83.1% and 90.2% to 78.5% and 83.3%.With the increasing glycine content, no decrease in CH4 and CO2 conversions was observed for Ni-G-2/SiO2-P and Ni-G-3/SiO2-P catalyzed CDR, indicating their higher stability and no inactivation.This behavior might be due to the smaller Ni particle sizes and stronger interaction between Ni and SiO2 support, in compliance with the previously described results of XRD, TEM, and H2-TPR.
As shown in Figure 5C, H2/CO ratio was always less than 1 due to unavoidable RWGS reaction in CDR, i.e., H2 yield was lower than CO for all the catalysts.Interestingly, H2/CO ratio for Ni/SiO2 was far less than the Ni-G/SiO2-C and Ni-G/SiO2-P catalysts.Simultaneously, the H2 yield exhibited the same trend with CH4 conversion for all catalysts,

Catalytic Activity and Stability Tests
The CH 4 and CO 2 conversions, H 2 /CO, and H 2 yield for all catalysts toward CDR at 750 • C using CH 4 to CO 2 molar ratio of 1 are presented in Figure 5.The results demonstrate that the CH 4 and CO 2 conversions significantly improved with the preparation method and the addition of glycine.The Ni/SiO 2 catalyst exhibited initial CH 4 and CO 2 conversion of 62.7% and 75.1%, respectively.As expected, Ni-G/SiO 2 -C and Ni-G/SiO 2 -P catalysts showed higher conversions of CH 4 and CO 2 , which were about 83% and 90%, respectively.The higher activity of these catalysts can be attributed to the synergistic effect of high surface area and smaller Ni particle size that substantially improved the accessibility of active sites.
Catalysts 2021, 11, 1221 7 of 13 while it was slightly lower than the CH4 conversion shown in Figure 5D, which was also caused by the RWGS reaction in CDR.All these results signify the importance of Ni particle size play in RWGS in CDR reaction, which is regulated by the preparation method and glycine amount employed for the synthesis.

Characterization of Used Catalysts
After the stability tests at the reaction conditions of P = 1.0 atm, T = 750 °C, CH4/CO2 = 1.0, and gas hourly space velocity (GHSV) = 120,000 mL•g −1 •h −1 for 20 h, the used catalysts were subjected to XRD and TG characterizations to investigate the coking and sintering Moreover, the CH 4 and CO 2 conversions over all the catalysts varied upon extending the reaction time, as shown in Figure 5a,b.The conversions of CH 4 and CO 2 over Ni/SiO 2 catalyst linearly decreased after 20 h to 24% and 28% from the initial 62.5% and 75.1%, respectively.Interestingly, the stability of Ni/SiO 2 prepared by the addition of glycine significantly improved.The stability of Ni-G-2/SiO 2 -C decreased during the 20 h CDR reaction, accompanied by the reduction in CH 4 and CO 2 conversions from 82.5% and 87.5% to 65.4% and 70.3%, respectively.However, the stability of Ni-G-1/SiO 2 -P decreased slightly, and the CH 4 and CO 2 conversions dropped marginally from 83.1% and 90.2% to 78.5% and 83.3%.With the increasing glycine content, no decrease in CH 4 and CO 2 conversions was observed for Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P catalyzed CDR, indicating their higher stability and no inactivation.This behavior might be due to the smaller Ni particle sizes and stronger interaction between Ni and SiO 2 support, in compliance with the previously described results of XRD, TEM, and H 2 -TPR.
As shown in Figure 5c, H 2 /CO ratio was always less than 1 due to unavoidable RWGS reaction in CDR, i.e., H 2 yield was lower than CO for all the catalysts.Interestingly, H 2 /CO ratio for Ni/SiO 2 was far less than the Ni-G/SiO 2 -C and Ni-G/SiO 2 -P catalysts.Simultaneously, the H 2 yield exhibited the same trend with CH 4 conversion for all catalysts, while it was slightly lower than the CH 4 conversion shown in Figure 5d, which was also caused by the RWGS reaction in CDR.All these results signify the importance of Ni particle size play in RWGS in CDR reaction, which is regulated by the preparation method and glycine amount employed for the synthesis.

Characterization of Used Catalysts
After the stability tests at the reaction conditions of P = 1.0 atm, T = 750 • C, CH 4 /CO 2 = 1.0, and gas hourly space velocity (GHSV) = 120,000 mL•g −1 •h −1 for 20 h, the used catalysts were subjected to XRD and TG characterizations to investigate the coking and sintering of Ni.
XRD patterns of the used catalysts are displayed in Figure 6.The characteristic diffraction peaks of metallic Ni and amorphous silica were obtained for all the used Nibased catalysts, similar to the XRD patterns of the reduced catalysts shown in Figure 2, implying the same crystal structure of Ni-based catalysts after CDR.Moreover, Ni/SiO 2 displayed a prominent shoulder peak at 2θ of approximately 26 • , a typical diffraction peak of graphitic carbon.Although a small graphitic carbon diffraction peak was obtained for used Ni-G-2/SiO 2 -C and Ni-G-1/SiO 2 -P catalysts, no such peak was found in the XRD patterns of used Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P.Thus, the XRD data infer that relatively more coke was deposited on Ni/SiO 2 , causing the linear inactivation for CDR, which was attributed to larger Ni particle size [10,15].Large Ni particle sizes are prone to the formation of carbon deposits on the surface of the catalyst, which can cover the active components, thereby deactivating the catalyst [14].The sintering of Ni after CDR is the primary reason for the deactivation of Ni-based catalysts.Therefore, Ni particle sizes of the used and reduced catalysts were carefully compared.The peak intensities of the used Ni/SiO 2 , Ni-G-2/SiO 2 -C, and Ni-G-1/SiO 2 -P were more than those of the corresponding reduced catalysts, indicating Ni sintering after CDR [39], while Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P did not exhibit any diffraction peak of the metallic nickel, suggesting a smaller Ni particle size and superior anti-sintering property.Combined with the front results of TPR (Figure 4), Ni-G/SiO 2 -C and Ni-G/SiO 2 -P exhibited better interaction between the Ni particles and the support, as compared with Ni/SiO 2 .Likewise, the strongest interaction between the Ni particles and SiO 2 existed in Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P catalysts, resulting in good CDR performance because of superior anti-sintering characteristics [10,18].
Coke deposition of the used catalysts after the CDR reaction was studied by TGA, and the results are shown in Figure 7.All the catalysts showed a distinct weight gain around 200-400 • C, during the oxidation of Ni to NiO in the air atmosphere [37,45].Further, a weight loss from 300 • C was observed for all the catalysts utilized in 20 h CDR due to the gasification of the coke deposited.The amount of the deposited coke on used catalysts decreased in the order of Ni/SiO 2 << Ni-G-2/SiO 2 -C < Ni-G-1/SiO 2 -P ≈ Ni-G-2/SiO 2 -P ≈ Ni-G-3/SiO 2 -P, i.e., 3.2% for Ni/SiO 2 , 2.2% for Ni-G-2/SiO 2 -C, and about 1.7-1.5% for Ni-G/SiO 2 -P.The amount of deposited coke is affected by the smaller Ni particle size of Ni-G/SiO 2 -P (Figures 2 and 3), leading to its stable performance after 20 h CDR testing (Figure 4).The highest weight loss of about 3.2% was obtained for used Ni/SiO 2 , consistent with the literature that the large Ni particle sizes of the catalyst are susceptible to the coke deposition during the CDR, which can deteriorate the activity of the catalyst [14,39].Thus, adding glycine and changing preparation methods significantly decrease coke deposition due to smaller Ni particle formation, leading to the high stability of the catalysts for CDR.

Preparation of Catalysts
Ni(NO3)2•6H2O (98.5%), glycine (99%), and sodium hydroxide (NaOH, 96%) were purchased (Shanghai, Sinopharm Chemical Reagent Co. Ltd., China) and used without further purification.Commercial SiO2 (Q-15 was purchased (Fuji, Fujisilysia Chemical Ltd., Japan) was used as support for the catalysts.Before use, SiO2 was calcined at 750 o C for 4 h to remove moisture and impurities.The SiO2-supported Ni catalysts were prepared by the precipitation-impregnation method using Ni complex as the precursor, prepared by Ni(OH)2 and glycine.The Ni loading for all of the catalysts was 10 wt.%.The detailed preparation of catalysts is as follows: The Ni(NO3)2 solution was added to the NaOH solution, where the molar ratio of NaOH to Ni(NO3)2 of 2.0 was used to form Ni(OH)2 precipitate.The Ni(OH)2 precipitate was filtered and washed with deionized water, and the light green solid Ni(OH)2 was dried at 80 o C for 12 h.In order to prepare three different Ni(II)-glycine complexes, the Ni(OH)2 powder was slowly added to three 0.2 mol/L aqueous glycine solutions at 80 o C while stirring for 2 h with final glycine to Ni(OH)2 ratio of

Preparation of Catalysts
Ni(NO 3 ) 2 •6H 2 O (98.5%), glycine (99%), and sodium hydroxide (NaOH, 96%) were purchased (Sinopharm Chemical Reagent Co. Ltd., Shanghai, China) and used without further purification.Commercial SiO 2 (Q-15 was purchased (Fuji, Fujisilysia Chemical Ltd., Kyoto, Japan) was used as support for the catalysts.Before use, SiO 2 was calcined at 750 • C for 4 h to remove moisture and impurities.The SiO 2 -supported Ni catalysts were prepared by the precipitation-impregnation method using Ni complex as the precursor, prepared by Ni(OH) 2 and glycine.The Ni loading for all of the catalysts was 10 wt.%.The detailed preparation of catalysts is as follows: The Ni(NO 3 ) 2 solution was added to the NaOH solution, where the molar ratio of NaOH to Ni(NO 3 ) 2 of 2.0 was used to form Ni(OH) 2 precipitate.The Ni(OH) 2 precipitate was filtered and washed with deionized water, and the light green solid Ni(OH) 2 was dried at 80 • C for 12 h.In order to prepare three different Ni(II)-glycine complexes, the Ni(OH) 2 powder was slowly added to three 0.2 mol/L aqueous glycine solutions at 80 • C while stirring for 2 h with final glycine to Ni(OH) 2 ratio of 1, 2, and 3.Then, the SiO 2 powder was impregnated with the obtained solutions.Finally, all the samples were dried at 100 • C for 12 h and calcined at 500 • C for 4 h.The obtained catalysts were denoted as Ni-G-x/SiO 2 -P, where x = molar ratio of glycine to Ni(OH) 2 .For comparative studies, a SiO 2 -supported Ni catalyst was also prepared by the complexation-impregnation method using a mixed solution of Ni(NO 3 ) 2 and glycine with a glycine-to-Ni molar ratio of 2, listed as Ni-G-2/SiO 2 -C.Moreover, the Ni/Q-50 catalyst was prepared following the same procedure with the same preparation parameters (conventional incipient impregnation method) but without the addition of citric acid.

Characterization Techniques
Specific BET surface areas of the prepared catalysts were measured by a surface area analyzer (Bel, BelSorp-Max, Osaka, Japan) at liquid nitrogen temperature (−196 • C).Before measurements, samples were heated to 300 • C for 12 h under vacuum (10 −2 kPa) to remove contaminants and water.The BET surface areas were acquired by the BET method, while pore volumes were obtained from the nitrogen adsorption data at the relative pressure (P/P0) = 0.99.
X-ray powder diffraction (XRD) patterns of the catalysts were recorded by X-ray diffractor (D8 advanced, Bruker, Germany) using CuKα (λ = 1.5406Å) radiation.The measurement was conditioned at 40 kV and 40 mA.The crystalline size of NiO was calculated according to the (200) diffraction peak according to the formula of Scherrer (D = Kλ/Bcosθ).
Transmission electron microscopy (TEM) images of the reduced catalysts were taken by a transmission electron microscope (JEM-2100, JEOL, Tokey, Japan).Specimens were prepared by dispersing samples in ethanol by ultrasonication for 2 min, and droplets of the suspension were deposited on a copper grid, which was dried in air.
H 2 temperature-programmed reduction (H 2 -TPR) was performed on a chemisorption apparatus (Micromeritics Autochem 2920, GA, USA) with a thermal conductivity detector to test the reduction behaviors of the catalysts.Prior to the experiments, approximately 0.0500 g sample was pretreated at 300 • C for 2 h under flowing Ar (30 mL/min) to remove impurities.After cooling to 100 • C, a 10% H 2 /Ar (30 mL/min) mixture was introduced to obtain a stable baseline.Then, the H 2 -TPR profile was recorded from 100 • C to 1000 • C at the heating rate of 10 • C min −1 .
Thermogravimetric (TG) analysis was carried out on a thermoanalyzer system (TA Instruments, Q1000DSC + LNCS + FACS Q600SDT, DE, USA) to analyze the coke deposited on used catalysts, which was performed from 100 to 900 • C at a heating rate of 10 • C min −1 under air atmosphere.

Activity Evaluation of the Catalysts
Catalytic activity measurements of the prepared Ni-based catalysts for CDR were carried out in a fixed-bed reactor with a quartz reaction tube (8 mm internal diameter).About 0.1 g catalyst (40-60 mesh) mixed with quartz sand (5 times of catalysts weight) was loaded in the reactor and in situ reduced by 10% H 2 /N 2 (50 mL/min) at 700 • C for 150 min.The reactor was then heated to 750 • C by purging N 2 (40 mL/min) with a heating rate of 5 • C min −1 .After reaching 750 • C temperature, an equimolar CH 4 /CO 2 mixture was passed through the catalyst (100 mL/min).The product gases in the outlet were separated by molecular sieve 5A and Porapak Q capillary column and then analyzed by an online gas chromatography (GC9720 II, Zhejiang Fuli chromatographic analysis Co., Ltd., Zhejiang, China) equipped with a TCD.Conversions of CH 4 and CO 2 , H 2 yield, and the molar ratio of H 2 /CO were determined from corresponding flow rates in the inlet and outlet.
The conversion of CH 4 and CO 2 (abbreviated as X i ), the molar ratio of H 2 to CO (H 2 /CO), and H 2 yield (Y) were calculated as follows:

Conclusions
The Ni-G/SiO 2 -C and Ni-G/SiO 2 -P catalysts were successfully prepared by complexationimpregnation and precipitation-impregnation methods using different amounts of glycine as the complexing agent.Compared with Ni/SiO 2 , a smaller Ni particle size and a narrower Ni distribution were obtained, suggesting that sintering of Ni was significantly inhibited through the complexation of metallic Ni by glycine in the preparation process.All the catalysts were tested for CDR at P = 0.1 MPa, T = 750 • C, CH 4 /CO 2 = 1, and GHSV = 120,000 mL•g −1 •h −1 for 20 h time on stream.The results demonstrated that Nisupported catalysts prepared with glycine as complexing agents possess enhanced activity (conversions of CH 4 and CO 2 above 82% and 87%, respectively) and improved stability.Moreover, the preparation methods and the amount of glycine used to prepare the Ni-based catalysts also play an important role in the CDR performance because of the Ni particle size and Ni dispersion.The Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P catalysts showed smaller Ni particle sizes and higher dispersion of Ni, resulting in high CDR activity and stability.The analyses of used catalysts showed that Ni-G-2/SiO 2 -P and Ni-G-3/SiO 2 -P exhibit carbon deposition and sintering of Ni, which is the reason for the deactivation of the catalysts.Thus, the results conclude that the CDR catalysts with small particle sizes and narrower Ni distribution can be easily prepared with the precipitation-impregnation method, a promising strategy for CDR.

Figure 4 .
Only Ni/SiO2 shows a significant reduction peak at about 370 o C, while Ni-G-2/SiO2-C and all Ni-G/SiO2-P display two broad reduction peaks at about 370 o C and 530 o C. Generally, the reduction ability of the larger NiO particles should be lower than that of the smaller NiO particles.The reduction occurs below 400 o C for larger NiO particles and at higher temperatures for smaller NiO particle sizes and NiO with

Figure 6 .
Figure 6.X-ray diffraction patterns of the used Ni/SiO2 (a), Ni-G-2/SiO2-C (b), Ni-G-1/SiO2-P (c), Ni-G-2/SiO2-P (d), and Ni-G-3/SiO2-P (e) catalysts.The sintering of Ni after CDR is the primary reason for the deactivation of Ni-based catalysts.Therefore, Ni particle sizes of the used and reduced catalysts were carefully compared.The peak intensities of the used Ni/SiO2, Ni-G-2/SiO2-C, and Ni-G-1/SiO2-P were more than those of the corresponding reduced catalysts, indicating Ni sintering after

Table 1 .
Textural properties of support and catalysts.