Encapsulated Ni@La2O3/SiO2 Catalyst with a One-Pot Method for the Dry Reforming of Methane

Ni nanoparticles encapsulated within La2O3 porous system (Ni@La2O3), the latter supported on SiO2 (Ni@La2O3)/SiO2), effectively inhibit carbon deposition for the dry reforming of methane. In this study, Ni@La2O3/SiO2 catalyst was prepared using a one-pot colloidal solution combustion method. Catalyst characterization demonstrates that the amorphous La2O3 layer was coated on SiO2, and small Ni nanoparticles were encapsulated within the layer of amorphous La2O3. During 50 h of dry reforming of methane at 700 °C and using a weight hourly space velocity (WHSV) of 120,000 mL gcat−1 h−1, the CH4 conversion obtained was maintained at 80%, which is near the equilibrium value, while that of impregnated Ni–La2O3/SiO2 catalyst decreased from 63% to 49%. The Ni@La2O3/SiO2 catalyst exhibited very good resistance to carbon deposition, and only 1.6 wt% carbon was formed on the Ni@La2O3/SiO2 catalyst after 50 h of reaction, far lower than that of 11.5 wt% deposited on the Ni–La2O3/SiO2 catalyst. This was mainly attributed to the encapsulated Ni nanoparticles in the amorphous La2O3 layer. In addition, after reaction at 700 °C for 80 h with a high WHSV of 600,000 mL gcat−1 h−1, the Ni@La2O3/SiO2 catalyst exhibited high CH4 conversion rate, ca. 10.10 mmol gNi−1 s−1. These findings outline a simple synthesis method to prepare supported encapsulated Ni within a metal oxide porous structure catalyst for the dry reforming of methane reaction.


Introduction
Dry reforming of methane (DRM) is a promising process, as it can simultaneously convert CO 2 and CH 4 present in CO 2 -rich natural gas reservoirs to produce syngas. The latter serves as the raw material to produce liquid fuels through gas-to-liquid technology (via Fischer-Tropsch synthesis) [1]. Due to the strong endothermic nature of the DRM reaction, most of previous studies were conducted at temperatures higher than 600 • C for high conversions of CO 2 and CH 4 . Therefore, developing a robust catalyst that possesses good stability and excellent resistance to coke plays a crucial role in the DRM reaction [2][3][4]. CO 2 + CH 4 ↔ 2CO + 2H 2 , ∆H 298 K = +247 kJ mol −1 (1)

Characterization of Fresh and Reduced Catalysts
The N 2 adsorption results of the Ni@La 2 O 3 /SiO 2 and Ni-La 2 O 3 /SiO 2 catalysts are listed in Table 1. The specific surface area (S BET ) of the Ni@La 2 O 3 /SiO 2 is 19.0 m 2 g −1 , which is smaller than that of the Ni-La 2 O 3 /SiO 2 catalyst. The pore volume and average pore size of Ni@La 2 O 3 /SiO 2 catalyst are 0.21 cm 3 g −1 and 43.9 nm, respectively, which are larger than those of Ni-La 2 O 3 /SiO 2 catalyst.
It is noted that the Ni@La 2 O 3 /SiO 2 catalyst has a significantly lower specific surface area compared with the recently reported mesoporous Ni-La 2 O 3 (172 m 2 .g −1 ) [42] and Ni-La 2 O 3 /SiO 2 (190 m 2 g −1 ) [38] catalysts for DRM. In our previous report [43], mesoporous Ni-La 2 O 3 (70.4 m 2 g −1 ) had been synthesized by the same colloidal solution combustion method with colloidal SiO 2 as a template, and the silica was then removed by NaOH etching to form mesopores. Compared with our previously reported mesoporous Ni-La 2 O 3 catalysts, the Ni@La 2 O 3 /SiO 2 catalyst may be more suitable for high-temperature reactions due to the use of silica as support of La 2 O 3 and Ni. a NiO size in the fresh catalyst. b Ni size in the reduced and used catalysts. Data in brackets correspond to the Ni size of used catalyst. c Crystallite size determined by XRD using the Scherrer equation. d Mean particle size determined by TEM images analysis. e The particle in TEM image was too small to be observed, or the peak in XRD pattern was too weak to be used for calculations. Figure 1 shows the powder XRD patterns of the fresh and reduced catalysts. For the fresh Ni-La 2 O 3 /SiO 2 , the peaks at 2θ = 37.2 • , 43.3 • , and 62.9 • are attributed to NiO [43,44]. The reduced Ni-La 2 O 3 /SiO 2 showed weak Ni peaks at 44.5 • and 51.7 • [43,44]. As shown in Table 1, the crystallite size of Ni in reduced Ni-La 2 O 3 /SiO 2 is 16.4 nm, which is about twice that of NiO in fresh Ni-La 2 O 3 /SiO 2 . This indicates that NiO in the fresh Ni-La 2 O 3 /SiO 2 catalyst is unstable during the reduction process and sintering. In contrast, there are no obviously NiO peaks found in fresh Ni@La 2 O 3 /SiO 2 catalyst, and a broad Ni peak at 44.5 • is found in reduced Ni@La 2 O 3 /SiO 2 catalyst. This indicates that Ni particle size in the reduced Ni@La 2 O 3 /SiO 2 catalyst is smaller than that in the reduced Ni-La 2 O 3 /SiO 2 catalyst. It should be noted here that no La 2 O 3 peak was found in any catalyst. This might be due to the very small La 2 O 3 crystals formed not able to detect by XRD, or that the La 2 O 3 was in the amorphous phase [37]. The results indicate that La 2 O 3 was highly dispersed or amorphous in these catalytic systems. Similar results were reported in the literature [38]. The morphology of La 2 O 3 (small particle or amorphous phase) needs to be further confirmed by TEM analysis.
Catalysts 2019, 9, x FOR PEER REVIEW 3 of 16 a NiO size in the fresh catalyst. b Ni size in the reduced and used catalysts. Data in brackets correspond to the Ni size of used catalyst. c Crystallite size determined by XRD using the Scherrer equation. d Mean particle size determined by TEM images analysis. e The particle in TEM image was too small to be observed, or the peak in XRD pattern was too weak to be used for calculations. Figure 1 shows the powder XRD patterns of the fresh and reduced catalysts. For the fresh Ni-La2O3/SiO2, the peaks at 2θ = 37.2°, 43.3°, and 62.9° are attributed to NiO [43,44]. The reduced Ni-La2O3/SiO2 showed weak Ni peaks at 44.5° and 51.7° [43,44]. As shown in Table 1, the crystallite size of Ni in reduced Ni-La2O3/SiO2 is 16.4 nm, which is about twice that of NiO in fresh Ni-La2O3/SiO2. This indicates that NiO in the fresh Ni-La2O3/SiO2 catalyst is unstable during the reduction process and sintering. In contrast, there are no obviously NiO peaks found in fresh Ni@La2O3/SiO2 catalyst, and a broad Ni peak at 44.5° is found in reduced Ni@La2O3/SiO2 catalyst. This indicates that Ni particle size in the reduced Ni@La2O3/SiO2 catalyst is smaller than that in the reduced Ni-La2O3/SiO2 catalyst. It should be noted here that no La2O3 peak was found in any catalyst. This might be due to the very small La2O3 crystals formed not able to detect by XRD, or that the La2O3 was in the amorphous phase [37]. The results indicate that La2O3 was highly dispersed or amorphous in these catalytic systems. Similar results were reported in the literature [38]. The morphology of La2O3 (small particle or amorphous phase) needs to be further confirmed by TEM analysis. The TEM images of the fresh and reduced catalysts are shown in Figure 2. For the fresh Ni-La2O3/SiO2 catalyst, dark aggregated NiO particles and SiO2 particles ~ 20 nm with a smooth surface are shown in Figure 2a, indicating that most of NiO was not loaded onto SiO2 supported but aggregated instead. The particle size distribution of NiO is displayed in the inset of Figure 2a. The representative high-resolution TEM images in Figure 2b show lattice fringes corresponding to La2O3 and NiO, thus illustrating the formation of NiO and La2O3 in the fresh Ni-La2O3/SiO2 catalyst. The TEM images of the fresh and reduced catalysts are shown in Figure 2. For the fresh Ni-La 2 O 3 /SiO 2 catalyst, dark aggregated NiO particles and SiO 2 particles~20 nm with a smooth surface are shown in Figure 2a, indicating that most of NiO was not loaded onto SiO 2 supported but aggregated instead. The particle size distribution of NiO is displayed in the inset of Figure 2a. The representative high-resolution TEM images in Figure 2b Figure 2e. As observed, SiO2 is coated with an amorphous La2O3 layer, on which metallic Ni particles are encapsulated within the amorphous La2O3 layer. An average Ni particle size about 3.5 nm was obtained by counting more than 100 Ni particles as shown in Figure 2f. This result   Figure 2e. As observed, SiO 2 is coated with an amorphous La 2 O 3 layer, on which metallic Ni particles are encapsulated within the amorphous La 2 O 3 layer. An average Ni particle size about 3.5 nm was obtained by counting more than 100 Ni particles as shown in Figure 2f. This result is consistent with our previous report of mesoporous Ni-La 2 O 3 prepared via colloidal solution combustion method [43].
Combined with TEM and S BET results, it can be concluded that the lower S BET of Ni@La 2 O 3 /SiO 2 may be due to the fact that the surface of SiO 2 nanoparticles was covered with La 2 O 3 and NiO. On the contrary, the SiO 2 nanoparticles in the Ni@La 2 O 3 /SiO 2 catalyst structure may not be completely covered by La 2 O 3 and NiO, and the exposed SiO 2 surface resulted in a slightly larger S BET .
The BF-STEM images and the element distribution profiles of Si, La, and Ni are shown in Figure 3.  is consistent with our previous report of mesoporous Ni-La2O3 prepared via colloidal solution combustion method [43]. Combined with TEM and SBET results, it can be concluded that the lower SBET of Ni@La2O3/SiO2 may be due to the fact that the surface of SiO2 nanoparticles was covered with La2O3 and NiO. On the contrary, the SiO2 nanoparticles in the Ni@La2O3/SiO2 catalyst structure may not be completely covered by La2O3 and NiO, and the exposed SiO2 surface resulted in a slightly larger SBET.
The BF-STEM images and the element distribution profiles of Si, La, and Ni are shown in Figure  3. SiO2 is surrounded by La2O3 and Ni nanoparticles. The signal of Ni is accompanied by the existence of La, but the signal of La is not necessarily accompanied by Ni, indicating that the nickel is encapsulated by La2O3. The signal of La2O3 is distributed around the signal of silica, indicating that SiO2 is encapsulated by La2O3. The Ni particles are encapsulated by amorphous La2O3 and wrapped on silica. The H2-TPR profiles of the catalysts are shown in Figure 4. The Ni-La2O3/SiO2 catalyst exhibits two reduction peaks. The first peak at 340 °C corresponds to the reduction of free NiO [21,39,45]. The second peak at 385 °C corresponds to the reduction of NiO with weak interaction with La2O3 or SiO2 [40,42,46,47]. These results indicate that NiO is weakly interacting or not interacting at all with La2O3 or SiO2 in the Ni-La2O3/SiO2 catalyst.
The Ni@La2O3/SiO2 catalyst displays a broad reduction peak at 615 °C, suggesting that Ni-based species have a strong interaction with the support [42,48]. Li et al. [42] found that the small NiO particle confined into mesoporous La2O3 strongly interacts with La2O3 support, resulting in a high reduction temperature for NiO. Also, the Ni@La2O3/SiO2 catalyst, which possesses the encapsulated The H 2 -TPR profiles of the catalysts are shown in Figure 4. The Ni-La 2 O 3 /SiO 2 catalyst exhibits two reduction peaks. The first peak at 340 • C corresponds to the reduction of free NiO [21,39,45]. The second peak at 385 • C corresponds to the reduction of NiO with weak interaction with La 2 O 3 or SiO 2 [40,42,46,47]. These results indicate that NiO is weakly interacting or not interacting at all with La 2 O 3 or SiO 2 in the Ni-La 2 O 3 /SiO 2 catalyst.
The Ni@La 2 O 3 /SiO 2 catalyst displays a broad reduction peak at 615 • C, suggesting that Ni-based species have a strong interaction with the support [42,48]. Li et al. [42] found that the small NiO particle confined into mesoporous La 2 O 3 strongly interacts with La 2 O 3 support, resulting in a high reduction temperature for NiO. Also, the Ni@La 2 O 3 /SiO 2 catalyst, which possesses the encapsulated structure of metal Ni by La 2 O 3 layer on SiO 2 (in Figure 3a), exhibits high reduction temperatures. Thus, the high reduction temperature at 615 • C is due to the reduction of NiO, which presents strong interactions with the encapsulated La 2 O 3 layer.
Catalysts 2019, 9, x FOR PEER REVIEW 6 of 16 structure of metal Ni by La2O3 layer on SiO2 (in Figure 3a), exhibits high reduction temperatures. Thus, the high reduction temperature at 615 °C is due to the reduction of NiO, which presents strong interactions with the encapsulated La2O3 layer.  Figure 5 shows results of the catalytic performance tests conducted over the Ni@La2O3/SiO2 and Ni-La2O3/SiO2 catalysts in the DRM at 700 °C with a weight hourly space velocity (WHSV) of 120,000 mL g −1 h −1 and after 50 h of reaction. At 700 °C, the thermodynamic equilibrium conversion of CO2 and CH4 were 90.2% and 84.4%, respectively. It should be noted that this thermodynamic equilibrium consists of the DRM reaction and reverse water-gas shift reaction (RWGS: CO2 + H2  CO + H2O). As shown in Figure 5a, the CO2 conversion on the Ni@La2O3/SiO2 catalyst is 90% and reached the equilibrium conversion. The CO2 conversion in the two catalysts are higher than the CH4 conversion, a result which is mainly due to the RWGS reaction [31]. The Ni@La2O3/SiO2 catalyst exhibits stable CO2 and CH4 conversions during the DRM reaction period of 50 h. In contrast, the CO2 and CH4 conversions of the Ni-La2O3/SiO2 catalyst decrease from 75% to 62% and from 63% to 49%, within 50 h, respectively. Therefore, the Ni@La2O3/SiO2 catalyst has better activity and stability behavior than the Ni-La2O3/SiO2 catalyst during DRM.

Catalytic Performance Studies
As shown in Figure 5b, the H2/CO ratio is lower than one for both catalysts. This result is mainly due to the RWGS reaction and to a less degree to other side reactions [31], such as the reverse Boudouard reaction (C + CO2  2 CO) [49]. The CO and H2 yields are shown in Figure 5c,d, respectively. For each catalyst, the CO yield is higher than the H2 yield. Comparing the two catalysts, the Ni@La2O3/SiO2 catalyst exhibits higher H2 and CO yields than the Ni-La2O3/SiO2 catalyst.  Figure 5 shows results of the catalytic performance tests conducted over the Ni@La 2 O 3 /SiO 2 and Ni-La 2 O 3 /SiO 2 catalysts in the DRM at 700 • C with a weight hourly space velocity (WHSV) of 120,000 mL g −1 h −1 and after 50 h of reaction. At 700 • C, the thermodynamic equilibrium conversion of CO 2 and CH 4 were 90.2% and 84.4%, respectively. It should be noted that this thermodynamic equilibrium consists of the DRM reaction and reverse water-gas shift reaction (RWGS: CO 2 + H 2 ↔ CO + H 2 O). As shown in Figure 5a, the CO 2 conversion on the Ni@La 2 O 3 /SiO 2 catalyst is 90% and reached the equilibrium conversion. The CO 2 conversion in the two catalysts are higher than the CH 4 conversion, a result which is mainly due to the RWGS reaction [31]. The Ni@La 2 O 3 /SiO 2 catalyst exhibits stable CO 2 and CH 4 conversions during the DRM reaction period of 50 h. In contrast, the CO 2 and CH 4 conversions of the Ni-La 2 O 3 /SiO 2 catalyst decrease from 75% to 62% and from 63% to 49%, within 50 h, respectively. Therefore, the Ni@La 2 O 3 /SiO 2 catalyst has better activity and stability behavior than the Ni-La 2 O 3 /SiO 2 catalyst during DRM.

Catalytic Performance Studies
As shown in Figure 5b, the H 2 /CO ratio is lower than one for both catalysts. This result is mainly due to the RWGS reaction and to a less degree to other side reactions [31], such as the reverse Boudouard reaction (C + CO 2 ↔ 2 CO) [49]. The CO and H 2 yields are shown in Figure 5c,d, respectively. For each catalyst, the CO yield is higher than the H 2 yield. Comparing the two catalysts, the Ni@La 2 O 3 /SiO 2 catalyst exhibits higher H 2 and CO yields than the Ni-La 2 O 3 /SiO 2 catalyst. Catalysts 2019, 9, x FOR PEER REVIEW 7 of 16 No sign of deactivation was observed for the Ni@La2O3/SiO2 catalyst with a WHSV of 120,000 mL g −1 h −1 , as shown in Figure 5, and Figure 6 illustrates the effect of WHSV on the catalytic performance of the Ni@La2O3/SiO2 catalyst in the range of 120,000 to 1,200,000 mL g −1 h −1 . As shown in Figure 6a, in the 120,000-300,000 mL g −1 h −1 range, the CO2 and CH4 conversions obviously remain constant. When the WHSV increased to 600,000 mL g −1 h −1 , the conversions start to decrease slightly, but the CH4 conversion is still higher than 70%, suggesting the very good activity of the catalyst. In particular, when the WHSV is increased to 1,200,000 mL g −1 h −1 , the conversions of CH4 and CO2 are decreased to 51.7% and 65.7%. Accordingly, the CO and H2 yields show the same No sign of deactivation was observed for the Ni@La 2 O 3 /SiO 2 catalyst with a WHSV of 120,000 mL g −1 h −1 , as shown in Figure 5, and Figure 6 illustrates the effect of WHSV on the catalytic performance of the Ni@La 2 O 3 /SiO 2 catalyst in the range of 120,000 to 1,200,000 mL g −1 h −1 . No sign of deactivation was observed for the Ni@La2O3/SiO2 catalyst with a WHSV of 120,000 mL g −1 h −1 , as shown in Figure 5, and Figure 6 illustrates the effect of WHSV on the catalytic performance of the Ni@La2O3/SiO2 catalyst in the range of 120,000 to 1,200,000 mL g −1 h −1 .   As shown in Figure 6a, in the 120,000-300,000 mL g −1 h −1 range, the CO2 and CH4 conversions obviously remain constant. When the WHSV increased to 600,000 mL g −1 h −1 , the conversions start to decrease slightly, but the CH4 conversion is still higher than 70%, suggesting the very good activity of the catalyst. In particular, when the WHSV is increased to 1,200,000 mL g −1 h −1 , the conversions of CH4 and CO2 are decreased to 51.7% and 65.7%. Accordingly, the CO and H2 yields show the same trend, which are decreased significantly when the WHSV becomes larger than 600,000 mL g −1 h −1 . This behavior with WHSV is largely related to external mass transport effects established within the catalytic bed. Figure 7 shows results of the stability test of Ni@La2O3/SiO2 catalyst conducted at a high WHSV of 600,000 mL g −1 h −1 . Although the CH4 and CO2 conversions are slightly decreased over 80 h of As shown in Figure 6a, in the 120,000-300,000 mL g −1 h −1 range, the CO 2 and CH 4 conversions obviously remain constant. When the WHSV increased to 600,000 mL g −1 h −1 , the conversions start to decrease slightly, but the CH 4 conversion is still higher than 70%, suggesting the very good activity of the catalyst. In particular, when the WHSV is increased to 1,200,000 mL g −1 h −1 , the conversions of CH 4 and CO 2 are decreased to 51.7% and 65.7%. Accordingly, the CO and H 2 yields show the same trend, which are decreased significantly when the WHSV becomes larger than 600,000 mL g −1 h −1 .
This behavior with WHSV is largely related to external mass transport effects established within the catalytic bed. Figure 7 shows results of the stability test of Ni@La 2 O 3 /SiO 2 catalyst conducted at a high WHSV of 600,000 mL g −1 h −1 . Although the CH 4 and CO 2 conversions are slightly decreased over 80 h of reaction, the CH 4 conversion is still as high as 65% after reaction, indicating the very good activity and stability of the Ni@La 2 O 3 /SiO 2 catalyst.  Table 2 lists the methane conversion rates obtained over Ni@La2O3/SiO2 and some representative Ni-based catalysts reported in the literature. As listed in Table 2, the CH4 conversion rate of Ni@La2O3/SiO2 catalyst is five times higher than that of Ni-La2O3/SiO2 catalyst. Although the literature in Table 2 is limited, it appears that the Ni@La2O3/SiO2 catalyst has a significantly better methane conversion rate and coke resistance in the DRM reaction at the conditions applied.

Characterization of Used Catalysts
To measure the amount of deposited carbon on the used catalysts, TG and DTA tests were conducted, and the obtained results are shown in Figure 8a,b. For the used Ni-La2O3/SiO2 catalyst, the weight loss is 11.5 wt% in the range 500-700 °C, and the DTA exhibits an obvious exothermic peak due to the oxidation of deposited carbon. The weight loss of the used Ni@La2O3/SiO2 catalyst was only 1.6 wt% over 50 h in DRM, which is significantly lower than that obtained in the used Ni-La2O3/SiO2 catalyst (11.5 wt% carbon deposition). These results indicate that Ni@La2O3/SiO2 catalyst  Table 2 lists the methane conversion rates obtained over Ni@La 2 O 3 /SiO 2 and some representative Ni-based catalysts reported in the literature. As listed in Table 2, the CH 4 conversion rate of Ni@La 2 O 3 /SiO 2 catalyst is five times higher than that of Ni-La 2 O 3 /SiO 2 catalyst. Although the literature in Table 2 is limited, it appears that the Ni@La 2 O 3 /SiO 2 catalyst has a significantly better methane conversion rate and coke resistance in the DRM reaction at the conditions applied.

Characterization of Used Catalysts
To measure the amount of deposited carbon on the used catalysts, TG and DTA tests were conducted, and the obtained results are shown in Figure 8a,b. For the used Ni-La 2 O 3 /SiO 2 catalyst, the weight loss is 11.5 wt% in the range 500-700 • C, and the DTA exhibits an obvious exothermic peak due to the oxidation of deposited carbon. The weight loss of the used Ni@La 2 O 3 /SiO 2 catalyst was only 1.6 wt% over 50 h in DRM, which is significantly lower than that obtained in the used Ni-La 2 O 3 /SiO 2 catalyst (11.5 wt% carbon deposition). These results indicate that Ni@La 2 O 3 /SiO 2 catalyst has a very good resistance to carbon deposition. The reason for the very good resistance to carbon deposition for the aforementioned catalyst might be partly due to the formed La2O3 layer, within which the Ni metallic active sites are well dispersed and less prone to carbon accumulation. As recently reported [51], the Ce0.8Pr0.2O2-δsupported Ni catalyst prepared by the citrate sol-gel method, due to the presence of mobile active oxygen species in the Ce0.8Pr0.2O2-δ support, largely participates in the carbon removal via gasification to CO(g). Moreover, Ni particles smaller in size can reduce carbon accumulation [52]. Therefore, the good carbon resistance exhibited by Ni@La2O3/SiO2 catalyst seems to be largely related to the smaller nickel particle size and the presence of La2O3 coating layer on the Ni particles.
Based on the weight loss range of temperatures ca. 480-730 °C along the exothermic peak, the carbon deposited is composed of whisker carbon and encapsulated graphitic carbon. It should be noted that the encapsulated graphitic carbon is usually responsible for catalyst deactivation [53]. However, the whisker carbon, which possesses hollow structure, has little effect on the active sites of metallic Ni, and therefore is not the main reason for catalyst deactivation. Figure 9 shows the XRD patterns of the used catalysts. After a 50 h DRM reaction, the used Ni-La2O3/SiO2 catalyst shows an obvious graphitic peak (2θ = 26.5°) [31] and Ni peaks (2θ = 44.5 and 51.8°) as well. As shown in Table 1, the crystallite size of Ni in the used catalyst increased to 44.6 nm, indicating that Ni particles in the Ni-La2O3/SiO2 catalyst are not stable during the DRM reaction, and large Ni particles and carbon deposition are formed during reaction. The weak Ni diffraction peaks found in the used Ni@La2O3/SiO2 catalyst indicate that Ni particles become smaller after DRM for 50 h. These results indicate that Ni particles in the Ni@La2O3/SiO2 catalyst are more stable than in Ni@La2O3/SiO2 catalyst. In addition, no obvious diffraction peak of graphitic carbon is observed in the used Ni@La2O3/SiO2 catalyst. As proved by the TG-DTA analysis, the amount of carbon deposition (1.6 wt%) is rather small to be detected by powder XRD. The reason for the very good resistance to carbon deposition for the aforementioned catalyst might be partly due to the formed La 2 O 3 layer, within which the Ni metallic active sites are well dispersed and less prone to carbon accumulation. As recently reported [51], the Ce 0.8 Pr 0.2 O 2-δ -supported Ni catalyst prepared by the citrate sol-gel method, due to the presence of mobile active oxygen species in the Ce 0.8 Pr 0.2 O 2-δ support, largely participates in the carbon removal via gasification to CO(g). Moreover, Ni particles smaller in size can reduce carbon accumulation [52]. Therefore, the good carbon resistance exhibited by Ni@La 2 O 3 /SiO 2 catalyst seems to be largely related to the smaller nickel particle size and the presence of La 2 O 3 coating layer on the Ni particles.
Based on the weight loss range of temperatures ca. 480-730 • C along the exothermic peak, the carbon deposited is composed of whisker carbon and encapsulated graphitic carbon. It should be noted that the encapsulated graphitic carbon is usually responsible for catalyst deactivation [53]. However, the whisker carbon, which possesses hollow structure, has little effect on the active sites of metallic Ni, and therefore is not the main reason for catalyst deactivation. Figure 9 shows the XRD patterns of the used catalysts. After a 50 h DRM reaction, the used Ni-La 2 O 3 /SiO 2 catalyst shows an obvious graphitic peak (2θ = 26.5 • ) [31] and Ni peaks (2θ = 44.5 • and 51.8 • ) as well. As shown in Table 1, the crystallite size of Ni in the used catalyst increased to 44.6 nm, indicating that Ni particles in the Ni-La 2 O 3 /SiO 2 catalyst are not stable during the DRM reaction, and large Ni particles and carbon deposition are formed during reaction. The weak Ni diffraction peaks found in the used Ni@La 2 O 3 /SiO 2 catalyst indicate that Ni particles become smaller after DRM for 50 h. These results indicate that Ni particles in the Ni@La 2 O 3 /SiO 2 catalyst are more stable than in Ni@La 2 O 3 /SiO 2 catalyst. In addition, no obvious diffraction peak of graphitic carbon is observed in the used Ni@La 2 O 3 /SiO 2 catalyst. As proved by the TG-DTA analysis, the amount of carbon deposition (1.6 wt%) is rather small to be detected by powder XRD. TEM images of the used catalysts are shown in Figure 10. No whisker carbon was found in the used Ni@La2O3/SiO2 (Figure 10a, b). The Ni mean particle size is about 5 nm, which is slightly larger than that in the reduced Ni@La2O3/SiO2, indicating that the small Ni particles encapsulated into the amorphous La2O3 layer are thermally stable and largely contribute to the inhibition of carbon deposition.
As shown in Figure 10, accessible nickel particles (without encapsulation) favor the formation of carbon. Carbon nanofibers and encapsulated graphitic carbon were formed over the used Ni-La2O3/SiO2 catalyst. Methane decomposes on the nickel surface forming atomic hydrogen and carbon, the latter diffusing to free surface sites on the nickel particle to form graphitic carbon (the graphitic carbon peak is precisely seen in Figure 9) [54,55]. As shown in Figure 10c, Ni particles in the used Ni-La2O3/SiO2 catalyst are in the 10-50 nm range, which is much wider than that found in the used Ni@La2O3/SiO2 catalyst. These results are consistent with the TG and XRD results of the used catalysts. It is mentioned here that deposited carbon can plug the reactor and reduce the lifetime of the catalyst as well. However, in general, the formation of carbon nanofibers does not decrease the exposed Ni surface area of the catalyst, thereby maintaining stable catalytic activity. Ni sintering and the formation of encapsulated carbon can reduce the exposed Ni surface area and thus result in catalyst deactivation.
Based on the Ni@La2O3/SiO2 catalyst structure features, it is reasonable to propose that its excellent activity, stability, and high resistance to carbon deposition are much related to the small Ni particles encapsulated within the amorphous La2O3 layer deposited on SiO2. Chen et al. [38] has prepared Ni-La2O3/SiO2 via one-pot sol-gel method with large specific surface area (190 m 2 g −1 ), which exhibited high activity and excellent stability for DRM at 700 ℃ . However, the rate of deposited carbon on Ni-La2O3/SiO2 was 5.9 mg C gcat −1 h −1 , which is much higher than that found in the present Ni@La2O3/SiO2 catalyst (0.32 mg C gcat −1 h −1 ). Although the BET surface area of the Ni@La2O3/SiO2 catalyst is relatively low, the amorphous La2O3 layer can encapsulate and stabilize the small nickel particles formed, thus resulting in an active and stable Ni@La2O3/SiO2 catalyst. Based on the XRD, TEM, and TPR results, it can be concluded that the formed Ni@La2O3/SiO2 catalyst structure cannot only stabilize small nickel particles and reduce carbon accumulation, but also provides more interface between Ni and La2O3. The latter can promote CO2 activation on oxygen vacant sites and on highly basic nature oxygen sites (lanthana oxycarbonates), which was found to be beneficial for inhibiting carbon deposition and enhancing catalytic performance [42,56]. TEM images of the used catalysts are shown in Figure 10. No whisker carbon was found in the used Ni@La 2 O 3 /SiO 2 (Figure 10a,b). The Ni mean particle size is about 5 nm, which is slightly larger than that in the reduced Ni@La 2 O 3 /SiO 2 , indicating that the small Ni particles encapsulated into the amorphous La 2 O 3 layer are thermally stable and largely contribute to the inhibition of carbon deposition.
As shown in Figure 10, accessible nickel particles (without encapsulation) favor the formation of carbon. Carbon nanofibers and encapsulated graphitic carbon were formed over the used Ni-La 2 O 3 /SiO 2 catalyst. Methane decomposes on the nickel surface forming atomic hydrogen and carbon, the latter diffusing to free surface sites on the nickel particle to form graphitic carbon (the graphitic carbon peak is precisely seen in Figure 9) [54,55]. As shown in Figure 10c, Ni particles in the used Ni-La 2 O 3 /SiO 2 catalyst are in the 10-50 nm range, which is much wider than that found in the used Ni@La 2 O 3 /SiO 2 catalyst. These results are consistent with the TG and XRD results of the used catalysts. It is mentioned here that deposited carbon can plug the reactor and reduce the lifetime of the catalyst as well. However, in general, the formation of carbon nanofibers does not decrease the exposed Ni surface area of the catalyst, thereby maintaining stable catalytic activity. Ni sintering and the formation of encapsulated carbon can reduce the exposed Ni surface area and thus result in catalyst deactivation.
Based on the Ni@La 2 O 3 /SiO 2 catalyst structure features, it is reasonable to propose that its excellent activity, stability, and high resistance to carbon deposition are much related to the small Ni particles encapsulated within the amorphous La 2 O 3 layer deposited on SiO 2 . Chen et al. [38] has prepared Ni-La 2 O 3 /SiO 2 via one-pot sol-gel method with large specific surface area (190 m 2 g −1 ), which exhibited high activity and excellent stability for DRM at 700°C. However, the rate of deposited carbon on Ni-La 2 O 3 /SiO 2 was 5.9 mg C g cat −1 h −1 , which is much higher than that found in the present  Even though the encapsulation of nanometal particles in core-shell or yolk-shell structures for stabilizing nanometal particles and inhibiting carbon deposition for high-temperature reactions have been reported in the literature, encapsulated metal catalyst using an inert shell, such as SiO2, can always result in lower activity due to the blockage of active sites [57]. There is still research demand to develop a simple method for preparing encapsulated metal catalysts with high activity. In this work, a simple colloidal solution combustion method was used to prepare a Ni@La2O3/SiO2 catalyst with small Ni particles encapsulated within amorphous La2O3 layer supported on SiO2. The Ni@La2O3/SiO2 catalyst obtained exhibited high activity and low carbon deposition rate for the DRM reaction conducted at 700 °C and using 15% CH4, CH4/CO2 = 1. This method is a simple approach that can be widely applied for the preparation of encapsulated metal catalysts.

Synthesis of Catalysts
A Ni@La2O3/SiO2 catalyst with Ni encapsulated within amorphous La2O3 layer on SiO2 support was prepared via a one-pot colloidal solution combustion method. La(NO3)3·6H2O (2.39 g), Ni(NO3)2·6H2O (0.50 g), and glycine (0.60 g) were added in deionized water (6.30 mL). After 20 min of ultrasonic stirring, 1.26 mL of aqueous colloidal SiO2 LUDOX TMA (34 wt%, diameter of 22 nm; Even though the encapsulation of nanometal particles in core-shell or yolk-shell structures for stabilizing nanometal particles and inhibiting carbon deposition for high-temperature reactions have been reported in the literature, encapsulated metal catalyst using an inert shell, such as SiO 2 , can always result in lower activity due to the blockage of active sites [57]. There is still research demand to develop a simple method for preparing encapsulated metal catalysts with high activity. In this work, a simple colloidal solution combustion method was used to prepare a Ni@La 2 O 3 /SiO 2 catalyst with small Ni particles encapsulated within amorphous La 2 O 3 layer supported on SiO 2 . The Ni@La 2 O 3 /SiO 2 catalyst obtained exhibited high activity and low carbon deposition rate for the DRM reaction conducted at 700 • C and using 15% CH 4 , CH 4 /CO 2 = 1. This method is a simple approach that can be widely applied for the preparation of encapsulated metal catalysts.

Synthesis of Catalysts
A Ni@La 2 O 3 /SiO 2 catalyst with Ni encapsulated within amorphous La 2 O 3 layer on SiO 2 support was prepared via a one-pot colloidal solution combustion method. La(NO 3 ) 3 ·6H 2 O (2.39 g), Ni(NO 3 ) 2 ·6H 2 O (0.50 g), and glycine (0.60 g) were added in deionized water (6.30 mL). After 20 min of ultrasonic stirring, 1.26 mL of aqueous colloidal SiO 2 LUDOX TMA (34 wt%, diameter of 22 nm; Sigma-Aldrich, St. Louis, MO, USA) was added to the solution. After 30 min of ultrasonic stirring, the solution was heated to 250 • C. After a few minutes of heating, glycine and nitrate began to react to form metal oxides and release a large amount of gas. In the combustion reaction, glycine and nitrate were used as the fuel and oxidizer, respectively. After the solid was formed, it was calcined at 700 • C for 4 h, and the Ni@La 2 O 3 /SiO 2 catalyst was obtained. The weight contents of Ni and La 2 O 3 in this catalyst were 6.7% and 60.0%, respectively. For comparison, a Ni-La 2 O 3 /SiO 2 catalyst with the same Ni and La 2 O 3 contents was prepared using the same method without adding glycine.

Characterization of Catalysts
N 2 adsorption/desorption curves were obtained using an Autosorb-iQ analyzer (Quantachrome Instruments, Boynton Beach, FL, USA) at −196 • C, to quantify the specific surface area, pore size distribution/mean pore size, and pore volume of the catalysts. The crystal structure of the catalysts was determined by powder X-ray diffraction (XRD). The spectra were collected using a Rigaku-Miniflex 6 (Rigaku Corporation, Tokyo, Japan) powder X-ray diffractometer equipped with CuKα (λ = 0.15406 nm), between 20 • and 80 • (2θ) at a scanning speed of 10 • min −1 .
H 2 -temperature programmed reduction (H 2 -TPR) was applied on a TP-5080 multifunctional adsorption apparatus (Xianquan, Tianjin, China) in 5% H 2 /Ar gas mixture with a heating rate of 10 • C min −1 . Transmission electron microscopy (TEM) was performed on a Tecnai G2 F20 microscope (FEI Company, Hillsboro, OR, USA), to directly observe the morphology and size of Ni particles and of deposited carbon after DRM. To determine the amount of carbon accumulation of the used catalysts, thermogravimetric (TG) and differential thermal analysis (DTA) were conducted on an HCT-1 TG thermal analyzer (Henven Scientific Instruments, Beijing, China).

Catalysts Performance Evaluation for the DRM Reaction
The catalytic performance of the Ni-based solids was evaluated at atmospheric pressure in a fixed-bed tubular quartz reactor (internal diameter 8 mm, length 300 mm). The total flow rate of reaction gases was 100 mL min −1 and the amount of catalyst used was in the 5-50 mg range. The corresponding WHSV was in the 120,000-1,200,000 mL g −1 h −1 range. CH 4 , CO 2 , H 2 , and Ar (99.999% purity) were purchased from Shanghai Maytor special Gas Co. Ltd. These gases contained less than 1 ppm of water vapor and were used without further purification. Before reaction, the catalyst was in-situ reduced in 20% H 2 /Ar gas mixture at 700 • C for 1.5 h. The reaction gases of CO 2 , CH 4 , and Ar at a molar ratio of 15/15/70 (vol%), were introduced into the reactor at 700 • C, and Ar gas was used as the internal standard. The gas effluent was analyzed using two on-line gas chromatographs of G5 (Beijing Purkinje General Instrument Co., Ltd., Beijing, China). One chromatograph used hydrogen as a carrier gas to detect Ar, CO, CH 4 , and CO 2 . Another chromatograph used N 2 as a carrier gas to detect H 2 .
The CO 2 conversion (X CO2 ), CH 4 conversion (X CH4 ) and the H 2 /CO gas product ratio were calculated based on the following Equations (1)-(3): where [x] in and [x] out represent the mole fraction of x gaseous species in the inlet feed and outlet from reactor gas mixture, respectively.
The H 2 and CO product yields were calculated based on the following Equations (4) and (5):

Conclusions
In this study, we prepared a Ni@La 2 O 3 /SiO 2 catalyst with encapsulated Ni nanoparticles via the colloidal solution combustion method tested for the DRM reaction at 700 • C. In the Ni@La 2 O 3 /SiO 2 catalyst, small Ni particles were encapsulated within an amorphous La 2 O 3 layer, where this was coated on SiO 2 . Due to the encapsulated Ni micro-structure, more interface between Ni and La 2 O 3 was formed, and the Ni@La 2 O 3 /SiO 2 catalyst exhibited excellent activity and stability and strong resistance to carbon deposition during DRM reaction. The catalytic performance results indicated that the CH 4 conversion rate of the Ni@La 2 O 3 /SiO 2 catalyst was five times higher than that of Ni-La 2 O 3 /SiO 2 catalyst. More importantly, the Ni@La 2 O 3 /SiO 2 catalyst exhibited excellent catalytic stability and only a slight deactivation for 80 h on reaction stream. TG-DTA studies revealed that 1.6 wt% carbon was deposited on the Ni@La 2 O 3 /SiO 2 catalyst after 50 h of DRM, which was much lower than that of 11.5 wt% obtained on the Ni-La 2 O 3 /SiO 2 catalyst.