CeZrOx Promoted Water-Gas Shift Reaction under Steam–Methane Reforming Conditions on Ni-HTASO5

: Ni-based catalysts (Ni- γ -Al 2 O 3 , Ni-HTASO5 and Ni-CeZrO x ) were prepared by impregnation method and characterized by BET, AAS, XRD, H 2 -TPR, CO-TPD, NH 3 -TPD, XPS, TG-DSC-MS and Raman spectroscopies. Using CeZrO x -modified Al 2 O 3 (HTASO5) as support, the catalyst exhibited good catalytic performance (TOF CH4 = 8.0 × 10 − 2 s − 1 , TOF H2 = 10.5 × 10 − 2 s − 1 ) and carbon resistance for steam-methane reforming (SMR) reaction. Moreover, CeZrO x was able to enhance water-gas shift (WGS) reaction for more hydrogen production. It was found that the addition of CeZrO x could increase the content of active nickel precursor on the surface of the catalyst, which was beneficial to the decomposition of water and methane on Ni-HTASO5. Furthermore, Ni-HTASO5 could decrease the strong acid sites of the catalyst, which would not only contribute to the formation of low graphited carbon, but also decrease the amount of carbon deposition.


Introduction
Hydrogen is considered an important part of future energy systems. With the development of the hydrogen fuel cell, the application of H 2 in vehicles and energy fields has aroused the interest of many researchers. Using Ni-based catalysts, methane can react with H 2 O, O 2 or CO 2 to produce hydrogen and carbon monoxide [1][2][3]. Because of its high H 2 /CO ratio, steam-methane reforming reaction (Equation (1)) is the main approach of hydrogen production in industry. When water-gas shift (WGS) reaction (Equation (2)) occurs simultaneously, it will increase the yield of hydrogen. Removing the products (H 2 , CO 2 ) or enhancing adsorption of CO and H 2 O on the catalyst are both beneficial to WGS reaction for hydrogen production.
CO + H 2 O H 2 + CO 2 (2) Although compared with precious metals, nickel is not the most active catalyst, it is the most attractive because of its low cost and promising catalytic performance [4]. It has been reported that coke formation and metal sintering are the main reasons that lead to the deactivation of Ni-based catalysts in SMR reaction [5]. With the stoichiometric steam-to-methane ratio (H 2 O/CH 4 = 1), graphite carbon is formed on the nickel-based catalyst, leading to reactor blockage and further deactivation of the catalyst [6]. High water-methane ratio can decrease the formation of carbon. In actual industrial

The Catalytic Activity of the Catalysts
The turnover frequency (TOF) of CH 4 and H 2 at 600 • C for 9 h are depicted in Figure 1. The initial TOF of CH 4 conversion (for one hour) on Ni-γ-Al 2 O 3 , Ni-HTASO5 and Ni-CeZrO x are 5.1 × 10 −2 s −1 , 8.0 × 10 −2 s −1 and 5.5 × 10 −2 s −1 , respectively. The corresponding initial TOF of H 2 (for one hour) are 7.0 × 10 −2 s −1 , 10.5 × 10 −2 s −1 and 6.8 × 10 −2 s −1 , respectively. The higher TOF of H 2 implies that WGS Catalysts 2020, 10, 1110 3 of 13 reaction may occur simultaneously. In the activity test, the Ni-HTASO5 exhibited the best catalytic performance while that of the other two catalysts were close. The repeatability of the catalyst was good, and the relative error between the two repeated experiments was 3.22% (Supplementary Materials Figure S2, Table S2). In terms of stability, the TOF of H 2 on Ni-γ-Al 2 O 3 and Ni-HTASO5 were stable at around 6.5 × 10 −2 s −1 and 8.7 × 10 −2 s −1 , respectively after reaction for five hours, while that of Ni-CeZrO x gradually decreased. Meanwhile, the ratio of the content of CO 2 to CO on Ni-CeZrO x was higher than those on the other two catalysts and that on Ni-HTASO5 was higher than on Ni-γ-Al 2 O 3 . It could be inferred that CeZrO x as additive or support can promote WGS reaction, which is beneficial to H 2 production. Compared to other results from the literature, the catalytic effect of CeZrO x on WGS reaction was generally observed, and the modified Ni-HTASO5 catalyst was more conducive to the simultaneous production of hydrogen in the two-step reaction [14,26,28].
Catalysts 2020, 10, x FOR PEER REVIEW 3 of 13 performance while that of the other two catalysts were close. The repeatability of the catalyst was good, and the relative error between the two repeated experiments was 3.22% ( Figure S2, Table S2). In terms of stability, the TOF of H2 on Ni-γ-Al2O3 and Ni-HTASO5 were stable at around 6.5 × 10 −2 s −1 and 8.7 × 10 −2 s −1 , respectively after reaction for five hours, while that of Ni-CeZrOx gradually decreased. Meanwhile, the ratio of the content of CO2 to CO on Ni-CeZrOx was higher than those on the other two catalysts and that on Ni-HTASO5 was higher than on Ni-γ-Al2O3. It could be inferred that CeZrOx as additive or support can promote WGS reaction, which is beneficial to H2 production. Compared to other results from the literature, the catalytic effect of CeZrOx on WGS reaction was generally observed, and the modified Ni-HTASO5 catalyst was more conducive to the simultaneous production of hydrogen in the two-step reaction [14,26,28].

BET Results of the Catalysts
The specific surface area (SBET) and dispersion of pore volume (Vp) and pore size (Dp) of the catalysts are shown in Figure 2 and Table 1. The isotherm curves of the three catalysts are all type Ⅳ and the ratio of P to P0 for the hysteresis loop is over the range of 0.45 to 1.0, which indicates that they are all mesoporous materials [33]. The BET result (Table 1) showed that HTASO5 retained a relatively larger surface area of the catalyst than CeZrOx. The specific surface area of Ni-γ-Al2O3, Ni-HTASO5 and Ni-CeZrOx are 141.9 m 2 g −1 , 67.6 m 2 g −1 and 49.0 m 2 g −1 , respectively. The addition of Ce, Zr in Al2O3 support decreased the surface area of the catalyst. The pore volume also decreased significantly compared with Ni-γ-Al2O3. However, the Ni-HTASO5 catalyst remained relatively larger surface area of than Ni-CeZrOx catalyst, which could promote the dispersion of nickel. The pore size distribution of Ni-HTASO5 was concentrated around 10-20 nm, while the other two catalysts were more widely distributed from 10 to 80 nm. As seen in Figure 2A, after being reduced in H2 atmosphere at 600 °C, the mesopores in Ni-γ-Al2O3 and Ni-CeZrOx could be classified into H3-type, which indicated an irregular pore structure and it is similar with that observed in nanorod and/or nanofiber [34]. However, the shape of hysteresis loop for Ni-HTASO5 was grouped within H1-type, which is typically found in spheroidal particles of uniform size and array [35,36].

BET Results of the Catalysts
The specific surface area (S BET ) and dispersion of pore volume (Vp) and pore size (Dp) of the catalysts are shown in Figure 2 and Table 1. The isotherm curves of the three catalysts are all type IV and the ratio of P to P 0 for the hysteresis loop is over the range of 0.45 to 1.0, which indicates that they are all mesoporous materials [33]. The BET result (Table 1) showed that HTASO5 retained a relatively larger surface area of the catalyst than CeZrO x . The specific surface area of Ni-γ-Al 2 O 3 , Ni-HTASO5 and Ni-CeZrO x are 141.9 m 2 g −1 , 67.6 m 2 g −1 and 49.0 m 2 g −1 , respectively. The addition of Ce, Zr in Al 2 O 3 support decreased the surface area of the catalyst. The pore volume also decreased significantly compared with Ni-γ-Al 2 O 3 . However, the Ni-HTASO5 catalyst remained relatively larger surface area of than Ni-CeZrO x catalyst, which could promote the dispersion of nickel. The pore size distribution of Ni-HTASO5 was concentrated around 10-20 nm, while the other two catalysts were more widely distributed from 10 to 80 nm. As seen in Figure 2A, after being reduced in H 2 atmosphere at 600 • C, the mesopores in Ni-γ-Al 2 O 3 and Ni-CeZrO x could be classified into H3-type, which indicated an irregular pore structure and it is similar with that observed in nanorod and/or nanofiber [34]. However, the shape of hysteresis loop for Ni-HTASO5 was grouped within H1-type, which is typically found in spheroidal particles of uniform size and array [35,36].  The actual loading of nickel was characterized by atom adsorption spectrum, and the results are shown in Table 1. The content of Ni on Ni-γ-Al2O3, Ni-HTASO5 and Ni-CeZrOx are 9.8%, 8.9% and 7.0%, respectively. Combining this result with that of surface area, the lower surface area was responsible for lower Ni-loading, which would influence the conversion of CH4.

The Reducibility and Distribution of Ni Species on the Catalysts
The reducibility of the Ni-based catalysts with different supports is shown in Figure 3A. Two peaks around 286 °C and 401 °C are observed on Ni-CeZrOx catalyst. The reducibility curve of Ni-HTASO5 is similar to Ni-γ-Al2O3. The peaks are around 450 °C (A peak) and 750 °C (B peak). Compared to the other two catalysts, the Ni species on Ni-CeZrOx could be reduced at lower temperature, which is attributed to weak interaction between metal Ni and CeZrOx support. The two reduction peaks of Ni-HTASO5 both shifted to lower temperatures when compared with Ni-γ-Al2O3. Furthermore, the ratio of A peak to B peak on Ni-γ-Al2O3 was lower than that of Ni-HTASO5. They both indicated that the addition of CeZrOx to Al2O3 can weaken the interaction between Ni species and the support.  Table 1. Physical properties of different Ni-based catalysts and different supports, the surface area was determined by BET method, the pore size (Dp) and the pore volume (Vp) were both determined by BJH method. The loading content of Ni was detected by atomic absorption spectrum (AAS). The actual loading of nickel was characterized by atom adsorption spectrum, and the results are shown in Table 1. The content of Ni on Ni-γ-Al 2 O 3 , Ni-HTASO5 and Ni-CeZrO x are 9.8%, 8.9% and 7.0%, respectively. Combining this result with that of surface area, the lower surface area was responsible for lower Ni-loading, which would influence the conversion of CH 4 .

The Reducibility and Distribution of Ni Species on the Catalysts
The reducibility of the Ni-based catalysts with different supports is shown in Figure 3A. Two peaks around 286 • C and 401 • C are observed on Ni-CeZrO x catalyst. The reducibility curve of Ni-HTASO5 is similar to Ni-γ-Al 2 O 3 . The peaks are around 450 • C (A peak) and 750 • C (B peak). Compared to the other two catalysts, the Ni species on Ni-CeZrO x could be reduced at lower temperature, which is attributed to weak interaction between metal Ni and CeZrO x support. The two reduction peaks of Ni-HTASO5 both shifted to lower temperatures when compared with Ni-γ-Al 2 O 3 . Furthermore, the ratio of A peak to B peak on Ni-γ-Al 2 O 3 was lower than that of Ni-HTASO5. They both indicated that the addition of CeZrO x to Al 2 O 3 can weaken the interaction between Ni species and the support. The XRD result of Ni-based catalysts with different supports is shown in Figure 3B. It can be seen that the diffraction peak of 2θ = 44.5°, 51.9°, 76.4°(PDF# 04-0850)of metal Ni 0 appears on all the catalysts after reduction and reaction. The Ce0.6Zr0.4O2 crystalline phase appeared on Ni-CeZrOx catalyst, while Ce0.16Zr0.84O2 crystalline phase was shown on Ni-HTASO5. Therefore, different CeZrOx were formed on Ni-CeZrOx and Ni-HTASO5. There was no significant change of the crystal size of Ni 0 on the catalysts before and after the reaction (see in Table 2). The crystal size of Ni 0 on Ni-γ-Al2O3 was smaller than the other two catalysts, which is owed to the high dispersion of Ni on the support. Furthermore, the carbon diffraction peak (PDF# 41-1487) appeared on Ni-γ-Al2O3 after reaction while it cannot be seen on the other catalysts, which means graphite carbon did form on Ni-γ-Al2O3. The grain size of Ni did not change significantly before and after the reaction of Ni-HTASO5 and Ni-CeZrOx and the grain size of them was similar. Table 2. Crystal size of nickel on Ni-γ-Al2O3, Ni-HTASO5 and Ni-CeZrOx after reduction and reaction as determined by XRD. The intensity ratio of G bond to D bond of deposited carbon after reaction at 600 °C for 9 h, tested by Raman spectroscopy.

Catalyst
Ni 0 (nm) IG/ID Reduction Reaction Ni-γ-Al2O3 12 12 0.58 Ni-HTASO5 18 19 0.47 Ni-CeZrOx 19 18 0.20 The information of different surface species on the catalyst was determined by X-ray photoelectron spectrum (XPS). As seen in Figure 4, for Ni 2p, all the catalysts have three peaks. The binding energy at 854.5 eV, 855.8 eV were regarded as NiOx species and Ni 2+ species, respectively [14,25,37]. Furthermore, NiOx species was regarded as the precursor of active nickel species [27,38]. From quantitative analysis results in Table 3, among the three catalysts, Ni-HTASO5 catalyst had the highest content of precursor of the active species nickel (NiOx), which play very important roles in the catalytic activity and stability [27,38]. Comparing the binding energy of Zr, Ce and O on Ni-HTASO5 and Ni-CeZrOx, it was found that the peak location of Ce had little change while that of Zr on Ni-HTASO5 shifted to higher binding energy. This means that Zr on Ni-HTASO5 had a stronger ability to donate electrons, which may contribute to the reduction of Ni species. The result of O 1sbinding energy showed that the ratio of lattice oxygen (O 2− , 529.6 eV) to surface oxygen (OH − , 531.6 eV) on Ni-CeZrOx was much higher than that on Ni-HTASO5 [39,40]. Therefore, CeZrOx had the property of storing oxygen while the HTASO5 support contained Al2O3, which made the surface of Ni-HTASO5 contain more OH − . It has been reported that the lattice oxygen could promote activation The XRD result of Ni-based catalysts with different supports is shown in Figure 3B. It can be seen that the diffraction peak of 2θ = 44.5 • , 51.9 • , 76.4 • (PDF# 04-0850) of metal Ni 0 appears on all the catalysts after reduction and reaction. The Ce 0.6 Zr 0.4 O 2 crystalline phase appeared on Ni-CeZrO x catalyst, while Ce 0.16 Zr 0.84 O 2 crystalline phase was shown on Ni-HTASO5. Therefore, different CeZrO x were formed on Ni-CeZrO x and Ni-HTASO5. There was no significant change of the crystal size of Ni 0 on the catalysts before and after the reaction (see in Table 2). The crystal size of Ni 0 on Ni-γ-Al 2 O 3 was smaller than the other two catalysts, which is owed to the high dispersion of Ni on the support. Furthermore, the carbon diffraction peak (PDF# 41-1487) appeared on Ni-γ-Al 2 O 3 after reaction while it cannot be seen on the other catalysts, which means graphite carbon did form on Ni-γ-Al 2 O 3 . The grain size of Ni did not change significantly before and after the reaction of Ni-HTASO5 and Ni-CeZrO x and the grain size of them was similar. Table 2. Crystal size of nickel on Ni-γ-Al 2 O 3 , Ni-HTASO5 and Ni-CeZrO x after reduction and reaction as determined by XRD. The intensity ratio of G bond to D bond of deposited carbon after reaction at 600 • C for 9 h, tested by Raman spectroscopy.

Catalyst
Ni 0 (nm) I G /I D Reduction Reaction The information of different surface species on the catalyst was determined by X-ray photoelectron spectrum (XPS). As seen in Figure 4, for Ni 2p, all the catalysts have three peaks. The binding energy at 854.5 eV, 855.8 eV were regarded as NiO x species and Ni 2+ species, respectively [14,25,37]. Furthermore, NiO x species was regarded as the precursor of active nickel species [27,38]. From quantitative analysis results in Table 3, among the three catalysts, Ni-HTASO5 catalyst had the highest content of precursor of the active species nickel (NiO x ), which play very important roles in the catalytic activity and stability [27,38]. Comparing the binding energy of Zr, Ce and O on Ni-HTASO5 and Ni-CeZrO x , it was found that the peak location of Ce had little change while that of Zr on Ni-HTASO5 shifted to higher binding energy. This means that Zr on Ni-HTASO5 had a stronger ability to donate electrons, which may contribute to the reduction of Ni species. The result of O 1s-binding energy showed that the ratio of lattice oxygen (O 2− , 529.6 eV) to surface oxygen (OH − , 531.6 eV) on Ni-CeZrO x was much higher than that on Ni-HTASO5 [39,40]. Therefore, CeZrO x had the property of storing oxygen while the HTASO5 support contained Al 2 O 3 , which made the surface of Ni-HTASO5 contain more OH − . It has been reported that the lattice oxygen could promote activation of water [21,23], which may promote WGS reaction. The high surface content of active nickel could promote methane decomposition and hydrogen production, which made high catalytic activity of SMR reaction. Although there was more active lattice oxygen to promote the WGS reaction on Ni-CeZrO x , the low content of active nickel limited the SMR reaction for H 2 production. Therefore, the TOF of hydrogen on Ni-HTASO5 was higher than that on Ni-CeZrO x .
Catalysts 2020, 10, x FOR PEER REVIEW 6 of 13 of water [21,23], which may promote WGS reaction. The high surface content of active nickel could promote methane decomposition and hydrogen production, which made high catalytic activity of SMR reaction. Although there was more active lattice oxygen to promote the WGS reaction on Ni-CeZrOx, the low content of active nickel limited the SMR reaction for H2 production. Therefore, the TOF of hydrogen on Ni-HTASO5 was higher than that on Ni-CeZrOx.

The Acidity and Carbon Deposition of the Catalysts
The acidity of Ni-γ-Al2O3, Ni-HTASO5 and Ni-CeZrOx catalysts was characterized by temperature programmed NH3 desorption experiment. Three NH3 desorption peaks, which are ascribed to weak, medium and strong acid sites, over the range of 50-500 °C on all the catalysts are shown in Figure 5. The desorption peak at the range of 50-150 °C (peak 1) was attributed to weak Lewis acid site, and the peak with maximum at 150-250 °C (peak 2) was regarded as NH4 + bound to medium acid site [41]. The peak at above 250 °C (peak 3) was attributed to NH3 bound to strong Lewis acid site and NH4 + bound to strong Brønsted acid site [41][42][43], which may be due to the presence of surface OH − on the support. Moreover, the amount of desorbed ammonia was calculated from NH3 desorption peak area and the result is shown in Table 4. The tendency of acid strength was as follows:

The Acidity and Carbon Deposition of the Catalysts
The acidity of Ni-γ-Al 2 O 3 , Ni-HTASO5 and Ni-CeZrO x catalysts was characterized by temperature programmed NH 3 desorption experiment. Three NH 3 desorption peaks, which are ascribed to weak, medium and strong acid sites, over the range of 50-500 • C on all the catalysts are shown in Figure 5. The desorption peak at the range of 50-150 • C (peak 1) was attributed to weak Lewis acid site, and the peak with maximum at 150-250 • C (peak 2) was regarded as NH 4 + bound to medium acid site [41].
The peak at above 250 • C (peak 3) was attributed to NH 3 bound to strong Lewis acid site and NH 4 + Catalysts 2020, 10, 1110 7 of 13 bound to strong Brønsted acid site [41][42][43], which may be due to the presence of surface OH − on the support. Moreover, the amount of desorbed ammonia was calculated from NH 3 desorption peak area and the result is shown in Table 4. The tendency of acid strength was as follows: Ni-γ-Al 2 O 3 > Ni-HTASO5 > Ni-CeZrO x . The medium and strong acid site (peak 2 and peak 3) gradually increased with the increase of Al 2 O 3 content. That was due to the acidity of Al 2 O 3 support. The weak acid site (peak 1) increased by the addition of CeZrO x . The acidity of Ni-HTASO5 was lower than Ni-γ-Al 2 O 3 , which was due to the addition of CeZrOx to the support.
Catalysts 2020, 10, x FOR PEER REVIEW 7 of 13 Ni-γ-Al2O3 > Ni-HTASO5 > Ni-CeZrOx. The medium and strong acid site (peak 2 and peak 3) gradually increased with the increase of Al2O3 content. That was due to the acidity of Al2O3 support. The weak acid site (peak 1) increased by the addition of CeZrOx. The acidity of Ni-HTASO5 was lower than Ni-γ-Al2O3, which was due to the addition of CeZrOx to the support.  The amount and type of carbon deposition were investigated by thermogravimetric analysis combined with differential scanning calorimeter and mass spectrometer (TG-DSC-MS). The result is shown in Figure 6. In terms of the amount of deposited carbon, there was the highest amount of coke on Ni-γ-Al2O3 (24.01%) catalyst and the least on Ni-CeZrOx (9.43%) catalyst. It had previously been indicated that carbon was more likely to deposit on Ni-γ-Al2O3. This may be attributed to the acidic site of Al2O3, which could catalyze polymerization and cracking, resulting in more carbon deposition on the catalyst surface [44]. Meanwhile, the addition of Zr or Ce could increase the ability of carbon resistance [44][45][46]. Comparing with Ni-γ-Al2O3 catalyst, Ni-HTASO5 catalyst had less carbon deposition. Meanwhile, the CO2 MS signal on Ni-HTASO5 shifted to lower temperature. It means that the deposited carbon on Ni-HTASO5 was more easily removed. There was the least amount of coke on Ni-CeZrOx, furthermore the coke was of easy removal character. Combined with the result of NH3-TPD, it was inferred that the increase of weak acid sites would decrease the amount of carbon deposition and inhibit the growth of coke, which could then easily to be removed. That is to say, the  The amount and type of carbon deposition were investigated by thermogravimetric analysis combined with differential scanning calorimeter and mass spectrometer (TG-DSC-MS). The result is shown in Figure 6. In terms of the amount of deposited carbon, there was the highest amount of coke on Ni-γ-Al 2 O 3 (24.01%) catalyst and the least on Ni-CeZrO x (9.43%) catalyst. It had previously been indicated that carbon was more likely to deposit on Ni-γ-Al 2 O 3 . This may be attributed to the acidic site of Al 2 O 3 , which could catalyze polymerization and cracking, resulting in more carbon deposition on the catalyst surface [44]. Meanwhile, the addition of Zr or Ce could increase the ability of carbon resistance [44][45][46]. Comparing with Ni-γ-Al 2 O 3 catalyst, Ni-HTASO5 catalyst had less carbon deposition. Meanwhile, the CO 2 MS signal on Ni-HTASO5 shifted to lower temperature. It means that the deposited carbon on Ni-HTASO5 was more easily removed. There was the least amount of coke on Ni-CeZrO x , furthermore the coke was of easy removal character. Combined with the result of NH 3 -TPD, it was inferred that the increase of weak acid sites would decrease the amount of carbon Catalysts 2020, 10, 1110 8 of 13 deposition and inhibit the growth of coke, which could then easily to be removed. That is to say, the Ni-HTASO5 modified by CeZrO x could not only form easily removable carbon, but also reduce the amount of carbon formation.
Ni-HTASO5 modified by CeZrOx could not only form easily removable carbon, but also reduce the amount of carbon formation.
In order to identify the graphitization degree of coke, the Raman spectroscopy analysis was performed ( Figure 6D). It could be seen that there were four peaks on all the catalysts. The peak at 1343 cm −1 (D bond) was considered to C-C stretch vibration of disordered carbon while the peak at 1579 cm −1 (G bond) was regarded as C-C stretch vibration of well-ordered carbon. The ratio of IG/ID of Ni-γ-Al2O3, Ni-HTASO5 and Ni-CeZrOx was 0.58, 0.47 and 0.20 (see in Table 2), respectively. As is shown, the graphitization degree increased with the increase content of Al2O3 in the support. Adding CeO2 and ZrO2 to the composite support could change the property of carbon deposition, making it decrease the graphitization, which was beneficial to carbon removal. Combined with the characterization of the catalysts and the results of activity, it could be seen that the highest turnover frequency of hydrogen on Ni-HTASO5 is owed to its high surface content of active nickel and promising carbon resistance.
The CeZrOx support was prepared by co-precipitation method. Cerium nitrate (99.5%, Kelong, Chengdu, China) and zirconium nitrate (99.5%, Kelong, Chengdu, China) with the mole-ratio of Ce:Zr = 3:2 were dissolved in deionized water. The resulting solution was transferred to a flask and the pH adjusted to 9 using 2.5 M NH3•H2O with constantly stirring. The mixture was aged for 2 h at 70 °C, then cooled to room temperature and stored for 12 h. The resulting gel was rinsed thoroughly In order to identify the graphitization degree of coke, the Raman spectroscopy analysis was performed ( Figure 6D). It could be seen that there were four peaks on all the catalysts. The peak at 1343 cm −1 (D bond) was considered to C-C stretch vibration of disordered carbon while the peak at 1579 cm −1 (G bond) was regarded as C-C stretch vibration of well-ordered carbon. The ratio of I G /I D of Ni-γ-Al 2 O 3 , Ni-HTASO5 and Ni-CeZrO x was 0.58, 0.47 and 0.20 (see in Table 2), respectively. As is shown, the graphitization degree increased with the increase content of Al 2 O 3 in the support. Adding CeO 2 and ZrO 2 to the composite support could change the property of carbon deposition, making it decrease the graphitization, which was beneficial to carbon removal.
Combined with the characterization of the catalysts and the results of activity, it could be seen that the highest turnover frequency of hydrogen on Ni-HTASO5 is owed to its high surface content of active nickel and promising carbon resistance.
The CeZrO x support was prepared by co-precipitation method. Cerium nitrate (99.5%, Kelong, Chengdu, China) and zirconium nitrate (99.5%, Kelong, Chengdu, China) with the mole-ratio of Ce:Zr = 3:2 were dissolved in deionized water. The resulting solution was transferred to a flask and the pH adjusted to 9 using 2.5 M NH 3 ·H 2 O with constantly stirring. The mixture was aged for 2 h at Catalysts 2020, 10, 1110 9 of 13 70 • C, then cooled to room temperature and stored for 12 h. The resulting gel was rinsed thoroughly with deionized water and then it was dried in air overnight. The support was finally calcined at 450 • C for 4 h.
Ni-γ-Al 2 O 3 , Ni-HTASO5 and Ni-CeZrO x catalysts were prepared by impregnation method. First, Ni(NO 3 ) 4 ·6H 2 O (2.1846 g) was added to 20 mL deionized water. The above support (4 g) was added after 20 min ultrasound treatment of the solution. The mixture was impregnated by stirring constantly at room temperature for 24 h, then dried at 80 • C using oil bath and at 110 • C in oven for 4 h. The three catalysts precursors were obtained after being calcined at 600 • C for 4 h.

Catalytic Activity Test
The catalyst activity was tested in a stainless steel fixed bed reactor at atmospheric pressure. Ni-γ-Al 2 O 3 and Ni-HTASO5 were reduced at 800 • C for 1 h under the atmosphere of H 2 and Ar mixture (H 2 /Ar = 1) and passivated in the mixture of oxygen and argon (5% O 2 in Ar) at room temperature, while the reduction temperature of Ni-CeZrO x was 600 • C and passivated under the same conditions. Before the activity test, all the catalysts were reduced again at 600 • C for 1 h. The reaction gases methane (F = 30 mL/min) and H 2 O (nH 2 O/nCH 4 = 3) were injected after ten min of argon purge. Meanwhile, the gas products were analyzed by gas chromatography (TDX-01 packed column).
The CH 4 (or H 2 ) turnover frequencies were calculated by the molar number of CH 4 (or H 2 ) converted or produced per second per mole exposed Ni atom. The number of exposed Ni atoms was determined by CO temperature programmed experiments (seen in Table S1). The methane conversion rate (X CH 4 ), H 2 production rate (Y H 2 ) and the CO 2 , CO (C CO 2 , C CO ) content were calculated as follows: where F referred to the flow of gas, mL/min and m(CH 4 ) (or m(H 2 )) was the molar number of CH 4 (or H 2 ) converted per second and n(Ni) was the molar number of exposed Ni atoms per gram of catalyst.

Characterization Methods
The BET specific area and the distribution of pore volume and pore size of samples were tested in a Micromeritics Tristar II 3020 instrument (Micromeritics, Norcross, GA, USA) by adsorption-desorption of N 2 at −196 • C. Before the test, the sample was activated at 120 • C and 300 • C to eliminate any adsorbed substance.
The actual loading was tested by atom adsorption spectrum in a SpectrAA 220FS instrument (Varian, Palo Alto, CA, USA). The sample was dissolved in aqua regia (HCl:HNO 3 = 3:1) and a small amount of HF. Deionized water was added while heating to nearly complete dissolution. Then water was added repeatedly to remove the acid completely. The grating was holographic diffraction grating with 1200 lines/mm (240 nm).
The temperature-programmed reduction (TPR) was performed to attain the reduction property of the catalysts by the Micromeritics Autochem II 2920 instrument (Micromeritics, Norcross, GA, USA). The test was from 50 • C to 900 • C with a heating rate of 5 • C/min under an atmosphere of 10% H 2 /Ar.
The temperature programmed NH 3 desorption (NH 3 -TPD) was used to study the acid sites of the sample. First, NH 3 was adsorbed on the reduced catalysts at 50 • C for 1 h under a mixture of 10% NH 3 in He. Then, He was used to clean the excessive unadsorbed NH 3 for 1 h. After this, with a heating rate of 10 • C/min under He flow, the sample was heated to 600 • C. The NH 3 desorbed curve at different temperature was presented.
The CO temperature programmed desorption (CO-TPD) was carried out to study the amount of active centers on the sample. First, the catalyst was reduced at 600 • C under an atmosphere of 10% H 2 /Ar. After reduction, the samples were blown with He for 2 h. Then, CO was adsorbed on the reduced catalysts at 50 • C for 1 h under a mixture of 3% CO in He. He was used to clean the excessive unadsorbed CO for 1 h. After this, with a heating rate of 10 • C/min under a He flow, the sample was heated to 800 • C. The CO desorbed curve was presented.
X-ray diffraction (XRD) was conducted to probe the type and the size of formed crystal by an XRD-6100 (SHIMADZU, Japan) instrument. Cu Kα radiation of 40 kV and 25 mA was used. The diffraction angle ranged from 5 • to 80 • .
Thermogravimetric, differential scanning calorimeter and mass combination (TG-DSC-MS) analysis were performed to characterize carbon deposition on used catalysts by TG209F1 (NETZSCH, Selb, Germany) instrument. With the heating rate of 10 • C/min, the test started from 30 • C to 800 • C under an atmosphere of air/N 2 (20/60 mL/min).
X-ray photoelectron spectroscopy (XPS) was carried out to study the surface state of the element on AXIS Ultra DLD (KRATOS, Manchester, UK) instrument equipped with a neutralizer. With monochromatic Al Kα as the light source, the acceleration power of 25 W, the binding energies were calibrated using C1s 284.6 eV, and the peak separation was performed with a Lorenz-Gaussian ratio (L/G) of 20%.
Raman spectroscopy analysis was conducted to characterize the deposited carbon on used catalysts. He-Ne laser source (532 nm) was used on LabRAM HR instrument (HORIBA, Kyoto, Japan) for the test. The filter was D1 and the aperture was 200 mm.

Conclusions
Nickel-based catalysts were prepared using γ-Al 2 O 3 , HTASO5 and CeZrO x as supports and used for steam-methane reforming reaction. Ni-HTASO5 showed good catalytic performance at 600 • C for 9 h while being more responsive to WGS reaction and having a promoting effect on hydrogen production. The high catalytic activity of Ni-HTASO5 was due to the presence of a high amount of active Ni precursor species on its surface as compared with Ni-γ-Al 2 O 3 and Ni-CeZrO x . This could contribute to the decomposition of methane and water. The presence of CeZrO x promoted the WGS reaction under steam-methane reforming conditions. Furthermore, it was seen that the weak acid sites would decrease on the Ni-based catalyst-doping with CeZrO x in Al 2 O 3 as support, which was beneficial to decrease the amount of carbon deposition and make it easier to be removed with a low graphited structure.