Alumina Support for Cobalt Catalyst in a Methane Dry Reforming Reaction: The Role of Water Content in a Solvent Medium

This study aimed to synthesize alumina from an inorganic aluminum nitrate precursor in various binary solvent systems of ethanol and water using the sol-gel self-assembly (SSA) method, employing a triblock copolymer, pluronic P123, as the pore-directing agent. The resulting materials were implemented as a support for the cobalt (Co) catalyst in a methane dry reforming (MDR) reaction at 1073 K under 1 atm. Regardless of the water percentage used in the support synthesis, the methane dry reforming reaction over Co catalysts on alumina supports showed the negligible change in conversion during the 12 h reaction. Moreover, there was evidence of large quantities of amorphous carbon and graphitic carbon on the spent catalyst surface. However, the low oxidation temperature of these deposited carbons could help maintain the balance between the carbon formation and the carbon elimination processes on the catalyst surface during the reforming reaction, hence prolonging the lifetime of the catalyst. The high conversion of methane (CH4) from 64.6% to 82.8% and carbon dioxide (CO2) from 70.7% to 86.6% for the MDR reaction over the as-prepared alumina-supported Co catalyst demonstrated a significant improvement in catalyst production for the MDR reaction from the viewpoint of large-scale applications.


Introduction
e increase in the average temperature and climate change caused by greenhouse gases has become serious global issues.
e human activities such as the burning of oil, coal, and gas, as well as deforestation, are associated with energy-related carbon dioxide (CO 2 ) emissions in the atmosphere.
erefore, in addition to the need to find new eco-energy sources, the use of mineral resources in an ecological and environmental approach is also a significant concern. Consequently, the methane dry reforming (MDR) method has emerged as a potential approach for producing syngas from CO 2 and methane (CH 4 ), which is a significant feedstock for downstream petrochemical processes [1][2][3][4][5][6]. Although this approach could have environmental and economic benefits, the catalyst limitations have impeded it from wide-ranging applications in large-scale industrial production.
Noble metals, such as rhodium (Rh), ruthenium (Ru), and platinum (Pt), have significant catalytic activity for the MDR process in terms of conversion and coking inhibition [7]. However, the unavailability and expensive cost of these catalysts are major drawbacks preventing their use in industrial applications. To date, cobalt-(Co-) based catalysts have garnered considerable attention since they possess comparable catalytic activity and higher stability against temperature variations in comparison to noble metals [8,9]. Moreover, different metal oxides have been evaluated as the support for a Co-catalyzed MDR reaction, such as oxides of the alkaline Earth elements, including magnesium oxide (MgO), calcium oxide (CaO) [8], ceric dioxide (CeO 2 ) [10], lanthanum dioxide (LaO 2 ) [11], strontium oxide (SrO) [12], aluminum oxide (Al 2 O 3 ) [13][14][15][16], and Santa Barbara Amorphous-15 (SBA-15) [17][18][19]. Among of these materials, mesoporous alumina (MA) has been proven to be a potential support because of its availability. Recent contributions in heterogeneous catalysis regarding porous support and mesoporous structure materials have been widely used as catalyst supports since they can facilitate the dispersion of the catalysts and confine the active particles inside their matrix, preventing them from aggregating during the reaction [17,[19][20][21]. Mesoporous alumina support is one of the materials that has a high potential for being screened for the same effects in a MDR reaction.
Sol-gel self-assembly (SSA) is a common approach used for mesoporous Al 2 O 3 production due to its easy, accessible, and reproducible characteristics in fabricating mesoporous structures [22]. Most SSA processes have been conducted by employing an organic salt precursor dispersed on a soft template dissolved in anhydrous ethanol (C 2 H 5 OH) [22,23]. However, the toxicity and the high cost of organic salt and anhydrous (C 2 H 5 OH) ethanol have made them the most unlikely substances for this purpose. us, interest in using a less expensive and readily available inorganic aluminum precursor in large-scale applications is increasing. Additionally, the presence of water in the solvent has a significant influence on the pore structure of the resulting alumina materials [24], thus enabling intrapellet diffusion of the active nanoparticles in the porous framework [25]. To the best of our knowledge, only a few studies have investigated the combination of an inorganic aluminum precursor and a binary solvent mixture (C 2 H 5 OH in H 2 O) for synthesis of alumina using the SSA method, to act as a support for a Co catalyst in an MDR reaction. erefore, instead of using an organic salt precursor and anhydrous ethanol, the present study employed a combination of an inorganic aluminum precursor and a binary solvent mixture (C 2 H 5 OH in H 2 O) for alumina synthesis using the SSA method. e performance of the Co catalyst on the as-prepared supports for the MDR reaction was evaluated to determine the amount of water needed in the solvent when preparing the support. were obtained from Sigma-Aldrich (St. Louis, Missouri, US). C 2 H 5 OH (99.9%) was obtained from VWR Chemicals (Darmstadt, Germany). All the reagents were used directly without any further purification. All the gases, including CH 4 , CO 2 , nitrogen (N 2 ), and hydrogen (H 2 ), were of analytical grade and provided by Air Products and Chemicals, Inc.

Catalyst
Synthesis. Al 2 O 3 was prepared by dissolving 0.98 g of P-123 in 14.6 ml of the C 2 H 5 OH-H 2 O solvent mixture with the following proportions of water: 0%, 25%, 50%, and 75%. is solvent mixture was mixed at an ambient temperature for 30 min, followed by adding 3.68 g of Al(NO 3 ) 3 ·9H 2 O and dropping 1.6 ml HCl (37%) solution.
e obtained blend was additionally stirred for 60 min prior to undergoing the hydrothermal process at 373 K for 24 h in an autoclave. e obtained mixture was dried in the oven for 48 h at 333 K and then calcined in a furnace for 5 h at 1073 K. e Al 2 O 3 -supported Co catalyst was prepared using the incipient wetness impregnation method. In particular, 0.28 g of the Co(NO 3 ) 2 ·6H 2 O precursor was mixed with 0.25 ml of C 2 H 5 OH, and the resulting solution was sprayed on 0.5 g of as-synthesized alumina. e resulting mixture was dried overnight at 373 K and then calcined at 873 K for 5 h. e individual alumina support, prepared from the binary solvent system with a water content of 0%, 25%, 50% and 75%, was labelled as MA00, MA25, MA50, and MA75, respectively. Consequently, the supported Co catalysts synthesized from the abovementioned support were denoted as 10Co/ MA00, 10Co/MA25, 10Co/MA50, and 10Co/MA75.

Catalyst Properties.
e phases and crystalline structure of the selected spent catalysts were determined using a Rigaku Miniflex 600 X-ray diffraction instrument, which employed a copper (Cu) target as the radiation source at the wavelength of 1.5418Å. e test specimen was scanned from 3°to 80°with the speed of 1°min −1 .
e amount of deposited carbons on the spent catalysts was quantified via the thermal programmed oxidation conducted on a TA TGA Q500 equipment (TA Instruments, Newcastle, DE, USA). In particular, the sample was heated at 373 K in N 2 atmosphere for 30 min to eliminate the volatile compounds, followed by increasing the temperature to 1023 K in a mixture flow of 20% O 2 in N 2 with a ramping rate of 10 K min −1 . e oxidation stage at 1023 K was left for an additional 30 min prior to cooling to room temperature. e surface morphology of the selected catalysts was elucidated using scanning electron microscope (SEM) equipped with an energy-dispersive X-ray (EDX) detector (Hitachi Tabletop Microscope TM3030 Plus unit, Hitachi High Technologies Corp., Tokyo, Japan) and a Raman spectrometer employing 532 nm laser excitation (JASCO NRS-3100, Tokyo, Japan).

Catalytic Activity Evaluation.
e fixed-bed reactor with a 3/8 diameter was used to evaluate the MDR reaction catalyzed by the alumina-supported Co at a fixed gas hourly space velocity (GHSV) of 36 L·g cat −1 ·h −1 . Prior to each assessment, H 2 reduction was done in situ at 1073 K for 1 h using a mixed flow of 50% H 2 in N 2 . e output products were analyzed in an Agilent 6890 gas chromatography (Agilent Technologies, Santa Clara, CA, USA). e reactant conversions (X i with i: CH 4 and CO 2 ) and product yields 2 International Journal of Chemical Engineering (Y CO and Y H 2 ) were estimated using the following equations, respectively: where F in and F out are the inlet and outlet molar flow rates (mol s −1 ), respectively. , were applied as a support for the Co catalyst, and the catalytic performance of the substrates for the MDR reaction was evaluated at a stoichiometric feed ratio and temperature of 1073 K. As seen in Figures 1-4, the conversion of the reactants and the product yield were unchanged within the 12 h reaction, indicating the firm stability of the four catalysts under MDR reaction conditions. e conversion of CO 2 and CH 4 in the MDR reaction showed a decreasing trend from 86.6% to 70.7% and from 82.8% to 64.6%, respectively, in the order of 10Co/ MA00 > 10Co/MA25 > 10Co/MA50 > 10Co/MA75 catalysts. is decrease in catalyst performance along with an increase in the water concentration in the solvent for the support preparation could be due to the modification in the support's pore structure, which was caused by the increase in the water content.

Results and Discussion
us, the MDR reaction performance strongly depends on the support's features and the active Co metal properties, such as crystal size and dispersion [26]. Furthermore, it has been reported that the percentage of water in the solvent mixture has a significant impact on enlarging the pore diameter of MA produced using the SSA method [23], hence facilitating the generation of active metals with an appropriate size and enhancing the intrapellet diffusion of both the reactants and the products [25]. e effectiveness of the MDR reaction was consolidated via the time on stream (TOS) plot of the H 2 and CO formation yields, as seen in Figures 3 and 4. e support in catalyst activity was observed to play a significant role in both the CO and H 2 yields. In particular, the H 2 and CO yields were highest in the case of the 10% Co/MA00 catalyst; they were about 62.4-68.7% and 71.2-76.0%, respectively. When the MDR reaction was conducted with the 10% Co/ MA75 catalyst, the CO yield decreased by approximately 10-20% and the H 2 yield decreased by approximately 20-30%. Regardless of the type of support, the H 2 /CO ratio was always <1, thus proving the coexistence of the reverse water-gas-shift process in the MDR reaction [27].

Effect of CH 4 and CO 2 Partial Pressure on the MDR Reaction.
To evaluate the partial pressure influence of CH 4 (P CH 4 ) and CO 2 (P CO 2 ) on MDR, the reactions over the 10% Co/MA00 catalyst were conducted with P CH 4 and P CO 2 in the range of 10-40 kPa at a temperature of 1023 K. Figure 5 shows the correlation between the CH 4 and CO 2 conversions with the change in P CH 4 at P CO 2 of 20 kPa (Figure 5(a)) and the change in P CO 2 at a fixed P CH 4 of 20 kPa (Figure 5(b)). e CH 4 conversion gradually decreased by approximately 15.0% as P CH 4 increased from 10 kPa to 40 kPa (see Figure 5(a)). is decrease in the CH 4 conversion was due to the upsurge in carbon formation caused by extreme CH 4 cracking in the surplus of the CH 4 feedstock. e formed carbon induced active sites blocking and hindering the catalytic performance of the MDR reaction [28]. In contrast, the CO 2 conversion increased and reached 80.6% when the P CH 4 increased from 10 kPa to 40 kPa. us, the increase in P CH 4 resulted in superior CH 4 adsorption on the catalyst, therefore improving the CO 2 consumption through the MDR reaction [29]. A similar trend was also observed for the MDR reaction over the CeO 2 -supported Co catalyst [28].
However, the CH 4 conversion showed continuous improvement to 82.7% when P CO 2 increased from 10 kPa to 40 kPa, while P CH 4 was kept at 20 kPa (see Figure 5(b)). is behaviour can be linked to the enhanced elimination of the deposited carbon, as depicted in equation (4) [30], and the coinciding existence of the CH 4 steam reforming process that was due to the increase in H 2 O via the reverse water-gasshift process, as follows [31]: A decrease in the CO 2 conversion was found when P CO 2 increased from 10 kPa to 40 kPa (see Figure 5(b)), which could be due to the deficiency of CH 4 in reacting with the CO 2 -rich feedstock. Furthermore, the presence of excess CO 2 in the reactor could intensify the possibility for active Co particles to be oxidized as follows, which resulted in a decrease in CO 2 adsorption [27]:

XRD Analysis.
e XRD spectrum of the synthesized 10% Co/MA00 after 12 h MDR reaction at 1073 K is shown in Figure 6. e peaks at 2θ of 37.6°, 45.6°, and 67.0°were assigned to the Al 2 O 3 phase (JCPDS card no. 04-0858); the signals at 31.7°, 37.5°, and 44.6°corresponded to the Co 3 O 4 crystalline phase (JCPDS card no. 74-2120). Moreover, the metallic Co presence was verified via the detection of a peak at 51.6°(JCPDS card no. . e copresence of Co metallic and oxide particles in the 10% Co/MA00 sample indicated the occurrence of the redox cycle of the Co species during the MDR reaction. Notably, a broad peak at 2θ from International Journal of Chemical Engineering 15.0°to 30.0°was deconvoluted into two separate diffraction signals, denoted as α and β, and displayed in a small inserted picture in Figure 6. e α peak represents amorphous carbon, and the β peak represents graphitic carbon [32]. Figure 7 shows the Raman spectrum of the 10% Co/MA00 after 12 h MDR reaction at 1073 K. Four active Raman peaks at 473.9 cm −1 , 517.7 cm −1 , 609.7 cm −1 , and 678.5 cm −1 , illustrated in the embedded picture, were assigned to the corresponding Co 3 O 4 symmetric vibrational modes of E g , F 2g , F 2g , and A 1g . e signal at 678.5 cm −1 represents the vibration of the octahedral site, and the E g and F 2g modes are likely related to the mixed vibrations of the octahedral site and the tetrahedral oxygen movements [33]. ese typical peaks were observed in the Raman spectra of the spent catalyst, suggesting that the reduced Co 0 species in the H 2 pretreatment were reoxidized to Co 3 O 4 during the MDR reaction. Two peaks at 1338.5 cm −1 and 1573.5 cm −1 demonstrate the heterogeneity of the surface carbons including ordered carbon-like graphite (G-band) and amorphous carbon (D-band). e D-band was attributed to amorphous carbon or carbon nanofibers, while the G-band arises from the stretching of the C-C bond in graphitic carbon [34].

Surface Morphology Analyses.
e SEM-EDX measurements of the 10% Co/MA00 catalyst after MDR at 1073 K suggest the presence of filamentous carbons or  carbon fibers, as shown in Figure 8. e percentage of total carbon in the catalyst was around 64.03% based on the normalized EDX results. is could be due to the fact that the alumina-supported Co catalyst is known to boost the formation of filamentous carbon in a CH 4 atmosphere at high temperature [35].

TPO Measurements.
ermal-programmed oxidation (TPO) was performed to quantify the sum of the carbon deposits on the 10% Co/MA00 catalyst after MDR at 1073 K. As seen in Figure 9, about 77.1% weight of the sample was lost after the oxidation at temperatures ranging from 700 K to 850 K, which is in agreement with the EDX results (see Table 1). e differences in the results obtained from the two analytical methods could be due to the decomposition of the other elements in the TPO measurement.
Moreover, the low oxidation temperature of the deposited carbon at temperatures ranging from 750 K to 820 K, as determined from the derivative weight curves, suggest that all the deposits were well gasified in the reforming conditions; hence, they did not cause a loss of intrinsic catalyst activity or lengthen the lifetime of the catalyst [13].

Conclusions
e performance of the Co catalysts supported on MA for the MDR reaction was investigated in terms of the support contribution. e water content in the solvent mixture applied for the Al 2 O 3 support synthesis plays a crucial role in assembling the structure; hence, it influences the catalytic activity in the MDR reaction. Regardless of the type of support used, the Co catalysts showed good stability under 12 h MDR reaction. Notably, the 10% Co/MA00 catalyst demonstrated the highest activity for the MDR reaction with a carbon monoxide yield of 71.2-76.0%, and the deposit on the spent 10% Co/MA00 catalyst surface was found to consist of both amorphous carbon and graphitic carbon possessing low-oxidation temperature property and hence easily be eliminated in situ the reaction process. Moreover, the reactant partial pressure was found to have a significant impact on the CO 2 and CH 4 conversions as well as the product yields when the MDR reaction was conducted at 1023 K.

Data Availability
All experimental data used to support the findings of this study are available within the article.

Conflicts of Interest
e authors declare that they have no conflicts of interest.