Effect of Calcination Temperature on Cu-Modified Ni Catalysts Supported on Mesocellular Silica for Methane Decomposition.

Catalytic methane decomposition has been considered suitable for the green and sustainable production of high-purity H2 to help reduce greenhouse gas emissions. This research developed a copper-modified nickel-supported mesocellular silica NiCu/MS(x) catalyst synthesized at different calcination temperatures to improve the activity and stability in the CH4 decomposition reaction at 600 °C. Ni and Cu metals were loaded on a mesocellular silica (MS) support using a co-impregnation method and calcined at different temperatures (500, 600, 700, and 800 °C). The NiCu/MS(600) catalyst not only had the highest H2 yield (32.78%), which was 1.47-3.87 times higher than those of the other NiCu/MS(x) catalysts, but also showed better stability during the reaction. Calcination at 600 °C helps improve the active nickel dispersion, the reducibility of the NiCu catalyst, and the interaction of the NiCu-MS support, leading to the formation of fishbone and platelet carbon nanofibers via a tip-growth mechanism, resulting in the NiCu metals remaining active during the reaction.

Introduction

Hydrogen (H2) gas has been considered as one of the key energy carriers and a key fuel for the 21st century, as it not only shows great potential in a wide variety of industrial applications but can also reduce environmental impacts.[1−3] H2 has been widely used in the chemical process industry, electronics, and food processing.[3−5] At present, a large-scale hydrogen production is based on methane steam reforming (MSR) and partial oxidation of methane (POM). In MSR and POM reactions, carbon dioxide (CO2) and carbon monoxide (CO) are simultaneously produced during the processes.[6−8] Therefore, these processes are not environmentally friendly, as they need considerable investment in both equipment and energy consumption to separate the CO2 and CO from the H2.

Catalytic methane decomposition is a green and sustainable process that is consistent with reducing greenhouse gas emission because a methane molecule is directly decomposed to COx-free H2 and carbon filament products.[2,3,9−13] Carbon filaments, such as carbon nanofibers and carbon nanotubes produced through this process have been widely used in many industrial applications, including functional materials, catalysts, hydrogen storage, and electrodes.[14−16] Catalytic methane decomposition can occur in a relatively lower temperature range compared to MSR and POM reactions.[3,4] However, breaking the methane molecule requires a lot of energy (high temperature) due to its symmetrical structure and the strong C–H bond (440 kJ/mol). Therefore, the use of an appropriate metal catalyst is necessary. Non-novel transition metals, such as nickel (Ni), cobalt (Co), and iron (Fe), have been reported to be active in methane decomposition due to their common, unique property of partly filled 3d orbitals.[2,17,18] In particular, a Ni catalyst is highly active at low reaction temperatures (500–600 °C) compared to the other catalysts because it has the greatest decrease in the dissociation energy of the C–H bonds;[2−4] however, a Ni-based catalyst is easily deactivated by the sintering effect at a higher temperature (greater than 600 °C) and by carbon encapsulation, resulting in lower catalytic activity during the reaction.[3,4,19]

Notably, the activity and stability of a Ni-based catalyst can be improved by selecting suitable supports and promoters because these supports have a substantial role in the catalytic reaction as they help to increase the metal catalyst dispersion and therefore, improve the activity of catalysts while minimizing the catalyst sintering. Many types of materials such as silica[20−23] (SiO2), alumina[21] (Al2O3), zeolite,[21] and metal oxides,[21,23−25] such as ZrO2, MgO, and TiO2, have been used as catalyst supports for catalytic methane decomposition. SiO2 material has been widely used as support because the interaction between metal and silica support is quite weak compared to the other supports in which the metal–support interaction is one of the factors determining the activity and lifetime of the catalyst.[3,4,22] Our previous work found that the SiO2 structure support directly affected activity and stability in the methane decomposition reaction. Ni supported on bimodal porous silica with mesopore and macropore structures produced higher activity and stability than that of Ni supported on monomodal porous silica with straight and sinusoidal mesopore[26] because a larger pore size of the support enhanced the diffusion of the reactant gas during the reaction. However, the activity of catalysts with a larger pore size still decreased gradually during the reaction due to their lower active metal dispersion.[26] Mesocellular silica (MS) is a type of porous silica that has a 3D structure with ultralarge (15–50 nm) pore diameters that are accessible via large windows (5–20 nm). In addition, the pore characteristic of MS can be controlled depending on the synthesis conditions. The MS structure has the advantage of better diffusion of the reactant for its 3D continuous mesopore system and interconnected windows.[27,28] Hence, MS material is of interest for use as a catalyst support.

The addition of a promoter metal can help to enhance the metal-catalyst dispersion and reducibility, assuring improving catalyst activity and simultaneously prolonging the catalyst’s lifetime during the reaction. Various types of metals such as Cu, Pt, Pd, Fe, Co, and Mo, have been used as a promoter for catalyst methane decomposition.[3,4] In particular, Cu metal has the greatest potential as a Ni-based catalyst. Many researchers have reported the effect of bimetallic NiCu supported on different types of supports, including Al2O3, MgO, MgO·Al2O3, MCM-41, SiO2, and CNTs in the CH4 decomposition reaction.[2,14,17,29−33] It was found that the addition of Cu could noticeably enhance catalytic stability compared to the monometallic Ni catalyst because Cu could improve the metal dispersion and reducibility of the nickel catalyst.[14,17,25,33] In addition, the Ni–Cu alloy is notable for enhancing the stability and activity of the catalyst at high temperatures; furthermore, it promotes the carbon formation rate because of favoring carbon diffusion through the higher lattice constant of the NiCu alloys.[30−32] However, the NiCu catalyst was still deactivated during the methane decomposition reaction.

Therefore, this research developed Ni–Cu supported mesocellular silica [NiCu/MS(x)] catalysts synthesized at different calcination temperatures to improve the activity and stability in the CH4 decomposition reaction at 600 °C. The effect of calcination temperature (500, 600, 700, and 800 °C) on the structural and chemical surface properties of Ni and Cu species on hydrogen production, carbon product allotrope, and the deactivation behaviors of NiCu/MS(x) catalysts were discussed in detail. It was found that the NiCu/MS(x) catalyst calcined at 600 °C produced the highest CH4 conversion and H2 yield; concurrently, the activity was stable during the reaction. The calcination temperature significantly affected the reduction behavior of the NiCu catalyst, active metal dispersion, and the interaction of the Ni and MS support, directly resulting in the decomposition of CH4 to H2, and the formation of carbon nanofibers (CNFs), as well as deactivation of the catalyst.

Results and Discussion

Structural and Chemical Surface Properties of NiCu/MS(x) Catalysts

The textural properties of the nitrogen adsorption–desorption isotherms and pore size distributions of NiCu/MS(x) catalysts are shown in Figure S1. All NiCu/MS(x) catalysts had the type IV isotherm with an H3 hysteresis loop, indicating the presence of a mesoporous structure with aggregated, plate-like particles consisting of macropores that were not completely filled with pore condensate.[34] As shown in Figure S1B, each NiCu/MS(x) catalyst had two small mesopore sizes at the main peaks of 3.71 and 6.78 nm, with a larger mesopore size at the main peak of 45 nm; however, the pore size at the peak intensities of 6.78 and 45 nm decreased when the calcination temperature increased. The specific surface area, pore size, and pore volume of catalysts are reported in Table . With the increase in calcination temperature, the specific surface area and pore volume tended to decrease because the nickel–copper particles easily agglomerated into larger particle size and blocked the MS support pores.

Table 1

Textural and Chemical Properties of NiCu/MS(x) Catalysts

catalysts	surface area (m2/g)	pore diameter (nm)	pore volume (cm3/g)	metallic Ni crystallinea size (nm)	active Ni surface areab (m2/gmetal)
NiCu/MS(500)	181	3.70, 6.76	0.28	39.8	0.083
NiCu/MS(600)	164	3.71, 6.77	0.25	35.4	0.052
NiCu/MS(700)	141	3.70, 6.77	0.25	37.3	0.024
NiCu/MS(800)	88	3.70, 7.37	0.20	36.6	0.022

Calculated from the Scherrer equation.

Calculated based on CO pulse chemisorption.

The structures of the NiCu/MS(x) catalysts and the nickel and copper dispersions on the MS support calcined using different temperatures were observed using the TEM technique. Figure S2a,c focused on the MS support structure; a foam-like structure was found in all catalysts, with this structure comprising various-sized interconnected spherical cells.[35] The size and dispersion (Figure ) of nickel oxide and copper oxide consisted of NiO nanoparticle sizes of NiCu/MS(500), NiCu/MS(600), NiCu/MS(700), and NiCu/MS(800) in the ranges 11.2–47.0, 5.1–37.7, 14.6–55.0, and 8.0–42.9 nm, respectively. The nickel–metal was well dispersed on the MS support in all NiCu/MS(x) catalysts. At higher calcination temperature, the larger NiO particle size located at the edge and on the external surface of the MS support.

Figure 1

TEM image and EDS-mapping of nickel and copper in the NiCu/MS catalyst calcined at different temperatures. (a) NiCu/MS(500), (b) NiCu/MS(600), (c) NiCu/MS(700), and (d) NiCu/MS(800).

The particle size of CuO could not be measured because the amount of copper loading (5 wt %) was quite low compared to the nickel loading (50 wt %). However, the copper metal was well dispersed on the NiCu/MS catalyst.

The reduction temperature and interaction between nickel, copper, and the MS support of Ni–Cu/MS(x) catalysts calcined at different temperatures were analyzed using the H2-temperature-programmed reduction (H2-TPR) technique, as shown in Figure . All NiCu/MS(x) catalysts had three reduction peaks. The first region in the range 70–190 °C was attributed to reduction of the CuO phase to metallic copper (Cu0).[2,14] The second region in the range 190–260 °C corresponded to the reduction of bulk NiO on the MS support.[2,25] In this region, the reduction peak of the NiCu/MS(600) catalyst shifted to a lower reduction temperature due to the smaller size of its nickel oxides on the surface. The last reduction range of 260–350 °C indicated the reduction of NiO that strongly interacted with the MS support.[2,29] In our previous work,[36] the reduction temperatures of monometallic Ni loaded on the silica support were in the range 400–600 °C, implying that nickel strongly interacted with the silica support. The addition of Cu into the Ni catalyst could facilitate the reducibility of NiO at a lower temperature through the adsorption of hydrogen on copper oxide.[14,37] Furthermore, the calcination temperature affected the interaction between Ni–Cu and the MS support. At calcination temperatures of 500 and 800 °C, the interactions between Ni–Cu and MS support were much stronger compared to the catalysts calcined at 600 and 700 °C.

Figure 2

H2-TPR profiles of fresh NiCu/MS(x) catalysts. (a) NiCu/MS(500), (b) NiCu/MS(600), (c) NiCu/MS(700), and (d) NiCu/MS(800).

The effect of the calcination temperature on the phase transformation of the nickel metal on the NiCu/MS(x) catalyst during the reduction process was proved using in situ X-ray absorption near edge structure (XANES) spectroscopy. The normalized Ni K-edge XANES spectra of each catalyst compared with the spectrum of NiO and Ni foil (metallic Ni) standards are presented in Figure A. The nickel species of all calcined NiCu/MS catalysts was the NiO phase. The NiO phase was slightly changed to metallic nickel during the reduction process under a hydrogen atmosphere. The NiO on NiCu/MS(600) catalyst began to change to metallic nickel species at a lower temperature compared to those of the NiCu/MS(x) catalysts due to the weaker interaction of Ni and the MS surface (confirmed by the H2-TPR data, Figure ). After the catalyst had been reduced using hydrogen gas at 700 °C for 1.5 h, the NiO metal was completely changed to metallic nickel (Ni0) as an active species in all NiCu/MS(x) catalysts.

Figure 3

In situ Ni K-edge XANES spectra of fresh catalysts during the reduction process from 50 to 700 °C and 90 min holding time: (a) NiCu/MS(500), (b) NiCu/MS(600), (c) NiCu/MS(700), and (d) NiCu/MS(800). Inset shows spectra of nickel standards.

The copper phases on the NiCu/MS(x) catalysts after the calcination process and after the reduction process at 700 °C with 90 min holding time were confirmed using in situ XANES spectroscopy, by comparing with each standard spectra of copper oxide and copper foil, as shown in Figure A,B, respectively. All the calcined NiCu/MS(x) catalysts had a white line peak at 8996.7 eV, which was the same as for the CuO standard spectra. After the reduction process, all the reduced NiCu/MS(x) catalysts had a pre-edge peak (8994.5 eV) and an edge jump (9003.3 eV) similar to those of metallic copper (Cu0), showing that the CuO species on the fresh NiCu/MS(x) catalysts were completely changed to metallic copper prior to the methane decomposition reaction. The fine structure of the white line was slightly different due to the effect of copper particle size.[36]

Figure 4

In situ Cu K-edge XANES spectrum of catalysts (A) after calcination process and (B) after reduction process at 700 °C and 90 min holding time: (a) Cu foil, (b) Cu2O, (c) CuO, (d) NiCu/MS(500), (e) NiCu/MS(600), (f) NiCu/MS(700), and (g) NiCu/MS(800).

The chemical states of nickel and copper on the reduced NiCu/MS(x) catalysts calcined at different temperatures were characterized in the binding energy regions of Ni 2p3/2 and Cu 2p3/2, as shown in Figure A,B, respectively. From the peak characteristics in the Ni 2p3/2 core level spectra, the peaks at 852.9 ± 0.3, 854.6 ± 0.2, and 856.2 ± 0.4 eV, corresponding to metallic nickel, NiO (Ni2+), and Ni(OH)2 (Ni2+),[38,39] respectively, were found in all Ni–Cu/MS(x) catalysts. With the Cu 2p3/2 core level spectra, the XPS spectra of all NiCu/MS(x) catalysts were deconvoluted into three Gaussian peaks at 932.4 ± 0.6 and 934.3 ± 1.1 eV, ascribed to the reduced copper species (Cu0/Cu+) and CuO (Cu2+),[40,41] respectively. NiO and CuO peaks were found in all reduced catalysts, probably due to unavoidable air oxidation of the catalyst during loading into the instrument. The binding energies of Ni 2p and Cu 2p of the reduced NiCu/MS(600) catalyst shifted lower, which could be explained by a charge transfer from Cu in close contact with Ni metal by means of a high portion of Ni–Cu alloy formation.[20] For the reduced NiCu/MS(700) and NiCu/MS(800) catalysts, the binding energies of Ni 2p and Cu 2p shifted higher, which could be explained by the charge compensation phenomenon of the binary NiCu-supported catalysts. Naghash et al. reported that at temperatures above 700 °C, the interfacial energy of copper and nickel was high; therefore, the copper and nickel atoms tended to segregate rather than be uniformly dispersed on the support surface.[42]

Figure 5

XPS narrow-scan spectra of (A) Ni 2p3/2 and (B) Cu 2p3/2 regions of reduced NiCu/MS(x) catalysts. (a) NiCu/MS(500), (b) NiCu/MS(600), (c) NiCu/MS(700), and (d) NiCu/MS(800).

The metal crystalline phases of the reduced NiCu/MS(x) catalysts calcined at different temperatures were characterized using XRD measurements (Figure ). All NiCu/MS(x) catalysts had four diffraction peaks. The diffraction peaks at 44.42° and 51.76° were assigned to the crystal planes (111) and (200) of the Ni–Cu alloy (JCPDS. 47-1406), respectively. The shoulders, which were located at 44.51° and 51.92° corresponded to the crystal planes (111) and (200) of the metallic nickel (JCPDS. 4-0850), respectively.[14,42,43] The diffraction peaks of copper were not found in any catalysts because the crystallite size of copper could be below the detection limit of the XRD technique (<5 nm). The crystallite size of metallic nickel on the NiCu/MS catalysts was calculated using the Scherrer equation and the result is presented in Table . The smallest and largest metallic nickel crystallite sizes were observed with the NiCu/MS(600) and NiCu/MS(500) catalysts, respectively.

Figure 6

XRD patterns of reduced NiCu/MS(x) catalysts. (a) NiCu/MS(500), (b) NiCu/MS(600), (c) NiCu/MS(700), and (d) NiCu/MS(800).

The active surface area of nickel on the NiCu/MS(x) catalysts with different calcination temperatures after the reduction process with H2 gas at 700 °C for 1.5 h was examined using a CO chemisorption technique; the results are presented in Table . The highest and lowest nickel active surface areas were for NiCu/MS(500) (the lowest calcination temperature) and Ni–Cu/MS(800) catalyst (the highest calcination temperature), respectively, because the higher calcination temperature resulted in the agglomeration of nickel and copper metals of a larger size, and therefore, the active nickel surface area decreased.

The calcination temperature directly affected the structural and chemical properties of the NiCu/MS(x) catalysts in terms of nickel metal dispersion, a portion of Ni–Cu alloy formation, the reduction behavior of nickel, and NiCu–MS support interaction. A higher temperature resulted in higher interfacial energies of copper and nickel, resulting in the formation of nickel and copper clusters rather than the nickel–copper alloy phase on the support surface.

NiCu/MS(x) Catalyst Performance in Methane Decomposition

The effect of calcination temperature on the activity and stability of NiCu/MS(x) catalysts in methane decomposition was investigated at 600 °C under atmospheric pressure in a packed-bed reactor. The CH4 conversion and H2 yield of the NiCu/MS(x) catalysts calcined at different temperatures are shown in Figure a,b, respectively. The NiCu/MS(600) catalyst had the highest CH4 conversion (58%) at commencement and the activity decreased to 40% conversion within 60 min and was quite stable until 3 h of reaction time. In addition, the NiCu/MS(600) catalyst had the highest H2 yield. The NiCu/MS(800) catalyst had a higher CH4 conversion and H2 yield at reaction commencement, but its performance continuously decreased with the increase in reaction time and ultimately was lower than those of the NiCu/MS(600) and NiCu/MS(700) catalysts. The NiCu/MS(700) catalyst had higher activities in terms of CH4 conversion and H2 yield compared to the NiCu/MS(800) catalyst, especially during 60–180 min. The NiCu/MS(500) catalyst had the lowest CH4 conversion and H2 yield, as well catalyst activity being significantly deactivated during the reaction time.

Figure 7

(A) CH4 conversion and (B) H2 yield of the NiCu/MS(x) catalyst calcined at different temperatures in methane decomposition at 600 °C. ● (red) NiCu/MS(500), ▲ (blue) NiCu/MS(600), ▼ (green) NiCu/MS(700), and ⧫ (gray) NiCu/MS(800).

At the end of the reaction period (180 min), NiCu/MS(600) catalyst produced H2 gas expressed in terms of H2 yield (approximately 32.78%) that was 3.87, 1.47, and 1.75 times higher than for the NiCu/MS(500), NiCu/MS(700), and NiCu/MS(800) catalysts, respectively. The amount of carbon on the NiCu/MS(x) catalysts was calculated based on the thermogravimetric analysis (TGA) data, as presented in Table . The weight loss curves of the NiCu/MS(500), NiCu/MS(600), NiCu/MS(700), and NiCu/MS(800) catalysts were in the ranges 400–665, 400–690, 400–680, and 400–730 °C, respectively (Figure S3), which corresponded to the decomposition of carbon nanofibers (CNFs).[44] The decomposition temperatures of the CNFs on NiCu/MS(x) catalysts at a higher calcination temperature shifted higher because a stronger CNF structure formed on the NiCu catalyst. The NiCu/MS(600) catalyst had the highest amount of carbon that was 1.02, 1.03, and 1.34 times higher than for NiCu/MS(700), NiCu/MS(800), and NiCu/MS(500) catalysts, respectively.

Table 2

Amounts of Carbon, TOF, Deactivation Rate of Catalysts in the CH4 Decomposition Reaction, Intensity Ratio of D and G Bands, and Metallic Ni Crystallite Size of Used Catalysts

catalyst	amount of carbona (%)	TOFb (1/s)	deactivation ratec	ID/IGd	metallic Ni crystallite sizee (nm)
NiCu/MS(500)	61.59	24.88	0.415	1.18	33.21
NiCu/MS(600)	82.25	65.30	0.947	1.43	22.98
NiCu/MS(700)	80.31	97.81	1.056	2.05	27.16
NiCu/MS(800)	79.86	141.53	0.562	1.52	28.01

Calculated from TGA data based on weights of the used NiCu/MS(x) catalysts.

Calculated as TOF (1/s) at initial reaction time = (% conversion × molar flow rate of CH4 inlet)/(100 × mole of active nickel).

Calculated from the reaction rate of CH4 at the time on stream of 180 min compared to that of 60 min.

Calculated from the Raman data.

Calculated from the Scherrer equation.

The turnover frequency (TOF) at the initial reaction time was calculated to evaluate the catalytic activity based on the amount of active nickel metal, and the results are shown in Table . The NiCu/MS(800) catalyst had the highest TOF value among all the catalysts that was approximately 1.4, 2.2, and 5.7 times higher than those of the NiCu/MS(700), NiCu/MS(600), and NiCu/MS(500) catalysts, respectively. Considering catalyst stability in terms of the deactivation rate calculated from the CH4 reaction rate at 180 min of reaction time compared to that at 60 min (Table ), a value less than one indicates that the catalyst is unstable, whereas a value greater than or equal to one indicates catalyst stability. The deactivation rates of the NiCu/MS(800) and NiCu/MS(500) catalysts were lower than one, indicating that both catalysts were continuously deactivated during the reaction. The results indicated that the NiCu/MS(800) catalyst had outstanding activity initially but its stability was quite poor. In contrast, the NiCu/MS(700) and NiCu/MS(600) catalysts had moderate TOF values, but the activities of both catalysts were quite stable during the reaction—the deactivation value of the NiCu/MS(700) catalyst was more than one, while the deactivation value of NiCu/MS(600) catalyst was almost equal to one. The phenomenon of catalyst deactivation is caused by the carbon product formation covering the active nickel metal and the nickel–metal sintering effect, which are discussed later.

Carbon Allotrope on NiCu/MS(x) Catalysts

The carbon allotrope on the NiCu/MS(x) catalysts used in this study after the CH4 decomposition reaction at 600 °C was characterized using TEM, Raman spectroscopy, and C1s XPS measurement. As shown in the TEM image in Figure , CNFs with different structures formed on the NiCu/MS(x) catalysts synthesized at different calcination temperatures. Fishbone CNFs with average diameters in the range 30–60 nm formed on the NiCu/MS(500) catalyst, whereas mixed fishbone and platelet CNF structures formed on the NiCu/MS(600), NiCu/MS(700), and NiCu/MS(800) catalysts. The average diameters of the fishbone and platelet CNFs of NiCu/MS(600), NiCu/MS(700), and NiCu/MS(800) catalysts were 20–30, 10–30, and 20–40 nm, respectively. The diameter of CNFs depended on the size of Ni and Cu particles because the CNFs formed at the active metal located on the MS support.

Figure 8

TEM image and histograms of CNF diameter distributions of used NiCu/MS(x) catalysts after the CH4 decomposition reaction. (a) NiCu/MS(500), (b) NiCu/MS(600), (c) NiCu/MS(700), and (d) NiCu/MS(800).

Most CNFs formed on the NiCu/MS(500) and NiCu/MS(800) catalysts through the base-growth mechanism, where the active Ni–Cu metals were partially covered by CNFs,[22,36] arising from a strong interaction between Ni–Cu and the MS support (in accordance with the H2-TPR results). Therefore, the activity and stability of both catalysts were deactivated during the reaction. The fishbone and platelet CNFs on the NiCu/MS(600) and NiCu/MS(700) catalysts were mostly formed through the tip-growth mechanism, where the active sites of nickel–copper were on the tips of the CNFs,[22,36] resulting from a weak interaction between Ni–Cu on the MS support (H2-TPR results; Figure ). This phenomenon resulted in higher catalyst activity and prolonged catalyst stability during the reaction. The metal particle shape and the interface between the metal and support directly affected the formation of the CNF structures. With the fishbone CNF formation, with the metal particles having a rhombic morphology or polygonal cluster, the carbon was precipitated from a faceted catalyst particle to generate a structure in which the graphitic platelets were aligned at an angle to the fiber axis.[45,46] For platelet CNF formation, the graphite platelets were stacked in a direction parallel to the base of the metal particle and perpendicular to the fiber axis via the tip-growth mechanism.[47,48]

Considering the MS pore structure after the reaction, Figure S2b,d focused on the support structure. It was found that the MS support was covered by carbon, with the mesocellular structure (foam-like structure) of the NiCu/MS(x) catalyst not changing during the reaction.

Raman spectra of the used NiCu/MS(x) catalysts obtained from the different calcination temperatures are shown in Figure . The positions were observed in the used NiCu/MS(x) catalysts of the D band at 1330 cm–1 and the G band at 1575 cm–1, corresponding to scattering from structural defects or disorders in the carbon and the in-plane vibrations of the carbon atom sp2 hybridization in the graphitic structure, respectively.[49,50] The higher intensity ratio of the D and G bands of the NiCu/MS(700) catalyst compared to the other catalyst clearly depicted the defective nature of the platelet carbon nanofibers. The lowest intensity ratio of the D and G bands was found in the NiCu/MS(500) catalyst, indicating a higher crystalline structure of the fishbone CNFs. The alignment and crystalline perfection of the platelets and fishbones were related to the nature and shape of the metal catalyst particle and the orientation of the precipitating faces.

Figure 9

Raman spectra of used NiCu/MS(x) catalysts after the CH4 decomposition reaction. (a) NiCu/MS(500), (b) NiCu/MS(600), (c) NiCu/MS(700), and (d) NiCu/MS(800).

To clarify the carbon bonding, the used NiCu/MS(x) catalysts after the CH4 decomposition reaction were examined using the XPS measurement, and the binding energy region of the C 1s core level is shown in Figure S4. All the used catalysts had similar C 1s peak deconvoluted into five peaks at 284.2 ± 0.1, 284.9 ± 0.1, 285.6 ± 0.2, 286.4 ± 0.2, and 287.4 ± 0.3 eV, which corresponded to metal carbide, sp2 carbon, sp3 carbon, carbonyl groups (C–O), and carboxyl groups (C=O),[49,51,52] respectively. The two main peaks of sp2 carbon and sp3 carbon were associated with sp2 and sp3 hybridization of the C–C or C–H bonds in the graphite structure of the CNFs formed on the used catalyst. The relative intensity ratio of sp[2] carbon and sp[3] carbon of used catalysts is presented in Table . The lowest intensity ratio of sp2 and sp3 carbons was for the NiCu/MS(600) catalyst, resulting from a higher proportion of sp3 hybridization in the structure.

Table 3

Binding Energy and Fitting Results of C 1s XPS Spectra of Used NiCu/MS(x) Catalysts after CH4 Decomposition Reaction

catalyst	metal carbide (%)	sp2-carbon (%)	sp3-carbon (%)	C–O (%)	C=O (%)	sp2/sp3
NiCu/MS(500)	284.2 (3.2)	284.9 (63.9)	285.6 (15.7)	286.4 (10.0)	287.4 (7.2)	4.07
NiCu/MS(600)	284.3 (3.6)	284.9 (56.2)	285.6 (22.9)	286.5 (10.5)	287.5 (6.8)	2.45
NiCu/MS(700)	284.1 (2.3)	285.1 (60.6)	285.8 (16.5)	286.6 (13.6)	287.7 (7.1)	3.67
NiCu/MS(800)	284.2 (2.4)	284.9 (67.6)	285.6 (15.2)	286.4 (8.8)	287.3 (5.9)	4.45

Characterization of Used NiCu/MS(x) Catalysts

Figure shows the XRD pattern of the used NiCu/MS(x) catalysts after the CH4 decomposition reaction at 600 °C. The diffraction patterns at 2θ of 26.3°, 42.8°, and 54.1° indexed as the (002), (10), and (11) graphite-type reflections[50] were found in all the used NiCu/MS(x) catalysts. Considering the metal catalyst, the diffraction peaks at 2θ of 44.42° and 51.76° were assigned to the Ni–Cu alloy (JCPDS. 47-1406), and the shoulders peak located at 2θ of 44.51° and 51.92° corresponded to the metallic nickel (JCPDS. 4-0850).[14,42,43] However, the diffraction peak of copper was not identified in the used catalysts because the copper metal was well dispersed on the catalyst surface.

Figure 10

XRD patterns of used NiCu/MS(x) catalysts after the CH4 decomposition reaction. (a) NiCu/MS(500), (b) NiCu/MS(600), (c) NiCu/MS(700), and (d) NiCu/MS(800).

The crystallite sizes of the metallic nickel of the used catalysts calculated using the Scherrer equation are presented in Table . The smallest and largest metallic nickel crystallite sizes were in the used NiCu/MS(600) and NiCu/MS(500) catalysts, respectively. In addition, the crystallite sizes of the metallic nickel of the used catalysts were smaller than those of the fresh catalysts. Comparing the fresh and used NiCu/MS(x) catalysts, the decreasing metallic Ni crystallite sizes of NiCu/MS(500), NiCu/MS(600), NiCu/MS(700), and NiCu/MS(800) were 16.55, 35.08, 30.54, and 23.47%, respectively. Among them, the Ni crystallite size of the NiCu/MS(600) catalyst significantly decreased after the reaction. This could be explained by the addition of Cu in the NiCu/MS catalyst promoting fragmentation of the parent Ni–Cu particles when in contact with hydrocarbons.[47] Furthermore, the Cu-modified Ni-supported mesocellular silica catalyst calcined at 600 °C resulted in the interaction between Ni and the MS support becoming weaker due to selective Ni fragmentation when the methane molecule decomposed into carbon species and simultaneously formed CNFs through the tip-growth mechanism on the catalyst surface (as shown in Figure ). Accordingly, the activity of the NiCu/MS(600) catalyst was quite high and stable during the reaction.

Conclusions

Cu-modified Ni-supported mesocellular silica [NiCu/MS(x)] catalysts were synthesized using the co-impregnation method, focusing on the role of the calcination temperature in determining hydrogen production and carbon allotropes during the catalytic CH4 decomposition reaction. The results showed that the nickel and copper reduction behaviors of the NiCu/MS(600) catalyst occurred at lower temperature, indicating a weaker interaction in the NiCu–MS support compared to the other catalysts. With the increase in calcination temperature, the NiCu–MS support interaction of the NiCu/MS(x) catalysts became stronger. Regarding catalyst performance, the maximum CH4 conversion and H2 yield occurred with the NiCu-supported mesocellular silica catalyst, with an optimum calcination temperature of 600 °C. Furthermore, the activity of the NiCu/MS(600) catalyst was stable during the reaction, perhaps due to the high number of active nickel sites and the weak interaction between nickel–copper and the MS support selectively forming fishbone and platelet carbon nanofibers via a tip-growth mechanism. In contrast, the activities of the NiCu/MS(500) and NiCu/MS(800) catalysts significantly decreased with the increase in reaction time because of their strong interaction between NiCu and the MS support, resulting in active NiCu sites being covered by the selective CNF formation via the based-growth mechanism.

Experimental Section

Catalyst Preparation

The bimetallic Ni–Cu loaded on the mesocellular silica (NiCu/MS) catalyst with different calcination temperatures was prepared using two steps: mesocellular silica synthesis and a bimetallic loading step. The mesocellular silica (MS) synthesis was synthesized using the sol–gel process with sodium silicate (29.45% silica; C. Thai Chemical Company) as the silica source, Pluronic P123 as the pore structure agent and 1,3,5-trimethylbenzene (TMB) as the swelling agent. First, Pluronic P123 was dissolved in deionized water under stirring at room temperature until the solution was clear. After that, hydrochloric acid (37 wt.%) and TMB (TMB and Pluronic P123 at a weight ratio of 2:1) were consecutively added into the Pluronic P123 solution while stirring at 40 °C. Then, the sodium silicate was dropped into the solution and stirring continued at 40 °C for 24 h. The mixed solution was hydrothermally treated in an autoclave at 100 °C for 24 h. The obtained product was filtered and washed with deionized water, dried at 100 °C overnight, and calcined in air at 550 °C for 6 h.

For the bimetallic loading process, amounts of 50 wt % nickel (Ni) and 5 wt % copper (Cu) were loaded onto the MS support using the incipient wetness co-impregnation method. Nickel nitrate [Ni(NO3)2·6H2O] and copper nitrate [Cu(NO3)2·3H2O] were dissolved in deionized water. Then, the nickel–copper solution was dropped into the MS support while stirring at room temperature for 1 h. After that, the mixture was dried in a microwave oven at 800 W for 1 min, before being calcined at different temperatures (500, 600, 700, and 800 °C) for 4 h. The catalysts were denoted as NiCu/MS(x), where x is the calcination temperature of the catalyst.

Catalyst Characterization

The textural properties of the NiCu/MS(x) catalysts were examined using N2 physisorption with a 3 Flex Surface Characterization Analyzer. The surface area and pore size distributions were calculated using the Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) methods under relative pressure (P/P0) at −196 °C. Specific pore volume was obtained at a relative pressure of 0.995.

The structure of fresh NiCu/MS(x) and carbon product on the used NiCu/MS(x) catalysts were observed using transmission electron microscopy (TEM: JEOL JEM2010 instrument) with an acceleration voltage of 200 kV.

The reduction temperature and interaction between nickel, copper, and the MS support were identified using the hydrogen temperature-programmed reduction (H2-TPR) technique (Micromeritics, Autochem II Chemisorption analyzer). During the process, an H2/Ar gas mixture (10% H2, Ar balance) was introduced into the catalyst bed at a total flow rate of 50 ml/min at room temperature; the temperature was increased to 1000 °C using a heating rate of 5 °C/min. The amount of H2 consumption was analyzed using a thermal conductivity detector (TCD).

Nickel and copper species were examined by X-ray absorption near edge structure (XANES) measurements using time-resolved X-ray absorption spectroscopy (TR-XAS) at Beamline 2.2 of the Synchrotron Light Research Institute (SLRI), Thailand. The beamline uses an energy-dispersive monochromator equipped with a Si (1 1 1) single crystal.

The chemical states and surface atomic concentrations of the reduced and used NiCu/MS(x) catalysts were analyzed using X-ray photoelectron spectroscopy (XPS: AXIS Ultra DLD). The C 1s peak at 284.8 eV was applied as a reference peak to calibrate the binding energy.

The crystalline structures of reduced NiCu/MS(x) catalysts were examined by using X-ray diffraction spectroscopy (Bruker D8 Advance) operated with monochromated Cu Kα radiation (40 kV and 40 mA) in the 2θ range of 10°–70°.

The active Ni surface area on the catalyst was determined by using the CO chemisorption technique (Micromeritics: Autochem II Chemisorption analyzer). Prior to chemisorption analysis, the catalyst was activated under H2 flow at 700 °C for 1.5 h. The CO chemisorption was operated at 50 °C by pulse-dosing. The amount of CO adsorbed on the catalyst was monitored by using TCD.

The amount and type of carbon product formed on the NiCu/MS(x) catalysts after the methane decomposition reaction were investigated using a thermogravimetric analysis (TGA) technique (METTLER TOLEDO, TGA/DSC 3+) from 30 to 800 °C at a heating rate of 5 °C/min in air flow.

The crystallinity of the carbon formed on the used NiCu/MS(x) catalyst was characterized using Raman spectroscopy (NTEGRA spectra, NT-MDT) with a laser wavelength of 632.8 nm.

Catalytic Activity Test

The effect of calcination temperature on the NiCu/MS(x) catalysts in the methane decomposition reaction was investigated using a packed-bed reactor. In the experiment, 0.1 g of catalyst was packed in an Inconel tube reactor (Inconel-600, O.D. 3/8″) in the center of a tubular furnace equipped with a K-type thermocouple. The catalyst was activated using H2 gas at a flow rate of 60 mL/min at 700 °C for 1.5 h. In the methane decomposition reaction, CH4 and N2 reactant gases (volume ratio of 1:9) with a total flow rate of 100 mL/min [a gas hourly space velocity (GHSV) of 60,000 mL/h·gcatalyst] were fed into the reactor. The reaction temperature was operated at 600 °C under atmospheric pressure. During the reaction, the remaining reactants and products were analyzed using a Shimadzu gas chromatograph (GC14A) equipped with TCD and using a Unibead-C packed column for H2, CO, CH4, and CO2 analysis using Ar as a carrier gas.